diff -Nru bzip2-1.0.2/AUTHORS bzip2-1.0.2.new/AUTHORS --- bzip2-1.0.2/AUTHORS Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/AUTHORS Fri Feb 1 04:18:59 2002 @@ -0,0 +1 @@ +Julian Seward diff -Nru bzip2-1.0.2/CHANGES bzip2-1.0.2.new/CHANGES --- bzip2-1.0.2/CHANGES Sat Jan 26 01:02:57 2002 +++ bzip2-1.0.2.new/CHANGES Thu Jan 1 01:00:00 1970 @@ -1,253 +0,0 @@ - - -0.9.0 -~~~~~ -First version. - - -0.9.0a -~~~~~~ -Removed 'ranlib' from Makefile, since most modern Unix-es -don't need it, or even know about it. - - -0.9.0b -~~~~~~ -Fixed a problem with error reporting in bzip2.c. This does not effect -the library in any way. Problem is: versions 0.9.0 and 0.9.0a (of the -program proper) compress and decompress correctly, but give misleading -error messages (internal panics) when an I/O error occurs, instead of -reporting the problem correctly. This shouldn't give any data loss -(as far as I can see), but is confusing. - -Made the inline declarations disappear for non-GCC compilers. - - -0.9.0c -~~~~~~ -Fixed some problems in the library pertaining to some boundary cases. -This makes the library behave more correctly in those situations. The -fixes apply only to features (calls and parameters) not used by -bzip2.c, so the non-fixedness of them in previous versions has no -effect on reliability of bzip2.c. - -In bzlib.c: - * made zero-length BZ_FLUSH work correctly in bzCompress(). - * fixed bzWrite/bzRead to ignore zero-length requests. - * fixed bzread to correctly handle read requests after EOF. - * wrong parameter order in call to bzDecompressInit in - bzBuffToBuffDecompress. Fixed. - -In compress.c: - * changed setting of nGroups in sendMTFValues() so as to - do a bit better on small files. This _does_ effect - bzip2.c. - - -0.9.5a -~~~~~~ -Major change: add a fallback sorting algorithm (blocksort.c) -to give reasonable behaviour even for very repetitive inputs. -Nuked --repetitive-best and --repetitive-fast since they are -no longer useful. - -Minor changes: mostly a whole bunch of small changes/ -bugfixes in the driver (bzip2.c). Changes pertaining to the -user interface are: - - allow decompression of symlink'd files to stdout - decompress/test files even without .bz2 extension - give more accurate error messages for I/O errors - when compressing/decompressing to stdout, don't catch control-C - read flags from BZIP2 and BZIP environment variables - decline to break hard links to a file unless forced with -f - allow -c flag even with no filenames - preserve file ownerships as far as possible - make -s -1 give the expected block size (100k) - add a flag -q --quiet to suppress nonessential warnings - stop decoding flags after --, so files beginning in - can be handled - resolved inconsistent naming: bzcat or bz2cat ? - bzip2 --help now returns 0 - -Programming-level changes are: - - fixed syntax error in GET_LL4 for Borland C++ 5.02 - let bzBuffToBuffDecompress return BZ_DATA_ERROR{_MAGIC} - fix overshoot of mode-string end in bzopen_or_bzdopen - wrapped bzlib.h in #ifdef __cplusplus ... extern "C" { ... } - close file handles under all error conditions - added minor mods so it compiles with DJGPP out of the box - fixed Makefile so it doesn't give problems with BSD make - fix uninitialised memory reads in dlltest.c - -0.9.5b -~~~~~~ -Open stdin/stdout in binary mode for DJGPP. - -0.9.5c -~~~~~~ -Changed BZ_N_OVERSHOOT to be ... + 2 instead of ... + 1. The + 1 -version could cause the sorted order to be wrong in some extremely -obscure cases. Also changed setting of quadrant in blocksort.c. - -0.9.5d -~~~~~~ -The only functional change is to make bzlibVersion() in the library -return the correct string. This has no effect whatsoever on the -functioning of the bzip2 program or library. Added a couple of casts -so the library compiles without warnings at level 3 in MS Visual -Studio 6.0. Included a Y2K statement in the file Y2K_INFO. All other -changes are minor documentation changes. - -1.0 -~~~ -Several minor bugfixes and enhancements: - -* Large file support. The library uses 64-bit counters to - count the volume of data passing through it. bzip2.c - is now compiled with -D_FILE_OFFSET_BITS=64 to get large - file support from the C library. -v correctly prints out - file sizes greater than 4 gigabytes. All these changes have - been made without assuming a 64-bit platform or a C compiler - which supports 64-bit ints, so, except for the C library - aspect, they are fully portable. - -* Decompression robustness. The library/program should be - robust to any corruption of compressed data, detecting and - handling _all_ corruption, instead of merely relying on - the CRCs. What this means is that the program should - never crash, given corrupted data, and the library should - always return BZ_DATA_ERROR. - -* Fixed an obscure race-condition bug only ever observed on - Solaris, in which, if you were very unlucky and issued - control-C at exactly the wrong time, both input and output - files would be deleted. - -* Don't run out of file handles on test/decompression when - large numbers of files have invalid magic numbers. - -* Avoid library namespace pollution. Prefix all exported - symbols with BZ2_. - -* Minor sorting enhancements from my DCC2000 paper. - -* Advance the version number to 1.0, so as to counteract the - (false-in-this-case) impression some people have that programs - with version numbers less than 1.0 are in some way, experimental, - pre-release versions. - -* Create an initial Makefile-libbz2_so to build a shared library. - Yes, I know I should really use libtool et al ... - -* Make the program exit with 2 instead of 0 when decompression - fails due to a bad magic number (ie, an invalid bzip2 header). - Also exit with 1 (as the manual claims :-) whenever a diagnostic - message would have been printed AND the corresponding operation - is aborted, for example - bzip2: Output file xx already exists. - When a diagnostic message is printed but the operation is not - aborted, for example - bzip2: Can't guess original name for wurble -- using wurble.out - then the exit value 0 is returned, unless some other problem is - also detected. - - I think it corresponds more closely to what the manual claims now. - - -1.0.1 -~~~~~ -* Modified dlltest.c so it uses the new BZ2_ naming scheme. -* Modified makefile-msc to fix minor build probs on Win2k. -* Updated README.COMPILATION.PROBLEMS. - -There are no functionality changes or bug fixes relative to version -1.0.0. This is just a documentation update + a fix for minor Win32 -build problems. For almost everyone, upgrading from 1.0.0 to 1.0.1 is -utterly pointless. Don't bother. - - -1.0.2 -~~~~~ -A bug fix release, addressing various minor issues which have appeared -in the 18 or so months since 1.0.1 was released. Most of the fixes -are to do with file-handling or documentation bugs. To the best of my -knowledge, there have been no data-loss-causing bugs reported in the -compression/decompression engine of 1.0.0 or 1.0.1. - -Note that this release does not improve the rather crude build system -for Unix platforms. The general plan here is to autoconfiscate/ -libtoolise 1.0.2 soon after release, and release the result as 1.1.0 -or perhaps 1.2.0. That, however, is still just a plan at this point. - -Here are the changes in 1.0.2. Bug-reporters and/or patch-senders in -parentheses. - -* Fix an infinite segfault loop in 1.0.1 when a directory is - encountered in -f (force) mode. - (Trond Eivind Glomsrod, Nicholas Nethercote, Volker Schmidt) - -* Avoid double fclose() of output file on certain I/O error paths. - (Solar Designer) - -* Don't fail with internal error 1007 when fed a long stream (> 48MB) - of byte 251. Also print useful message suggesting that 1007s may be - caused by bad memory. - (noticed by Juan Pedro Vallejo, fixed by me) - -* Fix uninitialised variable silly bug in demo prog dlltest.c. - (Jorj Bauer) - -* Remove 512-MB limitation on recovered file size for bzip2recover - on selected platforms which support 64-bit ints. At the moment - all GCC supported platforms, and Win32. - (me, Alson van der Meulen) - -* Hard-code header byte values, to give correct operation on platforms - using EBCDIC as their native character set (IBM's OS/390). - (Leland Lucius) - -* Copy file access times correctly. - (Marty Leisner) - -* Add distclean and check targets to Makefile. - (Michael Carmack) - -* Parameterise use of ar and ranlib in Makefile. Also add $(LDFLAGS). - (Rich Ireland, Bo Thorsen) - -* Pass -p (create parent dirs as needed) to mkdir during make install. - (Jeremy Fusco) - -* Dereference symlinks when copying file permissions in -f mode. - (Volker Schmidt) - -* Majorly simplify implementation of uInt64_qrm10. - (Bo Lindbergh) - -* Check the input file still exists before deleting the output one, - when aborting in cleanUpAndFail(). - (Joerg Prante, Robert Linden, Matthias Krings) - -Also a bunch of patches courtesy of Philippe Troin, the Debian maintainer -of bzip2: - -* Wrapper scripts (with manpages): bzdiff, bzgrep, bzmore. - -* Spelling changes and minor enhancements in bzip2.1. - -* Avoid race condition between creating the output file and setting its - interim permissions safely, by using fopen_output_safely(). - No changes to bzip2recover since there is no issue with file - permissions there. - -* do not print senseless report with -v when compressing an empty - file. - -* bzcat -f works on non-bzip2 files. - -* do not try to escape shell meta-characters on unix (the shell takes - care of these). - -* added --fast and --best aliases for -1 -9 for gzip compatibility. - diff -Nru bzip2-1.0.2/COPYING bzip2-1.0.2.new/COPYING --- bzip2-1.0.2/COPYING Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/COPYING Wed Jan 2 05:12:07 2002 @@ -0,0 +1,39 @@ + +This program, "bzip2" and associated library "libbzip2", are +copyright (C) 1996-2002 Julian R Seward. All rights reserved. + +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. The origin of this software must not be misrepresented; you must + not claim that you wrote the original software. If you use this + software in a product, an acknowledgment in the product + documentation would be appreciated but is not required. + +3. Altered source versions must be plainly marked as such, and must + not be misrepresented as being the original software. + +4. The name of the author may not be used to endorse or promote + products derived from this software without specific prior written + permission. + +THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS +OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED +WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY +DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE +GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, +WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +Julian Seward, Cambridge, UK. +jseward@acm.org +bzip2/libbzip2 version 1.0.2 of 30 December 2001 + diff -Nru bzip2-1.0.2/ChangeLog bzip2-1.0.2.new/ChangeLog --- bzip2-1.0.2/ChangeLog Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/ChangeLog Fri Feb 1 04:19:04 2002 @@ -0,0 +1 @@ + diff -Nru bzip2-1.0.2/INSTALL bzip2-1.0.2.new/INSTALL --- bzip2-1.0.2/INSTALL Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/INSTALL Fri Feb 1 04:19:04 2002 @@ -0,0 +1,182 @@ +Basic Installation +================== + + These are generic installation instructions. + + The `configure' shell script attempts to guess correct values for +various system-dependent variables used during compilation. It uses +those values to create a `Makefile' in each directory of the package. +It may also create one or more `.h' files containing system-dependent +definitions. Finally, it creates a shell script `config.status' that +you can run in the future to recreate the current configuration, a file +`config.cache' that saves the results of its tests to speed up +reconfiguring, and a file `config.log' containing compiler output +(useful mainly for debugging `configure'). + + If you need to do unusual things to compile the package, please try +to figure out how `configure' could check whether to do them, and mail +diffs or instructions to the address given in the `README' so they can +be considered for the next release. If at some point `config.cache' +contains results you don't want to keep, you may remove or edit it. + + The file `configure.in' is used to create `configure' by a program +called `autoconf'. You only need `configure.in' if you want to change +it or regenerate `configure' using a newer version of `autoconf'. + +The simplest way to compile this package is: + + 1. `cd' to the directory containing the package's source code and type + `./configure' to configure the package for your system. If you're + using `csh' on an old version of System V, you might need to type + `sh ./configure' instead to prevent `csh' from trying to execute + `configure' itself. + + Running `configure' takes awhile. While running, it prints some + messages telling which features it is checking for. + + 2. Type `make' to compile the package. + + 3. Optionally, type `make check' to run any self-tests that come with + the package. + + 4. Type `make install' to install the programs and any data files and + documentation. + + 5. You can remove the program binaries and object files from the + source code directory by typing `make clean'. To also remove the + files that `configure' created (so you can compile the package for + a different kind of computer), type `make distclean'. There is + also a `make maintainer-clean' target, but that is intended mainly + for the package's developers. If you use it, you may have to get + all sorts of other programs in order to regenerate files that came + with the distribution. + +Compilers and Options +===================== + + Some systems require unusual options for compilation or linking that +the `configure' script does not know about. You can give `configure' +initial values for variables by setting them in the environment. Using +a Bourne-compatible shell, you can do that on the command line like +this: + CC=c89 CFLAGS=-O2 LIBS=-lposix ./configure + +Or on systems that have the `env' program, you can do it like this: + env CPPFLAGS=-I/usr/local/include LDFLAGS=-s ./configure + +Compiling For Multiple Architectures +==================================== + + You can compile the package for more than one kind of computer at the +same time, by placing the object files for each architecture in their +own directory. To do this, you must use a version of `make' that +supports the `VPATH' variable, 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 the directory that `configure' is in and in `..'