bmf

bmf (Bayesian Mail Filter) 0.9.4 fork + patches
git clone git://git.codemadness.org/bmf
Log | Files | Refs | README | LICENSE

commit 0983b0f64c3e1bf7fa03f2a4060e6f25e9e79cef
Author: Hiltjo Posthuma <hiltjo@codemadness.org>
Date:   Sat, 22 Sep 2018 17:46:14 +0200

import bmf 0.9.4

Diffstat:
AAUTHORS | 4++++
AChangeLog | 95+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
ALICENSE | 340+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
AMakefile.in | 81+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
AREADME | 130+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
ATODO | 4++++
Abmf.1 | 148+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Abmf.c | 339+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Abmf.spec.in | 64++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Abmfconv.1 | 81+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Abmfconv.c | 169+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Aconfig.h | 80+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Aconfigure | 354+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbdb.c | 684+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbdb.h | 61+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbg.c | 302++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbg.h | 35+++++++++++++++++++++++++++++++++++
Adbh.c | 74++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbh.h | 56++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbmysql.c | 545+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbmysql.h | 60++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbtext.c | 591+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Adbtext.h | 53+++++++++++++++++++++++++++++++++++++++++++++++++++++
Afilt.c | 175+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Afilt.h | 31+++++++++++++++++++++++++++++++
Alex.c | 787+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Alex.h | 44++++++++++++++++++++++++++++++++++++++++++++
Astr.c | 78++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Astr.h | 30++++++++++++++++++++++++++++++
Avec.c | 345+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Avec.h | 58++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
31 files changed, 5898 insertions(+), 0 deletions(-)

diff --git a/AUTHORS b/AUTHORS @@ -0,0 +1,4 @@ +# $Id: AUTHORS,v 1.1.1.1 2002/09/30 21:08:29 tommy Exp $ + +Tom Marshall <tommy@home.tig-grr.com> + Initial version diff --git a/ChangeLog b/ChangeLog @@ -0,0 +1,95 @@ +Revision history for bmf: + +0.9.4: 20 Oct 2002 + * Remove X-RBL-Warning from ignored headers. + +0.9.4pre7: 20 Oct 2002 + * Update documentation. + +0.9.4pre6: 20 Oct 2002 + * Move Bayes stuff into its own file. + +0.9.4pre5: 20 Oct 2002 + * Fix NaN exception: if list is empty, use zero for probability. + * Make extrema array (keepers) variable size. Needs more work. + +0.9.4pre4: 19 Oct 2002 + * Add configure section for Darwin. + * Don't use file locking on Darwin, it is not supported. + +0.9.4pre3: 19 Oct 2002 + * Fixup configure script for OSF1. + * which(1) always returns 0 on OSF1, use type(1) instead. + * Add SYSLIBS to the makefile. + * Fix gcc-ism in dbg.c (ptr arithmetic on void*). + * Fix off-by-one in html tag check. + * Fix unaligned access in libdb. + +0.9.4pre2: 18 Oct 2002 + * Fix bug in -d handling for text and libdb. + (Found by Björn Kalkbrenn) + +0.9.4pre1: 17 Oct 2002 + * Autodetect mailbox type and deprecate the -m option. + +0.9.3: 14 Oct 2002 + * Ditch the builtin libdb locks, use fcntl instead. + * Fix memory leak in dbtext. + * Fix some trivial issues with the lexer: + - Be more strict about recognizing IP addresses. + - Do case-insensitive header name comparisons. + * Fix multiple database closure with mbox format. + * Fix a bogus assert in passthrough. + * Add verbose flag (no functionality yet). + * Add heap checking in debug mode. + * Fix bug in -N mode which made it act the same as -S. + * Add X-RBL-Warning to ignored headers. + * Support maildir style folders. + +0.9.2: 12 Oct 2002 + * Fix bug in multiple message registration. + +0.9.1: 12 Oct 2002 + * Improve error reporting and clarify some messages. + * Package preformatted manpage instead of XML. + * Remove single message per invocation restriction. + +0.84: 09 Oct 2002 + * Fix linker flags for autodetected libdb 4.1 in /usr/local on BSD. + +0.84pre3: 07 Oct 2002 + * Yet another libdb api fix. DB->open() in 4.0 is the same as 3.x. + +0.84pre2: 07 Oct 2002 + * Fix bug in dbdb (v1 only) that prevented file locking on FreeBSD. + +0.84pre1: 07 Oct 2002 + * Fix bug in dbtext that caused segfault searching an empty list. + +0.83: 07 Oct 2002 + * Fix bug preventing creation of libdb files when using -n or -s. + * Fix bug in libdb unmergeclose function logic. Users are strongly + encouraged to delete and rebuild word lists if possible. If not + possible, at least export to text and remove entries + that represent unsigned underflow (eg. 4.2 billion). + +0.82: 06 Oct 2002 + * Support GNU style --with-package=path options. + (copied from autoconf output) + * Use mysql_config in configure script. + * Support libdb v1 and v4. + * Find and use BerkeleyDB 4.1 in *BSD. + (thanks to ted@php.net) + * Clarify and robustify argument handling (fixes pr618875). + +0.81: 03 Oct 2002 + * Add configure script and conditionally compile libdb and mysql. + * Add manpage for bmfconv. + * Cleanup manpage for bmf. + * Rearrange makefile a bit. + * Remove -f text options in bmfconv, it is not supported. + * Fix mysql typo that prevented linking. + * Tweak X-Spam headers a bit. + +0.80: 02 Oct 2002 + * Initial release. diff --git a/LICENSE b/LICENSE @@ -0,0 +1,340 @@ + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 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. + + Preamble + + The licenses for most software are designed to take away your +freedom to share and change it. By contrast, the GNU General Public +License is intended to guarantee your freedom to share and change free +software--to make sure the software is free for all its users. This +General Public License applies to most of the Free Software +Foundation's software and to any other program whose authors commit to +using it. (Some other Free Software Foundation software is covered by +the GNU Library General Public License instead.) You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, 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 or use pieces of it +in new free programs; and that you know you can do these things. + + To protect your rights, we need to make restrictions that forbid +anyone to deny you these rights or to ask you to surrender the rights. +These restrictions translate to certain responsibilities for you if you +distribute copies of the software, or if you modify it. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must give the recipients all the rights that +you have. You must make sure that they, too, receive or can get the +source code. And you must show them these terms so they know their +rights. + + We protect your rights with two steps: (1) copyright the software, and +(2) offer you this license which gives you legal permission to copy, +distribute and/or modify the software. + + Also, for each author's protection and ours, we want to make certain +that everyone understands that there is no warranty for this free +software. If the software is modified by someone else and passed on, we +want its recipients to know that what they have is not the original, so +that any problems introduced by others will not reflect on the original +authors' reputations. + + Finally, any free program is threatened constantly by software +patents. We wish to avoid the danger that redistributors of a free +program will individually obtain patent licenses, in effect making the +program proprietary. To prevent this, we have made it clear that any +patent must be licensed for everyone's free use or not licensed at all. + + The precise terms and conditions for copying, distribution and +modification follow. + + GNU GENERAL PUBLIC LICENSE + TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION + + 0. This License applies to any program or other work which contains +a notice placed by the copyright holder saying it may be distributed +under the terms of this General Public License. The "Program", below, +refers to any such program or work, and a "work based on the Program" +means either the Program or any derivative work under copyright law: +that is to say, a work containing the Program or a portion of it, +either verbatim or with modifications and/or translated into another +language. (Hereinafter, translation is included without limitation in +the term "modification".) Each licensee is addressed as "you". + +Activities other than copying, distribution and modification are not +covered by this License; they are outside its scope. The act of +running the Program is not restricted, and the output from the Program +is covered only if its contents constitute a work based on the +Program (independent of having been made by running the Program). +Whether that is true depends on what the Program does. + + 1. You may copy and distribute verbatim copies of the Program's +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 give any other recipients of the Program a copy of this License +along with the Program. + +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 Program or any portion +of it, thus forming a work based on the Program, 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) You must cause the modified files to carry prominent notices + stating that you changed the files and the date of any change. + + b) You must cause any work that you distribute or publish, that in + whole or in part contains or is derived from the Program or any + part thereof, to be licensed as a whole at no charge to all third + parties under the terms of this License. + + c) If the modified program normally reads commands interactively + when run, you must cause it, when started running for such + interactive use in the most ordinary way, to print or display an + announcement including an appropriate copyright notice and a + notice that there is no warranty (or else, saying that you provide + a warranty) and that users may redistribute the program under + these conditions, and telling the user how to view a copy of this + License. (Exception: if the Program itself is interactive but + does not normally print such an announcement, your work based on + the Program is not required to print an announcement.) + +These requirements apply to the modified work as a whole. If +identifiable sections of that work are not derived from the Program, +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 Program, 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 Program. + +In addition, mere aggregation of another work not based on the Program +with the Program (or with a work based on the Program) on a volume of +a storage or distribution medium does not bring the other work under +the scope of this License. + + 3. You may copy and distribute the Program (or a work based on it, +under Section 2) in object code or executable form under the terms of +Sections 1 and 2 above provided that you also do one of the following: + + a) 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; or, + + b) Accompany it with a written offer, valid for at least three + years, to give any third party, for a charge no more than your + cost of physically performing source distribution, a complete + machine-readable copy of the corresponding source code, to be + distributed under the terms of Sections 1 and 2 above on a medium + customarily used for software interchange; or, + + c) Accompany it with the information you received as to the offer + to distribute corresponding source code. (This alternative is + allowed only for noncommercial distribution and only if you + received the program in object code or executable form with such + an offer, in accord with Subsection b above.) + +The source code for a work means the preferred form of the work for +making modifications to it. For an executable work, 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 executable. However, as a +special exception, the source code 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. + +If distribution of executable or 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 counts as +distribution of the source code, even though third parties are not +compelled to copy the source along with the object code. + + 4. You may not copy, modify, sublicense, or distribute the Program +except as expressly provided under this License. Any attempt +otherwise to copy, modify, sublicense or distribute the Program 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. + + 5. 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 Program or its derivative works. These actions are +prohibited by law if you do not accept this License. Therefore, by +modifying or distributing the Program (or any work based on the +Program), you indicate your acceptance of this License to do so, and +all its terms and conditions for copying, distributing or modifying +the Program or works based on it. + + 6. Each time you redistribute the Program (or any work based on the +Program), the recipient automatically receives a license from the +original licensor to copy, distribute or modify the Program 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 to +this License. + + 7. 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 Program at all. For example, if a patent +license would not permit royalty-free redistribution of the Program 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 Program. + +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. + + 8. If the distribution and/or use of the Program is restricted in +certain countries either by patents or by copyrighted interfaces, the +original copyright holder who places the Program 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. + + 9. The Free Software Foundation may publish revised and/or new versions +of the 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 Program +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 Program does not specify a version number of +this License, you may choose any version ever published by the Free Software +Foundation. + + 10. If you wish to incorporate parts of the Program into other free +programs whose distribution conditions are different, 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 + + 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY +FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN +OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES +PROVIDE THE PROGRAM "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 PROGRAM IS WITH YOU. SHOULD THE +PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, +REPAIR OR CORRECTION. + + 12. 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 PROGRAM 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 PROGRAM (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 PROGRAM TO OPERATE WITH ANY OTHER +PROGRAMS), 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 Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. 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 program's name and a brief idea of what it does.