[Box Backup-commit] COMMIT r2590 - in box/trunk: . qdbm qdbm/misc
subversion at boxbackup.org
subversion at boxbackup.org
Sat Jan 9 16:15:41 GMT 2010
Author: chris
Date: 2010-01-09 16:15:40 +0000 (Sat, 09 Jan 2010)
New Revision: 2590
Added:
box/trunk/qdbm/
box/trunk/qdbm/COPYING
box/trunk/qdbm/ChangeLog
box/trunk/qdbm/LTmakefile.in
box/trunk/qdbm/Makefile.in
box/trunk/qdbm/NEWS
box/trunk/qdbm/README
box/trunk/qdbm/RISCmakefile
box/trunk/qdbm/THANKS
box/trunk/qdbm/VCmakefile
box/trunk/qdbm/cabin.c
box/trunk/qdbm/cabin.h
box/trunk/qdbm/cbcodec.c
box/trunk/qdbm/cbtest.c
box/trunk/qdbm/configure
box/trunk/qdbm/configure.in
box/trunk/qdbm/crmgr.c
box/trunk/qdbm/crtest.c
box/trunk/qdbm/crtsv.c
box/trunk/qdbm/curia.c
box/trunk/qdbm/curia.h
box/trunk/qdbm/depot.c
box/trunk/qdbm/depot.h
box/trunk/qdbm/dpmgr.c
box/trunk/qdbm/dptest.c
box/trunk/qdbm/dptsv.c
box/trunk/qdbm/hovel.c
box/trunk/qdbm/hovel.h
box/trunk/qdbm/hvmgr.c
box/trunk/qdbm/hvtest.c
box/trunk/qdbm/misc/
box/trunk/qdbm/misc/COPYING.txt
box/trunk/qdbm/misc/README-win32.txt
box/trunk/qdbm/misc/VCmakefile-old
box/trunk/qdbm/misc/benchmark.pdf
box/trunk/qdbm/misc/icon16.png
box/trunk/qdbm/misc/icon20.png
box/trunk/qdbm/misc/index.html
box/trunk/qdbm/misc/index.ja.html
box/trunk/qdbm/misc/logo.png
box/trunk/qdbm/misc/makevcdef
box/trunk/qdbm/misc/mymemo-ja.html
box/trunk/qdbm/misc/tutorial-ja.html
box/trunk/qdbm/misc/win32check.bat
box/trunk/qdbm/myconf.c
box/trunk/qdbm/myconf.h
box/trunk/qdbm/odeum.c
box/trunk/qdbm/odeum.h
box/trunk/qdbm/odidx.c
box/trunk/qdbm/odmgr.c
box/trunk/qdbm/odtest.c
box/trunk/qdbm/qdbm.def
box/trunk/qdbm/qdbm.pc.in
box/trunk/qdbm/qdbm.spec.in
box/trunk/qdbm/qmttest.c
box/trunk/qdbm/relic.c
box/trunk/qdbm/relic.h
box/trunk/qdbm/rlmgr.c
box/trunk/qdbm/rltest.c
box/trunk/qdbm/spex-ja.html
box/trunk/qdbm/spex.html
box/trunk/qdbm/villa.c
box/trunk/qdbm/villa.h
box/trunk/qdbm/vista.c
box/trunk/qdbm/vista.h
box/trunk/qdbm/vlmgr.c
box/trunk/qdbm/vltest.c
box/trunk/qdbm/vltsv.c
Log:
Add QDBM 1.8.77 dbm-compatible database library to sources, will
probably be used as a default for the reference tracking implementation
as it's smaller than BDB and apparently faster.
Added: box/trunk/qdbm/COPYING
===================================================================
--- box/trunk/qdbm/COPYING (rev 0)
+++ box/trunk/qdbm/COPYING 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,504 @@
+ GNU LESSER GENERAL PUBLIC LICENSE
+ Version 2.1, February 1999
+
+ Copyright (C) 1991, 1999 Free Software Foundation, Inc.
+ 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+[This is the first released version of the Lesser GPL. It also counts
+ as the successor of the GNU Library Public License, version 2, hence
+ the version number 2.1.]
+
+ Preamble
+
+ The licenses for most software are designed to take away your
+freedom to share and change it. By contrast, the GNU General Public
+Licenses are intended to guarantee your freedom to share and change
+free software--to make sure the software is free for all its users.
+
+ This license, the Lesser General Public License, applies to some
+specially designated software packages--typically libraries--of the
+Free Software Foundation and other authors who decide to use it. You
+can use it too, but we suggest you first think carefully about whether
+this license or the ordinary General Public License is the better
+strategy to use in any particular case, based on the explanations below.
+
+ When we speak of free software, we are referring to freedom of use,
+not price. Our General Public Licenses are designed to make sure that
+you have the freedom to distribute copies of free software (and charge
+for this service if you wish); that you receive source code or can get
+it if you want it; that you can change the software and use pieces of
+it in new free programs; and that you are informed that you can do
+these things.
+
+ To protect your rights, we need to make restrictions that forbid
+distributors to deny you these rights or to ask you to surrender these
+rights. These restrictions translate to certain responsibilities for
+you if you distribute copies of the library or if you modify it.
+
+ For example, if you distribute copies of the library, whether gratis
+or for a fee, you must give the recipients all the rights that we gave
+you. You must make sure that they, too, receive or can get the source
+code. If you link other code with the library, you must provide
+complete object files to the recipients, so that they can relink them
+with the library after making changes to the library and recompiling
+it. And you must show them these terms so they know their rights.
+
+ We protect your rights with a two-step method: (1) we copyright the
+library, and (2) we offer you this license, which gives you legal
+permission to copy, distribute and/or modify the library.
+
+ To protect each distributor, we want to make it very clear that
+there is no warranty for the free library. Also, if the library is
+modified by someone else and passed on, the recipients should know
+that what they have is not the original version, so that the original
+author's reputation will not be affected by problems that might be
+introduced by others.
+
+ Finally, software patents pose a constant threat to the existence of
+any free program. We wish to make sure that a company cannot
+effectively restrict the users of a free program by obtaining a
+restrictive license from a patent holder. Therefore, we insist that
+any patent license obtained for a version of the library must be
+consistent with the full freedom of use specified in this license.
+
+ Most GNU software, including some libraries, is covered by the
+ordinary GNU General Public License. This license, the GNU Lesser
+General Public License, applies to certain designated libraries, and
+is quite different from the ordinary General Public License. We use
+this license for certain libraries in order to permit linking those
+libraries into non-free programs.
+
+ When a program is linked with a library, whether statically or using
+a shared library, the combination of the two is legally speaking a
+combined work, a derivative of the original library. The ordinary
+General Public License therefore permits such linking only if the
+entire combination fits its criteria of freedom. The Lesser General
+Public License permits more lax criteria for linking other code with
+the library.
+
+ We call this license the "Lesser" General Public License because it
+does Less to protect the user's freedom than the ordinary General
+Public License. It also provides other free software developers Less
+of an advantage over competing non-free programs. These disadvantages
+are the reason we use the ordinary General Public License for many
+libraries. However, the Lesser license provides advantages in certain
+special circumstances.
+
+ For example, on rare occasions, there may be a special need to
+encourage the widest possible use of a certain library, so that it becomes
+a de-facto standard. To achieve this, non-free programs must be
+allowed to use the library. A more frequent case is that a free
+library does the same job as widely used non-free libraries. In this
+case, there is little to gain by limiting the free library to free
+software only, so we use the Lesser General Public License.
+
+ In other cases, permission to use a particular library in non-free
+programs enables a greater number of people to use a large body of
+free software. For example, permission to use the GNU C Library in
+non-free programs enables many more people to use the whole GNU
+operating system, as well as its variant, the GNU/Linux operating
+system.
+
+ Although the Lesser General Public License is Less protective of the
+users' freedom, it does ensure that the user of a program that is
+linked with the Library has the freedom and the wherewithal to run
+that program using a modified version of the Library.
+
+ The precise terms and conditions for copying, distribution and
+modification follow. Pay close attention to the difference between a
+"work based on the library" and a "work that uses the library". The
+former contains code derived from the library, whereas the latter must
+be combined with the library in order to run.
+
+ GNU LESSER GENERAL PUBLIC LICENSE
+ TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+ 0. This License Agreement applies to any software library or other
+program which contains a notice placed by the copyright holder or
+other authorized party saying it may be distributed under the terms of
+this Lesser General Public License (also called "this License").
+Each licensee is addressed as "you".
+
+ A "library" means a collection of software functions and/or data
+prepared so as to be conveniently linked with application programs
+(which use some of those functions and data) to form executables.
+
+ The "Library", below, refers to any such software library or work
+which has been distributed under these terms. A "work based on the
+Library" means either the Library or any derivative work under
+copyright law: that is to say, a work containing the Library or a
+portion of it, either verbatim or with modifications and/or translated
+straightforwardly into another language. (Hereinafter, translation is
+included without limitation in the term "modification".)
+
+ "Source code" for a work means the preferred form of the work for
+making modifications to it. For a library, complete source code means
+all the source code for all modules it contains, plus any associated
+interface definition files, plus the scripts used to control compilation
+and installation of the library.
+
+ Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope. The act of
+running a program using the Library is not restricted, and output from
+such a program is covered only if its contents constitute a work based
+on the Library (independent of the use of the Library in a tool for
+writing it). Whether that is true depends on what the Library does
+and what the program that uses the Library does.
+
+ 1. You may copy and distribute verbatim copies of the Library's
+complete source code as you receive it, in any medium, provided that
+you conspicuously and appropriately publish on each copy an
+appropriate copyright notice and disclaimer of warranty; keep intact
+all the notices that refer to this License and to the absence of any
+warranty; and distribute a copy of this License along with the
+Library.
+
+ You may charge a fee for the physical act of transferring a copy,
+and you may at your option offer warranty protection in exchange for a
+fee.
+
+ 2. You may modify your copy or copies of the Library or any portion
+of it, thus forming a work based on the Library, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+ a) The modified work must itself be a software library.
+
+ b) You must cause the files modified to carry prominent notices
+ stating that you changed the files and the date of any change.
+
+ c) You must cause the whole of the work to be licensed at no
+ charge to all third parties under the terms of this License.
+
+ d) If a facility in the modified Library refers to a function or a
+ table of data to be supplied by an application program that uses
+ the facility, other than as an argument passed when the facility
+ is invoked, then you must make a good faith effort to ensure that,
+ in the event an application does not supply such function or
+ table, the facility still operates, and performs whatever part of
+ its purpose remains meaningful.
+
+ (For example, a function in a library to compute square roots has
+ a purpose that is entirely well-defined independent of the
+ application. Therefore, Subsection 2d requires that any
+ application-supplied function or table used by this function must
+ be optional: if the application does not supply it, the square
+ root function must still compute square roots.)
+
+These requirements apply to the modified work as a whole. If
+identifiable sections of that work are not derived from the Library,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works. But when you
+distribute the same sections as part of a whole which is a work based
+on the Library, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote
+it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Library.
+
+In addition, mere aggregation of another work not based on the Library
+with the Library (or with a work based on the Library) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+ 3. You may opt to apply the terms of the ordinary GNU General Public
+License instead of this License to a given copy of the Library. To do
+this, you must alter all the notices that refer to this License, so
+that they refer to the ordinary GNU General Public License, version 2,
+instead of to this License. (If a newer version than version 2 of the
+ordinary GNU General Public License has appeared, then you can specify
+that version instead if you wish.) Do not make any other change in
+these notices.
+
+ Once this change is made in a given copy, it is irreversible for
+that copy, so the ordinary GNU General Public License applies to all
+subsequent copies and derivative works made from that copy.
+
+ This option is useful when you wish to copy part of the code of
+the Library into a program that is not a library.
+
+ 4. You may copy and distribute the Library (or a portion or
+derivative of it, under Section 2) in object code or executable form
+under the terms of Sections 1 and 2 above provided that you accompany
+it with the complete corresponding machine-readable source code, which
+must be distributed under the terms of Sections 1 and 2 above on a
+medium customarily used for software interchange.
+
+ If distribution of object code is made by offering access to copy
+from a designated place, then offering equivalent access to copy the
+source code from the same place satisfies the requirement to
+distribute the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+ 5. A program that contains no derivative of any portion of the
+Library, but is designed to work with the Library by being compiled or
+linked with it, is called a "work that uses the Library". Such a
+work, in isolation, is not a derivative work of the Library, and
+therefore falls outside the scope of this License.
+
+ However, linking a "work that uses the Library" with the Library
+creates an executable that is a derivative of the Library (because it
+contains portions of the Library), rather than a "work that uses the
+library". The executable is therefore covered by this License.
+Section 6 states terms for distribution of such executables.
+
+ When a "work that uses the Library" uses material from a header file
+that is part of the Library, the object code for the work may be a
+derivative work of the Library even though the source code is not.
+Whether this is true is especially significant if the work can be
+linked without the Library, or if the work is itself a library. The
+threshold for this to be true is not precisely defined by law.
+
+ If such an object file uses only numerical parameters, data
+structure layouts and accessors, and small macros and small inline
+functions (ten lines or less in length), then the use of the object
+file is unrestricted, regardless of whether it is legally a derivative
+work. (Executables containing this object code plus portions of the
+Library will still fall under Section 6.)
+
+ Otherwise, if the work is a derivative of the Library, you may
+distribute the object code for the work under the terms of Section 6.
+Any executables containing that work also fall under Section 6,
+whether or not they are linked directly with the Library itself.
+
+ 6. As an exception to the Sections above, you may also combine or
+link a "work that uses the Library" with the Library to produce a
+work containing portions of the Library, and distribute that work
+under terms of your choice, provided that the terms permit
+modification of the work for the customer's own use and reverse
+engineering for debugging such modifications.
+
+ You must give prominent notice with each copy of the work that the
+Library is used in it and that the Library and its use are covered by
+this License. You must supply a copy of this License. If the work
+during execution displays copyright notices, you must include the
+copyright notice for the Library among them, as well as a reference
+directing the user to the copy of this License. Also, you must do one
+of these things:
+
+ a) Accompany the work with the complete corresponding
+ machine-readable source code for the Library including whatever
+ changes were used in the work (which must be distributed under
+ Sections 1 and 2 above); and, if the work is an executable linked
+ with the Library, with the complete machine-readable "work that
+ uses the Library", as object code and/or source code, so that the
+ user can modify the Library and then relink to produce a modified
+ executable containing the modified Library. (It is understood
+ that the user who changes the contents of definitions files in the
+ Library will not necessarily be able to recompile the application
+ to use the modified definitions.)
+
+ b) Use a suitable shared library mechanism for linking with the
+ Library. A suitable mechanism is one that (1) uses at run time a
+ copy of the library already present on the user's computer system,
+ rather than copying library functions into the executable, and (2)
+ will operate properly with a modified version of the library, if
+ the user installs one, as long as the modified version is
+ interface-compatible with the version that the work was made with.
+
+ c) Accompany the work with a written offer, valid for at
+ least three years, to give the same user the materials
+ specified in Subsection 6a, above, for a charge no more
+ than the cost of performing this distribution.
+
+ d) If distribution of the work is made by offering access to copy
+ from a designated place, offer equivalent access to copy the above
+ specified materials from the same place.
+
+ e) Verify that the user has already received a copy of these
+ materials or that you have already sent this user a copy.
+
+ For an executable, the required form of the "work that uses the
+Library" must include any data and utility programs needed for
+reproducing the executable from it. However, as a special exception,
+the materials to be distributed need not include anything that is
+normally distributed (in either source or binary form) with the major
+components (compiler, kernel, and so on) of the operating system on
+which the executable runs, unless that component itself accompanies
+the executable.
+
+ It may happen that this requirement contradicts the license
+restrictions of other proprietary libraries that do not normally
+accompany the operating system. Such a contradiction means you cannot
+use both them and the Library together in an executable that you
+distribute.
+
+ 7. You may place library facilities that are a work based on the
+Library side-by-side in a single library together with other library
+facilities not covered by this License, and distribute such a combined
+library, provided that the separate distribution of the work based on
+the Library and of the other library facilities is otherwise
+permitted, and provided that you do these two things:
+
+ a) Accompany the combined library with a copy of the same work
+ based on the Library, uncombined with any other library
+ facilities. This must be distributed under the terms of the
+ Sections above.
+
+ b) Give prominent notice with the combined library of the fact
+ that part of it is a work based on the Library, and explaining
+ where to find the accompanying uncombined form of the same work.
+
+ 8. You may not copy, modify, sublicense, link with, or distribute
+the Library except as expressly provided under this License. Any
+attempt otherwise to copy, modify, sublicense, link with, or
+distribute the Library is void, and will automatically terminate your
+rights under this License. However, parties who have received copies,
+or rights, from you under this License will not have their licenses
+terminated so long as such parties remain in full compliance.
+
+ 9. You are not required to accept this License, since you have not
+signed it. However, nothing else grants you permission to modify or
+distribute the Library or its derivative works. These actions are
+prohibited by law if you do not accept this License. Therefore, by
+modifying or distributing the Library (or any work based on the
+Library), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Library or works based on it.
+
+ 10. Each time you redistribute the Library (or any work based on the
+Library), the recipient automatically receives a license from the
+original licensor to copy, distribute, link with or modify the Library
+subject to these terms and conditions. You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties with
+this License.
+
+ 11. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License. If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Library at all. For example, if a patent
+license would not permit royalty-free redistribution of the Library by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Library.
+
+If any portion of this section is held invalid or unenforceable under any
+particular circumstance, the balance of the section is intended to apply,
+and the section as a whole is intended to apply in other circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system which is
+implemented by public license practices. Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+ 12. If the distribution and/or use of the Library is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Library under this License may add
+an explicit geographical distribution limitation excluding those countries,
+so that distribution is permitted only in or among countries not thus
+excluded. In such case, this License incorporates the limitation as if
+written in the body of this License.
+
+ 13. The Free Software Foundation may publish revised and/or new
+versions of the Lesser General Public License from time to time.
+Such new versions will be similar in spirit to the present version,
+but may differ in detail to address new problems or concerns.
+
+Each version is given a distinguishing version number. If the Library
+specifies a version number of this License which applies to it and
+"any later version", you have the option of following the terms and
+conditions either of that version or of any later version published by
+the Free Software Foundation. If the Library does not specify a
+license version number, you may choose any version ever published by
+the Free Software Foundation.
+
+ 14. If you wish to incorporate parts of the Library into other free
+programs whose distribution conditions are incompatible with these,
+write to the author to ask for permission. For software which is
+copyrighted by the Free Software Foundation, write to the Free
+Software Foundation; we sometimes make exceptions for this. Our
+decision will be guided by the two goals of preserving the free status
+of all derivatives of our free software and of promoting the sharing
+and reuse of software generally.
+
+ NO WARRANTY
+
+ 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
+WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
+EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
+OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
+KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
+LIBRARY IS WITH YOU. SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
+THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+ 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
+WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
+AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
+FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
+CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
+LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
+RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
+FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
+SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+DAMAGES.
+
+ END OF TERMS AND CONDITIONS
+
+ How to Apply These Terms to Your New Libraries
+
+ If you develop a new library, and you want it to be of the greatest
+possible use to the public, we recommend making it free software that
+everyone can redistribute and change. You can do so by permitting
+redistribution under these terms (or, alternatively, under the terms of the
+ordinary General Public License).
+
+ To apply these terms, attach the following notices to the library. It is
+safest to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least the
+"copyright" line and a pointer to where the full notice is found.
+
+ <one line to give the library's name and a brief idea of what it does.>
+ Copyright (C) <year> <name of author>
+
+ This library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ This library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with this library; if not, write to the Free Software
+ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+
+Also add information on how to contact you by electronic and paper mail.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a "copyright disclaimer" for the library, if
+necessary. Here is a sample; alter the names:
+
+ Yoyodyne, Inc., hereby disclaims all copyright interest in the
+ library `Frob' (a library for tweaking knobs) written by James Random Hacker.
+
+ <signature of Ty Coon>, 1 April 1990
+ Ty Coon, President of Vice
+
+That's all there is to it!
+
+
Added: box/trunk/qdbm/ChangeLog
===================================================================
--- box/trunk/qdbm/ChangeLog (rev 0)
+++ box/trunk/qdbm/ChangeLog 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,990 @@
+2007-10-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A potential defect related to leaf division of B+ tree was cleared.
+ - Release: 1.8.77
+
+2007-03-07 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The hash function of on-memory map was optimized.
+ - A bug related to large object was fixed.
+ - Release: 1.8.76
+
+2006-11-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - A bug related to B+ tree API for Ruby was fixed.
+ - Release: 1.8.75
+
+2006-11-02 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug related to optimization on Intel Mac series was escaped.
+ - Release: 1.8.74
+
+2006-10-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug related to BZIP2 encoding was fixed.
+ - Release: 1.8.73
+
+2006-10-06 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.8.72
+
+2006-08-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefile was modified to assure portability.
+ - Silent flag was added to the C++ API, the Java API, and the Ruby API.
+ - Release: 1.8.71
+
+2006-08-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of segmentation fault on BSD was fixed.
+ - A test command for checking multi thread safety was added.
+ - Release: 1.8.70
+
+2006-08-15 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Mutex controll in C++ API became refined.
+ - Release: 1.8.69
+
+2006-08-08 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug about memory alignment was fixed.
+ - A bug of handling meta data on big endian platforms was fixed.
+ - Release: 1.8.68
+
+2006-08-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug about memory alignment was fixed.
+ - A bug about parsing MIME was fixed.
+ - Release: 1.8.67
+
+2006-08-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.8.66
+
+2006-08-03 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The extended API was enhanced.
+ - The extended advanced API was enhanced.
+ - Release: 1.8.65
+
+2006-07-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of Makefile ralated to optimization was fixed.
+ - Release: 1.8.64
+
+2006-07-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A lot of functions were replaced by macros.
+ - The utility API was enhanced.
+ - Release: 1.8.63
+
+2006-07-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A lot of functions were replaced by macros.
+ - The URL resolver was to allow unescaped meta characters.
+ - The advanced API was enhanced.
+ - Release: 1.8.62
+
+2006-07-14 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of the size checking function of the extended API was fixed.
+ - The advanced API was enhanced.
+ - Release: 1.8.61
+
+2006-06-03 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The basic API was enhanced.
+ - The extended API was enhanced.
+ - The advanced API was enhanced.
+ - Multiple cursor class is now supported on the java API.
+ - Release: 1.8.60
+
+2006-05-30 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The basic API was enhanced.
+ - The extended API was enhanced.
+ - The advanced API was enhanced.
+ - A bug of Makefile for Mac OS X support was fixed.
+ - Release: 1.8.59
+
+2006-05-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The basic API was enhanced.
+ - The advanced API was enhanced.
+ - Release: 1.8.58
+
+2006-05-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The basic API was enhanced.
+ - The utility API was enhanced.
+ - Release: 1.8.57
+
+2006-05-17 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of URL decoder was fixed.
+ - Release: 1.8.56
+
+2006-05-15 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The extended API was enhanced.
+ - The utility API was enhanced.
+ - Release: 1.8.55
+
+2006-05-15 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The basic API was enhanced.
+ - Release: 1.8.54
+
+2006-05-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - AIX is now supported.
+ - The utility API was enhanced.
+ - Release: 1.8.53
+
+2006-05-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug about evaluating RFC822 date format was fixed.
+ - Release: 1.8.52
+
+2006-05-02 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug about evaluating RFC822 date format was fixed.
+ - Warings from GCC 4.1.x were dealt with.
+ - Missing functions in qdbm.def were supplied.
+ - Release: 1.8.51
+
+2006-04-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.8.50
+
+2006-04-19 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A parameter of dynamic linking library on Mac OS X was modified.
+ - The utility API was enhanced.
+ - Release: 1.8.49
+
+2006-03-27 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.8.48
+
+2006-03-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - LTmakefile was modified.
+ - The utility API was enhanced.
+ - Release: 1.8.47
+
+2006-02-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.8.46
+
+2006-01-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Alignment algorithm was improved.
+ - A bug of mmap emulation on Windows was fixed.
+ - Release: 1.8.45
+
+2006-01-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of handling meta data on big endian platforms was fixed.
+ - The advanced API was enhanced.
+ - Release: 1.8.44
+
+2006-01-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of mmap emulation on Windows was fixed.
+ - Release: 1.8.43
+
+2006-01-22 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - mmap emulation on Windows was enhanced.
+ - Release: 1.8.42
+
+2006-01-13 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Compression of pages of B+ tree with LZO and BZIP was added.
+ - Release: 1.8.41
+
+2006-01-08 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Configuration for VC++ was to build DLL.
+ - Release: 1.8.40
+
+2006-01-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The advanced API was enhanced.
+ - Release: 1.8.39
+
+2006-01-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The advanced API was enhanced.
+ - Release: 1.8.38
+
+2005-12-26 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Functions to set the size of the free block pool were added.
+ - Commands line tools were enhanced.
+ - Release: 1.8.37
+
+2005-12-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Free block pool was enhanced.
+ - Commands line tools were enhanced.
+ - Release: 1.8.36
+
+2005-11-30 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The extended advanced API was enhanced.
+ - All stat calls were replaced with lstat calls.
+ - Release: 1.8.35
+
+2005-11-22 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of i-node duplication on MinGW was fixed.
+ - Release: 1.8.34
+
+2005-09-09 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Compressing options of ZLIB was changed.
+ - The utility API was enhanced.
+ - Release: 1.8.33
+
+2005-08-30 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of backward string matching was fixed.
+ - The utility API was enhanced.
+ - Release: 1.8.32
+
+2005-06-19 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug of the MIME parser was fixed.
+ - A bug about database repair on Win32 was fixed.
+ - A bug of Makefile for Mac OS X support was fixed.
+ - The type of error codes of GDBM-compatibility API was changed.
+ - Release: 1.8.31
+
+2005-06-09 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A Bug of APIs for C++ related to namespace was fixed.
+ - The library of shared objects of C++ APIs was added.
+ - The utility API was enhanced.
+ - Release: 1.8.30
+
+2005-06-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug about file locking on Win32 was fixed.
+ - Manuals of command line tools were added.
+ - The utility API was enhanced.
+ - Release: 1.8.29
+
+2005-05-27 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A compilation problem on hppa-linux was fixed.
+ - A configuration file for pkg-config was added.
+ - Release: 1.8.28
+
+2005-05-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Non-blocking lock was featured.
+ - Release: 1.8.27
+
+2005-05-12 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - The utility API was enhanced.
+ - Release: 1.8.26
+
+2005-05-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - The utility API was enhanced.
+ - Release: 1.8.25
+
+2005-04-25 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - A bug about database repair on Win32 was fixed.
+ - Release: 1.8.24
+
+2005-04-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The extended advanced API was enhanced.
+ - Release: 1.8.23
+
+2005-03-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The extended advanced API was enhanced.
+ - The utility API was enhanced.
+ - The inverted API was enhanced.
+ - Test commands were enhanced.
+ - Release: 1.8.22
+
+2005-01-22 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Functions to dump endian independent data was added.
+ - Release: 1.8.21
+
+2005-01-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Sparse file support was added.
+ - Release: 1.8.20
+
+2005-01-02 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Building configurations were enhanced.
+ - Release: 1.8.19
+
+2004-11-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The extended advanced API was enhanced.
+ - The inverted API was enhanced.
+ - Release: 1.8.18
+
+2004-09-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The CGI script for file upload was enhanced.
+ - The spec file for RPM packages was enhanced.
+ - Release: 1.8.17
+
+2004-08-17 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug in the extended API was fixed.
+ - A bug about seek on Windows 9x was fixed.
+ - The CGI script for file upload was enhanced.
+ - Release: 1.8.16
+
+2004-08-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - A bug about mmap on Mac OS X was fixed.
+ - A CGI script for file upload was added.
+ - Building configurations were enhanced.
+ - Release: 1.8.15
+
+2004-07-19 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The basic API was enhanced.
+ - The extended API was enhanced.
+ - The advanced API was enhanced.
+ - The inverted API was enhanced.
+ - Building configurations were enhanced.
+ - Release: 1.8.14
+
+2004-07-07 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The inverted API was enhanced.
+ - Release: 1.8.13
+
+2004-05-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The inverted API was enhanced.
+ - Building configurations were enhanced.
+ - Release: 1.8.12
+
+2004-05-12 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Source codes were cleaned up for Visual C++.
+ - Test commands were enhanced.
+ - Building configurations were enhanced.
+ - Release: 1.8.11
+
+2004-04-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Utility for performance test was enhanced.
+ - It was escaped from a bug about mmap on FreeBSD and NetBSD.
+ - Release: 1.8.10
+
+2004-04-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Logger for debugging is now featured.
+ - It was escaped from a bug about mmap on OpenBSD.
+ - Release: 1.8.9
+
+2004-04-11 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Processing speed was improved.
+ - The basic API was enhanced.
+ - The extended API was enhanced.
+ - The configuration for VC++ was enhanced.
+ - Release: 1.8.8
+
+2004-03-25 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The configuration of C++ API escaped from a bug of pthread.
+ - Release: 1.8.7
+
+2004-03-21 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The extended API was enhanced.
+ - Bugs in the utility API about memory management were fixed.
+ - Release: 1.8.6
+
+2004-03-12 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug in the utility API for 64-bit platforms was fixed.
+ - The utility API was enhanced.
+ - Release: 1.8.5
+
+2004-03-09 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Runtime switching of whether to use ZLIB was featured.
+ - The binary package for Win32 now features ZLIB and ICONV.
+ - The utility API was enhanced.
+ - Release: 1.8.4
+
+2004-03-06 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - A utility class of Java API was enhanced.
+ - Test commands were enhanced.
+ - Release: 1.8.3
+
+2004-03-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.8.2
+
+2004-02-22 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Processing speed of multi-thread support was improved.
+ - Release: 1.8.1
+
+2004-02-21 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Switch to make API for C thread-safe was added.
+ - Release: 1.8.0
+
+2004-02-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Potential bugs in APIs for C++ were fixed.
+ - Release: 1.7.34
+
+2004-01-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The extended API was enhanced.
+ - The inverted API was enhanced.
+ - Release: 1.7.33
+
+2004-01-17 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - Release: 1.7.32
+
+2004-01-16 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - Release: 1.7.31
+
+2004-01-12 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - Release: 1.7.30
+
+2004-01-11 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.7.29
+
+2004-01-09 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug in the utility API was fixed.
+ - Release: 1.7.28
+
+2004-01-06 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug in the advanced API was fixed.
+ - The utility API was enhanced.
+ - Release: 1.7.27
+
+2004-01-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - Release: 1.7.26
+
+2004-01-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - The utility API was enhanced.
+ - Release: 1.7.25
+
+2003-12-26 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - C++ export was supported in header files.
+ - The utility API was enhanced.
+ - Release: 1.7.24
+
+2003-12-21 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Repairing utility was added to the advanced API.
+ - The basic API was enhanced.
+ - The utility API was enhanced.
+ - Release: 1.7.23
+
+2003-12-16 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Repairing utility was added to the extended API.
+ - The utility API was enhanced.
+ - Release: 1.7.22
+
+2003-12-14 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A utility to repair a broken database file was added.
+ - A ulitity command for full-text search was enhanced.
+ - Release: 1.7.21
+
+2003-12-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Character encoding converter was added.
+ - Release: 1.7.20
+
+2003-12-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Format converter for older versions than 1.7.13 was added.
+ - Tuning paramters of the inverted API were modified.
+ - A CGI script for full-text search was enhanced.
+ - Release: 1.7.19
+
+2003-12-08 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The GDBM-compatible API was enhanced.
+ - Release: 1.7.18
+
+2003-12-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - A CGI script for full-text search was enhanced.
+ - Release: 1.7.17
+
+2003-12-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A CGI script for full-text search was enhanced.
+ - A ulitity command for full-text search was enhanced.
+ - Building configuration files were modified.
+ - Release: 1.7.16
+
+2003-11-30 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A CGI script for full-text search was enhanced.
+ - A ulitity command for full-text search was enhanced.
+ - Release: 1.7.15
+
+2003-11-29 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - A ulitity command for full-text search was enhanced.
+ - A CGI script for full-text search was enhanced.
+ - Release: 1.7.14
+
+2003-11-27 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Format of headers of database was modified.
+ - Supporting endian conversion was abolished.
+ - Import/export features were added to utility commands.
+ - A ulitity command for full-text search was enhanced.
+ - A CGI script for full-text search was enhanced.
+ - Release: 1.7.13
+
+2003-11-25 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - The inverted API was enhanced.
+ - A ulitity command for full-text search was enhanced.
+ - A CGI script for full-text search was enhanced.
+ - Release: 1.7.12
+
+2003-11-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API was enhanced.
+ - A ulitity command for full-text search was enhanced.
+ - A CGI script for full-text search was enhanced.
+ - A utility class of Java API was enhanced.
+ - Release: 1.7.11
+
+2003-11-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A utility class of Java API was enhanced.
+ - Building configuration for CGI scripts was modified.
+ - Release: 1.7.10
+
+2003-11-20 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Tuning paramters of the inverted API were modified.
+ - A ulitity command for full-text search was enhanced.
+ - Release: 1.7.9
+
+2003-11-16 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Optional features of compressing with ZLIB were added.
+ - Release: 1.7.8
+
+2003-11-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The extended advanced API, Vista was added.
+ - Release: 1.7.7
+
+2003-11-03 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - C API compilation using Visual C++ was supported.
+ - Odeum and its commands were enhanced.
+ - Release: 1.7.6
+
+2003-10-25 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A ulitity command for full-text search was enhanced.
+ - A CGI script for full-text search was enhanced.
+ - Release: 1.7.5
+
+2003-10-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A CGI script for full-text search was added.
+ - A bug of BSD support in Makefile was fixed.
+ - Release: 1.7.4
+
+2003-10-19 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Processing speed of the inverted API was improved.
+ - A ulitity command for full-text search was enhanced.
+ - Release: 1.7.3
+
+2003-10-17 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The license was changed to LGPL.
+ - An indexing command for the inverted API was added.
+ - Release: 1.7.2
+
+2003-10-14 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A test command for the inverted API was added.
+ - A bug about optimization in the inverted API was fixed.
+ - Tuning paramters of the inverted API were modified.
+ - Release: 1.7.1
+
+2003-10-13 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The inverted API, Odeum was added.
+ - Comments in source codes increased.
+ - Release: 1.7.0
+
+2003-10-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Executable commands were modified for RISC OS.
+ - Release: 1.6.22
+
+2003-10-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Executable commands were modified for RISC OS.
+ - Release: 1.6.21
+
+2003-09-30 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Naming file in RISC OS style was supported.
+ - Hash functions were enhanced.
+ - Release: 1.6.20
+
+2003-09-26 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Test commands were added.
+ - Release: 1.6.19
+
+2003-09-22 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefile for RISC OS was added.
+ - C++ API was enhanced.
+ - Release: 1.6.18
+
+2003-09-14 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Test commands were enhanced.
+ - Source codes and Makefile were modified for RISC OS.
+ - The CGI script was enhanced.
+ - Release: 1.6.17
+
+2003-09-13 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The CGI script for administration was added.
+ - C++ API was enhanced.
+ - Release: 1.6.16
+
+2003-08-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.6.15
+
+2003-08-16 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.6.14
+
+2003-08-11 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.6.13
+
+2003-08-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - A bug in the utility API was fixed.
+ - Test commands were enhanced.
+ - Release: 1.6.12
+
+2003-07-31 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - C and Java APIs support compilation using MinGW.
+ - Release: 1.6.11
+
+2003-07-27 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Perl and Ruby APIs support transaction.
+ - Release: 1.6.10
+
+2003-07-26 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Speed of transaction was improved.
+ - Release: 1.6.9
+
+2003-07-24 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - C++ and Java APIs of B+ tree support transaction.
+ - Release: 1.6.8
+
+2003-07-21 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - C API of B+ tree supports transaction.
+ - Bugs about Mutex in C++ and Java APIs were fixed.
+ - Release: 1.6.7
+
+2003-07-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - C++ and Java APIs of database abstraction were simplified.
+ - Ruby API of B+ tree was simplified.
+ - Release: 1.6.6
+
+2003-07-16 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Ruby API of B+ tree was added.
+ - C, C++, Java and Perl APIs of B+ tree were enhanced.
+ - Release: 1.6.5
+
+2003-07-13 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Perl API of B+ tree was added.
+ - C, C++ and Java APIs of B+ tree were enhanced.
+ - Release: 1.6.4
+
+2003-07-08 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Java API of B++ tree was added.
+ - C API of B+ tree was enhanced.
+ - C++ API of B+ tree was enhanced.
+ - Release: 1.6.3
+
+2003-07-05 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Debug routines were removed and speed was improved.
+ - No locking mode was added.
+ - Endian converter of Villa was added.
+ - C++ API of B++ tree was added.
+ - Release: 1.6.2
+
+2003-07-03 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The advanced API was enhanced to improve its speed.
+ - Makefiles were enhanced for HP-UX.
+ - Release: 1.6.1
+
+2003-06-28 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The advanced API for B+ tree database is added.
+ - Release: 1.6.0
+
+2003-06-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefiles were enhanced for Mac OS X.
+ - Release: 1.5.13
+
+2003-06-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefile was enhanced for Solaris and Mac OS X.
+ - Release: 1.5.12
+
+2003-06-06 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.5.11
+
+2003-05-31 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was enhanced.
+ - Release: 1.5.10
+
+2003-05-29 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was changed.
+ - Release: 1.5.9
+
+2003-05-25 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Alignment mechanism was simplified.
+ - Release: 1.5.8
+
+2003-05-17 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefile was updated
+ - Release: 1.5.7
+
+2003-05-16 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Test tools were enhanced.
+ - Release: 1.5.6
+
+2003-05-14 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefile using libtool was added.
+ - Release: 1.5.5
+
+2003-05-12 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - API for Ruby was enhanced.
+ - Release: 1.5.4
+
+2003-05-08 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - API for Ruby was added.
+ - Release: 1.5.3
+
+2003-05-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - API for Perl was enhanced.
+ - Release: 1.5.2
+
+2003-04-30 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - API for Perl was enhanced.
+ - Release: 1.5.1
+
+2003-04-29 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - API for Perl was added.
+ - Release: 1.5.0
+
+2003-04-25 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The package was reorganized to be a GNU package.
+ - Release: 1.4.11
+
+2003-04-21 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The utility API was added.
+ - Release: 1.4.10
+
+2003-04-15 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Endian converters were added.
+ - Release: 1.4.9
+
+2003-04-12 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The GDBM-compatible API was enhanced.
+ - Release: 1.4.8
+
+2003-04-10 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Some bugs were fixed and C++ API was enhanced.
+ - Release: 1.4.7
+
+2003-04-09 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Coalescence of dead regions and reuse of them were featured.
+ - Release: 1.4.6
+
+2003-04-06 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The GDBM-compatible API and the C++ API were enhanced.
+ - Release: 1.4.5
+
+2003-04-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The API for C++ and the API for Java were enhanced.
+ - Release: 1.4.4
+
+2003-04-02 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Documents for C++ was enhanced.
+ - Release: 1.4.3
+
+2003-04-01 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The API for Java was enhanced.
+ - Release: 1.4.2
+
+2003-03-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Makefiles were modified.
+ - Release: 1.4.1
+
+2003-03-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - APIs for C++ and Java was enhanced.
+ - Release: 1.4.0
+
+2003-03-15 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The API for C++ was added.
+ - Release: 1.3.2
+
+2003-03-11 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The API for Java was supported on Solaris and Windows.
+ - Release: 1.3.1
+
+2003-03-04 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The API for Java was added.
+ - Release: 1.3.0
+
+2003-02-23 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The GDBM-compatible API was added.
+ - Release: 1.2.0
+
+2003-02-21 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Compatibility with NDBM was improved.
+ - Release: 1.1.3
+
+2003-02-18 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Optimizing suppoted on Windows. DLL is supported.
+ - Release: 1.1.2
+
+2003-02-15 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - Windows is now supported.
+ - The compatible API was enhanced.
+ - Release: 1.1.1
+
+2003-02-11 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The NDBM-compatible API was added.
+ - Release: 1.1.0
+
+2003-02-08 Mikio Hirabayashi <mikio at users.sourceforge.net>
+
+ - The initial version.
+ - Release: 1.0.0
+
Added: box/trunk/qdbm/LTmakefile.in
===================================================================
--- box/trunk/qdbm/LTmakefile.in (rev 0)
+++ box/trunk/qdbm/LTmakefile.in 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,318 @@
+# Makefile to build QDBM using libtool
+
+
+
+#================================================================
+# Setting variables
+#================================================================
+
+
+# Generic settings
+SHELL = @SHELL@
+
+# Packaging
+PACKAGE = @PACKAGE_NAME@
+VERSION = @PACKAGE_VERSION@
+LIBVER = @LIBVER@
+LIBREV = @LIBREV@
+
+# Targets
+MYHEADS = depot.h curia.h relic.h hovel.h cabin.h villa.h vista.h odeum.h
+MYLIBOBJS = depot.lo curia.lo relic.lo hovel.lo cabin.lo villa.lo vista.lo odeum.lo myconf.lo
+MYLIBS = libqdbm.la
+MYBINS = dpmgr dptest dptsv crmgr crtest crtsv rlmgr rltest hvmgr hvtest \
+ cbtest cbcodec vlmgr vltest vltsv odmgr odtest odidx qmttest
+MYMAN1S = dpmgr.1 dptest.1 dptsv.1 crmgr.1 crtest.1 crtsv.1 rlmgr.1 rltest.1 hvmgr.1 hvtest.1 \
+ cbtest.1 cbcodec.1 vlmgr.1 vltest.1 vltsv.1 odmgr.1 odtest.1 odidx.1 qmttest.1
+MYMAN3S = qdbm.3 depot.3 dpopen.3 curia.3 cropen.3 relic.3 hovel.3 \
+ cabin.3 villa.3 vlopen.3 vista.3 odeum.3 odopen.3
+MYDOCS = spex.html spex-ja.html COPYING ChangeLog NEWS THANKS
+MYPCS = qdbm.pc
+
+# Install destinations
+prefix = @prefix@
+exec_prefix = @exec_prefix@
+MYHEADDIR = @includedir@
+MYLIBDIR = @libdir@
+MYBINDIR = @bindir@
+MYMAN1DIR = @mandir@/man1
+MYMAN3DIR = @mandir@/man3
+MYSHAREDIR = $(prefix)/share/$(PACKAGE)
+MYPCDIR = @libdir@/pkgconfig
+
+# Building binaries
+LIBTOOL = libtool
+CC = gcc
+CFLAGS = -I. -I$(MYHEADDIR) -I$(HOME)/include -I/usr/local/include @MYDEFS@ \
+ -D_XOPEN_SOURCE_EXTENDED=1 -D_GNU_SOURCE=1 -D__EXTENSIONS__=1 -D_HPUX_SOURCE=1 \
+ -D_POSIX_MAPPED_FILES=1 -D_POSIX_SYNCHRONIZED_IO=1 \
+ -DPIC=1 -D_THREAD_SAFE=1 -D_REENTRANT=1 -DNDEBUG -O3
+LD = gcc
+LIBLDFLAGS = -rpath $(MYLIBDIR) -R $(MYLIBDIR) \
+ -version-info $$(($(LIBVER)+$(LIBREV))):0:$(LIBREV) @LIBS@
+LDFLAGS = -rpath $(MYLIBDIR) -R $(MYLIBDIR) -L. -lqdbm @LIBS@
+INSTALL = install
+MKDIR = mkdir -p
+CP = cp -rf
+RM = rm -rf
+
+
+
+#================================================================
+# Suffix rules
+#================================================================
+
+
+.SUFFIXES :
+.SUFFIXES : .c .lo
+
+.c.lo :
+ $(LIBTOOL) --mode=compile --tag=CC $(CC) -c $(CFLAGS) $<
+
+
+
+#================================================================
+# Actions
+#================================================================
+
+
+all : $(MYLIBS) $(MYBINS)
+ @printf '\n'
+ @printf '#================================================================\n'
+ @printf '# Ready to install.\n'
+ @printf '#================================================================\n'
+
+
+clean :
+ $(RM) $(MYLIBS) $(MYBINS) *.o *.a *.so *.lo *.la a.out .libs \
+ *.exe *.dll.a *.dll TAGS srcdoc gmon.out leak.log casket casket.* *~
+
+
+install :
+ $(LIBTOOL) --mode=install $(INSTALL) $(MYHEADS) $(MYHEADDIR)
+ $(LIBTOOL) --mode=install $(INSTALL) $(MYLIBS) $(MYLIBDIR)
+ $(LIBTOOL) --mode=install $(INSTALL) $(MYBINS) $(MYBINDIR)
+ $(MKDIR) $(MYMAN1DIR)
+ cd man && $(CP) $(MYMAN1S) $(MYMAN1DIR)
+ $(MKDIR) $(MYMAN3DIR)
+ cd man && $(CP) $(MYMAN3S) $(MYMAN3DIR)
+ $(MKDIR) $(MYSHAREDIR)
+ $(CP) $(MYDOCS) $(MYSHAREDIR)
+ $(MKDIR) $(MYPCDIR)
+ $(CP) $(MYPCS) $(MYPCDIR)
+ @printf '\n'
+ @printf '#================================================================\n'
+ @printf '# Thanks for using QDBM.\n'
+ @printf '#================================================================\n'
+
+
+uninstall :
+ cd $(MYHEADDIR) && $(LIBTOOL) --mode=uninstall $(RM) $(MYHEADS)
+ cd $(MYLIBDIR) && $(LIBTOOL) --mode=uninstall $(RM) $(MYLIBS)
+ cd $(MYBINDIR) && $(LIBTOOL) --mode=uninstall $(RM) $(MYBINS)
+ cd $(MYMAN1DIR) && $(RM) $(MYMAN1S)
+ cd $(MYMAN3DIR) && $(RM) $(MYMAN3S)
+ $(RM) $(MYSHAREDIR)
+ cd $(MYPCDIR) && $(RM) $(MYPCS)
+
+
+distclean : clean
+ $(RM) Makefile LTmakefile rpmspec config.cache config.log config.status autom4te.cache
+
+
+check :
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./dptest write -s casket 500 500000
+ $(LIBTOOL) --mode=execute ./dptest write casket 50000 5000
+ $(LIBTOOL) --mode=execute ./dptest read casket
+ $(LIBTOOL) --mode=execute ./dptest read -wb casket
+ $(LIBTOOL) --mode=execute ./dptest rcat -c casket 50000 50 500 32 8
+ $(LIBTOOL) --mode=execute ./dptest combo casket
+ $(LIBTOOL) --mode=execute ./dptest wicked -c casket 5000
+ $(LIBTOOL) --mode=execute ./dptest wicked casket 500
+ $(LIBTOOL) --mode=execute ./dpmgr repair casket
+ $(LIBTOOL) --mode=execute ./dpmgr optimize casket
+ $(LIBTOOL) --mode=execute ./dpmgr list casket
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./crtest write -s casket 500 100000 5
+ $(LIBTOOL) --mode=execute ./crtest write casket 50000 500 10
+ $(LIBTOOL) --mode=execute ./crtest read casket
+ $(LIBTOOL) --mode=execute ./crtest read -wb casket
+ $(LIBTOOL) --mode=execute ./crtest rcat -c casket 50000 5 10 500 32 8
+ $(LIBTOOL) --mode=execute ./crtest combo casket
+ $(LIBTOOL) --mode=execute ./crtest wicked -c casket 5000
+ $(LIBTOOL) --mode=execute ./crtest wicked casket 500
+ $(LIBTOOL) --mode=execute ./crmgr repair casket
+ $(LIBTOOL) --mode=execute ./crmgr optimize casket
+ $(LIBTOOL) --mode=execute ./crmgr list casket
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./crtest write -lob casket 1000 50 10
+ $(LIBTOOL) --mode=execute ./crtest read -lob casket
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./rltest write casket 5000
+ $(LIBTOOL) --mode=execute ./rltest read casket 5000
+ $(LIBTOOL) --mode=execute ./rlmgr list casket
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./hvtest write casket 5000
+ $(LIBTOOL) --mode=execute ./hvtest read casket 5000
+ $(LIBTOOL) --mode=execute ./hvmgr optimize casket
+ $(LIBTOOL) --mode=execute ./hvmgr list casket
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./hvtest write -qdbm -s casket 500
+ $(LIBTOOL) --mode=execute ./hvtest write -qdbm casket 5000
+ $(LIBTOOL) --mode=execute ./hvtest read -qdbm casket 5000
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./cbtest sort 5000
+ $(LIBTOOL) --mode=execute ./cbtest strstr 500
+ $(LIBTOOL) --mode=execute ./cbtest list 50000
+ $(LIBTOOL) --mode=execute ./cbtest list -d 500
+ $(LIBTOOL) --mode=execute ./cbtest map 50000 500
+ $(LIBTOOL) --mode=execute ./cbtest map -d 500 5
+ $(LIBTOOL) --mode=execute ./cbtest heap 50000 500
+ $(LIBTOOL) --mode=execute ./cbtest heap -d 500 50
+ $(LIBTOOL) --mode=execute ./cbtest wicked 5000
+ $(LIBTOOL) --mode=execute ./cbtest misc
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./vltest write -tune 32 31 32 32 casket 50000
+ $(LIBTOOL) --mode=execute ./vltest read casket
+ $(LIBTOOL) --mode=execute ./vltest rdup -tune 32 31 512 256 casket 50000 50000
+ $(LIBTOOL) --mode=execute ./vltest combo casket
+ $(LIBTOOL) --mode=execute ./vltest wicked -c casket 5000
+ $(LIBTOOL) --mode=execute ./vltest wicked casket 500
+ $(LIBTOOL) --mode=execute ./vlmgr repair casket
+ $(LIBTOOL) --mode=execute ./vlmgr optimize casket
+ $(LIBTOOL) --mode=execute ./vlmgr list casket
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./vltest write -int -cz -tune 32 31 32 32 casket 50000
+ $(LIBTOOL) --mode=execute ./vltest read -int -vc casket
+ $(LIBTOOL) --mode=execute ./vltest rdup -int -cz -cc -tune 32 31 512 256 casket 50000 50000
+ $(LIBTOOL) --mode=execute ./vltest combo -cz casket
+ $(LIBTOOL) --mode=execute ./vltest wicked -cz -c casket 5000
+ $(LIBTOOL) --mode=execute ./vltest combo -cy casket
+ $(LIBTOOL) --mode=execute ./vltest wicked -cy -c casket 5000
+ $(LIBTOOL) --mode=execute ./vltest combo -cx casket
+ $(LIBTOOL) --mode=execute ./vltest wicked -cx -c casket 5000
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./odtest write casket 500 50 5000
+ $(LIBTOOL) --mode=execute ./odtest read casket
+ $(LIBTOOL) --mode=execute ./odtest combo casket
+ $(LIBTOOL) --mode=execute ./odtest wicked casket 500
+ $(RM) casket*
+ $(LIBTOOL) --mode=execute ./qmttest casket 50000 10
+ $(RM) casket*
+ @printf '\n'
+ @printf '#================================================================\n'
+ @printf '# Checking completed.\n'
+ @printf '#================================================================\n'
+
+
+.PHONY : all clean install check
+
+
+
+#================================================================
+# Building binaries
+#================================================================
+
+
+libqdbm.la : $(MYLIBOBJS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) $(MYLIBOBJS) -o $@ $(LIBLDFLAGS)
+
+
+dpmgr : dpmgr.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ dpmgr.lo $(LDFLAGS)
+
+
+dptest : dptest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ dptest.lo $(LDFLAGS)
+
+
+dptsv : dptsv.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ dptsv.lo $(LDFLAGS)
+
+
+crmgr : crmgr.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ crmgr.lo $(LDFLAGS)
+
+
+crtest : crtest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ crtest.lo $(LDFLAGS)
+
+
+crtsv : crtsv.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ crtsv.lo $(LDFLAGS)
+
+
+rlmgr : rlmgr.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ rlmgr.lo $(LDFLAGS)
+
+
+rltest : rltest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ rltest.lo $(LDFLAGS)
+
+
+hvmgr : hvmgr.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ hvmgr.lo $(LDFLAGS)
+
+
+hvtest : hvtest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ hvtest.lo $(LDFLAGS)
+
+
+cbtest : cbtest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ cbtest.lo $(LDFLAGS)
+
+
+cbcodec : cbcodec.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ cbcodec.lo $(LDFLAGS)
+
+
+vlmgr : vlmgr.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ vlmgr.lo $(LDFLAGS)
+
+
+vltest : vltest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ vltest.lo $(LDFLAGS)
+
+
+vltsv : vltsv.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ vltsv.lo $(LDFLAGS)
+
+
+odmgr : odmgr.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ odmgr.lo $(LDFLAGS)
+
+
+odtest : odtest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ odtest.lo $(LDFLAGS)
+
+
+odidx : odidx.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ odidx.lo $(LDFLAGS)
+
+
+qmttest : qmttest.lo $(MYLIBS)
+ $(LIBTOOL) --mode=link --tag=CC $(LD) -o $@ qmttest.lo $(LDFLAGS)
+
+
+depot.lo dpmgr.lo dptest.lo dptsv.lo : depot.h myconf.h
+
+curia.lo crmgr.lo crtest.lo crtsv.lo : curia.h depot.h myconf.h
+
+relic.lo rlmgr.lo rltest.lo : relic.h depot.h myconf.h
+
+hovel.lo hvmgr.lo hvtest.lo : hovel.h depot.h curia.h myconf.h
+
+cabin.lo cbtest.lo : cabin.h myconf.h
+
+villa.lo vlmgr.lo vltest.lo vltsv.lo : villa.h depot.h cabin.h myconf.h
+
+vista.lo : vista.h villa.h depot.h curia.h cabin.h myconf.h villa.c
+
+odeum.lo odmgr.lo odtest.lo odidx.lo : odeum.h depot.h curia.h cabin.h villa.h myconf.h
+
+myconf.lo : myconf.h
+
+
+
+# END OF FILE
Added: box/trunk/qdbm/Makefile.in
===================================================================
--- box/trunk/qdbm/Makefile.in (rev 0)
+++ box/trunk/qdbm/Makefile.in 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,646 @@
+# Makefile for QDBM
+
+
+
+#================================================================
+# Setting variables
+#================================================================
+
+
+# Generic settings
+SHELL = @SHELL@
+srcdir = @srcdir@
+VPATH = @srcdir@
+SUBDIRS =
+
+# Packaging
+PACKAGE = @PACKAGE_NAME@
+VERSION = @PACKAGE_VERSION@
+PACKAGEDIR = $(PACKAGE)-$(VERSION)
+PACKAGETGZ = $(PACKAGE)-$(VERSION).tar.gz
+LIBVER = @LIBVER@
+LIBREV = @LIBREV@
+
+# Targets
+MYHEADS = depot.h curia.h relic.h hovel.h cabin.h villa.h vista.h odeum.h
+MYLIBOBJS = depot.o curia.o relic.o hovel.o cabin.o villa.o vista.o odeum.o myconf.o
+MYLIBS = libqdbm.a libqdbm.so.$(LIBVER).$(LIBREV).0 libqdbm.so.$(LIBVER) libqdbm.so
+MYBINS = dpmgr dptest dptsv crmgr crtest crtsv rlmgr rltest hvmgr hvtest \
+ cbtest cbcodec vlmgr vltest vltsv odmgr odtest odidx qmttest
+MYMAN1S = dpmgr.1 dptest.1 dptsv.1 crmgr.1 crtest.1 crtsv.1 rlmgr.1 rltest.1 hvmgr.1 hvtest.1 \
+ cbtest.1 cbcodec.1 vlmgr.1 vltest.1 vltsv.1 odmgr.1 odtest.1 odidx.1 qmttest.1
+MYMAN3S = qdbm.3 depot.3 dpopen.3 curia.3 cropen.3 relic.3 hovel.3 \
+ cabin.3 villa.3 vlopen.3 vista.3 odeum.3 odopen.3
+MYDOCS = spex.html spex-ja.html COPYING ChangeLog NEWS THANKS
+MYPCS = qdbm.pc
+MYWINLIBS = libqdbm.a libqdbm.dll.a
+MYMACLIBS = libqdbm.a libqdbm.$(LIBVER).$(LIBREV).0.dylib libqdbm.$(LIBVER).dylib libqdbm.dylib
+MYHPUXLIBS = libqdbm.a libqdbm.sl
+
+# Install destinations
+prefix = @prefix@
+exec_prefix = @exec_prefix@
+MYHEADDIR = @includedir@
+MYLIBDIR = @libdir@
+MYBINDIR = @bindir@
+MYMAN1DIR = @mandir@/man1
+MYMAN3DIR = @mandir@/man3
+MYDATADIR = @datadir@/$(PACKAGE)
+MYPCDIR = @libdir@/pkgconfig
+DESTDIR =
+
+# Building binaries
+CC = @CC@
+CPPFLAGS = @CPPFLAGS@ -I$(srcdir) -I$(MYHEADDIR) \
+ -I$(HOME)/include -I/usr/local/include @MYDEFS@ \
+ -D_XOPEN_SOURCE_EXTENDED=1 -D_GNU_SOURCE=1 -D__EXTENSIONS__=1 -D_HPUX_SOURCE=1 \
+ -D_POSIX_MAPPED_FILES=1 -D_POSIX_SYNCHRONIZED_IO=1 \
+ -DPIC=1 -D_THREAD_SAFE=1 -D_REENTRANT=1 -DNDEBUG
+CFLAGS = -Wall -pedantic -fPIC -fsigned-char -O3 -fomit-frame-pointer -fforce-addr @MYOPTS@
+LD = @LD@
+LIBS = -lqdbm @LIBS@
+LIBLDFLAGS = @LDFLAGS@ -L. -L$(MYLIBDIR) -L$(HOME)/lib -L/usr/local/lib @LIBS@
+LDFLAGS = @LDFLAGS@ -L. -L$(MYLIBDIR) -L$(HOME)/lib -L/usr/local/lib $(LIBS)
+LDENV = LD_RUN_PATH=/lib:/usr/lib:$(MYLIBDIR):$(HOME)/lib:/usr/local/lib
+AR = @AR@
+ARFLAGS = rcsv
+RUNENV = LD_LIBRARY_PATH=.:/lib:/usr/lib:$(MYLIBDIR):$(HOME)/lib:/usr/local/lib
+
+
+
+#================================================================
+# Suffix rules
+#================================================================
+
+
+.SUFFIXES :
+.SUFFIXES : .c .o
+
+.c.o :
+ $(CC) -c $(CPPFLAGS) $(CFLAGS) $<
+
+
+
+#================================================================
+# Actions
+#================================================================
+
+
+targets : @TARGETS@
+
+
+all : $(MYLIBS) $(MYBINS)
+ @printf '\n'
+ @printf '#================================================================\n'
+ @printf '# Ready to install.\n'
+ @printf '#================================================================\n'
+
+
+static :
+ make MYLIBS="$(MYLIBS)" LDFLAGS="-static $(LDFLAGS)" all
+
+
+debug :
+ make MYLIBS="$(MYLIBS)" CPPFLAGS="$(CPPFLAGS) -UNDEBUG" \
+ CFLAGS="-Wall -ansi -pedantic -fPIC -fsigned-char -O2 -g" \
+ LDFLAGS="-static $(LDFLAGS)" all
+
+
+devel :
+ make MYLIBS="$(MYLIBS)" CPPFLAGS="$(CPPFLAGS) -UNDEBUG" \
+ CFLAGS="-Wall -ansi -pedantic -fPIC -fsigned-char -O2 -g -pipe" all
+ sync ; sync
+
+
+stable :
+ make MYLIBS="$(MYLIBS)" CFLAGS="-Wall -ansi -pedantic -fPIC -fsigned-char -O2" all
+
+
+profile :
+ make MYLIBS="$(MYLIBS)" \
+ CFLAGS="-Wall -pedantic -fPIC -fsigned-char -O3 -pg -g -Werror" \
+ LDFLAGS="-static $(LDFLAGS)" all
+
+
+unsigned :
+ make MYLIBS="$(MYLIBS)" CPPFLAGS="$(CPPFLAGS) -UNDEBUG" \
+ CFLAGS="-Wall -ansi -pedantic -fPIC -funsigned-char -g -O2" all
+
+
+m64 :
+ make MYLIBS="$(MYLIBS)" CPPFLAGS="$(CPPFLAGS) -UNDEBUG" \
+ CFLAGS="-Wall -ansi -pedantic -fPIC -fsigned-char -O2 -m64 -g" all
+
+
+pen4 :
+ stdopt="-O3 -fomit-frame-pointer -minline-all-stringops" ; \
+ exopt="-march=pentium4 -minline-all-stringops -fprefetch-loop-arrays" ; \
+ make MYLIBS="$(MYLIBS)" \
+ CFLAGS="-Wall -pedantic -fPIC -fsigned-char $$stdopt $$exopt" all
+
+
+k8 :
+ stdopt="-O3 -fomit-frame-pointer -minline-all-stringops" ; \
+ exopt="-march=k8 -minline-all-stringops -fprefetch-loop-arrays" ; \
+ make MYLIBS="$(MYLIBS)" \
+ CFLAGS="-Wall -pedantic -fPIC -fsigned-char $$stdopt $$exopt" all
+
+
+clean :
+ rm -rf $(MYLIBS) $(MYBINS) *.o *.a *.so *.lo *.la a.out .libs \
+ *.exe *.dll.a *.dll *.dylib *.sl TAGS srcdoc gmon.out leak.log \
+ casket casket.* casket-* *~
+
+
+version :
+ vernum=`expr $(LIBVER)00 + $(LIBREV)` ; \
+ sed -e 's/_QDBM_VERSION.*/_QDBM_VERSION "$(VERSION)"/' \
+ -e "s/_QDBM_LIBVER.*/_QDBM_LIBVER $$vernum/" depot.h > depot.h~
+ [ -f depot.h~ ] && mv -f depot.h~ depot.h
+
+
+install :
+ mkdir -p $(DESTDIR)$(MYHEADDIR)
+ cd $(srcdir) && cp -Rf $(MYHEADS) $(DESTDIR)$(MYHEADDIR)
+ mkdir -p $(DESTDIR)$(MYLIBDIR)
+ cp -Rf $(MYLIBS) $(DESTDIR)$(MYLIBDIR)
+ mkdir -p $(DESTDIR)$(MYBINDIR)
+ cp -Rf $(MYBINS) $(DESTDIR)$(MYBINDIR)
+ mkdir -p $(DESTDIR)$(MYMAN1DIR)
+ cd $(srcdir)/man && cp -Rf $(MYMAN1S) $(DESTDIR)$(MYMAN1DIR)
+ mkdir -p $(DESTDIR)$(MYMAN3DIR)
+ cd $(srcdir)/man && cp -Rf $(MYMAN3S) $(DESTDIR)$(MYMAN3DIR)
+ mkdir -p $(DESTDIR)$(MYDATADIR)
+ cd $(srcdir) && cp -Rf $(MYDOCS) $(DESTDIR)$(MYDATADIR)
+ mkdir -p $(DESTDIR)$(MYPCDIR)
+ cd $(srcdir) && cp -Rf $(MYPCS) $(DESTDIR)$(MYPCDIR)
+ @printf '\n'
+ @printf '#================================================================\n'
+ @printf '# Thanks for using QDBM.\n'
+ @printf '#================================================================\n'
+
+
+install-strip :
+ make DESTDIR=$(DESTDIR) install
+ cd $(DESTDIR)$(MYBINDIR) && strip $(MYBINS)
+
+
+uninstall :
+ cd $(DESTDIR)$(MYHEADDIR) && rm -f $(MYHEADS)
+ cd $(DESTDIR)$(MYLIBDIR) && rm -f $(MYLIBS)
+ cd $(DESTDIR)$(MYBINDIR) && rm -f $(MYBINS)
+ cd $(DESTDIR)$(MYMAN1DIR) && rm -f $(MYMAN1S)
+ cd $(DESTDIR)$(MYMAN3DIR) && rm -f $(MYMAN3S)
+ rm -rf $(DESTDIR)$(MYDATADIR)
+ cd $(DESTDIR)$(MYPCDIR) && rm -f $(MYPCS)
+
+
+dist :
+ sync ; sync
+ for dir in $(SUBDIRS) ; \
+ do \
+ if [ -d $$dir ] ; \
+ then \
+ echo Making $@ in $$dir ; \
+ ( cd $$dir && if [ -f configure.in ] ; then autoconf ; ./configure ; \
+ make dist ; fi ) ; \
+ fi ; \
+ done
+ make version
+ make distclean
+ cd .. && tar cvf - $(PACKAGEDIR) | gzip -c > $(PACKAGETGZ)
+ sync ; sync
+
+
+distclean : clean
+ for dir in $(SUBDIRS) ; \
+ do \
+ if [ -d $$dir ] ; \
+ then \
+ echo Making $@ in $$dir ; \
+ ( cd $$dir && if [ -f Makefile ] ; then make distclean ; fi ) ; \
+ fi ; \
+ done
+ rm -rf Makefile LTmakefile qdbm.spec qdbm.pc config.cache config.log config.status \
+ autom4te.cache rpm-tmp *-win32
+
+
+TAGS :
+ etags -o $@ *.c *.h
+
+
+sdoc :
+ rm -rf srcdoc
+ ./lab/ccdoc -d srcdoc -t "Source Documents of QDBM" *.h *.c
+
+
+check :
+ sync ; sync
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./dptest write -s casket 500 500000
+ $(RUNENV) $(RUNCMD) ./dptest write casket 50000 5000
+ $(RUNENV) $(RUNCMD) ./dptest read casket
+ $(RUNENV) $(RUNCMD) ./dptest read -wb casket
+ $(RUNENV) $(RUNCMD) ./dptest rcat -c casket 50000 50 500 32 8
+ $(RUNENV) $(RUNCMD) ./dptest combo casket
+ $(RUNENV) $(RUNCMD) ./dptest wicked -c casket 5000
+ $(RUNENV) $(RUNCMD) ./dptest wicked casket 500
+ $(RUNENV) $(RUNCMD) ./dpmgr repair casket
+ $(RUNENV) $(RUNCMD) ./dpmgr optimize casket
+ $(RUNENV) $(RUNCMD) ./dpmgr list casket
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./crtest write -s casket 500 100000 5
+ $(RUNENV) $(RUNCMD) ./crtest write casket 50000 500 10
+ $(RUNENV) $(RUNCMD) ./crtest read casket
+ $(RUNENV) $(RUNCMD) ./crtest read -wb casket
+ $(RUNENV) $(RUNCMD) ./crtest rcat -c casket 50000 5 10 500 32 8
+ $(RUNENV) $(RUNCMD) ./crtest combo casket
+ $(RUNENV) $(RUNCMD) ./crtest wicked -c casket 5000
+ $(RUNENV) $(RUNCMD) ./crtest wicked casket 500
+ $(RUNENV) $(RUNCMD) ./crmgr repair casket
+ $(RUNENV) $(RUNCMD) ./crmgr optimize casket
+ $(RUNENV) $(RUNCMD) ./crmgr list casket
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./crtest write -lob casket 1000 50 10
+ $(RUNENV) $(RUNCMD) ./crtest read -lob casket
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./rltest write casket 5000
+ $(RUNENV) $(RUNCMD) ./rltest read casket 5000
+ $(RUNENV) $(RUNCMD) ./rlmgr list casket
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./hvtest write casket 5000
+ $(RUNENV) $(RUNCMD) ./hvtest read casket 5000
+ $(RUNENV) $(RUNCMD) ./hvmgr optimize casket
+ $(RUNENV) $(RUNCMD) ./hvmgr list casket
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./hvtest write -qdbm -s casket 500
+ $(RUNENV) $(RUNCMD) ./hvtest write -qdbm casket 5000
+ $(RUNENV) $(RUNCMD) ./hvtest read -qdbm casket 5000
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./cbtest sort 5000
+ $(RUNENV) $(RUNCMD) ./cbtest strstr 500
+ $(RUNENV) $(RUNCMD) ./cbtest list 50000
+ $(RUNENV) $(RUNCMD) ./cbtest list -d 500
+ $(RUNENV) $(RUNCMD) ./cbtest map 50000 500
+ $(RUNENV) $(RUNCMD) ./cbtest map -d 500 5
+ $(RUNENV) $(RUNCMD) ./cbtest heap 50000 500
+ $(RUNENV) $(RUNCMD) ./cbtest heap -d 500 50
+ $(RUNENV) $(RUNCMD) ./cbtest wicked 5000
+ $(RUNENV) $(RUNCMD) ./cbtest misc
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./vltest write -tune 32 31 32 32 casket 50000
+ $(RUNENV) $(RUNCMD) ./vltest read casket
+ $(RUNENV) $(RUNCMD) ./vltest rdup -tune 32 31 512 256 casket 50000 50000
+ $(RUNENV) $(RUNCMD) ./vltest combo casket
+ $(RUNENV) $(RUNCMD) ./vltest wicked -c casket 5000
+ $(RUNENV) $(RUNCMD) ./vltest wicked casket 500
+ $(RUNENV) $(RUNCMD) ./vlmgr repair casket
+ $(RUNENV) $(RUNCMD) ./vlmgr optimize casket
+ $(RUNENV) $(RUNCMD) ./vlmgr list casket
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./vltest write -int -cz -tune 32 31 32 32 casket 50000
+ $(RUNENV) $(RUNCMD) ./vltest read -int -vc casket
+ $(RUNENV) $(RUNCMD) ./vltest rdup -int -cz -cc -tune 32 31 512 256 casket 50000 50000
+ $(RUNENV) $(RUNCMD) ./vltest combo -cz casket
+ $(RUNENV) $(RUNCMD) ./vltest wicked -cz -c casket 5000
+ $(RUNENV) $(RUNCMD) ./vltest combo -cy casket
+ $(RUNENV) $(RUNCMD) ./vltest wicked -cy -c casket 5000
+ $(RUNENV) $(RUNCMD) ./vltest combo -cx casket
+ $(RUNENV) $(RUNCMD) ./vltest wicked -cx -c casket 5000
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./odtest write casket 500 50 5000
+ $(RUNENV) $(RUNCMD) ./odtest read casket
+ $(RUNENV) $(RUNCMD) ./odtest combo casket
+ $(RUNENV) $(RUNCMD) ./odtest wicked casket 500
+ rm -rf casket*
+ $(RUNENV) $(RUNCMD) ./qmttest casket 50000 10
+ rm -rf casket*
+ @printf '\n'
+ @printf '#================================================================\n'
+ @printf '# Checking completed.\n'
+ @printf '#================================================================\n'
+
+
+check-valgrind :
+ make RUNCMD="valgrind --tool=memcheck --log-fd=1" check | tee leak.log
+ grep ERROR leak.log
+ grep 'at exit' leak.log
+
+
+world :
+ make clean ; make
+ cd plus ; [ -f Makefile ] || ./configure ; make clean ; make
+ cd java ; [ -f Makefile ] || ./configure ; make clean ; make
+ cd perl ; [ -f Makefile ] || ./configure ; make clean ; make
+ cd ruby ; [ -f Makefile ] || ./configure ; make clean ; make
+ cd cgi ; [ -f Makefile ] || ./configure ; make clean ; make
+
+
+install-world :
+ make install
+ cd plus ; [ -f Makefile ] || ./configure ; make install
+ cd java ; [ -f Makefile ] || ./configure ; make install
+ cd perl ; [ -f Makefile ] || ./configure ; make install
+ cd ruby ; [ -f Makefile ] || ./configure ; make install
+ cd cgi ; [ -f Makefile ] || ./configure ; make install
+
+
+uninstall-world :
+ make uninstall
+ cd plus ; [ -f Makefile ] || ./configure ; make uninstall
+ cd java ; [ -f Makefile ] || ./configure ; make uninstall
+ cd perl ; [ -f Makefile ] || ./configure ; make uninstall
+ cd ruby ; [ -f Makefile ] || ./configure ; make uninstall
+ cd cgi ; [ -f Makefile ] || ./configure ; make uninstall
+
+
+check-world :
+ make check
+ cd plus ; [ -f Makefile ] || ./configure ; make check
+ cd java ; [ -f Makefile ] || ./configure ; make check
+ cd perl ; [ -f Makefile ] || ./configure ; make check
+ cd ruby ; [ -f Makefile ] || ./configure ; make check
+
+
+rpm : ../$(PACKAGETGZ) qdbm.spec
+ mkdir -p rpm-tmp/{BUILD,RPMS,SOURCES,SPECS,SRPMS}
+ mkdir -p rpm-tmp/RPMS/i386
+ cp ../$(PACKAGETGZ) rpm-tmp/SOURCES
+ rpmbuild -bb --target i386 --define "_topdir `pwd`/rpm-tmp" qdbm.spec
+ mv -f rpm-tmp/RPMS/i386/$(PACKAGE)-*$(VERSION)*.rpm ..
+ rm -rf rpm-tmp
+
+
+win32pkg :
+ test -f /bin/mgwz.dll
+ test -f /bin/libiconv-2.dll
+ make uninstall && make uninstall-win && make clean
+ make mingw && strip *.exe && make install-win
+ cd java && ./configure
+ cd java && make uninstall && make uninstall-win && make clean
+ cd java && make mingw
+ cd cgi && ./configure
+ cd cgi && make clean
+ cd cgi && make mingw && strip *.cgi
+ mkdir -p $(PACKAGE)-$(VERSION)-win32
+ cp -Rf $(MYHEADS) libqdbm.dll.a qdbm.dll *.exe *.html \
+ misc/README-win32.txt misc/COPYING.txt misc/win32check.bat \
+ /bin/mgwz.dll /bin/libiconv-2.dll \
+ $(PACKAGE)-$(VERSION)-win32
+ cp -Rf java/jqdbm.dll java/qdbm.jar java/*.html java/japidoc $(PACKAGE)-$(VERSION)-win32
+ mkdir -p $(PACKAGE)-$(VERSION)-win32/cgi
+ cp -Rf cgi/*.cgi cgi/*.conf cgi/*.html $(PACKAGE)-$(VERSION)-win32/cgi
+ zip -r $(PACKAGE)-$(VERSION)-win32.zip $(PACKAGE)-$(VERSION)-win32
+ mv -f $(PACKAGE)-$(VERSION)-win32.zip ..
+ rm -rf $(PACKAGE)-$(VERSION)-win32
+ make uninstall && make uninstall-win && make clean
+ cd java ; make uninstall && make uninstall-win && make clean
+
+
+win :
+ make MYLIBS="$(MYWINLIBS)" CFLAGS="-Wall -ansi -pedantic -fsigned-char -O2"
+
+
+mingw :
+ make CC="gcc -mno-cygwin" MYLIBS="$(MYWINLIBS)" \
+ CFLAGS="-Wall -fsigned-char -O2" LIBLDFLAGS="@MGWLIBS@" LDFLAGS="-L. -lqdbm @MGWLIBS@"
+
+
+check-win :
+ make check
+
+
+install-win :
+ make MYBINS="`for file in $(MYBINS) ; do echo $$file.exe ; done | tr '\n' ' '`" \
+ MYLIBS="$(MYWINLIBS)" install
+ cp -Rf qdbm.dll $(DESTDIR)$(MYBINDIR)
+
+
+uninstall-win :
+ make MYBINS="`for file in $(MYBINS) ; do echo $$file.exe ; done | tr '\n' ' '`" \
+ MYLIBS="$(MYWINLIBS)" uninstall
+ rm -f $(DESTDIR)$(MYBINDIR)/qdbm.dll
+
+
+def : libqdbm.a
+ ./misc/makevcdef libqdbm.a > qdbm.def
+
+
+mac :
+ make MYLIBS="$(MYMACLIBS)" CFLAGS="-Wall -fsigned-char -fno-common -O2"
+
+
+check-mac :
+ make RUNENV="DYLD_LIBRARY_PATH=." check
+
+
+install-mac :
+ make MYLIBS="$(MYMACLIBS)" install
+
+
+uninstall-mac :
+ make MYLIBS="$(MYMACLIBS)" uninstall
+
+
+hpux :
+ make MYLIBS="$(MYHPUXLIBS)"
+
+
+check-hpux :
+ make RUNENV="SHLIB_PATH=." check
+
+
+install-hpux :
+ make MYLIBS="$(MYHPUXLIBS)" install
+
+
+uninstall-hpux :
+ make MYLIBS="$(MYHPUXLIBS)" uninstall
+
+
+no-so :
+ make MYLIBS="libqdbm.a" all
+
+
+install-no-so :
+ make MYLIBS="libqdbm.a" install
+
+
+uninstall-no-so :
+ make MYLIBS="libqdbm.a" uninstall
+
+
+.PHONY : all clean install check
+
+
+
+#================================================================
+# Building binaries
+#================================================================
+
+
+libqdbm.a : $(MYLIBOBJS)
+ $(AR) $(ARFLAGS) $@ $(MYLIBOBJS)
+
+
+libqdbm.so.$(LIBVER).$(LIBREV).0 : $(MYLIBOBJS)
+ if uname -a | egrep -i 'SunOS' > /dev/null ; \
+ then \
+ $(CC) -shared -Wl,-G,-h,libqdbm.so.$(LIBVER) -o $@ $(MYLIBOBJS) $(LIBLDFLAGS) ; \
+ else \
+ $(CC) -shared -Wl,-soname,libqdbm.so.$(LIBVER) -o $@ $(MYLIBOBJS) $(LIBLDFLAGS) ; \
+ fi
+
+
+libqdbm.so.$(LIBVER) : libqdbm.so.$(LIBVER).$(LIBREV).0
+ ln -f -s libqdbm.so.$(LIBVER).$(LIBREV).0 $@
+
+
+libqdbm.so : libqdbm.so.$(LIBVER).$(LIBREV).0
+ ln -f -s libqdbm.so.$(LIBVER).$(LIBREV).0 $@
+
+
+libqdbm.dll.a : qdbm.dll
+
+
+qdbm.dll : $(MYLIBOBJS)
+ $(CC) -shared -o $@ \
+ -Wl,--out-implib=lib$@.a \
+ -Wl,--export-all-symbols \
+ -Wl,--enable-auto-import \
+ -Wl,--add-stdcall-alias \
+ -Wl,--whole-archive \
+ -Wl,--no-whole-archive $(MYLIBOBJS) $(LIBLDFLAGS)
+
+
+libqdbm.$(LIBVER).$(LIBREV).0.dylib : $(MYLIBOBJS)
+ $(CC) -dynamiclib -o $@ \
+ -install_name $(MYLIBDIR)/libqdbm.$(LIBVER).dylib \
+ -current_version $(LIBVER).$(LIBREV).0 \
+ -compatibility_version $(LIBVER) \
+ $(MYLIBOBJS) $(LIBLDFLAGS)
+
+
+libqdbm.$(LIBVER).dylib : libqdbm.$(LIBVER).$(LIBREV).0.dylib
+ ln -f -s libqdbm.$(LIBVER).$(LIBREV).0.dylib $@
+
+
+libqdbm.dylib : libqdbm.$(LIBVER).$(LIBREV).0.dylib
+ ln -f -s libqdbm.$(LIBVER).$(LIBREV).0.dylib $@
+
+
+libqdbm.sl : $(MYLIBOBJS)
+ $(CC) -shared -Wl,-b -o $@ $(MYLIBOBJS) $(LIBLDFLAGS)
+
+
+dpmgr : dpmgr.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ dpmgr.o $(LDFLAGS)
+
+
+dptest : dptest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ dptest.o $(LDFLAGS)
+
+
+dptsv : dptsv.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ dptsv.o $(LDFLAGS)
+
+
+crmgr : crmgr.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ crmgr.o $(LDFLAGS)
+
+
+crtest : crtest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ crtest.o $(LDFLAGS)
+
+
+crtsv : crtsv.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ crtsv.o $(LDFLAGS)
+
+
+rlmgr : rlmgr.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ rlmgr.o $(LDFLAGS)
+
+
+rltest : rltest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ rltest.o $(LDFLAGS)
+
+
+hvmgr : hvmgr.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ hvmgr.o $(LDFLAGS)
+
+
+hvtest : hvtest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ hvtest.o $(LDFLAGS)
+
+
+cbtest : cbtest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ cbtest.o $(LDFLAGS)
+
+
+cbcodec : cbcodec.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ cbcodec.o $(LDFLAGS)
+
+
+vlmgr : vlmgr.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ vlmgr.o $(LDFLAGS)
+
+
+vltest : vltest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ vltest.o $(LDFLAGS)
+
+
+vltsv : vltsv.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ vltsv.o $(LDFLAGS)
+
+
+odmgr : odmgr.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ odmgr.o $(LDFLAGS)
+
+
+odtest : odtest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ odtest.o $(LDFLAGS)
+
+
+odidx : odidx.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ odidx.o $(LDFLAGS)
+
+
+qmttest : qmttest.o $(MYLIBS)
+ $(LDENV) $(CC) $(CFLAGS) -o $@ qmttest.o $(LDFLAGS)
+
+
+depot.o : depot.h myconf.h
+
+curia.o : depot.h curia.h myconf.h
+
+relic.o : depot.h relic.h myconf.h
+
+hovel.o : depot.h curia.h hovel.h myconf.h
+
+cabin.o : cabin.h myconf.h
+
+villa.o : depot.h cabin.h villa.h myconf.h
+
+vista.o : depot.h curia.h cabin.h villa.h vista.h myconf.h villa.c
+
+odeum.o : depot.h curia.h cabin.h villa.h myconf.h
+
+myconf.o : myconf.h
+
+dpmgr.o dptest.o dptsv.o : depot.h cabin.h
+
+crmgr.o crtest.o crtsv.o : depot.h curia.h cabin.h
+
+rlmgr.o rltest.o : depot.h relic.h cabin.h
+
+hvmgr.o hvtest.o : depot.h curia.h hovel.h cabin.h
+
+cbtest.o cbcodec.o : cabin.h
+
+vlmgr.o vltest.o vltsv.o : depot.h cabin.h villa.h
+
+odmgr.o odtest.o odidx.o : depot.h curia.h cabin.h villa.h odeum.h
+
+
+
+# END OF FILE
Added: box/trunk/qdbm/NEWS
===================================================================
--- box/trunk/qdbm/NEWS (rev 0)
+++ box/trunk/qdbm/NEWS 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,43 @@
+== Thu, 08 Sep 2005 13:13:58 +0900 ==
+
+Compressing options of ZLIB was changed. If you use villa with the option
+`VL_OZCOMP', databases of the earlier versions are not compatible with the
+current version of QDBM. To convert the old database to new format,
+export endian independent data by "vlmgr exportdb" with the old version,
+and then import it by "vlmgr importdb" with the latest version.
+
+
+
+== Wed, 10 Mar 2004 23:24:24 +0900 ==
+
+API of B+ tree was changed a bit. Even if you build QDBM with ZLIB
+enabled, records are not compressed. Instead of it, the function `vlopen'
+has the option `VL_OZCOMP'. If it is specified, records are compressed.
+So, you can switch whether to compress records or not, on runtime.
+
+Users who have used ZLIB feature should modify their source codes to
+specify that option.
+
+
+
+== Wed, 10 Dec 2003 09:24:12 +0900 ==
+
+The database format was changed with QDBM 1.7.13. Newer versions do not
+have backward compatibility to old format. You can convert old databases
+with the command `fmtcnv031127' in the sub directory `lab'. To build it,
+perform the following steps.
+
+ cd lab
+ make fmtcnv031127
+
+About usage of it, see the file `README' in `lab'. Typically, to convert
+a Depot database named as `old' and create a database named as `new',
+perform the following step.
+
+ ./fmtcnv031127 < old > new
+
+I'm sorry for bothering you.
+
+
+
+== END OF FILE ==
Added: box/trunk/qdbm/README
===================================================================
--- box/trunk/qdbm/README (rev 0)
+++ box/trunk/qdbm/README 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,50 @@
+================================================================
+ QDBM: Quick Database Manager
+ Copyright (C) 2000-2007 Mikio Hirabayashi
+================================================================
+
+
+Please read the following documents with a WWW browser.
+How to install QDBM is explained in the specification.
+
+ README - this file
+ COPYING - license
+ ChangeLog - history of enhancement
+ NEWS - news for users
+ THANKS - list of contributors
+ spex.html - specification
+ spex-ja.html - specification in Japanese
+
+
+Contents of the directory tree is below.
+
+ ./ - sources of QDBM
+ ./plus/ - API for C++ (read `./plus/xspex.html')
+ ./java/ - API for Java (read `./java/jspex.html')
+ ./perl/ - API for Perl (read `./perl/plspex.html')
+ ./ruby/ - API for Ruby (read `./ruby/rbspex.html')
+ ./cgi/ - CGI scripts (read `./cgi/cgispex.html')
+ ./man1/ - manual pages for commands
+ ./man3/ - manual pages for C API
+ ./lab/ - for test and experiment
+ ./bros/ - for comparison with other database managers
+ ./misc/ - miscellaneous files
+
+
+QDBM is released under the terms of the GNU Lesser General Public
+License. See the file `COPYING' for details.
+
+QDBM was written by Mikio Hirabayashi. You can contact the author
+by e-mail to `mikio at users.sourceforge.net'. However, as for
+topics which can be shared among other users, pleae send it to
+the mailing list. To join the mailing list, refer to the following
+URL.
+
+ http://lists.sourceforge.net/lists/listinfo/qdbm-users
+
+
+Thanks.
+
+
+
+== END OF FILE ==
Added: box/trunk/qdbm/RISCmakefile
===================================================================
--- box/trunk/qdbm/RISCmakefile (rev 0)
+++ box/trunk/qdbm/RISCmakefile 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,140 @@
+# Makefile for the RISC OS version of QDBM
+
+
+# Define which compiler to use:
+
+CC = cc
+#CC = gcc
+
+
+#########################################
+# DO NOT EDIT ANYTHING BELOW THIS LINE! #
+#########################################
+
+ifeq (${CC},cc)
+CC = cc
+LD = link
+AR = libfile
+DEPEND = -depend !Depend
+CC_FLAGS = -Wdp -throwback -Otime -I@,Unix: -JUnix
+UNIXLIB = Unix:o.UnixLib
+else
+ifeq (${CC},gcc)
+CC = gcc
+LD = gcc
+AR = ar
+CC_FLAGS = -mthrowback -O3 -I.
+else
+# No other compiler supported!
+endif
+endif
+
+QDBM_OBJS = o.depot o.curia o.relic o.hovel o.cabin o.villa o.vista o.odeum o.myconf
+
+.INIT :
+ @cdir o
+
+## Rule Patterns ##
+
+.SUFFIXES : .c .o
+
+.c.o :
+ $(CC) $(CC_FLAGS) $(DEPEND) -c -o $@ $<
+
+# Static dependencies:
+
+all : libqdbm testcases managers converters
+
+libqdbm : $(QDBM_OBJS)
+ $(AR) $(AR_FLAGS) -c -o libqdbm $(QDBM_OBJS)
+
+testcases : dptest crtest rltest hvtest cbtest vltest odtest
+ create testcases
+
+managers : dpmgr crmgr rlmgr hvmgr vlmgr odmgr
+ create managers
+
+converters : dptsv crtsv cbcodec vltsv odidx
+ create converters
+
+dptest : o.dptest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+crtest : o.crtest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+rltest : o.rltest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+hvtest : o.hvtest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+cbtest : o.cbtest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+vltest : o.vltest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+odtest : o.odtest libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+dpmgr : o.dpmgr libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+crmgr : o.crmgr libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+rlmgr : o.rlmgr libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+hvmgr : o.hvmgr libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+vlmgr : o.vlmgr libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+odmgr : o.odmgr libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+dptsv : o.dptsv libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+crtsv : o.crtsv libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+cbcodec : o.cbcodec libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+vltsv : o.vltsv libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+odidx : o.odidx libqdbm
+ $(LD) $(LD_FLAGS) -o $@ o.$* libqdbm $(UNIXLIB)
+
+clean:
+ -ifthere libqdbm then wipe libqdbm ~CFR~V
+ -ifthere dptest then wipe dptest ~CFR~V
+ -ifthere crtest then wipe crtest ~CFR~V
+ -ifthere rltest then wipe rltest ~CFR~V
+ -ifthere hvtest then wipe hvtest ~CFR~V
+ -ifthere cbtest then wipe cbtest ~CFR~V
+ -ifthere vltest then wipe vltest ~CFR~V
+ -ifthere odtest then wipe odtest ~CFR~V
+ -ifthere dpmgr then wipe dpmgr ~CFR~V
+ -ifthere crmgr then wipe crmgr ~CFR~V
+ -ifthere rlmgr then wipe rlmgr ~CFR~V
+ -ifthere hvmgr then wipe hvmgr ~CFR~V
+ -ifthere cbmgr then wipe cbmgr ~CFR~V
+ -ifthere vlmgr then wipe vlmgr ~CFR~V
+ -ifthere odmgr then wipe odmgr ~CFR~V
+ -ifthere dptsv then wipe dptsv ~CFR~V
+ -ifthere crtsv then wipe crtsv ~CFR~V
+ -ifthere cbcodec then wipe cbcodec ~CFR~V
+ -ifthere vltsv then wipe vltsv ~CFR~V
+ -ifthere odidx then wipe odidx ~CFR~V
+ -ifthere testcases then wipe testcases ~CFR~V
+ -ifthere managers then wipe managers ~CFR~V
+ -ifthere converters then wipe converters ~CFR~V
+ -ifthere o.* then wipe o.* ~CFR~V
+
+# Dynamic dependencies:
Added: box/trunk/qdbm/THANKS
===================================================================
--- box/trunk/qdbm/THANKS (rev 0)
+++ box/trunk/qdbm/THANKS 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,45 @@
+================================================================
+ Thanks to all of the following for their valuable suggestions
+ or contributions.
+================================================================
+
+
+Kang-Jin Lee
+ - suggestions about the GDBM-compatible API
+ - contributions about Makefile
+
+Pat Podenski
+ - suggestions about porting to Mac OS X, Solaris, and HP-UX
+
+BERO
+ - contributions about supporting MinGW
+
+Stefan Bellon
+ - contributions about porting to RISC OS
+
+Donald Gobin
+ - contributions about supporting Visual C++
+
+Emanuel Dejanu
+ - contributions about supporting Visual C++
+
+Keith Bostic
+ - suggestions about the performance test suite
+
+William Lachance
+ - contributions about RPM spec file
+
+Zed A. Shaw
+ - contributions about a text analyzer in Odeum
+ - contributions about a query language in Odeum
+
+Chris Bilderback
+ - contributions about cursor functions in Villa.
+
+Fumitoshi Ukai
+ - contributions of troff manuals
+ - making Debian packages
+
+
+
+== END OF FILE ==
Added: box/trunk/qdbm/VCmakefile
===================================================================
--- box/trunk/qdbm/VCmakefile (rev 0)
+++ box/trunk/qdbm/VCmakefile 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,248 @@
+# Makefile to build QDBM using Microsoft Visual C++
+
+
+
+#================================================================
+# Setting variables
+#================================================================
+
+
+# VC++ directory
+VCPATH = C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7
+
+# User options
+YOUR_CLFLAGS =
+YOUR_LIBFLAGS =
+YOUR_LINKFLAGS=
+
+# Configurations
+!IF "$(CFG)" == "ld"
+!MESSAGE Build using static debug configuration
+BASE_FLAGS = /MLd /W3 /ZI /Od /FD /GZ /Fo"$(OUTDIR)\\" /Fd"$(OUTDIR)\\"
+BASE_DEFS = /D_DEBUG /D__DEBUG__
+OUTDIR = .\tmp_ld
+LIB_APPEND = _ld
+EXE_APPEND = _ld
+!ELSEIF "$(CFG)" == "l"
+!MESSAGE Build using static release configuration
+BASE_DEFS = /DNDEBUG
+BASE_FLAGS = /ML /W3 /O2 /FD /Fo"$(OUTDIR)\\" /Fd"$(OUTDIR)\\"
+OUTDIR = .\tmp_l
+LIB_APPEND = _l
+EXE_APPEND = _l
+!ELSEIF "$(CFG)" == "td"
+!MESSAGE Build using static threaded debug configuration
+BASE_FLAGS = /MTd /W3 /ZI /Od /FD /GZ /Fo"$(OUTDIR)\\" /Fd"$(OUTDIR)\\"
+BASE_DEFS = /D_DEBUG /D__DEBUG__
+OUTDIR = .\tmp_td
+LIB_APPEND = _td
+EXE_APPEND = _td
+!ELSEIF "$(CFG)" == "t"
+!MESSAGE Build using static threaded release configuration
+BASE_DEFS = /DNDEBUG
+BASE_FLAGS = /MT /W3 /O2 /FD /Fo"$(OUTDIR)\\" /Fd"$(OUTDIR)\\"
+OUTDIR = .\tmp_t
+LIB_APPEND = _t
+EXE_APPEND = _t
+!ELSEIF "$(CFG)" == "dd"
+!MESSAGE Build using dynamic threaded debug configuration
+BASE_FLAGS = /MDd /W3 /ZI /Od /FD /GZ /Fo"$(OUTDIR)\\" /Fd"$(OUTDIR)\\"
+BASE_DEFS = /D_DEBUG /D__DEBUG__
+OUTDIR = .\tmp_dd
+LIB_APPEND = _dd
+EXE_APPEND = _dd
+!ELSE
+!MESSAGE Build using dynamic threaded release configuration
+BASE_DEFS = /DNDEBUG
+BASE_FLAGS = /MD /W3 /O2 /FD /Fo"$(OUTDIR)\\" /Fd"$(OUTDIR)\\"
+OUTDIR = .\tmp
+LIB_APPEND =
+EXE_APPEND =
+!ENDIF
+
+# Building binaries
+CLFLAGS = /I "$(VCPATH)\Include" /I "$(VCPATH)\PlatformSDK\Include" /I "." \
+ /nologo $(YOUR_CLFLAGS) $(BASE_FLAGS) $(BASE_DEFS) /D_CRT_SECURE_NO_DEPRECATE=1
+LIBFLAGS = /libpath:"$(VCPATH)\lib" /libpath:"$(VCPATH)\PlatformSDK\Lib" /libpath:"." \
+ /nologo $(YOUR_LIBFLAGS)
+LINKFLAGS = /libpath:"$(VCPATH)\lib" /libpath:"$(VCPATH)\PlatformSDK\Lib" /libpath:"." \
+ /nologo $(YOUR_LINKFLAGS)
+
+# Targets
+MYLIBS = qdbm$(LIB_APPEND).dll qdbm$(LIB_APPEND).lib qdbm$(LIB_APPEND)_static.lib
+LIBOBJS = $(OUTDIR)\depot.obj $(OUTDIR)\curia.obj $(OUTDIR)\relic.obj \
+ $(OUTDIR)\hovel.obj $(OUTDIR)\cabin.obj $(OUTDIR)\villa.obj \
+ $(OUTDIR)\vista.obj $(OUTDIR)\odeum.obj $(OUTDIR)\myconf.obj
+MYBINS = dpmgr$(EXE_APPEND).exe dptest$(EXE_APPEND).exe dptsv$(EXE_APPEND).exe \
+ crmgr$(EXE_APPEND).exe crtest$(EXE_APPEND).exe crtsv$(EXE_APPEND).exe \
+ rlmgr$(EXE_APPEND).exe rltest$(EXE_APPEND).exe hvmgr$(EXE_APPEND).exe \
+ hvtest$(EXE_APPEND).exe cbtest$(EXE_APPEND).exe cbcodec$(EXE_APPEND).exe \
+ vlmgr$(EXE_APPEND).exe vltest$(EXE_APPEND).exe vltsv$(EXE_APPEND).exe \
+ odmgr$(EXE_APPEND).exe odtest$(EXE_APPEND).exe odidx$(EXE_APPEND).exe
+
+
+
+#================================================================
+# Suffix rules
+#================================================================
+
+
+.SUFFIXES :
+.SUFFIXES : .c .obj
+
+.c{$(OUTDIR)}.obj :
+ cl /c $(CLFLAGS) $<
+
+.c.obj:
+ cl /c $(CLFLAGS) $<
+
+
+
+#================================================================
+# Actions
+#================================================================
+
+
+all : $(OUTDIR) $(MYLIBS) $(MYBINS)
+
+
+allcfg:
+ nmake /NOLOGO /f VCmakefile CFG=ld
+ nmake /NOLOGO /f VCmakefile CFG=l
+ nmake /NOLOGO /f VCmakefile CFG=td
+ nmake /NOLOGO /f VCmakefile CFG=t
+ nmake /NOLOGO /f VCmakefile CFG=dd
+ nmake /NOLOGO /f VCmakefile
+
+
+clean :
+ -rd tmp_ld /S /Q > NUL: 2>&1
+ -rd tmp_l /S /Q > NUL: 2>&1
+ -rd tmp_td /S /Q > NUL: 2>&1
+ -rd tmp_t /S /Q > NUL: 2>&1
+ -rd tmp_dd /S /Q > NUL: 2>&1
+ -rd tmp /S /Q > NUL: 2>&1
+ -del *.obj *.lib *.dll *.exp *.exe casket /F /Q > NUL: 2>&1
+
+
+"$(OUTDIR)" :
+ if not exist "$(OUTDIR)/" mkdir "$(OUTDIR)"
+
+
+
+#================================================================
+# Building binaries
+#================================================================
+
+
+qdbm$(LIB_APPEND).dll : $(LIBOBJS) qdbm.def
+ link /DLL /DEF:qdbm.def $(LINKFLAGS) /OUT:$@ /IMPLIB:qdbm$(LIB_APPEND).lib $(LIBOBJS)
+
+
+qdbm$(LIB_APPEND).lib : qdbm$(LIB_APPEND).dll
+
+
+qdbm$(LIB_APPEND)_static.lib : $(LIBOBJS)
+ lib $(LIBFLAGS) /OUT:$@ $(LIBOBJS)
+
+
+dpmgr$(EXE_APPEND).exe : $(OUTDIR)\dpmgr.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\dpmgr.obj qdbm$(LIB_APPEND).lib
+
+
+dptest$(EXE_APPEND).exe : $(OUTDIR)\dptest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\dptest.obj qdbm$(LIB_APPEND).lib
+
+
+dptsv$(EXE_APPEND).exe : $(OUTDIR)\dptsv.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\dptsv.obj qdbm$(LIB_APPEND).lib
+
+
+crmgr$(EXE_APPEND).exe : $(OUTDIR)\crmgr.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\crmgr.obj qdbm$(LIB_APPEND).lib
+
+
+crtest$(EXE_APPEND).exe : $(OUTDIR)\crtest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\crtest.obj qdbm$(LIB_APPEND).lib
+
+
+crtsv$(EXE_APPEND).exe : $(OUTDIR)\crtsv.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\crtsv.obj qdbm$(LIB_APPEND).lib
+
+
+rlmgr$(EXE_APPEND).exe : $(OUTDIR)\rlmgr.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\rlmgr.obj qdbm$(LIB_APPEND).lib
+
+
+rltest$(EXE_APPEND).exe : $(OUTDIR)\rltest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\rltest.obj qdbm$(LIB_APPEND).lib
+
+
+hvmgr$(EXE_APPEND).exe : $(OUTDIR)\hvmgr.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\hvmgr.obj qdbm$(LIB_APPEND).lib
+
+
+hvtest$(EXE_APPEND).exe : $(OUTDIR)\hvtest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\hvtest.obj qdbm$(LIB_APPEND).lib
+
+
+cbtest$(EXE_APPEND).exe : $(OUTDIR)\cbtest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\cbtest.obj qdbm$(LIB_APPEND).lib
+
+
+cbcodec$(EXE_APPEND).exe : $(OUTDIR)\cbcodec.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\cbcodec.obj qdbm$(LIB_APPEND).lib
+
+
+vlmgr$(EXE_APPEND).exe : $(OUTDIR)\vlmgr.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\vlmgr.obj qdbm$(LIB_APPEND).lib
+
+
+vltest$(EXE_APPEND).exe : $(OUTDIR)\vltest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\vltest.obj qdbm$(LIB_APPEND).lib
+
+
+vltsv$(EXE_APPEND).exe : $(OUTDIR)\vltsv.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\vltsv.obj qdbm$(LIB_APPEND).lib
+
+
+odmgr$(EXE_APPEND).exe : $(OUTDIR)\odmgr.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\odmgr.obj qdbm$(LIB_APPEND).lib
+
+
+odtest$(EXE_APPEND).exe : $(OUTDIR)\odtest.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\odtest.obj qdbm$(LIB_APPEND).lib
+
+
+odidx$(EXE_APPEND).exe : $(OUTDIR)\odidx.obj qdbm$(LIB_APPEND).lib
+ link $(LINKFLAGS) /OUT:$@ $(OUTDIR)\odidx.obj qdbm$(LIB_APPEND).lib
+
+
+$(OUTDIR)\depot.obj $(OUTDIR)\dpmgr.obj $(OUTDIR)\dptest.obj $(OUTDIR)\dptsv.obj : \
+ depot.h myconf.h
+
+$(OUTDIR)\curia.obj $(OUTDIR)\crmgr.obj $(OUTDIR)\crtest.obj $(OUTDIR)\crtsv.obj : \
+ curia.h depot.h myconf.h
+
+$(OUTDIR)\relic.obj $(OUTDIR)\rlmgr.obj $(OUTDIR)\rltest.obj : \
+ relic.h depot.h myconf.h
+
+$(OUTDIR)\hovel.obj $(OUTDIR)\hvmgr.obj $(OUTDIR)\hvtest.obj : \
+ hovel.h depot.h curia.h myconf.h
+
+$(OUTDIR)\cabin.obj $(OUTDIR)\cbtest.obj $(OUTDIR)\cbcodec.obj : \
+ cabin.h myconf.h
+
+$(OUTDIR)\villa.obj $(OUTDIR)\vlmgr.obj $(OUTDIR)\vltest.obj $(OUTDIR)\vltsv.obj : \
+ villa.h depot.h cabin.h myconf.h
+
+$(OUTDIR)\vista.obj : vista.h villa.h depot.h curia.h cabin.h myconf.h
+
+$(OUTDIR)\odeum.obj $(OUTDIR)\odmgr.obj $(OUTDIR)\odtest.obj $(OUTDIR)\odidx.obj : \
+ odeum.h depot.h curia.h cabin.h villa.h myconf.h
+
+$(OUTDIR)\myconf.obj : myconf.h
+
+
+
+# END OF FILE
Added: box/trunk/qdbm/cabin.c
===================================================================
--- box/trunk/qdbm/cabin.c (rev 0)
+++ box/trunk/qdbm/cabin.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,3529 @@
+/*************************************************************************************************
+ * Implementation of Cabin
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "cabin.h"
+#include "myconf.h"
+
+#define CB_GCUNIT 64 /* allocation unit size of a buffer in gc */
+#define CB_SPBUFSIZ 32 /* size of a buffer for sprintf */
+#define CB_SPMAXWIDTH 128 /* max width of a column for sprintf */
+#define CB_MAPPBNUM 251 /* bucket size of a petit map handle */
+#define CB_MAPCSUNIT 52 /* small allocation unit size of map concatenation */
+#define CB_MAPCBUNIT 252 /* big allocation unit size of map concatenation */
+#define CB_MSGBUFSIZ 256 /* size of a buffer for log message */
+#define CB_IOBUFSIZ 8192 /* size of an I/O buffer */
+#define CB_FILEMODE 00644 /* permission of a creating file */
+#define CB_NUMBUFSIZ 32 /* size of a buffer for a number */
+#define CB_ENCBUFSIZ 32 /* size of a buffer for encoding name */
+#define CB_DATEBUFSIZ 64 /* size of a buffer for date expression */
+#define CB_VNUMBUFSIZ 8 /* size of a buffer for variable length number */
+
+/* set a buffer for a variable length number */
+#define CB_SETVNUMBUF(CB_len, CB_buf, CB_num) \
+ do { \
+ int _CB_num; \
+ _CB_num = (CB_num); \
+ if(_CB_num == 0){ \
+ ((signed char *)(CB_buf))[0] = 0; \
+ (CB_len) = 1; \
+ } else { \
+ (CB_len) = 0; \
+ while(_CB_num > 0){ \
+ int _CB_rem = _CB_num & 0x7f; \
+ _CB_num >>= 7; \
+ if(_CB_num > 0){ \
+ ((signed char *)(CB_buf))[(CB_len)] = -_CB_rem - 1; \
+ } else { \
+ ((signed char *)(CB_buf))[(CB_len)] = _CB_rem; \
+ } \
+ (CB_len)++; \
+ } \
+ } \
+ } while(FALSE)
+
+/* read a variable length buffer */
+#define CB_READVNUMBUF(CB_buf, CB_size, CB_num, CB_step) \
+ do { \
+ int _CB_i, _CB_base; \
+ CB_num = 0; \
+ _CB_base = 1; \
+ if((size) < 2){ \
+ CB_num = ((signed char *)(CB_buf))[0]; \
+ (CB_step) = 1; \
+ } else { \
+ for(_CB_i = 0; _CB_i < (size); _CB_i++){ \
+ if(((signed char *)(CB_buf))[_CB_i] >= 0){ \
+ CB_num += ((signed char *)(CB_buf))[_CB_i] * _CB_base; \
+ break; \
+ } \
+ CB_num += _CB_base * (((signed char *)(CB_buf))[_CB_i] + 1) * -1; \
+ _CB_base *= 128; \
+ } \
+ (CB_step) = _CB_i + 1; \
+ } \
+ } while(FALSE)
+
+/* get the first hash value */
+#define CB_FIRSTHASH(CB_res, CB_kbuf, CB_ksiz) \
+ do { \
+ const unsigned char *_CB_p; \
+ int _CB_ksiz; \
+ _CB_p = (const unsigned char *)(CB_kbuf); \
+ _CB_ksiz = CB_ksiz; \
+ for((CB_res) = 19780211; _CB_ksiz--;){ \
+ (CB_res) = (CB_res) * 37 + *(_CB_p)++; \
+ } \
+ (CB_res) &= INT_MAX; \
+ } while(FALSE)
+
+/* get the second hash value */
+#define CB_SECONDHASH(CB_res, CB_kbuf, CB_ksiz) \
+ do { \
+ const unsigned char *_CB_p; \
+ int _CB_ksiz; \
+ _CB_p = (const unsigned char *)(CB_kbuf) + CB_ksiz - 1; \
+ _CB_ksiz = CB_ksiz; \
+ for((CB_res) = 0x13579bdf; _CB_ksiz--;){ \
+ (CB_res) = (CB_res) * 31 + *(_CB_p)--; \
+ } \
+ (CB_res) &= INT_MAX; \
+ } while(FALSE)
+
+
+/* private function prototypes */
+static void cbggchandler(void);
+static void cbggckeeper(void *ptr, void (*func)(void *));
+static void cbqsortsub(char *bp, int nmemb, int size, char *pswap, char *vswap,
+ int(*compar)(const void *, const void *));
+static int cblistelemcmp(const void *a, const void *b);
+static int cbkeycmp(const char *abuf, int asiz, const char *bbuf, int bsiz);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* Call back function for handling a fatal error. */
+void (*cbfatalfunc)(const char *message) = NULL;
+
+
+/* Allocate a region on memory. */
+void *cbmalloc(size_t size){
+ char *p;
+ assert(size > 0 && size < INT_MAX);
+ if(!(p = malloc(size))) cbmyfatal("out of memory");
+ return p;
+}
+
+
+/* Re-allocate a region on memory. */
+void *cbrealloc(void *ptr, size_t size){
+ char *p;
+ assert(size > 0);
+ if(!(p = realloc(ptr, size))) cbmyfatal("out of memory");
+ return p;
+}
+
+
+/* Duplicate a region on memory. */
+char *cbmemdup(const char *ptr, int size){
+ char *p;
+ assert(ptr);
+ if(size < 0) size = strlen(ptr);
+ CB_MALLOC(p, size + 1);
+ memcpy(p, ptr, size);
+ p[size] = '\0';
+ return p;
+}
+
+
+/* Free a region on memory. */
+void cbfree(void *ptr){
+ free(ptr);
+}
+
+
+/* Register the pointer or handle of an object to the global garbage collector. */
+void cbglobalgc(void *ptr, void (*func)(void *)){
+ assert(ptr && func);
+ cbggckeeper(ptr, func);
+}
+
+
+/* Exercise the global garbage collector explicitly. */
+void cbggcsweep(void){
+ cbggckeeper(NULL, NULL);
+}
+
+
+/* Check availability of allocation of the virtual memory. */
+int cbvmemavail(size_t size){
+ assert(size >= 0);
+ return _qdbm_vmemavail(size);
+}
+
+
+/* Sort an array using insert sort. */
+void cbisort(void *base, int nmemb, int size, int(*compar)(const void *, const void *)){
+ char *bp, *swap;
+ int i, j;
+ assert(base && nmemb >= 0 && size > 0 && compar);
+ bp = (char *)base;
+ CB_MALLOC(swap, size);
+ for(i = 1; i < nmemb; i++){
+ if(compar(bp + (i - 1) * size, bp + i * size) > 0){
+ memcpy(swap, bp + i * size, size);
+ for(j = i; j > 0; j--){
+ if(compar(bp + (j - 1) * size, swap) < 0) break;
+ memcpy(bp + j * size, bp + (j - 1) * size, size);
+ }
+ memcpy(bp + j * size, swap, size);
+ }
+ }
+ free(swap);
+}
+
+
+/* Sort an array using shell sort. */
+void cbssort(void *base, int nmemb, int size, int(*compar)(const void *, const void *)){
+ char *bp, *swap;
+ int step, bottom, i, j;
+ assert(base && nmemb >= 0 && size > 0 && compar);
+ bp = (char *)base;
+ CB_MALLOC(swap, size);
+ for(step = (nmemb - 1) / 3; step >= 0; step = (step - 1) / 3){
+ if(step < 5) step = 1;
+ for(bottom = 0; bottom < step; bottom++){
+ for(i = bottom + step; i < nmemb; i += step){
+ if(compar(bp + (i - step) * size, bp + i * size) > 0){
+ memcpy(swap, bp + i * size, size);
+ for(j = i; j > step - 1; j -= step){
+ if(compar(bp + (j - step) * size, swap) < 0) break;
+ memcpy(bp + j * size, bp + (j - step) * size, size);
+ }
+ memcpy(bp + j * size, swap, size);
+ }
+ }
+ }
+ if(step < 2) break;
+ }
+ free(swap);
+}
+
+
+/* Sort an array using heap sort. */
+void cbhsort(void *base, int nmemb, int size, int(*compar)(const void *, const void *)){
+ char *bp, *swap;
+ int top, bottom, mybot, i;
+ assert(base && nmemb >= 0 && size > 0 && compar);
+ bp = (char *)base;
+ nmemb--;
+ bottom = nmemb / 2 + 1;
+ top = nmemb;
+ CB_MALLOC(swap, size);
+ while(bottom > 0){
+ bottom--;
+ mybot = bottom;
+ i = 2 * mybot;
+ while(i <= top) {
+ if(i < top && compar(bp + (i + 1) * size, bp + i * size) > 0) i++;
+ if(compar(bp + mybot * size, bp + i * size) >= 0) break;
+ memcpy(swap, bp + mybot * size, size);
+ memcpy(bp + mybot * size, bp + i * size, size);
+ memcpy(bp + i * size, swap, size);
+ mybot = i;
+ i = 2 * mybot;
+ }
+ }
+ while(top > 0){
+ memcpy(swap, bp, size);
+ memcpy(bp, bp + top * size, size);
+ memcpy(bp + top * size, swap, size);
+ top--;
+ mybot = bottom;
+ i = 2 * mybot;
+ while(i <= top){
+ if(i < top && compar(bp + (i + 1) * size, bp + i * size) > 0) i++;
+ if(compar(bp + mybot * size, bp + i * size) >= 0) break;
+ memcpy(swap, bp + mybot * size, size);
+ memcpy(bp + mybot * size, bp + i * size, size);
+ memcpy(bp + i * size, swap, size);
+ mybot = i;
+ i = 2 * mybot;
+ }
+ }
+ free(swap);
+}
+
+
+/* Sort an array using quick sort. */
+void cbqsort(void *base, int nmemb, int size, int(*compar)(const void *, const void *)){
+ char *pswap, *vswap;
+ assert(base && nmemb >= 0 && size > 0 && compar);
+ CB_MALLOC(pswap, size);
+ CB_MALLOC(vswap, size);
+ cbqsortsub(base, nmemb, size, pswap, vswap, compar);
+ free(vswap);
+ free(pswap);
+}
+
+
+/* Compare two strings with case insensitive evaluation. */
+int cbstricmp(const char *astr, const char *bstr){
+ int ac, bc;
+ assert(astr && bstr);
+ while(*astr != '\0'){
+ if(*bstr == '\0') return 1;
+ ac = (*astr >= 'A' && *astr <= 'Z') ? *astr + ('a' - 'A') : *(unsigned char *)astr;
+ bc = (*bstr >= 'A' && *bstr <= 'Z') ? *bstr + ('a' - 'A') : *(unsigned char *)bstr;
+ if(ac != bc) return ac - bc;
+ astr++;
+ bstr++;
+ }
+ return *bstr == '\0' ? 0 : -1;
+}
+
+
+/* Check whether a string begins with a key. */
+int cbstrfwmatch(const char *str, const char *key){
+ assert(str && key);
+ while(*key != '\0'){
+ if(*str != *key || *str == '\0') return FALSE;
+ key++;
+ str++;
+ }
+ return TRUE;
+}
+
+
+/* Check whether a string begins with a key, with case insensitive evaluation. */
+int cbstrfwimatch(const char *str, const char *key){
+ int sc, kc;
+ assert(str && key);
+ while(*key != '\0'){
+ if(*str == '\0') return FALSE;
+ sc = *str;
+ if(sc >= 'A' && sc <= 'Z') sc += 'a' - 'A';
+ kc = *key;
+ if(kc >= 'A' && kc <= 'Z') kc += 'a' - 'A';
+ if(sc != kc) return FALSE;
+ key++;
+ str++;
+ }
+ return TRUE;
+}
+
+
+/* Check whether a string ends with a key. */
+int cbstrbwmatch(const char *str, const char *key){
+ int slen, klen, i;
+ assert(str && key);
+ slen = strlen(str);
+ klen = strlen(key);
+ for(i = 1; i <= klen; i++){
+ if(i > slen || str[slen-i] != key[klen-i]) return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Check whether a string ends with a key, with case insensitive evaluation. */
+int cbstrbwimatch(const char *str, const char *key){
+ int slen, klen, i, sc, kc;
+ assert(str && key);
+ slen = strlen(str);
+ klen = strlen(key);
+ for(i = 1; i <= klen; i++){
+ if(i > slen) return FALSE;
+ sc = str[slen-i];
+ if(sc >= 'A' && sc <= 'Z') sc += 'a' - 'A';
+ kc = key[klen-i];
+ if(kc >= 'A' && kc <= 'Z') kc += 'a' - 'A';
+ if(sc != kc) return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Locate a substring in a string using KMP method. */
+char *cbstrstrkmp(const char *haystack, const char *needle){
+ int i, j, hlen, nlen;
+ signed char tbl[0x100];
+ assert(haystack && needle);
+ nlen = strlen(needle);
+ if(nlen >= 0x100) return strstr(haystack, needle);
+ tbl[0] = -1;
+ i = 0;
+ j = -1;
+ while(i < nlen){
+ while((j >= 0) && (needle[i] != needle[j])){
+ j = tbl[j];
+ }
+ i++;
+ j++;
+ tbl[i] = j;
+ }
+ hlen = strlen(haystack);
+ i = 0;
+ j = 0;
+ while(i < hlen && j < nlen){
+ while((j >= 0) && (haystack[i] != needle[j])){
+ j = tbl[j];
+ }
+ i++;
+ j++;
+ }
+ if(j == nlen) return (char *)(haystack + i - nlen);
+ return NULL;
+}
+
+
+/* Locate a substring in a string using BM method. */
+char *cbstrstrbm(const char *haystack, const char *needle){
+ const unsigned char *rp;
+ const char *ep;
+ unsigned char tbl[0x100];
+ int i, j, nlen, len, idx;
+ assert(haystack && needle);
+ nlen = strlen(needle);
+ if(nlen < 3 || nlen >= 0x100) return strstr(haystack, needle);
+ for(i = 0; i < 0x100; i++){
+ tbl[i] = nlen;
+ }
+ len = nlen;
+ rp = (const unsigned char *)needle;
+ while(len > 0){
+ tbl[*rp++] = --len;
+ }
+ nlen--;
+ ep = haystack + strlen(haystack) - nlen;
+ while(haystack < ep){
+ for(i = nlen; haystack[i] == needle[i]; i--){
+ if(i == 0) return (char *)haystack;
+ }
+ idx = ((unsigned char *)haystack)[i];
+ j = tbl[idx] - nlen + i;
+ haystack += j > 0 ? j : 2;
+ }
+ return NULL;
+}
+
+
+/* Convert the letters of a string to upper case. */
+char *cbstrtoupper(char *str){
+ int i;
+ assert(str);
+ for(i = 0; str[i] != '\0'; i++){
+ if(str[i] >= 'a' && str[i] <= 'z') str[i] -= 'a' - 'A';
+ }
+ return str;
+}
+
+
+/* Convert the letters of a string to lower case. */
+char *cbstrtolower(char *str){
+ int i;
+ assert(str);
+ for(i = 0; str[i] != '\0'; i++){
+ if(str[i] >= 'A' && str[i] <= 'Z') str[i] += 'a' - 'A';
+ }
+ return str;
+}
+
+
+/* Cut space characters at head or tail of a string. */
+char *cbstrtrim(char *str){
+ char *wp;
+ int i, head;
+ assert(str);
+ wp = str;
+ head = TRUE;
+ for(i = 0; str[i] != '\0'; i++){
+ if((str[i] >= 0x07 && str[i] <= 0x0d) || str[i] == 0x20){
+ if(!head) *(wp++) = str[i];
+ } else {
+ *(wp++) = str[i];
+ head = FALSE;
+ }
+ }
+ *wp = '\0';
+ while(wp > str && ((wp[-1] >= 0x07 && wp[-1] <= 0x0d) || wp[-1] == 0x20)){
+ *(--wp) = '\0';
+ }
+ return str;
+}
+
+
+/* Squeeze space characters in a string and trim it. */
+char *cbstrsqzspc(char *str){
+ char *wp;
+ int i, spc;
+ assert(str);
+ wp = str;
+ spc = TRUE;
+ for(i = 0; str[i] != '\0'; i++){
+ if(str[i] > 0 && str[i] <= ' '){
+ if(!spc) *(wp++) = str[i];
+ spc = TRUE;
+ } else {
+ *(wp++) = str[i];
+ spc = FALSE;
+ }
+ }
+ *wp = '\0';
+ for(wp--; wp >= str; wp--){
+ if(*wp > 0 && *wp <= ' '){
+ *wp = '\0';
+ } else {
+ break;
+ }
+ }
+ return str;
+}
+
+
+/* Count the number of characters in a string of UTF-8. */
+int cbstrcountutf(const char *str){
+ const unsigned char *rp;
+ int cnt;
+ assert(str);
+ rp = (unsigned char *)str;
+ cnt = 0;
+ while(*rp != '\0'){
+ if((*rp & 0x80) == 0x00 || (*rp & 0xe0) == 0xc0 ||
+ (*rp & 0xf0) == 0xe0 || (*rp & 0xf8) == 0xf0) cnt++;
+ rp++;
+ }
+ return cnt;
+}
+
+
+/* Cut a string of UTF-8 at the specified number of characters. */
+char *cbstrcututf(char *str, int num){
+ unsigned char *wp;
+ int cnt;
+ assert(str && num >= 0);
+ wp = (unsigned char *)str;
+ cnt = 0;
+ while(*wp != '\0'){
+ if((*wp & 0x80) == 0x00 || (*wp & 0xe0) == 0xc0 ||
+ (*wp & 0xf0) == 0xe0 || (*wp & 0xf8) == 0xf0){
+ cnt++;
+ if(cnt > num){
+ *wp = '\0';
+ break;
+ }
+ }
+ wp++;
+ }
+ return str;
+}
+
+
+/* Get a datum handle. */
+CBDATUM *cbdatumopen(const char *ptr, int size){
+ CBDATUM *datum;
+ CB_MALLOC(datum, sizeof(*datum));
+ CB_MALLOC(datum->dptr, CB_DATUMUNIT);
+ datum->dptr[0] = '\0';
+ datum->dsize = 0;
+ datum->asize = CB_DATUMUNIT;
+ if(ptr) CB_DATUMCAT(datum, ptr, (size >= 0 ? size : strlen(ptr)));
+ return datum;
+}
+
+
+/* Copy a datum. */
+CBDATUM *cbdatumdup(const CBDATUM *datum){
+ assert(datum);
+ return cbdatumopen(datum->dptr, datum->dsize);
+}
+
+
+/* Free a datum handle. */
+void cbdatumclose(CBDATUM *datum){
+ assert(datum);
+ free(datum->dptr);
+ free(datum);
+}
+
+
+/* Concatenate a datum and a region. */
+void cbdatumcat(CBDATUM *datum, const char *ptr, int size){
+ assert(datum && ptr);
+ if(size < 0) size = strlen(ptr);
+ if(datum->dsize + size >= datum->asize){
+ datum->asize = datum->asize * 2 + size + 1;
+ CB_REALLOC(datum->dptr, datum->asize);
+ }
+ memcpy(datum->dptr + datum->dsize, ptr, size);
+ datum->dsize += size;
+ datum->dptr[datum->dsize] = '\0';
+}
+
+
+/* Get the pointer of the region of a datum. */
+const char *cbdatumptr(const CBDATUM *datum){
+ assert(datum);
+ return datum->dptr;
+}
+
+
+/* Get the size of the region of a datum. */
+int cbdatumsize(const CBDATUM *datum){
+ assert(datum);
+ return datum->dsize;
+}
+
+
+/* Set the size of the region of a datum. */
+void cbdatumsetsize(CBDATUM *datum, int size){
+ assert(datum && size >= 0);
+ if(size <= datum->dsize){
+ datum->dsize = size;
+ datum->dptr[size] = '\0';
+ } else {
+ if(size >= datum->asize){
+ datum->asize = datum->asize * 2 + size + 1;
+ CB_REALLOC(datum->dptr, datum->asize);
+ }
+ memset(datum->dptr + datum->dsize, 0, (size - datum->dsize) + 1);
+ datum->dsize = size;
+ }
+}
+
+
+/* Perform formatted output into a datum. */
+void cbdatumprintf(CBDATUM *datum, const char *format, ...){
+ va_list ap;
+ char *tmp, cbuf[CB_NUMBUFSIZ], tbuf[CB_NUMBUFSIZ*2];
+ unsigned char c;
+ int cblen, tlen;
+ assert(datum && format);
+ va_start(ap, format);
+ while(*format != '\0'){
+ if(*format == '%'){
+ cbuf[0] = '%';
+ cblen = 1;
+ format++;
+ while(strchr("0123456789 .+-", *format) && *format != '\0' && cblen < CB_NUMBUFSIZ - 1){
+ cbuf[cblen++] = *format;
+ format++;
+ }
+ cbuf[cblen++] = *format;
+ cbuf[cblen] = '\0';
+ switch(*format){
+ case 's':
+ tmp = va_arg(ap, char *);
+ if(!tmp) tmp = "(null)";
+ cbdatumcat(datum, tmp, -1);
+ break;
+ case 'd':
+ tlen = sprintf(tbuf, cbuf, va_arg(ap, int));
+ cbdatumcat(datum, tbuf, tlen);
+ break;
+ case 'o': case 'u': case 'x': case 'X': case 'c':
+ tlen = sprintf(tbuf, cbuf, va_arg(ap, unsigned int));
+ cbdatumcat(datum, tbuf, tlen);
+ break;
+ case 'e': case 'E': case 'f': case 'g': case 'G':
+ tlen = sprintf(tbuf, cbuf, va_arg(ap, double));
+ cbdatumcat(datum, tbuf, tlen);
+ break;
+ case '@':
+ tmp = va_arg(ap, char *);
+ if(!tmp) tmp = "(null)";
+ while(*tmp){
+ switch(*tmp){
+ case '&': cbdatumcat(datum, "&", 5); break;
+ case '<': cbdatumcat(datum, "<", 4); break;
+ case '>': cbdatumcat(datum, ">", 4); break;
+ case '"': cbdatumcat(datum, """, 6); break;
+ default:
+ if(!((*tmp >= 0 && *tmp <= 0x8) || (*tmp >= 0x0e && *tmp <= 0x1f)))
+ cbdatumcat(datum, tmp, 1);
+ break;
+ }
+ tmp++;
+ }
+ break;
+ case '?':
+ tmp = va_arg(ap, char *);
+ if(!tmp) tmp = "(null)";
+ while(*tmp){
+ c = *(unsigned char *)tmp;
+ if((c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z') ||
+ (c >= '0' && c <= '9') || (c != '\0' && strchr("_-.", c))){
+ cbdatumcat(datum, tmp, 1);
+ } else {
+ tlen = sprintf(tbuf, "%%%02X", c);
+ cbdatumcat(datum, tbuf, tlen);
+ }
+ tmp++;
+ }
+ break;
+ case ':':
+ tmp = va_arg(ap, char *);
+ if(!tmp) tmp = "";
+ tmp = cbmimeencode(tmp, "UTF-8", TRUE);
+ cbdatumcat(datum, tmp, -1);
+ free(tmp);
+ break;
+ case '%':
+ cbdatumcat(datum, "%", 1);
+ break;
+ }
+ } else {
+ cbdatumcat(datum, format, 1);
+ }
+ format++;
+ }
+ va_end(ap);
+}
+
+
+/* Convert a datum to an allocated region. */
+char *cbdatumtomalloc(CBDATUM *datum, int *sp){
+ char *ptr;
+ assert(datum);
+ ptr = datum->dptr;
+ if(sp) *sp = datum->dsize;
+ free(datum);
+ return ptr;
+}
+
+
+/* Get a list handle. */
+CBLIST *cblistopen(void){
+ CBLIST *list;
+ CB_MALLOC(list, sizeof(*list));
+ list->anum = CB_LISTUNIT;
+ CB_MALLOC(list->array, sizeof(list->array[0]) * list->anum);
+ list->start = 0;
+ list->num = 0;
+ return list;
+}
+
+
+/* Copy a list. */
+CBLIST *cblistdup(const CBLIST *list){
+ CBLIST *newlist;
+ int i, size;
+ const char *val;
+ assert(list);
+ CB_LISTOPEN2(newlist, CB_LISTNUM(list));
+ for(i = 0; i < CB_LISTNUM(list); i++){
+ val = CB_LISTVAL2(list, i, size);
+ CB_LISTPUSH(newlist, val, size);
+ }
+ return newlist;
+}
+
+
+/* Close a list handle. */
+void cblistclose(CBLIST *list){
+ int i, end;
+ assert(list);
+ end = list->start + list->num;
+ for(i = list->start; i < end; i++){
+ free(list->array[i].dptr);
+ }
+ free(list->array);
+ free(list);
+}
+
+
+/* Get the number of elements of a list. */
+int cblistnum(const CBLIST *list){
+ assert(list);
+ return list->num;
+}
+
+
+/* Get the pointer to the region of an element. */
+const char *cblistval(const CBLIST *list, int index, int *sp){
+ assert(list && index >= 0);
+ if(index >= list->num) return NULL;
+ index += list->start;
+ if(sp) *sp = list->array[index].dsize;
+ return list->array[index].dptr;
+}
+
+
+/* Add an element at the end of a list. */
+void cblistpush(CBLIST *list, const char *ptr, int size){
+ int index;
+ assert(list && ptr);
+ if(size < 0) size = strlen(ptr);
+ index = list->start + list->num;
+ if(index >= list->anum){
+ list->anum *= 2;
+ CB_REALLOC(list->array, list->anum * sizeof(list->array[0]));
+ }
+ CB_MALLOC(list->array[index].dptr, (size < CB_DATUMUNIT ? CB_DATUMUNIT : size) + 1);
+ memcpy(list->array[index].dptr, ptr, size);
+ list->array[index].dptr[size] = '\0';
+ list->array[index].dsize = size;
+ list->num++;
+}
+
+
+/* Remove an element of the end of a list. */
+char *cblistpop(CBLIST *list, int *sp){
+ int index;
+ assert(list);
+ if(list->num < 1) return NULL;
+ index = list->start + list->num - 1;
+ list->num--;
+ if(sp) *sp = list->array[index].dsize;
+ return list->array[index].dptr;
+}
+
+
+/* Add an element at the top of a list. */
+void cblistunshift(CBLIST *list, const char *ptr, int size){
+ int index;
+ assert(list && ptr);
+ if(size < 0) size = strlen(ptr);
+ if(list->start < 1){
+ if(list->start + list->num >= list->anum){
+ list->anum *= 2;
+ CB_REALLOC(list->array, list->anum * sizeof(list->array[0]));
+ }
+ list->start = list->anum - list->num;
+ memmove(list->array + list->start, list->array, list->num * sizeof(list->array[0]));
+ }
+ index = list->start - 1;
+ CB_MALLOC(list->array[index].dptr, (size < CB_DATUMUNIT ? CB_DATUMUNIT : size) + 1);
+ memcpy(list->array[index].dptr, ptr, size);
+ list->array[index].dptr[size] = '\0';
+ list->array[index].dsize = size;
+ list->start--;
+ list->num++;
+}
+
+
+/* Remove an element of the top of a list. */
+char *cblistshift(CBLIST *list, int *sp){
+ int index;
+ assert(list);
+ if(list->num < 1) return NULL;
+ index = list->start;
+ list->start++;
+ list->num--;
+ if(sp) *sp = list->array[index].dsize;
+ return list->array[index].dptr;
+}
+
+
+/* Add an element at the specified location of a list. */
+void cblistinsert(CBLIST *list, int index, const char *ptr, int size){
+ assert(list && index >= 0);
+ if(index > list->num) return;
+ if(size < 0) size = strlen(ptr);
+ index += list->start;
+ if(list->start + list->num >= list->anum){
+ list->anum *= 2;
+ CB_REALLOC(list->array, list->anum * sizeof(list->array[0]));
+ }
+ memmove(list->array + index + 1, list->array + index,
+ sizeof(list->array[0]) * (list->start + list->num - index));
+ CB_MEMDUP(list->array[index].dptr, ptr, size);
+ list->array[index].dsize = size;
+ list->num++;
+}
+
+
+/* Remove an element at the specified location of a list. */
+char *cblistremove(CBLIST *list, int index, int *sp){
+ char *dptr;
+ assert(list && index >= 0);
+ if(index >= list->num) return NULL;
+ index += list->start;
+ dptr = list->array[index].dptr;
+ if(sp) *sp = list->array[index].dsize;
+ list->num--;
+ memmove(list->array + index, list->array + index + 1,
+ sizeof(list->array[0]) * (list->start + list->num - index));
+ return dptr;
+}
+
+
+/* Overwrite an element at the specified location of a list. */
+void cblistover(CBLIST *list, int index, const char *ptr, int size){
+ assert(list && index >= 0);
+ if(index >= list->num) return;
+ if(size < 0) size = strlen(ptr);
+ index += list->start;
+ if(size > list->array[index].dsize)
+ CB_REALLOC(list->array[index].dptr, size + 1);
+ memcpy(list->array[index].dptr, ptr, size);
+ list->array[index].dsize = size;
+ list->array[index].dptr[size] = '\0';
+}
+
+
+/* Sort elements of a list in lexical order. */
+void cblistsort(CBLIST *list){
+ assert(list);
+ qsort(list->array + list->start, list->num, sizeof(list->array[0]), cblistelemcmp);
+}
+
+
+/* Search a list for an element using liner search. */
+int cblistlsearch(const CBLIST *list, const char *ptr, int size){
+ int i, end;
+ assert(list && ptr);
+ if(size < 0) size = strlen(ptr);
+ end = list->start + list->num;
+ for(i = list->start; i < end; i++){
+ if(list->array[i].dsize == size && !memcmp(list->array[i].dptr, ptr, size))
+ return i - list->start;
+ }
+ return -1;
+}
+
+
+/* Search a list for an element using binary search. */
+int cblistbsearch(const CBLIST *list, const char *ptr, int size){
+ CBLISTDATUM key, *res;
+ assert(list && ptr);
+ if(size < 0) size = strlen(ptr);
+ CB_MEMDUP(key.dptr, ptr, size);
+ key.dsize = size;
+ res = bsearch(&key, list->array + list->start, list->num, sizeof(list->array[0]), cblistelemcmp);
+ free(key.dptr);
+ return res ? (res - list->array - list->start) : -1;
+}
+
+
+/* Serialize a list into a byte array. */
+char *cblistdump(const CBLIST *list, int *sp){
+ char *buf, vnumbuf[CB_VNUMBUFSIZ];
+ const char *vbuf;
+ int i, bsiz, vnumsiz, ln, vsiz;
+ assert(list && sp);
+ ln = CB_LISTNUM(list);
+ CB_SETVNUMBUF(vnumsiz, vnumbuf, ln);
+ CB_MALLOC(buf, vnumsiz + 1);
+ memcpy(buf, vnumbuf, vnumsiz);
+ bsiz = vnumsiz;
+ for(i = 0; i < ln; i++){
+ vbuf = CB_LISTVAL2(list, i, vsiz);
+ CB_SETVNUMBUF(vnumsiz, vnumbuf, vsiz);
+ CB_REALLOC(buf, bsiz + vnumsiz + vsiz + 1);
+ memcpy(buf + bsiz, vnumbuf, vnumsiz);
+ bsiz += vnumsiz;
+ memcpy(buf + bsiz, vbuf, vsiz);
+ bsiz += vsiz;
+ }
+ *sp = bsiz;
+ return buf;
+}
+
+
+/* Redintegrate a serialized list. */
+CBLIST *cblistload(const char *ptr, int size){
+ CBLIST *list;
+ const char *rp;
+ int i, anum, step, ln, vsiz;
+ assert(ptr && size >= 0);
+ anum = size / (sizeof(CBLISTDATUM) + 1);
+ CB_LISTOPEN2(list, anum);
+ rp = ptr;
+ CB_READVNUMBUF(rp, size, ln, step);
+ rp += step;
+ size -= step;
+ if(ln > size) return list;
+ for(i = 0; i < ln; i++){
+ if(size < 1) break;
+ CB_READVNUMBUF(rp, size, vsiz, step);
+ rp += step;
+ size -= step;
+ if(vsiz > size) break;
+ CB_LISTPUSH(list, rp, vsiz);
+ rp += vsiz;
+ }
+ return list;
+}
+
+
+/* Get a map handle. */
+CBMAP *cbmapopen(void){
+ return cbmapopenex(CB_MAPBNUM);
+}
+
+
+/* Copy a map. */
+CBMAP *cbmapdup(CBMAP *map){
+ CBMAP *newmap;
+ const char *kbuf, *vbuf;
+ int ksiz, vsiz;
+ assert(map);
+ cbmapiterinit(map);
+ newmap = map->rnum > CB_MAPPBNUM ? cbmapopen() : cbmapopenex(CB_MAPPBNUM);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ CB_MAPITERVAL(vbuf, kbuf, vsiz);
+ cbmapput(newmap, kbuf, ksiz, vbuf, vsiz, FALSE);
+ }
+ cbmapiterinit(map);
+ return newmap;
+}
+
+
+/* Close a map handle. */
+void cbmapclose(CBMAP *map){
+ CBMAPDATUM *datum, *next;
+ datum = map->first;
+ while(datum){
+ next = datum->next;
+ free(datum);
+ datum = next;
+ }
+ free(map->buckets);
+ free(map);
+}
+
+
+/* Store a record. */
+int cbmapput(CBMAP *map, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int over){
+ CBMAPDATUM *datum, **entp, *old;
+ char *dbuf;
+ int bidx, hash, kcmp, psiz;
+ assert(map && kbuf && vbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(vsiz < 0) vsiz = strlen(vbuf);
+ CB_FIRSTHASH(hash, kbuf, ksiz);
+ bidx = hash % map->bnum;
+ datum = map->buckets[bidx];
+ entp = map->buckets + bidx;
+ CB_SECONDHASH(hash, kbuf, ksiz);
+ while(datum){
+ if(hash > datum->hash){
+ entp = &(datum->left);
+ datum = datum->left;
+ } else if(hash < datum->hash){
+ entp = &(datum->right);
+ datum = datum->right;
+ } else {
+ dbuf = (char *)datum + sizeof(*datum);
+ kcmp = cbkeycmp(kbuf, ksiz, dbuf, datum->ksiz);
+ if(kcmp < 0){
+ entp = &(datum->left);
+ datum = datum->left;
+ } else if(kcmp > 0){
+ entp = &(datum->right);
+ datum = datum->right;
+ } else {
+ if(!over) return FALSE;
+ psiz = CB_ALIGNPAD(ksiz);
+ if(vsiz > datum->vsiz){
+ old = datum;
+ CB_REALLOC(datum, sizeof(*datum) + ksiz + psiz + vsiz + 1);
+ if(datum != old){
+ if(map->first == old) map->first = datum;
+ if(map->last == old) map->last = datum;
+ if(*entp == old) *entp = datum;
+ if(datum->prev) datum->prev->next = datum;
+ if(datum->next) datum->next->prev = datum;
+ dbuf = (char *)datum + sizeof(*datum);
+ }
+ }
+ memcpy(dbuf + ksiz + psiz, vbuf, vsiz);
+ dbuf[ksiz+psiz+vsiz] = '\0';
+ datum->vsiz = vsiz;
+ return TRUE;
+ }
+ }
+ }
+ psiz = CB_ALIGNPAD(ksiz);
+ CB_MALLOC(datum, sizeof(*datum) + ksiz + psiz + vsiz + 1);
+ dbuf = (char *)datum + sizeof(*datum);
+ memcpy(dbuf, kbuf, ksiz);
+ dbuf[ksiz] = '\0';
+ datum->ksiz = ksiz;
+ memcpy(dbuf + ksiz + psiz, vbuf, vsiz);
+ dbuf[ksiz+psiz+vsiz] = '\0';
+ datum->vsiz = vsiz;
+ datum->hash = hash;
+ datum->left = NULL;
+ datum->right = NULL;
+ datum->prev = map->last;
+ datum->next = NULL;
+ *entp = datum;
+ if(!map->first) map->first = datum;
+ if(map->last) map->last->next = datum;
+ map->last = datum;
+ map->rnum++;
+ return TRUE;
+}
+
+
+/* Concatenate a value at the end of the value of the existing record. */
+void cbmapputcat(CBMAP *map, const char *kbuf, int ksiz, const char *vbuf, int vsiz){
+ CBMAPDATUM *datum, **entp, *old;
+ char *dbuf;
+ int bidx, hash, kcmp, psiz, asiz, unit;
+ assert(map && kbuf && vbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(vsiz < 0) vsiz = strlen(vbuf);
+ CB_FIRSTHASH(hash, kbuf, ksiz);
+ bidx = hash % map->bnum;
+ datum = map->buckets[bidx];
+ entp = map->buckets + bidx;
+ CB_SECONDHASH(hash, kbuf, ksiz);
+ while(datum){
+ if(hash > datum->hash){
+ entp = &(datum->left);
+ datum = datum->left;
+ } else if(hash < datum->hash){
+ entp = &(datum->right);
+ datum = datum->right;
+ } else {
+ dbuf = (char *)datum + sizeof(*datum);
+ kcmp = cbkeycmp(kbuf, ksiz, dbuf, datum->ksiz);
+ if(kcmp < 0){
+ entp = &(datum->left);
+ datum = datum->left;
+ } else if(kcmp > 0){
+ entp = &(datum->right);
+ datum = datum->right;
+ } else {
+ psiz = CB_ALIGNPAD(ksiz);
+ asiz = sizeof(*datum) + ksiz + psiz + datum->vsiz + vsiz + 1;
+ unit = asiz <= CB_MAPCSUNIT ? CB_MAPCSUNIT : CB_MAPCBUNIT;
+ asiz = (asiz - 1) + unit - (asiz - 1) % unit;
+ old = datum;
+ CB_REALLOC(datum, asiz);
+ if(datum != old){
+ if(map->first == old) map->first = datum;
+ if(map->last == old) map->last = datum;
+ if(*entp == old) *entp = datum;
+ if(datum->prev) datum->prev->next = datum;
+ if(datum->next) datum->next->prev = datum;
+ dbuf = (char *)datum + sizeof(*datum);
+ }
+ memcpy(dbuf + ksiz + psiz + datum->vsiz, vbuf, vsiz);
+ dbuf[ksiz+psiz+datum->vsiz+vsiz] = '\0';
+ datum->vsiz += vsiz;
+ return;
+ }
+ }
+ }
+ psiz = CB_ALIGNPAD(ksiz);
+ asiz = sizeof(*datum) + ksiz + psiz + vsiz + 1;
+ unit = asiz <= CB_MAPCSUNIT ? CB_MAPCSUNIT : CB_MAPCBUNIT;
+ asiz = (asiz - 1) + unit - (asiz - 1) % unit;
+ CB_MALLOC(datum, asiz);
+ dbuf = (char *)datum + sizeof(*datum);
+ memcpy(dbuf, kbuf, ksiz);
+ dbuf[ksiz] = '\0';
+ datum->ksiz = ksiz;
+ memcpy(dbuf + ksiz + psiz, vbuf, vsiz);
+ dbuf[ksiz+psiz+vsiz] = '\0';
+ datum->vsiz = vsiz;
+ datum->hash = hash;
+ datum->left = NULL;
+ datum->right = NULL;
+ datum->prev = map->last;
+ datum->next = NULL;
+ *entp = datum;
+ if(!map->first) map->first = datum;
+ if(map->last) map->last->next = datum;
+ map->last = datum;
+ map->rnum++;
+}
+
+
+/* Delete a record. */
+int cbmapout(CBMAP *map, const char *kbuf, int ksiz){
+ CBMAPDATUM *datum, **entp, *tmp;
+ char *dbuf;
+ int bidx, hash, kcmp;
+ assert(map && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ CB_FIRSTHASH(hash, kbuf, ksiz);
+ bidx = hash % map->bnum;
+ datum = map->buckets[bidx];
+ entp = map->buckets + bidx;
+ CB_SECONDHASH(hash, kbuf, ksiz);
+ while(datum){
+ if(hash > datum->hash){
+ entp = &(datum->left);
+ datum = datum->left;
+ } else if(hash < datum->hash){
+ entp = &(datum->right);
+ datum = datum->right;
+ } else {
+ dbuf = (char *)datum + sizeof(*datum);
+ kcmp = cbkeycmp(kbuf, ksiz, dbuf, datum->ksiz);
+ if(kcmp < 0){
+ entp = &(datum->left);
+ datum = datum->left;
+ } else if(kcmp > 0){
+ entp = &(datum->right);
+ datum = datum->right;
+ } else {
+ if(datum->prev) datum->prev->next = datum->next;
+ if(datum->next) datum->next->prev = datum->prev;
+ if(datum == map->first) map->first = datum->next;
+ if(datum == map->last) map->last = datum->prev;
+ if(datum->left && !datum->right){
+ *entp = datum->left;
+ } else if(!datum->left && datum->right){
+ *entp = datum->right;
+ } else if(!datum->left && !datum->left){
+ *entp = NULL;
+ } else {
+ *entp = datum->left;
+ tmp = *entp;
+ while(TRUE){
+ if(hash > tmp->hash){
+ if(tmp->left){
+ tmp = tmp->left;
+ } else {
+ tmp->left = datum->right;
+ break;
+ }
+ } else if(hash < tmp->hash){
+ if(tmp->right){
+ tmp = tmp->right;
+ } else {
+ tmp->right = datum->right;
+ break;
+ }
+ } else {
+ kcmp = cbkeycmp(kbuf, ksiz, dbuf, datum->ksiz);
+ if(kcmp < 0){
+ if(tmp->left){
+ tmp = tmp->left;
+ } else {
+ tmp->left = datum->right;
+ break;
+ }
+ } else {
+ if(tmp->right){
+ tmp = tmp->right;
+ } else {
+ tmp->right = datum->right;
+ break;
+ }
+ }
+ }
+ }
+ }
+ free(datum);
+ map->rnum--;
+ return TRUE;
+ }
+ }
+ }
+ return FALSE;
+}
+
+
+/* Retrieve a record. */
+const char *cbmapget(const CBMAP *map, const char *kbuf, int ksiz, int *sp){
+ CBMAPDATUM *datum;
+ char *dbuf;
+ int hash, kcmp;
+ assert(map && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ CB_FIRSTHASH(hash, kbuf, ksiz);
+ datum = map->buckets[hash%map->bnum];
+ CB_SECONDHASH(hash, kbuf, ksiz);
+ while(datum){
+ if(hash > datum->hash){
+ datum = datum->left;
+ } else if(hash < datum->hash){
+ datum = datum->right;
+ } else {
+ dbuf = (char *)datum + sizeof(*datum);
+ kcmp = cbkeycmp(kbuf, ksiz, dbuf, datum->ksiz);
+ if(kcmp < 0){
+ datum = datum->left;
+ } else if(kcmp > 0){
+ datum = datum->right;
+ } else {
+ if(sp) *sp = datum->vsiz;
+ return dbuf + datum->ksiz + CB_ALIGNPAD(datum->ksiz);
+ }
+ }
+ }
+ return NULL;
+}
+
+
+/* Move a record to the edge. */
+int cbmapmove(CBMAP *map, const char *kbuf, int ksiz, int head){
+ CBMAPDATUM *datum;
+ char *dbuf;
+ int hash, kcmp;
+ assert(map && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ CB_FIRSTHASH(hash, kbuf, ksiz);
+ datum = map->buckets[hash%map->bnum];
+ CB_SECONDHASH(hash, kbuf, ksiz);
+ while(datum){
+ if(hash > datum->hash){
+ datum = datum->left;
+ } else if(hash < datum->hash){
+ datum = datum->right;
+ } else {
+ dbuf = (char *)datum + sizeof(*datum);
+ kcmp = cbkeycmp(kbuf, ksiz, dbuf, datum->ksiz);
+ if(kcmp < 0){
+ datum = datum->left;
+ } else if(kcmp > 0){
+ datum = datum->right;
+ } else {
+ if(head){
+ if(map->first == datum) return TRUE;
+ if(map->last == datum) map->last = datum->prev;
+ if(datum->prev) datum->prev->next = datum->next;
+ if(datum->next) datum->next->prev = datum->prev;
+ datum->prev = NULL;
+ datum->next = map->first;
+ map->first->prev = datum;
+ map->first = datum;
+ } else {
+ if(map->last == datum) return TRUE;
+ if(map->first == datum) map->first = datum->next;
+ if(datum->prev) datum->prev->next = datum->next;
+ if(datum->next) datum->next->prev = datum->prev;
+ datum->prev = map->last;
+ datum->next = NULL;
+ map->last->next = datum;
+ map->last = datum;
+ }
+ return TRUE;
+ }
+ }
+ }
+ return FALSE;
+}
+
+
+/* Initialize the iterator of a map handle. */
+void cbmapiterinit(CBMAP *map){
+ assert(map);
+ map->cur = map->first;
+}
+
+
+/* Get the next key of the iterator. */
+const char *cbmapiternext(CBMAP *map, int *sp){
+ CBMAPDATUM *datum;
+ assert(map);
+ if(!map->cur) return NULL;
+ datum = map->cur;
+ map->cur = datum->next;
+ if(sp) *sp = datum->ksiz;
+ return (char *)datum + sizeof(*datum);
+}
+
+
+/* Get the value binded to the key fetched from the iterator. */
+const char *cbmapiterval(const char *kbuf, int *sp){
+ CBMAPDATUM *datum;
+ assert(kbuf);
+ datum = (CBMAPDATUM *)(kbuf - sizeof(*datum));
+ if(sp) *sp = datum->vsiz;
+ return (char *)datum + sizeof(*datum) + datum->ksiz + CB_ALIGNPAD(datum->ksiz);
+}
+
+
+/* Get the number of the records stored in a map. */
+int cbmaprnum(const CBMAP *map){
+ assert(map);
+ return map->rnum;
+}
+
+
+/* Get the list handle contains all keys in a map. */
+CBLIST *cbmapkeys(CBMAP *map){
+ CBLIST *list;
+ const char *kbuf;
+ int anum, ksiz;
+ assert(map);
+ anum = cbmaprnum(map);
+ CB_LISTOPEN2(list, anum);
+ cbmapiterinit(map);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ CB_LISTPUSH(list, kbuf, ksiz);
+ }
+ return list;
+}
+
+
+/* Get the list handle contains all values in a map. */
+CBLIST *cbmapvals(CBMAP *map){
+ CBLIST *list;
+ const char *kbuf, *vbuf;
+ int anum, ksiz, vsiz;
+ assert(map);
+ anum = cbmaprnum(map);
+ CB_LISTOPEN2(list, anum);
+ cbmapiterinit(map);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ CB_MAPITERVAL(vbuf, kbuf, vsiz);
+ CB_LISTPUSH(list, vbuf, vsiz);
+ }
+ return list;
+}
+
+
+/* Serialize a map into a byte array. */
+char *cbmapdump(CBMAP *map, int *sp){
+ char *buf, vnumbuf[CB_VNUMBUFSIZ];
+ const char *kbuf, *vbuf;
+ int bsiz, vnumsiz, rn, ksiz, vsiz;
+ assert(map && sp);
+ rn = cbmaprnum(map);
+ CB_SETVNUMBUF(vnumsiz, vnumbuf, rn);
+ CB_MALLOC(buf, vnumsiz + 1);
+ memcpy(buf, vnumbuf, vnumsiz);
+ bsiz = vnumsiz;
+ cbmapiterinit(map);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ CB_MAPITERVAL(vbuf, kbuf, vsiz);
+ CB_SETVNUMBUF(vnumsiz, vnumbuf, ksiz);
+ CB_REALLOC(buf, bsiz + vnumsiz + ksiz + 1);
+ memcpy(buf + bsiz, vnumbuf, vnumsiz);
+ bsiz += vnumsiz;
+ memcpy(buf + bsiz, kbuf, ksiz);
+ bsiz += ksiz;
+ CB_SETVNUMBUF(vnumsiz, vnumbuf, vsiz);
+ CB_REALLOC(buf, bsiz + vnumsiz + vsiz + 1);
+ memcpy(buf + bsiz, vnumbuf, vnumsiz);
+ bsiz += vnumsiz;
+ memcpy(buf + bsiz, vbuf, vsiz);
+ bsiz += vsiz;
+ }
+ *sp = bsiz;
+ return buf;
+}
+
+
+/* Redintegrate a serialized map. */
+CBMAP *cbmapload(const char *ptr, int size){
+ CBMAP *map;
+ const char *rp, *kbuf, *vbuf;
+ int i, step, rn, ksiz, vsiz;
+ assert(ptr && size >= 0);
+ map = cbmapopenex(CB_MAPPBNUM);
+ rp = ptr;
+ CB_READVNUMBUF(rp, size, rn, step);
+ rp += step;
+ size -= step;
+ if(rn > size) return map;
+ for(i = 0; i < rn; i++){
+ if(size < 1) break;
+ CB_READVNUMBUF(rp, size, ksiz, step);
+ rp += step;
+ size -= step;
+ if(ksiz > size) break;
+ kbuf = rp;
+ rp += ksiz;
+ if(size < 1) break;
+ CB_READVNUMBUF(rp, size, vsiz, step);
+ rp += step;
+ size -= step;
+ if(vsiz > size) break;
+ vbuf = rp;
+ rp += vsiz;
+ cbmapput(map, kbuf, ksiz, vbuf, vsiz, TRUE);
+ }
+ return map;
+}
+
+
+/* Redintegrate a serialized map and get one of the records. */
+char *cbmaploadone(const char *ptr, int size, const char *kbuf, int ksiz, int *sp){
+ const char *rp, *tkbuf, *vbuf;
+ char *rv;
+ int i, step, rn, tksiz, vsiz;
+ assert(ptr && size >= 0 && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ rp = ptr;
+ CB_READVNUMBUF(rp, size, rn, step);
+ rp += step;
+ size -= step;
+ if(rn > size) return NULL;
+ for(i = 0; i < rn; i++){
+ if(size < 1) break;
+ CB_READVNUMBUF(rp, size, tksiz, step);
+ rp += step;
+ size -= step;
+ if(tksiz > size) break;
+ tkbuf = rp;
+ rp += tksiz;
+ if(size < 1) break;
+ CB_READVNUMBUF(rp, size, vsiz, step);
+ rp += step;
+ size -= step;
+ if(vsiz > size) break;
+ vbuf = rp;
+ rp += vsiz;
+ if(tksiz == ksiz && !memcmp(tkbuf, kbuf, ksiz)){
+ if(sp) *sp = vsiz;
+ CB_MEMDUP(rv, vbuf, vsiz);
+ return rv;
+ }
+ }
+ return NULL;
+}
+
+
+/* Get a heap handle. */
+CBHEAP *cbheapopen(int size, int max, int(*compar)(const void *, const void *)){
+ CBHEAP *heap;
+ assert(size > 0 && max >= 0 && compar);
+ CB_MALLOC(heap, sizeof(*heap));
+ CB_MALLOC(heap->base, size * max + 1);
+ CB_MALLOC(heap->swap, size);
+ heap->size = size;
+ heap->num = 0;
+ heap->max = max;
+ heap->compar = compar;
+ return heap;
+}
+
+
+/* Copy a heap. */
+CBHEAP *cbheapdup(CBHEAP *heap){
+ CBHEAP *newheap;
+ assert(heap);
+ CB_MALLOC(newheap, sizeof(*newheap));
+ CB_MEMDUP(newheap->base, heap->base, heap->size * heap->max);
+ CB_MALLOC(newheap->swap, heap->size);
+ newheap->size = heap->size;
+ newheap->num = heap->num;
+ newheap->max = heap->max;
+ newheap->compar = heap->compar;
+ return newheap;
+}
+
+
+/* Close a heap handle. */
+void cbheapclose(CBHEAP *heap){
+ assert(heap);
+ free(heap->swap);
+ free(heap->base);
+ free(heap);
+}
+
+
+/* Get the number of the records stored in a heap. */
+int cbheapnum(CBHEAP *heap){
+ assert(heap);
+ return heap->num;
+}
+
+
+/* Insert a record into a heap. */
+int cbheapinsert(CBHEAP *heap, const void *ptr){
+ char *base;
+ int size, pidx, cidx, bot;
+ assert(heap && ptr);
+ if(heap->max < 1) return FALSE;
+ base = heap->base;
+ size = heap->size;
+ if(heap->num >= heap->max){
+ if(heap->compar(ptr, base) > 0) return FALSE;
+ memcpy(base, ptr, size);
+ pidx = 0;
+ bot = heap->num / 2;
+ while(pidx < bot){
+ cidx = pidx * 2 + 1;
+ if(cidx < heap->num - 1 && heap->compar(base + cidx * size, base + (cidx + 1) * size) < 0)
+ cidx++;
+ if(heap->compar(base + pidx * size, base + cidx * size) > 0) break;
+ memcpy(heap->swap, base + pidx * size, size);
+ memcpy(base + pidx * size, base + cidx * size, size);
+ memcpy(base + cidx * size, heap->swap, size);
+ pidx = cidx;
+ }
+ } else {
+ memcpy(base + size * heap->num, ptr, size);
+ cidx = heap->num;
+ while(cidx > 0){
+ pidx = (cidx - 1) / 2;
+ if(heap->compar(base + cidx * size, base + pidx * size) <= 0) break;
+ memcpy(heap->swap, base + cidx * size, size);
+ memcpy(base + cidx * size, base + pidx * size, size);
+ memcpy(base + pidx * size, heap->swap, size);
+ cidx = pidx;
+ }
+ heap->num++;
+ }
+ return TRUE;
+}
+
+
+/* Get the pointer to the region of a record in a heap. */
+const void *cbheapval(CBHEAP *heap, int index){
+ assert(heap && index >= 0);
+ if(index >= heap->num) return NULL;
+ return heap->base + index * heap->size;
+}
+
+
+/* Convert a heap to an allocated region. */
+void *cbheaptomalloc(CBHEAP *heap, int *np){
+ char *ptr;
+ assert(heap);
+ qsort(heap->base, heap->num, heap->size, heap->compar);
+ ptr = heap->base;
+ if(np) *np = heap->num;
+ free(heap->swap);
+ free(heap);
+ return ptr;
+}
+
+
+/* Allocate a formatted string on memory. */
+char *cbsprintf(const char *format, ...){
+ va_list ap;
+ char *buf, cbuf[CB_SPBUFSIZ], *str;
+ int len, cblen, num, slen;
+ unsigned int unum;
+ double dnum;
+ va_start(ap, format);
+ assert(format);
+ CB_MALLOC(buf, 1);
+ len = 0;
+ while(*format != '\0'){
+ if(*format == '%'){
+ cbuf[0] = '%';
+ cblen = 1;
+ format++;
+ while(strchr("0123456789 .+-", *format) && *format != '\0' && cblen < CB_SPBUFSIZ - 1){
+ cbuf[cblen++] = *format;
+ format++;
+ }
+ cbuf[cblen] = '\0';
+ if(atoi(cbuf + 1) > CB_SPMAXWIDTH - 16){
+ sprintf(cbuf, "(err)");
+ } else {
+ cbuf[cblen++] = *format;
+ cbuf[cblen] = '\0';
+ }
+ switch(*format){
+ case 'd':
+ num = va_arg(ap, int);
+ CB_REALLOC(buf, len + CB_SPMAXWIDTH + 2);
+ len += sprintf(buf + len, cbuf, num);
+ break;
+ case 'o': case 'u': case 'x': case 'X': case 'c':
+ unum = va_arg(ap, unsigned int);
+ CB_REALLOC(buf, len + CB_SPMAXWIDTH + 2);
+ len += sprintf(buf + len, cbuf, unum);
+ break;
+ case 'e': case 'E': case 'f': case 'g': case 'G':
+ dnum = va_arg(ap, double);
+ CB_REALLOC(buf, len + CB_SPMAXWIDTH + 2);
+ len += sprintf(buf + len, cbuf, dnum);
+ break;
+ case 's':
+ str = va_arg(ap, char *);
+ slen = strlen(str);
+ CB_REALLOC(buf, len + slen + 2);
+ memcpy(buf + len, str, slen);
+ len += slen;
+ break;
+ case '%':
+ CB_REALLOC(buf, len + 2);
+ buf[len++] = '%';
+ break;
+ default:
+ break;
+ }
+ } else {
+ CB_REALLOC(buf, len + 2);
+ buf[len++] = *format;
+ }
+ format++;
+ }
+ buf[len] = '\0';
+ va_end(ap);
+ return buf;
+}
+
+
+/* Replace some patterns in a string. */
+char *cbreplace(const char *str, CBMAP *pairs){
+ int i, bsiz, wi, rep, ksiz, vsiz;
+ char *buf;
+ const char *key, *val;
+ assert(str && pairs);
+ bsiz = CB_DATUMUNIT;
+ CB_MALLOC(buf, bsiz);
+ wi = 0;
+ while(*str != '\0'){
+ rep = FALSE;
+ cbmapiterinit(pairs);
+ while((key = cbmapiternext(pairs, &ksiz)) != NULL){
+ for(i = 0; i < ksiz; i++){
+ if(str[i] == '\0' || str[i] != key[i]) break;
+ }
+ if(i == ksiz){
+ CB_MAPITERVAL(val, key, vsiz);
+ if(wi + vsiz >= bsiz){
+ bsiz = bsiz * 2 + vsiz;
+ CB_REALLOC(buf, bsiz);
+ }
+ memcpy(buf + wi, val, vsiz);
+ wi += vsiz;
+ str += ksiz;
+ rep = TRUE;
+ break;
+ }
+ }
+ if(!rep){
+ if(wi + 1 >= bsiz){
+ bsiz = bsiz * 2 + 1;
+ CB_REALLOC(buf, bsiz);
+ }
+ buf[wi++] = *str;
+ str++;
+ }
+ }
+ CB_REALLOC(buf, wi + 1);
+ buf[wi] = '\0';
+ return buf;
+}
+
+
+/* Make a list by split a serial datum. */
+CBLIST *cbsplit(const char *ptr, int size, const char *delim){
+ CBLIST *list;
+ int bi, step;
+ assert(ptr);
+ CB_LISTOPEN(list);
+ if(size < 0) size = strlen(ptr);
+ if(delim){
+ for(bi = 0; bi < size; bi += step){
+ step = 0;
+ while(bi + step < size && !strchr(delim, ptr[bi+step])){
+ step++;
+ }
+ cblistpush(list, ptr + bi, step);
+ step++;
+ }
+ if(size > 0 && strchr(delim, ptr[size-1])) cblistpush(list, "", 0);
+ } else {
+ for(bi = 0; bi < size; bi += step){
+ step = 0;
+ while(bi + step < size && ptr[bi+step]){
+ step++;
+ }
+ cblistpush(list, ptr + bi, step);
+ step++;
+ }
+ if(size > 0 && ptr[size-1] == 0) cblistpush(list, "", 0);
+ }
+ return list;
+}
+
+
+/* Read whole data of a file. */
+char *cbreadfile(const char *name, int *sp){
+ struct stat sbuf;
+ char iobuf[CB_IOBUFSIZ], *buf;
+ int fd, size, asiz, rv;
+ asiz = CB_IOBUFSIZ * 2;
+ if(name){
+ if((fd = open(name, O_RDONLY, 0)) == -1) return NULL;
+ if(fstat(fd, &sbuf) != -1) asiz = sbuf.st_size + 1;
+ } else {
+ fd = 0;
+ }
+ CB_MALLOC(buf, asiz + 1);
+ size = 0;
+ while((rv = read(fd, iobuf, CB_IOBUFSIZ)) > 0){
+ if(size + rv >= asiz){
+ asiz = asiz * 2 + size;
+ CB_REALLOC(buf, asiz + 1);
+ }
+ memcpy(buf + size, iobuf, rv);
+ size += rv;
+ }
+ buf[size] = '\0';
+ if(close(fd) == -1 || rv == -1){
+ free(buf);
+ return NULL;
+ }
+ if(sp) *sp = size;
+ return buf;
+}
+
+
+/* Write data of a region into a file. */
+int cbwritefile(const char *name, const char *ptr, int size){
+ int fd, err, wb;
+ assert(ptr);
+ if(size < 0) size = strlen(ptr);
+ if(name){
+ if((fd = open(name, O_WRONLY | O_CREAT | O_TRUNC, CB_FILEMODE)) == -1) return FALSE;
+ } else {
+ fd = 1;
+ }
+ err = FALSE;
+ wb = 0;
+ do {
+ wb = write(fd, ptr, size);
+ switch(wb){
+ case -1: err = errno != EINTR ? TRUE : FALSE; break;
+ case 0: break;
+ default:
+ ptr += wb;
+ size -= wb;
+ break;
+ }
+ } while(size > 0);
+ if(close(fd) == -1) err = TRUE;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Read every line of a file. */
+CBLIST *cbreadlines(const char *name){
+ char *buf, *tmp;
+ int vsiz;
+ CBMAP *pairs;
+ CBLIST *list;
+ if(!(buf = cbreadfile(name, NULL))) return NULL;
+ pairs = cbmapopenex(3);
+ cbmapput(pairs, "\r\n", 2, "\n", 1, TRUE);
+ cbmapput(pairs, "\r", 1, "\n", 1, TRUE);
+ tmp = cbreplace(buf, pairs);
+ list = cbsplit(tmp, strlen(tmp), "\n");
+ free(tmp);
+ cbmapclose(pairs);
+ free(buf);
+ if(CB_LISTNUM(list) > 0){
+ cblistval(list, CB_LISTNUM(list) - 1, &vsiz);
+ if(vsiz < 1) CB_LISTDROP(list);
+ }
+ return list;
+}
+
+
+/* Read names of files in a directory. */
+CBLIST *cbdirlist(const char *name){
+ DIR *DD;
+ struct dirent *dp;
+ CBLIST *list;
+ assert(name);
+ if(!(DD = opendir(name))) return NULL;
+ CB_LISTOPEN(list);
+ while((dp = readdir(DD)) != NULL){
+ CB_LISTPUSH(list, dp->d_name, strlen(dp->d_name));
+ }
+ if(closedir(DD) == -1){
+ CB_LISTCLOSE(list);
+ return NULL;
+ }
+ return list;
+}
+
+
+/* Get the status of a file or a directory. */
+int cbfilestat(const char *name, int *isdirp, int *sizep, time_t *mtimep){
+ struct stat sbuf;
+ assert(name);
+ if(lstat(name, &sbuf) == -1) return FALSE;
+ if(isdirp) *isdirp = S_ISDIR(sbuf.st_mode);
+ if(sizep) *sizep = (int)sbuf.st_size;
+ if(mtimep) *mtimep = sbuf.st_mtime;
+ return TRUE;
+}
+
+
+/* Remove a file or a directory and its sub ones recursively. */
+int cbremove(const char *name){
+ CBLIST *list;
+ const char *elem;
+ char *path;
+ int i, err, tail;
+ struct stat sbuf;
+ assert(name);
+ if(lstat(name, &sbuf) == -1) return FALSE;
+ if(unlink(name) == 0) return TRUE;
+ if(!S_ISDIR(sbuf.st_mode) || !(list = cbdirlist(name))) return FALSE;
+ err = FALSE;
+ tail = name[0] != '\0' && name[strlen(name)-1] == MYPATHCHR;
+ for(i = 0; i < CB_LISTNUM(list); i++){
+ elem = CB_LISTVAL(list, i);
+ if(!strcmp(MYCDIRSTR, elem) || !strcmp(MYPDIRSTR, elem)) continue;
+ if(tail){
+ path = cbsprintf("%s%s", name, elem);
+ } else {
+ path = cbsprintf("%s%c%s", name, MYPATHCHR, elem);
+ }
+ if(!cbremove(path)) err = TRUE;
+ free(path);
+ }
+ CB_LISTCLOSE(list);
+ return rmdir(name) == 0 ? TRUE : FALSE;
+}
+
+
+/* Break up a URL into elements. */
+CBMAP *cburlbreak(const char *str){
+ CBMAP *map;
+ char *tmp, *ep;
+ const char *rp;
+ int serv;
+ assert(str);
+ map = cbmapopenex(CB_MAPPBNUM);
+ CB_MEMDUP(tmp, str, strlen(str));
+ rp = cbstrtrim(tmp);
+ cbmapput(map, "self", -1, rp, -1, TRUE);
+ serv = FALSE;
+ if(cbstrfwimatch(rp, "http://")){
+ cbmapput(map, "scheme", -1, "http", -1, TRUE);
+ rp += 7;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "https://")){
+ cbmapput(map, "scheme", -1, "https", -1, TRUE);
+ rp += 8;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "ftp://")){
+ cbmapput(map, "scheme", -1, "ftp", -1, TRUE);
+ rp += 6;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "sftp://")){
+ cbmapput(map, "scheme", -1, "sftp", -1, TRUE);
+ rp += 7;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "ftps://")){
+ cbmapput(map, "scheme", -1, "ftps", -1, TRUE);
+ rp += 7;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "tftp://")){
+ cbmapput(map, "scheme", -1, "tftp", -1, TRUE);
+ rp += 7;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "ldap://")){
+ cbmapput(map, "scheme", -1, "ldap", -1, TRUE);
+ rp += 7;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "ldaps://")){
+ cbmapput(map, "scheme", -1, "ldaps", -1, TRUE);
+ rp += 8;
+ serv = TRUE;
+ } else if(cbstrfwimatch(rp, "file://")){
+ cbmapput(map, "scheme", -1, "file", -1, TRUE);
+ rp += 7;
+ serv = TRUE;
+ }
+ if((ep = strchr(rp, '#')) != NULL){
+ cbmapput(map, "fragment", -1, ep + 1, -1, TRUE);
+ *ep = '\0';
+ }
+ if((ep = strchr(rp, '?')) != NULL){
+ cbmapput(map, "query", -1, ep + 1, -1, TRUE);
+ *ep = '\0';
+ }
+ if(serv){
+ if((ep = strchr(rp, '/')) != NULL){
+ cbmapput(map, "path", -1, ep, -1, TRUE);
+ *ep = '\0';
+ } else {
+ cbmapput(map, "path", -1, "/", -1, TRUE);
+ }
+ if((ep = strchr(rp, '@')) != NULL){
+ *ep = '\0';
+ if(rp[0] != '\0') cbmapput(map, "authority", -1, rp, -1, TRUE);
+ rp = ep + 1;
+ }
+ if((ep = strchr(rp, ':')) != NULL){
+ if(ep[1] != '\0') cbmapput(map, "port", -1, ep + 1, -1, TRUE);
+ *ep = '\0';
+ }
+ if(rp[0] != '\0') cbmapput(map, "host", -1, rp, -1, TRUE);
+ } else {
+ cbmapput(map, "path", -1, rp, -1, TRUE);
+ }
+ free(tmp);
+ if((rp = cbmapget(map, "path", -1, NULL)) != NULL){
+ if((ep = strrchr(rp, '/')) != NULL){
+ if(ep[1] != '\0') cbmapput(map, "file", -1, ep + 1, -1, TRUE);
+ } else {
+ cbmapput(map, "file", -1, rp, -1, TRUE);
+ }
+ }
+ if((rp = cbmapget(map, "file", -1, NULL)) != NULL && (!strcmp(rp, ".") || !strcmp(rp, "..")))
+ cbmapout(map, "file", -1);
+ return map;
+}
+
+
+/* Resolve a relative URL with another absolute URL. */
+char *cburlresolve(const char *base, const char *target){
+ CBMAP *telems, *belems;
+ CBLIST *bpaths, *opaths, *qelems;
+ CBDATUM *rbuf;
+ const char *vbuf, *path;
+ char *tmp, *wp, *enc, numbuf[CB_NUMBUFSIZ];
+ int i, vsiz, port, num;
+ assert(base && target);
+ while(*base > '\0' && *base <= ' '){
+ base++;
+ }
+ while(*target > '\0' && *target <= ' '){
+ target++;
+ }
+ if(*target == '\0') target = base;
+ CB_DATUMOPEN(rbuf);
+ telems = cburlbreak(target);
+ port = 80;
+ belems = cburlbreak(cbmapget(telems, "scheme", -1, &vsiz) ? target : base);
+ if((vbuf = cbmapget(belems, "scheme", -1, &vsiz)) != NULL){
+ CB_DATUMCAT(rbuf, vbuf, vsiz);
+ CB_DATUMCAT(rbuf, "://", 3);
+ if(!cbstricmp(vbuf, "https")){
+ port = 443;
+ } else if(!cbstricmp(vbuf, "ftp")){
+ port = 21;
+ } else if(!cbstricmp(vbuf, "sftp")){
+ port = 115;
+ } else if(!cbstricmp(vbuf, "ftps")){
+ port = 22;
+ } else if(!cbstricmp(vbuf, "tftp")){
+ port = 69;
+ } else if(!cbstricmp(vbuf, "ldap")){
+ port = 389;
+ } else if(!cbstricmp(vbuf, "ldaps")){
+ port = 636;
+ }
+ } else {
+ CB_DATUMCAT(rbuf, "http://", 7);
+ }
+ if((vbuf = cbmapget(belems, "authority", -1, &vsiz)) != NULL){
+ if((wp = strchr(vbuf, ':')) != NULL){
+ *wp = '\0';
+ tmp = cburldecode(vbuf, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ CB_DATUMCAT(rbuf, ":", 1);
+ wp++;
+ tmp = cburldecode(wp, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ } else {
+ tmp = cburldecode(vbuf, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ }
+ CB_DATUMCAT(rbuf, "@", 1);
+ }
+ if((vbuf = cbmapget(belems, "host", -1, &vsiz)) != NULL){
+ tmp = cburldecode(vbuf, NULL);
+ cbstrtolower(tmp);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ } else {
+ CB_DATUMCAT(rbuf, "localhost", 9);
+ }
+ if((vbuf = cbmapget(belems, "port", -1, &vsiz)) != NULL &&
+ (num = atoi(vbuf)) != port && num > 1){
+ sprintf(numbuf, ":%d", num);
+ CB_DATUMCAT(rbuf, numbuf, strlen(numbuf));
+ }
+ if(!(path = cbmapget(telems, "path", -1, NULL))) path = "/";
+ if(path[0] == '\0' && (vbuf = cbmapget(belems, "path", -1, NULL)) != NULL) path = vbuf;
+ if(path[0] == '\0') path = "/";
+ CB_LISTOPEN(bpaths);
+ if(path[0] != '/' && (vbuf = cbmapget(belems, "path", -1, &vsiz)) != NULL){
+ opaths = cbsplit(vbuf, vsiz, "/");
+ } else {
+ opaths = cbsplit("/", 1, "/");
+ }
+ CB_LISTDROP(opaths);
+ for(i = 0; i < CB_LISTNUM(opaths); i++){
+ vbuf = CB_LISTVAL2(opaths, i, vsiz);
+ if(vsiz < 1 || !strcmp(vbuf, ".")) continue;
+ if(!strcmp(vbuf, "..")){
+ CB_LISTDROP(bpaths);
+ } else {
+ CB_LISTPUSH(bpaths, vbuf, vsiz);
+ }
+ }
+ CB_LISTCLOSE(opaths);
+ opaths = cbsplit(path, -1, "/");
+ for(i = 0; i < CB_LISTNUM(opaths); i++){
+ vbuf = CB_LISTVAL2(opaths, i, vsiz);
+ if(vsiz < 1 || !strcmp(vbuf, ".")) continue;
+ if(!strcmp(vbuf, "..")){
+ CB_LISTDROP(bpaths);
+ } else {
+ CB_LISTPUSH(bpaths, vbuf, vsiz);
+ }
+ }
+ CB_LISTCLOSE(opaths);
+ for(i = 0; i < CB_LISTNUM(bpaths); i++){
+ vbuf = CB_LISTVAL(bpaths, i);
+ if(strchr(vbuf, '%')){
+ tmp = cburldecode(vbuf, NULL);
+ } else {
+ CB_MEMDUP(tmp, vbuf, strlen(vbuf));
+ }
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, "/", 1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ }
+ if(cbstrbwmatch(path, "/")) CB_DATUMCAT(rbuf, "/", 1);
+ CB_LISTCLOSE(bpaths);
+ if((vbuf = cbmapget(telems, "query", -1, &vsiz)) != NULL){
+ CB_DATUMCAT(rbuf, "?", 1);
+ qelems = cbsplit(vbuf, vsiz, "&;");
+ for(i = 0; i < CB_LISTNUM(qelems); i++){
+ vbuf = CB_LISTVAL(qelems, i);
+ if(i > 0) CB_DATUMCAT(rbuf, "&", 1);
+ if((wp = strchr(vbuf, '=')) != NULL){
+ *wp = '\0';
+ tmp = cburldecode(vbuf, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ CB_DATUMCAT(rbuf, "=", 1);
+ wp++;
+ tmp = cburldecode(wp, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ } else {
+ tmp = cburldecode(vbuf, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ }
+ }
+ CB_LISTCLOSE(qelems);
+ }
+ if((vbuf = cbmapget(telems, "fragment", -1, &vsiz)) != NULL){
+ tmp = cburldecode(vbuf, NULL);
+ enc = cburlencode(tmp, -1);
+ CB_DATUMCAT(rbuf, "#", 1);
+ CB_DATUMCAT(rbuf, enc, strlen(enc));
+ free(enc);
+ free(tmp);
+ }
+ cbmapclose(belems);
+ cbmapclose(telems);
+ return cbdatumtomalloc(rbuf, NULL);
+}
+
+
+/* Encode a serial object with URL encoding. */
+char *cburlencode(const char *ptr, int size){
+ char *buf, *wp;
+ int i, c;
+ assert(ptr);
+ if(size < 0) size = strlen(ptr);
+ CB_MALLOC(buf, size * 3 + 1);
+ wp = buf;
+ for(i = 0; i < size; i++){
+ c = ((unsigned char *)ptr)[i];
+ if((c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z') ||
+ (c >= '0' && c <= '9') || (c != '\0' && strchr("_-.!~*'()", c))){
+ *(wp++) = c;
+ } else {
+ wp += sprintf(wp, "%%%02X", c);
+ }
+ }
+ *wp = '\0';
+ return buf;
+}
+
+
+/* Decode a string encoded with URL encoding. */
+char *cburldecode(const char *str, int *sp){
+ char *buf, *wp;
+ unsigned char c;
+ CB_MEMDUP(buf, str, strlen(str));
+ wp = buf;
+ while(*str != '\0'){
+ if(*str == '%'){
+ str++;
+ if(((str[0] >= '0' && str[0] <= '9') || (str[0] >= 'A' && str[0] <= 'F') ||
+ (str[0] >= 'a' && str[0] <= 'f')) &&
+ ((str[1] >= '0' && str[1] <= '9') || (str[1] >= 'A' && str[1] <= 'F') ||
+ (str[1] >= 'a' && str[1] <= 'f'))){
+ c = *str;
+ if(c >= 'A' && c <= 'Z') c += 'a' - 'A';
+ if(c >= 'a' && c <= 'z'){
+ *wp = c - 'a' + 10;
+ } else {
+ *wp = c - '0';
+ }
+ *wp *= 0x10;
+ str++;
+ c = *str;
+ if(c >= 'A' && c <= 'Z') c += 'a' - 'A';
+ if(c >= 'a' && c <= 'z'){
+ *wp += c - 'a' + 10;
+ } else {
+ *wp += c - '0';
+ }
+ str++;
+ wp++;
+ } else {
+ break;
+ }
+ } else if(*str == '+'){
+ *wp = ' ';
+ str++;
+ wp++;
+ } else {
+ *wp = *str;
+ str++;
+ wp++;
+ }
+ }
+ *wp = '\0';
+ if(sp) *sp = wp - buf;
+ return buf;
+}
+
+
+/* Encode a serial object with Base64 encoding. */
+char *cbbaseencode(const char *ptr, int size){
+ char *tbl = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
+ char *buf, *wp;
+ const unsigned char *obj;
+ int i;
+ assert(ptr);
+ if(size < 0) size = strlen(ptr);
+ CB_MALLOC(buf, 4 * (size + 2) / 3 + 1);
+ obj = (const unsigned char *)ptr;
+ wp = buf;
+ for(i = 0; i < size; i += 3){
+ switch(size - i){
+ case 1:
+ *wp++ = tbl[obj[0] >> 2];
+ *wp++ = tbl[(obj[0] & 3) << 4];
+ *wp++ = '=';
+ *wp++ = '=';
+ break;
+ case 2:
+ *wp++ = tbl[obj[0] >> 2];
+ *wp++ = tbl[((obj[0] & 3) << 4) + (obj[1] >> 4)];
+ *wp++ = tbl[(obj[1] & 0xf) << 2];
+ *wp++ = '=';
+ break;
+ default:
+ *wp++ = tbl[obj[0] >> 2];
+ *wp++ = tbl[((obj[0] & 3) << 4) + (obj[1] >> 4)];
+ *wp++ = tbl[((obj[1] & 0xf) << 2) + (obj[2] >> 6)];
+ *wp++ = tbl[obj[2] & 0x3f];
+ break;
+ }
+ obj += 3;
+ }
+ *wp = '\0';
+ return buf;
+}
+
+
+/* Decode a string encoded with Base64 encoding. */
+char *cbbasedecode(const char *str, int *sp){
+ unsigned char *obj, *wp;
+ int len, cnt, bpos, i, bits, eqcnt;
+ assert(str);
+ cnt = 0;
+ bpos = 0;
+ eqcnt = 0;
+ len = strlen(str);
+ CB_MALLOC(obj, len + 4);
+ wp = obj;
+ while(bpos < len && eqcnt == 0){
+ bits = 0;
+ for(i = 0; bpos < len && i < 4; bpos++){
+ if(str[bpos] >= 'A' && str[bpos] <= 'Z'){
+ bits = (bits << 6) | (str[bpos] - 'A');
+ i++;
+ } else if(str[bpos] >= 'a' && str[bpos] <= 'z'){
+ bits = (bits << 6) | (str[bpos] - 'a' + 26);
+ i++;
+ } else if(str[bpos] >= '0' && str[bpos] <= '9'){
+ bits = (bits << 6) | (str[bpos] - '0' + 52);
+ i++;
+ } else if(str[bpos] == '+'){
+ bits = (bits << 6) | 62;
+ i++;
+ } else if(str[bpos] == '/'){
+ bits = (bits << 6) | 63;
+ i++;
+ } else if(str[bpos] == '='){
+ bits <<= 6;
+ i++;
+ eqcnt++;
+ }
+ }
+ if(i == 0 && bpos >= len) continue;
+ switch(eqcnt){
+ case 0:
+ *wp++ = (bits >> 16) & 0xff;
+ *wp++ = (bits >> 8) & 0xff;
+ *wp++ = bits & 0xff;
+ cnt += 3;
+ break;
+ case 1:
+ *wp++ = (bits >> 16) & 0xff;
+ *wp++ = (bits >> 8) & 0xff;
+ cnt += 2;
+ break;
+ case 2:
+ *wp++ = (bits >> 16) & 0xff;
+ cnt += 1;
+ break;
+ }
+ }
+ obj[cnt] = '\0';
+ if(sp) *sp = cnt;
+ return (char *)obj;
+}
+
+
+/* Encode a serial object with quoted-printable encoding. */
+char *cbquoteencode(const char *ptr, int size){
+ const unsigned char *rp;
+ char *buf, *wp;
+ int i, cols;
+ assert(ptr);
+ if(size < 0) size = strlen(ptr);
+ rp = (const unsigned char *)ptr;
+ CB_MALLOC(buf, size * 3 + 1);
+ wp = buf;
+ cols = 0;
+ for(i = 0; i < size; i++){
+ if(rp[i] == '=' || (rp[i] < 0x20 && rp[i] != '\r' && rp[i] != '\n' && rp[i] != '\t') ||
+ rp[i] > 0x7e){
+ wp += sprintf(wp, "=%02X", rp[i]);
+ cols += 3;
+ } else {
+ *(wp++) = rp[i];
+ cols++;
+ }
+ }
+ *wp = '\0';
+ return buf;
+}
+
+
+/* Decode a string encoded with quoted-printable encoding. */
+char *cbquotedecode(const char *str, int *sp){
+ char *buf, *wp;
+ assert(str);
+ CB_MALLOC(buf, strlen(str) + 1);
+ wp = buf;
+ for(; *str != '\0'; str++){
+ if(*str == '='){
+ str++;
+ if(*str == '\0'){
+ break;
+ } else if(str[0] == '\r' && str[1] == '\n'){
+ str++;
+ } else if(str[0] != '\n' && str[0] != '\r'){
+ if(*str >= 'A' && *str <= 'Z'){
+ *wp = (*str - 'A' + 10) * 16;
+ } else if(*str >= 'a' && *str <= 'z'){
+ *wp = (*str - 'a' + 10) * 16;
+ } else {
+ *wp = (*str - '0') * 16;
+ }
+ str++;
+ if(*str == '\0') break;
+ if(*str >= 'A' && *str <= 'Z'){
+ *wp += *str - 'A' + 10;
+ } else if(*str >= 'a' && *str <= 'z'){
+ *wp += *str - 'a' + 10;
+ } else {
+ *wp += *str - '0';
+ }
+ wp++;
+ }
+ } else {
+ *wp = *str;
+ wp++;
+ }
+ }
+ *wp = '\0';
+ if(sp) *sp = wp - buf;
+ return buf;
+}
+
+
+/* Split a string of MIME into headers and the body. */
+char *cbmimebreak(const char *ptr, int size, CBMAP *attrs, int *sp){
+ CBLIST *list;
+ const char *head, *line, *pv, *ep;
+ char *hbuf, *name, *rv;
+ int i, j, wi, hlen;
+ assert(ptr);
+ if(size < 0) size = strlen(ptr);
+ head = NULL;
+ hlen = 0;
+ for(i = 0; i < size; i++){
+ if(i < size - 4 && ptr[i] == '\r' && ptr[i+1] == '\n' &&
+ ptr[i+2] == '\r' && ptr[i+3] == '\n'){
+ head = ptr;
+ hlen = i;
+ ptr += i + 4;
+ size -= i + 4;
+ break;
+ } else if(i < size - 2 && ptr[i] == '\n' && ptr[i+1] == '\n'){
+ head = ptr;
+ hlen = i;
+ ptr += i + 2;
+ size -= i + 2;
+ break;
+ }
+ }
+ if(head && attrs){
+ CB_MALLOC(hbuf, hlen + 1);
+ wi = 0;
+ for(i = 0; i < hlen; i++){
+ if(head[i] == '\r') continue;
+ if(i < hlen - 1 && head[i] == '\n' && (head[i+1] == ' ' || head[i+1] == '\t')){
+ hbuf[wi++] = ' ';
+ i++;
+ } else {
+ hbuf[wi++] = head[i];
+ }
+ }
+ list = cbsplit(hbuf, wi, "\n");
+ for(i = 0; i < CB_LISTNUM(list); i++){
+ line = CB_LISTVAL(list, i);
+ if((pv = strchr(line, ':')) != NULL){
+ CB_MEMDUP(name, line, pv - line);
+ for(j = 0; name[j] != '\0'; j++){
+ if(name[j] >= 'A' && name[j] <= 'Z') name[j] -= 'A' - 'a';
+ }
+ pv++;
+ while(*pv == ' ' || *pv == '\t'){
+ pv++;
+ }
+ cbmapput(attrs, name, -1, pv, -1, TRUE);
+ free(name);
+ }
+
+ }
+ CB_LISTCLOSE(list);
+ free(hbuf);
+ if((pv = cbmapget(attrs, "content-type", -1, NULL)) != NULL){
+ if((ep = strchr(pv, ';')) != NULL){
+ cbmapput(attrs, "TYPE", -1, pv, ep - pv, TRUE);
+ do {
+ ep++;
+ while(ep[0] == ' '){
+ ep++;
+ }
+ if(cbstrfwimatch(ep, "charset=")){
+ ep += 8;
+ while(*ep > '\0' && *ep <= ' '){
+ ep++;
+ }
+ if(ep[0] == '"') ep++;
+ pv = ep;
+ while(ep[0] != '\0' && ep[0] != ' ' && ep[0] != '"' && ep[0] != ';'){
+ ep++;
+ }
+ cbmapput(attrs, "CHARSET", -1, pv, ep - pv, TRUE);
+ } else if(cbstrfwimatch(ep, "boundary=")){
+ ep += 9;
+ while(*ep > '\0' && *ep <= ' '){
+ ep++;
+ }
+ if(ep[0] == '"'){
+ ep++;
+ pv = ep;
+ while(ep[0] != '\0' && ep[0] != '"'){
+ ep++;
+ }
+ } else {
+ pv = ep;
+ while(ep[0] != '\0' && ep[0] != ' ' && ep[0] != '"' && ep[0] != ';'){
+ ep++;
+ }
+ }
+ cbmapput(attrs, "BOUNDARY", -1, pv, ep - pv, TRUE);
+ }
+ } while((ep = strchr(ep, ';')) != NULL);
+ } else {
+ cbmapput(attrs, "TYPE", -1, pv, -1, TRUE);
+ }
+ }
+ if((pv = cbmapget(attrs, "content-disposition", -1, NULL)) != NULL){
+ if((ep = strchr(pv, ';')) != NULL){
+ cbmapput(attrs, "DISPOSITION", -1, pv, ep - pv, TRUE);
+ do {
+ ep++;
+ while(ep[0] == ' '){
+ ep++;
+ }
+ if(cbstrfwimatch(ep, "filename=")){
+ ep += 9;
+ if(ep[0] == '"') ep++;
+ pv = ep;
+ while(ep[0] != '\0' && ep[0] != '"'){
+ ep++;
+ }
+ cbmapput(attrs, "FILENAME", -1, pv, ep - pv, TRUE);
+ } else if(cbstrfwimatch(ep, "name=")){
+ ep += 5;
+ if(ep[0] == '"') ep++;
+ pv = ep;
+ while(ep[0] != '\0' && ep[0] != '"'){
+ ep++;
+ }
+ cbmapput(attrs, "NAME", -1, pv, ep - pv, TRUE);
+ }
+ } while((ep = strchr(ep, ';')) != NULL);
+ } else {
+ cbmapput(attrs, "DISPOSITION", -1, pv, -1, TRUE);
+ }
+ }
+ }
+ if(sp) *sp = size;
+ CB_MEMDUP(rv, ptr, size);
+ return rv;
+}
+
+
+/* Split multipart data in MIME into its parts. */
+CBLIST *cbmimeparts(const char *ptr, int size, const char *boundary){
+ CBLIST *list;
+ const char *pv, *ep;
+ int i, blen;
+ assert(ptr && boundary);
+ if(size < 0) size = strlen(ptr);
+ CB_LISTOPEN(list);
+ if((blen = strlen(boundary)) < 1) return list;
+ pv = NULL;
+ for(i = 0; i < size; i++){
+ if(ptr[i] == '-' && ptr[i+1] == '-' && i + 2 + blen < size &&
+ cbstrfwmatch(ptr + i + 2, boundary) && strchr("\t\n\v\f\r ", ptr[i+2+blen])){
+ pv = ptr + i + 2 + blen;
+ if(*pv == '\r') pv++;
+ if(*pv == '\n') pv++;
+ size -= pv - ptr;
+ ptr = pv;
+ break;
+ }
+ }
+ if(!pv) return list;
+ for(i = 0; i < size; i++){
+ if(ptr[i] == '-' && ptr[i+1] == '-' && i + 2 + blen < size &&
+ cbstrfwmatch(ptr + i + 2, boundary) && strchr("\t\n\v\f\r -", ptr[i+2+blen])){
+ ep = ptr + i;
+ if(ep > ptr && ep[-1] == '\n') ep--;
+ if(ep > ptr && ep[-1] == '\r') ep--;
+ if(ep > pv) CB_LISTPUSH(list, pv, ep - pv);
+ pv = ptr + i + 2 + blen;
+ if(*pv == '\r') pv++;
+ if(*pv == '\n') pv++;
+ }
+ }
+ return list;
+}
+
+
+/* Encode a string with MIME encoding. */
+char *cbmimeencode(const char *str, const char *encname, int base){
+ char *buf, *wp, *enc;
+ int len;
+ assert(str && encname);
+ len = strlen(str);
+ CB_MALLOC(buf, len * 3 + strlen(encname) + 16);
+ wp = buf;
+ wp += sprintf(wp, "=?%s?%c?", encname, base ? 'B' : 'Q');
+ enc = base ? cbbaseencode(str, len) : cbquoteencode(str, len);
+ wp += sprintf(wp, "%s?=", enc);
+ free(enc);
+ return buf;
+}
+
+
+/* Decode a string encoded with MIME encoding. */
+char *cbmimedecode(const char *str, char *enp){
+ char *buf, *wp, *tmp, *dec;
+ const char *pv, *ep;
+ int quoted;
+ assert(str);
+ if(enp) sprintf(enp, "US-ASCII");
+ CB_MALLOC(buf, strlen(str) + 1);
+ wp = buf;
+ while(*str != '\0'){
+ if(cbstrfwmatch(str, "=?")){
+ str += 2;
+ pv = str;
+ if(!(ep = strchr(str, '?'))) continue;
+ if(enp && ep - pv < CB_ENCBUFSIZ){
+ memcpy(enp, pv, ep - pv);
+ enp[ep-pv] = '\0';
+ }
+ pv = ep + 1;
+ quoted = (*pv == 'Q' || *pv == 'q');
+ if(*pv != '\0') pv++;
+ if(*pv != '\0') pv++;
+ if(!(ep = strchr(pv, '?'))) continue;
+ CB_MEMDUP(tmp, pv, ep - pv);
+ dec = quoted ? cbquotedecode(tmp, NULL) : cbbasedecode(tmp, NULL);
+ wp += sprintf(wp, "%s", dec);
+ free(dec);
+ free(tmp);
+ str = ep + 1;
+ if(*str != '\0') str++;
+ } else {
+ *(wp++) = *str;
+ str++;
+ }
+ }
+ *wp = '\0';
+ return buf;
+}
+
+
+/* Split a string of CSV into rows. */
+CBLIST *cbcsvrows(const char *str){
+ CBLIST *list;
+ const char *pv;
+ int quoted;
+ assert(str);
+ CB_LISTOPEN(list);
+ pv = str;
+ quoted = FALSE;
+ while(TRUE){
+ if(*str == '"') quoted = !quoted;
+ if(!quoted && (*str == '\r' || *str == '\n')){
+ CB_LISTPUSH(list, pv, str - pv);
+ if(str[0] == '\r' && str[1] == '\n') str++;
+ str++;
+ pv = str;
+ } else if(*str == '\0'){
+ if(str > pv) CB_LISTPUSH(list, pv, str - pv);
+ break;
+ } else {
+ str++;
+ }
+ }
+ return list;
+}
+
+
+/* Split a string of a row of CSV into cells. */
+CBLIST *cbcsvcells(const char *str){
+ CBLIST *list, *uelist;
+ const char *pv;
+ char *tmp;
+ int i, quoted;
+ assert(str);
+ CB_LISTOPEN(list);
+ pv = str;
+ quoted = FALSE;
+ while(TRUE){
+ if(*str == '"') quoted = !quoted;
+ if(!quoted && *str == ','){
+ CB_LISTPUSH(list, pv, str - pv);
+ str++;
+ pv = str;
+ } else if(*str == '\0'){
+ CB_LISTPUSH(list, pv, str - pv);
+ break;
+ } else {
+ str++;
+ }
+ }
+ CB_LISTOPEN(uelist);
+ for(i = 0; i < CB_LISTNUM(list); i++){
+ tmp = cbcsvunescape(CB_LISTVAL(list, i));
+ CB_LISTPUSH(uelist, tmp, strlen(tmp));
+ free(tmp);
+ }
+ CB_LISTCLOSE(list);
+ return uelist;
+}
+
+
+/* Escape a string with the meta characters of CSV. */
+char *cbcsvescape(const char *str){
+ char *buf, *wp;
+ int i;
+ assert(str);
+ CB_MALLOC(buf, strlen(str) * 2 + 3);
+ wp = buf;
+ *(wp++) = '"';
+ for(i = 0; str[i] != '\0'; i++){
+ if(str[i] == '"') *(wp++) = '"';
+ *(wp++) = str[i];
+ }
+ *(wp++) = '"';
+ *wp = '\0';
+ return buf;
+}
+
+
+/* Unescape a string with the escaped meta characters of CSV. */
+char *cbcsvunescape(const char *str){
+ char *buf, *wp;
+ int i, len;
+ assert(str);
+ len = strlen(str);
+ if(str[0] == '"'){
+ str++;
+ len--;
+ if(str[len-1] == '"') len--;
+ }
+ CB_MALLOC(buf, len + 1);
+ wp = buf;
+ for(i = 0; i < len; i++){
+ if(str[i] == '"'){
+ if(str[i+1] == '"') *(wp++) = str[i++];
+ } else {
+ *(wp++) = str[i];
+ }
+ }
+ *wp = '\0';
+ return buf;
+}
+
+
+/* Split a string of XML into tags and text sections. */
+CBLIST *cbxmlbreak(const char *str, int cr){
+ CBLIST *list;
+ CBDATUM *datum;
+ int i, pv, tag;
+ char *ep;
+ assert(str);
+ CB_LISTOPEN(list);
+ i = 0;
+ pv = 0;
+ tag = FALSE;
+ while(TRUE){
+ if(str[i] == '\0'){
+ if(i > pv) CB_LISTPUSH(list, str + pv, i - pv);
+ break;
+ } else if(!tag && str[i] == '<'){
+ if(str[i+1] == '!' && str[i+2] == '-' && str[i+3] == '-'){
+ if(i > pv) CB_LISTPUSH(list, str + pv, i - pv);
+ if((ep = strstr(str + i, "-->")) != NULL){
+ if(!cr) CB_LISTPUSH(list, str + i, ep - str - i + 3);
+ i = ep - str + 2;
+ pv = i + 1;
+ }
+ } else if(str[i+1] == '!' && str[i+2] == '[' && cbstrfwimatch(str + i, "<![CDATA[")){
+ if(i > pv) CB_LISTPUSH(list, str + pv, i - pv);
+ if((ep = strstr(str + i, "]]>")) != NULL){
+ i += 9;
+ CB_DATUMOPEN(datum);
+ while(str + i < ep){
+ if(str[i] == '&'){
+ CB_DATUMCAT(datum, "&", 5);
+ } else if(str[i] == '<'){
+ CB_DATUMCAT(datum, "<", 4);
+ } else if(str[i] == '>'){
+ CB_DATUMCAT(datum, ">", 4);
+ } else {
+ CB_DATUMCAT(datum, str + i, 1);
+ }
+ i++;
+ }
+ if(CB_DATUMSIZE(datum) > 0) CB_LISTPUSH(list, CB_DATUMPTR(datum), CB_DATUMSIZE(datum));
+ CB_DATUMCLOSE(datum);
+ i = ep - str + 2;
+ pv = i + 1;
+ }
+ } else {
+ if(i > pv) CB_LISTPUSH(list, str + pv, i - pv);
+ tag = TRUE;
+ pv = i;
+ }
+ } else if(tag && str[i] == '>'){
+ if(i > pv) CB_LISTPUSH(list, str + pv, i - pv + 1);
+ tag = FALSE;
+ pv = i + 1;
+ }
+ i++;
+ }
+ return list;
+}
+
+
+/* Get the map of attributes of a XML tag. */
+CBMAP *cbxmlattrs(const char *str){
+ CBMAP *map;
+ const unsigned char *rp, *key, *val;
+ char *copy, *raw;
+ int ksiz, vsiz;
+ assert(str);
+ map = cbmapopenex(CB_MAPPBNUM);
+ rp = (unsigned char *)str;
+ while(*rp == '<' || *rp == '/' || *rp == '?' || *rp == '!' || *rp == ' '){
+ rp++;
+ }
+ key = rp;
+ while(*rp > 0x20 && *rp != '/' && *rp != '>'){
+ rp++;
+ }
+ cbmapput(map, "", -1, (char *)key, rp - key, FALSE);
+ while(*rp != '\0'){
+ while(*rp != '\0' && (*rp <= 0x20 || *rp == '/' || *rp == '?' || *rp == '>')){
+ rp++;
+ }
+ key = rp;
+ while(*rp > 0x20 && *rp != '/' && *rp != '>' && *rp != '='){
+ rp++;
+ }
+ ksiz = rp - key;
+ while(*rp != '\0' && (*rp == '=' || *rp <= 0x20)){
+ rp++;
+ }
+ if(*rp == '"'){
+ rp++;
+ val = rp;
+ while(*rp != '\0' && *rp != '"'){
+ rp++;
+ }
+ vsiz = rp - val;
+ } else if(*rp == '\''){
+ rp++;
+ val = rp;
+ while(*rp != '\0' && *rp != '\''){
+ rp++;
+ }
+ vsiz = rp - val;
+ } else {
+ val = rp;
+ while(*rp > 0x20 && *rp != '"' && *rp != '\'' && *rp != '>'){
+ rp++;
+ }
+ vsiz = rp - val;
+ }
+ if(*rp != '\0') rp++;
+ if(ksiz > 0){
+ CB_MEMDUP(copy, (char *)val, vsiz);
+ raw = cbxmlunescape(copy);
+ cbmapput(map, (char *)key, ksiz, raw, -1, FALSE);
+ free(raw);
+ free(copy);
+ }
+ }
+ return map;
+}
+
+
+/* Escape a string with the meta characters of XML. */
+char *cbxmlescape(const char *str){
+ CBDATUM *datum;
+ assert(str);
+ CB_DATUMOPEN(datum);
+ while(*str != '\0'){
+ switch(*str){
+ case '&':
+ CB_DATUMCAT(datum, "&", 5);
+ break;
+ case '<':
+ CB_DATUMCAT(datum, "<", 4);
+ break;
+ case '>':
+ CB_DATUMCAT(datum, ">", 4);
+ break;
+ case '"':
+ CB_DATUMCAT(datum, """, 6);
+ break;
+ case '\'':
+ CB_DATUMCAT(datum, "'", 6);
+ break;
+ default:
+ CB_DATUMCAT(datum, str, 1);
+ break;
+ }
+ str++;
+ }
+ return cbdatumtomalloc(datum, NULL);
+}
+
+
+/* Unescape a string with the entity references of XML. */
+char *cbxmlunescape(const char *str){
+ CBDATUM *datum;
+ assert(str);
+ CB_DATUMOPEN(datum);
+ while(*str != '\0'){
+ if(*str == '&'){
+ if(cbstrfwmatch(str, "&")){
+ CB_DATUMCAT(datum, "&", 1);
+ str += 5;
+ } else if(cbstrfwmatch(str, "<")){
+ CB_DATUMCAT(datum, "<", 1);
+ str += 4;
+ } else if(cbstrfwmatch(str, ">")){
+ CB_DATUMCAT(datum, ">", 1);
+ str += 4;
+ } else if(cbstrfwmatch(str, """)){
+ CB_DATUMCAT(datum, "\"", 1);
+ str += 6;
+ } else if(cbstrfwmatch(str, "'")){
+ CB_DATUMCAT(datum, "'", 1);
+ str += 6;
+ } else {
+ CB_DATUMCAT(datum, str, 1);
+ str++;
+ }
+ } else {
+ CB_DATUMCAT(datum, str, 1);
+ str++;
+ }
+ }
+ return cbdatumtomalloc(datum, NULL);
+}
+
+
+/* Compress a serial object with ZLIB. */
+char *cbdeflate(const char *ptr, int size, int *sp){
+ assert(ptr && sp);
+ if(!_qdbm_deflate) return NULL;
+ return _qdbm_deflate(ptr, size, sp, _QDBM_ZMZLIB);
+}
+
+
+/* Decompress a serial object compressed with ZLIB. */
+char *cbinflate(const char *ptr, int size, int *sp){
+ assert(ptr && size >= 0);
+ if(!_qdbm_inflate) return NULL;
+ return _qdbm_inflate(ptr, size, sp, _QDBM_ZMZLIB);
+}
+
+
+/* Compress a serial object with GZIP. */
+char *cbgzencode(const char *ptr, int size, int *sp){
+ assert(ptr && sp);
+ if(!_qdbm_deflate) return NULL;
+ return _qdbm_deflate(ptr, size, sp, _QDBM_ZMGZIP);
+}
+
+
+/* Decompress a serial object compressed with GZIP. */
+char *cbgzdecode(const char *ptr, int size, int *sp){
+ assert(ptr && size >= 0);
+ if(!_qdbm_inflate) return NULL;
+ return _qdbm_inflate(ptr, size, sp, _QDBM_ZMGZIP);
+}
+
+
+/* Get the CRC32 checksum of a serial object. */
+unsigned int cbgetcrc(const char *ptr, int size){
+ assert(ptr);
+ if(!_qdbm_inflate) return 0;
+ return _qdbm_getcrc(ptr, size);
+}
+
+
+/* Compress a serial object with LZO. */
+char *cblzoencode(const char *ptr, int size, int *sp){
+ assert(ptr && sp);
+ if(!_qdbm_lzoencode) return NULL;
+ return _qdbm_lzoencode(ptr, size, sp);
+}
+
+
+/* Decompress a serial object compressed with LZO. */
+char *cblzodecode(const char *ptr, int size, int *sp){
+ assert(ptr && size >= 0);
+ if(!_qdbm_lzodecode) return NULL;
+ return _qdbm_lzodecode(ptr, size, sp);
+}
+
+
+/* Compress a serial object with BZIP2. */
+char *cbbzencode(const char *ptr, int size, int *sp){
+ assert(ptr && sp);
+ if(!_qdbm_bzencode) return NULL;
+ return _qdbm_bzencode(ptr, size, sp);
+}
+
+
+/* Decompress a serial object compressed with BZIP2. */
+char *cbbzdecode(const char *ptr, int size, int *sp){
+ assert(ptr && size >= 0);
+ if(!_qdbm_bzdecode) return NULL;
+ return _qdbm_bzdecode(ptr, size, sp);
+}
+
+
+/* Convert the character encoding of a string. */
+char *cbiconv(const char *ptr, int size, const char *icode, const char *ocode, int *sp, int *mp){
+ char *res;
+ assert(ptr && icode && ocode);
+ if(!_qdbm_iconv) return NULL;
+ if((res = _qdbm_iconv(ptr, size, icode, ocode, sp, mp)) != NULL) return res;
+ if(!cbstricmp(icode, ocode)){
+ if(sp) *sp = size;
+ if(mp) *mp = 0;
+ CB_MEMDUP(res, ptr, size < 0 ? strlen(ptr) : size);
+ return res;
+ }
+ return NULL;
+}
+
+
+/* Detect the encoding of a string automatically. */
+const char *cbencname(const char *ptr, int size){
+ assert(ptr);
+ if(!_qdbm_encname) return "ISO-8859-1";
+ return _qdbm_encname(ptr, size);
+}
+
+
+/* Get the jet lag of the local time in seconds. */
+int cbjetlag(void){
+ struct tm ts, *tp;
+ time_t t, gt, lt;
+ if((t = time(NULL)) < 0) return 0;
+ if(!(tp = _qdbm_gmtime(&t, &ts))) return 0;
+ if((gt = mktime(tp)) < 0) return 0;
+ if(!(tp = _qdbm_localtime(&t, &ts))) return 0;
+ if((lt = mktime(tp)) < 0) return 0;
+ return lt - gt;
+}
+
+
+/* Get the Gregorian calendar of a time. */
+void cbcalendar(time_t t, int jl, int *yearp, int *monp, int *dayp,
+ int *hourp, int *minp, int *secp){
+ struct tm ts, *tp;
+ if(t < 0) t = time(NULL);
+ t += jl;
+ if(!(tp = _qdbm_gmtime(&t, &ts))) return;
+ if(yearp) *yearp = tp->tm_year + 1900;
+ if(monp) *monp = tp->tm_mon + 1;
+ if(dayp) *dayp = tp->tm_mday;
+ if(hourp) *hourp = tp->tm_hour;
+ if(minp) *minp = tp->tm_min;
+ if(secp) *secp = tp->tm_sec;
+}
+
+
+/* Get the day of week of a date. */
+int cbdayofweek(int year, int mon, int day){
+ if(mon < 3){
+ year--;
+ mon += 12;
+ }
+ return (day + ((8 + (13 * mon)) / 5) + (year + (year / 4) - (year / 100) + (year / 400))) % 7;
+}
+
+
+/* Get the string for a date in W3CDTF. */
+char *cbdatestrwww(time_t t, int jl){
+ char date[CB_DATEBUFSIZ], tzone[CB_DATEBUFSIZ], *rv;
+ int year, mon, day, hour, min, sec;
+ cbcalendar(t, jl, &year, &mon, &day, &hour, &min, &sec);
+ jl /= 60;
+ if(jl == 0){
+ sprintf(tzone, "Z");
+ } else if(jl < 0){
+ jl *= -1;
+ sprintf(tzone, "-%02d:%02d", jl / 60, jl % 60);
+ } else {
+ sprintf(tzone, "+%02d:%02d", jl / 60, jl % 60);
+ }
+ sprintf(date, "%04d-%02d-%02dT%02d:%02d:%02d%s", year, mon, day, hour, min, sec, tzone);
+ CB_MEMDUP(rv, date, strlen(date));
+ return rv;
+}
+
+
+/* Get the string for a date in RFC 1123 format. */
+char *cbdatestrhttp(time_t t, int jl){
+ char date[CB_DATEBUFSIZ], *wp, *rv;
+ int year, mon, day, hour, min, sec;
+ cbcalendar(t, jl, &year, &mon, &day, &hour, &min, &sec);
+ jl /= 60;
+ wp = date;
+ switch(cbdayofweek(year, mon, day)){
+ case 0: wp += sprintf(wp, "Sun, "); break;
+ case 1: wp += sprintf(wp, "Mon, "); break;
+ case 2: wp += sprintf(wp, "Tue, "); break;
+ case 3: wp += sprintf(wp, "Wed, "); break;
+ case 4: wp += sprintf(wp, "Thu, "); break;
+ case 5: wp += sprintf(wp, "Fri, "); break;
+ case 6: wp += sprintf(wp, "Sat, "); break;
+ }
+ wp += sprintf(wp, "%02d ", day);
+ switch(mon){
+ case 1: wp += sprintf(wp, "Jan "); break;
+ case 2: wp += sprintf(wp, "Feb "); break;
+ case 3: wp += sprintf(wp, "Mar "); break;
+ case 4: wp += sprintf(wp, "Apr "); break;
+ case 5: wp += sprintf(wp, "May "); break;
+ case 6: wp += sprintf(wp, "Jun "); break;
+ case 7: wp += sprintf(wp, "Jul "); break;
+ case 8: wp += sprintf(wp, "Aug "); break;
+ case 9: wp += sprintf(wp, "Sep "); break;
+ case 10: wp += sprintf(wp, "Oct "); break;
+ case 11: wp += sprintf(wp, "Nov "); break;
+ case 12: wp += sprintf(wp, "Dec "); break;
+ }
+ wp += sprintf(wp, "%04d %02d:%02d:%02d ", year, hour, min, sec);
+ if(jl == 0){
+ wp += sprintf(wp, "GMT");
+ } else if(jl < 0){
+ jl *= -1;
+ wp += sprintf(wp, "-%02d%02d", jl / 60, jl % 60);
+ } else {
+ wp += sprintf(wp, "+%02d%02d", jl / 60, jl % 60);
+ }
+ CB_MEMDUP(rv, date, strlen(date));
+ return rv;
+}
+
+
+/* Get the time value of a date string in decimal, W3CDTF, or RFC 1123. */
+time_t cbstrmktime(const char *str){
+ const char *crp;
+ char *pv, *rp;
+ int len, clen;
+ time_t t;
+ struct tm ts;
+ assert(str);
+ while(*str > '\0' && *str <= ' '){
+ str++;
+ }
+ if(*str == '\0') return -1;
+ if(str[0] == '0' && (str[1] == 'x' || str[1] == 'X')) return (time_t)strtol(str + 2, NULL, 16);
+ memset(&ts, 0, sizeof(struct tm));
+ ts.tm_year = 70;
+ ts.tm_mon = 0;
+ ts.tm_mday = 1;
+ ts.tm_hour = 0;
+ ts.tm_min = 0;
+ ts.tm_sec = 0;
+ ts.tm_isdst = 0;
+ len = strlen(str);
+ t = (time_t)strtol(str, &pv, 10);
+ if(*(signed char *)pv >= '\0' && *pv <= ' '){
+ while(*pv > '\0' && *pv <= ' '){
+ pv++;
+ }
+ if(*pv == '\0') return t;
+ }
+ if((pv[0] == 's' || pv[0] == 'S') && ((signed char *)pv)[1] >= '\0' && pv[1] <= ' ')
+ return t;
+ if((pv[0] == 'm' || pv[0] == 'M') && ((signed char *)pv)[1] >= '\0' && pv[1] <= ' ')
+ return t * 60;
+ if((pv[0] == 'h' || pv[0] == 'H') && ((signed char *)pv)[1] >= '\0' && pv[1] <= ' ')
+ return t * 60 * 60;
+ if((pv[0] == 'd' || pv[0] == 'D') && ((signed char *)pv)[1] >= '\0' && pv[1] <= ' ')
+ return t * 60 * 60 * 24;
+ if(len > 4 && str[4] == '-'){
+ ts.tm_year = atoi(str) - 1900;
+ if((pv = strchr(str, '-')) != NULL && pv - str == 4){
+ rp = pv + 1;
+ ts.tm_mon = atoi(rp) - 1;
+ if((pv = strchr(rp, '-')) != NULL && pv - str == 7){
+ rp = pv + 1;
+ ts.tm_mday = atoi(rp);
+ if((pv = strchr(rp, 'T')) != NULL && pv - str == 10){
+ rp = pv + 1;
+ ts.tm_hour = atoi(rp);
+ if((pv = strchr(rp, ':')) != NULL && pv - str == 13){
+ rp = pv + 1;
+ ts.tm_min = atoi(rp);
+ }
+ if((pv = strchr(rp, ':')) != NULL && pv - str == 16){
+ rp = pv + 1;
+ ts.tm_sec = atoi(rp);
+ }
+ if((pv = strchr(rp, '.')) != NULL && pv - str >= 19) rp = pv + 1;
+ strtol(rp, &pv, 10);
+ if((*pv == '+' || *pv == '-') && strlen(pv) >= 6 && pv[3] == ':')
+ ts.tm_sec -= (atoi(pv + 1) * 3600 + atoi(pv + 4) * 60) * (pv[0] == '+' ? 1 : -1);
+ }
+ }
+ }
+ ts.tm_sec += cbjetlag();
+ return mktime(&ts);
+ }
+ if(len > 4 && str[4] == '/'){
+ ts.tm_year = atoi(str) - 1900;
+ if((pv = strchr(str, '/')) != NULL && pv - str == 4){
+ rp = pv + 1;
+ ts.tm_mon = atoi(rp) - 1;
+ if((pv = strchr(rp, '/')) != NULL && pv - str == 7){
+ rp = pv + 1;
+ ts.tm_mday = atoi(rp);
+ if((pv = strchr(rp, ' ')) != NULL && pv - str == 10){
+ rp = pv + 1;
+ ts.tm_hour = atoi(rp);
+ if((pv = strchr(rp, ':')) != NULL && pv - str == 13){
+ rp = pv + 1;
+ ts.tm_min = atoi(rp);
+ }
+ if((pv = strchr(rp, ':')) != NULL && pv - str == 16){
+ rp = pv + 1;
+ ts.tm_sec = atoi(rp);
+ }
+ if((pv = strchr(rp, '.')) != NULL && pv - str >= 19) rp = pv + 1;
+ strtol(rp, &pv, 10);
+ if((*pv == '+' || *pv == '-') && strlen(pv) >= 6 && pv[3] == ':')
+ ts.tm_sec -= (atoi(pv + 1) * 3600 + atoi(pv + 4) * 60) * (pv[0] == '+' ? 1 : -1);
+ }
+ }
+ }
+ ts.tm_sec += cbjetlag();
+ return mktime(&ts);
+ }
+ crp = str;
+ if(len >= 4 && str[3] == ',') crp = str + 4;
+ while(*crp == ' '){
+ crp++;
+ }
+ ts.tm_mday = atoi(crp);
+ while((*crp >= '0' && *crp <= '9') || *crp == ' '){
+ crp++;
+ }
+ if(cbstrfwimatch(crp, "Jan")){
+ ts.tm_mon = 0;
+ } else if(cbstrfwimatch(crp, "Feb")){
+ ts.tm_mon = 1;
+ } else if(cbstrfwimatch(crp, "Mar")){
+ ts.tm_mon = 2;
+ } else if(cbstrfwimatch(crp, "Apr")){
+ ts.tm_mon = 3;
+ } else if(cbstrfwimatch(crp, "May")){
+ ts.tm_mon = 4;
+ } else if(cbstrfwimatch(crp, "Jun")){
+ ts.tm_mon = 5;
+ } else if(cbstrfwimatch(crp, "Jul")){
+ ts.tm_mon = 6;
+ } else if(cbstrfwimatch(crp, "Aug")){
+ ts.tm_mon = 7;
+ } else if(cbstrfwimatch(crp, "Sep")){
+ ts.tm_mon = 8;
+ } else if(cbstrfwimatch(crp, "Oct")){
+ ts.tm_mon = 9;
+ } else if(cbstrfwimatch(crp, "Nov")){
+ ts.tm_mon = 10;
+ } else if(cbstrfwimatch(crp, "Dec")){
+ ts.tm_mon = 11;
+ } else {
+ ts.tm_mon = -1;
+ }
+ if(ts.tm_mon >= 0) crp += 3;
+ while(*crp == ' '){
+ crp++;
+ }
+ ts.tm_year = atoi(crp);
+ if(ts.tm_year >= 1969) ts.tm_year -= 1900;
+ while(*crp >= '0' && *crp <= '9'){
+ crp++;
+ }
+ while(*crp == ' '){
+ crp++;
+ }
+ if(ts.tm_mday > 0 && ts.tm_mon >= 0 && ts.tm_year >= 0){
+ clen = strlen(crp);
+ if(clen >= 8 && crp[2] == ':' && crp[5] == ':'){
+ ts.tm_hour = atoi(crp + 0);
+ ts.tm_min = atoi(crp + 3);
+ ts.tm_sec = atoi(crp + 6);
+ if(clen >= 14 && crp[8] == ' ' && (crp[9] == '+' || crp[9] == '-')){
+ ts.tm_sec -= ((crp[10] - '0') * 36000 + (crp[11] - '0') * 3600 +
+ (crp[12] - '0') * 600 + (crp[13] - '0') * 60) * (crp[9] == '+' ? 1 : -1);
+ } else if(clen > 9){
+ if(!strcmp(crp + 9, "JST")){
+ ts.tm_sec -= 9 * 3600;
+ } else if(!strcmp(crp + 9, "CCT")){
+ ts.tm_sec -= 8 * 3600;
+ } else if(!strcmp(crp + 9, "KST")){
+ ts.tm_sec -= 9 * 3600;
+ } else if(!strcmp(crp + 9, "EDT")){
+ ts.tm_sec -= -4 * 3600;
+ } else if(!strcmp(crp + 9, "EST")){
+ ts.tm_sec -= -5 * 3600;
+ } else if(!strcmp(crp + 9, "CDT")){
+ ts.tm_sec -= -5 * 3600;
+ } else if(!strcmp(crp + 9, "CST")){
+ ts.tm_sec -= -6 * 3600;
+ } else if(!strcmp(crp + 9, "MDT")){
+ ts.tm_sec -= -6 * 3600;
+ } else if(!strcmp(crp + 9, "MST")){
+ ts.tm_sec -= -7 * 3600;
+ } else if(!strcmp(crp + 9, "PDT")){
+ ts.tm_sec -= -7 * 3600;
+ } else if(!strcmp(crp + 9, "PST")){
+ ts.tm_sec -= -8 * 3600;
+ } else if(!strcmp(crp + 9, "HDT")){
+ ts.tm_sec -= -9 * 3600;
+ } else if(!strcmp(crp + 9, "HST")){
+ ts.tm_sec -= -10 * 3600;
+ }
+ }
+ }
+ ts.tm_sec += cbjetlag();
+ return mktime(&ts);
+ }
+ return -1;
+}
+
+
+/* Get user and system processing times. */
+void cbproctime(double *usrp, double *sysp){
+ struct tms buf;
+ times(&buf);
+ if(usrp) *usrp = (double)buf.tms_utime / sysconf(_SC_CLK_TCK);
+ if(sysp) *sysp = (double)buf.tms_stime / sysconf(_SC_CLK_TCK);
+}
+
+
+/* Ensure that the standard I/O is binary mode. */
+void cbstdiobin(void){
+ if(setmode(0, O_BINARY) == -1 || setmode(1, O_BINARY) == -1 || setmode(2, O_BINARY) == -1)
+ cbmyfatal("setmode failed");
+}
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Show error message on the standard error output and exit. */
+void *cbmyfatal(const char *message){
+ char buf[CB_MSGBUFSIZ];
+ assert(message);
+ if(cbfatalfunc){
+ cbfatalfunc(message);
+ } else {
+ sprintf(buf, "fatal error: %s\n", message);
+ write(2, buf, strlen(buf));
+ }
+ exit(1);
+ return NULL;
+}
+
+
+/* Create a datum handle from an allocated region. */
+CBDATUM *cbdatumopenbuf(char *ptr, int size){
+ CBDATUM *datum;
+ assert(ptr && size >= 0);
+ CB_REALLOC(ptr, size + 1);
+ CB_MALLOC(datum, sizeof(*datum));
+ datum->dptr = ptr;
+ datum->dptr[size] = '\0';
+ datum->dsize = size;
+ datum->asize = size;
+ return datum;
+}
+
+
+/* Set a buffer to a datum handle. */
+void cbdatumsetbuf(CBDATUM *datum, char *ptr, int size){
+ assert(datum && ptr && size >= 0);
+ free(datum->dptr);
+ CB_REALLOC(ptr, size + 1);
+ datum->dptr = ptr;
+ datum->dptr[size] = '\0';
+ datum->dsize = size;
+ datum->asize = size;
+}
+
+
+/* Add an allocated element at the end of a list. */
+void cblistpushbuf(CBLIST *list, char *ptr, int size){
+ int index;
+ assert(list && ptr && size >= 0);
+ index = list->start + list->num;
+ if(index >= list->anum){
+ list->anum *= 2;
+ CB_REALLOC(list->array, list->anum * sizeof(list->array[0]));
+ }
+ list->array[index].dptr = ptr;
+ list->array[index].dsize = size;
+ list->num++;
+}
+
+
+/* Get a map handle with specifying the number of buckets. */
+CBMAP *cbmapopenex(int bnum){
+ CBMAP *map;
+ int i;
+ assert(bnum > 0);
+ CB_MALLOC(map, sizeof(*map));
+ CB_MALLOC(map->buckets, sizeof(map->buckets[0]) * bnum);
+ for(i = 0; i < bnum; i++){
+ map->buckets[i] = NULL;
+ }
+ map->first = NULL;
+ map->last = NULL;
+ map->cur = NULL;
+ map->bnum = bnum;
+ map->rnum = 0;
+ return map;
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Handler to invoke the global garbage collector. */
+static void cbggchandler(void){
+ cbggckeeper(NULL, NULL);
+}
+
+
+/* Manage resources of the global garbage collector.
+ `ptr' specifies the pointer to add to the collection. If it is `NULL', all resources are
+ released.
+ `func' specifies the pointer to the function to release the resources. */
+static void cbggckeeper(void *ptr, void (*func)(void *)){
+ static void **parray = NULL;
+ static void (**farray)(void *) = NULL;
+ static int onum = 0;
+ static int asiz = CB_GCUNIT;
+ int i;
+ if(!ptr){
+ if(!parray) return;
+ for(i = onum - 1; i >= 0; i--){
+ farray[i](parray[i]);
+ }
+ free(parray);
+ free(farray);
+ parray = NULL;
+ farray = NULL;
+ onum = 0;
+ asiz = CB_GCUNIT;
+ return;
+ }
+ if(!parray){
+ CB_MALLOC(parray, sizeof(void *) * asiz);
+ CB_MALLOC(farray, sizeof(void *) * asiz);
+ if(atexit(cbggchandler) != 0) cbmyfatal("gc failed");
+ }
+ if(onum >= asiz){
+ asiz *= 2;
+ CB_REALLOC(parray, sizeof(void *) * asiz);
+ CB_REALLOC(farray, sizeof(void *) * asiz);
+ }
+ parray[onum] = ptr;
+ farray[onum] = func;
+ onum++;
+}
+
+
+/* Utility function for quick sort.
+ `bp' specifies the pointer to the pointer to an array.
+ `nmemb' specifies the number of elements of the array.
+ `size' specifies the size of each element.
+ `pswap' specifies the pointer to the swap region for a pivot.
+ `vswap' specifies the pointer to the swap region for elements.
+ `compar' specifies the pointer to comparing function. */
+static void cbqsortsub(char *bp, int nmemb, int size, char *pswap, char *vswap,
+ int(*compar)(const void *, const void *)){
+ int top, bottom;
+ assert(bp && nmemb >= 0 && size > 0 && pswap && vswap && compar);
+ if(nmemb < 10){
+ if(nmemb > 1) cbisort(bp, nmemb, size, compar);
+ return;
+ }
+ top = 0;
+ bottom = nmemb - 1;
+ memcpy(pswap, bp + (nmemb / 2) * size, size);
+ while(top - 1 < bottom){
+ if(compar(bp + top * size, pswap) < 0){
+ top++;
+ } else if(compar(bp + bottom * size, pswap) > 0){
+ bottom--;
+ } else {
+ if(top != bottom){
+ memcpy(vswap, bp + top * size, size);
+ memcpy(bp + top * size, bp + bottom * size, size);
+ memcpy(bp + bottom * size, vswap, size);
+ }
+ top++;
+ bottom--;
+ }
+ }
+ cbqsortsub(bp, top, size, pswap, vswap, compar);
+ cbqsortsub(bp + (bottom + 1) * size, nmemb - bottom - 1, size, pswap, vswap, compar);
+}
+
+
+/* Compare two list elements.
+ `a' specifies the pointer to one element.
+ `b' specifies the pointer to the other element.
+ The return value is positive if a is big, negative if b is big, else, it is 0. */
+static int cblistelemcmp(const void *a, const void *b){
+ int i, size;
+ CBLISTDATUM *ap, *bp;
+ char *ao, *bo;
+ assert(a && b);
+ ap = (CBLISTDATUM *)a;
+ bp = (CBLISTDATUM *)b;
+ ao = ap->dptr;
+ bo = bp->dptr;
+ size = ap->dsize < bp->dsize ? ap->dsize : bp->dsize;
+ for(i = 0; i < size; i++){
+ if(ao[i] > bo[i]) return 1;
+ if(ao[i] < bo[i]) return -1;
+ }
+ return ap->dsize - bp->dsize;
+}
+
+
+/* Compare two keys.
+ `abuf' specifies the pointer to the region of the former.
+ `asiz' specifies the size of the region.
+ `bbuf' specifies the pointer to the region of the latter.
+ `bsiz' specifies the size of the region.
+ The return value is 0 if two equals, positive if the formar is big, else, negative. */
+static int cbkeycmp(const char *abuf, int asiz, const char *bbuf, int bsiz){
+ assert(abuf && asiz >= 0 && bbuf && bsiz >= 0);
+ if(asiz > bsiz) return 1;
+ if(asiz < bsiz) return -1;
+ return memcmp(abuf, bbuf, asiz);
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/cabin.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/cabin.h
===================================================================
--- box/trunk/qdbm/cabin.h (rev 0)
+++ box/trunk/qdbm/cabin.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,1544 @@
+/*************************************************************************************************
+ * The utitlity API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _CABIN_H /* duplication check */
+#define _CABIN_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <stdlib.h>
+#include <time.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+#define CB_DATUMUNIT 12 /* allocation unit size of a datum handle */
+#define CB_LISTUNIT 64 /* allocation unit number of a list handle */
+#define CB_MAPBNUM 4093 /* bucket size of a map handle */
+
+typedef struct { /* type of structure for a basic datum */
+ char *dptr; /* pointer to the region */
+ int dsize; /* size of the region */
+ int asize; /* size of the allocated region */
+} CBDATUM;
+
+typedef struct { /* type of structure for an element of a list */
+ char *dptr; /* pointer to the region */
+ int dsize; /* size of the effective region */
+} CBLISTDATUM;
+
+typedef struct { /* type of structure for a list */
+ CBLISTDATUM *array; /* array of data */
+ int anum; /* number of the elements of the array */
+ int start; /* start index of using elements */
+ int num; /* number of using elements */
+} CBLIST;
+
+typedef struct _CBMAPDATUM { /* type of structure for an element of a map */
+ int ksiz; /* size of the region of the key */
+ int vsiz; /* size of the region of the value */
+ int hash; /* second hash value */
+ struct _CBMAPDATUM *left; /* pointer to the left child */
+ struct _CBMAPDATUM *right; /* pointer to the right child */
+ struct _CBMAPDATUM *prev; /* pointer to the previous element */
+ struct _CBMAPDATUM *next; /* pointer to the next element */
+} CBMAPDATUM;
+
+typedef struct { /* type of structure for a map */
+ CBMAPDATUM **buckets; /* bucket array */
+ CBMAPDATUM *first; /* pointer to the first element */
+ CBMAPDATUM *last; /* pointer to the last element */
+ CBMAPDATUM *cur; /* pointer to the current element */
+ int bnum; /* number of buckets */
+ int rnum; /* number of records */
+} CBMAP;
+
+typedef struct { /* type of structure for a heap */
+ char *base; /* base pointer */
+ char *swap; /* region for swapping */
+ int size; /* size of each record */
+ int num; /* currnet number of records */
+ int max; /* maximum number of records */
+ int(*compar)(const void *, const void *); /* comparing function */
+} CBHEAP;
+
+
+/* Call back function for handling a fatal error.
+ The argument specifies the error message. The initial value of this variable is `NULL'.
+ If the value is `NULL', the default function is called when a fatal error occurs. A fatal
+ error occurs when memory allocation is failed. */
+MYEXTERN void (*cbfatalfunc)(const char *);
+
+
+/* Allocate a region on memory.
+ `size' specifies the size of the region.
+ The return value is the pointer to the allocated region.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+void *cbmalloc(size_t size);
+
+
+/* Re-allocate a region on memory.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region.
+ The return value is the pointer to the re-allocated region.
+ Because the region of the return value is allocated with the `realloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+void *cbrealloc(void *ptr, size_t size);
+
+
+/* Duplicate a region on memory.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the pointer to the allocated region of the duplicate.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if
+ it is no longer in use. */
+char *cbmemdup(const char *ptr, int size);
+
+
+/* Free a region on memory.
+ `ptr' specifies the pointer to a region. If it is `NULL', this function has no effect.
+ Although this function is just a wrapper of `free' call, this is useful in applications using
+ another package of the `malloc' series. */
+void cbfree(void *ptr);
+
+
+/* Register the pointer or handle of an object to the global garbage collector.
+ `ptr' specifies the pointer or handle of an object.
+ `func' specifies the pointer to a function to release resources of the object. Its argument
+ is the pointer or handle of the object to release.
+ This function assures that resources of an object are released when the process exits
+ normally by returning from the `main' function or calling the `exit' function. */
+void cbglobalgc(void *ptr, void (*func)(void *));
+
+
+/* Exercise the global garbage collector explicitly.
+ Note that you should not use objects registered to the global garbage collector any longer
+ after calling this function. Because the global garbage collecter is initialized and you
+ can register new objects into it. */
+void cbggcsweep(void);
+
+
+/* Check availability of allocation of the virtual memory.
+ `size' specifies the size of region to be allocated newly.
+ The return value is true if allocation should be success, or false if not. */
+int cbvmemavail(size_t size);
+
+
+/* Sort an array using insert sort.
+ `base' spacifies the pointer to an array.
+ `nmemb' specifies the number of elements of the array.
+ `size' specifies the size of each element.
+ `compar' specifies the pointer to comparing function. The two arguments specify the pointers
+ of elements. The comparing function should returns positive if the former is big, negative
+ if the latter is big, 0 if both are equal.
+ Insert sort is useful only if most elements have been sorted already. */
+void cbisort(void *base, int nmemb, int size, int(*compar)(const void *, const void *));
+
+
+/* Sort an array using shell sort.
+ `base' spacifies the pointer to an array.
+ `nmemb' specifies the number of elements of the array.
+ `size' specifies the size of each element.
+ `compar' specifies the pointer to comparing function. The two arguments specify the pointers
+ of elements. The comparing function should returns positive if the former is big, negative
+ if the latter is big, 0 if both are equal.
+ If most elements have been sorted, shell sort may be faster than heap sort or quick sort. */
+void cbssort(void *base, int nmemb, int size, int(*compar)(const void *, const void *));
+
+
+/* Sort an array using heap sort.
+ `base' spacifies the pointer to an array.
+ `nmemb' specifies the number of elements of the array.
+ `size' specifies the size of each element.
+ `compar' specifies the pointer to comparing function. The two arguments specify the pointers
+ of elements. The comparing function should returns positive if the former is big, negative
+ if the latter is big, 0 if both are equal.
+ Although heap sort is robust against bias of input, quick sort is faster in most cases. */
+void cbhsort(void *base, int nmemb, int size, int(*compar)(const void *, const void *));
+
+
+/* Sort an array using quick sort.
+ `base' spacifies the pointer to an array.
+ `nmemb' specifies the number of elements of the array.
+ `size' specifies the size of each element.
+ `compar' specifies the pointer to comparing function. The two arguments specify the pointers
+ of elements. The comparing function should returns positive if the former is big, negative
+ if the latter is big, 0 if both are equal.
+ Being sensitive to bias of input, quick sort is the fastest sorting algorithm. */
+void cbqsort(void *base, int nmemb, int size, int(*compar)(const void *, const void *));
+
+
+/* Compare two strings with case insensitive evaluation.
+ `astr' specifies the pointer of one string.
+ `astr' specifies the pointer of the other string.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent.
+ Upper cases and lower cases of alphabets in ASCII code are not distinguished. */
+int cbstricmp(const char *astr, const char *bstr);
+
+
+/* Check whether a string begins with a key.
+ `str' specifies the pointer of a target string.
+ `key' specifies the pointer of a forward matching key string.
+ The return value is true if the target string begins with the key, else, it is false. */
+int cbstrfwmatch(const char *str, const char *key);
+
+
+/* Check whether a string begins with a key, with case insensitive evaluation.
+ `str' specifies the pointer of a target string.
+ `key' specifies the pointer of a forward matching key string.
+ The return value is true if the target string begins with the key, else, it is false.
+ Upper cases and lower cases of alphabets in ASCII code are not distinguished. */
+int cbstrfwimatch(const char *str, const char *key);
+
+
+/* Check whether a string ends with a key.
+ `str' specifies the pointer of a target string.
+ `key' specifies the pointer of a backward matching key string.
+ The return value is true if the target string ends with the key, else, it is false. */
+int cbstrbwmatch(const char *str, const char *key);
+
+
+/* Check whether a string ends with a key, with case insensitive evaluation.
+ `str' specifies the pointer of a target string.
+ `key' specifies the pointer of a backward matching key string.
+ The return value is true if the target string ends with the key, else, it is false.
+ Upper cases and lower cases of alphabets in ASCII code are not distinguished. */
+int cbstrbwimatch(const char *str, const char *key);
+
+
+/* Locate a substring in a string using KMP method.
+ `haystack' specifies the pointer of a target string.
+ `needle' specifies the pointer of a substring to be found.
+ The return value is the pointer to the beginning of the substring or `NULL' if the substring
+ is not found.
+ In most cases, `strstr' as a built-in function of the compiler is faster than this function. */
+char *cbstrstrkmp(const char *haystack, const char *needle);
+
+
+/* Locate a substring in a string using BM method.
+ `haystack' specifies the pointer of a target string.
+ `needle' specifies the pointer of a substring to be found.
+ The return value is the pointer to the beginning of the substring or `NULL' if the substring
+ is not found.
+ In most cases, `strstr' as a built-in function of the compiler is faster than this function. */
+char *cbstrstrbm(const char *haystack, const char *needle);
+
+
+/* Convert the letters of a string to upper case.
+ `str' specifies the pointer of a string to convert.
+ The return value is the pointer to the string. */
+char *cbstrtoupper(char *str);
+
+
+/* Convert the letters of a string to lower case.
+ `str' specifies the pointer of a string to convert.
+ The return value is the pointer to the string. */
+char *cbstrtolower(char *str);
+
+
+/* Cut space characters at head or tail of a string.
+ `str' specifies the pointer of a string to convert.
+ The return value is the pointer to the string. */
+char *cbstrtrim(char *str);
+
+
+/* Squeeze space characters in a string and trim it.
+ `str' specifies the pointer of a string to convert.
+ The return value is the pointer to the string. */
+char *cbstrsqzspc(char *str);
+
+
+/* Count the number of characters in a string of UTF-8.
+ `str' specifies the pointer of a string of UTF-8.
+ The return value is the number of characters in the string. */
+int cbstrcountutf(const char *str);
+
+
+/* Cut a string of UTF-8 at the specified number of characters.
+ `str' specifies the pointer of a string of UTF-8.
+ `num' specifies the number of characters to be kept.
+ The return value is the pointer to the string. */
+char *cbstrcututf(char *str, int num);
+
+
+/* Get a datum handle.
+ `ptr' specifies the pointer to the region of the initial content. If it is `NULL', an empty
+ datum is created.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is a datum handle. */
+CBDATUM *cbdatumopen(const char *ptr, int size);
+
+
+/* Copy a datum.
+ `datum' specifies a datum handle.
+ The return value is a new datum handle. */
+CBDATUM *cbdatumdup(const CBDATUM *datum);
+
+
+/* Free a datum handle.
+ `datum' specifies a datum handle.
+ Because the region of a closed handle is released, it becomes impossible to use the handle. */
+void cbdatumclose(CBDATUM *datum);
+
+
+/* Concatenate a datum and a region.
+ `datum' specifies a datum handle.
+ `ptr' specifies the pointer to the region to be appended.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'. */
+void cbdatumcat(CBDATUM *datum, const char *ptr, int size);
+
+
+/* Get the pointer of the region of a datum.
+ `datum' specifies a datum handle.
+ The return value is the pointer of the region of a datum.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. */
+const char *cbdatumptr(const CBDATUM *datum);
+
+
+/* Get the size of the region of a datum.
+ `datum' specifies a datum handle.
+ The return value is the size of the region of a datum. */
+int cbdatumsize(const CBDATUM *datum);
+
+
+/* Change the size of the region of a datum.
+ `datum' specifies a datum handle.
+ `size' specifies the new size of the region.
+ If the new size is bigger than the one of old, the surplus region is filled with zero codes. */
+void cbdatumsetsize(CBDATUM *datum, int size);
+
+
+/* Perform formatted output into a datum.
+ `format' specifies a printf-like format string.
+ The conversion character `%' can be used with such flag characters as `s', `d', `o', `u',
+ `x', `X', `c', `e', `E', `f', `g', `G', `@', `?', `:', `%'. `@' works as with `s' but escapes
+ meta characters of XML. `?' works as with `s' but escapes meta characters of URL. `:' works
+ as with `s' but performs MIME encoding as UTF-8. The other conversion character work as with
+ each original. */
+void cbdatumprintf(CBDATUM *datum, const char *format, ...);
+
+
+/* Convert a datum to an allocated region.
+ `datum' specifies a datum handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the datum.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. Because the region of the original datam is released, it should not be
+ released again. */
+char *cbdatumtomalloc(CBDATUM *datum, int *sp);
+
+
+/* Get a list handle.
+ The return value is a list handle. */
+CBLIST *cblistopen(void);
+
+
+/* Copy a list.
+ `list' specifies a list handle.
+ The return value is a new list handle. */
+CBLIST *cblistdup(const CBLIST *list);
+
+
+/* Close a list handle.
+ `list' specifies a list handle.
+ Because the region of a closed handle is released, it becomes impossible to use the handle. */
+void cblistclose(CBLIST *list);
+
+
+/* Get the number of elements of a list.
+ `list' specifies a list handle.
+ The return value is the number of elements of the list. */
+int cblistnum(const CBLIST *list);
+
+
+/* Get the pointer to the region of an element of a list.
+ `list' specifies a list handle.
+ `index' specifies the index of an element.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the value.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. If `index' is equal to or more than
+ the number of elements, the return value is `NULL'. */
+const char *cblistval(const CBLIST *list, int index, int *sp);
+
+
+/* Add an element at the end of a list.
+ `list' specifies a list handle.
+ `ptr' specifies the pointer to the region of an element.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'. */
+void cblistpush(CBLIST *list, const char *ptr, int size);
+
+
+/* Remove an element of the end of a list.
+ `list' specifies a list handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the value.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. If the list is empty, the return value is `NULL'. */
+char *cblistpop(CBLIST *list, int *sp);
+
+
+/* Add an element at the top of a list.
+ `list' specifies a list handle.
+ `ptr' specifies the pointer to the region of an element.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'. */
+void cblistunshift(CBLIST *list, const char *ptr, int size);
+
+
+/* Remove an element of the top of a list.
+ `list' specifies a list handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the value.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. If the list is empty, the return value is `NULL'. */
+char *cblistshift(CBLIST *list, int *sp);
+
+
+/* Add an element at the specified location of a list.
+ `list' specifies a list handle.
+ `index' specifies the index of an element.
+ `ptr' specifies the pointer to the region of the element.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'. */
+void cblistinsert(CBLIST *list, int index, const char *ptr, int size);
+
+
+/* Remove an element at the specified location of a list.
+ `list' specifies a list handle.
+ `index' specifies the index of an element.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the value.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. If `index' is equal to or more than the number of elements, no element
+ is removed and the return value is `NULL'. */
+char *cblistremove(CBLIST *list, int index, int *sp);
+
+
+/* Overwrite an element at the specified location of a list.
+ `list' specifies a list handle.
+ `index' specifies the index of an element.
+ `ptr' specifies the pointer to the region of the new content.
+ `size' specifies the size of the new content. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ If `index' is equal to or more than the number of elements, this function has no effect. */
+void cblistover(CBLIST *list, int index, const char *ptr, int size);
+
+
+/* Sort elements of a list in lexical order.
+ `list' specifies a list handle.
+ Quick sort is used for sorting. */
+void cblistsort(CBLIST *list);
+
+
+/* Search a list for an element using liner search.
+ `list' specifies a list handle.
+ `ptr' specifies the pointer to the region of a key.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the index of a corresponding element or -1 if there is no corresponding
+ element. If two or more elements corresponds, the former returns. */
+int cblistlsearch(const CBLIST *list, const char *ptr, int size);
+
+
+/* Search a list for an element using binary search.
+ `list' specifies a list handle. It should be sorted in lexical order.
+ `ptr' specifies the pointer to the region of a key.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the index of a corresponding element or -1 if there is no corresponding
+ element. If two or more elements corresponds, which returns is not defined. */
+int cblistbsearch(const CBLIST *list, const char *ptr, int size);
+
+
+/* Serialize a list into a byte array.
+ `list' specifies a list handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned.
+ The return value is the pointer to the region of the result serial region.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cblistdump(const CBLIST *list, int *sp);
+
+
+/* Redintegrate a serialized list.
+ `ptr' specifies the pointer to a byte array.
+ `size' specifies the size of the region.
+ The return value is a new list handle. */
+CBLIST *cblistload(const char *ptr, int size);
+
+
+/* Get a map handle.
+ The return value is a map handle. */
+CBMAP *cbmapopen(void);
+
+
+/* Copy a map.
+ `map' specifies a map handle.
+ The return value is a new map handle.
+ The iterator of the source map is initialized. */
+CBMAP *cbmapdup(CBMAP *map);
+
+
+/* Close a map handle.
+ `map' specifies a map handle.
+ Because the region of a closed handle is released, it becomes impossible to use the handle. */
+void cbmapclose(CBMAP *map);
+
+
+/* Store a record into a map.
+ `map' specifies a map handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ `over' specifies whether the value of the duplicated record is overwritten or not.
+ If `over' is false and the key is duplicated, the return value is false, else, it is true. */
+int cbmapput(CBMAP *map, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int over);
+
+
+/* Concatenate a value at the end of the value of the existing record.
+ `map' specifies a map handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ If there is no corresponding record, a new record is created. */
+void cbmapputcat(CBMAP *map, const char *kbuf, int ksiz, const char *vbuf, int vsiz);
+
+
+/* Delete a record in a map.
+ `map' specifies a map handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is true. False is returned when no record corresponds to
+ the specified key. */
+int cbmapout(CBMAP *map, const char *kbuf, int ksiz);
+
+
+/* Retrieve a record in a map.
+ `map' specifies a map handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record. `NULL' is returned when no record corresponds.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. */
+const char *cbmapget(const CBMAP *map, const char *kbuf, int ksiz, int *sp);
+
+
+/* Move a record to the edge of a map.
+ `map' specifies a map handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `head' specifies the destination which is head if it is true or tail if else.
+ If successful, the return value is true. False is returned when no record corresponds to
+ the specified key. */
+int cbmapmove(CBMAP *map, const char *kbuf, int ksiz, int head);
+
+
+/* Initialize the iterator of a map.
+ `map' specifies a map handle.
+ The iterator is used in order to access the key of every record stored in a map. */
+void cbmapiterinit(CBMAP *map);
+
+
+/* Get the next key of the iterator of a map.
+ `map' specifies a map handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the next key, else, it is
+ `NULL'. `NULL' is returned when no record is to be get out of the iterator.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. The order of iteration is assured
+ to be the same of the one of storing. */
+const char *cbmapiternext(CBMAP *map, int *sp);
+
+
+/* Get the value binded to the key fetched from the iterator of a map.
+ `kbuf' specifies the pointer to the region of a iteration key.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the value of the corresponding record.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. */
+const char *cbmapiterval(const char *kbuf, int *sp);
+
+
+/* Get the number of the records stored in a map.
+ `map' specifies a map handle.
+ The return value is the number of the records stored in the map. */
+int cbmaprnum(const CBMAP *map);
+
+
+/* Get the list handle contains all keys in a map.
+ `map' specifies a map handle.
+ The return value is the list handle contains all keys in the map.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *cbmapkeys(CBMAP *map);
+
+
+/* Get the list handle contains all values in a map.
+ `map' specifies a map handle.
+ The return value is the list handle contains all values in the map.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *cbmapvals(CBMAP *map);
+
+
+/* Serialize a map into a byte array.
+ `map' specifies a map handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned.
+ The return value is the pointer to the region of the result serial region.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbmapdump(CBMAP *map, int *sp);
+
+
+/* Redintegrate a serialized map.
+ `ptr' specifies the pointer to a byte array.
+ `size' specifies the size of the region.
+ The return value is a new map handle. */
+CBMAP *cbmapload(const char *ptr, int size);
+
+
+/* Extract a record from a serialized map.
+ `ptr' specifies the pointer to a byte array.
+ `size' specifies the size of the region.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record. `NULL' is returned when no record corresponds.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. */
+char *cbmaploadone(const char *ptr, int size, const char *kbuf, int ksiz, int *sp);
+
+
+/* Get a heap handle.
+ `size' specifies the size of each record.
+ `max' specifies the maximum number of records in the heap.
+ `compar' specifies the pointer to comparing function. The two arguments specify the pointers
+ of records. The comparing function should returns positive if the former is big, negative
+ if the latter is big, 0 if both are equal.
+ The return value is a heap handle. */
+CBHEAP *cbheapopen(int size, int max, int(*compar)(const void *, const void *));
+
+
+/* Copy a heap.
+ `heap' specifies a heap handle.
+ The return value is a new heap handle. */
+CBHEAP *cbheapdup(CBHEAP *heap);
+
+
+/* Close a heap handle.
+ `heap' specifies a heap handle.
+ Because the region of a closed handle is released, it becomes impossible to use the handle. */
+void cbheapclose(CBHEAP *heap);
+
+
+/* Get the number of the records stored in a heap.
+ `heap' specifies a heap handle.
+ The return value is the number of the records stored in the heap. */
+int cbheapnum(CBHEAP *heap);
+
+
+/* Insert a record into a heap.
+ `heap' specifies a heap handle.
+ `ptr' specifies the pointer to the region of a record.
+ The return value is true if the record is added, else false.
+ If the new record is bigger than the biggest existing regord, the new record is not added.
+ If the new record is added and the number of records exceeds the maximum number, the biggest
+ existing record is removed. */
+int cbheapinsert(CBHEAP *heap, const void *ptr);
+
+
+/* Get the pointer to the region of a record in a heap.
+ `heap' specifies a heap handle.
+ `index' specifies the index of a record.
+ The return value is the pointer to the region of the record.
+ If `index' is equal to or more than the number of records, the return value is `NULL'. Note
+ that records are organized by the nagative order the comparing function. */
+const void *cbheapval(CBHEAP *heap, int index);
+
+
+/* Convert a heap to an allocated region.
+ `heap' specifies a heap handle.
+ `np' specifies the pointer to a variable to which the number of records of the return value
+ is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the heap. Records are sorted.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. Because the region of the original
+ heap is released, it should not be released again. */
+void *cbheaptomalloc(CBHEAP *heap, int *np);
+
+
+/* Allocate a formatted string on memory.
+ `format' specifies a printf-like format string. The conversion character `%' can be used
+ with such flag characters as `d', `o', `u', `x', `X', `e', `E', `f', `g', `G', `c', `s', and
+ `%'. Specifiers of the field length and the precision can be put between the conversion
+ characters and the flag characters. The specifiers consist of decimal characters, `.', `+',
+ `-', and the space character.
+ The other arguments are used according to the format string.
+ The return value is the pointer to the allocated region of the result string. Because the
+ region of the return value is allocated with the `malloc' call, it should be released with
+ the `free' call if it is no longer in use. */
+char *cbsprintf(const char *format, ...);
+
+
+/* Replace some patterns in a string.
+ `str' specifies the pointer to a source string.
+ `pairs' specifies the handle of a map composed of pairs of replacement. The key of each pair
+ specifies a pattern before replacement and its value specifies the pattern after replacement.
+ The return value is the pointer to the allocated region of the result string.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbreplace(const char *str, CBMAP *pairs);
+
+
+/* Make a list by splitting a serial datum.
+ `ptr' specifies the pointer to the region of the source content.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `delim' specifies a string containing delimiting characters. If it is `NULL', zero code is
+ used as a delimiter.
+ The return value is a list handle.
+ If two delimiters are successive, it is assumed that an empty element is between the two.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose'. */
+CBLIST *cbsplit(const char *ptr, int size, const char *delim);
+
+
+/* Read whole data of a file.
+ `name' specifies the name of a file. If it is `NULL', the standard input is specified.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the allocated region of the read data. Because an
+ additional zero code is appended at the end of the region of the return value, the return
+ value can be treated as a character string. Because the region of the return value is
+ allocated with the `malloc' call, it should be released with the `free' call if it is no
+ longer in use. */
+char *cbreadfile(const char *name, int *sp);
+
+
+/* Write a serial datum into a file.
+ `name specifies the name of a file. If it is `NULL', the standard output is specified.
+ `ptr' specifies the pointer to the region of the source content.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ If successful, the return value is true, else, it is false.
+ If the file exists, it is overwritten. Else, a new file is created. */
+int cbwritefile(const char *name, const char *ptr, int size);
+
+
+/* Read every line of a file.
+ `name' specifies the name of a file. If it is `NULL', the standard input is specified.
+ The return value is a list handle of the lines if successful, else it is NULL. Line
+ separators are cut out. Because the handle of the return value is opened with the function
+ `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *cbreadlines(const char *name);
+
+
+/* Read names of files in a directory.
+ `name' specifies the name of a directory.
+ The return value is a list handle of names if successful, else it is NULL.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *cbdirlist(const char *name);
+
+
+/* Get the status of a file or a directory.
+ `name' specifies the name of a file or a directory.
+ `dirp' specifies the pointer to a variable to which whether the file is a directory is
+ assigned. If it is `NULL', it is not used.
+ `sizep' specifies the pointer to a variable to which the size of the file is assigned. If it
+ is `NULL', it is not used.
+ `mtimep' specifies the pointer to a variable to which the last modified time of the file is
+ assigned. If it is `NULL', it is not used.
+ If successful, the return value is true, else, false. False is returned when the file does
+ not exist or the permission is denied. */
+int cbfilestat(const char *name, int *isdirp, int *sizep, time_t *mtimep);
+
+
+/* Remove a file or a directory and its sub ones recursively.
+ `name' specifies the name of a file or a directory.
+ If successful, the return value is true, else, false. False is returned when the file does
+ not exist or the permission is denied. */
+int cbremove(const char *name);
+
+
+/* Break up a URL into elements.
+ `str' specifies the pointer to a string of URL.
+ The return value is a map handle. Each key of the map is the name of an element. The key
+ "self" specifies the URL itself. The key "scheme" specifies the scheme. The key "host"
+ specifies the host of the server. The key "port" specifies the port number of the server.
+ The key "authority" specifies the authority information. The key "path" specifies the path
+ of the resource. The key "file" specifies the file name without the directory section. The
+ key "query" specifies the query string. The key "fragment" specifies the fragment string.
+ Supported schema are HTTP, HTTPS, FTP, and FILE. Absolute URL and relative URL are supported.
+ Because the handle of the return value is opened with the function `cbmapopen', it should
+ be closed with the function `cbmapclose' if it is no longer in use. */
+CBMAP *cburlbreak(const char *str);
+
+
+/* Resolve a relative URL with another absolute URL.
+ `base' specifies an absolute URL of a base location.
+ `target' specifies a URL to be resolved.
+ The return value is a resolved URL. If the target URL is relative, a new URL of relative
+ location from the base location is returned. Else, a copy of the target URL is returned.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cburlresolve(const char *base, const char *target);
+
+
+/* Encode a serial object with URL encoding.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the pointer to the result string.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cburlencode(const char *ptr, int size);
+
+
+/* Decode a string encoded with URL encoding.
+ `str' specifies the pointer to an encoded string.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the result.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if
+ it is no longer in use. */
+char *cburldecode(const char *str, int *sp);
+
+
+/* Encode a serial object with Base64 encoding.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the pointer to the result string.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbbaseencode(const char *ptr, int size);
+
+
+/* Decode a string encoded with Base64 encoding.
+ `str' specifies the pointer to an encoded string.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the result.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if
+ it is no longer in use. */
+char *cbbasedecode(const char *str, int *sp);
+
+
+/* Encode a serial object with quoted-printable encoding.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the pointer to the result string.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbquoteencode(const char *ptr, int size);
+
+
+/* Decode a string encoded with quoted-printable encoding.
+ `str' specifies the pointer to an encoded string.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer to the region of the result.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if
+ it is no longer in use. */
+char *cbquotedecode(const char *str, int *sp);
+
+
+/* Split a string of MIME into headers and the body.
+ `ptr' specifies the pointer to the region of MIME data.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `attrs' specifies a map handle to store attributes. If it is `NULL', it is not used. Each
+ key of the map is an attribute name uncapitalized.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ The return value is the pointer of the body data.
+ If the content type is defined, the attribute map has the key "TYPE" specifying the type. If
+ the character encoding is defined, the key "CHARSET" specifies the encoding name. If the
+ boundary string of multipart is defined, the key "BOUNDARY" specifies the string. If the
+ content disposition is defined, the key "DISPOSITION" specifies the direction. If the file
+ name is defined, the key "FILENAME" specifies the name. If the attribute name is defined,
+ the key "NAME" specifies the name. Because the region of the return value is allocated with
+ the `malloc' call, it should be released with the `free' call if it is no longer in use. */
+char *cbmimebreak(const char *ptr, int size, CBMAP *attrs, int *sp);
+
+
+/* Split multipart data of MIME into its parts.
+ `ptr' specifies the pointer to the region of multipart data of MIME.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `boundary' specifies the pointer to the region of the boundary string.
+ The return value is a list handle. Each element of the list is the string of a part.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *cbmimeparts(const char *ptr, int size, const char *boundary);
+
+
+/* Encode a string with MIME encoding.
+ `str' specifies the pointer to a string.
+ `encname' specifies a string of the name of the character encoding.
+ The return value is the pointer to the result string.
+ `base' specifies whether to use Base64 encoding. If it is false, quoted-printable is used.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbmimeencode(const char *str, const char *encname, int base);
+
+
+/* Decode a string encoded with MIME encoding.
+ `str' specifies the pointer to an encoded string.
+ `enp' specifies the pointer to a region into which the name of encoding is written. If it is
+ `NULL', it is not used. The size of the buffer should be equal to or more than 32 bytes.
+ The return value is the pointer to the result string.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbmimedecode(const char *str, char *enp);
+
+
+/* Split a string of CSV into rows.
+ `str' specifies the pointer to the region of an CSV string.
+ The return value is a list handle. Each element of the list is a string of a row.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. The character encoding
+ of the input string should be US-ASCII, UTF-8, ISO-8859-*, EUC-*, or Shift_JIS. Being
+ compatible with MS-Excel, these functions for CSV can handle cells including such meta
+ characters as comma, between double quotation marks. */
+CBLIST *cbcsvrows(const char *str);
+
+
+/* Split the string of a row of CSV into cells.
+ `str' specifies the pointer to the region of a row of CSV.
+ The return value is a list handle. Each element of the list is the unescaped string of a
+ cell of the row.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *cbcsvcells(const char *str);
+
+
+/* Escape a string with the meta characters of CSV.
+ `str' specifies the pointer to the region of a string.
+ The return value is the pointer to the escaped string sanitized of meta characters.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbcsvescape(const char *str);
+
+
+/* Unescape a string with the escaped meta characters of CSV.
+ `str' specifies the pointer to the region of a string with meta characters.
+ The return value is the pointer to the unescaped string.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbcsvunescape(const char *str);
+
+
+/* Split a string of XML into tags and text sections.
+ `str' specifies the pointer to the region of an XML string.
+ `cr' specifies whether to remove comments.
+ The return value is a list handle. Each element of the list is the string of a tag or a
+ text section.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. The character encoding
+ of the input string should be US-ASCII, UTF-8, ISO-8859-*, EUC-*, or Shift_JIS. Because
+ these functions for XML are not XML parser with validation check, it can handle also HTML
+ and SGML. */
+CBLIST *cbxmlbreak(const char *str, int cr);
+
+
+/* Get the map of attributes of an XML tag.
+ `str' specifies the pointer to the region of a tag string.
+ The return value is a map handle. Each key of the map is the name of an attribute. Each
+ value is unescaped. You can get the name of the tag with the key of an empty string.
+ Because the handle of the return value is opened with the function `cbmapopen', it should
+ be closed with the function `cbmapclose' if it is no longer in use. */
+CBMAP *cbxmlattrs(const char *str);
+
+
+/* Escape a string with the meta characters of XML.
+ `str' specifies the pointer to the region of a string.
+ The return value is the pointer to the escaped string sanitized of meta characters.
+ This function converts only `&', `<', `>', and `"'. Because the region of the return value
+ is allocated with the `malloc' call, it should be released with the `free' call if it is no
+ longer in use. */
+char *cbxmlescape(const char *str);
+
+
+/* Unescape a string with the entity references of XML.
+ `str' specifies the pointer to the region of a string with meta characters.
+ The return value is the pointer to the unescaped string.
+ This function restores only `&', `<', `>', and `"'. Because the region of the
+ return value is allocated with the `malloc' call, it should be released with the `free' call
+ if it is no longer in use. */
+char *cbxmlunescape(const char *str);
+
+
+/* Compress a serial object with ZLIB.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. This function is available only if
+ QDBM was built with ZLIB enabled. */
+char *cbdeflate(const char *ptr, int size, int *sp);
+
+
+/* Decompress a serial object compressed with ZLIB.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. This function is available only if QDBM was built with ZLIB enabled. */
+char *cbinflate(const char *ptr, int size, int *sp);
+
+
+/* Compress a serial object with GZIP.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. This function is available only if
+ QDBM was built with ZLIB enabled. */
+char *cbgzencode(const char *ptr, int size, int *sp);
+
+
+/* Decompress a serial object compressed with GZIP.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. This function is available only if QDBM was built with ZLIB enabled. */
+char *cbgzdecode(const char *ptr, int size, int *sp);
+
+
+/* Get the CRC32 checksum of a serial object.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the CRC32 checksum of the object.
+ This function is available only if QDBM was built with ZLIB enabled. */
+unsigned int cbgetcrc(const char *ptr, int size);
+
+
+/* Compress a serial object with LZO.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. This function is available only if
+ QDBM was built with LZO enabled. */
+char *cblzoencode(const char *ptr, int size, int *sp);
+
+
+/* Decompress a serial object compressed with LZO.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. This function is available only if QDBM was built with LZO enabled. */
+char *cblzodecode(const char *ptr, int size, int *sp);
+
+
+/* Compress a serial object with BZIP2.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. This function is available only if
+ QDBM was built with LZO enabled. */
+char *cbbzencode(const char *ptr, int size, int *sp);
+
+
+/* Decompress a serial object compressed with BZIP2.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. This function is available only if QDBM was built with LZO enabled. */
+char *cbbzdecode(const char *ptr, int size, int *sp);
+
+
+/* Convert the character encoding of a string.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ `icode' specifies the name of encoding of the input string.
+ `ocode' specifies the name of encoding of the output string.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ `mp' specifies the pointer to a variable to which the number of missing characters by failure
+ of conversion is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the result object, else, it is `NULL'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. This function is available only if QDBM was built with ICONV enabled. */
+char *cbiconv(const char *ptr, int size, const char *icode, const char *ocode, int *sp, int *mp);
+
+
+/* Detect the encoding of a string automatically.
+ `ptr' specifies the pointer to a region.
+ `size' specifies the size of the region. If it is negative, the size is assigned with
+ `strlen(ptr)'.
+ The return value is the string of the encoding name of the string.
+ As it stands, US-ASCII, ISO-2022-JP, Shift_JIS, CP932, EUC-JP, UTF-8, UTF-16, UTF-16BE,
+ and UTF-16LE are supported. If none of them matches, ISO-8859-1 is selected. This function
+ is available only if QDBM was built with ICONV enabled. */
+const char *cbencname(const char *ptr, int size);
+
+
+/* Get the jet lag of the local time in seconds.
+ The return value is the jet lag of the local time in seconds. */
+int cbjetlag(void);
+
+
+/* Get the Gregorian calendar of a time.
+ `t' specifies a source time. If it is negative, the current time is specified.
+ `jl' specifies the jet lag of a location in seconds.
+ `yearp' specifies the pointer to a variable to which the year is assigned. If it is `NULL',
+ it is not used.
+ `monp' specifies the pointer to a variable to which the month is assigned. If it is `NULL',
+ it is not used. 1 means January and 12 means December.
+ `dayp' specifies the pointer to a variable to which the day of the month is assigned. If it
+ is `NULL', it is not used.
+ `hourp' specifies the pointer to a variable to which the hours is assigned. If it is `NULL',
+ it is not used.
+ `minp' specifies the pointer to a variable to which the minutes is assigned. If it is `NULL',
+ it is not used.
+ `secp' specifies the pointer to a variable to which the seconds is assigned. If it is `NULL',
+ it is not used. */
+void cbcalendar(time_t t, int jl, int *yearp, int *monp, int *dayp,
+ int *hourp, int *minp, int *secp);
+
+
+/* Get the day of week of a date.
+ `year' specifies the year of a date.
+ `mon' specifies the month of the date.
+ `day' specifies the day of the date.
+ The return value is the day of week of the date. 0 means Sunday and 6 means Saturday. */
+int cbdayofweek(int year, int mon, int day);
+
+
+/* Get the string for a date in W3CDTF.
+ `t' specifies a source time. If it is negative, the current time is specified.
+ `jl' specifies the jet lag of a location in seconds.
+ The return value is the string of the date in W3CDTF (YYYY-MM-DDThh:mm:ddTZD).
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbdatestrwww(time_t t, int jl);
+
+
+/* Get the string for a date in RFC 1123 format.
+ `t' specifies a source time. If it is negative, the current time is specified.
+ `jl' specifies the jet lag of a location in seconds.
+ The return value is the string of the date in RFC 1123 format (Wdy, DD-Mon-YYYY hh:mm:dd TZD).
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *cbdatestrhttp(time_t t, int jl);
+
+
+/* Get the time value of a date string in decimal, hexadecimal, W3CDTF, or RFC 822 (1123).
+ `str' specifies a date string in decimal, hexadecimal, W3CDTF, or RFC 822 (1123).
+ The return value is the time value of the date or -1 if the format is invalid.
+ Decimal can be trailed by "s" for in seconds, "m" for in minutes, "h" for in hours,
+ and "d" for in days. */
+time_t cbstrmktime(const char *str);
+
+
+/* Get user and system processing times.
+ `usrp' specifies the pointer to a variable to which the user processing time is assigned.
+ If it is `NULL', it is not used. The unit of time is seconds.
+ `sysp' specifies the pointer to a variable to which the system processing time is assigned.
+ If it is `NULL', it is not used. The unit of time is seconds. */
+void cbproctime(double *usrp, double *sysp);
+
+
+/* Ensure that the standard I/O is binary mode.
+ This function is useful for applications on dosish file systems. */
+void cbstdiobin(void);
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Show error message on the standard error output and exit.
+ `message' specifies an error message.
+ This function does not return. */
+void *cbmyfatal(const char *message);
+
+
+/* Create a datum handle from an allocated region.
+ `ptr' specifies the pointer to the region of an element. The region should be allocated with
+ malloc and it is released by the function.
+ `size' specifies the size of the region. */
+CBDATUM *cbdatumopenbuf(char *ptr, int size);
+
+
+/* Set a buffer to a datum handle.
+ `ptr' specifies the pointer to the region of an element. The region should be allocated with
+ malloc and it is released by the function.