. + + If you have to use a `make' that does not supports the `VPATH' +variable, you have to compile the package for one architecture at a time +in the source code directory. After you have installed the package for +one architecture, use `make distclean' before reconfiguring for another +architecture. + +Installation Names +================== + + By default, `make install' will install the package's files in +`/usr/local/bin', `/usr/local/man', etc. You can specify an +installation prefix other than `/usr/local' by giving `configure' the +option `--prefix=PATH'. + + You can specify separate installation prefixes for +architecture-specific files and architecture-independent files. If you +give `configure' the option `--exec-prefix=PATH', the package will use +PATH as the prefix for installing programs and libraries. +Documentation and other data files will still use the regular prefix. + + In addition, if you use an unusual directory layout you can give +options like `--bindir=PATH' to specify different values for particular +kinds of files. Run `configure --help' for a list of the directories +you can set and what kinds of files go in them. + + If the package supports it, you can cause programs to be installed +with an extra prefix or suffix on their names by giving `configure' the +option `--program-prefix=PREFIX' or `--program-suffix=SUFFIX'. + +Optional Features +================= + + Some packages pay attention to `--enable-FEATURE' options to +`configure', where FEATURE indicates an optional part of the package. +They may also pay attention to `--with-PACKAGE' options, where PACKAGE +is something like `gnu-as' or `x' (for the X Window System). The +`README' should mention any `--enable-' and `--with-' options that the +package recognizes. + + For packages that use the X Window System, `configure' can usually +find the X include and library files automatically, but if it doesn't, +you can use the `configure' options `--x-includes=DIR' and +`--x-libraries=DIR' to specify their locations. + +Specifying the System Type +========================== + + There may be some features `configure' can not figure out +automatically, but needs to determine by the type of host the package +will run on. Usually `configure' can figure that out, but if it prints +a message saying it can not guess the host type, give it the +`--host=TYPE' option. TYPE can either be a short name for the system +type, such as `sun4', or a canonical name with three fields: + CPU-COMPANY-SYSTEM + +See the file `config.sub' for the possible values of each field. If +`config.sub' isn't included in this package, then this package doesn't +need to know the host type. + + If you are building compiler tools for cross-compiling, you can also +use the `--target=TYPE' option to select the type of system they will +produce code for and the `--build=TYPE' option to select the type of +system on which you are compiling the package. + +Sharing Defaults +================ + + If you want to set default values for `configure' scripts to share, +you can create a site shell script called `config.site' that gives +default values for variables like `CC', `cache_file', and `prefix'. +`configure' looks for `PREFIX/share/config.site' if it exists, then +`PREFIX/etc/config.site' if it exists. Or, you can set the +`CONFIG_SITE' environment variable to the location of the site script. +A warning: not all `configure' scripts look for a site script. + +Operation Controls +================== + + `configure' recognizes the following options to control how it +operates. + +`--cache-file=FILE' + Use and save the results of the tests in FILE instead of + `./config.cache'. Set FILE to `/dev/null' to disable caching, for + debugging `configure'. + +`--help' + Print a summary of the options to `configure', and exit. + +`--quiet' +`--silent' +`-q' + Do not print messages saying which checks are being made. To + suppress all normal output, redirect it to `/dev/null' (any error + messages will still be shown). + +`--srcdir=DIR' + Look for the package's source code in directory DIR. Usually + `configure' can determine that directory automatically. + +`--version' + Print the version of Autoconf used to generate the `configure' + script, and exit. + +`configure' also accepts some other, not widely useful, options. diff -Nru bzip2-1.0.2/LICENSE bzip2-1.0.2.new/LICENSE --- bzip2-1.0.2/LICENSE Wed Jan 2 05:12:07 2002 +++ bzip2-1.0.2.new/LICENSE Thu Jan 1 01:00:00 1970 @@ -1,39 +0,0 @@ - -This program, "bzip2" and associated library "libbzip2", are -copyright (C) 1996-2002 Julian R Seward. All rights reserved. - -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. The origin of this software must not be misrepresented; you must - not claim that you wrote the original software. If you use this - software in a product, an acknowledgment in the product - documentation would be appreciated but is not required. - -3. Altered source versions must be plainly marked as such, and must - not be misrepresented as being the original software. - -4. The name of the author may not be used to endorse or promote - products derived from this software without specific prior written - permission. - -THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS -OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED -WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE -ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY -DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL -DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE -GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS -INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, -WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING -NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS -SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - -Julian Seward, Cambridge, UK. -jseward@acm.org -bzip2/libbzip2 version 1.0.2 of 30 December 2001 - diff -Nru bzip2-1.0.2/Makefile-libbz2_so bzip2-1.0.2.new/Makefile-libbz2_so --- bzip2-1.0.2/Makefile-libbz2_so Sun Dec 30 03:15:59 2001 +++ bzip2-1.0.2.new/Makefile-libbz2_so Thu Jan 1 01:00:00 1970 @@ -1,44 +0,0 @@ - -# This Makefile builds a shared version of the library, -# libbz2.so.1.0.2, with soname libbz2.so.1.0, -# at least on x86-Linux (RedHat 7.2), -# with gcc-2.96 20000731 (Red Hat Linux 7.1 2.96-98). -# Please see the README file for some -# important info about building the library like this. - -SHELL=/bin/sh -CC=gcc -BIGFILES=-D_FILE_OFFSET_BITS=64 -CFLAGS=-fpic -fPIC -Wall -Winline -O2 -fomit-frame-pointer -fno-strength-reduce $(BIGFILES) - -OBJS= blocksort.o \ - huffman.o \ - crctable.o \ - randtable.o \ - compress.o \ - decompress.o \ - bzlib.o - -all: $(OBJS) - $(CC) -shared -Wl,-soname -Wl,libbz2.so.1.0 -o libbz2.so.1.0.2 $(OBJS) - $(CC) $(CFLAGS) -o bzip2-shared bzip2.c libbz2.so.1.0.2 - rm -f libbz2.so.1.0 - ln -s libbz2.so.1.0.2 libbz2.so.1.0 - -clean: - rm -f $(OBJS) bzip2.o libbz2.so.1.0.2 libbz2.so.1.0 bzip2-shared - -blocksort.o: blocksort.c - $(CC) $(CFLAGS) -c blocksort.c -huffman.o: huffman.c - $(CC) $(CFLAGS) -c huffman.c -crctable.o: crctable.c - $(CC) $(CFLAGS) -c crctable.c -randtable.o: randtable.c - $(CC) $(CFLAGS) -c randtable.c -compress.o: compress.c - $(CC) $(CFLAGS) -c compress.c -decompress.o: decompress.c - $(CC) $(CFLAGS) -c decompress.c -bzlib.o: bzlib.c - $(CC) $(CFLAGS) -c bzlib.c diff -Nru bzip2-1.0.2/Makefile.am bzip2-1.0.2.new/Makefile.am --- bzip2-1.0.2/Makefile.am Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/Makefile.am Fri Feb 1 04:19:09 2002 @@ -0,0 +1,31 @@ +SUBDIRS = doc + +bin_PROGRAMS = bzip2 bzip2recover +bzip2_SOURCES = bzip2.c + +bzip2_LDADD = libbz2.la +bzip2recover_SOURCES = bzip2recover.c +lib_LTLIBRARIES = libbz2.la +libbz2_la_SOURCES = \ + blocksort.c \ + huffman.c \ + crctable.c \ + randtable.c \ + compress.c \ + decompress.c \ + bzlib.c \ + bzlib.h \ + bzlib_private.h + +libbz2_la_LDFLAGS = -version-info 1:0:0 +include_HEADERS = bzlib.h bzlib_private.h + +bin_SCRIPTS = bzless bzgrep + +EXTRA_DIST = README README.COMPILATION.PROBLEMS \ + Y2K_INFO libbz2.def libbz2.dsp \ + sample1.bz2 sample1.ref sample2.bz2 sample2.ref sample3.bz2 sample3.ref + +install-exec-hook: + $(LN_S) -f bzip2 $(DESTDIR)$(bindir)/bunzip2 + $(LN_S) -f bzip2 $(DESTDIR)$(bindir)/bzcat diff -Nru bzip2-1.0.2/NEWS bzip2-1.0.2.new/NEWS --- bzip2-1.0.2/NEWS Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/NEWS Sat Jan 26 01:02:57 2002 @@ -0,0 +1,253 @@ + + +0.9.0 +~~~~~ +First version. + + +0.9.0a +~~~~~~ +Removed 'ranlib' from Makefile, since most modern Unix-es +don't need it, or even know about it. + + +0.9.0b +~~~~~~ +Fixed a problem with error reporting in bzip2.c. This does not effect +the library in any way. Problem is: versions 0.9.0 and 0.9.0a (of the +program proper) compress and decompress correctly, but give misleading +error messages (internal panics) when an I/O error occurs, instead of +reporting the problem correctly. This shouldn't give any data loss +(as far as I can see), but is confusing. + +Made the inline declarations disappear for non-GCC compilers. + + +0.9.0c +~~~~~~ +Fixed some problems in the library pertaining to some boundary cases. +This makes the library behave more correctly in those situations. The +fixes apply only to features (calls and parameters) not used by +bzip2.c, so the non-fixedness of them in previous versions has no +effect on reliability of bzip2.c. + +In bzlib.c: + * made zero-length BZ_FLUSH work correctly in bzCompress(). + * fixed bzWrite/bzRead to ignore zero-length requests. + * fixed bzread to correctly handle read requests after EOF. + * wrong parameter order in call to bzDecompressInit in + bzBuffToBuffDecompress. Fixed. + +In compress.c: + * changed setting of nGroups in sendMTFValues() so as to + do a bit better on small files. This _does_ effect + bzip2.c. + + +0.9.5a +~~~~~~ +Major change: add a fallback sorting algorithm (blocksort.c) +to give reasonable behaviour even for very repetitive inputs. +Nuked --repetitive-best and --repetitive-fast since they are +no longer useful. + +Minor changes: mostly a whole bunch of small changes/ +bugfixes in the driver (bzip2.c). Changes pertaining to the +user interface are: + + allow decompression of symlink'd files to stdout + decompress/test files even without .bz2 extension + give more accurate error messages for I/O errors + when compressing/decompressing to stdout, don't catch control-C + read flags from BZIP2 and BZIP environment variables + decline to break hard links to a file unless forced with -f + allow -c flag even with no filenames + preserve file ownerships as far as possible + make -s -1 give the expected block size (100k) + add a flag -q --quiet to suppress nonessential warnings + stop decoding flags after --, so files beginning in - can be handled + resolved inconsistent naming: bzcat or bz2cat ? + bzip2 --help now returns 0 + +Programming-level changes are: + + fixed syntax error in GET_LL4 for Borland C++ 5.02 + let bzBuffToBuffDecompress return BZ_DATA_ERROR{_MAGIC} + fix overshoot of mode-string end in bzopen_or_bzdopen + wrapped bzlib.h in #ifdef __cplusplus ... extern "C" { ... } + close file handles under all error conditions + added minor mods so it compiles with DJGPP out of the box + fixed Makefile so it doesn't give problems with BSD make + fix uninitialised memory reads in dlltest.c + +0.9.5b +~~~~~~ +Open stdin/stdout in binary mode for DJGPP. + +0.9.5c +~~~~~~ +Changed BZ_N_OVERSHOOT to be ... + 2 instead of ... + 1. The + 1 +version could cause the sorted order to be wrong in some extremely +obscure cases. Also changed setting of quadrant in blocksort.c. + +0.9.5d +~~~~~~ +The only functional change is to make bzlibVersion() in the library +return the correct string. This has no effect whatsoever on the +functioning of the bzip2 program or library. Added a couple of casts +so the library compiles without warnings at level 3 in MS Visual +Studio 6.0. Included a Y2K statement in the file Y2K_INFO. All other +changes are minor documentation changes. + +1.0 +~~~ +Several minor bugfixes and enhancements: + +* Large file support. The library uses 64-bit counters to + count the volume of data passing through it. bzip2.c + is now compiled with -D_FILE_OFFSET_BITS=64 to get large + file support from the C library. -v correctly prints out + file sizes greater than 4 gigabytes. All these changes have + been made without assuming a 64-bit platform or a C compiler + which supports 64-bit ints, so, except for the C library + aspect, they are fully portable. + +* Decompression robustness. The library/program should be + robust to any corruption of compressed data, detecting and + handling _all_ corruption, instead of merely relying on + the CRCs. What this means is that the program should + never crash, given corrupted data, and the library should + always return BZ_DATA_ERROR. + +* Fixed an obscure race-condition bug only ever observed on + Solaris, in which, if you were very unlucky and issued + control-C at exactly the wrong time, both input and output + files would be deleted. + +* Don't run out of file handles on test/decompression when + large numbers of files have invalid magic numbers. + +* Avoid library namespace pollution. Prefix all exported + symbols with BZ2_. + +* Minor sorting enhancements from my DCC2000 paper. + +* Advance the version number to 1.0, so as to counteract the + (false-in-this-case) impression some people have that programs + with version numbers less than 1.0 are in some way, experimental, + pre-release versions. + +* Create an initial Makefile-libbz2_so to build a shared library. + Yes, I know I should really use libtool et al ... + +* Make the program exit with 2 instead of 0 when decompression + fails due to a bad magic number (ie, an invalid bzip2 header). + Also exit with 1 (as the manual claims :-) whenever a diagnostic + message would have been printed AND the corresponding operation + is aborted, for example + bzip2: Output file xx already exists. + When a diagnostic message is printed but the operation is not + aborted, for example + bzip2: Can't guess original name for wurble -- using wurble.out + then the exit value 0 is returned, unless some other problem is + also detected. + + I think it corresponds more closely to what the manual claims now. + + +1.0.1 +~~~~~ +* Modified dlltest.c so it uses the new BZ2_ naming scheme. +* Modified makefile-msc to fix minor build probs on Win2k. +* Updated README.COMPILATION.PROBLEMS. + +There are no functionality changes or bug fixes relative to version +1.0.0. This is just a documentation update + a fix for minor Win32 +build problems. For almost everyone, upgrading from 1.0.0 to 1.0.1 is +utterly pointless. Don't bother. + + +1.0.2 +~~~~~ +A bug fix release, addressing various minor issues which have appeared +in the 18 or so months since 1.0.1 was released. Most of the fixes +are to do with file-handling or documentation bugs. To the best of my +knowledge, there have been no data-loss-causing bugs reported in the +compression/decompression engine of 1.0.0 or 1.0.1. + +Note that this release does not improve the rather crude build system +for Unix platforms. The general plan here is to autoconfiscate/ +libtoolise 1.0.2 soon after release, and release the result as 1.1.0 +or perhaps 1.2.0. That, however, is still just a plan at this point. + +Here are the changes in 1.0.2. Bug-reporters and/or patch-senders in +parentheses. + +* Fix an infinite segfault loop in 1.0.1 when a directory is + encountered in -f (force) mode. + (Trond Eivind Glomsrod, Nicholas Nethercote, Volker Schmidt) + +* Avoid double fclose() of output file on certain I/O error paths. + (Solar Designer) + +* Don't fail with internal error 1007 when fed a long stream (> 48MB) + of byte 251. Also print useful message suggesting that 1007s may be + caused by bad memory. + (noticed by Juan Pedro Vallejo, fixed by me) + +* Fix uninitialised variable silly bug in demo prog dlltest.c. + (Jorj Bauer) + +* Remove 512-MB limitation on recovered file size for bzip2recover + on selected platforms which support 64-bit ints. At the moment + all GCC supported platforms, and Win32. + (me, Alson van der Meulen) + +* Hard-code header byte values, to give correct operation on platforms + using EBCDIC as their native character set (IBM's OS/390). + (Leland Lucius) + +* Copy file access times correctly. + (Marty Leisner) + +* Add distclean and check targets to Makefile. + (Michael Carmack) + +* Parameterise use of ar and ranlib in Makefile. Also add $(LDFLAGS). + (Rich Ireland, Bo Thorsen) + +* Pass -p (create parent dirs as needed) to mkdir during make install. + (Jeremy Fusco) + +* Dereference symlinks when copying file permissions in -f mode. + (Volker Schmidt) + +* Majorly simplify implementation of uInt64_qrm10. + (Bo Lindbergh) + +* Check the input file still exists before deleting the output one, + when aborting in cleanUpAndFail(). + (Joerg Prante, Robert Linden, Matthias Krings) + +Also a bunch of patches courtesy of Philippe Troin, the Debian maintainer +of bzip2: + +* Wrapper scripts (with manpages): bzdiff, bzgrep, bzmore. + +* Spelling changes and minor enhancements in bzip2.1. + +* Avoid race condition between creating the output file and setting its + interim permissions safely, by using fopen_output_safely(). + No changes to bzip2recover since there is no issue with file + permissions there. + +* do not print senseless report with -v when compressing an empty + file. + +* bzcat -f works on non-bzip2 files. + +* do not try to escape shell meta-characters on unix (the shell takes + care of these). + +* added --fast and --best aliases for -1 -9 for gzip compatibility. + diff -Nru bzip2-1.0.2/bzdiff.1 bzip2-1.0.2.new/bzdiff.1 --- bzip2-1.0.2/bzdiff.1 Sun Dec 30 03:12:35 2001 +++ bzip2-1.0.2.new/bzdiff.1 Thu Jan 1 01:00:00 1970 @@ -1,47 +0,0 @@ -\"Shamelessly copied from zmore.1 by Philippe Troin -\"for Debian GNU/Linux -.TH BZDIFF 1 -.SH NAME -bzcmp, bzdiff \- compare bzip2 compressed files -.SH SYNOPSIS -.B bzcmp -[ cmp_options ] file1 -[ file2 ] -.br -.B bzdiff -[ diff_options ] file1 -[ file2 ] -.SH DESCRIPTION -.I Bzcmp -and -.I bzdiff -are used to invoke the -.I cmp -or the -.I diff -program on bzip2 compressed files. All options specified are passed -directly to -.I cmp -or -.IR diff "." -If only 1 file is specified, then the files compared are -.I file1 -and an uncompressed -.IR file1 ".bz2." -If two files are specified, then they are uncompressed if necessary and fed to -.I cmp -or -.IR diff "." -The exit status from -.I cmp -or -.I diff -is preserved. -.SH "SEE ALSO" -cmp(1), diff(1), bzmore(1), bzless(1), bzgrep(1), bzip2(1) -.SH BUGS -Messages from the -.I cmp -or -.I diff -programs refer to temporary filenames instead of those specified. diff -Nru bzip2-1.0.2/bzgrep.1 bzip2-1.0.2.new/bzgrep.1 --- bzip2-1.0.2/bzgrep.1 Sun Dec 30 03:12:35 2001 +++ bzip2-1.0.2.new/bzgrep.1 Thu Jan 1 01:00:00 1970 @@ -1,56 +0,0 @@ -\"Shamelessly copied from zmore.1 by Philippe Troin -\"for Debian GNU/Linux -.TH BZGREP 1 -.SH NAME -bzgrep, bzfgrep, bzegrep \- search possibly bzip2 compressed files for a regular expression -.SH SYNOPSIS -.B bzgrep -[ grep_options ] -.BI [\ -e\ ] " pattern" -.IR filename ".\|.\|." -.br -.B bzegrep -[ egrep_options ] -.BI [\ -e\ ] " pattern" -.IR filename ".\|.\|." -.br -.B bzfgrep -[ fgrep_options ] -.BI [\ -e\ ] " pattern" -.IR filename ".\|.\|." -.SH DESCRIPTION -.IR Bzgrep -is used to invoke the -.I grep -on bzip2-compressed files. All options specified are passed directly to -.I grep. -If no file is specified, then the standard input is decompressed -if necessary and fed to grep. -Otherwise the given files are uncompressed if necessary and fed to -.I grep. -.PP -If -.I bzgrep -is invoked as -.I bzegrep -or -.I bzfgrep -then -.I egrep -or -.I fgrep -is used instead of -.I grep. -If the GREP environment variable is set, -.I bzgrep -uses it as the -.I grep -program to be invoked. For example: - - for sh: GREP=fgrep bzgrep string files - for csh: (setenv GREP fgrep; bzgrep string files) -.SH AUTHOR -Charles Levert (charles@comm.polymtl.ca). Adapted to bzip2 by Philippe -Troin for Debian GNU/Linux. -.SH "SEE ALSO" -grep(1), egrep(1), fgrep(1), bzdiff(1), bzmore(1), bzless(1), bzip2(1) diff -Nru bzip2-1.0.2/bzip2.1 bzip2-1.0.2.new/bzip2.1 --- bzip2-1.0.2/bzip2.1 Thu Jan 3 00:14:36 2002 +++ bzip2-1.0.2.new/bzip2.1 Thu Jan 1 01:00:00 1970 @@ -1,453 +0,0 @@ -.PU -.TH bzip2 1 -.SH NAME -bzip2, bunzip2 \- a block-sorting file compressor, v1.0.2 -.br -bzcat \- decompresses files to stdout -.br -bzip2recover \- recovers data from damaged bzip2 files - -.SH SYNOPSIS -.ll +8 -.B bzip2 -.RB [ " \-cdfkqstvzVL123456789 " ] -[ -.I "filenames \&..." -] -.ll -8 -.br -.B bunzip2 -.RB [ " \-fkvsVL " ] -[ -.I "filenames \&..." -] -.br -.B bzcat -.RB [ " \-s " ] -[ -.I "filenames \&..." -] -.br -.B bzip2recover -.I "filename" - -.SH DESCRIPTION -.I bzip2 -compresses files using the Burrows-Wheeler block sorting -text compression algorithm, and Huffman coding. Compression is -generally considerably better than that achieved by more conventional -LZ77/LZ78-based compressors, and approaches the performance of the PPM -family of statistical compressors. - -The command-line options are deliberately very similar to -those of -.I GNU gzip, -but they are not identical. - -.I bzip2 -expects a list of file names to accompany the -command-line flags. Each file is replaced by a compressed version of -itself, with the name "original_name.bz2". -Each compressed file -has the same modification date, permissions, and, when possible, -ownership as the corresponding original, so that these properties can -be correctly restored at decompression time. File name handling is -naive in the sense that there is no mechanism for preserving original -file names, permissions, ownerships or dates in filesystems which lack -these concepts, or have serious file name length restrictions, such as -MS-DOS. - -.I bzip2 -and -.I bunzip2 -will by default not overwrite existing -files. If you want this to happen, specify the \-f flag. - -If no file names are specified, -.I bzip2 -compresses from standard -input to standard output. In this case, -.I bzip2 -will decline to -write compressed output to a terminal, as this would be entirely -incomprehensible and therefore pointless. - -.I bunzip2 -(or -.I bzip2 \-d) -decompresses all -specified files. Files which were not created by -.I bzip2 -will be detected and ignored, and a warning issued. -.I bzip2 -attempts to guess the filename for the decompressed file -from that of the compressed file as follows: - - filename.bz2 becomes filename - filename.bz becomes filename - filename.tbz2 becomes filename.tar - filename.tbz becomes filename.tar - anyothername becomes anyothername.out - -If the file does not end in one of the recognised endings, -.I .bz2, -.I .bz, -.I .tbz2 -or -.I .tbz, -.I bzip2 -complains that it cannot -guess the name of the original file, and uses the original name -with -.I .out -appended. - -As with compression, supplying no -filenames causes decompression from -standard input to standard output. - -.I bunzip2 -will correctly decompress a file which is the -concatenation of two or more compressed files. The result is the -concatenation of the corresponding uncompressed files. Integrity -testing (\-t) -of concatenated -compressed files is also supported. - -You can also compress or decompress files to the standard output by -giving the \-c flag. Multiple files may be compressed and -decompressed like this. The resulting outputs are fed sequentially to -stdout. Compression of multiple files -in this manner generates a stream -containing multiple compressed file representations. Such a stream -can be decompressed correctly only by -.I bzip2 -version 0.9.0 or -later. Earlier versions of -.I bzip2 -will stop after decompressing -the first file in the stream. - -.I bzcat -(or -.I bzip2 -dc) -decompresses all specified files to -the standard output. - -.I bzip2 -will read arguments from the environment variables -.I BZIP2 -and -.I BZIP, -in that order, and will process them -before any arguments read from the command line. This gives a -convenient way to supply default arguments. - -Compression is always performed, even if the compressed -file is slightly -larger than the original. Files of less than about one hundred bytes -tend to get larger, since the compression mechanism has a constant -overhead in the region of 50 bytes. Random data (including the output -of most file compressors) is coded at about 8.05 bits per byte, giving -an expansion of around 0.5%. - -As a self-check for your protection, -.I -bzip2 -uses 32-bit CRCs to -make sure that the decompressed version of a file is identical to the -original. This guards against corruption of