> + Copyright (C) <year> <name of author> + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program 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 General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; 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. + +If the program is interactive, make it output a short notice like this +when it starts in an interactive mode: + + Gnomovision version 69, Copyright (C) year name of author + Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, the commands you use may +be called something other than `show w' and `show c'; they could even be +mouse-clicks or menu items--whatever suits your program. + +You should also get your employer (if you work as a programmer) or your +school, if any, to sign a "copyright disclaimer" for the program, if +necessary. Here is a sample; alter the names: + + Yoyodyne, Inc., hereby disclaims all copyright interest in the program + `Gnomovision' (which makes passes at compilers) written by James Hacker. + + <signature of Ty Coon>, 1 April 1989 + Ty Coon, President of Vice + +This General Public License does not permit incorporating your program into +proprietary programs. If your program is a subroutine library, you may +consider it more useful to permit linking proprietary applications with the +library. If this is what you want to do, use the GNU Library General +Public License instead of this License. diff --git a/Makefile.in b/Makefile.in @@ -0,0 +1,81 @@ +# Makefile for bmf + +BINDIR=/usr/bin +MANDIR=/usr/share/man + +VERSION=0.9.4 + +CC=@CC@ +CFLAGS=@CFLAGS@ +LDFLAGS=@LDFLAGS@ +SYSLIBS=@SYSLIBS@ + +# For creating rpm packages +RPMROOT=/usr/src/rpm +RPM = rpm +RPMFLAGS = -ba +ARCH=`arch|sed 's/i[4-9]86/i386/'` + +all: bmf bmfconv + +bmf: bmf.o filt.o dbmysql.o dbdb.o dbtext.o dbh.o lex.o vec.o str.o dbg.o + $(CC) -o $@ bmf.o filt.o dbmysql.o dbdb.o dbtext.o dbh.o lex.o vec.o str.o dbg.o $(LDFLAGS) $(SYSLIBS) + +bmf.o: bmf.c + $(CC) $(CFLAGS) -DPACKAGE=\"bmf\" -DVERSION=\"$(VERSION)\" -c $< + +bmfconv: bmfconv.o dbmysql.o dbdb.o dbtext.o dbh.o vec.o str.o dbg.o + $(CC) -o $@ bmfconv.o dbmysql.o dbdb.o dbtext.o dbh.o vec.o str.o dbg.o $(LDFLAGS) $(SYSLIBS) + +bmfconv.o: bmfconv.c + $(CC) $(CFLAGS) -DPACKAGE=\"bmfconv\" -DVERSION=\"$(VERSION)\" -c $< + +install: checkroot bmf bmf.1 bmfconv bmfconv.1 + [ -d $(DESTDIR)$(BINDIR) ] || mkdir -p $(DESTDIR)$(BINDIR) + [ -d $(DESTDIR)$(MANDIR)/man1 ] || mkdir -p $(DESTDIR)$(MANDIR)/man1 + cp bmf $(DESTDIR)$(BINDIR) + cp bmf.1 $(DESTDIR)$(MANDIR)/man1 + cp bmfconv $(DESTDIR)$(BINDIR) + cp bmfconv.1 $(DESTDIR)$(MANDIR)/man1 + +uninstall: checkroot + rm -f $(DESTDIR)$(BINDIR)/bmf + rm -f $(DESTDIR)$(MANDIR)/man1/bmf.1 + rm -f $(DESTDIR)$(BINDIR)/bmfconv + rm -f $(DESTDIR)$(MANDIR)/man1/bmfconv.1 + +clean: + rm -f core *.o bmf bmfconv + +distclean: clean + rm -f Makefile + +dist: tarball rpmpkg debpkg + +tarball: distclean + (cd ..; \ + cp -ar bmf bmf-$(VERSION); \ + tar czvf bmf-$(VERSION).tar.gz `find bmf-$(VERSION) -type f | egrep -v "CVS|debian"`; \ + rm -rf bmf-$(VERSION); \ + cd $(PWD)) + +debpkg: checkroot + debian/rules binary + rm -rf debian/tmp + +rpmpkg: checkroot + (mkdir -p $(RPMROOT); \ + mkdir -p $(RPMROOT)/SOURCES; \ + mkdir -p $(RPMROOT)/SPECS; \ + mkdir -p $(RPMROOT)/RPMS; \ + mkdir -p $(RPMROOT)/SRPMS; \ + cp ../bmf-$(VERSION).tar.gz $(RPMROOT)/SOURCES; \ + cat bmf.spec.in | sed 's/VERSION/$(VERSION)/' > $(RPMROOT)/SPECS/bmf.spec; \ + cd $(RPMROOT)/SPECS; \ + $(RPM) $(RPMFLAGS) bmf.spec; \ + cp $(RPMROOT)/RPMS/$(ARCH)/bmf-$(VERSION)-*.rpm $(PWD)/..; \ + cp $(RPMROOT)/SRPMS/bmf-$(VERSION)-*.src.rpm $(PWD)/..; \ + cd $(PWD)) + +checkroot: + [ "`whoami`" = root ] || (echo Need root; exit 1) diff --git a/README b/README @@ -0,0 +1,130 @@ + bmf -- Bayesian Mail Filter + +About bmf +========= + +This is a mail filter which uses the Bayes algorithm as explained in Paul +Graham's article "A Plan for Spam". It aims to be faster, smaller, and more +versatile than similar applicatios. Implementation is ANSI C and uses POSIX +functions. Supported platforms are (in theory) all POSIX systems. Support +for win32 is undecided. + +This project provides features which are not available in other filters: + +(1) Independence from external programs and libraries. Tokens are stored in +memory using simple vectors which require no heavyweight external data +structure libraries. Multiple token database formats are supported, +including flat files, libdb, and mysql. Conversion between formats will +always be possible with the included import/export utility and flat files +will always remain an option. + +(2) Efficient processing. Input data is parsed by a handcrafted parser +which weighs in under 3% of the equivalent code generated by flex. No +portion of the input is ever copied and all i/o and memory allocation are +done in large chunks. Updated token lists are merged and written in one +step. Hashing is being considered for the next version to improve lookup +speed. + +(3) Simple and elegant implementation. No heavyweight, copy-intensive mime +decoding routines are used. Decoding of quoted-printable text for selected +mime types is being considered for the next version. + +Note: the core filter function is from esr's bogofilter v0.6 (available at +http://sourceforge.net/projects/bogofilter/) with bugfix updates. + +For the most recent version of this software, see: + + http://sourceforge.net/projects/bmf/ + +How to integrate bmf +==================== + +The following procmail recipes will invoke bmf for each incoming email and +place spam into $MAILDIR/spam. The first sample invokes bmf in its normal +mode of operation and the second invokes bmf as a filter. + + ### begin sample one ### + # Invoke bmf and use return code to filter spam in one step + :0HB + * ? bmf + | formail -A"X-Spam-Status: Yes, tests=bmf" >>$MAILDIR/spam + + ### begin sample two ### + # Invoke bmf as a filter + :0 fw + | bmf -p + + # Filter spam + :0: + ^X-Spam-Status: Yes + $MAILDIR/spam + +The following maildrop equivalents are suggested by Christian Kurz. + + ### begin sample one ### + # Invoke bmf and use return code to filter spam in one step + exception { + `bmf` + if ( $RETURNCODE == 0 ) + to $MAILDIR/spam + } + + ### begin sample two ### + # Invoke bmf as a filter + exception { + xfilter "bmf -p" + if (/^X-Stam-Status: Yes/) + to $MAILDIR/spam + } + + +If you put bmf in your procmail or maildrop scripts as suggested above, it +will always register an email as either spam or non-spam. To reverse this +registration and train bmf, the following mutt macros may be useful: + + macro index \ed "<enter-command>unset wait_key\n<pipe-entry>bmf -S\n<enter-command>set wait_key\n<save-message>=spam\n" + macro index \et "<enter-command>unset wait_key\n<pipe-entry>bmf -t\n<enter-command>set wait_key\n" + macro index \eu "<enter-command>unset wait_key\n<pipe-entry>bmf -N\n<enter-command>set wait_key\n<save-message>=inbox\n" + +These will override these commands: + + <Esc>d = de-register as non-spam, register as spam, and move to spam folder. + <Esc>t = test for spamicity. + <Esc>u = de-register as spam, register as non-spam, and move to inbox folder. + +How to train bmf +================ + +First, please keep in mind that bmf "learns" how to recognize spam from the +input that you give it. It works best if you give it exactly the email that +you receive, or have received in the recent past. + +Here are some good techniques for training bmf: + + - If you keep a history of email that you have received, use your current + and/or saved emails. It is fairly easy to create a small shell script + that will pass all of your normal email to "bmf -n" and all of your spam + to "bmf -s". Note that if you do not use the mbox storage format, you + MUST invoke bmf exactly once per email. Using "cat * | bmf -n" will NOT + work properly because bmf sees the entire input as one big email. + + - If you already use spamassassin, you can use it to train bmf for a + couple of days or weeks. If spamassassin tags it as spam, run it + through "bmf -s". If not, run it through "bmf -n". This can be + automated with procmail or maildrop recipes. + +Here are some things that you should NOT do: + + - Get impatient with the training process and repeatedly pass one email + through "bmf -s". + + - Manually move words around between lists and/or adjust the word counts. + +Final words +=========== + +Thanks for trying bmf. If you have any problems, comments, or suggestions, +please direct them to the bmf mailing list, bmf-user@lists.sourceforge.net. + + Tom Marshall + 20 Oct 2002 diff --git a/TODO b/TODO @@ -0,0 +1,4 @@ +* Make extrema size configurable, and default to ~5% of tokens. +* Teach lexer about multiline MIME headers and case (in)sensitivity. +* Teach lexer about MIME quoted-printable and base64 encodings. +* Make a pop3 proxy (?) diff --git a/bmf.1 b/bmf.1 @@ -0,0 +1,148 @@ +.\"Generated by db2man.xsl. Don't modify this, modify the source. +.de Sh \" Subsection +.br +.if t .Sp +.ne 5 +.PP +\fB\\$1\fR +.PP +.. +.de Sp \" Vertical space (when we can't use .PP) +.if t .sp .5v +.if n .sp +.. +.de Ip \" List item +.br +.ie \\n(.$>=3 .ne \\$3 +.el .ne 3 +.IP "\\$1" \\$2 +.. +.TH "BMF" 1 "" "" "" +.SH NAME +bmf \- efficient Bayesian mail filter +.SH "SYNOPSIS" + +.nf +\fBbmf\fR [-t] [-n] [-s] [-N] [-S] [-f fmt] [-d db] [-i file] [-k n] [-m type] [-p] + [-v] [-V] [-h] +.fi + +.SH "DESCRIPTION" + +.PP +bmf is a Bayesian mail filter. In its normal mode of operation, it takes an email message or other text on standard input, does a statistical check against lists of "good" and "spam" words, registers the new data, and returns a status code indicating whether or not the message is spam. BMF is written with fast, zero-copy algorithms, coded directly in C, and tuned for speed. It aims to be faster, smaller, and more versatile than similar applications. + +.PP +bmf supports both mbox and maildir mail storage formats. It will automatically process multiple messages within an mbox file separately. + +.SH "OPTIONS" + +.PP +Without command-line options, bmf processes the input, registers it as either "good" or "spam", and returns the appropriate error code. The wordlist directory and nonexistent wordfiles are created if absent. + +.PP +\fB-t\fR Test to see if the input is spam. The word lists are not updated. A report is written to stdout showing the final score and the tokens with the highest deviation form a mean of 0.5. + +.PP +\fB-n\fR Register the input as non-spam. + +.PP +\fB-s\fR Register the input as spam. + +.PP +\fB-N\fR Register the input as non-spam and undo a prior registration as spam. + +.PP +\fB-S\fR Register the input as spam and undo a prior registration as non-spam. + +.PP +\fB-f fmt\fR Specify database format. Valid formats are text, db, and mysql. Text is always valid. The others may not be available if the corresponding option was not enabled at compile time. The default is db if available, else text. + +.PP +\fB-d db\fR Specify database or directory for loading and saving word lists. The default is \fI~/.bmf\fR in text mode. + +.PP +\fB-i file\fR Use file for input instead of stdin. + +.PP +\fB-k n\fR Specify the number of extrema (keepers) to use in the Bayes calculation. The default is 15. + +.PP +\fB-m fmt\fR Specify mail storage format. Valid formats are mbox and maildir. The default is to automatically detect the mail storage format. This option is deprecated. + +.PP +\fB-p\fR Copy the input to the output (passthrough) and insert spam headers in the style of SpamAssassin. An X-Spam-Status header is always inserted with processing details. The contents of this header always begin with either "Yes" or "No". If the input is judged to be spam, the header "X-Spam-Flag: YES" is also inserted. + +.PP +\fB-v\fR Be more verbose. This option is not well supported yet. + +.PP +\fB-V\fR Display version information. + +.PP +\fB-h\fR Display usage information. + +.SH "THEORY OF OPERATION" + +.PP +bmf treats its input as a bag of tokens. Each token is checked against "good" and "bad" wordlists, which maintain counts of the numbers of times it has occurred in non-spam and spam mails. These numbers are used to compute the probability that a mail in which the token occurs is spam. After probabilities for all input tokens have been computed, a fixed number of the probabilities that deviate furthest from average are combined using Bayes's theorem on conditional probabilities. + +.PP +While this method sounds crude compared to the more usual pattern-matching approach, it turns out to be extremely effective. Paul Graham's paper A Plan For Spam: \fIhttp://www.paulgraham.com/spam.html\fR is recommended reading. + +.PP +bmf improves on Paul's proposal by doing smarter lexical analysis. In particular, hostnames and IP addresses are not discarded, and certain types of MTA information are discarded (such as message ids and dates). + +.PP +MIME and other attachments are not decoded. Experience from watching the token streams suggests that spam with enclosures invariably gives itself away through cues in the headers and non-enclosure parts. Nonetheless, I would like to add the ability to decode quoted-printable and perhaps base64 encodings for textual attachments. + +.SH "INTEGRATION WITH OTHER TOOLS" + +.PP +Please see the README for samples and suggestions. + +.SH "RETURN VALUES" + +.PP +In passthrough mode: zero for success, nonzero for failure. + +.PP +In non-passthrough mode: 0 for spam; 1 for non-spam; 2 for I/O or other errors. + +.SH "FILES" + +.TP +\fI~/.bmf/goodlist.txt\fR +List of good tokens for text mode. + +.TP +\fI~/.bmf/spamlist.txt\fR +List of bad tokens for text mode. + +.TP +\fI~/.bmf/goodlist.db\fR +List of good tokens for libdb mode. + +.TP +\fI~/.bmf/spamlist.db\fR +List of bad tokens for libdb mode. + +.SH "BUGS" + +.PP +The lexer should recognize multiline headers. + +.PP +The lexer should recognize MIME attachments. + +.PP +Content-Transfer-Encoding is not decoded. + +.SH "AUTHOR" + +.PP +Tom Marshall <tommy@tig-grr.com>. + +.PP +The Bayes algorithm is from bogofilter by Eric S. Raymond <esr@thyrsus.com>. bogofilter can be found at the bogofilter project page: \fIhttp://bogofilter.sourceforge.net/\fR. + diff --git a/bmf.c b/bmf.c @@ -0,0 +1,339 @@ +/* $Id: bmf.c,v 1.20 2002/10/20 18:19:17 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * bmf.c: top level Bayesian mail filter app. + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" +#include "dbh.h" +#include "filt.h" + +/* modes of operation (mutually exclusive) */ +typedef enum +{ + mode_test, /* test and produce report */ + mode_normal, /* test and register result */ + mode_reg_s, /* register as spam */ + mode_reg_n, /* register as non-spam */ + mode_n_to_s, /* undo non-spam registration and register as spam */ + mode_s_to_n /* undo spam registration and register as non-spam */ +} runmode_t; + +static void usage( void ) +{ + printf( "\n" + "Usage: " PACKAGE " [mode] [options]\n" + "\n" + "Modes of operation (mutually exclusive; the last one specified is used):\n" + "\t\tRegister message using historical data if no mode is specified.\n" + "\t-n\tRegister message as non-spam.\n" + "\t-s\tRegister message as spam.\n" + "\t-N\tRegister message as non-spam and undo prior registration as spam.\n" + "\t-S\tRegister message as spam and undo prior registration as non-spam.\n" + "\t-t\tTest mode, print report and do not save results.\n" + "\n" + "Other options:\n" + "\t-f fmt\tSpecify database format (text|db|mysql).\n" + "\t-d db\tSpecify database or directory name.\n" + "\t-i file\tSpecify file to read instead of stdin.\n" + "\t-k n\tSpecify count of extrema to use (keepers), default is 15.\n" + "\t-m type\t[DEPRECATED] Specify mail storage format (mbox|maildir)\n" + "\t-p\tPassthrough mode, like SpamAssassin.\n" + "\t-v\tIncrease verbosity level.\n" + "\t-V\tShow version information and exit.\n" + "\t-h\tShow this message and exit.\n" + "\n" ); + exit( 2 ); +} + +static void version( void ) +{ + printf( "\n" + PACKAGE " version " VERSION " - a Bayesian mail filter\n" + "Copyright (c) 2002 Tom Marshall\n" + "\n" + PACKAGE " comes with ABSOLUTELY NO WARRANTY.