From ba33de40d6674e5d94bfae3d1a20648cd5ba4fba Mon Sep 17 00:00:00 2001 From: Daniel Baumann Date: Sat, 4 May 2024 16:25:58 +0200 Subject: Adding upstream version 1.13. Signed-off-by: Daniel Baumann --- AUTHORS | 7 + COPYING | 17 + COPYING.GPL | 338 +++++++++++ ChangeLog | 255 +++++++++ INSTALL | 80 +++ Makefile.in | 202 +++++++ NEWS | 15 + README | 103 ++++ bbexample.c | 367 ++++++++++++ carg_parser.c | 319 +++++++++++ carg_parser.h | 97 ++++ cbuffer.c | 143 +++++ configure | 239 ++++++++ decoder.c | 145 +++++ decoder.h | 463 +++++++++++++++ doc/lzlib.info | 1323 +++++++++++++++++++++++++++++++++++++++++++ doc/lzlib.texi | 1395 ++++++++++++++++++++++++++++++++++++++++++++++ doc/minilzip.1 | 134 +++++ encoder.c | 586 +++++++++++++++++++ encoder.h | 326 +++++++++++ encoder_base.c | 196 +++++++ encoder_base.h | 612 ++++++++++++++++++++ fast_encoder.c | 175 ++++++ fast_encoder.h | 70 +++ ffexample.c | 300 ++++++++++ lzcheck.c | 367 ++++++++++++ lzip.h | 298 ++++++++++ lzlib.c | 601 ++++++++++++++++++++ lzlib.h | 110 ++++ minilzip.c | 1290 ++++++++++++++++++++++++++++++++++++++++++ testsuite/check.sh | 444 +++++++++++++++ testsuite/fox.lz | Bin 0 -> 80 bytes testsuite/fox_bcrc.lz | Bin 0 -> 80 bytes testsuite/fox_crc0.lz | Bin 0 -> 80 bytes testsuite/fox_das46.lz | Bin 0 -> 80 bytes testsuite/fox_de20.lz | Bin 0 -> 80 bytes testsuite/fox_lf | 9 + testsuite/fox_mes81.lz | Bin 0 -> 80 bytes testsuite/fox_s11.lz | Bin 0 -> 80 bytes testsuite/fox_v2.lz | Bin 0 -> 80 bytes testsuite/test.txt | 676 ++++++++++++++++++++++ testsuite/test.txt.lz | Bin 0 -> 7376 bytes testsuite/test_em.txt.lz | Bin 0 -> 14024 bytes testsuite/test_sync.lz | Bin 0 -> 7568 bytes 44 files changed, 11702 insertions(+) create mode 100644 AUTHORS create mode 100644 COPYING create mode 100644 COPYING.GPL create mode 100644 ChangeLog create mode 100644 INSTALL create mode 100644 Makefile.in create mode 100644 NEWS create mode 100644 README create mode 100644 bbexample.c create mode 100644 carg_parser.c create mode 100644 carg_parser.h create mode 100644 cbuffer.c create mode 100755 configure create mode 100644 decoder.c create mode 100644 decoder.h create mode 100644 doc/lzlib.info create mode 100644 doc/lzlib.texi create mode 100644 doc/minilzip.1 create mode 100644 encoder.c create mode 100644 encoder.h create mode 100644 encoder_base.c create mode 100644 encoder_base.h create mode 100644 fast_encoder.c create mode 100644 fast_encoder.h create mode 100644 ffexample.c create mode 100644 lzcheck.c create mode 100644 lzip.h create mode 100644 lzlib.c create mode 100644 lzlib.h create mode 100644 minilzip.c create mode 100755 testsuite/check.sh create mode 100644 testsuite/fox.lz create mode 100644 testsuite/fox_bcrc.lz create mode 100644 testsuite/fox_crc0.lz create mode 100644 testsuite/fox_das46.lz create mode 100644 testsuite/fox_de20.lz create mode 100644 testsuite/fox_lf create mode 100644 testsuite/fox_mes81.lz create mode 100644 testsuite/fox_s11.lz create mode 100644 testsuite/fox_v2.lz create mode 100644 testsuite/test.txt create mode 100644 testsuite/test.txt.lz create mode 100644 testsuite/test_em.txt.lz create mode 100644 testsuite/test_sync.lz diff --git a/AUTHORS b/AUTHORS new file mode 100644 index 0000000..03a0e01 --- /dev/null +++ b/AUTHORS @@ -0,0 +1,7 @@ +Lzlib was written by Antonio Diaz Diaz. + +The ideas embodied in lzlib are due to (at least) the following people: +Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the +definition of Markov chains), G.N.N. Martin (for the definition of range +encoding), Igor Pavlov (for putting all the above together in LZMA), and +Julian Seward (for bzip2's CLI). diff --git a/COPYING b/COPYING new file mode 100644 index 0000000..a6511c8 --- /dev/null +++ b/COPYING @@ -0,0 +1,17 @@ + Lzlib - Compression library for the lzip format + Copyright (C) Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. diff --git a/COPYING.GPL b/COPYING.GPL new file mode 100644 index 0000000..4ad17ae --- /dev/null +++ b/COPYING.GPL @@ -0,0 +1,338 @@ + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 Free Software Foundation, Inc., + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 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 Lesser 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. + + + Copyright (C) + + 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, see . + +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) + 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. + + , 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 Lesser General +Public License instead of this License. diff --git a/ChangeLog b/ChangeLog new file mode 100644 index 0000000..8d7da96 --- /dev/null +++ b/ChangeLog @@ -0,0 +1,255 @@ +2022-01-23 Antonio Diaz Diaz + + * Version 1.13 released. + * Set variables AR and ARFLAGS from configure. + (Reported by Hoël Bézier). + * main.c: Rename to minilzip.c. + * minilzip.c (getnum): Show option name and valid range if error. + (check_lib): Check that LZ_API_VERSION and LZ_version_string match. + * Improve several descriptions in manual, '--help', and man page. + * lzlib.texi: Change GNU Texinfo category to 'Compression'. + (Reported by Alfred M. Szmidt). + +2021-01-02 Antonio Diaz Diaz + + * Version 1.12 released. + * lzlib.h: Define LZ_API_VERSION as 1000 * major + minor. 1.12 = 1012. + This change does not affect the soversion. + * lzlib.h, lzlib.c: New function LZ_api_version. + * LZd_try_verify_trailer: Return 2 if EOF at trailer or EOS marker. + * Decompression speed has been slightly increased. + * decoder.h: Increase 'rd_min_available_bytes' from 8 to 10. + * encoder_base.c (LZeb_try_sync_flush): + Compensate for the increase in 'rd_min_available_bytes'. + * main.c (do_decompress): Fix false report about library stall. + * main.c: New option '--check-lib'. + * main.c (main): Report an error if a file name is empty. + Make '-o' behave like '-c', but writing to file instead of stdout. + Make '-c' and '-o' check whether the output is a terminal only once. + Do not open output if input is a terminal. + Replace 'decompressed', 'compressed' with 'out', 'in' in output. + Set a valid invocation_name even if argc == 0. + * lzlib.texi: Document the new way of verifying the library version. + Document that 'LZ_(de)compress_close' and 'LZ_(de)compress_errno' + can be called with a null argument. + Document that sync flush marker is not allowed in lzip files. + Document the consequences of not calling 'LZ_decompress_finish'. + Document that 'LZ_decompress_read' returns at least once per member. + Document that 'LZ_(de)compress_read' can be called with a null + buffer pointer argument. + Real code examples for common uses have been added to the tutorial. + * bbexample.c: Don't use 'LZ_(de)compress_write_size'. + * lzcheck.c: New options '-s' (sync) and '-m' (member by member). + Test member by member without 'LZ_decompress_finish'. + * ffexample.c: New file containing example functions for file-to-file + compression/decompression. + * Document extraction from tar.lz in '--help' output and man page. + * Makefile.in: 'install-bin' no longer installs the man page. + New targets 'install-bin-compress' and 'install-bin-strip-compress'. + * testsuite: Add 9 new test files. + +2019-01-02 Antonio Diaz Diaz + + * Version 1.11 released. + * Rename File_* to Lzip_*. + * LZ_decompress_read: Don't return error until all data is read. + * decoder.c (LZd_decode_member): Decode truncated data until EOF. + * cbuffer.c (Cb_read_data): Allow a null buffer pointer. + * main.c: Don't allow mixing different operations (-d and -t). + * main.c: Check return value of close( infd ). + * main.c: Compile on DOS with DJGPP. + * lzlib.texi: Improve descriptions of '-0..-9', '-m', and '-s'. + Document that 'LZ_(de)compress_finish' can be called repeatedly. + * configure: Accept appending to CFLAGS; 'CFLAGS+=OPTIONS'. + * Makefile.in: Rename targets 'install-bin*' to 'install-lib*'. + * Makefile.in: Targets 'install-bin*' now install minilzip. + * INSTALL: Document use of CFLAGS+='-D __USE_MINGW_ANSI_STDIO'. + +2018-02-07 Antonio Diaz Diaz + + * Version 1.10 released. + * LZ_compress_finish now adjusts dictionary size for each member. + (Older versions can adjust dictionary size only once). + * lzlib.c (LZ_decompress_read): Detect corrupt header with HD=3. + * main.c: New option '--loose-trailing'. + * main.c (main): Option '-S, --volume-size' now keeps input files. + * main.c: Replace 'bits/byte' with inverse compression ratio. + * main.c: Show final diagnostic when testing multiple files. + * main.c: Do not add a second .lz extension to the arg of -o. + * main.c: Show dictionary size at verbosity level 4 (-vvvv). + * lzlib.texi: New chapter 'Invoking minilzip'. + +2017-04-11 Antonio Diaz Diaz + + * Version 1.9 released. + * Compression time of option '-0' has been reduced by 3%. + * Compression time of options -1 to -9 has been reduced by 1%. + * Decompression time has been reduced by 3%. + * main.c: Continue testing if any input file is a terminal. + * Change the license of the library to "2-clause BSD". + +2016-05-17 Antonio Diaz Diaz + + * Version 1.8 released. + * lzlib.h: Define LZ_API_VERSION to 1. + * lzlib.c (LZ_decompress_sync_to_member): Add skipped size to in_size. + * decoder.c (LZd_verify_trailer): Remove test of final code. + * main.c: New option '-a, --trailing-error'. + * main.c (main): Delete '--output' file if infd is a terminal. + * main.c (main): Don't use stdin more than once. + * configure: Avoid warning on some shells when testing for gcc. + * Makefile.in: Detect the existence of install-info. + * check.sh: A POSIX shell is required to run the tests. + * check.sh: Don't check error messages. + +2015-07-08 Antonio Diaz Diaz + + * Version 1.7 released. + * Port fast encoder and option '-0' from lzip. + * If open-->write-->finish, produce same dictionary size as lzip. + * Makefile.in: New targets 'install*-compress'. + +2014-08-27 Antonio Diaz Diaz + + * Version 1.6 released. + * Compression ratio of option '-9' has been slightly increased. + * configure: New options '--disable-static' and '--disable-ldconfig'. + * Makefile.in: Ignore errors from ldconfig. + * Makefile.in: Use 'CFLAGS' in every invocation of 'CC'. + * main.c (close_and_set_permissions): Behave like 'cp -p'. + * lzlib.texinfo: Rename to lzlib.texi. + * Change license to "GPL version 2 or later with link exception". + +2013-09-15 Antonio Diaz Diaz + + * Version 1.5 released. + * Remove decompression support for version 0 files. + * The LZ_compress_sync_flush mechanism has been fixed (again). + * Minor fixes. + +2013-05-28 Antonio Diaz Diaz + + * Version 1.4 released. + * Multi-step trials have been implemented. + * Compression ratio has been slightly increased. + * Compression time has been reduced by 8%. + * Decompression time has been reduced by 7%. + * lzlib.h: Change 'long long' values to 'unsigned long long'. + * encoder.c (Mf_init): Reduce minimum buffer size to 64KiB. + * lzlib.c (LZ_decompress_read): Tell LZ_header_error from + LZ_unexpected_eof the same way as lzip does. + * Makefile.in: New targets 'install-as-lzip' and 'install-bin'. + * main.c: Use 'setmode' instead of '_setmode' on Windows and OS/2. + * main.c: Define 'strtoull' to 'strtoul' on Windows. + +2012-02-29 Antonio Diaz Diaz + + * Version 1.3 released. + * Translated to C from the C++ source of lzlib 1.2. + * configure: Rename 'datadir' to 'datarootdir'. + +2011-10-25 Antonio Diaz Diaz + + * Version 1.2 released. + * encoder.h (Lee_update_prices): Update high length symbol prices + independently of the value of 'pos_state'. This gives better + compression for large values of '--match-length' without being + slower. + * encoder.h, encoder.cc: Optimize pair price calculations, reducing + compression time for large values of '--match-length' by up to 6%. + * main.cc: New option '-F, --recompress'. + * Makefile.in: 'make install' no longer tries to run + '/sbin/ldconfig' on systems lacking it. + +2011-01-03 Antonio Diaz Diaz + + * Version 1.1 released. + * Compression time has been reduced by 2%. + * All declarations not belonging to the API have been + encapsulated in the namespace 'Lzlib'. + * testsuite: Rename 'test1' to 'test.txt'. New tests. + * Match length limits set by options -1 to -9 of minilzip have + been changed to match those of lzip 1.11. + * main.cc: Set stdin/stdout in binary mode on OS2. + * bbexample.cc: New file containing example functions for + buffer-to-buffer compression/decompression. + +2010-05-08 Antonio Diaz Diaz + + * Version 1.0 released. + * New functions LZ_decompress_member_version, LZ_decompress_data_crc, + LZ_decompress_member_finished, and LZ_decompress_dictionary_size. + * Variables declared 'extern' have been encapsulated in a namespace. + * main.cc: Fix warning about fchown's return value being ignored. + * decoder.h: Integrate Input_buffer in Range_decoder. + +2010-02-10 Antonio Diaz Diaz + + * Version 0.9 released. + * Compression time has been reduced by 8%. + * main.cc: New constant 'o_binary'. + +2010-01-17 Antonio Diaz Diaz + + * Version 0.8 released. + * New functions LZ_decompress_reset, LZ_decompress_sync_to_member, + LZ_decompress_write_size, and LZ_strerror. + * lzlib.h: API change. Replace 'enum' with functions for values of + dictionary size limits to make interface names consistent. + * lzlib.h: API change. Rename 'LZ_errno' to 'LZ_Errno'. + * lzlib.h: API change. Replace 'void *' with 'struct LZ_Encoder *' + and 'struct LZ_Decoder *' to make interface type safe. + * decoder.cc: A truncated member trailer is now correctly detected. + * encoder.cc: Matchfinder::reset now also clears at_stream_end_, + allowing LZ_compress_restart_member to restart a finished stream. + * lzlib.cc: Accept only query or close operations after a fatal + error has occurred. + * The shared version of lzlib is no longer built by default. + * check.sh: Use 'test1' instead of 'COPYING' for testing. + +2009-10-20 Antonio Diaz Diaz + + * Version 0.7 released. + * Compression time has been reduced by 4%. + * check.sh: Remove -9 to run in less than 256MiB of RAM. + * lzcheck.cc: Read files of any size up to 2^63 bytes. + +2009-09-02 Antonio Diaz Diaz + + * Version 0.6 released. + * The LZ_compress_sync_flush mechanism has been fixed. + +2009-07-03 Antonio Diaz Diaz + + * Version 0.5 released. + * Decompression speed has been improved. + * main.cc (signal_handler): Declare as 'extern "C"'. + +2009-06-03 Antonio Diaz Diaz + + * Version 0.4 released. + * New functions LZ_compress_sync_flush and LZ_compress_write_size. + * Decompression speed has been improved. + * lzlib.texinfo: New chapter 'Buffering'. + +2009-05-03 Antonio Diaz Diaz + + * Version 0.3 released. + * Lzlib is now built as a shared library (in addition to static). + +2009-04-26 Antonio Diaz Diaz + + * Version 0.2 released. + * Fix a segfault when decompressing trailing garbage. + * Fix a false positive in LZ_(de)compress_finished. + +2009-04-21 Antonio Diaz Diaz + + * Version 0.1 released. + + +Copyright (C) 2009-2022 Antonio Diaz Diaz. + +This file is a collection of facts, and thus it is not copyrightable, +but just in case, you have unlimited permission to copy, distribute, and +modify it. diff --git a/INSTALL b/INSTALL new file mode 100644 index 0000000..f3b0b94 --- /dev/null +++ b/INSTALL @@ -0,0 +1,80 @@ +Requirements +------------ +You will need a C99 compiler. (gcc 3.3.6 or newer is recommended). +I use gcc 6.1.0 and 3.3.6, but the code should compile with any standards +compliant compiler. +Gcc is available at http://gcc.gnu.org. + +The operating system must allow signal handlers read access to objects with +static storage duration so that the cleanup handler for Control-C can delete +the partial output file. (This requirement is for minilzip only). + + +Procedure +--------- +1. Unpack the archive if you have not done so already: + + tar -xf lzlib[version].tar.lz +or + lzip -cd lzlib[version].tar.lz | tar -xf - + +This creates the directory ./lzlib[version] containing the source from +the main archive. + +2. Change to lzlib directory and run configure. + (Try 'configure --help' for usage instructions). + + cd lzlib[version] + ./configure + + If you are compiling on MinGW, use: + + ./configure CFLAGS+='-D __USE_MINGW_ANSI_STDIO' + +3. Run make. + + make + +4. Optionally, type 'make check' to run the tests that come with lzlib. + +5. Type 'make install' to install the library and any data files and + documentation. (You may need to run ldconfig also). + + Or type 'make install-compress', which additionally compresses the + info manual after installation. + (Installing compressed docs may become the default in the future). + + You can install only the library or the info manual by typing + 'make install-lib' or 'make install-info' respectively. + + 'make install-bin install-man' installs the program minilzip and its man + page. 'install-bin' installs a shared minilzip if the shared library has + been configured. Else it installs a static minilzip. + 'make install-bin-compress' additionally compresses the man page after + installation. + + 'make install-as-lzip' runs 'make install-bin' and then links minilzip to + the name 'lzip'. + + +Another way +----------- +You can also compile lzlib into a separate directory. +To do this, you must use a version of 'make' that supports the variable +'VPATH', such as GNU 'make'. 'cd' to the directory where you want the +object files and executables to go and run the 'configure' script. +'configure' automatically checks for the source code in '.', in '..', and +in the directory that 'configure' is in. + +'configure' recognizes the option '--srcdir=DIR' to control where to +look for the sources. Usually 'configure' can determine that directory +automatically. + +After running 'configure', you can run 'make' and 'make install' as +explained above. + + +Copyright (C) 2009-2022 Antonio Diaz Diaz. + +This file is free documentation: you have unlimited permission to copy, +distribute, and modify it. diff --git a/Makefile.in b/Makefile.in new file mode 100644 index 0000000..81b404b --- /dev/null +++ b/Makefile.in @@ -0,0 +1,202 @@ + +DISTNAME = $(pkgname)-$(pkgversion) +INSTALL = install +INSTALL_PROGRAM = $(INSTALL) -m 755 +INSTALL_DATA = $(INSTALL) -m 644 +INSTALL_DIR = $(INSTALL) -d -m 755 +LDCONFIG = /sbin/ldconfig +SHELL = /bin/sh +CAN_RUN_INSTALLINFO = $(SHELL) -c "install-info --version" > /dev/null 2>&1 + +objs = carg_parser.o minilzip.o + + +.PHONY : all install install-bin install-info install-man \ + install-strip install-compress install-strip-compress \ + install-bin-strip install-info-compress install-man-compress \ + install-bin-compress install-bin-strip-compress \ + install-lib install-lib-strip \ + install-as-lzip \ + uninstall uninstall-bin uninstall-lib uninstall-info uninstall-man \ + doc info man check dist clean distclean + +all : $(progname_static) $(progname_shared) + +lib$(libname).a : lzlib.o + $(AR) $(ARFLAGS) $@ $< + +lib$(libname).so.$(pkgversion) : lzlib_sh.o + $(CC) $(CFLAGS) $(LDFLAGS) -fpic -fPIC -shared -Wl,--soname=lib$(libname).so.$(soversion) -o $@ $< + +$(progname) : $(objs) lib$(libname).a + $(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(objs) lib$(libname).a + +$(progname)_shared : $(objs) lib$(libname).so.$(pkgversion) + $(CC) $(CFLAGS) $(LDFLAGS) -o $@ $(objs) lib$(libname).so.$(pkgversion) + +bbexample : bbexample.o lib$(libname).a + $(CC) $(CFLAGS) $(LDFLAGS) -o $@ bbexample.o lib$(libname).a + +ffexample : ffexample.o lib$(libname).a + $(CC) $(CFLAGS) $(LDFLAGS) -o $@ ffexample.o lib$(libname).a + +lzcheck : lzcheck.o lib$(libname).a + $(CC) $(CFLAGS) $(LDFLAGS) -o $@ lzcheck.o lib$(libname).a + +minilzip.o : minilzip.c + $(CC) $(CPPFLAGS) $(CFLAGS) -DPROGVERSION=\"$(pkgversion)\" -c -o $@ $< + +lzlib_sh.o : lzlib.c + $(CC) $(CPPFLAGS) $(CFLAGS) -fpic -fPIC -c -o $@ $< + +%.o : %.c + $(CC) $(CPPFLAGS) $(CFLAGS) -c -o $@ $< + +lzdeps = lzlib.h lzip.h cbuffer.c decoder.h decoder.c encoder_base.h \ + encoder_base.c encoder.h encoder.c fast_encoder.h fast_encoder.c + +$(objs) : Makefile +carg_parser.o : carg_parser.h +lzlib.o : Makefile $(lzdeps) +lzlib_sh.o : Makefile $(lzdeps) +minilzip.o : carg_parser.h lzlib.h +bbexample.o : Makefile lzlib.h +ffexample.o : Makefile lzlib.h +lzcheck.o : Makefile lzlib.h + + +doc : info man + +info : $(VPATH)/doc/$(pkgname).info + +$(VPATH)/doc/$(pkgname).info : $(VPATH)/doc/$(pkgname).texi + cd $(VPATH)/doc && makeinfo $(pkgname).texi + +man : $(VPATH)/doc/$(progname).1 + +$(VPATH)/doc/$(progname).1 : $(progname) + help2man -n 'reduces the size of files' -o $@ --info-page=$(pkgname) ./$(progname) + +Makefile : $(VPATH)/configure $(VPATH)/Makefile.in + ./config.status + +check : $(progname) bbexample ffexample lzcheck + @$(VPATH)/testsuite/check.sh $(VPATH)/testsuite $(pkgversion) + +install : install-lib install-info +install-strip : install-lib-strip install-info +install-compress : install-lib install-info-compress +install-strip-compress : install-lib-strip install-info-compress +install-bin-compress : install-bin install-man-compress +install-bin-strip-compress : install-bin-strip install-man-compress + +install-bin : all + if [ ! -d "$(DESTDIR)$(bindir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(bindir)" ; fi + $(INSTALL_PROGRAM) ./$(progname_lzip) "$(DESTDIR)$(bindir)/$(progname)" + +install-bin-strip : all + $(MAKE) INSTALL_PROGRAM='$(INSTALL_PROGRAM) -s' install-bin + +install-lib : all + if [ ! -d "$(DESTDIR)$(includedir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(includedir)" ; fi + if [ ! -d "$(DESTDIR)$(libdir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(libdir)" ; fi + $(INSTALL_DATA) $(VPATH)/$(libname)lib.h "$(DESTDIR)$(includedir)/$(libname)lib.h" + if [ -n "$(progname_static)" ] ; then \ + $(INSTALL_DATA) ./lib$(libname).a "$(DESTDIR)$(libdir)/lib$(libname).a" ; \ + fi + if [ -n "$(progname_shared)" ] ; then \ + $(INSTALL_PROGRAM) ./lib$(libname).so.$(pkgversion) "$(DESTDIR)$(libdir)/lib$(libname).so.$(pkgversion)" ; \ + if [ -e "$(DESTDIR)$(libdir)/lib$(libname).so.$(soversion)" ] ; then \ + run_ldconfig=no ; \ + else run_ldconfig=yes ; \ + fi ; \ + rm -f "$(DESTDIR)$(libdir)/lib$(libname).so" ; \ + rm -f "$(DESTDIR)$(libdir)/lib$(libname).so.$(soversion)" ; \ + cd "$(DESTDIR)$(libdir)" && ln -s lib$(libname).so.$(pkgversion) lib$(libname).so ; \ + cd "$(DESTDIR)$(libdir)" && ln -s lib$(libname).so.$(pkgversion) lib$(libname).so.$(soversion) ; \ + if [ "${disable_ldconfig}" != yes ] && [ $${run_ldconfig} = yes ] && \ + [ -x "$(LDCONFIG)" ] ; then "$(LDCONFIG)" -n "$(DESTDIR)$(libdir)" || true ; fi ; \ + fi + +install-lib-strip : all + $(MAKE) INSTALL_PROGRAM='$(INSTALL_PROGRAM) -s' install-lib + +install-info : + if [ ! -d "$(DESTDIR)$(infodir)" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(infodir)" ; fi + -rm -f "$(DESTDIR)$(infodir)/$(pkgname).info"* + $(INSTALL_DATA) $(VPATH)/doc/$(pkgname).info "$(DESTDIR)$(infodir)/$(pkgname).info" + -if $(CAN_RUN_INSTALLINFO) ; then \ + install-info --info-dir="$(DESTDIR)$(infodir)" "$(DESTDIR)$(infodir)/$(pkgname).info" ; \ + fi + +install-info-compress : install-info + lzip -v -9 "$(DESTDIR)$(infodir)/$(pkgname).info" + +install-man : + if [ ! -d "$(DESTDIR)$(mandir)/man1" ] ; then $(INSTALL_DIR) "$(DESTDIR)$(mandir)/man1" ; fi + -rm -f "$(DESTDIR)$(mandir)/man1/$(progname).1"* + $(INSTALL_DATA) $(VPATH)/doc/$(progname).1 "$(DESTDIR)$(mandir)/man1/$(progname).1" + +install-man-compress : install-man + lzip -v -9 "$(DESTDIR)$(mandir)/man1/$(progname).1" + +install-as-lzip : install-bin + -rm -f "$(DESTDIR)$(bindir)/lzip" + cd "$(DESTDIR)$(bindir)" && ln -s $(progname) lzip + +uninstall : uninstall-info uninstall-lib + +uninstall-bin : + -rm -f "$(DESTDIR)$(bindir)/$(progname)" + +uninstall-lib : + -rm -f "$(DESTDIR)$(includedir)/$(libname)lib.h" + -rm -f "$(DESTDIR)$(libdir)/lib$(libname).a" + -rm -f "$(DESTDIR)$(libdir)/lib$(libname).so" + -rm -f "$(DESTDIR)$(libdir)/lib$(libname).so.$(soversion)" + -rm -f "$(DESTDIR)$(libdir)/lib$(libname).so.$(pkgversion)" + +uninstall-info : + -if $(CAN_RUN_INSTALLINFO) ; then \ + install-info --info-dir="$(DESTDIR)$(infodir)" --remove "$(DESTDIR)$(infodir)/$(pkgname).info" ; \ + fi + -rm -f "$(DESTDIR)$(infodir)/$(pkgname).info"* + +uninstall-man : + -rm -f "$(DESTDIR)$(mandir)/man1/$(progname).1"* + +dist : doc + ln -sf $(VPATH) $(DISTNAME) + tar -Hustar --owner=root --group=root -cvf $(DISTNAME).tar \ + $(DISTNAME)/AUTHORS \ + $(DISTNAME)/COPYING \ + $(DISTNAME)/COPYING.GPL \ + $(DISTNAME)/ChangeLog \ + $(DISTNAME)/INSTALL \ + $(DISTNAME)/Makefile.in \ + $(DISTNAME)/NEWS \ + $(DISTNAME)/README \ + $(DISTNAME)/configure \ + $(DISTNAME)/doc/$(progname).1 \ + $(DISTNAME)/doc/$(pkgname).info \ + $(DISTNAME)/doc/$(pkgname).texi \ + $(DISTNAME)/*.h \ + $(DISTNAME)/*.c \ + $(DISTNAME)/testsuite/check.sh \ + $(DISTNAME)/testsuite/test.txt \ + $(DISTNAME)/testsuite/fox_lf \ + $(DISTNAME)/testsuite/fox.lz \ + $(DISTNAME)/testsuite/fox_*.lz \ + $(DISTNAME)/testsuite/test_sync.lz \ + $(DISTNAME)/testsuite/test.txt.lz \ + $(DISTNAME)/testsuite/test_em.txt.lz + rm -f $(DISTNAME) + lzip -v -9 $(DISTNAME).tar + +clean : + -rm -f $(progname) $(objs) lzlib.o lib$(libname).a + -rm -f $(progname)_shared lzlib_sh.o lib$(libname).so* + -rm -f bbexample bbexample.o ffexample ffexample.o lzcheck lzcheck.o + +distclean : clean + -rm -f Makefile config.status *.tar *.tar.lz diff --git a/NEWS b/NEWS new file mode 100644 index 0000000..e2ceb38 --- /dev/null +++ b/NEWS @@ -0,0 +1,15 @@ +Changes in version 1.13: + +The variables AR and ARFLAGS can now be set from configure. (Before you +needed to run 'make AR='. (Reported by Hoël Bézier). + +In case of error in a numerical argument to a command line option, minilzip +now shows the name of the option and the range of valid values. + +'minilzip --check-lib' now checks that LZ_API_VERSION and LZ_version_string +match. + +Several descriptions have been improved in manual, '--help', and man page. + +The texinfo category of the manual has been changed from 'Data Compression' +to 'Compression' to match that of gzip. (Reported by Alfred M. Szmidt). diff --git a/README b/README new file mode 100644 index 0000000..86a0cb5 --- /dev/null +++ b/README @@ -0,0 +1,103 @@ +Description + +Lzlib is a data compression library providing in-memory LZMA compression and +decompression functions, including integrity checking of the decompressed +data. The compressed data format used by the library is the lzip format. +Lzlib is written in C. + +The lzip file format is designed for data sharing and long-term archiving, +taking into account both data integrity and decoder availability: + + * The lzip format provides very safe integrity checking and some data + recovery means. The program lziprecover can repair bit flip errors + (one of the most common forms of data corruption) in lzip files, and + provides data recovery capabilities, including error-checked merging + of damaged copies of a file. + + * The lzip format is as simple as possible (but not simpler). The lzip + manual provides the source code of a simple decompressor along with a + detailed explanation of how it works, so that with the only help of the + lzip manual it would be possible for a digital archaeologist to extract + the data from a lzip file long after quantum computers eventually + render LZMA obsolete. + + * Additionally the lzip reference implementation is copylefted, which + guarantees that it will remain free forever. + +A nice feature of the lzip format is that a corrupt byte is easier to repair +the nearer it is from the beginning of the file. Therefore, with the help of +lziprecover, losing an entire archive just because of a corrupt byte near +the beginning is a thing of the past. + +The functions and variables forming the interface of the compression library +are declared in the file 'lzlib.h'. Usage examples of the library are given +in the files 'bbexample.c', 'ffexample.c', and 'minilzip.c' from the source +distribution. + +All the library functions are thread safe. The library does not install any +signal handler. The decoder checks the consistency of the compressed data, +so the library should never crash even in case of corrupted input. + +Compression/decompression is done by repeatedly calling a couple of +read/write functions until all the data have been processed by the library. +This interface is safer and less error prone than the traditional zlib +interface. + +Compression/decompression is done when the read function is called. This +means the value returned by the position functions will not be updated until +a read call, even if a lot of data are written. If you want the data to be +compressed in advance, just call the read function with a size equal to 0. + +If all the data to be compressed are written in advance, lzlib will +automatically adjust the header of the compressed data to use the largest +dictionary size that does not exceed neither the data size nor the limit +given to 'LZ_compress_open'. This feature reduces the amount of memory +needed for decompression and allows minilzip to produce identical compressed +output as lzip. + +Lzlib will correctly decompress a data stream which is the concatenation of +two or more compressed data streams. The result is the concatenation of the +corresponding decompressed data streams. Integrity testing of concatenated +compressed data streams is also supported. + +Lzlib is able to compress and decompress streams of unlimited size by +automatically creating multimember output. The members so created are large, +about 2 PiB each. + +In spite of its name (Lempel-Ziv-Markov chain-Algorithm), LZMA is not a +concrete algorithm; it is more like "any algorithm using the LZMA coding +scheme". For example, the option '-0' of lzip uses the scheme in almost the +simplest way possible; issuing the longest match it can find, or a literal +byte if it can't find a match. Inversely, a much more elaborated way of +finding coding sequences of minimum size than the one currently used by lzip +could be developed, and the resulting sequence could also be coded using the +LZMA coding scheme. + +Lzlib currently implements two variants of the LZMA algorithm: fast (used by +option '-0' of minilzip) and normal (used by all other compression levels). + +The high compression of LZMA comes from combining two basic, well-proven +compression ideas: sliding dictionaries (LZ77/78) and markov models (the +thing used by every compression algorithm that uses a range encoder or +similar order-0 entropy coder as its last stage) with segregation of +contexts according to what the bits are used for. + +The ideas embodied in lzlib are due to (at least) the following people: +Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the +definition of Markov chains), G.N.N. Martin (for the definition of range +encoding), Igor Pavlov (for putting all the above together in LZMA), and +Julian Seward (for bzip2's CLI). + +LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never have +been compressed. Decompressed is used to refer to data which have undergone +the process of decompression. + + +Copyright (C) 2009-2022 Antonio Diaz Diaz. + +This file is free documentation: you have unlimited permission to copy, +distribute, and modify it. + +The file Makefile.in is a data file used by configure to produce the +Makefile. It has the same copyright owner and permissions that configure +itself. diff --git a/bbexample.c b/bbexample.c new file mode 100644 index 0000000..074f7ae --- /dev/null +++ b/bbexample.c @@ -0,0 +1,367 @@ +/* Buffer to buffer example - Test program for the library lzlib + Copyright (C) 2010-2022 Antonio Diaz Diaz. + + This program is free software: you have unlimited permission + to copy, distribute, and modify it. + + Usage: bbexample filename + + This program is an example of how buffer-to-buffer + compression/decompression can be implemented using lzlib. +*/ + +#define _FILE_OFFSET_BITS 64 + +#include +#include +#include +#include +#include +#include +#include +#include + +#include "lzlib.h" + +#ifndef min + #define min(x,y) ((x) <= (y) ? (x) : (y)) +#endif + + +/* Return the address of a malloc'd buffer containing the file data and + the file size in '*file_sizep'. + In case of error, return 0 and do not modify '*file_sizep'. +*/ +uint8_t * read_file( const char * const name, long * const file_sizep ) + { + long buffer_size = 1 << 20, file_size; + uint8_t * buffer, * tmp; + FILE * const f = fopen( name, "rb" ); + if( !f ) + { fprintf( stderr, "bbexample: Can't open input file '%s': %s\n", + name, strerror( errno ) ); return 0; } + + buffer = (uint8_t *)malloc( buffer_size ); + if( !buffer ) + { fputs( "bbexample: read_file: Not enough memory.\n", stderr ); + fclose( f ); return 0; } + file_size = fread( buffer, 1, buffer_size, f ); + while( file_size >= buffer_size ) + { + if( buffer_size >= LONG_MAX ) + { + fprintf( stderr, "bbexample: Input file '%s' is too large.\n", name ); + free( buffer ); fclose( f ); return 0; + } + buffer_size = ( buffer_size <= LONG_MAX / 2 ) ? 2 * buffer_size : LONG_MAX; + tmp = (uint8_t *)realloc( buffer, buffer_size ); + if( !tmp ) + { fputs( "bbexample: read_file: Not enough memory.\n", stderr ); + free( buffer ); fclose( f ); return 0; } + buffer = tmp; + file_size += fread( buffer + file_size, 1, buffer_size - file_size, f ); + } + if( ferror( f ) || !feof( f ) ) + { + fprintf( stderr, "bbexample: Error reading file '%s': %s\n", + name, strerror( errno ) ); + free( buffer ); fclose( f ); return 0; + } + fclose( f ); + *file_sizep = file_size; + return buffer; + } + + +/* Compress 'insize' bytes from 'inbuf'. + Return the address of a malloc'd buffer containing the compressed data, + and the size of the data in '*outlenp'. + In case of error, return 0 and do not modify '*outlenp'. +*/ +uint8_t * bbcompressl( const uint8_t * const inbuf, const long insize, + const int level, long * const outlenp ) + { + struct Lzma_options + { + int dictionary_size; /* 4 KiB .. 512 MiB */ + int match_len_limit; /* 5 .. 273 */ + }; + /* Mapping from gzip/bzip2 style 1..9 compression modes + to the corresponding LZMA compression modes. */ + const struct Lzma_options option_mapping[] = + { + { 65535, 16 }, /* -0 (65535,16 chooses fast encoder) */ + { 1 << 20, 5 }, /* -1 */ + { 3 << 19, 6 }, /* -2 */ + { 1 << 21, 8 }, /* -3 */ + { 3 << 20, 12 }, /* -4 */ + { 1 << 22, 20 }, /* -5 */ + { 1 << 23, 36 }, /* -6 */ + { 1 << 24, 68 }, /* -7 */ + { 3 << 23, 132 }, /* -8 */ + { 1 << 25, 273 } }; /* -9 */ + struct Lzma_options encoder_options; + struct LZ_Encoder * encoder; + uint8_t * outbuf; + const long delta_size = ( insize / 4 ) + 64; /* insize may be zero */ + long outsize = delta_size; /* initial outsize */ + long inpos = 0; + long outpos = 0; + bool error = false; + + if( level < 0 || level > 9 ) return 0; + encoder_options = option_mapping[level]; + + if( encoder_options.dictionary_size > insize && level != 0 ) + encoder_options.dictionary_size = insize; /* saves memory */ + if( encoder_options.dictionary_size < LZ_min_dictionary_size() ) + encoder_options.dictionary_size = LZ_min_dictionary_size(); + encoder = LZ_compress_open( encoder_options.dictionary_size, + encoder_options.match_len_limit, INT64_MAX ); + outbuf = (uint8_t *)malloc( outsize ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok || !outbuf ) + { free( outbuf ); LZ_compress_close( encoder ); return 0; } + + while( true ) + { + int ret = LZ_compress_write( encoder, inbuf + inpos, + min( INT_MAX, insize - inpos ) ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_compress_finish( encoder ); + ret = LZ_compress_read( encoder, outbuf + outpos, + min( INT_MAX, outsize - outpos ) ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_compress_finished( encoder ) == 1 ) break; + if( outpos >= outsize ) + { + uint8_t * tmp; + if( outsize > LONG_MAX - delta_size ) { error = true; break; } + outsize += delta_size; + tmp = (uint8_t *)realloc( outbuf, outsize ); + if( !tmp ) { error = true; break; } + outbuf = tmp; + } + } + + if( LZ_compress_close( encoder ) < 0 ) error = true; + if( error ) { free( outbuf ); return 0; } + *outlenp = outpos; + return outbuf; + } + + +/* Decompress 'insize' bytes from 'inbuf'. + Return the address of a malloc'd buffer containing the decompressed + data, and the size of the data in '*outlenp'. + In case of error, return 0 and do not modify '*outlenp'. +*/ +uint8_t * bbdecompressl( const uint8_t * const inbuf, const long insize, + long * const outlenp ) + { + struct LZ_Decoder * const decoder = LZ_decompress_open(); + const long delta_size = insize; /* insize must be > zero */ + long outsize = delta_size; /* initial outsize */ + uint8_t * outbuf = (uint8_t *)malloc( outsize ); + long inpos = 0; + long outpos = 0; + bool error = false; + if( !decoder || LZ_decompress_errno( decoder ) != LZ_ok || !outbuf ) + { free( outbuf ); LZ_decompress_close( decoder ); return 0; } + + while( true ) + { + int ret = LZ_decompress_write( decoder, inbuf + inpos, + min( INT_MAX, insize - inpos ) ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_decompress_finish( decoder ); + ret = LZ_decompress_read( decoder, outbuf + outpos, + min( INT_MAX, outsize - outpos ) ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_decompress_finished( decoder ) == 1 ) break; + if( outpos >= outsize ) + { + uint8_t * tmp; + if( outsize > LONG_MAX - delta_size ) { error = true; break; } + outsize += delta_size; + tmp = (uint8_t *)realloc( outbuf, outsize ); + if( !tmp ) { error = true; break; } + outbuf = tmp; + } + } + + if( LZ_decompress_close( decoder ) < 0 ) error = true; + if( error ) { free( outbuf ); return 0; } + *outlenp = outpos; + return outbuf; + } + + +/* Test the whole file at all levels. */ +int full_test( const uint8_t * const inbuf, const long insize ) + { + int level; + for( level = 0; level <= 9; ++level ) + { + long midsize = 0, outsize = 0; + uint8_t * outbuf; + uint8_t * midbuf = bbcompressl( inbuf, insize, level, &midsize ); + if( !midbuf ) + { fputs( "bbexample: full_test: Not enough memory or compress error.\n", + stderr ); return 1; } + + outbuf = bbdecompressl( midbuf, midsize, &outsize ); + free( midbuf ); + if( !outbuf ) + { fputs( "bbexample: full_test: Not enough memory or decompress error.\n", + stderr ); return 1; } + + if( insize != outsize || + ( insize > 0 && memcmp( inbuf, outbuf, insize ) != 0 ) ) + { fputs( "bbexample: full_test: Decompressed data differs from original.\n", + stderr ); free( outbuf ); return 1; } + + free( outbuf ); + } + return 0; + } + + +/* Compress 'insize' bytes from 'inbuf' to 'outbuf'. + Return the size of the compressed data in '*outlenp'. + In case of error, or if 'outsize' is too small, return false and do not + modify '*outlenp'. +*/ +bool bbcompress( const uint8_t * const inbuf, const int insize, + const int dictionary_size, const int match_len_limit, + uint8_t * const outbuf, const int outsize, + int * const outlenp ) + { + int inpos = 0, outpos = 0; + bool error = false; + struct LZ_Encoder * const encoder = + LZ_compress_open( dictionary_size, match_len_limit, INT64_MAX ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { LZ_compress_close( encoder ); return false; } + + while( true ) + { + int ret = LZ_compress_write( encoder, inbuf + inpos, insize - inpos ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_compress_finish( encoder ); + ret = LZ_compress_read( encoder, outbuf + outpos, outsize - outpos ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_compress_finished( encoder ) == 1 ) break; + if( outpos >= outsize ) { error = true; break; } + } + + if( LZ_compress_close( encoder ) < 0 ) error = true; + if( error ) return false; + *outlenp = outpos; + return true; + } + + +/* Decompress 'insize' bytes from 'inbuf' to 'outbuf'. + Return the size of the decompressed data in '*outlenp'. + In case of error, or if 'outsize' is too small, return false and do not + modify '*outlenp'. +*/ +bool bbdecompress( const uint8_t * const inbuf, const int insize, + uint8_t * const outbuf, const int outsize, + int * const outlenp ) + { + int inpos = 0, outpos = 0; + bool error = false; + struct LZ_Decoder * const decoder = LZ_decompress_open(); + if( !decoder || LZ_decompress_errno( decoder ) != LZ_ok ) + { LZ_decompress_close( decoder ); return false; } + + while( true ) + { + int ret = LZ_decompress_write( decoder, inbuf + inpos, insize - inpos ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_decompress_finish( decoder ); + ret = LZ_decompress_read( decoder, outbuf + outpos, outsize - outpos ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_decompress_finished( decoder ) == 1 ) break; + if( outpos >= outsize ) { error = true; break; } + } + + if( LZ_decompress_close( decoder ) < 0 ) error = true; + if( error ) return false; + *outlenp = outpos; + return true; + } + + +/* Test at most INT_MAX bytes from the file with buffers of fixed size. */ +int fixed_test( const uint8_t * const inbuf, const int insize ) + { + int dictionary_size = 65535; /* fast encoder */ + int midsize = min( INT_MAX, ( insize / 8 ) * 9LL + 44 ), outsize = insize; + uint8_t * midbuf = (uint8_t *)malloc( midsize ); + uint8_t * outbuf = (uint8_t *)malloc( outsize ); + if( !midbuf || !outbuf ) + { fputs( "bbexample: fixed_test: Not enough memory.\n", stderr ); + free( outbuf ); free( midbuf ); return 1; } + + for( ; dictionary_size <= 8 << 20; dictionary_size += 8323073 ) + { + int midlen, outlen; + if( !bbcompress( inbuf, insize, dictionary_size, 16, midbuf, midsize, &midlen ) ) + { fputs( "bbexample: fixed_test: Not enough memory or compress error.\n", + stderr ); free( outbuf ); free( midbuf ); return 1; } + + if( !bbdecompress( midbuf, midlen, outbuf, outsize, &outlen ) ) + { fputs( "bbexample: fixed_test: Not enough memory or decompress error.\n", + stderr ); free( outbuf ); free( midbuf ); return 1; } + + if( insize != outlen || + ( insize > 0 && memcmp( inbuf, outbuf, insize ) != 0 ) ) + { fputs( "bbexample: fixed_test: Decompressed data differs from original.\n", + stderr ); free( outbuf ); free( midbuf ); return 1; } + + } + free( outbuf ); + free( midbuf ); + return 0; + } + + +int main( const int argc, const char * const argv[] ) + { + int retval = 0, i; + int open_failures = 0; + const bool verbose = ( argc > 2 ); + + if( argc < 2 ) + { + fputs( "Usage: bbexample filename\n", stderr ); + return 1; + } + + for( i = 1; i < argc && retval == 0; ++i ) + { + long insize; + uint8_t * const inbuf = read_file( argv[i], &insize ); + if( !inbuf ) { ++open_failures; continue; } + if( verbose ) fprintf( stderr, " Testing file '%s'\n", argv[i] ); + + retval = full_test( inbuf, insize ); + if( retval == 0 ) retval = fixed_test( inbuf, min( INT_MAX, insize ) ); + free( inbuf ); + } + if( open_failures > 0 && verbose ) + fprintf( stderr, "bbexample: warning: %d %s failed to open.