the compressed data, and -against undetected bugs in -.I bzip2 -(hopefully very unlikely). The -chances of data corruption going undetected is microscopic, about one -chance in four billion for each file processed. Be aware, though, that -the check occurs upon decompression, so it can only tell you that -something is wrong. It can't help you -recover the original uncompressed -data. You can use -.I bzip2recover -to try to recover data from -damaged files. - -Return values: 0 for a normal exit, 1 for environmental problems (file -not found, invalid flags, I/O errors, &c), 2 to indicate a corrupt -compressed file, 3 for an internal consistency error (eg, bug) which -caused -.I bzip2 -to panic. - -.SH OPTIONS -.TP -.B \-c --stdout -Compress or decompress to standard output. -.TP -.B \-d --decompress -Force decompression. -.I bzip2, -.I bunzip2 -and -.I bzcat -are -really the same program, and the decision about what actions to take is -done on the basis of which name is used. This flag overrides that -mechanism, and forces -.I bzip2 -to decompress. -.TP -.B \-z --compress -The complement to \-d: forces compression, regardless of the -invocation name. -.TP -.B \-t --test -Check integrity of the specified file(s), but don't decompress them. -This really performs a trial decompression and throws away the result. -.TP -.B \-f --force -Force overwrite of output files. Normally, -.I bzip2 -will not overwrite -existing output files. Also forces -.I bzip2 -to break hard links -to files, which it otherwise wouldn't do. - -bzip2 normally declines to decompress files which don't have the -correct magic header bytes. If forced (-f), however, it will pass -such files through unmodified. This is how GNU gzip behaves. -.TP -.B \-k --keep -Keep (don't delete) input files during compression -or decompression. -.TP -.B \-s --small -Reduce memory usage, for compression, decompression and testing. Files -are decompressed and tested using a modified algorithm which only -requires 2.5 bytes per block byte. This means any file can be -decompressed in 2300k of memory, albeit at about half the normal speed. - -During compression, \-s selects a block size of 200k, which limits -memory use to around the same figure, at the expense of your compression -ratio. In short, if your machine is low on memory (8 megabytes or -less), use \-s for everything. See MEMORY MANAGEMENT below. -.TP -.B \-q --quiet -Suppress non-essential warning messages. Messages pertaining to -I/O errors and other critical events will not be suppressed. -.TP -.B \-v --verbose -Verbose mode -- show the compression ratio for each file processed. -Further \-v's increase the verbosity level, spewing out lots of -information which is primarily of interest for diagnostic purposes. -.TP -.B \-L --license -V --version -Display the software version, license terms and conditions. -.TP -.B \-1 (or \-\-fast) to \-9 (or \-\-best) -Set the block size to 100 k, 200 k .. 900 k when compressing. Has no -effect when decompressing. See MEMORY MANAGEMENT below. -The \-\-fast and \-\-best aliases are primarily for GNU gzip -compatibility. In particular, \-\-fast doesn't make things -significantly faster. -And \-\-best merely selects the default behaviour. -.TP -.B \-- -Treats all subsequent arguments as file names, even if they start -with a dash. This is so you can handle files with names beginning -with a dash, for example: bzip2 \-- \-myfilename. -.TP -.B \--repetitive-fast --repetitive-best -These flags are redundant in versions 0.9.5 and above. They provided -some coarse control over the behaviour of the sorting algorithm in -earlier versions, which was sometimes useful. 0.9.5 and above have an -improved algorithm which renders these flags irrelevant. - -.SH MEMORY MANAGEMENT -.I bzip2 -compresses large files in blocks. The block size affects -both the compression ratio achieved, and the amount of memory needed for -compression and decompression. The flags \-1 through \-9 -specify the block size to be 100,000 bytes through 900,000 bytes (the -default) respectively. At decompression time, the block size used for -compression is read from the header of the compressed file, and -.I bunzip2 -then allocates itself just enough memory to decompress -the file. Since block sizes are stored in compressed files, it follows -that the flags \-1 to \-9 are irrelevant to and so ignored -during decompression. - -Compression and decompression requirements, -in bytes, can be estimated as: - - Compression: 400k + ( 8 x block size ) - - Decompression: 100k + ( 4 x block size ), or - 100k + ( 2.5 x block size ) - -Larger block sizes give rapidly diminishing marginal returns. Most of -the compression comes from the first two or three hundred k of block -size, a fact worth bearing in mind when using -.I bzip2 -on small machines. -It is also important to appreciate that the decompression memory -requirement is set at compression time by the choice of block size. - -For files compressed with the default 900k block size, -.I bunzip2 -will require about 3700 kbytes to decompress. To support decompression -of any file on a 4 megabyte machine, -.I bunzip2 -has an option to -decompress using approximately half this amount of memory, about 2300 -kbytes. Decompression speed is also halved, so you should use this -option only where necessary. The relevant flag is -s. - -In general, try and use the largest block size memory constraints allow, -since that maximises the compression achieved. Compression and -decompression speed are virtually unaffected by block size. - -Another significant point applies to files which fit in a single block --- that means most files you'd encounter using a large block size. The -amount of real memory touched is proportional to the size of the file, -since the file is smaller than a block. For example, compressing a file -20,000 bytes long with the flag -9 will cause the compressor to -allocate around 7600k of memory, but only touch 400k + 20000 * 8 = 560 -kbytes of it. Similarly, the decompressor will allocate 3700k but only -touch 100k + 20000 * 4 = 180 kbytes. - -Here is a table which summarises the maximum memory usage for different -block sizes. Also recorded is the total compressed size for 14 files of -the Calgary Text Compression Corpus totalling 3,141,622 bytes. This -column gives some feel for how compression varies with block size. -These figures tend to understate the advantage of larger block sizes for -larger files, since the Corpus is dominated by smaller files. - - Compress Decompress Decompress Corpus - Flag usage usage -s usage Size - - -1 1200k 500k 350k 914704 - -2 2000k 900k 600k 877703 - -3 2800k 1300k 850k 860338 - -4 3600k 1700k 1100k 846899 - -5 4400k 2100k 1350k 845160 - -6 5200k 2500k 1600k 838626 - -7 6100k 2900k 1850k 834096 - -8 6800k 3300k 2100k 828642 - -9 7600k 3700k 2350k 828642 - -.SH RECOVERING DATA FROM DAMAGED FILES -.I bzip2 -compresses files in blocks, usually 900kbytes long. Each -block is handled independently. If a media or transmission error causes -a multi-block .bz2 -file to become damaged, it may be possible to -recover data from the undamaged blocks in the file. - -The compressed representation of each block is delimited by a 48-bit -pattern, which makes it possible to find the block boundaries with -reasonable certainty. Each block also carries its own 32-bit CRC, so -damaged blocks can be distinguished from undamaged ones. - -.I bzip2recover -is a simple program whose purpose is to search for -blocks in .bz2 files, and write each block out into its own .bz2 -file. You can then use -.I bzip2 -\-t -to test the -integrity of the resulting files, and decompress those which are -undamaged. - -.I bzip2recover -takes a single argument, the name of the damaged file, -and writes a number of files "rec00001file.bz2", -"rec00002file.bz2", etc, containing the extracted blocks. -The output filenames are designed so that the use of -wildcards in subsequent processing -- for example, -"bzip2 -dc rec*file.bz2 > recovered_data" -- processes the files in -the correct order. - -.I bzip2recover -should be of most use dealing with large .bz2 -files, as these will contain many blocks. It is clearly -futile to use it on damaged single-block files, since a -damaged block cannot be recovered. If you wish to minimise -any potential data loss through media or transmission errors, -you might consider compressing with a smaller -block size. - -.SH PERFORMANCE NOTES -The sorting phase of compression gathers together similar strings in the -file. Because of this, files containing very long runs of repeated -symbols, like "aabaabaabaab ..." (repeated several hundred times) may -compress more slowly than normal. Versions 0.9.5 and above fare much -better than previous versions in this respect. The ratio between -worst-case and average-case compression time is in the region of 10:1. -For previous versions, this figure was more like 100:1. You can use the -\-vvvv option to monitor progress in great detail, if you want. - -Decompression speed is unaffected by these phenomena. - -.I bzip2 -usually allocates several megabytes of memory to operate -in, and then charges all over it in a fairly random fashion. This means -that performance, both for compressing and decompressing, is largely -determined by the speed at which your machine can service cache misses. -Because of this, small changes to the code to reduce the miss rate have -been observed to give disproportionately large performance improvements. -I imagine -.I bzip2 -will perform best on machines with very large caches. - -.SH CAVEATS -I/O error messages are not as helpful as they could be. -.I bzip2 -tries hard to detect I/O errors and exit cleanly, but the details of -what the problem is sometimes seem rather misleading. - -This manual page pertains to version 1.0.2 of -.I bzip2. -Compressed data created by this version is entirely forwards and -backwards compatible with the previous public releases, versions -0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1, but with the following -exception: 0.9.0 and above can correctly decompress multiple -concatenated compressed files. 0.1pl2 cannot do this; it will stop -after decompressing just the first file in the stream. - -.I bzip2recover -versions prior to this one, 1.0.2, used 32-bit integers to represent -bit positions in compressed files, so it could not handle compressed -files more than 512 megabytes long. Version 1.0.2 and above uses -64-bit ints on some platforms which support them (GNU supported -targets, and Windows). To establish whether or not bzip2recover was -built with such a limitation, run it without arguments. In any event -you can build yourself an unlimited version if you can recompile it -with MaybeUInt64 set to be an unsigned 64-bit integer. - - - -.SH AUTHOR -Julian Seward, jseward@acm.org. - -http://sources.redhat.com/bzip2 - -The ideas embodied in -.I bzip2 -are due to (at least) the following -people: Michael Burrows and David Wheeler (for the block sorting -transformation), David Wheeler (again, for the Huffman coder), Peter -Fenwick (for the structured coding model in the original -.I bzip, -and many refinements), and Alistair Moffat, Radford Neal and Ian Witten -(for the arithmetic coder in the original -.I bzip). -I am much -indebted for their help, support and advice. See the manual in the -source distribution for pointers to sources of documentation. Christian -von Roques encouraged me to look for faster sorting algorithms, so as to -speed up compression. Bela Lubkin encouraged me to improve the -worst-case compression performance. -The bz* scripts are derived from those of GNU gzip. -Many people sent patches, helped -with portability problems, lent machines, gave advice and were generally -helpful. diff -Nru bzip2-1.0.2/bzip2.1.preformatted bzip2-1.0.2.new/bzip2.1.preformatted --- bzip2-1.0.2/bzip2.1.preformatted Thu Jan 3 00:15:56 2002 +++ bzip2-1.0.2.new/bzip2.1.preformatted Thu Jan 1 01:00:00 1970 @@ -1,398 +0,0 @@ -bzip2(1) bzip2(1) - - - -NNAAMMEE - bzip2, bunzip2 - a block-sorting file compressor, v1.0.2 - bzcat - decompresses files to stdout - bzip2recover - recovers data from damaged bzip2 files - - -SSYYNNOOPPSSIISS - bbzziipp22 [ --ccddffkkqqssttvvzzVVLL112233445566778899 ] [ _f_i_l_e_n_a_m_e_s _._._. ] - bbuunnzziipp22 [ --ffkkvvssVVLL ] [ _f_i_l_e_n_a_m_e_s _._._. ] - bbzzccaatt [ --ss ] [ _f_i_l_e_n_a_m_e_s _._._. ] - bbzziipp22rreeccoovveerr _f_i_l_e_n_a_m_e - - -DDEESSCCRRIIPPTTIIOONN - _b_z_i_p_2 compresses files using the Burrows-Wheeler block - sorting text compression algorithm, and Huffman coding. - Compression is generally considerably better than that - achieved by more conventional LZ77/LZ78-based compressors, - and approaches the performance of the PPM family of sta­ - tistical compressors. - - The command-line options are deliberately very similar to - those of _G_N_U _g_z_i_p_, but they are not identical. - - _b_z_i_p_2 expects a list of file names to accompany the com­ - mand-line flags. Each file is replaced by a compressed - version of itself, with the name "original_name.bz2". - Each compressed file has the same modification