\n" + "This is free software. You are welcome to redistribute it under the terms\n" + "of the GNU General Public License. See the file LICENSE in the source\n" + "distribution, or visit http://www.gnu.org/licenses/gpl.html\n" + "\n" ); + exit( 2 ); +} + +int main( int argc, char** argv ) +{ + int ch; + dbfmt_t dbfmt = db_db; + char* dbname = NULL; + bool_t rdonly; + + runmode_t mode = mode_normal; + mbox_t mboxtype = detect; + bool_t do_passthru = false; + + dbh_t* pdb; + dbt_t* pblist; + dbt_t* pglist; + dbt_t* ptable; + vec_t mlist; + stats_t stats; + lex_t lex; + tok_t tok; + bool_t is_spam; + + int fd = STDIN_FILENO; + char* infile = NULL; + + srand(time(NULL)); + atexit( dump_alloc_heap ); + +#ifdef HAVE_LIBDB + dbfmt = db_db; +#else + dbfmt = db_text; +#endif + + stats.keepers = DEF_KEEPERS; + while( (ch = getopt( argc, argv, "NSVd:f:i:hk:m:npstv" )) != EOF ) + { + switch( ch ) + { + case 'N': + mode = mode_s_to_n; + break; + case 'S': + mode = mode_n_to_s; + break; + case 'V': + version(); + break; /* notreached */ + case 'd': + free( dbname ); + dbname = strdup( optarg ); + break; + case 'f': + if( strcasecmp( optarg, "text" ) == 0 ) + { + dbfmt = db_text; + } + else if( strcasecmp( optarg, "db" ) == 0 ) + { + dbfmt = db_db; + } + else if( strcasecmp( optarg, "mysql" ) == 0 ) + { + dbfmt = db_mysql; + } + else + { + usage(); + } + break; + case 'h': + usage(); + break; /* notreached */ + case 'i': + free( infile ); + infile = strdup( optarg ); + break; + case 'k': + stats.keepers = atoi( optarg ); + break; + case 'm': + if( strcasecmp( optarg, "mbox" ) == 0 ) + { + mboxtype = mbox; + } + else if( strcasecmp( optarg, "maildir" ) == 0 ) + { + mboxtype = maildir; + } + else + { + usage(); + } + break; + case 'n': + mode = mode_reg_n; + break; + case 'p': + do_passthru = true; + break; + case 's': + mode = mode_reg_s; + break; + case 't': + mode = mode_test; + break; + case 'v': + g_verbose++; + verbose( 1, "Verbose level now %u\n", g_verbose ); + break; + default: + usage(); + } + } + stats.extrema = (discrim_t*)malloc( stats.keepers*sizeof(discrim_t) ); + + if( infile != NULL ) + { + fd = open( infile, O_RDONLY ); + if( fd == -1 ) + { + fprintf( stderr, "%s: cannot open input file '%s': %s\n", + argv[0], infile, strerror(errno) ); + exit( 2 ); + } + } + + pdb = dbh_open( dbfmt, "localhost", dbname, DB_USER, DB_PASS ); + if( pdb == NULL ) + { + fprintf( stderr, "%s: cannot open database\n", argv[0] ); + exit( 2 ); + } + + lex_create( &lex, mboxtype ); + if( !lex_load( &lex, fd ) ) + { + fprintf( stderr, "%s: cannot read input\n", argv[0] ); + exit( 2 ); + } + lex_nexttoken( &lex, &tok ); + if( tok.tt == eof ) + { + fprintf( stderr, "%s: no input available\n", argv[0] ); + exit( 2 ); + } + + while( tok.tt != eof ) + { + if( mboxtype == mbox && tok.tt != from ) + { + fprintf( stderr, "%s: input does not look like an mbox message\n", argv[0] ); + exit( 2 ); + } + + rdonly = (mode == mode_test || mode == mode_reg_n); + pblist = pdb->opentable( pdb, "spamlist", rdonly ); + if( pblist == NULL ) + { + fprintf( stderr, "%s: cannot open spamlist\n", argv[0] ); + exit( 2 ); + } + + rdonly = (mode == mode_test || mode == mode_reg_s); + pglist = pdb->opentable( pdb, "goodlist", rdonly ); + if( pglist == NULL ) + { + fprintf( stderr, "%s: cannot open goodlist\n", argv[0] ); + exit( 2 ); + } + + vec_create( &mlist ); + bvec_loadmsg( &mlist, &lex, &tok ); + + switch( mode ) + { + case mode_test: + bayesfilt( pglist, pblist, &mlist, &stats ); + is_spam = (stats.spamicity > SPAM_CUTOFF); + break; + case mode_normal: + bayesfilt( pglist, pblist, &mlist, &stats ); + is_spam = (stats.spamicity > SPAM_CUTOFF); + ptable = (is_spam ? pblist : pglist); + svec_sort( &mlist ); + if( !ptable->mergeclose( ptable, &mlist ) ) + { + fprintf( stderr, "%s: cannot merge/save list\n", argv[0] ); + exit( 2 ); + } + break; + case mode_reg_s: + stats.spamicity = 1.0; + is_spam = true; + svec_sort( &mlist ); + if( !pblist->mergeclose( pblist, &mlist ) ) + { + fprintf( stderr, "%s: cannot merge/save list\n", argv[0] ); + exit( 2 ); + } + break; + case mode_reg_n: + stats.spamicity = 0.0; + is_spam = false; + svec_sort( &mlist ); + if( !pglist->mergeclose( pglist, &mlist ) ) + { + fprintf( stderr, "%s: cannot merge/save list\n", argv[0] ); + exit( 2 ); + } + break; + case mode_n_to_s: + stats.spamicity = 1.0; + is_spam = true; + svec_sort( &mlist ); + if( !pblist->mergeclose( pblist, &mlist ) || + !pglist->unmergeclose( pglist, &mlist ) ) + { + fprintf( stderr, "%s: cannot merge/save list\n", argv[0] ); + exit( 2 ); + } + break; + case mode_s_to_n: + stats.spamicity = 0.0; + is_spam = false; + svec_sort( &mlist ); + if( !pblist->unmergeclose( pblist, &mlist ) || + !pglist->mergeclose( pglist, &mlist ) ) + { + fprintf( stderr, "%s: cannot merge/save list\n", argv[0] ); + exit( 2 ); + } + break; + default: + usage(); + } + + if( mode == mode_test ) + { + statdump( &stats, STDOUT_FILENO ); + } + + if( do_passthru ) + { + lex_passthru( &lex, is_spam, stats.spamicity ); + } + + vec_destroy( &mlist ); + + pglist->close( pglist ); + free( pglist ); + pblist->close( pblist ); + free( pblist ); + } + + lex_destroy( &lex ); + + pdb->close( pdb ); + free( pdb ); + + if( infile != NULL ) + { + free( infile ); + close( fd ); + } + free( stats.extrema ); + + return ( (do_passthru || is_spam) ? 0 : 1 ); +} diff --git a/bmf.spec.in b/bmf.spec.in @@ -0,0 +1,64 @@ +Name: bmf +Version: VERSION +Release: 1 +URL: http://www.sourceforge.net/projects/bmf +Source0: %{name}-%{version}.tar.gz +License: GPL +Group: Applications/Internet +Summary: fast anti-spam filtering by Bayesian statistical analysis +Buildroot: %{_tmppath}/%{name}-%{version}-root + +%description +bmf is a Bayesian mail filter. It takes an email message or other text on +stdin, does a statistical check against lists of "good" and "spam" words, +and returns a status code indicating whether or not the message is spam. +bmf is efficient, small, and self-contained. + +%prep + +%setup + +%build +./configure --with-libdb --without-mysql +make + +%install +[ -n "$RPM_BUILD_ROOT" -a "$RPM_BUILD_ROOT" != / ] && rm -rf $RPM_BUILD_ROOT +make DESTDIR=${RPM_BUILD_ROOT} install +gzip $RPM_BUILD_ROOT/%{_mandir}/*/*.? + + +%files +%{_bindir}/bmf +%{_mandir}/man1/bmf.1.gz +%{_bindir}/bmfconv +%{_mandir}/man1/bmfconv.1.gz +%doc README LICENSE + +%changelog +* Mon Oct 14 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.9.3. + +* Sat Oct 12 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.9.2. + +* Sat Oct 12 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.9.1. + +* Wed Oct 09 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.84. + +* Mon Oct 07 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.83. + +* Sat Oct 05 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.82. + +* Thu Oct 03 2002 Tom Marshall <tommy@tig-grr.com> +- Update to version 0.81. +- Add bmfconv. +- Use new configure script. + +* Fri Sep 27 2002 Tom Marshall <tommy@tig-grr.com> +- Initial build. + diff --git a/bmfconv.1 b/bmfconv.1 @@ -0,0 +1,81 @@ +.\"Generated by db2man.xsl. Don't modify this, modify the source. +.de Sh \" Subsection +.br +.if t .Sp +.ne 5 +.PP +\fB\\$1\fR +.PP +.. +.de Sp \" Vertical space (when we can't use .PP) +.if t .sp .5v +.if n .sp +.. +.de Ip \" List item +.br +.ie \\n(.$>=3 .ne \\$3 +.el .ne 3 +.IP "\\$1" \\$2 +.. +.TH "BMFCONV" 1 "" "" "" +.SH NAME +bmfconv \- Database converter for bmf +.SH "SYNOPSIS" + +.nf +\fBbmfconv\fR [-f fmt] [-d db] [-e] [-i] [-v] [-h] +.fi + +.SH "DESCRIPTION" + +.PP +bmfconv converts bmf token databases between the supported formats. It can import flat text files into databases and export databases into flat text files. + +.PP +PLEASE NOTE that the text files used in import and export operations are read and written in the current directory. + +.SH "OPTIONS" + +.PP +\fB-f fmt\fR Specify database format. Supported formats are "db" for libdb and "mysql" for MySQL. + +.PP +\fB-d db\fR Specify database name. + +.PP +\fB-e\fR Export the database to text files. + +.PP +\fB-i\fR Import the database from text files. + +.PP +\fB-v\fR Display version information. + +.PP +\fB-h\fR Display usage information. + +.SH "RETURN VALUES" + +.PP +0 if conversion succeeds, nonzero if conversion fails. + +.SH "FILES" + +.TP +\fIgoodlist.txt\fR +Text file for import or export of good tokens. + +.TP +\fIspamlist.txt\fR +Text file for import or export of spam tokens. + +.SH "BUGS" + +.PP +Should be more robust. + +.SH "AUTHOR" + +.PP +Tom Marshall <tommy@tig-grr.com>. bmfconv is a part of the bmf package. + diff --git a/bmfconv.c b/bmfconv.c @@ -0,0 +1,169 @@ +/* $Id: bmfconv.c,v 1.9 2002/10/20 18:19:17 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * bmfconv.c: bmf database converter + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "vec.h" +#include "dbh.h" + +typedef enum +{ + none, + db2text, + text2db +} dir_t; + +static void usage( void ) +{ + printf( "\n" + "Usage: " PACKAGE " [options]\n" + "\t-f fmt\tSpecify database format (db|mysql).\n" + "\t-d db\tSpecify database or directory name.\n" + "\t-e\tExport to text files goodlist.txt and spamlist.txt.\n" + "\t-i\tImport from text files goodlist.txt and spamlist.txt.\n" + "\t-v\tShow version information and exit\n" + "\t-h\tShow this message and exit\n" + "\n" ); + exit( 2 ); +} + +static void version( void ) +{ + printf( "\n" + PACKAGE " version " VERSION " - a Bayesian mail filter\n" + "Copyright (c) 2002 Tom Marshall\n" + "\n" + PACKAGE " comes with ABSOLUTELY NO WARRANTY.\n" + "This is free software. You are welcome to redistribute it under the terms\n" + "of the GNU General Public License. See the file LICENSE in the source\n" + "distribution, or visit http://www.gnu.org/licenses/gpl.html\n" + "\n" ); + exit( 2 ); +} + +int main( int argc, char** argv ) +{ + int ch; + dbfmt_t dbfmt = db_db; + char* dbname = NULL; + bool_t rdonly; + + dbh_t* pdb; + dbt_t* ptable; + dir_t dir = none; + + while( (ch = getopt( argc, argv, "d:ef:ihv" )) != EOF ) + { + switch( ch ) + { + case 'd': + free( dbname ); + dbname = strdup( optarg ); + break; + case 'e': + dir = db2text; + break; + case 'f': + if( strcasecmp( optarg, "db" ) == 0 ) + { + dbfmt = db_db; + } + else if( strcasecmp( optarg, "mysql" ) == 0 ) + { + dbfmt = db_mysql; + } + else + { + usage(); + } + break; + case 'h': + usage(); + break; /* notreached */ + case 'i': + dir = text2db; + break; + case 'v': + version(); + break; /* notreached */ + default: + usage(); + } + } + if( dir == none ) + { + usage(); + } + + pdb = dbh_open( dbfmt, "localhost", dbname, DB_USER, DB_PASS ); + if( pdb == NULL ) + { + fprintf( stderr, "cannot open database\n" ); + exit( 1 ); + } + rdonly = (dir == db2text ? true : false); + + ptable = pdb->opentable( pdb, "spamlist", rdonly ); + if( ptable == NULL ) + { + fprintf( stderr, "cannot open spamlist\n" ); + exit( 1 ); + } + if( dir == db2text ) + { + if( !ptable->export( ptable, "spamlist.txt" ) ) + { + fprintf( stderr, "cannot export spamlist\n" ); + exit( 1 ); + } + } + else + { + if( !ptable->import( ptable, "spamlist.txt" ) ) + { + fprintf( stderr, "cannot import spamlist\n" ); + exit( 1 ); + } + } + ptable->close( ptable ); + free( ptable ); + + ptable = pdb->opentable( pdb, "goodlist", rdonly ); + if( ptable == NULL ) + { + fprintf( stderr, "cannot open goodlist\n" ); + exit( 1 ); + } + if( dir == db2text ) + { + if( !ptable->export( ptable, "goodlist.txt" ) ) + { + fprintf( stderr, "cannot export goodlist\n" ); + exit( 1 ); + } + } + else + { + if( !ptable->import( ptable, "goodlist.txt" ) ) + { + fprintf( stderr, "cannot import goodlist\n" ); + exit( 1 ); + } + } + ptable->close( ptable ); + free( ptable ); + + pdb->close( pdb ); + free( pdb ); + + return 0; +} diff --git a/config.h b/config.h @@ -0,0 +1,80 @@ +/* $Id: config.h,v 1.8 2002/10/20 07:16:57 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _CONFIG_H +#define _CONFIG_H + +/************************************** + * Standard headers + */ +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include <errno.h> +#include <math.h> +#include <ctype.h> +#include <assert.h> + +/************************************** + * System headers + */ +#include <sys/types.h> +#include <limits.h> +#include <unistd.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <sys/file.h> +#include <time.h> + +/************************************** + * For convenience + */ +typedef unsigned char byte; +typedef const char* cpchar; +typedef const byte* cpbyte; +typedef const void* cpvoid; +typedef enum { false, true } bool_t; + +#define min(a,b) ( (a)<(b) ? (a) : (b) ) +#define max(a,b) ( (a)<(b) ? (b) : (a) ) +#define minmax(v,a,b) ( (v)<(a)?(a) : (v)>(b)?(b) : (v) ) + +/* XXX: need to figure out MH and any others (MMDF?) */ +typedef enum { detect, mbox, maildir } mbox_t; + +/************************************** + * Tweakables + */ + +/* If you have the mysql client libs installed and wish to use them... */ +/* #define HAVE_MYSQL */ + +#define MSGCOUNT_KEY ".MSGCOUNT" +#define MSGCOUNT_KEY_LEN (sizeof(MSGCOUNT_KEY)-1) + +#define DB_USER "username" +#define DB_PASS "password" + +#define IOBUFSIZE 4096 /* chunk size for file buffers */ +#define MAXWORDLEN 20 /* max word length, inclusive */ +#define MAXFREQ 4 /* max times to count word per email */ +#define GOOD_BIAS 2.0 /* give good words more weight */ +#define DEF_KEEPERS 15 /* how many extrema to keep by default */ +#define MINIMUM_FREQ 5 /* min word count for consideration in filter */ +#define UNKNOWN_WORD 0.4 /* odds that unknown word is spammish */ +#define SPAM_CUTOFF 0.9 /* if it's spammier than this... */ + +/* + * If NON_EQUIPROBABLE is defined, use ratio of spamcount/goodcount instead + * of UNKNOWN_WORDS, and as a factor in the known word calculation. This is + * merely copied from bogofilter. I didn't write it and I cannot explain the + * relative merits of using it or not. Please don't ask. :-) + */ + +#endif /* ndef _CONFIG_H */ diff --git a/configure b/configure @@ -0,0 +1,354 @@ +#!/bin/sh + +echo "" + +# defaults +DEBUG=no +with_libdb=test +with_mysql=test + +# parse options +# --with/--without parsing copied from autoconf's output +while [ $# -gt 0 ]; do + case $1 in + --with-*) + ac_option=$1 + ac_optarg=`expr "x$ac_option" : 'x[^=]*=\(.*\)'` + ac_package=`expr "x$ac_option" : 'x-*with-\([^=]*\)'` + # Reject names that are not valid shell variable names. + expr "x$ac_package" : ".*[^A-Za-z0-9_-]" >/dev/null && + { echo "error: invalid package name: $ac_package"; 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-*) + ac_option=$1 + ac_optarg=`expr "x$ac_option" : 'x[^=]*=\(.*\)'` + ac_package=`expr "x$ac_option" : 'x-*without-\([^=]*\)'` + # Reject names that are not valid shell variable names. + expr "x$ac_package" : ".*[^A-Za-z0-9_-]" >/dev/null && + { echo "error: invalid package name: $ac_package"; exit 1; } + ac_package=`echo $ac_package | sed 's/-/_/g'` + eval "with_$ac_package=no" + ;; + --debug=yes|--debug|-d) + echo "Debug mode enabled." + with_debug=yes + ;; + --debug=no) + echo "Debug mode disabled." + with_debug=no + ;; + --help|-help|-h) + echo "usage: $0 [ options ]" + echo "available options:" + echo " --debug=yes (or -d) Enable debugging support." + echo " --debug=no Disable debugging support." + echo " --with-package Enable support for package in default location." + echo " --with-package=path Enable support for package installed in path." + echo " --without-package Disable support for package." + echo " --help (or -h) Show this message." + echo "" + echo "relevant packages:" + echo " libdb = BerkeleyDB" + echo " mysql = MySQL database" + exit 1 + ;; + *) + echo "Unknown option '$1', try -h for help" + exit 1 + ;; + esac + shift +done + +echo "Examining system setup..." + +# Some known configs: +# +# uname -s uname -r uname -m uname -p +# ======== =========== ======== ======== +# SunOS 5.6 sun4u sparc +# Linux 2.2.17 i686 unknown +# FreeBSD 4.1-RELEASE i386 i386 + +UNAME_S=`uname -s` +UNAME_R=`uname -r` + +case "$UNAME_S" in + Linux) + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=gcc + CFLAGS="$CCDBG -D_UNIX -D_LINUX -Wall" + CXX=g++ + CXXFLAGS=${CFLAGS} + AR=ar + ARFLAGS="-rc" + LD=gcc + LDFLAGS="$LDDBG" + SYSLIBS="" + LIBDB_LIB="-ldb" + MYSQL_LIB="-lmysqlclient" + ;; + FreeBSD) + case "$UNAME_R" in + 2.*) + OSVER=20 + ;; + 3.*) + OSVER=30 + ;; + 4.