+ `size' specifies the size of the region. */
+void cbdatumsetbuf(CBDATUM *datum, char *ptr, int size);
+
+
+/* Add an allocated element at the end of a list.
+ `list' specifies a list handle.
+ `ptr' specifies the pointer to the region of an element. The region should be allocated with
+ malloc and it is released by the function.
+ `size' specifies the size of the region. */
+void cblistpushbuf(CBLIST *list, char *ptr, int size);
+
+
+/* Get a map handle with specifying the number of buckets.
+ `bnum' specifies the number of buckets.
+ The return value is a map handle. */
+CBMAP *cbmapopenex(int bnum);
+
+
+/* Alias of `cbmalloc'. */
+#define CB_MALLOC(CB_ptr, CB_size) \
+ (((CB_ptr) = malloc(CB_size)) ? (CB_ptr) : cbmyfatal("out of memory"))
+
+
+/* Alias of `cbrealloc'. */
+#define CB_REALLOC(CB_ptr, CB_size) \
+ (((CB_ptr) = realloc((CB_ptr), (CB_size))) ? (CB_ptr) : cbmyfatal("out of memory"))
+
+
+/* Alias of `cbmemdup'.
+ However, `size' should not be negative. */
+#define CB_MEMDUP(CB_res, CB_ptr, CB_size) \
+ do { \
+ CB_MALLOC((CB_res), (CB_size) + 1); \
+ memcpy((CB_res), (CB_ptr), (CB_size)); \
+ (CB_res)[(CB_size)] = '\0'; \
+ } while(FALSE)
+
+
+/* Get the size of padding bytes for pointer alignment.
+ `hsiz' specifies the header size of the object.
+ The return value is the size of padding bytes. */
+#define CB_ALIGNPAD(CB_hsiz) \
+ (((CB_hsiz | ~-(int)sizeof(void *)) + 1) - CB_hsiz)
+
+
+/* Alias of `cbdatumopen'.
+ However, no dafault data is specified. */
+#define CB_DATUMOPEN(CB_datum) \
+ do { \
+ CB_MALLOC((CB_datum), sizeof(*(CB_datum))); \
+ CB_MALLOC((CB_datum)->dptr, CB_DATUMUNIT); \
+ (CB_datum)->dptr[0] = '\0'; \
+ (CB_datum)->dsize = 0; \
+ (CB_datum)->asize = CB_DATUMUNIT; \
+ } while(FALSE)
+
+
+/* Alias of `cbdatumopen'.
+ However, `size' should not be negative. */
+#define CB_DATUMOPEN2(CB_datum, CB_ptr, CB_size) \
+ do { \
+ CB_DATUMOPEN((CB_datum)); \
+ CB_DATUMCAT((CB_datum), (CB_ptr), (CB_size)); \
+ } while(FALSE)
+
+
+/* Alias of `cbdatumclose'. */
+#define CB_DATUMCLOSE(CB_datum) \
+ do { \
+ free((CB_datum)->dptr); \
+ free((CB_datum)); \
+ } while(FALSE)
+
+
+/* Alias of `cbdatumcat'.
+ However, `size' should not be negative. */
+#define CB_DATUMCAT(CB_datum, CB_ptr, CB_size) \
+ do { \
+ if((CB_datum)->dsize + (CB_size) >= (CB_datum)->asize){ \
+ (CB_datum)->asize = (CB_datum)->asize * 2 + (CB_size) + 1; \
+ CB_REALLOC((CB_datum)->dptr, (CB_datum)->asize); \
+ } \
+ memcpy((CB_datum)->dptr + (CB_datum)->dsize, (CB_ptr), (CB_size)); \
+ (CB_datum)->dsize += (CB_size); \
+ (CB_datum)->dptr[(CB_datum)->dsize] = '\0'; \
+ } while(FALSE)
+
+
+/* Alias of `cbdatumptr'. */
+#define CB_DATUMPTR(CB_datum) ((const char *)((CB_datum)->dptr))
+
+
+/* Alias of `cbdatumsize'. */
+#define CB_DATUMSIZE(CB_datum) ((int)((CB_datum)->dsize))
+
+
+/* Alias of `cbdatumsetsize'. */
+#define CB_DATUMSETSIZE(CB_datum, CB_size) \
+ do { \
+ if((CB_size) <= (CB_datum)->dsize){ \
+ (CB_datum)->dsize = (CB_size); \
+ (CB_datum)->dptr[(CB_size)] = '\0'; \
+ } else { \
+ if((CB_size) >= (CB_datum)->asize){ \
+ (CB_datum)->asize = (CB_datum)->asize * 2 + (CB_size) + 1; \
+ CB_REALLOC((CB_datum)->dptr, (CB_datum)->asize); \
+ } \
+ memset((CB_datum)->dptr + (CB_datum)->dsize, 0, ((CB_size) - (CB_datum)->dsize) + 1); \
+ (CB_datum)->dsize = (CB_size); \
+ } \
+ } while(FALSE)
+
+
+/* Alias of `cbdatumtomalloc'. */
+#define CB_DATUMTOMALLOC(CB_datum, CB_ptr, CB_size) \
+ do { \
+ (CB_ptr) = (CB_datum)->dptr; \
+ (CB_size) = (CB_datum)->dsize; \
+ free((CB_datum)); \
+ } while(FALSE)
+
+
+/* Alias of `cblistopen'. */
+#define CB_LISTOPEN(CB_list) \
+ do { \
+ CB_MALLOC((CB_list), sizeof(*(CB_list))); \
+ (CB_list)->anum = CB_LISTUNIT; \
+ CB_MALLOC((CB_list)->array, sizeof((CB_list)->array[0]) * (CB_list)->anum); \
+ (CB_list)->start = 0; \
+ (CB_list)->num = 0; \
+ } while(FALSE)
+
+
+/* Alias of `cblistopen'.
+ However, `anum' is specified for the number of initial allocated elements. */
+#define CB_LISTOPEN2(CB_list, CB_anum) \
+ do { \
+ CB_MALLOC((CB_list), sizeof(*(CB_list))); \
+ (CB_list)->anum = (CB_anum) > 4 ? (CB_anum) : 4; \
+ CB_MALLOC((CB_list)->array, sizeof((CB_list)->array[0]) * (CB_list)->anum); \
+ (CB_list)->start = 0; \
+ (CB_list)->num = 0; \
+ } while(FALSE)
+
+
+/* Alias of `cblistclose'. */
+#define CB_LISTCLOSE(CB_list) \
+ do { \
+ int _CB_i, _CB_end; \
+ _CB_end = (CB_list)->start + (CB_list)->num; \
+ for(_CB_i = (CB_list)->start; _CB_i < _CB_end; _CB_i++){ \
+ free((CB_list)->array[_CB_i].dptr); \
+ } \
+ free((CB_list)->array); \
+ free((CB_list)); \
+ } while(FALSE)
+
+
+/* Alias of `cblistnum'. */
+#define CB_LISTNUM(CB_list) \
+ ((int)((CB_list)->num))
+
+
+/* Alias of `cblistval'.
+ However, `sp' is ignored. */
+#define CB_LISTVAL(CB_list, CB_index) \
+ ((const char *)((CB_list)->array[(CB_list)->start+(CB_index)].dptr))
+
+
+/* Alias of `cblistval'.
+ However, `size' is used instead of `sp'. */
+#define CB_LISTVAL2(CB_list, CB_index, CB_size) \
+ ((CB_size) = (CB_list)->array[(CB_list)->start+(CB_index)].dsize, \
+ (const char *)((CB_list)->array[(CB_list)->start+(CB_index)].dptr))
+
+
+/* Alias of `cblistpush'.
+ However, `size' should not be negative. */
+#define CB_LISTPUSH(CB_list, CB_ptr, CB_size) \
+ do { \
+ int _CB_index; \
+ _CB_index = (CB_list)->start + (CB_list)->num; \
+ if(_CB_index >= (CB_list)->anum){ \
+ (CB_list)->anum *= 2; \
+ CB_REALLOC((CB_list)->array, (CB_list)->anum * sizeof((CB_list)->array[0])); \
+ } \
+ CB_MALLOC((CB_list)->array[_CB_index].dptr, \
+ ((CB_size) < CB_DATUMUNIT ? CB_DATUMUNIT : (CB_size)) + 1); \
+ memcpy((CB_list)->array[_CB_index].dptr, (CB_ptr), (CB_size)); \
+ (CB_list)->array[_CB_index].dptr[(CB_size)] = '\0'; \
+ (CB_list)->array[_CB_index].dsize = (CB_size); \
+ (CB_list)->num++; \
+ } while(FALSE)
+
+
+/* Remove and free an element of the end of a list.
+ `list' specifies a list handle. */
+#define CB_LISTDROP(CB_list) \
+ do { \
+ if((CB_list)->num > 0){ \
+ free((CB_list)->array[(CB_list)->start+(CB_list)->num-1].dptr); \
+ (CB_list)->num--; \
+ } \
+ } while(FALSE)
+
+
+/* Alias of `cblistinsert'.
+ However, `index' is not checked and `size' should not be negative. */
+#define CB_LISTINSERT(CB_list, CB_index, CB_ptr, CB_size) \
+ do { \
+ int _CB_index = (CB_index); \
+ _CB_index += (CB_list)->start; \
+ if((CB_list)->start + (CB_list)->num >= (CB_list)->anum){ \
+ (CB_list)->anum *= 2; \
+ CB_REALLOC((CB_list)->array, (CB_list)->anum * sizeof((CB_list)->array[0])); \
+ } \
+ memmove((CB_list)->array + _CB_index + 1, (CB_list)->array + _CB_index, \
+ sizeof((CB_list)->array[0]) * ((CB_list)->start + (CB_list)->num - _CB_index)); \
+ CB_MEMDUP((CB_list)->array[_CB_index].dptr, (CB_ptr), (CB_size)); \
+ (CB_list)->array[_CB_index].dsize = (CB_size); \
+ (CB_list)->num++; \
+ } while(FALSE)
+
+
+/* Alias of `cblistpushbuf'. */
+#define CB_LISTPUSHBUF(CB_list, CB_ptr, CB_size) \
+ do{ \
+ int _CB_index; \
+ _CB_index = (CB_list)->start + (CB_list)->num; \
+ if(_CB_index >= (CB_list)->anum){ \
+ (CB_list)->anum *= 2; \
+ CB_REALLOC((CB_list)->array, (CB_list)->anum * sizeof((CB_list)->array[0])); \
+ } \
+ (CB_list)->array[_CB_index].dptr = (CB_ptr); \
+ (CB_list)->array[_CB_index].dsize = (CB_size); \
+ (CB_list)->num++; \
+ } while(FALSE) \
+
+
+/* Alias of `cbmapiterval'.
+ However, `size' is used instead of `sp'. */
+#define CB_MAPITERVAL(CB_vbuf, CB_kbuf, CB_vsiz) \
+ do { \
+ CBMAPDATUM *_CB_datum; \
+ _CB_datum = (CBMAPDATUM *)((CB_kbuf) - sizeof(*_CB_datum)); \
+ (CB_vsiz) = _CB_datum->vsiz; \
+ (CB_vbuf) = (char *)_CB_datum + sizeof(*_CB_datum) + \
+ _CB_datum->ksiz + CB_ALIGNPAD(_CB_datum->ksiz); \
+ } while(FALSE)
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/cabin.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/cbcodec.c
===================================================================
--- box/trunk/qdbm/cbcodec.c (rev 0)
+++ box/trunk/qdbm/cbcodec.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,1079 @@
+/*************************************************************************************************
+ * Popular encoders and decoders
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <cabin.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define DEFCODE "UTF-8" /* default encoding */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *readstdin(int *sp);
+int runurl(int argc, char **argv);
+int runbase(int argc, char **argv);
+int runquote(int argc, char **argv);
+int runmime(int argc, char **argv);
+int runcsv(int argc, char **argv);
+int runxml(int argc, char **argv);
+int runzlib(int argc, char **argv);
+int runlzo(int argc, char **argv);
+int runbzip(int argc, char **argv);
+int runiconv(int argc, char **argv);
+int rundate(int argc, char **argv);
+void shouucsmap(void);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "url")){
+ rv = runurl(argc, argv);
+ } else if(!strcmp(argv[1], "base")){
+ rv = runbase(argc, argv);
+ } else if(!strcmp(argv[1], "quote")){
+ rv = runquote(argc, argv);
+ } else if(!strcmp(argv[1], "mime")){
+ rv = runmime(argc, argv);
+ } else if(!strcmp(argv[1], "csv")){
+ rv = runcsv(argc, argv);
+ } else if(!strcmp(argv[1], "xml")){
+ rv = runxml(argc, argv);
+ } else if(!strcmp(argv[1], "zlib")){
+ rv = runzlib(argc, argv);
+ } else if(!strcmp(argv[1], "lzo")){
+ rv = runlzo(argc, argv);
+ } else if(!strcmp(argv[1], "bzip")){
+ rv = runbzip(argc, argv);
+ } else if(!strcmp(argv[1], "iconv")){
+ rv = runiconv(argc, argv);
+ } else if(!strcmp(argv[1], "date")){
+ rv = rundate(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ char *tmp;
+ int tsiz;
+ fprintf(stderr, "%s: popular encoders and decoders\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s url [-d] [-br] [-rs base target] [-l] [-e expr] [file]\n", progname);
+ fprintf(stderr, " %s base [-d] [-l] [-c num] [-e expr] [file]\n", progname);
+ fprintf(stderr, " %s quote [-d] [-l] [-c num] [-e expr] [file]\n", progname);
+ fprintf(stderr, " %s mime [-d] [-hd] [-bd] [-part num] [-l] [-ec code] [-qp] [-dc] [-e expr]"
+ " [file]\n", progname);
+ fprintf(stderr, " %s csv [-d] [-t] [-l] [-e expr] [-html] [file]\n", progname);
+ fprintf(stderr, " %s xml [-d] [-p] [-l] [-e expr] [-tsv] [file]\n", progname);
+ if((tmp = cbdeflate("", 0, &tsiz)) != NULL){
+ fprintf(stderr, " %s zlib [-d] [-gz] [-crc] [file]\n", progname);
+ free(tmp);
+ }
+ if((tmp = cblzoencode("", 0, &tsiz)) != NULL){
+ fprintf(stderr, " %s lzo [-d] [file]\n", progname);
+ free(tmp);
+ }
+ if((tmp = cbbzencode("", 0, &tsiz)) != NULL){
+ fprintf(stderr, " %s bzip [-d] [file]\n", progname);
+ free(tmp);
+ }
+ if((tmp = cbiconv("", 0, "US-ASCII", "US-ASCII", NULL, NULL)) != NULL){
+ fprintf(stderr, " %s iconv [-ic code] [-oc code] [-ol ltype] [-cn] [-wc] [-um] [file]\n",
+ progname);
+ free(tmp);
+ }
+ fprintf(stderr, " %s date [-wf] [-rf] [-utc] [str]\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* read the standard input */
+char *readstdin(int *sp){
+ char *buf;
+ int i, blen, c;
+ blen = 256;
+ buf = cbmalloc(blen);
+ for(i = 0; (c = getchar()) != EOF; i++){
+ if(i >= blen - 1) buf = cbrealloc(buf, blen *= 2);
+ buf[i] = c;
+ }
+ buf[i] = '\0';
+ if(sp) *sp = i;
+ return buf;
+}
+
+
+/* parse arguments of url command */
+int runurl(int argc, char **argv){
+ CBMAP *map;
+ int i, size, dec, br, line;
+ const char *val;
+ char *base, *target, *expr, *file, *buf, *res;
+ dec = FALSE;
+ br = FALSE;
+ line = FALSE;
+ base = NULL;
+ target = NULL;
+ expr = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-br")){
+ br = TRUE;
+ } else if(!strcmp(argv[i], "-rs")){
+ if(++i >= argc) usage();
+ base = argv[i];
+ if(++i >= argc) usage();
+ target = argv[i];
+ } else if(!strcmp(argv[i], "-l")){
+ line = TRUE;
+ } else if(!strcmp(argv[i], "-e")){
+ if(++i >= argc) usage();
+ expr = argv[i];
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(base){
+ size = strlen(base);
+ buf = cbmemdup(base, size);
+ } else if(expr){
+ size = strlen(expr);
+ buf = cbmemdup(expr, size);
+ } else if(file){
+ if(!(buf = cbreadfile(file, &size))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&size);
+ }
+ if(target){
+ res = cburlresolve(base, target);
+ printf("%s", res);
+ free(res);
+ } else if(br){
+ map = cburlbreak(buf);
+ if((val = cbmapget(map, "self", -1, NULL))) printf("self\t%s\n", val);
+ if((val = cbmapget(map, "scheme", -1, NULL))) printf("scheme\t%s\n", val);
+ if((val = cbmapget(map, "host", -1, NULL))) printf("host\t%s\n", val);
+ if((val = cbmapget(map, "port", -1, NULL))) printf("port\t%s\n", val);
+ if((val = cbmapget(map, "authority", -1, NULL))) printf("authority\t%s\n", val);
+ if((val = cbmapget(map, "path", -1, NULL))) printf("path\t%s\n", val);
+ if((val = cbmapget(map, "file", -1, NULL))) printf("file\t%s\n", val);
+ if((val = cbmapget(map, "query", -1, NULL))) printf("query\t%s\n", val);
+ if((val = cbmapget(map, "fragment", -1, NULL))) printf("fragment\t%s\n", val);
+ cbmapclose(map);
+ } else if(dec){
+ res = cburldecode(buf, &size);
+ for(i = 0; i < size; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ res = cburlencode(buf, size);
+ for(i = 0; res[i] != '\0'; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ }
+ if(line) putchar('\n');
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of base command */
+int runbase(int argc, char **argv){
+ int i, ci, size, dec, line, cols;
+ char *expr, *file, *buf, *res;
+ dec = FALSE;
+ line = FALSE;
+ cols = -1;
+ expr = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-l")){
+ line = TRUE;
+ } else if(!strcmp(argv[i], "-c")){
+ if(++i >= argc) usage();
+ cols = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-e")){
+ if(++i >= argc) usage();
+ expr = argv[i];
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(expr){
+ size = strlen(expr);
+ buf = cbmemdup(expr, size);
+ } else if(file){
+ if(!(buf = cbreadfile(file, &size))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&size);
+ }
+ if(dec){
+ res = cbbasedecode(buf, &size);
+ for(i = 0; i < size; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ res = cbbaseencode(buf, size);
+ ci = 0;
+ for(i = 0; res[i] != '\0'; i++){
+ if(cols > 0 && ci >= cols){
+ putchar('\n');
+ ci = 0;
+ }
+ putchar(res[i]);
+ ci++;
+ }
+ free(res);
+ }
+ if(line) putchar('\n');
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of quote command */
+int runquote(int argc, char **argv){
+ int i, ci, size, dec, line, cols;
+ char *expr, *file, *buf, *res;
+ dec = FALSE;
+ line = FALSE;
+ cols = -1;
+ expr = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-l")){
+ line = TRUE;
+ } else if(!strcmp(argv[i], "-c")){
+ if(++i >= argc) usage();
+ cols = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-e")){
+ if(++i >= argc) usage();
+ expr = argv[i];
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(expr){
+ size = strlen(expr);
+ buf = cbmemdup(expr, size);
+ } else if(file){
+ if(!(buf = cbreadfile(file, &size))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&size);
+ }
+ if(dec){
+ res = cbquotedecode(buf, &size);
+ for(i = 0; i < size; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ res = cbquoteencode(buf, size);
+ ci = 0;
+ for(i = 0; res[i] != '\0'; i++){
+ if(cols > 0 && (ci >= cols || (ci >= cols - 2 && res[i] == '='))){
+ printf("=\n");
+ ci = 0;
+ }
+ if(res[i] == '\r' || res[i] == '\n') ci = 0;
+ putchar(res[i]);
+ ci++;
+ }
+ free(res);
+ }
+ if(line) putchar('\n');
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of mime command */
+int runmime(int argc, char **argv){
+ CBMAP *attrs;
+ CBLIST *parts;
+ int i, size, dec, line, qp, dc, hd, bd, pnum, rsiz, bsiz;
+ const char *key, *body;
+ char *code, *expr, *file, *buf, *res, renc[64];
+ dec = FALSE;
+ hd = FALSE;
+ bd = FALSE;
+ pnum = 0;
+ line = FALSE;
+ dc = FALSE;
+ qp = FALSE;
+ code = NULL;
+ expr = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-hd")){
+ hd = TRUE;
+ } else if(!strcmp(argv[i], "-bd")){
+ bd = TRUE;
+ } else if(!strcmp(argv[i], "-part")){
+ if(++i >= argc) usage();
+ pnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-l")){
+ line = TRUE;
+ } else if(!strcmp(argv[i], "-ec")){
+ if(++i >= argc) usage();
+ code = argv[i];
+ } else if(!strcmp(argv[i], "-qp")){
+ qp = TRUE;
+ } else if(!strcmp(argv[i], "-dc")){
+ dc = TRUE;
+ } else if(!strcmp(argv[i], "-e")){
+ if(++i >= argc) usage();
+ expr = argv[i];
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(expr){
+ size = strlen(expr);
+ buf = cbmemdup(expr, size);
+ } else if(file){
+ if(!(buf = cbreadfile(file, &size))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&size);
+ }
+ if(hd || bd || pnum > 0){
+ attrs = cbmapopen();
+ res = cbmimebreak(buf, size, attrs, &rsiz);
+ if(pnum > 0){
+ parts = NULL;
+ if(!(key = cbmapget(attrs, "TYPE", -1, NULL)) || !cbstrfwimatch(key, "multipart/") ||
+ !(key = cbmapget(attrs, "BOUNDARY", -1, NULL)) ||
+ !(parts = cbmimeparts(res, rsiz, key)) || cblistnum(parts) < pnum){
+ fprintf(stderr, "%s: not multipart or no such part\n", progname);
+ if(parts) cblistclose(parts);
+ free(res);
+ cbmapclose(attrs);
+ free(buf);
+ return 1;
+ }
+ body = cblistval(parts, pnum - 1, &bsiz);
+ for(i = 0; i < bsiz; i++){
+ putchar(body[i]);
+ }
+ cblistclose(parts);
+ } else if(hd){
+ cbmapiterinit(attrs);
+ while((key = cbmapiternext(attrs, NULL)) != NULL){
+ printf("%s\t%s\n", key, cbmapget(attrs, key, -1, NULL));
+ }
+ } else {
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ }
+ free(res);
+ cbmapclose(attrs);
+ } else if(dec){
+ res = cbmimedecode(buf, renc);
+ printf("%s", dc ? renc : res);
+ free(res);
+ } else {
+ res = cbmimeencode(buf, code ? code : DEFCODE, !qp);
+ printf("%s", res);
+ free(res);
+ }
+ if(line) putchar('\n');
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of csv command */
+int runcsv(int argc, char **argv){
+ CBLIST *rows, *cells;
+ int i, j, k, dec, tb, line, html;
+ const char *row, *cell;
+ char *expr, *file, *buf, *res;
+ dec = FALSE;
+ tb = FALSE;
+ line = FALSE;
+ html = FALSE;
+ expr = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-t")){
+ tb = TRUE;
+ } else if(!strcmp(argv[i], "-l")){
+ line = TRUE;
+ } else if(!strcmp(argv[i], "-e")){
+ if(++i >= argc) usage();
+ expr = argv[i];
+ } else if(!strcmp(argv[i], "-html")){
+ html = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(expr){
+ buf = cbmemdup(expr, -1);
+ } else if(file){
+ if(!(buf = cbreadfile(file, NULL))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(NULL);
+ }
+ if(tb || html){
+ if(html) printf("<table border=\"1\">\n");
+ rows = cbcsvrows(buf);
+ for(i = 0; i < cblistnum(rows); i++){
+ if(html) printf("<tr>");
+ row = cblistval(rows, i, NULL);
+ cells = cbcsvcells(row);
+ for(j = 0; j < cblistnum(cells); j++){
+ cell = cblistval(cells, j, NULL);
+ if(html){
+ printf("<td>");
+ for(k = 0; cell[k] != '\0'; k++){
+ if(cell[k] == '\r' || cell[k] == '\n'){
+ printf("<br>");
+ if(cell[k] == '\r' && cell[k] == '\n') k++;
+ } else {
+ switch(cell[k]){
+ case '&': printf("&"); break;
+ case '<': printf("<"); break;
+ case '>': printf(">"); break;
+ default: putchar(cell[k]); break;
+ }
+ }
+ }
+ printf("</td>");
+ } else {
+ if(j > 0) putchar('\t');
+ for(k = 0; cell[k] != '\0'; k++){
+ if(((unsigned char *)cell)[k] >= 0x20) putchar(cell[k]);
+ }
+ }
+ }
+ cblistclose(cells);
+ if(html) printf("</tr>");
+ putchar('\n');
+ }
+ cblistclose(rows);
+ if(html) printf("</table>\n");
+ } else if(dec){
+ res = cbcsvunescape(buf);
+ for(i = 0; res[i] != '\0'; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ res = cbcsvescape(buf);
+ for(i = 0; res[i] != '\0'; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ }
+ if(line) putchar('\n');
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of xml command */
+int runxml(int argc, char **argv){
+ CBLIST *elems;
+ CBMAP *attrs;
+ int i, j, dec, pb, line, tsv, div;
+ const char *elem, *attr;
+ char *expr, *file, *buf, *res;
+ dec = FALSE;
+ pb = FALSE;
+ line = FALSE;
+ tsv = FALSE;
+ expr = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-p")){
+ pb = TRUE;
+ } else if(!strcmp(argv[i], "-l")){
+ line = TRUE;
+ } else if(!strcmp(argv[i], "-e")){
+ if(++i >= argc) usage();
+ expr = argv[i];
+ } else if(!strcmp(argv[i], "-tsv")){
+ tsv = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(expr){
+ buf = cbmemdup(expr, -1);
+ } else if(file){
+ if(!(buf = cbreadfile(file, NULL))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(NULL);
+ }
+ if(pb || tsv){
+ elems = cbxmlbreak(buf, FALSE);
+ for(i = 0; i < cblistnum(elems); i++){
+ elem = cblistval(elems, i, NULL);
+ div = FALSE;
+ if(elem[0] == '<'){
+ if(cbstrfwimatch(elem, "<?xml")){
+ printf("XMLDECL");
+ div = TRUE;
+ } else if(cbstrfwimatch(elem, "<!DOCTYPE")){
+ printf("DOCTYPE");
+ } else if(cbstrfwimatch(elem, "<!--")){
+ printf("COMMENT");
+ } else if(cbstrfwimatch(elem, "</")){
+ printf("ENDTAG");
+ div = TRUE;
+ } else if(cbstrbwimatch(elem, "/>")){
+ printf("EMPTAG");
+ div = TRUE;
+ } else {
+ printf("BEGTAG");
+ div = TRUE;
+ }
+ } else {
+ printf("TEXT");
+ }
+ putchar('\t');
+ if(tsv){
+ if(div){
+ attrs = cbxmlattrs(elem);
+ cbmapiterinit(attrs);
+ for(j = 0; (attr = cbmapiternext(attrs, NULL)) != NULL; j++){
+ if(j < 1){
+ printf("%s", cbmapget(attrs, attr, -1, NULL));
+ } else {
+ printf("\t%s\t%s", attr, cbmapget(attrs, attr, -1, NULL));
+ }
+ }
+ cbmapclose(attrs);
+ } else {
+ res = cbxmlunescape(elem);
+ for(j = 0; elem[j] != '\0'; j++){
+ if(((unsigned char *)elem)[j] < 0x20 || elem[j] == '%'){
+ printf("%%%02X", elem[j]);
+ } else {
+ putchar(elem[j]);
+ }
+ }
+ free(res);
+ }
+ } else {
+ printf("%s", elem);
+ }
+ putchar('\n');
+ }
+ cblistclose(elems);
+ } else if(dec){
+ res = cbxmlunescape(buf);
+ for(i = 0; res[i] != '\0'; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ res = cbxmlescape(buf);
+ for(i = 0; res[i] != '\0'; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ }
+ if(line) putchar('\n');
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of zlib command */
+int runzlib(int argc, char **argv){
+ unsigned int sum;
+ int i, bsiz, rsiz, dec, gz, crc;
+ char *file, *buf, *res;
+ dec = FALSE;
+ gz = FALSE;
+ crc = FALSE;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else if(!strcmp(argv[i], "-gz")){
+ gz = TRUE;
+ } else if(!strcmp(argv[i], "-crc")){
+ crc = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(file){
+ if(!(buf = cbreadfile(file, &bsiz))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&bsiz);
+ }
+ if(crc){
+ sum = cbgetcrc(buf, bsiz);
+ for(i = 0; i < sizeof(int); i++){
+ printf("%02x", sum / 0x1000000);
+ sum <<= 8;
+ }
+ putchar('\n');
+ } else if(dec){
+ if(!(res = gz ? cbgzdecode(buf, bsiz, &rsiz) : cbinflate(buf, bsiz, &rsiz))){
+ fprintf(stderr, "%s: inflate failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ if(!(res = gz ? cbgzencode(buf, bsiz, &rsiz) : cbdeflate(buf, bsiz, &rsiz))){
+ fprintf(stderr, "%s: deflate failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ }
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of lzo command */
+int runlzo(int argc, char **argv){
+ int i, bsiz, rsiz, dec;
+ char *file, *buf, *res;
+ dec = FALSE;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(file){
+ if(!(buf = cbreadfile(file, &bsiz))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&bsiz);
+ }
+ if(dec){
+ if(!(res = cblzodecode(buf, bsiz, &rsiz))){
+ fprintf(stderr, "%s: decode failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ if(!(res = cblzoencode(buf, bsiz, &rsiz))){
+ fprintf(stderr, "%s: encode failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ }
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of bzip command */
+int runbzip(int argc, char **argv){
+ int i, bsiz, rsiz, dec;
+ char *file, *buf, *res;
+ dec = FALSE;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ dec = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(file){
+ if(!(buf = cbreadfile(file, &bsiz))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&bsiz);
+ }
+ if(dec){
+ if(!(res = cbbzdecode(buf, bsiz, &rsiz))){
+ fprintf(stderr, "%s: decode failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ } else {
+ if(!(res = cbbzencode(buf, bsiz, &rsiz))){
+ fprintf(stderr, "%s: encode failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ free(res);
+ }
+ free(buf);
+ return 0;
+}
+
+
+/* parse arguments of iconv command */
+int runiconv(int argc, char **argv){
+ CBDATUM *datum;
+ const char *rcode;
+ char *icode, *ocode, *ltype, *file, *buf, *res, *norm, *orig;
+ int i, cn, wc, bsiz, rsiz, nsiz, osiz, miss;
+ icode = NULL;
+ ocode = NULL;
+ ltype = NULL;
+ file = NULL;
+ cn = FALSE;
+ wc = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-ic")){
+ if(++i >= argc) usage();
+ icode = argv[i];
+ } else if(!strcmp(argv[i], "-oc")){
+ if(++i >= argc) usage();
+ ocode = argv[i];
+ } else if(!strcmp(argv[i], "-ol")){
+ if(++i >= argc) usage();
+ ltype = argv[i];
+ } else if(!strcmp(argv[i], "-cn")){
+ cn = TRUE;
+ } else if(!strcmp(argv[i], "-wc")){
+ wc = TRUE;
+ } else if(!strcmp(argv[i], "-um")){
+ shouucsmap();
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ buf = NULL;
+ if(file){
+ if(!(buf = cbreadfile(file, &bsiz))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ buf = readstdin(&bsiz);
+ }
+ miss = 0;
+ if(cn){
+ printf("%s\n", cbencname(buf, bsiz));
+ } else if(wc){
+ printf("%d\n", cbstrcountutf(buf));
+ } else {
+ rcode = icode ? icode : cbencname(buf, bsiz);
+ if(!(res = cbiconv(buf, bsiz, rcode, ocode ? ocode : DEFCODE,
+ &rsiz, &miss))){
+ fprintf(stderr, "%s: iconv failed\n", progname);
+ free(buf);
+ return 1;
+ }
+ if(miss > 0) fprintf(stderr, "%s: missing %d characters\n", progname, miss);
+ if(ltype && (!cbstricmp(ltype, "u") || !cbstricmp(ltype, "unix") ||
+ !cbstricmp(ltype, "lf"))){
+ ltype = "\n";
+ } else if(ltype && (!cbstricmp(ltype, "d") || !cbstricmp(ltype, "dos") ||
+ !cbstricmp(ltype, "crlf"))){
+ ltype = "\r\n";
+ } else if(ltype && (!cbstricmp(ltype, "m") || !cbstricmp(ltype, "mac") ||
+ !cbstricmp(ltype, "cr"))){
+ ltype = "\r";
+ } else {
+ ltype = NULL;
+ }
+ if(ltype){
+ if(!(norm = cbiconv(res, rsiz, ocode, "UTF-8", &nsiz, NULL))){
+ fprintf(stderr, "%s: iconv failed\n", progname);
+ free(res);
+ free(buf);
+ return 1;
+ }
+ datum = cbdatumopen(NULL, -1);
+ for(i = 0; i < nsiz; i++){
+ if(norm[i] == '\r'){
+ if(norm[i+1] == '\n') i++;
+ cbdatumcat(datum, ltype, -1);
+ } else if(norm[i] == '\n'){
+ cbdatumcat(datum, ltype, -1);
+ } else {
+ cbdatumcat(datum, norm + i, 1);
+ }
+ }
+ if(!(orig = cbiconv(cbdatumptr(datum), cbdatumsize(datum), "UTF-8", ocode, &osiz, NULL))){
+ fprintf(stderr, "%s: iconv failed\n", progname);
+ cbdatumclose(datum);
+ free(norm);
+ free(res);
+ free(buf);
+ return 1;
+ }
+ for(i = 0; i < osiz; i++){
+ putchar(orig[i]);
+ }
+ free(orig);
+ cbdatumclose(datum);
+ free(norm);
+ } else {
+ for(i = 0; i < rsiz; i++){
+ putchar(res[i]);
+ }
+ }
+ free(res);
+ }
+ free(buf);
+ return miss > 0 ? 1 : 0;
+}
+
+
+/* parse arguments of date command */
+int rundate(int argc, char **argv){
+ int i, wb, rb, utc, jl;
+ char *date, *res;
+ time_t t;
+ wb = FALSE;
+ rb = FALSE;
+ utc = FALSE;
+ date = NULL;
+ for(i = 2; i < argc; i++){
+ if(!date && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-wf")){
+ wb = TRUE;
+ } else if(!strcmp(argv[i], "-rf")){
+ rb = TRUE;
+ } else if(!strcmp(argv[i], "-utc")){
+ utc = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!date){
+ date = argv[i];
+ } else {
+ usage();
+ }
+ }
+ jl = utc ? 0 : cbjetlag();
+ if(date){
+ t = cbstrmktime(date);
+ } else {
+ t = time(NULL);
+ }
+ if(wb){
+ res = cbdatestrwww(t, jl);
+ } else if(rb){
+ res = cbdatestrhttp(t, jl);
+ } else {
+ res = cbsprintf("%d", (int)t);
+ }
+ if(t >= 0){
+ printf("%s\n", res);
+ } else {
+ if(date){
+ fprintf(stderr, "%s: %s: invalid date format\n", progname, date);
+ } else {
+ fprintf(stderr, "%s: invalid time setting\n", progname);
+ }
+ }
+ free(res);
+ return 0;
+}
+
+
+/* show mapping of UCS-2 and exit. */
+void shouucsmap(void){
+ unsigned char buf[2], *tmp;
+ int i, j, tsiz;
+ for(i = 0; i < 65536; i++){
+ buf[0] = i / 256;
+ buf[1] = i % 256;
+ printf("%d\t", i);
+ printf("U+%02X%02X\t", buf[0], buf[1]);
+ printf("\"\\x%x\\x%x\"\t", buf[0], buf[1]);
+ if((tmp = (unsigned char *)cbiconv((char *)buf, 2, "UTF-16BE", "UTF-8",
+ &tsiz, NULL)) != NULL){
+ if(tsiz > 0){
+ printf("\"");
+ for(j = 0; j < tsiz; j++){
+ printf("\\x%x", tmp[j]);
+ }
+ printf("\"");
+ } else {
+ printf("NULL");
+ }
+ free(tmp);
+ }
+ printf("\n");
+ }
+ exit(0);
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/cbcodec.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/cbtest.c
===================================================================
--- box/trunk/qdbm/cbtest.c (rev 0)
+++ box/trunk/qdbm/cbtest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,924 @@
+/*************************************************************************************************
+ * Test cases of Cabin
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <cabin.h>
+#include <stdio.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+#include <time.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define RECBUFSIZ 32 /* buffer for records */
+#define TEXTBUFSIZ 262144 /* buffer for text */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runsort(int argc, char **argv);
+int runstrstr(int argc, char **argv);
+int runlist(int argc, char **argv);
+int runmap(int argc, char **argv);
+int runheap(int argc, char **argv);
+int runwicked(int argc, char **argv);
+int runmisc(int argc, char **argv);
+int printfflush(const char *format, ...);
+int strpcmp(const void *ap, const void *bp);
+int intpcmp(const void *ap, const void *bp);
+int myrand(void);
+int dosort(int rnum, int disp);
+int dostrstr(int rnum, int disp);
+int dolist(int rnum, int disp);
+int domap(int rnum, int bnum, int disp);
+int doheap(int rnum, int max, int disp);
+int dowicked(int rnum);
+int domisc(void);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "sort")){
+ rv = runsort(argc, argv);
+ } else if(!strcmp(argv[1], "strstr")){
+ rv = runstrstr(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else if(!strcmp(argv[1], "map")){
+ rv = runmap(argc, argv);
+ } else if(!strcmp(argv[1], "heap")){
+ rv = runheap(argc, argv);
+ } else if(!strcmp(argv[1], "wicked")){
+ rv = runwicked(argc, argv);
+ } else if(!strcmp(argv[1], "misc")){
+ rv = runmisc(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Cabin\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s sort [-d] rnum\n", progname);
+ fprintf(stderr, " %s strstr [-d] rnum\n", progname);
+ fprintf(stderr, " %s list [-d] rnum\n", progname);
+ fprintf(stderr, " %s map [-d] rnum [bnum]\n", progname);
+ fprintf(stderr, " %s heap [-d] rnum [top]\n", progname);
+ fprintf(stderr, " %s wicked rnum\n", progname);
+ fprintf(stderr, " %s misc\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of sort command */
+int runsort(int argc, char **argv){
+ int i, rnum, disp, rv;
+ char *rstr;
+ rstr = NULL;
+ rnum = 0;
+ disp = FALSE;
+ for(i = 2; i < argc; i++){
+ if(argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ disp = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dosort(rnum, disp);
+ return rv;
+}
+
+
+/* parse arguments of strstr command */
+int runstrstr(int argc, char **argv){
+ int i, rnum, disp, rv;
+ char *rstr;
+ rstr = NULL;
+ rnum = 0;
+ disp = FALSE;
+ for(i = 2; i < argc; i++){
+ if(argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ disp = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dostrstr(rnum, disp);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ int i, rnum, disp, rv;
+ char *rstr;
+ rstr = NULL;
+ rnum = 0;
+ disp = FALSE;
+ for(i = 2; i < argc; i++){
+ if(argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ disp = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dolist(rnum, disp);
+ return rv;
+}
+
+
+/* parse arguments of map command */
+int runmap(int argc, char **argv){
+ int i, rnum, bnum, disp, rv;
+ char *rstr, *bstr;
+ rstr = NULL;
+ bstr = NULL;
+ rnum = 0;
+ bnum = -1;
+ disp = FALSE;
+ for(i = 2; i < argc; i++){
+ if(argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ disp = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!bstr){
+ bstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ if(bstr) bnum = atoi(bstr);
+ rv = domap(rnum, bnum, disp);
+ return rv;
+}
+
+
+/* parse arguments of heap command */
+int runheap(int argc, char **argv){
+ int i, rnum, max, disp, rv;
+ char *rstr, *mstr;
+ rstr = NULL;
+ mstr = NULL;
+ rnum = 0;
+ max = -1;
+ disp = FALSE;
+ for(i = 2; i < argc; i++){
+ if(argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-d")){
+ disp = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!mstr){
+ mstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ if(mstr) max = atoi(mstr);
+ if(max < 0) max = rnum;
+ rv = doheap(rnum, max, disp);
+ rv = 0;
+ return rv;
+}
+
+
+/* parse arguments of wicked command */
+int runwicked(int argc, char **argv){
+ int i, rnum, rv;
+ char *rstr;
+ rstr = NULL;
+ rnum = 0;
+ for(i = 2; i < argc; i++){
+ if(argv[i][0] == '-'){
+ usage();
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowicked(rnum);
+ return rv;
+}
+
+
+/* parse arguments of misc command */
+int runmisc(int argc, char **argv){
+ int rv;
+ rv = domisc();
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* comparing function for strings */
+int strpcmp(const void *ap, const void *bp){
+ return strcmp(*(char **)ap, *(char **)bp);
+}
+
+
+/* comparing function for integers */
+int intpcmp(const void *ap, const void *bp){
+ return *(int *)ap - *(int *)bp;
+}
+
+
+/* pseudo random number generator */
+int myrand(void){
+ static int cnt = 0;
+ if(cnt == 0) srand(time(NULL));
+ return (rand() * rand() + (rand() >> (sizeof(int) * 4)) + (cnt++)) & INT_MAX;
+}
+
+
+/* perform sort command */
+int dosort(int rnum, int disp){
+ char **ivector1, **ivector2, **ivector3, **ivector4, **ivector5;
+ char buf[RECBUFSIZ];
+ int i, len, err;
+ if(!disp) printfflush("<Sorting Test>\n rnum=%d\n\n", rnum);
+ ivector1 = cbmalloc(rnum * sizeof(ivector1[0]));
+ ivector2 = cbmalloc(rnum * sizeof(ivector2[0]));
+ ivector3 = cbmalloc(rnum * sizeof(ivector3[0]));
+ ivector4 = cbmalloc(rnum * sizeof(ivector4[0]));
+ ivector5 = cbmalloc(rnum * sizeof(ivector5[0]));
+ err = FALSE;
+ for(i = 0; i < rnum; i++){
+ len = sprintf(buf, "%08d", myrand() % rnum + 1);
+ ivector1[i] = cbmemdup(buf, len);
+ ivector2[i] = cbmemdup(buf, len);
+ ivector3[i] = cbmemdup(buf, len);
+ ivector4[i] = cbmemdup(buf, len);
+ ivector5[i] = cbmemdup(buf, len);
+ }
+ if(!disp) printfflush("Sorting with insert sort ... ");
+ cbisort(ivector1, rnum, sizeof(ivector1[0]), strpcmp);
+ if(!disp) printfflush("ok\n");
+ if(!disp) printfflush("Sorting with shell sort ... ");
+ cbssort(ivector2, rnum, sizeof(ivector2[0]), strpcmp);
+ if(!disp) printfflush("ok\n");
+ if(!disp) printfflush("Sorting with heap sort ... ");
+ cbhsort(ivector3, rnum, sizeof(ivector3[0]), strpcmp);
+ if(!disp) printfflush("ok\n");
+ if(!disp) printfflush("Sorting with quick sort ... ");
+ cbqsort(ivector4, rnum, sizeof(ivector4[0]), strpcmp);
+ if(!disp) printfflush("ok\n");
+ for(i = 0; i < rnum; i++){
+ if(disp) printfflush("%s\t%s\t%s\t%s\t[%s]\n",
+ ivector1[i], ivector2[i], ivector3[i], ivector4[i], ivector5[i]);
+ if(strcmp(ivector1[i], ivector2[i])) err = TRUE;
+ if(strcmp(ivector1[i], ivector3[i])) err = TRUE;
+ if(strcmp(ivector1[i], ivector4[i])) err = TRUE;
+ free(ivector1[i]);
+ free(ivector2[i]);
+ free(ivector3[i]);
+ free(ivector4[i]);
+ free(ivector5[i]);
+ }
+ free(ivector1);
+ free(ivector2);
+ free(ivector3);
+ free(ivector4);
+ free(ivector5);
+ if(err) fprintf(stderr, "%s: sorting failed\n", progname);
+ if(!disp && !err) printfflush("all ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform strstr command */
+int dostrstr(int rnum, int disp){
+ char *text, buf[RECBUFSIZ], *std, *kmp, *bm;
+ int i, j, len, err;
+ text = cbmalloc(TEXTBUFSIZ);
+ for(i = 0; i < TEXTBUFSIZ - 1; i++){
+ text[i] = 'a' + myrand() % ('z' - 'a');
+ }
+ text[i] = '\0';
+ err = FALSE;
+ if(!disp) printfflush("Locating substrings ... ");
+ for(i = 0; i < rnum; i++){
+ len = myrand() % (RECBUFSIZ - 1);
+ for(j = 0; j < len; j++){
+ buf[j] = 'a' + myrand() % ('z' - 'a');
+ }
+ buf[j] = 0;
+ std = strstr(text, buf);
+ kmp = cbstrstrkmp(text, buf);
+ bm = cbstrstrbm(text, buf);
+ if(kmp != std || bm != std){
+ err = TRUE;
+ break;
+ }
+ if(disp && std) printf("%s\n", buf);
+ }
+ if(err) fprintf(stderr, "%s: string scanning failed\n", progname);
+ if(!disp && !err){
+ printfflush("ok\n");
+ printfflush("all ok\n\n");
+ }
+ free(text);
+ return err ? 1 : 0;
+}
+
+
+/* perform list command */
+int dolist(int rnum, int disp){
+ CBLIST *list;
+ const char *vbuf;
+ char buf[RECBUFSIZ], *tmp;
+ int i, err, len, vsiz;
+ if(!disp) printfflush("<List Writing Test>\n rnum=%d\n\n", rnum);
+ list = cblistopen();
+ err = FALSE;
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ cblistpush(list, buf, len);
+ if(!disp && rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ if(disp){
+ for(i = 0; i < cblistnum(list); i++){
+ if((vbuf = cblistval(list, i, &vsiz)) != NULL){
+ printfflush("%s:%d\n", vbuf, vsiz);
+ } else {
+ fprintf(stderr, "%s: val error\n", progname);
+ err = TRUE;
+ break;
+ }
+ }
+ printfflush("\n");
+ while((tmp = cblistpop(list, &vsiz)) != NULL){
+ printfflush("%s:%d\n", tmp, vsiz);
+ free(tmp);
+ }
+ }
+ cblistclose(list);
+ if(!disp && !err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform list command */
+int domap(int rnum, int bnum, int disp){
+ CBMAP *map;
+ const char *kbuf, *vbuf;
+ char buf[RECBUFSIZ];
+ int i, err, len, ksiz, vsiz;
+ if(!disp) printfflush("<Map Writing Test>\n rnum=%d bnum=%d\n\n", rnum, bnum);
+ map = bnum > 0 ? cbmapopenex(bnum) : cbmapopen();
+ err = FALSE;
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!cbmapput(map, buf, len, buf, len, FALSE)){
+ fprintf(stderr, "%s: put error\n", progname);
+ err = TRUE;
+ break;
+ }
+ if(!disp && rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ if(disp){
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ if((vbuf = cbmapget(map, buf, len, &vsiz)) != NULL){
+ printfflush("%s:%d\t%s:%d\n", buf, len, vbuf, vsiz);
+ } else {
+ fprintf(stderr, "%s: get error\n", progname);
+ }
+ }
+ printfflush("\n");
+ cbmapiterinit(map);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ vbuf = cbmapiterval(kbuf, &vsiz);
+ printfflush("%s:%d\t%s:%d\n", kbuf, ksiz, vbuf, vsiz);
+ }
+ }
+ cbmapclose(map);
+ if(!disp && !err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform heap command */
+int doheap(int rnum, int max, int disp){
+ CBHEAP *heap;
+ int *orig, *ary;
+ int i, err, num, anum;
+ if(!disp) printfflush("<Heap Writing Test>\n rnum=%d max=%d\n\n", rnum, max);
+ orig = disp ? cbmalloc(rnum * sizeof(int) + 1) : NULL;
+ heap = cbheapopen(sizeof(int), max, intpcmp);
+ err = FALSE;
+ for(i = 1; i <= rnum; i++){
+ num = myrand() % rnum + 1;
+ if(orig) orig[i-1] = num;
+ cbheapinsert(heap, &num);
+ if(!disp && rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ if(disp){
+ for(i = 0; i < cbheapnum(heap); i++){
+ printf("%d\n", *(int *)cbheapval(heap, i));
+ }
+ printf("\n");
+ qsort(orig, rnum, sizeof(int), intpcmp);
+ ary = (int *)cbheaptomalloc(heap, &anum);
+ if(anum != rnum && anum != max) err = TRUE;
+ for(i = 0; i < anum; i++){
+ printf("%d\t%d\n", ary[i], orig[i]);
+ if(ary[i] != orig[i]) err = TRUE;
+ }
+ free(ary);
+ } else {
+ cbheapclose(heap);
+ }
+ free(orig);
+ if(!disp && !err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform wicked command */
+int dowicked(int rnum){
+ CBLIST *list;
+ CBMAP *map;
+ int i, len;
+ char buf[RECBUFSIZ], *tmp;
+ printfflush("<Wicked Writing Test>\n rnum=%d\n\n", rnum);
+ list = cblistopen();
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%d", myrand() % rnum + 1);
+ switch(myrand() % 16){
+ case 0:
+ free(cblistpop(list, NULL));
+ putchar('O');
+ break;
+ case 1:
+ cblistunshift(list, buf, len);
+ putchar('U');
+ break;
+ case 2:
+ free(cblistshift(list, NULL));
+ putchar('S');
+ break;
+ case 3:
+ cblistinsert(list, myrand() % (cblistnum(list) + 1), buf, len);
+ putchar('I');
+ break;
+ case 4:
+ free(cblistremove(list, myrand() % (cblistnum(list) + 1), NULL));
+ putchar('R');
+ break;
+ case 5:
+ cblistover(list, myrand() % (cblistnum(list) + 1), buf, len);
+ putchar('V');
+ break;
+ case 6:
+ tmp = cbmemdup(buf, len);
+ cblistpushbuf(list, tmp, len);
+ putchar('B');
+ break;
+ default:
+ cblistpush(list, buf, len);
+ putchar('P');
+ break;
+ }
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ cblistclose(list);
+ map = cbmapopen();
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%d", myrand() % rnum + 1);
+ switch(myrand() % 16){
+ case 0:
+ cbmapput(map, buf, len, buf, len, FALSE);
+ putchar('I');
+ break;
+ case 1:
+ cbmapputcat(map, buf, len, buf, len);
+ putchar('C');
+ break;
+ case 2:
+ cbmapget(map, buf, len, NULL);
+ putchar('V');
+ break;
+ case 3:
+ cbmapout(map, buf, len);
+ putchar('D');
+ break;
+ case 4:
+ cbmapmove(map, buf, len, myrand() % 2);
+ putchar('M');
+ break;
+ default:
+ cbmapput(map, buf, len, buf, len, TRUE);
+ putchar('O');
+ break;
+ }
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ cbmapclose(map);
+ printfflush("ok\n\n");
+ return 0;
+}
+
+
+/* perform misc command */
+int domisc(void){
+ CBDATUM *odatum, *ndatum;
+ CBLIST *olist, *nlist, *elems, *glist;
+ CBMAP *omap, *nmap, *pairs, *gmap;
+ int i, j, ssiz, osiz, tsiz, jl;
+ char kbuf[RECBUFSIZ], vbuf[RECBUFSIZ], *sbuf, spbuf[1024], *tmp, *orig, renc[64];
+ const char *op, *np;
+ time_t t;
+ printfflush("<Miscellaneous Test>\n\n");
+ printfflush("Checking memory allocation ... ");
+ tmp = cbmalloc(1024);
+ for(i = 1; i <= 65536; i *= 2){
+ tmp = cbrealloc(tmp, i);
+ }
+ cbfree(tmp);
+ printfflush("ok\n");
+ printfflush("Checking basic datum ... ");
+ odatum = cbdatumopen("x", -1);
+ for(i = 0; i < 1000; i++){
+ cbdatumcat(odatum, "x", 1);
+ }
+ cbdatumclose(odatum);
+ tmp = cbmalloc(3);
+ memcpy(tmp, "abc", 3);
+ odatum = cbdatumopenbuf(tmp, 3);
+ for(i = 0; i < 1000; i++){
+ cbdatumcat(odatum, ".", 1);
+ }
+ ndatum = cbdatumdup(odatum);
+ for(i = 0; i < 1000; i++){
+ cbdatumcat(ndatum, "*", 1);
+ }
+ for(i = 0; i < 1000; i++){
+ tmp = cbmalloc(3);
+ memcpy(tmp, "123", 3);
+ cbdatumsetbuf(ndatum, tmp, 3);
+ }
+ cbdatumprintf(ndatum, "[%s\t%08d\t%08o\t%08x\t%08.1f\t%@\t%?\t%:]",
+ "mikio", 1978, 1978, 1978, 1978.0211, "<>&#!+-*/%", "<>&#!+-*/%", "<>&#!+-*/%");
+ cbdatumclose(ndatum);
+ cbdatumclose(odatum);
+ printfflush("ok\n");
+ printfflush("Checking serialization of list ... ");
+ olist = cblistopen();
+ for(i = 1; i <= 1000; i++){
+ sprintf(vbuf, "%d", i);
+ cblistpush(olist, vbuf, -1);
+ }
+ sbuf = cblistdump(olist, &ssiz);
+ nlist = cblistload(sbuf, ssiz);
+ free(sbuf);
+ for(i = 0; i < cblistnum(olist); i++){
+ op = cblistval(olist, i, NULL);
+ np = cblistval(nlist, i, NULL);
+ if(!op || !np || strcmp(op, np)){
+ cblistclose(nlist);
+ cblistclose(olist);
+ fprintf(stderr, "%s: validation failed\n", progname);
+ return 1;
+ }
+ }
+ cblistclose(nlist);
+ cblistclose(olist);
+ printfflush("ok\n");
+ printfflush("Checking serialization of map ... ");
+ omap = cbmapopen();
+ for(i = 1; i <= 1000; i++){
+ sprintf(kbuf, "%X", i);
+ sprintf(vbuf, "[%d]", i);
+ cbmapput(omap, kbuf, -1, vbuf, -1, TRUE);
+ }
+ sbuf = cbmapdump(omap, &ssiz);
+ nmap = cbmapload(sbuf, ssiz);
+ free(cbmaploadone(sbuf, ssiz, "1", 2, &tsiz));
+ free(cbmaploadone(sbuf, ssiz, "33", 2, &tsiz));
+ free(sbuf);
+ cbmapiterinit(omap);
+ while((op = cbmapiternext(omap, NULL)) != NULL){
+ if(!(np = cbmapget(nmap, op, -1, NULL))){
+ cbmapclose(nmap);
+ cbmapclose(omap);
+ fprintf(stderr, "%s: validation failed\n", progname);
+ return 1;
+ }
+ }
+ cbmapclose(nmap);
+ cbmapclose(omap);
+ printfflush("ok\n");
+ printfflush("Checking string utilities ... ");
+ sprintf(spbuf, "[%08d/%08o/%08u/%08x/%08X/%08.3e/%08.3E/%08.3f/%08.3g/%08.3G/%c/%s/%%]",
+ 123456, 123456, 123456, 123456, 123456,
+ 123456.789, 123456.789, 123456.789, 123456.789, 123456.789,
+ 'A', "hoge");
+ tmp = cbsprintf("[%08d/%08o/%08u/%08x/%08X/%08.3e/%08.3E/%08.3f/%08.3g/%08.3G/%c/%s/%%]",
+ 123456, 123456, 123456, 123456, 123456,
+ 123456.789, 123456.789, 123456.789, 123456.789, 123456.789,
+ 'A', "hoge");
+ while(strcmp(spbuf, tmp)){
+ free(tmp);
+ fprintf(stderr, "%s: cbsprintf is invalid\n", progname);
+ return 1;
+ }
+ free(tmp);
+ pairs = cbmapopen();
+ cbmapput(pairs, "aa", -1, "AAA", -1, TRUE);
+ cbmapput(pairs, "bb", -1, "BBB", -1, TRUE);
+ cbmapput(pairs, "cc", -1, "CCC", -1, TRUE);
+ cbmapput(pairs, "ZZZ", -1, "z", -1, TRUE);
+ tmp = cbreplace("[aaaaabbbbbcccccdddddZZZZ]", pairs);
+ if(strcmp(tmp, "[AAAAAAaBBBBBBbCCCCCCcdddddzZ]")){
+ free(tmp);
+ cbmapclose(pairs);
+ fprintf(stderr, "%s: cbreplace is invalid\n", progname);
+ return 1;
+ }
+ free(tmp);
+ cbmapclose(pairs);
+ elems = cbsplit("aa bb,ccc-dd,", -1, " ,-");
+ if(cblistnum(elems) != 5 || strcmp(cblistval(elems, 0, NULL), "aa") ||
+ strcmp(cblistval(elems, 1, NULL), "bb") || strcmp(cblistval(elems, 2, NULL), "ccc") ||
+ strcmp(cblistval(elems, 3, NULL), "dd") || strcmp(cblistval(elems, 4, NULL), "")){
+ cblistclose(elems);
+ fprintf(stderr, "%s: cbsplit is invalid\n", progname);
+ return 1;
+ }
+ cblistclose(elems);
+ if(cbstricmp("abc", "ABC") || !cbstricmp("abc", "abcd")){
+ fprintf(stderr, "%s: cbstricmp is invalid\n", progname);
+ return 1;
+ }
+ if(!cbstrfwmatch("abcd", "abc") || cbstrfwmatch("abc", "abcd")){
+ fprintf(stderr, "%s: cbstrfwmatch is invalid\n", progname);
+ return 1;
+ }
+ if(!cbstrfwimatch("abcd", "ABC") || cbstrfwmatch("abc", "ABCD")){
+ fprintf(stderr, "%s: cbstrfwimatch is invalid\n", progname);
+ return 1;
+ }
+ if(!cbstrbwmatch("dcba", "cba") || cbstrbwmatch("cba", "dcba")){
+ fprintf(stderr, "%s: cbstrbwmatch is invalid\n", progname);
+ return 1;
+ }
+ if(!cbstrbwimatch("dcba", "CBA") || cbstrbwimatch("cba", "DCBA")){
+ fprintf(stderr, "%s: cbstrbwimatch is invalid\n", progname);
+ return 1;
+ }
+ tmp = cbmemdup(" \r\n[Quick Database Manager]\r\n ", -1);
+ if(cbstrtoupper(tmp) != tmp || strcmp(tmp, " \r\n[QUICK DATABASE MANAGER]\r\n ")){
+ free(tmp);
+ fprintf(stderr, "%s: cbstrtoupper is invalid\n", progname);
+ return 1;
+ }
+ if(cbstrtolower(tmp) != tmp || strcmp(tmp, " \r\n[quick database manager]\r\n ")){
+ free(tmp);
+ fprintf(stderr, "%s: cbstrtolower is invalid\n", progname);
+ return 1;
+ }
+ if(cbstrtrim(tmp) != tmp || strcmp(tmp, "[quick database manager]")){
+ free(tmp);
+ fprintf(stderr, "%s: cbstrtrim is invalid\n", progname);
+ return 1;
+ }
+ if(cbstrsqzspc(tmp) != tmp || strcmp(tmp, "[quick database manager]")){
+ free(tmp);
+ fprintf(stderr, "%s: cbstrsqzspc is invalid\n", progname);
+ return 1;
+ }
+ cbstrcututf(tmp, 5);
+ if(cbstrcountutf(tmp) != 5){
+ free(tmp);
+ fprintf(stderr, "%s: cbstrcututf or cbstrcountutf is invalid\n", progname);
+ return 1;
+ }
+ free(tmp);
+ printfflush("ok\n");
+ printfflush("Checking encoding utilities ... ");
+ strcpy(spbuf, "My name is \xca\xbf\xce\xd3\xb4\xb4\xcd\xba.\n\n<Love & Peace!>\n");
+ tmp = cbbaseencode(spbuf, -1);
+ orig = cbbasedecode(tmp, &osiz);
+ if(osiz != strlen(spbuf) || strcmp(orig, spbuf)){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: Base64 encoding is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ strcpy(spbuf, "My name is \xca\xbf\xce\xd3\xb4\xb4\xcd\xba.\n\n<Love & Peace!>\n");
+ tmp = cbquoteencode(spbuf, -1);
+ orig = cbquotedecode(tmp, &osiz);
+ if(osiz != strlen(spbuf) || strcmp(orig, spbuf)){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: quoted-printable encoding is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ strcpy(spbuf, "My name is \xca\xbf\xce\xd3\xb4\xb4\xcd\xba.\n\n<Love & Peace!>\n");
+ tmp = cbmimeencode(spbuf, "ISO-8859-1", TRUE);
+ orig = cbmimedecode(tmp, renc);
+ if(osiz != strlen(spbuf) || strcmp(orig, spbuf) || strcmp(renc, "ISO-8859-1")){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: MIME encoding is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ strcpy(spbuf, "\"He says...\r\n\"\"What PROGRAM are they watching?\"\"\"");
+ tmp = cbcsvunescape(spbuf);
+ orig = cbcsvescape(tmp);
+ if(strcmp(orig, spbuf)){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: CSV escaping is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ strcpy(spbuf, "<Nuts&Milk> is "very" surfin'!");
+ tmp = cbxmlunescape(spbuf);
+ orig = cbxmlescape(tmp);
+ if(strcmp(orig, spbuf)){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: XML escaping is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ printfflush("ok\n");
+ printfflush("Checking date utilities ... ");
+ for(i = 0; i < 200; i++){
+ jl = (myrand() % 23) * 1800;
+ if(myrand() % 2 == 0) jl *= -1;
+ t = myrand() % (INT_MAX - 3600 * 24 * 365 * 6) + 3600 * 24 * 365 * 5;
+ tmp = cbdatestrwww(t, jl);
+ t = cbstrmktime(tmp);
+ orig = cbdatestrwww(t, jl);
+ if(strcmp(orig, tmp)){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: W3CDTF formatter is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ tmp = cbdatestrhttp(t, jl);
+ t = cbstrmktime(tmp);
+ orig = cbdatestrhttp(t, jl);
+ if(strcmp(orig, tmp)){
+ free(orig);
+ free(tmp);
+ fprintf(stderr, "%s: RFC 822 date formatter is invalid\n", progname);
+ return 1;
+ }
+ free(orig);
+ free(tmp);
+ }
+ printfflush("ok\n");
+ printfflush("Checking the global garbage collector ... ");
+ for(i = 0; i < 512; i++){
+ glist = cblistopen();
+ cbglobalgc(glist, (void (*)(void *))cblistclose);
+ for(j = 0; j < 10; j++){
+ sprintf(kbuf, "%08d", j);
+ cblistpush(glist, kbuf, -1);
+ }
+ gmap = cbmapopen();
+ cbglobalgc(gmap, (void (*)(void *))cbmapclose);
+ for(j = 0; j < 10; j++){
+ sprintf(kbuf, "%08d", j);
+ cbmapput(gmap, kbuf, -1, kbuf, -1, TRUE);
+ }
+ if(myrand() % 64 == 0){
+ cbvmemavail(100);
+ cbggcsweep();
+ }
+ }
+ printfflush("ok\n");
+ printfflush("all ok\n\n");
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/cbtest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/configure
===================================================================
--- box/trunk/qdbm/configure (rev 0)
+++ box/trunk/qdbm/configure 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,3913 @@
+#! /bin/sh
+# Guess values for system-dependent variables and create Makefiles.
+# Generated by GNU Autoconf 2.59 for qdbm 1.8.77.
+#
+# Copyright (C) 2003 Free Software Foundation, Inc.
+# This configure script is free software; the Free Software Foundation
+# gives unlimited permission to copy, distribute and modify it.
+## --------------------- ##
+## M4sh Initialization. ##
+## --------------------- ##
+
+# Be Bourne compatible
+if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then
+ emulate sh
+ NULLCMD=:
+ # Zsh 3.x and 4.x performs word splitting on ${1+"$@"}, which
+ # is contrary to our usage. Disable this feature.
+ alias -g '${1+"$@"}'='"$@"'
+elif test -n "${BASH_VERSION+set}" && (set -o posix) >/dev/null 2>&1; then
+ set -o posix
+fi
+DUALCASE=1; export DUALCASE # for MKS sh
+
+# Support unset when possible.
+if ( (MAIL=60; unset MAIL) || exit) >/dev/null 2>&1; then
+ as_unset=unset
+else
+ as_unset=false
+fi
+
+
+# Work around bugs in pre-3.0 UWIN ksh.
+$as_unset ENV MAIL MAILPATH
+PS1='$ '
+PS2='> '
+PS4='+ '
+
+# NLS nuisances.
+for as_var in \
+ LANG LANGUAGE LC_ADDRESS LC_ALL LC_COLLATE LC_CTYPE LC_IDENTIFICATION \
+ LC_MEASUREMENT LC_MESSAGES LC_MONETARY LC_NAME LC_NUMERIC LC_PAPER \
+ LC_TELEPHONE LC_TIME
+do
+ if (set +x; test -z "`(eval $as_var=C; export $as_var) 2>&1`"); then
+ eval $as_var=C; export $as_var
+ else
+ $as_unset $as_var
+ fi
+done
+
+# Required to use basename.
+if expr a : '\(a\)' >/dev/null 2>&1; then
+ as_expr=expr
+else
+ as_expr=false
+fi
+
+if (basename /) >/dev/null 2>&1 && test "X`basename / 2>&1`" = "X/"; then
+ as_basename=basename
+else
+ as_basename=false
+fi
+
+
+# Name of the executable.
+as_me=`$as_basename "$0" ||
+$as_expr X/"$0" : '.*/\([^/][^/]*\)/*$' \| \
+ X"$0" : 'X\(//\)$' \| \
+ X"$0" : 'X\(/\)$' \| \
+ . : '\(.\)' 2>/dev/null ||
+echo X/"$0" |
+ sed '/^.*\/\([^/][^/]*\)\/*$/{ s//\1/; q; }
+ /^X\/\(\/\/\)$/{ s//\1/; q; }
+ /^X\/\(\/\).*/{ s//\1/; q; }
+ s/.*/./; q'`
+
+
+# PATH needs CR, and LINENO needs CR and PATH.
+# Avoid depending upon Character Ranges.
+as_cr_letters='abcdefghijklmnopqrstuvwxyz'
+as_cr_LETTERS='ABCDEFGHIJKLMNOPQRSTUVWXYZ'
+as_cr_Letters=$as_cr_letters$as_cr_LETTERS
+as_cr_digits='0123456789'
+as_cr_alnum=$as_cr_Letters$as_cr_digits
+
+# The user is always right.
+if test "${PATH_SEPARATOR+set}" != set; then
+ echo "#! /bin/sh" >conf$$.sh
+ echo "exit 0" >>conf$$.sh
+ chmod +x conf$$.sh
+ if (PATH="/nonexistent;."; conf$$.sh) >/dev/null 2>&1; then
+ PATH_SEPARATOR=';'
+ else
+ PATH_SEPARATOR=:
+ fi
+ rm -f conf$$.sh
+fi
+
+
+ as_lineno_1=$LINENO
+ as_lineno_2=$LINENO
+ as_lineno_3=`(expr $as_lineno_1 + 1) 2>/dev/null`
+ test "x$as_lineno_1" != "x$as_lineno_2" &&
+ test "x$as_lineno_3" = "x$as_lineno_2" || {
+ # Find who we are. Look in the path if we contain no path at all
+ # relative or not.
+ case $0 in
+ *[\\/]* ) as_myself=$0 ;;
+ *) as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ test -r "$as_dir/$0" && as_myself=$as_dir/$0 && break
+done
+
+ ;;
+ esac
+ # We did not find ourselves, most probably we were run as `sh COMMAND'
+ # in which case we are not to be found in the path.
+ if test "x$as_myself" = x; then
+ as_myself=$0
+ fi
+ if test ! -f "$as_myself"; then
+ { echo "$as_me: error: cannot find myself; rerun with an absolute path" >&2
+ { (exit 1); exit 1; }; }
+ fi
+ case $CONFIG_SHELL in
+ '')
+ as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in /bin$PATH_SEPARATOR/usr/bin$PATH_SEPARATOR$PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for as_base in sh bash ksh sh5; do
+ case $as_dir in
+ /*)
+ if ("$as_dir/$as_base" -c '
+ as_lineno_1=$LINENO
+ as_lineno_2=$LINENO
+ as_lineno_3=`(expr $as_lineno_1 + 1) 2>/dev/null`
+ test "x$as_lineno_1" != "x$as_lineno_2" &&
+ test "x$as_lineno_3" = "x$as_lineno_2" ') 2>/dev/null; then
+ $as_unset BASH_ENV || test "${BASH_ENV+set}" != set || { BASH_ENV=; export BASH_ENV; }
+ $as_unset ENV || test "${ENV+set}" != set || { ENV=; export ENV; }
+ CONFIG_SHELL=$as_dir/$as_base
+ export CONFIG_SHELL
+ exec "$CONFIG_SHELL" "$0" ${1+"$@"}
+ fi;;
+ esac
+ done
+done
+;;
+ esac
+
+ # Create $as_me.lineno as a copy of $as_myself, but with $LINENO
+ # uniformly replaced by the line number. The first 'sed' inserts a
+ # line-number line before each line; the second 'sed' does the real
+ # work. The second script uses 'N' to pair each line-number line
+ # with the numbered line, and appends trailing '-' during
+ # substitution so that $LINENO is not a special case at line end.
+ # (Raja R Harinath suggested sed '=', and Paul Eggert wrote the
+ # second 'sed' script. Blame Lee E. McMahon for sed's syntax. :-)
+ sed '=' <$as_myself |
+ sed '
+ N
+ s,$,-,
+ : loop
+ s,^\(['$as_cr_digits']*\)\(.*\)[$]LINENO\([^'$as_cr_alnum'_]\),\1\2\1\3,
+ t loop
+ s,-$,,
+ s,^['$as_cr_digits']*\n,,
+ ' >$as_me.lineno &&
+ chmod +x $as_me.lineno ||
+ { echo "$as_me: error: cannot create $as_me.lineno; rerun with a POSIX shell" >&2
+ { (exit 1); exit 1; }; }
+
+ # Don't try to exec as it changes $[0], causing all sort of problems
+ # (the dirname of $[0] is not the place where we might find the
+ # original and so on. Autoconf is especially sensible to this).
+ . ./$as_me.lineno
+ # Exit status is that of the last command.
+ exit
+}
+
+
+case `echo "testing\c"; echo 1,2,3`,`echo -n testing; echo 1,2,3` in
+ *c*,-n*) ECHO_N= ECHO_C='
+' ECHO_T=' ' ;;
+ *c*,* ) ECHO_N=-n ECHO_C= ECHO_T= ;;
+ *) ECHO_N= ECHO_C='\c' ECHO_T= ;;
+esac
+
+if expr a : '\(a\)' >/dev/null 2>&1; then
+ as_expr=expr
+else
+ as_expr=false
+fi
+
+rm -f conf$$ conf$$.exe conf$$.file
+echo >conf$$.file
+if ln -s conf$$.file conf$$ 2>/dev/null; then
+ # We could just check for DJGPP; but this test a) works b) is more generic
+ # and c) will remain valid once DJGPP supports symlinks (DJGPP 2.04).
+ if test -f conf$$.exe; then
+ # Don't use ln at all; we don't have any links
+ as_ln_s='cp -p'
+ else
+ as_ln_s='ln -s'
+ fi
+elif ln conf$$.file conf$$ 2>/dev/null; then
+ as_ln_s=ln
+else
+ as_ln_s='cp -p'
+fi
+rm -f conf$$ conf$$.exe conf$$.file
+
+if mkdir -p . 2>/dev/null; then
+ as_mkdir_p=:
+else
+ test -d ./-p && rmdir ./-p
+ as_mkdir_p=false
+fi
+
+as_executable_p="test -f"
+
+# Sed expression to map a string onto a valid CPP name.
+as_tr_cpp="eval sed 'y%*$as_cr_letters%P$as_cr_LETTERS%;s%[^_$as_cr_alnum]%_%g'"
+
+# Sed expression to map a string onto a valid variable name.
+as_tr_sh="eval sed 'y%*+%pp%;s%[^_$as_cr_alnum]%_%g'"
+
+
+# IFS
+# We need space, tab and new line, in precisely that order.
+as_nl='
+'
+IFS=" $as_nl"
+
+# CDPATH.
+$as_unset CDPATH
+
+
+# Name of the host.
+# hostname on some systems (SVR3.2, Linux) returns a bogus exit status,
+# so uname gets run too.
+ac_hostname=`(hostname || uname -n) 2>/dev/null | sed 1q`
+
+exec 6>&1
+
+#
+# Initializations.
+#
+ac_default_prefix=/usr/local
+ac_config_libobj_dir=.
+cross_compiling=no
+subdirs=
+MFLAGS=
+MAKEFLAGS=
+SHELL=${CONFIG_SHELL-/bin/sh}
+
+# Maximum number of lines to put in a shell here document.
+# This variable seems obsolete. It should probably be removed, and
+# only ac_max_sed_lines should be used.
+: ${ac_max_here_lines=38}
+
+# Identity of this package.
+PACKAGE_NAME='qdbm'
+PACKAGE_TARNAME='qdbm'
+PACKAGE_VERSION='1.8.77'
+PACKAGE_STRING='qdbm 1.8.77'
+PACKAGE_BUGREPORT=''
+
+ac_subst_vars='SHELL PATH_SEPARATOR PACKAGE_NAME PACKAGE_TARNAME PACKAGE_VERSION PACKAGE_STRING PACKAGE_BUGREPORT exec_prefix prefix program_transform_name bindir sbindir libexecdir datadir sysconfdir sharedstatedir localstatedir libdir includedir oldincludedir infodir mandir build_alias host_alias target_alias DEFS ECHO_C ECHO_N ECHO_T LIBS CC CFLAGS LDFLAGS CPPFLAGS ac_ct_CC EXEEXT OBJEXT LIBVER LIBREV TARGETS MYDEFS MYOPTS MGWLIBS LD AR LIBOBJS LTLIBOBJS'
+ac_subst_files=''
+
+# Initialize some variables set by options.
+ac_init_help=
+ac_init_version=false
+# The variables have the same names as the options, with
+# dashes changed to underlines.
+cache_file=/dev/null
+exec_prefix=NONE
+no_create=
+no_recursion=
+prefix=NONE
+program_prefix=NONE
+program_suffix=NONE
+program_transform_name=s,x,x,
+silent=
+site=
+srcdir=
+verbose=
+x_includes=NONE
+x_libraries=NONE
+
+# Installation directory options.
+# These are left unexpanded so users can "make install exec_prefix=/foo"
+# and all the variables that are supposed to be based on exec_prefix
+# by default will actually change.
+# Use braces instead of parens because sh, perl, etc. also accept them.
+bindir='${exec_prefix}/bin'
+sbindir='${exec_prefix}/sbin'
+libexecdir='${exec_prefix}/libexec'
+datadir='${prefix}/share'
+sysconfdir='${prefix}/etc'
+sharedstatedir='${prefix}/com'
+localstatedir='${prefix}/var'
+libdir='${exec_prefix}/lib'
+includedir='${prefix}/include'
+oldincludedir='/usr/include'
+infodir='${prefix}/info'
+mandir='${prefix}/man'
+
+ac_prev=
+for ac_option
+do
+ # If the previous option needs an argument, assign it.
+ if test -n "$ac_prev"; then
+ eval "$ac_prev=\$ac_option"
+ ac_prev=
+ continue
+ fi
+
+ ac_optarg=`expr "x$ac_option" : 'x[^=]*=\(.*\)'`
+
+ # Accept the important Cygnus configure options, so we can diagnose typos.
+
+ case $ac_option in
+
+ -bindir | --bindir | --bindi | --bind | --bin | --bi)
+ ac_prev=bindir ;;
+ -bindir=* | --bindir=* | --bindi=* | --bind=* | --bin=* | --bi=*)
+ bindir=$ac_optarg ;;
+
+ -build | --build | --buil | --bui | --bu)
+ ac_prev=build_alias ;;
+ -build=* | --build=* | --buil=* | --bui=* | --bu=*)
+ build_alias=$ac_optarg ;;
+
+ -cache-file | --cache-file | --cache-fil | --cache-fi \
+ | --cache-f | --cache- | --cache | --cach | --cac | --ca | --c)
+ ac_prev=cache_file ;;
+ -cache-file=* | --cache-file=* | --cache-fil=* | --cache-fi=* \
+ | --cache-f=* | --cache-=* | --cache=* | --cach=* | --cac=* | --ca=* | --c=*)
+ cache_file=$ac_optarg ;;
+
+ --config-cache | -C)
+ cache_file=config.cache ;;
+
+ -datadir | --datadir | --datadi | --datad | --data | --dat | --da)
+ ac_prev=datadir ;;
+ -datadir=* | --datadir=* | --datadi=* | --datad=* | --data=* | --dat=* \
+ | --da=*)
+ datadir=$ac_optarg ;;
+
+ -disable-* | --disable-*)
+ ac_feature=`expr "x$ac_option" : 'x-*disable-\(.*\)'`
+ # Reject names that are not valid shell variable names.
+ expr "x$ac_feature" : ".*[^-_$as_cr_alnum]" >/dev/null &&
+ { echo "$as_me: error: invalid feature name: $ac_feature" >&2
+ { (exit 1); exit 1; }; }
+ ac_feature=`echo $ac_feature | sed 's/-/_/g'`
+ eval "enable_$ac_feature=no" ;;
+
+ -enable-* | --enable-*)
+ ac_feature=`expr "x$ac_option" : 'x-*enable-\([^=]*\)'`
+ # Reject names that are not valid shell variable names.
+ expr "x$ac_feature" : ".*[^-_$as_cr_alnum]" >/dev/null &&
+ { echo "$as_me: error: invalid feature name: $ac_feature" >&2
+ { (exit 1); exit 1; }; }
+ ac_feature=`echo $ac_feature | sed 's/-/_/g'`
+ case $ac_option in
+ *=*) ac_optarg=`echo "$ac_optarg" | sed "s/'/'\\\\\\\\''/g"`;;
+ *) ac_optarg=yes ;;
+ esac
+ eval "enable_$ac_feature='$ac_optarg'" ;;
+
+ -exec-prefix | --exec_prefix | --exec-prefix | --exec-prefi \
+ | --exec-pref | --exec-pre | --exec-pr | --exec-p | --exec- \
+ | --exec | --exe | --ex)
+ ac_prev=exec_prefix ;;
+ -exec-prefix=* | --exec_prefix=* | --exec-prefix=* | --exec-prefi=* \
+ | --exec-pref=* | --exec-pre=* | --exec-pr=* | --exec-p=* | --exec-=* \
+ | --exec=* | --exe=* | --ex=*)
+ exec_prefix=$ac_optarg ;;
+
+ -gas | --gas | --ga | --g)
+ # Obsolete; use --with-gas.
+ with_gas=yes ;;
+
+ -help | --help | --hel | --he | -h)
+ ac_init_help=long ;;
+ -help=r* | --help=r* | --hel=r* | --he=r* | -hr*)
+ ac_init_help=recursive ;;
+ -help=s* | --help=s* | --hel=s* | --he=s* | -hs*)
+ ac_init_help=short ;;
+
+ -host | --host | --hos | --ho)
+ ac_prev=host_alias ;;
+ -host=* | --host=* | --hos=* | --ho=*)
+ host_alias=$ac_optarg ;;
+
+ -includedir | --includedir | --includedi | --included | --include \
+ | --includ | --inclu | --incl | --inc)
+ ac_prev=includedir ;;
+ -includedir=* | --includedir=* | --includedi=* | --included=* | --include=* \
+ | --includ=* | --inclu=* | --incl=* | --inc=*)
+ includedir=$ac_optarg ;;
+
+ -infodir | --infodir | --infodi | --infod | --info | --inf)
+ ac_prev=infodir ;;
+ -infodir=* | --infodir=* | --infodi=* | --infod=* | --info=* | --inf=*)
+ infodir=$ac_optarg ;;
+
+ -libdir | --libdir | --libdi | --libd)
+ ac_prev=libdir ;;
+ -libdir=* | --libdir=* | --libdi=* | --libd=*)
+ libdir=$ac_optarg ;;
+
+ -libexecdir | --libexecdir | --libexecdi | --libexecd | --libexec \
+ | --libexe | --libex | --libe)
+ ac_prev=libexecdir ;;
+ -libexecdir=* | --libexecdir=* | --libexecdi=* | --libexecd=* | --libexec=* \
+ | --libexe=* | --libex=* | --libe=*)
+ libexecdir=$ac_optarg ;;
+
+ -localstatedir | --localstatedir | --localstatedi | --localstated \
+ | --localstate | --localstat | --localsta | --localst \
+ | --locals | --local | --loca | --loc | --lo)
+ ac_prev=localstatedir ;;
+ -localstatedir=* | --localstatedir=* | --localstatedi=* | --localstated=* \
+ | --localstate=* | --localstat=* | --localsta=* | --localst=* \
+ | --locals=* | --local=* | --loca=* | --loc=* | --lo=*)
+ localstatedir=$ac_optarg ;;
+
+ -mandir | --mandir | --mandi | --mand | --man | --ma | --m)
+ ac_prev=mandir ;;
+ -mandir=* | --mandir=* | --mandi=* | --mand=* | --man=* | --ma=* | --m=*)
+ mandir=$ac_optarg ;;
+
+ -nfp | --nfp | --nf)
+ # Obsolete; use --without-fp.
+ with_fp=no ;;
+
+ -no-create | --no-create | --no-creat | --no-crea | --no-cre \
+ | --no-cr | --no-c | -n)
+ no_create=yes ;;
+
+ -no-recursion | --no-recursion | --no-recursio | --no-recursi \
+ | --no-recurs | --no-recur | --no-recu | --no-rec | --no-re | --no-r)
+ no_recursion=yes ;;
+
+ -oldincludedir | --oldincludedir | --oldincludedi | --oldincluded \
+ | --oldinclude | --oldinclud | --oldinclu | --oldincl | --oldinc \
+ | --oldin | --oldi | --old | --ol | --o)
+ ac_prev=oldincludedir ;;
+ -oldincludedir=* | --oldincludedir=* | --oldincludedi=* | --oldincluded=* \
+ | --oldinclude=* | --oldinclud=* | --oldinclu=* | --oldincl=* | --oldinc=* \
+ | --oldin=* | --oldi=* | --old=* | --ol=* | --o=*)
+ oldincludedir=$ac_optarg ;;
+
+ -prefix | --prefix | --prefi | --pref | --pre | --pr | --p)
+ ac_prev=prefix ;;
+ -prefix=* | --prefix=* | --prefi=* | --pref=* | --pre=* | --pr=* | --p=*)
+ prefix=$ac_optarg ;;
+
+ -program-prefix | --program-prefix | --program-prefi | --program-pref \
+ | --program-pre | --program-pr | --program-p)
+ ac_prev=program_prefix ;;
+ -program-prefix=* | --program-prefix=* | --program-prefi=* \
+ | --program-pref=* | --program-pre=* | --program-pr=* | --program-p=*)
+ program_prefix=$ac_optarg ;;
+
+ -program-suffix | --program-suffix | --program-suffi | --program-suff \
+ | --program-suf | --program-su | --program-s)
+ ac_prev=program_suffix ;;
+ -program-suffix=* | --program-suffix=* | --program-suffi=* \
+ | --program-suff=* | --program-suf=* | --program-su=* | --program-s=*)
+ program_suffix=$ac_optarg ;;
+
+ -program-transform-name | --program-transform-name \
+ | --program-transform-nam | --program-transform-na \
+ | --program-transform-n | --program-transform- \
+ | --program-transform | --program-transfor \
+ | --program-transfo | --program-transf \
+ | --program-trans | --program-tran \
+ | --progr-tra | --program-tr | --program-t)
+ ac_prev=program_transform_name ;;
+ -program-transform-name=* | --program-transform-name=* \
+ | --program-transform-nam=* | --program-transform-na=* \
+ | --program-transform-n=* | --program-transform-=* \
+ | --program-transform=* | --program-transfor=* \
+ | --program-transfo=* | --program-transf=* \
+ | --program-trans=* | --program-tran=* \
+ | --progr-tra=* | --program-tr=* | --program-t=*)
+ program_transform_name=$ac_optarg ;;
+
+ -q | -quiet | --quiet | --quie | --qui | --qu | --q \
+ | -silent | --silent | --silen | --sile | --sil)
+ silent=yes ;;
+
+ -sbindir | --sbindir | --sbindi | --sbind | --sbin | --sbi | --sb)
+ ac_prev=sbindir ;;
+ -sbindir=* | --sbindir=* | --sbindi=* | --sbind=* | --sbin=* \
+ | --sbi=* | --sb=*)
+ sbindir=$ac_optarg ;;
+
+ -sharedstatedir | --sharedstatedir | --sharedstatedi \
+ | --sharedstated | --sharedstate | --sharedstat | --sharedsta \
+ | --sharedst | --shareds | --shared | --share | --shar \
+ | --sha | --sh)
+ ac_prev=sharedstatedir ;;
+ -sharedstatedir=* | --sharedstatedir=* | --sharedstatedi=* \
+ | --sharedstated=* | --sharedstate=* | --sharedstat=* | --sharedsta=* \
+ | --sharedst=* | --shareds=* | --shared=* | --share=* | --shar=* \
+ | --sha=* | --sh=*)
+ sharedstatedir=$ac_optarg ;;
+
+ -site | --site | --sit)
+ ac_prev=site ;;
+ -site=* | --site=* | --sit=*)
+ site=$ac_optarg ;;
+
+ -srcdir | --srcdir | --srcdi | --srcd | --src | --sr)
+ ac_prev=srcdir ;;
+ -srcdir=* | --srcdir=* | --srcdi=* | --srcd=* | --src=* | --sr=*)
+ srcdir=$ac_optarg ;;
+
+ -sysconfdir | --sysconfdir | --sysconfdi | --sysconfd | --sysconf \
+ | --syscon | --sysco | --sysc | --sys | --sy)
+ ac_prev=sysconfdir ;;
+ -sysconfdir=* | --sysconfdir=* | --sysconfdi=* | --sysconfd=* | --sysconf=* \
+ | --syscon=* | --sysco=* | --sysc=* | --sys=* | --sy=*)
+ sysconfdir=$ac_optarg ;;
+
+ -target | --target | --targe | --targ | --tar | --ta | --t)
+ ac_prev=target_alias ;;
+ -target=* | --target=* | --targe=* | --targ=* | --tar=* | --ta=* | --t=*)
+ target_alias=$ac_optarg ;;
+
+ -v | -verbose | --verbose | --verbos | --verbo | --verb)
+ verbose=yes ;;
+
+ -version | --version | --versio | --versi | --vers | -V)
+ ac_init_version=: ;;
+
+ -with-* | --with-*)
+ ac_package=`expr "x$ac_option" : 'x-*with-\([^=]*\)'`
+ # Reject names that are not valid shell variable names.
+ expr "x$ac_package" : ".*[^-_$as_cr_alnum]" >/dev/null &&
+ { echo "$as_me: error: invalid package name: $ac_package" >&2
+ { (exit 1); exit 1; }; }
+ ac_package=`echo $ac_package| sed 's/-/_/g'`
+ case $ac_option in
+ *=*) ac_optarg=`echo "$ac_optarg" | sed "s/'/'\\\\\\\\''/g"`;;
+ *) ac_optarg=yes ;;
+ esac
+ eval "with_$ac_package='$ac_optarg'" ;;
+
+ -without-* | --without-*)
+ ac_package=`expr "x$ac_option" : 'x-*without-\(.*\)'`
+ # Reject names that are not valid shell variable names.
+ expr "x$ac_package" : ".*[^-_$as_cr_alnum]" >/dev/null &&
+ { echo "$as_me: error: invalid package name: $ac_package" >&2
+ { (exit 1); exit 1; }; }
+ ac_package=`echo $ac_package | sed 's/-/_/g'`
+ eval "with_$ac_package=no" ;;
+
+ --x)
+ # Obsolete; use --with-x.
+ with_x=yes ;;
+
+ -x-includes | --x-includes | --x-include | --x-includ | --x-inclu \
+ | --x-incl | --x-inc | --x-in | --x-i)
+ ac_prev=x_includes ;;
+ -x-includes=* | --x-includes=* | --x-include=* | --x-includ=* | --x-inclu=* \
+ | --x-incl=* | --x-inc=* | --x-in=* | --x-i=*)
+ x_includes=$ac_optarg ;;
+
+ -x-libraries | --x-libraries | --x-librarie | --x-librari \
+ | --x-librar | --x-libra | --x-libr | --x-lib | --x-li | --x-l)
+ ac_prev=x_libraries ;;
+ -x-libraries=* | --x-libraries=* | --x-librarie=* | --x-librari=* \
+ | --x-librar=* | --x-libra=* | --x-libr=* | --x-lib=* | --x-li=* | --x-l=*)
+ x_libraries=$ac_optarg ;;
+
+ -*) { echo "$as_me: error: unrecognized option: $ac_option
+Try \`$0 --help' for more information." >&2
+ { (exit 1); exit 1; }; }
+ ;;
+
+ *=*)
+ ac_envvar=`expr "x$ac_option" : 'x\([^=]*\)='`
+ # Reject names that are not valid shell variable names.
+ expr "x$ac_envvar" : ".*[^_$as_cr_alnum]" >/dev/null &&
+ { echo "$as_me: error: invalid variable name: $ac_envvar" >&2
+ { (exit 1); exit 1; }; }
+ ac_optarg=`echo "$ac_optarg" | sed "s/'/'\\\\\\\\''/g"`
+ eval "$ac_envvar='$ac_optarg'"
+ export $ac_envvar ;;
+
+ *)
+ # FIXME: should be removed in autoconf 3.0.
+ echo "$as_me: WARNING: you should use --build, --host, --target" >&2
+ expr "x$ac_option" : ".*[^-._$as_cr_alnum]" >/dev/null &&
+ echo "$as_me: WARNING: invalid host type: $ac_option" >&2
+ : ${build_alias=$ac_option} ${host_alias=$ac_option} ${target_alias=$ac_option}
+ ;;
+
+ esac
+done
+
+if test -n "$ac_prev"; then
+ ac_option=--`echo $ac_prev | sed 's/_/-/g'`
+ { echo "$as_me: error: missing argument to $ac_option" >&2
+ { (exit 1); exit 1; }; }
+fi
+
+# Be sure to have absolute paths.
+for ac_var in exec_prefix prefix
+do
+ eval ac_val=$`echo $ac_var`
+ case $ac_val in
+ [\\/$]* | ?:[\\/]* | NONE | '' ) ;;
+ *) { echo "$as_me: error: expected an absolute directory name for --$ac_var: $ac_val" >&2
+ { (exit 1); exit 1; }; };;
+ esac
+done
+
+# Be sure to have absolute paths.
+for ac_var in bindir sbindir libexecdir datadir sysconfdir sharedstatedir \
+ localstatedir libdir includedir oldincludedir infodir mandir
+do
+ eval ac_val=$`echo $ac_var`
+ case $ac_val in
+ [\\/$]* | ?:[\\/]* ) ;;
+ *) { echo "$as_me: error: expected an absolute directory name for --$ac_var: $ac_val" >&2
+ { (exit 1); exit 1; }; };;
+ esac
+done
+
+# There might be people who depend on the old broken behavior: `$host'
+# used to hold the argument of --host etc.
+# FIXME: To remove some day.
+build=$build_alias
+host=$host_alias
+target=$target_alias
+
+# FIXME: To remove some day.
+if test "x$host_alias" != x; then
+ if test "x$build_alias" = x; then
+ cross_compiling=maybe
+ echo "$as_me: WARNING: If you wanted to set the --build type, don't use --host.
+ If a cross compiler is detected then cross compile mode will be used." >&2
+ elif test "x$build_alias" != "x$host_alias"; then
+ cross_compiling=yes
+ fi
+fi
+
+ac_tool_prefix=
+test -n "$host_alias" && ac_tool_prefix=$host_alias-
+
+test "$silent" = yes && exec 6>/dev/null
+
+
+# Find the source files, if location was not specified.
+if test -z "$srcdir"; then
+ ac_srcdir_defaulted=yes
+ # Try the directory containing this script, then its parent.
+ ac_confdir=`(dirname "$0") 2>/dev/null ||
+$as_expr X"$0" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \
+ X"$0" : 'X\(//\)[^/]' \| \
+ X"$0" : 'X\(//\)$' \| \
+ X"$0" : 'X\(/\)' \| \
+ . : '\(.\)' 2>/dev/null ||
+echo X"$0" |
+ sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/; q; }
+ /^X\(\/\/\)[^/].*/{ s//\1/; q; }
+ /^X\(\/\/\)$/{ s//\1/; q; }
+ /^X\(\/\).*/{ s//\1/; q; }
+ s/.*/./; q'`
+ srcdir=$ac_confdir
+ if test ! -r $srcdir/$ac_unique_file; then
+ srcdir=..
+ fi
+else
+ ac_srcdir_defaulted=no
+fi
+if test ! -r $srcdir/$ac_unique_file; then
+ if test "$ac_srcdir_defaulted" = yes; then
+ { echo "$as_me: error: cannot find sources ($ac_unique_file) in $ac_confdir or .." >&2
+ { (exit 1); exit 1; }; }
+ else
+ { echo "$as_me: error: cannot find sources ($ac_unique_file) in $srcdir" >&2
+ { (exit 1); exit 1; }; }
+ fi
+fi
+(cd $srcdir && test -r ./$ac_unique_file) 2>/dev/null ||
+ { echo "$as_me: error: sources are in $srcdir, but \`cd $srcdir' does not work" >&2
+ { (exit 1); exit 1; }; }
+srcdir=`echo "$srcdir" | sed 's%\([^\\/]\)[\\/]*$%\1%'`
+ac_env_build_alias_set=${build_alias+set}
+ac_env_build_alias_value=$build_alias
+ac_cv_env_build_alias_set=${build_alias+set}
+ac_cv_env_build_alias_value=$build_alias
+ac_env_host_alias_set=${host_alias+set}
+ac_env_host_alias_value=$host_alias
+ac_cv_env_host_alias_set=${host_alias+set}
+ac_cv_env_host_alias_value=$host_alias
+ac_env_target_alias_set=${target_alias+set}
+ac_env_target_alias_value=$target_alias
+ac_cv_env_target_alias_set=${target_alias+set}
+ac_cv_env_target_alias_value=$target_alias
+ac_env_CC_set=${CC+set}
+ac_env_CC_value=$CC
+ac_cv_env_CC_set=${CC+set}
+ac_cv_env_CC_value=$CC
+ac_env_CFLAGS_set=${CFLAGS+set}
+ac_env_CFLAGS_value=$CFLAGS
+ac_cv_env_CFLAGS_set=${CFLAGS+set}
+ac_cv_env_CFLAGS_value=$CFLAGS
+ac_env_LDFLAGS_set=${LDFLAGS+set}
+ac_env_LDFLAGS_value=$LDFLAGS
+ac_cv_env_LDFLAGS_set=${LDFLAGS+set}
+ac_cv_env_LDFLAGS_value=$LDFLAGS
+ac_env_CPPFLAGS_set=${CPPFLAGS+set}
+ac_env_CPPFLAGS_value=$CPPFLAGS
+ac_cv_env_CPPFLAGS_set=${CPPFLAGS+set}
+ac_cv_env_CPPFLAGS_value=$CPPFLAGS
+
+#
+# Report the --help message.
+#
+if test "$ac_init_help" = "long"; then
+ # Omit some internal or obsolete options to make the list less imposing.
+ # This message is too long to be a string in the A/UX 3.1 sh.
+ cat <<_ACEOF
+\`configure' configures qdbm 1.8.77 to adapt to many kinds of systems.
+
+Usage: $0 [OPTION]... [VAR=VALUE]...
+
+To assign environment variables (e.g., CC, CFLAGS...), specify them as
+VAR=VALUE. See below for descriptions of some of the useful variables.
+
+Defaults for the options are specified in brackets.
+
+Configuration:
+ -h, --help display this help and exit
+ --help=short display options specific to this package
+ --help=recursive display the short help of all the included packages
+ -V, --version display version information and exit
+ -q, --quiet, --silent do not print \`checking...' messages
+ --cache-file=FILE cache test results in FILE [disabled]
+ -C, --config-cache alias for \`--cache-file=config.cache'
+ -n, --no-create do not create output files
+ --srcdir=DIR find the sources in DIR [configure dir or \`..']
+
+_ACEOF
+
+ cat <<_ACEOF
+Installation directories:
+ --prefix=PREFIX install architecture-independent files in PREFIX
+ [$ac_default_prefix]
+ --exec-prefix=EPREFIX install architecture-dependent files in EPREFIX
+ [PREFIX]
+
+By default, \`make install' will install all the files in
+\`$ac_default_prefix/bin', \`$ac_default_prefix/lib' etc. You can specify
+an installation prefix other than \`$ac_default_prefix' using \`--prefix',
+for instance \`--prefix=\$HOME'.
+
+For better control, use the options below.
+
+Fine tuning of the installation directories:
+ --bindir=DIR user executables [EPREFIX/bin]
+ --sbindir=DIR system admin executables [EPREFIX/sbin]
+ --libexecdir=DIR program executables [EPREFIX/libexec]
+ --datadir=DIR read-only architecture-independent data [PREFIX/share]
+ --sysconfdir=DIR read-only single-machine data [PREFIX/etc]
+ --sharedstatedir=DIR modifiable architecture-independent data [PREFIX/com]
+ --localstatedir=DIR modifiable single-machine data [PREFIX/var]
+ --libdir=DIR object code libraries [EPREFIX/lib]
+ --includedir=DIR C header files [PREFIX/include]
+ --oldincludedir=DIR C header files for non-gcc [/usr/include]
+ --infodir=DIR info documentation [PREFIX/info]
+ --mandir=DIR man documentation [PREFIX/man]
+_ACEOF
+
+ cat <<\_ACEOF
+_ACEOF
+fi
+
+if test -n "$ac_init_help"; then
+ case $ac_init_help in
+ short | recursive ) echo "Configuration of qdbm 1.8.77:";;
+ esac
+ cat <<\_ACEOF
+
+Optional Features:
+ --disable-FEATURE do not include FEATURE (same as --enable-FEATURE=no)
+ --enable-FEATURE[=ARG] include FEATURE [ARG=yes]
+ --enable-debug build for debugging
+ --enable-devel build for development
+ --enable-stable build for stable release
+ --enable-pthread use POSIX thread and make APIs thread-safe
+ --disable-lock build for environments without file locking
+ --disable-mmap build for environments without memory mapping
+ --enable-zlib feature ZLIB for B+ tree and inverted index
+ --enable-lzo feature LZO for B+ tree and inverted index
+ --enable-bzip feature BZIP2 for B+ tree and inverted index
+ --enable-iconv feature ICONV utilities
+ --disable-warn hide warnings in the configuration
+
+Some influential environment variables:
+ CC C compiler command
+ CFLAGS C compiler flags
+ LDFLAGS linker flags, e.g. -L<lib dir> if you have libraries in a
+ nonstandard directory <lib dir>
+ CPPFLAGS C/C++ preprocessor flags, e.g. -I<include dir> if you have
+ headers in a nonstandard directory <include dir>
+
+Use these variables to override the choices made by `configure' or to help
+it to find libraries and programs with nonstandard names/locations.
+
+_ACEOF
+fi
+
+if test "$ac_init_help" = "recursive"; then
+ # If there are subdirs, report their specific --help.
+ ac_popdir=`pwd`
+ for ac_dir in : $ac_subdirs_all; do test "x$ac_dir" = x: && continue
+ test -d $ac_dir || continue
+ ac_builddir=.
+
+if test "$ac_dir" != .; then
+ ac_dir_suffix=/`echo "$ac_dir" | sed 's,^\.[\\/],,'`
+ # A "../" for each directory in $ac_dir_suffix.
+ ac_top_builddir=`echo "$ac_dir_suffix" | sed 's,/[^\\/]*,../,g'`
+else
+ ac_dir_suffix= ac_top_builddir=
+fi
+
+case $srcdir in
+ .) # No --srcdir option. We are building in place.
+ ac_srcdir=.
+ if test -z "$ac_top_builddir"; then
+ ac_top_srcdir=.
+ else
+ ac_top_srcdir=`echo $ac_top_builddir | sed 's,/$,,'`
+ fi ;;
+ [\\/]* | ?:[\\/]* ) # Absolute path.
+ ac_srcdir=$srcdir$ac_dir_suffix;
+ ac_top_srcdir=$srcdir ;;
+ *) # Relative path.
+ ac_srcdir=$ac_top_builddir$srcdir$ac_dir_suffix
+ ac_top_srcdir=$ac_top_builddir$srcdir ;;
+esac
+
+# Do not use `cd foo && pwd` to compute absolute paths, because
+# the directories may not exist.
+case `pwd` in
+.) ac_abs_builddir="$ac_dir";;
+*)
+ case "$ac_dir" in
+ .) ac_abs_builddir=`pwd`;;
+ [\\/]* | ?:[\\/]* ) ac_abs_builddir="$ac_dir";;
+ *) ac_abs_builddir=`pwd`/"$ac_dir";;
+ esac;;
+esac
+case $ac_abs_builddir in
+.) ac_abs_top_builddir=${ac_top_builddir}.;;
+*)
+ case ${ac_top_builddir}. in
+ .) ac_abs_top_builddir=$ac_abs_builddir;;
+ [\\/]* | ?:[\\/]* ) ac_abs_top_builddir=${ac_top_builddir}.;;
+ *) ac_abs_top_builddir=$ac_abs_builddir/${ac_top_builddir}.;;
+ esac;;
+esac
+case $ac_abs_builddir in
+.) ac_abs_srcdir=$ac_srcdir;;
+*)
+ case $ac_srcdir in
+ .) ac_abs_srcdir=$ac_abs_builddir;;
+ [\\/]* | ?:[\\/]* ) ac_abs_srcdir=$ac_srcdir;;
+ *) ac_abs_srcdir=$ac_abs_builddir/$ac_srcdir;;
+ esac;;
+esac
+case $ac_abs_builddir in
+.) ac_abs_top_srcdir=$ac_top_srcdir;;
+*)
+ case $ac_top_srcdir in
+ .) ac_abs_top_srcdir=$ac_abs_builddir;;
+ [\\/]* | ?:[\\/]* ) ac_abs_top_srcdir=$ac_top_srcdir;;
+ *) ac_abs_top_srcdir=$ac_abs_builddir/$ac_top_srcdir;;
+ esac;;
+esac
+
+ cd $ac_dir
+ # Check for guested configure; otherwise get Cygnus style configure.
+ if test -f $ac_srcdir/configure.gnu; then
+ echo
+ $SHELL $ac_srcdir/configure.gnu --help=recursive
+ elif test -f $ac_srcdir/configure; then
+ echo
+ $SHELL $ac_srcdir/configure --help=recursive
+ elif test -f $ac_srcdir/configure.ac ||
+ test -f $ac_srcdir/configure.in; then
+ echo
+ $ac_configure --help
+ else
+ echo "$as_me: WARNING: no configuration information is in $ac_dir" >&2
+ fi
+ cd $ac_popdir
+ done
+fi
+
+test -n "$ac_init_help" && exit 0
+if $ac_init_version; then
+ cat <<\_ACEOF
+qdbm configure 1.8.77
+generated by GNU Autoconf 2.59
+
+Copyright (C) 2003 Free Software Foundation, Inc.
+This configure script is free software; the Free Software Foundation
+gives unlimited permission to copy, distribute and modify it.
+_ACEOF
+ exit 0
+fi
+exec 5>config.log
+cat >&5 <<_ACEOF
+This file contains any messages produced by compilers while
+running configure, to aid debugging if configure makes a mistake.
+
+It was created by qdbm $as_me 1.8.77, which was
+generated by GNU Autoconf 2.59. Invocation command line was
+
+ $ $0 $@
+
+_ACEOF
+{
+cat <<_ASUNAME
+## --------- ##
+## Platform. ##
+## --------- ##
+
+hostname = `(hostname || uname -n) 2>/dev/null | sed 1q`
+uname -m = `(uname -m) 2>/dev/null || echo unknown`
+uname -r = `(uname -r) 2>/dev/null || echo unknown`
+uname -s = `(uname -s) 2>/dev/null || echo unknown`
+uname -v = `(uname -v) 2>/dev/null || echo unknown`
+
+/usr/bin/uname -p = `(/usr/bin/uname -p) 2>/dev/null || echo unknown`
+/bin/uname -X = `(/bin/uname -X) 2>/dev/null || echo unknown`
+
+/bin/arch = `(/bin/arch) 2>/dev/null || echo unknown`
+/usr/bin/arch -k = `(/usr/bin/arch -k) 2>/dev/null || echo unknown`
+/usr/convex/getsysinfo = `(/usr/convex/getsysinfo) 2>/dev/null || echo unknown`
+hostinfo = `(hostinfo) 2>/dev/null || echo unknown`
+/bin/machine = `(/bin/machine) 2>/dev/null || echo unknown`
+/usr/bin/oslevel = `(/usr/bin/oslevel) 2>/dev/null || echo unknown`
+/bin/universe = `(/bin/universe) 2>/dev/null || echo unknown`
+
+_ASUNAME
+
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ echo "PATH: $as_dir"
+done
+
+} >&5
+
+cat >&5 <<_ACEOF
+
+
+## ----------- ##
+## Core tests. ##
+## ----------- ##
+
+_ACEOF
+
+
+# Keep a trace of the command line.
+# Strip out --no-create and --no-recursion so they do not pile up.
+# Strip out --silent because we don't want to record it for future runs.
+# Also quote any args containing shell meta-characters.
+# Make two passes to allow for proper duplicate-argument suppression.
+ac_configure_args=
+ac_configure_args0=
+ac_configure_args1=
+ac_sep=
+ac_must_keep_next=false
+for ac_pass in 1 2
+do
+ for ac_arg
+ do
+ case $ac_arg in
+ -no-create | --no-c* | -n | -no-recursion | --no-r*) continue ;;
+ -q | -quiet | --quiet | --quie | --qui | --qu | --q \
+ | -silent | --silent | --silen | --sile | --sil)
+ continue ;;
+ *" "*|*" "*|*[\[\]\~\#\$\^\&\*\(\)\{\}\\\|\;\<\>\?\"\']*)
+ ac_arg=`echo "$ac_arg" | sed "s/'/'\\\\\\\\''/g"` ;;
+ esac
+ case $ac_pass in
+ 1) ac_configure_args0="$ac_configure_args0 '$ac_arg'" ;;
+ 2)
+ ac_configure_args1="$ac_configure_args1 '$ac_arg'"
+ if test $ac_must_keep_next = true; then
+ ac_must_keep_next=false # Got value, back to normal.
+ else
+ case $ac_arg in
+ *=* | --config-cache | -C | -disable-* | --disable-* \
+ | -enable-* | --enable-* | -gas | --g* | -nfp | --nf* \
+ | -q | -quiet | --q* | -silent | --sil* | -v | -verb* \
+ | -with-* | --with-* | -without-* | --without-* | --x)
+ case "$ac_configure_args0 " in
+ "$ac_configure_args1"*" '$ac_arg' "* ) continue ;;
+ esac
+ ;;
+ -* ) ac_must_keep_next=true ;;
+ esac
+ fi
+ ac_configure_args="$ac_configure_args$ac_sep'$ac_arg'"
+ # Get rid of the leading space.
+ ac_sep=" "
+ ;;
+ esac
+ done
+done
+$as_unset ac_configure_args0 || test "${ac_configure_args0+set}" != set || { ac_configure_args0=; export ac_configure_args0; }
+$as_unset ac_configure_args1 || test "${ac_configure_args1+set}" != set || { ac_configure_args1=; export ac_configure_args1; }
+
+# When interrupted or exit'd, cleanup temporary files, and complete
+# config.log. We remove comments because anyway the quotes in there
+# would cause problems or look ugly.
+# WARNING: Be sure not to use single quotes in there, as some shells,
+# such as our DU 5.0 friend, will then `close' the trap.
+trap 'exit_status=$?
+ # Save into config.log some information that might help in debugging.
+ {
+ echo
+
+ cat <<\_ASBOX
+## ---------------- ##
+## Cache variables. ##
+## ---------------- ##
+_ASBOX
+ echo
+ # The following way of writing the cache mishandles newlines in values,
+{
+ (set) 2>&1 |
+ case `(ac_space='"'"' '"'"'; set | grep ac_space) 2>&1` in
+ *ac_space=\ *)
+ sed -n \
+ "s/'"'"'/'"'"'\\\\'"'"''"'"'/g;
+ s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1='"'"'\\2'"'"'/p"
+ ;;
+ *)
+ sed -n \
+ "s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1=\\2/p"
+ ;;
+ esac;
+}
+ echo
+
+ cat <<\_ASBOX
+## ----------------- ##
+## Output variables. ##
+## ----------------- ##
+_ASBOX
+ echo
+ for ac_var in $ac_subst_vars
+ do
+ eval ac_val=$`echo $ac_var`
+ echo "$ac_var='"'"'$ac_val'"'"'"
+ done | sort
+ echo
+
+ if test -n "$ac_subst_files"; then
+ cat <<\_ASBOX
+## ------------- ##
+## Output files. ##
+## ------------- ##
+_ASBOX
+ echo
+ for ac_var in $ac_subst_files
+ do
+ eval ac_val=$`echo $ac_var`
+ echo "$ac_var='"'"'$ac_val'"'"'"
+ done | sort
+ echo
+ fi
+
+ if test -s confdefs.h; then
+ cat <<\_ASBOX
+## ----------- ##
+## confdefs.h. ##
+## ----------- ##
+_ASBOX
+ echo
+ sed "/^$/d" confdefs.h | sort
+ echo
+ fi
+ test "$ac_signal" != 0 &&
+ echo "$as_me: caught signal $ac_signal"
+ echo "$as_me: exit $exit_status"
+ } >&5
+ rm -f core *.core &&
+ rm -rf conftest* confdefs* conf$$* $ac_clean_files &&
+ exit $exit_status
+ ' 0
+for ac_signal in 1 2 13 15; do
+ trap 'ac_signal='$ac_signal'; { (exit 1); exit 1; }' $ac_signal
+done
+ac_signal=0
+
+# confdefs.h avoids OS command line length limits that DEFS can exceed.
+rm -rf conftest* confdefs.h
+# AIX cpp loses on an empty file, so make sure it contains at least a newline.
+echo >confdefs.h
+
+# Predefined preprocessor variables.
+
+cat >>confdefs.h <<_ACEOF
+#define PACKAGE_NAME "$PACKAGE_NAME"
+_ACEOF
+
+
+cat >>confdefs.h <<_ACEOF
+#define PACKAGE_TARNAME "$PACKAGE_TARNAME"
+_ACEOF
+
+
+cat >>confdefs.h <<_ACEOF
+#define PACKAGE_VERSION "$PACKAGE_VERSION"
+_ACEOF
+
+
+cat >>confdefs.h <<_ACEOF
+#define PACKAGE_STRING "$PACKAGE_STRING"
+_ACEOF
+
+
+cat >>confdefs.h <<_ACEOF
+#define PACKAGE_BUGREPORT "$PACKAGE_BUGREPORT"
+_ACEOF
+
+
+# Let the site file select an alternate cache file if it wants to.
+# Prefer explicitly selected file to automatically selected ones.
+if test -z "$CONFIG_SITE"; then
+ if test "x$prefix" != xNONE; then
+ CONFIG_SITE="$prefix/share/config.site $prefix/etc/config.site"
+ else
+ CONFIG_SITE="$ac_default_prefix/share/config.site $ac_default_prefix/etc/config.site"
+ fi
+fi
+for ac_site_file in $CONFIG_SITE; do
+ if test -r "$ac_site_file"; then
+ { echo "$as_me:$LINENO: loading site script $ac_site_file" >&5
+echo "$as_me: loading site script $ac_site_file" >&6;}
+ sed 's/^/| /' "$ac_site_file" >&5
+ . "$ac_site_file"
+ fi
+done
+
+if test -r "$cache_file"; then
+ # Some versions of bash will fail to source /dev/null (special
+ # files actually), so we avoid doing that.
+ if test -f "$cache_file"; then
+ { echo "$as_me:$LINENO: loading cache $cache_file" >&5
+echo "$as_me: loading cache $cache_file" >&6;}
+ case $cache_file in
+ [\\/]* | ?:[\\/]* ) . $cache_file;;
+ *) . ./$cache_file;;
+ esac
+ fi
+else
+ { echo "$as_me:$LINENO: creating cache $cache_file" >&5
+echo "$as_me: creating cache $cache_file" >&6;}
+ >$cache_file
+fi
+
+# Check that the precious variables saved in the cache have kept the same
+# value.
+ac_cache_corrupted=false
+for ac_var in `(set) 2>&1 |
+ sed -n 's/^ac_env_\([a-zA-Z_0-9]*\)_set=.*/\1/p'`; do
+ eval ac_old_set=\$ac_cv_env_${ac_var}_set
+ eval ac_new_set=\$ac_env_${ac_var}_set
+ eval ac_old_val="\$ac_cv_env_${ac_var}_value"
+ eval ac_new_val="\$ac_env_${ac_var}_value"
+ case $ac_old_set,$ac_new_set in
+ set,)
+ { echo "$as_me:$LINENO: error: \`$ac_var' was set to \`$ac_old_val' in the previous run" >&5
+echo "$as_me: error: \`$ac_var' was set to \`$ac_old_val' in the previous run" >&2;}
+ ac_cache_corrupted=: ;;
+ ,set)
+ { echo "$as_me:$LINENO: error: \`$ac_var' was not set in the previous run" >&5
+echo "$as_me: error: \`$ac_var' was not set in the previous run" >&2;}
+ ac_cache_corrupted=: ;;
+ ,);;
+ *)
+ if test "x$ac_old_val" != "x$ac_new_val"; then
+ { echo "$as_me:$LINENO: error: \`$ac_var' has changed since the previous run:" >&5
+echo "$as_me: error: \`$ac_var' has changed since the previous run:" >&2;}
+ { echo "$as_me:$LINENO: former value: $ac_old_val" >&5
+echo "$as_me: former value: $ac_old_val" >&2;}
+ { echo "$as_me:$LINENO: current value: $ac_new_val" >&5
+echo "$as_me: current value: $ac_new_val" >&2;}
+ ac_cache_corrupted=:
+ fi;;
+ esac
+ # Pass precious variables to config.status.
+ if test "$ac_new_set" = set; then
+ case $ac_new_val in
+ *" "*|*" "*|*[\[\]\~\#\$\^\&\*\(\)\{\}\\\|\;\<\>\?\"\']*)
+ ac_arg=$ac_var=`echo "$ac_new_val" | sed "s/'/'\\\\\\\\''/g"` ;;
+ *) ac_arg=$ac_var=$ac_new_val ;;
+ esac
+ case " $ac_configure_args " in
+ *" '$ac_arg' "*) ;; # Avoid dups. Use of quotes ensures accuracy.
+ *) ac_configure_args="$ac_configure_args '$ac_arg'" ;;
+ esac
+ fi
+done
+if $ac_cache_corrupted; then
+ { echo "$as_me:$LINENO: error: changes in the environment can compromise the build" >&5
+echo "$as_me: error: changes in the environment can compromise the build" >&2;}
+ { { echo "$as_me:$LINENO: error: run \`make distclean' and/or \`rm $cache_file' and start over" >&5
+echo "$as_me: error: run \`make distclean' and/or \`rm $cache_file' and start over" >&2;}
+ { (exit 1); exit 1; }; }
+fi
+
+ac_ext=c
+ac_cpp='$CPP $CPPFLAGS'
+ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5'
+ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
+ac_compiler_gnu=$ac_cv_c_compiler_gnu
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+# Export variables
+LIBVER=14
+LIBREV=13
+TARGETS="all"
+MYDEFS=""
+MYOPTS=""
+MGWLIBS=""
+LD="ld"
+AR="ar"
+
+# Building paths
+pathtmp="$PATH"
+PATH="$HOME/bin:/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/local/sbin"
+PATH="$PATH:/usr/ccs/bin:/usr/ucb:/usr/xpg4/bin:/usr/xpg6/bin:$pathtmp"
+LIBRARY_PATH="$HOME/lib:/usr/local/lib:$LIBRARY_PATH"
+LD_LIBRARY_PATH="$HOME/lib:/usr/local/lib:$LD_LIBRARY_PATH"
+CPATH="$HOME/include:/usr/local/include:$CPATH"
+export PATH LIBRARY_PATH LD_LIBRARY_PATH CPATH
+
+
+
+#================================================================
+# Options
+#================================================================
+
+
+# Internal variables
+enables=""
+ispthread=""
+iszlib=""
+isiconv=""
+isnowarn=""
+
+# Debug mode
+# Check whether --enable-debug or --disable-debug was given.
+if test "${enable_debug+set}" = set; then
+ enableval="$enable_debug"
+
+fi;
+if test "$enable_debug" = "yes"
+then
+ TARGETS="debug"
+ enables="$enables (debug)"
+fi
+
+# Developping mode
+# Check whether --enable-devel or --disable-devel was given.
+if test "${enable_devel+set}" = set; then
+ enableval="$enable_devel"
+
+fi;
+if test "$enable_devel" = "yes"
+then
+ TARGETS="devel"
+ enables="$enables (devel)"
+fi
+
+# Stable mode
+# Check whether --enable-stable or --disable-stable was given.
+if test "${enable_stable+set}" = set; then
+ enableval="$enable_stable"
+
+fi;
+if test "$enable_stable" = "yes"
+then
+ TARGETS="stable"
+ enables="$enables (stable)"
+fi
+
+# Enable POSIX thread
+# Check whether --enable-pthread or --disable-pthread was given.
+if test "${enable_pthread+set}" = set; then
+ enableval="$enable_pthread"
+
+fi;
+if test "$enable_pthread" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYPTHREAD"
+ enables="$enables (pthread)"
+ ispthread="yes"
+fi
+
+# Disable file locking
+# Check whether --enable-lock or --disable-lock was given.
+if test "${enable_lock+set}" = set; then
+ enableval="$enable_lock"
+
+fi;
+if test "$enable_lock" = "no"
+then
+ MYDEFS="$MYDEFS -DMYNOLOCK"
+ enables="$enables (no-lock)"
+fi
+
+# Disable memory mapping
+# Check whether --enable-mmap or --disable-mmap was given.
+if test "${enable_mmap+set}" = set; then
+ enableval="$enable_mmap"
+
+fi;
+if test "$enable_mmap" = "no"
+then
+ MYDEFS="$MYDEFS -DMYNOMMAP"
+ enables="$enables (no-mmap)"
+fi
+
+# Enable ZLIB compression
+# Check whether --enable-zlib or --disable-zlib was given.
+if test "${enable_zlib+set}" = set; then
+ enableval="$enable_zlib"
+
+fi;
+if test "$enable_zlib" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYZLIB"
+ MGWLIBS="-lz $MGWLIBS"
+ enables="$enables (zlib)"
+ iszlib="yes"
+fi
+
+# Enable LZO compression
+# Check whether --enable-lzo or --disable-lzo was given.
+if test "${enable_lzo+set}" = set; then
+ enableval="$enable_lzo"
+
+fi;
+if test "$enable_lzo" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYLZO"
+ MGWLIBS="-llzo2 $MGWLIBS"
+ enables="$enables (lzo)"
+ islzo="yes"
+fi
+
+# Enable BZIP2 compression
+# Check whether --enable-bzip or --disable-bzip was given.
+if test "${enable_bzip+set}" = set; then
+ enableval="$enable_bzip"
+
+fi;
+if test "$enable_bzip" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYBZIP"
+ MGWLIBS="-lbz2 $MGWLIBS"
+ enables="$enables (bzip)"
+ isbzip="yes"
+fi
+
+# Enable ICONV utilities
+# Check whether --enable-iconv or --disable-iconv was given.
+if test "${enable_iconv+set}" = set; then
+ enableval="$enable_iconv"
+
+fi;
+if test "$enable_iconv" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYICONV"
+ MGWLIBS="-liconv $MGWLIBS"
+ enables="$enables (iconv)"
+ isiconv="yes"
+fi
+
+# No warning configuration
+# Check whether --enable-warn or --disable-warn was given.
+if test "${enable_warn+set}" = set; then
+ enableval="$enable_warn"
+
+fi;
+if test "$enable_warn" = "no"
+then
+ isnowarn="yes"
+fi
+
+# Messages
+printf '#================================================================\n'
+printf '# Configuring QDBM version %s%s.\n' "$PACKAGE_VERSION" "$enables"
+printf '#================================================================\n'
+
+
+
+#================================================================
+# Checking Commands to Build with
+#================================================================
+
+
+# C compiler
+ac_ext=c
+ac_cpp='$CPP $CPPFLAGS'
+ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5'
+ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
+ac_compiler_gnu=$ac_cv_c_compiler_gnu
+if test -n "$ac_tool_prefix"; then
+ # Extract the first word of "${ac_tool_prefix}gcc", so it can be a program name with args.
+set dummy ${ac_tool_prefix}gcc; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$CC"; then
+ ac_cv_prog_CC="$CC" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ ac_cv_prog_CC="${ac_tool_prefix}gcc"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+fi
+fi
+CC=$ac_cv_prog_CC
+if test -n "$CC"; then
+ echo "$as_me:$LINENO: result: $CC" >&5
+echo "${ECHO_T}$CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+fi
+if test -z "$ac_cv_prog_CC"; then
+ ac_ct_CC=$CC
+ # Extract the first word of "gcc", so it can be a program name with args.
+set dummy gcc; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_ac_ct_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$ac_ct_CC"; then
+ ac_cv_prog_ac_ct_CC="$ac_ct_CC" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ ac_cv_prog_ac_ct_CC="gcc"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+fi
+fi
+ac_ct_CC=$ac_cv_prog_ac_ct_CC
+if test -n "$ac_ct_CC"; then
+ echo "$as_me:$LINENO: result: $ac_ct_CC" >&5
+echo "${ECHO_T}$ac_ct_CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+ CC=$ac_ct_CC
+else
+ CC="$ac_cv_prog_CC"
+fi
+
+if test -z "$CC"; then
+ if test -n "$ac_tool_prefix"; then
+ # Extract the first word of "${ac_tool_prefix}cc", so it can be a program name with args.
+set dummy ${ac_tool_prefix}cc; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$CC"; then
+ ac_cv_prog_CC="$CC" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ ac_cv_prog_CC="${ac_tool_prefix}cc"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+fi
+fi
+CC=$ac_cv_prog_CC
+if test -n "$CC"; then
+ echo "$as_me:$LINENO: result: $CC" >&5
+echo "${ECHO_T}$CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+fi
+if test -z "$ac_cv_prog_CC"; then
+ ac_ct_CC=$CC
+ # Extract the first word of "cc", so it can be a program name with args.
+set dummy cc; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_ac_ct_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$ac_ct_CC"; then
+ ac_cv_prog_ac_ct_CC="$ac_ct_CC" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ ac_cv_prog_ac_ct_CC="cc"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+fi
+fi
+ac_ct_CC=$ac_cv_prog_ac_ct_CC
+if test -n "$ac_ct_CC"; then
+ echo "$as_me:$LINENO: result: $ac_ct_CC" >&5
+echo "${ECHO_T}$ac_ct_CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+ CC=$ac_ct_CC
+else
+ CC="$ac_cv_prog_CC"
+fi
+
+fi
+if test -z "$CC"; then
+ # Extract the first word of "cc", so it can be a program name with args.
+set dummy cc; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$CC"; then
+ ac_cv_prog_CC="$CC" # Let the user override the test.
+else
+ ac_prog_rejected=no
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ if test "$as_dir/$ac_word$ac_exec_ext" = "/usr/ucb/cc"; then
+ ac_prog_rejected=yes
+ continue
+ fi
+ ac_cv_prog_CC="cc"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+if test $ac_prog_rejected = yes; then
+ # We found a bogon in the path, so make sure we never use it.
+ set dummy $ac_cv_prog_CC
+ shift
+ if test $# != 0; then
+ # We chose a different compiler from the bogus one.
+ # However, it has the same basename, so the bogon will be chosen
+ # first if we set CC to just the basename; use the full file name.
+ shift
+ ac_cv_prog_CC="$as_dir/$ac_word${1+' '}$@"
+ fi
+fi
+fi
+fi
+CC=$ac_cv_prog_CC
+if test -n "$CC"; then
+ echo "$as_me:$LINENO: result: $CC" >&5
+echo "${ECHO_T}$CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+fi
+if test -z "$CC"; then
+ if test -n "$ac_tool_prefix"; then
+ for ac_prog in cl
+ do
+ # Extract the first word of "$ac_tool_prefix$ac_prog", so it can be a program name with args.
+set dummy $ac_tool_prefix$ac_prog; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$CC"; then
+ ac_cv_prog_CC="$CC" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ ac_cv_prog_CC="$ac_tool_prefix$ac_prog"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+fi
+fi
+CC=$ac_cv_prog_CC
+if test -n "$CC"; then
+ echo "$as_me:$LINENO: result: $CC" >&5
+echo "${ECHO_T}$CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+ test -n "$CC" && break
+ done
+fi
+if test -z "$CC"; then
+ ac_ct_CC=$CC
+ for ac_prog in cl
+do
+ # Extract the first word of "$ac_prog", so it can be a program name with args.
+set dummy $ac_prog; ac_word=$2
+echo "$as_me:$LINENO: checking for $ac_word" >&5
+echo $ECHO_N "checking for $ac_word... $ECHO_C" >&6
+if test "${ac_cv_prog_ac_ct_CC+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ if test -n "$ac_ct_CC"; then
+ ac_cv_prog_ac_ct_CC="$ac_ct_CC" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for ac_exec_ext in '' $ac_executable_extensions; do
+ if $as_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+ ac_cv_prog_ac_ct_CC="$ac_prog"
+ echo "$as_me:$LINENO: found $as_dir/$ac_word$ac_exec_ext" >&5
+ break 2
+ fi
+done
+done
+
+fi
+fi
+ac_ct_CC=$ac_cv_prog_ac_ct_CC
+if test -n "$ac_ct_CC"; then
+ echo "$as_me:$LINENO: result: $ac_ct_CC" >&5
+echo "${ECHO_T}$ac_ct_CC" >&6
+else
+ echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
+
+ test -n "$ac_ct_CC" && break
+done
+
+ CC=$ac_ct_CC
+fi
+
+fi
+
+
+test -z "$CC" && { { echo "$as_me:$LINENO: error: no acceptable C compiler found in \$PATH
+See \`config.log' for more details." >&5
+echo "$as_me: error: no acceptable C compiler found in \$PATH
+See \`config.log' for more details." >&2;}
+ { (exit 1); exit 1; }; }
+
+# Provide some information about the compiler.
+echo "$as_me:$LINENO:" \
+ "checking for C compiler version" >&5
+ac_compiler=`set X $ac_compile; echo $2`
+{ (eval echo "$as_me:$LINENO: \"$ac_compiler --version </dev/null >&5\"") >&5
+ (eval $ac_compiler --version </dev/null >&5) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }
+{ (eval echo "$as_me:$LINENO: \"$ac_compiler -v </dev/null >&5\"") >&5
+ (eval $ac_compiler -v </dev/null >&5) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }
+{ (eval echo "$as_me:$LINENO: \"$ac_compiler -V </dev/null >&5\"") >&5
+ (eval $ac_compiler -V </dev/null >&5) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }
+
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+int
+main ()
+{
+
+ ;
+ return 0;
+}
+_ACEOF
+ac_clean_files_save=$ac_clean_files
+ac_clean_files="$ac_clean_files a.out a.exe b.out"
+# Try to create an executable without -o first, disregard a.out.
+# It will help us diagnose broken compilers, and finding out an intuition
+# of exeext.
+echo "$as_me:$LINENO: checking for C compiler default output file name" >&5
+echo $ECHO_N "checking for C compiler default output file name... $ECHO_C" >&6
+ac_link_default=`echo "$ac_link" | sed 's/ -o *conftest[^ ]*//'`
+if { (eval echo "$as_me:$LINENO: \"$ac_link_default\"") >&5
+ (eval $ac_link_default) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; then
+ # Find the output, starting from the most likely. This scheme is
+# not robust to junk in `.', hence go to wildcards (a.*) only as a last
+# resort.
+
+# Be careful to initialize this variable, since it used to be cached.
+# Otherwise an old cache value of `no' led to `EXEEXT = no' in a Makefile.
+ac_cv_exeext=
+# b.out is created by i960 compilers.
+for ac_file in a_out.exe a.exe conftest.exe a.out conftest a.* conftest.* b.out
+do
+ test -f "$ac_file" || continue
+ case $ac_file in
+ *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.o | *.obj )
+ ;;
+ conftest.$ac_ext )
+ # This is the source file.
+ ;;
+ [ab].out )
+ # We found the default executable, but exeext='' is most
+ # certainly right.
+ break;;
+ *.* )
+ ac_cv_exeext=`expr "$ac_file" : '[^.]*\(\..*\)'`
+ # FIXME: I believe we export ac_cv_exeext for Libtool,
+ # but it would be cool to find out if it's true. Does anybody
+ # maintain Libtool? --akim.
+ export ac_cv_exeext
+ break;;
+ * )
+ break;;
+ esac
+done
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+{ { echo "$as_me:$LINENO: error: C compiler cannot create executables
+See \`config.log' for more details." >&5
+echo "$as_me: error: C compiler cannot create executables
+See \`config.log' for more details." >&2;}
+ { (exit 77); exit 77; }; }
+fi
+
+ac_exeext=$ac_cv_exeext
+echo "$as_me:$LINENO: result: $ac_file" >&5
+echo "${ECHO_T}$ac_file" >&6
+
+# Check the compiler produces executables we can run. If not, either
+# the compiler is broken, or we cross compile.
+echo "$as_me:$LINENO: checking whether the C compiler works" >&5
+echo $ECHO_N "checking whether the C compiler works... $ECHO_C" >&6
+# FIXME: These cross compiler hacks should be removed for Autoconf 3.0
+# If not cross compiling, check that we can run a simple program.
+if test "$cross_compiling" != yes; then
+ if { ac_try='./$ac_file'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ cross_compiling=no
+ else
+ if test "$cross_compiling" = maybe; then
+ cross_compiling=yes
+ else
+ { { echo "$as_me:$LINENO: error: cannot run C compiled programs.
+If you meant to cross compile, use \`--host'.
+See \`config.log' for more details." >&5
+echo "$as_me: error: cannot run C compiled programs.
+If you meant to cross compile, use \`--host'.
+See \`config.log' for more details." >&2;}
+ { (exit 1); exit 1; }; }
+ fi
+ fi
+fi
+echo "$as_me:$LINENO: result: yes" >&5
+echo "${ECHO_T}yes" >&6
+
+rm -f a.out a.exe conftest$ac_cv_exeext b.out
+ac_clean_files=$ac_clean_files_save
+# Check the compiler produces executables we can run. If not, either
+# the compiler is broken, or we cross compile.
+echo "$as_me:$LINENO: checking whether we are cross compiling" >&5
+echo $ECHO_N "checking whether we are cross compiling... $ECHO_C" >&6
+echo "$as_me:$LINENO: result: $cross_compiling" >&5
+echo "${ECHO_T}$cross_compiling" >&6
+
+echo "$as_me:$LINENO: checking for suffix of executables" >&5
+echo $ECHO_N "checking for suffix of executables... $ECHO_C" >&6
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; then
+ # If both `conftest.exe' and `conftest' are `present' (well, observable)
+# catch `conftest.exe'. For instance with Cygwin, `ls conftest' will
+# work properly (i.e., refer to `conftest.exe'), while it won't with
+# `rm'.
+for ac_file in conftest.exe conftest conftest.*; do
+ test -f "$ac_file" || continue
+ case $ac_file in
+ *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg | *.o | *.obj ) ;;
+ *.* ) ac_cv_exeext=`expr "$ac_file" : '[^.]*\(\..*\)'`
+ export ac_cv_exeext
+ break;;
+ * ) break;;
+ esac
+done
+else
+ { { echo "$as_me:$LINENO: error: cannot compute suffix of executables: cannot compile and link
+See \`config.log' for more details." >&5
+echo "$as_me: error: cannot compute suffix of executables: cannot compile and link
+See \`config.log' for more details." >&2;}
+ { (exit 1); exit 1; }; }
+fi
+
+rm -f conftest$ac_cv_exeext
+echo "$as_me:$LINENO: result: $ac_cv_exeext" >&5
+echo "${ECHO_T}$ac_cv_exeext" >&6
+
+rm -f conftest.$ac_ext
+EXEEXT=$ac_cv_exeext
+ac_exeext=$EXEEXT
+echo "$as_me:$LINENO: checking for suffix of object files" >&5
+echo $ECHO_N "checking for suffix of object files... $ECHO_C" >&6
+if test "${ac_cv_objext+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+int
+main ()
+{
+
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.o conftest.obj
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; then
+ for ac_file in `(ls conftest.o conftest.obj; ls conftest.*) 2>/dev/null`; do
+ case $ac_file in
+ *.$ac_ext | *.xcoff | *.tds | *.d | *.pdb | *.xSYM | *.bb | *.bbg ) ;;
+ *) ac_cv_objext=`expr "$ac_file" : '.*\.\(.*\)'`
+ break;;
+ esac
+done
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+{ { echo "$as_me:$LINENO: error: cannot compute suffix of object files: cannot compile
+See \`config.log' for more details." >&5
+echo "$as_me: error: cannot compute suffix of object files: cannot compile
+See \`config.log' for more details." >&2;}
+ { (exit 1); exit 1; }; }
+fi
+
+rm -f conftest.$ac_cv_objext conftest.$ac_ext
+fi
+echo "$as_me:$LINENO: result: $ac_cv_objext" >&5
+echo "${ECHO_T}$ac_cv_objext" >&6
+OBJEXT=$ac_cv_objext
+ac_objext=$OBJEXT
+echo "$as_me:$LINENO: checking whether we are using the GNU C compiler" >&5
+echo $ECHO_N "checking whether we are using the GNU C compiler... $ECHO_C" >&6
+if test "${ac_cv_c_compiler_gnu+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+int
+main ()
+{
+#ifndef __GNUC__
+ choke me
+#endif
+
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest.$ac_objext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_compiler_gnu=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_compiler_gnu=no
+fi
+rm -f conftest.err conftest.$ac_objext conftest.$ac_ext
+ac_cv_c_compiler_gnu=$ac_compiler_gnu
+
+fi
+echo "$as_me:$LINENO: result: $ac_cv_c_compiler_gnu" >&5
+echo "${ECHO_T}$ac_cv_c_compiler_gnu" >&6
+GCC=`test $ac_compiler_gnu = yes && echo yes`
+ac_test_CFLAGS=${CFLAGS+set}
+ac_save_CFLAGS=$CFLAGS
+CFLAGS="-g"
+echo "$as_me:$LINENO: checking whether $CC accepts -g" >&5
+echo $ECHO_N "checking whether $CC accepts -g... $ECHO_C" >&6
+if test "${ac_cv_prog_cc_g+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+int
+main ()
+{
+
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest.$ac_objext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_prog_cc_g=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_prog_cc_g=no
+fi
+rm -f conftest.err conftest.$ac_objext conftest.$ac_ext
+fi
+echo "$as_me:$LINENO: result: $ac_cv_prog_cc_g" >&5
+echo "${ECHO_T}$ac_cv_prog_cc_g" >&6
+if test "$ac_test_CFLAGS" = set; then
+ CFLAGS=$ac_save_CFLAGS
+elif test $ac_cv_prog_cc_g = yes; then
+ if test "$GCC" = yes; then
+ CFLAGS="-g -O2"
+ else
+ CFLAGS="-g"
+ fi
+else
+ if test "$GCC" = yes; then
+ CFLAGS="-O2"
+ else
+ CFLAGS=
+ fi
+fi
+echo "$as_me:$LINENO: checking for $CC option to accept ANSI C" >&5
+echo $ECHO_N "checking for $CC option to accept ANSI C... $ECHO_C" >&6
+if test "${ac_cv_prog_cc_stdc+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_cv_prog_cc_stdc=no
+ac_save_CC=$CC
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+#include <stdarg.h>
+#include <stdio.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+/* Most of the following tests are stolen from RCS 5.7's src/conf.sh. */
+struct buf { int x; };
+FILE * (*rcsopen) (struct buf *, struct stat *, int);
+static char *e (p, i)
+ char **p;
+ int i;
+{
+ return p[i];
+}
+static char *f (char * (*g) (char **, int), char **p, ...)
+{
+ char *s;
+ va_list v;
+ va_start (v,p);
+ s = g (p, va_arg (v,int));
+ va_end (v);
+ return s;
+}
+
+/* OSF 4.0 Compaq cc is some sort of almost-ANSI by default. It has
+ function prototypes and stuff, but not '\xHH' hex character constants.
+ These don't provoke an error unfortunately, instead are silently treated
+ as 'x'. The following induces an error, until -std1 is added to get
+ proper ANSI mode. Curiously '\x00'!='x' always comes out true, for an
+ array size at least. It's necessary to write '\x00'==0 to get something
+ that's true only with -std1. */
+int osf4_cc_array ['\x00' == 0 ? 1 : -1];
+
+int test (int i, double x);
+struct s1 {int (*f) (int a);};
+struct s2 {int (*f) (double a);};
+int pairnames (int, char **, FILE *(*)(struct buf *, struct stat *, int), int, int);
+int argc;
+char **argv;
+int
+main ()
+{
+return f (e, argv, 0) != argv[0] || f (e, argv, 1) != argv[1];
+ ;
+ return 0;
+}
+_ACEOF
+# Don't try gcc -ansi; that turns off useful extensions and
+# breaks some systems' header files.
+# AIX -qlanglvl=ansi
+# Ultrix and OSF/1 -std1
+# HP-UX 10.20 and later -Ae
+# HP-UX older versions -Aa -D_HPUX_SOURCE
+# SVR4 -Xc -D__EXTENSIONS__
+for ac_arg in "" -qlanglvl=ansi -std1 -Ae "-Aa -D_HPUX_SOURCE" "-Xc -D__EXTENSIONS__"
+do
+ CC="$ac_save_CC $ac_arg"
+ rm -f conftest.$ac_objext
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest.$ac_objext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_prog_cc_stdc=$ac_arg
+break
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+fi
+rm -f conftest.err conftest.$ac_objext
+done
+rm -f conftest.$ac_ext conftest.$ac_objext
+CC=$ac_save_CC
+
+fi
+
+case "x$ac_cv_prog_cc_stdc" in
+ x|xno)
+ echo "$as_me:$LINENO: result: none needed" >&5
+echo "${ECHO_T}none needed" >&6 ;;
+ *)
+ echo "$as_me:$LINENO: result: $ac_cv_prog_cc_stdc" >&5
+echo "${ECHO_T}$ac_cv_prog_cc_stdc" >&6
+ CC="$CC $ac_cv_prog_cc_stdc" ;;
+esac
+
+# Some people use a C++ compiler to compile C. Since we use `exit',
+# in C++ we need to declare it. In case someone uses the same compiler
+# for both compiling C and C++ we need to have the C++ compiler decide
+# the declaration of exit, since it's the most demanding environment.
+cat >conftest.$ac_ext <<_ACEOF
+#ifndef __cplusplus
+ choke me
+#endif
+_ACEOF
+rm -f conftest.$ac_objext
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest.$ac_objext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ for ac_declaration in \
+ '' \
+ 'extern "C" void std::exit (int) throw (); using std::exit;' \
+ 'extern "C" void std::exit (int); using std::exit;' \
+ 'extern "C" void exit (int) throw ();' \
+ 'extern "C" void exit (int);' \
+ 'void exit (int);'
+do
+ cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+$ac_declaration
+#include <stdlib.h>
+int
+main ()
+{
+exit (42);
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest.$ac_objext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ :
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+continue
+fi
+rm -f conftest.err conftest.$ac_objext conftest.$ac_ext
+ cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+$ac_declaration
+int
+main ()
+{
+exit (42);
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext
+if { (eval echo "$as_me:$LINENO: \"$ac_compile\"") >&5
+ (eval $ac_compile) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest.$ac_objext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ break
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+fi
+rm -f conftest.err conftest.$ac_objext conftest.$ac_ext
+done
+rm -f conftest*
+if test -n "$ac_declaration"; then
+ echo '#ifdef __cplusplus' >>confdefs.h
+ echo $ac_declaration >>confdefs.h
+ echo '#endif' >>confdefs.h
+fi
+
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+fi
+rm -f conftest.err conftest.$ac_objext conftest.$ac_ext
+ac_ext=c
+ac_cpp='$CPP $CPPFLAGS'
+ac_compile='$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5'
+ac_link='$CC -o conftest$ac_exeext $CFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ext $LIBS >&5'
+ac_compiler_gnu=$ac_cv_c_compiler_gnu
+
+if test "$GCC" != "yes"
+then
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: GCC is required to build this package.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+fi
+if uname | egrep -i 'Linux' > /dev/null 2>&1 &&
+ uname -m | egrep '(x|i)(3|4|5|6|7|8|9)?86' > /dev/null 2>&1
+then
+ MYOPTS="-minline-all-stringops"
+fi
+if uname | egrep -i 'SunOS' > /dev/null 2>&1
+then
+ MYOPTS="-O1 -fno-omit-frame-pointer -fno-force-addr"
+fi
+if uname | egrep -i 'BSD' > /dev/null 2>&1
+then
+ MYOPTS="-O1 -fno-omit-frame-pointer -fno-force-addr"
+fi
+if gcc --version | egrep -i '^2\.(8|9)' > /dev/null 2>&1
+then
+ MYOPTS="-O1 -fno-omit-frame-pointer -fno-force-addr"
+fi
+
+# Linker
+printf 'checking for ld... '
+if which ld | grep '/ld$' > /dev/null 2>&1
+then
+ LD=`which ld`
+ printf '%s\n' "$LD"
+else
+ printf 'no\n'
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: ld is not found in PATH.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+fi
+
+# Archiver
+printf 'checking for ar... '
+if which ar | grep '/ar$' > /dev/null 2>&1
+then
+ AR=`which ar`
+ printf '%s\n' "$AR"
+else
+ printf 'no\n'
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: ar is not found in PATH.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+fi
+
+
+
+#================================================================
+# Checking Libraries
+#================================================================
+
+
+# Underlying libraries
+
+
+echo "$as_me:$LINENO: checking for main in -lc" >&5
+echo $ECHO_N "checking for main in -lc... $ECHO_C" >&6
+if test "${ac_cv_lib_c_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-lc $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_c_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_c_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_c_main" >&5
+echo "${ECHO_T}$ac_cv_lib_c_main" >&6
+if test $ac_cv_lib_c_main = yes; then
+ cat >>confdefs.h <<_ACEOF
+#define HAVE_LIBC 1
+_ACEOF
+
+ LIBS="-lc $LIBS"
+
+fi
+
+
+# for pthread
+if test "$ispthread" = "yes"
+then
+
+echo "$as_me:$LINENO: checking for main in -lpthread" >&5
+echo $ECHO_N "checking for main in -lpthread... $ECHO_C" >&6
+if test "${ac_cv_lib_pthread_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-lpthread $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_pthread_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_pthread_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_pthread_main" >&5
+echo "${ECHO_T}$ac_cv_lib_pthread_main" >&6
+if test $ac_cv_lib_pthread_main = yes; then
+ cat >>confdefs.h <<_ACEOF
+#define HAVE_LIBPTHREAD 1
+_ACEOF
+
+ LIBS="-lpthread $LIBS"
+
+fi
+
+fi
+
+# for zlib
+if test "$iszlib" = "yes"
+then
+
+echo "$as_me:$LINENO: checking for main in -lz" >&5
+echo $ECHO_N "checking for main in -lz... $ECHO_C" >&6
+if test "${ac_cv_lib_z_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-lz $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_z_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_z_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_z_main" >&5
+echo "${ECHO_T}$ac_cv_lib_z_main" >&6
+if test $ac_cv_lib_z_main = yes; then
+ cat >>confdefs.h <<_ACEOF
+#define HAVE_LIBZ 1
+_ACEOF
+
+ LIBS="-lz $LIBS"
+
+fi
+
+fi
+
+# for lzo
+if test "$islzo" = "yes"
+then
+
+echo "$as_me:$LINENO: checking for main in -llzo2" >&5
+echo $ECHO_N "checking for main in -llzo2... $ECHO_C" >&6
+if test "${ac_cv_lib_lzo2_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-llzo2 $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_lzo2_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_lzo2_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_lzo2_main" >&5
+echo "${ECHO_T}$ac_cv_lib_lzo2_main" >&6
+if test $ac_cv_lib_lzo2_main = yes; then
+ cat >>confdefs.h <<_ACEOF
+#define HAVE_LIBLZO2 1
+_ACEOF
+
+ LIBS="-llzo2 $LIBS"
+
+fi
+
+fi
+
+# for bzip
+if test "$isbzip" = "yes"
+then
+
+echo "$as_me:$LINENO: checking for main in -lbz2" >&5
+echo $ECHO_N "checking for main in -lbz2... $ECHO_C" >&6
+if test "${ac_cv_lib_bz2_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-lbz2 $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_bz2_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_bz2_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_bz2_main" >&5
+echo "${ECHO_T}$ac_cv_lib_bz2_main" >&6
+if test $ac_cv_lib_bz2_main = yes; then
+ cat >>confdefs.h <<_ACEOF
+#define HAVE_LIBBZ2 1
+_ACEOF
+
+ LIBS="-lbz2 $LIBS"
+
+fi
+
+fi
+
+# for iconv
+if test "$isiconv" = "yes"
+then
+
+echo "$as_me:$LINENO: checking for main in -liconv" >&5
+echo $ECHO_N "checking for main in -liconv... $ECHO_C" >&6
+if test "${ac_cv_lib_iconv_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-liconv $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_iconv_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_iconv_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_iconv_main" >&5
+echo "${ECHO_T}$ac_cv_lib_iconv_main" >&6
+if test $ac_cv_lib_iconv_main = yes; then
+ cat >>confdefs.h <<_ACEOF
+#define HAVE_LIBICONV 1
+_ACEOF
+
+ LIBS="-liconv $LIBS"
+
+fi
+
+fi
+
+# For old BSDs
+if test "$ispthread" = "yes" && uname -a | grep BSD > /dev/null &&
+ test -f /usr/lib/libc_r.a && test ! -f /usr/lib/libpthread.a
+then
+ LIBS=`printf '%s' "$LIBS" | sed 's/-lc/-lc_r/g'`
+fi
+
+# Duplication of QDBM for C
+echo "$as_me:$LINENO: checking for main in -lqdbm" >&5
+echo $ECHO_N "checking for main in -lqdbm... $ECHO_C" >&6
+if test "${ac_cv_lib_qdbm_main+set}" = set; then
+ echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+ ac_check_lib_save_LIBS=$LIBS
+LIBS="-lqdbm $LIBS"
+cat >conftest.$ac_ext <<_ACEOF
+/* confdefs.h. */
+_ACEOF
+cat confdefs.h >>conftest.$ac_ext
+cat >>conftest.$ac_ext <<_ACEOF
+/* end confdefs.h. */
+
+
+int
+main ()
+{
+main ();
+ ;
+ return 0;
+}
+_ACEOF
+rm -f conftest.$ac_objext conftest$ac_exeext
+if { (eval echo "$as_me:$LINENO: \"$ac_link\"") >&5
+ (eval $ac_link) 2>conftest.er1
+ ac_status=$?
+ grep -v '^ *+' conftest.er1 >conftest.err
+ rm -f conftest.er1
+ cat conftest.err >&5
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); } &&
+ { ac_try='test -z "$ac_c_werror_flag"
+ || test ! -s conftest.err'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; } &&
+ { ac_try='test -s conftest$ac_exeext'
+ { (eval echo "$as_me:$LINENO: \"$ac_try\"") >&5
+ (eval $ac_try) 2>&5
+ ac_status=$?
+ echo "$as_me:$LINENO: \$? = $ac_status" >&5
+ (exit $ac_status); }; }; then
+ ac_cv_lib_qdbm_main=yes
+else
+ echo "$as_me: failed program was:" >&5
+sed 's/^/| /' conftest.$ac_ext >&5
+
+ac_cv_lib_qdbm_main=no
+fi
+rm -f conftest.err conftest.$ac_objext \
+ conftest$ac_exeext conftest.$ac_ext
+LIBS=$ac_check_lib_save_LIBS
+fi
+echo "$as_me:$LINENO: result: $ac_cv_lib_qdbm_main" >&5
+echo "${ECHO_T}$ac_cv_lib_qdbm_main" >&6
+if test $ac_cv_lib_qdbm_main = yes; then
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: The existing library was detected.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+
+fi
+
+
+
+
+#================================================================
+# Generic Settings
+#================================================================
+
+
+# Export variables
+
+
+
+
+
+
+
+
+
+
+
+# Targets
+ ac_config_files="$ac_config_files Makefile LTmakefile qdbm.spec qdbm.pc"
+cat >confcache <<\_ACEOF
+# This file is a shell script that caches the results of configure
+# tests run on this system so they can be shared between configure
+# scripts and configure runs, see configure's option --config-cache.
+# It is not useful on other systems. If it contains results you don't
+# want to keep, you may remove or edit it.
+#
+# config.status only pays attention to the cache file if you give it
+# the --recheck option to rerun configure.
+#
+# `ac_cv_env_foo' variables (set or unset) will be overridden when
+# loading this file, other *unset* `ac_cv_foo' will be assigned the
+# following values.
+
+_ACEOF
+
+# The following way of writing the cache mishandles newlines in values,
+# but we know of no workaround that is simple, portable, and efficient.
+# So, don't put newlines in cache variables' values.
+# Ultrix sh set writes to stderr and can't be redirected directly,
+# and sets the high bit in the cache file unless we assign to the vars.
+{
+ (set) 2>&1 |
+ case `(ac_space=' '; set | grep ac_space) 2>&1` in
+ *ac_space=\ *)
+ # `set' does not quote correctly, so add quotes (double-quote
+ # substitution turns \\\\ into \\, and sed turns \\ into \).
+ sed -n \
+ "s/'/'\\\\''/g;
+ s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1='\\2'/p"
+ ;;
+ *)
+ # `set' quotes correctly as required by POSIX, so do not add quotes.
+ sed -n \
+ "s/^\\([_$as_cr_alnum]*_cv_[_$as_cr_alnum]*\\)=\\(.*\\)/\\1=\\2/p"
+ ;;
+ esac;
+} |
+ sed '
+ t clear
+ : clear
+ s/^\([^=]*\)=\(.*[{}].*\)$/test "${\1+set}" = set || &/
+ t end
+ /^ac_cv_env/!s/^\([^=]*\)=\(.*\)$/\1=${\1=\2}/
+ : end' >>confcache
+if diff $cache_file confcache >/dev/null 2>&1; then :; else
+ if test -w $cache_file; then
+ test "x$cache_file" != "x/dev/null" && echo "updating cache $cache_file"
+ cat confcache >$cache_file
+ else
+ echo "not updating unwritable cache $cache_file"
+ fi
+fi
+rm -f confcache
+
+test "x$prefix" = xNONE && prefix=$ac_default_prefix
+# Let make expand exec_prefix.
+test "x$exec_prefix" = xNONE && exec_prefix='${prefix}'
+
+# VPATH may cause trouble with some makes, so we remove $(srcdir),
+# ${srcdir} and @srcdir@ from VPATH if srcdir is ".", strip leading and
+# trailing colons and then remove the whole line if VPATH becomes empty
+# (actually we leave an empty line to preserve line numbers).
+if test "x$srcdir" = x.; then
+ ac_vpsub='/^[ ]*VPATH[ ]*=/{
+s/:*\$(srcdir):*/:/;
+s/:*\${srcdir}:*/:/;
+s/:*@srcdir@:*/:/;
+s/^\([^=]*=[ ]*\):*/\1/;
+s/:*$//;
+s/^[^=]*=[ ]*$//;
+}'
+fi
+
+# Transform confdefs.h into DEFS.
+# Protect against shell expansion while executing Makefile rules.
+# Protect against Makefile macro expansion.
+#
+# If the first sed substitution is executed (which looks for macros that
+# take arguments), then we branch to the quote section. Otherwise,
+# look for a macro that doesn't take arguments.
+cat >confdef2opt.sed <<\_ACEOF
+t clear
+: clear
+s,^[ ]*#[ ]*define[ ][ ]*\([^ (][^ (]*([^)]*)\)[ ]*\(.*\),-D\1=\2,g
+t quote
+s,^[ ]*#[ ]*define[ ][ ]*\([^ ][^ ]*\)[ ]*\(.*\),-D\1=\2,g
+t quote
+d
+: quote
+s,[ `~#$^&*(){}\\|;'"<>?],\\&,g
+s,\[,\\&,g
+s,\],\\&,g
+s,\$,$$,g
+p
+_ACEOF
+# We use echo to avoid assuming a particular line-breaking character.
+# The extra dot is to prevent the shell from consuming trailing
+# line-breaks from the sub-command output. A line-break within
+# single-quotes doesn't work because, if this script is created in a
+# platform that uses two characters for line-breaks (e.g., DOS), tr
+# would break.
+ac_LF_and_DOT=`echo; echo .`
+DEFS=`sed -n -f confdef2opt.sed confdefs.h | tr "$ac_LF_and_DOT" ' .'`
+rm -f confdef2opt.sed
+
+
+ac_libobjs=
+ac_ltlibobjs=
+for ac_i in : $LIBOBJS; do test "x$ac_i" = x: && continue
+ # 1. Remove the extension, and $U if already installed.
+ ac_i=`echo "$ac_i" |
+ sed 's/\$U\././;s/\.o$//;s/\.obj$//'`
+ # 2. Add them.
+ ac_libobjs="$ac_libobjs $ac_i\$U.$ac_objext"
+ ac_ltlibobjs="$ac_ltlibobjs $ac_i"'$U.lo'
+done
+LIBOBJS=$ac_libobjs
+
+LTLIBOBJS=$ac_ltlibobjs
+
+
+
+: ${CONFIG_STATUS=./config.status}
+ac_clean_files_save=$ac_clean_files
+ac_clean_files="$ac_clean_files $CONFIG_STATUS"
+{ echo "$as_me:$LINENO: creating $CONFIG_STATUS" >&5
+echo "$as_me: creating $CONFIG_STATUS" >&6;}
+cat >$CONFIG_STATUS <<_ACEOF
+#! $SHELL
+# Generated by $as_me.
+# Run this file to recreate the current configuration.
+# Compiler output produced by configure, useful for debugging
+# configure, is in config.log if it exists.
+
+debug=false
+ac_cs_recheck=false
+ac_cs_silent=false
+SHELL=\${CONFIG_SHELL-$SHELL}
+_ACEOF
+
+cat >>$CONFIG_STATUS <<\_ACEOF
+## --------------------- ##
+## M4sh Initialization. ##
+## --------------------- ##
+
+# Be Bourne compatible
+if test -n "${ZSH_VERSION+set}" && (emulate sh) >/dev/null 2>&1; then
+ emulate sh
+ NULLCMD=:
+ # Zsh 3.x and 4.x performs word splitting on ${1+"$@"}, which
+ # is contrary to our usage. Disable this feature.
+ alias -g '${1+"$@"}'='"$@"'
+elif test -n "${BASH_VERSION+set}" && (set -o posix) >/dev/null 2>&1; then
+ set -o posix
+fi
+DUALCASE=1; export DUALCASE # for MKS sh
+
+# Support unset when possible.
+if ( (MAIL=60; unset MAIL) || exit) >/dev/null 2>&1; then
+ as_unset=unset
+else
+ as_unset=false
+fi
+
+
+# Work around bugs in pre-3.0 UWIN ksh.
+$as_unset ENV MAIL MAILPATH
+PS1='$ '
+PS2='> '
+PS4='+ '
+
+# NLS nuisances.
+for as_var in \
+ LANG LANGUAGE LC_ADDRESS LC_ALL LC_COLLATE LC_CTYPE LC_IDENTIFICATION \
+ LC_MEASUREMENT LC_MESSAGES LC_MONETARY LC_NAME LC_NUMERIC LC_PAPER \
+ LC_TELEPHONE LC_TIME
+do
+ if (set +x; test -z "`(eval $as_var=C; export $as_var) 2>&1`"); then
+ eval $as_var=C; export $as_var
+ else
+ $as_unset $as_var
+ fi
+done
+
+# Required to use basename.
+if expr a : '\(a\)' >/dev/null 2>&1; then
+ as_expr=expr
+else
+ as_expr=false
+fi
+
+if (basename /) >/dev/null 2>&1 && test "X`basename / 2>&1`" = "X/"; then
+ as_basename=basename
+else
+ as_basename=false
+fi
+
+
+# Name of the executable.
+as_me=`$as_basename "$0" ||
+$as_expr X/"$0" : '.*/\([^/][^/]*\)/*$' \| \
+ X"$0" : 'X\(//\)$' \| \
+ X"$0" : 'X\(/\)$' \| \
+ . : '\(.\)' 2>/dev/null ||
+echo X/"$0" |
+ sed '/^.*\/\([^/][^/]*\)\/*$/{ s//\1/; q; }
+ /^X\/\(\/\/\)$/{ s//\1/; q; }
+ /^X\/\(\/\).*/{ s//\1/; q; }
+ s/.*/./; q'`
+
+
+# PATH needs CR, and LINENO needs CR and PATH.
+# Avoid depending upon Character Ranges.
+as_cr_letters='abcdefghijklmnopqrstuvwxyz'
+as_cr_LETTERS='ABCDEFGHIJKLMNOPQRSTUVWXYZ'
+as_cr_Letters=$as_cr_letters$as_cr_LETTERS
+as_cr_digits='0123456789'
+as_cr_alnum=$as_cr_Letters$as_cr_digits
+
+# The user is always right.
+if test "${PATH_SEPARATOR+set}" != set; then
+ echo "#! /bin/sh" >conf$$.sh
+ echo "exit 0" >>conf$$.sh
+ chmod +x conf$$.sh
+ if (PATH="/nonexistent;."; conf$$.sh) >/dev/null 2>&1; then
+ PATH_SEPARATOR=';'
+ else
+ PATH_SEPARATOR=:
+ fi
+ rm -f conf$$.sh
+fi
+
+
+ as_lineno_1=$LINENO
+ as_lineno_2=$LINENO
+ as_lineno_3=`(expr $as_lineno_1 + 1) 2>/dev/null`
+ test "x$as_lineno_1" != "x$as_lineno_2" &&
+ test "x$as_lineno_3" = "x$as_lineno_2" || {
+ # Find who we are. Look in the path if we contain no path at all
+ # relative or not.
+ case $0 in
+ *[\\/]* ) as_myself=$0 ;;
+ *) as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ test -r "$as_dir/$0" && as_myself=$as_dir/$0 && break
+done
+
+ ;;
+ esac
+ # We did not find ourselves, most probably we were run as `sh COMMAND'
+ # in which case we are not to be found in the path.
+ if test "x$as_myself" = x; then
+ as_myself=$0
+ fi
+ if test ! -f "$as_myself"; then
+ { { echo "$as_me:$LINENO: error: cannot find myself; rerun with an absolute path" >&5
+echo "$as_me: error: cannot find myself; rerun with an absolute path" >&2;}
+ { (exit 1); exit 1; }; }
+ fi
+ case $CONFIG_SHELL in
+ '')
+ as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in /bin$PATH_SEPARATOR/usr/bin$PATH_SEPARATOR$PATH
+do
+ IFS=$as_save_IFS
+ test -z "$as_dir" && as_dir=.
+ for as_base in sh bash ksh sh5; do
+ case $as_dir in
+ /*)
+ if ("$as_dir/$as_base" -c '
+ as_lineno_1=$LINENO
+ as_lineno_2=$LINENO
+ as_lineno_3=`(expr $as_lineno_1 + 1) 2>/dev/null`
+ test "x$as_lineno_1" != "x$as_lineno_2" &&
+ test "x$as_lineno_3" = "x$as_lineno_2" ') 2>/dev/null; then
+ $as_unset BASH_ENV || test "${BASH_ENV+set}" != set || { BASH_ENV=; export BASH_ENV; }
+ $as_unset ENV || test "${ENV+set}" != set || { ENV=; export ENV; }
+ CONFIG_SHELL=$as_dir/$as_base
+ export CONFIG_SHELL
+ exec "$CONFIG_SHELL" "$0" ${1+"$@"}
+ fi;;
+ esac
+ done
+done
+;;
+ esac
+
+ # Create $as_me.lineno as a copy of $as_myself, but with $LINENO
+ # uniformly replaced by the line number. The first 'sed' inserts a
+ # line-number line before each line; the second 'sed' does the real
+ # work. The second script uses 'N' to pair each line-number line
+ # with the numbered line, and appends trailing '-' during
+ # substitution so that $LINENO is not a special case at line end.
+ # (Raja R Harinath suggested sed '=', and Paul Eggert wrote the
+ # second 'sed' script. Blame Lee E. McMahon for sed's syntax. :-)
+ sed '=' <$as_myself |
+ sed '
+ N
+ s,$,-,
+ : loop
+ s,^\(['$as_cr_digits']*\)\(.*\)[$]LINENO\([^'$as_cr_alnum'_]\),\1\2\1\3,
+ t loop
+ s,-$,,
+ s,^['$as_cr_digits']*\n,,
+ ' >$as_me.lineno &&
+ chmod +x $as_me.lineno ||
+ { { echo "$as_me:$LINENO: error: cannot create $as_me.lineno; rerun with a POSIX shell" >&5
+echo "$as_me: error: cannot create $as_me.lineno; rerun with a POSIX shell" >&2;}
+ { (exit 1); exit 1; }; }
+
+ # Don't try to exec as it changes $[0], causing all sort of problems
+ # (the dirname of $[0] is not the place where we might find the
+ # original and so on. Autoconf is especially sensible to this).
+ . ./$as_me.lineno
+ # Exit status is that of the last command.
+ exit
+}
+
+
+case `echo "testing\c"; echo 1,2,3`,`echo -n testing; echo 1,2,3` in
+ *c*,-n*) ECHO_N= ECHO_C='
+' ECHO_T=' ' ;;
+ *c*,* ) ECHO_N=-n ECHO_C= ECHO_T= ;;
+ *) ECHO_N= ECHO_C='\c' ECHO_T= ;;
+esac
+
+if expr a : '\(a\)' >/dev/null 2>&1; then
+ as_expr=expr
+else
+ as_expr=false
+fi
+
+rm -f conf$$ conf$$.exe conf$$.file
+echo >conf$$.file
+if ln -s conf$$.file conf$$ 2>/dev/null; then
+ # We could just check for DJGPP; but this test a) works b) is more generic
+ # and c) will remain valid once DJGPP supports symlinks (DJGPP 2.04).
+ if test -f conf$$.exe; then
+ # Don't use ln at all; we don't have any links
+ as_ln_s='cp -p'
+ else
+ as_ln_s='ln -s'
+ fi
+elif ln conf$$.file conf$$ 2>/dev/null; then
+ as_ln_s=ln
+else
+ as_ln_s='cp -p'
+fi
+rm -f conf$$ conf$$.exe conf$$.file
+
+if mkdir -p . 2>/dev/null; then
+ as_mkdir_p=:
+else
+ test -d ./-p && rmdir ./-p
+ as_mkdir_p=false
+fi
+
+as_executable_p="test -f"
+
+# Sed expression to map a string onto a valid CPP name.
+as_tr_cpp="eval sed 'y%*$as_cr_letters%P$as_cr_LETTERS%;s%[^_$as_cr_alnum]%_%g'"
+
+# Sed expression to map a string onto a valid variable name.
+as_tr_sh="eval sed 'y%*+%pp%;s%[^_$as_cr_alnum]%_%g'"
+
+
+# IFS
+# We need space, tab and new line, in precisely that order.
+as_nl='
+'
+IFS=" $as_nl"
+
+# CDPATH.
+$as_unset CDPATH
+
+exec 6>&1
+
+# Open the log real soon, to keep \$[0] and so on meaningful, and to
+# report actual input values of CONFIG_FILES etc. instead of their
+# values after options handling. Logging --version etc. is OK.
+exec 5>>config.log
+{
+ echo
+ sed 'h;s/./-/g;s/^.../## /;s/...$/ ##/;p;x;p;x' <<_ASBOX
+## Running $as_me. ##
+_ASBOX
+} >&5
+cat >&5 <<_CSEOF
+
+This file was extended by qdbm $as_me 1.8.77, which was
+generated by GNU Autoconf 2.59. Invocation command line was
+
+ CONFIG_FILES = $CONFIG_FILES
+ CONFIG_HEADERS = $CONFIG_HEADERS
+ CONFIG_LINKS = $CONFIG_LINKS
+ CONFIG_COMMANDS = $CONFIG_COMMANDS
+ $ $0 $@
+
+_CSEOF
+echo "on `(hostname || uname -n) 2>/dev/null | sed 1q`" >&5
+echo >&5
+_ACEOF
+
+# Files that config.status was made for.
+if test -n "$ac_config_files"; then
+ echo "config_files=\"$ac_config_files\"" >>$CONFIG_STATUS
+fi
+
+if test -n "$ac_config_headers"; then
+ echo "config_headers=\"$ac_config_headers\"" >>$CONFIG_STATUS
+fi
+
+if test -n "$ac_config_links"; then
+ echo "config_links=\"$ac_config_links\"" >>$CONFIG_STATUS
+fi
+
+if test -n "$ac_config_commands"; then
+ echo "config_commands=\"$ac_config_commands\"" >>$CONFIG_STATUS
+fi
+
+cat >>$CONFIG_STATUS <<\_ACEOF
+
+ac_cs_usage="\
+\`$as_me' instantiates files from templates according to the
+current configuration.
+
+Usage: $0 [OPTIONS] [FILE]...
+
+ -h, --help print this help, then exit
+ -V, --version print version number, then exit
+ -q, --quiet do not print progress messages
+ -d, --debug don't remove temporary files
+ --recheck update $as_me by reconfiguring in the same conditions
+ --file=FILE[:TEMPLATE]
+ instantiate the configuration file FILE
+
+Configuration files:
+$config_files
+
+Report bugs to <bug-autoconf at gnu.org>."
+_ACEOF
+
+cat >>$CONFIG_STATUS <<_ACEOF
+ac_cs_version="\\
+qdbm config.status 1.8.77
+configured by $0, generated by GNU Autoconf 2.59,
+ with options \\"`echo "$ac_configure_args" | sed 's/[\\""\`\$]/\\\\&/g'`\\"
+
+Copyright (C) 2003 Free Software Foundation, Inc.
+This config.status script is free software; the Free Software Foundation
+gives unlimited permission to copy, distribute and modify it."
+srcdir=$srcdir
+_ACEOF
+
+cat >>$CONFIG_STATUS <<\_ACEOF
+# If no file are specified by the user, then we need to provide default
+# value. By we need to know if files were specified by the user.
+ac_need_defaults=:
+while test $# != 0
+do
+ case $1 in
+ --*=*)
+ ac_option=`expr "x$1" : 'x\([^=]*\)='`
+ ac_optarg=`expr "x$1" : 'x[^=]*=\(.*\)'`
+ ac_shift=:
+ ;;
+ -*)
+ ac_option=$1
+ ac_optarg=$2
+ ac_shift=shift
+ ;;
+ *) # This is not an option, so the user has probably given explicit
+ # arguments.
+ ac_option=$1
+ ac_need_defaults=false;;
+ esac
+
+ case $ac_option in
+ # Handling of the options.
+_ACEOF
+cat >>$CONFIG_STATUS <<\_ACEOF
+ -recheck | --recheck | --rechec | --reche | --rech | --rec | --re | --r)
+ ac_cs_recheck=: ;;
+ --version | --vers* | -V )
+ echo "$ac_cs_version"; exit 0 ;;
+ --he | --h)
+ # Conflict between --help and --header
+ { { echo "$as_me:$LINENO: error: ambiguous option: $1
+Try \`$0 --help' for more information." >&5
+echo "$as_me: error: ambiguous option: $1
+Try \`$0 --help' for more information." >&2;}
+ { (exit 1); exit 1; }; };;
+ --help | --hel | -h )
+ echo "$ac_cs_usage"; exit 0 ;;
+ --debug | --d* | -d )
+ debug=: ;;
+ --file | --fil | --fi | --f )
+ $ac_shift
+ CONFIG_FILES="$CONFIG_FILES $ac_optarg"
+ ac_need_defaults=false;;
+ --header | --heade | --head | --hea )
+ $ac_shift
+ CONFIG_HEADERS="$CONFIG_HEADERS $ac_optarg"
+ ac_need_defaults=false;;
+ -q | -quiet | --quiet | --quie | --qui | --qu | --q \
+ | -silent | --silent | --silen | --sile | --sil | --si | --s)
+ ac_cs_silent=: ;;
+
+ # This is an error.
+ -*) { { echo "$as_me:$LINENO: error: unrecognized option: $1
+Try \`$0 --help' for more information." >&5
+echo "$as_me: error: unrecognized option: $1
+Try \`$0 --help' for more information." >&2;}
+ { (exit 1); exit 1; }; } ;;
+
+ *) ac_config_targets="$ac_config_targets $1" ;;
+
+ esac
+ shift
+done
+
+ac_configure_extra_args=
+
+if $ac_cs_silent; then
+ exec 6>/dev/null
+ ac_configure_extra_args="$ac_configure_extra_args --silent"
+fi
+
+_ACEOF
+cat >>$CONFIG_STATUS <<_ACEOF
+if \$ac_cs_recheck; then
+ echo "running $SHELL $0 " $ac_configure_args \$ac_configure_extra_args " --no-create --no-recursion" >&6
+ exec $SHELL $0 $ac_configure_args \$ac_configure_extra_args --no-create --no-recursion
+fi
+
+_ACEOF
+
+
+
+
+
+cat >>$CONFIG_STATUS <<\_ACEOF
+for ac_config_target in $ac_config_targets
+do
+ case "$ac_config_target" in
+ # Handling of arguments.
+ "Makefile" ) CONFIG_FILES="$CONFIG_FILES Makefile" ;;
+ "LTmakefile" ) CONFIG_FILES="$CONFIG_FILES LTmakefile" ;;
+ "qdbm.spec" ) CONFIG_FILES="$CONFIG_FILES qdbm.spec" ;;
+ "qdbm.pc" ) CONFIG_FILES="$CONFIG_FILES qdbm.pc" ;;
+ *) { { echo "$as_me:$LINENO: error: invalid argument: $ac_config_target" >&5
+echo "$as_me: error: invalid argument: $ac_config_target" >&2;}
+ { (exit 1); exit 1; }; };;
+ esac
+done
+
+# If the user did not use the arguments to specify the items to instantiate,
+# then the envvar interface is used. Set only those that are not.
+# We use the long form for the default assignment because of an extremely
+# bizarre bug on SunOS 4.1.3.
+if $ac_need_defaults; then
+ test "${CONFIG_FILES+set}" = set || CONFIG_FILES=$config_files
+fi
+
+# Have a temporary directory for convenience. Make it in the build tree
+# simply because there is no reason to put it here, and in addition,
+# creating and moving files from /tmp can sometimes cause problems.
+# Create a temporary directory, and hook for its removal unless debugging.
+$debug ||
+{
+ trap 'exit_status=$?; rm -rf $tmp && exit $exit_status' 0
+ trap '{ (exit 1); exit 1; }' 1 2 13 15
+}
+
+# Create a (secure) tmp directory for tmp files.
+
+{
+ tmp=`(umask 077 && mktemp -d -q "./confstatXXXXXX") 2>/dev/null` &&
+ test -n "$tmp" && test -d "$tmp"
+} ||
+{
+ tmp=./confstat$$-$RANDOM
+ (umask 077 && mkdir $tmp)
+} ||
+{
+ echo "$me: cannot create a temporary directory in ." >&2
+ { (exit 1); exit 1; }
+}
+
+_ACEOF
+
+cat >>$CONFIG_STATUS <<_ACEOF
+
+#
+# CONFIG_FILES section.
+#
+
+# No need to generate the scripts if there are no CONFIG_FILES.
+# This happens for instance when ./config.status config.h
+if test -n "\$CONFIG_FILES"; then
+ # Protect against being on the right side of a sed subst in config.status.
+ sed 's/,@/@@/; s/@,/@@/; s/,;t t\$/@;t t/; /@;t t\$/s/[\\\\&,]/\\\\&/g;
+ s/@@/,@/; s/@@/@,/; s/@;t t\$/,;t t/' >\$tmp/subs.sed <<\\CEOF
+s, at SHELL@,$SHELL,;t t
+s, at PATH_SEPARATOR@,$PATH_SEPARATOR,;t t
+s, at PACKAGE_NAME@,$PACKAGE_NAME,;t t
+s, at PACKAGE_TARNAME@,$PACKAGE_TARNAME,;t t
+s, at PACKAGE_VERSION@,$PACKAGE_VERSION,;t t
+s, at PACKAGE_STRING@,$PACKAGE_STRING,;t t
+s, at PACKAGE_BUGREPORT@,$PACKAGE_BUGREPORT,;t t
+s, at exec_prefix@,$exec_prefix,;t t
+s, at prefix@,$prefix,;t t
+s, at program_transform_name@,$program_transform_name,;t t
+s, at bindir@,$bindir,;t t
+s, at sbindir@,$sbindir,;t t
+s, at libexecdir@,$libexecdir,;t t
+s, at datadir@,$datadir,;t t
+s, at sysconfdir@,$sysconfdir,;t t
+s, at sharedstatedir@,$sharedstatedir,;t t
+s, at localstatedir@,$localstatedir,;t t
+s, at libdir@,$libdir,;t t
+s, at includedir@,$includedir,;t t
+s, at oldincludedir@,$oldincludedir,;t t
+s, at infodir@,$infodir,;t t
+s, at mandir@,$mandir,;t t
+s, at build_alias@,$build_alias,;t t
+s, at host_alias@,$host_alias,;t t
+s, at target_alias@,$target_alias,;t t
+s, at DEFS@,$DEFS,;t t
+s, at ECHO_C@,$ECHO_C,;t t
+s, at ECHO_N@,$ECHO_N,;t t
+s, at ECHO_T@,$ECHO_T,;t t
+s, at LIBS@,$LIBS,;t t
+s, at CC@,$CC,;t t
+s, at CFLAGS@,$CFLAGS,;t t
+s, at LDFLAGS@,$LDFLAGS,;t t
+s, at CPPFLAGS@,$CPPFLAGS,;t t
+s, at ac_ct_CC@,$ac_ct_CC,;t t
+s, at EXEEXT@,$EXEEXT,;t t
+s, at OBJEXT@,$OBJEXT,;t t
+s, at LIBVER@,$LIBVER,;t t
+s, at LIBREV@,$LIBREV,;t t
+s, at TARGETS@,$TARGETS,;t t
+s, at MYDEFS@,$MYDEFS,;t t
+s, at MYOPTS@,$MYOPTS,;t t
+s, at MGWLIBS@,$MGWLIBS,;t t
+s, at LD@,$LD,;t t
+s, at AR@,$AR,;t t
+s, at LIBOBJS@,$LIBOBJS,;t t
+s, at LTLIBOBJS@,$LTLIBOBJS,;t t
+CEOF
+
+_ACEOF
+
+ cat >>$CONFIG_STATUS <<\_ACEOF
+ # Split the substitutions into bite-sized pieces for seds with
+ # small command number limits, like on Digital OSF/1 and HP-UX.
+ ac_max_sed_lines=48
+ ac_sed_frag=1 # Number of current file.
+ ac_beg=1 # First line for current file.
+ ac_end=$ac_max_sed_lines # Line after last line for current file.
+ ac_more_lines=:
+ ac_sed_cmds=
+ while $ac_more_lines; do
+ if test $ac_beg -gt 1; then
+ sed "1,${ac_beg}d; ${ac_end}q" $tmp/subs.sed >$tmp/subs.frag
+ else
+ sed "${ac_end}q" $tmp/subs.sed >$tmp/subs.frag
+ fi
+ if test ! -s $tmp/subs.frag; then
+ ac_more_lines=false
+ else
+ # The purpose of the label and of the branching condition is to
+ # speed up the sed processing (if there are no `@' at all, there
+ # is no need to browse any of the substitutions).
+ # These are the two extra sed commands mentioned above.
+ (echo ':t
+ /@[a-zA-Z_][a-zA-Z_0-9]*@/!b' && cat $tmp/subs.frag) >$tmp/subs-$ac_sed_frag.sed
+ if test -z "$ac_sed_cmds"; then
+ ac_sed_cmds="sed -f $tmp/subs-$ac_sed_frag.sed"
+ else
+ ac_sed_cmds="$ac_sed_cmds | sed -f $tmp/subs-$ac_sed_frag.sed"
+ fi
+ ac_sed_frag=`expr $ac_sed_frag + 1`
+ ac_beg=$ac_end
+ ac_end=`expr $ac_end + $ac_max_sed_lines`
+ fi
+ done
+ if test -z "$ac_sed_cmds"; then
+ ac_sed_cmds=cat
+ fi
+fi # test -n "$CONFIG_FILES"
+
+_ACEOF
+cat >>$CONFIG_STATUS <<\_ACEOF
+for ac_file in : $CONFIG_FILES; do test "x$ac_file" = x: && continue
+ # Support "outfile[:infile[:infile...]]", defaulting infile="outfile.in".
+ case $ac_file in
+ - | *:- | *:-:* ) # input from stdin
+ cat >$tmp/stdin
+ ac_file_in=`echo "$ac_file" | sed 's,[^:]*:,,'`
+ ac_file=`echo "$ac_file" | sed 's,:.*,,'` ;;
+ *:* ) ac_file_in=`echo "$ac_file" | sed 's,[^:]*:,,'`
+ ac_file=`echo "$ac_file" | sed 's,:.*,,'` ;;
+ * ) ac_file_in=$ac_file.in ;;
+ esac
+
+ # Compute @srcdir@, @top_srcdir@, and @INSTALL@ for subdirectories.
+ ac_dir=`(dirname "$ac_file") 2>/dev/null ||
+$as_expr X"$ac_file" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \
+ X"$ac_file" : 'X\(//\)[^/]' \| \
+ X"$ac_file" : 'X\(//\)$' \| \
+ X"$ac_file" : 'X\(/\)' \| \
+ . : '\(.\)' 2>/dev/null ||
+echo X"$ac_file" |
+ sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/; q; }
+ /^X\(\/\/\)[^/].*/{ s//\1/; q; }
+ /^X\(\/\/\)$/{ s//\1/; q; }
+ /^X\(\/\).*/{ s//\1/; q; }
+ s/.*/./; q'`
+ { if $as_mkdir_p; then
+ mkdir -p "$ac_dir"
+ else
+ as_dir="$ac_dir"
+ as_dirs=
+ while test ! -d "$as_dir"; do
+ as_dirs="$as_dir $as_dirs"
+ as_dir=`(dirname "$as_dir") 2>/dev/null ||
+$as_expr X"$as_dir" : 'X\(.*[^/]\)//*[^/][^/]*/*$' \| \
+ X"$as_dir" : 'X\(//\)[^/]' \| \
+ X"$as_dir" : 'X\(//\)$' \| \
+ X"$as_dir" : 'X\(/\)' \| \
+ . : '\(.\)' 2>/dev/null ||
+echo X"$as_dir" |
+ sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/; q; }
+ /^X\(\/\/\)[^/].*/{ s//\1/; q; }
+ /^X\(\/\/\)$/{ s//\1/; q; }
+ /^X\(\/\).*/{ s//\1/; q; }
+ s/.*/./; q'`
+ done
+ test ! -n "$as_dirs" || mkdir $as_dirs
+ fi || { { echo "$as_me:$LINENO: error: cannot create directory \"$ac_dir\"" >&5
+echo "$as_me: error: cannot create directory \"$ac_dir\"" >&2;}
+ { (exit 1); exit 1; }; }; }
+
+ ac_builddir=.
+
+if test "$ac_dir" != .; then
+ ac_dir_suffix=/`echo "$ac_dir" | sed 's,^\.[\\/],,'`
+ # A "../" for each directory in $ac_dir_suffix.
+ ac_top_builddir=`echo "$ac_dir_suffix" | sed 's,/[^\\/]*,../,g'`
+else
+ ac_dir_suffix= ac_top_builddir=
+fi
+
+case $srcdir in
+ .) # No --srcdir option. We are building in place.
+ ac_srcdir=.
+ if test -z "$ac_top_builddir"; then
+ ac_top_srcdir=.
+ else
+ ac_top_srcdir=`echo $ac_top_builddir | sed 's,/$,,'`
+ fi ;;
+ [\\/]* | ?:[\\/]* ) # Absolute path.
+ ac_srcdir=$srcdir$ac_dir_suffix;
+ ac_top_srcdir=$srcdir ;;
+ *) # Relative path.
+ ac_srcdir=$ac_top_builddir$srcdir$ac_dir_suffix
+ ac_top_srcdir=$ac_top_builddir$srcdir ;;
+esac
+
+# Do not use `cd foo && pwd` to compute absolute paths, because
+# the directories may not exist.
+case `pwd` in
+.) ac_abs_builddir="$ac_dir";;
+*)
+ case "$ac_dir" in
+ .) ac_abs_builddir=`pwd`;;
+ [\\/]* | ?:[\\/]* ) ac_abs_builddir="$ac_dir";;
+ *) ac_abs_builddir=`pwd`/"$ac_dir";;
+ esac;;
+esac
+case $ac_abs_builddir in
+.) ac_abs_top_builddir=${ac_top_builddir}.;;
+*)
+ case ${ac_top_builddir}. in
+ .) ac_abs_top_builddir=$ac_abs_builddir;;
+ [\\/]* | ?:[\\/]* ) ac_abs_top_builddir=${ac_top_builddir}.;;
+ *) ac_abs_top_builddir=$ac_abs_builddir/${ac_top_builddir}.;;
+ esac;;
+esac
+case $ac_abs_builddir in
+.) ac_abs_srcdir=$ac_srcdir;;
+*)
+ case $ac_srcdir in
+ .) ac_abs_srcdir=$ac_abs_builddir;;
+ [\\/]* | ?:[\\/]* ) ac_abs_srcdir=$ac_srcdir;;
+ *) ac_abs_srcdir=$ac_abs_builddir/$ac_srcdir;;
+ esac;;
+esac
+case $ac_abs_builddir in
+.) ac_abs_top_srcdir=$ac_top_srcdir;;
+*)
+ case $ac_top_srcdir in
+ .) ac_abs_top_srcdir=$ac_abs_builddir;;
+ [\\/]* | ?:[\\/]* ) ac_abs_top_srcdir=$ac_top_srcdir;;
+ *) ac_abs_top_srcdir=$ac_abs_builddir/$ac_top_srcdir;;
+ esac;;
+esac
+
+
+
+ if test x"$ac_file" != x-; then
+ { echo "$as_me:$LINENO: creating $ac_file" >&5
+echo "$as_me: creating $ac_file" >&6;}
+ rm -f "$ac_file"
+ fi
+ # Let's still pretend it is `configure' which instantiates (i.e., don't
+ # use $as_me), people would be surprised to read:
+ # /* config.h. Generated by config.status. */
+ if test x"$ac_file" = x-; then
+ configure_input=
+ else
+ configure_input="$ac_file. "
+ fi
+ configure_input=$configure_input"Generated from `echo $ac_file_in |
+ sed 's,.*/,,'` by configure."
+
+ # First look for the input files in the build tree, otherwise in the
+ # src tree.
+ ac_file_inputs=`IFS=:
+ for f in $ac_file_in; do
+ case $f in
+ -) echo $tmp/stdin ;;
+ [\\/$]*)
+ # Absolute (can't be DOS-style, as IFS=:)
+ test -f "$f" || { { echo "$as_me:$LINENO: error: cannot find input file: $f" >&5
+echo "$as_me: error: cannot find input file: $f" >&2;}
+ { (exit 1); exit 1; }; }
+ echo "$f";;
+ *) # Relative
+ if test -f "$f"; then
+ # Build tree
+ echo "$f"
+ elif test -f "$srcdir/$f"; then
+ # Source tree
+ echo "$srcdir/$f"
+ else
+ # /dev/null tree
+ { { echo "$as_me:$LINENO: error: cannot find input file: $f" >&5
+echo "$as_me: error: cannot find input file: $f" >&2;}
+ { (exit 1); exit 1; }; }
+ fi;;
+ esac
+ done` || { (exit 1); exit 1; }
+_ACEOF
+cat >>$CONFIG_STATUS <<_ACEOF
+ sed "$ac_vpsub
+$extrasub
+_ACEOF
+cat >>$CONFIG_STATUS <<\_ACEOF
+:t
+/@[a-zA-Z_][a-zA-Z_0-9]*@/!b
+s, at configure_input@,$configure_input,;t t
+s, at srcdir@,$ac_srcdir,;t t
+s, at abs_srcdir@,$ac_abs_srcdir,;t t
+s, at top_srcdir@,$ac_top_srcdir,;t t
+s, at abs_top_srcdir@,$ac_abs_top_srcdir,;t t
+s, at builddir@,$ac_builddir,;t t
+s, at abs_builddir@,$ac_abs_builddir,;t t
+s, at top_builddir@,$ac_top_builddir,;t t
+s, at abs_top_builddir@,$ac_abs_top_builddir,;t t
+" $ac_file_inputs | (eval "$ac_sed_cmds") >$tmp/out
+ rm -f $tmp/stdin
+ if test x"$ac_file" != x-; then
+ mv $tmp/out $ac_file
+ else
+ cat $tmp/out
+ rm -f $tmp/out
+ fi
+
+done
+_ACEOF
+
+cat >>$CONFIG_STATUS <<\_ACEOF
+
+{ (exit 0); exit 0; }
+_ACEOF
+chmod +x $CONFIG_STATUS
+ac_clean_files=$ac_clean_files_save
+
+
+# configure is writing to config.log, and then calls config.status.
+# config.status does its own redirection, appending to config.log.
+# Unfortunately, on DOS this fails, as config.log is still kept open
+# by configure, so config.status won't be able to write to it; its
+# output is simply discarded. So we exec the FD to /dev/null,
+# effectively closing config.log, so it can be properly (re)opened and
+# appended to by config.status. When coming back to configure, we
+# need to make the FD available again.
+if test "$no_create" != yes; then
+ ac_cs_success=:
+ ac_config_status_args=
+ test "$silent" = yes &&
+ ac_config_status_args="$ac_config_status_args --quiet"
+ exec 5>/dev/null
+ $SHELL $CONFIG_STATUS $ac_config_status_args || ac_cs_success=false
+ exec 5>>config.log
+ # Use ||, not &&, to avoid exiting from the if with $? = 1, which
+ # would make configure fail if this is the last instruction.
+ $ac_cs_success || { (exit 1); exit 1; }
+fi
+
+
+# Messages
+printf '#================================================================\n'
+printf '# Ready to make.\n'
+printf '#================================================================\n'
+
+
+
+# END OF FILE
Property changes on: box/trunk/qdbm/configure
___________________________________________________________________
Added: svn:executable
+ *
Added: box/trunk/qdbm/configure.in
===================================================================
--- box/trunk/qdbm/configure.in (rev 0)
+++ box/trunk/qdbm/configure.in 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,312 @@
+# Source of configuration for QDBM
+
+
+
+#================================================================
+# Generic Settings
+#================================================================
+
+
+# Targets
+AC_INIT(qdbm, 1.8.77)
+
+# Export variables
+LIBVER=14
+LIBREV=13
+TARGETS="all"
+MYDEFS=""
+MYOPTS=""
+MGWLIBS=""
+LD="ld"
+AR="ar"
+
+# Building paths
+pathtmp="$PATH"
+PATH="$HOME/bin:/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin:/usr/local/sbin"
+PATH="$PATH:/usr/ccs/bin:/usr/ucb:/usr/xpg4/bin:/usr/xpg6/bin:$pathtmp"
+LIBRARY_PATH="$HOME/lib:/usr/local/lib:$LIBRARY_PATH"
+LD_LIBRARY_PATH="$HOME/lib:/usr/local/lib:$LD_LIBRARY_PATH"
+CPATH="$HOME/include:/usr/local/include:$CPATH"
+export PATH LIBRARY_PATH LD_LIBRARY_PATH CPATH
+
+
+
+#================================================================
+# Options
+#================================================================
+
+
+# Internal variables
+enables=""
+ispthread=""
+iszlib=""
+isiconv=""
+isnowarn=""
+
+# Debug mode
+AC_ARG_ENABLE(debug,
+ AC_HELP_STRING([--enable-debug], [build for debugging]))
+if test "$enable_debug" = "yes"
+then
+ TARGETS="debug"
+ enables="$enables (debug)"
+fi
+
+# Developping mode
+AC_ARG_ENABLE(devel,
+ AC_HELP_STRING([--enable-devel], [build for development]))
+if test "$enable_devel" = "yes"
+then
+ TARGETS="devel"
+ enables="$enables (devel)"
+fi
+
+# Stable mode
+AC_ARG_ENABLE(stable,
+ AC_HELP_STRING([--enable-stable], [build for stable release]))
+if test "$enable_stable" = "yes"
+then
+ TARGETS="stable"
+ enables="$enables (stable)"
+fi
+
+# Enable POSIX thread
+AC_ARG_ENABLE(pthread,
+ AC_HELP_STRING([--enable-pthread], [use POSIX thread and make APIs thread-safe]))
+if test "$enable_pthread" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYPTHREAD"
+ enables="$enables (pthread)"
+ ispthread="yes"
+fi
+
+# Disable file locking
+AC_ARG_ENABLE(lock,
+ AC_HELP_STRING([--disable-lock], [build for environments without file locking]))
+if test "$enable_lock" = "no"
+then
+ MYDEFS="$MYDEFS -DMYNOLOCK"
+ enables="$enables (no-lock)"
+fi
+
+# Disable memory mapping
+AC_ARG_ENABLE(mmap,
+ AC_HELP_STRING([--disable-mmap], [build for environments without memory mapping]))
+if test "$enable_mmap" = "no"
+then
+ MYDEFS="$MYDEFS -DMYNOMMAP"
+ enables="$enables (no-mmap)"
+fi
+
+# Enable ZLIB compression
+AC_ARG_ENABLE(zlib,
+ AC_HELP_STRING([--enable-zlib], [feature ZLIB for B+ tree and inverted index]))
+if test "$enable_zlib" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYZLIB"
+ MGWLIBS="-lz $MGWLIBS"
+ enables="$enables (zlib)"
+ iszlib="yes"
+fi
+
+# Enable LZO compression
+AC_ARG_ENABLE(lzo,
+ AC_HELP_STRING([--enable-lzo], [feature LZO for B+ tree and inverted index]))
+if test "$enable_lzo" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYLZO"
+ MGWLIBS="-llzo2 $MGWLIBS"
+ enables="$enables (lzo)"
+ islzo="yes"
+fi
+
+# Enable BZIP2 compression
+AC_ARG_ENABLE(bzip,
+ AC_HELP_STRING([--enable-bzip], [feature BZIP2 for B+ tree and inverted index]))
+if test "$enable_bzip" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYBZIP"
+ MGWLIBS="-lbz2 $MGWLIBS"
+ enables="$enables (bzip)"
+ isbzip="yes"
+fi
+
+# Enable ICONV utilities
+AC_ARG_ENABLE(iconv,
+ AC_HELP_STRING([--enable-iconv], [feature ICONV utilities]))
+if test "$enable_iconv" = "yes"
+then
+ MYDEFS="$MYDEFS -DMYICONV"
+ MGWLIBS="-liconv $MGWLIBS"
+ enables="$enables (iconv)"
+ isiconv="yes"
+fi
+
+# No warning configuration
+AC_ARG_ENABLE(warn,
+ AC_HELP_STRING([--disable-warn], [hide warnings in the configuration]))
+if test "$enable_warn" = "no"
+then
+ isnowarn="yes"
+fi
+
+# Messages
+printf '#================================================================\n'
+printf '# Configuring QDBM version %s%s.\n' "$PACKAGE_VERSION" "$enables"
+printf '#================================================================\n'
+
+
+
+#================================================================
+# Checking Commands to Build with
+#================================================================
+
+
+# C compiler
+AC_PROG_CC
+if test "$GCC" != "yes"
+then
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: GCC is required to build this package.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+fi
+if uname | egrep -i 'Linux' > /dev/null 2>&1 &&
+ uname -m | egrep '(x|i)(3|4|5|6|7|8|9)?86' > /dev/null 2>&1
+then
+ MYOPTS="-minline-all-stringops"
+fi
+if uname | egrep -i 'SunOS' > /dev/null 2>&1
+then
+ MYOPTS="-O1 -fno-omit-frame-pointer -fno-force-addr"
+fi
+if uname | egrep -i 'BSD' > /dev/null 2>&1
+then
+ MYOPTS="-O1 -fno-omit-frame-pointer -fno-force-addr"
+fi
+if gcc --version | egrep -i '^2\.(8|9)' > /dev/null 2>&1
+then
+ MYOPTS="-O1 -fno-omit-frame-pointer -fno-force-addr"
+fi
+
+# Linker
+printf 'checking for ld... '
+if which ld | grep '/ld$' > /dev/null 2>&1
+then
+ LD=`which ld`
+ printf '%s\n' "$LD"
+else
+ printf 'no\n'
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: ld is not found in PATH.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+fi
+
+# Archiver
+printf 'checking for ar... '
+if which ar | grep '/ar$' > /dev/null 2>&1
+then
+ AR=`which ar`
+ printf '%s\n' "$AR"
+else
+ printf 'no\n'
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: ar is not found in PATH.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+fi
+
+
+
+#================================================================
+# Checking Libraries
+#================================================================
+
+
+# Underlying libraries
+AC_CHECK_LIB(c, main)
+
+# for pthread
+if test "$ispthread" = "yes"
+then
+ AC_CHECK_LIB(pthread, main)
+fi
+
+# for zlib
+if test "$iszlib" = "yes"
+then
+ AC_CHECK_LIB(z, main)
+fi
+
+# for lzo
+if test "$islzo" = "yes"
+then
+ AC_CHECK_LIB(lzo2, main)
+fi
+
+# for bzip
+if test "$isbzip" = "yes"
+then
+ AC_CHECK_LIB(bz2, main)
+fi
+
+# for iconv
+if test "$isiconv" = "yes"
+then
+ AC_CHECK_LIB(iconv, main)
+fi
+
+# For old BSDs
+if test "$ispthread" = "yes" && uname -a | grep BSD > /dev/null &&
+ test -f /usr/lib/libc_r.a && test ! -f /usr/lib/libpthread.a
+then
+ LIBS=`printf '%s' "$LIBS" | sed 's/-lc/-lc_r/g'`
+fi
+
+# Duplication of QDBM for C
+AC_CHECK_LIB(qdbm, main,
+ if test "$isnowarn" != "yes"
+ then
+ printf '#================================================================\n' 1>&2
+ printf '# WARNING: The existing library was detected.\n' 1>&2
+ printf '#================================================================\n' 1>&2
+ fi
+)
+
+
+
+#================================================================
+# Generic Settings
+#================================================================
+
+
+# Export variables
+AC_SUBST(LIBVER)
+AC_SUBST(LIBREV)
+AC_SUBST(TARGETS)
+AC_SUBST(MYDEFS)
+AC_SUBST(MYOPTS)
+AC_SUBST(MGWLIBS)
+AC_SUBST(LD)
+AC_SUBST(AR)
+AC_SUBST(CPPFLAGS)
+AC_SUBST(LDFLAGS)
+
+# Targets
+AC_OUTPUT(Makefile LTmakefile qdbm.spec qdbm.pc)
+
+# Messages
+printf '#================================================================\n'
+printf '# Ready to make.\n'
+printf '#================================================================\n'
+
+
+
+# END OF FILE
Added: box/trunk/qdbm/crmgr.c
===================================================================
--- box/trunk/qdbm/crmgr.c (rev 0)
+++ box/trunk/qdbm/crmgr.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,956 @@
+/*************************************************************************************************
+ * Utility for debugging Curia and its applications
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <curia.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define ALIGNSIZ 32 /* basic size of alignment */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *hextoobj(const char *str, int *sp);
+char *dectoiobj(const char *str, int *sp);
+int runcreate(int argc, char **argv);
+int runput(int argc, char **argv);
+int runout(int argc, char **argv);
+int runget(int argc, char **argv);
+int runlist(int argc, char **argv);
+int runoptimize(int argc, char **argv);
+int runinform(int argc, char **argv);
+int runremove(int argc, char **argv);
+int runrepair(int argc, char **argv);
+int runexportdb(int argc, char **argv);
+int runimportdb(int argc, char **argv);
+int runsnaffle(int argc, char **argv);
+void pdperror(const char *name);
+void printobj(const char *obj, int size);
+void printobjhex(const char *obj, int size);
+int docreate(const char *name, int bnum, int dnum, int sparse);
+int doput(const char *name, const char *kbuf, int ksiz,
+ const char *vbuf, int vsiz, int dmode, int lob, int align);
+int doout(const char *name, const char *kbuf, int ksiz, int lob);
+int doget(const char *name, int opts, const char *kbuf, int ksiz,
+ int start, int max, int ox, int lob, int nb);
+int dolist(const char *name, int opts, int kb, int vb, int ox);
+int dooptimize(const char *name, int bnum, int align);
+int doinform(const char *name, int opts);
+int doremove(const char *name);
+int dorepair(const char *name);
+int doexportdb(const char *name, const char *dir);
+int doimportdb(const char *name, const char *dir, int bnum, int dnum);
+int dosnaffle(const char *name, const char *kbuf, int ksiz, int ox, int nb);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "create")){
+ rv = runcreate(argc, argv);
+ } else if(!strcmp(argv[1], "put")){
+ rv = runput(argc, argv);
+ } else if(!strcmp(argv[1], "out")){
+ rv = runout(argc, argv);
+ } else if(!strcmp(argv[1], "get")){
+ rv = runget(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else if(!strcmp(argv[1], "optimize")){
+ rv = runoptimize(argc, argv);
+ } else if(!strcmp(argv[1], "inform")){
+ rv = runinform(argc, argv);
+ } else if(!strcmp(argv[1], "remove")){
+ rv = runremove(argc, argv);
+ } else if(!strcmp(argv[1], "repair")){
+ rv = runrepair(argc, argv);
+ } else if(!strcmp(argv[1], "exportdb")){
+ rv = runexportdb(argc, argv);
+ } else if(!strcmp(argv[1], "importdb")){
+ rv = runimportdb(argc, argv);
+ } else if(!strcmp(argv[1], "snaffle")){
+ rv = runsnaffle(argc, argv);
+ } else if(!strcmp(argv[1], "version") || !strcmp(argv[1], "--version")){
+ printf("Powered by QDBM version %s on %s%s\n",
+ dpversion, dpsysname, dpisreentrant ? " (reentrant)" : "");
+ printf("Copyright (c) 2000-2007 Mikio Hirabayashi\n");
+ rv = 0;
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: administration utility for Curia\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s create [-s] [-bnum num] [-dnum num] name\n", progname);
+ fprintf(stderr, " %s put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat] [-lob] [-na] "
+ "name key val\n", progname);
+ fprintf(stderr, " %s out [-kx|-ki] [-lob] name key\n", progname);
+ fprintf(stderr, " %s get [-nl] [-kx|-ki] [-start num] [-max num] [-ox] [-lob] [-n] name key\n",
+ progname);
+ fprintf(stderr, " %s list [-nl] [-k|-v] [-ox] name\n", progname);
+ fprintf(stderr, " %s optimize [-bnum num] [-na] name\n", progname);
+ fprintf(stderr, " %s inform [-nl] name\n", progname);
+ fprintf(stderr, " %s remove name\n", progname);
+ fprintf(stderr, " %s repair name\n", progname);
+ fprintf(stderr, " %s exportdb name dir\n", progname);
+ fprintf(stderr, " %s importdb [-bnum num] [-dnum num] name dir\n", progname);
+ fprintf(stderr, " %s snaffle [-kx|-ki] [-ox] [-n] name key\n", progname);
+ fprintf(stderr, " %s version\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* create a binary object from a hexadecimal string */
+char *hextoobj(const char *str, int *sp){
+ char *buf, mbuf[3];
+ int len, i, j;
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ j = 0;
+ for(i = 0; i < len; i += 2){
+ while(strchr(" \n\r\t\f\v", str[i])){
+ i++;
+ }
+ if((mbuf[0] = str[i]) == '\0') break;
+ if((mbuf[1] = str[i+1]) == '\0') break;
+ mbuf[2] = '\0';
+ buf[j++] = (char)strtol(mbuf, NULL, 16);
+ }
+ buf[j] = '\0';
+ *sp = j;
+ return buf;
+}
+
+
+/* create a integer object from a decimal string */
+char *dectoiobj(const char *str, int *sp){
+ char *buf;
+ int num;
+ num = atoi(str);
+ if(!(buf = malloc(sizeof(int)))) return NULL;
+ *(int *)buf = num;
+ *sp = sizeof(int);
+ return buf;
+}
+
+
+/* parse arguments of create command */
+int runcreate(int argc, char **argv){
+ char *name;
+ int i, sb, bnum, dnum, rv;
+ name = NULL;
+ sb = FALSE;
+ bnum = -1;
+ dnum = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-dnum")){
+ if(++i >= argc) usage();
+ dnum = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docreate(name, bnum, dnum, sb);
+ return rv;
+}
+
+
+/* parse arguments of put command */
+int runput(int argc, char **argv){
+ char *name, *key, *val, *kbuf, *vbuf;
+ int i, kx, ki, vx, vi, vf, lob, ksiz, vsiz, align, rv;
+ int dmode;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ vx = FALSE;
+ vi = FALSE;
+ vf = FALSE;
+ lob = FALSE;
+ align = ALIGNSIZ;
+ key = NULL;
+ val = NULL;
+ dmode = CR_DOVER;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-vx")){
+ vx = TRUE;
+ } else if(!strcmp(argv[i], "-vi")){
+ vi = TRUE;
+ } else if(!strcmp(argv[i], "-vf")){
+ vf = TRUE;
+ } else if(!strcmp(argv[i], "-keep")){
+ dmode = CR_DKEEP;
+ } else if(!strcmp(argv[i], "-cat")){
+ dmode = CR_DCAT;
+ } else if(!strcmp(argv[i], "-lob")){
+ lob = TRUE;
+ } else if(!strcmp(argv[i], "-na")){
+ align = 0;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else if(!val){
+ val = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || !val) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(vx){
+ vbuf = hextoobj(val, &vsiz);
+ } else if(vi){
+ vbuf = dectoiobj(val, &vsiz);
+ } else if(vf){
+ vbuf = cbreadfile(val, &vsiz);
+ } else {
+ vbuf = cbmemdup(val, -1);
+ vsiz = -1;
+ }
+ if(kbuf && vbuf){
+ rv = doput(name, kbuf, ksiz, vbuf, vsiz, dmode, lob, align);
+ } else {
+ if(vf){
+ fprintf(stderr, "%s: %s: cannot read\n", progname, val);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ }
+ rv = 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ return rv;
+}
+
+
+/* parse arguments of out command */
+int runout(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ki, lob, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ lob = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-lob")){
+ lob = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = doout(name, kbuf, ksiz, lob);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of get command */
+int runget(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, opts, kx, ki, ox, lob, nb, start, max, ksiz, rv;
+ name = NULL;
+ opts = 0;
+ kx = FALSE;
+ ki = FALSE;
+ ox = FALSE;
+ lob = FALSE;
+ nb = FALSE;
+ start = 0;
+ max = -1;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= CR_ONOLCK;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-lob")){
+ lob = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else if(!strcmp(argv[i], "-start")){
+ if(++i >= argc) usage();
+ start = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-max")){
+ if(++i >= argc) usage();
+ max = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || start < 0) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = doget(name, opts, kbuf, ksiz, start, max, ox, lob, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ char *name;
+ int i, opts, kb, vb, ox, rv;
+ name = NULL;
+ opts = 0;
+ kb = FALSE;
+ vb = FALSE;
+ ox = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= CR_ONOLCK;
+ } else if(!strcmp(argv[i], "-k")){
+ kb = TRUE;
+ } else if(!strcmp(argv[i], "-v")){
+ vb = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dolist(name, opts, kb, vb, ox);
+ return rv;
+}
+
+
+/* parse arguments of optimize command */
+int runoptimize(int argc, char **argv){
+ char *name;
+ int i, bnum, align, rv;
+ name = NULL;
+ bnum = -1;
+ align = ALIGNSIZ;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-na")){
+ align = 0;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dooptimize(name, bnum, align);
+ return rv;
+}
+
+
+/* parse arguments of inform command */
+int runinform(int argc, char **argv){
+ char *name;
+ int i, opts, rv;
+ name = NULL;
+ opts = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= CR_ONOLCK;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doinform(name, opts);
+ return rv;
+}
+
+
+/* parse arguments of remove command */
+int runremove(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doremove(name);
+ return rv;
+}
+
+
+/* parse arguments of repair command */
+int runrepair(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dorepair(name);
+ return rv;
+}
+
+
+/* parse arguments of exportdb command */
+int runexportdb(int argc, char **argv){
+ char *name, *dir;
+ int i, rv;
+ name = NULL;
+ dir = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else if(!dir){
+ dir = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !dir) usage();
+ rv = doexportdb(name, dir);
+ return rv;
+}
+
+
+/* parse arguments of importdb command */
+int runimportdb(int argc, char **argv){
+ char *name, *dir;
+ int i, bnum, dnum, rv;
+ name = NULL;
+ dir = NULL;
+ bnum = -1;
+ dnum = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-dnum")){
+ if(++i >= argc) usage();
+ dnum = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!dir){
+ dir = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !dir) usage();
+ rv = doimportdb(name, dir, bnum, dnum);
+ return rv;
+}
+
+
+/* parse arguments of snaffle command */
+int runsnaffle(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ki, ox, nb, start, max, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ ox = FALSE;
+ nb = FALSE;
+ start = 0;
+ max = -1;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || start < 0) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = dosnaffle(name, kbuf, ksiz, ox, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* print an object */
+void printobj(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ putchar(obj[i]);
+ }
+}
+
+
+/* print an object as a hexadecimal string */
+void printobjhex(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ printf("%s%02X", i > 0 ? " " : "", ((const unsigned char *)obj)[i]);
+ }
+}
+
+
+/* perform create command */
+int docreate(const char *name, int bnum, int dnum, int sparse){
+ CURIA *curia;
+ int omode;
+ omode = CR_OWRITER | CR_OCREAT | CR_OTRUNC | (sparse ? CR_OSPARSE : 0);
+ if(!(curia = cropen(name, omode, bnum, dnum))){
+ pdperror(name);
+ return 1;
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform put command */
+int doput(const char *name, const char *kbuf, int ksiz,
+ const char *vbuf, int vsiz, int dmode, int lob, int align){
+ CURIA *curia;
+ if(!(curia = cropen(name, CR_OWRITER, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(align > 0 && !crsetalign(curia, ALIGNSIZ)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ if(lob){
+ if(!crputlob(curia, kbuf, ksiz, vbuf, vsiz, dmode)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ } else {
+ if(!crput(curia, kbuf, ksiz, vbuf, vsiz, dmode)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform out command */
+int doout(const char *name, const char *kbuf, int ksiz, int lob){
+ CURIA *curia;
+ if(!(curia = cropen(name, CR_OWRITER, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(lob){
+ if(!croutlob(curia, kbuf, ksiz)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ } else {
+ if(!crout(curia, kbuf, ksiz)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform get command */
+int doget(const char *name, int opts, const char *kbuf, int ksiz,
+ int start, int max, int ox, int lob, int nb){
+ CURIA *curia;
+ char *vbuf;
+ int vsiz;
+ if(!(curia = cropen(name, CR_OREADER | opts, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(lob){
+ if(!(vbuf = crgetlob(curia, kbuf, ksiz, start, max, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ } else {
+ if(!(vbuf = crget(curia, kbuf, ksiz, start, max, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ if(ox){
+ printobjhex(vbuf, vsiz);
+ } else {
+ printobj(vbuf, vsiz);
+ }
+ free(vbuf);
+ if(!nb) putchar('\n');
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform list command */
+int dolist(const char *name, int opts, int kb, int vb, int ox){
+ CURIA *curia;
+ char *kbuf, *vbuf;
+ int ksiz, vsiz;
+ if(!(curia = cropen(name, CR_OREADER | opts, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ criterinit(curia);
+ while((kbuf = criternext(curia, &ksiz)) != NULL){
+ if(!(vbuf = crget(curia, kbuf, ksiz, 0, -1, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ break;
+ }
+ if(ox){
+ if(!vb) printobjhex(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobjhex(vbuf, vsiz);
+ } else {
+ if(!vb) printobj(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobj(vbuf, vsiz);
+ }
+ putchar('\n');
+ free(vbuf);
+ free(kbuf);
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform optimize command */
+int dooptimize(const char *name, int bnum, int align){
+ CURIA *curia;
+ if(!(curia = cropen(name, CR_OWRITER, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(align > 0 && !crsetalign(curia, ALIGNSIZ)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ if(!croptimize(curia, bnum)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform inform command */
+int doinform(const char *name, int opts){
+ CURIA *curia;
+ if(!(curia = cropen(name, CR_OREADER | opts, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ printf("name: %s\n", crname(curia));
+ printf("file size: %.0f\n", crfsizd(curia));
+ printf("all buckets: %d\n", crbnum(curia));
+ printf("used buckets: %d\n", crbusenum(curia));
+ printf("records: %d\n", crrnum(curia));
+ printf("inode number: %d\n", crinode(curia));
+ printf("modified time: %.0f\n", (double)crmtime(curia));
+ printf("LOB records: %d\n", crrnumlob(curia));
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform remove command */
+int doremove(const char *name){
+ if(!crremove(name)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform repair command */
+int dorepair(const char *name){
+ if(!crrepair(name)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform exportdb command */
+int doexportdb(const char *name, const char *dir){
+ CURIA *curia;
+ if(!(curia = cropen(name, CR_OREADER, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(!crexportdb(curia, dir)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform importdb command */
+int doimportdb(const char *name, const char *dir, int bnum, int dnum){
+ CURIA *curia;
+ if(!(curia = cropen(name, CR_OWRITER | CR_OCREAT | CR_OTRUNC, bnum, dnum))){
+ pdperror(name);
+ return 1;
+ }
+ if(!crimportdb(curia, dir)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform snaffle command */
+int dosnaffle(const char *name, const char *kbuf, int ksiz, int ox, int nb){
+ char *vbuf;
+ int vsiz;
+ if(!(vbuf = crsnaffle(name, kbuf, ksiz, &vsiz))){
+ pdperror(name);
+ return 1;
+ }
+ if(ox){
+ printobjhex(vbuf, vsiz);
+ } else {
+ printobj(vbuf, vsiz);
+ }
+ free(vbuf);
+ if(!nb) putchar('\n');
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/crmgr.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/crtest.c
===================================================================
--- box/trunk/qdbm/crtest.c (rev 0)
+++ box/trunk/qdbm/crtest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,873 @@
+/*************************************************************************************************
+ * Test cases of Curia
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <curia.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+#include <time.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define RECBUFSIZ 32 /* buffer for records */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runwrite(int argc, char **argv);
+int runread(int argc, char **argv);
+int runrcat(int argc, char **argv);
+int runcombo(int argc, char **argv);
+int runwicked(int argc, char **argv);
+int printfflush(const char *format, ...);
+void pdperror(const char *name);
+int myrand(void);
+int dowrite(const char *name, int rnum, int bnum, int dnum, int sparse, int lob);
+int doread(const char *name, int wb, int lob);
+int dorcat(const char *name, int rnum, int bnum, int dnum, int pnum,
+ int align, int fbpsiz, int cb);
+int docombo(const char *name);
+int dowicked(const char *name, int rnum, int cb);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "write")){
+ rv = runwrite(argc, argv);
+ } else if(!strcmp(argv[1], "read")){
+ rv = runread(argc, argv);
+ } else if(!strcmp(argv[1], "rcat")){
+ rv = runrcat(argc, argv);
+ } else if(!strcmp(argv[1], "combo")){
+ rv = runcombo(argc, argv);
+ } else if(!strcmp(argv[1], "wicked")){
+ rv = runwicked(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Curia\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s write [-s] [-lob] name rnum bnum dnum\n", progname);
+ fprintf(stderr, " %s read [-wb] [-lob] name\n", progname);
+ fprintf(stderr, " %s rcat [-c] name rnum bnum dnum pnum align fbpsiz\n", progname);
+ fprintf(stderr, " %s combo name\n", progname);
+ fprintf(stderr, " %s wicked [-c] name rnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of write command */
+int runwrite(int argc, char **argv){
+ char *name, *rstr, *bstr, *dstr;
+ int i, rnum, bnum, dnum, sb, lob, rv;
+ name = NULL;
+ rstr = NULL;
+ bstr = NULL;
+ dstr = NULL;
+ rnum = 0;
+ bnum = 0;
+ dnum = 0;
+ sb = FALSE;
+ lob = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else if(!strcmp(argv[i], "-lob")){
+ lob = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!bstr){
+ bstr = argv[i];
+ } else if(!dstr){
+ dstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr || !bstr || !dstr) usage();
+ rnum = atoi(rstr);
+ bnum = atoi(bstr);
+ dnum = atoi(dstr);
+ if(rnum < 1 || bnum < 1 || dnum < 1) usage();
+ rv = dowrite(name, rnum, bnum, dnum, sb, lob);
+ return rv;
+}
+
+
+/* parse arguments of read command */
+int runread(int argc, char **argv){
+ char *name;
+ int i, wb, lob, rv;
+ name = NULL;
+ wb = FALSE;
+ lob = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-wb")){
+ wb = TRUE;
+ } else if(!strcmp(argv[i], "-lob")){
+ lob = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doread(name, wb, lob);
+ return rv;
+}
+
+
+/* parse arguments of rcat command */
+int runrcat(int argc, char **argv){
+ char *name, *rstr, *bstr, *dstr, *pstr, *astr, *fstr;
+ int i, rnum, bnum, dnum, pnum, align, fbpsiz, cb, rv;
+ name = NULL;
+ rstr = NULL;
+ bstr = NULL;
+ dstr = NULL;
+ pstr = NULL;
+ astr = NULL;
+ fstr = NULL;
+ rnum = 0;
+ bnum = 0;
+ pnum = 0;
+ align = 0;
+ fbpsiz = 0;
+ cb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-c")){
+ cb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!bstr){
+ bstr = argv[i];
+ } else if(!dstr){
+ dstr = argv[i];
+ } else if(!pstr){
+ pstr = argv[i];
+ } else if(!astr){
+ astr = argv[i];
+ } else if(!fstr){
+ fstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr || !bstr || !dstr || !pstr || !astr || !fstr) usage();
+ rnum = atoi(rstr);
+ bnum = atoi(bstr);
+ dnum = atoi(dstr);
+ pnum = atoi(pstr);
+ align = atoi(astr);
+ fbpsiz = atoi(fstr);
+ if(rnum < 1 || bnum < 1 || dnum < 1 || pnum < 1 || fbpsiz < 0) usage();
+ rv = dorcat(name, rnum, bnum, dnum, pnum, align, fbpsiz, cb);
+ return rv;
+}
+
+
+/* parse arguments of combo command */
+int runcombo(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docombo(name);
+ return rv;
+}
+
+
+/* parse arguments of wicked command */
+int runwicked(int argc, char **argv){
+ char *name, *rstr;
+ int i, rnum, cb, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ cb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-c")){
+ cb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowicked(name, rnum, cb);
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* pseudo random number generator */
+int myrand(void){
+ static int cnt = 0;
+ if(cnt == 0) srand(time(NULL));
+ return (rand() * rand() + (rand() >> (sizeof(int) * 4)) + (cnt++)) & INT_MAX;
+}
+
+
+/* perform write command */
+int dowrite(const char *name, int rnum, int bnum, int dnum, int sparse, int lob){
+ CURIA *curia;
+ int i, omode, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Writing Test>\n name=%s rnum=%d bnum=%d dnum=%d s=%d lob=%d\n\n",
+ name, rnum, bnum, dnum, sparse, lob);
+ /* open a database */
+ omode = CR_OWRITER | CR_OCREAT | CR_OTRUNC | (sparse ? CR_OSPARSE : 0);
+ if(!(curia = cropen(name, omode, bnum, dnum))){
+ pdperror(name);
+ return 1;
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* store a record */
+ len = sprintf(buf, "%08d", i);
+ if(lob){
+ if(!crputlob(curia, buf, len, buf, len, CR_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ if(!crput(curia, buf, len, buf, len, CR_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform read command */
+int doread(const char *name, int wb, int lob){
+ CURIA *curia;
+ int i, rnum, err, len;
+ char buf[RECBUFSIZ], vbuf[RECBUFSIZ], *val;
+ printfflush("<Reading Test>\n name=%s wb=%d lob=%d\n\n", name, wb, lob);
+ /* open a database */
+ if(!(curia = cropen(name, CR_OREADER, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ /* get the number of records */
+ if(lob){
+ rnum = crrnumlob(curia);
+ } else {
+ rnum = crrnum(curia);
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* retrieve a record */
+ len = sprintf(buf, "%08d", i);
+ if(lob){
+ if(!(val = crgetlob(curia, buf, len, 0, -1, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ } else if(wb){
+ if(crgetwb(curia, buf, len, 0, RECBUFSIZ, vbuf) == -1){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ if(!(val = crget(curia, buf, len, 0, -1, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform rcat command */
+int dorcat(const char *name, int rnum, int bnum, int dnum, int pnum,
+ int align, int fbpsiz, int cb){
+ CURIA *curia;
+ CBMAP *map;
+ int i, err, len, ksiz, vsiz, rsiz;
+ const char *kbuf, *vbuf;
+ char buf[RECBUFSIZ], *rbuf;
+ printfflush("<Random Writing Test>\n name=%s rnum=%d bnum=%d dnum=%d pnum=%d"
+ " align=%d fbpsiz=%d c=%d\n\n", name, rnum, bnum, dnum, pnum, align, fbpsiz, cb);
+ if(!(curia = cropen(name, CR_OWRITER | CR_OCREAT | CR_OTRUNC, bnum, dnum))){
+ pdperror(name);
+ return 1;
+ }
+ if(!crsetalign(curia, align) || !crsetfbpsiz(curia, fbpsiz)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ map = NULL;
+ if(cb) map = cbmapopen();
+ err = FALSE;
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", myrand() % pnum + 1);
+ if(!crput(curia, buf, len, buf, len, CR_DCAT)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d: fsiz=%d rnum=%d)\n", i, crfsiz(curia), crrnum(curia));
+ }
+ }
+ }
+ if(map){
+ printfflush("Matching records ... ");
+ cbmapiterinit(map);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ vbuf = cbmapget(map, kbuf, ksiz, &vsiz);
+ if(!(rbuf = crget(curia, kbuf, ksiz, 0, -1, &rsiz))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(rsiz != vsiz || memcmp(rbuf, vbuf, rsiz)){
+ fprintf(stderr, "%s: %s: unmatched record\n", progname, name);
+ free(rbuf);
+ err = TRUE;
+ break;
+ }
+ free(rbuf);
+ }
+ cbmapclose(map);
+ if(!err) printfflush("ok\n");
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform combo command */
+int docombo(const char *name){
+ CURIA *curia;
+ char buf[RECBUFSIZ], wbuf[RECBUFSIZ], *vbuf;
+ int i, len, wlen, vsiz;
+ printfflush("<Combination Test>\n name=%s\n\n", name);
+ printfflush("Creating a database with bnum 3 ... ");
+ if(!(curia = cropen(name, CR_OWRITER | CR_OCREAT | CR_OTRUNC, 3, 3))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Setting alignment as 16 ... ");
+ if(!crsetalign(curia, 16)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 20 records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!crput(curia, buf, len, buf, len, CR_DOVER)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = crget(curia, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != crvsiz(curia, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 10 records without moving rooms ... ");
+ for(i = 1; i <= 10; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!crput(curia, buf, len, buf, len, CR_DOVER)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 5 records with moving rooms ... ");
+ for(i = 1; i <= 5; i++){
+ len = sprintf(buf, "%08d", i);
+ wlen = sprintf(wbuf, "%024d", i);
+ if(!crput(curia, buf, len, wbuf, wlen, CR_DOVER)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 15 records in concatenation with moving rooms ... ");
+ for(i = 1; i <= 15; i++){
+ len = sprintf(buf, "%08d", i);
+ wlen = sprintf(wbuf, "========================");
+ if(!crput(curia, buf, len, wbuf, wlen, CR_DCAT)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = crget(curia, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != crvsiz(curia, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting top 10 records ... ");
+ for(i = 1; i <= 10; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!crout(curia, buf, len)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking deleted records ... ");
+ for(i = 1; i <= 10; i++){
+ len = sprintf(buf, "%08d", i);
+ vbuf = crget(curia, buf, len, 0, -1, &vsiz);
+ free(vbuf);
+ if(vbuf || dpecode != DP_ENOITEM){
+ fprintf(stderr, "%s: %s: deleting failed\n", progname, name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 15 records in concatenation with moving rooms ... ");
+ for(i = 1; i <= 15; i++){
+ len = sprintf(buf, "%08d", i);
+ wlen = sprintf(wbuf, "========================");
+ if(!crput(curia, buf, len, wbuf, wlen, CR_DCAT)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = crget(curia, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != crvsiz(curia, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Optimizing the database ... ");
+ if(!croptimize(curia, -1)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = crget(curia, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != crvsiz(curia, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Closing database ... ");
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Creating a database with bnum 1000000 ... ");
+ if(!(curia = cropen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, 1000000, -1))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 1000 records ... ");
+ for(i = 1; i <= 1000; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!crput(curia, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Adding 64 records ... ");
+ for(i = 1; i <= 64; i++){
+ len = sprintf(buf, "%o", i);
+ if(!crput(curia, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Syncing the database ... ");
+ if(!crsync(curia)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Retrieving records directly ... ");
+ for(i = 1; i <= 64; i++){
+ len = sprintf(buf, "%o", i);
+ if(!(vbuf = crsnaffle(name, buf, len, &vsiz))){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ if(strcmp(vbuf, buf)){
+ fprintf(stderr, "%s: %s: invalid content\n", progname, name);
+ free(vbuf);
+ crclose(curia);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != crvsiz(curia, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ crclose(curia);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Optimizing the database ... ");
+ if(!croptimize(curia, -1)){
+ pdperror(name);
+ crclose(curia);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("all ok\n\n");
+ return 0;
+}
+
+
+/* perform wicked command */
+int dowicked(const char *name, int rnum, int cb){
+ CURIA *curia;
+ CBMAP *map;
+ int i, len, err, align, mksiz, mvsiz, rsiz;
+ const char *mkbuf, *mvbuf;
+ char buf[RECBUFSIZ], vbuf[RECBUFSIZ], *val;
+ printfflush("<Wicked Writing Test>\n name=%s rnum=%d\n\n", name, rnum);
+ err = FALSE;
+ if(!(curia = cropen(name, CR_OWRITER | CR_OCREAT | CR_OTRUNC, rnum / 10, 5))){
+ pdperror(name);
+ return 1;
+ }
+ if(!crsetalign(curia, 16) || !crsetfbpsiz(curia, 256)){
+ pdperror(name);
+ err = TRUE;
+ }
+ map = NULL;
+ if(cb) map = cbmapopen();
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", myrand() % rnum + 1);
+ switch(myrand() % 16){
+ case 0:
+ putchar('O');
+ if(!crput(curia, buf, len, buf, len, CR_DOVER)) err = TRUE;
+ if(map) cbmapput(map, buf, len, buf, len, TRUE);
+ break;
+ case 1:
+ putchar('K');
+ if(!crput(curia, buf, len, buf, len, CR_DKEEP) && dpecode != DP_EKEEP) err = TRUE;
+ if(map) cbmapput(map, buf, len, buf, len, FALSE);
+ break;
+ case 2:
+ putchar('D');
+ if(!crout(curia, buf, len) && dpecode != DP_ENOITEM) err = TRUE;
+ if(map) cbmapout(map, buf, len);
+ break;
+ case 3:
+ putchar('G');
+ if(crgetwb(curia, buf, len, 2, RECBUFSIZ, vbuf) == -1 && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 4:
+ putchar('V');
+ if(crvsiz(curia, buf, len) == -1 && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ default:
+ putchar('C');
+ if(!crput(curia, buf, len, buf, len, CR_DCAT)) err = TRUE;
+ if(map) cbmapputcat(map, buf, len, buf, len);
+ break;
+ }
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ if(!err && rnum > 100 && myrand() % (rnum / 100) == 0){
+ if(myrand() % 10 == 0){
+ align = (myrand() % 4 + 1) * -1;
+ } else {
+ align = myrand() % 32;
+ }
+ if(!crsetalign(curia, align)) err = TRUE;
+ }
+ if(err){
+ pdperror(name);
+ break;
+ }
+ }
+ if(!croptimize(curia, -1)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!crput(curia, buf, len, ":", -1, CR_DCAT)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(map) cbmapputcat(map, buf, len, ":", -1);
+ putchar(':');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!croptimize(curia, -1)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(val = crget(curia, buf, len, 0, -1, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ putchar('=');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!criterinit(curia)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= rnum; i++){
+ if(!(val = criternext(curia, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ putchar('@');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(map){
+ printfflush("Matching records ... ");
+ cbmapiterinit(map);
+ while((mkbuf = cbmapiternext(map, &mksiz)) != NULL){
+ mvbuf = cbmapget(map, mkbuf, mksiz, &mvsiz);
+ if(!(val = crget(curia, mkbuf, mksiz, 0, -1, &rsiz))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(rsiz != mvsiz || memcmp(val, mvbuf, rsiz)){
+ fprintf(stderr, "%s: %s: unmatched record\n", progname, name);
+ free(val);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ }
+ cbmapclose(map);
+ if(!err) printfflush("ok\n");
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/crtest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/crtsv.c
===================================================================
--- box/trunk/qdbm/crtsv.c (rev 0)
+++ box/trunk/qdbm/crtsv.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,266 @@
+/*************************************************************************************************
+ * Mutual converter between a database of Curia and a TSV text
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <curia.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runimport(int argc, char **argv);
+int runexport(int argc, char **argv);
+void pdperror(const char *name);
+char *getl(void);
+int doimport(const char *name, int bnum, int dnum, int bin);
+int doexport(const char *name, int bin);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "import")){
+ rv = runimport(argc, argv);
+ } else if(!strcmp(argv[1], "export")){
+ rv = runexport(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: mutual converter between TSV and Curia database\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s import [-bnum num] [-dnum num] [-bin] name\n", progname);
+ fprintf(stderr, " %s export [-bin] name\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of import command */
+int runimport(int argc, char **argv){
+ char *name;
+ int i, bnum, dnum, bin, rv;
+ name = NULL;
+ bnum = -1;
+ dnum = -1;
+ bin = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-dnum")){
+ if(++i >= argc) usage();
+ dnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-bin")){
+ bin = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doimport(name, bnum, dnum, bin);
+ return rv;
+}
+
+
+/* parse arguments of export command */
+int runexport(int argc, char **argv){
+ char *name;
+ int i, bin, rv;
+ name = NULL;
+ bin = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bin")){
+ bin = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doexport(name, bin);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* read a line */
+char *getl(void){
+ char *buf;
+ int c, len, blen;
+ buf = NULL;
+ len = 0;
+ blen = 256;
+ while((c = getchar()) != EOF){
+ if(blen <= len) blen *= 2;
+ buf = cbrealloc(buf, blen + 1);
+ if(c == '\n') c = '\0';
+ buf[len++] = c;
+ if(c == '\0') break;
+ }
+ if(!buf) return NULL;
+ buf[len] = '\0';
+ return buf;
+}
+
+
+/* perform import command */
+int doimport(const char *name, int bnum, int dnum, int bin){
+ CURIA *curia;
+ char *buf, *kbuf, *vbuf, *ktmp, *vtmp;
+ int i, err, ktsiz, vtsiz;
+ /* open a database */
+ if(!(curia = cropen(name, CR_OWRITER | CR_OCREAT, bnum, dnum))){
+ pdperror(name);
+ return 1;
+ }
+ /* loop for each line */
+ err = FALSE;
+ for(i = 1; (buf = getl()) != NULL; i++){
+ kbuf = buf;
+ if((vbuf = strchr(buf, '\t')) != NULL){
+ *vbuf = '\0';
+ vbuf++;
+ /* store a record */
+ if(bin){
+ ktmp = cbbasedecode(kbuf, &ktsiz);
+ vtmp = cbbasedecode(vbuf, &vtsiz);
+ if(!crput(curia, ktmp, ktsiz, vtmp, vtsiz, CR_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ }
+ free(vtmp);
+ free(ktmp);
+ } else {
+ if(!crput(curia, kbuf, -1, vbuf, -1, CR_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ }
+ }
+ } else {
+ fprintf(stderr, "%s: %s: invalid format in line %d\n", progname, name, i);
+ }
+ free(buf);
+ if(err) break;
+ }
+ /* close the database */
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return err ? 1 : 0;
+}
+
+
+/* perform export command */
+int doexport(const char *name, int bin){
+ CURIA *curia;
+ char *kbuf, *vbuf, *tmp;
+ int err, ksiz, vsiz;
+ /* open a database */
+ if(!(curia = cropen(name, CR_OREADER, -1, -1))){
+ pdperror(name);
+ return 1;
+ }
+ /* initialize the iterator */
+ criterinit(curia);
+ /* loop for each key */
+ err = FALSE;
+ while((kbuf = criternext(curia, &ksiz)) != NULL){
+ /* retrieve a value with a key */
+ if(!(vbuf = crget(curia, kbuf, ksiz, 0, -1, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ err = TRUE;
+ break;
+ }
+ /* output data */
+ if(bin){
+ tmp = cbbaseencode(kbuf, ksiz);
+ printf("%s\t", tmp);
+ free(tmp);
+ tmp = cbbaseencode(vbuf, vsiz);
+ printf("%s\n", tmp);
+ free(tmp);
+ } else {
+ printf("%s\t%s\n", kbuf, vbuf);
+ }
+ /* free resources */
+ free(vbuf);
+ free(kbuf);
+ }
+ /* check whether all records were retrieved */
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ /* close the database */
+ if(!crclose(curia)){
+ pdperror(name);
+ return 1;
+ }
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/crtsv.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/curia.c
===================================================================
--- box/trunk/qdbm/curia.c (rev 0)
+++ box/trunk/qdbm/curia.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,1192 @@
+/*************************************************************************************************
+ * Implementation of Curia
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "curia.h"
+#include "myconf.h"
+
+#define CR_NAMEMAX 512 /* max size of a database name */
+#define CR_DPMAX 512 /* max number of division of a database */
+#define CR_DIRMODE 00755 /* permission of a creating directory */
+#define CR_FILEMODE 00644 /* permission of a creating file */
+#define CR_PATHBUFSIZ 1024 /* size of a path buffer */
+#define CR_DEFDNUM 5 /* default number of division of a database */
+#define CR_ATTRBNUM 16 /* bucket number of attrubute database */
+#define CR_DPNAME "depot" /* name of each sub database */
+#define CR_KEYDNUM "dnum" /* key of division number */
+#define CR_KEYLRNUM "lrnum" /* key of the number of large objects */
+#define CR_TMPFSUF MYEXTSTR "crtmp" /* suffix of a temporary directory */
+#define CR_LOBDIR "lob" /* name of the directory of large objects */
+#define CR_LOBDDEPTH 2 /* depth of the directories of large objects */
+#define CR_NUMBUFSIZ 32 /* size of a buffer for a number */
+#define CR_IOBUFSIZ 8192 /* size of an I/O buffer */
+
+
+/* private function prototypes */
+static char *crstrdup(const char *str);
+static int crdpgetnum(DEPOT *depot, const char *kbuf, int ksiz);
+static char *crgetlobpath(CURIA *curia, const char *kbuf, int ksiz);
+static int crmklobdir(const char *path);
+static int crrmlobdir(const char *path);
+static int crcplobdir(CURIA *curia, const char *path);
+static int crwrite(int fd, const void *buf, int size);
+static int crread(int fd, void *buf, int size);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* Get a database handle. */
+CURIA *cropen(const char *name, int omode, int bnum, int dnum){
+ DEPOT *attr, **depots;
+ char path[CR_PATHBUFSIZ], *tname;
+ int i, j, dpomode, inode, lrnum;
+ struct stat sbuf;
+ CURIA *curia;
+ assert(name);
+ if(dnum < 1) dnum = CR_DEFDNUM;
+ if(dnum > CR_DPMAX) dnum = CR_DPMAX;
+ if(strlen(name) > CR_NAMEMAX){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return NULL;
+ }
+ dpomode = DP_OREADER;
+ if(omode & CR_OWRITER){
+ dpomode = DP_OWRITER;
+ if(omode & CR_OCREAT) dpomode |= DP_OCREAT;
+ if(omode & CR_OTRUNC) dpomode |= DP_OTRUNC;
+ if(omode & CR_OSPARSE) dpomode |= DP_OSPARSE;
+ }
+ if(omode & CR_ONOLCK) dpomode |= DP_ONOLCK;
+ if(omode & CR_OLCKNB) dpomode |= DP_OLCKNB;
+ attr = NULL;
+ lrnum = 0;
+ if((omode & CR_OWRITER) && (omode & CR_OCREAT)){
+ if(mkdir(name, CR_DIRMODE) == -1 && errno != EEXIST){
+ dpecodeset(DP_EMKDIR, __FILE__, __LINE__);
+ return NULL;
+ }
+ sprintf(path, "%s%c%s", name, MYPATHCHR, CR_DPNAME);
+ if(!(attr = dpopen(path, dpomode, CR_ATTRBNUM))) return NULL;
+ if(dprnum(attr) > 0){
+ if((dnum = crdpgetnum(attr, CR_KEYDNUM, -1)) < 1 ||
+ (lrnum = crdpgetnum(attr, CR_KEYLRNUM, -1)) < 0){
+ dpclose(attr);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ } else {
+ if(!dpput(attr, CR_KEYDNUM, -1, (char *)&dnum, sizeof(int), DP_DOVER) ||
+ !dpput(attr, CR_KEYLRNUM, -1, (char *)&lrnum, sizeof(int), DP_DOVER)){
+ dpclose(attr);
+ return NULL;
+ }
+ for(i = 0; i < dnum; i++){
+ sprintf(path, "%s%c%04d", name, MYPATHCHR, i + 1);
+ if(mkdir(path, CR_DIRMODE) == -1 && errno != EEXIST){
+ dpclose(attr);
+ dpecodeset(DP_EMKDIR, __FILE__, __LINE__);
+ return NULL;
+ }
+ }
+ }
+ }
+ if(!attr){
+ sprintf(path, "%s%c%s", name, MYPATHCHR, CR_DPNAME);
+ if(!(attr = dpopen(path, dpomode, 1))) return NULL;
+ if(!(omode & CR_OTRUNC)){
+ if((dnum = crdpgetnum(attr, CR_KEYDNUM, -1)) < 1 ||
+ (lrnum = crdpgetnum(attr, CR_KEYLRNUM, -1)) < 0){
+ dpclose(attr);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ }
+ }
+ if(omode & CR_OTRUNC){
+ for(i = 0; i < CR_DPMAX; i++){
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, i + 1, MYPATHCHR, CR_DPNAME);
+ if(unlink(path) == -1 && errno != ENOENT){
+ dpclose(attr);
+ dpecodeset(DP_EUNLINK, __FILE__, __LINE__);
+ return NULL;
+ }
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, i + 1, MYPATHCHR, CR_LOBDIR);
+ if(!crrmlobdir(path)){
+ dpclose(attr);
+ return NULL;
+ }
+ if(i >= dnum){
+ sprintf(path, "%s%c%04d", name, MYPATHCHR, i + 1);
+ if(rmdir(path) == -1 && errno != ENOENT){
+ dpclose(attr);
+ dpecodeset(DP_ERMDIR, __FILE__, __LINE__);
+ return NULL;
+ }
+ }
+ }
+ errno = 0;
+ }
+ if(lstat(name, &sbuf) == -1){
+ dpclose(attr);
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return NULL;
+ }
+ inode = sbuf.st_ino;
+ if(!(depots = malloc(dnum * sizeof(DEPOT *)))){
+ dpclose(attr);
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ for(i = 0; i < dnum; i++){
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, i + 1, MYPATHCHR, CR_DPNAME);
+ if(!(depots[i] = dpopen(path, dpomode, bnum))){
+ for(j = 0; j < i; j++){
+ dpclose(depots[j]);
+ }
+ free(depots);
+ dpclose(attr);
+ return NULL;
+ }
+ }
+ curia = malloc(sizeof(CURIA));
+ tname = crstrdup(name);
+ if(!curia || !tname){
+ free(curia);
+ free(tname);
+ for(i = 0; i < dnum; i++){
+ dpclose(depots[i]);
+ }
+ free(depots);
+ dpclose(attr);
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ curia->name = tname;
+ curia->wmode = (omode & CR_OWRITER);
+ curia->inode = inode;
+ curia->attr = attr;
+ curia->depots = depots;
+ curia->dnum = dnum;
+ curia->inum = 0;
+ curia->lrnum = lrnum;
+ return curia;
+}
+
+
+/* Close a database handle. */
+int crclose(CURIA *curia){
+ int i, err;
+ assert(curia);
+ err = FALSE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpclose(curia->depots[i])) err = TRUE;
+ }
+ free(curia->depots);
+ if(curia->wmode){
+ if(!dpput(curia->attr, CR_KEYLRNUM, -1, (char *)&(curia->lrnum), sizeof(int), DP_DOVER))
+ err = TRUE;
+ }
+ if(!dpclose(curia->attr)) err = TRUE;
+ free(curia->name);
+ free(curia);
+ return err ? FALSE : TRUE;
+}
+
+
+/* Store a record. */
+int crput(CURIA *curia, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode){
+ int dpdmode;
+ int tnum;
+ assert(curia && kbuf && vbuf);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ switch(dmode){
+ case CR_DKEEP: dpdmode = DP_DKEEP; break;
+ case CR_DCAT: dpdmode = DP_DCAT; break;
+ default: dpdmode = DP_DOVER; break;
+ }
+ tnum = dpouterhash(kbuf, ksiz) % curia->dnum;
+ return dpput(curia->depots[tnum], kbuf, ksiz, vbuf, vsiz, dpdmode);
+}
+
+
+/* Delete a record. */
+int crout(CURIA *curia, const char *kbuf, int ksiz){
+ int tnum;
+ assert(curia && kbuf);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ tnum = dpouterhash(kbuf, ksiz) % curia->dnum;
+ return dpout(curia->depots[tnum], kbuf, ksiz);
+}
+
+
+/* Retrieve a record. */
+char *crget(CURIA *curia, const char *kbuf, int ksiz, int start, int max, int *sp){
+ int tnum;
+ assert(curia && kbuf && start >= 0);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ tnum = dpouterhash(kbuf, ksiz) % curia->dnum;
+ return dpget(curia->depots[tnum], kbuf, ksiz, start, max, sp);
+}
+
+
+/* Retrieve a record and write the value into a buffer. */
+int crgetwb(CURIA *curia, const char *kbuf, int ksiz, int start, int max, char *vbuf){
+ int tnum;
+ assert(curia && kbuf && start >= 0 && max >= 0 && vbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ tnum = dpouterhash(kbuf, ksiz) % curia->dnum;
+ return dpgetwb(curia->depots[tnum], kbuf, ksiz, start, max, vbuf);
+}
+
+
+/* Get the size of the value of a record. */
+int crvsiz(CURIA *curia, const char *kbuf, int ksiz){
+ int tnum;
+ assert(curia && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ tnum = dpouterhash(kbuf, ksiz) % curia->dnum;
+ return dpvsiz(curia->depots[tnum], kbuf, ksiz);
+}
+
+
+/* Initialize the iterator of a database handle. */
+int criterinit(CURIA *curia){
+ int i, err;
+ assert(curia);
+ err = FALSE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpiterinit(curia->depots[i])){
+ err = TRUE;
+ break;
+ }
+ }
+ curia->inum = 0;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Get the next key of the iterator. */
+char *criternext(CURIA *curia, int *sp){
+ char *kbuf;
+ assert(curia);
+ kbuf = NULL;
+ while(curia->inum < curia->dnum && !(kbuf = dpiternext(curia->depots[curia->inum], sp))){
+ if(dpecode != DP_ENOITEM) return NULL;
+ (curia->inum)++;
+ }
+ return kbuf;
+}
+
+
+/* Set alignment of a database handle. */
+int crsetalign(CURIA *curia, int align){
+ int i, err;
+ assert(curia);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpsetalign(curia->depots[i], align)){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Set the size of the free block pool of a database handle. */
+int crsetfbpsiz(CURIA *curia, int size){
+ int i, err;
+ assert(curia && size >= 0);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpsetfbpsiz(curia->depots[i], size)){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Synchronize contents of updating a database with the files and the devices. */
+int crsync(CURIA *curia){
+ int i, err;
+ assert(curia);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ if(!dpput(curia->attr, CR_KEYLRNUM, -1, (char *)&(curia->lrnum), sizeof(int), DP_DOVER) ||
+ !dpsync(curia->attr)) err = TRUE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpsync(curia->depots[i])){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Optimize a database. */
+int croptimize(CURIA *curia, int bnum){
+ int i, err;
+ assert(curia);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpoptimize(curia->depots[i], bnum)){
+ err = TRUE;
+ break;
+ }
+ }
+ curia->inum = 0;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Get the name of a database. */
+char *crname(CURIA *curia){
+ char *name;
+ assert(curia);
+ if(!(name = crstrdup(curia->name))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ return name;
+}
+
+
+/* Get the total size of database files. */
+int crfsiz(CURIA *curia){
+ int i, sum, rv;
+ assert(curia);
+ if((sum = dpfsiz(curia->attr)) == -1) return -1;
+ for(i = 0; i < curia->dnum; i++){
+ if((rv = dpfsiz(curia->depots[i])) == -1) return -1;
+ sum += rv;
+ }
+ return sum;
+}
+
+
+/* Get the total size of database files as double-precision value. */
+double crfsizd(CURIA *curia){
+ double sum;
+ int i, rv;
+ assert(curia);
+ sum = 0.0;
+ if((sum = dpfsiz(curia->attr)) < 0) return -1.0;
+ for(i = 0; i < curia->dnum; i++){
+ if((rv = dpfsiz(curia->depots[i])) == -1) return -1.0;
+ sum += rv;
+ }
+ return sum;
+}
+
+
+/* Get the total number of the elements of each bucket array. */
+int crbnum(CURIA *curia){
+ int i, sum, rv;
+ assert(curia);
+ sum = 0;
+ for(i = 0; i < curia->dnum; i++){
+ rv = dpbnum(curia->depots[i]);
+ if(rv == -1) return -1;
+ sum += rv;
+ }
+ return sum;
+}
+
+
+/* Get the total number of the used elements of each bucket array. */
+int crbusenum(CURIA *curia){
+ int i, sum, rv;
+ assert(curia);
+ sum = 0;
+ for(i = 0; i < curia->dnum; i++){
+ rv = dpbusenum(curia->depots[i]);
+ if(rv == -1) return -1;
+ sum += rv;
+ }
+ return sum;
+}
+
+
+/* Get the number of the records stored in a database. */
+int crrnum(CURIA *curia){
+ int i, sum, rv;
+ assert(curia);
+ sum = 0;
+ for(i = 0; i < curia->dnum; i++){
+ rv = dprnum(curia->depots[i]);
+ if(rv == -1) return -1;
+ sum += rv;
+ }
+ return sum;
+}
+
+
+/* Check whether a database handle is a writer or not. */
+int crwritable(CURIA *curia){
+ assert(curia);
+ return curia->wmode;
+}
+
+
+/* Check whether a database has a fatal error or not. */
+int crfatalerror(CURIA *curia){
+ int i;
+ assert(curia);
+ if(dpfatalerror(curia->attr)) return TRUE;
+ for(i = 0; i < curia->dnum; i++){
+ if(dpfatalerror(curia->depots[i])) return TRUE;
+ }
+ return FALSE;
+}
+
+
+/* Get the inode number of a database directory. */
+int crinode(CURIA *curia){
+ assert(curia);
+ return curia->inode;
+}
+
+
+/* Get the last modified time of a database. */
+time_t crmtime(CURIA *curia){
+ assert(curia);
+ return dpmtime(curia->attr);
+}
+
+
+/* Remove a database directory. */
+int crremove(const char *name){
+ struct stat sbuf;
+ CURIA *curia;
+ char path[CR_PATHBUFSIZ];
+ assert(name);
+ if(lstat(name, &sbuf) == -1){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if((curia = cropen(name, CR_OWRITER | CR_OTRUNC, 1, 1)) != NULL) crclose(curia);
+ sprintf(path, "%s%c0001%c%s", name, MYPATHCHR, MYPATHCHR, CR_DPNAME);
+ dpremove(path);
+ sprintf(path, "%s%c0001", name, MYPATHCHR);
+ if(rmdir(path) == -1){
+ dpecodeset(DP_ERMDIR, __FILE__, __LINE__);
+ return FALSE;
+ }
+ sprintf(path, "%s%c%s", name, MYPATHCHR, CR_DPNAME);
+ if(!dpremove(path)) return FALSE;
+ if(rmdir(name) == -1){
+ dpecodeset(DP_ERMDIR, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Repair a broken database directory. */
+int crrepair(const char *name){
+ CURIA *tcuria;
+ DEPOT *tdepot;
+ char path[CR_PATHBUFSIZ], *kbuf, *vbuf;
+ struct stat sbuf;
+ int i, j, err, flags, bnum, dnum, ksiz, vsiz;
+ assert(name);
+ err = FALSE;
+ flags = 0;
+ bnum = 0;
+ dnum = 0;
+ sprintf(path, "%s%c%s", name, MYPATHCHR, CR_DPNAME);
+ if(lstat(path, &sbuf) != -1){
+ if((tdepot = dpopen(path, DP_OREADER, -1)) != NULL){
+ flags = dpgetflags(tdepot);
+ dpclose(tdepot);
+ }
+ }
+ for(i = 1; i <= CR_DPMAX; i++){
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, i, MYPATHCHR, CR_DPNAME);
+ if(lstat(path, &sbuf) != -1){
+ dnum++;
+ if(!dprepair(path)) err = TRUE;
+ if((tdepot = dpopen(path, DP_OREADER, -1)) != NULL){
+ bnum += dpbnum(tdepot);
+ dpclose(tdepot);
+ }
+ }
+ }
+ if(dnum < CR_DEFDNUM) dnum = CR_DEFDNUM;
+ bnum /= dnum;
+ sprintf(path, "%s%s", name, CR_TMPFSUF);
+ if((tcuria = cropen(path, CR_OWRITER | CR_OCREAT | CR_OTRUNC, bnum, dnum)) != NULL){
+ if(!crsetflags(tcuria, flags)) err = TRUE;
+ for(i = 1; i <= CR_DPMAX; i++){
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, i, MYPATHCHR, CR_DPNAME);
+ if(lstat(path, &sbuf) != -1){
+ if((tdepot = dpopen(path, DP_OREADER, -1)) != NULL){
+ if(!dpiterinit(tdepot)) err = TRUE;
+ while((kbuf = dpiternext(tdepot, &ksiz)) != NULL){
+ if((vbuf = dpget(tdepot, kbuf, ksiz, 0, -1, &vsiz)) != NULL){
+ if(!crput(tcuria, kbuf, ksiz, vbuf, vsiz, CR_DKEEP)) err = TRUE;
+ free(vbuf);
+ }
+ free(kbuf);
+ }
+ dpclose(tdepot);
+ } else {
+ err = TRUE;
+ }
+ }
+ for(j = 0; j <= CR_DPMAX; j++){
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, j, MYPATHCHR, CR_LOBDIR);
+ if(lstat(path, &sbuf) != -1){
+ if(!crcplobdir(tcuria, path)) err = TRUE;
+ }
+ }
+ }
+ if(!crclose(tcuria)) err = TRUE;
+ if(!crremove(name)) err = TRUE;
+ sprintf(path, "%s%s", name, CR_TMPFSUF);
+ if(rename(path, name) == -1){
+ if(!err) dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ } else {
+ err = TRUE;
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Dump all records as endian independent data. */
+int crexportdb(CURIA *curia, const char *name){
+ char path[CR_PATHBUFSIZ], *kbuf, *vbuf, *pbuf;
+ int i, err, *fds, ksiz, vsiz, psiz;
+ assert(curia && name);
+ if(!(criterinit(curia))) return FALSE;
+ if(mkdir(name, CR_DIRMODE) == -1 && errno != EEXIST){
+ dpecodeset(DP_EMKDIR, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ fds = malloc(sizeof(int) * curia->dnum);
+ for(i = 0; i < curia->dnum; i++){
+ sprintf(path, "%s%c%04d", name, MYPATHCHR, i + 1);
+ if((fds[i] = open(path, O_RDWR | O_CREAT | O_TRUNC, CR_FILEMODE)) == -1){
+ if(!err) dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ }
+ while(!err && (kbuf = criternext(curia, &ksiz)) != NULL){
+ if((vbuf = crget(curia, kbuf, ksiz, 0, -1, &vsiz)) != NULL){
+ if((pbuf = malloc(ksiz + vsiz + CR_NUMBUFSIZ * 2)) != NULL){
+ psiz = 0;
+ psiz += sprintf(pbuf + psiz, "%X\n%X\n", ksiz, vsiz);
+ memcpy(pbuf + psiz, kbuf, ksiz);
+ psiz += ksiz;
+ pbuf[psiz++] = '\n';
+ memcpy(pbuf + psiz, vbuf, vsiz);
+ psiz += vsiz;
+ pbuf[psiz++] = '\n';
+ if(!crwrite(fds[curia->inum], pbuf, psiz)){
+ dpecodeset(DP_EWRITE, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(pbuf);
+ } else {
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ }
+ for(i = 0; i < curia->dnum; i++){
+ if(fds[i] != -1 && close(fds[i]) == -1){
+ if(!err) dpecodeset(DP_ECLOSE, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ }
+ free(fds);
+ return !err && !crfatalerror(curia);
+}
+
+
+/* Load all records from endian independent data. */
+int crimportdb(CURIA *curia, const char *name){
+ DEPOT *depot;
+ char ipath[CR_PATHBUFSIZ], opath[CR_PATHBUFSIZ], *kbuf, *vbuf;
+ int i, err, ksiz, vsiz;
+ struct stat sbuf;
+ assert(curia && name);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(crrnum(curia) > 0){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ for(i = 0; !err && i < CR_DPMAX; i++){
+ sprintf(ipath, "%s%c%04d", name, MYPATHCHR, i + 1);
+ if(lstat(ipath, &sbuf) == -1 || !S_ISREG(sbuf.st_mode)) break;
+ sprintf(opath, "%s%c%04d%s", curia->name, MYPATHCHR, i + 1, CR_TMPFSUF);
+ if((depot = dpopen(opath, DP_OWRITER | DP_OCREAT | DP_OTRUNC, -1)) != NULL){
+ if(dpimportdb(depot, ipath)){
+ dpiterinit(depot);
+ while((kbuf = dpiternext(depot, &ksiz)) != NULL){
+ if((vbuf = dpget(depot, kbuf, ksiz, 0, -1, &vsiz)) != NULL){
+ if(!crput(curia, kbuf, ksiz, vbuf, vsiz, CR_DKEEP)) err = TRUE;
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ }
+ } else {
+ err = TRUE;
+ }
+ if(!dpclose(depot)) err = TRUE;
+ if(!dpremove(opath)) err = TRUE;
+ } else {
+ err = TRUE;
+ }
+ }
+ return !err && !crfatalerror(curia);
+}
+
+
+/* Retrieve a record directly from a database directory. */
+char *crsnaffle(const char *name, const char *kbuf, int ksiz, int *sp){
+ char path[CR_PATHBUFSIZ], *vbuf;
+ int dnum, vsiz, tnum;
+ assert(name && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ sprintf(path, "%s%c%s", name, MYPATHCHR, CR_DPNAME);
+ if(!(vbuf = dpsnaffle(path, CR_KEYDNUM, -1, &vsiz)) || vsiz != sizeof(int) ||
+ (dnum = *(int *)vbuf) < 1){
+ free(vbuf);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ free(vbuf);
+ tnum = dpouterhash(kbuf, ksiz) % dnum;
+ sprintf(path, "%s%c%04d%c%s", name, MYPATHCHR, tnum + 1, MYPATHCHR, CR_DPNAME);
+ return dpsnaffle(path, kbuf, ksiz, sp);
+}
+
+
+/* Store a large object. */
+int crputlob(CURIA *curia, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode){
+ char *path;
+ int mode, fd, err, be;
+ struct stat sbuf;
+ assert(curia && kbuf && vbuf);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(vsiz < 0) vsiz = strlen(vbuf);
+ if(!(path = crgetlobpath(curia, kbuf, ksiz))) return FALSE;
+ if(!crmklobdir(path)){
+ free(path);
+ return FALSE;
+ }
+ be = lstat(path, &sbuf) != -1 && S_ISREG(sbuf.st_mode);
+ mode = O_RDWR | O_CREAT;
+ if(dmode & CR_DKEEP) mode |= O_EXCL;
+ if(dmode & CR_DCAT){
+ mode |= O_APPEND;
+ } else {
+ mode |= O_TRUNC;
+ }
+ if((fd = open(path, mode, CR_FILEMODE)) == -1){
+ free(path);
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ if(dmode == CR_DKEEP) dpecodeset(DP_EKEEP, __FILE__, __LINE__);
+ return FALSE;
+ }
+ free(path);
+ err = FALSE;
+ if(crwrite(fd, vbuf, vsiz) == -1){
+ err = TRUE;
+ dpecodeset(DP_EWRITE, __FILE__, __LINE__);
+ }
+ if(close(fd) == -1){
+ err = TRUE;
+ dpecodeset(DP_ECLOSE, __FILE__, __LINE__);
+ }
+ if(!err && !be) (curia->lrnum)++;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Delete a large object. */
+int croutlob(CURIA *curia, const char *kbuf, int ksiz){
+ char *path;
+ int err, be;
+ struct stat sbuf;
+ assert(curia && kbuf);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(!(path = crgetlobpath(curia, kbuf, ksiz))) return FALSE;
+ be = lstat(path, &sbuf) != -1 && S_ISREG(sbuf.st_mode);
+ err = FALSE;
+ if(unlink(path) == -1){
+ err = TRUE;
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ }
+ free(path);
+ if(!err && be) (curia->lrnum)--;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Retrieve a large object. */
+char *crgetlob(CURIA *curia, const char *kbuf, int ksiz, int start, int max, int *sp){
+ char *path, *buf;
+ struct stat sbuf;
+ int fd, size;
+ assert(curia && kbuf && start >= 0);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(!(path = crgetlobpath(curia, kbuf, ksiz))) return NULL;
+ if((fd = open(path, O_RDONLY, CR_FILEMODE)) == -1){
+ free(path);
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ free(path);
+ if(fstat(fd, &sbuf) == -1){
+ close(fd);
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(start > sbuf.st_size){
+ close(fd);
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(lseek(fd, start, SEEK_SET) == -1){
+ close(fd);
+ dpecodeset(DP_ESEEK, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(max < 0) max = sbuf.st_size;
+ if(!(buf = malloc(max + 1))){
+ close(fd);
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ size = crread(fd, buf, max);
+ close(fd);
+ if(size == -1){
+ free(buf);
+ dpecodeset(DP_EREAD, __FILE__, __LINE__);
+ return NULL;
+ }
+ buf[size] = '\0';
+ if(sp) *sp = size;
+ return buf;
+}
+
+
+/* Get the file descriptor of a large object. */
+int crgetlobfd(CURIA *curia, const char *kbuf, int ksiz){
+ char *path;
+ int fd;
+ assert(curia && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(!(path = crgetlobpath(curia, kbuf, ksiz))) return -1;
+ if((fd = open(path, curia->wmode ? O_RDWR: O_RDONLY, CR_FILEMODE)) == -1){
+ free(path);
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return -1;
+ }
+ free(path);
+ return fd;
+}
+
+
+/* Get the size of the value of a large object. */
+int crvsizlob(CURIA *curia, const char *kbuf, int ksiz){
+ char *path;
+ struct stat sbuf;
+ assert(curia && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(!(path = crgetlobpath(curia, kbuf, ksiz))) return -1;
+ if(lstat(path, &sbuf) == -1){
+ free(path);
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return -1;
+ }
+ free(path);
+ return sbuf.st_size;
+}
+
+
+/* Get the number of the large objects stored in a database. */
+int crrnumlob(CURIA *curia){
+ assert(curia);
+ return curia->lrnum;
+}
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Synchronize updating contents on memory. */
+int crmemsync(CURIA *curia){
+ int i, err;
+ assert(curia);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ if(!dpput(curia->attr, CR_KEYLRNUM, -1, (char *)&(curia->lrnum), sizeof(int), DP_DOVER) ||
+ !dpmemsync(curia->attr)) err = TRUE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpmemsync(curia->depots[i])){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Synchronize updating contents on memory, not physically. */
+int crmemflush(CURIA *curia){
+ int i, err;
+ assert(curia);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ if(!dpput(curia->attr, CR_KEYLRNUM, -1, (char *)&(curia->lrnum), sizeof(int), DP_DOVER) ||
+ !dpmemsync(curia->attr)) err = TRUE;
+ for(i = 0; i < curia->dnum; i++){
+ if(!dpmemflush(curia->depots[i])){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Get flags of a database. */
+int crgetflags(CURIA *curia){
+ assert(curia);
+ return dpgetflags(curia->attr);
+}
+
+
+/* Set flags of a database. */
+int crsetflags(CURIA *curia, int flags){
+ assert(curia);
+ if(!curia->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return dpsetflags(curia->attr, flags);
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Get a copied string.
+ `str' specifies an original string.
+ The return value is a copied string whose region is allocated by `malloc'. */
+static char *crstrdup(const char *str){
+ int len;
+ char *buf;
+ assert(str);
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ memcpy(buf, str, len + 1);
+ return buf;
+}
+
+
+/* Get an integer from a database.
+ `depot' specifies an inner database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the key.
+ The return value is the integer of the corresponding record. */
+static int crdpgetnum(DEPOT *depot, const char *kbuf, int ksiz){
+ char *vbuf;
+ int vsiz, rv;
+ if(!(vbuf = dpget(depot, kbuf, ksiz, 0, -1, &vsiz)) || vsiz != sizeof(int)){
+ free(vbuf);
+ return INT_MIN;
+ }
+ rv = *(int *)vbuf;
+ free(vbuf);
+ return rv;
+}
+
+
+/* Get the path of a large object.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the key.
+ The return value is a path string whose region is allocated by `malloc'. */
+static char *crgetlobpath(CURIA *curia, const char *kbuf, int ksiz){
+ char prefix[CR_PATHBUFSIZ], *wp, *path;
+ int i, hash;
+ assert(curia && kbuf && ksiz >= 0);
+ wp = prefix;
+ wp += sprintf(wp, "%s%c%04d%c%s%c",
+ curia->name, MYPATHCHR, dpouterhash(kbuf, ksiz) % curia->dnum + 1,
+ MYPATHCHR, CR_LOBDIR, MYPATHCHR);
+ hash = dpinnerhash(kbuf, ksiz);
+ for(i = 0; i < CR_LOBDDEPTH; i++){
+ wp += sprintf(wp, "%02X%c", hash % 0x100, MYPATHCHR);
+ hash /= 0x100;
+ }
+ if(!(path = malloc(strlen(prefix) + ksiz * 2 + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ wp = path;
+ wp += sprintf(path, "%s", prefix);
+ for(i = 0; i < ksiz; i++){
+ wp += sprintf(wp, "%02X", ((unsigned char *)kbuf)[i]);
+ }
+ return path;
+}
+
+
+/* Create directories included in a path.
+ `path' specifies a path.
+ The return value is true if successful, else, it is false. */
+static int crmklobdir(const char *path){
+ char elem[CR_PATHBUFSIZ], *wp;
+ const char *dp;
+ int err, len;
+ wp = elem;
+ err = FALSE;
+ while(*path != '\0' && (dp = strchr(path, MYPATHCHR)) != NULL){
+ len = dp - path;
+ if((wp != elem || dp == path)) wp += sprintf(wp, "%c", MYPATHCHR);
+ memcpy(wp, path, len);
+ wp[len] = '\0';
+ wp += len;
+ if(mkdir(elem, CR_DIRMODE) == -1 && errno != EEXIST) err = TRUE;
+ path = dp + 1;
+ }
+ if(err) dpecodeset(DP_EMKDIR, __FILE__, __LINE__);
+ return err ? FALSE : TRUE;
+}
+
+
+/* Remove file and directories under a directory.
+ `path' specifies a path.
+ The return value is true if successful, else, it is false. */
+static int crrmlobdir(const char *path){
+ char elem[CR_PATHBUFSIZ];
+ DIR *DD;
+ struct dirent *dp;
+ assert(path);
+ if(unlink(path) != -1){
+ return TRUE;
+ } else {
+ if(errno == ENOENT) return TRUE;
+ if(!(DD = opendir(path))){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ while((dp = readdir(DD)) != NULL){
+ if(!strcmp(dp->d_name, MYCDIRSTR) || !strcmp(dp->d_name, MYPDIRSTR)) continue;
+ sprintf(elem, "%s%c%s", path, MYPATHCHR, dp->d_name);
+ if(!crrmlobdir(elem)){
+ closedir(DD);
+ return FALSE;
+ }
+ }
+ }
+ if(closedir(DD) == -1){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(rmdir(path) == -1){
+ dpecodeset(DP_ERMDIR, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Copy file and directories under a directory for repairing.
+ `path' specifies a path.
+ The return value is true if successful, else, it is false. */
+static int crcplobdir(CURIA *curia, const char *path){
+ char elem[CR_PATHBUFSIZ], numbuf[3], *rp, *kbuf, *vbuf;
+ DIR *DD;
+ struct dirent *dp;
+ struct stat sbuf;
+ int i, ksiz, vsiz, fd;
+ assert(curia && path);
+ if(lstat(path, &sbuf) == -1){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(S_ISREG(sbuf.st_mode)){
+ rp = strrchr(path, MYPATHCHR) + 1;
+ for(i = 0; rp[i] != '\0'; i += 2){
+ numbuf[0] = rp[i];
+ numbuf[1] = rp[i+1];
+ numbuf[2] = '\0';
+ elem[i/2] = (int)strtol(numbuf, NULL, 16);
+ }
+ kbuf = elem;
+ ksiz = i / 2;
+ vsiz = sbuf.st_size;
+ if(!(vbuf = malloc(vsiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if((fd = open(path, O_RDONLY, CR_FILEMODE)) == -1){
+ free(vbuf);
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(crread(fd, vbuf, vsiz) == -1){
+ close(fd);
+ free(vbuf);
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!crputlob(curia, kbuf, ksiz, vbuf, vsiz, DP_DOVER)){
+ close(fd);
+ free(vbuf);
+ return FALSE;
+ }
+ close(fd);
+ free(vbuf);
+ return TRUE;
+ }
+ if(!(DD = opendir(path))){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ while((dp = readdir(DD)) != NULL){
+ if(!strcmp(dp->d_name, MYCDIRSTR) || !strcmp(dp->d_name, MYPDIRSTR)) continue;
+ sprintf(elem, "%s%c%s", path, MYPATHCHR, dp->d_name);
+ if(!crcplobdir(curia, elem)){
+ closedir(DD);
+ return FALSE;
+ }
+ }
+ if(closedir(DD) == -1){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Write into a file.