\n", + open_failures, ( open_failures == 1 ) ? "file" : "files" ); + if( retval == 0 && open_failures ) retval = 1; + return retval; + } diff --git a/carg_parser.c b/carg_parser.c new file mode 100644 index 0000000..181ba23 --- /dev/null +++ b/carg_parser.c @@ -0,0 +1,319 @@ +/* Arg_parser - POSIX/GNU command line argument parser. (C version) + Copyright (C) 2006-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +#include +#include + +#include "carg_parser.h" + + +/* assure at least a minimum size for buffer 'buf' */ +static void * ap_resize_buffer( void * buf, const int min_size ) + { + if( buf ) buf = realloc( buf, min_size ); + else buf = malloc( min_size ); + return buf; + } + + +static char push_back_record( struct Arg_parser * const ap, const int code, + const char * const long_name, + const char * const argument ) + { + struct ap_Record * p; + void * tmp = ap_resize_buffer( ap->data, + ( ap->data_size + 1 ) * sizeof (struct ap_Record) ); + if( !tmp ) return 0; + ap->data = (struct ap_Record *)tmp; + p = &(ap->data[ap->data_size]); + p->code = code; + if( long_name ) + { + const int len = strlen( long_name ); + p->parsed_name = (char *)malloc( len + 2 + 1 ); + if( !p->parsed_name ) return 0; + p->parsed_name[0] = p->parsed_name[1] = '-'; + strncpy( p->parsed_name + 2, long_name, len + 1 ); + } + else if( code > 0 && code < 256 ) + { + p->parsed_name = (char *)malloc( 2 + 1 ); + if( !p->parsed_name ) return 0; + p->parsed_name[0] = '-'; p->parsed_name[1] = code; p->parsed_name[2] = 0; + } + else p->parsed_name = 0; + if( argument ) + { + const int len = strlen( argument ); + p->argument = (char *)malloc( len + 1 ); + if( !p->argument ) { free( p->parsed_name ); return 0; } + strncpy( p->argument, argument, len + 1 ); + } + else p->argument = 0; + ++ap->data_size; + return 1; + } + + +static char add_error( struct Arg_parser * const ap, const char * const msg ) + { + const int len = strlen( msg ); + void * tmp = ap_resize_buffer( ap->error, ap->error_size + len + 1 ); + if( !tmp ) return 0; + ap->error = (char *)tmp; + strncpy( ap->error + ap->error_size, msg, len + 1 ); + ap->error_size += len; + return 1; + } + + +static void free_data( struct Arg_parser * const ap ) + { + int i; + for( i = 0; i < ap->data_size; ++i ) + { free( ap->data[i].argument ); free( ap->data[i].parsed_name ); } + if( ap->data ) { free( ap->data ); ap->data = 0; } + ap->data_size = 0; + } + + +/* Return 0 only if out of memory. */ +static char parse_long_option( struct Arg_parser * const ap, + const char * const opt, const char * const arg, + const struct ap_Option options[], + int * const argindp ) + { + unsigned len; + int index = -1, i; + char exact = 0, ambig = 0; + + for( len = 0; opt[len+2] && opt[len+2] != '='; ++len ) ; + + /* Test all long options for either exact match or abbreviated matches. */ + for( i = 0; options[i].code != 0; ++i ) + if( options[i].long_name && + strncmp( options[i].long_name, &opt[2], len ) == 0 ) + { + if( strlen( options[i].long_name ) == len ) /* Exact match found */ + { index = i; exact = 1; break; } + else if( index < 0 ) index = i; /* First nonexact match found */ + else if( options[index].code != options[i].code || + options[index].has_arg != options[i].has_arg ) + ambig = 1; /* Second or later nonexact match found */ + } + + if( ambig && !exact ) + { + add_error( ap, "option '" ); add_error( ap, opt ); + add_error( ap, "' is ambiguous" ); + return 1; + } + + if( index < 0 ) /* nothing found */ + { + add_error( ap, "unrecognized option '" ); add_error( ap, opt ); + add_error( ap, "'" ); + return 1; + } + + ++*argindp; + + if( opt[len+2] ) /* '--=' syntax */ + { + if( options[index].has_arg == ap_no ) + { + add_error( ap, "option '--" ); add_error( ap, options[index].long_name ); + add_error( ap, "' doesn't allow an argument" ); + return 1; + } + if( options[index].has_arg == ap_yes && !opt[len+3] ) + { + add_error( ap, "option '--" ); add_error( ap, options[index].long_name ); + add_error( ap, "' requires an argument" ); + return 1; + } + return push_back_record( ap, options[index].code, + options[index].long_name, &opt[len+3] ); + } + + if( options[index].has_arg == ap_yes ) + { + if( !arg || !arg[0] ) + { + add_error( ap, "option '--" ); add_error( ap, options[index].long_name ); + add_error( ap, "' requires an argument" ); + return 1; + } + ++*argindp; + return push_back_record( ap, options[index].code, + options[index].long_name, arg ); + } + + return push_back_record( ap, options[index].code, + options[index].long_name, 0 ); + } + + +/* Return 0 only if out of memory. */ +static char parse_short_option( struct Arg_parser * const ap, + const char * const opt, const char * const arg, + const struct ap_Option options[], + int * const argindp ) + { + int cind = 1; /* character index in opt */ + + while( cind > 0 ) + { + int index = -1, i; + const unsigned char c = opt[cind]; + char code_str[2]; + code_str[0] = c; code_str[1] = 0; + + if( c != 0 ) + for( i = 0; options[i].code; ++i ) + if( c == options[i].code ) + { index = i; break; } + + if( index < 0 ) + { + add_error( ap, "invalid option -- '" ); add_error( ap, code_str ); + add_error( ap, "'" ); + return 1; + } + + if( opt[++cind] == 0 ) { ++*argindp; cind = 0; } /* opt finished */ + + if( options[index].has_arg != ap_no && cind > 0 && opt[cind] ) + { + if( !push_back_record( ap, c, 0, &opt[cind] ) ) return 0; + ++*argindp; cind = 0; + } + else if( options[index].has_arg == ap_yes ) + { + if( !arg || !arg[0] ) + { + add_error( ap, "option requires an argument -- '" ); + add_error( ap, code_str ); add_error( ap, "'" ); + return 1; + } + ++*argindp; cind = 0; + if( !push_back_record( ap, c, 0, arg ) ) return 0; + } + else if( !push_back_record( ap, c, 0, 0 ) ) return 0; + } + return 1; + } + + +char ap_init( struct Arg_parser * const ap, + const int argc, const char * const argv[], + const struct ap_Option options[], const char in_order ) + { + const char ** non_options = 0; /* skipped non-options */ + int non_options_size = 0; /* number of skipped non-options */ + int argind = 1; /* index in argv */ + char done = 0; /* false until success */ + + ap->data = 0; + ap->error = 0; + ap->data_size = 0; + ap->error_size = 0; + if( argc < 2 || !argv || !options ) return 1; + + while( argind < argc ) + { + const unsigned char ch1 = argv[argind][0]; + const unsigned char ch2 = ch1 ? argv[argind][1] : 0; + + if( ch1 == '-' && ch2 ) /* we found an option */ + { + const char * const opt = argv[argind]; + const char * const arg = ( argind + 1 < argc ) ? argv[argind+1] : 0; + if( ch2 == '-' ) + { + if( !argv[argind][2] ) { ++argind; break; } /* we found "--" */ + else if( !parse_long_option( ap, opt, arg, options, &argind ) ) goto out; + } + else if( !parse_short_option( ap, opt, arg, options, &argind ) ) goto out; + if( ap->error ) break; + } + else + { + if( in_order ) + { if( !push_back_record( ap, 0, 0, argv[argind++] ) ) goto out; } + else + { + void * tmp = ap_resize_buffer( non_options, + ( non_options_size + 1 ) * sizeof *non_options ); + if( !tmp ) goto out; + non_options = (const char **)tmp; + non_options[non_options_size++] = argv[argind++]; + } + } + } + if( ap->error ) free_data( ap ); + else + { + int i; + for( i = 0; i < non_options_size; ++i ) + if( !push_back_record( ap, 0, 0, non_options[i] ) ) goto out; + while( argind < argc ) + if( !push_back_record( ap, 0, 0, argv[argind++] ) ) goto out; + } + done = 1; +out: if( non_options ) free( non_options ); + return done; + } + + +void ap_free( struct Arg_parser * const ap ) + { + free_data( ap ); + if( ap->error ) { free( ap->error ); ap->error = 0; } + ap->error_size = 0; + } + + +const char * ap_error( const struct Arg_parser * const ap ) + { return ap->error; } + + +int ap_arguments( const struct Arg_parser * const ap ) + { return ap->data_size; } + + +int ap_code( const struct Arg_parser * const ap, const int i ) + { + if( i < 0 || i >= ap_arguments( ap ) ) return 0; + return ap->data[i].code; + } + + +const char * ap_parsed_name( const struct Arg_parser * const ap, const int i ) + { + if( i < 0 || i >= ap_arguments( ap ) || !ap->data[i].parsed_name ) return ""; + return ap->data[i].parsed_name; + } + + +const char * ap_argument( const struct Arg_parser * const ap, const int i ) + { + if( i < 0 || i >= ap_arguments( ap ) || !ap->data[i].argument ) return ""; + return ap->data[i].argument; + } diff --git a/carg_parser.h b/carg_parser.h new file mode 100644 index 0000000..0c64861 --- /dev/null +++ b/carg_parser.h @@ -0,0 +1,97 @@ +/* Arg_parser - POSIX/GNU command line argument parser. (C version) + Copyright (C) 2006-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +/* Arg_parser reads the arguments in 'argv' and creates a number of + option codes, option arguments, and non-option arguments. + + In case of error, 'ap_error' returns a non-null pointer to an error + message. + + 'options' is an array of 'struct ap_Option' terminated by an element + containing a code which is zero. A null long_name means a short-only + option. A code value outside the unsigned char range means a long-only + option. + + Arg_parser normally makes it appear as if all the option arguments + were specified before all the non-option arguments for the purposes + of parsing, even if the user of your program intermixed option and + non-option arguments. If you want the arguments in the exact order + the user typed them, call 'ap_init' with 'in_order' = true. + + The argument '--' terminates all options; any following arguments are + treated as non-option arguments, even if they begin with a hyphen. + + The syntax for optional option arguments is '-' + (without whitespace), or '--='. +*/ + +#ifdef __cplusplus +extern "C" { +#endif + +enum ap_Has_arg { ap_no, ap_yes, ap_maybe }; + +struct ap_Option + { + int code; /* Short option letter or code ( code != 0 ) */ + const char * long_name; /* Long option name (maybe null) */ + enum ap_Has_arg has_arg; + }; + + +struct ap_Record + { + int code; + char * parsed_name; + char * argument; + }; + + +struct Arg_parser + { + struct ap_Record * data; + char * error; + int data_size; + int error_size; + }; + + +char ap_init( struct Arg_parser * const ap, + const int argc, const char * const argv[], + const struct ap_Option options[], const char in_order ); + +void ap_free( struct Arg_parser * const ap ); + +const char * ap_error( const struct Arg_parser * const ap ); + +/* The number of arguments parsed. May be different from argc. */ +int ap_arguments( const struct Arg_parser * const ap ); + +/* If ap_code( i ) is 0, ap_argument( i ) is a non-option. + Else ap_argument( i ) is the option's argument (or empty). */ +int ap_code( const struct Arg_parser * const ap, const int i ); + +/* Full name of the option parsed (short or long). */ +const char * ap_parsed_name( const struct Arg_parser * const ap, const int i ); + +const char * ap_argument( const struct Arg_parser * const ap, const int i ); + +#ifdef __cplusplus +} +#endif diff --git a/cbuffer.c b/cbuffer.c new file mode 100644 index 0000000..812de42 --- /dev/null +++ b/cbuffer.c @@ -0,0 +1,143 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +struct Circular_buffer + { + uint8_t * buffer; + unsigned buffer_size; /* capacity == buffer_size - 1 */ + unsigned get; /* buffer is empty when get == put */ + unsigned put; + }; + +static inline bool Cb_init( struct Circular_buffer * const cb, + const unsigned buf_size ) + { + cb->buffer_size = buf_size + 1; + cb->get = 0; + cb->put = 0; + cb->buffer = + ( cb->buffer_size > 1 ) ? (uint8_t *)malloc( cb->buffer_size ) : 0; + return ( cb->buffer != 0 ); + } + +static inline void Cb_free( struct Circular_buffer * const cb ) + { free( cb->buffer ); cb->buffer = 0; } + +static inline void Cb_reset( struct Circular_buffer * const cb ) + { cb->get = 0; cb->put = 0; } + +static inline unsigned Cb_empty( const struct Circular_buffer * const cb ) + { return cb->get == cb->put; } + +static inline unsigned Cb_used_bytes( const struct Circular_buffer * const cb ) + { return ( (cb->get <= cb->put) ? 0 : cb->buffer_size ) + cb->put - cb->get; } + +static inline unsigned Cb_free_bytes( const struct Circular_buffer * const cb ) + { return ( (cb->get <= cb->put) ? cb->buffer_size : 0 ) - cb->put + cb->get - 1; } + +static inline uint8_t Cb_get_byte( struct Circular_buffer * const cb ) + { + const uint8_t b = cb->buffer[cb->get]; + if( ++cb->get >= cb->buffer_size ) cb->get = 0; + return b; + } + +static inline void Cb_put_byte( struct Circular_buffer * const cb, + const uint8_t b ) + { + cb->buffer[cb->put] = b; + if( ++cb->put >= cb->buffer_size ) cb->put = 0; + } + + +static bool Cb_unread_data( struct Circular_buffer * const cb, + const unsigned size ) + { + if( size > Cb_free_bytes( cb ) ) return false; + if( cb->get >= size ) cb->get -= size; + else cb->get = cb->buffer_size - size + cb->get; + return true; + } + + +/* Copy up to 'out_size' bytes to 'out_buffer' and update 'get'. + If 'out_buffer' is null, the bytes are discarded. + Return the number of bytes copied or discarded. +*/ +static unsigned Cb_read_data( struct Circular_buffer * const cb, + uint8_t * const out_buffer, + const unsigned out_size ) + { + unsigned size = 0; + if( out_size == 0 ) return 0; + if( cb->get > cb->put ) + { + size = min( cb->buffer_size - cb->get, out_size ); + if( size > 0 ) + { + if( out_buffer ) memcpy( out_buffer, cb->buffer + cb->get, size ); + cb->get += size; + if( cb->get >= cb->buffer_size ) cb->get = 0; + } + } + if( cb->get < cb->put ) + { + const unsigned size2 = min( cb->put - cb->get, out_size - size ); + if( size2 > 0 ) + { + if( out_buffer ) memcpy( out_buffer + size, cb->buffer + cb->get, size2 ); + cb->get += size2; + size += size2; + } + } + return size; + } + + +/* Copy up to 'in_size' bytes from 'in_buffer' and update 'put'. + Return the number of bytes copied. +*/ +static unsigned Cb_write_data( struct Circular_buffer * const cb, + const uint8_t * const in_buffer, + const unsigned in_size ) + { + unsigned size = 0; + if( in_size == 0 ) return 0; + if( cb->put >= cb->get ) + { + size = min( cb->buffer_size - cb->put - (cb->get == 0), in_size ); + if( size > 0 ) + { + memcpy( cb->buffer + cb->put, in_buffer, size ); + cb->put += size; + if( cb->put >= cb->buffer_size ) cb->put = 0; + } + } + if( cb->put < cb->get ) + { + const unsigned size2 = min( cb->get - cb->put - 1, in_size - size ); + if( size2 > 0 ) + { + memcpy( cb->buffer + cb->put, in_buffer + size, size2 ); + cb->put += size2; + size += size2; + } + } + return size; + } diff --git a/configure b/configure new file mode 100755 index 0000000..4060472 --- /dev/null +++ b/configure @@ -0,0 +1,239 @@ +#! /bin/sh +# configure script for Lzlib - Compression library for the lzip format +# Copyright (C) 2009-2022 Antonio Diaz Diaz. +# +# This configure script is free software: you have unlimited permission +# to copy, distribute, and modify it. + +pkgname=lzlib +pkgversion=1.13 +soversion=1 +progname=minilzip +progname_static=${progname} +progname_shared= +progname_lzip=${progname} +disable_ldconfig= +libname=lz +srctrigger=doc/${pkgname}.texi + +# clear some things potentially inherited from environment. +LC_ALL=C +export LC_ALL +srcdir= +prefix=/usr/local +exec_prefix='$(prefix)' +bindir='$(exec_prefix)/bin' +datarootdir='$(prefix)/share' +includedir='$(prefix)/include' +infodir='$(datarootdir)/info' +libdir='$(exec_prefix)/lib' +mandir='$(datarootdir)/man' +CC=gcc +AR=ar +CPPFLAGS= +CFLAGS='-Wall -W -O2' +LDFLAGS= +ARFLAGS=-rcs + +# checking whether we are using GNU C. +/bin/sh -c "${CC} --version" > /dev/null 2>&1 || { CC=cc ; CFLAGS=-O2 ; } + +# Loop over all args +args= +no_create= +while [ $# != 0 ] ; do + + # Get the first arg, and shuffle + option=$1 ; arg2=no + shift + + # Add the argument quoted to args + if [ -z "${args}" ] ; then args="\"${option}\"" + else args="${args} \"${option}\"" ; fi + + # Split out the argument for options that take them + case ${option} in + *=*) optarg=`echo "${option}" | sed -e 's,^[^=]*=,,;s,/$,,'` ;; + esac + + # Process the options + case ${option} in + --help | -h) + echo "Usage: $0 [OPTION]... [VAR=VALUE]..." + echo + echo "To assign makefile variables (e.g., CC, CFLAGS...), specify them as" + echo "arguments to configure in the form VAR=VALUE." + echo + echo "Options and variables: [defaults in brackets]" + echo " -h, --help display this help and exit" + echo " -V, --version output version information and exit" + echo " --srcdir=DIR find the sources in DIR [. or ..]" + echo " --prefix=DIR install into DIR [${prefix}]" + echo " --exec-prefix=DIR base directory for arch-dependent files [${exec_prefix}]" + echo " --bindir=DIR user executables directory [${bindir}]" + echo " --datarootdir=DIR base directory for doc and data [${datarootdir}]" + echo " --includedir=DIR C header files [${includedir}]" + echo " --infodir=DIR info files directory [${infodir}]" + echo " --libdir=DIR object code libraries [${libdir}]" + echo " --mandir=DIR man pages directory [${mandir}]" + echo " --disable-static don't build a static library [enable]" + echo " (implies --enable-shared)" + echo " --enable-shared build also a shared library [disable]" + echo " --disable-ldconfig don't run ldconfig after install" + echo " CC=COMPILER C compiler to use [${CC}]" + echo " AR=ARCHIVER library archiver to use [${AR}]" + echo " CPPFLAGS=OPTIONS command line options for the preprocessor [${CPPFLAGS}]" + echo " CFLAGS=OPTIONS command line options for the C compiler [${CFLAGS}]" + echo " CFLAGS+=OPTIONS append options to the current value of CFLAGS" + echo " LDFLAGS=OPTIONS command line options for the linker [${LDFLAGS}]" + echo " ARFLAGS=OPTIONS command line options for the library archiver [${ARFLAGS}]" + echo + exit 0 ;; + --version | -V) + echo "Configure script for ${pkgname} version ${pkgversion}" + exit 0 ;; + --srcdir) srcdir=$1 ; arg2=yes ;; + --prefix) prefix=$1 ; arg2=yes ;; + --exec-prefix) exec_prefix=$1 ; arg2=yes ;; + --bindir) bindir=$1 ; arg2=yes ;; + --datarootdir) datarootdir=$1 ; arg2=yes ;; + --includedir) includedir=$1 ; arg2=yes ;; + --infodir) infodir=$1 ; arg2=yes ;; + --libdir) libdir=$1 ; arg2=yes ;; + --mandir) mandir=$1 ; arg2=yes ;; + + --srcdir=*) srcdir=${optarg} ;; + --prefix=*) prefix=${optarg} ;; + --exec-prefix=*) exec_prefix=${optarg} ;; + --bindir=*) bindir=${optarg} ;; + --datarootdir=*) datarootdir=${optarg} ;; + --includedir=*) includedir=${optarg} ;; + --infodir=*) infodir=${optarg} ;; + --libdir=*) libdir=${optarg} ;; + --mandir=*) mandir=${optarg} ;; + --no-create) no_create=yes ;; + --disable-static) + progname_static= + progname_shared=${progname}_shared + progname_lzip=${progname}_shared ;; + --enable-shared) + progname_shared=${progname}_shared + progname_lzip=${progname}_shared ;; + --disable-ldconfig) disable_ldconfig=yes ;; + + CC=*) CC=${optarg} ;; + AR=*) AR=${optarg} ;; + CPPFLAGS=*) CPPFLAGS=${optarg} ;; + CFLAGS=*) CFLAGS=${optarg} ;; + CFLAGS+=*) CFLAGS="${CFLAGS} ${optarg}" ;; + LDFLAGS=*) LDFLAGS=${optarg} ;; + ARFLAGS=*) ARFLAGS=${optarg} ;; + + --*) + echo "configure: WARNING: unrecognized option: '${option}'" 1>&2 ;; + *=* | *-*-*) ;; + *) + echo "configure: unrecognized option: '${option}'" 1>&2 + echo "Try 'configure --help' for more information." 1>&2 + exit 1 ;; + esac + + # Check if the option took a separate argument + if [ "${arg2}" = yes ] ; then + if [ $# != 0 ] ; then args="${args} \"$1\"" ; shift + else echo "configure: Missing argument to '${option}'" 1>&2 + exit 1 + fi + fi +done + +# Find the source files, if location was not specified. +srcdirtext= +if [ -z "${srcdir}" ] ; then + srcdirtext="or . or .." ; srcdir=. + if [ ! -r "${srcdir}/${srctrigger}" ] ; then srcdir=.. ; fi + if [ ! -r "${srcdir}/${srctrigger}" ] ; then + ## the sed command below emulates the dirname command + srcdir=`echo "$0" | sed -e 's,[^/]*$,,;s,/$,,;s,^$,.,'` + fi +fi + +if [ ! -r "${srcdir}/${srctrigger}" ] ; then + echo "configure: Can't find sources in ${srcdir} ${srcdirtext}" 1>&2 + echo "configure: (At least ${srctrigger} is missing)." 1>&2 + exit 1 +fi + +# Set srcdir to . if that's what it is. +if [ "`pwd`" = "`cd "${srcdir}" ; pwd`" ] ; then srcdir=. ; fi + +echo +if [ -z "${no_create}" ] ; then + echo "creating config.status" + rm -f config.status + cat > config.status << EOF +#! /bin/sh +# This file was generated automatically by configure. Don't edit. +# Run this file to recreate the current configuration. +# +# This script is free software: you have unlimited permission +# to copy, distribute, and modify it. + +exec /bin/sh $0 ${args} --no-create +EOF + chmod +x config.status +fi + +echo "creating Makefile" +echo "VPATH = ${srcdir}" +echo "prefix = ${prefix}" +echo "exec_prefix = ${exec_prefix}" +echo "bindir = ${bindir}" +echo "datarootdir = ${datarootdir}" +echo "includedir = ${includedir}" +echo "infodir = ${infodir}" +echo "libdir = ${libdir}" +echo "mandir = ${mandir}" +echo "CC = ${CC}" +echo "AR = ${AR}" +echo "CPPFLAGS = ${CPPFLAGS}" +echo "CFLAGS = ${CFLAGS}" +echo "LDFLAGS = ${LDFLAGS}" +echo "ARFLAGS = ${ARFLAGS}" +rm -f Makefile +cat > Makefile << EOF +# Makefile for Lzlib - Compression library for the lzip format +# Copyright (C) 2009-2022 Antonio Diaz Diaz. +# This file was generated automatically by configure. Don't edit. +# +# This Makefile is free software: you have unlimited permission +# to copy, distribute, and modify it. + +pkgname = ${pkgname} +pkgversion = ${pkgversion} +soversion = ${soversion} +progname = ${progname} +progname_static = ${progname_static} +progname_shared = ${progname_shared} +progname_lzip = ${progname_lzip} +disable_ldconfig = ${disable_ldconfig} +libname = ${libname} +VPATH = ${srcdir} +prefix = ${prefix} +exec_prefix = ${exec_prefix} +bindir = ${bindir} +datarootdir = ${datarootdir} +includedir = ${includedir} +infodir = ${infodir} +libdir = ${libdir} +mandir = ${mandir} +CC = ${CC} +AR = ${AR} +CPPFLAGS = ${CPPFLAGS} +CFLAGS = ${CFLAGS} +LDFLAGS = ${LDFLAGS} +ARFLAGS = ${ARFLAGS} +EOF +cat "${srcdir}/Makefile.in" >> Makefile + +echo "OK. Now you can run make." diff --git a/decoder.c b/decoder.c new file mode 100644 index 0000000..16f6532 --- /dev/null +++ b/decoder.c @@ -0,0 +1,145 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +static int LZd_try_verify_trailer( struct LZ_decoder * const d ) + { + Lzip_trailer trailer; + if( Rd_available_bytes( d->rdec ) < Lt_size ) + { if( !d->rdec->at_stream_end ) return 0; else return 2; } + d->verify_trailer_pending = false; + d->member_finished = true; + + if( Rd_read_data( d->rdec, trailer, Lt_size ) == Lt_size && + Lt_get_data_crc( trailer ) == LZd_crc( d ) && + Lt_get_data_size( trailer ) == LZd_data_position( d ) && + Lt_get_member_size( trailer ) == d->rdec->member_position ) return 0; + return 3; + } + + +/* Return value: 0 = OK, 1 = decoder error, 2 = unexpected EOF, + 3 = trailer error, 4 = unknown marker found, + 5 = library error. */ +static int LZd_decode_member( struct LZ_decoder * const d ) + { + struct Range_decoder * const rdec = d->rdec; + State * const state = &d->state; + /* unsigned old_mpos = rdec->member_position; */ + + if( d->member_finished ) return 0; + if( !Rd_try_reload( rdec ) ) + { if( !rdec->at_stream_end ) return 0; else return 2; } + if( d->verify_trailer_pending ) return LZd_try_verify_trailer( d ); + + while( !Rd_finished( rdec ) ) + { + /* const unsigned mpos = rdec->member_position; + if( mpos - old_mpos > rd_min_available_bytes ) return 5; + old_mpos = mpos; */ + if( !Rd_enough_available_bytes( rdec ) ) /* check unexpected EOF */ + { if( !rdec->at_stream_end ) return 0; + if( Cb_empty( &rdec->cb ) ) break; } /* decode until EOF */ + if( !LZd_enough_free_bytes( d ) ) return 0; + const int pos_state = LZd_data_position( d ) & pos_state_mask; + if( Rd_decode_bit( rdec, &d->bm_match[*state][pos_state] ) == 0 ) /* 1st bit */ + { + /* literal byte */ + Bit_model * const bm = d->bm_literal[get_lit_state(LZd_peek_prev( d ))]; + if( ( *state = St_set_char( *state ) ) < 4 ) + LZd_put_byte( d, Rd_decode_tree8( rdec, bm ) ); + else + LZd_put_byte( d, Rd_decode_matched( rdec, bm, LZd_peek( d, d->rep0 ) ) ); + continue; + } + /* match or repeated match */ + int len; + if( Rd_decode_bit( rdec, &d->bm_rep[*state] ) != 0 ) /* 2nd bit */ + { + if( Rd_decode_bit( rdec, &d->bm_rep0[*state] ) == 0 ) /* 3rd bit */ + { + if( Rd_decode_bit( rdec, &d->bm_len[*state][pos_state] ) == 0 ) /* 4th bit */ + { *state = St_set_short_rep( *state ); + LZd_put_byte( d, LZd_peek( d, d->rep0 ) ); continue; } + } + else + { + unsigned distance; + if( Rd_decode_bit( rdec, &d->bm_rep1[*state] ) == 0 ) /* 4th bit */ + distance = d->rep1; + else + { + if( Rd_decode_bit( rdec, &d->bm_rep2[*state] ) == 0 ) /* 5th bit */ + distance = d->rep2; + else + { distance = d->rep3; d->rep3 = d->rep2; } + d->rep2 = d->rep1; + } + d->rep1 = d->rep0; + d->rep0 = distance; + } + *state = St_set_rep( *state ); + len = Rd_decode_len( rdec, &d->rep_len_model, pos_state ); + } + else /* match */ + { + len = Rd_decode_len( rdec, &d->match_len_model, pos_state ); + unsigned distance = Rd_decode_tree6( rdec, d->bm_dis_slot[get_len_state(len)] ); + if( distance >= start_dis_model ) + { + const unsigned dis_slot = distance; + const int direct_bits = ( dis_slot >> 1 ) - 1; + distance = ( 2 | ( dis_slot & 1 ) ) << direct_bits; + if( dis_slot < end_dis_model ) + distance += Rd_decode_tree_reversed( rdec, + d->bm_dis + ( distance - dis_slot ), direct_bits ); + else + { + distance += + Rd_decode( rdec, direct_bits - dis_align_bits ) << dis_align_bits; + distance += Rd_decode_tree_reversed4( rdec, d->bm_align ); + if( distance == 0xFFFFFFFFU ) /* marker found */ + { + Rd_normalize( rdec ); + /* const unsigned mpos = rdec->member_position; + if( mpos - old_mpos > rd_min_available_bytes ) return 5; + old_mpos = mpos; */ + if( len == min_match_len ) /* End Of Stream marker */ + { + d->verify_trailer_pending = true; + return LZd_try_verify_trailer( d ); + } + if( len == min_match_len + 1 ) /* Sync Flush marker */ + { + rdec->reload_pending = true; + if( Rd_try_reload( rdec ) ) continue; + else { if( !rdec->at_stream_end ) return 0; else break; } + } + return 4; + } + } + } + d->rep3 = d->rep2; d->rep2 = d->rep1; d->rep1 = d->rep0; d->rep0 = distance; + *state = St_set_match( *state ); + if( d->rep0 >= d->dictionary_size || + ( d->rep0 >= d->cb.put && !d->pos_wrapped ) ) return 1; + } + LZd_copy_block( d, d->rep0, len ); + } + return 2; + } diff --git a/decoder.h b/decoder.h new file mode 100644 index 0000000..27de9cb --- /dev/null +++ b/decoder.h @@ -0,0 +1,463 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +enum { rd_min_available_bytes = 10 }; + +struct Range_decoder + { + struct Circular_buffer cb; /* input buffer */ + unsigned long long member_position; + uint32_t code; + uint32_t range; + bool at_stream_end; + bool reload_pending; + }; + +static inline bool Rd_init( struct Range_decoder * const rdec ) + { + if( !Cb_init( &rdec->cb, 65536 + rd_min_available_bytes ) ) return false; + rdec->member_position = 0; + rdec->code = 0; + rdec->range = 0xFFFFFFFFU; + rdec->at_stream_end = false; + rdec->reload_pending = false; + return true; + } + +static inline void Rd_free( struct Range_decoder * const rdec ) + { Cb_free( &rdec->cb ); } + +static inline bool Rd_finished( const struct Range_decoder * const rdec ) + { return rdec->at_stream_end && Cb_empty( &rdec->cb ); } + +static inline void Rd_finish( struct Range_decoder * const rdec ) + { rdec->at_stream_end = true; } + +static inline bool Rd_enough_available_bytes( const struct Range_decoder * const rdec ) + { return ( Cb_used_bytes( &rdec->cb ) >= rd_min_available_bytes ); } + +static inline unsigned Rd_available_bytes( const struct Range_decoder * const rdec ) + { return Cb_used_bytes( &rdec->cb ); } + +static inline unsigned Rd_free_bytes( const struct Range_decoder * const rdec ) + { return rdec->at_stream_end ? 0 : Cb_free_bytes( &rdec->cb ); } + +static inline unsigned long long Rd_purge( struct Range_decoder * const rdec ) + { + const unsigned long long size = + rdec->member_position + Cb_used_bytes( &rdec->cb ); + Cb_reset( &rdec->cb ); + rdec->member_position = 0; rdec->at_stream_end = true; + return size; + } + +static inline void Rd_reset( struct Range_decoder * const rdec ) + { Cb_reset( &rdec->cb ); + rdec->member_position = 0; rdec->at_stream_end = false; } + + +/* Seek for a member header and update 'get'. Set '*skippedp' to the number + of bytes skipped. Return true if a valid header is found. +*/ +static bool Rd_find_header( struct Range_decoder * const rdec, + unsigned * const skippedp ) + { + *skippedp = 0; + while( rdec->cb.get != rdec->cb.put ) + { + if( rdec->cb.buffer[rdec->cb.get] == lzip_magic[0] ) + { + unsigned get = rdec->cb.get; + int i; + Lzip_header header; + for( i = 0; i < Lh_size; ++i ) + { + if( get == rdec->cb.put ) return false; /* not enough data */ + header[i] = rdec->cb.buffer[get]; + if( ++get >= rdec->cb.buffer_size ) get = 0; + } + if( Lh_verify( header ) ) return true; + } + if( ++rdec->cb.get >= rdec->cb.buffer_size ) rdec->cb.get = 0; + ++*skippedp; + } + return false; + } + + +static inline int Rd_write_data( struct Range_decoder * const rdec, + const uint8_t * const inbuf, const int size ) + { + if( rdec->at_stream_end || size <= 0 ) return 0; + return Cb_write_data( &rdec->cb, inbuf, size ); + } + +static inline uint8_t Rd_get_byte( struct Range_decoder * const rdec ) + { + /* 0xFF avoids decoder error if member is truncated at EOS marker */ + if( Rd_finished( rdec ) ) return 0xFF; + ++rdec->member_position; + return Cb_get_byte( &rdec->cb ); + } + +static inline int Rd_read_data( struct Range_decoder * const rdec, + uint8_t * const outbuf, const int size ) + { + const int sz = Cb_read_data( &rdec->cb, outbuf, size ); + if( sz > 0 ) rdec->member_position += sz; + return sz; + } + +static inline bool Rd_unread_data( struct Range_decoder * const rdec, + const unsigned size ) + { + if( size > rdec->member_position || !Cb_unread_data( &rdec->cb, size ) ) + return false; + rdec->member_position -= size; + return true; + } + +static bool Rd_try_reload( struct Range_decoder * const rdec ) + { + if( rdec->reload_pending && Rd_available_bytes( rdec ) >= 5 ) + { + int i; + rdec->reload_pending = false; + rdec->code = 0; + for( i = 0; i < 5; ++i ) rdec->code = (rdec->code << 8) | Rd_get_byte( rdec ); + rdec->range = 0xFFFFFFFFU; + rdec->code &= rdec->range; /* make sure that first byte is discarded */ + } + return !rdec->reload_pending; + } + +static inline void Rd_normalize( struct Range_decoder * const rdec ) + { + if( rdec->range <= 0x00FFFFFFU ) + { rdec->range <<= 8; rdec->code = (rdec->code << 8) | Rd_get_byte( rdec ); } + } + +static inline unsigned Rd_decode( struct Range_decoder * const rdec, + const int num_bits ) + { + unsigned symbol = 0; + int i; + for( i = num_bits; i > 0; --i ) + { + Rd_normalize( rdec ); + rdec->range >>= 1; +/* symbol <<= 1; */ +/* if( rdec->code >= rdec->range ) { rdec->code -= rdec->range; symbol |= 1; } */ + const bool bit = ( rdec->code >= rdec->range ); + symbol <<= 1; symbol += bit; + rdec->code -= rdec->range & ( 0U - bit ); + } + return symbol; + } + +static inline unsigned Rd_decode_bit( struct Range_decoder * const rdec, + Bit_model * const probability ) + { + Rd_normalize( rdec ); + const uint32_t bound = ( rdec->range >> bit_model_total_bits ) * *probability; + if( rdec->code < bound ) + { + rdec->range = bound; + *probability += ( bit_model_total - *probability ) >> bit_model_move_bits; + return 0; + } + else + { + rdec->code -= bound; + rdec->range -= bound; + *probability -= *probability >> bit_model_move_bits; + return 1; + } + } + +static inline void Rd_decode_symbol_bit( struct Range_decoder * const rdec, + Bit_model * const probability, unsigned * symbol ) + { + Rd_normalize( rdec ); + *symbol <<= 1; + const uint32_t bound = ( rdec->range >> bit_model_total_bits ) * *probability; + if( rdec->code < bound ) + { + rdec->range = bound; + *probability += ( bit_model_total - *probability ) >> bit_model_move_bits; + } + else + { + rdec->code -= bound; + rdec->range -= bound; + *probability -= *probability >> bit_model_move_bits; + *symbol |= 1; + } + } + +static inline void Rd_decode_symbol_bit_reversed( struct Range_decoder * const rdec, + Bit_model * const probability, unsigned * model, + unsigned * symbol, const int i ) + { + Rd_normalize( rdec ); + *model <<= 1; + const uint32_t bound = ( rdec->range >> bit_model_total_bits ) * *probability; + if( rdec->code < bound ) + { + rdec->range = bound; + *probability += ( bit_model_total - *probability ) >> bit_model_move_bits; + } + else + { + rdec->code -= bound; + rdec->range -= bound; + *probability -= *probability >> bit_model_move_bits; + *model |= 1; + *symbol |= 1 << i; + } + } + +static inline unsigned Rd_decode_tree6( struct Range_decoder * const rdec, + Bit_model bm[] ) + { + unsigned symbol = 1; + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + return symbol & 0x3F; + } + +static inline unsigned Rd_decode_tree8( struct Range_decoder * const rdec, + Bit_model bm[] ) + { + unsigned symbol = 1; + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + return symbol & 0xFF; + } + +static inline unsigned +Rd_decode_tree_reversed( struct Range_decoder * const rdec, + Bit_model bm[], const int num_bits ) + { + unsigned model = 1; + unsigned symbol = 0; + int i; + for( i = 0; i < num_bits; ++i ) + Rd_decode_symbol_bit_reversed( rdec, &bm[model], &model, &symbol, i ); + return symbol; + } + +static inline unsigned +Rd_decode_tree_reversed4( struct Range_decoder * const rdec, Bit_model bm[] ) + { + unsigned model = 1; + unsigned symbol = 0; + Rd_decode_symbol_bit_reversed( rdec, &bm[model], &model, &symbol, 0 ); + Rd_decode_symbol_bit_reversed( rdec, &bm[model], &model, &symbol, 1 ); + Rd_decode_symbol_bit_reversed( rdec, &bm[model], &model, &symbol, 2 ); + Rd_decode_symbol_bit_reversed( rdec, &bm[model], &model, &symbol, 3 ); + return symbol; + } + +static inline unsigned Rd_decode_matched( struct Range_decoder * const rdec, + Bit_model bm[], unsigned match_byte ) + { + unsigned symbol = 1; + unsigned mask = 0x100; + while( true ) + { + const unsigned match_bit = ( match_byte <<= 1 ) & mask; + const unsigned bit = Rd_decode_bit( rdec, &bm[symbol+match_bit+mask] ); + symbol <<= 1; symbol += bit; + if( symbol > 0xFF ) return symbol & 0xFF; + mask &= ~(match_bit ^ (bit << 8)); /* if( match_bit != bit ) mask = 0; */ + } + } + +static inline unsigned Rd_decode_len( struct Range_decoder * const rdec, + struct Len_model * const lm, + const int pos_state ) + { + Bit_model * bm; + unsigned mask, offset, symbol = 1; + + if( Rd_decode_bit( rdec, &lm->choice1 ) == 0 ) + { bm = lm->bm_low[pos_state]; mask = 7; offset = 0; goto len3; } + if( Rd_decode_bit( rdec, &lm->choice2 ) == 0 ) + { bm = lm->bm_mid[pos_state]; mask = 7; offset = len_low_symbols; goto len3; } + bm = lm->bm_high; mask = 0xFF; offset = len_low_symbols + len_mid_symbols; + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); +len3: + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + Rd_decode_symbol_bit( rdec, &bm[symbol], &symbol ); + return ( symbol & mask ) + min_match_len + offset; + } + + +enum { lzd_min_free_bytes = max_match_len }; + +struct LZ_decoder + { + struct Circular_buffer cb; + unsigned long long partial_data_pos; + struct Range_decoder * rdec; + unsigned dictionary_size; + uint32_t crc; + bool member_finished; + bool verify_trailer_pending; + bool pos_wrapped; + unsigned rep0; /* rep[0-3] latest four distances */ + unsigned rep1; /* used for efficient coding of */ + unsigned rep2; /* repeated distances */ + unsigned rep3; + State state; + + Bit_model bm_literal[1<cb ) >= lzd_min_free_bytes; } + +static inline uint8_t LZd_peek_prev( const struct LZ_decoder * const d ) + { return d->cb.buffer[((d->cb.put > 0) ? d->cb.put : d->cb.buffer_size)-1]; } + +static inline uint8_t LZd_peek( const struct LZ_decoder * const d, + const unsigned distance ) + { + const unsigned i = ( ( d->cb.put > distance ) ? 0 : d->cb.buffer_size ) + + d->cb.put - distance - 1; + return d->cb.buffer[i]; + } + +static inline void LZd_put_byte( struct LZ_decoder * const d, const uint8_t b ) + { + CRC32_update_byte( &d->crc, b ); + d->cb.buffer[d->cb.put] = b; + if( ++d->cb.put >= d->cb.buffer_size ) + { d->partial_data_pos += d->cb.put; d->cb.put = 0; d->pos_wrapped = true; } + } + +static inline void LZd_copy_block( struct LZ_decoder * const d, + const unsigned distance, unsigned len ) + { + unsigned lpos = d->cb.put, i = lpos - distance - 1; + bool fast, fast2; + if( lpos > distance ) + { + fast = ( len < d->cb.buffer_size - lpos ); + fast2 = ( fast && len <= lpos - i ); + } + else + { + i += d->cb.buffer_size; + fast = ( len < d->cb.buffer_size - i ); /* (i == pos) may happen */ + fast2 = ( fast && len <= i - lpos ); + } + if( fast ) /* no wrap */ + { + const unsigned tlen = len; + if( fast2 ) /* no wrap, no overlap */ + memcpy( d->cb.buffer + lpos, d->cb.buffer + i, len ); + else + for( ; len > 0; --len ) d->cb.buffer[lpos++] = d->cb.buffer[i++]; + CRC32_update_buf( &d->crc, d->cb.buffer + d->cb.put, tlen ); + d->cb.put += tlen; + } + else for( ; len > 0; --len ) + { + LZd_put_byte( d, d->cb.buffer[i] ); + if( ++i >= d->cb.buffer_size ) i = 0; + } + } + +static inline bool LZd_init( struct LZ_decoder * const d, + struct Range_decoder * const rde, + const unsigned dict_size ) + { + if( !Cb_init( &d->cb, max( 65536, dict_size ) + lzd_min_free_bytes ) ) + return false; + d->partial_data_pos = 0; + d->rdec = rde; + d->dictionary_size = dict_size; + d->crc = 0xFFFFFFFFU; + d->member_finished = false; + d->verify_trailer_pending = false; + d->pos_wrapped = false; + /* prev_byte of first byte; also for LZd_peek( 0 ) on corrupt file */ + d->cb.buffer[d->cb.buffer_size-1] = 0; + d->rep0 = 0; + d->rep1 = 0; + d->rep2 = 0; + d->rep3 = 0; + d->state = 0; + + Bm_array_init( d->bm_literal[0], (1 << literal_context_bits) * 0x300 ); + Bm_array_init( d->bm_match[0], states * pos_states ); + Bm_array_init( d->bm_rep, states ); + Bm_array_init( d->bm_rep0, states ); + Bm_array_init( d->bm_rep1, states ); + Bm_array_init( d->bm_rep2, states ); + Bm_array_init( d->bm_len[0], states * pos_states ); + Bm_array_init( d->bm_dis_slot[0], len_states * (1 << dis_slot_bits) ); + Bm_array_init( d->bm_dis, modeled_distances - end_dis_model + 1 ); + Bm_array_init( d->bm_align, dis_align_size ); + Lm_init( &d->match_len_model ); + Lm_init( &d->rep_len_model ); + return true; + } + +static inline void LZd_free( struct LZ_decoder * const d ) + { Cb_free( &d->cb ); } + +static inline bool LZd_member_finished( const struct LZ_decoder * const d ) + { return ( d->member_finished && Cb_empty( &d->cb ) ); } + +static inline unsigned LZd_crc( const struct LZ_decoder * const d ) + { return d->crc ^ 0xFFFFFFFFU; } + +static inline unsigned long long +LZd_data_position( const struct LZ_decoder * const d ) + { return d->partial_data_pos + d->cb.put; } diff --git a/doc/lzlib.info b/doc/lzlib.info new file mode 100644 index 0000000..d81bc88 --- /dev/null +++ b/doc/lzlib.info @@ -0,0 +1,1323 @@ +This is lzlib.info, produced by makeinfo version 4.13+ from lzlib.texi. + +INFO-DIR-SECTION Compression +START-INFO-DIR-ENTRY +* Lzlib: (lzlib). Compression library for the lzip format +END-INFO-DIR-ENTRY + + +File: lzlib.info, Node: Top, Next: Introduction, Up: (dir) + +Lzlib Manual +************ + +This manual is for Lzlib (version 1.13, 23 January 2022). + +* Menu: + +* Introduction:: Purpose and features of lzlib +* Library version:: Checking library version +* Buffering:: Sizes of lzlib's buffers +* Parameter limits:: Min / max values for some parameters +* Compression functions:: Descriptions of the compression functions +* Decompression functions:: Descriptions of the decompression functions +* Error codes:: Meaning of codes returned by functions +* Error messages:: Error messages corresponding to error codes +* Invoking minilzip:: Command line interface of the test program +* Data format:: Detailed format of the compressed data +* Examples:: A small tutorial with examples +* Problems:: Reporting bugs +* Concept index:: Index of concepts + + + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This manual is free documentation: you have unlimited permission to copy, +distribute, and modify it. + + +File: lzlib.info, Node: Introduction, Next: Library version, Prev: Top, Up: Top + +1 Introduction +************** + +Lzlib is a data compression library providing in-memory LZMA compression and +decompression functions, including integrity checking of the decompressed +data. The compressed data format used by the library is the lzip format. +Lzlib is written in C. + + The lzip file format is designed for data sharing and long-term +archiving, taking into account both data integrity and decoder availability: + + * The lzip format provides very safe integrity checking and some data + recovery means. The program lziprecover can repair bit flip errors + (one of the most common forms of data corruption) in lzip files, and + provides data recovery capabilities, including error-checked merging + of damaged copies of a file. *Note Data safety: (lziprecover)Data + safety. + + * The lzip format is as simple as possible (but not simpler). The lzip + manual provides the source code of a simple decompressor along with a + detailed explanation of how it works, so that with the only help of the + lzip manual it would be possible for a digital archaeologist to extract + the data from a lzip file long after quantum computers eventually + render LZMA obsolete. + + * Additionally the lzip reference implementation is copylefted, which + guarantees that it will remain free forever. + + A nice feature of the lzip format is that a corrupt byte is easier to +repair the nearer it is from the beginning of the file. Therefore, with the +help of lziprecover, losing an entire archive just because of a corrupt +byte near the beginning is a thing of the past. + + The functions and variables forming the interface of the compression +library are declared in the file 'lzlib.h'. Usage examples of the library +are given in the files 'bbexample.c', 'ffexample.c', and 'minilzip.c' from +the source distribution. + + All the library functions are thread safe. The library does not install +any signal handler. The decoder checks the consistency of the compressed +data, so the library should never crash even in case of corrupted input. + + Compression/decompression is done by repeatedly calling a couple of +read/write functions until all the data have been processed by the library. +This interface is safer and less error prone than the traditional zlib +interface. + + Compression/decompression is done when the read function is called. This +means the value returned by the position functions will not be updated until +a read call, even if a lot of data are written. If you want the data to be +compressed in advance, just call the read function with a SIZE equal to 0. + + If all the data to be compressed are written in advance, lzlib will +automatically adjust the header of the compressed data to use the largest +dictionary size that does not exceed neither the data size nor the limit +given to 'LZ_compress_open'. This feature reduces the amount of memory +needed for decompression and allows minilzip to produce identical compressed +output as lzip. + + Lzlib will correctly decompress a data stream which is the concatenation +of two or more compressed data streams. The result is the concatenation of +the corresponding decompressed data streams. Integrity testing of +concatenated compressed data streams is also supported. + + Lzlib is able to compress and decompress streams of unlimited size by +automatically creating multimember output. The members so created are large, +about 2 PiB each. + + In spite of its name (Lempel-Ziv-Markov chain-Algorithm), LZMA is not a +concrete algorithm; it is more like "any algorithm using the LZMA coding +scheme". For example, the option '-0' of lzip uses the scheme in almost the +simplest way possible; issuing the longest match it can find, or a literal +byte if it can't find a match. Inversely, a much more elaborated way of +finding coding sequences of minimum size than the one currently used by lzip +could be developed, and the resulting sequence could also be coded using the +LZMA coding scheme. + + Lzlib currently implements two variants of the LZMA algorithm: fast +(used by option '-0' of minilzip) and normal (used by all other compression +levels). + + The high compression of LZMA comes from combining two basic, well-proven +compression ideas: sliding dictionaries (LZ77/78) and markov models (the +thing used by every compression algorithm that uses a range encoder or +similar order-0 entropy coder as its last stage) with segregation of +contexts according to what the bits are used for. + + The ideas embodied in lzlib are due to (at least) the following people: +Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the +definition of Markov chains), G.N.N. Martin (for the definition of range +encoding), Igor Pavlov (for putting all the above together in LZMA), and +Julian Seward (for bzip2's CLI). + + LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never +have been compressed. Decompressed is used to refer to data which have +undergone the process of decompression. + + +File: lzlib.info, Node: Library version, Next: Buffering, Prev: Introduction, Up: Top + +2 Library version +***************** + +One goal of lzlib is to keep perfect backward compatibility with older +versions of itself down to 1.0. Any application working with an older lzlib +should work with a newer lzlib. Installing a newer lzlib should not break +anything. This chapter describes the constants and functions that the +application can use to discover the version of the library being used. All +of them are declared in 'lzlib.h'. + + -- Constant: LZ_API_VERSION + This constant is defined in 'lzlib.h' and works as a version test + macro. The application should verify at compile time that + LZ_API_VERSION is greater than or equal to the version required by the + application: + + #if !defined LZ_API_VERSION || LZ_API_VERSION < 1012 + #error "lzlib 1.12 or newer needed." + #endif + + Before version 1.8, lzlib didn't define LZ_API_VERSION. + LZ_API_VERSION was first defined in lzlib 1.8 to 1. + Since lzlib 1.12, LZ_API_VERSION is defined as (major * 1000 + minor). + + NOTE: Version test macros are the library's way of announcing +functionality to the application. They should not be confused with feature +test macros, which allow the application to announce to the library its +desire to have certain symbols and prototypes exposed. + + -- Function: int LZ_api_version ( void ) + If LZ_API_VERSION >= 1012, this function is declared in 'lzlib.h' (else + it doesn't exist). It returns the LZ_API_VERSION of the library object + code being used. The application should verify at run time that the + value returned by 'LZ_api_version' is greater than or equal to the + version required by the application. An application may be dinamically + linked at run time with a different version of lzlib than the one it + was compiled for, and this should not break the program as long as the + library used provides the functionality required by the application. + + #if defined LZ_API_VERSION && LZ_API_VERSION >= 1012 + if( LZ_api_version() < 1012 ) + show_error( "lzlib 1.12 or newer needed." ); + #endif + + -- Constant: const char * LZ_version_string + This string constant is defined in the header file 'lzlib.h' and + represents the version of the library being used at compile time. + + -- Function: const char * LZ_version ( void ) + This function returns a string representing the version of the library + being used at run time. + + +File: lzlib.info, Node: Buffering, Next: Parameter limits, Prev: Library version, Up: Top + +3 Buffering +*********** + +Lzlib internal functions need access to a memory chunk at least as large as +the dictionary size (sliding window). For efficiency reasons, the input +buffer for compression is twice or sixteen times as large as the dictionary +size. + + Finally, for safety reasons, lzlib uses two more internal buffers. + + These are the four buffers used by lzlib, and their guaranteed minimum +sizes: + + * Input compression buffer. Written to by the function + 'LZ_compress_write'. For the normal variant of LZMA, its size is two + times the dictionary size set with the function 'LZ_compress_open' or + 64 KiB, whichever is larger. For the fast variant, its size is 1 MiB. + + * Output compression buffer. Read from by the function + 'LZ_compress_read'. Its size is 64 KiB. + + * Input decompression buffer. Written to by the function + 'LZ_decompress_write'. Its size is 64 KiB. + + * Output decompression buffer. Read from by the function + 'LZ_decompress_read'. Its size is the dictionary size set in the header + of the member currently being decompressed or 64 KiB, whichever is + larger. + + +File: lzlib.info, Node: Parameter limits, Next: Compression functions, Prev: Buffering, Up: Top + +4 Parameter limits +****************** + +These functions provide minimum and maximum values for some parameters. +Current values are shown in square brackets. + + -- Function: int LZ_min_dictionary_bits ( void ) + Returns the base 2 logarithm of the smallest valid dictionary size + [12]. + + -- Function: int LZ_min_dictionary_size ( void ) + Returns the smallest valid dictionary size [4 KiB]. + + -- Function: int LZ_max_dictionary_bits ( void ) + Returns the base 2 logarithm of the largest valid dictionary size [29]. + + -- Function: int LZ_max_dictionary_size ( void ) + Returns the largest valid dictionary size [512 MiB]. + + -- Function: int LZ_min_match_len_limit ( void ) + Returns the smallest valid match length limit [5]. + + -- Function: int LZ_max_match_len_limit ( void ) + Returns the largest valid match length limit [273]. + + +File: lzlib.info, Node: Compression functions, Next: Decompression functions, Prev: Parameter limits, Up: Top + +5 Compression functions +*********************** + +These are the functions used to compress data. In case of error, all of +them return -1 or 0, for signed and unsigned return values respectively, +except 'LZ_compress_open' whose return value must be verified by calling +'LZ_compress_errno' before using it. + + -- Function: struct LZ_Encoder * LZ_compress_open ( const int + DICTIONARY_SIZE, const int MATCH_LEN_LIMIT, const unsigned long + long MEMBER_SIZE ) + Initializes the internal stream state for compression and returns a + pointer that can only be used as the ENCODER argument for the other + LZ_compress functions, or a null pointer if the encoder could not be + allocated. + + The returned pointer must be verified by calling 'LZ_compress_errno' + before using it. If 'LZ_compress_errno' does not return 'LZ_ok', the + returned pointer must not be used and should be freed with + 'LZ_compress_close' to avoid memory leaks. + + DICTIONARY_SIZE sets the dictionary size to be used, in bytes. Valid + values range from 4 KiB to 512 MiB. Note that dictionary sizes are + quantized. If the size specified does not match one of the valid + sizes, it will be rounded upwards by adding up to + (DICTIONARY_SIZE / 8) to it. + + MATCH_LEN_LIMIT sets the match length limit in bytes. Valid values + range from 5 to 273. Larger values usually give better compression + ratios but longer compression times. + + If DICTIONARY_SIZE is 65535 and MATCH_LEN_LIMIT is 16, the fast + variant of LZMA is chosen, which produces identical compressed output + as 'lzip -0'. (The dictionary size used will be rounded upwards to + 64 KiB). + + MEMBER_SIZE sets the member size limit in bytes. Valid values range + from 4 KiB to 2 PiB. A small member size may degrade compression + ratio, so use it only when needed. To produce a single-member data + stream, give MEMBER_SIZE a value larger than the amount of data to be + produced. Values larger than 2 PiB will be reduced to 2 PiB to prevent + the uncompressed size of the member from overflowing. + + -- Function: int LZ_compress_close ( struct LZ_Encoder * const ENCODER ) + Frees all dynamically allocated data structures for this stream. This + function discards any unprocessed input and does not flush any pending + output. After a call to 'LZ_compress_close', ENCODER can no longer be + used as an argument to any LZ_compress function. It is safe to call + 'LZ_compress_close' with a null argument. + + -- Function: int LZ_compress_finish ( struct LZ_Encoder * const ENCODER ) + Use this function to tell 'lzlib' that all the data for this member + have already been written (with the function 'LZ_compress_write'). It + is safe to call 'LZ_compress_finish' as many times as needed. After + all the compressed data have been read with 'LZ_compress_read' and + 'LZ_compress_member_finished' returns 1, a new member can be started + with 'LZ_compress_restart_member'. + + -- Function: int LZ_compress_restart_member ( struct LZ_Encoder * const + ENCODER, const unsigned long long MEMBER_SIZE ) + Use this function to start a new member in a multimember data stream. + Call this function only after 'LZ_compress_member_finished' indicates + that the current member has been fully read (with the function + 'LZ_compress_read'). *Note member_size::, for a description of + MEMBER_SIZE. + + -- Function: int LZ_compress_sync_flush ( struct LZ_Encoder * const + ENCODER ) + Use this function to make available to 'LZ_compress_read' all the data + already written with the function 'LZ_compress_write'. First call + 'LZ_compress_sync_flush'. Then call 'LZ_compress_read' until it + returns 0. + + This function writes at least one LZMA marker '3' ("Sync Flush" marker) + to the compressed output. Note that the sync flush marker is not + allowed in lzip files; it is a device for interactive communication + between applications using lzlib, but is useless and wasteful in a + file, and is excluded from the media type 'application/lzip'. The LZMA + marker '2' ("End Of Stream" marker) is the only marker allowed in lzip + files. *Note Data format::. + + Repeated use of 'LZ_compress_sync_flush' may degrade compression + ratio, so use it only when needed. If the interval between calls to + 'LZ_compress_sync_flush' is large (comparable to dictionary size), + creating a multimember data stream with 'LZ_compress_restart_member' + may be an alternative. + + Combining multimember stream creation with flushing may be tricky. If + there are more bytes available than those needed to complete + MEMBER_SIZE, 'LZ_compress_restart_member' needs to be called when + 'LZ_compress_member_finished' returns 1, followed by a new call to + 'LZ_compress_sync_flush'. + + -- Function: int LZ_compress_read ( struct LZ_Encoder * const ENCODER, + uint8_t * const BUFFER, const int SIZE ) + Reads up to SIZE bytes from the stream pointed to by ENCODER, storing + the results in BUFFER. If LZ_API_VERSION >= 1012, BUFFER may be a null + pointer, in which case the bytes read are discarded. + + Returns the number of bytes actually read. This might be less than + SIZE; for example, if there aren't that many bytes left in the stream + or if more bytes have to be yet written with the function + 'LZ_compress_write'. Note that reading less than SIZE bytes is not an + error. + + -- Function: int LZ_compress_write ( struct LZ_Encoder * const ENCODER, + uint8_t * const BUFFER, const int SIZE ) + Writes up to SIZE bytes from BUFFER to the stream pointed to by + ENCODER. Returns the number of bytes actually written. This might be + less than SIZE. Note that writing less than SIZE bytes is not an error. + + -- Function: int LZ_compress_write_size ( struct LZ_Encoder * const + ENCODER ) + Returns the maximum number of bytes that can be immediately written + through 'LZ_compress_write'. For efficiency reasons, once the input + buffer is full and 'LZ_compress_write_size' returns 0, almost all the + buffer must be compressed before a size greater than 0 is returned + again. (This is done to minimize the amount of data that must be + copied to the beginning of the buffer before new data can be accepted). + + It is guaranteed that an immediate call to 'LZ_compress_write' will + accept a SIZE up to the returned number of bytes. + + -- Function: enum LZ_Errno LZ_compress_errno ( struct LZ_Encoder * const + ENCODER ) + Returns the current error code for ENCODER. *Note Error codes::. It is + safe to call 'LZ_compress_errno' with a null argument, in which case + it returns 'LZ_bad_argument'. + + -- Function: int LZ_compress_finished ( struct LZ_Encoder * const ENCODER ) + Returns 1 if all the data have been read and 'LZ_compress_close' can + be safely called. Otherwise it returns 0. 'LZ_compress_finished' + implies 'LZ_compress_member_finished'. + + -- Function: int LZ_compress_member_finished ( struct LZ_Encoder * const + ENCODER ) + Returns 1 if the current member, in a multimember data stream, has been + fully read and 'LZ_compress_restart_member' can be safely called. + Otherwise it returns 0. + + -- Function: unsigned long long LZ_compress_data_position ( struct + LZ_Encoder * const ENCODER ) + Returns the number of input bytes already compressed in the current + member. + + -- Function: unsigned long long LZ_compress_member_position ( struct + LZ_Encoder * const ENCODER ) + Returns the number of compressed bytes already produced, but perhaps + not yet read, in the current member. + + -- Function: unsigned long long LZ_compress_total_in_size ( struct + LZ_Encoder * const ENCODER ) + Returns the total number of input bytes already compressed. + + -- Function: unsigned long long LZ_compress_total_out_size ( struct + LZ_Encoder * const ENCODER ) + Returns the total number of compressed bytes already produced, but + perhaps not yet read. + + +File: lzlib.info, Node: Decompression functions, Next: Error codes, Prev: Compression functions, Up: Top + +6 Decompression functions +************************* + +These are the functions used to decompress data. In case of error, all of +them return -1 or 0, for signed and unsigned return values respectively, +except 'LZ_decompress_open' whose return value must be verified by calling +'LZ_decompress_errno' before using it. + + -- Function: struct LZ_Decoder * LZ_decompress_open ( void ) + Initializes the internal stream state for decompression and returns a + pointer that can only be used as the DECODER argument for the other + LZ_decompress functions, or a null pointer if the decoder could not be + allocated. + + The returned pointer must be verified by calling 'LZ_decompress_errno' + before using it. If 'LZ_decompress_errno' does not return 'LZ_ok', the + returned pointer must not be used and should be freed with + 'LZ_decompress_close' to avoid memory leaks. + + -- Function: int LZ_decompress_close ( struct LZ_Decoder * const DECODER ) + Frees all dynamically allocated data structures for this stream. This + function discards any unprocessed input and does not flush any pending + output. After a call to 'LZ_decompress_close', DECODER can no longer + be used as an argument to any LZ_decompress function. It is safe to + call 'LZ_decompress_close' with a null argument. + + -- Function: int LZ_decompress_finish ( struct LZ_Decoder * const DECODER ) + Use this function to tell 'lzlib' that all the data for this stream + have already been written (with the function 'LZ_decompress_write'). + It is safe to call 'LZ_decompress_finish' as many times as needed. It + is not required to call 'LZ_decompress_finish' if the input stream + only contains whole members, but not calling it prevents lzlib from + detecting a truncated member. + + -- Function: int LZ_decompress_reset ( struct LZ_Decoder * const DECODER ) + Resets the internal state of DECODER as it was just after opening it + with the function 'LZ_decompress_open'. Data stored in the internal + buffers is discarded. Position counters are set to 0. + + -- Function: int LZ_decompress_sync_to_member ( struct LZ_Decoder * const + DECODER ) + Resets the error state of DECODER and enters a search state that lasts + until a new member header (or the end of the stream) is found. After a + successful call to 'LZ_decompress_sync_to_member', data written with + 'LZ_decompress_write' will be consumed and 'LZ_decompress_read' will + return 0 until a header is found. + + This function is useful to discard any data preceding the first member, + or to discard the rest of the current member, for example in case of a + data error. If the decoder is already at the beginning of a member, + this function does nothing. + + -- Function: int LZ_decompress_read ( struct LZ_Decoder * const DECODER, + uint8_t * const BUFFER, const int SIZE ) + Reads up to SIZE bytes from the stream pointed to by DECODER, storing + the results in BUFFER. If LZ_API_VERSION >= 1012, BUFFER may be a null + pointer, in which case the bytes read are discarded. + + Returns the number of bytes actually read. This might be less than + SIZE; for example, if there aren't that many bytes left in the stream + or if more bytes have to be yet written with the function + 'LZ_decompress_write'. Note that reading less than SIZE bytes is not + an error. + + 'LZ_decompress_read' returns at least once per member so that + 'LZ_decompress_member_finished' can be called (and trailer data + retrieved) for each member, even for empty members. Therefore, + 'LZ_decompress_read' returning 0 does not mean that the end of the + stream has been reached. The increase in the value returned by + 'LZ_decompress_total_in_size' can be used to tell the end of the stream + from an empty member. + + In case of decompression error caused by corrupt or truncated data, + 'LZ_decompress_read' does not signal the error immediately to the + application, but waits until all the bytes decoded have been read. This + allows tools like tarlz to recover as much data as possible from each + damaged member. *Note tarlz manual: (tarlz)Top. + + -- Function: int LZ_decompress_write ( struct LZ_Decoder * const DECODER, + uint8_t * const BUFFER, const int SIZE ) + Writes up to SIZE bytes from BUFFER to the stream pointed to by + DECODER. Returns the number of bytes actually written. This might be + less than SIZE. Note that writing less than SIZE bytes is not an error. + + -- Function: int LZ_decompress_write_size ( struct LZ_Decoder * const + DECODER ) + Returns the maximum number of bytes that can be immediately written + through 'LZ_decompress_write'. This number varies smoothly; each + compressed byte consumed may be overwritten immediately, increasing by + 1 the value returned. + + It is guaranteed that an immediate call to 'LZ_decompress_write' will + accept a SIZE up to the returned number of bytes. + + -- Function: enum LZ_Errno LZ_decompress_errno ( struct LZ_Decoder * const + DECODER ) + Returns the current error code for DECODER. *Note Error codes::. It is + safe to call 'LZ_decompress_errno' with a null argument, in which case + it returns 'LZ_bad_argument'. + + -- Function: int LZ_decompress_finished ( struct LZ_Decoder * const + DECODER ) + Returns 1 if all the data have been read and 'LZ_decompress_close' can + be safely called. Otherwise it returns 0. 'LZ_decompress_finished' + does not imply 'LZ_decompress_member_finished'. + + -- Function: int LZ_decompress_member_finished ( struct LZ_Decoder * const + DECODER ) + Returns 1 if the previous call to 'LZ_decompress_read' finished reading + the current member, indicating that final values for the member are + available through 'LZ_decompress_data_crc', + 'LZ_decompress_data_position', and 'LZ_decompress_member_position'. + Otherwise it returns 0. + + -- Function: int LZ_decompress_member_version ( struct LZ_Decoder * const + DECODER ) + Returns the version of the current member, read from the member header. + + -- Function: int LZ_decompress_dictionary_size ( struct LZ_Decoder * const + DECODER ) + Returns the dictionary size of the current member, read from the + member header. + + -- Function: unsigned LZ_decompress_data_crc ( struct LZ_Decoder * const + DECODER ) + Returns the 32 bit Cyclic Redundancy Check of the data decompressed + from the current member. The value returned is valid only when + 'LZ_decompress_member_finished' returns 1. + + -- Function: unsigned long long LZ_decompress_data_position ( struct + LZ_Decoder * const DECODER ) + Returns the number of decompressed bytes already produced, but perhaps + not yet read, in the current member. + + -- Function: unsigned long long LZ_decompress_member_position ( struct + LZ_Decoder * const DECODER ) + Returns the number of input bytes already decompressed in the current + member. + + -- Function: unsigned long long LZ_decompress_total_in_size ( struct + LZ_Decoder * const DECODER ) + Returns the total number of input bytes already decompressed. + + -- Function: unsigned long long LZ_decompress_total_out_size ( struct + LZ_Decoder * const DECODER ) + Returns the total number of decompressed bytes already produced, but + perhaps not yet read. + + +File: lzlib.info, Node: Error codes, Next: Error messages, Prev: Decompression functions, Up: Top + +7 Error codes +************* + +Most library functions return -1 to indicate that they have failed. But +this return value only tells you that an error has occurred. To find out +what kind of error it was, you need to verify the error code by calling +'LZ_(de)compress_errno'. + + Library functions don't change the value returned by +'LZ_(de)compress_errno' when they succeed; thus, the value returned by +'LZ_(de)compress_errno' after a successful call is not necessarily LZ_ok, +and you should not use 'LZ_(de)compress_errno' to determine whether a call +failed. If the call failed, then you can examine 'LZ_(de)compress_errno'. + + The error codes are defined in the header file 'lzlib.h'. + + -- Constant: enum LZ_Errno LZ_ok + The value of this constant is 0 and is used to indicate that there is + no error. + + -- Constant: enum LZ_Errno LZ_bad_argument + At least one of the arguments passed to the library function was + invalid. + + -- Constant: enum LZ_Errno LZ_mem_error + No memory available. The system cannot allocate more virtual memory + because its capacity is full. + + -- Constant: enum LZ_Errno LZ_sequence_error + A library function was called in the wrong order. For example + 'LZ_compress_restart_member' was called before + 'LZ_compress_member_finished' indicates that the current member is + finished. + + -- Constant: enum LZ_Errno LZ_header_error + An invalid member header (one with the wrong magic bytes) was read. If + this happens at the end of the data stream it may indicate trailing + data. + + -- Constant: enum LZ_Errno LZ_unexpected_eof + The end of the data stream was reached in the middle of a member. + + -- Constant: enum LZ_Errno LZ_data_error + The data stream is corrupt. If 'LZ_decompress_member_position' is 6 or + less, it indicates either a format version not supported, an invalid + dictionary size, a corrupt header in a multimember data stream, or + trailing data too similar to a valid lzip header. Lziprecover can be + used to remove conflicting trailing data from a file. + + -- Constant: enum LZ_Errno LZ_library_error + A bug was detected in the library. Please, report it. *Note Problems::. + + +File: lzlib.info, Node: Error messages, Next: Invoking minilzip, Prev: Error codes, Up: Top + +8 Error messages +**************** + + -- Function: const char * LZ_strerror ( const enum LZ_Errno LZ_ERRNO ) + Returns the standard error message for a given error code. The messages + are fairly short; there are no multi-line messages or embedded + newlines. This function makes it easy for your program to report + informative error messages about the failure of a library call. + + The value of LZ_ERRNO normally comes from a call to + 'LZ_(de)compress_errno'. + + +File: lzlib.info, Node: Invoking minilzip, Next: Data format, Prev: Error messages, Up: Top + +9 Invoking minilzip +******************* + +Minilzip is a test program for the compression library lzlib, fully +compatible with lzip 1.4 or newer. + + Lzip is a lossless data compressor with a user interface similar to the +one of gzip or bzip2. Lzip uses a simplified form of the 'Lempel-Ziv-Markov +chain-Algorithm' (LZMA) stream format and provides a 3 factor integrity +checking to maximize interoperability and optimize safety. Lzip can compress +about as fast as gzip (lzip -0) or compress most files more than bzip2 +(lzip -9). Decompression speed is intermediate between gzip and bzip2. Lzip +is better than gzip and bzip2 from a data recovery perspective. Lzip has +been designed, written, and tested with great care to replace gzip and +bzip2 as the standard general-purpose compressed format for unix-like +systems. + +The format for running minilzip is: + + minilzip [OPTIONS] [FILES] + +If no file names are specified, minilzip compresses (or decompresses) from +standard input to standard output. A hyphen '-' used as a FILE argument +means standard input. It can be mixed with other FILES and is read just +once, the first time it appears in the command line. + + minilzip supports the following options: *Note Argument syntax: +(arg_parser)Argument syntax. + +'-h' +'--help' + Print an informative help message describing the options and exit. + +'-V' +'--version' + Print the version number of minilzip on the standard output and exit. + This version number should be included in all bug reports. + +'-a' +'--trailing-error' + Exit with error status 2 if any remaining input is detected after + decompressing the last member. Such remaining input is usually trailing + garbage that can be safely ignored. + +'-b BYTES' +'--member-size=BYTES' + When compressing, set the member size limit to BYTES. It is advisable + to keep members smaller than RAM size so that they can be repaired with + lziprecover in case of corruption. A small member size may degrade + compression ratio, so use it only when needed. Valid values range from + 100 kB to 2 PiB. Defaults to 2 PiB. + +'-c' +'--stdout' + Compress or decompress to standard output; keep input files unchanged. + If compressing several files, each file is compressed independently. + (The output consists of a sequence of independently compressed + members). This option (or '-o') is needed when reading from a named + pipe (fifo) or from a device. Use it also to recover as much of the + decompressed data as possible when decompressing a corrupt file. '-c' + overrides '-o' and '-S'. '-c' has no effect when testing or listing. + +'-d' +'--decompress' + Decompress the files specified. If a file does not exist, can't be + opened, or the destination file already exists and '--force' has not + been specified, minilzip continues decompressing the rest of the files + and exits with error status 1. If a file fails to decompress, or is a + terminal, minilzip exits immediately with error status 2 without + decompressing the rest of the files. A terminal is considered an + uncompressed file, and therefore invalid. + +'-f' +'--force' + Force overwrite of output files. + +'-F' +'--recompress' + When compressing, force re-compression of files whose name already has + the '.lz' or '.tlz' suffix. + +'-k' +'--keep' + Keep (don't delete) input files during compression or decompression. + +'-m BYTES' +'--match-length=BYTES' + When compressing, set the match length limit in bytes. After a match + this long is found, the search is finished. Valid values range from 5 + to 273. Larger values usually give better compression ratios but longer + compression times. + +'-o FILE' +'--output=FILE' + If '-c' has not been also specified, write the (de)compressed output to + FILE; keep input files unchanged. If compressing several files, each + file is compressed independently. (The output consists of a sequence of + independently compressed members). This option (or '-c') is needed when + reading from a named pipe (fifo) or from a device. '-o -' is + equivalent to '-c'. '-o' has no effect when testing or listing. + + When compressing and splitting the output in volumes, FILE is used as + a prefix, and several files named 'FILE00001.lz', 'FILE00002.lz', etc, + are created. In this case, only one input file is allowed. + +'-q' +'--quiet' + Quiet operation. Suppress all messages. + +'-s BYTES' +'--dictionary-size=BYTES' + When compressing, set the dictionary size limit in bytes. Minilzip + will use for each file the largest dictionary size that does not + exceed neither the file size nor this limit. Valid values range from + 4 KiB to 512 MiB. Values 12 to 29 are interpreted as powers of two, + meaning 2^12 to 2^29 bytes. Dictionary sizes are quantized so that + they can be coded in just one byte (*note coded-dict-size::). If the + size specified does not match one of the valid sizes, it will be + rounded upwards by adding up to (BYTES / 8) to it. + + For maximum compression you should use a dictionary size limit as large + as possible, but keep in mind that the decompression memory requirement + is affected at compression time by the choice of dictionary size limit. + +'-S BYTES' +'--volume-size=BYTES' + When compressing, and '-c' has not been also specified, split the + compressed output into several volume files with names + 'original_name00001.lz', 'original_name00002.lz', etc, and set the + volume size limit to BYTES. Input files are kept unchanged. Each + volume is a complete, maybe multimember, lzip file. A small volume + size may degrade compression ratio, so use it only when needed. Valid + values range from 100 kB to 4 EiB. + +'-t' +'--test' + Check integrity of the files specified, but don't decompress them. This + really performs a trial decompression and throws away the result. Use + it together with '-v' to see information about the files. If a file + fails the test, does not exist, can't be opened, or is a terminal, + minilzip continues checking the rest of the files. A final diagnostic + is shown at verbosity level 1 or higher if any file fails the test + when testing multiple files. + +'-v' +'--verbose' + Verbose mode. + When compressing, show the compression ratio and size for each file + processed. + When decompressing or testing, further -v's (up to 4) increase the + verbosity level, showing status, compression ratio, dictionary size, + and trailer contents (CRC, data size, member size). + +'-0 .. -9' + Compression level. Set the compression parameters (dictionary size and + match length limit) as shown in the table below. The default + compression level is '-6', equivalent to '-s8MiB -m36'. Note that '-9' + can be much slower than '-0'. These options have no effect when + decompressing or testing. + + The bidimensional parameter space of LZMA can't be mapped to a linear + scale optimal for all files. If your files are large, very repetitive, + etc, you may need to use the options '--dictionary-size' and + '--match-length' directly to achieve optimal performance. + + If several compression levels or '-s' or '-m' options are given, the + last setting is used. For example '-9 -s64MiB' is equivalent to + '-s64MiB -m273' + + Level Dictionary size (-s) Match length limit (-m) + -0 64 KiB 16 bytes + -1 1 MiB 5 bytes + -2 1.5 MiB 6 bytes + -3 2 MiB 8 bytes + -4 3 MiB 12 bytes + -5 4 MiB 20 bytes + -6 8 MiB 36 bytes + -7 16 MiB 68 bytes + -8 24 MiB 132 bytes + -9 32 MiB 273 bytes + +'--fast' +'--best' + Aliases for GNU gzip compatibility. + +'--loose-trailing' + When decompressing or testing, allow trailing data whose first bytes + are so similar to the magic bytes of a lzip header that they can be + confused with a corrupt header. Use this option if a file triggers a + "corrupt header" error and the cause is not indeed a corrupt header. + +'--check-lib' + Compare the version of lzlib used to compile minilzip with the version + actually being used at run time and exit. Report any differences + found. Exit with error status 1 if differences are found. A mismatch + may indicate that lzlib is not correctly installed or that a different + version of lzlib has been installed after compiling the shared version + of minilzip. Exit with error status 2 if LZ_API_VERSION and + LZ_version_string don't match. 'minilzip -v --check-lib' shows the + version of lzlib being used and the value of LZ_API_VERSION (if + defined). *Note Library version::. + + + Numbers given as arguments to options may be followed by a multiplier +and an optional 'B' for "byte". + + Table of SI and binary prefixes (unit multipliers): + +Prefix Value | Prefix Value +k kilobyte (10^3 = 1000) | Ki kibibyte (2^10 = 1024) +M megabyte (10^6) | Mi mebibyte (2^20) +G gigabyte (10^9) | Gi gibibyte (2^30) +T terabyte (10^12) | Ti tebibyte (2^40) +P petabyte (10^15) | Pi pebibyte (2^50) +E exabyte (10^18) | Ei exbibyte (2^60) +Z zettabyte (10^21) | Zi zebibyte (2^70) +Y yottabyte (10^24) | Yi yobibyte (2^80) + + + Exit status: 0 for a normal exit, 1 for environmental problems (file not +found, invalid flags, I/O errors, etc), 2 to indicate a corrupt or invalid +input file, 3 for an internal consistency error (e.g., bug) which caused +minilzip to panic. + + +File: lzlib.info, Node: Data format, Next: Examples, Prev: Invoking minilzip, Up: Top + +10 Data format +************** + +Perfection is reached, not when there is no longer anything to add, but +when there is no longer anything to take away. +-- Antoine de Saint-Exupery + + + In the diagram below, a box like this: + ++---+ +| | <-- the vertical bars might be missing ++---+ + + represents one byte; a box like this: + ++==============+ +| | ++==============+ + + represents a variable number of bytes. + + + Lzip data consist of a series of independent "members" (compressed data +sets). The members simply appear one after another in the data stream, with +no additional information before, between, or after them. Each member can +encode in compressed form up to 16 EiB - 1 byte of uncompressed data. The +size of a multimember data stream is unlimited. + + Each member has the following structure: + ++--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +| ID string | VN | DS | LZMA stream | CRC32 | Data size | Member size | ++--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + All multibyte values are stored in little endian order. + +'ID string (the "magic" bytes)' + A four byte string, identifying the lzip format, with the value "LZIP" + (0x4C, 0x5A, 0x49, 0x50). + +'VN (version number, 1 byte)' + Just in case something needs to be modified in the future. 1 for now. + +'DS (coded dictionary size, 1 byte)' + The dictionary size is calculated by taking a power of 2 (the base + size) and subtracting from it a fraction between 0/16 and 7/16 of the + base size. + Bits 4-0 contain the base 2 logarithm of the base size (12 to 29). + Bits 7-5 contain the numerator of the fraction (0 to 7) to subtract + from the base size to obtain the dictionary size. + Example: 0xD3 = 2^19 - 6 * 2^15 = 512 KiB - 6 * 32 KiB = 320 KiB + Valid values for dictionary size range from 4 KiB to 512 MiB. + +'LZMA stream' + The LZMA stream, finished by an "End Of Stream" marker. Uses default + values for encoder properties. *Note Stream format: (lzip)Stream + format, for a complete description. + Lzip only uses the LZMA marker '2' ("End Of Stream" marker). Lzlib + also uses the LZMA marker '3' ("Sync Flush" marker). *Note + sync_flush::. + +'CRC32 (4 bytes)' + Cyclic Redundancy Check (CRC) of the original uncompressed data. + +'Data size (8 bytes)' + Size of the original uncompressed data. + +'Member size (8 bytes)' + Total size of the member, including header and trailer. This field acts + as a distributed index, allows the verification of stream integrity, + and facilitates the safe recovery of undamaged members from + multimember files. Member size should be limited to 2 PiB to prevent + the data size field from overflowing. + + + +File: lzlib.info, Node: Examples, Next: Problems, Prev: Data format, Up: Top + +11 A small tutorial with examples +********************************* + +This chapter provides real code examples for the most common uses of the +library. See these examples in context in the files 'bbexample.c' and +'ffexample.c' from the source distribution of lzlib. + + Note that the interface of lzlib is symmetrical. That is, the code for +normal compression and decompression is identical except because one calls +LZ_compress* functions while the other calls LZ_decompress* functions. + +* Menu: + +* Buffer compression:: Buffer-to-buffer single-member compression +* Buffer decompression:: Buffer-to-buffer decompression +* File compression:: File-to-file single-member compression +* File decompression:: File-to-file decompression +* File compression mm:: File-to-file multimember compression +* Skipping data errors:: Decompression with automatic resynchronization + + +File: lzlib.info, Node: Buffer compression, Next: Buffer decompression, Up: Examples + +11.1 Buffer compression +======================= + +Buffer-to-buffer single-member compression (MEMBER_SIZE > total output). + +/* Compress 'insize' bytes from 'inbuf' to 'outbuf'. + Return the size of the compressed data in '*outlenp'. + In case of error, or if 'outsize' is too small, return false and do not + modify '*outlenp'. +*/ +bool bbcompress( const uint8_t * const inbuf, const int insize, + const int dictionary_size, const int match_len_limit, + uint8_t * const outbuf, const int outsize, + int * const outlenp ) + { + int inpos = 0, outpos = 0; + bool error = false; + struct LZ_Encoder * const encoder = + LZ_compress_open( dictionary_size, match_len_limit, INT64_MAX ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { LZ_compress_close( encoder ); return false; } + + while( true ) + { + int ret = LZ_compress_write( encoder, inbuf + inpos, insize - inpos ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_compress_finish( encoder ); + ret = LZ_compress_read( encoder, outbuf + outpos, outsize - outpos ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_compress_finished( encoder ) == 1 ) break; + if( outpos >= outsize ) { error = true; break; } + } + + if( LZ_compress_close( encoder ) < 0 ) error = true; + if( error ) return false; + *outlenp = outpos; + return true; + } + + +File: lzlib.info, Node: Buffer decompression, Next: File compression, Prev: Buffer compression, Up: Examples + +11.2 Buffer decompression +========================= + +Buffer-to-buffer decompression. + +/* Decompress 'insize' bytes from 'inbuf' to 'outbuf'. + Return the size of the decompressed data in '*outlenp'. + In case of error, or if 'outsize' is too small, return false and do not + modify '*outlenp'. +*/ +bool bbdecompress( const uint8_t * const inbuf, const int insize, + uint8_t * const outbuf, const int outsize, + int * const outlenp ) + { + int inpos = 0, outpos = 0; + bool error = false; + struct LZ_Decoder * const decoder = LZ_decompress_open(); + if( !decoder || LZ_decompress_errno( decoder ) != LZ_ok ) + { LZ_decompress_close( decoder ); return false; } + + while( true ) + { + int ret = LZ_decompress_write( decoder, inbuf + inpos, insize - inpos ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_decompress_finish( decoder ); + ret = LZ_decompress_read( decoder, outbuf + outpos, outsize - outpos ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_decompress_finished( decoder ) == 1 ) break; + if( outpos >= outsize ) { error = true; break; } + } + + if( LZ_decompress_close( decoder ) < 0 ) error = true; + if( error ) return false; + *outlenp = outpos; + return true; + } + + +File: lzlib.info, Node: File compression, Next: File decompression, Prev: Buffer decompression, Up: Examples + +11.3 File compression +===================== + +File-to-file compression using LZ_compress_write_size. + +int ffcompress( struct LZ_Encoder * const encoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_finished( encoder ) == 1 ) return 0; + } + return 1; + } + + +File: lzlib.info, Node: File decompression, Next: File compression mm, Prev: File compression, Up: Examples + +11.4 File decompression +======================= + +File-to-file decompression using LZ_decompress_write_size. + +int ffdecompress( struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_decompress_write( decoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } + + +File: lzlib.info, Node: File compression mm, Next: Skipping data errors, Prev: File decompression, Up: Examples + +11.5 File-to-file multimember compression +========================================= + +Example 1: Multimember compression with members of fixed size +(MEMBER_SIZE < total output). + +int ffmmcompress( FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384, member_size = 4096 }; + uint8_t buffer[buffer_size]; + bool done = false; + struct LZ_Encoder * const encoder = + LZ_compress_open( 65535, 16, member_size ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { fputs( "ffexample: Not enough memory.\n", stderr ); + LZ_compress_close( encoder ); return 1; } + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_member_finished( encoder ) == 1 ) + { + if( LZ_compress_finished( encoder ) == 1 ) { done = true; break; } + if( LZ_compress_restart_member( encoder, member_size ) < 0 ) break; + } + } + if( LZ_compress_close( encoder ) < 0 ) done = false; + return done; + } + + +Example 2: Multimember compression (user-restarted members). (Call +LZ_compress_open with MEMBER_SIZE > largest member). + +/* Compress 'infile' to 'outfile' as a multimember stream with one member + for each line of text terminated by a newline character or by EOF. + Return 0 if success, 1 if error. +*/ +int fflfcompress( struct LZ_Encoder * const encoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + for( len = 0; len < size; ) + { + int ch = getc( infile ); + if( ch == EOF || ( buffer[len++] = ch ) == '\n' ) break; + } + /* avoid writing an empty member to outfile */ + if( len == 0 && LZ_compress_data_position( encoder ) == 0 ) return 0; + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) || buffer[len-1] == '\n' ) + LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_member_finished( encoder ) == 1 ) + { + if( feof( infile ) && LZ_compress_finished( encoder ) == 1 ) return 0; + if( LZ_compress_restart_member( encoder, INT64_MAX ) < 0 ) break; + } + } + return 1; + } + + +File: lzlib.info, Node: Skipping data errors, Prev: File compression mm, Up: Examples + +11.6 Skipping data errors +========================= + +/* Decompress 'infile' to 'outfile' with automatic resynchronization to + next member in case of data error, including the automatic removal of + leading garbage. +*/ +int ffrsdecompress( struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_decompress_write( decoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) + { + if( LZ_decompress_errno( decoder ) == LZ_header_error || + LZ_decompress_errno( decoder ) == LZ_data_error ) + { LZ_decompress_sync_to_member( decoder ); continue; } + else break; + } + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } + + +File: lzlib.info, Node: Problems, Next: Concept index, Prev: Examples, Up: Top + +12 Reporting bugs +***************** + +There are probably bugs in lzlib. There are certainly errors and omissions +in this manual. If you report them, they will get fixed. If you don't, no +one will ever know about them and they will remain unfixed for all +eternity, if not longer. + + If you find a bug in lzlib, please send electronic mail to +. Include the version number, which you can find by +running 'minilzip --version' and 'minilzip -v --check-lib'. + + +File: lzlib.info, Node: Concept index, Prev: Problems, Up: Top + +Concept index +************* + +[index] +* Menu: + +* buffer compression: Buffer compression. (line 6) +* buffer decompression: Buffer decompression. (line 6) +* buffering: Buffering. (line 6) +* bugs: Problems. (line 6) +* compression functions: Compression functions. (line 6) +* data format: Data format. (line 6) +* decompression functions: Decompression functions. (line 6) +* error codes: Error codes. (line 6) +* error messages: Error messages. (line 6) +* examples: Examples. (line 6) +* file compression: File compression. (line 6) +* file decompression: File decompression. (line 6) +* getting help: Problems. (line 6) +* introduction: Introduction. (line 6) +* invoking: Invoking minilzip. (line 6) +* library version: Library version. (line 6) +* multimember compression: File compression mm. (line 6) +* options: Invoking minilzip. (line 6) +* parameter limits: Parameter limits. (line 6) +* skipping data errors: Skipping data errors. (line 6) + + + +Tag Table: +Node: Top215 +Node: Introduction1338 +Node: Library version6413 +Node: Buffering8957 +Node: Parameter limits10182 +Node: Compression functions11136 +Ref: member_size12946 +Ref: sync_flush14712 +Node: Decompression functions19400 +Node: Error codes26968 +Node: Error messages29259 +Node: Invoking minilzip29838 +Node: Data format39786 +Ref: coded-dict-size41232 +Node: Examples42641 +Node: Buffer compression43602 +Node: Buffer decompression45122 +Node: File compression46536 +Node: File decompression47519 +Node: File compression mm48523 +Node: Skipping data errors51552 +Node: Problems52862 +Node: Concept index53423 + +End Tag Table + + +Local Variables: +coding: iso-8859-15 +End: diff --git a/doc/lzlib.texi b/doc/lzlib.texi new file mode 100644 index 0000000..3caf9dd --- /dev/null +++ b/doc/lzlib.texi @@ -0,0 +1,1395 @@ +\input texinfo @c -*-texinfo-*- +@c %**start of header +@setfilename lzlib.info +@documentencoding ISO-8859-15 +@settitle Lzlib Manual +@finalout +@c %**end of header + +@set UPDATED 23 January 2022 +@set VERSION 1.13 + +@dircategory Compression +@direntry +* Lzlib: (lzlib). Compression library for the lzip format +@end direntry + + +@ifnothtml +@titlepage +@title Lzlib +@subtitle Compression library for the lzip format +@subtitle for Lzlib version @value{VERSION}, @value{UPDATED} +@author by Antonio Diaz Diaz + +@page +@vskip 0pt plus 1filll +@end titlepage + +@contents +@end ifnothtml + +@ifnottex +@node Top +@top + +This manual is for Lzlib (version @value{VERSION}, @value{UPDATED}). + +@menu +* Introduction:: Purpose and features of lzlib +* Library version:: Checking library version +* Buffering:: Sizes of lzlib's buffers +* Parameter limits:: Min / max values for some parameters +* Compression functions:: Descriptions of the compression functions +* Decompression functions:: Descriptions of the decompression functions +* Error codes:: Meaning of codes returned by functions +* Error messages:: Error messages corresponding to error codes +* Invoking minilzip:: Command line interface of the test program +* Data format:: Detailed format of the compressed data +* Examples:: A small tutorial with examples +* Problems:: Reporting bugs +* Concept index:: Index of concepts +@end menu + +@sp 1 +Copyright @copyright{} 2009-2022 Antonio Diaz Diaz. + +This manual is free documentation: you have unlimited permission to copy, +distribute, and modify it. +@end ifnottex + + +@node Introduction +@chapter Introduction +@cindex introduction + +@uref{http://www.nongnu.org/lzip/lzlib.html,,Lzlib} +is a data compression library providing in-memory LZMA compression and +decompression functions, including integrity checking of the decompressed +data. The compressed data format used by the library is the lzip format. +Lzlib is written in C. + +The lzip file format is designed for data sharing and long-term archiving, +taking into account both data integrity and decoder availability: + +@itemize @bullet +@item +The lzip format provides very safe integrity checking and some data +recovery means. The program +@uref{http://www.nongnu.org/lzip/manual/lziprecover_manual.html#Data-safety,,lziprecover} +can repair bit flip errors (one of the most common forms of data corruption) +in lzip files, and provides data recovery capabilities, including +error-checked merging of damaged copies of a file. +@ifnothtml +@xref{Data safety,,,lziprecover}. +@end ifnothtml + +@item +The lzip format is as simple as possible (but not simpler). The lzip +manual provides the source code of a simple decompressor along with a +detailed explanation of how it works, so that with the only help of the +lzip manual it would be possible for a digital archaeologist to extract +the data from a lzip file long after quantum computers eventually +render LZMA obsolete. + +@item +Additionally the lzip reference implementation is copylefted, which +guarantees that it will remain free forever. +@end itemize + +A nice feature of the lzip format is that a corrupt byte is easier to repair +the nearer it is from the beginning of the file. Therefore, with the help of +lziprecover, losing an entire archive just because of a corrupt byte near +the beginning is a thing of the past. + +The functions and variables forming the interface of the compression library +are declared in the file @samp{lzlib.h}. Usage examples of the library are +given in the files @samp{bbexample.c}, @samp{ffexample.c}, and +@samp{minilzip.c} from the source distribution. + +All the library functions are thread safe. The library does not install any +signal handler. The decoder checks the consistency of the compressed data, +so the library should never crash even in case of corrupted input. + +Compression/decompression is done by repeatedly calling a couple of +read/write functions until all the data have been processed by the library. +This interface is safer and less error prone than the traditional zlib +interface. + +Compression/decompression is done when the read function is called. This +means the value returned by the position functions will not be updated until +a read call, even if a lot of data are written. If you want the data to be +compressed in advance, just call the read function with a @var{size} equal +to 0. + +If all the data to be compressed are written in advance, lzlib will +automatically adjust the header of the compressed data to use the largest +dictionary size that does not exceed neither the data size nor the limit +given to @samp{LZ_compress_open}. This feature reduces the amount of memory +needed for decompression and allows minilzip to produce identical compressed +output as lzip. + +Lzlib will correctly decompress a data stream which is the concatenation of +two or more compressed data streams. The result is the concatenation of the +corresponding decompressed data streams. Integrity testing of concatenated +compressed data streams is also supported. + +Lzlib is able to compress and decompress streams of unlimited size by +automatically creating multimember output. The members so created are large, +about @w{2 PiB} each. + +In spite of its name (Lempel-Ziv-Markov chain-Algorithm), LZMA is not a +concrete algorithm; it is more like "any algorithm using the LZMA coding +scheme". For example, the option @samp{-0} of lzip uses the scheme in almost the +simplest way possible; issuing the longest match it can find, or a literal +byte if it can't find a match. Inversely, a much more elaborated way of +finding coding sequences of minimum size than the one currently used by lzip +could be developed, and the resulting sequence could also be coded using the +LZMA coding scheme. + +Lzlib currently implements two variants of the LZMA algorithm: fast (used by +option @samp{-0} of minilzip) and normal (used by all other compression levels). + +The high compression of LZMA comes from combining two basic, well-proven +compression ideas: sliding dictionaries (LZ77/78) and markov models (the +thing used by every compression algorithm that uses a range encoder or +similar order-0 entropy coder as its last stage) with segregation of +contexts according to what the bits are used for. + +The ideas embodied in lzlib are due to (at least) the following people: +Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the +definition of Markov chains), G.N.N. Martin (for the definition of range +encoding), Igor Pavlov (for putting all the above together in LZMA), and +Julian Seward (for bzip2's CLI). + +LANGUAGE NOTE: Uncompressed = not compressed = plain data; it may never have +been compressed. Decompressed is used to refer to data which have undergone +the process of decompression. + + +@node Library version +@chapter Library version +@cindex library version + +One goal of lzlib is to keep perfect backward compatibility with older +versions of itself down to 1.0. Any application working with an older lzlib +should work with a newer lzlib. Installing a newer lzlib should not break +anything. This chapter describes the constants and functions that the +application can use to discover the version of the library being used. All +of them are declared in @samp{lzlib.h}. + +@defvr Constant LZ_API_VERSION +This constant is defined in @samp{lzlib.h} and works as a version test +macro. The application should verify at compile time that LZ_API_VERSION is +greater than or equal to the version required by the application: + +@example +#if !defined LZ_API_VERSION || LZ_API_VERSION < 1012 +#error "lzlib 1.12 or newer needed." +#endif +@end example + +Before version 1.8, lzlib didn't define LZ_API_VERSION.