*) + OSVER=40 + ;; + *) + # Assume 5.0 + + OSVER=50 + ;; + esac + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=gcc + CFLAGS="$CCDBG -D_UNIX -D_BSD=$OSVER -Wall" + CXX=g++ + CXXFLAGS=${CFLAGS} + AR=ar + ARFLAGS="-rc" + LD=gcc + LDFLAGS="$LDDBG" + SYSLIBS="" + LIBDB_LIB="" + MYSQL_LIB="-lmysqlclient" + ;; + OpenBSD) + # I'm guessing OpenBSD looks mostly like FreeBSD 4.x + OSVER=40 + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=gcc + CFLAGS="$CCDBG -D_UNIX -D_BSD=$OSVER -Wall" + CXX=g++ + CXXFLAGS=${CFLAGS} + AR=ar + ARFLAGS="-rc" + LD=gcc + LDFLAGS="$LDDBG" + SYSLIBS="" + LIBDB_LIB="" + MYSQL_LIB="-lmysqlclient" + ;; + Darwin) + # I'm guessing Darwin looks mostly like FreeBSD 4.x + OSVER=40 + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=cc + CFLAGS="$CCDBG -D_UNIX -D_BSD=$OSVER -DNOLOCK -Wall" + CXX=c++ + CXXFLAGS=${CFLAGS} + AR=ar + ARFLAGS="-rc" + LD=cc + LDFLAGS="$LDDBG" + SYSLIBS="" + LIBDB_LIB="" + MYSQL_LIB="-lmysqlclient" + ;; + SunOS) + case "$UNAME_R" in + 5.6) + OSVER=56 + ;; + 5.7) + OSVER=57 + ;; + 5.8) + OSVER=58 + ;; + *) + # Assume 9.0+ + OSVER=59 + ;; + esac + # Prefer gcc to the native cc here because I haven't tested with the + # native compiler yet. + type gcc >/dev/null 2>&1 + if [ $? -eq 0 ]; then + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=gcc + CFLAGS="$CCDBG -D_UNIX -D_SOLARIS=$OSVER -Wall" + CXX=g++ + CXXFLAGS=${CFLAGS} + LD=gcc + LDFLAGS="$LDDBG" + else + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=cc + CFLAGS="-DNDEBUG -D_UNIX -D_SOLARIS=$OSVER" + CXX=c++ + CXXFLAGS=${CFLAGS} + LD=ld + LDFLAGS="" + fi + AR=ar + ARFLAGS="-rc" + SYSLIBS="" + LIBDB_LIB="-ldb" + MYSQL_LIB="-lmysqlclient" + ;; + OSF1) + if [ "$with_debug" = "yes" ]; then + CCDBG="-g2" + LDDBG="-g2" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=cc + CFLAGS="$CCDBG -D_UNIX -D_OSF" + CXX=cxx + CXXFLAGS="-noexceptions ${CFLAGS}" + AR=ar + ARFLAGS="-rc" + LD=ld + LDFLAGS="" + SYSLIBS="-lm" + LIBDB_LIB="-ldb" + MYSQL_LIB="-lmysqlclient" + ;; + *) + if [ "$with_debug" = "yes" ]; then + CCDBG="-g" + LDDBG="-g" + else + CCDBG="-DNDEBUG" + LDDBG="" + fi + CC=cc + CFLAGS="$CCDBG -DNDEBUG -D_UNIX" + CXX=c++ + CXXFLAGS=${CFLAGS} + AR=ar + ARFLAGS="-rc" + LD=ld + LDFLAGS="$LDDBG" + SYSLIBS="" + LIBDB_LIB="-ldb" + MYSQL_LIB="-lmysqlclient" + ;; +esac + +echo -n "Looking for compiler... " +type $CC >/dev/null 2>&1 +if [ $? -ne 0 ]; then + echo "compiler '$CC' not found!" + exit 1 +else + echo "$CC is executable." +fi + +echo -n "Checking for BerkeleyDB... " +if [ "$with_libdb" = "test" ]; then + if [ -e /usr/local/BerkeleyDB.4.1/include/db.h ]; then + with_libdb=/usr/local/BerkeleyDB.4.1 + LIBDB_LIB="-ldb" + elif [ -e /usr/include/db.h -o \ + -e /usr/local/include/db.h ]; then + with_libdb=yes + else + with_libdb=no + fi +fi +if [ "$with_libdb" = "no" ]; then + echo "disabled." +else + if [ "$with_libdb" != "yes" ]; then + CFLAGS="$CFLAGS -I$with_libdb/include" + LDFLAGS="$LDFLAGS -L$with_libdb/lib" + fi + CFLAGS="$CFLAGS -DHAVE_LIBDB" + LDFLAGS="$LDFLAGS $LIBDB_LIB" + echo "enabled." +fi + +echo -n "Checking for MySQL... " +if [ "$with_mysql" = "test" ]; then + type mysql_config >/dev/null 2>&1 + if [ $? -eq 0 ]; then + with_mysql=yes + CFLAGS="$CFLAGS `mysql_config --cflags`" + LDFLAGS="$LDFLAGS `mysql_config --libs`" + # mysql_config will add -lmysqlclient, don't add it twice + MYSQL_LIB="" + else + with_mysql=no + fi +fi +if [ "$with_mysql" = "no" ]; then + echo "disabled." +else + if [ "$with_mysql" != "yes" ]; then + CFLAGS="$CFLAGS -I$with_mysql/include" + LDFLAGS="$LDFLAGS -L$with_mysql/lib" + fi + CFLAGS="$CFLAGS -DHAVE_MYSQL" + LDFLAGS="$LDFLAGS $MYSQL_LIB" + echo "enabled." +fi + +for D in .; do + cat $D/Makefile.in | sed "s%@CC@%${CC}%g + s%@CXX@%${CXX}%g + s%@AR@%${AR}%g + s%@LD@%${LD}%g + s%@CFLAGS@%${CFLAGS}%g + s%@CXXFLAGS@%${CXXFLAGS}%g + s%@ARFLAGS@%${ARFLAGS}%g + s%@LDFLAGS@%${LDFLAGS}%g + s%@SYSLIBS@%${SYSLIBS}%" > $D/Makefile || exit 1 +done + +echo "Configuration successful." +echo "Now run 'make all' and 'make install'." +echo "" diff --git a/dbdb.c b/dbdb.c @@ -0,0 +1,684 @@ +/* $Id: dbdb.c,v 1.22 2002/10/19 09:59:35 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * dbdb.c: berkeley database handler + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" + +#include "dbh.h" +#include "dbdb.h" + +#ifdef HAVE_LIBDB + +#define DBT_init( pdbt ) memset( pdbt, 0, sizeof(DBT) ) + +#if !defined(DB_VERSION_MAJOR) /* v1 */ +#define dbx_get(dbp,kp,vp) dbp->get( dbp, kp, vp, 0 ) +#define dbx_put(dbp,kp,vp) dbp->put( dbp, kp, vp, 0 ) +#define dbx_fd(dbp,fd) fd = dbp->fd( dbp ) +#else /* v2+ */ +#define dbx_get(dbp,kp,vp) dbp->get( dbp, NULL, kp, vp, 0 ) +#define dbx_put(dbp,kp,vp) dbp->put( dbp, NULL, kp, vp, 0 ) +#define dbx_fd(dbp,fd) dbp->fd( dbp, &fd ) +#endif /* DB_VERSION_MAJOR */ + +#if !defined(DB_VERSION_MAJOR) /* v1 */ +typedef DB DBC; /* no separate cursor type */ +#define dbx_createcursor(dbp,dbcp) ((dbcp = dbp) ? 0 : -1) +#define dbx_destroycursor(dbcp) (dbcp = NULL) +#define dbx_first(dbcp,kp,vp) dbcp->seq(dbcp,kp,vp,R_FIRST) +#define dbx_next(dbcp,kp,vp) dbcp->seq(dbcp,kp,vp,R_NEXT) +#define dbx_prev(dbcp,kp,vp) dbcp->seq(dbcp,kp,vp,R_PREV) +#define dbx_last(dbcp,kp,vp) dbcp->seq(dbcp,kp,vp,R_LAST) +#elif DB_VERSION_MAJOR == 2 +#define dbx_createcursor(dbp,dbcp) dbp->cursor(dbp,NULL,&csrp) +#define dbx_destroycursor(dbcp) dbcp->c_close(dbcp) +#define dbx_first(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_FIRST) +#define dbx_next(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_NEXT) +#define dbx_prev(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_PREV) +#define dbx_last(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_LAST) +#else /* v3+ */ +#define dbx_createcursor(dbp,dbcp) dbp->cursor(dbp,NULL,&csrp,0) +#define dbx_destroycursor(dbcp) dbcp->c_close(dbcp) +#define dbx_first(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_FIRST) +#define dbx_next(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_NEXT) +#define dbx_prev(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_PREV) +#define dbx_last(dbcp,kp,vp) dbcp->c_get(dbcp,kp,vp,DB_LAST) +#endif /* DB_VERSION_MAJOR */ + +static void char2DBT( DBT* pdbt, char* p ) +{ + pdbt->data = p; + pdbt->size = strlen(p); +} + +static void uint2DBT( DBT* pdbt, uint* p ) +{ + pdbt->data = p; + pdbt->size = sizeof(uint); +} + +static uint DBT2uint( DBT* pdbt ) +{ + uint n; + memcpy( &n, pdbt->data, sizeof(n) ); + return n; +} + +dbh_t* dbdb_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ) +{ + dbhdb_t* pthis; + + uint dirlen; + cpchar phome; + struct stat st; + + pthis = (dbhdb_t*)malloc( sizeof(dbhdb_t) ); + if( pthis == NULL ) + { + goto bail; + } + pthis->close = dbdb_db_close; + pthis->opentable = dbdb_db_opentable; + if( dbname != NULL && *dbname != '\0' ) + { + dirlen = strlen( dbname ); + pthis->dir = strdup( dbname ); + if( pthis->dir[dirlen-1] == '/' ) + { + pthis->dir[dirlen-1] = '\0'; + } + } + else + { + phome = getenv( "HOME" ); + if( phome == NULL || *phome == '\0' ) + { + phome = "."; + } + pthis->dir = (char*)malloc( strlen(phome)+5+1 ); + if( pthis->dir == NULL ) + { + goto bail; + } + sprintf( pthis->dir, "%s/.bmf", phome ); + } + + /* ensure config directory exists */ + if( stat( pthis->dir, &st ) != 0 ) + { + if( errno == ENOENT ) + { + if( mkdir( pthis->dir, S_IRUSR|S_IWUSR|S_IXUSR ) != 0 ) + { + goto bail; + } + } + else + { + goto bail; + } + } + else + { + if( !S_ISDIR( st.st_mode ) ) + { + goto bail; + } + } + +#if !defined(DB_VERSION_MAJOR) || DB_VERSION_MAJOR < 3 + /* no initialization */ +#else /* DB_VERSION_MAJOR >= 3 */ + if( db_env_create( &pthis->envp, 0 ) != 0 ) + { + goto bail; + } + if( pthis->envp->open( pthis->envp, pthis->dir, DB_INIT_LOCK|DB_INIT_MPOOL|DB_CREATE, 0644 ) != 0 ) + { + goto bail; + } +#endif /* DB_VERSION_MAJOR */ + + return (dbh_t*)pthis; + +bail: + free( pthis ); + return NULL; +} + +bool_t dbdb_db_close( dbhdb_t* pthis ) +{ +#if !defined(DB_VERSION_MAJOR) || DB_VERSION_MAJOR < 3 + /* no cleanup */ +#else /* DB_VERSION_MAJOR >= 3 */ + pthis->envp->close( pthis->envp, 0 ); +#endif /* DB_VERSION_MAJOR */ + + free( pthis->dir ); + pthis->dir = NULL; + + return true; +} + +dbt_t* dbdb_db_opentable( dbhdb_t* pthis, cpchar table, bool_t rdonly ) +{ + dbtdb_t* ptable; + DB* dbp; + DBT key; + DBT val; + + char szpath[PATH_MAX]; + + ptable = (dbtdb_t*)malloc( sizeof(dbtdb_t) ); + if( ptable == NULL ) + { + return NULL; + } + ptable->close = dbdb_table_close; + ptable->mergeclose = dbdb_table_mergeclose; + ptable->unmergeclose = dbdb_table_unmergeclose; + ptable->import = dbdb_table_import; + ptable->export = dbdb_table_export; + ptable->getmsgcount = dbdb_table_getmsgcount; + ptable->getcount = dbdb_table_getcount; + ptable->dbp = NULL; + + sprintf( szpath, "%s/%s.db", pthis->dir, table ); +#if !defined(DB_VERSION_MAJOR) + if( (dbp = dbopen( szpath, O_CREAT|O_RDWR, 0644, DB_BTREE, NULL)) == NULL ) + { + goto bail; + } +#elif DB_VERSION_MAJOR == 2 + if( db_open( szpath, DB_BTREE, DB_CREATE, 0644, NULL, NULL, &dbp ) != 0 ) + { + goto bail; + } +#elif (DB_VERSION_MAJOR == 3) || (DB_VERSION_MAJOR == 4 && DB_VERSION_MINOR == 0) + ptable->envp = pthis->envp; + if( db_create( &dbp, NULL, 0 ) != 0 ) + { + goto bail; + } + if( dbp->open( dbp, szpath, NULL, DB_BTREE, DB_CREATE, 0644 ) != 0 ) + { + goto bail; + } +#else /* v4.1+ */ + ptable->envp = pthis->envp; + if( db_create( &dbp, NULL, 0 ) != 0 ) + { + goto bail; + } + if( dbp->open( dbp, NULL, szpath, NULL, DB_BTREE, DB_CREATE, 0644 ) != 0 ) + { + goto bail; + } +#endif /* DB_VERSION_MAJOR */ + ptable->dbp = dbp; + + DBT_init( &key ); + DBT_init( &val ); + ptable->nmsgs = 0; + char2DBT( &key, MSGCOUNT_KEY ); + if( dbx_get( dbp, &key, &val ) == 0 ) + { + ptable->nmsgs = DBT2uint( &val ); + } + + return (dbt_t*)ptable; + +bail: + free( ptable ); + return NULL; +} + +static bool_t dbdb_table_lock( dbtdb_t* pthis ) +{ +#ifndef NOLOCK + struct flock lock; + int fd; + + dbx_fd( pthis->dbp, fd ); + memset( &lock, 0, sizeof(lock) ); + lock.l_type = F_WRLCK; + lock.l_start = 0; + lock.l_whence = SEEK_SET; + lock.l_len = 0; + if( fcntl( fd, F_SETLKW, &lock ) != 0 ) + { + return false; + } +#endif /* ndef NOLOCK */ + return true; +} + +static bool_t dbdb_table_unlock( dbtdb_t* pthis ) +{ +#ifndef NOLOCK + struct flock lock; + int fd; + + dbx_fd( pthis->dbp, fd ); + memset( &lock, 0, sizeof(lock) ); + lock.l_type = F_UNLCK; + lock.l_start = 0; + lock.l_whence = SEEK_SET; + lock.l_len = 0; + if( fcntl( fd, F_SETLK, &lock ) != 0 ) + { + return false; + } +#endif /* ndef NOLOCK */ + return true; +} + +bool_t dbdb_table_close( dbtdb_t* pthis ) +{ + DB* dbp = pthis->dbp; + + if( dbp != NULL ) + { +#if !defined(DB_VERSION_MAJOR) /* v1 */ + dbp->close( dbp ); +#else /* v2+ */ + dbp->close( dbp, 0 ); +#endif /* DB_VERSION_MAJOR */ + pthis->dbp = NULL; + } + + return true; +} + +bool_t dbdb_table_mergeclose( dbtdb_t* pthis, vec_t* pmsg ) +{ + DB* dbp = pthis->dbp; + DBT key; + DBT val; + + char szword[MAXWORDLEN+1]; + uint count; + veciter_t msgiter; + str_t* pmsgstr; + + if( pthis->dbp == NULL ) + { + return false; + } + + if( !dbdb_table_lock( pthis ) ) + { + return false; + } + + pthis->nmsgs++; + + DBT_init( &key ); + DBT_init( &val ); + + char2DBT( &key, MSGCOUNT_KEY ); + uint2DBT( &val, &pthis->nmsgs ); + dbx_put( dbp, &key, &val ); + + vec_first( pmsg, &msgiter ); + pmsgstr = veciter_get( &msgiter ); + + while( pmsgstr != NULL ) + { + assert( pmsgstr->len <= MAXWORDLEN ); + strncpylwr( szword, pmsgstr->p, pmsgstr->len ); + szword[pmsgstr->len] = '\0'; + count = db_getnewcount( &msgiter ); + + char2DBT( &key, szword ); + if( dbx_get( dbp, &key, &val ) == 0 ) + { + count += DBT2uint( &val ); + } + char2DBT( &key, szword ); + uint2DBT( &val, &count ); + if( dbx_put( dbp, &key, &val ) != 0 ) + { + goto bail; + } + + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + + veciter_destroy( &msgiter ); + dbdb_table_unlock( pthis ); + return dbdb_table_close( pthis ); + +bail: + return false; +} + +bool_t dbdb_table_unmergeclose( dbtdb_t* pthis, vec_t* pmsg ) +{ + DB* dbp = pthis->dbp; + DBT key; + DBT val; + + char szword[MAXWORDLEN+1]; + uint count; + veciter_t msgiter; + str_t* pmsgstr; + + if( pthis->dbp == NULL ) + { + return false; + } + + if( pthis->nmsgs > 0 ) + { + pthis->nmsgs--; + } + + if( !dbdb_table_lock( pthis ) ) + { + return false; + } + + DBT_init( &key ); + DBT_init( &val ); + + char2DBT( &key, MSGCOUNT_KEY ); + uint2DBT( &val, &pthis->nmsgs ); + dbx_put( dbp, &key, &val ); + + vec_first( pmsg, &msgiter ); + pmsgstr = veciter_get( &msgiter ); + + while( pmsgstr != NULL ) + { + assert( pmsgstr->len <= MAXWORDLEN ); + strncpylwr( szword, pmsgstr->p, pmsgstr->len ); + szword[pmsgstr->len] = '\0'; + count = db_getnewcount( &msgiter ); + + char2DBT( &key, szword ); + if( dbx_get( dbp, &key, &val ) == 0 ) + { + uint n = DBT2uint( &val ); + n = (n > count) ? (n - count) : 0; + char2DBT( &key, szword ); + uint2DBT( &val, &n ); + if( dbx_put( dbp, &key, &val ) != 0 ) + { + goto bail; + } + } + + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + + veciter_destroy( &msgiter ); + dbdb_table_unlock( pthis ); + return dbdb_table_close( pthis ); + +bail: + return false; +} + +bool_t dbdb_table_import( dbtdb_t* pthis, cpchar filename ) +{ + DB* dbp = pthis->dbp; + int fd; + struct stat st; + char* pbuf; + char* pbegin; + char* pend; + rec_t r; + DBT key; + DBT val; + char szword[MAXWORDLEN+1]; + + if( pthis->dbp == NULL ) + { + return false; + } + if( (fd = open( filename, O_RDONLY, 0644 )) < 0 ) + { + return false; + } + if( fstat( fd, &st ) != 0 ) + { + goto bail; + } + if( st.st_size == 0 ) + { + goto bail; + } + pbuf = (char*)malloc( st.st_size ); + if( pbuf == NULL ) + { + goto bail; + } + if( read( fd, pbuf, st.st_size ) != st.st_size ) + { + goto bail; + } + + DBT_init( &key ); + DBT_init( &val ); + + if( sscanf( pbuf, BOGOFILTER_HEADER, &pthis->nmsgs ) != 1 ) + { + goto bail; + } + pbegin = pbuf; + while( *pbegin != '\n' ) pbegin++; + pbegin++; + + char2DBT( &key, MSGCOUNT_KEY ); + uint2DBT( &val, &pthis->nmsgs ); + if( dbx_put( dbp, &key, &val ) != 0 ) + { + goto bail; + } + + while( pbegin < pbuf + st.st_size ) + { + pend = pbegin; + r.w.p = pbegin; + r.w.len = 0; + r.n = 0; + + while( *pend != '\n' ) + { + if( pend >= pbuf + st.st_size ) + { + goto bail; + } + *pend = tolower(*pend); + if( *pend == ' ' ) + { + r.w.len = (pend-pbegin); + r.n = strtol( pend+1, NULL, 10 ); + } + pend++; + } + if( pend > pbegin && *pbegin != '#' && *pbegin != ';' ) + { + if( r.w.len == 0 || r.w.len > MAXWORDLEN ) + { + fprintf( stderr, "dbh_loadfile: bad file format\n" ); + goto bail; + } + strncpylwr( szword, r.w.p, r.w.len ); + szword[r.w.len] = '\0'; + char2DBT( &key, szword ); + uint2DBT( &val, &r.n ); + if( dbx_put( dbp, &key, &val ) != 0 ) + { + goto bail; + } + } + pbegin = pend+1; + } + + free( pbuf ); + close( fd ); + + return true; + +bail: + return false; +} + +bool_t dbdb_table_export( dbtdb_t* pthis, cpchar filename ) +{ + DB* dbp = pthis->dbp; + int fd; + char iobuf[IOBUFSIZE]; + char* p; + + DBC* csrp; + int rc; + DBT key; + DBT val; + + if( (fd = open( filename, O_CREAT|O_WRONLY|O_TRUNC, 0644 )) < 0 ) + { + goto bail; + } + if( dbx_createcursor( dbp, csrp ) != 0 ) + { + goto bail; + } + + DBT_init( &key ); + DBT_init( &val ); + + p = iobuf; + p += sprintf( p, BOGOFILTER_HEADER, pthis->nmsgs ); + + rc = dbx_first( csrp, &key, &val ); + while( rc == 0 ) + { + assert( key.data != NULL && key.size <= MAXWORDLEN ); + assert( val.data != NULL && val.size == sizeof(uint) ); + if( key.size != MSGCOUNT_KEY_LEN || + memcmp( key.data, MSGCOUNT_KEY, MSGCOUNT_KEY_LEN ) != 0 ) + { + memcpy( p, key.data, key.size ); p += key.size; + *p++ = ' '; + p += sprintf( p, "%u\n", DBT2uint(&val) ); + if( p+TEXTDB_MAXLINELEN > (iobuf+1) ) + { + write( fd, iobuf, p-iobuf ); + p = iobuf; + } + } + rc = dbx_next( csrp, &key, &val ); + } + dbx_destroycursor( csrp ); + if( p != iobuf ) + { + write( fd, iobuf, p-iobuf ); + } + close( fd ); + return true; + +bail: + return false; +} + +uint dbdb_table_getmsgcount( dbtdb_t* pthis ) +{ + return pthis->nmsgs; +} + +uint dbdb_table_getcount( dbtdb_t* pthis, str_t* pword ) +{ + DB* dbp = pthis->dbp; + DBT key; + DBT val; + + char szword[MAXWORDLEN+1]; + uint count = 0; + + assert( pword->len <= MAXWORDLEN ); + strncpylwr( szword, pword->p, pword->len ); + szword[pword->len] = '\0'; + count = 0; + + DBT_init( &key ); + DBT_init( &val ); + + char2DBT( &key, szword ); + if( dbx_get( dbp, &key, &val ) == 0 ) + { + count = DBT2uint( &val ); + } + + return count; +} + +#else /* def HAVE_LIBDB */ + +dbh_t* dbdb_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ) +{ + return NULL; +} + +#endif /* def HAVE_LIBDB */ + +#ifdef UNIT_TEST +int main( int argc, char** argv ) +{ + dbh_t* pdb; + veciter_t iter; + str_t* pstr; + uint n; + + if( argc != 2 ) + { + fprintf( stderr, "usage: %s <file>\n", argv[0] ); + return 1; + } + + for( n = 0; n < 100; n++ ) + { + pdb = dbh_open( "testlist", true ); + + vec_first( &db, &iter ); + while( (pstr = veciter_get( &iter )) != NULL ) + { + char buf[MAXWORDLEN+32]; + char* p; + if( pstr->len > 200 ) + { + fprintf( stderr, "str too long: %u chars\n", pstr->len ); + break; + } + p = buf; + strcpy( buf, "str: " ); + p += 6; + memcpy( p, pstr->p, pstr->len ); + p += pstr->len; + sprintf( p, " %u", pstr->count ); + puts( buf ); + + veciter_next( &iter ); + } + + dbh_close( &db ); + } + + return 0; +} +#endif /* def UNIT_TEST */ diff --git