+ `fd' specifies a file descriptor.
+ `buf' specifies a buffer to write.
+ `size' specifies the size of the buffer.
+ The return value is the size of the written buffer, or, -1 on failure. */
+static int crwrite(int fd, const void *buf, int size){
+ char *lbuf;
+ int rv, wb;
+ assert(fd >= 0 && buf && size >= 0);
+ lbuf = (char *)buf;
+ rv = 0;
+ do {
+ wb = write(fd, lbuf, size);
+ switch(wb){
+ case -1: if(errno != EINTR) return -1;
+ case 0: break;
+ default:
+ lbuf += wb;
+ size -= wb;
+ rv += wb;
+ break;
+ }
+ } while(size > 0);
+ return rv;
+}
+
+
+/* Read from a file and store the data into a buffer.
+ `fd' specifies a file descriptor.
+ `buffer' specifies a buffer to store into.
+ `size' specifies the size to read with.
+ The return value is the size read with, or, -1 on failure. */
+static int crread(int fd, void *buf, int size){
+ char *lbuf;
+ int i, bs;
+ assert(fd >= 0 && buf && size >= 0);
+ lbuf = buf;
+ for(i = 0; i < size && (bs = read(fd, lbuf + i, size - i)) != 0; i += bs){
+ if(bs == -1 && errno != EINTR) return -1;
+ }
+ return i;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/curia.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/curia.h
===================================================================
--- box/trunk/qdbm/curia.h (rev 0)
+++ box/trunk/qdbm/curia.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,474 @@
+/*************************************************************************************************
+ * The extended API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _CURIA_H /* duplication check */
+#define _CURIA_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <depot.h>
+#include <stdlib.h>
+#include <time.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+typedef struct { /* type of structure for the database handle */
+ char *name; /* name of the database directory */
+ int wmode; /* whether to be writable */
+ int inode; /* inode of the database directory */
+ DEPOT *attr; /* database handle for attributes */
+ DEPOT **depots; /* handles of the record database */
+ int dnum; /* number of record database handles */
+ int inum; /* number of the database of the using iterator */
+ int lrnum; /* number of large objects */
+} CURIA;
+
+enum { /* enumeration for open modes */
+ CR_OREADER = 1 << 0, /* open as a reader */
+ CR_OWRITER = 1 << 1, /* open as a writer */
+ CR_OCREAT = 1 << 2, /* a writer creating */
+ CR_OTRUNC = 1 << 3, /* a writer truncating */
+ CR_ONOLCK = 1 << 4, /* open without locking */
+ CR_OLCKNB = 1 << 5, /* lock without blocking */
+ CR_OSPARSE = 1 << 6 /* create as sparse files */
+};
+
+enum { /* enumeration for write modes */
+ CR_DOVER, /* overwrite an existing value */
+ CR_DKEEP, /* keep an existing value */
+ CR_DCAT /* concatenate values */
+};
+
+
+/* Get a database handle.
+ `name' specifies the name of a database directory.
+ `omode' specifies the connection mode: `CR_OWRITER' as a writer, `CR_OREADER' as a reader.
+ If the mode is `CR_OWRITER', the following may be added by bitwise or: `CR_OCREAT', which
+ means it creates a new database if not exist, `CR_OTRUNC', which means it creates a new
+ database regardless if one exists. Both of `CR_OREADER' and `CR_OWRITER' can be added to by
+ bitwise or: `CR_ONOLCK', which means it opens a database directory without file locking, or
+ `CR_OLCKNB', which means locking is performed without blocking. `CR_OCREAT' can be added to
+ by bitwise or: `CR_OSPARSE', which means it creates database files as sparse files.
+ `bnum' specifies the number of elements of each bucket array. If it is not more than 0,
+ the default value is specified. The size of each bucket array is determined on creating,
+ and can not be changed except for by optimization of the database. Suggested size of each
+ bucket array is about from 0.5 to 4 times of the number of all records to store.
+ `dnum' specifies the number of division of the database. If it is not more than 0, the
+ default value is specified. The number of division can not be changed from the initial value.
+ The max number of division is 512.
+ The return value is the database handle or `NULL' if it is not successful.
+ While connecting as a writer, an exclusive lock is invoked to the database directory.
+ While connecting as a reader, a shared lock is invoked to the database directory.
+ The thread blocks until the lock is achieved. If `CR_ONOLCK' is used, the application is
+ responsible for exclusion control. */
+CURIA *cropen(const char *name, int omode, int bnum, int dnum);
+
+
+/* Close a database handle.
+ `curia' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ Because the region of a closed handle is released, it becomes impossible to use the handle.
+ Updating a database is assured to be written when the handle is closed. If a writer opens
+ a database but does not close it appropriately, the database will be broken. */
+int crclose(CURIA *curia);
+
+
+/* Store a record.
+ `curia' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ `dmode' specifies behavior when the key overlaps, by the following values: `CR_DOVER',
+ which means the specified value overwrites the existing one, `CR_DKEEP', which means the
+ existing value is kept, `CR_DCAT', which means the specified value is concatenated at the
+ end of the existing value.
+ If successful, the return value is true, else, it is false. */
+int crput(CURIA *curia, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode);
+
+
+/* Delete a record.
+ `curia' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is true, else, it is false. False is returned when no
+ record corresponds to the specified key. */
+int crout(CURIA *curia, const char *kbuf, int ksiz);
+
+
+/* Retrieve a record.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. If it is negative, the size to read is unlimited.
+ `sp' specifies the pointer to a variable to which the size of the region of the return value
+ is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the specified key or the size of the value of the corresponding record is less than `start'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. */
+char *crget(CURIA *curia, const char *kbuf, int ksiz, int start, int max, int *sp);
+
+
+/* Retrieve a record and write the value into a buffer.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. It shuld be equal to or less than the size of the
+ writing buffer.
+ `vbuf' specifies the pointer to a buffer into which the value of the corresponding record is
+ written.
+ If successful, the return value is the size of the written data, else, it is -1. -1 is
+ returned when no record corresponds to the specified key or the size of the value of the
+ corresponding record is less than `start'.
+ Note that no additional zero code is appended at the end of the region of the writing buffer. */
+int crgetwb(CURIA *curia, const char *kbuf, int ksiz, int start, int max, char *vbuf);
+
+
+/* Get the size of the value of a record.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is the size of the value of the corresponding record, else,
+ it is -1.
+ Because this function does not read the entity of a record, it is faster than `crget'. */
+int crvsiz(CURIA *curia, const char *kbuf, int ksiz);
+
+
+/* Initialize the iterator of a database handle.
+ `curia' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ The iterator is used in order to access the key of every record stored in a database. */
+int criterinit(CURIA *curia);
+
+
+/* Get the next key of the iterator.
+ `curia' specifies a database handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the next key, else, it is
+ `NULL'. `NULL' is returned when no record is to be get out of the iterator.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. It is possible to access every record by iteration of calling this
+ function. However, it is not assured if updating the database is occurred while the
+ iteration. Besides, the order of this traversal access method is arbitrary, so it is not
+ assured that the order of storing matches the one of the traversal access. */
+char *criternext(CURIA *curia, int *sp);
+
+
+/* Set alignment of a database handle.
+ `curia' specifies a database handle connected as a writer.
+ `align' specifies the size of alignment.
+ If successful, the return value is true, else, it is false.
+ If alignment is set to a database, the efficiency of overwriting values is improved.
+ The size of alignment is suggested to be average size of the values of the records to be
+ stored. If alignment is positive, padding whose size is multiple number of the alignment
+ is placed. If alignment is negative, as `vsiz' is the size of a value, the size of padding
+ is calculated with `(vsiz / pow(2, abs(align) - 1))'. Because alignment setting is not
+ saved in a database, you should specify alignment every opening a database. */
+int crsetalign(CURIA *curia, int align);
+
+
+/* Set the size of the free block pool of a database handle.
+ `curia' specifies a database handle connected as a writer.
+ `size' specifies the size of the free block pool of a database.
+ If successful, the return value is true, else, it is false.
+ The default size of the free block pool is 16. If the size is greater, the space efficiency
+ of overwriting values is improved with the time efficiency sacrificed. */
+int crsetfbpsiz(CURIA *curia, int size);
+
+
+/* Synchronize updating contents with the files and the devices.
+ `curia' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ This function is useful when another process uses the connected database directory. */
+int crsync(CURIA *curia);
+
+
+/* Optimize a database.
+ `curia' specifies a database handle connected as a writer.
+ `bnum' specifies the number of the elements of each bucket array. If it is not more than 0,
+ the default value is specified.
+ If successful, the return value is true, else, it is false.
+ In an alternating succession of deleting and storing with overwrite or concatenate,
+ dispensable regions accumulate. This function is useful to do away with them. */
+int croptimize(CURIA *curia, int bnum);
+
+/* Get the name of a database.
+ `curia' specifies a database handle.
+ If successful, the return value is the pointer to the region of the name of the database,
+ else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *crname(CURIA *curia);
+
+
+/* Get the total size of database files.
+ `curia' specifies a database handle.
+ If successful, the return value is the total size of the database files, else, it is -1.
+ If the total size is more than 2GB, the return value overflows. */
+int crfsiz(CURIA *curia);
+
+
+/* Get the total size of database files as double-precision floating-point number.
+ `curia' specifies a database handle.
+ If successful, the return value is the total size of the database files, else, it is -1.0. */
+double crfsizd(CURIA *curia);
+
+
+/* Get the total number of the elements of each bucket array.
+ `curia' specifies a database handle.
+ If successful, the return value is the total number of the elements of each bucket array,
+ else, it is -1. */
+int crbnum(CURIA *curia);
+
+
+/* Get the total number of the used elements of each bucket array.
+ `curia' specifies a database handle.
+ If successful, the return value is the total number of the used elements of each bucket
+ array, else, it is -1.
+ This function is inefficient because it accesses all elements of each bucket array. */
+int crbusenum(CURIA *curia);
+
+
+/* Get the number of the records stored in a database.
+ `curia' specifies a database handle.
+ If successful, the return value is the number of the records stored in the database, else,
+ it is -1. */
+int crrnum(CURIA *curia);
+
+
+/* Check whether a database handle is a writer or not.
+ `curia' specifies a database handle.
+ The return value is true if the handle is a writer, false if not. */
+int crwritable(CURIA *curia);
+
+
+/* Check whether a database has a fatal error or not.
+ `curia' specifies a database handle.
+ The return value is true if the database has a fatal error, false if not. */
+int crfatalerror(CURIA *curia);
+
+
+/* Get the inode number of a database directory.
+ `curia' specifies a database handle.
+ The return value is the inode number of the database directory. */
+int crinode(CURIA *curia);
+
+
+/* Get the last modified time of a database.
+ `curia' specifies a database handle.
+ The return value is the last modified time of the database. */
+time_t crmtime(CURIA *curia);
+
+
+/* Remove a database directory.
+ `name' specifies the name of a database directory.
+ If successful, the return value is true, else, it is false. */
+int crremove(const char *name);
+
+
+/* Repair a broken database directory.
+ `name' specifies the name of a database directory.
+ If successful, the return value is true, else, it is false.
+ There is no guarantee that all records in a repaired database directory correspond to the
+ original or expected state. */
+int crrepair(const char *name);
+
+
+/* Dump all records as endian independent data.
+ `curia' specifies a database handle.
+ `name' specifies the name of an output directory.
+ If successful, the return value is true, else, it is false.
+ Note that large objects are ignored. */
+int crexportdb(CURIA *curia, const char *name);
+
+
+/* Load all records from endian independent data.
+ `curia' specifies a database handle connected as a writer. The database of the handle must
+ be empty.
+ `name' specifies the name of an input directory.
+ If successful, the return value is true, else, it is false.
+ Note that large objects are ignored. */
+int crimportdb(CURIA *curia, const char *name);
+
+
+/* Retrieve a record directly from a database directory.
+ `name' specifies the name of a database directory.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the specified key.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. Although this function can be used even while the database directory is
+ locked by another process, it is not assured that recent updated is reflected. */
+char *crsnaffle(const char *name, const char *kbuf, int ksiz, int *sp);
+
+
+/* Store a large object.
+ `curia' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ `dmode' specifies behavior when the key overlaps, by the following values: `CR_DOVER',
+ which means the specified value overwrites the existing one, `CR_DKEEP', which means the
+ existing value is kept, `CR_DCAT', which means the specified value is concatenated at the
+ end of the existing value.
+ If successful, the return value is true, else, it is false. */
+int crputlob(CURIA *curia, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode);
+
+
+/* Delete a large object.
+ `curia' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is true, else, it is false. false is returned when no large
+ object corresponds to the specified key. */
+int croutlob(CURIA *curia, const char *kbuf, int ksiz);
+
+
+/* Retrieve a large object.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. If it is negative, the size to read is unlimited.
+ `sp' specifies the pointer to a variable to which the size of the region of the return value
+ is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding large object, else, it is `NULL'. `NULL' is returned when no large object
+ corresponds to the specified key or the size of the value of the corresponding large object
+ is less than `start'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. */
+char *crgetlob(CURIA *curia, const char *kbuf, int ksiz, int start, int max, int *sp);
+
+
+/* Get the file descriptor of a large object.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is the file descriptor of the corresponding large object,
+ else, it is -1. -1 is returned when no large object corresponds to the specified key. The
+ returned file descriptor is opened with the `open' call. If the database handle was opened
+ as a writer, the descriptor is writable (O_RDWR), else, it is not writable (O_RDONLY). The
+ descriptor should be closed with the `close' call if it is no longer in use. */
+int crgetlobfd(CURIA *curia, const char *kbuf, int ksiz);
+
+
+/* Get the size of the value of a large object.
+ `curia' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is the size of the value of the corresponding large object,
+ else, it is -1.
+ Because this function does not read the entity of a large object, it is faster than
+ `crgetlob'. */
+int crvsizlob(CURIA *curia, const char *kbuf, int ksiz);
+
+
+/* Get the number of the large objects stored in a database.
+ `curia' specifies a database handle.
+ If successful, the return value is the number of the large objects stored in the database,
+ else, it is -1. */
+int crrnumlob(CURIA *curia);
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Synchronize updating contents on memory.
+ `curia' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. */
+int crmemsync(CURIA *curia);
+
+
+/* Synchronize updating contents on memory, not physically.
+ `curia' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. */
+int crmemflush(CURIA *curia);
+
+
+/* Get flags of a database.
+ `curia' specifies a database handle.
+ The return value is the flags of a database. */
+int crgetflags(CURIA *curia);
+
+
+/* Set flags of a database.
+ `curia' specifies a database handle connected as a writer.
+ `flags' specifies flags to set. Least ten bits are reserved for internal use.
+ If successful, the return value is true, else, it is false. */
+int crsetflags(CURIA *curia, int flags);
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/curia.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/depot.c
===================================================================
--- box/trunk/qdbm/depot.c (rev 0)
+++ box/trunk/qdbm/depot.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,2219 @@
+/*************************************************************************************************
+ * Implementation of Depot
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "depot.h"
+#include "myconf.h"
+
+#define DP_FILEMODE 00644 /* permission of a creating file */
+#define DP_MAGICNUMB "[DEPOT]\n\f" /* magic number on environments of big endian */
+#define DP_MAGICNUML "[depot]\n\f" /* magic number on environments of little endian */
+#define DP_HEADSIZ 48 /* size of the reagion of the header */
+#define DP_LIBVEROFF 12 /* offset of the region for the library version */
+#define DP_FLAGSOFF 16 /* offset of the region for flags */
+#define DP_FSIZOFF 24 /* offset of the region for the file size */
+#define DP_BNUMOFF 32 /* offset of the region for the bucket number */
+#define DP_RNUMOFF 40 /* offset of the region for the record number */
+#define DP_DEFBNUM 8191 /* default bucket number */
+#define DP_FBPOOLSIZ 16 /* size of free block pool */
+#define DP_ENTBUFSIZ 128 /* size of the entity buffer */
+#define DP_STKBUFSIZ 256 /* size of the stack key buffer */
+#define DP_WRTBUFSIZ 8192 /* size of the writing buffer */
+#define DP_FSBLKSIZ 4096 /* size of a block of the file system */
+#define DP_TMPFSUF MYEXTSTR "dptmp" /* suffix of a temporary file */
+#define DP_OPTBLOAD 0.25 /* ratio of bucket loading at optimization */
+#define DP_OPTRUNIT 256 /* number of records in a process of optimization */
+#define DP_NUMBUFSIZ 32 /* size of a buffer for a number */
+#define DP_IOBUFSIZ 8192 /* size of an I/O buffer */
+
+/* get the first hash value */
+#define DP_FIRSTHASH(DP_res, DP_kbuf, DP_ksiz) \
+ do { \
+ const unsigned char *_DP_p; \
+ int _DP_ksiz; \
+ _DP_p = (const unsigned char *)(DP_kbuf); \
+ _DP_ksiz = DP_ksiz; \
+ if((_DP_ksiz) == sizeof(int)){ \
+ memcpy(&(DP_res), (DP_kbuf), sizeof(int)); \
+ } else { \
+ (DP_res) = 751; \
+ } \
+ while(_DP_ksiz--){ \
+ (DP_res) = (DP_res) * 31 + *(_DP_p)++; \
+ } \
+ (DP_res) = ((DP_res) * 87767623) & INT_MAX; \
+ } while(FALSE)
+
+/* get the second hash value */
+#define DP_SECONDHASH(DP_res, DP_kbuf, DP_ksiz) \
+ do { \
+ const unsigned char *_DP_p; \
+ int _DP_ksiz; \
+ _DP_p = (const unsigned char *)(DP_kbuf) + DP_ksiz - 1; \
+ _DP_ksiz = DP_ksiz; \
+ for((DP_res) = 19780211; _DP_ksiz--;){ \
+ (DP_res) = (DP_res) * 37 + *(_DP_p)--; \
+ } \
+ (DP_res) = ((DP_res) * 43321879) & INT_MAX; \
+ } while(FALSE)
+
+/* get the third hash value */
+#define DP_THIRDHASH(DP_res, DP_kbuf, DP_ksiz) \
+ do { \
+ int _DP_i; \
+ (DP_res) = 774831917; \
+ for(_DP_i = (DP_ksiz) - 1; _DP_i >= 0; _DP_i--){ \
+ (DP_res) = (DP_res) * 29 + ((const unsigned char *)(DP_kbuf))[_DP_i]; \
+ } \
+ (DP_res) = ((DP_res) * 5157883) & INT_MAX; \
+ } while(FALSE)
+
+enum { /* enumeration for a record header */
+ DP_RHIFLAGS, /* offset of flags */
+ DP_RHIHASH, /* offset of value of the second hash function */
+ DP_RHIKSIZ, /* offset of the size of the key */
+ DP_RHIVSIZ, /* offset of the size of the value */
+ DP_RHIPSIZ, /* offset of the size of the padding bytes */
+ DP_RHILEFT, /* offset of the offset of the left child */
+ DP_RHIRIGHT, /* offset of the offset of the right child */
+ DP_RHNUM /* number of elements of a header */
+};
+
+enum { /* enumeration for the flag of a record */
+ DP_RECFDEL = 1 << 0, /* deleted */
+ DP_RECFREUSE = 1 << 1 /* reusable */
+};
+
+
+/* private function prototypes */
+static int dpbigendian(void);
+static char *dpstrdup(const char *str);
+static int dplock(int fd, int ex, int nb);
+static int dpwrite(int fd, const void *buf, int size);
+static int dpseekwrite(int fd, int off, const void *buf, int size);
+static int dpseekwritenum(int fd, int off, int num);
+static int dpread(int fd, void *buf, int size);
+static int dpseekread(int fd, int off, void *buf, int size);
+static int dpfcopy(int destfd, int destoff, int srcfd, int srcoff);
+static int dpgetprime(int num);
+static int dppadsize(DEPOT *depot, int ksiz, int vsiz);
+static int dprecsize(int *head);
+static int dprechead(DEPOT *depot, int off, int *head, char *ebuf, int *eep);
+static char *dpreckey(DEPOT *depot, int off, int *head);
+static char *dprecval(DEPOT *depot, int off, int *head, int start, int max);
+static int dprecvalwb(DEPOT *depot, int off, int *head, int start, int max, char *vbuf);
+static int dpkeycmp(const char *abuf, int asiz, const char *bbuf, int bsiz);
+static int dprecsearch(DEPOT *depot, const char *kbuf, int ksiz, int hash, int *bip, int *offp,
+ int *entp, int *head, char *ebuf, int *eep, int delhit);
+static int dprecrewrite(DEPOT *depot, int off, int rsiz, const char *kbuf, int ksiz,
+ const char *vbuf, int vsiz, int hash, int left, int right);
+static int dprecappend(DEPOT *depot, const char *kbuf, int ksiz, const char *vbuf, int vsiz,
+ int hash, int left, int right);
+static int dprecover(DEPOT *depot, int off, int *head, const char *vbuf, int vsiz, int cat);
+static int dprecdelete(DEPOT *depot, int off, int *head, int reusable);
+static void dpfbpoolcoal(DEPOT *depot);
+static int dpfbpoolcmp(const void *a, const void *b);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* String containing the version information. */
+const char *dpversion = _QDBM_VERSION;
+
+
+/* Get a message string corresponding to an error code. */
+const char *dperrmsg(int ecode){
+ switch(ecode){
+ case DP_ENOERR: return "no error";
+ case DP_EFATAL: return "with fatal error";
+ case DP_EMODE: return "invalid mode";
+ case DP_EBROKEN: return "broken database file";
+ case DP_EKEEP: return "existing record";
+ case DP_ENOITEM: return "no item found";
+ case DP_EALLOC: return "memory allocation error";
+ case DP_EMAP: return "memory mapping error";
+ case DP_EOPEN: return "open error";
+ case DP_ECLOSE: return "close error";
+ case DP_ETRUNC: return "trunc error";
+ case DP_ESYNC: return "sync error";
+ case DP_ESTAT: return "stat error";
+ case DP_ESEEK: return "seek error";
+ case DP_EREAD: return "read error";
+ case DP_EWRITE: return "write error";
+ case DP_ELOCK: return "lock error";
+ case DP_EUNLINK: return "unlink error";
+ case DP_EMKDIR: return "mkdir error";
+ case DP_ERMDIR: return "rmdir error";
+ case DP_EMISC: return "miscellaneous error";
+ }
+ return "(invalid ecode)";
+}
+
+
+/* Get a database handle. */
+DEPOT *dpopen(const char *name, int omode, int bnum){
+ char hbuf[DP_HEADSIZ], *map, c, *tname;
+ int i, mode, fd, inode, fsiz, rnum, msiz, *fbpool;
+ struct stat sbuf;
+ time_t mtime;
+ DEPOT *depot;
+ assert(name);
+ mode = O_RDONLY;
+ if(omode & DP_OWRITER){
+ mode = O_RDWR;
+ if(omode & DP_OCREAT) mode |= O_CREAT;
+ }
+ if((fd = open(name, mode, DP_FILEMODE)) == -1){
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!(omode & DP_ONOLCK)){
+ if(!dplock(fd, omode & DP_OWRITER, omode & DP_OLCKNB)){
+ close(fd);
+ return NULL;
+ }
+ }
+ if((omode & DP_OWRITER) && (omode & DP_OTRUNC)){
+ if(ftruncate(fd, 0) == -1){
+ close(fd);
+ dpecodeset(DP_ETRUNC, __FILE__, __LINE__);
+ return NULL;
+ }
+ }
+ if(fstat(fd, &sbuf) == -1 || !S_ISREG(sbuf.st_mode) ||
+ (sbuf.st_ino == 0 && lstat(name, &sbuf) == -1)){
+ close(fd);
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return NULL;
+ }
+ inode = sbuf.st_ino;
+ mtime = sbuf.st_mtime;
+ fsiz = sbuf.st_size;
+ if((omode & DP_OWRITER) && fsiz == 0){
+ memset(hbuf, 0, DP_HEADSIZ);
+ if(dpbigendian()){
+ memcpy(hbuf, DP_MAGICNUMB, strlen(DP_MAGICNUMB));
+ } else {
+ memcpy(hbuf, DP_MAGICNUML, strlen(DP_MAGICNUML));
+ }
+ sprintf(hbuf + DP_LIBVEROFF, "%d", _QDBM_LIBVER / 100);
+ bnum = bnum < 1 ? DP_DEFBNUM : bnum;
+ bnum = dpgetprime(bnum);
+ memcpy(hbuf + DP_BNUMOFF, &bnum, sizeof(int));
+ rnum = 0;
+ memcpy(hbuf + DP_RNUMOFF, &rnum, sizeof(int));
+ fsiz = DP_HEADSIZ + bnum * sizeof(int);
+ memcpy(hbuf + DP_FSIZOFF, &fsiz, sizeof(int));
+ if(!dpseekwrite(fd, 0, hbuf, DP_HEADSIZ)){
+ close(fd);
+ return NULL;
+ }
+ if(omode & DP_OSPARSE){
+ c = 0;
+ if(!dpseekwrite(fd, fsiz - 1, &c, 1)){
+ close(fd);
+ return NULL;
+ }
+ } else {
+ if(!(map = malloc(bnum * sizeof(int)))){
+ close(fd);
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ memset(map, 0, bnum * sizeof(int));
+ if(!dpseekwrite(fd, DP_HEADSIZ, map, bnum * sizeof(int))){
+ free(map);
+ close(fd);
+ return NULL;
+ }
+ free(map);
+ }
+ }
+ if(!dpseekread(fd, 0, hbuf, DP_HEADSIZ)){
+ close(fd);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!(omode & DP_ONOLCK) &&
+ ((dpbigendian() ? memcmp(hbuf, DP_MAGICNUMB, strlen(DP_MAGICNUMB)) != 0 :
+ memcmp(hbuf, DP_MAGICNUML, strlen(DP_MAGICNUML)) != 0) ||
+ *((int *)(hbuf + DP_FSIZOFF)) != fsiz)){
+ close(fd);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ bnum = *((int *)(hbuf + DP_BNUMOFF));
+ rnum = *((int *)(hbuf + DP_RNUMOFF));
+ if(bnum < 1 || rnum < 0 || fsiz < DP_HEADSIZ + bnum * sizeof(int)){
+ close(fd);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ msiz = DP_HEADSIZ + bnum * sizeof(int);
+ map = mmap(0, msiz, PROT_READ | ((mode & DP_OWRITER) ? PROT_WRITE : 0), MAP_SHARED, fd, 0);
+ if(map == MAP_FAILED){
+ close(fd);
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ return NULL;
+ }
+ tname = NULL;
+ fbpool = NULL;
+ if(!(depot = malloc(sizeof(DEPOT))) || !(tname = dpstrdup(name)) ||
+ !(fbpool = malloc(DP_FBPOOLSIZ * 2 * sizeof(int)))){
+ free(fbpool);
+ free(tname);
+ free(depot);
+ munmap(map, msiz);
+ close(fd);
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ depot->name = tname;
+ depot->wmode = (mode & DP_OWRITER);
+ depot->inode = inode;
+ depot->mtime = mtime;
+ depot->fd = fd;
+ depot->fsiz = fsiz;
+ depot->map = map;
+ depot->msiz = msiz;
+ depot->buckets = (int *)(map + DP_HEADSIZ);
+ depot->bnum = bnum;
+ depot->rnum = rnum;
+ depot->fatal = FALSE;
+ depot->ioff = 0;
+ depot->fbpool = fbpool;
+ for(i = 0; i < DP_FBPOOLSIZ * 2; i += 2){
+ depot->fbpool[i] = -1;
+ depot->fbpool[i+1] = -1;
+ }
+ depot->fbpsiz = DP_FBPOOLSIZ * 2;
+ depot->fbpinc = 0;
+ depot->align = 0;
+ return depot;
+}
+
+
+/* Close a database handle. */
+int dpclose(DEPOT *depot){
+ int fatal, err;
+ assert(depot);
+ fatal = depot->fatal;
+ err = FALSE;
+ if(depot->wmode){
+ *((int *)(depot->map + DP_FSIZOFF)) = depot->fsiz;
+ *((int *)(depot->map + DP_RNUMOFF)) = depot->rnum;
+ }
+ if(depot->map != MAP_FAILED){
+ if(munmap(depot->map, depot->msiz) == -1){
+ err = TRUE;
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ }
+ }
+ if(close(depot->fd) == -1){
+ err = TRUE;
+ dpecodeset(DP_ECLOSE, __FILE__, __LINE__);
+ }
+ free(depot->fbpool);
+ free(depot->name);
+ free(depot);
+ if(fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Store a record. */
+int dpput(DEPOT *depot, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode){
+ int head[DP_RHNUM], next[DP_RHNUM];
+ int i, hash, bi, off, entoff, ee, newoff, rsiz, nsiz, fdel, mroff, mrsiz, mi, min;
+ char ebuf[DP_ENTBUFSIZ], *tval, *swap;
+ assert(depot && kbuf && vbuf);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(vsiz < 0) vsiz = strlen(vbuf);
+ newoff = -1;
+ DP_SECONDHASH(hash, kbuf, ksiz);
+ switch(dprecsearch(depot, kbuf, ksiz, hash, &bi, &off, &entoff, head, ebuf, &ee, TRUE)){
+ case -1:
+ depot->fatal = TRUE;
+ return FALSE;
+ case 0:
+ fdel = head[DP_RHIFLAGS] & DP_RECFDEL;
+ if(dmode == DP_DKEEP && !fdel){
+ dpecodeset(DP_EKEEP, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(fdel){
+ head[DP_RHIPSIZ] += head[DP_RHIVSIZ];
+ head[DP_RHIVSIZ] = 0;
+ }
+ rsiz = dprecsize(head);
+ nsiz = DP_RHNUM * sizeof(int) + ksiz + vsiz;
+ if(dmode == DP_DCAT) nsiz += head[DP_RHIVSIZ];
+ if(off + rsiz >= depot->fsiz){
+ if(rsiz < nsiz){
+ head[DP_RHIPSIZ] += nsiz - rsiz;
+ rsiz = nsiz;
+ depot->fsiz = off + rsiz;
+ }
+ } else {
+ while(nsiz > rsiz && off + rsiz < depot->fsiz){
+ if(!dprechead(depot, off + rsiz, next, NULL, NULL)) return FALSE;
+ if(!(next[DP_RHIFLAGS] & DP_RECFREUSE)) break;
+ head[DP_RHIPSIZ] += dprecsize(next);
+ rsiz += dprecsize(next);
+ }
+ for(i = 0; i < depot->fbpsiz; i += 2){
+ if(depot->fbpool[i] >= off && depot->fbpool[i] < off + rsiz){
+ depot->fbpool[i] = -1;
+ depot->fbpool[i+1] = -1;
+ }
+ }
+ }
+ if(nsiz <= rsiz){
+ if(!dprecover(depot, off, head, vbuf, vsiz, dmode == DP_DCAT)){
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ } else {
+ tval = NULL;
+ if(dmode == DP_DCAT){
+ if(ee && DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + head[DP_RHIVSIZ] <= DP_ENTBUFSIZ){
+ if(!(tval = malloc(head[DP_RHIVSIZ] + vsiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ memcpy(tval, ebuf + (DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ]), head[DP_RHIVSIZ]);
+ } else {
+ if(!(tval = dprecval(depot, off, head, 0, -1))){
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(!(swap = realloc(tval, head[DP_RHIVSIZ] + vsiz + 1))){
+ free(tval);
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ tval = swap;
+ }
+ memcpy(tval + head[DP_RHIVSIZ], vbuf, vsiz);
+ vsiz += head[DP_RHIVSIZ];
+ vbuf = tval;
+ }
+ mi = -1;
+ min = -1;
+ for(i = 0; i < depot->fbpsiz; i += 2){
+ if(depot->fbpool[i+1] < nsiz) continue;
+ if(mi == -1 || depot->fbpool[i+1] < min){
+ mi = i;
+ min = depot->fbpool[i+1];
+ }
+ }
+ if(mi >= 0){
+ mroff = depot->fbpool[mi];
+ mrsiz = depot->fbpool[mi+1];
+ depot->fbpool[mi] = -1;
+ depot->fbpool[mi+1] = -1;
+ } else {
+ mroff = -1;
+ mrsiz = -1;
+ }
+ if(!dprecdelete(depot, off, head, TRUE)){
+ free(tval);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(mroff > 0 && nsiz <= mrsiz){
+ if(!dprecrewrite(depot, mroff, mrsiz, kbuf, ksiz, vbuf, vsiz,
+ hash, head[DP_RHILEFT], head[DP_RHIRIGHT])){
+ free(tval);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ newoff = mroff;
+ } else {
+ if((newoff = dprecappend(depot, kbuf, ksiz, vbuf, vsiz,
+ hash, head[DP_RHILEFT], head[DP_RHIRIGHT])) == -1){
+ free(tval);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ }
+ free(tval);
+ }
+ if(fdel) depot->rnum++;
+ break;
+ default:
+ if((newoff = dprecappend(depot, kbuf, ksiz, vbuf, vsiz, hash, 0, 0)) == -1){
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ depot->rnum++;
+ break;
+ }
+ if(newoff > 0){
+ if(entoff > 0){
+ if(!dpseekwritenum(depot->fd, entoff, newoff)){
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ } else {
+ depot->buckets[bi] = newoff;
+ }
+ }
+ return TRUE;
+}
+
+
+/* Delete a record. */
+int dpout(DEPOT *depot, const char *kbuf, int ksiz){
+ int head[DP_RHNUM], hash, bi, off, entoff, ee;
+ char ebuf[DP_ENTBUFSIZ];
+ assert(depot && kbuf);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ DP_SECONDHASH(hash, kbuf, ksiz);
+ switch(dprecsearch(depot, kbuf, ksiz, hash, &bi, &off, &entoff, head, ebuf, &ee, FALSE)){
+ case -1:
+ depot->fatal = TRUE;
+ return FALSE;
+ case 0:
+ break;
+ default:
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!dprecdelete(depot, off, head, FALSE)){
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ depot->rnum--;
+ return TRUE;
+}
+
+
+/* Retrieve a record. */
+char *dpget(DEPOT *depot, const char *kbuf, int ksiz, int start, int max, int *sp){
+ int head[DP_RHNUM], hash, bi, off, entoff, ee, vsiz;
+ char ebuf[DP_ENTBUFSIZ], *vbuf;
+ assert(depot && kbuf && start >= 0);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ DP_SECONDHASH(hash, kbuf, ksiz);
+ switch(dprecsearch(depot, kbuf, ksiz, hash, &bi, &off, &entoff, head, ebuf, &ee, FALSE)){
+ case -1:
+ depot->fatal = TRUE;
+ return NULL;
+ case 0:
+ break;
+ default:
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(start > head[DP_RHIVSIZ]){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(ee && DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + head[DP_RHIVSIZ] <= DP_ENTBUFSIZ){
+ head[DP_RHIVSIZ] -= start;
+ if(max < 0){
+ vsiz = head[DP_RHIVSIZ];
+ } else {
+ vsiz = max < head[DP_RHIVSIZ] ? max : head[DP_RHIVSIZ];
+ }
+ if(!(vbuf = malloc(vsiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return NULL;
+ }
+ memcpy(vbuf, ebuf + (DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + start), vsiz);
+ vbuf[vsiz] = '\0';
+ } else {
+ if(!(vbuf = dprecval(depot, off, head, start, max))){
+ depot->fatal = TRUE;
+ return NULL;
+ }
+ }
+ if(sp){
+ if(max < 0){
+ *sp = head[DP_RHIVSIZ];
+ } else {
+ *sp = max < head[DP_RHIVSIZ] ? max : head[DP_RHIVSIZ];
+ }
+ }
+ return vbuf;
+}
+
+
+/* Retrieve a record and write the value into a buffer. */
+int dpgetwb(DEPOT *depot, const char *kbuf, int ksiz, int start, int max, char *vbuf){
+ int head[DP_RHNUM], hash, bi, off, entoff, ee, vsiz;
+ char ebuf[DP_ENTBUFSIZ];
+ assert(depot && kbuf && start >= 0 && max >= 0 && vbuf);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ DP_SECONDHASH(hash, kbuf, ksiz);
+ switch(dprecsearch(depot, kbuf, ksiz, hash, &bi, &off, &entoff, head, ebuf, &ee, FALSE)){
+ case -1:
+ depot->fatal = TRUE;
+ return -1;
+ case 0:
+ break;
+ default:
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return -1;
+ }
+ if(start > head[DP_RHIVSIZ]){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return -1;
+ }
+ if(ee && DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + head[DP_RHIVSIZ] <= DP_ENTBUFSIZ){
+ head[DP_RHIVSIZ] -= start;
+ vsiz = max < head[DP_RHIVSIZ] ? max : head[DP_RHIVSIZ];
+ memcpy(vbuf, ebuf + (DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + start), vsiz);
+ } else {
+ if((vsiz = dprecvalwb(depot, off, head, start, max, vbuf)) == -1){
+ depot->fatal = TRUE;
+ return -1;
+ }
+ }
+ return vsiz;
+}
+
+
+/* Get the size of the value of a record. */
+int dpvsiz(DEPOT *depot, const char *kbuf, int ksiz){
+ int head[DP_RHNUM], hash, bi, off, entoff, ee;
+ char ebuf[DP_ENTBUFSIZ];
+ assert(depot && kbuf);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ DP_SECONDHASH(hash, kbuf, ksiz);
+ switch(dprecsearch(depot, kbuf, ksiz, hash, &bi, &off, &entoff, head, ebuf, &ee, FALSE)){
+ case -1:
+ depot->fatal = TRUE;
+ return -1;
+ case 0:
+ break;
+ default:
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return -1;
+ }
+ return head[DP_RHIVSIZ];
+}
+
+
+/* Initialize the iterator of a database handle. */
+int dpiterinit(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ depot->ioff = 0;
+ return TRUE;
+}
+
+
+/* Get the next key of the iterator. */
+char *dpiternext(DEPOT *depot, int *sp){
+ int off, head[DP_RHNUM], ee;
+ char ebuf[DP_ENTBUFSIZ], *kbuf;
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ off = DP_HEADSIZ + depot->bnum * sizeof(int);
+ off = off > depot->ioff ? off : depot->ioff;
+ while(off < depot->fsiz){
+ if(!dprechead(depot, off, head, ebuf, &ee)){
+ depot->fatal = TRUE;
+ return NULL;
+ }
+ if(head[DP_RHIFLAGS] & DP_RECFDEL){
+ off += dprecsize(head);
+ } else {
+ if(ee && DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] <= DP_ENTBUFSIZ){
+ if(!(kbuf = malloc(head[DP_RHIKSIZ] + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return NULL;
+ }
+ memcpy(kbuf, ebuf + (DP_RHNUM * sizeof(int)), head[DP_RHIKSIZ]);
+ kbuf[head[DP_RHIKSIZ]] = '\0';
+ } else {
+ if(!(kbuf = dpreckey(depot, off, head))){
+ depot->fatal = TRUE;
+ return NULL;
+ }
+ }
+ depot->ioff = off + dprecsize(head);
+ if(sp) *sp = head[DP_RHIKSIZ];
+ return kbuf;
+ }
+ }
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+}
+
+
+/* Set alignment of a database handle. */
+int dpsetalign(DEPOT *depot, int align){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ depot->align = align;
+ return TRUE;
+}
+
+
+/* Set the size of the free block pool of a database handle. */
+int dpsetfbpsiz(DEPOT *depot, int size){
+ int *fbpool;
+ int i;
+ assert(depot && size >= 0);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ size *= 2;
+ if(!(fbpool = realloc(depot->fbpool, size * sizeof(int) + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ for(i = 0; i < size; i += 2){
+ fbpool[i] = -1;
+ fbpool[i+1] = -1;
+ }
+ depot->fbpool = fbpool;
+ depot->fbpsiz = size;
+ return TRUE;
+}
+
+
+
+/* Synchronize contents of updating a database with the file and the device. */
+int dpsync(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ *((int *)(depot->map + DP_FSIZOFF)) = depot->fsiz;
+ *((int *)(depot->map + DP_RNUMOFF)) = depot->rnum;
+ if(msync(depot->map, depot->msiz, MS_SYNC) == -1){
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(fsync(depot->fd) == -1){
+ dpecodeset(DP_ESYNC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Optimize a database. */
+int dpoptimize(DEPOT *depot, int bnum){
+ DEPOT *tdepot;
+ char *name;
+ int i, err, off, head[DP_RHNUM], ee, ksizs[DP_OPTRUNIT], vsizs[DP_OPTRUNIT], unum;
+ char ebuf[DP_ENTBUFSIZ], *kbufs[DP_OPTRUNIT], *vbufs[DP_OPTRUNIT];
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(name = malloc(strlen(depot->name) + strlen(DP_TMPFSUF) + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ depot->fatal = FALSE;
+ return FALSE;
+ }
+ sprintf(name, "%s%s", depot->name, DP_TMPFSUF);
+ if(bnum < 0){
+ bnum = (int)(depot->rnum * (1.0 / DP_OPTBLOAD)) + 1;
+ if(bnum < DP_DEFBNUM / 2) bnum = DP_DEFBNUM / 2;
+ }
+ if(!(tdepot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, bnum))){
+ free(name);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ free(name);
+ if(!dpsetflags(tdepot, dpgetflags(depot))){
+ dpclose(tdepot);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ tdepot->align = depot->align;
+ err = FALSE;
+ off = DP_HEADSIZ + depot->bnum * sizeof(int);
+ unum = 0;
+ while(off < depot->fsiz){
+ if(!dprechead(depot, off, head, ebuf, &ee)){
+ err = TRUE;
+ break;
+ }
+ if(!(head[DP_RHIFLAGS] & DP_RECFDEL)){
+ if(ee && DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] <= DP_ENTBUFSIZ){
+ if(!(kbufs[unum] = malloc(head[DP_RHIKSIZ] + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ memcpy(kbufs[unum], ebuf + (DP_RHNUM * sizeof(int)), head[DP_RHIKSIZ]);
+ if(DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + head[DP_RHIVSIZ] <= DP_ENTBUFSIZ){
+ if(!(vbufs[unum] = malloc(head[DP_RHIVSIZ] + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ memcpy(vbufs[unum], ebuf + (DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ]),
+ head[DP_RHIVSIZ]);
+ } else {
+ vbufs[unum] = dprecval(depot, off, head, 0, -1);
+ }
+ } else {
+ kbufs[unum] = dpreckey(depot, off, head);
+ vbufs[unum] = dprecval(depot, off, head, 0, -1);
+ }
+ ksizs[unum] = head[DP_RHIKSIZ];
+ vsizs[unum] = head[DP_RHIVSIZ];
+ unum++;
+ if(unum >= DP_OPTRUNIT){
+ for(i = 0; i < unum; i++){
+ if(kbufs[i] && vbufs[i]){
+ if(!dpput(tdepot, kbufs[i], ksizs[i], vbufs[i], vsizs[i], DP_DKEEP)) err = TRUE;
+ } else {
+ err = TRUE;
+ }
+ free(kbufs[i]);
+ free(vbufs[i]);
+ if(err) break;
+ }
+ unum = 0;
+ }
+ }
+ off += dprecsize(head);
+ if(err) break;
+ }
+ for(i = 0; i < unum; i++){
+ if(kbufs[i] && vbufs[i]){
+ if(!dpput(tdepot, kbufs[i], ksizs[i], vbufs[i], vsizs[i], DP_DKEEP)) err = TRUE;
+ } else {
+ err = TRUE;
+ }
+ free(kbufs[i]);
+ free(vbufs[i]);
+ if(err) break;
+ }
+ if(!dpsync(tdepot)) err = TRUE;
+ if(err){
+ unlink(tdepot->name);
+ dpclose(tdepot);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(munmap(depot->map, depot->msiz) == -1){
+ dpclose(tdepot);
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ depot->map = MAP_FAILED;
+ if(ftruncate(depot->fd, 0) == -1){
+ dpclose(tdepot);
+ unlink(tdepot->name);
+ dpecodeset(DP_ETRUNC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(dpfcopy(depot->fd, 0, tdepot->fd, 0) == -1){
+ dpclose(tdepot);
+ unlink(tdepot->name);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ depot->fsiz = tdepot->fsiz;
+ depot->bnum = tdepot->bnum;
+ depot->ioff = 0;
+ for(i = 0; i < depot->fbpsiz; i += 2){
+ depot->fbpool[i] = -1;
+ depot->fbpool[i+1] = -1;
+ }
+ depot->msiz = tdepot->msiz;
+ depot->map = mmap(0, depot->msiz, PROT_READ | PROT_WRITE, MAP_SHARED, depot->fd, 0);
+ if(depot->map == MAP_FAILED){
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ depot->buckets = (int *)(depot->map + DP_HEADSIZ);
+ if(!(name = dpname(tdepot))){
+ dpclose(tdepot);
+ unlink(tdepot->name);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(!dpclose(tdepot)){
+ free(name);
+ unlink(tdepot->name);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ if(unlink(name) == -1){
+ free(name);
+ dpecodeset(DP_EUNLINK, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ free(name);
+ return TRUE;
+}
+
+
+/* Get the name of a database. */
+char *dpname(DEPOT *depot){
+ char *name;
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!(name = dpstrdup(depot->name))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return NULL;
+ }
+ return name;
+}
+
+
+/* Get the size of a database file. */
+int dpfsiz(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return depot->fsiz;
+}
+
+
+/* Get the number of the elements of the bucket array. */
+int dpbnum(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return depot->bnum;
+}
+
+
+/* Get the number of the used elements of the bucket array. */
+int dpbusenum(DEPOT *depot){
+ int i, hits;
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ hits = 0;
+ for(i = 0; i < depot->bnum; i++){
+ if(depot->buckets[i]) hits++;
+ }
+ return hits;
+}
+
+
+/* Get the number of the records stored in a database. */
+int dprnum(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return depot->rnum;
+}
+
+
+/* Check whether a database handle is a writer or not. */
+int dpwritable(DEPOT *depot){
+ assert(depot);
+ return depot->wmode;
+}
+
+
+/* Check whether a database has a fatal error or not. */
+int dpfatalerror(DEPOT *depot){
+ assert(depot);
+ return depot->fatal;
+}
+
+
+/* Get the inode number of a database file. */
+int dpinode(DEPOT *depot){
+ assert(depot);
+ return depot->inode;
+}
+
+
+/* Get the last modified time of a database. */
+time_t dpmtime(DEPOT *depot){
+ assert(depot);
+ return depot->mtime;
+}
+
+
+/* Get the file descriptor of a database file. */
+int dpfdesc(DEPOT *depot){
+ assert(depot);
+ return depot->fd;
+}
+
+
+/* Remove a database file. */
+int dpremove(const char *name){
+ struct stat sbuf;
+ DEPOT *depot;
+ assert(name);
+ if(lstat(name, &sbuf) == -1){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if((depot = dpopen(name, DP_OWRITER | DP_OTRUNC, -1)) != NULL) dpclose(depot);
+ if(unlink(name) == -1){
+ dpecodeset(DP_EUNLINK, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Repair a broken database file. */
+int dprepair(const char *name){
+ DEPOT *tdepot;
+ char dbhead[DP_HEADSIZ], *tname, *kbuf, *vbuf;
+ int fd, fsiz, flags, bnum, tbnum, err, head[DP_RHNUM], off, rsiz, ksiz, vsiz;
+ struct stat sbuf;
+ assert(name);
+ if(lstat(name, &sbuf) == -1){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return FALSE;
+ }
+ fsiz = sbuf.st_size;
+ if((fd = open(name, O_RDWR, DP_FILEMODE)) == -1){
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!dpseekread(fd, 0, dbhead, DP_HEADSIZ)){
+ close(fd);
+ return FALSE;
+ }
+ flags = *(int *)(dbhead + DP_FLAGSOFF);
+ bnum = *(int *)(dbhead + DP_BNUMOFF);
+ tbnum = *(int *)(dbhead + DP_RNUMOFF) * 2;
+ if(tbnum < DP_DEFBNUM) tbnum = DP_DEFBNUM;
+ if(!(tname = malloc(strlen(name) + strlen(DP_TMPFSUF) + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ sprintf(tname, "%s%s", name, DP_TMPFSUF);
+ if(!(tdepot = dpopen(tname, DP_OWRITER | DP_OCREAT | DP_OTRUNC, tbnum))){
+ free(tname);
+ close(fd);
+ return FALSE;
+ }
+ err = FALSE;
+ off = DP_HEADSIZ + bnum * sizeof(int);
+ while(off < fsiz){
+ if(!dpseekread(fd, off, head, DP_RHNUM * sizeof(int))) break;
+ if(head[DP_RHIFLAGS] & DP_RECFDEL){
+ if((rsiz = dprecsize(head)) < 0) break;
+ off += rsiz;
+ continue;
+ }
+ ksiz = head[DP_RHIKSIZ];
+ vsiz = head[DP_RHIVSIZ];
+ if(ksiz >= 0 && vsiz >= 0){
+ kbuf = malloc(ksiz + 1);
+ vbuf = malloc(vsiz + 1);
+ if(kbuf && vbuf){
+ if(dpseekread(fd, off + DP_RHNUM * sizeof(int), kbuf, ksiz) &&
+ dpseekread(fd, off + DP_RHNUM * sizeof(int) + ksiz, vbuf, vsiz)){
+ if(!dpput(tdepot, kbuf, ksiz, vbuf, vsiz, DP_DKEEP)) err = TRUE;
+ } else {
+ err = TRUE;
+ }
+ } else {
+ if(!err) dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(vbuf);
+ free(kbuf);
+ } else {
+ if(!err) dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ if((rsiz = dprecsize(head)) < 0) break;
+ off += rsiz;
+ }
+ if(!dpsetflags(tdepot, flags)) err = TRUE;
+ if(!dpsync(tdepot)) err = TRUE;
+ if(ftruncate(fd, 0) == -1){
+ if(!err) dpecodeset(DP_ETRUNC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ if(dpfcopy(fd, 0, tdepot->fd, 0) == -1) err = TRUE;
+ if(!dpclose(tdepot)) err = TRUE;
+ if(close(fd) == -1){
+ if(!err) dpecodeset(DP_ECLOSE, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ if(unlink(tname) == -1){
+ if(!err) dpecodeset(DP_EUNLINK, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(tname);
+ return err ? FALSE : TRUE;
+}
+
+
+/* Dump all records as endian independent data. */
+int dpexportdb(DEPOT *depot, const char *name){
+ char *kbuf, *vbuf, *pbuf;
+ int fd, err, ksiz, vsiz, psiz;
+ assert(depot && name);
+ if(!(dpiterinit(depot))) return FALSE;
+ if((fd = open(name, O_RDWR | O_CREAT | O_TRUNC, DP_FILEMODE)) == -1){
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ while(!err && (kbuf = dpiternext(depot, &ksiz)) != NULL){
+ if((vbuf = dpget(depot, kbuf, ksiz, 0, -1, &vsiz)) != NULL){
+ if((pbuf = malloc(ksiz + vsiz + DP_NUMBUFSIZ * 2)) != NULL){
+ psiz = 0;
+ psiz += sprintf(pbuf + psiz, "%X\n%X\n", ksiz, vsiz);
+ memcpy(pbuf + psiz, kbuf, ksiz);
+ psiz += ksiz;
+ pbuf[psiz++] = '\n';
+ memcpy(pbuf + psiz, vbuf, vsiz);
+ psiz += vsiz;
+ pbuf[psiz++] = '\n';
+ if(!dpwrite(fd, pbuf, psiz)){
+ dpecodeset(DP_EWRITE, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(pbuf);
+ } else {
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ }
+ if(close(fd) == -1){
+ if(!err) dpecodeset(DP_ECLOSE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return !err && !dpfatalerror(depot);
+}
+
+
+/* Load all records from endian independent data. */
+int dpimportdb(DEPOT *depot, const char *name){
+ char mbuf[DP_IOBUFSIZ], *rbuf;
+ int i, j, fd, err, fsiz, off, msiz, ksiz, vsiz, hlen;
+ struct stat sbuf;
+ assert(depot && name);
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(dprnum(depot) > 0){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if((fd = open(name, O_RDONLY, DP_FILEMODE)) == -1){
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(fstat(fd, &sbuf) == -1 || !S_ISREG(sbuf.st_mode)){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ close(fd);
+ return FALSE;
+ }
+ err = FALSE;
+ fsiz = sbuf.st_size;
+ off = 0;
+ while(!err && off < fsiz){
+ msiz = fsiz - off;
+ if(msiz > DP_IOBUFSIZ) msiz = DP_IOBUFSIZ;
+ if(!dpseekread(fd, off, mbuf, msiz)){
+ err = TRUE;
+ break;
+ }
+ hlen = 0;
+ ksiz = -1;
+ vsiz = -1;
+ for(i = 0; i < msiz; i++){
+ if(mbuf[i] == '\n'){
+ mbuf[i] = '\0';
+ ksiz = strtol(mbuf, NULL, 16);
+ for(j = i + 1; j < msiz; j++){
+ if(mbuf[j] == '\n'){
+ mbuf[j] = '\0';
+ vsiz = strtol(mbuf + i + 1, NULL, 16);
+ hlen = j + 1;
+ break;
+ }
+ }
+ break;
+ }
+ }
+ if(ksiz < 0 || vsiz < 0 || hlen < 4){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ if(hlen + ksiz + vsiz + 2 < DP_IOBUFSIZ){
+ if(!dpput(depot, mbuf + hlen, ksiz, mbuf + hlen + ksiz + 1, vsiz, DP_DKEEP)) err = TRUE;
+ } else {
+ if((rbuf = malloc(ksiz + vsiz + 2)) != NULL){
+ if(dpseekread(fd, off + hlen, rbuf, ksiz + vsiz + 2)){
+ if(!dpput(depot, rbuf, ksiz, rbuf + ksiz + 1, vsiz, DP_DKEEP)) err = TRUE;
+ } else {
+ err = TRUE;
+ }
+ free(rbuf);
+ } else {
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ }
+ off += hlen + ksiz + vsiz + 2;
+ }
+ if(close(fd) == -1){
+ if(!err) dpecodeset(DP_ECLOSE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return !err && !dpfatalerror(depot);
+}
+
+
+/* Retrieve a record directly from a database file. */
+char *dpsnaffle(const char *name, const char* kbuf, int ksiz, int *sp){
+ char hbuf[DP_HEADSIZ], *map, *vbuf, *tkbuf;
+ int fd, fsiz, bnum, msiz, *buckets, hash, thash, head[DP_RHNUM], err, off, vsiz, tksiz, kcmp;
+ struct stat sbuf;
+ assert(name && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if((fd = open(name, O_RDONLY, DP_FILEMODE)) == -1){
+ dpecodeset(DP_EOPEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(fstat(fd, &sbuf) == -1 || !S_ISREG(sbuf.st_mode)){
+ close(fd);
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return NULL;
+ }
+ fsiz = sbuf.st_size;
+ if(!dpseekread(fd, 0, hbuf, DP_HEADSIZ)){
+ close(fd);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(dpbigendian() ? memcmp(hbuf, DP_MAGICNUMB, strlen(DP_MAGICNUMB)) != 0 :
+ memcmp(hbuf, DP_MAGICNUML, strlen(DP_MAGICNUML)) != 0){
+ close(fd);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ bnum = *((int *)(hbuf + DP_BNUMOFF));
+ if(bnum < 1 || fsiz < DP_HEADSIZ + bnum * sizeof(int)){
+ close(fd);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ msiz = DP_HEADSIZ + bnum * sizeof(int);
+ map = mmap(0, msiz, PROT_READ, MAP_SHARED, fd, 0);
+ if(map == MAP_FAILED){
+ close(fd);
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ return NULL;
+ }
+ buckets = (int *)(map + DP_HEADSIZ);
+ err = FALSE;
+ vbuf = NULL;
+ vsiz = 0;
+ DP_SECONDHASH(hash, kbuf, ksiz);
+ DP_FIRSTHASH(thash, kbuf, ksiz);
+ off = buckets[thash%bnum];
+ while(off != 0){
+ if(!dpseekread(fd, off, head, DP_RHNUM * sizeof(int))){
+ err = TRUE;
+ break;
+ }
+ if(head[DP_RHIKSIZ] < 0 || head[DP_RHIVSIZ] < 0 || head[DP_RHIPSIZ] < 0 ||
+ head[DP_RHILEFT] < 0 || head[DP_RHIRIGHT] < 0){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ thash = head[DP_RHIHASH];
+ if(hash > thash){
+ off = head[DP_RHILEFT];
+ } else if(hash < thash){
+ off = head[DP_RHIRIGHT];
+ } else {
+ tksiz = head[DP_RHIKSIZ];
+ if(!(tkbuf = malloc(tksiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ if(!dpseekread(fd, off + DP_RHNUM * sizeof(int), tkbuf, tksiz)){
+ free(tkbuf);
+ err = TRUE;
+ break;
+ }
+ tkbuf[tksiz] = '\0';
+ kcmp = dpkeycmp(kbuf, ksiz, tkbuf, tksiz);
+ free(tkbuf);
+ if(kcmp > 0){
+ off = head[DP_RHILEFT];
+ } else if(kcmp < 0){
+ off = head[DP_RHIRIGHT];
+ } else if(head[DP_RHIFLAGS] & DP_RECFDEL){
+ break;
+ } else {
+ vsiz = head[DP_RHIVSIZ];
+ if(!(vbuf = malloc(vsiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ err = TRUE;
+ break;
+ }
+ if(!dpseekread(fd, off + DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ], vbuf, vsiz)){
+ free(vbuf);
+ vbuf = NULL;
+ err = TRUE;
+ break;
+ }
+ vbuf[vsiz] = '\0';
+ break;
+ }
+ }
+ }
+ if(vbuf){
+ if(sp) *sp = vsiz;
+ } else if(!err){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ }
+ munmap(map, msiz);
+ close(fd);
+ return vbuf;
+}
+
+
+/* Hash function used inside Depot. */
+int dpinnerhash(const char *kbuf, int ksiz){
+ int res;
+ assert(kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ DP_FIRSTHASH(res, kbuf, ksiz);
+ return res;
+}
+
+
+/* Hash function which is independent from the hash functions used inside Depot. */
+int dpouterhash(const char *kbuf, int ksiz){
+ int res;
+ assert(kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ DP_THIRDHASH(res, kbuf, ksiz);
+ return res;
+}
+
+
+/* Get a natural prime number not less than a number. */
+int dpprimenum(int num){
+ assert(num > 0);
+ return dpgetprime(num);
+}
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Name of the operating system. */
+const char *dpsysname = _QDBM_SYSNAME;
+
+
+/* File descriptor for debugging output. */
+int dpdbgfd = -1;
+
+
+/* Whether this build is reentrant. */
+const int dpisreentrant = _qdbm_ptsafe;
+
+
+/* Set the last happened error code. */
+void dpecodeset(int ecode, const char *file, int line){
+ char iobuf[DP_IOBUFSIZ];
+ assert(ecode >= DP_ENOERR && file && line >= 0);
+ dpecode = ecode;
+ if(dpdbgfd >= 0){
+ fflush(stdout);
+ fflush(stderr);
+ sprintf(iobuf, "* dpecodeset: %s:%d: [%d] %s\n", file, line, ecode, dperrmsg(ecode));
+ dpwrite(dpdbgfd, iobuf, strlen(iobuf));
+ }
+}
+
+
+/* Get the pointer of the variable of the last happened error code. */
+int *dpecodeptr(void){
+ static int defdpecode = DP_ENOERR;
+ void *ptr;
+ if(_qdbm_ptsafe){
+ if(!(ptr = _qdbm_settsd(&defdpecode, sizeof(int), &defdpecode))){
+ defdpecode = DP_EMISC;
+ return &defdpecode;
+ }
+ return (int *)ptr;
+ }
+ return &defdpecode;
+}
+
+
+/* Synchronize updating contents on memory. */
+int dpmemsync(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ *((int *)(depot->map + DP_FSIZOFF)) = depot->fsiz;
+ *((int *)(depot->map + DP_RNUMOFF)) = depot->rnum;
+ if(msync(depot->map, depot->msiz, MS_SYNC) == -1){
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Synchronize updating contents on memory, not physically. */
+int dpmemflush(DEPOT *depot){
+ assert(depot);
+ if(depot->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ *((int *)(depot->map + DP_FSIZOFF)) = depot->fsiz;
+ *((int *)(depot->map + DP_RNUMOFF)) = depot->rnum;
+ if(mflush(depot->map, depot->msiz, MS_SYNC) == -1){
+ dpecodeset(DP_EMAP, __FILE__, __LINE__);
+ depot->fatal = TRUE;
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Get flags of a database. */
+int dpgetflags(DEPOT *depot){
+ int flags;
+ assert(depot);
+ memcpy(&flags, depot->map + DP_FLAGSOFF, sizeof(int));
+ return flags;
+}
+
+
+/* Set flags of a database. */
+int dpsetflags(DEPOT *depot, int flags){
+ assert(depot);
+ if(!depot->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ memcpy(depot->map + DP_FLAGSOFF, &flags, sizeof(int));
+ return TRUE;
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Check whether the byte order of the platform is big endian or not.
+ The return value is true if bigendian, else, it is false. */
+static int dpbigendian(void){
+ char buf[sizeof(int)];
+ *(int *)buf = 1;
+ return !buf[0];
+}
+
+
+/* Get a copied string.
+ `str' specifies an original string.
+ The return value is a copied string whose region is allocated by `malloc'. */
+static char *dpstrdup(const char *str){
+ int len;
+ char *buf;
+ assert(str);
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ memcpy(buf, str, len + 1);
+ return buf;
+}
+
+
+/* Lock a file descriptor.
+ `fd' specifies a file descriptor.
+ `ex' specifies whether an exclusive lock or a shared lock is performed.
+ `nb' specifies whether to request with non-blocking.
+ The return value is true if successful, else, it is false. */
+static int dplock(int fd, int ex, int nb){
+ struct flock lock;
+ assert(fd >= 0);
+ memset(&lock, 0, sizeof(struct flock));
+ lock.l_type = ex ? F_WRLCK : F_RDLCK;
+ lock.l_whence = SEEK_SET;
+ lock.l_start = 0;
+ lock.l_len = 0;
+ lock.l_pid = 0;
+ while(fcntl(fd, nb ? F_SETLK : F_SETLKW, &lock) == -1){
+ if(errno != EINTR){
+ dpecodeset(DP_ELOCK, __FILE__, __LINE__);
+ return FALSE;
+ }
+ }
+ return TRUE;
+}
+
+
+/* Write into a file.
+ `fd' specifies a file descriptor.
+ `buf' specifies a buffer to write.
+ `size' specifies the size of the buffer.
+ The return value is the size of the written buffer, or, -1 on failure. */
+static int dpwrite(int fd, const void *buf, int size){
+ const char *lbuf;
+ int rv, wb;
+ assert(fd >= 0 && buf && size >= 0);
+ lbuf = buf;
+ rv = 0;
+ do {
+ wb = write(fd, lbuf, size);
+ switch(wb){
+ case -1: if(errno != EINTR) return -1;
+ case 0: break;
+ default:
+ lbuf += wb;
+ size -= wb;
+ rv += wb;
+ break;
+ }
+ } while(size > 0);
+ return rv;
+}
+
+
+/* Write into a file at an offset.
+ `fd' specifies a file descriptor.
+ `off' specifies an offset of the file.
+ `buf' specifies a buffer to write.
+ `size' specifies the size of the buffer.
+ The return value is true if successful, else, it is false. */
+static int dpseekwrite(int fd, int off, const void *buf, int size){
+ assert(fd >= 0 && buf && size >= 0);
+ if(size < 1) return TRUE;
+ if(off < 0){
+ if(lseek(fd, 0, SEEK_END) == -1){
+ dpecodeset(DP_ESEEK, __FILE__, __LINE__);
+ return FALSE;
+ }
+ } else {
+ if(lseek(fd, off, SEEK_SET) != off){
+ dpecodeset(DP_ESEEK, __FILE__, __LINE__);
+ return FALSE;
+ }
+ }
+ if(dpwrite(fd, buf, size) != size){
+ dpecodeset(DP_EWRITE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Write an integer into a file at an offset.
+ `fd' specifies a file descriptor.
+ `off' specifies an offset of the file.
+ `num' specifies an integer.
+ The return value is true if successful, else, it is false. */
+static int dpseekwritenum(int fd, int off, int num){
+ assert(fd >= 0);
+ return dpseekwrite(fd, off, &num, sizeof(int));
+}
+
+
+/* Read from a file and store the data into a buffer.
+ `fd' specifies a file descriptor.
+ `buffer' specifies a buffer to store into.
+ `size' specifies the size to read with.
+ The return value is the size read with, or, -1 on failure. */
+static int dpread(int fd, void *buf, int size){
+ char *lbuf;
+ int i, bs;
+ assert(fd >= 0 && buf && size >= 0);
+ lbuf = buf;
+ for(i = 0; i < size && (bs = read(fd, lbuf + i, size - i)) != 0; i += bs){
+ if(bs == -1 && errno != EINTR) return -1;
+ }
+ return i;
+}
+
+
+/* Read from a file at an offset and store the data into a buffer.
+ `fd' specifies a file descriptor.
+ `off' specifies an offset of the file.
+ `buffer' specifies a buffer to store into.
+ `size' specifies the size to read with.
+ The return value is true if successful, else, it is false. */
+static int dpseekread(int fd, int off, void *buf, int size){
+ char *lbuf;
+ assert(fd >= 0 && off >= 0 && buf && size >= 0);
+ lbuf = (char *)buf;
+ if(lseek(fd, off, SEEK_SET) != off){
+ dpecodeset(DP_ESEEK, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(dpread(fd, lbuf, size) != size){
+ dpecodeset(DP_EREAD, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Copy data between files.
+ `destfd' specifies a file descriptor of a destination file.
+ `destoff' specifies an offset of the destination file.
+ `srcfd' specifies a file descriptor of a source file.
+ `srcoff' specifies an offset of the source file.
+ The return value is the size copied with, or, -1 on failure. */
+static int dpfcopy(int destfd, int destoff, int srcfd, int srcoff){
+ char iobuf[DP_IOBUFSIZ];
+ int sum, iosiz;
+ if(lseek(srcfd, srcoff, SEEK_SET) == -1 || lseek(destfd, destoff, SEEK_SET) == -1){
+ dpecodeset(DP_ESEEK, __FILE__, __LINE__);
+ return -1;
+ }
+ sum = 0;
+ while((iosiz = dpread(srcfd, iobuf, DP_IOBUFSIZ)) > 0){
+ if(dpwrite(destfd, iobuf, iosiz) == -1){
+ dpecodeset(DP_EWRITE, __FILE__, __LINE__);
+ return -1;
+ }
+ sum += iosiz;
+ }
+ if(iosiz < 0){
+ dpecodeset(DP_EREAD, __FILE__, __LINE__);
+ return -1;
+ }
+ return sum;
+}
+
+
+/* Get a natural prime number not less than a number.
+ `num' specified a natural number.
+ The return value is a prime number not less than the specified number. */
+static int dpgetprime(int num){
+ int primes[] = {
+ 1, 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 43, 47, 53, 59, 61, 71, 79, 83,
+ 89, 103, 109, 113, 127, 139, 157, 173, 191, 199, 223, 239, 251, 283, 317, 349,
+ 383, 409, 443, 479, 509, 571, 631, 701, 761, 829, 887, 953, 1021, 1151, 1279,
+ 1399, 1531, 1663, 1789, 1913, 2039, 2297, 2557, 2803, 3067, 3323, 3583, 3833,
+ 4093, 4603, 5119, 5623, 6143, 6653, 7159, 7673, 8191, 9209, 10223, 11261,
+ 12281, 13309, 14327, 15359, 16381, 18427, 20479, 22511, 24571, 26597, 28669,
+ 30713, 32749, 36857, 40949, 45053, 49139, 53239, 57331, 61417, 65521, 73727,
+ 81919, 90107, 98299, 106487, 114679, 122869, 131071, 147451, 163819, 180221,
+ 196597, 212987, 229373, 245759, 262139, 294911, 327673, 360439, 393209, 425977,
+ 458747, 491503, 524287, 589811, 655357, 720887, 786431, 851957, 917503, 982981,
+ 1048573, 1179641, 1310719, 1441771, 1572853, 1703903, 1835003, 1966079,
+ 2097143, 2359267, 2621431, 2883577, 3145721, 3407857, 3670013, 3932153,
+ 4194301, 4718579, 5242877, 5767129, 6291449, 6815741, 7340009, 7864301,
+ 8388593, 9437179, 10485751, 11534329, 12582893, 13631477, 14680063, 15728611,
+ 16777213, 18874367, 20971507, 23068667, 25165813, 27262931, 29360087, 31457269,
+ 33554393, 37748717, 41943023, 46137319, 50331599, 54525917, 58720253, 62914549,
+ 67108859, 75497467, 83886053, 92274671, 100663291, 109051903, 117440509,
+ 125829103, 134217689, 150994939, 167772107, 184549373, 201326557, 218103799,
+ 234881011, 251658227, 268435399, 301989881, 335544301, 369098707, 402653171,
+ 436207613, 469762043, 503316469, 536870909, 603979769, 671088637, 738197503,
+ 805306357, 872415211, 939524087, 1006632947, 1073741789, 1207959503,
+ 1342177237, 1476394991, 1610612711, 1744830457, 1879048183, 2013265907, -1
+ };
+ int i;
+ assert(num > 0);
+ for(i = 0; primes[i] > 0; i++){
+ if(num <= primes[i]) return primes[i];
+ }
+ return primes[i-1];
+}
+
+
+/* Get the padding size of a record.
+ `vsiz' specifies the size of the value of a record.
+ The return value is the padding size of a record. */
+static int dppadsize(DEPOT *depot, int ksiz, int vsiz){
+ int pad;
+ assert(depot && vsiz >= 0);
+ if(depot->align > 0){
+ return depot->align - (depot->fsiz + DP_RHNUM * sizeof(int) + ksiz + vsiz) % depot->align;
+ } else if(depot->align < 0){
+ pad = (int)(vsiz * (2.0 / (1 << -(depot->align))));
+ if(vsiz + pad >= DP_FSBLKSIZ){
+ if(vsiz <= DP_FSBLKSIZ) pad = 0;
+ if(depot->fsiz % DP_FSBLKSIZ == 0){
+ return (pad / DP_FSBLKSIZ) * DP_FSBLKSIZ + DP_FSBLKSIZ -
+ (depot->fsiz + DP_RHNUM * sizeof(int) + ksiz + vsiz) % DP_FSBLKSIZ;
+ } else {
+ return (pad / (DP_FSBLKSIZ / 2)) * (DP_FSBLKSIZ / 2) + (DP_FSBLKSIZ / 2) -
+ (depot->fsiz + DP_RHNUM * sizeof(int) + ksiz + vsiz) % (DP_FSBLKSIZ / 2);
+ }
+ } else {
+ return pad >= DP_RHNUM * sizeof(int) ? pad : DP_RHNUM * sizeof(int);
+ }
+ }
+ return 0;
+}
+
+
+/* Get the size of a record in a database file.
+ `head' specifies the header of a record.
+ The return value is the size of a record in a database file. */
+static int dprecsize(int *head){
+ assert(head);
+ return DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + head[DP_RHIVSIZ] + head[DP_RHIPSIZ];
+}
+
+
+/* Read the header of a record.
+ `depot' specifies a database handle.
+ `off' specifies an offset of the database file.
+ `head' specifies a buffer for the header.
+ `ebuf' specifies the pointer to the entity buffer.
+ `eep' specifies the pointer to a variable to which whether ebuf was used is assigned.
+ The return value is true if successful, else, it is false. */
+static int dprechead(DEPOT *depot, int off, int *head, char *ebuf, int *eep){
+ assert(depot && off >= 0 && head);
+ if(off > depot->fsiz){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ebuf){
+ *eep = FALSE;
+ if(off < depot->fsiz - DP_ENTBUFSIZ){
+ *eep = TRUE;
+ if(!dpseekread(depot->fd, off, ebuf, DP_ENTBUFSIZ)) return FALSE;
+ memcpy(head, ebuf, DP_RHNUM * sizeof(int));
+ if(head[DP_RHIKSIZ] < 0 || head[DP_RHIVSIZ] < 0 || head[DP_RHIPSIZ] < 0 ||
+ head[DP_RHILEFT] < 0 || head[DP_RHIRIGHT] < 0){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+ }
+ }
+ if(!dpseekread(depot->fd, off, head, DP_RHNUM * sizeof(int))) return FALSE;
+ if(head[DP_RHIKSIZ] < 0 || head[DP_RHIVSIZ] < 0 || head[DP_RHIPSIZ] < 0 ||
+ head[DP_RHILEFT] < 0 || head[DP_RHIRIGHT] < 0){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Read the entitiy of the key of a record.
+ `depot' specifies a database handle.
+ `off' specifies an offset of the database file.
+ `head' specifies the header of a record.
+ The return value is a key data whose region is allocated by `malloc', or NULL on failure. */
+static char *dpreckey(DEPOT *depot, int off, int *head){
+ char *kbuf;
+ int ksiz;
+ assert(depot && off >= 0);
+ ksiz = head[DP_RHIKSIZ];
+ if(!(kbuf = malloc(ksiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!dpseekread(depot->fd, off + DP_RHNUM * sizeof(int), kbuf, ksiz)){
+ free(kbuf);
+ return NULL;
+ }
+ kbuf[ksiz] = '\0';
+ return kbuf;
+}
+
+
+/* Read the entitiy of the value of a record.
+ `depot' specifies a database handle.
+ `off' specifies an offset of the database file.
+ `head' specifies the header of a record.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. If it is negative, the size to read is unlimited.
+ The return value is a value data whose region is allocated by `malloc', or NULL on failure. */
+static char *dprecval(DEPOT *depot, int off, int *head, int start, int max){
+ char *vbuf;
+ int vsiz;
+ assert(depot && off >= 0 && start >= 0);
+ head[DP_RHIVSIZ] -= start;
+ if(max < 0){
+ vsiz = head[DP_RHIVSIZ];
+ } else {
+ vsiz = max < head[DP_RHIVSIZ] ? max : head[DP_RHIVSIZ];
+ }
+ if(!(vbuf = malloc(vsiz + 1))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!dpseekread(depot->fd, off + DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + start, vbuf, vsiz)){
+ free(vbuf);
+ return NULL;
+ }
+ vbuf[vsiz] = '\0';
+ return vbuf;
+}
+
+
+/* Read the entitiy of the value of a record and write it into a given buffer.
+ `depot' specifies a database handle.
+ `off' specifies an offset of the database file.
+ `head' specifies the header of a record.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. It shuld be less than the size of the writing buffer.
+ If successful, the return value is the size of the written data, else, it is -1. */
+static int dprecvalwb(DEPOT *depot, int off, int *head, int start, int max, char *vbuf){
+ int vsiz;
+ assert(depot && off >= 0 && start >= 0 && max >= 0 && vbuf);
+ head[DP_RHIVSIZ] -= start;
+ vsiz = max < head[DP_RHIVSIZ] ? max : head[DP_RHIVSIZ];
+ if(!dpseekread(depot->fd, off + DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + start, vbuf, vsiz))
+ return -1;
+ return vsiz;
+}
+
+
+/* Compare two keys.
+ `abuf' specifies the pointer to the region of the former.
+ `asiz' specifies the size of the region.
+ `bbuf' specifies the pointer to the region of the latter.
+ `bsiz' specifies the size of the region.
+ The return value is 0 if two equals, positive if the formar is big, else, negative. */
+static int dpkeycmp(const char *abuf, int asiz, const char *bbuf, int bsiz){
+ assert(abuf && asiz >= 0 && bbuf && bsiz >= 0);
+ if(asiz > bsiz) return 1;
+ if(asiz < bsiz) return -1;
+ return memcmp(abuf, bbuf, asiz);
+}
+
+
+/* Search for a record.
+ `depot' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region.
+ `hash' specifies the second hash value of the key.
+ `bip' specifies the pointer to the region to assign the index of the corresponding record.
+ `offp' specifies the pointer to the region to assign the last visited node in the hash chain,
+ or, -1 if the hash chain is empty.
+ `entp' specifies the offset of the last used joint, or, -1 if the hash chain is empty.
+ `head' specifies the pointer to the region to store the header of the last visited record in.
+ `ebuf' specifies the pointer to the entity buffer.
+ `eep' specifies the pointer to a variable to which whether ebuf was used is assigned.
+ `delhit' specifies whether a deleted record corresponds or not.
+ The return value is 0 if successful, 1 if there is no corresponding record, -1 on error. */
+static int dprecsearch(DEPOT *depot, const char *kbuf, int ksiz, int hash, int *bip, int *offp,
+ int *entp, int *head, char *ebuf, int *eep, int delhit){
+ int off, entoff, thash, kcmp;
+ char stkey[DP_STKBUFSIZ], *tkey;
+ assert(depot && kbuf && ksiz >= 0 && hash >= 0 && bip && offp && entp && head && ebuf && eep);
+ DP_FIRSTHASH(thash, kbuf, ksiz);
+ *bip = thash % depot->bnum;
+ off = depot->buckets[*bip];
+ *offp = -1;
+ *entp = -1;
+ entoff = -1;
+ *eep = FALSE;
+ while(off != 0){
+ if(!dprechead(depot, off, head, ebuf, eep)) return -1;
+ thash = head[DP_RHIHASH];
+ if(hash > thash){
+ entoff = off + DP_RHILEFT * sizeof(int);
+ off = head[DP_RHILEFT];
+ } else if(hash < thash){
+ entoff = off + DP_RHIRIGHT * sizeof(int);
+ off = head[DP_RHIRIGHT];
+ } else {
+ if(*eep && DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] <= DP_ENTBUFSIZ){
+ kcmp = dpkeycmp(kbuf, ksiz, ebuf + (DP_RHNUM * sizeof(int)), head[DP_RHIKSIZ]);
+ } else if(head[DP_RHIKSIZ] > DP_STKBUFSIZ){
+ if(!(tkey = dpreckey(depot, off, head))) return -1;
+ kcmp = dpkeycmp(kbuf, ksiz, tkey, head[DP_RHIKSIZ]);
+ free(tkey);
+ } else {
+ if(!dpseekread(depot->fd, off + DP_RHNUM * sizeof(int), stkey, head[DP_RHIKSIZ]))
+ return -1;
+ kcmp = dpkeycmp(kbuf, ksiz, stkey, head[DP_RHIKSIZ]);
+ }
+ if(kcmp > 0){
+ entoff = off + DP_RHILEFT * sizeof(int);
+ off = head[DP_RHILEFT];
+ } else if(kcmp < 0){
+ entoff = off + DP_RHIRIGHT * sizeof(int);
+ off = head[DP_RHIRIGHT];
+ } else {
+ if(!delhit && (head[DP_RHIFLAGS] & DP_RECFDEL)){
+ entoff = off + DP_RHILEFT * sizeof(int);
+ off = head[DP_RHILEFT];
+ } else {
+ *offp = off;
+ *entp = entoff;
+ return 0;
+ }
+ }
+ }
+ }
+ *offp = off;
+ *entp = entoff;
+ return 1;
+}
+
+
+/* Overwrite a record.
+ `depot' specifies a database handle.
+ `off' specifies the offset of the database file.
+ `rsiz' specifies the size of the existing record.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region.
+ `hash' specifies the second hash value of the key.
+ `left' specifies the offset of the left child.
+ `right' specifies the offset of the right child.
+ The return value is true if successful, or, false on failure. */
+static int dprecrewrite(DEPOT *depot, int off, int rsiz, const char *kbuf, int ksiz,
+ const char *vbuf, int vsiz, int hash, int left, int right){
+ char ebuf[DP_WRTBUFSIZ];
+ int i, head[DP_RHNUM], asiz, hoff, koff, voff, mi, min, size;
+ assert(depot && off >= 1 && rsiz > 0 && kbuf && ksiz >= 0 && vbuf && vsiz >= 0);
+ head[DP_RHIFLAGS] = 0;
+ head[DP_RHIHASH] = hash;
+ head[DP_RHIKSIZ] = ksiz;
+ head[DP_RHIVSIZ] = vsiz;
+ head[DP_RHIPSIZ] = rsiz - sizeof(head) - ksiz - vsiz;
+ head[DP_RHILEFT] = left;
+ head[DP_RHIRIGHT] = right;
+ asiz = sizeof(head) + ksiz + vsiz;
+ if(depot->fbpsiz > DP_FBPOOLSIZ * 4 && head[DP_RHIPSIZ] > asiz){
+ rsiz = (head[DP_RHIPSIZ] - asiz) / 2 + asiz;
+ head[DP_RHIPSIZ] -= rsiz;
+ } else {
+ rsiz = 0;
+ }
+ if(asiz <= DP_WRTBUFSIZ){
+ memcpy(ebuf, head, sizeof(head));
+ memcpy(ebuf + sizeof(head), kbuf, ksiz);
+ memcpy(ebuf + sizeof(head) + ksiz, vbuf, vsiz);
+ if(!dpseekwrite(depot->fd, off, ebuf, asiz)) return FALSE;
+ } else {
+ hoff = off;
+ koff = hoff + sizeof(head);
+ voff = koff + ksiz;
+ if(!dpseekwrite(depot->fd, hoff, head, sizeof(head)) ||
+ !dpseekwrite(depot->fd, koff, kbuf, ksiz) || !dpseekwrite(depot->fd, voff, vbuf, vsiz))
+ return FALSE;
+ }
+ if(rsiz > 0){
+ off += sizeof(head) + ksiz + vsiz + head[DP_RHIPSIZ];
+ head[DP_RHIFLAGS] = DP_RECFDEL | DP_RECFREUSE;
+ head[DP_RHIHASH] = hash;
+ head[DP_RHIKSIZ] = ksiz;
+ head[DP_RHIVSIZ] = vsiz;
+ head[DP_RHIPSIZ] = rsiz - sizeof(head) - ksiz - vsiz;
+ head[DP_RHILEFT] = 0;
+ head[DP_RHIRIGHT] = 0;
+ if(!dpseekwrite(depot->fd, off, head, sizeof(head))) return FALSE;
+ size = dprecsize(head);
+ mi = -1;
+ min = -1;
+ for(i = 0; i < depot->fbpsiz; i += 2){
+ if(depot->fbpool[i] == -1){
+ depot->fbpool[i] = off;
+ depot->fbpool[i+1] = size;
+ dpfbpoolcoal(depot);
+ mi = -1;
+ break;
+ }
+ if(mi == -1 || depot->fbpool[i+1] < min){
+ mi = i;
+ min = depot->fbpool[i+1];
+ }
+ }
+ if(mi >= 0 && size > min){
+ depot->fbpool[mi] = off;
+ depot->fbpool[mi+1] = size;
+ dpfbpoolcoal(depot);
+ }
+ }
+ return TRUE;
+}
+
+
+/* Write a record at the end of a database file.
+ `depot' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region.
+ `hash' specifies the second hash value of the key.
+ `left' specifies the offset of the left child.
+ `right' specifies the offset of the right child.
+ The return value is the offset of the record, or, -1 on failure. */
+static int dprecappend(DEPOT *depot, const char *kbuf, int ksiz, const char *vbuf, int vsiz,
+ int hash, int left, int right){
+ char ebuf[DP_WRTBUFSIZ], *hbuf;
+ int head[DP_RHNUM], asiz, psiz, off;
+ assert(depot && kbuf && ksiz >= 0 && vbuf && vsiz >= 0);
+ psiz = dppadsize(depot, ksiz, vsiz);
+ head[DP_RHIFLAGS] = 0;
+ head[DP_RHIHASH] = hash;
+ head[DP_RHIKSIZ] = ksiz;
+ head[DP_RHIVSIZ] = vsiz;
+ head[DP_RHIPSIZ] = psiz;
+ head[DP_RHILEFT] = left;
+ head[DP_RHIRIGHT] = right;
+ asiz = sizeof(head) + ksiz + vsiz + psiz;
+ off = depot->fsiz;
+ if(asiz <= DP_WRTBUFSIZ){
+ memcpy(ebuf, head, sizeof(head));
+ memcpy(ebuf + sizeof(head), kbuf, ksiz);
+ memcpy(ebuf + sizeof(head) + ksiz, vbuf, vsiz);
+ memset(ebuf + sizeof(head) + ksiz + vsiz, 0, psiz);
+ if(!dpseekwrite(depot->fd, off, ebuf, asiz)) return -1;
+ } else {
+ if(!(hbuf = malloc(asiz))){
+ dpecodeset(DP_EALLOC, __FILE__, __LINE__);
+ return -1;
+ }
+ memcpy(hbuf, head, sizeof(head));
+ memcpy(hbuf + sizeof(head), kbuf, ksiz);
+ memcpy(hbuf + sizeof(head) + ksiz, vbuf, vsiz);
+ memset(hbuf + sizeof(head) + ksiz + vsiz, 0, psiz);
+ if(!dpseekwrite(depot->fd, off, hbuf, asiz)){
+ free(hbuf);
+ return -1;
+ }
+ free(hbuf);
+ }
+ depot->fsiz += asiz;
+ return off;
+}
+
+
+/* Overwrite the value of a record.
+ `depot' specifies a database handle.
+ `off' specifies the offset of the database file.
+ `head' specifies the header of the record.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region.
+ `cat' specifies whether it is concatenate mode or not.
+ The return value is true if successful, or, false on failure. */
+static int dprecover(DEPOT *depot, int off, int *head, const char *vbuf, int vsiz, int cat){
+ int i, hsiz, hoff, voff;
+ assert(depot && off >= 0 && head && vbuf && vsiz >= 0);
+ for(i = 0; i < depot->fbpsiz; i += 2){
+ if(depot->fbpool[i] == off){
+ depot->fbpool[i] = -1;
+ depot->fbpool[i+1] = -1;
+ break;
+ }
+ }
+ if(cat){
+ head[DP_RHIFLAGS] = 0;
+ head[DP_RHIPSIZ] -= vsiz;
+ head[DP_RHIVSIZ] += vsiz;
+ hsiz = DP_RHNUM * sizeof(int);
+ hoff = off;
+ voff = hoff + DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ] + head[DP_RHIVSIZ] - vsiz;
+ } else {
+ head[DP_RHIFLAGS] = 0;
+ head[DP_RHIPSIZ] += head[DP_RHIVSIZ] - vsiz;
+ head[DP_RHIVSIZ] = vsiz;
+ hsiz = DP_RHNUM * sizeof(int);
+ hoff = off;
+ voff = hoff + DP_RHNUM * sizeof(int) + head[DP_RHIKSIZ];
+ }
+ if(!dpseekwrite(depot->fd, hoff, head, hsiz) ||
+ !dpseekwrite(depot->fd, voff, vbuf, vsiz)) return FALSE;
+ return TRUE;
+}
+
+
+/* Delete a record.
+ `depot' specifies a database handle.
+ `off' specifies the offset of the database file.
+ `head' specifies the header of the record.
+ `reusable' specifies whether the region is reusable or not.
+ The return value is true if successful, or, false on failure. */
+static int dprecdelete(DEPOT *depot, int off, int *head, int reusable){
+ int i, mi, min, size;
+ assert(depot && off >= 0 && head);
+ if(reusable){
+ size = dprecsize(head);
+ mi = -1;
+ min = -1;
+ for(i = 0; i < depot->fbpsiz; i += 2){
+ if(depot->fbpool[i] == -1){
+ depot->fbpool[i] = off;
+ depot->fbpool[i+1] = size;
+ dpfbpoolcoal(depot);
+ mi = -1;
+ break;
+ }
+ if(mi == -1 || depot->fbpool[i+1] < min){
+ mi = i;
+ min = depot->fbpool[i+1];
+ }
+ }
+ if(mi >= 0 && size > min){
+ depot->fbpool[mi] = off;
+ depot->fbpool[mi+1] = size;
+ dpfbpoolcoal(depot);
+ }
+ }
+ return dpseekwritenum(depot->fd, off + DP_RHIFLAGS * sizeof(int),
+ DP_RECFDEL | (reusable ? DP_RECFREUSE : 0));
+}
+
+
+/* Make contiguous records of the free block pool coalesce.
+ `depot' specifies a database handle. */
+static void dpfbpoolcoal(DEPOT *depot){
+ int i;
+ assert(depot);
+ if(depot->fbpinc++ <= depot->fbpsiz / 4) return;
+ depot->fbpinc = 0;
+ qsort(depot->fbpool, depot->fbpsiz / 2, sizeof(int) * 2, dpfbpoolcmp);
+ for(i = 2; i < depot->fbpsiz; i += 2){
+ if(depot->fbpool[i-2] > 0 &&
+ depot->fbpool[i-2] + depot->fbpool[i-1] - depot->fbpool[i] == 0){
+ depot->fbpool[i] = depot->fbpool[i-2];
+ depot->fbpool[i+1] += depot->fbpool[i-1];
+ depot->fbpool[i-2] = -1;
+ depot->fbpool[i-1] = -1;
+ }
+ }
+}
+
+
+/* Compare two records of the free block pool.
+ `a' specifies the pointer to one record.
+ `b' specifies the pointer to the other record.
+ The return value is 0 if two equals, positive if the formar is big, else, negative. */
+static int dpfbpoolcmp(const void *a, const void *b){
+ assert(a && b);
+ return *(int *)a - *(int *)b;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/depot.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/depot.h
===================================================================
--- box/trunk/qdbm/depot.h (rev 0)
+++ box/trunk/qdbm/depot.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,492 @@
+/*************************************************************************************************
+ * The basic API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _DEPOT_H /* duplication check */
+#define _DEPOT_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <stdlib.h>
+#include <time.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+typedef struct { /* type of structure for a database handle */
+ char *name; /* name of the database file */
+ int wmode; /* whether to be writable */
+ int inode; /* inode of the database file */
+ time_t mtime; /* last modified time of the database */
+ int fd; /* file descriptor of the database file */
+ int fsiz; /* size of the database file */
+ char *map; /* pointer to the mapped memory */
+ int msiz; /* size of the mapped memory */
+ int *buckets; /* pointer to the bucket array */
+ int bnum; /* number of the bucket array */
+ int rnum; /* number of records */
+ int fatal; /* whether a fatal error occured */
+ int ioff; /* offset of the iterator */
+ int *fbpool; /* free block pool */
+ int fbpsiz; /* size of the free block pool */
+ int fbpinc; /* incrementor of update of the free block pool */
+ int align; /* basic size of alignment */
+} DEPOT;
+
+enum { /* enumeration for error codes */
+ DP_ENOERR, /* no error */
+ DP_EFATAL, /* with fatal error */
+ DP_EMODE, /* invalid mode */
+ DP_EBROKEN, /* broken database file */
+ DP_EKEEP, /* existing record */
+ DP_ENOITEM, /* no item found */
+ DP_EALLOC, /* memory allocation error */
+ DP_EMAP, /* memory mapping error */
+ DP_EOPEN, /* open error */
+ DP_ECLOSE, /* close error */
+ DP_ETRUNC, /* trunc error */
+ DP_ESYNC, /* sync error */
+ DP_ESTAT, /* stat error */
+ DP_ESEEK, /* seek error */
+ DP_EREAD, /* read error */
+ DP_EWRITE, /* write error */
+ DP_ELOCK, /* lock error */
+ DP_EUNLINK, /* unlink error */
+ DP_EMKDIR, /* mkdir error */
+ DP_ERMDIR, /* rmdir error */
+ DP_EMISC /* miscellaneous error */
+};
+
+enum { /* enumeration for open modes */
+ DP_OREADER = 1 << 0, /* open as a reader */
+ DP_OWRITER = 1 << 1, /* open as a writer */
+ DP_OCREAT = 1 << 2, /* a writer creating */
+ DP_OTRUNC = 1 << 3, /* a writer truncating */
+ DP_ONOLCK = 1 << 4, /* open without locking */
+ DP_OLCKNB = 1 << 5, /* lock without blocking */
+ DP_OSPARSE = 1 << 6 /* create as a sparse file */
+};
+
+enum { /* enumeration for write modes */
+ DP_DOVER, /* overwrite an existing value */
+ DP_DKEEP, /* keep an existing value */
+ DP_DCAT /* concatenate values */
+};
+
+
+/* String containing the version information. */
+MYEXTERN const char *dpversion;
+
+
+/* Last happened error code. */
+#define dpecode (*dpecodeptr())
+
+
+/* Get a message string corresponding to an error code.
+ `ecode' specifies an error code.
+ The return value is the message string of the error code. The region of the return value
+ is not writable. */
+const char *dperrmsg(int ecode);
+
+
+/* Get a database handle.
+ `name' specifies the name of a database file.
+ `omode' specifies the connection mode: `DP_OWRITER' as a writer, `DP_OREADER' as a reader.
+ If the mode is `DP_OWRITER', the following may be added by bitwise or: `DP_OCREAT', which
+ means it creates a new database if not exist, `DP_OTRUNC', which means it creates a new
+ database regardless if one exists. Both of `DP_OREADER' and `DP_OWRITER' can be added to by
+ bitwise or: `DP_ONOLCK', which means it opens a database file without file locking, or
+ `DP_OLCKNB', which means locking is performed without blocking. `DP_OCREAT' can be added to
+ by bitwise or: `DP_OSPARSE', which means it creates a database file as a sparse file.
+ `bnum' specifies the number of elements of the bucket array. If it is not more than 0,
+ the default value is specified. The size of a bucket array is determined on creating,
+ and can not be changed except for by optimization of the database. Suggested size of a
+ bucket array is about from 0.5 to 4 times of the number of all records to store.
+ The return value is the database handle or `NULL' if it is not successful.
+ While connecting as a writer, an exclusive lock is invoked to the database file.
+ While connecting as a reader, a shared lock is invoked to the database file. The thread
+ blocks until the lock is achieved. If `DP_ONOLCK' is used, the application is responsible
+ for exclusion control. */
+DEPOT *dpopen(const char *name, int omode, int bnum);
+
+
+/* Close a database handle.
+ `depot' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ Because the region of a closed handle is released, it becomes impossible to use the handle.
+ Updating a database is assured to be written when the handle is closed. If a writer opens
+ a database but does not close it appropriately, the database will be broken. */
+int dpclose(DEPOT *depot);
+
+
+/* Store a record.
+ `depot' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ `dmode' specifies behavior when the key overlaps, by the following values: `DP_DOVER',
+ which means the specified value overwrites the existing one, `DP_DKEEP', which means the
+ existing value is kept, `DP_DCAT', which means the specified value is concatenated at the
+ end of the existing value.
+ If successful, the return value is true, else, it is false. */
+int dpput(DEPOT *depot, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode);
+
+
+/* Delete a record.
+ `depot' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is true, else, it is false. False is returned when no
+ record corresponds to the specified key. */
+int dpout(DEPOT *depot, const char *kbuf, int ksiz);
+
+
+/* Retrieve a record.
+ `depot' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. If it is negative, the size to read is unlimited.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the specified key or the size of the value of the corresponding record is less than `start'.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. */
+char *dpget(DEPOT *depot, const char *kbuf, int ksiz, int start, int max, int *sp);
+
+
+/* Retrieve a record and write the value into a buffer.
+ `depot' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `start' specifies the offset address of the beginning of the region of the value to be read.
+ `max' specifies the max size to be read. It shuld be equal to or less than the size of the
+ writing buffer.
+ `vbuf' specifies the pointer to a buffer into which the value of the corresponding record is
+ written.
+ If successful, the return value is the size of the written data, else, it is -1. -1 is
+ returned when no record corresponds to the specified key or the size of the value of the
+ corresponding record is less than `start'.
+ Note that no additional zero code is appended at the end of the region of the writing buffer. */
+int dpgetwb(DEPOT *depot, const char *kbuf, int ksiz, int start, int max, char *vbuf);
+
+
+/* Get the size of the value of a record.
+ `depot' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is the size of the value of the corresponding record, else,
+ it is -1.
+ Because this function does not read the entity of a record, it is faster than `dpget'. */
+int dpvsiz(DEPOT *depot, const char *kbuf, int ksiz);
+
+
+/* Initialize the iterator of a database handle.
+ `depot' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ The iterator is used in order to access the key of every record stored in a database. */
+int dpiterinit(DEPOT *depot);
+
+
+/* Get the next key of the iterator.
+ `depot' specifies a database handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the next key, else, it is
+ `NULL'. `NULL' is returned when no record is to be get out of the iterator.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if
+ it is no longer in use. It is possible to access every record by iteration of calling
+ this function. However, it is not assured if updating the database is occurred while the
+ iteration. Besides, the order of this traversal access method is arbitrary, so it is not
+ assured that the order of storing matches the one of the traversal access. */
+char *dpiternext(DEPOT *depot, int *sp);
+
+
+/* Set alignment of a database handle.
+ `depot' specifies a database handle connected as a writer.
+ `align' specifies the size of alignment.
+ If successful, the return value is true, else, it is false.
+ If alignment is set to a database, the efficiency of overwriting values is improved.
+ The size of alignment is suggested to be average size of the values of the records to be
+ stored. If alignment is positive, padding whose size is multiple number of the alignment
+ is placed. If alignment is negative, as `vsiz' is the size of a value, the size of padding
+ is calculated with `(vsiz / pow(2, abs(align) - 1))'. Because alignment setting is not
+ saved in a database, you should specify alignment every opening a database. */
+int dpsetalign(DEPOT *depot, int align);
+
+
+/* Set the size of the free block pool of a database handle.
+ `depot' specifies a database handle connected as a writer.
+ `size' specifies the size of the free block pool of a database.
+ If successful, the return value is true, else, it is false.
+ The default size of the free block pool is 16. If the size is greater, the space efficiency
+ of overwriting values is improved with the time efficiency sacrificed. */
+int dpsetfbpsiz(DEPOT *depot, int size);
+
+
+/* Synchronize updating contents with the file and the device.
+ `depot' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ This function is useful when another process uses the connected database file. */
+int dpsync(DEPOT *depot);
+
+
+/* Optimize a database.
+ `depot' specifies a database handle connected as a writer.
+ `bnum' specifies the number of the elements of the bucket array. If it is not more than 0,
+ the default value is specified.
+ If successful, the return value is true, else, it is false.
+ In an alternating succession of deleting and storing with overwrite or concatenate,
+ dispensable regions accumulate. This function is useful to do away with them. */
+int dpoptimize(DEPOT *depot, int bnum);
+
+
+/* Get the name of a database.
+ `depot' specifies a database handle.
+ If successful, the return value is the pointer to the region of the name of the database,
+ else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *dpname(DEPOT *depot);
+
+
+/* Get the size of a database file.
+ `depot' specifies a database handle.
+ If successful, the return value is the size of the database file, else, it is -1. */
+int dpfsiz(DEPOT *depot);
+
+
+/* Get the number of the elements of the bucket array.
+ `depot' specifies a database handle.
+ If successful, the return value is the number of the elements of the bucket array, else, it
+ is -1. */
+int dpbnum(DEPOT *depot);
+
+
+/* Get the number of the used elements of the bucket array.
+ `depot' specifies a database handle.
+ If successful, the return value is the number of the used elements of the bucket array,
+ else, it is -1.
+ This function is inefficient because it accesses all elements of the bucket array. */
+int dpbusenum(DEPOT *depot);
+
+
+/* Get the number of the records stored in a database.
+ `depot' specifies a database handle.
+ If successful, the return value is the number of the records stored in the database, else,
+ it is -1. */
+int dprnum(DEPOT *depot);
+
+
+/* Check whether a database handle is a writer or not.
+ `depot' specifies a database handle.
+ The return value is true if the handle is a writer, false if not. */
+int dpwritable(DEPOT *depot);
+
+
+/* Check whether a database has a fatal error or not.
+ `depot' specifies a database handle.
+ The return value is true if the database has a fatal error, false if not. */
+int dpfatalerror(DEPOT *depot);
+
+
+/* Get the inode number of a database file.
+ `depot' specifies a database handle.
+ The return value is the inode number of the database file. */
+int dpinode(DEPOT *depot);
+
+
+/* Get the last modified time of a database.
+ `depot' specifies a database handle.
+ The return value is the last modified time of the database. */
+time_t dpmtime(DEPOT *depot);
+
+
+/* Get the file descriptor of a database file.
+ `depot' specifies a database handle.
+ The return value is the file descriptor of the database file.
+ Handling the file descriptor of a database file directly is not suggested. */
+int dpfdesc(DEPOT *depot);
+
+
+/* Remove a database file.
+ `name' specifies the name of a database file.
+ If successful, the return value is true, else, it is false. */
+int dpremove(const char *name);
+
+
+/* Repair a broken database file.
+ `name' specifies the name of a database file.
+ If successful, the return value is true, else, it is false.
+ There is no guarantee that all records in a repaired database file correspond to the original
+ or expected state. */
+int dprepair(const char *name);
+
+
+/* Dump all records as endian independent data.
+ `depot' specifies a database handle.
+ `name' specifies the name of an output file.
+ If successful, the return value is true, else, it is false. */
+int dpexportdb(DEPOT *depot, const char *name);
+
+
+/* Load all records from endian independent data.
+ `depot' specifies a database handle connected as a writer. The database of the handle must
+ be empty.
+ `name' specifies the name of an input file.
+ If successful, the return value is true, else, it is false. */
+int dpimportdb(DEPOT *depot, const char *name);
+
+
+/* Retrieve a record directly from a database file.
+ `name' specifies the name of a database file.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the specified key.
+ Because an additional zero code is appended at the end of the region of the return value,
+ the return value can be treated as a character string. Because the region of the return
+ value is allocated with the `malloc' call, it should be released with the `free' call if it
+ is no longer in use. Although this function can be used even while the database file is
+ locked by another process, it is not assured that recent updated is reflected. */
+char *dpsnaffle(const char *name, const char *kbuf, int ksiz, int *sp);
+
+
+/* Hash function used inside Depot.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ The return value is the hash value of 31 bits length computed from the key.
+ This function is useful when an application calculates the state of the inside bucket array. */
+int dpinnerhash(const char *kbuf, int ksiz);
+
+
+/* Hash function which is independent from the hash functions used inside Depot.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ The return value is the hash value of 31 bits length computed from the key.
+ This function is useful when an application uses its own hash algorithm outside Depot. */
+int dpouterhash(const char *kbuf, int ksiz);
+
+
+/* Get a natural prime number not less than a number.
+ `num' specified a natural number.
+ The return value is a natural prime number not less than the specified number.
+ This function is useful when an application determines the size of a bucket array of its
+ own hash algorithm. */
+int dpprimenum(int num);
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+#define _QDBM_VERSION "1.8.77"
+#define _QDBM_LIBVER 1413
+
+
+/* Name of the operating system. */
+MYEXTERN const char *dpsysname;
+
+
+/* File descriptor for debugging output. */
+MYEXTERN int dpdbgfd;
+
+
+/* Whether this build is reentrant. */
+MYEXTERN const int dpisreentrant;
+
+
+/* Set the last happened error code.
+ `ecode' specifies the error code.
+ `line' specifies the number of the line where the error happened. */
+void dpecodeset(int ecode, const char *file, int line);
+
+
+/* Get the pointer of the variable of the last happened error code.
+ The return value is the pointer of the variable. */
+int *dpecodeptr(void);
+
+
+/* Synchronize updating contents on memory.
+ `depot' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. */
+int dpmemsync(DEPOT *depot);
+
+
+/* Synchronize updating contents on memory, not physically.
+ `depot' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. */
+int dpmemflush(DEPOT *depot);
+
+
+/* Get flags of a database.
+ `depot' specifies a database handle.
+ The return value is the flags of a database. */
+int dpgetflags(DEPOT *depot);
+
+
+/* Set flags of a database.
+ `depot' specifies a database handle connected as a writer.
+ `flags' specifies flags to set. Least ten bits are reserved for internal use.
+ If successful, the return value is true, else, it is false. */
+int dpsetflags(DEPOT *depot, int flags);
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/depot.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/dpmgr.c
===================================================================
--- box/trunk/qdbm/dpmgr.c (rev 0)
+++ box/trunk/qdbm/dpmgr.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,916 @@
+/*************************************************************************************************
+ * Utility for debugging Depot and its applications
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define ALIGNSIZ 32 /* basic size of alignment */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *hextoobj(const char *str, int *sp);
+char *dectoiobj(const char *str, int *sp);
+int runcreate(int argc, char **argv);
+int runput(int argc, char **argv);
+int runout(int argc, char **argv);
+int runget(int argc, char **argv);
+int runlist(int argc, char **argv);
+int runoptimize(int argc, char **argv);
+int runinform(int argc, char **argv);
+int runremove(int argc, char **argv);
+int runrepair(int argc, char **argv);
+int runexportdb(int argc, char **argv);
+int runimportdb(int argc, char **argv);
+int runsnaffle(int argc, char **argv);
+void pdperror(const char *name);
+void printobj(const char *obj, int size);
+void printobjhex(const char *obj, int size);
+int docreate(const char *name, int bnum, int sparse);
+int doput(const char *name, const char *kbuf, int ksiz, const char *vbuf, int vsiz,
+ int dmode, int align);
+int doout(const char *name, const char *kbuf, int ksiz);
+int doget(const char *name, int opts, const char *kbuf, int ksiz,
+ int start, int max, int ox, int nb);
+int dolist(const char *name, int opts, int kb, int vb, int ox);
+int dooptimize(const char *name, int bnum, int align);
+int doinform(const char *name, int opts);
+int doremove(const char *name);
+int dorepair(const char *name);
+int doexportdb(const char *name, const char *file);
+int doimportdb(const char *name, const char *file, int bnum);
+int dosnaffle(const char *name, const char *kbuf, int ksiz, int ox, int nb);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "create")){
+ rv = runcreate(argc, argv);
+ } else if(!strcmp(argv[1], "put")){
+ rv = runput(argc, argv);
+ } else if(!strcmp(argv[1], "out")){
+ rv = runout(argc, argv);
+ } else if(!strcmp(argv[1], "get")){
+ rv = runget(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else if(!strcmp(argv[1], "optimize")){
+ rv = runoptimize(argc, argv);
+ } else if(!strcmp(argv[1], "inform")){
+ rv = runinform(argc, argv);
+ } else if(!strcmp(argv[1], "remove")){
+ rv = runremove(argc, argv);
+ } else if(!strcmp(argv[1], "repair")){
+ rv = runrepair(argc, argv);
+ } else if(!strcmp(argv[1], "exportdb")){
+ rv = runexportdb(argc, argv);
+ } else if(!strcmp(argv[1], "importdb")){
+ rv = runimportdb(argc, argv);
+ } else if(!strcmp(argv[1], "snaffle")){
+ rv = runsnaffle(argc, argv);
+ } else if(!strcmp(argv[1], "version") || !strcmp(argv[1], "--version")){
+ printf("Powered by QDBM version %s on %s%s\n",
+ dpversion, dpsysname, dpisreentrant ? " (reentrant)" : "");
+ printf("Copyright (c) 2000-2007 Mikio Hirabayashi\n");
+ rv = 0;
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: administration utility for Depot\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s create [-s] [-bnum num] name\n", progname);
+ fprintf(stderr, " %s put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat] [-na] name key val\n",
+ progname);
+ fprintf(stderr, " %s out [-kx|-ki] name key\n", progname);
+ fprintf(stderr, " %s get [-nl] [-kx|-ki] [-start num] [-max num] [-ox] [-n] name key\n",
+ progname);
+ fprintf(stderr, " %s list [-nl] [-k|-v] [-ox] name\n", progname);
+ fprintf(stderr, " %s optimize [-bnum num] [-na] name\n", progname);
+ fprintf(stderr, " %s inform [-nl] name\n", progname);
+ fprintf(stderr, " %s remove name\n", progname);
+ fprintf(stderr, " %s repair name\n", progname);
+ fprintf(stderr, " %s exportdb name file\n", progname);
+ fprintf(stderr, " %s importdb [-bnum num] name file\n", progname);
+ fprintf(stderr, " %s snaffle [-kx|-ki] [-ox] [-n] name key\n", progname);
+ fprintf(stderr, " %s version\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* create a binary object from a hexadecimal string */
+char *hextoobj(const char *str, int *sp){
+ char *buf, mbuf[3];
+ int len, i, j;
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ j = 0;
+ for(i = 0; i < len; i += 2){
+ while(strchr(" \n\r\t\f\v", str[i])){
+ i++;
+ }
+ if((mbuf[0] = str[i]) == '\0') break;
+ if((mbuf[1] = str[i+1]) == '\0') break;
+ mbuf[2] = '\0';
+ buf[j++] = (char)strtol(mbuf, NULL, 16);
+ }
+ buf[j] = '\0';
+ *sp = j;
+ return buf;
+}
+
+
+/* create a integer object from a decimal string */
+char *dectoiobj(const char *str, int *sp){
+ char *buf;
+ int num;
+ num = atoi(str);
+ if(!(buf = malloc(sizeof(int)))) return NULL;
+ *(int *)buf = num;
+ *sp = sizeof(int);
+ return buf;
+}
+
+
+/* parse arguments of create command */
+int runcreate(int argc, char **argv){
+ char *name;
+ int i, sb, bnum, rv;
+ name = NULL;
+ sb = FALSE;
+ bnum = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docreate(name, bnum, sb);
+ return rv;
+}
+
+
+/* parse arguments of put command */
+int runput(int argc, char **argv){
+ char *name, *key, *val, *kbuf, *vbuf;
+ int i, kx, ki, vx, vi, vf, align, ksiz, vsiz, rv;
+ int dmode;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ vx = FALSE;
+ vi = FALSE;
+ vf = FALSE;
+ align = ALIGNSIZ;
+ key = NULL;
+ val = NULL;
+ dmode = DP_DOVER;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-vx")){
+ vx = TRUE;
+ } else if(!strcmp(argv[i], "-vi")){
+ vi = TRUE;
+ } else if(!strcmp(argv[i], "-vf")){
+ vf = TRUE;
+ } else if(!strcmp(argv[i], "-keep")){
+ dmode = DP_DKEEP;
+ } else if(!strcmp(argv[i], "-cat")){
+ dmode = DP_DCAT;
+ } else if(!strcmp(argv[i], "-na")){
+ align = 0;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else if(!val){
+ val = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || !val) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(vx){
+ vbuf = hextoobj(val, &vsiz);
+ } else if(vi){
+ vbuf = dectoiobj(val, &vsiz);
+ } else if(vf){
+ vbuf = cbreadfile(val, &vsiz);
+ } else {
+ vbuf = cbmemdup(val, -1);
+ vsiz = -1;
+ }
+ if(kbuf && vbuf){
+ rv = doput(name, kbuf, ksiz, vbuf, vsiz, dmode, align);
+ } else {
+ if(vf){
+ fprintf(stderr, "%s: %s: cannot read\n", progname, val);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ }
+ rv = 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ return rv;
+}
+
+
+/* parse arguments of out command */
+int runout(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ki, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = doout(name, kbuf, ksiz);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of get command */
+int runget(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, opts, kx, ki, ox, nb, start, max, ksiz, rv;
+ name = NULL;
+ opts = 0;
+ kx = FALSE;
+ ki = FALSE;
+ ox = FALSE;
+ nb = FALSE;
+ start = 0;
+ max = -1;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= DP_ONOLCK;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else if(!strcmp(argv[i], "-start")){
+ if(++i >= argc) usage();
+ start = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-max")){
+ if(++i >= argc) usage();
+ max = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || start < 0) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = doget(name, opts, kbuf, ksiz, start, max, ox, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ char *name;
+ int i, opts, kb, vb, ox, rv;
+ name = NULL;
+ opts = 0;
+ kb = FALSE;
+ vb = FALSE;
+ ox = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= DP_ONOLCK;
+ } else if(!strcmp(argv[i], "-k")){
+ kb = TRUE;
+ } else if(!strcmp(argv[i], "-v")){
+ vb = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dolist(name, opts, kb, vb, ox);
+ return rv;
+}
+
+
+/* parse arguments of optimize command */
+int runoptimize(int argc, char **argv){
+ char *name;
+ int i, bnum, align, rv;
+ name = NULL;
+ bnum = -1;
+ align = ALIGNSIZ;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-na")){
+ align = 0;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dooptimize(name, bnum, align);
+ return rv;
+}
+
+
+/* parse arguments of inform command */
+int runinform(int argc, char **argv){
+ char *name;
+ int i, opts, rv;
+ name = NULL;
+ opts = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= DP_ONOLCK;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doinform(name, opts);
+ return rv;
+}
+
+
+/* parse arguments of remove command */
+int runremove(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doremove(name);
+ return rv;
+}
+
+
+/* parse arguments of repair command */
+int runrepair(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dorepair(name);
+ return rv;
+}
+
+
+/* parse arguments of exportdb command */
+int runexportdb(int argc, char **argv){
+ char *name, *file;
+ int i, rv;
+ name = NULL;
+ file = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !file) usage();
+ rv = doexportdb(name, file);
+ return rv;
+}
+
+
+/* parse arguments of importdb command */
+int runimportdb(int argc, char **argv){
+ char *name, *file;
+ int i, bnum, rv;
+ name = NULL;
+ file = NULL;
+ bnum = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !file) usage();
+ rv = doimportdb(name, file, bnum);
+ return rv;
+}
+
+
+/* parse arguments of snaffle command */
+int runsnaffle(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ki, ox, nb, start, max, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ ox = FALSE;
+ nb = FALSE;
+ start = 0;
+ max = -1;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || start < 0) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = dosnaffle(name, kbuf, ksiz, ox, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* print an object */
+void printobj(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ putchar(obj[i]);
+ }
+}
+
+
+/* print an object as a hexadecimal string */
+void printobjhex(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ printf("%s%02X", i > 0 ? " " : "", ((const unsigned char *)obj)[i]);
+ }
+}
+
+
+/* perform create command */
+int docreate(const char *name, int bnum, int sparse){
+ DEPOT *depot;
+ int omode;
+ omode = DP_OWRITER | DP_OCREAT | DP_OTRUNC | (sparse ? DP_OSPARSE : 0);
+ if(!(depot = dpopen(name, omode, bnum))){
+ pdperror(name);
+ return 1;
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform put command */
+int doput(const char *name, const char *kbuf, int ksiz, const char *vbuf, int vsiz,
+ int dmode, int align){
+ DEPOT *depot;
+ if(!(depot = dpopen(name, DP_OWRITER, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(align > 0 && !dpsetalign(depot, align)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpput(depot, kbuf, ksiz, vbuf, vsiz, dmode)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform out command */
+int doout(const char *name, const char *kbuf, int ksiz){
+ DEPOT *depot;
+ if(!(depot = dpopen(name, DP_OWRITER, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(!dpout(depot, kbuf, ksiz)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform get command */
+int doget(const char *name, int opts, const char *kbuf, int ksiz,
+ int start, int max, int ox, int nb){
+ DEPOT *depot;
+ char *vbuf;
+ int vsiz;
+ if(!(depot = dpopen(name, DP_OREADER | opts, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(!(vbuf = dpget(depot, kbuf, ksiz, start, max, &vsiz))){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(ox){
+ printobjhex(vbuf, vsiz);
+ } else {
+ printobj(vbuf, vsiz);
+ }
+ free(vbuf);
+ if(!nb) putchar('\n');
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform list command */
+int dolist(const char *name, int opts, int kb, int vb, int ox){
+ DEPOT *depot;
+ char *kbuf, *vbuf;
+ int ksiz, vsiz;
+ if(!(depot = dpopen(name, DP_OREADER | opts, -1))){
+ pdperror(name);
+ return 1;
+ }
+ dpiterinit(depot);
+ while((kbuf = dpiternext(depot, &ksiz)) != NULL){
+ if(!(vbuf = dpget(depot, kbuf, ksiz, 0, -1, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ break;
+ }
+ if(ox){
+ if(!vb) printobjhex(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobjhex(vbuf, vsiz);
+ } else {
+ if(!vb) printobj(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobj(vbuf, vsiz);
+ }
+ putchar('\n');
+ free(vbuf);
+ free(kbuf);
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform optimize command */
+int dooptimize(const char *name, int bnum, int align){
+ DEPOT *depot;
+ if(!(depot = dpopen(name, DP_OWRITER, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(align > 0 && !dpsetalign(depot, align)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpoptimize(depot, bnum)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform inform command */
+int doinform(const char *name, int opts){
+ DEPOT *depot;
+ char *tmp;
+ if(!(depot = dpopen(name, DP_OREADER | opts, -1))){
+ pdperror(name);
+ return 1;
+ }
+ tmp = dpname(depot);
+ printf("name: %s\n", tmp ? tmp : "(null)");
+ free(tmp);
+ printf("file size: %d\n", dpfsiz(depot));
+ printf("all buckets: %d\n", dpbnum(depot));
+ printf("used buckets: %d\n", dpbusenum(depot));
+ printf("records: %d\n", dprnum(depot));
+ printf("inode number: %d\n", dpinode(depot));
+ printf("modified time: %.0f\n", (double)dpmtime(depot));
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform remove command */
+int doremove(const char *name){
+ if(!dpremove(name)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform repair command */
+int dorepair(const char *name){
+ if(!dprepair(name)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform exportdb command */
+int doexportdb(const char *name, const char *file){
+ DEPOT *depot;
+ if(!(depot = dpopen(name, DP_OREADER, -1))){
+ pdperror(name);
+ return 1;
+ }
+ if(!dpexportdb(depot, file)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform importdb command */
+int doimportdb(const char *name, const char *file, int bnum){
+ DEPOT *depot;
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, bnum))){
+ pdperror(name);
+ return 1;
+ }
+ if(!dpimportdb(depot, file)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform snaffle command */
+int dosnaffle(const char *name, const char *kbuf, int ksiz, int ox, int nb){
+ char *vbuf;
+ int vsiz;
+ if(!(vbuf = dpsnaffle(name, kbuf, ksiz, &vsiz))){
+ pdperror(name);
+ return 1;
+ }
+ if(ox){
+ printobjhex(vbuf, vsiz);
+ } else {
+ printobj(vbuf, vsiz);
+ }
+ free(vbuf);
+ if(!nb) putchar('\n');
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/dpmgr.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/dptest.c
===================================================================
--- box/trunk/qdbm/dptest.c (rev 0)
+++ box/trunk/qdbm/dptest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,836 @@
+/*************************************************************************************************
+ * Test cases of Depot
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+#include <time.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define RECBUFSIZ 32 /* buffer for records */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runwrite(int argc, char **argv);
+int runread(int argc, char **argv);
+int runrcat(int argc, char **argv);
+int runcombo(int argc, char **argv);
+int runwicked(int argc, char **argv);
+int printfflush(const char *format, ...);
+void pdperror(const char *name);
+int myrand(void);
+int dowrite(const char *name, int rnum, int bnum, int sparse);
+int doread(const char *name, int wb);
+int dorcat(const char *name, int rnum, int bnum, int pnum, int align, int fbpsiz, int cb);
+int docombo(const char *name);
+int dowicked(const char *name, int rnum, int cb);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "write")){
+ rv = runwrite(argc, argv);
+ } else if(!strcmp(argv[1], "read")){
+ rv = runread(argc, argv);
+ } else if(!strcmp(argv[1], "rcat")){
+ rv = runrcat(argc, argv);
+ } else if(!strcmp(argv[1], "combo")){
+ rv = runcombo(argc, argv);
+ } else if(!strcmp(argv[1], "wicked")){
+ rv = runwicked(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Depot\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s write [-s] name rnum bnum\n", progname);
+ fprintf(stderr, " %s read [-wb] name\n", progname);
+ fprintf(stderr, " %s rcat [-c] name rnum bnum pnum align fbpsiz\n", progname);
+ fprintf(stderr, " %s combo name\n", progname);
+ fprintf(stderr, " %s wicked [-c] name rnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of write command */
+int runwrite(int argc, char **argv){
+ char *name, *rstr, *bstr;
+ int i, rnum, bnum, sb, rv;
+ name = NULL;
+ rstr = NULL;
+ bstr = NULL;
+ rnum = 0;
+ bnum = 0;
+ sb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!bstr){
+ bstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr || !bstr) usage();
+ rnum = atoi(rstr);
+ bnum = atoi(bstr);
+ if(rnum < 1 || bnum < 1) usage();
+ rv = dowrite(name, rnum, bnum, sb);
+ return rv;
+}
+
+
+/* parse arguments of read command */
+int runread(int argc, char **argv){
+ char *name;
+ int i, wb, rv;
+ name = NULL;
+ wb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-wb")){
+ wb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doread(name, wb);
+ return rv;
+}
+
+
+/* parse arguments of rcat command */
+int runrcat(int argc, char **argv){
+ char *name, *rstr, *bstr, *pstr, *astr, *fstr;
+ int i, rnum, bnum, pnum, align, fbpsiz, cb, rv;
+ name = NULL;
+ rstr = NULL;
+ bstr = NULL;
+ pstr = NULL;
+ astr = NULL;
+ fstr = NULL;
+ rnum = 0;
+ bnum = 0;
+ pnum = 0;
+ align = 0;
+ fbpsiz = 0;
+ cb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-c")){
+ cb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!bstr){
+ bstr = argv[i];
+ } else if(!pstr){
+ pstr = argv[i];
+ } else if(!astr){
+ astr = argv[i];
+ } else if(!fstr){
+ fstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr || !bstr || !pstr || !astr || !fstr) usage();
+ rnum = atoi(rstr);
+ bnum = atoi(bstr);
+ pnum = atoi(pstr);
+ align = atoi(astr);
+ fbpsiz= atoi(fstr);
+ if(rnum < 1 || bnum < 1 || pnum < 1 || fbpsiz < 0) usage();
+ rv = dorcat(name, rnum, bnum, pnum, align, fbpsiz, cb);
+ return rv;
+}
+
+
+/* parse arguments of combo command */
+int runcombo(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docombo(name);
+ return rv;
+}
+
+
+/* parse arguments of wicked command */
+int runwicked(int argc, char **argv){
+ char *name, *rstr;
+ int i, rnum, cb, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ cb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-c")){
+ cb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowicked(name, rnum, cb);
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* pseudo random number generator */
+int myrand(void){
+ static int cnt = 0;
+ if(cnt == 0) srand(time(NULL));
+ return (rand() * rand() + (rand() >> (sizeof(int) * 4)) + (cnt++)) & INT_MAX;
+}
+
+
+/* perform write command */
+int dowrite(const char *name, int rnum, int bnum, int sparse){
+ DEPOT *depot;
+ int i, omode, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Writing Test>\n name=%s rnum=%d bnum=%d s=%d\n\n", name, rnum, bnum, sparse);
+ /* open a database */
+ omode = DP_OWRITER | DP_OCREAT | DP_OTRUNC | (sparse ? DP_OSPARSE : 0);
+ if(!(depot = dpopen(name, omode, bnum))){
+ pdperror(name);
+ return 1;
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* store a record */
+ len = sprintf(buf, "%08d", i);
+ if(!dpput(depot, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform read command */
+int doread(const char *name, int wb){
+ DEPOT *depot;
+ int i, rnum, err, len;
+ char buf[RECBUFSIZ], vbuf[RECBUFSIZ], *val;
+ printfflush("<Reading Test>\n name=%s wb=%d\n\n", name, wb);
+ /* open a database */
+ if(!(depot = dpopen(name, DP_OREADER, -1))){
+ pdperror(name);
+ return 1;
+ }
+ /* get the number of records */
+ rnum = dprnum(depot);
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* retrieve a record */
+ len = sprintf(buf, "%08d", i);
+ if(wb){
+ if(dpgetwb(depot, buf, len, 0, RECBUFSIZ, vbuf) == -1){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ if(!(val = dpget(depot, buf, len, 0, -1, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform rcat command */
+int dorcat(const char *name, int rnum, int bnum, int pnum, int align, int fbpsiz, int cb){
+ DEPOT *depot;
+ CBMAP *map;
+ int i, err, len, ksiz, vsiz, rsiz;
+ const char *kbuf, *vbuf;
+ char buf[RECBUFSIZ], *rbuf;
+ printfflush("<Random Writing Test>\n name=%s rnum=%d bnum=%d pnum=%d align=%d"
+ " fbpsiz=%d c=%d\n\n", name, rnum, bnum, pnum, align, fbpsiz, cb);
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, bnum))){
+ pdperror(name);
+ return 1;
+ }
+ if(!dpsetalign(depot, align) || !dpsetfbpsiz(depot, fbpsiz)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ map = NULL;
+ if(cb) map = cbmapopen();
+ err = FALSE;
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", myrand() % pnum + 1);
+ if(!dpput(depot, buf, len, buf, len, DP_DCAT)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(map) cbmapputcat(map, buf, len, buf, len);
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d: fsiz=%d rnum=%d)\n", i, dpfsiz(depot), dprnum(depot));
+ }
+ }
+ }
+ if(map){
+ printfflush("Matching records ... ");
+ cbmapiterinit(map);
+ while((kbuf = cbmapiternext(map, &ksiz)) != NULL){
+ vbuf = cbmapget(map, kbuf, ksiz, &vsiz);
+ if(!(rbuf = dpget(depot, kbuf, ksiz, 0, -1, &rsiz))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(rsiz != vsiz || memcmp(rbuf, vbuf, rsiz)){
+ fprintf(stderr, "%s: %s: unmatched record\n", progname, name);
+ free(rbuf);
+ err = TRUE;
+ break;
+ }
+ free(rbuf);
+ }
+ cbmapclose(map);
+ if(!err) printfflush("ok\n");
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform combo command */
+int docombo(const char *name){
+ DEPOT *depot;
+ char buf[RECBUFSIZ], wbuf[RECBUFSIZ], *vbuf;
+ int i, len, wlen, vsiz;
+ printfflush("<Combination Test>\n name=%s\n\n", name);
+ printfflush("Creating a database with bnum 3 ... ");
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, 3))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Setting alignment as 16 ... ");
+ if(!dpsetalign(depot, 16)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 20 records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!dpput(depot, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = dpget(depot, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != dpvsiz(depot, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 10 records without moving rooms ... ");
+ for(i = 1; i <= 10; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!dpput(depot, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 5 records with moving rooms ... ");
+ for(i = 1; i <= 5; i++){
+ len = sprintf(buf, "%08d", i);
+ wlen = sprintf(wbuf, "%024d", i);
+ if(!dpput(depot, buf, len, wbuf, wlen, DP_DOVER)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 15 records in concatenation with moving rooms ... ");
+ for(i = 1; i <= 15; i++){
+ len = sprintf(buf, "%08d", i);
+ wlen = sprintf(wbuf, "========================");
+ if(!dpput(depot, buf, len, wbuf, wlen, DP_DCAT)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = dpget(depot, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != dpvsiz(depot, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting top 10 records ... ");
+ for(i = 1; i <= 10; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!dpout(depot, buf, len)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking deleted records ... ");
+ for(i = 1; i <= 10; i++){
+ len = sprintf(buf, "%08d", i);
+ vbuf = dpget(depot, buf, len, 0, -1, &vsiz);
+ free(vbuf);
+ if(vbuf || dpecode != DP_ENOITEM){
+ fprintf(stderr, "%s: %s: deleting failed\n", progname, name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting top 15 records in concatenation with moving rooms ... ");
+ for(i = 1; i <= 15; i++){
+ len = sprintf(buf, "%08d", i);
+ wlen = sprintf(wbuf, "========================");
+ if(!dpput(depot, buf, len, wbuf, wlen, DP_DCAT)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = dpget(depot, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != dpvsiz(depot, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Optimizing the database ... ");
+ if(!dpoptimize(depot, -1)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 20; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = dpget(depot, buf, len, 0, -1, &vsiz))){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != dpvsiz(depot, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Creating a database with bnum 1000000 ... ");
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, 1000000))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 1000 records ... ");
+ for(i = 1; i <= 1000; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!dpput(depot, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Adding 64 records ... ");
+ for(i = 1; i <= 1000; i++){
+ len = sprintf(buf, "%o", i);
+ if(!dpput(depot, buf, len, buf, len, DP_DOVER)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Syncing the database ... ");
+ if(!dpsync(depot)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Retrieving records directly ... ");
+ for(i = 1; i <= 64; i++){
+ len = sprintf(buf, "%o", i);
+ if(!(vbuf = dpsnaffle(name, buf, len, &vsiz))){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ if(strcmp(vbuf, buf)){
+ fprintf(stderr, "%s: %s: invalid content\n", progname, name);
+ free(vbuf);
+ dpclose(depot);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != dpvsiz(depot, buf, len)){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ dpclose(depot);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Optimizing the database ... ");
+ if(!dpoptimize(depot, -1)){
+ pdperror(name);
+ dpclose(depot);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("all ok\n\n");
+ return 0;
+}
+
+
+/* perform wicked command */
+int dowicked(const char *name, int rnum, int cb){
+ DEPOT *depot;
+ CBMAP *map;
+ int i, len, err, align, mksiz, mvsiz, rsiz;
+ const char *mkbuf, *mvbuf;
+ char buf[RECBUFSIZ], vbuf[RECBUFSIZ], *val;
+ printfflush("<Wicked Writing Test>\n name=%s rnum=%d\n\n", name, rnum);
+ err = FALSE;
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, rnum / 10))){
+ pdperror(name);
+ return 1;
+ }
+ if(!dpsetalign(depot, 16) || !dpsetfbpsiz(depot, 256)){
+ pdperror(name);
+ err = TRUE;
+ }
+ map = NULL;
+ if(cb) map = cbmapopen();
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", myrand() % rnum + 1);
+ switch(myrand() % 16){
+ case 0:
+ putchar('O');
+ if(!dpput(depot, buf, len, buf, len, DP_DOVER)) err = TRUE;
+ if(map) cbmapput(map, buf, len, buf, len, TRUE);
+ break;
+ case 1:
+ putchar('K');
+ if(!dpput(depot, buf, len, buf, len, DP_DKEEP) && dpecode != DP_EKEEP) err = TRUE;
+ if(map) cbmapput(map, buf, len, buf, len, FALSE);
+ break;
+ case 2:
+ putchar('D');
+ if(!dpout(depot, buf, len) && dpecode != DP_ENOITEM) err = TRUE;
+ if(map) cbmapout(map, buf, len);
+ break;
+ case 3:
+ putchar('G');
+ if(dpgetwb(depot, buf, len, 2, RECBUFSIZ, vbuf) == -1 && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 4:
+ putchar('V');
+ if(dpvsiz(depot, buf, len) == -1 && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ default:
+ putchar('C');
+ if(!dpput(depot, buf, len, buf, len, DP_DCAT)) err = TRUE;
+ if(map) cbmapputcat(map, buf, len, buf, len);
+ break;
+ }
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ if(!err && rnum > 100 && myrand() % (rnum / 100) == 0){
+ if(myrand() % 10 == 0){
+ align = (myrand() % 4 + 1) * -1;
+ } else {
+ align = myrand() % 32;
+ }
+ if(!dpsetalign(depot, align)) err = TRUE;
+ }
+ if(err){
+ pdperror(name);
+ break;
+ }
+ }
+ if(!dpoptimize(depot, -1)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!dpput(depot, buf, len, ":", -1, DP_DCAT)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(map) cbmapputcat(map, buf, len, ":", -1);
+ putchar(':');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!dpoptimize(depot, -1)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(val = dpget(depot, buf, len, 0, -1, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ putchar('=');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!dpiterinit(depot)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= rnum; i++){
+ if(!(val = dpiternext(depot, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ putchar('@');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(map){
+ printfflush("Matching records ... ");
+ cbmapiterinit(map);
+ while((mkbuf = cbmapiternext(map, &mksiz)) != NULL){
+ mvbuf = cbmapget(map, mkbuf, mksiz, &mvsiz);
+ if(!(val = dpget(depot, mkbuf, mksiz, 0, -1, &rsiz))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(rsiz != mvsiz || memcmp(val, mvbuf, rsiz)){
+ fprintf(stderr, "%s: %s: unmatched record\n", progname, name);
+ free(val);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ }
+ cbmapclose(map);
+ if(!err) printfflush("ok\n");
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/dptest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/dptsv.c
===================================================================
--- box/trunk/qdbm/dptsv.c (rev 0)
+++ box/trunk/qdbm/dptsv.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,261 @@
+/*************************************************************************************************
+ * Mutual converter between a database of Depot and a TSV text
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runimport(int argc, char **argv);
+int runexport(int argc, char **argv);
+void pdperror(const char *name);
+char *getl(void);
+int doimport(const char *name, int bnum, int bin);
+int doexport(const char *name, int bin);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "import")){
+ rv = runimport(argc, argv);
+ } else if(!strcmp(argv[1], "export")){
+ rv = runexport(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: mutual converter between TSV and Depot database\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s import [-bnum num] [-bin] name\n", progname);
+ fprintf(stderr, " %s export [-bin] name\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of import command */
+int runimport(int argc, char **argv){
+ char *name;
+ int i, bnum, bin, rv;
+ name = NULL;
+ bnum = -1;
+ bin = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bnum")){
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-bin")){
+ bin = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doimport(name, bnum, bin);
+ return rv;
+}
+
+
+/* parse arguments of export command */
+int runexport(int argc, char **argv){
+ char *name;
+ int i, bin, rv;
+ name = NULL;
+ bin = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bin")){
+ bin = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doexport(name, bin);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* read a line */
+char *getl(void){
+ char *buf;
+ int c, len, blen;
+ buf = NULL;
+ len = 0;
+ blen = 256;
+ while((c = getchar()) != EOF){
+ if(blen <= len) blen *= 2;
+ buf = cbrealloc(buf, blen + 1);
+ if(c == '\n') c = '\0';
+ buf[len++] = c;
+ if(c == '\0') break;
+ }
+ if(!buf) return NULL;
+ buf[len] = '\0';
+ return buf;
+}
+
+
+/* perform import command */
+int doimport(const char *name, int bnum, int bin){
+ DEPOT *depot;
+ char *buf, *kbuf, *vbuf, *ktmp, *vtmp;
+ int i, err, ktsiz, vtsiz;
+ /* open a database */
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT, bnum))){
+ pdperror(name);
+ return 1;
+ }
+ /* loop for each line */
+ err = FALSE;
+ for(i = 1; (buf = getl()) != NULL; i++){
+ kbuf = buf;
+ if((vbuf = strchr(buf, '\t')) != NULL){
+ *vbuf = '\0';
+ vbuf++;
+ /* store a record */
+ if(bin){
+ ktmp = cbbasedecode(kbuf, &ktsiz);
+ vtmp = cbbasedecode(vbuf, &vtsiz);
+ if(!dpput(depot, ktmp, ktsiz, vtmp, vtsiz, DP_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ }
+ free(vtmp);
+ free(ktmp);
+ } else {
+ if(!dpput(depot, kbuf, -1, vbuf, -1, DP_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ }
+ }
+ } else {
+ fprintf(stderr, "%s: %s: invalid format in line %d\n", progname, name, i);
+ }
+ free(buf);
+ if(err) break;
+ }
+ /* close the database */
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return err ? 1 : 0;
+}
+
+
+/* perform export command */
+int doexport(const char *name, int bin){
+ DEPOT *depot;
+ char *kbuf, *vbuf, *tmp;
+ int err, ksiz, vsiz;
+ /* open a database */
+ if(!(depot = dpopen(name, DP_OREADER, -1))){
+ pdperror(name);
+ return 1;
+ }
+ /* initialize the iterator */
+ dpiterinit(depot);
+ /* loop for each key */
+ err = FALSE;
+ while((kbuf = dpiternext(depot, &ksiz)) != NULL){
+ /* retrieve a value with a key */
+ if(!(vbuf = dpget(depot, kbuf, ksiz, 0, -1, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ err = TRUE;
+ break;
+ }
+ /* output data */
+ if(bin){
+ tmp = cbbaseencode(kbuf, ksiz);
+ printf("%s\t", tmp);
+ free(tmp);
+ tmp = cbbaseencode(vbuf, vsiz);
+ printf("%s\n", tmp);
+ free(tmp);
+ } else {
+ printf("%s\t%s\n", kbuf, vbuf);
+ }
+ /* free resources */
+ free(vbuf);
+ free(kbuf);
+ }
+ /* check whether all records were retrieved */
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ /* close the database */
+ if(!dpclose(depot)){
+ pdperror(name);
+ return 1;
+ }
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/dptsv.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/hovel.c
===================================================================
--- box/trunk/qdbm/hovel.c (rev 0)
+++ box/trunk/qdbm/hovel.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,568 @@
+/*************************************************************************************************
+ * Implementation of Hovel
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "hovel.h"
+#include "myconf.h"
+
+#define HV_INITBNUM 32749 /* initial bucket number */
+#define HV_ALIGNSIZ 16 /* size of alignment */
+
+
+/* private function prototypes */
+static int gdbm_geterrno(int ecode);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* String containing the version information. */
+char *gdbm_version = "Powered by QDBM";
+
+
+/* Get a message string corresponding to an error code. */
+char *gdbm_strerror(gdbm_error gdbmerrno){
+ switch(gdbmerrno){
+ case GDBM_NO_ERROR: return "No error";
+ case GDBM_MALLOC_ERROR: return "Malloc error";
+ case GDBM_BLOCK_SIZE_ERROR: return "Block size error";
+ case GDBM_FILE_OPEN_ERROR: return "File open error";
+ case GDBM_FILE_WRITE_ERROR: return "File write error";
+ case GDBM_FILE_SEEK_ERROR: return "File seek error";
+ case GDBM_FILE_READ_ERROR: return "File read error";
+ case GDBM_BAD_MAGIC_NUMBER: return "Bad magic number";
+ case GDBM_EMPTY_DATABASE: return "Empty database";
+ case GDBM_CANT_BE_READER: return "Can't be reader";
+ case GDBM_CANT_BE_WRITER: return "Can't be writer";
+ case GDBM_READER_CANT_DELETE: return "Reader can't delete";
+ case GDBM_READER_CANT_STORE: return "Reader can't store";
+ case GDBM_READER_CANT_REORGANIZE: return "Reader can't reorganize";
+ case GDBM_UNKNOWN_UPDATE: return "Unknown update";
+ case GDBM_ITEM_NOT_FOUND: return "Item not found";
+ case GDBM_REORGANIZE_FAILED: return "Reorganize failed";
+ case GDBM_CANNOT_REPLACE: return "Cannot replace";
+ case GDBM_ILLEGAL_DATA: return "Illegal data";
+ case GDBM_OPT_ALREADY_SET: return "Option already set";
+ case GDBM_OPT_ILLEGAL: return "Illegal option";
+ }
+ return "(invalid ecode)";
+}
+
+
+/* Get a database handle after the fashion of GDBM. */
+GDBM_FILE gdbm_open(char *name, int block_size, int read_write, int mode,
+ void (*fatal_func)(void)){
+ GDBM_FILE dbf;
+ int dpomode;
+ DEPOT *depot;
+ int flags, fd;
+ assert(name);
+ dpomode = DP_OREADER;
+ flags = O_RDONLY;
+ if(read_write & GDBM_READER){
+ dpomode = GDBM_READER;
+ if(read_write & GDBM_NOLOCK) dpomode |= DP_ONOLCK;
+ if(read_write & GDBM_LOCKNB) dpomode |= DP_OLCKNB;
+ flags = O_RDONLY;
+ } else if(read_write & GDBM_WRITER){
+ dpomode = DP_OWRITER;
+ if(read_write & GDBM_NOLOCK) dpomode |= DP_ONOLCK;
+ if(read_write & GDBM_LOCKNB) dpomode |= DP_OLCKNB;
+ flags = O_RDWR;
+ } else if(read_write & GDBM_WRCREAT){
+ dpomode = DP_OWRITER | DP_OCREAT;
+ if(read_write & GDBM_NOLOCK) dpomode |= DP_ONOLCK;
+ if(read_write & GDBM_LOCKNB) dpomode |= DP_OLCKNB;
+ if(read_write & GDBM_SPARSE) dpomode |= DP_OSPARSE;
+ flags = O_RDWR | O_CREAT;
+ } else if(read_write & GDBM_NEWDB){
+ dpomode = DP_OWRITER | DP_OCREAT | DP_OTRUNC;
+ if(read_write & GDBM_NOLOCK) dpomode |= DP_ONOLCK;
+ if(read_write & GDBM_LOCKNB) dpomode |= DP_OLCKNB;
+ if(read_write & GDBM_SPARSE) dpomode |= DP_OSPARSE;
+ flags = O_RDWR | O_CREAT | O_TRUNC;
+ } else {
+ gdbm_errno = GDBM_ILLEGAL_DATA;
+ return NULL;
+ }
+ mode |= 00600;
+ if((fd = open(name, flags, mode)) == -1 || close(fd) == -1){
+ gdbm_errno = GDBM_FILE_OPEN_ERROR;
+ return NULL;
+ }
+ if(!(depot = dpopen(name, dpomode, HV_INITBNUM))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ if(dpecode == DP_ESTAT) gdbm_errno = GDBM_FILE_OPEN_ERROR;
+ return NULL;
+ }
+ if(dpomode & DP_OWRITER){
+ if(!dpsetalign(depot, HV_ALIGNSIZ)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ dpclose(depot);
+ }
+ }
+ if((dpomode & DP_OWRITER) && (read_write & GDBM_SYNC)){
+ if(!dpsync(depot)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ dpclose(depot);
+ }
+ }
+ if(!(dbf = malloc(sizeof(GDBM)))){
+ gdbm_errno = GDBM_MALLOC_ERROR;
+ dpclose(depot);
+ return NULL;
+ }
+ dbf->depot = depot;
+ dbf->curia = NULL;
+ dbf->syncmode = (dpomode & DP_OWRITER) && (read_write & GDBM_SYNC) ? TRUE : FALSE;
+ return dbf;
+}
+
+
+/* Get a database handle after the fashion of QDBM. */
+GDBM_FILE gdbm_open2(char *name, int read_write, int mode, int bnum, int dnum, int align){
+ GDBM_FILE dbf;
+ int dpomode, cromode, flags, fd;
+ struct stat sbuf;
+ DEPOT *depot;
+ CURIA *curia;
+ assert(name);
+ dpomode = DP_OREADER;
+ cromode = CR_OREADER;
+ flags = O_RDONLY;
+ if(read_write & GDBM_READER){
+ dpomode = DP_OREADER;
+ cromode = CR_OREADER;
+ if(read_write & GDBM_NOLOCK){
+ dpomode |= DP_ONOLCK;
+ cromode |= CR_ONOLCK;
+ }
+ if(read_write & GDBM_LOCKNB){
+ dpomode |= DP_OLCKNB;
+ cromode |= CR_OLCKNB;
+ }
+ flags = O_RDONLY;
+ } else if(read_write & GDBM_WRITER){
+ dpomode = DP_OWRITER;
+ cromode = CR_OWRITER;
+ if(read_write & GDBM_NOLOCK){
+ dpomode |= DP_ONOLCK;
+ cromode |= CR_ONOLCK;
+ }
+ if(read_write & GDBM_LOCKNB){
+ dpomode |= DP_OLCKNB;
+ cromode |= CR_OLCKNB;
+ }
+ flags = O_RDWR;
+ } else if(read_write & GDBM_WRCREAT){
+ dpomode = DP_OWRITER | DP_OCREAT;
+ cromode = CR_OWRITER | CR_OCREAT;
+ if(read_write & GDBM_NOLOCK){
+ dpomode |= DP_ONOLCK;
+ cromode |= CR_ONOLCK;
+ }
+ if(read_write & GDBM_LOCKNB){
+ dpomode |= DP_OLCKNB;
+ cromode |= CR_OLCKNB;
+ }
+ if(read_write & GDBM_SPARSE){
+ dpomode |= DP_OSPARSE;
+ cromode |= CR_OSPARSE;
+ }
+ flags = O_RDWR | O_CREAT;
+ } else if(read_write & GDBM_NEWDB){
+ dpomode = DP_OWRITER | DP_OCREAT | DP_OTRUNC;
+ cromode = CR_OWRITER | CR_OCREAT | CR_OTRUNC;
+ if(read_write & GDBM_NOLOCK){
+ dpomode |= DP_ONOLCK;
+ cromode |= CR_ONOLCK;
+ }
+ if(read_write & GDBM_LOCKNB){
+ dpomode |= DP_OLCKNB;
+ cromode |= CR_OLCKNB;
+ }
+ if(read_write & GDBM_SPARSE){
+ dpomode |= DP_OSPARSE;
+ cromode |= CR_OSPARSE;
+ }
+ flags = O_RDWR | O_CREAT | O_TRUNC;
+ } else {
+ gdbm_errno = GDBM_ILLEGAL_DATA;
+ return NULL;
+ }
+ if(lstat(name, &sbuf) != -1){
+ if(S_ISDIR(sbuf.st_mode)){
+ if(dnum < 1) dnum = 1;
+ } else {
+ dnum = 0;
+ }
+ }
+ depot = NULL;
+ curia = NULL;
+ if(dnum > 0){
+ mode |= 00700;
+ if((cromode & CR_OCREAT)){
+ if(mkdir(name, mode) == -1 && errno != EEXIST){
+ gdbm_errno = GDBM_FILE_OPEN_ERROR;
+ return NULL;
+ }
+ }
+ if(!(curia = cropen(name, cromode, bnum, dnum))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return NULL;
+ }
+ if(cromode & CR_OWRITER) crsetalign(curia, align);
+ if((cromode & CR_OWRITER) && (read_write & GDBM_SYNC)) crsync(curia);
+ } else {
+ mode |= 00600;
+ if(dpomode & DP_OWRITER){
+ if((fd = open(name, flags, mode)) == -1 || close(fd) == -1){
+ gdbm_errno = GDBM_FILE_OPEN_ERROR;
+ return NULL;
+ }
+ }
+ if(!(depot = dpopen(name, dpomode, bnum))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return NULL;
+ }
+ if(dpomode & DP_OWRITER) dpsetalign(depot, align);
+ if((dpomode & DP_OWRITER) && (read_write & GDBM_SYNC)) dpsync(depot);
+ }
+ if(!(dbf = malloc(sizeof(GDBM)))){
+ gdbm_errno = GDBM_MALLOC_ERROR;
+ if(depot) dpclose(depot);
+ if(curia) crclose(curia);
+ return NULL;
+ }
+ dbf->depot = depot;
+ dbf->curia = curia;
+ dbf->syncmode = (dpomode & DP_OWRITER) && (read_write & GDBM_SYNC) ? TRUE : FALSE;
+ return dbf;
+}
+
+
+/* Close a database handle. */
+void gdbm_close(GDBM_FILE dbf){
+ assert(dbf);
+ if(dbf->depot){
+ if(dbf->syncmode) dpsync(dbf->depot);
+ dpclose(dbf->depot);
+ } else {
+ if(dbf->syncmode) crsync(dbf->curia);
+ crclose(dbf->curia);
+ }
+ free(dbf);
+}
+
+
+/* Store a record. */
+int gdbm_store(GDBM_FILE dbf, datum key, datum content, int flag){
+ int dmode;
+ assert(dbf);
+ if(!key.dptr || key.dsize < 0 || !content.dptr || content.dsize < 0){
+ gdbm_errno = GDBM_ILLEGAL_DATA;
+ return -1;
+ }
+ if(dbf->depot){
+ if(!dpwritable(dbf->depot)){
+ gdbm_errno = GDBM_READER_CANT_STORE;
+ return -1;
+ }
+ dmode = (flag == GDBM_INSERT) ? DP_DKEEP : DP_DOVER;
+ if(!dpput(dbf->depot, key.dptr, key.dsize, content.dptr, content.dsize, dmode)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ if(dpecode == DP_EKEEP) return 1;
+ return -1;
+ }
+ if(dbf->syncmode && !dpsync(dbf->depot)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ } else {
+ if(!crwritable(dbf->curia)){
+ gdbm_errno = GDBM_READER_CANT_STORE;
+ return -1;
+ }
+ dmode = (flag == GDBM_INSERT) ? CR_DKEEP : CR_DOVER;
+ if(!crput(dbf->curia, key.dptr, key.dsize, content.dptr, content.dsize, dmode)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ if(dpecode == DP_EKEEP) return 1;
+ return -1;
+ }
+ if(dbf->syncmode && !crsync(dbf->curia)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ }
+ return 0;
+}
+
+
+/* Delete a record. */
+int gdbm_delete(GDBM_FILE dbf, datum key){
+ assert(dbf);
+ if(!key.dptr || key.dsize < 0){
+ gdbm_errno = GDBM_ILLEGAL_DATA;
+ return -1;
+ }
+ if(dbf->depot){
+ if(!dpwritable(dbf->depot)){
+ gdbm_errno = GDBM_READER_CANT_DELETE;
+ return -1;
+ }
+ if(!dpout(dbf->depot, key.dptr, key.dsize)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ if(dbf->syncmode && !dpsync(dbf->depot)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ } else {
+ if(!crwritable(dbf->curia)){
+ gdbm_errno = GDBM_READER_CANT_DELETE;
+ return -1;
+ }
+ if(!crout(dbf->curia, key.dptr, key.dsize)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ if(dbf->syncmode && !crsync(dbf->curia)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ }
+ return 0;
+}
+
+
+/* Retrieve a record. */
+datum gdbm_fetch(GDBM_FILE dbf, datum key){
+ datum content;
+ char *vbuf;
+ int vsiz;
+ assert(dbf);
+ if(!key.dptr || key.dsize < 0){
+ gdbm_errno = GDBM_ILLEGAL_DATA;
+ content.dptr = NULL;
+ content.dsize = 0;
+ return content;
+ }
+ if(dbf->depot){
+ if(!(vbuf = dpget(dbf->depot, key.dptr, key.dsize, 0, -1, &vsiz))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ content.dptr = NULL;
+ content.dsize = 0;
+ return content;
+ }
+ } else {
+ if(!(vbuf = crget(dbf->curia, key.dptr, key.dsize, 0, -1, &vsiz))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ content.dptr = NULL;
+ content.dsize = 0;
+ return content;
+ }
+ }
+ content.dptr = vbuf;
+ content.dsize = vsiz;
+ return content;
+}
+
+
+/* Check whether a record exists or not. */
+int gdbm_exists(GDBM_FILE dbf, datum key){
+ assert(dbf);
+ if(!key.dptr || key.dsize < 0){
+ gdbm_errno = GDBM_ILLEGAL_DATA;
+ return FALSE;
+ }
+ if(dbf->depot){
+ if(dpvsiz(dbf->depot, key.dptr, key.dsize) == -1){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return FALSE;
+ }
+ } else {
+ if(crvsiz(dbf->curia, key.dptr, key.dsize) == -1){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return FALSE;
+ }
+ }
+ return TRUE;
+}
+
+
+/* Get the first key of a database. */
+datum gdbm_firstkey(GDBM_FILE dbf){
+ datum key;
+ assert(dbf);
+ memset(&key, 0, sizeof(datum));
+ if(dbf->depot){
+ if(dprnum(dbf->depot) < 1){
+ gdbm_errno = GDBM_EMPTY_DATABASE;
+ key.dptr = NULL;
+ key.dsize = 0;
+ return key;
+ }
+ dpiterinit(dbf->depot);
+ return gdbm_nextkey(dbf, key);
+ } else {
+ if(crrnum(dbf->curia) < 1){
+ gdbm_errno = GDBM_EMPTY_DATABASE;
+ key.dptr = NULL;
+ key.dsize = 0;
+ return key;
+ }
+ criterinit(dbf->curia);
+ return gdbm_nextkey(dbf, key);
+ }
+}
+
+
+/* Get the next key of a database. */
+datum gdbm_nextkey(GDBM_FILE dbf, datum key){
+ char *kbuf;
+ int ksiz;
+ assert(dbf);
+ if(dbf->depot){
+ if(!(kbuf = dpiternext(dbf->depot, &ksiz))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ key.dptr = NULL;
+ key.dsize = 0;
+ return key;
+ }
+ } else {
+ if(!(kbuf = criternext(dbf->curia, &ksiz))){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ key.dptr = NULL;
+ key.dsize = 0;
+ return key;
+ }
+ }
+ key.dptr = kbuf;
+ key.dsize = ksiz;
+ return key;
+}
+
+
+/* Synchronize contents of updating with the file and the device. */
+void gdbm_sync(GDBM_FILE dbf){
+ assert(dbf);
+ if(dbf->depot){
+ if(!dpsync(dbf->depot)) gdbm_errno = gdbm_geterrno(dpecode);
+ } else {
+ if(!crsync(dbf->curia)) gdbm_errno = gdbm_geterrno(dpecode);
+ }
+}
+
+
+/* Reorganize a database. */
+int gdbm_reorganize(GDBM_FILE dbf){
+ assert(dbf);
+ if(dbf->depot){
+ if(!dpwritable(dbf->depot)){
+ gdbm_errno = GDBM_READER_CANT_REORGANIZE;
+ return -1;
+ }
+ if(!dpoptimize(dbf->depot, dprnum(dbf->depot) >= HV_INITBNUM ? -1 : HV_INITBNUM)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ } else {
+ if(!crwritable(dbf->curia)){
+ gdbm_errno = GDBM_READER_CANT_REORGANIZE;
+ return -1;
+ }
+ if(!croptimize(dbf->curia, crrnum(dbf->curia) >= HV_INITBNUM ? -1 : HV_INITBNUM)){
+ gdbm_errno = gdbm_geterrno(dpecode);
+ return -1;
+ }
+ }
+ return 0;
+}
+
+
+/* Get the file descriptor of a database file. */
+int gdbm_fdesc(GDBM_FILE dbf){
+ assert(dbf);
+ if(dbf->depot){
+ return dpfdesc(dbf->depot);
+ } else {
+ return -1;
+ }
+}
+
+
+/* No effect. */
+int gdbm_setopt(GDBM_FILE dbf, int option, int *value, int size){
+ assert(dbf);
+ return 0;
+}
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Get the pointer of the last happened error code. */
+int *gdbm_errnoptr(void){
+ static int deferrno = GDBM_NO_ERROR;
+ void *ptr;
+ if(_qdbm_ptsafe){
+ if(!(ptr = _qdbm_settsd(&deferrno, sizeof(int), &deferrno))){
+ deferrno = GDBM_ILLEGAL_DATA;
+ return &deferrno;
+ }
+ return (int *)ptr;
+ }
+ return &deferrno;
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Get the error code of GDBM corresponding to an error code of Depot.
+ `ecode' specifies an error code of Depot.
+ The return value is the error code of GDBM. */
+static int gdbm_geterrno(int ecode){
+ switch(ecode){
+ case DP_ENOERR: return GDBM_NO_ERROR;
+ case DP_EBROKEN: return GDBM_BAD_MAGIC_NUMBER;
+ case DP_EKEEP: return GDBM_CANNOT_REPLACE;
+ case DP_ENOITEM: return GDBM_ITEM_NOT_FOUND;
+ case DP_EALLOC: return GDBM_MALLOC_ERROR;
+ case DP_EOPEN: return GDBM_FILE_OPEN_ERROR;
+ case DP_ESEEK: return GDBM_FILE_SEEK_ERROR;
+ case DP_EREAD: return GDBM_FILE_READ_ERROR;
+ case DP_EWRITE: return GDBM_FILE_WRITE_ERROR;
+ case DP_EMKDIR: return GDBM_FILE_OPEN_ERROR;
+ default: break;
+ }
+ return GDBM_ILLEGAL_DATA;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/hovel.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/hovel.h
===================================================================
--- box/trunk/qdbm/hovel.h (rev 0)
+++ box/trunk/qdbm/hovel.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,278 @@
+/*************************************************************************************************
+ * The GDBM-compatible API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _HOVEL_H /* duplication check */
+#define _HOVEL_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <depot.h>
+#include <curia.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+enum { /* enumeration for error codes */
+ GDBM_NO_ERROR, /* no error */
+ GDBM_MALLOC_ERROR, /* malloc error */
+ GDBM_BLOCK_SIZE_ERROR, /* block size error */
+ GDBM_FILE_OPEN_ERROR, /* file open error */
+ GDBM_FILE_WRITE_ERROR, /* file write error */
+ GDBM_FILE_SEEK_ERROR, /* file seek error */
+ GDBM_FILE_READ_ERROR, /* file read error */
+ GDBM_BAD_MAGIC_NUMBER, /* bad magic number */
+ GDBM_EMPTY_DATABASE, /* empty database */
+ GDBM_CANT_BE_READER, /* can't be reader */
+ GDBM_CANT_BE_WRITER, /* can't be writer */
+ GDBM_READER_CANT_DELETE, /* reader can't delete */
+ GDBM_READER_CANT_STORE, /* reader can't store */
+ GDBM_READER_CANT_REORGANIZE, /* reader can't reorganize */
+ GDBM_UNKNOWN_UPDATE, /* unknown update */
+ GDBM_ITEM_NOT_FOUND, /* item not found */
+ GDBM_REORGANIZE_FAILED, /* reorganize failed */
+ GDBM_CANNOT_REPLACE, /* cannot replace */
+ GDBM_ILLEGAL_DATA, /* illegal data */
+ GDBM_OPT_ALREADY_SET, /* option already set */
+ GDBM_OPT_ILLEGAL /* option illegal */
+};
+
+typedef int gdbm_error; /* type of error codes */
+
+typedef struct { /* type of structure for a database handle */
+ DEPOT *depot; /* internal database handle of Depot */
+ CURIA *curia; /* internal database handle of Curia */
+ int syncmode; /* whether to be besyncronous mode */
+} GDBM;
+
+typedef GDBM *GDBM_FILE; /* type of pointer to a database handle */
+
+typedef struct { /* type of structure for a key or a value */
+ char *dptr; /* pointer to the region */
+ size_t dsize; /* size of the region */
+} datum;
+
+enum { /* enumeration for open modes */
+ GDBM_READER = 1 << 0, /* open as a reader */
+ GDBM_WRITER = 1 << 1, /* open as a writer */
+ GDBM_WRCREAT = 1 << 2, /* a writer creating */
+ GDBM_NEWDB = 1 << 3, /* a writer creating and truncating */
+ GDBM_SYNC = 1 << 4, /* syncronous mode */
+ GDBM_NOLOCK = 1 << 5, /* no lock mode */
+ GDBM_LOCKNB = 1 << 6, /* non-blocking lock mode */
+ GDBM_FAST = 1 << 7, /* fast mode */
+ GDBM_SPARSE = 1 << 8 /* create as sparse file */
+};
+
+enum { /* enumeration for write modes */
+ GDBM_INSERT, /* keep an existing value */
+ GDBM_REPLACE /* overwrite an existing value */
+};
+
+enum { /* enumeration for options */
+ GDBM_CACHESIZE, /* set cache size */
+ GDBM_FASTMODE, /* set fast mode */
+ GDBM_SYNCMODE, /* set syncronous mode */
+ GDBM_CENTFREE, /* set free block pool */
+ GDBM_COALESCEBLKS /* set free block marging */
+};
+
+
+/* String containing the version information. */
+MYEXTERN char *gdbm_version;
+
+
+/* Last happened error code. */
+#define gdbm_errno (*gdbm_errnoptr())
+
+
+/* Get a message string corresponding to an error code.
+ `gdbmerrno' specifies an error code.
+ The return value is the message string of the error code. The region of the return value
+ is not writable. */
+char *gdbm_strerror(gdbm_error gdbmerrno);
+
+
+/* Get a database handle after the fashion of GDBM.
+ `name' specifies a name of a database.
+ `read_write' specifies the connection mode: `GDBM_READER' as a reader, `GDBM_WRITER',
+ `GDBM_WRCREAT' and `GDBM_NEWDB' as a writer. `GDBM_WRCREAT' makes a database file or
+ directory if it does not exist. `GDBM_NEWDB' makes a new database even if it exists.
+ You can add the following to writer modes by bitwise or: `GDBM_SYNC', `GDBM_NOLOCK',
+ `GDBM_LCKNB', `GDBM_FAST', and `GDBM_SPARSE'. `GDBM_SYNC' means a database is synchronized
+ after every updating method. `GDBM_NOLOCK' means a database is opened without file locking.
+ `GDBM_LOCKNB' means file locking is performed without blocking. `GDBM_FAST' is ignored.
+ `GDBM_SPARSE' is an original mode of QDBM and makes database a sparse file.
+ `mode' specifies a mode of a database file or a database directory as the one of `open'
+ or `mkdir' call does.
+ `bnum' specifies the number of elements of each bucket array. If it is not more than 0,
+ the default value is specified.
+ `dnum' specifies the number of division of the database. If it is not more than 0, the
+ returning handle is created as a wrapper of Depot, else, it is as a wrapper of Curia.
+ The return value is the database handle or `NULL' if it is not successful.
+ If the database already exists, whether it is one of Depot or Curia is measured
+ automatically. */
+GDBM_FILE gdbm_open(char *name, int block_size, int read_write, int mode,
+ void (*fatal_func)(void));
+
+
+/* Get a database handle after the fashion of QDBM.
+ `name' specifies a name of a database.
+ `read_write' specifies the connection mode: `GDBM_READER' as a reader, `GDBM_WRITER',
+ `GDBM_WRCREAT' and `GDBM_NEWDB' as a writer. `GDBM_WRCREAT' makes a database file or
+ directory if it does not exist. `GDBM_NEWDB' makes a new database even if it exists.
+ You can add the following to writer modes by bitwise or: `GDBM_SYNC', `GDBM_NOLOCK',
+ `GDBM_LOCKNB', `GDBM_FAST', and `GDBM_SPARSE'. `GDBM_SYNC' means a database is synchronized
+ after every updating method. `GDBM_NOLOCK' means a database is opened without file locking.
+ `GDBM_LOCKNB' means file locking is performed without blocking. `GDBM_FAST' is ignored.
+ `GDBM_SPARSE' is an original mode of QDBM and makes database sparse files.
+ `mode' specifies a mode of a database file as the one of `open' or `mkdir' call does.
+ `bnum' specifies the number of elements of each bucket array. If it is not more than 0,
+ the default value is specified.
+ `dnum' specifies the number of division of the database. If it is not more than 0, the
+ returning handle is created as a wrapper of Depot, else, it is as a wrapper of Curia.
+ `align' specifies the basic size of alignment.
+ The return value is the database handle or `NULL' if it is not successful. */
+GDBM_FILE gdbm_open2(char *name, int read_write, int mode, int bnum, int dnum, int align);
+
+
+/* Close a database handle.
+ `dbf' specifies a database handle.
+ Because the region of the closed handle is released, it becomes impossible to use the
+ handle. */
+void gdbm_close(GDBM_FILE dbf);
+
+
+/* Store a record.
+ `dbf' specifies a database handle connected as a writer.
+ `key' specifies a structure of a key. `content' specifies a structure of a value.
+ `flag' specifies behavior when the key overlaps, by the following values: `GDBM_REPLACE',
+ which means the specified value overwrites the existing one, `GDBM_INSERT', which means
+ the existing value is kept.
+ The return value is 0 if it is successful, 1 if it gives up because of overlaps of the key,
+ -1 if other error occurs. */
+int gdbm_store(GDBM_FILE dbf, datum key, datum content, int flag);
+
+
+/* Delete a record.
+ `dbf' specifies a database handle connected as a writer.
+ `key' specifies a structure of a key.
+ The return value is 0 if it is successful, -1 if some errors occur. */
+int gdbm_delete(GDBM_FILE dbf, datum key);
+
+
+/* Retrieve a record.
+ `dbf' specifies a database handle.
+ `key' specifies a structure of a key.
+ The return value is a structure of the result.
+ If a record corresponds, the member `dptr' of the structure is the pointer to the region
+ of the value. If no record corresponds or some errors occur, `dptr' is `NULL'. Because
+ the region pointed to by `dptr' is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+datum gdbm_fetch(GDBM_FILE dbf, datum key);
+
+
+/* Check whether a record exists or not.
+ `dbf' specifies a database handle.
+ `key' specifies a structure of a key.
+ The return value is true if a record corresponds and no error occurs, or false, else,
+ it is false. */
+int gdbm_exists(GDBM_FILE dbf, datum key);
+
+
+/* Get the first key of a database.
+ `dbf' specifies a database handle.
+ The return value is a structure of the result.
+ If a record corresponds, the member `dptr' of the structure is the pointer to the region
+ of the first key. If no record corresponds or some errors occur, `dptr' is `NULL'.
+ Because the region pointed to by `dptr' is allocated with the `malloc' call, it should
+ be released with the `free' call if it is no longer in use. */
+datum gdbm_firstkey(GDBM_FILE dbf);
+
+
+/* Get the next key of a database.
+ `dbf' specifies a database handle.
+ The return value is a structure of the result.
+ If a record corresponds, the member `dptr' of the structure is the pointer to the region
+ of the next key. If no record corresponds or some errors occur, `dptr' is `NULL'.
+ Because the region pointed to by `dptr' is allocated with the `malloc' call, it should
+ be released with the `free' call if it is no longer in use. */
+datum gdbm_nextkey(GDBM_FILE dbf, datum key);
+
+
+/* Synchronize updating contents with the file and the device.
+ `dbf' specifies a database handle connected as a writer. */
+void gdbm_sync(GDBM_FILE dbf);
+
+
+/* Reorganize a database.
+ `dbf' specifies a database handle connected as a writer.
+ If successful, the return value is 0, else -1. */
+int gdbm_reorganize(GDBM_FILE dbf);
+
+
+/* Get the file descriptor of a database file.
+ `dbf' specifies a database handle connected as a writer.
+ The return value is the file descriptor of the database file.
+ If the database is a directory the return value is -1. */
+int gdbm_fdesc(GDBM_FILE dbf);
+
+
+/* No effect.
+ `dbf' specifies a database handle.
+ `option' is ignored. `size' is ignored.
+ The return value is 0.
+ The function is only for compatibility. */
+int gdbm_setopt(GDBM_FILE dbf, int option, int *value, int size);
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Get the pointer of the last happened error code. */
+int *gdbm_errnoptr(void);
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/hovel.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/hvmgr.c
===================================================================
--- box/trunk/qdbm/hvmgr.c (rev 0)
+++ box/trunk/qdbm/hvmgr.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,582 @@
+/*************************************************************************************************
+ * Utility for debugging Hovel and its applications
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <hovel.h>
+#include <cabin.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define ALIGNSIZ 16 /* basic size of alignment */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *hextoobj(const char *str, int *sp);
+int runcreate(int argc, char **argv);
+int runstore(int argc, char **argv);
+int rundelete(int argc, char **argv);
+int runfetch(int argc, char **argv);
+int runlist(int argc, char **argv);
+int runoptimize(int argc, char **argv);
+void pgerror(const char *name);
+void printobj(const char *obj, int size);
+void printobjhex(const char *obj, int size);
+int docreate(char *name, int qdbm, int bnum, int dnum, int sparse);
+int dostore(char *name, int qdbm, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int ins);
+int dodelete(char *name, int qdbm, const char *kbuf, int ksiz);
+int dofetch(char *name, int qdbm, const char *kbuf, int ksiz, int ox, int nb);
+int dolist(char *name, int qdbm, int ox);
+int dooptimize(char *name, int qdbm);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "create")){
+ rv = runcreate(argc, argv);
+ } else if(!strcmp(argv[1], "store")){
+ rv = runstore(argc, argv);
+ } else if(!strcmp(argv[1], "delete")){
+ rv = rundelete(argc, argv);
+ } else if(!strcmp(argv[1], "fetch")){
+ rv = runfetch(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else if(!strcmp(argv[1], "optimize")){
+ rv = runoptimize(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: administration utility for Hovel\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s create [-qdbm bnum dnum] [-s] name\n", progname);
+ fprintf(stderr, " %s store [-qdbm] [-kx] [-vx|-vf] [-insert] name key val\n", progname);
+ fprintf(stderr, " %s delete [-qdbm] [-kx] name key\n", progname);
+ fprintf(stderr, " %s fetch [-qdbm] [-kx] [-ox] [-n] name key\n", progname);
+ fprintf(stderr, " %s list [-qdbm] [-ox] name\n", progname);
+ fprintf(stderr, " %s optimize [-qdbm] name\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* create a binary object from a hexadecimal string */
+char *hextoobj(const char *str, int *sp){
+ char *buf, mbuf[3];
+ int len, i, j;
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ j = 0;
+ for(i = 0; i < len; i += 2){
+ while(strchr(" \n\r\t\f\v", str[i])){
+ i++;
+ }
+ if((mbuf[0] = str[i]) == '\0') break;
+ if((mbuf[1] = str[i+1]) == '\0') break;
+ mbuf[2] = '\0';
+ buf[j++] = (char)strtol(mbuf, NULL, 16);
+ }
+ buf[j] = '\0';
+ *sp = j;
+ return buf;
+}
+
+
+/* parse arguments of create command */
+int runcreate(int argc, char **argv){
+ char *name;
+ int i, sb, qdbm, bnum, dnum, rv;
+ name = NULL;
+ sb = FALSE;
+ qdbm = FALSE;
+ bnum = -1;
+ dnum = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ if(++i >= argc) usage();
+ bnum = atoi(argv[i]);
+ if(++i >= argc) usage();
+ dnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docreate(name, qdbm, bnum, dnum, sb);
+ return rv;
+}
+
+
+/* parse arguments of store command */
+int runstore(int argc, char **argv){
+ char *name, *key, *val, *kbuf, *vbuf;
+ int i, qdbm, kx, vx, vf, ins, ksiz, vsiz, rv;
+ name = NULL;
+ qdbm = FALSE;
+ kx = FALSE;
+ vx = FALSE;
+ vf = FALSE;
+ ins = FALSE;
+ key = NULL;
+ val = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-vx")){
+ vx = TRUE;
+ } else if(!strcmp(argv[i], "-vf")){
+ vf = TRUE;
+ } else if(!strcmp(argv[i], "-insert")){
+ ins = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else if(!val){
+ val = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || !val) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = strlen(kbuf);
+ }
+ if(vx){
+ vbuf = hextoobj(val, &vsiz);
+ } else if(vf){
+ vbuf = cbreadfile(val, &vsiz);
+ } else {
+ vbuf = cbmemdup(val, -1);
+ vsiz = strlen(vbuf);
+ }
+ if(kbuf && vbuf){
+ rv = dostore(name, qdbm, kbuf, ksiz, vbuf, vsiz, ins);
+ } else {
+ if(vf){
+ fprintf(stderr, "%s: %s: cannot read\n", progname, val);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ }
+ rv = 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ return rv;
+}
+
+
+/* parse arguments of delete command */
+int rundelete(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, qdbm, kx, ksiz, rv;
+ name = NULL;
+ qdbm = FALSE;
+ kx = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = strlen(kbuf);
+ }
+ if(kbuf){
+ rv = dodelete(name, qdbm, kbuf, ksiz);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of fetch command */
+int runfetch(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, qdbm, kx, ox, nb, ksiz, rv;
+ name = NULL;
+ qdbm = FALSE;
+ kx = FALSE;
+ ox = FALSE;
+ nb = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = strlen(kbuf);
+ }
+ if(kbuf){
+ rv = dofetch(name, qdbm, kbuf, ksiz, ox, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ char *name;
+ int i, qdbm, ox, rv;
+ name = NULL;
+ qdbm = FALSE;
+ ox = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dolist(name, qdbm, ox);
+ return rv;
+}
+
+
+/* parse arguments of optimize command */
+int runoptimize(int argc, char **argv){
+ char *name;
+ int i, qdbm, rv;
+ name = NULL;
+ qdbm = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dooptimize(name, qdbm);
+ return rv;
+}
+
+
+/* print an error message */
+void pgerror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, gdbm_strerror(gdbm_errno));
+}
+
+
+/* print an object */
+void printobj(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ putchar(obj[i]);
+ }
+}
+
+
+/* print an object as a hexadecimal string */
+void printobjhex(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ printf("%s%02X", i > 0 ? " " : "", ((const unsigned char *)obj)[i]);
+ }
+}
+
+
+/* perform create command */
+int docreate(char *name, int qdbm, int bnum, int dnum, int sparse){
+ GDBM_FILE dbf;
+ int rwmode;
+ rwmode = GDBM_NEWDB | (sparse ? GDBM_SPARSE : 0);
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, rwmode, 00644, bnum, dnum, ALIGNSIZ))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, rwmode, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ gdbm_close(dbf);
+ return 0;
+}
+
+
+/* perform store command */
+int dostore(char *name, int qdbm, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int ins){
+ GDBM_FILE dbf;
+ datum key, content;
+ int rv;
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, GDBM_WRITER, 00644, -1, -1, ALIGNSIZ))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, GDBM_WRITER, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ key.dptr = (char *)kbuf;
+ key.dsize = ksiz;
+ content.dptr = (char *)vbuf;
+ content.dsize = vsiz;
+ rv = 0;
+ if(gdbm_store(dbf, key, content, ins ? GDBM_INSERT : GDBM_REPLACE) != 0){
+ pgerror(name);
+ rv = 1;
+ }
+ gdbm_close(dbf);
+ return rv;
+}
+
+
+/* perform delete command */
+int dodelete(char *name, int qdbm, const char *kbuf, int ksiz){
+ GDBM_FILE dbf;
+ datum key;
+ int rv;
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, GDBM_WRITER, 00644, -1, -1, ALIGNSIZ))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, GDBM_WRITER, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ key.dptr = (char *)kbuf;
+ key.dsize = ksiz;
+ if(gdbm_delete(dbf, key) == 0){
+ rv = 0;
+ } else {
+ pgerror(name);
+ rv = 1;
+ }
+ gdbm_close(dbf);
+ return rv;
+}
+
+
+/* perform fetch command */
+int dofetch(char *name, int qdbm, const char *kbuf, int ksiz, int ox, int nb){
+ GDBM_FILE dbf;
+ datum key, content;
+ int rv;
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, GDBM_READER, 00644, -1, -1, -1))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, GDBM_READER, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ key.dptr = (char *)kbuf;
+ key.dsize = ksiz;
+ content = gdbm_fetch(dbf, key);
+ if(content.dptr){
+ if(ox){
+ printobjhex(content.dptr, content.dsize);
+ } else {
+ printobj(content.dptr, content.dsize);
+ }
+ if(!nb) putchar('\n');
+ rv = 0;
+ free(content.dptr);
+ } else {
+ pgerror(name);
+ rv = 1;
+ }
+ gdbm_close(dbf);
+ return rv;
+}
+
+
+/* perform list command */
+int dolist(char *name, int qdbm, int ox){
+ GDBM_FILE dbf;
+ datum key, val;
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, GDBM_READER, 00644, -1, -1, -1))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, GDBM_READER, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ for(key = gdbm_firstkey(dbf); key.dptr != NULL; key = gdbm_nextkey(dbf, key)){
+ val = gdbm_fetch(dbf, key);
+ if(!val.dptr){
+ free(key.dptr);
+ break;
+ }
+ if(ox){
+ printobjhex(key.dptr, key.dsize);
+ putchar('\t');
+ printobjhex(val.dptr, val.dsize);
+ } else {
+ printobj(key.dptr, key.dsize);
+ putchar('\t');
+ printobj(val.dptr, val.dsize);
+ }
+ putchar('\n');
+ free(val.dptr);
+ free(key.dptr);
+ }
+ gdbm_close(dbf);
+ return 0;
+}
+
+
+/* perform optimize command */
+int dooptimize(char *name, int qdbm){
+ GDBM_FILE dbf;
+ int rv;
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, GDBM_WRITER, 00644, -1, -1, ALIGNSIZ))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, GDBM_WRITER, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ rv = 0;
+ if(gdbm_reorganize(dbf) != 0){
+ pgerror(name);
+ rv = 1;
+ }
+ gdbm_close(dbf);
+ return rv;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/hvmgr.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/hvtest.c
===================================================================
--- box/trunk/qdbm/hvtest.c (rev 0)
+++ box/trunk/qdbm/hvtest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,272 @@
+/*************************************************************************************************
+ * Test cases of Hovel
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <hovel.h>
+#include <cabin.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdarg.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define DIVNUM 5 /* number of division */
+#define ALIGNSIZ 16 /* basic size of alignment */
+#define RECBUFSIZ 32 /* buffer for records */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runwrite(int argc, char **argv);
+int runread(int argc, char **argv);
+int printfflush(const char *format, ...);
+void pgerror(const char *name);
+int dowrite(char *name, int rnum, int qdbm, int sparse);
+int doread(char *name, int rnum, int qdbm);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "write")){
+ rv = runwrite(argc, argv);
+ } else if(!strcmp(argv[1], "read")){
+ rv = runread(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Hovel\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s write [-qdbm] [-s] name rnum\n", progname);
+ fprintf(stderr, " %s read [-qdbm] name rnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of write command */
+int runwrite(int argc, char **argv){
+ char *name, *rstr;
+ int i, sb, qdbm, rnum, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ sb = FALSE;
+ qdbm = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowrite(name, rnum, qdbm, sb);
+ return rv;
+}
+
+
+/* parse arguments of read command */
+int runread(int argc, char **argv){
+ char *name, *rstr;
+ int i, qdbm, rnum, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ qdbm = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-qdbm")){
+ qdbm = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = doread(name, rnum, qdbm);
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* print an error message */
+void pgerror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, gdbm_strerror(gdbm_errno));
+}
+
+
+/* perform write command */
+int dowrite(char *name, int rnum, int qdbm, int sparse){
+ GDBM_FILE dbf;
+ datum key, content;
+ int i, rwmode, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Writing Test>\n name=%s rnum=%d qdbm=%d\n\n", name, rnum, qdbm);
+ /* open a database */
+ rwmode = GDBM_NEWDB | (sparse ? GDBM_SPARSE : 0);
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, rwmode, 00644, rnum / DIVNUM, DIVNUM, ALIGNSIZ))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, rwmode, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ key.dptr = buf;
+ key.dsize = len;
+ content.dptr = buf;
+ content.dsize = len;
+ /* store a record */
+ if(gdbm_store(dbf, key, content, GDBM_REPLACE) != 0){
+ pgerror(name);
+ err = TRUE;
+ break;
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ gdbm_close(dbf);
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform read command */
+int doread(char *name, int rnum, int qdbm){
+ GDBM_FILE dbf;
+ datum key, content;
+ int i, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Reading Test>\n name=%s rnum=%d qdbm=%d\n\n", name, rnum, qdbm);
+ /* open a database */
+ if(qdbm){
+ if(!(dbf = gdbm_open2(name, GDBM_READER, 00644, -1, -1, -1))){
+ pgerror(name);
+ return 1;
+ }
+ } else {
+ if(!(dbf = gdbm_open(name, 0, GDBM_READER, 00644, NULL))){
+ pgerror(name);
+ return 1;
+ }
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* retrieve a record */
+ len = sprintf(buf, "%08d", i);
+ key.dptr = buf;
+ key.dsize = len;
+ content = gdbm_fetch(dbf, key);
+ if(!content.dptr){
+ pgerror(name);
+ err = TRUE;
+ break;
+ }
+ free(content.dptr);
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ gdbm_close(dbf);
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/hvtest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/misc/COPYING.txt
===================================================================
--- box/trunk/qdbm/misc/COPYING.txt (rev 0)
+++ box/trunk/qdbm/misc/COPYING.txt 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,504 @@
+ GNU LESSER GENERAL PUBLIC LICENSE
+ Version 2.1, February 1999
+
+ Copyright (C) 1991, 1999 Free Software Foundation, Inc.
+ 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+[This is the first released version of the Lesser GPL. It also counts
+ as the successor of the GNU Library Public License, version 2, hence
+ the version number 2.1.]
+
+ Preamble
+
+ The licenses for most software are designed to take away your
+freedom to share and change it. By contrast, the GNU General Public
+Licenses are intended to guarantee your freedom to share and change
+free software--to make sure the software is free for all its users.
+
+ This license, the Lesser General Public License, applies to some
+specially designated software packages--typically libraries--of the
+Free Software Foundation and other authors who decide to use it. You
+can use it too, but we suggest you first think carefully about whether
+this license or the ordinary General Public License is the better
+strategy to use in any particular case, based on the explanations below.
+
+ When we speak of free software, we are referring to freedom of use,
+not price. Our General Public Licenses are designed to make sure that
+you have the freedom to distribute copies of free software (and charge
+for this service if you wish); that you receive source code or can get
+it if you want it; that you can change the software and use pieces of
+it in new free programs; and that you are informed that you can do
+these things.
+
+ To protect your rights, we need to make restrictions that forbid
+distributors to deny you these rights or to ask you to surrender these
+rights. These restrictions translate to certain responsibilities for
+you if you distribute copies of the library or if you modify it.
+
+ For example, if you distribute copies of the library, whether gratis
+or for a fee, you must give the recipients all the rights that we gave
+you. You must make sure that they, too, receive or can get the source
+code. If you link other code with the library, you must provide
+complete object files to the recipients, so that they can relink them
+with the library after making changes to the library and recompiling
+it. And you must show them these terms so they know their rights.
+
+ We protect your rights with a two-step method: (1) we copyright the
+library, and (2) we offer you this license, which gives you legal
+permission to copy, distribute and/or modify the library.
+
+ To protect each distributor, we want to make it very clear that
+there is no warranty for the free library. Also, if the library is
+modified by someone else and passed on, the recipients should know
+that what they have is not the original version, so that the original
+author's reputation will not be affected by problems that might be
+introduced by others.
+
+ Finally, software patents pose a constant threat to the existence of
+any free program. We wish to make sure that a company cannot
+effectively restrict the users of a free program by obtaining a
+restrictive license from a patent holder. Therefore, we insist that
+any patent license obtained for a version of the library must be
+consistent with the full freedom of use specified in this license.
+
+ Most GNU software, including some libraries, is covered by the
+ordinary GNU General Public License. This license, the GNU Lesser
+General Public License, applies to certain designated libraries, and
+is quite different from the ordinary General Public License. We use
+this license for certain libraries in order to permit linking those
+libraries into non-free programs.
+
+ When a program is linked with a library, whether statically or using
+a shared library, the combination of the two is legally speaking a
+combined work, a derivative of the original library. The ordinary
+General Public License therefore permits such linking only if the
+entire combination fits its criteria of freedom. The Lesser General
+Public License permits more lax criteria for linking other code with
+the library.
+
+ We call this license the "Lesser" General Public License because it
+does Less to protect the user's freedom than the ordinary General
+Public License. It also provides other free software developers Less
+of an advantage over competing non-free programs. These disadvantages
+are the reason we use the ordinary General Public License for many
+libraries. However, the Lesser license provides advantages in certain
+special circumstances.
+
+ For example, on rare occasions, there may be a special need to
+encourage the widest possible use of a certain library, so that it becomes
+a de-facto standard. To achieve this, non-free programs must be
+allowed to use the library. A more frequent case is that a free
+library does the same job as widely used non-free libraries. In this
+case, there is little to gain by limiting the free library to free
+software only, so we use the Lesser General Public License.
+
+ In other cases, permission to use a particular library in non-free
+programs enables a greater number of people to use a large body of
+free software. For example, permission to use the GNU C Library in
+non-free programs enables many more people to use the whole GNU
+operating system, as well as its variant, the GNU/Linux operating
+system.
+
+ Although the Lesser General Public License is Less protective of the
+users' freedom, it does ensure that the user of a program that is
+linked with the Library has the freedom and the wherewithal to run
+that program using a modified version of the Library.
+
+ The precise terms and conditions for copying, distribution and
+modification follow. Pay close attention to the difference between a
+"work based on the library" and a "work that uses the library". The
+former contains code derived from the library, whereas the latter must
+be combined with the library in order to run.
+
+ GNU LESSER GENERAL PUBLIC LICENSE
+ TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
+
+ 0. This License Agreement applies to any software library or other
+program which contains a notice placed by the copyright holder or
+other authorized party saying it may be distributed under the terms of
+this Lesser General Public License (also called "this License").
+Each licensee is addressed as "you".
+
+ A "library" means a collection of software functions and/or data
+prepared so as to be conveniently linked with application programs
+(which use some of those functions and data) to form executables.
+
+ The "Library", below, refers to any such software library or work
+which has been distributed under these terms. A "work based on the
+Library" means either the Library or any derivative work under
+copyright law: that is to say, a work containing the Library or a
+portion of it, either verbatim or with modifications and/or translated
+straightforwardly into another language. (Hereinafter, translation is
+included without limitation in the term "modification".)
+
+ "Source code" for a work means the preferred form of the work for
+making modifications to it. For a library, complete source code means
+all the source code for all modules it contains, plus any associated
+interface definition files, plus the scripts used to control compilation
+and installation of the library.
+
+ Activities other than copying, distribution and modification are not
+covered by this License; they are outside its scope. The act of
+running a program using the Library is not restricted, and output from
+such a program is covered only if its contents constitute a work based
+on the Library (independent of the use of the Library in a tool for
+writing it). Whether that is true depends on what the Library does
+and what the program that uses the Library does.
+
+ 1. You may copy and distribute verbatim copies of the Library's
+complete source code as you receive it, in any medium, provided that
+you conspicuously and appropriately publish on each copy an
+appropriate copyright notice and disclaimer of warranty; keep intact
+all the notices that refer to this License and to the absence of any
+warranty; and distribute a copy of this License along with the
+Library.
+
+ You may charge a fee for the physical act of transferring a copy,
+and you may at your option offer warranty protection in exchange for a
+fee.
+
+ 2. You may modify your copy or copies of the Library or any portion
+of it, thus forming a work based on the Library, and copy and
+distribute such modifications or work under the terms of Section 1
+above, provided that you also meet all of these conditions:
+
+ a) The modified work must itself be a software library.
+
+ b) You must cause the files modified to carry prominent notices
+ stating that you changed the files and the date of any change.
+
+ c) You must cause the whole of the work to be licensed at no
+ charge to all third parties under the terms of this License.
+
+ d) If a facility in the modified Library refers to a function or a
+ table of data to be supplied by an application program that uses
+ the facility, other than as an argument passed when the facility
+ is invoked, then you must make a good faith effort to ensure that,
+ in the event an application does not supply such function or
+ table, the facility still operates, and performs whatever part of
+ its purpose remains meaningful.
+
+ (For example, a function in a library to compute square roots has
+ a purpose that is entirely well-defined independent of the
+ application. Therefore, Subsection 2d requires that any
+ application-supplied function or table used by this function must
+ be optional: if the application does not supply it, the square
+ root function must still compute square roots.)
+
+These requirements apply to the modified work as a whole. If
+identifiable sections of that work are not derived from the Library,
+and can be reasonably considered independent and separate works in
+themselves, then this License, and its terms, do not apply to those
+sections when you distribute them as separate works. But when you
+distribute the same sections as part of a whole which is a work based
+on the Library, the distribution of the whole must be on the terms of
+this License, whose permissions for other licensees extend to the
+entire whole, and thus to each and every part regardless of who wrote
+it.
+
+Thus, it is not the intent of this section to claim rights or contest
+your rights to work written entirely by you; rather, the intent is to
+exercise the right to control the distribution of derivative or
+collective works based on the Library.
+
+In addition, mere aggregation of another work not based on the Library
+with the Library (or with a work based on the Library) on a volume of
+a storage or distribution medium does not bring the other work under
+the scope of this License.
+
+ 3. You may opt to apply the terms of the ordinary GNU General Public
+License instead of this License to a given copy of the Library. To do
+this, you must alter all the notices that refer to this License, so
+that they refer to the ordinary GNU General Public License, version 2,
+instead of to this License. (If a newer version than version 2 of the
+ordinary GNU General Public License has appeared, then you can specify
+that version instead if you wish.) Do not make any other change in
+these notices.
+
+ Once this change is made in a given copy, it is irreversible for
+that copy, so the ordinary GNU General Public License applies to all
+subsequent copies and derivative works made from that copy.
+
+ This option is useful when you wish to copy part of the code of
+the Library into a program that is not a library.
+
+ 4. You may copy and distribute the Library (or a portion or
+derivative of it, under Section 2) in object code or executable form
+under the terms of Sections 1 and 2 above provided that you accompany
+it with the complete corresponding machine-readable source code, which
+must be distributed under the terms of Sections 1 and 2 above on a
+medium customarily used for software interchange.
+
+ If distribution of object code is made by offering access to copy
+from a designated place, then offering equivalent access to copy the
+source code from the same place satisfies the requirement to
+distribute the source code, even though third parties are not
+compelled to copy the source along with the object code.
+
+ 5. A program that contains no derivative of any portion of the
+Library, but is designed to work with the Library by being compiled or
+linked with it, is called a "work that uses the Library". Such a
+work, in isolation, is not a derivative work of the Library, and
+therefore falls outside the scope of this License.
+
+ However, linking a "work that uses the Library" with the Library
+creates an executable that is a derivative of the Library (because it
+contains portions of the Library), rather than a "work that uses the
+library". The executable is therefore covered by this License.
+Section 6 states terms for distribution of such executables.
+
+ When a "work that uses the Library" uses material from a header file
+that is part of the Library, the object code for the work may be a
+derivative work of the Library even though the source code is not.
+Whether this is true is especially significant if the work can be
+linked without the Library, or if the work is itself a library. The
+threshold for this to be true is not precisely defined by law.
+
+ If such an object file uses only numerical parameters, data
+structure layouts and accessors, and small macros and small inline
+functions (ten lines or less in length), then the use of the object
+file is unrestricted, regardless of whether it is legally a derivative
+work. (Executables containing this object code plus portions of the
+Library will still fall under Section 6.)
+
+ Otherwise, if the work is a derivative of the Library, you may
+distribute the object code for the work under the terms of Section 6.
+Any executables containing that work also fall under Section 6,
+whether or not they are linked directly with the Library itself.
+
+ 6. As an exception to the Sections above, you may also combine or
+link a "work that uses the Library" with the Library to produce a
+work containing portions of the Library, and distribute that work
+under terms of your choice, provided that the terms permit
+modification of the work for the customer's own use and reverse
+engineering for debugging such modifications.
+
+ You must give prominent notice with each copy of the work that the
+Library is used in it and that the Library and its use are covered by
+this License. You must supply a copy of this License. If the work
+during execution displays copyright notices, you must include the
+copyright notice for the Library among them, as well as a reference
+directing the user to the copy of this License. Also, you must do one
+of these things:
+
+ a) Accompany the work with the complete corresponding
+ machine-readable source code for the Library including whatever
+ changes were used in the work (which must be distributed under
+ Sections 1 and 2 above); and, if the work is an executable linked
+ with the Library, with the complete machine-readable "work that
+ uses the Library", as object code and/or source code, so that the
+ user can modify the Library and then relink to produce a modified
+ executable containing the modified Library. (It is understood
+ that the user who changes the contents of definitions files in the
+ Library will not necessarily be able to recompile the application
+ to use the modified definitions.)
+
+ b) Use a suitable shared library mechanism for linking with the
+ Library. A suitable mechanism is one that (1) uses at run time a
+ copy of the library already present on the user's computer system,
+ rather than copying library functions into the executable, and (2)
+ will operate properly with a modified version of the library, if
+ the user installs one, as long as the modified version is
+ interface-compatible with the version that the work was made with.
+
+ c) Accompany the work with a written offer, valid for at
+ least three years, to give the same user the materials
+ specified in Subsection 6a, above, for a charge no more
+ than the cost of performing this distribution.
+
+ d) If distribution of the work is made by offering access to copy
+ from a designated place, offer equivalent access to copy the above
+ specified materials from the same place.
+
+ e) Verify that the user has already received a copy of these
+ materials or that you have already sent this user a copy.
+
+ For an executable, the required form of the "work that uses the
+Library" must include any data and utility programs needed for
+reproducing the executable from it. However, as a special exception,
+the materials to be distributed need not include anything that is
+normally distributed (in either source or binary form) with the major
+components (compiler, kernel, and so on) of the operating system on
+which the executable runs, unless that component itself accompanies
+the executable.
+
+ It may happen that this requirement contradicts the license
+restrictions of other proprietary libraries that do not normally
+accompany the operating system. Such a contradiction means you cannot
+use both them and the Library together in an executable that you
+distribute.
+
+ 7. You may place library facilities that are a work based on the
+Library side-by-side in a single library together with other library
+facilities not covered by this License, and distribute such a combined
+library, provided that the separate distribution of the work based on
+the Library and of the other library facilities is otherwise
+permitted, and provided that you do these two things:
+
+ a) Accompany the combined library with a copy of the same work
+ based on the Library, uncombined with any other library
+ facilities. This must be distributed under the terms of the
+ Sections above.
+
+ b) Give prominent notice with the combined library of the fact
+ that part of it is a work based on the Library, and explaining
+ where to find the accompanying uncombined form of the same work.
+
+ 8. You may not copy, modify, sublicense, link with, or distribute
+the Library except as expressly provided under this License. Any
+attempt otherwise to copy, modify, sublicense, link with, or
+distribute the Library is void, and will automatically terminate your
+rights under this License. However, parties who have received copies,
+or rights, from you under this License will not have their licenses
+terminated so long as such parties remain in full compliance.
+
+ 9. You are not required to accept this License, since you have not
+signed it. However, nothing else grants you permission to modify or
+distribute the Library or its derivative works. These actions are
+prohibited by law if you do not accept this License. Therefore, by
+modifying or distributing the Library (or any work based on the
+Library), you indicate your acceptance of this License to do so, and
+all its terms and conditions for copying, distributing or modifying
+the Library or works based on it.
+
+ 10. Each time you redistribute the Library (or any work based on the
+Library), the recipient automatically receives a license from the
+original licensor to copy, distribute, link with or modify the Library
+subject to these terms and conditions. You may not impose any further
+restrictions on the recipients' exercise of the rights granted herein.
+You are not responsible for enforcing compliance by third parties with
+this License.
+
+ 11. If, as a consequence of a court judgment or allegation of patent
+infringement or for any other reason (not limited to patent issues),
+conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License. If you cannot
+distribute so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you
+may not distribute the Library at all. For example, if a patent
+license would not permit royalty-free redistribution of the Library by
+all those who receive copies directly or indirectly through you, then
+the only way you could satisfy both it and this License would be to
+refrain entirely from distribution of the Library.
+
+If any portion of this section is held invalid or unenforceable under any
+particular circumstance, the balance of the section is intended to apply,
+and the section as a whole is intended to apply in other circumstances.
+
+It is not the purpose of this section to induce you to infringe any
+patents or other property right claims or to contest validity of any
+such claims; this section has the sole purpose of protecting the
+integrity of the free software distribution system which is
+implemented by public license practices. Many people have made
+generous contributions to the wide range of software distributed
+through that system in reliance on consistent application of that
+system; it is up to the author/donor to decide if he or she is willing
+to distribute software through any other system and a licensee cannot
+impose that choice.
+
+This section is intended to make thoroughly clear what is believed to
+be a consequence of the rest of this License.
+
+ 12. If the distribution and/or use of the Library is restricted in
+certain countries either by patents or by copyrighted interfaces, the
+original copyright holder who places the Library under this License may add
+an explicit geographical distribution limitation excluding those countries,
+so that distribution is permitted only in or among countries not thus
+excluded. In such case, this License incorporates the limitation as if
+written in the body of this License.
+
+ 13. The Free Software Foundation may publish revised and/or new
+versions of the Lesser General Public License from time to time.
+Such new versions will be similar in spirit to the present version,
+but may differ in detail to address new problems or concerns.
+
+Each version is given a distinguishing version number. If the Library
+specifies a version number of this License which applies to it and
+"any later version", you have the option of following the terms and
+conditions either of that version or of any later version published by
+the Free Software Foundation. If the Library does not specify a
+license version number, you may choose any version ever published by
+the Free Software Foundation.
+
+ 14. If you wish to incorporate parts of the Library into other free
+programs whose distribution conditions are incompatible with these,
+write to the author to ask for permission. For software which is
+copyrighted by the Free Software Foundation, write to the Free
+Software Foundation; we sometimes make exceptions for this. Our
+decision will be guided by the two goals of preserving the free status
+of all derivatives of our free software and of promoting the sharing
+and reuse of software generally.
+
+ NO WARRANTY
+
+ 15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
+WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
+EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
+OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
+KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
+LIBRARY IS WITH YOU. SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
+THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+ 16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
+WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
+AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
+FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
+CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
+LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
+RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
+FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
+SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
+DAMAGES.
+
+ END OF TERMS AND CONDITIONS
+
+ How to Apply These Terms to Your New Libraries
+
+ If you develop a new library, and you want it to be of the greatest
+possible use to the public, we recommend making it free software that
+everyone can redistribute and change. You can do so by permitting
+redistribution under these terms (or, alternatively, under the terms of the
+ordinary General Public License).
+
+ To apply these terms, attach the following notices to the library. It is
+safest to attach them to the start of each source file to most effectively
+convey the exclusion of warranty; and each file should have at least the
+"copyright" line and a pointer to where the full notice is found.
+
+ <one line to give the library's name and a brief idea of what it does.>
+ Copyright (C) <year> <name of author>
+
+ This library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ This library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with this library; if not, write to the Free Software
+ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+
+Also add information on how to contact you by electronic and paper mail.
+
+You should also get your employer (if you work as a programmer) or your
+school, if any, to sign a "copyright disclaimer" for the library, if
+necessary. Here is a sample; alter the names:
+
+ Yoyodyne, Inc., hereby disclaims all copyright interest in the
+ library `Frob' (a library for tweaking knobs) written by James Random Hacker.
+
+ <signature of Ty Coon>, 1 April 1990
+ Ty Coon, President of Vice
+
+That's all there is to it!
+
+
Property changes on: box/trunk/qdbm/misc/COPYING.txt
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/misc/README-win32.txt
===================================================================
--- box/trunk/qdbm/misc/README-win32.txt (rev 0)
+++ box/trunk/qdbm/misc/README-win32.txt 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,101 @@
+================================================================
+ QDBM: Quick Database Manager
+ Copyright (C) 2000-2007 Mikio Hirabayashi
+================================================================
+
+This is a package of Win32 binaries of QDBM. It contains C/Java
+APIs, their utility commands, and CGI scripts.
+
+See http://qdbm.sourceforge.net/ for more information.
+
+
+The following are documents of specifications.
+
+ spex.html : fundamental specifications
+ spex-ja.html : fundamental specifications in Japanese
+ jspex.html : specifications of Java API
+ jspex-ja.html : specifications of Java API in Japanese
+ japidoc/ : documents of Java API
+
+
+The following are header files of C language.
+Include them at source codes of your applications.
+
+ depot.h
+ curia.h
+ relic.h
+ hovel.h
+ cabin.h
+ villa.h
+ vista.h
+ odeum.h
+
+
+The following are dynamic linking libraries for the API of C.
+Copy them to the system directory or a directory of your project.
+
+ qdbm.dll : QDBM itself
+ libqdbm.dll.a : import library for `qdbm.dll'
+ mgwz.dll : ZLIB
+ libiconv-2.dll : ICONV
+
+
+The following is a dynamic linking library for the API of Java.
+Copy it to the system directory or a directory of your project.
+
+ jqdbm.dll
+
+
+The following is a Java archive of the classes.
+Include it in the CLASSPATH of your environment.
+
+ qdbm.jar
+
+
+The following are utility commands for testing and debugging.
+
+ dpmgr.exe
+ dptest.exe
+ dptsv.exe
+ crmgr.exe
+ crtest.exe
+ crtsv.exe
+ rlmgr.exe
+ rltest.exe
+ hvmgr.exe
+ hvtest.exe
+ cbtest.exe
+ cbcodec.exe
+ vlmgr.exe
+ vltest.exe
+ vltsv.exe
+ odmgr.exe
+ odtest.exe
+ odidx.exe
+ qmttest.exe
+
+
+The sub directory `cgi' contains CGI scripts, their configuration
+files, and their specifications.
+
+If you want an import library or a static library for Visual C++,
+please obtain the source package and use VCmakefile in it.
+
+
+QDBM was released under the terms of the GNU Lesser General Public
+License. See the file `COPYING.txt' for details.
+
+QDBM was written by Mikio Hirabayashi. You can contact the author
+by e-mail to `mikio at users.sourceforge.net'. However, as for
+topics which can be shared among other users, pleae send it to
+the mailing list. To join the mailing list, refer to the following
+URL.
+
+ http://lists.sourceforge.net/lists/listinfo/qdbm-users
+
+
+Thanks.
+
+
+
+== END OF FILE ==
Property changes on: box/trunk/qdbm/misc/README-win32.txt
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/misc/VCmakefile-old
===================================================================
--- box/trunk/qdbm/misc/VCmakefile-old (rev 0)
+++ box/trunk/qdbm/misc/VCmakefile-old 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,169 @@
+# Makefile to build QDBM using Microsoft Visual C++
+
+
+
+#================================================================
+# Setting Variables
+#================================================================
+
+
+# Targets
+MYLIBS = qdbm.lib
+LIBOBJS = depot.obj curia.obj relic.obj hovel.obj \
+ cabin.obj villa.obj vista.obj odeum.obj myconf.obj
+MYBINS = dpmgr.exe dptest.exe dptsv.exe crmgr.exe crtest.exe crtsv.exe \
+ rlmgr.exe rltest.exe hvmgr.exe hvtest.exe cbtest.exe cbcodec.exe \
+ vlmgr.exe vltest.exe vltsv.exe odmgr.exe odtest.exe odidx.exe
+
+# VC++ directory
+VCPATH = C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7
+
+# Building binaries
+CLFLAGS = /I "$(VCPATH)\Include" /I "$(VCPATH)\PlatformSDK\Include" /I "." /O2 /nologo
+LIBFLAGS = /libpath:"$(VCPATH)\lib" /libpath:"$(VCPATH)\PlatformSDK\Lib" /libpath:"." /nologo
+LINKFLAGS = /libpath:"$(VCPATH)\lib" /libpath:"$(VCPATH)\PlatformSDK\Lib" /libpath:"." /nologo
+
+
+
+#================================================================
+# Suffix rules
+#================================================================
+
+
+.SUFFIXES :
+.SUFFIXES : .c .obj
+
+.c.obj :
+ cl /c $(CLFLAGS) $<
+
+
+
+#================================================================
+# Actions
+#================================================================
+
+
+all : $(MYLIBS) $(MYBINS)
+
+
+clean :
+ del *.obj *.lib *.dll *.exp *.exe
+
+
+
+#================================================================
+# Building binaries
+#================================================================
+
+
+qdbm.lib : $(LIBOBJS)
+ lib $(LIBFLAGS) /OUT:$@ $(LIBOBJS)
+
+
+dpmgr.exe : dpmgr.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ dpmgr.obj qdbm.lib
+
+
+dptest.exe : dptest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ dptest.obj qdbm.lib
+
+
+dptsv.exe : dptsv.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ dptsv.obj qdbm.lib
+
+
+crmgr.exe : crmgr.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ crmgr.obj qdbm.lib
+
+
+crtest.exe : crtest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ crtest.obj qdbm.lib
+
+
+crtsv.exe : crtsv.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ crtsv.obj qdbm.lib
+
+
+rlmgr.exe : rlmgr.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ rlmgr.obj qdbm.lib
+
+
+rltest.exe : rltest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ rltest.obj qdbm.lib
+
+
+hvmgr.exe : hvmgr.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ hvmgr.obj qdbm.lib
+
+
+hvtest.exe : hvtest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ hvtest.obj qdbm.lib
+
+
+cbtest.exe : cbtest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ cbtest.obj qdbm.lib
+
+
+cbcodec.exe : cbcodec.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ cbcodec.obj qdbm.lib
+
+
+vlmgr.exe : vlmgr.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ vlmgr.obj qdbm.lib
+
+
+vltest.exe : vltest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ vltest.obj qdbm.lib
+
+
+vltsv.exe : vltsv.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ vltsv.obj qdbm.lib
+
+
+odmgr.exe : odmgr.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ odmgr.obj qdbm.lib
+
+
+odtest.exe : odtest.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ odtest.obj qdbm.lib
+
+
+odidx.exe : odidx.obj qdbm.lib
+ link $(LINKFLAGS) /OUT:$@ odidx.obj qdbm.lib
+
+
+depot.obj : depot.h myconf.h
+
+curia.obj : depot.h curia.h myconf.h
+
+relic.obj : depot.h relic.h myconf.h
+
+hovel.obj : depot.h curia.h hovel.h myconf.h
+
+cabin.obj : cabin.h myconf.h
+
+villa.obj : depot.h cabin.h villa.h myconf.h
+
+vista.obj : depot.h curia.h cabin.h villa.h vista.h myconf.h
+
+odeum.obj : depot.h curia.h cabin.h villa.h myconf.h
+
+myconf.obj : myconf.h
+
+dpmgr.obj dptest.obj dptsv.obj : depot.h cabin.h
+
+crmgr.obj crtest.obj crtsv.obj : depot.h curia.h cabin.h
+
+rlmgr.obj rltest.obj : depot.h relic.h cabin.h
+
+hvmgr.obj hvtest.obj : depot.h curia.h hovel.h cabin.h
+
+cbtest.obj cbcodec.obj : cabin.h
+
+vlmgr.obj vltest.obj vltsv.obj : depot.h cabin.h villa.h
+
+odmgr.obj odtest.obj odidx.obj : depot.h curia.h cabin.h villa.h odeum.h
+
+
+
+# END OF FILE
Added: box/trunk/qdbm/misc/benchmark.pdf
===================================================================
--- box/trunk/qdbm/misc/benchmark.pdf (rev 0)
+++ box/trunk/qdbm/misc/benchmark.pdf 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,588 @@
+%PDF-1.2
+%âãÏÓ
+3 0 obj
+<<
+/Linearized 1
+/O 5
+/H [ 997 194 ]
+/L 52196
+/E 51873
+/N 1
+/T 52019
+>>
+endobj
+ xref
+3 30
+0000000016 00000 n
+0000000944 00000 n
+0000001191 00000 n
+0000001395 00000 n
+0000001539 00000 n
+0000001745 00000 n
+0000002240 00000 n
+0000002419 00000 n
+0000002440 00000 n
+0000003096 00000 n
+0000003117 00000 n
+0000003679 00000 n
+0000003700 00000 n
+0000004424 00000 n
+0000004445 00000 n
+0000004888 00000 n
+0000004909 00000 n
+0000005424 00000 n
+0000005632 00000 n
+0000005979 00000 n
+0000006000 00000 n
+0000006581 00000 n
+0000006602 00000 n
+0000007090 00000 n
+0000007111 00000 n
+0000007620 00000 n
+0000029117 00000 n
+0000051731 00000 n
+0000000997 00000 n
+0000001171 00000 n
+trailer
+<<
+/Size 33
+/Info 2 0 R
+/Root 4 0 R
+/Prev 52010
+/ID[<21b37df8f836949f6a724722f5ba950d><21b37df8f836949f6a724722f5ba950d>]
+>>
+startxref
+0
+%%EOF
+
+4 0 obj
+<<
+/Type /Catalog
+/Pages 1 0 R
+>>
+endobj
+31 0 obj
+<< /S 36 /Filter /FlateDecode /Length 32 0 R >>
+stream
+H‰b``a``~Ê Gw1` PZ ˆ¥ ˜A”Ÿ1ñ
+ë ÙÂmX¬€¢?L00|gØÍÔÃÀ ØÀÀ`•ÀÀð%øɆ&†' } ¢ÔV
+endstream
+endobj
+32 0 obj
+90
+endobj
+5 0 obj
+<<
+/Type /Page
+/Parent 1 0 R
+/Resources 6 0 R
+/Contents [ 11 0 R 13 0 R 15 0 R 17 0 R 19 0 R 23 0 R 25 0 R 27 0 R ]
+/MediaBox [ 0 0 595 842 ]
+/CropBox [ 0 0 595 842 ]
+/Rotate 0
+>>
+endobj
+6 0 obj
+<<
+/ProcSet [ /PDF /Text ]
+/Font << /TT2 8 0 R /TT4 21 0 R >>
+/ExtGState << /GS1 30 0 R >>
+/ColorSpace << /Cs5 9 0 R >>
+>>
+endobj
+7 0 obj
+<<
+/Type /FontDescriptor
+/Ascent 985
+/CapHeight 0
+/Descent -216
+/Flags 34
+/FontBBox [ -165 -307 1246 1201 ]
+/FontName /ADKHFG+Century
+/ItalicAngle 0
+/StemV 0
+/FontFile2 28 0 R
+>>
+endobj
+8 0 obj
+<<
+/Type /Font
+/Subtype /TrueType
+/FirstChar 32
+/LastChar 122
+/Widths [ 278 0 0 0 0 0 0 204 333 333 0 606 278 333 278 0 556 556 556 556 556
+556 556 556 556 556 278 0 0 0 0 0 0 722 722 722 778 722 667 778
+833 407 0 0 667 944 815 778 667 778 722
630 667 815 0 981 704 0
+611 0 0 0 0 0 333 556 556 444 574 500 333 537 611 315 0 593 315
+889 611 500 574 0 444 463 389 611 537 778 537 537 481 ]
+/Encoding /WinAnsiEncoding
+/BaseFont /ADKHFG+Century
+/FontDescriptor 7 0 R
+>>
+endobj
+9 0 obj
+[
+/CalRGB << /WhitePoint [ 0.9505 1 1.089 ] /Gamma [ 2.22221 2.22221 2.22221 ]
+/Matrix [ 0.4124 0.2126 0.0193 0.3576 0.71519 0.1192 0.1805 0.0722 0.9505 ] >>
+
+]
+endobj
+10 0 obj
+578
+endobj
+11 0 obj
+<< /Filter /FlateDecode /Length 10 0 R >>
+stream
+H‰”TQoÚ0~÷¯¸Çä!Æg;¶óXJ†2
+¶‚§J«ú¢Ð²®T#Aýû;§$”Áh§HŽåÜ}÷ÝwŸ3¸¬SXÔ xêôÁZ/Ø`<G¸¯Â
+˜Q\ “Ž´U\hHPre`S±%Üê Þ®ÑeiÅ©SŽ†²”æv—6ôlཿd¨ª çug¨˜¬CJ
+üíwÂàÛå…EÃj½xx*7ૺç%Œ†nž›‡jSÇþ'Khº· }$k,>ÞÿW
+ľçvgRn õlŒ&í^{ŽfU½ýÕt5$ZèÞ¡ÈþëîPüuÜŸÌÉxfÚúí†Ê+ÊëVû¶üM4½˜äq’Ê”òùå¬ø拯Ó8qNf\F׳Âçà‹†( Ï¢Y~1êNB3}*¾ä0/~äñÿÌrÏ~³¶ DÉ`«°š ƒtÁ ×°þøA¥x&wú\…Fðê8>8ÕXwOC®“¾Ý Î:hÍî ·«Å#ŒÊ¦¼+ë
+&庼¯6€Ü‘ÉõSF„± ò»‰(<#i²$Ò
+‹ít)%Um+HBJiRº \J—µsš¾CUʽK¦ÕË1QºrçXJ vÛ±$lÛ³´Üe-³0s‡ZŸ£9‡&ŠÞMÑ|{W7«fÛT§d¥Ò'ø’ñ\{+DƵîUE‚„ã*"w‹ž0ý6¬=aa¥: <>O8ÍìÞ-ãé÷ÓPg˜)z¦ô?Þ)YØôpÒ:ßþ` 6gDé
+endstream
+endobj
+12 0 obj
+484
+endobj
+13 0 obj
+<< /Filter /FlateDecode /Length 12 0 R >>
+stream
+H‰””Mo›@@ïüŠ9‚*F³ß»Ç`zh¥Vu‹ÔƒÓIå6Á*F•òï;P ÄN—]13ûôfv¡x ¥2 9© (¥PäQ\äYRüŒ<’â¯_ PI²`¼AëIAñÈ¡ÍáÏ¡|€¼lËïå©„¶ËlŸÝ/”%ÔÔgkÞÿËÞÅNRÃåPÆ’ ’T{í1ÄF±É·ÑJkž!n–Bψ‡ú 6ÇúÔ–u;:³Äé$:yæ$ônää
+œ*5¯sfËœ–Cià̪æWõP=AžFÍ?–𺅟i”#žé7ž¦0h|[D¿#kø`F[ölIIÌà¡©¢¯PGYqqbŸÑ Èå¤føuÛ‘qÁíu<×$ëüóø<ûfEzóe³$£¶go mª
+n>½ƒã=l¹ÜÆå鮪÷‡ú ›}ÕÜ&K†f ;ž-55P ó£!©ékCŠ Î ×k)‹tîåAóøAÐçù’ Î8äk‚ZhÊz||QÑèýê.„:7ΣŠ9dš Á‘r î«w‚‡{E[¥ 'õ’š üoW¥B7> ¬rz>:¯SWáû³t-Wørá—É×éæ2r¦›ÔL·0t ô_ P'6ì
+endstream
+endobj
+14 0 obj
+646
+endobj
+15 0 obj
+<< /Filter /FlateDecode /Length 14 0 R >>
+stream
+H‰”TMoÛ0 ½ûWðÖfpùÛémi»ÃŠ‹íÐÃ[©µØr É[Ó_?JŽÓfëR,A Z|â# )AñÉ Bc ø ¬4!iNSˆÓˆä9 h½Ó+Å65t+X¨®äZ ù …hù¤øáM)¡aÁø_\ Çï`ôMàðÿâNCJ’CðÑp¸Oo§{€|=ÑWƒ½kæžYuCŠêfIF’p§nQ
+K.˺ej
+†k¸b:(YSö
+3 6Ï,îOg³ï'Àd+˧‘Ï’WÌ°%Óœ¸B`ˆ¦C‚_•0.Ì36âø”RûÅËNUš ÜpVý^qSÖÀšÆò£÷hKU¼óNçq˜šÃšo]rÖþÉšÞ%ÇvCbÝã{>]n±Nmòi`¾Óáœø{;|aG'„ìI‹^Ú<7L±–®ôžèb~=°pcÓ¯„Þ4lë2_Úª6\:Õ2Yò}´jnÏà³ý#„$&Q€Õ(É .¿áNï6¼õaÁ¥}C@²«Op¾¸ó! a|=‡›÷×>`åzÁ*(âpd9ïÚ@&ÑÉ3x(KˆHDBlmïú<ýÝO|xhÄ²Ä Ði7æãqÍw§•ÚÓšPJòx7OwR× ;(Ø4RK×ÎÑãÆ=kÑt`u×/%ˆb7(¶÷C Üqke,™ÙÔíy2+|+Y¯¹Stœ;|©º5—ð«ÆGkÌÔ¸?p½;šËÂKC2 ,'%)ž¦YŠ™Ä–„@oåY;H o§xHñ† Ípí v ÿ ÄÉ í#€$J0Ÿ# waŽ5Ä™ÈÞ2;pÇ8Jù 0 åt‹c
+endstream
+endobj
+16 0 obj
+365
+endobj
+17 0 obj
+<< /Filter /FlateDecode /Length 16 0 R >>
+stream
+H‰|”»NÅ0 †÷>…g¤câĹÍgCb at g`@L\Hp^ »4 AÔêP©þ¿Ï“–ç%,… ä³Çêáà°&ø|úU§*w£î2&Þ¯ÇÑ*—€”ŒzvZدǂÉx½È Ùê -½'}´_wó<^ç
+:&àÈmL$ClM^ ÏŽ'pëu:
+z%è 8]—/ ×pwïàqIÚ…ÀË7™¨«°ºNËÇBÒ‡4#só?
+Šukv ï=À¾*ýö‡xËzŠ[ß{ß³†g"šÇ»(31<-Ð=3Ñ=0›ž-0<Ñ=Q¶k±<[`x&¢{dß&gy¶ÀðLDóÙ@ÑZ¯èž™èž ÇÌZ¯ž‰èžD¬9·ÀðLD÷”ŒÞšs ÏD4S at o͹ºg&º'Tt¦g ÏDôsá³@íi(º¿LÌ£ #TÒßZªà“œŠµÉ…vI¼ áÖå[€ £µð
+endstream
+endobj
+18 0 obj
+437
+endobj
+19 0 obj
+<< /Filter /FlateDecode /Length 18 0 R >>
+stream
+H‰t“»nÝ0 †w?ç o’¨µy‚ C è$Ú}ý’ŽeË©² ãüüÄ;a³jØ…A¡ÀëfÐÁ¤‡o§0€çíiûµ®ØêP Y€KÇ&ðûçö ÞOàôs1üTùˆ^ ¬)’÷‚½¦#3Åo´3r…ÚAjã!‚Êè×ïóãFðÂÁ_ØX …FPuwS¢™Ãÿ>þüXú0‘|fžº!yzøöpÙ)¾ Ô3Ã…]»buøˆ/Ö³JžøˆYU±=ÿPöID¢e™¢jÁº¿—uŠÂ)VF[Û©¦gqA].Y}ó/”Ö°D‚1[…„Í:Ì¿1„³±ìY¶Q.’ «qk¼íUÇ¢® m%C±úW@¥¬«44_Úm_;‰ÔêÒ®œe±•t³¸AúX{'ɸ,†K»xK1¶C—õKœMÌ‹iïíìËpÛàO·x¯±ûxûü$ˆ’Mž‰S„S4i&Ná „ò°'âl7âCª>§0;›k¹„ƒÐ¸Ñ2×r ƒˆC±¹–KD ‰Î™^ bl2gz a¬·®ÿ` åqÕ
+endstream
+endobj
+20 0 obj
+<<
+/Type /FontDescriptor
+/Ascent 859
+/CapHeight 0
+/Descent -140
+/Flags 32
+/FontBBox [ -121 -137 996 859 ]
+/FontName /ADKHNA+MS-PGothic
+/ItalicAngle 0
+/StemV 0
+/FontFile2 29 0 R
+>>
+endobj
+21 0 obj
+<<
+/Type /Font
+/Subtype /TrueType
+/FirstChar 32
+/LastChar 87
+/Widths [ 305 0 0 0 0 0 0 0 0 0 0 0 0 500 0 0 500 500 500 500 500 500 500 500
+500 500 0 0 0 0 0 0 0 633 637 664 648 566 0 680 0 246 0 0 0 742
+641 0 0 707 625 602 590 0 0 742 ]
+/Encoding /WinAnsiEncoding
+/BaseFont /ADKHNA
+MS-PGothic
+/FontDescriptor 20 0 R
+>>
+endobj
+22 0 obj
+503
+endobj
+23 0 obj
+<< /Filter /FlateDecode /Length 22 0 R >>
+stream
+H‰””=oã0 †wý
+vð(‰úë:(z@ \" CÑ©¸E³ô†ûû¥lŦsY é} Š”)E+2,M6C¶úS‘Ë€“Á[}™ÄI,™ȉ$±ð&@ ‚X…F›da§E°á:àØ kK$i.|äê’fa%\
+à‚ š ˆhx‡’˜AøNÖÙAYÆ<~ýÝ ³ îfÜÔ9 +asfCMD"ˆ²Î&|^t@ #G5õ£ÒF—w• I#?ÓÀX>$ƒ&ùà Óå¤pòÿ¼)Ë\ÖåM½th¬ë
+Ó ùSîw:Ó[fXÅÞ#®SÓû ^»õÝvJÛ©ï_ËO¸´©”i` S-Õ"ÐR*×ý‡Ç¾|( Yï,sÁê2ªîiu ùulÖîߨû%*Yî“´Ê8œƒª$»Ùº’jhAüA(¹UKß
+ew{¼«H¹Y¥ÃÓØŠ0è.×»ˆiJ Ùet}¸‡DBª×+ò:$˜väëeüúÞ'Àó wÙ€ àKfC= ì 9¿G«å›WÅÚ‘¿Z‘'¾qõn&î‰&N5¿8t¸©'o{Þ(`;oë`{ÞºçÃCÙëòð¸oÛ@ˆ”µ|×-µ¤.ó?Fú¤5 ñ2éa;Šœß È»ú
+endstream
+endobj
+24 0 obj
+410
+endobj
+25 0 obj
+<< /Filter /FlateDecode /Length 24 0 R >>
+stream
+H‰|”MNÄ0 …÷=…×HX±ãümá£Yp « Xp}ìvœ6e¦ªÔÑôù}q ;0$†_˜2ccNUAŠýj «À÷ÛtžìùyrÄ ÉãYᑪ [໓DŒì •#/ ¯‰JÀ” RB®0g@±)Fõøìº- *|ì …œÃ¡b÷1®wÌhèɘø sÕWÌ`èÝâѦ\_1ƒÁ1‘¥ÞǸÞ1£¡cr at 9¨ë+f0tL«(r€¹ê+f08FbÆxP×;f4tL‘ÃM¹¾bƒc±6ã}Œë 3Œ6d`ؾÏÏú±e&ŠŒ¹W!–†¹®kœÍ¬cµ}ë\°G)mh84,4—yŸ l‚<$=|ˆ•0¤ÜgÈ#JT)†Æ–ÊrÚb’ØGmZcPœŒœ #Õœ‘d·TˆI›Cëñ¶®¬µ0ŽÝ,vqìh9-kâ[„¦å€[uUNÏ!Íôr!;>
+rž²Ì9Œº_¡=Ë}ß\?\6¬£ ¸
+endstream
+endobj
+26 0 obj
+431
+endobj
+27 0 obj
+<< /Filter /FlateDecode /Length 26 0 R >>
+stream
+H‰t“ÁŽÜ †ï<ÇL¥XÆ€ãf§ªT©+µÃê©ê¥Ú¹´ï/Õ d0ÕîÌ!øÿ?lãkœ'àà¬+0ÛWãBä] yÚ»:A˜QC8‰ÀIá$r˜« ¡žBVÄN‚‚îc'Q2„ ‰‡Ð‰à¼îc'‘Â\eˆŽÀëÓáœé<äûCˆè @Æ·RdÉÍšh‚"¸ÈSMPDd(¤‰&("˜€#V¾wò.4ÐE¸ y"š dpÓIš0ˆâåe) ÅÃÏIwÙâᳮϺv›û0Ï÷p3[5Y®ªEù GEºB¶„R>£·õnððÿþ4ß¼¬TäÒðâP~—Õ–¤ Q”®?yaòâñD¦ÉË“WzÄ%µêxùQ?›´_ýÁc±U’Ê"ËÕ¦v€åëuûr©¿
+¦ìê€3E[¯fy–wi²nÝZßØöé±-ç'«^·÷v=7KÕbN‡³õM
+ç9ßÞüºÕõéö¼#õþ½\{çJ’ï¿=]é[>VÃ$Ÿ³rÙ ÙöH¦djöÏ/¹ ÿ åè·
+endstream
+endobj
+28 0 obj
+<< /Filter /FlateDecode /Length 21406 /Length1 30692 >>
+stream
+H‰|V
+pTÕþî½ïm $²‚yË#AHB4D„ðÈn `Ÿ»Hénþ ?1I#¨Ó†ò×¾P¦X„–ߪ(¼¤¡¢0t*C+ ëÔ heDÑj¦yýÞ&a Ú¾›³÷œ{Î=÷|çœûò $¢
+e÷ÍÎÉ(/Gpufe]¤aBÕÑ ±•MÆó“ §îo€žVÓð`Ýåñ å“ÿÐV>¸tEMϩӹýHmu¤ê”l»HÏqÏèZ.x[zžâ §<´¶®qùÙíçR(ï$-_Z_Á±íyÀb͵©‹,oˆû‰Æ³ O{ã¡H]uaÚé ДÅxnk¨_Öȸù4%¸ú†‡«j§?ž $Ó§'OýôS¬¯†PÏ
+p.vSÇœ«Ôù8_Sk‰à#|ûYJúe'ëü§ÛÔY#G‡‡;clûw[ð)ŧ¤fÒB<,¼IhÃýø!v‰¸F‹ #éÅß“˜Š{ñ0þA>„Fá×TL£ïW…‰9ø5á)œq~‹8x‘„*T£†>wãW0è#ŽáM- èt܉r´à8®ŠR±NÝBÑÛ
+¼‰w¤’V~Ì@V£M~O>‰ ž=
+K° ûÅu‘$.ieÎ0g£³É9åü §b&ê°/ãÇï9®]d‹÷•¡šœFg§ó#ÎÁ"œG«ãx %š‹<q@¼.®È9N¶«ûôNñÈäy!„QÏ3›ðVbZ±{èý$>Š1ôƒÄp‘+F‰ZÑ 6Óχr !säj¤jS¿Qç´<ýiç:ÅÓ F0 ¹˜€ÉD\N|ÑóZæf#†ÈŽá }Î5b‘h»ÄAqX|.óätù±¼¦ªÕj“: ë‘oÞîØÚñ§Ž+N g®sÄá½ ÿD Æ] “Q¥ÌIÑ~Ÿ'-bv ÁrâX [°
+;X™6ìÇa"z§ñG|ˆËÌG’Hwˆ{Ľb’˜'Ö
+K<ÍHN2GŸˆËRÉÞ²¿ Ê1œHKe¹¬—äV¹WÚ²CPeê¨zO}©Ý®¥h†–«= ÒžÒjoéóõ%ú*}‹¾Ë3$né¿þðÍÈŽ„ŽšŽ ˆ"Í™èÌs uÖ;ç™-ðv)x8zÕ@öÏPöf6F¢€yó³§bÇlÌÅT¢–7æ!VÊ•‹ñ1ü˜([˜×'° ›±»ˆt {t?^ÄQVÿÖð5üçØïs\ÄGøWð5 8¬mœH‰ÌD31T c.Ɖ‰b¦˜+ˆ*V§^<̱Ž£Ul¿`~öŠ6ñ¢8Å,½%Ή…#5ÙW¦ÊÑr‚œ"ýr6GP6ÈuòI¹M¶Écòwòmù®</?UšJW÷¨õ3µ‡c?‡Nª/Õ7´^Úhm¥vIëСûôÉz1³xR?¥¿¡ÿE¿à™ëyÅóϸPØIJÿÖˆ¨&Ô‹lT«—ÙYç°GL>9L\Ãí-í]™£%£D'ú¨Åfæ5ÞÙó"QÈ"‡7ø¸ªÑWy† 3n)LÞ ² Å-M¬»µ29_6É ªãT®6C¾Ý.ôHÑŠéj_æ+ñïàBácoe%LÏX `lþØ1÷æʽû®œ‘ÙY™#†ß9,#}¨9Äg ¾cPZêÀ)·÷Oî×·O’·÷m‰ ½â{öˆóèš’Y³(lØa[Ë0§MËve3Â…ÈM aÛàRÑ6¶Ž™·Zвæ¿, :- nX
+¯1 ã³³Œ€iØgü¦Ñ.æ—ɯ÷›!þãgÄx-#&$Rðù¸Ã¤Ôú
+[„€]ÔTkÂ~ú‹öŠ/4 «ã³³ïE¶9»ÈlˆŠ¢‰"ÆÈ¢@~T¢G"£²KLÀ.6ýn¶JDªì²ò`ÀŸêó…²³lQXiVØ0§Ø½3c&(Œ c{
+í¸Ø1Æ"ZŒhÖ «µÝ‹ŠpfB•YY´U$äž‘”iO5ýöÔG/¦dgµ‹¶9A»ga»ÀœàQ”8ÍÑâf¿?äžÖ§0¸îfóTeR®hYë {gyðfÏý
+…è4;«tVÐǨÍ@«á˜Œ! S‘’à Ý5f'àj3஄vOsŠYk-³X-³Vø^XRpÔy%Ú4}ö¤T3ñ§EûÁšµâPqQ|«&;+êMêÌtô¶Þ]LBâÍLõ
+]Œ‹™» £îNµp#2‹Ù"¶Qi0’ iËô1îOõX•chÆ'$˜ÑEÌ_Øòæ»…ÐÓ½¦a}6‚yé³[W"]+žtïpY·]n´ õݼ™iávJ\!KËÈ&Æä{²³šìR³ÁkØ¥LÊ‚ÜÊÏaÊ}>·Ê-í¨ `7—;e©/ '3dË°«9ÑIžëjš»57¶‡M¶óa¸ŸlÉvŒ½½ýûjómÑÿÿ¨«;õ¼>#ªééVY0#bµ¤f„ÖKSÄ«hYE¦Qd…H»Ó\a^ÓŠ––Z
+p7¤vçDKª]ЪLª=ª3vß J•¡NN¦*r¥³ÍÒòùA#`…»
+×µ2&&¹/FF÷AG€ßz3ùÿçç=¿ŠÅ~Ó#·{º–äØ.Zˆ#jêµeÎym6yžÅ"Ïð=ˆ%òY¬&õäÇc…vÚ¯¡üSÎ{ܽ´¯&] …I©Ž´‘ôÒ6Òã¤&Ú¶‘6ÓG‘ëÇÕz”ë "A?å|ÎïÆiߣEœ×êó°˜|-c «)˜N ƒ)ÏæzAÜz a|s(×S.ãÜÈ=yz<ÖPΧn4}~w¦¾éúnçºöJ÷Už;„ó휟Óà\'¿² ƒ—ñLжü‡órâºÂø¹ó^¯×»8Œ ˜C¤°¥ ˆgð` %Ýo ãÆ»<TQÜȆ5´Po+Ùª6‰Ò–‚Á#Š ……Ô8-6ª,Qd§
+nQ+ ¢©‹€R‚*ÕýÎÌŽ¡¦ý£]ëçoæÞ;wÎ=÷œswa?QÚK•—i3ž‹H4Sª îUÎÓl\ÏÒ'büMZŠqË1þE<Û€g_ÏÒvøq6úæáÙ"Ø5jˆ|•6aÎW•«øfqœt¼ÿ%wÝÎÚßà5{kbû³6=…c#ì{Ç>—`ìˆm£a»ž„h¾<ûšÄ¾$‡oâ¾U¹*À¶¬k1ö¦„1ˆžƒÍ—À•-4÷عZ=‹õm!2ðõXåµÉÈF_D{Ç¿Xš‰;I¦ôU
+hø6«ÕSÛƒµl@ìu9ñ°…ÖÂŽÙÐIÊgT$o wñ®K~ùx¾bÿ mâ¢6?0°‡ðy¯_e˜c<làý‹ðÞ‹uÿ|sçJùˆß|ú‚㟛ˆC¢“ïá™åü<æ}>» mëAü=‰³GÌjü"Êî—‡³ÇéE €|w9ÎÀV°¼ !V‹0~Ç-ÇÇ(Ç Çæ
+þÅNìò:ÖÑtìÉølîìÀóßÓ@ ¾øi–¼¼áØe[½¹9Æ8v<uãœã_œ¼^Ž/O<œC¯² ìŽ3Oõ³ÈÇ›4“s‘óU ëªt˜fq ³=Uÿ0|›ýÄ{Îyʹ2¢ž}È[Î O @Œz:⛬ªÓ)¡•Á÷@YD1ùöp}Eù1ô%š¦$8æ±÷?AßDZ|(Ï 3¿´‘¯¸®€MFéýŠØ¦^ ^ÇßWè-é·TÇàú t|6ôØgÿîÛÑþòÔÛ—§5ŽëŒ§¼oRëðÙ¬ÞÊê=Wi&×B®G\¹&)?¼#¾¯Å\'¡Y¿?åÿQ~ íïÑʵ”ë™Ò~ìUvýN=@NsMàwÀ7ˆ«á?s?×}mΤu8+ðÅYÛ
+ºÁÔ„ûýگ<„½wÜœã=f_°ÍÙ|[Ã㼺¯>rⳑÇpýçñ¼¶M» £æ ‘ói•“lóvÄÅyªW>¡¥Ò}ÔäwShS¸–¿Lcœz„~¹mƒuwƱ7¡^gkºü ·–ªr`áðïû x¶1Çs>Âý!Ò´í4O
+ÁÞ»¸þ¾ë¶9µ~+-à5«Ç¨ÅÙ‹lmá½5.ÒTã%ô98Ëò0fÈ
+ÇWçY]iÕ\©Í˜ƒ O5Z/Ú<?Muü•õ£ç+õW8kØW<ç#ªf_iG1æyR
+Ìm\ÂõQZ`”cüvpõöô>=çœAÅøU…3M>Žu½I¹N\7âþ<-†M–Ãap1‰1ª‚k^ûm(ÎJÎù3œˆ!-í|>þ ìCL< ‹Q P¡‚Îż‡Q eçäóŽóóʇàsÄoöÜëäÃI<ã0ÆÏu“/Õ”þH–ãՋiÍ+p® ‚‡‘K—©óž¿ýà"Ú¸uAš¢\B=1|K=!íuñ®ŸÔш^·•y²ïmÿÕˆ,\ßäß{ʵ¤ÀtƒÄw½’BÑШÚÔ
+æ‚pû3
+í•Èÿ6P¢ÀÆÜMð×Þ%9”’^£4è = _zÍÎSk›ûÌøà‘Áëƒò[Í]}`¶—™æ/Ä^|Só=³»ÏìïW.0ÃKò¤/‘R
+î•,ÑM¶è¶Ÿ‘~TÛg&@õ_úÌ[L ³ž1ï´Al¿Ù¶§Ï¼_dÆÛù>Ç´ÖÛë¥õñ¼ úgí}fhZR nP
+¤Á=€åá„l Š:;ã·ñ3
+ LZ2_ZF¥ N~p4½’PÇ—a‰–¸†¹®Q ô‚ph# ÐR cÔ¨{âš=œs¢¡Ï¤ÎPg¸Óê´;cU†*ÕV¥]«Ôè`è`ø uÐ>;¨QC¨!Ü`5Ø
+±ªBUá*«Ê®²‡bU±ÆØ/<d
+ÙC²•gí<;87Ô¬]ö.ÉÚcï‘R»Ò»¤Ôžô é^§ðÝó{wì–RíévÉj·Û¥T<u! O_PLÅTK•RU1%S–L2…d
+SùBÁRQ*‡}á`™(“-Ÿ¬²í³ƒq—c¾X°Qn‘÷ËãE—¶ˆ&_Sp¿HùRÁ#"íKsO‰_G°Gd|™`¿èõÙ õ¯‹ß@0×ÔCþR=ì/Ó-…nûãzÌߨ'ü-z“¿žò ÑÓþSz‡¿GÏøûõ^ÿu}À_Ô‚z©Vª+ÞE™V¦Çµ¸n• VÂj™R¦6*j‹Ò¢ QŽ¨¹¦
+”ªá@™j*T;WcF5hQ›ûÕTàˆšœR;=j&Яö®«"*Ø©…V¬@Æå1ì•á.qãÌÊY+®|êÊo\¹ìʯ]ù¥+u®ÔºRíJ•+•®¼âJ…+“\)v¥È•ñöfèð7ðWpÜ·À'àèçÁiÐÞm | ì»À6°ÄÁ° T;ïù“ûºK®ô¹ò±= z|
+Ž‚wVÎããKw‰zÛÖÓÇô4öêzúzz‡žÞ¬§7êéuúTcŠ6&ÏEF¡1Î(0ò‘gä9†ah†bHrLdž‘£RtM¹ˆfz7StS8ópMI—Èy¥6£–”‹L~”¢kË 3ó"QTÕ™¹‘hF¯>-Ä5é{]‚ÖVw‰ |ßZœÉ_Z}uhnë¾bÖáÖ}554.òô§päJDc»Ï!êŸýP7?×Íoêh‹®ASš›ÒŸëi§©Pœ‰Ñ÷£_OL¤‘GÿÃÄÿå#–«>mPyÍÒ:Wß—ü909Q<¹¦|\¨i±cÿÂÉ…{‹»ÇÈ ©Éä–”g€»f,™±„»Pä¸+ÍÁlWáÞ…“‹»Å±lWÍcà@JîL&›wâ“L&E²9Ù\Ÿ¬oF>ܺ3ÒœÜÙœ„…ù¶yg$²3’¤Œ´|k& ì×™pɲŒ¶|k—ôõlÔÔH<QñG2é4í¥Žô:Ñ%EþEuõÇ6qÝñ÷½óÝùÇ™»³}wq~Ú v~‘8þÝ€q.ËïAÚ-!Yjp5ʦ(„1Ú€Ú”@
+…°t ¬k‡ÒM]†Ã¦ñ#h‚vUò6¦¶†ˆM³Zi)ÝT|ì]ZèfKß{ÏÒ½ï÷}??üÞEdÄêûäáBÈ£Þãï¡*-z+œ‚Spá x; 䕇
+…¾À÷©+Øk»É „©AÄ¢’)lè»™Ö±Èð"yÒAVºV2NäÏÌüƒÞXª²U¥¼VÒ „ü>I´1äTXxÊWe…ñJ+^)ŽD|® ¨·ðÚ:‰³ô ²¢²Id4"$Á©Ø¨Ýˆã9WÁm䆹1Žáþaã ÄÒ"xyÇ[2M¢Í"ÉŒ»ÐM–pˆ8˜yô`Cóë2Ï>sàоW×Ó3áÑEõý÷§ÕÒGG7ƒM×.CÇ™ñ.gv/enUÊ ¤±“v¦„,a:‰N²“éÔï%÷2‡ÉÃÌ)â4yš9¥·hõY©Ý¦KàD 8ßMXÁ:?A¸²Þ )ËãÒbVI–,¢`h¦0²„»°àÂÓ¯&]?Í <ÒÜxð=3öÓ깫¿S~î»#×UðMÿ‚ŸÐj«CÇt#º¿ ŠN!‚€I 0pߘ$@É4PiR é ¦„iÜô “®ÕtÌD˜.±üƒ±ÅH:2<‹© ™Ç(8¿àBଃ{àžšu@Í">‡5sPÍ…¿álƒê›ð#ôg܉µŠ¿
+£1LœÜ-Úˆ€$- HXæÖ±=ì0{‡Õ
+³cìev©<<$Ø)hÕz°3Ö›úg¯¥Ò#h ˜áì…¡`À]ÀÐùOÈp~ùU“I²²e5µÁot¨oVü°‡çtd½Ï[»¥sWóqƒÈ%㸄ìÉ 3âæ¯Iè¢È. ¿À/ OJ[ïê5Q»ÉøIüÖ Íú;V¢ØÑm"ç<†*Ã:CáŽ6ŒÆ1ÆSÄ„Û„ ]â¨K+è«â ÇW_ï÷54ÜnôýþF|ÖxþѼ. 3ÊB£Jî9D5znQâc¶ÏÍê Bת‹ë:.I4)²eŽ79LŠ‰Œ›4|Ž-!D›¦ˆF”½$ô—ɵô5}Š²—Ù-» ®,°¿Ì2Š$°Gâ»7
+‹C%Ž"«‡ÿóº}Ÿ, <*È/t |8$çÓoñûÂ!'ˆÈôþ³êýß_NŒö}0püÔ¾ýõß<C
+žü: Ž©Ã§§al˜y«1îT÷ª_ †½0FƒÅöq‹êFjúe°´Y“ )…’ÎЦg%Ð_73×a£å:ov˜[Íqóó¬™F8fs†,`ôÆæÓóü¼EcCÕbŠ`i6YZ¢3è÷…‚BÀ]
+Á(„‰‹¶œ’kiï[_ùòÎ\k¦º¿pdsbK‹·˜¼STg« áŠâøÜ0 tIi9ÅNêo ?šþäù»çsô™þ?¬Aï±›ì —¾°<ˆM¦¹ +Œ\§u+ÿ‚uWÆ@ÆñŒ¥7Ê •nУž ÈÉa„‰å•$¾}mdXZ·JBdrÙ *‚¢j#щ2‰dƒ)||å°•"X¯ˆz½ƒQ"Îì`Ì1fœ™Å2×ðÐðð)~1•J÷¦SÚÆ1и1¡Ò[zcÐ[îô…Â~ÒK8
+š<òiwE~BÅ(ø}¹€gx]Š,¶(ŵŽ‡ (ø—·^êzÇÌ=_\ýì¶[{úf¶|í]ê¿7¯ý¥ú\Z at _ûÜdóÞ#û›ìf³1cƒ¿¦·kóÛÝ[Ïoª;í^óÑ·4taf‹¸—ÍÔúÑ8ä;ï#2èFßzj}ýVßKõt« •¹ú¥ýemNg ÇÕUE×E‰hT h«£²Úôfä-*ZUík<Ÿ;×. eÊjvWx |<L†·÷¯‚U±Ùê»ÕŸV“ÕçwÉòq™”·÷[À²ÉÁ*ì§,‰ð#Î&Ø+ì,{—e–Œ¤eI%ø»¸ùt wRxì*»‰gž/ɵ$£Êúær¶ÚM¸²™Üvð×{ÛÁ¹í¤§ÙÛNæPYíÈ×hLJig;@ESy;jFžr ûdiäkIíÃ
+¨,!¢YW9<ö.ŒˆhË»nO|-°Ü¾†ò1á¯Þ"f‡[7ï9|çÀÕxóSþ¨£¨tõK]oŒîëÏ÷[ryáŠ~EÙÓ×3àE<¾ª5/ÞüíÝ>âェ·œ|gëË '6¬é¨YéòK²˜ÛRvzâÝQ=m·Š2ATW½}¤5_rùBÅO©}á™]ïtŸAع¶cø±zE”n*R¡X$…Ä&©SB'³„D1 5J–,ÙdY’DDÊm:›
+™—™ÚÌ, ˆ2'‰g›ÐÃö³Ÿ°X ¾(° ¶ŠÕ±I¢K1Ò²ˆh©•2 · ½QêÆ„O’ò@
+I‚V‡¨ˆD\Ü!&ÄqqV¤DM,v
+æE¬—t„Ÿa|-•ü|oŒŸGU‘*ü£†¸ö¨¬<X^zð•ëÚß*Š_ð‹âÿ˜Èjp
+à$Ùê…ÉQÒ–þpKÀ›~©9ð1ÕýpmÛ‡Ù‚ÜÒ²¢d:NœLÇ4îcî[1÷Ð+•
+ dÈÅrCqcÙú⎲meÌ÷åÍÅD¾^AJU4ªðŽ vEÛÒ$JZóòìá$Ì)åÞ «ÕN÷#Ž3Ï3ˆ*°›<‡¢
+½re~¶Þ`(Ÿ‹K¤é¿l—}Pç ÇŸg÷v÷îàv÷ööÞ–ã^ààx9ïÌÖ7`•8¢ä**(hEßÐh¸¤XÄÌÀ51*ZÆ ¾ ‘ 5 at lÓ©Vmì$˜É`L¦1alÒ–DglÒoíó Z¦ÓàŽûãæ÷û~¿¿Ï÷;“Ê„ ’bó…MÍI0©¦ÑñD=¤g“Ó&Û¾°Ýµ©ªmÛ(~cÃN˜÷È (RbáX¿ Å<§è_¬~ì>6ƒ¹ •òeîá/3lˆaùDfñe,Èo@´x-¾
+ÑŸYæÌ
+à7"sLJ~JìèÇ„f‰õjYˆN ‰CE0ãrÿúirRæy†ÙÐë)mqé)íÊ£µÛÞr-Ùÿäe§ŒyöÖ¢Ù¹ \^ohçÒÎc§[BM!Æm-éñ”™¥öÎI4h¸ä- ÝÊ+©©Êƒ>U¾ýݾå+Š 2¼ÞäüQÖ¯Ç.