@* +LZ_API_VERSION was first defined in lzlib 1.8 to 1.@* +Since lzlib 1.12, LZ_API_VERSION is defined as (major * 1000 + minor). +@end defvr + +NOTE: Version test macros are the library's way of announcing functionality +to the application. They should not be confused with feature test macros, +which allow the application to announce to the library its desire to have +certain symbols and prototypes exposed. + +@deftypefun int LZ_api_version ( void ) +If LZ_API_VERSION >= 1012, this function is declared in @samp{lzlib.h} (else +it doesn't exist). It returns the LZ_API_VERSION of the library object code +being used. The application should verify at run time that the value +returned by @code{LZ_api_version} is greater than or equal to the version +required by the application. An application may be dinamically linked at run +time with a different version of lzlib than the one it was compiled for, and +this should not break the program as long as the library used provides the +functionality required by the application. + +@example +#if defined LZ_API_VERSION && LZ_API_VERSION >= 1012 + if( LZ_api_version() < 1012 ) + show_error( "lzlib 1.12 or newer needed." ); +#endif +@end example +@end deftypefun + +@deftypevr Constant {const char *} LZ_version_string +This string constant is defined in the header file @samp{lzlib.h} and +represents the version of the library being used at compile time. +@end deftypevr + +@deftypefun {const char *} LZ_version ( void ) +This function returns a string representing the version of the library being +used at run time. +@end deftypefun + + +@node Buffering +@chapter Buffering +@cindex buffering + +Lzlib internal functions need access to a memory chunk at least as large +as the dictionary size (sliding window). For efficiency reasons, the +input buffer for compression is twice or sixteen times as large as the +dictionary size. + +Finally, for safety reasons, lzlib uses two more internal buffers. + +These are the four buffers used by lzlib, and their guaranteed minimum sizes: + +@itemize @bullet +@item Input compression buffer. Written to by the function +@samp{LZ_compress_write}. For the normal variant of LZMA, its size is two +times the dictionary size set with the function @samp{LZ_compress_open} or +@w{64 KiB}, whichever is larger. For the fast variant, its size is @w{1 MiB}. + +@item Output compression buffer. Read from by the function +@samp{LZ_compress_read}. Its size is @w{64 KiB}. + +@item Input decompression buffer. Written to by the function +@samp{LZ_decompress_write}. Its size is @w{64 KiB}. + +@item Output decompression buffer. Read from by the function +@samp{LZ_decompress_read}. Its size is the dictionary size set in the header +of the member currently being decompressed or @w{64 KiB}, whichever is larger. +@end itemize + + +@node Parameter limits +@chapter Parameter limits +@cindex parameter limits + +These functions provide minimum and maximum values for some parameters. +Current values are shown in square brackets. + +@deftypefun int LZ_min_dictionary_bits ( void ) +Returns the base 2 logarithm of the smallest valid dictionary size [12]. +@end deftypefun + +@deftypefun int LZ_min_dictionary_size ( void ) +Returns the smallest valid dictionary size [4 KiB]. +@end deftypefun + +@deftypefun int LZ_max_dictionary_bits ( void ) +Returns the base 2 logarithm of the largest valid dictionary size [29]. +@end deftypefun + +@deftypefun int LZ_max_dictionary_size ( void ) +Returns the largest valid dictionary size [512 MiB]. +@end deftypefun + +@deftypefun int LZ_min_match_len_limit ( void ) +Returns the smallest valid match length limit [5]. +@end deftypefun + +@deftypefun int LZ_max_match_len_limit ( void ) +Returns the largest valid match length limit [273]. +@end deftypefun + + +@node Compression functions +@chapter Compression functions +@cindex compression functions + +These are the functions used to compress data. In case of error, all of +them return -1 or 0, for signed and unsigned return values respectively, +except @samp{LZ_compress_open} whose return value must be verified by +calling @samp{LZ_compress_errno} before using it. + + +@deftypefun {struct LZ_Encoder *} LZ_compress_open ( const int @var{dictionary_size}, const int @var{match_len_limit}, const unsigned long long @var{member_size} ) +Initializes the internal stream state for compression and returns a +pointer that can only be used as the @var{encoder} argument for the +other LZ_compress functions, or a null pointer if the encoder could not +be allocated. + +The returned pointer must be verified by calling +@samp{LZ_compress_errno} before using it. If @samp{LZ_compress_errno} +does not return @samp{LZ_ok}, the returned pointer must not be used and +should be freed with @samp{LZ_compress_close} to avoid memory leaks. + +@var{dictionary_size} sets the dictionary size to be used, in bytes. +Valid values range from @w{4 KiB} to @w{512 MiB}. Note that dictionary +sizes are quantized. If the size specified does not match one of the +valid sizes, it will be rounded upwards by adding up to +@w{(@var{dictionary_size} / 8)} to it. + +@var{match_len_limit} sets the match length limit in bytes. Valid values +range from 5 to 273. Larger values usually give better compression ratios +but longer compression times. + +If @var{dictionary_size} is 65535 and @var{match_len_limit} is 16, the fast +variant of LZMA is chosen, which produces identical compressed output as +@w{@samp{lzip -0}}. (The dictionary size used will be rounded upwards to +@w{64 KiB}). + +@anchor{member_size} +@var{member_size} sets the member size limit in bytes. Valid values range +from @w{4 KiB} to @w{2 PiB}. A small member size may degrade compression +ratio, so use it only when needed. To produce a single-member data stream, +give @var{member_size} a value larger than the amount of data to be +produced. Values larger than @w{2 PiB} will be reduced to @w{2 PiB} to +prevent the uncompressed size of the member from overflowing. +@end deftypefun + + +@deftypefun int LZ_compress_close ( struct LZ_Encoder * const @var{encoder} ) +Frees all dynamically allocated data structures for this stream. This +function discards any unprocessed input and does not flush any pending +output. After a call to @samp{LZ_compress_close}, @var{encoder} can no +longer be used as an argument to any LZ_compress function. +It is safe to call @samp{LZ_compress_close} with a null argument. +@end deftypefun + + +@deftypefun int LZ_compress_finish ( struct LZ_Encoder * const @var{encoder} ) +Use this function to tell @samp{lzlib} that all the data for this member +have already been written (with the function @samp{LZ_compress_write}). +It is safe to call @samp{LZ_compress_finish} as many times as needed. +After all the compressed data have been read with @samp{LZ_compress_read} +and @samp{LZ_compress_member_finished} returns 1, a new member can be +started with @samp{LZ_compress_restart_member}. +@end deftypefun + + +@deftypefun int LZ_compress_restart_member ( struct LZ_Encoder * const @var{encoder}, const unsigned long long @var{member_size} ) +Use this function to start a new member in a multimember data stream. Call +this function only after @samp{LZ_compress_member_finished} indicates that +the current member has been fully read (with the function +@samp{LZ_compress_read}). @xref{member_size}, for a description of +@var{member_size}. +@end deftypefun + + +@anchor{sync_flush} +@deftypefun int LZ_compress_sync_flush ( struct LZ_Encoder * const @var{encoder} ) +Use this function to make available to @samp{LZ_compress_read} all the data +already written with the function @samp{LZ_compress_write}. First call +@samp{LZ_compress_sync_flush}. Then call @samp{LZ_compress_read} until it +returns 0. + +This function writes at least one LZMA marker @samp{3} ("Sync Flush" marker) +to the compressed output. Note that the sync flush marker is not allowed in +lzip files; it is a device for interactive communication between +applications using lzlib, but is useless and wasteful in a file, and is +excluded from the media type @samp{application/lzip}. The LZMA marker +@samp{2} ("End Of Stream" marker) is the only marker allowed in lzip files. +@xref{Data format}. + +Repeated use of @samp{LZ_compress_sync_flush} may degrade compression +ratio, so use it only when needed. If the interval between calls to +@samp{LZ_compress_sync_flush} is large (comparable to dictionary size), +creating a multimember data stream with @samp{LZ_compress_restart_member} +may be an alternative. + +Combining multimember stream creation with flushing may be tricky. If there +are more bytes available than those needed to complete @var{member_size}, +@samp{LZ_compress_restart_member} needs to be called when +@samp{LZ_compress_member_finished} returns 1, followed by a new call to +@samp{LZ_compress_sync_flush}. +@end deftypefun + + +@deftypefun int LZ_compress_read ( struct LZ_Encoder * const @var{encoder}, uint8_t * const @var{buffer}, const int @var{size} ) +Reads up to @var{size} bytes from the stream pointed to by @var{encoder}, +storing the results in @var{buffer}. If @w{LZ_API_VERSION >= 1012}, +@var{buffer} may be a null pointer, in which case the bytes read are +discarded. + +Returns the number of bytes actually read. This might be less than +@var{size}; for example, if there aren't that many bytes left in the stream +or if more bytes have to be yet written with the function +@samp{LZ_compress_write}. Note that reading less than @var{size} bytes is +not an error. +@end deftypefun + + +@deftypefun int LZ_compress_write ( struct LZ_Encoder * const @var{encoder}, uint8_t * const @var{buffer}, const int @var{size} ) +Writes up to @var{size} bytes from @var{buffer} to the stream pointed to by +@var{encoder}. Returns the number of bytes actually written. This might be +less than @var{size}. Note that writing less than @var{size} bytes is not an +error. +@end deftypefun + + +@deftypefun int LZ_compress_write_size ( struct LZ_Encoder * const @var{encoder} ) +Returns the maximum number of bytes that can be immediately written through +@samp{LZ_compress_write}. For efficiency reasons, once the input buffer is +full and @samp{LZ_compress_write_size} returns 0, almost all the buffer must +be compressed before a size greater than 0 is returned again. (This is done +to minimize the amount of data that must be copied to the beginning of the +buffer before new data can be accepted). + +It is guaranteed that an immediate call to @samp{LZ_compress_write} will +accept a @var{size} up to the returned number of bytes. +@end deftypefun + + +@deftypefun {enum LZ_Errno} LZ_compress_errno ( struct LZ_Encoder * const @var{encoder} ) +Returns the current error code for @var{encoder}. @xref{Error codes}. +It is safe to call @samp{LZ_compress_errno} with a null argument, in which +case it returns @samp{LZ_bad_argument}. +@end deftypefun + + +@deftypefun int LZ_compress_finished ( struct LZ_Encoder * const @var{encoder} ) +Returns 1 if all the data have been read and @samp{LZ_compress_close} +can be safely called. Otherwise it returns 0. @samp{LZ_compress_finished} +implies @samp{LZ_compress_member_finished}. +@end deftypefun + + +@deftypefun int LZ_compress_member_finished ( struct LZ_Encoder * const @var{encoder} ) +Returns 1 if the current member, in a multimember data stream, has been +fully read and @samp{LZ_compress_restart_member} can be safely called. +Otherwise it returns 0. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_compress_data_position ( struct LZ_Encoder * const @var{encoder} ) +Returns the number of input bytes already compressed in the current member. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_compress_member_position ( struct LZ_Encoder * const @var{encoder} ) +Returns the number of compressed bytes already produced, but perhaps not +yet read, in the current member. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_compress_total_in_size ( struct LZ_Encoder * const @var{encoder} ) +Returns the total number of input bytes already compressed. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_compress_total_out_size ( struct LZ_Encoder * const @var{encoder} ) +Returns the total number of compressed bytes already produced, but +perhaps not yet read. +@end deftypefun + + +@node Decompression functions +@chapter Decompression functions +@cindex decompression functions + +These are the functions used to decompress data. In case of error, all of +them return -1 or 0, for signed and unsigned return values respectively, +except @samp{LZ_decompress_open} whose return value must be verified by +calling @samp{LZ_decompress_errno} before using it. + + +@deftypefun {struct LZ_Decoder *} LZ_decompress_open ( void ) +Initializes the internal stream state for decompression and returns a +pointer that can only be used as the @var{decoder} argument for the +other LZ_decompress functions, or a null pointer if the decoder could +not be allocated. + +The returned pointer must be verified by calling +@samp{LZ_decompress_errno} before using it. If +@samp{LZ_decompress_errno} does not return @samp{LZ_ok}, the returned +pointer must not be used and should be freed with +@samp{LZ_decompress_close} to avoid memory leaks. +@end deftypefun + + +@deftypefun int LZ_decompress_close ( struct LZ_Decoder * const @var{decoder} ) +Frees all dynamically allocated data structures for this stream. This +function discards any unprocessed input and does not flush any pending +output. After a call to @samp{LZ_decompress_close}, @var{decoder} can no +longer be used as an argument to any LZ_decompress function. +It is safe to call @samp{LZ_decompress_close} with a null argument. +@end deftypefun + + +@deftypefun int LZ_decompress_finish ( struct LZ_Decoder * const @var{decoder} ) +Use this function to tell @samp{lzlib} that all the data for this stream +have already been written (with the function @samp{LZ_decompress_write}). +It is safe to call @samp{LZ_decompress_finish} as many times as needed. +It is not required to call @samp{LZ_decompress_finish} if the input stream +only contains whole members, but not calling it prevents lzlib from +detecting a truncated member. +@end deftypefun + + +@deftypefun int LZ_decompress_reset ( struct LZ_Decoder * const @var{decoder} ) +Resets the internal state of @var{decoder} as it was just after opening +it with the function @samp{LZ_decompress_open}. Data stored in the +internal buffers is discarded. Position counters are set to 0. +@end deftypefun + + +@deftypefun int LZ_decompress_sync_to_member ( struct LZ_Decoder * const @var{decoder} ) +Resets the error state of @var{decoder} and enters a search state that +lasts until a new member header (or the end of the stream) is found. +After a successful call to @samp{LZ_decompress_sync_to_member}, data +written with @samp{LZ_decompress_write} will be consumed and +@samp{LZ_decompress_read} will return 0 until a header is found. + +This function is useful to discard any data preceding the first member, +or to discard the rest of the current member, for example in case of a +data error. If the decoder is already at the beginning of a member, this +function does nothing. +@end deftypefun + + +@deftypefun int LZ_decompress_read ( struct LZ_Decoder * const @var{decoder}, uint8_t * const @var{buffer}, const int @var{size} ) +Reads up to @var{size} bytes from the stream pointed to by @var{decoder}, +storing the results in @var{buffer}. If @w{LZ_API_VERSION >= 1012}, +@var{buffer} may be a null pointer, in which case the bytes read are +discarded. + +Returns the number of bytes actually read. This might be less than +@var{size}; for example, if there aren't that many bytes left in the stream +or if more bytes have to be yet written with the function +@samp{LZ_decompress_write}. Note that reading less than @var{size} bytes is +not an error. + +@samp{LZ_decompress_read} returns at least once per member so that +@samp{LZ_decompress_member_finished} can be called (and trailer data +retrieved) for each member, even for empty members. Therefore, +@samp{LZ_decompress_read} returning 0 does not mean that the end of the +stream has been reached. The increase in the value returned by +@samp{LZ_decompress_total_in_size} can be used to tell the end of the stream +from an empty member. + +In case of decompression error caused by corrupt or truncated data, +@samp{LZ_decompress_read} does not signal the error immediately to the +application, but waits until all the bytes decoded have been read. This +allows tools like +@uref{http://www.nongnu.org/lzip/manual/tarlz_manual.html,,tarlz} to +recover as much data as possible from each damaged member. +@ifnothtml +@xref{Top,tarlz manual,,tarlz}. +@end ifnothtml +@end deftypefun + + +@deftypefun int LZ_decompress_write ( struct LZ_Decoder * const @var{decoder}, uint8_t * const @var{buffer}, const int @var{size} ) +Writes up to @var{size} bytes from @var{buffer} to the stream pointed to by +@var{decoder}. Returns the number of bytes actually written. This might be +less than @var{size}. Note that writing less than @var{size} bytes is not an +error. +@end deftypefun + + +@deftypefun int LZ_decompress_write_size ( struct LZ_Decoder * const @var{decoder} ) +Returns the maximum number of bytes that can be immediately written through +@samp{LZ_decompress_write}. This number varies smoothly; each compressed +byte consumed may be overwritten immediately, increasing by 1 the value +returned. + +It is guaranteed that an immediate call to @samp{LZ_decompress_write} will +accept a @var{size} up to the returned number of bytes. +@end deftypefun + + +@deftypefun {enum LZ_Errno} LZ_decompress_errno ( struct LZ_Decoder * const @var{decoder} ) +Returns the current error code for @var{decoder}. @xref{Error codes}. +It is safe to call @samp{LZ_decompress_errno} with a null argument, in which +case it returns @samp{LZ_bad_argument}. +@end deftypefun + + +@deftypefun int LZ_decompress_finished ( struct LZ_Decoder * const @var{decoder} ) +Returns 1 if all the data have been read and @samp{LZ_decompress_close} +can be safely called. Otherwise it returns 0. @samp{LZ_decompress_finished} +does not imply @samp{LZ_decompress_member_finished}. +@end deftypefun + + +@deftypefun int LZ_decompress_member_finished ( struct LZ_Decoder * const @var{decoder} ) +Returns 1 if the previous call to @samp{LZ_decompress_read} finished reading +the current member, indicating that final values for the member are available +through @samp{LZ_decompress_data_crc}, @samp{LZ_decompress_data_position}, +and @samp{LZ_decompress_member_position}. Otherwise it returns 0. +@end deftypefun + + +@deftypefun int LZ_decompress_member_version ( struct LZ_Decoder * const @var{decoder} ) +Returns the version of the current member, read from the member header. +@end deftypefun + + +@deftypefun int LZ_decompress_dictionary_size ( struct LZ_Decoder * const @var{decoder} ) +Returns the dictionary size of the current member, read from the member header. +@end deftypefun + + +@deftypefun {unsigned} LZ_decompress_data_crc ( struct LZ_Decoder * const @var{decoder} ) +Returns the 32 bit Cyclic Redundancy Check of the data decompressed from +the current member. The value returned is valid only when +@samp{LZ_decompress_member_finished} returns 1. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_decompress_data_position ( struct LZ_Decoder * const @var{decoder} ) +Returns the number of decompressed bytes already produced, but perhaps +not yet read, in the current member. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_decompress_member_position ( struct LZ_Decoder * const @var{decoder} ) +Returns the number of input bytes already decompressed in the current member. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_decompress_total_in_size ( struct LZ_Decoder * const @var{decoder} ) +Returns the total number of input bytes already decompressed. +@end deftypefun + + +@deftypefun {unsigned long long} LZ_decompress_total_out_size ( struct LZ_Decoder * const @var{decoder} ) +Returns the total number of decompressed bytes already produced, but +perhaps not yet read. +@end deftypefun + + +@node Error codes +@chapter Error codes +@cindex error codes + +Most library functions return -1 to indicate that they have failed. But +this return value only tells you that an error has occurred. To find out +what kind of error it was, you need to verify the error code by calling +@samp{LZ_(de)compress_errno}. + +Library functions don't change the value returned by +@samp{LZ_(de)compress_errno} when they succeed; thus, the value returned +by @samp{LZ_(de)compress_errno} after a successful call is not +necessarily LZ_ok, and you should not use @samp{LZ_(de)compress_errno} +to determine whether a call failed. If the call failed, then you can +examine @samp{LZ_(de)compress_errno}. + +The error codes are defined in the header file @samp{lzlib.h}. + +@deftypevr Constant {enum LZ_Errno} LZ_ok +The value of this constant is 0 and is used to indicate that there is no error. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_bad_argument +At least one of the arguments passed to the library function was invalid. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_mem_error +No memory available. The system cannot allocate more virtual memory +because its capacity is full. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_sequence_error +A library function was called in the wrong order. For example +@samp{LZ_compress_restart_member} was called before +@samp{LZ_compress_member_finished} indicates that the current member is +finished. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_header_error +An invalid member header (one with the wrong magic bytes) was read. If +this happens at the end of the data stream it may indicate trailing data. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_unexpected_eof +The end of the data stream was reached in the middle of a member. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_data_error +The data stream is corrupt. If @samp{LZ_decompress_member_position} is 6 +or less, it indicates either a format version not supported, an invalid +dictionary size, a corrupt header in a multimember data stream, or +trailing data too similar to a valid lzip header. Lziprecover can be +used to remove conflicting trailing data from a file. +@end deftypevr + +@deftypevr Constant {enum LZ_Errno} LZ_library_error +A bug was detected in the library. Please, report it. @xref{Problems}. +@end deftypevr + + +@node Error messages +@chapter Error messages +@cindex error messages + +@deftypefun {const char *} LZ_strerror ( const enum LZ_Errno @var{lz_errno} ) +Returns the standard error message for a given error code. The messages +are fairly short; there are no multi-line messages or embedded newlines. +This function makes it easy for your program to report informative error +messages about the failure of a library call. + +The value of @var{lz_errno} normally comes from a call to +@samp{LZ_(de)compress_errno}. +@end deftypefun + + +@node Invoking minilzip +@chapter Invoking minilzip +@cindex invoking +@cindex options + +Minilzip is a test program for the compression library lzlib, fully +compatible with lzip 1.4 or newer. + +@uref{http://www.nongnu.org/lzip/lzip.html,,Lzip} +is a lossless data compressor with a user interface similar to the one +of gzip or bzip2. Lzip uses a simplified form of the 'Lempel-Ziv-Markov +chain-Algorithm' (LZMA) stream format and provides a 3 factor integrity +checking to maximize interoperability and optimize safety. Lzip can compress +about as fast as gzip @w{(lzip -0)} or compress most files more than bzip2 +@w{(lzip -9)}. Decompression speed is intermediate between gzip and bzip2. +Lzip is better than gzip and bzip2 from a data recovery perspective. Lzip +has been designed, written, and tested with great care to replace gzip and +bzip2 as the standard general-purpose compressed format for unix-like +systems. + +@noindent +The format for running minilzip is: + +@example +minilzip [@var{options}] [@var{files}] +@end example + +@noindent +If no file names are specified, minilzip compresses (or decompresses) from +standard input to standard output. A hyphen @samp{-} used as a @var{file} +argument means standard input. It can be mixed with other @var{files} and is +read just once, the first time it appears in the command line. + +minilzip supports the following +@uref{http://www.nongnu.org/arg-parser/manual/arg_parser_manual.html#Argument-syntax,,options}: +@ifnothtml +@xref{Argument syntax,,,arg_parser}. +@end ifnothtml + +@table @code +@item -h +@itemx --help +Print an informative help message describing the options and exit. + +@item -V +@itemx --version +Print the version number of minilzip on the standard output and exit. +This version number should be included in all bug reports. + +@item -a +@itemx --trailing-error +Exit with error status 2 if any remaining input is detected after +decompressing the last member. Such remaining input is usually trailing +garbage that can be safely ignored. + +@item -b @var{bytes} +@itemx --member-size=@var{bytes} +When compressing, set the member size limit to @var{bytes}. It is advisable +to keep members smaller than RAM size so that they can be repaired with +lziprecover in case of corruption. A small member size may degrade +compression ratio, so use it only when needed. Valid values range from +@w{100 kB} to @w{2 PiB}. Defaults to @w{2 PiB}. + +@item -c +@itemx --stdout +Compress or decompress to standard output; keep input files unchanged. If +compressing several files, each file is compressed independently. (The +output consists of a sequence of independently compressed members). This +option (or @samp{-o}) is needed when reading from a named pipe (fifo) or +from a device. Use it also to recover as much of the decompressed data as +possible when decompressing a corrupt file. @samp{-c} overrides @samp{-o} +and @samp{-S}. @samp{-c} has no effect when testing or listing. + +@item -d +@itemx --decompress +Decompress the files specified. If a file does not exist, can't be opened, +or the destination file already exists and @samp{--force} has not been +specified, minilzip continues decompressing the rest of the files and exits with +error status 1. If a file fails to decompress, or is a terminal, minilzip exits +immediately with error status 2 without decompressing the rest of the files. +A terminal is considered an uncompressed file, and therefore invalid. + +@item -f +@itemx --force +Force overwrite of output files. + +@item -F +@itemx --recompress +When compressing, force re-compression of files whose name already has +the @samp{.lz} or @samp{.tlz} suffix. + +@item -k +@itemx --keep +Keep (don't delete) input files during compression or decompression. + +@item -m @var{bytes} +@itemx --match-length=@var{bytes} +When compressing, set the match length limit in bytes. After a match +this long is found, the search is finished. Valid values range from 5 to +273. Larger values usually give better compression ratios but longer +compression times. + +@item -o @var{file} +@itemx --output=@var{file} +If @samp{-c} has not been also specified, write the (de)compressed output to +@var{file}; keep input files unchanged. If compressing several files, each +file is compressed independently. (The output consists of a sequence of +independently compressed members). This option (or @samp{-c}) is needed when +reading from a named pipe (fifo) or from a device. @w{@samp{-o -}} is +equivalent to @samp{-c}. @samp{-o} has no effect when testing or listing. + +When compressing and splitting the output in volumes, @var{file} is used as +a prefix, and several files named @samp{@var{file}00001.lz}, +@samp{@var{file}00002.lz}, etc, are created. In this case, only one input +file is allowed. + +@item -q +@itemx --quiet +Quiet operation. Suppress all messages. + +@item -s @var{bytes} +@itemx --dictionary-size=@var{bytes} +When compressing, set the dictionary size limit in bytes. Minilzip will use +for each file the largest dictionary size that does not exceed neither +the file size nor this limit. Valid values range from @w{4 KiB} to +@w{512 MiB}. Values 12 to 29 are interpreted as powers of two, meaning +2^12 to 2^29 bytes. Dictionary sizes are quantized so that they can be +coded in just one byte (@pxref{coded-dict-size}). If the size specified +does not match one of the valid sizes, it will be rounded upwards by +adding up to @w{(@var{bytes} / 8)} to it. + +For maximum compression you should use a dictionary size limit as large +as possible, but keep in mind that the decompression memory requirement +is affected at compression time by the choice of dictionary size limit. + +@item -S @var{bytes} +@itemx --volume-size=@var{bytes} +When compressing, and @samp{-c} has not been also specified, split the +compressed output into several volume files with names +@samp{original_name00001.lz}, @samp{original_name00002.lz}, etc, and set the +volume size limit to @var{bytes}. Input files are kept unchanged. Each +volume is a complete, maybe multimember, lzip file. A small volume size may +degrade compression ratio, so use it only when needed. Valid values range +from @w{100 kB} to @w{4 EiB}. + +@item -t +@itemx --test +Check integrity of the files specified, but don't decompress them. This +really performs a trial decompression and throws away the result. Use it +together with @samp{-v} to see information about the files. If a file +fails the test, does not exist, can't be opened, or is a terminal, minilzip +continues checking the rest of the files. A final diagnostic is shown at +verbosity level 1 or higher if any file fails the test when testing +multiple files. + +@item -v +@itemx --verbose +Verbose mode.@* +When compressing, show the compression ratio and size for each file +processed.@* +When decompressing or testing, further -v's (up to 4) increase the +verbosity level, showing status, compression ratio, dictionary size, +and trailer contents (CRC, data size, member size). + +@item -0 .. -9 +Compression level. Set the compression parameters (dictionary size and +match length limit) as shown in the table below. The default compression +level is @samp{-6}, equivalent to @w{@samp{-s8MiB -m36}}. Note that +@samp{-9} can be much slower than @samp{-0}. These options have no +effect when decompressing or testing. + +The bidimensional parameter space of LZMA can't be mapped to a linear +scale optimal for all files. If your files are large, very repetitive, +etc, you may need to use the options @samp{--dictionary-size} and +@samp{--match-length} directly to achieve optimal performance. + +If several compression levels or @samp{-s} or @samp{-m} options are +given, the last setting is used. For example @w{@samp{-9 -s64MiB}} is +equivalent to @w{@samp{-s64MiB -m273}} + +@multitable {Level} {Dictionary size (-s)} {Match length limit (-m)} +@item Level @tab Dictionary size (-s) @tab Match length limit (-m) +@item -0 @tab 64 KiB @tab 16 bytes +@item -1 @tab 1 MiB @tab 5 bytes +@item -2 @tab 1.5 MiB @tab 6 bytes +@item -3 @tab 2 MiB @tab 8 bytes +@item -4 @tab 3 MiB @tab 12 bytes +@item -5 @tab 4 MiB @tab 20 bytes +@item -6 @tab 8 MiB @tab 36 bytes +@item -7 @tab 16 MiB @tab 68 bytes +@item -8 @tab 24 MiB @tab 132 bytes +@item -9 @tab 32 MiB @tab 273 bytes +@end multitable + +@item --fast +@itemx --best +Aliases for GNU gzip compatibility. + +@item --loose-trailing +When decompressing or testing, allow trailing data whose first bytes are +so similar to the magic bytes of a lzip header that they can be confused +with a corrupt header. Use this option if a file triggers a "corrupt +header" error and the cause is not indeed a corrupt header. + +@item --check-lib +Compare the @uref{#Library-version,,version of lzlib} used to compile +minilzip with the version actually being used at run time and exit. Report +any differences found. Exit with error status 1 if differences are found. A +mismatch may indicate that lzlib is not correctly installed or that a +different version of lzlib has been installed after compiling the shared +version of minilzip. Exit with error status 2 if LZ_API_VERSION and +LZ_version_string don't match. @w{@samp{minilzip -v --check-lib}} shows the +version of lzlib being used and the value of LZ_API_VERSION (if defined). +@ifnothtml +@xref{Library version}. +@end ifnothtml + +@end table + +Numbers given as arguments to options may be followed by a multiplier +and an optional @samp{B} for "byte". + +Table of SI and binary prefixes (unit multipliers): + +@multitable {Prefix} {kilobyte (10^3 = 1000)} {|} {Prefix} {kibibyte (2^10 = 1024)} +@item Prefix @tab Value @tab | @tab Prefix @tab Value +@item k @tab kilobyte (10^3 = 1000) @tab | @tab Ki @tab kibibyte (2^10 = 1024) +@item M @tab megabyte (10^6) @tab | @tab Mi @tab mebibyte (2^20) +@item G @tab gigabyte (10^9) @tab | @tab Gi @tab gibibyte (2^30) +@item T @tab terabyte (10^12) @tab | @tab Ti @tab tebibyte (2^40) +@item P @tab petabyte (10^15) @tab | @tab Pi @tab pebibyte (2^50) +@item E @tab exabyte (10^18) @tab | @tab Ei @tab exbibyte (2^60) +@item Z @tab zettabyte (10^21) @tab | @tab Zi @tab zebibyte (2^70) +@item Y @tab yottabyte (10^24) @tab | @tab Yi @tab yobibyte (2^80) +@end multitable + +@sp 1 +Exit status: 0 for a normal exit, 1 for environmental problems (file not +found, invalid flags, I/O errors, etc), 2 to indicate a corrupt or invalid +input file, 3 for an internal consistency error (e.g., bug) which caused +minilzip to panic. + + +@node Data format +@chapter Data format +@cindex data format + +Perfection is reached, not when there is no longer anything to add, but +when there is no longer anything to take away.@* +--- Antoine de Saint-Exupery + +@sp 1 +In the diagram below, a box like this: + +@verbatim ++---+ +| | <-- the vertical bars might be missing ++---+ +@end verbatim + +represents one byte; a box like this: + +@verbatim ++==============+ +| | ++==============+ +@end verbatim + +represents a variable number of bytes. + +@sp 1 +Lzip data consist of a series of independent "members" (compressed data +sets). The members simply appear one after another in the data stream, with +no additional information before, between, or after them. Each member can +encode in compressed form up to @w{16 EiB - 1 byte} of uncompressed data. +The size of a multimember data stream is unlimited. + +Each member has the following structure: + +@verbatim ++--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +| ID string | VN | DS | LZMA stream | CRC32 | Data size | Member size | ++--+--+--+--+----+----+=============+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +@end verbatim + +All multibyte values are stored in little endian order. + +@table @samp +@item ID string (the "magic" bytes) +A four byte string, identifying the lzip format, with the value "LZIP" +(0x4C, 0x5A, 0x49, 0x50). + +@item VN (version number, 1 byte) +Just in case something needs to be modified in the future. 1 for now. + +@anchor{coded-dict-size} +@item DS (coded dictionary size, 1 byte) +The dictionary size is calculated by taking a power of 2 (the base size) +and subtracting from it a fraction between 0/16 and 7/16 of the base size.@* +Bits 4-0 contain the base 2 logarithm of the base size (12 to 29).@* +Bits 7-5 contain the numerator of the fraction (0 to 7) to subtract +from the base size to obtain the dictionary size.@* +Example: 0xD3 = 2^19 - 6 * 2^15 = 512 KiB - 6 * 32 KiB = 320 KiB@* +Valid values for dictionary size range from 4 KiB to 512 MiB. + +@item LZMA stream +The LZMA stream, finished by an "End Of Stream" marker. Uses default values +for encoder properties. +@ifnothtml +@xref{Stream format,,,lzip}, +@end ifnothtml +@ifhtml +See +@uref{http://www.nongnu.org/lzip/manual/lzip_manual.html#Stream-format,,Stream format} +@end ifhtml +for a complete description.@* +Lzip only uses the LZMA marker @samp{2} ("End Of Stream" marker). Lzlib +also uses the LZMA marker @samp{3} ("Sync Flush" marker). @xref{sync_flush}. + +@item CRC32 (4 bytes) +Cyclic Redundancy Check (CRC) of the original uncompressed data. + +@item Data size (8 bytes) +Size of the original uncompressed data. + +@item Member size (8 bytes) +Total size of the member, including header and trailer. This field acts +as a distributed index, allows the verification of stream integrity, and +facilitates the safe recovery of undamaged members from multimember files. +Member size should be limited to @w{2 PiB} to prevent the data size field +from overflowing. + +@end table + + +@node Examples +@chapter A small tutorial with examples +@cindex examples + +This chapter provides real code examples for the most common uses of the +library. See these examples in context in the files @samp{bbexample.c} and +@samp{ffexample.c} from the source distribution of lzlib. + +Note that the interface of lzlib is symmetrical. That is, the code for +normal compression and decompression is identical except because one calls +LZ_compress* functions while the other calls LZ_decompress* functions. + +@menu +* Buffer compression:: Buffer-to-buffer single-member compression +* Buffer decompression:: Buffer-to-buffer decompression +* File compression:: File-to-file single-member compression +* File decompression:: File-to-file decompression +* File compression mm:: File-to-file multimember compression +* Skipping data errors:: Decompression with automatic resynchronization +@end menu + + +@node Buffer compression +@section Buffer compression +@cindex buffer compression + +Buffer-to-buffer single-member compression +@w{(@var{member_size} > total output)}. + +@verbatim +/* Compress 'insize' bytes from 'inbuf' to 'outbuf'. + Return the size of the compressed data in '*outlenp'. + In case of error, or if 'outsize' is too small, return false and do not + modify '*outlenp'. +*/ +bool bbcompress( const uint8_t * const inbuf, const int insize, + const int dictionary_size, const int match_len_limit, + uint8_t * const outbuf, const int outsize, + int * const outlenp ) + { + int inpos = 0, outpos = 0; + bool error = false; + struct LZ_Encoder * const encoder = + LZ_compress_open( dictionary_size, match_len_limit, INT64_MAX ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { LZ_compress_close( encoder ); return false; } + + while( true ) + { + int ret = LZ_compress_write( encoder, inbuf + inpos, insize - inpos ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_compress_finish( encoder ); + ret = LZ_compress_read( encoder, outbuf + outpos, outsize - outpos ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_compress_finished( encoder ) == 1 ) break; + if( outpos >= outsize ) { error = true; break; } + } + + if( LZ_compress_close( encoder ) < 0 ) error = true; + if( error ) return false; + *outlenp = outpos; + return true; + } +@end verbatim + + +@node Buffer decompression +@section Buffer decompression +@cindex buffer decompression + +Buffer-to-buffer decompression. + +@verbatim +/* Decompress 'insize' bytes from 'inbuf' to 'outbuf'. + Return the size of the decompressed data in '*outlenp'. + In case of error, or if 'outsize' is too small, return false and do not + modify '*outlenp'. +*/ +bool bbdecompress( const uint8_t * const inbuf, const int insize, + uint8_t * const outbuf, const int outsize, + int * const outlenp ) + { + int inpos = 0, outpos = 0; + bool error = false; + struct LZ_Decoder * const decoder = LZ_decompress_open(); + if( !decoder || LZ_decompress_errno( decoder ) != LZ_ok ) + { LZ_decompress_close( decoder ); return false; } + + while( true ) + { + int ret = LZ_decompress_write( decoder, inbuf + inpos, insize - inpos ); + if( ret < 0 ) { error = true; break; } + inpos += ret; + if( inpos >= insize ) LZ_decompress_finish( decoder ); + ret = LZ_decompress_read( decoder, outbuf + outpos, outsize - outpos ); + if( ret < 0 ) { error = true; break; } + outpos += ret; + if( LZ_decompress_finished( decoder ) == 1 ) break; + if( outpos >= outsize ) { error = true; break; } + } + + if( LZ_decompress_close( decoder ) < 0 ) error = true; + if( error ) return false; + *outlenp = outpos; + return true; + } +@end verbatim + + +@node File compression +@section File compression +@cindex file compression + +File-to-file compression using LZ_compress_write_size. + +@verbatim +int ffcompress( struct LZ_Encoder * const encoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_finished( encoder ) == 1 ) return 0; + } + return 1; + } +@end verbatim + + +@node File decompression +@section File decompression +@cindex file decompression + +File-to-file decompression using LZ_decompress_write_size. + +@verbatim +int ffdecompress( struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_decompress_write( decoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } +@end verbatim + + +@node File compression mm +@section File-to-file multimember compression +@cindex multimember compression + +Example 1: Multimember compression with members of fixed size +@w{(@var{member_size} < total output)}. + +@verbatim +int ffmmcompress( FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384, member_size = 4096 }; + uint8_t buffer[buffer_size]; + bool done = false; + struct LZ_Encoder * const encoder = + LZ_compress_open( 65535, 16, member_size ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { fputs( "ffexample: Not enough memory.