date, per­ - missions, and, when possible, ownership as the correspond­ - ing original, so that these properties can be correctly - restored at decompression time. File name handling is - naive in the sense that there is no mechanism for preserv­ - ing original file names, permissions, ownerships or dates - in filesystems which lack these concepts, or have serious - file name length restrictions, such as MS-DOS. - - _b_z_i_p_2 and _b_u_n_z_i_p_2 will by default not overwrite existing - files. If you want this to happen, specify the -f flag. - - If no file names are specified, _b_z_i_p_2 compresses from - standard input to standard output. In this case, _b_z_i_p_2 - will decline to write compressed output to a terminal, as - this would be entirely incomprehensible and therefore - pointless. - - _b_u_n_z_i_p_2 (or _b_z_i_p_2 _-_d_) decompresses all specified files. - Files which were not created by _b_z_i_p_2 will be detected and - ignored, and a warning issued. _b_z_i_p_2 attempts to guess - the filename for the decompressed file from that of the - compressed file as follows: - - filename.bz2 becomes filename - filename.bz becomes filename - filename.tbz2 becomes filename.tar - filename.tbz becomes filename.tar - anyothername becomes anyothername.out - - If the file does not end in one of the recognised endings, - _._b_z_2_, _._b_z_, _._t_b_z_2 or _._t_b_z_, _b_z_i_p_2 complains that it cannot - guess the name of the original file, and uses the original - name with _._o_u_t appended. - - As with compression, supplying no filenames causes decom­ - pression from standard input to standard output. - - _b_u_n_z_i_p_2 will correctly decompress a file which is the con­ - catenation of two or more compressed files. The result is - the concatenation of the corresponding uncompressed files. - Integrity testing (-t) of concatenated compressed files is - also supported. - - You can also compress or decompress files to the standard - output by giving the -c flag. Multiple files may be com­ - pressed and decompressed like this. The resulting outputs - are fed sequentially to stdout. Compression of multiple - files in this manner generates a stream containing multi­ - ple compressed file representations. Such a stream can be - decompressed correctly only by _b_z_i_p_2 version 0.9.0 or - later. Earlier versions of _b_z_i_p_2 will stop after decom­ - pressing the first file in the stream. - - _b_z_c_a_t (or _b_z_i_p_2 _-_d_c_) decompresses all specified files to - the standard output. - - _b_z_i_p_2 will read arguments from the environment variables - _B_Z_I_P_2 and _B_Z_I_P_, in that order, and will process them - before any arguments read from the command line. This - gives a convenient way to supply default arguments. - - Compression is always performed, even if the compressed - file is slightly larger than the original. Files of less - than about one hundred bytes tend to get larger, since the - compression mechanism has a constant overhead in the - region of 50 bytes. Random data (including the output of - most file compressors) is coded at about 8.05 bits per - byte, giving an expansion of around 0.5%. - - As a self-check for your protection, _b_z_i_p_2 uses 32-bit - CRCs to make sure that the decompressed version of a file - is identical to the original. This guards against corrup­ - tion of the compressed data, and against undetected bugs - in _b_z_i_p_2 (hopefully very unlikely). The chances of data - corruption going undetected is microscopic, about one - chance in four billion for each file processed. Be aware, - though, that the check occurs upon decompression, so it - can only tell you that something is wrong. It can't help - you recover the original uncompressed data. You can use - _b_z_i_p_2_r_e_c_o_v_e_r to try to recover data from damaged files. - - Return values: 0 for a normal exit, 1 for environmental - problems (file not found, invalid flags, I/O errors, &c), - 2 to indicate a corrupt compressed file, 3 for an internal - consistency error (eg, bug) which caused _b_z_i_p_2 to panic. - - -OOPPTTIIOONNSS - --cc ----ssttddoouutt - Compress or decompress to standard output. - - --dd ----ddeeccoommpprreessss - Force decompression. _b_z_i_p_2_, _b_u_n_z_i_p_2 and _b_z_c_a_t are - really the same program, and the decision about - what actions to take is done on the basis of which - name is used. This flag overrides that mechanism, - and forces _b_z_i_p_2 to decompress. - - --zz ----ccoommpprreessss - The complement to -d: forces compression, - regardless of the invocation name. - - --tt ----tteesstt - Check integrity of the specified file(s), but don't - decompress them. This really performs a trial - decompression and throws away the result. - - --ff ----ffoorrccee - Force overwrite of output files. Normally, _b_z_i_p_2 - will not overwrite existing output files. Also - forces _b_z_i_p_2 to break hard links to files, which it - otherwise wouldn't do. - - bzip2 normally declines to decompress files which - don't have the correct magic header bytes. If - forced (-f), however, it will pass such files - through unmodified. This is how GNU gzip behaves. - - --kk ----kkeeeepp - Keep (don't delete) input files during compression - or decompression. - - --ss ----ssmmaallll - Reduce memory usage, for compression, decompression - and testing. Files are decompressed and tested - using a modified algorithm which only requires 2.5 - bytes per block byte. This means any file can be - decompressed in 2300k of memory, albeit at about - half the normal speed. - - During compression, -s selects a block size of - 200k, which limits memory use to around the same - figure, at the expense of your compression ratio. - In short, if your machine is low on memory (8 - megabytes or less), use -s for everything. See - MEMORY MANAGEMENT below. - - --qq ----qquuiieett - Suppress non-essential warning messages. Messages - pertaining to I/O errors and other critical events - will not be suppressed. - - --vv ----vveerrbboossee - Verbose mode -- show the compression ratio for each - file processed. Further -v's increase the ver­ - bosity level, spewing out lots of information which - is primarily of interest for diagnostic purposes. - - --LL ----lliicceennssee --VV ----vveerrssiioonn - Display the software version, license terms and - conditions. - - --11 ((oorr ----ffaasstt)) ttoo --99 ((oorr ----bbeesstt)) - Set the block size to 100 k, 200 k .. 900 k when - compressing. Has no effect when decompressing. - See MEMORY MANAGEMENT below. The --fast and --best - aliases are primarily for GNU gzip compatibility. - In particular, --fast doesn't make things signifi­ - cantly faster. And --best merely selects the - default behaviour. - - ---- Treats all subsequent arguments as file names, even - if they start with a dash. This is so you can han­ - dle files with names beginning with a dash, for - example: bzip2 -- -myfilename. - - ----rreeppeettiittiivvee--ffaasstt ----rreeppeettiittiivvee--bbeesstt - These flags are redundant in versions 0.9.5 and - above. They provided some coarse control over the - behaviour of the sorting algorithm in earlier ver­ - sions, which was sometimes useful. 0.9.5 and above - have an improved algorithm which renders these - flags irrelevant. - - -MMEEMMOORRYY MMAANNAAGGEEMMEENNTT - _b_z_i_p_2 compresses large files in blocks. The block size - affects both the compression ratio achieved, and the - amount of memory needed for compression and decompression. - The flags -1 through -9 specify the block size to be - 100,000 bytes through 900,000 bytes (the default) respec­ - tively. At decompression time, the block size used for - compression is read from the header of the compressed - file, and _b_u_n_z_i_p_2 then allocates itself just enough memory - to decompress the file. Since block sizes are stored in - compressed files, it follows that the flags -1 to -9 are - irrelevant to and so ignored during decompression. - - Compression and decompression requirements, in bytes, can - be estimated as: - - Compression: 400k + ( 8 x block size ) - - Decompression: 100k + ( 4 x block size ), or - 100k + ( 2.5 x block size ) - - Larger block sizes give rapidly diminishing marginal - returns. Most of the compression comes from the first two - or three hundred k of block size, a fact worth bearing in - mind when using _b_z_i_p_2 on small machines. It is also - important to appreciate that the decompression memory - requirement is set at compression time by the choice of - block size. - - For files compressed with the default 900k block size, - _b_u_n_z_i_p_2 will require about 3700 kbytes to decompress. To - support decompression of any file on a 4 megabyte machine, - _b_u_n_z_i_p_2 has an option to decompress using approximately - half this amount of memory, about 2300 kbytes. Decompres­ - sion speed is also halved, so you should use this option - only where necessary. The relevant flag is -s. - - In general, try and use the largest block size memory con­ - straints allow, since that maximises the compression - achieved. Compression and decompression speed are virtu­ - ally unaffected by block size. - - Another significant point applies to files which fit in a - single block -- that means most files you'd encounter - using a large block size. The amount of real memory - touched is proportional to the size of the file, since the - file is smaller than a block. For example, compressing a - file 20,000 bytes long with the flag -9 will cause the - compressor to allocate around 7600k of memory, but only - touch 400k + 20000 * 8 = 560 kbytes of it. Similarly, the - decompressor will allocate 3700k but only touch 100k + - 20000 * 4 = 180 kbytes. - - Here is a table which summarises the maximum memory usage - for different block sizes. Also recorded is the total - compressed size for 14 files of the Calgary Text Compres­ - sion Corpus totalling 3,141,622 bytes. This column gives - some feel for how compression varies with block size. - These figures tend to understate the advantage of larger - block sizes for larger files, since the Corpus is domi­ - nated by smaller files. - - Compress Decompress Decompress Corpus - Flag usage usage -s usage Size - - -1 1200k 500k 350k 914704 - -2 2000k 900k 600k 877703 - -3 2800k 1300k 850k 860338 - -4 3600k 1700k 1100k 846899 - -5 4400k 2100k 1350k 845160 - -6 5200k 2500k 1600k 838626 - -7 6100k 2900k 1850k 834096 - -8 6800k 3300k 2100k 828642 - -9 7600k 3700k 2350k 828642 - - -RREECCOOVVEERRIINNGG DDAATTAA FFRROOMM DDAAMMAAGGEEDD FFIILLEESS - _b_z_i_p_2 compresses files in blocks, usually 900kbytes long. - Each block is handled independently. If a media or trans­ - mission error causes a multi-block .bz2 file to become - damaged, it may be possible to recover data from the - undamaged blocks in the file. - - The compressed representation of each block is delimited - by a 48-bit pattern, which makes it possible to find the - block boundaries with reasonable certainty. Each block - also carries its own 32-bit CRC, so damaged blocks can be - distinguished from undamaged ones. - - _b_z_i_p_2_r_e_c_o_v_e_r is a simple program whose purpose is to - search for blocks in .bz2 files, and write each block out - into its own .bz2 file. You can then use _b_z_i_p_2 -t to test - the integrity of the resulting files, and decompress those - which are undamaged. - - _b_z_i_p_2_r_e_c_o_v_e_r takes a single argument, the name of the dam­ - aged file, and writes a number of files - "rec00001file.bz2", "rec00002file.bz2", etc, containing - the extracted blocks. The output filenames are - designed so that the use of wildcards in subsequent pro­ - cessing -- for example, "bzip2 -dc rec*file.bz2 > recov­ - ered_data" -- processes the files in the correct order. - - _b_z_i_p_2_r_e_c_o_v_e_r should be of most use dealing with large .bz2 - files, as these will contain many blocks. It is clearly - futile to use it on damaged single-block files, since a - damaged block cannot be recovered. If you wish to min­ - imise any potential data loss through media or transmis­ - sion errors, you might consider compressing with a smaller - block size. - - -PPEERRFFOORRMMAANNCCEE NNOOTTEESS - The sorting phase of compression gathers together similar - strings in the file. Because of this, files containing - very long runs of repeated symbols, like "aabaabaabaab - ..." (repeated several hundred times) may compress more - slowly than normal. Versions 0.9.5 and above fare much - better than previous versions in this respect. The ratio - between worst-case and average-case compression time is in - the region of 10:1. For previous versions, this figure - was more like 100:1. You can use the -vvvv option to mon­ - itor progress in great detail, if you want. - - Decompression speed is unaffected by these phenomena. - - _b_z_i_p_2 usually allocates several megabytes of memory to - operate in, and then charges all over it in a fairly ran­ - dom fashion. This means that performance, both for com­ - pressing and decompressing, is largely determined by the - speed at which your machine can service cache misses. - Because of this, small changes to the code to reduce the - miss rate have