+z½E~›Õ›©ÓÂÔw°º·¡
+úÐÓ@8)W]án²7¹oR~`)†ÍÏcɼ¬â¬Ê¬ÁuLmÖ†`]ŽšÖæÅV+©²/=]“ܧ§ü}]‡£ÛA8 5ÆSÙc4 ^@Ë4 Ëèj:B«è!Ø,[½cøÊÉ:²Zר‹è¢º ]?ºtŒ+9wrMaL><ʪ‰ð£œB¿xc’ ♚¦fŠÕ Ìž™¹9 <Uµ)8FLSs†EKϬï½|yÿ ÛNBªýµu女³‹-¦wáêžÍpÖžî#;¶G÷Ѻ•·Ï|¬|û:Ì ½
+Þ+›W5wéüYöÔdÑ&˜w)7‡2~ùöÀÅw/žÄ.8Šf˜Bµ ¬ V¸ZÖšVfB¯3’˜ÁuõÂXë’,•IÕRTê‘hiÖÈLçÝ\?÷ ÷£\”åHŽ³%L]~ÞH´÷‘X \3]FZ –˜’TžP]HDÉ_yhhÛW„ÊWž7;ß·|SÙþØ—¡'ÒS-Vß=å¼2úá‰e™ Kæ]ú-T½¡<s÷šÝèÛ‘Ì ø¡EnÌôÏÖéVØV8+“*Ò*¼Uþ:}½mƒs]rmZw¿ŸoRïÒírïJÙíoÓ´Å·é^e÷ŠûÝmþ!Ýñ‚eØÍ«u05>5å ö)v=\G4iv²ÛÄ^ãUc IBmn…Jç¶ •?"(}VöùYÑïgÙdüÔÕŒ½O 2>~Â%T5›:LÝ(°ýe9~VU€Ä})BªHLŒqZWU#A$9ü«§Š×:µ²–¬Ö6j#Ú¨¶GÛ¯ Õ2ÚÇŒ1¦;yÂyú*¢`}™,Š €Ò ¥åóa¤º0Ä ñèNå¹'Œ QÿE“™…åFR=íë÷~ëÁÏ”N\Ø3¼Z¹R¿¸½õ§¯5/io8ÑÕ;ÔMêç¼ßr¬éè_l;7÷ù¿u~^qàé‹kCÛVíÙ×¼®¶éëÐÁMí]]/okB>ÝŽ¶äF[rƒ xK~6?®À“—ž (ñ§W²I+<éõÌFOmzËæ` [GÕëH“‰ä±M33ù”>
+åìãu‘Äh"‘¶ÖNÑÀ7†ÁÄ—
MÒ:Gؤic<íDæ%«éFdÝ(ÝC÷Ó£(´ðܲ ³wÿ-
+ñ¸àt¢”}ˆžgÖfÚ’B.œáNò¹yïüäàß!wñc(v>x³ólñ쬅Ãü’º7×(w£Ñ[ºŽ YûÉ ®B³rRù½¢(]g[üä’ù>WšU°¥oÌ
+âÚûÏœºðöék8Ýza<¢z€ºŒäË)ñ÷ú È1†ó ËŒ›2R ÆntPI‡WY¦ê¢OTA'xT.àŒnc˜Ùs Á¢EÁ`qqê:ÙÙ%‹UŠ‚9‹Š³³ÇŒÅAT}PýAßâá= ¨O¨;À\àšü¢dÏ°W•öJg=¨êí‰ïrpz§Ã!D ÁAlà "Ç9€Ý¾Upˆ‚à° ¡Jp9Ëy½^‚€¤Ô–&¡\R:ŽwQ ´C/ˆv'eu8¤BfÓÌt3*æÞd-ð.MÑCD΀õ3Éö7¶
+“÷m ½ K!nº!Ä SÀ <>…> ~Élå/_fB!|Õá#t˜L\Ò5¹ØÉGè "’ ®/ ε¾Ø€ÿ¹T‰y16´(¡ª2;öuEM¼¥ü@ÄQwþõÞòÒR‘MÔàœÁÅYE¥¥¥õ䇱M8™V£¹µQW¤€,0(»KÔŒTž•š
+2ÊùrNTn-㱩T´h¡vÒ²E”<] ÁÈ5ðb@,IYì#’â0Q ²qk #\Š!%#åê§ë–qhbŠƒs½›‹Ð9QŽÙhľ.¡°W‡¿Ú7MÁ´ƒLB ‡9@ƒ8@Íð˜C3Øàšº÷“-D•<£AÖ½âä°B<¸)XM’_Ý;yêÆ Ê_ûc⮼`và8-/w–¼Ú·sŽîùÓýG©«×½Ô ôþ*Úpz{©Û¬:–‘;~øìƦ¼do¬7v%-«÷ð9tÚ ~FÕ õ9ÁuY®§klõÎ üv} ßI aòô‡ÄCÆÃæ~DÏyŒOÑ+é–—¨ b;u„:¬;‘pÁ6f‹ßL6“ $IŽZ@Àc² >n'‚,T BTèha*2Ï 7i@œ
+–[Tåäôð]Y 0ȆjC¿A%
+i"–Ë–¸1^rJ£ÒÒ]‰þ‡ • 4L<
+\È„Ó)sª´€ÂØx¸½Oók^ÐœŠï<¡çÍ“Ì$¸§yÀ¤:¶¨PZ0¿f÷™‘¿üì…ï§;”OLJ•ëà ÀŒ3ÕƒÔ3 ±³ÿÞñ†2¨lUΟ;ðço”(ã#0<ú9l¹4 ¥GÒߤ? ȯË.ç,x(ùp!«¤;'WΪ=c9ç%‰gŸ¬KD¹†C·á?l—kl×ÇçÞ™Ù—wwfg×;ë}Ž÷ýðck0cØÇ.éìÖ,
+~4ZS³ã l"¶@©!B¨ÁMHi#hÈZAU«ŽZP+ÑF<‡”4u Q”¤˜5=wlc*Uë ï; æüî9ÿó?ØJòŒÿ²Òm‚$DQ j†UùãبÓêµm‘–Ñ„ÂÁp Lk–f±&ö#Ý(ŽáÖ–vSQ6Ñ=m4ñx Ì& V\¢] E5øöxn(¶¨)ŽrA5˜ðƒ³;Ž\_Zù»ƒ ç
+«VÇÊ"Ç°£<QU»ûç}͉Å{2wÙñÚ•S'¿ø`¤7WáWüAæŒ/O®†ž®éxù>êþå!ȧý@åY G^¹a¿yÐzÔ|Üúªû÷—n-ÇPîFFa™dÙ
+f%»Ê½Žiuw˜:¬ î¬)kͺÝÇs£Ï“ädµ[k¸ ~J¶ù„J!%4
+?N Nð X \QÚn—´ñ4U‰Sã n’}QØ•.Mó^É›ôÊÞ&o¿wÈ«¥¼<,ho7Ê I'pi Ï{E´Ncìñmó¤%ØïWŸHËîA –2T&<: º¢© ’BÉPWˆ
+‘£K^W›W¥¢e"¿ ôÁ1‘Ÿ¨@3ˆ[MÕòÿhá D4'Ô±«eÖ2XK¨oõ0Ôü´À2Ôƒ‘¬óèÇl?sÄ]}ûÀÔÝ@°oc [_ÏîÙÒ“¨ú=·hEÉä’
+‹“ñò,;>eúû¯Ã‘¿ +4Ÿ}¥ðžäG· )±s$†³Ùgó-8›(u].{’}†Ír;mƒÜ€í§ÌqV×Æ¢W¹×m÷YZÏ®„ã¡%>ÉËüÏð9pa£ÏZiMY$y5Vdý‹1-Š^ʇ zÚ‡?žáïDH›Ž¦ÉøÒ^ omµ‚õæ begÀ3VäöÚàù âR =и`kìÖ<_˜È?ÂI$hn%Cì Lâ&ºU˜8¨þ ÍâÇPnbÞµÕW4·ŸH}›v³+VhŸïìK.~½°èµŠHlÞ! H!]SKii᳃‡ Þ|¹p*¦(¡8óZÔOüü ÞÑ PÌ$õ7yù:öûš¶SÓËîÒœ`ÏDߊ){?ú~™QaÌö²ôQªÐQ:&v•¤ ŽŠ0G7ªbÉl¡OÑç@3sØ G1Ö® ¥>þtXO1ÆÒ@@rd¬ÃÒŒRö»ÈR—„¥ n–MÆ›|\Š'ãtS ÅGqÃ#Ó•'jȲEª(NäIמ˜AV«z0U]¤H¹ž3„ÊÂæp4¤/7vS\.E ]72Åøn*1×¥(u¦z$¨PT(êD€V—“ eMbêæ¶ï{u¡eÊÆßÚsmóæ«{Î]™Wóî\½0¹<Q®ï½<õÛÑK¨îÖ_°Í¯ÞyØ}èðÔí©¯§ NÝ ;q 4ØÐ >Ü—íí½ŠZnßALúí¥ùäp
+5ð+6‡lò|P!O
+6ñˆO#W–ùÜérŽËx2¢9£÷eôÁá”Ø%^ÿ)2¼(‰I‘s()ÛªQF£)Ï3P™é¯FMÕ¨šdàbÒ_ óµ|ž¸ ’‰ðÉS)ÈDb¦ÐtfÂÈ5ƒ Ô2DK$V|ÌÁîܚȮúp;Þѳ$„ß,ëÄ=Û"ÉY¿±p°SM…gŒ%Kz
+ƒY)¬gCK7±ãJ¢yj!ú`MTQì.ËŠÉ¿>Åû%`ý.Æ…¯Vñ¥ŠâtðõLYs‚Z æò+Õ ’ÍIÂÄqÆ´I~áeD·ÞÑsä'…nÉ–>µ® Z™FëëuhÄ„“d!‘Ô›Hc™Ž÷±x R}+ñÙçVÅGjoaÍž'¢‚[é·Ž¤&Gê,¢¢Dë˜v™áÝdP¢*x75*/\güÈø™qÒÈ,75™6˜¶›ö™Ž˜~oÒjŒ² Ó®4ˆùlChR[‚–âxsDÆ#–"!mÑñÛÌiž‘¹vxâ §¡ÔE,Ȩ́å.C0v
+¡[ïj´Ã–Ù mâˆý¼ý*ŒAö!ûˆý†ýžµ“¸ÝkÔ–>)¤”zâDtæ0¨æì[Æoÿ,ªüö7ÿœlðO&³K«â±£À"V6¹5 ¸ƒ,…q¥r>³7DtdEû4ЈQ7/¤èF¹‡»ävˆmÅضaL£‡ã
+FðÙq ͳ±4C³Å Ž“|:‡¨C‘vÒÅ™ß@ |
+¡šíÅÃPò)C»á”1,+Âí0-ú`¿„r¢÷J‚çôÉ ÄCA ÅmTœŒ;3ZjÕ:È“Á@X$¨3ÔBíLMÔÎ
+/½ ð/¬TT`øçhÍæ
+€@{‘ë¥hœ M Qt }Ó»¬|Áò…u»ÌŸŸ(|‘\ûߟHífÇ'»˜!%\9¹9V”˜Ä|z »}òímó4Š¢_ßG?H¬h®~êS¹ç¼u°ëY¶×¸ÝÚã<À 0î·ö;O°'ŒCÖ!§Ùâö€= ³ïµ-@”“{¬¿Ø‘æEÖàOëÀÝZx ^›vù“vÙÞd§íc‚ö$6œäÚD$fÚüÈŸqáì9q9´_Ž
+YÞ%¹’®&W¿kÈ5âºçú—KO¹ÈMn²<\†\´k}Nfí4·| ¯ê¨7™Ã ¸ª?ä2=˜½¨Y`m cÆÆfI#RŠP!¿,“8 îâÕèô®šêÌ6®ªi}°áùþµM»,UÅžŠ:`»~²oÉB¬(e=õ“ Ìèƒ}Ù °bö¶ÑŸ fp¾[`fø: Gy¨¹³Y³VXíÅ49ó¸ön-^c^c[ëÞIïdèvÀuÆÓ^4\°\.Ú._/9.».yþÃ~m°ó§)Ú~Z§wÞ¬ddæ
+C'™v3 [oòFɘ4vÿËw•7qÝá}{XÒêص´²¤•V‡×Zk©>‘±ãÅ€œ`1E&Øø™R è@0¥„£´@B'¡)ÇŒ›Bè'6±qiÔÀ´C˜IiÓ”ÀLM‡éÔ´
+$@‹Ö}»²¡É ýCowÞŽfgß÷û.<iØg@ÕÅ Ø{æÔÄ SÁLä/ÀaØEñY²)Ñß çÌ\L£FšÑÓòØûïÉLJÁSgÎ}ö`¸ïƒu/\$þzA>rýºü£‰s`Íõ?¶‰gä3òëò:ùÃÝ z ²J)’„š•§jV yMòµ®”Nç4)ò“—rBmuIòóä¡ ïµ àJÔ1#]0ƒgw
+ä :Ç Så%Hßëm¹MdRAV—L¾ÉÄDÕ ]n Õd ¹‚[ãm<
+Á5">,¯q&èGÔþˆÃ€ÓýWÄXi…Odýø>À7¼ó™üÀêŠ OÏ=
+Õ¾Æ/}÷é7ŽwÏ%&nž:Õ¶MÖD-vè(nüm1˜î82Ü·«*\„`ÈKr/q"îCæ Õ€“J³ð ]fmö=Ë/‹½Æì¨:n§ƒCáÁØûU_ £†Óúj4ØuÔ1Ê ãQ"¦E`Ø+¤ •`(*QFÚ’‹$âxÅ‚~‹®¸¿¢ÂiAHšô’qR" r\‘¤¬õÆP¿ËE(ÌO'~©5 ÷‹ºV²Ï-æ‘<g=XÜ<7Äkgï±áI[»mŸ
+Kµ)óRód^Ô²KgCçVù4KOfÂQ唩„9¡*ØfòÉ®%!jws~È¿Cºü°t±âZDp‡Ö‚¨ ._NJ½ ¥×ŸcË™©œJ"zœ‹èÒ_fFŒE¨š¥¬ŒÙ&•Â®†eò••Qÿ ¦-¼×JÑr°»û›[s:b§å £W Øq0çÌWJ
+º.¾ºéBwU5zdë·Ÿš‡Š‰õÕ,._½"T³=L=?tâ¿™L¿U“WÎÊo_¿,Ÿ×¬º3øø÷ì‘ÿ0=$Ë/å»6 ]V%:´¤cC[ëR[™÷¢éýpÞ §KÔí,Ðaèt ä0Q«7$ÖÅ°¬‹uéxŽ¶0åõøC© vnª@1m„¡™$ƒ1#àšä£éLÀéÛ9:miÀÅÂ?á>K+Í fGé5•.‚XèòÁ8ì’ÄñW"†ˆ´èÛÅ ±OüP¼)j“"
+Uª fBZzê^ËTF$!¦éÞIhÓ锘1¥Çº‹Ëj%„bMi#V†–6‚¢¨Ñhç`s! tù,¿] ˱2& q£qEX…@>ª4ŒÿG¼"ôѧŸ?7|~ï óÂî\Öí)ÎZ`¹]·Æie©y»Û7Zì…ßO/Ù ócûÁ¹“õß\?‡˜¸üã•›™@ý+ÃÛâþæuòÏÕ–ÜÚÞ³ =æ§,õõ‚?,éþí'^ÞYà *^·}zÿdh>HIâ ÿná?Vj¨1H¬ä\N4º:‰U†U®Äz׉K¶Kö߸þ ‹µöNÇ(tv²ìTƒ8 ‰Ø€}Ð
+:Qe ïÄ°3(m„/±F4µ›Òb^`x^Œæ®àø5ü~/ˆÿ ƒÿ?Í“ >ÆWÂ8ÏkÜ‚‘§p ©ébî~«®kúa :‚70І( ïI_³dn7÷˜Ïšqóš’hv Ù Û`Ⱥúgˆ¢KýLcP5'!úªúÕÚ¡6åGLk;BÿýÞ !åŸÜ©ÒªG<^ £Õi#PcZãS®@Ö¨W¯Å=
+Z$Cí[3ÄV
+PT%²¸!s>7 á³ sÄU¹œƒ
+6ÕeU=u÷ðwŽK[îO¬”ïþþÖ»½µµa7|`¬úŸ/ßüðƒ–ãIJé[&¯Y3_(é:ÖsQ~x`7Z1%ß9¹¹fM]QYíª±s« ÂeÐ òsiA þ þ‘[èµø"¶‰mr ÄßrÞÇ5•x9[îlÀX‚°c .ÆUr
+\·—;ÂMÃVÄÍnܺTʶCã2d§(“`©$
+ÐÓ 1Á Hê1ˆ+ÕÆNÙ¢” ‰0G™KBÊ2 wæÌ'!÷ÐÔ‹P*§d˜é™˜Ü dáQßW³_”:M–†o¸… ä›e‘püé%'róò ÷?—ä눉ôU0ÌÿÏÊ<Xrr]øI!°«÷®2ë%Ó“j®‹ G¥Š€±„,6–˜ªÉ…ÆjÓR²ÑøuÓ›Æ7MÇÈcÆŸ˜†IZ£/&Kõ u5Ægø¬Ž<PªVÔIåÀ õåÓÙº8kµ1ÉäPN m¹Ú¥:ð
+qaŸpTD ¯Ð#ô - a]ŒDU‚.®Äc¸L*©æa˜Ià|öþOWrƒÌwç£|ñì&tnEH2ºQ‚þtWiÔ–{êõË 9bµKÑX°uÅ΃òŸ %"±Ø÷`yÈó?ß6ÿùõWO¢ÇšZŠ õõßhO‡îlÙ»ùåå;°Í¹x6üôuÍoáÙÄA»”?ßúj{]´“ít†:¢¯Ø7q;Cßê
+†¹†§
+uÖÚ Gô=iL7:BAïÈô
+éU2ñêÄht5©cHRg‡B« vÆá°‚Á& ïXȤH©‡Ïõ:Row°°¨Ã¢>á½f
+ÃÑ
+†t‘¨®Ý ÌfEUR
+FQa2ÀòöPsôá°¾&‚“ñv/ð~,):‚ò#°"ê½Gƒ£Û±ÁC
+I È1°ñB9zQʦâ}q´'èx2ŽÆã_SæqmKïëH·°vøS¯Sj|v@ËO(VHØÕÁ…P@`š®Tð*‡[žýåþUd¦Æ̶™ˆFK—kgÁÁ–ñ È0+c(‚ºaB5€×Á¢¥ƒwøÅ* ˆÿåºêc›8ïðû¾w>ßù#g_ü™óÇ]b; 8¶ã'€/KB Z‚ihHÚ Ê ŽZ… Ñ6U”‘¬E¬…?í«ÛªI Ú-0T˜J7¡UK
+-cZ%@¤:5M51¨ _ö{€ºI¾×ö½'Y~žç}~σùêXk'ùž¶!ÎáÄ}qùòj<µ|URÖ¬Ô£ŸÍè=Þt 0î)‘Ê‹ïw0½G♌,;¥ûµLk¬ª)Éð‹Û»î5²Ëï½gP×Æ€ù0øÃO€ù: ×Nf±Î¾Øóν±C±c1~ÐýŠ“‚Éíq»<¼Ûí G¡ˆKøD<Q7…ÑDm"Î{<φj ¡P+錛 ñ¸‰L.·'ŽÀƒfëluÀ õWG;ªùDµ[‡}Õ.OÈÄŠRm<³¤®
+L真ñâÿyM¢•Ú$Fúµ¯†h$G†É(17ëê4˜í¤ŠRnÕ|9߬}A ñMàØ)WGf º¥ÌtÑÜ<c›y؈¼”Jà•vÑ<åÐöC©|Èå‰ðÆÃkÙº^8ðÏzÓÿpWèO`Ož¯§ƒ/Ç <1ïa+¨€ã‡I÷£k~,ÿcѪ^âÝ”h©×§ðeGCkþ÷ëªú¥ëÉËêó lz¥ §ô 5«×%q‘ÞƒWkååÕ8“ TlÔ›ó[ü^øìw¹—à ýÝ5õ±d&ÃdŸÆHsK9¦¾·!ö5`Øfµ§›¹fcÆž‘6pŒ›í›¥mÆíÒ qH:À½h<h?( åŽß° “®pWmŸq_Øoqwí%,g¶k\›ýXºc'F;¶†9g±:,«MP„¡] HÐ"Làœ&™ƒŽ¨#íèvŒ9 ¢£ÍA ôvk Ûaæ e°…±›M|‡p£âÇsÆx‚dµ¢6syØ<j>až2sæÓd#rÒ!Þu Æ\”(8ˆùe)zÙ¾è²Ý Çð[6ŠÕ8€ž`Ô$VKÇÊŽ2³÷OáÆ—Oì¬Z)ß»Ätè“/´lô¬„ïNvžºœÿ¸Éò—_ü`·Ÿ]I|¤üÁœ¿wct‡Ð0bq;ôŽÛ%ù‚¸?i /DrâsëÿÚ•PëªØnï_×' &ãµZ˜÷Q5J¡ImÍveH!Ü#Ñã¢(œ L®10 Á Ü-î éÐU9†Ìs•)Ðà[jOE{eå\%¬ «œ¬¼VÉVNfÍci—{äœ<,Ê×å¯e#’5øÂÈgÈHg š2®žPIŠU51¨X4 I[Ú,Ý–a Ûm³ËiÒ…êç[ Ñ3 ¨
+êϲB"¦}ÇNG5 m¨xŒe;ýëa7÷ ³ ¡pr~PÓRâtp…›…²B
G„aÌÑ=šŠ“¬j}¢¹«¬{tëèQœÅþ=[²é*¿>oºöΛ›6¹\eåúMýٶǾ3øÛg–fVw}·iïS¯þI9¼fGH.•Š`Uª9ÙûóUÛuÔ¬¨ø²¶e»ÖfO ¨ÛæþÍ®"8ÝMM‘5Öæ±½î¬g?BÞ k ö‰6“Ùf*Š ™%Q”¬VŸmDBø§Z‘oâ©ñhžvOÇà™À^IŠ4,J¬"ÕHšÔ^È ÃÒ¸4%]—¾–Ì6¸5
+¶u P:òŸÙ¦²½py } tZJ’éB½L¥
+1³D 5 6”õbE€%Äy{q) ì]¨Œ+[7| b®L™x‡ %ŒoF†þhù°¯€È@û"%„3–/ô‹$.‡¼Ž:ÞÚ0Úñ‹éÿ¼³ñÐêT²Â×’^DÖ >4°mźƒ†«ùU?¦ë—?Õ?o^¿¹¾è¹Ø²ØÒLþ—ïß¼zæÉçÿþÑ9ê+Ë禙vгŠNiÀ)â]oŸ`Dw” cF®n©_Ú'HcÒ¬4'E˜æTŠ.\¢¹,#
+®Áíx ŸÇ×1Ò¶~³òæ¸Qnœ;Á禸ëœq
+Wõs9ø;˜ë/i –[T¥´¿ÍPMŽó(–Y^Y/*áa ”^ä3”PÄnH—$Y )ü¡D ÐÌë€bÄÆuúí÷¾Á¶‰—ßnª®ZQj[×´'Ý9¾÷-ì*Öo`rål½kiëŠ%K*ŠÍþº'ö_¾8 šÛ šÛ
+š“Qú\K´bŸ¬•Z¼Kð•ÊY´ #_®ØXš³˜¹¬ÑÄó0bÃdŸ<"Áy½ä ç¸aø« w)¼‡PIJÁAElIT I·8&‘žÖÊ f¦aÆMÓ#…W¡à<ÔÕ< ~Õ*Ù#~ t¤Za Ø| °¢Òÿ—UYÈd™J‡Ø2sx›„²*ìJ‘
+‡³<qú† Øwºæ÷)žá… ^0BÁdÝí1¥N{my„lÄ»R©Fý¼'õF÷øÑ×:G’-Mõ;öýpÛÎã¯2üÅ~WÏh™sÛžÁ–O¾:Üß¹™ØÅò·O^þô ^|wg=×YPßPŸÐÏ´*MÁ
+<o°±ñŽÉåDÎã|}6£!g2yï ÏñxŠÇ6 óQÑêă³V,ZƒVbíR$ó`nMQ£›Iç!ìÑX )Ý>ã’’€ÙAØb‡ä ‘íÞÝØ/:{kÆ 4‡o7°Äb À]¼ª :_ • JÖJ —AÏKŽcNæŸ2d6¾xawã«ú—gõ¹üé—þ /Ë÷⣡Šƒ]ÛŒÿveSï5Ÿ>m Mcï¡ =ÙkØuì =‰Q„˜„áKh6okò6ÿ ÿ%åu…uñ$‡±”sž)æ‚h€&ðZ "ëÁ‚%k6…7w«ýê>uD SgÕ9Õ(Âh˜ [þ€ä Œü_²«>¶‰óŒß{çÏ|ØçøÛΓøÛß]ìàÄ$à˧YM‰ÉBB˜LJC£R‡$ ”vÐòGÒi°j
+ªÚ?h·¡•¦0DÕÓÔIt¶ÝGþ±ŠÄ–¥ÒÐ:MØÝóžÓd,Jôæ}ïì‹ý<¿ç÷±€¼R
+G Ò RäèÚ‹^SŹªs+4"èyš¤é4=KSô诫&Ÿo¥!ÑÀ u€J¬Î0“™ÄPÄe¹~º`Èb
+Z¹~"d ôƒùÅÀÉ¢¨Ùua=æ¹7EK%ú¿ÉT™Ñé,lìdᣞM¡7ß¼}!;½E5y˜px×$uy Ð'!áx|äÍî#å¡ÁÍ ± ÀÚ* ÕçHPÄp1«ØBÝ#Zˆ$±‹øDÚ<9¹ùµñ7M ·©šÝÍžæ-MRS[RJ¶©_¯zÝ8ÛD…§ØémS.—/Ú~°Æœ6“=æ%3i^ +%o\GÄ5+O‘Óãꞃ>}ïh]RGºÊʔ约2,6ÉMLÅb14š:é¥ÅQ%Q—1ñCñKQ)þ´ ;Õ‚<·à)
+™åü1Ð
+<ÅŠò\áAÁº\ˆãW=\–h6ƒbßÔË9ŒEf“*¼ªÂÑ&|Sª…» ÅM<Z[ámêõõÉÃj- xDö÷EÚÃlܤ1Ö¸š¾_ŽNø|.WÿW7ÞÞ{ºøèγ½Ž
+ïß¹Wÿ ßí~ñ“Á‰‰={O² ?nÙ1û“ŸÅ¦¨‰g$ÞÛÐ͆["/Øì¢ßlV =Ú;ѳõÉ–§m¶L—4T'¼Üv!`;=uâÜ™“g åF¶uúÇ»ƒI/Æ8.ö=àTø\.Ú+) ŒFÉÔq_©öÐîÏí®ªÒÐe¹BzƒË0o¸ePn ;a!hd¿jÉ•c‘цÏ%Ïå1¿êÊù°¦›¦…aN „E"„èP:4¢BÞ"†w>“/ ËÕÒ%¢ÅðÆÉ^Æé!ñ°„ð*›]¥U{í*ë¤ÑÚ4Ž#ÐàÈ5
+u2”¢ÚËPŽ£J§‚=JPä…£WVo+4o•MK󀱯Ûp
+Æ
+?ñÎОmÍ=õBº¨j–²»»¦zCbºhhŽ‹‰ÖN ÓÿÔW÷ÎD±ƒó§RÁ º9yú‰ó}ÅW#¡T*,(ö¶¶ Ÿ6 ©B}Ë ‘‘*EB‚‹¿W¹*N©Ðߤ`¥ªÊ2`.3.êf´sZ2¡í³Ÿ¡(]μX±H8æ dÚ1ë ¸d5rÉòP!l"ó2|ñ2XÜâúXÇ(ïヿj w¶ó|Gç 4A[ÝÏþ>\ ”O*ïvòáöŽp¸³ðYù£Ïku¦TŠ ’;šã QÀÇyøü
+ˆ–zûô}ôîêAF¡×Ó ”Ì0ßÓÓ&8)t,;I*L$© ™jV¯Sš2_à +X¡VTù=ã€ÉköšÊ˜2^aVèhªl¾W^.˜Lý‹’uÈtË´dZ1)Ó dœy8|mR™ÈS’C$[ã`íI{f=”–ö 8†šTíšöÏøçü”Ñ•›n˜i˜k W8”†fâÂE p™| z~ofïxB©<외PÀéÿbWáVÌy"¿,ÇVˆ>ºã¿:£ƒÜªƒàªÖÑšV(÷8
+ÉÒ‹ó©JÐzÁý”ÚüxüЂ(A
+õÿnôÏ ì¸ƒ25=E"÷—_1:Ŷ¢vgDdBÉ( ëBZ?c×™ÈînM÷b
+—J™t: Úd™ÑS©[ñRØÌâ E c» G&"&±Pd¨¹ õ"ÊiM‹Då|%™®œAƵ0¯‚è¡ <`È°ñ³nBoìžáÂ?öòa§7Ö§¼ûŸ«ÕzÀF}@1±Ö¤ð~† Tw¨ëD
+m—Z|º˜Ž¼àºÐHj=UÛ»¯ª½_Q¡R§0›¨Úêë ’Äó g·[¬@&-I?Bµ$-ÐÃ÷
+9½}ÞNÚq?Í3Ç,1+€4ÆÅô0CÌ £dHÄDŽæj8‘»Ä]á>äÔcÜIŽ¹47ËQ ~/ëÏéy/ð ~ž¿ÅÊ«§ùžLðCü Oñ¥ÇGæ"K‘•ˆB qEz"C‘™ˆ2¿–$¤K)aújÉu\‚KáG&’cÉÙä—I‘¤a3Ÿü4ùuRE'Óp ’¸¨ÛWµ“ˆM ¼ÈØ–öÁþ@˜íd´É¢|e±bÂ*«ñîqÚÂIo—É× ¥ÿ8ª°¨·_øξu oí;)õš«›„bPhG¾a¸z^*ºA¬öõ>ßitÆÅb àwP—
+†í9<úüöýo…ÞšJè³gÇ;¶n-Žõ.tå•cÛFÎ_åh3>“#шѼñõ Е»Rà¬ý¢tJ*'«¡ÜÎû*µ¬/*½&,”¬,¸7ååë³P’–ëd'!ÀuyÁÊã"|pÒ•Û +Åñ¬,'t¾€Ûk/Ä×ËŽ ^XužU6‡ZëPÛ°ªØmv³¤+ ß8÷jF¡dΣÊj%ÈŠ‚Zsîk]Ú *~ÙRmP•ÔdÔ&Þ}>sb; k/V íæñìy!ÖV¤Û¢m[Dê²q÷ÔÈÈ·ÞxºØ …‰…Ðk§{®}¿ø6>X2Óº¤ \ú
+pé„8# ’ HmC›5• µñ ºÜ2%QcÔ,u…ú ¥”(ä‚=I-]ïÓSD*{ÁNÿ++G½’ÿ)ùpdSúL*Ë0ªÒÐÄ
+ÁNW;³Â8L´°XÉÊa¤/¯ –=¸¼p§ˆ,ÊdQ„ÀÖH)UÖÿq>¨p¡ÃøîµBñúØìLvòµêãâÊÒî@¦?ÿ™’¿ýÓÒíÛî}ßlKqZÑ ŒRÈùJò;%¥Û "ÉiH:è SªkêNªàXe9æ IúØð”ö:yÅŠì?7\òålxnk ^‚a_`)‘•Øôù®ÖØ(®+|ïÌììÌ>fvwfgf ìû;³ë5žuüÊ'žàÒà°ÍckRA!Æk‹’ØjÀ!Å&À‘ELT©)Q[¨iŠ Xj¢” -N_TQ+Pë´j+
+-Šb»gÆ<”FªVš=wæ®vî9çûÎ÷Eº#C‘™ÈÈ|„ñDF#7#$à|äþÐÔ"æX°ü¨í¹ûHµ$Hƒ©
+ EJuÚeO9ÕlJ§WîÄ¥.¸”3‰¸Ì‘¹oú Ú>-ÍñŸ Òœ:€xÏ—lŸ'™"©™°%)õ¡béþ ¯@YÂ@×¾êjÛ½üZ æ±Ýçz†{Ç;_«õ‡rÍ‹îæ\nõNâéžS¯÷íÉ· &/sǾí/ÎŽ|<:°uËâ‘2Ÿ²f%?jΓë~óÁµéÏÏ|dùž*ð={ Yô#·K9 ¼\T>Pn(Sþ0¶€?ð~àwªFj‘ˆ¤a‰¨3k:êÄNÓ •{gmö’hGÌûí.®—äF¸ î6·ÄÙy.ÊéܸqŽ»ÊÝä˜(‡90G†£›)2W™ë Řú߯F÷…"t0{¼äx›Š‘zN%Ô{¦¨oÿìì\^7]QaÎê`Ý×…K⇚éUü’Rr@
+t€ïØüTð΃4ƒ}„¶í»¯
+¿d‰L&µÓ+ðÿ"¸
+?]3V|rÛcå+k=¼©Z«Õçê,\.蹺š@™‹G®Lžjh©¯Œ¯Úø\cÿÏ{®•$ׯO‰ÍM
+fvû »_'¯ £õïh‰ÚÄzךīîauÌu2õGË8ñ.| ÿ§ŒLÆãÃNVt:Ù2½ÜO’Yv»ž}•}ƒ}‹ý)û {ƒý{—uҬijf)–
+‘¨MïÖ‹úy}ZŸÑçu;Ò=º¡/ß ²n² XŽB@韗¯t”;†Å“"!*ƒòM™Ø wɽò <‹%Ù …å„”<…?1$)´×ç§}v‰&÷–Â{Å“ Q¦°Ï&äsòUù:ì½-Ó¼Œå_ƒ¾õî¢6ÍhdTkÐí
+!!’à›ÔG"æ㈇Ã1ÎàÚ¸yŽBœ‡‹q9kÙÍ
+q w‰xÅ„f_a?2S¯ÍÁ8µ¥IZ
+
+06M
+«Ï ¶"·}KÑd³è›‚BGð=ó¥•fa²ÖV
+‰rœX¶fð1!g:={‚¼ðø[/öôgŸœhlÞræàîÕ™`cÙÉ}]Ï?R1òƒG6nýÅÉãZ«E¼ãÇGºZK¢‘ÈêšvmÚûµ–ÅÉ×^*4®ÒÒrXˆ¼´õжoÂý¦–^ú”¸mÓP õáSNŒ¢HIÚ…Ü wÌÝæ¦ÜŸ³EAB'AI_!žE4raß…Q7vO;&éƒ~sFz‘à áŒ0-Ì´ „‚Ö±gç žY…B=Ðܶ$E¥7î{-F®4IZö&ªkjkp¾ ë‹ ®Ë7u3ô›/¿ðÊ÷{ëÚmÚâw¶/¼ýÔ3/Ÿ¸òîà.¢¿EX× 'xtéSê';ÿÁÈ %NB
+J 2~ʹÙ5èv#qÐ×!8€¤Mý]qˆ"ç÷Û}ðÒ?ãx‹ “öƒ€ø¦I|0nž%e„ÛÂ3á[a
+AÐ .†‡Âg`9 f<áÑ0ÁzÂ8 ¾„ÿ„8zny®Ãá,–^¦èYÛ€ÿežöBIp;
+†ˆ ÕŽ,·xŸx%ÙFÓu‡²Éö;Xò?œéx9GË$›Z á6ÍL˜¦Z“]«Æ‡¿µxºùYÜZ_³jñ}qÓ; ƒí ÙH¼¼ðÝý‡Þ=†»rÕd‹÷‹¥ê|#siçs›7]çj ä!ƒÁ¥Yª†< =ðOC\+n„aû1ÿ˜ÿ¢ÿ—vÆ`ðÔÒQ£ñ‹1»ä÷3v‘aì¬ Šý !+Ù’a`Ò:Ýãžqž'ÆãÚѬÀñ’&¡…ÓÌ£Ì7˜cÌc»É`Øn Õõ¾‹1>ÇOó$?E
+Wñ(>§±
+O½†€Ð º…æåAmh‘h$èù Ôø¤¹BÀ‚V>ïµÐ•GJh©¹9h7°ð ¼’ížY²Y2ƒ,4 h!îçàµä嗪͕ '0 1ˆµ½Òq÷¯¶Q|xU‹°øÚÅQ1ÿy¶‰X’¤…„Âq¾¦&ºªì‹ÉŠ²69´zi–¶CËð)ã¹g@t¥¶ªß“m¬sD±NG§Ö‘ÞíØ
+8ŠåCå7$‡¹©êT éBæ
+çëÒ›©ÓéÓ§ªjŽ”ƒM91Ct`^òtðÑD¬#Š2e ˆsº;¸ˆ²¢#bL- º §’1ø6ö°i-¦jjJëg "Ë:2éô0Ë@Ä8S©aYeYJ$“ý²‘’PdÉé`™ŒêHIð<˜8ŒèT2™NkN&“Ö(”‚ ªM&R–£jŽ}‚%XE§pöÂ9«S„d„cCpr”C‚2Ù&ÐMD°QÔ‹ˆ<º„ G+ïã#€rAA
+^«*À–P2(ײú}`<|Ë3)Ò*ØW‚l6¨ Ï öLÿÿëQs3c’íþB Ú%_¦Õ”9Qýpe
+–‡- at Z]PÇkÒ%Žß^TëZñ{j•Æþþ¬·j-îijQçëï~R·xÞ“yÜ ðK…á»mZssHñ)ßÐã0››
+O,8‰ ½d½" W9ð +¡K*ñ>ãDF_Cw
+±ÝÂîØúÍ016G´ûÚ…-±Îø¶Ü–
+:—«°3Œn÷ ‹Çb>Ÿ/ñzKJúuFÔu¨v8V!½$&PÙœÏîehº"Çñx–w•Ð4n·YÞlÚÔ§²¢d2¥Ž¸ÝGŒ~»àoôQé4Qžoà7ðäU 󼬉!b”˜!(Däƒ ”þÊÊLé 2¢L(׊W”
+ÿe¿Zc›ªâøïœSnÛÛn½íº® Û:Ø£íF÷r
+”{°Ž996NŽÌmÂtˆO at tfÓì ¯á‰1(1HŒ ú‰ Dcˆ5†CD·Îÿ½· 1<Œ übOïïžÿyÜÓsþÿó¹…ûþ,2PLkXú3ÆÈïŠ ùÜSYvøúÎú¥E>õZTªþs’ …^êHT±Í‘ªÐü©~+¦C¡Ð_Huž>áŽ7¥:qCR'L¡Ý©ÊQJH{éâ Põ„êê.Ù#ï«…Rjß|fÔ=.õZÌXAaMô9Þ8` f¯Œ¶dyÙ‘úxž·Ê£¤Ïuå´Æ7µÕW½ Ÿ,]K×À¯x.w9ý–ÆÆŒžïâ{æåD£“Õ’yû\±Ç“â`ѨÑ]\= ªÅðÇ4»Ä·áp—oÈÌ1 æãé K‚T7à”øåt—Ëp•?é
+ít¹¶øe§ß/—˜d'X~?<¤–Ÿ„KŠz<6‡-¨;E'¤Þ1/÷¾˜•…2§lŠÝTÂ`’¸7NŠ¬œbN†í;ìíÂ~Šœ`WÔMæ}(Óò°Ø¤j]k5ê#£Ö +ÍÆ’T¦õD6agU $¨Bi©‘˜¿Û”Ð?R¿Ø]õciwN22ɨǸhñ®*ƒŠòE¬ZUÁRö‘o_b³Î¥×u±]…ñëWwoŒ®©ñT 9y+ZºãûÓj«É4çd§¤L_`˃óŠ¥HÄÒ=uåÓeíËC%uåù–4›Ñ |æÿ½¨ÐHG\K¹ÄñÝ/ÂÍF
+mùn~„‹ Q-šE—x^Œ‹ýÒ^ÛqÔz™..[SŒÂlåVòÞ»,VŠ}0uPj°“ƒßä ŽQw¨¡è|Æ)½L²C¯L ·r“ÍÆ^Ó(ƒ½×1zˆÜÏ%Ïv• …1ro‘°W‘så –;äC²¹Gî—wÊcòUùgÙØA—OóÅV7Õ¤‚ÌØÉZ
+)jõxs:T«B™ŒÑ£yº—TnÄ v³<5´yU+mË çßO¿ËrÖ
+å6¸¦®ñ¶ÛíËjÜâhhê«íoÇ_kÙþuø[Úÿ!ŽÑ½™K8«ƒ—ήŸ[t"„K€äL 0Ïä«€•¾³~
+¤îl Jäþ°o 瀴) ݸŽ7 Ï ¹˜]ó'²š€ìÀœ›@Î5 ÷Àû%0÷e `3P8½®Ã7út” »hùô]ð<PFû® ýTv U´§jÊMköµ+uÔý
+, =-¤uý,ž –´è¨? Dt›Þši¼UÚ²€e´Fû{I$‘DI$‘Dÿ?€SŒ®'„J±L‚„‡¡W2…ç©6ÅŽ4gº+ÃíÉœ•ÌWPXTìó(– –•WTR¢WƒºX¨
+7D›š[–¶FÛÐþhÇcËW óñ®U«»×<{øŸÿÅ€èí†BG5!Å(E –¢«ÐMèǶ™š‘‹Bø)«\‚6tÑHz1833óͽ~ ^ß»ˆ‡îÉ„
+‰Rm :5AKDe©’4˜©' ¾Í©¿)A ê_™
+D÷'h‰è‘%‘hKS³¿a}ß–ƒÃjщ#ˆ¢…Öm&4`=ú°[1ˆa,§ÖF¢7ã)j?hæ¿ ÓNÏà:BÀ,:¡‚|, vtó[ú]6€ÑˆÉ@”ÚºSc§Œ†Ý-gób*“d·™hY\0ïÏ%¸Í**>»øªïI[èÌ6i³ßªy'¨Ö§ouÇO˜o™)?‚Êmå? nõ€ý
+endstream
+endobj
+29 0 obj
+<< /Filter /FlateDecode /Length 22522 /Length1 168840 >>
+stream
+H‰¼WIo\Ç®7ûpö• ŽÈ™)’g¸s,R%.¢$kId!R Ä"´±bÆ,%']r džo‰}ð!‡Ä0 £àKŒ\òÄHnFùJÕ¼¯ôJ/ ‰J” ðuõ«î®®®®¥‡ "JÑC
+Ò…ó§ç6ï¯ýœ 2“Î]¿»µ}uìóÏ)ðË:Qàáõ÷ïÕéïôW {ÂøìÖöí»8õç+DÎC ÿáíw~zëËÇÿü19ëD}WïÜܺñÕýàË»ÀóÛw˜áüåÂ×DÁð÷è»÷ |=¶ñ{¢P–œc¿}çÝë[ô· ‘sþ ÿîîÖƒí·®+øÕ7<¿þ£»7ï/jøà1ïYÛ~÷'÷XoÞúƒOe|û½›Ûzoá[rV ñø5’³ý'¿¹-R‰Û2õsÛO|"ÊRŽÛ å¹M²ÕˆÒ”á6J1nãÔ÷¬5ž>íÊLð
+ômÛ¡f£¾Þ©_»UïÐ[—cÕŽ³uåæZ«ã4 Q`}dãÚf«à~pý`«äN Ì÷ã s' f„ûaF¹“ 3Æý¤0ãÜé³ûÂLp§f’û%a¦¸3fšûUaf¸“3ËýŒ0sÜÉ™ç~^˜…fÇ)ç[bÓÉ>nuJMRn:õÜH=¿1²qéòRµq¥Õéo’Ÿ5ЬoˆI:õ-–Yq¿n»_ƒÍzg`²Sç=ªÜL>zú
+÷÷4ë´ä,µ:CÍF«3ÜüÂÙ³¾öE`ˆ›à07¡7á:7‘7ѽÜÄF¸‰rÓ·›Ä7ÉqnRܤ÷s“9ÀMv’›\“›|‹›Â7ÅinJ3Ü”g¹éŸãf`ž›Ê7ƒ‹ÜTÛë|‹5Vå
+Qå ¨²$ª,‹*‡D•Ã¢ÊQ娨rLTYUŽ‹*'D•UQeMTYU6D•“¢Ê¦¨rJT9-ªœUÞUΊ*çD•ó¢ÊQå;¬
+uãáÉÓ'n‡›§þ9ÿ-3nt»»˜cHvˆ€F2ãúaŒ—AKàÅF’‘À·Ð4#‹¾ÈœâX 3{†°¾4ƒ1ý^€¼¤áë^iP»> Ý aŒÍƒ— rö†‘›À¼”9È_Ĺå<mè›1såLUÃKƒ@㘷ŒõY³ÿ¬±·ê\äû9Çø.øƒfŽB.5ÅØËýÆî_ä2ÝÀ~Æ$Ö¨ qèê Æ3>ˆ¼)F1dæÅü` Ž˜{±›»ì…0y~cQÅyt½Ç>?xÿœ9탵úŸªõ¢IȶvJí‚îFnrÕw
+¸Aß‹†öÁ6iŒ'17‹}§@Û;PÙ§†ý3;Ш¹ÿ˜¹oKÕw“ÐÃÒ¼±Q/jc×R¥ä.i€¼˜“÷© ¨ÌwØOÆ'$ÖßcŒ3Ž2«/ ‹ GŒßÉˉQcŸ y¹0D^ «¿ /gd¤ _ó‡æIuF ÷(v– 5ƒ³iÞ“} s÷ªc=öÄÉËSè¥~$/Ç1W}Ec"Œ}tžž#I^þÑý5æ3fžæ<]225G
+8»Æ‡ÄÃ8¹uKd
+ëÿSàÍ£?92_ë`ÔôeLòð îùMF“qßËŒiÆûŒôÅ—–Ä'ð=…¹«à‹ïL€ÊØ ôÐÚšƒ’g£<~rV ó£
+=θŒ9mÈ;̘Ç<¡â1ŒŸ…,Ù{?cR¨îE^N“}OâžuLì¸ ]ÓäåU‚ÎÝ{†|=ïiØChcË€è¾ãËflö[½al¥kJ¨yCŒ˜/çìÇO1ÖP;/"Ž7¤€žÎæGÍ [oQ3c Ïb {œÄ·Ì߀Ma— ük±…Ê A«;A›!̉‘W¯4v5Æ,ßÖu]¯¾!/¯ê~JµþD ŸÌ+!®´Æh,fSaèêÐóo¶4hçÔœœ…¼ ¨Öw•ck òíNçhîÌÂÆicë4Î 0tÌØ'GÞ["odøë¿æq=“¤‘§g×ûHC®Ò^õRïÓOí™l=óCïÝ}Øú*à5o3 ƽW}òþ(ñF®ß–¹ÿ1¾å]UB<‰¿K›&÷Ý)±+¾ßzɾ] DÜöãL
+sæ,ì”çñè¨q£¹_ëá0ì!¹¬ úÕ@ÇÀ߇¾oàÕ1w?¾G_q}„¼¸Ó¢þa}Xã;fl5wjc<E^õûžÊÓÚ+ù=aÆäÅC^L "<×ê§ï!û¿¢[oyÞ8yõ^ø%s?qPñUúÎÕw×NTm¥ýnç½~€\Ü`QÄŸb'F9?Lj#çΣöÕ÷ ÷íÖF.–ú'µzòæ ·~JÍ«:nÍÞ@
+؃=‚ÊÙ/aÿK°A‘¼w‰Þµú`„¼üTÁwÅŒ+?
+Z4÷4k… ?:`üIß“QÈõS‰ÁaÛ5ß1Ÿ®V¦ÆÑs5± bš'Ï_ë|=Kúêù•§{ûõ[K¬ÉËÛä|µkþ –ùr'MôeÎ(|QîHbSãNkZö“8ž•;ÕqøD÷[3ö8,>̼+Øo ÐF—w@º /6ÊQÄ;Cò†äÄ”ï~ºÿå`WåÏáÎ_?`G•ÝiNucÆÆÏÖ#Gônc5òü[ßü/¢ö®5Îú}ÔÈ eÌšýå5iγӹRäÓ›<_ürÓfo±Ç" ü¬¡|Íì9ô|N°z•}¾ ûØ8Ò7šØ·
+žÿ\þ¾ß°oŽÚ\ç
+þÍþ¾µv½gÏ×ÞÁ~òþ-‘›7"¯Ñ× ù¯4…:!ÿ%Z€Ô•iô³»Dì _/ìÆV¡×€×}/?cl#Ö+Œƒ&n»ß2|éo;ôÜÛ’v3îŸãŸï_ã[ÿZ¿wê÷‚oŽÚ‡¯“—»ì¼WåïG|}}oFn
+‚žV_?ýo r5Ï…À{à¸õ[ÿW´ÖûsiÐwD К0ÀÚ«~©ïé@òvÒÌË ê·IÉòÿé¹÷ã¾ ºï,Ç}³êÛo'¤qfÿÛ jlÞk`ô%ãšgäý-y]Î&o£qPy;m:nŽ^ÅqßççÇ1.Ø‹³ìÅx
+òä
+ý3°›ü¬ƒÊ;IÞ[뎛•°—¼áä}(ïy+”þ }ÿùdÿS+:7ñ=
+™ Žû¿@þ+Ìb¿¾Ø3þh?ú%¬@çAøá&Ö˼EðeŸ“ðYý ¨>üŒ'ò˜–0·;‡ûW™Ê>S°uöMÀÞ#8‹Ô²³wöÖ;Ú}Ç çôž o ²Îª¾7þ(öSŒâ -†
+-Cv ²uÝŒo¿1œMÐÂ9X·sZX7:Lš¾è3îƒÈ¾Ï¸Å¸è¸ÿ /AÎο [°·úÆÿ³.ZpWuèX¯Œû”þ Æ‚œsh‡»Y ä|ÜúE ¶®‚W…Œ
+dæA <Á ÇõÁ!ÜøÖ‡Žëw¿€œOñý±ãúìoð-{/3~ï ÿìõ‰9§È’8ÖÜ$o–ãÀ
+p º„pFëÇÀ ª9®ÿÍê‹G!g ëf€&°
+è[0
+¨=ú0‰û0sÔÏ5¯• w vT ¤
+Rvþ0xš'pŽÆ¥µq®6Î)±Ø‡¹5³^ö…mŒîk°õôj‚? ¹ÓŽçwÇó»’ãåÊ=Ž—«' Ï3ŽÇE@ýYí¯v`½ìÛ TÇ-ÆÛŽëCU =¬í–
+&
+J=PÁÜü¿x/—;Ž+ sðÌÜÛÝ÷ýšI#YŽ-bAÀ;CdáeÈ6›l²Î6ÿ ’B磾>j'8^]]u꼟©·fÓïòÿCòùSôøjzj>Ò|¯å˜ SÖ]Ú»ÅÜÇäý‡¼{¯ûfÏ&ŽÖ«¶üô[Í ¿æ|Boñ¥õßæÆûÿáÍÏfÛMò²ŽÞ r¿>óeÚ¿õjrî¥{òØcÊÙ|ø˜pÄ yß¹} ó¿á¤«M¾Û%î}žm„ ·ÿæÿHØ7‰kL\ë|ßÎZL¼O[L sËï)ùÁf§|‡ k='?¼ ’öû<§<$ äj÷ßæ{Ûry†~ÿ—øÓ/Yw k+Ý`t{ÔÙ]ÊÕèÞËÖÙEþ½OÝb'×܇¤y/Ý[ã+ï£×Êmê‹xÅ' fÈ…½îÅëA´Î¢Óο¿[íWâmH:ƒlóœßßçêxó‡–§›ý8®Ñ}N¹ ’Æcü}àçuó)zsÎók~9ÃöÇ„g®Ýæ™íH|±g5<ÿJÙZýluöuÂ^òÿ}êðGÙû’ß?Eï…8cvy¿Ø'¿´³æSÌ=| ¢ÇâF6F¶:GmRßO=±ÏÝY¸¯ÑýÏ}Ê>æþH<²ç½mr<²²v1Ï›ð8F÷£cž!˦ì‘ϱ ÎIzàŸEì=EwèVØ©ÐZ-Üm„üÆø‡…{¾Äï*zmYâÇ| ¸· ðuz3,Àù ñκ÷›»ò¿-t¬§Ê˺Ð^:ßÜ«b¿ÞÝ댼4–3÷p«¢Kú¢!¿Öéÿu_}eˆ¹¾ˆ
+ =ÃY—Sùb'z‹•ÎŸµÞ¬Ê¿Ï‰ë%û®DçNïLÃz¦ÖÙFká ùˆ'׳/éܼT½¹V;TŸ²—ì9 ~UþM»ÚØv©±eü_•;Ë .rÁRìT}WYè‘ÛÙý‚^Ö…–mrÎu¹³ÍüEŸW˜/Ÿ-ñÆuÚîïÊÿ’ŒÌYöË5êûP` ë;ïYÌu¾¢¯ _mÿ ÖÒ³ôè,úËSâ<æ{ò"5˜¾ oæŸÏ Ô{hmµˆÇ—øÔû°½.¾®Wùÿ.>õ.ï¢×íç¤ošž•nÑç4ß·÷
+ëĵž^d_ÖƒöÐÞéß½C}»/pøÖ˜çôFá¡ït¾ ¼6 >¥ÎžSooŠ®Ìߢ÷X®]èé"yè!ÐüÓSŽ:㟾ýÂm8ûÃYûct¿lÿô‡¢[öø þ)æµbŒÞC"¯{Ý•p]¤oû½{r13Î^tà‹zI~¡÷Z:œ’ç!û1
+ÏÒ×Fx\ߟ£Ï!ØÙáa“|ìÅ/z£OAwWÙ…ù“<}íAúÚE÷ëuÌíáüxµà™a˜3ÈIžlóC¡\ÎSÁaë+/iƒ§¤wÍÿ¦ÇW:ÉuÍó«þ¯Âó’onÉó«<{“|]óû”çç´ÅE¶w\"Û¿‹¹o8Ùoÿ‘÷ óû6ú ùSô ‚/¼Ä|6²Þ¨ßÎCУƸ†»&ì³'>Vå ¿&>׺§^]¤/ôóTød&%Ç¢×IïíØdŒ^ë¨ÅÑã£Á~ =¶ sSæsž¹ µï«èqø’0SÒöëäËõú»<{[dy–]]G óØ«`8cOÜ ‹KôØ{L sjê½å 9z=žÎÑýø,?æwSþ#Öý;^£ûþE²]t~Š¹¯¢EÞæXh°' ÊýQxLã,Zôc»óªÀ¶û×1qì²/o¡{Kø[Ì{?ä?¶~¡ ?jîeaU½Ö¤÷²Ù–UOöbïXèY_¶|£ú5Ó6.òívá?Û”·Û˜ûÕ—t‚ÚWˆõ*/u{\°ãÄò/ù™}?±-‰ëù*xü½ÛÈlßs¬X æäÝÄ<§ë¸—ìm ÏeUÙÑ×V´–ôÃ9¹›µ]À
+_ð¯'«ÖÿAt_b.;ýoµ÷^pžÅ襘ÑÜ{»ß±ÏÂE¢û$}áëè}
+ûzÞ¾ot×ÖSù#Ø›öÏϼ½jo=ëþ¶ Knñ?¾jÛàgô†CÌgQÃ3_á»ô §òÛÙÏ7ºßè
+u•;÷¡Çè=2ùγ&0Äæ=gðvˆÞû;†ÁznÛê
+} =ÓA0Ìž¯¸[If|ʳ³2
+…ŽçV÷‚ƒÞç{I÷³mï>p'|ž!‰ ðÃCÌeÏQÿ£à†¢?Þš7ë{/ æù Z—˜ëjÒzpOÍc,ëȲ sr0¸ ï;ûÒ½æMÂÉþª7£îØ“<7 ŸQßQ÷ðÏÛ˜Ï {ᆽécgxhÚ»v?ôQäYj!z™¢çkæ·kÌk~tŠ>OA£½WtKÞ!g“ÓS®Ñk¦sŠ÷¬CÊw’¬ØÓ3 < §g6j1Üþ dÏ©ì‰kÛ›mu‡ð1çÞ8ïR+n¿µ¡»£ö•'jtÕ[Õ¡õUõ\ù°žì·^ÈTë…u3Ų^¶ÂceOþ!·ú¯yß2ÂÓQpÛg; ãsù‡˜×Iò™ãØ~I.äËù¶àDç•ïAtÐSÕ•k½ë¹s)¾aÿ$<êŒØÙÇ<?Õú^FáqýsÎæÜ9
ú«rï:l›8¿‘]Scï-·}ŒÞwzÞ¡¯pcùñSÇÕ$ ëèýÁ5zŸ³Õþšz}Š>»>F÷÷bà=ç?°;Ù†··è³=׳ðâ'‡|ó¨ÿ½ðºï?þVôÎœQûôE¾ÿYøØßbîSÔMrWµù^°Ø×ù`+ltO¼Â õ
+}ŸÒ.5Ÿño uŠ<زáy s_?Jƒ¾Ô™c~á{÷ëÔ¾æþÓ²££š?Wœ“6‚«1líÊæû˜×ôç¼{Ž>ÔÙöœï<ëšëôU±ùþâs]TœðymëÉ}K{ó.qzŽÝæ¬3âªÁ¾Žž/ÜÓÓCx6² ´ýSÌãÅütF ¡–~ óxs}}8ó£gŽŠëÝGñPÞÔøs Ô¾ä=®<Ïí¿ðï=tÝ÷W[X'ž+ìçîÉÀµIý ¢ç8pã7ì¿ËwÀ´Øþú?ëC®uÂ=ŠÞ.á¬Z»klA n¸¾‰îïoÓ†ðl>Ýà ëIøà½Ö6ô¾QòìtÏ “p^ôùÜ×Ó/1Ï®CÎýÕ§àv
+Þêܳ_ía=O¹p Ôü¸)ø<;,õÉ«ò¯‘×í§îQè‘Vº_ż†µýC.l{Œ¹ ø6BçΓ®“Ñë1ŠÏÿ.>ùø›|ßòÆà>&çLòÝQÿÔqbÀý¹Ð}õÕoíØ òý59á›ø<·WØ ½ù6zܬŽ8vt AïcÌýüP¾“öŽ‹% v?lyí£ž
+ð r“çºÚßWúÄ9°Ði´›¢Ï¬ôb‘þŽy¿Šîs¶ÇÝŸÑó‹ãã=f “h¯tçûIïF¤SŸÁ/½÷óx¥¶¸'cnEFú6ôg=Ы9OŠ|žC°ü9‡y¦@^÷€îg&Ña®Å¦î«Fѵ˜Ö”rš·¶î¢û8™É¦xÏm®»_²)¾ãü
+^ûù\7íÇònÔ~-ÝS»˜ÃVIßùw§}ýŽÑó%ü1>Hf÷Ø °#>ÆÞ3¥qX.ì€N¬SÿëôJøè.æþ7è Þ DÛtêr_ë|s;¡_⃹ˆú®Z¡½OžjMµáExpÿÅÝ%æsÔ:áÃcù÷²|»S{0û¢u·)gèÆg¼?K ƒàª-ñ]–óÐAïWe¹‡aö<Ü®!{½% <ŸR¿È—îC]‡©
+ȺÕóü]Ìóòcobœ~§ö¸µ×vŸ Ïuá?î£\g—úBÏeîùk=ÞèŽu-/äB–“<d<Ð gŷ޻㓎å³pXž)æ±SûeËàYÑqo=¹÷c®öžuÙŸ ñ í“Þœ…9èÍký 'uŽâ‹ ¸—[ÒçZïÚ“[Ð o8Ûêߺ©³.gàÙ—kŽý£ÖW÷Îo¦Í[ò¼yäß½·íÁ9øè5Ü÷âkÀ;·¹o2Ÿ•æºàs.EOö?Ï_üÛ^ìÑßXîmߪqáÜðþŸâsžœÓƚĈóóº¬af¥í™]ú _>ýñzfÿf¼\– 7Ž(3d“ øÉîiÙ–dY +B;y§ ðÎÿÿ9n„p—ű(Ô#3+Ÿ7GìøRkl cçKÛÇX© ÚÖÑ^4ÀìÆ–3±¨ëôIô퟽h³¿×}xR¯ñ3ǘsª{ÇBÞñy☊»ÎÕ`´/µä4bé»Zr<‘[#;1f9ݯ b=mëÞ/uÌpݯ‚#À ä^÷Óÿæc¼….†y-r9½ µâ0ŸÍäšxoôêÚÇ[;ÍqÇ~eúĸezÛoóÿ×ñOéË5ú4ëíë,ßu oó:ukwésš¿Îÿ?|Œ÷úÃ7Àf÷y ó¹{-~„-|î^k|=ñÿ¢õ«æÈDn¿×Rk¨Mw½×ò¾ÍçÌ÷ ^û^wïµ– Á;˜ƒk¯¢“ïÄ7Fñ ¤?t=dà ~8ˆ?ï„g¯{Œ[Ђôzî:Ë>q5ÖZŸ£ÞߣhŽµøÓI<L<F>@æŸkñäFæÓ¬°ÏµÆªä›)^Þk©×ÓÙm-õØu˜å¿iÏõïVK#¯œkñÇsó¤ûH0ø³ Ïý"43rÆyÝ5‡8pØ×:“·ÑçIô]§ÜWdžß7ÖüG]5Æ5ïsÐÅÿ4~±žé]S§ÖÝ Ž‰åKÜ3Ö4Fð¾q«k"û—ÿ£Ÿ½ÖÐo1†3Î4½goÎ5ãY×WlbŒ¶•Œ¶“}âXºÍ7&γÿ{lB†ô
5듸´ïïëQ–¾±ßC¬;&OºŸ6³/¢9Ô‚cÇY-Û3©µ_}û»Ða+9þì››ù^ö±œµñc¬ÄŒÆ·öý–,ƒßË›óü û22æZþÎû^tï$ Æã¾÷9þ÷µö¯×'oå=ž¥ËŠlƒNƵcÁ¾vžï’~ï!ŸíÌÚ'ý;v9{¯ÌøǦ–x¶-÷µŽküÕ9 t_KïãxdÏyƱ…]yÿK=êÐwÒ_˜û\<NúÇ'3ǯÐzÄßO’±OÓe ãÿ³è÷Ö'9Ýç[ /ø›o؉O×àñ´¯±Özn½ïT²a£]ðá½ô¶ }Â:ç ÙÒ‡ssåKÌÁÙÈ
+nÍüB_ƒ
+è¿œCÉσîÓÔSè8æ-§c‡˜»ÎoO€·ÑÏQ_çê·Z×pd#ÈìºÈšc˜ âÚØŠGöÍ÷µ`ÿV½¡¯ul׌ãŽqï\K0
+°âÛù¦»Æ ß¹öØ6]-}¤ã×ZüÛ€Oµô1Ñûoñ/{Øýµ–>ÊövßešÆ´þG°ÿ¡q;Þ;kmÐÚ1î¶dË·»ç³.õ¨›ëLÏ{Çøbósuq
+^ø›iõ(›eä tkØ Ù}þëÌÁ‹ƒøpçÜàß×£\‡ }¬õ;m3bÊ6%Üo /½ãE4¹;ÿ^ë}ð>Š²:ï·žµvmx]ãÞ¨}tòáAƒ³Ö“ó°÷Ø';öýNrgµŽ)Ç ù ý
+ÒyµøìU÷y“å½JŸÈOøX'ðékmG㟮ÖyÒ6äíÈ ¬)¾{¨µ,È;ÖÚ>Ö‹}ýZK½kñQ° :…Ÿuæ8to“ºH?±ÏÕö¥^{®7®ÿТV¡ômæwé—ó}ýÒvKûåþ± ß™oÉ7µ|sÔ:: ‚Ÿó
+±i,’r%Zò´Þh 8 ’»û¸Pëm °Ü¾ cþé³®ÖþsˆõÖ[^¤{ë˸ê™^°Cæîñvæ-»yÞÅ7ãÒ}á4¾…{½gý¼ˆÆ>øž‚¿}³•WœL{N5¼‰ÉÔϱ–ÚÎ ù ¸„ºæ8cÍùŽœîû÷à>í:Š c¬ó½‰g o]%ÿ-h ‚÷Ð ÞýûYÊët.9ˆgÜ\EÏ؇|ïzŠ c-=é›dmaï¾ñ“ù Æ2ÈI/ÒiϺp<Rs©Od³¿»¿p€5a¿rŽpþÎ vÖ=bÇ1p”
+àOÅÿÖ^æ@ÞÅÛ%<èuñùÉ×ûoµäæàWrÙIô† ŸyÄ{ôNàZdtžÇ÷±É^:ÿt… :9Gf^tnrîôžýèg2ÿÙŽîIù¶ðªóò-ö:9ĺ嵾}î,ÐÝÕ’OÈÿ¦cyxcˆ;Çú¶þvqϵÞï~‘]yÛv^w¸Iä¦[Cùãݬ϶ ñdl†$O^êѯ’vöbÎቛZgÎ?Çør&ó k~—í“üK¸Ïmí¥þŽ1ZXèÙÈXLü0¯¼O $ $Öq¾Îrn¹H?Ûy°¿Ç7[ïÜiå^ßÛ6†í³¿.xoÅ u M¼F 47b à þ,‡/µä çbb;Ï©!à‰»t·•LŒŸjñŸæóÓ÷ ã_ ãçùÿUr‚Ϧ1倿Í_zѯµÔ1p#²õµ`Îéì{-Xà*šÆ‹Î)?ÎûìMÑG·× møŒóü¦ñ*šà ð!q‹=ñ_ûŸçÏpAæs¯;'%nÿVÞsO´³¦akå‹ìÏ’ÿI<Zgû gˆóú³9 ùÝ¿¤Ì~{ w’s2±á3-Ü’÷2æ‰oäp®¢&{9œ°õ/õ‡¯ý:Ûàu>3žz¨%6&ÿ~«¥n{ÜÀ]p%±@¼c¬ûÈIüÒcZõ)c ãºO,·|ý¨5Û{¿I~êyå{¯µô y¯µ>œ/œë¾Ô’{κÕùäcY˜ßu‡w^Dcü Fæ6ÖRfó<ü ÚY¬«ÖøÒ8COúLÖ=ôJ?bß=Ç?ú<jŸ^ošƒ ‡½ Ë`c×g|üRë^Õ½Ütïç‰}ò>`nrÀf¦á»›ùÜN|^jÁ0ÐØÕç˜þv¦Ÿç¡Ã;ö¢VÞHÓÙ…lí“^4¶3=ä³¼±/ódt
+ðü¬·ùÝžoê'ò†^{|Á»ÆñútþÓ,Ç]zà|Úƒ»Ÿ
t–~ÙÝüû÷Z×:×N#ûÕN´§¯ó:xĽk)=vö}ÓÜÇ¿e!F]mÓϲ]Ò‡6y CðèôŸ›èOØõûúç¾Õ’gOâ…|Üc›Ÿñú^ý‚ùgŸˆN ·Žûl§|bÍë¬ù
+y—¼³‹/{ b-ßê¾'óIÞã»}²ß’ãoÞã=pÅ}¶éô}ŸçôFÓþ8Ï©ÓüU÷.óÿ«ÞFÍ¢¦ŒóúµÖxçK1uf7Ó»Öâ?Ès ° ?;kœô=jݘŠºNï‚nNâE/Ãp¯À9baˆÿCe2¿Qre a¬ä÷ ò"ÓI4ü¾sÐDÆŽcõtÚø+ØáûÆ¥'Ñkg/:Ÿ}¹Z~×QgÐ/¶ f=]ƒßE´™á|Ïðsú€õj Ï ŒÆ}äéDœf¼ÏýN¶…6ÿÖ}ö‡öSì£Ï^|Ì›9¹‰±“îgÞîkmg÷ƈƑæé<iŸp±M,×1î8†ûZç»^¼Ï:‹_ëyî>Ç |‹§±‰w°v¿¾ëø!O“×7ñ6ö‡yo#ú{ý3ïtn3ßÝè>6eÿP Ž¥N‚³‘o/ºèf]¡· >ØG÷õèüÑñ˜ùaÐ}æ‡Øs]…ß]gÀføæ$ß{ýÍþyh¬ç™´µ¿¬»þ'ÞÛìÛ©¾§Z¿³¯µžÌ¯^)Cî·ôžzMZ- ìÏ幬™–k¯ÿ¾Á—ówÍ—ñ1â)sĵ¾ýnãyëÁ½*¸Ï=¸ð^>â‘÷Ó‡œ?wq}9wCƒ¼ì5ûíyž¿I^è;wÃ7ÛŸ‡Òwvµä¥}-µ:s§k
+6ëõ¬u26ŸÎÝ4wŸg?÷°O¢ûÌ鶅ã=ók“ÿ>ïjÁi—Zò!xë:¯qg¬µÍÀÈt«%ù¬Œ\çÀä@x‚ð¯Q¼Ðõ¼Úéž}Ñ8ɵŒ`\æx õ¥'C_öcœ'–·:EÆ_Öñärœ§/œD÷^ NCŽV‡ÿÛO÷¹c„:˜zgÎ÷¬3™£ ×Ó÷UçOOî8f:‡Ý çÆåÎÍØÿ¯µø ~û«ìeÜc\Š ‘Ÿ=ôs›÷.Ò~þËÇø¡Öx4sÑX¾—ù¾³¾¿oÌÉKƆèÄ”ž€Ü…]‚ï :¹n ±üÝsŽ³LîÿEzÚÇ]ódïÐXË÷>«Û®7Çà…?Nó«lœXŽXû ãeöjËQÅá÷î½{O½Çsî¹I Qãt À¨$ÑGA ø øâ“ ‚>û‰£Ýœú读Îñ>,ª»ªÖPk^Ž«¬)w{í¹^gO²)sÝÊMR·¾»Žkoþ°<öš¯êÞøÞOêÞøfòÊ¥ÊI6O÷† ‰þˆC¼sÔsæ¥ÎOe>orÆ;\ãÈyÇ2åÔìgÓ:Ëž÷¹~1{ãV?± öܬµ?êã=Ù‚š°:'GÝÕ•Zžx÷I6:•©þÊ”“Á ïßëî.h^Ä¯× ìw.“ØÔ;ì}bçØ K®2íc™¿wj ýÈE8WÑ÷;¡Ažßí‹pð{ÞBÞwmÝëß+ºä Ô
+ôs¾œ'žû-ëÆqy ÔÕ}™Ûy|Áa層‡ ÷ˆyør϶¥."çöû’û¼ôÛ ;Á÷ ~ÖÉU|,—óú^|œƒl7t”}6ïpßw(óºâžÅ>Ê[¬CëÒvr,уñ jõp-î8Wðfznç1î˜<Ý×’Ñc)yQ^ûð wÇv©ƒS™û.rìtnºøsØQ4ðÌ3¼‘ÚÜ Óž±lCÇæ®C? ïV|Xhð.¯™OˬÞ~>Øž¼á Ñ1æ~Ûy zž™2çk‘Óó&sÈUo݉fÎà":ÀÏÑk/›r_αögî¸OÖÎý'ØÊ3¨mt,Ó¬€ ¯·aÏ èÕõYÜ«¦ŸZGø-ïà.çžcsnB¦ì<·ñv닳cƒ®ëCæ»ôlaÿÏ l¼³Ò >Å»8¿ tö:èŒô_”y¿èþÍ}©W÷þ}ì9'º®ÿ“=£{bâç¾AS¦üÙ—¹ -£m†KŽç”Õôl{î§%¾ç6û‹kMê™»2×OÎæåp
<ƒ½î›¯ßóÕ>èãŸÛàiº¹Ÿqsоï ãÞE4 _Ûøvíå½á Dœ£¾œ=›·5·&=øZOûjW÷Më2ÿ]¹}§g¼Ö;ø&·R#6e^—-wÖ ÷ŠäçMïY÷¶7yš7àk]ã®çGbÿ»ÒíNLa›´ýŸß…ì‹JQ¡¥SûÌxg)¹r¦…ߢÌí â†X]
+õÜë¡žágøŠ÷»ú݉ΘÇÉû଴·
+ì‡í^ênÖtb™_ÖÕ4–²!~‹ŽœW:ÁJºZI×Ô‹½þñ‰•p×A/y'¯¥ð_ÆÙNßèu]æz¤¾»†;Ÿ£ãÌgÞóÜë8Á?ñ—‹dÉz‚]õOÎ<Š½Ï&ô8Ò½ÓYçÄȾÌm¶®z{QÁoq í œY#ÏÝ«»ô —ôÜ_z&õ|ÇÝÎ=_øÜròVÿÛ¾¦e¹ ç)¯å¶ox¾ðüs(í÷ú=)sêeSnßlHû´ô‘½‚kß.ðzñN^æÿºÌ;¦ëù)é$Žõï î>Óµ}[Ç» ¿)s[¤ YóÖåV>óEoîG[:4Ïm¹µ•mêý”Od_—ö¶¾Ý«ÚŸ¬ û@îÙŸ2¶Òæ¾Gêü³6çì辸5K éÁÓsíH Ïu™zz úÏHÔ‰“îŽwÞÕ½¯—¹S/vñïóœ•óŽa¬WÉpW¦™õ,ž—z‡zAëü9â½W¦\²Ñ]¥wŽûy÷ò¹>æ}ê.<ƽÉüºL=(z9 ×==Èé Ý!—s=½·k12Ͻ qxÑÛZóÈVº²?Û‡œ;¼—wò ù—qv½ô—ô¿<gÏúØŠ–í•1ÝòKâ"g<Ó´,™»çÿÖœ‹$ îùM)ÏsàYÐ<“žýn ÷ܯã¬+mÝÁ/í»®ó…g<çþäg›ç»¶ñ_Ú¹©åçëøÏ=øõúfõLš>•µÍyt§;®Kö+Çú¾ÌëLÖËìS]»ì·zfš‡þS|ûÔ¹ñ“÷Q¸†Ä¥Î:Çeý!#¿sä£/Þ¬Yì]…GÎvïå~ý¨wô»ÆýCœo«}æµCÐÛ~_n{oËpŒ»Ùïä >Ï\—½ZòIÝa÷¬ îe³W4=ÇŒc$ßÝŠtIÞñÌ’=cÆ_îƒçÚáqoi[QßíôÎùÙ:^Zþô•>–~^…MF½+S·nÝßweÞ'n*ÞB{#îI8Ö£k«2ù <–xGÖ¤¾Ìs/]™ûúà]îe–õü…ä[‹žý‚^Ø~n g-£ß¦î+ tÒ•)¯ÎC9¦·ât÷ľßÃÛ
+ZI¿EÛk椔uWneoÕŸüwî°
+œ¿mÿçýeÕ¹s±kÀªÜöƒ¦³¯w:ýg¿’ÐÅÿ*öüïÉs÷}Ô«>î¶h<'_ Z4V±·Oõ?íåY'ÝñŸz²Ìû ±,m¹â·|âžû*ògòÎ Îónú =EÖ‹çè¥-[ëÿ“ÇvZ”É¿Ÿ³£g…Í4[w÷qº”²>õ6
{®DïŽÃ®´eÜŠ÷¸®Ê|fj½¯Ó¹ëÑsºÝ”¶_[®]œç ó\ ù]-½·d‚vß~÷xn‚9§å›–ÁrbMÝÙ~}¹µ—Ï3ö×mþíg^Ó“qñ
+Ë¿næ'û¥çÖmÜ]–[»íb¿‹oדUœµäN¾èd,Ž5À=þëòØG0ömã¼2úÆCyœI êÞ½ÎFx¿þ_+ «é\ëú ïÖ:Ü»
+ÿÏu…8gÑÿ¬ïÖÞ%øÜé žÆ½ ¼Q—¿oððýûªã«è=4îÞWÝÃ<Ô="+úkÉy¯ûWÑ;éÌ:³¼ôí)CK?W½-÷ ôNd<Uö‡;í ´Ö¾unyÏÚ?‹/óǹL}/½qÏö«îùœOöÇyh_æ9(g·èÑ'ïÊäÇÇzç®Lý:³ºF–“h¯ÞëEã¬ÿ“Ö]™ü‚w Cä…ÿ©Lý½õÁXÝY·Ö4s¾9”yÎt_¯“ÃЫûù½¾½9Ãú̳}ß®±GÎî·Uk{eo‘¸Þ?ç¾÷ô<€ûÒÜß=ĶæýÖwÊ0žß×õ®Ì}þîw©ÍîÝo¸Ÿw¿H¬ Ø;Ýs-ÅDzŸ"66º·.ó÷BóT÷ðÙE™×ãœIWz«{¥ƒtÁì³hØ~ûÿžýÜAw]ß·(SÌwõ|%|¾7UŸô@ís‡ü—ýà&ÎwÂY6ønâ»ËÆ~Wne],KûMyïmð’Û±c<Ï]™ò >Þ }ã{÷ºç\H ´7þ¿÷2¶ùN?=ÇŸóè:À2îw×G¼›ö¡A3ûÒ½öú°;÷úú~lÀ½^t‰^ûèú(œNàÚî™Æ3‹ca-¾Ò½ó¢Wç[ç÷}cßµ¦sê":ÂfðæÝ}™û#µÚg½xd=ÄNÇ2·i>%gë½¾ãœk ÷Ü_ºçóÿeª îl b`- îí£Ü?gZbŒ;ä+;7â3£]<»97uÂuîÛëÞFgî:}Û_Ý"ƒß‚ŒKÑ÷LºŽkC'\÷(ìuÕn¼k <-û²ÌåöYÖ£ÄC—]¹•?sÈ"ä ³ ò?u9m•µóeè…\”ušäSð|YÏ» Ç¿ýp_æ²[o+ñIþèë.Ο«ÍÖ%12Æ ù†ý“ÎG¸”iÞÀŽ™ ÈßÌ6Ùÿž+E™÷uônà^u~”Lî÷¨Ù·‘§Ø#î\?3×_‚¶ã-ks =ƒsoqŠÜ|/ËTOGËŽŽs¥÷¸ŠÞªÌõ mf3èÑcc«…ðxÛûåq†ÄNî™ð³ìœ±#°ÿc™ò 6ãmGÙl ÷ˆùñîeòêáQoÏ}³ýß3–óÔA8}ÜwÝ8WÝ» fÞ2xÖAgîçLc%ºÔLüŸ^t+:ñ ï± À#ç ÇcWæ¾
+ìªí dKbž‡œ*xÏ} osíz[ÀGÜ <Žm÷9oÔ[Û‚¸ÆOÈe¼ÕñpОñlÛ«=ôŽOÎü뿳¡<ƃõ±ò\Ê<f;Ñâ¾û‚ÌGñï˼ÿyUñ¨k Éó·¾ à£~P óÄ7øÇ ÿ-¾óÙ ðZëøþ ð»~6ÀOêÿ§ãþßøë Vü‘×÷+¯
+ðf€ï
+ð£ŠóŠߨòü¥ÒÏ~]i¼©xŸÖu<ûv½?ÂWë;?à;üp€ÿ”©^þv€wøC}ÛÈïWü¼®àó~9À?øã ¿à_å1Ïþb€Ÿ
+çÏõûËŠóy•ó³úýEÕ%ø¯Ý x3Àv€~€oÕõ8ÀE0îmêý÷8
+ÐUø ïU#ÕuD_O/¯{º§{¦ga†uØc0Ãxc`°±{0B`a ”
+ÇF†àCl°ÃÂb $f‚iZÊOÄO„È">#G?–">Bb#G(È! |$çÔ=·ûÑÌp¤wTw©[·ªnݺõê„ _R2Cñ&ÄË~JòòËh|‚æ2’‘O
+ЦvB O“ö)ˆ·Fë›5Ö&}Z¥CAzr¯¼ü7JqÈ/´¿Kº7iOÒ¢éÔÆéeÝCù) ŽBŸÈïí±qÛ ¶§§Þ¶!U>«“<꘳6ùÝgÓÁz´KÀ)Øt¼}à; zØ ]Î@ÏÙ¹|ÇÐÞƒñ ?yô$|Ô z²9w
+>½ ÿõBÇ# !cÖnÅ^ËÐ_
+{hûj¬y s?4seʹ
+v… úLÿï_ѯ³öy³£lƒOÉO,‡î°îE¬Ûb2ã°%±Ú5ÁÐ&ú3¡ÃLÌÆXÅ÷!ü
+¶C'êÚew©ÇΨ„¹æ'ÚBÿÐÚ%;_Þ»QÁ,è9Ë|:ó;«óy ã9›ãší°Óc½üÞ J Ïâ ÎÂcdtßØ>+_—°g˜C¬}r¢ }{qf{ѦßûCÉüž» ±ßBèÍý_Ýn¾sçPì?4;RÏ‚çEJ[¹Ÿ§´ûÇ }&77 eŒ ‡Œ³±ô âün”þ`L
+DÏؤÏâv÷×ã _ƒ_³p~˜š™³A%З±Žwä5Äï c•à}™‡$;BºkH—
+‚EØëŒÝûBD‡¢|?‘ïJòu”î4ÊøJØ ä=p4ct-æz;܀”w–÷†ôÛ¢/šNÅ~uŠÒå>ý[cwŽ1Æóõ÷žw¯š2v>T¼û³/é,7o–svm—6™ìtpÂúYC úë /Ѽ‰5«@n6¬c~K݇óÍÆ¡x©pvq´ÃàëÁä³äTpþ8®5[m¿0ØönãwØd4d y€{ÖÞ€Þ§mÿ8Αg7ýVBî.Ë µà+ at f²SÁŒíAû0Æ7 ´g‡Ñ¬ù„túÛÀ·{m4[3¦ß÷ ûm¬ýÚ'"kúÌ7 “ß'?±Ïs ¾)Ú }çY.ËãŽÖB^<#<®7¤Ûb„¹¤Æú¤<÷µFË£ì#Ç«vað¬êžnõ»TÃlPÁö¸HÅþ#â§ñÕD š›b9ˆ1¶v½½Þ±{ZŸ†ª&Kþ,Õrª]檦™®±GT‹M é$žæŸ“,îý0‘w,Á½à»Ðè‡V£.@»ÆÞÏ#Я±:÷ƒùz¯½#Œó<ò\ sÅà=ŒŸ°š}rÄÞnù£;Rïu Ô}¬æ»"sóäïª ½¯üšf½WÀ ⟦Z¥Su ë— `I¤ -Ú-,’O¿YãçækŒµâˆÈÜóÒÇ×Ä3´ÇõÉÿ”ÕV«à›5ðÉ‹Ç‚ùíÄÕ!øôåÈĸ{?~^ÞÿãV›Äqgønò¾å,·l²8ÏÁÇu qä*ö³ÈAyÔWiÌónÔ€?´»îîwÂî–£IФ̻ý*Ô½I{ÑíïCæZè¹ye
+bàП*G Cݸ pw1 }Yƒ¥ƒï@ÇcX·Y9f·éÀú)
‰œÄº«ˆ>ã¥hºœ´üŒßéÀ¼ý« ùÆ'Ë”¹ï=ÝûÀIì³Ãd¸7ç¤jYÚ¹k<kþIØ:ê°ÙüX„Œ†òZú™9ÆSç«4r{s÷©ÖôÔË; ¬‡^»Í7µVî‚LOû,çfû¡î_ÀÑX™z}ú@no²{‹ª©›¶Xý”éR`08 <´õ®¿[ðV,Ò¿ÇÖr=XÜ‚ À†JÍÝbýåh¯1=“ªkÁ™ÇTÐÝ Œ~C¿û øs)ÝÛAÏÏêÿíõ&›>á]o
+¾e1éþ3\ÞO›Î¤^ÿ•ˆÓ•èw2¢Y Ôª –Ñiuxof^¼âÉÜ×ú3ªõˆ}滜—ûÌVÆ£ÃLóÁ!Óç Cl2~2²?¥5)‹…É{ÿ.âò¨Åf ž¸â´ ²²ºq»•·2üÂâ©Yà¾äK`¯ä…æç ûû°gôX ŽEð7 NûÁ÷
+ÐmºWïÚ ¨A½Ã}‰á ÆQ/°ïÿA“ÐTäÿ!euƱ×Jåâ‘@“ÞÆkp¾GÊù|c2ʹÌû|ûÚõ–
+‰¹·ø)åö¡Z;]4,•Ü¥ö†2?eägOΧI»þ
+ãKŠÆÅëÇ]]½ª<WM“Ègõö.¸¸þËy¤ÏòaXUÛ¸þœß‚²~!ÚEÀÍ .§Ê1W‘áiRv„ÈŸ¼sÍÌ£ EåÓ¬x¼Íaµa…Áíµ·ÏlþÛo´Öè~{ëÇÀFž%kŽÇUç&ésô‡óí¶‚-v&«Mž1ÁÏÜÿÆÇ*…BJ<Ë[ –r~žyhÑùŽPÌqn¦òÈrí7VèÔÎ÷½çú„ΧFùŽ2šT?ŒF{
+eI¯c¸ÑüZŸÊ±ñªñT§ª!?z™a<¬ñ±Ëí[ÍG#eG\>k—¯Ú´6.ÿ~Mˆú±´—WM©?ðhäŒX;MŒØÓŸ]u‘±è|ú P-7 Ù›þxLàx}„Æ€@4VegV4ª£—ßR Ü'¡¸ñ4%ÿvj¬Ú®êv"ÒOVñxŸ{¾xwø¿jmt}”/j_çmü'ì ) cQ
+Œðƒÿ@üâ¿ÿ& |/&«]H ú
+„{ñUâàË8ÃàK¾·€uz3ˆÅ@·j£Zß'_¤½ŽkÖé«È¹Û¼xº=O´íõ‰®©Zÿ@ûýµBóOÉêÉnݽn!ˆà~Çï·wTj]GÈ¿þ<ƒÛh©jíý÷Y§%Êpã§bîý ©ü*oýí¹ôIÌ•:¢]ðoB+Ш7€È †×Ý?i„¯ ø÷Æñ¿îþ ÁÓ Têƒí¶G³ôc]ÄZ¢µŸw*ú^Í 6‚n´¶¯
+BwhmhuáÍo)çÖÕÌë´µÑxQÖNOÇ\ŽäX[ÌÕÝ=À\Í£dË(Í×I kó…j/”ßø×.Ê:‰õÖ“1—=mÔ^¬áX ²~aÐ" íjÏ‚û—xFr©óÃê‘ÌŽ˜û/à¿Â£Ú¯Yý‘Ú³AãÃD©Ý¬µmÒy°âði'ßcç> ôÿ—Æ¿ÿÓA¯¨Æ)ÿŸa„—í—ø¿soXA¾¬Ó™Ñߣe ß²9‘½›äoFIßqÒs†ôþŠÆÆIÖs^_É›ªñ1ÚÏcŒÎ1ŠáÚ"Ùí’í×M©Úoœl#&ÉŽ‘Z7V<“´nªèhab¤M}ÆW²7/K€bîŸoªlœ¦ó¢/µ·ÿç»E£Îª]:¶h¬EçÉö`Í# Ãú9›öµél|\4Ë×C46D2Ú$³Hª|ÛF ÿ„xxnC•—šW¯”W›Tslxàþ¸oÚÃÔßãÛoëàÿ¼7#{AÏaìwÀÕ ãM ¹ûØ|Œù«À´ÿŒ×w {ük?."_^.C 7Q×Þ*ç¹|p
+:Þ®Aq9ú*Àý?Çü¿M~Ápò® W!ãºÙN=²¢Ã< ¯ÞÃï«ÜW
+ÿßÍ×uЫ(bïìÊ›¯?F]~ ¹á7h_6¿´ÂGÃÑæ|-æ h×Kf‹Ùôèwþºd¹¯Íp ï⯱öXó)æ~‰wóÖþ raKpcŸ`íŸ@ÿû/^nB¯ÿ€Ò ×ÚXþ-@ïœâ¤`2?µ3ª6øÚ· ÿ³—
+>ƒ ®?öe« þ=yg+ç`÷ùÌôÏÙ~çÊaõº"Ü€<îõùÿh¯¶Ø¨®+zçyçá™±ÇØ`HÁIx”ðp)`ÂÃÁBRÔ´…&(’“Äy˜gx'<í Ü´‘ÊW“¨ªøˆ”æ³ý®Ô~VªÔ~¥jÒµöYwæ2±HDºKûœ{öÙgŸ}öÞg ØŽàÙ•1þ?à3Ó3‡yÔ/çýÅôÉ{×ÑÿðgóAžK½÷Wøóß¼IX¯:MÔf1· ÿi‡±˜“ó¾Äžÿ|…µþƒ³ú'Ú}xW~e²Z ÆæÝjú¬â«!R½7 æ9潌@ æEÚ¤^í ê7HNƒú’•‘¬6µíܼ—Ç=€hÄÕÂäéOÏÂ>f+>çíŠù€æ¼[s ý[ˆ‹[ð]×o¬A2æÊAíZ‡ýïS¼3¿„-oã >ý¸y‚¬/Ð^
+»Â¦ôÕëðááßvöðŽ7 ø ü·±ÎUÈâا^ç…ŸGàWQ€±03âê0ËOÏê¿fïC¬; ÙïCÞ Î—t øÏaÍkÛ þίÿë]‚üNôÿ
+ÞkÀ xApÎσ~ 9[°>ÇÙžbÞà]èpÐÖgŽìµzõa˜
+Ô©ýˆw
+ô(ð*b°ôuÐÃÀQà °¶` 9Œ±«@?0„õ.ƒÞÄÿ÷_£}¸&Y%à¢g½_g€÷€z`GŽQÎ9àcìý É<Œ=S¯+Z—ûÞü8|¤58v8 œ߀¿ 9í,Ú;YØ/é-6g½Ç:€5
+}o¶· …•ÚÝWííÃîixOãiÀ¯Ôõóí~ (kkÚf‚ú9ÉkP]ž7~7'£uRØO9"¿‹aÏ
+ØO 6h0ýOãü]†qþË ó$ø.£}ÿésüÿ{È»
+ Ò!»·
+سoc7koà ãEÐ~ÈØ‚¹°Öôwa?Üû.ÌycïšÎuºÞâ3e÷iôYøõßÑÏYûz{ð ÈOl…î=˜÷¬ù7eÆ°—8þ%ÑŽz¯Aÿ%è/K0¶ ÿª¶g-ÊóK˜®‹,ö×ãÑjÑmf'îˇþ¾µËv~Œ§ÉÞRè¹Ôl: ã;«ùfó<þ×ÙXÙÞ@é
+öÉî#¡ ð,úu¶@F7ð2ÀöDzuÙrý €^K= 9ap'qf'ѦÝGB9ôž c-Ö[ ½¹þs ‡Ìvî j€õ§€¦BïÄ‚Î%jçEʽr½€rßï–Ü[sTJ „ æ¸ã ýòó»QÚ£ÇÞ #ÓaËÙ´Y 4Š3Jx/a¯?6 pvXÉÜ™Ë@éeИÇy >Å8¡¯Œ—ðAÒeˆÒM˜CºÅ¨çmÀZÃ÷…EÙÌ·Üܧ³¥}Ãô¨QúWÜbqàhÚè Œ
+ØþêF¥ŒYÆ
+éÏEŸ5Š#ê¦G+}Ú7j1G ãùqÏØ«¥‡¬þˆÝqöeýæ
+Ëc[ç
+å„^“BNf?kà›r/ôÝëÞŒàÍx;@/ƒ¾e2Ò˜³œ öնLJ¼8ƒÇÛÓû êž«+…pöÔœ¶ž»¤ ¿#¿C¯Ñ<†ä®Yox
+zß´õc8GžIÌôÛ¹Ç,'dÀW€ÌdóÞñá)ïþ÷ ÜÏa£Y³ i/ú¼[÷cý¶×´é÷*d¿¹¿@ûJhNÉl7ù%Ù‰}žñSÑè»ÂrY 1š<ß[#,Ö Òm>Â\µ>)Ï}QÏò(ûÈñªãV ë Ö!Ýê/Š¸÷fЩ6ß|‹.U¶xWè?릇E96Çr}Œ5 Ðë ‹Ó(lÊzs¡ätJæ2Õx¬é–?O‡ä-æ‰ožÕÇLÇŸ’,®Í7ã RÄXœkÁv¾ÑëVgv¡ µûó"ôã[nâƒùú¤Ý#ôó<ò\ cEÔ >ΪAòƒývË ¤ -¨ûT/
+½i« „lÌiŠTß`Šžjîqê“õÀ&àq`5ó¸h·°A6}<4'[©«€I¡±§¥±@:r.õÉ¿>ÌØÈ{»a“ óÇ‚ÙíøÕyØô¼åÈ(|ÜÝ GÀËø ´Ú$†˜á½Éx«³ÜÒk~^ç@cÈU¾Õlý˜w óz-6¢à÷-Ö]|Ç-¶ M€&lnuÜW¥îN:‰þÛhÿ2÷@Ï=È+»á/€~ qu|/àb1}Yƒ¥P³á»9Œ×#g¹ sÜt`ý”@NbÝUDŸþR4]†,?ùVw:0o¿nµ#ïøD…2÷SÜŸ†°Îa“áîœ!» ¶Ï˜ÐåfŸ¸Í£o™ ‹ÑP™K;3ÇÔÙ*…\çîÜÓª5È» ìƒ^ÇÍ6«?Af at K–s³#ÐŒÕWŽF*4Чú‘ÝÉî.ª¥ÁÝÔgõ›W¡›NàQ`xhê]¡þ
+ôá®Ø`{Jª&trÊ6Ö4…jîfëoE{·é™P˜g wPA±; ý†×ðÑç’ŠÛ±æÏëô~(Ù%¡{“±>Îû™ùdÆj]—÷S¦3i ÿvøévô»
+iÑ,2jûtX žÅ™o£xÒ÷5Xµ qÚlW§xsõrÉöJ î0ÓßlpÞtçùGà›ôŸ´öŸÔœ¤ùÂQÈdÜŸ‚_^2ßôÀ“ŸÖCVVq³8¨Þ•q¼
+éOM×%_ kÅ!Ï7;_ƒîW±ŸcÐct¸äÁ7Ýt¸œßó W§,¢¨W"ˆ—â0†zý¸Þn‰Qh2ô~HZÝ€ÿXk»rq0F÷ý‰58ï£Få|Þ1iå\æ}Þ}ºË&DÜ]¼Z¹ý!Í](Zl–ÜÍv‡í0;¥egçOΦ ýwýmÆ—‰7øïêê •±Zš@>«·{Á½Abz4VòHÉò¡_SÛ¸~ί«¢Ÿvpc*ÿ“Ÿ«ÊhBûð‘?sMÌ£ EåÓ¬x‚=û^xÛ nð~K¶ßê¿3F3FÏØ]ߎ=ò,Ys,V› ÍÑŸÈ»ÛÞ}v&3Õ&O»÷¡{áÿTå!_HŠ'myk¾å†º`œyhÖùN’Ïql‰òÈV7UèÐ Îû½…çú„Î'ª|GcT?LA{eI—@'úp£Ùµú.ÿŠ§ö²²c Ó ‰g”yÓB6v¹ý€Ù¨MûˆÉf²U‹æÆdßo£qÑà_*WK©?ðXèŒX;Ííg¤}åBÿÂã©{@Ü|hmÚcÀÿõ!<ÑHÍ>³¢a ùÍ5¾À
uâò›€&eßý«ÝWm; ê'jx›|±¾aÿš¹áùa¾ðþ:îà¿bo¤|„DÙÛз 4ü ¾øâ[o„Yï‹ï«]H}ú†{±Uüà»8ÃÑ”|o{ugnÕF™ O¾P{/çZSªÊ¹Û¸xºžp;Ð'<§fþíÔ
+5<Î>e«'»{Ý‚Âýþ¿;ÜÚa=juÝ‚'ûçéÝAË5sï¿Ï:-^û#âîï6åï¼W½ëïÌ¥«0FTëˆV!¸Æºˆ:aœá÷&
+ñ„à¾qü¯¸w!xš€j}pÈÖh’~¬‹XKŒáž
+ßWÞ~ÐýÖjßþ;Œ6×·ºðÛÆû*y†u5ó:÷ÆÚhº(k§'#.Gò_KÄÕÝëå'&k/“5ž“<ÖækÕ^+»ñ×*Ê:‰õÖªˆËÇmÔZ¬áX ²~aÐ,Œ ¥íjϳ‚{K¬‘\ê<SývÉœ qï¾ ÓzMê·iÍýÿžèXµ›4·E:— >©ùä[ ÿ\§Kï¿ÞV| ¦¼Qý§<¼?ý/ÛÏñ½qwXA¶ÌéÌhï)Ú ï²ÎÐÚcdïàŒ ‘¾Ó¤ç¥÷Ãú7M²ž
+ô•¼¹úß®õ´ë آ͒Ý*ÙÁ¼95ëMÓÞˆYÚG›æMÏ,Í›+:E˜jSŸé5 ìýÀN`ðLĽùæjót^´EãÿÙ¯²Ø¨®3|îؾ÷Î̽wöWìñ†0^f 0;f+"dÚª4u“ÚRJi¢ª MH¤6$¡MH€6”¦…R5`Ó:!Q9$J$·âñاJ•Ú¾´ªú~ÿï0×i¤Vªò[ßýÿ³ýçßÎ?çpoÿÏßÅ RŒÕê˜e_–ñ¾ŽãuD* ±³ñ±"ľ<c£ó"C_׳¯ž2ò”™Êz› ½s$n
+¬KiU¾¯ÇXWÓ¼K_“*¿dßUàÙ>fÈ»íyÜ;ç1ÿêã<j⃸—žDÿYÜ®aýÈœW ÀU7€9Ì» µ7€¼cß4}ïFÁM¬/ÃõñäþΗ]ÆUìuÍß/ *º8¤Q¿ï5àjð5ȼ úG´eìtÒ˜‡üÄ<ÎõpKÅocΗã ßóíºŠ¾´oÒÞ[ðéIì1~~? _Šž²÷UôÏ€ÞôõŠ 6æÅ!'¬þà×¾¼Wñ»økøóúç ÷8~{^{¾¹'ý~ûYê-ü¦¾ù×}â°Cô‰ÂQ?¢ã Ö¼íÛWŽÃ¼ªÃZÇ÷Y%> äx˜ Á~âcÏÇSà_ÅúëÀ[¾œ˜šUwaÎ(ú"ˆ‹
+¾ì“ Œ•øûô<øÎ"毾€¾9î£ýWöqœþû~;‰xÍûz8hG 3êϽ ^èг¾]|’öý7ççLZ=½æà»yÚ> ™o@Þ5È}
+rD¿?aýß±æ ØûÏ࿃wå¼?/ÈÙü ;}”çK~gôï¦Ç:'u/BÈ ©‹ò '_Ïv‚rl§(+BY-äe,¦¾Œs| 5Êwa™Sd-èƒ Ëx>圷ñÌkêªKs íKðÑ%ø¤ÜN-B2†¬AmÜGÚyäZ•zG凞zôEà§÷[?Çrj~=
+ŸÎAÖYÄá/è“þw1þ"rî]èñ&æ¿Ž}žƒ,{I…Ô¿€ó0ë è9ÿ,ôå{˜_Ÿ åßÿ²ê‡Ø÷d? y§O¡[1_rô äŽbþè{ÿTÏBþ(ÚÅÜ3ÀiÌ=
+z
+9s è g7ö—±‡|›ªÔQàÛÐá»þþR#ƒú÷Ü£@;à_ª }8ˆ»â7@ }x8 €/ä ù3Œ= ~Âúð<úŸÎƒÿ p†²~ <ãGÕÏÇGS † eLäHNŸƒíߢ̇a³èõcî+v òó>
+»¢êGÜCÆŽO aÍ/0ÿ8êáàŸñý…ÿLµÚÏw¹ïÉ=@î(Š¿ _#÷åû˽ÌåÛoF£ò†\ =þAóþÛõ—€_ñ 9<o‚õÆÂû³îßv‡¹ºí÷ß>g”ï!ò.&9{kU +Ñ}¯•>¹«æÜ)î=Á¶øUîIYê`“J
+Ñ÷ZKÒ†æ ²œ'1aŸðMÔ_b”1*wÒ$i‚ñ•}¬€ÍRƒº2e]÷é <‘¡ï¹iêÛÌv-mIt–zÔ˜ïQv.`wåèw‚M]‚ï‡N®O°-úöÖi{KÆÂ:ÝàOì ÄXr§&@%æî>Î8éü‡þ<AªóàÃ|.þ׳!y/÷÷ͤ’¿C´»9@„~;¬¥üN®ë'z‰1"Nˆy"Gè<—sVKè9úM¢ß ê.´1€º ¼ Ü ‚óõÙѹ×E[$žrÎí€/Š´±•ó$÷Úè§Ô‚QÒ:RéßDãÜWÖʹÒïŒQyÃê3¨ßlúm3*ï«4¡ßÚÏÚ‡y®—ZÔBh ïöåóZOd©GÐO«è ó>Èsn’>’øf[ûTçSý$¹°Î(¿ý¶ÑV‰«¼ƒº©·®_0ÊçRjßJ£R‹šc]Ït K•:ªë‰¶íNuBdÞC}îb{Í"HÿW€/åº"û
+SÿΙd,$_>Ë5#ŒÅÎïâÜ.Žõs~‘ºJmZ¬0*5N ¹(g|·Q>'“ôÏ(©ÄUÎÙ8}õ £œ{‡é“ô¥^s?ù)®£ŽÈO ÕËÿT/Å– R‰—Ü Ö±Ý mı•Ðgjå"L„®ÝD;±œxHNw•š<BÈ<y»ÈYj0*õK£#€æ÷Ag }ìk§M’û›©ßnê'gCêµäØ8ul`ßùÊYε‹k¨ä—œ£½:À~Ña×k jýmŸ>W:+=_÷÷z|)ñI£òû£ý¿hg{9ùÚÒB:F{ÆW]»´ï´ouì–Ñ_íôëRÚ)sw‘~Þ(××5e#eKm׿KmäKôǧŸê臤Žª-WÔèÎ)|;§®ï VÇê}j ÛS?ŽîÙrÅP»¦¦iãñ=—‡‡TÏåø,_†Oo>K»ðéhǧЊϒf|TÏ+ªI5ª†ñÂø§7¾¢ZÔÕL¾MT+ùNÕ¡ÚÉwãÝÛE¾Oõª òýj¹ZF~P
+¨ä‹jX
+ùüå Û±ë¶ølÞ„ÏÝøl Ãgý(>k×à³z>#%Ql¥Q%
+¹ ÏÐUäשµj
+ù
+j½%?®6ª1ò“ên5A~‹Ú¬6‘ÿ˜Ú¦¶’ß©v¨í>gïÇÿiüW!üÙý¥ý¥’ÁR»Ôƒj úÝé&ÙØP¬»4˜)Xî‹«
+_W3*Õø³JC¥âà@.Ûh¤-³ÐÚÙ1\´L4rÙÁRq1ÔÙQ§½LÆK—¿}¶ëÙáˆíºvd_oÚóÒ½é !=Ûm·-Wðl!m®-:Õ«j¨>£"*#»ÍV‡jB!«[ÍÖX¦ŠvE¹áÂ0ö3œ\SسÏE_8êØXëBý¯ªiÕªíIÆSñÑ5Î@aÿ_DKø e[-3“ˆ»É¬ëƼ˜m[5[>Šåã Û ÛfôPsÜuÝTm2‡m;•ˆ$ë“É´c›f´ÖOf}‹*íë\[—¯“ËÖÙ…ÛgËTü¶ÆÌÐk¢G1¬YQÉJÄÜTÖu"– ± ÏËg</¶"®¿ˆ¨fÛf$lÙÆ7s1×uÒéˆeTYN:ëá/e[.Æ윛°í˜k›–ôãY¥ €®hÝ,ÚŽb™mD
+ר†Ê£q¸
+‡ª
+ÉxÊwwÉÜšËdrcc}}Ó_Jõö¦'7l¿¨z俶¯3·M“;³ƒ}+ ÅáR~÷˜2ÜXÚõ’^&íºU¹ t¬hØ´°J5SU©dU•ãzõIÏuëR°(Z—›a³ªºj›‚m t9ۖ༖sµHÏtœ‹¥úHêcPðjdÊqs¦›ñj3ì&XÜÉ$cÉ +êX6ö [ÖcµŽã¦=Ó4ã]455ÇXrг¢©Z=u•ïŸÜq þ1µjÀ–ýcJ…ÄÞþlÿ¦‹Ù§ŸÅ<óAÿµG¡’;™@ò #©Jà w0Á\ÔNxN&ƒ©±Âö¿Ù·ß8®:€ã¿ÝÙËÎmgÎÎÎ®í¬¯/!N’í˜LSÕ)©Ò:¤Rú©\4RÕJPÄ R*‘È””QT ”Ë•
+i UB⥠則ŠwÄ£ùÍîØq¬š…ÇïÌ~fÏœÙÙ=çw~çÌQØï‡Q Ýpü9ÍÇ l=–Y¥WµÞsÂw¢ »I ¯îj^i´u*®ž:~{à¸qä:¶Ç®ó%MX3ÑÓcÙßÿ6;
+_²«ÆÙAê›4̺‘1~˜×¼†ë9íFÓoÇöQ»å:N`û¾ã½úa¶¼Òô-ôN{ÑnúöägìV¿£-ôÛ}ß)Ûdéþ[m“Sf}5
+:Ž:
+OÔ–òáÂÐÑô¶ÎuÎõ²'³SËÖÝ»ËËo>¥ý±Fã÷¶Ž_ªO’£r|4¿Å¤¦œßkQ2N½2è íصckÚÆdjòÐÂüQ+×!Þ7ék«Ùn/ûÅd½~™ðA ôžËOéœ×üÿüR¬ƒâÅ[:¹Û¾f¹¤Ý8Ø\qÜÐwíó»^¹°Ê(Ïʸÿz”;D= ÎW™¿á¥°G}3$iW÷ƒ$Ò44 ?êÕLrx`ú7ú&I§2?ˆÒ¹¹4ô£‰“˜þbÏ2&ïhLt–ÙqDü–?ŠHË Z¯õº|üY³«c²úžH<äÌ@3ÚLŸÉgM`´[Y÷ø±ÜlÞ ÌÌlíΟú[¾ Æ¿dùÖ8öå¸Ú
+§Ñr5Õ‚SI^¹réÒÕïl™ó¸Ç§Ê ˜“©ñ=ÕŸ=Wú““š ÕZ}PØVËGN6\‡®gƒi×z5ß7&LM˜'¾_?´œØÉÄáNÿFž™vY¦ó³&ôý m^H’Nÿب-å|jOŸ<«µ¿OzͶvû´²´2ôNø¾¾¶º×»Ýëã>¥úéô{IÇ+¹´Ê›ƒV¹’¯.å£G¥3\ȆéôÉ“ÓùÉöÉ|ëæÍ;?ÐmçþŸêýÞîýN½a5<§ºÔc½¹¿žëkkcãÎÆF9þñö¿å=¹§÷ÏífÛÂj~ÿiZm~wÚçN¶k^7îIlLœèñBO£˜õ23Æh¡ãL†íéÙ$Ž“Y=¾Ò˲ޱù uZZÑú¢™¥y~Ob™#ãŒãpœ{A&m×ÜûS¿7QÛ;Ëiž
+÷d_ö¼>hB'\èÏ.çùc-Gçœ/ ó<têŸyfó1ÇÑ…®ü_«·4wVzªhÇûýEnÏ‚¶³¯˜(:4ÝŽr×3™ë4ša~Èuü(ž9’†¹ëç™>ÖË,Õ„ö~EíÈ芬³Èk4›Ðrü¸ÅiØO]¿¥‹®Ýl6 ·iÙ^,ãÿs–ÜÖñtdb½2ΫÖ(¯´¾QÍ}¤;oœÎOo™ÍÍq.ÕÕG£\ªþå•é£/M gXïN é ™ìÎw†ÃÎÖuÝîÞº%»[ àÙãõ‰Hí+"õ@]Uzn}_½/ÒxY¤)¡ßÓü‡ˆ½®>q•Û•¿>ÈÕ6¸ë{üEÄ[Sú»Ù=®
Àÿ‡Ô¥&åÖ«,Õb×[eŸúQ¶â{^¹1:n?,Þ(~TÜ*Þ,~\¼Uü¤¸½ïs›Å/Š_¿*ÞÕòoÔïªúßïÞß/þPü±¸3*o=rƒÙ Üò’”™ÑÖÜhÈa9!§å)yZ¾(ÏÊóre{[¯–µŸ“³£Ú‹r©¬-®ìßL&ë¡mpå›ÕÝ–$²“–
+-'UÙÖÒL™½
+OkRYªÊu‰´µã²¥Ÿ¸\•ZþnU¶µüó³ç.œöìâÅË'ÖŸ~õµ—®}ýáÚásrAÎk ÎÊ¢vü²|Y¾!ßÒ/~Y¾*ßþ”ë'd]ƒôª¼¦A½&Ÿö
+ÿû'Ê<÷ÏÚ'zQvcs¸ Oí³zVž¿^»-MâëºËè8~äI½awÛ ö't“5ý®+ziMúòqm¦ú…ºXé>¼÷b|æ_29¾ñÃ¥¯ ßyßþûößd[>ÖÓþÎêòŸ í
+r
+endstream
+endobj
+30 0 obj
+<<
+/Type /ExtGState
+/SA false
+/SM 0.02
+/TR /Identity
+>>
+endobj
+1 0 obj
+<<
+/Type /Pages
+/Kids [ 5 0 R ]
+/Count 1
+>>
+endobj
+2 0 obj
+<<
+/CreationDate (D:20061105011133)
+/Producer (Acrobat Distiller 4.0 for Windows)
+/ModDate (D:20061105011215+09'00')
+>>
+endobj
+xref
+0 3
+0000000000 65535 f
+0000051809 00000 n
+0000051873 00000 n
+trailer
+<<
+/Size 3
+/ID[<21b37df8f836949f6a724722f5ba950d><21b37df8f836949f6a724722f5ba950d>]
+>>
+startxref
+173
+%%EOF
Added: box/trunk/qdbm/misc/icon16.png
===================================================================
(Binary files differ)
Property changes on: box/trunk/qdbm/misc/icon16.png
___________________________________________________________________
Added: svn:mime-type
+ image/png
Added: box/trunk/qdbm/misc/icon20.png
===================================================================
(Binary files differ)
Property changes on: box/trunk/qdbm/misc/icon20.png
___________________________________________________________________
Added: svn:mime-type
+ image/png
Added: box/trunk/qdbm/misc/index.html
===================================================================
--- box/trunk/qdbm/misc/index.html (rev 0)
+++ box/trunk/qdbm/misc/index.html 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,202 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+
+<head>
+<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
+<meta http-equiv="Content-Style-Type" content="text/css" />
+<meta http-equiv="Content-Script-Type" content="text/javascript" />
+<meta name="author" content="Mikio Hirabayashi" />
+<link rel="contents" href="./" />
+<link rel="icon" href="icon16.png" />
+<link rev="made" href="mailto:mikio at users.sourceforge.net" />
+<title>QDBM: Quick Database Manager</title>
+<style type="text/css">html { margin: 0em 0em; padding: 0em 0em; background: #eeeeee none; }
+body { margin: 2em 2em; padding: 0em 0em;
+ background: #eeeeee none; color: #111111;
+ font-style: normal; font-weight: normal; }
+h1 { margin-top: 1.8em; margin-bottom: 1.3em; font-weight: bold; }
+h2 { margin-top: 1.8em; margin-bottom: 1.1em; font-weight: bold;
+ border-left: solid 0.6em #445555; border-bottom: solid 1pt #bbbbbb;
+ padding: 0.5em 0.5em; width: 60%; }
+h3 { margin-top: 1.8em; margin-bottom: 0.8em; font-weight: bold; }
+hr { margin-top: 2.5em; margin-bottom: 1.5em; height: 1pt;
+ color: #999999; background-color: #999999; border: none; }
+div.note,div.navi { text-align: right; }
+div.logo { text-align: center; margin: 3em 0em; }
+div.logo img { border: inset 2pt #ccccdd; }
+p { margin: 0.8em 0em; line-height: 140%; }
+p,dd { text-indent: 0.8em; }
+div,pre { margin-left: 1.7em; margin-right: 1.7em; }
+pre { background-color: #ddddee; padding: 0.2em; border: 1pt solid #bbbbcc; font-size: smaller; }
+kbd { color: #111111; font-style: normal; font-weight: bold; }
+a { color: #0022aa; text-decoration: none; }
+a:hover,a:focus { color: #0033ee; text-decoration: underline; }
+a.head { color: #111111; text-decoration: none; }
+table { padding: 1pt 2pt 1pt 2pt; border: none; margin-left: 1.7em; border-collapse: collapse; }
+th { padding: 1pt 4pt 1pt 4pt; border-style: none;
+ text-align: left; vertical-align: bottom; }
+td { padding: 1pt 4pt 1pt 4pt; border: 1pt solid #333333;
+ text-align: left; vertical-align: top; }
+ul,ol,dl { line-height: 140%; }
+dt { margin-left: 1.2em; }
+dd { margin-left: 2.0em; }
+ul.lines { list-style-type: none; }
+ at media print {
+ html,body { margin: 0em 0em; background-color: #ffffff; color: #000000; }
+ h1 { padding: 8em 0em 0.5em 0em; text-align: center; }
+ h2 { page-break-before: always; }
+ div.note { text-align: center; }
+ div.navi,div.logo { display: none }
+ hr { display: none; }
+ pre { margin: 0.8em 0.8em; background-color: #ffffff;
+ border: 1pt solid #aaaaaa; font-size: smaller; }
+ a,kbd { color: #000000; text-decoration: none; }
+ h1,h2,h3 { font-family: sans-serif; }
+ p,div,li,dt,dd { font-family: serif; }
+ pre,kbd { font-family: monospace; }
+ dd { font-size: smaller; }
+}
+</style>
+<script type="text/javascript">function startup(){
+ var elem = document.getElementById("headline");
+ if(elem){
+ var now = new Date();
+ if((now.getFullYear() + now.getMonth() + now.getDate() + now.getHours()) % 4 == 0){
+ var text;
+ switch((now.getMonth() + now.getDate() + now.getMinutes()) % 10){
+ default: text = "QDBM: Quicker Database Manager"; break;
+ case 1: text = "QDBM: Quantumtheoretically Dangerous Blast Machine"; break;
+ case 3: text = "QDBM: Quite Dramatic Body Metabolism"; break;
+ case 5: text = "QDBM: Quality Documentation of Black Market"; break;
+ case 7: text = "QDBM: Quake Damage Between Mountains"; break;
+ case 9: text = "QDBM: Quiet Down Before Millennium"; break;
+ }
+ elem.firstChild.nodeValue = text;
+ }
+ }
+}
+</script>
+</head>
+
+<body onload="startup();">
+
+<h1 id="headline">QDBM: Quick Database Manager</h1>
+
+<div class="note">Copyright (C) 2000-2007 Mikio Hirabayashi</div>
+<div class="note">Last Update: Thu, 26 Oct 2006 15:00:20 +0900</div>
+<div class="navi">[<span class="void">English</span>/<a href="index.ja.html" hreflang="ja">Japanese</a>]</div>
+
+<div class="logo"><img src="logo.png" alt="QDBM" width="300" height="110" /></div>
+
+<hr />
+
+<h2>Overview</h2>
+
+<p>QDBM is a library of routines for managing a database. The database is a simple data file containing records, each is a pair of a key and a value. Every key and value is serial bytes with variable length. Both binary data and character string can be used as a key and a value. There is neither concept of data tables nor data types. Records are organized in hash table or B+ tree.</p>
+
+<p>As for database of hash table, each key must be unique within a database, so it is impossible to store two or more records with a key overlaps. The following access methods are provided to the database: storing a record with a key and a value, deleting a record by a key, retrieving a record by a key. Moreover, traversal access to every key are provided, although the order is arbitrary. These access methods are similar to ones of DBM (or its followers: NDBM and GDBM) library defined in the UNIX standard. QDBM is an alternative for DBM because of its higher performance.</p>
+
+<p>As for database of B+ tree, records whose keys are duplicated can be stored. Access methods of storing, deleting, and retrieving are provided as with the database of hash table. Records are stored in order by a comparing function assigned by a user. It is possible to access each record with the cursor in ascending or descending order. According to this mechanism, forward matching search for strings and range search for integers are realized. Moreover, transaction is available in database of B+ tree.</p>
+
+<p>QDBM is written in C, and provided as APIs of C, C++, Java, Perl, and Ruby. QDBM is available on platforms which have API conforming to POSIX. QDBM is a free software licensed under the GNU Lesser General Public License.</p>
+
+<hr />
+
+<h2>Documents</h2>
+
+<p>The following are documents of QDBM. They are contained also in the source package.</p>
+
+<ul>
+<li><a href="spex.html">Fundamental Specifications</a></li>
+<li><a href="xspex.html">Specifications of C++ API</a></li>
+<li><a href="jspex.html">Specifications of Java API</a></li>
+<li><a href="plspex.html">Specifications of Perl API</a></li>
+<li><a href="rbspex.html">Specifications of Ruby API</a></li>
+<li><a href="cgispex.html">Specifications of CGI scripts</a></li>
+</ul>
+
+<hr />
+
+<h2>Packages</h2>
+
+<p>The following are packages of QDBM. The Linux binary package contains APIs for C, C++, and Java. It contains CGI scripts also. The Windows binary package contains APIs for C and Java. It contains CGI scripts also.</p>
+
+<ul>
+<li><a href="qdbm-1.8.77.tar.gz">Latest Source Package (version 1.8.77)</a></li>
+</ul>
+
+<ul>
+<li><a href="past/">Past Versions</a></li>
+<li><a href="rpm/">RPM Binary Packages for Linux</a></li>
+<li><a href="win/">Binary Packages for Windows</a></li>
+<li><a href="misc/">Related Packages</a></li>
+</ul>
+
+<ul>
+<li><a href="http://www.freshports.org/databases/qdbm/">Binary Packages for FreeBSD</a></li>
+<li><a href="http://www.blastwave.org/packages.php/qdbm">Binary Packages for Solaris</a></li>
+<li><a href="http://hpux.cs.utah.edu/hppd/cgi-bin/search?package=on&description=on&term=qdbm&Search=Search">Binary Packages for HP-UX</a></li>
+<li><a href="http://www.sbellon.de/sw-ports.html">Binary Packages for RISC OS</a></li>
+</ul>
+
+<hr />
+
+<h2>Applications</h2>
+
+<p>The following are links to applications of QDBM. If you run or find a project using QDBM, please tell it, and it will added to this list.</p>
+
+<ul>
+<li><a href="http://hyperestraier.sourceforge.net/">Hyper Estraier</a> : Full-text Search System for Communities written by Mikio Hirabayashi</li>
+<li><a href="http://estraier.sourceforge.net/">Estraier</a> : Personal Full-text Search System written by Mikio Hirabayashi</li>
+<li><a href="http://diqt.sourceforge.net/">Diqt</a> : Multilingual Dictionary Searcher written by Mikio Hirabayashi</li>
+<li><a href="http://rbbs.sourceforge.jp/">RBBS</a> : WWW-based Bulletin Board System written by Mikio Hirabayashi</li>
+<li><a href="http://harvest.sourceforge.net/">Harvest</a> : Distributed Search System written by Kang-Jin Lee et al</li>
+<li><a href="http://bogofilter.sourceforge.net/">Bogofilter</a> : Bayesian Spam Mail Filter written by Eric S. Raymond et al</li>
+<li><a href="http://www.foo.be/mqs/">MQS</a> : Minimalist Queue Services written by Alexandre Dulaunoy</li>
+<li><a href="http://www.zedshaw.com/projects/ruby_odeum/">Ruby/Odeum</a> : Ruby Binding to the Inverted API written by Zed A. Shaw</li>
+<li><a href="http://www.surfulater.com/">Surfulater</a> : Utility to Save and Organize Web Pages written by Soft As it Gets Pty Ltd</li>
+<li><a href="http://mutt-ng.berlios.de/">Mutt-NG</a> : Text Based Mail Client written by Andreas Krennmair et al</li>
+<li><a href="http://smfs.sourceforge.net/">SMFS</a> : Smart Sendmail Filters written by Eugene Kurmanin</li>
+<li><a href="http://dixit.sourceforge.net/">Dixit</a> : Romanian Dictionary Searcher written by Octavian Procopiuc</li>
+</ul>
+
+<hr />
+
+<h2>Brothers</h2>
+
+<p>There are many followers of UNIX DBM. Select the best suited one for your products. NDBM is ancient and you should not use it. SDBM is maintained by Apache Project, and GDBM is maintained by GNU Project. They are most popular and time-tested. TDB is maintained by Samba Team. It allows multiple simultaneous writers. While CDB does not support updating at a runtime, it is the fastest. Berkeley DB is very multifunctional and ACID compliant. It is used in many commercial products. Finally, QDBM is balanced of performance, functionality, portability, and usability.</p>
+
+<ul>
+<li><a href="bros/ndbm-bsd-5.1.m1.tar.gz">NDBM</a> : New DBM written by BSD</li>
+<li><a href="bros/sdbm-1.0.2.tar.gz">SDBM</a> : Substitute DBM written by Ozan S. Yigit</li>
+<li><a href="bros/gdbm-1.8.3.tar.gz">GDBM</a> : GNU Database Manager written by Philip A. Nelson et al</li>
+<li><a href="bros/tdb-1.0.6.tar.gz">TDB</a> : Trivial Database written by Andrew Tridgell et al</li>
+<li><a href="bros/tinycdb-0.75.tar.gz">TinyCDB</a> : Constant Database written by Michael Tokarev</li>
+<li><a href="bros/db-4.4.20.tar.gz">Berkeley DB</a> : Berkeley DB written by Sleepycat Software</li>
+</ul>
+
+<ul>
+<li><a href="benchmark.pdf">Report of Benchmark Test</a></li>
+<li><a href="http://tokyocabinet.sourceforge.net/">Tokyo Cabinet</a> : the successor of QDBM</li>
+</ul>
+
+<hr />
+
+<h2>Information</h2>
+
+<p>QDBM was written by <a href="http://qdbm.sourceforge.net/mikio/">Mikio Hirabayashi</a>. You can contact the author by e-mail to `mikio at users.sourceforge.net'. However, as for topics which can be shared among other users, pleae send it to the mailing list. To join the mailing list, refer to `<a href="http://lists.sourceforge.net/lists/listinfo/qdbm-users">http://lists.sourceforge.net/lists/listinfo/qdbm-users</a>'.</p>
+
+<p>The project page on SourceForge.net is `<a href="http://sourceforge.net/projects/qdbm/">http://sourceforge.net/projects/qdbm/</a>'.</p>
+
+<p>Update of this project is announced on Freshmeat.net at `<a href="http://freshmeat.net/projects/qdbm/">http://freshmeat.net/projects/qdbm/</a>'.</p>
+
+<hr />
+
+</body>
+
+</html>
+
+<!-- END OF FILE -->
Added: box/trunk/qdbm/misc/index.ja.html
===================================================================
--- box/trunk/qdbm/misc/index.ja.html (rev 0)
+++ box/trunk/qdbm/misc/index.ja.html 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,209 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="ja" lang="ja">
+
+<head>
+<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
+<meta http-equiv="Content-Style-Type" content="text/css" />
+<meta http-equiv="Content-Script-Type" content="text/javascript" />
+<meta name="author" content="Mikio Hirabayashi" />
+<link rel="contents" href="./" />
+<link rel="icon" href="icon16.png" />
+<link rev="made" href="mailto:mikio at users.sourceforge.net" />
+<title>データベースライブラリ QDBM</title>
+<style type="text/css">html { margin: 0em 0em; padding: 0em 0em; background: #eeeeee none; }
+body { margin: 2em 2em; padding: 0em 0em;
+ background: #eeeeee none; color: #111111;
+ font-style: normal; font-weight: normal; }
+h1 { margin-top: 1.8em; margin-bottom: 1.3em; font-weight: bold; }
+h2 { margin-top: 1.8em; margin-bottom: 1.1em; font-weight: bold;
+ border-left: solid 0.6em #445555; border-bottom: solid 1pt #bbbbbb;
+ padding: 0.5em 0.5em; width: 60%; }
+h3 { margin-top: 1.8em; margin-bottom: 0.8em; font-weight: bold; }
+hr { margin-top: 2.5em; margin-bottom: 1.5em; height: 1pt;
+ color: #999999; background-color: #999999; border: none; }
+div.note,div.navi { text-align: right; }
+div.logo { text-align: center; margin: 3em 0em; }
+div.logo img { border: inset 2pt #ccccdd; }
+p { margin: 0.8em 0em; line-height: 140%; }
+p,dd { text-indent: 0.8em; }
+div,pre { margin-left: 1.7em; margin-right: 1.7em; }
+pre { background-color: #ddddee; padding: 0.2em; border: 1pt solid #bbbbcc; font-size: smaller; }
+kbd { color: #111111; font-style: normal; font-weight: bold; }
+a { color: #0022aa; text-decoration: none; }
+a:hover,a:focus { color: #0033ee; text-decoration: underline; }
+a.head { color: #111111; text-decoration: none; }
+table { padding: 1pt 2pt 1pt 2pt; border: none; margin-left: 1.7em; border-collapse: collapse; }
+th { padding: 1pt 4pt 1pt 4pt; border-style: none;
+ text-align: left; vertical-align: bottom; }
+td { padding: 1pt 4pt 1pt 4pt; border: 1pt solid #333333;
+ text-align: left; vertical-align: top; }
+ul,ol,dl { line-height: 140%; }
+dt { margin-left: 1.2em; }
+dd { margin-left: 2.0em; }
+ul.lines { list-style-type: none; }
+ at media print {
+ html,body { margin: 0em 0em; background-color: #ffffff; color: #000000; }
+ h1 { padding: 8em 0em 0.5em 0em; text-align: center; }
+ h2 { page-break-before: always; }
+ div.note { text-align: center; }
+ div.navi,div.logo { display: none }
+ hr { display: none; }
+ pre { margin: 0.8em 0.8em; background-color: #ffffff;
+ border: 1pt solid #aaaaaa; font-size: smaller; }
+ a,kbd { color: #000000; text-decoration: none; }
+ h1,h2,h3 { font-family: sans-serif; }
+ p,div,li,dt,dd { font-family: serif; }
+ pre,kbd { font-family: monospace; }
+ dd { font-size: smaller; }
+}
+</style>
+<script type="text/javascript">function startup(){
+ var elem = document.getElementById("headline");
+ if(elem){
+ var now = new Date();
+ if((now.getFullYear() + now.getMonth() + now.getDate() + now.getHours()) % 4 == 0){
+ var text;
+ switch((now.getMonth() + now.getDate() + now.getMinutes()) % 24){
+ default: text = "QDBM: 高速資料基盤管ç†å™¨"; break;
+ case 1: text = "QDBM: NDBMã¨ã¯é•ã†ã®ã よã€NDBMã¨ã¯ï¼"; break;
+ case 3: text = "QDBM: QDBMã¯ä¼Šé”ã˜ã‚ƒãªã„ã£ï¼"; break;
+ case 5: text = "QDBM: æ•¢ãˆã¦è¨€ãŠã†ã€é«˜é€Ÿã§ã‚ã‚‹ã¨ï¼"; break;
+ case 7: text = "QDBM: ã—ã‹ã—ã€ç§ã‚‚DBMã®ã¯ãšã ï¼"; break;
+ case 9: text = "QDBM: 今ã®ç§ã¯QDBMã 。ãれ以上ã§ã‚‚ãれ以下ã§ã‚‚ãªã„。"; break;
+ case 11: text = "QDBM: ç§ã¯Berkeleyã¨ã¯é–¢ä¿‚ãªã„。ç§ã¯ã„ã¤ã‚‚一人ã®DBMã ã£ãŸã€‚"; break;
+ case 13: text = "QDBM: 見ã›ã¦ã‚‚らãŠã†ã‹ï¼Berkeleyã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ€§èƒ½ã¨ã‚„らをï¼"; break;
+ case 15: text = "QDBM: 悲ã—ã„ã‘ã©ã“ã‚Œã€DBMãªã®ã‚ˆã。"; break;
+ case 17: text = "QDBM: æ–°ã—ã„時代を創るã®ã¯SQLã§ã¯ãªã„ï¼ï¼"; break;
+ case 19: text = "QDBM: ãˆã‡ãƒã€ãƒžã‚¤ãƒŠãƒ¼ãƒã‚§ãƒ³ã‚¸ã®ãã›ã«ã£ï¼"; break;
+ case 21: text = "QDBM: よãã‚‚ã“ã‚“ãªããŸã³ã‚ŒãŸDBMãŒç¾å½¹ã§ã„られるもã®ã 。"; break;
+ case 22: text = "QDBM: 見ãˆã‚‹ï¼ç§ã«ã‚‚レコードãŒè¦‹ãˆã‚‹"; break;
+ case 23: text = "QDBM: ãã‚“ãªRDBã¿ãŸã„ãªå£ã®åˆ©ãæ–¹ã€ãŠã‚„ã‚ãªã•ã„ï¼"; break;
+ }
+ elem.firstChild.nodeValue = text;
+ }
+ }
+}
+</script>
+</head>
+
+<body onload="startup();">
+
+<h1 id="headline">QDBM: Quick Database Manager</h1>
+
+<div class="note">Copyright (C) 2000-2007 Mikio Hirabayashi</div>
+<div class="note">Last Update: Thu, 26 Oct 2006 15:00:20 +0900</div>
+<div class="navi">[<a href="index.html" hreflang="en">English</a>/<span class="void">Japanese</span>]</div>
+
+<div class="logo"><img src="logo.png" alt="QDBM" width="300" height="110" /></div>
+
+<hr />
+
+<h2><a name="overview" id="overview" class="head">概è¦</a></h2>
+
+<p>QDBMã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†ãƒ«ãƒ¼ãƒãƒ³ç¾¤ã®ãƒ©ã‚¤ãƒ–ラリã§ã™ã€‚データベースã¨ã„ã£ã¦ã‚‚å˜ç´”ãªã‚‚ã®ã§ã€ã‚ーã¨å€¤ã®ãƒšã‚¢ã‹ã‚‰ãªã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ç¾¤ã‚’æ ¼ç´ã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ•ã‚¡ã‚¤ãƒ«ã§ã™ã€‚ã‚ーã¨å€¤ã¯ä»»æ„ã®é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆåˆ—ã§ã‚ã‚Šã€æ–‡å—列ã§ã‚‚ãƒã‚¤ãƒŠãƒªã§ã‚‚扱ã†ã“ã¨ãŒã§ãã¾ã™ã€‚テーブルやデータ型ã®æ¦‚念ã¯ã‚ã‚Šã¾ã›ã‚“。レコードã¯ãƒãƒƒã‚·ãƒ¥è¡¨ã¾ãŸã¯B+木ã§ç·¨æˆã•ã‚Œã¾ã™ã€‚</p>
+
+<p>ãƒãƒƒã‚·ãƒ¥è¡¨ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ã€ã‚ーã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã§ä¸€æ„ã§ã‚ã‚Šã€ã‚ーãŒé‡è¤‡ã™ã‚‹è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ã¯ã§ãã¾ã›ã‚“。ã“ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«å¯¾ã—ã¦ã¯ã€ã‚ーã¨å€¤ã‚’指定ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã—ãŸã‚Šã€ã‚ーを指定ã—ã¦å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã—ãŸã‚Šã€ã‚ーを指定ã—ã¦å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ã¾ãŸã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã—ã¦ã‚ã‚‹å…¨ã¦ã®ã‚ãƒ¼ã‚’é †ä¸åŒã«ä¸€ã¤ãšã¤å–り出ã™ã“ã¨ã‚‚ã§ãã¾ã™ã€‚ã“ã®ã‚ˆã†ãªæ“作ã¯ã€UNIX標準ã§å®šç¾©ã•ã‚Œã¦ã„ã‚‹DBMライブラリãŠã‚ˆã³ãã®è¿½å¾“ã§ã‚ã‚‹NDBMã‚„GDBMã«é¡žã™ã‚‹ã‚‚ã®ã§ã™ã€‚QDBMã¯DBMã®ã‚ˆã‚Šè‰¯ã„代替ã¨ã—ã¦åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚</p>
+
+<p>B+木ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ã€ã‚ーãŒé‡è¤‡ã™ã‚‹è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ã“ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«å¯¾ã—ã¦ã¯ã€ãƒãƒƒã‚·ãƒ¥è¡¨ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¨åŒæ§˜ã«ã€ã‚ーを指定ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã—ãŸã‚Šå–り出ã—ãŸã‚Šå‰Šé™¤ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚レコードã¯ãƒ¦ãƒ¼ã‚¶ãŒæŒ‡ç¤ºã—ãŸæ¯”較関数ã«åŸºã¥ã„ã¦æ•´åˆ—ã•ã‚Œã¦æ ¼ç´ã•ã‚Œã¾ã™ã€‚カーソルを用ã„ã¦å„ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ˜‡é †ã¾ãŸã¯é™é †ã§å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ã“ã®æ©Ÿæ§‹ã«ã‚ˆã£ã¦ã€æ–‡å—列ã®å‰æ–¹ä¸€è‡´æ¤œç´¢ã‚„数値ã®ç¯„囲検索ãŒå¯èƒ½ã«ãªã‚Šã¾ã™ã€‚ã¾ãŸã€B+木ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ãŒåˆ©ç”¨ã§ãã¾ã™ã€‚</p>
+
+<p>QDBMã¯Cã§è¨˜è¿°ã•ã‚Œã€Cã€C++ã€Javaã€PerlãŠã‚ˆã³Rubyã®APIã¨ã—ã¦æä¾›ã•ã‚Œã¾ã™ã€‚QDBMã¯POSIXæº–æ‹ ã®APIã‚’å‚™ãˆã‚‹ãƒ—ラットフォームã§åˆ©ç”¨ã§ãã¾ã™ã€‚QDBMã¯GNU Lesser General Public Licenseã«åŸºã¥ãフリーソフトウェアã§ã™ã€‚</p>
+
+<hr />
+
+<h2>文書</h2>
+
+<p>以下ã®æ–‡æ›¸ã‚’ãŠèªã¿ãã ã•ã„。ソースパッケージã«ã‚‚åŒã˜ã‚‚ã®ãŒå«ã¾ã‚Œã¦ã„ã¾ã™ã€‚</p>
+
+<ul>
+<li><a href="spex-ja.html">基本仕様書</a></li>
+<li><a href="xspex-ja.html">C++用API仕様書</a></li>
+<li><a href="jspex-ja.html">Java用API仕様書</a></li>
+<li><a href="plspex-ja.html">Perl用API仕様書</a></li>
+<li><a href="rbspex-ja.html">Ruby用API仕様書</a></li>
+<li><a href="cgispex-ja.html">CGIスクリプト仕様書</a></li>
+<li><a href="tutorial-ja.html">ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«</a></li>
+</ul>
+
+<hr />
+
+<h2>ダウンãƒãƒ¼ãƒ‰</h2>
+
+<p>以下ã®ãƒ‘ッケージをダウンãƒãƒ¼ãƒ‰ã—ã¦ãã ã•ã„。Linux用ãƒã‚¤ãƒŠãƒªãƒ‘ッケージã¯ã€Cã¨C++ã¨Javaã®APIã¨ã€CGIスクリプトをç´ã‚ã¦ã„ã¾ã™ã€‚Windows用ãƒã‚¤ãƒŠãƒªãƒ‘ッケージã¯ã€Cã¨Javaã®APIã¨ã€CGIスクリプトをç´ã‚ã¦ã„ã¾ã™ã€‚</p>
+
+<ul>
+<li><a href="qdbm-1.8.77.tar.gz">最新ã®ã‚½ãƒ¼ã‚¹ãƒ‘ッケージ（ãƒãƒ¼ã‚¸ãƒ§ãƒ³1.8.77）</a></li>
+</ul>
+
+<ul>
+<li><a href="past/">éŽåŽ»ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³</a></li>
+<li><a href="rpm/">Linux用ãƒã‚¤ãƒŠãƒªãƒ‘ッケージ（RPM）</a></li>
+<li><a href="win/">Windows用ãƒã‚¤ãƒŠãƒªãƒ‘ッケージ</a></li>
+<li><a href="misc/">関連ã™ã‚‹ãƒ‘ッケージ</a></li>
+</ul>
+
+<ul>
+<li><a href="http://www.freshports.org/databases/qdbm/">FreeBSD用パッケージ</a></li>
+<li><a href="http://www.blastwave.org/packages.php/qdbm">Solaris用パッケージ</a></li>
+<li><a href="http://hpux.cs.utah.edu/hppd/cgi-bin/search?package=on&description=on&term=qdbm&Search=Search">HP-UX用パッケージ</a></li>
+<li><a href="http://www.sbellon.de/sw-ports.html">RISC OS用パッケージ</a></li>
+</ul>
+
+<hr />
+
+<h2>アプリケーション</h2>
+
+<p>QDBMã®ã‚¢ãƒ—リケーションã«ã¯ä»¥ä¸‹ã®ã‚‚ã®ãŒã‚ã‚Šã¾ã™ã€‚ã‚‚ã—ã‚ãªãŸã®ãƒ—ãƒã‚¸ã‚§ã‚¯ãƒˆãŒQDBMを使ã£ã¦ã„ã‚‹ãªã‚‰ã€é€£çµ¡ã‚’ã„ãŸã ã‘ã‚Œã°ã“ã®ãƒªã‚¹ãƒˆã«è¿½åŠ ã„ãŸã—ã¾ã™ã€‚</p>
+
+<ul>
+<li><a href="http://hyperestraier.sourceforge.net/">Hyper Estraier</a> : 平林幹雄ã«ã‚ˆã‚‹å…±åŒä½“的全文検索システム</li>
+<li><a href="http://estraier.sourceforge.net/">Estraier</a> : 平林幹雄ã«ã‚ˆã‚‹å€‹äººç”¨å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ </li>
+<li><a href="http://diqt.sourceforge.net/">Diqt</a> : 平林幹雄ã«ã‚ˆã‚‹å¤šè¨€èªžè¾žæ›¸æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ </li>
+<li><a href="http://rbbs.sourceforge.jp/">RBBS</a> : 平林幹雄ã«ã‚ˆã‚‹Web掲示æ¿ã‚·ã‚¹ãƒ†ãƒ </li>
+<li><a href="http://harvest.sourceforge.net/">Harvest</a> : Kang-Jin Leeらã«ã‚ˆã‚‹åˆ†æ•£æƒ…å ±æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ </li>
+<li><a href="http://bogofilter.sourceforge.net/">Bogofilter</a> : Eric S. Raymondらã«ã‚ˆã‚‹ãƒ™ã‚¤ã‚ºå¼ã‚¹ãƒ‘ムメールフィルタ</li>
+<li><a href="http://www.foo.be/mqs/">MQS</a> : Alexandre Dulaunoyã«ã‚ˆã‚‹ã‚ュー管ç†ã‚µãƒ¼ãƒ“ス</li>
+<li><a href="http://www.zedshaw.com/projects/ruby_odeum/">Ruby/Odeum</a> : Zed A. Shawã«ã‚ˆã‚‹è»¢ç½®APIã®Ruby用インターフェイス</li>
+<li><a href="http://www.surfulater.com/">Surfulater</a> : Soft As it Gets Pty社ã«ã‚ˆã‚‹Webページスクラップブック</li>
+<li><a href="http://mutt-ng.berlios.de/">Mutt-NG</a> : Andreas Krennmairらã«ã‚ˆã‚‹ãƒ†ã‚スト型メールクライアント</li>
+<li><a href="http://smfs.sourceforge.net/">SMFS</a> : Eugene Kurmaninã«ã‚ˆã‚‹Sendmail用フィルタシステム</li>
+<li><a href="http://dixit.sourceforge.net/">Dixit</a> : Octavian Procopiucã«ã‚ˆã‚‹ãƒ«ãƒ¼ãƒžãƒ‹ã‚¢èªžè¾žæ›¸æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ </li>
+</ul>
+
+<hr />
+
+<h2>兄弟</h2>
+
+<p>UNIX標準ã®DBMã«è¿½å¾“ã—ãŸãƒ©ã‚¤ãƒ–ラリã¯ãŸãã•ã‚“ã‚ã‚Šã¾ã™ã€‚ãã‚Œãžã‚Œç‰¹å¾´ãŒã‚ã‚Šã¾ã™ã®ã§ã€ã‚ãªãŸã®è£½å“ã«ã¨ã£ã¦éƒ½åˆãŒã„ã„ã‚‚ã®ã‚’é¸ã‚“ã§ãã ã•ã„。NDBMã¯å¤ã„ã®ã§ã€ã‚‚ã¯ã‚„使ã†ç†ç”±ã¯ãªã„ã§ã—ょã†ã€‚SDBMã¯Apache Projectã§ç®¡ç†ã•ã‚Œã€GDBMã¯GNU Projectã§ç®¡ç†ã•ã‚Œã¦ã„ã¾ã™ã€‚ãれらã¯æœ€ã‚‚人気ãŒã‚ã‚Šã€ã¾ãŸæ•°å¤šãテストã•ã‚Œã¦ã„ã‚‹ã‚‚ã®ã§ã—ょã†ã€‚TDBã¯Samba Teamã§ç®¡ç†ã•ã‚Œã¦ãŠã‚Šã€è¤‡æ•°åŒæ™‚ã®æ›¸ãè¾¼ã¿ãŒå¯èƒ½ã§ã™ã€‚TinyCDBã¯éšæ™‚ã®æ›´æ–°ã¯ã§ãã¾ã›ã‚“ãŒã€æœ€ã‚‚高速ã«å‹•ä½œã—ã¾ã™ã€‚Berkeley DBã¯éžå¸¸ã«å¤šæ©Ÿèƒ½ã§ã€ACIDæº–æ‹ ã§ã€å¤šãã®å•†ç”¨è£½å“ã§ä½¿ã‚ã‚Œã¦ã„ã¾ã™ã€‚最後ã«ã€QDBMã¯æ€§èƒ½ã¨æ©Ÿèƒ½ã¨ç§»æ¤æ€§ã¨åˆ©ä¾¿æ€§ã®ãƒãƒ©ãƒ³ã‚¹ãŒã¨ã‚ŒãŸè£½å“ã§ã™ã€‚</p>
+
+<ul>
+<li><a href="bros/ndbm-bsd-5.1.m1.tar.gz">NDBM</a> : BSDã«ã‚ˆã‚‹New DBM</li>
+<li><a href="bros/sdbm-1.0.2.tar.gz">SDBM</a> : Ozan S. Yigitã«ã‚ˆã‚‹Substitute DBM</li>
+<li><a href="bros/gdbm-1.8.3.tar.gz">GDBM</a> : Philip A. Nelsonらã«ã‚ˆã‚‹GNU Database Manager</li>
+<li><a href="bros/tdb-1.0.6.tar.gz">TDB</a> : Andrew Tridgellらã«ã‚ˆã‚‹Trivial Database</li>
+<li><a href="bros/tinycdb-0.75.tar.gz">TinyCDB</a> : Michael Tokarevã«ã‚ˆã‚‹Constant Database</li>
+<li><a href="bros/db-4.4.20.tar.gz">Berkeley DB</a> : Sleepycat Software社ã«ã‚ˆã‚‹Berkeley DB</li>
+</ul>
+
+<ul>
+<li><a href="benchmark.pdf">ベンãƒãƒžãƒ¼ã‚¯ãƒ†ã‚¹ãƒˆã®ãƒ¬ãƒãƒ¼ãƒˆ</a></li>
+<li><a href="http://tokyocabinet.sourceforge.net/">Tokyo Cabinet</a> : QDBMã®å¾Œç¶™è£½å“</li>
+</ul>
+
+<hr />
+
+<h2>ãã®ä»–ã®æƒ…å ±</h2>
+
+<p>QDBMã¯<a href="http://qdbm.sourceforge.net/mikio/">平林幹雄</a>ãŒä½œæˆã—ã¾ã—ãŸã€‚作者ã¨é€£çµ¡ã‚’ã¨ã‚‹ã«ã¯ã€`mikio at users.sourceforge.net' å®›ã«é›»åメールをé€ã£ã¦ãã ã•ã„。ãŸã ã—ã€è³ªå•ã‚„ãƒã‚°ãƒ¬ãƒãƒ¼ãƒˆãªã©ã€ä»–ã®ãƒ¦ãƒ¼ã‚¶ã¨å…±æœ‰ã§ãる話題ã¯ãƒ¡ãƒ¼ãƒªãƒ³ã‚°ãƒªã‚¹ãƒˆã«é€ã£ã¦ãã ã•ã„。メーリングリストã®å‚åŠ æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€`<a href="http://lists.sourceforge.net/lists/listinfo/qdbm-users">http://lists.sourceforge.net/lists/listinfo/qdbm-users</a>' ã‚’å‚ç…§ã—ã¦ãã ã•ã„。</p>
+
+<p>SourceForge.netã«ãŠã‘るプãƒã‚¸ã‚§ã‚¯ãƒˆãƒšãƒ¼ã‚¸ã¯ `<a href="http://sourceforge.net/projects/qdbm/">http://sourceforge.net/projects/qdbm/</a>' ã«ã‚ã‚Šã¾ã™ã€‚Freshmeatã«ãŠã‘ã‚‹æ›´æ–°æƒ…å ±ã¯ `<a href="http://freshmeat.net/projects/qdbm/">http://freshmeat.net/projects/qdbm/</a>' ã‹ã‚‰å‚ç…§ã§ãã¾ã™ã€‚</p>
+
+<hr />
+
+</body>
+
+</html>
+
+<!-- END OF FILE -->
Added: box/trunk/qdbm/misc/logo.png
===================================================================
(Binary files differ)
Property changes on: box/trunk/qdbm/misc/logo.png
___________________________________________________________________
Added: svn:mime-type
+ image/png
Added: box/trunk/qdbm/misc/makevcdef
===================================================================
--- box/trunk/qdbm/misc/makevcdef (rev 0)
+++ box/trunk/qdbm/misc/makevcdef 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,48 @@
+#! /bin/sh
+
+#================================================================
+# makevcdef
+# Generator of module definition file for Visual C++
+#================================================================
+
+
+# check arguments
+file="$1"
+if [ -f "$file" ]
+then
+ true
+else
+ printf 'usage: makevcdef library\n' 1>&2
+ exit 1
+fi
+
+
+# show headers
+name=`echo $file | sed -e 's/^lib//' -e 's/\..*//'`
+printf 'EXPORTS\r\n'
+
+
+# show variables
+nm -g "$file" | grep ' [BDR] ' | sed 's/.* [BDR] //' | grep -v '^_' | sort | uniq |
+while read name
+do
+ num=$((num + 1))
+ printf ' %s = %s DATA\r\n' "$name" "$name"
+done
+
+
+# show functions
+nm -g "$file" | grep ' T ' | sed 's/.* T //' | grep -v '^_' | sort | uniq |
+while read name
+do
+ num=$((num + 1))
+ printf ' %s = %s\r\n' "$name" "$name"
+done
+
+
+# exit normally
+exit 0
+
+
+
+# END OF FILE
Property changes on: box/trunk/qdbm/misc/makevcdef
___________________________________________________________________
Added: svn:executable
+ *
Added: box/trunk/qdbm/misc/mymemo-ja.html
===================================================================
--- box/trunk/qdbm/misc/mymemo-ja.html (rev 0)
+++ box/trunk/qdbm/misc/mymemo-ja.html 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,34 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="ja" lang="ja">
+
+<head>
+<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
+<meta name="author" content="Mikio Hirabayashi" />
+<link rel="contents" href="./" />
+<title>My Private Memo for QDBM</title>
+</head>
+
+<body>
+
+<h1>QDBMã®ãŸã‚ã®ç§çš„メモ</h1>
+
+<hr />
+
+<h2>cygwin+mingwã§ã®ãƒ“ルド環境ã®è¨å®šæ–¹æ³•</h2>
+
+<ul>
+<li>Cygwinã®ã‚»ãƒƒãƒˆã‚¢ãƒƒãƒ—時ã«ã€mingwã®gccã¨mingwã®zlibã‚‚é¸æŠžã™ã‚‹ã€‚</li>
+<li>mingwã®libiconvã¯mingwã®ã‚µã‚¤ãƒˆã‹ã‚‰æ‰‹ã«å…¥ã‚Œã‚‹ã€‚ãã—ã¦ã€*.h 㯠/usr/include/mingw ã®ä¸‹ã«ã€*.dll.a 㯠/lib/mingw ã®ä¸‹ã«ã€*.dll 㯠/binã®ä¸‹ã«ç§»å‹•ã•ã›ã‚‹ã€‚</li>
+<li>QDBMã®ãƒ“ルド環境ã®è¨å®šã¯ ./configure --enable-zlib --enable-iconv ã§è¡Œã„ã€ãƒ‘ッケージã®ä½œæˆã¯ make win32pkg ã§è¡Œã†ã€‚</li>
+</ul>
+
+<hr />
+
+</body>
+
+</html>
+
+<!-- END OF FILE -->
Added: box/trunk/qdbm/misc/tutorial-ja.html
===================================================================
--- box/trunk/qdbm/misc/tutorial-ja.html (rev 0)
+++ box/trunk/qdbm/misc/tutorial-ja.html 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,622 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="ja" lang="ja">
+
+<head>
+<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
+<meta http-equiv="Content-Style-Type" content="text/css" />
+<meta name="author" content="Mikio Hirabayashi" />
+<meta name="keywords" content="QDBM, tutrial" />
+<meta name="description" content="tutorial of QDBM" />
+<link rel="contents" href="./" />
+<link rev="made" href="mailto:mikio at users.sourceforge.net" />
+<title>Tutorial of QDBM Version 1 (Japanese)</title>
+<style type="text/css">html { margin: 0em 0em; padding: 0em 0em; background: #eeeeee none; }
+body { margin: 2em 2em; padding: 0em 0em;
+ background: #eeeeee none; color: #111111;
+ font-style: normal; font-weight: normal; }
+h1 { margin-top: 1.8em; margin-bottom: 1.3em; font-weight: bold; }
+h2 { margin-top: 1.8em; margin-bottom: 1.1em; font-weight: bold;
+ border-left: solid 0.6em #445555; border-bottom: solid 1pt #bbbbbb;
+ padding: 0.5em 0.5em; width: 60%; }
+h3 { margin-top: 1.8em; margin-bottom: 0.8em; font-weight: bold; }
+hr { margin-top: 2.5em; margin-bottom: 1.5em; height: 1pt;
+ color: #999999; background-color: #999999; border: none; }
+div.note,div.navi { text-align: right; }
+div.logo { text-align: center; margin: 3em 0em; }
+div.logo img { border: inset 2pt #ccccdd; }
+p { margin: 0.8em 0em; line-height: 140%; }
+p,dd { text-indent: 0.8em; }
+div,pre { margin-left: 1.7em; margin-right: 1.7em; }
+pre { background-color: #ddddee; padding: 0.2em; border: 1pt solid #bbbbcc; font-size: smaller; }
+kbd { color: #111111; font-style: normal; font-weight: bold; }
+a { color: #0022aa; text-decoration: none; }
+a:hover,a:focus { color: #0033ee; text-decoration: underline; }
+a.head { color: #111111; text-decoration: none; }
+table { padding: 1pt 2pt 1pt 2pt; border: none; margin-left: 1.7em; border-collapse: collapse; }
+th { padding: 1pt 4pt 1pt 4pt; border-style: none;
+ text-align: left; vertical-align: bottom; }
+td { padding: 1pt 4pt 1pt 4pt; border: 1pt solid #333333;
+ text-align: left; vertical-align: top; }
+ul,ol,dl { line-height: 140%; }
+dt { margin-left: 1.2em; }
+dd { margin-left: 2.0em; }
+ul.lines { list-style-type: none; }
+ at media print {
+ html,body { margin: 0em 0em; background-color: #ffffff; color: #000000; }
+ h1 { padding: 8em 0em 0.5em 0em; text-align: center; }
+ h2 { page-break-before: always; }
+ div.note { text-align: center; }
+ div.navi,div.logo { display: none }
+ hr { display: none; }
+ pre { margin: 0.8em 0.8em; background-color: #ffffff;
+ border: 1pt solid #aaaaaa; font-size: smaller; }
+ a,kbd { color: #000000; text-decoration: none; }
+ h1,h2,h3 { font-family: sans-serif; }
+ p,div,li,dt,dd { font-family: serif; }
+ pre,kbd { font-family: monospace; }
+ dd { font-size: smaller; }
+}
+</style>
+</head>
+
+<body>
+
+<h1>QDBMã®ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«</h1>
+
+<div class="note">Copyright (C) 2000-2007 Mikio Hirabayashi</div>
+<div class="note">Last Update: Thu, 26 Oct 2006 15:00:20 +0900</div>
+<div class="navi">[<a href="http://qdbm.sourceforge.net/">Home</a>]</div>
+
+<hr />
+
+<h2>目次</h2>
+
+<ol>
+<li><a href="#introduction">イントãƒãƒ€ã‚¯ã‚·ãƒ§ãƒ³</a></li>
+<li><a href="#depot">Depot: 基本API</a></li>
+<li><a href="#curia">Curia: æ‹¡å¼µAPI</a></li>
+<li><a href="#relic">Relic: NDBM互æ›API</a></li>
+<li><a href="#hovel">Hovel: GDBM互æ›API</a></li>
+<li><a href="#cabin">Cabin: ユーティリティAPI</a></li>
+<li><a href="#villa">Villa: 上級API</a></li>
+<li><a href="#odeum">Odeum: 転置API</a></li>
+</ol>
+
+<hr />
+
+<h2><a name="introduction" id="introduction" class="head">イントãƒãƒ€ã‚¯ã‚·ãƒ§ãƒ³</a></h2>
+
+<p>QDBMã¯ã€ã‚·ãƒ³ãƒ—ルãªãŒã‚‰ä¾¿åˆ©ãªãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ©ã‚¤ãƒ–ラリã§ã™ã€‚データベースã¨ã„ã†ã¨SQLやリレーショナルデータベースをæ€ã„æµ®ã‹ã¹ã‚‹äººãŒå¤šã„ã¨æ€ã„ã¾ã™ãŒã€QDBMã¯ãã‚“ãªé«˜æ©Ÿèƒ½ãªã‚‚ã®ã§ã¯ã‚ã‚Šã¾ã›ã‚“。「ã‚ーã€ã¨ã€Œå€¤ã€ã®çµ„ã‹ã‚‰ãªã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ファイルã«ä¿å˜ã—ãŸã‚Šã€ä¿å˜ã—ã¦ãŠã„ãŸãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸ã‹ã‚‰ç‰¹å®šã®ã‚ーをæŒã¤ã‚‚ã®ã‚’å–り出ã™æ©Ÿèƒ½ã‚’æä¾›ã™ã‚‹ã ã‘ã§ã™ã€‚ãã®ã‚ˆã†ãªæ©Ÿèƒ½ã‚’ã“ã“ã§ã¯ã€Œãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã€ã¨å‘¼ã¶ã“ã¨ã«ã—ã¾ã™ã€‚ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ç‰¹é•·ã¯ã€ä½¿ã„æ–¹ãŒç°¡å˜ã§ã€ãƒ‘フォーマンスãŒé«˜ã„ã“ã¨ã§ã™ã€‚</p>
+
+<p>QDBMã¯C言語ã®ãƒ©ã‚¤ãƒ–ラリã§ã™ï¼ˆä»–ã®è¨€èªžã®APIã‚‚ã‚ã‚Šã¾ã™ãŒï¼‰ã€‚QDBMã«ã¯ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ©Ÿèƒ½ã ã‘ã§ãªãã€ç§ï¼ˆQDBMã®ä½œè€…）ãŒãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ã‚’ã™ã‚‹æ™‚ã«ã‚ˆã使ã†æ©Ÿèƒ½ãŒè©°ã‚è¾¼ã¾ã‚Œã¦ã„ã¾ã™ã€‚Cã§æ›¸ã„ãŸãƒ—ãƒã‚°ãƒ©ãƒ ã¯é«˜é€Ÿã«å‹•ä½œã™ã‚‹ã®ãŒåˆ©ç‚¹ã§ã™ãŒã€C++ã€Javaã€Perlã€Rubyã¨ã„ã£ãŸæ¯”較的高級ãªè¨€èªžã§ã¯æ¨™æº–çš„ã«ã‚µãƒãƒ¼ãƒˆã•ã‚Œã‚‹ãƒ‡ãƒ¼ã‚¿æ§‹é€ やアルゴリズムを自分ã§å®Ÿè£…ã—ãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“。ãã†ã„ã£ãŸä½œæ¥ã¯é¢å€’ã§ã™ã—ã€ãƒã‚°ã‚’生ã¿ã‚„ã™ã„ã‚‚ã®ã§ã™ã€‚ãã“ã§ã€QDBMã®ç™»å ´ã§ã™ã€‚QDBMã‚’å†åˆ©ç”¨ã™ã‚Œã°ã€Cã§ã®ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ãŒPerlを使ã£ã¦ã„ã‚‹ã‹ã®ã‚ˆã†ã«æ‰‹è»½ã«ãªã‚Šã¾ã™ã€‚ã—ã‹ã‚‚ã€UNIXã§ã‚‚Windowsã§ã‚‚Mac OS Xã§ã‚‚利用ã§ãã‚‹ã®ã§ã€ç§»æ¤æ€§ã®ã‚るプãƒã‚°ãƒ©ãƒ ãŒæ›¸ãã‚„ã™ããªã‚Šã¾ã™ã€‚</p>
+
+<p>シンプルã¨ã„ã„ãªãŒã‚‰ã€QDBMã®æ©Ÿèƒ½ã¯ãªã‹ãªã‹è±Šå¯Œã§ã™ã€‚データベースã¨ã—ã¦ã¯ã€ãƒãƒƒã‚·ãƒ¥è¡¨ã¨B+木ãŒåˆ©ç”¨ã§ãã¾ã™ã€‚メモリ上ã§æ‰±ã†ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã¨ã—ã¦ã¯ã€ãƒªã‚¹ãƒˆã‚„マップãªã©ãŒã‚ã‚Šã¾ã™ã€‚MIMEã‚„CSVã‚„XMLã®è§£æžã‚‚ã§ãã¾ã™ã€‚ã—ã¾ã„ã«ã¯å…¨æ–‡æ¤œç´¢ã¾ã§ã§ããŸã‚Šã™ã‚‹ã®ã§é©šãã§ã™ã€‚</p>
+
+<p>ã“ã®ãƒãƒ¥ãƒ¼ãƒˆãƒªã‚¢ãƒ«ã§ã¯QDBMã®ä½¿ã„方を簡å˜ã«èª¬æ˜Žã™ã‚‹ã¨ã¨ã‚‚ã«ã€åŸºæœ¬ä»•æ§˜æ›¸ã®è£œè¶³ã‚’è¿°ã¹ã¾ã™ã€‚QDBMã®è©³ç´°ã«ã¤ã„ã¦ã¯åŸºæœ¬ä»•æ§˜æ›¸ã‚’御覧ãã ã•ã„。ãªãŠã€ã“ã“ã§ã¯ãƒãƒƒã‚·ãƒ¥è¡¨ã‚„B+木ã¨ã„ã£ãŸãƒ‡ãƒ¼ã‚¿æ§‹é€ ã®èª¬æ˜Žã¯ã—ã¾ã›ã‚“ã®ã§ã€ä¸æ…£ã‚Œãªæ–¹ã¯é©å½“ãªæœ¬ã‚„Webサイトã§èª¿ã¹ã¦ãŠã„ã¦ãã ã•ã„。</p>
+
+<hr />
+
+<h2><a name="depot" id="depot" class="head">Depot: 基本API</a></h2>
+
+<p>データベースアプリケーションã®å…¸åž‹çš„ãªä¾‹ã¨ã—ã¦ã€ã€Œç¤¾å“¡ç•ªå·ã‚’入力ã™ã‚‹ã¨ã€ãã®å†…線番å·ãŒã‚ã‹ã‚‹ã€ã¨ã„ã†ãƒ—ãƒã‚°ãƒ©ãƒ を考ãˆã¦ã¿ã¾ã—ょã†ã€‚社員番å·ã‚’ã‚ーã¨ã—ã¦ã€ãã‚Œã«å¯¾å¿œã™ã‚‹å€¤ã§ã‚る内線番å·ã‚’検索ã™ã‚‹ã¨ã„ã†ã“ã¨ã§ã™ã€‚</p>
+
+<p>ã¾ãšã¯QDBMを使ã‚ãªã„ã§ã‚„ã£ã¦ã¿ã¾ã™ã€‚社員番å·ã¨å†…線番å·ã®å¯¾å¿œè¡¨ã¯ã€CSVテã‚ストã§ãƒ•ã‚¡ã‚¤ãƒ«ã«ä¿å˜ã™ã‚‹ã“ã¨ã«ã—ã¾ã™ã€‚書å¼ã®ä¾‹ã‚’以下ã«ç¤ºã—ã¾ã™ã€‚</p>
+
+<pre>00001,8-902-1234
+00002,7-938-834
+00008,4-214-491
+</pre>
+
+<p>ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’åŠ ãˆã‚‹é–¢æ•° `putphonenumber' ã¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹é–¢æ•° `getphonenumber' を実装ã—ã¾ã™ã€‚</p>
+
+<pre>#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#define PHONEFILE "phone"
+#define LINEBUFSIZ 256
+
+int putphonenumber(const char *id, const char *phone){
+ FILE *OUT;
+ /* ファイルを追記モードã§é–‹ã */
+ if(!(OUT = fopen(PHONEFILE, "a"))) return -1;
+ /* レコードを書ã込む */
+ fprintf(OUT, "%s,%s\n", id, phone);
+ /* ファイルを閉ã˜ã‚‹ */
+ if(fclose(OUT) != 0) return -1;
+ return 0;
+}
+
+char *getphonenumber(const char *id){
+ FILE *IN;
+ char line[LINEBUFSIZ], *pivot, *phone;
+ int len;
+ /* ファイルをèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã */
+ if(!(IN = fopen(PHONEFILE, "r"))) return NULL;
+ /* å„行をèªã¿è¾¼ã‚€ */
+ while(fscanf(IN, "%s", line) == 1){
+ /* 区切り文å—を処ç†ã™ã‚‹ */
+ if(!(pivot = strchr(line, ','))) continue;
+ *pivot = '\0';
+ pivot++;
+ /* ã‚ーã®ä¸€è‡´åˆ¤å®š */
+ if(strcmp(line, id) == 0){
+ /* ファイルを閉ã˜ã‚‹ */
+ if(fclose(IN) != 0) return NULL;
+ /* メモリを確ä¿ã—ã¦æˆ»ã‚Šå€¤ã‚’生æˆã™ã‚‹ */
+ len = strlen(pivot);
+ if(!(phone = malloc(len + 1))) return NULL;
+ memcpy(phone, pivot, len + 1);
+ return phone;
+ }
+ }
+ /* ファイルを閉ã˜ã‚‹ */
+ fclose(IN);
+ return NULL;
+}
+</pre>
+
+<p>`fscanf' を使ã£ã¦ã„る時点ã§ã‹ãªã‚Šè²§å¼±ã§ã™ãŒã€ãã¡ã‚“ã¨æ›¸ã“ã†ã¨ã™ã‚‹ã¨éžå¸¸ã«é•·ããªã‚‹ã®ã§å¦¥å”ã—ã¾ã—ãŸï¼ˆã¡ãªã¿ã«ã€255æ–‡å—を越ãˆã‚‹è¡ŒãŒã‚ã£ãŸã‚‰æš´èµ°ã—ã¾ã™ï¼‰ã€‚ã¨ã«ã‹ãã€ã“ã®ç¨‹åº¦ã®å‡¦ç†ã§ã‚„ãŸã‚‰é•·ã„コードを書ã‹ãã°ãªã‚‰ãªã„ã®ã§ã¯æ‚²ã—ããªã‚Šã¾ã™ã€‚ã•ã‚‰ã«é‡å¤§ãªæ¬ 点ã¯ã€æ¤œç´¢ã®å‡¦ç†ãŒé…ã„ã¨ã„ã†ã“ã¨ã§ã™ã€‚ファイルã®æœ€åˆã‹ã‚‰æœ€å¾Œã¾ã§ï¼ˆå¹³å‡çš„ã«ã¯åŠåˆ†ã¾ã§ï¼‰èªã¾ãªã‘ã‚Œã°ãªã‚‰ãªã„ã‹ã‚‰ã§ã™ã€‚æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ä¿®æ£ã™ã‚‹æ™‚ã«ã‚‚ã‹ãªã‚Šé¢å€’ãªã“ã¨ã‚’ã—ãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“。</p>
+
+<p>QDBMを使ãˆã°ã‚‚ã£ã¨ã‚¨ãƒ¬ã‚¬ãƒ³ãƒˆãªã‚³ãƒ¼ãƒ‰ãŒæ›¸ã‘ã¾ã™ã€‚上記ã¨åŒã˜æ©Ÿèƒ½ã®é–¢æ•°ã‚’実装ã—ã¦ã¿ã¾ã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <stdlib.h>
+
+#define PHONEFILE "phone"
+
+int putphonenumber(const char *id, const char *phone){
+ DEPOT *depot;
+ /* データベースを追記モードã§é–‹ã */
+ if(!(depot = dpopen(PHONEFILE, DP_OWRITER | DP_OCREAT, -1))) return -1;
+ /* レコードを書ã込む */
+ dpput(depot, id, -1, phone, -1, DP_DOVER);
+ /* データベースを閉ã˜ã‚‹ */
+ if(!dpclose(depot)) return -1;
+ return 0;
+}
+
+char *getphonenumber(const char *id){
+ DEPOT *depot;
+ char *phone;
+ /* データベースをèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã */
+ if(!(depot = dpopen(PHONEFILE, DP_OREADER, -1))) return NULL;
+ /* レコードを探索ã—ã¦æˆ»ã‚Šå€¤ã‚’生æˆã™ã‚‹ */
+ phone = dpget(depot, id, -1, 0, -1, NULL);
+ /* データベースを閉ã˜ã‚‹ */
+ dpclose(depot);
+ return phone;
+}
+</pre>
+
+<p>ã‚‚ã¯ã‚„ファイル形å¼ã¯CSVファイルã§ã¯ãªãã€åŒºåˆ‡ã‚Šæ–‡å—ãŒä½•ã§ã‚ã‚‹ã‹æ°—ã«ã™ã‚‹å¿…è¦ã¯ã‚ã‚Šã¾ã›ã‚“。プãƒã‚°ãƒ©ãƒžã¯ãƒ•ã‚¡ã‚¤ãƒ«å½¢å¼ãŒã©ã†ã§ã‚ã‚‹ã‹ãªã©è€ƒãˆãªãã¦ã‚‚よã„ã®ã§ã™ã€‚メモリã®ç¢ºä¿ãªã©ã‚‚QDBMã®å†…部ã§ã‚„ã£ã¦ãれるã®ã§ã€ãƒãƒƒãƒ•ã‚¡ã®ã‚µã‚¤ã‚ºã‚’æ°—ã«ã™ã‚‹å¿…è¦ã¯ã‚ã‚Šã¾ã›ã‚“（解放ã¯å¿…è¦ã§ã™ï¼‰ã€‚処ç†é€Ÿåº¦ã‚’æ°—ã«ã™ã‚‹å¿…è¦ã‚‚ã‚ã‚Šã¾ã›ã‚“。データベースãŒã©ã‚“ãªã«å¤§ããã¦ã‚‚ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®è¿½åŠ や削除や検索ãŒä¸€çž¬ã§ã§ãã¾ã™ã€‚ã“ã®ã‚ˆã†ã«ã€ãƒ—ãƒã‚°ãƒ©ãƒžã‚’データ管ç†ã®è‹¦æ‚©ã‹ã‚‰è§£æ”¾ã™ã‚‹ã®ãŒQDBMã®å½¹å‰²ã§ã™ã€‚</p>
+
+<p>上記ã®ä¾‹ã§ã¯QDBMã®åŸºæœ¬APIã§ã‚ã‚‹Depotを利用ã—ã¦ã„ã¾ã™ã€‚ã¾ãšã€`DEPOT' ã¨ã„ã†åž‹ãŒç™»å ´ã—ã¦ã„ã¾ã™ã€‚ã“ã‚Œã¯æ¨™æº–ライブラリ㮠`FILE' ã¨åŒæ§˜ã«ã€æ“作対象ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®æƒ…å ±ã‚’æ ¼ç´ã—ã¦ã„ã‚‹æ§‹é€ ä½“ã®åž‹ã§ã™ã€‚ã“ã®åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦å„種ã®é–¢æ•°ã«æ¸¡ã™ã“ã¨ã«ãªã‚Šã¾ã™ã€‚関数ã¨ã—ã¦ã¯ã€`dpopen'ã€`dpclose'ã€`dpput' ãŠã‚ˆã³ `dpget' ãŒç™»å ´ã—ã¦ã„ã¾ã™ã€‚ã“ã®å››ã¤ã®ä½¿ã„方を覚ãˆã‚Œã°QDBMã®åŠåˆ†ã¯ç†è§£ã—ãŸã‚ˆã†ãªã‚‚ã®ã§ã™ã€‚</p>
+
+<div><kbd>DEPOT *dpopen(const char *<var>name</var>, int <var>omode</var>, int <var>bnum</var>);</kbd></div>
+
+<p>`dpopen' ã¯ãã®åã®é€šã‚Šã€Depotã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã関数ã§ã™ã€‚ãã®çµæžœã¨ã—㦠`DEPOT' åž‹ã®æ§‹é€ 体ã¸ã®ãƒã‚¤ãƒ³ã‚¿ãŒè¿”ã•ã‚Œã¾ã™ã€‚第1引数ã«ã¯ãƒ‡ãƒ¼ã‚¿ã‚’æ ¼ç´ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«åを指定ã—ã¾ã™ã€‚ã“ã‚Œã¯ç›¸å¯¾ãƒ‘スã§æŒ‡å®šã—ã¦ã‚‚絶対パスã§æŒ‡å®šã—ã¦ã‚‚構ã„ã¾ã›ã‚“。第2引数ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã¾ã™ã€‚èªã¿è¾¼ã¿ã¨æ›¸ãè¾¼ã¿ã®ä¸¡æ–¹ã‚’ã™ã‚‹ãªã‚‰ `DP_OWRITER' を指定ã—ã¾ã™ã€‚ãŸã ã—ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒå˜åœ¨ã—ãªã„å ´åˆã«æ–°è¦ä½œæˆã™ã‚‹ãªã‚‰ã°åŒæ™‚ã« `DP_OCREAT' をビットORã¨ã—ã¦åŠ ãˆã‚‹å¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚ã“ã®äºŒã¤ã‚’指定ã™ã‚‹ã¨ã€`fopen' ã® `a+' モードã¨ã»ã¼åŒã˜æ„味ã«ãªã‚Šã¾ã™ã€‚ä»–ã« `DP_OTRUNC' ã¨ã„ã†ãƒ•ãƒ©ã‚°ãŒã‚ã‚‹ã®ã§ã™ãŒã€ãã‚Œã¯ãƒ•ã‚¡ã‚¤ãƒ«ã‚’切り詰ã‚ã‚‹ã“ã¨ã‚’指示ã—ã¾ã™ã€‚ãã‚Œã‚‚åŠ ãˆã¦ä¸‰ã¤ã‚’指定ã™ã‚‹ã¨ `fopen' ã® `w+' モードã¨ã»ã¼åŒã˜æ„味ã«ãªã‚Šã¾ã™ã€‚èªã¿è¾¼ã¿ã
ã‘ã‚’ã™ã‚‹å ´åˆã€`DP_OREADER' を指定ã—ã¾ã™ã€‚ã“れ㯠`fopen' ã® `r' モードã¨ã»ã¼åŒã˜æ„味ã§ã™ã€‚第3引数ã¯ãƒãƒƒã‚·ãƒ¥è¡¨ã®ãƒã‚±ãƒƒãƒˆæ•°ã‚’指定ã—ã¾ã™ã€‚ã¨ã‚Šã‚ãˆãšãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ã‚’æ„味ã™ã‚‹ `-1' を指定ã—ã¦ãŠã‘ã°ã„ã„ã§ã—ょã†ã€‚</p>
+
+<p>複数ã®ãƒ—ãƒã‚»ã‚¹ãŒåŒã˜ãƒ•ã‚¡ã‚¤ãƒ«ã‚’èªã¿æ›¸ãã™ã‚‹å ´åˆã€ã€Œãƒ¬ãƒ¼ã‚¹ã‚³ãƒ³ãƒ‡ã‚£ã‚·ãƒ§ãƒ³ã€ã¨ã„ã†å•é¡ŒãŒèµ·ã“ã‚Šã¾ã™ã€‚åŒæ™‚ã«ãƒ•ã‚¡ã‚¤ãƒ«ã«æ›¸ã込むã¨ã€å†…容ãŒæ··ã–ã£ã¦å¤‰ã«ãªã£ã¦ã—ã¾ã†å¯èƒ½æ€§ãŒã‚ã‚‹ã®ã§ã™ã€‚QDBMã§ã¯ãã‚Œã«å¯¾å‡¦ã™ã‚‹ãŸã‚ã«ã€Œãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã€ã‚’ã‹ã‘ã¾ã™ã€‚ã‚るプãƒã‚»ã‚¹ãŒãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’書ãè¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã„ã¦ã„ã‚‹å ´åˆã¯ã€ä»–ã®ãƒ—ãƒã‚»ã‚¹ãŒãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã“ã†ã¨ã—ã¦ã‚‚ブãƒãƒƒã‚¯ã•ã‚Œã‚‹ã®ã§ã™ã€‚処ç†ãŒå¤±æ•—ã™ã‚‹ã‚ã‘ã§ã¯ãªãã€æ—¢ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã„ã¦ã„るプãƒã‚»ã‚¹ãŒãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ã‚‹ã¾ã§å¾…ã£ã¦ãれるã®ã§ã™ã€‚ãªãŠã€èªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰åŒå£«ã§ã‚ã‚Œã°ãƒ¬ãƒ¼ã‚¹ã‚³ãƒ³ãƒ‡ã‚£ã‚·ãƒ§ãƒ³ã¯èµ·ã“らãªã„ã®ã§åŒæ™‚ã«ã‚¢ã‚¯ã‚»ã‚¹ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚</p>
+
+<div><kbd>int dpclose(DEPOT *<var>depot</var>);</kbd></div>
+
+<p>`dpclose' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ã‚‹é–¢æ•°ã§ã™ã€‚第1引数ã«ã¯ `dpopen' ã§é–‹ã„ãŸãƒãƒ³ãƒ‰ãƒ«ã‚’渡ã—ã¾ã™ã€‚é–‹ã„ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯å¿…ãšé–‰ã˜ã¦ãã ã•ã„。ãã†ã—ãªã„ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒå£Šã‚Œã¾ã™ï¼ˆèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã®å ´åˆã¯å£Šã‚Œã¾ã›ã‚“ãŒã€ãƒ¡ãƒ¢ãƒªãƒªãƒ¼ã‚¯ã«ãªã‚Šã¾ã™ï¼‰ã€‚</p>
+
+<p>ã¨ã“ã‚ã§ã€QDBMを使ã‚ãªã„例ã§ã¯ã€æ›¸ãè¾¼ã¿ã®éš›ã«å‘¼ã³å‡ºã™ `fprintf' ã®æˆ»ã‚Šå€¤ã‚’ãƒã‚§ãƒƒã‚¯ã—ã¦ã„ã¾ã›ã‚“。`fprintf' ãŒå¤±æ•—ã—ãŸå ´åˆã¯ `fclose' もエラーを返ã™ã¨è¦å®šã•ã‚Œã¦ã„ã‚‹ã‹ã‚‰ã§ã™ã€‚åŒæ§˜ã«ã€QDBMã§ã‚‚データベースã«ä¸€åº¦ã§ã‚‚致命的ãªã‚¨ãƒ©ãƒ¼ãŒèµ·ããŸå ´åˆã¯ `dpclose' ãŒã‚¨ãƒ©ãƒ¼ã‚’è¿”ã™ã®ã§ã€ã‚¨ãƒ©ãƒ¼ãƒã‚§ãƒƒã‚¯ã‚’簡略化ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ã®ã§ã™ã€‚</p>
+
+<div><kbd>int dpput(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></div>
+
+<p>`dpput' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹é–¢æ•°ã§ã™ã€‚第1引数ã«ã¯ `dpopen' ã§é–‹ã„ãŸãƒãƒ³ãƒ‰ãƒ«ã‚’渡ã—ã¾ã™ã€‚第2引数ã«ã¯ã€æ›¸ã込むレコードã®ã‚ーã®å†…容をä¿æŒã™ã‚‹é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã—ã¾ã™ã€‚第3引数ã«ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã—ã¾ã™ã€‚ãã‚ŒãŒè² æ•°ã®å ´åˆã¯ã€ç¬¬2引数を文å—列ã¨ã—ã¦æ‰±ã£ã¦ã€`strlen' ã®å€¤ã‚’サイズã¨ã—ã¦åˆ¤å®šã—ã¾ã™ï¼ˆæ–‡å—列ã¨ãƒã‚¤ãƒŠãƒªã®ä¸¡æ–¹ã‚’ç°¡å˜ã«æ‰±ãˆã‚‹ã‚ˆã†ã«ã™ã‚‹ãŸã‚ã§ã™ï¼‰ã€‚第4引数ã¨ç¬¬5引数ã¯ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã«é–¢ã—ã¦åŒæ§˜ã«ãƒã‚¤ãƒ³ã‚¿ã¨ã‚µã‚¤ã‚ºã‚’指定ã—ã¾ã™ã€‚第6引数ã¯æ›¸ãè¾¼ã¿ã®ãƒ¢ãƒ¼ãƒ‰ã§ã™ã€‚データベース内ã«ã¯åŒã˜ã‚ーをæŒã¤è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ãŒã§ããªã„ã®ã§ã€æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨åŒä¸€ã®ã‚ーをæŒã¤ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã—よã†ã¨ã—ãŸéš›ã«ã©ã†ã™ã‚‹ã‹ã‚’指示ã™ã‚‹å¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚`DP_DOVER' ã¨ã—ã
Ÿå ´åˆã¯ã€æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ–°ã—ã„レコードã§ä¸Šæ›¸ãã—ã¾ã™ã€‚`DP_DKEEP' ã¨ã—ãŸå ´åˆã¯ã€æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’優先ã—ã€ã‚¨ãƒ©ãƒ¼ãŒè¿”ã•ã‚Œã¾ã™ï¼ˆ`DP_EKEEP' ã¨ã„ã†ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒå¤–部変数 `dpecode' ã«è¨å®šã•ã‚Œã¾ã™ï¼‰ã€‚</p>
+
+<p>`dpput' ã®æ›¸ãè¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã«ã¯ã€`DP_DCAT' ã«ã‚ˆã‚‹ã€Œé€£çµãƒ¢ãƒ¼ãƒ‰ã€ã‚‚ã‚ã‚Šã¾ã™ã€‚ã“れを利用ã™ã‚‹ã“ã¨ã¯ã‚ã¾ã‚Šãªã„ã‹ã‚‚ã—ã‚Œã¾ã›ã‚“ãŒã€ä»–ã®DBMã«ã¯ãªã„特徴ãªã®ã§èª¬æ˜Žã—ã¾ã™ã€‚連çµãƒ¢ãƒ¼ãƒ‰ã¯ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã¨ã—ã¦é…列を入れる際ã«ä¾¿åˆ©ãªã®ã§ã™ã€‚例ãˆã°ã€[10,11,12] ã¨ã„ã†ä¸‰ã¤ã®æ•°å€¤ã‚’è¦ç´ ã«æŒã¤é…åˆ—ã‚’æ ¼ç´ã—ã¦ã„ã¦ã€ãれを [10,11,12,13] ã«ã—ãŸã„å ´åˆã‚’考ãˆã¦ã¿ã¾ã™ã€‚ä»–ã®DBMã§ã¯ã€ã¾ãšãã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã—ã¦ã€[10,11,12] ã‚’ç²å¾—ã—ã¦ã‹ã‚‰ã€ãれを [10,11,12,13] ã«åŠ å·¥ã—ã¦å€¤ã‚’生æˆã—ã€å…ƒã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ä¸Šæ›¸ãã§æ›¸ã込むã¨ã„ã£ãŸã“ã¨ã‚’ã—ãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“。QDBMã®å ´åˆã¯ã€é€£çµãƒ¢ãƒ¼ãƒ‰ã§ [13] ã‚’æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«æ›¸ã込むã ã‘ã§åŒã˜ã“ã¨ãŒã§ãã¾ã™ã€‚é…列ã®è¦ç´ æ•°ãŒå¤§ãã„å ´åˆã«ã¯ã“ã®é•ã„ã¯ãƒ‘フォーマンスã«å¤§ããªå½±éŸ¿ã‚’åŠã¼ã—ã¾ã™ã€‚</p>
+
+<div><kbd>char *dpget(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></div>
+
+<p>`dpget' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’検索ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–り出ã™é–¢æ•°ã§ã™ã€‚第1引数ã¯ä»–ã¨ä¸€ç·’ã§ã™ã。第2引数ã¨ç¬¬3引数㯠`dpput' ã¨åŒæ§˜ã«æ¤œç´¢ã‚ーã®ãƒã‚¤ãƒ³ã‚¿ã¨ã‚µã‚¤ã‚ºã‚’指定ã—ã¾ã™ã€‚第4引数ã¨ç¬¬5引数ã¯ã¨ã‚Šã‚ãˆãš `0' 㨠'-1' ã«ã™ã‚‹ã‚‚ã®ã ã¨æ€ã£ã¦ã„ãŸã ã„ã¦çµæ§‹ã§ã™ã€‚一応説明ã™ã‚‹ã¨ã€å–り出ã™é ˜åŸŸã®é–‹å§‹ã‚ªãƒ•ã‚»ãƒƒãƒˆã¨æœ€å¤§ã‚µã‚¤ã‚ºã‚’指定ã—ã¾ã™ã€‚`-1' ã¯ç„¡åˆ¶é™ã¨ã„ã†æ„味ã§ã™ã€‚例ãˆã°ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ãŒ `abcdef' ã®å ´åˆã«é–‹å§‹ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’ `1'ã€ã‚µã‚¤ã‚ºã‚’ `3' ã«ã—ãŸå ´åˆã€`bcd' ãŒå–り出ã•ã‚Œã¾ã™ã€‚戻り値ã¯å–り出ã—ãŸå€¤ã®å†…容を記録ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã™ã€‚ãã®é ˜åŸŸã¯ `malloc' ã§ãƒ’ープã«ç¢ºä¿ã•ã‚Œã¦ã„ã‚‹ã®ã§ã€ã‚¢ãƒ—リケーションã®è²¬ä»»ã§ `free' ã«æ¸¡ã—ã¦è§£æ”¾ã™ã‚‹å¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚戻り値ã®é ˜åŸŸã¯çµ‚端ã«ã‚¼ãƒã‚³ãƒ¼ãƒ‰ãŒä»˜åŠ ã•ã‚Œã¦ã„ã‚‹ã“ã¨ãŒä¿è¨¼ã•ã‚Œã¦ã
„ã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãã¾ã™ã€‚ãŸã ã—ã€ãƒã‚¤ãƒŠãƒªã‚’扱ã†å ´åˆã«ã¯æ˜Žç¤ºçš„ã«ã‚µã‚¤ã‚ºãŒçŸ¥ã‚ŠãŸã„ã§ã—ょã†ã‹ã‚‰ã€ãã®ç‚ºã«ç¬¬6引数ã«ã‚µã‚¤ã‚ºã‚’å—ã‘å–る変数ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ï¼ˆã“ã®ã‚µã‚¤ã‚ºã«ã¯çµ‚端ã®ã‚¼ãƒã‚³ãƒ¼ãƒ‰ã¯å‹˜å®šã•ã‚Œã¾ã›ã‚“）。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã«ã¯ `NULL' ãŒè¿”ã•ã‚Œã¾ã™ï¼ˆ`DP_ENOITEM' ã¨ã„ã†ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒå¤–部変数 `dpecode' ã«è¨å®šã•ã‚Œã¾ã™ï¼‰ã€‚</p>
+
+<p>Depotã®é–¢æ•°ã§ä»–ã«ã‚ˆã使ã†ã‚‚ã®ã¨ã—ã¦ã€`dpout'ã€`dpiterinit'ã€`dpiternext' ãŒã‚ã‚Šã¾ã™ã€‚ã“れらã«ã¤ã„ã¦ã‚‚説明ã—ã¦ãŠãã¾ã™ã€‚</p>
+
+<div><kbd>int dpout(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></div>
+
+<p>`dpout' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‹ã‚‰ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹é–¢æ•°ã§ã™ã€‚三ã¤ã®å¼•æ•°ã®æ‰±ã„㯠`dpget' ã®ã‚‚ã®ã¨åŒã˜ã§ã™ã€‚該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã—ã¾ã™ï¼ˆ`DP_ENOITEM' ã¨ã„ã†ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒå¤–部変数 `dpecode' ã«è¨å®šã•ã‚Œã¾ã™ï¼‰ã€‚</p>
+
+<p>`dpiterinit' 㨠`dpiternext' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ä¸ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’一ã¤ä¸€ã¤è¦‹ã¦ã„ãå ´åˆã«ä½¿ã„ã¾ã™ã€‚先程ã®å†…線番å·ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’表示ã™ã‚‹é–¢æ•°ã®ä¾‹ã‚’以下ã«ç¤ºã—ã¾ã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define PHONEFILE "phone"
+
+int printphonenumbers(void){
+ DEPOT *depot;
+ char *kbuf, *vbuf;
+ /* データベースをèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã */
+ if(!(depot = dpopen(PHONEFILE, DP_OREADER, -1))) return -1;
+ /* イテレータをåˆæœŸåŒ–ã™ã‚‹ */
+ dpiterinit(depot);
+ /* å„レコードã®ã‚ーをå–り出㙠*/
+ while((kbuf = dpiternext(depot, NULL)) != NULL){
+ /* å„レコードã®å€¤ã‚’å–り出㙠*/
+ if((vbuf = dpget(depot, kbuf, -1, 0, -1, NULL)) != NULL){
+ printf("%s: %s\n", kbuf, vbuf);
+ /* 値ã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ free(vbuf);
+ }
+ /* ã‚ーã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ free(kbuf);
+ }
+ /* データベースを閉ã˜ã‚‹ */
+ return dpclose(depot) ? 0 : -1;
+}
+</pre>
+
+<p>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’横æ–çš„ã«è¦‹ã¦ã„ãã“ã¨ã‚’「トラãƒãƒ¼ã‚µãƒ«ã‚¢ã‚¯ã‚»ã‚¹ã€ã¨å‘¼ã¶ã“ã¨ã«ã—ã¾ã™ã€‚トラãƒãƒ¼ã‚µãƒ«ã‚¢ã‚¯ã‚»ã‚¹ã®éš›ã«ã¯ã€ã€Œã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã€ã‚’利用ã—ã¾ã™ã€‚イテレータã¯é›†åˆã«å«ã¾ã‚Œã‚‹å€‹ã€…ã®è¦ç´ を処ç†ã™ã‚‹éš›ã«ã€ãã®è¦ç´ を一ã¤ãšã¤å–り出ã™æ©Ÿèƒ½ã§ã™ã€‚トラãƒãƒ¼ã‚µãƒ«ã‚¢ã‚¯ã‚»ã‚¹ã‚’ã¯ã˜ã‚ã‚‹å‰ã«ã¯ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’åˆæœŸåŒ–ã—ã¾ã™ã€‚ãã—ã¦ã€ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ãŒã€Œæ‰“ã¡æ¢ã‚ã€ã®åˆå›³ã‚’è¿”ã™ã¾ã§ç¹°ã‚Šè¿”ã—ã¦å‘¼ã³å‡ºã—ã¾ã™ã€‚</p>
+
+<div><kbd>int dpiterinit(DEPOT *<var>depot</var>);</kbd></div>
+
+<p>`dpiterinit' ã¯ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’åˆæœŸåŒ–ã™ã‚‹é–¢æ•°ã§ã™ã€‚特ã«èª¬æ˜Žã¯å¿…è¦ã‚ã‚Šã¾ã›ã‚“ã。</p>
+
+<div><kbd>char *dpiternext(DEPOT *<var>depot</var>, int *<var>sp</var>);</kbd></div>
+
+<p>`dpiternext' ã¯ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰æ¬¡ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーをå–り出ã™é–¢æ•°ã§ã™ã€‚第1引数ã¯ä»–ã¨åŒã˜ã§ã™ã€‚例ã«ã‚ˆã£ã¦æˆ»ã‚Šå€¤ã¯ã‚¼ãƒã‚³ãƒ¼ãƒ‰ãŒä»˜åŠ ã•ã‚ŒãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã™ã€‚é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã¦ã„ã‚‹ã®ã§ã€ã‚¢ãƒ—リケーションã®è²¬ä»»ã§ `free' ã—ã¦ãã ã•ã„。明示的ã«ã‚µã‚¤ã‚ºã‚’知りãŸã„å ´åˆã¯ç¬¬2引数ã«ãれをå—ã‘å–る変数ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã—ã¾ã™ã€‚ã‚‚ã†å–り出ã™ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªããªã£ãŸã‚‰ `NULL' ãŒè¿”ã•ã‚Œã¾ã™ï¼ˆ`DP_ENOITEM' ã¨ã„ã†ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒå¤–部変数 `dpecode' ã«è¨å®šã•ã‚Œã¾ã™ï¼‰ã€‚</p>
+
+<p>トラãƒãƒ¼ã‚µãƒ«ã‚¢ã‚¯ã‚»ã‚¹ã§å„ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¾¿ã‚‹é †ç•ªã«ã¤ã„ã¦ã¯è¦å®šã•ã‚Œã¦ã„ã¾ã›ã‚“ã®ã§ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é †åºã«ä¾å˜ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ã‚’ã—ã¦ã¯ã„ã‘ã¾ã›ã‚“（ãã®ã‚ˆã†ãªå ´åˆã¯B+木データベースを使ã„ã¾ã—ょã†ï¼‰ã€‚ã¾ãŸã€ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’ç¹°ã‚Šè¿”ã—ã¦ã„る途ä¸ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’上書ãã—ãŸå ´åˆã€æ—¢ã«èªã‚“ã ã¯ãšã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒã¾ãŸå–り出ã•ã‚Œã‚‹å¯èƒ½æ€§ãŒã‚ã‚Šã¾ã™ã€‚途ä¸ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã“ã¨ã«é–¢ã—ã¦ã¯å•é¡Œã‚ã‚Šã¾ã›ã‚“。</p>
+
+<p>ã“ã“ã¾ã§ã§Depotã®åŸºæœ¬çš„ãªä½¿ã„æ–¹ã¯èª¬æ˜Žã—終ãˆã¾ã—ãŸã€‚QDBMã®ãれ以外ã®APIã‚‚Depotã«ä¼¼ãŸã‚ˆã†ãªä½¿ã„方をã—ã¾ã™ã®ã§ã€åŸºæœ¬ä»•æ§˜æ›¸ã®ã‚µãƒ³ãƒ—ルコードを見れã°ç†è§£ã—ã¦ã‚‚らãˆã‚‹ã¨æ€ã„ã¾ã™ã€‚</p>
+
+<hr />
+
+<h2><a name="curia" id="curia" class="head">Curia: æ‹¡å¼µAPI</a></h2>
+
+<p>Curiaã¯ã€Depotã¨ã»ã¨ã‚“ã©åŒã˜æ©Ÿèƒ½ã¨ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã‚’å‚™ãˆã¾ã™ãŒã€ãƒ•ã‚¡ã‚¤ãƒ«ã§ãªãディレクトリã¨ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†APIã§ã™ã€‚ã‚„ãŸã‚‰ã¨å¤§é‡ã®ãƒ‡ãƒ¼ã‚¿ã‚’扱ã‚ãªã‘ã‚Œã°ãªã‚‰ãªã„å ´åˆã«ã¯CuriaãŒãŠè–¦ã‚ã§ã™ã€‚データをディレクトリã®ä¸ã®è¤‡æ•°ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«åˆ†æ•£ã—ã¦æ ¼ç´ã™ã‚‹ã®ã§ã€Depotよりも大ããªï¼ˆ2GB以上ã®ï¼‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†ã“ã¨ãŒã§ãã¾ã™ã€‚Depotã§ã¯ãƒãƒ³ãƒ‰ãƒ«ã« `DEPOT' ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’使ã„ã¾ã—ãŸãŒã€Curiaã§ã¯ `CURIA' ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’使ã„ã¾ã™ã€‚ãã—ã¦ã€Depotã§ã¯ `dp' ã§å§‹ã¾ã£ã¦ã„ãŸé–¢æ•°åãŒã€Curiaã§ã¯ `cr' ã§å§‹ã¾ã‚Šã¾ã™ã€‚åå‰ãŒé•ã†ã ã‘ã§ã€ä½¿ã„æ–¹ã¯å…¨ã一緒ã§ã™ã€‚ãŸã ã—ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã `cropen' ã¨ã„ã†é–¢æ•°ã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®åˆ†å‰²æ•°ã‚’指定ã™ã‚‹å¼•æ•°ãŒå¢—ãˆã¦ã„ã¾ã™ã€‚</p>
+
+<div><kbd>CURIA *cropen(const char *<var>name</var>, int <var>omode</var>, int <var>bnum</var>, int <var>dnum</var>);</kbd></div>
+
+<p>第1ã€ç¬¬2ã€ç¬¬3引数㯠`dpopen' ã®ã‚‚ã®ã¨å…¨ã一緒ã§ã™ã€‚第4引数ã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®åˆ†å‰²æ•°ã‚’指定ã—ã¾ã™ã€‚ディレクトリã®ä¸ã«ã€æŒ‡å®šã—ãŸæ•°ã ã‘ã®ãƒ•ã‚¡ã‚¤ãƒ«ãŒä½œã‚‰ã‚Œã¾ã™ã€‚ãªãŠã€ç¬¬3引数ã§æŒ‡å®šã—ãŸå€¤ã¯ãã®å„々ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒæŒã¤ãƒã‚±ãƒƒãƒˆæ•°ã«ãªã‚Šã¾ã™ã€‚戻り値ã¯Curiaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã¨ãªã‚Šã¾ã™ã€‚</p>
+
+<p>Curiaã«ã¯ã€Œãƒ©ãƒ¼ã‚¸ã‚ªãƒ–ジェクトã€ã‚’扱ã†æ©Ÿèƒ½ã‚‚ã‚ã‚Šã¾ã™ã€‚ラージオブジェクトã¨ã¯ã€å„レコードをファイルã¨ã—ã¦ç‹¬ç«‹ã•ã›ã¦ä¿å˜ã™ã‚‹ä»•çµ„ã¿ã§ã™ã€‚ラージオブジェクトã«ã™ã‚‹ã¨é€šå¸¸ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚ˆã‚Šå‡¦ç†é€Ÿåº¦ãŒè½ã¡ã¾ã™ãŒã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ã‚µã‚¤ã‚ºã‚’ç„¡é™ã«ï¼ˆãƒ‡ã‚£ã‚¹ã‚¯ãŒè¨±ã™é™ã‚Šï¼‰å¤§ããã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ãªãŠã€ãƒ©ãƒ¼ã‚¸ã‚ªãƒ–ジェクトã®ãƒˆãƒ©ãƒãƒ¼ã‚µãƒ«ã‚¢ã‚¯ã‚»ã‚¹ã¯ã‚µãƒãƒ¼ãƒˆã•ã‚Œã¾ã›ã‚“。</p>
+
+<hr />
+
+<h2><a name="relic" id="relic" class="head">Relic: NDBM互æ›API</a></h2>
+
+<p>Relicã¯ã€NDBMã®ã‚¢ãƒ—リケーションをã™ã¹ã‹ã‚‰ãQDBMã«ä¹—ã‚Šæ›ãˆã•ã›ã‚‹ã¨ã„ã†é‡Žæœ›ã®ä¸‹ã«ä½œã‚‰ã‚ŒãŸAPIã§ã™ã€‚パフォーマンスã¯ã‚ªãƒªã‚¸ãƒŠãƒ«ã®NDBMã®æ•°å€ã¯å‡ºã¾ã™ã€‚NDBMã®ã‚¢ãƒ—リケーションã¯ã‚ã¾ã‚Šè¦‹ãªã„ã§ã™ãŒã€Perlç‰ã®ã‚¹ã‚¯ãƒªãƒ—ト言語ãŒãã®ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã‚’å‚™ãˆã¦ã„ã¾ã™ã€‚ã¤ã¾ã‚ŠRelicã«ã‚ˆã£ã¦å„種ã®ã‚¹ã‚¯ãƒªãƒ—ト言語ã§QDBMãŒä½¿ãˆã‚‹ã“ã¨ãŒä¿è¨¼ã•ã‚Œã‚‹ã‚ã‘ã§ã™ã€‚</p>
+
+<p>ã‚ãªãŸã®ã‚¢ãƒ—リケーションã®ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ä¸ã§ `ndbm.h' をインクルードã—ã¦ã„る部分を `relic.h' ã«æ›¸ãæ›ãˆã€ãƒªãƒ³ã‚¯å¯¾è±¡ã‚’ `ndbm' ã‹ã‚‰ `qdbm' ã«æ›ãˆã¦å†ã‚³ãƒ³ãƒ‘イルã—ãã ã•ã„。ãã‚Œã ã‘ã§ã‚ãªãŸã®ã‚¢ãƒ—リケーションã¯QDBMã«ä¹—ã‚Šæ›ãˆã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ãªãŠã€æ–°ãŸã«ã‚¢ãƒ—リケーションを書ãéš›ã«ã¯ã€Relicã§ãªãDepotを利用ã™ã‚‹ã“ã¨ã‚’ãŠè–¦ã‚ã—ã¾ã™ã€‚</p>
+
+<hr />
+
+<h2><a name="hovel" id="hovel" class="head">Hovel: GDBM互æ›API</a></h2>
+
+<p>Hovelã¯ã€GDBMã®ã‚¢ãƒ—リケーションをã™ã¹ã‹ã‚‰ãQDBMã«ä¹—ã‚Šæ›ãˆã•ã›ã‚‹ã¨ã„ã†é‡Žæœ›ã®ä¸‹ã«ä½œã‚‰ã‚ŒãŸAPIã§ã™ã€‚パフォーマンスã¯ã‚ªãƒªã‚¸ãƒŠãƒ«ã®GDBMã®æ•°å€ã¯å‡ºã¾ã™ã€‚GDBMã®ã‚¢ãƒ—リケーションã¯å¸‚å ´ã«å¤šã見られã¾ã™ãŒã€ãれらをãŠä½¿ã„ã®å ´åˆã¯ã€ãœã²ã¨ã‚‚QDBMã«ç§»æ¤ã—ã¦ã‚ã’ã¦ãã ã•ã„。パフォーマンスãŒç›®ã«è¦‹ãˆã¦ã«æ”¹å–„ã•ã‚Œã‚‹ã“ã¨ã†ã‘ã‚ã„ã§ã™ã€‚</p>
+
+<p>ã‚ãªãŸã®ã‚¢ãƒ—リケーションã®ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ä¸ã§ `gdbm.h' をインクルードã—ã¦ã„る部分を `hovel.h' ã«æ›¸ãæ›ãˆã€ãƒªãƒ³ã‚¯å¯¾è±¡ã‚’ `gdbm' ã‹ã‚‰ `qdbm' ã«æ›ãˆã¦å†ã‚³ãƒ³ãƒ‘イルã—ãã ã•ã„。ãã‚Œã ã‘ã§ã‚ãªãŸã®ã‚¢ãƒ—リケーションã¯QDBMã«ä¹—ã‚Šæ›ãˆã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ãªãŠã€æ–°ãŸã«ã‚¢ãƒ—リケーションを書ãéš›ã«ã¯ã€Hovelã§ãªãDepotを利用ã™ã‚‹ã“ã¨ã‚’ãŠè–¦ã‚ã—ã¾ã™ã€‚</p>
+
+<p>通常ã€HovelãŒç”Ÿæˆã™ã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã¯Depotã®ã‚‚ã®ã¨å…¨ãåŒã˜ã‚‚ã®ã§ã™ã€‚ã—ã‹ã—ã€ã¡ã‚‡ã£ã¨ç´°å·¥ã™ã‚‹ã¨ãれをCuriaã«ã‚ˆã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã«å¤‰æ›´ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚データベースãƒãƒ³ãƒ‰ãƒ«ã‚’å–å¾—ã™ã‚‹é–¢æ•° `gdbm_open' ã‚’ã€`gdbm_open2' ã«æ›¸ãæ›ãˆã‚Œã°ã‚ˆã„ã®ã§ã™ã€‚å˜ä¸€ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’æ ¼ç´ã—ã¦ã„ã‚‹å ´åˆã¯ã€ãƒ•ã‚¡ã‚¤ãƒ«ã‚µã‚¤ã‚ºãŒ2GBã¾ã§ã¨ã„ã†åˆ¶é™ã«ã²ã£ã‹ã‹ã£ã¦ã—ã¾ã„ã¾ã™ãŒã€`gdbm_open2' を使ãˆã°ãれを乗り越ãˆã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚`gdbm_open' を呼ã³å‡ºã—ã¦ã„ã‚‹ã¨ã“ã‚以外ã¯å…¨ã変更ã™ã‚‹å¿…è¦ãŒãªã„ã¨ã„ã†ã®ãŒå¬‰ã—ã„ã¨ã“ã‚ã§ã™ã€‚</p>
+
+<div><kbd>GDBM_FILE gdbm_open2(char *<var>name</var>, int <var>read_write</var>, int <var>mode</var>, int <var>bnum</var>, int <var>dnum</var>, int <var>align</var>);</kbd></div>
+
+<p>第1引数ã¯ç”Ÿæˆã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã‹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã§ã™ã€‚第2引数ã¨ç¬¬3引数㯠`gdbm_open' ã®ç¬¬3引数ã¨ç¬¬4引数ã¨ã—ã¦æ¸¡ã™ã‚‚ã®ã¨ä¸€ç·’ã§ã™ã€‚第4引数ã¯ãƒã‚±ãƒƒãƒˆã®è¦ç´ æ•°ã§ã™ã€‚第5引数ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åˆ†å‰²æ•°ã§ã™ã€‚第6引数ã¯å„レコードã®ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã§ã™ã€‚戻り値㯠`gdbm_open' ã¨åŒã˜ãデータベースãƒãƒ³ãƒ‰ãƒ«ã§ã™ã€‚</p>
+
+<hr />
+
+<h2><a name="cabin" id="cabin" class="head">Cabin: ユーティリティAPI</a></h2>
+
+<p>Cabinã¯ã€ãƒ‡ãƒ¼ã‚¿ã®æ“作を簡å˜ã«è¡Œã†ãŸã‚ã®ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã‚’集ã‚ãŸAPIã§ã™ã€‚密ã‹ã«QDBMã®APIã®ä¸ã§æœ€ã‚‚充実ã—ã¦ã„ã¾ã™ã€‚特ã«ãƒªã‚¹ãƒˆã¨ãƒžãƒƒãƒ—ã«é–¢é€£ã™ã‚‹é–¢æ•°ãŒé‡å®ã—ã¾ã™ã€‚ä»–ã«ã‚‚ã€ãƒ•ã‚¡ã‚¤ãƒ«ã‚„ディレクトリをèªã‚“ã ã‚Šã€æ–‡å—列を分割ã—ãŸã‚Šã€CSVã‚„XMLを解æžã—ãŸã‚Šã€å„種ã®ç¬¦å·åŒ–ã¨å¾©å·ã‚‚ã§ãã¾ã™ã€‚ã“ã“ã§ã¯ãƒªã‚¹ãƒˆã¨ãƒžãƒƒãƒ—ã¨XMLã«ã¤ã„ã¦è©³ã—ã説明ã—ã¾ã™ã€‚</p>
+
+<p>リストã¨ã¯ã€é †åºã‚’æŒã£ãŸé›†åˆã®ã“ã¨ã§ã™ã€‚CabinãŒæ‰±ã†ãƒªã‚¹ãƒˆã«ã¯ä»»æ„ã®æ–‡å—列やãƒã‚¤ãƒŠãƒªã‚’è¦ç´ ã¨ã—ã¦åŠ ãˆã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚リストã®å…ˆé ã«å¯¾ã—ã¦è¦ç´ ã®è¿½åŠ ã¨å‰Šé™¤ãŒã§ãã‚‹ã¨ã¨ã‚‚ã«ã€ãƒªã‚¹ãƒˆã®æœ«å°¾ã«å¯¾ã—ã¦è¦ç´ ã®è¿½åŠ ã¨å‰Šé™¤ã‚’ã™ã‚‹ã“ã¨ã‚‚ã§ãã¾ã™ï¼ˆã¤ã¾ã‚Šãƒ‡ã‚¯ã§ã™ï¼‰ã€‚ã¾ãŸã€é…列を使ã£ã¦å®Ÿè£…ã•ã‚Œã¦ã„ã‚‹ã®ã§ã€ä»»æ„ã®é †ç•ªã®è¦ç´ ã®å€¤ã‚’高速ã«å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚以下ã®ä¾‹ã§ã¯ã€`first'ã€`second'ã€`third' ã¨ã„ã†æ–‡å—åˆ—ã‚’é †ã«æœ«å°¾ã‹ã‚‰è¿½åŠ ã—ãŸä¸Šã§ã€å…ˆé ã‹ã‚‰æœ«å°¾ã¾ã§è¦ç´ ã®å†…容を表示ã—ã¦ã„ã¾ã™ã€‚</p>
+
+<pre>#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+void listtest(void){
+ CBLIST *list;
+ int i;
+ /* リストを開ã */
+ list = cblistopen();
+ /* è¦ç´ を末尾ã‹ã‚‰è¿½åŠ ã™ã‚‹ */
+ cblistpush(list, "first", -1);
+ cblistpush(list, "second", -1);
+ cblistpush(list, "third", -1);
+ /* å…ˆé ã‹ã‚‰è¦ç´ ã®å†…容を表示ã™ã‚‹ */
+ for(i = 0; i < cblistnum(list); i++){
+ printf("%s\n", cblistval(list, i, NULL));
+ }
+ /* リストを閉ã˜ã‚‹ */
+ cblistclose(list);
+}
+</pre>
+
+<p>`CBLIST' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’リストã®ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„ã¾ã™ã€‚実際ã®ãƒãƒ³ãƒ‰ãƒ«ã¯ `cblistopen' を呼ã³å‡ºã—ã¦ç²å¾—ã—ã¾ã™ã€‚ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã¦ãƒ¡ãƒ¢ãƒªã‚’解放ã™ã‚‹ã«ã¯ã€`cblistclose' を呼ã³å‡ºã—ã¾ã™ã€‚`cblistpush' ã¯æœ«å°¾ã«è¦ç´ ã‚’è¿½åŠ ã—ã¾ã™ã€‚`cblistnum' ã¯ãƒªã‚¹ãƒˆã®è¦ç´ æ•°ã‚’è¿”ã—ã¾ã™ã€‚`cblistval' ã¯ãƒªã‚¹ãƒˆå†…ã®ç‰¹å®šã®ç•ªå·ï¼ˆã‚¼ãƒã‹ã‚‰ã¯ã˜ã¾ã‚‹ï¼‰ã®è¦ç´ ã‚’è¿”ã—ã¾ã™ã€‚リストæ“作ã®é–¢æ•°ã¯ãã®ä»–ã«ã‚‚ã„ãã¤ã‹ã‚ã‚Šã¾ã™ã€‚</p>
+
+<p>マップã¨ã¯ã€ã‚ーã¨å€¤ã‹ã‚‰ãªã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é›†åˆã§ã™ã€‚CabinãŒæ‰±ã†ãƒžãƒƒãƒ—ã«ã¯ä»»æ„ã®æ–‡å—列やãƒã‚¤ãƒŠãƒªã‚’ã‚ーや値ã«æŒã¤ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ã‚ーãŒå®Œå…¨ã«ä¸€è‡´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã—ã¦å€¤ã‚’å–り出ã™ã“ã¨ãŒã§ãã¾ã™ï¼ˆå®Ÿè£…ã¯ãƒãƒƒã‚·ãƒ¥è¡¨ã§ã™ï¼‰ã€‚マップ内ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å…ˆé ã‹ã‚‰ä¸€ã¤ãšã¤å–り出ã™ã“ã¨ã‚‚ã§ãã¾ã™ã€‚ãªãŠã€å„è¦ç´ ã¯æ ¼ç´ã—ãŸé †ç•ªã§ä¸¦ã‚“ã§ã„ã‚‹ã“ã¨ãŒä¿è¨¼ã•ã‚Œã¦ã„ã¾ã™ã€‚以下ã®ä¾‹ã§ã¯ã€ã‚ー `one' ã¨å€¤ `first'ã€ã‚ー `two' ã¨å€¤ `second'ã€ã‚ー `three' ã¨å€¤ `third' ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’é †ã«æ ¼ç´ã—ãŸä¸Šã§ã€ãã®å„々を検索ã—ã¦è¡¨ç¤ºã—ã¦ã„ã¾ã™ã€‚</p>
+
+<pre>#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+void maptest(void){
+ CBMAP *map;
+ /* マップを開ã */
+ map = cbmapopen();
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ */
+ cbmapput(map, "one", -1, "first", -1, 1);
+ cbmapput(map, "two", -1, "second", -1, 1);
+ cbmapput(map, "three", -1, "third", -1, 1);
+ /* レコードを検索ã—ã¦å†…容を表示ã™ã‚‹ */
+ printf("one: %s\n", cbmapget(map, "one", -1, NULL));
+ printf("two: %s\n", cbmapget(map, "two", -1, NULL));
+ printf("three: %s\n", cbmapget(map, "three", -1, NULL));
+ /* マップを閉ã˜ã‚‹ */
+ cbmapclose(map);
+}
+</pre>
+
+<p>`CBMAP' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’マップã®ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„ã¾ã™ã€‚実際ã®ãƒãƒ³ãƒ‰ãƒ«ã¯ `cbmapopen' を呼ã³å‡ºã—ã¦ç²å¾—ã—ã¾ã™ã€‚ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã¦ãƒ¡ãƒ¢ãƒªã‚’解放ã™ã‚‹ã«ã¯ã€`cbmapclose' を呼ã³å‡ºã—ã¾ã™ã€‚`cbmapput' ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã—ã¾ã™ã€‚`cbmapget' ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã—ã¾ã™ã€‚マップæ“作ã®é–¢æ•°ã¯ãã®ä»–ã«ã‚‚ã„ãã¤ã‹ã‚ã‚Šã¾ã™ã€‚</p>
+
+<p>XMLã‚’ç°¡å˜ã«å‡¦ç†ã™ã‚‹ãŸã‚ã«ã€ç°¡æ˜“çš„ãªãƒ‘ーザãŒç”¨æ„ã•ã‚Œã¦ã„ã¾ã™ã€‚ã“ã®ãƒ‘ーザã¯å¦¥å½“性検証をã›ãšã€æ›¸å¼ã®æ¤œæŸ»ã‚‚厳密ã§ãªã„ã®ãŒç‰¹å¾´ã§ã™ã€‚ã—ãŸãŒã£ã¦ã€ä¸€èˆ¬çš„ãªHTMLã‚„SGMLã®è§£æžã«ã‚‚用ã„ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚å˜ç´”ãªæ§‹é€ ã®XML文書を処ç†ã™ã‚‹éš›ã«ã¯ã€DOMã‚„SAXã¨ã„ã£ãŸAPIを使ã†ã‚ˆã‚Šã‚‚便利ã§ã™ã€‚以下ã®ä¾‹ã§ã¯ã€XML文書ã®ä¸ã‹ã‚‰ `id' 属性ã®å€¤ã§è¦ç´ を指定ã—ã¦ã€ãã®ãƒ†ã‚ストをå–り出ã—ã¦è¡¨ç¤ºã—ã¾ã™ã€‚</p>
+
+<pre>#include <cabin.h>
+#include <stdlib.h>
+#include <string.h>
+#include <stdio.h>
+
+void showtextbyid(const char *xml, const char *id){
+ CBLIST *elems;
+ CBMAP *attrs;
+ const char *elem, *attr;
+ char *orig;
+ int i;
+ /* ã‚¿ã‚°ã¨ãƒ†ã‚ストã®ãƒªã‚¹ãƒˆã‚’å–å¾—ã™ã‚‹ */
+ elems = cbxmlbreak(xml, 1);
+ /* リストã®å„è¦ç´ ã‚’ãŸã©ã‚‹ */
+ for(i = 0; i < cblistnum(elems); i++){
+ /* è¦ç´ ã‚’å–り出㙠*/
+ elem = cblistval(elems, i, NULL);
+ /* ã‚¿ã‚°ã§ãªã„å ´åˆã¯èªã¿é£›ã°ã™ */
+ if(elem[0] != '<' || elem[1] == '?' || elem[1] == '!' || elem[1] == '/') continue;
+ /* 属性ã®ãƒžãƒƒãƒ—ã‚’å–å¾—ã™ã‚‹ */
+ attrs = cbxmlattrs(elem);
+ /* IDè¦ç´ ã®å€¤ã‚’å–り出ã—ã€ä¸€è‡´ã‚’検査ã™ã‚‹ */
+ attr = cbmapget(attrs, "id", -1, NULL);
+ if(attr && !strcmp(attr, id)){
+ /* 次ã®è¦ç´ ã‚’å–り出㙠*/
+ elem = cblistval(elems, i + 1, NULL);
+ if(elem){
+ /* 実体å‚照を復元ã—ã¦è¡¨ç¤ºã™ã‚‹ */
+ orig = cbxmlunescape(elem);
+ printf("%s\n", orig);
+ free(orig);
+ }
+ }
+ /* 属性マップを閉ã˜ã‚‹ */
+ cbmapclose(attrs);
+ }
+ /* è¦ç´ リストを閉ã˜ã‚‹ */
+ cblistclose(elems);
+}
+</pre>
+
+<p>処ç†å¯¾è±¡ã®XML文書ã®ãƒ†ã‚ストを `cbxmlbreak' ã§åˆ†è§£ã—ã¾ã™ã€‚例ãˆã° `<body><p id="nuts">NUTS&MILK</p></body>' を分解ã™ã‚‹ã¨ã€`<body>'ã€`<p id="nuts">'ã€`NUTS&MILK'ã€`</p>'ã€`</body>' ãŒå¾—られã¾ã™ã€‚ãã—ã¦ã€å„è¦ç´ を巡回ã—ã¾ã™ã€‚1æ–‡å—目㌠'<' ã§ã‚ã‚Œã°ã‚¿ã‚°ã‹å„種ã®å®£è¨€ã§ã‚ã‚Šã€ã‹ã¤2æ–‡å—目㌠'?'ã€`!'ã€`/' ã®ã„ãšã‚Œã§ã‚‚ãªã‘ã‚Œã°é–‹å§‹ã‚¿ã‚°ã¾ãŸã¯ç©ºã‚¿ã‚°ã§ã‚ã‚‹ã¨åˆ¤æ–ã§ãã¾ã™ã€‚ã‚¿ã‚°ã«å¯¾ã—㦠`cbxmlattrs' を呼ã¶ã“ã¨ã§å±žæ€§ã®ãƒžãƒƒãƒ—ãŒå¾—られã¾ã™ã€‚ã“ã®ãƒžãƒƒãƒ—ã¯å±žæ€§åã‚’ã‚ーã«ã—ã¦å€¤ã‚’å–り出ã™ã“ã¨ãŒã§ãã¾ã™ã€‚属性値やテã‚ストセクションã®æ–‡å—列ã¯æ–‡æ›¸å†…ã«å‡ºç¾ã—ãŸã¾ã¾ã®å½¢å¼ã«ãªã£ã¦ã„ã¾ã™ã€‚実体å‚照をå«ã‚“ã æ–‡å—列を復元ã™ã‚‹ã«ã¯ `cbxmlunescape' を用ã„ã¾ã™ã€‚</p>
+
+<p>GTK+ã«ä»˜å±žã™ã‚‹GLibã‚„Apacheã«ä»˜å±žã™ã‚‹APRãªã©ã®ä¾¿åˆ©ãªãƒ©ã‚¤ãƒ–ラリãŒä¸–ã®ä¸ã«ã¯ã‚ã‚Šã¾ã™ã®ã§ã€å˜ä½“ã§Cabinを利用ã™ã‚‹ä¾¡å€¤ã¯ã‚ã¾ã‚Šã‚ã‚Šã¾ã›ã‚“。æ£ç›´è¨€ã£ã¦ã€GLibã‚„APRã®æ–¹ãŒé«˜æ©Ÿèƒ½ã§ã€ãƒ¦ãƒ¼ã‚¶æ•°ã‚‚多ãã€å‚考ã«ãªã‚‹æƒ…å ±ã‚‚å¤šã„ã§ã™ã€‚ã¨ã¯ã„ãˆã€Cabinã®æ–¹ãŒæ‰‹è»½ã«ä½¿ãˆã‚‹ã®ã§ç§ã¯å¥½ãã§ã™ã€‚</p>
+
+<hr />
+
+<h2><a name="villa" id="villa" class="head">Villa: 上級API</a></h2>
+
+<p>Villaã¯ã€B+木ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†APIã§ã™ã€‚B+木データベースã«ã¯ãƒ¦ãƒ¼ã‚¶ãŒæŒ‡å®šã—ãŸé †åºã§ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒä¸¦ã¹ã‚‰ã‚Œã¦æ ¼ç´ã•ã‚Œã¾ã™ã€‚Depotã‚„Curiaã¯ã‚ーã®å®Œå…¨ä¸€è‡´ã«ã‚ˆã‚‹æ¤œç´¢ã—ã‹ã§ãã¾ã›ã‚“ãŒã€Villaを用ã„ã‚‹ã¨ç¯„囲を指定ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹ã“ã¨ãŒã§ãã¾ã™ã€‚ã¾ãŸã€åŒã˜ã‚ーをæŒã¤è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ã‚‚ã§ãã¾ã™ã€‚例ãˆã°ã€ã‚ーを文å—列ã®è¾žæ›¸é †ï¼ˆABCé †ã¨ã‹ã‚¢ã‚¤ã‚¦ã‚¨ã‚ªé †ã¨åŒã˜ã»ã¼æ„味ã§ã™ï¼‰ã§ä¸¦ã¹ã‚‹ã‚ˆã†ã«æŒ‡å®šã—ãŸå ´åˆã¯ã€æ–‡å—列ã®å‰æ–¹ä¸€è‡´æ¤œç´¢ãŒã§ãã‚‹ã®ã§ã™ã€‚</p>
+
+<p>ã¨ã¯ã„ãˆã€åŸºæœ¬çš„ãªä½¿ã„æ–¹ã¯Depotã¨ä¸€ç·’ã§ã™ã€‚Depotã®èª¬æ˜Žã§æŒ™ã’ãŸé–¢æ•°ã‚’Villaを使ã£ã¦å®Ÿè£…ã—ãªãŠã—ã¦ã¿ã¾ã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+
+#define PHONEFILE "phone"
+
+int putphonenumber(const char *id, const char *phone){
+ VILLA *villa;
+ /* データベースを追記モードã§é–‹ã */
+ if(!(villa = vlopen(PHONEFILE, VL_OWRITER | VL_OCREAT, VL_CMPLEX))) return -1;
+ /* レコードを書ã込む */
+ vlput(villa, id, -1, phone, -1, VL_DOVER);
+ /* データベースを閉ã˜ã‚‹ */
+ if(!vlclose(villa)) return -1;
+ return 0;
+}
+
+char *getphonenumber(const char *id){
+ VILLA *villa;
+ char *phone;
+ /* データベースをèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã */
+ if(!(villa = vlopen(PHONEFILE, VL_OREADER, VL_CMPLEX))) return NULL;
+ /* レコードを検索ã—ã¦æˆ»ã‚Šå€¤ã‚’生æˆã™ã‚‹ */
+ phone = vlget(villa, id, -1, NULL);
+ /* データベースを閉ã˜ã‚‹ */
+ vlclose(villa);
+ return phone;
+}
+</pre>
+
+<p>`VILLA' ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã¯ä¾‹ã«ã‚ˆã£ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã™ã€‚`vlopen' ã§ãã®ãƒãƒ³ãƒ‰ãƒ«ã‚’ç²å¾—ã—ã¾ã™ã€‚ãã®ç¬¬3引数㮠`VL_CMPLEX' ã¯è¾žæ›¸é †ã®æ¯”較を行ã†é–¢æ•°ã§ã™ã€‚é–‹ã„ãŸãƒãƒ³ãƒ‰ãƒ«ã¯ `vlclose' ã§é–‰ã˜ã¾ã™ã€‚`vlput' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹é–¢æ•°ã§ã€`vlget' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹é–¢æ•°ã§ã™ã€‚</p>
+
+<p>上記ã®ä¾‹ã§ã¯ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¨å…¨ãåŒã˜ä½¿ã„方をã—ã¾ã—ãŸãŒã€é †ç•ªã«åŸºã¥ã„ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ã‚¢ã‚¯ã‚»ã‚¹ã™ã‚‹æ©Ÿèƒ½ãŒB+木データベースã®ç‰¹å¾´ã§ã™ã€‚è¾žæ›¸é †ã‚’ä¾‹ã«ã¨ã£ã¦èª¬æ˜Žã—ã¾ã™ã€‚ã‚ーãŒãã‚Œãžã‚Œ `one'ã€`two'ã€`three'ã€`four' ã¨ã„ã†ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã—ãŸã¨ã™ã‚Œã°ã€ãれ㯠`four'ã€`one'ã€`three'ã€`two' ã¨ã„ã†é †ç•ªã§ä¸¦ã¹ã‚‰ã‚Œã¦ä¿å˜ã•ã‚Œã¾ã™ã€‚検索ã«ã¯ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¨åŒæ§˜ã«å®Œå…¨ä¸€è‡´æ¡ä»¶ã‚‚使ãˆã¾ã™ã€‚ã•ã‚‰ã«ã€ã€Œã‚«ãƒ¼ã‚½ãƒ«ã€ã¨ã„ã†æ©Ÿèƒ½ã‚’使ã£ã¦ç¯„囲を指定ã—ãŸæ¤œç´¢ãŒã§ãã¾ã™ã€‚カーソルã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä½ç½®ã‚’指ã—示ã—ã¾ã™ã€‚例ãˆã°ã€`one' ã®å ´æ‰€ã«ã‚«ãƒ¼ã‚½ãƒ«ã‚’飛ã°ã™ã¨ã„ã£ãŸæŒ‡å®šãŒã§ãã¾ã™ã€‚`one' ãŒãªã„å ´åˆã¯ã€`one' ã®ç›´å¾Œã® `three' ã®ä½ç½®ã«ã‚«ãƒ¼ã‚½ãƒ«ãŒé£›ã¶ã“ã¨ã«ãªã‚Šã¾ã™ï¼ˆ`one' ãŒè¤‡æ•°ã‚ã£ãŸå ´åˆã¯ã€ãã®æœ€åˆã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«é£›ã³ã¾ã™ï¼‰ã€‚ã‚«ã
ƒ¼ã‚½ãƒ«ã¯ã€ç¾åœ¨ä½ç½®ã‹ã‚‰å‰ã«é€²ã‚ãŸã‚Šå¾Œã‚ã«æˆ»ã—ãŸã‚Šã™ã‚‹ã“ã¨ã‚‚ã§ãã¾ã™ã€‚ãã—ã¦ã€ã‚«ãƒ¼ã‚½ãƒ«ã®ä½ç½®ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å†…容をèªã¿å‡ºã›ã°ã€ç¯„囲を指定ã—ãŸæ¤œç´¢ãŒã§ãã‚‹ã¨ã„ã†ã‚ã‘ã§ã™ã€‚以下ã®ä¾‹ã¯ã€`one' ã‹ã‚‰ `three' ã¾ã§ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨å€¤ã‚’表示ã™ã‚‹é–¢æ•°ã§ã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define WORDFILE "word"
+#define TOPWORD "one"
+#define BOTTOMWORD "three"
+
+void printwords(void){
+ VILLA *villa;
+ char *kbuf, *vbuf;
+ int ksiz;
+ /* データベースをèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã */
+ if(!(villa = vlopen(WORDFILE, VL_OREADER, VL_CMPLEX))) return;
+ /* カーソルを候補ã®å…ˆé ã«é£›ã°ã™ */
+ vlcurjump(villa, TOPWORD, -1, VL_JFORWARD);
+ /* å„候補を処ç†ã™ã‚‹ */
+ do {
+ /* レコードã®ã‚ーをå–り出㙠*/
+ kbuf = vlcurkey(villa, &ksiz);
+ /* 候補ãŒç¯„囲外ã§ã‚ã‚Œã°æŠœã‘ã‚‹ */
+ if(!kbuf || VL_CMPLEX(kbuf, ksiz, BOTTOMWORD, sizeof(BOTTOMWORD) - 1) > 0){
+ free(kbuf);
+ break;
+ }
+ /* レコードã®å€¤ã‚’å–り出ã—ã¦è¡¨ç¤ºã™ã‚‹ */
+ vbuf = vlcurval(villa, NULL);
+ if(kbuf && vbuf) printf("%s: %s\n", kbuf, vbuf);
+ /* ã‚ーã¨å€¤ã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ free(vbuf);
+ free(kbuf);
+ /* カーソルを次ã«é€²ã‚ã‚‹ */
+ } while(vlcurnext(villa));
+ /* データベースを閉ã˜ã‚‹ */
+ vlclose(villa);
+}
+</pre>
+
+<p>`vlcurjump' ã§ã‚«ãƒ¼ã‚½ãƒ«ã‚’候補ã®å…ˆé ã«é£›ã°ã—ã¦ã„ã¾ã™ã€‚`VL_JFORWARD' ã¯ã“ã‚Œã‹ã‚‰ã‚«ãƒ¼ã‚½ãƒ«ã‚’å‰ã«é€²ã‚ã¦ã„ãå ´åˆã«æŒ‡å®šã—ã¾ã™ï¼ˆå€™è£œã®æœ«å°¾ã«é£›ã°ã—ã¦ã‹ã‚‰å¾Œã‚ã«æˆ»ã—ã¦è¡Œãå ´åˆã¯ `VL_JBACKWARD' を指定ã—ã¾ã™ï¼‰ã€‚do-whileループã®æ¡ä»¶éƒ¨ã§ `vlcurnext' を呼んã§ã„ã¾ã™ãŒã€ã“ã‚ŒãŒã‚«ãƒ¼ã‚½ãƒ«ã‚’å‰ã«é€²ã‚ã¦ã„ã¾ã™ï¼ˆãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æœ€å¾Œã¾ã§æ¥ãŸã‚‰å½ã‚’è¿”ã™ã®ã§ãƒ«ãƒ¼ãƒ—ã‹ã‚‰æŠœã‘ã¾ã™ï¼‰ã€‚`vlcurkey' 㨠`vlcurval' ã¯ãã‚Œãžã‚Œã‚«ãƒ¼ã‚½ãƒ«ã®ã‚ーã¨å€¤ã‚’å–り出ã—ã¾ã™ã€‚`VL_CMPLEX' を明示的ã«å‘¼ã‚“ã§ã„ã‚‹å ´æ‰€ãŒã‚ã‚Šã¾ã™ãŒã€ã“ã“ã§å€™è£œã®æœ«å°¾ã«æ¥ãŸã‹ã©ã†ã‹åˆ¤å®šã—ã¦ã„ã¾ã™ã€‚比較関数ã¯ã€äºŒã¤ã®ã‚ーã®ãƒã‚¤ãƒ³ã‚¿ã¨ã‚µã‚¤ã‚ºã‚’渡ã—ã¦ã€å‰è€…ãŒå¤§ãã„（ã¤ã¾ã‚Šå¾Œã‚ã«ä½ç½®ã™ã¹ã）ãªã‚‰æ£ã®å€¤ã€å‰è€…ãŒå°ã•ã„（ã¤ã¾ã‚Šå‰ã«ä½ç½®ã™ã¹ã）ãªã‚‰è² ã®å€¤ã€ä¸¡è€…ãŒåŒã˜ãªã‚‰ã‚¼ãƒã‚’è¿”ã™ã¨è¦å®š
ã•ã‚Œã¦ã„ã¾ã™ã€‚ã—ãŸãŒã£ã¦ã€æ£ã®å€¤ãŒè¿”ã•ã‚ŒãŸå ´åˆã€ä»Šå–り出ã—ãŸã‚ーã¯å€™è£œã®æœ«å°¾ã‚ˆã‚Šã‚‚後ã‚ã ã¨ã„ã†ã“ã¨ã«ãªã‚Šã¾ã™ã€‚</p>
+
+<p>è¾žæ›¸é †ä»¥å¤–ã®æ¯”較関数も使ãˆã‚‹ã¨ã“ã‚ãŒVillaã®ãƒŸã‚½ã§ã™ã€‚intåž‹ã®æ•°å€¤ã‚’比較ã™ã‚‹ `VL_CMPINT' ã‚„ã€10進数ã®æ–‡å—列を比較ã™ã‚‹ `VL_CMPDEC' ã¨ã„ã£ãŸé–¢æ•°ãŒæœ€åˆã‹ã‚‰ç”¨æ„ã•ã‚Œã¦ã„ã¾ã™ã€‚ã•ã‚‰ã«ã€ã‚ãªãŸãŒè‡ªåˆ†ã§å®šç¾©ã—ãŸé–¢æ•°ã‚‚比較関数ã¨ã—ã¦ä½¿ã†ã“ã¨ãŒã§ãã¾ã™ã€‚使ã„æ–¹ãŒé¢å€’ã ã¨ã„ã†æ¬ 点を除ã‘ã°ã€B+木データベースã¯ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚ˆã‚Šã‚‚多ãã®ã‚·ãƒ¼ãƒ³ã§æ´»ç”¨ã§ãã‚‹ã¨æ€ã„ã¾ã™ã€‚ファイルãŒãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚ˆã‚Šã‚‚å°ã•ã‹ã£ãŸã‚Šã€ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ãŒä½¿ãˆã‚‹ã¨ã„ã£ãŸç‰¹å¾´ã‚‚ã€äººã«ã‚ˆã£ã¦ã¯å¬‰ã—ã„ã‹ã‚‚ã—ã‚Œã¾ã›ã‚“。</p>
+
+<p>Villaã¯ã€ã‚ュー（FIFO）を永続化ã™ã‚‹ç›®çš„ã§ã‚‚利用ã§ãã¾ã™ã€‚`VL_CMPINT' を比較関数ã«ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã¾ã™ã€‚å„レコードã®ã‚ーã¯intåž‹ã¨ã—ã€å€¤ã«ã¯ä»»æ„ã®ã‚ªãƒ–ジェクトを入れるã“ã¨ã«ã—ã¾ã™ã€‚ã‚ューã«è¦ç´ ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€æœ«å°¾ã®ã‚ーã®æ•°å€¤ï¼ˆãªã‘ã‚Œã°ã‚¼ãƒï¼‰ã‚’インクリメントã—ã¦ã‚ーを生æˆã—ã¦æ ¼ç´ã—ã¾ã™ã€‚ã‚ューã‹ã‚‰è¦ç´ ã‚’å–り出ã™ã«ã¯ã€å…ˆé ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–り出ã—ã¦ã‹ã‚‰å‰Šé™¤ã™ã‚Œã°ã‚ˆã„ã®ã§ã™ã€‚ãã®ã‚ˆã†ãªæ©Ÿèƒ½ã‚’æŒã¤ãƒ©ãƒƒãƒ‘関数をã€`qopen'ã€`qclose'ã€`qappend'ã€`qconsume' ã¨ã„ã£ãŸåå‰ã§ä½œã£ã¦ãŠãã¨å°ç²‹ã§ã™ã。</p>
+
+<hr />
+
+<h2><a name="odeum" id="odeum" class="head">Odeum: 転置API</a></h2>
+
+<p>Odeumã¯ã€å…¨æ–‡æ¤œç´¢ç”¨ã®è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’扱ã†APIã§ã™ã€‚テã‚ストファイル（ã¾ãŸã¯ãƒ†ã‚ストをå«ã‚€HTMLã‚„MS-Wordã®æ–‡æ›¸ãªã©ï¼‰ã¯å˜èªžã®é›†åˆã¨ã¿ãªã›ã¾ã™ãŒã€ã‚ã‚‹å˜èªžãŒã©ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«å«ã¾ã‚Œã‚‹ã‹ã¨ã„ã†æƒ…å ±ã‚’ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ã—ãŸã‚‚ã®ã‚’転置インデックスã¨å‘¼ã³ã¾ã™ã€‚本ã®å·»æœ«ã«ã‚る索引ã¯ã€ã‚ã‚‹å˜èªžãŒã©ã®ãƒšãƒ¼ã‚¸ã«å«ã¾ã‚Œã‚‹ã‹ã¨ã„ã†æƒ…å ±ã‚’æŒã£ã¦ã„ã¾ã™ãŒã€ãã‚Œã¨ä¼¼ãŸã‚ˆã†ãªã‚‚ã®ã§ã™ã€‚ç§ãŒQDBMを開発ã™ã‚‹å¥‘æ©Ÿã¨ãªã£ãŸã®ã¯ã€ã¨ã‚る全文検索システムã®é–‹ç™ºã§ä½¿ã£ã¦ã„ãŸGDBMã®ãƒ‘フォーマンスã«é™ç•Œã‚’æ„Ÿã˜ãŸã“ã¨ã§ã™ã€‚ãã®çµŒç·¯ã‹ã‚‰ã€QDBM（特ã«Curia）ã«ã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®å®Ÿç¾ã«éƒ½åˆã®ã‚ˆã„特徴ãŒã„ãã¤ã‹å‚™ã‚ã£ã¦ã„ã¾ã™ã€‚</p>
+
+<p>転置インデックスã®æ ¸ã¨ãªã‚‹ã®ã¯ã€ã‚ã‚‹å˜èªžã‚’ã‚ーã¨ã—ã€ãã®å˜èªžã‚’å«ã‚€ãƒ•ã‚¡ã‚¤ãƒ«ã®IDã®é…列を値ã¨ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‹ã‚‰ãªã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã™ã€‚å˜èªžã¯å®Œå…¨ä¸€è‡´ã§æ¤œç´¢ã§ãã‚Œã°ã‚ˆã„ã®ã§ã€ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã«ã¯ãƒãƒƒã‚·ãƒ¥è¡¨ã‚’採用ã—ã¦ã„ã¾ã™ã€‚転置インデックスã®ã‚¤ãƒ¡ãƒ¼ã‚¸ã‚’例示ã—ã¾ã™ã€‚ファイル `penguin.txt' ã«ã¯ã€Œflightless marine birdsã€ã¨ã„ã†ãƒ†ã‚ストãŒæ ¼ç´ã•ã‚Œã¦ã„ã¦ã€ãƒ•ã‚¡ã‚¤ãƒ« `ostrich.txt' ã«ã¯ã€ã€Œflightless birds in africaã€ã¨ã„ã†ãƒ†ã‚ストãŒæ ¼ç´ã•ã‚Œã¦ã„ã‚‹ã¨ã—ã¾ã™ã€‚å„ファイルã«ã¯ã€èªã¿è¾¼ã‚“ã é †ç•ªã§IDã‚’ã¤ã‘ã‚‹ã“ã¨ã«ã—ã¾ã™ã€‚ã“れらを対象ã¨ã—ã¦è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’作æˆã™ã‚‹ã¨ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ãªã‚Šã¾ã™ã€‚IDã¨ãƒ•ã‚¡ã‚¤ãƒ«åã®å¯¾å¿œã¥ã‘ã¯åˆ¥ã®ã€Œæ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã€ã«ä¿å˜ã—ã¦ãŠãã¾ã™ã€‚</p>
+
+<table summary="index example">
+<tr><td>flightless</td><td>1,2</td></tr>
+<tr><td>marine</td><td>1</td></tr>
+<tr><td>birds</td><td>1,2</td></tr>
+<tr><td>in</td><td>2</td></tr>
+<tr><td>africa</td><td>2</td></tr>
+</table>
+
+<p>ãã®æ¬¡ã«èªã¿è¾¼ã‚“ã 文書ã«ã€Œbirdsã€ã¨ã„ã†å˜èªžãŒå«ã¾ã‚Œã¦ã„ãŸå ´åˆã«ã¯ã€`birds' ã®å€¤ã‚’ [1,2,3] ã«å¤‰æ›´ã™ã‚‹ã“ã¨ã«ãªã‚Šã¾ã™ã€‚ã“ã®ã‚ˆã†ã«ã€æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã®æœ«å°¾ã«ãƒ‡ãƒ¼ã‚¿ã‚’è¿½åŠ ã™ã‚‹ã¨ã„ã†æ“作ãŒé »ç¹ã«ç™ºç”Ÿã—ã¾ã™ã€‚ã“ã®å‡¦ç†ã‚’効率的ã«è¡Œã†ãŸã‚ã«ã€Depotã‚„Curiaã®æ›¸ãè¾¼ã¿æ“作ã«ã¯ã€Œé€£çµãƒ¢ãƒ¼ãƒ‰ã€ãŒã‚ã‚‹ã®ã§ã™ã€‚</p>
+
+<p>検索時ã«ã¯ã€æ¤œç´¢èªžã‚’ã‚ーã«ã—ã¦å€¤ã®é…列をå–り出ã—ã€å€‹ã€…ã®IDã«å¯¾å¿œã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«åを文書データベースã‹ã‚‰å–り出ã—ã¦æ示ã™ã‚‹ã“ã¨ã«ãªã‚Šã¾ã™ã€‚ãŸã ã—該当ã®æ–‡æ›¸ãŒå¤šã„å ´åˆã«ãã®å…¨ã¦ã‚’æ示ã—ã¦ã‚‚ユーザã¯å›°æƒ‘ã™ã‚‹ã®ã§ã€å…ˆé ã®ä½•ä»¶ã‹ã«çµžã£ã¦å–り出ã—ã¾ã™ã€‚Depotã‚„Curiaã®èªã¿è¾¼ã¿æ“作ã§ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‹ã‚‰ç‰¹å®šã®éƒ¨åˆ†ã®ã¿ã‚’å–り出ã™ã“ã¨ãŒã§ãã‚‹ã®ã§ã€ã“ã®å‡¦ç†ã‚’効率良ãè¡Œã†ã“ã¨ãŒã§ãã¾ã™ã€‚実際ã¯ã€ãƒ•ã‚¡ã‚¤ãƒ«ã«ãŠã‘ã‚‹å˜èªžã®ã‚¹ã‚³ã‚¢ï¼ˆé‡è¦åº¦ï¼‰ã‚‚é…列è¦ç´ ã®ä¸€éƒ¨ã¨ã—ã¦æ ¼ç´ã—ã¦ãŠã„ã¦ã€ãれを基準ã«ã‚½ãƒ¼ãƒˆã—ã¦ãŠãã“ã¨ã«ã‚ˆã£ã¦ã€å…ˆé ã®è¦ç´ ãŒãƒ¦ãƒ¼ã‚¶ã«ã¨ã£ã¦æ„味ã®ã‚ã‚‹ã‚‚ã®ã«ã—ã¦ã„ã¾ã™ã€‚</p>
+
+<p>ファイルã«ã¯åå‰ä»¥å¤–ã«ã‚‚ã€ã‚¿ã‚¤ãƒˆãƒ«ã‚„作者åや更新日時ã¨ã„ã£ãŸå±žæ€§ã‚’ã¤ã‘ãŸã„å ´åˆãŒã‚ã‚Šã¾ã™ã€‚ãれらã®æƒ…å ±ã¯æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ä¿å˜ã—ã¾ã™ã€‚Odeumã§æ‰±ã†ãƒ•ã‚¡ã‚¤ãƒ«ï¼ˆä»¥å¾Œã¯æ–‡æ›¸ã¨å‘¼ã³ã¾ã™ï¼‰ã¯ã€ãƒ•ã‚¡ã‚¤ãƒ«åã®ä»£ã‚ã‚Šã«URIã‚’ã¤ã‘ã‚‹ã‚‚ã®ã¨ã—ã¦ã„ã¾ã™ã€‚URI以外ã®å±žæ€§ã¯ãƒ¦ãƒ¼ã‚¶ãŒãƒ©ãƒ™ãƒ«ã‚’ã¤ã‘ã¦ä»»æ„ã®ã‚‚ã®ã‚’æ ¼ç´ã§ãã¾ã™ã€‚</p>
+
+<p>Odeumã¯ã€æ–‡æ›¸ã®ãƒ†ã‚ストや属性を元ã®ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰æŠ½å‡ºã™ã‚‹æ©Ÿèƒ½ã«ã¤ã„ã¦ã¯æä¾›ã—ã¾ã›ã‚“。ãれらã¯ãƒ‰ãƒ¡ã‚¤ãƒ³ã«å¼·ãä¾å˜ã™ã‚‹ã®ã§ã€å…±é€šåŒ–ã™ã‚‹ã“ã¨ãŒé›£ã—ã„ã‹ã‚‰ã§ã™ã€‚ã—ãŸãŒã£ã¦ã€ã‚¢ãƒ—リケーションãŒãれを実装ã™ã‚‹å¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚Odeumã®ã‚µãƒ³ãƒ—ルアプリケーションã¯ã€ãƒãƒ¼ã‚«ãƒ«ã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ ã«ã‚るプレーンテã‚ストã¨HTMLã‹ã‚‰ãƒ†ã‚ストã¨å±žæ€§ã‚’抽出ã§ãã¾ã™ã€‚ã‚ãªãŸã®ã‚¢ãƒ—リケーションã§ã¯ã€PDFã‚„MS-Wordã®æ–‡æ›¸ã«å¯¾å¿œã™ã‚‹ã®ã‚‚よã„ã§ã—ょã†ã€‚Webã‹ã‚‰æ–‡æ›¸ã‚’å–å¾—ã—ã¦ã‚‚よã„ã§ã—ょã†ã€‚åŒã˜ç†ç”±ã‹ã‚‰ã€Odeumã¯ãƒ†ã‚ストã‹ã‚‰å˜èªžã‚’抽出ã™ã‚‹æ©Ÿèƒ½ã«ã¤ã„ã¦ã‚‚æä¾›ã—ã¾ã›ã‚“。英語ã¨æ—¥æœ¬èªžã§ã¯å…¨ãç•°ãªã‚‹æ‰‹æ³•ã§ãƒ†ã‚ストã®è§£æžã‚’ã—ãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“ãŒã€ãã†ã„ã£ãŸã“ã¨ã‚‚アプリケーションã«ä»»ã•ã‚Œã¾ã™ã€‚Odeumã®ã‚µãƒ³ãƒ—ルアプリケーションã
§ã¯ã€å˜ã«ç©ºç™½ã§èªžã‚’区切るã¨ã„ã†æ‰‹æ³•ã‚’ã¨ã£ã¦ã„ã¾ã™ã€‚ã‚ãªãŸã®ã‚¢ãƒ—リケーションã§ã¯ã€æ—¥æœ¬èªžã®å½¢æ…‹ç´ 解æžã‚’è¡Œã†ã®ã‚‚よã„ã§ã—ょã†ã€‚英å˜èªžã®ã‚¹ãƒ†ãƒŸãƒ³ã‚°å‡¦ç†ã‚’è¡Œã£ã¦ã‚‚よã„ã§ã—ょã†ã€‚</p>
+
+<p>多ãã®è¨€èªžã§ã¯ã€åŒã˜å˜èªžã«å¯¾ã—ã¦ç•°ä½“や活用ãŒå˜åœ¨ã—ã¾ã™ã€‚例ãˆã°ã€Œä½¿ã†/使ã£ï¼ˆãŸï¼‰ã€ã€Œgo/wentã€ã€Œchild/childrenã€ãªã©ã§ã™ã€‚ãã“ã§ã€Odeumã¯å„å˜èªžã‚’「æ£è¦å½¢ã€ãŠã‚ˆã³ã€Œå‡ºç¾å½¢ã€ã®çµ„ã¨ã—ã¦æ‰±ã„ã¾ã™ã€‚テã‚ストã‹ã‚‰å‡ºç¾å½¢ã®å˜èªžã‚’切り出ã—ãŸã‚Šã€ãれらã®æ£è¦å½¢ã‚’生æˆã™ã‚‹å‡¦ç†ã¯ã‚¢ãƒ—リケーションã«ä»»ã•ã‚Œã¾ã™ã€‚転置インデックスã§ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã«ã¯æ£è¦å½¢ãŒä½¿ã‚ã‚Œã¾ã™ã€‚「childã€ã§æ¤œç´¢ã™ã‚Œã°ã€ã€Œchildrenã€ã‚’å«ã‚€æ–‡æ›¸ã‚‚該当ã•ã›ã‚‰ã‚Œã‚‹ã¨ã„ã†ã“ã¨ã§ã™ã€‚検索語ã«å¯¾ã—ã¦ã‚‚æ£è¦åŒ–ã®å‡¦ç†ã‚’è¡Œãˆã°ã€ã€Œchildrenã€ã§æ¤œç´¢ã—ã¦ã€Œchildã€ã‚’該当ã«ã§ãã¾ã™ã€‚ãªãŠã€å‡ºç¾å½¢ã‚‚データベースã«è¨˜éŒ²ã•ã‚Œã¾ã™ãŒã€ãã‚Œã¯æ¤œç´¢çµæžœã¨ã—ã¦æ–‡æ›¸ã®è¦ç´„ã‚’æ示ã™ã‚‹ãªã©ã®ç”¨é€”ã§åˆ©ç”¨ã•ã‚Œã¾ã™ï¼ˆä¸è¦ãªå ´åˆã¯å‡ºç¾å½¢ã‚’å…¨ã¦ç©ºæ–‡å—列ã«ã—ã¦è¨˜æ†¶é ˜åŸŸã‚’節約ã§
ãã¾ã™ï¼‰ã€‚</p>
+
+<p>全文検索システムã®å–„ã—悪ã—を評価ã™ã‚‹éš›ã«ã¯ã€ã‚¹ã‚³ã‚¢ãƒªãƒ³ã‚°ï¼ˆãƒ©ãƒ³ã‚ング）ãŒé‡è¦ãªè¦ç´ ã«ãªã‚Šã¾ã™ã€‚検索çµæžœãŒå¤šã„å ´åˆã«ã€ãã®ä¸ã‹ã‚‰çµžã‚Šè¾¼ã‚“ã§ãƒ¦ãƒ¼ã‚¶ã«æ示ã™ã‚‹æ–‡æ›¸ã‚’ã©ã†ã‚„ã£ã¦é¸æŠžã™ã‚‹ã‹ã¨ã„ã†ã“ã¨ã§ã™ã€‚ã„ãら検索速度ãŒé€Ÿãã¦ã‚‚ã€æº€è¶³ã®ã„ã検索çµæžœãŒæ示ã•ã‚Œãšã«ä½•åº¦ã‚‚検索ã—ãŸã‚Šã€ç„¡é§„ãªæ–‡æ›¸ã‚’閲覧ã—ãªã‘ã‚Œã°ãªã‚‰ãªã„ã®ã§ã¯ã€çµå±€ã¯æ™‚é–“ãŒã‹ã‹ã‚‹ã“ã¨ã«ãªã£ã¦ã—ã¾ã„ã¾ã™ã€‚Odeumã§ã¯ã€æ–‡æ›¸ã‚’登録ã™ã‚‹éš›ã«ã€ãã“ã«å«ã¾ã‚Œã‚‹å„å˜èªžã«ã¤ã„ã¦ä»¥ä¸‹ã®å¼ã§ã‚¹ã‚³ã‚¢ã‚’算出ã—ã¦ã€æ–‡æ›¸IDã¨ã¨ã‚‚ã«è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã«æ ¼ç´ã—ã¦ã„ã¾ã™ã€‚</p>
+
+<div><kbd>スコア = (該当語ã®å‡ºç¾æ•° * 10000) / log(文書内ã®ç·èªžæ•°) ^ 3</kbd></div>
+
+<p>ã‚ã‚‹å˜èªžãŒãŸãã•ã‚“出ç¾ã™ã‚‹ã¨ã„ã†ã“ã¨ã¯ã€ãã®æ–‡æ›¸ã§ãã®å˜èªžã®ã“ã¨ã«ã¤ã„ã¦è©³ã—ã説明ã•ã‚Œã¦ã„ã‚‹å¯èƒ½æ€§ãŒé«˜ã„ã¨åˆ¤æ–ã§ãã¾ã™ã€‚ãŸã ã—ã€å¤§ããªæ–‡æ›¸ã¯å°ã•ã„文書よりもå˜èªžã®å‡ºç¾æ•°ãŒå¤šããªã‚Šã¾ã™ã®ã§ã€ãã®èª¿æ•´ã‚’ã™ã‚‹å¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚ãã“ã§ã€è©²å½“語ã®å‡ºç¾æ•°ã‚’ã€ç·èªžæ•°ã®è‡ªç„¶å¯¾æ•°ã®ä¸‰ä¹—ã§å‰²ã£ã¦ã„ã‚‹ã®ã§ã™ã€‚ãªãŠã€ãƒ†ã‚ストã®å…ˆé 10%以内ã«å‡ºã¦æ¥ã‚‹å˜èªžã¯ã€Œãƒˆãƒ”ックセンテンスã€ã‚’構æˆã™ã‚‹ã¨ã¿ãªã—ã¦ã€åˆå‡ºæ™‚ã«é™ã‚Š10000ã§ãªã15000ã‚’åŠ ç®—ã—ã¦ã„ã¾ã™ã€‚</p>
+
+<p>全文検索ã¯é€šå¸¸ã€è¤‡æ•°ã®æ¤œç´¢èªžã‚’使ã£ã¦è¡Œã‚ã‚Œã¾ã™ã€‚ãã®éš›ã«ã¯ã€å„検索語ã®ã‚¹ã‚³ã‚¢ã‚’調整ã—ãªã„ã¨ã€ç‰¹å®šã®æ¤œç´¢èªžã®ã‚¹ã‚³ã‚¢ã®å½±éŸ¿ãŒå¼·ã™ãŽã‚‹ã¨ã„ã†äº‹æ…‹ãŒç™ºç”Ÿã—ã¾ã™ã€‚例ãˆã°ã€Œthe beatlesã€ã§æ¤œç´¢ã—ãŸå ´åˆã€ã€Œtheã€ã®ã‚¹ã‚³ã‚¢ãŒã€Œbeatlesã€ã®ã‚¹ã‚³ã‚¢ã‚’圧倒ã—ã¦ã€ã€Œbeatlesã€ã«ã¤ã„ã¦çŸ¥ã‚ŠãŸã„ã®ã«ã€é–¢ä¿‚ãªã„文書ã°ã‹ã‚ŠãŒæ示ã•ã‚Œã‚‹ã“ã¨ã«ãªã£ã¦ã—ã¾ã„ã¾ã™ã€‚Odeumã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®å†…容をæ示ã™ã‚‹ã ã‘ã§ã€ãれ以後ã®å‡¦ç†ã¯ã‚¢ãƒ—リケーションã«ä»»ã•ã‚Œã¾ã™ã€‚Odeumã«ä»˜å±žã™ã‚‹ã‚µãƒ³ãƒ—ルアプリケーションã§ã¯ä»¥ä¸‹ã®å¼ã§èª¿æ•´ã‚’è¡Œã£ã¦ã„ã¾ã™ã€‚</p>
+
+<div><kbd>調整済ã¿ã‚¹ã‚³ã‚¢ = 登録ã•ã‚ŒãŸã‚¹ã‚³ã‚¢ / log(ãã®èªžã‚’å«ã‚€æ–‡æ›¸æ•°) ^ 2</kbd></div>
+
+<p>ã‚ã‚Šãµã‚ŒãŸå˜èªžã¯ã‚¹ã‚³ã‚¢ã‚’下ã’ã€ç‰¹å¾´çš„ãªå˜èªžã¯ã‚¹ã‚³ã‚¢ã‚’ç¶æŒã™ã¹ãã§ã™ã€‚ãã“ã§ã€ç™»éŒ²ã•ã‚ŒãŸã‚¹ã‚³ã‚¢ã‚’ã€ãã®èªžã‚’å«ã‚€æ–‡æ›¸æ•°ã®è‡ªç„¶å¯¾æ•°ã®äºŒä¹—ã§å‰²ã£ã¦ã„ã‚‹ã®ã§ã™ï¼ˆTF-IDF法を強化ã—ãŸã‚‚ã®ã§ã™ï¼‰ã€‚å„検索語ã®èª¿æ•´æ¸ˆã¿ã‚¹ã‚³ã‚¢ã‚’足ã—ãŸã‚‚ã®ã‚’文書ã®ã‚¹ã‚³ã‚¢ã¨ã—ã€ãã®é™é †ã§æ¤œç´¢çµæžœã‚’æ示ã—ã¾ã™ã€‚ä»–ã«ã‚‚ドメインã«ä¾å˜ã—ãŸæ§˜ã€…ãªã‚¹ã‚³ã‚¢ãƒªãƒ³ã‚°æ‰‹æ³•ãŒã‚ã‚‹ã¨æ€ã„ã¾ã™ãŒã€ã‚¢ãƒ—リケーションãŒãれを実装ã§ãã‚‹ã®ãŒå¬‰ã—ã„ã¨ã“ã‚ã§ã™ã€‚</p>
+
+<p>Odeumã®ã‚µãƒ³ãƒ—ルアプリケーションã§ã¯ã€é–¢é€£æ–‡æ›¸æ¤œç´¢ï¼ˆé¡žä¼¼æ–‡æ›¸æ¤œç´¢ã¨å‘¼ã‚“ã æ–¹ãŒé©åˆ‡ã‹ã‚‚ã—ã‚Œã¾ã›ã‚“）ã®æ©Ÿèƒ½ã‚‚実装ã—ã¦ã„ã¾ã™ã€‚ã‚る文書（種文書ã¨å‘¼ã³ã¾ã™ï¼‰ã«é–¢é€£ã—ãŸæ–‡æ›¸ã®ä¸€è¦§ã‚’æ示ã™ã‚‹ã‚‚ã®ã§ã™ã€‚å˜èªžã®å‡ºç¾å‚¾å‘ãŒä¼¼é€šã£ãŸæ–‡æ›¸ã¯äº’ã„ã«é–¢é€£ã—ã¦ã„ã‚‹ã¨ã„ã†è€ƒãˆæ–¹ï¼ˆãƒ™ã‚¯ãƒˆãƒ«ç©ºé–“モデル）ã«åŸºã¥ã„ã¦ã„ã¾ã™ã€‚文書をデータベースã«ç™»éŒ²ã™ã‚‹éš›ã«ã¯ã€å„文書ã«å«ã¾ã‚Œã‚‹å…¨å˜èªžã«å¯¾ã—ã¦èª¿æ•´æ¸ˆã¿ã‚¹ã‚³ã‚¢ã‚’計算ã—ã€ãã®ä¸Šä½32語（ã‚ーワードã¨å‘¼ã³ã¾ã™ï¼‰ã®æƒ…å ±ã‚’æ–‡æ›¸ã¨å¯¾å¿œã¥ã‘ã¦ç™»éŒ²ã—ã¦ãŠãã¾ã™ã€‚検索時ã«ã¯ã€ç¨®æ–‡æ›¸ã®ã‚ーワードã¨ã‚¹ã‚³ã‚¢ã‚’å–り出ã—ã€ãれを32次元ã®ãƒ™ã‚¯ãƒˆãƒ«ã¨ã—ã¦è¡¨ç¾ã—ã¾ã™ã€‚関連度を判定ã™ã‚‹å¯¾è±¡ã®å„文書ã‹ã‚‰ã‚‚ã‚ーワードã¨ã‚¹ã‚³ã‚¢ã‚’å–り出ã—ã€ç¨®æ–‡æ›¸ã®ãƒ™ã‚¯ãƒˆãƒ«ç©ºé–“ã«å¯¾å¿œã—ãŸãƒ™ã‚¯ãƒˆãƒ«ã‚’生æˆã—ã¾ã™ã€
‚ãã†ã—ã¦ã§ããŸäºŒã¤ã®ãƒ™ã‚¯ãƒˆãƒ«ã®ãªã™è§’ãŒå°ã•ã„ã‚‚ã®ã¯é–¢é€£åº¦ãŒé«˜ã„ã¨åˆ¤å®šã—ã¾ã™ã€‚実際ã«ã¯ã€ãªã™è§’ã®ä½™å¼¦ï¼ˆ0ã‹ã‚‰1ã®ç¯„囲ã§ã€å®Œå…¨ä¸€è‡´ã™ã‚‹å ´åˆã¯1ã«ãªã‚‹ï¼‰ãŒå¤§ãã„ã‚‚ã®ã‹ã‚‰æ示ã•ã‚Œã‚‹ã“ã¨ã«ãªã‚Šã¾ã™ã€‚ãªãŠã€ç™»éŒ²ã•ã‚ŒãŸå…¨ã¦ã®æ–‡æ›¸ã‚’対象ã¨ã—ã¦é¡žä¼¼åº¦ã®åˆ¤å®šã‚’è¡Œã†ã¨ã‚ã¾ã‚Šã«æ™‚é–“ãŒã‹ã‹ã‚‹ã®ã§ã€ã‚ーワードã§OR検索を行ã£ãŸçµæžœã®ä¸Šä½ã®æ–‡æ›¸ã®ã¿ã‚’関連度算出ã®å¯¾è±¡ã¨ã—ã¦ã„ã¾ã™ã€‚</p>
+
+<p>å‰ç½®ãã¯ã“ã“ã¾ã§ã«ã—ã¦ã€Odeumã®ä½¿ç”¨æ–¹æ³•ã«ã¤ã„ã¦ã®èª¬æ˜Žã«å…¥ã‚Šã¾ã™ã€‚URI㌠`http://tako.ika/uni.txt' ã§ã€ãã“ã‹ã‚‰å–り出ã—ãŸãƒ†ã‚ストãŒã€ŒThe sun is driven by the Grateful Dead.ã€ã§ã€åˆ‡ã‚Šå‡ºã—ãŸå˜èªžï¼ˆæ£è¦å½¢/出ç¾å½¢ï¼‰ãŒã€Œthe/Theã€ã€Œsun/sunã€ã€Œbe/isã€ã€Œdrive/drivenã€ã€Œby/byã€ã€Œthe/theã€ã€Œgrateful/Gratefulã€ã€Œdie/deadã€ã€Œï¼ˆæ£è¦å½¢ãªã—）/.ã€ã ã¨ã—ã¾ã—ょã†ï¼ˆã“ã®å‡¦ç†ã¯ã‚¢ãƒ—リケーションãŒç‹¬è‡ªã«å®Ÿè£…ã—ã¦ãã ã•ã„）（ã„ã‚ゆるストップワードã¯æ£è¦å½¢ã‚’空文å—列ã«ã—ã¦è¡¨ç¾ã—ã¾ã™ï¼‰ã€‚ãれを `index' ã¨ã„ã†åå‰ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ç™»éŒ²ã™ã‚‹ä¾‹ã‚’示ã—ã¾ã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+
+int docregistertest(void){
+ ODEUM *odeum;
+ ODDOC *doc;
+ /* データベースを開ã */
+ if(!(odeum = odopen("index", OD_OWRITER | OD_OCREAT))) return -1;
+ /* 文書を表ç¾ã™ã‚‹ */
+ doc = oddocopen("http://tako.ika/uni.txt");
+ oddocaddword(doc, "the", "the");
+ oddocaddword(doc, "sun", "sun");
+ oddocaddword(doc, "be", "is");
+ oddocaddword(doc, "drive", "driven");
+ oddocaddword(doc, "by", "by");
+ oddocaddword(doc, "the", "the");
+ oddocaddword(doc, "grateful", "Grateful");
+ oddocaddword(doc, "die", "Dead");
+ oddocaddword(doc, "", ".");
+ /* 文書を登録ã™ã‚‹ */
+ odput(odeum, doc, -1, 1);
+ /* 文書ã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ oddocclose(doc);
+ /* データベースを閉ã˜ã‚‹ */
+ if(!odclose(odeum)) return -1;
+ return 0;
+}
+</pre>
+
+<p>`ODEUM' ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã¯ä¾‹ã«ã‚ˆã£ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã™ã€‚`odopen' ã§ãã®ãƒãƒ³ãƒ‰ãƒ«ã‚’ç²å¾—ã—ã¾ã™ã€‚é–‹ã„ãŸãƒãƒ³ãƒ‰ãƒ«ã¯ `odclose' ã§é–‰ã˜ã¾ã™ã€‚`ODDOC' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã§ã™ã€‚å„文書ã®å†…容ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã«ã‚ˆã£ã¦è¡¨ç¾ã•ã‚Œã¾ã™ã€‚`oddocopen' ã¯ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ã関数ã§ã™ã€‚ãã®ç¬¬1引数ã§æ–‡æ›¸ã®URIを指定ã—ã¾ã™ã€‚文書ãƒãƒ³ãƒ‰ãƒ«ã¯ä¸è¦ã«ãªã£ãŸã‚‰ `oddocclose' ã§è§£æ”¾ã—ã¾ã™ã€‚テã‚ストã®å„å˜èªžã‚’文書ã«ç™»éŒ²ã™ã‚‹ã«ã¯ã€`oddocaddword' を用ã„ã¾ã™ã€‚ãã®ç¬¬2引数ã¯å˜èªžã®æ£è¦å½¢ã§ã€ç¬¬3引数ã¯å‡ºç¾å½¢ã§ã™ã€‚文書を表ç¾ã—ãŸã‚‰ã€ãれを `odput' ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ç™»éŒ²ã—ã¾ã™ã€‚第3引数ã¯ã€æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ç™»éŒ²ã™ã‚‹èªžæ•°ã‚’指定ã—ã¾ã™ã€‚`-1' ã«ã™ã‚‹ã¨å…¨éƒ¨ã®èªžãŒç™»éŒ²ã•ã‚Œã¾ã™ã€‚第4引数ã¯ã€åŒã˜URIã®æ—¢å˜ã®æ–‡æ›¸ãŒã‚ã‚‹å ´åˆã«ã€ãれを上書ãã™ã‚‹ã‹å¦ã‹æŒ‡å®šã—ã¾ã™ã€‚</p>
+
+<p>`odput' ã®ç¬¬3引数ã®æŒ‡å®šãŒå°‘ã—難ã—ã„ã®ã§è£œè¶³ã—ã¾ã™ã€‚ã‚る文書ãŒé©åˆ‡ã«æ¤œç´¢ã•ã‚Œã‚‹ãŸã‚ã«ã€è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã«ãŠã„ã¦ã€ãã®æ–‡æ›¸ã«å«ã¾ã‚Œã‚‹å…¨ã¦ã®å˜èªžã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã«ãã®æ–‡æ›¸ã®IDãŒç„¡æ¡ä»¶ã§è¿½åŠ ã•ã‚Œã¾ã™ã€‚ã¨ã“ã‚ã§ã€å¤šãã®å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ ã§ã¯ã€æ¤œç´¢çµæžœã®ç”»é¢ã§è©²å½“ã®æ–‡æ›¸ã®è¦ç´„を表示ã—ã¾ã™ã€‚ãã®ãŸã‚ã«ã¯ã€æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ä¸ã«ã€å„文書ã¨é–¢é€£ã¥ã‘ã¦å«ã¾ã‚Œã‚‹å˜èªžã‚’é †ç•ªã«è¨˜éŒ²ã—ã¦ãŠãå¿…è¦ãŒã‚ã‚Šã¾ã™ã€‚検索語ã®å‘¨è¾ºã®æ–‡ã‚’切り出ã—ã¦è¡¨ç¤ºã™ã‚‹å ´åˆã‚’考ãˆã‚‹ã¨ã€æ¤œç´¢èªžãŒæ–‡æ›¸ä¸ã®ã©ã“ã«ç¾ã‚Œã‚‹ã‹ã¯äºˆæƒ³ã§ããªã„ã®ã§ã€å…¨ã¦ã®å˜èªžã‚’文書データベースã«è¨˜éŒ²ã—ã¦ãŠã‹ãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“。ã‚ã‚‹ã„ã¯ã€å†’é ã®ä½•èªžã‹ã ã‘表示ã™ã‚‹å ´åˆã«ã¯ã€ãã®èªžæ•°åˆ†ã®èªžã‚’登録ã—ã¦ãŠã‘ã°ã‚ˆã„ã“ã¨ã«ãªã‚Šã¾ã™ã€‚ãã†ã„ã£ãŸæ±ºå®šã‚’アプリケ
ーションã«ä»»ã›ã‚‹ãŸã‚ã«ã€æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ç™»éŒ²ã™ã‚‹èªžæ•°ã‚’指定ã§ãるよã†ã«ãªã£ã¦ã„ã‚‹ã®ã§ã™ã€‚</p>
+
+<p>以下ã®ä¾‹ã§ã¯ã€`grateful' ã¨ã„ã†å˜èªžã‚’å«ã‚€æ–‡æ›¸ã‚’検索ã—ã¦ã€ãã®URIã¨ã‚¹ã‚³ã‚¢ã‚’表示ã—ã¾ã™ã€‚ã¾ãšè»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’検索ã—ã¦çµæžœã®é…列をå—ã‘å–ã‚Šã¾ã™ã€‚ãã®é…列ã®å„è¦ç´ ã¯æ–‡æ›¸ã®IDã¨ã‚¹ã‚³ã‚¢ã®çµ„ã§ã™ã€‚文書ã®å†…容をå–å¾—ã™ã‚‹ã«ã¯ã€æ–‡æ›¸IDを使ã£ã¦æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«å•ã„åˆã‚ã›ã¾ã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+void docsearchtest(void){
+ ODEUM *odeum;
+ ODPAIR *pairs;
+ ODDOC *doc;
+ int i, pnum;
+ /* データベースをèªã¿è¾¼ã¿ãƒ¢ãƒ¼ãƒ‰ã§é–‹ã */
+ if(!(odeum = odopen("index", OD_OREADER))) return;
+ /* 転置インデックスを検索ã™ã‚‹ */
+ pairs = odsearch(odeum, "grateful", -1, &pnum);
+ if(pairs && pnum > 0){
+ /* 該当ã®å„文書を処ç†ã™ã‚‹ */
+ for(i = 0; i < pnum; i++){
+ /* 文書データベースã‹ã‚‰æ–‡æ›¸ã‚’å–り出㙠*/
+ if(!(doc = odgetbyid(odeum, pairs[i].id))) continue;
+ /* 文書ã®å†…容を表示ã™ã‚‹ */
+ printf("URI: %s\n", oddocuri(doc));
+ printf("SCORE: %d\n", pairs[i].score);
+ /* 文書ã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ oddocclose(doc);
+ }
+ }
+ /* 検索çµæžœã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ free(pairs);
+ /* データベースを閉ã˜ã‚‹ */
+ odclose(odeum);
+}
+</pre>
+
+<p>転置インデックスを検索ã™ã‚‹ã®ãŒ `odsearch' ã¨ã„ã†é–¢æ•°ã§ã™ã€‚第2引数ã«ã¯æ£è¦å½¢ã®æ¤œç´¢èªžã‚’指定ã—ã¾ã™ã€‚第3引数ã«ã¯ã€æ¤œç´¢ã™ã‚‹æ–‡æ›¸ã®æœ€å¤§æ•°ã‚’指定ã—ã¾ã™ãŒã€å…¨éƒ¨å–り出ã™å ´åˆã¯ `-1' ã¨ã—ã¾ã™ã€‚çµæžœã®é…列ã¯ã‚¹ã‚³ã‚¢ã®é™é †ã§ã‚½ãƒ¼ãƒˆã•ã‚Œã¦ã„ã‚‹ã“ã¨ãŒè¦å®šã•ã‚Œã¦ã„ã¾ã™ã€‚第4引数ã«ã¯ã€çµæžœã®é…列ã®è¦ç´ æ•°ã‚’æ ¼ç´ã™ã‚‹å¤‰æ•°ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã—ã¾ã™ã€‚次ã«ã€é…列ã®å„è¦ç´ を処ç†ã—ã¦ã„ãã¾ã™ã€‚`odgetbyid' ã¯æ–‡æ›¸IDを用ã„ã¦æ–‡æ›¸ã®å†…容をå•ã„åˆã‚ã›ã‚‹é–¢æ•°ã§ã™ã€‚転置インデックスã®ä¸ã«ã¯æ—¢ã«å‰Šé™¤ã•ã‚ŒãŸã‚Šä¸Šæ›¸ãã•ã‚Œã¦IDãŒå¤‰æ›´ã•ã‚ŒãŸæ–‡æ›¸ã®æƒ…å ±ã‚‚å…¥ã£ã¦ã„ã‚‹ã®ã§ï¼ˆæœ€é©åŒ–ã™ã‚Œã°ä¸è¦ãªæƒ…å ±ã¯ãªããªã‚Šã¾ã™ãŒï¼‰ã€`odgetbyid' ã¯å¤±æ•—ã™ã‚‹å¯èƒ½æ€§ãŒã‚ã‚Šã¾ã™ã€‚ãã†ã„ã†æ™‚ã¯å˜ã«ç„¡è¦–ã—ã¦æ¬¡ã®ãƒ«ãƒ¼ãƒ—ã«é€²ã‚“ã§ãã ã•ã„。文書ãŒå–å¾—ã§ããŸã‚‰ã€ã‚ã¨ã¯ãã‚Œã
‚’表示ã—ã¾ã™ã€‚`oddocuri' ã¯æ–‡æ›¸ã®URIã‚’è¿”ã™é–¢æ•°ã§ã™ã€‚ä»–ã«ã‚‚文書ã®æƒ…å ±ã‚’å–å¾—ã™ã‚‹é–¢æ•°ãŒã„ãã¤ã‹ç”¨æ„ã•ã‚Œã¦ã„ã¾ã™ã€‚</p>
+
+<p>Odeumã§ã¯ã€è¤‡æ•°ã®æ¤œç´¢èªžã‚’用ã„ã¦ã€ANDæ¡ä»¶ï¼ˆæ¤œç´¢èªžã®å…¨ã¦ã‚’å«ã‚€ï¼‰ã‚„ORæ¡ä»¶ï¼ˆæ¤œç´¢èªžã®ã„ãšã‚Œã‹ã‚’å«ã‚€ï¼‰ã‚„NOTANDæ¡ä»¶ï¼ˆæ¤œç´¢èªžã®å‰è€…ã‚’å«ã‚€ãŒå¾Œè€…ã¯å«ã¾ãªã„）ã¨ã„ã£ãŸé›†åˆæ¼”算を処ç†ã™ã‚‹ãŸã‚ã®é–¢æ•°ãŒç”¨æ„ã•ã‚Œã¦ã„ã‚‹ã»ã‹ã€å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ ã®å®Ÿè£…ã«ä¾¿åˆ©ãªãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£é–¢æ•°ãŒå¤šæ•°æä¾›ã•ã‚Œã¾ã™ã€‚全文検索システムを実装ã™ã‚‹éš›ã«ã¯æ€§èƒ½ã¨ç²¾åº¦ã®ãƒãƒ©ãƒ³ã‚¹ã‚’考ãˆãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“ãŒã€Odeumã®APIã¯ã‚¢ãƒ—リケーションãŒãれを任æ„ã«æ±ºã‚られるよã†ã«è¨è¨ˆã•ã‚Œã¦ã„ã¾ã™ã€‚大è¦æ¨¡ãªã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’扱ã†éš›ã«ã¯ã€ã¾ãšç²¾åº¦ã‚’è½ã—ãŸæ¤œç´¢ã‚’è¡Œã£ã¦ã€ãã®çµæžœãŒãƒ¦ãƒ¼ã‚¶ã®è¦æ±‚を満ãŸã•ãªã‘ã‚Œã°ç²¾åº¦ã‚’高ã‚ãŸãƒ‘ラメータã§å†æ¤œç´¢ã™ã‚‹æ‰‹æ³•ãŒæœ‰åŠ¹ã§ã—ょã†ã€‚</p>
+
+<hr />
+
+</body>
+
+</html>
+
+<!-- END OF FILE -->
Added: box/trunk/qdbm/misc/win32check.bat
===================================================================
--- box/trunk/qdbm/misc/win32check.bat (rev 0)
+++ box/trunk/qdbm/misc/win32check.bat 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,111 @@
+dptest write casket 50000 5000
+if errorlevel 1 goto error
+dptest read casket
+if errorlevel 1 goto error
+dptest read -wb casket
+if errorlevel 1 goto error
+dptest rcat casket 50000 50 500 32 32
+if errorlevel 1 goto error
+dptest combo casket
+if errorlevel 1 goto error
+dptest wicked casket 5000
+if errorlevel 1 goto error
+del /Q casket
+
+crtest write casket 50000 500 10
+if errorlevel 1 goto error
+crtest read casket
+if errorlevel 1 goto error
+crtest read -wb casket
+if errorlevel 1 goto error
+crtest rcat casket 50000 5 10 500 32 32
+if errorlevel 1 goto error
+crtest combo casket
+if errorlevel 1 goto error
+crtest wicked casket 5000
+if errorlevel 1 goto error
+rd /S /Q casket
+
+crtest write -lob casket 1000 50 10
+if errorlevel 1 goto error
+crtest read -lob casket
+if errorlevel 1 goto error
+rd /S /Q casket
+
+rltest write casket 50000
+if errorlevel 1 goto error
+rltest read casket 50000
+if errorlevel 1 goto error
+del /Q casket*
+
+hvtest write casket 50000
+if errorlevel 1 goto error
+hvtest read casket 50000
+if errorlevel 1 goto error
+del /Q casket
+
+hvtest write -qdbm casket 50000
+if errorlevel 1 goto error
+hvtest read -qdbm casket 50000
+if errorlevel 1 goto error
+rd /S /Q casket
+
+cbtest sort 5000
+if errorlevel 1 goto error
+cbtest strstr 500
+if errorlevel 1 goto error
+cbtest list 50000
+if errorlevel 1 goto error
+cbtest map 50000
+if errorlevel 1 goto error
+cbtest wicked 5000
+if errorlevel 1 goto error
+cbtest misc
+if errorlevel 1 goto error
+
+vltest write -tune 25 64 32 32 casket 50000
+if errorlevel 1 goto error
+vltest read casket
+if errorlevel 1 goto error
+vltest rdup -tune 25 64 256 256 casket 50000 50000
+if errorlevel 1 goto error
+vltest combo casket
+if errorlevel 1 goto error
+vltest wicked casket 5000
+if errorlevel 1 goto error
+del /Q casket
+
+vltest write -int -cz -tune 25 64 32 32 casket 50000
+if errorlevel 1 goto error
+vltest read -int casket
+if errorlevel 1 goto error
+vltest rdup -int -cz -tune 25 64 256 256 casket 50000 50000
+if errorlevel 1 goto error
+vltest combo -cz casket
+if errorlevel 1 goto error
+vltest wicked -cz casket 5000
+if errorlevel 1 goto error
+del /Q casket
+
+odtest write casket 500 50 5000
+if errorlevel 1 goto error
+odtest read casket
+if errorlevel 1 goto error
+odtest combo casket
+if errorlevel 1 goto error
+odtest wicked casket 500
+if errorlevel 1 goto error
+rd /S /Q casket
+
+ at echo off
+echo #================================
+echo # SUCCESS
+echo #================================
+goto :EOF
+
+:error
+ at echo off
+echo #================================
+echo # ERROR
+echo #================================
+goto :EOF
Added: box/trunk/qdbm/myconf.c
===================================================================
--- box/trunk/qdbm/myconf.c (rev 0)
+++ box/trunk/qdbm/myconf.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,1113 @@
+/*************************************************************************************************
+ * Emulation of system calls
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include "myconf.h"
+
+
+
+/*************************************************************************************************
+ * for dosish filesystems
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_) || defined(_SYS_CYGWIN_)
+
+
+#define DOSPATHBUFSIZ 8192
+
+
+int _qdbm_win32_lstat(const char *pathname, struct stat *buf){
+ char pbuf[DOSPATHBUFSIZ], *p;
+ int inode;
+ if(stat(pathname, buf) == -1) return -1;
+ if(GetFullPathName(pathname, DOSPATHBUFSIZ, pbuf, &p) != 0){
+ inode = 11003;
+ for(p = pbuf; *p != '\0'; p++){
+ inode = inode * 31 + *(unsigned char *)p;
+ }
+ buf->st_ino = (inode * 911) & 0x7FFF;
+ }
+ return 0;
+}
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for POSIX thread
+ *************************************************************************************************/
+
+
+#if defined(MYPTHREAD)
+
+
+#include <pthread.h>
+
+
+#define PTKEYMAX 8
+
+
+struct { void *ptr; pthread_key_t key; } _qdbm_ptkeys[PTKEYMAX];
+int _qdbm_ptknum = 0;
+
+
+static void *_qdbm_gettsd(void *ptr, int size, const void *initval);
+
+
+void *_qdbm_settsd(void *ptr, int size, const void *initval){
+ static pthread_mutex_t mutex = PTHREAD_MUTEX_INITIALIZER;
+ char *val;
+ if((val = _qdbm_gettsd(ptr, size, initval)) != NULL) return val;
+ if(pthread_mutex_lock(&mutex) != 0) return NULL;
+ if((val = _qdbm_gettsd(ptr, size, initval)) != NULL){
+ pthread_mutex_unlock(&mutex);
+ return val;
+ }
+ if(_qdbm_ptknum >= PTKEYMAX){
+ pthread_mutex_unlock(&mutex);
+ return NULL;
+ }
+ _qdbm_ptkeys[_qdbm_ptknum].ptr = ptr;
+ if(pthread_key_create(&(_qdbm_ptkeys[_qdbm_ptknum].key), free) != 0){
+ pthread_mutex_unlock(&mutex);
+ return NULL;
+ }
+ if(!(val = malloc(size))){
+ pthread_key_delete(_qdbm_ptkeys[_qdbm_ptknum].key);
+ pthread_mutex_unlock(&mutex);
+ return NULL;
+ }
+ memcpy(val, initval, size);
+ if(pthread_setspecific(_qdbm_ptkeys[_qdbm_ptknum].key, val) != 0){
+ free(val);
+ pthread_key_delete(_qdbm_ptkeys[_qdbm_ptknum].key);
+ pthread_mutex_unlock(&mutex);
+ return NULL;
+ }
+ _qdbm_ptknum++;
+ pthread_mutex_unlock(&mutex);
+ return val;
+}
+
+
+static void *_qdbm_gettsd(void *ptr, int size, const void *initval){
+ char *val;
+ int i;
+ for(i = 0; i < _qdbm_ptknum; i++){
+ if(_qdbm_ptkeys[i].ptr == ptr){
+ if(!(val = pthread_getspecific(_qdbm_ptkeys[i].key))){
+ if(!(val = malloc(size))) return NULL;
+ memcpy(val, initval, size);
+ if(pthread_setspecific(_qdbm_ptkeys[i].key, val) != 0){
+ free(val);
+ return NULL;
+ }
+ }
+ return val;
+ }
+ }
+ return NULL;
+}
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for systems without mmap
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_)
+
+
+#define MMFDESCMAX 2048
+
+
+struct { void *start; HANDLE handle; } mmhandles[MMFDESCMAX];
+int mmhnum = 0;
+CRITICAL_SECTION mmcsec;
+
+
+static void _qdbm_delete_mmap_env(void);
+
+
+void *_qdbm_mmap(void *start, size_t length, int prot, int flags, int fd, off_t offset){
+ static volatile long first = TRUE;
+ static volatile long ready = FALSE;
+ HANDLE handle;
+ int i;
+ if(InterlockedExchange((void *)&first, FALSE)){
+ InitializeCriticalSection(&mmcsec);
+ atexit(_qdbm_delete_mmap_env);
+ InterlockedExchange((void *)&ready, TRUE);
+ }
+ while(!InterlockedCompareExchange((void *)&ready, TRUE, TRUE)){
+ Sleep(1);
+ }
+ if(fd < 0 || flags & MAP_FIXED) return MAP_FAILED;
+ if(!(handle = CreateFileMapping((HANDLE)_get_osfhandle(fd), NULL,
+ (prot & PROT_WRITE) ? PAGE_READWRITE : PAGE_READONLY,
+ 0, length, NULL))) return MAP_FAILED;
+ if(!(start = MapViewOfFile(handle, (prot & PROT_WRITE) ? FILE_MAP_WRITE : FILE_MAP_READ,
+ 0, 0, length))){
+ CloseHandle(handle);
+ return MAP_FAILED;
+ }
+ EnterCriticalSection(&mmcsec);
+ if(mmhnum >= MMFDESCMAX - 1){
+ UnmapViewOfFile(start);
+ CloseHandle(handle);
+ LeaveCriticalSection(&mmcsec);
+ return MAP_FAILED;
+ }
+ for(i = 0; i < MMFDESCMAX; i++){
+ if(!mmhandles[i].start){
+ mmhandles[i].start = start;
+ mmhandles[i].handle = handle;
+ break;
+ }
+ }
+ mmhnum++;
+ LeaveCriticalSection(&mmcsec);
+ return start;
+}
+
+
+int _qdbm_munmap(void *start, size_t length){
+ HANDLE handle;
+ int i;
+ EnterCriticalSection(&mmcsec);
+ handle = NULL;
+ for(i = 0; i < MMFDESCMAX; i++){
+ if(mmhandles[i].start == start){
+ handle = mmhandles[i].handle;
+ mmhandles[i].start = NULL;
+ mmhandles[i].handle = NULL;
+ break;
+ }
+ }
+ if(!handle){
+ LeaveCriticalSection(&mmcsec);
+ return -1;
+ }
+ mmhnum--;
+ LeaveCriticalSection(&mmcsec);
+ if(!UnmapViewOfFile(start)){
+ CloseHandle(handle);
+ return -1;
+ }
+ if(!CloseHandle(handle)) return -1;
+ return 0;
+}
+
+
+int _qdbm_msync(const void *start, size_t length, int flags){
+ if(!FlushViewOfFile(start, length)) return -1;
+ return 0;
+}
+
+
+static void _qdbm_delete_mmap_env(void){
+ DeleteCriticalSection(&mmcsec);
+}
+
+
+#elif defined(_SYS_FREEBSD_) || defined(_SYS_NETBSD_) || defined(_SYS_OPENBSD_) || \
+ defined(_SYS_AIX_) || defined(_SYS_RISCOS_) || defined(MYNOMMAP)
+
+
+void *_qdbm_mmap(void *start, size_t length, int prot, int flags, int fd, off_t offset){
+ char *buf, *wp;
+ int rv, rlen;
+ if(flags & MAP_FIXED) return MAP_FAILED;
+ if(lseek(fd, SEEK_SET, offset) == -1) return MAP_FAILED;
+ if(!(buf = malloc(sizeof(int) * 3 + length))) return MAP_FAILED;
+ wp = buf;
+ *(int *)wp = fd;
+ wp += sizeof(int);
+ *(int *)wp = offset;
+ wp += sizeof(int);
+ *(int *)wp = prot;
+ wp += sizeof(int);
+ rlen = 0;
+ while((rv = read(fd, wp + rlen, length - rlen)) > 0){
+ rlen += rv;
+ }
+ if(rv == -1 || rlen != length){
+ free(buf);
+ return MAP_FAILED;
+ }
+ return wp;
+}
+
+
+int _qdbm_munmap(void *start, size_t length){
+ char *buf, *rp;
+ int fd, offset, prot, rv, wlen;
+ buf = (char *)start - sizeof(int) * 3;
+ rp = buf;
+ fd = *(int *)rp;
+ rp += sizeof(int);
+ offset = *(int *)rp;
+ rp += sizeof(int);
+ prot = *(int *)rp;
+ rp += sizeof(int);
+ if(prot & PROT_WRITE){
+ if(lseek(fd, offset, SEEK_SET) == -1){
+ free(buf);
+ return -1;
+ }
+ wlen = 0;
+ while(wlen < (int)length){
+ rv = write(fd, rp + wlen, length - wlen);
+ if(rv == -1){
+ if(errno == EINTR) continue;
+ free(buf);
+ return -1;
+ }
+ wlen += rv;
+ }
+ }
+ free(buf);
+ return 0;
+}
+
+
+int _qdbm_msync(const void *start, size_t length, int flags){
+ char *buf, *rp;
+ int fd, offset, prot, rv, wlen;
+ buf = (char *)start - sizeof(int) * 3;
+ rp = buf;
+ fd = *(int *)rp;
+ rp += sizeof(int);
+ offset = *(int *)rp;
+ rp += sizeof(int);
+ prot = *(int *)rp;
+ rp += sizeof(int);
+ if(prot & PROT_WRITE){
+ if(lseek(fd, offset, SEEK_SET) == -1) return -1;
+ wlen = 0;
+ while(wlen < (int)length){
+ rv = write(fd, rp + wlen, length - wlen);
+ if(rv == -1){
+ if(errno == EINTR) continue;
+ return -1;
+ }
+ wlen += rv;
+ }
+ }
+ return 0;
+}
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for reentrant time routines
+ *************************************************************************************************/
+
+
+#if defined(_SYS_LINUX_) || defined(_SYS_FREEBSD_) || defined(_SYS_OPENBSD_) || \
+ defined(_SYS_NETBSD_) || defined(_SYS_SUNOS_) || defined(_SYS_HPUX_) || \
+ defined(_SYS_MACOSX_) || defined(_SYS_CYGWIN_)
+
+
+struct tm *_qdbm_gmtime(const time_t *timep, struct tm *result){
+ return gmtime_r(timep, result);
+}
+
+
+struct tm *_qdbm_localtime(const time_t *timep, struct tm *result){
+ return localtime_r(timep, result);
+}
+
+
+# else
+
+
+struct tm *_qdbm_gmtime(const time_t *timep, struct tm *result){
+ return gmtime(timep);
+}
+
+
+struct tm *_qdbm_localtime(const time_t *timep, struct tm *result){
+ return localtime(timep);
+}
+
+
+# endif
+
+
+
+/*************************************************************************************************
+ * for systems without times
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_)
+
+
+clock_t _qdbm_times(struct tms *buf){
+ buf->tms_utime = clock();
+ buf->tms_stime = 0;
+ buf->tms_cutime = 0;
+ buf->tms_cstime = 0;
+ return 0;
+}
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for Win32
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_)
+
+
+#define WINLOCKWAIT 100
+
+
+int _qdbm_win32_fcntl(int fd, int cmd, struct flock *lock){
+ HANDLE fh;
+ DWORD opt;
+ OVERLAPPED ol;
+ fh = (HANDLE)_get_osfhandle(fd);
+ opt = (cmd == F_SETLK) ? LOCKFILE_FAIL_IMMEDIATELY : 0;
+ if(lock->l_type == F_WRLCK) opt |= LOCKFILE_EXCLUSIVE_LOCK;
+ memset(&ol, 0, sizeof(OVERLAPPED));
+ ol.Offset = INT_MAX;
+ ol.OffsetHigh = 0;
+ ol.hEvent = 0;
+ if(!LockFileEx(fh, opt, 0, 1, 0, &ol)){
+ if(GetLastError() == ERROR_CALL_NOT_IMPLEMENTED){
+ while(TRUE){
+ if(LockFile(fh, 0, 0, 1, 0)) return 0;
+ Sleep(WINLOCKWAIT);
+ }
+ }
+ return -1;
+ }
+ return 0;
+}
+
+
+#endif
+
+
+#if defined(_SYS_MSVC_)
+
+
+DIR *_qdbm_win32_opendir(const char *name){
+ char expr[8192];
+ int len;
+ DIR *dir;
+ HANDLE fh;
+ WIN32_FIND_DATA data;
+ len = strlen(name);
+ if(len > 0 && name[len-1] == MYPATHCHR){
+ sprintf(expr, "%s*", name);
+ } else {
+ sprintf(expr, "%s%c*", name, MYPATHCHR);
+ }
+ if((fh = FindFirstFile(expr, &data)) == INVALID_HANDLE_VALUE) return NULL;
+ if(!(dir = malloc(sizeof(DIR)))){
+ FindClose(fh);
+ return NULL;
+ }
+ dir->fh = fh;
+ dir->data = data;
+ dir->first = TRUE;
+ return dir;
+}
+
+
+int _qdbm_win32_closedir(DIR *dir){
+ if(!FindClose(dir->fh)){
+ free(dir);
+ return -1;
+ }
+ free(dir);
+ return 0;
+}
+
+
+struct dirent *_qdbm_win32_readdir(DIR *dir){
+ if(dir->first){
+ sprintf(dir->de.d_name, "%s", dir->data.cFileName);
+ dir->first = FALSE;
+ return &(dir->de);
+ }
+ if(!FindNextFile(dir->fh, &(dir->data))) return NULL;
+ sprintf(dir->de.d_name, "%s", dir->data.cFileName);
+ return &(dir->de);
+}
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for checking information of the system
+ *************************************************************************************************/
+
+
+#if defined(_SYS_LINUX_)
+
+
+int _qdbm_vmemavail(size_t size){
+ char buf[4096], *rp;
+ int fd, rv, bsiz;
+ double avail;
+ if((fd = open("/proc/meminfo", O_RDONLY, 00644)) == -1) return TRUE;
+ rv = TRUE;
+ if((bsiz = read(fd, buf, sizeof(buf) - 1)) > 0){
+ buf[bsiz] = '\0';
+ avail = -1;
+ if((rp = strstr(buf, "MemFree:")) != NULL){
+ rp = strchr(rp, ':') + 1;
+ avail = strtod(rp, NULL) * 1024.0;
+ if((rp = strstr(buf, "SwapFree:")) != NULL){
+ rp = strchr(rp, ':') + 1;
+ avail += strtod(rp, NULL) * 1024.0;
+ }
+ if(size >= avail) rv = FALSE;
+ }
+ }
+ close(fd);
+ return rv;
+}
+
+
+#elif defined(_SYS_MSVC_) || defined(_SYS_MINGW_) || defined(_SYS_CYGWIN_)
+
+
+int _qdbm_vmemavail(size_t size){
+ MEMORYSTATUS sbuf;
+ sbuf.dwLength = sizeof(MEMORYSTATUS);
+ GlobalMemoryStatus(&sbuf);
+ return size < sbuf.dwAvailVirtual;
+}
+
+
+#else
+
+
+int _qdbm_vmemavail(size_t size){
+ return TRUE;
+}
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for ZLIB
+ *************************************************************************************************/
+
+
+#if defined(MYZLIB)
+
+
+#include <zlib.h>
+
+#define ZLIBBUFSIZ 8192
+
+
+static char *_qdbm_deflate_impl(const char *ptr, int size, int *sp, int mode);
+static char *_qdbm_inflate_impl(const char *ptr, int size, int *sp, int mode);
+static unsigned int _qdbm_getcrc_impl(const char *ptr, int size);
+
+
+char *(*_qdbm_deflate)(const char *, int, int *, int) = _qdbm_deflate_impl;
+char *(*_qdbm_inflate)(const char *, int, int *, int) = _qdbm_inflate_impl;
+unsigned int (*_qdbm_getcrc)(const char *, int) = _qdbm_getcrc_impl;
+
+
+static char *_qdbm_deflate_impl(const char *ptr, int size, int *sp, int mode){
+ z_stream zs;
+ char *buf, *swap;
+ unsigned char obuf[ZLIBBUFSIZ];
+ int rv, asiz, bsiz, osiz;
+ if(size < 0) size = strlen(ptr);
+ zs.zalloc = Z_NULL;
+ zs.zfree = Z_NULL;
+ zs.opaque = Z_NULL;
+ switch(mode){
+ case _QDBM_ZMRAW:
+ if(deflateInit2(&zs, 5, Z_DEFLATED, -15, 7, Z_DEFAULT_STRATEGY) != Z_OK)
+ return NULL;
+ break;
+ case _QDBM_ZMGZIP:
+ if(deflateInit2(&zs, 6, Z_DEFLATED, 15 + 16, 9, Z_DEFAULT_STRATEGY) != Z_OK)
+ return NULL;
+ break;
+ default:
+ if(deflateInit2(&zs, 6, Z_DEFLATED, 15, 8, Z_DEFAULT_STRATEGY) != Z_OK)
+ return NULL;
+ break;
+ }
+ asiz = size + 16;
+ if(asiz < ZLIBBUFSIZ) asiz = ZLIBBUFSIZ;
+ if(!(buf = malloc(asiz))){
+ deflateEnd(&zs);
+ return NULL;
+ }
+ bsiz = 0;
+ zs.next_in = (unsigned char *)ptr;
+ zs.avail_in = size;
+ zs.next_out = obuf;
+ zs.avail_out = ZLIBBUFSIZ;
+ while((rv = deflate(&zs, Z_FINISH)) == Z_OK){
+ osiz = ZLIBBUFSIZ - zs.avail_out;
+ if(bsiz + osiz > asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ deflateEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ zs.next_out = obuf;
+ zs.avail_out = ZLIBBUFSIZ;
+ }
+ if(rv != Z_STREAM_END){
+ free(buf);
+ deflateEnd(&zs);
+ return NULL;
+ }
+ osiz = ZLIBBUFSIZ - zs.avail_out;
+ if(bsiz + osiz + 1 > asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ deflateEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ buf[bsiz] = '\0';
+ if(mode == _QDBM_ZMRAW) bsiz++;
+ *sp = bsiz;
+ deflateEnd(&zs);
+ return buf;
+}
+
+
+static char *_qdbm_inflate_impl(const char *ptr, int size, int *sp, int mode){
+ z_stream zs;
+ char *buf, *swap;
+ unsigned char obuf[ZLIBBUFSIZ];
+ int rv, asiz, bsiz, osiz;
+ zs.zalloc = Z_NULL;
+ zs.zfree = Z_NULL;
+ zs.opaque = Z_NULL;
+ switch(mode){
+ case _QDBM_ZMRAW:
+ if(inflateInit2(&zs, -15) != Z_OK) return NULL;
+ break;
+ case _QDBM_ZMGZIP:
+ if(inflateInit2(&zs, 15 + 16) != Z_OK) return NULL;
+ break;
+ default:
+ if(inflateInit2(&zs, 15) != Z_OK) return NULL;
+ break;
+ }
+ asiz = size * 2 + 16;
+ if(asiz < ZLIBBUFSIZ) asiz = ZLIBBUFSIZ;
+ if(!(buf = malloc(asiz))){
+ inflateEnd(&zs);
+ return NULL;
+ }
+ bsiz = 0;
+ zs.next_in = (unsigned char *)ptr;
+ zs.avail_in = size;
+ zs.next_out = obuf;
+ zs.avail_out = ZLIBBUFSIZ;
+ while((rv = inflate(&zs, Z_NO_FLUSH)) == Z_OK){
+ osiz = ZLIBBUFSIZ - zs.avail_out;
+ if(bsiz + osiz >= asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ inflateEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ zs.next_out = obuf;
+ zs.avail_out = ZLIBBUFSIZ;
+ }
+ if(rv != Z_STREAM_END){
+ free(buf);
+ inflateEnd(&zs);
+ return NULL;
+ }
+ osiz = ZLIBBUFSIZ - zs.avail_out;
+ if(bsiz + osiz >= asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ inflateEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ buf[bsiz] = '\0';
+ if(sp) *sp = bsiz;
+ inflateEnd(&zs);
+ return buf;
+}
+
+
+static unsigned int _qdbm_getcrc_impl(const char *ptr, int size){
+ int crc;
+ if(size < 0) size = strlen(ptr);
+ crc = crc32(0, Z_NULL, 0);
+ return crc32(crc, (unsigned char *)ptr, size);
+}
+
+
+#else
+
+
+char *(*_qdbm_deflate)(const char *, int, int *, int) = NULL;
+char *(*_qdbm_inflate)(const char *, int, int *, int) = NULL;
+unsigned int (*_qdbm_getcrc)(const char *, int) = NULL;
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for LZO
+ *************************************************************************************************/
+
+
+#if defined(MYLZO)
+
+
+#include <lzo/lzo1x.h>
+
+
+static char *_qdbm_lzoencode_impl(const char *ptr, int size, int *sp);
+static char *_qdbm_lzodecode_impl(const char *ptr, int size, int *sp);
+
+
+int _qdbm_lzo_init = FALSE;
+char *(*_qdbm_lzoencode)(const char *, int, int *) = _qdbm_lzoencode_impl;
+char *(*_qdbm_lzodecode)(const char *, int, int *) = _qdbm_lzodecode_impl;
+
+
+static char *_qdbm_lzoencode_impl(const char *ptr, int size, int *sp){
+ char wrkmem[LZO1X_1_MEM_COMPRESS];
+ lzo_bytep buf;
+ lzo_uint bsiz;
+ if(!_qdbm_lzo_init){
+ if(lzo_init() != LZO_E_OK) return NULL;
+ _qdbm_lzo_init = TRUE;
+ }
+ if(size < 0) size = strlen(ptr);
+ if(!(buf = malloc(size + size / 16 + 80))) return NULL;
+ if(lzo1x_1_compress((lzo_bytep)ptr, size, buf, &bsiz, wrkmem) != LZO_E_OK){
+ free(buf);
+ return NULL;
+ }
+ buf[bsiz] = '\0';
+ *sp = bsiz;
+ return (char *)buf;
+}
+
+
+static char *_qdbm_lzodecode_impl(const char *ptr, int size, int *sp){
+ lzo_bytep buf;
+ lzo_uint bsiz;
+ int rat, rv;
+ if(!_qdbm_lzo_init){
+ if(lzo_init() != LZO_E_OK) return NULL;
+ _qdbm_lzo_init = TRUE;
+ }
+ rat = 6;
+ while(TRUE){
+ bsiz = (size + 256) * rat + 3;
+ if(!(buf = malloc(bsiz + 1))) return NULL;
+ rv = lzo1x_decompress_safe((lzo_bytep)(ptr), size, buf, &bsiz, NULL);
+ if(rv == LZO_E_OK){
+ break;
+ } else if(rv == LZO_E_OUTPUT_OVERRUN){
+ free(buf);
+ rat *= 2;
+ } else {
+ free(buf);
+ return NULL;
+ }
+ }
+ buf[bsiz] = '\0';
+ if(sp) *sp = bsiz;
+ return (char *)buf;
+}
+
+
+#else
+
+
+char *(*_qdbm_lzoencode)(const char *, int, int *) = NULL;
+char *(*_qdbm_lzodecode)(const char *, int, int *) = NULL;
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for BZIP2
+ *************************************************************************************************/
+
+
+#if defined(MYBZIP)
+
+
+#include <bzlib.h>
+
+#define BZIPBUFSIZ 8192
+
+
+static char *_qdbm_bzencode_impl(const char *ptr, int size, int *sp);
+static char *_qdbm_bzdecode_impl(const char *ptr, int size, int *sp);
+
+
+char *(*_qdbm_bzencode)(const char *, int, int *) = _qdbm_bzencode_impl;
+char *(*_qdbm_bzdecode)(const char *, int, int *) = _qdbm_bzdecode_impl;
+
+
+static char *_qdbm_bzencode_impl(const char *ptr, int size, int *sp){
+ bz_stream zs;
+ char *buf, *swap, obuf[BZIPBUFSIZ];
+ int rv, asiz, bsiz, osiz;
+ if(size < 0) size = strlen(ptr);
+ zs.bzalloc = NULL;
+ zs.bzfree = NULL;
+ zs.opaque = NULL;
+ if(BZ2_bzCompressInit(&zs, 9, 0, 30) != BZ_OK) return NULL;
+ asiz = size + 16;
+ if(asiz < BZIPBUFSIZ) asiz = BZIPBUFSIZ;
+ if(!(buf = malloc(asiz))){
+ BZ2_bzCompressEnd(&zs);
+ return NULL;
+ }
+ bsiz = 0;
+ zs.next_in = (char *)ptr;
+ zs.avail_in = size;
+ zs.next_out = obuf;
+ zs.avail_out = BZIPBUFSIZ;
+ while((rv = BZ2_bzCompress(&zs, BZ_FINISH)) == BZ_FINISH_OK){
+ osiz = BZIPBUFSIZ - zs.avail_out;
+ if(bsiz + osiz > asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ BZ2_bzCompressEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ zs.next_out = obuf;
+ zs.avail_out = BZIPBUFSIZ;
+ }
+ if(rv != BZ_STREAM_END){
+ free(buf);
+ BZ2_bzCompressEnd(&zs);
+ return NULL;
+ }
+ osiz = BZIPBUFSIZ - zs.avail_out;
+ if(bsiz + osiz + 1 > asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ BZ2_bzCompressEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ buf[bsiz] = '\0';
+ *sp = bsiz;
+ BZ2_bzCompressEnd(&zs);
+ return buf;
+}
+
+
+static char *_qdbm_bzdecode_impl(const char *ptr, int size, int *sp){
+ bz_stream zs;
+ char *buf, *swap, obuf[BZIPBUFSIZ];
+ int rv, asiz, bsiz, osiz;
+ zs.bzalloc = NULL;
+ zs.bzfree = NULL;
+ zs.opaque = NULL;
+ if(BZ2_bzDecompressInit(&zs, 0, 0) != BZ_OK) return NULL;
+ asiz = size * 2 + 16;
+ if(asiz < BZIPBUFSIZ) asiz = BZIPBUFSIZ;
+ if(!(buf = malloc(asiz))){
+ BZ2_bzDecompressEnd(&zs);
+ return NULL;
+ }
+ bsiz = 0;
+ zs.next_in = (char *)ptr;
+ zs.avail_in = size;
+ zs.next_out = obuf;
+ zs.avail_out = BZIPBUFSIZ;
+ while((rv = BZ2_bzDecompress(&zs)) == BZ_OK){
+ osiz = BZIPBUFSIZ - zs.avail_out;
+ if(bsiz + osiz >= asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ BZ2_bzDecompressEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ zs.next_out = obuf;
+ zs.avail_out = BZIPBUFSIZ;
+ }
+ if(rv != BZ_STREAM_END){
+ free(buf);
+ BZ2_bzDecompressEnd(&zs);
+ return NULL;
+ }
+ osiz = BZIPBUFSIZ - zs.avail_out;
+ if(bsiz + osiz >= asiz){
+ asiz = asiz * 2 + osiz;
+ if(!(swap = realloc(buf, asiz))){
+ free(buf);
+ BZ2_bzDecompressEnd(&zs);
+ return NULL;
+ }
+ buf = swap;
+ }
+ memcpy(buf + bsiz, obuf, osiz);
+ bsiz += osiz;
+ buf[bsiz] = '\0';
+ if(sp) *sp = bsiz;
+ BZ2_bzDecompressEnd(&zs);
+ return buf;
+}
+
+
+#else
+
+
+char *(*_qdbm_bzencode)(const char *, int, int *) = NULL;
+char *(*_qdbm_bzdecode)(const char *, int, int *) = NULL;
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for ICONV
+ *************************************************************************************************/
+
+
+#if defined(MYICONV)
+
+
+#include <iconv.h>
+
+#define ICONVCHECKSIZ 32768
+#define ICONVMISSMAX 256
+#define ICONVALLWRAT 0.001
+
+
+static char *_qdbm_iconv_impl(const char *ptr, int size,
+ const char *icode, const char *ocode, int *sp, int *mp);
+static const char *_qdbm_encname_impl(const char *ptr, int size);
+static int _qdbm_encmiss(const char *ptr, int size, const char *icode, const char *ocode);
+
+
+char *(*_qdbm_iconv)(const char *, int, const char *, const char *,
+ int *, int *) = _qdbm_iconv_impl;
+const char *(*_qdbm_encname)(const char *, int) = _qdbm_encname_impl;
+
+
+static char *_qdbm_iconv_impl(const char *ptr, int size,
+ const char *icode, const char *ocode, int *sp, int *mp){
+ iconv_t ic;
+ char *obuf, *wp, *rp;
+ size_t isiz, osiz;
+ int miss;
+ if(size < 0) size = strlen(ptr);
+ isiz = size;
+ if((ic = iconv_open(ocode, icode)) == (iconv_t)-1) return NULL;
+ osiz = isiz * 5;
+ if(!(obuf = malloc(osiz + 1))){
+ iconv_close(ic);
+ return NULL;
+ }
+ wp = obuf;
+ rp = (char *)ptr;
+ miss = 0;
+ while(isiz > 0){
+ if(iconv(ic, (void *)&rp, &isiz, &wp, &osiz) == -1){
+ if(errno == EILSEQ && (*rp == 0x5c || *rp == 0x7e)){
+ *wp = *rp;
+ wp++;
+ rp++;
+ isiz--;
+ } else if(errno == EILSEQ || errno == EINVAL){
+ rp++;
+ isiz--;
+ miss++;
+ } else {
+ break;
+ }
+ }
+ }
+ *wp = '\0';
+ if(iconv_close(ic) == -1){
+ free(obuf);
+ return NULL;
+ }
+ if(sp) *sp = wp - obuf;
+ if(mp) *mp = miss;
+ return obuf;
+}
+
+
+static const char *_qdbm_encname_impl(const char *ptr, int size){
+ const char *hypo;
+ int i, miss, cr;
+ if(size < 0) size = strlen(ptr);
+ if(size > ICONVCHECKSIZ) size = ICONVCHECKSIZ;
+ if(size >= 2 && (!memcmp(ptr, "\xfe\xff", 2) || !memcmp(ptr, "\xff\xfe", 2))) return "UTF-16";
+ for(i = 0; i < size - 1; i += 2){
+ if(ptr[i] == 0 && ptr[i+1] != 0) return "UTF-16BE";
+ if(ptr[i+1] == 0 && ptr[i] != 0) return "UTF-16LE";
+ }
+ for(i = 0; i < size - 3; i++){
+ if(ptr[i] == 0x1b){
+ i++;
+ if(ptr[i] == '(' && strchr("BJHI", ptr[i+1])) return "ISO-2022-JP";
+ if(ptr[i] == '$' && strchr("@B(", ptr[i+1])) return "ISO-2022-JP";
+ }
+ }
+ if(_qdbm_encmiss(ptr, size, "US-ASCII", "UTF-16BE") < 1) return "US-ASCII";
+ if(_qdbm_encmiss(ptr, size, "UTF-8", "UTF-16BE") < 1) return "UTF-8";
+ hypo = NULL;
+ cr = FALSE;
+ for(i = 0; i < size; i++){
+ if(ptr[i] == 0xd){
+ cr = TRUE;
+ break;
+ }
+ }
+ if(cr){
+ if((miss = _qdbm_encmiss(ptr, size, "Shift_JIS", "EUC-JP")) < 1) return "Shift_JIS";
+ if(!hypo && miss / (double)size <= ICONVALLWRAT) hypo = "Shift_JIS";
+ if((miss = _qdbm_encmiss(ptr, size, "EUC-JP", "UTF-16BE")) < 1) return "EUC-JP";
+ if(!hypo && miss / (double)size <= ICONVALLWRAT) hypo = "EUC-JP";
+ } else {
+ if((miss = _qdbm_encmiss(ptr, size, "EUC-JP", "UTF-16BE")) < 1) return "EUC-JP";
+ if(!hypo && miss / (double)size <= ICONVALLWRAT) hypo = "EUC-JP";
+ if((miss = _qdbm_encmiss(ptr, size, "Shift_JIS", "EUC-JP")) < 1) return "Shift_JIS";
+ if(!hypo && miss / (double)size <= ICONVALLWRAT) hypo = "Shift_JIS";
+ }
+ if((miss = _qdbm_encmiss(ptr, size, "UTF-8", "UTF-16BE")) < 1) return "UTF-8";
+ if(!hypo && miss / (double)size <= ICONVALLWRAT) hypo = "UTF-8";
+ if((miss = _qdbm_encmiss(ptr, size, "CP932", "UTF-16BE")) < 1) return "CP932";
+ if(!hypo && miss / (double)size <= ICONVALLWRAT) hypo = "CP932";
+ return hypo ? hypo : "ISO-8859-1";
+}
+
+
+static int _qdbm_encmiss(const char *ptr, int size, const char *icode, const char *ocode){
+ iconv_t ic;
+ char obuf[ICONVCHECKSIZ], *wp, *rp;
+ size_t isiz, osiz;
+ int miss;
+ isiz = size;
+ if((ic = iconv_open(ocode, icode)) == (iconv_t)-1) return ICONVMISSMAX;
+ miss = 0;
+ rp = (char *)ptr;
+ while(isiz > 0){
+ osiz = ICONVCHECKSIZ;
+ wp = obuf;
+ if(iconv(ic, (void *)&rp, &isiz, &wp, &osiz) == -1){
+ if(errno == EILSEQ || errno == EINVAL){
+ rp++;
+ isiz--;
+ miss++;
+ if(miss >= ICONVMISSMAX) break;
+ } else {
+ break;
+ }
+ }
+ }
+ if(iconv_close(ic) == -1) return ICONVMISSMAX;
+ return miss;
+}
+
+
+#else
+
+
+char *(*_qdbm_iconv)(const char *, int, const char *, const char *, int *, int *) = NULL;
+const char *(*_qdbm_encname)(const char *, int) = NULL;
+
+
+#endif
+
+
+
+/*************************************************************************************************
+ * common settings
+ *************************************************************************************************/
+
+
+int _qdbm_dummyfunc(void){
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/myconf.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/myconf.h
===================================================================
--- box/trunk/qdbm/myconf.h (rev 0)
+++ box/trunk/qdbm/myconf.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,593 @@
+/*************************************************************************************************
+ * System configurations for QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _MYCONF_H /* duplication check */
+#define _MYCONF_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+
+/*************************************************************************************************
+ * system discrimination
+ *************************************************************************************************/
+
+
+#if defined(__linux__)
+
+#define _SYS_LINUX_
+#define _QDBM_SYSNAME "Linux"
+
+#elif defined(__FreeBSD__)
+
+#define _SYS_FREEBSD_
+#define _QDBM_SYSNAME "FreeBSD"
+
+#elif defined(__NetBSD__)
+
+#define _SYS_NETBSD_
+#define _QDBM_SYSNAME "NetBSD"
+
+#elif defined(__OpenBSD__)
+
+#define _SYS_OPENBSD_
+#define _QDBM_SYSNAME "OpenBSD"
+
+#elif defined(__sun__)
+
+#define _SYS_SUNOS_
+#define _QDBM_SYSNAME "SunOS"
+
+#elif defined(__hpux)
+
+#define _SYS_HPUX_
+#define _QDBM_SYSNAME "HP-UX"
+
+#elif defined(__osf)
+
+#define _SYS_TRU64_
+#define _QDBM_SYSNAME "Tru64"
+
+#elif defined(_AIX)
+
+#define _SYS_AIX_
+#define _QDBM_SYSNAME "AIX"
+
+#elif defined(__APPLE__) && defined(__MACH__)
+
+#define _SYS_MACOSX_
+#define _QDBM_SYSNAME "Mac OS X"
+
+#elif defined(_MSC_VER)
+
+#define _SYS_MSVC_
+#define _QDBM_SYSNAME "Windows (VC++)"
+
+#elif defined(_WIN32)
+
+#define _SYS_MINGW_
+#define _QDBM_SYSNAME "Windows (MinGW)"
+
+#elif defined(__CYGWIN__)
+
+#define _SYS_CYGWIN_
+#define _QDBM_SYSNAME "Windows (Cygwin)"
+
+#elif defined(__riscos__) || defined(__riscos)
+
+#define _SYS_RISCOS_
+#define _QDBM_SYSNAME "RISC OS"
+
+#else
+
+#define _SYS_GENERIC_
+#define _QDBM_SYSNAME "Generic"
+
+#endif
+
+
+
+/*************************************************************************************************
+ * general headers
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_)
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <limits.h>
+#include <time.h>
+#include <assert.h>
+#include <stdarg.h>
+#include <errno.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <direct.h>
+#include <windows.h>
+#include <io.h>
+
+#elif defined(_SYS_MINGW_)
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <limits.h>
+#include <time.h>
+#include <assert.h>
+#include <stdarg.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <dirent.h>
+#include <windows.h>
+#include <io.h>
+
+#elif defined(_SYS_CYGWIN_)
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <limits.h>
+#include <time.h>
+#include <assert.h>
+#include <stdarg.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/mman.h>
+#include <sys/times.h>
+#include <fcntl.h>
+#include <dirent.h>
+#include <windows.h>
+#include <io.h>
+
+#else
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <limits.h>
+#include <time.h>
+#include <assert.h>
+#include <stdarg.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/mman.h>
+#include <sys/times.h>
+#include <fcntl.h>
+#include <dirent.h>
+
+#endif
+
+
+
+/*************************************************************************************************
+ * notation of filesystems
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_)
+
+#define MYPATHCHR '\\'
+#define MYPATHSTR "\\"
+#define MYEXTCHR '.'