\n", stderr ); + LZ_compress_close( encoder ); return 1; } + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_member_finished( encoder ) == 1 ) + { + if( LZ_compress_finished( encoder ) == 1 ) { done = true; break; } + if( LZ_compress_restart_member( encoder, member_size ) < 0 ) break; + } + } + if( LZ_compress_close( encoder ) < 0 ) done = false; + return done; + } +@end verbatim + +@sp 1 +@noindent +Example 2: Multimember compression (user-restarted members). +(Call LZ_compress_open with @var{member_size} > largest member). + +@verbatim +/* Compress 'infile' to 'outfile' as a multimember stream with one member + for each line of text terminated by a newline character or by EOF. + Return 0 if success, 1 if error. +*/ +int fflfcompress( struct LZ_Encoder * const encoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + for( len = 0; len < size; ) + { + int ch = getc( infile ); + if( ch == EOF || ( buffer[len++] = ch ) == '\n' ) break; + } + /* avoid writing an empty member to outfile */ + if( len == 0 && LZ_compress_data_position( encoder ) == 0 ) return 0; + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) || buffer[len-1] == '\n' ) + LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_member_finished( encoder ) == 1 ) + { + if( feof( infile ) && LZ_compress_finished( encoder ) == 1 ) return 0; + if( LZ_compress_restart_member( encoder, INT64_MAX ) < 0 ) break; + } + } + return 1; + } +@end verbatim + + +@node Skipping data errors +@section Skipping data errors +@cindex skipping data errors + +@verbatim +/* Decompress 'infile' to 'outfile' with automatic resynchronization to + next member in case of data error, including the automatic removal of + leading garbage. +*/ +int ffrsdecompress( struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_decompress_write( decoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) + { + if( LZ_decompress_errno( decoder ) == LZ_header_error || + LZ_decompress_errno( decoder ) == LZ_data_error ) + { LZ_decompress_sync_to_member( decoder ); continue; } + else break; + } + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } +@end verbatim + + +@node Problems +@chapter Reporting bugs +@cindex bugs +@cindex getting help + +There are probably bugs in lzlib. There are certainly errors and +omissions in this manual. If you report them, they will get fixed. If +you don't, no one will ever know about them and they will remain unfixed +for all eternity, if not longer. + +If you find a bug in lzlib, please send electronic mail to +@email{lzip-bug@@nongnu.org}. Include the version number, which you can +find by running @w{@samp{minilzip --version}} and +@w{@samp{minilzip -v --check-lib}}. + + +@node Concept index +@unnumbered Concept index + +@printindex cp + +@bye diff --git a/doc/minilzip.1 b/doc/minilzip.1 new file mode 100644 index 0000000..0c4c06d --- /dev/null +++ b/doc/minilzip.1 @@ -0,0 +1,134 @@ +.\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.47.16. +.TH MINILZIP "1" "January 2022" "minilzip 1.13" "User Commands" +.SH NAME +minilzip \- reduces the size of files +.SH SYNOPSIS +.B minilzip +[\fI\,options\/\fR] [\fI\,files\/\fR] +.SH DESCRIPTION +Minilzip is a test program for the compression library lzlib, fully +compatible with lzip 1.4 or newer. +.PP +Lzip is a lossless data compressor with a user interface similar to the one +of gzip or bzip2. Lzip uses a simplified form of the 'Lempel\-Ziv\-Markov +chain\-Algorithm' (LZMA) stream format and provides a 3 factor integrity +checking to maximize interoperability and optimize safety. Lzip can compress +about as fast as gzip (lzip \fB\-0\fR) or compress most files more than bzip2 +(lzip \fB\-9\fR). Decompression speed is intermediate between gzip and bzip2. +Lzip is better than gzip and bzip2 from a data recovery perspective. Lzip +has been designed, written, and tested with great care to replace gzip and +bzip2 as the standard general\-purpose compressed format for unix\-like +systems. +.SH OPTIONS +.TP +\fB\-h\fR, \fB\-\-help\fR +display this help and exit +.TP +\fB\-V\fR, \fB\-\-version\fR +output version information and exit +.TP +\fB\-a\fR, \fB\-\-trailing\-error\fR +exit with error status if trailing data +.TP +\fB\-b\fR, \fB\-\-member\-size=\fR +set member size limit in bytes +.TP +\fB\-c\fR, \fB\-\-stdout\fR +write to standard output, keep input files +.TP +\fB\-d\fR, \fB\-\-decompress\fR +decompress +.TP +\fB\-f\fR, \fB\-\-force\fR +overwrite existing output files +.TP +\fB\-F\fR, \fB\-\-recompress\fR +force re\-compression of compressed files +.TP +\fB\-k\fR, \fB\-\-keep\fR +keep (don't delete) input files +.TP +\fB\-m\fR, \fB\-\-match\-length=\fR +set match length limit in bytes [36] +.TP +\fB\-o\fR, \fB\-\-output=\fR +write to , keep input files +.TP +\fB\-q\fR, \fB\-\-quiet\fR +suppress all messages +.TP +\fB\-s\fR, \fB\-\-dictionary\-size=\fR +set dictionary size limit in bytes [8 MiB] +.TP +\fB\-S\fR, \fB\-\-volume\-size=\fR +set volume size limit in bytes +.TP +\fB\-t\fR, \fB\-\-test\fR +test compressed file integrity +.TP +\fB\-v\fR, \fB\-\-verbose\fR +be verbose (a 2nd \fB\-v\fR gives more) +.TP +\fB\-0\fR .. \fB\-9\fR +set compression level [default 6] +.TP +\fB\-\-fast\fR +alias for \fB\-0\fR +.TP +\fB\-\-best\fR +alias for \fB\-9\fR +.TP +\fB\-\-loose\-trailing\fR +allow trailing data seeming corrupt header +.TP +\fB\-\-check\-lib\fR +compare version of lzlib.h with liblz.{a,so} +.PP +If no file names are given, or if a file is '\-', minilzip compresses or +decompresses from standard input to standard output. +Numbers may be followed by a multiplier: k = kB = 10^3 = 1000, +Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc... +Dictionary sizes 12 to 29 are interpreted as powers of two, meaning 2^12 +to 2^29 bytes. +.PP +The bidimensional parameter space of LZMA can't be mapped to a linear +scale optimal for all files. If your files are large, very repetitive, +etc, you may need to use the options \fB\-\-dictionary\-size\fR and \fB\-\-match\-length\fR +directly to achieve optimal performance. +.PP +To extract all the files from archive 'foo.tar.lz', use the commands +\&'tar \fB\-xf\fR foo.tar.lz' or 'minilzip \fB\-cd\fR foo.tar.lz | tar \fB\-xf\fR \-'. +.PP +Exit status: 0 for a normal exit, 1 for environmental problems (file +not found, invalid flags, I/O errors, etc), 2 to indicate a corrupt or +invalid input file, 3 for an internal consistency error (e.g., bug) which +caused minilzip to panic. +.PP +The ideas embodied in lzlib are due to (at least) the following people: +Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the +definition of Markov chains), G.N.N. Martin (for the definition of range +encoding), Igor Pavlov (for putting all the above together in LZMA), and +Julian Seward (for bzip2's CLI). +.SH "REPORTING BUGS" +Report bugs to lzip\-bug@nongnu.org +.br +Lzlib home page: http://www.nongnu.org/lzip/lzlib.html +.SH COPYRIGHT +Copyright \(co 2022 Antonio Diaz Diaz. +Using lzlib 1.13 +License GPLv2+: GNU GPL version 2 or later +.br +This is free software: you are free to change and redistribute it. +There is NO WARRANTY, to the extent permitted by law. +.SH "SEE ALSO" +The full documentation for +.B minilzip +is maintained as a Texinfo manual. If the +.B info +and +.B minilzip +programs are properly installed at your site, the command +.IP +.B info lzlib +.PP +should give you access to the complete manual. diff --git a/encoder.c b/encoder.c new file mode 100644 index 0000000..b76dafa --- /dev/null +++ b/encoder.c @@ -0,0 +1,586 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +static int LZe_get_match_pairs( struct LZ_encoder * const e, struct Pair * pairs ) + { + int32_t * ptr0 = e->eb.mb.pos_array + ( e->eb.mb.cyclic_pos << 1 ); + int32_t * ptr1 = ptr0 + 1; + int len_limit = e->match_len_limit; + if( len_limit > Mb_available_bytes( &e->eb.mb ) ) + { + e->been_flushed = true; + len_limit = Mb_available_bytes( &e->eb.mb ); + if( len_limit < 4 ) { *ptr0 = *ptr1 = 0; return 0; } + } + + int maxlen = 3; /* only used if pairs != 0 */ + int num_pairs = 0; + const int min_pos = ( e->eb.mb.pos > e->eb.mb.dictionary_size ) ? + e->eb.mb.pos - e->eb.mb.dictionary_size : 0; + const uint8_t * const data = Mb_ptr_to_current_pos( &e->eb.mb ); + + unsigned tmp = crc32[data[0]] ^ data[1]; + const int key2 = tmp & ( num_prev_positions2 - 1 ); + tmp ^= (unsigned)data[2] << 8; + const int key3 = num_prev_positions2 + ( tmp & ( num_prev_positions3 - 1 ) ); + const int key4 = num_prev_positions2 + num_prev_positions3 + + ( ( tmp ^ ( crc32[data[3]] << 5 ) ) & e->eb.mb.key4_mask ); + + if( pairs ) + { + const int np2 = e->eb.mb.prev_positions[key2]; + const int np3 = e->eb.mb.prev_positions[key3]; + if( np2 > min_pos && e->eb.mb.buffer[np2-1] == data[0] ) + { + pairs[0].dis = e->eb.mb.pos - np2; + pairs[0].len = maxlen = 2 + ( np2 == np3 ); + num_pairs = 1; + } + if( np2 != np3 && np3 > min_pos && e->eb.mb.buffer[np3-1] == data[0] ) + { + maxlen = 3; + pairs[num_pairs++].dis = e->eb.mb.pos - np3; + } + if( num_pairs > 0 ) + { + const int delta = pairs[num_pairs-1].dis + 1; + while( maxlen < len_limit && data[maxlen-delta] == data[maxlen] ) + ++maxlen; + pairs[num_pairs-1].len = maxlen; + if( maxlen < 3 ) maxlen = 3; + if( maxlen >= len_limit ) pairs = 0; /* done. now just skip */ + } + } + + const int pos1 = e->eb.mb.pos + 1; + e->eb.mb.prev_positions[key2] = pos1; + e->eb.mb.prev_positions[key3] = pos1; + int newpos1 = e->eb.mb.prev_positions[key4]; + e->eb.mb.prev_positions[key4] = pos1; + + int len = 0, len0 = 0, len1 = 0; + + int count; + for( count = e->cycles; ; ) + { + if( newpos1 <= min_pos || --count < 0 ) { *ptr0 = *ptr1 = 0; break; } + + if( e->been_flushed ) len = 0; + const int delta = pos1 - newpos1; + int32_t * const newptr = e->eb.mb.pos_array + + ( ( e->eb.mb.cyclic_pos - delta + + ( (e->eb.mb.cyclic_pos >= delta) ? 0 : e->eb.mb.dictionary_size + 1 ) ) << 1 ); + if( data[len-delta] == data[len] ) + { + while( ++len < len_limit && data[len-delta] == data[len] ) {} + if( pairs && maxlen < len ) + { + pairs[num_pairs].dis = delta - 1; + pairs[num_pairs].len = maxlen = len; + ++num_pairs; + } + if( len >= len_limit ) + { + *ptr0 = newptr[0]; + *ptr1 = newptr[1]; + break; + } + } + if( data[len-delta] < data[len] ) + { + *ptr0 = newpos1; + ptr0 = newptr + 1; + newpos1 = *ptr0; + len0 = len; if( len1 < len ) len = len1; + } + else + { + *ptr1 = newpos1; + ptr1 = newptr; + newpos1 = *ptr1; + len1 = len; if( len0 < len ) len = len0; + } + } + return num_pairs; + } + + +static void LZe_update_distance_prices( struct LZ_encoder * const e ) + { + int dis, len_state; + for( dis = start_dis_model; dis < modeled_distances; ++dis ) + { + const int dis_slot = dis_slots[dis]; + const int direct_bits = ( dis_slot >> 1 ) - 1; + const int base = ( 2 | ( dis_slot & 1 ) ) << direct_bits; + const int price = price_symbol_reversed( e->eb.bm_dis + ( base - dis_slot ), + dis - base, direct_bits ); + for( len_state = 0; len_state < len_states; ++len_state ) + e->dis_prices[len_state][dis] = price; + } + + for( len_state = 0; len_state < len_states; ++len_state ) + { + int * const dsp = e->dis_slot_prices[len_state]; + const Bit_model * const bmds = e->eb.bm_dis_slot[len_state]; + int slot = 0; + for( ; slot < end_dis_model; ++slot ) + dsp[slot] = price_symbol6( bmds, slot ); + for( ; slot < e->num_dis_slots; ++slot ) + dsp[slot] = price_symbol6( bmds, slot ) + + (((( slot >> 1 ) - 1 ) - dis_align_bits ) << price_shift_bits ); + + int * const dp = e->dis_prices[len_state]; + for( dis = 0; dis < start_dis_model; ++dis ) + dp[dis] = dsp[dis]; + for( ; dis < modeled_distances; ++dis ) + dp[dis] += dsp[dis_slots[dis]]; + } + } + + +/* Return the number of bytes advanced (ahead). + trials[0]..trials[ahead-1] contain the steps to encode. + ( trials[0].dis4 == -1 ) means literal. + A match/rep longer or equal than match_len_limit finishes the sequence. +*/ +static int LZe_sequence_optimizer( struct LZ_encoder * const e, + const int reps[num_rep_distances], + const State state ) + { + int num_pairs, num_trials; + int i, rep, len; + + if( e->pending_num_pairs > 0 ) /* from previous call */ + { + num_pairs = e->pending_num_pairs; + e->pending_num_pairs = 0; + } + else + num_pairs = LZe_read_match_distances( e ); + const int main_len = ( num_pairs > 0 ) ? e->pairs[num_pairs-1].len : 0; + + int replens[num_rep_distances]; + int rep_index = 0; + for( i = 0; i < num_rep_distances; ++i ) + { + replens[i] = Mb_true_match_len( &e->eb.mb, 0, reps[i] + 1 ); + if( replens[i] > replens[rep_index] ) rep_index = i; + } + if( replens[rep_index] >= e->match_len_limit ) + { + e->trials[0].price = replens[rep_index]; + e->trials[0].dis4 = rep_index; + if( !LZe_move_and_update( e, replens[rep_index] ) ) return 0; + return replens[rep_index]; + } + + if( main_len >= e->match_len_limit ) + { + e->trials[0].price = main_len; + e->trials[0].dis4 = e->pairs[num_pairs-1].dis + num_rep_distances; + if( !LZe_move_and_update( e, main_len ) ) return 0; + return main_len; + } + + const int pos_state = Mb_data_position( &e->eb.mb ) & pos_state_mask; + const int match_price = price1( e->eb.bm_match[state][pos_state] ); + const int rep_match_price = match_price + price1( e->eb.bm_rep[state] ); + const uint8_t prev_byte = Mb_peek( &e->eb.mb, 1 ); + const uint8_t cur_byte = Mb_peek( &e->eb.mb, 0 ); + const uint8_t match_byte = Mb_peek( &e->eb.mb, reps[0] + 1 ); + + e->trials[1].price = price0( e->eb.bm_match[state][pos_state] ); + if( St_is_char( state ) ) + e->trials[1].price += LZeb_price_literal( &e->eb, prev_byte, cur_byte ); + else + e->trials[1].price += LZeb_price_matched( &e->eb, prev_byte, cur_byte, match_byte ); + e->trials[1].dis4 = -1; /* literal */ + + if( match_byte == cur_byte ) + Tr_update( &e->trials[1], rep_match_price + + LZeb_price_shortrep( &e->eb, state, pos_state ), 0, 0 ); + + num_trials = max( main_len, replens[rep_index] ); + + if( num_trials < min_match_len ) + { + e->trials[0].price = 1; + e->trials[0].dis4 = e->trials[1].dis4; + if( !Mb_move_pos( &e->eb.mb ) ) return 0; + return 1; + } + + e->trials[0].state = state; + for( i = 0; i < num_rep_distances; ++i ) + e->trials[0].reps[i] = reps[i]; + + for( len = min_match_len; len <= num_trials; ++len ) + e->trials[len].price = infinite_price; + + for( rep = 0; rep < num_rep_distances; ++rep ) + { + if( replens[rep] < min_match_len ) continue; + const int price = rep_match_price + LZeb_price_rep( &e->eb, rep, state, pos_state ); + for( len = min_match_len; len <= replens[rep]; ++len ) + Tr_update( &e->trials[len], price + + Lp_price( &e->rep_len_prices, len, pos_state ), rep, 0 ); + } + + if( main_len > replens[0] ) + { + const int normal_match_price = match_price + price0( e->eb.bm_rep[state] ); + int i = 0, len = max( replens[0] + 1, min_match_len ); + while( len > e->pairs[i].len ) ++i; + while( true ) + { + const int dis = e->pairs[i].dis; + Tr_update( &e->trials[len], normal_match_price + + LZe_price_pair( e, dis, len, pos_state ), + dis + num_rep_distances, 0 ); + if( ++len > e->pairs[i].len && ++i >= num_pairs ) break; + } + } + + int cur = 0; + while( true ) /* price optimization loop */ + { + if( !Mb_move_pos( &e->eb.mb ) ) return 0; + if( ++cur >= num_trials ) /* no more initialized trials */ + { + LZe_backward( e, cur ); + return cur; + } + + const int num_pairs = LZe_read_match_distances( e ); + const int newlen = ( num_pairs > 0 ) ? e->pairs[num_pairs-1].len : 0; + if( newlen >= e->match_len_limit ) + { + e->pending_num_pairs = num_pairs; + LZe_backward( e, cur ); + return cur; + } + + /* give final values to current trial */ + struct Trial * cur_trial = &e->trials[cur]; + State cur_state; + { + const int dis4 = cur_trial->dis4; + int prev_index = cur_trial->prev_index; + const int prev_index2 = cur_trial->prev_index2; + + if( prev_index2 == single_step_trial ) + { + cur_state = e->trials[prev_index].state; + if( prev_index + 1 == cur ) /* len == 1 */ + { + if( dis4 == 0 ) cur_state = St_set_short_rep( cur_state ); + else cur_state = St_set_char( cur_state ); /* literal */ + } + else if( dis4 < num_rep_distances ) cur_state = St_set_rep( cur_state ); + else cur_state = St_set_match( cur_state ); + } + else + { + if( prev_index2 == dual_step_trial ) /* dis4 == 0 (rep0) */ + --prev_index; + else /* prev_index2 >= 0 */ + prev_index = prev_index2; + cur_state = St_set_char_rep(); + } + cur_trial->state = cur_state; + for( i = 0; i < num_rep_distances; ++i ) + cur_trial->reps[i] = e->trials[prev_index].reps[i]; + mtf_reps( dis4, cur_trial->reps ); /* literal is ignored */ + } + + const int pos_state = Mb_data_position( &e->eb.mb ) & pos_state_mask; + const uint8_t prev_byte = Mb_peek( &e->eb.mb, 1 ); + const uint8_t cur_byte = Mb_peek( &e->eb.mb, 0 ); + const uint8_t match_byte = Mb_peek( &e->eb.mb, cur_trial->reps[0] + 1 ); + + int next_price = cur_trial->price + + price0( e->eb.bm_match[cur_state][pos_state] ); + if( St_is_char( cur_state ) ) + next_price += LZeb_price_literal( &e->eb, prev_byte, cur_byte ); + else + next_price += LZeb_price_matched( &e->eb, prev_byte, cur_byte, match_byte ); + + /* try last updates to next trial */ + struct Trial * next_trial = &e->trials[cur+1]; + + Tr_update( next_trial, next_price, -1, cur ); /* literal */ + + const int match_price = cur_trial->price + price1( e->eb.bm_match[cur_state][pos_state] ); + const int rep_match_price = match_price + price1( e->eb.bm_rep[cur_state] ); + + if( match_byte == cur_byte && next_trial->dis4 != 0 && + next_trial->prev_index2 == single_step_trial ) + { + const int price = rep_match_price + + LZeb_price_shortrep( &e->eb, cur_state, pos_state ); + if( price <= next_trial->price ) + { + next_trial->price = price; + next_trial->dis4 = 0; /* rep0 */ + next_trial->prev_index = cur; + } + } + + const int triable_bytes = + min( Mb_available_bytes( &e->eb.mb ), max_num_trials - 1 - cur ); + if( triable_bytes < min_match_len ) continue; + + const int len_limit = min( e->match_len_limit, triable_bytes ); + + /* try literal + rep0 */ + if( match_byte != cur_byte && next_trial->prev_index != cur ) + { + const uint8_t * const data = Mb_ptr_to_current_pos( &e->eb.mb ); + const int dis = cur_trial->reps[0] + 1; + const int limit = min( e->match_len_limit + 1, triable_bytes ); + int len = 1; + while( len < limit && data[len-dis] == data[len] ) ++len; + if( --len >= min_match_len ) + { + const int pos_state2 = ( pos_state + 1 ) & pos_state_mask; + const State state2 = St_set_char( cur_state ); + const int price = next_price + + price1( e->eb.bm_match[state2][pos_state2] ) + + price1( e->eb.bm_rep[state2] ) + + LZe_price_rep0_len( e, len, state2, pos_state2 ); + while( num_trials < cur + 1 + len ) + e->trials[++num_trials].price = infinite_price; + Tr_update2( &e->trials[cur+1+len], price, cur + 1 ); + } + } + + int start_len = min_match_len; + + /* try rep distances */ + for( rep = 0; rep < num_rep_distances; ++rep ) + { + const uint8_t * const data = Mb_ptr_to_current_pos( &e->eb.mb ); + const int dis = cur_trial->reps[rep] + 1; + + if( data[0-dis] != data[0] || data[1-dis] != data[1] ) continue; + for( len = min_match_len; len < len_limit; ++len ) + if( data[len-dis] != data[len] ) break; + while( num_trials < cur + len ) + e->trials[++num_trials].price = infinite_price; + int price = rep_match_price + LZeb_price_rep( &e->eb, rep, cur_state, pos_state ); + for( i = min_match_len; i <= len; ++i ) + Tr_update( &e->trials[cur+i], price + + Lp_price( &e->rep_len_prices, i, pos_state ), rep, cur ); + + if( rep == 0 ) start_len = len + 1; /* discard shorter matches */ + + /* try rep + literal + rep0 */ + int len2 = len + 1; + const int limit = min( e->match_len_limit + len2, triable_bytes ); + while( len2 < limit && data[len2-dis] == data[len2] ) ++len2; + len2 -= len + 1; + if( len2 < min_match_len ) continue; + + int pos_state2 = ( pos_state + len ) & pos_state_mask; + State state2 = St_set_rep( cur_state ); + price += Lp_price( &e->rep_len_prices, len, pos_state ) + + price0( e->eb.bm_match[state2][pos_state2] ) + + LZeb_price_matched( &e->eb, data[len-1], data[len], data[len-dis] ); + pos_state2 = ( pos_state2 + 1 ) & pos_state_mask; + state2 = St_set_char( state2 ); + price += price1( e->eb.bm_match[state2][pos_state2] ) + + price1( e->eb.bm_rep[state2] ) + + LZe_price_rep0_len( e, len2, state2, pos_state2 ); + while( num_trials < cur + len + 1 + len2 ) + e->trials[++num_trials].price = infinite_price; + Tr_update3( &e->trials[cur+len+1+len2], price, rep, cur + len + 1, cur ); + } + + /* try matches */ + if( newlen >= start_len && newlen <= len_limit ) + { + const int normal_match_price = match_price + + price0( e->eb.bm_rep[cur_state] ); + + while( num_trials < cur + newlen ) + e->trials[++num_trials].price = infinite_price; + + int i = 0; + while( e->pairs[i].len < start_len ) ++i; + int dis = e->pairs[i].dis; + for( len = start_len; ; ++len ) + { + int price = normal_match_price + LZe_price_pair( e, dis, len, pos_state ); + Tr_update( &e->trials[cur+len], price, dis + num_rep_distances, cur ); + + /* try match + literal + rep0 */ + if( len == e->pairs[i].len ) + { + const uint8_t * const data = Mb_ptr_to_current_pos( &e->eb.mb ); + const int dis2 = dis + 1; + int len2 = len + 1; + const int limit = min( e->match_len_limit + len2, triable_bytes ); + while( len2 < limit && data[len2-dis2] == data[len2] ) ++len2; + len2 -= len + 1; + if( len2 >= min_match_len ) + { + int pos_state2 = ( pos_state + len ) & pos_state_mask; + State state2 = St_set_match( cur_state ); + price += price0( e->eb.bm_match[state2][pos_state2] ) + + LZeb_price_matched( &e->eb, data[len-1], data[len], data[len-dis2] ); + pos_state2 = ( pos_state2 + 1 ) & pos_state_mask; + state2 = St_set_char( state2 ); + price += price1( e->eb.bm_match[state2][pos_state2] ) + + price1( e->eb.bm_rep[state2] ) + + LZe_price_rep0_len( e, len2, state2, pos_state2 ); + + while( num_trials < cur + len + 1 + len2 ) + e->trials[++num_trials].price = infinite_price; + Tr_update3( &e->trials[cur+len+1+len2], price, + dis + num_rep_distances, cur + len + 1, cur ); + } + if( ++i >= num_pairs ) break; + dis = e->pairs[i].dis; + } + } + } + } + } + + +static bool LZe_encode_member( struct LZ_encoder * const e ) + { + const bool best = ( e->match_len_limit > 12 ); + const int dis_price_count = best ? 1 : 512; + const int align_price_count = best ? 1 : dis_align_size; + const int price_count = ( e->match_len_limit > 36 ) ? 1013 : 4093; + int i; + State * const state = &e->eb.state; + + if( e->eb.member_finished ) return true; + if( Re_member_position( &e->eb.renc ) >= e->eb.member_size_limit ) + { LZeb_try_full_flush( &e->eb ); return true; } + + if( Mb_data_position( &e->eb.mb ) == 0 && + !Mb_data_finished( &e->eb.mb ) ) /* encode first byte */ + { + if( !Mb_enough_available_bytes( &e->eb.mb ) || + !Re_enough_free_bytes( &e->eb.renc ) ) return true; + const uint8_t prev_byte = 0; + const uint8_t cur_byte = Mb_peek( &e->eb.mb, 0 ); + Re_encode_bit( &e->eb.renc, &e->eb.bm_match[*state][0], 0 ); + LZeb_encode_literal( &e->eb, prev_byte, cur_byte ); + CRC32_update_byte( &e->eb.crc, cur_byte ); + LZe_get_match_pairs( e, 0 ); + if( !Mb_move_pos( &e->eb.mb ) ) return false; + } + + while( !Mb_data_finished( &e->eb.mb ) ) + { + if( !Mb_enough_available_bytes( &e->eb.mb ) || + !Re_enough_free_bytes( &e->eb.renc ) ) return true; + if( e->price_counter <= 0 && e->pending_num_pairs == 0 ) + { + e->price_counter = price_count; /* recalculate prices every these bytes */ + if( e->dis_price_counter <= 0 ) + { e->dis_price_counter = dis_price_count; LZe_update_distance_prices( e ); } + if( e->align_price_counter <= 0 ) + { + e->align_price_counter = align_price_count; + for( i = 0; i < dis_align_size; ++i ) + e->align_prices[i] = price_symbol_reversed( e->eb.bm_align, i, dis_align_bits ); + } + Lp_update_prices( &e->match_len_prices ); + Lp_update_prices( &e->rep_len_prices ); + } + + int ahead = LZe_sequence_optimizer( e, e->eb.reps, *state ); + e->price_counter -= ahead; + + for( i = 0; ahead > 0; ) + { + const int pos_state = + ( Mb_data_position( &e->eb.mb ) - ahead ) & pos_state_mask; + const int len = e->trials[i].price; + int dis = e->trials[i].dis4; + + bool bit = ( dis < 0 ); + Re_encode_bit( &e->eb.renc, &e->eb.bm_match[*state][pos_state], !bit ); + if( bit ) /* literal byte */ + { + const uint8_t prev_byte = Mb_peek( &e->eb.mb, ahead + 1 ); + const uint8_t cur_byte = Mb_peek( &e->eb.mb, ahead ); + CRC32_update_byte( &e->eb.crc, cur_byte ); + if( ( *state = St_set_char( *state ) ) < 4 ) + LZeb_encode_literal( &e->eb, prev_byte, cur_byte ); + else + { + const uint8_t match_byte = Mb_peek( &e->eb.mb, ahead + e->eb.reps[0] + 1 ); + LZeb_encode_matched( &e->eb, prev_byte, cur_byte, match_byte ); + } + } + else /* match or repeated match */ + { + CRC32_update_buf( &e->eb.crc, Mb_ptr_to_current_pos( &e->eb.mb ) - ahead, len ); + mtf_reps( dis, e->eb.reps ); + bit = ( dis < num_rep_distances ); + Re_encode_bit( &e->eb.renc, &e->eb.bm_rep[*state], bit ); + if( bit ) /* repeated match */ + { + bit = ( dis == 0 ); + Re_encode_bit( &e->eb.renc, &e->eb.bm_rep0[*state], !bit ); + if( bit ) + Re_encode_bit( &e->eb.renc, &e->eb.bm_len[*state][pos_state], len > 1 ); + else + { + Re_encode_bit( &e->eb.renc, &e->eb.bm_rep1[*state], dis > 1 ); + if( dis > 1 ) + Re_encode_bit( &e->eb.renc, &e->eb.bm_rep2[*state], dis > 2 ); + } + if( len == 1 ) *state = St_set_short_rep( *state ); + else + { + Re_encode_len( &e->eb.renc, &e->eb.rep_len_model, len, pos_state ); + Lp_decrement_counter( &e->rep_len_prices, pos_state ); + *state = St_set_rep( *state ); + } + } + else /* match */ + { + dis -= num_rep_distances; + LZeb_encode_pair( &e->eb, dis, len, pos_state ); + if( dis >= modeled_distances ) --e->align_price_counter; + --e->dis_price_counter; + Lp_decrement_counter( &e->match_len_prices, pos_state ); + *state = St_set_match( *state ); + } + } + ahead -= len; i += len; + if( Re_member_position( &e->eb.renc ) >= e->eb.member_size_limit ) + { + if( !Mb_dec_pos( &e->eb.mb, ahead ) ) return false; + LZeb_try_full_flush( &e->eb ); + return true; + } + } + } + LZeb_try_full_flush( &e->eb ); + return true; + } diff --git a/encoder.h b/encoder.h new file mode 100644 index 0000000..f17bb99 --- /dev/null +++ b/encoder.h @@ -0,0 +1,326 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +struct Len_prices + { + const struct Len_model * lm; + int len_symbols; + int count; + int prices[pos_states][max_len_symbols]; + int counters[pos_states]; /* may decrement below 0 */ + }; + +static inline void Lp_update_low_mid_prices( struct Len_prices * const lp, + const int pos_state ) + { + int * const pps = lp->prices[pos_state]; + int tmp = price0( lp->lm->choice1 ); + int len = 0; + for( ; len < len_low_symbols && len < lp->len_symbols; ++len ) + pps[len] = tmp + price_symbol3( lp->lm->bm_low[pos_state], len ); + if( len >= lp->len_symbols ) return; + tmp = price1( lp->lm->choice1 ) + price0( lp->lm->choice2 ); + for( ; len < len_low_symbols + len_mid_symbols && len < lp->len_symbols; ++len ) + pps[len] = tmp + + price_symbol3( lp->lm->bm_mid[pos_state], len - len_low_symbols ); + } + +static inline void Lp_update_high_prices( struct Len_prices * const lp ) + { + const int tmp = price1( lp->lm->choice1 ) + price1( lp->lm->choice2 ); + int len; + for( len = len_low_symbols + len_mid_symbols; len < lp->len_symbols; ++len ) + /* using 4 slots per value makes "Lp_price" faster */ + lp->prices[3][len] = lp->prices[2][len] = + lp->prices[1][len] = lp->prices[0][len] = tmp + + price_symbol8( lp->lm->bm_high, len - len_low_symbols - len_mid_symbols ); + } + +static inline void Lp_reset( struct Len_prices * const lp ) + { int i; for( i = 0; i < pos_states; ++i ) lp->counters[i] = 0; } + +static inline void Lp_init( struct Len_prices * const lp, + const struct Len_model * const lm, + const int match_len_limit ) + { + lp->lm = lm; + lp->len_symbols = match_len_limit + 1 - min_match_len; + lp->count = ( match_len_limit > 12 ) ? 1 : lp->len_symbols; + Lp_reset( lp ); + } + +static inline void Lp_decrement_counter( struct Len_prices * const lp, + const int pos_state ) + { --lp->counters[pos_state]; } + +static inline void Lp_update_prices( struct Len_prices * const lp ) + { + int pos_state; + bool high_pending = false; + for( pos_state = 0; pos_state < pos_states; ++pos_state ) + if( lp->counters[pos_state] <= 0 ) + { lp->counters[pos_state] = lp->count; + Lp_update_low_mid_prices( lp, pos_state ); high_pending = true; } + if( high_pending && lp->len_symbols > len_low_symbols + len_mid_symbols ) + Lp_update_high_prices( lp ); + } + +static inline int Lp_price( const struct Len_prices * const lp, + const int len, const int pos_state ) + { return lp->prices[pos_state][len - min_match_len]; } + + +struct Pair /* distance-length pair */ + { + int dis; + int len; + }; + +enum { infinite_price = 0x0FFFFFFF, + max_num_trials = 1 << 13, + single_step_trial = -2, + dual_step_trial = -1 }; + +struct Trial + { + State state; + int price; /* dual use var; cumulative price, match length */ + int dis4; /* -1 for literal, or rep, or match distance + 4 */ + int prev_index; /* index of prev trial in trials[] */ + int prev_index2; /* -2 trial is single step */ + /* -1 literal + rep0 */ + /* >= 0 ( rep or match ) + literal + rep0 */ + int reps[num_rep_distances]; + }; + +static inline void Tr_update( struct Trial * const trial, const int pr, + const int distance4, const int p_i ) + { + if( pr < trial->price ) + { trial->price = pr; trial->dis4 = distance4; trial->prev_index = p_i; + trial->prev_index2 = single_step_trial; } + } + +static inline void Tr_update2( struct Trial * const trial, const int pr, + const int p_i ) + { + if( pr < trial->price ) + { trial->price = pr; trial->dis4 = 0; trial->prev_index = p_i; + trial->prev_index2 = dual_step_trial; } + } + +static inline void Tr_update3( struct Trial * const trial, const int pr, + const int distance4, const int p_i, + const int p_i2 ) + { + if( pr < trial->price ) + { trial->price = pr; trial->dis4 = distance4; trial->prev_index = p_i; + trial->prev_index2 = p_i2; } + } + + +struct LZ_encoder + { + struct LZ_encoder_base eb; + int cycles; + int match_len_limit; + struct Len_prices match_len_prices; + struct Len_prices rep_len_prices; + int pending_num_pairs; + struct Pair pairs[max_match_len+1]; + struct Trial trials[max_num_trials]; + + int dis_slot_prices[len_states][2*max_dictionary_bits]; + int dis_prices[len_states][modeled_distances]; + int align_prices[dis_align_size]; + int num_dis_slots; + int price_counter; /* counters may decrement below 0 */ + int dis_price_counter; + int align_price_counter; + bool been_flushed; + }; + +static inline bool Mb_dec_pos( struct Matchfinder_base * const mb, + const int ahead ) + { + if( ahead < 0 || mb->pos < ahead ) return false; + mb->pos -= ahead; + if( mb->cyclic_pos < ahead ) mb->cyclic_pos += mb->dictionary_size + 1; + mb->cyclic_pos -= ahead; + return true; + } + +static int LZe_get_match_pairs( struct LZ_encoder * const e, struct Pair * pairs ); + + /* move-to-front dis in/into reps; do nothing if( dis4 <= 0 ) */ +static inline void mtf_reps( const int dis4, int reps[num_rep_distances] ) + { + if( dis4 >= num_rep_distances ) /* match */ + { + reps[3] = reps[2]; reps[2] = reps[1]; reps[1] = reps[0]; + reps[0] = dis4 - num_rep_distances; + } + else if( dis4 > 0 ) /* repeated match */ + { + const int distance = reps[dis4]; + int i; for( i = dis4; i > 0; --i ) reps[i] = reps[i-1]; + reps[0] = distance; + } + } + +static inline int LZeb_price_shortrep( const struct LZ_encoder_base * const eb, + const State state, const int pos_state ) + { + return price0( eb->bm_rep0[state] ) + price0( eb->bm_len[state][pos_state] ); + } + +static inline int LZeb_price_rep( const struct LZ_encoder_base * const eb, + const int rep, const State state, + const int pos_state ) + { + if( rep == 0 ) return price0( eb->bm_rep0[state] ) + + price1( eb->bm_len[state][pos_state] ); + int price = price1( eb->bm_rep0[state] ); + if( rep == 1 ) + price += price0( eb->bm_rep1[state] ); + else + { + price += price1( eb->bm_rep1[state] ); + price += price_bit( eb->bm_rep2[state], rep - 2 ); + } + return price; + } + +static inline int LZe_price_rep0_len( const struct LZ_encoder * const e, + const int len, const State state, + const int pos_state ) + { + return LZeb_price_rep( &e->eb, 0, state, pos_state ) + + Lp_price( &e->rep_len_prices, len, pos_state ); + } + +static inline int LZe_price_pair( const struct LZ_encoder * const e, + const int dis, const int len, + const int pos_state ) + { + const int price = Lp_price( &e->match_len_prices, len, pos_state ); + const int len_state = get_len_state( len ); + if( dis < modeled_distances ) + return price + e->dis_prices[len_state][dis]; + else + return price + e->dis_slot_prices[len_state][get_slot( dis )] + + e->align_prices[dis & (dis_align_size - 1)]; + } + +static inline int LZe_read_match_distances( struct LZ_encoder * const e ) + { + const int num_pairs = LZe_get_match_pairs( e, e->pairs ); + if( num_pairs > 0 ) + { + const int len = e->pairs[num_pairs-1].len; + if( len == e->match_len_limit && len < max_match_len ) + e->pairs[num_pairs-1].len = + Mb_true_match_len( &e->eb.mb, len, e->pairs[num_pairs-1].dis + 1 ); + } + return num_pairs; + } + +static inline bool LZe_move_and_update( struct LZ_encoder * const e, int n ) + { + while( true ) + { + if( !Mb_move_pos( &e->eb.mb ) ) return false; + if( --n <= 0 ) break; + LZe_get_match_pairs( e, 0 ); + } + return true; + } + +static inline void LZe_backward( struct LZ_encoder * const e, int cur ) + { + int dis4 = e->trials[cur].dis4; + while( cur > 0 ) + { + const int prev_index = e->trials[cur].prev_index; + struct Trial * const prev_trial = &e->trials[prev_index]; + + if( e->trials[cur].prev_index2 != single_step_trial ) + { + prev_trial->dis4 = -1; /* literal */ + prev_trial->prev_index = prev_index - 1; + prev_trial->prev_index2 = single_step_trial; + if( e->trials[cur].prev_index2 >= 0 ) + { + struct Trial * const prev_trial2 = &e->trials[prev_index-1]; + prev_trial2->dis4 = dis4; dis4 = 0; /* rep0 */ + prev_trial2->prev_index = e->trials[cur].prev_index2; + prev_trial2->prev_index2 = single_step_trial; + } + } + prev_trial->price = cur - prev_index; /* len */ + cur = dis4; dis4 = prev_trial->dis4; prev_trial->dis4 = cur; + cur = prev_index; + } + } + +enum { num_prev_positions3 = 1 << 16, + num_prev_positions2 = 1 << 10 }; + +static inline bool LZe_init( struct LZ_encoder * const e, + const int dict_size, const int len_limit, + const unsigned long long member_size ) + { + enum { before_size = max_num_trials, + /* bytes to keep in buffer after pos */ + after_size = max_num_trials + ( 2 * max_match_len ) + 1, + dict_factor = 2, + num_prev_positions23 = num_prev_positions2 + num_prev_positions3, + pos_array_factor = 2, + min_free_bytes = 2 * max_num_trials }; + + if( !LZeb_init( &e->eb, before_size, dict_size, after_size, dict_factor, + num_prev_positions23, pos_array_factor, min_free_bytes, + member_size ) ) return false; + e->cycles = ( len_limit < max_match_len ) ? 16 + ( len_limit / 2 ) : 256; + e->match_len_limit = len_limit; + Lp_init( &e->match_len_prices, &e->eb.match_len_model, e->match_len_limit ); + Lp_init( &e->rep_len_prices, &e->eb.rep_len_model, e->match_len_limit ); + e->pending_num_pairs = 0; + e->num_dis_slots = 2 * real_bits( e->eb.mb.dictionary_size - 1 ); + e->trials[1].prev_index = 0; + e->trials[1].prev_index2 = single_step_trial; + e->price_counter = 0; + e->dis_price_counter = 0; + e->align_price_counter = 0; + e->been_flushed = false; + return true; + } + +static inline void LZe_reset( struct LZ_encoder * const e, + const unsigned long long member_size ) + { + LZeb_reset( &e->eb, member_size ); + Lp_reset( &e->match_len_prices ); + Lp_reset( &e->rep_len_prices ); + e->pending_num_pairs = 0; + e->price_counter = 0; + e->dis_price_counter = 0; + e->align_price_counter = 0; + e->been_flushed = false; + } diff --git a/encoder_base.c b/encoder_base.c new file mode 100644 index 0000000..4535352 --- /dev/null +++ b/encoder_base.c @@ -0,0 +1,196 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +static bool Mb_normalize_pos( struct Matchfinder_base * const mb ) + { + if( mb->pos > mb->stream_pos ) + { mb->pos = mb->stream_pos; return false; } + if( !mb->at_stream_end ) + { + int i; + /* offset is int32_t for the min below */ + const int32_t offset = mb->pos - mb->before_size - mb->dictionary_size; + const int size = mb->stream_pos - offset; + memmove( mb->buffer, mb->buffer + offset, size ); + mb->partial_data_pos += offset; + mb->pos -= offset; /* pos = before_size + dictionary_size */ + mb->stream_pos -= offset; + for( i = 0; i < mb->num_prev_positions; ++i ) + mb->prev_positions[i] -= min( mb->prev_positions[i], offset ); + for( i = 0; i < mb->pos_array_size; ++i ) + mb->pos_array[i] -= min( mb->pos_array[i], offset ); + } + return true; + } + + +static bool Mb_init( struct Matchfinder_base * const mb, const int before_size, + const int dict_size, const int after_size, + const int dict_factor, const int num_prev_positions23, + const int pos_array_factor ) + { + const int buffer_size_limit = + ( dict_factor * dict_size ) + before_size + after_size; + int i; + + mb->partial_data_pos = 0; + mb->before_size = before_size; + mb->after_size = after_size; + mb->pos = 0; + mb->cyclic_pos = 0; + mb->stream_pos = 0; + mb->num_prev_positions23 = num_prev_positions23; + mb->at_stream_end = false; + mb->sync_flush_pending = false; + + mb->buffer_size = max( 65536, buffer_size_limit ); + mb->buffer = (uint8_t *)malloc( mb->buffer_size ); + if( !mb->buffer ) return false; + mb->saved_dictionary_size = dict_size; + mb->dictionary_size = dict_size; + mb->pos_limit = mb->buffer_size - after_size; + unsigned size = 1 << max( 16, real_bits( mb->dictionary_size - 1 ) - 2 ); + if( mb->dictionary_size > 1 << 26 ) size >>= 1; /* 64 MiB */ + mb->key4_mask = size - 1; /* increases with dictionary size */ + size += num_prev_positions23; + mb->num_prev_positions = size; + + mb->pos_array_size = pos_array_factor * ( mb->dictionary_size + 1 ); + size += mb->pos_array_size; + if( size * sizeof mb->prev_positions[0] <= size ) mb->prev_positions = 0; + else mb->prev_positions = + (int32_t *)malloc( size * sizeof mb->prev_positions[0] ); + if( !mb->prev_positions ) { free( mb->buffer ); return false; } + mb->pos_array = mb->prev_positions + mb->num_prev_positions; + for( i = 0; i < mb->num_prev_positions; ++i ) mb->prev_positions[i] = 0; + return true; + } + + +static void Mb_adjust_array( struct Matchfinder_base * const mb ) + { + int size = 1 << max( 16, real_bits( mb->dictionary_size - 1 ) - 2 ); + if( mb->dictionary_size > 1 << 26 ) size >>= 1; /* 64 MiB */ + mb->key4_mask = size - 1; + size += mb->num_prev_positions23; + mb->num_prev_positions = size; + mb->pos_array = mb->prev_positions + mb->num_prev_positions; + } + + +static void Mb_adjust_dictionary_size( struct Matchfinder_base * const mb ) + { + if( mb->stream_pos < mb->dictionary_size ) + { + mb->dictionary_size = max( min_dictionary_size, mb->stream_pos ); + Mb_adjust_array( mb ); + mb->pos_limit = mb->buffer_size; + } + } + + +static void Mb_reset( struct Matchfinder_base * const mb ) + { + int i; + if( mb->stream_pos > mb->pos ) + memmove( mb->buffer, mb->buffer + mb->pos, mb->stream_pos - mb->pos ); + mb->partial_data_pos = 0; + mb->stream_pos -= mb->pos; + mb->pos = 0; + mb->cyclic_pos = 0; + mb->at_stream_end = false; + mb->sync_flush_pending = false; + mb->dictionary_size = mb->saved_dictionary_size; + Mb_adjust_array( mb ); + mb->pos_limit = mb->buffer_size - mb->after_size; + for( i = 0; i < mb->num_prev_positions; ++i ) mb->prev_positions[i] = 0; + } + + +/* End Of Stream marker => (dis == 0xFFFFFFFFU, len == min_match_len) */ +static void LZeb_try_full_flush( struct LZ_encoder_base * const eb ) + { + if( eb->member_finished || + Cb_free_bytes( &eb->renc.cb ) < max_marker_size + eb->renc.ff_count + Lt_size ) + return; + eb->member_finished = true; + const int pos_state = Mb_data_position( &eb->mb ) & pos_state_mask; + const State state = eb->state; + Re_encode_bit( &eb->renc, &eb->bm_match[state][pos_state], 1 ); + Re_encode_bit( &eb->renc, &eb->bm_rep[state], 0 ); + LZeb_encode_pair( eb, 0xFFFFFFFFU, min_match_len, pos_state ); + Re_flush( &eb->renc ); + Lzip_trailer trailer; + Lt_set_data_crc( trailer, LZeb_crc( eb ) ); + Lt_set_data_size( trailer, Mb_data_position( &eb->mb ) ); + Lt_set_member_size( trailer, Re_member_position( &eb->renc ) + Lt_size ); + int i; + for( i = 0; i < Lt_size; ++i ) + Cb_put_byte( &eb->renc.cb, trailer[i] ); + } + + +/* Sync Flush marker => (dis == 0xFFFFFFFFU, len == min_match_len + 1) */ +static void LZeb_try_sync_flush( struct LZ_encoder_base * const eb ) + { + const unsigned min_size = eb->renc.ff_count + max_marker_size; + if( eb->member_finished || + Cb_free_bytes( &eb->renc.cb ) < min_size + max_marker_size ) return; + eb->mb.sync_flush_pending = false; + const unsigned long long old_mpos = Re_member_position( &eb->renc ); + const int pos_state = Mb_data_position( &eb->mb ) & pos_state_mask; + const State state = eb->state; + do { /* size of markers must be >= rd_min_available_bytes + 5 */ + Re_encode_bit( &eb->renc, &eb->bm_match[state][pos_state], 1 ); + Re_encode_bit( &eb->renc, &eb->bm_rep[state], 0 ); + LZeb_encode_pair( eb, 0xFFFFFFFFU, min_match_len + 1, pos_state ); + Re_flush( &eb->renc ); + } + while( Re_member_position( &eb->renc ) - old_mpos < min_size ); + } + + +static void LZeb_reset( struct LZ_encoder_base * const eb, + const unsigned long long member_size ) + { + const unsigned long long min_member_size = min_dictionary_size; + const unsigned long long max_member_size = 0x0008000000000000ULL; /* 2 PiB */ + int i; + Mb_reset( &eb->mb ); + eb->member_size_limit = + min( max( min_member_size, member_size ), max_member_size ) - + Lt_size - max_marker_size; + eb->crc = 0xFFFFFFFFU; + Bm_array_init( eb->bm_literal[0], (1 << literal_context_bits) * 0x300 ); + Bm_array_init( eb->bm_match[0], states * pos_states ); + Bm_array_init( eb->bm_rep, states ); + Bm_array_init( eb->bm_rep0, states ); + Bm_array_init( eb->bm_rep1, states ); + Bm_array_init( eb->bm_rep2, states ); + Bm_array_init( eb->bm_len[0], states * pos_states ); + Bm_array_init( eb->bm_dis_slot[0], len_states * (1 << dis_slot_bits) ); + Bm_array_init( eb->bm_dis, modeled_distances - end_dis_model + 1 ); + Bm_array_init( eb->bm_align, dis_align_size ); + Lm_init( &eb->match_len_model ); + Lm_init( &eb->rep_len_model ); + Re_reset( &eb->renc, eb->mb.dictionary_size ); + for( i = 0; i < num_rep_distances; ++i ) eb->reps[i] = 0; + eb->state = 0; + eb->member_finished = false; + } diff --git a/encoder_base.h b/encoder_base.h new file mode 100644 index 0000000..17ffc93 --- /dev/null +++ b/encoder_base.h @@ -0,0 +1,612 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +enum { price_shift_bits = 6, + price_step_bits = 2 }; + +static const uint8_t dis_slots[1<<10] = + { + 0, 1, 2, 3, 4, 4, 5, 5, 6, 6, 6, 6, 7, 7, 7, 7, + 8, 8, 8, 8, 8, 8, 8, 8, 9, 9, 9, 9, 9, 9, 9, 9, + 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, + 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, + 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, + 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, + 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, + 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, + 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, + 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, + 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, + 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, 14, + 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, + 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, + 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, + 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, 15, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, 18, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, + 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19, 19 }; + +static inline uint8_t get_slot( const unsigned dis ) + { + if( dis < (1 << 10) ) return dis_slots[dis]; + if( dis < (1 << 19) ) return dis_slots[dis>> 9] + 18; + if( dis < (1 << 28) ) return dis_slots[dis>>18] + 36; + return dis_slots[dis>>27] + 54; + } + + +static const short prob_prices[bit_model_total >> price_step_bits] = +{ +640, 539, 492, 461, 438, 419, 404, 390, 379, 369, 