been observed to give disproportionately - large performance improvements. I imagine _b_z_i_p_2 will per­ - form best on machines with very large caches. - - -CCAAVVEEAATTSS - I/O error messages are not as helpful as they could be. - _b_z_i_p_2 tries hard to detect I/O errors and exit cleanly, - but the details of what the problem is sometimes seem - rather misleading. - - This manual page pertains to version 1.0.2 of _b_z_i_p_2_. Com­ - pressed data created by this version is entirely forwards - and backwards compatible with the previous public - releases, versions 0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1, - but with the following exception: 0.9.0 and above can cor­ - rectly decompress multiple concatenated compressed files. - 0.1pl2 cannot do this; it will stop after decompressing - just the first file in the stream. - - _b_z_i_p_2_r_e_c_o_v_e_r versions prior to this one, 1.0.2, used - 32-bit integers to represent bit positions in compressed - files, so it could not handle compressed files more than - 512 megabytes long. Version 1.0.2 and above uses 64-bit - ints on some platforms which support them (GNU supported - targets, and Windows). To establish whether or not - bzip2recover was built with such a limitation, run it - without arguments. In any event you can build yourself an - unlimited version if you can recompile it with MaybeUInt64 - set to be an unsigned 64-bit integer. - - - - -AAUUTTHHOORR - Julian Seward, jseward@acm.org. - - http://sources.redhat.com/bzip2 - - The ideas embodied in _b_z_i_p_2 are due to (at least) the fol­ - lowing people: Michael Burrows and David Wheeler (for the - block sorting transformation), David Wheeler (again, for - the Huffman coder), Peter Fenwick (for the structured cod­ - ing model in the original _b_z_i_p_, and many refinements), and - Alistair Moffat, Radford Neal and Ian Witten (for the - arithmetic coder in the original _b_z_i_p_)_. I am much - indebted for their help, support and advice. See the man­ - ual in the source distribution for pointers to sources of - documentation. Christian von Roques encouraged me to look - for faster sorting algorithms, so as to speed up compres­ - sion. Bela Lubkin encouraged me to improve the worst-case - compression performance. The bz* scripts are derived from - those of GNU gzip. Many people sent patches, helped with - portability problems, lent machines, gave advice and were - generally helpful. - - - - bzip2(1) diff -Nru bzip2-1.0.2/bzless bzip2-1.0.2.new/bzless --- bzip2-1.0.2/bzless Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/bzless Fri Feb 1 04:19:11 2002 @@ -0,0 +1,2 @@ +#!/bin/sh +%{_bindir}/bunzip2 -c "$@" | %{_bindir}/less diff -Nru bzip2-1.0.2/bzmore.1 bzip2-1.0.2.new/bzmore.1 --- bzip2-1.0.2/bzmore.1 Sun Dec 30 03:12:35 2001 +++ bzip2-1.0.2.new/bzmore.1 Thu Jan 1 01:00:00 1970 @@ -1,152 +0,0 @@ -.\"Shamelessly copied from zmore.1 by Philippe Troin -.\"for Debian GNU/Linux -.TH BZMORE 1 -.SH NAME -bzmore, bzless \- file perusal filter for crt viewing of bzip2 compressed text -.SH SYNOPSIS -.B bzmore -[ name ... ] -.br -.B bzless -[ name ... ] -.SH NOTE -In the following description, -.I bzless -and -.I less -can be used interchangeably with -.I bzmore -and -.I more. -.SH DESCRIPTION -.I Bzmore -is a filter which allows examination of compressed or plain text files -one screenful at a time on a soft-copy terminal. -.I bzmore -works on files compressed with -.I bzip2 -and also on uncompressed files. -If a file does not exist, -.I bzmore -looks for a file of the same name with the addition of a .bz2 suffix. -.PP -.I Bzmore -normally pauses after each screenful, printing --More-- -at the bottom of the screen. -If the user then types a carriage return, one more line is displayed. -If the user hits a space, -another screenful is displayed. Other possibilities are enumerated later. -.PP -.I Bzmore -looks in the file -.I /etc/termcap -to determine terminal characteristics, -and to determine the default window size. -On a terminal capable of displaying 24 lines, -the default window size is 22 lines. -Other sequences which may be typed when -.I bzmore -pauses, and their effects, are as follows (\fIi\fP is an optional integer -argument, defaulting to 1) : -.PP -.IP \fIi\|\fP -display -.I i -more lines, (or another screenful if no argument is given) -.PP -.IP ^D -display 11 more lines (a ``scroll''). -If -.I i -is given, then the scroll size is set to \fIi\|\fP. -.PP -.IP d -same as ^D (control-D) -.PP -.IP \fIi\|\fPz -same as typing a space except that \fIi\|\fP, if present, becomes the new -window size. Note that the window size reverts back to the default at the -end of the current file. -.PP -.IP \fIi\|\fPs -skip \fIi\|\fP lines and print a screenful of lines -.PP -.IP \fIi\|\fPf -skip \fIi\fP screenfuls and print a screenful of lines -.PP -.IP "q or Q" -quit reading the current file; go on to the next (if any) -.PP -.IP "e or q" -When the prompt --More--(Next file: -.IR file ) -is printed, this command causes bzmore to exit. -.PP -.IP s -When the prompt --More--(Next file: -.IR file ) -is printed, this command causes bzmore to skip the next file and continue. -.PP -.IP = -Display the current line number. -.PP -.IP \fIi\|\fP/expr -search for the \fIi\|\fP-th occurrence of the regular expression \fIexpr.\fP -If the pattern is not found, -.I bzmore -goes on to the next file (if any). -Otherwise, a screenful is displayed, starting two lines before the place -where the expression was found. -The user's erase and kill characters may be used to edit the regular -expression. -Erasing back past the first column cancels the search command. -.PP -.IP \fIi\|\fPn -search for the \fIi\|\fP-th occurrence of the last regular expression entered. -.PP -.IP !command -invoke a shell with \fIcommand\|\fP. -The character `!' in "command" are replaced with the -previous shell command. The sequence "\\!" is replaced by "!". -.PP -.IP ":q or :Q" -quit reading the current file; go on to the next (if any) -(same as q or Q). -.PP -.IP . -(dot) repeat the previous command. -.PP -The commands take effect immediately, i.e., it is not necessary to -type a carriage return. -Up to the time when the command character itself is given, -the user may hit the line kill character to cancel the numerical -argument being formed. -In addition, the user may hit the erase character to redisplay the ---More-- message. -.PP -At any time when output is being sent to the terminal, the user can -hit the quit key (normally control\-\\). -.I Bzmore -will stop sending output, and will display the usual --More-- -prompt. -The user may then enter one of the above commands in the normal manner. -Unfortunately, some output is lost when this is done, due to the -fact that any characters waiting in the terminal's output queue -are flushed when the quit signal occurs. -.PP -The terminal is set to -.I noecho -mode by this program so that the output can be continuous. -What you type will thus not show on your terminal, except for the / and ! -commands. -.PP -If the standard output is not a teletype, then -.I bzmore -acts just like -.I bzcat, -except that a header is printed before each file. -.SH FILES -.DT -/etc/termcap Terminal data base -.SH "SEE ALSO" -more(1), less(1), bzip2(1), bzdiff(1), bzgrep(1) diff -Nru bzip2-1.0.2/config.h.in bzip2-1.0.2.new/config.h.in --- bzip2-1.0.2/config.h.in Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/config.h.in Fri Feb 1 04:19:11 2002 @@ -0,0 +1,17 @@ +/* config.h.in. Generated automatically from configure.in by autoheader. */ + +/* Name of package */ +#undef PACKAGE + +/* Version number of package */ +#undef VERSION + +/* Number of bits in a file offset, on hosts where this is settable. */ +#undef _FILE_OFFSET_BITS + +/* Define to make fseeko etc. visible, on some hosts. */ +#undef _LARGEFILE_SOURCE + +/* Define for large files, on AIX-style hosts. */ +#undef _LARGE_FILES + diff -Nru bzip2-1.0.2/configure.in bzip2-1.0.2.new/configure.in --- bzip2-1.0.2/configure.in Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/configure.in Fri Feb 1 04:19:11 2002 @@ -0,0 +1,10 @@ +AC_INIT(bzip2.c) +AM_INIT_AUTOMAKE(bzip2,1.0.1) +AM_CONFIG_HEADER(config.h) +AC_PROG_CC +AM_PROG_LIBTOOL +AC_PROG_LN_S +AC_SYS_LARGEFILE +AC_OUTPUT(Makefile + doc/Makefile + doc/pl/Makefile) diff -Nru bzip2-1.0.2/crctable.c bzip2-1.0.2.new/crctable.c --- bzip2-1.0.2/crctable.c Sun Dec 30 03:19:28 2001 +++ bzip2-1.0.2.new/crctable.c Fri Feb 1 04:19:11 2002 @@ -58,6 +58,10 @@ For more information on these sources, see the manual. --*/ +#ifdef HAVE_CONFIG_H +#include +#endif + #include "bzlib_private.h" diff -Nru bzip2-1.0.2/decompress.c bzip2-1.0.2.new/decompress.c --- bzip2-1.0.2/decompress.c Sun Dec 30 21:45:53 2001 +++ bzip2-1.0.2.new/decompress.c Fri Feb 1 04:19:11 2002 @@ -58,6 +58,10 @@ For more information on these sources, see the manual. --*/ +#ifdef HAVE_CONFIG_H +#include +#endif + #include "bzlib_private.h" diff -Nru bzip2-1.0.2/dlltest.c bzip2-1.0.2.new/dlltest.c --- bzip2-1.0.2/dlltest.c Sun Dec 30 20:44:07 2001 +++ bzip2-1.0.2.new/dlltest.c Fri Feb 1 04:19:11 2002 @@ -8,6 +8,10 @@ usage: minibz2 [-d] [-{1,2,..9}] [[srcfilename] destfilename] */ +#ifdef HAVE_CONFIG_H +#include +#endif + #define BZ_IMPORT #include #include diff -Nru bzip2-1.0.2/doc/Makefile.am bzip2-1.0.2.new/doc/Makefile.am --- bzip2-1.0.2/doc/Makefile.am Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/Makefile.am Fri Feb 1 04:27:21 2002 @@ -0,0 +1,6 @@ + +SUBDIRS = pl + +man_MANS = bunzip2.1 bzcat.1 bzdiff.1 bzgrep.1 bzip2.1 \ + bzip2recover.1 bzmore.1 +#info_TEXINFOS = bzip2.texi diff -Nru bzip2-1.0.2/doc/bunzip2.1 bzip2-1.0.2.new/doc/bunzip2.1 --- bzip2-1.0.2/doc/bunzip2.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bunzip2.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1 @@ +.so bzip2.1 \ No newline at end of file diff -Nru bzip2-1.0.2/doc/bzcat.1 bzip2-1.0.2.new/doc/bzcat.1 --- bzip2-1.0.2/doc/bzcat.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzcat.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1 @@ +.so bzip2.1 \ No newline at end of file diff -Nru bzip2-1.0.2/doc/bzdiff.1 bzip2-1.0.2.new/doc/bzdiff.1 --- bzip2-1.0.2/doc/bzdiff.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzdiff.1 Sun Dec 30 03:12:35 2001 @@ -0,0 +1,47 @@ +\"Shamelessly copied from zmore.1 by Philippe Troin +\"for Debian GNU/Linux +.TH BZDIFF 1 +.SH NAME +bzcmp, bzdiff \- compare bzip2 compressed files +.SH SYNOPSIS +.B bzcmp +[ cmp_options ] file1 +[ file2 ] +.br +.B bzdiff +[ diff_options ] file1 +[ file2 ] +.SH DESCRIPTION +.I Bzcmp +and +.I bzdiff +are used to invoke the +.I cmp +or the +.I diff +program on bzip2 compressed files. All options specified are passed +directly to +.I cmp +or +.IR diff "." +If only 1 file is specified, then the files compared are +.I file1 +and an uncompressed +.IR file1 ".bz2." +If two files are specified, then they are uncompressed if necessary and fed to +.I cmp +or +.IR diff "." +The exit status from +.I cmp +or +.I diff +is preserved. +.SH "SEE ALSO" +cmp(1), diff(1), bzmore(1), bzless(1), bzgrep(1), bzip2(1) +.SH BUGS +Messages from the +.I cmp +or +.I diff +programs refer to temporary filenames instead of those specified. diff -Nru bzip2-1.0.2/doc/bzgrep.1 bzip2-1.0.2.new/doc/bzgrep.1 --- bzip2-1.0.2/doc/bzgrep.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzgrep.1 Sun Dec 30 03:12:35 2001 @@ -0,0 +1,56 @@ +\"Shamelessly copied from zmore.1 by Philippe Troin +\"for Debian GNU/Linux +.TH BZGREP 1 +.SH NAME +bzgrep, bzfgrep, bzegrep \- search possibly bzip2 compressed files for a regular expression +.SH SYNOPSIS +.B bzgrep +[ grep_options ] +.BI [\ -e\ ] " pattern" +.IR filename ".\|.\|." +.br +.B bzegrep +[ egrep_options ] +.BI [\ -e\ ] " pattern" +.IR filename ".\|.\|." +.br +.B bzfgrep +[ fgrep_options ] +.BI [\ -e\ ] " pattern" +.IR filename ".\|.\|." +.SH DESCRIPTION +.IR Bzgrep +is used to invoke the +.I grep +on bzip2-compressed files. All options specified are passed directly to +.I grep. +If no file is specified, then the standard input is decompressed +if necessary and fed to grep. +Otherwise the given files are uncompressed if necessary and fed to +.I grep. +.PP +If +.I bzgrep +is invoked as +.I bzegrep +or +.I bzfgrep +then +.I egrep +or +.I fgrep +is used instead of +.I grep. +If the GREP environment variable is set, +.I bzgrep +uses it as the +.I grep +program to be invoked. For example: + + for sh: GREP=fgrep bzgrep string files + for csh: (setenv GREP fgrep; bzgrep string files) +.SH AUTHOR +Charles Levert (charles@comm.polymtl.ca). Adapted to bzip2 by Philippe +Troin for Debian GNU/Linux. +.SH "SEE ALSO" +grep(1), egrep(1), fgrep(1), bzdiff(1), bzmore(1), bzless(1), bzip2(1) diff -Nru bzip2-1.0.2/doc/bzip2.1 bzip2-1.0.2.new/doc/bzip2.1 --- bzip2-1.0.2/doc/bzip2.