+#define MYEXTSTR "."
+#define MYCDIRSTR "."
+#define MYPDIRSTR ".."
+
+#elif defined(_SYS_RISCOS_)
+
+#define MYPATHCHR '.'
+#define MYPATHSTR "."
+#define MYEXTCHR '/'
+#define MYEXTSTR "/"
+#define MYCDIRSTR "@"
+#define MYPDIRSTR "^"
+
+#else
+
+#define MYPATHCHR '/'
+#define MYPATHSTR "/"
+#define MYEXTCHR '.'
+#define MYEXTSTR "."
+#define MYCDIRSTR "."
+#define MYPDIRSTR ".."
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for dosish filesystems
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_) || defined(_SYS_CYGWIN_)
+
+#undef UNICODE
+#undef open
+
+#define \
+ open(pathname, flags, mode) \
+ open(pathname, flags | O_BINARY, mode)
+
+#define \
+ lstat(pathname, buf) \
+ _qdbm_win32_lstat(pathname, buf)
+
+int _qdbm_win32_lstat(const char *pathname, struct stat *buf);
+
+#else
+
+#undef O_BINARY
+#undef O_TEXT
+#undef setmode
+
+#define O_BINARY 0
+#define O_TEXT 1
+
+#define \
+ setmode(fd, mode) \
+ (O_BINARY)
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for POSIX thread
+ *************************************************************************************************/
+
+
+#if defined(MYPTHREAD)
+
+#define _qdbm_ptsafe TRUE
+
+void *_qdbm_settsd(void *ptr, int size, const void *initval);
+
+#else
+
+#define _qdbm_ptsafe FALSE
+
+#define \
+ _qdbm_settsd(ptr, size, initval) \
+ (NULL)
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for systems without file locking
+ *************************************************************************************************/
+
+
+#if defined(_SYS_RISCOS_) || defined(MYNOLOCK)
+
+#undef fcntl
+
+#define \
+ fcntl(fd, cmd, lock) \
+ (0)
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for systems without mmap
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_) || \
+ defined(_SYS_FREEBSD_) || defined(_SYS_NETBSD_) || defined(_SYS_OPENBSD_) || \
+ defined(_SYS_AIX_) || defined(_SYS_RISCOS_) || defined(MYNOMMAP)
+
+#undef PROT_EXEC
+#undef PROT_READ
+#undef PROT_WRITE
+#undef PROT_NONE
+#undef MAP_FIXED
+#undef MAP_SHARED
+#undef MAP_PRIVATE
+#undef MAP_FAILED
+#undef MS_ASYNC
+#undef MS_SYNC
+#undef MS_INVALIDATE
+#undef mmap
+#undef munmap
+#undef msync
+#undef mflush
+
+#define PROT_EXEC (1 << 0)
+#define PROT_READ (1 << 1)
+#define PROT_WRITE (1 << 2)
+#define PROT_NONE (1 << 3)
+#define MAP_FIXED 1
+#define MAP_SHARED 2
+#define MAP_PRIVATE 3
+#define MAP_FAILED ((void *)-1)
+#define MS_ASYNC (1 << 0)
+#define MS_SYNC (1 << 1)
+#define MS_INVALIDATE (1 << 2)
+
+#define \
+ mmap(start, length, prot, flags, fd, offset) \
+ _qdbm_mmap(start, length, prot, flags, fd, offset)
+
+#define \
+ munmap(start, length) \
+ _qdbm_munmap(start, length)
+
+#define \
+ msync(start, length, flags) \
+ _qdbm_msync(start, length, flags)
+
+#define \
+ mflush(start, length, flags) \
+ _qdbm_msync(start, length, flags)
+
+void *_qdbm_mmap(void *start, size_t length, int prot, int flags, int fd, off_t offset);
+int _qdbm_munmap(void *start, size_t length);
+int _qdbm_msync(const void *start, size_t length, int flags);
+
+#else
+
+#undef mflush
+#define \
+ mflush(start, length, flags) \
+ (0)
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for reentrant time routines
+ *************************************************************************************************/
+
+
+struct tm *_qdbm_gmtime(const time_t *timep, struct tm *result);
+struct tm *_qdbm_localtime(const time_t *timep, struct tm *result);
+
+
+
+/*************************************************************************************************
+ * for systems without times
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_)
+
+#undef times
+#undef sysconf
+
+struct tms {
+ clock_t tms_utime;
+ clock_t tms_stime;
+ clock_t tms_cutime;
+ clock_t tms_cstime;
+};
+
+#define \
+ times(buf) \
+ _qdbm_times(buf)
+
+#define \
+ sysconf(name) \
+ (CLOCKS_PER_SEC)
+
+clock_t _qdbm_times(struct tms *buf);
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for Win32
+ *************************************************************************************************/
+
+
+#if defined(_SYS_MSVC_) || defined(_SYS_MINGW_)
+
+#undef F_WRLCK
+#undef F_RDLCK
+#undef F_SETLK
+#undef F_SETLKW
+#undef fcntl
+#undef ftruncate
+#undef fsync
+#undef mkdir
+#undef rename
+
+#define F_WRLCK 0
+#define F_RDLCK 1
+#define F_SETLK 0
+#define F_SETLKW 1
+
+struct flock {
+ int l_type;
+ int l_whence;
+ int l_start;
+ int l_len;
+ int l_pid;
+};
+
+#define \
+ fcntl(fd, cmd, lock) \
+ _qdbm_win32_fcntl(fd, cmd, lock)
+
+#define \
+ ftruncate(fd, length) \
+ _chsize(fd, length)
+
+#define \
+ fsync(fd) \
+ (0)
+
+#define \
+ mkdir(pathname, mode) \
+ mkdir(pathname)
+
+#define \
+ rename(oldpath, newpath) \
+ (unlink(newpath), rename(oldpath, newpath))
+
+int _qdbm_win32_fcntl(int fd, int cmd, struct flock *lock);
+
+#endif
+
+
+#if defined(_SYS_MSVC_)
+
+#undef S_ISDIR
+#undef S_ISREG
+#undef opendir
+#undef closedir
+#undef readdir
+
+#define S_ISDIR(x) (x & _S_IFDIR)
+#define S_ISREG(x) (x & _S_IFREG)
+
+struct dirent {
+ char d_name[1024];
+};
+
+typedef struct {
+ HANDLE fh;
+ WIN32_FIND_DATA data;
+ struct dirent de;
+ int first;
+} DIR;
+
+#define \
+ opendir(name) \
+ _qdbm_win32_opendir(name)
+
+#define \
+ closedir(dir) \
+ _qdbm_win32_closedir(dir)
+
+#define \
+ readdir(dir) \
+ _qdbm_win32_readdir(dir)
+
+DIR *_qdbm_win32_opendir(const char *name);
+
+int _qdbm_win32_closedir(DIR *dir);
+
+struct dirent *_qdbm_win32_readdir(DIR *dir);
+
+#endif
+
+
+
+/*************************************************************************************************
+ * for checking information of the system
+ *************************************************************************************************/
+
+
+int _qdbm_vmemavail(size_t size);
+
+
+
+/*************************************************************************************************
+ * for ZLIB
+ *************************************************************************************************/
+
+
+enum {
+ _QDBM_ZMZLIB,
+ _QDBM_ZMRAW,
+ _QDBM_ZMGZIP
+};
+
+
+extern char *(*_qdbm_deflate)(const char *, int, int *, int);
+
+extern char *(*_qdbm_inflate)(const char *, int, int *, int);
+
+extern unsigned int (*_qdbm_getcrc)(const char *, int);
+
+
+
+/*************************************************************************************************
+ * for LZO
+ *************************************************************************************************/
+
+
+extern char *(*_qdbm_lzoencode)(const char *, int, int *);
+
+extern char *(*_qdbm_lzodecode)(const char *, int, int *);
+
+
+
+/*************************************************************************************************
+ * for BZIP2
+ *************************************************************************************************/
+
+
+extern char *(*_qdbm_bzencode)(const char *, int, int *);
+
+extern char *(*_qdbm_bzdecode)(const char *, int, int *);
+
+
+
+/*************************************************************************************************
+ * for ICONV
+ *************************************************************************************************/
+
+
+extern char *(*_qdbm_iconv)(const char *, int, const char *, const char *, int *, int *);
+
+extern const char *(*_qdbm_encname)(const char *, int);
+
+
+
+/*************************************************************************************************
+ * common settings
+ *************************************************************************************************/
+
+
+#undef TRUE
+#define TRUE 1
+#undef FALSE
+#define FALSE 0
+
+#define sizeof(a) ((int)sizeof(a))
+
+int _qdbm_dummyfunc(void);
+
+
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/myconf.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/odeum.c
===================================================================
--- box/trunk/qdbm/odeum.c (rev 0)
+++ box/trunk/qdbm/odeum.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,2090 @@
+/*************************************************************************************************
+ * Implementation of Odeum
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "odeum.h"
+#include "myconf.h"
+
+#define OD_NAMEMAX 256 /* max size of a database name */
+#define OD_DIRMODE 00755 /* permission of a creating directory */
+#define OD_PATHBUFSIZ 1024 /* size of a path buffer */
+#define OD_NUMBUFSIZ 32 /* size of a buffer for a number */
+#define OD_MAPPBNUM 127 /* bucket size of a petit map handle */
+#define OD_DOCSNAME "docs" /* name of the database for documents */
+#define OD_INDEXNAME "index" /* name of the database for inverted index */
+#define OD_RDOCSNAME "rdocs" /* name of the database for reverse dictionary */
+#define OD_DOCSBNUM 2039 /* initial bucket number of document database */
+#define OD_DOCSDNUM 17 /* division number of document database */
+#define OD_DOCSALIGN -4 /* alignment of document database */
+#define OD_DOCSFBP 32 /* size of free block pool of document database */
+#define OD_INDEXBNUM 32749 /* initial bucket number of inverted index */
+#define OD_INDEXDNUM 7 /* division number of inverted index */
+#define OD_INDEXALIGN -2 /* alignment of inverted index */
+#define OD_INDEXFBP 32 /* size of free block pool of inverted index */
+#define OD_RDOCSLRM 81 /* records in a leaf node of reverse dictionary */
+#define OD_RDOCSNIM 192 /* records in a non-leaf node of reverse dictionary */
+#define OD_RDOCSLCN 128 /* number of leaf cache of reverse dictionary */
+#define OD_RDOCSNCN 32 /* number of non-leaf cache of reverse dictionary */
+#define OD_CACHEBNUM 262139 /* number of buckets for dirty buffers */
+#define OD_CACHESIZ 8388608 /* max bytes to use memory for dirty buffers */
+#define OD_CFLIVERAT 0.8 /* ratio of usable cache region */
+#define OD_CFBEGSIZ 2048 /* beginning size of flushing frequent words */
+#define OD_CFENDSIZ 64 /* lower limit of flushing frequent words */
+#define OD_CFRFRAT 0.2 /* ratio of flushing rare words a time */
+#define OD_OTCBBUFSIZ 1024 /* size of a buffer for call back functions */
+#define OD_OTPERWORDS 10000 /* frequency of call back in merging index */
+#define OD_OTPERDOCS 1000 /* frequency of call back in merging docs */
+#define OD_MDBRATIO 2.5 /* ratio of bucket number and document number */
+#define OD_MIBRATIO 1.5 /* ratio of bucket number and word number */
+#define OD_MIARATIO 0.75 /* ratio of alignment to the first words */
+#define OD_MIWUNIT 32 /* writing unit of merging inverted index */
+#define OD_DMAXEXPR "dmax" /* key of max number of the document ID */
+#define OD_DNUMEXPR "dnum" /* key of number of the documents */
+#define OD_URIEXPR "1" /* map key of URI */
+#define OD_ATTRSEXPR "2" /* map key of attributes */
+#define OD_NWORDSEXPR "3" /* map key of normal words */
+#define OD_AWORDSEXPR "4" /* map key of as-is words */
+#define OD_WTOPRATE 0.1 /* ratio of top words */
+#define OD_WTOPBONUS 5000 /* bonus points of top words */
+#define OD_KEYCRATIO 1.75 /* ratio of number to max of keyword candidates */
+#define OD_WOCCRPOINT 10000 /* points per occurence */
+#define OD_SPACECHARS "\t\n\v\f\r " /* space characters */
+#define OD_DELIMCHARS "!\"#$%&'()*/<=>?[\\]^`{|}~" /* delimiter characters */
+#define OD_GLUECHARS "+,-.:;@" /* glueing characters */
+#define OD_MAXWORDLEN 48 /* max length of a word */
+
+typedef struct { /* type of structure for word counting */
+ const char *word; /* pointer to the word */
+ int num; /* frequency of the word */
+} ODWORD;
+
+enum { /* enumeration for events binded to each character */
+ OD_EVWORD, /* word */
+ OD_EVSPACE, /* space */
+ OD_EVDELIM, /* delimiter */
+ OD_EVGLUE /* glue */
+};
+
+
+/* private global variables */
+int odindexbnum = OD_INDEXBNUM;
+int odindexdnum = OD_INDEXDNUM;
+int odcachebnum = OD_CACHEBNUM;
+int odcachesiz = OD_CACHESIZ;
+void (*odotcb)(const char *, ODEUM *, const char *) = NULL;
+
+
+/* private function prototypes */
+static ODEUM *odopendb(const char *name, int omode, int docsbnum, int indexbnum,
+ const char *fname);
+static int odcacheflush(ODEUM *odeum, const char *fname);
+static int odcacheflushfreq(ODEUM *odeum, const char *fname, int min);
+static int odcacheflushrare(ODEUM *odeum, const char *fname, double ratio);
+static int odsortindex(ODEUM *odeum, const char *fname);
+static int odsortcompare(const void *a, const void *b);
+static int odpurgeindex(ODEUM *odeum, const char *fname);
+static CBMAP *odpairsmap(const ODPAIR *pairs, int num);
+static int odwordcompare(const void *a, const void *b);
+static int odmatchoperator(ODEUM *odeum, CBLIST *tokens);
+static ODPAIR *odparsesubexpr(ODEUM *odeum, CBLIST *tokens, CBLIST *nwords, int *np,
+ CBLIST *errors);
+static ODPAIR *odparseexpr(ODEUM *odeum, CBLIST *tokens, CBLIST *nwords, int *np,
+ CBLIST *errors);
+static void odfixtokens(ODEUM *odeum, CBLIST *tokens);
+static void odcleannormalized(ODEUM *odeum, CBLIST *nwords);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* Get a database handle. */
+ODEUM *odopen(const char *name, int omode){
+ assert(name);
+ return odopendb(name, omode, OD_DOCSBNUM, odindexbnum, "odopen");
+}
+
+
+/* Close a database handle. */
+int odclose(ODEUM *odeum){
+ char numbuf[OD_NUMBUFSIZ];
+ int err;
+ assert(odeum);
+ err = FALSE;
+ if(odotcb) odotcb("odclose", odeum, "closing the connection");
+ if(odeum->wmode){
+ if(odotcb) odotcb("odclose", odeum, "writing meta information");
+ sprintf(numbuf, "%d", odeum->dmax);
+ if(!vlput(odeum->rdocsdb, OD_DMAXEXPR, sizeof(OD_DMAXEXPR), numbuf, -1, VL_DOVER)) err = TRUE;
+ sprintf(numbuf, "%d", odeum->dnum);
+ if(!vlput(odeum->rdocsdb, OD_DNUMEXPR, sizeof(OD_DNUMEXPR), numbuf, -1, VL_DOVER)) err = TRUE;
+ if(!odcacheflushfreq(odeum, "odclose", OD_CFENDSIZ)) err = TRUE;
+ if(!odcacheflushrare(odeum, "odclose", OD_CFRFRAT)) err = TRUE;
+ if(!odcacheflush(odeum, "odclose")) err = TRUE;
+ if(!odsortindex(odeum, "odclose")) err = TRUE;
+ cbmapclose(odeum->cachemap);
+ cbmapclose(odeum->sortmap);
+ }
+ if(!vlclose(odeum->rdocsdb)) err = TRUE;
+ if(!crclose(odeum->indexdb)) err = TRUE;
+ if(!crclose(odeum->docsdb)) err = TRUE;
+ free(odeum->name);
+ free(odeum);
+ return err ? FALSE : TRUE;
+}
+
+
+/* Store a document. */
+int odput(ODEUM *odeum, ODDOC *doc, int wmax, int over){
+ char *tmp, *zbuf;
+ const char *word, *ctmp;
+ int i, docid, tsiz, wsiz, wnum, tmax, num, zsiz;
+ double ival;
+ ODPAIR pair;
+ CBMAP *map;
+ CBLIST *tlist;
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!odeum->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if((tmp = vlget(odeum->rdocsdb, doc->uri, -1, &tsiz)) != NULL){
+ if(!over){
+ free(tmp);
+ dpecodeset(DP_EKEEP, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(tsiz != sizeof(int) || !odoutbyid(odeum, *(int *)tmp)){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ free(tmp);
+ }
+ odeum->dmax++;
+ odeum->dnum++;
+ docid = odeum->dmax;
+ map = cbmapopen();
+ cbmapput(map, OD_URIEXPR, sizeof(OD_URIEXPR), doc->uri, -1, TRUE);
+ tmp = cbmapdump(doc->attrs, &tsiz);
+ cbmapput(map, OD_ATTRSEXPR, sizeof(OD_ATTRSEXPR), tmp, tsiz, TRUE);
+ free(tmp);
+ if(wmax < 0 || wmax > cblistnum(doc->nwords)) wmax = cblistnum(doc->nwords);
+ tlist = cblistopen();
+ for(i = 0; i < wmax; i++){
+ ctmp = cblistval(doc->nwords, i, &wsiz);
+ cblistpush(tlist, ctmp, wsiz);
+ }
+ tmp = cblistdump(tlist, &tsiz);
+ cbmapput(map, OD_NWORDSEXPR, sizeof(OD_NWORDSEXPR), tmp, tsiz, TRUE);
+ free(tmp);
+ cblistclose(tlist);
+ tlist = cblistopen();
+ for(i = 0; i < wmax; i++){
+ ctmp = cblistval(doc->awords, i, &wsiz);
+ if(strcmp(ctmp, cblistval(doc->nwords, i, NULL))){
+ cblistpush(tlist, ctmp, wsiz);
+ } else {
+ cblistpush(tlist, "\0", 1);
+ }
+ }
+ tmp = cblistdump(tlist, &tsiz);
+ cbmapput(map, OD_AWORDSEXPR, sizeof(OD_AWORDSEXPR), tmp, tsiz, TRUE);
+ free(tmp);
+ cblistclose(tlist);
+ tmp = cbmapdump(map, &tsiz);
+ cbmapclose(map);
+ if(_qdbm_deflate){
+ if(!(zbuf = _qdbm_deflate(tmp, tsiz, &zsiz, _QDBM_ZMRAW))){
+ free(tmp);
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ free(tmp);
+ tmp = zbuf;
+ tsiz = zsiz;
+ }
+ if(!crput(odeum->docsdb, (char *)&docid, sizeof(int), tmp, tsiz, CR_DKEEP)){
+ free(tmp);
+ if(dpecode == DP_EKEEP) dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ free(tmp);
+ if(!vlput(odeum->rdocsdb, doc->uri, -1, (char *)&docid, sizeof(int), VL_DOVER)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ map = cbmapopen();
+ wnum = cblistnum(doc->nwords);
+ tmax = (int)(wnum * OD_WTOPRATE);
+ for(i = 0; i < wnum; i++){
+ word = cblistval(doc->nwords, i, &wsiz);
+ if(wsiz < 1) continue;
+ if((ctmp = cbmapget(map, word, wsiz, NULL)) != NULL){
+ num = *(int *)ctmp + OD_WOCCRPOINT;
+ } else {
+ num = i <= tmax ? OD_WTOPBONUS + OD_WOCCRPOINT : OD_WOCCRPOINT;
+ }
+ cbmapput(map, word, wsiz, (char *)&num, sizeof(int), TRUE);
+ }
+ ival = odlogarithm(wnum);
+ ival = (ival * ival * ival) / 8.0;
+ if(ival < 8.0) ival = 8.0;
+ cbmapiterinit(map);
+ while((word = cbmapiternext(map, &wsiz)) != NULL){
+ pair.id = docid;
+ pair.score = (int)(*(int *)cbmapget(map, word, wsiz, NULL) / ival);
+ cbmapputcat(odeum->cachemap, word, wsiz, (char *)&pair, sizeof(pair));
+ cbmapmove(odeum->cachemap, word, wsiz, FALSE);
+ odeum->cacheasiz += sizeof(pair);
+ cbmapput(odeum->sortmap, word, wsiz, "", 0, FALSE);
+ }
+ cbmapclose(map);
+ if(odeum->cacheasiz > odcachesiz){
+ for(i = OD_CFBEGSIZ; odeum->cacheasiz > odcachesiz * OD_CFLIVERAT && i >= OD_CFENDSIZ;
+ i /= 2){
+ if(!odcacheflushfreq(odeum, "odput", i)) return FALSE;
+ }
+ while(odeum->cacheasiz > odcachesiz * OD_CFLIVERAT){
+ if(!odcacheflushrare(odeum, "odput", OD_CFRFRAT)) return FALSE;
+ }
+ }
+ doc->id = docid;
+ odeum->ldid = docid;
+ return TRUE;
+}
+
+
+/* Delete a document by a URL. */
+int odout(ODEUM *odeum, const char *uri){
+ char *tmp;
+ int tsiz, docid;
+ assert(odeum && uri);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!odeum->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(tmp = vlget(odeum->rdocsdb, uri, -1, &tsiz))){
+ if(dpecode != DP_ENOITEM) odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(tsiz != sizeof(int)){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ docid = *(int *)tmp;
+ free(tmp);
+ return odoutbyid(odeum, docid);
+}
+
+
+/* Delete a document specified by an ID number. */
+int odoutbyid(ODEUM *odeum, int id){
+ char *tmp, *zbuf;
+ const char *uritmp;
+ int tsiz, uritsiz, zsiz;
+ CBMAP *map;
+ assert(odeum && id > 0);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!odeum->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(tmp = crget(odeum->docsdb, (char *)&id, sizeof(int), 0, -1, &tsiz))){
+ if(dpecode != DP_ENOITEM) odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(_qdbm_inflate){
+ if(!(zbuf = _qdbm_inflate(tmp, tsiz, &zsiz, _QDBM_ZMRAW))){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ free(tmp);
+ tmp = zbuf;
+ tsiz = zsiz;
+ }
+ map = cbmapload(tmp, tsiz);
+ free(tmp);
+ uritmp = cbmapget(map, OD_URIEXPR, sizeof(OD_URIEXPR), &uritsiz);
+ if(!uritmp || !vlout(odeum->rdocsdb, uritmp, uritsiz)){
+ cbmapclose(map);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ cbmapclose(map);
+ if(!crout(odeum->docsdb, (char *)&id, sizeof(int))){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ odeum->dnum--;
+ return TRUE;
+}
+
+
+/* Retrieve a document by a URI. */
+ODDOC *odget(ODEUM *odeum, const char *uri){
+ char *tmp;
+ int tsiz, docid;
+ assert(odeum && uri);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!(tmp = vlget(odeum->rdocsdb, uri, -1, &tsiz))){
+ if(dpecode != DP_ENOITEM) odeum->fatal = TRUE;
+ return NULL;
+ }
+ if(tsiz != sizeof(int)){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ docid = *(int *)tmp;
+ free(tmp);
+ return odgetbyid(odeum, docid);
+}
+
+
+/* Retrieve a document by an ID number. */
+ODDOC *odgetbyid(ODEUM *odeum, int id){
+ char *tmp, *zbuf;
+ const char *uritmp, *attrstmp, *nwordstmp, *awordstmp, *asis, *normal;
+ int i, tsiz, uritsiz, attrstsiz, nwordstsiz, awordstsiz, zsiz, asiz, nsiz;
+ ODDOC *doc;
+ CBMAP *map;
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(id < 1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!(tmp = crget(odeum->docsdb, (char *)&id, sizeof(int), 0, -1, &tsiz))){
+ if(dpecode != DP_ENOITEM) odeum->fatal = TRUE;
+ return NULL;
+ }
+ if(_qdbm_inflate){
+ if(!(zbuf = _qdbm_inflate(tmp, tsiz, &zsiz, _QDBM_ZMRAW))){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ free(tmp);
+ tmp = zbuf;
+ tsiz = zsiz;
+ }
+ map = cbmapload(tmp, tsiz);
+ free(tmp);
+ uritmp = cbmapget(map, OD_URIEXPR, sizeof(OD_URIEXPR), &uritsiz);
+ attrstmp = cbmapget(map, OD_ATTRSEXPR, sizeof(OD_ATTRSEXPR), &attrstsiz);
+ nwordstmp = cbmapget(map, OD_NWORDSEXPR, sizeof(OD_NWORDSEXPR), &nwordstsiz);
+ awordstmp = cbmapget(map, OD_AWORDSEXPR, sizeof(OD_AWORDSEXPR), &awordstsiz);
+ if(!uritmp || !attrstmp || !nwordstmp || !awordstmp){
+ cbmapclose(map);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ doc = cbmalloc(sizeof(ODDOC));
+ doc->id = id;
+ doc->uri = cbmemdup(uritmp, uritsiz);
+ doc->attrs = cbmapload(attrstmp, attrstsiz);
+ doc->nwords = cblistload(nwordstmp, nwordstsiz);
+ doc->awords = cblistload(awordstmp, awordstsiz);
+ cbmapclose(map);
+ for(i = 0; i < cblistnum(doc->awords); i++){
+ asis = cblistval(doc->awords, i, &asiz);
+ if(asiz == 1 && asis[0] == '\0'){
+ normal = cblistval(doc->nwords, i, &nsiz);
+ cblistover(doc->awords, i, normal, nsiz);
+ }
+ }
+ return doc;
+}
+
+
+/* Retrieve the ID of the document specified by a URI. */
+int odgetidbyuri(ODEUM *odeum, const char *uri){
+ char *tmp;
+ int tsiz, docid;
+ assert(odeum && uri);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ if(!(tmp = vlget(odeum->rdocsdb, uri, -1, &tsiz))){
+ if(dpecode != DP_ENOITEM) odeum->fatal = TRUE;
+ return -1;
+ }
+ if(tsiz != sizeof(int)){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return -1;
+ }
+ docid = *(int *)tmp;
+ free(tmp);
+ return docid;
+}
+
+
+/* Check whether the document specified by an ID number exists. */
+int odcheck(ODEUM *odeum, int id){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(id < 1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return crvsiz(odeum->docsdb, (char *)&id, sizeof(int)) != -1;
+}
+
+
+/* Search the inverted index for documents including a word. */
+ODPAIR *odsearch(ODEUM *odeum, const char *word, int max, int *np){
+ char *tmp;
+ int tsiz;
+ assert(odeum && word && np);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(odeum->wmode && cbmaprnum(odeum->sortmap) > 0 &&
+ (!odcacheflush(odeum, "odsearch") || !odsortindex(odeum, "odsearch"))){
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ max = max < 0 ? -1 : max * sizeof(ODPAIR);
+ if(!(tmp = crget(odeum->indexdb, word, -1, 0, max, &tsiz))){
+ if(dpecode != DP_ENOITEM){
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ *np = 0;
+ return cbmalloc(1);
+ }
+ *np = tsiz / sizeof(ODPAIR);
+ return (ODPAIR *)tmp;
+}
+
+
+/* Get the number of documents including a word. */
+int odsearchdnum(ODEUM *odeum, const char *word){
+ int rv;
+ assert(odeum && word);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ rv = crvsiz(odeum->indexdb, word, -1);
+ return rv < 0 ? -1 : rv / sizeof(ODPAIR);
+}
+
+
+/* Initialize the iterator of a database handle. */
+int oditerinit(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return criterinit(odeum->docsdb);
+}
+
+
+/* Get the next key of the iterator. */
+ODDOC *oditernext(ODEUM *odeum){
+ char *tmp;
+ int tsiz, docsid;
+ ODDOC *doc;
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ doc = NULL;
+ while(TRUE){
+ if(!(tmp = criternext(odeum->docsdb, &tsiz))){
+ if(dpecode != DP_ENOITEM) odeum->fatal = TRUE;
+ return NULL;
+ }
+ if(tsiz != sizeof(int)){
+ free(tmp);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ docsid = *(int *)tmp;
+ free(tmp);
+ if((doc = odgetbyid(odeum, docsid)) != NULL) break;
+ if(dpecode != DP_ENOITEM){
+ odeum->fatal = TRUE;
+ return NULL;
+ }
+ }
+ return doc;
+}
+
+
+/* Synchronize updating contents with the files and the devices. */
+int odsync(ODEUM *odeum){
+ char numbuf[OD_NUMBUFSIZ];
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!odeum->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(odotcb) odotcb("odsync", odeum, "writing meta information");
+ sprintf(numbuf, "%d", odeum->dmax);
+ if(!vlput(odeum->rdocsdb, OD_DMAXEXPR, sizeof(OD_DMAXEXPR), numbuf, -1, VL_DOVER)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ sprintf(numbuf, "%d", odeum->dnum);
+ if(!vlput(odeum->rdocsdb, OD_DNUMEXPR, sizeof(OD_DNUMEXPR), numbuf, -1, VL_DOVER)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(!odcacheflush(odeum, "odsync")){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(!odsortindex(odeum, "odsync")){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odotcb) odotcb("odsync", odeum, "synchronizing the document database");
+ if(!crsync(odeum->docsdb)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odotcb) odotcb("odsync", odeum, "synchronizing the inverted index");
+ if(!crsync(odeum->indexdb)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odotcb) odotcb("odsync", odeum, "synchronizing the reverse dictionary");
+ if(!vlsync(odeum->rdocsdb)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Optimize a database. */
+int odoptimize(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!odeum->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!odcacheflush(odeum, "odoptimize")){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odeum->ldid < 1 || odeum->ldid != odeum->dnum){
+ if(!odpurgeindex(odeum, "odoptimize")){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ }
+ if(odeum->ldid > 0){
+ if(!odsortindex(odeum, "odoptimize")){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ }
+ if(odotcb) odotcb("odoptimize", odeum, "optimizing the document database");
+ if(!croptimize(odeum->docsdb, -1)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odotcb) odotcb("odoptimize", odeum, "optimizing the inverted index");
+ if(!croptimize(odeum->indexdb, -1)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odotcb) odotcb("odoptimize", odeum, "optimizing the reverse dictionary");
+ if(!vloptimize(odeum->rdocsdb)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Get the name of a database. */
+char *odname(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return NULL;
+ }
+ return cbmemdup(odeum->name, -1);
+}
+
+
+/* Get the total size of database files. */
+double odfsiz(ODEUM *odeum){
+ double fsiz, rv;
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ fsiz = 0;
+ if((rv = crfsizd(odeum->docsdb)) < 0) return -1.0;
+ fsiz += rv;
+ if((rv = crfsizd(odeum->indexdb)) < 0) return -1.0;
+ fsiz += rv;
+ if((rv = vlfsiz(odeum->rdocsdb)) == -1) return -1.0;
+ fsiz += rv;
+ return fsiz;
+}
+
+
+/* Get the total number of the elements of the bucket arrays for the inverted index. */
+int odbnum(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return crbnum(odeum->indexdb);
+}
+
+
+/* Get the total number of the used elements of the bucket arrays in the inverted index. */
+int odbusenum(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return crbusenum(odeum->indexdb);
+}
+
+
+/* Get the number of the documents stored in a database. */
+int oddnum(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return odeum->dnum;
+}
+
+
+/* Get the number of the words stored in a database. */
+int odwnum(ODEUM *odeum){
+ assert(odeum);
+ if(odeum->fatal){
+ dpecodeset(DP_EFATAL, __FILE__, __LINE__);
+ return -1;
+ }
+ return crrnum(odeum->indexdb);
+}
+
+
+/* Check whether a database handle is a writer or not. */
+int odwritable(ODEUM *odeum){
+ assert(odeum);
+ return odeum->wmode;
+}
+
+
+/* Check whether a database has a fatal error or not. */
+int odfatalerror(ODEUM *odeum){
+ assert(odeum);
+ return odeum->fatal;
+}
+
+
+/* Get the inode number of a database directory. */
+int odinode(ODEUM *odeum){
+ assert(odeum);
+ return odeum->inode;
+}
+
+
+/* Get the last modified time of a database. */
+time_t odmtime(ODEUM *odeum){
+ assert(odeum);
+ return crmtime(odeum->indexdb);
+}
+
+
+/* Merge plural database directories. */
+int odmerge(const char *name, const CBLIST *elemnames){
+ ODEUM *odeum, **elems;
+ CURIA *curia, *ecuria;
+ VILLA *villa, *evilla;
+ ODPAIR *pairs;
+ char *word, *kbuf, *vbuf, *dbuf, otmsg[OD_OTCBBUFSIZ];
+ char *wpunit[OD_MIWUNIT], *vpunit[OD_MIWUNIT];
+ int i, j, k, num, dnum, wnum, dbnum, ibnum, tnum, wsunit[OD_MIWUNIT], vsunit[OD_MIWUNIT];
+ int err, *bases, sum, max, wsiz, ksiz, vsiz, uend, unum, pnum, align, id, nid, dsiz;
+ assert(name && elemnames);
+ num = cblistnum(elemnames);
+ elems = cbmalloc(num * sizeof(ODEUM *) + 1);
+ dnum = 0;
+ wnum = 0;
+ for(i = 0; i < num; i++){
+ if(!(elems[i] = odopen(cblistval(elemnames, i, NULL), OD_OREADER))){
+ for(i -= 1; i >= 0; i--){
+ odclose(elems[i]);
+ }
+ free(elems);
+ return FALSE;
+ }
+ dnum += oddnum(elems[i]);
+ wnum += odwnum(elems[i]);
+ }
+ dbnum = (int)(dnum * OD_MDBRATIO / OD_DOCSDNUM);
+ ibnum = (int)(wnum * OD_MIBRATIO / odindexdnum);
+ if(!(odeum = odopendb(name, OD_OWRITER | OD_OCREAT | OD_OTRUNC, dbnum, ibnum, "odmerge"))){
+ for(i = 0; i < num; i++){
+ odclose(elems[i]);
+ }
+ free(elems);
+ return FALSE;
+ }
+ err = FALSE;
+ if(odotcb) odotcb("odmerge", odeum, "calculating the base ID numbers");
+ bases = cbmalloc(num * sizeof(int) + 1);
+ sum = 0;
+ for(i = 0; i < num; i++){
+ ecuria = elems[i]->docsdb;
+ max = 0;
+ if(!criterinit(ecuria) && dpecode != DP_ENOITEM) err = TRUE;
+ while((kbuf = criternext(ecuria, &ksiz)) != NULL){
+ if(ksiz == sizeof(int)){
+ if(*(int *)kbuf > max) max = *(int *)kbuf;
+ }
+ free(kbuf);
+ }
+ bases[i] = sum;
+ sum += max;
+ }
+ curia = odeum->indexdb;
+ for(i = 0; i < num; i++){
+ if(odotcb){
+ sprintf(otmsg, "merging the inverted index (%d/%d)", i + 1, num);
+ odotcb("odmerge", odeum, otmsg);
+ }
+ ecuria = elems[i]->indexdb;
+ tnum = 0;
+ uend = FALSE;
+ if(!criterinit(ecuria) && dpecode != DP_ENOITEM) err = TRUE;
+ while(!uend){
+ for(unum = 0; unum < OD_MIWUNIT; unum++){
+ if(!(word = criternext(ecuria, &wsiz))){
+ uend = TRUE;
+ break;
+ }
+ if(!(vbuf = crget(ecuria, word, wsiz, 0, -1, &vsiz))){
+ err = TRUE;
+ free(word);
+ break;
+ }
+ wpunit[unum] = word;
+ wsunit[unum] = wsiz;
+ vpunit[unum] = vbuf;
+ vsunit[unum] = vsiz;
+ }
+ for(j = 0; j < unum; j++){
+ word = wpunit[j];
+ wsiz = wsunit[j];
+ vbuf = vpunit[j];
+ vsiz = vsunit[j];
+ pairs = (ODPAIR *)vbuf;
+ pnum = vsiz / sizeof(ODPAIR);
+ for(k = 0; k < pnum; k++){
+ pairs[k].id += bases[i];
+ }
+ align = (int)(i < num - 1 ? vsiz * (num - i) * OD_MIARATIO : OD_INDEXALIGN);
+ if(!crsetalign(curia, align)) err = TRUE;
+ if(!crput(curia, word, wsiz, vbuf, vsiz, CR_DCAT)) err = TRUE;
+ free(vbuf);
+ free(word);
+ if(odotcb && (tnum + 1) % OD_OTPERWORDS == 0){
+ sprintf(otmsg, "... (%d/%d)", tnum + 1, crrnum(ecuria));
+ odotcb("odmerge", odeum, otmsg);
+ }
+ tnum++;
+ }
+ }
+ }
+ if(odotcb) odotcb("odmerge", odeum, "sorting the inverted index");
+ tnum = 0;
+ if(!criterinit(curia) && dpecode != DP_ENOITEM) err = TRUE;
+ while((word = criternext(curia, &wsiz)) != NULL){
+ if((vbuf = crget(curia, word, wsiz, 0, -1, &vsiz)) != NULL){
+ if(vsiz > sizeof(ODPAIR)){
+ pairs = (ODPAIR *)vbuf;
+ pnum = vsiz / sizeof(ODPAIR);
+ qsort(pairs, pnum, sizeof(ODPAIR), odsortcompare);
+ if(!crput(curia, word, wsiz, vbuf, vsiz, CR_DOVER)) err = TRUE;
+ }
+ free(vbuf);
+ }
+ free(word);
+ if(odotcb && (tnum + 1) % OD_OTPERWORDS == 0){
+ sprintf(otmsg, "... (%d/%d)", tnum + 1, crrnum(curia));
+ odotcb("odmerge", odeum, otmsg);
+ }
+ tnum++;
+ }
+ if(odotcb) odotcb("odmerge", odeum, "synchronizing the inverted index");
+ if(!crsync(curia)) err = TRUE;
+ dnum = 0;
+ curia = odeum->docsdb;
+ villa = odeum->rdocsdb;
+ for(i = 0; i < num; i++){
+ if(odotcb){
+ sprintf(otmsg, "merging the document database (%d/%d)", i + 1, num);
+ odotcb("odmerge", odeum, otmsg);
+ }
+ evilla = elems[i]->rdocsdb;
+ ecuria = elems[i]->docsdb;
+ tnum = 0;
+ if(!vlcurfirst(evilla) && dpecode != DP_ENOITEM) err = TRUE;
+ while(TRUE){
+ if(!(kbuf = vlcurkey(evilla, &ksiz))) break;
+ if((ksiz == sizeof(OD_DMAXEXPR) && !memcmp(kbuf, OD_DMAXEXPR, ksiz)) ||
+ (ksiz == sizeof(OD_DNUMEXPR) && !memcmp(kbuf, OD_DNUMEXPR, ksiz))){
+ free(kbuf);
+ if(!vlcurnext(evilla)) break;
+ continue;
+ }
+ if(!(vbuf = vlcurval(evilla, &vsiz))){
+ free(kbuf);
+ if(!vlcurnext(evilla)) break;
+ continue;
+ }
+ if(vsiz != sizeof(int)){
+ free(vbuf);
+ free(kbuf);
+ if(!vlcurnext(evilla)) break;
+ continue;
+ }
+ id = *(int *)vbuf;
+ nid = id + bases[i];
+ if(vlput(villa, kbuf, ksiz, (char *)&nid, sizeof(int), VL_DKEEP)){
+ if((dbuf = crget(ecuria, (char *)&id, sizeof(int), 0, -1, &dsiz)) != NULL){
+ if(crput(curia, (char *)&nid, sizeof(int), dbuf, dsiz, CR_DKEEP)){
+ dnum++;
+ } else {
+ err = TRUE;
+ }
+ free(dbuf);
+ } else {
+ err = TRUE;
+ }
+ } else if(dpecode != DP_EKEEP){
+ err = TRUE;
+ }
+ free(vbuf);
+ free(kbuf);
+ odeum->dnum++;
+ if(odotcb && (tnum + 1) % OD_OTPERDOCS == 0){
+ sprintf(otmsg, "... (%d/%d)", tnum + 1, crrnum(ecuria));
+ odotcb("odmerge", odeum, otmsg);
+ }
+ tnum++;
+ if(!vlcurnext(evilla)) break;
+ }
+ }
+ odeum->dnum = dnum;
+ odeum->dmax = dnum;
+ free(bases);
+ if(odotcb) odotcb("odmerge", odeum, "synchronizing the document index");
+ if(!crsync(curia)) err = TRUE;
+ if(!odclose(odeum)) err = TRUE;
+ for(i = 0; i < num; i++){
+ if(!odclose(elems[i])) err = TRUE;
+ }
+ free(elems);
+ return err ? FALSE : TRUE;
+}
+
+
+/* Remove a database directory. */
+int odremove(const char *name){
+ char docsname[OD_PATHBUFSIZ], indexname[OD_PATHBUFSIZ], rdocsname[OD_PATHBUFSIZ];
+ char path[OD_PATHBUFSIZ];
+ const char *file;
+ struct stat sbuf;
+ CBLIST *list;
+ int i;
+ assert(name);
+ sprintf(docsname, "%s%c%s", name, MYPATHCHR, OD_DOCSNAME);
+ sprintf(indexname, "%s%c%s", name, MYPATHCHR, OD_INDEXNAME);
+ sprintf(rdocsname, "%s%c%s", name, MYPATHCHR, OD_RDOCSNAME);
+ if(lstat(name, &sbuf) == -1){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(lstat(docsname, &sbuf) != -1 && !crremove(docsname)) return FALSE;
+ if(lstat(indexname, &sbuf) != -1 && !crremove(indexname)) return FALSE;
+ if(lstat(rdocsname, &sbuf) != -1 && !vlremove(rdocsname)) return FALSE;
+ if((list = cbdirlist(name)) != NULL){
+ for(i = 0; i < cblistnum(list); i++){
+ file = cblistval(list, i, NULL);
+ if(!strcmp(file, MYCDIRSTR) || !strcmp(file, MYPDIRSTR)) continue;
+ sprintf(path, "%s%c%s", name, MYPATHCHR, file);
+ if(lstat(path, &sbuf) == -1) continue;
+ if(S_ISDIR(sbuf.st_mode)){
+ if(!crremove(path)) return FALSE;
+ } else {
+ if(!dpremove(path)) return FALSE;
+ }
+ }
+ cblistclose(list);
+ }
+ if(rmdir(name) == -1){
+ dpecodeset(DP_ERMDIR, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Get a document handle. */
+ODDOC *oddocopen(const char *uri){
+ ODDOC *doc;
+ assert(uri);
+ doc = cbmalloc(sizeof(ODDOC));
+ doc->id = -1;
+ doc->uri = cbmemdup(uri, -1);
+ doc->attrs = cbmapopenex(OD_MAPPBNUM);
+ doc->nwords = cblistopen();
+ doc->awords = cblistopen();
+ return doc;
+}
+
+
+/* Close a document handle. */
+void oddocclose(ODDOC *doc){
+ assert(doc);
+ cblistclose(doc->awords);
+ cblistclose(doc->nwords);
+ cbmapclose(doc->attrs);
+ free(doc->uri);
+ free(doc);
+}
+
+
+/* Add an attribute to a document. */
+void oddocaddattr(ODDOC *doc, const char *name, const char *value){
+ assert(doc && name && value);
+ cbmapput(doc->attrs, name, -1, value, -1, TRUE);
+}
+
+
+/* Add a word to a document. */
+void oddocaddword(ODDOC *doc, const char *normal, const char *asis){
+ assert(doc && normal && asis);
+ cblistpush(doc->nwords, normal, -1);
+ cblistpush(doc->awords, asis, -1);
+}
+
+
+/* Get the ID number of a document. */
+int oddocid(const ODDOC *doc){
+ assert(doc);
+ return doc->id;
+}
+
+
+/* Get the URI of a document. */
+const char *oddocuri(const ODDOC *doc){
+ assert(doc);
+ return doc->uri;
+}
+
+
+/* Get the value of an attribute of a document. */
+const char *oddocgetattr(const ODDOC *doc, const char *name){
+ assert(doc && name);
+ return cbmapget(doc->attrs, name, -1, NULL);
+}
+
+
+/* Get the list handle contains words in normalized form of a document. */
+const CBLIST *oddocnwords(const ODDOC *doc){
+ assert(doc);
+ return doc->nwords;
+}
+
+
+/* Get the list handle contains words in appearance form of a document. */
+const CBLIST *oddocawords(const ODDOC *doc){
+ assert(doc);
+ return doc->awords;
+}
+
+
+/* Get the map handle contains keywords in normalized form and their scores. */
+CBMAP *oddocscores(const ODDOC *doc, int max, ODEUM *odeum){
+ const CBLIST *nwords;
+ CBMAP *map, *kwmap;
+ const char *word, *ctmp;
+ char numbuf[OD_NUMBUFSIZ];
+ ODWORD *owords;
+ int i, wsiz, wnum, hnum, mnum, nbsiz;
+ double ival;
+ assert(doc && max >= 0);
+ map = cbmapopen();
+ nwords = oddocnwords(doc);
+ for(i = 0; i < cblistnum(nwords); i++){
+ word = cblistval(nwords, i, &wsiz);
+ if(wsiz < 1) continue;
+ if((ctmp = cbmapget(map, word, wsiz, NULL)) != NULL){
+ wnum = *(int *)ctmp + OD_WOCCRPOINT;
+ } else {
+ wnum = OD_WOCCRPOINT;
+ }
+ cbmapput(map, word, wsiz, (char *)&wnum, sizeof(int), TRUE);
+ }
+ mnum = cbmaprnum(map);
+ owords = cbmalloc(mnum * sizeof(ODWORD) + 1);
+ cbmapiterinit(map);
+ for(i = 0; (word = cbmapiternext(map, &wsiz)) != NULL; i++){
+ owords[i].word = word;
+ owords[i].num = *(int *)cbmapget(map, word, wsiz, NULL);
+ }
+ qsort(owords, mnum, sizeof(ODWORD), odwordcompare);
+ if(odeum){
+ if(mnum > max * OD_KEYCRATIO) mnum = (int)(max * OD_KEYCRATIO);
+ for(i = 0; i < mnum; i++){
+ if((hnum = odsearchdnum(odeum, owords[i].word)) < 0) hnum = 0;
+ ival = odlogarithm(hnum);
+ ival = (ival * ival * ival) / 8.0;
+ if(ival < 8.0) ival = 8.0;
+ owords[i].num = (int)(owords[i].num / ival);
+ }
+ qsort(owords, mnum, sizeof(ODWORD), odwordcompare);
+ }
+ if(mnum > max) mnum = max;
+ kwmap = cbmapopenex(OD_MAPPBNUM);
+ for(i = 0; i < mnum; i++){
+ nbsiz = sprintf(numbuf, "%d", owords[i].num);
+ cbmapput(kwmap, owords[i].word, -1, numbuf, nbsiz, TRUE);
+ }
+ free(owords);
+ cbmapclose(map);
+ return kwmap;
+}
+
+
+/* Break a text into words in appearance form. */
+CBLIST *odbreaktext(const char *text){
+ const char *word;
+ CBLIST *elems, *words;
+ int i, j, dif, wsiz, pv, delim;
+ assert(text);
+ words = cblistopen();
+ elems = cbsplit(text, -1, OD_SPACECHARS);
+ for(i = 0; i < cblistnum(elems); i++){
+ word = cblistval(elems, i, &wsiz);
+ delim = FALSE;
+ j = 0;
+ pv = 0;
+ while(TRUE){
+ dif = j - pv;
+ if(j >= wsiz){
+ if(dif > 0 && dif <= OD_MAXWORDLEN) cblistpush(words, word + pv, j - pv);
+ break;
+ }
+ if(delim){
+ if(!strchr(OD_DELIMCHARS, word[j])){
+ if(dif > 0 && dif <= OD_MAXWORDLEN) cblistpush(words, word + pv, j - pv);
+ pv = j;
+ delim = FALSE;
+ }
+ } else {
+ if(strchr(OD_DELIMCHARS, word[j])){
+ if(dif > 0 && dif <= OD_MAXWORDLEN) cblistpush(words, word + pv, j - pv);
+ pv = j;
+ delim = TRUE;
+ }
+ }
+ j++;
+ }
+ }
+ cblistclose(elems);
+ return words;
+}
+
+
+/* Make the normalized form of a word. */
+char *odnormalizeword(const char *asis){
+ char *nword;
+ int i;
+ assert(asis);
+ for(i = 0; asis[i] != '\0'; i++){
+ if(!strchr(OD_DELIMCHARS, asis[i])) break;
+ }
+ if(asis[i] == '\0') return cbmemdup("", 0);
+ nword = cbmemdup(asis, -1);
+ for(i = 0; nword[i] != '\0'; i++){
+ if(nword[i] >= 'A' && nword[i] <= 'Z') nword[i] += 'a' - 'A';
+ }
+ while(i >= 0){
+ if(strchr(OD_GLUECHARS, nword[i])){
+ nword[i] = '\0';
+ } else {
+ break;
+ }
+ i--;
+ }
+ return nword;
+}
+
+
+/* Get the common elements of two sets of documents. */
+ODPAIR *odpairsand(ODPAIR *apairs, int anum, ODPAIR *bpairs, int bnum, int *np){
+ CBMAP *map;
+ ODPAIR *result;
+ const char *tmp;
+ int i, rnum;
+ assert(apairs && anum >= 0 && bpairs && bnum >= 0);
+ map = odpairsmap(bpairs, bnum);
+ result = cbmalloc(sizeof(ODPAIR) * anum + 1);
+ rnum = 0;
+ for(i = 0; i < anum; i++){
+ if(!(tmp = cbmapget(map, (char *)&(apairs[i].id), sizeof(int), NULL))) continue;
+ result[rnum].id = apairs[i].id;
+ result[rnum].score = apairs[i].score + *(int *)tmp;
+ rnum++;
+ }
+ cbmapclose(map);
+ qsort(result, rnum, sizeof(ODPAIR), odsortcompare);
+ *np = rnum;
+ return result;
+}
+
+
+/* Get the sum of elements of two sets of documents. */
+ODPAIR *odpairsor(ODPAIR *apairs, int anum, ODPAIR *bpairs, int bnum, int *np){
+ CBMAP *map;
+ ODPAIR *result;
+ const char *tmp;
+ int i, score, rnum;
+ assert(apairs && anum >= 0 && bpairs && bnum >= 0);
+ map = odpairsmap(bpairs, bnum);
+ for(i = 0; i < anum; i++){
+ score = 0;
+ if((tmp = cbmapget(map, (char *)&(apairs[i].id), sizeof(int), NULL)) != NULL)
+ score = *(int *)tmp;
+ score += apairs[i].score;
+ cbmapput(map, (char *)&(apairs[i].id), sizeof(int),
+ (char *)&score, sizeof(int), TRUE);
+ }
+ rnum = cbmaprnum(map);
+ result = cbmalloc(rnum * sizeof(ODPAIR) + 1);
+ cbmapiterinit(map);
+ for(i = 0; (tmp = cbmapiternext(map, NULL)) != NULL; i++){
+ result[i].id = *(int *)tmp;
+ result[i].score = *(int *)cbmapget(map, tmp, sizeof(int), NULL);
+ }
+ cbmapclose(map);
+ qsort(result, rnum, sizeof(ODPAIR), odsortcompare);
+ *np = rnum;
+ return result;
+}
+
+
+/* Get the difference set of documents. */
+ODPAIR *odpairsnotand(ODPAIR *apairs, int anum, ODPAIR *bpairs, int bnum, int *np){
+ CBMAP *map;
+ ODPAIR *result;
+ const char *tmp;
+ int i, rnum;
+ assert(apairs && anum >= 0 && bpairs && bnum >= 0);
+ map = odpairsmap(bpairs, bnum);
+ result = cbmalloc(sizeof(ODPAIR) * anum + 1);
+ rnum = 0;
+ for(i = 0; i < anum; i++){
+ if((tmp = cbmapget(map, (char *)&(apairs[i].id), sizeof(int), NULL)) != NULL) continue;
+ result[rnum].id = apairs[i].id;
+ result[rnum].score = apairs[i].score;
+ rnum++;
+ }
+ cbmapclose(map);
+ qsort(result, rnum, sizeof(ODPAIR), odsortcompare);
+ *np = rnum;
+ return result;
+}
+
+
+/* Sort a set of documents in descending order of scores. */
+void odpairssort(ODPAIR *pairs, int pnum){
+ assert(pairs && pnum >= 0);
+ qsort(pairs, pnum, sizeof(ODPAIR), odsortcompare);
+}
+
+
+/* Get the natural logarithm of a number. */
+double odlogarithm(double x){
+ int i;
+ if(x <= 1.0) return 0.0;
+ x = x * x * x * x * x * x * x * x * x * x;
+ for(i = 0; x > 1.0; i++){
+ x /= 2.718281828459;
+ }
+ return (double)i / 10.0;
+}
+
+
+/* Get the cosine of the angle of two vectors. */
+double odvectorcosine(const int *avec, const int *bvec, int vnum){
+ double rv;
+ assert(avec && bvec && vnum >= 0);
+ rv = odvecinnerproduct(avec, bvec, vnum) /
+ ((odvecabsolute(avec, vnum) * odvecabsolute(bvec, vnum)));
+ return rv > 0.0 ? rv : 0.0;
+}
+
+
+/* Set the global tuning parameters. */
+void odsettuning(int ibnum, int idnum, int cbnum, int csiz){
+ if(ibnum > 0) odindexbnum = ibnum;
+ if(idnum > 0) odindexdnum = idnum;
+ if(cbnum > 0) odcachebnum = dpprimenum(cbnum);
+ if(csiz > 0) odcachesiz = csiz;
+}
+
+
+/* Break a text into words and store appearance forms and normalized form into lists. */
+void odanalyzetext(ODEUM *odeum, const char *text, CBLIST *awords, CBLIST *nwords){
+ char aword[OD_MAXWORDLEN+1], *wp;
+ int lev, wsiz;
+ assert(odeum && text && awords);
+ lev = OD_EVSPACE;
+ wsiz = 0;
+ for(; *text != '\0'; text++){
+ switch(odeum->statechars[*(unsigned char *)text]){
+ case OD_EVWORD:
+ if(wsiz > 0 && lev == OD_EVDELIM){
+ cblistpush(awords, aword, wsiz);
+ if(nwords) cblistpush(nwords, "", 0);
+ wsiz = 0;
+ }
+ if(wsiz <= OD_MAXWORDLEN){
+ aword[wsiz++] = *text;
+ }
+ lev = OD_EVWORD;
+ break;
+ case OD_EVGLUE:
+ if(wsiz > 0 && lev == OD_EVDELIM){
+ cblistpush(awords, aword, wsiz);
+ if(nwords) cblistpush(nwords, "", 0);
+ wsiz = 0;
+ }
+ if(wsiz <= OD_MAXWORDLEN){
+ aword[wsiz++] = *text;
+ }
+ lev = OD_EVGLUE;
+ break;
+ case OD_EVDELIM:
+ if(wsiz > 0 && lev != OD_EVDELIM){
+ cblistpush(awords, aword, wsiz);
+ if(nwords){
+ wp = aword;
+ aword[wsiz] = '\0';
+ while(*wp != '\0'){
+ if(*wp >= 'A' && *wp <= 'Z') *wp += 'a' - 'A';
+ wp++;
+ }
+ wp--;
+ while(wp >= aword && odeum->statechars[*(unsigned char *)wp] == OD_EVGLUE){
+ wsiz--;
+ wp--;
+ }
+ cblistpush(nwords, aword, wsiz);
+ }
+ wsiz = 0;
+ }
+ if(wsiz <= OD_MAXWORDLEN){
+ aword[wsiz++] = *text;
+ }
+ lev = OD_EVDELIM;
+ break;
+ default:
+ if(wsiz > 0){
+ cblistpush(awords, aword, wsiz);
+ if(nwords){
+ if(lev == OD_EVDELIM){
+ cblistpush(nwords, "", 0);
+ } else {
+ wp = aword;
+ aword[wsiz] = '\0';
+ while(*wp != '\0'){
+ if(*wp >= 'A' && *wp <= 'Z') *wp += 'a' - 'A';
+ wp++;
+ }
+ wp--;
+ while(wp >= aword && odeum->statechars[*(unsigned char *)wp] == OD_EVGLUE){
+ wsiz--;
+ wp--;
+ }
+ cblistpush(nwords, aword, wsiz);
+ }
+ }
+ wsiz = 0;
+ }
+ lev = OD_EVSPACE;
+ break;
+ }
+ }
+ if(wsiz > 0){
+ cblistpush(awords, aword, wsiz);
+ if(nwords){
+ if(lev == OD_EVDELIM){
+ cblistpush(nwords, "", 0);
+ } else {
+ wp = aword;
+ aword[wsiz] = '\0';
+ while(*wp != '\0'){
+ if(*wp >= 'A' && *wp <= 'Z') *wp += 'a' - 'A';
+ wp++;
+ }
+ wp--;
+ while(wp >= aword && odeum->statechars[*(unsigned char *)wp] == OD_EVGLUE){
+ wsiz--;
+ wp--;
+ }
+ cblistpush(nwords, aword, wsiz);
+ }
+ }
+ wsiz = 0;
+ }
+}
+
+
+/* Set the classes of characters used by `odanalyzetext'. */
+void odsetcharclass(ODEUM *odeum, const char *spacechars, const char *delimchars,
+ const char *gluechars){
+ assert(odeum && spacechars && delimchars && gluechars);
+ memset(odeum->statechars, OD_EVWORD, sizeof(odeum->statechars));
+ for(; *spacechars != '\0'; spacechars++){
+ odeum->statechars[*(unsigned char *)spacechars] = OD_EVSPACE;
+ }
+ for(; *delimchars != '\0'; delimchars++){
+ odeum->statechars[*(unsigned char *)delimchars] = OD_EVDELIM;
+ }
+ for(; *gluechars != '\0'; gluechars++){
+ odeum->statechars[*(unsigned char *)gluechars] = OD_EVGLUE;
+ }
+}
+
+
+/* Query a database using a small boolean query language. */
+ODPAIR *odquery(ODEUM *odeum, const char *query, int *np, CBLIST *errors){
+ CBLIST *tokens = cblistopen();
+ CBLIST *nwords = cblistopen();
+ ODPAIR *results = NULL;
+ assert(odeum && query && np);
+ odanalyzetext(odeum, query, tokens, nwords);
+ odcleannormalized(odeum, nwords);
+ odfixtokens(odeum, tokens);
+ results = odparseexpr(odeum, tokens, nwords, np, errors);
+ cblistclose(tokens);
+ cblistclose(nwords);
+ return results;
+}
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Get the internal database handle for documents. */
+CURIA *odidbdocs(ODEUM *odeum){
+ assert(odeum);
+ return odeum->docsdb;
+}
+
+
+/* Get the internal database handle for the inverted index. */
+CURIA *odidbindex(ODEUM *odeum){
+ assert(odeum);
+ return odeum->indexdb;
+}
+
+
+/* Get the internal database handle for the reverse dictionary. */
+VILLA *odidbrdocs(ODEUM *odeum){
+ assert(odeum);
+ return odeum->rdocsdb;
+}
+
+
+/* Set the call back function called in merging. */
+void odsetotcb(void (*otcb)(const char *, ODEUM *, const char *)){
+ odotcb = otcb;
+}
+
+
+/* Get the positive one of square roots of a number. */
+double odsquareroot(double x){
+ double c, rv;
+ if(x <= 0.0) return 0.0;
+ c = x > 1.0 ? x : 1;
+ do {
+ rv = c;
+ c = (x / c + c) * 0.5;
+ } while(c < rv);
+ return rv;
+}
+
+
+/* Get the absolute of a vector. */
+double odvecabsolute(const int *vec, int vnum){
+ double rv;
+ int i;
+ assert(vec && vnum >= 0);
+ rv = 0;
+ for(i = 0; i < vnum; i++){
+ rv += (double)vec[i] * (double)vec[i];
+ }
+ return odsquareroot(rv);
+}
+
+
+/* Get the inner product of two vectors. */
+double odvecinnerproduct(const int *avec, const int *bvec, int vnum){
+ double rv;
+ int i;
+ assert(avec && bvec && vnum >= 0);
+ rv = 0;
+ for(i = 0; i < vnum; i++){
+ rv += (double)avec[i] * (double)bvec[i];
+ }
+ return rv;
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Get a database handle.
+ `name' specifies the name of a database directory.
+ `omode' specifies the connection mode.
+ `docsbnum` specifies the number of buckets of the document database.
+ `indexbnum` specifies the number of buckets of the index database.
+ `fname' specifies the name of caller function.
+ The return value is the database handle or `NULL' if it is not successful. */
+static ODEUM *odopendb(const char *name, int omode, int docsbnum, int indexbnum,
+ const char *fname){
+ int cromode, vlomode, inode, dmax, dnum;
+ char docsname[OD_PATHBUFSIZ], indexname[OD_PATHBUFSIZ], rdocsname[OD_PATHBUFSIZ], *tmp;
+ struct stat sbuf;
+ CURIA *docsdb, *indexdb;
+ VILLA *rdocsdb;
+ CBMAP *cachemap;
+ CBMAP *sortmap;
+ ODEUM *odeum;
+ assert(name);
+ if(strlen(name) > OD_NAMEMAX){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return NULL;
+ }
+ cromode = CR_OREADER;
+ vlomode = VL_OREADER;
+ if(omode & OD_OWRITER){
+ cromode = CR_OWRITER;
+ vlomode = VL_OWRITER | VL_OZCOMP | VL_OYCOMP;
+ if(omode & OD_OCREAT){
+ cromode |= CR_OCREAT;
+ vlomode |= VL_OCREAT;
+ }
+ if(omode & OD_OTRUNC){
+ cromode |= CR_OTRUNC;
+ vlomode |= VL_OTRUNC;
+ }
+ }
+ if(omode & OD_ONOLCK){
+ cromode |= CR_ONOLCK;
+ vlomode |= VL_ONOLCK;
+ }
+ if(omode & OD_OLCKNB){
+ cromode |= CR_OLCKNB;
+ vlomode |= VL_OLCKNB;
+ }
+ sprintf(docsname, "%s%c%s", name, MYPATHCHR, OD_DOCSNAME);
+ sprintf(indexname, "%s%c%s", name, MYPATHCHR, OD_INDEXNAME);
+ sprintf(rdocsname, "%s%c%s", name, MYPATHCHR, OD_RDOCSNAME);
+ docsdb = NULL;
+ indexdb = NULL;
+ rdocsdb = NULL;
+ if((omode & OD_OWRITER) && (omode & OD_OCREAT)){
+ if(mkdir(name, OD_DIRMODE) == -1 && errno != EEXIST){
+ dpecodeset(DP_EMKDIR, __FILE__, __LINE__);
+ return NULL;
+ }
+ }
+ if(lstat(name, &sbuf) == -1){
+ dpecodeset(DP_ESTAT, __FILE__, __LINE__);
+ return NULL;
+ }
+ inode = sbuf.st_ino;
+ if(!(docsdb = cropen(docsname, cromode, docsbnum, OD_DOCSDNUM))) return NULL;
+ if(!(indexdb = cropen(indexname, cromode, indexbnum, odindexdnum))){
+ crclose(docsdb);
+ return NULL;
+ }
+ if(omode & OD_OWRITER){
+ if(!crsetalign(docsdb, OD_DOCSALIGN) || !crsetfbpsiz(docsdb, OD_DOCSFBP) ||
+ !crsetalign(indexdb, OD_INDEXALIGN) || !crsetfbpsiz(indexdb, OD_INDEXFBP)){
+ crclose(indexdb);
+ crclose(docsdb);
+ return NULL;
+ }
+ }
+ if(!(rdocsdb = vlopen(rdocsname, vlomode, VL_CMPLEX))){
+ crclose(indexdb);
+ crclose(docsdb);
+ return NULL;
+ }
+ vlsettuning(rdocsdb, OD_RDOCSLRM, OD_RDOCSNIM, OD_RDOCSLCN, OD_RDOCSNCN);
+ if(omode & OD_OWRITER){
+ cachemap = cbmapopenex(odcachebnum);
+ sortmap = cbmapopenex(odcachebnum);
+ } else {
+ cachemap = NULL;
+ sortmap = NULL;
+ }
+ if(vlrnum(rdocsdb) > 0){
+ dmax = -1;
+ dnum = -1;
+ if((tmp = vlget(rdocsdb, OD_DMAXEXPR, sizeof(OD_DMAXEXPR), NULL)) != NULL){
+ dmax = atoi(tmp);
+ free(tmp);
+ }
+ if((tmp = vlget(rdocsdb, OD_DNUMEXPR, sizeof(OD_DNUMEXPR), NULL)) != NULL){
+ dnum = atoi(tmp);
+ free(tmp);
+ }
+ if(dmax < 0 || dnum < 0){
+ if(sortmap) cbmapclose(sortmap);
+ if(cachemap) cbmapclose(cachemap);
+ vlclose(rdocsdb);
+ crclose(indexdb);
+ crclose(docsdb);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ } else {
+ dmax = 0;
+ dnum = 0;
+ }
+ odeum = cbmalloc(sizeof(ODEUM));
+ odeum->name = cbmemdup(name, -1);
+ odeum->wmode = omode & OD_OWRITER;
+ odeum->fatal = FALSE;
+ odeum->inode = inode;
+ odeum->docsdb = docsdb;
+ odeum->indexdb = indexdb;
+ odeum->rdocsdb = rdocsdb;
+ odeum->cachemap = cachemap;
+ odeum->cacheasiz = 0;
+ odeum->sortmap = sortmap;
+ odeum->dmax = dmax;
+ odeum->dnum = dnum;
+ odeum->ldid = -1;
+ odsetcharclass(odeum, OD_SPACECHARS, OD_DELIMCHARS, OD_GLUECHARS);
+ if(odotcb) odotcb(fname, odeum, "the connection was established");
+ return odeum;
+}
+
+
+/* Flush the cache for dirty buffer of words.
+ `odeum' specifies a database handle.
+ `fname' specifies the name of caller function.
+ If successful, the return value is true, else, it is false. */
+static int odcacheflush(ODEUM *odeum, const char *fname){
+ const char *kbuf, *vbuf;
+ char otmsg[OD_OTCBBUFSIZ];
+ int i, rnum, ksiz, vsiz;
+ assert(odeum);
+ if((rnum = cbmaprnum(odeum->cachemap)) < 1) return TRUE;
+ if(odotcb) odotcb(fname, odeum, "flushing caches");
+ cbmapiterinit(odeum->cachemap);
+ for(i = 0; (kbuf = cbmapiternext(odeum->cachemap, &ksiz)) != NULL; i++){
+ vbuf = cbmapget(odeum->cachemap, kbuf, ksiz, &vsiz);
+ if(!crput(odeum->indexdb, kbuf, ksiz, vbuf, vsiz, CR_DCAT)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ if(odotcb && (i + 1) % OD_OTPERWORDS == 0){
+ sprintf(otmsg, "... (%d/%d)", i + 1, rnum);
+ odotcb(fname, odeum, otmsg);
+ }
+ }
+ cbmapclose(odeum->cachemap);
+ odeum->cachemap = cbmapopenex(odcachebnum);
+ odeum->cacheasiz = 0;
+ return TRUE;
+}
+
+
+/* Flush all frequent words in the cache for dirty buffer of words.
+ `odeum' specifies a database handle.
+ `fname' specifies the name of caller function.
+ `min' specifies the minimum size of frequent words.
+ If successful, the return value is true, else, it is false. */
+static int odcacheflushfreq(ODEUM *odeum, const char *fname, int min){
+ const char *kbuf, *vbuf;
+ char otmsg[OD_OTCBBUFSIZ];
+ int rnum, ksiz, vsiz;
+ assert(odeum);
+ if((rnum = cbmaprnum(odeum->cachemap)) < 1) return TRUE;
+ if(odotcb){
+ sprintf(otmsg, "flushing frequent words: min=%d asiz=%d rnum=%d)",
+ min, odeum->cacheasiz, rnum);
+ odotcb(fname, odeum, otmsg);
+ }
+ cbmapiterinit(odeum->cachemap);
+ while((kbuf = cbmapiternext(odeum->cachemap, &ksiz)) != NULL){
+ vbuf = cbmapget(odeum->cachemap, kbuf, ksiz, &vsiz);
+ if(vsiz >= sizeof(ODPAIR) * min){
+ if(!crput(odeum->indexdb, kbuf, ksiz, vbuf, vsiz, CR_DCAT)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ cbmapout(odeum->cachemap, kbuf, ksiz);
+ odeum->cacheasiz -= vsiz;
+ }
+ }
+ if(odotcb){
+ sprintf(otmsg, "... (done): min=%d asiz=%d rnum=%d)",
+ min, odeum->cacheasiz, cbmaprnum(odeum->cachemap));
+ odotcb(fname, odeum, otmsg);
+ }
+ return TRUE;
+}
+
+
+/* Flush the half of rare words in the cache for dirty buffer of words.
+ `odeum' specifies a database handle.
+ `fname' specifies the name of caller function.
+ `ratio' specifies the ratio of rare words.
+ If successful, the return value is true, else, it is false. */
+static int odcacheflushrare(ODEUM *odeum, const char *fname, double ratio){
+ const char *kbuf, *vbuf;
+ char otmsg[OD_OTCBBUFSIZ];
+ int i, rnum, limit, ksiz, vsiz;
+ assert(odeum);
+ if((rnum = cbmaprnum(odeum->cachemap)) < 1) return TRUE;
+ if(odotcb){
+ sprintf(otmsg, "flushing rare words: ratio=%.2f asiz=%d rnum=%d)",
+ ratio, odeum->cacheasiz, rnum);
+ odotcb(fname, odeum, otmsg);
+ }
+ cbmapiterinit(odeum->cachemap);
+ limit = (int)(rnum * ratio);
+ for(i = 0; i < limit && (kbuf = cbmapiternext(odeum->cachemap, &ksiz)) != NULL; i++){
+ vbuf = cbmapget(odeum->cachemap, kbuf, ksiz, &vsiz);
+ if(!crput(odeum->indexdb, kbuf, ksiz, vbuf, vsiz, CR_DCAT)){
+ odeum->fatal = TRUE;
+ return FALSE;
+ }
+ cbmapout(odeum->cachemap, kbuf, ksiz);
+ odeum->cacheasiz -= vsiz;
+ }
+ if(odotcb){
+ sprintf(otmsg, "... (done): ratio=%.2f asiz=%d rnum=%d)",
+ ratio, odeum->cacheasiz, cbmaprnum(odeum->cachemap));
+ odotcb(fname, odeum, otmsg);
+ }
+ return TRUE;
+}
+
+
+/* Sort the records of inverted index.
+ `odeum' specifies a database handle.
+ `fname' specifies the name of caller function.
+ If successful, the return value is true, else, it is false. */
+static int odsortindex(ODEUM *odeum, const char *fname){
+ const char *word;
+ char *tmp, otmsg[OD_OTCBBUFSIZ];
+ int i, rnum, wsiz, tsiz;
+ ODPAIR *pairs;
+ assert(odeum);
+ if((rnum = cbmaprnum(odeum->sortmap)) < 1) return TRUE;
+ if(odotcb) odotcb(fname, odeum, "sorting the inverted index");
+ cbmapiterinit(odeum->sortmap);
+ for(i = 0; (word = cbmapiternext(odeum->sortmap, &wsiz)) != NULL; i++){
+ if((tmp = crget(odeum->indexdb, word, wsiz, 0, -1, &tsiz)) != NULL){
+ if(tsiz > sizeof(ODPAIR)){
+ pairs = (ODPAIR *)tmp;
+ qsort(pairs, tsiz / sizeof(ODPAIR), sizeof(ODPAIR), odsortcompare);
+ if(!crput(odeum->indexdb, word, wsiz, tmp, tsiz, CR_DOVER)){
+ free(tmp);
+ return FALSE;
+ }
+ }
+ free(tmp);
+ } else if(dpecode != DP_ENOITEM){
+ return FALSE;
+ }
+ if(odotcb && (i + 1) % OD_OTPERWORDS == 0){
+ sprintf(otmsg, "... (%d/%d)", i + 1, rnum);
+ odotcb(fname, odeum, otmsg);
+ }
+ }
+ cbmapclose(odeum->sortmap);
+ odeum->sortmap = cbmapopenex(odcachebnum);
+ return TRUE;
+}
+
+
+/* Compare two pairs of structures of a search result.
+ `a' specifies the pointer to the region of one pair.
+ `b' specifies the pointer to the region of the other pair.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+static int odsortcompare(const void *a, const void *b){
+ ODPAIR *ap, *bp;
+ int rv;
+ assert(a && b);
+ ap = (ODPAIR *)a;
+ bp = (ODPAIR *)b;
+ rv = bp->score - ap->score;
+ if(rv != 0) return rv;
+ return ap->id - bp->id;
+}
+
+
+/* Purge the elements of the deleted documents from the inverted index.
+ `odeum' specifies a database handle.
+ `fname' specifies the name of caller function.
+ If successful, the return value is true, else, it is false. */
+static int odpurgeindex(ODEUM *odeum, const char *fname){
+ ODPAIR *pairs;
+ char *kbuf, *vbuf, otmsg[OD_OTCBBUFSIZ];
+ int i, rnum, tnum, ksiz, vsiz, pnum, wi;
+ assert(odeum);
+ if((rnum = crrnum(odeum->indexdb)) < 1) return TRUE;
+ if(odotcb) odotcb(fname, odeum, "purging dispensable regions");
+ if(!criterinit(odeum->indexdb)) return FALSE;
+ tnum = 0;
+ while(TRUE){
+ if(!(kbuf = criternext(odeum->indexdb, &ksiz))){
+ if(dpecode != DP_ENOITEM) return FALSE;
+ break;
+ }
+ if(!(vbuf = crget(odeum->indexdb, kbuf, ksiz, 0, -1, &vsiz))){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ free(kbuf);
+ return FALSE;
+ }
+ pairs = (ODPAIR *)vbuf;
+ pnum = vsiz / sizeof(ODPAIR);
+ wi = 0;
+ for(i = 0; i < pnum; i++){
+ if(crvsiz(odeum->docsdb, (char *)&(pairs[i].id), sizeof(int)) != -1){
+ pairs[wi++] = pairs[i];
+ }
+ }
+ if(wi > 0){
+ if(!crput(odeum->indexdb, kbuf, ksiz, vbuf, wi * sizeof(ODPAIR), CR_DOVER)){
+ free(vbuf);
+ free(kbuf);
+ return FALSE;
+ }
+ } else {
+ if(!crout(odeum->indexdb, kbuf, ksiz)){
+ free(vbuf);
+ free(kbuf);
+ return FALSE;
+ }
+ }
+ free(vbuf);
+ free(kbuf);
+ if(odotcb && (tnum + 1) % OD_OTPERWORDS == 0){
+ sprintf(otmsg, "... (%d/%d)", tnum + 1, rnum);
+ odotcb(fname, odeum, otmsg);
+ }
+ tnum++;
+ }
+ return TRUE;
+}
+
+
+/* Create a map of a document array.
+ `pairs' specifies the pointer to a document array.
+ `num' specifies the number of elements of the array.
+ The return value is a map of the document array. */
+static CBMAP *odpairsmap(const ODPAIR *pairs, int num){
+ CBMAP *map;
+ int i;
+ assert(pairs && num >= 0);
+ map = cbmapopen();
+ for(i = 0; i < num; i++){
+ cbmapput(map, (char *)&(pairs[i].id), sizeof(int),
+ (char *)&(pairs[i].score), sizeof(int), TRUE);
+ }
+ return map;
+}
+
+
+/* compare two pairs of structures of words in a document.
+ `a' specifies the pointer to the region of one word.
+ `b' specifies the pointer to the region of the other word.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+static int odwordcompare(const void *a, const void *b){
+ ODWORD *ap, *bp;
+ int rv;
+ assert(a && b);
+ ap = (ODWORD *)a;
+ bp = (ODWORD *)b;
+ if((rv = bp->num - ap->num) != 0) return rv;
+ if((rv = strlen(bp->word) - strlen(ap->word)) != 0) return rv;
+ return strcmp(ap->word, bp->word);
+}
+
+
+/* Match an operator without taking it off the token list.
+ `odeum' specifies a database handle.
+ `tokens' specifies a list handle of tokens.
+ The return value is whether the next token is an operator. */
+static int odmatchoperator(ODEUM *odeum, CBLIST *tokens){
+ const char *tk = NULL;
+ int tk_len = 0;
+ tk = cblistval(tokens, 0, &tk_len);
+ if(tk && (tk[0] == '&' || tk[0] == '|' || tk[0] == '!')) return 1;
+ return 0;
+}
+
+
+/* Implements the subexpr part of the grammar.
+ `odeum' specifies a database handle.
+ `tokens' specifies a list handle of tokens.
+ `nwords' specifies a list handle of normalized words.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ `errors' specifies a list handle into which error messages are stored.
+ The return value is the pointer to an array of document IDs. */
+static ODPAIR *odparsesubexpr(ODEUM *odeum, CBLIST *tokens, CBLIST *nwords, int *np,
+ CBLIST *errors){
+ char *tk = NULL;
+ int tk_len = 0;
+ char *nword = NULL; /* used to do the actual search, should match with tokens */
+ ODPAIR *result = NULL;
+ int result_num = 0;
+ int i;
+ double ival;
+ if((tk = cblistshift(tokens, &tk_len)) != NULL){
+ assert(tk != NULL);
+ if(tk[0] == '('){
+ free(tk);
+ /* recurse into expr */
+ result = odparseexpr(odeum, tokens, nwords, &result_num, errors);
+ /* match right token RPAREN */
+ tk = cblistshift(tokens, &tk_len);
+ /* print an error if either we didn't get anything or we didn't get a ) */
+ if(tk == NULL){
+ if(errors) cblistpush(errors, "Expression ended without closing ')'", -1);
+ } else if(tk[0] != ')'){
+ if(errors) cblistpush(errors, "Un-balanced parenthesis.", -1);
+ }
+ } else if(odeum->statechars[*(unsigned char *)tk] == 0){
+ /* Perform odsearch with the next norm word that isn't an operator. */
+ nword = cblistshift(nwords, NULL);
+ assert(nword != NULL);
+ if((result = odsearch(odeum, nword, -1, &result_num)) != NULL){
+ /* TF-IDF tuning */
+ ival = odlogarithm(result_num);
+ ival = (ival * ival) / 4.0;
+ if(ival < 4.0) ival = 4.0;
+ for(i = 0; i < result_num; i++){
+ result[i].score = (int)(result[i].score / ival);
+ }
+ }
+ free(nword);
+ } else {
+ if(errors) cblistpush(errors, "Invalid sub-expression. Expected '(' or WORD.", -1);
+ result = cbmalloc(1);
+ result_num = 0;
+ }
+ /* done with the token */
+ free(tk);
+ }
+ *np = result_num;
+ return result;
+}
+
+
+/* Implements the actual recursive decent parser for the mini query language.
+ `odeum' specifies a database handle.
+ `tokens' specifies a list handle of tokens.
+ `nwords' specifies a list handle of normalized words.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ `errors' specifies a list handle into which error messages are stored.
+ The return value is the pointer to an array of document IDs.
+ It simply parses an initial subexpr, and then loops over as many (operator subexpr)
+ sequences as it can find. The odmatchoperator function handles injecting a default &
+ between consecutive words. */
+static ODPAIR *odparseexpr(ODEUM *odeum, CBLIST *tokens, CBLIST *nwords, int *np,
+ CBLIST *errors){
+ ODPAIR *left = NULL;
+ ODPAIR *right = NULL;
+ ODPAIR *temp = NULL;
+ int left_num = 0;
+ int right_num = 0;
+ int temp_num = 0;
+ char *op = NULL;
+ int op_len = 0;
+ if(!(left = odparsesubexpr(odeum, tokens, nwords, &left_num, errors))) return NULL;
+ /* expr ::= subexpr ( op subexpr )* */
+ while(odmatchoperator(odeum, tokens)){
+ op = cblistshift(tokens, &op_len);
+ if(!(right = odparsesubexpr(odeum, tokens, nwords, &right_num, errors))){
+ free(op);
+ free(left);
+ return NULL;
+ }
+ switch(op[0]){
+ case '&':
+ temp = odpairsand(left, left_num, right, right_num, &temp_num);
+ break;
+ case '|':
+ temp = odpairsor(left, left_num, right, right_num, &temp_num);
+ break;
+ case '!':
+ temp = odpairsnotand(left, left_num, right, right_num, &temp_num);
+ break;
+ default:
+ if(errors) cblistpush(errors, "Invalid operator. Expected '&', '|', or '!'.", -1);
+ break;
+ }
+ if(temp){
+ /* an operator was done so we must swap it with the left */
+ free(left); left = NULL;
+ left = temp;
+ left_num = temp_num;
+ }
+ free(op);
+ if(right) free(right);
+ }
+ *np = left_num;
+ return left;
+}
+
+
+/* Processes the tokens in order to break them up further.
+ `odeum' specifies a database handle.
+ `tokens' specifies a list handle of tokens. */
+static void odfixtokens(ODEUM *odeum, CBLIST *tokens){
+ const char *tk = NULL;
+ int tk_len = 0;
+ int i = 0;
+ int lastword = 0;
+ for(i = 0; i < cblistnum(tokens); i++){
+ tk = cblistval(tokens, i, &tk_len);
+ assert(tk);
+ if(tk[0] == '&' || tk[0] == '|' || tk[0] == '!' || tk[0] == '(' || tk[0] == ')'){
+ lastword = 0;
+ if(tk_len > 1){
+ /* need to break it up for the next loop around */
+ tk = cblistremove(tokens, i, &tk_len);
+ cblistinsert(tokens, i, tk, 1);
+ cblistinsert(tokens, i+1, tk+1, tk_len-1);
+ free((char *)tk);
+ }
+ } else if(odeum->statechars[*(unsigned char *)tk] == 0){
+ /* if the last one was a word and this is a word then we need a default & between them */
+ if(lastword){
+ cblistinsert(tokens, i, "&", 1);
+ i++;
+ }
+ lastword = 1;
+ }
+ }
+}
+
+
+/* Cleans out the parts of the normalized word list that are not considered words.
+ `odeum' specifies a database handle.
+ `tokens' specifies a list handle of tokens. */
+static void odcleannormalized(ODEUM *odeum, CBLIST *nwords){
+ char *tk = NULL;
+ int tk_len = 0;
+ int i = 0;
+ for(i = 0; i < cblistnum(nwords); i++){
+ tk = (char *)cblistval(nwords, i, &tk_len);
+ if(tk_len == 0 || (!odeum->statechars[*(unsigned char *)tk] == 0)){
+ /* not a word so delete it */
+ tk = cblistremove(nwords, i, &tk_len);
+ free(tk);
+ i--;
+ }
+ }
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/odeum.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/odeum.h
===================================================================
--- box/trunk/qdbm/odeum.h (rev 0)
+++ box/trunk/qdbm/odeum.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,590 @@
+/*************************************************************************************************
+ * The inverted API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _ODEUM_H /* duplication check */
+#define _ODEUM_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <depot.h>
+#include <curia.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <time.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+typedef struct { /* type of structure for a database handle */
+ char *name; /* name of the database directory */
+ int wmode; /* whether to be writable */
+ int fatal; /* whether a fatal error occured */
+ int inode; /* inode of the database directory */
+ CURIA *docsdb; /* database handle for documents */
+ CURIA *indexdb; /* database handle for the inverted index */
+ VILLA *rdocsdb; /* database handle for the reverse dictionary */
+ CBMAP *cachemap; /* cache for dirty buffers of words */
+ int cacheasiz; /* total allocated size of dirty buffers */
+ CBMAP *sortmap; /* map handle for candidates of sorting */
+ int dmax; /* max number of the document ID */
+ int dnum; /* number of the documents */
+ int ldid; /* ID number of the last registered document */
+ char statechars[256]; /* state of single byte characters */
+} ODEUM;
+
+typedef struct { /* type of structure for a document handle */
+ int id; /* ID number */
+ char *uri; /* uniform resource identifier */
+ CBMAP *attrs; /* map handle for attrubutes */
+ CBLIST *nwords; /* list handle for words in normalized form */
+ CBLIST *awords; /* list handle for words in appearance form */
+} ODDOC;
+
+typedef struct { /* type of structure for an element of search result */
+ int id; /* ID number of the document */
+ int score; /* score of the document */
+} ODPAIR;
+
+enum { /* enumeration for open modes */
+ OD_OREADER = 1 << 0, /* open as a reader */
+ OD_OWRITER = 1 << 1, /* open as a writer */
+ OD_OCREAT = 1 << 2, /* a writer creating */
+ OD_OTRUNC = 1 << 3, /* a writer truncating */
+ OD_ONOLCK = 1 << 4, /* open without locking */
+ OD_OLCKNB = 1 << 5 /* lock without blocking */
+};
+
+
+/* Get a database handle.
+ `name' specifies the name of a database directory.
+ `omode' specifies the connection mode: `OD_OWRITER' as a writer, `OD_OREADER' as a reader.
+ If the mode is `OD_OWRITER', the following may be added by bitwise or: `OD_OCREAT', which
+ means it creates a new database if not exist, `OD_OTRUNC', which means it creates a new
+ database regardless if one exists. Both of `OD_OREADER' and `OD_OWRITER' can be added to by
+ bitwise or: `OD_ONOLCK', which means it opens a database directory without file locking, or
+ `OD_OLCKNB', which means locking is performed without blocking.
+ The return value is the database handle or `NULL' if it is not successful.
+ While connecting as a writer, an exclusive lock is invoked to the database directory.
+ While connecting as a reader, a shared lock is invoked to the database directory.
+ The thread blocks until the lock is achieved. If `OD_ONOLCK' is used, the application is
+ responsible for exclusion control. */
+ODEUM *odopen(const char *name, int omode);
+
+
+/* Close a database handle.
+ `odeum' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ Because the region of a closed handle is released, it becomes impossible to use the handle.
+ Updating a database is assured to be written when the handle is closed. If a writer opens
+ a database but does not close it appropriately, the database will be broken. */
+int odclose(ODEUM *odeum);
+
+
+/* Store a document.
+ `odeum' specifies a database handle connected as a writer.
+ `doc' specifies a document handle.
+ `wmax' specifies the max number of words to be stored in the document database. If it is
+ negative, the number is unlimited.
+ `over' specifies whether the data of the duplicated document is overwritten or not. If it
+ is false and the URI of the document is duplicated, the function returns as an error.
+ If successful, the return value is true, else, it is false. */
+int odput(ODEUM *odeum, ODDOC *doc, int wmax, int over);
+
+
+/* Delete a document specified by a URI.
+ `odeum' specifies a database handle connected as a writer.
+ `uri' specifies the string of the URI of a document.
+ If successful, the return value is true, else, it is false. False is returned when no
+ document corresponds to the specified URI. */
+int odout(ODEUM *odeum, const char *uri);
+
+
+/* Delete a document specified by an ID number.
+ `odeum' specifies a database handle connected as a writer.
+ `id' specifies the ID number of a document.
+ If successful, the return value is true, else, it is false. False is returned when no
+ document corresponds to the specified ID number. */
+int odoutbyid(ODEUM *odeum, int id);
+
+
+/* Retrieve a document specified by a URI.
+ `odeum' specifies a database handle.
+ `uri' specifies the string the URI of a document.
+ If successful, the return value is the handle of the corresponding document, else, it is
+ `NULL'. `NULL' is returned when no document corresponds to the specified URI.
+ Because the handle of the return value is opened with the function `oddocopen', it should
+ be closed with the function `oddocclose'. */
+ODDOC *odget(ODEUM *odeum, const char *uri);
+
+
+/* Retrieve a document by an ID number.
+ `odeum' specifies a database handle.
+ `id' specifies the ID number of a document.
+ If successful, the return value is the handle of the corresponding document, else, it is
+ `NULL'. `NULL' is returned when no document corresponds to the specified ID number.
+ Because the handle of the return value is opened with the function `oddocopen', it should
+ be closed with the function `oddocclose'. */
+ODDOC *odgetbyid(ODEUM *odeum, int id);
+
+
+/* Retrieve the ID of the document specified by a URI.
+ `odeum' specifies a database handle.
+ `uri' specifies the string the URI of a document.
+ If successful, the return value is the ID number of the document, else, it is -1. -1 is
+ returned when no document corresponds to the specified URI. */
+int odgetidbyuri(ODEUM *odeum, const char *uri);
+
+
+/* Check whether the document specified by an ID number exists.
+ `odeum' specifies a database handle.
+ `id' specifies the ID number of a document.
+ The return value is true if the document exists, else, it is false. */
+int odcheck(ODEUM *odeum, int id);
+
+
+/* Search the inverted index for documents including a particular word.
+ `odeum' specifies a database handle.
+ `word' specifies a searching word.
+ `max' specifies the max number of documents to be retrieve.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ If successful, the return value is the pointer to an array, else, it is `NULL'. Each
+ element of the array is a pair of the ID number and the score of a document, and sorted in
+ descending order of their scores. Even if no document corresponds to the specified word,
+ it is not error but returns an dummy array.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. Note that each element of the array
+ of the return value can be data of a deleted document. */
+ODPAIR *odsearch(ODEUM *odeum, const char *word, int max, int *np);
+
+
+/* Get the number of documents including a word.
+ `odeum' specifies a database handle.
+ `word' specifies a searching word.
+ If successful, the return value is the number of documents including the word, else, it is -1.
+ Because this function does not read the entity of the inverted index, it is faster than
+ `odsearch'. */
+int odsearchdnum(ODEUM *odeum, const char *word);
+
+
+/* Initialize the iterator of a database handle.
+ `odeum' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ The iterator is used in order to access every document stored in a database. */
+int oditerinit(ODEUM *odeum);
+
+
+/* Get the next key of the iterator.
+ `odeum' specifies a database handle.
+ If successful, the return value is the handle of the next document, else, it is `NULL'.
+ `NULL' is returned when no document is to be get out of the iterator.
+ It is possible to access every document by iteration of calling this function. However,
+ it is not assured if updating the database is occurred while the iteration. Besides, the
+ order of this traversal access method is arbitrary, so it is not assured that the order of
+ string matches the one of the traversal access. Because the handle of the return value is
+ opened with the function `oddocopen', it should be closed with the function `oddocclose'. */
+ODDOC *oditernext(ODEUM *odeum);
+
+
+/* Synchronize updating contents with the files and the devices.
+ `odeum' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ This function is useful when another process uses the connected database directory. */
+int odsync(ODEUM *odeum);
+
+
+/* Optimize a database.
+ `odeum' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ Elements of the deleted documents in the inverted index are purged. */
+int odoptimize(ODEUM *odeum);
+
+
+/* Get the name of a database.
+ `odeum' specifies a database handle.
+ If successful, the return value is the pointer to the region of the name of the database,
+ else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *odname(ODEUM *odeum);
+
+
+/* Get the total size of database files.
+ `odeum' specifies a database handle.
+ If successful, the return value is the total size of the database files, else, it is -1.0. */
+double odfsiz(ODEUM *odeum);
+
+
+/* Get the total number of the elements of the bucket arrays in the inverted index.
+ `odeum' specifies a database handle.
+ If successful, the return value is the total number of the elements of the bucket arrays,
+ else, it is -1. */
+int odbnum(ODEUM *odeum);
+
+
+/* Get the total number of the used elements of the bucket arrays in the inverted index.
+ `odeum' specifies a database handle.
+ If successful, the return value is the total number of the used elements of the bucket
+ arrays, else, it is -1. */
+int odbusenum(ODEUM *odeum);
+
+
+/* Get the number of the documents stored in a database.
+ `odeum' specifies a database handle.
+ If successful, the return value is the number of the documents stored in the database, else,
+ it is -1. */
+int oddnum(ODEUM *odeum);
+
+
+/* Get the number of the words stored in a database.
+ `odeum' specifies a database handle.
+ If successful, the return value is the number of the words stored in the database, else,
+ it is -1.
+ Because of the I/O buffer, the return value may be less than the hard number. */
+int odwnum(ODEUM *odeum);
+
+
+/* Check whether a database handle is a writer or not.
+ `odeum' specifies a database handle.
+ The return value is true if the handle is a writer, false if not. */
+int odwritable(ODEUM *odeum);
+
+
+/* Check whether a database has a fatal error or not.
+ `odeum' specifies a database handle.
+ The return value is true if the database has a fatal error, false if not. */
+int odfatalerror(ODEUM *odeum);
+
+
+/* Get the inode number of a database directory.
+ `odeum' specifies a database handle.
+ The return value is the inode number of the database directory. */
+int odinode(ODEUM *odeum);
+
+
+/* Get the last modified time of a database.
+ `odeum' specifies a database handle.
+ The return value is the last modified time of the database. */
+time_t odmtime(ODEUM *odeum);
+
+
+/* Merge plural database directories.
+ `name' specifies the name of a database directory to create.
+ `elemnames' specifies a list of names of element databases.
+ If successful, the return value is true, else, it is false.
+ If two or more documents which have the same URL come in, the first one is adopted and the
+ others are ignored. */
+int odmerge(const char *name, const CBLIST *elemnames);
+
+
+/* Remove a database directory.
+ `name' specifies the name of a database directory.
+ If successful, the return value is true, else, it is false.
+ A database directory can contain databases of other APIs of QDBM, they are also removed by
+ this function. */
+int odremove(const char *name);
+
+
+/* Get a document handle.
+ `uri' specifies the URI of a document.
+ The return value is a document handle.
+ The ID number of a new document is not defined. It is defined when the document is stored
+ in a database. */
+ODDOC *oddocopen(const char *uri);
+
+
+/* Close a document handle.
+ `doc' specifies a document handle.
+ Because the region of a closed handle is released, it becomes impossible to use the handle. */
+void oddocclose(ODDOC *doc);
+
+
+/* Add an attribute to a document.
+ `doc' specifies a document handle.
+ `name' specifies the string of the name of an attribute.
+ `value' specifies the string of the value of the attribute. */
+void oddocaddattr(ODDOC *doc, const char *name, const char *value);
+
+
+/* Add a word to a document.
+ `doc' specifies a document handle.
+ `normal' specifies the string of the normalized form of a word. Normalized forms are
+ treated as keys of the inverted index. If the normalized form of a word is an empty
+ string, the word is not reflected in the inverted index.
+ `asis' specifies the string of the appearance form of the word. Appearance forms are used
+ after the document is retrieved by an application. */
+void oddocaddword(ODDOC *doc, const char *normal, const char *asis);
+
+
+/* Get the ID number of a document.
+ `doc' specifies a document handle.
+ The return value is the ID number of a document. */
+int oddocid(const ODDOC *doc);
+
+
+/* Get the URI of a document.
+ `doc' specifies a document handle.
+ The return value is the string of the URI of a document. */
+const char *oddocuri(const ODDOC *doc);
+
+
+/* Get the value of an attribute of a document.
+ `doc' specifies a document handle.
+ `name' specifies the string of the name of an attribute.
+ The return value is the string of the value of the attribute, or `NULL' if no attribute
+ corresponds. */
+const char *oddocgetattr(const ODDOC *doc, const char *name);
+
+
+/* Get the list handle contains words in normalized form of a document.
+ `doc' specifies a document handle.
+ The return value is the list handle contains words in normalized form. */
+const CBLIST *oddocnwords(const ODDOC *doc);
+
+
+/* Get the list handle contains words in appearance form of a document.
+ `doc' specifies a document handle.
+ The return value is the list handle contains words in appearance form. */
+const CBLIST *oddocawords(const ODDOC *doc);
+
+
+/* Get the map handle contains keywords in normalized form and their scores.
+ `doc' specifies a document handle.
+ `max' specifies the max number of keywords to get.
+ `odeum' specifies a database handle with which the IDF for weighting is calculate.
+ If it is `NULL', it is not used.
+ The return value is the map handle contains keywords and their scores. Scores are expressed
+ as decimal strings.
+ Because the handle of the return value is opened with the function `cbmapopen', it should
+ be closed with the function `cbmapclose' if it is no longer in use. */
+CBMAP *oddocscores(const ODDOC *doc, int max, ODEUM *odeum);
+
+
+/* Break a text into words in appearance form.
+ `text' specifies the string of a text.
+ The return value is the list handle contains words in appearance form.
+ Words are separated with space characters and such delimiters as period, comma and so on.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *odbreaktext(const char *text);
+
+
+/* Make the normalized form of a word.
+ `asis' specifies the string of the appearance form of a word.
+ The return value is is the string of the normalized form of the word.
+ Alphabets of the ASCII code are unified into lower cases. Words composed of only delimiters
+ are treated as empty strings. Because the region of the return value is allocated with the
+ `malloc' call, it should be released with the `free' call if it is no longer in use. */
+char *odnormalizeword(const char *asis);
+
+
+/* Get the common elements of two sets of documents.
+ `apairs' specifies the pointer to the former document array.
+ `anum' specifies the number of the elements of the former document array.
+ `bpairs' specifies the pointer to the latter document array.
+ `bnum' specifies the number of the elements of the latter document array.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ The return value is the pointer to a new document array whose elements commonly belong to
+ the specified two sets.
+ Elements of the array are sorted in descending order of their scores. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+ODPAIR *odpairsand(ODPAIR *apairs, int anum, ODPAIR *bpairs, int bnum, int *np);
+
+
+/* Get the sum of elements of two sets of documents.
+ `apairs' specifies the pointer to the former document array.
+ `anum' specifies the number of the elements of the former document array.
+ `bpairs' specifies the pointer to the latter document array.
+ `bnum' specifies the number of the elements of the latter document array.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ The return value is the pointer to a new document array whose elements belong to both or
+ either of the specified two sets.
+ Elements of the array are sorted in descending order of their scores. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+ODPAIR *odpairsor(ODPAIR *apairs, int anum, ODPAIR *bpairs, int bnum, int *np);
+
+
+/* Get the difference set of documents.
+ `apairs' specifies the pointer to the former document array.
+ `anum' specifies the number of the elements of the former document array.
+ `bpairs' specifies the pointer to the latter document array of the sum of elements.
+ `bnum' specifies the number of the elements of the latter document array.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ The return value is the pointer to a new document array whose elements belong to the former
+ set but not to the latter set.
+ Elements of the array are sorted in descending order of their scores. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+ODPAIR *odpairsnotand(ODPAIR *apairs, int anum, ODPAIR *bpairs, int bnum, int *np);
+
+
+/* Sort a set of documents in descending order of scores.
+ `pairs' specifies the pointer to a document array.
+ `pnum' specifies the number of the elements of the document array. */
+void odpairssort(ODPAIR *pairs, int pnum);
+
+
+/* Get the natural logarithm of a number.
+ `x' specifies a number.
+ The return value is the natural logarithm of the number. If the number is equal to or less
+ than 1.0, the return value is 0.0.
+ This function is useful when an application calculates the IDF of search results. */
+double odlogarithm(double x);
+
+
+/* Get the cosine of the angle of two vectors.
+ `avec' specifies the pointer to one array of numbers.
+ `bvec' specifies the pointer to the other array of numbers.
+ `vnum' specifies the number of elements of each array.
+ The return value is the cosine of the angle of two vectors.
+ This function is useful when an application calculates similarity of documents. */
+double odvectorcosine(const int *avec, const int *bvec, int vnum);
+
+
+/* Set the global tuning parameters.
+ `ibnum' specifies the number of buckets for inverted indexes.
+ `idnum' specifies the division number of inverted index.
+ `cbnum' specifies the number of buckets for dirty buffers.
+ `csiz' specifies the maximum bytes to use memory for dirty buffers.
+ The default setting is equivalent to `odsettuning(32749, 7, 262139, 8388608)'. This function
+ should be called before opening a handle. */
+void odsettuning(int ibnum, int idnum, int cbnum, int csiz);
+
+
+/* Break a text into words and store appearance forms and normalized form into lists.
+ `odeum' specifies a database handle.
+ `text' specifies the string of a text.
+ `awords' specifies a list handle into which appearance form is store.
+ `nwords' specifies a list handle into which normalized form is store. If it is `NULL', it is
+ ignored.
+ Words are separated with space characters and such delimiters as period, comma and so on. */
+void odanalyzetext(ODEUM *odeum, const char *text, CBLIST *awords, CBLIST *nwords);
+
+
+/* Set the classes of characters used by `odanalyzetext'.
+ `odeum' specifies a database handle.
+ `spacechars' spacifies a string contains space characters.
+ `delimchars' spacifies a string contains delimiter characters.
+ `gluechars' spacifies a string contains glue characters. */
+void odsetcharclass(ODEUM *odeum, const char *spacechars, const char *delimchars,
+ const char *gluechars);
+
+
+/* Query a database using a small boolean query language.
+ `odeum' specifies a database handle.
+ 'query' specifies the text of the query.
+ `np' specifies the pointer to a variable to which the number of the elements of the return
+ value is assigned.
+ `errors' specifies a list handle into which error messages are stored. If it is `NULL', it
+ is ignored.
+ If successful, the return value is the pointer to an array, else, it is `NULL'. Each
+ element of the array is a pair of the ID number and the score of a document, and sorted in
+ descending order of their scores. Even if no document corresponds to the specified condition,
+ it is not error but returns an dummy array.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. Note that each element of the array
+ of the return value can be data of a deleted document. */
+ODPAIR *odquery(ODEUM *odeum, const char *query, int *np, CBLIST *errors);
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Get the internal database handle for documents.
+ `odeum' specifies a database handle.
+ The return value is the internal database handle for documents.
+ Note that the the returned handle should not be updated. */
+CURIA *odidbdocs(ODEUM *odeum);
+
+
+/* Get the internal database handle for the inverted index.
+ `odeum' specifies a database handle.
+ The return value is the internal database handle for the inverted index.
+ Note that the the returned handle should not be updated. */
+CURIA *odidbindex(ODEUM *odeum);
+
+
+/* Get the internal database handle for the reverse dictionary.
+ `odeum' specifies a database handle.
+ The return value is the internal database handle for the reverse dictionary.
+ Note that the the returned handle should not be updated. */
+VILLA *odidbrdocs(ODEUM *odeum);
+
+
+/* Set the call back function called in merging.
+ `otcb' specifires the pointer to a function to report outturn. Its first argument is the name
+ of processing function. Its second argument is the handle of the database being processed.
+ Its third argument is ths string of a log message. If it is `NULL', the call back function is
+ cleared. */
+void odsetotcb(void (*otcb)(const char *, ODEUM *, const char *));
+
+
+/* Get the positive one of square roots of a number.
+ `x' specifies a number.
+ The return value is the positive one of square roots of a number. If the number is equal to
+ or less than 0.0, the return value is 0.0. */
+double odsquareroot(double x);
+
+
+/* Get the absolute of a vector.
+ `vec' specifies the pointer to an array of numbers.
+ `vnum' specifies the number of elements of the array.
+ The return value is the absolute of a vector. */
+double odvecabsolute(const int *vec, int vnum);
+
+
+/* Get the inner product of two vectors.
+ `avec' specifies the pointer to one array of numbers.
+ `bvec' specifies the pointer to the other array of numbers.
+ `vnum' specifies the number of elements of each array.
+ The return value is the inner product of two vectors. */
+double odvecinnerproduct(const int *avec, const int *bvec, int vnum);
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/odeum.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/odidx.c
===================================================================
--- box/trunk/qdbm/odidx.c (rev 0)
+++ box/trunk/qdbm/odidx.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,890 @@
+/*************************************************************************************************
+ * Utility for indexing document files into a database of Odeum
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+#include <time.h>
+#include <signal.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define PATHCHR '/' /* delimiter character of path */
+#define EXTCHR '.' /* delimiter character of extension */
+#define CDIRSTR "." /* string of current directory */
+#define PDIRSTR ".." /* string of parent directory */
+#define MTDBNAME "_mtime" /* name of the database for last modified times */
+#define MTDBLRM 81 /* records in a leaf node of time database */
+#define MTDBNIM 192 /* records in a non-leaf node of time database */
+#define MTDBLCN 64 /* number of leaf cache of time database */
+#define MTDBNCN 32 /* number of non-leaf cache of time database */
+#define SCDBNAME "_score" /* name of the database for scores */
+#define SCDBBNUM 32749 /* bucket number of the score database */
+#define SCDBALIGN -3 /* alignment of the score database */
+#define PATHBUFSIZ 2048 /* size of a path buffer */
+#define MAXLOAD 0.85 /* max ratio of bucket loading */
+#define KEYNUM 32 /* number of keywords to store */
+
+
+/* for Win32 and RISC OS */
+#if defined(_WIN32)
+#undef PATHCHR
+#define PATHCHR '\\'
+#undef EXTCHR
+#define EXTCHR '.'
+#undef CDIRSTR
+#define CDIRSTR "."
+#undef PDIRSTR
+#define PDIRSTR ".."
+#elif defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#undef PATHCHR
+#define PATHCHR '.'
+#undef EXTCHR
+#define EXTCHR '/'
+#undef CDIRSTR
+#define CDIRSTR "@"
+#undef PDIRSTR
+#define PDIRSTR "^"
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+int sigterm; /* flag for termination signal */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void setsignals(void);
+void sigtermhandler(int num);
+void usage(void);
+int runregister(int argc, char **argv);
+int runrelate(int argc, char **argv);
+int runpurge(int argc, char **argv);
+int bwimatchlist(const char *str, const CBLIST *keys);
+char *fgetl(FILE *ifp);
+void otcb(const char *fname, ODEUM *odeum, const char *msg);
+void pdperror(const char *name);
+void printferror(const char *format, ...);
+void printfinfo(const char *format, ...);
+const char *datestr(time_t t);
+int proclist(const char *name, const char *lfile, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist);
+int procdir(const char *name, const char *dir, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist);
+int indexdir(ODEUM *odeum, VILLA *mtdb, const char *name, const char *dir, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist);
+int indexfile(ODEUM *odeum, VILLA *mtdb, const char *name, const char *file, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist);
+char *filetouri(const char *file);
+ODDOC *makedocplain(const char *uri, const char *text, const char *date);
+ODDOC *makedochtml(const char *uri, const char *html, const char *date);
+CBMAP *htmlescpairs(void);
+int procrelate(const char *name);
+int procpurge(const char *name);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ sigterm = FALSE;
+ setsignals();
+ if(argc < 2) usage();
+ odsetotcb(otcb);
+ rv = 0;
+ if(!strcmp(argv[1], "register")){
+ rv = runregister(argc, argv);
+ } else if(!strcmp(argv[1], "relate")){
+ rv = runrelate(argc, argv);
+ } else if(!strcmp(argv[1], "purge")){
+ rv = runpurge(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* set signal handlers */
+void setsignals(void){
+ signal(1, sigtermhandler);
+ signal(2, sigtermhandler);
+ signal(3, sigtermhandler);
+ signal(13, sigtermhandler);
+ signal(15, sigtermhandler);
+}
+
+
+/* handler of termination signal */
+void sigtermhandler(int num){
+ signal(num, SIG_DFL);
+ sigterm = TRUE;
+ printfinfo("the termination signal %d catched", num);
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: indexer of document files\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s register [-l file] [-wmax num] [-tsuf sufs] [-hsuf sufs] name [dir]\n",
+ progname);
+ fprintf(stderr, " %s relate name\n", progname);
+ fprintf(stderr, " %s purge name\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of register command */
+int runregister(int argc, char **argv){
+ char *name, *dir, *lfile, *tsuf, *hsuf, path[PATHBUFSIZ];
+ int i, wmax, plen, rv;
+ CBLIST *tsuflist, *hsuflist;
+ name = NULL;
+ dir = NULL;
+ lfile = NULL;
+ tsuf = NULL;
+ hsuf = NULL;
+ wmax = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-l")){
+ if(++i >= argc) usage();
+ lfile = argv[i];
+ } else if(!strcmp(argv[i], "-wmax")){
+ if(++i >= argc) usage();
+ wmax = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-tsuf")){
+ if(++i >= argc) usage();
+ tsuf = argv[i];
+ } else if(!strcmp(argv[i], "-hsuf")){
+ if(++i >= argc) usage();
+ hsuf = argv[i];
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!dir){
+ dir = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ if(!dir) dir = CDIRSTR;
+ plen = sprintf(path, "%s", dir);
+ if(plen > 1 && path[plen-1] == PATHCHR) path[plen-1] = '\0';
+ tsuflist = cbsplit(tsuf ? tsuf : ".txt,.text", -1, ",");
+ hsuflist = cbsplit(hsuf ? hsuf : ".html,.htm", -1, ",");
+ if(lfile){
+ rv = proclist(name, lfile, wmax, tsuflist, hsuflist);
+ } else {
+ rv = procdir(name, path, wmax, tsuflist, hsuflist);
+ }
+ cblistclose(hsuflist);
+ cblistclose(tsuflist);
+ return rv;
+}
+
+
+/* parse arguments of relate command */
+int runrelate(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = procrelate(name);
+ return rv;
+}
+
+
+/* parse arguments of purge command */
+int runpurge(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = procpurge(name);
+ return rv;
+}
+
+
+/* case insensitive backward matching with a list */
+int bwimatchlist(const char *str, const CBLIST *keys){
+ int i;
+ for(i = 0; i < cblistnum(keys); i++){
+ if(cbstrbwimatch(str, cblistval(keys, i, NULL))) return TRUE;
+ }
+ return FALSE;
+}
+
+
+/* read a line */
+char *fgetl(FILE *ifp){
+ char *buf;
+ int c, len, blen;
+ buf = NULL;
+ len = 0;
+ blen = 256;
+ while((c = fgetc(ifp)) != EOF){
+ if(blen <= len) blen *= 2;
+ buf = cbrealloc(buf, blen + 1);
+ if(c == '\n') c = '\0';
+ buf[len++] = c;
+ if(c == '\0') break;
+ }
+ if(!buf) return NULL;
+ buf[len] = '\0';
+ return buf;
+}
+
+
+/* report the outturn */
+void otcb(const char *fname, ODEUM *odeum, const char *msg){
+ char *name;
+ name = odname(odeum);
+ printf("%s: %s: %s: %s\n", progname, fname, name, msg);
+ free(name);
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ printf("%s: ERROR: %s: %s\n", progname, name, dperrmsg(dpecode));
+ fflush(stdout);
+}
+
+
+/* print formatted error string and flush the buffer */
+void printferror(const char *format, ...){
+ va_list ap;
+ va_start(ap, format);
+ printf("%s: ERROR: ", progname);
+ vprintf(format, ap);
+ putchar('\n');
+ fflush(stdout);
+ va_end(ap);
+}
+
+
+/* print formatted information string and flush the buffer */
+void printfinfo(const char *format, ...){
+ va_list ap;
+ va_start(ap, format);
+ printf("%s: INFO: ", progname);
+ vprintf(format, ap);
+ putchar('\n');
+ fflush(stdout);
+ va_end(ap);
+}
+
+
+/* get static string of the date */
+const char *datestr(time_t t){
+ static char buf[32];
+ struct tm *stp;
+ if(!(stp = localtime(&t))) return "0000/00/00 00:00:00";
+ sprintf(buf, "%04d/%02d/%02d %02d:%02d:%02d",
+ stp->tm_year + 1900, stp->tm_mon + 1, stp->tm_mday,
+ stp->tm_hour, stp->tm_min, stp->tm_sec);
+ return buf;
+}
+
+
+/* processing with finding files in a list file */
+int proclist(const char *name, const char *lfile, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist){
+ ODEUM *odeum;
+ VILLA *mtdb;
+ FILE *ifp;
+ char *line, path[PATHBUFSIZ];
+ int err, fatal;
+ if(!strcmp(lfile, "-")){
+ ifp = stdin;
+ } else {
+ if(!(ifp = fopen(lfile, "rb"))){
+ printferror("%s: file cannot be opened", lfile);
+ return 1;
+ }
+ }
+ printfinfo("%s: registration started", name);
+ if(!(odeum = odopen(name, OD_OWRITER | OD_OCREAT))){
+ pdperror(name);
+ if(ifp != stdin) fclose(ifp);
+ return 1;
+ }
+ sprintf(path, "%s%c%s", name, PATHCHR, MTDBNAME);
+ if(!(mtdb = vlopen(path, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ pdperror(name);
+ odclose(odeum);
+ if(ifp != stdin) fclose(ifp);
+ return 1;
+ }
+ vlsettuning(mtdb, MTDBLRM, MTDBNIM, MTDBLCN, MTDBNCN);
+ printfinfo("%s: database opened: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ err = FALSE;
+ while((line = fgetl(ifp)) != NULL){
+ if(sigterm){
+ printferror("aborting due to a termination signal");
+ free(line);
+ err = TRUE;
+ break;
+ }
+ if(!indexfile(odeum, mtdb, name, line, wmax, tsuflist, hsuflist)) err = TRUE;
+ free(line);
+ }
+ fatal = odfatalerror(odeum);
+ printfinfo("%s: database closing: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ if(!vlclose(mtdb)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(ifp != stdin) fclose(ifp);
+ if(err){
+ printfinfo("%s: registration was over%s", name, fatal ? " with fatal error" : "");
+ } else {
+ printfinfo("%s: registration completed successfully", name);
+ }
+ return err ? 1 : 0;
+}
+
+
+/* processing with finding files in a directory */
+int procdir(const char *name, const char *dir, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist){
+ ODEUM *odeum;
+ VILLA *mtdb;
+ char path[PATHBUFSIZ];
+ int err, fatal;
+ printfinfo("%s: registration started", name);
+ if(!(odeum = odopen(name, OD_OWRITER | OD_OCREAT))){
+ pdperror(name);
+ return 1;
+ }
+ sprintf(path, "%s%c%s", name, PATHCHR, MTDBNAME);
+ if(!(mtdb = vlopen(path, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ vlsettuning(mtdb, MTDBLRM, MTDBNIM, MTDBLCN, MTDBNCN);
+ printfinfo("%s: database opened: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ err = FALSE;
+ if(!indexdir(odeum, mtdb, name, dir, wmax, tsuflist, hsuflist)) err = TRUE;
+ fatal = odfatalerror(odeum);
+ printfinfo("%s: database closing: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ if(!vlclose(mtdb)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(err){
+ printfinfo("%s: registration was over%s", name, fatal ? " with fatal error" : "");
+ } else {
+ printfinfo("%s: registration completed successfully", name);
+ }
+ return err ? 1 : 0;
+}
+
+
+/* find and index files in a directory */
+int indexdir(ODEUM *odeum, VILLA *mtdb, const char *name, const char *dir, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist){
+ CBLIST *files;
+ const char *file;
+ char path[PATHBUFSIZ];
+ int i, isroot, isdir, err;
+ if(!(files = cbdirlist(dir))){
+ printferror("%s: directory cannot be opened", dir);
+ return FALSE;
+ }
+ isroot = dir[0] == PATHCHR && dir[1] == '\0';
+ err = FALSE;
+ for(i = 0; i < cblistnum(files); i++){
+ if(sigterm){
+ printferror("aborting due to a termination signal");
+ cblistclose(files);
+ return FALSE;
+ }
+ file = cblistval(files, i, NULL);
+ if(!strcmp(file, CDIRSTR) || !strcmp(file, PDIRSTR)) continue;
+ if(isroot){
+ sprintf(path, "%s%s", dir, file);
+ } else {
+ sprintf(path, "%s%c%s", dir, PATHCHR, file);
+ }
+ if(!cbfilestat(path, &isdir, NULL, NULL)){
+ printferror("%s: file does not exist", file);
+ err = TRUE;
+ continue;
+ }
+ if(isdir){
+ if(!indexdir(odeum, mtdb, name, path, wmax, tsuflist, hsuflist)) err = TRUE;
+ } else {
+ if(!indexfile(odeum, mtdb, name, path, wmax, tsuflist, hsuflist)) err = TRUE;
+ }
+ }
+ cblistclose(files);
+ return err ? FALSE : TRUE;
+}
+
+
+/* index a file into the database */
+int indexfile(ODEUM *odeum, VILLA *mtdb, const char *name, const char *file, int wmax,
+ const CBLIST *tsuflist, const CBLIST *hsuflist){
+ static int cnt = 0;
+ char *vbuf, *buf, *uri;
+ const char *title;
+ int size, hot, vsiz, wnum, bnum;
+ time_t mtime;
+ ODDOC *doc;
+ if(!cbfilestat(file, NULL, &size, &mtime)){
+ printferror("%s: file does not exist", file);
+ return FALSE;
+ }
+ hot = TRUE;
+ if((vbuf = vlget(mtdb, file, -1, &vsiz)) != NULL){
+ if(vsiz == sizeof(int) && mtime <= *(int *)vbuf) hot = FALSE;
+ free(vbuf);
+ }
+ if(!hot){
+ printfinfo("%s: passed", file);
+ return TRUE;
+ }
+ doc = NULL;
+ uri = filetouri(file);
+ if(bwimatchlist(file, tsuflist)){
+ if(!(buf = cbreadfile(file, NULL))){
+ printferror("%s: file cannot be opened", file);
+ return FALSE;
+ }
+ doc = makedocplain(uri, buf, datestr(mtime));
+ free(buf);
+ } else if(bwimatchlist(file, hsuflist)){
+ if(!(buf = cbreadfile(file, NULL))){
+ printferror("%s: file cannot be opened", file);
+ return FALSE;
+ }
+ doc = makedochtml(uri, buf, datestr(mtime));
+ free(buf);
+ }
+ free(uri);
+ if(doc){
+ if(!(title = oddocgetattr(doc, "title")) || strlen(title) < 1){
+ if((title = strrchr(file, PATHCHR)) != NULL){
+ title++;
+ } else {
+ title = file;
+ }
+ oddocaddattr(doc, "title", title);
+ }
+ if(odput(odeum, doc, wmax, TRUE) &&
+ vlput(mtdb, file, -1, (char *)&mtime, sizeof(int), VL_DOVER)){
+ printfinfo("%s: registered: id=%d wnum=%d",
+ file, oddocid(doc), cblistnum(oddocnwords(doc)));
+ cnt++;
+ } else {
+ pdperror(file);
+ }
+ oddocclose(doc);
+ }
+ wnum = odwnum(odeum);
+ bnum = odbnum(odeum);
+ if(wnum != -1 && bnum != -1 && (double)wnum / (double)bnum > MAXLOAD){
+ printfinfo("%s: optimizing started: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ if(!odoptimize(odeum)){
+ pdperror(file);
+ return FALSE;
+ }
+ printfinfo("%s: optimizing completed: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ }
+ if(cnt >= 256){
+ printfinfo("%s: database status: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ cnt = 0;
+ }
+ return TRUE;
+}
+
+
+/* make the url from file path */
+char *filetouri(const char *file){
+ CBLIST *list;
+ char str[PATHBUFSIZ], *wp, *enc;
+ const char *name;
+ int i, nsiz;
+ sprintf(str, "%c", PATHCHR);
+ list = cbsplit(file, -1, str);
+ wp = str;
+ for(i = 0; i < cblistnum(list); i++){
+ if(i > 0) *(wp++) = '/';
+ name = cblistval(list, i, &nsiz);
+ enc = cburlencode(name, nsiz);
+ wp += sprintf(wp, "%s", enc);
+ free(enc);
+ }
+ cblistclose(list);
+ *wp = '\0';
+ return cbmemdup(str, -1);
+}
+
+
+/* make a document of plain text */
+ODDOC *makedocplain(const char *uri, const char *text, const char *date){
+ ODDOC *doc;
+ CBLIST *awords;
+ const char *asis;
+ char *normal;
+ int i;
+ doc = oddocopen(uri);
+ if(date) oddocaddattr(doc, "date", date);
+ awords = odbreaktext(text);
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ normal = odnormalizeword(asis);
+ oddocaddword(doc, normal, asis);
+ free(normal);
+ }
+ cblistclose(awords);
+ return doc;
+}
+
+
+/* make a document of HTML */
+ODDOC *makedochtml(const char *uri, const char *html, const char *date){
+ ODDOC *doc;
+ CBMAP *pairs;
+ CBLIST *elems, *awords;
+ const char *text, *asis;
+ char *rtext, *normal;
+ int i, j, body;
+ pairs = htmlescpairs();
+ doc = oddocopen(uri);
+ if(date) oddocaddattr(doc, "date", date);
+ elems = cbxmlbreak(html, TRUE);
+ body = FALSE;
+ for(i = 0; i < cblistnum(elems); i++){
+ text = cblistval(elems, i, NULL);
+ if(cbstrfwimatch(text, "<title")){
+ i++;
+ if(i < cblistnum(elems)){
+ text = cblistval(elems, i, NULL);
+ if(text[0] == '<') text = "";
+ rtext = cbreplace(text, pairs);
+ for(j = 0; rtext[j] != '\0'; j++){
+ if(strchr("\t\n\v\f\r", rtext[j])) rtext[j] = ' ';
+ }
+ while(--j >= 0){
+ if(rtext[j] != ' ') break;
+ rtext[j] = '\0';
+ }
+ for(j = 0; rtext[j] != '\0'; j++){
+ if(rtext[j] != ' ') break;
+ }
+ oddocaddattr(doc, "title", rtext + j);
+ awords = odbreaktext(rtext);
+ for(j = 0; j < cblistnum(awords); j++){
+ asis = cblistval(awords, j, NULL);
+ normal = odnormalizeword(asis);
+ oddocaddword(doc, normal, "");
+ free(normal);
+ }
+ cblistclose(awords);
+ free(rtext);
+ }
+ } else if(cbstrfwimatch(text, "<body")){
+ body = TRUE;
+ } else if(body && text[0] != '<'){
+ rtext = cbreplace(text, pairs);
+ awords = odbreaktext(rtext);
+ for(j = 0; j < cblistnum(awords); j++){
+ asis = cblistval(awords, j, NULL);
+ normal = odnormalizeword(asis);
+ oddocaddword(doc, normal, asis);
+ free(normal);
+ }
+ cblistclose(awords);
+ free(rtext);
+ }
+ }
+ if(!body){
+ for(i = 0; i < cblistnum(elems); i++){
+ text = cblistval(elems, i, NULL);
+ if(cbstrfwimatch(text, "<title")){
+ i++;
+ } else if(text[0] != '<'){
+ rtext = cbreplace(text, pairs);
+ awords = odbreaktext(rtext);
+ for(j = 0; j < cblistnum(awords); j++){
+ asis = cblistval(awords, j, NULL);
+ normal = odnormalizeword(asis);
+ oddocaddword(doc, normal, asis);
+ free(normal);
+ }
+ cblistclose(awords);
+ free(rtext);
+ }
+ }
+ }
+ cblistclose(elems);
+ return doc;
+}
+
+
+/* get pairs of escaping characters */
+CBMAP *htmlescpairs(void){
+ char *latinext[] = {
+ " ", "!", "(cent)", "(pound)", "(currency)", "(yen)", "|", "(section)", "\"", "(C)",
+ "", "<<", "(not)", "-", "(R)", "~", "(degree)", "+-", "^2", "^3",
+ "'", "(u)", "(P)", "*", ",", "^1", "", ">>", "(1/4)", "(1/2)",
+ "(3/4)", "?", "A", "A", "A", "A", "A", "A", "AE", "C",
+ "E", "E", "E", "E", "I", "I", "I", "I", "D", "N",
+ "O", "O", "O", "O", "O", "*", "O", "U", "U", "U",
+ "U", "Y", "P", "s", "a", "a", "a", "a", "a", "a",
+ "ae", "c", "e", "e", "e", "e", "i", "i", "i", "i",
+ "o", "n", "o", "o", "o", "o", "o", "/", "o", "u",
+ "u", "u", "u", "y", "p", "y", NULL
+ };
+ static CBMAP *pairs = NULL;
+ char kbuf[8], vbuf[8];
+ int i, ksiz, vsiz;
+ if(pairs) return pairs;
+ pairs = cbmapopen();
+ cbglobalgc(pairs, (void (*)(void *))cbmapclose);
+ cbmapput(pairs, "&", -1, "&", -1, TRUE);
+ cbmapput(pairs, "<", -1, "<", -1, TRUE);
+ cbmapput(pairs, ">", -1, ">", -1, TRUE);
+ cbmapput(pairs, """, -1, "\"", -1, TRUE);
+ cbmapput(pairs, "'", -1, "'", -1, TRUE);
+ cbmapput(pairs, " ", -1, " ", -1, TRUE);
+ cbmapput(pairs, "©", -1, "(C)", -1, TRUE);
+ cbmapput(pairs, "®", -1, "(R)", -1, TRUE);
+ cbmapput(pairs, "™", -1, "(TM)", -1, TRUE);
+ for(i = 1; i <= 127; i++){
+ ksiz = sprintf(kbuf, "&#%d;", i);
+ vsiz = sprintf(vbuf, "%c", i);
+ cbmapput(pairs, kbuf, ksiz, vbuf, vsiz, TRUE);
+ }
+ cbmapput(pairs, "‚", -1, ",", -1, TRUE);
+ cbmapput(pairs, "„", -1, ",,", -1, TRUE);
+ cbmapput(pairs, "…", -1, "...", -1, TRUE);
+ cbmapput(pairs, "‹", -1, "<", -1, TRUE);
+ cbmapput(pairs, "‘", -1, "'", -1, TRUE);
+ cbmapput(pairs, "’", -1, "'", -1, TRUE);
+ cbmapput(pairs, "“", -1, "\"", -1, TRUE);
+ cbmapput(pairs, "”", -1, "\"", -1, TRUE);
+ cbmapput(pairs, "–", -1, "-", -1, TRUE);
+ cbmapput(pairs, "—", -1, "-", -1, TRUE);
+ cbmapput(pairs, "˜", -1, "~", -1, TRUE);
+ cbmapput(pairs, "™", -1, "(TM)", -1, TRUE);
+ cbmapput(pairs, "›", -1, ">", -1, TRUE);
+ for(i = 0; latinext[i]; i++){
+ ksiz = sprintf(kbuf, "&#%d;", i + 160);
+ cbmapput(pairs, kbuf, ksiz, latinext[i], -1, TRUE);
+ }
+ return pairs;
+}
+
+
+/* register scores of documents */
+int procrelate(const char *name){
+ ODEUM *odeum;
+ DEPOT *scdb;
+ ODDOC *doc;
+ CBMAP *scores;
+ const char *file;
+ char path[PATHBUFSIZ], *mbuf;
+ int err, fatal, id, msiz;
+ printfinfo("%s: relating started", name);
+ if(!(odeum = odopen(name, OD_OWRITER))){
+ pdperror(name);
+ return 1;
+ }
+ sprintf(path, "%s%c%s", name, PATHCHR, SCDBNAME);
+ if(!(scdb = dpopen(path, OD_OWRITER | OD_OCREAT, SCDBBNUM))){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ if(!dpsetalign(scdb, SCDBALIGN)){
+ pdperror(name);
+ dpclose(scdb);
+ odclose(odeum);
+ return 1;
+ }
+ printfinfo("%s: database opened: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ err = FALSE;
+ if(!oditerinit(odeum)){
+ pdperror(name);
+ err = TRUE;
+ } else {
+ while(TRUE){
+ if(sigterm){
+ printferror("aborting due to a termination signal");
+ err = TRUE;
+ break;
+ }
+ if(!(doc = oditernext(odeum))){
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ break;
+ }
+ file = oddocuri(doc);
+ id = oddocid(doc);
+ scores = oddocscores(doc, KEYNUM, odeum);
+ mbuf = cbmapdump(scores, &msiz);
+ if(!dpput(scdb, (char *)&id, sizeof(int), mbuf, msiz, DP_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ } else {
+ printfinfo("%s: related", file);
+ }
+ free(mbuf);
+ cbmapclose(scores);
+ oddocclose(doc);
+ if(err) break;
+ }
+ }
+ fatal = odfatalerror(odeum);
+ printfinfo("%s: database closing: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ if(!dpclose(scdb)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(err){
+ printfinfo("%s: relating was over%s", name, fatal ? " with fatal error" : "");
+ } else {
+ printfinfo("%s: relating completed successfully", name);
+ }
+ return err ? 1 : 0;
+}
+
+
+/* purge documents which is not existing. */
+int procpurge(const char *name){
+ ODEUM *odeum;
+ ODDOC *doc;
+ const char *file;
+ int err, fatal;
+ printfinfo("%s: purging started", name);
+ if(!(odeum = odopen(name, OD_OWRITER))){
+ pdperror(name);
+ return 1;
+ }
+ printfinfo("%s: database opened: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ err = FALSE;
+ if(!oditerinit(odeum)){
+ pdperror(name);
+ err = TRUE;
+ } else {
+ while(TRUE){
+ if(sigterm){
+ printferror("aborting due to a termination signal");
+ err = TRUE;
+ break;
+ }
+ if(!(doc = oditernext(odeum))){
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ break;
+ }
+ file = oddocuri(doc);
+ if(cbfilestat(file, NULL, NULL, NULL)){
+ printfinfo("%s: passed", file);
+ } else {
+ if(!odout(odeum, file)){
+ pdperror(file);
+ err = TRUE;
+ }
+ printfinfo("%s: purged", file);
+ }
+ oddocclose(doc);
+ }
+ }
+ fatal = odfatalerror(odeum);
+ printfinfo("%s: database closing: fsiz=%.0f dnum=%d wnum=%d bnum=%d",
+ name, odfsiz(odeum), oddnum(odeum), odwnum(odeum), odbnum(odeum));
+ if(!odclose(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(err){
+ printfinfo("%s: purging was over%s", name, fatal ? " with fatal error" : "");
+ } else {
+ printfinfo("%s: purging completed successfully", name);
+ }
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/odidx.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/odmgr.c
===================================================================
--- box/trunk/qdbm/odmgr.c (rev 0)
+++ box/trunk/qdbm/odmgr.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,1085 @@
+/*************************************************************************************************
+ * Utility for debugging Odeum and its applications
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define MAXSRCHWORDS 256 /* max number of search words */
+#define WOCCRPOINT 10000 /* points per occurence */
+#define MAXKEYWORDS 8 /* max number of keywords */
+#define SUMMARYWIDTH 16 /* width of each phrase in a summary */
+#define MAXSUMMARY 128 /* max number of words in a summary */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *readstdin(int *sp);
+void otcb(const char *fname, ODEUM *odeum, const char *msg);
+int runcreate(int argc, char **argv);
+int runput(int argc, char **argv);
+int runout(int argc, char **argv);
+int runget(int argc, char **argv);
+int runsearch(int argc, char **argv);
+int runlist(int argc, char **argv);
+int runoptimize(int argc, char **argv);
+int runinform(int argc, char **argv);
+int runmerge(int argc, char **argv);
+int runremove(int argc, char **argv);
+int runbreak(int argc, char **argv);
+void pdperror(const char *name);
+void printdoc(const ODDOC *doc, int tb, int hb, int score, ODEUM *odeum, const CBLIST *skeys);
+char *docsummary(const ODDOC *doc, const CBLIST *kwords, int num, int hilight);
+CBMAP *listtomap(const CBLIST *list);
+int docreate(const char *name);
+int doput(const char *name, const char *text, const char *uri, const char *title,
+ const char *author, const char *date, int wmax, int keep);
+int doout(const char *name, const char *uri, int id);
+int doget(const char *name, const char *uri, int id, int tb, int hb);
+int dosearch(const char *name, const char *text, int max, int or, int idf, int ql,
+ int tb, int hb, int nb);
+int dolist(const char *name, int tb, int hb);
+int dooptimize(const char *name);
+int doinform(const char *name);
+int domerge(const char *name, const CBLIST *elems);
+int doremove(const char *name);
+int dobreak(const char *text, int hb, int kb, int sb);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "create")){
+ odsetotcb(otcb);
+ rv = runcreate(argc, argv);
+ } else if(!strcmp(argv[1], "put")){
+ odsetotcb(otcb);
+ rv = runput(argc, argv);
+ } else if(!strcmp(argv[1], "out")){
+ odsetotcb(otcb);
+ rv = runout(argc, argv);
+ } else if(!strcmp(argv[1], "get")){
+ rv = runget(argc, argv);
+ } else if(!strcmp(argv[1], "search")){
+ rv = runsearch(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else if(!strcmp(argv[1], "optimize")){
+ odsetotcb(otcb);
+ rv = runoptimize(argc, argv);
+ } else if(!strcmp(argv[1], "inform")){
+ rv = runinform(argc, argv);
+ } else if(!strcmp(argv[1], "merge")){
+ odsetotcb(otcb);
+ rv = runmerge(argc, argv);
+ } else if(!strcmp(argv[1], "remove")){
+ rv = runremove(argc, argv);
+ } else if(!strcmp(argv[1], "break")){
+ rv = runbreak(argc, argv);
+ } else if(!strcmp(argv[1], "version") || !strcmp(argv[1], "--version")){
+ printf("Powered by QDBM version %s\n", dpversion);
+ printf("Copyright (c) 2000-2007 Mikio Hirabayashi\n");
+ rv = 0;
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: administration utility for Odeum\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s create name\n", progname);
+ fprintf(stderr, " %s put [-uri str] [-title str] [-author str] [-date str]"
+ " [-wmax num] [-keep] name [file]\n", progname);
+ fprintf(stderr, " %s out [-id] name expr\n", progname);
+ fprintf(stderr, " %s get [-id] [-t|-h] name expr\n", progname);
+ fprintf(stderr, " %s search [-max num] [-or] [-idf] [-t|-h|-n] name words...\n", progname);
+ fprintf(stderr, " %s list [-t|-h] name\n", progname);
+ fprintf(stderr, " %s optimize name\n", progname);
+ fprintf(stderr, " %s inform name\n", progname);
+ fprintf(stderr, " %s merge name elems...\n", progname);
+ fprintf(stderr, " %s remove name\n", progname);
+ fprintf(stderr, " %s break [-h|-k|-s] [file]\n", progname);
+ fprintf(stderr, " %s version\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* read the standard input */
+char *readstdin(int *sp){
+ char *buf;
+ int i, blen, c;
+ blen = 256;
+ buf = cbmalloc(blen);
+ for(i = 0; (c = getchar()) != EOF; i++){
+ if(i >= blen - 1) buf = cbrealloc(buf, blen *= 2);
+ buf[i] = c;
+ }
+ buf[i] = '\0';
+ *sp = i;
+ return buf;
+}
+
+
+/* report the outturn */
+void otcb(const char *fname, ODEUM *odeum, const char *msg){
+ char *name;
+ name = odname(odeum);
+ printf("%s: %s: %s: %s\n", progname, fname, name, msg);
+ free(name);
+}
+
+
+/* parse arguments of create command */
+int runcreate(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docreate(name);
+ return rv;
+}
+
+
+/* parse arguments of put command */
+int runput(int argc, char **argv){
+ char *name, *file, *uri, *title, *author, *date, *text;
+ int i, wmax, keep, size, rv;
+ name = NULL;
+ file = NULL;
+ uri = NULL;
+ title = NULL;
+ author = NULL;
+ date = NULL;
+ wmax = -1;
+ keep = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-uri")){
+ if(++i >= argc) usage();
+ uri = argv[i];
+ } else if(!strcmp(argv[i], "-uri")){
+ if(++i >= argc) usage();
+ uri = argv[i];
+ } else if(!strcmp(argv[i], "-title")){
+ if(++i >= argc) usage();
+ title = argv[i];
+ } else if(!strcmp(argv[i], "-author")){
+ if(++i >= argc) usage();
+ author = argv[i];
+ } else if(!strcmp(argv[i], "-date")){
+ if(++i >= argc) usage();
+ date = argv[i];
+ } else if(!strcmp(argv[i], "-wmax")){
+ if(++i >= argc) usage();
+ wmax = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-keep")){
+ keep = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ if(!uri) uri = file;
+ if(!uri) usage();
+ if(file){
+ if(!(text = cbreadfile(file, &size))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ text = readstdin(&size);
+ }
+ rv = doput(name, text, uri, title, author, date, wmax, keep);
+ free(text);
+ return rv;
+}
+
+
+/* parse arguments of out command */
+int runout(int argc, char **argv){
+ char *name, *expr;
+ int i, ib, id, rv;
+ name = NULL;
+ expr = NULL;
+ ib = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-id")){
+ ib = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!expr){
+ expr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !expr) usage();
+ id = -1;
+ if(ib){
+ id = atoi(expr);
+ if(id < 1) usage();
+ }
+ rv = doout(name, expr, id);
+ return rv;
+}
+
+
+/* parse arguments of get command */
+int runget(int argc, char **argv){
+ char *name, *expr;
+ int i, ib, tb, hb, id, rv;
+ name = NULL;
+ expr = NULL;
+ ib = FALSE;
+ tb = FALSE;
+ hb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-id")){
+ ib = TRUE;
+ } else if(!strcmp(argv[i], "-t")){
+ tb = TRUE;
+ } else if(!strcmp(argv[i], "-h")){
+ hb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!expr){
+ expr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !expr) usage();
+ id = -1;
+ if(ib){
+ id = atoi(expr);
+ if(id < 1) usage();
+ }
+ rv = doget(name, expr, id, tb, hb);
+ return rv;
+}
+
+
+/* parse arguments of search command */
+int runsearch(int argc, char **argv){
+ char *name, *srchwords[MAXSRCHWORDS];
+ int i, wnum, max, or, idf, ql, tb, hb, nb, rv;
+ CBDATUM *text;
+ name = NULL;
+ wnum = 0;
+ max = -1;
+ or = FALSE;
+ idf = FALSE;
+ ql = FALSE;
+ tb = FALSE;
+ hb = FALSE;
+ nb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-max")){
+ if(++i >= argc) usage();
+ max = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-or")){
+ or = TRUE;
+ } else if(!strcmp(argv[i], "-idf")){
+ idf = TRUE;
+ } else if(!strcmp(argv[i], "-ql")){
+ ql = TRUE;
+ } else if(!strcmp(argv[i], "-t")){
+ tb = TRUE;
+ } else if(!strcmp(argv[i], "-h")){
+ hb = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(wnum < MAXSRCHWORDS){
+ srchwords[wnum++] = argv[i];
+ }
+ }
+ if(!name) usage();
+ text = cbdatumopen(NULL, -1);
+ for(i = 0; i < wnum; i++){
+ if(i > 0) cbdatumcat(text, " ", 1);
+ cbdatumcat(text, srchwords[i], -1);
+ }
+ rv = dosearch(name, cbdatumptr(text), max, or, idf, ql, tb, hb, nb);
+ cbdatumclose(text);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ char *name;
+ int i, tb, hb, rv;
+ name = NULL;
+ tb = FALSE;
+ hb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-t")){
+ tb = TRUE;
+ } else if(!strcmp(argv[i], "-h")){
+ hb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dolist(name, tb, hb);
+ return rv;
+}
+
+
+/* parse arguments of optimize command */
+int runoptimize(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dooptimize(name);
+ return rv;
+}
+
+
+/* parse arguments of inform command */
+int runinform(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doinform(name);
+ return rv;
+}
+
+
+/* parse arguments of merge command */
+int runmerge(int argc, char **argv){
+ char *name;
+ CBLIST *elems;
+ int i, rv;
+ name = NULL;
+ elems = cblistopen();
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ cblistpush(elems, argv[i], -1);
+ }
+ }
+ if(!name) usage();
+ if(cblistnum(elems) < 1){
+ cblistclose(elems);
+ usage();
+ }
+ rv = domerge(name, elems);
+ cblistclose(elems);
+ return rv;
+}
+
+
+/* parse arguments of remove command */
+int runremove(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doremove(name);
+ return rv;
+}
+
+
+/* parse arguments of break command */
+int runbreak(int argc, char **argv){
+ char *file, *text;
+ int i, hb, kb, sb, size, rv;
+ file = NULL;
+ hb = FALSE;
+ kb = FALSE;
+ sb = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!file && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-h")){
+ hb = TRUE;
+ } else if(!strcmp(argv[i], "-k")){
+ kb = TRUE;
+ } else if(!strcmp(argv[i], "-s")){
+ sb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(file){
+ if(!(text = cbreadfile(file, &size))){
+ fprintf(stderr, "%s: %s: cannot open\n", progname, file);
+ return 1;
+ }
+ } else {
+ text = readstdin(&size);
+ }
+ rv = dobreak(text, hb, kb, sb);
+ free(text);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* print the contents of a document */
+void printdoc(const ODDOC *doc, int tb, int hb, int score, ODEUM *odeum, const CBLIST *skeys){
+ const CBLIST *words;
+ CBMAP *scores;
+ CBLIST *kwords;
+ const char *title, *author, *word, *date;
+ char *summary;
+ int i, wsiz;
+ title = oddocgetattr(doc, "title");
+ author = oddocgetattr(doc, "author");
+ date = oddocgetattr(doc, "date");
+ if(hb){
+ printf("ID: %d\n", oddocid(doc));
+ printf("URI: %s\n", oddocuri(doc));
+ if(title) printf("TITLE: %s\n", title);
+ if(author) printf("AUTHOR: %s\n", author);
+ if(date) printf("DATE: %s\n", date);
+ if(score >= 0) printf("SCORE: %d\n", score);
+ scores = oddocscores(doc, MAXKEYWORDS, odeum);
+ kwords = cblistopen();
+ printf("KEYWORDS: ");
+ cbmapiterinit(scores);
+ while((word = cbmapiternext(scores, &wsiz)) != NULL){
+ if(cblistnum(kwords) > 0) printf(", ");
+ printf("%s (%s)", word, cbmapget(scores, word, wsiz, NULL));
+ cblistpush(kwords, word, wsiz);
+ }
+ putchar('\n');
+ summary = docsummary(doc, skeys ? skeys : kwords, MAXSUMMARY, skeys != NULL);
+ printf("SUMMARY: %s\n", summary);
+ free(summary);
+ cblistclose(kwords);
+ cbmapclose(scores);
+ printf("\n\n");
+ } else if(tb){
+ printf("%d\t%s\t%s\t%s\t%s\t%d\n", oddocid(doc), oddocuri(doc),
+ title ? title : "", author ? author : "", date ? date : "", score);
+ words = oddocnwords(doc);
+ for(i = 0; i < cblistnum(words); i++){
+ word = cblistval(words, i, &wsiz);
+ if(i > 0) putchar('\t');
+ printf("%s", word);
+ }
+ putchar('\n');
+ words = oddocawords(doc);
+ for(i = 0; i < cblistnum(words); i++){
+ word = cblistval(words, i, &wsiz);
+ if(i > 0) putchar('\t');
+ printf("%s", word);
+ }
+ putchar('\n');
+ } else {
+ printf("%d\t%s\t%d\n", oddocid(doc), oddocuri(doc), score);
+ }
+}
+
+
+/* get a list handle contains summary of a document */
+char *docsummary(const ODDOC *doc, const CBLIST *kwords, int num, int hilight){
+ const CBLIST *nwords, *awords;
+ CBMAP *kmap, *map;
+ const char *normal, *asis;
+ char *sbuf;
+ int i, j, bsiz, ssiz, lnum, nwsiz, awsiz, pv, bi, first;
+ bsiz = 256;
+ sbuf = cbmalloc(bsiz);
+ ssiz = 0;
+ nwords = oddocnwords(doc);
+ awords = oddocawords(doc);
+ kmap = listtomap(kwords);
+ map = listtomap(kwords);
+ lnum = cblistnum(nwords);
+ first = TRUE;
+ for(i = 0; i < lnum && i < SUMMARYWIDTH; i++){
+ normal = cblistval(nwords, i, &nwsiz);
+ asis = cblistval(awords, i, &awsiz);
+ if(awsiz < 1) continue;
+ cbmapout(map, normal, nwsiz);
+ if(ssiz + awsiz + 16 >= bsiz){
+ bsiz = bsiz * 2 + awsiz;
+ sbuf = cbrealloc(sbuf, bsiz);
+ }
+ if(!first) ssiz += sprintf(sbuf + ssiz, " ");
+ if(hilight && normal[0] != '\0' && cbmapget(kmap, normal, nwsiz, NULL)){
+ ssiz += sprintf(sbuf + ssiz, "<<%s>>", asis);
+ } else {
+ ssiz += sprintf(sbuf + ssiz, "%s", asis);
+ }
+ first = FALSE;
+ num--;
+ }
+ ssiz += sprintf(sbuf + ssiz, " ...");
+ pv = i;
+ while(i < lnum){
+ if(cbmaprnum(map) < 1){
+ cbmapclose(map);
+ map = listtomap(kwords);
+ }
+ normal = cblistval(nwords, i, &nwsiz);
+ if(cbmapget(map, normal, nwsiz, NULL)){
+ bi = i - SUMMARYWIDTH / 2;
+ bi = bi > pv ? bi : pv;
+ for(j = bi; j < lnum && j <= bi + SUMMARYWIDTH; j++){
+ normal = cblistval(nwords, j, &nwsiz);
+ asis = cblistval(awords, j, &awsiz);
+ if(awsiz < 1) continue;
+ cbmapout(map, normal, nwsiz);
+ if(ssiz + awsiz + 16 >= bsiz){
+ bsiz = bsiz * 2 + awsiz;
+ sbuf = cbrealloc(sbuf, bsiz);
+ }
+ ssiz += sprintf(sbuf + ssiz, " ");
+ if(hilight && normal[0] != '\0' && cbmapget(kmap, normal, nwsiz, NULL)){
+ ssiz += sprintf(sbuf + ssiz, "<<%s>>", asis);
+ } else {
+ ssiz += sprintf(sbuf + ssiz, "%s", asis);
+ }
+ num--;
+ }
+ ssiz += sprintf(sbuf + ssiz, " ...");
+ i = j;
+ pv = i;
+ } else {
+ i++;
+ }
+ if(num <= 0) break;
+ }
+ cbmapclose(map);
+ cbmapclose(kmap);
+ return sbuf;
+}
+
+
+/* get a map made from a list */
+CBMAP *listtomap(const CBLIST *list){
+ CBMAP *map;
+ const char *tmp;
+ int i, tsiz;
+ map = cbmapopen();
+ for(i = 0; i < cblistnum(list); i++){
+ tmp = cblistval(list, i, &tsiz);
+ cbmapput(map, tmp, tsiz, "", 0, FALSE);
+ }
+ return map;
+}
+
+
+/* perform create command */
+int docreate(const char *name){
+ ODEUM *odeum;
+ if(!(odeum = odopen(name, OD_OWRITER | OD_OCREAT | OD_OTRUNC))){
+ pdperror(name);
+ return 1;
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform put command */
+int doput(const char *name, const char *text, const char *uri, const char *title,
+ const char *author, const char *date, int wmax, int keep){
+ ODEUM *odeum;
+ ODDOC *doc;
+ CBLIST *awords;
+ const char *asis;
+ char *normal;
+ int i;
+ if(!(odeum = odopen(name, OD_OWRITER))){
+ pdperror(name);
+ return 1;
+ }
+ doc = oddocopen(uri);
+ if(title) oddocaddattr(doc, "title", title);
+ if(author) oddocaddattr(doc, "author", author);
+ if(date) oddocaddattr(doc, "date", date);
+ awords = odbreaktext(text);
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ normal = odnormalizeword(asis);
+ oddocaddword(doc, normal, asis);
+ free(normal);
+ }
+ cblistclose(awords);
+ if(!odput(odeum, doc, wmax, keep ? FALSE : TRUE)){
+ pdperror(name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ oddocclose(doc);
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform out command */
+int doout(const char *name, const char *uri, int id){
+ ODEUM *odeum;
+ if(!(odeum = odopen(name, OD_OWRITER))){
+ pdperror(name);
+ return 1;
+ }
+ if(id > 0){
+ if(!odoutbyid(odeum, id)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ } else {
+ if(!odout(odeum, uri)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform get command */
+int doget(const char *name, const char *uri, int id, int tb, int hb){
+ ODEUM *odeum;
+ ODDOC *doc;
+ if(!(odeum = odopen(name, OD_OREADER))){
+ pdperror(name);
+ return 1;
+ }
+ if(id > 0){
+ if(!(doc = odgetbyid(odeum, id))){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ } else {
+ if(!(doc = odget(odeum, uri))){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ printdoc(doc, tb, hb, -1, odeum, NULL);
+ oddocclose(doc);
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform search command */
+int dosearch(const char *name, const char *text, int max, int or, int idf, int ql,
+ int tb, int hb, int nb){
+ ODEUM *odeum;
+ CBLIST *awords, *nwords, *uris, *hits;
+ ODPAIR *pairs, *last, *tmp;
+ ODDOC *doc;
+ const char *asis;
+ char *normal, numbuf[32];
+ int i, j, pnum, lnum, hnum, tnum, shows;
+ double ival;
+ if(!(odeum = odopen(name, OD_OREADER))){
+ pdperror(name);
+ return 1;
+ }
+ awords = odbreaktext(text);
+ nwords = cblistopen();
+ uris = cblistopen();
+ hits = cblistopen();
+ last = NULL;
+ lnum = 0;
+ if(ql){
+ last= odquery(odeum, text, &lnum, NULL);
+ } else {
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ normal = odnormalizeword(asis);
+ cblistpush(nwords, normal, -1);
+ if(strlen(normal) < 1){
+ free(normal);
+ continue;
+ }
+ if(!(pairs = odsearch(odeum, normal, or ? max : -1, &pnum))){
+ pdperror(name);
+ free(normal);
+ continue;
+ }
+ if((hnum = odsearchdnum(odeum, normal)) < 0) hnum = 0;
+ if(idf){
+ ival = odlogarithm(hnum);
+ ival = (ival * ival) / 4.0;
+ if(ival < 4.0) ival = 4.0;
+ for(j = 0; j < pnum; j++){
+ pairs[j].score = (int)(pairs[j].score / ival);
+ }
+ }
+ cblistpush(uris, normal, -1);
+ sprintf(numbuf, "%d", hnum);
+ cblistpush(hits, numbuf, -1);
+ if(last){
+ if(or){
+ tmp = odpairsor(last, lnum, pairs, pnum, &tnum);
+ } else {
+ tmp = odpairsand(last, lnum, pairs, pnum, &tnum);
+ }
+ free(last);
+ free(pairs);
+ last = tmp;
+ lnum = tnum;
+ } else {
+ last = pairs;
+ lnum = pnum;
+ }
+ free(normal);
+ }
+ }
+ if(hb){
+ printf("TOTAL: %d\n", lnum);
+ printf("EACHWORD: ");
+ } else {
+ printf("%d", lnum);
+ }
+ for(i = 0; i < cblistnum(uris); i++){
+ if(hb){
+ if(i > 0) printf(", ");
+ printf("%s(%s)", cblistval(uris, i, NULL), cblistval(hits, i, NULL));
+ } else {
+ printf("\t%s\t%s", cblistval(uris, i, NULL), cblistval(hits, i, NULL));
+ }
+ }
+ putchar('\n');
+ if(hb) putchar('\n');
+ if(last){
+ if(max < 0) max = lnum;
+ shows = 0;
+ for(i = 0; i < lnum && shows < max; i++){
+ if(nb){
+ printf("%d\t%d\n", last[i].id, last[i].score);
+ shows++;
+ } else {
+ if(!(doc = odgetbyid(odeum, last[i].id))) continue;
+ printdoc(doc, tb, hb, last[i].score, odeum, nwords);
+ oddocclose(doc);
+ shows++;
+ }
+ }
+ free(last);
+ }
+ cblistclose(uris);
+ cblistclose(hits);
+ cblistclose(nwords);
+ cblistclose(awords);
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform list command */
+int dolist(const char *name, int tb, int hb){
+ ODEUM *odeum;
+ ODDOC *doc;
+ if(!(odeum = odopen(name, OD_OREADER))){
+ pdperror(name);
+ return 1;
+ }
+ if(!oditerinit(odeum)){
+ odclose(odeum);
+ pdperror(name);
+ return 1;
+ }
+ while(TRUE){
+ if(!(doc = oditernext(odeum))){
+ if(dpecode == DP_ENOITEM) break;
+ odclose(odeum);
+ pdperror(name);
+ return 1;
+ }
+ printdoc(doc, tb, hb, -1, odeum, NULL);
+ oddocclose(doc);
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform optimize command */
+int dooptimize(const char *name){
+ ODEUM *odeum;
+ if(!(odeum = odopen(name, OD_OWRITER))){
+ pdperror(name);
+ return 1;
+ }
+ if(!odoptimize(odeum)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform inform command */
+int doinform(const char *name){
+ ODEUM *odeum;
+ char *tmp;
+ if(!(odeum = odopen(name, OD_OREADER))){
+ pdperror(name);
+ return 1;
+ }
+ tmp = odname(odeum);
+ printf("name: %s\n", tmp ? tmp : "(null)");
+ free(tmp);
+ printf("file size: %.0f\n", odfsiz(odeum));
+ printf("index buckets: %d\n", odbnum(odeum));
+ printf("used buckets: %d\n", odbusenum(odeum));
+ printf("all documents: %d\n", oddnum(odeum));
+ printf("all words: %d\n", odwnum(odeum));
+ printf("inode number: %d\n", odinode(odeum));
+ printf("modified time: %.0f\n", (double)odmtime(odeum));
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform merge command */
+int domerge(const char *name, const CBLIST *elems){
+ if(!odmerge(name, elems)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform remove command */
+int doremove(const char *name){
+ if(!odremove(name)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform break command */
+int dobreak(const char *text, int hb, int kb, int sb){
+ CBLIST *awords, *kwords;
+ CBMAP *scores;
+ ODDOC *doc;
+ const char *asis;
+ char *normal, *summary;
+ int i, first;
+ awords = odbreaktext(text);
+ if(kb || sb){
+ doc = oddocopen("");
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ normal = odnormalizeword(asis);
+ oddocaddword(doc, normal, asis);
+ free(normal);
+ }
+ scores = oddocscores(doc, MAXKEYWORDS, NULL);
+ cbmapiterinit(scores);
+ kwords = cbmapkeys(scores);
+ if(kb){
+ for(i = 0; i < cblistnum(kwords); i++){
+ if(i > 0) putchar('\t');
+ printf("%s", cblistval(kwords, i, NULL));
+ }
+ putchar('\n');
+ } else {
+ summary = docsummary(doc, kwords, MAXSUMMARY, FALSE);
+ printf("%s\n", summary);
+ free(summary);
+ }
+ cblistclose(kwords);
+ cbmapclose(scores);
+ oddocclose(doc);
+ } else if(hb){
+ printf("NWORDS: ");
+ first = TRUE;
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ normal = odnormalizeword(asis);
+ if(normal[0] == '\0'){
+ free(normal);
+ continue;
+ }
+ if(!first) putchar(' ');
+ first = FALSE;
+ printf("%s", normal);
+ free(normal);
+ }
+ putchar('\n');
+ printf("AWORDS: ");
+ first = TRUE;
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ if(asis[0] == '\0') continue;
+ if(!first) putchar(' ');
+ first = FALSE;
+ printf("%s", asis);
+ }
+ putchar('\n');
+ } else {
+ for(i = 0; i < cblistnum(awords); i++){
+ asis = cblistval(awords, i, NULL);
+ normal = odnormalizeword(asis);
+ printf("%s\t%s\n", normal, asis);
+ free(normal);
+ }
+ }
+ cblistclose(awords);
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/odmgr.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/odtest.c
===================================================================
--- box/trunk/qdbm/odtest.c (rev 0)
+++ box/trunk/qdbm/odtest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,694 @@
+/*************************************************************************************************
+ * Test cases of Odeum
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+#include <time.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define DOCBUFSIZ 256 /* buffer for documents */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runwrite(int argc, char **argv);
+int runread(int argc, char **argv);
+int runcombo(int argc, char **argv);
+int runwicked(int argc, char **argv);
+int printfflush(const char *format, ...);
+void pdperror(const char *name);
+int myrand(void);
+ODDOC *makedoc(int id, int wnum, int pnum);
+int dowrite(const char *name, int dnum, int wnum, int pnum,
+ int ibnum, int idnum, int cbnum, int csiz);
+int doread(const char *name);
+int docombo(const char *name);
+int dowicked(const char *name, int dnum);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "write")){
+ rv = runwrite(argc, argv);
+ } else if(!strcmp(argv[1], "read")){
+ rv = runread(argc, argv);
+ } else if(!strcmp(argv[1], "combo")){
+ rv = runcombo(argc, argv);
+ } else if(!strcmp(argv[1], "wicked")){
+ rv = runwicked(argc, argv);
+ } else {
+ usage();
+ }
+ return 0;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Odeum\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s write [-tune ibnum idnum cbnum csiz] name dnum wnum pnum\n", progname);
+ fprintf(stderr, " %s read name\n", progname);
+ fprintf(stderr, " %s combo name\n", progname);
+ fprintf(stderr, " %s wicked name dnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of write command */
+int runwrite(int argc, char **argv){
+ char *name, *dstr, *wstr, *pstr;
+ int i, dnum, wnum, pnum, ibnum, idnum, cbnum, csiz, rv;
+ name = NULL;
+ dstr = NULL;
+ wstr = NULL;
+ pstr = NULL;
+ dnum = 0;
+ wnum = 0;
+ pnum = 0;
+ ibnum = -1;
+ idnum = -1;
+ cbnum = -1;
+ csiz = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-tune")){
+ if(++i >= argc) usage();
+ ibnum = atoi(argv[i]);
+ if(++i >= argc) usage();
+ idnum = atoi(argv[i]);
+ if(++i >= argc) usage();
+ cbnum = atoi(argv[i]);
+ if(++i >= argc) usage();
+ csiz = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!dstr){
+ dstr = argv[i];
+ } else if(!wstr){
+ wstr = argv[i];
+ } else if(!pstr){
+ pstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !dstr || !wstr || !pstr) usage();
+ dnum = atoi(dstr);
+ wnum = atoi(wstr);
+ pnum = atoi(pstr);
+ if(dnum < 1 || wnum < 1 || pnum < 1) usage();
+ rv = dowrite(name, dnum, wnum, pnum, ibnum, idnum, cbnum, csiz);
+ return rv;
+}
+
+
+/* parse arguments of read command */
+int runread(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doread(name);
+ return rv;
+}
+
+
+/* parse arguments of combo command */
+int runcombo(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docombo(name);
+ return rv;
+}
+
+
+/* parse arguments of wicked command */
+int runwicked(int argc, char **argv){
+ char *name, *dstr;
+ int i, dnum, rv;
+ name = NULL;
+ dstr = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else if(!dstr){
+ dstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !dstr) usage();
+ dnum = atoi(dstr);
+ if(dnum < 1) usage();
+ rv = dowicked(name, dnum);
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* pseudo random number generator */
+int myrand(void){
+ static int cnt = 0;
+ if(cnt == 0) srand(time(NULL));
+ return (rand() * rand() + (rand() >> (sizeof(int) * 4)) + (cnt++)) & INT_MAX;
+}
+
+
+/* create a document */
+ODDOC *makedoc(int id, int wnum, int pnum){
+ ODDOC *doc;
+ char buf[DOCBUFSIZ];
+ int i;
+ sprintf(buf, "%08d", id);
+ doc = oddocopen(buf);
+ oddocaddattr(doc, "title", buf);
+ oddocaddattr(doc, "author", buf);
+ oddocaddattr(doc, "date", buf);
+ for(i = 0; i < wnum; i++){
+ sprintf(buf, "%08d", myrand() % pnum);
+ oddocaddword(doc, buf, buf);
+ }
+ return doc;
+}
+
+
+/* perform write command */
+int dowrite(const char *name, int dnum, int wnum, int pnum,
+ int ibnum, int idnum, int cbnum, int csiz){
+ ODEUM *odeum;
+ ODDOC *doc;
+ int i, err;
+ printfflush("<Writing Test>\n name=%s dnum=%d wnum=%d pnum=%d"
+ " ibnum=%d idnum=%d cbnum=%d csiz=%d\n\n",
+ name, dnum, wnum, pnum, ibnum, idnum, cbnum, csiz);
+ /* open a database */
+ if(ibnum > 0) odsettuning(ibnum, idnum, cbnum, csiz);
+ if(!(odeum = odopen(name, OD_OWRITER | OD_OCREAT | OD_OTRUNC))){
+ pdperror(name);
+ return 1;
+ }
+ err = FALSE;
+ /* loop for each document */
+ for(i = 1; i <= dnum; i++){
+ /* store a document */
+ doc = makedoc(i, wnum, pnum);
+ if(!odput(odeum, doc, -1, FALSE)){
+ pdperror(name);
+ oddocclose(doc);
+ err = TRUE;
+ break;
+ }
+ oddocclose(doc);
+ /* print progression */
+ if(dnum > 250 && i % (dnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == dnum || i % (dnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform read command */
+int doread(const char *name){
+ ODEUM *odeum;
+ ODDOC *doc;
+ char buf[DOCBUFSIZ];
+ int i, dnum, err;
+ printfflush("<Reading Test>\n name=%s\n\n", name);
+ /* open a database */
+ if(!(odeum = odopen(name, OD_OREADER))){
+ pdperror(name);
+ return 1;
+ }
+ /* get the number of documents */
+ dnum = oddnum(odeum);
+ err = FALSE;
+ /* loop for each document */
+ for(i = 1; i <= dnum; i++){
+ /* retrieve a document */
+ sprintf(buf, "%08d", i);
+ if(!(doc = odget(odeum, buf))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ oddocclose(doc);
+ /* print progression */
+ if(dnum > 250 && i % (dnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == dnum || i % (dnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform combo command */
+int docombo(const char *name){
+ ODEUM *odeum;
+ ODDOC *doc;
+ const CBLIST *nwords, *awords;
+ CBLIST *tawords, *tnwords, *oawords;
+ ODPAIR *pairs;
+ const char *asis;
+ char buf[DOCBUFSIZ], *normal;
+ int i, j, pnum;
+ printfflush("<Combination Test>\n name=%s\n\n", name);
+ printfflush("Creating a database with ... ");
+ if(!(odeum = odopen(name, OD_OWRITER | OD_OCREAT | OD_OTRUNC))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 20 documents including about 200 words ... ");
+ for(i = 1; i <= 20; i++){
+ sprintf(buf, "%08d", i);
+ doc = makedoc(i, 120 + myrand() % 160, myrand() % 500 + 500);
+ if(!odput(odeum, doc, 180 + myrand() % 40, FALSE)){
+ pdperror(name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ oddocclose(doc);
+ }
+ printfflush("ok\n");
+ printfflush("Checking documents ... ");
+ for(i = 1; i <= 20; i++){
+ sprintf(buf, "%08d", i);
+ if(!(doc = odget(odeum, buf))){
+ pdperror(name);
+ return 1;
+ }
+ nwords = oddocnwords(doc);
+ awords = oddocawords(doc);
+ if(!oddocuri(doc) || !oddocgetattr(doc, "title") || cblistnum(nwords) != cblistnum(awords)){
+ fprintf(stderr, "%s: %s: invalid document\n", progname, name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ for(j = 0; j < cblistnum(nwords); j++){
+ if(strcmp(cblistval(nwords, j, NULL), cblistval(nwords, j, NULL))){
+ fprintf(stderr, "%s: %s: invalid words\n", progname, name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ oddocclose(doc);
+ }
+ printfflush("ok\n");
+ printfflush("Syncing the database ... ");
+ if(!odsync(odeum)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Overwriting 1 - 10 documents ... ");
+ for(i = 1; i <= 10; i++){
+ sprintf(buf, "%08d", i);
+ doc = makedoc(i, 120 + myrand() % 160, myrand() % 500 + 500);
+ if(!odput(odeum, doc, 180 + myrand() % 40, TRUE)){
+ pdperror(name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ oddocclose(doc);
+ }
+ printfflush("ok\n");
+ printfflush("Deleting 11 - 20 documents ... ");
+ for(i = 11; i <= 20; i++){
+ sprintf(buf, "%08d", i);
+ if(!odout(odeum, buf)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking documents ... ");
+ for(i = 1; i <= 10; i++){
+ sprintf(buf, "%08d", i);
+ if(!(doc = odget(odeum, buf))){
+ pdperror(name);
+ return 1;
+ }
+ nwords = oddocnwords(doc);
+ awords = oddocawords(doc);
+ if(!oddocuri(doc) || !oddocgetattr(doc, "title") || cblistnum(nwords) != cblistnum(awords)){
+ fprintf(stderr, "%s: %s: invalid document\n", progname, name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ for(j = 0; j < cblistnum(nwords); j++){
+ if(strcmp(cblistval(nwords, j, NULL), cblistval(nwords, j, NULL))){
+ fprintf(stderr, "%s: %s: invalid words\n", progname, name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ oddocclose(doc);
+ }
+ if(oddnum(odeum) != 10){
+ fprintf(stderr, "%s: %s: invalid document number\n", progname, name);
+ odclose(odeum);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Optimizing the database ... ");
+ if(!odoptimize(odeum)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 10 documents including about 200 words ... ");
+ for(i = 11; i <= 20; i++){
+ sprintf(buf, "%08d", i);
+ doc = makedoc(i, 120 + myrand() % 160, myrand() % 500 + 500);
+ if(!odput(odeum, doc, 180 + myrand() % 40, FALSE)){
+ pdperror(name);
+ oddocclose(doc);
+ odclose(odeum);
+ return 1;
+ }
+ oddocclose(doc);
+ }
+ printfflush("ok\n");
+ printfflush("Deleting 6 - 15 documents ... ");
+ for(i = 6; i <= 15; i++){
+ sprintf(buf, "%08d", i);
+ if(!odout(odeum, buf)){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Retrieving documents 100 times ... ");
+ for(i = 1; i <= 100; i++){
+ sprintf(buf, "%08d", myrand() % 1000 + 1);
+ if((pairs = odsearch(odeum, buf, -1, &pnum)) != NULL){
+ for(j = 0; j < pnum; j++){
+ if((doc = odgetbyid(odeum, pairs[j].id)) != NULL){
+ oddocclose(doc);
+ } else if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ free(pairs);
+ } else if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ odclose(odeum);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Analyzing text ... ");
+ tawords = cblistopen();
+ tnwords = cblistopen();
+ odanalyzetext(odeum, "I'd like to ++see++ Mr. X-men tomorrow.", tawords, tnwords);
+ odanalyzetext(odeum, "=== :-) SMILE . @ . SAD :-< ===", tawords, tnwords);
+ for(i = 0; i < DOCBUFSIZ - 1; i++){
+ buf[i] = myrand() % 255 + 1;
+ }
+ buf[DOCBUFSIZ-1] = '\0';
+ cblistclose(tnwords);
+ cblistclose(tawords);
+ for(i = 0; i < 1000; i++){
+ for(j = 0; j < DOCBUFSIZ - 1; j++){
+ if((j + 1) % 32 == 0){
+ buf[j] = ' ';
+ } else {
+ buf[j] = myrand() % 255 + 1;
+ }
+ }
+ buf[DOCBUFSIZ-1] = '\0';
+ tawords = cblistopen();
+ tnwords = cblistopen();
+ odanalyzetext(odeum, buf, tawords, tnwords);
+ oawords = odbreaktext(buf);
+ if(cblistnum(tawords) != cblistnum(oawords) || cblistnum(tnwords) != cblistnum(oawords)){
+ fprintf(stderr, "%s: %s: invalid analyzing\n", progname, name);
+ cblistclose(oawords);
+ cblistclose(tnwords);
+ cblistclose(tawords);
+ odclose(odeum);
+ return 1;
+ }
+ for(j = 0; j < cblistnum(oawords); j++){
+ asis = cblistval(oawords, j, NULL);
+ normal = odnormalizeword(asis);
+ if(strcmp(asis, cblistval(oawords, j, NULL)) || strcmp(normal, cblistval(tnwords, j, NULL))){
+ fprintf(stderr, "%s: %s: invalid analyzing\n", progname, name);
+ free(normal);
+ cblistclose(oawords);
+ cblistclose(tnwords);
+ cblistclose(tawords);
+ odclose(odeum);
+ return 1;
+ }
+ free(normal);
+ }
+ cblistclose(oawords);
+ cblistclose(tnwords);
+ cblistclose(tawords);
+ }
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("all ok\n\n");
+ return 0;
+}
+
+
+/* perform wicked command */
+int dowicked(const char *name, int dnum){
+ ODEUM *odeum;
+ ODDOC *doc;
+ ODPAIR *pairs;
+ char buf[DOCBUFSIZ];
+ int i, j, pnum, err;
+ printfflush("<Wicked Writing Test>\n name=%s dnum=%d\n\n", name, dnum);
+ err = FALSE;
+ if(!(odeum = odopen(name, OD_OWRITER | OD_OCREAT | OD_OTRUNC))){
+ pdperror(name);
+ return 1;
+ }
+ for(i = 1; i <= dnum; i++){
+ switch(myrand() % 8){
+ case 1:
+ putchar('K');
+ doc = makedoc(myrand() % dnum + 1, myrand() % 10 + 10, myrand() % dnum + 500);
+ if(!odput(odeum, doc, 5, FALSE) && dpecode != DP_EKEEP) err = TRUE;
+ oddocclose(doc);
+ break;
+ case 3:
+ putchar('D');
+ if(!odoutbyid(odeum, myrand() % dnum + 1) && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 4:
+ putchar('R');
+ sprintf(buf, "%08d", myrand() % (dnum + 500) + 1);
+ if((pairs = odsearch(odeum, buf, 5, &pnum)) != NULL){
+ if(myrand() % 5 == 0){
+ for(j = 0; j < pnum; j++){
+ if((doc = odgetbyid(odeum, pairs[j].id)) != NULL){
+ oddocclose(doc);
+ } else if(dpecode != DP_ENOITEM){
+ err = TRUE;
+ break;
+ }
+ }
+ }
+ free(pairs);
+ } else if(dpecode != DP_ENOITEM){
+ err = TRUE;
+ }
+ break;
+ default:
+ putchar('O');
+ doc = makedoc(myrand() % dnum + 1, myrand() % 10 + 10, myrand() % dnum + 500);
+ if(!odput(odeum, doc, 5, TRUE)) err = TRUE;
+ oddocclose(doc);
+ break;
+ }
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ if(err){
+ pdperror(name);
+ break;
+ }
+ }
+ if(!odoptimize(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= dnum; i++){
+ doc = makedoc(i, 5, 5);
+ if(!odput(odeum, doc, 5, FALSE) && dpecode != DP_EKEEP){
+ pdperror(name);
+ oddocclose(doc);
+ err = TRUE;
+ break;
+ }
+ oddocclose(doc);
+ putchar(':');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!odoptimize(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= dnum; i++){
+ sprintf(buf, "%08d", i);
+ if(!(doc = odget(odeum, buf))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ oddocclose(doc);
+ putchar('=');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!oditerinit(odeum)){
+ pdperror(name);
+ err = TRUE;
+ }
+ for(i = 1; i <= dnum; i++){
+ if(!(doc = oditernext(odeum))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ oddocclose(doc);
+ putchar('@');
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ }
+ if(!odclose(odeum)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/odtest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/qdbm.def
===================================================================
--- box/trunk/qdbm/qdbm.def (rev 0)
+++ box/trunk/qdbm/qdbm.def 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,424 @@
+EXPORTS
+ VL_CMPDEC = VL_CMPDEC DATA
+ VL_CMPINT = VL_CMPINT DATA
+ VL_CMPLEX = VL_CMPLEX DATA
+ VL_CMPNUM = VL_CMPNUM DATA
+ VST_CMPDEC = VST_CMPDEC DATA
+ VST_CMPINT = VST_CMPINT DATA
+ VST_CMPLEX = VST_CMPLEX DATA
+ VST_CMPNUM = VST_CMPNUM DATA
+ cbfatalfunc = cbfatalfunc DATA
+ dpdbgfd = dpdbgfd DATA
+ dpisreentrant = dpisreentrant DATA
+ dpsysname = dpsysname DATA
+ dpversion = dpversion DATA
+ gdbm_version = gdbm_version DATA
+ odcachebnum = odcachebnum DATA
+ odcachesiz = odcachesiz DATA
+ odindexbnum = odindexbnum DATA
+ odindexdnum = odindexdnum DATA
+ odotcb = odotcb DATA
+ cbbasedecode = cbbasedecode
+ cbbaseencode = cbbaseencode
+ cbbzdecode = cbbzdecode
+ cbbzencode = cbbzencode
+ cbcalendar = cbcalendar
+ cbcsvcells = cbcsvcells
+ cbcsvescape = cbcsvescape
+ cbcsvrows = cbcsvrows
+ cbcsvunescape = cbcsvunescape
+ cbdatestrhttp = cbdatestrhttp
+ cbdatestrwww = cbdatestrwww
+ cbdatumcat = cbdatumcat
+ cbdatumclose = cbdatumclose
+ cbdatumdup = cbdatumdup
+ cbdatumopen = cbdatumopen
+ cbdatumopenbuf = cbdatumopenbuf
+ cbdatumprintf = cbdatumprintf
+ cbdatumptr = cbdatumptr
+ cbdatumsetbuf = cbdatumsetbuf
+ cbdatumsetsize = cbdatumsetsize
+ cbdatumsize = cbdatumsize
+ cbdatumtomalloc = cbdatumtomalloc
+ cbdayofweek = cbdayofweek
+ cbdeflate = cbdeflate
+ cbdirlist = cbdirlist
+ cbencname = cbencname
+ cbfilestat = cbfilestat
+ cbfree = cbfree
+ cbgetcrc = cbgetcrc
+ cbggcsweep = cbggcsweep
+ cbglobalgc = cbglobalgc
+ cbgzdecode = cbgzdecode
+ cbgzencode = cbgzencode
+ cbheapclose = cbheapclose
+ cbheapdup = cbheapdup
+ cbheapinsert = cbheapinsert
+ cbheapnum = cbheapnum
+ cbheapopen = cbheapopen
+ cbheaptomalloc = cbheaptomalloc
+ cbheapval = cbheapval
+ cbhsort = cbhsort
+ cbiconv = cbiconv
+ cbinflate = cbinflate
+ cbisort = cbisort
+ cbjetlag = cbjetlag
+ cblistbsearch = cblistbsearch
+ cblistclose = cblistclose
+ cblistdump = cblistdump
+ cblistdup = cblistdup
+ cblistinsert = cblistinsert
+ cblistload = cblistload
+ cblistlsearch = cblistlsearch
+ cblistnum = cblistnum
+ cblistopen = cblistopen
+ cblistover = cblistover
+ cblistpop = cblistpop
+ cblistpush = cblistpush
+ cblistpushbuf = cblistpushbuf
+ cblistremove = cblistremove
+ cblistshift = cblistshift
+ cblistsort = cblistsort
+ cblistunshift = cblistunshift
+ cblistval = cblistval
+ cblzodecode = cblzodecode
+ cblzoencode = cblzoencode
+ cbmalloc = cbmalloc
+ cbmapclose = cbmapclose
+ cbmapdump = cbmapdump
+ cbmapdup = cbmapdup
+ cbmapget = cbmapget
+ cbmapiterinit = cbmapiterinit
+ cbmapiternext = cbmapiternext
+ cbmapiterval = cbmapiterval
+ cbmapkeys = cbmapkeys
+ cbmapload = cbmapload
+ cbmaploadone = cbmaploadone
+ cbmapmove = cbmapmove
+ cbmapopen = cbmapopen
+ cbmapopenex = cbmapopenex
+ cbmapout = cbmapout
+ cbmapput = cbmapput
+ cbmapputcat = cbmapputcat
+ cbmaprnum = cbmaprnum
+ cbmapvals = cbmapvals
+ cbmemdup = cbmemdup
+ cbmimebreak = cbmimebreak
+ cbmimedecode = cbmimedecode
+ cbmimeencode = cbmimeencode
+ cbmimeparts = cbmimeparts
+ cbmyfatal = cbmyfatal
+ cbproctime = cbproctime
+ cbqsort = cbqsort
+ cbquotedecode = cbquotedecode
+ cbquoteencode = cbquoteencode
+ cbreadfile = cbreadfile
+ cbreadlines = cbreadlines
+ cbrealloc = cbrealloc
+ cbremove = cbremove
+ cbreplace = cbreplace
+ cbsplit = cbsplit
+ cbsprintf = cbsprintf
+ cbssort = cbssort
+ cbstdiobin = cbstdiobin
+ cbstrbwimatch = cbstrbwimatch
+ cbstrbwmatch = cbstrbwmatch
+ cbstrcountutf = cbstrcountutf
+ cbstrcututf = cbstrcututf
+ cbstrfwimatch = cbstrfwimatch
+ cbstrfwmatch = cbstrfwmatch
+ cbstricmp = cbstricmp
+ cbstrmktime = cbstrmktime
+ cbstrsqzspc = cbstrsqzspc
+ cbstrstrbm = cbstrstrbm
+ cbstrstrkmp = cbstrstrkmp
+ cbstrtolower = cbstrtolower
+ cbstrtoupper = cbstrtoupper
+ cbstrtrim = cbstrtrim
+ cburlbreak = cburlbreak
+ cburldecode = cburldecode
+ cburlencode = cburlencode
+ cburlresolve = cburlresolve
+ cbvmemavail = cbvmemavail
+ cbwritefile = cbwritefile
+ cbxmlattrs = cbxmlattrs
+ cbxmlbreak = cbxmlbreak
+ cbxmlescape = cbxmlescape
+ cbxmlunescape = cbxmlunescape
+ crbnum = crbnum
+ crbusenum = crbusenum
+ crclose = crclose
+ crexportdb = crexportdb
+ crfatalerror = crfatalerror
+ crfsiz = crfsiz
+ crfsizd = crfsizd
+ crget = crget
+ crgetflags = crgetflags
+ crgetlob = crgetlob
+ crgetlobfd = crgetlobfd
+ crgetwb = crgetwb
+ crimportdb = crimportdb
+ crinode = crinode
+ criterinit = criterinit
+ criternext = criternext
+ crmemflush = crmemflush
+ crmemsync = crmemsync
+ crmtime = crmtime
+ crname = crname
+ cropen = cropen
+ croptimize = croptimize
+ crout = crout
+ croutlob = croutlob
+ crput = crput
+ crputlob = crputlob
+ crremove = crremove
+ crrepair = crrepair
+ crrnum = crrnum
+ crrnumlob = crrnumlob
+ crsetalign = crsetalign
+ crsetfbpsiz = crsetfbpsiz
+ crsetflags = crsetflags
+ crsnaffle = crsnaffle
+ crsync = crsync
+ crvsiz = crvsiz
+ crvsizlob = crvsizlob
+ crwritable = crwritable
+ dbm_clearerr = dbm_clearerr
+ dbm_close = dbm_close
+ dbm_delete = dbm_delete
+ dbm_dirfno = dbm_dirfno
+ dbm_error = dbm_error
+ dbm_fetch = dbm_fetch
+ dbm_firstkey = dbm_firstkey
+ dbm_nextkey = dbm_nextkey
+ dbm_open = dbm_open
+ dbm_pagfno = dbm_pagfno
+ dbm_rdonly = dbm_rdonly
+ dbm_store = dbm_store
+ dpbnum = dpbnum
+ dpbusenum = dpbusenum
+ dpclose = dpclose
+ dpecodeptr = dpecodeptr
+ dpecodeset = dpecodeset
+ dperrmsg = dperrmsg
+ dpexportdb = dpexportdb
+ dpfatalerror = dpfatalerror
+ dpfdesc = dpfdesc
+ dpfsiz = dpfsiz
+ dpget = dpget
+ dpgetflags = dpgetflags
+ dpgetwb = dpgetwb
+ dpimportdb = dpimportdb
+ dpinnerhash = dpinnerhash
+ dpinode = dpinode
+ dpiterinit = dpiterinit
+ dpiternext = dpiternext
+ dpmemflush = dpmemflush
+ dpmemsync = dpmemsync
+ dpmtime = dpmtime
+ dpname = dpname
+ dpopen = dpopen
+ dpoptimize = dpoptimize
+ dpout = dpout
+ dpouterhash = dpouterhash
+ dpprimenum = dpprimenum
+ dpput = dpput
+ dpremove = dpremove
+ dprepair = dprepair
+ dprnum = dprnum
+ dpsetalign = dpsetalign
+ dpsetfbpsiz = dpsetfbpsiz
+ dpsetflags = dpsetflags
+ dpsnaffle = dpsnaffle
+ dpsync = dpsync
+ dpvsiz = dpvsiz
+ dpwritable = dpwritable
+ gdbm_close = gdbm_close
+ gdbm_delete = gdbm_delete
+ gdbm_errnoptr = gdbm_errnoptr
+ gdbm_exists = gdbm_exists
+ gdbm_fdesc = gdbm_fdesc
+ gdbm_fetch = gdbm_fetch
+ gdbm_firstkey = gdbm_firstkey
+ gdbm_nextkey = gdbm_nextkey
+ gdbm_open = gdbm_open
+ gdbm_open2 = gdbm_open2
+ gdbm_reorganize = gdbm_reorganize
+ gdbm_setopt = gdbm_setopt
+ gdbm_store = gdbm_store
+ gdbm_strerror = gdbm_strerror
+ gdbm_sync = gdbm_sync
+ odanalyzetext = odanalyzetext
+ odbnum = odbnum
+ odbreaktext = odbreaktext
+ odbusenum = odbusenum
+ odcheck = odcheck
+ odclose = odclose
+ oddnum = oddnum
+ oddocaddattr = oddocaddattr
+ oddocaddword = oddocaddword
+ oddocawords = oddocawords
+ oddocclose = oddocclose
+ oddocgetattr = oddocgetattr
+ oddocid = oddocid
+ oddocnwords = oddocnwords
+ oddocopen = oddocopen
+ oddocscores = oddocscores
+ oddocuri = oddocuri
+ odfatalerror = odfatalerror
+ odfsiz = odfsiz
+ odget = odget
+ odgetbyid = odgetbyid
+ odgetidbyuri = odgetidbyuri
+ odidbdocs = odidbdocs
+ odidbindex = odidbindex
+ odidbrdocs = odidbrdocs
+ odinode = odinode
+ oditerinit = oditerinit
+ oditernext = oditernext
+ odlogarithm = odlogarithm
+ odmerge = odmerge
+ odmtime = odmtime
+ odname = odname
+ odnormalizeword = odnormalizeword
+ odopen = odopen
+ odoptimize = odoptimize
+ odout = odout
+ odoutbyid = odoutbyid
+ odpairsand = odpairsand
+ odpairsnotand = odpairsnotand
+ odpairsor = odpairsor
+ odpairssort = odpairssort
+ odput = odput
+ odquery = odquery
+ odremove = odremove
+ odsearch = odsearch
+ odsearchdnum = odsearchdnum
+ odsetcharclass = odsetcharclass
+ odsetotcb = odsetotcb
+ odsettuning = odsettuning
+ odsquareroot = odsquareroot
+ odsync = odsync
+ odvecabsolute = odvecabsolute
+ odvecinnerproduct = odvecinnerproduct
+ odvectorcosine = odvectorcosine
+ odwnum = odwnum
+ odwritable = odwritable
+ vlclose = vlclose
+ vlcrdnumptr = vlcrdnumptr
+ vlcurfirst = vlcurfirst
+ vlcurjump = vlcurjump
+ vlcurkey = vlcurkey
+ vlcurkeycache = vlcurkeycache
+ vlcurlast = vlcurlast
+ vlcurnext = vlcurnext
+ vlcurout = vlcurout
+ vlcurprev = vlcurprev
+ vlcurput = vlcurput
+ vlcurval = vlcurval
+ vlcurvalcache = vlcurvalcache
+ vlexportdb = vlexportdb
+ vlfatalerror = vlfatalerror
+ vlfsiz = vlfsiz
+ vlget = vlget
+ vlgetcache = vlgetcache
+ vlgetcat = vlgetcat
+ vlgetflags = vlgetflags
+ vlgetlist = vlgetlist
+ vlimportdb = vlimportdb
+ vlinode = vlinode
+ vllnum = vllnum
+ vlmemflush = vlmemflush
+ vlmemsync = vlmemsync
+ vlmtime = vlmtime
+ vlmulcurclose = vlmulcurclose
+ vlmulcurfirst = vlmulcurfirst
+ vlmulcurjump = vlmulcurjump
+ vlmulcurkey = vlmulcurkey
+ vlmulcurkeycache = vlmulcurkeycache
+ vlmulcurlast = vlmulcurlast
+ vlmulcurnext = vlmulcurnext
+ vlmulcuropen = vlmulcuropen
+ vlmulcurprev = vlmulcurprev
+ vlmulcurval = vlmulcurval
+ vlmulcurvalcache = vlmulcurvalcache
+ vlname = vlname
+ vlnnum = vlnnum
+ vlopen = vlopen
+ vloptimize = vloptimize
+ vlout = vlout
+ vloutlist = vloutlist
+ vlput = vlput
+ vlputlist = vlputlist
+ vlremove = vlremove
+ vlrepair = vlrepair
+ vlrnum = vlrnum
+ vlsetfbpsiz = vlsetfbpsiz
+ vlsetflags = vlsetflags
+ vlsettuning = vlsettuning
+ vlsync = vlsync
+ vltranabort = vltranabort
+ vltranbegin = vltranbegin
+ vltrancommit = vltrancommit
+ vlvnum = vlvnum
+ vlvsiz = vlvsiz
+ vlwritable = vlwritable
+ vstclose = vstclose
+ vstcrdnumptr = vstcrdnumptr
+ vstcurfirst = vstcurfirst
+ vstcurjump = vstcurjump
+ vstcurkey = vstcurkey
+ vstcurkeycache = vstcurkeycache
+ vstcurlast = vstcurlast
+ vstcurnext = vstcurnext
+ vstcurout = vstcurout
+ vstcurprev = vstcurprev
+ vstcurput = vstcurput
+ vstcurval = vstcurval
+ vstcurvalcache = vstcurvalcache
+ vstexportdb = vstexportdb
+ vstfatalerror = vstfatalerror
+ vstfsiz = vstfsiz
+ vstget = vstget
+ vstgetcache = vstgetcache
+ vstgetcat = vstgetcat
+ vstgetflags = vstgetflags
+ vstgetlist = vstgetlist
+ vstimportdb = vstimportdb
+ vstinode = vstinode
+ vstlnum = vstlnum
+ vstmemflush = vstmemflush
+ vstmemsync = vstmemsync
+ vstmtime = vstmtime
+ vstmulcurclose = vstmulcurclose
+ vstmulcurfirst = vstmulcurfirst
+ vstmulcurjump = vstmulcurjump
+ vstmulcurkey = vstmulcurkey
+ vstmulcurkeycache = vstmulcurkeycache
+ vstmulcurlast = vstmulcurlast
+ vstmulcurnext = vstmulcurnext
+ vstmulcuropen = vstmulcuropen
+ vstmulcurprev = vstmulcurprev
+ vstmulcurval = vstmulcurval
+ vstmulcurvalcache = vstmulcurvalcache
+ vstname = vstname
+ vstnnum = vstnnum
+ vstopen = vstopen
+ vstoptimize = vstoptimize
+ vstout = vstout
+ vstoutlist = vstoutlist
+ vstput = vstput
+ vstputlist = vstputlist
+ vstremove = vstremove
+ vstrepair = vstrepair
+ vstrnum = vstrnum
+ vstsetfbpsiz = vstsetfbpsiz
+ vstsetflags = vstsetflags
+ vstsettuning = vstsettuning
+ vstsync = vstsync
+ vsttranabort = vsttranabort
+ vsttranbegin = vsttranbegin
+ vsttrancommit = vsttrancommit
+ vstvnum = vstvnum
+ vstvsiz = vstvsiz
+ vstwritable = vstwritable
Added: box/trunk/qdbm/qdbm.pc.in
===================================================================
--- box/trunk/qdbm/qdbm.pc.in (rev 0)
+++ box/trunk/qdbm/qdbm.pc.in 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,14 @@
+prefix=@prefix@
+exec_prefix=@exec_prefix@
+bindir=@bindir@
+libdir=@libdir@
+libexecdir=@libexecdir@
+includedir=@includedir@
+datadir=@datadir@
+
+Name: QDBM
+Description: a high performance embedded database library
+Version: @PACKAGE_VERSION@
+Requires:
+Libs: -L${libdir} -lqdbm @LIBS@
+Cflags: -I${includedir}
Added: box/trunk/qdbm/qdbm.spec.in
===================================================================
--- box/trunk/qdbm/qdbm.spec.in (rev 0)
+++ box/trunk/qdbm/qdbm.spec.in 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,218 @@
+#================================================================
+# RPM Specification for QDBM
+#================================================================
+
+
+
+%define name @PACKAGE_NAME@
+%define version @PACKAGE_VERSION@
+%define release 1
+%define libver @LIBVER@
+%define librev @LIBREV@
+%define disturl http://qdbm.sourceforge.net/
+%define homeurl http://qdbm.sourceforge.net/
+
+Summary: Quick Database Manager
+Name: %{name}
+Version: %{version}
+Release: %{release}
+Source: %{disturl}%{name}-%{version}.tar.gz
+Copyright: LGPL
+Group: Development/Libraries
+Packager: Mikio Hirabayashi <mikio at users.sourceforge.net>
+Distribution: Private
+Vendor: Private
+Url: %{homeurl}
+Requires: zlib
+BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root
+
+%description
+QDBM is an embeded database library compatible with GDBM and NDBM.
+It features hash database and B+ tree database and is developed referring
+to GDBM for the purpose of the following three points: higher processing
+speed, smaller size of a database file, and simpler API.
+This package includes APIs for C, C++, and Java. CGI scripts are also
+contained. APIs for Perl and Ruby should be installed with a source package.
+
+%package devel
+Summary: Headers, libraries, utilities and documentation for QDBM
+Group: Development/Libraries
+Requires: %{name} = %{version}
+
+%description devel
+This package contains header files and libraries needed to develop programs
+using the QDBM library. Some utility commands are also provided.
+
+%package plus
+Summary: C++ libraries for QDBM
+Group: Development/Libraries
+Requires: %{name} = %{version}
+
+%description plus
+This package contains libraries needed to develop and run programs using the
+QDBM C++ bindings.
+
+%package java
+Summary: Java libraries for QDBM
+Group: Development/Libraries
+Requires: %{name} = %{version}
+
+%description java
+This package contains libraries needed to develop and run programs using the
+QDBM Java bindings.
+
+%package cgi
+Summary: CGI scripts with QDBM
+Group: Development/Libraries
+Requires: %{name} = %{version}
+
+%description cgi
+This package contains CGI scripts with QDBM, for administration of databases,
+file uploading, and full-text search.
+
+%prep
+rm -rf $RPM_BUILD_ROOT
+
+%setup -q
+
+%build
+( cd . ; %{configure} --prefix=%{_usr} --mandir=%{_mandir} \
+ --enable-stable --enable-zlib --enable-iconv --enable-pthread ; make )
+( cd plus ; %{configure} --prefix=%{_usr} --mandir=%{_mandir} ; make )
+( cd java ; %{configure} --prefix=%{_usr} --mandir=%{_mandir} ; make )
+( cd cgi ; %{configure} --prefix=%{_usr} --mandir=%{_mandir} ; make )
+
+%install
+rm -rf $RPM_BUILD_ROOT
+( cd . ; make DESTDIR=$RPM_BUILD_ROOT install-strip )
+( cd plus ; make DESTDIR=$RPM_BUILD_ROOT install-strip )
+( cd java ; make DESTDIR=$RPM_BUILD_ROOT install )
+( cd cgi ; make DESTDIR=$RPM_BUILD_ROOT install-strip )
+( mkdir -p $RPM_BUILD_ROOT%{_datadir}/doc/qdbm && \
+ cd $RPM_BUILD_ROOT%{_usr}/share/qdbm && \
+ cp -Rf spex-ja.html spex.html COPYING ChangeLog NEWS THANKS \
+ plus/xspex.html plus/xspex-ja.html plus/xapidoc \
+ java/jspex.html java/jspex-ja.html java/japidoc \
+ cgi/cgispex-ja.html cgi/cgispex.html $RPM_BUILD_ROOT%{_datadir}/doc/qdbm )
+
+%clean
+rm -rf $RPM_BUILD_ROOT
+
+%files
+%defattr(-,root,root)
+%{_libdir}/libqdbm.so.%{libver}.%{librev}.0
+%{_libdir}/libqdbm.so.%{libver}
+%{_libdir}/libqdbm.so
+
+%files devel
+%defattr(-,root,root)
+%{_includedir}/depot.h
+%{_includedir}/curia.h
+%{_includedir}/relic.h
+%{_includedir}/hovel.h
+%{_includedir}/cabin.h
+%{_includedir}/villa.h
+%{_includedir}/vista.h
+%{_includedir}/odeum.h
+%{_libdir}/libqdbm.a
+%{_bindir}/dpmgr
+%{_bindir}/dptest
+%{_bindir}/dptsv
+%{_bindir}/crmgr
+%{_bindir}/crtest
+%{_bindir}/crtsv
+%{_bindir}/rlmgr
+%{_bindir}/rltest
+%{_bindir}/hvmgr
+%{_bindir}/hvtest
+%{_bindir}/cbtest
+%{_bindir}/cbcodec
+%{_bindir}/vlmgr
+%{_bindir}/vltest
+%{_bindir}/vltsv
+%{_bindir}/odmgr
+%{_bindir}/odtest
+%{_bindir}/odidx
+%{_bindir}/qmttest
+%{_mandir}/man1/dpmgr.1.gz
+%{_mandir}/man1/dptest.1.gz
+%{_mandir}/man1/dptsv.1.gz
+%{_mandir}/man1/crmgr.1.gz
+%{_mandir}/man1/crtest.1.gz
+%{_mandir}/man1/crtsv.1.gz
+%{_mandir}/man1/rlmgr.1.gz
+%{_mandir}/man1/rltest.1.gz
+%{_mandir}/man1/hvmgr.1.gz
+%{_mandir}/man1/hvtest.1.gz
+%{_mandir}/man1/cbtest.1.gz
+%{_mandir}/man1/cbcodec.1.gz
+%{_mandir}/man1/vlmgr.1.gz
+%{_mandir}/man1/vltest.1.gz
+%{_mandir}/man1/vltsv.1.gz
+%{_mandir}/man1/odmgr.1.gz
+%{_mandir}/man1/odtest.1.gz
+%{_mandir}/man1/odidx.1.gz
+%{_mandir}/man1/qmttest.1.gz
+%{_mandir}/man3/qdbm.3.gz
+%{_mandir}/man3/depot.3.gz
+%{_mandir}/man3/dpopen.3.gz
+%{_mandir}/man3/curia.3.gz
+%{_mandir}/man3/cropen.3.gz
+%{_mandir}/man3/relic.3.gz
+%{_mandir}/man3/hovel.3.gz
+%{_mandir}/man3/cabin.3.gz
+%{_mandir}/man3/villa.3.gz
+%{_mandir}/man3/vlopen.3.gz
+%{_mandir}/man3/vista.3.gz
+%{_mandir}/man3/odeum.3.gz
+%{_mandir}/man3/odopen.3.gz
+%{_datadir}/doc/qdbm/spex.html
+%{_datadir}/doc/qdbm/spex-ja.html
+%{_datadir}/doc/qdbm/COPYING
+%{_datadir}/doc/qdbm/ChangeLog
+%{_datadir}/doc/qdbm/NEWS
+%{_datadir}/doc/qdbm/THANKS
+%{_libdir}/pkgconfig/qdbm.pc
+
+%files plus
+%defattr(-,root,root)
+%{_includedir}/xqdbm.h
+%{_includedir}/xadbm.h
+%{_includedir}/xdepot.h
+%{_includedir}/xcuria.h
+%{_includedir}/xvilla.h
+%{_libdir}/libxqdbm.a
+%{_libdir}/libxqdbm.so.3.0.0
+%{_libdir}/libxqdbm.so.3
+%{_libdir}/libxqdbm.so
+%{_bindir}/xdptest
+%{_bindir}/xcrtest
+%{_bindir}/xvltest
+%{_datadir}/doc/qdbm/xspex.html
+%{_datadir}/doc/qdbm/xspex-ja.html
+%{_datadir}/doc/qdbm/xapidoc/
+
+%files java
+%defattr(-,root,root)
+%{_libdir}/qdbm.jar
+%{_libdir}/libjqdbm.so.1.0.0
+%{_libdir}/libjqdbm.so.1
+%{_libdir}/libjqdbm.so
+%{_datadir}/doc/qdbm/jspex.html
+%{_datadir}/doc/qdbm/jspex-ja.html
+%{_datadir}/doc/qdbm/japidoc/
+
+%files cgi
+%defattr(-,root,root)
+%{_libexecdir}/qadm.cgi
+%{_libexecdir}/qupl.cgi
+%{_libexecdir}/qfts.cgi
+%{_datadir}/qdbm/cgi/qadm.conf
+%{_datadir}/qdbm/cgi/qupl.conf
+%{_datadir}/qdbm/cgi/qfts.conf
+%{_datadir}/doc/qdbm/cgispex.html
+%{_datadir}/doc/qdbm/cgispex-ja.html
+
+
+
+# END OF FILE
Added: box/trunk/qdbm/qmttest.c
===================================================================
--- box/trunk/qdbm/qmttest.c (rev 0)
+++ box/trunk/qdbm/qmttest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,300 @@
+/*************************************************************************************************
+ * Test cases for thread-safety
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <curia.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+#include <time.h>
+
+#if defined(MYPTHREAD)
+#include <sys/types.h>
+#include <pthread.h>
+#endif
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define PATHBUFSIZ 1024 /* buffer for paths */
+#define RECBUFSIZ 32 /* buffer for records */
+
+typedef struct { /* type of structure of thread arguments */
+ int id; /* ID of the thread */
+ const char *name; /* prefix of the database */
+ int rnum; /* number of records */
+ int alive; /* alive or not */
+} MYARGS;
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+void pdperror(const char *name);
+void *procthread(void *args);
+int dotest(const char *name, int rnum, int tnum);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env, *name;
+ int rv, rnum, tnum;
+ cbstdiobin();
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ progname = argv[0];
+ srand(time(NULL));
+ if(argc < 4) usage();
+ name = argv[1];
+ if((rnum = atoi(argv[2])) < 1) usage();
+ if((tnum = atoi(argv[3])) < 1) usage();
+ rv = dotest(name, rnum, tnum);
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for thread-safety\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s name rnum tnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+#if defined(MYPTHREAD)
+ static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
+ va_list ap;
+ int rv;
+ if(pthread_mutex_lock(&mymutex) != 0) return -1;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ pthread_mutex_unlock(&mymutex);
+ return rv;
+#else
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+#endif
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* pseudo random number generator */
+int myrand(void){
+ static int cnt = 0;
+ return (rand() * rand() + (rand() >> (sizeof(int) * 4)) + (cnt++)) & INT_MAX;
+}
+
+
+/* process the test */
+void *procthread(void *args){
+ MYARGS *myargs;
+ DEPOT *depot;
+ CURIA *curia;
+ VILLA *villa;
+ CBLIST *list;
+ CBMAP *map;
+ char name[PATHBUFSIZ], buf[RECBUFSIZ];
+ int i, err, len;
+ myargs = (MYARGS *)args;
+ err = FALSE;
+ sprintf(name, "%s-%04d", myargs->name, myargs->id);
+ dpremove(name);
+ crremove(name);
+ vlremove(name);
+ switch(myrand() % 4){
+ case 0:
+ printfflush("\n[Depot Test] name=%s rnum=%d\n", name, myargs->rnum);
+ if(!(depot = dpopen(name, DP_OWRITER | DP_OCREAT | DP_OTRUNC, -1))){
+ pdperror(name);
+ return "error";
+ }
+ for(i = 1; i <= myargs->rnum; i++){
+ len = sprintf(buf, "%d", myrand() % i + 1);
+ if(!dpput(depot, buf, len, buf, len, i % 2 == 0 ? DP_DOVER : DP_DCAT)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(myargs->rnum > 250 && i % (myargs->rnum / 250) == 0){
+ printfflush(".");
+ if(i == myargs->rnum || i % (myargs->rnum / 10) == 0){
+ printfflush("\n%s: (%d)\n", name, i);
+ }
+ }
+ }
+ if(!dpclose(depot)){
+ pdperror(name);
+ err = TRUE;
+ }
+ printfflush("\n%s: finished\n", name);
+ break;
+ case 1:
+ printfflush("\n[Curia Test] name=%s rnum=%d\n", name, myargs->rnum);
+ if(!(curia = cropen(name, CR_OWRITER | CR_OCREAT | CR_OTRUNC, -1, -1))){
+ pdperror(name);
+ return "error";
+ }
+ for(i = 1; i <= myargs->rnum; i++){
+ len = sprintf(buf, "%d", myrand() % i + 1);
+ if(!crput(curia, buf, len, buf, len, i % 2 == 0 ? CR_DOVER : CR_DCAT)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(myargs->rnum > 250 && i % (myargs->rnum / 250) == 0){
+ printfflush(".");
+ if(i == myargs->rnum || i % (myargs->rnum / 10) == 0){
+ printfflush("\n%s: (%d)\n", name, i);
+ }
+ }
+ }
+ if(!crclose(curia)){
+ pdperror(name);
+ err = TRUE;
+ }
+ printfflush("\n%s: finished\n", name);
+ break;
+ case 2:
+ printfflush("\n[Villa Test] name=%s rnum=%d\n", name, myargs->rnum);
+ if(!(villa = vlopen(name, VL_OWRITER | VL_OCREAT | VL_OTRUNC, VL_CMPLEX))){
+ pdperror(name);
+ return "error";
+ }
+ for(i = 1; i <= myargs->rnum; i++){
+ len = sprintf(buf, "%d", myrand() % i + 1);
+ if(!vlput(villa, buf, len, buf, len, i % 2 == 0 ? VL_DOVER : VL_DDUP)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(myargs->rnum > 250 && i % (myargs->rnum / 250) == 0){
+ printfflush(".");
+ if(i == myargs->rnum || i % (myargs->rnum / 10) == 0){
+ printfflush("\n%s: (%d)\n", name, i);
+ }
+ }
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ err = TRUE;
+ }
+ printfflush("\n%s: finished\n", name);
+ break;
+ case 3:
+ printfflush("\n[Cabin Test] name=%s rnum=%d\n", name, myargs->rnum);
+ list = cblistopen();
+ map = cbmapopen();
+ for(i = 1; i <= myargs->rnum; i++){
+ len = sprintf(buf, "%d", myrand() % i + 1);
+ cblistpush(list, buf, len);
+ cbmapput(map, buf, len, buf, len, i % 2 == 0 ? TRUE : FALSE);
+ if(myargs->rnum > 250 && i % (myargs->rnum / 250) == 0){
+ printfflush(".");
+ if(i == myargs->rnum || i % (myargs->rnum / 10) == 0){
+ printfflush("\n%s: (%d)\n", name, i);
+ }
+ }
+ }
+ cbmapclose(map);
+ cblistclose(list);
+ printfflush("\n%s: finished\n", name);
+ break;
+ }
+ return err ? "error" : NULL;
+}
+
+
+/* drive the test */
+int dotest(const char *name, int rnum, int tnum){
+#if defined(MYPTHREAD)
+ pthread_t *thary;
+ MYARGS *argsary;
+ char *rv;
+ int i, err;
+ printfflush("<Thread-Safety Test>\n name=%s rnum=%d tnum=%d\n", name, rnum, tnum);
+ err = FALSE;
+ thary = cbmalloc(tnum * sizeof(pthread_t));
+ argsary = cbmalloc(tnum * sizeof(MYARGS));
+ for(i = 0; i < tnum; i++){
+ argsary[i].id = i + 1;
+ argsary[i].name = name;
+ argsary[i].rnum = rnum;
+ argsary[i].alive = TRUE;
+ if(pthread_create(thary + i, NULL, procthread, argsary + i) != 0){
+ argsary[i].alive = FALSE;
+ err = TRUE;
+ }
+ }
+ for(i = 0; i < tnum; i++){
+ if(!argsary[i].alive) continue;
+ if(pthread_join(thary[i], (void *)&rv) != 0 || rv) err = TRUE;
+ }
+ free(argsary);
+ free(thary);
+ if(!err) printfflush("\nall ok\n");
+ return err ? 1 : 0;
+#else
+ MYARGS *argsary;
+ int i, err;
+ printfflush("<Thread-Safety Test>\n name=%s rnum=%d tnum=%d\n", name, rnum, tnum);
+ err = FALSE;
+ argsary = cbmalloc(tnum * sizeof(MYARGS));
+ for(i = 0; i < tnum; i++){
+ argsary[i].id = i + 1;
+ argsary[i].name = name;
+ argsary[i].rnum = rnum;
+ argsary[i].alive = TRUE;
+ if(procthread(argsary + i)) err = TRUE;
+ }
+ free(argsary);
+ if(!err) printfflush("\nall ok\n");
+ return err ? 1 : 0;
+#endif
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/qmttest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/relic.c
===================================================================
--- box/trunk/qdbm/relic.c (rev 0)
+++ box/trunk/qdbm/relic.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,266 @@
+/*************************************************************************************************
+ * Implementation of Relic
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "relic.h"
+#include "myconf.h"
+
+#define RL_NAMEMAX 512 /* max size of a database name */
+#define RL_DIRFSUF MYEXTSTR "dir" /* suffix of a directory file */
+#define RL_DATAFSUF MYEXTSTR "pag" /* suffix of a page file */
+#define RL_PATHBUFSIZ 1024 /* size of a path buffer */
+#define RL_INITBNUM 1913 /* initial bucket number */
+#define RL_ALIGNSIZ 16 /* size of alignment */
+#define RL_MAXLOAD 1.25 /* max ratio of bucket loading */
+#define RL_DIRMAGIC "[depot]\0\v" /* magic number of a directory file */
+
+
+/* private function prototypes */
+static void dbm_writedummy(int fd);
+static int dbm_writestr(int fd, const char *str);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* Get a database handle. */
+DBM *dbm_open(char *name, int flags, int mode){
+ DBM *db;
+ DEPOT *depot;
+ int dpomode;
+ char path[RL_PATHBUFSIZ];
+ int dfd, fd;
+ assert(name);
+ if(strlen(name) > RL_NAMEMAX) return NULL;
+ dpomode = DP_OREADER;
+ if((flags & O_WRONLY) || (flags & O_RDWR)){
+ dpomode = DP_OWRITER;
+ if(flags & O_CREAT) dpomode |= DP_OCREAT;
+ if(flags & O_TRUNC) dpomode |= DP_OTRUNC;
+ }
+ mode |= 00600;
+ sprintf(path, "%s%s", name, RL_DIRFSUF);
+ if((dfd = open(path, flags, mode)) == -1) return NULL;
+ dbm_writedummy(dfd);
+ sprintf(path, "%s%s", name, RL_DATAFSUF);
+ if((fd = open(path, flags, mode)) == -1 || close(fd) == -1){
+ close(dfd);
+ return NULL;
+ }
+ if(!(depot = dpopen(path, dpomode, RL_INITBNUM))){
+ close(dfd);
+ return NULL;
+ }
+ if(dpomode & DP_OWRITER){
+ if(!dpsetalign(depot, RL_ALIGNSIZ)){
+ close(dfd);
+ dpclose(depot);
+ return NULL;
+ }
+ }
+ if(!(db = malloc(sizeof(DBM)))){
+ close(dfd);
+ dpclose(depot);
+ return NULL;
+ }
+ db->depot = depot;
+ db->dfd = dfd;
+ db->dbm_fetch_vbuf = NULL;
+ db->dbm_nextkey_kbuf = NULL;
+ return db;
+}
+
+
+/* Close a database handle. */
+void dbm_close(DBM *db){
+ assert(db);
+ free(db->dbm_fetch_vbuf);
+ free(db->dbm_nextkey_kbuf);
+ close(db->dfd);
+ dpclose(db->depot);
+ free(db);
+}
+
+
+/* Store a record. */
+int dbm_store(DBM *db, datum key, datum content, int flags){
+ int dmode;
+ int bnum, rnum;
+ assert(db);
+ if(!key.dptr || key.dsize < 0 || !content.dptr || content.dsize < 0) return -1;
+ dmode = (flags == DBM_INSERT) ? DP_DKEEP : DP_DOVER;
+ if(!dpput(db->depot, key.dptr, key.dsize, content.dptr, content.dsize, dmode)){
+ if(dpecode == DP_EKEEP) return 1;
+ return -1;
+ }
+ bnum = dpbnum(db->depot);
+ rnum = dprnum(db->depot);
+ if(bnum > 0 && rnum > 0 && ((double)rnum / (double)bnum > RL_MAXLOAD)){
+ if(!dpoptimize(db->depot, -1)) return -1;
+ }
+ return 0;
+}
+
+
+/* Delete a record. */
+int dbm_delete(DBM *db, datum key){
+ assert(db);
+ if(!key.dptr || key.dsize < 0) return -1;
+ if(!dpout(db->depot, key.dptr, key.dsize)) return -1;
+ return 0;
+}
+
+
+/* Retrieve a record. */
+datum dbm_fetch(DBM *db, datum key){
+ datum content;
+ char *vbuf;
+ int vsiz;
+ assert(db);
+ if(!key.dptr || key.dsize < 0 ||
+ !(vbuf = dpget(db->depot, key.dptr, key.dsize, 0, -1, &vsiz))){
+ content.dptr = NULL;
+ content.dsize = 0;
+ return content;
+ }
+ free(db->dbm_fetch_vbuf);
+ db->dbm_fetch_vbuf = vbuf;
+ content.dptr = vbuf;
+ content.dsize = vsiz;
+ return content;
+}
+
+
+/* Get the first key of a database. */
+datum dbm_firstkey(DBM *db){
+ assert(db);
+ dpiterinit(db->depot);
+ return dbm_nextkey(db);
+}
+
+
+/* Get the next key of a database. */
+datum dbm_nextkey(DBM *db){
+ datum key;
+ char *kbuf;
+ int ksiz;
+ if(!(kbuf = dpiternext(db->depot, &ksiz))){
+ key.dptr = NULL;
+ key.dsize = 0;
+ return key;
+ }
+ free(db->dbm_nextkey_kbuf);
+ db->dbm_nextkey_kbuf = kbuf;
+ key.dptr = kbuf;
+ key.dsize = ksiz;
+ return key;
+}
+
+
+/* Check whether a database has a fatal error or not. */
+int dbm_error(DBM *db){
+ assert(db);
+ return dpfatalerror(db->depot) ? TRUE : FALSE;
+}
+
+
+/* No effect. */
+int dbm_clearerr(DBM *db){
+ assert(db);
+ return 0;
+}
+
+
+/* Check whether a handle is read-only or not. */
+int dbm_rdonly(DBM *db){
+ assert(db);
+ return dpwritable(db->depot) ? FALSE : TRUE;
+}
+
+
+/* Get the file descriptor of a directory file. */
+int dbm_dirfno(DBM *db){
+ assert(db);
+ return db->dfd;
+}
+
+
+/* Get the file descriptor of a data file. */
+int dbm_pagfno(DBM *db){
+ assert(db);
+ return dpfdesc(db->depot);
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Write dummy data into a dummy file.
+ `fd' specifies a file descriptor. */
+static void dbm_writedummy(int fd){
+ struct stat sbuf;
+ if(fstat(fd, &sbuf) == -1 || sbuf.st_size > 0) return;
+ write(fd, RL_DIRMAGIC, sizeof(RL_DIRMAGIC) - 1);
+ dbm_writestr(fd, "\n\n");
+ dbm_writestr(fd, "\x20\x20\xa2\xca\xa1\xb2\xa2\xca\x20\x20\x20\x20\x20\xa1\xbf\xa1");
+ dbm_writestr(fd, "\xb1\xa1\xb1\xa1\xb1\xa1\xb1\xa1\xb1\xa1\xb1\xa1\xb1\xa1\xb1\xa1");
+ dbm_writestr(fd, "\xb1\x0a\xa1\xca\x20\xa1\xad\xa2\xcf\xa1\xae\xa1\xcb\xa1\xe3\x20");
+ dbm_writestr(fd, "\x20\x4e\x44\x42\x4d\x20\x43\x6f\x6d\x70\x61\x74\x69\x62\x69\x6c");
+ dbm_writestr(fd, "\x69\x74\x79\x0a\xa1\xca\x20\x20\x20\x20\x20\x20\x20\xa1\xcb\x20");
+ dbm_writestr(fd, "\x20\xa1\xc0\xa1\xb2\xa1\xb2\xa1\xb2\xa1\xb2\xa1\xb2\xa1\xb2\xa1");
+ dbm_writestr(fd, "\xb2\xa1\xb2\xa1\xb2\x0a\x20\xa1\xc3\x20\x20\xa1\xc3\x20\xa1\xc3");
+ dbm_writestr(fd, "\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20");
+ dbm_writestr(fd, "\x20\x20\x20\x20\x20\x20\x20\x0a\xa1\xca\x5f\x5f\xa1\xb2\xa1\xcb");
+ dbm_writestr(fd, "\x5f\xa1\xcb\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20\x20");
+ dbm_writestr(fd, "\x20\x20\x20\x20\x20\x20\x20\x20\x20\x0a");
+}
+
+
+/* Write a string into a file.
+ `fd' specifies a file descriptor.
+ `str' specifies a string. */
+static int dbm_writestr(int fd, const char *str){
+ const char *lbuf;
+ int size, rv, wb;
+ assert(fd >= 0 && str);
+ lbuf = str;
+ size = strlen(str);
+ rv = 0;
+ do {
+ wb = write(fd, lbuf, size);
+ switch(wb){
+ case -1: if(errno != EINTR) return -1;
+ case 0: break;
+ default:
+ lbuf += wb;
+ size -= wb;
+ rv += wb;
+ break;
+ }
+ } while(size > 0);
+ return rv;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/relic.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/relic.h
===================================================================
--- box/trunk/qdbm/relic.h (rev 0)
+++ box/trunk/qdbm/relic.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,170 @@
+/*************************************************************************************************
+ * The NDBM-compatible API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _RELIC_H /* duplication check */
+#define _RELIC_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <depot.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+typedef struct { /* type of structure for a database handle */
+ DEPOT *depot; /* internal database handle */
+ int dfd; /* file descriptor of a dummy file */
+ char *dbm_fetch_vbuf; /* buffer for dbm_fetch */
+ char *dbm_nextkey_kbuf; /* buffer for dbm_nextkey */
+} DBM;
+
+typedef struct { /* type of structure for a key or a value */
+ void *dptr; /* pointer to the region */
+ size_t dsize; /* size of the region */
+} datum;
+
+enum { /* enumeration for write modes */
+ DBM_INSERT, /* keep an existing value */
+ DBM_REPLACE /* overwrite an existing value */
+};
+
+
+/* Get a database handle.
+ `name' specifies the name of a database. The file names are concatenated with suffixes.
+ `flags' is the same as the one of `open' call, although `O_WRONLY' is treated as `O_RDWR'
+ and additional flags except for `O_CREAT' and `O_TRUNC' have no effect.
+ `mode' specifies the mode of the database file as the one of `open' call does.
+ The return value is the database handle or `NULL' if it is not successful. */
+DBM *dbm_open(char *name, int flags, int mode);
+
+
+/* Close a database handle.
+ `db' specifies a database handle.
+ Because the region of the closed handle is released, it becomes impossible to use the
+ handle. */
+void dbm_close(DBM *db);
+
+
+/* Store a record.
+ `db' specifies a database handle.
+ `key' specifies a structure of a key.
+ `content' specifies a structure of a value.
+ `flags' specifies behavior when the key overlaps, by the following values: `DBM_REPLACE',
+ which means the specified value overwrites the existing one, `DBM_INSERT', which means the
+ existing value is kept.
+ The return value is 0 if it is successful, 1 if it gives up because of overlaps of the key,
+ -1 if other error occurs. */
+int dbm_store(DBM *db, datum key, datum content, int flags);
+
+
+/* Delete a record.
+ `db' specifies a database handle.
+ `key' specifies a structure of a key.
+ The return value is 0 if it is successful, -1 if some errors occur. */
+int dbm_delete(DBM *db, datum key);
+
+
+/* Retrieve a record.
+ `db' specifies a database handle.
+ `key' specifies a structure of a key.
+ The return value is a structure of the result.
+ If a record corresponds, the member `dptr' of the structure is the pointer to the region of
+ the value. If no record corresponds or some errors occur, `dptr' is `NULL'. `dptr' points
+ to the region related with the handle. The region is available until the next time of
+ calling this function with the same handle. */
+datum dbm_fetch(DBM *db, datum key);
+
+
+/* Get the first key of a database.
+ `db' specifies a database handle.
+ The return value is a structure of the result.
+ If a record corresponds, the member `dptr' of the structure is the pointer to the region of
+ the first key. If no record corresponds or some errors occur, `dptr' is `NULL'. `dptr'
+ points to the region related with the handle. The region is available until the next time
+ of calling this function or the function `dbm_nextkey' with the same handle. */
+datum dbm_firstkey(DBM *db);
+
+
+/* Get the next key of a database.
+ `db' specifies a database handle.
+ The return value is a structure of the result.
+ If a record corresponds, the member `dptr' of the structure is the pointer to the region of
+ the next key. If no record corresponds or some errors occur, `dptr' is `NULL'. `dptr'
+ points to the region related with the handle. The region is available until the next time
+ of calling this function or the function `dbm_firstkey' with the same handle. */
+datum dbm_nextkey(DBM *db);
+
+
+/* Check whether a database has a fatal error or not.
+ `db' specifies a database handle.
+ The return value is true if the database has a fatal error, false if not. */
+int dbm_error(DBM *db);
+
+
+/* No effect.
+ `db' specifies a database handle.
+ The return value is 0.
+ The function is only for compatibility. */
+int dbm_clearerr(DBM *db);
+
+
+/* Check whether a handle is read-only or not.
+ `db' specifies a database handle.
+ The return value is true if the handle is read-only, or false if not read-only. */
+int dbm_rdonly(DBM *db);
+
+
+/* Get the file descriptor of a directory file.
+ `db' specifies a database handle.
+ The return value is the file descriptor of the directory file. */
+int dbm_dirfno(DBM *db);
+
+
+/* Get the file descriptor of a data file.
+ `db' specifies a database handle.