359, 351, 343, 336, 330, 323, +318, 312, 307, 302, 298, 293, 289, 285, 281, 277, 274, 270, 267, 264, 261, 258, +255, 252, 250, 247, 244, 242, 239, 237, 235, 232, 230, 228, 226, 224, 222, 220, +218, 216, 214, 213, 211, 209, 207, 206, 204, 202, 201, 199, 198, 196, 195, 193, +192, 190, 189, 188, 186, 185, 184, 182, 181, 180, 178, 177, 176, 175, 174, 172, +171, 170, 169, 168, 167, 166, 165, 164, 163, 162, 161, 159, 158, 157, 157, 156, +155, 154, 153, 152, 151, 150, 149, 148, 147, 146, 145, 145, 144, 143, 142, 141, +140, 140, 139, 138, 137, 136, 136, 135, 134, 133, 133, 132, 131, 130, 130, 129, +128, 127, 127, 126, 125, 125, 124, 123, 123, 122, 121, 121, 120, 119, 119, 118, +117, 117, 116, 115, 115, 114, 114, 113, 112, 112, 111, 111, 110, 109, 109, 108, +108, 107, 106, 106, 105, 105, 104, 104, 103, 103, 102, 101, 101, 100, 100, 99, + 99, 98, 98, 97, 97, 96, 96, 95, 95, 94, 94, 93, 93, 92, 92, 91, + 91, 90, 90, 89, 89, 88, 88, 88, 87, 87, 86, 86, 85, 85, 84, 84, + 83, 83, 83, 82, 82, 81, 81, 80, 80, 80, 79, 79, 78, 78, 77, 77, + 77, 76, 76, 75, 75, 75, 74, 74, 73, 73, 73, 72, 72, 71, 71, 71, + 70, 70, 70, 69, 69, 68, 68, 68, 67, 67, 67, 66, 66, 65, 65, 65, + 64, 64, 64, 63, 63, 63, 62, 62, 61, 61, 61, 60, 60, 60, 59, 59, + 59, 58, 58, 58, 57, 57, 57, 56, 56, 56, 55, 55, 55, 54, 54, 54, + 53, 53, 53, 53, 52, 52, 52, 51, 51, 51, 50, 50, 50, 49, 49, 49, + 48, 48, 48, 48, 47, 47, 47, 46, 46, 46, 45, 45, 45, 45, 44, 44, + 44, 43, 43, 43, 43, 42, 42, 42, 41, 41, 41, 41, 40, 40, 40, 40, + 39, 39, 39, 38, 38, 38, 38, 37, 37, 37, 37, 36, 36, 36, 35, 35, + 35, 35, 34, 34, 34, 34, 33, 33, 33, 33, 32, 32, 32, 32, 31, 31, + 31, 31, 30, 30, 30, 30, 29, 29, 29, 29, 28, 28, 28, 28, 27, 27, + 27, 27, 26, 26, 26, 26, 26, 25, 25, 25, 25, 24, 24, 24, 24, 23, + 23, 23, 23, 22, 22, 22, 22, 22, 21, 21, 21, 21, 20, 20, 20, 20, + 20, 19, 19, 19, 19, 18, 18, 18, 18, 18, 17, 17, 17, 17, 17, 16, + 16, 16, 16, 15, 15, 15, 15, 15, 14, 14, 14, 14, 14, 13, 13, 13, + 13, 13, 12, 12, 12, 12, 12, 11, 11, 11, 11, 10, 10, 10, 10, 10, + 9, 9, 9, 9, 9, 9, 8, 8, 8, 8, 8, 7, 7, 7, 7, 7, + 6, 6, 6, 6, 6, 5, 5, 5, 5, 5, 4, 4, 4, 4, 4, 4, + 3, 3, 3, 3, 3, 2, 2, 2, 2, 2, 1, 1, 1, 1, 1, 1 }; + +static inline int get_price( const int probability ) + { return prob_prices[probability >> price_step_bits]; } + + +static inline int price0( const Bit_model probability ) + { return get_price( probability ); } + +static inline int price1( const Bit_model probability ) + { return get_price( bit_model_total - probability ); } + +static inline int price_bit( const Bit_model bm, const bool bit ) + { return ( bit ? price1( bm ) : price0( bm ) ); } + + +static inline int price_symbol3( const Bit_model bm[], int symbol ) + { + bool bit = symbol & 1; + symbol |= 8; symbol >>= 1; + int price = price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + return price + price_bit( bm[1], symbol & 1 ); + } + + +static inline int price_symbol6( const Bit_model bm[], unsigned symbol ) + { + bool bit = symbol & 1; + symbol |= 64; symbol >>= 1; + int price = price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + return price + price_bit( bm[1], symbol & 1 ); + } + + +static inline int price_symbol8( const Bit_model bm[], int symbol ) + { + bool bit = symbol & 1; + symbol |= 0x100; symbol >>= 1; + int price = price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + bit = symbol & 1; symbol >>= 1; price += price_bit( bm[symbol], bit ); + return price + price_bit( bm[1], symbol & 1 ); + } + + +static inline int price_symbol_reversed( const Bit_model bm[], int symbol, + const int num_bits ) + { + int price = 0; + int model = 1; + int i; + for( i = num_bits; i > 0; --i ) + { + const bool bit = symbol & 1; + symbol >>= 1; + price += price_bit( bm[model], bit ); + model <<= 1; model |= bit; + } + return price; + } + + +static inline int price_matched( const Bit_model bm[], unsigned symbol, + unsigned match_byte ) + { + int price = 0; + unsigned mask = 0x100; + symbol |= mask; + while( true ) + { + const unsigned match_bit = ( match_byte <<= 1 ) & mask; + const bool bit = ( symbol <<= 1 ) & 0x100; + price += price_bit( bm[(symbol>>9)+match_bit+mask], bit ); + if( symbol >= 0x10000 ) return price; + mask &= ~(match_bit ^ symbol); /* if( match_bit != bit ) mask = 0; */ + } + } + + +struct Matchfinder_base + { + unsigned long long partial_data_pos; + uint8_t * buffer; /* input buffer */ + int32_t * prev_positions; /* 1 + last seen position of key. else 0 */ + int32_t * pos_array; /* may be tree or chain */ + int before_size; /* bytes to keep in buffer before dictionary */ + int after_size; /* bytes to keep in buffer after pos */ + int buffer_size; + int dictionary_size; /* bytes to keep in buffer before pos */ + int pos; /* current pos in buffer */ + int cyclic_pos; /* cycles through [0, dictionary_size] */ + int stream_pos; /* first byte not yet read from file */ + int pos_limit; /* when reached, a new block must be read */ + int key4_mask; + int num_prev_positions23; + int num_prev_positions; /* size of prev_positions */ + int pos_array_size; + int saved_dictionary_size; /* dictionary_size restored by Mb_reset */ + bool at_stream_end; /* stream_pos shows real end of file */ + bool sync_flush_pending; + }; + +static bool Mb_normalize_pos( struct Matchfinder_base * const mb ); + +static bool Mb_init( struct Matchfinder_base * const mb, const int before_size, + const int dict_size, const int after_size, + const int dict_factor, const int num_prev_positions23, + const int pos_array_factor ); + +static inline void Mb_free( struct Matchfinder_base * const mb ) + { free( mb->prev_positions ); free( mb->buffer ); } + +static inline uint8_t Mb_peek( const struct Matchfinder_base * const mb, + const int distance ) + { return mb->buffer[mb->pos-distance]; } + +static inline int Mb_available_bytes( const struct Matchfinder_base * const mb ) + { return mb->stream_pos - mb->pos; } + +static inline unsigned long long +Mb_data_position( const struct Matchfinder_base * const mb ) + { return mb->partial_data_pos + mb->pos; } + +static inline void Mb_finish( struct Matchfinder_base * const mb ) + { mb->at_stream_end = true; mb->sync_flush_pending = false; } + +static inline bool Mb_data_finished( const struct Matchfinder_base * const mb ) + { return mb->at_stream_end && mb->pos >= mb->stream_pos; } + +static inline bool Mb_flushing_or_end( const struct Matchfinder_base * const mb ) + { return mb->at_stream_end || mb->sync_flush_pending; } + +static inline int Mb_free_bytes( const struct Matchfinder_base * const mb ) + { if( Mb_flushing_or_end( mb ) ) return 0; + return mb->buffer_size - mb->stream_pos; } + +static inline bool +Mb_enough_available_bytes( const struct Matchfinder_base * const mb ) + { return ( mb->pos + mb->after_size <= mb->stream_pos || + ( Mb_flushing_or_end( mb ) && mb->pos < mb->stream_pos ) ); } + +static inline const uint8_t * +Mb_ptr_to_current_pos( const struct Matchfinder_base * const mb ) + { return mb->buffer + mb->pos; } + +static int Mb_write_data( struct Matchfinder_base * const mb, + const uint8_t * const inbuf, const int size ) + { + const int sz = min( mb->buffer_size - mb->stream_pos, size ); + if( Mb_flushing_or_end( mb ) || sz <= 0 ) return 0; + memcpy( mb->buffer + mb->stream_pos, inbuf, sz ); + mb->stream_pos += sz; + return sz; + } + +static inline int Mb_true_match_len( const struct Matchfinder_base * const mb, + const int index, const int distance ) + { + const uint8_t * const data = mb->buffer + mb->pos; + int i = index; + const int len_limit = min( Mb_available_bytes( mb ), max_match_len ); + while( i < len_limit && data[i-distance] == data[i] ) ++i; + return i; + } + +static inline bool Mb_move_pos( struct Matchfinder_base * const mb ) + { + if( ++mb->cyclic_pos > mb->dictionary_size ) mb->cyclic_pos = 0; + if( ++mb->pos >= mb->pos_limit ) return Mb_normalize_pos( mb ); + return true; + } + + +struct Range_encoder + { + struct Circular_buffer cb; + unsigned min_free_bytes; + uint64_t low; + unsigned long long partial_member_pos; + uint32_t range; + unsigned ff_count; + uint8_t cache; + Lzip_header header; + }; + +static inline void Re_shift_low( struct Range_encoder * const renc ) + { + if( renc->low >> 24 != 0xFF ) + { + const bool carry = ( renc->low > 0xFFFFFFFFU ); + Cb_put_byte( &renc->cb, renc->cache + carry ); + for( ; renc->ff_count > 0; --renc->ff_count ) + Cb_put_byte( &renc->cb, 0xFF + carry ); + renc->cache = renc->low >> 24; + } + else ++renc->ff_count; + renc->low = ( renc->low & 0x00FFFFFFU ) << 8; + } + +static inline void Re_reset( struct Range_encoder * const renc, + const unsigned dictionary_size ) + { + int i; + Cb_reset( &renc->cb ); + renc->low = 0; + renc->partial_member_pos = 0; + renc->range = 0xFFFFFFFFU; + renc->ff_count = 0; + renc->cache = 0; + Lh_set_dictionary_size( renc->header, dictionary_size ); + for( i = 0; i < Lh_size; ++i ) + Cb_put_byte( &renc->cb, renc->header[i] ); + } + +static inline bool Re_init( struct Range_encoder * const renc, + const unsigned dictionary_size, + const unsigned min_free_bytes ) + { + if( !Cb_init( &renc->cb, 65536 + min_free_bytes ) ) return false; + renc->min_free_bytes = min_free_bytes; + Lh_set_magic( renc->header ); + Re_reset( renc, dictionary_size ); + return true; + } + +static inline void Re_free( struct Range_encoder * const renc ) + { Cb_free( &renc->cb ); } + +static inline unsigned long long +Re_member_position( const struct Range_encoder * const renc ) + { return renc->partial_member_pos + Cb_used_bytes( &renc->cb ) + renc->ff_count; } + +static inline bool Re_enough_free_bytes( const struct Range_encoder * const renc ) + { return Cb_free_bytes( &renc->cb ) >= renc->min_free_bytes + renc->ff_count; } + +static inline int Re_read_data( struct Range_encoder * const renc, + uint8_t * const out_buffer, const int out_size ) + { + const int size = Cb_read_data( &renc->cb, out_buffer, out_size ); + if( size > 0 ) renc->partial_member_pos += size; + return size; + } + +static inline void Re_flush( struct Range_encoder * const renc ) + { + int i; for( i = 0; i < 5; ++i ) Re_shift_low( renc ); + renc->low = 0; + renc->range = 0xFFFFFFFFU; + renc->ff_count = 0; + renc->cache = 0; + } + +static inline void Re_encode( struct Range_encoder * const renc, + const int symbol, const int num_bits ) + { + unsigned mask; + for( mask = 1 << ( num_bits - 1 ); mask > 0; mask >>= 1 ) + { + renc->range >>= 1; + if( symbol & mask ) renc->low += renc->range; + if( renc->range <= 0x00FFFFFFU ) + { renc->range <<= 8; Re_shift_low( renc ); } + } + } + +static inline void Re_encode_bit( struct Range_encoder * const renc, + Bit_model * const probability, const bool bit ) + { + const uint32_t bound = ( renc->range >> bit_model_total_bits ) * *probability; + if( !bit ) + { + renc->range = bound; + *probability += (bit_model_total - *probability) >> bit_model_move_bits; + } + else + { + renc->low += bound; + renc->range -= bound; + *probability -= *probability >> bit_model_move_bits; + } + if( renc->range <= 0x00FFFFFFU ) { renc->range <<= 8; Re_shift_low( renc ); } + } + +static inline void Re_encode_tree3( struct Range_encoder * const renc, + Bit_model bm[], const int symbol ) + { + bool bit = ( symbol >> 2 ) & 1; + Re_encode_bit( renc, &bm[1], bit ); + int model = 2 | bit; + bit = ( symbol >> 1 ) & 1; + Re_encode_bit( renc, &bm[model], bit ); model <<= 1; model |= bit; + Re_encode_bit( renc, &bm[model], symbol & 1 ); + } + +static inline void Re_encode_tree6( struct Range_encoder * const renc, + Bit_model bm[], const unsigned symbol ) + { + bool bit = ( symbol >> 5 ) & 1; + Re_encode_bit( renc, &bm[1], bit ); + int model = 2 | bit; + bit = ( symbol >> 4 ) & 1; + Re_encode_bit( renc, &bm[model], bit ); model <<= 1; model |= bit; + bit = ( symbol >> 3 ) & 1; + Re_encode_bit( renc, &bm[model], bit ); model <<= 1; model |= bit; + bit = ( symbol >> 2 ) & 1; + Re_encode_bit( renc, &bm[model], bit ); model <<= 1; model |= bit; + bit = ( symbol >> 1 ) & 1; + Re_encode_bit( renc, &bm[model], bit ); model <<= 1; model |= bit; + Re_encode_bit( renc, &bm[model], symbol & 1 ); + } + +static inline void Re_encode_tree8( struct Range_encoder * const renc, + Bit_model bm[], const int symbol ) + { + int model = 1; + int i; + for( i = 7; i >= 0; --i ) + { + const bool bit = ( symbol >> i ) & 1; + Re_encode_bit( renc, &bm[model], bit ); + model <<= 1; model |= bit; + } + } + +static inline void Re_encode_tree_reversed( struct Range_encoder * const renc, + Bit_model bm[], int symbol, const int num_bits ) + { + int model = 1; + int i; + for( i = num_bits; i > 0; --i ) + { + const bool bit = symbol & 1; + symbol >>= 1; + Re_encode_bit( renc, &bm[model], bit ); + model <<= 1; model |= bit; + } + } + +static inline void Re_encode_matched( struct Range_encoder * const renc, + Bit_model bm[], unsigned symbol, + unsigned match_byte ) + { + unsigned mask = 0x100; + symbol |= mask; + while( true ) + { + const unsigned match_bit = ( match_byte <<= 1 ) & mask; + const bool bit = ( symbol <<= 1 ) & 0x100; + Re_encode_bit( renc, &bm[(symbol>>9)+match_bit+mask], bit ); + if( symbol >= 0x10000 ) break; + mask &= ~(match_bit ^ symbol); /* if( match_bit != bit ) mask = 0; */ + } + } + +static inline void Re_encode_len( struct Range_encoder * const renc, + struct Len_model * const lm, + int symbol, const int pos_state ) + { + bool bit = ( ( symbol -= min_match_len ) >= len_low_symbols ); + Re_encode_bit( renc, &lm->choice1, bit ); + if( !bit ) + Re_encode_tree3( renc, lm->bm_low[pos_state], symbol ); + else + { + bit = ( ( symbol -= len_low_symbols ) >= len_mid_symbols ); + Re_encode_bit( renc, &lm->choice2, bit ); + if( !bit ) + Re_encode_tree3( renc, lm->bm_mid[pos_state], symbol ); + else + Re_encode_tree8( renc, lm->bm_high, symbol - len_mid_symbols ); + } + } + + +enum { max_marker_size = 16, + num_rep_distances = 4 }; /* must be 4 */ + +struct LZ_encoder_base + { + struct Matchfinder_base mb; + unsigned long long member_size_limit; + uint32_t crc; + + Bit_model bm_literal[1<mb, before_size, dict_size, after_size, dict_factor, + num_prev_positions23, pos_array_factor ) ) return false; + if( !Re_init( &eb->renc, eb->mb.dictionary_size, min_free_bytes ) ) + return false; + LZeb_reset( eb, member_size ); + return true; + } + +static inline bool LZeb_member_finished( const struct LZ_encoder_base * const eb ) + { return ( eb->member_finished && Cb_empty( &eb->renc.cb ) ); } + +static inline void LZeb_free( struct LZ_encoder_base * const eb ) + { Re_free( &eb->renc ); Mb_free( &eb->mb ); } + +static inline unsigned LZeb_crc( const struct LZ_encoder_base * const eb ) + { return eb->crc ^ 0xFFFFFFFFU; } + +static inline int LZeb_price_literal( const struct LZ_encoder_base * const eb, + const uint8_t prev_byte, const uint8_t symbol ) + { return price_symbol8( eb->bm_literal[get_lit_state(prev_byte)], symbol ); } + +static inline int LZeb_price_matched( const struct LZ_encoder_base * const eb, + const uint8_t prev_byte, const uint8_t symbol, const uint8_t match_byte ) + { return price_matched( eb->bm_literal[get_lit_state(prev_byte)], symbol, + match_byte ); } + +static inline void LZeb_encode_literal( struct LZ_encoder_base * const eb, + const uint8_t prev_byte, const uint8_t symbol ) + { Re_encode_tree8( &eb->renc, eb->bm_literal[get_lit_state(prev_byte)], symbol ); } + +static inline void LZeb_encode_matched( struct LZ_encoder_base * const eb, + const uint8_t prev_byte, const uint8_t symbol, const uint8_t match_byte ) + { Re_encode_matched( &eb->renc, eb->bm_literal[get_lit_state(prev_byte)], + symbol, match_byte ); } + +static inline void LZeb_encode_pair( struct LZ_encoder_base * const eb, + const unsigned dis, const int len, + const int pos_state ) + { + Re_encode_len( &eb->renc, &eb->match_len_model, len, pos_state ); + const unsigned dis_slot = get_slot( dis ); + Re_encode_tree6( &eb->renc, eb->bm_dis_slot[get_len_state(len)], dis_slot ); + + if( dis_slot >= start_dis_model ) + { + const int direct_bits = ( dis_slot >> 1 ) - 1; + const unsigned base = ( 2 | ( dis_slot & 1 ) ) << direct_bits; + const unsigned direct_dis = dis - base; + + if( dis_slot < end_dis_model ) + Re_encode_tree_reversed( &eb->renc, eb->bm_dis + ( base - dis_slot ), + direct_dis, direct_bits ); + else + { + Re_encode( &eb->renc, direct_dis >> dis_align_bits, + direct_bits - dis_align_bits ); + Re_encode_tree_reversed( &eb->renc, eb->bm_align, direct_dis, dis_align_bits ); + } + } + } diff --git a/fast_encoder.c b/fast_encoder.c new file mode 100644 index 0000000..618c3d6 --- /dev/null +++ b/fast_encoder.c @@ -0,0 +1,175 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +static int FLZe_longest_match_len( struct FLZ_encoder * const fe, int * const distance ) + { + enum { len_limit = 16 }; + int32_t * ptr0 = fe->eb.mb.pos_array + fe->eb.mb.cyclic_pos; + const int available = min( Mb_available_bytes( &fe->eb.mb ), max_match_len ); + if( available < len_limit ) { *ptr0 = 0; return 0; } + + const uint8_t * const data = Mb_ptr_to_current_pos( &fe->eb.mb ); + fe->key4 = ( ( fe->key4 << 4 ) ^ data[3] ) & fe->eb.mb.key4_mask; + const int pos1 = fe->eb.mb.pos + 1; + int newpos1 = fe->eb.mb.prev_positions[fe->key4]; + fe->eb.mb.prev_positions[fe->key4] = pos1; + int maxlen = 0, count; + + for( count = 4; ; ) + { + int delta; + if( newpos1 <= 0 || --count < 0 || + ( delta = pos1 - newpos1 ) > fe->eb.mb.dictionary_size ) + { *ptr0 = 0; break; } + int32_t * const newptr = fe->eb.mb.pos_array + + ( fe->eb.mb.cyclic_pos - delta + + ( ( fe->eb.mb.cyclic_pos >= delta ) ? 0 : fe->eb.mb.dictionary_size + 1 ) ); + + if( data[maxlen-delta] == data[maxlen] ) + { + int len = 0; + while( len < available && data[len-delta] == data[len] ) ++len; + if( maxlen < len ) + { maxlen = len; *distance = delta - 1; + if( maxlen >= len_limit ) { *ptr0 = *newptr; break; } } + } + + *ptr0 = newpos1; + ptr0 = newptr; + newpos1 = *ptr0; + } + return maxlen; + } + + +static bool FLZe_encode_member( struct FLZ_encoder * const fe ) + { + int rep = 0, i; + State * const state = &fe->eb.state; + + if( fe->eb.member_finished ) return true; + if( Re_member_position( &fe->eb.renc ) >= fe->eb.member_size_limit ) + { LZeb_try_full_flush( &fe->eb ); return true; } + + if( Mb_data_position( &fe->eb.mb ) == 0 && + !Mb_data_finished( &fe->eb.mb ) ) /* encode first byte */ + { + if( !Mb_enough_available_bytes( &fe->eb.mb ) || + !Re_enough_free_bytes( &fe->eb.renc ) ) return true; + const uint8_t prev_byte = 0; + const uint8_t cur_byte = Mb_peek( &fe->eb.mb, 0 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_match[*state][0], 0 ); + LZeb_encode_literal( &fe->eb, prev_byte, cur_byte ); + CRC32_update_byte( &fe->eb.crc, cur_byte ); + FLZe_reset_key4( fe ); + if( !FLZe_update_and_move( fe, 1 ) ) return false; + } + + while( !Mb_data_finished( &fe->eb.mb ) && + Re_member_position( &fe->eb.renc ) < fe->eb.member_size_limit ) + { + if( !Mb_enough_available_bytes( &fe->eb.mb ) || + !Re_enough_free_bytes( &fe->eb.renc ) ) return true; + int match_distance = 0; /* avoid warning from gcc 6.1.0 */ + const int main_len = FLZe_longest_match_len( fe, &match_distance ); + const int pos_state = Mb_data_position( &fe->eb.mb ) & pos_state_mask; + int len = 0; + + for( i = 0; i < num_rep_distances; ++i ) + { + const int tlen = Mb_true_match_len( &fe->eb.mb, 0, fe->eb.reps[i] + 1 ); + if( tlen > len ) { len = tlen; rep = i; } + } + if( len > min_match_len && len + 3 > main_len ) + { + CRC32_update_buf( &fe->eb.crc, Mb_ptr_to_current_pos( &fe->eb.mb ), len ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_match[*state][pos_state], 1 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep[*state], 1 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep0[*state], rep != 0 ); + if( rep == 0 ) + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_len[*state][pos_state], 1 ); + else + { + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep1[*state], rep > 1 ); + if( rep > 1 ) + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep2[*state], rep > 2 ); + const int distance = fe->eb.reps[rep]; + for( i = rep; i > 0; --i ) fe->eb.reps[i] = fe->eb.reps[i-1]; + fe->eb.reps[0] = distance; + } + *state = St_set_rep( *state ); + Re_encode_len( &fe->eb.renc, &fe->eb.rep_len_model, len, pos_state ); + if( !Mb_move_pos( &fe->eb.mb ) ) return false; + if( !FLZe_update_and_move( fe, len - 1 ) ) return false; + continue; + } + + if( main_len > min_match_len ) + { + CRC32_update_buf( &fe->eb.crc, Mb_ptr_to_current_pos( &fe->eb.mb ), main_len ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_match[*state][pos_state], 1 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep[*state], 0 ); + *state = St_set_match( *state ); + for( i = num_rep_distances - 1; i > 0; --i ) fe->eb.reps[i] = fe->eb.reps[i-1]; + fe->eb.reps[0] = match_distance; + LZeb_encode_pair( &fe->eb, match_distance, main_len, pos_state ); + if( !Mb_move_pos( &fe->eb.mb ) ) return false; + if( !FLZe_update_and_move( fe, main_len - 1 ) ) return false; + continue; + } + + const uint8_t prev_byte = Mb_peek( &fe->eb.mb, 1 ); + const uint8_t cur_byte = Mb_peek( &fe->eb.mb, 0 ); + const uint8_t match_byte = Mb_peek( &fe->eb.mb, fe->eb.reps[0] + 1 ); + if( !Mb_move_pos( &fe->eb.mb ) ) return false; + CRC32_update_byte( &fe->eb.crc, cur_byte ); + + if( match_byte == cur_byte ) + { + const int short_rep_price = price1( fe->eb.bm_match[*state][pos_state] ) + + price1( fe->eb.bm_rep[*state] ) + + price0( fe->eb.bm_rep0[*state] ) + + price0( fe->eb.bm_len[*state][pos_state] ); + int price = price0( fe->eb.bm_match[*state][pos_state] ); + if( St_is_char( *state ) ) + price += LZeb_price_literal( &fe->eb, prev_byte, cur_byte ); + else + price += LZeb_price_matched( &fe->eb, prev_byte, cur_byte, match_byte ); + if( short_rep_price < price ) + { + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_match[*state][pos_state], 1 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep[*state], 1 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_rep0[*state], 0 ); + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_len[*state][pos_state], 0 ); + *state = St_set_short_rep( *state ); + continue; + } + } + + /* literal byte */ + Re_encode_bit( &fe->eb.renc, &fe->eb.bm_match[*state][pos_state], 0 ); + if( ( *state = St_set_char( *state ) ) < 4 ) + LZeb_encode_literal( &fe->eb, prev_byte, cur_byte ); + else + LZeb_encode_matched( &fe->eb, prev_byte, cur_byte, match_byte ); + } + + LZeb_try_full_flush( &fe->eb ); + return true; + } diff --git a/fast_encoder.h b/fast_encoder.h new file mode 100644 index 0000000..54756bd --- /dev/null +++ b/fast_encoder.h @@ -0,0 +1,70 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +struct FLZ_encoder + { + struct LZ_encoder_base eb; + unsigned key4; /* key made from latest 4 bytes */ + }; + +static inline void FLZe_reset_key4( struct FLZ_encoder * const fe ) + { + int i; + fe->key4 = 0; + for( i = 0; i < 3 && i < Mb_available_bytes( &fe->eb.mb ); ++i ) + fe->key4 = ( fe->key4 << 4 ) ^ fe->eb.mb.buffer[i]; + } + +static inline bool FLZe_update_and_move( struct FLZ_encoder * const fe, int n ) + { + struct Matchfinder_base * const mb = &fe->eb.mb; + while( --n >= 0 ) + { + if( Mb_available_bytes( mb ) >= 4 ) + { + fe->key4 = ( ( fe->key4 << 4 ) ^ mb->buffer[mb->pos+3] ) & mb->key4_mask; + mb->pos_array[mb->cyclic_pos] = mb->prev_positions[fe->key4]; + mb->prev_positions[fe->key4] = mb->pos + 1; + } + else mb->pos_array[mb->cyclic_pos] = 0; + if( !Mb_move_pos( mb ) ) return false; + } + return true; + } + +static inline bool FLZe_init( struct FLZ_encoder * const fe, + const unsigned long long member_size ) + { + enum { before_size = 0, + dict_size = 65536, + /* bytes to keep in buffer after pos */ + after_size = max_match_len, + dict_factor = 16, + min_free_bytes = max_marker_size, + num_prev_positions23 = 0, + pos_array_factor = 1 }; + + return LZeb_init( &fe->eb, before_size, dict_size, after_size, dict_factor, + num_prev_positions23, pos_array_factor, min_free_bytes, + member_size ); + } + +static inline void FLZe_reset( struct FLZ_encoder * const fe, + const unsigned long long member_size ) + { LZeb_reset( &fe->eb, member_size ); } diff --git a/ffexample.c b/ffexample.c new file mode 100644 index 0000000..59345ee --- /dev/null +++ b/ffexample.c @@ -0,0 +1,300 @@ +/* File to file example - Test program for the library lzlib + Copyright (C) 2010-2022 Antonio Diaz Diaz. + + This program is free software: you have unlimited permission + to copy, distribute, and modify it. + + Try 'ffexample -h' for usage information. + + This program is an example of how file-to-file + compression/decompression can be implemented using lzlib. +*/ + +#define _FILE_OFFSET_BITS 64 + +#include +#include +#include +#include +#include +#include +#include +#include +#if defined __MSVCRT__ || defined __OS2__ || defined __DJGPP__ +#include +#include +#endif + +#include "lzlib.h" + +#ifndef min + #define min(x,y) ((x) <= (y) ? (x) : (y)) +#endif + + +static void show_help( void ) + { + printf( "ffexample is an example program showing how file-to-file (de)compression can\n" + "be implemented using lzlib. The content of infile is compressed,\n" + "decompressed, or both, and then written to outfile.\n" + "\nUsage: ffexample operation [infile [outfile]]\n" ); + printf( "\nOperation:\n" + " -h display this help and exit\n" + " -c compress infile to outfile\n" + " -d decompress infile to outfile\n" + " -b both (compress then decompress) infile to outfile\n" + " -m compress (multimember) infile to outfile\n" + " -l compress (1 member per line) infile to outfile\n" + " -r decompress with resync if data error or leading garbage\n" + "\nIf infile or outfile are omitted, or are specified as '-', standard input or\n" + "standard output are used in their place respectively.\n" + "\nReport bugs to lzip-bug@nongnu.org\n" + "Lzlib home page: http://www.nongnu.org/lzip/lzlib.html\n" ); + } + + +int ffcompress( struct LZ_Encoder * const encoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_finished( encoder ) == 1 ) return 0; + } + return 1; + } + + +int ffdecompress( struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_decompress_write( decoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } + + +int ffboth( struct LZ_Encoder * const encoder, + struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + ret = LZ_compress_read( encoder, buffer, size ); + if( ret < 0 ) break; + ret = LZ_decompress_write( decoder, buffer, ret ); + if( ret < 0 ) break; + if( LZ_compress_finished( encoder ) == 1 ) + LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } + + +int ffmmcompress( FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384, member_size = 4096 }; + uint8_t buffer[buffer_size]; + bool done = false; + struct LZ_Encoder * const encoder = + LZ_compress_open( 65535, 16, member_size ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { fputs( "ffexample: Not enough memory.\n", stderr ); + LZ_compress_close( encoder ); return 1; } + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_member_finished( encoder ) == 1 ) + { + if( LZ_compress_finished( encoder ) == 1 ) { done = true; break; } + if( LZ_compress_restart_member( encoder, member_size ) < 0 ) break; + } + } + if( LZ_compress_close( encoder ) < 0 ) done = false; + return done; + } + + +/* Compress 'infile' to 'outfile' as a multimember stream with one member + for each line of text terminated by a newline character or by EOF. + Return 0 if success, 1 if error. +*/ +int fflfcompress( struct LZ_Encoder * const encoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_compress_write_size( encoder ) ); + if( size > 0 ) + { + for( len = 0; len < size; ) + { + int ch = getc( infile ); + if( ch == EOF || ( buffer[len++] = ch ) == '\n' ) break; + } + /* avoid writing an empty member to outfile */ + if( len == 0 && LZ_compress_data_position( encoder ) == 0 ) return 0; + ret = LZ_compress_write( encoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) || buffer[len-1] == '\n' ) + LZ_compress_finish( encoder ); + } + ret = LZ_compress_read( encoder, buffer, buffer_size ); + if( ret < 0 ) break; + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_compress_member_finished( encoder ) == 1 ) + { + if( feof( infile ) && LZ_compress_finished( encoder ) == 1 ) return 0; + if( LZ_compress_restart_member( encoder, INT64_MAX ) < 0 ) break; + } + } + return 1; + } + + +/* Decompress 'infile' to 'outfile' with automatic resynchronization to + next member in case of data error, including the automatic removal of + leading garbage. +*/ +int ffrsdecompress( struct LZ_Decoder * const decoder, + FILE * const infile, FILE * const outfile ) + { + enum { buffer_size = 16384 }; + uint8_t buffer[buffer_size]; + while( true ) + { + int len, ret; + int size = min( buffer_size, LZ_decompress_write_size( decoder ) ); + if( size > 0 ) + { + len = fread( buffer, 1, size, infile ); + ret = LZ_decompress_write( decoder, buffer, len ); + if( ret < 0 || ferror( infile ) ) break; + if( feof( infile ) ) LZ_decompress_finish( decoder ); + } + ret = LZ_decompress_read( decoder, buffer, buffer_size ); + if( ret < 0 ) + { + if( LZ_decompress_errno( decoder ) == LZ_header_error || + LZ_decompress_errno( decoder ) == LZ_data_error ) + { LZ_decompress_sync_to_member( decoder ); continue; } + else break; + } + len = fwrite( buffer, 1, ret, outfile ); + if( len < ret ) break; + if( LZ_decompress_finished( decoder ) == 1 ) return 0; + } + return 1; + } + + +int main( const int argc, const char * const argv[] ) + { +#if defined __MSVCRT__ || defined __OS2__ || defined __DJGPP__ + setmode( STDIN_FILENO, O_BINARY ); + setmode( STDOUT_FILENO, O_BINARY ); +#endif + + struct LZ_Encoder * const encoder = LZ_compress_open( 65535, 16, INT64_MAX ); + struct LZ_Decoder * const decoder = LZ_decompress_open(); + FILE * const infile = ( argc >= 3 && strcmp( argv[2], "-" ) != 0 ) ? + fopen( argv[2], "rb" ) : stdin; + FILE * const outfile = ( argc >= 4 && strcmp( argv[3], "-" ) != 0 ) ? + fopen( argv[3], "wb" ) : stdout; + int retval; + + if( argc < 2 || argc > 4 || strlen( argv[1] ) != 2 || argv[1][0] != '-' ) + { show_help(); return 1; } + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok || + !decoder || LZ_decompress_errno( decoder ) != LZ_ok ) + { fputs( "ffexample: Not enough memory.\n", stderr ); + LZ_compress_close( encoder ); LZ_decompress_close( decoder ); return 1; } + if( !infile ) + { fprintf( stderr, "ffexample: Can't open input file '%s': %s\n", + argv[2], strerror( errno ) ); return 1; } + if( !outfile ) + { fprintf( stderr, "ffexample: Can't open output file '%s': %s\n", + argv[3], strerror( errno ) ); return 1; } + + switch( argv[1][1] ) + { + case 'c': retval = ffcompress( encoder, infile, outfile ); break; + case 'd': retval = ffdecompress( decoder, infile, outfile ); break; + case 'b': retval = ffboth( encoder, decoder, infile, outfile ); break; + case 'm': retval = ffmmcompress( infile, outfile ); break; + case 'l': retval = fflfcompress( encoder, infile, outfile ); break; + case 'r': retval = ffrsdecompress( decoder, infile, outfile ); break; + default: show_help(); return ( argv[1][1] != 'h' ); + } + + if( LZ_decompress_close( decoder ) < 0 || LZ_compress_close( encoder ) < 0 || + fclose( outfile ) != 0 || fclose( infile ) != 0 ) retval = 1; + return retval; + } diff --git a/lzcheck.c b/lzcheck.c new file mode 100644 index 0000000..88dd4c9 --- /dev/null +++ b/lzcheck.c @@ -0,0 +1,367 @@ +/* Lzcheck - Test program for the library lzlib + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This program is free software: you have unlimited permission + to copy, distribute, and modify it. + + Usage: lzcheck [-m|-s] filename.txt... + + This program reads each text file specified and then compresses it, + line by line, to test the flushing mechanism and the member + restart/reset/sync functions. +*/ + +#define _FILE_OFFSET_BITS 64 + +#include +#include +#include +#include +#include +#include +#include +#include + +#include "lzlib.h" + + +const unsigned long long member_size = INT64_MAX; +enum { buffer_size = 32768 }; +uint8_t in_buffer[buffer_size]; +uint8_t mid_buffer[buffer_size]; +uint8_t out_buffer[buffer_size]; + + +static void show_line( const uint8_t * const buffer, const int size ) + { + int i; + for( i = 0; i < size; ++i ) + fputc( isprint( buffer[i] ) ? buffer[i] : '.', stderr ); + fputc( '\n', stderr ); + } + + +static struct LZ_Encoder * xopen_encoder( const int dictionary_size ) + { + const int match_len_limit = 16; + struct LZ_Encoder * const encoder = + LZ_compress_open( dictionary_size, match_len_limit, member_size ); + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { + const bool bad_arg = + encoder && ( LZ_compress_errno( encoder ) == LZ_bad_argument ); + LZ_compress_close( encoder ); + if( bad_arg ) + { + fputs( "lzcheck: internal error: Invalid argument to encoder.\n", stderr ); + exit( 3 ); + } + fputs( "lzcheck: Not enough memory.\n", stderr ); + exit( 1 ); + } + return encoder; + } + +static struct LZ_Decoder * xopen_decoder( void ) + { + struct LZ_Decoder * const decoder = LZ_decompress_open(); + if( !decoder || LZ_decompress_errno( decoder ) != LZ_ok ) + { + LZ_decompress_close( decoder ); + fputs( "lzcheck: Not enough memory.\n", stderr ); + exit( 1 ); + } + return decoder; + } + + +static void xclose_encoder( struct LZ_Encoder * const encoder, + const bool finish ) + { + if( finish ) + { + unsigned long long size = 0; + LZ_compress_finish( encoder ); + while( true ) + { + const int rd = LZ_compress_read( encoder, mid_buffer, buffer_size ); + if( rd < 0 ) + { + fprintf( stderr, "lzcheck: xclose: LZ_compress_read error: %s\n", + LZ_strerror( LZ_compress_errno( encoder ) ) ); + exit( 3 ); + } + size += rd; + if( LZ_compress_finished( encoder ) == 1 ) break; + } + if( size > 0 ) + { + fprintf( stderr, "lzcheck: %lld bytes remain in encoder.\n", size ); + exit( 3 ); + } + } + if( LZ_compress_close( encoder ) < 0 ) exit( 1 ); + } + + +static void xclose_decoder( struct LZ_Decoder * const decoder, + const bool finish ) + { + if( finish ) + { + unsigned long long size = 0; + LZ_decompress_finish( decoder ); + while( true ) + { + const int rd = LZ_decompress_read( decoder, out_buffer, buffer_size ); + if( rd < 0 ) + { + fprintf( stderr, "lzcheck: xclose: LZ_decompress_read error: %s\n", + LZ_strerror( LZ_decompress_errno( decoder ) ) ); + exit( 3 ); + } + size += rd; + if( LZ_decompress_finished( decoder ) == 1 ) break; + } + if( size > 0 ) + { + fprintf( stderr, "lzcheck: %lld bytes remain in decoder.\n", size ); + exit( 3 ); + } + } + if( LZ_decompress_close( decoder ) < 0 ) exit( 1 ); + } + + +/* Return the next (usually newline-terminated) chunk of data from file. + The size returned in *sizep is always <= buffer_size. + If sizep is a null pointer, rewind the file, reset state, and return. + If file is at EOF, return an empty line. +*/ +static const uint8_t * next_line( FILE * const file, int * const sizep ) + { + static int l = 0; + static int read_size = 0; + int r; + + if( !sizep ) { rewind( file ); l = read_size = 0; return in_buffer; } + if( l >= read_size ) + { + l = 0; read_size = fread( in_buffer, 1, buffer_size, file ); + if( l >= read_size ) { *sizep = 0; return in_buffer; } /* end of file */ + } + + for( r = l + 1; r < read_size && in_buffer[r-1] != '\n'; ++r ); + *sizep = r - l; l = r; + return in_buffer + l - *sizep; + } + + +static int check_sync_flush( FILE * const file, const int dictionary_size ) + { + struct LZ_Encoder * const encoder = xopen_encoder( dictionary_size ); + struct LZ_Decoder * const decoder = xopen_decoder(); + int retval = 0; + + while( retval <= 1 ) /* test LZ_compress_sync_flush */ + { + int in_size, mid_size, out_size; + int line_size; + const uint8_t * const line_buf = next_line( file, &line_size ); + if( line_size <= 0 ) break; /* end of file */ + + in_size = LZ_compress_write( encoder, line_buf, line_size ); + if( in_size < line_size ) + fprintf( stderr, "lzcheck: sync: LZ_compress_write only accepted %d of %d bytes\n", + in_size, line_size ); + LZ_compress_sync_flush( encoder ); + if( line_buf[0] & 1 ) /* read all data at once or byte by byte */ + mid_size = LZ_compress_read( encoder, mid_buffer, buffer_size ); + else for( mid_size = 0; mid_size < buffer_size; ) + { + const int rd = LZ_compress_read( encoder, mid_buffer + mid_size, 1 ); + if( rd > 0 ) mid_size += rd; + else { if( rd < 0 ) { mid_size = -1; } break; } + } + if( mid_size < 0 ) + { + fprintf( stderr, "lzcheck: LZ_compress_read error: %s\n", + LZ_strerror( LZ_compress_errno( encoder ) ) ); + retval = 3; break; + } + LZ_decompress_write( decoder, mid_buffer, mid_size ); + out_size = LZ_decompress_read( decoder, out_buffer, buffer_size ); + if( out_size < 0 ) + { + fprintf( stderr, "lzcheck: LZ_decompress_read error: %s\n", + LZ_strerror( LZ_decompress_errno( decoder ) ) ); + retval = 3; break; + } + + if( out_size != in_size || memcmp( line_buf, out_buffer, out_size ) ) + { + fprintf( stderr, "lzcheck: LZ_compress_sync_flush error: " + "in_size = %d, out_size = %d\n", in_size, out_size ); + show_line( line_buf, in_size ); + show_line( out_buffer, out_size ); + retval = 1; + } + } + + if( retval <= 1 ) + { + int rd = 0; + if( LZ_compress_finish( encoder ) < 0 || + ( rd = LZ_compress_read( encoder, mid_buffer, buffer_size ) ) < 0 ) + { + fprintf( stderr, "lzcheck: Can't drain encoder: %s\n", + LZ_strerror( LZ_compress_errno( encoder ) ) ); + retval = 3; + } + LZ_decompress_write( decoder, mid_buffer, rd ); + } + + xclose_decoder( decoder, retval == 0 ); + xclose_encoder( encoder, retval == 0 ); + return retval; + } + + +/* Test member by member decompression without calling LZ_decompress_finish, + inserting leading garbage before some members, and resetting the + decompressor sometimes. Test that the increase in total_in_size when + syncing to member is equal to the size of the leading garbage skipped. +*/ +static int check_members( FILE * const file, const int dictionary_size ) + { + struct LZ_Encoder * const encoder = xopen_encoder( dictionary_size ); + struct LZ_Decoder * const decoder = xopen_decoder(); + int retval = 0; + + while( retval <= 1 ) /* test LZ_compress_restart_member */ + { + unsigned long long garbage_begin = 0; /* avoid warning from gcc 3.3.6 */ + int leading_garbage, in_size, mid_size, out_size; + int line_size; + const uint8_t * const line_buf = next_line( file, &line_size ); + if( line_size <= 0 && /* end of file, write at least 1 member */ + LZ_decompress_total_in_size( decoder ) != 0 ) break; + + if( LZ_compress_finished( encoder ) == 1 ) + { + if( LZ_compress_restart_member( encoder, member_size ) < 0 ) + { + fprintf( stderr, "lzcheck: Can't restart member: %s\n", + LZ_strerror( LZ_compress_errno( encoder ) ) ); + retval = 3; break; + } + if( line_size >= 2 && line_buf[1] == 'h' ) + LZ_decompress_reset( decoder ); + } + in_size = LZ_compress_write( encoder, line_buf, line_size ); + if( in_size < line_size ) + fprintf( stderr, "lzcheck: member: LZ_compress_write only accepted %d of %d bytes\n", + in_size, line_size ); + LZ_compress_finish( encoder ); + if( line_size * 3 < buffer_size && line_buf[0] == 't' ) + { leading_garbage = line_size; + memset( mid_buffer, in_buffer[0], leading_garbage ); + garbage_begin = LZ_decompress_total_in_size( decoder ); } + else leading_garbage = 0; + mid_size = LZ_compress_read( encoder, mid_buffer + leading_garbage, + buffer_size - leading_garbage ); + if( mid_size < 0 ) + { + fprintf( stderr, "lzcheck: member: LZ_compress_read error: %s\n", + LZ_strerror( LZ_compress_errno( encoder ) ) ); + retval = 3; break; + } + LZ_decompress_write( decoder, mid_buffer, leading_garbage + mid_size ); + out_size = LZ_decompress_read( decoder, out_buffer, buffer_size ); + if( out_size < 0 ) + { + if( leading_garbage && + ( LZ_decompress_errno( decoder ) == LZ_header_error || + LZ_decompress_errno( decoder ) == LZ_data_error ) ) + { + LZ_decompress_sync_to_member( decoder ); /* skip leading garbage */ + const unsigned long long garbage_end = + LZ_decompress_total_in_size( decoder ); + if( garbage_end - garbage_begin != (unsigned)leading_garbage ) + { + fprintf( stderr, "lzcheck: member: LZ_decompress_sync_to_member error:\n" + " garbage_begin = %llu garbage_end = %llu " + "difference = %llu expected = %d\n", garbage_begin, + garbage_end, garbage_end - garbage_begin, leading_garbage ); + retval = 3; break; + } + out_size = LZ_decompress_read( decoder, out_buffer, buffer_size ); + } + if( out_size < 0 ) + { + fprintf( stderr, "lzcheck: member: LZ_decompress_read error: %s\n", + LZ_strerror( LZ_decompress_errno( decoder ) ) ); + retval = 3; break; + } + } + + if( out_size != in_size || memcmp( line_buf, out_buffer, out_size ) ) + { + fprintf( stderr, "lzcheck: LZ_compress_restart_member error: " + "in_size = %d, out_size = %d\n", in_size, out_size ); + show_line( line_buf, in_size ); + show_line( out_buffer, out_size ); + retval = 1; + } + } + + xclose_decoder( decoder, retval == 0 ); + xclose_encoder( encoder, retval == 0 ); + return retval; + } + + +int main( const int argc, const char * const argv[] ) + { + int retval = 0, i; + int open_failures = 0; + const char opt = ( argc > 2 && + ( strcmp( argv[1], "-m" ) == 0 || strcmp( argv[1], "-s" ) == 0 ) ) ? + argv[1][1] : 0; + const int first = opt ? 2 : 1; + const bool verbose = ( opt != 0 || argc > first + 1 ); + + if( argc < 2 ) + { + fputs( "Usage: lzcheck [-m|-s] filename.txt...\n", stderr ); + return 1; + } + + for( i = first; i < argc && retval == 0; ++i ) + { + struct stat st; + if( stat( argv[i], &st ) != 0 || !S_ISREG( st.st_mode ) ) continue; + FILE * file = fopen( argv[i], "rb" ); + if( !file ) + { + fprintf( stderr, "lzcheck: Can't open file '%s' for reading.\n", argv[i] ); + ++open_failures; continue; + } + if( verbose ) fprintf( stderr, " Testing file '%s'\n", argv[i] ); + + /* 65535,16 chooses fast encoder */ + if( opt != 'm' ) retval = check_sync_flush( file, 65535 ); + if( retval == 0 && opt != 'm' ) + { next_line( file, 0 ); retval = check_sync_flush( file, 1 << 20 ); } + if( retval == 0 && opt != 's' ) + { next_line( file, 0 ); retval = check_members( file, 65535 ); } + if( retval == 0 && opt != 's' ) + { next_line( file, 0 ); retval = check_members( file, 1 << 20 ); } + fclose( file ); + } + if( open_failures > 0 && verbose ) + fprintf( stderr, "lzcheck: warning: %d %s failed to open.\n", + open_failures, ( open_failures == 1 ) ? "file" : "files" ); + if( retval == 0 && open_failures ) retval = 1; + return retval; + } diff --git a/lzip.h b/lzip.h new file mode 100644 index 0000000..aad7e00 --- /dev/null +++ b/lzip.h @@ -0,0 +1,298 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +#ifndef max + #define max(x,y) ((x) >= (y) ? (x) : (y)) +#endif +#ifndef min + #define min(x,y) ((x) <= (y) ? (x) : (y)) +#endif + +typedef int State; + +enum { states = 12 }; + +static inline bool St_is_char( const State st ) { return st < 7; } + +static inline State St_set_char( const State st ) + { + static const State next[states] = { 0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 4, 5 }; + return next[st]; + } + +static inline State St_set_char_rep() { return 8; } + +static inline State St_set_match( const State st ) + { return ( ( st < 7 ) ? 7 : 10 ); } + +static inline State St_set_rep( const State st ) + { return ( ( st < 7 ) ? 8 : 11 ); } + +static inline State St_set_short_rep( const State st ) + { return ( ( st < 7 ) ? 