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzip2.1 Thu Jan 3 00:14:36 2002 @@ -0,0 +1,453 @@ +.PU +.TH bzip2 1 +.SH NAME +bzip2, bunzip2 \- a block-sorting file compressor, v1.0.2 +.br +bzcat \- decompresses files to stdout +.br +bzip2recover \- recovers data from damaged bzip2 files + +.SH SYNOPSIS +.ll +8 +.B bzip2 +.RB [ " \-cdfkqstvzVL123456789 " ] +[ +.I "filenames \&..." +] +.ll -8 +.br +.B bunzip2 +.RB [ " \-fkvsVL " ] +[ +.I "filenames \&..." +] +.br +.B bzcat +.RB [ " \-s " ] +[ +.I "filenames \&..." +] +.br +.B bzip2recover +.I "filename" + +.SH DESCRIPTION +.I bzip2 +compresses files using the Burrows-Wheeler block sorting +text compression algorithm, and Huffman coding. Compression is +generally considerably better than that achieved by more conventional +LZ77/LZ78-based compressors, and approaches the performance of the PPM +family of statistical compressors. + +The command-line options are deliberately very similar to +those of +.I GNU gzip, +but they are not identical. + +.I bzip2 +expects a list of file names to accompany the +command-line flags. Each file is replaced by a compressed version of +itself, with the name "original_name.bz2". +Each compressed file +has the same modification date, permissions, and, when possible, +ownership as the corresponding original, so that these properties can +be correctly restored at decompression time. File name handling is +naive in the sense that there is no mechanism for preserving original +file names, permissions, ownerships or dates in filesystems which lack +these concepts, or have serious file name length restrictions, such as +MS-DOS. + +.I bzip2 +and +.I bunzip2 +will by default not overwrite existing +files. If you want this to happen, specify the \-f flag. + +If no file names are specified, +.I bzip2 +compresses from standard +input to standard output. In this case, +.I bzip2 +will decline to +write compressed output to a terminal, as this would be entirely +incomprehensible and therefore pointless. + +.I bunzip2 +(or +.I bzip2 \-d) +decompresses all +specified files. Files which were not created by +.I bzip2 +will be detected and ignored, and a warning issued. +.I bzip2 +attempts to guess the filename for the decompressed file +from that of the compressed file as follows: + + filename.bz2 becomes filename + filename.bz becomes filename + filename.tbz2 becomes filename.tar + filename.tbz becomes filename.tar + anyothername becomes anyothername.out + +If the file does not end in one of the recognised endings, +.I .bz2, +.I .bz, +.I .tbz2 +or +.I .tbz, +.I bzip2 +complains that it cannot +guess the name of the original file, and uses the original name +with +.I .out +appended. + +As with compression, supplying no +filenames causes decompression from +standard input to standard output. + +.I bunzip2 +will correctly decompress a file which is the +concatenation of two or more compressed files. The result is the +concatenation of the corresponding uncompressed files. Integrity +testing (\-t) +of concatenated +compressed files is also supported. + +You can also compress or decompress files to the standard output by +giving the \-c flag. Multiple files may be compressed and +decompressed like this. The resulting outputs are fed sequentially to +stdout. Compression of multiple files +in this manner generates a stream +containing multiple compressed file representations. Such a stream +can be decompressed correctly only by +.I bzip2 +version 0.9.0 or +later. Earlier versions of +.I bzip2 +will stop after decompressing +the first file in the stream. + +.I bzcat +(or +.I bzip2 -dc) +decompresses all specified files to +the standard output. + +.I bzip2 +will read arguments from the environment variables +.I BZIP2 +and +.I BZIP, +in that order, and will process them +before any arguments read from the command line. This gives a +convenient way to supply default arguments. + +Compression is always performed, even if the compressed +file is slightly +larger than the original. Files of less than about one hundred bytes +tend to get larger, since the compression mechanism has a constant +overhead in the region of 50 bytes. Random data (including the output +of most file compressors) is coded at about 8.05 bits per byte, giving +an expansion of around 0.5%. + +As a self-check for your protection, +.I +bzip2 +uses 32-bit CRCs to +make sure that the decompressed version of a file is identical to the +original. This guards against corruption of the compressed data, and +against undetected bugs in +.I bzip2 +(hopefully very unlikely). The +chances of data corruption going undetected is microscopic, about one +chance in four billion for each file processed. Be aware, though, that +the check occurs upon decompression, so it can only tell you that +something is wrong. It can't help you +recover the original uncompressed +data. You can use +.I bzip2recover +to try to recover data from +damaged files. + +Return values: 0 for a normal exit, 1 for environmental problems (file +not found, invalid flags, I/O errors, &c), 2 to indicate a corrupt +compressed file, 3 for an internal consistency error (eg, bug) which +caused +.I bzip2 +to panic. + +.SH OPTIONS +.TP +.B \-c --stdout +Compress or decompress to standard output. +.TP +.B \-d --decompress +Force decompression. +.I bzip2, +.I bunzip2 +and +.I bzcat +are +really the same program, and the decision about what actions to take is +done on the basis of which name is used. This flag overrides that +mechanism, and forces +.I bzip2 +to decompress. +.TP +.B \-z --compress +The complement to \-d: forces compression, regardless of the +invocation name. +.TP +.B \-t --test +Check integrity of the specified file(s), but don't decompress them. +This really performs a trial decompression and throws away the result. +.TP +.B \-f --force +Force overwrite of output files. Normally, +.I bzip2 +will not overwrite +existing output files. Also forces +.I bzip2 +to break hard links +to files, which it otherwise wouldn't do. + +bzip2 normally declines to decompress files which don't have the +correct magic header bytes. If forced (-f), however, it will pass +such files through unmodified. This is how GNU gzip behaves. +.TP +.B \-k --keep +Keep (don't delete) input files during compression +or decompression. +.TP +.B \-s --small +Reduce memory usage, for compression, decompression and testing. Files +are decompressed and tested using a modified algorithm which only +requires 2.5 bytes per block byte. This means any file can be +decompressed in 2300k of memory, albeit at about half the normal speed. + +During compression, \-s selects a block size of 200k, which limits +memory use to around the same figure, at the expense of your compression +ratio. In short, if your machine is low on memory (8 megabytes or +less), use \-s for everything. See MEMORY MANAGEMENT below. +.TP +.B \-q --quiet +Suppress non-essential warning messages. Messages pertaining to +I/O errors and other critical events will not be suppressed. +.TP +.B \-v --verbose +Verbose mode -- show the compression ratio for each file processed. +Further \-v's increase the verbosity level, spewing out lots of +information which is primarily of interest for diagnostic purposes. +.TP +.B \-L --license -V --version +Display the software version, license terms and conditions. +.TP +.B \-1 (or \-\-fast) to \-9 (or \-\-best) +Set the block size to 100 k, 200 k .. 900 k when compressing. Has no +effect when decompressing. See MEMORY MANAGEMENT below. +The \-\-fast and \-\-best aliases are primarily for GNU gzip +compatibility. In particular, \-\-fast doesn't make things +significantly faster. +And \-\-best merely selects the default behaviour. +.TP +.B \-- +Treats all subsequent arguments as file names, even if they start +with a dash. This is so you can handle files with names beginning +with a dash, for example: bzip2 \-- \-myfilename. +.TP +.B \--repetitive-fast --repetitive-best +These flags are redundant in versions 0.9.5 and above. They provided +some coarse control over the behaviour of the sorting algorithm in +earlier versions, which was sometimes useful. 0.9.5 and above have an +improved algorithm which renders these flags irrelevant. + +.SH MEMORY MANAGEMENT +.I bzip2 +compresses large files in blocks. The block size affects +both the compression ratio achieved, and the amount of memory needed for +compression and decompression. The flags \-1 through \-9 +specify the block size to be 100,000 bytes through 900,000 bytes (the +default) respectively. At decompression time, the block size used for +compression is read from the header of the compressed file, and +.I bunzip2 +then allocates itself just enough memory to decompress +the file. Since block sizes are stored in compressed files, it follows +that the flags \-1 to \-9 are irrelevant to and so ignored +during decompression. + +Compression and decompression requirements, +in bytes, can be estimated as: + + Compression: 400k + ( 8 x block size ) + + Decompression: 100k + ( 4 x block size ), or + 100k + ( 2.5 x block size ) + +Larger block sizes give rapidly diminishing marginal returns. Most of +the compression comes from the first two or three hundred k of block +size, a fact worth bearing in mind when using +.I bzip2 +on small machines. +It is also important to appreciate that the decompression memory +requirement is set at compression time by the choice of block size. + +For files compressed with the default 900k block size, +.I bunzip2 +will require about 3700 kbytes to decompress. To support decompression +of any file on a 4 megabyte machine, +.I bunzip2 +has an option to +decompress using approximately half this amount of memory, about 2300 +kbytes. Decompression speed is also halved, so you should use this +option only where necessary. The relevant flag is -s. + +In general, try and use the largest block size memory constraints allow, +since that maximises the compression achieved. Compression and +decompression speed are virtually unaffected by block size. + +Another significant point applies to files which fit in a single block +-- that means most files you'd encounter using a large block size. The +amount of real memory touched is proportional to the size of the file, +since the file is smaller than a block. For example, compressing a file +20,000 bytes long with the flag -9 will cause the compressor to +allocate around 7600k of memory, but only touch 400k + 20000 * 8 = 560 +kbytes of it. Similarly, the decompressor will allocate 3700k but only +touch 100k + 20000 * 4 = 180 kbytes. + +Here is a table which summarises the maximum memory usage for different +block sizes. Also recorded is the total compressed size for 14 files of +the Calgary Text Compression Corpus totalling 3,141,622 bytes. This +column gives some feel for how compression varies with block size. +These figures tend to understate the advantage of larger block sizes for +larger files, since the Corpus is dominated by smaller files. + + Compress Decompress Decompress Corpus + Flag usage usage -s usage Size + + -1 1200k 500k 350k 914704 + -2 2000k 900k 600k 877703 + -3 2800k 1300k 850k 860338 + -4 3600k 1700k 1100k 846899 + -5 4400k 2100k 1350k 845160 + -6 5200k 2500k 1600k 838626 + -7 6100k 2900k 1850k 834096 + -8 6800k 3300k 2100k 828642 + -9 7600k 3700k 2350k 828642 + +.SH RECOVERING DATA FROM DAMAGED FILES +.I bzip2 +compresses files in blocks, usually 900kbytes long. Each +block is handled independently. If a media or transmission error causes +a multi-block .bz2 +file to become damaged, it may be possible to +recover data from the undamaged blocks in the file. + +The compressed representation of each block is delimited by a 48-bit +pattern, which makes it possible to find the block boundaries with +reasonable certainty. Each block also carries its own 32-bit CRC, so +damaged blocks can be distinguished from undamaged ones. + +.I bzip2recover +is a simple program whose purpose is to search for +blocks in .bz2 files, and write each block out into its own .bz2 +file. You can then use +.I bzip2 +\-t +to test the +integrity of the resulting files, and decompress those which are +undamaged. + +.I bzip2recover +takes a single argument, the name of the damaged file, +and writes a number of files "rec00001file.bz2", +"rec00002file.bz2", etc, containing the extracted blocks. +The output filenames are designed so that the use of +wildcards in subsequent processing -- for example, +"bzip2 -dc rec*file.bz2 > recovered_data" -- processes the files in +the correct order. + +.I bzip2recover +should be of most use dealing with large .bz2 +files, as these will contain many blocks. It is clearly +futile to use it on damaged single-block files, since a +damaged block cannot be recovered. If you wish to minimise +any potential data loss through media or transmission errors, +you might consider compressing with a smaller +block size. + +.SH PERFORMANCE NOTES +The sorting phase of compression gathers together similar strings in the +file. Because of this, files containing very long runs of repeated +symbols, like "aabaabaabaab ..." (repeated several hundred times) may +compress more slowly than normal. Versions 0.9.5 and above fare much +better than previous versions in this respect. The ratio between +worst-case and average-case compression time is in the region of 10:1. +For previous versions, this figure was more like 100:1. You can use the +\-vvvv option to monitor progress in great detail, if you want. + +Decompression speed is unaffected by these phenomena. + +.I bzip2 +usually allocates several megabytes of memory to operate +in, and then charges all over it in a fairly random fashion. This means +that performance, both for compressing and decompressing, is largely +determined by the speed at which your machine can service cache misses. +Because of this, small changes to the code to reduce the miss rate have +been observed to give disproportionately large performance improvements. +I imagine +.I bzip2 +will perform best on machines with very large caches. + +.SH CAVEATS +I/O error messages are not as helpful as they could be. +.I bzip2 +tries hard to detect I/O errors and exit cleanly, but the details of +what the problem is sometimes seem rather misleading. + +This manual page pertains to version 1.0.2 of +.I bzip2. +Compressed data created by this version is entirely forwards and +backwards compatible with the previous public releases, versions +0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1, but with the following +exception: 0.9.0 and above can correctly decompress multiple +concatenated compressed files. 