+ The return value is the file descriptor of the data file. */
+int dbm_pagfno(DBM *db);
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/relic.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/rlmgr.c
===================================================================
--- box/trunk/qdbm/rlmgr.c (rev 0)
+++ box/trunk/qdbm/rlmgr.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,465 @@
+/*************************************************************************************************
+ * Utility for debugging Relic and its applications
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <relic.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *hextoobj(const char *str, int *sp);
+int runcreate(int argc, char **argv);
+int runstore(int argc, char **argv);
+int rundelete(int argc, char **argv);
+int runfetch(int argc, char **argv);
+int runlist(int argc, char **argv);
+void pmyerror(const char *name, const char *msg);
+void printobj(const char *obj, int size);
+void printobjhex(const char *obj, int size);
+int docreate(char *name);
+int dostore(char *name, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int ins);
+int dodelete(char *name, const char *kbuf, int ksiz);
+int dofetch(char *name, const char *kbuf, int ksiz, int ox, int nb);
+int dolist(char *name, int ox);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "create")){
+ rv = runcreate(argc, argv);
+ } else if(!strcmp(argv[1], "store")){
+ rv = runstore(argc, argv);
+ } else if(!strcmp(argv[1], "delete")){
+ rv = rundelete(argc, argv);
+ } else if(!strcmp(argv[1], "fetch")){
+ rv = runfetch(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: administration utility for Relic\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s create name\n", progname);
+ fprintf(stderr, " %s store [-kx] [-vx|-vf] [-insert] name key val\n", progname);
+ fprintf(stderr, " %s delete [-kx] name key\n", progname);
+ fprintf(stderr, " %s fetch [-kx] [-ox] [-n] name key\n", progname);
+ fprintf(stderr, " %s list [-ox] name\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* create a binary object from a hexadecimal string */
+char *hextoobj(const char *str, int *sp){
+ char *buf, mbuf[3];
+ int len, i, j;
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ j = 0;
+ for(i = 0; i < len; i += 2){
+ while(strchr(" \n\r\t\f\v", str[i])){
+ i++;
+ }
+ if((mbuf[0] = str[i]) == '\0') break;
+ if((mbuf[1] = str[i+1]) == '\0') break;
+ mbuf[2] = '\0';
+ buf[j++] = (char)strtol(mbuf, NULL, 16);
+ }
+ buf[j] = '\0';
+ *sp = j;
+ return buf;
+}
+
+
+/* parse arguments of create command */
+int runcreate(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docreate(name);
+ return rv;
+}
+
+
+/* parse arguments of store command */
+int runstore(int argc, char **argv){
+ char *name, *key, *val, *kbuf, *vbuf;
+ int i, kx, vx, vf, ins, ksiz, vsiz, rv;
+ name = NULL;
+ kx = FALSE;
+ vx = FALSE;
+ vf = FALSE;
+ ins = FALSE;
+ key = NULL;
+ val = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-vx")){
+ vx = TRUE;
+ } else if(!strcmp(argv[i], "-vf")){
+ vf = TRUE;
+ } else if(!strcmp(argv[i], "-insert")){
+ ins = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else if(!val){
+ val = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || !val) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = strlen(kbuf);
+ }
+ if(vx){
+ vbuf = hextoobj(val, &vsiz);
+ } else if(vf){
+ vbuf = cbreadfile(val, &vsiz);
+ } else {
+ vbuf = cbmemdup(val, -1);
+ vsiz = strlen(vbuf);
+ }
+ if(kbuf && vbuf){
+ rv = dostore(name, kbuf, ksiz, vbuf, vsiz, ins);
+ } else {
+ if(vf){
+ fprintf(stderr, "%s: %s: cannot read\n", progname, val);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ }
+ rv = 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ return rv;
+}
+
+
+/* parse arguments of delete command */
+int rundelete(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = strlen(kbuf);
+ }
+ if(kbuf){
+ rv = dodelete(name, kbuf, ksiz);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of fetch command */
+int runfetch(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ox, nb, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ ox = FALSE;
+ nb = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = strlen(kbuf);
+ }
+ if(kbuf){
+ rv = dofetch(name, kbuf, ksiz, ox, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ char *name;
+ int i, ox, rv;
+ name = NULL;
+ ox = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dolist(name, ox);
+ return rv;
+}
+
+
+/* print an error message */
+void pmyerror(const char *name, const char *msg){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, msg);
+}
+
+
+/* print an object */
+void printobj(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ putchar(obj[i]);
+ }
+}
+
+
+/* print an object as a hexadecimal string */
+void printobjhex(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ printf("%s%02X", i > 0 ? " " : "", ((const unsigned char *)obj)[i]);
+ }
+}
+
+
+/* perform create command */
+int docreate(char *name){
+ DBM *db;
+ if(!(db = dbm_open(name, O_RDWR | O_CREAT | O_TRUNC, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ dbm_close(db);
+ return 0;
+}
+
+
+/* perform store command */
+int dostore(char *name, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int ins){
+ DBM *db;
+ datum key, content;
+ int rv;
+ if(!(db = dbm_open(name, O_RDWR, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ key.dptr = (char *)kbuf;
+ key.dsize = ksiz;
+ content.dptr = (char *)vbuf;
+ content.dsize = vsiz;
+ switch(dbm_store(db, key, content, ins ? DBM_INSERT : DBM_REPLACE)){
+ case 0:
+ rv = 0;
+ break;
+ case 1:
+ pmyerror(name, "dbm_store failed by insert");
+ rv = 1;
+ break;
+ default:
+ pmyerror(name, "dbm_store failed");
+ rv = 1;
+ break;
+ }
+ dbm_close(db);
+ return rv;
+}
+
+
+/* perform delete command */
+int dodelete(char *name, const char *kbuf, int ksiz){
+ DBM *db;
+ datum key;
+ int rv;
+ if(!(db = dbm_open((char *)name, O_RDWR, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ key.dptr = (char *)kbuf;
+ key.dsize = ksiz;
+ if(dbm_delete(db, key) == 0){
+ rv = 0;
+ } else {
+ pmyerror(name, "dbm_delete failed");
+ rv = 1;
+ }
+ dbm_close(db);
+ return rv;
+}
+
+
+/* perform fetch command */
+int dofetch(char *name, const char *kbuf, int ksiz, int ox, int nb){
+ DBM *db;
+ datum key, content;
+ int rv;
+ if(!(db = dbm_open((char *)name, O_RDONLY, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ key.dptr = (char *)kbuf;
+ key.dsize = ksiz;
+ content = dbm_fetch(db, key);
+ if(content.dptr){
+ if(ox){
+ printobjhex(content.dptr, content.dsize);
+ } else {
+ printobj(content.dptr, content.dsize);
+ }
+ if(!nb) putchar('\n');
+ rv = 0;
+ } else {
+ pmyerror(name, "dbm_fetch failed");
+ rv = 1;
+ }
+ dbm_close(db);
+ return rv;
+}
+
+
+/* perform list command */
+int dolist(char *name, int ox){
+ DBM *db;
+ datum key, val;
+ if(!(db = dbm_open((char *)name, O_RDONLY, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ for(key = dbm_firstkey(db); key.dptr != NULL; key = dbm_nextkey(db)){
+ val = dbm_fetch(db, key);
+ if(!val.dptr) break;
+ if(ox){
+ printobjhex(key.dptr, key.dsize);
+ putchar('\t');
+ printobjhex(val.dptr, val.dsize);
+ } else {
+ printobj(key.dptr, key.dsize);
+ putchar('\t');
+ printobj(val.dptr, val.dsize);
+ }
+ putchar('\n');
+ }
+ dbm_close(db);
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/rlmgr.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/rltest.c
===================================================================
--- box/trunk/qdbm/rltest.c (rev 0)
+++ box/trunk/qdbm/rltest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,241 @@
+/*************************************************************************************************
+ * Test cases of Relic
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <relic.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define RECBUFSIZ 32 /* buffer for records */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runwrite(int argc, char **argv);
+int runread(int argc, char **argv);
+int printfflush(const char *format, ...);
+void pmyerror(const char *name, const char *msg);
+int dowrite(char *name, int rnum);
+int doread(char *name, int rnum);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "write")){
+ rv = runwrite(argc, argv);
+ } else if(!strcmp(argv[1], "read")){
+ rv = runread(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Relic\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s write name rnum\n", progname);
+ fprintf(stderr, " %s read name rnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of write command */
+int runwrite(int argc, char **argv){
+ char *name, *rstr;
+ int i, rnum, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowrite(name, rnum);
+ return rv;
+}
+
+
+/* parse arguments of read command */
+int runread(int argc, char **argv){
+ char *name, *rstr;
+ int i, rnum, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = doread(name, rnum);
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* print an error message */
+void pmyerror(const char *name, const char *msg){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, msg);
+}
+
+
+/* perform write command */
+int dowrite(char *name, int rnum){
+ DBM *db;
+ datum key, content;
+ int i, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Writing Test>\n name=%s rnum=%d\n\n", name, rnum);
+ /* open a database */
+ if(!(db = dbm_open(name, O_RDWR | O_CREAT | O_TRUNC, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", i);
+ key.dptr = buf;
+ key.dsize = len;
+ content.dptr = buf;
+ content.dsize = len;
+ /* store a record */
+ if(dbm_store(db, key, content, DBM_REPLACE) < 0){
+ pmyerror(name, "dbm_store failed");
+ err = TRUE;
+ break;
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ dbm_close(db);
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+/* perform read command */
+int doread(char *name, int rnum){
+ DBM *db;
+ datum key, content;
+ int i, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Reading Test>\n name=%s rnum=%d\n\n", name, rnum);
+ /* open a database */
+ if(!(db = dbm_open(name, O_RDONLY, 00644))){
+ pmyerror(name, "dbm_open failed");
+ return 1;
+ }
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* retrieve a record */
+ len = sprintf(buf, "%08d", i);
+ key.dptr = buf;
+ key.dsize = len;
+ content = dbm_fetch(db, key);
+ if(!content.dptr){
+ pmyerror(name, "dbm_fetch failed");
+ err = TRUE;
+ break;
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ dbm_close(db);
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/rltest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/spex-ja.html
===================================================================
--- box/trunk/qdbm/spex-ja.html (rev 0)
+++ box/trunk/qdbm/spex-ja.html 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,4348 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="ja" lang="ja">
+
+<head>
+<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
+<meta http-equiv="Content-Style-Type" content="text/css" />
+<meta name="author" content="Mikio Hirabayashi" />
+<meta name="keywords" content="QDBM, DBM, database, hash, B+ tree" />
+<meta name="description" content="fundamental specifications of QDBM" />
+<link rel="contents" href="./" />
+<link rel="alternate" href="spex.html" hreflang="en" title="the English version" />
+<link rev="made" href="mailto:mikio at users.sourceforge.net" />
+<title>Specifications of QDBM Version 1 (Japanese)</title>
+<style type="text/css">html { margin: 0em 0em; padding: 0em 0em; background: #eeeeee none; }
+body { margin: 2em 2em; padding: 0em 0em;
+ background: #eeeeee none; color: #111111;
+ font-style: normal; font-weight: normal; }
+h1 { margin-top: 1.8em; margin-bottom: 1.3em; font-weight: bold; }
+h2 { margin-top: 1.8em; margin-bottom: 1.1em; font-weight: bold;
+ border-left: solid 0.6em #445555; border-bottom: solid 1pt #bbbbbb;
+ padding: 0.5em 0.5em; width: 60%; }
+h3 { margin-top: 1.8em; margin-bottom: 0.8em; font-weight: bold; }
+hr { margin-top: 2.5em; margin-bottom: 1.5em; height: 1pt;
+ color: #999999; background-color: #999999; border: none; }
+div.note,div.navi { text-align: right; }
+div.logo { text-align: center; margin: 3em 0em; }
+div.logo img { border: inset 2pt #ccccdd; }
+p { margin: 0.8em 0em; line-height: 140%; }
+p,dd { text-indent: 0.8em; }
+div,pre { margin-left: 1.7em; margin-right: 1.7em; }
+pre { background-color: #ddddee; padding: 0.2em; border: 1pt solid #bbbbcc; font-size: smaller; }
+kbd { color: #111111; font-style: normal; font-weight: bold; }
+a { color: #0022aa; text-decoration: none; }
+a:hover,a:focus { color: #0033ee; text-decoration: underline; }
+a.head { color: #111111; text-decoration: none; }
+table { padding: 1pt 2pt 1pt 2pt; border: none; margin-left: 1.7em; border-collapse: collapse; }
+th { padding: 1pt 4pt 1pt 4pt; border-style: none;
+ text-align: left; vertical-align: bottom; }
+td { padding: 1pt 4pt 1pt 4pt; border: 1pt solid #333333;
+ text-align: left; vertical-align: top; }
+ul,ol,dl { line-height: 140%; }
+dt { margin-left: 1.2em; }
+dd { margin-left: 2.0em; }
+ul.lines { list-style-type: none; }
+ at media print {
+ html,body { margin: 0em 0em; background-color: #ffffff; color: #000000; }
+ h1 { padding: 8em 0em 0.5em 0em; text-align: center; }
+ h2 { page-break-before: always; }
+ div.note { text-align: center; }
+ div.navi,div.logo { display: none }
+ hr { display: none; }
+ pre { margin: 0.8em 0.8em; background-color: #ffffff;
+ border: 1pt solid #aaaaaa; font-size: smaller; }
+ a,kbd { color: #000000; text-decoration: none; }
+ h1,h2,h3 { font-family: sans-serif; }
+ p,div,li,dt,dd { font-family: serif; }
+ pre,kbd { font-family: monospace; }
+ dd { font-size: smaller; }
+}
+</style>
+</head>
+
+<body>
+
+<h1>QDBMãƒãƒ¼ã‚¸ãƒ§ãƒ³ï¼‘基本仕様書</h1>
+
+<div class="note">Copyright (C) 2000-2007 Mikio Hirabayashi</div>
+<div class="note">Last Update: Thu, 26 Oct 2006 15:00:20 +0900</div>
+<div class="navi">[<a href="spex.html" hreflang="en">English</a>] [<a href="http://qdbm.sourceforge.net/">Home</a>]</div>
+
+<hr />
+
+<h2>目次</h2>
+
+<ol>
+<li><a href="#overview">概è¦</a></li>
+<li><a href="#features">特徴</a></li>
+<li><a href="#installation">インストール</a></li>
+<li><a href="#depotapi">Depot: 基本API</a></li>
+<li><a href="#depotcli">Depot用コマンド</a></li>
+<li><a href="#curiaapi">Curia: æ‹¡å¼µAPI</a></li>
+<li><a href="#curiacli">Curia用コマンド</a></li>
+<li><a href="#relicapi">Relic: NDBM互æ›API</a></li>
+<li><a href="#reliccli">Relic用コマンド</a></li>
+<li><a href="#hovelapi">Hovel: GDBM互æ›API</a></li>
+<li><a href="#hovelcli">Hovel用コマンド</a></li>
+<li><a href="#cabinapi">Cabin: ユーティリティAPI</a></li>
+<li><a href="#cabincli">Cabin用コマンド</a></li>
+<li><a href="#villaapi">Villa: 上級API</a></li>
+<li><a href="#villacli">Villa用コマンド</a></li>
+<li><a href="#odeumapi">Odeum: 転置API</a></li>
+<li><a href="#odeumcli">Odeum用コマンド</a></li>
+<li><a href="#fileformat">ファイルフォーマット</a></li>
+<li><a href="#porting">移æ¤æ–¹æ³•</a></li>
+<li><a href="#bugs">ãƒã‚°</a></li>
+<li><a href="#faq">よãèžã‹ã‚Œã‚‹è³ªå•</a></li>
+<li><a href="#copying">ライセンス</a></li>
+</ol>
+
+<hr />
+
+<h2><a name="overview" id="overview" class="head">概è¦</a></h2>
+
+<p>QDBMã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†ãƒ«ãƒ¼ãƒãƒ³ç¾¤ã®ãƒ©ã‚¤ãƒ–ラリã§ã‚る。データベースã¨ã„ã£ã¦ã‚‚å˜ç´”ãªã‚‚ã®ã§ã‚ã‚Šã€ã‚ーã¨å€¤ã®ãƒšã‚¢ã‹ã‚‰ãªã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ç¾¤ã‚’æ ¼ç´ã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ•ã‚¡ã‚¤ãƒ«ã§ã‚る。ã‚ーã¨å€¤ã¯ä»»æ„ã®é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆåˆ—ã§ã‚ã‚Šã€æ–‡å—列ã§ã‚‚ãƒã‚¤ãƒŠãƒªã§ã‚‚扱ã†ã“ã¨ãŒã§ãる。テーブルやデータ型ã®æ¦‚念ã¯ãªã„。レコードã¯ãƒãƒƒã‚·ãƒ¥è¡¨ã¾ãŸã¯B+木ã§ç·¨æˆã•ã‚Œã‚‹ã€‚</p>
+
+<p>ãƒãƒƒã‚·ãƒ¥è¡¨ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ã€ã‚ーã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã§ä¸€æ„ã§ã‚ã‚Šã€ã‚ーãŒé‡è¤‡ã™ã‚‹è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ã¯ã§ããªã„。ã“ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«å¯¾ã—ã¦ã¯ã€ã‚ーã¨å€¤ã‚’指定ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã—ãŸã‚Šã€ã‚ーを指定ã—ã¦å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã—ãŸã‚Šã€ã‚ーを指定ã—ã¦å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãる。ã¾ãŸã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã—ã¦ã‚ã‚‹å…¨ã¦ã®ã‚ãƒ¼ã‚’é †ä¸åŒã«ä¸€ã¤ãšã¤å–り出ã™ã“ã¨ã‚‚ã§ãる。ã“ã®ã‚ˆã†ãªæ“作ã¯ã€UNIX標準ã§å®šç¾©ã•ã‚Œã¦ã„ã‚‹DBMライブラリãŠã‚ˆã³ãã®è¿½å¾“ã§ã‚ã‚‹NDBMã‚„GDBMã«é¡žã™ã‚‹ã‚‚ã®ã§ã‚る。QDBMã¯DBMã®ã‚ˆã‚Šè‰¯ã„代替ã¨ã—ã¦åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>B+木ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ã€ã‚ーãŒé‡è¤‡ã™ã‚‹è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ãŒã§ãる。ã“ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«å¯¾ã—ã¦ã¯ã€ãƒãƒƒã‚·ãƒ¥è¡¨ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¨åŒæ§˜ã«ã€ã‚ーを指定ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã—ãŸã‚Šå–り出ã—ãŸã‚Šå‰Šé™¤ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãる。レコードã¯ãƒ¦ãƒ¼ã‚¶ãŒæŒ‡ç¤ºã—ãŸæ¯”較関数ã«åŸºã¥ã„ã¦æ•´åˆ—ã•ã‚Œã¦æ ¼ç´ã•ã‚Œã‚‹ã€‚カーソルを用ã„ã¦å„ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ˜‡é †ã¾ãŸã¯é™é †ã§å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãる。ã“ã®æ©Ÿæ§‹ã«ã‚ˆã£ã¦ã€æ–‡å—列ã®å‰æ–¹ä¸€è‡´æ¤œç´¢ã‚„数値ã®ç¯„囲検索ãŒå¯èƒ½ã«ãªã‚‹ã€‚ã¾ãŸã€B+木ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ãŒåˆ©ç”¨ã§ãる。</p>
+
+<p>QDBMã¯Cã§è¨˜è¿°ã•ã‚Œã€Cã€C++ã€Javaã€PerlãŠã‚ˆã³Rubyã®APIã¨ã—ã¦æä¾›ã•ã‚Œã‚‹ã€‚QDBMã¯POSIXæº–æ‹ ã®APIã‚’å‚™ãˆã‚‹ãƒ—ラットフォームã§åˆ©ç”¨ã§ãる。QDBMã¯GNU Lesser General Public Licenseã«åŸºã¥ãフリーソフトウェアã§ã‚る。</p>
+
+<hr />
+
+<h2><a name="features" id="features" class="head">特徴</a></h2>
+
+<h3>効率的ãªãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å®Ÿè£…</h3>
+
+<p>QDBMã¯GDBMã‚’å‚考ã«æ¬¡ã®ä¸‰ç‚¹ã‚’目標ã¨ã—ã¦é–‹ç™ºã•ã‚ŒãŸã€‚処ç†ãŒã‚ˆã‚Šé«˜é€Ÿã§ã‚ã‚‹ã“ã¨ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒã‚ˆã‚Šå°ã•ã„ã“ã¨ã€APIãŒã‚ˆã‚Šå˜ç´”ã§ã‚ã‚‹ã“ã¨ã€‚ã“れらã®ç›®æ¨™ã¯é”æˆã•ã‚Œã¦ã„る。ã¾ãŸã€GDBMã¨åŒæ§˜ã«ã€ä¼çµ±çš„ãªDBMãŒæŠ±ãˆã‚‹ä¸‰ã¤ã®åˆ¶é™äº‹é …を回é¿ã—ã¦ã„る。ã™ãªã‚ã¡ã€ãƒ—ãƒã‚»ã‚¹å†…ã§è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†ã“ã¨ãŒã§ãã€ã‚ーã¨å€¤ã®ã‚µã‚¤ã‚ºã«åˆ¶é™ãŒãªãã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒã‚¹ãƒ‘ースã§ã¯ãªã„。</p>
+
+<p>QDBMã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æŽ¢ç´¢ã«ãƒãƒƒã‚·ãƒ¥ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ を用ã„る。ãƒã‚±ãƒƒãƒˆé…列ã«å分ãªè¦ç´ æ•°ãŒã‚ã‚Œã°ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æŽ¢ç´¢ã«ã‹ã‹ã‚‹æ™‚間計算é‡ã¯ O(1) ã§ã‚る。ã™ãªã‚ã¡ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æŽ¢ç´¢ã«å¿…è¦ãªæ™‚é–“ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®è¦æ¨¡ã«é–¢ã‚らãšä¸€å®šã§ã‚ã‚‹ã€‚è¿½åŠ ã‚„å‰Šé™¤ã«é–¢ã—ã¦ã‚‚åŒæ§˜ã§ã‚る。ãƒãƒƒã‚·ãƒ¥å€¤ã®è¡çªã¯ã‚»ãƒ‘レートãƒã‚§ãƒ¼ãƒ³æ³•ã§ç®¡ç†ã™ã‚‹ã€‚ãƒã‚§ãƒ¼ãƒ³ã®ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã¯äºŒåˆ†æŽ¢ç´¢æœ¨ã§ã‚る。ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ãŒè‘—ã—ãå°‘ãªã„å ´åˆã§ã‚‚ã€æŽ¢ç´¢ç‰ã®æ™‚間計算é‡ã¯ O(log n) ã«æŠ‘ãˆã‚‰ã‚Œã‚‹ã€‚</p>
+
+<p>QDBMã¯ãƒã‚±ãƒƒãƒˆé…列を全ã¦RAM上ã«ä¿æŒã™ã‚‹ã“ã¨ã«ã‚ˆã£ã¦ã€å‡¦ç†ã®é«˜é€ŸåŒ–を図る。ãƒã‚±ãƒƒãƒˆé…列ãŒRAM上ã«ã‚ã‚Œã°ã€ã»ã¼1パスã®ãƒ•ã‚¡ã‚¤ãƒ«æ“作ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã«è©²å½“ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ä¸Šã®é ˜åŸŸã‚’å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãる。ファイルã«è¨˜éŒ²ã•ã‚ŒãŸãƒã‚±ãƒƒãƒˆé…列㯠`read' コールã§RAM上ã«èªã¿è¾¼ã‚€ã®ã§ã¯ãªãã€`mmap' コールã§RAMã«ç›´æŽ¥ãƒžãƒƒãƒ”ングã•ã‚Œã‚‹ã€‚ã—ãŸãŒã£ã¦ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æŽ¥ç¶šã™ã‚‹éš›ã®æº–備時間ãŒæ¥µã‚ã¦çŸãã€ã¾ãŸã€è¤‡æ•°ã®ãƒ—ãƒã‚»ã‚¹ã§ãƒ¡ãƒ¢ãƒªãƒžãƒƒãƒ—を共有ã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ãŒæ ¼ç´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã®åŠåˆ†ã»ã©ã§ã‚ã‚Œã°ã€ãƒ‡ãƒ¼ã‚¿ã®æ€§è³ªã«ã‚ˆã£ã¦å¤šå°‘å‰å¾Œã™ã‚‹ãŒã€ãƒãƒƒã‚·ãƒ¥å€¤ã®è¡çªçŽ‡ã¯56.7%ã»ã©ã§ã‚る（ç‰å€ã ã¨36.8%ã€2å€ã ã¨21.3%ã€4å€ã ã¨11.5%ã€8å€ã ã¨6.0%ã»ã©ï¼‰ã€‚ãã®ã‚ˆã†ãªå ´åˆã€å¹³å‡2パス以下ã®ãƒ•ã‚¡ã‚¤ãƒ«æ“作ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’探索ã™ã‚‹ã“ã¨ãŒã§ãる。ã“れを性能指標ã¨ã™ã‚‹ãªã‚‰ã°ã€ä¾‹ãˆã°100万個ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ãŸã‚ã«ã¯50万è¦ç´ ã®ãƒã‚±ãƒƒãƒˆé…列ãŒæ±‚ã‚られる。ãƒã‚±ãƒƒãƒˆé…列ã®å„è¦ç´ ã¯4ãƒã‚¤ãƒˆã§ã‚る。ã™ãªã‚ã¡ã€2Mãƒã‚¤ãƒˆã®RAMãŒåˆ©ç”¨ã§ãã‚Œã°100万レコードã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒæ§‹ç¯‰ã§ãる。</p>
+
+<p>QDBMã«ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æŽ¥ç¶šã™ã‚‹ãƒ¢ãƒ¼ãƒ‰ã¨ã—ã¦ã€ã€Œãƒªãƒ¼ãƒ€ã€ã¨ã€Œãƒ©ã‚¤ã‚¿ã€ã®äºŒç¨®é¡žãŒã‚る。リーダã¯èªã¿è¾¼ã¿å°‚用ã§ã‚ã‚Šã€ãƒ©ã‚¤ã‚¿ã¯èªã¿æ›¸ã両用ã§ã‚る。データベースã«ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã«ã‚ˆã£ã¦ãƒ—ãƒã‚»ã‚¹é–“ã§ã®æŽ’他制御ãŒè¡Œã‚れる。ライタãŒæŽ¥ç¶šã—ã¦ã„ã‚‹é–“ã¯ã€ä»–ã®ãƒ—ãƒã‚»ã‚¹ã¯ãƒªãƒ¼ãƒ€ã¨ã—ã¦ã‚‚ライタã¨ã—ã¦ã‚‚接続ã§ããªã„。リーダãŒæŽ¥ç¶šã—ã¦ã„ã‚‹é–“ã¯ã€ä»–ã®ãƒ—ãƒã‚»ã‚¹ã®ãƒªãƒ¼ãƒ€ã¯æŽ¥ç¶šã§ãã‚‹ãŒã€ãƒ©ã‚¤ã‚¿ã¯æŽ¥ç¶šã§ããªã„。ã“ã®æ©Ÿæ§‹ã«ã‚ˆã£ã¦ã€ãƒžãƒ«ãƒã‚¿ã‚¹ã‚¯ç’°å¢ƒã§ã®åŒæ™‚接続ã«ä¼´ã†ãƒ‡ãƒ¼ã‚¿ã®æ•´åˆæ€§ãŒä¿è¨¼ã•ã‚Œã‚‹ã€‚</p>
+
+<p>ä¼çµ±çš„ãªDBMã«ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®è¿½åŠ æ“作ã«é–¢ã—ã¦ã€ŒæŒ¿å…¥ã€ãƒ¢ãƒ¼ãƒ‰ã¨ã€Œç½®æ›ã€ãƒ¢ãƒ¼ãƒ‰ãŒã‚る。å‰è€…ã§ã¯ã€ã‚ーãŒæ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨é‡è¤‡ã™ã‚‹éš›ã«æ—¢å˜ã®å€¤ã‚’残ã™ã€‚後者ã§ã¯ã€ã‚ーãŒæ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨é‡è¤‡ã—ãŸéš›ã«æ–°ã—ã„値ã«ç½®ãæ›ãˆã‚‹ã€‚QDBMã¯ãã®ï¼’ã¤ã«åŠ ãˆã¦ã€Œé€£çµã€ãƒ¢ãƒ¼ãƒ‰ãŒã‚る。既å˜ã®å€¤ã®æœ«å°¾ã«æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’連çµã—ã¦æ ¼ç´ã™ã‚‹æ“作ã§ã‚る。レコードã®å€¤ã‚’é…列ã¨ã—ã¦æ‰±ã†å ´åˆã€è¦ç´ ã‚’è¿½åŠ ã™ã‚‹ã«ã¯é€£çµãƒ¢ãƒ¼ãƒ‰ãŒå½¹ã«ç«‹ã¤ã€‚ã¾ãŸã€DBMã§ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–り出ã™éš›ã«ã¯ãã®å…¨ã¦ã®é ˜åŸŸã‚’処ç†å¯¾è±¡ã«ã™ã‚‹ã—ã‹æ–¹æ³•ãŒãªã„ãŒã€QDBMã§ã¯å€¤ã®é ˜åŸŸã®ä¸€éƒ¨ã®ã¿ã‚’é¸æŠžã—ã¦å–り出ã™ã“ã¨ãŒã§ãる。レコードã®å€¤ã‚’é…列ã¨ã—ã¦æ‰±ã†å ´åˆã«ã¯ã“ã®æ©Ÿèƒ½ã‚‚å½¹ã«ç«‹ã¤ã€‚</p>
+
+<p>一般的ã«ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´æ–°å‡¦ç†ã‚’続ã‘ã‚‹ã¨ãƒ•ã‚¡ã‚¤ãƒ«å†…ã®åˆ©ç”¨å¯èƒ½é ˜åŸŸã®æ–片化ãŒèµ·ãã€ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºãŒè‚¥å¤§åŒ–ã—ã¦ã—ã¾ã†ã€‚QDBMã¯éš£æŽ¥ã™ã‚‹ä¸è¦é ˜åŸŸã‚’連çµã—ã¦å†åˆ©ç”¨ã—ã€ã¾ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æœ€é©åŒ–機能を備ãˆã‚‹ã“ã¨ã«ã‚ˆã£ã¦ã“ã®å•é¡Œã«å¯¾å‡¦ã™ã‚‹ã€‚æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’より大ããªã‚µã‚¤ã‚ºã®å€¤ã«ä¸Šæ›¸ãã™ã‚‹å ´åˆã€ãã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é ˜åŸŸã‚’ファイルä¸ã®åˆ¥ã®ä½ç½®ã«ç§»å‹•ã•ã›ã‚‹å¿…è¦ãŒã‚る。ã“ã®å‡¦ç†ã®æ™‚間計算é‡ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚µã‚¤ã‚ºã«ä¾å˜ã™ã‚‹ã®ã§ã€å€¤ã‚’æ‹¡å¼µã—ã¦ã„ãå ´åˆã«ã¯åŠ¹çŽ‡ãŒæ‚ªã„。ã—ã‹ã—ã€QDBMã¯ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã«ã‚ˆã£ã¦ã“ã®å•é¡Œã«å¯¾å‡¦ã™ã‚‹ã€‚増分ãŒãƒ‘ディングã«åŽã¾ã‚Œã°é ˜åŸŸã‚’移動ã•ã›ã‚‹å¿…è¦ã¯ãªã„。</p>
+
+<p>多ãã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ ã§ã¯ã€2GBを越ãˆã‚‹ã‚µã‚¤ã‚ºã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚’扱ã†ã“ã¨ãŒã§ããªã„。ã“ã®å•é¡Œã«å¯¾å‡¦ã™ã‚‹ãŸã‚ã«ã€QDBMã¯è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’å«ã‚€ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‹ã‚‰ãªã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†æ©Ÿèƒ½ã‚’æä¾›ã™ã‚‹ã€‚レコードをã©ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«æ ¼ç´ã™ã‚‹ã‹ã¯ã‚ーã«åˆ¥ã®ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã‚’é©ç”¨ã™ã‚‹ã“ã¨ã«ã‚ˆã£ã¦æ±ºã‚られる。ã“ã®æ©Ÿèƒ½ã«ã‚ˆã£ã¦ã€ç†è«–çš„ã«ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®åˆè¨ˆã‚µã‚¤ã‚ºãŒ1TBã¾ã§ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’構築ã™ã‚‹ã“ã¨ãŒã§ãる。ã¾ãŸã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’複数ã®ãƒ‡ã‚£ã‚¹ã‚¯ã«æŒ¯ã‚Šåˆ†ã‘ã‚‹ã“ã¨ãŒã§ãã‚‹ãŸã‚ã€RAID-0（ストライピング）ã«è¦‹ã‚‰ã‚Œã‚‹ã‚ˆã†ãªæ›´æ–°æ“作ã®é«˜é€ŸåŒ–ãŒæœŸå¾…ã§ãる。NFSç‰ã‚’利用ã™ã‚Œã°è¤‡æ•°ã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚µãƒ¼ãƒã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’分散ã•ã›ã‚‹ã“ã¨ã‚‚ã§ãる。</p>
+
+<h3>便利ãªB+木データベースã®å®Ÿè£…</h3>
+
+<p>B+木データベースã¯ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚ˆã‚Šé…ã„ãŒã€ãƒ¦ãƒ¼ã‚¶ãŒå®šç¾©ã—ãŸé †åºã«åŸºã¥ã„ã¦å„レコードをå‚ç…§ã§ãã‚‹ã“ã¨ãŒç‰¹é•·ã§ã‚る。B+木ã¯è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’整列ã•ã›ãŸçŠ¶æ…‹ã§è«–ç†çš„ãªãƒšãƒ¼ã‚¸ã«ã¾ã¨ã‚ã¦ç®¡ç†ã™ã‚‹ã€‚å„ページã«å¯¾ã—ã¦ã¯B木ã™ãªã‚ã¡å¤šé€²å¹³è¡¡æœ¨ã«ã‚ˆã£ã¦éšŽå±¤åŒ–ã•ã‚ŒãŸç–Žã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ãŒç¶æŒã•ã‚Œã‚‹ã€‚ã—ãŸãŒã£ã¦ã€å„レコードã®æŽ¢ç´¢ç‰ã«ã‹ã‹ã‚‹æ™‚間計算é‡ã¯ O(log n) ã§ã‚る。å„ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’é †ç•ªã«å‚ç…§ã™ã‚‹ãŸã‚ã«ã‚«ãƒ¼ã‚½ãƒ«ãŒæä¾›ã•ã‚Œã‚‹ã€‚カーソルã®å ´æ‰€ã¯ã‚ーを指定ã—ã¦é£›ã°ã™ã“ã¨ãŒã§ãã€ã¾ãŸç¾åœ¨ã®å ´æ‰€ã‹ã‚‰æ¬¡ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«é€²ã‚ãŸã‚Šå‰ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«æˆ»ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãる。å„ページã¯åŒæ–¹å‘リンクリストã§ç·¨æˆã•ã‚Œã‚‹ã®ã§ã€ã‚«ãƒ¼ã‚½ãƒ«ã‚’å‰å¾Œã«ç§»å‹•ã•ã›ã‚‹æ™‚間計算é‡ã¯ O(1) ã§ã‚る。</p>
+
+<p>B+木データベースã¯ä¸Šè¿°ã®ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’基盤ã¨ã—ã¦å®Ÿè£…ã•ã‚Œã‚‹ã€‚B+木ã®å„ページã¯ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã—ã¦è¨˜éŒ²ã•ã‚Œã‚‹ã®ã§ã€ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®è¨˜æ†¶ç®¡ç†ã®åŠ¹çŽ‡æ€§ã‚’継承ã—ã¦ã„る。B+木ã§ã¯å„レコードã®ãƒ˜ãƒƒãƒ€ãŒå°ã•ãã€å„ページã®ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã¯ãƒšãƒ¼ã‚¸ã‚µã‚¤ã‚ºã«å¿œã˜ã¦èª¿æ•´ã•ã‚Œã‚‹ã®ã§ã€ã»ã¨ã‚“ã©ã®å ´åˆã€ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¼ƒã¹ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºãŒåŠæ¸›ã™ã‚‹ã€‚B+木を更新ã™ã‚‹éš›ã«ã¯å¤šãã®ãƒšãƒ¼ã‚¸ã‚’æ“作ã™ã‚‹å¿…è¦ãŒã‚ã‚‹ãŒã€QDBMã¯ãƒšãƒ¼ã‚¸ã‚’ã‚ャッシュã™ã‚‹ã“ã¨ã«ã‚ˆã£ã¦ãƒ•ã‚¡ã‚¤ãƒ«æ“作を減らã—ã¦å‡¦ç†ã‚’効率化ã™ã‚‹ã€‚ã»ã¨ã‚“ã©ã®å ´åˆã€ç–Žã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹å…¨ä½“ãŒãƒ¡ãƒ¢ãƒªä¸Šã«ã‚ャッシュã•ã‚Œã‚‹ã®ã§ã€å„レコードをå‚ç…§ã™ã‚‹ã®ã«å¿…è¦ãªãƒ•ã‚¡ã‚¤ãƒ«æ“作ã¯å¹³å‡1パス以下ã§ã‚る。</p>
+
+<p>B+木データベースã¯ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³æ©Ÿæ§‹ã‚’æä¾›ã™ã‚‹ã€‚トランザクションを開始ã—ã¦ã‹ã‚‰çµ‚了ã™ã‚‹ã¾ã§ã®ä¸€é€£ã®æ“作を一括ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ã‚³ãƒŸãƒƒãƒˆã—ãŸã‚Šã€ä¸€é€£ã®æ›´æ–°æ“ä½œã‚’ç ´æ£„ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’トランザクションã®é–‹å§‹å‰ã®çŠ¶æ…‹ã«ãƒãƒ¼ãƒ«ãƒãƒƒã‚¯ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãる。トランザクションã®é–“ã«ã‚¢ãƒ—リケーションã®ãƒ—ãƒã‚»ã‚¹ãŒã‚¯ãƒ©ãƒƒã‚·ãƒ¥ã—ã¦ã‚‚データベースファイルã¯ç ´å£Šã•ã‚Œãªã„。</p>
+
+<p>å¯é€†ãƒ‡ãƒ¼ã‚¿åœ§ç¸®ãƒ©ã‚¤ãƒ–ラリã§ã‚ã‚‹ZLIBã‹LZOã‹BZIP2を有効化ã—ã¦QDBMをビルドã™ã‚‹ã¨ã€B+木ã®å„ページã®å†…容ã¯åœ§ç¸®ã•ã‚Œã¦ãƒ•ã‚¡ã‚¤ãƒ«ã«æ›¸ãè¾¼ã¾ã‚Œã‚‹ã€‚åŒä¸€ãƒšãƒ¼ã‚¸å†…ã®å„レコードã¯ä¼¼ãŸã‚ˆã†ãªãƒ‘ターンをæŒã¤ãŸã‚ã€Lempel-Zivãªã©ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’é©ç”¨ã™ã‚‹ã¨é«˜ã„圧縮効率ãŒæœŸå¾…ã§ãる。テã‚ストデータを扱ã†å ´åˆã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ã‚µã‚¤ã‚ºãŒå…ƒã®25%程度ã«ãªã‚‹ã€‚データベースã®è¦æ¨¡ãŒå¤§ããディスクI/OãŒãƒœãƒˆãƒ«ãƒãƒƒã‚¯ã¨ãªã‚‹å ´åˆã¯ã€åœ§ç¸®æ©Ÿèƒ½ã‚’有効化ã™ã‚‹ã¨å‡¦ç†é€Ÿåº¦ãŒå¤§å¹…ã«æ”¹å–„ã•ã‚Œã‚‹ã€‚</p>
+
+<h3>å˜ç´”ã ãŒå¤šæ§˜ãªã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ç¾¤</h3>
+
+<p>QDBMã®APIã¯éžå¸¸ã«å˜ç´”ã§ã‚る。ANSI Cã§å®šç¾©ã•ã‚ŒãŸ `FILE' ãƒã‚¤ãƒ³ã‚¿ã‚’用ã„ãŸé€šå¸¸ã®ãƒ•ã‚¡ã‚¤ãƒ«å…¥å‡ºåŠ›ã¨åŒã˜ã‚ˆã†ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã«å¯¾ã™ã‚‹å…¥å‡ºåŠ›ã‚’è¡Œã†ã“ã¨ãŒã§ãる。QDBMã®åŸºæœ¬APIã§ã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å®Ÿä½“ã¯å˜ä¸€ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«è¨˜éŒ²ã•ã‚Œã‚‹ã€‚æ‹¡å¼µAPIã§ã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å®Ÿä½“ã¯å˜ä¸€ã®ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã«å«ã¾ã‚Œã‚‹è¤‡æ•°ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«è¨˜éŒ²ã•ã‚Œã‚‹ã€‚二ã¤ã®APIã¯äº’ã„ã«é…·ä¼¼ã—ã¦ã„ã‚‹ã®ã§ã€ã‚¢ãƒ—リケーションを一方ã‹ã‚‰ä»–æ–¹ã«ç§»æ¤ã™ã‚‹ã“ã¨ã¯ãŸã‚„ã™ã„。</p>
+
+<p>NDBMãŠã‚ˆã³GDBMã«äº’æ›ã™ã‚‹APIã‚‚æä¾›ã•ã‚Œã‚‹ã€‚NDBMã‚„GDBMã®ã‚¢ãƒ—リケーションã¯å¸‚å ´ã«æ•°å¤šãå˜åœ¨ã™ã‚‹ãŒã€ãれらをQDBMã«ç§»æ¤ã™ã‚‹ã®ã¯ãŸã‚„ã™ã„。ã»ã¨ã‚“ã©ã®å ´åˆã€ãƒ˜ãƒƒãƒ€ãƒ•ã‚¡ã‚¤ãƒ«ã®å–ã‚Šè¾¼ã¿ï¼ˆ#include）を書ãæ›ãˆã¦ã‚³ãƒ³ãƒ‘イルã—ãªãŠã›ã°ã‚ˆã„。ãŸã ã—ã€ã‚ªãƒªã‚¸ãƒŠãƒ«ã®NDBMã‚„GDBMã§ä½œæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’QDBMã§æ‰±ã†ã“ã¨ã¯ã§ããªã„。</p>
+
+<p>メモリ上ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ç°¡å˜ã«æ‰±ã†ãŸã‚ã«ã€ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£APIãŒæä¾›ã•ã‚Œã‚‹ã€‚メモリ確ä¿é–¢æ•°ã¨ã‚½ãƒ¼ãƒˆé–¢æ•°ã¨æ‹¡å¼µå¯èƒ½ãªãƒ‡ãƒ¼ã‚¿ã¨é…列リストã¨ãƒãƒƒã‚·ãƒ¥ãƒžãƒƒãƒ—ç‰ã®å®Ÿè£…ã§ã‚る。ãれらを用ã„ã‚‹ã¨ã€C言語ã§ã‚‚Perlã‚„Rubyç‰ã®ã‚¹ã‚¯ãƒªãƒ—ト言語ã®ã‚ˆã†ãªæ‰‹è»½ã•ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’扱ã†ã“ã¨ãŒã§ãる。</p>
+
+<p>B+木データベースã¯ä¸Šç´šAPIを介ã—ã¦åˆ©ç”¨ã™ã‚‹ã€‚上級APIã¯åŸºæœ¬APIã¨ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£APIを利用ã—ã¦å®Ÿè£…ã•ã‚Œã‚‹ã€‚上級APIも基本APIã‚„æ‹¡å¼µAPIã«é¡žä¼¼ã—ãŸæ›¸å¼ã‚’æŒã¤ã®ã§ã€ä½¿ã„方を覚ãˆã‚‹ã®ã¯å®¹æ˜“ã§ã‚る。</p>
+
+<p>全文検索システムã§åˆ©ç”¨ã•ã‚Œã‚‹è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’扱ã†ãŸã‚ã«ã€è»¢ç½®APIãŒæä¾›ã•ã‚Œã‚‹ã€‚文書群ã®è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’容易ã«æ‰±ã†ã“ã¨ãŒã§ãã‚Œã°ã€ã‚¢ãƒ—リケーションã¯ãƒ†ã‚スト処ç†ã‚„自然言語処ç†ã«æ³¨åŠ›ã§ãる。ã“ã®APIã¯æ–‡å—コードや言語ã«ä¾å˜ã—ãªã„ã®ã§ã€ãƒ¦ãƒ¼ã‚¶ã®å¤šæ§˜ãªè¦æ±‚ã«å¿œãˆã‚‹å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ を実装ã™ã‚‹ã“ã¨ãŒå¯èƒ½ã¨ãªã‚‹ã€‚</p>
+
+<p>QDBMã¯C言語ã®ä»–ã«ã‚‚ã€C++ã€Javaã€PerlãŠã‚ˆã³Rubyã®APIã‚’æä¾›ã™ã‚‹ã€‚C言語ã®APIã«ã¯ã€åŸºæœ¬APIã€æ‹¡å¼µAPIã€NDBM互æ›APIã€GDBM互æ›APIã€ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£APIã€ä¸Šç´šAPIãŠã‚ˆã³è»¢ç½®APIã®ä¸ƒç¨®é¡žãŒã‚る。å„APIã«å¯¾å¿œã—ãŸã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã‚‚用æ„ã•ã‚Œã¦ã„る。ãれらã¯ãƒ—ãƒãƒˆã‚¿ã‚¤ãƒ”ングやテストやデãƒãƒƒã‚°ãªã©ã§æ´»èºã™ã‚‹ã€‚C++用APIã¯åŸºæœ¬APIã¨æ‹¡å¼µAPIã¨ä¸Šç´šAPIã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹æ“作関数群をC++ã®ã‚¯ãƒ©ã‚¹æ©Ÿæ§‹ã§ã‚«ãƒ—セル化ã—ãŸã‚‚ã®ã§ã‚る。Java用APIã¯Java Native Interfaceを用ã„ã¦åŸºæœ¬APIã¨æ‹¡å¼µAPIã¨ä¸Šç´šAPIを呼ã³å‡ºã™ã‚‚ã®ã§ã‚る。Perl用APIã¯XS言語を用ã„ã¦åŸºæœ¬APIã¨æ‹¡å¼µAPIã¨ä¸Šç´šAPIを呼ã³å‡ºã™ã‚‚ã®ã§ã‚る。Ruby用APIã¯Rubyã®ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«ã¨ã—ã¦åŸºæœ¬APIã¨æ‹¡å¼µAPIã¨ä¸Šç´šAPIを呼ã³å‡ºã™ã‚‚ã®ã§ã‚る。データベースã®ç®¡ç†ã¨ãƒ•ã‚¡ã‚¤ãƒ«ã‚¢ãƒƒãƒ—ãƒãƒ¼ãƒ‰ã¨å…¨æ–‡æ¤œç´¢ã®ãŸã‚ã®CGIスクリプ
トもæä¾›ã•ã‚Œã‚‹ã€‚</p>
+
+<h3>幅広ã„移æ¤æ€§</h3>
+
+<p>QDBMã¯ANSI C（C89）ã®è¨˜æ³•ã«å¾“ã„ã€ANSI Cã¾ãŸã¯POSIXã§å®šç¾©ã•ã‚ŒãŸAPIã®ã¿ã‚’用ã„ã¦å®Ÿè£…ã•ã‚Œã‚‹ã€‚ã—ãŸãŒã£ã¦ã€ã»ã¨ã‚“ã©ã®UNIXãŠã‚ˆã³ãã®äº’æ›ã‚’ã†ãŸã†OSã§å‹•ä½œã•ã›ã‚‹ã“ã¨ãŒã§ãる。C言語ã®APIã«é–¢ã—ã¦ã¯ã€å°‘ãªãã¨ã‚‚以下ã®ãƒ—ラットフォームã§å‹•ä½œç¢ºèªã•ã‚Œã¦ã„る。</p>
+
+<ul>
+<li>Linux (2.2, 2.4, 2.6) (IA32, IA64, AMD64, PA-RISC, Alpha, PowerPC, M68000, ARM)</li>
+<li>FreeBSD (4.9, 5.0, 5.1, 5.2, 5.3) (IA32, IA64, SPARC, Alpha)</li>
+<li>NetBSD (1.6) (IA32)</li>
+<li>OpenBSD (3.4) (IA32)</li>
+<li>SunOS (5.6, 5.7, 5.8, 5.9, 5.10) (IA32, SPARC)</li>
+<li>HP-UX (11.11, 11.23) (IA64, PA-RISC)</li>
+<li>AIX (5.2) (POWER)</li>
+<li>Windows (2000, XP) (IA32, IA64, AMD64) (Cygwin, MinGW, Visual C++)</li>
+<li>Mac OS X (10.2, 10.3, 10.4) (IA32, PowerPC)</li>
+<li>Tru64 (5.1) (Alpha)</li>
+<li>RISC OS (5.03) (ARM)</li>
+</ul>
+
+<p>QDBMãŒä½œæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã¯å‡¦ç†ç³»ã®ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã«ä¾å˜ã™ã‚‹ãŒã€ãã®å¯¾ç–ã¨ã—ã¦ã€ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã«ä¾å˜ã—ãªã„å½¢å¼ã®ãƒ‡ãƒ¼ã‚¿ã‚’ダンプã™ã‚‹ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ãŒæä¾›ã•ã‚Œã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="installation" id="installation" class="head">インストール</a></h2>
+
+<h3>準備</h3>
+
+<p>ソースパッケージを用ã„ã¦QDBMをインストールã™ã‚‹ã«ã¯ã€GCCã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³2.8以é™ã¨ `make' ãŒå¿…è¦ã§ã‚る。</p>
+
+<p>QDBMã®é…布用アーカイブファイルを展開ã—ãŸã‚‰ã€ç”Ÿæˆã•ã‚ŒãŸãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã«å…¥ã£ã¦ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ä½œæ¥ã‚’è¡Œã†ã€‚</p>
+
+<h3>普通ã®æ‰‹é †</h3>
+
+<p>Linuxã¨BSDã¨SunOSã§ã¯ä»¥ä¸‹ã®æ‰‹é †ã«å¾“ã†ã€‚</p>
+
+<p>ビルド環境をè¨å®šã™ã‚‹ã€‚</p>
+
+<pre>./configure
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã€‚</p>
+
+<pre>make
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ ã®è‡ªå·±è¨ºæ–テストを行ã†ã€‚</p>
+
+<pre>make check
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をインストールã™ã‚‹ã€‚作æ¥ã¯ `root' ユーザã§è¡Œã†ã€‚</p>
+
+<pre>make install
+</pre>
+
+<h3>GNU Libtoolを使ã†å ´åˆ</h3>
+
+<p>上記ã®æ–¹æ³•ã§ã†ã¾ãã„ã‹ãªã„å ´åˆã€ä»¥ä¸‹ã®æ‰‹é †ã«å¾“ã†ã€‚ã“ã®æ‰‹é †ã«ã¯ã€GNU Libtoolã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³1.5以é™ãŒå¿…è¦ã§ã‚る。</p>
+
+<p>ビルド環境をè¨å®šã™ã‚‹ã€‚</p>
+
+<pre>./configure
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã€‚</p>
+
+<pre>make -f LTmakefile
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ ã®è‡ªå·±è¨ºæ–テストを行ã†ã€‚</p>
+
+<pre>make -f LTmakefile check
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をインストールã™ã‚‹ã€‚作æ¥ã¯ `root' ユーザã§è¡Œã†ã€‚</p>
+
+<pre>make -f LTmakefile install
+</pre>
+
+<h3>çµæžœ</h3>
+
+<p>一連ã®ä½œæ¥ãŒçµ‚ã‚‹ã¨ã€ä»¥ä¸‹ã®ãƒ•ã‚¡ã‚¤ãƒ«ãŒã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã•ã‚Œã‚‹ã€‚ãã®ä»–ã«ã‚‚ã€ãƒžãƒ‹ãƒ¥ã‚¢ãƒ«ãŒ `/usr/local/man/man1' 㨠`/usr/local/man/man3' ã«ã€ãれ以外ã®æ–‡æ›¸ãŒ `/usr/local/share/qdbm' ã«ã€`pkg-config' 用ã®è¨å®šãƒ•ã‚¡ã‚¤ãƒ«ãŒ `/usr/local/lib/pkgconfig' ã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã•ã‚Œã‚‹ã€‚</p>
+
+<pre>/usr/local/include/depot.h
+/usr/local/include/curia.h
+/usr/local/include/relic.h
+/usr/local/include/hovel.h
+/usr/local/include/cabin.h
+/usr/local/include/villa.h
+/usr/local/include/vista.h
+/usr/local/include/odeum.h
+/usr/local/lib/libqdbm.a
+/usr/local/lib/libqdbm.so.14.13.0
+/usr/local/lib/libqdbm.so.14
+/usr/local/lib/libqdbm.so
+/usr/local/bin/dpmgr
+/usr/local/bin/dptest
+/usr/local/bin/dptsv
+/usr/local/bin/crmgr
+/usr/local/bin/crtest
+/usr/local/bin/crtsv
+/usr/local/bin/rlmgr
+/usr/local/bin/rltest
+/usr/local/bin/hvmgr
+/usr/local/bin/hvtest
+/usr/local/bin/cbtest
+/usr/local/bin/cbcodec
+/usr/local/bin/vlmgr
+/usr/local/bin/vltest
+/usr/local/bin/vltsv
+/usr/local/bin/odmgr
+/usr/local/bin/odtest
+/usr/local/bin/odidx
+/usr/local/bin/qmttest
+</pre>
+
+<p>`libqdbm.so' ã¨å‹•çš„ã«ãƒªãƒ³ã‚¯ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ を実行ã™ã‚‹éš›ã«ã¯ã€ãƒ©ã‚¤ãƒ–ラリã®æ¤œç´¢ãƒ‘ス㫠`/usr/local/lib' ã‚’å«ã‚ã‚‹ã¹ãã§ã‚る。環境変数 `LD_LIBRARY_PATH' ã§ãƒ©ã‚¤ãƒ–ラリã®æ¤œç´¢ãƒ‘スをè¨å®šã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>QDBMをアンインストールã™ã‚‹ã«ã¯ã€`./configure' ã‚’ã—ãŸå¾Œã®çŠ¶æ…‹ã§ä»¥ä¸‹ã®ã‚³ãƒžãƒ³ãƒ‰ã‚’実行ã™ã‚‹ã€‚作æ¥ã¯ `root' ユーザã§è¡Œã†ã€‚</p>
+
+<pre>make uninstall
+</pre>
+
+<p>QDBMã®å¤ã„ãƒãƒ¼ã‚¸ãƒ§ãƒ³ãŒã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã•ã‚Œã¦ã„ã‚‹å ´åˆã€ãれをアンインストールã—ã¦ã‹ã‚‰ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ä½œæ¥ã‚’è¡Œã†ã¹ãã§ã‚る。</p>
+
+<p>C言語以外ã®APIã¨CGIスクリプトã¯ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã§ã¯ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã•ã‚Œãªã„。C++用APIã®ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€ã‚µãƒ–ディレクトリ `plus' ã«ã‚ã‚‹ `xspex-ja.html' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚JAVA用APIã®ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€ã‚µãƒ–ディレクトリ `java' ã«ã‚ã‚‹ `jspex-ja.html' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚Perl用APIã®ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€ã‚µãƒ–ディレクトリ `perl' ã«ã‚ã‚‹ `plspex-ja.html' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚Ruby用APIã®ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€ã‚µãƒ–ディレクトリ `ruby' ã«ã‚ã‚‹ `rbspex-ja.html' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚CGIスクリプトã®ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€ã‚µãƒ–ディレクトリ `cgi' ã«ã‚ã‚‹ `cgispex.html' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚</p>
+
+<p>RPMç‰ã®ãƒã‚¤ãƒŠãƒªãƒ‘ッケージを用ã„ã¦ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã‚’è¡Œã†å ´åˆã¯ã€ãã‚Œãžã‚Œã®ãƒ‘ッケージマãƒãƒ¼ã‚¸ãƒ£ã®ãƒžãƒ‹ãƒ¥ã‚¢ãƒ«ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚例ãˆã°ã€RPMを用ã„ã‚‹å ´åˆã€ä»¥ä¸‹ã®ã‚ˆã†ãªã‚³ãƒžãƒ³ãƒ‰ã‚’ `root' ユーザã§å®Ÿè¡Œã™ã‚‹ã€‚</p>
+
+<pre>rpm -ivh qdbm-1.x.x-x.i386.rpm
+</pre>
+
+<h3>Windowsã®å ´åˆ</h3>
+
+<p>Windows（Cygwin）ã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã€ä»¥ä¸‹ã®æ‰‹é †ã«å¾“ã†ã€‚</p>
+
+<p>ビルド環境をè¨å®šã™ã‚‹ã€‚</p>
+
+<pre>./configure
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã€‚</p>
+
+<pre>make win
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ ã®è‡ªå·±è¨ºæ–テストを行ã†ã€‚</p>
+
+<pre>make check-win
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をインストールã™ã‚‹ã€‚ãªãŠã€ã‚¢ãƒ³ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã¯ `make uninstall-win' ã¨ã™ã‚‹ã€‚</p>
+
+<pre>make install-win
+</pre>
+
+<p>Windowsã§ã¯ã€é™çš„ライブラリ `libqdbm.a' ã«åŠ ãˆã¦ã‚¤ãƒ³ãƒãƒ¼ãƒˆãƒ©ã‚¤ãƒ–ラリ `libqdbm.dll.a' ãŒç”Ÿæˆã•ã‚Œã€å‹•çš„ライブラリ `libqdbm.so' ç‰ã®ä»£ã‚ã‚Šã«ãƒ€ã‚¤ãƒŠãƒŸãƒƒã‚¯ãƒªãƒ³ã‚¯ãƒ©ã‚¤ãƒ–ラリ `qdbm.dll' ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚`qdbm.dll' 㯠`/usr/local/bin' ã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã•ã‚Œã‚‹ã€‚</p>
+
+<p>Cygwin環境ã§MinGWを用ã„ã¦ãƒ“ルドã™ã‚‹ã«ã¯ã€`make win' ã®ä»£ã‚ã‚Šã« `make mingw' を用ã„る。Cygwinã®UNIXエミュレーション層を用ã„ã‚‹å ´åˆã€ç”Ÿæˆã•ã‚Œã‚‹ãƒ—ãƒã‚°ãƒ©ãƒ 㯠`cygwin1.dll' ã«ä¾å˜ã—ãŸã‚‚ã®ã«ãªã‚‹ï¼ˆGNU GPLã®å½±éŸ¿ã‚’å—ã‘る）。MinGWã«ã‚ˆã£ã¦Win32ã®ãƒã‚¤ãƒ†ã‚£ãƒ–DLLã¨ãƒªãƒ³ã‚¯ã•ã›ã‚Œã°ã“ã®å•é¡Œã‚’回é¿ã§ãる。</p>
+
+<p>Visual C++を用ã„ã¦ãƒ“ルドã™ã‚‹ã«ã¯ã€`VCmakefile' を編集ã—ã¦ãƒ˜ãƒƒãƒ€ã¨ãƒ©ã‚¤ãƒ–ラリã®æ¤œç´¢ãƒ‘スをè¨å®šã—ãŸä¸Šã§ã€`nmake /f VCMakefile' ã¨ã™ã‚Œã°ã‚ˆã„。生æˆã•ã‚ŒãŸ `qdbm.dll' ã¨ãƒªãƒ³ã‚¯ã™ã‚‹ã‚¢ãƒ—リケーションã¯ã€ã‚³ãƒ³ãƒ‘イラ㮠`/MD' ã¾ãŸã¯ `/MDd' オプションを用ã„㦠`msvcrt.dll' ã¨ãƒªãƒ³ã‚¯ã•ã›ã‚‹å¿…è¦ãŒã‚る。詳細è¨å®šã«é–¢ã—ã¦ã¯ `VCmakefile' ã‚’å‚ç…§ã®ã“ã¨ã€‚</p>
+
+<h3>Mac OS Xã®å ´åˆ</h3>
+
+<p>Mac OS X（Darwin）ã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã€ä»¥ä¸‹ã®æ‰‹é †ã«å¾“ã†ã€‚</p>
+
+<p>ビルド環境をè¨å®šã™ã‚‹ã€‚</p>
+
+<pre>./configure
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã€‚</p>
+
+<pre>make mac
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ ã®è‡ªå·±è¨ºæ–テストを行ã†ã€‚</p>
+
+<pre>make check-mac
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をインストールã™ã‚‹ã€‚ãªãŠã€ã‚¢ãƒ³ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã¯ `make uninstall-mac' ã¨ã™ã‚‹ã€‚</p>
+
+<pre>make install-mac
+</pre>
+
+<p>Mac OS Xã§ã¯ã€`libqdbm.so' ç‰ã®ä»£ã‚ã‚Šã« `libqdbm.dylib' ç‰ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚ライブラリã®æ¤œç´¢ãƒ‘スã®æŒ‡å®šã¯ç’°å¢ƒå¤‰æ•° `DYLD_LIBRARY_PATH' ã§è¡Œã†ã“ã¨ãŒã§ãる。</p>
+
+<h3>HP-UXã®å ´åˆ</h3>
+
+<p>HP-UXã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã€ä»¥ä¸‹ã®æ‰‹é †ã«å¾“ã†ã€‚</p>
+
+<p>ビルド環境をè¨å®šã™ã‚‹ã€‚</p>
+
+<pre>./configure
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã€‚</p>
+
+<pre>make hpux
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ ã®è‡ªå·±è¨ºæ–テストを行ã†ã€‚</p>
+
+<pre>make check-hpux
+</pre>
+
+<p>プãƒã‚°ãƒ©ãƒ をインストールã™ã‚‹ã€‚ãªãŠã€ã‚¢ãƒ³ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã¯ `make uninstall-hpux' ã¨ã™ã‚‹ã€‚</p>
+
+<pre>make install-hpux
+</pre>
+
+<p>HP-UXã§ã¯ã€`libqdbm.so' ç‰ã®ä»£ã‚ã‚Šã« `libqdbm.sl' ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚ライブラリã®æ¤œç´¢ãƒ‘スã®æŒ‡å®šã¯ç’°å¢ƒå¤‰æ•° `SHLIB_PATH' ã§è¡Œã†ã“ã¨ãŒã§ãる。</p>
+
+<h3>RISC OSã®å ´åˆ</h3>
+
+<p>RISC OSã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹å ´åˆã€ä»¥ä¸‹ã®æ‰‹é †ã«å¾“ã†ã€‚</p>
+
+<p>プãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã€‚デフォルトã§ã¯ã‚³ãƒ³ãƒ‘イラ㫠`cc' を用ã„るよã†ã«ãªã£ã¦ã„ã‚‹ãŒã€`gcc' を用ã„ãŸã‘れ㰠`CC=gcc' ã¨ã„ã†å¼•æ•°ã‚’åŠ ãˆã‚Œã°ã‚ˆã„。</p>
+
+<pre>make -f RISCmakefile
+</pre>
+
+<p>一連ã®ä½œæ¥ãŒçµ‚ã‚‹ã¨ã€`libqdbm' ã¨ã„ã†ãƒ©ã‚¤ãƒ–ラリファイル㨠`dpmgr' ç‰ã®ã‚³ãƒžãƒ³ãƒ‰ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚ãれらã®ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«æ–¹æ³•ã¯å®šç¾©ã•ã‚Œã¦ã„ãªã„ã®ã§ã€æ‰‹å‹•ã§ä»»æ„ã®å ´æ‰€ã«ã‚³ãƒ”ーã—ã¦ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹ã“ã¨ã€‚`depot.h' ç‰ã®ãƒ˜ãƒƒãƒ€ãƒ•ã‚¡ã‚¤ãƒ«ã‚‚åŒæ§˜ã«æ‰‹å‹•ã§ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã™ã‚‹ã“ã¨ã€‚</p>
+
+<h3>詳細è¨å®š</h3>
+
+<p>`./configure' を実行ã™ã‚‹éš›ã«ä»¥ä¸‹ã®ã‚ªãƒ—ション引数を指定ã™ã‚‹ã“ã¨ã§ã€ãƒ“ルド方法ã®è©³ç´°ãªè¨å®šã‚’è¡Œã†ã“ã¨ãŒã§ãる。</p>
+
+<ul class="lines">
+<li><kbd>--enable-debug</kbd> : デãƒãƒƒã‚°ç”¨ã«ãƒ“ルドã™ã‚‹ã€‚デãƒãƒƒã‚°ã‚·ãƒ³ãƒœãƒ«ã‚’有効化ã—ã€æœ€é©åŒ–ã‚’è¡Œã‚ãšã€é™çš„ã«ãƒªãƒ³ã‚¯ã™ã‚‹ã€‚</li>
+<li><kbd>--enable-devel</kbd> : 開発用ã«ãƒ“ルドã™ã‚‹ã€‚デãƒãƒƒã‚°ã‚·ãƒ³ãƒœãƒ«ã‚’有効化ã—ã€æœ€é©åŒ–ã‚’è¡Œã„ã€å‹•çš„ã«ãƒªãƒ³ã‚¯ã™ã‚‹ã€‚</li>
+<li><kbd>--enable-stable</kbd> : 安定版ã®ãƒªãƒªãƒ¼ã‚¹ç”¨ã«ãƒ“ルドã™ã‚‹ã€‚ä¿å®ˆçš„ãªæœ€é©åŒ–ã‚’è¡Œã„ã€å‹•çš„ã«ãƒªãƒ³ã‚¯ã™ã‚‹ã€‚</li>
+<li><kbd>--enable-pthread</kbd> : POSIXスレッドを用ã„ã€ã‚°ãƒãƒ¼ãƒãƒ«å¤‰æ•°ã‚’スレッド固有データã¨ã—ã¦æ‰±ã†ã€‚</li>
+<li><kbd>--disable-lock</kbd> : ファイルãƒãƒƒã‚ングãŒå®Ÿè£…ã•ã‚Œã¦ã„ãªã„環境用ã«ãƒ“ルドã™ã‚‹ã€‚</li>
+<li><kbd>--disable-mmap</kbd> : メモリマッピングãŒå®Ÿè£…ã•ã‚Œã¦ã„ãªã„環境用ã«ãƒ“ルドã™ã‚‹ã€‚</li>
+<li><kbd>--enable-zlib</kbd> : ZLIBã«ã‚ˆã‚‹B+木ã¨è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰åœ§ç¸®ã‚’機能ã•ã›ã‚‹ã€‚</li>
+<li><kbd>--enable-lzo</kbd> : LZOã«ã‚ˆã‚‹B+木ã¨è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰åœ§ç¸®ã‚’機能ã•ã›ã‚‹ã€‚</li>
+<li><kbd>--enable-bzip</kbd> : BZIP2ã«ã‚ˆã‚‹B+木ã¨è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰åœ§ç¸®ã‚’機能ã•ã›ã‚‹ã€‚</li>
+<li><kbd>--enable-iconv</kbd> : ICONVã«ã‚ˆã‚‹æ–‡å—コード変æ›ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã‚’機能ã•ã›ã‚‹ã€‚</li>
+</ul>
+
+<p>通常ã€QDBMã¨ãã®ã‚¢ãƒ—リケーション㯠`libqdbm.*' 以外ã®éžæ¨™æº–ã®ãƒ©ã‚¤ãƒ–ラリã«ã¯ä¾å˜ã—ãªã„ã§ãƒ“ルドã™ã‚‹ã“ã¨ãŒã§ãる。ãŸã ã—ã€POSIXスレッドを有効ã«ã—ãŸå ´åˆã¯ `libpthread.*' ã«ä¾å˜ã—ã€ZLIBを有効ã«ã—ãŸå ´åˆã¯ `libz.*' ã«ä¾å˜ã—ã€LZOを有効ã«ã—ãŸå ´åˆã¯ `liblzo2.*' ã«ä¾å˜ã—ã€BZIP2を有効ã«ã—ãŸå ´åˆã¯ `libbz2.*' ã«ä¾å˜ã—ã€ICONVを有効ã«ã—ãŸå ´åˆã¯ `libiconv.*' ã«ä¾å˜ã™ã‚‹ã‚ˆã†ã«ãªã‚‹ã€‚</p>
+
+<p>LZOã®ãƒ©ã‚¤ã‚»ãƒ³ã‚¹ã¯GNU GPLãªã®ã§ã€`liblzo2.*' ã¨ãƒªãƒ³ã‚¯ã—ãŸã‚¢ãƒ—リケーションã¯GNU GPLã®åˆ¶ç´„ã‚’å—ã‘ã‚‹ã“ã¨ã«æ³¨æ„ã™ã‚‹ã“ã¨ã€‚</p>
+
+<hr />
+
+<h2><a name="depotapi" id="depotapi" class="head">Depot: 基本API</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Depotã¯QDBMã®åŸºæœ¬APIã§ã‚る。QDBMãŒæä¾›ã™ã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ç®¡ç†æ©Ÿèƒ½ã®ã»ã¼å…¨ã¦ãŒDepotã«ã‚ˆã£ã¦å®Ÿè£…ã•ã‚Œã‚‹ã€‚ãã®ä»–ã®APIã¯Depotã®ãƒ©ãƒƒãƒ‘ーã«ã™ãŽãªã„。ã—ãŸãŒã£ã¦ã€QDBMã®APIã®ä¸ã§DepotãŒæœ€ã‚‚高速ã«å‹•ä½œã™ã‚‹ã€‚</p>
+
+<p>Depotを使ã†ãŸã‚ã«ã¯ã€`depot.h' 㨠`stdlib.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>Depotã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†éš›ã«ã¯ã€`DEPOT' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ã“ã‚Œã¯ã€`stdio.h' ã®å„種ルーãƒãƒ³ãŒãƒ•ã‚¡ã‚¤ãƒ«å…¥å‡ºåŠ›ã« `FILE' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’用ã„ã‚‹ã®ã«ä¼¼ã¦ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `dpopen' ã§é–‹ãã€é–¢æ•° `dpclose' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。データベースã«è‡´å‘½çš„ãªã‚¨ãƒ©ãƒ¼ãŒèµ·ããŸå ´åˆã¯ã€ä»¥å¾Œãã®ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã™ã‚‹ `dpclose' を除ãå…¨ã¦ã®æ“作ã¯ä½•ã‚‚ã›ãšã«ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã™ã€‚ã²ã¨ã¤ã®ãƒ—ãƒã‚»ã‚¹ã§è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’åŒæ™‚ã«åˆ©ç”¨ã™ã‚‹ã“ã¨ã¯å¯èƒ½ã§ã‚ã‚‹ãŒã€åŒã˜ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®è¤‡æ•°ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’利用ã—ã¦ã¯ãªã‚‰ãªã„。</p>
+
+<h3>API</h3>
+
+<p>外部変数 `dpversion' ã¯ãƒãƒ¼ã‚¸ãƒ§ãƒ³æƒ…å ±ã®æ–‡å—列ã§ã‚る。</p>
+
+<dl>
+<dt><kbd>extern const char *dpversion;</kbd></dt>
+<dd>ã“ã®å¤‰æ•°ã®æŒ‡ã™é ˜åŸŸã¯æ›¸ãè¾¼ã¿ç¦æ¢ã§ã‚る。</dd>
+</dl>
+
+<p>外部変数 `dpecode' ã«ã¯ç›´å‰ã®ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚エラーコードã®è©³ç´°ã«ã¤ã„ã¦ã¯ `depot.h' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚</p>
+
+<dl>
+<dt><kbd>extern int dpecode;</kbd></dt>
+<dd>ã“ã®å¤‰æ•°ã®åˆæœŸå€¤ã¯ `DP_ENOERR' ã§ã‚る。ãã®ä»–ã®å€¤ã¨ã—ã¦ã€`DP_EFATAL'ã€`DP_EMODE'ã€`DP_EBROKEN'ã€`DP_EKEEP'ã€`DP_ENOITEM'ã€`DP_EALLOC'ã€`DP_EMAP'ã€`DP_EOPEN'ã€`DP_ECLOSE'ã€`DP_ETRUNC'ã€`DP_ESYNC'ã€`DP_ESTAT'ã€`DP_ESEEK'ã€`DP_EREAD'ã€`DP_EWRITE'ã€`DP_ELOCK'ã€`DP_EUNLINK'ã€`DP_EMKDIR'ã€`DP_ERMDIR' ãŠã‚ˆã³ `DP_EMISC' ãŒã‚る。</dd>
+</dl>
+
+<p>エラーコードã«å¯¾å¿œã™ã‚‹ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸æ–‡å—列を得るã«ã¯ã€é–¢æ•° `dperrmsg' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *dperrmsg(int <var>ecode</var>);</kbd></dt>
+<dd>`ecode' ã¯ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã‚¨ãƒ©ãƒ¼ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸ã®æ–‡å—列ã§ã‚ã‚Šã€ãã®é ˜åŸŸã¯æ›¸ãè¾¼ã¿ç¦æ¢ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `dpopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>DEPOT *dpopen(const char *<var>name</var>, int <var>omode</var>, int <var>bnum</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚`omode' ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã€`DP_OREADER' ãªã‚‰ãƒªãƒ¼ãƒ€ã€`DP_OWRITER' ãªã‚‰ãƒ©ã‚¤ã‚¿ã¨ãªã‚‹ã€‚`DP_OWRITER' ã®å ´åˆã€`DP_OCREAT' ã¾ãŸã¯ `DP_OTRUNC' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãる。`DP_OCREAT' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒç„¡ã„å ´åˆã«æ–°è¦ä½œæˆã™ã‚‹ã“ã¨ã‚’指示ã—ã€`DP_OTRUNC' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒå˜åœ¨ã—ã¦ã‚‚作り直ã™ã“ã¨ã‚’指示ã™ã‚‹ã€‚`DP_OREADER' 㨠`DP_OWRITER' ã®ä¸¡æ–¹ã§ `DP_ONOLCK' ã¾ãŸã¯ `DP_OLCKNB' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€å‰è€…ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã“ã¨ã‚’指示ã—ã€å¾Œè€…ã¯ãƒ–ãƒãƒƒã‚¯ã›ãšã«ãƒãƒƒã‚¯ã‚’ã‹ã‘ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚`DP_OCREAT' 㯠`DP_OSPARSE' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€ãã‚Œã¯ç”Ÿæˆã•ã‚Œã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’スパースã«ã™ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚`bnum' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®ç›®å®‰ã‚’指å
®šã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹æ™‚ã«æ±ºã‚られã€æœ€é©åŒ–以外ã®æ‰‹æ®µã§å¤‰æ›´ã™ã‚‹ã“ã¨ã¯ã§ããªã„。ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã¯ã€æ ¼ç´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã®åŠåˆ†ã‹ã‚‰4å€ç¨‹åº¦ã«ã™ã‚‹ã®ãŒã‚ˆã„。戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚‹ã‹ã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。ライタ（èªã¿æ›¸ã両用モード）ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’é–‹ãéš›ã«ã¯ãã®ãƒ•ã‚¡ã‚¤ãƒ«ã«å¯¾ã—ã¦æŽ’ä»–ãƒãƒƒã‚¯ãŒã‹ã‘られã€ãƒªãƒ¼ãƒ€ï¼ˆèªã¿è¾¼ã¿å°‚用モード）ã§é–‹ãéš›ã«ã¯å…±æœ‰ãƒãƒƒã‚¯ãŒã‹ã‘られる。ãã®éš›ã«ã¯è©²å½“ã®ãƒãƒƒã‚¯ãŒã‹ã‘られるã¾ã§åˆ¶å¾¡ãŒãƒ–ãƒãƒƒã‚¯ã™ã‚‹ã€‚`DP_ONOLCK' を使ã†å ´åˆã€ã‚¢ãƒ—リケーションãŒæŽ’他制御ã®è²¬ä»»ã‚’è² ã†ã€‚</dd>
+</dl>
+
+<p>データベースã¨ã®æŽ¥ç¶šã‚’é–‰ã˜ã¦ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpclose(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースã®æ›´æ–°å†…容ã¯ã€æŽ¥ç¶šã‚’é–‰ã˜ãŸæ™‚点ã§åˆã‚ã¦ãƒ•ã‚¡ã‚¤ãƒ«ã¨åŒæœŸã•ã‚Œã‚‹ã€‚ライタã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã„ãŸå ´åˆã€é©åˆ‡ã«æŽ¥ç¶šã‚’é–‰ã˜ãªã„ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒç ´å£Šã•ã‚Œã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpput' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpput(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`dmode' ã¯ã‚ーãŒæ—¢å˜ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨é‡è¤‡ã—ãŸéš›ã®åˆ¶å¾¡ã‚’指定ã™ã‚‹ã€‚`DP_DOVER' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’上書ãã—ã€`DP_DKEEP' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’残ã—ã¦ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã—ã€`DP_DCAT' ã¯æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’æ—¢å˜ã®å€¤ã®æœ«å°¾ã«åŠ ãˆã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>レコードを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpout' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpout(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>レコードをå–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpget' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *dpget(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ç„¡åˆ¶é™ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `mall
oc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>レコードをå–å¾—ã—ã¦ãƒãƒƒãƒ•ã‚¡ã«æ›¸ã込むã«ã¯ã€é–¢æ•° `dpgetwb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpgetwb(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, char *<var>vbuf</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚ãã‚Œã¯æ›¸ãè¾¼ã¿ç”¨ã®ãƒãƒƒãƒ•ã‚¡ã®ã‚µã‚¤ã‚ºä»¥ä¸‹ã§ã‚ã‚‹å¿…è¦ãŒã‚る。`vbuf' ã¯æŠ½å‡ºã—ãŸãƒ‡ãƒ¼ã‚¿ã‚’書ã込むãƒãƒƒãƒ•ã‚¡ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒãƒƒãƒ•ã‚¡ã«æ›¸ãè¾¼ã¾ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ã®ã‚µã‚¤ã‚ºã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ -1 ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。書ãè¾¼ã¿ç”¨ãƒãƒƒãƒ•ã‚¡ã®æœ«å°¾ã«çµ‚端文å—ãŒè¿½åŠ ã•ã‚Œãªã„ã“ã¨ã«æ³¨æ„ã™ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>レコードã®å€¤ã®ã‚µã‚¤ã‚ºã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpvsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpvsiz(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“レコードã®å€¤ã®ã‚µã‚¤ã‚ºã§ã‚ã‚‹ãŒã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯ -1 ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ‰ç„¡ã‚’調ã¹ã‚‹ã®ã«ã‚‚便利ã§ã‚る。`dpget' ã¨é•ã£ã¦å®Ÿãƒ‡ãƒ¼ã‚¿ã‚’èªã¿è¾¼ã¾ãªã„ã®ã§åŠ¹çŽ‡ãŒã‚ˆã„。</dd>
+</dl>
+
+<p>データベースã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’åˆæœŸåŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpiterinit' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpiterinit(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。イテレータã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚ŒãŸå…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å‚ç…§ã™ã‚‹ãŸã‚ã«ç”¨ã„られる。</dd>
+</dl>
+
+<p>データベースã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰æ¬¡ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーをå–り出ã™ã«ã¯ã€é–¢æ•° `dpiternext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *dpiternext(DEPOT *<var>depot</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ã‚ãƒ¼ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。イテレータãŒæœ€å¾Œã¾ã§ãã¦è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã‚’ç¹°ã‚Šè¿”ã—ã¦å‘¼ã¶ã“ã¨ã«ã‚ˆã£ã¦å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’一度ãšã¤å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãる。ãŸã ã—ã€ç¹°ã‚Šè¿”ã—ã®é–“ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´æ–°ãŒã‚ã£ãŸå ´åˆã¯ãã®é™ã‚Šã§ã¯ãªã„。ãªãŠã€å–り出ã™ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é †åºã¯åˆ¶å¾¡ã§ããšã€æ ¼ç´ã—ãŸé †ç•ªã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–り出ã›ã‚‹ã¨ã¯é™ã‚‰ã
ªã„。</dd>
+</dl>
+
+<p>データベースã®ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã‚’è¨å®šã™ã‚‹ã«ã¯ã€é–¢æ•° `dpsetalign' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpsetalign(DEPOT *<var>depot</var>, int <var>align</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`align' ã¯ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。アラインメントをè¨å®šã—ã¦ãŠãã¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸Šæ›¸ãã‚’é »ç¹ã«ã™ã‚‹å ´åˆã®å‡¦ç†åŠ¹çŽ‡ãŒè‰¯ããªã‚‹ã€‚アラインメントã«ã¯ã€ä¸€é€£ã®æ›´æ–°æ“作をã—ãŸå¾Œã®çŠ¶æ…‹ã§ã®æ¨™æº–çš„ãªå€¤ã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã®ãŒã‚ˆã„。アラインメントãŒæ£æ•°ã®å ´åˆã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é ˜åŸŸã®ã‚µã‚¤ã‚ºãŒã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®å€æ•°ã«ãªã‚‹ã‚ˆã†ã«ãƒ‘ディングãŒã¨ã‚‰ã‚Œã‚‹ã€‚アラインメントãŒè² æ•°ã®å ´åˆã€`vsiz' を値ã®ã‚µã‚¤ã‚ºã¨ã—ã¦ã€ãƒ‘ディングã®ã‚µã‚¤ã‚ºã¯ `(vsiz / pow(2, abs(align) - 1))' ã¨ã—ã¦ç®—出ã•ã‚Œã‚‹ã€‚アラインメントã®è¨å®šã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ä¿å˜ã•ã‚Œãªã„ã®ã§ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã度ã«æŒ‡å®šã™ã‚‹å¿…è¦ãŒã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºè¨å®šã™ã‚‹ã«ã¯ã€é–¢æ•° `dpsetfbpsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpsetfbpsiz(DEPOT *<var>depot</var>, int <var>size</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。フリーブãƒãƒƒã‚¯ãƒ—ールã®ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã®ã‚µã‚¤ã‚ºã¯16ã§ã‚る。サイズをより大ããã™ã‚‹ã¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸Šæ›¸ãã‚’ç¹°ã‚Šè¿”ã™éš›ã®ç©ºé–“効率ã¯ä¸ŠãŒã‚‹ãŒã€æ™‚間効率ãŒä¸‹ãŒã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースを更新ã—ãŸå†…容をファイルã¨ãƒ‡ãƒã‚¤ã‚¹ã«åŒæœŸã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `dpsync' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpsync(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ãªã„ã†ã¡ã«åˆ¥ãƒ—ãƒã‚»ã‚¹ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’利用ã•ã›ã‚‹å ´åˆã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>データベースを最é©åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpoptimize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpoptimize(DEPOT *<var>depot</var>, int <var>bnum</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`bnum' ã¯æ–°ãŸãªãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ 数を指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ç¾åœ¨ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã«æœ€é©ãªå€¤ãŒæŒ‡å®šã•ã‚Œã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。レコードを削除ã—ãŸã‚Šã€ç½®æ›ãƒ¢ãƒ¼ãƒ‰ã‚„連çµãƒ¢ãƒ¼ãƒ‰ã§æ›¸ãè¾¼ã¿ã‚’ç¹°ã‚Šè¿”ã—ãŸã‚Šã™ã‚‹å ´åˆã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã«ä¸è¦ãªé ˜åŸŸãŒè“„ç©ã™ã‚‹ãŒã€ã“ã®é–¢æ•°ã¯ãれを解消ã™ã‚‹ã®ã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>データベースã®åå‰ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpname' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *dpname(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰åå‰ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®ã‚µã‚¤ã‚ºã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpfsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpfsiz(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpbnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpbnum(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒã‚±ãƒƒãƒˆé…列ã®åˆ©ç”¨æ¸ˆã¿ã®è¦ç´ æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpbusenum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpbusenum(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒã‚±ãƒƒãƒˆé…列ã®åˆ©ç”¨æ¸ˆã¿ã®è¦ç´ æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒã‚±ãƒƒãƒˆé…列ã®å…¨ã¦ã®è¦ç´ ã‚’å‚ç…§ã™ã‚‹ã®ã§ã€åŠ¹çŽ‡ãŒæ‚ªã„。</dd>
+</dl>
+
+<p>データベースã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dprnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dprnum(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースãƒãƒ³ãƒ‰ãƒ«ãŒãƒ©ã‚¤ã‚¿ã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `dpwritable' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpwritable(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ©ã‚¤ã‚¿ãªã‚‰çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã«è‡´å‘½çš„エラーãŒèµ·ããŸã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `dpfatalerror' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpfatalerror(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯è‡´å‘½çš„エラーãŒã‚ã‚Œã°çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®inode番å·ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpinode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpinode(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®inode番å·ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®æœ€çµ‚更新時刻を得るã«ã¯ã€é–¢æ•° `dpmtime' を用ã„る。</p>
+
+<dl>
+<dt><kbd>time_t dpmtime(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æœ€çµ‚更新時刻ã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpfdesc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpfdesc(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã§ã‚る。データベースã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを直接æ“ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。</dd>
+</dl>
+
+<p>データベースファイルを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpremove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>壊れãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’修復ã™ã‚‹ã«ã¯ã€é–¢æ•° `dprepair' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dprepair(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。修復ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒå…ƒæ¥ã‚‚ã—ãã¯æœŸå¾…ã•ã‚Œã‚‹çŠ¶æ…‹ã«æˆ»ã‚‹ä¿è¨¼ã¯ãªã„。</dd>
+</dl>
+
+<p>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’エンディアンéžä¾å˜ã®ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦ãƒ€ãƒ³ãƒ—ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpexportdb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpexportdb(DEPOT *<var>depot</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`name' ã¯å‡ºåŠ›ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>エンディアンéžä¾å˜ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ãƒãƒ¼ãƒ‰ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpimportdb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpimportdb(DEPOT *<var>depot</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`depot' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚データベースã¯ç©ºã§ãªã‘ã‚Œã°ãªã‚‰ãªã„。`name' ã¯å…¥åŠ›ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã‹ã‚‰ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’直接å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `dpsnaffle' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *dpsnaffle(const char *<var>name</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ç„¡åˆ¶é™ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã
¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒåˆ¥ã®ãƒ—ãƒã‚»ã‚¹ã«ãƒãƒƒã‚¯ã•ã‚Œã¦ã„ã¦ã‚‚利用ã§ãã‚‹ãŒã€æœ€æ–°ã®æ›´æ–°ãŒåæ˜ ã•ã‚Œã¦ã„ã‚‹ä¿è¨¼ã¯ãªã„。</dd>
+</dl>
+
+<p>データベースã®å†…部ã§ç”¨ã„ã‚‹ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã¨ã—ã¦ã€é–¢æ•° `dpinnerhash' ãŒã‚る。</p>
+
+<dl>
+<dt><kbd>int dpinnerhash(const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯ã‚ーã‹ã‚‰31ビット長ã®ãƒãƒƒã‚·ãƒ¥å€¤ã‚’算出ã—ãŸå€¤ã§ã‚る。ã“ã®é–¢æ•°ã¯ã‚¢ãƒ—リケーションãŒãƒã‚±ãƒƒãƒˆé…列ã®çŠ¶æ…‹ã‚’予測ã™ã‚‹éš›ã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>データベースã®å†…部ã§ç”¨ã„ã‚‹ãƒãƒƒã‚·ãƒ¥é–¢æ•°ã¨ç‹¬ç«‹ã—ãŸãƒãƒƒã‚·ãƒ¥é–¢æ•°ã¨ã—ã¦ã€é–¢æ•° `dpouterhash' ãŒã‚る。</p>
+
+<dl>
+<dt><kbd>int dpouterhash(const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯ã‚ーã‹ã‚‰31ビット長ã®ãƒãƒƒã‚·ãƒ¥å€¤ã‚’算出ã—ãŸå€¤ã§ã‚る。ã“ã®é–¢æ•°ã¯ã‚¢ãƒ—リケーションãŒãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´ã«ä¸Šã§ãƒãƒƒã‚·ãƒ¥ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ を利用ã™ã‚‹éš›ã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>ã‚る数以上ã®è‡ªç„¶æ•°ã®ç´ æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dpprimenum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dpprimenum(int <var>num</var>);</kbd></dt>
+<dd>`num' ã¯é©å½“ãªè‡ªç„¶æ•°ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã€æŒ‡å®šã—ãŸæ•°ã¨åŒã˜ã‹ã‚ˆã‚Šå¤§ããã‹ã¤ãªã‚‹ã¹ãå°ã•ã„自然数ã®ç´ æ•°ã§ã‚る。ã“ã®é–¢æ•°ã¯ã‚¢ãƒ—リケーションãŒåˆ©ç”¨ã™ã‚‹ãƒã‚±ãƒƒãƒˆé…列ã®ã‚µã‚¤ã‚ºã‚’決ã‚ã‚‹å ´åˆã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>åå‰ã¨å¯¾å¿œã•ã›ã¦é›»è©±ç•ªå·ã‚’æ ¼ç´ã—ã€ãれを検索ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ DEPOT *depot;
+ char *val;
+
+ /* データベースを開ã */
+ if(!(depot = dpopen(DBNAME, DP_OWRITER | DP_OCREAT, -1))){
+ fprintf(stderr, "dpopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ */
+ if(!dpput(depot, NAME, -1, NUMBER, -1, DP_DOVER)){
+ fprintf(stderr, "dpput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* レコードをå–å¾—ã™ã‚‹ */
+ if(!(val = dpget(depot, NAME, -1, 0, -1, NULL))){
+ fprintf(stderr, "dpget: %s\n", dperrmsg(dpecode));
+ } else {
+ printf("Name: %s\n", NAME);
+ printf("Number: %s\n", val);
+ free(val);
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!dpclose(depot)){
+ fprintf(stderr, "dpclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>データベースã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’表示ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ DEPOT *depot;
+ char *key, *val;
+
+ /* データベースを開ã */
+ if(!(depot = dpopen(DBNAME, DP_OREADER, -1))){
+ fprintf(stderr, "dpopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* イテレータをåˆæœŸåŒ–ã™ã‚‹ */
+ if(!dpiterinit(depot)){
+ fprintf(stderr, "dpiterinit: %s\n", dperrmsg(dpecode));
+ }
+
+ /* イテレータを走査ã™ã‚‹ */
+ while((key = dpiternext(depot, NULL)) != NULL){
+ if(!(val = dpget(depot, key, -1, 0, -1, NULL))){
+ fprintf(stderr, "dpget: %s\n", dperrmsg(dpecode));
+ free(key);
+ break;
+ }
+ printf("%s: %s\n", key, val);
+ free(val);
+ free(key);
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!dpclose(depot)){
+ fprintf(stderr, "dpclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Depotを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹ã«ã¯ã€ãƒ©ã‚¤ãƒ–ラリ `libqdbm.a' ã¾ãŸã¯ `libqdbm.so' をリンク対象ã«åŠ ãˆã‚‹å¿…è¦ãŒã‚る。例ãˆã°ã€`sample.c' ã‹ã‚‰ `sample' を作るã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ãƒ“ルドを行ã†ã€‚</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>POSIXスレッドを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã€å¤–部変数 `dpecode' ã¯ã‚¹ãƒ¬ãƒƒãƒ‰å›ºæœ‰ãƒ‡ãƒ¼ã‚¿ã¸ã®å‚ç…§ã¨ã—ã¦æ‰±ã‚ã‚Œã€Depotã®å„関数ã¯ãƒªã‚¨ãƒ³ãƒˆãƒ©ãƒ³ãƒˆã«ãªã‚‹ã€‚ãã®å ´åˆã€ã‚¹ãƒ¬ãƒƒãƒ‰é–“ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="depotcli" id="depotcli" class="head">Depot用コマンド</a></h2>
+
+<p>Depotã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `dpmgr' ã¯Depotã‚„ãã®ã‚¢ãƒ—リケーションã®ãƒ‡ãƒãƒƒã‚°ã«å½¹ç«‹ã¤ãƒ„ールã§ã‚る。データベースを更新ã—ãŸã‚Šã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’調ã¹ãŸã‚Šã™ã‚‹æ©Ÿèƒ½ã‚’æŒã¤ã€‚シェルスクリプトã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚¢ãƒ—リケーションを作るã®ã«ã‚‚利用ã§ãる。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`key' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã€`val' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>dpmgr create [-s] [-bnum <var>num</var>] <var>name</var></kbd></dt>
+<dd>データベースファイルを作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat] [-na] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>ã‚ーã¨å€¤ã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr out [-kx|-ki] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr get [-nl] [-kx|-ki] [-start <var>num</var>] [-max <var>num</var>] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr list [-nl] [-k|-v] [-ox] <var>name</var></kbd></dt>
+<dd>データベース内ã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨å€¤ã‚’タブã¨æ”¹è¡Œã§åŒºåˆ‡ã£ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr optimize [-bnum <var>num</var>] [-na] <var>name</var></kbd></dt>
+<dd>データベースを最é©åŒ–ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr inform [-nl] <var>name</var></kbd></dt>
+<dd>データベースã®é›‘多ãªæƒ…å ±ã‚’å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr remove <var>name</var></kbd></dt>
+<dd>データベースファイルを削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr repair <var>name</var></kbd></dt>
+<dd>壊れãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’修復ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr exportdb <var>name</var> <var>file</var></kbd></dt>
+<dd>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’エンディアンéžä¾å˜ã®ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦ãƒ€ãƒ³ãƒ—ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr importdb [-bnum <var>num</var>] <var>name</var> <var>file</var></kbd></dt>
+<dd>エンディアンéžä¾å˜ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ãƒãƒ¼ãƒ‰ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr snaffle [-kx|-ki] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ãƒãƒƒã‚¯ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‹ã‚‰ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>dpmgr version</kbd></dt>
+<dd>QDBMã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³æƒ…å ±ã‚’æ¨™æº–å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : ファイルをスパースã«ã™ã‚‹ã€‚</li>
+<li><kbd>-bnum <var>num</var></kbd> : ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã‚’ `num' ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-kx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-ki</kbd> : 10進数ã«ã‚ˆã‚‹æ•°å€¤è¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-vx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vi</kbd> : 10進数ã«ã‚ˆã‚‹æ•°å€¤è¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vf</kbd> : åå‰ãŒ `val' ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ‡ãƒ¼ã‚¿ã‚’値ã¨ã—ã¦èªã¿è¾¼ã‚€ã€‚</li>
+<li><kbd>-keep</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«ä¸Šæ›¸ãã›ãšã«ã‚¨ãƒ©ãƒ¼ã«ã™ã‚‹ã€‚</li>
+<li><kbd>-cat</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«å€¤ã‚’末尾ã«è¿½åŠ ã™ã‚‹ã€‚</li>
+<li><kbd>-na</kbd> : アラインメントをè¨å®šã—ãªã„。</li>
+<li><kbd>-nl</kbd> : ファイルãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã。</li>
+<li><kbd>-start</kbd> : 値ã‹ã‚‰å–り出ã™ãƒ‡ãƒ¼ã‚¿ã®é–‹å§‹ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-max</kbd> : 値ã‹ã‚‰å–り出ã™ãƒ‡ãƒ¼ã‚¿ã®æœ€å¤§ã®é•·ã•ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-ox</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—ã¦æ¨™æº–出力を行ã†ã€‚</li>
+<li><kbd>-n</kbd> : 標準出力ã®æœ«å°¾ã«ä»˜åŠ ã•ã‚Œã‚‹æ”¹è¡Œæ–‡å—ã®å‡ºåŠ›ã‚’抑制ã™ã‚‹ã€‚</li>
+<li><kbd>-k</kbd> : ã‚ーã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+<li><kbd>-v</kbd> : 値ã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `dptest' ã¯Depotã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã«ã‚ˆã£ã¦ç”Ÿæˆã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ `dpmgr' ã«ã‚ˆã£ã¦è§£æžã—ãŸã‚Šã€`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計ã£ãŸã‚Šã™ã‚‹ã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`rnum' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã€`bnum' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã€`pnum' ã¯ã‚ーã®ãƒ‘ターン数ã€`align' ã¯ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®åŸºæœ¬ã‚µã‚¤ã‚ºã€`fbpsiz' ã¯ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>dptest write [-s] <var>name</var> <var>rnum</var> <var>bnum</var></kbd></dt>
+<dd>`00000001'ã€`00000002' ã®ã‚ˆã†ã«å¤‰åŒ–ã™ã‚‹8ãƒã‚¤ãƒˆã®ã‚ーã¨é©å½“ãª8ãƒã‚¤ãƒˆã®å€¤ã‚’連続ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>dptest read [-wb] <var>name</var></kbd></dt>
+<dd>上記ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹ã€‚</dd>
+<dt><kbd>dptest rcat [-c] <var>name</var> <var>rnum</var> <var>bnum</var> <var>pnum</var> <var>align</var> <var>fbpsiz</var></kbd></dt>
+<dd>ã‚ーãŒã‚る程度é‡è¤‡ã™ã‚‹ã‚ˆã†ã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã®è¿½åŠ ã‚’è¡Œã„ã€é€£çµãƒ¢ãƒ¼ãƒ‰ã§å‡¦ç†ã™ã‚‹ã€‚</dd>
+<dt><kbd>dptest combo <var>name</var></kbd></dt>
+<dd>å„種æ“作ã®çµ„ã¿åˆã‚ã›ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>dptest wicked [-c] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>å„種更新æ“作を無作為ã«é¸æŠžã—ã¦å®Ÿè¡Œã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : ファイルをスパースã«ã™ã‚‹ã€‚</li>
+<li><kbd>-wb</kbd> : 関数 `dpget' ã®ä»£ã‚ã‚Šã«é–¢æ•° `dpgetwb' を用ã„る。</li>
+<li><kbd>-c</kbd> : Cabinã®ãƒžãƒƒãƒ—を使ã£ã¦æ¯”較テストを行ã†ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `dptsv' ã¯ã‚¿ãƒ–区切りã§ã‚ーã¨å€¤ã‚’表ç¾ã—ãŸè¡Œã‹ã‚‰ãªã‚‹TSVファイルã¨Depotã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’相互変æ›ã™ã‚‹ã€‚ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯ã€QDBMã®ä»–ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã‚„ä»–ã®DBMã¨ã®é–“ã§ãƒ‡ãƒ¼ã‚¿ã®äº¤æ›ã‚’è¡Œã†éš›ã«å½¹ç«‹ã¤ã€‚ã¾ãŸã€ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã®é•ã†ã‚·ã‚¹ãƒ†ãƒ ã®é–“ã§ãƒ‡ãƒ¼ã‚¿ã‚’交æ›ã™ã‚‹éš›ã«ã‚‚役立ã¤ã€‚以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åを指定ã™ã‚‹ã€‚`export' サブコマンドã§ã¯TSVã®ãƒ‡ãƒ¼ã‚¿ã¯æ¨™æº–入力ã‹ã‚‰èªã¿è¾¼ã‚€ã€‚ã‚ーãŒé‡è¤‡ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã¯å¾Œè€…を優先ã™ã‚‹ã€‚`-bnum' オプションã®å¼•æ•° `num' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`import' サブコマンドã§ã¯TSVã®ãƒ‡ãƒ¼ã‚¿ãŒæ¨™æº–出力ã«æ›¸ã出ã•ã‚Œã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>dptsv import [-bnum <var>num</var>] [-bin] <var>name</var></kbd></dt>
+<dd>TSVファイルをèªã¿è¾¼ã‚“ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>dptsv export [-bin] <var>name</var></kbd></dt>
+<dd>データベースã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’TSVファイルã¨ã—ã¦å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-bnum <var>num</var></kbd> : ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã‚’ `num' ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-bin</kbd> : Base64å½¢å¼ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’扱ã†ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>Depotã®ã‚³ãƒžãƒ³ãƒ‰ç¾¤ã‚’駆使ã™ã‚‹ã¨ã€ç°¡å˜ãªãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚·ã‚¹ãƒ†ãƒ ãŒæ§‹ç¯‰ã§ãる。例ãˆã° `/etc/password' をユーザåã§æ¤œç´¢ã™ã‚‹ãŸã‚ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>cat /etc/passwd | tr ':' '\t' | dptsv import casket
+</pre>
+
+<p>ãã—ã¦ã€`mikio' ã¨ã„ã†ãƒ¦ãƒ¼ã‚¶ã®æƒ…å ±ã‚’å–り出ã™ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>dpmgr get casket mikio
+</pre>
+
+<p>ã“れらã®ã‚³ãƒžãƒ³ãƒ‰ã¨åŒç‰ã®æ©Ÿèƒ½ã‚’Depotã®APIを用ã„ã¦å®Ÿè£…ã™ã‚‹ã“ã¨ã‚‚容易ã§ã‚る。</p>
+
+<hr />
+
+<h2><a name="curiaapi" id="curiaapi" class="head">Curia: æ‹¡å¼µAPI</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Curiaã¯QDBMã®æ‹¡å¼µAPIã§ã‚ã‚Šã€è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ディレクトリã§ä¸€æ‹¬ã—ã¦æ‰±ã†æ©Ÿèƒ½ã‚’æä¾›ã™ã‚‹ã€‚データベースを複数ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«åˆ†å‰²ã™ã‚‹ã“ã¨ã§ã€ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ ã«ã‚ˆã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã‚µã‚¤ã‚ºã®åˆ¶é™ã‚’回é¿ã™ã‚‹ã“ã¨ãŒã§ãる。複数ã®ãƒ‡ãƒã‚¤ã‚¹ã«ãƒ•ã‚¡ã‚¤ãƒ«ã‚’分散ã•ã›ã‚Œã°ã€ã‚¹ã‚±ãƒ¼ãƒ©ãƒ“リティをå‘上ã•ã›ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>Depotã§ã¯ãƒ•ã‚¡ã‚¤ãƒ«åを指定ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’構築ã™ã‚‹ãŒã€Curiaã§ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªåを指定ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’構築ã™ã‚‹ã€‚指定ã—ãŸãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®ç›´ä¸‹ã«ã¯ã€`depot' ã¨ã„ã†åå‰ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚ã“ã‚Œã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®å±žæ€§ã‚’ä¿æŒã™ã‚‹ã‚‚ã®ã§ã‚ã‚Šã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å®Ÿãƒ‡ãƒ¼ã‚¿ã¯æ ¼ç´ã•ã‚Œãªã„。ãã‚Œã¨ã¯åˆ¥ã«ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’分割ã—ãŸå€‹æ•°ã ã‘ã€4æ¡ã®10進数値ã®åå‰ã‚’æŒã¤ã‚µãƒ–ディレクトリãŒç”Ÿæˆã•ã‚Œã€å„々ã®ã‚µãƒ–ディレクトリã®ä¸ã«ã¯ `depot' ã¨ã„ã†åå‰ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚レコードã®å®Ÿãƒ‡ãƒ¼ã‚¿ã¯ãれらã«æ ¼ç´ã•ã‚Œã‚‹ã€‚例ãˆã°ã€`casket' ã¨ã„ã†åå‰ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã—ã€åˆ†å‰²æ•°ã‚’3ã«ã™ã‚‹å ´åˆã€`casket/depot'ã€`casket/0001/depot'ã€`casket/0002/depot'ã€`casket/0003/depot' ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚データベースを作æˆã™ã‚‹éš›ã«ã™ã
§ã«ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãŒå˜åœ¨ã—ã¦ã„ã¦ã‚‚エラーã¨ã¯ãªã‚‰ãªã„。ã—ãŸãŒã£ã¦ã€äºˆã‚サブディレクトリを生æˆã—ã¦ãŠã„ã¦ã€å„々ã«ç•°ãªã‚‹ãƒ‡ãƒã‚¤ã‚¹ã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ をマウントã—ã¦ãŠã‘ã°ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’複数ã®ãƒ‡ãƒã‚¤ã‚¹ã«åˆ†æ•£ã•ã›ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>Curiaã«ã¯ãƒ©ãƒ¼ã‚¸ã‚ªãƒ–ジェクトを扱ã†æ©Ÿèƒ½ãŒã‚る。通常ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ãƒ‡ãƒ¼ã‚¿ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã«æ ¼ç´ã•ã‚Œã‚‹ãŒã€ãƒ©ãƒ¼ã‚¸ã‚ªãƒ–ジェクトã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ãƒ‡ãƒ¼ã‚¿ã¯å€‹åˆ¥ã®ãƒ•ã‚¡ã‚¤ãƒ«ã«æ ¼ç´ã•ã‚Œã‚‹ã€‚ラージオブジェクトã®ãƒ•ã‚¡ã‚¤ãƒ«ã¯ãƒãƒƒã‚·ãƒ¥å€¤ã‚’å…ƒã«ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã«åˆ†ã‘ã¦æ ¼ç´ã•ã‚Œã‚‹ã®ã§ã€é€šå¸¸ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ã¯åŠ£ã‚‹ãŒã€ãã‚Œãªã‚Šã®é€Ÿåº¦ã§å‚ç…§ã§ãる。サイズãŒå¤§ããå‚ç…§é »åº¦ãŒä½Žã„データã¯ã€ãƒ©ãƒ¼ã‚¸ã‚ªãƒ–ジェクトã¨ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‹ã‚‰åˆ†é›¢ã™ã¹ãã§ã‚る。ãã†ã™ã‚Œã°ã€é€šå¸¸ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«å¯¾ã™ã‚‹å‡¦ç†é€Ÿåº¦ãŒå‘上ã™ã‚‹ã€‚ラージオブジェクトã®ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªéšŽå±¤ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãŒæ ¼ç´ã•ã‚Œã‚‹ã‚µãƒ–ディレクトリã®ä¸ã® `lob' ã¨ã„ã†åå‰ã®ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®ä¸ã«ä½œã‚‰ã‚Œã‚‹ã€‚通常ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¨ãƒ©ãƒ¼ã‚¸ã‚ªãƒ–ジェクトã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯ã‚ãƒ
¼ç©ºé–“ãŒç•°ãªã‚Šã€äº’ã„ã«å¹²æ¸‰ã™ã‚‹ã“ã¨ã¯ãªã„。</p>
+
+<p>Curiaを使ã†ãŸã‚ã«ã¯ã€`depot.h' 㨠`curia.h' 㨠`stdlib.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <curia.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>Curiaã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†éš›ã«ã¯ã€`CURIA' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ã“ã‚Œã¯ã€`stdio.h' ã®å„種ルーãƒãƒ³ãŒãƒ•ã‚¡ã‚¤ãƒ«å…¥å‡ºåŠ›ã« `FILE' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’用ã„ã‚‹ã®ã«ä¼¼ã¦ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `cropen' ã§é–‹ãã€é–¢æ•° `crclose' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。データベースã«è‡´å‘½çš„ãªã‚¨ãƒ©ãƒ¼ãŒèµ·ããŸå ´åˆã¯ã€ä»¥å¾Œãã®ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã™ã‚‹ `crclose' を除ãå…¨ã¦ã®æ“作ã¯ä½•ã‚‚ã›ãšã«ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã™ã€‚ã²ã¨ã¤ã®ãƒ—ãƒã‚»ã‚¹ã§è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’åŒæ™‚ã«åˆ©ç”¨ã™ã‚‹ã“ã¨ã¯å¯èƒ½ã§ã‚ã‚‹ãŒã€åŒã˜ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®è¤‡æ•°ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’利用ã—ã¦ã¯ãªã‚‰ãªã„。</p>
+
+<p>Curiaã§ã‚‚Depotã¨åŒã˜ã外部変数 `dpecode' ã«ç›´å‰ã®ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚エラーコードã«å¯¾å¿œã™ã‚‹ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸æ–‡å—列を得るã«ã¯ã€é–¢æ•° `dperrmsg' を用ã„る。</p>
+
+<h3>API</h3>
+
+<p>データベースã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `cropen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CURIA *cropen(const char *<var>name</var>, int <var>omode</var>, int <var>bnum</var>, int <var>dnum</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚`omode' ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã€`CR_OREADER' ãªã‚‰ãƒªãƒ¼ãƒ€ã€`CR_OWRITER' ãªã‚‰ãƒ©ã‚¤ã‚¿ã¨ãªã‚‹ã€‚`CR_OWRITER' ã®å ´åˆã€`CR_OCREAT' ã¾ãŸã¯ `CR_OTRUNC' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãる。`CR_OCREAT' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒç„¡ã„å ´åˆã«æ–°è¦ä½œæˆã™ã‚‹ã“ã¨ã‚’指示ã—ã€`CR_OTRUNC' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒå˜åœ¨ã—ã¦ã‚‚作り直ã™ã“ã¨ã‚’指示ã™ã‚‹ã€‚`CR_OREADER' 㨠`CR_OWRITER' ã®ä¸¡æ–¹ã§ `CR_ONOLCK' ã¾ãŸã¯ `CR_OLCKNB' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€å‰è€…ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã“ã¨ã‚’指示ã—ã€å¾Œè€…ã¯ãƒ–ãƒãƒƒã‚¯ã›ãšã«ãƒãƒƒã‚¯ã‚’ã‹ã‘ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚`CR_OCREAT' 㯠`CR_OSPARSE' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€ãã‚Œã¯ç”Ÿæˆã•ã‚Œã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’スパースã«ã™ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚`bnum' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®ç›®å®‰ã
‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹æ™‚ã«æ±ºã‚られã€æœ€é©åŒ–以外ã®æ‰‹æ®µã§å¤‰æ›´ã™ã‚‹ã“ã¨ã¯ã§ããªã„。ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã¯ã€æ ¼ç´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã®åŠåˆ†ã‹ã‚‰4å€ç¨‹åº¦ã«ã™ã‚‹ã®ãŒã‚ˆã„。`dnum' ã¯è¦ç´ データベースã®æ•°ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。データベースファイルã®åˆ†å‰²æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹æ™‚ã«æŒ‡å®šã—ãŸã‚‚ã®ã‹ã‚‰å¤‰æ›´ã™ã‚‹ã“ã¨ã¯ã§ããªã„。データベースファイルã®åˆ†å‰²æ•°ã®æœ€å¤§å€¤ã¯ 512 個ã§ã‚る。戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚‹ã‹ã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。ライタ（èªã¿æ›¸ã両用モード）ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’é–‹ãéš›ã«ã¯ãã®ãƒ•ã‚¡ã‚¤ãƒ«ã«å¯¾ã—ã¦æŽ’ä»–ãƒãƒƒã‚¯ãŒã‹ã‘られã€ãƒªãƒ¼ãƒ€ï¼ˆèªã¿è¾¼ã¿å°‚用モード）ã§é–‹ãéš›ã«ã¯å…±æœ‰ãƒãƒƒã‚¯ã
Œã‹ã‘られる。ãã®éš›ã«ã¯è©²å½“ã®ãƒãƒƒã‚¯ãŒã‹ã‘られるã¾ã§åˆ¶å¾¡ãŒãƒ–ãƒãƒƒã‚¯ã™ã‚‹ã€‚`CR_ONOLCK' を使ã†å ´åˆã€ã‚¢ãƒ—リケーションãŒæŽ’他制御ã®è²¬ä»»ã‚’è² ã†ã€‚</dd>
+</dl>
+
+<p>データベースã¨ã®æŽ¥ç¶šã‚’é–‰ã˜ã¦ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `crclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crclose(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースã®æ›´æ–°å†…容ã¯ã€æŽ¥ç¶šã‚’é–‰ã˜ãŸæ™‚点ã§åˆã‚ã¦ãƒ•ã‚¡ã‚¤ãƒ«ã¨åŒæœŸã•ã‚Œã‚‹ã€‚ライタã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã„ãŸå ´åˆã€é©åˆ‡ã«æŽ¥ç¶šã‚’é–‰ã˜ãªã„ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒç ´å£Šã•ã‚Œã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `crput' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crput(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`dmode' ã¯ã‚ーãŒæ—¢å˜ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨é‡è¤‡ã—ãŸéš›ã®åˆ¶å¾¡ã‚’指定ã™ã‚‹ã€‚`CR_DOVER' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’上書ãã—ã€`CR_DKEEP' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’残ã—ã¦ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã—ã€`DP_DCAT' ã¯æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’æ—¢å˜ã®å€¤ã®æœ«å°¾ã«åŠ ãˆã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>レコードを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `crout' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crout(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>レコードをå–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crget' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *crget(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ç„¡åˆ¶é™ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `mall
oc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>レコードをå–å¾—ã—ã¦ãƒãƒƒãƒ•ã‚¡ã«æ›¸ã込むã«ã¯ã€é–¢æ•° `crgetwb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crgetwb(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, char *<var>vbuf</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚ãã‚Œã¯æ›¸ãè¾¼ã¿ç”¨ã®ãƒãƒƒãƒ•ã‚¡ã®ã‚µã‚¤ã‚ºä»¥ä¸‹ã§ã‚ã‚‹å¿…è¦ãŒã‚る。`vbuf' ã¯æŠ½å‡ºã—ãŸãƒ‡ãƒ¼ã‚¿ã‚’書ã込むãƒãƒƒãƒ•ã‚¡ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒãƒƒãƒ•ã‚¡ã«æ›¸ãè¾¼ã¾ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ã®ã‚µã‚¤ã‚ºã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ -1 ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。書ãè¾¼ã¿ç”¨ãƒãƒƒãƒ•ã‚¡ã®æœ«å°¾ã«çµ‚端文å—ãŒè¿½åŠ ã•ã‚Œãªã„ã“ã¨ã«æ³¨æ„ã™ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>レコードã®å€¤ã®ã‚µã‚¤ã‚ºã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crvsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crvsiz(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“レコードã®å€¤ã®ã‚µã‚¤ã‚ºã§ã‚ã‚‹ãŒã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯ -1 ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ‰ç„¡ã‚’調ã¹ã‚‹ã®ã«ã‚‚便利ã§ã‚る。`crget' ã¨é•ã£ã¦å®Ÿãƒ‡ãƒ¼ã‚¿ã‚’èªã¿è¾¼ã¾ãªã„ã®ã§åŠ¹çŽ‡ãŒã‚ˆã„。</dd>
+</dl>
+
+<p>データベースã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’åˆæœŸåŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `criterinit' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int criterinit(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。イテレータã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚ŒãŸå…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å‚ç…§ã™ã‚‹ãŸã‚ã«ç”¨ã„られる。</dd>
+</dl>
+
+<p>データベースã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰æ¬¡ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーをå–り出ã™ã«ã¯ã€é–¢æ•° `criternext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *criternext(CURIA *<var>curia</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ã‚ãƒ¼ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。イテレータãŒæœ€å¾Œã¾ã§ãã¦è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã‚’ç¹°ã‚Šè¿”ã—ã¦å‘¼ã¶ã“ã¨ã«ã‚ˆã£ã¦å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’一度ãšã¤å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãる。ãŸã ã—ã€ç¹°ã‚Šè¿”ã—ã®é–“ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´æ–°ãŒã‚ã£ãŸå ´åˆã¯ãã®é™ã‚Šã§ã¯ãªã„。ãªãŠã€å–り出ã™ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é †åºã¯åˆ¶å¾¡ã§ããšã€æ ¼ç´ã—ãŸé †ç•ªã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–り出ã›ã‚‹ã¨ã¯é™ã‚‰ã
ªã„。</dd>
+</dl>
+
+<p>データベースã®ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã‚’è¨å®šã™ã‚‹ã«ã¯ã€é–¢æ•° `crsetalign' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crsetalign(CURIA *<var>curia</var>, int <var>align</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`align' ã¯ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。アラインメントをè¨å®šã—ã¦ãŠãã¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸Šæ›¸ãã‚’é »ç¹ã«ã™ã‚‹å ´åˆã®å‡¦ç†åŠ¹çŽ‡ãŒè‰¯ããªã‚‹ã€‚アラインメントã«ã¯ã€ä¸€é€£ã®æ›´æ–°æ“作をã—ãŸå¾Œã®çŠ¶æ…‹ã§ã®æ¨™æº–çš„ãªå€¤ã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã®ãŒã‚ˆã„。アラインメントãŒæ£æ•°ã®å ´åˆã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é ˜åŸŸã®ã‚µã‚¤ã‚ºãŒã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®å€æ•°ã«ãªã‚‹ã‚ˆã†ã«ãƒ‘ディングãŒã¨ã‚‰ã‚Œã‚‹ã€‚アラインメントãŒè² æ•°ã®å ´åˆã€`vsiz' を値ã®ã‚µã‚¤ã‚ºã¨ã—ã¦ã€ãƒ‘ディングã®ã‚µã‚¤ã‚ºã¯ `(vsiz / pow(2, abs(align) - 1))' ã¨ã—ã¦ç®—出ã•ã‚Œã‚‹ã€‚アラインメントã®è¨å®šã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ä¿å˜ã•ã‚Œãªã„ã®ã§ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã度ã«æŒ‡å®šã™ã‚‹å¿…è¦ãŒã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºè¨å®šã™ã‚‹ã«ã¯ã€é–¢æ•° `crsetfbpsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crsetfbpsiz(CURIA *<var>curia</var>, int <var>size</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。フリーブãƒãƒƒã‚¯ãƒ—ールã®ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã®ã‚µã‚¤ã‚ºã¯16ã§ã‚る。サイズをより大ããã™ã‚‹ã¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸Šæ›¸ãã‚’ç¹°ã‚Šè¿”ã™éš›ã®ç©ºé–“効率ã¯ä¸ŠãŒã‚‹ãŒã€æ™‚間効率ãŒä¸‹ãŒã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースを更新ã—ãŸå†…容をファイルã¨ãƒ‡ãƒã‚¤ã‚¹ã«åŒæœŸã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `crsync' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crsync(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ãªã„ã†ã¡ã«åˆ¥ãƒ—ãƒã‚»ã‚¹ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’利用ã•ã›ã‚‹å ´åˆã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>データベースを最é©åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `croptimize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int croptimize(CURIA *<var>curia</var>, int <var>bnum</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`bnum' ã¯æ–°ãŸãªãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ 数を指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ç¾åœ¨ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã«æœ€é©ãªå€¤ãŒæŒ‡å®šã•ã‚Œã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。レコードを削除ã—ãŸã‚Šã€ç½®æ›ãƒ¢ãƒ¼ãƒ‰ã‚„連çµãƒ¢ãƒ¼ãƒ‰ã§æ›¸ãè¾¼ã¿ã‚’ç¹°ã‚Šè¿”ã—ãŸã‚Šã™ã‚‹å ´åˆã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã«ä¸è¦ãªé ˜åŸŸãŒè“„ç©ã™ã‚‹ãŒã€ã“ã®é–¢æ•°ã¯ãれを解消ã™ã‚‹ã®ã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>データベースã®åå‰ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crname' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *crname(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰åå‰ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®ã‚µã‚¤ã‚ºã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crfsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crfsiz(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。戻り値ãŒ2GBを越ãˆãŸå ´åˆã¯æ¡æº¢ã‚ŒãŒèµ·ã“る。</dd>
+</dl>
+
+<p>データベースファイルã®ã‚µã‚¤ã‚ºã®åˆè¨ˆã‚’å€ç²¾åº¦æµ®å‹•å°æ•°ã¨ã—ã¦å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crfsizd' を用ã„る。</p>
+
+<dl>
+<dt><kbd>double crfsizd(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã®åˆè¨ˆã®å€ç²¾åº¦å€¤ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crbnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crbnum(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒã‚±ãƒƒãƒˆé…列ã®åˆ©ç”¨æ¸ˆã¿ã®è¦ç´ æ•°ã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crbusenum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crbusenum(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒã‚±ãƒƒãƒˆé…列ã®åˆ©ç”¨æ¸ˆã¿ã®è¦ç´ æ•°ã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒã‚±ãƒƒãƒˆé…列ã®å…¨ã¦ã®è¦ç´ ã‚’å‚ç…§ã™ã‚‹ã®ã§ã€åŠ¹çŽ‡ãŒæ‚ªã„。</dd>
+</dl>
+
+<p>データベースã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crrnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crrnum(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースãƒãƒ³ãƒ‰ãƒ«ãŒãƒ©ã‚¤ã‚¿ã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `crwritable' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crwritable(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ©ã‚¤ã‚¿ãªã‚‰çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã«è‡´å‘½çš„エラーãŒèµ·ããŸã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `crfatalerror' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crfatalerror(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯è‡´å‘½çš„エラーãŒã‚ã‚Œã°çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースディレクトリã®inode番å·ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crinode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crinode(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®inode番å·ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®æœ€çµ‚更新時刻を得るã«ã¯ã€é–¢æ•° `crmtime' を用ã„る。</p>
+
+<dl>
+<dt><kbd>time_t crmtime(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æœ€çµ‚更新時刻ã§ã‚る。</dd>
+</dl>
+
+<p>データベースディレクトリを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `crremove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>壊れãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’修復ã™ã‚‹ã«ã¯ã€é–¢æ•° `crrepair' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crrepair(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。修復ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒå…ƒæ¥ã‚‚ã—ãã¯æœŸå¾…ã•ã‚Œã‚‹çŠ¶æ…‹ã«æˆ»ã‚‹ä¿è¨¼ã¯ãªã„。</dd>
+</dl>
+
+<p>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’エンディアンéžä¾å˜ã®ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦ãƒ€ãƒ³ãƒ—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crexportdb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crexportdb(CURIA *<var>curia</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`name' ã¯å‡ºåŠ›ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>エンディアンéžä¾å˜ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ãƒãƒ¼ãƒ‰ã™ã‚‹ã«ã¯ã€é–¢æ•° `crimportdb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crimportdb(CURIA *<var>curia</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚データベースã¯ç©ºã§ãªã‘ã‚Œã°ãªã‚‰ãªã„。`name' ã¯å…¥åŠ›ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースディレクトリã‹ã‚‰ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’直接å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crsnaffle' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *crsnaffle(const char *<var>name</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ç„¡åˆ¶é™ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é
˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãŒåˆ¥ã®ãƒ—ãƒã‚»ã‚¹ã«ãƒãƒƒã‚¯ã•ã‚Œã¦ã„ã¦ã‚‚利用ã§ãã‚‹ãŒã€æœ€æ–°ã®æ›´æ–°ãŒåæ˜ ã•ã‚Œã¦ã„ã‚‹ä¿è¨¼ã¯ãªã„。</dd>
+</dl>
+
+<p>ラージオブジェクト用データベースã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `crputlob' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crputlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`dmode' ã¯ã‚ーãŒæ—¢å˜ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨é‡è¤‡ã—ãŸéš›ã®åˆ¶å¾¡ã‚’指定ã™ã‚‹ã€‚`CR_DOVER' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’上書ãã—ã€`CR_DKEEP' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’残ã—ã¦ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã—ã€`DP_DCAT' ã¯æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’æ—¢å˜ã®å€¤ã®æœ«å°¾ã«åŠ ãˆã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>ラージオブジェクト用データベースã‹ã‚‰ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `croutlob' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int croutlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>ラージオブジェクト用データベースã‹ã‚‰ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crgetlob' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *crgetlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`start' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹æœ€åˆã®ãƒã‚¤ãƒˆã®ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚`max' ã¯å€¤ã®é ˜åŸŸã‹ã‚‰æŠ½å‡ºã™ã‚‹ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ç„¡åˆ¶é™ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚å–り出ãã†ã¨ã—ãŸå€¤ã®ã‚µã‚¤ã‚ºãŒ `start' よりå°ã•ã‹ã£ãŸå ´åˆã«ã¯è©²å½“ã¨ã¿ãªã•ãªã„。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `mall
oc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ラージオブジェクト用データベースã«ã‚るレコードã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crgetlobfd' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crgetlobfd(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰è©²å½“ã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ -1 ã‚’è¿”ã™ã€‚戻り値ã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タ㯠`open' コールã§é–‹ã‹ã‚Œã‚‹ã€‚データベースãŒãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã•ã‚ŒãŸå ´åˆã¯ãã®ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã¯æ›¸ãè¾¼ã¿å¯èƒ½ï¼ˆO_RDWR）ã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°æ›¸ãè¾¼ã¿ä¸å¯èƒ½ï¼ˆO_RDONLY）ã§ã‚る。ディスクリプタãŒä¸è¦ã«ãªã£ãŸã‚‰ `close' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ラージオブジェクト用データベースã«ã‚るレコードã®å€¤ã®ã‚µã‚¤ã‚ºã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `crvsizlob' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crvsizlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“レコードã®å€¤ã®ã‚µã‚¤ã‚ºã§ã‚ã‚Šã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯ -1 ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ‰ç„¡ã‚’調ã¹ã‚‹ã®ã«ã‚‚便利ã§ã‚る。`crgetlob' ã¨é•ã£ã¦å®Ÿãƒ‡ãƒ¼ã‚¿ã‚’èªã¿è¾¼ã¾ãªã„ã®ã§åŠ¹çŽ‡ãŒã‚ˆã„。</dd>
+</dl>
+
+<p>ラージオブジェクト用データベースã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `crrnumlob' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int crrnumlob(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>åå‰ã¨å¯¾å¿œã•ã›ã¦é›»è©±ç•ªå·ã‚’æ ¼ç´ã—ã€ãれを検索ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <curia.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ CURIA *curia;
+ char *val;
+
+ /* データベースを開ã */
+ if(!(curia = cropen(DBNAME, CR_OWRITER | CR_OCREAT, -1, -1))){
+ fprintf(stderr, "cropen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ */
+ if(!crput(curia, NAME, -1, NUMBER, -1, CR_DOVER)){
+ fprintf(stderr, "crput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* レコードをå–å¾—ã™ã‚‹ */
+ if(!(val = crget(curia, NAME, -1, 0, -1, NULL))){
+ fprintf(stderr, "crget: %s\n", dperrmsg(dpecode));
+ } else {
+ printf("Name: %s\n", NAME);
+ printf("Number: %s\n", val);
+ free(val);
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!crclose(curia)){
+ fprintf(stderr, "crclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>データベースã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’表示ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <curia.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ CURIA *curia;
+ char *key, *val;
+
+ /* データベースを開ã */
+ if(!(curia = cropen(DBNAME, CR_OREADER, -1, -1))){
+ fprintf(stderr, "cropen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* イテレータをåˆæœŸåŒ–ã™ã‚‹ */
+ if(!criterinit(curia)){
+ fprintf(stderr, "criterinit: %s\n", dperrmsg(dpecode));
+ }
+
+ /* イテレータを走査ã™ã‚‹ */
+ while((key = criternext(curia, NULL)) != NULL){
+ if(!(val = crget(curia, key, -1, 0, -1, NULL))){
+ fprintf(stderr, "crget: %s\n", dperrmsg(dpecode));
+ free(key);
+ break;
+ }
+ printf("%s: %s\n", key, val);
+ free(val);
+ free(key);
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!crclose(curia)){
+ fprintf(stderr, "crclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Curiaを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹æ–¹æ³•ã¯ã€Depotã®å ´åˆã¨å…¨ãåŒã˜ã§ã‚る。</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>POSIXスレッドを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã€å¤–部変数 `dpecode' ã¯ã‚¹ãƒ¬ãƒƒãƒ‰å›ºæœ‰ãƒ‡ãƒ¼ã‚¿ã¸ã®å‚ç…§ã¨ã—ã¦æ‰±ã‚ã‚Œã€Curiaã®å„関数ã¯ãƒªã‚¨ãƒ³ãƒˆãƒ©ãƒ³ãƒˆã«ãªã‚‹ã€‚ãã®å ´åˆã€ã‚¹ãƒ¬ãƒƒãƒ‰é–“ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="curiacli" id="curiacli" class="head">Curia用コマンド</a></h2>
+
+<p>Curiaã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `crmgr' ã¯Curiaã‚„ãã®ã‚¢ãƒ—リケーションã®ãƒ‡ãƒãƒƒã‚°ã«å½¹ç«‹ã¤ãƒ„ールã§ã‚る。データベースを更新ã—ãŸã‚Šã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’調ã¹ãŸã‚Šã™ã‚‹æ©Ÿèƒ½ã‚’æŒã¤ã€‚シェルスクリプトã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚¢ãƒ—リケーションを作るã®ã«ã‚‚利用ã§ãる。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`key' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã€`val' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>crmgr create [-s] [-bnum <var>num</var>] [-dnum <var>num</var>] <var>name</var></kbd></dt>
+<dd>データベースディレクトリを作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat] [-lob] [-na] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>ã‚ーã¨å€¤ã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr out [-kx|-ki] [-lob] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr get [-nl] [-kx|-ki] [-start <var>num</var>] [-max <var>num</var>] [-ox] [-lob] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr list [-nl] [-k|-v] [-ox] <var>name</var></kbd></dt>
+<dd>データベース内ã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨å€¤ã‚’タブã¨æ”¹è¡Œã§åŒºåˆ‡ã£ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr optimize [-bnum <var>num</var>] [-na] <var>name</var></kbd></dt>
+<dd>データベースを最é©åŒ–ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr inform [-nl] <var>name</var></kbd></dt>
+<dd>データベースã®é›‘多ãªæƒ…å ±ã‚’å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr remove <var>name</var></kbd></dt>
+<dd>データベースディレクトリを削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr repair <var>name</var></kbd></dt>
+<dd>壊れãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’修復ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr exportdb <var>name</var> <var>dir</var></kbd></dt>
+<dd>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’エンディアンéžä¾å˜ã®ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦ãƒ€ãƒ³ãƒ—ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr importdb [-bnum <var>num</var>] [-dnum <var>num</var>] <var>name</var> <var>dir</var></kbd></dt>
+<dd>エンディアンéžä¾å˜ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ãƒãƒ¼ãƒ‰ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr snaffle [-kx|-ki] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ãƒãƒƒã‚¯ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‹ã‚‰ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>crmgr version</kbd></dt>
+<dd>QDBMã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³æƒ…å ±ã‚’æ¨™æº–å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : ファイルをスパースã«ã™ã‚‹ã€‚</li>
+<li><kbd>-bnum <var>num</var></kbd> : ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã‚’ `num' ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-dnum <var>num</var></kbd> : データベースファイルã®åˆ†å‰²æ•°ã‚’ `num' ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-kx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-ki</kbd> : 10進数ã«ã‚ˆã‚‹æ•°å€¤è¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-vx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vi</kbd> : 10進数ã«ã‚ˆã‚‹æ•°å€¤è¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vf</kbd> : åå‰ãŒ `val' ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ‡ãƒ¼ã‚¿ã‚’値ã¨ã—ã¦èªã¿è¾¼ã‚€ã€‚</li>
+<li><kbd>-keep</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«ä¸Šæ›¸ãã›ãšã«ã‚¨ãƒ©ãƒ¼ã«ã™ã‚‹ã€‚</li>
+<li><kbd>-cat</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«å€¤ã‚’末尾ã«è¿½åŠ ã™ã‚‹ã€‚</li>
+<li><kbd>-na</kbd> : アラインメントをè¨å®šã—ãªã„。</li>
+<li><kbd>-nl</kbd> : ファイルãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã。</li>
+<li><kbd>-start</kbd> : 値ã‹ã‚‰å–り出ã™ãƒ‡ãƒ¼ã‚¿ã®é–‹å§‹ã‚ªãƒ•ã‚»ãƒƒãƒˆã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-max</kbd> : 値ã‹ã‚‰å–り出ã™ãƒ‡ãƒ¼ã‚¿ã®æœ€å¤§ã®é•·ã•ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-ox</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—ã¦æ¨™æº–出力を行ã†ã€‚</li>
+<li><kbd>-lob</kbd> : ラージオブジェクトを扱ã†ã€‚</li>
+<li><kbd>-n</kbd> : 標準出力ã®æœ«å°¾ã«ä»˜åŠ ã•ã‚Œã‚‹æ”¹è¡Œæ–‡å—ã®å‡ºåŠ›ã‚’抑制ã™ã‚‹ã€‚</li>
+<li><kbd>-k</kbd> : ã‚ーã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+<li><kbd>-v</kbd> : 値ã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `crtest' ã¯Curiaã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。`crtest' ã«ã‚ˆã£ã¦ç”Ÿæˆã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’ `crmgr' ã«ã‚ˆã£ã¦è§£æžã—ãŸã‚Šã€`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計ã£ãŸã‚Šã™ã‚‹ã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`rnum' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã€`bnum' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã€`dnum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åˆ†å‰²æ•°ã€`pnum' ã¯ã‚ーã®ãƒ‘ターン数ã€`align' ã¯ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®åŸºæœ¬ã‚µã‚¤ã‚ºã€`fbpsiz' ã¯ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>crtest write [-s] [-lob] <var>name</var> <var>rnum</var> <var>bnum</var> <var>dnum</var></kbd></dt>
+<dd>`00000001'ã€`00000002' ã®ã‚ˆã†ã«å¤‰åŒ–ã™ã‚‹8ãƒã‚¤ãƒˆã®ã‚ーã¨é©å½“ãª8ãƒã‚¤ãƒˆã®å€¤ã‚’連続ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>crtest read [-wb] [-lob] <var>name</var></kbd></dt>
+<dd>上記ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹ã€‚</dd>
+<dt><kbd>crtest rcat [-c] <var>name</var> <var>rnum</var> <var>bnum</var> <var>dnum</var> <var>pnum</var> <var>align</var> <var>fbpsiz</var></kbd></dt>
+<dd>ã‚ーãŒã‚る程度é‡è¤‡ã™ã‚‹ã‚ˆã†ã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã®è¿½åŠ ã‚’è¡Œã„ã€é€£çµãƒ¢ãƒ¼ãƒ‰ã§å‡¦ç†ã™ã‚‹ã€‚</dd>
+<dt><kbd>crtest combo <var>name</var></kbd></dt>
+<dd>å„種æ“作ã®çµ„ã¿åˆã‚ã›ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>crtest wicked [-c] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>å„種更新æ“作を無作為ã«é¸æŠžã—ã¦å®Ÿè¡Œã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : ファイルをスパースã«ã™ã‚‹ã€‚</li>
+<li><kbd>-lob</kbd> : ラージオブジェクトを扱ã†ã€‚</li>
+<li><kbd>-wb</kbd> : 関数 `crget' ã®ä»£ã‚ã‚Šã«é–¢æ•° `crgetwb' を用ã„る。</li>
+<li><kbd>-c</kbd> : Cabinã®ãƒžãƒƒãƒ—を使ã£ã¦æ¯”較テストを行ã†ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `crtsv' ã¯ã‚¿ãƒ–区切りã§ã‚ーã¨å€¤ã‚’表ç¾ã—ãŸè¡Œã‹ã‚‰ãªã‚‹TSVファイルã¨Curiaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’相互変æ›ã™ã‚‹ã€‚ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯ã€QDBMã®ä»–ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã‚„ä»–ã®DBMã¨ã®é–“ã§ãƒ‡ãƒ¼ã‚¿ã®äº¤æ›ã‚’è¡Œã†éš›ã«å½¹ç«‹ã¤ã€‚ã¾ãŸã€ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã®é•ã†ã‚·ã‚¹ãƒ†ãƒ ã®é–“ã§ãƒ‡ãƒ¼ã‚¿ã‚’交æ›ã™ã‚‹éš›ã«ã‚‚役立ã¤ã€‚以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åを指定ã™ã‚‹ã€‚`export' サブコマンドã§ã¯TSVã®ãƒ‡ãƒ¼ã‚¿ã¯æ¨™æº–入力ã‹ã‚‰èªã¿è¾¼ã‚€ã€‚ã‚ーãŒé‡è¤‡ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã¯å¾Œè€…を優先ã™ã‚‹ã€‚`-bnum' オプションã®å¼•æ•° `num' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`-dnum' オプションã®å¼•æ•° `num' ã¯è¦ç´ データベースã®æ•°ã‚’指定ã™ã‚‹ã€‚`import' サブコマンドã§ã¯TSVã®ãƒ‡ãƒ¼ã‚¿ãŒæ¨™æº–出力ã«æ›¸ã出ã•ã‚Œã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>crtsv import [-bnum <var>num</var>] [-dnum <var>num</var>] [-bin] <var>name</var></kbd></dt>
+<dd>TSVファイルをèªã¿è¾¼ã‚“ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>crtsv export [-bin] <var>name</var></kbd></dt>
+<dd>データベースã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’TSVファイルã¨ã—ã¦å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-bnum <var>num</var></kbd> : ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã‚’ `num' ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-dnum <var>num</var></kbd> : データベースファイルã®åˆ†å‰²æ•°ã‚’ `num' ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-bin</kbd> : Base64å½¢å¼ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’扱ã†ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>Curiaã®ã‚³ãƒžãƒ³ãƒ‰ç¾¤ã‚’駆使ã™ã‚‹ã¨ã€ç°¡å˜ãªãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚·ã‚¹ãƒ†ãƒ ãŒæ§‹ç¯‰ã§ãる。例ãˆã° `/etc/password' をユーザåã§æ¤œç´¢ã™ã‚‹ãŸã‚ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>cat /etc/passwd | tr ':' '\t' | crtsv import casket
+</pre>
+
+<p>ãã—ã¦ã€`mikio' ã¨ã„ã†ãƒ¦ãƒ¼ã‚¶ã®æƒ…å ±ã‚’å–り出ã™ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>crmgr get casket mikio
+</pre>
+
+<p>ã“れらã®ã‚³ãƒžãƒ³ãƒ‰ã¨åŒç‰ã®æ©Ÿèƒ½ã‚’Curiaã®APIを用ã„ã¦å®Ÿè£…ã™ã‚‹ã“ã¨ã‚‚容易ã§ã‚る。</p>
+
+<hr />
+
+<h2><a name="relicapi" id="relicapi" class="head">Relic: NDBM互æ›API</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Relicã¯ã€NDBMã¨äº’æ›ã™ã‚‹APIã§ã‚る。ã™ãªã‚ã¡ã€Depotã®é–¢æ•°ç¾¤ã‚’NDBMã®APIã§åŒ…ã‚“ã ã‚‚ã®ã§ã‚る。Relicを使ã£ã¦NDBMã®ã‚¢ãƒ—リケーションをQDBMã«ç§»æ¤ã™ã‚‹ã®ã¯ãŸã‚„ã™ã„。ã»ã¨ã‚“ã©ã®å ´åˆã€ã‚¤ãƒ³ã‚¯ãƒ«ãƒ¼ãƒ‰ã™ã‚‹ãƒ˜ãƒƒãƒ€ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ `ndbm.h' ã‹ã‚‰ `relic.h' ã«æ›ãˆã€ãƒ“ルドã®éš›ã®ãƒªãƒ³ã‚«ã‚ªãƒ—ションを `-lndbm' ã‹ã‚‰ `-lqdbm' ã«æ›ãˆã‚‹ã ã‘ã§ã‚ˆã„。</p>
+
+<p>オリジナルã®NDBMã§ã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯äºŒã¤ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®å¯¾ã‹ã‚‰ãªã‚‹ã€‚ã²ã¨ã¤ã¯æŽ¥å°¾è¾žã« `.dir' ãŒã¤ãåå‰ã§ã€ã‚ーã®ãƒ“ãƒƒãƒˆãƒžãƒƒãƒ—ã‚’æ ¼ç´ã™ã‚‹ã€Œãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãƒ•ã‚¡ã‚¤ãƒ«ã€ã§ã‚る。もã†ã²ã¨ã¤ã¯æŽ¥å°¾è¾žã« `.pag' ãŒã¤ãåå‰ã§ã€ãƒ‡ãƒ¼ã‚¿ã®å®Ÿä½“ã‚’æ ¼ç´ã™ã‚‹ã€Œãƒ‡ãƒ¼ã‚¿ãƒ•ã‚¡ã‚¤ãƒ«ã€ã§ã‚る。Relicã§ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãƒ•ã‚¡ã‚¤ãƒ«ã¯å˜ãªã‚‹ãƒ€ãƒŸãƒ¼ã¨ã—ã¦ä½œæˆã—ã€ãƒ‡ãƒ¼ã‚¿ãƒ•ã‚¡ã‚¤ãƒ«ã‚’データベースã¨ã™ã‚‹ã€‚Relicã§ã¯ã‚ªãƒªã‚¸ãƒŠãƒ«ã®NDBMã¨é•ã„ã€æ ¼ç´ã™ã‚‹ãƒ‡ãƒ¼ã‚¿ã®ã‚µã‚¤ã‚ºã«åˆ¶é™ã¯ãªã„。ãªãŠã€ã‚ªãƒªã‚¸ãƒŠãƒ«ã®NDBMã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’Relicã§æ‰±ã†ã“ã¨ã¯ã§ããªã„。</p>
+
+<p>Relicを使ã†ãŸã‚ã«ã¯ã€`relic.h' 㨠`stdlib.h' 㨠`sys/types.h' 㨠`sys/stat.h' 㨠`fcntl.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <relic.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+<dt><kbd>#include <sys/types.h></kbd></dt>
+<dt><kbd>#include <sys/stat.h></kbd></dt>
+<dt><kbd>#include <fcntl.h></kbd></dt>
+</dl>
+
+<p>Relicã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†éš›ã«ã¯ã€`DBM' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `dbm_open' ã§é–‹ãã€é–¢æ•° `dbm_close' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。</p>
+
+<h3>API</h3>
+
+<p>データã®æ ¼ç´ã€å‰Šé™¤ã€æ¤œç´¢ã«ç”¨ã„る関数ã¨ã®ãƒ‡ãƒ¼ã‚¿ã®æŽˆå—ã«ã¯ã€ã‚ーã¨å€¤ã‚’表ç¾ã™ã‚‹ã®ã« `datum' åž‹ã®æ§‹é€ 体を用ã„る。</p>
+
+<dl>
+<dt><kbd>typedef struct { void *dptr; size_t dsize; } datum;</kbd></dt>
+<dd>`dptr' ã¯ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。`dsize' ã¯ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `dbm_open' を用ã„る。</p>
+
+<dl>
+<dt><kbd>DBM *dbm_open(char *<var>name</var>, int <var>flags</var>, int <var>mode</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®åå‰ã‚’指定ã™ã‚‹ãŒã€ãƒ•ã‚¡ã‚¤ãƒ«åã¯ãã‚Œã«æŽ¥å°¾è¾žã‚’ã¤ã‘ãŸã‚‚ã®ã«ãªã‚‹ã€‚`flags' 㯠`open' コールã«æ¸¡ã™ã‚‚ã®ã¨åŒã˜ã ãŒã€`O_WRONLY' 㯠`O_RDWR' ã¨åŒã˜ã«ãªã‚Šã€è¿½åŠ フラグã§ã¯ `O_CREAT' 㨠`O_TRUNC' ã®ã¿ãŒæœ‰åŠ¹ã§ã‚る。`mode' 㯠`open' コールã«æ¸¡ã™ã‚‚ã®ã¨åŒã˜ã§ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ¢ãƒ¼ãƒ‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã¨ã®æŽ¥ç¶šã‚’é–‰ã˜ã¦ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `dbm_close' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void dbm_close(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `dbm_store' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dbm_store(DBM *<var>db</var>, datum <var>key</var>, datum <var>content</var>, int <var>flags</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚`content' ã¯å€¤ã®æ§‹é€ 体を指定ã™ã‚‹ã€‚`frags' ㌠`DBM_INSERT' ãªã‚‰ã‚ーã®é‡è¤‡æ™‚ã«æ›¸ãè¾¼ã¿ã‚’æ–念ã—ã€`DBM_REPLACE' ãªã‚‰ä¸Šæ›¸ãã‚’è¡Œã†ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ 0 ã§ã‚ã‚Šã€é‡è¤‡ã§ã®æ–念ãªã‚‰ 1 ã§ã‚ã‚Šã€ãã®ä»–ã®ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>レコードを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `dbm_delete' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dbm_delete(DBM *<var>db</var>, datum <var>key</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ 0 ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>レコードをå–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `dbm_fetch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>datum dbm_fetch(DBM *<var>db</var>, datum <var>key</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚戻り値ã¯å€¤ã®æ§‹é€ 体ã§ã‚る。該当ãŒã‚ã‚Œã°ãƒ¡ãƒ³ãƒ `dptr' ãŒãã®é ˜åŸŸã‚’指ã—ã€ãƒ¡ãƒ³ãƒ `dsize' ãŒãã®ã‚µã‚¤ã‚ºã‚’示ã™ã€‚該当ãŒãªã‘れ㰠`dptr' ã®å€¤ã¯ `NULL' ã¨ãªã‚‹ã€‚`dptr' ã®æŒ‡ã™é ˜åŸŸã¯ãƒãƒ³ãƒ‰ãƒ«ã«é–¢é€£ã¥ã‘られã¦ç¢ºä¿ã•ã‚Œã€åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã—ã¦æ¬¡ã«ã“ã®é–¢æ•°ã‚’呼ã³å‡ºã™ã‹ã€ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã‚‹ã¾ã§ã€æœ‰åŠ¹ãªãƒ‡ãƒ¼ã‚¿ã‚’ä¿æŒã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>最åˆã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーを得るã«ã¯ã€é–¢æ•° `dbm_firstkey' を用ã„る。</p>
+
+<dl>
+<dt><kbd>datum dbm_firstkey(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã‚ーã®æ§‹é€ 体ã§ã‚る。該当ãŒã‚ã‚Œã°ãƒ¡ãƒ³ãƒ `dptr' ãŒãã®é ˜åŸŸã‚’指ã—ã€`dsize' ãŒãã®ã‚µã‚¤ã‚ºã‚’示ã™ã€‚該当ãŒãªã‘れ㰠`dptr' ã®å€¤ã¯ `NULL' ã¨ãªã‚‹ã€‚`dptr' ã®æŒ‡ã™é ˜åŸŸã¯ãƒãƒ³ãƒ‰ãƒ«ã«é–¢é€£ã¥ã‘られã¦ç¢ºä¿ã•ã‚Œã€åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã—ã¦æ¬¡ã«ã“ã®é–¢æ•°ã‚‚ã—ãã¯é–¢æ•° `dbm_nextkey' を呼ã³å‡ºã™ã‹ã€ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã‚‹ã¾ã§ã€æœ‰åŠ¹ãªãƒ‡ãƒ¼ã‚¿ã‚’ä¿æŒã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>次レコードã®ã‚ーを得るã«ã¯ã€é–¢æ•° `dbm_nextkey' を用ã„る。</p>
+
+<dl>
+<dt><kbd>datum dbm_nextkey(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã‚ーã®æ§‹é€ 体ã§ã‚る。該当ãŒã‚ã‚Œã°ãƒ¡ãƒ³ãƒ `dptr' ãŒãã®é ˜åŸŸã‚’指ã—ã€`dsize' ãŒãã®ã‚µã‚¤ã‚ºã‚’示ã™ã€‚該当ãŒãªã‘れ㰠`dptr' ã®å€¤ã¯ `NULL' ã¨ãªã‚‹ã€‚`dptr' ã®æŒ‡ã™é ˜åŸŸã¯ãƒãƒ³ãƒ‰ãƒ«ã«é–¢é€£ã¥ã‘られã¦ç¢ºä¿ã•ã‚Œã€åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã—ã¦æ¬¡ã«ã“ã®é–¢æ•°ã‚‚ã—ãã¯é–¢æ•° `dbm_firstkey' を呼ã³å‡ºã™ã‹ã€ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã‚‹ã¾ã§ã€æœ‰åŠ¹ãªãƒ‡ãƒ¼ã‚¿ã‚’ä¿æŒã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースã«è‡´å‘½çš„エラーãŒèµ·ããŸã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `dbm_error' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dbm_error(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯è‡´å‘½çš„エラーãŒã‚ã‚Œã°çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>関数 `dbm_clearerr' ã¯ä½•ã‚‚ã—ãªã„。</p>
+
+<dl>
+<dt><kbd>int dbm_clearerr(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値㯠0 ã§ã‚る。ã“ã®é–¢æ•°ã¯äº’æ›æ€§ã®ãŸã‚ã«ã®ã¿å˜åœ¨ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースãŒèªã¿è¾¼ã¿å°‚用ã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `dbm_rdonly' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dbm_rdonly(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯èªã¿è¾¼ã¿å°‚用ãªã‚‰çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>ディレクトリファイルã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dbm_dirfno' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dbm_dirfno(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<p>データファイルã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `dbm_pagfno' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int dbm_pagfno(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>åå‰ã¨å¯¾å¿œã•ã›ã¦é›»è©±ç•ªå·ã‚’æ ¼ç´ã—ã€ãれを検索ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <relic.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <stdio.h>
+#include <string.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ DBM *db;
+ datum key, val;
+ int i;
+
+ /* データベースを開ã */
+ if(!(db = dbm_open(DBNAME, O_RDWR | O_CREAT, 00644))){
+ perror("dbm_open");
+ return 1;
+ }
+
+ /* レコードを準備ã™ã‚‹ */
+ key.dptr = NAME;
+ key.dsize = strlen(NAME);
+ val.dptr = NUMBER;
+ val.dsize = strlen(NUMBER);
+
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ */
+ if(dbm_store(db, key, val, DBM_REPLACE) != 0){
+ perror("dbm_store");
+ }
+
+ /* レコードを検索ã™ã‚‹ */
+ val = dbm_fetch(db, key);
+ if(val.dptr){
+ printf("Name: %s\n", NAME);
+ printf("Number: ");
+ for(i = 0; i < val.dsize; i++){
+ putchar(((char *)val.dptr)[i]);
+ }
+ putchar('\n');
+ } else {
+ perror("dbm_fetch");
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ dbm_close(db);
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Relicを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹æ–¹æ³•ã¯ã€Depotã®å ´åˆã¨å…¨ãåŒã˜ã§ã‚る。リンカã«æ¸¡ã™ã‚ªãƒ—ション㯠`-lndbm' ã§ã¯ãªã `-lqdbm' ã§ã‚る。</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>スレッド間ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€Relicã®å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="reliccli" id="reliccli" class="head">Relic用コマンド</a></h2>
+
+<p>Relicã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `rlmgr' ã¯Relicã‚„ãã®ã‚¢ãƒ—リケーションã®ãƒ‡ãƒãƒƒã‚°ã«å½¹ç«‹ã¤ãƒ„ールã§ã‚る。データベースを更新ã—ãŸã‚Šã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’調ã¹ãŸã‚Šã™ã‚‹æ©Ÿèƒ½ã‚’æŒã¤ã€‚シェルスクリプトã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚¢ãƒ—リケーションを作るã®ã«ã‚‚利用ã§ãる。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`key' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã€`val' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>rlmgr create <var>name</var></kbd></dt>
+<dd>データベースファイルを作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>rlmgr store [-kx] [-vx|-vf] [-insert] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>ã‚ーã¨å€¤ã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>rlmgr delete [-kx] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>rlmgr fetch [-kx] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>rlmgr list [-ox] <var>name</var></kbd></dt>
+<dd>データベース内ã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨å€¤ã‚’タブã¨æ”¹è¡Œã§åŒºåˆ‡ã£ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-kx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-vx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vf</kbd> : åå‰ãŒ `val' ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ‡ãƒ¼ã‚¿ã‚’値ã¨ã—ã¦èªã¿è¾¼ã‚€ã€‚</li>
+<li><kbd>-insert</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«ä¸Šæ›¸ãã›ãšã«ã‚¨ãƒ©ãƒ¼ã«ã™ã‚‹ã€‚</li>
+<li><kbd>-ox</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—ã¦æ¨™æº–出力を行ã†ã€‚</li>
+<li><kbd>-n</kbd> : 標準出力ã®æœ«å°¾ã«ä»˜åŠ ã•ã‚Œã‚‹æ”¹è¡Œæ–‡å—ã®å‡ºåŠ›ã‚’抑制ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>コマンド `rltest' ã¯Relicã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã«ã‚ˆã£ã¦ç”Ÿæˆã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ `rlmgr' ã«ã‚ˆã£ã¦è§£æžã—ãŸã‚Šã€`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計ã£ãŸã‚Šã™ã‚‹ã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`rnum' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>rltest write <var>name</var> <var>rnum</var></kbd></dt>
+<dd>`00000001'ã€`00000002' ã®ã‚ˆã†ã«å¤‰åŒ–ã™ã‚‹8ãƒã‚¤ãƒˆã®ã‚ーã¨é©å½“ãª8ãƒã‚¤ãƒˆã®å€¤ã‚’連続ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>rltest read <var>name</var> <var>rnum</var></kbd></dt>
+<dd>上記ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’検索ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="hovelapi" id="hovelapi" class="head">Hovel: GDBM互æ›API</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Hovelã¯ã€GDBMã¨äº’æ›ã™ã‚‹APIã§ã‚る。ã™ãªã‚ã¡ã€DepotãŠã‚ˆã³Curiaã®é–¢æ•°ç¾¤ã‚’GDBMã®APIã§åŒ…ã‚“ã ã‚‚ã®ã§ã‚る。Hovelを使ã£ã¦GDBMã®ã‚¢ãƒ—リケーションをQDBMã«ç§»æ¤ã™ã‚‹ã®ã¯ãŸã‚„ã™ã„。ã»ã¨ã‚“ã©ã®å ´åˆã€ã‚¤ãƒ³ã‚¯ãƒ«ãƒ¼ãƒ‰ã™ã‚‹ãƒ˜ãƒƒãƒ€ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ `gdbm.h' ã‹ã‚‰ `hovel.h' ã«æ›ãˆã€ãƒ“ルドã®éš›ã®ãƒªãƒ³ã‚«ã‚ªãƒ—ションを `-lgdbm' ã‹ã‚‰ `-lqdbm' ã«æ›ãˆã‚‹ã ã‘ã§ã‚ˆã„。ãªãŠã€ã‚ªãƒªã‚¸ãƒŠãƒ«ã®GDBMã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’Hovelã§æ‰±ã†ã“ã¨ã¯ã§ããªã„。</p>
+
+<p>Hovelを使ã†ãŸã‚ã«ã¯ã€`hovel.h' 㨠`stdlib.h' 㨠`sys/types.h' 㨠`sys/stat.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <hovel.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+<dt><kbd>#include <sys/types.h></kbd></dt>
+<dt><kbd>#include <sys/stat.h></kbd></dt>
+</dl>
+
+<p>Hovelã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†éš›ã«ã¯ã€`GDBM_FILE' åž‹ã®ã‚ªãƒ–ジェクト（ãれ自体ãŒãƒã‚¤ãƒ³ã‚¿åž‹ï¼‰ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `gdbm_open' ã§é–‹ãã€é–¢æ•° `gdbm_close' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。Hovelã¯é€šå¸¸ã¯Depotã®ãƒ©ãƒƒãƒ‘ーã¨ã—ã¦å‹•ä½œã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’扱ã†ãŒã€ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ãéš›ã«é–¢æ•° `gdbm_open2' を用ã„ã‚‹ã“ã¨ã«ã‚ˆã£ã¦Curiaã®ãƒ©ãƒƒãƒ‘ーã¨ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’扱ã†ã‚ˆã†ã«ã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<h3>API</h3>
+
+<p>データã®æ ¼ç´ã€å‰Šé™¤ã€æ¤œç´¢ã«ç”¨ã„る関数ã¨ã®ãƒ‡ãƒ¼ã‚¿ã®æŽˆå—ã«ã¯ã€ã‚ーã¨å€¤ã‚’表ç¾ã™ã‚‹ã®ã« `datum' åž‹ã®æ§‹é€ 体を用ã„る。</p>
+
+<dl>
+<dt><kbd>typedef struct { char *dptr; size_t dsize; } datum;</kbd></dt>
+<dd>`dptr' ã¯ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。`dsize' ã¯ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã§ã‚る。</dd>
+</dl>
+
+<p>外部変数 `gdbm_version' ã¯ãƒãƒ¼ã‚¸ãƒ§ãƒ³æƒ…å ±ã®æ–‡å—列ã§ã‚る。</p>
+
+<dl>
+<dt><kbd>extern char *gdbm_version;</kbd></dt>
+<dd>ã“ã®å¤‰æ•°ã®æŒ‡ã™é ˜åŸŸã¯æ›¸ãè¾¼ã¿ç¦æ¢ã§ã‚る。</dd>
+</dl>
+
+<p>外部変数 `gdbm_errno' ã«ã¯ç›´å‰ã®ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚エラーコードã®è©³ç´°ã«ã¤ã„ã¦ã¯ `hovel.h' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚</p>
+
+<dl>
+<dt><kbd>extern gdbm_error gdbm_errno;</kbd></dt>
+<dd>ã“ã®å¤‰æ•°ã®åˆæœŸå€¤ã¯ `GDBM_NO_ERROR' ã§ã‚る。ãã®ä»–ã®å€¤ã¨ã—ã¦ã€`GDBM_MALLOC_ERROR'ã€`GDBM_BLOCK_SIZE_ERROR'ã€`GDBM_FILE_OPEN_ERROR'ã€`GDBM_FILE_WRITE_ERROR'ã€`GDBM_FILE_SEEK_ERROR'ã€`GDBM_FILE_READ_ERROR'ã€`GDBM_BAD_MAGIC_NUMBER'ã€`GDBM_EMPTY_DATABASE'ã€`GDBM_CANT_BE_READER'ã€`GDBM_CANT_BE_WRITER'ã€`GDBM_READER_CANT_DELETE'ã€`GDBM_READER_CANT_STORE'ã€`GDBM_READER_CANT_REORGANIZE'ã€`GDBM_UNKNOWN_UPDATE'ã€`GDBM_ITEM_NOT_FOUND'ã€`GDBM_REORGANIZE_FAILED'ã€`GDBM_CANNOT_REPLACE'ã€`GDBM_ILLEGAL_DATA'ã€`GDBM_OPT_ALREADY_SET' ãŠã‚ˆã³ `GDBM_OPT_ILLEGAL' ãŒã‚る。</dd>
+</dl>
+
+<p>エラーコードã«å¯¾å¿œã™ã‚‹ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸æ–‡å—列を得るã«ã¯ã€é–¢æ•° `gdbm_strerror' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *gdbm_strerror(gdbm_error <var>gdbmerrno</var>);</kbd></dt>
+<dd>`gdbmerrno' ã¯ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã‚¨ãƒ©ãƒ¼ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸ã®æ–‡å—列ã§ã‚ã‚Šã€ãã®é ˜åŸŸã¯æ›¸ãè¾¼ã¿ç¦æ¢é ˜åŸŸã§ã‚る。</dd>
+</dl>
+
+<p>GDBMæµã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_open' を用ã„る。</p>
+
+<dl>
+<dt><kbd>GDBM_FILE gdbm_open(char *<var>name</var>, int <var>block_size</var>, int <var>read_write</var>, int <var>mode</var>, void (*<var>fatal_func</var>)(void));</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®åå‰ã‚’指定ã™ã‚‹ã€‚`block_size' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚`read_write' ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã€`GDBM_READER' ãªã‚‰ãƒªãƒ¼ãƒ€ã€`GDBM_WRITER' 㨠`GDBM_WRCREAT' 㨠`GDBM_NEWDB' ãªã‚‰ãƒ©ã‚¤ã‚¿ã¨ãªã‚‹ã€‚`GDBM_WRCREAT' ã®å ´åˆã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒå˜åœ¨ã—ãªã‘ã‚Œã°ä½œæˆã—ã€`GDBM_NEWDB' ã®å ´åˆã¯æ—¢ã«å˜åœ¨ã—ã¦ã„ã¦ã‚‚æ–°ã—ã„データベースを作æˆã™ã‚‹ã€‚ライタã«å¯¾ã—ã¦ã¯ã€`GDBM_SYNC' ã‹ `GDBM_NOLOCK' ã‹ `GDBM_LOCKNB' ã‹ `GDBM_FAST' ã‹ `GDBM_SPARSE' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãる。`GDBM_SYNC' ã¯å…¨ã¦ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹æ“作をディスクã¨åŒæœŸã•ã›ã€`GDBM_NOLOCK' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã‚’ä¼´ã‚ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã€`GDBM_LOCKNB' ã¯ãƒ–ãƒãƒƒã‚¯ãªã—ã®ãƒãƒƒã‚¯ã‚’è¡Œã„ã€`GDBM_FAST' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚`GDBM_SPARSE' ã¯QDBM独自ã®ã‚‚ã®ã§ã‚ã‚Šã€ä½œæˆã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’スパースã«ã™ã‚‹ã€‚`mode' 㯠`open' コールã«æ¸¡ã™ã‚‚ã®ã¨
åŒã˜ã§ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ¢ãƒ¼ãƒ‰ã‚’指定ã™ã‚‹ã€‚`fatal_func' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>QDBMæµã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_open2' を用ã„る。</p>
+
+<dl>
+<dt><kbd>GDBM_FILE gdbm_open2(char *<var>name</var>, int <var>read_write</var>, int <var>mode</var>, int <var>bnum</var>, int <var>dnum</var>, int <var>align</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®åå‰ã‚’指定ã™ã‚‹ã€‚`read_write' ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã€`GDBM_READER' ãªã‚‰ãƒªãƒ¼ãƒ€ã€`GDBM_WRITER' 㨠`GDBM_WRCREAT' 㨠`GDBM_NEWDB' ãªã‚‰ãƒ©ã‚¤ã‚¿ã¨ãªã‚‹ã€‚`GDBM_WRCREAT' ã®å ´åˆã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒå˜åœ¨ã—ãªã‘ã‚Œã°ä½œæˆã—ã€`GDBM_NEWDB' ã®å ´åˆã¯æ—¢ã«å˜åœ¨ã—ã¦ã„ã¦ã‚‚æ–°ã—ã„データベースを作æˆã™ã‚‹ã€‚ライタã«å¯¾ã—ã¦ã¯ã€`GDBM_SYNC' ã‹ `GDBM_NOLOCK' ã‹ `GDBM_LOCKNB' ã‹ `GDBM_FAST' ã‹ `GDBM_SPARSE' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãる。`GDBM_SYNC' ã¯å…¨ã¦ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹æ“作をディスクã¨åŒæœŸã•ã›ã€`GDBM_NOLOCK' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã‚’ä¼´ã‚ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã€`GDBM_LOCKNB' ã¯ãƒ–ãƒãƒƒã‚¯ãªã—ã®ãƒãƒƒã‚¯ã‚’è¡Œã„ã€`GDBM_FAST' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚`GDBM_SPARSE' ã¯QDBM独自ã®ã‚‚ã®ã§ã‚ã‚Šã€ä½œæˆã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’スパースã«ã™ã‚‹ã€‚`mode' 㯠`open' コールもã—ã㯠`mkdir' コールã«æ¸¡ã™ã‚‚ã®ã¨åŒã
˜ã§ãƒ•ã‚¡ã‚¤ãƒ«ã‚„ディレクトリã®ãƒ¢ãƒ¼ãƒ‰ã‚’指定ã™ã‚‹ã€‚`bnum' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®ç›®å®‰ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。`dnum' ã¯è¦ç´ データベースã®æ•°ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰è¿”ã•ã‚Œã‚‹ãƒãƒ³ãƒ‰ãƒ«ã¯Depotã®ãƒ©ãƒƒãƒ‘ーã¨ã—ã¦ç”Ÿæˆã•ã‚Œã€ãã†ã§ãªã‘ã‚Œã°Curiaã®ãƒ©ãƒƒãƒ‘ーã«ãªã‚‹ã€‚`align' ã¯ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®åŸºæœ¬ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。既ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒå˜åœ¨ã™ã‚‹å ´åˆã€ãã‚ŒãŒDepotã®ã‚‚ã®ã‹Curiaã®ã‚‚ã®ã‹ãŒè‡ªå‹•çš„ã«åˆ¤æ–ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースã¨ã®æŽ¥ç¶šã‚’é–‰ã˜ã¦ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_close' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void gdbm_close(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_store' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int gdbm_store(GDBM_FILE <var>dbf</var>, datum <var>key</var>, datum <var>content</var>, int <var>flag</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚`content' ã¯å€¤ã®æ§‹é€ 体を指定ã™ã‚‹ã€‚`frags' ㌠`GDBM_INSERT' ãªã‚‰ã‚ーã®é‡è¤‡æ™‚ã«æ›¸ãè¾¼ã¿ã‚’æ–念ã—ã€`GDBM_REPLACE' ãªã‚‰ä¸Šæ›¸ãã‚’è¡Œã†ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ 0 ã€é‡è¤‡ã§ã®æ–念ãªã‚‰ 1 ã€ãã®ä»–ã®ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>レコードを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_delete' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int gdbm_delete(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ 0 ã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>レコードをå–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_fetch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>datum gdbm_fetch(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚戻り値ã¯å€¤ã®æ§‹é€ 体ã§ã‚る。該当ãŒã‚ã‚Œã°ãƒ¡ãƒ³ãƒ `dptr' ãŒãã®é ˜åŸŸã‚’指ã—ã€`dsize' ãŒãã®ã‚µã‚¤ã‚ºã‚’示ã™ã€‚該当ãŒãªã‘れ㰠`dptr' ã®å€¤ã¯ `NULL' ã¨ãªã‚‹ã€‚戻り値ã®ãƒ¡ãƒ³ãƒ `dptr' ã®æŒ‡ã™é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>レコードãŒå˜åœ¨ã™ã‚‹ã‹èª¿ã¹ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_exists' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int gdbm_exists(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ã‚ーã®æ§‹é€ 体を指定ã™ã‚‹ã€‚戻り値ã¯è©²å½“ãŒã‚ã‚Œã°çœŸã§ã‚ã‚Šã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯å½ã§ã‚る。</dd>
+</dl>
+
+<p>最åˆã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーを得るã«ã¯ã€é–¢æ•° `gdbm_firstkey' を用ã„る。</p>
+
+<dl>
+<dt><kbd>datum gdbm_firstkey(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã‚ーã®æ§‹é€ 体ã§ã‚る。該当ãŒã‚ã‚Œã°ãƒ¡ãƒ³ãƒ `dptr' ãŒãã®é ˜åŸŸã‚’指ã—ã€`dsize' ãŒãã®ã‚µã‚¤ã‚ºã‚’示ã™ã€‚該当ãŒãªã‘れ㰠`dptr' ã®å€¤ã¯ `NULL' ã¨ãªã‚‹ã€‚戻り値ã®ãƒ¡ãƒ³ãƒ `dptr' ã®æŒ‡ã™é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>次ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーを得るã«ã¯ã€é–¢æ•° gdbm_nextkey を用ã„る。</p>
+
+<dl>
+<dt><kbd>datum gdbm_nextkey(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`key' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚戻り値ã¯ã‚ーã®æ§‹é€ 体ã§ã‚る。該当ãŒã‚ã‚Œã°ãƒ¡ãƒ³ãƒ `dptr' ãŒãã®é ˜åŸŸã‚’指ã—ã€`dsize' ãŒãã®ã‚µã‚¤ã‚ºã‚’示ã™ã€‚該当ãŒãªã‘れ㰠`dptr' ã®å€¤ã¯ `NULL' ã¨ãªã‚‹ã€‚戻り値ã®ãƒ¡ãƒ³ãƒ `dptr' ã®æŒ‡ã™é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>データベースを更新ã—ãŸå†…容をファイルã¨ãƒ‡ãƒã‚¤ã‚¹ã«åŒæœŸã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_sync' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void gdbm_sync(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースを最é©åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_reorganize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int gdbm_reorganize(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ 0 ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `gdbm_fdesc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int gdbm_fdesc(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—ã‚¿ã§ã‚る。データベースãŒãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãªã‚‰ã€æˆ»ã‚Šå€¤ã¯ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>関数 `gdbm_setopt' ã¯ä½•ã‚‚ã—ãªã„。</p>
+
+<dl>
+<dt><kbd>int gdbm_setopt(GDBM_FILE <var>dbf</var>, int <var>option</var>, int *<var>value</var>, int <var>size</var>);</kbd></dt>
+<dd>`dbf' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`option' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚`value' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚`size' ã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚戻り値㯠0 ã§ã‚る。ã“ã®é–¢æ•°ã¯äº’æ›æ€§ã®ãŸã‚ã«ã®ã¿å˜åœ¨ã™ã‚‹ã€‚</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>åå‰ã¨å¯¾å¿œã•ã›ã¦é›»è©±ç•ªå·ã‚’æ ¼ç´ã—ã€ãれを検索ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <hovel.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <stdio.h>
+#include <string.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ GDBM_FILE dbf;
+ datum key, val;
+ int i;
+
+ /* データベースを開ã */
+ if(!(dbf = gdbm_open(DBNAME, 0, GDBM_WRCREAT, 00644, NULL))){
+ fprintf(stderr, "gdbm_open: %s\n", gdbm_strerror(gdbm_errno));
+ return 1;
+ }
+
+ /* レコードを準備ã™ã‚‹ */
+ key.dptr = NAME;
+ key.dsize = strlen(NAME);
+ val.dptr = NUMBER;
+ val.dsize = strlen(NUMBER);
+
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ */
+ if(gdbm_store(dbf, key, val, GDBM_REPLACE) != 0){
+ fprintf(stderr, "gdbm_store: %s\n", gdbm_strerror(gdbm_errno));
+ }
+
+ /* レコードを検索ã™ã‚‹ */
+ val = gdbm_fetch(dbf, key);
+ if(val.dptr){
+ printf("Name: %s\n", NAME);
+ printf("Number: ");
+ for(i = 0; i < val.dsize; i++){
+ putchar(val.dptr[i]);
+ }
+ putchar('\n');
+ free(val.dptr);
+ } else {
+ fprintf(stderr, "gdbm_fetch: %s\n", gdbm_strerror(gdbm_errno));
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ gdbm_close(dbf);
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Hovelを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹æ–¹æ³•ã¯ã€Depotã®å ´åˆã¨å…¨ãåŒã˜ã§ã‚る。リンカã«æ¸¡ã™ã‚ªãƒ—ション㯠`-lgdbm' ã§ã¯ãªã `-lqdbm' ã§ã‚る。</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>POSIXスレッドを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã€å¤–部変数 `gdbm_errno' ã¯ã‚¹ãƒ¬ãƒƒãƒ‰å›ºæœ‰ãƒ‡ãƒ¼ã‚¿ã¸ã®å‚ç…§ã¨ã—ã¦æ‰±ã‚ã‚Œã€Hovelã®å„関数ã¯ãƒªã‚¨ãƒ³ãƒˆãƒ©ãƒ³ãƒˆã«ãªã‚‹ã€‚ãã®å ´åˆã€ã‚¹ãƒ¬ãƒƒãƒ‰é–“ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="hovelcli" id="hovelcli" class="head">Hovel用コマンド</a></h2>
+
+<p>Hovelã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `hvmgr' ã¯Hovelã‚„ãã®ã‚¢ãƒ—リケーションã®ãƒ‡ãƒãƒƒã‚°ã«å½¹ç«‹ã¤ãƒ„ールã§ã‚る。データベースを更新ã—ãŸã‚Šã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’調ã¹ãŸã‚Šã™æ©Ÿèƒ½ã‚’æŒã¤ã€‚シェルスクリプトã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚¢ãƒ—リケーションを作るã®ã«ã‚‚利用ã§ãる。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`key' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã€`val' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>hvmgr [-qdbm bnum dnum] [-s] create <var>name</var></kbd></dt>
+<dd>データベースファイルを作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>hvmgr store [-qdbm] [-kx] [-vx|-vf] [-insert] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>ã‚ーã¨å€¤ã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>hvmgr delete [-qdbm] [-kx] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>hvmgr fetch [-qdbm] [-kx] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>hvmgr list [-qdbm] [-ox] <var>name</var></kbd></dt>
+<dd>データベース内ã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨å€¤ã‚’タブã¨æ”¹è¡Œã§åŒºåˆ‡ã£ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>hvmgr optimize [-qdbm] <var>name</var></kbd></dt>
+<dd>データベースを最é©åŒ–ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-qdbm [<var>bnum</var> <var>dnum</var>]</kbd> : `gdbm_open2' ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã。`bnum' 㨠`dnum' ã¯ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®åˆ†å‰²æ•°ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-s</kbd> : ファイルをスパースã«ã™ã‚‹ã€‚</li>
+<li><kbd>-kx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-vx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vf</kbd> : åå‰ãŒ `val' ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ‡ãƒ¼ã‚¿ã‚’値ã¨ã—ã¦èªã¿è¾¼ã‚€ã€‚</li>
+<li><kbd>-insert</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«ä¸Šæ›¸ãã›ãšã«ã‚¨ãƒ©ãƒ¼ã«ã™ã‚‹ã€‚</li>
+<li><kbd>-ox</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—ã¦æ¨™æº–出力を行ã†ã€‚</li>
+<li><kbd>-n</kbd> : 標準出力ã®æœ«å°¾ã«ä»˜åŠ ã•ã‚Œã‚‹æ”¹è¡Œæ–‡å—ã®å‡ºåŠ›ã‚’抑制ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>コマンド `hvtest' ã¯Hovelã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã«ã‚ˆã£ã¦ç”Ÿæˆã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ `hvmgr' ã«ã‚ˆã£ã¦è§£æžã—ãŸã‚Šã€`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計ã£ãŸã‚Šã™ã‚‹ã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`rnum' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>hvtest write [-qdbm] [-s] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>`00000001'ã€`00000002' ã®ã‚ˆã†ã«å¤‰åŒ–ã™ã‚‹8ãƒã‚¤ãƒˆã®ã‚ーã¨é©å½“ãª8ãƒã‚¤ãƒˆã®å€¤ã‚’連続ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>hvtest read [-qdbm] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>上記ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’検索ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-qdbm</kbd> : `gdbm_open2' を用ã„ã¦Curiaã®ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ã。</li>
+<li><kbd>-s</kbd> : ファイルをスパースã«ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="cabinapi" id="cabinapi" class="head">Cabin: ユーティリティAPI</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Cabinã¯ãƒ¡ãƒ¢ãƒªä¸Šã§ç°¡å˜ã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’扱ã†ãŸã‚ã®ãƒ¡ãƒ¢ãƒªç¢ºä¿é–¢æ•°ã‚„整列関数や拡張å¯èƒ½ãªãƒ‡ãƒ¼ã‚¿ã‚„é…列リストやãƒãƒƒã‚·ãƒ¥ãƒžãƒƒãƒ—やヒープé…列ãªã©æä¾›ã™ã‚‹ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã®APIã§ã‚る。MIMEã‚„CSVã‚„XMLを解æžã™ã‚‹æ©Ÿèƒ½ã‚„ã€å„種ã®ç¬¦å·åŒ–ã¨å¾©å·ã‚’è¡Œã†æ©Ÿèƒ½ã‚‚å‚™ãˆã‚‹ã€‚</p>
+
+<p>Cabinを使ã†ãŸã‚ã«ã¯ã€`cabin.h' 㨠`stdlib.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <cabin.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>æ‹¡å¼µå¯èƒ½ãªãƒ‡ãƒ¼ã‚¿ã‚’扱ã†éš›ã«ã¯ã€`CBDATUM' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。データãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `cbdatumopen' ã§é–‹ãã€é–¢æ•° `cbdatumclose' ã§é–‰ã˜ã‚‹ã€‚リストを扱ã†éš›ã«ã¯ã€`CBLIST' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。リストãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `cblistopen' ã§é–‹ãã€é–¢æ•° `cblistclose' ã§é–‰ã˜ã‚‹ã€‚マップを扱ã†éš›ã«ã¯ã€`CBMAP' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。マップãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `cbmapopen' ã§é–‹ãã€é–¢æ•° `cbmapclose' ã§é–‰ã˜ã‚‹ã€‚ヒープé…列を扱ã†éš›ã«ã¯ `CBHEAP' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ヒープãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cbheapopen' ã§é–‹ãã€é–¢æ•° `cbheapclose' ã§é–‰ã˜ã‚‹ã€‚å„ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。
+
+</p>
+
+<h3>API</h3>
+
+<p>外部変数 `cbfatalfunc' ã¯è‡´å‘½çš„エラーをãƒãƒ³ãƒ‰ãƒªãƒ³ã‚°ã™ã‚‹ã‚³ãƒ¼ãƒ«ãƒãƒƒã‚¯é–¢æ•°ã§ã‚る。</p>
+
+<dl>
+<dt><kbd>extern void (*cbfatalfunc)(const char *<var>message</var>);</kbd></dt>
+<dd>引数ã¯ã‚¨ãƒ©ãƒ¼ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸ã‚’指定ã™ã‚‹ã€‚ã“ã®å¤‰æ•°ã®åˆæœŸå€¤ã¯ `NULL' ã§ã‚ã‚Šã€`NULL' ãªã‚‰ã°è‡´å‘½çš„エラーã®ç™ºç”Ÿæ™‚ã«ã¯ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã®é–¢æ•°ãŒå‘¼ã°ã‚Œã‚‹ã€‚致命的エラーã¯ãƒ¡ãƒ¢ãƒªã®å‰²ã‚Šå½“ã¦ã«å¤±æ•—ã—ãŸéš›ã«èµ·ã“る。</dd>
+</dl>
+
+<p>メモリ上ã«é ˜åŸŸã‚’確ä¿ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmalloc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void *cbmalloc(size_t <var>size</var>);</kbd></dt>
+<dd>`size' ã¯é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯ç¢ºä¿ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>メモリ上ã®é ˜åŸŸã‚’å†ç¢ºä¿ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbrealloc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void *cbrealloc(void *<var>ptr</var>, size_t <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯å†ç¢ºä¿ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `remalloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>メモリ上ã®é ˜åŸŸã‚’複製ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmemdup' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbmemdup(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯å†ç¢ºä¿ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã¯è¤‡è£½ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>メモリ上ã®é–‹æ”¾ã‚’解放ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbfree' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbfree(void *<var>ptr</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ãŒã€`NULL' ã®å ´åˆã¯ä½•ã‚‚ã—ãªã„。ã“ã®é–¢æ•°ã¯ `free' ã®ãƒ©ãƒƒãƒ‘ーã«éŽãŽãªã„ãŒã€`malloc' シリーズã®åˆ¥ã®ãƒ‘ッケージを使ã†ã‚¢ãƒ—リケーションã«ãŠã„ã¦QDBMãŒç¢ºä¿ã—ãŸé ˜åŸŸã‚’解放ã™ã‚‹ã®ã«ä¾¿åˆ©ã§ã‚る。</dd>
+</dl>
+
+<p>オブジェクトã®ãƒã‚¤ãƒ³ã‚¿ã‹ãƒãƒ³ãƒ‰ãƒ«ã‚’ã‚°ãƒãƒ¼ãƒãƒ«ã‚¬ãƒ™ãƒ¼ã‚¸ã‚³ãƒ¬ã‚¯ã‚¿ã«ç™»éŒ²ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbglobalgc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbglobalgc(void *<var>ptr</var>, void (*<var>func</var>)(void *));</kbd></dt>
+<dd>`ptr' ã¯ã‚ªãƒ–ジェクトã®ãƒã‚¤ãƒ³ã‚¿ã‹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`func' ã¯ã‚ªãƒ–ジェクトã®ãƒªã‚½ãƒ¼ã‚¹ã‚’解放ã™ã‚‹é–¢æ•°ã‚’指定ã™ã‚‹ã€‚ãã®å¼•æ•°ã¯è§£æ”¾ã™ã‚‹ã‚ªãƒ–ジェクトã®ãƒã‚¤ãƒ³ã‚¿ã‹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。ã“ã®é–¢æ•°ã¯ã€`main' 関数ãŒãƒªã‚¿ãƒ¼ãƒ³ã™ã‚‹ã‹ `exit' 関数ãŒå‘¼ã°ã‚Œã¦ãƒ—ãƒã‚»ã‚¹ãŒæ£å¸¸çµ‚了ã™ã‚‹éš›ã«ã€ã‚ªãƒ–ジェクトã®ãƒªã‚½ãƒ¼ã‚¹ãŒè§£æ”¾ã•ã‚Œã‚‹ã“ã¨ã‚’ä¿è¨¼ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ã‚°ãƒãƒ¼ãƒãƒ«ã‚¬ãƒ™ãƒ¼ã‚¸ã‚³ãƒ¬ã‚¯ã‚¿ã‚’明示的ã«ç™ºå‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cbggcsweep' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbggcsweep(void);</kbd></dt>
+<dd>ã“ã®é–¢æ•°ã‚’呼んã 後ã¯ã‚°ãƒãƒ¼ãƒãƒ«ã‚¬ãƒ™ãƒ¼ã‚¸ã‚³ãƒ¬ã‚¯ã‚¿ã«ç™»éŒ²ã—ã¦ã‚ã£ãŸã‚ªãƒ–ジェクトã¯åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ããªããªã‚‹ã“ã¨ã«æ³¨æ„ã™ã‚‹ã“ã¨ã€‚ã‚°ãƒãƒ¼ãƒãƒ«ã‚¬ãƒ™ãƒ¼ã‚¸ã‚³ãƒ¬ã‚¯ã‚¿ã¯åˆæœŸåŒ–ã•ã‚Œã‚‹ã®ã§ã€æ–°ã—ã„オブジェクトを入れるã“ã¨ãŒã§ãるよã†ã«ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>仮想メモリã®å‰²ã‚Šå½“ã¦å¯èƒ½æ€§ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `cbvmemavail' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbvmemavail(size_t <var>size</var>);</kbd></dt>
+<dd>`size' ã¯æ–°ãŸã«å‰²ã‚Šå½“ã¦å¯èƒ½ã§ã‚ã‚‹ã¹ãé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯å‰²ã‚Šå½“ã¦ãŒå¯èƒ½ã§ã‚ã‚Œã°çœŸã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>é…列ã®å„è¦ç´ を挿入ソートã§æ•´åˆ—ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cbisort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbisort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' ã¯é…列ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`nmemb' ã¯é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`size' ã¯å„è¦ç´ ã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`compar' ã¯æ¯”較関数を指定ã™ã‚‹ã€‚二ã¤ã®å¼•æ•°ã¯è¦ç´ ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。比較関数ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£æ•°ã‚’ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² æ•°ã‚’ã€ä¸¡è€…ãŒç‰ã—ã‘れ㰠0 ã‚’è¿”ã™ã¹ãã§ã‚る。挿入ソートã¯ã€ã»ã¨ã‚“ã©ã®è¦ç´ ãŒæ—¢ã«æ•´åˆ—済ã¿ã®å ´åˆã«ã®ã¿æœ‰ç”¨ã§ã‚る。</dd>
+</dl>
+
+<p>é…列ã®å„è¦ç´ をシェルソートã§æ•´åˆ—ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cbssort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbssort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' ã¯é…列ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`nmemb' ã¯é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`size' ã¯å„è¦ç´ ã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`compar' ã¯æ¯”較関数を指定ã™ã‚‹ã€‚二ã¤ã®å¼•æ•°ã¯è¦ç´ ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。比較関数ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£æ•°ã‚’ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² æ•°ã‚’ã€ä¸¡è€…ãŒç‰ã—ã‘れ㰠0 ã‚’è¿”ã™ã¹ãã§ã‚る。ã»ã¨ã‚“ã©ã®è¦ç´ ãŒæ•´åˆ—済ã¿ã®å ´åˆã€ã‚·ã‚§ãƒ«ã‚½ãƒ¼ãƒˆã®æ–¹ãŒãƒ’ープソートやクイックソートより速ã„ã‹ã‚‚ã—ã‚Œãªã„。</dd>
+</dl>
+
+<p>é…列ã®å„è¦ç´ をヒープソートã§æ•´åˆ—ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cbhsort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbhsort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' ã¯é…列ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`nmemb' ã¯é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`size' ã¯å„è¦ç´ ã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`compar' ã¯æ¯”較関数を指定ã™ã‚‹ã€‚二ã¤ã®å¼•æ•°ã¯è¦ç´ ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。比較関数ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£æ•°ã‚’ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² æ•°ã‚’ã€ä¸¡è€…ãŒç‰ã—ã‘れ㰠0 ã‚’è¿”ã™ã¹ãã§ã‚る。ヒープソートã¯å…¥åŠ›ã®åã‚Šã«å¯¾ã—ã¦é ‘丈ã§ã‚ã‚‹ãŒã€ã»ã¨ã‚“ã©ã®å ´åˆã§ã‚¯ã‚¤ãƒƒã‚¯ã‚½ãƒ¼ãƒˆã®æ–¹ãŒé€Ÿã„。</dd>
+</dl>
+
+<p>é…列ã®å„è¦ç´ をクイックソートã§æ•´åˆ—ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cbqsort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbqsort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' ã¯é…列ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`nmemb' ã¯é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`size' ã¯å„è¦ç´ ã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`compar' ã¯æ¯”較関数を指定ã™ã‚‹ã€‚二ã¤ã®å¼•æ•°ã¯è¦ç´ ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。比較関数ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£æ•°ã‚’ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² æ•°ã‚’ã€ä¸¡è€…ãŒç‰ã—ã‘れ㰠0 ã‚’è¿”ã™ã¹ãã§ã‚る。入力ã®åã‚Šã«æ•æ„Ÿã§ã¯ã‚ã‚‹ãŒã€ã‚¯ã‚¤ãƒƒã‚¯ã‚½ãƒ¼ãƒˆã¯æœ€é€Ÿã®æ•´åˆ—アルゴリズムã§ã‚る。</dd>
+</dl>
+
+<p>大文å—ã¨å°æ–‡å—ã®é•ã„を無視ã—ã¦äºŒã¤ã®æ–‡å—列を比較ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbstricmp' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbstricmp(const char *<var>astr</var>, const char *<var>bstr</var>);</kbd></dt>
+<dd>`astr' ã¯ä¸€æ–¹ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`bstr' ã¯ä»–æ–¹ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² ã€ä¸¡è€…ãŒç‰ä¾¡ãªã‚‰ 0 ã§ã‚る。ASCIIコードä¸ã®ã‚¢ãƒ«ãƒ•ã‚¡ãƒ™ãƒƒãƒˆã®å¤§æ–‡å—ã¨å°æ–‡å—ã¯åŒºåˆ¥ã•ã‚Œãªã„。</dd>
+</dl>
+
+<p>æ–‡å—列ãŒã‚ã‚‹ã‚ーã§å§‹ã¾ã£ã¦ã„ã‚‹ã‹èª¿ã¹ã‚‹ã«ã¯ã€é–¢æ•° `cbstrfwmatch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbstrfwmatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`key' ã¯å‰æ–¹ä¸€è‡´ã®ã‚ーã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯å¯¾è±¡ã®æ–‡å—列ãŒã‚ーã§å§‹ã¾ã£ã¦ã„ã‚Œã°çœŸã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>大文å—ã¨å°æ–‡å—ã®é•ã„を無視ã—ã¤ã¤ã€æ–‡å—列ãŒã‚ã‚‹ã‚ーã§å§‹ã¾ã£ã¦ã„ã‚‹ã‹èª¿ã¹ã‚‹ã«ã¯ã€é–¢æ•° `cbstrfwimatch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbstrfwimatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`key' ã¯å‰æ–¹ä¸€è‡´ã®ã‚ーã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯å¯¾è±¡ã®æ–‡å—列ãŒã‚ーã§å§‹ã¾ã£ã¦ã„ã‚Œã°çœŸã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。ASCIIコードä¸ã®ã‚¢ãƒ«ãƒ•ã‚¡ãƒ™ãƒƒãƒˆã®å¤§æ–‡å—ã¨å°æ–‡å—ã¯åŒºåˆ¥ã•ã‚Œãªã„。</dd>
+</dl>
+
+<p>æ–‡å—列ãŒã‚ã‚‹ã‚ーã§çµ‚ã£ã¦ã„ã‚‹ã‹èª¿ã¹ã‚‹ã«ã¯ã€é–¢æ•° `cbstrbwmatch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbstrbwmatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`key' ã¯å¾Œæ–¹ä¸€è‡´ã®ã‚ーã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯å¯¾è±¡ã®æ–‡å—列ãŒã‚ーã§çµ‚ã£ã¦ã„ã‚Œã°çœŸã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>大文å—ã¨å°æ–‡å—ã®é•ã„を無視ã—ã¤ã¤ã€æ–‡å—列ãŒã‚ã‚‹ã‚ーã§çµ‚ã£ã¦ã„ã‚‹ã‹èª¿ã¹ã‚‹ã«ã¯ã€é–¢æ•° `cbstrbwimatch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbstrbwimatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`key' ã¯å¾Œæ–¹ä¸€è‡´ã®ã‚ーã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯å¯¾è±¡ã®æ–‡å—列ãŒã‚ーã§çµ‚ã£ã¦ã„ã‚Œã°çœŸã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。ASCIIコードä¸ã®ã‚¢ãƒ«ãƒ•ã‚¡ãƒ™ãƒƒãƒˆã®å¤§æ–‡å—ã¨å°æ–‡å—ã¯åŒºåˆ¥ã•ã‚Œãªã„。</dd>
+</dl>
+
+<p>KMP法を用ã„ã¦æ–‡å—列ã®éƒ¨åˆ†æ–‡å—列ã®ä½ç½®ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbstrstrkmp' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrstrkmp(const char *<var>haystack</var>, const char *<var>needle</var>);</kbd></dt>
+<dd>`haystack' ã¯æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`needle' ã¯æŽ¢ã™ã¹ã部分文å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯éƒ¨åˆ†æ–‡å—列ã®é–‹å§‹ã‚’指ã™ãƒã‚¤ãƒ³ã‚¿ã‹ã€è¦‹ã¤ã‹ã‚‰ãªã‘れ㰠`NULL' ã§ã‚る。大抵ã®å ´åˆã€ã“ã®é–¢æ•°ã‚ˆã‚Šã‚³ãƒ³ãƒ‘イラã®ãƒ“ルドインã§ã‚ã‚‹ `strstr' ã®æ–¹ãŒé«˜é€Ÿã§ã‚る。</dd>
+</dl>
+
+<p>BM法を用ã„ã¦æ–‡å—列ã®éƒ¨åˆ†æ–‡å—列ã®ä½ç½®ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbstrstrbm' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrstrbm(const char *<var>haystack</var>, const char *<var>needle</var>);</kbd></dt>
+<dd>`haystack' ã¯æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`needle' ã¯æŽ¢ã™ã¹ã部分文å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯éƒ¨åˆ†æ–‡å—列ã®é–‹å§‹ã‚’指ã™ãƒã‚¤ãƒ³ã‚¿ã‹ã€è¦‹ã¤ã‹ã‚‰ãªã‘れ㰠`NULL' ã§ã‚る。大抵ã®å ´åˆã€ã“ã®é–¢æ•°ã‚ˆã‚Šã‚³ãƒ³ãƒ‘イラã®ãƒ“ルドインã§ã‚ã‚‹ `strstr' ã®æ–¹ãŒé«˜é€Ÿã§ã‚る。</dd>
+</dl>
+
+<p>æ–‡å—列ã®å…¨ã¦ã®æ–‡å—を大文å—ã«å¤‰æ›ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbstrtoupper' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrtoupper(char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¤‰æ›å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<p>æ–‡å—列ã®å…¨ã¦ã®æ–‡å—ã‚’å°æ–‡å—ã«å¤‰æ›ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbstrtolower' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrtolower(char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¤‰æ›å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<p>æ–‡å—列ã®å…ˆé ã¨æœ«å°¾ã«ã‚る空白文å—を削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbstrtrim' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrtrim(char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¤‰æ›å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<p>æ–‡å—列内ã®é€£ç¶šã™ã‚‹ç©ºç™½ã‚’絞ã£ã¦æ•´å½¢ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbstrsqzcpc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrsqzspc(char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¤‰æ›å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<p>UTF-8ã®æ–‡å—列ã«å«ã¾ã‚Œã‚‹æ–‡å—æ•°ã‚’æ•°ãˆã‚‹ã«ã¯ã€é–¢æ•° `cbstrcountutf' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbstrcountutf(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯UTF-8ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ–‡å—列ã«å«ã¾ã‚Œã‚‹æ–‡å—æ•°ã§ã‚る。</dd>
+</dl>
+
+<p>UTF-8ã®æ–‡å—列を指定ã—ãŸæ–‡å—æ•°ã§åˆ‡ã‚‹ã«ã¯ã€é–¢æ•° `cbstrcututf' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbstrcututf(char *<var>str</var>, int <var>num</var>);</kbd></dt>
+<dd>`str' ã¯UTF-8ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`num' ã¯ä¿æŒã™ã‚‹æ–‡å—数を指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。</dd>
+</dl>
+
+<p>データãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBDATUM *cbdatumopen(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯åˆæœŸå†…容ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã‹ã€`NULL' ãªã‚‰ç©ºã®ãƒ‡ãƒ¼ã‚¿ã‚’作æˆã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>データを複製ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumdup' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBDATUM *cbdatumdup(const CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–°ã—ã„データãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>データãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbdatumclose(CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>データã«åˆ¥ã®é ˜åŸŸã‚’連çµã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumcat' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbdatumcat(CBDATUM *<var>datum</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯é€£çµã™ã‚‹é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>データã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumptr' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *cbdatumptr(const CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>データã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumsize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbdatumsize(const CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã§ã‚る。</dd>
+</dl>
+
+<p>データã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’変更ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumsetsize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbdatumsetsize(CBDATUM *<var>datum</var>, int <var>size</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`size' ã¯é ˜åŸŸã®æ–°ã—ã„サイズを指定ã™ã‚‹ã€‚æ–°ã—ã„サイズãŒæ—¢å˜ã®ã‚µã‚¤ã‚ºã‚ˆã‚Šå¤§ãã„å ´åˆã€ä½™ã£ãŸé ˜åŸŸã¯çµ‚端文å—ã§åŸ‹ã‚られる。</dd>
+</dl>
+
+<p>データã«æ›¸å¼å‡ºåŠ›ã‚’è¡Œãªã†ã«ã¯ã€é–¢æ•° `cbdatumprintf' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbdatumprintf(CBDATUM *<var>datum</var>, const char *<var>format</var>, ...);</kbd></dt>
+<dd>`format' ã¯printf風ã®æ›¸å¼æ–‡å—列を指定ã™ã‚‹ã€‚変æ›æ–‡å— `%' ã‚’ `s'ã€`d'ã€`o'ã€`u'ã€`x'ã€`X'ã€`c'ã€`e'ã€`E'ã€`f'ã€`g'ã€`G'ã€`@'ã€`?'ã€`:'ã€`%' ã¨ä½µã›ã¦åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãる。`@' 㯠`s' ã¨åŒæ§˜ã«æ©Ÿèƒ½ã™ã‚‹ãŒã€XMLã®ãƒ¡ã‚¿æ–‡å—をエスケープã™ã‚‹ã€‚`?' 㯠`s' ã¨åŒæ§˜ã«æ©Ÿèƒ½ã™ã‚‹ãŒã€URLã®ãƒ¡ã‚¿æ–‡å—をエスケープã™ã‚‹ã€‚`:' 㯠`s' ã¨åŒæ§˜ã«æ©Ÿèƒ½ã™ã‚‹ãŒã€UTF-8ã¨ã—ã¦ã®MIMEエンコーディングを施ã™ã€‚ãれ以外ã®å¤‰æ›æ–‡å—ã¯å…ƒæ¥ã®ã‚‚ã®ã¨åŒæ§˜ã«æ©Ÿèƒ½ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>データを確ä¿ã•ã‚ŒãŸé ˜åŸŸã«å¤‰æ›ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdatumtomalloc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbdatumtomalloc(CBDATUM *<var>datum</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`datum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。元ã®ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ãれをå†ã³è§£æ”¾ã—ã¦ã¯ãªã‚‰ãªã„。</dd>
+</dl>
+
+<p>リストãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cblistopen(void);</kbd></dt>
+<dd>戻り値ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>リストを複製ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistdup' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cblistdup(const CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–°ã—ã„リストãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>リストãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cblistclose(CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>リストã«æ ¼ç´ã•ã‚ŒãŸè¦ç´ æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cblistnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cblistnum(const CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆã«æ ¼ç´ã•ã‚ŒãŸè¦ç´ æ•°ã§ã‚る。</dd>
+</dl>
+
+<p>リスト内ã®ã‚ã‚‹è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cblistval' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *cblistval(const CBLIST *<var>list</var>, int <var>index</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`index' ã¯å–り出ã™è¦ç´ ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯è©²å½“è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。`index' ãŒè¦ç´ 数以上ãªã‚‰ã°ã€æˆ»ã‚Šå€¤ã¯ `NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>è¦ç´ をリストã®æœ«å°¾ã«åŠ ãˆã‚‹ã«ã¯ã€é–¢æ•° `cblistpush' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cblistpush(CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯è¿½åŠ ã™ã‚‹è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>リストã®æœ«å°¾ã®è¦ç´ を削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistpop' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cblistpop(CBLIST *<var>list</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯è©²å½“è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。リストãŒç©ºãªã‚‰ã°ã€æˆ»ã‚Šå€¤ã¯ `NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>è¦ç´ をリストã®å…ˆé ã«åŠ ãˆã‚‹ã«ã¯ã€é–¢æ•° `cblistunshift' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cblistunshift(CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯è¿½åŠ ã™ã‚‹è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>リストã®å…ˆé ã®è¦ç´ を削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistshift' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cblistshift(CBLIST *<var>list</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯è©²å½“è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。リストãŒç©ºãªã‚‰ã°ã€æˆ»ã‚Šå€¤ã¯ `NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>リスト内ã®æŒ‡å®šã—ãŸä½ç½®ã«è¦ç´ ã‚’åŠ ãˆã‚‹ã«ã¯ã€é–¢æ•° `cblistinsert' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cblistinsert(CBLIST *<var>list</var>, int <var>index</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`index' ã¯è¿½åŠ ã™ã‚‹è¦ç´ ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯è¿½åŠ ã™ã‚‹è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>リスト内ã®æŒ‡å®šã—ãŸä½ç½®ã®è¦ç´ を削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistremove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cblistremove(CBLIST *<var>list</var>, int <var>index</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`index' ã¯å‰Šé™¤ã™ã‚‹è¦ç´ ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯è©²å½“è¦ç´ ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。`index' ãŒè¦ç´ 数以上ãªã‚‰ã°ã€è¦ç´ ã¯å‰Šé™¤ã•ã‚Œãšã€æˆ»ã‚Šå€¤ã¯ `NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>リスト内ã®æŒ‡å®šã—ãŸä½ç½®ã®è¦ç´ を上書ãã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistover' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cblistover(CBLIST *<var>list</var>, int <var>index</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`index' ã¯å‰Šé™¤ã™ã‚‹è¦ç´ ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯æ–°ã—ã„内容ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`index' ãŒè¦ç´ 数以上ãªã‚‰ã°ã€ã“ã®é–¢æ•°ã¯ä½•ã‚‚ã—ãªã„。</dd>
+</dl>
+
+<p>リストã®è¦ç´ ã‚’è¾žæ›¸é †ã§æ•´åˆ—ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cblistsort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cblistsort(CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚整列ã«ã¯ã‚¯ã‚¤ãƒƒã‚¯ã‚½ãƒ¼ãƒˆãŒç”¨ã„られる。</dd>
+</dl>
+
+<p>リストã®è¦ç´ を線形探索を使ã£ã¦æ¤œç´¢ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistlsearch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cblistlsearch(const CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯æ¤œç´¢ã‚ーã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“ã®è¦ç´ ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã§ã‚ã‚‹ãŒã€è©²å½“ãŒãªã‘れ㰠-1 ã§ã‚る。複数ã®è¦ç´ ãŒè©²å½“ã—ãŸå ´åˆã€å‰è€…ãŒè¿”ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>リストã®è¦ç´ を二分探索を使ã£ã¦æ¤œç´¢ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistbsearch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cblistbsearch(const CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚リストã¯è¾žæ›¸é †ã«ã‚½ãƒ¼ãƒˆã•ã‚Œã¦ã„ã‚‹å¿…è¦ãŒã‚る。`ptr' ã¯æ¤œç´¢ã‚ーã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“ã®è¦ç´ ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã§ã‚ã‚‹ãŒã€è©²å½“ãŒãªã‘れ㰠-1 ã§ã‚る。複数ã®è¦ç´ ãŒè©²å½“ã—ãŸå ´åˆã«ã©ã¡ã‚‰ãŒè¿”ã‚‹ã‹ã¯æœªå®šç¾©ã§ã‚る。</dd>
+</dl>
+
+<p>リストを直列化ã—ã¦ãƒã‚¤ãƒˆé…列ã«ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistdump' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cblistdump(const CBLIST *<var>list</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ã¯æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ç›´åˆ—化ã•ã‚ŒãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>直列化ã•ã‚ŒãŸãƒªã‚¹ãƒˆã‚’復元ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblistload' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cblistload(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯ãƒã‚¤ãƒˆé…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–°ã—ã„リストãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>マップãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBMAP *cbmapopen(void);</kbd></dt>
+<dd>戻り値ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>マップを複製ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapdup' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBMAP *cbmapdup(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–°ã—ã„マップãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。コピー元ã®ãƒžãƒƒãƒ—ã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã¯åˆæœŸåŒ–ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>マップãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbmapclose(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>マップã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapput' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbmapput(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>over</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`over' ã¯é‡è¤‡ã—ãŸãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’上書ãã™ã‚‹ã‹å¦ã‹ã‚’指定ã™ã‚‹ã€‚`over' ãŒå½ã§ã‚ーãŒé‡è¤‡ã—ãŸå ´åˆã¯æˆ»ã‚Šå€¤ã¯å½ã§ã‚ã‚‹ãŒã€ãã†ã§ãªã„å ´åˆã¯çœŸã§ã‚る。</dd>
+</dl>
+
+<p>æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã®æœ«å°¾ã«å€¤ã‚’連çµã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapputcat' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbmapputcat(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒå˜åœ¨ã—ãªã„å ´åˆã¯æ–°ã—ã„レコードãŒä½œã‚‰ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>マップã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapout' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbmapout(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯å½ã§ã‚る。</dd>
+</dl>
+
+<p>マップã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapget' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *cbmapget(const CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>レコードをマップã®ç«¯ã«ç§»å‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `cbmapmove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbmapmove(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>head</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`head' ã¯ç§»å‹•å…ˆã‚’指定ã—ã€çœŸãªã‚‰å…ˆé ã€å½ãªã‚‰æœ«å°¾ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯å½ã§ã‚る。</dd>
+</dl>
+
+<p>マップã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’åˆæœŸåŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapiterinit' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbmapiterinit(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚イテレータã¯ã€ãƒžãƒƒãƒ—ã«æ ¼ç´ã•ã‚ŒãŸå…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å‚ç…§ã™ã‚‹ãŸã‚ã«ç”¨ã„られる。</dd>
+</dl>
+
+<p>マップã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰æ¬¡ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーをå–り出ã™ã«ã¯ã€é–¢æ•° `cbmapiternext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *cbmapiternext(CBMAP *<var>map</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ã‚ãƒ¼ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。イテレータãŒæœ€å¾Œã¾ã§ãã¦è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。å–り出ã™é †ç•ªã¯æ ¼ç´ã—ãŸéš›ã®é †ç•ªã«ä¸€è‡´ã™ã‚‹ã“ã¨ãŒä¿è¨¼ã•ã‚Œã¦ã„る。</dd>
+</dl>
+
+<p>マップã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰å–り出ã—ãŸã‚ーã«å¯¾å¿œã™ã‚‹å€¤ã‚’å–り出ã™ã«ã¯ã€é–¢æ•°ã€€`cbmapiterval' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *cbmapiterval(const char *<var>kbuf</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`kbuf' ã¯ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰å–り出ã—ãŸã‚ーã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>マップã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbmaprnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbmaprnum(const CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã§ã‚る。</dd>
+</dl>
+
+<p>マップã«å«ã¾ã‚Œã‚‹å…¨ã¦ã®ã‚ーをå«ã‚€ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbmapkeys' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbmapkeys(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒžãƒƒãƒ—ã«å«ã¾ã‚Œã‚‹å…¨ã¦ã®ã‚ーをå«ã‚€ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>マップã«å«ã¾ã‚Œã‚‹å…¨ã¦ã®å€¤ã‚’å«ã‚€ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbmapvals' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbmapvals(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒžãƒƒãƒ—ã«å«ã¾ã‚Œã‚‹å…¨ã¦ã®å€¤ã‚’å«ã‚€ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>マップを直列化ã—ã¦ãƒã‚¤ãƒˆé…列ã«ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapdump' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbmapdump(const CBMAP *<var>map</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`map' ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ã¯æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ç›´åˆ—化ã•ã‚ŒãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>直列化ã•ã‚ŒãŸãƒžãƒƒãƒ—を復元ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmapload' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBMAP *cbmapload(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯ãƒã‚¤ãƒˆé…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–°ã—ã„マップãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>直列化ã—ãŸãƒžãƒƒãƒ—ã‹ã‚‰ã²ã¨ã¤ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’抽出ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmaploadone' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbmaploadone(const char *<var>ptr</var>, int <var>size</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯ãƒã‚¤ãƒˆé…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€è©²å½“ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>ヒープãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `cbheapopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBHEAP *cbheapopen(int <var>size</var>, int <var>max</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`size' ã¯å„レコードã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`max' ã¯ãƒ’ープã«æ ¼ç´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚`compar' ã¯æ¯”較関数を指定ã™ã‚‹ã€‚二ã¤ã®å¼•æ•°ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。比較関数ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£æ•°ã‚’ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² æ•°ã‚’ã€ä¸¡è€…ãŒç‰ã—ã‘れ㰠0 ã‚’è¿”ã™ã¹ãã§ã‚る。戻り値ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã€‚</dd>
+</dl>
+
+<p>ヒープãƒãƒ³ãƒ‰ãƒ«ã‚’複製ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbheapdup' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBHEAP *cbheapdup(CBHEAP *<var>heap</var>);</kbd></dt>
+<dd>`heap' ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–°ã—ã„ヒープãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>ヒープãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbheapclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbheapclose(CBHEAP *<var>heap</var>);</kbd></dt>
+<dd>`heap' ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>ヒープã«æ ¼ç´ã•ã‚ŒãŸãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbheapnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbheapnum(CBHEAP *<var>heap</var>);</kbd></dt>
+<dd>`heap' ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆã«æ ¼ç´ã•ã‚ŒãŸãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã§ã‚る。</dd>
+</dl>
+
+<p>ヒープã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’挿入ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbheapinsert' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbheapinsert(CBHEAP *<var>heap</var>, const void *<var>ptr</var>);</kbd></dt>
+<dd>`heap' ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`ptr' ã¯è¿½åŠ ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒè¿½åŠ ã•ã‚Œã‚Œã°çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。新ã—ã„レコードã®å€¤ãŒæ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ€å¤§å€¤ã‚ˆã‚Šå¤§ãã‘ã‚Œã°ã€æ–°ã—ã„レコードã¯è¿½åŠ ã•ã‚Œãªã„。新ã—ã„レコードãŒè¿½åŠ ã•ã‚Œã¦ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ãŒæœ€å¤§æ•°ã‚’越ãˆãŸå ´åˆã€æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ€å¤§å€¤ã®ã‚‚ã®ãŒå‰Šé™¤ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>ヒープ内ã®ã‚るレコードã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbheapget' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void *cbheapval(CBHEAP *<var>heap</var>, int <var>index</var>);</kbd></dt>
+<dd>`heap' ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`index' ã¯å–り出ã™ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’指定ã™ã‚‹ã€‚戻り値ã¯è©²å½“レコードã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。`index' ãŒè¦ç´ 数以上ãªã‚‰ã°ã€æˆ»ã‚Šå€¤ã¯ `NULL' ã§ã‚る。レコードã¯å†…部的ã«ã¯æ¯”較関数ã®è² 値ã«åŸºã¥ã„ã¦çµ„織化ã•ã‚Œã¦ã„ã‚‹ã“ã¨ã«æ³¨æ„ã™ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ヒープを確ä¿ã•ã‚ŒãŸé ˜åŸŸã«å¤‰æ›ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbheaptomalloc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void *cbheaptomalloc(CBHEAP *<var>heap</var>, int *<var>np</var>);</kbd></dt>
+<dd>`heap' ã¯ãƒ’ープãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`np' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯ãƒ’ープã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。レコードã¯ã‚½ãƒ¼ãƒˆæ¸ˆã¿ã«ãªã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。元ã®ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ãれをå†ã³è§£æ”¾ã—ã¦ã¯ãªã‚‰ãªã„。</dd>
+</dl>
+
+<p>書å¼ã«åŸºã¥ã„ãŸæ–‡å—列をメモリ上ã§ç¢ºä¿ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbsprintf' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbsprintf(const char *<var>format</var>, ...);</kbd></dt>
+<dd>`format' ã¯printf風ã®æ›¸å¼æ–‡å—列を指定ã™ã‚‹ã€‚変æ›æ–‡å— `%' ã‚’ãƒ•ãƒ©ã‚°æ–‡å— `d'ã€`o'ã€`u'ã€`x'ã€`X'ã€`e'ã€`E'ã€`f'ã€`g'ã€`G'ã€`c'ã€`s' ãŠã‚ˆã³ `%' ã‚’ä¼´ã‚ã›ã¦ä½¿ç”¨ã™ã‚‹ã“ã¨ãŒã§ãる。フィールドã®å¹…ã¨ç²¾åº¦ã®æŒ‡ç¤ºåを変æ›æ–‡å—ã¨ãƒ•ãƒ©ã‚°æ–‡å—ã®é–“ã«ç½®ãã“ã¨ãŒã§ãる。ãã®æŒ‡ç¤ºåã¯10進数å—ã€`.'ã€`+'ã€`-' ãŠã‚ˆã³ã‚¹ãƒšãƒ¼ã‚¹æ–‡å—ã‹ã‚‰ãªã‚‹ã€‚ãã®ä»–ã®å¼•æ•°ã¯æ›¸å¼æ–‡å—列ã«ã‚ˆã£ã¦åˆ©ç”¨ã•ã‚Œã‚‹ã€‚戻り値ã¯çµæžœã®æ–‡å—列ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>æ–‡å—列ä¸ã®ãƒ‘ターンを置æ›ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbreplace' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbreplace(const char *<var>str</var>, CBMAP *<var>pairs</var>);</kbd></dt>
+<dd>`str' ã¯ç½®æ›å‰ã®æ–‡å—列を指定ã™ã‚‹ã€‚`pairs' ã¯ç½®æ›ã®ãƒšã‚¢ã‹ã‚‰ãªã‚‹ãƒžãƒƒãƒ—ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚å„ペアã®ã‚ーã¯ç½®æ›å‰ã®ãƒ‘ターンを指定ã—ã€å€¤ã¯ç½®æ›å¾Œã®ãƒ‘ターンを指定ã™ã‚‹ã€‚戻り値ã¯çµæžœã®æ–‡å—列ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>一連ã®ãƒ‡ãƒ¼ã‚¿ã‚’分割ã—ã¦ãƒªã‚¹ãƒˆã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `cbsplit' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbsplit(const char *<var>ptr</var>, int <var>size</var>, const char *<var>delim</var>);</kbd></dt>
+<dd>`ptr' ã¯å†…容ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`delim' ã¯åŒºåˆ‡ã‚Šæ–‡å—ã‚’å«ã‚€æ–‡å—列を指定ã™ã‚‹ã‹ã€`NULL' ãªã‚‰çµ‚端文å—を区切り文å—ã¨ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。区切り文å—ãŒé€£ç¶šã—ã¦ã„ã‚‹å ´åˆã§ã‚‚ã€ãã®é–“ã«ç©ºã®è¦ç´ ãŒã‚ã‚‹ã¨ã¿ãªã™ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ファイルã®å…¨ãƒ‡ãƒ¼ã‚¿ã‚’èªã¿è¾¼ã‚€ã«ã¯ã€é–¢æ•° `cbreadfile' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbreadfile(const char *<var>name</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ãŒã€`NULL' ãªã‚‰æ¨™æº–入力をèªã¿è¾¼ã‚€ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰èªã¿è¾¼ã‚“ã ãƒ‡ãƒ¼ã‚¿ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>一連ã®ãƒ‡ãƒ¼ã‚¿ã‚’ファイルã«æ›¸ã込むã«ã¯ã€é–¢æ•° `cbwritefile' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbwritefile(const char *<var>name</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ•ã‚¡ã‚¤ãƒ«åを指定ã™ã‚‹ãŒã€`NULL' ãªã‚‰æ¨™æº–出力ã«æ›¸ã出ã•ã‚Œã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€è©²å½“ã®ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ファイルãŒå˜åœ¨ã™ã‚‹å ´åˆã«ã¯ä¸Šæ›¸ãã•ã‚Œã€ãã†ã§ãªã„å ´åˆã¯æ–°ã—ã„ファイルãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>ファイルã®å„行をèªã¿è¾¼ã‚€ã«ã¯ã€é–¢æ•° `cbreadlines' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbreadlines(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ãŒã€`NULL' ãªã‚‰æ¨™æº–入力をèªã¿è¾¼ã‚€ã€‚æˆåŠŸã™ã‚Œã°æˆ»ã‚Šå€¤ã¯å„è¡Œã®ãƒ‡ãƒ¼ã‚¿ã‚’ä¿æŒã™ã‚‹ãƒªã‚¹ãƒˆã®ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€å¤±æ•—ãªã‚‰ `NULL' ã§ã‚る。改行文å—ã¯å‰Šé™¤ã•ã‚Œã‚‹ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ディレクトリã«å«ã¾ã‚Œã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã®ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbdirlist' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbdirlist(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚æˆåŠŸã™ã‚Œã°æˆ»ã‚Šå€¤ã¯å„ファイルã®åå‰ã‚’ä¿æŒã™ã‚‹ãƒªã‚¹ãƒˆã®ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€å¤±æ•—ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ファイルやディレクトリã®çŠ¶æ…‹ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbfilestat' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbfilestat(const char *<var>name</var>, int *<var>isdirp</var>, int *<var>sizep</var>, int *<var>mtimep</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã‚„ディレクトリã®åå‰ã‚’指定ã™ã‚‹ã€‚`dirp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«ãƒ•ã‚¡ã‚¤ãƒ«ãŒãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‹å¦ã‹ã‚’æ ¼ç´ã™ã‚‹ã€‚`sizep' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚`mtimep' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«ãƒ•ã‚¡ã‚¤ãƒ«ã®æœ€çµ‚æ›´æ–°æ™‚åˆ»ã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ファイルãŒå˜åœ¨ã—ãªã„å ´åˆã‚„パーミッションãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>ファイルã‹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã¨ãã®å†…容を削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbremove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã‚„ディレクトリã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ファイルãŒå˜åœ¨ã—ãªã„å ´åˆã‚„パーミッションãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>URLを構æˆè¦ç´ ã«åˆ†è§£ã™ã‚‹ã«ã¯ã€é–¢æ•° `cburlbreak' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBMAP *cburlbreak(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯URLã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。マップã®å„ã‚ーã¯è¦ç´ åã§ã‚る。ã‚ー "self" ã¯URLãれ自体を指示ã™ã‚‹ã€‚ã‚ー "scheme" ã¯URLã®ã‚¹ã‚ームを指示ã™ã‚‹ã€‚ã‚ー "host" ã¯ã‚µãƒ¼ãƒã®ãƒ›ã‚¹ãƒˆåを指示ã™ã‚‹ã€‚ã‚ー "port" ã¯ã‚µãƒ¼ãƒã®ãƒãƒ¼ãƒˆç•ªå·ã‚’指示ã™ã‚‹ã€‚ã‚ー "authority" ã¯èªè¨¼æƒ…å ±ã‚’æŒ‡ç¤ºã™ã‚‹ã€‚ã‚ー "path" ã¯ãƒªã‚½ãƒ¼ã‚¹ã®ãƒ‘スを指示ã™ã‚‹ã€‚ã‚ー "file" ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªéƒ¨åˆ†ã‚’除ã„ãŸãƒ•ã‚¡ã‚¤ãƒ«åを指示ã™ã‚‹ã€‚ã‚ー "query" ã¯ã‚¯ã‚¨ãƒªæ–‡å—列を指示ã™ã‚‹ã€‚ã‚ー "fragment" ã¯ãƒ•ãƒ©ã‚°ãƒ¡ãƒ³ãƒˆæ–‡å—列を指示ã™ã‚‹ã€‚サãƒãƒ¼ãƒˆã•ã‚Œã‚‹ã‚¹ã‚ームã¯HTTPã¨HTTPSã¨FTPã¨FILEã§ã‚る。絶対URLã«ã‚‚相対URLã«ã‚‚対応ã™ã‚‹ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cbmapopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cbmapclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>相対URLを絶対URLを用ã„ã¦è§£æ±ºã™ã‚‹ã«ã¯ã€é–¢æ•° `cburlresolve' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cburlresolve(const char *<var>base</var>, const char *<var>target</var>);</kbd></dt>
+<dd>`base' ã¯ãƒ™ãƒ¼ã‚¹ãƒã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã®çµ¶å¯¾URLを指定ã™ã‚‹ã€‚`target' ã¯è§£æ±ºã™ã¹ãURLを指定ã™ã‚‹ã€‚戻り値ã¯è§£æ±ºã•ã‚ŒãŸURLã§ã‚る。ターゲットURLãŒç›¸å¯¾URLã®å ´åˆã€ãƒ™ãƒ¼ã‚¹ãƒã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‹ã‚‰ã®ç›¸å¯¾ä½ç½®ã®URLã‚’è¿”ã™ã€‚ターゲットURLãŒçµ¶å¯¾URLã®å ´åˆã€ã‚¿ãƒ¼ã‚²ãƒƒãƒˆURLã®ã‚³ãƒ”ーを返ã™ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>一連ã®ã‚ªãƒ–ジェクトをURLエンコーディングã§ç¬¦å·åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `cburlencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cburlencode(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯çµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>URLエンコーディングã§ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列を復元ã™ã‚‹ã«ã¯ã€é–¢æ•° `cburldecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cburldecode(const char *<var>str</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`str' ã¯ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>一連ã®ã‚ªãƒ–ジェクトをBase64エンコーディングã§ç¬¦å·åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbbaseencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbbaseencode(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯çµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>Base64エンコーディングã§ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列を復元ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbbasedecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbbasedecode(const char *<var>str</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`str' ã¯ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>一連ã®ã‚ªãƒ–ジェクトをquoted-printableエンコーディングã§ç¬¦å·åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbquoteencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbquoteencode(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯çµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>quoted-printableエンコーディングã§ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列を復元ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbquotedecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbquotedecode(const char *<var>str</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`str' ã¯ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>MIMEã®æ–‡å—列をヘッダã¨ãƒœãƒ‡ã‚£ã«åˆ†å‰²ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmimebreak' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbmimebreak(const char *<var>ptr</var>, int <var>size</var>, CBMAP *<var>attrs</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯MIMEã®ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`attrs' ã¯å±žæ€§ã‚’ä¿å˜ã™ã‚‹ãŸã‚ã®ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ãŒã€`NULL' ã®å ´åˆã«ã¯åˆ©ç”¨ã•ã‚Œãªã„。マップã®å„ã‚ーã¯å°æ–‡å—ã«æ£è¦åŒ–ã•ã‚ŒãŸå±žæ€§åã§ã‚る。`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯ãƒœãƒ‡ã‚£ã®æ–‡å—列ã§ã‚る。コンテントタイプãŒæŒ‡å®šã•ã‚Œã¦ã„ã‚‹å ´åˆã€ãƒžãƒƒãƒ—ã®ã‚ー "TYPE" ã¯ãã®ã‚¿ã‚¤ãƒ—を指示ã™ã‚‹ã€‚æ–‡å—コードãŒæŒ‡å®šã•ã‚Œã¦ã„ã‚‹å ´åˆã€ã‚ー "CHARSET" ã¯ãã®ã‚³ãƒ¼ãƒ‰åを指示ã™ã‚‹ã€‚マルãƒãƒ‘ートã®åŒºåˆ‡ã‚Šæ–‡å—列ãŒæŒ‡å®šã•ã‚Œã¦ã„ã‚‹å ´åˆã€ã‚ー "BOUNDARY" ã¯ãã®æ–‡å—列を指示ã™ã‚‹ã€‚コンテントディスãƒã‚¸ã‚·ãƒ§ãƒ³ãŒæŒ‡å®šã•ã‚Œã¦ã„ã‚‹å ´åˆã€ã‚ー "DISPOSITION" ã¯ãã®æ–¹é‡ã‚’指示ã™
る。ファイルåãŒæŒ‡å®šã•ã‚Œã¦ã„ã‚‹å ´åˆã€ã‚ー "FILENAME" ã¯ãã®åå‰ã‚’指示ã™ã‚‹ã€‚属性åãŒæŒ‡å®šã•ã‚Œã¦ã„ã‚‹å ´åˆã€ã‚ー "NAME" ã¯ãã®åå‰ã‚’指示ã™ã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>MIMEã®ãƒžãƒ«ãƒãƒ‘ートデータã®æ–‡å—列をå„パートã«åˆ†å‰²ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmimeparts' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbmimeparts(const char *<var>ptr</var>, int <var>size</var>, const char *<var>boundary</var>);</kbd></dt>
+<dd>`ptr' ã¯MIMEã®ãƒ‡ãƒ¼ã‚¿ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`boundary' ã¯åŒºåˆ‡ã‚Šæ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。リストã®å„è¦ç´ ã¯ãƒ‘ートã®æ–‡å—列ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>æ–‡å—列をMIMEエンコーディングã§ç¬¦å·åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmimeencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbmimeencode(const char *<var>str</var>, const char *<var>encname</var>, int <var>base</var>);</kbd></dt>
+<dd>`str' ã¯æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`encname' ã¯æ–‡å—コードåã®æ–‡å—列を指定ã™ã‚‹ã€‚`base' ã¯Base64エンコードを使ã†ã‹å¦ã‹ã‚’指定ã™ã‚‹ã€‚å½ã®å ´åˆã¯quoted-printableãŒç”¨ã„られる。戻り値ã¯çµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>MIMEエンコーディングã§ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列を復元ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbmimedecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbmimedecode(const char *<var>str</var>, char *<var>enp</var>);</kbd></dt>
+<dd>`str' ã¯ç¬¦å·åŒ–ã•ã‚ŒãŸæ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`enp' ã¯æ–‡å—コードåを書ã込むãŸã‚ã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ãŒã€`NULL' ã®å ´åˆã¯åˆ©ç”¨ã•ã‚Œãªã„。ãƒãƒƒãƒ•ã‚¡ã®ã‚µã‚¤ã‚ºã¯32ãƒã‚¤ãƒˆä»¥ä¸Šã§ãªã‘ã‚Œã°ãªã‚‰ãªã„。戻り値ã¯çµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>CSVã®æ–‡å—列を行ã«åˆ†å‰²ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbcsvrows' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbcsvrows(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯CSVã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。リストã®å„è¦ç´ ã¯è¡Œã®æ–‡å—列ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。入力文å—列ã®æ–‡å—コードã¯US-ASCIIã€UTF-8ã€ISO-8859-*ã€EUC-*ã€Shift_JISã®ã©ã‚Œã‹ã§ã‚ã‚‹å¿…è¦ãŒã‚る。MS-Excelã¨äº’æ›ã—ã¦ã€ã“れらCSV用関数群ã¯ã€ãƒ€ãƒ–ルクォートã§å›²ã‚“ã§ã‚³ãƒ³ãƒžãªã©ã®ãƒ¡ã‚¿æ–‡å—ã‚’å«ã‚ãŸã‚»ãƒ«ã‚’扱ã†ã“ã¨ãŒã§ãる。</dd>
+</dl>
+
+<p>CSVã®è¡Œã®æ–‡å—列をセルã«åˆ†å‰²ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbcsvcells' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbcsvcells(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯CSVã®è¡Œã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。リストã®å„è¦ç´ ã¯ã‚»ãƒ«å†…容をアンエスケープã—ãŸæ–‡å—列ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>CSVã®ãƒ¡ã‚¿æ–‡å—をエスケープã—ãŸæ–‡å—列を得るã«ã¯ã€é–¢æ•° `cbcsvescape' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbcsvescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ¡ã‚¿æ–‡å—を無効化ã—ãŸçµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>XMLã®å®Ÿä½“å‚照をアンエスケープã—ãŸæ–‡å—列を得るã«ã¯ã€é–¢æ•° `cbcsvunescape' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbcsvunescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ¡ã‚¿æ–‡å—ã‚’ä¼´ã£ãŸçµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>XMLã®æ–‡å—列をタグã¨ãƒ†ã‚ストセクションã«åˆ†å‰²ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbxmlbreak' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *cbxmlbreak(const char *<var>str</var>, int <var>cr</var>);</kbd></dt>
+<dd>`str' ã¯XMLã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`cr' ã¯ã‚³ãƒ¡ãƒ³ãƒˆã‚’削除ã™ã‚‹ã‹å¦ã‹ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。リストã®å„è¦ç´ ã¯ã‚¿ã‚°ã‹ãƒ†ã‚ストセクションã®æ–‡å—列ã§ã‚る。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。入力文å—列ã®æ–‡å—コードã¯US-ASCIIã€UTF-8ã€ISO-8859-*ã€EUC-*ã€Shift_JISã®ã©ã‚Œã‹ã§ã‚ã‚‹å¿…è¦ãŒã‚る。ã“れらXML用関数群ã¯å¦¥å½“性検証を行ã†XMLパーザã§ã¯ãªã„ã®ã§ã€HTMLã‚„SGMLも扱ã†ã“ã¨ãŒã§ãる。</dd>
+</dl>
+
+<p>XMLã®ã‚¿ã‚°ã®å±žæ€§ã®ãƒžãƒƒãƒ—ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbxmlattrs' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBMAP *cbxmlattrs(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯ã‚¿ã‚°ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。マップã®å„ã‚ーã¯å±žæ€§ã®åå‰ã§ã‚る。å„値ã¯ã‚¢ãƒ³ã‚¨ã‚¹ã‚±ãƒ¼ãƒ—ã•ã‚Œã‚‹ã€‚空文å—列をã‚ーã«ã™ã‚‹ã¨ã‚¿ã‚°ã®åå‰ã‚’å–り出ã™ã“ã¨ãŒã§ãる。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cbmapopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cbmapclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>XMLã®ãƒ¡ã‚¿æ–‡å—をエスケープã—ãŸæ–‡å—列を得るã«ã¯ã€é–¢æ•° `cbxmlescape' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbxmlescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ¡ã‚¿æ–‡å—を無効化ã—ãŸçµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。ã“ã®é–¢æ•°ã¯ `&'〠`<'ã€`>'ã€`"' ã®ã¿ã‚’変æ›ã™ã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>XMLã®å®Ÿä½“å‚照をアンエスケープã—ãŸæ–‡å—列を得るã«ã¯ã€é–¢æ•° `cbxmlunescape' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbxmlunescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯å¯¾è±¡ã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ¡ã‚¿æ–‡å—ã‚’ä¼´ã£ãŸçµæžœã®æ–‡å—列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚る。ã“ã®é–¢æ•°ã¯ `&'ã€`<'ã€`>'ã€`"' ã®ã¿ã‚’復元ã™ã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ZLIBを用ã„ã¦ä¸€é€£ã®ã‚ªãƒ–ジェクトを圧縮ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbdeflate' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbdeflate(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ã®å‚照先ã«ã¯æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒZLIBを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>ZLIBを用ã„ã¦åœ§ç¸®ã•ã‚ŒãŸã‚ªãƒ–ジェクトを伸長ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbinflate' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbinflate(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒZLIBを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>GZIPを用ã„ã¦ä¸€é€£ã®ã‚ªãƒ–ジェクトを圧縮ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbgzencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbgzencode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ã®å‚照先ã«ã¯æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒZLIBを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>GZIPを用ã„ã¦åœ§ç¸®ã•ã‚ŒãŸã‚ªãƒ–ジェクトを伸長ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbgzdecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbgzdecode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒZLIBを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>一連ã®ã‚ªãƒ–ジェクトã®CRC32ãƒã‚§ãƒƒã‚¯ã‚µãƒ ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbgetcrc' を用ã„る。</p>
+
+<dl>
+<dt><kbd>unsigned int cbgetcrc(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯ã‚ªãƒ–ジェクトã®CRC32ãƒã‚§ãƒƒã‚¯ã‚µãƒ ã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒZLIBを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>LZOを用ã„ã¦ä¸€é€£ã®ã‚ªãƒ–ジェクトを圧縮ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblzoencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cblzoencode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ã®å‚照先ã«ã¯æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒLZOを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>LZOを用ã„ã¦åœ§ç¸®ã•ã‚ŒãŸã‚ªãƒ–ジェクトを伸長ã™ã‚‹ã«ã¯ã€é–¢æ•° `cblzodecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cblzodecode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒLZOを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>BZIP2を用ã„ã¦ä¸€é€£ã®ã‚ªãƒ–ジェクトを圧縮ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbbzencode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbbzencode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ã®å‚照先ã«ã¯æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒBZIP2を有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>BZIP2を用ã„ã¦åœ§ç¸®ã•ã‚ŒãŸã‚ªãƒ–ジェクトを伸長ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbbzdecode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbbzdecode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒBZIP2を有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>æ–‡å—列ã®æ–‡å—コードを変æ›ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbiconv' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbiconv(const char *<var>ptr</var>, int <var>size</var>, const char *<var>icode</var>, const char *<var>ocode</var>, int *<var>sp</var>, int *<var>mp</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`icode' ã¯å…¥åŠ›æ–‡å—列ã®æ–‡å—コードã®åå‰ã‚’指定ã™ã‚‹ã€‚`outcode' ã¯å‡ºåŠ›æ–‡å—列ã®æ–‡å—コードã®åå‰ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚`mp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«å¤‰æ›ã«å¤±æ•—ã—ãŸæ–‡å—æ•°ã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çµæžœã®é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。ã“ã®é–¢æ•°ã¯QDBMãŒICONVを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>æ–‡å—列ã®æ–‡å—コードを自動判定ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbencname' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *cbencname(const char *<var>str</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' ã¯é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãã®é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(ptr)' ã®å€¤ã¨ãªã‚‹ã€‚`icode' ã¯å…¥åŠ›æ–‡å—列ã®æ–‡å—コードã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ã‚¨ãƒ³ã‚³ãƒ¼ãƒ‰ã®åå‰ã®æ–‡å—列ã§ã‚る。ç¾çŠ¶ã§ã¯ã€US-ASCIIã€ISO-2022-JPã€Shift_JISã€CP932ã€EUC-JPã€UTF-8ã€UTF-16ã€UTF-16BEãŠã‚ˆã³UTF-16LEã«å¯¾å¿œã—ã¦ã„る。ãã®ã„ãšã‚Œã§ã‚‚ãªã„å ´åˆã¯ã€ISO-8859-1ã¨åˆ¤å®šã™ã‚‹ã€‚ã“ã®é–¢æ•°ã¯QDBMãŒICONVを有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã®ã¿åˆ©ç”¨ã§ãる。</dd>
+</dl>
+
+<p>ãƒãƒ¼ã‚«ãƒ«æ™‚é–“ã®ç§’å˜ä½ã®æ™‚差を得るã«ã¯ã€é–¢æ•° `cbjetlag' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbjetlag(void);</kbd></dt>
+<dd>戻り値ã¯ãƒãƒ¼ã‚«ãƒ«æ™‚é–“ã®ç§’å˜ä½ã®æ™‚å·®ã§ã‚る。</dd>
+</dl>
+
+<p>ã‚る時間ã®ã‚°ãƒ¬ã‚´ãƒªã‚ªæš¦ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbcalendar' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbcalendar(time_t <var>t</var>, int <var>jl</var>, int *<var>yearp</var>, int *<var>monp</var>, int *<var>dayp</var>, int *<var>hourp</var>, int *<var>minp</var>, int *<var>secp</var>);</kbd></dt>
+<dd>`t' ã¯å¯¾è±¡ã®æ™‚間を指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ç¾åœ¨æ™‚é–“ãŒä½¿ã‚れる。`jl' ã¯ã‚ã‚‹å ´æ‰€ã®æ™‚差を秒å˜ä½ã§æŒ‡å®šã™ã‚‹ã€‚`yearp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«å¹´ã‚’æ ¼ç´ã™ã‚‹ã€‚`monp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æœˆã‚’æ ¼ç´ã™ã‚‹ã€‚1ã¯1月をæ„味ã—ã€12ã¯12月をæ„味ã™ã‚‹ã€‚`dayp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æ—¥ã‚’æ ¼ç´ã™ã‚‹ã€‚`hourp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æ™‚ã‚’æ ¼ç´ã™ã‚‹ã€‚`minp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«åˆ†ã‚’æ ¼ç´ã™ã‚‹ã€‚`secp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«ç§’ã‚’æ ¼ç´ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ã‚る日付ã®æ›œæ—¥ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbdayofweek' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int cbdayofweek(int <var>year</var>, int <var>mon</var>, int <var>day</var>);</kbd></dt>
+<dd>`year' ã¯æ—¥ä»˜ã®å¹´ã‚’指定ã™ã‚‹ã€‚`mon' ã¯æ—¥ä»˜ã®æœˆã‚’指定ã™ã‚‹ã€‚`day' ã¯æ—¥ä»˜ã®æ—¥ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ›œæ—¥ã®å€¤ã§ã‚る。0ã¯æ—¥æ›œã‚’æ„味ã—ã€6ã¯åœŸæ›œã‚’æ„味ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ã‚る日付をW3CDTFã®æ›¸å¼ã§è¡¨ã—ãŸæ–‡å—列を得るã«ã¯ã€é–¢æ•° `cbdatestrwww' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbdatestrwww(time_t <var>t</var>, int <var>jl</var>);</kbd></dt>
+<dd>`t' ã¯å¯¾è±¡ã®æ™‚間を指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ç¾åœ¨æ™‚é–“ãŒä½¿ã‚れる。`jl' ã¯ã‚ã‚‹å ´æ‰€ã®æ™‚差を秒å˜ä½ã§æŒ‡å®šã™ã‚‹ã€‚戻り値ã¯W3CDTFã®æ›¸å¼ï¼ˆYYYY-MM-DDThh:mm:ddTZD）ã®æ–‡å—列ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ã‚る日付をRFC 1123ã®æ›¸å¼ã§è¡¨ã—ãŸæ–‡å—列を得るã«ã¯ã€é–¢æ•° `cbdatestrhttp' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *cbdatestrhttp(time_t <var>t</var>, int <var>jl</var>);</kbd></dt>
+<dd>`t' ã¯å¯¾è±¡ã®æ™‚間を指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ç¾åœ¨æ™‚é–“ãŒä½¿ã‚れる。`jl' ã¯ã‚ã‚‹å ´æ‰€ã®æ™‚差を秒å˜ä½ã§æŒ‡å®šã™ã‚‹ã€‚戻り値ã¯RFC 1123ã®æ›¸å¼ï¼ˆWdy, DD-Mon-YYYY hh:mm:dd TZD）ã®æ–‡å—列ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>10進数ã‹16進数ã‹W3CDTFã‹RFC 822（1123）ã®æ›¸å¼ã§è¡¨ã—ãŸæ–‡å—列ã‹ã‚‰æ™‚é–“ã®å€¤ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbstrmktime' を用ã„る。</p>
+
+<dl>
+<dt><kbd>time_t cbstrmktime(const char *<var>str</var>);</kbd></dt>
+<dd>`str' ã¯10進数ã‹16進数ã‹W3CDTFã‹RFC 822（1123）ã®æ›¸å¼ã§è¡¨ã—ãŸæ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ™‚é–“ã®å€¤ã‹ã€æ›¸å¼ãŒä¸æ£ãªå ´åˆã¯ -1 ã§ã‚る。10進数㫠"s" を後置ã™ã‚‹ã¨ç§’å˜ä½ã‚’æ„味ã—ã€"m" を後置ã™ã‚‹ã¨åˆ†å˜ä½ã‚’æ„味ã—ã€"h" を後置ã™ã‚‹ã¨æ™‚å˜ä½ã‚’æ„味ã—ã€"d" を後置ã™ã‚‹ã¨æ—¥å˜ä½ã‚’æ„味ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ユーザã¨ã‚·ã‚¹ãƒ†ãƒ ã®ãƒ—ãƒã‚»ã‚¹æ™‚é–“ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `cbproctime' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbproctime(double *<var>usrp</var>, double *<var>sysp</var>);</kbd></dt>
+<dd>`usrp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«ãƒ¦ãƒ¼ã‚¶æ™‚é–“ã‚’æ ¼ç´ã™ã‚‹ã€‚時間ã®å˜ä½ã¯ç§’ã§ã‚る。`sysp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«ã‚·ã‚¹ãƒ†ãƒ æ™‚é–“ã‚’æ ¼ç´ã™ã‚‹ã€‚時間ã®å˜ä½ã¯ç§’ã§ã‚る。</dd>
+</dl>
+
+<p>標準入出力ãŒãƒã‚¤ãƒŠãƒªãƒ¢ãƒ¼ãƒ‰ã§ã‚ã‚‹ã“ã¨ã‚’ä¿è¨¼ã™ã‚‹ã«ã¯ã€é–¢æ•° `cbstdiobin' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void cbstdiobin(void);</kbd></dt>
+<dd>ã“ã®é–¢æ•°ã¯DOSçš„ãªãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ ã®ä¸Šã®ã‚¢ãƒ—リケーションã§æœ‰ç”¨ã§ã‚る。</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>以下ã®ã‚µãƒ³ãƒ—ルコードã¯å…¸åž‹çš„ãªåˆ©ç”¨ä¾‹ã§ã‚る。</p>
+
+<pre>#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+int main(int argc, char **argv){
+ CBDATUM *datum;
+ CBLIST *list;
+ CBMAP *map;
+ char *buf1, *buf2;
+ int i;
+
+ /* データãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ã */
+ datum = cbdatumopen("123", -1);
+ /* データを連çµã™ã‚‹ */
+ cbdatumcat(datum, "abc", -1);
+ /* データを表示ã™ã‚‹ */
+ printf("%s\n", cbdatumptr(datum));
+ /* データãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã‚‹ */
+ cbdatumclose(datum);
+
+ /* リストãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ã */
+ list = cblistopen();
+ /* リストã«è¦ç´ ã‚’è¿½åŠ ã™ã‚‹ */
+ cblistpush(list, "apple", -1);
+ cblistpush(list, "orange", -1);
+ /* å…¨ã¦ã®è¦ç´ を表示ã™ã‚‹ */
+ for(i = 0; i < cblistnum(list); i++){
+ printf("%s\n", cblistval(list, i, NULL));
+ }
+ /* リストãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã‚‹ */
+ cblistclose(list);
+
+ /* マップãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ã */
+ map = cbmapopen();
+ /* マップã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ */
+ cbmapput(map, "dog", -1, "bowwow", -1, 1);
+ cbmapput(map, "cat", -1, "meow", -1, 1);
+ /* 値をå–å¾—ã—ã¦è¡¨ç¤ºã™ã‚‹ */
+ printf("%s\n", cbmapget(map, "dog", -1, NULL));
+ printf("%s\n", cbmapget(map, "cat", -1, NULL));
+ /* マップãƒãƒ³ãƒ‰ãƒ«ã‚’é–‰ã˜ã‚‹ */
+ cbmapclose(map);
+
+ /* Base64ã®ç¬¦å·åŒ–を行ㆠ*/
+ buf1 = cbbaseencode("I miss you.", -1);
+ printf("%s\n", buf1);
+ /* Base64ã®å¾©å…ƒã‚’行ㆠ*/
+ buf2 = cbbasedecode(buf1, NULL);
+ printf("%s\n", buf2);
+ /* リソースを解放ã™ã‚‹ */
+ free(buf2);
+ free(buf1);
+
+ /* å˜ç´”ãªãƒã‚¤ãƒ³ã‚¿ã‚’ã‚°ãƒãƒ¼ãƒãƒ«ã‚¬ãƒ™ãƒ¼ã‚¸ã‚³ãƒ¬ã‚¯ã‚¿ã«ç™»éŒ²ã™ã‚‹ */
+ buf1 = cbmemdup("Take it easy.", -1);
+ cbglobalgc(buf1, free);
+ /* ãƒã‚¤ãƒ³ã‚¿ã¯åˆ©ç”¨ã§ãã‚‹ãŒè§£æ”¾ã™ã‚‹å¿…è¦ã¯ãªã„ */
+ printf("%s\n", buf1);
+
+ /* リストをグãƒãƒ¼ãƒãƒ«ã‚¬ãƒ™ãƒ¼ã‚¸ã‚³ãƒ¬ã‚¯ã‚¿ã«ç™»éŒ²ã™ã‚‹ */
+ list = cblistopen();
+ cbglobalgc(list, (void (*)(void *))cblistclose);
+ /* ãƒãƒ³ãƒ‰ãƒ«ã¯åˆ©ç”¨ã§ãã‚‹ãŒé–‰ã˜ã‚‹å¿…è¦ã¯ãªã„ */
+ cblistpush(list, "Don't hesitate.", -1);
+ for(i = 0; i < cblistnum(list); i++){
+ printf("%s\n", cblistval(list, i, NULL));
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Cabinを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹æ–¹æ³•ã¯ã€Depotã®å ´åˆã¨å…¨ãåŒã˜ã§ã‚る。</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>スレッド間ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€`cbglobalgc' を除ã„ãŸCabinã®å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="cabincli" id="cabincli" class="head">Cabin用コマンド</a></h2>
+
+<p>Cabinã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `cbtest' ã¯Cabinã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計るã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`rnum' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>cbtest sort [-d] <var>rnum</var></kbd></dt>
+<dd>ソートアルゴリズムã®ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbtest strstr [-d] <var>rnum</var></kbd></dt>
+<dd>æ–‡å—列探索アルゴリズムã®ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbtest list [-d] <var>rnum</var></kbd></dt>
+<dd>リストã®æ›¸ãè¾¼ã¿ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbtest map [-d] <var>rnum</var></kbd></dt>
+<dd>マップã®æ›¸ãè¾¼ã¿ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbtest wicked <var>rnum</var></kbd></dt>
+<dd>リストã¨ãƒžãƒƒãƒ—ã®å„種更新æ“作を無作為ã«é¸æŠžã—ã¦å®Ÿè¡Œã™ã‚‹ã€‚</dd>
+<dt><kbd>cbtest misc</kbd></dt>
+<dd>雑多ãªãƒ«ãƒ¼ãƒãƒ³ã®ãƒ†ã‚¹ãƒˆã‚’実行ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-d</kbd> : çµæžœã®ãƒ‡ãƒ¼ã‚¿ã‚’èªã¿è¾¼ã‚“ã§è¡¨ç¤ºã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>コマンド `cbcodec' ã¯CabinãŒæä¾›ã™ã‚‹ã‚¨ãƒ³ã‚³ãƒ¼ãƒ‰ãŠã‚ˆã³ãƒ‡ã‚³ãƒ¼ãƒ‰ã®æ©Ÿèƒ½ã‚’利用ã™ã‚‹ãƒ„ールã§ã‚る。以下ã®æ›¸å¼ã§ç”¨ã„る。`file' ã¯å…¥åŠ›ãƒ•ã‚¡ã‚¤ãƒ«ã‚’指定ã™ã‚‹ãŒã€çœç•¥ã•ã‚Œã‚Œã°æ¨™æº–入力をèªã¿è¾¼ã‚€ã€‚</p>
+
+<dl>
+<dt><kbd>cbcodec url [-d] [-br] [-rs <var>base</var> <var>target</var>] [-l] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>URLエンコードã¨ãã®ãƒ‡ã‚³ãƒ¼ãƒ‰ã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbcodec base [-d] [-l] [-c <var>num</var>] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>Base64エンコードã¨ãã®ãƒ‡ã‚³ãƒ¼ãƒ‰ã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbcodec quote [-d] [-l] [-c <var>num</var>] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>quoted-printableエンコードã¨ãã®ãƒ‡ã‚³ãƒ¼ãƒ‰ã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbcodec mime [-d] [-hd] [-bd] [-part <var>num</var>] [-l] [-ec <var>code</var>] [-qp] [-dc] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>MIMEエンコードã¨ãã®ãƒ‡ã‚³ãƒ¼ãƒ‰ã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbcodec csv [-d] [-t] [-l] [-e <var>expr</var>] [-html] [<var>file</var>]</kbd></dt>
+<dd>CSVã®å‡¦ç†ã‚’è¡Œã†ã€‚デフォルトã§ã¯ãƒ¡ã‚¿æ–‡å—ã®ã‚¨ã‚¹ã‚±ãƒ¼ãƒ—ã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbcodec xml [-d] [-p] [-l] [-e <var>expr</var>] [-tsv] [<var>file</var>]</kbd></dt>
+<dd>XMLã®å‡¦ç†ã‚’è¡Œã†ã€‚デフォルトã§ã¯ãƒ¡ã‚¿æ–‡å—ã®ã‚¨ã‚¹ã‚±ãƒ¼ãƒ—ã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>cbcodec zlib [-d] [-gz] [-crc] [<var>file</var>]</kbd></dt>
+<dd>ZLIBã®åœ§ç¸®ã¨ãã®ä¼¸é•·ã‚’è¡Œã†ã€‚ZLIBを有効化ã—ã¦QDBMをビルドã—ãŸå ´åˆã«ã®ã¿åˆ©ç”¨å¯èƒ½ã§ã‚る。</dd>
+<dt><kbd>cbcodec lzo [-d] [<var>file</var>]</kbd></dt>
+<dd>LZOã®åœ§ç¸®ã¨ãã®ä¼¸é•·ã‚’è¡Œã†ã€‚LZOを有効化ã—ã¦QDBMをビルドã—ãŸå ´åˆã«ã®ã¿åˆ©ç”¨å¯èƒ½ã§ã‚る。</dd>
+<dt><kbd>cbcodec bzip [-d] [<var>file</var>]</kbd></dt>
+<dd>BZIP2ã®åœ§ç¸®ã¨ãã®ä¼¸é•·ã‚’è¡Œã†ã€‚BZIP2を有効化ã—ã¦QDBMをビルドã—ãŸå ´åˆã«ã®ã¿åˆ©ç”¨å¯èƒ½ã§ã‚る。</dd>
+<dt><kbd>cbcodec iconv [-ic <var>code</var>] [-oc <var>code</var>] [-ol <var>ltype</var>] [-cn] [-um] [-wc] [<var>file</var>]</kbd></dt>
+<dd>ICONVã«ã‚ˆã‚‹æ–‡å—コードã®å¤‰æ›ã‚’è¡Œã†ã€‚ICONVを有効化ã—ã¦QDBMをビルドã—ãŸå ´åˆã«ã®ã¿åˆ©ç”¨å¯èƒ½ã§ã‚る。</dd>
+<dt><kbd>cbcodec date [-wf] [-rf] [-utc] [<var>str</var>]</kbd></dt>
+<dd>`str' ã§æŒ‡å®šã—ãŸæ—¥ä»˜ã®æ–‡å—列ã®æ›¸å¼ã‚’変æ›ã™ã‚‹ã€‚デフォルトã§ã¯ã€UNIX時間を出力ã™ã‚‹ã€‚`str' ãŒçœç•¥ã•ã‚Œã‚‹ã¨ã€ç¾åœ¨æ—¥æ™‚を扱ã†ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-d</kbd> : エンコード（エスケープ）ã§ã¯ãªãã€ãƒ‡ã‚³ãƒ¼ãƒ‰ï¼ˆã‚¢ãƒ³ã‚¨ã‚¹ã‚±ãƒ¼ãƒ—）を行ã†ã€‚</li>
+<li><kbd>-br</kbd> : URLを構æˆè¦ç´ ã«åˆ†è§£ã™ã‚‹ã€‚</li>
+<li><kbd>-rs</kbd> : 相対URLを解決ã™ã‚‹ã€‚</li>
+<li><kbd>-l</kbd> : 出力ã®æœ«å°¾ã«æ”¹è¡Œæ–‡å—ã‚’åŠ ãˆã‚‹ã€‚</li>
+<li><kbd>-e <var>expr</var></kbd> : 入力データを直接指定ã™ã‚‹ã€‚</li>
+<li><kbd>-c <var>num</var></kbd> : エンコードã®éš›ã®æ¡æ•°åˆ¶é™ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-hd</kbd> : MIMEã®æ§‹æ–‡è§£æžã‚’è¡Œã„ã€ãƒ˜ãƒƒãƒ€ã‚’TSVå½¢å¼ã§æŠ½å‡ºã™ã‚‹ã€‚</li>
+<li><kbd>-bd</kbd> : MIMEã®æ§‹æ–‡è§£æžã‚’è¡Œã„ã€ãƒœãƒ‡ã‚£ã‚’抽出ã™ã‚‹ã€‚</li>
+<li><kbd>-part <var>num</var></kbd> : MIMEã®æ§‹æ–‡è§£æžã‚’è¡Œã„ã€ç‰¹å®šã®ãƒ‘ートを抽出ã™ã‚‹ã€‚</li>
+<li><kbd>-ec <var>code</var></kbd> : 入力ã®æ–‡å—コードを指定ã™ã‚‹ã€‚デフォルトã¯UTF-8ã§ã‚る。</li>
+<li><kbd>-qp</kbd> : quoted-printableエンコードを用ã„る。デフォルトã¯Base64ã§ã‚る。</li>
+<li><kbd>-dc</kbd> : デコードçµæžœã®æ–‡å—列ã§ãªãã€æ–‡å—コードåを出力ã™ã‚‹ã€‚</li>
+<li><kbd>-t</kbd> : CSVã®æ§‹é€ を解æžã™ã‚‹ã€‚TSVã«å¤‰æ›ã—ã¦è¡¨ç¤ºã™ã‚‹ã€‚セル内ã®ã‚¿ãƒ–ã¨æ”¹è¡Œã¯å‰Šé™¤ã•ã‚Œã‚‹ã€‚</li>
+<li><kbd>-html</kbd> : CSVã®æ§‹é€ を解æžã™ã‚‹ã€‚HTMLã«å¤‰æ›ã—ã¦å‡ºåŠ›ã™ã‚‹ã€‚</li>
+<li><kbd>-p</kbd> : XMLã®æ§‹æ–‡è§£æžã‚’è¡Œã†ã€‚ã‚¿ã‚°ã¨ãƒ†ã‚ストセクションをヘッダã§åˆ†ã‘ã¦è¡¨ç¤ºã™ã‚‹ã€‚</li>
+<li><kbd>-tsv</kbd> : XMLã®æ§‹æ–‡è§£æžã‚’è¡Œã†ã€‚çµæžœã‚’TSVå½¢å¼ã§å‡ºåŠ›ã™ã‚‹ã€‚テã‚ストセクションã®ã‚¿ãƒ–ã¨æ”¹è¡Œã¯URLエンコードã•ã‚Œã‚‹ã€‚</li>
+<li><kbd>-gz</kbd> : GZIPå½¢å¼ã‚’用ã„る。</li>
+<li><kbd>-crc</kbd> : CRC32ã®ãƒã‚§ãƒƒã‚¯ã‚µãƒ をビッグエンディアンã®16進数ã§å‡ºåŠ›ã™ã‚‹ã€‚</li>
+<li><kbd>-ic <var>code</var></kbd> : 入力ã®æ–‡å—コードを指定ã™ã‚‹ã€‚デフォルトã ã¨è‡ªå‹•åˆ¤å®šã™ã‚‹ã€‚</li>
+<li><kbd>-oc <var>code</var></kbd> : 出力ã®æ–‡å—コードを指定ã™ã‚‹ã€‚デフォルトã ã¨UTF-8ã«ãªã‚‹ã€‚</li>
+<li><kbd>-ol <var>ltype</var></kbd> : 改行文å—を変æ›ã™ã‚‹ã€‚`unix'(LF)ã€`dos'(CRLF)ã€`mac'(CR) ã®ã©ã‚Œã‹ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-cn</kbd> : 入力ã®æ–‡å—コードを自動判定ã—ã€ãã®åå‰ã‚’表示ã™ã‚‹ã€‚</li>
+<li><kbd>-wc</kbd> : 入力ã®æ–‡å—コードをUTF-8ã¨ä»®å®šã—ã€ãã®æ–‡å—数を表示ã™ã‚‹ã€‚</li>
+<li><kbd>-um</kbd> : UCS-2ã®æ–‡å—ã¨ã€C言語ã®æ–‡å—列表ç¾ã§ã®UTF-16BEã¨UTF-8ã®å¯¾å¿œè¡¨ã‚’出力ã™ã‚‹ã€‚</li>
+<li><kbd>-wf</kbd> : W3CDTFã®æ›¸å¼ã§å‡ºåŠ›ã™ã‚‹ã€‚</li>
+<li><kbd>-rf</kbd> : RFC 1123ã®æ›¸å¼ã§å‡ºåŠ›ã™ã‚‹ã€‚</li>
+<li><kbd>-utc</kbd> : å”定世界時を出力ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="villaapi" id="villaapi" class="head">Villa: 上級API</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Villaã¯QDBMã®ä¸Šç´šAPIã§ã‚ã‚Šã€B+木ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’管ç†ã™ã‚‹ãƒ«ãƒ¼ãƒãƒ³ã‚’æä¾›ã™ã‚‹ã€‚å„レコードã¯ãƒ¦ãƒ¼ã‚¶ãŒæŒ‡å®šã—ãŸé †åºã§æ•´åˆ—ã•ã‚Œã¦æ ¼ç´ã•ã‚Œã‚‹ã€‚ãƒãƒƒã‚·ãƒ¥ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã§ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æ¤œç´¢ã¯ã‚ーã®å®Œå…¨ä¸€è‡´ã«ã‚ˆã‚‹ã—ã‹ãªã‹ã£ãŸã€‚ã—ã‹ã—ã€Villaを用ã„ã‚‹ã¨ç¯„囲を指定ã—ã¦ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹ã“ã¨ãŒã§ãる。å„ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’é †ç•ªã«å‚ç…§ã™ã‚‹ã«ã¯ã‚«ãƒ¼ã‚½ãƒ«ã‚’用ã„る。データベースã«ã¯ã‚ーãŒé‡è¤‡ã™ã‚‹è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ã“ã¨ãŒã§ãる。ã¾ãŸã€ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³æ©Ÿæ§‹ã«ã‚ˆã£ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ“作を一括ã—ã¦åæ˜ ã•ã›ãŸã‚Šç ´æ£„ã—ãŸã‚Šã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>Villaã¯DepotãŠã‚ˆã³Cabinを基盤ã¨ã—ã¦å®Ÿè£…ã•ã‚Œã‚‹ã€‚Villaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã¯Depotã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ãã®ã‚‚ã®ã§ã‚る。レコードã®æ¤œç´¢ã‚„æ ¼ç´ã®å‡¦ç†é€Ÿåº¦ã¯Depotよりé…ã„ãŒã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã¯ã‚ˆã‚Šå°ã•ã„。</p>
+
+<p>Villaを使ã†ãŸã‚ã«ã¯ã€`depot.h' 㨠`cabin.h' 㨠`villa.h' 㨠`stdlib.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <cabin.h></kbd></dt>
+<dt><kbd>#include <villa.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>Villaã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†éš›ã«ã¯ã€`VILLA' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ã“ã‚Œã¯ã€`stdio.h' ã®å„種ルーãƒãƒ³ãŒãƒ•ã‚¡ã‚¤ãƒ«å…¥å‡ºåŠ›ã« `FILE' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’用ã„ã‚‹ã®ã«ä¼¼ã¦ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `vlopen' ã§é–‹ãã€é–¢æ•° `vlclose' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。データベースã«è‡´å‘½çš„ãªã‚¨ãƒ©ãƒ¼ãŒèµ·ããŸå ´åˆã¯ã€ä»¥å¾Œãã®ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã™ã‚‹ `vlclose' を除ãå…¨ã¦ã®æ“作ã¯ä½•ã‚‚ã›ãšã«ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã™ã€‚ã²ã¨ã¤ã®ãƒ—ãƒã‚»ã‚¹ã§è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’åŒæ™‚ã«åˆ©ç”¨ã™ã‚‹ã“ã¨ã¯å¯èƒ½ã§ã‚ã‚‹ãŒã€åŒã˜ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®è¤‡æ•°ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’利用ã—ã¦ã¯ãªã‚‰ãªã„。カーソルを使ã†å‰ã«ã¯ `vlcurfirst' ã‹ `vlcurlast' ã‹ `vlcurjump' ã®ã©ã‚Œã‹ã§åˆæœŸåŒ–ã™ã‚‹å¿…è¦ãŒã‚る。`vlcurput' 㨠`vlcurout' 以外ã®é–¢æ•°ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æ›´æ–°ã‚„削除ã‚
’ã—ãŸå¾Œã«ã‚‚カーソルをåˆæœŸåŒ–ã™ã‚‹å¿…è¦ãŒã‚る。</p>
+
+<p>Villaã§ã‚‚Depotã¨åŒã˜ã外部変数 `dpecode' ã«ç›´å‰ã®ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚エラーコードã«å¯¾å¿œã™ã‚‹ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸æ–‡å—列を得るã«ã¯ã€é–¢æ•° `dperrmsg' を用ã„る。</p>
+
+<h3>API</h3>
+
+<p>レコードã®é †ç•ªã‚’指定ã™ã‚‹ãŸã‚ã«ã¯ã€æ¯”較関数を定義ã™ã‚‹ã€‚比較関数ã«ã¯ä»¥ä¸‹ã®åž‹ã‚’用ã„る。</p>
+
+<dl>
+<dt><kbd>typedef int(*VLCFUNC)(const char *<var>aptr</var>, int <var>asiz</var>, const char *<var>bptr</var>, int <var>bsiz</var>);</kbd></dt>
+<dd>`aptr' ã¯ä¸€æ–¹ã®ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`asiz' ã¯ãã®ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚`bptr' ã¯ä»–æ–¹ã®ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`bsiz' ã¯ãã®ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯å‰è€…ãŒå¤§ãã‘ã‚Œã°æ£ã€å¾Œè€…ãŒå¤§ãã‘ã‚Œã°è² ã€ä¸¡è€…ãŒç‰ä¾¡ãªã‚‰ 0 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `vlopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>VILLA *vlopen(const char *<var>name</var>, int <var>omode</var>, VLCFUNC <var>cmp</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚`omode' ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã€`VL_OREADER' ãªã‚‰ãƒªãƒ¼ãƒ€ã€`VL_OWRITER' ãªã‚‰ãƒ©ã‚¤ã‚¿ã¨ãªã‚‹ã€‚`VL_OWRITER' ã®å ´åˆã€`VL_OCREAT' ã¾ãŸã¯ `VL_OTRUNC' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãる。`VL_OCREAT' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒç„¡ã„å ´åˆã«æ–°è¦ä½œæˆã™ã‚‹ã“ã¨ã‚’指示ã—ã€`VL_OTRUNC' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒå˜åœ¨ã—ã¦ã‚‚作り直ã™ã“ã¨ã‚’指示ã—ã€`VL_OZCOMP' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã®ãƒªãƒ¼ãƒ•ã‚’ZLIBã§åœ§ç¸®ã™ã‚‹ã“ã¨ã‚’指示ã—ã€`VL_OYCOMP' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã®ãƒªãƒ¼ãƒ•ã‚’LZOã§åœ§ç¸®ã™ã‚‹ã“ã¨ã‚’指示ã—ã€`VL_OXCOMP' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã®ãƒªãƒ¼ãƒ•ã‚’BZIP2ã§åœ§ç¸®ã™ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚`VL_OREADER' 㨠`VL_OWRITER' ã®ä¸¡æ–¹ã§ `VL_ONOLCK' ã¾ãŸã¯ `VL_OLCKNB' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€å‰è€…ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã“ã¨ã‚’指示ã—ã€å¾Œè€…ã¯ãƒ–ãƒãƒƒã‚¯ã›ãšã
«ãƒãƒƒã‚¯ã‚’ã‹ã‘ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚`cmp' ã¯æ¯”較関数を指定ã™ã‚‹ã€‚`VL_CMPLEX' ã¯ã‚ãƒ¼ã‚’è¾žæ›¸é †ã§æ¯”較ã™ã‚‹ã€‚`VL_CMPINT' ã¯ã‚ーを `int' åž‹ã®ã‚ªãƒ–ジェクトã¨ã¿ãªã—ã¦æ¯”較ã™ã‚‹ã€‚`VL_CMPNUM' ã¯ã‚ーをビッグエンディアンã®æ•°å€¤ã¨ã¿ãªã—ã¦æ¯”較ã™ã‚‹ã€‚`VL_CMPDEC' ã¯ã‚ーを10進数ã®æ•°å€¤ã‚’表ã™æ–‡å—列ã¨ã¿ãªã—ã¦æ¯”較ã™ã‚‹ã€‚`VLCFUNC' åž‹ã®å®£è¨€ã«åŸºã¥ã関数ã§ã‚ã‚Œã°æ¯”較関数ã¨ã—ã¦ç”¨ã„ã‚‹ã“ã¨ãŒã§ãる。åŒã˜ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ã¯å¸¸ã«åŒã˜æ¯”較関数を用ã„ã‚‹å¿…è¦ãŒã‚る。戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚‹ã‹ã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。ライタ（èªã¿æ›¸ã両用モード）ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’é–‹ãéš›ã«ã¯ãã®ãƒ•ã‚¡ã‚¤ãƒ«ã«å¯¾ã—ã¦æŽ’ä»–ãƒãƒƒã‚¯ãŒã‹ã‘られã€ãƒªãƒ¼ãƒ€ï¼ˆèªã¿è¾¼ã¿å°‚用モード）ã§é–‹ãéš›ã«ã¯å…±æœ‰ãƒãƒƒã‚¯ãŒã‹ã‘られる。ãã®éš›ã«ã¯è©²å½“ã®ãƒãƒƒã‚¯ãŒã‹ã‘られるã¾ã§
制御ãŒãƒ–ãƒãƒƒã‚¯ã™ã‚‹ã€‚`VL_OZCOMP' 㨠`VL_OYCOMP' 㨠`VL_OXCOMP' ã¯QDBMãŒãã‚Œãžã‚ŒZLIBã¨LZOã¨BZIP2を有効ã«ã—ã¦ãƒ“ルドã•ã‚ŒãŸå ´åˆã«ã®ã¿åˆ©ç”¨ã§ãる。`VL_ONOLCK' を使ã†å ´åˆã€ã‚¢ãƒ—リケーションãŒæŽ’他制御ã®è²¬ä»»ã‚’è² ã†ã€‚</dd>
+</dl>
+
+<p>データベースã¨ã®æŽ¥ç¶šã‚’é–‰ã˜ã¦ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlclose(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースã®æ›´æ–°å†…容ã¯ã€æŽ¥ç¶šã‚’é–‰ã˜ãŸæ™‚点ã§åˆã‚ã¦ãƒ•ã‚¡ã‚¤ãƒ«ã¨åŒæœŸã•ã‚Œã‚‹ã€‚ライタã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã„ãŸå ´åˆã€é©åˆ‡ã«æŽ¥ç¶šã‚’é–‰ã˜ãªã„ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒç ´å£Šã•ã‚Œã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã§ããªããªã‚‹ã€‚トランザクションãŒæœ‰åŠ¹ã§ã‚³ãƒŸãƒƒãƒˆã•ã‚Œã¦ã„ãªã„å ´åˆã€ãã‚Œã¯ç ´æ£„ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlput' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlput(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`dmode' ã¯ã‚ーãŒæ—¢å˜ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨é‡è¤‡ã—ãŸéš›ã®åˆ¶å¾¡ã‚’指定ã™ã‚‹ã€‚`VL_DOVER' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’上書ãã—ã€`VL_DKEEP' ã¯æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’残ã—ã¦ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã—ã€`VL_DCAT' ã¯æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’æ—¢å˜ã®å€¤ã®æœ«å°¾ã«åŠ ãˆã€`VL_DDUP' ã¯ã‚ーã®é‡è¤‡ã‚’許ã—ã¦æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’最後ã®å€¤ã¨ã—ã¦åŠ ãˆã€`VL_DDUPR' ã¯ã‚ーã®é‡è¤‡ã‚’許ã—ã¦æŒ‡å®šã•ã‚ŒãŸå€¤ã‚’最åˆã®å€¤ã¨ã—ã¦åŠ ãˆã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データã
ƒ™ãƒ¼ã‚¹ã®æ›´æ–°ã«ã‚ˆã£ã¦ã‚«ãƒ¼ã‚½ãƒ«ã¯ä½¿ç”¨ä¸èƒ½ã«ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>レコードを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlout' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlout(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚é‡è¤‡ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ç¾¤ã®ã‚ーãŒæŒ‡å®šã•ã‚ŒãŸå ´åˆã€ãã®æœ€åˆã®ã‚‚ã®ãŒå‰Šé™¤ã•ã‚Œã‚‹ã€‚データベースã®æ›´æ–°ã«ã‚ˆã£ã¦ã‚«ãƒ¼ã‚½ãƒ«ã¯ä½¿ç”¨ä¸èƒ½ã«ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>レコードをå–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlget' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *vlget(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚é‡è¤‡ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ç¾¤ã®ã‚ーãŒæŒ‡å®šã•ã‚ŒãŸå ´åˆã€ãã®æœ€åˆã®ã‚‚ã®ãŒé¸æŠžã•ã‚Œã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>レコードã®å€¤ã®ã‚µã‚¤ã‚ºã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlvsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlvsiz(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“レコードã®å€¤ã®ã‚µã‚¤ã‚ºã§ã‚ã‚‹ãŒã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯ -1 ã§ã‚る。複数ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒè©²å½“ã™ã‚‹å ´åˆã¯ã€æœ€åˆã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã®ã‚µã‚¤ã‚ºã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>ã‚ーã«ä¸€è‡´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æ•°ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlvnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlvnum(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯è©²å½“レコードã®å€¤ã®æ•°ã§ã‚ã‚Šã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯ 0 ã§ã‚る。</dd>
+</dl>
+
+<p>ã‚ーã«ä¸€è‡´ã™ã‚‹è¤‡æ•°ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlputlist' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlputlist(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const CBLIST *<var>vals</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`vals' ã¯å€¤ã®ãƒªã‚¹ãƒˆã®ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚リストã¯ç©ºã§ã‚ã£ã¦ã¯ãªã‚‰ãªã„。戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースã®æ›´æ–°ã«ã‚ˆã£ã¦ã‚«ãƒ¼ã‚½ãƒ«ã¯ä½¿ç”¨ä¸èƒ½ã«ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>ã‚ーã«ä¸€è‡´ã™ã‚‹å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `vloutlist' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vloutlist(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚データベースã®æ›´æ–°ã«ã‚ˆã£ã¦ã‚«ãƒ¼ã‚½ãƒ«ã¯ä½¿ç”¨ä¸èƒ½ã«ãªã‚‹ã€‚</dd>
+</dl>
+
+<p>ã‚ーã«ä¸€è‡´ã™ã‚‹å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlgetlist' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *vlgetlist(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚戻り値ã¯ä¸€è‡´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã®ãƒªã‚¹ãƒˆã®ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚‹ã‹ã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ã‚ーã«ä¸€è‡´ã™ã‚‹å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é€£çµã—ãŸå€¤ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlgetcat' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *vlgetcat(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰é€£çµã—ãŸå€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>カーソルを最åˆã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ç§»å‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `vlcurfirst' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurfirst(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースãŒç©ºã®å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>カーソルを最後ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ç§»å‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `vlcurlast' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurlast(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースãŒç©ºã®å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>カーソルをå‰ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ç§»å‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `vlcurprev' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurprev(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。å‰ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>カーソルを次ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ç§»å‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `vlcurnext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurnext(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。次ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>カーソルを特定ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å‰å¾Œã«ç§»å‹•ã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `vlcurjump' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurjump(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>jmode</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`kbuf' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`ksiz' ã¯ã‚ーã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(kbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`jmode' ã¯è©³ç´°ãªèª¿æ•´ã‚’指定ã™ã‚‹ã€‚`VL_JFORWARD' ã¯ã‚ーãŒåŒã˜ãƒ¬ã‚³ãƒ¼ãƒ‰ç¾¤ã®æœ€åˆã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ã‚«ãƒ¼ã‚½ãƒ«ãŒè¨å®šã•ã‚Œã€ã¾ãŸå®Œå…¨ã«ä¸€è‡´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯æ¬¡ã®å€™è£œã«ã‚«ãƒ¼ã‚½ãƒ«ãŒè¨å®šã•ã‚Œã‚‹ã“ã¨ã‚’æ„味ã™ã‚‹ã€‚`VL_JBACKWORD' ã¯ã‚ーãŒåŒã˜ãƒ¬ã‚³ãƒ¼ãƒ‰ç¾¤ã®æœ€å¾Œã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ã‚«ãƒ¼ã‚½ãƒ«ãŒè¨å®šã•ã‚Œã€ã¾ãŸå®Œå…¨ã«ä¸€è‡´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã¯å‰ã®å€™è£œã«ã‚«ãƒ¼ã‚½ãƒ«ãŒè¨å®šã•ã‚Œã‚‹ã“ã¨ã‚’æ„味ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。æ¡ä»¶ã«ä¸€è‡´ã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>カーソルã®ã‚るレコードã®ã‚ーをå–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlcurkey' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *vlcurkey(VILLA *<var>villa</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>カーソルã®ã‚るレコードã®å€¤ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlcurkey' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *vlcurval(VILLA *<var>villa</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`sp' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãã®å‚照先ã«æŠ½å‡ºã—ãŸé ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’æ ¼ç´ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰å€¤ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®é ˜åŸŸã¯ã€å®Ÿéš›ã«ã¯1ãƒã‚¤ãƒˆå¤šã確ä¿ã—ã¦çµ‚端文å—ãŒç½®ã‹ã‚Œã‚‹ã®ã§ã€æ–‡å—列ã¨ã—ã¦åˆ©ç”¨ã§ãる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>カーソルã®ã‚るレコードã®å‘¨è¾ºã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’挿入ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlcurput' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurput(VILLA *<var>villa</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>cpmode</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`vbuf' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’指定ã™ã‚‹ã€‚`vsiz' ã¯å€¤ã®ãƒ‡ãƒ¼ã‚¿é ˜åŸŸã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã‹ã€è² æ•°ãªã‚‰ `strlen(vbuf)' ã®å€¤ã¨ãªã‚‹ã€‚`cpmode' ã¯è©³ç´°ãªèª¿æ•´ã‚’指定ã™ã‚‹ã€‚`VL_CPCURRENT' ã¯ç¾åœ¨ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’上書ãã™ã‚‹ã“ã¨ã‚’指示ã—ã€`VL_CPBEFORE' ã¯ã‚«ãƒ¼ã‚½ãƒ«ã®ç›´å‰ã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’挿入ã™ã‚‹ã“ã¨ã‚’指示ã—ã€`VL_CPAFTER' ã¯ã‚«ãƒ¼ã‚½ãƒ«ã®ç›´å¾Œã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’挿入ã™ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚挿入æ“作ã®å¾Œã«ã¯ã€ã‚«ãƒ¼ã‚½ãƒ«ã¯æŒ¿å…¥ã•ã‚ŒãŸãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä½ç½®ã«ç§»å‹•ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>カーソルã®ã‚るレコードを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlcurout' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlcurout(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚削除æ“作ã®å¾Œã«ã¯ã€å¯èƒ½ã§ã‚ã‚Œã°ã‚«ãƒ¼ã‚½ãƒ«ã¯æ¬¡ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ç§»å‹•ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>性能を調整ã™ã‚‹ãƒ‘ラメータを指定ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlsettuning' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void vlsettuning(VILLA *<var>villa</var>, int <var>lrecmax</var>, int <var>nidxmax</var>, int <var>lcnum</var>, int <var>ncnum</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`lrecmax' ã¯B+木ã®ã²ã¨ã¤ã®ãƒªãƒ¼ãƒ•ã«å…¥ã‚Œã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。`nidxmax' ã¯B+木ã®éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã«å…¥ã‚Œã‚‹ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。`lcnum' ã¯ã‚ャッシュã«å…¥ã‚Œã‚‹ãƒªãƒ¼ãƒ•ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。`ncnum' ã¯ã‚ャッシュã«å…¥ã‚Œã‚‹éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ãŒã€0 以下ãªã‚‰ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆå€¤ãŒä½¿ã‚れる。デフォルトã®è¨å®šã¯ `vlsettuning(49, 192, 1024, 512)' ã«ç›¸å½“ã™ã‚‹ã€‚性能調整ã®ãƒ‘ラメータã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ä¿å˜ã•ã‚Œãªã„ã®ã§ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã度ã«æŒ‡å®šã™ã‚‹å¿…è¦ãŒã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºè¨å®šã™ã‚‹ã«ã¯ã€é–¢æ•° `vlsetfbpsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlsetfbpsiz(VILLA *<var>villa</var>, int <var>size</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`size' ã¯ãƒ•ãƒªãƒ¼ãƒ–ãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。フリーブãƒãƒƒã‚¯ãƒ—ールã®ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã®ã‚µã‚¤ã‚ºã¯256ã§ã‚る。サイズをより大ããã™ã‚‹ã¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸Šæ›¸ãã‚’ç¹°ã‚Šè¿”ã™éš›ã®ç©ºé–“効率ã¯ä¸ŠãŒã‚‹ãŒã€æ™‚間効率ãŒä¸‹ãŒã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースを更新ã—ãŸå†…容をファイルã¨ãƒ‡ãƒã‚¤ã‚¹ã«åŒæœŸã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `vlsync' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlsync(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ãªã„ã†ã¡ã«åˆ¥ãƒ—ãƒã‚»ã‚¹ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’利用ã•ã›ã‚‹å ´åˆã«å½¹ç«‹ã¤ã€‚トランザクションãŒæœ‰åŠ¹ãªé–“ã¯ã“ã®é–¢æ•°ã‚’使用ã™ã¹ãã§ã¯ãªã„。</dd>
+</dl>
+
+<p>データベースを最é©åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `vloptimize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vloptimize(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。レコードを削除ã—ãŸã‚Šã€ç½®æ›ãƒ¢ãƒ¼ãƒ‰ã‚„連çµãƒ¢ãƒ¼ãƒ‰ã§æ›¸ãè¾¼ã¿ã‚’ç¹°ã‚Šè¿”ã—ãŸã‚Šã™ã‚‹å ´åˆã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã«ä¸è¦ãªé ˜åŸŸãŒè“„ç©ã™ã‚‹ãŒã€ã“ã®é–¢æ•°ã¯ãれを解消ã™ã‚‹ã®ã«å½¹ç«‹ã¤ã€‚トランザクションãŒæœ‰åŠ¹ãªé–“ã¯ã“ã®é–¢æ•°ã‚’使用ã™ã¹ãã§ã¯ãªã„。</dd>
+</dl>
+
+<p>データベースã®åå‰ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `vlname' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *vlname(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰åå‰ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®ã‚µã‚¤ã‚ºã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `vlfsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlfsiz(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。I/Oãƒãƒƒãƒ•ã‚¡ã«ã‚ˆã‚Šã€æˆ»ã‚Šå€¤ã¯å®Ÿéš›ã®ã‚µã‚¤ã‚ºã‚ˆã‚Šå°ã•ããªã‚‹å ´åˆãŒã‚る。</dd>
+</dl>
+
+<p>B+木ã®ãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã®æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `vllnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vllnum(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã®æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>B+木ã®éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã®æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `vlnnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlnnum(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã®æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `vlrnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlrnum(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースãƒãƒ³ãƒ‰ãƒ«ãŒãƒ©ã‚¤ã‚¿ã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `vlwritable' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlwritable(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ©ã‚¤ã‚¿ãªã‚‰çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã«è‡´å‘½çš„エラーãŒèµ·ããŸã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `vlfatalerror' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlfatalerror(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯è‡´å‘½çš„エラーãŒã‚ã‚Œã°çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®inode番å·ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `vlinode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlinode(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®inode番å·ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®æœ€çµ‚更新時刻を得るã«ã¯ã€é–¢æ•° `vlmtime' を用ã„る。</p>
+
+<dl>
+<dt><kbd>time_t vlmtime(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æœ€çµ‚更新時刻ã§ã‚る。</dd>
+</dl>
+
+<p>トランザクションを開始ã™ã‚‹ã«ã¯ã€é–¢æ•° `vltranbegin' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vltranbegin(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒžãƒ«ãƒã‚¹ãƒ¬ãƒƒãƒ‰ã§ã®ç›¸äº’排他制御を行ã‚ãªã„ã®ã§ã€ã‚¢ãƒ—リケーションãŒãã®è²¬ä»»ã‚’è² ã†ã€‚一ã¤ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§åŒæ™‚ã«æœ‰åŠ¹ã«ã§ãるトランザクションã¯ä¸€ã¤ã ã‘ã§ã‚る。</dd>
+</dl>
+
+<p>トランザクションをコミットã™ã‚‹ã«ã¯ã€é–¢æ•° `vltrancommit' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vltrancommit(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。トランザクションã®ä¸ã§ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´æ–°ã¯ã‚³ãƒŸãƒƒãƒˆãŒæˆåŠŸã—ãŸæ™‚ã«ç¢ºå®šã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `vltranabort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vltranabort(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。トランザクションã®ä¸ã§ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´æ–°ã¯ã‚¢ãƒœãƒ¼ãƒˆã—ãŸæ™‚ã«ã¯ç ´æ£„ã•ã‚Œã‚‹ã€‚データベースã®çŠ¶æ…‹ã¯ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ã®å‰ã®çŠ¶æ…‹ã«ãƒãƒ¼ãƒ«ãƒãƒƒã‚¯ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースファイルを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlremove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>壊れãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’修復ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlrepair' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlrepair(const char *<var>name</var>, VLCFUNC <var>cmp</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚`cmp' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®æ¯”較関数を指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。修復ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ãŒå…ƒæ¥ã‚‚ã—ãã¯æœŸå¾…ã•ã‚Œã‚‹çŠ¶æ…‹ã«æˆ»ã‚‹ä¿è¨¼ã¯ãªã„。</dd>
+</dl>
+
+<p>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’エンディアンéžä¾å˜ã®ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦ãƒ€ãƒ³ãƒ—ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlexportdb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlexportdb(VILLA *<var>villa</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`name' ã¯å‡ºåŠ›ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>エンディアンéžä¾å˜ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ãƒãƒ¼ãƒ‰ã™ã‚‹ã«ã¯ã€é–¢æ•° `vlimportdb' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int vlimportdb(VILLA *<var>villa</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`villa' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚データベースã¯ç©ºã§ãªã‘ã‚Œã°ãªã‚‰ãªã„。`name' ã¯å…¥åŠ›ãƒ•ã‚¡ã‚¤ãƒ«ã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>åå‰ã¨å¯¾å¿œã•ã›ã¦é›»è©±ç•ªå·ã‚’æ ¼ç´ã—ã€ãれを検索ã™ã‚‹ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ VILLA *villa;
+ char *val;
+
+ /* データベースを開ã */
+ if(!(villa = vlopen(DBNAME, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ fprintf(stderr, "vlopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ */
+ if(!vlput(villa, NAME, -1, NUMBER, -1, VL_DOVER)){
+ fprintf(stderr, "vlput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* レコードをå–å¾—ã™ã‚‹ */
+ if(!(val = vlget(villa, NAME, -1, NULL))){
+ fprintf(stderr, "vlget: %s\n", dperrmsg(dpecode));
+ } else {
+ printf("Name: %s\n", NAME);
+ printf("Number: %s\n", val);
+ free(val);
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!vlclose(villa)){
+ fprintf(stderr, "vlclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>æ–‡å—列ã®å‰æ–¹ä¸€è‡´æ¤œç´¢ã‚’è¡Œã†ã‚¢ãƒ—リケーションã®ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#define DBNAME "words"
+#define PREFIX "apple"
+
+int main(int argc, char **argv){
+ VILLA *villa;
+ char *key, *val;
+
+ /* データベースを開ã */
+ if(!(villa = vlopen(DBNAME, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ fprintf(stderr, "vlopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’æ ¼ç´ã™ã‚‹ */
+ if(!vlput(villa, "applet", -1, "little application", -1, VL_DDUP) ||
+ !vlput(villa, "aurora", -1, "polar wonderwork", -1, VL_DDUP) ||
+ !vlput(villa, "apple", -1, "delicious fruit", -1, VL_DDUP) ||
+ !vlput(villa, "amigo", -1, "good friend", -1, VL_DDUP) ||
+ !vlput(villa, "apple", -1, "big city", -1, VL_DDUP)){
+ fprintf(stderr, "vlput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* カーソルを候補ã®å…ˆé ã«ç½®ã */
+ vlcurjump(villa, PREFIX, -1, VL_JFORWARD);
+
+ /* カーソルを走査ã™ã‚‹ */
+ while((key = vlcurkey(villa, NULL)) != NULL){
+ if(strstr(key, PREFIX) != key){
+ free(key);
+ break;
+ }
+ if(!(val = vlcurval(villa, NULL))){
+ fprintf(stderr, "vlcurval: %s\n", dperrmsg(dpecode));
+ free(key);
+ break;
+ }
+ printf("%s: %s\n", key, val);
+ free(val);
+ free(key);
+ vlcurnext(villa);
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!vlclose(villa)){
+ fprintf(stderr, "vlclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Villaを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹æ–¹æ³•ã¯ã€Depotã®å ´åˆã¨å…¨ãåŒã˜ã§ã‚る。</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>POSIXスレッドを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã€å¤–部変数 `dpecode' ã¯ã‚¹ãƒ¬ãƒƒãƒ‰å›ºæœ‰ãƒ‡ãƒ¼ã‚¿ã¸ã®å‚ç…§ã¨ã—ã¦æ‰±ã‚ã‚Œã€Villaã®å„関数ã¯ãƒªã‚¨ãƒ³ãƒˆãƒ©ãƒ³ãƒˆã«ãªã‚‹ã€‚ãã®å ´åˆã€ã‚¹ãƒ¬ãƒƒãƒ‰é–“ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<h3>Vista: 拡張上級API</h3>
+
+<p>Vistaã¯Villaã‚’æ‹¡å¼µã—ãŸAPIã§ã‚る。VillaãŒ2GB以上ã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚’扱ã†ã“ã¨ãŒã§ããªã„ã¨ã„ã†æ¬ 点を補ã†ãŸã‚ã«ã€Vistaã§ã¯Depotã§ã¯ãªãCuriaを用ã„ã¦å†…部データベースを管ç†ã™ã‚‹ã€‚Vistaã¯Villaã¨åŒã˜ãB+木ã®ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã¨ãã®æ“作をæä¾›ã™ã‚‹ãŒã€ãã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã§å®Ÿç¾ã•ã‚Œã‚‹ã€‚</p>
+
+<p>Vistaを使ã†ã«ã¯ã€`villa.h' ã®ä»£ã‚ã‚Šã« `vista.h' をインクルードã™ã‚Œã°ã‚ˆã„。Vistaã¯Villaã®ã‚·ãƒ³ãƒœãƒ«ã‚’マクãƒã§ã‚ªãƒ¼ãƒãƒ¼ãƒ©ã‚¤ãƒ‰ã—ã¦å®Ÿè£…ã•ã‚Œã¦ã„ã‚‹ãŸã‚ã€Villaã¨å…¨ãåŒæ§˜ã®APIã§åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãる。ã™ãªã‚ã¡ã€åŒæ–¹ã®ã‚·ã‚°ãƒãƒãƒ£ã¯å…¨ãåŒã˜ã§ã‚る。ãŸã ã—ã€ãã®å‰¯ä½œç”¨ã¨ã—ã¦ã€Vistaを使ã†ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«ï¼ˆã‚³ãƒ³ãƒ‘イルユニット）ã¯Villaを利用ã™ã‚‹ã“ã¨ãŒã§ããªã„（`villa.h' をインクルードã—ã¦ã¯ãªã‚‰ãªã„）。</p>
+
+<hr />
+
+<h2><a name="villacli" id="villacli" class="head">Villa用コマンド</a></h2>
+
+<p>Villaã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `vlmgr' ã¯Villaã‚„ãã®ã‚¢ãƒ—リケーションã®ãƒ‡ãƒãƒƒã‚°ã«å½¹ç«‹ã¤ãƒ„ールã§ã‚る。データベースを更新ã—ãŸã‚Šã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’調ã¹ãŸã‚Šã™ã‚‹æ©Ÿèƒ½ã‚’æŒã¤ã€‚シェルスクリプトã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚¢ãƒ—リケーションを作るã®ã«ã‚‚利用ã§ãる。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`key' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã€`val' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>vlmgr create [-cz|-cy|-cx] <var>name</var></kbd></dt>
+<dd>データベースファイルを作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat|-dup] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>ã‚ーã¨å€¤ã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr out [-l] [-kx|-ki] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr get [-nl] [-l] [-kx|-ki] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>ã‚ーã«å¯¾å¿œã™ã‚‹ãƒ¬ã‚³ãƒ¼ãƒ‰ã®å€¤ã‚’å–å¾—ã—ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr list [-nl] [-k|-v] [-kx|-ki] [-ox] [-top <var>key</var>] [-bot <var>key</var>] [-gt] [-lt] [-max <var>num</var>] [-desc] <var>name</var></kbd></dt>
+<dd>データベース内ã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ã‚ーã¨å€¤ã‚’タブã¨æ”¹è¡Œã§åŒºåˆ‡ã£ã¦æ¨™æº–出力ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr optimize <var>name</var></kbd></dt>
+<dd>データベースを最é©åŒ–ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr inform [-nl] <var>name</var></kbd></dt>
+<dd>データベースã®é›‘多ãªæƒ…å ±ã‚’å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr remove <var>name</var></kbd></dt>
+<dd>データベースファイルを削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr repair [-ki] <var>name</var></kbd></dt>
+<dd>壊れãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’修復ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr exportdb [-ki] <var>name</var> <var>file</var></kbd></dt>
+<dd>å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’エンディアンéžä¾å˜ã®ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦ãƒ€ãƒ³ãƒ—ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr importdb [-ki] <var>name</var> <var>file</var></kbd></dt>
+<dd>エンディアンéžä¾å˜ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ãƒãƒ¼ãƒ‰ã™ã‚‹ã€‚</dd>
+<dt><kbd>vlmgr version</kbd></dt>
+<dd>QDBMã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³æƒ…å ±ã‚’æ¨™æº–å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-cz</kbd> : データベースã®ãƒªãƒ¼ãƒ•ã‚’ZLIBã§åœ§ç¸®ã™ã‚‹ã€‚</li>
+<li><kbd>-cy</kbd> : データベースã®ãƒªãƒ¼ãƒ•ã‚’LZOã§åœ§ç¸®ã™ã‚‹ã€‚</li>
+<li><kbd>-cx</kbd> : データベースã®ãƒªãƒ¼ãƒ•ã‚’BZIP2ã§åœ§ç¸®ã™ã‚‹ã€‚</li>
+<li><kbd>-l</kbd> : ã‚ーã«ä¸€è‡´ã™ã‚‹å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’処ç†å¯¾è±¡ã¨ã™ã‚‹ã€‚</li>
+<li><kbd>-kx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-ki</kbd> : 10進数ã«ã‚ˆã‚‹æ•°å€¤è¡¨ç¾ã¨ã—㦠`key' を扱ã†ã€‚</li>
+<li><kbd>-vx</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vi</kbd> : 10進数ã«ã‚ˆã‚‹æ•°å€¤è¡¨ç¾ã¨ã—㦠`val' を扱ã†ã€‚</li>
+<li><kbd>-vf</kbd> : åå‰ãŒ `val' ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ‡ãƒ¼ã‚¿ã‚’値ã¨ã—ã¦èªã¿è¾¼ã‚€ã€‚</li>
+<li><kbd>-keep</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«ä¸Šæ›¸ãã›ãšã«ã‚¨ãƒ©ãƒ¼ã«ã™ã‚‹ã€‚</li>
+<li><kbd>-cat</kbd> : æ—¢å˜ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã‚ーãŒé‡è¤‡æ™‚ã«å€¤ã‚’末尾ã«è¿½åŠ ã™ã‚‹ã€‚</li>
+<li><kbd>-dup</kbd> : レコードã®ã‚ーãŒé‡è¤‡ã™ã‚‹ã®ã‚’許ã™ã€‚</li>
+<li><kbd>-nl</kbd> : ファイルãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã。</li>
+<li><kbd>-top <var>key</var></kbd> : リストã®æœ€å°ã®ã‚ーを指定ã™ã‚‹ã€‚</li>
+<li><kbd>-bot <var>key</var></kbd> : リストã®æœ€å¤§ã®ã‚ーを指定ã™ã‚‹ã€‚</li>
+<li><kbd>-gt</kbd> : 最å°ã®ã‚ーをリストã«å«ã‚ãªã„。</li>
+<li><kbd>-lt</kbd> : 最大ã®ã‚ーをリストã«å«ã‚ãªã„。</li>
+<li><kbd>-max <var>num</var></kbd> : リストã™ã‚‹æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-desc</kbd> : リストをé™é †ã§è¡Œã†ã€‚</li>
+<li><kbd>-ox</kbd> : 2æ¡å˜ä½ã®16進数ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªè¡¨ç¾ã¨ã—ã¦æ¨™æº–出力を行ã†ã€‚</li>
+<li><kbd>-n</kbd> : 標準出力ã®æœ«å°¾ã«ä»˜åŠ ã•ã‚Œã‚‹æ”¹è¡Œæ–‡å—ã®å‡ºåŠ›ã‚’抑制ã™ã‚‹ã€‚</li>
+<li><kbd>-k</kbd> : ã‚ーã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+<li><kbd>-v</kbd> : 値ã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `vltest' ã¯Villaã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã«ã‚ˆã£ã¦ç”Ÿæˆã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ `vlmgr' ã«ã‚ˆã£ã¦è§£æžã—ãŸã‚Šã€`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計ã£ãŸã‚Šã™ã‚‹ã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`rnum' ã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã€`pnum' ã¯ã‚ーã®ãƒ‘ターン数を指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>vltest write [-int] [-cz|-cy|-cx] [-tune <var>lrecmax</var> <var>nidxmax</var> <var>lcnum</var> <var>ncnum</var>] [-fbp <var>num</var>] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>`00000001'ã€`00000002' ã®ã‚ˆã†ã«å¤‰åŒ–ã™ã‚‹8ãƒã‚¤ãƒˆã®ã‚ーã¨é©å½“ãª8ãƒã‚¤ãƒˆã®å€¤ã‚’連続ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>vltest read [-int] [-vc] <var>name</var></kbd></dt>
+<dd>上記ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’検索ã™ã‚‹ã€‚</dd>
+<dt><kbd>vltest rdup [-int] [-cz|-cy|-cx] [-cc] [-tune <var>lrecmax</var> <var>nidxmax</var> <var>lcnum</var> <var>ncnum</var>] [-fbp <var>num</var>] <var>name</var> <var>rnum</var> <var>pnum</var></kbd></dt>
+<dd>ã‚ーãŒã‚る程度é‡è¤‡ã™ã‚‹ã‚ˆã†ã«ãƒ¬ã‚³ãƒ¼ãƒ‰ã®è¿½åŠ ã‚’è¡Œã„ã€é‡è¤‡ãƒ¢ãƒ¼ãƒ‰ã§å‡¦ç†ã™ã‚‹ã€‚</dd>
+<dt><kbd>vltest combo [-cz|-cy|-cx] <var>name</var></kbd></dt>
+<dd>å„種æ“作ã®çµ„ã¿åˆã‚ã›ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>vltest wicked [-cz|-cy|-cx] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>å„種更新æ“作を無作為ã«é¸æŠžã—ã¦å®Ÿè¡Œã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-int</kbd> : `int' åž‹ã®ã‚ªãƒ–ジェクトをã‚ーã¨å€¤ã«ç”¨ã„ã€æ¯”較関数もãã‚Œã«åˆã‚ã›ã‚‹ã€‚</li>
+<li><kbd>-cz</kbd> : データベースã®ãƒªãƒ¼ãƒ•ã‚’ZLIBã§åœ§ç¸®ã™ã‚‹ã€‚</li>
+<li><kbd>-cy</kbd> : データベースã®ãƒªãƒ¼ãƒ•ã‚’LZOã§åœ§ç¸®ã™ã‚‹ã€‚</li>
+<li><kbd>-cx</kbd> : データベースã®ãƒªãƒ¼ãƒ•ã‚’BZIP2ã§åœ§ç¸®ã™ã‚‹ã€‚</li>
+<li><kbd>-vc</kbd> : æ®ç™ºæ€§ã‚ャッシュをå‚ç…§ã™ã‚‹ã€‚</li>
+<li><kbd>-cc</kbd> : 連çµãƒ¢ãƒ¼ãƒ‰ã¨é‡è¤‡ãƒ¢ãƒ¼ãƒ‰ã‚’無作為ã«é¸æŠžã™ã‚‹ã€‚</li>
+<li><kbd>-tune <var>lrecmax</var> <var>nidxmax</var> <var>lcnum</var> <var>ncnum</var></kbd> : 性能パラメータを指定ã™ã‚‹ã€‚</li>
+<li><kbd>-fbp <var>num</var></kbd> : フリーブãƒãƒƒã‚¯ãƒ—ールã®ã‚µã‚¤ã‚ºã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-c</kbd> : Cabinã®ãƒžãƒƒãƒ—を使ã£ã¦æ¯”較テストを行ã†ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `vltsv' ã¯ã‚¿ãƒ–区切りã§ã‚ーã¨å€¤ã‚’表ç¾ã—ãŸè¡Œã‹ã‚‰ãªã‚‹TSVファイルã¨Villaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’相互変æ›ã™ã‚‹ã€‚ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯ã€QDBMã®ä»–ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã‚„ä»–ã®DBMã¨ã®é–“ã§ãƒ‡ãƒ¼ã‚¿ã®äº¤æ›ã‚’è¡Œã†éš›ã«å½¹ç«‹ã¤ã€‚ã¾ãŸã€ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã®é•ã†ã‚·ã‚¹ãƒ†ãƒ ã®é–“ã§ãƒ‡ãƒ¼ã‚¿ã‚’交æ›ã™ã‚‹éš›ã«ã‚‚役立ã¤ã€‚以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åを指定ã™ã‚‹ã€‚`export' サブコマンドã§ã¯TSVã®ãƒ‡ãƒ¼ã‚¿ã¯æ¨™æº–入力ã‹ã‚‰èªã¿è¾¼ã‚€ã€‚`import' サブコマンドã§ã¯TSVã®ãƒ‡ãƒ¼ã‚¿ãŒæ¨™æº–出力ã«æ›¸ã出ã•ã‚Œã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>vltsv import [-bin] <var>name</var></kbd></dt>
+<dd>TSVファイルをèªã¿è¾¼ã‚“ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>vltsv export [-bin] <var>name</var></kbd></dt>
+<dd>データベースã®å…¨ã¦ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’TSVファイルã¨ã—ã¦å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-bin</kbd> : Base64å½¢å¼ã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’扱ã†ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>Villaã®ã‚³ãƒžãƒ³ãƒ‰ç¾¤ã‚’駆使ã™ã‚‹ã¨ã€ç°¡å˜ãªãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚·ã‚¹ãƒ†ãƒ ãŒæ§‹ç¯‰ã§ãる。例ãˆã° `/etc/password' をユーザåã§æ¤œç´¢ã™ã‚‹ãŸã‚ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>cat /etc/passwd | tr ':' '\t' | vltsv import casket
+</pre>
+
+<p>ãã—ã¦ã€`mikio' ã¨ã„ã†ãƒ¦ãƒ¼ã‚¶ã®æƒ…å ±ã‚’å–り出ã™ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>vlmgr get casket mikio
+</pre>
+
+<p>ã“れらã®ã‚³ãƒžãƒ³ãƒ‰ã¨åŒç‰ã®æ©Ÿèƒ½ã‚’Villaã®APIを用ã„ã¦å®Ÿè£…ã™ã‚‹ã“ã¨ã‚‚容易ã§ã‚る。</p>
+
+<p>コマンド `qmttest' ã¯Depotã¨Curiaã¨Villaã®ãƒžãƒ«ãƒã‚¹ãƒ¬ãƒƒãƒ‰ã«ãŠã‘る安全性を調査ã™ã‚‹ã€‚ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯POSIXスレッドを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã«ã®ã¿ãƒžãƒ«ãƒã‚¹ãƒ¬ãƒƒãƒ‰ã§å‹•ä½œã™ã‚‹ã€‚以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã®æŽ¥é 辞を指定ã™ã‚‹ã€‚`rnum' ã¯å„データベースã«æ›¸ã込むレコード数を指定ã™ã‚‹ã€‚`tnum' ã¯ã‚¹ãƒ¬ãƒƒãƒ‰æ•°ã‚’指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>qmttest <var>name</var> <var>rnum</var> <var>tnum</var></kbd></dt>
+<dd>マルãƒã‚¹ãƒ¬ãƒƒãƒ‰ã«ãŠã‘る安全性を調査ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<hr />
+
+<h2><a name="odeumapi" id="odeumapi" class="head">Odeum: 転置API</a></h2>
+
+<h3>概è¦</h3>
+
+<p>Odeumã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’扱ã†APIã§ã‚る。転置インデックスã¨ã¯ã€æ¯é›†å›£ã®æ–‡æ›¸ç¾¤ã«å«ã¾ã‚Œã‚‹èªžã‚’抽出ã—ã¦ã€å„語をã‚ーã¨ã—ã€ãã®èªžã‚’å«ã‚€æ–‡æ›¸ã®ãƒªã‚¹ãƒˆã‚’検索ã™ã‚‹ãŸã‚ã®ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã§ã‚る。転置インデックスを用ã„ã‚‹ã¨ã€å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ を容易ã«å®Ÿç¾ã™ã‚‹ã“ã¨ãŒã§ãる。Odeumã¯æ–‡æ›¸ã‚’語や属性ã®é›†åˆã¨ã—ã¦æ‰±ã†ãŸã‚ã®æŠ½è±¡ãƒ‡ãƒ¼ã‚¿åž‹ã‚’æä¾›ã™ã‚‹ã€‚ãã‚Œã¯ã€å„アプリケーションãŒOdeumã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ–‡æ›¸ã‚’æ ¼ç´ã™ã‚‹éš›ã‚„ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‹ã‚‰æ–‡æ›¸ã‚’検索ã™ã‚‹éš›ã«ç”¨ã„られる。</p>
+
+<p>Odeumã¯å…ƒæ¥ã®æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ã‹ã‚‰ãƒ†ã‚ストを抽出ã™ã‚‹æ–¹æ³•ã¯æä¾›ã—ãªã„。ãã‚Œã¯ã‚¢ãƒ—リケーションãŒå®Ÿè£…ã™ã‚‹å¿…è¦ãŒã‚る。Odeumã¯ãƒ†ã‚ストを分解ã—ã¦èªžç¾¤ã‚’抽出ã™ã‚‹ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã‚’æä¾›ã™ã‚‹ãŒã€ãã‚Œã¯ç©ºç™½ã§èªžãŒåˆ†å‰²ã•ã‚Œã‚‹è‹±èªžãªã©ã®è¨€èªžã‚’指å‘ã—ã¦ã„ã‚‹ã€‚å½¢æ…‹ç´ è§£æžã‚„N-gram解æžãŒå¿…è¦ãªæ—¥æœ¬èªžãªã©ã®è¨€èªžã‚’扱ã†éš›ã‚„ã€ã‚¹ãƒ†ãƒŸãƒ³ã‚°ãªã©ã®ã‚ˆã‚Šé«˜åº¦ãªè‡ªç„¶è¨€èªžå‡¦ç†ã‚’è¡Œã†éš›ã«ã¯ã€ã‚¢ãƒ—リケーションã¯ç‹¬è‡ªã®è§£æžæ–¹æ³•ã‚’é©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãる。検索çµæžœã¯æ–‡æ›¸ã®ID番å·ã¨ãã®ã‚¹ã‚³ã‚¢ã‚’メンãƒã«æŒã¤æ§‹é€ 体をè¦ç´ ã¨ã™ã‚‹é…列ã¨ã—ã¦å¾—られる。複数ã®èªžã‚’用ã„ã¦æ¤œç´¢ã‚’è¡Œã†ãŸã‚ã«ã€Odeumã¯çµæžœã®é…列ã®é›†åˆæ¼”ç®—ã‚’è¡Œã†ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã‚’æä¾›ã™ã‚‹ã€‚</p>
+
+<p>Odeumã¯Curiaã¨Cabinã¨Villaを基盤ã¨ã—ã¦å®Ÿè£…ã•ã‚Œã‚‹ã€‚Odeumã§ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªåを指定ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’構築ã™ã‚‹ã€‚特定ã®ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®ç›´ä¸‹ã«Curiaã‚„Villaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’構築ã™ã‚‹ã€‚例ãˆã°ã€`casket' ã¨ã„ã†åå‰ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’作æˆã™ã‚‹å ´åˆã€`casket/docs'ã€`casket/index' ãŠã‚ˆã³ `casket/rdocs' ãŒç”Ÿæˆã•ã‚Œã‚‹ã€‚`docs' ã¯Curiaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã§ã‚る。ã‚ーã¯æ–‡æ›¸ã®ID番å·ã§ã‚ã‚Šã€å€¤ã¯URIç‰ã®æ–‡æ›¸å±žæ€§ã§ã‚る。`index' ã¯Curiaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã§ã‚る。ã‚ーã¯èªžã®æ£è¦å½¢ã§ã‚ã‚Šã€å€¤ã¯ãã®èªžã‚’å«ã‚€æ–‡æ›¸ã®ID番å·ã¨ãã®ã‚¹ã‚³ã‚¢ã‚’è¦ç´ ã¨ã™ã‚‹é…列ã§ã‚る。`rdocs' ã¯Villaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã§ã‚る。ã‚ーã¯æ–‡æ›¸ã®URIã§ã‚ã‚Šã€å€¤ã¯æ–‡æ›¸ã®ID番å·ã§ã‚る。</p>
+
+<p>Odeumを使ã†ãŸã‚ã«ã¯ã€`depot.h' 㨠`cabin.h' 㨠`odeum.h' 㨠`stdlib.h' をインクルードã™ã¹ãã§ã‚る。通常ã€ã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†’é 付近ã§ä»¥ä¸‹ã®è¨˜è¿°ã‚’è¡Œã†ã€‚</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <cabin.h></kbd></dt>
+<dt><kbd>#include <odeum.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>Odeumã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’扱ã†éš›ã«ã¯ã€`ODEUM' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `odopen' ã§é–‹ãã€é–¢æ•° `odclose' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。データベースã«è‡´å‘½çš„ãªã‚¨ãƒ©ãƒ¼ãŒèµ·ããŸå ´åˆã¯ã€ä»¥å¾Œãã®ãƒãƒ³ãƒ‰ãƒ«ã«å¯¾ã™ã‚‹ `odclose' を除ãå…¨ã¦ã®æ“作ã¯ä½•ã‚‚ã›ãšã«ã‚¨ãƒ©ãƒ¼ã‚’è¿”ã™ã€‚ã²ã¨ã¤ã®ãƒ—ãƒã‚»ã‚¹ã§è¤‡æ•°ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’åŒæ™‚ã«åˆ©ç”¨ã™ã‚‹ã“ã¨ã¯å¯èƒ½ã§ã‚ã‚‹ãŒã€åŒã˜ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®è¤‡æ•°ã®ãƒãƒ³ãƒ‰ãƒ«ã‚’利用ã—ã¦ã¯ãªã‚‰ãªã„。</p>
+
+<p>å„文書を扱ã†éš›ã«ã¯ã€`ODDOC' åž‹ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã‚’ãƒãƒ³ãƒ‰ãƒ«ã¨ã—ã¦ç”¨ã„る。ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `oddocopen' ã§é–‹ãã€é–¢æ•° `oddocclose' ã§é–‰ã˜ã‚‹ã€‚ãƒãƒ³ãƒ‰ãƒ«ã®ãƒ¡ãƒ³ãƒã‚’直接å‚ç…§ã™ã‚‹ã“ã¨ã¯æŽ¨å¥¨ã•ã‚Œãªã„。文書ã¯å±žæ€§ã¨èªžã®é›†åˆã‹ã‚‰ãªã‚‹ã€‚語ã¯æ£è¦å½¢ã¨å‡ºç¾å½¢ã®ãƒšã‚¢ã¨ã—ã¦è¡¨ç¾ã•ã‚Œã‚‹ã€‚</p>
+
+<p>Odeumã§ã‚‚Depotã¨åŒã˜ã外部変数 `dpecode' ã«ç›´å‰ã®ã‚¨ãƒ©ãƒ¼ã‚³ãƒ¼ãƒ‰ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚エラーコードã«å¯¾å¿œã™ã‚‹ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸æ–‡å—列を得るã«ã¯ã€é–¢æ•° `dperrmsg' を用ã„る。</p>
+
+<h3>API</h3>
+
+<p>検索çµæžœã‚’扱ã†ãŸã‚ã«ã¯ã€`ODPAIR' åž‹ã®æ§‹é€ 体を用ã„る。</p>
+
+<dl>
+<dt><kbd>typedef struct { int id; int score; } ODPAIR;</kbd></dt>
+<dd>`id' ã¯æ–‡æ›¸ã®ID番å·ã§ã‚る。`score' ã¯æ–‡æ›¸ã«å«ã¾ã‚Œã‚‹æ¤œç´¢èªžã®æ•°ã‚’å…ƒã«ç®—出ã•ã‚Œã‚‹ã‚¹ã‚³ã‚¢ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `odopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODEUM *odopen(const char *<var>name</var>, int <var>omode</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚`omode' ã¯æŽ¥ç¶šãƒ¢ãƒ¼ãƒ‰ã‚’指定ã—ã€`OD_OREADER' ãªã‚‰ãƒªãƒ¼ãƒ€ã€`OD_OWRITER' ãªã‚‰ãƒ©ã‚¤ã‚¿ã¨ãªã‚‹ã€‚`OD_OWRITER' ã®å ´åˆã€`OD_OCREAT' ã¾ãŸã¯ `OD_OTRUNC' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãる。`OD_OCREAT' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒç„¡ã„å ´åˆã«æ–°è¦ä½œæˆã™ã‚‹ã“ã¨ã‚’指示ã—ã€`OD_OTRUNC' ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãŒå˜åœ¨ã—ã¦ã‚‚作り直ã™ã“ã¨ã‚’指示ã™ã‚‹ã€‚`OD_OREADER' 㨠`OD_OWRITER' ã®ä¸¡æ–¹ã§ `OD_ONOLCK' ã‹ `OD_OLCKNB' ã¨ã®ãƒ“ット論ç†å’Œã«ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€å‰è€…ã¯ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ã‚’ã‹ã‘ãšã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ãã“ã¨ã‚’指示ã—ã€å¾Œè€…ã¯ãƒ–ãƒãƒƒã‚¯ã›ãšã«ãƒãƒƒã‚¯ã‚’ã‹ã‘ã‚‹ã“ã¨ã‚’指示ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚‹ã‹ã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。ライタ（èªã¿æ›¸ã両用モード）ã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’é–‹ãéš›ã«ã¯ãã®ãƒ•ã‚¡ã‚¤ãƒ«ã«å¯¾ã—ã¦æŽ’ä»–ãƒãƒ
ƒã‚¯ãŒã‹ã‘られã€ãƒªãƒ¼ãƒ€ï¼ˆèªã¿è¾¼ã¿å°‚用モード）ã§é–‹ãéš›ã«ã¯å…±æœ‰ãƒãƒƒã‚¯ãŒã‹ã‘られる。ãã®éš›ã«ã¯è©²å½“ã®ãƒãƒƒã‚¯ãŒã‹ã‘られるã¾ã§åˆ¶å¾¡ãŒãƒ–ãƒãƒƒã‚¯ã™ã‚‹ã€‚`OD_ONOLCK' を使ã†å ´åˆã€ã‚¢ãƒ—リケーションãŒæŽ’他制御ã®è²¬ä»»ã‚’è² ã†ã€‚</dd>
+</dl>
+
+<p>データベースã¨ã®æŽ¥ç¶šã‚’é–‰ã˜ã¦ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `odclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odclose(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースã®æ›´æ–°å†…容ã¯ã€æŽ¥ç¶šã‚’é–‰ã˜ãŸæ™‚点ã§åˆã‚ã¦ãƒ•ã‚¡ã‚¤ãƒ«ã¨åŒæœŸã•ã‚Œã‚‹ã€‚ライタã§ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‹ã„ãŸå ´åˆã€é©åˆ‡ã«æŽ¥ç¶šã‚’é–‰ã˜ãªã„ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒç ´å£Šã•ã‚Œã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>æ–‡æ›¸ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `odput' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odput(ODEUM *<var>odeum</var>, const ODDOC *<var>doc</var>, int <var>wmax</var>, int <var>over</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`wmax' ã¯æ–‡æ›¸ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã™ã‚‹èªžã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ãŒã€è² æ•°ãªã‚‰ç„¡åˆ¶é™ã¨ãªã‚‹ã€‚`over' ã¯é‡è¤‡ã—ãŸæ–‡æ›¸ã®ä¸Šæ›¸ãã‚’è¡Œã†ã‹å¦ã‹ã‚’指定ã™ã‚‹ã€‚ãã‚ŒãŒå½ã§æ–‡æ›¸ã®URIãŒé‡è¤‡ã—ãŸå ´åˆã¯ã‚¨ãƒ©ãƒ¼ã¨ãªã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。</dd>
+</dl>
+
+<p>URIã§æŒ‡å®šã—ãŸæ–‡æ›¸ã‚’削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `odout' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odout(ODEUM *<var>odeum</var>, const char *<var>uri</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`uri' ã¯æ–‡æ›¸ã®URIã®æ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã™ã‚‹æ–‡æ›¸ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>ID番å·ã§æŒ‡å®šã—ãŸæ–‡æ›¸ã‚’削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `odoutbyid' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odoutbyid(ODEUM *<var>odeum</var>, int <var>id</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`id' ã¯æ–‡æ›¸ã®ID番å·ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。該当ã™ã‚‹æ–‡æ›¸ãŒãªã„å ´åˆã‚‚å½ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>URIã§æŒ‡å®šã—ãŸæ–‡æ›¸ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `odget' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODDOC *odget(ODEUM *<var>odeum</var>, const char *<var>uri</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`uri' ã¯æ–‡æ›¸ã®URIã®æ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰è©²å½“ã®æ–‡æ›¸ã®ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®æ–‡æ›¸ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `oddocopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `oddocclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>ID番å·ã§æŒ‡å®šã—ãŸæ–‡æ›¸ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `odgetbyid' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODDOC *odgetbyid(ODEUM *<var>odeum</var>, int <var>id</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`id' ã¯æ–‡æ›¸ã®ID番å·ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰è©²å½“ã®æ–‡æ›¸ã®ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。該当ã®æ–‡æ›¸ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `oddocopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `oddocclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>URIã§æŒ‡å®šã—ãŸæ–‡æ›¸ã®IDã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `odgetidbyuri' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odgetidbyuri(ODEUM *<var>odeum</var>, const char *<var>uri</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`uri' ã¯æ–‡æ›¸ã®URIã®æ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰è©²å½“ã®æ–‡æ›¸ã®IDã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。該当ã®æ–‡æ›¸ãŒãªã„å ´åˆã‚‚ -1 ã‚’è¿”ã™ã€‚</dd>
+</dl>
+