a/dbdb.h b/dbdb.h @@ -0,0 +1,61 @@ +/* $Id: dbdb.h,v 1.7 2002/10/14 22:17:19 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _DBDB_H +#define _DBDB_H + +#ifdef HAVE_LIBDB + +#include <db.h> + +typedef struct _dbtdb dbtdb_t; +struct _dbtdb +{ + bool_t (*close)(dbtdb_t*); + bool_t (*mergeclose)(dbtdb_t*,vec_t*); + bool_t (*unmergeclose)(dbtdb_t*,vec_t*); + bool_t (*import)(dbtdb_t*,cpchar); + bool_t (*export)(dbtdb_t*,cpchar); + uint (*getmsgcount)(dbtdb_t*); + uint (*getcount)(dbtdb_t*,str_t*); + + DB* dbp; /* db handle */ +#if defined(DB_VERSION_MAJOR) && DB_VERSION_MAJOR >= 3 + DB_ENV* envp; /* we don't own this */ +#endif /* DB_VERSION_MAJOR */ + uint nmsgs; /* number of messages in table (cached) */ +}; + +typedef struct _dbhdb dbhdb_t; +struct _dbhdb +{ + bool_t (*close)(dbhdb_t*); + dbt_t* (*opentable)(dbhdb_t*,cpchar,bool_t); + + char* dir; /* directory for db files */ +#if defined(DB_VERSION_MAJOR) && DB_VERSION_MAJOR >= 3 + DB_ENV* envp; /* db environment */ +#endif /* DB_VERSION_MAJOR */ +}; + +dbh_t* dbdb_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ); +bool_t dbdb_db_close( dbhdb_t* pthis ); +dbt_t* dbdb_db_opentable( dbhdb_t* pthis, cpchar table, bool_t rdonly ); + +bool_t dbdb_table_close( dbtdb_t* pthis ); +bool_t dbdb_table_mergeclose( dbtdb_t* pthis, vec_t* pmsg ); +bool_t dbdb_table_unmergeclose( dbtdb_t* pthis, vec_t* pmsg ); +bool_t dbdb_table_import( dbtdb_t* pthis, cpchar filename ); +bool_t dbdb_table_export( dbtdb_t* pthis, cpchar filename ); +uint dbdb_table_getmsgcount( dbtdb_t* pthis ); +uint dbdb_table_getcount( dbtdb_t* pthis, str_t* pword ); + +#endif /* def HAVE_LIBDB */ + +#endif /* ndef _DBDB_H */ diff --git a/dbg.c b/dbg.c @@ -0,0 +1,302 @@ +/* $Id: dbg.c,v 1.3 2002/10/19 08:30:57 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * dbg.c: debug functions for bmf. + */ + +#include "config.h" +#include "dbg.h" +#include <stdarg.h> + + +uint g_verbose = 0; + +void verbose( int level, const char* fmt, ... ) +{ + if( g_verbose >= level ) + { + char str[4096]; + va_list v; + va_start( v, fmt ); + vsnprintf( str, sizeof(str)-1, fmt, v ); + str[sizeof(str)-1] = '\0'; +#ifdef _UNIX + fputs( str, stderr ); +#endif +#ifdef _WIN32 + ::OutputDebugString( str ); +#endif + } +} + +#ifndef NDEBUG + +void dbgout( const char* fmt, ... ) +{ + char str[4096]; + va_list v; + va_start( v, fmt ); + vsnprintf( str, sizeof(str)-1, fmt, v ); + str[sizeof(str)-1] = '\0'; +#ifdef _UNIX + fputs( str, stderr ); +#endif +#ifdef _WIN32 + ::OutputDebugString( str ); +#endif +} + +/* + * Heap management routines. These routines use unbalanced binary trees to + * keep track of allocations in an attempt to make them fast yet simple. + * + * Each block of memory consists of an alloc_node header, the requested + * memory block, and guard bytes before and after the requested memory + * block. The requested memory block is filled with a semi-random byte + * value to ensure that the caller does not rely on any particular initial + * bit pattern (eg. a block of zeros or NULLs). It is refilled with a + * (possibly different) byte value after deallocation to ensure that the + * caller doesn't attempt to use the freed memory. + */ + +/* we need to use the real malloc and free */ +#undef malloc +#undef free + +typedef struct _alloc_node +{ + struct _alloc_node* lptr; + struct _alloc_node* rptr; + size_t len; + cpchar file; + uint line; +} alloc_node; + +static alloc_node* g_heap = NULL; + +/* Our magic guard bytes */ +static byte g_guard[] = +{ + 0xDE, 0xAD, 0xBE, 0xEF, 0xDE, 0xAD, 0xBE, 0xEF, + 0xDE, 0xAD, 0xBE, 0xEF, 0xDE, 0xAD, 0xBE, 0xEF +}; + +void* debug_malloc( cpchar file, uint line, size_t n, int fill ) +{ + byte* pmem = NULL; + alloc_node* pnode; + + pmem = NULL; + if( n == 0 ) + { + n = 1; + } + pnode = (alloc_node*)malloc( n + 2*sizeof(g_guard) + sizeof(alloc_node) ); + if( pnode != NULL ) + { + alloc_node** ppuplink; + alloc_node* pcur; + + pmem = (byte*)pnode + sizeof(alloc_node) + sizeof(g_guard); + memcpy( pmem - sizeof(g_guard), g_guard, sizeof(g_guard) ); + memset( pmem, fill, n ); + memcpy( pmem + n, g_guard, sizeof(g_guard) ); + + pnode->lptr = pnode->rptr = NULL; + pnode->len = n; + pnode->file = file; + pnode->line = line; + ppuplink = &g_heap; + pcur = g_heap; + while( pcur != NULL ) + { + if( pnode == pcur ) + { + dbgout( "%s(%u): *** FATAL: duplicate memory allocated ***\n", file, line ); + assert( false ); + exit( -1 ); + } + if( pnode < pcur ) + { + ppuplink = &pcur->lptr; + pcur = pcur->lptr; + } + else + { + ppuplink = &pcur->rptr; + pcur = pcur->rptr; + } + } + *ppuplink = pnode; + } + + return pmem; +} + +void debug_free( cpchar file, uint line, void* p ) +{ + alloc_node** ppuplink; + alloc_node* pcur; + + if( p == NULL ) + { + return; + } + if( g_heap == NULL ) + { + dbgout( "%s(%u): *** FATAL: delete with empty heap ***\n", file, line ); + assert( false ); + exit( -1 ); + } + + ppuplink = &g_heap; + pcur = g_heap; + while( pcur != NULL ) + { + void* pcurblk = (char*)pcur + sizeof(alloc_node) + sizeof(g_guard); + if( p == pcurblk ) + { + byte* pmem = (byte*)p; + if( memcmp( pmem - sizeof(g_guard), g_guard, sizeof(g_guard) ) != 0 || + memcmp( pmem + pcur->len, g_guard, sizeof(g_guard) ) != 0 ) + { + dbgout( "%s(%u): *** FATAL: corrupted memory at %p\n", file, line, p ); + assert( false ); + exit( -1 ); + } + memset( pmem, rand(), pcur->len ); + if( pcur->lptr && pcur->rptr ) + { + /* + * node has both ptrs so replace it with left child and move + * right child to bottom right of left child's tree + */ + alloc_node* pend = pcur->lptr; + while( pend->rptr ) pend = pend->rptr; + *ppuplink = pcur->lptr; + pend->rptr = pcur->rptr; + } + else + { + /* move child up */ + *ppuplink = (pcur->lptr) ? pcur->lptr : pcur->rptr; + } + free( pcur ); + return; + } + if( p < pcurblk ) + { + ppuplink = &pcur->lptr; + pcur = pcur->lptr; + } + else + { + ppuplink = &pcur->rptr; + pcur = pcur->rptr; + } + } + + dbgout( "%s(%u): *** FATAL: delete on unalloced memory ***\n", file, line ); + assert( false ); + exit( -1 ); +} + +void* debug_realloc( cpchar file, uint line, void* p, size_t n ) +{ + void* pnew; + + if( p == NULL ) + { + pnew = debug_malloc( file, line, n, rand() ); + } + else if( n == 0 ) + { + debug_free( file, line, p ); + pnew = NULL; + } + else + { + alloc_node* pnode = (alloc_node*)((char*)p-sizeof(g_guard)-sizeof(alloc_node)); + pnew = debug_malloc( file, line, n, rand() ); + if( pnew != NULL ) + { + memcpy( pnew, p, pnode->len ); + debug_free( file, line, p ); + } + } + + return pnew; +} + +char* debug_strdup( cpchar file, uint line, cpchar s ) +{ + char* s2; + uint sl = strlen(s); + + s2 = (char*)debug_malloc( file, line, sl+1, 0 ); + memcpy( s2, s, sl ); + s2[sl] = '\0'; + + return s2; +} + +char* debug_strndup( cpchar file, uint line, cpchar s, size_t n ) +{ + char* s2; + uint sl = strlen(s); + + sl = min( n-1, sl ); + s2 = (char*)debug_malloc( file, line, sl+1, 0 ); + memcpy( s2, s, sl ); + s2[sl] = '\0'; + + return s2; +} + +static void walk_alloc_tree( alloc_node* pcur, size_t* pttl ) +{ + if( pcur != NULL ) + { + walk_alloc_tree( pcur->lptr, pttl ); + dbgout( "%s(%u): %u bytes at %p\n", pcur->file, pcur->line, + pcur->len, pcur+sizeof(alloc_node)+sizeof(g_guard) ); + *pttl += pcur->len; + walk_alloc_tree( pcur->rptr, pttl ); + } +} + +void dump_alloc_heap( void ) +{ + if( g_heap != NULL ) + { + size_t ttl = 0; + dbgout( "\n" ); + dbgout( "Memory leaks detected\n" ); + dbgout( "=====================\n" ); + dbgout( "\n" ); + walk_alloc_tree( g_heap, &ttl ); + dbgout( "\n" ); + dbgout( "=====================\n" ); + dbgout( "Total bytes: %u\n", ttl ); + dbgout( "=====================\n" ); + } +} + +#else /* ndef NDEBUG */ + +void dbgout( const char* fmt, ... ) +{ + /* empty */ +} + +void dump_alloc_heap( void ) +{ + /* empty */ +} + +#endif /* ndef NDEBUG */ diff --git a/dbg.h b/dbg.h @@ -0,0 +1,35 @@ +/* $Id: dbg.h,v 1.1 2002/10/14 07:09:51 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _DBG_H +#define _DBG_H + +extern uint g_verbose; + +void verbose( int level, const char* fmt, ... ); + +void dbgout( const char* fmt, ... ); +void dump_alloc_heap( void ); + +#ifndef NDEBUG +void* debug_malloc ( cpchar file, uint line, size_t n, int fill ); +void debug_free ( cpchar file, uint line, void* p ); +void* debug_realloc ( cpchar file, uint line, void* p, size_t n ); +char* debug_strdup ( cpchar file, uint line, cpchar s ); +char* debug_strndup ( cpchar file, uint line, cpchar s, size_t n ); + +#define malloc(n) debug_malloc (__FILE__,__LINE__,n,rand()) +#define calloc(n) debug_calloc (__FILE__,__LINE__,n,0) +#define free(p) debug_free (__FILE__,__LINE__,p) +#define realloc(p,n) debug_realloc (__FILE__,__LINE__,p,n) +#define strdup(s) debug_strdup (__FILE__,__LINE__,s) +#define strndup(s,n) debug_strndup (__FILE__,__LINE__,s,n) +#endif /* ndef NDEBUG */ + +#endif /* ndef _DBG_H */ diff --git a/dbh.c b/dbh.c @@ -0,0 +1,74 @@ +/* $Id: dbh.c,v 1.2 2002/10/14 07:09:51 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * dbh.c: database handler interface + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" + +#include "dbh.h" +#include "dbtext.h" +#include "dbdb.h" +#include "dbmysql.h" + +/* + * get count for new (incoming) word. there may be duplicate entries for the + * str, so sum the counts and leave the iterator at the last one. + * + * the list referenced in the iterator must be sorted. + */ +uint db_getnewcount( veciter_t* piter ) +{ + str_t* pstr; + uint count; + veciter_t curiter; + str_t* pcurstr; + + pstr = &piter->plist->pitems[piter->index]; + count = 0; + + curiter.plist = piter->plist; + curiter.index = piter->index; + pcurstr = &curiter.plist->pitems[curiter.index]; + + while( curiter.index < curiter.plist->nitems && str_casecmp( pstr, pcurstr ) == 0 ) + { + piter->index = curiter.index; + count = min( MAXFREQ, count + 1 ); + veciter_next( &curiter ); + pcurstr = &curiter.plist->pitems[curiter.index]; + } + + return count; +} + +dbh_t* dbh_open( dbfmt_t dbfmt, cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ) +{ + dbh_t* pdb = NULL; + + switch( dbfmt ) + { + case db_text: + pdb = (dbh_t*)dbtext_db_open( dbhost, dbname, dbuser, dbpass ); + break; + case db_db: + pdb = (dbh_t*)dbdb_db_open( dbhost, dbname, dbuser, dbpass ); + break; + case db_mysql: + pdb = (dbh_t*) dbmysql_db_open( dbhost, dbname, dbuser, dbpass ); + break; + default: + assert(false); + } + + return pdb; +} diff --git a/dbh.h b/dbh.h @@ -0,0 +1,56 @@ +/* $Id: dbh.h,v 1.3 2002/10/02 04:45:40 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _DBH_H +#define _DBH_H + +/* database formats */ +typedef enum +{ + db_text, /* flat text */ + db_db, /* libdb */ + db_mysql /* mysql */ +} dbfmt_t; + +/* record/field structure */ +typedef struct _rec +{ + str_t w; + uint n; +} rec_t; + +/* database table */ +typedef struct _dbt dbt_t; +struct _dbt +{ + bool_t (*close)(dbt_t*); + bool_t (*mergeclose)(dbt_t*,vec_t*); + bool_t (*unmergeclose)(dbt_t*,vec_t*); + bool_t (*import)(dbt_t*,cpchar); + bool_t (*export)(dbt_t*,cpchar); + uint (*getmsgcount)(dbt_t*); + uint (*getcount)(dbt_t*,str_t*); +}; + +/* database instance */ +typedef struct _dbh dbh_t; +struct _dbh +{ + bool_t (*close)(dbh_t*); + dbt_t* (*opentable)(dbh_t*,cpchar,bool_t); +}; + +dbh_t* dbh_open( dbfmt_t dbfmt, cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ); + +#define BOGOFILTER_HEADER "# bogofilter wordlist (format version A): %u\n" +#define TEXTDB_MAXLINELEN (MAXWORDLEN+32) + +uint db_getnewcount( veciter_t* piter ); + +#endif /* ndef _DBH_H */ diff --git a/dbmysql.c b/dbmysql.c @@ -0,0 +1,545 @@ +/* $Id: dbmysql.c,v 1.9 2002/10/14 07:09:51 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * dbmysql.c: mysql database handler + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" + +#include "dbh.h" +#include "dbmysql.h" + +#ifdef HAVE_MYSQL + +#define MAXQUERY 256 + +static MYSQL* g_mysql = NULL; + +static void sql_escape( char* d, const char* s ) +{ + while( *s != '\0' ) + { + if( *s == '\'' ) + { + *d++ = '\''; + } + *d++ = tolower(*s++); + } +} + +dbh_t* dbmysql_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ) +{ + dbhmysql_t* pthis; + + if( g_mysql == NULL ) + { + g_mysql = mysql_init( NULL ); + if( g_mysql == NULL ) + { + return NULL; + } + } + + pthis = (dbhmysql_t*)malloc( sizeof(dbhmysql_t) ); + if( pthis == NULL ) + { + perror( "malloc()" ); + goto bail; + } + pthis->close = dbmysql_db_close; + pthis->opentable = dbmysql_db_opentable; + + pthis->dbh = mysql_real_connect( g_mysql, dbhost, dbuser, dbpass, dbname, 0, NULL, 0 ); + if( pthis->dbh == NULL ) + { + goto bail; + } + + + return (dbh_t*)pthis; + +bail: + fprintf( stderr, "cannot open mysql database '%s': %s\n", dbname, mysql_error(g_mysql) ); + free( pthis ); + return NULL; +} + +bool_t dbmysql_db_close( dbhmysql_t* pthis ) +{ + if( pthis->dbh != NULL ) + { + mysql_close( pthis->dbh ); + pthis->dbh = NULL; + } + return true; +} + +dbt_t* dbmysql_db_opentable( dbhmysql_t* pthis, cpchar table, bool_t rdonly ) +{ + dbtmysql_t* ptable; + + char query[MAXQUERY]; + MYSQL_RES* res; + MYSQL_ROW row; + + ptable = (dbtmysql_t*)malloc( sizeof(dbtmysql_t) ); + if( ptable == NULL ) + { + return NULL; + } + ptable->close = dbmysql_table_close; + ptable->mergeclose = dbmysql_table_mergeclose; + ptable->unmergeclose = dbmysql_table_unmergeclose; + ptable->import = dbmysql_table_import; + ptable->export = dbmysql_table_export; + ptable->getmsgcount = dbmysql_table_getmsgcount; + ptable->getcount = dbmysql_table_getcount; + ptable->pdb = pthis; + ptable->table = strdup( table ); + ptable->nmsgs = 0; + + sprintf( query, "SELECT count FROM %s WHERE name='%s'", + table, MSGCOUNT_KEY ); + if( mysql_query( pthis->dbh, query ) != 0 ) + { + goto bail; + } + if( (res = mysql_store_result( pthis->dbh )) == NULL ) + { + goto bail; + } + while( (row = mysql_fetch_row( res )) ) + { + ptable->nmsgs = atoi( row[0] ); + } + + return (dbt_t*)ptable; + +bail: + free( ptable->table ); + free( ptable ); + return NULL; +} + +bool_t dbmysql_table_close( dbtmysql_t* pthis ) +{ + if( pthis->pdb != NULL ) + { + free( pthis->table ); + pthis->table = NULL; + pthis->pdb = NULL; + } + return true; +} + +bool_t dbmysql_table_mergeclose( dbtmysql_t* pthis, vec_t* pmsg ) +{ + char szword[MAXWORDLEN+1]; + char szsqlword[MAXWORDLEN*2+1]; + veciter_t msgiter; + str_t* pmsgstr; + + char query[MAXQUERY]; + uint count; + + if( pthis->pdb == NULL || pthis->pdb->dbh == NULL ) + { + assert( false ); + return false; + } + + pthis->nmsgs++; + + sprintf( query, "UPDATE %s SET count=%u WHERE name='%s'", + pthis->table, pthis->nmsgs, MSGCOUNT_KEY ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + sprintf( query, "INSERT INTO %s ( name, count ) VALUES ( '%s', %u )", + pthis->table, MSGCOUNT_KEY, pthis->nmsgs ); + mysql_query( pthis->pdb->dbh, query ); + } + + vec_first( pmsg, &msgiter ); + pmsgstr = veciter_get( &msgiter ); + + while( pmsgstr != NULL ) + { + assert( pmsgstr->len <= MAXWORDLEN ); + strncpylwr( szword, pmsgstr->p, pmsgstr->len ); + szword[pmsgstr->len] = '\0'; + sql_escape( szsqlword, szword ); + count = db_getnewcount( &msgiter ); + + sprintf( query, "UPDATE %s SET count=count+%u WHERE name='%s'", + pthis->table, count, szsqlword ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + sprintf( query, "INSERT INTO %s ( name, count ) VALUES ( '%s', %u )", + pthis->table, szsqlword, count ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + goto bail; + } + } + + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + + veciter_destroy( &msgiter ); + return dbmysql_table_close( pthis ); + +bail: + return false; +} + +bool_t dbmysql_table_unmergeclose( dbtmysql_t* pthis, vec_t* pmsg ) +{ + char szword[MAXWORDLEN+1]; + char szsqlword[MAXWORDLEN*2+1]; + veciter_t msgiter; + str_t* pmsgstr; + + char query[MAXQUERY]; + uint count; + + if( pthis->pdb == NULL || pthis->pdb->dbh == NULL ) + { + assert( false ); + return false; + } + + if( pthis->nmsgs > 0 ) + { + pthis->nmsgs--; + } + + sprintf( query, "UPDATE %s SET count=%u WHERE name='%s'", + pthis->table, pthis->nmsgs, MSGCOUNT_KEY ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + sprintf( query, "INSERT INTO %s ( name, count ) VALUES ( '%s', %u )", + pthis->table, MSGCOUNT_KEY, pthis->nmsgs ); + mysql_query( pthis->pdb->dbh, query ); + } + + vec_first( pmsg, &msgiter ); + pmsgstr = veciter_get( &msgiter ); + + while( pmsgstr != NULL ) + { + assert( pmsgstr->len <= MAXWORDLEN ); + strncpylwr( szword, pmsgstr->p, pmsgstr->len ); + szword[pmsgstr->len] = '\0'; + sql_escape( szsqlword, szword ); + count = db_getnewcount( &msgiter ); + + sprintf( query, "UPDATE %s SET count=GREATEST(0,count-%u) WHERE name='%s'", + pthis->table, count, szsqlword ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + /* this should not happen, so write with count=0 */ + sprintf( query, "INSERT INTO %s ( name, count ) VALUES ( '%s', 0 )", + pthis->table, szsqlword ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + goto bail; + } + } + + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + + veciter_destroy( &msgiter ); + return dbmysql_table_close( pthis ); + +bail: + return false; +} + +bool_t dbmysql_table_import( dbtmysql_t* pthis, cpchar filename ) +{ + int fd; + struct stat st; + char* pbuf; + char* pbegin; + char* pend; + rec_t r; + char szword[MAXWORDLEN+1]; + char szsqlword[MAXWORDLEN*2+1]; + char query[MAXQUERY]; + + if( pthis->pdb == NULL || pthis->pdb->dbh == NULL ) + { + return false; + } + if( (fd = open( filename, O_RDONLY, 0644 )) < 0 ) + { + return false; + } + if( fstat( fd, &st ) != 0 ) + { + goto bail; + } + if( st.st_size == 0 ) + { + goto bail; + } + pbuf = (char*)malloc( st.st_size ); + if( pbuf == NULL ) + { + goto bail; + } + if( read( fd, pbuf, st.st_size ) != st.st_size ) + { + goto bail; + } + + if( sscanf( pbuf, BOGOFILTER_HEADER, &pthis->nmsgs ) != 1 ) + { + goto bail; + } + pbegin = pbuf; + while( *pbegin != '\n' ) pbegin++; + pbegin++; + + sprintf( query, "INSERT INTO %s ( name, count ) VALUES ( '%s', %u )", + pthis->table, MSGCOUNT_KEY, pthis->nmsgs ); + mysql_query( pthis->pdb->dbh, query ); + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + goto bail; + } + + while( pbegin < pbuf + st.st_size ) + { + pend = pbegin; + r.w.p = pbegin; + r.w.len = 0; + r.n = 0; + + while( *pend != '\n' ) + { + if( pend >= pbuf + st.st_size ) + { + goto bail; + } + *pend = tolower(*pend); + if( *pend == ' ' ) + { + r.w.len = (pend-pbegin); + r.n = strtol( pend+1, NULL, 10 ); + } + pend++; + } + if( pend > pbegin && *pbegin != '#' && *pbegin != ';' ) + { + if( r.w.len == 0 || r.w.len > MAXWORDLEN ) + { + fprintf( stderr, "dbh_loadfile: bad file format\n" ); + goto bail; + } + strncpylwr( szword, r.w.p, r.w.len ); + szword[r.w.len] = '\0'; + sql_escape( szsqlword, szword ); + + sprintf( query, "INSERT INTO %s ( name, count ) VALUES ( '%s', %u )", + pthis->table, szsqlword, r.n ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + goto bail; + } + } + pbegin = pend+1; + } + + return true; + +bail: + return false; +} + +bool_t dbmysql_table_export( dbtmysql_t* pthis, cpchar filename ) +{ + int fd; + char iobuf[IOBUFSIZE]; + char* p; + + char query[MAXQUERY]; + MYSQL_RES* res; + MYSQL_ROW row; + + if( (fd = open( filename, O_CREAT|O_WRONLY|O_TRUNC, 0644 )) < 0 ) + { + return false; + } + + p += sprintf( p, BOGOFILTER_HEADER, pthis->nmsgs ); + + sprintf( query, "SELECT name, count FROM %s", + pthis->table ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( mysql_affected_rows( pthis->pdb->dbh ) == 0 ) + { + goto bail; + } + + while( (row = mysql_fetch_row( res )) ) + { + if( strcmp( row[0], MSGCOUNT_KEY ) == 0 ) + { + continue; + } + + p += sprintf( p, "%s %s\n", row[0], row[1] ); + if( p+TEXTDB_MAXLINELEN > (iobuf+1) ) + { + write( fd, iobuf, p-iobuf ); + p = iobuf; + } + } + if( p != iobuf ) + { + write( fd, iobuf, p-iobuf ); + } + close( fd ); + + return true; + +bail: + return false; +} + +uint dbmysql_table_getmsgcount( dbtmysql_t* pthis ) +{ + return pthis->nmsgs; +} + +uint dbmysql_table_getcount( dbtmysql_t* pthis, str_t* pword ) +{ + uint count = 0; + char szword[MAXWORDLEN+1]; + char szsqlword[MAXWORDLEN*2+1]; + + char query[MAXQUERY]; + MYSQL_RES* res; + MYSQL_ROW row; + + assert( pword->len <= MAXWORDLEN ); + strncpylwr( szword, pword->p, pword->len ); + szword[pword->len] = '\0'; + sql_escape( szsqlword, szword ); + sprintf( query, "SELECT count FROM %s WHERE name='%s'", + pthis->table, szsqlword ); + if( mysql_query( pthis->pdb->dbh, query ) != 0 ) + { + goto bail; + } + if( (res = mysql_store_result( pthis->pdb->dbh )) == NULL ) + { + goto bail; + } + while( (row = mysql_fetch_row( res )) ) + { + count = atoi( row[0] ); + } + +bail: + return count; +} + +#else /* def HAVE_MYSQL */ + +dbh_t* dbmysql_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ) +{ + return NULL; +} + +#endif /* def HAVE_MYSQL */ + +#ifdef UNIT_TEST +int main( int argc, char** argv ) +{ + dbh_t* pdb; + veciter_t iter; + str_t* pstr; + uint n; + + if( argc != 2 ) + { + fprintf( stderr, "usage: %s <file>\n", argv[0] ); + return 1; + } + + for( n = 0; n < 100; n++ ) + { + pdb = dbh_open( "testlist", true ); + + vec_first( &db, &iter ); + while( (pstr = veciter_get( &iter )) != NULL ) + { + char buf[MAXWORDLEN+32]; + char* p; + if( pstr->len > 200 ) + { + fprintf( stderr, "str too long: %u chars\n", pstr->len ); + break; + } + p = buf; + strcpy( buf, "str: " ); + p += 6; + memcpy( p, pstr->p, pstr->len ); + p += pstr->len; + sprintf( p, " %u", pstr->count ); + puts( buf ); + + veciter_next( &iter ); + } + + dbh_close( &db ); + } + + return 0; +} +#endif /* def UNIT_TEST */ diff --git a/dbmysql.h b/dbmysql.h @@ -0,0 +1,60 @@ +/* $Id: dbmysql.h,v 1.4 2002/10/06 06:46:53 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _DBMYSQL_H +#define _DBMYSQL_H + +#ifdef HAVE_MYSQL + +#include "mysql.h" + +typedef struct _dbtmysql dbtmysql_t; +struct _dbtmysql +{ + bool_t (*close)(dbtmysql_t*); + bool_t (*mergeclose)(dbtmysql_t*,vec_t*); + bool_t (*unmergeclose)(dbtmysql_t*,vec_t*); + bool_t (*import)(dbtmysql_t*,cpchar); + bool_t (*export)(dbtmysql_t*,cpchar); + uint (*getmsgcount)(dbtmysql_t*); + uint (*getcount)(dbtmysql_t*,str_t*); + + struct _dbhmysql* pdb; + char* table; /* table name */ + uint nmsgs; /* number of messages in table (cached) */ +}; + +typedef struct _dbhmysql dbhmysql_t; +struct _dbhmysql +{ + bool_t (*close)(dbhmysql_t*); + dbt_t* (*opentable)(dbhmysql_t*,cpchar,bool_t); + + MYSQL* dbh; /* database handle, if currently open */ +}; + +dbh_t* dbmysql_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ); +bool_t dbmysql_db_close( dbhmysql_t* pthis ); +dbt_t* dbmysql_db_opentable( dbhmysql_t* pthis, cpchar table, bool_t rdonly ); + +bool_t dbmysql_table_close( dbtmysql_t* pthis ); +bool_t dbmysql_table_mergeclose( dbtmysql_t* pthis, vec_t* pmsg ); +bool_t dbmysql_table_unmergeclose( dbtmysql_t* pthis, vec_t* pmsg ); +bool_t dbmysql_table_import( dbtmysql_t* pthis, cpchar filename ); +bool_t dbmysql_table_export( dbtmysql_t* pthis, cpchar filename ); +uint dbmysql_table_getmsgcount( dbtmysql_t* pthis ); +uint dbmysql_table_getcount( dbtmysql_t* pthis, str_t* pword ); + +#else /* def HAVE_MYSQL */ + +dbh_t* dbmysql_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ); + +#endif /* def HAVE_MYSQL */ + +#endif /* ndef _DBMYSQL_H */ diff --git a/dbtext.c b/dbtext.c @@ -0,0 +1,591 @@ +/* $Id: dbtext.c,v 1.12 2002/10/19 09:59:35 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * dbtext.c: flatfile database handler + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" + +#include "dbh.h" +#include "dbtext.h" + +static void dbtext_table_setsize( dbttext_t* pthis, uint nsize ) +{ + if( nsize > pthis->nalloc ) + { + uint nnewalloc; + rec_t* pnewitems; + uint n; + + nnewalloc = pthis->nalloc * 2; + if( nnewalloc < nsize ) nnewalloc = nsize; + pnewitems = (rec_t*)realloc( pthis->pitems, nnewalloc*sizeof(rec_t) ); + if( pnewitems == NULL ) + { + exit( 2 ); + } + for( n = pthis->nitems; n < nsize; n++ ) + { + str_create( &pnewitems[n].w ); + pnewitems[n].n = 0; + } + pthis->pitems = pnewitems; + pthis->nalloc = nnewalloc; + } +} + +dbh_t* dbtext_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ) +{ + dbhtext_t* pthis; + + uint dirlen; + cpchar phome; + struct stat st; + + pthis = (dbhtext_t*)malloc( sizeof(dbhtext_t) ); + if( pthis == NULL ) + { + goto bail; + } + pthis->close = dbtext_db_close; + pthis->opentable = dbtext_db_opentable; + if( dbname != NULL && *dbname != '\0' ) + { + dirlen = strlen( dbname ); + pthis->dir = strdup( dbname ); + if( pthis->dir[dirlen-1] == '/' ) + { + pthis->dir[dirlen-1] = '\0'; + } + } + else + { + phome = getenv( "HOME" ); + if( phome == NULL || *phome == '\0' ) + { + phome = "."; + } + pthis->dir = (char*)malloc( strlen(phome)+5+1 ); + if( pthis->dir == NULL ) + { + goto bail; + } + sprintf( pthis->dir, "%s/.bmf", phome ); + } + + /* ensure config directory exists */ + if( stat( pthis->dir, &st ) != 0 ) + { + if( errno == ENOENT ) + { + if( mkdir( pthis->dir, S_IRUSR|S_IWUSR|S_IXUSR ) != 0 ) + { + goto bail; + } + } + else + { + goto bail; + } + } + else + { + if( !S_ISDIR( st.st_mode ) ) + { + goto bail; + } + } + + return (dbh_t*)pthis; + +bail: + return NULL; +} + +bool_t dbtext_db_close( dbhtext_t* pthis ) +{ + free( pthis->dir ); + pthis->dir = NULL; + return true; +} + +dbt_t* dbtext_db_opentable( dbhtext_t* pthis, cpchar table, bool_t rdonly ) +{ + dbttext_t* ptable = NULL; + +#ifndef NOLOCK + struct flock lock; +#endif /* ndef NOLOCK */ + char szpath[PATH_MAX]; + int flags; + struct stat st; + + char* pbegin; + char* pend; + rec_t r; + uint pos; + + if( pthis->dir == NULL ) + { + goto bail; + } + + ptable = (dbttext_t*)malloc( sizeof(dbttext_t) ); + if( ptable == NULL ) + { + perror( "malloc()" ); + goto bail; + } + ptable->close = dbtext_table_close; + ptable->mergeclose = dbtext_table_mergeclose; + ptable->unmergeclose = dbtext_table_unmergeclose; + ptable->import = dbtext_table_import; + ptable->export = dbtext_table_export; + ptable->getmsgcount = dbtext_table_getmsgcount; + ptable->getcount = dbtext_table_getcount; + ptable->fd = -1; + ptable->pbuf = NULL; + ptable->nmsgs = 0; + ptable->nalloc = 0; + ptable->nitems = 0; + ptable->pitems = NULL; + + sprintf( szpath, "%s/%s.txt", pthis->dir, table ); + flags = (rdonly ? O_RDONLY|O_CREAT : O_RDWR|O_CREAT); + ptable->fd = open( szpath, flags, 0644 ); + if( ptable->fd == -1 ) + { + perror( "open()" ); + goto bail; + } + +#ifndef NOLOCK + memset( &lock, 0, sizeof(lock) ); + lock.l_type = rdonly ? F_RDLCK : F_WRLCK; + lock.l_start = 0; + lock.l_whence = SEEK_SET; + lock.l_len = 0; + fcntl( ptable->fd, F_SETLKW, &lock ); +#endif /* ndef NOLOCK */ + + if( fstat( ptable->fd, &st ) != 0 ) + { + perror( "fstat()" ); + goto bail_uc; + } + + if( st.st_size == 0 ) + { + return (dbt_t*)ptable; + } + + ptable->pbuf = (char*)malloc( st.st_size ); + if( ptable->pbuf == NULL ) + { + perror( "malloc()" ); + goto bail_uc; + } + + if( read( ptable->fd, ptable->pbuf, st.st_size ) != st.st_size ) + { + perror( "read()" ); + goto bail_fuc; + } + + /* XXX: bogofilter compatibility */ + if( sscanf( ptable->pbuf, BOGOFILTER_HEADER, &ptable->nmsgs ) != 1 ) + { + goto bail_fuc; + } + pbegin = ptable->pbuf; + while( *pbegin != '\n' ) pbegin++; + pbegin++; + + pos = 0; + while( pbegin < ptable->pbuf + st.st_size ) + { + pend = pbegin; + r.w.p = pbegin; + r.w.len = 0; + r.n = 0; + + while( *pend != '\n' ) + { + if( pend >= ptable->pbuf + st.st_size ) + { + goto bail_fuc; + } + *pend = tolower(*pend); + if( *pend == ' ' ) + { + r.w.len = (pend-pbegin); + r.n = strtol( pend+1, NULL, 10 ); + } + pend++; + } + if( pend > pbegin && *pbegin != '#' && *pbegin != ';' ) + { + if( r.w.len == 0 || r.w.len > MAXWORDLEN ) + { + fprintf( stderr, "dbh_loadfile: bad file format\n" ); + goto bail_fuc; + } + dbtext_table_setsize( ptable, pos+1 ); + ptable->pitems[pos++] = r; + ptable->nitems = pos; + } + pbegin = pend+1; + } + + if( rdonly ) + { +#ifndef NOLOCK + lock.l_type = F_UNLCK; + fcntl( ptable->fd, F_SETLKW, &lock ); +#endif /* ndef NOLOCK */ + close( ptable->fd ); + ptable->fd = -1; + } + + return (dbt_t*)ptable; + +bail_fuc: + free( ptable->pbuf ); + +bail_uc: +#ifndef NOLOCK + lock.l_type = F_UNLCK; + fcntl( ptable->fd, F_SETLKW, &lock ); +#endif /* ndef NOLOCK */ + + close( ptable->fd ); + ptable->fd = -1; + +bail: + free( ptable ); + return NULL; +} + +bool_t dbtext_table_close( dbttext_t* pthis ) +{ + struct flock lockall; + + free( pthis->pbuf ); + pthis->pbuf = NULL; + free( pthis->pitems ); + pthis->pitems = NULL; + + if( pthis->fd != -1 ) + { +#ifndef NOLOCK + memset( &lockall, 0, sizeof(lockall) ); + lockall.l_type = F_UNLCK; + lockall.l_start = 0; + lockall.l_whence = SEEK_SET; + lockall.l_len = 0; + fcntl( pthis->fd, F_SETLKW, &lockall ); +#endif /* ndef NOLOCK */ + close( pthis->fd ); + pthis->fd = -1; + } + + return true; +} + +bool_t dbtext_table_mergeclose( dbttext_t* pthis, vec_t* pmsg ) +{ + /* note that we require both vectors to be sorted */ + + uint pos; + rec_t* prec; + veciter_t msgiter; + str_t* pmsgstr; + uint count; + char iobuf[IOBUFSIZE]; + char* p; + + if( pthis->fd == -1 ) + { + return false; + } + ftruncate( pthis->fd, 0 ); + lseek( pthis->fd, 0, SEEK_SET ); + + pthis->nmsgs++; + + p = iobuf; + p += sprintf( p, BOGOFILTER_HEADER, pthis->nmsgs ); + + vec_first( pmsg, &msgiter ); + pmsgstr = veciter_get( &msgiter ); + + pos = 0; + while( pos < pthis->nitems || pmsgstr != NULL ) + { + int cmp = 0; + prec = &pthis->pitems[pos]; + if( pmsgstr != NULL && pos < pthis->nitems ) + { + cmp = str_casecmp( &prec->w, pmsgstr ); + } + else + { + /* we exhausted one list or the other (but not both) */ + cmp = (pos < pthis->nitems) ? -1 : 1; + } + if( cmp < 0 ) + { + /* write existing str */ + assert( prec->w.p != NULL && prec->w.len > 0 ); + assert( prec->w.len <= MAXWORDLEN ); + count = prec->n; + strncpylwr( p, prec->w.p, prec->w.len ); p += prec->w.len; + *p++ = ' '; + p += sprintf( p, "%u\n", count ); + + pos++; + } + else if( cmp == 0 ) + { + /* same str, merge and write sum */ + assert( prec->w.p != NULL && prec->w.len > 0 ); + assert( pmsgstr->p != NULL && pmsgstr->len > 0 ); + assert( prec->w.len <= MAXWORDLEN ); + assert( pmsgstr->len <= MAXWORDLEN ); + count = db_getnewcount( &msgiter ); + count += prec->n; + strncpylwr( p, prec->w.p, prec->w.len ); p += prec->w.len; + *p++ = ' '; + p += sprintf( p, "%u\n", count ); + + pos++; + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + else /* cmp > 0 */ + { + /* write new str */ + assert( pmsgstr->p != NULL && pmsgstr->len > 0 ); + assert( pmsgstr->len <= MAXWORDLEN ); + count = db_getnewcount( &msgiter ); + strncpylwr( p, pmsgstr->p, pmsgstr->len ); p += pmsgstr->len; + *p++ = ' '; + p += sprintf( p, "%u\n", count ); + + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + + if( p+TEXTDB_MAXLINELEN > (iobuf+1) ) + { + write( pthis->fd, iobuf, p-iobuf ); + p = iobuf; + } + } + if( p != iobuf ) + { + write( pthis->fd, iobuf, p-iobuf ); + } + + veciter_destroy( &msgiter ); + return dbtext_table_close( pthis ); +} + +bool_t dbtext_table_unmergeclose( dbttext_t* pthis, vec_t* pmsg ) +{ + /* note that we require both vectors to be sorted */ + + uint pos; + rec_t* prec; + veciter_t msgiter; + str_t* pmsgstr; + uint count; + char iobuf[IOBUFSIZE]; + char* p; + + if( pthis->fd == -1 ) + { + return false; + } + ftruncate( pthis->fd, 0 ); + lseek( pthis->fd, 0, SEEK_SET ); + + pthis->nmsgs--; + + p = iobuf; + p += sprintf( p, BOGOFILTER_HEADER, pthis->nmsgs ); + + vec_first( pmsg, &msgiter ); + pmsgstr = veciter_get( &msgiter ); + + pos = 0; + while( pos < pthis->nitems || pmsgstr != NULL ) + { + int cmp = 0; + prec = &pthis->pitems[pos]; + if( pmsgstr != NULL && pos < pthis->nitems ) + { + cmp = str_casecmp( &prec->w, pmsgstr ); + } + else + { + /* we exhausted one list or the other (but not both) */ + cmp = (pos < pthis->nitems) ? -1 : 1; + } + if( cmp < 0 ) + { + /* write existing str */ + assert( prec->w.p != NULL && prec->w.len > 0 ); + assert( prec->w.len <= MAXWORDLEN ); + count = prec->n; + strncpylwr( p, prec->w.p, prec->w.len ); p += prec->w.len; + *p++ = ' '; + p += sprintf( p, "%u\n", count ); + + pos++; + } + else if( cmp == 0 ) + { + /* same str, merge and write difference */ + assert( prec->w.p != NULL && prec->w.len > 0 ); + assert( pmsgstr->p != NULL && pmsgstr->len > 0 ); + assert( prec->w.len <= MAXWORDLEN ); + assert( pmsgstr->len <= MAXWORDLEN ); + count = db_getnewcount( &msgiter ); + count = (prec->n > count) ? (prec->n - count) : 0; + strncpylwr( p, prec->w.p, prec->w.len ); p += prec->w.len; + *p++ = ' '; + p += sprintf( p, "%u\n", count ); + + pos++; + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + else /* cmp > 0 */ + { + /* this should not happen, so write with count=0 */ + assert( pmsgstr->p != NULL && pmsgstr->len > 0 ); + assert( pmsgstr->len <= MAXWORDLEN ); + db_getnewcount( &msgiter ); + count = 0; + strncpylwr( p, pmsgstr->p, pmsgstr->len ); p += pmsgstr->len; + *p++ = ' '; + p += sprintf( p, "%u\n", count ); + + veciter_next( &msgiter ); + pmsgstr = veciter_get( &msgiter ); + } + + if( p+TEXTDB_MAXLINELEN > (iobuf+1) ) + { + write( pthis->fd, iobuf, p-iobuf ); + p = iobuf; + } + } + if( p != iobuf ) + { + write( pthis->fd, iobuf, p-iobuf ); + } + + veciter_destroy( &msgiter ); + return dbtext_table_close( pthis ); +} + +bool_t dbtext_table_import( dbttext_t* pthis, cpchar filename ) +{ + return false; +} + +bool_t dbtext_table_export( dbttext_t* pthis, cpchar filename ) +{ + return false; +} + +uint dbtext_table_getmsgcount( dbttext_t* pthis ) +{ + return pthis->nmsgs; +} + +uint dbtext_table_getcount( dbttext_t* pthis, str_t* pword ) +{ + int lo, hi, mid; + + if( pthis->nitems == 0 ) + { + return 0; + } + + hi = pthis->nitems - 1; + lo = -1; + while( hi-lo > 1 ) + { + mid = (hi+lo)/2; + if( str_casecmp( pword, &pthis->pitems[mid].w ) <= 0 ) + hi = mid; + else + lo = mid; + } + assert( hi >= 0 && hi < pthis->nitems ); + + if( str_casecmp( pword, &pthis->pitems[hi].w ) != 0 ) + { + return 0; + } + + return pthis->pitems[hi].n; +} + +#ifdef UNIT_TEST +int main( int argc, char** argv ) +{ + dbh_t* pdb; + veciter_t iter; + str_t* pstr; + uint n; + + if( argc != 2 ) + { + fprintf( stderr, "usage: %s <file>\n", argv[0] ); + return 1; + } + + for( n = 0; n < 100; n++ ) + { + pdb = dbh_open( "testlist", true ); + + vec_first( &db, &iter ); + while( (pstr = veciter_get( &iter )) != NULL ) + { + char buf[MAXWORDLEN+32]; + char* p; + if( pstr->len > 200 ) + { + fprintf( stderr, "str too long: %u chars\n", pstr->len ); + break; + } + p = buf; + strcpy( buf, "str: " ); + p += 6; + memcpy( p, pstr->p, pstr->len ); + p += pstr->len; + sprintf( p, " %u", pstr->count ); + puts( buf ); + + veciter_next( &iter ); + } + + dbh_close( &db ); + } + + return 0; +} +#endif /* def UNIT_TEST */ diff --git a/dbtext.h b/dbtext.h @@ -0,0 +1,53 @@ +/* $Id: dbtext.h,v 1.3 2002/10/02 04:45:40 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _DBTEXT_H +#define _DBTEXT_H + +typedef struct _dbttext dbttext_t; +struct _dbttext +{ + bool_t (*close)(dbttext_t*); + bool_t (*mergeclose)(dbttext_t*,vec_t*); + bool_t (*unmergeclose)(dbttext_t*,vec_t*); + bool_t (*import)(dbttext_t*,cpchar); + bool_t (*export)(dbttext_t*,cpchar); + uint (*getmsgcount)(dbttext_t*); + uint (*getcount)(dbttext_t*,str_t*); + + int fd; /* file descriptor, if currently open */ + char* pbuf; /* data buffer, if currently open */ + uint nmsgs; /* number of messages represented in list */ + uint nalloc; /* items alloced in pitems */ + uint nitems; /* items available */ + rec_t* pitems; /* growing vector of items */ +}; + +typedef struct _dbhtext dbhtext_t; +struct _dbhtext +{ + bool_t (*close)(dbhtext_t*); + dbt_t* (*opentable)(dbhtext_t*,cpchar,bool_t); + + char* dir; +}; + +dbh_t* dbtext_db_open( cpchar dbhost, cpchar dbname, cpchar dbuser, cpchar dbpass ); +bool_t dbtext_db_close( dbhtext_t* pthis ); +dbt_t* dbtext_db_opentable( dbhtext_t* pthis, cpchar table, bool_t rdonly ); + +bool_t dbtext_table_close( dbttext_t* pthis ); +bool_t dbtext_table_mergeclose( dbttext_t* pthis, vec_t* pmsg ); +bool_t dbtext_table_unmergeclose( dbttext_t* pthis, vec_t* pmsg ); +bool_t dbtext_table_import( dbttext_t* pthis, cpchar filename ); +bool_t dbtext_table_export( dbttext_t* pthis, cpchar filename ); +uint dbtext_table_getmsgcount( dbttext_t* pthis ); +uint dbtext_table_getcount( dbttext_t* pthis, str_t* pword ); + +#endif /* ndef _DBTEXT_H */ diff --git a/filt.c b/filt.c @@ -0,0 +1,175 @@ +/* $Id: filt.c,v 1.1 2002/10/20 18:19:17 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * filt.c: The Bayes filter implementation. + * See http://www.paulgraham.com/spam.html for discussion. + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" +#include "dbh.h" +#include "filt.h" + +#define DEVIATION(n) fabs((n)-0.5f) + +/* Dump the contents of a statistics structure */ +void statdump( stats_t* pstat, int fd ) +{ + char iobuf[IOBUFSIZE]; + char* p; + discrim_t* pp; + + p = iobuf; + p += sprintf( iobuf, "# Spamicity: %f\n", pstat->spamicity ); + + for (pp = pstat->extrema; pp < pstat->extrema + pstat->keepers; pp++) + { + if (pp->key.len) + { + strcpy( p, "# '" ); p += 3; + strncpylwr( p, pp->key.p, pp->key.len ); p += pp->key.len; + p += snprintf( p, 28, "' -> %f\n", pp->prob ); + if( p+MAXWORDLEN+32 > (iobuf+1) ) + { + write( fd, iobuf, p-iobuf ); + p = iobuf; + } + } + } + if( p != iobuf ) + { + write( fd, iobuf, p-iobuf ); + } +} + +void bayesfilt( dbt_t* pglist, dbt_t* pblist, vec_t* pmlist, stats_t* pstats ) +{ + veciter_t iter; + str_t* pword; + + double prob, product, invproduct, dev; + double slotdev, hitdev; + +#ifdef NON_EQUIPROBABLE + /* There is an argument that we should (go?) by number of *words* here. */ + double msg_prob = ((double)pblist->nitems / (double)pglist->nitems); +#endif + + discrim_t* pp; + discrim_t* hit; + + for (pp = pstats->extrema; pp < pstats->extrema+pstats->keepers; pp++) + { + pp->key.p = NULL; + pp->key.len = 0; + pp->prob = 0.5f; + } + + vec_first( pmlist, &iter ); + while( (pword = veciter_get( &iter )) != NULL ) + { + double goodness = pglist->getcount( pglist, pword ); + double spamness = pblist->getcount( pblist, pword ); + uint goodtotal = pglist->getmsgcount( pglist ); + uint spamtotal = pblist->getmsgcount( pblist ); + + if( goodness + spamness < MINIMUM_FREQ ) + { +#ifdef NON_EQUIPROBABLE + /* + * In the absence of evidence, the probability that a new word will + * be spam is the historical ratio of spam words to nonspam words. + */ + prob = msg_prob; +#else + prob = UNKNOWN_WORD; +#endif + } + else + { + double goodprob = goodtotal ? min( 1.0, (goodness / goodtotal) ) : 0.0; + double spamprob = spamtotal ? min( 1.0, (spamness / spamtotal) ) : 0.0; + assert( goodtotal > 0 || spamtotal > 0 ); + +#ifdef NON_EQUIPROBABLE + prob = (spamprob * msg_prob) / ((goodprob * (1 - msg_prob)) + (spamprob * msg_prob)); +#else + prob = spamprob / (goodprob + spamprob); +#endif + + prob = minmax( prob, 0.01, 0.99 ); + } + + /* update the list of tokens with maximum deviation */ + dev = DEVIATION(prob); + hit = NULL; + hitdev = 0; + for (pp = pstats->extrema; pp < pstats->extrema+pstats->keepers; pp++) + { + /* don't allow duplicate tokens in the stats.extrema */ + if( pp->key.len > 0 && str_casecmp( pword, &pp->key ) == 0 ) + { + hit = NULL; + break; + } + + slotdev = DEVIATION(pp->prob); + if (dev>slotdev && dev>hitdev) + { + hit = pp; + hitdev = slotdev; + } + } + if (hit) + { + hit->prob = prob; + hit->key = *pword; + } + + veciter_next( &iter ); + } + veciter_destroy( &iter ); + + /* + * Bayes' theorem. + * For discussion, see <http://www.mathpages.com/home/kmath267.htm>. + */ + product = invproduct = 1.0f; + for (pp = pstats->extrema; pp < pstats->extrema+pstats->keepers; pp++) + { + if( pp->prob == 0 ) + { + break; + } + else + { + product *= pp->prob; + invproduct *= (1 - pp->prob); + } + } + pstats->spamicity = product / (product + invproduct); +} + +bool_t bvec_loadmsg( vec_t* pthis, lex_t* plex, tok_t* ptok ) +{ + str_t w; + + lex_nexttoken( plex, ptok ); + while( ptok->tt != eof && ptok->tt != from ) + { + w.p = ptok->p; + w.len = ptok->len; + vec_addtail( pthis, &w ); + lex_nexttoken( plex, ptok ); + } + + return true; +} diff --git a/filt.h b/filt.h @@ -0,0 +1,31 @@ +/* $Id: filt.h,v 1.1 2002/10/20 18:19:17 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _FILT_H +#define _FILT_H + +typedef struct +{ + str_t key; + double prob; +} discrim_t; + +typedef struct +{ + double spamicity; + uint keepers; + discrim_t* extrema; +} stats_t; + +void statdump( stats_t* pstat, int fd ); +void bayesfilt( dbt_t* pglist, dbt_t* pblist, vec_t* pmlist, stats_t* pstats ); + +bool_t bvec_loadmsg( vec_t* pthis, lex_t* plex, tok_t* ptok ); + +#endif /* ndef _FILT_H */ diff --git a/lex.c b/lex.c @@ -0,0 +1,787 @@ +/* $Id: lex.c,v 1.18 2002/10/20 20:29:15 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * lex.c: generate token stream for bmf. + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" + +static cpchar g_htmltags[] = +{ + "abbr", + "above", + "accesskey", + "acronym", + "align", + "alink", + "all", + "alt", + "applet", + "archive", + "axis", + "basefont", + "baseline", + "below", + "bgcolor", + "big", + "body", + "border", + "bottom", + "box", + "button", + "cellpadding", + "cellspacing", + "center", + "char", + "charoff", + "charset", + "circle", + "cite", + "class", + "classid", + "clear", + "codebase", + "codetype", + "color", + "cols", + "colspan", + "compact", + "content", + "coords", + "data", + "datetime", + "declare", + "default", + "defer", + "dfn", + "dir", + "disabled", + "face", + "font", + "frameborder", + "groups", + "head", + "headers", + "height", + "href", + "hreflang", + "hsides", + "hspace", + "http-equiv", + "iframe", + "img", + "input", + "ismap", + "justify", + "kbd", + "label", + "lang", + "language", + "left", + "lhs", + "link", + "longdesc", + "map", + "marginheight", + "marginwidth", + "media", + "meta", + "middle", + "multiple", + "name", + "nohref", + "none", + "noresize", + "noshade", + "nowrap", + "object", + "onblur", + "onchange", + "onclick", + "ondblclick", + "onfocus", + "onkeydown", + "onkeypress", + "onkeyup", + "onload", + "onmousedown", + "onmousemove", + "onmouseout", + "onmouseover", + "onmouseup", + "onselect", + "onunload", + "param", + "poly", + "profile", + "prompt", + "readonly", + "rect", + "rel", + "rev", + "rhs", + "right", + "rows", + "rowspan", + "rules", + "samp", + "scheme", + "scope", + "script", + "scrolling", + "select", + "selected", + "shape", + "size", + "small", + "span", + "src", + "standby", + "strike", + "strong", + "style", + "sub", + "summary", + "sup", + "tabindex", + "table", + "target", + "textarea", + "title", + "top", + "type", + "usemap", + "valign", + "value", + "valuetype", + "var", + "vlink", + "void", + "vsides", + "vspace", + "width" +}; +static const uint g_nhtmltags = sizeof(g_htmltags)/sizeof(cpchar); + +static cpchar g_ignoredheaders[] = +{ + "Date:", + "Delivery-date:", + "Message-ID:", + "X-Sorted:", + "X-Spam-" +}; +static const uint g_nignoredheaders = sizeof(g_ignoredheaders)/sizeof(cpchar); + +static inline bool_t is_whitespace( int c ) +{ + return ( c == ' ' || c == '\t' || c == '\r' ); +} + +static inline bool_t is_base64char(c) +{ + return ( isalnum(c) || (c == '/' || c == '+') ); +} + +static inline bool_t is_wordmidchar(c) +{ + return ( isalnum(c) || c == '$' || c == '\'' || c == '.' || c == '-' ); +} + +static inline bool_t is_wordendchar(c) +{ + return ( isalnum(c) || c == '$' ); +} + +static inline bool_t is_htmltag( cpchar p, uint len, uint* ptoklen ) +{ + int lo, hi, mid, minlen, cmp; + + *ptoklen = 0; + + hi = g_nhtmltags-1; + lo = -1; + while( hi-lo > 1 ) + { + mid = (hi+lo)/2; + minlen = min( strlen(g_htmltags[mid]), len ); + cmp = strncmp( g_htmltags[mid], p, minlen ); + if( cmp > 0 || (cmp == 0 && minlen < len && !islower(p[minlen])) ) + hi = mid; + else + lo = mid; + } + minlen = min( strlen(g_htmltags[hi]), len ); + if( len == minlen || strncmp(g_htmltags[hi], p, minlen) != 0 ) + { + return false; + } + + /* check if is_word() will have a longer match */ + if( is_wordendchar(p[minlen]) ) + { + return false; + } + if( is_wordmidchar(p[minlen]) && is_wordendchar(p[minlen+1]) ) + { + return false; + } + + *ptoklen = strlen(g_htmltags[hi]); + + return true; +} + +static inline bool_t is_htmlcomment( cpchar p, uint len, uint* ptoklen ) +{ + *ptoklen = 0; + + if( len >=4 && memcmp( p, "<!--", 4 ) == 0 ) + { + *ptoklen = 4; + return true; + } + if( len >= 3 && memcmp( p, "-->", 3 ) == 0 ) + { + *ptoklen = 3; + return true; + } + + return false; +} + +static inline bool_t is_base64( cpchar p, uint len, uint* ptoklen ) +{ + *ptoklen = 0; + while( len > 0 ) + { + if( *p != '\n' && *p != '\r' && !is_base64char(*p) ) + { + return false; + } + p++; + len--; + (*ptoklen)++; + } + return true; +} + +static inline bool_t is_mimeboundary( cpchar p, uint len, uint* ptoklen ) +{ + *ptoklen = 0; + + if( len < 3 || p[0] != '-' || p[1] != '-' ) + { + return false; + } + p += 2; + len -= 2; + *ptoklen += 2; + while( len > 0 ) + { + if( is_whitespace(*p) ) + { + return false; + } + if( *p == '\n' || *p == '\r' ) + { + break; + } + p++; + len--; + (*ptoklen)++; + } + return true; +} + +static inline bool_t is_ipaddr( cpchar p, uint len, uint* ptoklen ) +{ + uint noctets, ndigits; + + *ptoklen = 0; + + noctets = 0; + while( len > 0 && noctets < 4 ) + { + ndigits = 0; + while( len > 0 && isdigit(*p) ) + { + ndigits++; + p++; + len--; + (*ptoklen)++; + } + if( ndigits == 0 || ndigits > 3 ) + { + return false; + } + noctets++; + if( noctets < 4 ) + { + if( *p != '.' ) + { + return false; + } + p++; + len--; + (*ptoklen)++; + } + } + if( noctets < 4 ) + { + return false; + } + return true; +} + +static inline bool_t is_word( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 3 ) + { + return false; + } + if( !(isalpha(*p) || *p == '$') ) + { + return false; + } + *ptoklen = 1; + p++; + len--; + while( len > 0 ) + { + if( !is_wordmidchar(*p) ) + { + break; + } + (*ptoklen)++; + p++; + len--; + } + while( *ptoklen >= 3 && !is_wordendchar(*(p-1)) ) + { + (*ptoklen)--; + p--; + len++; + } + if( *ptoklen < 3 ) + { + return false; + } + + return true; +} + +static inline bool_t is_ignoredheader( cpchar p, uint len, uint* ptoklen ) +{ + int lo, hi, mid, minlen, cmp; + + hi = g_nignoredheaders-1; + lo = -1; + while( hi-lo > 1 ) + { + mid = (hi+lo)/2; + minlen = min( strlen(g_ignoredheaders[mid]), len ); + cmp = strncasecmp( g_ignoredheaders[mid], p, minlen ); + if( cmp >= 0 ) + hi = mid; + else + lo = mid; + } + minlen = min( strlen(g_ignoredheaders[hi]), len ); + if( len == minlen || strncasecmp(g_ignoredheaders[hi], p, minlen) != 0 ) + { + return false; + } + *ptoklen = len; + return true; +} + +static inline bool_t is_mailerid( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 4 || strncmp( p, "\tid ", 4 ) != 0 ) + { + return false; + } + *ptoklen = len; + return true; +} + +static inline bool_t is_spamtext( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 5 || strncmp( p, "SPAM:", 5 ) != 0 ) + { + return false; + } + *ptoklen = len; + return true; +} + +static inline bool_t is_smtpid( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 8 || strncmp( p, "SMTP id ", 8 ) != 0 ) + { + return false; + } + *ptoklen = len; + return true; +} + +static inline bool_t is_boundaryequal( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 9 || strncmp( p, "boundary=", 9 ) != 0 ) + { + return false; + } + *ptoklen = len; + return true; +} + +static inline bool_t is_nameequal( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 6 || strncmp( p, "name=\"", 6 ) != 0 ) + { + return false; + } + *ptoklen = 6; + return true; +} + +static inline bool_t is_filenameequal( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 10 || strncmp( p, "filename=\"", 10 ) != 0 ) + { + return false; + } + *ptoklen = 10; + return true; +} + +static inline bool_t is_from( cpchar p, uint len, uint* ptoklen ) +{ + if( len < 5 || strncmp( p, "From ", 5 ) != 0 ) + { + return false; + } + *ptoklen = 5; + return true; +} + +/*****************************************************************************/ + +void lex_create( lex_t* pthis, mbox_t mboxtype ) +{ + pthis->mboxtype = mboxtype; + pthis->section = envelope; + pthis->pos = 0; + pthis->bom = 0; + pthis->eom = 0; + pthis->lineend = 0; + pthis->buflen = 0; + pthis->pbuf = NULL; +} + +void lex_destroy( lex_t* pthis ) +{ + free( pthis->pbuf ); +} + +bool_t lex_load( lex_t* pthis, int fd ) +{ + uint nalloc; + ssize_t nread; + + nalloc = IOBUFSIZE; + pthis->pbuf = (char*)malloc( IOBUFSIZE ); + if( pthis->pbuf == NULL ) + { + return false; + } + + while( (nread = read( fd, pthis->pbuf + pthis->buflen, nalloc - pthis->buflen )) > 0 ) + { + pthis->buflen += nread; + if( pthis->buflen == nalloc ) + { + char* pnewbuf; + nalloc += IOBUFSIZE; + pnewbuf = (char*)realloc( pthis->pbuf, nalloc ); + if( pnewbuf == NULL ) + { + free( pthis->pbuf ); + pthis->pbuf = NULL; + return