9 : 11 ); } + + +enum { + min_dictionary_bits = 12, + min_dictionary_size = 1 << min_dictionary_bits, /* >= modeled_distances */ + max_dictionary_bits = 29, + max_dictionary_size = 1 << max_dictionary_bits, + literal_context_bits = 3, + literal_pos_state_bits = 0, /* not used */ + pos_state_bits = 2, + pos_states = 1 << pos_state_bits, + pos_state_mask = pos_states - 1, + + len_states = 4, + dis_slot_bits = 6, + start_dis_model = 4, + end_dis_model = 14, + modeled_distances = 1 << (end_dis_model / 2), /* 128 */ + dis_align_bits = 4, + dis_align_size = 1 << dis_align_bits, + + len_low_bits = 3, + len_mid_bits = 3, + len_high_bits = 8, + len_low_symbols = 1 << len_low_bits, + len_mid_symbols = 1 << len_mid_bits, + len_high_symbols = 1 << len_high_bits, + max_len_symbols = len_low_symbols + len_mid_symbols + len_high_symbols, + + min_match_len = 2, /* must be 2 */ + max_match_len = min_match_len + max_len_symbols - 1, /* 273 */ + min_match_len_limit = 5 }; + +static inline int get_len_state( const int len ) + { return min( len - min_match_len, len_states - 1 ); } + +static inline int get_lit_state( const uint8_t prev_byte ) + { return prev_byte >> ( 8 - literal_context_bits ); } + + +enum { bit_model_move_bits = 5, + bit_model_total_bits = 11, + bit_model_total = 1 << bit_model_total_bits }; + +typedef int Bit_model; + +static inline void Bm_init( Bit_model * const probability ) + { *probability = bit_model_total / 2; } + +static inline void Bm_array_init( Bit_model bm[], const int size ) + { int i; for( i = 0; i < size; ++i ) Bm_init( &bm[i] ); } + +struct Len_model + { + Bit_model choice1; + Bit_model choice2; + Bit_model bm_low[pos_states][len_low_symbols]; + Bit_model bm_mid[pos_states][len_mid_symbols]; + Bit_model bm_high[len_high_symbols]; + }; + +static inline void Lm_init( struct Len_model * const lm ) + { + Bm_init( &lm->choice1 ); + Bm_init( &lm->choice2 ); + Bm_array_init( lm->bm_low[0], pos_states * len_low_symbols ); + Bm_array_init( lm->bm_mid[0], pos_states * len_mid_symbols ); + Bm_array_init( lm->bm_high, len_high_symbols ); + } + + +/* Table of CRCs of all 8-bit messages. */ +static const uint32_t crc32[256] = + { + 0x00000000, 0x77073096, 0xEE0E612C, 0x990951BA, 0x076DC419, 0x706AF48F, + 0xE963A535, 0x9E6495A3, 0x0EDB8832, 0x79DCB8A4, 0xE0D5E91E, 0x97D2D988, + 0x09B64C2B, 0x7EB17CBD, 0xE7B82D07, 0x90BF1D91, 0x1DB71064, 0x6AB020F2, + 0xF3B97148, 0x84BE41DE, 0x1ADAD47D, 0x6DDDE4EB, 0xF4D4B551, 0x83D385C7, + 0x136C9856, 0x646BA8C0, 0xFD62F97A, 0x8A65C9EC, 0x14015C4F, 0x63066CD9, + 0xFA0F3D63, 0x8D080DF5, 0x3B6E20C8, 0x4C69105E, 0xD56041E4, 0xA2677172, + 0x3C03E4D1, 0x4B04D447, 0xD20D85FD, 0xA50AB56B, 0x35B5A8FA, 0x42B2986C, + 0xDBBBC9D6, 0xACBCF940, 0x32D86CE3, 0x45DF5C75, 0xDCD60DCF, 0xABD13D59, + 0x26D930AC, 0x51DE003A, 0xC8D75180, 0xBFD06116, 0x21B4F4B5, 0x56B3C423, + 0xCFBA9599, 0xB8BDA50F, 0x2802B89E, 0x5F058808, 0xC60CD9B2, 0xB10BE924, + 0x2F6F7C87, 0x58684C11, 0xC1611DAB, 0xB6662D3D, 0x76DC4190, 0x01DB7106, + 0x98D220BC, 0xEFD5102A, 0x71B18589, 0x06B6B51F, 0x9FBFE4A5, 0xE8B8D433, + 0x7807C9A2, 0x0F00F934, 0x9609A88E, 0xE10E9818, 0x7F6A0DBB, 0x086D3D2D, + 0x91646C97, 0xE6635C01, 0x6B6B51F4, 0x1C6C6162, 0x856530D8, 0xF262004E, + 0x6C0695ED, 0x1B01A57B, 0x8208F4C1, 0xF50FC457, 0x65B0D9C6, 0x12B7E950, + 0x8BBEB8EA, 0xFCB9887C, 0x62DD1DDF, 0x15DA2D49, 0x8CD37CF3, 0xFBD44C65, + 0x4DB26158, 0x3AB551CE, 0xA3BC0074, 0xD4BB30E2, 0x4ADFA541, 0x3DD895D7, + 0xA4D1C46D, 0xD3D6F4FB, 0x4369E96A, 0x346ED9FC, 0xAD678846, 0xDA60B8D0, + 0x44042D73, 0x33031DE5, 0xAA0A4C5F, 0xDD0D7CC9, 0x5005713C, 0x270241AA, + 0xBE0B1010, 0xC90C2086, 0x5768B525, 0x206F85B3, 0xB966D409, 0xCE61E49F, + 0x5EDEF90E, 0x29D9C998, 0xB0D09822, 0xC7D7A8B4, 0x59B33D17, 0x2EB40D81, + 0xB7BD5C3B, 0xC0BA6CAD, 0xEDB88320, 0x9ABFB3B6, 0x03B6E20C, 0x74B1D29A, + 0xEAD54739, 0x9DD277AF, 0x04DB2615, 0x73DC1683, 0xE3630B12, 0x94643B84, + 0x0D6D6A3E, 0x7A6A5AA8, 0xE40ECF0B, 0x9309FF9D, 0x0A00AE27, 0x7D079EB1, + 0xF00F9344, 0x8708A3D2, 0x1E01F268, 0x6906C2FE, 0xF762575D, 0x806567CB, + 0x196C3671, 0x6E6B06E7, 0xFED41B76, 0x89D32BE0, 0x10DA7A5A, 0x67DD4ACC, + 0xF9B9DF6F, 0x8EBEEFF9, 0x17B7BE43, 0x60B08ED5, 0xD6D6A3E8, 0xA1D1937E, + 0x38D8C2C4, 0x4FDFF252, 0xD1BB67F1, 0xA6BC5767, 0x3FB506DD, 0x48B2364B, + 0xD80D2BDA, 0xAF0A1B4C, 0x36034AF6, 0x41047A60, 0xDF60EFC3, 0xA867DF55, + 0x316E8EEF, 0x4669BE79, 0xCB61B38C, 0xBC66831A, 0x256FD2A0, 0x5268E236, + 0xCC0C7795, 0xBB0B4703, 0x220216B9, 0x5505262F, 0xC5BA3BBE, 0xB2BD0B28, + 0x2BB45A92, 0x5CB36A04, 0xC2D7FFA7, 0xB5D0CF31, 0x2CD99E8B, 0x5BDEAE1D, + 0x9B64C2B0, 0xEC63F226, 0x756AA39C, 0x026D930A, 0x9C0906A9, 0xEB0E363F, + 0x72076785, 0x05005713, 0x95BF4A82, 0xE2B87A14, 0x7BB12BAE, 0x0CB61B38, + 0x92D28E9B, 0xE5D5BE0D, 0x7CDCEFB7, 0x0BDBDF21, 0x86D3D2D4, 0xF1D4E242, + 0x68DDB3F8, 0x1FDA836E, 0x81BE16CD, 0xF6B9265B, 0x6FB077E1, 0x18B74777, + 0x88085AE6, 0xFF0F6A70, 0x66063BCA, 0x11010B5C, 0x8F659EFF, 0xF862AE69, + 0x616BFFD3, 0x166CCF45, 0xA00AE278, 0xD70DD2EE, 0x4E048354, 0x3903B3C2, + 0xA7672661, 0xD06016F7, 0x4969474D, 0x3E6E77DB, 0xAED16A4A, 0xD9D65ADC, + 0x40DF0B66, 0x37D83BF0, 0xA9BCAE53, 0xDEBB9EC5, 0x47B2CF7F, 0x30B5FFE9, + 0xBDBDF21C, 0xCABAC28A, 0x53B39330, 0x24B4A3A6, 0xBAD03605, 0xCDD70693, + 0x54DE5729, 0x23D967BF, 0xB3667A2E, 0xC4614AB8, 0x5D681B02, 0x2A6F2B94, + 0xB40BBE37, 0xC30C8EA1, 0x5A05DF1B, 0x2D02EF8D }; + + +static inline void CRC32_update_byte( uint32_t * const crc, const uint8_t byte ) + { *crc = crc32[(*crc^byte)&0xFF] ^ ( *crc >> 8 ); } + +/* about as fast as it is possible without messing with endianness */ +static inline void CRC32_update_buf( uint32_t * const crc, + const uint8_t * const buffer, + const int size ) + { + int i; + uint32_t c = *crc; + for( i = 0; i < size; ++i ) + c = crc32[(c^buffer[i])&0xFF] ^ ( c >> 8 ); + *crc = c; + } + + +static inline bool isvalid_ds( const unsigned dictionary_size ) + { return ( dictionary_size >= min_dictionary_size && + dictionary_size <= max_dictionary_size ); } + + +static inline int real_bits( unsigned value ) + { + int bits = 0; + while( value > 0 ) { value >>= 1; ++bits; } + return bits; + } + + +static const uint8_t lzip_magic[4] = { 0x4C, 0x5A, 0x49, 0x50 }; /* "LZIP" */ + +typedef uint8_t Lzip_header[6]; /* 0-3 magic bytes */ + /* 4 version */ + /* 5 coded dictionary size */ +enum { Lh_size = 6 }; + +static inline void Lh_set_magic( Lzip_header data ) + { memcpy( data, lzip_magic, 4 ); data[4] = 1; } + +static inline bool Lh_verify_magic( const Lzip_header data ) + { return ( memcmp( data, lzip_magic, 4 ) == 0 ); } + +/* detect (truncated) header */ +static inline bool Lh_verify_prefix( const Lzip_header data, const int sz ) + { + int i; for( i = 0; i < sz && i < 4; ++i ) + if( data[i] != lzip_magic[i] ) return false; + return ( sz > 0 ); + } + +/* detect corrupt header */ +static inline bool Lh_verify_corrupt( const Lzip_header data ) + { + int matches = 0; + int i; for( i = 0; i < 4; ++i ) + if( data[i] == lzip_magic[i] ) ++matches; + return ( matches > 1 && matches < 4 ); + } + +static inline uint8_t Lh_version( const Lzip_header data ) + { return data[4]; } + +static inline bool Lh_verify_version( const Lzip_header data ) + { return ( data[4] == 1 ); } + +static inline unsigned Lh_get_dictionary_size( const Lzip_header data ) + { + unsigned sz = ( 1 << ( data[5] & 0x1F ) ); + if( sz > min_dictionary_size ) + sz -= ( sz / 16 ) * ( ( data[5] >> 5 ) & 7 ); + return sz; + } + +static inline bool Lh_set_dictionary_size( Lzip_header data, const unsigned sz ) + { + if( !isvalid_ds( sz ) ) return false; + data[5] = real_bits( sz - 1 ); + if( sz > min_dictionary_size ) + { + const unsigned base_size = 1 << data[5]; + const unsigned fraction = base_size / 16; + unsigned i; + for( i = 7; i >= 1; --i ) + if( base_size - ( i * fraction ) >= sz ) + { data[5] |= ( i << 5 ); break; } + } + return true; + } + +static inline bool Lh_verify( const Lzip_header data ) + { + return Lh_verify_magic( data ) && Lh_verify_version( data ) && + isvalid_ds( Lh_get_dictionary_size( data ) ); + } + + +typedef uint8_t Lzip_trailer[20]; + /* 0-3 CRC32 of the uncompressed data */ + /* 4-11 size of the uncompressed data */ + /* 12-19 member size including header and trailer */ +enum { Lt_size = 20 }; + +static inline unsigned Lt_get_data_crc( const Lzip_trailer data ) + { + unsigned tmp = 0; + int i; for( i = 3; i >= 0; --i ) { tmp <<= 8; tmp += data[i]; } + return tmp; + } + +static inline void Lt_set_data_crc( Lzip_trailer data, unsigned crc ) + { int i; for( i = 0; i <= 3; ++i ) { data[i] = (uint8_t)crc; crc >>= 8; } } + +static inline unsigned long long Lt_get_data_size( const Lzip_trailer data ) + { + unsigned long long tmp = 0; + int i; for( i = 11; i >= 4; --i ) { tmp <<= 8; tmp += data[i]; } + return tmp; + } + +static inline void Lt_set_data_size( Lzip_trailer data, unsigned long long sz ) + { int i; for( i = 4; i <= 11; ++i ) { data[i] = (uint8_t)sz; sz >>= 8; } } + +static inline unsigned long long Lt_get_member_size( const Lzip_trailer data ) + { + unsigned long long tmp = 0; + int i; for( i = 19; i >= 12; --i ) { tmp <<= 8; tmp += data[i]; } + return tmp; + } + +static inline void Lt_set_member_size( Lzip_trailer data, unsigned long long sz ) + { int i; for( i = 12; i <= 19; ++i ) { data[i] = (uint8_t)sz; sz >>= 8; } } diff --git a/lzlib.c b/lzlib.c new file mode 100644 index 0000000..3084fbd --- /dev/null +++ b/lzlib.c @@ -0,0 +1,601 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +#include +#include +#include +#include + +#include "lzlib.h" +#include "lzip.h" +#include "cbuffer.c" +#include "decoder.h" +#include "decoder.c" +#include "encoder_base.h" +#include "encoder_base.c" +#include "encoder.h" +#include "encoder.c" +#include "fast_encoder.h" +#include "fast_encoder.c" + + +struct LZ_Encoder + { + unsigned long long partial_in_size; + unsigned long long partial_out_size; + struct LZ_encoder_base * lz_encoder_base; /* these 3 pointers make a */ + struct LZ_encoder * lz_encoder; /* polymorphic encoder */ + struct FLZ_encoder * flz_encoder; + enum LZ_Errno lz_errno; + bool fatal; + }; + +static void LZ_Encoder_init( struct LZ_Encoder * const e ) + { + e->partial_in_size = 0; + e->partial_out_size = 0; + e->lz_encoder_base = 0; + e->lz_encoder = 0; + e->flz_encoder = 0; + e->lz_errno = LZ_ok; + e->fatal = false; + } + + +struct LZ_Decoder + { + unsigned long long partial_in_size; + unsigned long long partial_out_size; + struct Range_decoder * rdec; + struct LZ_decoder * lz_decoder; + enum LZ_Errno lz_errno; + Lzip_header member_header; /* header of current member */ + bool fatal; + bool first_header; /* true until first header is read */ + bool seeking; + }; + +static void LZ_Decoder_init( struct LZ_Decoder * const d ) + { + int i; + d->partial_in_size = 0; + d->partial_out_size = 0; + d->rdec = 0; + d->lz_decoder = 0; + d->lz_errno = LZ_ok; + for( i = 0; i < Lh_size; ++i ) d->member_header[i] = 0; + d->fatal = false; + d->first_header = true; + d->seeking = false; + } + + +static bool verify_encoder( struct LZ_Encoder * const e ) + { + if( !e ) return false; + if( !e->lz_encoder_base || ( !e->lz_encoder && !e->flz_encoder ) || + ( e->lz_encoder && e->flz_encoder ) ) + { e->lz_errno = LZ_bad_argument; return false; } + return true; + } + + +static bool verify_decoder( struct LZ_Decoder * const d ) + { + if( !d ) return false; + if( !d->rdec ) + { d->lz_errno = LZ_bad_argument; return false; } + return true; + } + + +/* ------------------------- Misc Functions ------------------------- */ + +int LZ_api_version( void ) { return LZ_API_VERSION; } + +const char * LZ_version( void ) { return LZ_version_string; } + +const char * LZ_strerror( const enum LZ_Errno lz_errno ) + { + switch( lz_errno ) + { + case LZ_ok : return "ok"; + case LZ_bad_argument : return "Bad argument"; + case LZ_mem_error : return "Not enough memory"; + case LZ_sequence_error: return "Sequence error"; + case LZ_header_error : return "Header error"; + case LZ_unexpected_eof: return "Unexpected EOF"; + case LZ_data_error : return "Data error"; + case LZ_library_error : return "Library error"; + } + return "Invalid error code"; + } + + +int LZ_min_dictionary_bits( void ) { return min_dictionary_bits; } +int LZ_min_dictionary_size( void ) { return min_dictionary_size; } +int LZ_max_dictionary_bits( void ) { return max_dictionary_bits; } +int LZ_max_dictionary_size( void ) { return max_dictionary_size; } +int LZ_min_match_len_limit( void ) { return min_match_len_limit; } +int LZ_max_match_len_limit( void ) { return max_match_len; } + + +/* --------------------- Compression Functions --------------------- */ + +struct LZ_Encoder * LZ_compress_open( const int dictionary_size, + const int match_len_limit, + const unsigned long long member_size ) + { + Lzip_header header; + struct LZ_Encoder * const e = + (struct LZ_Encoder *)malloc( sizeof (struct LZ_Encoder) ); + if( !e ) return 0; + LZ_Encoder_init( e ); + if( !Lh_set_dictionary_size( header, dictionary_size ) || + match_len_limit < min_match_len_limit || + match_len_limit > max_match_len || + member_size < min_dictionary_size ) + e->lz_errno = LZ_bad_argument; + else + { + if( dictionary_size == 65535 && match_len_limit == 16 ) + { + e->flz_encoder = (struct FLZ_encoder *)malloc( sizeof (struct FLZ_encoder) ); + if( e->flz_encoder && FLZe_init( e->flz_encoder, member_size ) ) + { e->lz_encoder_base = &e->flz_encoder->eb; return e; } + free( e->flz_encoder ); e->flz_encoder = 0; + } + else + { + e->lz_encoder = (struct LZ_encoder *)malloc( sizeof (struct LZ_encoder) ); + if( e->lz_encoder && LZe_init( e->lz_encoder, Lh_get_dictionary_size( header ), + match_len_limit, member_size ) ) + { e->lz_encoder_base = &e->lz_encoder->eb; return e; } + free( e->lz_encoder ); e->lz_encoder = 0; + } + e->lz_errno = LZ_mem_error; + } + e->fatal = true; + return e; + } + + +int LZ_compress_close( struct LZ_Encoder * const e ) + { + if( !e ) return -1; + if( e->lz_encoder_base ) + { LZeb_free( e->lz_encoder_base ); + free( e->lz_encoder ); free( e->flz_encoder ); } + free( e ); + return 0; + } + + +int LZ_compress_finish( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) || e->fatal ) return -1; + Mb_finish( &e->lz_encoder_base->mb ); + /* if (open --> write --> finish) use same dictionary size as lzip. */ + /* this does not save any memory. */ + if( Mb_data_position( &e->lz_encoder_base->mb ) == 0 && + Re_member_position( &e->lz_encoder_base->renc ) == Lh_size ) + { + Mb_adjust_dictionary_size( &e->lz_encoder_base->mb ); + Lh_set_dictionary_size( e->lz_encoder_base->renc.header, + e->lz_encoder_base->mb.dictionary_size ); + e->lz_encoder_base->renc.cb.buffer[5] = e->lz_encoder_base->renc.header[5]; + } + return 0; + } + + +int LZ_compress_restart_member( struct LZ_Encoder * const e, + const unsigned long long member_size ) + { + if( !verify_encoder( e ) || e->fatal ) return -1; + if( !LZeb_member_finished( e->lz_encoder_base ) ) + { e->lz_errno = LZ_sequence_error; return -1; } + if( member_size < min_dictionary_size ) + { e->lz_errno = LZ_bad_argument; return -1; } + + e->partial_in_size += Mb_data_position( &e->lz_encoder_base->mb ); + e->partial_out_size += Re_member_position( &e->lz_encoder_base->renc ); + + if( e->lz_encoder ) LZe_reset( e->lz_encoder, member_size ); + else FLZe_reset( e->flz_encoder, member_size ); + e->lz_errno = LZ_ok; + return 0; + } + + +int LZ_compress_sync_flush( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) || e->fatal ) return -1; + if( !e->lz_encoder_base->mb.at_stream_end ) + e->lz_encoder_base->mb.sync_flush_pending = true; + return 0; + } + + +int LZ_compress_read( struct LZ_Encoder * const e, + uint8_t * const buffer, const int size ) + { + if( !verify_encoder( e ) || e->fatal ) return -1; + if( size < 0 ) return 0; + + { struct LZ_encoder_base * const eb = e->lz_encoder_base; + int out_size = Re_read_data( &eb->renc, buffer, size ); + /* minimize number of calls to encode_member */ + if( out_size < size || size == 0 ) + { + if( ( e->flz_encoder && !FLZe_encode_member( e->flz_encoder ) ) || + ( e->lz_encoder && !LZe_encode_member( e->lz_encoder ) ) ) + { e->lz_errno = LZ_library_error; e->fatal = true; return -1; } + if( eb->mb.sync_flush_pending && Mb_available_bytes( &eb->mb ) <= 0 ) + LZeb_try_sync_flush( eb ); + out_size += Re_read_data( &eb->renc, buffer + out_size, size - out_size ); + } + return out_size; } + } + + +int LZ_compress_write( struct LZ_Encoder * const e, + const uint8_t * const buffer, const int size ) + { + if( !verify_encoder( e ) || e->fatal ) return -1; + return Mb_write_data( &e->lz_encoder_base->mb, buffer, size ); + } + + +int LZ_compress_write_size( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) || e->fatal ) return -1; + return Mb_free_bytes( &e->lz_encoder_base->mb ); + } + + +enum LZ_Errno LZ_compress_errno( struct LZ_Encoder * const e ) + { + if( !e ) return LZ_bad_argument; + return e->lz_errno; + } + + +int LZ_compress_finished( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) ) return -1; + return ( Mb_data_finished( &e->lz_encoder_base->mb ) && + LZeb_member_finished( e->lz_encoder_base ) ); + } + + +int LZ_compress_member_finished( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) ) return -1; + return LZeb_member_finished( e->lz_encoder_base ); + } + + +unsigned long long LZ_compress_data_position( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) ) return 0; + return Mb_data_position( &e->lz_encoder_base->mb ); + } + + +unsigned long long LZ_compress_member_position( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) ) return 0; + return Re_member_position( &e->lz_encoder_base->renc ); + } + + +unsigned long long LZ_compress_total_in_size( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) ) return 0; + return e->partial_in_size + Mb_data_position( &e->lz_encoder_base->mb ); + } + + +unsigned long long LZ_compress_total_out_size( struct LZ_Encoder * const e ) + { + if( !verify_encoder( e ) ) return 0; + return e->partial_out_size + Re_member_position( &e->lz_encoder_base->renc ); + } + + +/* -------------------- Decompression Functions -------------------- */ + +struct LZ_Decoder * LZ_decompress_open( void ) + { + struct LZ_Decoder * const d = + (struct LZ_Decoder *)malloc( sizeof (struct LZ_Decoder) ); + if( !d ) return 0; + LZ_Decoder_init( d ); + + d->rdec = (struct Range_decoder *)malloc( sizeof (struct Range_decoder) ); + if( !d->rdec || !Rd_init( d->rdec ) ) + { + if( d->rdec ) { Rd_free( d->rdec ); free( d->rdec ); d->rdec = 0; } + d->lz_errno = LZ_mem_error; d->fatal = true; + } + return d; + } + + +int LZ_decompress_close( struct LZ_Decoder * const d ) + { + if( !d ) return -1; + if( d->lz_decoder ) + { LZd_free( d->lz_decoder ); free( d->lz_decoder ); } + if( d->rdec ) { Rd_free( d->rdec ); free( d->rdec ); } + free( d ); + return 0; + } + + +int LZ_decompress_finish( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) || d->fatal ) return -1; + if( d->seeking ) + { d->seeking = false; d->partial_in_size += Rd_purge( d->rdec ); } + else Rd_finish( d->rdec ); + return 0; + } + + +int LZ_decompress_reset( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) ) return -1; + if( d->lz_decoder ) + { LZd_free( d->lz_decoder ); free( d->lz_decoder ); d->lz_decoder = 0; } + d->partial_in_size = 0; + d->partial_out_size = 0; + Rd_reset( d->rdec ); + d->lz_errno = LZ_ok; + d->fatal = false; + d->first_header = true; + d->seeking = false; + return 0; + } + + +int LZ_decompress_sync_to_member( struct LZ_Decoder * const d ) + { + unsigned skipped = 0; + if( !verify_decoder( d ) ) return -1; + if( d->lz_decoder ) + { LZd_free( d->lz_decoder ); free( d->lz_decoder ); d->lz_decoder = 0; } + if( Rd_find_header( d->rdec, &skipped ) ) d->seeking = false; + else + { + if( !d->rdec->at_stream_end ) d->seeking = true; + else { d->seeking = false; d->partial_in_size += Rd_purge( d->rdec ); } + } + d->partial_in_size += skipped; + d->lz_errno = LZ_ok; + d->fatal = false; + return 0; + } + + +int LZ_decompress_read( struct LZ_Decoder * const d, + uint8_t * const buffer, const int size ) + { + int result; + if( !verify_decoder( d ) ) return -1; + if( size < 0 ) return 0; + if( d->fatal ) /* don't return error until pending bytes are read */ + { if( d->lz_decoder && !Cb_empty( &d->lz_decoder->cb ) ) goto get_data; + return -1; } + if( d->seeking ) return 0; + + if( d->lz_decoder && LZd_member_finished( d->lz_decoder ) ) + { + d->partial_out_size += LZd_data_position( d->lz_decoder ); + LZd_free( d->lz_decoder ); free( d->lz_decoder ); d->lz_decoder = 0; + } + if( !d->lz_decoder ) + { + int rd; + d->partial_in_size += d->rdec->member_position; + d->rdec->member_position = 0; + if( Rd_available_bytes( d->rdec ) < Lh_size + 5 && + !d->rdec->at_stream_end ) return 0; + if( Rd_finished( d->rdec ) && !d->first_header ) return 0; + rd = Rd_read_data( d->rdec, d->member_header, Lh_size ); + if( rd < Lh_size || Rd_finished( d->rdec ) ) /* End Of File */ + { + if( rd <= 0 || Lh_verify_prefix( d->member_header, rd ) ) + d->lz_errno = LZ_unexpected_eof; + else + d->lz_errno = LZ_header_error; + d->fatal = true; + return -1; + } + if( !Lh_verify_magic( d->member_header ) ) + { + /* unreading the header prevents sync_to_member from skipping a member + if leading garbage is shorter than a full header; "lgLZIP\x01\x0C" */ + if( Rd_unread_data( d->rdec, rd ) ) + { + if( d->first_header || !Lh_verify_corrupt( d->member_header ) ) + d->lz_errno = LZ_header_error; + else + d->lz_errno = LZ_data_error; /* corrupt header */ + } + else + d->lz_errno = LZ_library_error; + d->fatal = true; + return -1; + } + if( !Lh_verify_version( d->member_header ) || + !isvalid_ds( Lh_get_dictionary_size( d->member_header ) ) ) + { + /* Skip a possible "LZIP" leading garbage; "LZIPLZIP\x01\x0C". + Leave member_pos pointing to the first error. */ + if( Rd_unread_data( d->rdec, 1 + !Lh_verify_version( d->member_header ) ) ) + d->lz_errno = LZ_data_error; /* bad version or bad dict size */ + else + d->lz_errno = LZ_library_error; + d->fatal = true; + return -1; + } + d->first_header = false; + if( Rd_available_bytes( d->rdec ) < 5 ) + { + /* set position at EOF */ + d->rdec->member_position += Cb_used_bytes( &d->rdec->cb ); + Cb_reset( &d->rdec->cb ); + d->lz_errno = LZ_unexpected_eof; + d->fatal = true; + return -1; + } + d->lz_decoder = (struct LZ_decoder *)malloc( sizeof (struct LZ_decoder) ); + if( !d->lz_decoder || !LZd_init( d->lz_decoder, d->rdec, + Lh_get_dictionary_size( d->member_header ) ) ) + { /* not enough free memory */ + if( d->lz_decoder ) + { LZd_free( d->lz_decoder ); free( d->lz_decoder ); d->lz_decoder = 0; } + d->lz_errno = LZ_mem_error; + d->fatal = true; + return -1; + } + d->rdec->reload_pending = true; + } + result = LZd_decode_member( d->lz_decoder ); + if( result != 0 ) + { + if( result == 2 ) /* set input position at EOF */ + { d->rdec->member_position += Cb_used_bytes( &d->rdec->cb ); + Cb_reset( &d->rdec->cb ); + d->lz_errno = LZ_unexpected_eof; } + else if( result == 5 ) d->lz_errno = LZ_library_error; + else d->lz_errno = LZ_data_error; + d->fatal = true; + if( Cb_empty( &d->lz_decoder->cb ) ) return -1; + } +get_data: + return Cb_read_data( &d->lz_decoder->cb, buffer, size ); + } + + +int LZ_decompress_write( struct LZ_Decoder * const d, + const uint8_t * const buffer, const int size ) + { + int result; + if( !verify_decoder( d ) || d->fatal ) return -1; + if( size < 0 ) return 0; + + result = Rd_write_data( d->rdec, buffer, size ); + while( d->seeking ) + { + int size2; + unsigned skipped = 0; + if( Rd_find_header( d->rdec, &skipped ) ) d->seeking = false; + d->partial_in_size += skipped; + if( result >= size ) break; + size2 = Rd_write_data( d->rdec, buffer + result, size - result ); + if( size2 > 0 ) result += size2; + else break; + } + return result; + } + + +int LZ_decompress_write_size( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) || d->fatal ) return -1; + return Rd_free_bytes( d->rdec ); + } + + +enum LZ_Errno LZ_decompress_errno( struct LZ_Decoder * const d ) + { + if( !d ) return LZ_bad_argument; + return d->lz_errno; + } + + +int LZ_decompress_finished( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) || d->fatal ) return -1; + return ( Rd_finished( d->rdec ) && + ( !d->lz_decoder || LZd_member_finished( d->lz_decoder ) ) ); + } + + +int LZ_decompress_member_finished( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) || d->fatal ) return -1; + return ( d->lz_decoder && LZd_member_finished( d->lz_decoder ) ); + } + + +int LZ_decompress_member_version( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) ) return -1; + return Lh_version( d->member_header ); + } + + +int LZ_decompress_dictionary_size( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) ) return -1; + return Lh_get_dictionary_size( d->member_header ); + } + + +unsigned LZ_decompress_data_crc( struct LZ_Decoder * const d ) + { + if( verify_decoder( d ) && d->lz_decoder ) + return LZd_crc( d->lz_decoder ); + return 0; + } + + +unsigned long long LZ_decompress_data_position( struct LZ_Decoder * const d ) + { + if( verify_decoder( d ) && d->lz_decoder ) + return LZd_data_position( d->lz_decoder ); + return 0; + } + + +unsigned long long LZ_decompress_member_position( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) ) return 0; + return d->rdec->member_position; + } + + +unsigned long long LZ_decompress_total_in_size( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) ) return 0; + return d->partial_in_size + d->rdec->member_position; + } + + +unsigned long long LZ_decompress_total_out_size( struct LZ_Decoder * const d ) + { + if( !verify_decoder( d ) ) return 0; + if( d->lz_decoder ) + return d->partial_out_size + LZd_data_position( d->lz_decoder ); + return d->partial_out_size; + } diff --git a/lzlib.h b/lzlib.h new file mode 100644 index 0000000..ba29f97 --- /dev/null +++ b/lzlib.h @@ -0,0 +1,110 @@ +/* Lzlib - Compression library for the lzip format + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + This library is free software. Redistribution and use in source and + binary forms, with or without modification, are permitted provided + that the following conditions are met: + + 1. Redistributions of source code must retain the above copyright + notice, this list of conditions, and the following disclaimer. + + 2. Redistributions in binary form must reproduce the above copyright + notice, this list of conditions, and the following disclaimer in the + documentation and/or other materials provided with the distribution. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +*/ + +#ifdef __cplusplus +extern "C" { +#endif + +/* LZ_API_VERSION was first defined in lzlib 1.8 to 1. + Since lzlib 1.12, LZ_API_VERSION is defined as (major * 1000 + minor). */ + +#define LZ_API_VERSION 1013 + +static const char * const LZ_version_string = "1.13"; + +enum LZ_Errno { LZ_ok = 0, LZ_bad_argument, LZ_mem_error, + LZ_sequence_error, LZ_header_error, LZ_unexpected_eof, + LZ_data_error, LZ_library_error }; + + +int LZ_api_version( void ); /* new in 1.12 */ +const char * LZ_version( void ); +const char * LZ_strerror( const enum LZ_Errno lz_errno ); + +int LZ_min_dictionary_bits( void ); +int LZ_min_dictionary_size( void ); +int LZ_max_dictionary_bits( void ); +int LZ_max_dictionary_size( void ); +int LZ_min_match_len_limit( void ); +int LZ_max_match_len_limit( void ); + + +/* --------------------- Compression Functions --------------------- */ + +struct LZ_Encoder; + +struct LZ_Encoder * LZ_compress_open( const int dictionary_size, + const int match_len_limit, + const unsigned long long member_size ); +int LZ_compress_close( struct LZ_Encoder * const encoder ); + +int LZ_compress_finish( struct LZ_Encoder * const encoder ); +int LZ_compress_restart_member( struct LZ_Encoder * const encoder, + const unsigned long long member_size ); +int LZ_compress_sync_flush( struct LZ_Encoder * const encoder ); + +int LZ_compress_read( struct LZ_Encoder * const encoder, + uint8_t * const buffer, const int size ); +int LZ_compress_write( struct LZ_Encoder * const encoder, + const uint8_t * const buffer, const int size ); +int LZ_compress_write_size( struct LZ_Encoder * const encoder ); + +enum LZ_Errno LZ_compress_errno( struct LZ_Encoder * const encoder ); +int LZ_compress_finished( struct LZ_Encoder * const encoder ); +int LZ_compress_member_finished( struct LZ_Encoder * const encoder ); + +unsigned long long LZ_compress_data_position( struct LZ_Encoder * const encoder ); +unsigned long long LZ_compress_member_position( struct LZ_Encoder * const encoder ); +unsigned long long LZ_compress_total_in_size( struct LZ_Encoder * const encoder ); +unsigned long long LZ_compress_total_out_size( struct LZ_Encoder * const encoder ); + + +/* -------------------- Decompression Functions -------------------- */ + +struct LZ_Decoder; + +struct LZ_Decoder * LZ_decompress_open( void ); +int LZ_decompress_close( struct LZ_Decoder * const decoder ); + +int LZ_decompress_finish( struct LZ_Decoder * const decoder ); +int LZ_decompress_reset( struct LZ_Decoder * const decoder ); +int LZ_decompress_sync_to_member( struct LZ_Decoder * const decoder ); + +int LZ_decompress_read( struct LZ_Decoder * const decoder, + uint8_t * const buffer, const int size ); +int LZ_decompress_write( struct LZ_Decoder * const decoder, + const uint8_t * const buffer, const int size ); +int LZ_decompress_write_size( struct LZ_Decoder * const decoder ); + +enum LZ_Errno LZ_decompress_errno( struct LZ_Decoder * const decoder ); +int LZ_decompress_finished( struct LZ_Decoder * const decoder ); +int LZ_decompress_member_finished( struct LZ_Decoder * const decoder ); + +int LZ_decompress_member_version( struct LZ_Decoder * const decoder ); +int LZ_decompress_dictionary_size( struct LZ_Decoder * const decoder ); +unsigned LZ_decompress_data_crc( struct LZ_Decoder * const decoder ); + +unsigned long long LZ_decompress_data_position( struct LZ_Decoder * const decoder ); +unsigned long long LZ_decompress_member_position( struct LZ_Decoder * const decoder ); +unsigned long long LZ_decompress_total_in_size( struct LZ_Decoder * const decoder ); +unsigned long long LZ_decompress_total_out_size( struct LZ_Decoder * const decoder ); + +#ifdef __cplusplus +} +#endif diff --git a/minilzip.c b/minilzip.c new file mode 100644 index 0000000..f9313b2 --- /dev/null +++ b/minilzip.c @@ -0,0 +1,1290 @@ +/* Minilzip - Test program for the library lzlib + Copyright (C) 2009-2022 Antonio Diaz Diaz. + + 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, see . +*/ +/* + Exit status: 0 for a normal exit, 1 for environmental problems + (file not found, invalid flags, I/O errors, etc), 2 to indicate a + corrupt or invalid input file, 3 for an internal consistency error + (e.g., bug) which caused minilzip to panic. +*/ + +#define _FILE_OFFSET_BITS 64 + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#if defined __MSVCRT__ || defined __OS2__ || defined __DJGPP__ +#include +#if defined __MSVCRT__ +#define fchmod(x,y) 0 +#define fchown(x,y,z) 0 +#define strtoull strtoul +#define SIGHUP SIGTERM +#define S_ISSOCK(x) 0 +#ifndef S_IRGRP +#define S_IRGRP 0 +#define S_IWGRP 0 +#define S_IROTH 0 +#define S_IWOTH 0 +#endif +#endif +#if defined __DJGPP__ +#define S_ISSOCK(x) 0 +#define S_ISVTX 0 +#endif +#endif + +#include "carg_parser.h" +#include "lzlib.h" + +#ifndef O_BINARY +#define O_BINARY 0 +#endif + +#if CHAR_BIT != 8 +#error "Environments where CHAR_BIT != 8 are not supported." +#endif + +#if ( defined SIZE_MAX && SIZE_MAX < UINT_MAX ) || \ + ( defined SSIZE_MAX && SSIZE_MAX < INT_MAX ) +#error "Environments where 'size_t' is narrower than 'int' are not supported." +#endif + +#ifndef max + #define max(x,y) ((x) >= (y) ? (x) : (y)) +#endif +#ifndef min + #define min(x,y) ((x) <= (y) ? (x) : (y)) +#endif + +static void cleanup_and_fail( const int retval ); +static void show_error( const char * const msg, const int errcode, + const bool help ); +static void show_file_error( const char * const filename, + const char * const msg, const int errcode ); +static void internal_error( const char * const msg ); +static const char * const mem_msg = "Not enough memory."; + +int verbosity = 0; + +static const char * const program_name = "minilzip"; +static const char * const program_year = "2022"; +static const char * invocation_name = "minilzip"; /* default value */ + +static const struct { const char * from; const char * to; } known_extensions[] = { + { ".lz", "" }, + { ".tlz", ".tar" }, + { 0, 0 } }; + +struct Lzma_options + { + int dictionary_size; /* 4 KiB .. 512 MiB */ + int match_len_limit; /* 5 .. 273 */ + }; + +enum Mode { m_compress, m_decompress, m_test }; + +/* Variables used in signal handler context. + They are not declared volatile because the handler never returns. */ +static char * output_filename = 0; +static int outfd = -1; +static bool delete_output_on_interrupt = false; + + +static void show_help( void ) + { + printf( "Minilzip is a test program for the compression library lzlib, fully\n" + "compatible with lzip 1.4 or newer.\n" + "\nLzip is a lossless data compressor with a user interface similar to the one\n" + "of gzip or bzip2. Lzip uses a simplified form of the 'Lempel-Ziv-Markov\n" + "chain-Algorithm' (LZMA) stream format and provides a 3 factor integrity\n" + "checking to maximize interoperability and optimize safety. Lzip can compress\n" + "about as fast as gzip (lzip -0) or compress most files more than bzip2\n" + "(lzip -9). Decompression speed is intermediate between gzip and bzip2.\n" + "Lzip is better than gzip and bzip2 from a data recovery perspective. Lzip\n" + "has been designed, written, and tested with great care to replace gzip and\n" + "bzip2 as the standard general-purpose compressed format for unix-like\n" + "systems.\n" + "\nUsage: %s [options] [files]\n", invocation_name ); + printf( "\nOptions:\n" + " -h, --help display this help and exit\n" + " -V, --version output version information and exit\n" + " -a, --trailing-error exit with error status if trailing data\n" + " -b, --member-size= set member size limit in bytes\n" + " -c, --stdout write to standard output, keep input files\n" + " -d, --decompress decompress\n" + " -f, --force overwrite existing output files\n" + " -F, --recompress force re-compression of compressed files\n" + " -k, --keep keep (don't delete) input files\n" + " -m, --match-length= set match length limit in bytes [36]\n" + " -o, --output= write to , keep input files\n" + " -q, --quiet suppress all messages\n" + " -s, --dictionary-size= set dictionary size limit in bytes [8 MiB]\n" + " -S, --volume-size= set volume size limit in bytes\n" + " -t, --test test compressed file integrity\n" + " -v, --verbose be verbose (a 2nd -v gives more)\n" + " -0 .. -9 set compression level [default 6]\n" + " --fast alias for -0\n" + " --best alias for -9\n" + " --loose-trailing allow trailing data seeming corrupt header\n" + " --check-lib compare version of lzlib.h with liblz.{a,so}\n" + "\nIf no file names are given, or if a file is '-', minilzip compresses or\n" + "decompresses from standard input to standard output.\n" + "Numbers may be followed by a multiplier: k = kB = 10^3 = 1000,\n" + "Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc...\n" + "Dictionary sizes 12 to 29 are interpreted as powers of two, meaning 2^12\n" + "to 2^29 bytes.\n" + "\nThe bidimensional parameter space of LZMA can't be mapped to a linear\n" + "scale optimal for all files. If your files are large, very repetitive,\n" + "etc, you may need to use the options --dictionary-size and --match-length\n" + "directly to achieve optimal performance.\n" + "\nTo extract all the files from archive 'foo.tar.lz', use the commands\n" + "'tar -xf foo.tar.lz' or 'minilzip -cd foo.tar.lz | tar -xf -'.\n" + "\nExit status: 0 for a normal exit, 1 for environmental problems (file\n" + "not found, invalid flags, I/O errors, etc), 2 to indicate a corrupt or\n" + "invalid input file, 3 for an internal consistency error (e.g., bug) which\n" + "caused minilzip to panic.\n" + "\nThe ideas embodied in lzlib are due to (at least) the following people:\n" + "Abraham Lempel and Jacob Ziv (for the LZ algorithm), Andrey Markov (for the\n" + "definition of Markov chains), G.N.N. Martin (for the definition of range\n" + "encoding), Igor Pavlov (for putting all the above together in LZMA), and\n" + "Julian Seward (for bzip2's CLI).\n" + "\nReport bugs to lzip-bug@nongnu.org\n" + "Lzlib home page: http://www.nongnu.org/lzip/lzlib.html\n" ); + } + + +static void show_version( void ) + { + printf( "%s %s\n", program_name, PROGVERSION ); + printf( "Copyright (C) %s Antonio Diaz Diaz.\n", program_year ); + printf( "Using lzlib %s\n", LZ_version() ); + printf( "License GPLv2+: GNU GPL version 2 or later \n" + "This is free software: you are free to change and redistribute it.\n" + "There is NO WARRANTY, to the extent permitted by law.\n" ); + } + + +static inline void set_retval( int * retval, const int new_val ) + { if( *retval < new_val ) *retval = new_val; } + + +static int check_lzlib_ver() /* . or .[a-z.-]* */ + { +#if defined LZ_API_VERSION && LZ_API_VERSION >= 1012 + const unsigned char * p = (unsigned char *)LZ_version_string; + unsigned major = 0, minor = 0; + while( major < 100000 && isdigit( *p ) ) + { major *= 10; major += *p - '0'; ++p; } + if( *p == '.' ) ++p; + else +out: { show_error( "Invalid LZ_version_string in lzlib.h", 0, false ); return 2; } + while( minor < 100 && isdigit( *p ) ) + { minor *= 10; minor += *p - '0'; ++p; } + if( *p && *p != '-' && *p != '.' && !islower( *p ) ) goto out; + const unsigned version = major * 1000 + minor; + if( LZ_API_VERSION != version ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: Version mismatch in lzlib.h: " + "LZ_API_VERSION = %u, should be %u.\n", + program_name, LZ_API_VERSION, version ); + return 2; + } +#endif + return 0; + } + + +static int check_lib() + { + int retval = check_lzlib_ver(); + if( strcmp( LZ_version_string, LZ_version() ) != 0 ) + { set_retval( &retval, 1 ); + if( verbosity >= 0 ) + printf( "warning: LZ_version_string != LZ_version() (%s vs %s)\n", + LZ_version_string, LZ_version() ); } +#if defined LZ_API_VERSION && LZ_API_VERSION >= 1012 + if( LZ_API_VERSION != LZ_api_version() ) + { set_retval( &retval, 1 ); + if( verbosity >= 0 ) + printf( "warning: LZ_API_VERSION != LZ_api_version() (%u vs %u)\n", + LZ_API_VERSION, LZ_api_version() ); } +#endif + if( verbosity >= 1 ) + { + printf( "Using lzlib %s\n", LZ_version() ); +#if !defined LZ_API_VERSION + fputs( "LZ_API_VERSION is not defined.\n", stdout ); +#elif LZ_API_VERSION >= 1012 + printf( "Using LZ_API_VERSION = %u\n", LZ_api_version() ); +#else + printf( "Compiled with LZ_API_VERSION = %u. " + "Using an unknown LZ_API_VERSION\n", LZ_API_VERSION ); +#endif + } + return retval; + } + + +/* assure at least a minimum size for buffer 'buf' */ +static void * resize_buffer( void * buf, const unsigned min_size ) + { + if( buf ) buf = realloc( buf, min_size ); + else buf = malloc( min_size ); + if( !buf ) { show_error( mem_msg, 0, false ); cleanup_and_fail( 1 ); } + return buf; + } + + +struct Pretty_print + { + const char * name; + char * padded_name; + const char * stdin_name; + unsigned longest_name; + bool first_post; + }; + +static void Pp_init( struct Pretty_print * const pp, + const char * const filenames[], const int num_filenames ) + { + pp->name = 0; + pp->padded_name = 0; + pp->stdin_name = "(stdin)"; + pp->longest_name = 0; + pp->first_post = false; + + if( verbosity <= 0 ) return; + const unsigned stdin_name_len = strlen( pp->stdin_name ); + int i; + for( i = 0; i < num_filenames; ++i ) + { + const char * const s = filenames[i]; + const unsigned len = (strcmp( s, "-" ) == 0) ? stdin_name_len : strlen( s ); + if( pp->longest_name < len ) pp->longest_name = len; + } + if( pp->longest_name == 0 ) pp->longest_name = stdin_name_len; + } + +static void Pp_set_name( struct Pretty_print * const pp, + const char * const filename ) + { + unsigned name_len, padded_name_len, i = 0; + + if( filename && filename[0] && strcmp( filename, "-" ) != 0 ) + pp->name = filename; + else pp->name = pp->stdin_name; + name_len = strlen( pp->name ); + padded_name_len = max( name_len, pp->longest_name ) + 4; + pp->padded_name = resize_buffer( pp->padded_name, padded_name_len + 1 ); + while( i < 2 ) pp->padded_name[i++] = ' '; + while( i < name_len + 2 ) { pp->padded_name[i] = pp->name[i-2]; ++i; } + pp->padded_name[i++] = ':'; + while( i < padded_name_len ) pp->padded_name[i++] = ' '; + pp->padded_name[i] = 0; + pp->first_post = true; + } + +static void Pp_reset( struct Pretty_print * const pp ) + { if( pp->name && pp->name[0] ) pp->first_post = true; } + +static void Pp_show_msg( struct Pretty_print * const pp, const char * const msg ) + { + if( verbosity < 0 ) return; + if( pp->first_post ) + { + pp->first_post = false; + fputs( pp->padded_name, stderr ); + if( !msg ) fflush( stderr ); + } + if( msg ) fprintf( stderr, "%s\n", msg ); + } + + +static void show_header( const unsigned dictionary_size ) + { + enum { factor = 1024 }; + const char * const prefix[8] = + { "Ki", "Mi", "Gi", "Ti", "Pi", "Ei", "Zi", "Yi" }; + const char * p = ""; + const char * np = " "; + unsigned num = dictionary_size; + bool exact = ( num % factor == 0 ); + + int i; for( i = 0; i < 8 && ( num > 9999 || ( exact && num >= factor ) ); ++i ) + { num /= factor; if( num % factor != 0 ) exact = false; + p = prefix[i]; np = ""; } + fprintf( stderr, "dict %s%4u %sB, ", np, num, p ); + } + + +/* separate large numbers >= 100_000 in groups of 3 digits using '_' */ +static const char * format_num3( unsigned long long num ) + { + const char * const si_prefix = "kMGTPEZY"; + const char * const binary_prefix = "KMGTPEZY"; + enum { buffers = 8, bufsize = 4 * sizeof (long long) }; + static char buffer[buffers][bufsize]; /* circle of static buffers for printf */ + static int current = 0; + int i; + char * const buf = buffer[current++]; current %= buffers; + char * p = buf + bufsize - 1; /* fill the buffer backwards */ + *p = 0; /* terminator */ + if( num > 1024 ) + { + char prefix = 0; /* try binary first, then si */ + for( i = 0; i < 8 && num >= 1024 && num % 1024 == 0; ++i ) + { num /= 1024; prefix = binary_prefix[i]; } + if( prefix ) *(--p) = 'i'; + else + for( i = 0; i < 8 && num >= 1000 && num % 1000 == 0; ++i ) + { num /= 1000; prefix = si_prefix[i]; } + if( prefix ) *(--p) = prefix; + } + const bool split = num >= 100000; + + for( i = 0; ; ) + { + *(--p) = num % 10 + '0'; num /= 10; if( num == 0 ) break; + if( split && ++i >= 3 ) { i = 0; *(--p) = '_'; } + } + return p; + } + + +static unsigned long long getnum( const char * const arg, + const char * const option_name, + const unsigned long long llimit, + const unsigned long long ulimit ) + { + char * tail; + errno = 0; + unsigned long long result = strtoull( arg, &tail, 0 ); + if( tail == arg ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: Bad or missing numerical argument in " + "option '%s'.\n", program_name, option_name ); + exit( 1 ); + } + + if( !errno && tail[0] ) + { + const unsigned factor = ( tail[1] == 'i' ) ? 1024 : 1000; + int exponent = 0; /* 0 = bad multiplier */ + int i; + switch( tail[0] ) + { + case 'Y': exponent = 8; break; + case 'Z': exponent = 7; break; + case 'E': exponent = 6; break; + case 'P': exponent = 5; break; + case 'T': exponent = 4; break; + case 'G': exponent = 3; break; + case 'M': exponent = 2; break; + case 'K': if( factor == 1024 ) exponent = 1; break; + case 'k': if( factor == 1000 ) exponent = 1; break; + } + if( exponent <= 0 ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: Bad multiplier in numerical argument of " + "option '%s'.\n", program_name, option_name ); + exit( 1 ); + } + for( i = 0; i < exponent; ++i ) + { + if( ulimit / factor >= result ) result *= factor; + else { errno = ERANGE; break; } + } + } + if( !errno && ( result < llimit || result > ulimit ) ) errno = ERANGE; + if( errno ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: Numerical argument out of limits [%s,%s] " + "in option '%s'.\n", program_name, format_num3( llimit ), + format_num3( ulimit ), option_name ); + exit( 1 ); + } + return result; + } + + +static int get_dict_size( const char * const arg, const char * const option_name ) + { + char * tail; + const long bits = strtol( arg, &tail, 0 ); + if( bits >= LZ_min_dictionary_bits() && + bits <= LZ_max_dictionary_bits() && *tail == 0 ) + return 1 << bits; + int dictionary_size = getnum( arg, option_name, LZ_min_dictionary_size(), + LZ_max_dictionary_size() ); + if( dictionary_size == 65535 ) ++dictionary_size; /* no fast encoder */ + return dictionary_size; + } + + +static void set_mode( enum Mode * const program_modep, const enum Mode new_mode ) + { + if( *program_modep != m_compress && *program_modep != new_mode ) + { + show_error( "Only one operation can be specified.", 0, true ); + exit( 1 ); + } + *program_modep = new_mode; + } + + +static int extension_index( const char * const name ) + { + int eindex; + for( eindex = 0; known_extensions[eindex].from; ++eindex ) + { + const char * const ext = known_extensions[eindex].from; + const unsigned name_len = strlen( name ); + const unsigned ext_len = strlen( ext ); + if( name_len > ext_len && + strncmp( name + name_len - ext_len, ext, ext_len ) == 0 ) + return eindex; + } + return -1; + } + + +static void set_c_outname( const char * const name, const bool force_ext, + const bool multifile ) + { + output_filename = resize_buffer( output_filename, strlen( name ) + 5 + + strlen( known_extensions[0].from ) + 1 ); + strcpy( output_filename, name ); + if( multifile ) strcat( output_filename, "00001" ); + if( force_ext || multifile ) + strcat( output_filename, known_extensions[0].from ); + } + + +static void set_d_outname( const char * const name, const int eindex ) + { + const unsigned name_len = strlen( name ); + if( eindex >= 0 ) + { + const char * const from = known_extensions[eindex].from; + const unsigned from_len = strlen( from ); + if( name_len > from_len ) + { + output_filename = resize_buffer( output_filename, name_len + + strlen( known_extensions[eindex].to ) + 1 ); + strcpy( output_filename, name ); + strcpy( output_filename + name_len - from_len, known_extensions[eindex].to ); + return; + } + } + output_filename = resize_buffer( output_filename, name_len + 4 + 1 ); + strcpy( output_filename, name ); + strcat( output_filename, ".out" ); + if( verbosity >= 1 ) + fprintf( stderr, "%s: Can't guess original name for '%s' -- using '%s'\n", + program_name, name, output_filename ); + } + + +static int open_instream( const char * const name, struct stat * const in_statsp, + const enum Mode program_mode, const int eindex, + const bool one_to_one, const bool recompress ) + { + if( program_mode == m_compress && !recompress && eindex >= 0 ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: Input file '%s' already has '%s' suffix.