+<p>ID番å·ã§æŒ‡å®šã—ãŸæ–‡æ›¸ãŒå˜åœ¨ã—ã¦ã„ã‚‹ã‹èª¿ã¹ã‚‹ã«ã¯ã€é–¢æ•° `odcheck' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odcheck(ODEUM *<var>odeum</var>, int <var>id</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`id' ã¯æ–‡æ›¸ã®ID番å·ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–‡æ›¸ãŒå˜åœ¨ã™ã‚Œã°çœŸã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>転置インデックスを検索ã—ã¦ç‰¹å®šã®èªžã‚’å«ã‚€æ–‡æ›¸ç¾¤ã‚’知るã«ã¯ã€é–¢æ•° `odsearch' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODPAIR *odsearch(ODEUM *<var>odeum</var>, const char *<var>word</var>, int <var>max</var>, int *<var>np</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`word' ã¯æ¤œç´¢èªžã‚’指定ã™ã‚‹ã€‚`max' ã¯å–り出ã™æ–‡æ›¸ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚`np' ã®å‚照先ã«ã¯ã€æˆ»ã‚Šå€¤ã®é…列ã®è¦ç´ æ•°ãŒæ ¼ç´ã•ã‚Œã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰é…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。ãã®é…列ã®å„è¦ç´ ã¯æ–‡æ›¸ã®ID番å·ã¨ã‚¹ã‚³ã‚¢ã®ãƒšã‚¢ã§ã‚ã‚Šã€ã‚¹ã‚³ã‚¢ã®é™é †ã§ä¸¦ã¹ã‚‰ã‚Œã‚‹ã€‚検索語ã«è©²å½“ã™ã‚‹æ–‡æ›¸ãŒä¸€ã¤ã‚‚ãªã‹ã£ãŸã¨ã—ã¦ã‚‚エラーã«ã¯ãªã‚‰ãšã«ã€ç©ºã®é…列を返ã™ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。é…列ã®å„è¦ç´ ã«ã¯æ—¢ã«å‰Šé™¤ã•ã‚ŒãŸæ–‡æ›¸ã®ãƒ‡ãƒ¼ã‚¿ã‚‚å«ã¾ã‚Œã‚‹ã“ã¨ã«æ³¨æ„ã™ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>特定ã®èªžã‚’å«ã‚€æ–‡æ›¸ã®æ•°ã‚’知るã«ã¯ã€é–¢æ•° `odsearchdnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odsearchdnum(ODEUM *<var>odeum</var>, const char *<var>word</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`word' ã¯æ¤œç´¢èªžã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰æ¤œç´¢èªžã‚’å«ã‚€æ–‡æ›¸ã®æ•°ã§ã‚ã‚Šã€è©²å½“ãŒãªã„å ´åˆã‚„エラーã®å ´åˆã¯ -1 ã§ã‚る。ã“ã®é–¢æ•°ã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®å®Ÿãƒ‡ãƒ¼ã‚¿ã‚’èªã¿è¾¼ã¾ãªã„ã®ã§åŠ¹çŽ‡ãŒã‚ˆã„。</dd>
+</dl>
+
+<p>データベースã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‚’åˆæœŸåŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `oditerinit' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int oditerinit(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。イテレータã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚ŒãŸå…¨ã¦ã®æ–‡æ›¸ã‚’å‚ç…§ã™ã‚‹ãŸã‚ã«ç”¨ã„られる。</dd>
+</dl>
+
+<p>データベースã®ã‚¤ãƒ†ãƒ¬ãƒ¼ã‚¿ã‹ã‚‰æ¬¡ã®æ–‡æ›¸ã‚’å–り出ã™ã«ã¯ã€é–¢æ•° `oditernext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODDOC *oditernext(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。イテレータãŒæœ€å¾Œã¾ã§ãã¦è©²å½“ã®æ–‡æ›¸ãŒãªã„å ´åˆã‚‚ `NULL' ã‚’è¿”ã™ã€‚ã“ã®é–¢æ•°ã‚’ç¹°ã‚Šè¿”ã—ã¦å‘¼ã¶ã“ã¨ã«ã‚ˆã£ã¦å…¨ã¦ã®æ–‡æ›¸ã‚’一度ãšã¤å‚ç…§ã™ã‚‹ã“ã¨ãŒã§ãる。ãŸã ã—ã€ç¹°ã‚Šè¿”ã—ã®é–“ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æ›´æ–°ãŒã‚ã£ãŸå ´åˆã¯ãã®é™ã‚Šã§ã¯ãªã„。ãªãŠã€å–り出ã™ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é †åºã¯åˆ¶å¾¡ã§ããšã€æ ¼ç´ã—ãŸé †ç•ªã§ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’å–り出ã›ã‚‹ã¨ã¯é™ã‚‰ãªã„。戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯ã€é–¢æ•° `oddocopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `oddocclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>データベースを更新ã—ãŸå†…容をファイルã¨ãƒ‡ãƒã‚¤ã‚¹ã«åŒæœŸã•ã›ã‚‹ã«ã¯ã€é–¢æ•° `odsync' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odsync(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。ã“ã®é–¢æ•°ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ãªã„ã†ã¡ã«åˆ¥ãƒ—ãƒã‚»ã‚¹ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’利用ã•ã›ã‚‹å ´åˆã«å½¹ç«‹ã¤ã€‚</dd>
+</dl>
+
+<p>データベースを最é©åŒ–ã™ã‚‹ã«ã¯ã€é–¢æ•° `odoptimize' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odoptimize(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ©ã‚¤ã‚¿ã§æŽ¥ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。転置インデックスã«ãŠã‘る削除ã•ã‚ŒãŸæ–‡æ›¸ã®è¦ç´ ã¯å‰Šé™¤ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースã®åå‰ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odname' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *odname(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰åå‰ã‚’æ ¼ç´ã—ãŸé ˜åŸŸã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>データベースファイルã®ã‚µã‚¤ã‚ºã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odfsiz' を用ã„る。</p>
+
+<dl>
+<dt><kbd>double odfsiz(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1.0 ã§ã‚る。</dd>
+</dl>
+
+<p>転置インデックス内ã®ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odbnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odbnum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>転置インデックス内ã®ãƒã‚±ãƒƒãƒˆé…列ã®åˆ©ç”¨æ¸ˆã¿ã®è¦ç´ æ•°ã®åˆè¨ˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odbusenum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odbusenum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒã‚±ãƒƒãƒˆé…列ã®åˆ©ç”¨æ¸ˆã¿ã®è¦ç´ æ•°ã®åˆè¨ˆã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+
+<p>データベースã«æ ¼ç´ã•ã‚ŒãŸæ–‡æ›¸æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int oddnum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚ŒãŸæ–‡æ›¸ã®æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã«æ ¼ç´ã•ã‚ŒãŸå˜èªžæ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odwnum' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odwnum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚ŒãŸèªžã®æ•°ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ -1 ã§ã‚る。I/Oãƒãƒƒãƒ•ã‚¡ã«ã‚ˆã‚Šã€æˆ»ã‚Šå€¤ã¯å®Ÿéš›ã®ã‚µã‚¤ã‚ºã‚ˆã‚Šå°ã•ããªã‚‹å ´åˆãŒã‚る。</dd>
+</dl>
+
+<p>データベースãƒãƒ³ãƒ‰ãƒ«ãŒãƒ©ã‚¤ã‚¿ã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `odwritable' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odwritable(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ©ã‚¤ã‚¿ãªã‚‰çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã«è‡´å‘½çš„エラーãŒèµ·ããŸã‹ã©ã†ã‹ã‚’調ã¹ã‚‹ã«ã¯ã€é–¢æ•° `odfatalerror' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odfatalerror(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯è‡´å‘½çš„エラーãŒã‚ã‚Œã°çœŸã§ã‚ã‚Šã€ãã†ã§ãªã‘ã‚Œã°å½ã§ã‚る。</dd>
+</dl>
+
+<p>データベースディレクトリã®inode番å·ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odinode' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odinode(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®inode番å·ã§ã‚る。</dd>
+</dl>
+
+<p>データベースã®æœ€çµ‚更新時刻を得るã«ã¯ã€é–¢æ•° `odmtime' を用ã„る。</p>
+
+<dl>
+<dt><kbd>time_t odmtime(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®æœ€çµ‚更新時刻ã§ã‚る。</dd>
+</dl>
+
+<p>複数ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’マージã™ã‚‹ã«ã¯ã€é–¢æ•° `odmerge' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odmerge(const char *<var>name</var>, const CBLIST *<var>elemnames</var>);</kbd></dt>
+<dd>`name' ã¯ä½œæˆã™ã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚`elemnames' ã¯è¦ç´ データベースã®åå‰ã®ãƒªã‚¹ãƒˆã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。åŒã˜URLã‚’æŒã¤è¤‡æ•°ã®æ–‡æ›¸ãŒç¾ã‚ŒãŸå ´åˆã€æœ€åˆã«ç¾ã‚ŒãŸã‚‚ã®ãŒæŽ¡ç”¨ã•ã‚Œã€æ®‹ã‚Šã¯ç„¡è¦–ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>データベースディレクトリを削除ã™ã‚‹ã«ã¯ã€é–¢æ•° `odremove' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int odremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®åå‰ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰çœŸã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰å½ã§ã‚る。データベースディレクトリã®ä¸ã«ã¯QDBMã®ä»–ã®APIã«ã‚ˆã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’ç½®ãã“ã¨ãŒã§ãã‚‹ãŒã€ãれらもã“ã®é–¢æ•°ã«ã‚ˆã£ã¦å‰Šé™¤ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>文書ãƒãƒ³ãƒ‰ãƒ«ã‚’作æˆã™ã‚‹ã«ã¯ã€é–¢æ•° `oddocopen' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODDOC *oddocopen(const char *<var>uri</var>);</kbd></dt>
+<dd>`uri' ã¯æ–‡æ›¸ã®URIを指定ã™ã‚‹ã€‚戻り値ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。新ã—ã„文書ã®ID番å·ã¯å®šç¾©ã•ã‚Œãªã„。ãã‚Œã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã—ãŸéš›ã«å®šç¾©ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>文書ãƒãƒ³ãƒ‰ãƒ«ã‚’ç ´æ£„ã™ã‚‹ã«ã¯ã€é–¢æ•° `oddocclose' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void oddocclose(ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚é–‰ã˜ãŸãƒãƒ³ãƒ‰ãƒ«ã®é ˜åŸŸã¯è§£æ”¾ã•ã‚Œã‚‹ã®ã§ã€ä»¥å¾Œã¯åˆ©ç”¨ã§ããªããªã‚‹ã€‚</dd>
+</dl>
+
+<p>文書ã«å±žæ€§ã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `oddocaddattr' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void oddocaddattr(ODDOC *<var>doc</var>, const char *<var>name</var>, const char *<var>value</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`name' ã¯å±žæ€§åã®æ–‡å—列を指定ã™ã‚‹ã€‚`value' ã¯å±žæ€§å€¤ã®æ–‡å—列を指定ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>文書ã«èªžã‚’è¿½åŠ ã™ã‚‹ã«ã¯ã€é–¢æ•° `oddocaddword' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void oddocaddword(ODDOC *<var>doc</var>, const char *<var>normal</var>, const char *<var>asis</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`normal' ã¯èªžã®æ£è¦å½¢ã®æ–‡å—列を指定ã™ã‚‹ã€‚æ£è¦å½¢ã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®ã‚ーã¨ã—ã¦æ‰±ã‚れる。æ£è¦å½¢ãŒç©ºæ–‡å—列ã®å ´åˆã€ãã®èªžã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã«åæ˜ ã•ã‚Œãªã„。`asis' ã¯èªžã®å‡ºç¾å½¢ã®æ–‡å—列を指定ã™ã‚‹ã€‚出ç¾å½¢ã¯ã‚¢ãƒ—リケーションãŒæ–‡æ›¸ã‚’å–å¾—ã—ãŸéš›ã«åˆ©ç”¨ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>文書ã®IDã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddocid' を用ã„る。</p>
+
+<dl>
+<dt><kbd>int oddocid(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–‡æ›¸ã®ID番å·ã§ã‚る。</dd>
+</dl>
+
+<p>文書ã®URIã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddocuri' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *oddocuri(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ–‡æ›¸ã®URIã®æ–‡å—列ã§ã‚る。</dd>
+</dl>
+
+<p>文書ã®å±žæ€§å€¤ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddocgetattr' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const char *oddocgetattr(const ODDOC *<var>doc</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`name' ã¯å±žæ€§åã®æ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯å±žæ€§å€¤ã®æ–‡å—列ã§ã‚ã‚‹ã‹ã€è©²å½“ãŒãªã‘れ㰠`NULL' ã§ã‚る。</dd>
+</dl>
+
+<p>文書内ã®èªžç¾¤ã®æ£è¦å½¢ã®ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddocnwords' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const CBLIST *oddocnwords(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£è¦å½¢ã®èªžç¾¤ã‚’æ ¼ç´ã—ãŸãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>文書内ã®èªžç¾¤ã®å‡ºç¾å½¢ã®ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddocawords' を用ã„る。</p>
+
+<dl>
+<dt><kbd>const CBLIST *oddocawords(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯å‡ºç¾å½¢ã®èªžç¾¤ã‚’æ ¼ç´ã—ãŸãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。</dd>
+</dl>
+
+<p>文書ã®ã‚ーワードã®æ£è¦å½¢ã¨ãã®ã‚¹ã‚³ã‚¢ã®ãƒžãƒƒãƒ—ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `oddocscores' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBMAP *oddocscores(const ODDOC *<var>doc</var>, int <var>max</var>, ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`doc' ã¯æ–‡æ›¸ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`max' ã¯å–å¾—ã™ã‚‹ã‚ーワードã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚`odeum' ㌠`NULL' ã§ãªã‘ã‚Œã°ã€ãれを用ã„ã¦é‡ã¿ã¥ã‘ã®ãŸã‚ã®IDFãŒç®—出ã•ã‚Œã‚‹ã€‚戻り値ã¯ã‚ーワードã¨ãã®ã‚¹ã‚³ã‚¢ã‚’æ ¼ç´ã—ãŸãƒžãƒƒãƒ—ãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。スコアã¯10進数ã®æ–‡å—列ã§è¡¨ç¾ã•ã‚Œã‚‹ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cbmapopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cbmapclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>テã‚ストを分解ã—ã¦èªžç¾¤ã®å‡ºç¾å½¢ã®ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odbreaktext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>CBLIST *odbreaktext(const char *<var>text</var>);</kbd></dt>
+<dd>`text' ã¯ãƒ†ã‚ストã®æ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯èªžç¾¤ã®å‡ºç¾å½¢ã®ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã§ã‚る。語群ã¯ç©ºç™½æ–‡å—ã¨ãƒ”リオドやコンマç‰ã®åŒºåˆ‡ã‚Šæ–‡å—ã§åˆ†å‰²ã•ã‚Œã‚‹ã€‚戻り値ã®ãƒãƒ³ãƒ‰ãƒ«ã¯é–¢æ•° `cblistopen' ã§é–‹ã‹ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `cblistclose' ã§é–‰ã˜ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>語ã®æ£è¦å½¢ã‚’å–å¾—ã™ã‚‹ã«ã¯ã€é–¢æ•° `odnormalizeword' を用ã„る。</p>
+
+<dl>
+<dt><kbd>char *odnormalizeword(const char *<var>asis</var>);</kbd></dt>
+<dd>`asis' ã¯èªžã®å‡ºç¾å½¢ã®æ–‡å—列を指定ã™ã‚‹ã€‚戻り値ã¯èªžã®æ£è¦å½¢ã®æ–‡å—列ã§ã‚る。ASCIIコードã®ã‚¢ãƒ«ãƒ•ã‚¡ãƒ™ãƒƒãƒˆã¯å°æ–‡å—ã«çµ±ä¸€ã•ã‚Œã‚‹ã€‚区切り文å—ã®ã¿ã‹ã‚‰ãªã‚‹æ–‡å—ã¯ç©ºæ–‡å—列ã¨ã—ã¦æ‰±ã‚れる。戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>二ã¤ã®æ–‡æ›¸é›†åˆã‹ã‚‰ãã®å…±é€šé›†åˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odpairsand' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODPAIR *odpairsand(ODPAIR *<var>apairs</var>, int <var>anum</var>, ODPAIR *<var>bpairs</var>, int <var>bnum</var>, int *<var>np</var>);</kbd></dt>
+<dd>`apairs' ã¯å‰è€…ã®æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`anum' ã¯å‰è€…ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`apairs' ã¯å¾Œè€…ã®æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`anum' ã¯å¾Œè€…ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`np' ã®å‚照先ã«ã¯ã€æˆ»ã‚Šå€¤ã®é…列ã®è¦ç´ æ•°ãŒæ ¼ç´ã•ã‚Œã‚‹ã€‚戻り値ã¯æ–°ã—ã„文書集åˆã®é…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€å„è¦ç´ ã¯äºŒã¤ã®é›†åˆã«å…±é€šã—ã¦å±žã™ã‚‹ã‚‚ã®ã§ã‚る。å„è¦ç´ ã¯ã‚¹ã‚³ã‚¢ã®é™é †ã§ä¸¦ã¹ã‚‰ã‚Œã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>二ã¤ã®æ–‡æ›¸é›†åˆã‹ã‚‰ãã®å’Œé›†åˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odpairsor' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODPAIR *odpairsor(ODPAIR *<var>apairs</var>, int <var>anum</var>, ODPAIR *<var>bpairs</var>, int <var>bnum</var>, int *<var>np</var>);</kbd></dt>
+<dd>`apairs' ã¯å‰è€…ã®æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`anum' ã¯å‰è€…ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`apairs' ã¯å¾Œè€…ã®æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`anum' ã¯å¾Œè€…ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`np' ã®å‚照先ã«ã¯ã€æˆ»ã‚Šå€¤ã®é…列ã®è¦ç´ æ•°ãŒæ ¼ç´ã•ã‚Œã‚‹ã€‚戻り値ã¯æ–°ã—ã„文書集åˆã®é…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€å„è¦ç´ ã¯äºŒã¤ã®é›†åˆã®ä¸¡æ–¹ã‚ã‚‹ã„ã¯ã©ã¡ã‚‰ã‹ä¸€æ–¹ã«å±žã™ã‚‹ã‚‚ã®ã§ã‚る。å„è¦ç´ ã¯ã‚¹ã‚³ã‚¢ã®é™é †ã§ä¸¦ã¹ã‚‰ã‚Œã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>二ã¤ã®æ–‡æ›¸é›†åˆã‹ã‚‰ãã®å·®é›†åˆã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odpairsnotand' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODPAIR *odpairsnotand(ODPAIR *<var>apairs</var>, int <var>anum</var>, ODPAIR *<var>bpairs</var>, int <var>bnum</var>, int *<var>np</var>);</kbd></dt>
+<dd>`apairs' ã¯å‰è€…ã®æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`anum' ã¯å‰è€…ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`apairs' ã¯å¾Œè€…ã®æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`anum' ã¯å¾Œè€…ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚`np' ã®å‚照先ã«ã¯ã€æˆ»ã‚Šå€¤ã®é…列ã®è¦ç´ æ•°ãŒæ ¼ç´ã•ã‚Œã‚‹ã€‚戻り値ã¯æ–°ã—ã„文書集åˆã®é…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€å„è¦ç´ ã¯å‰è€…ã®é›†åˆã«ã¯å±žã™ã‚‹ãŒå¾Œè€…ã®é›†åˆã«ã¯å±žã•ãªã„ã‚‚ã®ã§ã‚る。å„è¦ç´ ã¯ã‚¹ã‚³ã‚¢ã®é™é †ã§ä¸¦ã¹ã‚‰ã‚Œã‚‹ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>文書集åˆã‚’スコアã®é™é †ã§ä¸¦ã¹ã‚‹ã«ã¯ã€é–¢æ•° `odpairssort' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void odpairssort(ODPAIR *<var>pairs</var>, int <var>pnum</var>);</kbd></dt>
+<dd>`pairs' ã¯æ–‡æ›¸é›†åˆã®é…列を指定ã™ã‚‹ã€‚`pnum' ã¯ãã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>ã‚ã‚‹æ•°ã®è‡ªç„¶å¯¾æ•°ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odlogarithm' を用ã„る。</p>
+
+<dl>
+<dt><kbd>double odlogarithm(double <var>x</var>);</kbd></dt>
+<dd>`x' ã¯ã‚る数を指定ã™ã‚‹ã€‚戻り値ã¯ãã®æ•°ã®è‡ªç„¶å¯¾æ•°ã§ã‚る。もã—ãã®æ•°ãŒ 1.0 以下ã§ã‚ã‚Œã°ã€æˆ»ã‚Šå€¤ã¯ 0.0 ã¨ãªã‚‹ã€‚ã“ã®é–¢æ•°ã¯ã‚¢ãƒ—リケーションãŒæ¤œç´¢çµæžœã®IDFを算出ã™ã‚‹éš›ã«ä¾¿åˆ©ã§ã‚る。</dd>
+</dl>
+
+<p>二ã¤ã®ãƒ™ã‚¯ãƒˆãƒ«ã®ãªã™è§’ã®ä½™å¼¦ã‚’å¾—ã‚‹ã«ã¯ã€é–¢æ•° `odvectorcosine' を用ã„る。</p>
+
+<dl>
+<dt><kbd>double odvectorcosine(const int *<var>avec</var>, const int *<var>bvec</var>, int <var>vnum</var>);</kbd></dt>
+<dd>`avec' ã¯å‰è€…ã®æ•°å€¤é…列を指定ã™ã‚‹ã€‚`bvec' ã¯å¾Œè€…ã®æ•°å€¤é…列を指定ã™ã‚‹ã€‚`vnum' ã¯å„々ã®é…列ã®è¦ç´ 数を指定ã™ã‚‹ã€‚戻り値ã¯äºŒã¤ã®ãƒ™ã‚¯ãƒˆãƒ«ã®ãªã™è§’ã®ä½™å¼¦ã§ã‚る。ã“ã®é–¢æ•°ã¯ã‚¢ãƒ—リケーションãŒæ–‡æ›¸ã®é¡žä¼¼åº¦ã‚’算出ã™ã‚‹éš›ã«ä¾¿åˆ©ã§ã‚る。</dd>
+</dl>
+
+<p>性能を調整ã™ã‚‹å¤§åŸŸçš„ãªãƒ‘ラメータを指定ã™ã‚‹ã«ã¯ã€é–¢æ•° `odsettuning' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void odsettuning(int <var>ibnum</var>, int <var>idnum</var>, int <var>cbnum</var>, int <var>csiz</var>);</kbd></dt>
+<dd>`ibnum' ã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®ãƒã‚±ãƒƒãƒˆæ•°ã‚’指定ã™ã‚‹ã€‚`idnum' ã¯è»¢ç½®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®åˆ†å‰²æ•°ã‚’指定ã™ã‚‹ã€‚`cbnum' ã¯ãƒ€ãƒ¼ãƒ†ã‚£ãƒãƒƒãƒ•ã‚¡ã®ãƒã‚±ãƒƒãƒˆæ•°ã‚’指定ã™ã‚‹ã€‚`csiz' ã¯ãƒ€ãƒ¼ãƒ†ã‚£ãƒãƒƒãƒ•ã‚¡ã«ä½¿ã†ãƒ¡ãƒ¢ãƒªã®æœ€å¤§ãƒã‚¤ãƒˆæ•°ã‚’指定ã™ã‚‹ã€‚デフォルトã®è¨å®šã¯ `odsettuning(32749, 7, 262139, 8388608)' ã«ç›¸å½“ã™ã‚‹ã€‚ã“ã®é–¢æ•°ã¯ãƒãƒ³ãƒ‰ãƒ«ã‚’é–‹ãå‰ã«å‘¼ã³å‡ºã™ã¹ãã§ã‚る。</dd>
+</dl>
+
+<p>テã‚ストを分解ã—ã¦å‡ºç¾å½¢ã¨æ£è¦å½¢ã‚’別々ã®ãƒªã‚¹ãƒˆã«æ ¼ç´ã™ã‚‹ã«ã¯ã€é–¢æ•° `odanalyzetext' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void odanalyzetext(ODEUM *<var>odeum</var>, const char *<var>text</var>, CBLIST *<var>awords</var>, CBLIST *<var>nwords</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`text' ã¯ãƒ†ã‚ストã®æ–‡å—列を指定ã™ã‚‹ã€‚`awords' ã¯å‡ºç¾åž‹ã‚’æ ¼ç´ã™ã‚‹ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`nwords' ã¯æ£è¦åž‹ã‚’æ ¼ç´ã™ã‚‹ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ãŒã€`NULL' ãªã‚‰ç„¡è¦–ã•ã‚Œã‚‹ã€‚語群ã¯ç©ºç™½æ–‡å—ã¨ãƒ”リオドやコンマç‰ã®åŒºåˆ‡ã‚Šæ–‡å—ã§åˆ†å‰²ã•ã‚Œã‚‹ã€‚</dd>
+</dl>
+
+<p>関数 `odanalyzetext' ã§ä½¿ã‚れる文å—ã®åˆ†é¡žã‚’è¨å®šã™ã‚‹ã«ã¯ã€é–¢æ•° `odsetcharclass' を用ã„る。</p>
+
+<dl>
+<dt><kbd>void odsetcharclass(ODEUM *<var>odeum</var>, const char *<var>spacechars</var>, const char *<var>delimchars</var>, const char *<var>gluechars</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`spachechars' ã¯ç©ºç™½æ–‡å—ã‚’å«ã‚€æ–‡å—列を指定ã™ã‚‹ã€‚`delimchars' ã¯åŒºåˆ‡ã‚Šæ–‡å—ã‚’å«ã‚€æ–‡å—列を指定ã™ã‚‹ã€‚`gluechars' ã¯æŽ¥ç€æ–‡å—ã‚’å«ã‚€æ–‡å—列を指定ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å°ã•ãªå•ã„åˆã‚ã›è¨€èªžã‚’使ã£ã¦æ¤œç´¢ã‚’è¡Œã†ã«ã¯ã€é–¢æ•° `odquery' を用ã„る。</p>
+
+<dl>
+<dt><kbd>ODPAIR *odquery(ODEUM *<var>odeum</var>, const char *<var>query</var>, int *<var>np</var>, CBLIST *<var>errors</var>);</kbd></dt>
+<dd>`odeum' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚`query' ã¯å•ã„åˆã‚ã›è¨€èªžã®æ–‡å—列を指定ã™ã‚‹ã€‚`np' ã®å‚照先ã«ã¯ã€æˆ»ã‚Šå€¤ã®é…列ã®è¦ç´ æ•°ãŒæ ¼ç´ã•ã‚Œã‚‹ã€‚`error' ã¯ã‚¨ãƒ©ãƒ¼ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸ã‚’æ ¼ç´ã™ã‚‹ãƒªã‚¹ãƒˆãƒãƒ³ãƒ‰ãƒ«ã‚’指定ã™ã‚‹ã€‚戻り値ã¯æ£å¸¸ãªã‚‰é…列ã¸ã®ãƒã‚¤ãƒ³ã‚¿ã§ã‚ã‚Šã€ã‚¨ãƒ©ãƒ¼ãªã‚‰ `NULL' ã§ã‚る。ãã®é…列ã®å„è¦ç´ ã¯æ–‡æ›¸ã®ID番å·ã¨ã‚¹ã‚³ã‚¢ã®ãƒšã‚¢ã§ã‚ã‚Šã€ã‚¹ã‚³ã‚¢ã®é™é †ã§ä¸¦ã¹ã‚‰ã‚Œã‚‹ã€‚検索語ã«è©²å½“ã™ã‚‹æ–‡æ›¸ãŒä¸€ã¤ã‚‚ãªã‹ã£ãŸã¨ã—ã¦ã‚‚エラーã«ã¯ãªã‚‰ãšã«ã€ç©ºã®é…列を返ã™ã€‚戻り値ã®é ˜åŸŸã¯ `malloc' ã§ç¢ºä¿ã•ã‚Œã‚‹ã®ã§ã€ä¸è¦ã«ãªã£ãŸã‚‰ `free' ã§è§£æ”¾ã™ã‚‹ã¹ãã§ã‚る。é…列ã®å„è¦ç´ ã«ã¯æ—¢ã«å‰Šé™¤ã•ã‚ŒãŸæ–‡æ›¸ã®ãƒ‡ãƒ¼ã‚¿ã‚‚å«ã¾ã‚Œã‚‹ã“ã¨ã«æ³¨æ„ã™ã¹ãã§ã‚る。</dd>
+</dl>
+
+<h3>サンプルコード</h3>
+
+<p>文書をデータベースã«æ ¼ç´ã™ã‚‹ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#define DBNAME "index"
+
+int main(int argc, char **argv){
+ ODEUM *odeum;
+ ODDOC *doc;
+ CBLIST *awords;
+ const char *asis;
+ char *normal;
+ int i;
+
+ /* データベースを開ã */
+ if(!(odeum = odopen(DBNAME, OD_OWRITER | OD_OCREAT))){
+ fprintf(stderr, "odopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* 文書ãƒãƒ³ãƒ‰ãƒ«ã‚’å–å¾—ã™ã‚‹ */
+ doc = oddocopen("http://www.foo.bar/baz.txt");
+
+ /* 文書ã®å±žæ€§ã‚’è¨å®šã™ã‚‹ */
+ oddocaddattr(doc, "title", "Balcony Scene");
+ oddocaddattr(doc, "author", "Shakespeare");
+
+ /* テã‚ストを分解ã—ã¦èªžã®ãƒªã‚¹ãƒˆã‚’å¾—ã‚‹ */
+ awords = odbreaktext("Parting is such sweet sorrow.");
+
+ /* å„語を文書ãƒãƒ³ãƒ‰ãƒ«ã«è¨å®šã™ã‚‹ */
+ for(i = 0; i < cblistnum(awords); i++){
+ /* 語ã®ãƒªã‚¹ãƒˆã‹ã‚‰ä¸€èªžã‚’å–り出㙠*/
+ asis = cblistval(awords, i, NULL);
+ /* 出ç¾å½¢ã‹ã‚‰æ£è¦å½¢ã‚’生æˆã™ã‚‹ */
+ normal = odnormalizeword(asis);
+ /* 語を文書ãƒãƒ³ãƒ‰ãƒ«ã«è¨å®šã™ã‚‹ */
+ oddocaddword(doc, normal, asis);
+ /* æ£è¦å½¢ã®é ˜åŸŸã‚’解放ã™ã‚‹ */
+ free(normal);
+ }
+
+ /* 文書をデータベースã«ç™»éŒ²ã™ã‚‹ */
+ if(!odput(odeum, doc, -1, 1)){
+ fprintf(stderr, "odput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* 語ã®ãƒªã‚¹ãƒˆã‚’解放ã™ã‚‹ */
+ cblistclose(awords);
+
+ /* 文書ãƒãƒ³ãƒ‰ãƒ«ã‚’解放ã™ã‚‹ */
+ oddocclose(doc);
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!odclose(odeum)){
+ fprintf(stderr, "odclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>データベース内ã®æ–‡æ›¸ã‚’検索ã™ã‚‹ã‚µãƒ³ãƒ—ルコードを以下ã«ç¤ºã™ã€‚</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#define DBNAME "index"
+
+int main(int argc, char **argv){
+ ODEUM *odeum;
+ ODPAIR *pairs;
+ ODDOC *doc;
+ const CBLIST *words;
+ const char *title, *author, *asis;
+ int i, j, pnum;
+
+ /* データベースをå–å¾—ã™ã‚‹ */
+ if(!(odeum = odopen(DBNAME, OD_OREADER))){
+ fprintf(stderr, "odopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* 文書ã®æ¤œç´¢ã‚’行ㆠ*/
+ if((pairs = odsearch(odeum, "sorrow", -1, &pnum)) != NULL){
+
+ /* 文書ã®é…列を走査ã™ã‚‹ */
+ for(i = 0; i < pnum; i++){
+ /* 文書ãƒãƒ³ãƒ‰ãƒ«ã‚’å–å¾—ã™ã‚‹ */
+ if(!(doc = odgetbyid(odeum, pairs[i].id))) continue;
+ /* 文書ã®å±žæ€§ã‚’表示ã™ã‚‹ */
+ printf("URI: %s\n", oddocuri(doc));
+ title = oddocgetattr(doc, "title");
+ if(title) printf("TITLE: %s\n", title);
+ author = oddocgetattr(doc, "author");
+ if(author) printf("AUTHOR: %s\n", author);
+ /* 文書内ã®èªžã‚’出ç¾å½¢ã§è¡¨ç¤ºã™ã‚‹ */
+ printf("WORDS:");
+ words = oddocawords(doc);
+ for(j = 0; j < cblistnum(words); j++){
+ asis = cblistval(words, j, NULL);
+ printf(" %s", asis);
+ }
+ putchar('\n');
+ /* 文書ãƒãƒ³ãƒ‰ãƒ«ã‚’解放ã™ã‚‹ */
+ oddocclose(doc);
+ }
+
+ /* 文書ã®é…列を解放ã™ã‚‹ */
+ free(pairs);
+
+ } else {
+ fprintf(stderr, "odsearch: %s\n", dperrmsg(dpecode));
+ }
+
+ /* データベースを閉ã˜ã‚‹ */
+ if(!odclose(odeum)){
+ fprintf(stderr, "odclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>注記</h3>
+
+<p>Odeumを利用ã—ãŸãƒ—ãƒã‚°ãƒ©ãƒ をビルドã™ã‚‹æ–¹æ³•ã¯ã€Depotã®å ´åˆã¨å…¨ãåŒã˜ã§ã‚る。</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>POSIXスレッドを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã€å¤–部変数 `dpecode' ã¯ã‚¹ãƒ¬ãƒƒãƒ‰å›ºæœ‰ãƒ‡ãƒ¼ã‚¿ã¸ã®å‚ç…§ã¨ã—ã¦æ‰±ã‚ã‚Œã€Odeumã®å„関数ã¯ãƒªã‚¨ãƒ³ãƒˆãƒ©ãƒ³ãƒˆã«ãªã‚‹ã€‚ãã®å ´åˆã€ã‚¹ãƒ¬ãƒƒãƒ‰é–“ã§åŒæ™‚ã«åŒã˜ãƒãƒ³ãƒ‰ãƒ«ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ãªã„é™ã‚Šã¯ã€å„関数ã¯ã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ã§ã‚る。ãŸã ã—ã€`errno' ã‚„ `malloc' ç‰ãŒã‚¹ãƒ¬ãƒƒãƒ‰ã‚»ãƒ¼ãƒ•ãªå‡¦ç†ç³»ã§ã‚ã‚‹ã“ã¨ãŒå‰æã¨ãªã‚‹ã€‚</p>
+
+<p>ZLIBを有効ã«ã—ã¦QDBMをビルドã—ãŸå ´åˆã€æ–‡æ›¸å±žæ€§ç”¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã¯åœ§ç¸®ã•ã‚Œã¦ä¿å˜ã•ã‚Œã‚‹ã€‚ãã®å ´åˆã€ã‚µã‚¤ã‚ºãŒ30%以下ã«ãªã‚‹ã€‚ã—ãŸãŒã£ã¦ã€Odeumを利用ã™ã‚‹å ´åˆã¯ZLIBを有効ã«ã—ãŸæ–¹ãŒã‚ˆã„。ZLIBを有効ã«ã—ã¦ä½œæˆã—ãŸOdeumã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’ã€ZLIBを有効ã«ã—ã¦ã„ãªã„環境ã§åˆ©ç”¨ã™ã‚‹ã“ã¨ã¯ã§ããšã€ã¾ãŸãã®é€†ã‚‚åŒæ§˜ã§ã‚る。ZLIBãŒæœ‰åŠ¹ã§ãªãLZOãŒæœ‰åŠ¹ãªå ´åˆã¯ã€ZLIBã®å¤‰ã‚ã‚Šã«LZOãŒç”¨ã„られる。</p>
+
+<h3>å•ã„åˆã‚ã›è¨€èªž</h3>
+
+<p>関数 `odquery' ã®å•ã„åˆã‚ã›è¨€èªžã¯ä»¥ä¸‹ã®æ–‡æ³•ã«åŸºã¥ã。</p>
+
+<pre>expr ::= subexpr ( op subexpr )*
+subexpr ::= WORD
+subexpr ::= LPAREN expr RPAREN
+</pre>
+
+<p>演算åã¨ã—ã¦ã¯ "&"（AND）㨠"|"（OR）㨠"!"（NOTAND）ãŒåˆ©ç”¨ã§ãる。ã¾ãŸã€æ‹¬å¼§ "()" を使ã†ã“ã¨ã§æ¼”ç®—åã®è©•ä¾¡é †åºã‚’制御ã™ã‚‹ã“ã¨ãŒã§ãる。å•ã„åˆã‚ã›ã®æ–‡å—列ã¯é–¢æ•° `odanalyzetext' を用ã„ã¦åˆ†è§£ã•ã‚Œã‚‹ã®ã§ã€"&"ã€"|"ã€"!"ã€"("ã€")" ã¯åŒºåˆ‡ã‚Šæ–‡å—ã¨ã—ã¦è¨å®šã•ã‚Œã¦ã„ã‚‹å¿…è¦ãŒã‚る。ã¾ãŸã€ç©ºç™½ã§å˜èªžã‚’区切ã£ã¦ã‚‚ "&" ã§åŒºåˆ‡ã£ãŸã®ã¨åŒã˜ã“ã¨ã«ãªã‚‹ã€‚ã¤ã¾ã‚Š "joe blow" 㯠"joe & blow" ã¨åŒã˜ã§ã‚る。</p>
+
+<p>å•ã„åˆã‚ã›æ–‡å—列ã®æ–‡å—コードã¯å¯¾è±¡æ–‡æ›¸ã®æ–‡å—コードã¨ä¸€è‡´ã—ã¦ã„ã‚‹å¿…è¦ãŒã‚る。ã¾ãŸã€ç©ºç™½æ–‡å—ã¨åŒºåˆ‡ã‚Šæ–‡å—ã¨æŽ¥ç€æ–‡å—ã«æŒ‡å®šã§ãã‚‹ã®ã¯1ãƒã‚¤ãƒˆæ–‡å—ã ã‘ã§ã‚る。</p>
+
+<hr />
+
+<h2><a name="odeumcli" id="odeumcli" class="head">Odeum用コマンド</a></h2>
+
+<p>Odeumã«å¯¾å¿œã™ã‚‹ã‚³ãƒžãƒ³ãƒ‰ãƒ©ã‚¤ãƒ³ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ã¯ä»¥ä¸‹ã®ã‚‚ã®ã§ã‚る。</p>
+
+<p>コマンド `odmgr' ã¯Odeumã‚„ãã®ã‚¢ãƒ—リケーションã®ãƒ‡ãƒãƒƒã‚°ã«å½¹ç«‹ã¤ãƒ„ールã§ã‚る。データベースを更新ã—ãŸã‚Šã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®çŠ¶æ…‹ã‚’調ã¹ãŸã‚Šã™ã‚‹æ©Ÿèƒ½ã‚’æŒã¤ã€‚シェルスクリプトã§å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ を作るã®ã«ã‚‚利用ã§ãる。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`file' ã¯ãƒ•ã‚¡ã‚¤ãƒ«åã€`expr' ã¯æ–‡æ›¸ã®URIã‹ID番å·ã€`words' ã¯æ¤œç´¢èªžã€`elems' ã¯è¦ç´ データベースを指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>odmgr create <var>name</var></kbd></dt>
+<dd>データベースファイルを作æˆã™ã‚‹ã€‚</dd>
+<dt><kbd>odmgr put [-uri <var>str</var>] [-title <var>str</var>] [-author <var>str</var>] [-date <var>str</var>] [-wmax <var>num</var>] [-keep] <var>name</var> [<var>file</var>]</kbd></dt>
+<dd>ファイルをèªã¿è¾¼ã‚“ã§æ–‡æ›¸ã‚’è¿½åŠ ã™ã‚‹ã€‚`file' ã‚’çœç•¥ã™ã‚‹ã¨æ¨™æº–入力をèªã¿è¾¼ã‚€ãŒã€ãã®å ´åˆã¯URIã®æŒ‡å®šãŒå¿…é ˆã¨ãªã‚‹ã€‚</dd>
+<dt><kbd>odmgr out [-id] <var>name</var> <var>expr</var></kbd></dt>
+<dd>URIã«å¯¾å¿œã™ã‚‹æ–‡æ›¸ã‚’削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>odmgr get [-id] [-t|-h] <var>name</var> <var>expr</var></kbd></dt>
+<dd>URIã«å¯¾å¿œã™ã‚‹æ–‡æ›¸ã‚’表示ã™ã‚‹ã€‚出力ã¯æ–‡æ›¸ã®ID番å·ã¨URIã¨ã‚¹ã‚³ã‚¢ã‚’タブã§åŒºåˆ‡ã£ãŸã‚‚ã®ã§ã‚る。</dd>
+<dt><kbd>odmgr search [-max <var>num</var>] [-or] [-idf] [-t|-h|-n] <var>name</var> <var>words</var>...</kbd></dt>
+<dd>指定ã—ãŸèªžã‚’å«ã‚€æ–‡æ›¸ã‚’検索ã™ã‚‹ã€‚出力ã®ç¬¬1è¡Œã¯ã€æ¤œç´¢èªžå…¨ä½“ã®è©²å½“æ•°ã¨å„検索語ãŠã‚ˆã³ãã®è©²å½“数をタブã§åŒºåˆ‡ã£ãŸã‚‚ã®ã§ã‚る。第2行以é™ã¯ã€è©²å½“ã®å„文書ã®ID番å·ã¨URIã¨ã‚¹ã‚³ã‚¢ã‚’タブã§åŒºåˆ‡ã£ãŸã‚‚ã®ã§ã‚る。</dd>
+<dt><kbd>odmgr list [-t|-h] <var>name</var></kbd></dt>
+<dd>データベース内ã®å…¨ã¦ã®æ–‡æ›¸ã‚’表示ã™ã‚‹ã€‚出力ã®å„è¡Œã¯æ–‡æ›¸ã®ID番å·ã¨URIã¨ã‚¹ã‚³ã‚¢ã‚’タブã§åŒºåˆ‡ã£ãŸã‚‚ã®ã§ã‚る。</dd>
+<dt><kbd>odmgr optimize <var>name</var></kbd></dt>
+<dd>データベースを最é©åŒ–ã™ã‚‹ã€‚</dd>
+<dt><kbd>odmgr inform <var>name</var></kbd></dt>
+<dd>データベースã®é›‘多ãªæƒ…å ±ã‚’å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+<dt><kbd>odmgr merge <var>name</var> <var>elems</var>...</kbd></dt>
+<dd>複数ã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’マージã™ã‚‹ã€‚</dd>
+<dt><kbd>odmgr remove <var>name</var></kbd></dt>
+<dd>データベースディレクトリを削除ã™ã‚‹ã€‚</dd>
+<dt><kbd>odmgr break [-h|-k|-s] [<var>file</var>]</kbd></dt>
+<dd>ファイルをèªã¿è¾¼ã‚“ã§ã€ãƒ†ã‚ストを語ã«åˆ†è§£ã—ã¦å‡ºåŠ›ã™ã‚‹ã€‚出力ã®å„è¡Œã¯å„語ã®æ£è¦å½¢ã¨å‡ºç¾å½¢ã‚’タブã§åŒºåˆ‡ã£ãŸã‚‚ã®ã§ã‚る。</dd>
+<dt><kbd>odmgr version</kbd></dt>
+<dd>QDBMã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³æƒ…å ±ã‚’å‡ºåŠ›ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-uri <var>str</var></kbd> : 文書ã®URIを明示的ã«æŒ‡å®šã™ã‚‹ã€‚</li>
+<li><kbd>-title <var>str</var></kbd> : 文書ã®ã‚¿ã‚¤ãƒˆãƒ«ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-author <var>str</var></kbd> : 文書ã®è‘—者åを指定ã™ã‚‹ã€‚</li>
+<li><kbd>-date <var>str</var></kbd> : 文書ã®æ›´æ–°æ—¥æ™‚を指定ã™ã‚‹ã€‚</li>
+<li><kbd>-wmax <var>num</var></kbd> : æ ¼ç´ã™ã‚‹èªžã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-keep</kbd> : åŒã˜URIã®æ–‡æ›¸ãŒæ—¢å˜ã§ã‚ã‚Œã°ä¸Šæ›¸ãã‚’è¡Œã‚ãªã„。</li>
+<li><kbd>-id</kbd> : URIã§ãªãID番å·ã§æ–‡æ›¸ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-t</kbd> : 文書ã®è©³ç´°æƒ…å ±ã‚’ã‚¿ãƒ–åŒºåˆ‡ã‚Šã§å‡ºåŠ›ã™ã‚‹ã€‚</li>
+<li><kbd>-h</kbd> : 文書ã®è©³ç´°æƒ…å ±ã‚’äººé–“ãŒèªã¿ã‚„ã™ã„å½¢å¼ã§å‡ºåŠ›ã™ã‚‹ã€‚</li>
+<li><kbd>-k</kbd> : 文書ã®ã‚ーワードã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+<li><kbd>-s</kbd> : 文書ã®è¦ç´„ã®ã¿ã‚’出力ã™ã‚‹ã€‚</li>
+<li><kbd>-max <var>num</var></kbd> : 出力ã™ã‚‹æ–‡æ›¸ã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-or</kbd> : AND検索ã§ãªãOR検索を行ã†ã€‚</li>
+<li><kbd>-idf</kbd> : IDFã§ã‚¹ã‚³ã‚¢ã‚’é‡ã¿ã¥ã‘ã™ã‚‹ã€‚</li>
+<li><kbd>-n</kbd> : 文書ã®IDã¨ã‚¹ã‚³ã‚¢ã®ã¿ã‚’表示ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `odtest' ã¯Odeumã®æ©Ÿèƒ½ãƒ†ã‚¹ãƒˆã‚„性能テストã«ç”¨ã„るツールã§ã‚る。ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã«ã‚ˆã£ã¦ç”Ÿæˆã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’ `odmgr' ã«ã‚ˆã£ã¦è§£æžã—ãŸã‚Šã€`time' コマンドã«ã‚ˆã£ã¦ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã®å®Ÿè¡Œæ™‚間を計ã£ãŸã‚Šã™ã‚‹ã¨ã‚ˆã„。以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`dnum' ã¯æ–‡æ›¸æ•°ã€`wnum' ã¯æ–‡æ›¸æ¯Žã®èªžæ•°ã€`pnum' ã¯èªžã®ãƒ‘ターン数を指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>odtest write [-tune <var>ibnum</var> <var>idnum</var> <var>cbnum</var> <var>csiz</var>] <var>name</var> <var>dnum</var> <var>wnum</var> <var>pnum</var></kbd></dt>
+<dd>無作為ãªå±žæ€§ã¨èªžã‚’æŒã¤æ–‡æ›¸ã‚’連続ã—ã¦ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¿½åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>odtest read <var>name</var></kbd></dt>
+<dd>上記ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å…¨æ–‡æ›¸ã‚’検索ã™ã‚‹ã€‚</dd>
+<dt><kbd>odtest combo <var>name</var></kbd></dt>
+<dd>å„種æ“作ã®çµ„ã¿åˆã‚ã›ãƒ†ã‚¹ãƒˆã‚’è¡Œã†ã€‚</dd>
+<dt><kbd>odtest wicked <var>name</var> <var>dnum</var></kbd></dt>
+<dd>å„種更新æ“作を無作為ã«é¸æŠžã—ã¦å®Ÿè¡Œã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-tune <var>ibnum</var> <var>idnum</var> <var>cbnum</var> <var>csiz</var></kbd> : 性能パラメータを指定ã™ã‚‹ã€‚</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚環境変数 `QDBMDBGFD' ã®å€¤ã¨ã—ã¦ã€å¤‰æ•° `dpecode' ã®å¤‰æ›´å±¥æ´ã‚’出力ã™ã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ—タを指定ã“ã¨ãŒã§ãる。</p>
+
+<p>コマンド `odidx' ã¯ãƒãƒ¼ã‚«ãƒ«ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ 上ã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚’èªã¿è¾¼ã‚“ã§Odeumã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ç™»éŒ²ã™ã‚‹ãƒ¦ãƒ¼ãƒ†ã‚£ãƒªãƒ†ã‚£ã§ã‚る。ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯Webサイトã®å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ を構築ã™ã‚‹éš›ã«å½¹ç«‹ã¤ã€‚サãƒãƒ¼ãƒˆã•ã‚Œã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒ•ã‚©ãƒ¼ãƒžãƒƒãƒˆã¯ãƒ—レーンテã‚ストã¨HTMLã§ã‚る。サãƒãƒ¼ãƒˆã•ã‚Œã‚‹æ–‡å—コードã¯US-ASCIIã¨ISO-8859-1ã§ã‚る。å„文書ã®URIã«ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã®ãƒ‘スãŒæŒ‡å®šã•ã‚Œã‚‹ã€‚å„文書ã«ã¯ã€`title' 㨠`date' ã¨ã„ã†å±žæ€§ãŒä»˜ä¸Žã•ã‚Œã‚‹ã€‚æ—¢ã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ç™»éŒ²ã—ã¦ã‚るファイルを登録ã—よã†ã¨ã—ãŸå ´åˆã€æ›´æ–°æ™‚刻ãŒæ–°ã—ã‘ã‚Œã°ç™»éŒ²ã•ã‚Œã€ãã†ã§ãªã‘ã‚Œã°ç„¡è¦–ã•ã‚Œã‚‹ã€‚更新時刻ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®ä¸ã® '_mtime' ã¨ã„ã†ã‚µãƒ–データベースã«è¨˜éŒ²ã•ã‚Œã‚‹ã€‚ã‚¹ã‚³ã‚¢æƒ…å ±ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã®ä¸ã® `_score' ã¨ã„ã†ã‚µãƒ–データベースã«è¨˜éŒ²ã•ã‚
Œã‚‹ã€‚以下ã®æ›¸å¼ã§ç”¨ã„る。`name' ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹åã€`dir' ã¯ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªåを指定ã™ã‚‹ã€‚</p>
+
+<dl>
+<dt><kbd>odidx register [-l <var>file</var>] [-wmax <var>num</var>] [-tsuf <var>sufs</var>] [-hsuf <var>sufs</var>] <var>name</var> [<var>dir</var>]</kbd></dt>
+<dd>特定ã®ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªä»¥ä¸‹ã®ãƒ•ã‚¡ã‚¤ãƒ«ç¾¤ã‚’データベース登録ã™ã‚‹ã€‚`dir' ãŒçœç•¥ã•ã‚ŒãŸå ´åˆã€ã‚«ãƒ¬ãƒ³ãƒˆãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªãŒæŒ‡å®šã•ã‚Œã‚‹ã€‚</dd>
+<dt><kbd>odidx relate <var>name</var></kbd></dt>
+<dd>データベースã®å„文書ã«é–¢é€£æ–‡æ›¸æ¤œç´¢ã®ãŸã‚ã®ã‚¹ã‚³ã‚¢æƒ…å ±ã‚’ä»˜åŠ ã™ã‚‹ã€‚</dd>
+<dt><kbd>odidx purge <var>name</var></kbd></dt>
+<dd>ファイルシステムã«å˜åœ¨ã—ãªã„文書をデータベースã‹ã‚‰å‰Šé™¤ã™ã‚‹ã€‚</dd>
+</dl>
+
+<p>å„オプションã¯ä»¥ä¸‹ã®æ©Ÿèƒ½ã‚’æŒã¤ã€‚</p>
+
+<ul class="lines">
+<li><kbd>-l <var>file</var></kbd> : 登録ã™ã¹ãファイルã®ãƒ‘スã®ãƒªã‚¹ãƒˆã‚’ファイルã‹ã‚‰èªã¿è¾¼ã‚€ã€‚`-' を指定ã—ãŸå ´åˆã€æ¨™æº–入力ãŒèªã¿è¾¼ã¾ã‚Œã‚‹ã€‚</li>
+<li><kbd>-wmax <var>num</var></kbd> : データベースã«æ ¼ç´ã™ã‚‹èªžã®æœ€å¤§æ•°ã‚’指定ã™ã‚‹ã€‚</li>
+<li><kbd>-tsuf <var>sufs</var></kbd> : プレーンテã‚ストファイルã®æ‹¡å¼µåをカンマ区切りã§æŒ‡å®šã™ã‚‹ã€‚デフォルト㯠`-tsuf .txt,.text' ã¨åŒæ„ã§ã‚る。</li>
+<li><kbd>-hsuf <var>sufs</var></kbd> : HTMLファイルã®æ‹¡å¼µåをカンマ区切りã§æŒ‡å®šã™ã‚‹ã€‚デフォルト㯠`-hsuf .html,.htm' ã¨åŒæ„ã§ã‚る。</li>
+</ul>
+
+<p>ã“ã®ã‚³ãƒžãƒ³ãƒ‰ã¯å‡¦ç†ãŒæ£å¸¸ã«çµ‚了ã™ã‚Œã° 0 ã‚’è¿”ã—ã€ã‚¨ãƒ©ãƒ¼ãŒã‚ã‚Œã°ãれ以外ã®å€¤ã‚’è¿”ã—ã¦çµ‚了ã™ã‚‹ã€‚</p>
+
+<p>Odeumã®ã‚³ãƒžãƒ³ãƒ‰ç¾¤ã‚’駆使ã™ã‚‹ã¨ã€å…¨æ–‡æ¤œç´¢ã‚·ã‚¹ãƒ†ãƒ ã‚’ç°¡å˜ã«å®Ÿç¾ã™ã‚‹ã“ã¨ãŒã§ãる。例ãˆã° `/home/mikio' 以下ã«ã‚ã‚Šã€ã‹ã¤ `.txt' ã‹ `.c' ã‹ `.h' ã¨ã„ã†æŽ¥å°¾è¾žã‚’æŒã¤ãƒ•ã‚¡ã‚¤ãƒ«ç¾¤ã‚’ `casket' ã¨ã„ã†åå‰ã®ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã«ç™»éŒ²ã™ã‚‹ãªã‚‰ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>odidx register -tsuf ".txt,.c,.h" -hsuf "" casket /home/mikio
+</pre>
+
+<p>ãã—ã¦ã€`unix' ãŠã‚ˆã³ `posix' ã¨ã„ã†èªžã‚’å«ã‚€æ–‡æ›¸ã‚’検索ã—ã€ä¸Šä½8件を表示ã™ã‚‹ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã™ã‚‹ã€‚</p>
+
+<pre>odmgr search -max 8 -h casket "unix posix"
+</pre>
+
+<p>`odidx' ã§ç”Ÿæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯ã€QDBMã«ä»˜éŒ²ã•ã‚Œã‚‹å…¨æ–‡æ¤œç´¢ã®ãŸã‚ã®CGIスクリプト `qfts.cgi' ã§ãã®ã¾ã¾åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<hr />
+
+<h2><a name="fileformat" id="fileformat" class="head">ファイルフォーマット</a></h2>
+
+<h3>Depotã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ•ã‚©ãƒ¼ãƒžãƒƒãƒˆ</h3>
+
+<p>DepotãŒç®¡ç†ã™ã‚‹ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®å†…容ã¯ã€ãƒ˜ãƒƒãƒ€éƒ¨ã€ãƒã‚±ãƒƒãƒˆéƒ¨ã€ãƒ¬ã‚³ãƒ¼ãƒ‰éƒ¨ã®ä¸‰ã¤ã«å¤§åˆ¥ã•ã‚Œã‚‹ã€‚</p>
+
+<p>ヘッダ部ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã®å…ˆé ã‹ã‚‰ 48 ãƒã‚¤ãƒˆã®å›ºå®šé•·ã§ã¨ã‚‰ã‚Œã€ä»¥ä¸‹ã®æƒ…å ±ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚</p>
+
+<ol>
+<li>マジックナンム: オフセット 0 ã‹ã‚‰å§‹ã¾ã‚‹ã€‚ビッグエンディアン用ãªã‚‰æ–‡å—列 "[DEPOT]\n\f" を内容ã¨ã—ã€ãƒªãƒˆãƒ«ã‚¨ãƒ³ãƒ‡ã‚£ã‚¢ãƒ³ç”¨ãªã‚‰æ–‡å—列 "[depot]\n\f" を内容ã¨ã™ã‚‹ã€‚</li>
+<li>ãƒãƒ¼ã‚¸ãƒ§ãƒ³ç•ªå· : オフセット 12 ã‹ã‚‰å§‹ã¾ã‚‹ã€‚ライブラリã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ç•ªå·ã‚’10進数ã§è¡¨ç¾ã—ãŸæ–‡å—列を内容ã¨ã™ã‚‹ã€‚</li>
+<li>ラッパー用フラグ : オフセット 16 ã‹ã‚‰å§‹ã¾ã‚‹ã€‚`int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>ファイルサイズ : オフセット 24 ã‹ã‚‰å§‹ã¾ã‚‹ã€‚`int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•° : オフセット 32 ã‹ã‚‰å§‹ã¾ã‚‹ã€‚`int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>レコード数 : オフセット 40 ã‹ã‚‰å§‹ã¾ã‚‹ã€‚`int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+</ol>
+
+<p>ãƒã‚±ãƒƒãƒˆéƒ¨ã¯ãƒ˜ãƒƒãƒ€éƒ¨ã®ç›´å¾Œã«ãƒã‚±ãƒƒãƒˆé…列ã®è¦ç´ æ•°ã«å¿œã˜ãŸå¤§ãã•ã§ã¨ã‚‰ã‚Œã€ãƒã‚§ãƒ¼ãƒ³ã®å…ˆé è¦ç´ ã®ã‚ªãƒ•ã‚»ãƒƒãƒˆãŒå„è¦ç´ ã«è¨˜éŒ²ã•ã‚Œã‚‹ã€‚</p>
+
+<p>レコード部ã¯ãƒã‚±ãƒƒãƒˆéƒ¨ã®ç›´å¾Œã‹ã‚‰ãƒ•ã‚¡ã‚¤ãƒ«ã®æœ«å°¾ã¾ã§ã‚’å ã‚ã€å„レコードã®ä»¥ä¸‹ã®æƒ…å ±ã‚’æŒã¤è¦ç´ ãŒè¨˜éŒ²ã•ã‚Œã‚‹ã€‚</p>
+
+<ol>
+<li>フラグ : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>ã‚ーã®ç¬¬äºŒãƒãƒƒã‚·ãƒ¥å€¤ : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>ã‚ーã®ã‚µã‚¤ã‚º : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>値ã®ã‚µã‚¤ã‚º : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>パディングã®ã‚µã‚¤ã‚º : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>å·¦ã®åã®ä½ç½® : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>å³ã®åã®ä½ç½® : `int' åž‹ã®æ•´æ•°ã§ã‚る。</li>
+<li>ã‚ーã®å®Ÿãƒ‡ãƒ¼ã‚¿ : ã‚ーã®ã‚µã‚¤ã‚ºã§å®šç¾©ã•ã‚Œã‚‹é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。</li>
+<li>値ã®å®Ÿãƒ‡ãƒ¼ã‚¿ : 値ã®ã‚µã‚¤ã‚ºã§å®šç¾©ã•ã‚Œã‚‹é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。</li>
+<li>パディング : 値ã®ã‚µã‚¤ã‚ºã¨ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã«ã‚ˆã‚Šç®—出ã•ã‚Œã‚‹é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。</li>
+</ol>
+
+<h3>Villaã®ãƒ•ã‚¡ã‚¤ãƒ«ãƒ•ã‚©ãƒ¼ãƒžãƒƒãƒˆ</h3>
+
+<p>Villaã®æ‰±ã†å…¨ã¦ã®ãƒ‡ãƒ¼ã‚¿ã¯Depotã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«è¨˜éŒ²ã•ã‚Œã‚‹ã€‚記録ã•ã‚Œã‚‹ãƒ‡ãƒ¼ã‚¿ã¯ã€ãƒ¡ã‚¿ãƒ‡ãƒ¼ã‚¿ã¨è«–ç†ãƒšãƒ¼ã‚¸ã«åˆ†é¡žã•ã‚Œã‚‹ã€‚è«–ç†ãƒšãƒ¼ã‚¸ã¯ãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã¨éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã«åˆ†é¡žã•ã‚Œã‚‹ã€‚メタデータã¯ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ç‰ã®ç®¡ç†æƒ…å ±ã‚’è¨˜éŒ²ã™ã‚‹ã‚‚ã®ã§ã€ã‚ーã¨å€¤ã¨ã‚‚ã« `int' åž‹ã§ã‚る。リーフノードã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’ä¿æŒã™ã‚‹ã€‚éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã¯ãƒšãƒ¼ã‚¸ã‚’å‚ç…§ã™ã‚‹ç–Žã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’ä¿æŒã™ã‚‹ã€‚</p>
+
+<p>Villaã¯ã€å°ã•ã„自然数を直列化ã—ã¦æ‰±ã†éš›ã«è¨˜æ†¶é ˜åŸŸã‚’節約ã™ã‚‹ãŸã‚ã«ã€å¯å¤‰é•·æ•´æ•°ãƒ•ã‚©ãƒ¼ãƒžãƒƒãƒˆï¼ˆBER圧縮）を用ã„る。å¯å¤‰é•·æ•´æ•°ã®ã‚ªãƒ–ジェクトã¯ã€é ˜åŸŸã®å…ˆé ã‹ã‚‰è§£æžã—ã€å€¤ãŒæ£ã®ãƒã‚¤ãƒˆã‚’èªã‚“ã らãã“ã§çµ‚端ã¨ã™ã‚‹ã€‚å„ãƒã‚¤ãƒˆã¯çµ¶å¯¾å€¤ã§è©•ä¾¡ã•ã‚Œã€ãƒªãƒˆãƒ«ã‚¨ãƒ³ãƒ‡ã‚£ã‚¢ãƒ³ã®128進数ã¨ã—ã¦ç®—出ã•ã‚Œã‚‹ã€‚</p>
+
+<p>レコードã¯ãƒ¦ãƒ¼ã‚¶ãƒ‡ãƒ¼ã‚¿ã®è«–ç†çš„ãªå˜ä½ã§ã‚る。ã‚ーãŒé‡è¤‡ã™ã‚‹è«–ç†ãƒ¬ã‚³ãƒ¼ãƒ‰ã¯ç‰©ç†çš„ã«ã¯å˜ä¸€ã®ãƒ¬ã‚³ãƒ¼ãƒ‰ã«ã¾ã¨ã‚られる。物ç†ãƒ¬ã‚³ãƒ¼ãƒ‰ã¯ä»¥ä¸‹ã®å½¢å¼ã§ç›´åˆ—化ã•ã‚Œã‚‹ã€‚</p>
+
+<ol>
+<li>ã‚ーã®ã‚µã‚¤ã‚º : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>ã‚ーã®å®Ÿãƒ‡ãƒ¼ã‚¿ : ã‚ーã®ã‚µã‚¤ã‚ºã§å®šç¾©ã•ã‚Œã‚‹é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。</li>
+<li>値ã®æ•° : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>値ã®ãƒªã‚¹ãƒˆ : 以下ã®è¡¨ç¾ã‚’値ã®æ•°ã ã‘ç¹°ã‚Šè¿”ã—ãŸä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。<ol>
+<li>サイズ : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>実データ : サイズã§å®šç¾©ã•ã‚Œã‚‹é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。</li>
+</ol></li>
+</ol>
+
+<p>リーフノードã¯ãƒ¬ã‚³ãƒ¼ãƒ‰ã®é›†åˆã‚’æ ¼ç´ã™ã‚‹ãŸã‚ã®ç‰©ç†çš„ãªå˜ä½ã§ã‚る。リーフノード㯠`int' åž‹ã®IDã‚’ã‚ーã¨ã—ã€ä»¥ä¸‹ã®å€¤ã‚’æŒã¤ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã—ã¦Depotã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚Œã‚‹ã€‚レコードã¯å¸¸ã«ã‚ーã®æ˜‡é †ã«æ•´åˆ—ã—ãŸçŠ¶æ…‹ã§ä¿æŒã•ã‚Œã‚‹ã€‚</p>
+
+<ol>
+<li>å‰ã®ãƒªãƒ¼ãƒ•ã®ID : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>次ã®ãƒªãƒ¼ãƒ•ã®ID : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>レコードã®ãƒªã‚¹ãƒˆ : 直列化ã—ãŸãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’連çµã—ãŸã‚‚ã®ã€‚</li>
+</ol>
+
+<p>インデックスã¯ãƒšãƒ¼ã‚¸ã‚’探索ã™ã‚‹ãŸã‚ã®ãƒã‚¤ãƒ³ã‚¿ã®è«–ç†çš„ãªå˜ä½ã§ã‚る。インデックスã¯ä»¥ä¸‹ã®å½¢å¼ã§ç›´åˆ—化ã•ã‚Œã‚‹ã€‚</p>
+
+<ol>
+<li>å‚照先ã®ãƒšãƒ¼ã‚¸ã®ID : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>ã‚ーã®ã‚µã‚¤ã‚º : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>ã‚ーã®å®Ÿãƒ‡ãƒ¼ã‚¿ : ã‚ーã®ã‚µã‚¤ã‚ºã§å®šç¾©ã•ã‚Œã‚‹é•·ã•ã‚’æŒã¤ä¸€é€£ã®ãƒã‚¤ãƒˆã§ã‚る。</li>
+</ol>
+
+<p>éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã¯ã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã®é›†åˆã‚’æ ¼ç´ã™ã‚‹ãŸã‚ã®ç‰©ç†çš„ãªå˜ä½ã§ã‚る。éžãƒªãƒ¼ãƒ•ãƒŽãƒ¼ãƒ‰ã¯ `int' åž‹ã®IDã‚’ã‚ーã¨ã—ã€ä»¥ä¸‹ã®å€¤ã‚’æŒã¤ãƒ¬ã‚³ãƒ¼ãƒ‰ã¨ã—ã¦Depotã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«æ ¼ç´ã•ã‚Œã‚‹ã€‚インデックスã¯å¸¸ã«ã‚ーã®æ˜‡é †ã«æ•´åˆ—ã—ãŸçŠ¶æ…‹ã§ä¿æŒã•ã‚Œã‚‹ã€‚</p>
+
+<ol>
+<li>最åˆã®åノードã®ID : å¯å¤‰é•·æ•´æ•°åž‹ã§ã‚る。</li>
+<li>インデックスã®ãƒªã‚¹ãƒˆ : 直列化ã—ãŸã‚¤ãƒ³ãƒ‡ãƒƒã‚¯ã‚¹ã‚’連çµã—ãŸã‚‚ã®ã€‚</li>
+</ol>
+
+<h3>注記</h3>
+
+<p>データベースファイルã¯ã‚¹ãƒ‘ースã§ã¯ãªã„ã®ã§ã€é€šå¸¸ã®ãƒ•ã‚¡ã‚¤ãƒ«ã¨åŒæ§˜ã«è¤‡è£½ç‰ã®æ“作を行ã†ã“ã¨ãŒã§ãる。Depotã¯ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã®èª¿æ•´ã‚’ã—ãªã„ã§ãƒ•ã‚¡ã‚¤ãƒ«ã®èªã¿æ›¸ãã‚’è¡Œã£ã¦ã„ã‚‹ã®ã§ã€ãƒã‚¤ãƒˆã‚ªãƒ¼ãƒ€ã®ç•°ãªã‚‹ç’°å¢ƒã«ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’移è¨ã—ã¦ã‚‚ãã®ã¾ã¾ã§ã¯åˆ©ç”¨ã§ããªã„。</p>
+
+<p>Depotã‚„Villaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã‚’ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã§é…布ã™ã‚‹éš›ã«ã¯ã€MIMEタイプを `application/x-qdbm' ã«ã—ã¦ã»ã—ã„。ファイルåã®æŽ¥å°¾è¾žã¯ `.qdb' ã«ã—ã¦ã»ã—ã„。Curiaã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ‡ã‚£ãƒ¬ã‚¯ãƒˆãƒªã‚’ãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã§é…布ã™ã‚‹éš›ã«ã¯ã€TARå½¢å¼ç‰ã‚’用ã„ãŸã‚¢ãƒ¼ã‚«ã‚¤ãƒ–ã«å¤‰æ›ã—ã¦è¡Œã†ã“ã¨ãŒã§ãる。</p>
+
+<p>データベースファイルã®ãƒžã‚¸ãƒƒã‚¯ãƒŠãƒ³ãƒã‚’ `file' コマンドã«è˜åˆ¥ã•ã›ãŸã„å ´åˆã¯ã€`magic' ファイルã«ä»¥ä¸‹ã®è¡Œã‚’追記ã™ã‚‹ã¨ã‚ˆã„。</p>
+
+<pre>0 string [DEPOT]\n\f QDBM, big endian
+>12 string x \b, version=%s
+>19 byte ^1 \b, Hash
+>19 byte &1 \b, B+ tree
+>19 byte &2 \b (deflated:ZLIB)
+>19 byte &4 \b (deflated:LZO)
+>19 byte &8 \b (deflated:BZIP2)
+>24 belong x \b, filesize=%d
+>32 belong x \b, buckets=%d
+>40 belong x \b, records=%d
+0 string [depot]\n\f QDBM, little endian
+>12 string x \b, version=%s
+>16 byte ^1 \b, Hash
+>16 byte &1 \b, B+ tree
+>16 byte &2 \b (deflated:ZLIB)
+>16 byte &4 \b (deflated:LZO)
+>16 byte &8 \b (deflated:BZIP2)
+>24 lelong x \b, filesize=%d
+>32 lelong x \b, buckets=%d
+>40 lelong x \b, records=%d
+</pre>
+
+<hr />
+
+<h2><a name="porting" id="porting" class="head">移æ¤æ–¹æ³•</a></h2>
+
+<p>QDBMã¯POSIX互æ›ã®å…¨ã¦ã®ãƒ—ラットフォームã§å‹•ä½œã™ã‚‹ã“ã¨ã‚’目標ã¨ã—ã¦ã„る。ãŸã ã—ã€ã„ãã¤ã‹ã®APIãŒå®Ÿè£…ã•ã‚Œã¦ã„ãªã„プラットフォームã§ã‚‚動作ã™ã‚‹ã“ã¨ãŒæœ›ã¾ã—ã„。ã¾ãŸã€GCC以外ã®ã‚³ãƒ³ãƒ‘イラを利用ã—ã¦ã‚‚ビルドãŒã§ãã‚‹ã“ã¨ãŒæœ›ã¾ã—ã„。様々ãªãƒ—ラットフォームã¸ã®ç§»æ¤ä½œæ¥ã¯ã€æ–°ã—ã„ `Makefile' ã‚’è¿½åŠ ã—ãŸã‚Šã‚½ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ä¸€éƒ¨ã‚’ä¿®æ£ã—ãŸã‚Šã™ã‚‹ã“ã¨ã«ã‚ˆã£ã¦ãªã•ã‚Œã‚‹ã€‚C言語ã®APIã§ã‚ã‚Œã°ã€ãŠãらã以下ã®ãƒ•ã‚¡ã‚¤ãƒ«ã®ã„ãã¤ã‹ã‚’ä¿®æ£ã™ã‚‹ã“ã¨ã«ãªã‚‹ã€‚ã‚‚ã—ãã¯ãれらを基ã«æ–°ã—ã„ファイルを作ã£ã¦ã‚‚よã„。</p>
+
+<ul class="lines">
+<li><kbd>Makefile.in</kbd> : `./configure' ã«åˆ©ç”¨ã•ã‚Œã€`Makefile' ã®ãƒ™ãƒ¼ã‚¹ã¨ãªã‚‹ã€‚</li>
+<li><kbd>myconf.h</kbd> : システムä¾å˜ã®è¨å®šãƒ•ã‚¡ã‚¤ãƒ«ã€‚</li>
+<li><kbd>depot.h</kbd> : 基本APIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>curia.h</kbd> : æ‹¡å¼µAPIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>relic.h</kbd> : NDBM互æ›APIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>hovel.h</kbd> : GDBM互æ›APIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>cabin.h</kbd> : ユーティリティAPIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>villa.h</kbd> : 上級APIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>vista.h</kbd> : 拡張上級APIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>odeum.h</kbd> : 転置APIã®ãƒ˜ãƒƒãƒ€ã€‚</li>
+<li><kbd>myconf.c</kbd> : システムä¾å˜ã®å®Ÿè£…。</li>
+<li><kbd>depot.c</kbd> : 基本APIã®å®Ÿè£…。</li>
+<li><kbd>curia.c</kbd> : æ‹¡å¼µAPIã®å®Ÿè£…。</li>
+<li><kbd>relic.c</kbd> : NDBM互æ›APIã®å®Ÿè£…。</li>
+<li><kbd>hovel.c</kbd> : GDBM互æ›APIã®å®Ÿè£…。</li>
+<li><kbd>cabin.c</kbd> : ユーティリティAPIã®å®Ÿè£…。</li>
+<li><kbd>villa.c</kbd> : 上級APIã®å®Ÿè£…。</li>
+<li><kbd>vista.c</kbd> : 拡張上級APIã®å®Ÿè£…。</li>
+<li><kbd>odeum.c</kbd> : 転置APIã®å®Ÿè£…。</li>
+</ul>
+
+<p>`fcntl' コールã«ã‚ˆã‚‹ãƒ•ã‚¡ã‚¤ãƒ«ãƒãƒƒã‚¯ãŒã‚µãƒãƒ¼ãƒˆã•ã‚Œã¦ã„ãªã„プラットフォームã§ã¯ã€`Makefile' ã§å®šç¾©ã•ã‚Œã‚‹ `CFLAGS' マクãƒã« `-DMYNOLOCK' ã‚’è¿½åŠ ã™ã‚‹ã¨ã‚ˆã„。ãã®éš›ã«ã¯ãƒ—ãƒã‚»ã‚¹é–“ã®æŽ’他制御を行ã†åˆ¥ã®æ–¹æ³•ã‚’考ãˆã‚‹å¿…è¦ãŒã‚る。åŒæ§˜ã«ã€`mmap' コールãŒãªã„プラットフォームã§ã¯ã€`CFLAGS' ã« `-DMYNOMMAP' ã‚’è¿½åŠ ã™ã‚‹ã¨ã‚ˆã„。`mmap' ã«é–¢ã—ã¦ã¯ `malloc' ç‰ã‚’用ã„ãŸã‚¨ãƒŸãƒ¥ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ãŒç”¨æ„ã•ã‚Œã¦ã„る。ãã®ä»–ã®ã‚·ã‚¹ãƒ†ãƒ コールãŒå®Ÿè£…ã•ã‚Œã¦ã„ãªã„å ´åˆã¯ã€`myconf.h' 㨠`myconf.c' ã‚’ä¿®æ£ã—ã¦è©²å½“ã®ã‚·ã‚¹ãƒ†ãƒ コールã®ã‚¨ãƒŸãƒ¥ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ã‚’è¡Œãˆã°ã‚ˆã„。</p>
+
+<p>C++用ã®APIã§ã¯POSIXスレッドを使ã£ã¦ã„ã‚‹ã®ã§ã€ãã®ãƒ‘ッケージãŒå®Ÿè£…ã•ã‚Œã¦ã„ãªã„環境ã«ã¯C++用APIã¯ç§»æ¤ã§ããªã„。Java用ã®APIã§ã¯JNIを使ã£ã¦ã„ã‚‹ã®ã§ã€ãã®ãƒ˜ãƒƒãƒ€ã‚„ライブラリã®å ´æ‰€ã«æ³¨æ„ã™ã¹ãã§ã‚る。ã¾ãŸã€`long long' ã‚„ `int64' ã¨ã„ã£ãŸåž‹å®šç¾©ã«ã‚‚注æ„ã™ã¹ãã§ã‚る。Perlã‚„Ruby用ã®APIã§ã¯å„々ã®è¨€èªžå‡¦ç†ç³»ã§ç”¨æ„ã•ã‚ŒãŸãƒ“ルドコマンドを用ã„ã¦ã„ã‚‹ã®ã§ã€ãã®ä»•æ§˜ã«ç²¾é€šã™ã¹ãã§ã‚る。</p>
+
+<hr />
+
+<h2><a name="bugs" id="bugs" class="head">ãƒã‚°</a></h2>
+
+<p>QDBMã®å„文書ã¯è‹±èªžã‚’æ¯å›½èªžã¨ã™ã‚‹äººé”ã«ã‚ˆã£ã¦æ ¡æ£ã•ã‚Œã‚‹ã¹ãã§ã‚る。</p>
+
+<p>segmentation faultã«ã‚ˆã‚‹ã‚¯ãƒ©ãƒƒã‚·ãƒ¥ã€äºˆæœŸã›ã¬ãƒ‡ãƒ¼ã‚¿ã®æ¶ˆå¤±ç‰ã®ä¸æ•´åˆã€ãƒ¡ãƒ¢ãƒªãƒªãƒ¼ã‚¯ã€ãã®ä»–諸々ã®ãƒã‚°ã«é–¢ã—ã¦ã€æ—¢çŸ¥ã®ã‚‚ã®ã§æœªä¿®æ£ã®ã‚‚ã®ã¯ãªã„。</p>
+
+<p>ãƒã‚°ã‚’発見ã—ãŸã‚‰ã€æ˜¯éžã¨ã‚‚作者ã«ãƒ•ã‚£ãƒ¼ãƒ‰ãƒãƒƒã‚¯ã—ã¦ã»ã—ã„。ãã®éš›ã€QDBMã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã¨ã€åˆ©ç”¨ç’°å¢ƒã®OSã¨ã‚³ãƒ³ãƒ‘イラã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã‚‚æ•™ãˆã¦ã»ã—ã„。</p>
+
+<p>1.7.13よりå‰ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã®QDBMã§ä½œæˆã—ãŸãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯ã€ãれ以後ã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã¨äº’æ›æ€§ãŒãªã„。</p>
+
+<hr />
+
+<h2><a name="faq" id="faq" class="head">よãèžã‹ã‚Œã‚‹è³ªå•</a></h2>
+
+<dl>
+<dt>Q. : QDBMã¯SQLをサãƒãƒ¼ãƒˆã™ã‚‹ã‹ã€‚</dt>
+<dd>A. : QDBMã¯SQLをサãƒãƒ¼ãƒˆã—ãªã„。QDBMã¯RDBMS（関係データベース管ç†ã‚·ã‚¹ãƒ†ãƒ ）ã§ã¯ãªã„。組ã¿è¾¼ã¿ã®RDBMSを求ã‚ã‚‹ãªã‚‰ã€SQLiteãªã©ã‚’利用ã™ã‚‹ã¨ã‚ˆã„。</dd>
+<dt>Q. : çµå±€ã®ã¨ã“ã‚ã€GDBM（NDBMã€SDBMã€Berkeley DB）ã¨ã©ã†é•ã†ã®ã‹ã€‚</dt>
+<dd>A. : 処ç†ãŒé€Ÿã„。データベースファイルãŒå°ã•ã„。APIãŒç°¡æ½”ã§ã‚る。特ç†ã™ã¹ãã¯ã€ãƒ¬ã‚³ãƒ¼ãƒ‰ã®ä¸Šæ›¸ãã‚’ç¹°ã‚Šè¿”ã™å ´åˆã®æ™‚é–“çš„ãŠã‚ˆã³ç©ºé–“的効率ãŒã¨ã¦ã‚‚よãã€å®Ÿç”¨ä¸Šã®ã‚¹ã‚±ãƒ¼ãƒ©ãƒ“リティãŒé«˜ã„ã“ã¨ã§ã‚る。ã¾ãŸã€ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ãŒ100万を越ãˆã‚‹ã‚ˆã†ãªå¤§è¦æ¨¡ãªãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’構築ã™ã‚‹éš›ã«ã‚‚ã€å‡¦ç†ãŒæ¥µç«¯ã«é…ããªã£ãŸã‚Šã€ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºãŒæ¥µç«¯ã«å¤§ãããªã£ãŸã‚Šã—ãªã„。ã¨ã¯ã„ãˆã€ç”¨é€”ã«ã‚ˆã£ã¦ã¯ä»–ã®DBMã‚„DBMSを使ã†æ–¹ãŒé©åˆ‡ã‹ã‚‚ã—ã‚Œãªã„ã®ã§ã€å„自ã§æ€§èƒ½ã‚„機能ã®æ¯”較をã—ã¦ã¿ã¦ã»ã—ã„。</dd>
+<dt>Q. : å‚考文献ã¯ä½•ã‹ã€‚</dt>
+<dd>A. : QDBMã®å„種アルゴリズムã¯ã€ä¸»ã«Ahoä»–ã® `Data Structures and Algorithms'（邦題ã¯ã€Œãƒ‡ãƒ¼ã‚¿æ§‹é€ ã¨ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã€ï¼‰ãŠã‚ˆã³Sedgewickã® `Algorithms in C'（邦題ã¯ã€Œã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ Cã€ï¼‰ã®è¨˜è¿°ã«åŸºç¤Žã‚’ç½®ã„ã¦ã„る。</dd>
+<dt>Q. : ã©ã®APIを使ãˆã°ã‚ˆã„ã®ã‹ã€‚</dt>
+<dd>A. : レコードã®æ¤œç´¢ãŒå®Œå…¨ä¸€è‡´ã ã‘ã§æ¸ˆã‚€ã®ãªã‚‰ã€Depotを試ã™ã¨ã‚ˆã„。ãã®è¦æ¨¡ãŒå¤§ãã„ãªã‚‰ã€Curiaを試ã™ã¨ã‚ˆã„ã€‚ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’é †åºã«åŸºã¥ã„ã¦å‚ç…§ã—ãŸã„ãªã‚‰ã€Villaを試ã™ã¨ã‚ˆã„。ãã®è¦æ¨¡ãŒå¤§ãã„ãªã‚‰ã€Vistaを試ã™ã¨ã‚ˆã„。最大ã®ãƒ¬ã‚³ãƒ¼ãƒ‰æ•°ã‚’追求ã™ã‚‹ãªã‚‰ã€ZLIBã‹LZOを有効ã«ã—ã¦QDBMをビルドã—ã€ãã®ä¸Šã§Vistaを用ã„ã‚‹ã®ãŒã‚ˆã„。</dd>
+<dt>Q. : アプリケーションã®è‰¯ã„サンプルコードã¯ã‚ã‚‹ã‹ã€‚</dt>
+<dd>A. : å„APIã®ã‚³ãƒžãƒ³ãƒ‰ã®ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ã‚’å‚考ã«ã—ã¦ã»ã—ã„。`dptsv.c' 㨠`crtsv.c' 㨠`vltsv.c' ãŒæœ€ã‚‚ç°¡æ½”ã§ã‚る。</dd>
+<dt>Q. : データベースãŒå£Šã‚ŒãŸã®ã ãŒã€ã©ã†ã—ã¦ã‹ã€‚</dt>
+<dd>A. : 大抵ã®å ´åˆã€ã‚ãªãŸã®ã‚¢ãƒ—リケーションãŒãã¡ã‚“ã¨ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ã¦ã„ãªã„ã®ãŒåŽŸå› ã§ã‚る。デーモンプãƒã‚»ã‚¹ã§ã‚ã‚ã†ãŒã€CGIスクリプトã§ã‚ã‚ã†ãŒã€ã‚¢ãƒ—リケーションãŒçµ‚了ã™ã‚‹éš›ã«ã¯å¿…ãšãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’é–‰ã˜ãªã‘ã‚Œã°ãªã‚‰ãªã„。ãªãŠã€CGIã®ãƒ—ãƒã‚»ã‚¹ã¯SIGPIPEã‚„SIGTERMã«ã‚ˆã£ã¦æ®ºã•ã‚Œã‚‹ã“ã¨ãŒã‚ã‚‹ã“ã¨ã«ã‚‚ç•™æ„ã™ã¹ãã§ã‚る。</dd>
+<dt>Q. : QDBMã®ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã¯ã©ã®ãらã„å …ç‰¢ãªã®ã‹ã€‚</dt>
+<dd>A. : QDBMã¯çµ¶å¯¾çš„ãªå …牢性ã¯ä¿è¨¼ã—ãªã„。オペレーティングシステムãŒæš´èµ°ã—ãŸéš›ã«ã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãŒå£Šã‚Œã‚‹å¯èƒ½æ€§ãŒã‚る。Villaã®ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ã¯ã‚¢ãƒ—リケーションã®æš´èµ°ã‹ã‚‰ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚’ä¿è·ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ãŒã€ã‚ªãƒšãƒ¬ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã‚·ã‚¹ãƒ†ãƒ ã®æš´èµ°ã«ã¯å¯¾å‡¦ã§ããªã„。ã—ãŸãŒã£ã¦ã€ãƒŸãƒƒã‚·ãƒ§ãƒ³ã‚¯ãƒªãƒ†ã‚£ã‚«ãƒ«ãªç”¨é€”ã«QDBMを利用ã™ã‚‹å ´åˆã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã®å¤šé‡åŒ–を考慮ã™ã¹ãã§ã‚る。</dd>
+<dt>Q. : Depotã¨Curiaã®ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã®ä½¿ã„æ–¹ãŒã‚ˆãã‚ã‹ã‚‰ãªã„ãŒã€‚</dt>
+<dd>A. : 上書ãモードや連çµãƒ¢ãƒ¼ãƒ‰ã§ã®æ›¸ãè¾¼ã¿ã‚’ç¹°ã‚Šè¿”ã™å ´åˆã«ã€ã‚¢ãƒ©ã‚¤ãƒ³ãƒ¡ãƒ³ãƒˆã¯ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ãƒ•ã‚¡ã‚¤ãƒ«ã®ã‚µã‚¤ã‚ºãŒæ€¥æ¿€ã«å¤§ãããªã‚‹ã®ã‚’防ã。アラインメントã®é©åˆ‡ãªã‚µã‚¤ã‚ºã¯ã‚¢ãƒ—リケーションã«ã‚ˆã£ã¦ç•°ãªã‚‹ã®ã§ã€å„自ã§å®Ÿé¨“ã—ã¦ã¿ã¦ã»ã—ã„。ã•ã—ã‚ãŸã‚Šã¯32ãらã„ã«ã—ã¦ãŠãã¨ã‚ˆã„。</dd>
+<dt>Q. : Villaã®æ€§èƒ½ãƒ‘ラメータã®èª¿æ•´ãŒã‚ˆãã‚ã‹ã‚‰ãªã„ãŒã€‚</dt>
+<dd>A. : ãƒ¬ã‚³ãƒ¼ãƒ‰ã‚’é †ç•ªã«å‚ç…§ã™ã‚‹ã“ã¨ãŒå¤šã„ãªã‚‰ã°ã€`lrecmax' 㨠`nidxmax' をより大ããã—ãŸæ–¹ãŒã‚ˆã„。レコードを無作為ã«å‚ç…§ã™ã‚‹ã“ã¨ãŒå¤šã„ãªã‚‰ã°ã€ãれらã¯å°ã•ãã—ãŸæ–¹ãŒã‚ˆã„。RAMã«ä½™è£•ãŒã‚ã‚‹ãªã‚‰ã°ã€`lcnum' 㨠`ncnum' を増やã™ã¨æ€§èƒ½ãŒã‹ãªã‚Šå‘上ã™ã‚‹ã€‚ZLIBã‚„LZOã‚„BZIP2を有効化ã—ãŸå ´åˆã€`lrecmax' を大ããã—ãŸæ–¹ãŒåœ§ç¸®åŠ¹çŽ‡ãŒã‚ˆããªã‚‹ã€‚</dd>
+<dt>Q. : Villaã®åœ§ç¸®æ–¹å¼ã¨ã—ã¦ã¯ZLIBã¨LZOã¨BZIP2ã®ã©ã‚ŒãŒã‚ˆã„ã®ã‹ã€‚</dt>
+<dd>A. : 圧縮率ãŒæœ€ã‚‚良ã„ã®ã¯BZIP2ã§ã€å‡¦ç†é€Ÿåº¦ãŒæœ€ã‚‚高ã„ã®ã¯LZOã§ã€ZLIBã¯ãã®ä¸é–“çš„ãªç‰¹æ€§ã‚’æŒã¤ã€‚特ã«ç†ç”±ãŒãªã„é™ã‚Šã¯ZLIBを使ã†ã¨ã‚ˆã„。ãŸã ã—ã€æ›´æ–°ãŒé »ç¹ãªãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã«ã¯LZOãŒé©åˆ‡ã§ã€æ›´æ–°ãŒã»ã¨ã‚“ã©ãªã„ãªã‚‰ã°BZIP2ãŒé©åˆ‡ã§ã‚る。圧縮ã—ãªã„ã¨ã„ã†é¸æŠžè‚¢ã‚ˆã‚Šã¯LZOを使ã†æ–¹ãŒã‚ˆã„。LZOã®ãƒ©ã‚¤ã‚»ãƒ³ã‚¹ã¯GNU GPLã§ã‚ã‚‹ã“ã¨ã«æ³¨æ„ã™ã‚‹ã“ã¨ã€‚</dd>
+<dt>Q. : スパースファイルã¨ã¯ä½•ã‹ã€‚</dt>
+<dd>A. : ホール（一度もデータãŒæ›¸ãè¾¼ã¾ã‚Œã¦ã„ãªã„ブãƒãƒƒã‚¯ï¼‰ãŒã‚るファイルã®ã“ã¨ã§ã‚る。ファイルシステムãŒã‚¹ãƒ‘ースファイルをサãƒãƒ¼ãƒˆã—ã¦ã„ã‚‹å ´åˆã€ãƒ›ãƒ¼ãƒ«ã¯ç‰©ç†çš„ãªè¨˜æ†¶è£…ç½®ã«å‰²ã‚Šå½“ã¦ã‚‰ã‚Œãªã„。QDBMã§ã¯ã€DP_OSPARSEãªã©ã®ãƒ•ãƒ©ã‚°ã‚’用ã„ã‚‹ã¨ã€ãƒãƒƒã‚·ãƒ¥ã®ãƒã‚±ãƒƒãƒˆé…列ã¯åˆæœŸåŒ–ã•ã‚Œãšã«ãƒ›ãƒ¼ãƒ«ã¨ãªã‚‹ã€‚ã“ã®ã“ã¨ã‚’利用ã—ã¦ã€éžå¸¸ã«å·¨å¤§ãªãƒãƒƒã‚·ãƒ¥è¡¨ã‚’実ç¾ã™ã‚‹ã“ã¨ãŒã§ãる。ãŸã ã—ã€ãã®æ€§èƒ½ã¯ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ ã®è¨å®šã«å¼·ãä¾å˜ã™ã‚‹ã€‚</dd>
+<dt>Q. : ãªãœDepotã¨Curiaã«ã¯ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³æ©Ÿèƒ½ãŒãªã„ã®ã‹ã€‚</dt>
+<dd>A. : アプリケーションãŒç‹¬è‡ªã®ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³æ©Ÿèƒ½ã‚’実装ã—ã¦ã„ã‚‹å ´åˆã«ã¯ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹å†…部ã®ãƒˆãƒ©ãƒ³ã‚¶ã‚¯ã‚·ãƒ§ãƒ³ã¯é‚ªé”ã«ãªã‚‹ã‹ã‚‰ã§ã‚る。トランザクションã¯Cabinã®ãƒžãƒƒãƒ—を使ãˆã°ç°¡å˜ã«å®Ÿè£…ã§ãる。</dd>
+<dt>Q. : 性能を引ã出ã™ã‚·ã‚¹ãƒ†ãƒ ã®è¨å®šã¯ã©ã†ã§ã‚ã‚‹ã‹ã€‚</dt>
+<dd>A. : データベースã®ã‚µã‚¤ã‚ºã¨åŒç‰ä»¥ä¸Šã®RAMをマシンã«æ載ã™ã‚‹ã“ã¨ãŒæœ›ã¾ã—ã„。ãã—ã¦ã€I/Oãƒãƒƒãƒ•ã‚¡ã®ã‚µã‚¤ã‚ºã‚’大ããã—ã€ãƒ€ãƒ¼ãƒ†ã‚£ãƒãƒƒãƒ•ã‚¡ã‚’フラッシュã™ã‚‹é »åº¦ãŒå°‘ãªãã™ã‚‹ã‚ˆã†ã«è¨å®šã™ã‚‹ã¨ã‚ˆã„。ファイルシステムã®é¸æŠžã‚‚é‡è¦ã§ã‚る。Linux上ã§ã¯ã€é€šå¸¸ã¯EXT2ãŒæœ€é«˜é€Ÿã§ã‚ã‚‹ãŒã€EXT3ã® `writeback' モードã®æ–¹ãŒé€Ÿã„ã“ã¨ã‚‚ã‚る。ReiserFSã¯ãã‚Œãªã‚Šã§ã‚る。EXT3ã®ãã®ä»–ã®ãƒ¢ãƒ¼ãƒ‰ã¯ã‹ãªã‚Šé…ã„。他ã®ãƒ•ã‚¡ã‚¤ãƒ«ã‚·ã‚¹ãƒ†ãƒ ã«é–¢ã—ã¦ã¯å„自ã§å®Ÿé¨“ã—ã¦ã¿ã¦ã»ã—ã„。</dd>
+<dt>Q. : `gcc' ã®ä»£ã‚ã‚Šã« `cc' を使ã£ã¦ãƒ“ルドã§ãã‚‹ã‹ã€‚</dt>
+<dd>A. : `LTmakefile' を使ãˆã°ã§ãる。</dd>
+<dt>Q. : Visual C++を使ã£ã¦ãƒ“ルドã§ãã‚‹ã‹ã€‚</dt>
+<dd>A. : ã§ãる。`Makefile' ã®ä»£ã‚ã‚Šã« `VCmakefile' を使ã†ã“ã¨ã€‚</dd>
+<dt>Q. : ä»–ã«QDBMを利用ã§ãる言語ã¯ã‚ã‚‹ã‹ã€‚</dt>
+<dd>A. : PHPã€Scheme（Gauche）ã€OCaml用ã®ã‚¤ãƒ³ã‚¿ãƒ•ã‚§ãƒ¼ã‚¹ãŒæ—¢ã«å…¬é–‹ã•ã‚Œã¦ã„るよã†ã§ã‚る。ãれ以外ã®è¨€èªžã«ã¤ã„ã¦ã¯ã€å¿…è¦ãªã‚‰è‡ªåˆ†ã§ä½œã£ã¦ã»ã—ã„。</dd>
+<dt>Q. : 「QDBMã€ã¨ã¯ã©ã†ã„ã†æ„味ãªã®ã‹ã€‚</dt>
+<dd>A. : 「QDBMã€ã¯ã€ŒQuick Database Managerã€ã®ç•¥ã§ã‚る。高速ã«å‹•ä½œã™ã‚‹ã¨ã„ã†æ„味ã¨ã€ã‚¢ãƒ—リケーションã®é–‹ç™ºãŒè¿…速ã«ã§ãã‚‹ã¨ã„ã†æ„味ãŒè¾¼ã‚られã¦ã„る。</dd>
+<dt>Q. : å„APIã®åå‰ã¯ã©ã†ã„ã†æ„味ãªã®ã‹ã€‚ã©ã†ç™ºéŸ³ã™ã‚‹ã®ã‹ã€‚</dt>
+<dd>A. : 5æ–‡å—ã®è‹±å˜èªžã‹ã‚‰é©å½“ã«é¸æŠžã—ãŸã ã‘ã§ã€æ·±ã„æ„味ã¯ãªã„。ãªãŠã€ã€Œdepotã€ã¯ã€ç©ºæ¸¯ã€å€‰åº«ã€è£œçµ¦æ‰€ãªã©ã€ç‰©è³ªãŒé›†ã¾ã‚‹å ´æ‰€ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œãƒ‡ã‚£ãƒ¼ãƒã‚¥ã€ãŒå¦¥å½“ã ã‚ã†ã€‚「curiaã€ã¯ã€å®®å»·ã€æ³•å»·ãªã©ã€æ¨©å¨ãŒé›†ã¾ã‚‹å ´æ‰€ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œã‚ュリアã€ãŒå¦¥å½“ã ã‚ã†ã€‚「relicã€ã¯ã€éºç‰©ã€éºè·¡ãªã©ã€éŽåŽ»ã®æ®‹éª¸ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œãƒ¬ãƒªãƒƒã‚¯ã€ãŒå¦¥å½“ã ã‚ã†ã€‚「hovelã€ã¯ã€å°å±‹ã€ç‰©ç½®ã€é›¢ã‚Œå®¶ãªã©ã€ç²—末ãªå»ºç‰©ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œãƒãƒ´ãƒ«ã€ãŒå¦¥å½“ã ã‚ã†ã€‚「cabinã€ã¯ã€æ©Ÿå®¤ã€å®¢å®¤ã€å°å±‹ãªã©ã€ç°¡æ˜“çš„ãªå±…ä½ç©ºé–“ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œã‚ャビンã€ãŒå¦¥å½“ã ã‚ã†ã€‚「villaã€ã¯ã€åˆ¥è˜ã
€éƒŠå¤–ä½å®…ãªã©ã€éƒ½ä¼šé¢¨ã§ãªã„ä½å±…ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œãƒ´ã‚£ãƒ©ã€ãŒå¦¥å½“ã ã‚ã†ã€‚「vistaã€ã¯ã€äºˆæƒ³ã€å±•æœ›ãªã©ã€é ãを見渡ã™ã“ã¨ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œãƒ´ã‚£ã‚¹ã‚¿ã€ãŒå¦¥å½“ã ã‚ã†ã€‚「odeumã€ã¯ã€éŸ³æ¥½å ‚ã€åŠ‡å ´ãªã©ã€éŸ³æ¥½ã‚„è©©åŸã‚’è¡Œã†å»ºç‰©ã‚’æ„味ã™ã‚‹ã‚‰ã—ã„。発音を片仮åã§è¡¨ç¾ã™ã‚‹ãªã‚‰ã€Œã‚ªãƒ‡ã‚£ã‚¢ãƒ ã€ãŒå¦¥å½“ã ã‚ã†ã€‚</dd>
+</dl>
+
+<hr />
+
+<h2><a name="copying" id="copying" class="head">ライセンス</a></h2>
+
+<p>QDBMã¯ãƒ•ãƒªãƒ¼ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã§ã‚る。ã‚ãªãŸã¯ã€Free Software FoundationãŒå…¬è¡¨ã—ãŸGNU Lesser General Public Licenseã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³2.1ã‚ã‚‹ã„ã¯ãれ以é™ã®å„ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã®ä¸ã‹ã‚‰ã„ãšã‚Œã‹ã‚’é¸æŠžã—ã€ãã®ãƒãƒ¼ã‚¸ãƒ§ãƒ³ãŒå®šã‚ã‚‹æ¡é …ã«å¾“ã£ã¦QDBMã‚’å†é ’布ã¾ãŸã¯å¤‰æ›´ã™ã‚‹ã“ã¨ãŒã§ãる。</p>
+
+<p>QDBMã¯æœ‰ç”¨ã§ã‚ã‚‹ã¨æ€ã‚れるãŒã€é ’布ã«ã‚ãŸã£ã¦ã¯ã€å¸‚å ´æ€§åŠã³ç‰¹å®šç›®çš„é©åˆæ€§ã«ã¤ã„ã¦ã®æš—é»™ã®ä¿è¨¼ã‚’å«ã‚ã¦ã€ã„ã‹ãªã‚‹ä¿è¨¼ã‚‚è¡Œãªã‚ãªã„。詳細ã«ã¤ã„ã¦ã¯GNU Lesser General Public Licenseã‚’èªã‚“ã§ã»ã—ã„。</p>
+
+<p>ã‚ãªãŸã¯ã€QDBMã¨ä¸€ç·’ã«GNU Lesser General Public Licenseã®å†™ã—ã‚’å—ã‘å–ã£ã¦ã„ã‚‹ã¯ãšã§ã‚る（`COPYING' ファイルをå‚照）。ãã†ã§ãªã„å ´åˆã¯ã€Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA ã¸é€£çµ¡ã—ã¦ã»ã—ã„。</p>
+
+<p>QDBMã¯å¹³æž—幹雄ãŒä½œæˆã—ãŸã€‚作者ã¨é€£çµ¡ã‚’ã¨ã‚‹ã«ã¯ã€`mikio at users.sourceforge.net' å®›ã«é›»åメールをé€ã£ã¦ã»ã—ã„。ãŸã ã—ã€è³ªå•ã‚„ãƒã‚°ãƒ¬ãƒãƒ¼ãƒˆãªã©ã€ä»–ã®ãƒ¦ãƒ¼ã‚¶ã¨å…±æœ‰ã§ãる話題ã¯ãƒ¡ãƒ¼ãƒªãƒ³ã‚°ãƒªã‚¹ãƒˆã«é€ã£ã¦ã»ã—ã„。メーリングリストã®å‚åŠ æ–¹æ³•ã«ã¤ã„ã¦ã¯ã€`http://lists.sourceforge.net/lists/listinfo/qdbm-users' ã‚’å‚ç…§ã™ã‚‹ã“ã¨ã€‚</p>
+
+<hr />
+
+</body>
+
+</html>
+
+<!-- END OF FILE -->
Added: box/trunk/qdbm/spex.html
===================================================================
--- box/trunk/qdbm/spex.html (rev 0)
+++ box/trunk/qdbm/spex.html 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,4343 @@
+<?xml version="1.0" encoding="UTF-8"?>
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+
+<head>
+<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
+<meta http-equiv="Content-Style-Type" content="text/css" />
+<meta name="author" content="Mikio Hirabayashi" />
+<meta name="keywords" content="QDBM, DBM, database, hash, B+ tree" />
+<meta name="description" content="fundamental specifications of QDBM" />
+<link rel="contents" href="./" />
+<link rel="alternate" href="spex-ja.html" hreflang="ja" title="the Japanese version" />
+<link rev="made" href="mailto:mikio at users.sourceforge.net" />
+<title>Specifications of QDBM Version 1</title>
+<style type="text/css">html { margin: 0em 0em; padding: 0em 0em; background: #eeeeee none; }
+body { margin: 2em 2em; padding: 0em 0em;
+ background: #eeeeee none; color: #111111;
+ font-style: normal; font-weight: normal; }
+h1 { margin-top: 1.8em; margin-bottom: 1.3em; font-weight: bold; }
+h2 { margin-top: 1.8em; margin-bottom: 1.1em; font-weight: bold;
+ border-left: solid 0.6em #445555; border-bottom: solid 1pt #bbbbbb;
+ padding: 0.5em 0.5em; width: 60%; }
+h3 { margin-top: 1.8em; margin-bottom: 0.8em; font-weight: bold; }
+hr { margin-top: 2.5em; margin-bottom: 1.5em; height: 1pt;
+ color: #999999; background-color: #999999; border: none; }
+div.note,div.navi { text-align: right; }
+div.logo { text-align: center; margin: 3em 0em; }
+div.logo img { border: inset 2pt #ccccdd; }
+p { margin: 0.8em 0em; line-height: 140%; }
+p,dd { text-indent: 0.8em; }
+div,pre { margin-left: 1.7em; margin-right: 1.7em; }
+pre { background-color: #ddddee; padding: 0.2em; border: 1pt solid #bbbbcc; font-size: smaller; }
+kbd { color: #111111; font-style: normal; font-weight: bold; }
+a { color: #0022aa; text-decoration: none; }
+a:hover,a:focus { color: #0033ee; text-decoration: underline; }
+a.head { color: #111111; text-decoration: none; }
+table { padding: 1pt 2pt 1pt 2pt; border: none; margin-left: 1.7em; border-collapse: collapse; }
+th { padding: 1pt 4pt 1pt 4pt; border-style: none;
+ text-align: left; vertical-align: bottom; }
+td { padding: 1pt 4pt 1pt 4pt; border: 1pt solid #333333;
+ text-align: left; vertical-align: top; }
+ul,ol,dl { line-height: 140%; }
+dt { margin-left: 1.2em; }
+dd { margin-left: 2.0em; }
+ul.lines { list-style-type: none; }
+ at media print {
+ html,body { margin: 0em 0em; background-color: #ffffff; color: #000000; }
+ h1 { padding: 8em 0em 0.5em 0em; text-align: center; }
+ h2 { page-break-before: always; }
+ div.note { text-align: center; }
+ div.navi,div.logo { display: none }
+ hr { display: none; }
+ pre { margin: 0.8em 0.8em; background-color: #ffffff;
+ border: 1pt solid #aaaaaa; font-size: smaller; }
+ a,kbd { color: #000000; text-decoration: none; }
+ h1,h2,h3 { font-family: sans-serif; }
+ p,div,li,dt,dd { font-family: serif; }
+ pre,kbd { font-family: monospace; }
+ dd { font-size: smaller; }
+}
+</style>
+</head>
+
+<body>
+
+<h1>Fundamental Specifications of QDBM Version 1</h1>
+
+<div class="note">Copyright (C) 2000-2007 Mikio Hirabayashi</div>
+<div class="note">Last Update: Thu, 26 Oct 2006 15:00:20 +0900</div>
+<div class="navi">[<a href="spex-ja.html" hreflang="ja">Japanese</a>] [<a href="http://qdbm.sourceforge.net/">Home</a>]</div>
+
+<hr />
+
+<h2>Table of Contents</h2>
+
+<ol>
+<li><a href="#overview">Overview</a></li>
+<li><a href="#features">Features</a></li>
+<li><a href="#installation">Installation</a></li>
+<li><a href="#depotapi">Depot: Basic API</a></li>
+<li><a href="#depotcli">Commands for Depot</a></li>
+<li><a href="#curiaapi">Curia: Extended API</a></li>
+<li><a href="#curiacli">Commands for Curia</a></li>
+<li><a href="#relicapi">Relic: NDBM-compatible API</a></li>
+<li><a href="#reliccli">Commands for Relic</a></li>
+<li><a href="#hovelapi">Hovel: GDBM-compatible API</a></li>
+<li><a href="#hovelcli">Commands for Hovel</a></li>
+<li><a href="#cabinapi">Cabin: Utility API</a></li>
+<li><a href="#cabincli">Commands for Cabin</a></li>
+<li><a href="#villaapi">Villa: Advanced API</a></li>
+<li><a href="#villacli">Commands for Villa</a></li>
+<li><a href="#odeumapi">Odeum: Inverted API</a></li>
+<li><a href="#odeumcli">Commands for Odeum</a></li>
+<li><a href="#fileformat">File Format</a></li>
+<li><a href="#porting">Porting</a></li>
+<li><a href="#bugs">Bugs</a></li>
+<li><a href="#faq">Frequently Asked Questions</a></li>
+<li><a href="#copying">Copying</a></li>
+</ol>
+
+<hr />
+
+<h2><a name="overview" id="overview" class="head">Overview</a></h2>
+
+<p>QDBM is a library of routines for managing a database. The database is a simple data file containing records, each is a pair of a key and a value. Every key and value is serial bytes with variable length. Both binary data and character string can be used as a key and a value. There is neither concept of data tables nor data types. Records are organized in hash table or B+ tree.</p>
+
+<p>As for database of hash table, each key must be unique within a database, so it is impossible to store two or more records with a key overlaps. The following access methods are provided to the database: storing a record with a key and a value, deleting a record by a key, retrieving a record by a key. Moreover, traversal access to every key are provided, although the order is arbitrary. These access methods are similar to ones of DBM (or its followers: NDBM and GDBM) library defined in the UNIX standard. QDBM is an alternative for DBM because of its higher performance.</p>
+
+<p>As for database of B+ tree, records whose keys are duplicated can be stored. Access methods of storing, deleting, and retrieving are provided as with the database of hash table. Records are stored in order by a comparing function assigned by a user. It is possible to access each record with the cursor in ascending or descending order. According to this mechanism, forward matching search for strings and range search for integers are realized. Moreover, transaction is available in database of B+ tree.</p>
+
+<p>QDBM is written in C, and provided as APIs of C, C++, Java, Perl, and Ruby. QDBM is available on platforms which have API conforming to POSIX. QDBM is a free software licensed under the GNU Lesser General Public License.</p>
+
+<hr />
+
+<h2><a name="features" id="features" class="head">Features</a></h2>
+
+<h3>Effective Implementation of Hash Database</h3>
+
+<p>QDBM is developed referring to GDBM for the purpose of the following three points: higher processing speed, smaller size of a database file, and simpler API. They have been achieved. Moreover, as with GDBM, the following three restrictions of traditional DBM: a process can handle only one database, the size of a key and a value is bounded, a database file is sparse, are cleared.</p>
+
+<p>QDBM uses hash algorithm to retrieve records. If a bucket array has sufficient number of elements, the time complexity of retrieval is `O(1)'. That is, time required for retrieving a record is constant, regardless of the scale of a database. It is also the same about storing and deleting. Collision of hash values is managed by separate chaining. Data structure of the chains is binary search tree. Even if a bucket array has unusually scarce elements, the time complexity of retrieval is `O(log n)'.</p>
+
+<p>QDBM attains improvement in retrieval by loading RAM with the whole of a bucket array. If a bucket array is on RAM, it is possible to access a region of a target record by about one path of file operations. A bucket array saved in a file is not read into RAM with the `read' call but directly mapped to RAM with the `mmap' call. Therefore, preparation time on connecting to a database is very short, and two or more processes can share the same memory map.</p>
+
+<p>If the number of elements of a bucket array is about half of records stored within a database, although it depends on characteristic of the input, the probability of collision of hash values is about 56.7% (36.8% if the same, 21.3% if twice, 11.5% if four times, 6.0% if eight times). In such case, it is possible to retrieve a record by two or less paths of file operations. If it is made into a performance index, in order to handle a database containing one million of records, a bucket array with half a million of elements is needed. The size of each element is 4 bytes. That is, if 2M bytes of RAM is available, a database containing one million records can be handled.</p>
+
+<p>QDBM provides two modes to connect to a database: `reader' and `writer'. A reader can perform retrieving but neither storing nor deleting. A writer can perform all access methods. Exclusion control between processes is performed when connecting to a database by file locking. While a writer is connected to a database, neither readers nor writers can be connected. While a reader is connected to a database, other readers can be connect, but writers can not. According to this mechanism, data consistency is guaranteed with simultaneous connections in multitasking environment.</p>
+
+<p>Traditional DBM provides two modes of the storing operations: `insert' and `replace'. In the case a key overlaps an existing record, the insert mode keeps the existing value, while the replace mode transposes it to the specified value. In addition to the two modes, QDBM provides `concatenate' mode. In the mode, the specified value is concatenated at the end of the existing value and stored. This feature is useful when adding a element to a value as an array. Moreover, although DBM has a method to fetch out a value from a database only by reading the whole of a region of a record, QDBM has a method to fetch out a part of a region of a value. When a value is treated as an array, this feature is also useful.</p>
+
+<p>Generally speaking, while succession of updating, fragmentation of available regions occurs, and the size of a database grows rapidly. QDBM deal with this problem by coalescence of dispensable regions and reuse of them, and featuring of optimization of a database. When overwriting a record with a value whose size is greater than the existing one, it is necessary to remove the region to another position of the file. Because the time complexity of the operation depends on the size of the region of a record, extending values successively is inefficient. However, QDBM deal with this problem by alignment. If increment can be put in padding, it is not necessary to remove the region.</p>
+
+<p>As for many file systems, it is impossible to handle a file whose size is more than 2GB. To deal with this problem, QDBM provides a directory database containing multiple database files. Due to this feature, it is possible to handle a database whose total size is up to 1TB in theory. Moreover, because database files can be deployed on multiple disks, the speed of updating operations can be improved as with RAID-0 (striping). It is also possible for the database files to deploy on multiple file servers using NFS and so on.</p>
+
+<h3>Useful Implementation of B+ Tree Database</h3>
+
+<p>Although B+ tree database is slower than hash database, it features ordering access to each record. The order can be assigned by users. Records of B+ tree are sorted and arranged in logical pages. Sparse index organized in B tree that is multiway balanced tree are maintained for each page. Thus, the time complexity of retrieval and so on is `O(log n)'. Cursor is provided to access each record in order. The cursor can jump to a position specified by a key and can step forward or backward from the current position. Because each page is arranged as double linked list, the time complexity of stepping cursor is `O(1)'.</p>
+
+<p>B+ tree database is implemented, based on above hash database. Because each page of B+ tree is stored as each record of hash database, B+ tree database inherits efficiency of storage management of hash database. Because the header of each record is smaller and alignment of each page is adjusted according to the page size, in most cases, the size of database file is cut by half compared to one of hash database. Although operation of many pages are required to update B+ tree, QDBM expedites the process by caching pages and reducing file operations. In most cases, because whole of the sparse index is cached on memory, it is possible to retrieve a record by one or less path of file operations.</p>
+
+<p>B+ tree database features transaction mechanism. It is possible to commit a series of operations between the beginning and the end of the transaction in a lump, or to abort the transaction and perform rollback to the state before the transaction. Even if the process of an application is crashed while the transaction, the database file is not broken.</p>
+
+<p>In case that QDBM was built with ZLIB, LZO, or BZIP2 enabled, a lossless data-compression library, the content of each page of B+ tree is compressed and stored in a file. Because each record in a page has similar patterns, high efficiency of compression is expected due to the Lempel-Ziv algorithm and the like. In case handling text data, the size of a database is reduced to about 25%. If the scale of a database is large and disk I/O is the bottleneck, featuring compression makes the processing speed improved to a large extent.</p>
+
+<h3>Simple but Various Interfaces</h3>
+
+<p>QDBM provides very simple APIs. You can perform database I/O as usual file I/O with `FILE' pointer defined in ANSI C. In the basic API of QDBM, entity of a database is recorded as one file. In the extended API, entity of a database is recorded as several files in one directory. Because the two APIs are very similar with each other, porting an application from one to the other is easy.</p>
+
+<p>APIs which are compatible with NDBM and GDBM are also provided. As there are a lot of applications using NDBM or GDBM, it is easy to port them onto QDBM. In most cases, it is completed only by replacement of header including (#include) and re-compiling. However, QDBM can not handle database files made by the original NDBM or GDBM.</p>
+
+<p>In order to handle records on memory easily, the utility API is provided. It implements memory allocating functions, sorting functions, extensible datum, array list, hash map, and so on. Using them, you can handle records in C language cheaply as in such script languages as Perl or Ruby.</p>
+
+<p>B+ tree database is used with the advanced API. The advanced API is implemented using the basic API and the utility API. Because the advanced API is also similar to the basic API and the extended API, it is easy to learn how to use it.</p>
+
+<p>In order to handle an inverted index which is used by full-text search systems, the inverted API is provided. If it is easy to handle an inverted index of documents, an application can focus on text processing and natural language processing. Because this API does not depend on character codes nor languages, it is possible to implement a full-text search system which can respond to various requests from users.</p>
+
+<p>Along with APIs for C, QDBM provides APIs for C++, Java, Perl, and Ruby. APIs for C are composed of seven kinds: the basic API, the extended API, the NDBM-compatible API, the GDBM-compatible API, the utility API, the advanced API, and the inverted API. Command line interfaces corresponding to each API are also provided. They are useful for prototyping, testing, debugging, and so on. The C++ API encapsulates database handling functions of the basic API, the extended API, and the advanced API with class mechanism of C++. The Java API has native methods calling the basic API, the extended API, and the advanced API with Java Native Interface. The Perl API has methods calling the basic API, the extended API, and the advanced API with XS language. The Ruby API has method calling the basic API, the extended API, and the advanced API as modules of Ruby. Moreover, CGI scripts for administration of databases, file uploading, and full-text search are provided.</p>
+
+<h3>Wide Portability</h3>
+
+<p>QDBM is implemented being based on syntax of ANSI C (C89) and using only APIs defined in ANSI C or POSIX. Thus, QDBM works on most UNIX and its compatible OSs. As for C API, checking operations have been done at least on the following platforms.</p>
+
+<ul>
+<li>Linux (2.2, 2.4, 2.6) (IA32, IA64, AMD64, PA-RISC, Alpha, PowerPC, M68000, ARM)</li>
+<li>FreeBSD (4.9, 5.0, 5.1, 5.2, 5.3) (IA32, IA64, SPARC, Alpha)</li>
+<li>NetBSD (1.6) (IA32)</li>
+<li>OpenBSD (3.4) (IA32)</li>
+<li>SunOS (5.6, 5.7, 5.8, 5.9, 5.10) (IA32, SPARC)</li>
+<li>HP-UX (11.11, 11.23) (IA64, PA-RISC)</li>
+<li>AIX (5.2) (POWER)</li>
+<li>Windows (2000, XP) (IA32, IA64, AMD64) (Cygwin, MinGW, Visual C++)</li>
+<li>Mac OS X (10.2, 10.3, 10.4) (IA32, PowerPC)</li>
+<li>Tru64 (5.1) (Alpha)</li>
+<li>RISC OS (5.03) (ARM)</li>
+</ul>
+
+<p>Although a database file created by QDBM depends on byte order of the processor, to do with it, utilities to dump data in format which is independent to byte orders are provided.</p>
+
+<hr />
+
+<h2><a name="installation" id="installation" class="head">Installation</a></h2>
+
+<h3>Preparation</h3>
+
+<p>To install QDBM from a source package, GCC of 2.8 or later version and `make' are required.</p>
+
+<p>When an archive file of QDBM is extracted, change the current working directory to the generated directory and perform installation.</p>
+
+<h3>Usual Steps</h3>
+
+<p>Follow the procedures below on Linux, BSD, or SunOS.</p>
+
+<p>Run the configuration script.</p>
+
+<pre>./configure
+</pre>
+
+<p>Build programs.</p>
+
+<pre>make
+</pre>
+
+<p>Perform self-diagnostic test.</p>
+
+<pre>make check
+</pre>
+
+<p>Install programs. This operation must be carried out by the root user.</p>
+
+<pre>make install
+</pre>
+
+<h3>Using GNU Libtool</h3>
+
+<p>If above steps do not work, try the following steps. This way needs GNU Libtool of 1.5 or later version.</p>
+
+<p>Run the configuration script.</p>
+
+<pre>./configure
+</pre>
+
+<p>Build programs.</p>
+
+<pre>make -f LTmakefile
+</pre>
+
+<p>Perform self-diagnostic test.</p>
+
+<pre>make -f LTmakefile check
+</pre>
+
+<p>Install programs. This operation must be carried out by the root user.</p>
+
+<pre>make -f LTmakefile install
+</pre>
+
+<h3>Result</h3>
+
+<p>When a series of work finishes, the following files will be installed. As for the rest, manuals will be installed under `/usr/local/man/man1' and '/usr/local/man/man3', other documents will be installed under `/usr/local/share/qdbm'. A configuration file for `pkg-config' will be installed under `/usr/local/lib/pkgconfig'.</p>
+
+<pre>/usr/local/include/depot.h
+/usr/local/include/curia.h
+/usr/local/include/relic.h
+/usr/local/include/hovel.h
+/usr/local/include/cabin.h
+/usr/local/include/villa.h
+/usr/local/include/vista.h
+/usr/local/include/odeum.h
+/usr/local/lib/libqdbm.a
+/usr/local/lib/libqdbm.so.14.13.0
+/usr/local/lib/libqdbm.so.14
+/usr/local/lib/libqdbm.so
+/usr/local/bin/dpmgr
+/usr/local/bin/dptest
+/usr/local/bin/dptsv
+/usr/local/bin/crmgr
+/usr/local/bin/crtest
+/usr/local/bin/crtsv
+/usr/local/bin/rlmgr
+/usr/local/bin/rltest
+/usr/local/bin/hvmgr
+/usr/local/bin/hvtest
+/usr/local/bin/cbtest
+/usr/local/bin/cbcodec
+/usr/local/bin/vlmgr
+/usr/local/bin/vltest
+/usr/local/bin/vltsv
+/usr/local/bin/odmgr
+/usr/local/bin/odtest
+/usr/local/bin/odidx
+/usr/local/bin/qmttest
+</pre>
+
+<p>When you run a program linked dynamically to `libqdbm.so', the library search path should include `/usr/local/lib'. You can set the library search path with the environment variable `LD_LIBRARY_PATH'.</p>
+
+<p>To uninstall QDBM, execute the following command after `./configure'. This operation must be carried out by the root user.</p>
+
+<pre>make uninstall
+</pre>
+
+<p>If an old version of QDBM is installed on your system, uninstall it before installation of a new one.</p>
+
+<p>The other APIs except for C nor CGI scripts are not installed by default. Refer to `plus/xspex.html' to know how to install the C++ API. Refer to `java/jspex.html' to know how to install the Java API. Refer to `perl/plspex.html' to know how to install the Perl API. Refer to `ruby/rbspex.html' to know how to install the Ruby API. Refer to `cgi/cgispex.html' to know how to install the CGI script.</p>
+
+<p>To install QDBM from such a binary package as RPM, refer to the manual of the package manager. For example, if you use RPM, execute like the following command by the root user.</p>
+
+<pre>rpm -ivh qdbm-1.x.x-x.i386.rpm
+</pre>
+
+<h3>For Windows</h3>
+
+<p>On Windows (Cygwin), you should follow the procedures below for installation.</p>
+
+<p>Run the configuration script.</p>
+
+<pre>./configure
+</pre>
+
+<p>Build programs.</p>
+
+<pre>make win
+</pre>
+
+<p>Perform self-diagnostic test.</p>
+
+<pre>make check-win
+</pre>
+
+<p>Install programs. As well, perform `make uninstall-win' to uninstall them.</p>
+
+<pre>make install-win
+</pre>
+
+<p>On Windows, the import library `libqdbm.dll.a' is created as well as the static library `libqdbm.a', and the dynamic linking library `qdbm.dll' is created instead of such shared libraries as `libqdbm.so'. `qdbm.dll' is installed into `/usr/local/bin'.</p>
+
+<p>In order to build QDBM using MinGW on Cygwin, you should perform `make mingw' instead of `make win'. With the UNIX emulation layer of Cygwin, generated programs depend on `cygwin1.dll' (they come under GNU GPL). This problem is solved by linking them to the Win32 native DLL with MinGW.</p>
+
+<p>In order to build QDBM using Visual C++, you should edit `VCmakefile' and set the search paths for libraries and headers. And perform `nmake /f VCMakefile'. Applications linking to `qdbm.dll' should link to `msvcrt.dll' by `/MD' or `/MDd' option of the compiler. Refer to `VCmakefile' for detail configurations.</p>
+
+<h3>For Mac OS X</h3>
+
+<p>On Mac OS X (Darwin), you should follow the procedures below for installation.</p>
+
+<p>Run the configuration script.</p>
+
+<pre>./configure
+</pre>
+
+<p>Build programs.</p>
+
+<pre>make mac
+</pre>
+
+<p>Perform self-diagnostic test.</p>
+
+<pre>make check-mac
+</pre>
+
+<p>Install programs. As well, perform `make uninstall-mac' to uninstall them.</p>
+
+<pre>make install-mac
+</pre>
+
+<p>On Mac OS X, `libqdbm.dylib' and so on are created instead of `libqdbm.so' and so on. You can set the library search path with the environment variable `DYLD_LIBRARY_PATH'.</p>
+
+<h3>For HP-UX</h3>
+
+<p>On HP-UX, you should follow the procedures below for installation.</p>
+
+<p>Run the configuration script.</p>
+
+<pre>./configure
+</pre>
+
+<p>Build programs.</p>
+
+<pre>make hpux
+</pre>
+
+<p>Perform self-diagnostic test.</p>
+
+<pre>make check-hpux
+</pre>
+
+<p>Install programs. As well, perform `make uninstall-hpux' to uninstall them.</p>
+
+<pre>make install-hpux
+</pre>
+
+<p>On HP-UX, `libqdbm.sl' is created instead of `libqdbm.so' and so on. You can set the library search path with the environment variable `SHLIB_PATH'.</p>
+
+<h3>For RISC OS</h3>
+
+<p>On RISC OS, you should follow the procedures below for installation.</p>
+
+<p>Build programs. As `cc' is used for compilation by default, if you want to use `gcc', add the argument `CC=gcc'.</p>
+
+<pre>make -f RISCmakefile
+</pre>
+
+<p>When a series of work finishes, the library file `libqdbm' and such commands as `dpmgr' are generated. Because how to install them is not defined, copy them manually for installation. As with it, such header files as `depot.h' should be installed manually.</p>
+
+<h3>Detail Configurations</h3>
+
+<p>You can configure building processes by the following optional arguments of `./configure'.</p>
+
+<ul class="lines">
+<li><kbd>--enable-debug</kbd> : build for debugging. Enable debugging symbols, do not perform optimization, and perform static linking.</li>
+<li><kbd>--enable-devel</kbd> : build for development. Enable debugging symbols, perform optimization, and perform dynamic linking.</li>
+<li><kbd>--enable-stable</kbd> : build for stable release. Perform conservative optimization, and perform dynamic linking.</li>
+<li><kbd>--enable-pthread</kbd> : feature POSIX thread and treat global variables as thread specific data.</li>
+<li><kbd>--disable-lock</kbd> : build for environments without file locking.</li>
+<li><kbd>--disable-mmap</kbd> : build for environments without memory mapping.</li>
+<li><kbd>--enable-zlib</kbd> : feature ZLIB compression for B+ tree and inverted index.</li>
+<li><kbd>--enable-lzo</kbd> : feature LZO compression for B+ tree and inverted index.</li>
+<li><kbd>--enable-bzip</kbd> : feature BZIP2 compression for B+ tree and inverted index.</li>
+<li><kbd>--enable-iconv</kbd> : feature ICONV utilities for conversion of character encodings.</li>
+</ul>
+
+<p>Usually, QDBM and its applications can be built without any dependency on non-standard libraries except for `libqdbm.*'. However, they depend on `libpthread.*' if POSIX thread is enabled, and they depend on `libz.*' if ZLIB is enabled, and they depend on `liblzo2.*' if LZO is enabled, and they depend on `libbz2.*' if BZIP2 is enabled, and they depend on `libiconv.*' if ICONV is enabled.</p>
+
+<p>Because the license of LZO is GNU GPL, note that applications linking to `liblzo2.*' should meet commitments of GNU GPL.</p>
+
+<hr />
+
+<h2><a name="depotapi" id="depotapi" class="head">Depot: Basic API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Depot is the basic API of QDBM. Almost all features for managing a database provided by QDBM are implemented by Depot. Other APIs are no more than wrappers of Depot. Depot is the fastest in all APIs of QDBM.</p>
+
+<p>In order to use Depot, you should include `depot.h' and `stdlib.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>A pointer to `DEPOT' is used as a database handle. It is like that some file I/O routines of `stdio.h' use a pointer to `FILE'. A database handle is opened with the function `dpopen' and closed with `dpclose'. You should not refer directly to any member of the handle. If a fatal error occurs in a database, any access method via the handle except `dpclose' will not work and return error status. Although a process is allowed to use multiple database handles at the same time, handles of the same database file should not be used.</p>
+
+<h3>API</h3>
+
+<p>The external variable `dpversion' is the string containing the version information.</p>
+
+<dl>
+<dt><kbd>extern const char *dpversion;</kbd></dt>
+</dl>
+
+<p>The external variable `dpecode' is assigned with the last happened error code. Refer to `depot.h' for details of the error codes.</p>
+
+<dl>
+<dt><kbd>extern int dpecode;</kbd></dt>
+<dd>The initial value of this variable is `DP_ENOERR'. The other values are `DP_EFATAL', `DP_EMODE', `DP_EBROKEN', `DP_EKEEP', `DP_ENOITEM', `DP_EALLOC', `DP_EMAP', `DP_EOPEN', `DP_ECLOSE', `DP_ETRUNC', `DP_ESYNC', `DP_ESTAT', `DP_ESEEK', `DP_EREAD', `DP_EWRITE', `DP_ELOCK', `DP_EUNLINK', `DP_EMKDIR', `DP_ERMDIR', and `DP_EMISC'.</dd>
+</dl>
+
+<p>The function `dperrmsg' is used in order to get a message string corresponding to an error code.</p>
+
+<dl>
+<dt><kbd>const char *dperrmsg(int <var>ecode</var>);</kbd></dt>
+<dd>`ecode' specifies an error code. The return value is the message string of the error code. The region of the return value is not writable.</dd>
+</dl>
+
+<p>The function `dpopen' is used in order to get a database handle.</p>
+
+<dl>
+<dt><kbd>DEPOT *dpopen(const char *<var>name</var>, int <var>omode</var>, int <var>bnum</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. `omode' specifies the connection mode: `DP_OWRITER' as a writer, `DP_OREADER' as a reader. If the mode is `DP_OWRITER', the following may be added by bitwise or: `DP_OCREAT', which means it creates a new database if not exist, `DP_OTRUNC', which means it creates a new database regardless if one exists. Both of `DP_OREADER' and `DP_OWRITER' can be added to by bitwise or: `DP_ONOLCK', which means it opens a database file without file locking, or `DP_OLCKNB', which means locking is performed without blocking. `DP_OCREAT' can be added to by bitwise or: `DP_OSPARSE', which means it creates a database file as a sparse file. `bnum' specifies the number of elements of the bucket array. If it is not more than 0, the default value is specified. The size of a bucket array is determined on creating, and can not be changed except for by optimization of the database. Suggested size of a bucket array is about from 0.5 to 4 times of
the number of all records to store. The return value is the database handle or `NULL' if it is not successful. While connecting as a writer, an exclusive lock is invoked to the database file. While connecting as a reader, a shared lock is invoked to the database file. The thread blocks until the lock is achieved. If `DP_ONOLCK' is used, the application is responsible for exclusion control.</dd>
+</dl>
+
+<p>The function `dpclose' is used in order to close a database handle.</p>
+
+<dl>
+<dt><kbd>int dpclose(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is true, else, it is false. Because the region of a closed handle is released, it becomes impossible to use the handle. Updating a database is assured to be written when the handle is closed. If a writer opens a database but does not close it appropriately, the database will be broken.</dd>
+</dl>
+
+<p>The function `dpput' is used in order to store a record.</p>
+
+<dl>
+<dt><kbd>int dpput(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. `dmode' specifies behavior when the key overlaps, by the following values: `DP_DOVER', which means the specified value overwrites the existing one, `DP_DKEEP', which means the existing value is kept, `DP_DCAT', which means the specified value is concatenated at the end of the existing value. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `dpout' is used in order to delete a record.</p>
+
+<dl>
+<dt><kbd>int dpout(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is true, else, it is false. False is returned when no record corresponds to the specified key.</dd>
+</dl>
+
+<p>The function `dpget' is used in order to retrieve a record.</p>
+
+<dl>
+<dt><kbd>char *dpget(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `start' specifies the offset address of the beginning of the region of the value to be read. `max' specifies the max size to be read. If it is negative, the size to read is unlimited. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key or the size of the value of the corresponding record is less than `start'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return valu
e is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `dpgetwb' is used in order to retrieve a record and write the value into a buffer.</p>
+
+<dl>
+<dt><kbd>int dpgetwb(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, char *<var>vbuf</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `start' specifies the offset address of the beginning of the region of the value to be read. `max' specifies the max size to be read. It shuld be equal to or less than the size of the writing buffer. `vbuf' specifies the pointer to a buffer into which the value of the corresponding record is written. If successful, the return value is the size of the written data, else, it is -1. -1 is returned when no record corresponds to the specified key or the size of the value of the corresponding record is less than `start'. Note that no additional zero code is appended at the end of the region of the writing buffer.</dd>
+</dl>
+
+<p>The function `dpvsiz' is used in order to get the size of the value of a record.</p>
+
+<dl>
+<dt><kbd>int dpvsiz(DEPOT *<var>depot</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is the size of the value of the corresponding record, else, it is -1. Because this function does not read the entity of a record, it is faster than `dpget'.</dd>
+</dl>
+
+<p>The function `dpiterinit' is used in order to initialize the iterator of a database handle.</p>
+
+<dl>
+<dt><kbd>int dpiterinit(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is true, else, it is false. The iterator is used in order to access the key of every record stored in a database.</dd>
+</dl>
+
+<p>The function `dpiternext' is used in order to get the next key of the iterator.</p>
+
+<dl>
+<dt><kbd>char *dpiternext(DEPOT *<var>depot</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the next key, else, it is `NULL'. `NULL' is returned when no record is to be get out of the iterator. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. It is possible to access every record by iteration of calling this function. However, it is not assured if updating the database is occurred while the iteration. Besides, the order of this traversal access method is arbitrary, so it is not assured that the order of storing matches the one of the traversal access.</dd>
+</dl>
+
+<p>The function `dpsetalign' is used in order to set alignment of a database handle.</p>
+
+<dl>
+<dt><kbd>int dpsetalign(DEPOT *<var>depot</var>, int <var>align</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. `align' specifies the size of alignment. If successful, the return value is true, else, it is false. If alignment is set to a database, the efficiency of overwriting values is improved. The size of alignment is suggested to be average size of the values of the records to be stored. If alignment is positive, padding whose size is multiple number of the alignment is placed. If alignment is negative, as `vsiz' is the size of a value, the size of padding is calculated with `(vsiz / pow(2, abs(align) - 1))'. Because alignment setting is not saved in a database, you should specify alignment every opening a database.</dd>
+</dl>
+
+<p>The function `dpsetfbpsiz' is used in order to set the size of the free block pool of a database handle.</p>
+
+<dl>
+<dt><kbd>int dpsetfbpsiz(DEPOT *<var>depot</var>, int <var>size</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. `size' specifies the size of the free block pool of a database. If successful, the return value is true, else, it is false. The default size of the free block pool is 16. If the size is greater, the space efficiency of overwriting values is improved with the time efficiency sacrificed.</dd>
+</dl>
+
+<p>The function `dpsync' is used in order to synchronize updating contents with the file and the device.</p>
+
+<dl>
+<dt><kbd>int dpsync(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. This function is useful when another process uses the connected database file.</dd>
+</dl>
+
+<p>The function `dpoptimize' is used in order to optimize a database.</p>
+
+<dl>
+<dt><kbd>int dpoptimize(DEPOT *<var>depot</var>, int <var>bnum</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. `bnum' specifies the number of the elements of the bucket array. If it is not more than 0, the default value is specified. If successful, the return value is true, else, it is false. In an alternating succession of deleting and storing with overwrite or concatenate, dispensable regions accumulate. This function is useful to do away with them.</dd>
+</dl>
+
+<p>The function `dpname' is used in order to get the name of a database.</p>
+
+<dl>
+<dt><kbd>char *dpname(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is the pointer to the region of the name of the database, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `dpfsiz' is used in order to get the size of a database file.</p>
+
+<dl>
+<dt><kbd>int dpfsiz(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is the size of the database file, else, it is -1.</dd>
+</dl>
+
+<p>The function `dpbnum' is used in order to get the number of the elements of the bucket array.</p>
+
+<dl>
+<dt><kbd>int dpbnum(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is the number of the elements of the bucket array, else, it is -1.</dd>
+</dl>
+
+<p>The function `dpbusenum' is used in order to get the number of the used elements of the bucket array.</p>
+
+<dl>
+<dt><kbd>int dpbusenum(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is the number of the used elements of the bucket array, else, it is -1. This function is inefficient because it accesses all elements of the bucket array.</dd>
+</dl>
+
+<p>The function `dprnum' is used in order to get the number of the records stored in a database.</p>
+
+<dl>
+<dt><kbd>int dprnum(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. If successful, the return value is the number of the records stored in the database, else, it is -1.</dd>
+</dl>
+
+<p>The function `dpwritable' is used in order to check whether a database handle is a writer or not.</p>
+
+<dl>
+<dt><kbd>int dpwritable(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. The return value is true if the handle is a writer, false if not.</dd>
+</dl>
+
+<p>The function `dpfatalerror' is used in order to check whether a database has a fatal error or not.</p>
+
+<dl>
+<dt><kbd>int dpfatalerror(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. The return value is true if the database has a fatal error, false if not.</dd>
+</dl>
+
+<p>The function `dpinode' is used in order to get the inode number of a database file.</p>
+
+<dl>
+<dt><kbd>int dpinode(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. The return value is the inode number of the database file.</dd>
+</dl>
+
+<p>The function `dpmtime' is used in order to get the last modified time of a database.</p>
+
+<dl>
+<dt><kbd>time_t dpmtime(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. The return value is the last modified time of the database.</dd>
+</dl>
+
+<p>The function `dpfdesc' is used in order to get the file descriptor of a database file.</p>
+
+<dl>
+<dt><kbd>int dpfdesc(DEPOT *<var>depot</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. The return value is the file descriptor of the database file. Handling the file descriptor of a database file directly is not suggested.</dd>
+</dl>
+
+<p>The function `dpremove' is used in order to remove a database file.</p>
+
+<dl>
+<dt><kbd>int dpremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `dprepair' is used in order to repair a broken database file.</p>
+
+<dl>
+<dt><kbd>int dprepair(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. If successful, the return value is true, else, it is false. There is no guarantee that all records in a repaired database file correspond to the original or expected state.</dd>
+</dl>
+
+<p>The function `dpexportdb' is used in order to dump all records as endian independent data.</p>
+
+<dl>
+<dt><kbd>int dpexportdb(DEPOT *<var>depot</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`depot' specifies a database handle. `name' specifies the name of an output file. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `dpimportdb' is used in order to load all records from endian independent data.</p>
+
+<dl>
+<dt><kbd>int dpimportdb(DEPOT *<var>depot</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`depot' specifies a database handle connected as a writer. The database of the handle must be empty. `name' specifies the name of an input file. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `dpsnaffle' is used in order to retrieve a record directly from a database file.</p>
+
+<dl>
+<dt><kbd>char *dpsnaffle(const char *<var>name</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. Although this function can be used even while the database file is locked by another process, it is not assured that recent updated is reflect
ed.</dd>
+</dl>
+
+<p>The function `dpinnerhash' is a hash function used inside Depot.</p>
+
+<dl>
+<dt><kbd>int dpinnerhash(const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. The return value is the hash value of 31 bits length computed from the key. This function is useful when an application calculates the state of the inside bucket array.</dd>
+</dl>
+
+<p>The function `dpouterhash' is a hash function which is independent from the hash functions used inside Depot.</p>
+
+<dl>
+<dt><kbd>int dpouterhash(const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. The return value is the hash value of 31 bits length computed from the key. This function is useful when an application uses its own hash algorithm outside Depot.</dd>
+</dl>
+
+<p>The function `dpprimenum' is used in order to get a natural prime number not less than a number.</p>
+
+<dl>
+<dt><kbd>int dpprimenum(int <var>num</var>);</kbd></dt>
+<dd>`num' specified a natural number. The return value is a natural prime number not less than the specified number. This function is useful when an application determines the size of a bucket array of its own hash algorithm.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example stores and retrieves a phone number, using the name as the key.</p>
+
+<pre>#include <depot.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ DEPOT *depot;
+ char *val;
+
+ /* open the database */
+ if(!(depot = dpopen(DBNAME, DP_OWRITER | DP_OCREAT, -1))){
+ fprintf(stderr, "dpopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* store the record */
+ if(!dpput(depot, NAME, -1, NUMBER, -1, DP_DOVER)){
+ fprintf(stderr, "dpput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* retrieve the record */
+ if(!(val = dpget(depot, NAME, -1, 0, -1, NULL))){
+ fprintf(stderr, "dpget: %s\n", dperrmsg(dpecode));
+ } else {
+ printf("Name: %s\n", NAME);
+ printf("Number: %s\n", val);
+ free(val);
+ }
+
+ /* close the database */
+ if(!dpclose(depot)){
+ fprintf(stderr, "dpclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>The following example shows all records of the database.</p>
+
+<pre>#include <depot.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ DEPOT *depot;
+ char *key, *val;
+
+ /* open the database */
+ if(!(depot = dpopen(DBNAME, DP_OREADER, -1))){
+ fprintf(stderr, "dpopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* initialize the iterator */
+ if(!dpiterinit(depot)){
+ fprintf(stderr, "dpiterinit: %s\n", dperrmsg(dpecode));
+ }
+
+ /* scan with the iterator */
+ while((key = dpiternext(depot, NULL)) != NULL){
+ if(!(val = dpget(depot, key, -1, 0, -1, NULL))){
+ fprintf(stderr, "dpget: %s\n", dperrmsg(dpecode));
+ free(key);
+ break;
+ }
+ printf("%s: %s\n", key, val);
+ free(val);
+ free(key);
+ }
+
+ /* close the database */
+ if(!dpclose(depot)){
+ fprintf(stderr, "dpclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>For building a program using Depot, the program should be linked with a library file `libqdbm.a' or `libqdbm.so'. For example, the following command is executed to build `sample' from `sample.c'.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>If QDBM was built with POSIX thread enabled, the global variable `dpecode' is treated as thread specific data, and functions of Depot are reentrant. In that case, they are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<hr />
+
+<h2><a name="depotcli" id="depotcli" class="head">Commands for Depot</a></h2>
+
+<p>Depot has the following command line interfaces.</p>
+
+<p>The command `dpmgr' is a utility for debugging Depot and its applications. It features editing and checking of a database. It can be used for database applications with shell scripts. This command is used in the following format. `name' specifies a database name. `key' specifies the key of a record. `val' specifies the value of a record.</p>
+
+<dl>
+<dt><kbd>dpmgr create [-s] [-bnum <var>num</var>] <var>name</var></kbd></dt>
+<dd>Create a database file.</dd>
+<dt><kbd>dpmgr put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat] [-na] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>Store a record with a key and a value.</dd>
+<dt><kbd>dpmgr out [-kx|-ki] <var>name</var> <var>key</var></kbd></dt>
+<dd>Delete a record with a key.</dd>
+<dt><kbd>dpmgr get [-nl] [-kx|-ki] [-start <var>num</var>] [-max <var>num</var>] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record with a key and output it to the standard output.</dd>
+<dt><kbd>dpmgr list [-nl] [-k|-v] [-ox] <var>name</var></kbd></dt>
+<dd>List all keys and values delimited with tab and line-feed to the standard output.</dd>
+<dt><kbd>dpmgr optimize [-bnum <var>num</var>] [-na] <var>name</var></kbd></dt>
+<dd>Optimize a database.</dd>
+<dt><kbd>dpmgr inform [-nl] <var>name</var></kbd></dt>
+<dd>Output miscellaneous information to the standard output.</dd>
+<dt><kbd>dpmgr remove <var>name</var></kbd></dt>
+<dd>Remove a database file.</dd>
+<dt><kbd>dpmgr repair <var>name</var></kbd></dt>
+<dd>Repair a broken database file.</dd>
+<dt><kbd>dpmgr exportdb <var>name</var> <var>file</var></kbd></dt>
+<dd>Dump all records as endian independent data.</dd>
+<dt><kbd>dpmgr importdb [-bnum <var>num</var>] <var>name</var> <var>file</var></kbd></dt>
+<dd>Load all records from endian independent data.</dd>
+<dt><kbd>dpmgr snaffle [-kx|-ki] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record from a locked database with a key and output it to the standard output.</dd>
+<dt><kbd>dpmgr version</kbd></dt>
+<dd>Output version information of QDBM to the standard output.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : make the file sparse.</li>
+<li><kbd>-bnum <var>num</var></kbd> : specify the number of the elements of the bucket array.</li>
+<li><kbd>-kx</kbd> : treat `key' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-ki</kbd> : treat `key' as an integer expression of decimal notation.</li>
+<li><kbd>-vx</kbd> : treat `val' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vi</kbd> : treat `val' as an integer expression of decimal notation.</li>
+<li><kbd>-vf</kbd> : read the value from a file specified with `val'.</li>
+<li><kbd>-keep</kbd> : specify the storing mode for `DP_DKEEP'.</li>
+<li><kbd>-cat</kbd> : specify the storing mode for `DP_DCAT'.</li>
+<li><kbd>-na</kbd> : do not set alignment.</li>
+<li><kbd>-nl</kbd> : open the database without file locking.</li>
+<li><kbd>-start</kbd> : specify the beginning offset of a value to fetch.</li>
+<li><kbd>-max</kbd> : specify the max size of a value to fetch.</li>
+<li><kbd>-ox</kbd> : treat the output as a binary expression of hexadecimal notation.</li>
+<li><kbd>-n</kbd> : do not output the tailing newline.</li>
+<li><kbd>-k</kbd> : output keys only.</li>
+<li><kbd>-v</kbd> : output values only.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `dptest' is a utility for facility test and performance test. Check a database generated by the command or measure the execution time of the command. This command is used in the following format. `name' specifies a database name. `rnum' specifies the number of the records. `bnum' specifies the number of the elements of the bucket array. `pnum' specifies the number of patterns of the keys. `align' specifies the basic size of alignment. `fbpsiz' specifies the size of the free block pool.</p>
+
+<dl>
+<dt><kbd>dptest write [-s] <var>name</var> <var>rnum</var> <var>bnum</var></kbd></dt>
+<dd>Store records with keys of 8 bytes. They change as `00000001', `00000002'...</dd>
+<dt><kbd>dptest read [-wb] <var>name</var></kbd></dt>
+<dd>Retrieve all records of the database above.</dd>
+<dt><kbd>dptest rcat [-c] <var>name</var> <var>rnum</var> <var>bnum</var> <var>pnum</var> <var>align</var> <var>fbpsiz</var></kbd></dt>
+<dd>Store records with partway duplicated keys using concatenate mode.</dd>
+<dt><kbd>dptest combo <var>name</var></kbd></dt>
+<dd>Perform combination test of various operations.</dd>
+<dt><kbd>dptest wicked [-c] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Perform updating operations selected at random.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : make the file sparse.</li>
+<li><kbd>-wb</kbd> : use the function `dpgetwb' instead of the function `dpget'.</li>
+<li><kbd>-c</kbd> : perform comparison test with map of Cabin.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `dptsv' features mutual conversion between a database of Depot and a TSV text. This command is useful when data exchange with another version of QDBM or another DBM, or when data exchange between systems which have different byte orders. This command is used in the following format. `name' specifies a database name. The subcommand `export' reads TSV data from the standard input. If a key overlaps, the latter is adopted. `-bnum' specifies the number of the elements of the bucket array. The subcommand `import' writes TSV data to the standard output.</p>
+
+<dl>
+<dt><kbd>dptsv import [-bnum <var>num</var>] [-bin] <var>name</var></kbd></dt>
+<dd>Create a database from TSV.</dd>
+<dt><kbd>dptsv export [-bin] <var>name</var></kbd></dt>
+<dd>Write TSV data of a database.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-bnum <var>num</var></kbd> : specify the number of the elements of the bucket array.</li>
+<li><kbd>-bin</kbd> : treat records as Base64 format.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>Commands of Depot realize a simple database system. For example, to make a database to search `/etc/password' by a user name, perform the following command.</p>
+
+<pre>cat /etc/passwd | tr ':' '\t' | dptsv import casket
+</pre>
+
+<p>Thus, to retrieve the information of a user `mikio', perform the following command.</p>
+
+<pre>dpmgr get casket mikio
+</pre>
+
+<p>It is easy to implement functions upsides with these commands, using the API of Depot.</p>
+
+<hr />
+
+<h2><a name="curiaapi" id="curiaapi" class="head">Curia: Extended API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Curia is the extended API of QDBM. It provides routines for managing multiple database files in a directory. Restrictions of some file systems that the size of each file is limited are escaped by dividing a database file into two or more. If the database files deploy on multiple devices, the scalability is improved.</p>
+
+<p>Although Depot creates a database with a file name, Curia creates a database with a directory name. A database file named as `depot' is placed in the specified directory. Although it keeps the attribute of the database, it does not keep the entities of the records. Besides, sub directories are created by the number of division of the database, named with 4 digits. The database files are placed in the subdirectories. The entities of the records are stored in the database file. For example, in the case that a database directory named as `casket' and the number of division is 3, `casket/depot', `casket/0001/depot', `casket/0002/depot' and `casket/0003/depot' are created. No error occurs even if the namesake directory exists when creating a database. So, if sub directories exists and some devices are mounted on the sub directories, the database files deploy on the multiple devices.</p>
+
+<p>Curia features managing large objects. Although usual records are stored in some database files, records of large objects are stored in individual files. Because the files of large objects are deployed in different directories named with the hash values, the access speed is part-way robust although it is slower than the speed of usual records. Large and not often accessed data should be secluded as large objects. By doing this, the access speed of usual records is improved. The directory hierarchies of large objects are placed in the directory named as `lob' in the sub directories of the database. Because the key spaces of the usual records and the large objects are different, the operations keep out of each other.</p>
+
+<p>In order to use Curia, you should include `depot.h', `curia.h' and `stdlib.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <curia.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>A pointer to `CURIA' is used as a database handle. It is like that some file I/O routines of `stdio.h' use a pointer to `FILE'. A database handle is opened with the function `cropen' and closed with `crclose'. You should not refer directly to any member of the handle. If a fatal error occurs in a database, any access method via the handle except `crclose' will not work and return error status. Although a process is allowed to use multiple database handles at the same time, handles of the same database directory should not be used.</p>
+
+<p>Curia also assign the external variable `dpecode' with the error code. The function `dperrmsg' is used in order to get the message of the error code.</p>
+
+<h3>API</h3>
+
+<p>The function `cropen' is used in order to get a database handle.</p>
+
+<dl>
+<dt><kbd>CURIA *cropen(const char *<var>name</var>, int <var>omode</var>, int <var>bnum</var>, int <var>dnum</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory. `omode' specifies the connection mode: `CR_OWRITER' as a writer, `CR_OREADER' as a reader. If the mode is `CR_OWRITER', the following may be added by bitwise or: `CR_OCREAT', which means it creates a new database if not exist, `CR_OTRUNC', which means it creates a new database regardless if one exists. Both of `CR_OREADER' and `CR_OWRITER' can be added to by bitwise or: `CR_ONOLCK', which means it opens a database directory without file locking, or `CR_OLCKNB', which means locking is performed without blocking. `CR_OCREAT' can be added to by bitwise or: `CR_OSPARSE', which means it creates database files as sparse files. `bnum' specifies the number of elements of each bucket array. If it is not more than 0, the default value is specified. The size of each bucket array is determined on creating, and can not be changed except for by optimization of the database. Suggested size of each bucket array is about from 0.5
to 4 times of the number of all records to store. `dnum' specifies the number of division of the database. If it is not more than 0, the default value is specified. The number of division can not be changed from the initial value. The max number of division is 512. The return value is the database handle or `NULL' if it is not successful. While connecting as a writer, an exclusive lock is invoked to the database directory. While connecting as a reader, a shared lock is invoked to the database directory. The thread blocks until the lock is achieved. If `CR_ONOLCK' is used, the application is responsible for exclusion control.</dd>
+</dl>
+
+<p>The function `crclose' is used in order to close a database handle.</p>
+
+<dl>
+<dt><kbd>int crclose(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is true, else, it is false. Because the region of a closed handle is released, it becomes impossible to use the handle. Updating a database is assured to be written when the handle is closed. If a writer opens a database but does not close it appropriately, the database will be broken.</dd>
+</dl>
+
+<p>The function `crput' is used in order to store a record.</p>
+
+<dl>
+<dt><kbd>int crput(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. `dmode' specifies behavior when the key overlaps, by the following values: `CR_DOVER', which means the specified value overwrites the existing one, `CR_DKEEP', which means the existing value is kept, `CR_DCAT', which means the specified value is concatenated at the end of the existing value. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `crout' is used in order to delete a record.</p>
+
+<dl>
+<dt><kbd>int crout(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is true, else, it is false. False is returned when no record corresponds to the specified key.</dd>
+</dl>
+
+<p>The function `crget' is used in order to retrieve a record.</p>
+
+<dl>
+<dt><kbd>char *crget(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `start' specifies the offset address of the beginning of the region of the value to be read. `max' specifies the max size to be read. If it is negative, the size to read is unlimited. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key or the size of the value of the corresponding record is less than `start'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return valu
e is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `crgetwb' is used in order to retrieve a record and write the value into a buffer.</p>
+
+<dl>
+<dt><kbd>int crgetwb(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, char *<var>vbuf</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `start' specifies the offset address of the beginning of the region of the value to be read. `max' specifies the max size to be read. It shuld be equal to or less than the size of the writing buffer. `vbuf' specifies the pointer to a buffer into which the value of the corresponding record is written. If successful, the return value is the size of the written data, else, it is -1. -1 is returned when no record corresponds to the specified key or the size of the value of the corresponding record is less than `start'. Note that no additional zero code is appended at the end of the region of the writing buffer.</dd>
+</dl>
+
+<p>The function `crvsiz' is used in order to get the size of the value of a record.</p>
+
+<dl>
+<dt><kbd>int crvsiz(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is the size of the value of the corresponding record, else, it is -1. Because this function does not read the entity of a record, it is faster than `crget'.</dd>
+</dl>
+
+<p>The function `criterinit' is used in order to initialize the iterator of a database handle.</p>
+
+<dl>
+<dt><kbd>int criterinit(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is true, else, it is false. The iterator is used in order to access the key of every record stored in a database.</dd>
+</dl>
+
+<p>The function `criternext' is used in order to get the next key of the iterator.</p>
+
+<dl>
+<dt><kbd>char *criternext(CURIA *<var>curia</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the next key, else, it is `NULL'. `NULL' is returned when no record is to be get out of the iterator. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. It is possible to access every record by iteration of calling this function. However, it is not assured if updating the database is occurred while the iteration. Besides, the order of this traversal access method is arbitrary, so it is not assured that the order of storing matches the one of the traversal access.</dd>
+</dl>
+
+<p>The function `crsetalign' is used in order to set alignment of a database handle.</p>
+
+<dl>
+<dt><kbd>int crsetalign(CURIA *<var>curia</var>, int <var>align</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `align' specifies the size of alignment. If successful, the return value is true, else, it is false. If alignment is set to a database, the efficiency of overwriting values is improved. The size of alignment is suggested to be average size of the values of the records to be stored. If alignment is positive, padding whose size is multiple number of the alignment is placed. If alignment is negative, as `vsiz' is the size of a value, the size of padding is calculated with `(vsiz / pow(2, abs(align) - 1))'. Because alignment setting is not saved in a database, you should specify alignment every opening a database.</dd>
+</dl>
+
+<p>The function `crsetfbpsiz' is used in order to set the size of the free block pool of a database handle.</p>
+
+<dl>
+<dt><kbd>int crsetfbpsiz(CURIA *<var>curia</var>, int <var>size</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `size' specifies the size of the free block pool of a database. If successful, the return value is true, else, it is false. The default size of the free block pool is 16. If the size is greater, the space efficiency of overwriting values is improved with the time efficiency sacrificed.</dd>
+</dl>
+
+<p>The function `crsync' is used in order to synchronize updating contents with the files and the devices.</p>
+
+<dl>
+<dt><kbd>int crsync(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. This function is useful when another process uses the connected database directory.</dd>
+</dl>
+
+<p>The function `croptimize' is used in order to optimize a database.</p>
+
+<dl>
+<dt><kbd>int croptimize(CURIA *<var>curia</var>, int <var>bnum</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `bnum' specifies the number of the elements of each bucket array. If it is not more than 0, the default value is specified. In an alternating succession of deleting and storing with overwrite or concatenate, dispensable regions accumulate. This function is useful to do away with them.</dd>
+</dl>
+
+<p>The function `crname' is used in order to get the name of a database.</p>
+
+<dl>
+<dt><kbd>char *crname(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the pointer to the region of the name of the database, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `crfsiz' is used in order to get the total size of database files.</p>
+
+<dl>
+<dt><kbd>int crfsiz(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the total size of the database files, else, it is -1. If the total size is more than 2GB, the return value overflows.</dd>
+</dl>
+
+<p>The function `crfsizd' is used in order to get the total size of database files as double-precision floating-point number.</p>
+
+<dl>
+<dt><kbd>double crfsizd(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the total size of the database files, else, it is -1.0.</dd>
+</dl>
+
+<p>The function `crbnum' is used in order to get the total number of the elements of each bucket array.</p>
+
+<dl>
+<dt><kbd>int crbnum(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the total number of the elements of each bucket array, else, it is -1.</dd>
+</dl>
+
+<p>The function `crbusenum' is used in order to get the total number of the used elements of each bucket array.</p>
+
+<dl>
+<dt><kbd>int crbusenum(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the total number of the used elements of each bucket array, else, it is -1. This function is inefficient because it accesses all elements of each bucket array.</dd>
+</dl>
+
+<p>The function `crrnum' is used in order to get the number of the records stored in a database.</p>
+
+<dl>
+<dt><kbd>int crrnum(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the number of the records stored in the database, else, it is -1.</dd>
+</dl>
+
+<p>The function `crwritable' is used in order to check whether a database handle is a writer or not.</p>
+
+<dl>
+<dt><kbd>int crwritable(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. The return value is true if the handle is a writer, false if not.</dd>
+</dl>
+
+<p>The function `crfatalerror' is used in order to check whether a database has a fatal error or not.</p>
+
+<dl>
+<dt><kbd>int crfatalerror(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. The return value is true if the database has a fatal error, false if not.</dd>
+</dl>
+
+<p>The function `crinode' is used in order to get the inode number of a database directory.</p>
+
+<dl>
+<dt><kbd>int crinode(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. The return value is the inode number of the database directory.</dd>
+</dl>
+
+<p>The function `crmtime' is used in order to get the last modified time of a database.</p>
+
+<dl>
+<dt><kbd>time_t crmtime(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. The return value is the last modified time of the database.</dd>
+</dl>
+
+<p>The function `crremove' is used in order to remove a database directory.</p>
+
+<dl>
+<dt><kbd>int crremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `crrepair' is used in order to repair a broken database directory.</p>
+
+<dl>
+<dt><kbd>int crrepair(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory. If successful, the return value is true, else, it is false. There is no guarantee that all records in a repaired database directory correspond to the original or expected state.</dd>
+</dl>
+
+<p>The function `crexportdb' is used in order to dump all records as endian independent data.</p>
+
+<dl>
+<dt><kbd>int crexportdb(CURIA *<var>curia</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `name' specifies the name of an output directory. If successful, the return value is true, else, it is false. Note that large objects are ignored.</dd>
+</dl>
+
+<p>The function `crimportdb' is used in order to load all records from endian independent data.</p>
+
+<dl>
+<dt><kbd>int crimportdb(CURIA *<var>curia</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. The database of the handle must be empty. `name' specifies the name of an input directory. If successful, the return value is true, else, it is false. Note that large objects are ignored.</dd>
+</dl>
+
+<p>The function `crsnaffle' is used in order to retrieve a record directly from a database directory.</p>
+
+<dl>
+<dt><kbd>char *crsnaffle(const char *<var>name</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. Although this function can be used even while the database directory is locked by another process, it is not assured that recent updated
is reflected.</dd>
+</dl>
+
+<p>The function `crputlob' is used in order to store a large object.</p>
+
+<dl>
+<dt><kbd>int crputlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. `dmode' specifies behavior when the key overlaps, by the following values: `CR_DOVER', which means the specified value overwrites the existing one, `CR_DKEEP', which means the existing value is kept, `CR_DCAT', which means the specified value is concatenated at the end of the existing value. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `croutlob' is used in order to delete a large object.</p>
+
+<dl>
+<dt><kbd>int croutlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is true, else, it is false. false is returned when no large object corresponds to the specified key.</dd>
+</dl>
+
+<p>The function `crgetlob' is used in order to retrieve a large object.</p>
+
+<dl>
+<dt><kbd>char *crgetlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>start</var>, int <var>max</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `start' specifies the offset address of the beginning of the region of the value to be read. `max' specifies the max size to be read. If it is negative, the size to read is unlimited. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding large object, else, it is `NULL'. `NULL' is returned when no large object corresponds to the specified key or the size of the value of the corresponding large object is less than `start'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region
of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `crgetlobfd' is used in order to get the file descriptor of a large object.</p>
+
+<dl>
+<dt><kbd>int crgetlobfd(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is the file descriptor of the corresponding large object, else, it is -1. -1 is returned when no large object corresponds to the specified key. The returned file descriptor is opened with the `open' call. If the database handle was opened as a writer, the descriptor is writable (O_RDWR), else, it is not writable (O_RDONLY). The descriptor should be closed with the `close' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `crvsizlob' is used in order to get the size of the value of a large object.</p>
+
+<dl>
+<dt><kbd>int crvsizlob(CURIA *<var>curia</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is the size of the value of the corresponding large object, else, it is -1. Because this function does not read the entity of a large object, it is faster than `crgetlob'.</dd>
+</dl>
+
+<p>The function `crrnumlob' is used in order to get the number of the large objects stored in a database.</p>
+
+<dl>
+<dt><kbd>int crrnumlob(CURIA *<var>curia</var>);</kbd></dt>
+<dd>`curia' specifies a database handle. If successful, the return value is the number of the large objects stored in the database, else, it is -1.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example stores and retrieves a phone number, using the name as the key.</p>
+
+<pre>#include <depot.h>
+#include <curia.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ CURIA *curia;
+ char *val;
+
+ /* open the database */
+ if(!(curia = cropen(DBNAME, CR_OWRITER | CR_OCREAT, -1, -1))){
+ fprintf(stderr, "cropen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* store the record */
+ if(!crput(curia, NAME, -1, NUMBER, -1, CR_DOVER)){
+ fprintf(stderr, "crput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* retrieve the record */
+ if(!(val = crget(curia, NAME, -1, 0, -1, NULL))){
+ fprintf(stderr, "crget: %s\n", dperrmsg(dpecode));
+ } else {
+ printf("Name: %s\n", NAME);
+ printf("Number: %s\n", val);
+ free(val);
+ }
+
+ /* close the database */
+ if(!crclose(curia)){
+ fprintf(stderr, "crclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>The following example shows all records of the database.</p>
+
+<pre>#include <depot.h>
+#include <curia.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ CURIA *curia;
+ char *key, *val;
+
+ /* open the database */
+ if(!(curia = cropen(DBNAME, CR_OREADER, -1, -1))){
+ fprintf(stderr, "cropen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* initialize the iterator */
+ if(!criterinit(curia)){
+ fprintf(stderr, "criterinit: %s\n", dperrmsg(dpecode));
+ }
+
+ /* scan with the iterator */
+ while((key = criternext(curia, NULL)) != NULL){
+ if(!(val = crget(curia, key, -1, 0, -1, NULL))){
+ fprintf(stderr, "crget: %s\n", dperrmsg(dpecode));
+ free(key);
+ break;
+ }
+ printf("%s: %s\n", key, val);
+ free(val);
+ free(key);
+ }
+
+ /* close the iterator */
+ if(!crclose(curia)){
+ fprintf(stderr, "crclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>How to build programs using Curia is the same as the case of Depot.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>If QDBM was built with POSIX thread enabled, the global variable `dpecode' is treated as thread specific data, and functions of Curia are reentrant. In that case, they are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<hr />
+
+<h2><a name="curiacli" id="curiacli" class="head">Commands for Curia</a></h2>
+
+<p>Curia has the following command line interfaces.</p>
+
+<p>The command `crmgr' is a utility for debugging Curia and its applications. It features editing and checking of a database. It can be used for the database applications with shell scripts. This command is used in the following format. `name' specifies a database name. `key' specifies the key of a record. `val' specifies the value of a record.</p>
+
+<dl>
+<dt><kbd>crmgr create [-s] [-bnum <var>num</var>] [-dnum <var>num</var>] <var>name</var></kbd></dt>
+<dd>Create a database file.</dd>
+<dt><kbd>crmgr put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat] [-lob] [-na] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>Store a record with a key and a value.</dd>
+<dt><kbd>crmgr out [-kx|-ki] [-lob] <var>name</var> <var>key</var></kbd></dt>
+<dd>Delete a record with a key.</dd>
+<dt><kbd>crmgr get [-nl] [-kx|-ki] [-start <var>num</var>] [-max <var>num</var>] [-ox] [-lob] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record with a key and output it to the standard output.</dd>
+<dt><kbd>crmgr list [-nl] [-k|-v] [-ox] <var>name</var></kbd></dt>
+<dd>List all keys and values delimited with tab and line-feed to the standard output.</dd>
+<dt><kbd>crmgr optimize [-bnum <var>num</var>] [-na] <var>name</var></kbd></dt>
+<dd>Optimize a database.</dd>
+<dt><kbd>crmgr inform [-nl] <var>name</var></kbd></dt>
+<dd>Output miscellaneous information to the standard output.</dd>
+<dt><kbd>crmgr remove <var>name</var></kbd></dt>
+<dd>Remove a database directory.</dd>
+<dt><kbd>crmgr repair <var>name</var></kbd></dt>
+<dd>Repair a broken database directory.</dd>
+<dt><kbd>crmgr exportdb <var>name</var> <var>dir</var></kbd></dt>
+<dd>Dump all records as endian independent data.</dd>
+<dt><kbd>crmgr importdb [-bnum <var>num</var>] [-dnum <var>num</var>] <var>name</var> <var>dir</var></kbd></dt>
+<dd>Load all records from endian independent data.</dd>
+<dt><kbd>crmgr snaffle [-kx|-ki] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record from a locked database with a key and output it to the standard output.</dd>
+<dt><kbd>crmgr version</kbd></dt>
+<dd>Output version information of QDBM to the standard output.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : make the files sparse.</li>
+<li><kbd>-bnum <var>num</var></kbd> : specify the number of elements of each bucket array.</li>
+<li><kbd>-dnum <var>num</var></kbd> : specify the number of division of the database.</li>
+<li><kbd>-kx</kbd> : treat `key' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-ki</kbd> : treat `key' as an integer expression of decimal notation.</li>
+<li><kbd>-vx</kbd> : treat `val' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vi</kbd> : treat `val' as an integer expression of decimal notation.</li>
+<li><kbd>-vf</kbd> : read the value from a file specified with `val'.</li>
+<li><kbd>-keep</kbd> : specify the storing mode for `CR_DKEEP'.</li>
+<li><kbd>-cat</kbd> : specify the storing mode for `CR_DCAT'.</li>
+<li><kbd>-na</kbd> : do not set alignment.</li>
+<li><kbd>-nl</kbd> : open the database without file locking.</li>
+<li><kbd>-start</kbd> : specify the beginning offset of a value to fetch.</li>
+<li><kbd>-max</kbd> : specify the max size of a value to fetch.</li>
+<li><kbd>-ox</kbd> : treat the output as a binary expression of hexadecimal notation.</li>
+<li><kbd>-lob</kbd> : handle large objects.</li>
+<li><kbd>-n</kbd> : do not output the tailing newline.</li>
+<li><kbd>-k</kbd> : output keys only.</li>
+<li><kbd>-v</kbd> : output values only.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `crtest' is a utility for facility test and performance test. Check a database generated by the command or measure the execution time of the command. This command is used in the following format. `name' specifies a database name. `rnum' specifies the number of records. `bnum' specifies the number of elements of a bucket array. `dnum' specifies the number of division of a database. `pnum' specifies the number of patterns of the keys. `align' specifies the basic size of alignment. `fbpsiz' specifies the size of the free block pool.</p>
+
+<dl>
+<dt><kbd>crtest write [-s] [-lob] <var>name</var> <var>rnum</var> <var>bnum</var> <var>dnum</var></kbd></dt>
+<dd>Store records with keys of 8 bytes. They change as `00000001', `00000002'...</dd>
+<dt><kbd>crtest read [-wb] [-lob] <var>name</var></kbd></dt>
+<dd>Retrieve all records of the database above.</dd>
+<dt><kbd>crtest rcat [-c] <var>name</var> <var>rnum</var> <var>bnum</var> <var>dnum</var> <var>pnum</var> <var>align</var> <var>fbpsiz</var></kbd></dt>
+<dd>Store records with partway duplicated keys using concatenate mode.</dd>
+<dt><kbd>crtest combo <var>name</var></kbd></dt>
+<dd>Perform combination test of various operations.</dd>
+<dt><kbd>crtest wicked [-c] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Perform updating operations selected at random.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-s</kbd> : make the files sparse.</li>
+<li><kbd>-lob</kbd> : handle large objects.</li>
+<li><kbd>-wb</kbd> : use the function `crgetwb' instead of the function `crget'.</li>
+<li><kbd>-c</kbd> : perform comparison test with map of Cabin.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `crtsv' features mutual conversion between a database of Curia and a TSV text. This command is useful when data exchange with another version of QDBM or another DBM, or when data exchange between systems which have different byte orders. This command is used in the following format. `name' specifies a database name. The subcommand `export' reads TSV data from the standard input. If a key overlaps, the latter is adopted. `-bnum' specifies the number of the elements of the bucket array. `-dnum' specifies the number of division of the database. The subcommand `import' writes TSV data to the standard output.</p>
+
+<dl>
+<dt><kbd>crtsv import [-bnum <var>num</var>] [-dnum <var>num</var>] [-bin] <var>name</var></kbd></dt>
+<dd>Create a database from TSV.</dd>
+<dt><kbd>crtsv export [-bin] <var>name</var></kbd></dt>
+<dd>Write TSV data of a database.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-bnum <var>num</var></kbd> : specify the number of the elements of the bucket array.</li>
+<li><kbd>-dnum <var>num</var></kbd> : specify the number of division of the database.</li>
+<li><kbd>-bin</kbd> : treat records as Base64 format.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>Commands of Curia realize a simple database system. For example, to make a database to search `/etc/password' by a user name, perform the following command.</p>
+
+<pre>cat /etc/passwd | tr ':' '\t' | crtsv import casket
+</pre>
+
+<p>Thus, to retrieve the information of a user `mikio', perform the following command.</p>
+
+<pre>crmgr get casket mikio
+</pre>
+
+<p>It is easy to implement functions upsides with these commands, using the API of Curia.</p>
+
+<hr />
+
+<h2><a name="relicapi" id="relicapi" class="head">Relic: NDBM-compatible API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Relic is the API which is compatible with NDBM. So, Relic wraps functions of Depot as API of NDBM. It is easy to port an application from NDBM to QDBM. In most cases, you should only replace the includings of `ndbm.h' with `relic.h' and replace the linking option `-lndbm' with `-lqdbm'.</p>
+
+<p>The original NDBM treats a database as a pair of files. One, `a directory file', has a name with suffix `.dir' and stores a bit map of keys. The other, `a data file', has a name with suffix `.pag' and stores entities of each records. Relic creates the directory file as a mere dummy file and creates the data file as a database. Relic has no restriction about the size of each record. Relic can not handle database files made by the original NDBM.</p>
+
+<p>In order to use Relic, you should include `relic.h', `stdlib.h', `sys/types.h', `sys/stat.h' and `fcntl.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <relic.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+<dt><kbd>#include <sys/types.h></kbd></dt>
+<dt><kbd>#include <sys/stat.h></kbd></dt>
+<dt><kbd>#include <fcntl.h></kbd></dt>
+</dl>
+
+<p>A pointer to `DBM' is used as a database handle. A database handle is opened with the function `dbm_open' and closed with `dbm_close'. You should not refer directly to any member of a handle.</p>
+
+<h3>API</h3>
+
+<p>Structures of `datum' type is used in order to give and receive data of keys and values with functions of Relic.</p>
+
+<dl>
+<dt><kbd>typedef struct { void *dptr; size_t dsize; } datum;</kbd></dt>
+<dd>`dptr' specifies the pointer to the region of a key or a value. `dsize' specifies the size of the region.</dd>
+</dl>
+
+<p>The function `dbm_open' is used in order to get a database handle.</p>
+
+<dl>
+<dt><kbd>DBM *dbm_open(char *<var>name</var>, int <var>flags</var>, int <var>mode</var>);</kbd></dt>
+<dd>`name' specifies the name of a database. The file names are concatenated with suffixes. `flags' is the same as the one of `open' call, although `O_WRONLY' is treated as `O_RDWR' and additional flags except for `O_CREAT' and `O_TRUNC' have no effect. `mode' specifies the mode of the database file as the one of `open' call does. The return value is the database handle or `NULL' if it is not successful.</dd>
+</dl>
+
+<p>The function `dbm_close' is used in order to close a database handle.</p>
+
+<dl>
+<dt><kbd>void dbm_close(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. Because the region of the closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `dbm_store' is used in order to store a record.</p>
+
+<dl>
+<dt><kbd>int dbm_store(DBM *<var>db</var>, datum <var>key</var>, datum <var>content</var>, int <var>flags</var>);</kbd></dt>
+<dd>`db' specifies a database handle. `key' specifies a structure of a key. `content' specifies a structure of a value. `flags' specifies behavior when the key overlaps, by the following values: `DBM_REPLACE', which means the specified value overwrites the existing one, `DBM_INSERT', which means the existing value is kept. The return value is 0 if it is successful, 1 if it gives up because of overlaps of the key, -1 if other error occurs.</dd>
+</dl>
+
+<p>The function `dbm_delete' is used in order to delete a record.</p>
+
+<dl>
+<dt><kbd>int dbm_delete(DBM *<var>db</var>, datum <var>key</var>);</kbd></dt>
+<dd>`db' specifies a database handle. `key' specifies a structure of a key. The return value is 0 if it is successful, -1 if some errors occur.</dd>
+</dl>
+
+<p>The function `dbm_fetch' is used in order to retrieve a record.</p>
+
+<dl>
+<dt><kbd>datum dbm_fetch(DBM *<var>db</var>, datum <var>key</var>);</kbd></dt>
+<dd>`db' specifies a database handle. `key' specifies a structure of a key. The return value is a structure of the result. If a record corresponds, the member `dptr' of the structure is the pointer to the region of the value. If no record corresponds or some errors occur, `dptr' is `NULL'. `dptr' points to the region related with the handle. The region is available until the next time of calling this function with the same handle.</dd>
+</dl>
+
+<p>The function `dbm_firstkey' is used in order to get the first key of a database.</p>
+
+<dl>
+<dt><kbd>datum dbm_firstkey(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is a structure of the result. If a record corresponds, the member `dptr' of the structure is the pointer to the region of the first key. If no record corresponds or some errors occur, `dptr' is `NULL'. `dptr' points to the region related with the handle. The region is available until the next time of calling this function or the function `dbm_nextkey' with the same handle.</dd>
+</dl>
+
+<p>The function `dbm_nextkey' is used in order to get the next key of a database.</p>
+
+<dl>
+<dt><kbd>datum dbm_nextkey(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is a structure of the result. If a record corresponds, the member `dptr' of the structure is the pointer to the region of the next key. If no record corresponds or some errors occur, `dptr' is `NULL'. `dptr' points to the region related with the handle. The region is available until the next time of calling this function or the function `dbm_firstkey' with the same handle.</dd>
+</dl>
+
+<p>The function `dbm_error' is used in order to check whether a database has a fatal error or not.</p>
+
+<dl>
+<dt><kbd>int dbm_error(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is true if the database has a fatal error, false if not.</dd>
+</dl>
+
+<p>The function `dbm_clearerr' has no effect.</p>
+
+<dl>
+<dt><kbd>int dbm_clearerr(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is 0. The function is only for compatibility.</dd>
+</dl>
+
+<p>The function `dbm_rdonly' is used in order to check whether a handle is read-only or not.</p>
+
+<dl>
+<dt><kbd>int dbm_rdonly(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is true if the handle is read-only, or false if not read-only.</dd>
+</dl>
+
+<p>The function `dbm_dirfno' is used in order to get the file descriptor of a directory file.</p>
+
+<dl>
+<dt><kbd>int dbm_dirfno(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is the file descriptor of the directory file.</dd>
+</dl>
+
+<p>The function `dbm_pagfno' is used in order to get the file descriptor of a data file.</p>
+
+<dl>
+<dt><kbd>int dbm_pagfno(DBM *<var>db</var>);</kbd></dt>
+<dd>`db' specifies a database handle. The return value is the file descriptor of the data file.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example stores and retrieves a phone number, using the name as the key.</p>
+
+<pre>#include <relic.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <stdio.h>
+#include <string.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ DBM *db;
+ datum key, val;
+ int i;
+
+ /* open the database */
+ if(!(db = dbm_open(DBNAME, O_RDWR | O_CREAT, 00644))){
+ perror("dbm_open");
+ return 1;
+ }
+
+ /* prepare the record */
+ key.dptr = NAME;
+ key.dsize = strlen(NAME);
+ val.dptr = NUMBER;
+ val.dsize = strlen(NUMBER);
+
+ /* store the record */
+ if(dbm_store(db, key, val, DBM_REPLACE) != 0){
+ perror("dbm_store");
+ }
+
+ /* retrieve the record */
+ val = dbm_fetch(db, key);
+ if(val.dptr){
+ printf("Name: %s\n", NAME);
+ printf("Number: ");
+ for(i = 0; i < val.dsize; i++){
+ putchar(((char *)val.dptr)[i]);
+ }
+ putchar('\n');
+ } else {
+ perror("dbm_fetch");
+ }
+
+ /* close the database */
+ dbm_close(db);
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>How to build programs using Relic is the same as the case of Depot. Note that an option to be given to a linker is not `-lndbm', but `-lqdbm'.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>Functions of Relic are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<hr />
+
+<h2><a name="reliccli" id="reliccli" class="head">Commands for Relic</a></h2>
+
+<p>Relic has the following command line interfaces.</p>
+
+<p>The command `rlmgr' is a utility for debugging Relic and its applications. It features editing and checking of a database. It can be used for database applications with shell scripts. This command is used in the following format. `name' specifies a database name. `key' specifies the key of a record. `val' specifies the value of a record.</p>
+
+<dl>
+<dt><kbd>rlmgr create <var>name</var></kbd></dt>
+<dd>Create a database file.</dd>
+<dt><kbd>rlmgr store [-kx] [-vx|-vf] [-insert] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>Store a record with a key and a value.</dd>
+<dt><kbd>rlmgr delete [-kx] <var>name</var> <var>key</var></kbd></dt>
+<dd>Delete a record with a key.</dd>
+<dt><kbd>rlmgr fetch [-kx] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record with a key and output to the standard output.</dd>
+<dt><kbd>rlmgr list [-ox] <var>name</var></kbd></dt>
+<dd>List all keys and values delimited with tab and line-feed to the standard output.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-kx</kbd> : treat `key' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vx</kbd> : treat `val' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vf</kbd> : read the value from a file specified with `val'.</li>
+<li><kbd>-insert</kbd> : specify the storing mode for `DBM_INSERT'.</li>
+<li><kbd>-ox</kbd> : treat the output as a binary expression of hexadecimal notation.</li>
+<li><kbd>-n</kbd> : do not output the tailing newline.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>The command `rltest' is a utility for facility test and performance test. Check a database generated by the command or measure the execution time of the command. This command is used in the following format. `name' specifies a database name. `rnum' specifies the number of records.</p>
+
+<dl>
+<dt><kbd>rltest write <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Store records with keys of 8 bytes. They change as `00000001', `00000002'...</dd>
+<dt><kbd>rltest read <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Retrieve records of the database above.</dd>
+</dl>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<hr />
+
+<h2><a name="hovelapi" id="hovelapi" class="head">Hovel: GDBM-compatible API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Hovel is the API which is compatible with GDBM. So, Hovel wraps functions of Depot and Curia as API of GDBM. It is easy to port an application from GDBM to QDBM. In most cases, you should only replace the includings of `gdbm.h' with `hovel.h' and replace the linking option `-lgdbm' with `-lqdbm'. Hovel can not handle database files made by the original GDBM.</p>
+
+<p>In order to use Hovel, you should include `hovel.h', `stdlib.h', `sys/types.h' and `sys/stat.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <hovel.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+<dt><kbd>#include <sys/types.h></kbd></dt>
+<dt><kbd>#include <sys/stat.h></kbd></dt>
+</dl>
+
+<p>An object of `GDBM_FILE' is used as a database handle. A database handle is opened with the function `gdbm_open' and closed with `gdbm_close'. You should not refer directly to any member of a handle. Although Hovel works as a wrapper of Depot and handles a database file usually, if you use the function `gdbm_open2' to open the handle, it is possible to make behavior of a handle as a wrapper of Curia and treat a database directory.</p>
+
+<h3>API</h3>
+
+<p>Structures of `datum' type is used in order to give and receive data of keys and values with functions of Hovel.</p>
+
+<dl>
+<dt><kbd>typedef struct { char *dptr; size_t dsize; } datum;</kbd></dt>
+<dd>`dptr' specifies the pointer to the region of a key or a value. `dsize' specifies the size of the region.</dd>
+</dl>
+
+<p>The external variable `gdbm_version' is the string containing the version information.</p>
+
+<dl>
+<dt><kbd>extern char *gdbm_version;</kbd></dt>
+</dl>
+
+<p>The external variable `gdbm_errno' is assigned with the last happened error code. Refer to `hovel.h' for details of the error codes.</p>
+
+<dl>
+<dt><kbd>extern gdbm_error gdbm_errno;</kbd></dt>
+<dd>The initial value of this variable is `GDBM_NO_ERROR'. The other values are `GDBM_MALLOC_ERROR', `GDBM_BLOCK_SIZE_ERROR', `GDBM_FILE_OPEN_ERROR', `GDBM_FILE_WRITE_ERROR', `GDBM_FILE_SEEK_ERROR', `GDBM_FILE_READ_ERROR', `GDBM_BAD_MAGIC_NUMBER', `GDBM_EMPTY_DATABASE', `GDBM_CANT_BE_READER', `GDBM_CANT_BE_WRITER', `GDBM_READER_CANT_DELETE', `GDBM_READER_CANT_STORE', `GDBM_READER_CANT_REORGANIZE', `GDBM_UNKNOWN_UPDATE', `GDBM_ITEM_NOT_FOUND', `GDBM_REORGANIZE_FAILED', `GDBM_CANNOT_REPLACE', `GDBM_ILLEGAL_DATA', `GDBM_OPT_ALREADY_SET', and `GDBM_OPT_ILLEGAL'.</dd>
+</dl>
+
+<p>The function `gdbm_strerror' is used in order to get a message string corresponding to an error code.</p>
+
+<dl>
+<dt><kbd>char *gdbm_strerror(gdbm_error <var>gdbmerrno</var>);</kbd></dt>
+<dd>`gdbmerrno' specifies an error code. The return value is the message string of the error code. The region of the return value is not writable.</dd>
+</dl>
+
+<p>The function `gdbm_open' is used in order to get a database handle after the fashion of GDBM.</p>
+
+<dl>
+<dt><kbd>GDBM_FILE gdbm_open(char *<var>name</var>, int <var>block_size</var>, int <var>read_write</var>, int <var>mode</var>, void (*<var>fatal_func</var>)(void));</kbd></dt>
+<dd>`name' specifies the name of a database. `block_size' is ignored. `read_write' specifies the connection mode: `GDBM_READER' as a reader, `GDBM_WRITER', `GDBM_WRCREAT' and `GDBM_NEWDB' as a writer. `GDBM_WRCREAT' makes a database file or directory if it does not exist. `GDBM_NEWDB' makes a new database even if it exists. You can add the following to writer modes by bitwise or: `GDBM_SYNC', `GDBM_NOLOCK', `GDBM_LOCKNB', `GDBM_FAST', and `GDBM_SPARSE'. `GDBM_SYNC' means a database is synchronized after every updating method. `GDBM_NOLOCK' means a database is opened without file locking. `GDBM_LOCKNB' means file locking is performed without blocking. `GDBM_FAST' is ignored. `GDBM_SPARSE' is an original mode of QDBM and makes database a sparse file. `mode' specifies mode of a database file as the one of `open' call does. `fatal_func' is ignored. The return value is the database handle or `NULL' if it is not successful.</dd>
+</dl>
+
+<p>The function `gdbm_open2' is used in order to get a database handle after the fashion of QDBM.</p>
+
+<dl>
+<dt><kbd>GDBM_FILE gdbm_open2(char *<var>name</var>, int <var>read_write</var>, int <var>mode</var>, int <var>bnum</var>, int <var>dnum</var>, int <var>align</var>);</kbd></dt>
+<dd>`name' specifies the name of a database. `read_write' specifies the connection mode: `GDBM_READER' as a reader, `GDBM_WRITER', `GDBM_WRCREAT' and `GDBM_NEWDB' as a writer. `GDBM_WRCREAT' makes a database file or directory if it does not exist. `GDBM_NEWDB' makes a new database even if it exists. You can add the following to writer modes by bitwise or: `GDBM_SYNC', `GDBM_NOLOCK', 'GDBM_LOCKNB', `GDBM_FAST', and `GDBM_SPARSE'. `GDBM_SYNC' means a database is synchronized after every updating method. `GDBM_NOLOCK' means a database is opened without file locking. `GDBM_LOCKNB' means file locking is performed without blocking. `GDBM_FAST' is ignored. `GDBM_SPARSE' is an original mode of QDBM and makes database sparse files. `mode' specifies a mode of a database file or a database directory as the one of `open' or `mkdir' call does. `bnum' specifies the number of elements of each bucket array. If it is not more than 0, the default value is specified. `dnum' specif
ies the number of division of the database. If it is not more than 0, the returning handle is created as a wrapper of Depot, else, it is as a wrapper of Curia. `align' specifies the basic size of alignment. The return value is the database handle or `NULL' if it is not successful. If the database already exists, whether it is one of Depot or Curia is measured automatically.</dd>
+</dl>
+
+<p>The function `gdbm_close' is used in order to close a database handle.</p>
+
+<dl>
+<dt><kbd>void gdbm_close(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle. Because the region of the closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `gdbm_store' is used in order to store a record.</p>
+
+<dl>
+<dt><kbd>int gdbm_store(GDBM_FILE <var>dbf</var>, datum <var>key</var>, datum <var>content</var>, int <var>flag</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle connected as a writer. `key' specifies a structure of a key. `content' specifies a structure of a value. `flag' specifies behavior when the key overlaps, by the following values: `GDBM_REPLACE', which means the specified value overwrites the existing one, `GDBM_INSERT', which means the existing value is kept. The return value is 0 if it is successful, 1 if it gives up because of overlaps of the key, -1 if other error occurs.</dd>
+</dl>
+
+<p>The function `gdbm_delete' is used in order to delete a record.</p>
+
+<dl>
+<dt><kbd>int gdbm_delete(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle connected as a writer. `key' specifies a structure of a key. The return value is 0 if it is successful, -1 if some errors occur.</dd>
+</dl>
+
+<p>The function `gdbm_fetch' is used in order to retrieve a record.</p>
+
+<dl>
+<dt><kbd>datum gdbm_fetch(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle. `key' specifies a structure of a key. The return value is a structure of the result. If a record corresponds, the member `dptr' of the structure is the pointer to the region of the value. If no record corresponds or some errors occur, `dptr' is `NULL'. Because the region pointed to by `dptr' is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `gdbm_exists' is used in order to check whether a record exists or not.</p>
+
+<dl>
+<dt><kbd>int gdbm_exists(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle. `key' specifies a structure of a key. The return value is true if a record corresponds and no error occurs, or false, else, it is false.</dd>
+</dl>
+
+<p>The function `gdbm_firstkey' is used in order to get the first key of a database.</p>
+
+<dl>
+<dt><kbd>datum gdbm_firstkey(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle. The return value is a structure of the result. If a record corresponds, the member `dptr' of the structure is the pointer to the region of the first key. If no record corresponds or some errors occur, `dptr' is `NULL'. Because the region pointed to by `dptr' is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `gdbm_nextkey' is used in order to get the next key of a database.</p>
+
+<dl>
+<dt><kbd>datum gdbm_nextkey(GDBM_FILE <var>dbf</var>, datum <var>key</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle. The return value is a structure of the result. If a record corresponds, the member `dptr' of the structure is the pointer to the region of the next key. If no record corresponds or some errors occur, `dptr' is `NULL'. Because the region pointed to by `dptr' is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `gdbm_sync' is used in order to synchronize updating contents with the file and the device.</p>
+
+<dl>
+<dt><kbd>void gdbm_sync(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle connected as a writer.</dd>
+</dl>
+
+<p>The function `gdbm_reorganize' is used in order to reorganize a database.</p>
+
+<dl>
+<dt><kbd>int gdbm_reorganize(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle connected as a writer. If successful, the return value is 0, else -1.</dd>
+</dl>
+
+<p>The function `gdbm_fdesc' is used in order to get the file descriptor of a database file.</p>
+
+<dl>
+<dt><kbd>int gdbm_fdesc(GDBM_FILE <var>dbf</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle connected as a writer. The return value is the file descriptor of the database file. If the database is a directory the return value is -1.</dd>
+</dl>
+
+<p>The function `gdbm_setopt' has no effect.</p>
+
+<dl>
+<dt><kbd>int gdbm_setopt(GDBM_FILE <var>dbf</var>, int <var>option</var>, int *<var>value</var>, int <var>size</var>);</kbd></dt>
+<dd>`dbf' specifies a database handle. `option' is ignored. `size' is ignored. The return value is 0. The function is only for compatibility.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example stores and retrieves a phone number, using the name as the key.</p>
+
+<pre>#include <hovel.h>
+#include <stdlib.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <stdio.h>
+#include <string.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ GDBM_FILE dbf;
+ datum key, val;
+ int i;
+
+ /* open the database */
+ if(!(dbf = gdbm_open(DBNAME, 0, GDBM_WRCREAT, 00644, NULL))){
+ fprintf(stderr, "gdbm_open: %s\n", gdbm_strerror(gdbm_errno));
+ return 1;
+ }
+
+ /* prepare the record */
+ key.dptr = NAME;
+ key.dsize = strlen(NAME);
+ val.dptr = NUMBER;
+ val.dsize = strlen(NUMBER);
+
+ /* store the record */
+ if(gdbm_store(dbf, key, val, GDBM_REPLACE) != 0){
+ fprintf(stderr, "gdbm_store: %s\n", gdbm_strerror(gdbm_errno));
+ }
+
+ /* retrieve the record */
+ val = gdbm_fetch(dbf, key);
+ if(val.dptr){
+ printf("Name: %s\n", NAME);
+ printf("Number: ");
+ for(i = 0; i < val.dsize; i++){
+ putchar(val.dptr[i]);
+ }
+ putchar('\n');
+ free(val.dptr);
+ } else {
+ fprintf(stderr, "gdbm_fetch: %s\n", gdbm_strerror(gdbm_errno));
+ }
+
+ /* close the database */
+ gdbm_close(dbf);
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>How to build programs using Hovel is the same as the case of Depot. Note that an option to be given to a linker is not `-lgdbm', but `-lqdbm'.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>If QDBM was built with POSIX thread enabled, the global variable `gdbm_errno' is treated as thread specific data, and functions of Hovel are reentrant. In that case, they are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<hr />
+
+<h2><a name="hovelcli" id="hovelcli" class="head">Commands for Hovel</a></h2>
+
+<p>Hovel has the following command line interfaces.</p>
+
+<p>The command `hvmgr' is a utility for debugging Hovel and its applications. It features editing and checking of a database. It can be used for database applications with shell scripts. This command is used in the following format. `name' specifies a database name. `key' specifies the key of a record. `val' specifies the value of a record.</p>
+
+<dl>
+<dt><kbd>hvmgr create [-qdbm bnum dnum] [-s] <var>name</var></kbd></dt>
+<dd>Create a database file.</dd>
+<dt><kbd>hvmgr store [-qdbm] [-kx] [-vx|-vf] [-insert] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>Store a record with a key and a value.</dd>
+<dt><kbd>hvmgr delete [-qdbm] [-kx] <var>name</var> <var>key</var></kbd></dt>
+<dd>Delete a record with a key.</dd>
+<dt><kbd>hvmgr fetch [-qdbm] [-kx] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record with a key and output to the standard output.</dd>
+<dt><kbd>hvmgr list [-qdbm] [-ox] <var>name</var></kbd></dt>
+<dd>List all keys and values delimited with tab and line-feed to the standard output.</dd>
+<dt><kbd>hvmgr optimize [-qdbm] <var>name</var></kbd></dt>
+<dd>Optimize a database.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-qdbm [<var>bnum</var> <var>dnum</var>]</kbd> : use `gdbm_open2' to open the database. `bnum' specifies the number of the elements of the bucket array. `dnum' specifies the number of division of the database.</li>
+<li><kbd>-s</kbd> : make the file sparse.</li>
+<li><kbd>-kx</kbd> : treat `key' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vx</kbd> : treat `val' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vf</kbd> : read the value from a file specified with `val'.</li>
+<li><kbd>-insert</kbd> : specify the storing mode for `GDBM_INSERT'.</li>
+<li><kbd>-ox</kbd> : treat the output as a binary expression of hexadecimal notation.</li>
+<li><kbd>-n</kbd> : do not output the trailing newline.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>The command `hvtest' is a utility for facility test and performance test. Check a database generated by the command or measure the execution time of the command. This command is used in the following format. `name' specifies a database name. `rnum' specifies the number of records.</p>
+
+<dl>
+<dt><kbd>hvtest write [-qdbm] [-s] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Store records with keys of 8 bytes. They changes as `00000001', `00000002'...</dd>
+<dt><kbd>hvtest read [-qdbm] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Retrieve records of the database above.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-qdbm</kbd> : use `gdbm_open2' and open the handle as Curia.</li>
+<li><kbd>-s</kbd> : make the file sparse.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<hr />
+
+<h2><a name="cabinapi" id="cabinapi" class="head">Cabin: Utility API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Cabin is the utility API which provides memory allocating functions, sorting functions, extensible datum, array list, hash map, heap array, and so on for handling records easily on memory. This API features also parsing MIME, CSV, and XML, and features various types of encoding and decoding.</p>
+
+<p>In order to use Cabin, you should include `cabin.h' and `stdlib.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <cabin.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>A pointer to `CBDATUM' is used as a handle of an extensible datum. A datum handle is opened with the function `cbdatumopen' and closed with `cbdatumclose'. A pointer to `CBLIST' is used as a handle of an array list. A list handle is opened with the function `cblistopen' and closed with `cblistclose'. A pointer to `CBMAP' is used as a handle of a hash map. A map handle is opened with the function `cbmapopen' and closed with `cbmapclose'. A pointer to `CBHEAP' is used as a handle of a heap array. A heap handle is opened with the function `cbheapopen' and closed with `cbheapclose'. You should not refer directly to any member of each handles.</p>
+
+<h3>API</h3>
+
+<p>The external variable `cbfatalfunc' is the pointer to call back function for handling a fatal error.</p>
+
+<dl>
+<dt><kbd>extern void (*cbfatalfunc)(const char *);</kbd></dt>
+<dd>The argument specifies the error message. The initial value of this variable is `NULL'. If the value is `NULL', the default function is called when a fatal error occurs. A fatal error occurs when memory allocation is failed.</dd>
+</dl>
+
+<p>The function `cbmalloc' is used in order to allocate a region on memory.</p>
+
+<dl>
+<dt><kbd>void *cbmalloc(size_t <var>size</var>);</kbd></dt>
+<dd>`size' specifies the size of the region. The return value is the pointer to the allocated region. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbrealloc' is used in order to re-allocate a region on memory.</p>
+
+<dl>
+<dt><kbd>void *cbrealloc(void *<var>ptr</var>, size_t <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. The return value is the pointer to the re-allocated region. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmemdup' is used in order to duplicate a region on memory.</p>
+
+<dl>
+<dt><kbd>char *cbmemdup(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the pointer to the allocated region of the duplicate. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbfree' is used in order to free a region on memory.</p>
+
+<dl>
+<dt><kbd>void cbfree(void *<var>ptr</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. If it is `NULL', this function has no effect. Although this function is just a wrapper of `free' call, this is useful in applications using another package of the `malloc' series.</dd>
+</dl>
+
+<p>The function `cbglobalgc' is used in order to register the pointer or handle of an object to the global garbage collector.</p>
+
+<dl>
+<dt><kbd>void cbglobalgc(void *<var>ptr</var>, void (*<var>func</var>)(void *));</kbd></dt>
+<dd>`ptr' specifies the pointer or handle of an object. `func' specifies the pointer to a function to release resources of the object. Its argument is the pointer or handle of the object to release. This function assures that resources of an object are released when the process exits normally by returning from the `main' function or calling the `exit' function.</dd>
+</dl>
+
+<p>The function `cbggcsweep' is used in order to exercise the global garbage collector explicitly.</p>
+
+<dl>
+<dt><kbd>void cbggcsweep(void);</kbd></dt>
+<dd>Note that you should not use objects registered to the global garbage collector any longer after calling this function. Because the global garbage collector is initialized and you can register new objects into it.</dd>
+</dl>
+
+<p>The function `cbvmemavail' is used in order to check availability of allocation of the virtual memory.</p>
+
+<dl>
+<dt><kbd>int cbvmemavail(size_t <var>size</var>);</kbd></dt>
+<dd>`size' specifies the size of region to be allocated newly. The return value is true if allocation should be success, or false if not.</dd>
+</dl>
+
+<p>The function `cbisort' is used in order to sort an array using insert sort.</p>
+
+<dl>
+<dt><kbd>void cbisort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' specifies the pointer to an array. `nmemb' specifies the number of elements of the array. `size' specifies the size of each element. `compar' specifies the pointer to comparing function. The two arguments specify the pointers of elements. The comparing function should returns positive if the former is big, negative if the latter is big, 0 if both are equal. Insert sort is useful only if most elements have been sorted already.</dd>
+</dl>
+
+<p>The function `cbssort' is used in order to sort an array using shell sort.</p>
+
+<dl>
+<dt><kbd>void cbssort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' specifies the pointer to an array. `nmemb' specifies the number of elements of the array. `size' specifies the size of each element. `compar' specifies the pointer to comparing function. The two arguments specify the pointers of elements. The comparing function should returns positive if the former is big, negative if the latter is big, 0 if both are equal. If most elements have been sorted, shell sort may be faster than heap sort or quick sort.</dd>
+</dl>
+
+<p>The function `cbhsort' is used in order to sort an array using heap sort.</p>
+
+<dl>
+<dt><kbd>void cbhsort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' specifies the pointer to an array. `nmemb' specifies the number of elements of the array. `size' specifies the size of each element. `compar' specifies the pointer to comparing function. The two arguments specify the pointers of elements. The comparing function should returns positive if the former is big, negative if the latter is big, 0 if both are equal. Although heap sort is robust against bias of input, quick sort is faster in most cases.</dd>
+</dl>
+
+<p>The function `cbqsort' is used in order to sort an array using quick sort.</p>
+
+<dl>
+<dt><kbd>void cbqsort(void *<var>base</var>, int <var>nmemb</var>, int <var>size</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`base' specifies the pointer to an array. `nmemb' specifies the number of elements of the array. `size' specifies the size of each element. `compar' specifies the pointer to comparing function. The two arguments specify the pointers of elements. The comparing function should returns positive if the former is big, negative if the latter is big, 0 if both are equal. Being sensitive to bias of input, quick sort is the fastest sorting algorithm.</dd>
+</dl>
+
+<p>The function `cbstricmp' is used in order to compare two strings with case insensitive evaluation.</p>
+
+<dl>
+<dt><kbd>int cbstricmp(const char *<var>astr</var>, const char *<var>bstr</var>);</kbd></dt>
+<dd>`astr' specifies the pointer of one string. `astr' specifies the pointer of the other string. The return value is positive if the former is big, negative if the latter is big, 0 if both are equivalent. Upper cases and lower cases of alphabets in ASCII code are not distinguished.</dd>
+</dl>
+
+<p>The function `cbstrfwmatch' is used in order to check whether a string begins with a key.</p>
+
+<dl>
+<dt><kbd>int cbstrfwmatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a target string. `key' specifies the pointer of a forward matching key string. The return value is true if the target string begins with the key, else, it is false.</dd>
+</dl>
+
+<p>The function `cbstrfwimatch' is used in order to check whether a string begins with a key, with case insensitive evaluation.</p>
+
+<dl>
+<dt><kbd>int cbstrfwimatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a target string. `key' specifies the pointer of a forward matching key string. The return value is true if the target string begins with the key, else, it is false. Upper cases and lower cases of alphabets in ASCII code are not distinguished.</dd>
+</dl>
+
+<p>The function `cbstrbwmatch' is used in order to check whether a string ends with a key.</p>
+
+<dl>
+<dt><kbd>int cbstrbwmatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a target string. `key' specifies the pointer of a backward matching key string. The return value is true if the target string ends with the key, else, it is false.</dd>
+</dl>
+
+<p>The function `cbstrbwimatch' is used in order to check whether a string ends with a key, with case insensitive evaluation.</p>
+
+<dl>
+<dt><kbd>int cbstrbwimatch(const char *<var>str</var>, const char *<var>key</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a target string. `key' specifies the pointer of a backward matching key string. The return value is true if the target string ends with the key, else, it is false. Upper cases and lower cases of alphabets in ASCII code are not distinguished.</dd>
+</dl>
+
+<p>The function `cbstrstrkmp' is used in order to locate a substring in a string using KMP method.</p>
+
+<dl>
+<dt><kbd>char *cbstrstrkmp(const char *<var>haystack</var>, const char *<var>needle</var>);</kbd></dt>
+<dd>`haystack' specifies the pointer of a target string. `needle' specifies the pointer of a substring to be found. The return value is the pointer to the beginning of the substring or `NULL' if the substring is not found. In most cases, `strstr' as a built-in function of the compiler is faster than this function.</dd>
+</dl>
+
+<p>The function `cbstrstrkmp' is used in order to locate a substring in a string using BM method.</p>
+
+<dl>
+<dt><kbd>char *cbstrstrbm(const char *<var>haystack</var>, const char *<var>needle</var>);</kbd></dt>
+<dd>`haystack' specifies the pointer of a target string. `needle' specifies the pointer of a substring to be found. The return value is the pointer to the beginning of the substring or `NULL' if the substring is not found. In most cases, `strstr' as a built-in function of the compiler is faster than this function.</dd>
+</dl>
+
+<p>The function `cbstrtoupper' is used in order to convert the letters of a string to upper case.</p>
+
+<dl>
+<dt><kbd>char *cbstrtoupper(char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a string to convert. The return value is the pointer to the string.</dd>
+</dl>
+
+<p>The function `cbstrtolower' is used in order to convert the letters of a string to lower case.</p>
+
+<dl>
+<dt><kbd>char *cbstrtolower(char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a string to convert. The return value is the pointer to the string.</dd>
+</dl>
+
+<p>The function `cbstrtrim' is used in order to cut space characters at head or tail of a string.</p>
+
+<dl>
+<dt><kbd>char *cbstrtrim(char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a string to convert. The return value is the pointer to the string.</dd>
+</dl>
+
+<p>The function `cbstrsqzspc' is used in order to squeeze space characters in a string and trim it.</p>
+
+<dl>
+<dt><kbd>char *cbstrsqzspc(char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a string to convert. The return value is the pointer to the string.</dd>
+</dl>
+
+<p>The function `cbstrcountutf' is used in order to count the number of characters in a string of UTF-8.</p>
+
+<dl>
+<dt><kbd>int cbstrcountutf(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a string of UTF-8. The return value is the number of characters in the string.</dd>
+</dl>
+
+<p>The function `cbstrcututf' is used in order to cut a string of UTF-8 at the specified number of characters.</p>
+
+<dl>
+<dt><kbd>char *cbstrcututf(char *<var>str</var>, int <var>num</var>);</kbd></dt>
+<dd>`str' specifies the pointer of a string of UTF-8. `num' specifies the number of characters to be kept. The return value is the pointer to the string.</dd>
+</dl>
+
+<p>The function `cbdatumopen' is used in order to get a datum handle.</p>
+
+<dl>
+<dt><kbd>CBDATUM *cbdatumopen(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to the region of the initial content. If it is `NULL', an empty datum is created. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is a datum handle.</dd>
+</dl>
+
+<p>The function `cbdatumdup' is used in order to copy a datum.</p>
+
+<dl>
+<dt><kbd>CBDATUM *cbdatumdup(const CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. The return value is a new datum handle.</dd>
+</dl>
+
+<p>The function `cbdatumclose' is used in order to free a datum handle.</p>
+
+<dl>
+<dt><kbd>void cbdatumclose(CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. Because the region of a closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `cbdatumcat' is used in order to concatenate a datum and a region.</p>
+
+<dl>
+<dt><kbd>void cbdatumcat(CBDATUM *<var>datum</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. `ptr' specifies the pointer to the region to be appended. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'.</dd>
+</dl>
+
+<p>The function `cbdatumptr' is used in order to get the pointer of the region of a datum.</p>
+
+<dl>
+<dt><kbd>const char *cbdatumptr(const CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. The return value is the pointer of the region of a datum. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string.</dd>
+</dl>
+
+<p>The function `cbdatumsize' is used in order to get the size of the region of a datum.</p>
+
+<dl>
+<dt><kbd>int cbdatumsize(const CBDATUM *<var>datum</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. The return value is the size of the region of a datum.</dd>
+</dl>
+
+<p>The function `cbdatumsetsize' is used in order to change the size of the region of a datum.</p>
+
+<dl>
+<dt><kbd>void cbdatumsetsize(CBDATUM *<var>datum</var>, int <var>size</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. `size' specifies the new size of the region. If the new size is bigger than the one of old, the surplus region is filled with zero codes.</dd>
+</dl>
+
+<p>The function `cbdatumprintf' is used in order to perform formatted output into a datum.</p>
+
+<dl>
+<dt><kbd>void cbdatumprintf(CBDATUM *<var>datum</var>, const char *<var>format</var>, ...);</kbd></dt>
+<dd>`format' specifies a printf-like format string. The conversion character `%' can be used with such flag characters as `s', `d', `o', `u', `x', `X', `c', `e', `E', `f', `g', `G', `@', `?', `:', `%'. `@' works as with `s' but escapes meta characters of XML. `?' works as with `s' but escapes meta characters of URL. `:' works as with `s' but performs MIME encoding as UTF-8. The other conversion character work as with each original.</dd>
+</dl>
+
+<p>The function `cbdatumtomalloc' is used in order to convert a datum to an allocated region.</p>
+
+<dl>
+<dt><kbd>char *cbdatumtomalloc(CBDATUM *<var>datum</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`datum' specifies a datum handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the datum. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. Because the region of the original datum is released, it should not be released again.</dd>
+</dl>
+
+<p>The function `cblistopen' is used in order to get a list handle.</p>
+
+<dl>
+<dt><kbd>CBLIST *cblistopen(void);</kbd></dt>
+<dd>The return value is a list handle.</dd>
+</dl>
+
+<p>The function `cblistdup' is used in order to copy a list.</p>
+
+<dl>
+<dt><kbd>CBLIST *cblistdup(const CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' specifies a list handle. The return value is a new list handle.</dd>
+</dl>
+
+<p>The function `cblistclose' is used in order to close a list handle.</p>
+
+<dl>
+<dt><kbd>void cblistclose(CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' specifies a list handle. Because the region of a closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `cblistnum' is used in order to get the number of elements of a list.</p>
+
+<dl>
+<dt><kbd>int cblistnum(const CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' specifies a list handle. The return value is the number of elements of the list.</dd>
+</dl>
+
+<p>The function `cblistval' is used in order to get the pointer to the region of an element of a list.</p>
+
+<dl>
+<dt><kbd>const char *cblistval(const CBLIST *<var>list</var>, int <var>index</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `index' specifies the index of an element. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the element. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. If `index' is equal to or more than the number of elements, the return value is `NULL'.</dd>
+</dl>
+
+<p>The function `cblistpush' is used in order to add an element at the end of a list.</p>
+
+<dl>
+<dt><kbd>void cblistpush(CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `ptr' specifies the pointer to the region of an element. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'.</dd>
+</dl>
+
+<p>The function `cblistpop' is used in order to remove an element of the end of a list.</p>
+
+<dl>
+<dt><kbd>char *cblistpop(CBLIST *<var>list</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the value. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. If the list is empty, the return value is `NULL'.</dd>
+</dl>
+
+<p>The function `cblistunshift' is used in order to add an element at the top of a list.</p>
+
+<dl>
+<dt><kbd>void cblistunshift(CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `ptr' specifies the pointer to the region of an element. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'.</dd>
+</dl>
+
+<p>The function `cblistshift' is used in order to remove an element of the top of a list.</p>
+
+<dl>
+<dt><kbd>char *cblistshift(CBLIST *<var>list</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the value. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. If the list is empty, the return value is `NULL'.</dd>
+</dl>
+
+<p>The function `cblistinsert' is used in order to add an element at the specified location of a list.</p>
+
+<dl>
+<dt><kbd>void cblistinsert(CBLIST *<var>list</var>, int <var>index</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `index' specifies the index of an element. `ptr' specifies the pointer to the region of the element. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'.</dd>
+</dl>
+
+<p>The function `cblistremove' is used in order to remove an element at the specified location of a list.</p>
+
+<dl>
+<dt><kbd>char *cblistremove(CBLIST *<var>list</var>, int <var>index</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `index' specifies the index of an element. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the value. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. If `index' is equal to or more than the number of elements, no element is removed and the return value is `NULL'.</dd>
+</dl>
+
+<p>The function `cblistover' is used in order to overwrite an element at the specified location of a list.</p>
+
+<dl>
+<dt><kbd>void cblistover(CBLIST *<var>list</var>, int <var>index</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `index' specifies the index of an element. `ptr' specifies the pointer to the region of the new content. `size' specifies the size of the new content. If it is negative, the size is assigned with `strlen(ptr)'. If `index' is equal to or more than the number of elements, this function has no effect.</dd>
+</dl>
+
+<p>The function `cblistsort' is used in order to sort elements of a list in lexical order.</p>
+
+<dl>
+<dt><kbd>void cblistsort(CBLIST *<var>list</var>);</kbd></dt>
+<dd>`list' specifies a list handle. Quick sort is used for sorting.</dd>
+</dl>
+
+<p>The function `cblistlsearch' is used in order to search a list for an element using liner search.</p>
+
+<dl>
+<dt><kbd>int cblistlsearch(const CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `ptr' specifies the pointer to the region of a key. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the index of a corresponding element or -1 if there is no corresponding element. If two or more elements corresponds, the former returns.</dd>
+</dl>
+
+<p>The function `cblistbsearch' is used in order to search a list for an element using binary search.</p>
+
+<dl>
+<dt><kbd>int cblistbsearch(const CBLIST *<var>list</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`list' specifies a list handle. It should be sorted in lexical order. `ptr' specifies the pointer to the region of a key. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the index of a corresponding element or -1 if there is no corresponding element. If two or more elements corresponds, which returns is not defined.</dd>
+</dl>
+
+<p>The function `cblistdump' is used in order to serialize a list into a byte array.</p>
+
+<dl>
+<dt><kbd>char *cblistdump(const CBLIST *<var>list</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`list' specifies a list handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. The return value is the pointer to the region of the result serial region. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cblistload' is used in order to redintegrate a serialized list.</p>
+
+<dl>
+<dt><kbd>CBLIST *cblistload(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a byte array. `size' specifies the size of the region. The return value is a new list handle.</dd>
+</dl>
+
+<p>The function `cbmapopen' is used in order to get a map handle.</p>
+
+<dl>
+<dt><kbd>CBMAP *cbmapopen(void);</kbd></dt>
+<dd>The return value is a map handle.</dd>
+</dl>
+
+<p>The function `cbmapdup' is used in order to copy a map.</p>
+
+<dl>
+<dt><kbd>CBMAP *cbmapdup(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' specifies a map handle. The return value is a new map handle. The iterator of the source map is initialized.</dd>
+</dl>
+
+<p>The function `cbmapclose' is used in order to close a map handle.</p>
+
+<dl>
+<dt><kbd>void cbmapclose(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' specifies a map handle. Because the region of a closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `cbmapput' is used in order to store a record into a map.</p>
+
+<dl>
+<dt><kbd>int cbmapput(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>over</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. `over' specifies whether the value of the duplicated record is overwritten or not. If `over' is false and the key duplicated, the return value is false, else, it is true.</dd>
+</dl>
+
+<p>The function `cbmapputcat' is used in order to concatenate a value at the end of the value of the existing record.</p>
+
+<dl>
+<dt><kbd>void cbmapputcat(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. If there is no corresponding record, a new record is created.</dd>
+</dl>
+
+<p>The function `cbmapout' is used in order to delete a record of a map.</p>
+
+<dl>
+<dt><kbd>int cbmapout(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is true. False is returned when no record corresponds to the specified key.</dd>
+</dl>
+
+<p>The function `cbmapget' is used in order to retrieve a record of a map.</p>
+
+<dl>
+<dt><kbd>const char *cbmapget(const CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record. `NULL' is returned when no record corresponds. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string.</dd>
+</dl>
+
+<p>The function `cbmapmove' is used in order to move a record to the edge of a map.</p>
+
+<dl>
+<dt><kbd>int cbmapmove(CBMAP *<var>map</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>head</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `head' specifies the destination which is head if it is true or tail if else. If successful, the return value is true. False is returned when no record corresponds to the specified key.</dd>
+</dl>
+
+<p>The function `cbmapiterinit' is used in order to initialize the iterator of a map.</p>
+
+<dl>
+<dt><kbd>void cbmapiterinit(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' specifies a map handle. The iterator is used in order to access the key of every record stored in a map.</dd>
+</dl>
+
+<p>The function `cbmapiternext' is used in order to get the next key of the iterator of a map.</p>
+
+<dl>
+<dt><kbd>const char *cbmapiternext(CBMAP *<var>map</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the next key, else, it is `NULL'. `NULL' is returned when no record is to be get out of the iterator. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. The order of iteration is assured to be the same of the one of storing.</dd>
+</dl>
+
+<p>The function `cbmapiterval' is used in order to get the value binded to the key fetched from the iterator of a map.</p>
+
+<dl>
+<dt><kbd>const char *cbmapiterval(const char *<var>kbuf</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`kbuf' specifies the pointer to the region of a iteration key. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the value of the corresponding record. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string.</dd>
+</dl>
+
+<p>The function `cbmaprnum' is used in order to get the number of the records stored in a map.</p>
+
+<dl>
+<dt><kbd>int cbmaprnum(const CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' specifies a map handle. The return value is the number of the records stored in the map.</dd>
+</dl>
+
+<p>The function `cbmapkeys' is used in order to get the list handle contains all keys in a map.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbmapkeys(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' specifies a map handle. The return value is the list handle contains all keys in the map. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmapvals' is used in order to get the list handle contains all values in a map.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbmapvals(CBMAP *<var>map</var>);</kbd></dt>
+<dd>`map' specifies a map handle. The return value is the list handle contains all values in the map. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmapdump' is used in order to serialize a map into a byte array.</p>
+
+<dl>
+<dt><kbd>char *cbmapdump(const CBMAP *<var>map</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`map' specifies a map handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. The return value is the pointer to the region of the result serial region. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmapload' is used in order to redintegrate a serialized map.</p>
+
+<dl>
+<dt><kbd>CBMAP *cbmapload(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a byte array. `size' specifies the size of the region. The return value is a new map handle.</dd>
+</dl>
+
+<p>The function `cbmaploadone' is used in order to extract a record from a serialized map.</p>
+
+<dl>
+<dt><kbd>char *cbmaploadone(const char *<var>ptr</var>, int <var>size</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a byte array. `size' specifies the size of the region. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record. `NULL' is returned when no record corresponds. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string.</dd>
+</dl>
+
+
+<p>The function `cbheapopen' is used in order to get a heap handle.</p>
+
+<dl>
+<dt><kbd>CBHEAP *cbheapopen(int <var>size</var>, int <var>max</var>, int(*<var>compar</var>)(const void *, const void *));</kbd></dt>
+<dd>`size' specifies the size of each record. `max' specifies the maximum number of records in the heap. `compar' specifies the pointer to comparing function. The two arguments specify the pointers of records. The comparing function should returns positive if the former is big, negative if the latter is big, 0 if both are equal. The return value is a heap handle.</dd>
+</dl>
+
+<p>The function `cbheapdup' is used in order to copy a heap.</p>
+
+<dl>
+<dt><kbd>CBHEAP *cbheapdup(CBHEAP *<var>heap</var>);</kbd></dt>
+<dd>`heap' specifies a heap handle. The return value is a new heap handle.</dd>
+</dl>
+
+<p>The function `cbheapclose' is used in order to close a heap handle.</p>
+
+<dl>
+<dt><kbd>void cbheapclose(CBHEAP *<var>heap</var>);</kbd></dt>
+<dd>`heap' specifies a heap handle. Because the region of a closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `cbheapnum' is used in order to get the number of the records stored in a heap.</p>
+
+<dl>
+<dt><kbd>int cbheapnum(CBHEAP *<var>heap</var>);</kbd></dt>
+<dd>`heap' specifies a heap handle. The return value is the number of the records stored in the heap.</dd>
+</dl>
+
+<p>The function `cbheapinsert' is used in order to insert a record into a heap.</p>
+
+<dl>
+<dt><kbd>int cbheapinsert(CBHEAP *<var>heap</var>, const void *<var>ptr</var>);</kbd></dt>
+<dd>`heap' specifies a heap handle. `ptr' specifies the pointer to the region of a record. The return value is true if the record is added, else false. If the new record is bigger than the biggest existing regord, the new record is not added. If the new record is added and the number of records exceeds the maximum number, the biggest existing record is removed.</dd>
+</dl>
+
+<p>The function `cbheapval' is used in order to get the pointer to the region of a record in a heap.</p>
+
+<dl>
+<dt><kbd>void *cbheapval(CBHEAP *<var>heap</var>, int <var>index</var>);</kbd></dt>
+<dd>`heap' specifies a heap handle. `index' specifies the index of a record. The return value is the pointer to the region of the record. If `index' is equal to or more than the number of records, the return value is `NULL'. Note that records are organized by the nagative order the comparing function.</dd>
+</dl>
+
+<p>The function `cbheaptomalloc' is used in order to convert a heap to an allocated region.</p>
+
+<dl>
+<dt><kbd>void *cbheaptomalloc(CBHEAP *<var>heap</var>, int *<var>np</var>);</kbd></dt>
+<dd>`heap' specifies a heap handle. `np' specifies the pointer to a variable to which the number of records of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the heap. Records are sorted. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. Because the region of the original heap is released, it should not be released again.</dd>
+</dl>
+
+<p>The function `cbsprintf' is used in order to allocate a formatted string on memory.</p>
+
+<dl>
+<dt><kbd>char *cbsprintf(const char *<var>format</var>, ...);</kbd></dt>
+<dd>`format' specifies a printf-like format string. The conversion character `%' can be used with such flag characters as `d', `o', `u', `x', `X', `e', `E', `f', `g', `G', `c', `s', and `%'. Specifiers of the field length and the precision can be put between the conversion characters and the flag characters. The specifiers consist of decimal characters, `.', `+', `-', and the space character. The other arguments are used according to the format string. The return value is the pointer to the allocated region of the result string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbreplace' is used in order to replace some patterns in a string.</p>
+
+<dl>
+<dt><kbd>char *cbreplace(const char *<var>str</var>, CBMAP *<var>pairs</var>);</kbd></dt>
+<dd>`str' specifies the pointer to a source string. `pairs' specifies the handle of a map composed of pairs of replacement. The key of each pair specifies a pattern before replacement and its value specifies the pattern after replacement. The return value is the pointer to the allocated region of the result string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbsplit' is used in order to make a list by splitting a serial datum.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbsplit(const char *<var>ptr</var>, int <var>size</var>, const char *<var>delim</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to the region of the source content. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `delim' specifies a string containing delimiting characters. If it is `NULL', zero code is used as a delimiter. The return value is a list handle. If two delimiters are successive, it is assumed that an empty element is between the two. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose'.</dd>
+</dl>
+
+<p>The function `cbreadfile' is used in order to read whole data of a file.</p>
+
+<dl>
+<dt><kbd>char *cbreadfile(const char *<var>name</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`name' specifies the name of a file. If it is `NULL', the standard input is specified. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the allocated region of the read data. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbwritefile' is used in order to write a serial datum into a file.</p>
+
+<dl>
+<dt><kbd>int cbwritefile(const char *<var>name</var>, const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`name specifies the name of a file. If it is `NULL', the standard output is specified. `ptr' specifies the pointer to the region of the source content. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. If successful, the return value is true, else, it is false. If the file exists, it is overwritten. Else, a new file is created.</dd>
+</dl>
+
+<p>The function `cbreadlines' is used in order to read every line of a file.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbreadlines(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a file. If it is `NULL', the standard input is specified. The return value is a list handle of the lines if successful, else it is NULL. Line separators are cut out. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbdirlist' is used in order to read names of files in a directory.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbdirlist(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a directory. The return value is a list handle of names if successful, else it is NULL. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbfilestat' is used in order to get the status of a file or a directory.</p>
+
+<dl>
+<dt><kbd>int cbfilestat(const char *<var>name</var>, int *<var>isdirp</var>, int *<var>sizep</var>, int *<var>mtimep</var>);</kbd></dt>
+<dd>`name' specifies the name of a file or a directory. `dirp' specifies the pointer to a variable to which whether the file is a directory is assigned. If it is `NULL', it is not used. `sizep' specifies the pointer to a variable to which the size of the file is assigned. If it is `NULL', it is not used. `mtimep' specifies the pointer to a variable to which the last modified time of the file is assigned. If it is `NULL', it is not used. If successful, the return value is true, else, false. False is returned when the file does not exist or the permission is denied.</dd>
+</dl>
+
+<p>The function `cbremove' is used in order to remove a file or a directory and its sub ones recursively.</p>
+
+<dl>
+<dt><kbd>int cbremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a file or a directory. If successful, the return value is true, else, false. False is returned when the file does not exist or the permission is denied.</dd>
+</dl>
+
+<p>The function `cburlbreak' is used in order to break up a URL into elements.</p>
+
+<dl>
+<dt><kbd>CBMAP *cburlbreak(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to a string of URL. The return value is a map handle. Each key of the map is the name of an element. The key "self" specifies the URL itself. The key "scheme" specifies the scheme. The key "host" specifies the host of the server. The key "port" specifies the port number of the server. The key "authority" specifies the authority information. The key "path" specifies the path of the resource. The key "file" specifies the file name without the directory section. The key "query" specifies the query string. The key "fragment" specifies the fragment string. Supported schema are HTTP, HTTPS, FTP, and FILE. Absolute URL and relative URL are supported. Because the handle of the return value is opened with the function `cbmapopen', it should be closed with the function `cbmapclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The runction `cburlresolve' is used in order to resolve a relative URL with another absolute URL.</p>
+
+<dl>
+<dt><kbd>char *cburlresolve(const char *<var>base</var>, const char *<var>target</var>);</kbd></dt>
+<dd>`base' specifies an absolute URL of a base location. `target' specifies a URL to be resolved. The return value is a resolved URL. If the target URL is relative, a new URL of relative location from the base location is returned. Else, a copy of the target URL is returned. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cburlencode' is used in order to encode a serial object with URL encoding.</p>
+
+<dl>
+<dt><kbd>char *cburlencode(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the pointer to the result string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cburldecode' is used in order to decode a string encoded with URL encoding.</p>
+
+<dl>
+<dt><kbd>char *cburldecode(const char *<var>str</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`str' specifies the pointer to a source string. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the result. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbbaseencode' is used in order to encode a serial object with Base64 encoding.</p>
+
+<dl>
+<dt><kbd>char *cbbaseencode(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the pointer to the result string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbbasedecode' is used in order to decode a string encoded with Base64 encoding.</p>
+
+<dl>
+<dt><kbd>char *cbbasedecode(const char *<var>str</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`str' specifies the pointer to a source string. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the result. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbquoteencode' is used in order to encode a serial object with quoted-printable encoding.</p>
+
+<dl>
+<dt><kbd>char *cbquoteencode(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the pointer to the result string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbquotedecode' is used in order to decode a string encoded with quoted-printable encoding.</p>
+
+<dl>
+<dt><kbd>char *cbquotedecode(const char *<var>str</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`str' specifies the pointer to a source string. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer to the region of the result. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+
+<p>The function `cbmimebreak' is used in order to split a string of MIME into headers and the body.</p>
+
+<dl>
+<dt><kbd>char *cbmimebreak(const char *<var>ptr</var>, int <var>size</var>, CBMAP *<var>attrs</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to the region of MIME data. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `attrs' specifies a map handle to store attributes. If it is `NULL', it is not used. Each key of the map is an attribute name uncapitalized. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. The return value is the pointer of the body data. If the content type is defined, the attribute map has the key "TYPE" specifying the type. If the character encoding is defined, the key "CHARSET" specifies the encoding name. If the boundary string of multipart is defined, the key "BOUNDARY" specifies the string. If the content disposition is defined, the key "DISPOSITION" specifies the direction. If the file name is defined, the key "FILENAME" specifies the name. If the attribute name is defined, the key "NAME" specifies the nam
e. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmimeparts' is used in order to split multipart data of MIME into its parts.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbmimeparts(const char *<var>ptr</var>, int <var>size</var>, const char *<var>boundary</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to the region of multipart data of MIME. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `boundary' specifies the pointer to the region of the boundary string. The return value is a list handle. Each element of the list is the string of a part. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmimeencode' is used in order to encode a string with MIME encoding.</p>
+
+<dl>
+<dt><kbd>char *cbmimeencode(const char *<var>str</var>, const char *<var>encname</var>, int <var>base</var>);</kbd></dt>
+<dd>`str' specifies the pointer to a string. `encname' specifies a string of the name of the character encoding. The return value is the pointer to the result string. `base' specifies whether to use Base64 encoding. If it is false, quoted-printable is used. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbmimedecode' is used in order to decode a string encoded with MIME encoding.</p>
+
+<dl>
+<dt><kbd>char *cbmimedecode(const char *<var>str</var>, char *<var>enp</var>);</kbd></dt>
+<dd>`str' specifies the pointer to an encoded string. `enp' specifies the pointer to a region into which the name of encoding is written. If it is `NULL', it is not used. The size of the buffer should be equal to or more than 32 bytes. The return value is the pointer to the result string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbcsvrows' is used in order to split a string of CSV into rows.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbcsvrows(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of an CSV string. The return value is a list handle. Each element of the list is a string of a row. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use. The character encoding of the input string should be US-ASCII, UTF-8, ISO-8859-*, EUC-*, or Shift_JIS. Being compatible with MS-Excel, these functions for CSV can handle cells including such meta characters as comma, between double quotation marks.</dd>
+</dl>
+
+<p>The function `cbcsvcells' is used in order to split the string of a row of CSV into cells.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbcsvcells(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of a row of CSV. The return value is a list handle. Each element of the list is the unescaped string of a cell of the row. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbcsvescape' is used in order to escape a string with the meta characters of CSV.</p>
+
+<dl>
+<dt><kbd>char *cbcsvescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of a string. The return value is the pointer to the escaped string sanitized of meta characters. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbcsvunescape' is used in order to unescape a string with the escaped meta characters of CSV.</p>
+
+<dl>
+<dt><kbd>char *cbcsvunescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of a string with meta characters. The return value is the pointer to the unescaped string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbxmlbreak' is used in order to split a string of XML into tags and text sections.</p>
+
+<dl>
+<dt><kbd>CBLIST *cbxmlbreak(const char *<var>str</var>, int <var>cr</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of an XML string. `cr' specifies whether to remove comments. The return value is a list handle. Each element of the list is the string of a tag or a text section. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use. The character encoding of the input string should be US-ASCII, UTF-8, ISO-8859-*, EUC-*, or Shift_JIS. Because these functions for XML are not XML parser with validation check, it can handle also HTML and SGML.</dd>
+</dl>
+
+<p>The function `cbxmlattrs' is used in order to get the map of attributes of an XML tag.</p>
+
+<dl>
+<dt><kbd>CBMAP *cbxmlattrs(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of a tag string. The return value is a map handle. Each key of the map is the name of an attribute. Each value is unescaped. You can get the name of the tag with the key of an empty string. Because the handle of the return value is opened with the function `cbmapopen', it should be closed with the function `cbmapclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbxmlescape' is used in order to escape a string with the meta characters of XML.</p>
+
+<dl>
+<dt><kbd>char *cbxmlescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of a string. The return value is the pointer to the escaped string sanitized of meta characters. This function converts only `&', `<', `>', and `"'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbxmlunescape' is used in order to unescape a string with the entity references of XML.</p>
+
+<dl>
+<dt><kbd>char *cbxmlunescape(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies the pointer to the region of a string. The return value is the pointer to the unescaped string. This function restores only `&', `<', `>', and `"'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbdeflate' is used in order to compress a serial object with ZLIB.</p>
+
+<dl>
+<dt><kbd>char *cbdeflate(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with ZLIB enabled.</dd>
+</dl>
+
+<p>The function `cbinflate' is used in order to decompress a serial object compressed with ZLIB.</p>
+
+<dl>
+<dt><kbd>char *cbinflate(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with ZLIB enabled.</dd>
+</dl>
+
+<p>The function `cbgzencode' is used in order to compress a serial object with GZIP.</p>
+
+<dl>
+<dt><kbd>char *cbgzencode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with ZLIB enabled.</dd>
+</dl>
+
+<p>The function `cbgzdecode' is used in order to decompress a serial object compressed with GZIP.</p>
+
+<dl>
+<dt><kbd>char *cbgzdecode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with ZLIB enabled.</dd>
+</dl>
+
+<p>The function `cbgetcrc' is used in order to get the CRC32 checksum of a serial object.</p>
+
+<dl>
+<dt><kbd>unsigned int cbgetcrc(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the CRC32 checksum of the object. This function is available only if QDBM was built with ZLIB enabled.</dd>
+</dl>
+
+<p>The function `cblzoencode' is used in order to compress a serial object with LZO.</p>
+
+<dl>
+<dt><kbd>char *cblzoencode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with LZO enabled.</dd>
+</dl>
+
+<p>The function `cblzodecode' is used in order to decompress a serial object compressed with LZO.</p>
+
+<dl>
+<dt><kbd>char *cblzodecode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with LZO enabled.</dd>
+</dl>
+
+<p>The function `cbbzencode' is used in order to compress a serial object with BZIP2.</p>
+
+<dl>
+<dt><kbd>char *cbbzencode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with BZIP2 enabled.</dd>
+</dl>
+
+<p>The function `cbbzdecode' is used in order to decompress a serial object compressed with BZIP2.</p>
+
+<dl>
+<dt><kbd>char *cbbzdecode(const char *<var>ptr</var>, int <var>size</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QDBM was built with BZIP2 enabled.</dd>
+</dl>
+
+<p>The function `cbiconv' is used in order to convert the character encoding of a string.</p>
+
+<dl>
+<dt><kbd>char *cbiconv(const char *<var>ptr</var>, int <var>size</var>, const char *<var>icode</var>, const char *<var>ocode</var>, int *<var>sp</var>, int *<var>mp</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. `icode' specifies the name of encoding of the input string. `ocode' specifies the name of encoding of the output string. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. `mp' specifies the pointer to a variable to which the number of missing characters by failure of conversion is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the result object, else, it is `NULL'. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. This function is available only if QD
BM was built with ICONV enabled.</dd>
+</dl>
+
+<p>The function `cbencname' is used in order to detect the encoding of a string automatically.</p>
+
+<dl>
+<dt><kbd>const char *cbencname(const char *<var>ptr</var>, int <var>size</var>);</kbd></dt>
+<dd>`ptr' specifies the pointer to a region. `size' specifies the size of the region. If it is negative, the size is assigned with `strlen(ptr)'. The return value is the string of the encoding name of the string. As it stands, US-ASCII, ISO-2022-JP, Shift_JIS, CP932, EUC-JP, UTF-8, UTF-16, UTF-16BE, and UTF-16LE are supported. If none of them matches, ISO-8859-1 is selected. This function is available only if QDBM was built with ICONV enabled.</dd>
+</dl>
+
+<p>The function `cbjetlag' is used in order to get the jet lag of the local time in seconds.</p>
+
+<dl>
+<dt><kbd>int cbjetlag(void);</kbd></dt>
+<dd>The return value is the jet lag of the local time in seconds.</dd>
+</dl>
+
+<p>The function `cbcalendar' is used in order to get the Gregorian calendar of a time.</p>
+
+<dl>
+<dt><kbd>void cbcalendar(time_t <var>t</var>, int <var>jl</var>, int *<var>yearp</var>, int *<var>monp</var>, int *<var>dayp</var>, int *<var>hourp</var>, int *<var>minp</var>, int *<var>secp</var>);</kbd></dt>
+<dd>`t' specifies a source time. If it is negative, the current time is specified. `jl' specifies the jet lag of a location in seconds. `yearp' specifies the pointer to a variable to which the year is assigned. If it is `NULL', it is not used. `monp' specifies the pointer to a variable to which the month is assigned. If it is `NULL', it is not used. 1 means January and 12 means December. `dayp' specifies the pointer to a variable to which the day of the month is assigned. If it is `NULL', it is not used. `hourp' specifies the pointer to a variable to which the hours is assigned. If it is `NULL', it is not used. `minp' specifies the pointer to a variable to which the minutes is assigned. If it is `NULL', it is not used. `secp' specifies the pointer to a variable to which the seconds is assigned. If it is `NULL', it is not used.</dd>
+</dl>
+
+<p>The function `cbdayofweek' is used in order to get the day of week of a date.</p>
+
+<dl>
+<dt><kbd>int cbdayofweek(int <var>year</var>, int <var>mon</var>, int <var>day</var>);</kbd></dt>
+<dd>`year' specifies the year of a date. `mon' specifies the month of the date. `day' specifies the day of the date. The return value is the day of week of the date. 0 means Sunday and 6 means Saturday.</dd>
+</dl>
+
+<p>The function `cbdatestrwww' is used in order to get the string for a date in W3CDTF.</p>
+
+<dl>
+<dt><kbd>char *cbdatestrwww(time_t <var>t</var>, int <var>jl</var>);</kbd></dt>
+<dd>`t' specifies a source time. If it is negative, the current time is specified. `jl' specifies the jet lag of a location in seconds. The return value is the string of the date in W3CDTF (YYYY-MM-DDThh:mm:ddTZD). Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbdatestrhttp' is used in order to get the string for a date in RFC 1123 format.</p>
+
+<dl>
+<dt><kbd>char *cbdatestrhttp(time_t <var>t</var>, int <var>jl</var>);</kbd></dt>
+<dd>`t' specifies a source time. If it is negative, the current time is specified. `jl' specifies the jet lag of a location in seconds. The return value is the string of the date in RFC 1123 format (Wdy, DD-Mon-YYYY hh:mm:dd TZD). Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `cbstrmktime' is used in order to get the time value of a date string in decimal, hexadecimal, W3CDTF, or RFC 822 (1123).</p>
+
+<dl>
+<dt><kbd>time_t cbstrmktime(const char *<var>str</var>);</kbd></dt>
+<dd>`str' specifies a date string in decimal, hexadecimal, W3CDTF, or RFC 822 (1123). The return value is the time value of the date or -1 if the format is invalid. Decimal can be trailed by "s" for in seconds, "m" for in minutes, "h" for in hours, and "d" for in days.</dd>
+</dl>
+
+<p>The function `cbproctime' is used in order to get user and system processing times.</p>
+
+<dl>
+<dt><kbd>void cbproctime(double *<var>usrp</var>, double *<var>sysp</var>);</kbd></dt>
+<dd>`usrp' specifies the pointer to a variable to which the user processing time is assigned. If it is `NULL', it is not used. The unit of time is seconds. `sysp' specifies the pointer to a variable to which the system processing time is assigned. If it is `NULL', it is not used. The unit of time is seconds.</dd>
+</dl>
+
+<p>The function `cbstdiobin' is used in order to ensure that the standard I/O is binary mode.</p>
+
+<dl>
+<dt><kbd>void cbstdiobin(void);</kbd></dt>
+<dd>This function is useful for applications on dosish file systems.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example is typical use.</p>
+
+<pre>#include <cabin.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+int main(int argc, char **argv){
+ CBDATUM *datum;
+ CBLIST *list;
+ CBMAP *map;
+ char *buf1, *buf2;
+ int i;
+
+ /* open the datum handle */
+ datum = cbdatumopen("123", -1);
+ /* concatenate the data */
+ cbdatumcat(datum, "abc", -1);
+ /* print the datum */
+ printf("%s\n", cbdatumptr(datum));
+ /* close the datum handle */
+ cbdatumclose(datum);
+
+ /* open the list handle */
+ list = cblistopen();
+ /* add elements into the list */
+ cblistpush(list, "apple", -1);
+ cblistpush(list, "orange", -1);
+ /* print all elements */
+ for(i = 0; i < cblistnum(list); i++){
+ printf("%s\n", cblistval(list, i, NULL));
+ }
+ /* close the list handle */
+ cblistclose(list);
+
+ /* open the map handle */
+ map = cbmapopen();
+ /* add records into the map */
+ cbmapput(map, "dog", -1, "bowwow", -1, 1);
+ cbmapput(map, "cat", -1, "meow", -1, 1);
+ /* search for values and print them */
+ printf("%s\n", cbmapget(map, "dog", -1, NULL));
+ printf("%s\n", cbmapget(map, "cat", -1, NULL));
+ /* close the map */
+ cbmapclose(map);
+
+ /* Base64 encoding */
+ buf1 = cbbaseencode("I miss you.", -1);
+ printf("%s\n", buf1);
+ /* Base64 decoding */
+ buf2 = cbbasedecode(buf1, NULL);
+ printf("%s\n", buf2);
+ /* release the resources */
+ free(buf2);
+ free(buf1);
+
+ /* register a plain pointer to the global garbage collector */
+ buf1 = cbmemdup("Take it easy.", -1);
+ cbglobalgc(buf1, free);
+ /* the pointer is available but you don't have to release it */
+ printf("%s\n", buf1);
+
+ /* register a list to the global garbage collector */
+ list = cblistopen();
+ cbglobalgc(list, (void (*)(void *))cblistclose);
+ /* the handle is available but you don't have to close it */
+ cblistpush(list, "Don't hesitate.", -1);
+ for(i = 0; i < cblistnum(list); i++){
+ printf("%s\n", cblistval(list, i, NULL));
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>How to build programs using Cabin is the same as the case of Depot.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>Functions of Cabin except for `cbglobalgc' are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<hr />
+
+<h2><a name="cabincli" id="cabincli" class="head">Commands for Cabin</a></h2>
+
+<p>Cabin has the following command line interfaces.</p>
+
+<p>The command `cbtest' is a utility for facility test and performance test. Measure the execution time of the command. This command is used in the following format. `rnum' specifies the number of records.</p>
+
+<dl>
+<dt><kbd>cbtest sort [-d] <var>rnum</var></kbd></dt>
+<dd>Perform test of sorting algorithms.</dd>
+<dt><kbd>cbtest strstr [-d] <var>rnum</var></kbd></dt>
+<dd>Perform test of string locating algorithms.</dd>
+<dt><kbd>cbtest list [-d] <var>rnum</var></kbd></dt>
+<dd>Perform writing test of list.</dd>
+<dt><kbd>cbtest map [-d] <var>rnum</var></kbd></dt>
+<dd>Perform writing test of map.</dd>
+<dt><kbd>cbtest wicked <var>rnum</var></kbd></dt>
+<dd>Perform updating operations of list and map selected at random.</dd>
+<dt><kbd>cbtest misc</kbd></dt>
+<dd>Perform test of miscellaneous routines.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-d</kbd> : read and show data of the result.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>The command `cbcodec' is a tool to use encoding and decoding features provided by Cabin. This command is used in the following format. `file' specifies a input file. If it is omitted, the standard input is read.</p>
+
+<dl>
+<dt><kbd>cbcodec url [-d] [-br] [-rs <var>base</var> <var>target</var>] [-l] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>Perform URL encoding and its decoding.</dd>
+<dt><kbd>cbcodec base [-d] [-l] [-c <var>num</var>] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>Perform Base64 encoding and its decoding.</dd>
+<dt><kbd>cbcodec quote [-d] [-l] [-c <var>num</var>] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>Perform quoted-printable encoding and its decoding.</dd>
+<dt><kbd>cbcodec mime [-d] [-hd] [-bd] [-part <var>num</var>] [-l] [-ec <var>code</var>] [-qp] [-dc] [-e <var>expr</var>] [<var>file</var>]</kbd></dt>
+<dd>Perform MIME encoding and its decoding.</dd>
+<dt><kbd>cbcodec csv [-d] [-t] [-l] [-e <var>expr</var>] [-html] [<var>file</var>]</kbd></dt>
+<dd>Process CSV. By default, escape meta characters.</dd>
+<dt><kbd>cbcodec xml [-d] [-p] [-l] [-e <var>expr</var>] [-tsv] [<var>file</var>]</kbd></dt>
+<dd>Process XML. By default, escape meta characters.</dd>
+<dt><kbd>cbcodec zlib [-d] [-gz] [-crc] [<var>file</var>]</kbd></dt>
+<dd>Perform deflation and inflation with ZLIB. It is available only if QDBM was built with ZLIB enabled.</dd>
+<dt><kbd>cbcodec lzo [-d] [<var>file</var>]</kbd></dt>
+<dd>Perform compression and decompression with LZO. It is available only if QDBM was built with LZO enabled.</dd>
+<dt><kbd>cbcodec bzip [-d] [<var>file</var>]</kbd></dt>
+<dd>Perform compression and decompression with BZIP2. It is available only if QDBM was built with BZIP2 enabled.</dd>
+<dt><kbd>cbcodec iconv [-ic <var>code</var>] [-oc <var>code</var>] [-ol <var>ltype</var>] [-cn] [-um] [-wc] [<var>file</var>]</kbd></dt>
+<dd>Convert character encoding with ICONV. It is available only if QDBM was built with ICONV enabled.</dd>
+<dt><kbd>cbcodec date [-wf] [-rf] [-utc] [<var>str</var>]</kbd></dt>
+<dd>Convert a date string specified `str'. By default, UNIX time is output. If `str' is omitted, the current date is dealt.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-d</kbd> : perform decoding (unescaping), not encoding (escaping).</li>
+<li><kbd>-br</kbd> : break up URL into elements.</li>
+<li><kbd>-rs</kbd> : resolve relative URL.</li>
+<li><kbd>-l</kbd> : output the tailing newline.</li>
+<li><kbd>-e <var>expr</var></kbd> : specify input data directly.</li>
+<li><kbd>-c <var>num</var></kbd> : limit the number of columns of the encoded data.</li>
+<li><kbd>-hd</kbd> : parse MIME and extract headers in TSV format.</li>
+<li><kbd>-bd</kbd> : parse MIME and extract the body.</li>
+<li><kbd>-part <var>num</var></kbd> : parse MIME and extract a part.</li>
+<li><kbd>-ec <var>code</var></kbd> : specify the input encoding, which is UTF-8 by default.</li>
+<li><kbd>-qp</kbd> : use quoted-printable encoding, which is Base64 by default.</li>
+<li><kbd>-dc</kbd> : output the encoding name instead of the result string when decoding.</li>
+<li><kbd>-t</kbd> : parse CSV. Convert the data into TSV. Tab and new-line in a cell are deleted.</li>
+<li><kbd>-html</kbd> : parse CSV. Convert the data into HTML.</li>
+<li><kbd>-p</kbd> : parse XML. Show tags and text sections with dividing headers.</li>
+<li><kbd>-tsv</kbd> : parse XML. Show the result in TSV format. Characters of tabs and new-lines are URL-encoded.</li>
+<li><kbd>-gz</kbd> : use GZIP format.</li>
+<li><kbd>-crc</kbd> : output the CRC32 checksum as hexadecimal and big endian.</li>
+<li><kbd>-ic <var>code</var></kbd> : specify the input encoding, which is detected automatically by default.</li>
+<li><kbd>-oc <var>code</var></kbd> : specify the output encoding, which is UTF-8 by default.</li>
+<li><kbd>-ol <var>ltype</var></kbd> : convert line feed characters, with `unix'(LF), `dos'(CRLF), and `mac'(CR).</li>
+<li><kbd>-cn</kbd> : detect the input encoding and show its name.</li>
+<li><kbd>-wc</kbd> : count the number of characters of the input string of UTF-8.</li>
+<li><kbd>-um</kbd> : output mappings of UCS-2 characters and C strings of UTF-16BE and UTF-8.</li>
+<li><kbd>-wf</kbd> : output in W3CDTF format.</li>
+<li><kbd>-rf</kbd> : output in RFC 1123 format.</li>
+<li><kbd>-utc</kbd> : output the coordinate universal time.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<hr />
+
+<h2><a name="villaapi" id="villaapi" class="head">Villa: Advanced API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Villa is the advanced API of QDBM. It provides routines for managing a database file of B+ tree. Each record is stored being sorted in order defined by a user. As for hash databases, retrieving method is provided only as complete accord. However, with Villa, it is possible to retrieve records specified by range. Cursor is used in order to access each record in order. It is possible to store records duplicating keys in a database. Moreover, according to the transaction mechanism, you can commit or abort operations of a database in a lump.</p>
+
+<p>Villa is implemented, based on Depot and Cabin. A database file of Villa is actual one of Depot. Although processing speed of retrieving and storing is slower than Depot, the size of a database is smaller.</p>
+
+<p>In order to use Villa, you should include `depot.h', `cabin.h', `villa.h' and `stdlib.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <cabin.h></kbd></dt>
+<dt><kbd>#include <villa.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>A pointer to `VILLA' is used as a database handle. It is like that some file I/O routines of `stdio.h' use a pointer to `FILE'. A database handle is opened with the function `vlopen' and closed with `vlclose'. You should not refer directly to any member of the handle. If a fatal error occurs in a database, any access method via the handle except `vlclose' will not work and return error status. Although a process is allowed to use multiple database handles at the same time, handles of the same database file should not be used. Before the cursor is used, it should be initialized by one of `vlcurfirst', `vlcurlast' or `vlcurjump'. Also after storing or deleting a record with functions except for `vlcurput' and `vlcurout', the cursor should be initialized.</p>
+
+<p>Villa also assign the external variable `dpecode' with the error code. The function `dperrmsg' is used in order to get the message of the error code.</p>
+
+<h3>API</h3>
+
+<p>You can define a comparing function to specify the order of records. The function should be the following type.</p>
+
+<dl>
+<dt><kbd>typedef int(*VLCFUNC)(const char *<var>aptr</var>, int <var>asiz</var>, const char *<var>bptr</var>, int <var>bsiz</var>);</kbd></dt>
+<dd>`aptr' specifies the pointer to the region of one key. `asiz' specifies the size of the region of one key. `bptr' specifies the pointer to the region of the other key. `bsiz' specifies the size of the region of the other key. The return value is positive if the former is big, negative if the latter is big, 0 if both are equivalent.</dd>
+</dl>
+
+<p>The function `vlopen' is used in order to get a database handle.</p>
+
+<dl>
+<dt><kbd>VILLA *vlopen(const char *<var>name</var>, int <var>omode</var>, VLCFUNC <var>cmp</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. `omode' specifies the connection mode: `VL_OWRITER' as a writer, `VL_OREADER' as a reader. If the mode is `VL_OWRITER', the following may be added by bitwise or: `VL_OCREAT', which means it creates a new database if not exist, `VL_OTRUNC', which means it creates a new database regardless if one exists, `VL_OZCOMP', which means leaves in the database are compressed with ZLIB, `VL_OYCOMP', which means leaves in the database are compressed with LZO, `VL_OXCOMP', which means leaves in the database are compressed with BZIP2. Both of `VL_OREADER' and `VL_OWRITER' can be added to by bitwise or: `VL_ONOLCK', which means it opens a database file without file locking, or `VL_OLCKNB', which means locking is performed without blocking. `cmp' specifies the comparing function: `VL_CMPLEX' comparing keys in lexical order, `VL_CMPINT' comparing keys as objects of `int' in native byte order, `VL_CMPNUM' comparing keys as numbers of big en
dian, `VL_CMPDEC' comparing keys as decimal strings. Any function based on the declaration of the type `VLCFUNC' can be assigned to the comparing function. The comparing function should be kept same in the life of a database. The return value is the database handle or `NULL' if it is not successful. While connecting as a writer, an exclusive lock is invoked to the database file. While connecting as a reader, a shared lock is invoked to the database file. The thread blocks until the lock is achieved. `VL_OZCOMP', `VL_OYCOMP', and `VL_OXCOMP' are available only if QDBM was built each with ZLIB, LZO, and BZIP2 enabled. If `VL_ONOLCK' is used, the application is responsible for exclusion control.</dd>
+</dl>
+
+<p>The function `vlclose' is used in order to close a database handle.</p>
+
+<dl>
+<dt><kbd>int vlclose(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is true, else, it is false. Because the region of a closed handle is released, it becomes impossible to use the handle. Updating a database is assured to be written when the handle is closed. If a writer opens a database but does not close it appropriately, the database will be broken. If the transaction is activated and not committed, it is aborted.</dd>
+</dl>
+
+<p>The function `vlput' is used in order to store a record.</p>
+
+<dl>
+<dt><kbd>int vlput(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>dmode</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. `dmode' specifies behavior when the key overlaps, by the following values: `VL_DOVER', which means the specified value overwrites the existing one, `VL_DKEEP', which means the existing value is kept, `VL_DCAT', which means the specified value is concatenated at the end of the existing value, `VL_DDUP', which means duplication of keys is allowed and the specified value is added as the last one, `VL_DDUPR', which means duplication of keys is allowed and the specified value is added as the first one. If successful, the return value is true, else, it is false. The cursor becomes
unavailable due to updating database.</dd>
+</dl>
+
+<p>The function `vlout' is used in order to delete a record.</p>
+
+<dl>
+<dt><kbd>int vlout(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is true, else, it is false. False is returned when no record corresponds to the specified key. When the key of duplicated records is specified, the first record of the same key is deleted. The cursor becomes unavailable due to updating database.</dd>
+</dl>
+
+<p>The function `vlget' is used in order to retrieve a record.</p>
+
+<dl>
+<dt><kbd>char *vlget(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key. When the key of duplicated records is specified, the value of the first record of the same key is selected. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `vlvsiz' is used in order to get the size of the value of a record.</p>
+
+<dl>
+<dt><kbd>int vlvsiz(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is the size of the value of the corresponding record, else, it is -1. If multiple records correspond, the size of the first is returned.</dd>
+</dl>
+
+<p>The function `vlvnum' is used in order to get the number of records corresponding a key.</p>
+
+<dl>
+<dt><kbd>int vlvnum(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. The return value is the number of corresponding records. If no record corresponds, 0 is returned.</dd>
+</dl>
+
+<p>The function `vlputlist' is used in order to store plural records corresponding a key.</p>
+
+<dl>
+<dt><kbd>int vlputlist(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, const CBLIST *<var>vals</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `vals' specifies a list handle of values. The list should not be empty. If successful, the return value is true, else, it is false. The cursor becomes unavailable due to updating database.</dd>
+</dl>
+
+<p>The function `vloutlist' is used in order to delete all records corresponding a key.</p>
+
+<dl>
+<dt><kbd>int vloutlist(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is true, else, it is false. False is returned when no record corresponds to the specified key. The cursor becomes unavailable due to updating database.</dd>
+</dl>
+
+<p>The function `vlgetlist' is used in order to retrieve values of all records corresponding a key.</p>
+
+<dl>
+<dt><kbd>CBLIST *vlgetlist(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. If successful, the return value is a list handle of the values of the corresponding records, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `vlgetcat' is used in order to retrieve concatenated values of all records corresponding a key.</p>
+
+<dl>
+<dt><kbd>char *vlgetcat(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the concatenated values of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `vlcurfirst' is used in order to move the cursor to the first record.</p>
+
+<dl>
+<dt><kbd>int vlcurfirst(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is true, else, it is false. False is returned if there is no record in the database.</dd>
+</dl>
+
+<p>The function `vlcurlast' is used in order to move the cursor to the last record.</p>
+
+<dl>
+<dt><kbd>int vlcurlast(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is true, else, it is false. False is returned if there is no record in the database.</dd>
+</dl>
+
+<p>The function `vlcurprev' is used in order to move the cursor to the previous record.</p>
+
+<dl>
+<dt><kbd>int vlcurprev(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is true, else, it is false. False is returned if there is no previous record.</dd>
+</dl>
+
+<p>The function `vlcurnext' is used in order to move the cursor to the next record.</p>
+
+<dl>
+<dt><kbd>int vlcurnext(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is true, else, it is false. False is returned if there is no next record.</dd>
+</dl>
+
+<p>The function `vlcurjump' is used in order to move the cursor to a position around a record.</p>
+
+<dl>
+<dt><kbd>int vlcurjump(VILLA *<var>villa</var>, const char *<var>kbuf</var>, int <var>ksiz</var>, int <var>jmode</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `kbuf' specifies the pointer to the region of a key. `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned with `strlen(kbuf)'. `jmode' specifies detail adjustment: `VL_JFORWARD', which means that the cursor is set to the first record of the same key and that the cursor is set to the next substitute if completely matching record does not exist, `VL_JBACKWARD', which means that the cursor is set to the last record of the same key and that the cursor is set to the previous substitute if completely matching record does not exist. If successful, the return value is true, else, it is false. False is returned if there is no record corresponding the condition.</dd>
+</dl>
+
+<p>The function `vlcurkey' is used in order to get the key of the record where the cursor is.</p>
+
+<dl>
+<dt><kbd>char *vlcurkey(VILLA *<var>villa</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the key of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the cursor. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `vlcurval' is used in order to get the value of the record where the cursor is.</p>
+
+<dl>
+<dt><kbd>char *vlcurval(VILLA *<var>villa</var>, int *<var>sp</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `sp' specifies the pointer to a variable to which the size of the region of the return value is assigned. If it is `NULL', it is not used. If successful, the return value is the pointer to the region of the value of the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to the cursor. Because an additional zero code is appended at the end of the region of the return value, the return value can be treated as a character string. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `vlcurput' is used in order to insert a record around the cursor.</p>
+
+<dl>
+<dt><kbd>int vlcurput(VILLA *<var>villa</var>, const char *<var>vbuf</var>, int <var>vsiz</var>, int <var>cpmode</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. `vbuf' specifies the pointer to the region of a value. `vsiz' specifies the size of the region of the value. If it is negative, the size is assigned with `strlen(vbuf)'. `cpmode' specifies detail adjustment: `VL_CPCURRENT', which means that the value of the current record is overwritten, `VL_CPBEFORE', which means that a new record is inserted before the current record, `VL_CPAFTER', which means that a new record is inserted after the current record. If successful, the return value is true, else, it is false. False is returned when no record corresponds to the cursor. After insertion, the cursor is moved to the inserted record.</dd>
+</dl>
+
+<p>The function `vlcurout' is used in order to delete the record where the cursor is.</p>
+
+<dl>
+<dt><kbd>int vlcurout(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. False is returned when no record corresponds to the cursor. After deletion, the cursor is moved to the next record if possible.</dd>
+</dl>
+
+<p>The function `vlsettuning' is used in order to set the tuning parameters for performance.</p>
+
+<dl>
+<dt><kbd>void vlsettuning(VILLA *<var>villa</var>, int <var>lrecmax</var>, int <var>nidxmax</var>, int <var>lcnum</var>, int <var>ncnum</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `lrecmax' specifies the max number of records in a leaf node of B+ tree. If it is not more than 0, the default value is specified. `nidxmax' specifies the max number of indexes in a non-leaf node of B+ tree. If it is not more than 0, the default value is specified. `lcnum' specifies the max number of caching leaf nodes. If it is not more than 0, the default value is specified. `ncnum' specifies the max number of caching non-leaf nodes. If it is not more than 0, the default value is specified. The default setting is equivalent to `vlsettuning(49, 192, 1024, 512)'. Because tuning parameters are not saved in a database, you should specify them every opening a database.</dd>
+</dl>
+
+<p>The function `vlsetfbpsiz' is used in order to set the size of the free block pool of a database handle.</p>
+
+<dl>
+<dt><kbd>int vlsetfbpsiz(VILLA *<var>villa</var>, int <var>size</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. `size' specifies the size of the free block pool of a database. If successful, the return value is true, else, it is false. The default size of the free block pool is 256. If the size is greater, the space efficiency of overwriting values is improved with the time efficiency sacrificed.</dd>
+</dl>
+
+<p>The function `vlsync' is used in order to synchronize updating contents with the file and the device.</p>
+
+<dl>
+<dt><kbd>int vlsync(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. This function is useful when another process uses the connected database file. This function should not be used while the transaction is activated.</dd>
+</dl>
+
+<p>The function `vloptimize' is used in order to optimize a database.</p>
+
+<dl>
+<dt><kbd>int vloptimize(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. In an alternating succession of deleting and storing with overwrite or concatenate, dispensable regions accumulate. This function is useful to do away with them. This function should not be used while the transaction is activated.</dd>
+</dl>
+
+<p>The function `vlname' is used in order to get the name of a database.</p>
+
+<dl>
+<dt><kbd>char *vlname(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is the pointer to the region of the name of the database, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `vlfsiz' is used in order to get the size of a database file.</p>
+
+<dl>
+<dt><kbd>int vlfsiz(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is the size of the database file, else, it is -1. Because of the I/O buffer, the return value may be less than the hard size.</dd>
+</dl>
+
+<p>The function `vllnum' is used in order to get the number of the leaf nodes of B+ tree.</p>
+
+<dl>
+<dt><kbd>int vllnum(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is the number of the leaf nodes, else, it is -1.</dd>
+</dl>
+