false; + } + pthis->pbuf = pnewbuf; + } + } + if( nread < 0 ) + { + free( pthis->pbuf ); + pthis->pbuf = NULL; + return false; + } + if( pthis->mboxtype == detect ) + { + if( pthis->buflen > 5 && memcmp( pthis->pbuf, "From ", 5 ) == 0 ) + { + verbose( 1, "Input looks like an mbox\n" ); + pthis->mboxtype = mbox; + } + else + { + verbose( 1, "Input looks like a maildir\n" ); + pthis->mboxtype = maildir; + } + } + + return true; +} + +static bool_t lex_nextline( lex_t* pthis ) +{ + cpchar pbuf; + uint len; + uint toklen; + +again: + /* XXX: use and update pthis->section */ + pthis->pos = pthis->lineend; + if( pthis->lineend == pthis->buflen ) + { + return false; + } + + pbuf = pthis->pbuf + pthis->pos; + len = 0; + while( pthis->pos + len < pthis->buflen && pbuf[len] != '\n' ) + { + len++; + } + if( pthis->pos + len < pthis->buflen ) + { + len++; /* bump past the LF */ + } + + pthis->lineend = pthis->pos + len; + + /* check beginning-of-line patterns */ + if( is_base64( pbuf, len, &toklen ) || + is_ignoredheader( pbuf, len, &toklen ) || + is_mailerid( pbuf, len, &toklen ) || + is_mimeboundary( pbuf, len, &toklen ) || + is_spamtext( pbuf, len, &toklen ) ) + { + /* ignore line */ + pthis->pos += toklen; + goto again; + } + + return true; +} + +void lex_nexttoken( lex_t* pthis, tok_t* ptok ) +{ + cpchar pbuf; + uint len; + uint toklen; + + assert( pthis->pbuf != NULL ); + + if( pthis->pos == pthis->eom ) + { + pthis->bom = pthis->pos; + } + +again: + /* skip whitespace between tokens */ + while( pthis->pos != pthis->lineend && is_whitespace(pthis->pbuf[pthis->pos]) ) + { + pthis->pos++; + } + + pbuf = pthis->pbuf + pthis->pos; + len = pthis->lineend - pthis->pos; + + /* possibilities: end-of-line, html-comment, ipaddr, word, junk */ + + if( pthis->pos == pthis->lineend ) + { + if( !lex_nextline( pthis ) ) + { + pthis->eom = pthis->pos; + ptok->tt = eof; + return; + } + + pbuf = pthis->pbuf + pthis->pos; + len = pthis->lineend - pthis->pos; + + if( pthis->mboxtype == mbox ) + { + if( is_from( pbuf, len, &toklen ) ) + { + pthis->eom = pthis->pos; + ptok->tt = from; + ptok->p = pthis->pbuf + pthis->pos; + ptok->len = toklen; + pthis->pos += toklen; + return; + } + } + + goto again; /* skip lws */ + } + + if( is_htmltag( pbuf, len, &toklen ) || + is_htmlcomment( pbuf, len, &toklen ) || + is_smtpid( pbuf, len, &toklen ) || + is_boundaryequal( pbuf, len, &toklen ) || + is_nameequal( pbuf, len, &toklen ) || + is_filenameequal( pbuf, len, &toklen ) ) + { + /* ignore it */ + pthis->pos += toklen; + goto again; + } + + if( is_ipaddr( pbuf, len, &toklen ) ) + { + ptok->tt = word; + ptok->p = pthis->pbuf + pthis->pos; + ptok->len = toklen; + pthis->pos += toklen; + return; + } + if( is_word( pbuf, len, &toklen ) ) + { + ptok->tt = word; + ptok->p = pthis->pbuf + pthis->pos; + ptok->len = toklen; + pthis->pos += toklen; + if( toklen > MAXWORDLEN ) + { + goto again; + } + return; + } + + /* junk */ + pthis->pos++; + goto again; +} + +/* SpamAssassin style passthru */ +void lex_passthru( lex_t* pthis, bool_t is_spam, double hits ) +{ + char szbuf[256]; + bool_t in_headers = true; + + assert( pthis->bom < pthis->buflen && pthis->eom <= pthis->buflen ); + assert( pthis->bom <= pthis->eom ); + + pthis->pos = pthis->bom; + if( is_spam ) + { + sprintf( szbuf, "X-Spam-Status: Yes, hits=%f required=%f, tests=bmf\n" + "X-Spam-Flag: YES\n", + hits, SPAM_CUTOFF ); + } + else + { + sprintf( szbuf, "X-Spam-Status: No, hits=%f required=%f\n", + hits, SPAM_CUTOFF ); + } + + /* existing headers */ + while( in_headers && pthis->pos < pthis->eom ) + { + cpchar pbuf = pthis->pbuf + pthis->pos; + uint len = 0; + while( pthis->pos + len < pthis->buflen && pbuf[len] != '\n' ) + { + len++; + } + if( pthis->pos + len < pthis->buflen ) + { + len++; /* bump past the LF */ + } + + /* check for end of headers */ + if( pbuf[0] == '\n' || (pbuf[0] == '\r' && pbuf[1] == '\n') ) + { + /* end of headers */ + break; + } + + /* write header, ignoring existing spam headers */ + if( strncasecmp( pbuf, "X-Spam-", 7 ) != 0 ) + { + write( STDOUT_FILENO, pbuf, len ); + } + + pthis->pos += len; + } + + /* new headers */ + write( STDOUT_FILENO, szbuf, strlen(szbuf) ); + + /* remainder */ + if( pthis->pos < pthis->eom ) + { + write( STDOUT_FILENO, pthis->pbuf+pthis->pos, pthis->eom-pthis->pos ); + } + pthis->bom = pthis->eom; +} + +#ifdef UNIT_TEST + +int main( int argc, char** argv ) +{ + int fd; + lex_t lex; + tok_t tok; + + fd = STDIN_FILENO; + if( argc == 2 ) + { + fd = open( argv[1], O_RDONLY ); + } + + lex_create( &lex ); + if( ! lex_load( &lex, fd ) ) + { + fprintf( stderr, "cannot load file\n" ); + exit( 1 ); + } + + lex_nexttoken( &lex, &tok ); + while( tok.tt != eof ) + { + char sztok[64]; + if( tok.len > MAXWORDLEN ) + { + printf( "*** token too long! ***\n" ); + exit( 1 ); + } + + memcpy( sztok, tok.p, tok.len ); + strlwr( sztok ); + sztok[tok.len] = '\0'; + printf( "get_token: %d '%s'\n", tok.tt, sztok ); + + lex_nexttoken( &lex, &tok ); + } + + lex_destroy( &lex ); + return 0; +} + +#endif /* def UNIT_TEST */ diff --git a/lex.h b/lex.h @@ -0,0 +1,44 @@ +/* $Id: lex.h,v 1.4 2002/10/12 17:36:41 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _TOK_H +#define _TOK_H + +typedef enum { from, eof, word } toktype_t; + +typedef struct _tok +{ + toktype_t tt; /* token type */ + char* p; + uint len; +} tok_t; + +typedef enum { envelope, hdrs, body } msgsec_t; + +typedef struct _lex +{ + mbox_t mboxtype; + msgsec_t section; /* current section (envelope, headers, body) */ + uint pos; /* current position */ + uint bom; /* beginning of message */ + uint eom; /* end of current message (start of next) */ + uint lineend; /* line end (actually, start of next line) */ + uint buflen; /* length of buffer */ + char* pbuf; +} lex_t; + +void lex_create ( lex_t* plex, mbox_t mboxtype ); +void lex_destroy ( lex_t* plex ); + +bool_t lex_load ( lex_t* plex, int fd ); +void lex_nexttoken( lex_t* plex, tok_t* ptok ); + +void lex_passthru ( lex_t* plex, bool_t is_spam, double hits ); + +#endif /* ndef TOK_H */ diff --git a/str.c b/str.c @@ -0,0 +1,78 @@ +/* $Id: str.c,v 1.2 2002/10/14 07:09:51 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" + +void strlwr( char* s ) +{ + while( *s != '\0' ) + { + *s = tolower(*s); + s++; + } +} + +void strcpylwr( char* d, const char* s ) +{ + while( *s != '\0' ) + { + *d++ = tolower(*s++); + } +} + +void strncpylwr( char* d, const char* s, int n ) +{ + while( n-- ) + { + *d++ = tolower(*s++); + } +} + +void str_create( str_t* pstr ) +{ + pstr->p = NULL; + pstr->len = 0; +} + +void str_destroy( str_t* pstr ) +{ + /* empty */ +} + +int str_cmp( const str_t* pthis, const str_t* pother ) +{ + uint minlen = min( pthis->len, pother->len ); + int cmp; + assert( pthis->p != NULL && pother->p != NULL && minlen != 0 ); + + cmp = strncmp( pthis->p, pother->p, minlen ); + + if( cmp == 0 && pthis->len != pother->len ) + { + cmp = (pthis->len < pother->len) ? -1 : 1; + } + return cmp; +} + +int str_casecmp( const str_t* pthis, const str_t* pother ) +{ + uint minlen = min( pthis->len, pother->len ); + int cmp; + assert( pthis->p != NULL && pother->p != NULL && minlen != 0 ); + + cmp = strncasecmp( pthis->p, pother->p, minlen ); + + if( cmp == 0 && pthis->len != pother->len ) + { + cmp = (pthis->len < pother->len) ? -1 : 1; + } + return cmp; +} diff --git a/str.h b/str.h @@ -0,0 +1,30 @@ +/* $Id: str.h,v 1.1.1.1 2002/09/30 21:08:29 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _STR_H +#define _STR_H + +/* a couple of generic string functions... */ +void strlwr( char* s ); +void strcpylwr( char* d, const char* s ); +void strncpylwr( char* d, const char* s, int n ); + +typedef struct _str +{ + char* p; + uint len; +} str_t; + +void str_create ( str_t* pthis ); +void str_destroy( str_t* pthis ); + +int str_cmp ( const str_t* pthis, const str_t* pother ); +int str_casecmp( const str_t* pthis, const str_t* pother ); + +#endif /* ndef _STR_H */ diff --git a/vec.c b/vec.c @@ -0,0 +1,345 @@ +/* $Id: vec.c,v 1.4 2002/10/20 18:19:17 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + * + * vec.c: vector functions for bmf. + * Vectors are used to hold token lists for input data and flatfile database + * entries in standalone mode. They dramatically reduce the number of small + * mallocs and, if used properly, have no performance penalty over fancier + * data structures. + */ + +#include "config.h" +#include "dbg.h" +#include "str.h" +#include "lex.h" +#include "vec.h" + +/***************************************************************************** + * vector + */ + +void vec_create( vec_t* pthis ) +{ + pthis->nalloc = VEC_INITIAL_SIZE; + pthis->nitems = 0; + pthis->pitems = (str_t*)malloc( VEC_INITIAL_SIZE*sizeof(str_t) ); +} + +void vec_destroy( vec_t* pthis ) +{ + free( pthis->pitems ); +} + +static void vec_setsize( vec_t* pthis, uint nsize ) +{ + if( nsize > pthis->nalloc ) + { + uint nnewalloc; + str_t* pnewitems; + uint n; + + nnewalloc = pthis->nalloc * 2; + if( nnewalloc < nsize ) nnewalloc = nsize; + pnewitems = (str_t*)realloc( pthis->pitems, nnewalloc*sizeof(str_t) ); + if( pnewitems == NULL ) + { + exit( 2 ); + } + for( n = pthis->nitems; n < nsize; n++ ) + { + str_create( &pnewitems[n] ); + } + pthis->pitems = pnewitems; + pthis->nalloc = nnewalloc; + } +} + +void vec_addhead( vec_t* pthis, str_t* pstr ) +{ + assert( pstr->p != NULL && pstr->len > 0 ); + + vec_setsize( pthis, pthis->nitems+1 ); + memmove( &pthis->pitems[1], &pthis->pitems[0], pthis->nitems*sizeof(str_t) ); + pthis->pitems[0] = *pstr; + pthis->nitems++; +} + +void vec_addtail( vec_t* pthis, str_t* pstr ) +{ + assert( pstr->p != NULL && pstr->len > 0 ); + + vec_setsize( pthis, pthis->nitems+1 ); + pthis->pitems[pthis->nitems] = *pstr; + pthis->nitems++; +} + +void vec_delhead( vec_t* pthis ) +{ + assert( pthis->nitems > 0 ); + pthis->nitems--; + memmove( &pthis->pitems[0], &pthis->pitems[1], pthis->nitems*sizeof(str_t) ); +} + +void vec_deltail( vec_t* pthis ) +{ + assert( pthis->nitems > 0 ); + pthis->nitems--; +} + +void vec_first( vec_t* pthis, veciter_t* piter ) +{ + piter->plist = pthis; + piter->index = 0; +} + +void vec_last( vec_t* pthis, veciter_t* piter ) +{ + piter->plist = pthis; + piter->index = pthis->nitems; +} + +/***************************************************************************** + * sorted vector + */ + +static int svec_compare( const void* p1, const void* p2 ) +{ + return str_casecmp( (const str_t*)p1, (const str_t*)p2 ); +} + +void svec_add( vec_t* pthis, str_t* pstr ) +{ + int lo, hi, mid; + veciter_t iter; + + if( pthis->nitems == 0 ) + { + vec_addtail( pthis, pstr ); + return; + } + + if( str_casecmp( pstr, &pthis->pitems[0] ) < 0 ) + { + vec_addhead( pthis, pstr ); + return; + } + + hi = pthis->nitems - 1; + lo = -1; + while( hi-lo > 1 ) + { + mid = (hi+lo)/2; + if( str_casecmp( pstr, &pthis->pitems[mid] ) <= 0 ) + hi = mid; + else + lo = mid; + } + assert( hi < pthis->nitems ); + + iter.plist = pthis; + iter.index = hi; + + if( str_casecmp( pstr, &pthis->pitems[hi] ) < 0 ) + { + veciter_addbefore( &iter, pstr ); + } + else + { + veciter_addafter( &iter, pstr ); + } +} + +str_t* svec_find( vec_t* pthis, str_t* pstr ) +{ + int lo, hi, mid; + + if( pthis->nitems == 0 ) + { + return NULL; + } + + hi = pthis->nitems - 1; + lo = -1; + while( hi-lo > 1 ) + { + mid = (hi+lo)/2; + if( str_casecmp( pstr, &pthis->pitems[mid] ) <= 0 ) + hi = mid; + else + lo = mid; + } + assert( hi >= 0 && hi < pthis->nitems ); + + if( str_casecmp( pstr, &pthis->pitems[hi] ) != 0 ) + { + return NULL; + } + + return &pthis->pitems[hi]; +} + +void svec_sort( vec_t* pthis ) +{ + if( pthis->nitems > 1 ) + { + qsort( pthis->pitems, pthis->nitems, sizeof(str_t), svec_compare ); + } +} + +/***************************************************************************** + * vector iterator + */ + +void veciter_destroy( veciter_t* pthis ) +{ + /* empty */ +} + +str_t* veciter_get( veciter_t* pthis ) +{ + if( pthis->plist == NULL || pthis->index >= pthis->plist->nitems ) + { + return NULL; + } + + return &pthis->plist->pitems[pthis->index]; +} + +bool_t veciter_equal( veciter_t* pthis, veciter_t* pthat ) +{ + if( pthis->plist != pthat->plist || + pthis->index != pthat->index ) + { + return false; + } + + return true; +} + +bool_t veciter_hasitem( veciter_t* pthis ) +{ + if( pthis->plist == NULL || pthis->index >= pthis->plist->nitems ) + { + return false; + } + return true; +} + +bool_t veciter_prev( veciter_t* pthis ) +{ + if( pthis->index == 0 ) + { + return false; + } + pthis->index--; + return true; +} + +bool_t veciter_next( veciter_t* pthis ) +{ + pthis->index++; + if( pthis->index == pthis->plist->nitems ) + { + return false; + } + return true; +} + +void veciter_addafter( veciter_t* pthis, str_t* pstr ) +{ + str_t* pitems; + + vec_setsize( pthis->plist, pthis->plist->nitems+1 ); + assert( pthis->index < pthis->plist->nitems ); + pitems = pthis->plist->pitems; + + if( pthis->index != pthis->plist->nitems-1 ) + { + memmove( &pitems[pthis->index+2], &pitems[pthis->index+1], + (pthis->plist->nitems-pthis->index-1) * sizeof(str_t) ); + } + + pitems[pthis->index+1] = *pstr; + pthis->plist->nitems++; +} + +void veciter_addbefore( veciter_t* pthis, str_t* pstr ) +{ + str_t* pitems; + + vec_setsize( pthis->plist, pthis->plist->nitems+1 ); + assert( pthis->index < pthis->plist->nitems ); + pitems = pthis->plist->pitems; + + memmove( &pitems[pthis->index+1], &pitems[pthis->index], + (pthis->plist->nitems-pthis->index) * sizeof(str_t) ); + + pitems[pthis->index] = *pstr; + pthis->plist->nitems++; +} + +void veciter_del( veciter_t* pthis ) +{ + str_t* pitems; + + assert( pthis->plist->nitems > 0 ); + pthis->plist->nitems--; + if( pthis->index < pthis->plist->nitems ) + { + pitems = pthis->plist->pitems; + memmove( &pitems[pthis->index], &pitems[pthis->index+1], + (pthis->plist->nitems-pthis->index) * sizeof(str_t) ); + } +} + +#ifdef UNIT_TEST +int main( int argc, char** argv ) +{ + vec_t vl; + veciter_t iter; + str_t* pstr; + uint n; + + if( argc != 2 ) + { + fprintf( stderr, "usage: %s <file>\n", argv[0] ); + return 1; + } + + for( n = 0; n < 100; n++ ) + { + vec_create( &vl ); + vec_load( &vl, argv[1] ); + + vec_first( &vl, &iter ); + while( (pstr = veciter_get( &iter )) != NULL ) + { + char buf[256]; + char* p; + if( pstr->len > 200 ) + { + fprintf( stderr, "str too long: %u chars\n", pstr->len ); + break; + } + p = buf; + strcpy( buf, "str: " ); + p += 6; + memcpy( p, pstr->p, pstr->len ); + p += pstr->len; + sprintf( p, " %u", pstr->count ); + puts( buf ); + + veciter_next( &iter ); + } + + vec_destroy( &vl ); + } + + return 0; +} +#endif /* def UNIT_TEST */ diff --git a/vec.h b/vec.h @@ -0,0 +1,58 @@ +/* $Id: vec.h,v 1.3 2002/10/20 18:19:17 tommy Exp $ */ + +/* + * Copyright (c) 2002 Tom Marshall <tommy@tig-grr.com> + * + * This program is free software. It may be distributed under the terms + * in the file LICENSE, found in the top level of the distribution. + */ + +#ifndef _VEC_H +#define _VEC_H + +/* item count for initial alloc */ +#define VEC_INITIAL_SIZE 256 + +typedef struct _vec +{ + uint nalloc; /* items alloced in pitems */ + uint nitems; /* items available */ + str_t* pitems; /* growing vector of items */ +} vec_t; + +typedef struct _veciter +{ + struct _vec* plist; + uint index; +} veciter_t; + +/* class vector */ +void vec_create ( vec_t* pthis ); +void vec_destroy ( vec_t* pthis ); + +void vec_addhead ( vec_t* pthis, str_t* pstr ); +void vec_addtail ( vec_t* pthis, str_t* pstr ); +void vec_delhead ( vec_t* pthis ); +void vec_deltail ( vec_t* pthis ); + +void vec_first ( vec_t* pthis, veciter_t* piter ); +void vec_last ( vec_t* pthis, veciter_t* piter ); + +/* class sorted_vector */ +void svec_add ( vec_t* pthis, str_t* pstr ); +str_t* svec_find ( vec_t* pthis, str_t* pstr ); +void svec_sort ( vec_t* ptthis ); + +/* veciter_create not needed */ +void veciter_destroy ( veciter_t* pthis ); + +str_t* veciter_get ( veciter_t* pthis ); +bool_t veciter_equal ( veciter_t* pthis, veciter_t* pthat ); +bool_t veciter_hasitem ( veciter_t* pthis ); +bool_t veciter_prev ( veciter_t* pthis ); +bool_t veciter_next ( veciter_t* pthis ); +void veciter_addafter ( veciter_t* pthis, str_t* pstr ); +void veciter_addbefore( veciter_t* pthis, str_t* pstr ); +void veciter_del ( veciter_t* pthis ); + +#endif /* ndef _VEC_H */