\n", + program_name, name, known_extensions[eindex].from ); + return -1; + } + int infd = open( name, O_RDONLY | O_BINARY ); + if( infd < 0 ) + show_file_error( name, "Can't open input file", errno ); + else + { + const int i = fstat( infd, in_statsp ); + const mode_t mode = in_statsp->st_mode; + const bool can_read = ( i == 0 && + ( S_ISBLK( mode ) || S_ISCHR( mode ) || + S_ISFIFO( mode ) || S_ISSOCK( mode ) ) ); + if( i != 0 || ( !S_ISREG( mode ) && ( !can_read || one_to_one ) ) ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: Input file '%s' is not a regular file%s.\n", + program_name, name, ( can_read && one_to_one ) ? + ",\n and neither '-c' nor '-o' were specified" : "" ); + close( infd ); + infd = -1; + } + } + return infd; + } + + +static bool open_outstream( const bool force, const bool protect ) + { + const mode_t usr_rw = S_IRUSR | S_IWUSR; + const mode_t all_rw = usr_rw | S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH; + const mode_t outfd_mode = protect ? usr_rw : all_rw; + int flags = O_CREAT | O_WRONLY | O_BINARY; + if( force ) flags |= O_TRUNC; else flags |= O_EXCL; + + outfd = open( output_filename, flags, outfd_mode ); + if( outfd >= 0 ) delete_output_on_interrupt = true; + else if( verbosity >= 0 ) + { + if( errno == EEXIST ) + fprintf( stderr, "%s: Output file '%s' already exists, skipping.\n", + program_name, output_filename ); + else + fprintf( stderr, "%s: Can't create output file '%s': %s\n", + program_name, output_filename, strerror( errno ) ); + } + return ( outfd >= 0 ); + } + + +static void set_signals( void (*action)(int) ) + { + signal( SIGHUP, action ); + signal( SIGINT, action ); + signal( SIGTERM, action ); + } + + +static void cleanup_and_fail( const int retval ) + { + set_signals( SIG_IGN ); /* ignore signals */ + if( delete_output_on_interrupt ) + { + delete_output_on_interrupt = false; + if( verbosity >= 0 ) + fprintf( stderr, "%s: Deleting output file '%s', if it exists.\n", + program_name, output_filename ); + if( outfd >= 0 ) { close( outfd ); outfd = -1; } + if( remove( output_filename ) != 0 && errno != ENOENT ) + show_error( "WARNING: deletion of output file (apparently) failed.", 0, false ); + } + exit( retval ); + } + + +static void signal_handler( int sig ) + { + if( sig ) {} /* keep compiler happy */ + show_error( "Control-C or similar caught, quitting.", 0, false ); + cleanup_and_fail( 1 ); + } + + +static bool check_tty_in( const char * const input_filename, const int infd, + const enum Mode program_mode, int * const retval ) + { + if( ( program_mode == m_decompress || program_mode == m_test ) && + isatty( infd ) ) /* for example /dev/tty */ + { show_file_error( input_filename, + "I won't read compressed data from a terminal.", 0 ); + close( infd ); set_retval( retval, 2 ); + if( program_mode != m_test ) cleanup_and_fail( *retval ); + return false; } + return true; + } + +static bool check_tty_out( const enum Mode program_mode ) + { + if( program_mode == m_compress && isatty( outfd ) ) + { show_file_error( output_filename[0] ? + output_filename : "(stdout)", + "I won't write compressed data to a terminal.", 0 ); + return false; } + return true; + } + + +/* Set permissions, owner, and times. */ +static void close_and_set_permissions( const struct stat * const in_statsp ) + { + bool warning = false; + if( in_statsp ) + { + const mode_t mode = in_statsp->st_mode; + /* fchown will in many cases return with EPERM, which can be safely ignored. */ + if( fchown( outfd, in_statsp->st_uid, in_statsp->st_gid ) == 0 ) + { if( fchmod( outfd, mode ) != 0 ) warning = true; } + else + if( errno != EPERM || + fchmod( outfd, mode & ~( S_ISUID | S_ISGID | S_ISVTX ) ) != 0 ) + warning = true; + } + if( close( outfd ) != 0 ) + { + show_error( "Error closing output file", errno, false ); + cleanup_and_fail( 1 ); + } + outfd = -1; + delete_output_on_interrupt = false; + if( in_statsp ) + { + struct utimbuf t; + t.actime = in_statsp->st_atime; + t.modtime = in_statsp->st_mtime; + if( utime( output_filename, &t ) != 0 ) warning = true; + } + if( warning && verbosity >= 1 ) + show_error( "Can't change output file attributes.", 0, false ); + } + + +/* Return the number of bytes really read. + If (value returned < size) and (errno == 0), means EOF was reached. +*/ +static int readblock( const int fd, uint8_t * const buf, const int size ) + { + int sz = 0; + errno = 0; + while( sz < size ) + { + const int n = read( fd, buf + sz, size - sz ); + if( n > 0 ) sz += n; + else if( n == 0 ) break; /* EOF */ + else if( errno != EINTR ) break; + errno = 0; + } + return sz; + } + + +/* Return the number of bytes really written. + If (value returned < size), it is always an error. +*/ +static int writeblock( const int fd, const uint8_t * const buf, const int size ) + { + int sz = 0; + errno = 0; + while( sz < size ) + { + const int n = write( fd, buf + sz, size - sz ); + if( n > 0 ) sz += n; + else if( n < 0 && errno != EINTR ) break; + errno = 0; + } + return sz; + } + + +static bool next_filename( void ) + { + const unsigned name_len = strlen( output_filename ); + const unsigned ext_len = strlen( known_extensions[0].from ); + int i, j; + if( name_len >= ext_len + 5 ) /* "*00001.lz" */ + for( i = name_len - ext_len - 1, j = 0; j < 5; --i, ++j ) + { + if( output_filename[i] < '9' ) { ++output_filename[i]; return true; } + else output_filename[i] = '0'; + } + return false; + } + + +static int do_compress( struct LZ_Encoder * const encoder, + const unsigned long long member_size, + const unsigned long long volume_size, const int infd, + struct Pretty_print * const pp, + const struct stat * const in_statsp ) + { + unsigned long long partial_volume_size = 0; + enum { buffer_size = 65536 }; + uint8_t buffer[buffer_size]; /* read/write buffer */ + if( verbosity >= 1 ) Pp_show_msg( pp, 0 ); + + while( true ) + { + int in_size = 0; + while( LZ_compress_write_size( encoder ) > 0 ) + { + const int size = min( LZ_compress_write_size( encoder ), buffer_size ); + const int rd = readblock( infd, buffer, size ); + if( rd != size && errno ) + { + Pp_show_msg( pp, 0 ); show_error( "Read error", errno, false ); + return 1; + } + if( rd > 0 && rd != LZ_compress_write( encoder, buffer, rd ) ) + internal_error( "library error (LZ_compress_write)." ); + if( rd < size ) LZ_compress_finish( encoder ); +/* else LZ_compress_sync_flush( encoder ); */ + in_size += rd; + } + const int out_size = LZ_compress_read( encoder, buffer, buffer_size ); + if( out_size < 0 ) + { + Pp_show_msg( pp, 0 ); + if( verbosity >= 0 ) + fprintf( stderr, "%s: LZ_compress_read error: %s\n", + program_name, LZ_strerror( LZ_compress_errno( encoder ) ) ); + return 1; + } + else if( out_size > 0 ) + { + const int wr = writeblock( outfd, buffer, out_size ); + if( wr != out_size ) + { + Pp_show_msg( pp, 0 ); show_error( "Write error", errno, false ); + return 1; + } + } + else if( in_size == 0 ) + internal_error( "library error (LZ_compress_read)." ); + if( LZ_compress_member_finished( encoder ) ) + { + unsigned long long size; + if( LZ_compress_finished( encoder ) == 1 ) break; + if( volume_size > 0 ) + { + partial_volume_size += LZ_compress_member_position( encoder ); + if( partial_volume_size >= volume_size - LZ_min_dictionary_size() ) + { + partial_volume_size = 0; + if( delete_output_on_interrupt ) + { + close_and_set_permissions( in_statsp ); + if( !next_filename() ) + { Pp_show_msg( pp, "Too many volume files." ); return 1; } + if( !open_outstream( true, in_statsp ) ) return 1; + } + } + size = min( member_size, volume_size - partial_volume_size ); + } + else + size = member_size; + if( LZ_compress_restart_member( encoder, size ) < 0 ) + { + Pp_show_msg( pp, 0 ); + if( verbosity >= 0 ) + fprintf( stderr, "%s: LZ_compress_restart_member error: %s\n", + program_name, LZ_strerror( LZ_compress_errno( encoder ) ) ); + return 1; + } + } + } + + if( verbosity >= 1 ) + { + const unsigned long long in_size = LZ_compress_total_in_size( encoder ); + const unsigned long long out_size = LZ_compress_total_out_size( encoder ); + if( in_size == 0 || out_size == 0 ) + fputs( " no data compressed.\n", stderr ); + else + fprintf( stderr, "%6.3f:1, %5.2f%% ratio, %5.2f%% saved, " + "%llu in, %llu out.\n", + (double)in_size / out_size, + ( 100.0 * out_size ) / in_size, + 100.0 - ( ( 100.0 * out_size ) / in_size ), + in_size, out_size ); + } + return 0; + } + + +static int compress( const unsigned long long member_size, + const unsigned long long volume_size, const int infd, + const struct Lzma_options * const encoder_options, + struct Pretty_print * const pp, + const struct stat * const in_statsp ) + { + struct LZ_Encoder * const encoder = + LZ_compress_open( encoder_options->dictionary_size, + encoder_options->match_len_limit, ( volume_size > 0 ) ? + min( member_size, volume_size ) : member_size ); + int retval; + + if( !encoder || LZ_compress_errno( encoder ) != LZ_ok ) + { + if( !encoder || LZ_compress_errno( encoder ) == LZ_mem_error ) + Pp_show_msg( pp, "Not enough memory. Try a smaller dictionary size." ); + else + internal_error( "invalid argument to encoder." ); + retval = 1; + } + else retval = do_compress( encoder, member_size, volume_size, + infd, pp, in_statsp ); + LZ_compress_close( encoder ); + return retval; + } + + +static int do_decompress( struct LZ_Decoder * const decoder, const int infd, + struct Pretty_print * const pp, const bool ignore_trailing, + const bool loose_trailing, const bool testing ) + { + enum { buffer_size = 65536 }; + uint8_t buffer[buffer_size]; /* read/write buffer */ + unsigned long long total_in = 0; /* to detect library stall */ + bool first_member; + + for( first_member = true; ; ) + { + const int max_in_size = + min( LZ_decompress_write_size( decoder ), buffer_size ); + int in_size = 0, out_size = 0; + if( max_in_size > 0 ) + { + in_size = readblock( infd, buffer, max_in_size ); + if( in_size != max_in_size && errno ) + { + Pp_show_msg( pp, 0 ); show_error( "Read error", errno, false ); + return 1; + } + if( in_size > 0 && in_size != LZ_decompress_write( decoder, buffer, in_size ) ) + internal_error( "library error (LZ_decompress_write)." ); + if( in_size < max_in_size ) LZ_decompress_finish( decoder ); + } + while( true ) + { + const int rd = + LZ_decompress_read( decoder, (outfd >= 0) ? buffer : 0, buffer_size ); + if( rd > 0 ) + { + out_size += rd; + if( outfd >= 0 ) + { + const int wr = writeblock( outfd, buffer, rd ); + if( wr != rd ) + { + Pp_show_msg( pp, 0 ); show_error( "Write error", errno, false ); + return 1; + } + } + } + else if( rd < 0 ) { out_size = rd; break; } + if( LZ_decompress_member_finished( decoder ) == 1 ) + { + if( verbosity >= 1 ) + { + const unsigned long long data_size = LZ_decompress_data_position( decoder ); + const unsigned long long member_size = LZ_decompress_member_position( decoder ); + if( verbosity >= 2 || ( verbosity == 1 && first_member ) ) + Pp_show_msg( pp, 0 ); + if( verbosity >= 2 ) + { + if( verbosity >= 4 ) + show_header( LZ_decompress_dictionary_size( decoder ) ); + if( data_size == 0 || member_size == 0 ) + fputs( "no data compressed. ", stderr ); + else + fprintf( stderr, "%6.3f:1, %5.2f%% ratio, %5.2f%% saved. ", + (double)data_size / member_size, + ( 100.0 * member_size ) / data_size, + 100.0 - ( ( 100.0 * member_size ) / data_size ) ); + if( verbosity >= 4 ) + fprintf( stderr, "CRC %08X, ", LZ_decompress_data_crc( decoder ) ); + if( verbosity >= 3 ) + fprintf( stderr, "%9llu out, %8llu in. ", data_size, member_size ); + fputs( testing ? "ok\n" : "done\n", stderr ); Pp_reset( pp ); + } + } + first_member = false; /* member decompressed successfully */ + } + if( rd <= 0 ) break; + } + if( out_size < 0 || ( first_member && out_size == 0 ) ) + { + const unsigned long long member_pos = LZ_decompress_member_position( decoder ); + const enum LZ_Errno lz_errno = LZ_decompress_errno( decoder ); + if( lz_errno == LZ_library_error ) + internal_error( "library error (LZ_decompress_read)." ); + if( member_pos <= 6 ) + { + if( lz_errno == LZ_unexpected_eof ) + { + if( first_member ) + show_file_error( pp->name, "File ends unexpectedly at member header.", 0 ); + else + Pp_show_msg( pp, "Truncated header in multimember file." ); + return 2; + } + else if( lz_errno == LZ_data_error ) + { + if( member_pos == 4 ) + { if( verbosity >= 0 ) + { Pp_show_msg( pp, 0 ); + fprintf( stderr, "Version %d member format not supported.\n", + LZ_decompress_member_version( decoder ) ); } } + else if( member_pos == 5 ) + Pp_show_msg( pp, "Invalid dictionary size in member header." ); + else if( first_member ) /* for lzlib older than 1.10 */ + Pp_show_msg( pp, "Bad version or dictionary size in member header." ); + else if( !loose_trailing ) + Pp_show_msg( pp, "Corrupt header in multimember file." ); + else if( !ignore_trailing ) + Pp_show_msg( pp, "Trailing data not allowed." ); + else break; /* trailing data */ + return 2; + } + } + if( lz_errno == LZ_header_error ) + { + if( first_member ) + show_file_error( pp->name, + "Bad magic number (file not in lzip format).", 0 ); + else if( !ignore_trailing ) + Pp_show_msg( pp, "Trailing data not allowed." ); + else break; /* trailing data */ + return 2; + } + if( lz_errno == LZ_mem_error ) { Pp_show_msg( pp, mem_msg ); return 1; } + if( verbosity >= 0 ) + { + Pp_show_msg( pp, 0 ); + fprintf( stderr, "%s at pos %llu\n", ( lz_errno == LZ_unexpected_eof ) ? + "File ends unexpectedly" : "Decoder error", + LZ_decompress_total_in_size( decoder ) ); + } + return 2; + } + if( LZ_decompress_finished( decoder ) == 1 ) break; + if( in_size == 0 && out_size == 0 ) + { + const unsigned long long size = LZ_decompress_total_in_size( decoder ); + if( total_in == size ) internal_error( "library error (stalled)." ); + total_in = size; + } + } + if( verbosity == 1 ) fputs( testing ? "ok\n" : "done\n", stderr ); + return 0; + } + + +static int decompress( const int infd, struct Pretty_print * const pp, + const bool ignore_trailing, + const bool loose_trailing, const bool testing ) + { + struct LZ_Decoder * const decoder = LZ_decompress_open(); + int retval; + + if( !decoder || LZ_decompress_errno( decoder ) != LZ_ok ) + { Pp_show_msg( pp, mem_msg ); retval = 1; } + else retval = do_decompress( decoder, infd, pp, ignore_trailing, + loose_trailing, testing ); + LZ_decompress_close( decoder ); + return retval; + } + + +static void show_error( const char * const msg, const int errcode, + const bool help ) + { + if( verbosity < 0 ) return; + if( msg && msg[0] ) + fprintf( stderr, "%s: %s%s%s\n", program_name, msg, + ( errcode > 0 ) ? ": " : "", + ( errcode > 0 ) ? strerror( errcode ) : "" ); + if( help ) + fprintf( stderr, "Try '%s --help' for more information.\n", + invocation_name ); + } + + +static void show_file_error( const char * const filename, + const char * const msg, const int errcode ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: %s: %s%s%s\n", program_name, filename, msg, + ( errcode > 0 ) ? ": " : "", + ( errcode > 0 ) ? strerror( errcode ) : "" ); + } + + +static void internal_error( const char * const msg ) + { + if( verbosity >= 0 ) + fprintf( stderr, "%s: internal error: %s\n", program_name, msg ); + exit( 3 ); + } + + +int main( const int argc, const char * const argv[] ) + { + /* Mapping from gzip/bzip2 style 1..9 compression modes + to the corresponding LZMA compression modes. */ + const struct Lzma_options option_mapping[] = + { + { 65535, 16 }, /* -0 (65535,16 chooses fast encoder) */ + { 1 << 20, 5 }, /* -1 */ + { 3 << 19, 6 }, /* -2 */ + { 1 << 21, 8 }, /* -3 */ + { 3 << 20, 12 }, /* -4 */ + { 1 << 22, 20 }, /* -5 */ + { 1 << 23, 36 }, /* -6 */ + { 1 << 24, 68 }, /* -7 */ + { 3 << 23, 132 }, /* -8 */ + { 1 << 25, 273 } }; /* -9 */ + struct Lzma_options encoder_options = option_mapping[6]; /* default = "-6" */ + const unsigned long long max_member_size = 0x0008000000000000ULL; /* 2 PiB */ + const unsigned long long max_volume_size = 0x4000000000000000ULL; /* 4 EiB */ + unsigned long long member_size = max_member_size; + unsigned long long volume_size = 0; + const char * default_output_filename = ""; + enum Mode program_mode = m_compress; + int i; + bool force = false; + bool ignore_trailing = true; + bool keep_input_files = false; + bool loose_trailing = false; + bool recompress = false; + bool to_stdout = false; + if( argc > 0 ) invocation_name = argv[0]; + + enum { opt_chk = 256, opt_lt }; + const struct ap_Option options[] = + { + { '0', "fast", ap_no }, + { '1', 0, ap_no }, + { '2', 0, ap_no }, + { '3', 0, ap_no }, + { '4', 0, ap_no }, + { '5', 0, ap_no }, + { '6', 0, ap_no }, + { '7', 0, ap_no }, + { '8', 0, ap_no }, + { '9', "best", ap_no }, + { 'a', "trailing-error", ap_no }, + { 'b', "member-size", ap_yes }, + { 'c', "stdout", ap_no }, + { 'd', "decompress", ap_no }, + { 'f', "force", ap_no }, + { 'F', "recompress", ap_no }, + { 'h', "help", ap_no }, + { 'k', "keep", ap_no }, + { 'm', "match-length", ap_yes }, + { 'n', "threads", ap_yes }, + { 'o', "output", ap_yes }, + { 'q', "quiet", ap_no }, + { 's', "dictionary-size", ap_yes }, + { 'S', "volume-size", ap_yes }, + { 't', "test", ap_no }, + { 'v', "verbose", ap_no }, + { 'V', "version", ap_no }, + { opt_chk, "check-lib", ap_no }, + { opt_lt, "loose-trailing", ap_no }, + { 0, 0, ap_no } }; + + /* static because valgrind complains and memory management in C sucks */ + static struct Arg_parser parser; + if( !ap_init( &parser, argc, argv, options, 0 ) ) + { show_error( mem_msg, 0, false ); return 1; } + if( ap_error( &parser ) ) /* bad option */ + { show_error( ap_error( &parser ), 0, true ); return 1; } + + int argind = 0; + for( ; argind < ap_arguments( &parser ); ++argind ) + { + const int code = ap_code( &parser, argind ); + if( !code ) break; /* no more options */ + const char * const pn = ap_parsed_name( &parser, argind ); + const char * const arg = ap_argument( &parser, argind ); + switch( code ) + { + case '0': case '1': case '2': case '3': case '4': + case '5': case '6': case '7': case '8': case '9': + encoder_options = option_mapping[code-'0']; break; + case 'a': ignore_trailing = false; break; + case 'b': member_size = getnum( arg, pn, 100000, max_member_size ); break; + case 'c': to_stdout = true; break; + case 'd': set_mode( &program_mode, m_decompress ); break; + case 'f': force = true; break; + case 'F': recompress = true; break; + case 'h': show_help(); return 0; + case 'k': keep_input_files = true; break; + case 'm': encoder_options.match_len_limit = + getnum( arg, pn, LZ_min_match_len_limit(), + LZ_max_match_len_limit() ); break; + case 'n': break; + case 'o': if( strcmp( arg, "-" ) == 0 ) to_stdout = true; + else { default_output_filename = arg; } break; + case 'q': verbosity = -1; break; + case 's': encoder_options.dictionary_size = get_dict_size( arg, pn ); + break; + case 'S': volume_size = getnum( arg, pn, 100000, max_volume_size ); break; + case 't': set_mode( &program_mode, m_test ); break; + case 'v': if( verbosity < 4 ) ++verbosity; break; + case 'V': show_version(); return 0; + case opt_chk: return check_lib(); + case opt_lt: loose_trailing = true; break; + default : internal_error( "uncaught option." ); + } + } /* end process options */ + + if( strcmp( PROGVERSION, LZ_version_string ) != 0 ) + internal_error( "wrong PROGVERSION." ); +#if !defined LZ_API_VERSION || LZ_API_VERSION < 1012 +#error "lzlib 1.12 or newer needed." +#else + if( LZ_api_version() < 1012 ) /* minilzip passes null to LZ_decompress_read */ + { show_error( "lzlib 1.12 or newer needed. Try --check-lib.", 0, false ); + return 1; } + if( LZ_api_version() != LZ_API_VERSION ) show_error( + "warning: wrong library API version. Try --check-lib.", 0, false ); + else +#endif + if( strcmp( LZ_version_string, LZ_version() ) != 0 ) show_error( + "warning: wrong library version_string. Try --check-lib.", 0, false ); + +#if defined __MSVCRT__ || defined __OS2__ || defined __DJGPP__ + setmode( STDIN_FILENO, O_BINARY ); + setmode( STDOUT_FILENO, O_BINARY ); +#endif + + static const char ** filenames = 0; + int num_filenames = max( 1, ap_arguments( &parser ) - argind ); + filenames = resize_buffer( filenames, num_filenames * sizeof filenames[0] ); + filenames[0] = "-"; + + bool filenames_given = false; + for( i = 0; argind + i < ap_arguments( &parser ); ++i ) + { + filenames[i] = ap_argument( &parser, argind + i ); + if( strcmp( filenames[i], "-" ) != 0 ) filenames_given = true; + } + + if( program_mode == m_compress ) + { + if( volume_size > 0 && !to_stdout && default_output_filename[0] && + num_filenames > 1 ) + { show_error( "Only can compress one file when using '-o' and '-S'.", + 0, true ); return 1; } + } + else volume_size = 0; + if( program_mode == m_test ) to_stdout = false; /* apply overrides */ + if( program_mode == m_test || to_stdout ) default_output_filename = ""; + + output_filename = resize_buffer( output_filename, 1 ); + output_filename[0] = 0; + if( to_stdout && program_mode != m_test ) /* check tty only once */ + { outfd = STDOUT_FILENO; if( !check_tty_out( program_mode ) ) return 1; } + else outfd = -1; + + const bool to_file = !to_stdout && program_mode != m_test && + default_output_filename[0]; + if( !to_stdout && program_mode != m_test && ( filenames_given || to_file ) ) + set_signals( signal_handler ); + + static struct Pretty_print pp; + Pp_init( &pp, filenames, num_filenames ); + + int failed_tests = 0; + int retval = 0; + const bool one_to_one = !to_stdout && program_mode != m_test && !to_file; + bool stdin_used = false; + for( i = 0; i < num_filenames; ++i ) + { + const char * input_filename = ""; + int infd; + struct stat in_stats; + + Pp_set_name( &pp, filenames[i] ); + if( strcmp( filenames[i], "-" ) == 0 ) + { + if( stdin_used ) continue; else stdin_used = true; + infd = STDIN_FILENO; + if( !check_tty_in( pp.name, infd, program_mode, &retval ) ) continue; + if( one_to_one ) { outfd = STDOUT_FILENO; output_filename[0] = 0; } + } + else + { + const int eindex = extension_index( input_filename = filenames[i] ); + infd = open_instream( input_filename, &in_stats, program_mode, + eindex, one_to_one, recompress ); + if( infd < 0 ) { set_retval( &retval, 1 ); continue; } + if( !check_tty_in( pp.name, infd, program_mode, &retval ) ) continue; + if( one_to_one ) /* open outfd after verifying infd */ + { + if( program_mode == m_compress ) + set_c_outname( input_filename, true, volume_size > 0 ); + else set_d_outname( input_filename, eindex ); + if( !open_outstream( force, true ) ) + { close( infd ); set_retval( &retval, 1 ); continue; } + } + } + + if( one_to_one && !check_tty_out( program_mode ) ) + { set_retval( &retval, 1 ); return retval; } /* don't delete a tty */ + + if( to_file && outfd < 0 ) /* open outfd after verifying infd */ + { + if( program_mode == m_compress ) set_c_outname( default_output_filename, + false, volume_size > 0 ); + else + { output_filename = resize_buffer( output_filename, + strlen( default_output_filename ) + 1 ); + strcpy( output_filename, default_output_filename ); } + if( !open_outstream( force, false ) || !check_tty_out( program_mode ) ) + return 1; /* check tty only once and don't try to delete a tty */ + } + + const struct stat * const in_statsp = + ( input_filename[0] && one_to_one ) ? &in_stats : 0; + int tmp; + if( program_mode == m_compress ) + tmp = compress( member_size, volume_size, infd, &encoder_options, &pp, + in_statsp ); + else + tmp = decompress( infd, &pp, ignore_trailing, + loose_trailing, program_mode == m_test ); + if( close( infd ) != 0 ) + { show_file_error( pp.name, "Error closing input file", errno ); + set_retval( &tmp, 1 ); } + set_retval( &retval, tmp ); + if( tmp ) + { if( program_mode != m_test ) cleanup_and_fail( retval ); + else ++failed_tests; } + + if( delete_output_on_interrupt && one_to_one ) + close_and_set_permissions( in_statsp ); + if( input_filename[0] && !keep_input_files && one_to_one && + ( program_mode != m_compress || volume_size == 0 ) ) + remove( input_filename ); + } + if( delete_output_on_interrupt ) close_and_set_permissions( 0 ); /* -o */ + else if( outfd >= 0 && close( outfd ) != 0 ) /* -c */ + { + show_error( "Error closing stdout", errno, false ); + set_retval( &retval, 1 ); + } + if( failed_tests > 0 && verbosity >= 1 && num_filenames > 1 ) + fprintf( stderr, "%s: warning: %d %s failed the test.\n", + program_name, failed_tests, + ( failed_tests == 1 ) ? "file" : "files" ); + free( output_filename ); + free( filenames ); + ap_free( &parser ); + return retval; + } diff --git a/testsuite/check.sh b/testsuite/check.sh new file mode 100755 index 0000000..e93697e --- /dev/null +++ b/testsuite/check.sh @@ -0,0 +1,444 @@ +#! /bin/sh +# check script for Lzlib - Compression library for the lzip format +# Copyright (C) 2009-2022 Antonio Diaz Diaz. +# +# This script is free software: you have unlimited permission +# to copy, distribute, and modify it. + +LC_ALL=C +export LC_ALL +objdir=`pwd` +testdir=`cd "$1" ; pwd` +LZIP="${objdir}"/minilzip +BBEXAMPLE="${objdir}"/bbexample +FFEXAMPLE="${objdir}"/ffexample +LZCHECK="${objdir}"/lzcheck +framework_failure() { echo "failure in testing framework" ; exit 1 ; } + +if [ ! -f "${LZIP}" ] || [ ! -x "${LZIP}" ] ; then + echo "${LZIP}: cannot execute" + exit 1 +fi + +[ -e "${LZIP}" ] 2> /dev/null || + { + echo "$0: a POSIX shell is required to run the tests" + echo "Try bash -c \"$0 $1 $2\"" + exit 1 + } + +if [ -d tmp ] ; then rm -rf tmp ; fi +mkdir tmp +cd "${objdir}"/tmp || framework_failure + +cat "${testdir}"/test.txt > in || framework_failure +in_lz="${testdir}"/test.txt.lz +in_em="${testdir}"/test_em.txt.lz +fox_lf="${testdir}"/fox_lf +fox_lz="${testdir}"/fox.lz +fail=0 +test_failed() { fail=1 ; printf " $1" ; [ -z "$2" ] || printf "($2)" ; } + +"${LZIP}" --check-lib # just print warning +[ $? != 2 ] || { test_failed $LINENO ; exit 2 ; } # unless bad lzlib.h +printf "testing lzlib-%s..." "$2" + +"${LZIP}" -fkqm4 in +[ $? = 1 ] || test_failed $LINENO +[ ! -e in.lz ] || test_failed $LINENO +"${LZIP}" -fkqm274 in +[ $? = 1 ] || test_failed $LINENO +[ ! -e in.lz ] || test_failed $LINENO +for i in bad_size -1 0 4095 513MiB 1G 1T 1P 1E 1Z 1Y 10KB ; do + "${LZIP}" -fkqs $i in + [ $? = 1 ] || test_failed $LINENO $i + [ ! -e in.lz ] || test_failed $LINENO $i +done +"${LZIP}" -tq in +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -tq < in +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -cdq in +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -cdq < in +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -dq -o in < "${in_lz}" +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -dq -o in "${in_lz}" +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -dq -o out nx_file.lz +[ $? = 1 ] || test_failed $LINENO +[ ! -e out ] || test_failed $LINENO +"${LZIP}" -q -o out.lz nx_file +[ $? = 1 ] || test_failed $LINENO +[ ! -e out.lz ] || test_failed $LINENO +"${LZIP}" -qf -S100k -o out in in +[ $? = 1 ] || test_failed $LINENO +# these are for code coverage +"${LZIP}" -cdt "${in_lz}" > out 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -t -- nx_file.lz 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -t "" < /dev/null 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" --help > /dev/null || test_failed $LINENO +"${LZIP}" -n1 -V > /dev/null || test_failed $LINENO +"${LZIP}" -m 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -z 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" --bad_option 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" --t 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" --test=2 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" --output= 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" --output 2> /dev/null +[ $? = 1 ] || test_failed $LINENO +printf "LZIP\001-.............................." | "${LZIP}" -t 2> /dev/null +printf "LZIP\002-.............................." | "${LZIP}" -t 2> /dev/null +printf "LZIP\001+.............................." | "${LZIP}" -t 2> /dev/null +rm -f out || framework_failure + +printf "\ntesting decompression..." + +for i in "${in_lz}" "${in_em}" "${testdir}"/test_sync.lz ; do + "${LZIP}" -t "$i" || test_failed $LINENO "$i" + "${LZIP}" -d "$i" -o copy || test_failed $LINENO "$i" + cmp in copy || test_failed $LINENO "$i" + "${LZIP}" -cd "$i" > copy || test_failed $LINENO "$i" + cmp in copy || test_failed $LINENO "$i" + "${LZIP}" -d "$i" -o - > copy || test_failed $LINENO "$i" + cmp in copy || test_failed $LINENO "$i" + "${LZIP}" -d < "$i" > copy || test_failed $LINENO "$i" + cmp in copy || test_failed $LINENO "$i" + rm -f copy || framework_failure +done + +lines=$("${LZIP}" -tvv "${in_em}" 2>&1 | wc -l) || test_failed $LINENO +[ "${lines}" -eq 8 ] || test_failed $LINENO "${lines}" + +"${LZIP}" -cd "${fox_lz}" > fox || test_failed $LINENO +cat "${in_lz}" > copy.lz || framework_failure +"${LZIP}" -dk copy.lz || test_failed $LINENO +cmp in copy || test_failed $LINENO +cat fox > copy || framework_failure +cat "${in_lz}" > out.lz || framework_failure +rm -f out || framework_failure +"${LZIP}" -d copy.lz out.lz 2> /dev/null # skip copy, decompress out +[ $? = 1 ] || test_failed $LINENO +cmp fox copy || test_failed $LINENO +cmp in out || test_failed $LINENO +"${LZIP}" -df copy.lz || test_failed $LINENO +[ ! -e copy.lz ] || test_failed $LINENO +cmp in copy || test_failed $LINENO +rm -f copy out || framework_failure + +cat "${in_lz}" > copy.lz || framework_failure +"${LZIP}" -d -S100k copy.lz || test_failed $LINENO # ignore -S +[ ! -e copy.lz ] || test_failed $LINENO +cmp in copy || test_failed $LINENO + +printf "to be overwritten" > copy || framework_failure +"${LZIP}" -df -o copy < "${in_lz}" || test_failed $LINENO +cmp in copy || test_failed $LINENO +rm -f out copy || framework_failure +"${LZIP}" -d -o ./- "${in_lz}" || test_failed $LINENO +cmp in ./- || test_failed $LINENO +rm -f ./- || framework_failure +"${LZIP}" -d -o ./- < "${in_lz}" || test_failed $LINENO +cmp in ./- || test_failed $LINENO +rm -f ./- || framework_failure + +cat "${in_lz}" > anyothername || framework_failure +"${LZIP}" -dv - anyothername - < "${in_lz}" > copy 2> /dev/null || + test_failed $LINENO +cmp in copy || test_failed $LINENO +cmp in anyothername.out || test_failed $LINENO +rm -f copy anyothername.out || framework_failure + +"${LZIP}" -tq in "${in_lz}" +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -tq nx_file.lz "${in_lz}" +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -cdq in "${in_lz}" > copy +[ $? = 2 ] || test_failed $LINENO +cat copy in | cmp in - || test_failed $LINENO # copy must be empty +"${LZIP}" -cdq nx_file.lz "${in_lz}" > copy +[ $? = 1 ] || test_failed $LINENO +cmp in copy || test_failed $LINENO +rm -f copy || framework_failure +cat "${in_lz}" > copy.lz || framework_failure +for i in 1 2 3 4 5 6 7 ; do + printf "g" >> copy.lz || framework_failure + "${LZIP}" -atvvvv copy.lz "${in_lz}" 2> /dev/null + [ $? = 2 ] || test_failed $LINENO $i +done +"${LZIP}" -dq in copy.lz +[ $? = 2 ] || test_failed $LINENO +[ -e copy.lz ] || test_failed $LINENO +[ ! -e copy ] || test_failed $LINENO +[ ! -e in.out ] || test_failed $LINENO +"${LZIP}" -dq nx_file.lz copy.lz +[ $? = 1 ] || test_failed $LINENO +[ ! -e copy.lz ] || test_failed $LINENO +[ ! -e nx_file ] || test_failed $LINENO +cmp in copy || test_failed $LINENO + +cat in in > in2 || framework_failure +"${LZIP}" -t "${in_lz}" "${in_lz}" || test_failed $LINENO +"${LZIP}" -cd "${in_lz}" "${in_lz}" -o out > copy2 || test_failed $LINENO +[ ! -e out ] || test_failed $LINENO # override -o +cmp in2 copy2 || test_failed $LINENO +rm -f copy2 || framework_failure +"${LZIP}" -d "${in_lz}" "${in_lz}" -o copy2 || test_failed $LINENO +cmp in2 copy2 || test_failed $LINENO +rm -f copy2 || framework_failure + +cat "${in_lz}" "${in_lz}" > copy2.lz || framework_failure +printf "\ngarbage" >> copy2.lz || framework_failure +"${LZIP}" -tvvvv copy2.lz 2> /dev/null || test_failed $LINENO +"${LZIP}" -atq copy2.lz +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -atq < copy2.lz +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -adkq copy2.lz +[ $? = 2 ] || test_failed $LINENO +[ ! -e copy2 ] || test_failed $LINENO +"${LZIP}" -adkq -o copy2 < copy2.lz +[ $? = 2 ] || test_failed $LINENO +[ ! -e copy2 ] || test_failed $LINENO +printf "to be overwritten" > copy2 || framework_failure +"${LZIP}" -df copy2.lz || test_failed $LINENO +cmp in2 copy2 || test_failed $LINENO +rm -f copy2 || framework_failure + +printf "\ntesting compression..." + +"${LZIP}" -c -0 in in in -S100k -o out3.lz > copy2.lz || test_failed $LINENO +[ ! -e out3.lz ] || test_failed $LINENO # override -o and -S +"${LZIP}" -0f in in --output=copy2.lz || test_failed $LINENO +"${LZIP}" -d copy2.lz -o out2 || test_failed $LINENO +cmp in2 out2 || test_failed $LINENO +rm -f out2 copy2.lz || framework_failure + +"${LZIP}" -cf "${in_lz}" > out 2> /dev/null # /dev/null is a tty on OS/2 +[ $? = 1 ] || test_failed $LINENO +"${LZIP}" -Fvvm36 -o - -s16 "${in_lz}" > out 2> /dev/null || test_failed $LINENO +"${LZIP}" -cd out | "${LZIP}" -d > copy || test_failed $LINENO +cmp in copy || test_failed $LINENO + +"${LZIP}" -0 -o ./- in || test_failed $LINENO +"${LZIP}" -cd ./- | cmp in - || test_failed $LINENO +rm -f ./- || framework_failure +"${LZIP}" -0 -o ./- < in || test_failed $LINENO # don't add .lz +[ ! -e ./-.lz ] || test_failed $LINENO +"${LZIP}" -cd ./- | cmp in - || test_failed $LINENO +rm -f ./- || framework_failure + +for i in s4Ki 0 1 2 3 4 5 6 7 8 9 ; do + "${LZIP}" -k -$i -s16 in || test_failed $LINENO $i + mv -f in.lz copy.lz || test_failed $LINENO $i + printf "garbage" >> copy.lz || framework_failure + "${LZIP}" -df copy.lz || test_failed $LINENO $i + cmp in copy || test_failed $LINENO $i + + "${LZIP}" -$i -s16 in -c > out || test_failed $LINENO $i + "${LZIP}" -$i -s16 in -o o_out || test_failed $LINENO $i # don't add .lz + [ ! -e o_out.lz ] || test_failed $LINENO + cmp out o_out || test_failed $LINENO $i + rm -f o_out || framework_failure + printf "g" >> out || framework_failure + "${LZIP}" -cd out > copy || test_failed $LINENO $i + cmp in copy || test_failed $LINENO $i + + "${LZIP}" -$i -s16 < in > out || test_failed $LINENO $i + "${LZIP}" -d < out > copy || test_failed $LINENO $i + cmp in copy || test_failed $LINENO $i + + rm -f out.lz || framework_failure + printf "to be overwritten" > out || framework_failure # don't add .lz + "${LZIP}" -f -$i -s16 -o out < in || test_failed $LINENO $i + [ ! -e out.lz ] || test_failed $LINENO + "${LZIP}" -df -o copy < out || test_failed $LINENO $i + cmp in copy || test_failed $LINENO $i +done +rm -f out out.lz || framework_failure + +cat in in in in in in in in > in8 || framework_failure +"${LZIP}" -1s12 -S100k in8 || test_failed $LINENO +"${LZIP}" -t in800001.lz in800002.lz || test_failed $LINENO +"${LZIP}" -cd in800001.lz in800002.lz | cmp in8 - || test_failed $LINENO +[ ! -e in800003.lz ] || test_failed $LINENO +rm -f in800001.lz in800002.lz || framework_failure +"${LZIP}" -1s12 -S100k -o out.lz in8 || test_failed $LINENO +# ignore -S +"${LZIP}" -d out.lz00001.lz out.lz00002.lz -S100k -o out || test_failed $LINENO +cmp in8 out || test_failed $LINENO +"${LZIP}" -t out.lz00001.lz out.lz00002.lz || test_failed $LINENO +[ ! -e out.lz00003.lz ] || test_failed $LINENO +rm -f out out.lz00001.lz out.lz00002.lz || framework_failure +"${LZIP}" -1ks4Ki -b100000 in8 || test_failed $LINENO +"${LZIP}" -t in8.lz || test_failed $LINENO +"${LZIP}" -cd in8.lz -o out | cmp in8 - || test_failed $LINENO # override -o +[ ! -e out ] || test_failed $LINENO +"${LZIP}" -0 -S100k -o out < in8.lz || test_failed $LINENO +"${LZIP}" -t out00001.lz out00002.lz || test_failed $LINENO +"${LZIP}" -cd out00001.lz out00002.lz | cmp in8.lz - || test_failed $LINENO +[ ! -e out00003.lz ] || test_failed $LINENO +rm -f out00001.lz || framework_failure +"${LZIP}" -1 -S100k -o out < in8.lz || test_failed $LINENO +"${LZIP}" -t out00001.lz out00002.lz || test_failed $LINENO +"${LZIP}" -cd out00001.lz out00002.lz | cmp in8.lz - || test_failed $LINENO +[ ! -e out00003.lz ] || test_failed $LINENO +rm -f out00001.lz out00002.lz || framework_failure +"${LZIP}" -0 -F -S100k in8.lz || test_failed $LINENO +"${LZIP}" -t in8.lz00001.lz in8.lz00002.lz || test_failed $LINENO +"${LZIP}" -cd in8.lz00001.lz in8.lz00002.lz | cmp in8.lz - || test_failed $LINENO +[ ! -e in8.lz00003.lz ] || test_failed $LINENO +rm -f in8.lz00001.lz in8.lz00002.lz || framework_failure +"${LZIP}" -0kF -b100k in8.lz || test_failed $LINENO +"${LZIP}" -t in8.lz.lz || test_failed $LINENO +"${LZIP}" -cd in8.lz.lz | cmp in8.lz - || test_failed $LINENO +rm -f in8.lz in8.lz.lz || framework_failure + +"${BBEXAMPLE}" in || test_failed $LINENO +"${BBEXAMPLE}" "${in_lz}" || test_failed $LINENO +"${BBEXAMPLE}" "${fox_lf}" || test_failed $LINENO + +"${FFEXAMPLE}" -h > /dev/null || test_failed $LINENO +"${FFEXAMPLE}" > /dev/null && test_failed $LINENO +rm -f out || framework_failure +"${FFEXAMPLE}" -b in out || test_failed $LINENO +cmp in out || test_failed $LINENO +"${FFEXAMPLE}" -b in | cmp in - || test_failed $LINENO +"${FFEXAMPLE}" -b in8 | cmp in8 - || test_failed $LINENO +"${FFEXAMPLE}" -b "${fox_lf}" | cmp "${fox_lf}" - || test_failed $LINENO +"${FFEXAMPLE}" -d "${in_lz}" - | cmp in - || test_failed $LINENO +"${FFEXAMPLE}" -d "${in_em}" - | cmp in - || test_failed $LINENO +"${FFEXAMPLE}" -c in | "${FFEXAMPLE}" -d | cmp in - || test_failed $LINENO +"${FFEXAMPLE}" -m in | "${FFEXAMPLE}" -d | cmp in - || test_failed $LINENO +"${FFEXAMPLE}" -l in | "${FFEXAMPLE}" -d | cmp in - || test_failed $LINENO +cat "${fox_lf}" "${in_lz}" | "${FFEXAMPLE}" -r | cmp in - || test_failed $LINENO +cat in8 "${in_lz}" | "${FFEXAMPLE}" -r | cmp in - || test_failed $LINENO +cat "${in_lz}" "${fox_lf}" "${in_lz}" | "${FFEXAMPLE}" -r - | cmp in2 - || + test_failed $LINENO +cat "${in_lz}" in8 "${in_lz}" | "${FFEXAMPLE}" -r - - | cmp in2 - || + test_failed $LINENO + +"${LZCHECK}" in || test_failed $LINENO +"${LZCHECK}" "${in_lz}" || test_failed $LINENO +"${LZCHECK}" "${fox_lf}" || test_failed $LINENO +rm -f in8 || framework_failure + +printf "\ntesting bad input..." + +headers='LZIp LZiP LZip LzIP LzIp LziP lZIP lZIp lZiP lzIP' +body='\001\014\000\203\377\373\377\377\300\000\000\000\000\000\000\000\000\000\000\000\000\000\000\000$\000\000\000\000\000\000\000' +cat "${in_lz}" > int.lz +printf "LZIP${body}" >> int.lz +if "${LZIP}" -tq int.lz ; then + for header in ${headers} ; do + printf "${header}${body}" > int.lz # first member + "${LZIP}" -tq int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -tq < int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -cdq int.lz > /dev/null + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -tq --loose-trailing int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -tq --loose-trailing < int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -cdq --loose-trailing int.lz > /dev/null + [ $? = 2 ] || test_failed $LINENO ${header} + cat "${in_lz}" > int.lz + printf "${header}${body}" >> int.lz # trailing data + "${LZIP}" -tq int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -tq < int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -cdq int.lz > /dev/null + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -t --loose-trailing int.lz || + test_failed $LINENO ${header} + "${LZIP}" -t --loose-trailing < int.lz || + test_failed $LINENO ${header} + "${LZIP}" -cd --loose-trailing int.lz > /dev/null || + test_failed $LINENO ${header} + "${LZIP}" -tq --loose-trailing --trailing-error int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -tq --loose-trailing --trailing-error < int.lz + [ $? = 2 ] || test_failed $LINENO ${header} + "${LZIP}" -cdq --loose-trailing --trailing-error int.lz > /dev/null + [ $? = 2 ] || test_failed $LINENO ${header} + done +else + printf "\nwarning: skipping header test: 'printf' does not work on your system." +fi +rm -f int.lz || framework_failure + +for i in fox_v2.lz fox_s11.lz fox_de20.lz \ + fox_bcrc.lz fox_crc0.lz fox_das46.lz fox_mes81.lz ; do + "${LZIP}" -tq "${testdir}"/$i + [ $? = 2 ] || test_failed $LINENO $i +done + +for i in fox_bcrc.lz fox_crc0.lz fox_das46.lz fox_mes81.lz ; do + "${LZIP}" -cdq "${testdir}"/$i > out + [ $? = 2 ] || test_failed $LINENO $i + cmp fox out || test_failed $LINENO $i +done +rm -f fox out || framework_failure + +cat "${in_lz}" "${in_lz}" > in2.lz || framework_failure +cat "${in_lz}" "${in_lz}" "${in_lz}" > in3.lz || framework_failure +if dd if=in3.lz of=trunc.lz bs=14752 count=1 2> /dev/null && + [ -e trunc.lz ] && cmp in2.lz trunc.lz > /dev/null 2>&1 ; then + for i in 6 20 14734 14753 14754 14755 14756 14757 14758 ; do + dd if=in3.lz of=trunc.lz bs=$i count=1 2> /dev/null + "${LZIP}" -tq trunc.lz + [ $? = 2 ] || test_failed $LINENO $i + "${LZIP}" -tq < trunc.lz + [ $? = 2 ] || test_failed $LINENO $i + "${LZIP}" -cdq trunc.lz > out + [ $? = 2 ] || test_failed $LINENO $i + "${LZIP}" -dq < trunc.lz > out + [ $? = 2 ] || test_failed $LINENO $i + done +else + printf "\nwarning: skipping truncation test: 'dd' does not work on your system." +fi +rm -f in2.lz in3.lz trunc.lz out || framework_failure + +cat "${in_lz}" > ingin.lz || framework_failure +printf "g" >> ingin.lz || framework_failure +cat "${in_lz}" >> ingin.lz || framework_failure +"${LZIP}" -atq ingin.lz +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -atq < ingin.lz +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -acdq ingin.lz > out +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -adq < ingin.lz > out +[ $? = 2 ] || test_failed $LINENO +"${LZIP}" -t ingin.lz || test_failed $LINENO +"${LZIP}" -t < ingin.lz || test_failed $LINENO +"${LZIP}" -cd ingin.lz > copy || test_failed $LINENO +cmp in copy || test_failed $LINENO +"${LZIP}" -d < ingin.lz > copy || test_failed $LINENO +cmp in copy || test_failed $LINENO +"${FFEXAMPLE}" -d ingin.lz | cmp in - || test_failed $LINENO +"${FFEXAMPLE}" -r ingin.lz | cmp in2 - || test_failed $LINENO +rm -f copy ingin.lz in2 out || framework_failure + +echo +if [ ${fail} = 0 ] ; then + echo "tests completed successfully." + cd "${objdir}" && rm -r tmp +else + echo "tests failed." +fi +exit ${fail} diff --git a/testsuite/fox.lz b/testsuite/fox.lz new file mode 100644 index 0000000..509da82 Binary files /dev/null and b/testsuite/fox.lz differ diff --git a/testsuite/fox_bcrc.lz b/testsuite/fox_bcrc.lz new file mode 100644 index 0000000..8f6a7c4 Binary files /dev/null and b/testsuite/fox_bcrc.lz differ diff --git a/testsuite/fox_crc0.lz b/testsuite/fox_crc0.lz new file mode 100644 index 0000000..1abe926 Binary files /dev/null and b/testsuite/fox_crc0.lz differ diff --git a/testsuite/fox_das46.lz b/testsuite/fox_das46.lz new file mode 100644 index 0000000..43ed9f9 Binary files /dev/null and b/testsuite/fox_das46.lz differ diff --git a/testsuite/fox_de20.lz b/testsuite/fox_de20.lz new file mode 100644 index 0000000..10949d8 Binary files /dev/null and b/testsuite/fox_de20.lz differ diff --git a/testsuite/fox_lf b/testsuite/fox_lf new file mode 100644 index 0000000..a0b11b5 --- /dev/null +++ b/testsuite/fox_lf @@ -0,0 +1,9 @@ +The +quick +brown +fox +jumps +over +the +lazy +dog. diff --git a/testsuite/fox_mes81.lz b/testsuite/fox_mes81.lz new file mode 100644 index 0000000..d50ef2e Binary files /dev/null and b/testsuite/fox_mes81.lz differ diff --git a/testsuite/fox_s11.lz b/testsuite/fox_s11.lz new file mode 100644 index 0000000..dca909c Binary files /dev/null and b/testsuite/fox_s11.lz differ diff --git a/testsuite/fox_v2.lz b/testsuite/fox_v2.lz new file mode 100644 index 0000000..8620981 Binary files /dev/null and b/testsuite/fox_v2.lz differ diff --git a/testsuite/test.txt b/testsuite/test.txt new file mode 100644 index 0000000..9196a3a --- /dev/null +++ b/testsuite/test.txt @@ -0,0 +1,676 @@ + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 Free Software Foundation, Inc., + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 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 Lesser 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. + + + Copyright (C) + + 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, see . + +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) + 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. + + , 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 Lesser General +Public License instead of this License. + GNU GENERAL PUBLIC LICENSE + Version 2, June 1991 + + Copyright (C) 1989, 1991 Free Software Foundation, Inc., + 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 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 Lesser 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. + + + Copyright (C) + + 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, see . + +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) + 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. + + , 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 Lesser General +Public License instead of this License. diff --git a/testsuite/test.txt.lz b/testsuite/test.txt.lz new file mode 100644 index 0000000..22cea6e Binary files /dev/null and b/testsuite/test.txt.lz differ diff --git a/testsuite/test_em.txt.lz b/testsuite/test_em.txt.lz new file mode 100644 index 0000000..7e96250 Binary files /dev/null and b/testsuite/test_em.txt.lz differ diff --git a/testsuite/test_sync.lz b/testsuite/test_sync.lz new file mode 100644 index 0000000..db680c3 Binary files /dev/null and b/testsuite/test_sync.lz differ -- cgit v1.2.3