+<p>The function `vlnnum' is used in order to get the number of the non-leaf nodes of B+ tree.</p>
+
+<dl>
+<dt><kbd>int vlnnum(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is the number of the non-leaf nodes, else, it is -1.</dd>
+</dl>
+
+<p>The function `vlrnum' is used in order to get the number of the records stored in a database.</p>
+
+<dl>
+<dt><kbd>int vlrnum(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. If successful, the return value is the number of the records stored in the database, else, it is -1.</dd>
+</dl>
+
+<p>The function `vlwritable' is used in order to check whether a database handle is a writer or not.</p>
+
+<dl>
+<dt><kbd>int vlwritable(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. The return value is true if the handle is a writer, false if not.</dd>
+</dl>
+
+<p>The function `vlfatalerror' is used in order to check whether a database has a fatal error or not.</p>
+
+<dl>
+<dt><kbd>int vlfatalerror(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. The return value is true if the database has a fatal error, false if not.</dd>
+</dl>
+
+<p>The function `vlinode' is used in order to get the inode number of a database file.</p>
+
+<dl>
+<dt><kbd>int vlinode(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. The return value is the inode number of the database file.</dd>
+</dl>
+
+<p>The function `vlmtime' is used in order to get the last modified time of a database.</p>
+
+<dl>
+<dt><kbd>time_t vlmtime(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. The return value is the last modified time of the database.</dd>
+</dl>
+
+<p>The function `vltranbegin' is used in order to begin the transaction.</p>
+
+<dl>
+<dt><kbd>int vltranbegin(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. Because this function does not perform mutual exclusion control in multi-thread, the application is responsible for it. Only one transaction can be activated with a database handle at the same time.</dd>
+</dl>
+
+<p>The function `vltrancommit' is used in order to commit the transaction.</p>
+
+<dl>
+<dt><kbd>int vltrancommit(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. Updating a database in the transaction is fixed when it is committed successfully.</dd>
+</dl>
+
+<p>The function `vltranabort' is used in order to abort the transaction.</p>
+
+<dl>
+<dt><kbd>int vltranabort(VILLA *<var>villa</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. Updating a database in the transaction is discarded when it is aborted. The state of the database is rollbacked to before transaction.</dd>
+</dl>
+
+<p>The function `vlremove' is used in order to remove a database file.</p>
+
+<dl>
+<dt><kbd>int vlremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `vlrepair' is used in order to repair a broken database file.</p>
+
+<dl>
+<dt><kbd>int vlrepair(const char *<var>name</var>, VLCFUNC <var>cmp</var>);</kbd></dt>
+<dd>`name' specifies the name of a database file. `cmp' specifies the comparing function of the database file. If successful, the return value is true, else, it is false. There is no guarantee that all records in a repaired database file correspond to the original or expected state.</dd>
+</dl>
+
+<p>The function `vlexportdb' is used in order to dump all records as endian independent data.</p>
+
+<dl>
+<dt><kbd>int vlexportdb(VILLA *<var>villa</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`villa' specifies a database handle. `name' specifies the name of an output file. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `vlimportdb' is used in order to load all records from endian independent data.</p>
+
+<dl>
+<dt><kbd>int vlimportdb(VILLA *<var>villa</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`villa' specifies a database handle connected as a writer. The database of the handle must be empty. `name' specifies the name of an input file. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example stores and retrieves a phone number, using the name as the key.</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#define NAME "mikio"
+#define NUMBER "000-1234-5678"
+#define DBNAME "book"
+
+int main(int argc, char **argv){
+ VILLA *villa;
+ char *val;
+
+ /* open the database */
+ if(!(villa = vlopen(DBNAME, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ fprintf(stderr, "vlopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* store the record */
+ if(!vlput(villa, NAME, -1, NUMBER, -1, VL_DOVER)){
+ fprintf(stderr, "vlput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* retrieve the record */
+ if(!(val = vlget(villa, NAME, -1, NULL))){
+ fprintf(stderr, "vlget: %s\n", dperrmsg(dpecode));
+ } else {
+ printf("Name: %s\n", NAME);
+ printf("Number: %s\n", val);
+ free(val);
+ }
+
+ /* close the database */
+ if(!vlclose(villa)){
+ fprintf(stderr, "vlclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>The following example performs forward matching search for strings.</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#define DBNAME "words"
+#define PREFIX "apple"
+
+int main(int argc, char **argv){
+ VILLA *villa;
+ char *key, *val;
+
+ /* open the database */
+ if(!(villa = vlopen(DBNAME, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ fprintf(stderr, "vlopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* store records */
+ if(!vlput(villa, "applet", -1, "little application", -1, VL_DDUP) ||
+ !vlput(villa, "aurora", -1, "polar wonderwork", -1, VL_DDUP) ||
+ !vlput(villa, "apple", -1, "delicious fruit", -1, VL_DDUP) ||
+ !vlput(villa, "amigo", -1, "good friend", -1, VL_DDUP) ||
+ !vlput(villa, "apple", -1, "big city", -1, VL_DDUP)){
+ fprintf(stderr, "vlput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* set the cursor at the top of candidates */
+ vlcurjump(villa, PREFIX, -1, VL_JFORWARD);
+
+ /* scan with the cursor */
+ while((key = vlcurkey(villa, NULL)) != NULL){
+ if(strstr(key, PREFIX) != key){
+ free(key);
+ break;
+ }
+ if(!(val = vlcurval(villa, NULL))){
+ fprintf(stderr, "vlcurval: %s\n", dperrmsg(dpecode));
+ free(key);
+ break;
+ }
+ printf("%s: %s\n", key, val);
+ free(val);
+ free(key);
+ vlcurnext(villa);
+ }
+
+ /* close the database */
+ if(!vlclose(villa)){
+ fprintf(stderr, "vlclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>How to build programs using Villa is the same as the case of Depot.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>If QDBM was built with POSIX thread enabled, the global variable `dpecode' is treated as thread specific data, and functions of Villa are reentrant. In that case, they are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<h3>Vista: Extended Advanced API</h3>
+
+<p>Vista is the extended API of Villa. To compensate for the defect that Villa can not handle a file whose size is more than 2GB, Vista does not use Depot but Curia for handling its internal database. While Vista provides data structure and operations of B+ tree as with Villa, its database is realized as a directory.</p>
+
+<p>In order to use Vista, you should include `vista.h' instead of `villa.h'. Because Vista is implemented by overriding symbols of Villa, it can be used as with Villa. That is, Signatures of Villa and Vista is all the same. However, as its adverse effect, modules (compilation unit) using Vista can not use Villa (do not include `villa.h').</p>
+
+<hr />
+
+<h2><a name="villacli" id="villacli" class="head">Commands for Villa</a></h2>
+
+<p>Villa has the following command line interfaces.</p>
+
+<p>The command `vlmgr' is a utility for debugging Villa and its applications. It features editing and checking of a database. It can be used for database applications with shell scripts. This command is used in the following format. `name' specifies a database name. `key' specifies the key of a record. `val' specifies the value of a record.</p>
+
+<dl>
+<dt><kbd>vlmgr create [-cz|-cy|-cx] <var>name</var></kbd></dt>
+<dd>Create a database file.</dd>
+<dt><kbd>vlmgr put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat|-dup] <var>name</var> <var>key</var> <var>val</var></kbd></dt>
+<dd>Store a record with a key and a value.</dd>
+<dt><kbd>vlmgr out [-l] [-kx|-ki] <var>name</var> <var>key</var></kbd></dt>
+<dd>Delete a record with a key.</dd>
+<dt><kbd>vlmgr get [-nl] [-l] [-kx|-ki] [-ox] [-n] <var>name</var> <var>key</var></kbd></dt>
+<dd>Retrieve a record with a key and output it to the standard output.</dd>
+<dt><kbd>vlmgr list [-nl] [-k|-v] [-kx|-ki] [-ox] [-top <var>key</var>] [-bot <var>key</var>] [-gt] [-lt] [-max <var>num</var>] [-desc] <var>name</var></kbd></dt>
+<dd>List all keys and values delimited with tab and line-feed to the standard output.</dd>
+<dt><kbd>vlmgr optimize <var>name</var></kbd></dt>
+<dd>Optimize a database.</dd>
+<dt><kbd>vlmgr inform [-nl] <var>name</var></kbd></dt>
+<dd>Output miscellaneous information to the standard output.</dd>
+<dt><kbd>vlmgr remove <var>name</var></kbd></dt>
+<dd>Remove a database file.</dd>
+<dt><kbd>vlmgr repair [-ki] <var>name</var></kbd></dt>
+<dd>Repair a broken database file.</dd>
+<dt><kbd>vlmgr exportdb [-ki] <var>name</var> <var>file</var></kbd></dt>
+<dd>Dump all records as endian independent data.</dd>
+<dt><kbd>vlmgr importdb [-ki] <var>name</var> <var>file</var></kbd></dt>
+<dd>Load all records from endian independent data.</dd>
+<dt><kbd>vlmgr version</kbd></dt>
+<dd>Output version information of QDBM to the standard output.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-cz</kbd> : compress leaves in the database with ZLIB.</li>
+<li><kbd>-cy</kbd> : compress leaves in the database with LZO.</li>
+<li><kbd>-cx</kbd> : compress leaves in the database with BZIP2.</li>
+<li><kbd>-l</kbd> : all records corresponding the key are dealt.</li>
+<li><kbd>-kx</kbd> : treat `key' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-ki</kbd> : treat `key' as an integer expression of decimal notation.</li>
+<li><kbd>-vx</kbd> : treat `val' as a binary expression of hexadecimal notation.</li>
+<li><kbd>-vi</kbd> : treat `val' as an integer expression of decimal notation.</li>
+<li><kbd>-vf</kbd> : read the value from a file specified with `val'.</li>
+<li><kbd>-keep</kbd> : specify the storing mode for `VL_DKEEP'.</li>
+<li><kbd>-cat</kbd> : specify the storing mode for `VL_DCAT'.</li>
+<li><kbd>-dup</kbd> : specify the storing mode for `VL_DDUP'.</li>
+<li><kbd>-nl</kbd> : open the database without file locking.</li>
+<li><kbd>-top <var>key</var></kbd> : specify the top key of listing.</li>
+<li><kbd>-bot <var>key</var></kbd> : specify the bottom key of listing.</li>
+<li><kbd>-gt</kbd> : do not include the top key of listing.</li>
+<li><kbd>-lt</kbd> : do not include the bottom key of listing.</li>
+<li><kbd>-max <var>num</var></kbd> : specify the max number of listing.</li>
+<li><kbd>-desc</kbd> : list in descending order.</li>
+<li><kbd>-ox</kbd> : treat the output as a binary expression of hexadecimal notation.</li>
+<li><kbd>-n</kbd> : do not output the tailing newline.</li>
+<li><kbd>-k</kbd> : output keys only.</li>
+<li><kbd>-v</kbd> : output values only.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `vltest' is a utility for facility test and performance test. Check a database generated by the command or measure the execution time of the command. This command is used in the following format. `name' specifies a database name. `rnum' specifies the number of the records.</p>
+
+<dl>
+<dt><kbd>vltest write [-int] [-cz|-cy|-cx] [-tune <var>lrecmax</var> <var>nidxmax</var> <var>lcnum</var> <var>ncnum</var>] [-fbp <var>num</var>] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Store records with keys of 8 bytes. They change as `00000001', `00000002'...</dd>
+<dt><kbd>vltest read [-int] [-vc] <var>name</var></kbd></dt>
+<dd>Retrieve all records of the database above.</dd>
+<dt><kbd>vltest rdup [-int] [-cz|-cy|-cx] [-cc] [-tune <var>lrecmax</var> <var>nidxmax</var> <var>lcnum</var> <var>ncnum</var>] [-fbp <var>num</var>] <var>name</var> <var>rnum</var> <var>pnum</var></kbd></dt>
+<dd>Store records with partway duplicated keys using duplicate mode.</dd>
+<dt><kbd>vltest combo [-cz|-cy|-cx] <var>name</var></kbd></dt>
+<dd>Perform combination test of various operations.</dd>
+<dt><kbd>vltest wicked [-cz|-cy|-cx] <var>name</var> <var>rnum</var></kbd></dt>
+<dd>Perform updating operations selected at random.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-int</kbd> : treat keys and values as objects of `int', and use comparing function `VL_CMPINT'.</li>
+<li><kbd>-cz</kbd> : compress leaves in the database with ZLIB.</li>
+<li><kbd>-cy</kbd> : compress leaves in the database with LZO.</li>
+<li><kbd>-cx</kbd> : compress leaves in the database with BZIP2.</li>
+<li><kbd>-vc</kbd> : refer to volatile cache.</li>
+<li><kbd>-cc</kbd> : select `VL_DCAT' or `VL_DDUP' at random.</li>
+<li><kbd>-tune <var>lrecmax</var> <var>nidxmax</var> <var>lcnum</var> <var>ncnum</var></kbd> : set tuning parameters.</li>
+<li><kbd>-fbp <var>num</var></kbd> : set the size of the free block pool.</li>
+<li><kbd>-c</kbd> : perform comparison test with map of Cabin.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `vltsv' features mutual conversion between a database of Villa and a TSV text. This command is useful when data exchange with another version of QDBM or another DBM, or when data exchange between systems which have different byte orders. This command is used in the following format. `name' specifies a database name. The subcommand `export' reads TSV data from the standard input. The subcommand `import' writes TSV data to the standard output.</p>
+
+<dl>
+<dt><kbd>vltsv import [-bin] <var>name</var></kbd></dt>
+<dd>Create a database from TSV.</dd>
+<dt><kbd>vltsv export [-bin] <var>name</var></kbd></dt>
+<dd>Write TSV data of a database.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-bin</kbd> : treat records as Base64 format.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>Commands of Villa realize a simple database system. For example, to make a database to search `/etc/password' by a user name, perform the following command.</p>
+
+<pre>cat /etc/passwd | tr ':' '\t' | vltsv import casket
+</pre>
+
+<p>Thus, to retrieve the information of a user `mikio', perform the following command.</p>
+
+<pre>vlmgr get casket mikio
+</pre>
+
+<p>It is easy to implement functions upsides with these commands, using the API of Villa.</p>
+
+<p>The command `qmttest' checks multi-thread safety of Depot, Curia, and Villa. This command works with multi threads only if QDBM was built with POSIX thread. This command is used in the following format. `name' specifies the prefix of each database. `rnum' specifies the number of records to be stored in each database. `tnum' specifies the number of threads.</p>
+
+<dl>
+<dt><kbd>qmttest <var>name</var> <var>rnum</var> <var>tnum</var></kbd></dt>
+<dd>Check multi-thread safety.</dd>
+</dl>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<hr />
+
+<h2><a name="odeumapi" id="odeumapi" class="head">Odeum: Inverted API</a></h2>
+
+<h3>Overview</h3>
+
+<p>Odeum is the API which handles an inverted index. An inverted index is a data structure to retrieve a list of some documents that include one of words which were extracted from a population of documents. It is easy to realize a full-text search system with an inverted index. Odeum provides an abstract data structure which consists of words and attributes of a document. It is used when an application stores a document into a database and when an application retrieves some documents from a database.</p>
+
+<p>Odeum does not provide methods to extract the text from the original data of a document. It should be implemented by applications. Although Odeum provides utilities to extract words from a text, it is oriented to such languages whose words are separated with space characters as English. If an application handles such languages which need morphological analysis or N-gram analysis as Japanese, or if an application perform more such rarefied analysis of natural languages as stemming, its own analyzing method can be adopted. Result of search is expressed as an array contains elements which are structures composed of the ID number of documents and its score. In order to search with two or more words, Odeum provides utilities of set operations.</p>
+
+<p>Odeum is implemented, based on Curia, Cabin, and Villa. Odeum creates a database with a directory name. Some databases of Curia and Villa are placed in the specified directory. For example, `casket/docs', `casket/index', and `casket/rdocs' are created in the case that a database directory named as `casket'. `docs' is a database directory of Curia. The key of each record is the ID number of a document, and the value is such attributes as URI. `index' is a database directory of Curia. The key of each record is the normalized form of a word, and the value is an array whose element is a pair of the ID number of a document including the word and its score. `rdocs' is a database file of Villa. The key of each record is the URI of a document, and the value is its ID number.</p>
+
+<p>In order to use Odeum, you should include `depot.h', `cabin.h', `odeum.h' and `stdlib.h' in the source files. Usually, the following description will be near the beginning of a source file.</p>
+
+<dl>
+<dt><kbd>#include <depot.h></kbd></dt>
+<dt><kbd>#include <cabin.h></kbd></dt>
+<dt><kbd>#include <odeum.h></kbd></dt>
+<dt><kbd>#include <stdlib.h></kbd></dt>
+</dl>
+
+<p>A pointer to `ODEUM' is used as a database handle. A database handle is opened with the function `odopen' and closed with `odclose'. You should not refer directly to any member of the handle. If a fatal error occurs in a database, any access method via the handle except `odclose' will not work and return error status. Although a process is allowed to use multiple database handles at the same time, handles of the same database file should not be used.</p>
+
+<p>A pointer to `ODDOC' is used as a document handle. A document handle is opened with the function `oddocopen' and closed with `oddocclose'. You should not refer directly to any member of the handle. A document consists of attributes and words. Each word is expressed as a pair of a normalized form and a appearance form.</p>
+
+<p>Odeum also assign the external variable `dpecode' with the error code. The function `dperrmsg' is used in order to get the message of the error code.</p>
+
+<h3>API</h3>
+
+<p>Structures of `ODPAIR' type is used in order to handle results of search.</p>
+
+<dl>
+<dt><kbd>typedef struct { int id; int score; } ODPAIR;</kbd></dt>
+<dd>`id' specifies the ID number of a document. `score' specifies the score calculated from the number of searching words in the document.</dd>
+</dl>
+
+<p>The function `odopen' is used in order to get a database handle.</p>
+
+<dl>
+<dt><kbd>ODEUM *odopen(const char *<var>name</var>, int <var>omode</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory. `omode' specifies the connection mode: `OD_OWRITER' as a writer, `OD_OREADER' as a reader. If the mode is `OD_OWRITER', the following may be added by bitwise or: `OD_OCREAT', which means it creates a new database if not exist, `OD_OTRUNC', which means it creates a new database regardless if one exists. Both of `OD_OREADER' and `OD_OWRITER' can be added to by bitwise or: `OD_ONOLCK', which means it opens a database directory without file locking, or `OD_OLCKNB', which means locking is performed without blocking. The return value is the database handle or `NULL' if it is not successful. While connecting as a writer, an exclusive lock is invoked to the database directory. While connecting as a reader, a shared lock is invoked to the database directory. The thread blocks until the lock is achieved. If `OD_ONOLCK' is used, the application is responsible for exclusion control.</dd>
+</dl>
+
+<p>The function `odclose' is used in order to close a database handle.</p>
+
+<dl>
+<dt><kbd>int odclose(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is true, else, it is false. Because the region of a closed handle is released, it becomes impossible to use the handle. Updating a database is assured to be written when the handle is closed. If a writer opens a database but does not close it appropriately, the database will be broken.</dd>
+</dl>
+
+<p>The function `odput' is used in order to store a document.</p>
+
+<dl>
+<dt><kbd>int odput(ODEUM *<var>odeum</var>, const ODDOC *<var>doc</var>, int <var>wmax</var>, int <var>over</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle connected as a writer. `doc' specifies a document handle. `wmax' specifies the max number of words to be stored in the document database. If it is negative, the number is unlimited. `over' specifies whether the data of the duplicated document is overwritten or not. If it is false and the URI of the document is duplicated, the function returns as an error. If successful, the return value is true, else, it is false.</dd>
+</dl>
+
+<p>The function `odout' is used in order to delete a document specified by a URI.</p>
+
+<dl>
+<dt><kbd>int odout(ODEUM *<var>odeum</var>, const char *<var>uri</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle connected as a writer. `uri' specifies the string of the URI of a document. If successful, the return value is true, else, it is false. False is returned when no document corresponds to the specified URI.</dd>
+</dl>
+
+<p>The function `odoutbyid' is used in order to delete a document specified by an ID number.</p>
+
+<dl>
+<dt><kbd>int odoutbyid(ODEUM *<var>odeum</var>, int <var>id</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle connected as a writer. `id' specifies the ID number of a document. If successful, the return value is true, else, it is false. False is returned when no document corresponds to the specified ID number.</dd>
+</dl>
+
+<p>The function `odget' is used in order to retrieve a document specified by a URI.</p>
+
+<dl>
+<dt><kbd>ODDOC *odget(ODEUM *<var>odeum</var>, const char *<var>uri</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `uri' specifies the string of the URI of a document. If successful, the return value is the handle of the corresponding document, else, it is `NULL'. `NULL' is returned when no document corresponds to the specified URI. Because the handle of the return value is opened with the function `oddocopen', it should be closed with the function `oddocclose'.</dd>
+</dl>
+
+<p>The function `odgetbyid' is used in order to retrieve a document by an ID number.</p>
+
+<dl>
+<dt><kbd>ODDOC *odgetbyid(ODEUM *<var>odeum</var>, int <var>id</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `id' specifies the ID number of a document. If successful, the return value is the handle of the corresponding document, else, it is `NULL'. `NULL' is returned when no document corresponds to the specified ID number. Because the handle of the return value is opened with the function `oddocopen', it should be closed with the function `oddocclose'.</dd>
+</dl>
+
+<p>The function `odgetidbyuri' is used in order to retrieve the ID of the document specified by a URI.</p>
+
+<dl>
+<dt><kbd>int odgetidbyuri(ODEUM *<var>odeum</var>, const char *<var>uri</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `uri' specifies the string the URI of a document. If successful, the return value is the ID number of the document, else, it is -1. -1 is returned when no document corresponds to the specified URI.</dd>
+</dl>
+
+<p>The function `odcheck' is used in order to check whether the document specified by an ID number exists.</p>
+
+<dl>
+<dt><kbd>int odcheck(ODEUM *<var>odeum</var>, int <var>id</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `id' specifies the ID number of a document. The return value is true if the document exists, else, it is false.</dd>
+</dl>
+
+<p>The function `odsearch' is used in order to search the inverted index for documents including a particular word.</p>
+
+<dl>
+<dt><kbd>ODPAIR *odsearch(ODEUM *<var>odeum</var>, const char *<var>word</var>, int <var>max</var>, int *<var>np</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `word' specifies a searching word. `max' specifies the max number of documents to be retrieve. `np' specifies the pointer to a variable to which the number of the elements of the return value is assigned. If successful, the return value is the pointer to an array, else, it is `NULL'. Each element of the array is a pair of the ID number and the score of a document, and sorted in descending order of their scores. Even if no document corresponds to the specified word, it is not error but returns an dummy array. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. Note that each element of the array of the return value can be data of a deleted document.</dd>
+</dl>
+
+<p>The function `odsearchnum' is used in order to get the number of documents including a word.</p>
+
+<dl>
+<dt><kbd>int odsearchdnum(ODEUM *<var>odeum</var>, const char *<var>word</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `word' specifies a searching word. If successful, the return value is the number of documents including the word, else, it is -1. Because this function does not read the entity of the inverted index, it is faster than `odsearch'.</dd>
+</dl>
+
+<p>The function `oditerinit' is used in order to initialize the iterator of a database handle.</p>
+
+<dl>
+<dt><kbd>int oditerinit(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is true, else, it is false. The iterator is used in order to access every document stored in a database.</dd>
+</dl>
+
+<p>The function `oditernext' is used in order to get the next key of the iterator.</p>
+
+<dl>
+<dt><kbd>ODDOC *oditernext(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the handle of the next document, else, it is `NULL'. `NULL' is returned when no document is to be get out of the iterator. It is possible to access every document by iteration of calling this function. However, it is not assured if updating the database is occurred while the iteration. Besides, the order of this traversal access method is arbitrary, so it is not assured that the order of string matches the one of the traversal access. Because the handle of the return value is opened with the function `oddocopen', it should be closed with the function `oddocclose'.</dd>
+</dl>
+
+<p>The function `odsync' is used in order to synchronize updating contents with the files and the devices.</p>
+
+<dl>
+<dt><kbd>int odsync(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. This function is useful when another process uses the connected database directory.</dd>
+</dl>
+
+<p>The function `odoptimize' is used in order to optimize a database.</p>
+
+<dl>
+<dt><kbd>int odoptimize(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle connected as a writer. If successful, the return value is true, else, it is false. Elements of the deleted documents in the inverted index are purged.</dd>
+</dl>
+
+<p>The function `odname' is used in order to get the name of a database.</p>
+
+<dl>
+<dt><kbd>char *odname(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the pointer to the region of the name of the database, else, it is `NULL'. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odfsiz' is used in order to get the total size of database files.</p>
+
+<dl>
+<dt><kbd>double odfsiz(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the total size of the database files, else, it is -1.0.</dd>
+</dl>
+
+<p>The function `odbnum' is used in order to get the total number of the elements of the bucket arrays in the inverted index.</p>
+
+<dl>
+<dt><kbd>int odbnum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the total number of the elements of the bucket arrays, else, it is -1.</dd>
+</dl>
+
+<p>The function `odbusenum' is used in order to get the total number of the used elements of the bucket arrays in the inverted index.</p>
+
+<dl>
+<dt><kbd>int odbusenum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the total number of the used elements of the bucket arrays, else, it is -1.</dd>
+</dl>
+
+<p>The function `oddnum' is used in order to get the number of the documents stored in a database.</p>
+
+<dl>
+<dt><kbd>int oddnum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the number of the documents stored in the database, else, it is -1.</dd>
+</dl>
+
+<p>The function `odwnum' is used in order to get the number of the words stored in a database.</p>
+
+<dl>
+<dt><kbd>int odwnum(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. If successful, the return value is the number of the words stored in the database, else, it is -1. Because of the I/O buffer, the return value may be less than the hard number.</dd>
+</dl>
+
+<p>The function `odwritable' is used in order to check whether a database handle is a writer or not.</p>
+
+<dl>
+<dt><kbd>int odwritable(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. The return value is true if the handle is a writer, false if not.</dd>
+</dl>
+
+<p>The function `odfatalerror' is used in order to check whether a database has a fatal error or not.</p>
+
+<dl>
+<dt><kbd>int odfatalerror(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. The return value is true if the database has a fatal error, false if not.</dd>
+</dl>
+
+<p>The function `odinode' is used in order to get the inode number of a database directory.</p>
+
+<dl>
+<dt><kbd>int odinode(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. The return value is the inode number of the database directory.</dd>
+</dl>
+
+<p>The function `odmtime' is used in order to get the last modified time of a database.</p>
+
+<dl>
+<dt><kbd>time_t odmtime(ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. The return value is the last modified time of the database.</dd>
+</dl>
+
+<p>The function `odmerge' is used in order to merge plural database directories.</p>
+
+<dl>
+<dt><kbd>int odmerge(const char *<var>name</var>, const CBLIST *<var>elemnames</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory to create. `elemnames' specifies a list of names of element databases. If successful, the return value is true, else, it is false. If two or more documents which have the same URL come in, the first one is adopted and the others are ignored.</dd>
+</dl>
+
+<p>The function `odremove' is used in order to remove a database directory.</p>
+
+<dl>
+<dt><kbd>int odremove(const char *<var>name</var>);</kbd></dt>
+<dd>`name' specifies the name of a database directory. If successful, the return value is true, else, it is false. A database directory can contain databases of other APIs of QDBM, they are also removed by this function.</dd>
+</dl>
+
+<p>The function `oddocopen' is used in order to get a document handle.</p>
+
+<dl>
+<dt><kbd>ODDOC *oddocopen(const char *<var>uri</var>);</kbd></dt>
+<dd>`uri' specifies the URI of a document. The return value is a document handle. The ID number of a new document is not defined. It is defined when the document is stored in a database.</dd>
+</dl>
+
+<p>The function `oddocclose' is used in order to close a document handle.</p>
+
+<dl>
+<dt><kbd>void oddocclose(ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. Because the region of a closed handle is released, it becomes impossible to use the handle.</dd>
+</dl>
+
+<p>The function `oddocaddattr' is used in order to add an attribute to a document.</p>
+
+<dl>
+<dt><kbd>void oddocaddattr(ODDOC *<var>doc</var>, const char *<var>name</var>, const char *<var>value</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. `name' specifies the string of the name of an attribute. `value' specifies the string of the value of the attribute.</dd>
+</dl>
+
+<p>The function `oddocaddword' is used in order to add a word to a document.</p>
+
+<dl>
+<dt><kbd>void oddocaddword(ODDOC *<var>doc</var>, const char *<var>normal</var>, const char *<var>asis</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. `normal' specifies the string of the normalized form of a word. Normalized forms are treated as keys of the inverted index. If the normalized form of a word is an empty string, the word is not reflected in the inverted index. `asis' specifies the string of the appearance form of the word. Appearance forms are used after the document is retrieved by an application.</dd>
+</dl>
+
+<p>The function `oddocid' is used in order to get the ID number of a document.</p>
+
+<dl>
+<dt><kbd>int oddocid(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. The return value is the ID number of a document.</dd>
+</dl>
+
+<p>The function `oddocuri' is used in order to get the URI of a document.</p>
+
+<dl>
+<dt><kbd>const char *oddocuri(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. The return value is the string of the URI of a document.</dd>
+</dl>
+
+<p>The function `oddocgetattr' is used in order to get the value of an attribute of a document.</p>
+
+<dl>
+<dt><kbd>const char *oddocgetattr(const ODDOC *<var>doc</var>, const char *<var>name</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. `name' specifies the string of the name of an attribute. The return value is the string of the value of the attribute, or `NULL' if no attribute corresponds.</dd>
+</dl>
+
+<p>The function `oddocnwords' is used in order to get the list handle contains words in normalized form of a document.</p>
+
+<dl>
+<dt><kbd>const CBLIST *oddocnwords(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. The return value is the list handle contains words in normalized form.</dd>
+</dl>
+
+<p>The function `oddocawords' is used in order to get the list handle contains words in appearance form of a document.</p>
+
+<dl>
+<dt><kbd>const CBLIST *oddocawords(const ODDOC *<var>doc</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. The return value is the list handle contains words in appearance form.</dd>
+</dl>
+
+<p>The function `oddocscores' is used in order to get the map handle contains keywords in normalized form and their scores.</p>
+
+<dl>
+<dt><kbd>CBMAP *oddocscores(const ODDOC *<var>doc</var>, int <var>max</var>, ODEUM *<var>odeum</var>);</kbd></dt>
+<dd>`doc' specifies a document handle. `max' specifies the max number of keywords to get. `odeum' specifies a database handle with which the IDF for weighting is calculate. If it is `NULL', it is not used. The return value is the map handle contains keywords and their scores. Scores are expressed as decimal strings. Because the handle of the return value is opened with the function `cbmapopen', it should be closed with the function `cbmapclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odbreaktext' is used in order to break a text into words in appearance form.</p>
+
+<dl>
+<dt><kbd>CBLIST *odbreaktext(const char *<var>text</var>);</kbd></dt>
+<dd>`text' specifies the string of a text. The return value is the list handle contains words in appearance form. Words are separated with space characters and such delimiters as period, comma and so on. Because the handle of the return value is opened with the function `cblistopen', it should be closed with the function `cblistclose' if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odnormalizeword' is used in order to make the normalized form of a word.</p>
+
+<dl>
+<dt><kbd>char *odnormalizeword(const char *<var>asis</var>);</kbd></dt>
+<dd>`asis' specifies the string of the appearance form of a word. The return value is is the string of the normalized form of the word. Alphabets of the ASCII code are unified into lower cases. Words composed of only delimiters are treated as empty strings. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odpairsand' is used in order to get the common elements of two sets of documents.</p>
+
+<dl>
+<dt><kbd>ODPAIR *odpairsand(ODPAIR *<var>apairs</var>, int <var>anum</var>, ODPAIR *<var>bpairs</var>, int <var>bnum</var>, int *<var>np</var>);</kbd></dt>
+<dd>`apairs' specifies the pointer to the former document array. `anum' specifies the number of the elements of the former document array. `bpairs' specifies the pointer to the latter document array. `bnum' specifies the number of the elements of the latter document array. `np' specifies the pointer to a variable to which the number of the elements of the return value is assigned. The return value is the pointer to a new document array whose elements commonly belong to the specified two sets. Elements of the array are sorted in descending order of their scores. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odpairsor' is used in order to get the sum of elements of two sets of documents.</p>
+
+<dl>
+<dt><kbd>ODPAIR *odpairsor(ODPAIR *<var>apairs</var>, int <var>anum</var>, ODPAIR *<var>bpairs</var>, int <var>bnum</var>, int *<var>np</var>);</kbd></dt>
+<dd>`apairs' specifies the pointer to the former document array. `anum' specifies the number of the elements of the former document array. `bpairs' specifies the pointer to the latter document array. `bnum' specifies the number of the elements of the latter document array. `np' specifies the pointer to a variable to which the number of the elements of the return value is assigned. The return value is the pointer to a new document array whose elements belong to both or either of the specified two sets. Elements of the array are sorted in descending order of their scores. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odpairsnotand' is used in order to get the difference set of documents.</p>
+
+<dl>
+<dt><kbd>ODPAIR *odpairsnotand(ODPAIR *<var>apairs</var>, int <var>anum</var>, ODPAIR *<var>bpairs</var>, int <var>bnum</var>, int *<var>np</var>);</kbd></dt>
+<dd>`apairs' specifies the pointer to the former document array. `anum' specifies the number of the elements of the former document array. `bpairs' specifies the pointer to the latter document array of the sum of elements. `bnum' specifies the number of the elements of the latter document array. `np' specifies the pointer to a variable to which the number of the elements of the return value is assigned. The return value is the pointer to a new document array whose elements belong to the former set but not to the latter set. Elements of the array are sorted in descending order of their scores. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use.</dd>
+</dl>
+
+<p>The function `odpairssort' is used in order to sort a set of documents in descending order of scores.</p>
+
+<dl>
+<dt><kbd>void odpairssort(ODPAIR *<var>pairs</var>, int <var>pnum</var>);</kbd></dt>
+<dd>`pairs' specifies the pointer to a document array. `pnum' specifies the number of the elements of the document array.</dd>
+</dl>
+
+<p>The function `odlogarithm' is used in order to get the natural logarithm of a number.</p>
+
+<dl>
+<dt><kbd>double odlogarithm(double <var>x</var>);</kbd></dt>
+<dd>`x' specifies a number. The return value is the natural logarithm of the number. If the number is equal to or less than 1.0, the return value is 0.0. This function is useful when an application calculates the IDF of search results.</dd>
+</dl>
+
+<p>The function `odvectorcosine' is used in order to get the cosine of the angle of two vectors.</p>
+
+<dl>
+<dt><kbd>double odvectorcosine(const int *<var>avec</var>, const int *<var>bvec</var>, int <var>vnum</var>);</kbd></dt>
+<dd>`avec' specifies the pointer to one array of numbers. `bvec' specifies the pointer to the other array of numbers. `vnum' specifies the number of elements of each array. The return value is the cosine of the angle of two vectors. This function is useful when an application calculates similarity of documents.</dd>
+</dl>
+
+<p>The function `odsettuning' is used in order to set the global tuning parameters.</p>
+
+<dl>
+<dt><kbd>void odsettuning(int <var>ibnum</var>, int <var>idnum</var>, int <var>cbnum</var>, int <var>csiz</var>);</kbd></dt>
+<dd>`ibnum' specifies the number of buckets for inverted indexes. `idnum' specifies the division number of inverted index. `cbnum' specifies the number of buckets for dirty buffers. `csiz' specifies the maximum bytes to use memory for dirty buffers. The default setting is equivalent to `odsettuning(32749, 7, 262139, 8388608)'. This function should be called before opening a handle.</dd>
+</dl>
+
+<p>The function `odanalyzetext' is used in order to break a text into words and store appearance forms and normalized form into lists.</p>
+
+<dl>
+<dt><kbd>void odanalyzetext(ODEUM *<var>odeum</var>, const char *<var>text</var>, CBLIST *<var>awords</var>, CBLIST *<var>nwords</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `text' specifies the string of a text. `awords' specifies a list handle into which appearance form is store. `nwords' specifies a list handle into which normalized form is store. If it is `NULL', it is ignored. Words are separated with space characters and such delimiters as period, comma and so on.</dd>
+</dl>
+
+<p>The function `odsetcharclass' is used in order to set the classes of characters used by `odanalyzetext'.</p>
+
+<dl>
+<dt><kbd>void odsetcharclass(ODEUM *<var>odeum</var>, const char *<var>spacechars</var>, const char *<var>delimchars</var>, const char *<var>gluechars</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. `spacechars' spacifies a string contains space characters. `delimchars' spacifies a string contains delimiter characters. `gluechars' spacifies a string contains glue characters.</dd>
+</dl>
+
+<p>The function `odquery' is used in order to query a database using a small boolean query language.</p>
+
+<dl>
+<dt><kbd>ODPAIR *odquery(ODEUM *<var>odeum</var>, const char *<var>query</var>, int *<var>np</var>, CBLIST *<var>errors</var>);</kbd></dt>
+<dd>`odeum' specifies a database handle. 'query' specifies the text of the query. `np' specifies the pointer to a variable to which the number of the elements of the return value is assigned. `errors' specifies a list handle into which error messages are stored. If it is `NULL', it is ignored. If successful, the return value is the pointer to an array, else, it is `NULL'. Each element of the array is a pair of the ID number and the score of a document, and sorted in descending order of their scores. Even if no document corresponds to the specified condition, it is not error but returns an dummy array. Because the region of the return value is allocated with the `malloc' call, it should be released with the `free' call if it is no longer in use. Note that each element of the array of the return value can be data of a deleted document.</dd>
+</dl>
+
+<h3>Examples</h3>
+
+<p>The following example stores a document into the database.</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#define DBNAME "index"
+
+int main(int argc, char **argv){
+ ODEUM *odeum;
+ ODDOC *doc;
+ CBLIST *awords;
+ const char *asis;
+ char *normal;
+ int i;
+
+ /* open the database */
+ if(!(odeum = odopen(DBNAME, OD_OWRITER | OD_OCREAT))){
+ fprintf(stderr, "odopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* get the document handle */
+ doc = oddocopen("http://www.foo.bar/baz.txt");
+
+ /* set attributes of the document */
+ oddocaddattr(doc, "title", "Balcony Scene");
+ oddocaddattr(doc, "author", "Shakespeare");
+
+ /* break the text and get the word list */
+ awords = odbreaktext("Parting is such sweet sorrow.");
+
+ /* set each word into the document handle */
+ for(i = 0; i < cblistnum(awords); i++){
+ /* get one word of the list */
+ asis = cblistval(awords, i, NULL);
+ /* get the normalized form from the appearance form */
+ normal = odnormalizeword(asis);
+ /* add the word into the document handle */
+ oddocaddword(doc, normal, asis);
+ /* release the region of the normalized form */
+ free(normal);
+ }
+
+ /* store the document into the database */
+ if(!odput(odeum, doc, -1, 1)){
+ fprintf(stderr, "odput: %s\n", dperrmsg(dpecode));
+ }
+
+ /* release the word list */
+ cblistclose(awords);
+
+ /* release the document handle */
+ oddocclose(doc);
+
+ /* close the database */
+ if(!odclose(odeum)){
+ fprintf(stderr, "odclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<p>The following example retrieves documents.</p>
+
+<pre>#include <depot.h>
+#include <cabin.h>
+#include <odeum.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#define DBNAME "index"
+
+int main(int argc, char **argv){
+ ODEUM *odeum;
+ ODPAIR *pairs;
+ ODDOC *doc;
+ const CBLIST *words;
+ const char *title, *author, *asis;
+ int i, j, pnum;
+
+ /* open the database */
+ if(!(odeum = odopen(DBNAME, OD_OREADER))){
+ fprintf(stderr, "odopen: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ /* retrieve documents */
+ if((pairs = odsearch(odeum, "sorrow", -1, &pnum)) != NULL){
+
+ /* scan each element of the document array */
+ for(i = 0; i < pnum; i++){
+ /* get the document handle */
+ if(!(doc = odgetbyid(odeum, pairs[i].id))) continue;
+ /* show the attributes */
+ printf("URI: %s\n", oddocuri(doc));
+ title = oddocgetattr(doc, "title");
+ if(title) printf("TITLE: %s\n", title);
+ author = oddocgetattr(doc, "author");
+ if(author) printf("AUTHOR: %s\n", author);
+ /* show words in appearance form */
+ printf("WORDS:");
+ words = oddocawords(doc);
+ for(j = 0; j < cblistnum(words); j++){
+ asis = cblistval(words, j, NULL);
+ printf(" %s", asis);
+ }
+ putchar('\n');
+ /* release the document handle */
+ oddocclose(doc);
+ }
+
+ /* release the document array */
+ free(pairs);
+
+ } else {
+ fprintf(stderr, "odsearch: %s\n", dperrmsg(dpecode));
+ }
+
+ /* close the database */
+ if(!odclose(odeum)){
+ fprintf(stderr, "odclose: %s\n", dperrmsg(dpecode));
+ return 1;
+ }
+
+ return 0;
+}
+</pre>
+
+<h3>Notes</h3>
+
+<p>How to build programs using Odeum is the same as the case of Depot.</p>
+
+<pre>gcc -I/usr/local/include -o sample sample.c -L/usr/local/lib -lqdbm
+</pre>
+
+<p>If QDBM was built with POSIX thread enabled, the global variable `dpecode' is treated as thread specific data, and functions of Odeum are reentrant. In that case, they are thread-safe as long as a handle is not accessed by threads at the same time, on the assumption that `errno', `malloc', and so on are thread-safe.</p>
+
+<p>If QDBM was built with ZLIB enabled, records in the database for document attributes are compressed. In that case, the size of the database is reduced to 30% or less. Thus, you should enable ZLIB if you use Odeum. A database of Odeum created without ZLIB enabled is not available on environment with ZLIB enabled, and vice versa. If ZLIB was not enabled but LZO, LZO is used instead.</p>
+
+<h3>Query Language</h3>
+
+<p>The query language of the function `odquery' is a basic language following this grammar:</p>
+
+<pre>expr ::= subexpr ( op subexpr )*
+subexpr ::= WORD
+subexpr ::= LPAREN expr RPAREN
+</pre>
+
+<p>Operators are "&" (AND), "|" (OR), and "!" (NOTAND). You can use parenthesis to group sub-expressions together in order to change order of operations. The given query is broken up using the function `odanalyzetext', so if you want to specify different text breaking rules, then make sure that you at least set "&", "|", "!", "(", and ")" to be delimiter characters. Consecutive words are treated as having an implicit "&" operator between them, so "zed shaw" is actually "zed & shaw".</p>
+
+<p>The encoding of the query text should be the same with the encoding of target documents. Moreover, each of space characters, delimiter characters, and glue characters should be single byte.</p>
+
+<hr />
+
+<h2><a name="odeumcli" id="odeumcli" class="head">Commands for Odeum</a></h2>
+
+<p>Odeum has the following command line interfaces.</p>
+
+<p>The command `odmgr' is a utility for debugging Odeum and its applications. It features editing and checking of a database. It can be used for full-text search systems with shell scripts. This command is used in the following format. `name' specifies a database name. `file' specifies a file name, `expr' specifies the URI or the ID number of a document, `words' specifies searching words. `elems' specifies element databases.</p>
+
+<dl>
+<dt><kbd>odmgr create <var>name</var></kbd></dt>
+<dd>Create a database file.</dd>
+<dt><kbd>odmgr put [-uri <var>str</var>] [-title <var>str</var>] [-author <var>str</var>] [-date <var>str</var>] [-wmax <var>num</var>] [-keep] <var>name</var> [<var>file</var>]</kbd></dt>
+<dd>Add a document by reading a file. If `file' is omitted, the standard input is read and URI is needed.</dd>
+<dt><kbd>odmgr out [-id] <var>name</var> <var>expr</var></kbd></dt>
+<dd>Delete a document specified by a URI.</dd>
+<dt><kbd>odmgr get [-id] [-t|-h] <var>name</var> <var>expr</var></kbd></dt>
+<dd>Show a document specified by a URI. The output is the ID number and the URI of a document, in tab separated format.</dd>
+<dt><kbd>odmgr search [-max <var>num</var>] [-or] [-idf] [-t|-h|-n] <var>name</var> <var>words</var>...</kbd></dt>
+<dd>Retrieve documents including specified words. The first line of the output is the total number of hits and each word with its number of hits, in tab separated format. The second line and below are the ID numbers and the scores of documents, in tab separated format.</dd>
+<dt><kbd>odmgr list [-t|-h] <var>name</var></kbd></dt>
+<dd>Show all documents in a database. Each line of the output is the ID number and the score of a document, in tab separated format.</dd>
+<dt><kbd>odmgr optimize <var>name</var></kbd></dt>
+<dd>Optimize a database.</dd>
+<dt><kbd>odmgr inform <var>name</var></kbd></dt>
+<dd>Output miscellaneous information.</dd>
+<dt><kbd>odmgr merge <var>name</var> <var>elems</var>...</kbd></dt>
+<dd>Merge plural databases.</dd>
+<dt><kbd>odmgr remove <var>name</var></kbd></dt>
+<dd>Remove a database directory.</dd>
+<dt><kbd>odmgr break [-h|-k|-s] [<var>file</var>]</kbd></dt>
+<dd>Read a file and output words in the text. Each line of the output is the appearance form and the normalized form of a word, in tab separated format.</dd>
+<dt><kbd>odmgr version</kbd></dt>
+<dd>Output version information of QDBM.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-uri <var>str</var></kbd> : specify the URI of the document explicitly.</li>
+<li><kbd>-title <var>str</var></kbd> : specify the title of the document.</li>
+<li><kbd>-author <var>str</var></kbd> : specify the author of the document.</li>
+<li><kbd>-date <var>str</var></kbd> : specify the modified date of the document.</li>
+<li><kbd>-wmax <var>num</var></kbd> : specify the max number of words to be stored.</li>
+<li><kbd>-keep</kbd> : the storing mode is not to be overwrite.</li>
+<li><kbd>-id</kbd> : specify a document not by a URI but by an ID number.</li>
+<li><kbd>-t</kbd> : output the details of a document in tab separated format.</li>
+<li><kbd>-h</kbd> : output the details of a document in human-readable format.</li>
+<li><kbd>-k</kbd> : output keywords of a document.</li>
+<li><kbd>-s</kbd> : output summary of a document.</li>
+<li><kbd>-max <var>num</var></kbd> : specify the max number of documents of the output.</li>
+<li><kbd>-or</kbd> : perform OR search, nut AND search.</li>
+<li><kbd>-idf</kbd> : tune scores with IDF.</li>
+<li><kbd>-n</kbd> : show ID numbers and scores only.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `odtest' is a utility for facility test and performance test. Check a database generated by the command or measure the execution time of the command. This command is used in the following format. `name' specifies a database name. `dnum' specifies the number of the documents. `wnum' specifies the number of words per document. `pnum' specifies the number of patterns of words.</p>
+
+<dl>
+<dt><kbd>odtest write [-tune <var>ibnum</var> <var>idnum</var> <var>cbnum</var> <var>csiz</var>] <var>name</var> <var>dnum</var> <var>wnum</var> <var>pnum</var></kbd></dt>
+<dd>Store documents with random attributes and random words.</dd>
+<dt><kbd>odtest read <var>name</var></kbd></dt>
+<dd>Retrieve all documents of the database above.</dd>
+<dt><kbd>odtest combo <var>name</var></kbd></dt>
+<dd>Perform combination test of various operations.</dd>
+<dt><kbd>odtest wicked <var>name</var> <var>dnum</var></kbd></dt>
+<dd>Perform updating operations selected at random.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-tune <var>ibnum</var> <var>idnum</var> <var>cbnum</var> <var>csiz</var></kbd> : set tuning parameters.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure. The environment variable `QDBMDBGFD' specifies the file descriptor to output the history of updating the variable `dpecode'.</p>
+
+<p>The command `odidx' is a utility for indexing files on the local file system. This command is useful for a full-text search system of a Web site. Supported file format are plain text and HTML. Supported character encoding are US-ASCII and ISO-8859-1. The URI of each document is specified with the path of its file. Attributes named as `title' and `date' are given to each documents. When a document is already existing in the database, it is registered if its last modification time is newer, or it is ignored. Modification times are stored in the sub database `_mtime' in the main database directory. Score information are stored in the sub database `_score' in the main database directory. This command is used in the following format. `name' specifies a database name. `dir' specifies a directory name.</p>
+
+<dl>
+<dt><kbd>odidx register [-l <var>file</var>] [-wmax <var>num</var>] [-tsuf <var>sufs</var>] [-hsuf <var>sufs</var>] <var>name</var> [<var>dir</var>]</kbd></dt>
+<dd>Register files in the specified directory. If `dir' is omitted, the current directory is specified.</dd>
+<dt><kbd>odidx relate <var>name</var></kbd></dt>
+<dd>Add score information for relational document search to each documents in the database.</dd>
+<dt><kbd>odidx purge <var>name</var></kbd></dt>
+<dd>Purge documents which are not existing on the local files system.</dd>
+</dl>
+
+<p>Options feature the following.</p>
+
+<ul class="lines">
+<li><kbd>-l <var>file</var></kbd> : read a file and get list of paths of files to register. If `-' is specified, the standard input is read.</li>
+<li><kbd>-wmax <var>num</var></kbd> : specify the max number of words to be stored in the document database.</li>
+<li><kbd>-tsuf <var>sufs</var></kbd> : specify suffixes of plain text files in comma separated format. The default is `-tsuf .txt,.text'.</li>
+<li><kbd>-hsuf <var>sufs</var></kbd> : specify suffixes of HTML files in comma separated format. The default is `-hsuf .html,.htm'.</li>
+</ul>
+
+<p>This command returns 0 on success, another on failure.</p>
+
+<p>Commands of Odeum make it easy to realize a full-text search system. For example, to register files which are under `/home/mikio' and whose suffix are `.txt', `.c', or `.h', perform the following command.</p>
+
+<pre>odidx register -tsuf ".txt,.c,.h" -hsuf "" casket /home/mikio
+</pre>
+
+<p>Thus, to retrieve documents which include `unix' and `posix' and show the top 8 terms, perform the following command.</p>
+
+<pre>odmgr search -max 8 -h casket "unix posix"
+</pre>
+
+<p>A database generated by `odidx' is available with the CGI script which is included in QDBM for full-text search.</p>
+
+<hr />
+
+<h2><a name="fileformat" id="fileformat" class="head">File Format</a></h2>
+
+<h3>File Format of Depot</h3>
+
+<p>The contents of a database file managed by Depot is divided roughly into the following three sections: the header section, the bucket section and the record section.</p>
+
+<p>The header section places at the beginning of the file and its length is constant 48 bytes. The following information are stored in the header section.</p>
+
+<ol>
+<li>magic number: from offset 0, contains "[DEPOT]\n\f" for big endian or "[depot]\n\f" for little endian.</li>
+<li>version number: decimal string of the version number of the library.</li>
+<li>flags for wrappers: from offset 16, type of `int'.</li>
+<li>file size: from offset 24, type of `int'.</li>
+<li>number of the bucket: from offset 32, type of `int'.</li>
+<li>number of records: from offset 40, type of `int'.</li>
+</ol>
+
+<p>The bucket section places after the header section and its length is determined according to the number of the bucket. Each element of the bucket stores an offset of the root node of each separate chain.</p>
+
+<p>The record section places after the bucket section and occupies to the end of the file. The element of the record section contains the following information.</p>
+
+<ol>
+<li>flags: type of `int'.</li>
+<li>second hash value: type of `int'.</li>
+<li>size of the key: type of `int'.</li>
+<li>size of the value: type of `int'.</li>
+<li>size of the padding: type of `int'.</li>
+<li>offset of the left child: type of `int'.</li>
+<li>offset of the right child: type of `int'.</li>
+<li>entity of the key: serial bytes with variable length.</li>
+<li>entity of the value: serial bytes with variable length.</li>
+<li>padding data: void serial bytes with variable length.</li>
+</ol>
+
+<h3>File Format of Villa</h3>
+
+<p>Every data handled by Villa is stored in a database of Depot. Storing data is divided into meta data and logical pages. Logical pages can be classified into leaf nodes and non-leaf nodes. Meta data are such managing information as the number of records. Both of its key and its value are type of `int'. Leaf nodes hold records. Non-leaf nodes hold sparse index referring to pages.</p>
+
+<p>Villa uses variable length numeric format (BER compression) to handle small natural number with frugal storage. A variable length numeric object is parsed from the top of the region and parsing ends at the byte of positive value. Each byte are evaluated as absolute value and calculated as little endian number based on the radix 128.</p>
+
+<p>Record is logical unit of user data. Some records overlapping keys are shaped into one physical record. A Physical record is serialized in the following format.</p>
+
+<ol>
+<li>size of the key: type of variable length number</li>
+<li>entity of the key: serial bytes with variable length</li>
+<li>number of values: type of variable length number</li>
+<li>list of values: serial bytes repeating the following expressions<ol>
+<li>size: type of variable length number</li>
+<li>entity of the key: serial bytes with variable length</li>
+</ol></li>
+</ol>
+
+<p>Leaf node is physical unit to store a set of records. The key of a leaf node is its ID whose type is `int'. A leaf node is stored in a database of Depot with the following values. Its records are sorted in ascending order of each key.</p>
+
+<ol>
+<li>ID of the previous leaf: type of variable length number</li>
+<li>ID of the next leaf: type of variable length number</li>
+<li>list of records: concatenation of serialized records</li>
+</ol>
+
+<p>Index is logical unit of a pointer to search for pages. An index is serialized int the following format.</p>
+
+<ol>
+<li>ID of the referring page: type of variable length number</li>
+<li>size of the key: type of variable length number</li>
+<li>entity of the key: serial bytes with variable length</li>
+</ol>
+
+<p>Non-leaf node is physical unit to store a set of indexes. The key of a non-leaf node is its ID whose type is `int'. A non-leaf node is stored in a database of Depot with the following values. Its indexes are sorted in ascending order of each key.</p>
+
+<ol>
+<li>ID of the first child node: type of variable length number</li>
+<li>list of indexes: concatenation of serialized indexes</li>
+</ol>
+
+<h3>Notes</h3>
+
+<p>Because the database file is not sparse, move, copy, unlink, ftp, and so on with the file are possible. Because Depot reads and writes data without normalization of byte order, it is impossible to share the same file between the environment with different byte order.</p>
+
+<p>When you distribute a database file of Depot or Villa via network, the MIME type suggested to be `application/x-qdbm'. Suffix of the file name is suggested to be `.qdb'. When you distribute a database directory of Curia, you may convert the directory tree to an archive of such type as TAR.</p>
+
+<p>For the command `file' to recognize database files, append the following expressions into `magic' file.</p>
+
+<pre>0 string [DEPOT]\n\f QDBM, big endian
+>12 string x \b, version=%s
+>19 byte ^1 \b, Hash
+>19 byte &1 \b, B+ tree
+>19 byte &2 \b (deflated:ZLIB)
+>19 byte &4 \b (deflated:LZO)
+>19 byte &8 \b (deflated:BZIP2)
+>24 belong x \b, filesize=%d
+>32 belong x \b, buckets=%d
+>40 belong x \b, records=%d
+0 string [depot]\n\f QDBM, little endian
+>12 string x \b, version=%s
+>16 byte ^1 \b, Hash
+>16 byte &1 \b, B+ tree
+>16 byte &2 \b (deflated:ZLIB)
+>16 byte &4 \b (deflated:LZO)
+>16 byte &8 \b (deflated:BZIP2)
+>24 lelong x \b, filesize=%d
+>32 lelong x \b, buckets=%d
+>40 lelong x \b, records=%d
+</pre>
+
+<hr />
+
+<h2><a name="porting" id="porting" class="head">Porting</a></h2>
+
+<p>One of the goal of QDBM is to work on all platforms which conform to POSIX. Even if some APIs are not implemented, QDBM should work. Moreover, it should be possible to build QDBM using compilers other than GCC. Porting to various platforms is performed to add a new `Makefile' or modify some parts of source files. As for APIs of C, some of the following files should be modified. Otherwise, you can create new files based on them.</p>
+
+<ul class="lines">
+<li><kbd>Makefile.in</kbd> : base of `Makefile', used by `./configure'.</li>
+<li><kbd>myconf.h</kbd> : configuration of system dependency.</li>
+<li><kbd>depot.h</kbd> : header of the basic API.</li>
+<li><kbd>curia.h</kbd> : header of the extended API.</li>
+<li><kbd>relic.h</kbd> : header of the NDBM-compatible API.</li>
+<li><kbd>hovel.h</kbd> : header of the GDBM-compatible API.</li>
+<li><kbd>cabin.h</kbd> : header of the utility API.</li>
+<li><kbd>villa.h</kbd> : header of the advanced API.</li>
+<li><kbd>vista.h</kbd> : header of the extended advanced API.</li>
+<li><kbd>odeum.h</kbd> : header of the inverted API.</li>
+<li><kbd>myconf.c</kbd> : implementation of system dependency.</li>
+<li><kbd>depot.c</kbd> : implementation of the basic API.</li>
+<li><kbd>curia.c</kbd> : implementation of the extended API.</li>
+<li><kbd>relic.c</kbd> : implementation of the NDBM-compatible API.</li>
+<li><kbd>hovel.c</kbd> : implementation of the GDBM-compatible API.</li>
+<li><kbd>cabin.c</kbd> : implementation of the utility API.</li>
+<li><kbd>villa.c</kbd> : implementation of the advanced API.</li>
+<li><kbd>vista.c</kbd> : implementation of the extended advanced API.</li>
+<li><kbd>odeum.c</kbd> : implementation of the inverted API.</li>
+</ul>
+
+<p>On platforms which do not support file locking with `fcntl' call, you should append `-DMYNOLOCK' to the macro `CFLAGS' defined in `Makefile'. In that case, you should consider another exclusion control. As with it, on platforms without `mmap' call, you should append `-DMYNOMMAP' to `CFLAGS'. As for `mmap', its emulation using `malloc' and so on is provided. If other system calls are not implemented, you should define emulation by modification of `myconf.h' and `myconf.c'.</p>
+
+<p>Because POSIX thread is used in C++ API, it is impossible to port C++ API to platforms without the package. Because JNI is used in Java API, you should pay attention to location of the headers and libraries. Moreover, you should consider such type definitions as `long long' or `int64'. Because APIs of Perl and Ruby use building commands provided with each language system, you should be knowledgeable about their specifications.</p>
+
+<hr />
+
+<h2><a name="bugs" id="bugs" class="head">Bugs</a></h2>
+
+<p>Each document of QDBM should be calibrated by native English speakers.</p>
+
+<p>There is no such bug which are found but not fixed, as crash by segmentation fault, unexpected data vanishing, memory leak and so on.</p>
+
+<p>If you find any bug, report it to the author, with the information of the version of QDBM, the operating system and the compiler.</p>
+
+<p>Databases created with QDBM version 1.7.13 or earlier are not compatible to ones of the later versions.</p>
+
+<hr />
+
+<h2><a name="faq" id="faq" class="head">Frequently Asked Questions</a></h2>
+
+<dl>
+<dt>Q. : Does QDBM support SQL?</dt>
+<dd>A. : No, it does not. QDBM is not a RDBMS (Relational Database Management System). If you want an embedded RDBMS, use SQLite and so on.</dd>
+<dt>Q. : After all, how different from GDBM (NDBM, SDBM, Berkeley DB)?</dt>
+<dd>A. : Processing speed is higher, a database file is smaller, API is simpler. A highly important thing is that efficiency in time and space is very good when records are frequently overwritten, so, scalability in practical use is high. Moreover, even when constructing such a large database that the number of storing record is more than one million, processing speed does not slowdown deathly, filesize does not grow extremely. However, because other DBM or DBMS may be more suitable in some cases, comparing performance and functionality by yourself is suggested.</dd>
+<dt>Q. : Which API should I use?</dt>
+<dd>A. : If you search for records as complete accord, try Depot. If the scale is large, try Curia. If you access records in some order, try Villa. If the scale is large, try Vista. If you pursue the greatest number of records, build QDBM with ZLIB or LZO enabled and use Vista.</dd>
+<dt>Q. : What is bibliography?</dt>
+<dd>A. : Algorithms of QDBM are mainly based on the descriptions in `Data Structures and Algorithms' by Aho et al and `Algorithms in C' by Sedgewick.</dd>
+<dt>Q. : Are there good sample codes for applications?</dt>
+<dd>A. : Refer to the source code of commands of each API. `dptsv.c', `crtsv.c' and `vltsv.c' are simplest.</dd>
+<dt>Q. : My database file has been broken. Why?</dt>
+<dd>A. : In most cases, the reason is that your application did not close the database on exit. No matter whether it is a demon process or a CGI script, any application should close handling databases when it exits. Moreover, we should remember that a process of CGI may be killed by SIGPIPE or SIGTERM.</dd>
+<dt>Q. : How robust are databases of QDBM?</dt>
+<dd>A. : QDBM does not assure absolute robustness. A database may be broken if your operating system crashes. Although transaction of Villa can save a database from crashes of applications, it is inadequate to crashes of operating systems. So, you should consider multiplexing of a database or backup system if you use QDBM for mission critical applications.</dd>
+<dt>Q: How should I use alignment of Depot and Curia?</dt>
+<dd>A: If your application repeats writing with overwrite or concatenate mode. Alignment saves the rapid growth of the size of the database file. Because the best suited size of alignment of each application is different, you should learn it by experiment. For the meantime, about 32 is suitable.</dd>
+<dt>Q. : How should I tune performance parameters of Villa?</dt>
+<dd>A. : If you perform mainly ordering access, `lrecmax' and `nidxmax' should be larger. If you perform mainly random access, they should be less. If RAM of your system is abundant, `lcnum' and `ncnum' should be increased in order to improve performance. If ZLIB, LZO, or BZIP2 is enabled, increase `lrecmax' and compression efficiency is improved.</dd>
+<dt>Q. : Which is the most preferable of ZLIB, LZO or BZIP2 for Villa?</dt>
+<dd>A. : BZIP2 has the best compression retio. LZO has the best compression speed. ZLIB takes a mean position of them. If you don't have any special reason, using ZLIB is suggested. However, if updating of the database is frequent, LZO is more preferable. If updating of the database is very infrequently, BZIP2 is more preferable. Note that the license of LZO is the GNU LGPL.</dd>
+<dt>Q. : What is `sparse file'?</dt>
+<dd>A. : It is a file where some holes are. `Hole' means a block where any data has never written in. If a file system supports sparse file, holes are not allocated into any physical storage. As for QDBM, if a database is created with such flags as DP_OSPARSE, the bucket array is not initialized and its blocks become holes. According to that mechanism, you can use greatly huge hash tables. However, its performance is strongly depends on the setting of the file system.</dd>
+<dt>Q. : Why Depot and Curia do not feature transaction?</dt>
+<dd>A. : If an application implements its own transaction, inner transaction of database is superfluous. You can implement transaction for application easily with hash map provided by Cabin.</dd>
+<dt>Q. : How should I tune the system for performance?</dt>
+<dd>A. : Install more RAM on your machine than the size of a database. Then, enlarge I/O buffer and cut down on flushing dirty buffers. File system is also important. On Linux, although EXT2 is usually fastest, EXT3 is faster in some cases. ReiserFS is okey. The other modes of EXT3 are very slow. About other file systems, you should learn them by experiment.</dd>
+<dt>Q. : Can I build QDBM using `cc' instead of `gcc'?</dt>
+<dd>A. : Yes. Try to build QDBM with `LTmakefile'.</dd>
+<dt>Q. : Can I build QDBM using Visual C++?</dt>
+<dd>A. : Yes. Use `VCmakefile' instead of `Makefile'.</dd>
+<dt>Q. : Can I use QDBM in other languages?</dt>
+<dd>A. : As for PHP, Scheme (Gauche), and OCaml, interfaces of QDBM have been released. If you need it for another language, try to turn it out.</dd>
+<dt>Q. : What does `QDBM' mean?</dt>
+<dd>A. : `QDBM' stands for `Quick Database Manager'. It means that processing speed is high, and that you can write applications quickly.</dd>
+</dl>
+
+<hr />
+
+<h2><a name="copying" id="copying" class="head">Copying</a></h2>
+
+<p>QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License or any later version.</p>
+
+<p>QDBM is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.</p>
+
+<p>You should have received a copy of the GNU Lesser General Public License along with QDBM (See the file `COPYING'); if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.</p>
+
+<p>QDBM was written by Mikio Hirabayashi. You can contact the author by e-mail to `mikio at users.sourceforge.net'. However, as for topics which can be shared among other users, please send it to the mailing list. To join the mailing list, refer to `http://lists.sourceforge.net/lists/listinfo/qdbm-users'.</p>
+
+<hr />
+
+</body>
+
+</html>
+
+<!-- END OF FILE -->
Added: box/trunk/qdbm/villa.c
===================================================================
--- box/trunk/qdbm/villa.c (rev 0)
+++ box/trunk/qdbm/villa.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,2666 @@
+/*************************************************************************************************
+ * Implementation of Villa
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+
+#include "villa.h"
+#include "myconf.h"
+
+#define VL_LEAFIDMIN 1 /* minimum number of leaf ID */
+#define VL_NODEIDMIN 100000000 /* minimum number of node ID */
+#define VL_VNUMBUFSIZ 8 /* size of a buffer for variable length number */
+#define VL_NUMBUFSIZ 32 /* size of a buffer for a number */
+#define VL_PAGEBUFSIZ 32768 /* size of a buffer to read each page */
+#define VL_MAXLEAFSIZ 49152 /* maximum size of each leaf */
+#define VL_DEFLRECMAX 49 /* default number of records in each leaf */
+#define VL_DEFNIDXMAX 192 /* default number of indexes in each node */
+#define VL_DEFLCNUM 1024 /* default number of leaf cache */
+#define VL_DEFNCNUM 512 /* default number of node cache */
+#define VL_CACHEOUT 8 /* number of pages in a process of cacheout */
+#define VL_INITBNUM 32749 /* initial bucket number */
+#define VL_PAGEALIGN -3 /* alignment for pages */
+#define VL_FBPOOLSIZ 128 /* size of free block pool */
+#define VL_PATHBUFSIZ 1024 /* size of a path buffer */
+#define VL_TMPFSUF MYEXTSTR "vltmp" /* suffix of a temporary file */
+#define VL_ROOTKEY -1 /* key of the root key */
+#define VL_LASTKEY -2 /* key of the last key */
+#define VL_LNUMKEY -3 /* key of the number of leaves */
+#define VL_NNUMKEY -4 /* key of the number of nodes */
+#define VL_RNUMKEY -5 /* key of the number of records */
+#define VL_CRDNUM 7 /* default division number for Vista */
+
+/* set a buffer for a variable length number */
+#define VL_SETVNUMBUF(VL_len, VL_buf, VL_num) \
+ do { \
+ int _VL_num; \
+ _VL_num = VL_num; \
+ if(_VL_num == 0){ \
+ ((signed char *)(VL_buf))[0] = 0; \
+ (VL_len) = 1; \
+ } else { \
+ (VL_len) = 0; \
+ while(_VL_num > 0){ \
+ int _VL_rem = _VL_num & 0x7f; \
+ _VL_num >>= 7; \
+ if(_VL_num > 0){ \
+ ((signed char *)(VL_buf))[(VL_len)] = -_VL_rem - 1; \
+ } else { \
+ ((signed char *)(VL_buf))[(VL_len)] = _VL_rem; \
+ } \
+ (VL_len)++; \
+ } \
+ } \
+ } while(FALSE)
+
+/* read a variable length buffer */
+#define VL_READVNUMBUF(VL_buf, VL_size, VL_num, VL_step) \
+ do { \
+ int _VL_i, _VL_base; \
+ (VL_num) = 0; \
+ _VL_base = 1; \
+ if((VL_size) < 2){ \
+ (VL_num) = ((signed char *)(VL_buf))[0]; \
+ (VL_step) = 1; \
+ } else { \
+ for(_VL_i = 0; _VL_i < (VL_size); _VL_i++){ \
+ if(((signed char *)(VL_buf))[_VL_i] >= 0){ \
+ (VL_num) += ((signed char *)(VL_buf))[_VL_i] * _VL_base; \
+ break; \
+ } \
+ (VL_num) += _VL_base * (((signed char *)(VL_buf))[_VL_i] + 1) * -1; \
+ _VL_base *= 128; \
+ } \
+ (VL_step) = _VL_i + 1; \
+ } \
+ } while(FALSE)
+
+enum { /* enumeration for flags */
+ VL_FLISVILLA = 1 << 0, /* whether for Villa */
+ VL_FLISZLIB = 1 << 1, /* whether with ZLIB */
+ VL_FLISLZO = 1 << 2, /* whether with LZO */
+ VL_FLISBZIP = 1 << 3 /* whether with BZIP2 */
+};
+
+
+/* private function prototypes */
+static int vllexcompare(const char *aptr, int asiz, const char *bptr, int bsiz);
+static int vlintcompare(const char *aptr, int asiz, const char *bptr, int bsiz);
+static int vlnumcompare(const char *aptr, int asiz, const char *bptr, int bsiz);
+static int vldeccompare(const char *aptr, int asiz, const char *bptr, int bsiz);
+static int vldpputnum(DEPOT *depot, int knum, int vnum);
+static int vldpgetnum(DEPOT *depot, int knum, int *vnp);
+static VLLEAF *vlleafnew(VILLA *villa, int prev, int next);
+static int vlleafcacheout(VILLA *villa, int id);
+static int vlleafsave(VILLA *villa, VLLEAF *leaf);
+static VLLEAF *vlleafload(VILLA *villa, int id);
+static VLLEAF *vlgethistleaf(VILLA *villa, const char *kbuf, int ksiz);
+static int vlleafaddrec(VILLA *villa, VLLEAF *leaf, int dmode,
+ const char *kbuf, int ksiz, const char *vbuf, int vsiz);
+static int vlleafdatasize(VLLEAF *leaf);
+static VLLEAF *vlleafdivide(VILLA *villa, VLLEAF *leaf);
+static VLNODE *vlnodenew(VILLA *villa, int heir);
+static int vlnodecacheout(VILLA *villa, int id);
+static int vlnodesave(VILLA *villa, VLNODE *node);
+static VLNODE *vlnodeload(VILLA *villa, int id);
+static void vlnodeaddidx(VILLA *villa, VLNODE *node, int order,
+ int pid, const char *kbuf, int ksiz);
+static int vlsearchleaf(VILLA *villa, const char *kbuf, int ksiz);
+static int vlcacheadjust(VILLA *villa);
+static VLREC *vlrecsearch(VILLA *villa, VLLEAF *leaf, const char *kbuf, int ksiz, int *ip);
+
+
+
+/*************************************************************************************************
+ * public objects
+ *************************************************************************************************/
+
+
+/* Comparing functions. */
+VLCFUNC VL_CMPLEX = vllexcompare;
+VLCFUNC VL_CMPINT = vlintcompare;
+VLCFUNC VL_CMPNUM = vlnumcompare;
+VLCFUNC VL_CMPDEC = vldeccompare;
+
+
+/* Get a database handle. */
+VILLA *vlopen(const char *name, int omode, VLCFUNC cmp){
+ DEPOT *depot;
+ int dpomode, flags, cmode, root, last, lnum, nnum, rnum;
+ VILLA *villa;
+ VLLEAF *leaf;
+ assert(name && cmp);
+ dpomode = DP_OREADER;
+ if(omode & VL_OWRITER){
+ dpomode = DP_OWRITER;
+ if(omode & VL_OCREAT) dpomode |= DP_OCREAT;
+ if(omode & VL_OTRUNC) dpomode |= DP_OTRUNC;
+ }
+ if(omode & VL_ONOLCK) dpomode |= DP_ONOLCK;
+ if(omode & VL_OLCKNB) dpomode |= DP_OLCKNB;
+ if(!(depot = dpopen(name, dpomode, VL_INITBNUM))) return NULL;
+ flags = dpgetflags(depot);
+ cmode = 0;
+ root = -1;
+ last = -1;
+ lnum = 0;
+ nnum = 0;
+ rnum = 0;
+ if(dprnum(depot) > 0){
+ if(!(flags & VL_FLISVILLA) ||
+ !vldpgetnum(depot, VL_ROOTKEY, &root) || !vldpgetnum(depot, VL_LASTKEY, &last) ||
+ !vldpgetnum(depot, VL_LNUMKEY, &lnum) || !vldpgetnum(depot, VL_NNUMKEY, &nnum) ||
+ !vldpgetnum(depot, VL_RNUMKEY, &rnum) || root < VL_LEAFIDMIN || last < VL_LEAFIDMIN ||
+ lnum < 0 || nnum < 0 || rnum < 0){
+ dpclose(depot);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(flags & VL_FLISZLIB){
+ cmode = VL_OZCOMP;
+ } else if(flags & VL_FLISLZO){
+ cmode = VL_OYCOMP;
+ } else if(flags & VL_FLISBZIP){
+ cmode = VL_OXCOMP;
+ }
+ } else if(omode & VL_OWRITER){
+ if(omode & VL_OZCOMP){
+ cmode = VL_OZCOMP;
+ } else if(omode & VL_OYCOMP){
+ cmode = VL_OYCOMP;
+ } else if(omode & VL_OXCOMP){
+ cmode = VL_OXCOMP;
+ }
+ }
+ if(omode & VL_OWRITER){
+ flags |= VL_FLISVILLA;
+ if(_qdbm_deflate && cmode == VL_OZCOMP){
+ flags |= VL_FLISZLIB;
+ } else if(_qdbm_lzoencode && cmode == VL_OYCOMP){
+ flags |= VL_FLISLZO;
+ } else if(_qdbm_bzencode && cmode == VL_OXCOMP){
+ flags |= VL_FLISBZIP;
+ }
+ if(!dpsetflags(depot, flags) || !dpsetalign(depot, VL_PAGEALIGN) ||
+ !dpsetfbpsiz(depot, VL_FBPOOLSIZ)){
+ dpclose(depot);
+ return NULL;
+ }
+ }
+ CB_MALLOC(villa, sizeof(VILLA));
+ villa->depot = depot;
+ villa->cmp = cmp;
+ villa->wmode = (omode & VL_OWRITER);
+ villa->cmode = cmode;
+ villa->root = root;
+ villa->last = last;
+ villa->lnum = lnum;
+ villa->nnum = nnum;
+ villa->rnum = rnum;
+ villa->leafc = cbmapopen();
+ villa->nodec = cbmapopen();
+ villa->hnum = 0;
+ villa->hleaf = -1;
+ villa->lleaf = -1;
+ villa->curleaf = -1;
+ villa->curknum = -1;
+ villa->curvnum = -1;
+ villa->leafrecmax = VL_DEFLRECMAX;
+ villa->nodeidxmax = VL_DEFNIDXMAX;
+ villa->leafcnum = VL_DEFLCNUM;
+ villa->nodecnum = VL_DEFNCNUM;
+ villa->tran = FALSE;
+ villa->rbroot = -1;
+ villa->rblast = -1;
+ villa->rblnum = -1;
+ villa->rbnnum = -1;
+ villa->rbrnum = -1;
+ if(root == -1){
+ leaf = vlleafnew(villa, -1, -1);
+ villa->root = leaf->id;
+ villa->last = leaf->id;
+ if(!vltranbegin(villa) || !vltranabort(villa)){
+ vlclose(villa);
+ return NULL;
+ }
+ }
+ return villa;
+}
+
+
+/* Close a database handle. */
+int vlclose(VILLA *villa){
+ int err, pid;
+ const char *tmp;
+ assert(villa);
+ err = FALSE;
+ if(villa->tran){
+ if(!vltranabort(villa)) err = TRUE;
+ }
+ cbmapiterinit(villa->leafc);
+ while((tmp = cbmapiternext(villa->leafc, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!vlleafcacheout(villa, pid)) err = TRUE;
+ }
+ cbmapiterinit(villa->nodec);
+ while((tmp = cbmapiternext(villa->nodec, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!vlnodecacheout(villa, pid)) err = TRUE;
+ }
+ if(villa->wmode){
+ if(!dpsetalign(villa->depot, 0)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_ROOTKEY, villa->root)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LASTKEY, villa->last)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LNUMKEY, villa->lnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_NNUMKEY, villa->nnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_RNUMKEY, villa->rnum)) err = TRUE;
+ }
+ cbmapclose(villa->leafc);
+ cbmapclose(villa->nodec);
+ if(!dpclose(villa->depot)) err = TRUE;
+ free(villa);
+ return err ? FALSE : TRUE;
+}
+
+
+/* Store a record. */
+int vlput(VILLA *villa, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode){
+ VLLEAF *leaf, *newleaf;
+ VLNODE *node, *newnode;
+ VLIDX *idxp;
+ CBDATUM *key;
+ int i, pid, todiv, heir, parent, mid;
+ assert(villa && kbuf && vbuf);
+ villa->curleaf = -1;
+ villa->curknum = -1;
+ villa->curvnum = -1;
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(vsiz < 0) vsiz = strlen(vbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return FALSE;
+ if(!(leaf = vlleafload(villa, pid))) return FALSE;
+ }
+ if(!vlleafaddrec(villa, leaf, dmode, kbuf, ksiz, vbuf, vsiz)){
+ dpecodeset(DP_EKEEP, __FILE__, __LINE__);
+ return FALSE;
+ }
+ todiv = FALSE;
+ switch(CB_LISTNUM(leaf->recs) % 4){
+ case 0:
+ if(CB_LISTNUM(leaf->recs) >= 4 &&
+ vlleafdatasize(leaf) > VL_MAXLEAFSIZ * (villa->cmode > 0 ? 2 : 1)){
+ todiv = TRUE;
+ break;
+ }
+ case 2:
+ if(CB_LISTNUM(leaf->recs) > villa->leafrecmax) todiv = TRUE;
+ break;
+ }
+ if(todiv){
+ if(!(newleaf = vlleafdivide(villa, leaf))) return FALSE;
+ if(leaf->id == villa->last) villa->last = newleaf->id;
+ heir = leaf->id;
+ pid = newleaf->id;
+ key = ((VLREC *)CB_LISTVAL(newleaf->recs, 0))->key;
+ key = cbdatumdup(key);
+ while(TRUE){
+ if(villa->hnum < 1){
+ node = vlnodenew(villa, heir);
+ vlnodeaddidx(villa, node, TRUE, pid, CB_DATUMPTR(key), CB_DATUMSIZE(key));
+ villa->root = node->id;
+ CB_DATUMCLOSE(key);
+ break;
+ }
+ parent = villa->hist[--villa->hnum];
+ if(!(node = vlnodeload(villa, parent))){
+ CB_DATUMCLOSE(key);
+ return FALSE;
+ }
+ vlnodeaddidx(villa, node, FALSE, pid, CB_DATUMPTR(key), CB_DATUMSIZE(key));
+ CB_DATUMCLOSE(key);
+ if(CB_LISTNUM(node->idxs) <= villa->nodeidxmax) break;
+ mid = CB_LISTNUM(node->idxs) / 2;
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, mid);
+ newnode = vlnodenew(villa, idxp->pid);
+ heir = node->id;
+ pid = newnode->id;
+ CB_DATUMOPEN2(key, CB_DATUMPTR(idxp->key), CB_DATUMSIZE(idxp->key));
+ for(i = mid + 1; i < CB_LISTNUM(node->idxs); i++){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ vlnodeaddidx(villa, newnode, TRUE, idxp->pid,
+ CB_DATUMPTR(idxp->key), CB_DATUMSIZE(idxp->key));
+ }
+ for(i = 0; i < CB_LISTNUM(newnode->idxs); i++){
+ idxp = (VLIDX *)cblistpop(node->idxs, NULL);
+ CB_DATUMCLOSE(idxp->key);
+ free(idxp);
+ }
+ node->dirty = TRUE;
+ }
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return FALSE;
+ return TRUE;
+}
+
+
+/* Delete a record. */
+int vlout(VILLA *villa, const char *kbuf, int ksiz){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid, ri, vsiz;
+ char *vbuf;
+ assert(villa && kbuf);
+ villa->curleaf = -1;
+ villa->curknum = -1;
+ villa->curvnum = -1;
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return FALSE;
+ if(!(leaf = vlleafload(villa, pid))) return FALSE;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, &ri))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(recp->rest){
+ CB_DATUMCLOSE(recp->first);
+ vbuf = cblistshift(recp->rest, &vsiz);
+ CB_DATUMOPEN2(recp->first, vbuf, vsiz);
+ free(vbuf);
+ if(CB_LISTNUM(recp->rest) < 1){
+ CB_LISTCLOSE(recp->rest);
+ recp->rest = NULL;
+ }
+ } else {
+ CB_DATUMCLOSE(recp->key);
+ CB_DATUMCLOSE(recp->first);
+ free(cblistremove(leaf->recs, ri, NULL));
+ }
+ leaf->dirty = TRUE;
+ villa->rnum--;
+ if(!villa->tran && !vlcacheadjust(villa)) return FALSE;
+ return TRUE;
+}
+
+
+/* Retrieve a record. */
+char *vlget(VILLA *villa, const char *kbuf, int ksiz, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ char *rv;
+ int pid;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return NULL;
+ if(!(leaf = vlleafload(villa, pid))) return NULL;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, NULL))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return NULL;
+ if(sp) *sp = CB_DATUMSIZE(recp->first);
+ CB_MEMDUP(rv, CB_DATUMPTR(recp->first), CB_DATUMSIZE(recp->first));
+ return rv;
+}
+
+
+/* Get the size of the value of a record. */
+int vlvsiz(VILLA *villa, const char *kbuf, int ksiz){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return -1;
+ if(!(leaf = vlleafload(villa, pid))) return -1;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, NULL))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return -1;
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return -1;
+ return CB_DATUMSIZE(recp->first);
+}
+
+
+/* Get the number of records corresponding a key. */
+int vlvnum(VILLA *villa, const char *kbuf, int ksiz){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return 0;
+ if(!(leaf = vlleafload(villa, pid))) return 0;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, NULL))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return 0;
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return 0;
+ return 1 + (recp->rest ? CB_LISTNUM(recp->rest) : 0);
+}
+
+
+/* Store plural records corresponding a key. */
+int vlputlist(VILLA *villa, const char *kbuf, int ksiz, const CBLIST *vals){
+ int i, vsiz;
+ const char *vbuf;
+ assert(villa && kbuf && vals);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(CB_LISTNUM(vals) < 1){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ for(i = 0; i < CB_LISTNUM(vals); i++){
+ vbuf = CB_LISTVAL2(vals, i, vsiz);
+ if(!vlput(villa, kbuf, ksiz, vbuf, vsiz, VL_DDUP)) return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Delete all records corresponding a key. */
+int vloutlist(VILLA *villa, const char *kbuf, int ksiz){
+ int i, vnum;
+ assert(villa && kbuf);
+ if(!villa->wmode){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if((vnum = vlvnum(villa, kbuf, ksiz)) < 1) return FALSE;
+ for(i = 0; i < vnum; i++){
+ if(!vlout(villa, kbuf, ksiz)) return FALSE;
+ }
+ return TRUE;
+}
+
+
+/* Retrieve values of all records corresponding a key. */
+CBLIST *vlgetlist(VILLA *villa, const char *kbuf, int ksiz){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid, i, vsiz;
+ CBLIST *vals;
+ const char *vbuf;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return NULL;
+ if(!(leaf = vlleafload(villa, pid))) return NULL;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, NULL))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ CB_LISTOPEN(vals);
+ CB_LISTPUSH(vals, CB_DATUMPTR(recp->first), CB_DATUMSIZE(recp->first));
+ if(recp->rest){
+ for(i = 0; i < CB_LISTNUM(recp->rest); i++){
+ vbuf = CB_LISTVAL2(recp->rest, i, vsiz);
+ CB_LISTPUSH(vals, vbuf, vsiz);
+ }
+ }
+ if(!villa->tran && !vlcacheadjust(villa)){
+ CB_LISTCLOSE(vals);
+ return NULL;
+ }
+ return vals;
+}
+
+
+/* Retrieve concatenated values of all records corresponding a key. */
+char *vlgetcat(VILLA *villa, const char *kbuf, int ksiz, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid, i, vsiz, rsiz;
+ char *rbuf;
+ const char *vbuf;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return NULL;
+ if(!(leaf = vlleafload(villa, pid))) return NULL;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, NULL))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ rsiz = CB_DATUMSIZE(recp->first);
+ CB_MALLOC(rbuf, rsiz + 1);
+ memcpy(rbuf, CB_DATUMPTR(recp->first), rsiz);
+ if(recp->rest){
+ for(i = 0; i < CB_LISTNUM(recp->rest); i++){
+ vbuf = CB_LISTVAL2(recp->rest, i, vsiz);
+ CB_REALLOC(rbuf, rsiz + vsiz + 1);
+ memcpy(rbuf + rsiz, vbuf, vsiz);
+ rsiz += vsiz;
+ }
+ }
+ rbuf[rsiz] = '\0';
+ if(!villa->tran && !vlcacheadjust(villa)){
+ free(rbuf);
+ return NULL;
+ }
+ if(sp) *sp = rsiz;
+ return rbuf;
+}
+
+
+/* Move the cursor to the first record. */
+int vlcurfirst(VILLA *villa){
+ VLLEAF *leaf;
+ assert(villa);
+ villa->curleaf = VL_LEAFIDMIN;
+ villa->curknum = 0;
+ villa->curvnum = 0;
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ while(CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = leaf->next;
+ villa->curknum = 0;
+ villa->curvnum = 0;
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ }
+ return TRUE;
+}
+
+
+/* Move the cursor to the last record. */
+int vlcurlast(VILLA *villa){
+ VLLEAF *leaf;
+ VLREC *recp;
+ assert(villa);
+ villa->curleaf = villa->last;
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ while(CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = leaf->prev;
+ if(villa->curleaf == -1){
+ villa->curleaf = -1;
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ }
+ villa->curknum = CB_LISTNUM(leaf->recs) - 1;
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ villa->curvnum = recp->rest ? CB_LISTNUM(recp->rest) : 0;
+ return TRUE;
+}
+
+
+/* Move the cursor to the previous record. */
+int vlcurprev(VILLA *villa){
+ VLLEAF *leaf;
+ VLREC *recp;
+ assert(villa);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf)) || CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ villa->curvnum--;
+ if(villa->curvnum < 0){
+ villa->curknum--;
+ if(villa->curknum < 0){
+ villa->curleaf = leaf->prev;
+ if(villa->curleaf == -1){
+ villa->curleaf = -1;
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ while(CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = leaf->prev;
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ }
+ villa->curknum = CB_LISTNUM(leaf->recs) - 1;
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ villa->curvnum = recp->rest ? CB_LISTNUM(recp->rest) : 0;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ villa->curvnum = recp->rest ? CB_LISTNUM(recp->rest) : 0;
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return FALSE;
+ return TRUE;
+}
+
+
+/* Move the cursor to the next record. */
+int vlcurnext(VILLA *villa){
+ VLLEAF *leaf;
+ VLREC *recp;
+ assert(villa);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf)) || CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ villa->curvnum++;
+ if(villa->curvnum > (recp->rest ? CB_LISTNUM(recp->rest) : 0)){
+ villa->curknum++;
+ villa->curvnum = 0;
+ }
+ if(villa->curknum >= CB_LISTNUM(leaf->recs)){
+ villa->curleaf = leaf->next;
+ villa->curknum = 0;
+ villa->curvnum = 0;
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ while(CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = leaf->next;
+ villa->curknum = 0;
+ villa->curvnum = 0;
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ }
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return FALSE;
+ return TRUE;
+}
+
+
+/* Move the cursor to a position around a record. */
+int vlcurjump(VILLA *villa, const char *kbuf, int ksiz, int jmode){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid, index;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, pid))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ while(CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = (jmode == VL_JFORWARD) ? leaf->next : leaf->prev;
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, &index))){
+ if(jmode == VL_JFORWARD){
+ villa->curleaf = leaf->id;
+ if(index >= CB_LISTNUM(leaf->recs)) index--;
+ villa->curknum = index;
+ villa->curvnum = 0;
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, index);
+ if(villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key)) < 0) return TRUE;
+ villa->curvnum = (recp->rest ? CB_LISTNUM(recp->rest) : 0);
+ return vlcurnext(villa);
+ } else {
+ villa->curleaf = leaf->id;
+ if(index >= CB_LISTNUM(leaf->recs)) index--;
+ villa->curknum = index;
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, index);
+ villa->curvnum = (recp->rest ? CB_LISTNUM(recp->rest) : 0);
+ if(villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key)) > 0) return TRUE;
+ villa->curvnum = 0;
+ return vlcurprev(villa);
+ }
+ }
+ if(jmode == VL_JFORWARD){
+ villa->curleaf = pid;
+ villa->curknum = index;
+ villa->curvnum = 0;
+ } else {
+ villa->curleaf = pid;
+ villa->curknum = index;
+ villa->curvnum = (recp->rest ? CB_LISTNUM(recp->rest) : 0);
+ }
+ return TRUE;
+}
+
+
+/* Get the key of the record where the cursor is. */
+char *vlcurkey(VILLA *villa, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ const char *kbuf;
+ char *rv;
+ int ksiz;
+ assert(villa);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ kbuf = CB_DATUMPTR(recp->key);
+ ksiz = CB_DATUMSIZE(recp->key);
+ if(sp) *sp = ksiz;
+ CB_MEMDUP(rv, kbuf, ksiz);
+ return rv;
+}
+
+
+/* Get the value of the record where the cursor is. */
+char *vlcurval(VILLA *villa, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ const char *vbuf;
+ char *rv;
+ int vsiz;
+ assert(villa);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ if(villa->curvnum < 1){
+ vbuf = CB_DATUMPTR(recp->first);
+ vsiz = CB_DATUMSIZE(recp->first);
+ } else {
+ vbuf = CB_LISTVAL2(recp->rest, villa->curvnum - 1, vsiz);
+ }
+ if(sp) *sp = vsiz;
+ CB_MEMDUP(rv, vbuf, vsiz);
+ return rv;
+}
+
+
+/* Insert a record around the cursor. */
+int vlcurput(VILLA *villa, const char *vbuf, int vsiz, int cpmode){
+ VLLEAF *leaf;
+ VLREC *recp;
+ char *tbuf;
+ int tsiz;
+ assert(villa && vbuf);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(vsiz < 0) vsiz = strlen(vbuf);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ switch(cpmode){
+ case VL_CPBEFORE:
+ if(villa->curvnum < 1){
+ if(!recp->rest){
+ CB_DATUMTOMALLOC(recp->first, tbuf, tsiz);
+ CB_DATUMOPEN2(recp->first, vbuf, vsiz);
+ CB_LISTOPEN(recp->rest);
+ CB_LISTPUSHBUF(recp->rest, tbuf, tsiz);
+ } else {
+ cblistunshift(recp->rest, CB_DATUMPTR(recp->first), CB_DATUMSIZE(recp->first));
+ CB_DATUMSETSIZE(recp->first, 0);
+ CB_DATUMCAT(recp->first, vbuf, vsiz);
+ }
+ } else {
+ CB_LISTINSERT(recp->rest, villa->curvnum - 1, vbuf, vsiz);
+ }
+ villa->rnum++;
+ break;
+ case VL_CPAFTER:
+ if(!recp->rest) CB_LISTOPEN(recp->rest);
+ CB_LISTINSERT(recp->rest, villa->curvnum, vbuf, vsiz);
+ villa->curvnum++;
+ villa->rnum++;
+ break;
+ default:
+ if(villa->curvnum < 1){
+ CB_DATUMSETSIZE(recp->first, 0);
+ CB_DATUMCAT(recp->first, vbuf, vsiz);
+ } else {
+ cblistover(recp->rest, villa->curvnum - 1, vbuf, vsiz);
+ }
+ break;
+ }
+ leaf->dirty = TRUE;
+ return TRUE;
+}
+
+
+/* Delete the record where the cursor is. */
+int vlcurout(VILLA *villa){
+ VLLEAF *leaf;
+ VLREC *recp;
+ char *vbuf;
+ int vsiz;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ if(villa->curvnum < 1){
+ if(recp->rest){
+ vbuf = cblistshift(recp->rest, &vsiz);
+ CB_DATUMSETSIZE(recp->first, 0);
+ CB_DATUMCAT(recp->first, vbuf, vsiz);
+ free(vbuf);
+ if(CB_LISTNUM(recp->rest) < 1){
+ CB_LISTCLOSE(recp->rest);
+ recp->rest = NULL;
+ }
+ } else {
+ CB_DATUMCLOSE(recp->first);
+ CB_DATUMCLOSE(recp->key);
+ free(cblistremove(leaf->recs, villa->curknum, NULL));
+ }
+ } else {
+ free(cblistremove(recp->rest, villa->curvnum - 1, NULL));
+ if(villa->curvnum - 1 >= CB_LISTNUM(recp->rest)){
+ villa->curknum++;
+ villa->curvnum = 0;
+ }
+ if(CB_LISTNUM(recp->rest) < 1){
+ CB_LISTCLOSE(recp->rest);
+ recp->rest = NULL;
+ }
+ }
+ villa->rnum--;
+ leaf->dirty = TRUE;
+ if(villa->curknum >= CB_LISTNUM(leaf->recs)){
+ villa->curleaf = leaf->next;
+ villa->curknum = 0;
+ villa->curvnum = 0;
+ while(villa->curleaf != -1 && (leaf = vlleafload(villa, villa->curleaf)) != NULL &&
+ CB_LISTNUM(leaf->recs) < 1){
+ villa->curleaf = leaf->next;
+ }
+ }
+ return TRUE;
+}
+
+
+/* Set the tuning parameters for performance. */
+void vlsettuning(VILLA *villa, int lrecmax, int nidxmax, int lcnum, int ncnum){
+ assert(villa);
+ if(lrecmax < 1) lrecmax = VL_DEFLRECMAX;
+ if(lrecmax < 3) lrecmax = 3;
+ if(nidxmax < 1) nidxmax = VL_DEFNIDXMAX;
+ if(nidxmax < 4) nidxmax = 4;
+ if(lcnum < 1) lcnum = VL_DEFLCNUM;
+ if(lcnum < VL_CACHEOUT * 2) lcnum = VL_CACHEOUT * 2;
+ if(ncnum < 1) ncnum = VL_DEFNCNUM;
+ if(ncnum < VL_CACHEOUT * 2) ncnum = VL_CACHEOUT * 2;
+ villa->leafrecmax = lrecmax;
+ villa->nodeidxmax = nidxmax;
+ villa->leafcnum = lcnum;
+ villa->nodecnum = ncnum;
+}
+
+
+/* Set the size of the free block pool of a database handle. */
+int vlsetfbpsiz(VILLA *villa, int size){
+ assert(villa && size >= 0);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return dpsetfbpsiz(villa->depot, size);
+}
+
+
+/* Synchronize updating contents with the file and the device. */
+int vlsync(VILLA *villa){
+ int err;
+ err = FALSE;
+ if(!vlmemsync(villa)) err = TRUE;
+ if(!dpsync(villa->depot)) err = TRUE;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Optimize a database. */
+int vloptimize(VILLA *villa){
+ int err;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(villa->tran){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ if(!vlsync(villa)) return FALSE;
+ if(!dpoptimize(villa->depot, -1)) err = TRUE;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Get the name of a database. */
+char *vlname(VILLA *villa){
+ assert(villa);
+ return dpname(villa->depot);
+}
+
+
+/* Get the size of a database file. */
+int vlfsiz(VILLA *villa){
+ return dpfsiz(villa->depot);
+}
+
+
+/* Get the number of the leaf nodes of B+ tree. */
+int vllnum(VILLA *villa){
+ assert(villa);
+ return villa->lnum;
+}
+
+
+/* Get the number of the non-leaf nodes of B+ tree. */
+int vlnnum(VILLA *villa){
+ assert(villa);
+ return villa->nnum;
+}
+
+
+/* Get the number of the records stored in a database. */
+int vlrnum(VILLA *villa){
+ assert(villa);
+ return villa->rnum;
+}
+
+
+/* Check whether a database handle is a writer or not. */
+int vlwritable(VILLA *villa){
+ assert(villa);
+ return villa->wmode;
+}
+
+
+/* Check whether a database has a fatal error or not. */
+int vlfatalerror(VILLA *villa){
+ assert(villa);
+ return dpfatalerror(villa->depot);
+}
+
+
+/* Get the inode number of a database file. */
+int vlinode(VILLA *villa){
+ assert(villa);
+ return dpinode(villa->depot);
+}
+
+
+/* Get the last modified time of a database. */
+time_t vlmtime(VILLA *villa){
+ assert(villa);
+ return dpmtime(villa->depot);
+}
+
+
+/* Begin the transaction. */
+int vltranbegin(VILLA *villa){
+ int err, pid;
+ const char *tmp;
+ VLLEAF *leaf;
+ VLNODE *node;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(villa->tran){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ cbmapiterinit(villa->leafc);
+ while((tmp = cbmapiternext(villa->leafc, NULL)) != NULL){
+ pid = *(int *)tmp;
+ leaf = (VLLEAF *)cbmapget(villa->leafc, (char *)&pid, sizeof(int), NULL);
+ if(leaf->dirty && !vlleafsave(villa, leaf)) err = TRUE;
+ }
+ cbmapiterinit(villa->nodec);
+ while((tmp = cbmapiternext(villa->nodec, NULL)) != NULL){
+ pid = *(int *)tmp;
+ node = (VLNODE *)cbmapget(villa->nodec, (char *)&pid, sizeof(int), NULL);
+ if(node->dirty && !vlnodesave(villa, node)) err = TRUE;
+ }
+ if(!dpsetalign(villa->depot, 0)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_ROOTKEY, villa->root)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LASTKEY, villa->last)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LNUMKEY, villa->lnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_NNUMKEY, villa->nnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_RNUMKEY, villa->rnum)) err = TRUE;
+ if(!dpmemsync(villa->depot)) err = TRUE;
+ if(!dpsetalign(villa->depot, VL_PAGEALIGN)) err = TRUE;
+ villa->tran = TRUE;
+ villa->rbroot = villa->root;
+ villa->rblast = villa->last;
+ villa->rblnum = villa->lnum;
+ villa->rbnnum = villa->nnum;
+ villa->rbrnum = villa->rnum;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Commit the transaction. */
+int vltrancommit(VILLA *villa){
+ int err, pid;
+ const char *tmp;
+ VLLEAF *leaf;
+ VLNODE *node;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!villa->tran){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ cbmapiterinit(villa->leafc);
+ while((tmp = cbmapiternext(villa->leafc, NULL)) != NULL){
+ pid = *(int *)tmp;
+ leaf = (VLLEAF *)cbmapget(villa->leafc, (char *)&pid, sizeof(int), NULL);
+ if(leaf->dirty && !vlleafsave(villa, leaf)) err = TRUE;
+ }
+ cbmapiterinit(villa->nodec);
+ while((tmp = cbmapiternext(villa->nodec, NULL)) != NULL){
+ pid = *(int *)tmp;
+ node = (VLNODE *)cbmapget(villa->nodec, (char *)&pid, sizeof(int), NULL);
+ if(node->dirty && !vlnodesave(villa, node)) err = TRUE;
+ }
+ if(!dpsetalign(villa->depot, 0)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_ROOTKEY, villa->root)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LASTKEY, villa->last)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LNUMKEY, villa->lnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_NNUMKEY, villa->nnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_RNUMKEY, villa->rnum)) err = TRUE;
+ if(!dpmemsync(villa->depot)) err = TRUE;
+ if(!dpsetalign(villa->depot, VL_PAGEALIGN)) err = TRUE;
+ villa->tran = FALSE;
+ villa->rbroot = -1;
+ villa->rblast = -1;
+ villa->rblnum = -1;
+ villa->rbnnum = -1;
+ villa->rbrnum = -1;
+ while(cbmaprnum(villa->leafc) > villa->leafcnum || cbmaprnum(villa->nodec) > villa->nodecnum){
+ if(!vlcacheadjust(villa)){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Abort the transaction. */
+int vltranabort(VILLA *villa){
+ int err, pid;
+ const char *tmp;
+ VLLEAF *leaf;
+ VLNODE *node;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!villa->tran){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ cbmapiterinit(villa->leafc);
+ while((tmp = cbmapiternext(villa->leafc, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!(leaf = (VLLEAF *)cbmapget(villa->leafc, (char *)&pid, sizeof(int), NULL))){
+ err = TRUE;
+ continue;
+ }
+ if(leaf->dirty){
+ leaf->dirty = FALSE;
+ if(!vlleafcacheout(villa, pid)) err = TRUE;
+ }
+ }
+ cbmapiterinit(villa->nodec);
+ while((tmp = cbmapiternext(villa->nodec, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!(node = (VLNODE *)cbmapget(villa->nodec, (char *)&pid, sizeof(int), NULL))){
+ err = TRUE;
+ continue;
+ }
+ if(node->dirty){
+ node->dirty = FALSE;
+ if(!vlnodecacheout(villa, pid)) err = TRUE;
+ }
+ }
+ villa->tran = FALSE;
+ villa->root = villa->rbroot;
+ villa->last = villa->rblast;
+ villa->lnum = villa->rblnum;
+ villa->nnum = villa->rbnnum;
+ villa->rnum = villa->rbrnum;
+ while(cbmaprnum(villa->leafc) > villa->leafcnum || cbmaprnum(villa->nodec) > villa->nodecnum){
+ if(!vlcacheadjust(villa)){
+ err = TRUE;
+ break;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Remove a database file. */
+int vlremove(const char *name){
+ assert(name);
+ return dpremove(name);
+}
+
+
+/* Repair a broken database file. */
+int vlrepair(const char *name, VLCFUNC cmp){
+ DEPOT *depot;
+ VILLA *tvilla;
+ char path[VL_PATHBUFSIZ], *kbuf, *vbuf, *zbuf, *rp, *tkbuf, *tvbuf;
+ int i, err, flags, omode, ksiz, vsiz, zsiz, size, step, tksiz, tvsiz, vnum;
+ assert(name && cmp);
+ err = FALSE;
+ if(!dprepair(name)) err = TRUE;
+ if(!(depot = dpopen(name, DP_OREADER, -1))) return FALSE;
+ flags = dpgetflags(depot);
+ if(!(flags & VL_FLISVILLA)){
+ dpclose(depot);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ sprintf(path, "%s%s", name, VL_TMPFSUF);
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC;
+ if(flags & VL_FLISZLIB){
+ omode |= VL_OZCOMP;
+ } else if(flags & VL_FLISLZO){
+ omode |= VL_OXCOMP;
+ } else if(flags & VL_FLISBZIP){
+ omode |= VL_OYCOMP;
+ }
+ if(!(tvilla = vlopen(path, omode, cmp))){
+ dpclose(depot);
+ return FALSE;
+ }
+ if(!dpiterinit(depot)) err = TRUE;
+ while((kbuf = dpiternext(depot, &ksiz)) != NULL){
+ if(ksiz == sizeof(int) && *(int *)kbuf < VL_NODEIDMIN && *(int *)kbuf > 0){
+ if((vbuf = dpget(depot, (char *)kbuf, sizeof(int), 0, -1, &vsiz)) != NULL){
+ if(_qdbm_inflate && (flags & VL_FLISZLIB) &&
+ (zbuf = _qdbm_inflate(vbuf, vsiz, &zsiz, _QDBM_ZMRAW)) != NULL){
+ free(vbuf);
+ vbuf = zbuf;
+ vsiz = zsiz;
+ } else if(_qdbm_lzodecode && (flags & VL_FLISLZO) &&
+ (zbuf = _qdbm_lzodecode(vbuf, vsiz, &zsiz)) != NULL){
+ free(vbuf);
+ vbuf = zbuf;
+ vsiz = zsiz;
+ } else if(_qdbm_bzdecode && (flags & VL_FLISBZIP) &&
+ (zbuf = _qdbm_bzdecode(vbuf, vsiz, &zsiz)) != NULL){
+ free(vbuf);
+ vbuf = zbuf;
+ vsiz = zsiz;
+ }
+ rp = vbuf;
+ size = vsiz;
+ if(size >= 1){
+ VL_READVNUMBUF(rp, size, vnum, step);
+ rp += step;
+ size -= step;
+ }
+ if(size >= 1){
+ VL_READVNUMBUF(rp, size, vnum, step);
+ rp += step;
+ size -= step;
+ }
+ while(size >= 1){
+ VL_READVNUMBUF(rp, size, tksiz, step);
+ rp += step;
+ size -= step;
+ if(size < tksiz) break;
+ tkbuf = rp;
+ rp += tksiz;
+ size -= tksiz;
+ if(size < 1) break;
+ VL_READVNUMBUF(rp, size, vnum, step);
+ rp += step;
+ size -= step;
+ if(vnum < 1 || size < 1) break;
+ for(i = 0; i < vnum && size >= 1; i++){
+ VL_READVNUMBUF(rp, size, tvsiz, step);
+ rp += step;
+ size -= step;
+ if(size < tvsiz) break;
+ tvbuf = rp;
+ rp += tvsiz;
+ size -= tvsiz;
+ if(!vlput(tvilla, tkbuf, tksiz, tvbuf, tvsiz, VL_DDUP)) err = TRUE;
+ }
+ }
+ free(vbuf);
+ }
+ }
+ free(kbuf);
+ }
+ if(!vlclose(tvilla)) err = TRUE;
+ if(!dpclose(depot)) err = TRUE;
+ if(!dpremove(name)) err = TRUE;
+ if(rename(path, name) == -1){
+ if(!err) dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Dump all records as endian independent data. */
+int vlexportdb(VILLA *villa, const char *name){
+ DEPOT *depot;
+ char path[VL_PATHBUFSIZ], *kbuf, *vbuf, *nkey;
+ int i, err, ksiz, vsiz, ki;
+ assert(villa && name);
+ sprintf(path, "%s%s", name, VL_TMPFSUF);
+ if(!(depot = dpopen(path, DP_OWRITER | DP_OCREAT | DP_OTRUNC, -1))) return FALSE;
+ err = FALSE;
+ vlcurfirst(villa);
+ for(i = 0; !err && (kbuf = vlcurkey(villa, &ksiz)) != NULL; i++){
+ if((vbuf = vlcurval(villa, &vsiz)) != NULL){
+ CB_MALLOC(nkey, ksiz + VL_NUMBUFSIZ);
+ ki = sprintf(nkey, "%X\t", i);
+ memcpy(nkey + ki, kbuf, ksiz);
+ if(!dpput(depot, nkey, ki + ksiz, vbuf, vsiz, DP_DKEEP)) err = TRUE;
+ free(nkey);
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ vlcurnext(villa);
+ }
+ if(!dpexportdb(depot, name)) err = TRUE;
+ if(!dpclose(depot)) err = TRUE;
+ if(!dpremove(path)) err = TRUE;
+ return !err && !vlfatalerror(villa);
+}
+
+
+/* Load all records from endian independent data. */
+int vlimportdb(VILLA *villa, const char *name){
+ DEPOT *depot;
+ char path[VL_PATHBUFSIZ], *kbuf, *vbuf, *rp;
+ int err, ksiz, vsiz;
+ assert(villa && name);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(vlrnum(villa) > 0){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ kbuf = dpname(villa->depot);
+ sprintf(path, "%s%s", kbuf, VL_TMPFSUF);
+ free(kbuf);
+ if(!(depot = dpopen(path, DP_OWRITER | DP_OCREAT | DP_OTRUNC, -1))) return FALSE;
+ err = FALSE;
+ if(!dpimportdb(depot, name)) err = TRUE;
+ dpiterinit(depot);
+ while(!err && (kbuf = dpiternext(depot, &ksiz)) != NULL){
+ if((vbuf = dpget(depot, kbuf, ksiz, 0, -1, &vsiz)) != NULL){
+ if((rp = strchr(kbuf, '\t')) != NULL){
+ rp++;
+ if(!vlput(villa, rp, ksiz - (rp - kbuf), vbuf, vsiz, VL_DDUP)) err = TRUE;
+ } else {
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ err = TRUE;
+ }
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ }
+ if(!dpclose(depot)) err = TRUE;
+ if(!dpremove(path)) err = TRUE;
+ return !err && !vlfatalerror(villa);
+}
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Number of division of the database for Vista. */
+int *vlcrdnumptr(void){
+ static int defvlcrdnum = VL_CRDNUM;
+ void *ptr;
+ if(_qdbm_ptsafe){
+ if(!(ptr = _qdbm_settsd(&defvlcrdnum, sizeof(int), &defvlcrdnum))){
+ defvlcrdnum = DP_EMISC;
+ return &defvlcrdnum;
+ }
+ return (int *)ptr;
+ }
+ return &defvlcrdnum;
+}
+
+
+/* Synchronize updating contents on memory. */
+int vlmemsync(VILLA *villa){
+ int err, pid;
+ const char *tmp;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(villa->tran){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ cbmapiterinit(villa->leafc);
+ while((tmp = cbmapiternext(villa->leafc, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!vlleafcacheout(villa, pid)) err = TRUE;
+ }
+ cbmapiterinit(villa->nodec);
+ while((tmp = cbmapiternext(villa->nodec, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!vlnodecacheout(villa, pid)) err = TRUE;
+ }
+ if(!dpsetalign(villa->depot, 0)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_ROOTKEY, villa->root)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LASTKEY, villa->last)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LNUMKEY, villa->lnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_NNUMKEY, villa->nnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_RNUMKEY, villa->rnum)) err = TRUE;
+ if(!dpsetalign(villa->depot, VL_PAGEALIGN)) err = TRUE;
+ if(!dpmemsync(villa->depot)) err = TRUE;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Synchronize updating contents on memory, not physically. */
+int vlmemflush(VILLA *villa){
+ int err, pid;
+ const char *tmp;
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(villa->tran){
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ err = FALSE;
+ cbmapiterinit(villa->leafc);
+ while((tmp = cbmapiternext(villa->leafc, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!vlleafcacheout(villa, pid)) err = TRUE;
+ }
+ cbmapiterinit(villa->nodec);
+ while((tmp = cbmapiternext(villa->nodec, NULL)) != NULL){
+ pid = *(int *)tmp;
+ if(!vlnodecacheout(villa, pid)) err = TRUE;
+ }
+ if(!dpsetalign(villa->depot, 0)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_ROOTKEY, villa->root)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LASTKEY, villa->last)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_LNUMKEY, villa->lnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_NNUMKEY, villa->nnum)) err = TRUE;
+ if(!vldpputnum(villa->depot, VL_RNUMKEY, villa->rnum)) err = TRUE;
+ if(!dpsetalign(villa->depot, VL_PAGEALIGN)) err = TRUE;
+ if(!dpmemflush(villa->depot)) err = TRUE;
+ return err ? FALSE : TRUE;
+}
+
+
+/* Refer to a volatile cache of a value of a record. */
+const char *vlgetcache(VILLA *villa, const char *kbuf, int ksiz, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int pid;
+ assert(villa && kbuf);
+ if(ksiz < 0) ksiz = strlen(kbuf);
+ if(villa->hleaf < VL_LEAFIDMIN || !(leaf = vlgethistleaf(villa, kbuf, ksiz))){
+ if((pid = vlsearchleaf(villa, kbuf, ksiz)) == -1) return NULL;
+ if(!(leaf = vlleafload(villa, pid))) return NULL;
+ }
+ if(!(recp = vlrecsearch(villa, leaf, kbuf, ksiz, NULL))){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(!villa->tran && !vlcacheadjust(villa)) return NULL;
+ if(sp) *sp = CB_DATUMSIZE(recp->first);
+ return CB_DATUMPTR(recp->first);
+}
+
+
+/* Refer to volatile cache of the key of the record where the cursor is. */
+const char *vlcurkeycache(VILLA *villa, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ const char *kbuf;
+ int ksiz;
+ assert(villa);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ kbuf = CB_DATUMPTR(recp->key);
+ ksiz = CB_DATUMSIZE(recp->key);
+ if(sp) *sp = ksiz;
+ return kbuf;
+}
+
+
+/* Refer to volatile cache of the value of the record where the cursor is. */
+const char *vlcurvalcache(VILLA *villa, int *sp){
+ VLLEAF *leaf;
+ VLREC *recp;
+ const char *vbuf;
+ int vsiz;
+ assert(villa);
+ if(villa->curleaf == -1){
+ dpecodeset(DP_ENOITEM, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!(leaf = vlleafload(villa, villa->curleaf))){
+ villa->curleaf = -1;
+ return FALSE;
+ }
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, villa->curknum);
+ if(villa->curvnum < 1){
+ vbuf = CB_DATUMPTR(recp->first);
+ vsiz = CB_DATUMSIZE(recp->first);
+ } else {
+ vbuf = CB_LISTVAL2(recp->rest, villa->curvnum - 1, vsiz);
+ }
+ if(sp) *sp = vsiz;
+ return vbuf;
+}
+
+
+/* Get a multiple cursor handle. */
+VLMULCUR *vlmulcuropen(VILLA *villa){
+ VLMULCUR *mulcur;
+ assert(villa);
+ if(villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return NULL;
+ }
+ CB_MALLOC(mulcur, sizeof(VLMULCUR));
+ mulcur->villa = villa;
+ mulcur->curleaf = -1;
+ mulcur->curknum = -1;
+ mulcur->curvnum = -1;
+ return mulcur;
+}
+
+
+/* Close a multiple cursor handle. */
+void vlmulcurclose(VLMULCUR *mulcur){
+ assert(mulcur);
+ free(mulcur);
+}
+
+
+/* Move a multiple cursor to the first record. */
+int vlmulcurfirst(VLMULCUR *mulcur){
+ VLMULCUR swap;
+ int rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurfirst(mulcur->villa);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Move a multiple cursor to the last record. */
+int vlmulcurlast(VLMULCUR *mulcur){
+ VLMULCUR swap;
+ int rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurlast(mulcur->villa);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Move a multiple cursor to the previous record. */
+int vlmulcurprev(VLMULCUR *mulcur){
+ VLMULCUR swap;
+ int rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurprev(mulcur->villa);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Move a multiple cursor to the next record. */
+int vlmulcurnext(VLMULCUR *mulcur){
+ VLMULCUR swap;
+ int rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurnext(mulcur->villa);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Move a multiple cursor to a position around a record. */
+int vlmulcurjump(VLMULCUR *mulcur, const char *kbuf, int ksiz, int jmode){
+ VLMULCUR swap;
+ int rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurjump(mulcur->villa, kbuf, ksiz, jmode);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Get the key of the record where a multiple cursor is. */
+char *vlmulcurkey(VLMULCUR *mulcur, int *sp){
+ VLMULCUR swap;
+ char *rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurkey(mulcur->villa, sp);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Get the value of the record where a multiple cursor is. */
+char *vlmulcurval(VLMULCUR *mulcur, int *sp){
+ VLMULCUR swap;
+ char *rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurval(mulcur->villa, sp);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Refer to volatile cache of the key of the record where a multiple cursor is. */
+const char *vlmulcurkeycache(VLMULCUR *mulcur, int *sp){
+ VLMULCUR swap;
+ const char *rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurkeycache(mulcur->villa, sp);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+/* Refer to volatile cache of the value of the record where a multiple cursor is. */
+const char *vlmulcurvalcache(VLMULCUR *mulcur, int *sp){
+ VLMULCUR swap;
+ const char *rv;
+ assert(mulcur);
+ swap.curleaf = mulcur->villa->curleaf;
+ swap.curknum = mulcur->villa->curknum;
+ swap.curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = mulcur->curleaf;
+ mulcur->villa->curknum = mulcur->curknum;
+ mulcur->villa->curvnum = mulcur->curvnum;
+ rv = vlcurvalcache(mulcur->villa, sp);
+ mulcur->curleaf = mulcur->villa->curleaf;
+ mulcur->curknum = mulcur->villa->curknum;
+ mulcur->curvnum = mulcur->villa->curvnum;
+ mulcur->villa->curleaf = swap.curleaf;
+ mulcur->villa->curknum = swap.curknum;
+ mulcur->villa->curvnum = swap.curvnum;
+ return rv;
+}
+
+
+
+/*************************************************************************************************
+ * private objects
+ *************************************************************************************************/
+
+
+/* Compare keys of two records by lexical order.
+ `aptr' specifies the pointer to the region of one key.
+ `asiz' specifies the size of the region of one key.
+ `bptr' specifies the pointer to the region of the other key.
+ `bsiz' specifies the size of the region of the other key.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+static int vllexcompare(const char *aptr, int asiz, const char *bptr, int bsiz){
+ int i, min;
+ assert(aptr && asiz >= 0 && bptr && bsiz >= 0);
+ min = asiz < bsiz ? asiz : bsiz;
+ for(i = 0; i < min; i++){
+ if(((unsigned char *)aptr)[i] != ((unsigned char *)bptr)[i])
+ return ((unsigned char *)aptr)[i] - ((unsigned char *)bptr)[i];
+ }
+ if(asiz == bsiz) return 0;
+ return asiz - bsiz;
+}
+
+
+/* Compare keys of two records as native integers.
+ `aptr' specifies the pointer to the region of one key.
+ `asiz' specifies the size of the region of one key.
+ `bptr' specifies the pointer to the region of the other key.
+ `bsiz' specifies the size of the region of the other key.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+static int vlintcompare(const char *aptr, int asiz, const char *bptr, int bsiz){
+ int anum, bnum;
+ assert(aptr && asiz >= 0 && bptr && bsiz >= 0);
+ if(asiz != bsiz) return asiz - bsiz;
+ anum = (asiz == sizeof(int) ? *(int *)aptr : INT_MIN);
+ bnum = (bsiz == sizeof(int) ? *(int *)bptr : INT_MIN);
+ return anum - bnum;
+}
+
+
+/* Compare keys of two records as numbers of big endian.
+ `aptr' specifies the pointer to the region of one key.
+ `asiz' specifies the size of the region of one key.
+ `bptr' specifies the pointer to the region of the other key.
+ `bsiz' specifies the size of the region of the other key.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+static int vlnumcompare(const char *aptr, int asiz, const char *bptr, int bsiz){
+ int i;
+ assert(aptr && asiz >= 0 && bptr && bsiz >= 0);
+ if(asiz != bsiz) return asiz - bsiz;
+ for(i = 0; i < asiz; i++){
+ if(aptr[i] != bptr[i]) return aptr[i] - bptr[i];
+ }
+ return 0;
+}
+
+
+/* Compare keys of two records as numeric strings of octal, decimal or hexadecimal.
+ `aptr' specifies the pointer to the region of one key.
+ `asiz' specifies the size of the region of one key.
+ `bptr' specifies the pointer to the region of the other key.
+ `bsiz' specifies the size of the region of the other key.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+static int vldeccompare(const char *aptr, int asiz, const char *bptr, int bsiz){
+ assert(aptr && asiz >= 0 && bptr && bsiz >= 0);
+ return (int)(strtod(aptr, NULL) - strtod(bptr, NULL));
+}
+
+
+/* Store a record composed of a pair of integers.
+ `depot' specifies an internal database handle.
+ `knum' specifies an integer of the key.
+ `vnum' specifies an integer of the value.
+ The return value is true if successful, else, it is false. */
+static int vldpputnum(DEPOT *depot, int knum, int vnum){
+ assert(depot);
+ return dpput(depot, (char *)&knum, sizeof(int), (char *)&vnum, sizeof(int), DP_DOVER);
+}
+
+
+/* Retrieve a record composed of a pair of integers.
+ `depot' specifies an internal database handle.
+ `knum' specifies an integer of the key.
+ `vip' specifies the pointer to a variable to assign the result to.
+ The return value is true if successful, else, it is false. */
+static int vldpgetnum(DEPOT *depot, int knum, int *vnp){
+ char *vbuf;
+ int vsiz;
+ assert(depot && vnp);
+ vbuf = dpget(depot, (char *)&knum, sizeof(int), 0, -1, &vsiz);
+ if(!vbuf || vsiz != sizeof(int)){
+ free(vbuf);
+ return FALSE;
+ }
+ *vnp = *(int *)vbuf;
+ free(vbuf);
+ return TRUE;
+}
+
+
+/* Create a new leaf.
+ `villa' specifies a database handle.
+ `prev' specifies the ID number of the previous leaf.
+ `next' specifies the ID number of the previous leaf.
+ The return value is a handle of the leaf. */
+static VLLEAF *vlleafnew(VILLA *villa, int prev, int next){
+ VLLEAF lent;
+ assert(villa);
+ lent.id = villa->lnum + VL_LEAFIDMIN;
+ lent.dirty = TRUE;
+ CB_LISTOPEN(lent.recs);
+ lent.prev = prev;
+ lent.next = next;
+ villa->lnum++;
+ cbmapput(villa->leafc, (char *)&(lent.id), sizeof(int), (char *)&lent, sizeof(VLLEAF), TRUE);
+ return (VLLEAF *)cbmapget(villa->leafc, (char *)&(lent.id), sizeof(int), NULL);
+}
+
+
+/* Remove a leaf from the cache.
+ `villa' specifies a database handle.
+ `id' specifies the ID number of the leaf.
+ The return value is true if successful, else, it is false. */
+static int vlleafcacheout(VILLA *villa, int id){
+ VLLEAF *leaf;
+ VLREC *recp;
+ CBLIST *recs;
+ int i, err, ln;
+ assert(villa && id >= VL_LEAFIDMIN);
+ if(!(leaf = (VLLEAF *)cbmapget(villa->leafc, (char *)&id, sizeof(int), NULL))) return FALSE;
+ err = FALSE;
+ if(leaf->dirty && !vlleafsave(villa, leaf)) err = TRUE;
+ recs = leaf->recs;
+ ln = CB_LISTNUM(recs);
+ for(i = 0; i < ln; i++){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ CB_DATUMCLOSE(recp->key);
+ CB_DATUMCLOSE(recp->first);
+ if(recp->rest) CB_LISTCLOSE(recp->rest);
+ }
+ CB_LISTCLOSE(recs);
+ cbmapout(villa->leafc, (char *)&id, sizeof(int));
+ return err ? FALSE : TRUE;
+}
+
+
+/* Save a leaf into the database.
+ `villa' specifies a database handle.
+ `leaf' specifies a leaf handle.
+ The return value is true if successful, else, it is false. */
+static int vlleafsave(VILLA *villa, VLLEAF *leaf){
+ VLREC *recp;
+ CBLIST *recs;
+ CBDATUM *buf;
+ char vnumbuf[VL_VNUMBUFSIZ], *zbuf;
+ const char *vbuf;
+ int i, j, ksiz, vnum, vsiz, prev, next, vnumsiz, ln, zsiz;
+ assert(villa && leaf);
+ CB_DATUMOPEN(buf);
+ prev = leaf->prev;
+ if(prev == -1) prev = VL_NODEIDMIN - 1;
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, prev);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ next = leaf->next;
+ if(next == -1) next = VL_NODEIDMIN - 1;
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, next);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ recs = leaf->recs;
+ ln = CB_LISTNUM(recs);
+ for(i = 0; i < ln; i++){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ ksiz = CB_DATUMSIZE(recp->key);
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, ksiz);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ CB_DATUMCAT(buf, CB_DATUMPTR(recp->key), ksiz);
+ vnum = 1 + (recp->rest ? CB_LISTNUM(recp->rest) : 0);
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, vnum);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ vsiz = CB_DATUMSIZE(recp->first);
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, vsiz);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ CB_DATUMCAT(buf, CB_DATUMPTR(recp->first), vsiz);
+ if(recp->rest){
+ for(j = 0; j < CB_LISTNUM(recp->rest); j++){
+ vbuf = CB_LISTVAL2(recp->rest, j, vsiz);
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, vsiz);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ CB_DATUMCAT(buf, vbuf, vsiz);
+ }
+ }
+ }
+ if(_qdbm_deflate && villa->cmode == VL_OZCOMP){
+ if(!(zbuf = _qdbm_deflate(CB_DATUMPTR(buf), CB_DATUMSIZE(buf), &zsiz, _QDBM_ZMRAW))){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!dpput(villa->depot, (char *)&(leaf->id), sizeof(int), zbuf, zsiz, DP_DOVER)){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ free(zbuf);
+ } else if(_qdbm_lzoencode && villa->cmode == VL_OYCOMP){
+ if(!(zbuf = _qdbm_lzoencode(CB_DATUMPTR(buf), CB_DATUMSIZE(buf), &zsiz))){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!dpput(villa->depot, (char *)&(leaf->id), sizeof(int), zbuf, zsiz, DP_DOVER)){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ free(zbuf);
+ } else if(_qdbm_bzencode && villa->cmode == VL_OXCOMP){
+ if(!(zbuf = _qdbm_bzencode(CB_DATUMPTR(buf), CB_DATUMSIZE(buf), &zsiz))){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EMISC, __FILE__, __LINE__);
+ return FALSE;
+ }
+ if(!dpput(villa->depot, (char *)&(leaf->id), sizeof(int), zbuf, zsiz, DP_DOVER)){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ free(zbuf);
+ } else {
+ if(!dpput(villa->depot, (char *)&(leaf->id), sizeof(int),
+ CB_DATUMPTR(buf), CB_DATUMSIZE(buf), DP_DOVER)){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ }
+ CB_DATUMCLOSE(buf);
+ leaf->dirty = FALSE;
+ return TRUE;
+}
+
+
+/* Load a leaf from the database.
+ `villa' specifies a database handle.
+ `id' specifies the ID number of the leaf.
+ If successful, the return value is the pointer to the leaf, else, it is `NULL'. */
+static VLLEAF *vlleafload(VILLA *villa, int id){
+ char wbuf[VL_PAGEBUFSIZ], *buf, *rp, *kbuf, *vbuf, *zbuf;
+ int i, size, step, ksiz, vnum, vsiz, prev, next, zsiz;
+ VLLEAF *leaf, lent;
+ VLREC rec;
+ assert(villa && id >= VL_LEAFIDMIN);
+ if((leaf = (VLLEAF *)cbmapget(villa->leafc, (char *)&id, sizeof(int), NULL)) != NULL){
+ cbmapmove(villa->leafc, (char *)&id, sizeof(int), FALSE);
+ return leaf;
+ }
+ ksiz = -1;
+ prev = -1;
+ next = -1;
+ if((size = dpgetwb(villa->depot, (char *)&id, sizeof(int), 0, VL_PAGEBUFSIZ, wbuf)) > 0 &&
+ size < VL_PAGEBUFSIZ){
+ buf = NULL;
+ } else if(!(buf = dpget(villa->depot, (char *)&id, sizeof(int), 0, -1, &size))){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ if(_qdbm_inflate && villa->cmode == VL_OZCOMP){
+ if(!(zbuf = _qdbm_inflate(buf ? buf : wbuf, size, &zsiz, _QDBM_ZMRAW))){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ free(buf);
+ return NULL;
+ }
+ free(buf);
+ buf = zbuf;
+ size = zsiz;
+ } else if(_qdbm_lzodecode && villa->cmode == VL_OYCOMP){
+ if(!(zbuf = _qdbm_lzodecode(buf ? buf : wbuf, size, &zsiz))){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ free(buf);
+ return NULL;
+ }
+ free(buf);
+ buf = zbuf;
+ size = zsiz;
+ } else if(_qdbm_bzdecode && villa->cmode == VL_OXCOMP){
+ if(!(zbuf = _qdbm_bzdecode(buf ? buf : wbuf, size, &zsiz))){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ free(buf);
+ return NULL;
+ }
+ free(buf);
+ buf = zbuf;
+ size = zsiz;
+ }
+ rp = buf ? buf : wbuf;
+ if(size >= 1){
+ VL_READVNUMBUF(rp, size, prev, step);
+ rp += step;
+ size -= step;
+ if(prev >= VL_NODEIDMIN - 1) prev = -1;
+ }
+ if(size >= 1){
+ VL_READVNUMBUF(rp, size, next, step);
+ rp += step;
+ size -= step;
+ if(next >= VL_NODEIDMIN - 1) next = -1;
+ }
+ lent.id = id;
+ lent.dirty = FALSE;
+ CB_LISTOPEN(lent.recs);
+ lent.prev = prev;
+ lent.next = next;
+ while(size >= 1){
+ VL_READVNUMBUF(rp, size, ksiz, step);
+ rp += step;
+ size -= step;
+ if(size < ksiz) break;
+ kbuf = rp;
+ rp += ksiz;
+ size -= ksiz;
+ VL_READVNUMBUF(rp, size, vnum, step);
+ rp += step;
+ size -= step;
+ if(vnum < 1 || size < 1) break;
+ for(i = 0; i < vnum && size >= 1; i++){
+ VL_READVNUMBUF(rp, size, vsiz, step);
+ rp += step;
+ size -= step;
+ if(size < vsiz) break;
+ vbuf = rp;
+ rp += vsiz;
+ size -= vsiz;
+ if(i < 1){
+ CB_DATUMOPEN2(rec.key, kbuf, ksiz);
+ CB_DATUMOPEN2(rec.first, vbuf, vsiz);
+ rec.rest = NULL;
+ } else {
+ if(!rec.rest) CB_LISTOPEN(rec.rest);
+ CB_LISTPUSH(rec.rest, vbuf, vsiz);
+ }
+ }
+ if(i > 0) CB_LISTPUSH(lent.recs, (char *)&rec, sizeof(VLREC));
+ }
+ free(buf);
+ cbmapput(villa->leafc, (char *)&(lent.id), sizeof(int), (char *)&lent, sizeof(VLLEAF), TRUE);
+ return (VLLEAF *)cbmapget(villa->leafc, (char *)&(lent.id), sizeof(int), NULL);
+}
+
+
+/* Load the historical leaf from the database.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key.
+ If successful, the return value is the pointer to the leaf, else, it is `NULL'. */
+static VLLEAF *vlgethistleaf(VILLA *villa, const char *kbuf, int ksiz){
+ VLLEAF *leaf;
+ VLREC *recp;
+ int ln, rv;
+ assert(villa && kbuf && ksiz >= 0);
+ if(!(leaf = vlleafload(villa, villa->hleaf))) return NULL;
+ if((ln = CB_LISTNUM(leaf->recs)) < 2) return NULL;
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, 0);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key));
+ if(rv == 0) return leaf;
+ if(rv < 0) return NULL;
+ recp = (VLREC *)CB_LISTVAL(leaf->recs, ln - 1);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key));
+ if(rv <= 0 || leaf->next < VL_LEAFIDMIN) return leaf;
+ return NULL;
+}
+
+
+/* Add a record to a leaf.
+ `villa' specifies a database handle.
+ `leaf' specifies a leaf handle.
+ `dmode' specifies behavior when the key overlaps.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value.
+ The return value is true if successful, else, it is false. */
+static int vlleafaddrec(VILLA *villa, VLLEAF *leaf, int dmode,
+ const char *kbuf, int ksiz, const char *vbuf, int vsiz){
+ VLREC *recp, rec;
+ CBLIST *recs;
+ int i, rv, left, right, ln, tsiz;
+ char *tbuf;
+ assert(villa && leaf && kbuf && ksiz >= 0 && vbuf && vsiz >= 0);
+ left = 0;
+ recs = leaf->recs;
+ ln = CB_LISTNUM(recs);
+ right = ln;
+ i = (left + right) / 2;
+ while(right >= left && i < ln){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key));
+ if(rv == 0){
+ break;
+ } else if(rv <= 0){
+ right = i - 1;
+ } else {
+ left = i + 1;
+ }
+ i = (left + right) / 2;
+ }
+ while(i < ln){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key));
+ if(rv == 0){
+ switch(dmode){
+ case VL_DKEEP:
+ return FALSE;
+ case VL_DCAT:
+ CB_DATUMCAT(recp->first, vbuf, vsiz);
+ break;
+ case VL_DDUP:
+ if(!recp->rest) CB_LISTOPEN(recp->rest);
+ CB_LISTPUSH(recp->rest, vbuf, vsiz);
+ villa->rnum++;
+ break;
+ case VL_DDUPR:
+ if(!recp->rest){
+ CB_DATUMTOMALLOC(recp->first, tbuf, tsiz);
+ CB_DATUMOPEN2(recp->first, vbuf, vsiz);
+ CB_LISTOPEN(recp->rest);
+ CB_LISTPUSHBUF(recp->rest, tbuf, tsiz);
+ } else {
+ cblistunshift(recp->rest, CB_DATUMPTR(recp->first), CB_DATUMSIZE(recp->first));
+ CB_DATUMSETSIZE(recp->first, 0);
+ CB_DATUMCAT(recp->first, vbuf, vsiz);
+ }
+ villa->rnum++;
+ break;
+ default:
+ CB_DATUMSETSIZE(recp->first, 0);
+ CB_DATUMCAT(recp->first, vbuf, vsiz);
+ break;
+ }
+ break;
+ } else if(rv < 0){
+ CB_DATUMOPEN2(rec.key, kbuf, ksiz);
+ CB_DATUMOPEN2(rec.first, vbuf, vsiz);
+ rec.rest = NULL;
+ CB_LISTINSERT(recs, i, (char *)&rec, sizeof(VLREC));
+ villa->rnum++;
+ break;
+ }
+ i++;
+ }
+ if(i >= ln){
+ CB_DATUMOPEN2(rec.key, kbuf, ksiz);
+ CB_DATUMOPEN2(rec.first, vbuf, vsiz);
+ rec.rest = NULL;
+ CB_LISTPUSH(recs, (char *)&rec, sizeof(VLREC));
+ villa->rnum++;
+ }
+ leaf->dirty = TRUE;
+ return TRUE;
+}
+
+
+/* Calculate the size of data of a leaf.
+ `leaf' specifies a leaf handle.
+ The return value is size of data of the leaf. */
+static int vlleafdatasize(VLLEAF *leaf){
+ VLREC *recp;
+ CBLIST *recs, *rest;
+ const char *vbuf;
+ int i, j, sum, rnum, restnum, vsiz;
+ assert(leaf);
+ sum = 0;
+ recs = leaf->recs;
+ rnum = CB_LISTNUM(recs);
+ for(i = 0; i < rnum; i++){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ sum += CB_DATUMSIZE(recp->key);
+ sum += CB_DATUMSIZE(recp->first);
+ if(recp->rest){
+ rest = recp->rest;
+ restnum = CB_LISTNUM(rest);
+ for(j = 0; j < restnum; j++){
+ vbuf = CB_LISTVAL2(rest, j, vsiz);
+ sum += vsiz;
+ }
+ }
+ }
+ return sum;
+}
+
+
+/* Divide a leaf into two.
+ `villa' specifies a database handle.
+ `leaf' specifies a leaf handle.
+ The return value is the handle of a new leaf, or `NULL' on failure. */
+static VLLEAF *vlleafdivide(VILLA *villa, VLLEAF *leaf){
+ VLLEAF *newleaf, *nextleaf;
+ VLREC *recp;
+ CBLIST *recs, *newrecs;
+ int i, mid, ln;
+ assert(villa && leaf);
+ villa->hleaf = -1;
+ recs = leaf->recs;
+ mid = CB_LISTNUM(recs) / 2;
+ recp = (VLREC *)CB_LISTVAL(recs, mid);
+ newleaf = vlleafnew(villa, leaf->id, leaf->next);
+ if(newleaf->next != -1){
+ if(!(nextleaf = vlleafload(villa, newleaf->next))) return NULL;
+ nextleaf->prev = newleaf->id;
+ nextleaf->dirty = TRUE;
+ }
+ leaf->next = newleaf->id;
+ leaf->dirty = TRUE;
+ ln = CB_LISTNUM(recs);
+ newrecs = newleaf->recs;
+ for(i = mid; i < ln; i++){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ CB_LISTPUSH(newrecs, (char *)recp, sizeof(VLREC));
+ }
+ ln = CB_LISTNUM(newrecs);
+ for(i = 0; i < ln; i++){
+ CB_LISTDROP(recs);
+ }
+ return newleaf;
+}
+
+
+/* Create a new node.
+ `villa' specifies a database handle.
+ `heir' specifies the ID of the child before the first index.
+ The return value is a handle of the node. */
+static VLNODE *vlnodenew(VILLA *villa, int heir){
+ VLNODE nent;
+ assert(villa && heir >= VL_LEAFIDMIN);
+ nent.id = villa->nnum + VL_NODEIDMIN;
+ nent.dirty = TRUE;
+ nent.heir = heir;
+ CB_LISTOPEN(nent.idxs);
+ villa->nnum++;
+ cbmapput(villa->nodec, (char *)&(nent.id), sizeof(int), (char *)&nent, sizeof(VLNODE), TRUE);
+ return (VLNODE *)cbmapget(villa->nodec, (char *)&(nent.id), sizeof(int), NULL);
+}
+
+
+/* Remove a node from the cache.
+ `villa' specifies a database handle.
+ `id' specifies the ID number of the node.
+ The return value is true if successful, else, it is false. */
+static int vlnodecacheout(VILLA *villa, int id){
+ VLNODE *node;
+ VLIDX *idxp;
+ int i, err, ln;
+ assert(villa && id >= VL_NODEIDMIN);
+ if(!(node = (VLNODE *)cbmapget(villa->nodec, (char *)&id, sizeof(int), NULL))) return FALSE;
+ err = FALSE;
+ if(node->dirty && !vlnodesave(villa, node)) err = TRUE;
+ ln = CB_LISTNUM(node->idxs);
+ for(i = 0; i < ln; i++){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ CB_DATUMCLOSE(idxp->key);
+ }
+ CB_LISTCLOSE(node->idxs);
+ cbmapout(villa->nodec, (char *)&id, sizeof(int));
+ return err ? FALSE : TRUE;
+}
+
+
+/* Save a node into the database.
+ `villa' specifies a database handle.
+ `node' specifies a node handle.
+ The return value is true if successful, else, it is false. */
+static int vlnodesave(VILLA *villa, VLNODE *node){
+ CBDATUM *buf;
+ char vnumbuf[VL_VNUMBUFSIZ];
+ VLIDX *idxp;
+ int i, heir, pid, ksiz, vnumsiz, ln;
+ assert(villa && node);
+ CB_DATUMOPEN(buf);
+ heir = node->heir;
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, heir);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ ln = CB_LISTNUM(node->idxs);
+ for(i = 0; i < ln; i++){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ pid = idxp->pid;
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, pid);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ ksiz = CB_DATUMSIZE(idxp->key);
+ VL_SETVNUMBUF(vnumsiz, vnumbuf, ksiz);
+ CB_DATUMCAT(buf, vnumbuf, vnumsiz);
+ CB_DATUMCAT(buf, CB_DATUMPTR(idxp->key), ksiz);
+ }
+ if(!dpput(villa->depot, (char *)&(node->id), sizeof(int),
+ CB_DATUMPTR(buf), CB_DATUMSIZE(buf), DP_DOVER)){
+ CB_DATUMCLOSE(buf);
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return FALSE;
+ }
+ CB_DATUMCLOSE(buf);
+ node->dirty = FALSE;
+ return TRUE;
+}
+
+
+/* Load a node from the database.
+ `villa' specifies a database handle.
+ `id' specifies the ID number of the node.
+ If successful, the return value is the pointer to the node, else, it is `NULL'. */
+static VLNODE *vlnodeload(VILLA *villa, int id){
+ char wbuf[VL_PAGEBUFSIZ], *buf, *rp, *kbuf;
+ int size, step, heir, pid, ksiz;
+ VLNODE *node, nent;
+ VLIDX idx;
+ assert(villa && id >= VL_NODEIDMIN);
+ if((node = (VLNODE *)cbmapget(villa->nodec, (char *)&id, sizeof(int), NULL)) != NULL){
+ cbmapmove(villa->nodec, (char *)&id, sizeof(int), FALSE);
+ return node;
+ }
+ heir = -1;
+ if((size = dpgetwb(villa->depot, (char *)&id, sizeof(int), 0, VL_PAGEBUFSIZ, wbuf)) > 0 &&
+ size < VL_PAGEBUFSIZ){
+ buf = NULL;
+ } else if(!(buf = dpget(villa->depot, (char *)&id, sizeof(int), 0, -1, &size))){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return NULL;
+ }
+ rp = buf ? buf : wbuf;
+ if(size >= 1){
+ VL_READVNUMBUF(rp, size, heir, step);
+ rp += step;
+ size -= step;
+ }
+ if(heir < 0){
+ free(buf);
+ return NULL;
+ }
+ nent.id = id;
+ nent.dirty = FALSE;
+ nent.heir = heir;
+ CB_LISTOPEN(nent.idxs);
+ while(size >= 1){
+ VL_READVNUMBUF(rp, size, pid, step);
+ rp += step;
+ size -= step;
+ if(size < 1) break;
+ VL_READVNUMBUF(rp, size, ksiz, step);
+ rp += step;
+ size -= step;
+ if(size < ksiz) break;
+ kbuf = rp;
+ rp += ksiz;
+ size -= ksiz;
+ idx.pid = pid;
+ CB_DATUMOPEN2(idx.key, kbuf, ksiz);
+ CB_LISTPUSH(nent.idxs, (char *)&idx, sizeof(VLIDX));
+ }
+ free(buf);
+ cbmapput(villa->nodec, (char *)&(nent.id), sizeof(int), (char *)&nent, sizeof(VLNODE), TRUE);
+ return (VLNODE *)cbmapget(villa->nodec, (char *)&(nent.id), sizeof(int), NULL);
+}
+
+
+/* Add an index to a node.
+ `villa' specifies a database handle.
+ `node' specifies a node handle.
+ `order' specifies whether the calling sequence is orderd or not.
+ `pid' specifies the ID number of referred page.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. */
+static void vlnodeaddidx(VILLA *villa, VLNODE *node, int order,
+ int pid, const char *kbuf, int ksiz){
+ VLIDX idx, *idxp;
+ int i, rv, left, right, ln;
+ assert(villa && node && pid >= VL_LEAFIDMIN && kbuf && ksiz >= 0);
+ idx.pid = pid;
+ CB_DATUMOPEN2(idx.key, kbuf, ksiz);
+ if(order){
+ CB_LISTPUSH(node->idxs, (char *)&idx, sizeof(VLIDX));
+ } else {
+ left = 0;
+ right = CB_LISTNUM(node->idxs);
+ i = (left + right) / 2;
+ ln = CB_LISTNUM(node->idxs);
+ while(right >= left && i < ln){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(idxp->key), CB_DATUMSIZE(idxp->key));
+ if(rv == 0){
+ break;
+ } else if(rv <= 0){
+ right = i - 1;
+ } else {
+ left = i + 1;
+ }
+ i = (left + right) / 2;
+ }
+ ln = CB_LISTNUM(node->idxs);
+ while(i < ln){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ if(villa->cmp(kbuf, ksiz, CB_DATUMPTR(idxp->key), CB_DATUMSIZE(idxp->key)) < 0){
+ CB_LISTINSERT(node->idxs, i, (char *)&idx, sizeof(VLIDX));
+ break;
+ }
+ i++;
+ }
+ if(i >= CB_LISTNUM(node->idxs)) CB_LISTPUSH(node->idxs, (char *)&idx, sizeof(VLIDX));
+ }
+ node->dirty = TRUE;
+}
+
+
+/* Search the leaf corresponding to a key.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key.
+ The return value is the ID number of the leaf, or -1 on failure. */
+static int vlsearchleaf(VILLA *villa, const char *kbuf, int ksiz){
+ VLNODE *node;
+ VLIDX *idxp;
+ int i, pid, rv, left, right, ln;
+ assert(villa && kbuf && ksiz >= 0);
+ pid = villa->root;
+ idxp = NULL;
+ villa->hnum = 0;
+ villa->hleaf = -1;
+ while(pid >= VL_NODEIDMIN){
+ if(!(node = vlnodeload(villa, pid)) || (ln = CB_LISTNUM(node->idxs)) < 1){
+ dpecodeset(DP_EBROKEN, __FILE__, __LINE__);
+ return -1;
+ }
+ villa->hist[villa->hnum++] = node->id;
+ left = 1;
+ right = ln;
+ i = (left + right) / 2;
+ while(right >= left && i < ln){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(idxp->key), CB_DATUMSIZE(idxp->key));
+ if(rv == 0){
+ break;
+ } else if(rv <= 0){
+ right = i - 1;
+ } else {
+ left = i + 1;
+ }
+ i = (left + right) / 2;
+ }
+ if(i > 0) i--;
+ while(i < ln){
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i);
+ if(villa->cmp(kbuf, ksiz, CB_DATUMPTR(idxp->key), CB_DATUMSIZE(idxp->key)) < 0){
+ if(i == 0){
+ pid = node->heir;
+ break;
+ }
+ idxp = (VLIDX *)CB_LISTVAL(node->idxs, i - 1);
+ pid = idxp->pid;
+ break;
+ }
+ i++;
+ }
+ if(i >= ln) pid = idxp->pid;
+ }
+ if(villa->lleaf == pid) villa->hleaf = pid;
+ villa->lleaf = pid;
+ return pid;
+}
+
+
+/* Adjust the caches for leaves and nodes.
+ `villa' specifies a database handle.
+ The return value is true if successful, else, it is false. */
+static int vlcacheadjust(VILLA *villa){
+ const char *tmp;
+ int i, pid, err;
+ err = FALSE;
+ if(cbmaprnum(villa->leafc) > villa->leafcnum){
+ cbmapiterinit(villa->leafc);
+ for(i = 0; i < VL_CACHEOUT; i++){
+ tmp = cbmapiternext(villa->leafc, NULL);
+ pid = *(int *)tmp;
+ if(!vlleafcacheout(villa, pid)) err = TRUE;
+ }
+ }
+ if(cbmaprnum(villa->nodec) > villa->nodecnum){
+ cbmapiterinit(villa->nodec);
+ for(i = 0; i < VL_CACHEOUT; i++){
+ tmp = cbmapiternext(villa->nodec, NULL);
+ pid = *(int *)tmp;
+ if(!vlnodecacheout(villa, pid)) err = TRUE;
+ }
+ }
+ return err ? FALSE : TRUE;
+}
+
+
+/* Search a record of a leaf.
+ `villa' specifies a database handle.
+ `leaf' specifies a leaf handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key.
+ `ip' specifies the pointer to a variable to fetch the index of the correspnding record.
+ The return value is the pointer to a corresponding record, or `NULL' on failure. */
+static VLREC *vlrecsearch(VILLA *villa, VLLEAF *leaf, const char *kbuf, int ksiz, int *ip){
+ VLREC *recp;
+ CBLIST *recs;
+ int i, rv, left, right, ln;
+ assert(villa && leaf && kbuf && ksiz >= 0);
+ recs = leaf->recs;
+ ln = CB_LISTNUM(recs);
+ left = 0;
+ right = ln;
+ i = (left + right) / 2;
+ while(right >= left && i < ln){
+ recp = (VLREC *)CB_LISTVAL(recs, i);
+ rv = villa->cmp(kbuf, ksiz, CB_DATUMPTR(recp->key), CB_DATUMSIZE(recp->key));
+ if(rv == 0){
+ if(ip) *ip = i;
+ return recp;
+ } else if(rv <= 0){
+ right = i - 1;
+ } else {
+ left = i + 1;
+ }
+ i = (left + right) / 2;
+ }
+ if(ip) *ip = i;
+ return NULL;
+}
+
+
+/* Get flags of a database. */
+int vlgetflags(VILLA *villa){
+ assert(villa);
+ return dpgetflags(villa->depot);
+}
+
+
+/* Set flags of a database.
+ `villa' specifies a database handle connected as a writer.
+ `flags' specifies flags to set. Lesser ten bits are reserved for internal use.
+ If successful, the return value is true, else, it is false. */
+int vlsetflags(VILLA *villa, int flags){
+ assert(villa);
+ if(!villa->wmode){
+ dpecodeset(DP_EMODE, __FILE__, __LINE__);
+ return FALSE;
+ }
+ return dpsetflags(villa->depot, flags);
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/villa.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/villa.h
===================================================================
--- box/trunk/qdbm/villa.h (rev 0)
+++ box/trunk/qdbm/villa.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,758 @@
+/*************************************************************************************************
+ * The advanced API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _VILLA_H /* duplication check */
+#define _VILLA_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <stdlib.h>
+#include <time.h>
+
+
+#if defined(_MSC_VER) && !defined(QDBM_INTERNAL) && !defined(QDBM_STATIC)
+#define MYEXTERN extern __declspec(dllimport)
+#else
+#define MYEXTERN extern
+#endif
+
+
+
+/*************************************************************************************************
+ * API
+ *************************************************************************************************/
+
+
+#define VL_LEVELMAX 64 /* max level of B+ tree */
+
+typedef struct { /* type of structure for a record */
+ CBDATUM *key; /* datum of the key */
+ CBDATUM *first; /* datum of the first value */
+ CBLIST *rest; /* list of the rest values */
+} VLREC;
+
+typedef struct { /* type of structure for index of a page */
+ int pid; /* ID number of the referring page */
+ CBDATUM *key; /* threshold key of the page */
+} VLIDX;
+
+typedef struct { /* type of structure for a leaf page */
+ int id; /* ID number of the leaf */
+ int dirty; /* whether to be written back */
+ CBLIST *recs; /* list of records */
+ int prev; /* ID number of the previous leaf */
+ int next; /* ID number of the next leaf */
+} VLLEAF;
+
+typedef struct { /* type of structure for a node page */
+ int id; /* ID number of the node */
+ int dirty; /* whether to be written back */
+ int heir; /* ID of the child before the first index */
+ CBLIST *idxs; /* list of indexes */
+} VLNODE;
+
+/* type of the pointer to a comparing function.
+ `aptr' specifies the pointer to the region of one key.
+ `asiz' specifies the size of the region of one key.
+ `bptr' specifies the pointer to the region of the other key.
+ `bsiz' specifies the size of the region of the other key.
+ The return value is positive if the former is big, negative if the latter is big, 0 if both
+ are equivalent. */
+typedef int (*VLCFUNC)(const char *aptr, int asiz, const char *bptr, int bsiz);
+MYEXTERN VLCFUNC VL_CMPLEX; /* lexical comparing function */
+MYEXTERN VLCFUNC VL_CMPINT; /* native integer comparing function */
+MYEXTERN VLCFUNC VL_CMPNUM; /* big endian number comparing function */
+MYEXTERN VLCFUNC VL_CMPDEC; /* decimal string comparing function */
+
+typedef struct { /* type of structure for a database handle */
+ DEPOT *depot; /* internal database handle */
+ VLCFUNC cmp; /* pointer to the comparing function */
+ int wmode; /* whether to be writable */
+ int cmode; /* compression mode for leaves */
+ int root; /* ID number of the root page */
+ int last; /* ID number of the last leaf */
+ int lnum; /* number of leaves */
+ int nnum; /* number of nodes */
+ int rnum; /* number of records */
+ CBMAP *leafc; /* cache for leaves */
+ CBMAP *nodec; /* cache for nodes */
+ int hist[VL_LEVELMAX]; /* array history of visited nodes */
+ int hnum; /* number of elements of the history */
+ int hleaf; /* ID number of the leaf referred by the history */
+ int lleaf; /* ID number of the last visited leaf */
+ int curleaf; /* ID number of the leaf where the cursor is */
+ int curknum; /* index of the key where the cursor is */
+ int curvnum; /* index of the value where the cursor is */
+ int leafrecmax; /* max number of records in a leaf */
+ int nodeidxmax; /* max number of indexes in a node */
+ int leafcnum; /* max number of caching leaves */
+ int nodecnum; /* max number of caching nodes */
+ int avglsiz; /* average size of each leave */
+ int avgnsiz; /* average size of each node */
+ int tran; /* whether in the transaction */
+ int rbroot; /* root for rollback */
+ int rblast; /* last for rollback */
+ int rblnum; /* lnum for rollback */
+ int rbnnum; /* nnum for rollback */
+ int rbrnum; /* rnum for rollback */
+} VILLA;
+
+typedef struct { /* type of structure for a multiple cursor handle */
+ VILLA *villa; /* database handle */
+ int curleaf; /* ID number of the leaf where the cursor is */
+ int curknum; /* index of the key where the cursor is */
+ int curvnum; /* index of the value where the cursor is */
+} VLMULCUR;
+
+enum { /* enumeration for open modes */
+ VL_OREADER = 1 << 0, /* open as a reader */
+ VL_OWRITER = 1 << 1, /* open as a writer */
+ VL_OCREAT = 1 << 2, /* a writer creating */
+ VL_OTRUNC = 1 << 3, /* a writer truncating */
+ VL_ONOLCK = 1 << 4, /* open without locking */
+ VL_OLCKNB = 1 << 5, /* lock without blocking */
+ VL_OZCOMP = 1 << 6, /* compress leaves with ZLIB */
+ VL_OYCOMP = 1 << 7, /* compress leaves with LZO */
+ VL_OXCOMP = 1 << 8 /* compress leaves with BZIP2 */
+};
+
+enum { /* enumeration for write modes */
+ VL_DOVER, /* overwrite the existing value */
+ VL_DKEEP, /* keep the existing value */
+ VL_DCAT, /* concatenate values */
+ VL_DDUP, /* allow duplication of keys */
+ VL_DDUPR /* allow duplication with reverse order */
+};
+
+enum { /* enumeration for jump modes */
+ VL_JFORWARD, /* step forward */
+ VL_JBACKWARD /* step backward */
+};
+
+enum { /* enumeration for insertion modes */
+ VL_CPCURRENT, /* overwrite the current record */
+ VL_CPBEFORE, /* insert before the current record */
+ VL_CPAFTER /* insert after the current record */
+};
+
+
+/* Get a database handle.
+ `name' specifies the name of a database file.
+ `omode' specifies the connection mode: `VL_OWRITER' as a writer, `VL_OREADER' as a reader.
+ If the mode is `VL_OWRITER', the following may be added by bitwise or: `VL_OCREAT', which
+ means it creates a new database if not exist, `VL_OTRUNC', which means it creates a new
+ database regardless if one exists, `VL_OZCOMP', which means leaves in the database are
+ compressed with ZLIB, `VL_OYCOMP', which means leaves in the database are compressed with LZO,
+ `VL_OXCOMP', which means leaves in the database are compressed with BZIP2. Both of
+ `VL_OREADER' and `VL_OWRITER' can be added to by bitwise or: `VL_ONOLCK', which means it opens
+ a database file without file locking, or `VL_OLCKNB', which means locking is performed without
+ blocking.
+ `cmp' specifies a comparing function: `VL_CMPLEX' comparing keys in lexical order,
+ `VL_CMPINT' comparing keys as objects of `int' in native byte order, `VL_CMPNUM' comparing
+ keys as numbers of big endian, `VL_CMPDEC' comparing keys as decimal strings. Any function
+ based on the declaration of the type `VLCFUNC' can be assigned to the comparing function.
+ The comparing function should be kept same in the life of a database.
+ The return value is the database handle or `NULL' if it is not successful.
+ While connecting as a writer, an exclusive lock is invoked to the database file.
+ While connecting as a reader, a shared lock is invoked to the database file. The thread
+ blocks until the lock is achieved. `VL_OZCOMP', `VL_OYCOMP', and `VL_OXCOMP' are available
+ only if QDBM was built each with ZLIB, LZO, and BZIP2 enabled. If `VL_ONOLCK' is used, the
+ application is responsible for exclusion control. */
+VILLA *vlopen(const char *name, int omode, VLCFUNC cmp);
+
+
+/* Close a database handle.
+ `villa' specifies a database handle.
+ If successful, the return value is true, else, it is false.
+ Because the region of a closed handle is released, it becomes impossible to use the handle.
+ Updating a database is assured to be written when the handle is closed. If a writer opens
+ a database but does not close it appropriately, the database will be broken. If the
+ transaction is activated and not committed, it is aborted. */
+int vlclose(VILLA *villa);
+
+
+/* Store a record.
+ `villa' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ `dmode' specifies behavior when the key overlaps, by the following values: `VL_DOVER',
+ which means the specified value overwrites the existing one, `VL_DKEEP', which means the
+ existing value is kept, `VL_DCAT', which means the specified value is concatenated at the
+ end of the existing value, `VL_DDUP', which means duplication of keys is allowed and the
+ specified value is added as the last one, `VL_DDUPR', which means duplication of keys is
+ allowed and the specified value is added as the first one.
+ If successful, the return value is true, else, it is false.
+ The cursor becomes unavailable due to updating database. */
+int vlput(VILLA *villa, const char *kbuf, int ksiz, const char *vbuf, int vsiz, int dmode);
+
+
+/* Delete a record.
+ `villa' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is true, else, it is false. False is returned when no
+ record corresponds to the specified key.
+ When the key of duplicated records is specified, the first record of the same key is deleted.
+ The cursor becomes unavailable due to updating database. */
+int vlout(VILLA *villa, const char *kbuf, int ksiz);
+
+
+/* Retrieve a record.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the specified key.
+ When the key of duplicated records is specified, the value of the first record of the same
+ key is selected. Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+char *vlget(VILLA *villa, const char *kbuf, int ksiz, int *sp);
+
+
+/* Get the size of the value of a record.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is the size of the value of the corresponding record, else,
+ it is -1. If multiple records correspond, the size of the first is returned. */
+int vlvsiz(VILLA *villa, const char *kbuf, int ksiz);
+
+
+/* Get the number of records corresponding a key.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ The return value is the number of corresponding records. If no record corresponds, 0 is
+ returned. */
+int vlvnum(VILLA *villa, const char *kbuf, int ksiz);
+
+
+/* Store plural records corresponding a key.
+ `villa' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `vals' specifies a list handle of values. The list should not be empty.
+ If successful, the return value is true, else, it is false.
+ The cursor becomes unavailable due to updating database. */
+int vlputlist(VILLA *villa, const char *kbuf, int ksiz, const CBLIST *vals);
+
+
+/* Delete all records corresponding a key.
+ `villa' specifies a database handle connected as a writer.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is true, else, it is false. False is returned when no
+ record corresponds to the specified key.
+ The cursor becomes unavailable due to updating database. */
+int vloutlist(VILLA *villa, const char *kbuf, int ksiz);
+
+
+/* Retrieve values of all records corresponding a key.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ If successful, the return value is a list handle of the values of the corresponding records,
+ else, it is `NULL'. `NULL' is returned when no record corresponds to the specified key.
+ Because the handle of the return value is opened with the function `cblistopen', it should
+ be closed with the function `cblistclose' if it is no longer in use. */
+CBLIST *vlgetlist(VILLA *villa, const char *kbuf, int ksiz);
+
+
+/* Retrieve concatenated values of all records corresponding a key.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the concatenated values of
+ the corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds
+ to the specified key. Because an additional zero code is appended at the end of the region of
+ the return value, the return value can be treated as a character string. Because the region
+ of the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+char *vlgetcat(VILLA *villa, const char *kbuf, int ksiz, int *sp);
+
+
+/* Move the cursor to the first record.
+ `villa' specifies a database handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no record in the database. */
+int vlcurfirst(VILLA *villa);
+
+
+/* Move the cursor to the last record.
+ `villa' specifies a database handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no record in the database. */
+int vlcurlast(VILLA *villa);
+
+
+/* Move the cursor to the previous record.
+ `villa' specifies a database handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no previous record. */
+int vlcurprev(VILLA *villa);
+
+
+/* Move the cursor to the next record.
+ `villa' specifies a database handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no next record. */
+int vlcurnext(VILLA *villa);
+
+
+/* Move the cursor to a position around a record.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `jmode' specifies detail adjustment: `VL_JFORWARD', which means that the cursor is set to
+ the first record of the same key and that the cursor is set to the next substitute if
+ completely matching record does not exist, `VL_JBACKWARD', which means that the cursor is
+ set to the last record of the same key and that the cursor is set to the previous substitute
+ if completely matching record does not exist.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no record corresponding the condition. */
+int vlcurjump(VILLA *villa, const char *kbuf, int ksiz, int jmode);
+
+
+/* Get the key of the record where the cursor is.
+ `villa' specifies a database handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the key of the corresponding
+ record, else, it is `NULL'. `NULL' is returned when no record corresponds to the cursor.
+ Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+char *vlcurkey(VILLA *villa, int *sp);
+
+
+/* Get the value of the record where the cursor is.
+ `villa' specifies a database handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the cursor.
+ Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+char *vlcurval(VILLA *villa, int *sp);
+
+
+/* Insert a record around the cursor.
+ `villa' specifies a database handle connected as a writer.
+ `vbuf' specifies the pointer to the region of a value.
+ `vsiz' specifies the size of the region of the value. If it is negative, the size is
+ assigned with `strlen(vbuf)'.
+ `cpmode' specifies detail adjustment: `VL_CPCURRENT', which means that the value of the
+ current record is overwritten, `VL_CPBEFORE', which means that a new record is inserted before
+ the current record, `VL_CPAFTER', which means that a new record is inserted after the current
+ record.
+ If successful, the return value is true, else, it is false. False is returned when no record
+ corresponds to the cursor.
+ After insertion, the cursor is moved to the inserted record. */
+int vlcurput(VILLA *villa, const char *vbuf, int vsiz, int cpmode);
+
+
+/* Delete the record where the cursor is.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. False is returned when no record
+ corresponds to the cursor.
+ After deletion, the cursor is moved to the next record if possible. */
+int vlcurout(VILLA *villa);
+
+
+/* Set the tuning parameters for performance.
+ `villa' specifies a database handle.
+ `lrecmax' specifies the max number of records in a leaf node of B+ tree. If it is not more
+ than 0, the default value is specified.
+ `nidxmax' specifies the max number of indexes in a non-leaf node of B+ tree. If it is not
+ more than 0, the default value is specified.
+ `lcnum' specifies the max number of caching leaf nodes. If it is not more than 0, the
+ default value is specified.
+ `ncnum' specifies the max number of caching non-leaf nodes. If it is not more than 0, the
+ default value is specified.
+ The default setting is equivalent to `vlsettuning(49, 192, 1024, 512)'. Because tuning
+ parameters are not saved in a database, you should specify them every opening a database. */
+void vlsettuning(VILLA *villa, int lrecmax, int nidxmax, int lcnum, int ncnum);
+
+
+/* Set the size of the free block pool of a database handle.
+ `villa' specifies a database handle connected as a writer.
+ `size' specifies the size of the free block pool of a database.
+ If successful, the return value is true, else, it is false.
+ The default size of the free block pool is 256. If the size is greater, the space efficiency
+ of overwriting values is improved with the time efficiency sacrificed. */
+int vlsetfbpsiz(VILLA *villa, int size);
+
+
+/* Synchronize updating contents with the file and the device.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ This function is useful when another process uses the connected database file. This function
+ should not be used while the transaction is activated. */
+int vlsync(VILLA *villa);
+
+
+/* Optimize a database.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ In an alternating succession of deleting and storing with overwrite or concatenate,
+ dispensable regions accumulate. This function is useful to do away with them. This function
+ should not be used while the transaction is activated. */
+int vloptimize(VILLA *villa);
+
+
+/* Get the name of a database.
+ `villa' specifies a database handle.
+ If successful, the return value is the pointer to the region of the name of the database,
+ else, it is `NULL'.
+ Because the region of the return value is allocated with the `malloc' call, it should be
+ released with the `free' call if it is no longer in use. */
+char *vlname(VILLA *villa);
+
+
+/* Get the size of a database file.
+ `villa' specifies a database handle.
+ If successful, the return value is the size of the database file, else, it is -1.
+ Because of the I/O buffer, the return value may be less than the hard size. */
+int vlfsiz(VILLA *villa);
+
+
+/* Get the number of the leaf nodes of B+ tree.
+ `villa' specifies a database handle.
+ If successful, the return value is the number of the leaf nodes, else, it is -1. */
+int vllnum(VILLA *villa);
+
+
+/* Get the number of the non-leaf nodes of B+ tree.
+ `villa' specifies a database handle.
+ If successful, the return value is the number of the non-leaf nodes, else, it is -1. */
+int vlnnum(VILLA *villa);
+
+
+/* Get the number of the records stored in a database.
+ `villa' specifies a database handle.
+ If successful, the return value is the number of the records stored in the database, else,
+ it is -1. */
+int vlrnum(VILLA *villa);
+
+
+/* Check whether a database handle is a writer or not.
+ `villa' specifies a database handle.
+ The return value is true if the handle is a writer, false if not. */
+int vlwritable(VILLA *villa);
+
+
+/* Check whether a database has a fatal error or not.
+ `villa' specifies a database handle.
+ The return value is true if the database has a fatal error, false if not. */
+int vlfatalerror(VILLA *villa);
+
+
+/* Get the inode number of a database file.
+ `villa' specifies a database handle.
+ The return value is the inode number of the database file. */
+int vlinode(VILLA *villa);
+
+
+/* Get the last modified time of a database.
+ `villa' specifies a database handle.
+ The return value is the last modified time of the database. */
+time_t vlmtime(VILLA *villa);
+
+
+/* Begin the transaction.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ Because this function does not perform mutual exclusion control in multi-thread, the
+ application is responsible for it. Only one transaction can be activated with a database
+ handle at the same time. */
+int vltranbegin(VILLA *villa);
+
+
+/* Commit the transaction.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ Updating a database in the transaction is fixed when it is committed successfully. */
+int vltrancommit(VILLA *villa);
+
+
+/* Abort the transaction.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false.
+ Updating a database in the transaction is discarded when it is aborted. The state of the
+ database is rollbacked to before transaction. */
+int vltranabort(VILLA *villa);
+
+
+/* Remove a database file.
+ `name' specifies the name of a database file.
+ If successful, the return value is true, else, it is false. */
+int vlremove(const char *name);
+
+
+/* Repair a broken database file.
+ `name' specifies the name of a database file.
+ `cmp' specifies the comparing function of the database file.
+ If successful, the return value is true, else, it is false.
+ There is no guarantee that all records in a repaired database file correspond to the original
+ or expected state. */
+int vlrepair(const char *name, VLCFUNC cmp);
+
+
+/* Dump all records as endian independent data.
+ `villa' specifies a database handle.
+ `name' specifies the name of an output file.
+ If successful, the return value is true, else, it is false. */
+int vlexportdb(VILLA *villa, const char *name);
+
+
+/* Load all records from endian independent data.
+ `villa' specifies a database handle connected as a writer. The database of the handle must
+ be empty.
+ `name' specifies the name of an input file.
+ If successful, the return value is true, else, it is false. */
+int vlimportdb(VILLA *villa, const char *name);
+
+
+
+/*************************************************************************************************
+ * features for experts
+ *************************************************************************************************/
+
+
+/* Number of division of the database for Vista. */
+#define vlcrdnum (*vlcrdnumptr())
+
+
+/* Get the pointer of the variable of the number of division of the database for Vista.
+ The return value is the pointer of the variable. */
+int *vlcrdnumptr(void);
+
+
+/* Synchronize updating contents on memory.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. */
+int vlmemsync(VILLA *villa);
+
+
+/* Synchronize updating contents on memory, not physically.
+ `villa' specifies a database handle connected as a writer.
+ If successful, the return value is true, else, it is false. */
+int vlmemflush(VILLA *villa);
+
+
+/* Refer to volatile cache of a value of a record.
+ `villa' specifies a database handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the specified key.
+ Because the region of the return value is volatile and it may be spoiled by another operation
+ of the database, the data should be copied into another involatile buffer immediately. */
+const char *vlgetcache(VILLA *villa, const char *kbuf, int ksiz, int *sp);
+
+
+/* Refer to volatile cache of the key of the record where the cursor is.
+ `villa' specifies a database handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the key of the corresponding
+ record, else, it is `NULL'. `NULL' is returned when no record corresponds to the cursor.
+ Because the region of the return value is volatile and it may be spoiled by another operation
+ of the database, the data should be copied into another involatile buffer immediately. */
+const char *vlcurkeycache(VILLA *villa, int *sp);
+
+
+/* Refer to volatile cache of the value of the record where the cursor is.
+ `villa' specifies a database handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the cursor.
+ Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+const char *vlcurvalcache(VILLA *villa, int *sp);
+
+
+/* Get a multiple cursor handle.
+ `villa' specifies a database handle connected as a reader.
+ The return value is a multiple cursor handle or `NULL' if it is not successful.
+ The returned object is should be closed before the database handle is closed. Even if plural
+ cursors are fetched out of a database handle, they does not share the locations with each
+ other. Note that this function can be used only if the database handle is connected as a
+ reader. */
+VLMULCUR *vlmulcuropen(VILLA *villa);
+
+
+/* Close a multiple cursor handle.
+ `mulcur' specifies a multiple cursor handle. */
+void vlmulcurclose(VLMULCUR *mulcur);
+
+
+/* Move a multiple cursor to the first record.
+ `mulcur' specifies a multiple cursor handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no record in the database. */
+int vlmulcurfirst(VLMULCUR *mulcur);
+
+
+/* Move a multiple cursor to the last record.
+ `mulcur' specifies a multiple cursor handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no record in the database. */
+int vlmulcurlast(VLMULCUR *mulcur);
+
+
+/* Move a multiple cursor to the previous record.
+ `mulcur' specifies a multiple cursor handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no previous record. */
+int vlmulcurprev(VLMULCUR *mulcur);
+
+
+/* Move a multiple cursor to the next record.
+ `mulcur' specifies a multiple cursor handle.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no next record. */
+int vlmulcurnext(VLMULCUR *mulcur);
+
+
+/* Move a multiple cursor to a position around a record.
+ `mulcur' specifies a multiple cursor handle.
+ `kbuf' specifies the pointer to the region of a key.
+ `ksiz' specifies the size of the region of the key. If it is negative, the size is assigned
+ with `strlen(kbuf)'.
+ `jmode' specifies detail adjustment: `VL_JFORWARD', which means that the cursor is set to
+ the first record of the same key and that the cursor is set to the next substitute if
+ completely matching record does not exist, `VL_JBACKWARD', which means that the cursor is
+ set to the last record of the same key and that the cursor is set to the previous substitute
+ if completely matching record does not exist.
+ If successful, the return value is true, else, it is false. False is returned if there is
+ no record corresponding the condition. */
+int vlmulcurjump(VLMULCUR *mulcur, const char *kbuf, int ksiz, int jmode);
+
+
+/* Get the key of the record where a multiple cursor is.
+ `mulcur' specifies a multiple cursor handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the key of the corresponding
+ record, else, it is `NULL'. `NULL' is returned when no record corresponds to the cursor.
+ Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+char *vlmulcurkey(VLMULCUR *mulcur, int *sp);
+
+
+/* Get the value of the record where a multiple cursor is.
+ `mulcur' specifies a multiple cursor handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the cursor.
+ Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+char *vlmulcurval(VLMULCUR *mulcur, int *sp);
+
+
+/* Refer to volatile cache of the key of the record where a multiple cursor is.
+ `mulcur' specifies a multiple cursor handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the key of the corresponding
+ record, else, it is `NULL'. `NULL' is returned when no record corresponds to the cursor.
+ Because the region of the return value is volatile and it may be spoiled by another operation
+ of the database, the data should be copied into another involatile buffer immediately. */
+const char *vlmulcurkeycache(VLMULCUR *mulcur, int *sp);
+
+
+/* Refer to volatile cache of the value of the record where a multiple cursor is.
+ `mulcur' specifies a multiple cursor handle.
+ `sp' specifies the pointer to a variable to which the size of the region of the return
+ value is assigned. If it is `NULL', it is not used.
+ If successful, the return value is the pointer to the region of the value of the
+ corresponding record, else, it is `NULL'. `NULL' is returned when no record corresponds to
+ the cursor.
+ Because an additional zero code is appended at the end of the region of the
+ return value, the return value can be treated as a character string. Because the region of
+ the return value is allocated with the `malloc' call, it should be released with the `free'
+ call if it is no longer in use. */
+const char *vlmulcurvalcache(VLMULCUR *mulcur, int *sp);
+
+
+/* Get flags of a database.
+ `villa' specifies a database handle.
+ The return value is the flags of a database. */
+int vlgetflags(VILLA *villa);
+
+
+/* Set flags of a database.
+ `villa' specifies a database handle connected as a writer.
+ `flags' specifies flags to set. Least ten bits are reserved for internal use.
+ If successful, the return value is true, else, it is false. */
+int vlsetflags(VILLA *villa, int flags);
+
+
+
+#undef MYEXTERN
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/villa.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/vista.c
===================================================================
--- box/trunk/qdbm/vista.c (rev 0)
+++ box/trunk/qdbm/vista.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,171 @@
+/*************************************************************************************************
+ * Implementation of Vista
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#define QDBM_INTERNAL 1
+#define _VISTA_C 1
+
+#include "vista.h"
+#include "myconf.h"
+
+
+
+/*************************************************************************************************
+ * macros to convert Depot to Curia
+ *************************************************************************************************/
+
+
+#define DEPOT CURIA
+
+#define \
+ dpopen(name, omode, bnum) \
+ cropen(name, omode, ((bnum / vlcrdnum) * 2), vlcrdnum)
+
+#define \
+ dpclose(db) \
+ crclose(db)
+
+#define \
+ dpput(db, kbuf, ksiz, vbuf, vsiz, dmode) \
+ crput(db, kbuf, ksiz, vbuf, vsiz, dmode)
+
+#define \
+ dpout(db, kbuf, ksiz) \
+ crout(db, kbuf, ksiz)
+
+#define \
+ dpget(db, kbuf, ksiz, start, max, sp) \
+ crget(db, kbuf, ksiz, start, max, sp)
+
+#define \
+ dpgetwb(db, kbuf, ksiz, start, max, vbuf) \
+ crgetwb(db, kbuf, ksiz, start, max, vbuf)
+
+#define \
+ dpvsiz(db, kbuf, ksiz) \
+ crvsiz(db, kbuf, ksiz)
+
+#define \
+ dpiterinit(db) \
+ criterinit(db)
+
+#define \
+ dpiternext(db, sp) \
+ criternext(db, sp)
+
+#define \
+ dpsetalign(db, align) \
+ crsetalign(db, align)
+
+#define \
+ dpsetfbpsiz(db, size) \
+ crsetfbpsiz(db, size)
+
+#define \
+ dpsync(db) \
+ crsync(db)
+
+#define \
+ dpoptimize(db, bnum) \
+ croptimize(db, bnum)
+
+#define \
+ dpname(db) \
+ crname(db)
+
+#define \
+ dpfsiz(db) \
+ crfsiz(db)
+
+#define \
+ dpbnum(db) \
+ crbnum(db)
+
+#define \
+ dpbusenum(db) \
+ crbusenum(db)
+
+#define \
+ dprnum(db) \
+ crrnum(db)
+
+#define \
+ dpwritable(db) \
+ crwritable(db)
+
+#define \
+ dpfatalerror(db) \
+ crfatalerror(db)
+
+#define \
+ dpinode(db) \
+ crinode(db)
+
+#define \
+ dpmtime(db) \
+ crmtime(db)
+
+#define \
+ dpfdesc(db) \
+ crfdesc(db)
+
+#define \
+ dpremove(db) \
+ crremove(db)
+
+#define \
+ dprepair(db) \
+ crrepair(db)
+
+#define \
+ dpexportdb(db, name) \
+ crexportdb(db, name)
+
+#define \
+ dpimportdb(db, name) \
+ crimportdb(db, name)
+
+#define \
+ dpsnaffle(db, name) \
+ crsnaffle(db, name)
+
+#define \
+ dpmemsync(db) \
+ crmemsync(db)
+
+#define \
+ dpmemflush(db) \
+ crmemflush(db)
+
+#define \
+ dpgetflags(db) \
+ crgetflags(db)
+
+#define \
+ dpsetflags(db, flags) \
+ crsetflags(db, flags)
+
+
+
+/*************************************************************************************************
+ * including real implementation
+ *************************************************************************************************/
+
+
+#include "villa.c"
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/vista.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/vista.h
===================================================================
--- box/trunk/qdbm/vista.h (rev 0)
+++ box/trunk/qdbm/vista.h 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,138 @@
+/*************************************************************************************************
+ * The extended advanced API of QDBM
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#ifndef _VISTA_H /* duplication check */
+#define _VISTA_H
+
+#if defined(__cplusplus) /* export for C++ */
+extern "C" {
+#endif
+
+
+
+/*************************************************************************************************
+ * macros to borrow symbols from Villa
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <curia.h>
+#include <cabin.h>
+#include <stdlib.h>
+
+#define VLREC VSTREC
+#define VLIDX VSTIDX
+#define VLLEAF VSTLEAF
+#define VLNODE VSTNODE
+#define VLCFUNC VSTCFUNC
+
+#define VL_CMPLEX VST_CMPLEX
+#define VL_CMPINT VST_CMPINT
+#define VL_CMPNUM VST_CMPNUM
+#define VL_CMPDEC VST_CMPDEC
+
+#define VILLA VISTA
+
+#define VL_OREADER VST_OREADER
+#define VL_OWRITER VST_OWRITER
+#define VL_OCREAT VST_OCREAT
+#define VL_OTRUNC VST_OTRUNC
+#define VL_ONOLCK VST_ONOLCK
+#define VL_OLCKNB VST_OLCKNB
+#define VL_OZCOMP VST_OZCOMP
+
+#define VL_DOVER VST_DOVER
+#define VL_DKEEP VST_DKEEP
+#define VL_DCAT VST_DCAT
+#define VL_DDUP VST_DDUP
+
+#define VL_JFORWARD VST_JFORWARD
+#define VL_JBACKWARD VST_JBACKWARD
+
+#define vlopen vstopen
+#define vlclose vstclose
+#define vlput vstput
+#define vlout vstout
+#define vlget vstget
+#define vlvsiz vstvsiz
+#define vlvnum vstvnum
+#define vlputlist vstputlist
+#define vloutlist vstoutlist
+#define vlgetlist vstgetlist
+#define vlgetcat vstgetcat
+#define vlcurfirst vstcurfirst
+#define vlcurlast vstcurlast
+#define vlcurprev vstcurprev
+#define vlcurnext vstcurnext
+#define vlcurjump vstcurjump
+#define vlcurkey vstcurkey
+#define vlcurval vstcurval
+#define vlcurput vstcurput
+#define vlcurout vstcurout
+#define vlsettuning vstsettuning
+#define vlsync vstsync
+#define vloptimize vstoptimize
+#define vlname vstname
+#define vlfsiz vstfsiz
+#define vllnum vstlnum
+#define vlnnum vstnnum
+#define vlrnum vstrnum
+#define vlwritable vstwritable
+#define vlfatalerror vstfatalerror
+#define vlinode vstinode
+#define vlmtime vstmtime
+#define vltranbegin vsttranbegin
+#define vltrancommit vsttrancommit
+#define vltranabort vsttranabort
+#define vlremove vstremove
+#define vlrepair vstrepair
+#define vlexportdb vstexportdb
+#define vlimportdb vstimportdb
+#define vlcrdnumptr vstcrdnumptr
+#define vlmemsync vstmemsync
+#define vlmemflush vstmemflush
+#define vlgetcache vstgetcache
+#define vlcurkeycache vstcurkeycache
+#define vlcurvalcache vstcurvalcache
+#define vlmulcuropen vstmulcuropen
+#define vlmulcurclose vstmulcurclose
+#define vlmulcurfirst vstmulcurfirst
+#define vlmulcurlast vstmulcurlast
+#define vlmulcurprev vstmulcurprev
+#define vlmulcurnext vstmulcurnext
+#define vlmulcurjump vstmulcurjump
+#define vlmulcurkey vstmulcurkey
+#define vlmulcurval vstmulcurval
+#define vlmulcurkeycache vstmulcurkeycache
+#define vlmulcurvalcache vstmulcurvalcache
+#define vlsetfbpsiz vstsetfbpsiz
+#define vlgetflags vstgetflags
+#define vlsetflags vstsetflags
+
+#if !defined(_VISTA_C)
+#include <villa.h>
+#endif
+
+
+
+#if defined(__cplusplus) /* export for C++ */
+}
+#endif
+
+#endif /* duplication check */
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/vista.h
___________________________________________________________________
Added: svn:mime-type
+ text/x-chdr
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/vlmgr.c
===================================================================
--- box/trunk/qdbm/vlmgr.c (rev 0)
+++ box/trunk/qdbm/vlmgr.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,968 @@
+/*************************************************************************************************
+ * Utility for debugging Villa and its applications
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+char *hextoobj(const char *str, int *sp);
+char *dectoiobj(const char *str, int *sp);
+int runcreate(int argc, char **argv);
+int runput(int argc, char **argv);
+int runout(int argc, char **argv);
+int runget(int argc, char **argv);
+int runlist(int argc, char **argv);
+int runoptimize(int argc, char **argv);
+int runinform(int argc, char **argv);
+int runremove(int argc, char **argv);
+int runrepair(int argc, char **argv);
+int runexportdb(int argc, char **argv);
+int runimportdb(int argc, char **argv);
+void pdperror(const char *name);
+void printobj(const char *obj, int size);
+void printobjhex(const char *obj, int size);
+int docreate(const char *name, int cmode);
+int doput(const char *name, const char *kbuf, int ksiz, const char *vbuf, int vsiz,
+ int dmode, VLCFUNC cmp);
+int doout(const char *name, const char *kbuf, int ksiz, VLCFUNC cmp, int lb);
+int doget(const char *name, int opts, const char *kbuf, int ksiz, VLCFUNC cmp,
+ int lb, int ox, int nb);
+int dolist(const char *name, int opts, const char *tbuf, int tsiz, const char *bbuf, int bsiz,
+ VLCFUNC cmp, int ki, int kb, int vb, int ox, int gt, int lt, int max, int desc);
+int dooptimize(const char *name);
+int doinform(const char *name, int opts);
+int doremove(const char *name);
+int dorepair(const char *name, VLCFUNC cmp);
+int doexportdb(const char *name, const char *file, VLCFUNC cmp);
+int doimportdb(const char *name, const char *file, VLCFUNC cmp);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "create")){
+ rv = runcreate(argc, argv);
+ } else if(!strcmp(argv[1], "put")){
+ rv = runput(argc, argv);
+ } else if(!strcmp(argv[1], "out")){
+ rv = runout(argc, argv);
+ } else if(!strcmp(argv[1], "get")){
+ rv = runget(argc, argv);
+ } else if(!strcmp(argv[1], "list")){
+ rv = runlist(argc, argv);
+ } else if(!strcmp(argv[1], "optimize")){
+ rv = runoptimize(argc, argv);
+ } else if(!strcmp(argv[1], "inform")){
+ rv = runinform(argc, argv);
+ } else if(!strcmp(argv[1], "remove")){
+ rv = runremove(argc, argv);
+ } else if(!strcmp(argv[1], "repair")){
+ rv = runrepair(argc, argv);
+ } else if(!strcmp(argv[1], "exportdb")){
+ rv = runexportdb(argc, argv);
+ } else if(!strcmp(argv[1], "importdb")){
+ rv = runimportdb(argc, argv);
+ } else if(!strcmp(argv[1], "version") || !strcmp(argv[1], "--version")){
+ printf("Powered by QDBM version %s on %s%s\n",
+ dpversion, dpsysname, dpisreentrant ? " (reentrant)" : "");
+ printf("Copyright (c) 2000-2007 Mikio Hirabayashi\n");
+ rv = 0;
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: administration utility for Villa\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s create [-cz|-cy|-cx] name\n", progname);
+ fprintf(stderr, " %s put [-kx|-ki] [-vx|-vi|-vf] [-keep|-cat|-dup] name key val\n", progname);
+ fprintf(stderr, " %s out [-l] [-kx|-ki] name key\n", progname);
+ fprintf(stderr, " %s get [-nl] [-l] [-kx|-ki] [-ox] [-n] name key\n", progname);
+ fprintf(stderr, " %s list [-nl] [-k|-v] [-kx|-ki] [-ox] [-top key] [-bot key] [-gt] [-lt]"
+ " [-max num] [-desc] name\n", progname);
+ fprintf(stderr, " %s optimize name\n", progname);
+ fprintf(stderr, " %s inform [-nl] name\n", progname);
+ fprintf(stderr, " %s remove name\n", progname);
+ fprintf(stderr, " %s repair [-ki] name\n", progname);
+ fprintf(stderr, " %s exportdb [-ki] name file\n", progname);
+ fprintf(stderr, " %s importdb [-ki] name file\n", progname);
+ fprintf(stderr, " %s version\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* create a binary object from a hexadecimal string */
+char *hextoobj(const char *str, int *sp){
+ char *buf, mbuf[3];
+ int len, i, j;
+ len = strlen(str);
+ if(!(buf = malloc(len + 1))) return NULL;
+ j = 0;
+ for(i = 0; i < len; i += 2){
+ while(strchr(" \n\r\t\f\v", str[i])){
+ i++;
+ }
+ if((mbuf[0] = str[i]) == '\0') break;
+ if((mbuf[1] = str[i+1]) == '\0') break;
+ mbuf[2] = '\0';
+ buf[j++] = (char)strtol(mbuf, NULL, 16);
+ }
+ buf[j] = '\0';
+ *sp = j;
+ return buf;
+}
+
+
+/* create a integer object from a decimal string */
+char *dectoiobj(const char *str, int *sp){
+ char *buf;
+ int num;
+ num = atoi(str);
+ if(!(buf = malloc(sizeof(int)))) return NULL;
+ *(int *)buf = num;
+ *sp = sizeof(int);
+ return buf;
+}
+
+
+/* parse arguments of create command */
+int runcreate(int argc, char **argv){
+ char *name;
+ int i, cmode, rv;
+ name = NULL;
+ cmode = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-cz")){
+ cmode |= VL_OZCOMP;
+ } else if(!strcmp(argv[i], "-cy")){
+ cmode |= VL_OYCOMP;
+ } else if(!strcmp(argv[i], "-cx")){
+ cmode |= VL_OXCOMP;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docreate(name, cmode);
+ return rv;
+}
+
+
+/* parse arguments of put command */
+int runput(int argc, char **argv){
+ char *name, *key, *val, *kbuf, *vbuf;
+ int i, kx, ki, vx, vi, vf, ksiz, vsiz, rv;
+ int dmode;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ vx = FALSE;
+ vi = FALSE;
+ vf = FALSE;
+ key = NULL;
+ val = NULL;
+ dmode = VL_DOVER;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-vx")){
+ vx = TRUE;
+ } else if(!strcmp(argv[i], "-vi")){
+ vi = TRUE;
+ } else if(!strcmp(argv[i], "-vf")){
+ vf = TRUE;
+ } else if(!strcmp(argv[i], "-keep")){
+ dmode = VL_DKEEP;
+ } else if(!strcmp(argv[i], "-cat")){
+ dmode = VL_DCAT;
+ } else if(!strcmp(argv[i], "-dup")){
+ dmode = VL_DDUP;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else if(!val){
+ val = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key || !val) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(vx){
+ vbuf = hextoobj(val, &vsiz);
+ } else if(vi){
+ vbuf = dectoiobj(val, &vsiz);
+ } else if(vf){
+ vbuf = cbreadfile(val, &vsiz);
+ } else {
+ vbuf = cbmemdup(val, -1);
+ vsiz = -1;
+ }
+ if(kbuf && vbuf){
+ rv = doput(name, kbuf, ksiz, vbuf, vsiz, dmode, ki ? VL_CMPINT : VL_CMPLEX);
+ } else {
+ if(vf){
+ fprintf(stderr, "%s: %s: cannot read\n", progname, val);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ }
+ rv = 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ return rv;
+}
+
+
+/* parse arguments of out command */
+int runout(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, kx, ki, lb, ksiz, rv;
+ name = NULL;
+ kx = FALSE;
+ ki = FALSE;
+ lb = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-l")){
+ lb = TRUE;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = doout(name, kbuf, ksiz, ki ? VL_CMPINT : VL_CMPLEX, lb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of get command */
+int runget(int argc, char **argv){
+ char *name, *key, *kbuf;
+ int i, opts, lb, kx, ki, ox, nb, ksiz, rv;
+ name = NULL;
+ opts = 0;
+ lb = FALSE;
+ kx = FALSE;
+ ki = FALSE;
+ ox = FALSE;
+ nb = FALSE;
+ key = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= VL_ONOLCK;
+ } else if(!strcmp(argv[i], "-l")){
+ lb = TRUE;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-n")){
+ nb = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!key){
+ key = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !key) usage();
+ if(kx){
+ kbuf = hextoobj(key, &ksiz);
+ } else if(ki){
+ kbuf = dectoiobj(key, &ksiz);
+ } else {
+ kbuf = cbmemdup(key, -1);
+ ksiz = -1;
+ }
+ if(kbuf){
+ rv = doget(name, opts, kbuf, ksiz, ki ? VL_CMPINT : VL_CMPLEX, lb, ox, nb);
+ } else {
+ fprintf(stderr, "%s: out of memory\n", progname);
+ rv = 1;
+ }
+ free(kbuf);
+ return rv;
+}
+
+
+/* parse arguments of list command */
+int runlist(int argc, char **argv){
+ char *name, *top, *bot, *tbuf, *bbuf, *nstr;
+ int i, opts, kb, vb, kx, ki, ox, gt, lt, max, desc, tsiz, bsiz, rv;
+ name = NULL;
+ opts = 0;
+ kb = FALSE;
+ vb = FALSE;
+ kx = FALSE;
+ ki = FALSE;
+ ox = FALSE;
+ gt = FALSE;
+ lt = FALSE;
+ max = -1;
+ desc = FALSE;
+ top = NULL;
+ bot = NULL;
+ nstr = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= VL_ONOLCK;
+ } else if(!strcmp(argv[i], "-k")){
+ kb = TRUE;
+ } else if(!strcmp(argv[i], "-v")){
+ vb = TRUE;
+ } else if(!strcmp(argv[i], "-kx")){
+ kx = TRUE;
+ } else if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else if(!strcmp(argv[i], "-ox")){
+ ox = TRUE;
+ } else if(!strcmp(argv[i], "-top")){
+ if(++i >= argc) usage();
+ top = argv[i];
+ } else if(!strcmp(argv[i], "-bot")){
+ if(++i >= argc) usage();
+ bot = argv[i];
+ } else if(!strcmp(argv[i], "-gt")){
+ gt = TRUE;
+ } else if(!strcmp(argv[i], "-lt")){
+ lt = TRUE;
+ } else if(!strcmp(argv[i], "-max")){
+ if(++i >= argc) usage();
+ max = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-desc")){
+ desc = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ tbuf = NULL;
+ bbuf = NULL;
+ if(kx){
+ if(top) tbuf = hextoobj(top, &tsiz);
+ if(bot) bbuf = hextoobj(bot, &bsiz);
+ } else if(ki){
+ if(top) tbuf = dectoiobj(top, &tsiz);
+ if(bot) bbuf = dectoiobj(bot, &bsiz);
+ } else {
+ if(top){
+ tbuf = cbmemdup(top, -1);
+ tsiz = strlen(tbuf);
+ }
+ if(bot){
+ bbuf = cbmemdup(bot, -1);
+ bsiz = strlen(bbuf);
+ }
+ }
+ rv = dolist(name, opts, tbuf, tsiz, bbuf, bsiz, ki ? VL_CMPINT : VL_CMPLEX, ki,
+ kb, vb, ox, gt, lt, max, desc);
+ free(tbuf);
+ free(bbuf);
+ return rv;
+}
+
+
+/* parse arguments of optimize command */
+int runoptimize(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dooptimize(name);
+ return rv;
+}
+
+
+/* parse arguments of inform command */
+int runinform(int argc, char **argv){
+ char *name;
+ int i, opts, rv;
+ name = NULL;
+ opts = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-nl")){
+ opts |= VL_ONOLCK;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doinform(name, opts);
+ return rv;
+}
+
+
+/* parse arguments of remove command */
+int runremove(int argc, char **argv){
+ char *name;
+ int i, rv;
+ name = NULL;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ usage();
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doremove(name);
+ return rv;
+}
+
+
+/* parse arguments of repair command */
+int runrepair(int argc, char **argv){
+ char *name;
+ int i, ki, rv;
+ name = NULL;
+ ki = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = dorepair(name, ki ? VL_CMPINT : VL_CMPLEX);
+ return rv;
+}
+
+
+/* parse arguments of exportdb command */
+int runexportdb(int argc, char **argv){
+ char *name, *file;
+ int i, ki, rv;
+ name = NULL;
+ file = NULL;
+ ki = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !file) usage();
+ rv = doexportdb(name, file, ki ? VL_CMPINT : VL_CMPLEX);
+ return rv;
+}
+
+
+/* parse arguments of importdb command */
+int runimportdb(int argc, char **argv){
+ char *name, *file;
+ int i, ki, rv;
+ name = NULL;
+ file = NULL;
+ ki = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-ki")){
+ ki = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!file){
+ file = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !file) usage();
+ rv = doimportdb(name, file, ki ? VL_CMPINT : VL_CMPLEX);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* print an object */
+void printobj(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ putchar(obj[i]);
+ }
+}
+
+
+/* print an object as a hexadecimal string */
+void printobjhex(const char *obj, int size){
+ int i;
+ for(i = 0; i < size; i++){
+ printf("%s%02X", i > 0 ? " " : "", ((const unsigned char *)obj)[i]);
+ }
+}
+
+
+/* perform create command */
+int docreate(const char *name, int cmode){
+ VILLA *villa;
+ int omode;
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC | cmode;
+ if(!(villa = vlopen(name, omode, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform put command */
+int doput(const char *name, const char *kbuf, int ksiz, const char *vbuf, int vsiz,
+ int dmode, VLCFUNC cmp){
+ VILLA *villa;
+ if(!(villa = vlopen(name, VL_OWRITER, cmp))){
+ pdperror(name);
+ return 1;
+ }
+ if(!vlput(villa, kbuf, ksiz, vbuf, vsiz, dmode)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform out command */
+int doout(const char *name, const char *kbuf, int ksiz, VLCFUNC cmp, int lb){
+ VILLA *villa;
+ if(!(villa = vlopen(name, VL_OWRITER, cmp))){
+ pdperror(name);
+ return 1;
+ }
+ if(lb){
+ if(!vloutlist(villa, kbuf, ksiz)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ } else {
+ if(!vlout(villa, kbuf, ksiz)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform get command */
+int doget(const char *name, int opts, const char *kbuf, int ksiz, VLCFUNC cmp,
+ int lb, int ox, int nb){
+ VILLA *villa;
+ CBLIST *vals;
+ char *vbuf;
+ int vsiz;
+ if(!(villa = vlopen(name, VL_OREADER | opts, cmp))){
+ pdperror(name);
+ return 1;
+ }
+ if(lb){
+ if(!(vals = vlgetlist(villa, kbuf, ksiz))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ while((vbuf = cblistshift(vals, &vsiz)) != NULL){
+ if(ox){
+ printobjhex(vbuf, vsiz);
+ } else {
+ printobj(vbuf, vsiz);
+ }
+ free(vbuf);
+ putchar('\n');
+ }
+ cblistclose(vals);
+ } else {
+ if(!(vbuf = vlget(villa, kbuf, ksiz, &vsiz))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ if(ox){
+ printobjhex(vbuf, vsiz);
+ } else {
+ printobj(vbuf, vsiz);
+ }
+ free(vbuf);
+ if(!nb) putchar('\n');
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform list command */
+int dolist(const char *name, int opts, const char *tbuf, int tsiz, const char *bbuf, int bsiz,
+ VLCFUNC cmp, int ki, int kb, int vb, int ox, int gt, int lt, int max, int desc){
+ VILLA *villa;
+ char *kbuf, *vbuf;
+ int ksiz, vsiz, show, rv;
+ if(!(villa = vlopen(name, VL_OREADER | opts, cmp))){
+ pdperror(name);
+ return 1;
+ }
+ if(max < 0) max = vlrnum(villa);
+ if(desc){
+ if(bbuf){
+ vlcurjump(villa, bbuf, bsiz, VL_JBACKWARD);
+ } else {
+ vlcurlast(villa);
+ }
+ show = 0;
+ while(show < max && (kbuf = vlcurkey(villa, &ksiz)) != NULL){
+ if(bbuf && lt){
+ if(cmp(kbuf, ksiz, bbuf, bsiz) == 0){
+ free(kbuf);
+ vlcurnext(villa);
+ continue;
+ }
+ lt = FALSE;
+ }
+ if(tbuf){
+ rv = cmp(kbuf, ksiz, tbuf, tsiz);
+ if(rv < 0 || (gt && rv == 0)){
+ free(kbuf);
+ break;
+ }
+ }
+ if(!(vbuf = vlcurval(villa, &vsiz))){
+ free(kbuf);
+ break;
+ }
+ if(ox){
+ if(!vb) printobjhex(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobjhex(vbuf, vsiz);
+ } else {
+ if(!vb) printobj(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobj(vbuf, vsiz);
+ }
+ putchar('\n');
+ free(kbuf);
+ free(vbuf);
+ show++;
+ vlcurprev(villa);
+ }
+ } else {
+ if(tbuf){
+ vlcurjump(villa, tbuf, tsiz, VL_JFORWARD);
+ } else {
+ vlcurfirst(villa);
+ }
+ show = 0;
+ while(show < max && (kbuf = vlcurkey(villa, &ksiz)) != NULL){
+ if(tbuf && gt){
+ if(cmp(kbuf, ksiz, tbuf, tsiz) == 0){
+ free(kbuf);
+ vlcurnext(villa);
+ continue;
+ }
+ gt = FALSE;
+ }
+ if(bbuf){
+ rv = cmp(kbuf, ksiz, bbuf, bsiz);
+ if(rv > 0 || (lt && rv == 0)){
+ free(kbuf);
+ break;
+ }
+ }
+ if(!(vbuf = vlcurval(villa, &vsiz))){
+ free(kbuf);
+ break;
+ }
+ if(ox){
+ if(!vb) printobjhex(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobjhex(vbuf, vsiz);
+ } else {
+ if(!vb) printobj(kbuf, ksiz);
+ if(!kb && !vb) putchar('\t');
+ if(!kb) printobj(vbuf, vsiz);
+ }
+ putchar('\n');
+ free(kbuf);
+ free(vbuf);
+ show++;
+ vlcurnext(villa);
+ }
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform optimize command */
+int dooptimize(const char *name){
+ VILLA *villa;
+ if(!(villa = vlopen(name, VL_OWRITER, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ if(!vloptimize(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform inform command */
+int doinform(const char *name, int opts){
+ VILLA *villa;
+ char *tmp;
+ if(!(villa = vlopen(name, VL_OREADER | opts, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ tmp = vlname(villa);
+ printf("name: %s\n", tmp ? tmp : "(null)");
+ free(tmp);
+ printf("file size: %d\n", vlfsiz(villa));
+ printf("leaf nodes: %d\n", vllnum(villa));
+ printf("non-leaf nodes: %d\n", vlnnum(villa));
+ printf("records: %d\n", vlrnum(villa));
+ printf("inode number: %d\n", vlinode(villa));
+ printf("modified time: %.0f\n", (double)vlmtime(villa));
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform remove command */
+int doremove(const char *name){
+ if(!vlremove(name)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform repair command */
+int dorepair(const char *name, VLCFUNC cmp){
+ if(!vlrepair(name, cmp)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform exportdb command */
+int doexportdb(const char *name, const char *file, VLCFUNC cmp){
+ VILLA *villa;
+ if(!(villa = vlopen(name, VL_OREADER, cmp))){
+ pdperror(name);
+ return 1;
+ }
+ if(!vlexportdb(villa, file)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+/* perform importdb command */
+int doimportdb(const char *name, const char *file, VLCFUNC cmp){
+ VILLA *villa;
+ if(!(villa = vlopen(name, VL_OWRITER | VL_OCREAT | VL_OTRUNC, cmp))){
+ pdperror(name);
+ return 1;
+ }
+ if(!vlimportdb(villa, file)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/vlmgr.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/vltest.c
===================================================================
--- box/trunk/qdbm/vltest.c (rev 0)
+++ box/trunk/qdbm/vltest.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,1507 @@
+/*************************************************************************************************
+ * Test cases of Villa
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdarg.h>
+#include <limits.h>
+#include <time.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+#define RECBUFSIZ 32 /* buffer for records */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runwrite(int argc, char **argv);
+int runread(int argc, char **argv);
+int runrdup(int argc, char **argv);
+int runcombo(int argc, char **argv);
+int runwicked(int argc, char **argv);
+int printfflush(const char *format, ...);
+void pdperror(const char *name);
+int myrand(void);
+int dowrite(const char *name, int rnum, int ii, int cmode,
+ int lrecmax, int nidxmax, int lcnum, int ncnum, int fbp);
+int doread(const char *name, int ii, int vc);
+int dordup(const char *name, int rnum, int pnum, int ii, int cmode, int cc,
+ int lrecmax, int nidxmax, int lcnum, int ncnum, int fbp);
+int docombo(const char *name, int cmode);
+int dowicked(const char *name, int rnum, int cb, int cmode);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ char *env;
+ int rv;
+ cbstdiobin();
+ if((env = getenv("QDBMDBGFD")) != NULL) dpdbgfd = atoi(env);
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "write")){
+ rv = runwrite(argc, argv);
+ } else if(!strcmp(argv[1], "read")){
+ rv = runread(argc, argv);
+ } else if(!strcmp(argv[1], "rdup")){
+ rv = runrdup(argc, argv);
+ } else if(!strcmp(argv[1], "combo")){
+ rv = runcombo(argc, argv);
+ } else if(!strcmp(argv[1], "wicked")){
+ rv = runwicked(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: test cases for Villa\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s write [-int] [-cz|-cy|-cx] [-tune lrecmax nidxmax lcnum ncnum]"
+ " [-fbp num] name rnum\n", progname);
+ fprintf(stderr, " %s read [-int] [-vc] name\n", progname);
+ fprintf(stderr, " %s rdup [-int] [-cz|-cy|-cx] [-cc] [-tune lrecmax nidxmax lcnum ncnum]"
+ " [-fbp num] name rnum pnum\n", progname);
+ fprintf(stderr, " %s combo [-cz|-cy|-cx] name\n", progname);
+ fprintf(stderr, " %s wicked [-c] [-cz|-cy|-cx] name rnum\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of write command */
+int runwrite(int argc, char **argv){
+ char *name, *rstr;
+ int i, rnum, ii, cmode, lrecmax, nidxmax, lcnum, ncnum, fbp, rv;
+ name = NULL;
+ rstr = NULL;
+ rnum = 0;
+ ii = FALSE;
+ cmode = 0;
+ lrecmax = -1;
+ nidxmax = -1;
+ lcnum = -1;
+ ncnum = -1;
+ fbp = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-int")){
+ ii = TRUE;
+ } else if(!strcmp(argv[i], "-cz")){
+ cmode |= VL_OZCOMP;
+ } else if(!strcmp(argv[i], "-cy")){
+ cmode |= VL_OYCOMP;
+ } else if(!strcmp(argv[i], "-cx")){
+ cmode |= VL_OXCOMP;
+ } else if(!strcmp(argv[i], "-tune")){
+ if(++i >= argc) usage();
+ lrecmax = atoi(argv[i]);
+ if(++i >= argc) usage();
+ nidxmax = atoi(argv[i]);
+ if(++i >= argc) usage();
+ lcnum = atoi(argv[i]);
+ if(++i >= argc) usage();
+ ncnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-fbp")){
+ if(++i >= argc) usage();
+ fbp = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowrite(name, rnum, ii, cmode, lrecmax, nidxmax, lcnum, ncnum, fbp);
+ return rv;
+}
+
+
+/* parse arguments of read command */
+int runread(int argc, char **argv){
+ char *name;
+ int i, ii, vc, rv;
+ name = NULL;
+ ii = FALSE;
+ vc = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-int")){
+ ii = TRUE;
+ } else if(!strcmp(argv[i], "-vc")){
+ vc = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doread(name, ii, vc);
+ return rv;
+}
+
+
+/* parse arguments of rdup command */
+int runrdup(int argc, char **argv){
+ char *name, *rstr, *pstr;
+ int i, rnum, pnum, ii, cmode, cc, lrecmax, nidxmax, lcnum, ncnum, fbp, rv;
+ name = NULL;
+ rstr = NULL;
+ pstr = NULL;
+ rnum = 0;
+ pnum = 0;
+ ii = FALSE;
+ cmode = 0;
+ cc = FALSE;
+ lrecmax = -1;
+ nidxmax = -1;
+ lcnum = -1;
+ ncnum = -1;
+ fbp = -1;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-int")){
+ ii = TRUE;
+ } else if(!strcmp(argv[i], "-cz")){
+ cmode |= VL_OZCOMP;
+ } else if(!strcmp(argv[i], "-cy")){
+ cmode |= VL_OYCOMP;
+ } else if(!strcmp(argv[i], "-cx")){
+ cmode |= VL_OXCOMP;
+ } else if(!strcmp(argv[i], "-cc")){
+ cc = TRUE;
+ } else if(!strcmp(argv[i], "-tune")){
+ if(++i >= argc) usage();
+ lrecmax = atoi(argv[i]);
+ if(++i >= argc) usage();
+ nidxmax = atoi(argv[i]);
+ if(++i >= argc) usage();
+ lcnum = atoi(argv[i]);
+ if(++i >= argc) usage();
+ ncnum = atoi(argv[i]);
+ } else if(!strcmp(argv[i], "-fbp")){
+ if(++i >= argc) usage();
+ fbp = atoi(argv[i]);
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else if(!pstr){
+ pstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr || !pstr) usage();
+ rnum = atoi(rstr);
+ pnum = atoi(pstr);
+ if(rnum < 1 || pnum < 1) usage();
+ rv = dordup(name, rnum, pnum, ii, cmode, cc, lrecmax, nidxmax, lcnum, ncnum, fbp);
+ return rv;
+}
+
+
+/* parse arguments of combo command */
+int runcombo(int argc, char **argv){
+ char *name;
+ int i, cmode, rv;
+ name = NULL;
+ cmode = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-cz")){
+ cmode |= VL_OZCOMP;
+ } else if(!strcmp(argv[i], "-cy")){
+ cmode |= VL_OYCOMP;
+ } else if(!strcmp(argv[i], "-cx")){
+ cmode |= VL_OXCOMP;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = docombo(name, cmode);
+ return rv;
+}
+
+
+/* parse arguments of wicked command */
+int runwicked(int argc, char **argv){
+ char *name, *rstr;
+ int i, cb, cmode, rnum, rv;
+ name = NULL;
+ rstr = NULL;
+ cb = FALSE;
+ cmode = 0;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-c")){
+ cb = TRUE;
+ } else if(!strcmp(argv[i], "-cz")){
+ cmode |= VL_OZCOMP;
+ } else if(!strcmp(argv[i], "-cy")){
+ cmode |= VL_OYCOMP;
+ } else if(!strcmp(argv[i], "-cx")){
+ cmode |= VL_OXCOMP;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else if(!rstr){
+ rstr = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name || !rstr) usage();
+ rnum = atoi(rstr);
+ if(rnum < 1) usage();
+ rv = dowicked(name, rnum, cb, cmode);
+ return rv;
+}
+
+
+/* print formatted string and flush the buffer */
+int printfflush(const char *format, ...){
+ va_list ap;
+ int rv;
+ va_start(ap, format);
+ rv = vprintf(format, ap);
+ if(fflush(stdout) == EOF) rv = -1;
+ va_end(ap);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* pseudo random number generator */
+int myrand(void){
+ static int cnt = 0;
+ if(cnt == 0) srand(time(NULL));
+ return (rand() * rand() + (rand() >> (sizeof(int) * 4)) + (cnt++)) & INT_MAX;
+}
+
+
+/* perform write command */
+int dowrite(const char *name, int rnum, int ii, int cmode,
+ int lrecmax, int nidxmax, int lcnum, int ncnum, int fbp){
+ VILLA *villa;
+ int i, omode, err, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Writing Test>\n name=%s rnum=%d int=%d cmode=%d "
+ "lrecmax=%d nidxmax=%d lcnum=%d ncnum=%d fbp=%d\n\n",
+ name, rnum, ii, cmode, lrecmax, nidxmax, lcnum, ncnum, fbp);
+ /* open a database */
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC | cmode;
+ if(!(villa = vlopen(name, omode, ii ? VL_CMPINT : VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ err = FALSE;
+ /* set tuning parameters */
+ if(lrecmax > 0) vlsettuning(villa, lrecmax, nidxmax, lcnum, ncnum);
+ if(fbp >= 0) vlsetfbpsiz(villa, fbp);
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* store a record */
+ if(ii){
+ if(!vlput(villa, (char *)&i, sizeof(int), (char *)&i, sizeof(int), VL_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ len = sprintf(buf, "%08d", i);
+ if(!vlput(villa, buf, len, buf, len, VL_DOVER)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return 0;
+}
+
+
+/* perform read command */
+int doread(const char *name, int ii, int vc){
+ VILLA *villa;
+ int i, rnum, err, len;
+ const char *cval;
+ char buf[RECBUFSIZ], *val;
+ printfflush("<Reading Test>\n name=%s int=%d\n\n", name, ii);
+ /* open a database */
+ if(!(villa = vlopen(name, VL_OREADER, ii ? VL_CMPINT : VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ /* get the number of records */
+ rnum = vlrnum(villa);
+ err = FALSE;
+ /* loop for each record */
+ for(i = 1; i <= rnum; i++){
+ /* retrieve a record */
+ if(ii){
+ if(vc){
+ if(!(cval = vlgetcache(villa, (char *)&i, sizeof(int), NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ if(!(val = vlget(villa, (char *)&i, sizeof(int), NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ }
+ } else {
+ len = sprintf(buf, "%08d", i);
+ if(vc){
+ if(!(cval = vlgetcache(villa, buf, len, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ if(!(val = vlget(villa, buf, len, NULL))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ free(val);
+ }
+ }
+ /* print progression */
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d)\n", i);
+ }
+ }
+ }
+ /* close the database */
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return 0;
+}
+
+
+/* perform rdup command */
+int dordup(const char *name, int rnum, int pnum, int ii, int cmode, int cc,
+ int lrecmax, int nidxmax, int lcnum, int ncnum, int fbp){
+ VILLA *villa;
+ int i, omode, err, dmode, vi, len;
+ char buf[RECBUFSIZ];
+ printfflush("<Random Writing Test>\n name=%s rnum=%d int=%d cmode=%d "
+ "lrecmax=%d nidxmax=%d lcnum=%d ncnum=%d fbp=%d\n\n",
+ name, rnum, ii, cmode, lrecmax, nidxmax, lcnum, ncnum, fbp);
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC | cmode;
+ if(!(villa = vlopen(name, omode, ii ? VL_CMPINT : VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ err = FALSE;
+ if(lrecmax > 0) vlsettuning(villa, lrecmax, nidxmax, lcnum, ncnum);
+ if(fbp >= 0) vlsetfbpsiz(villa, fbp);
+ for(i = 1; i <= rnum; i++){
+ dmode = i % 3 == 0 ? VL_DDUPR : VL_DDUP;
+ if(cc && myrand() % 2 == 0) dmode = VL_DCAT;
+ vi = myrand() % pnum + 1;
+ if(ii){
+ if(!vlput(villa, (char *)&vi, sizeof(int), (char *)&vi, sizeof(int), dmode)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ } else {
+ len = sprintf(buf, "%08d", vi);
+ if(!vlput(villa, buf, len, buf, len, dmode)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ }
+ if(rnum > 250 && i % (rnum / 250) == 0){
+ putchar('.');
+ fflush(stdout);
+ if(i == rnum || i % (rnum / 10) == 0){
+ printfflush(" (%08d: fsiz=%d lnum=%d nnum=%d)\n",
+ i, vlfsiz(villa), vllnum(villa), vlnnum(villa));
+ }
+ }
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return 0;
+}
+
+
+/* perform combo command */
+int docombo(const char *name, int cmode){
+ VILLA *villa;
+ VLMULCUR **mulcurs;
+ char buf[RECBUFSIZ], *vbuf, *kbuf;
+ int i, j, omode, len, vsiz, ksiz, fsiz, lnum, nnum, rnum;
+ CBLIST *alist, *dlist;
+ const char *ap, *dp;
+ printfflush("<Combination Test>\n name=%s cmode=%d\n\n", name, cmode);
+ printfflush("Creating a database with VL_CMPLEX ... ");
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC | cmode;
+ if(!(villa = vlopen(name, omode, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Setting tuning parameters with 3, 4, 16, 16 ... ");
+ vlsettuning(villa, 3, 4, 16, 16);
+ printfflush("ok\n");
+ printfflush("Adding 100 records with VL_DOVER ... ");
+ for(i = 1; i <= 100; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!vlput(villa, buf, len, buf, len, VL_DOVER)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 100; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = vlget(villa, buf, len, &vsiz))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != 8 || vlvsiz(villa, buf, len) != 8){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ if(vlvnum(villa, buf, len) != 1){
+ fprintf(stderr, "%s: %s: invalid vnum\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting x1 - x5 records ... ");
+ for(i = 1; i <= 100; i++){
+ if(i % 10 < 1 || i % 10 > 5) continue;
+ len = sprintf(buf, "%08d", i);
+ if(!vlout(villa, buf, len)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Adding 100 records with VL_DOVER ... ");
+ for(i = 1; i <= 100; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!vlput(villa, buf, len, buf, len, VL_DOVER)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting x1 - x5 records ... ");
+ for(i = 1; i <= 100; i++){
+ if(i % 10 < 1 || i % 10 > 5) continue;
+ len = sprintf(buf, "%08d", i);
+ if(!vlout(villa, buf, len)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Checking number of records ... ");
+ if(vlrnum(villa) != 50){
+ fprintf(stderr, "%s: %s: invalid rnum\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 100 records with VL_DDUP ... ");
+ for(i = 1; i <= 100; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!vlput(villa, buf, len, buf, len, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting x6 - x0 records ... ");
+ for(i = 1; i <= 100; i++){
+ if(i % 10 >= 1 && i % 10 <= 5) continue;
+ len = sprintf(buf, "%08d", i);
+ if(!vlout(villa, buf, len)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Optimizing the database ... ");
+ if(!vloptimize(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Checking number of records ... ");
+ if(vlrnum(villa) != 100){
+ fprintf(stderr, "%s: %s: invalid rnum\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Checking records ... ");
+ for(i = 1; i <= 100; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!(vbuf = vlget(villa, buf, len, &vsiz))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ if(vsiz != 8){
+ fprintf(stderr, "%s: %s: invalid vsiz\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ if(vlvnum(villa, buf, len) != 1){
+ fprintf(stderr, "%s: %s: invalid vnum\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting x6 - x0 records ... ");
+ for(i = 1; i <= 100; i++){
+ if(i % 10 >= 1 && i % 10 <= 5) continue;
+ len = sprintf(buf, "%08d", i);
+ if(!vlout(villa, buf, len)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Scanning with the cursor in ascending order ... ");
+ if(!vlcurfirst(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ i = 0;
+ do {
+ kbuf = NULL;
+ vbuf = NULL;
+ if(!(kbuf = vlcurkey(villa, &ksiz)) || !(vbuf = vlcurval(villa, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ free(vbuf);
+ vlclose(villa);
+ return 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ i++;
+ } while(vlcurnext(villa));
+ if(i != 50){
+ fprintf(stderr, "%s: %s: invalid cursor\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Scanning with the cursor in decending order ... ");
+ if(!vlcurlast(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ i = 0;
+ do {
+ kbuf = NULL;
+ vbuf = NULL;
+ if(!(kbuf = vlcurkey(villa, &ksiz)) || !(vbuf = vlcurval(villa, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ free(vbuf);
+ vlclose(villa);
+ return 1;
+ }
+ free(kbuf);
+ free(vbuf);
+ i++;
+ } while(vlcurprev(villa));
+ if(i != 50){
+ fprintf(stderr, "%s: %s: invalid cursor\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 50 random records with VL_DDUPR ... ");
+ for(i = 0; i < 50; i++){
+ len = sprintf(buf, "%08d", myrand() % 100 + 1);
+ if(!vlput(villa, buf, len, buf, len, VL_DDUPR)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Deleting 80 random records ... ");
+ i = 0;
+ while(i < 80){
+ len = sprintf(buf, "%08d", myrand() % 100 + 1);
+ if(!vlout(villa, buf, len)){
+ if(dpecode == DP_ENOITEM) continue;
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ i++;
+ }
+ printfflush("ok\n");
+ alist = cblistopen();
+ dlist = cblistopen();
+ printfflush("Scanning with the cursor in ascending order ... ");
+ if(!vlcurfirst(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ i = 0;
+ do {
+ kbuf = NULL;
+ vbuf = NULL;
+ if(!(kbuf = vlcurkey(villa, &ksiz)) || !(vbuf = vlcurval(villa, &vsiz))){
+ pdperror(name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ free(kbuf);
+ free(vbuf);
+ vlclose(villa);
+ return 1;
+ }
+ cblistpush(alist, kbuf, ksiz);
+ free(kbuf);
+ free(vbuf);
+ i++;
+ } while(vlcurnext(villa));
+ if(i != 20){
+ fprintf(stderr, "%s: %s: invalid cursor\n", progname, name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Scanning with the cursor in decending order ... ");
+ if(!vlcurlast(villa)){
+ pdperror(name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ i = 0;
+ do {
+ kbuf = NULL;
+ vbuf = NULL;
+ if(!(kbuf = vlcurkey(villa, &ksiz)) || !(vbuf = vlcurval(villa, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ free(vbuf);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ cblistunshift(dlist, kbuf, ksiz);
+ free(kbuf);
+ free(vbuf);
+ i++;
+ } while(vlcurprev(villa));
+ if(i != 20){
+ fprintf(stderr, "%s: %s: invalid cursor\n", progname, name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Matching result of ascending scan and desending scan ... ");
+ for(i = 0; i < cblistnum(alist); i++){
+ ap = cblistval(alist, i, NULL);
+ dp = cblistval(dlist, i, NULL);
+ if(strcmp(ap, dp)){
+ fprintf(stderr, "%s: %s: not match\n", progname, name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ cblistsort(alist);
+ for(i = 0; i < cblistnum(alist); i++){
+ ap = cblistval(alist, i, NULL);
+ dp = cblistval(dlist, i, NULL);
+ if(strcmp(ap, dp)){
+ fprintf(stderr, "%s: %s: not match\n", progname, name);
+ cblistclose(alist);
+ cblistclose(dlist);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ cblistclose(alist);
+ cblistclose(dlist);
+ printfflush("Resetting tuning parameters with 41, 80, 32, 32 ... ");
+ vlsettuning(villa, 41, 80, 32, 32);
+ printfflush("ok\n");
+ printfflush("Adding 1000 random records with VL_DDUP ... ");
+ for(i = 0; i < 1000; i++){
+ len = sprintf(buf, "%08d", myrand() % 1000 + 1);
+ if(!vlput(villa, buf, len, buf, len, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Resetting tuning parameters with 8, 5, 16, 16 ... ");
+ vlsettuning(villa, 8, 5, 16, 16);
+ printfflush("ok\n");
+ printfflush("Adding 1000 random records with VL_DDUP ... ");
+ for(i = 0; i < 1000; i++){
+ len = sprintf(buf, "%08d", myrand() % 1000 + 1);
+ if(!vlput(villa, buf, len, buf, len, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Beginning the transaction ... ");
+ if(!vltranbegin(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 100 random records with VL_DDUP ... ");
+ for(i = 0; i < 100; i++){
+ len = sprintf(buf, "%08d", myrand() % 1000 + 1);
+ if(!vlput(villa, buf, len, buf, len, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Scanning and checking ... ");
+ i = 0;
+ for(vlcurlast(villa); (kbuf = vlcurkey(villa, &ksiz)) != NULL; vlcurprev(villa)){
+ if(vlvnum(villa, kbuf, ksiz) < 1 || !(vbuf = vlcurval(villa, NULL))){
+ pdperror(name);
+ free(kbuf);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ free(kbuf);
+ i++;
+ }
+ if(i != vlrnum(villa)){
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Committing the transaction ... ");
+ if(!vltrancommit(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Scanning and checking ... ");
+ i = 0;
+ for(vlcurlast(villa); (kbuf = vlcurkey(villa, &ksiz)) != NULL; vlcurprev(villa)){
+ if(vlvnum(villa, kbuf, ksiz) < 1 || !(vbuf = vlcurval(villa, NULL))){
+ pdperror(name);
+ free(kbuf);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ free(kbuf);
+ i++;
+ }
+ if(i != vlrnum(villa)){
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ lnum = vllnum(villa);
+ nnum = vlnnum(villa);
+ rnum = vlrnum(villa);
+ fsiz = vlfsiz(villa);
+ printfflush("Beginning the transaction ... ");
+ if(!vltranbegin(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 100 random records with VL_DDUP ... ");
+ for(i = 0; i < 100; i++){
+ len = sprintf(buf, "%08d", myrand() % 1000 + 1);
+ if(!vlput(villa, buf, len, buf, len, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Aborting the transaction ... ");
+ if(!vltranabort(villa)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Checking rollback ... ");
+ if(vlfsiz(villa) != fsiz || vllnum(villa) != lnum ||
+ vlnnum(villa) != nnum || vlrnum(villa) != rnum){
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Scanning and checking ... ");
+ i = 0;
+ for(vlcurlast(villa); (kbuf = vlcurkey(villa, &ksiz)) != NULL; vlcurprev(villa)){
+ if(vlvnum(villa, kbuf, ksiz) < 1 || !(vbuf = vlcurval(villa, NULL))){
+ pdperror(name);
+ free(kbuf);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ free(kbuf);
+ i++;
+ }
+ if(i != vlrnum(villa)){
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Creating a database with VL_CMPLEX ... ");
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC | cmode;
+ if(!(villa = vlopen(name, omode, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Setting tuning parameters with 5, 6, 16, 16 ... ");
+ vlsettuning(villa, 5, 6, 16, 16);
+ printfflush("ok\n");
+ printfflush("Adding 3 * 3 records with VL_DDUP ... ");
+ for(i = 0; i < 3; i++){
+ for(j = 0; j < 3; j++){
+ len = sprintf(buf, "%08d", j);
+ if(!vlput(villa, buf, len, buf, -1, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Inserting records with the cursor ... ");
+ if(!vlcurjump(villa, "00000001", -1, VL_JFORWARD) ||
+ !vlcurput(villa, "first", -1, VL_CPAFTER) || !vlcurput(villa, "second", -1, VL_CPAFTER) ||
+ !vlcurnext(villa) ||
+ !vlcurput(villa, "third", -1, VL_CPAFTER) ||
+ strcmp(vlcurvalcache(villa, NULL), "third") ||
+ !vlcurput(villa, "fourth", -1, VL_CPCURRENT) ||
+ strcmp(vlcurvalcache(villa, NULL), "fourth") ||
+ !vlcurjump(villa, "00000001", -1, VL_JFORWARD) ||
+ strcmp(vlcurvalcache(villa, NULL), "00000001") ||
+ !vlcurput(villa, "one", -1, VL_CPBEFORE) || !vlcurput(villa, "two", -1, VL_CPBEFORE) ||
+ !vlcurput(villa, "three", -1, VL_CPBEFORE) || !vlcurput(villa, "five", -1, VL_CPBEFORE) ||
+ !vlcurnext(villa) ||
+ !vlcurput(villa, "four", -1, VL_CPBEFORE) ||
+ strcmp(vlcurvalcache(villa, NULL), "four") ||
+ !vlcurjump(villa, "00000001*", -1, VL_JBACKWARD) ||
+ strcmp(vlcurvalcache(villa, NULL), "00000001") ||
+ !vlcurput(villa, "omega", -1, VL_CPAFTER) ||
+ strcmp(vlcurkeycache(villa, NULL), "00000001") ||
+ strcmp(vlcurvalcache(villa, NULL), "omega") ||
+ !vlcurjump(villa, "00000000*", -1, VL_JFORWARD) ||
+ !vlcurput(villa, "alpha", -1, VL_CPBEFORE) ||
+ strcmp(vlcurvalcache(villa, NULL), "alpha") ||
+ !vlcurprev(villa) ||
+ strcmp(vlcurkeycache(villa, NULL), "00000000") ||
+ strcmp(vlcurvalcache(villa, NULL), "00000000") ||
+ !vlcurput(villa, "before", -1, VL_CPAFTER) ||
+ strcmp(vlcurvalcache(villa, NULL), "before") ||
+ !vlcurjump(villa, "00000001*", -1, VL_JFORWARD) ||
+ !vlcurput(villa, "after", -1, VL_CPBEFORE) ||
+ strcmp(vlcurvalcache(villa, NULL), "after") ||
+ !vlcurfirst(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "00000000") ||
+ !vlcurput(villa, "top", -1, VL_CPBEFORE) ||
+ strcmp(vlcurvalcache(villa, NULL), "top") ||
+ !vlcurlast(villa) ||
+ !vlcurput(villa, "bottom", -1, VL_CPAFTER) ||
+ strcmp(vlcurvalcache(villa, NULL), "bottom")){
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Deleting records with the cursor ... ");
+ if(!vlcurjump(villa, "00000000*", -1, VL_JBACKWARD) ||
+ strcmp(vlcurvalcache(villa, NULL), "before") ||
+ !vlcurout(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "alpha") ||
+ !vlcurout(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "five") ||
+ !vlcurfirst(villa) || !vlcurnext(villa) ||
+ !vlcurout(villa) || !vlcurout(villa) || !vlcurout(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "five") ||
+ !vlcurprev(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "top") ||
+ !vlcurout(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "five") ||
+ !vlcurjump(villa, "00000002", -1, VL_JBACKWARD) ||
+ strcmp(vlcurvalcache(villa, NULL), "bottom") ||
+ !vlcurout(villa) ||
+ !vlcurjump(villa, "00000001", -1, VL_JBACKWARD) ||
+ !vlcurout(villa) ||
+ !vlcurout(villa) || !vlcurout(villa) || !vlcurout(villa) ||
+ strcmp(vlcurkeycache(villa, NULL), "00000002") ||
+ strcmp(vlcurvalcache(villa, NULL), "00000002") ||
+ !vlcurout(villa) || vlcurout(villa) ||
+ !vlcurfirst(villa) ||
+ strcmp(vlcurvalcache(villa, NULL), "five")){
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ vlcurfirst(villa);
+ while(vlcurout(villa)){
+ free(vlcurval(villa, NULL));
+ }
+ if(vlrnum(villa) != 0){
+ printf("%d\n", vlrnum(villa));
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ for(i = 0; i < 1000; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!vlput(villa, buf, len, buf, -1, VL_DKEEP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ for(i = 200; i < 800; i++){
+ len = sprintf(buf, "%08d", i);
+ if(!vlout(villa, buf, len)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ vlcurfirst(villa);
+ while(vlcurout(villa)){
+ free(vlcurval(villa, NULL));
+ }
+ if(vlrnum(villa) != 0){
+ printf("%d\n", vlrnum(villa));
+ fprintf(stderr, "%s: %s: invalid\n", progname, name);
+ vlclose(villa);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Adding 3 * 100 records with VL_DDUP ... ");
+ for(i = 1; i <= 100; i++){
+ len = sprintf(buf, "%08d", i);
+ for(j = 0; j < 3; j++){
+ if(!vlput(villa, buf, len, buf, len, VL_DDUP)){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Opening the database as a reader ... ");
+ if(!(villa = vlopen(name, VL_OREADER, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("Opening multiple cursors ... ");
+ mulcurs = cbmalloc(sizeof(VLMULCUR *) * 8);
+ for(i = 0; i < 8; i++){
+ if(!(mulcurs[i] = vlmulcuropen(villa))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Scanning multiple cursors ... ");
+ for(i = 0; i < 8; i++){
+ if(i % 2 == 0){
+ vlmulcurfirst(mulcurs[i]);
+ } else {
+ vlmulcurlast(mulcurs[i]);
+ }
+ }
+ for(i = 0; i < 300; i++){
+ for(j = 0; j < 8; j++){
+ if(j % 2 == 0){
+ if(!(vbuf = vlmulcurkey(mulcurs[j], &vsiz))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ vlmulcurnext(mulcurs[j]);
+ } else {
+ if(!(vbuf = vlmulcurval(mulcurs[j], &vsiz))){
+ pdperror(name);
+ vlclose(villa);
+ return 1;
+ }
+ free(vbuf);
+ vlmulcurprev(mulcurs[j]);
+ }
+ }
+ }
+ printfflush("ok\n");
+ printfflush("Closing multiple cursors ... ");
+ for(i = 0; i < 8; i++){
+ vlmulcurclose(mulcurs[i]);
+ }
+ free(mulcurs);
+ printfflush("ok\n");
+ printfflush("Closing the database ... ");
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ printfflush("ok\n");
+ printfflush("all ok\n\n");
+ return 0;
+}
+
+
+/* perform wicked command */
+int dowicked(const char *name, int rnum, int cb, int cmode){
+ VILLA *villa;
+ CBMAP *map;
+ int i, j, omode, len, err, ksiz, vsiz, tran, mksiz, mvsiz, rsiz;
+ const char *mkbuf, *mvbuf;
+ char buf[32], *kbuf, *vbuf;
+ CBLIST *list;
+ printfflush("<Wicked Writing Test>\n name=%s rnum=%d\n\n", name, rnum);
+ omode = VL_OWRITER | VL_OCREAT | VL_OTRUNC | cmode;
+ if(!(villa = vlopen(name, omode, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ err = FALSE;
+ tran = FALSE;
+ vlsettuning(villa, 5, 10, 64, 64);
+ map = NULL;
+ if(cb) map = cbmapopen();
+ for(i = 1; i <= rnum; i++){
+ len = sprintf(buf, "%08d", myrand() % rnum + 1);
+ switch(cb ? (myrand() % 5) : myrand() % 16){
+ case 0:
+ putchar('O');
+ if(!vlput(villa, buf, len, buf, len, VL_DOVER)) err = TRUE;
+ if(map) cbmapput(map, buf, len, buf, len, TRUE);
+ break;
+ case 1:
+ putchar('K');
+ if(!vlput(villa, buf, len, buf, len, VL_DKEEP) && dpecode != DP_EKEEP) err = TRUE;
+ if(map) cbmapput(map, buf, len, buf, len, FALSE);
+ break;
+ case 2:
+ putchar('C');
+ if(!vlput(villa, buf, len, buf, len, VL_DCAT)) err = TRUE;
+ if(map) cbmapputcat(map, buf, len, buf, len);
+ break;
+ case 3:
+ putchar('D');
+ if(!vlout(villa, buf, len) && dpecode != DP_ENOITEM) err = TRUE;
+ if(map) cbmapout(map, buf, len);
+ break;
+ case 4:
+ putchar('G');
+ if((vbuf = vlget(villa, buf, len, NULL)) != NULL){
+ free(vbuf);
+ } else if(dpecode != DP_ENOITEM){
+ err = TRUE;
+ }
+ break;
+ case 5:
+ putchar('V');
+ if(vlvsiz(villa, buf, len) < 0 && dpecode != DP_ENOITEM) err = TRUE;
+ if(!vlvnum(villa, buf, len) && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 6:
+ putchar('X');
+ list = cblistopen();
+ cblistpush(list, buf, len);
+ cblistpush(list, buf, len);
+ if(!vlputlist(villa, buf, len, list)) err = TRUE;
+ cblistclose(list);
+ break;
+ case 7:
+ putchar('Y');
+ if(!vloutlist(villa, buf, len) && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 8:
+ putchar('Z');
+ if((list = vlgetlist(villa, buf, len)) != NULL){
+ cblistclose(list);
+ } else if(dpecode != DP_ENOITEM){
+ err = TRUE;
+ }
+ if((vbuf = vlgetcat(villa, buf, len, NULL)) != NULL){
+ free(vbuf);
+ } else if(dpecode != DP_ENOITEM){
+ err = TRUE;
+ }
+ break;
+ case 9:
+ putchar('Q');
+ if(vlcurjump(villa, buf, len, VL_JFORWARD)){
+ for(j = 0; j < 3 && (kbuf = vlcurkey(villa, &ksiz)); j++){
+ if(VL_CMPLEX(buf, len, kbuf, ksiz) > 0) err = TRUE;
+ if(strcmp(vlcurkeycache(villa, NULL), kbuf)) err = TRUE;
+ if((vbuf = vlcurval(villa, &vsiz)) != NULL){
+ if(strcmp(vlcurvalcache(villa, NULL), vbuf)) err = TRUE;
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ if(!vlcurnext(villa) && dpecode != DP_ENOITEM) err = TRUE;
+ }
+ } else {
+ if(dpecode != DP_ENOITEM) err = TRUE;
+ }
+ break;
+ case 10:
+ putchar('W');
+ if(vlcurjump(villa, buf, len, VL_JBACKWARD)){
+ for(j = 0; j < 3 && (kbuf = vlcurkey(villa, &ksiz)); j++){
+ if(VL_CMPLEX(buf, len, kbuf, ksiz) < 0) err = TRUE;
+ if(strcmp(vlcurkeycache(villa, NULL), kbuf)) err = TRUE;
+ if((vbuf = vlcurval(villa, &vsiz)) != NULL){
+ if(strcmp(vlcurvalcache(villa, NULL), vbuf)) err = TRUE;
+ free(vbuf);
+ } else {
+ err = TRUE;
+ }
+ free(kbuf);
+ if(!vlcurprev(villa) && dpecode != DP_ENOITEM) err = TRUE;
+ }
+ } else {
+ if(dpecode != DP_ENOITEM) err = TRUE;
+ }
+ break;
+ case 11:
+ putchar('L');
+ if(myrand() % 3 == 0 &&
+ !vlcurjump(villa, buf, len, i % 3 == 0 ? VL_JFORWARD : VL_JBACKWARD) &&
+ dpecode != DP_ENOITEM) err = TRUE;
+ for(j = myrand() % 5; j >= 0; j--){
+ switch(myrand() % 6){
+ case 0:
+ if(!vlcurput(villa, buf, len, VL_CPAFTER) && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 1:
+ if(!vlcurput(villa, buf, len, VL_CPBEFORE) && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ case 2:
+ if(!vlcurput(villa, buf, len, VL_CPCURRENT) && dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ default:
+ if(!vlcurout(villa)){
+ if(dpecode != DP_ENOITEM) err = TRUE;
+ break;
+ }
+ break;
+ }
+ }
+ break;
+ case 12:
+ if(tran ? myrand() % 32 != 0 : myrand() % 1024 != 0){
+ putchar('N');
+ break;
+ }
+ putchar('T');
+ if(tran){
+ if(myrand() % 5 == 0){
+ if(!vltranabort(villa)) err = TRUE;
+ } else {
+ if(!vltrancommit(villa)) err = TRUE;
+ }
+ tran = FALSE;
+ } else {
+ if(!vltranbegin(villa)) err = TRUE;
+ tran = TRUE;
+ }
+ break;
+ default:
+ putchar('P');
+ if(!vlput(villa, buf, len, buf, len, myrand() % 3 == 0 ? VL_DDUPR : VL_DDUP)) err = TRUE;
+ break;
+ }
+ if(i % 50 == 0) printfflush(" (%08d)\n", i);
+ if(err){
+ pdperror(name);
+ break;
+ }
+ }
+ if(tran){
+ if(!vltranabort(villa)) err = TRUE;
+ }
+ if(!vloptimize(villa)){
+ pdperror(name);
+ err = TRUE;
+ }
+ if((rnum = vlrnum(villa)) == -1){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(!vlcurfirst(villa)){
+ pdperror(name);
+ err = TRUE;
+ }
+ i = 0;
+ do {
+ kbuf = NULL;
+ vbuf = NULL;
+ if(!(kbuf = vlcurkey(villa, &ksiz)) || !(vbuf = vlcurval(villa, &vsiz)) ||
+ ksiz != 8 || vsiz % 8 != 0 || vlvnum(villa, kbuf, ksiz) < 1){
+ pdperror(name);
+ free(kbuf);
+ free(vbuf);
+ err = TRUE;
+ break;
+ }
+ free(kbuf);
+ free(vbuf);
+ i++;
+ } while(vlcurnext(villa));
+ if(i != rnum){
+ fprintf(stderr, "%s: %s: invalid cursor\n", progname, name);
+ err = TRUE;
+ }
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(!vlcurlast(villa)){
+ pdperror(name);
+ err = TRUE;
+ }
+ i = 0;
+ do {
+ kbuf = NULL;
+ vbuf = NULL;
+ if(!(kbuf = vlcurkey(villa, &ksiz)) || !(vbuf = vlcurval(villa, &vsiz)) ||
+ ksiz != 8 || vsiz % 8 != 0 || vlvnum(villa, kbuf, ksiz) < 1){
+ pdperror(name);
+ free(kbuf);
+ free(vbuf);
+ err = TRUE;
+ break;
+ }
+ free(kbuf);
+ free(vbuf);
+ i++;
+ } while(vlcurprev(villa));
+ if(i != rnum){
+ fprintf(stderr, "%s: %s: invalid cursor\n", progname, name);
+ err = TRUE;
+ }
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ if(map){
+ printfflush("Matching records ... ");
+ cbmapiterinit(map);
+ while((mkbuf = cbmapiternext(map, &mksiz)) != NULL){
+ mvbuf = cbmapget(map, mkbuf, mksiz, &mvsiz);
+ if(!(vbuf = vlget(villa, mkbuf, mksiz, &rsiz))){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ if(rsiz != mvsiz || memcmp(vbuf, mvbuf, rsiz)){
+ fprintf(stderr, "%s: %s: unmatched record\n", progname, name);
+ free(vbuf);
+ err = TRUE;
+ break;
+ }
+ free(vbuf);
+ }
+ cbmapclose(map);
+ if(!err) printfflush("ok\n");
+ }
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ if(!err) printfflush("ok\n\n");
+ return err ? 1 : 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/vltest.c
___________________________________________________________________
Added: svn:eol-style
+ native
Added: box/trunk/qdbm/vltsv.c
===================================================================
--- box/trunk/qdbm/vltsv.c (rev 0)
+++ box/trunk/qdbm/vltsv.c 2010-01-09 16:15:40 UTC (rev 2590)
@@ -0,0 +1,261 @@
+/*************************************************************************************************
+ * Mutual converter between a database of Villa and a TSV text
+ * Copyright (C) 2000-2007 Mikio Hirabayashi
+ * This file is part of QDBM, Quick Database Manager.
+ * QDBM is free software; you can redistribute it and/or modify it under the terms of the GNU
+ * Lesser General Public License as published by the Free Software Foundation; either version
+ * 2.1 of the License or any later version. QDBM is distributed in the hope that it will be
+ * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more
+ * details.
+ * You should have received a copy of the GNU Lesser General Public License along with QDBM; if
+ * not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ * 02111-1307 USA.
+ *************************************************************************************************/
+
+
+#include <depot.h>
+#include <cabin.h>
+#include <villa.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+
+#undef TRUE
+#define TRUE 1 /* boolean true */
+#undef FALSE
+#define FALSE 0 /* boolean false */
+
+
+/* for RISC OS */
+#if defined(__riscos__) || defined(__riscos)
+#include <unixlib/local.h>
+int __riscosify_control = __RISCOSIFY_NO_PROCESS;
+#endif
+
+
+/* global variables */
+const char *progname; /* program name */
+
+
+/* function prototypes */
+int main(int argc, char **argv);
+void usage(void);
+int runimport(int argc, char **argv);
+int runexport(int argc, char **argv);
+void pdperror(const char *name);
+char *getl(void);
+int doimport(const char *name, int bin);
+int doexport(const char *name, int bin);
+
+
+/* main routine */
+int main(int argc, char **argv){
+ int rv;
+ cbstdiobin();
+ progname = argv[0];
+ if(argc < 2) usage();
+ rv = 0;
+ if(!strcmp(argv[1], "import")){
+ rv = runimport(argc, argv);
+ } else if(!strcmp(argv[1], "export")){
+ rv = runexport(argc, argv);
+ } else {
+ usage();
+ }
+ return rv;
+}
+
+
+/* print the usage and exit */
+void usage(void){
+ fprintf(stderr, "%s: mutual converter between TSV and Villa database\n", progname);
+ fprintf(stderr, "\n");
+ fprintf(stderr, "usage:\n");
+ fprintf(stderr, " %s import [-bin] name\n", progname);
+ fprintf(stderr, " %s export [-bin] name\n", progname);
+ fprintf(stderr, "\n");
+ exit(1);
+}
+
+
+/* parse arguments of import command */
+int runimport(int argc, char **argv){
+ char *name;
+ int i, bin, rv;
+ name = NULL;
+ bin = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bin")){
+ bin = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doimport(name, bin);
+ return rv;
+}
+
+
+/* parse arguments of export command */
+int runexport(int argc, char **argv){
+ char *name;
+ int i, bin, rv;
+ name = NULL;
+ bin = FALSE;
+ for(i = 2; i < argc; i++){
+ if(!name && argv[i][0] == '-'){
+ if(!strcmp(argv[i], "-bin")){
+ bin = TRUE;
+ } else {
+ usage();
+ }
+ } else if(!name){
+ name = argv[i];
+ } else {
+ usage();
+ }
+ }
+ if(!name) usage();
+ rv = doexport(name, bin);
+ return rv;
+}
+
+
+/* print an error message */
+void pdperror(const char *name){
+ fprintf(stderr, "%s: %s: %s\n", progname, name, dperrmsg(dpecode));
+}
+
+
+/* read a line */
+char *getl(void){
+ char *buf;
+ int c, len, blen;
+ buf = NULL;
+ len = 0;
+ blen = 256;
+ while((c = getchar()) != EOF){
+ if(blen <= len) blen *= 2;
+ buf = cbrealloc(buf, blen + 1);
+ if(c == '\n') c = '\0';
+ buf[len++] = c;
+ if(c == '\0') break;
+ }
+ if(!buf) return NULL;
+ buf[len] = '\0';
+ return buf;
+}
+
+
+/* perform import command */
+int doimport(const char *name, int bin){
+ VILLA *villa;
+ char *buf, *kbuf, *vbuf, *ktmp, *vtmp;
+ int i, err, ktsiz, vtsiz;
+ /* open a database */
+ if(!(villa = vlopen(name, VL_OWRITER | VL_OCREAT, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ /* loop for each line */
+ err = FALSE;
+ for(i = 1; (buf = getl()) != NULL; i++){
+ kbuf = buf;
+ if((vbuf = strchr(buf, '\t')) != NULL){
+ *vbuf = '\0';
+ vbuf++;
+ /* store a record */
+ if(bin){
+ ktmp = cbbasedecode(kbuf, &ktsiz);
+ vtmp = cbbasedecode(vbuf, &vtsiz);
+ if(!vlput(villa, ktmp, ktsiz, vtmp, vtsiz, VL_DDUP)){
+ pdperror(name);
+ err = TRUE;
+ }
+ free(vtmp);
+ free(ktmp);
+ } else {
+ if(!vlput(villa, kbuf, -1, vbuf, -1, VL_DDUP)){
+ pdperror(name);
+ err = TRUE;
+ break;
+ }
+ }
+ } else {
+ fprintf(stderr, "%s: %s: invalid format in line %d\n", progname, name, i);
+ }
+ free(buf);
+ if(err) break;
+ }
+ /* close the database */
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return err ? 1 : 0;
+}
+
+
+/* perform export command */
+int doexport(const char *name, int bin){
+ VILLA *villa;
+ char *kbuf, *vbuf, *tmp;
+ int err, ksiz, vsiz;
+ /* open a database */
+ if(!(villa = vlopen(name, VL_OREADER, VL_CMPLEX))){
+ pdperror(name);
+ return 1;
+ }
+ /* initialize the cursor */
+ vlcurfirst(villa);
+ /* loop for each key */
+ err = FALSE;
+ while((kbuf = vlcurkey(villa, &ksiz)) != NULL){
+ /* retrieve a value with a key */
+ if(!(vbuf = vlcurval(villa, &vsiz))){
+ pdperror(name);
+ free(kbuf);
+ err = TRUE;
+ break;
+ }
+ /* output data */
+ if(bin){
+ tmp = cbbaseencode(kbuf, ksiz);
+ printf("%s\t", tmp);
+ free(tmp);
+ tmp = cbbaseencode(vbuf, vsiz);
+ printf("%s\n", tmp);
+ free(tmp);
+ } else {
+ printf("%s\t%s\n", kbuf, vbuf);
+ }
+ /* free resources */
+ free(vbuf);
+ free(kbuf);
+ /* step the cursor */
+ vlcurnext(villa);
+ }
+ /* check whether all records were retrieved */
+ if(dpecode != DP_ENOITEM){
+ pdperror(name);
+ err = TRUE;
+ }
+ /* close the database */
+ if(!vlclose(villa)){
+ pdperror(name);
+ return 1;
+ }
+ return 0;
+}
+
+
+
+/* END OF FILE */
Property changes on: box/trunk/qdbm/vltsv.c
___________________________________________________________________
Added: svn:eol-style
+ native
More information about the Boxbackup-commit
mailing list