0.1pl2 cannot do this; it will stop +after decompressing just the first file in the stream. + +.I bzip2recover +versions prior to this one, 1.0.2, used 32-bit integers to represent +bit positions in compressed files, so it could not handle compressed +files more than 512 megabytes long. Version 1.0.2 and above uses +64-bit ints on some platforms which support them (GNU supported +targets, and Windows). To establish whether or not bzip2recover was +built with such a limitation, run it without arguments. In any event +you can build yourself an unlimited version if you can recompile it +with MaybeUInt64 set to be an unsigned 64-bit integer. + + + +.SH AUTHOR +Julian Seward, jseward@acm.org. + +http://sources.redhat.com/bzip2 + +The ideas embodied in +.I bzip2 +are due to (at least) the following +people: Michael Burrows and David Wheeler (for the block sorting +transformation), David Wheeler (again, for the Huffman coder), Peter +Fenwick (for the structured coding model in the original +.I bzip, +and many refinements), and Alistair Moffat, Radford Neal and Ian Witten +(for the arithmetic coder in the original +.I bzip). +I am much +indebted for their help, support and advice. See the manual in the +source distribution for pointers to sources of documentation. Christian +von Roques encouraged me to look for faster sorting algorithms, so as to +speed up compression. Bela Lubkin encouraged me to improve the +worst-case compression performance. +The bz* scripts are derived from those of GNU gzip. +Many people sent patches, helped +with portability problems, lent machines, gave advice and were generally +helpful. diff -Nru bzip2-1.0.2/doc/bzip2.texi bzip2-1.0.2.new/doc/bzip2.texi --- bzip2-1.0.2/doc/bzip2.texi Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzip2.texi Fri Feb 1 04:26:21 2002 @@ -0,0 +1,2234 @@ +\input texinfo @c -*- Texinfo -*- +@setfilename bzip2.info + +@ignore +This file documents bzip2 version 1.0.2, and associated library +libbzip2, written by Julian Seward (jseward@acm.org). + +Copyright (C) 1996-2002 Julian R Seward + +Permission is granted to make and distribute verbatim copies of +this manual provided the copyright notice and this permission notice +are preserved on all copies. + +Permission is granted to copy and distribute translations of this manual +into another language, under the above conditions for verbatim copies. +@end ignore + +@iftex +@c @finalout +@settitle bzip2 and libbzip2 +@titlepage +@title bzip2 and libbzip2 +@subtitle a program and library for data compression +@subtitle copyright (C) 1996-2002 Julian Seward +@subtitle version 1.0.2 of 30 December 2001 +@author Julian Seward + +@end titlepage + +@parindent 0mm +@parskip 2mm + +@end iftex +@node Top,,, (dir) + +The following text is the License for this software. You should +find it identical to that contained in the file LICENSE in the +source distribution. + +------------------ START OF THE LICENSE ------------------ + +This program, @code{bzip2}, +and associated library @code{libbzip2}, are +Copyright (C) 1996-2002 Julian R Seward. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions +are met: +@itemize @bullet +@item + Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. +@item + The origin of this software must not be misrepresented; you must + not claim that you wrote the original software. If you use this + software in a product, an acknowledgment in the product + documentation would be appreciated but is not required. +@item + Altered source versions must be plainly marked as such, and must + not be misrepresented as being the original software. +@item + The name of the author may not be used to endorse or promote + products derived from this software without specific prior written + permission. +@end itemize +THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS +OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED +WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY +DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE +GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, +WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING +NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +Julian Seward, Cambridge, UK. + +@code{jseward@@acm.org} + +@code{bzip2}/@code{libbzip2} version 1.0.2 of 30 December 2001. + +------------------ END OF THE LICENSE ------------------ + +Web sites: + +@code{http://sources.redhat.com/bzip2} + +@code{http://www.cacheprof.org} + +PATENTS: To the best of my knowledge, @code{bzip2} does not use any patented +algorithms. However, I do not have the resources available to carry out +a full patent search. Therefore I cannot give any guarantee of the +above statement. + + + + + + + +@chapter Introduction + +@code{bzip2} compresses files using the Burrows-Wheeler +block-sorting text compression algorithm, and Huffman coding. +Compression is generally considerably better than that +achieved by more conventional LZ77/LZ78-based compressors, +and approaches the performance of the PPM family of statistical compressors. + +@code{bzip2} is built on top of @code{libbzip2}, a flexible library +for handling compressed data in the @code{bzip2} format. This manual +describes both how to use the program and +how to work with the library interface. Most of the +manual is devoted to this library, not the program, +which is good news if your interest is only in the program. + +Chapter 2 describes how to use @code{bzip2}; this is the only part +you need to read if you just want to know how to operate the program. +Chapter 3 describes the programming interfaces in detail, and +Chapter 4 records some miscellaneous notes which I thought +ought to be recorded somewhere. + + +@chapter How to use @code{bzip2} + +This chapter contains a copy of the @code{bzip2} man page, +and nothing else. + +@quotation + +@unnumberedsubsubsec NAME +@itemize +@item @code{bzip2}, @code{bunzip2} +- a block-sorting file compressor, v1.0.2 +@item @code{bzcat} +- decompresses files to stdout +@item @code{bzip2recover} +- recovers data from damaged bzip2 files +@end itemize + +@unnumberedsubsubsec SYNOPSIS +@itemize +@item @code{bzip2} [ -cdfkqstvzVL123456789 ] [ filenames ... ] +@item @code{bunzip2} [ -fkvsVL ] [ filenames ... ] +@item @code{bzcat} [ -s ] [ filenames ... ] +@item @code{bzip2recover} filename +@end itemize + +@unnumberedsubsubsec DESCRIPTION + +@code{bzip2} compresses files using the Burrows-Wheeler block sorting +text compression algorithm, and Huffman coding. Compression is +generally considerably better than that achieved by more conventional +LZ77/LZ78-based compressors, and approaches the performance of the PPM +family of statistical compressors. + +The command-line options are deliberately very similar to those of GNU +@code{gzip}, but they are not identical. + +@code{bzip2} expects a list of file names to accompany the command-line +flags. Each file is replaced by a compressed version of itself, with +the name @code{original_name.bz2}. Each compressed file has the same +modification date, permissions, and, when possible, ownership as the +corresponding original, so that these properties can be correctly +restored at decompression time. File name handling is naive in the +sense that there is no mechanism for preserving original file names, +permissions, ownerships or dates in filesystems which lack these +concepts, or have serious file name length restrictions, such as MS-DOS. + +@code{bzip2} and @code{bunzip2} will by default not overwrite existing +files. If you want this to happen, specify the @code{-f} flag. + +If no file names are specified, @code{bzip2} compresses from standard +input to standard output. In this case, @code{bzip2} will decline to +write compressed output to a terminal, as this would be entirely +incomprehensible and therefore pointless. + +@code{bunzip2} (or @code{bzip2 -d}) decompresses all +specified files. Files which were not created by @code{bzip2} +will be detected and ignored, and a warning issued. +@code{bzip2} attempts to guess the filename for the decompressed file +from that of the compressed file as follows: +@itemize +@item @code{filename.bz2 } becomes @code{filename} +@item @code{filename.bz } becomes @code{filename} +@item @code{filename.tbz2} becomes @code{filename.tar} +@item @code{filename.tbz } becomes @code{filename.tar} +@item @code{anyothername } becomes @code{anyothername.out} +@end itemize +If the file does not end in one of the recognised endings, +@code{.bz2}, @code{.bz}, +@code{.tbz2} or @code{.tbz}, @code{bzip2} complains that it cannot +guess the name of the original file, and uses the original name +with @code{.out} appended. + +As with compression, supplying no +filenames causes decompression from standard input to standard output. + +@code{bunzip2} will correctly decompress a file which is the +concatenation of two or more compressed files. The result is the +concatenation of the corresponding uncompressed files. Integrity +testing (@code{-t}) of concatenated compressed files is also supported. + +You can also compress or decompress files to the standard output by +giving the @code{-c} flag. Multiple files may be compressed and +decompressed like this. The resulting outputs are fed sequentially to +stdout. Compression of multiple files in this manner generates a stream +containing multiple compressed file representations. Such a stream +can be decompressed correctly only by @code{bzip2} version 0.9.0 or +later. Earlier versions of @code{bzip2} will stop after decompressing +the first file in the stream. + +@code{bzcat} (or @code{bzip2 -dc}) decompresses all specified files to +the standard output. + +@code{bzip2} will read arguments from the environment variables +@code{BZIP2} and @code{BZIP}, in that order, and will process them +before any arguments read from the command line. This gives a +convenient way to supply default arguments. + +Compression is always performed, even if the compressed file is slightly +larger than the original. Files of less than about one hundred bytes +tend to get larger, since the compression mechanism has a constant +overhead in the region of 50 bytes. Random data (including the output +of most file compressors) is coded at about 8.05 bits per byte, giving +an expansion of around 0.5%. + +As a self-check for your protection, @code{bzip2} uses 32-bit CRCs to +make sure that the decompressed version of a file is identical to the +original. This guards against corruption of the compressed data, and +against undetected bugs in @code{bzip2} (hopefully very unlikely). The +chances of data corruption going undetected is microscopic, about one +chance in four billion for each file processed. Be aware, though, that +the check occurs upon decompression, so it can only tell you that +something is wrong. It can't help you recover the original uncompressed +data. You can use @code{bzip2recover} to try to recover data from +damaged files. + +Return values: 0 for a normal exit, 1 for environmental problems (file +not found, invalid flags, I/O errors, &c), 2 to indicate a corrupt +compressed file, 3 for an internal consistency error (eg, bug) which +caused @code{bzip2} to panic. + + +@unnumberedsubsubsec OPTIONS +@table @code +@item -c --stdout +Compress or decompress to standard output. +@item -d --decompress +Force decompression. @code{bzip2}, @code{bunzip2} and @code{bzcat} are +really the same program, and the decision about what actions to take is +done on the basis of which name is used. This flag overrides that +mechanism, and forces bzip2 to decompress. +@item -z --compress +The complement to @code{-d}: forces compression, regardless of the +invokation name. +@item -t --test +Check integrity of the specified file(s), but don't decompress them. +This really performs a trial decompression and throws away the result. +@item -f --force +Force overwrite of output files. Normally, @code{bzip2} will not overwrite +existing output files. Also forces @code{bzip2} to break hard links +to files, which it otherwise wouldn't do. + +@code{bzip2} normally declines to decompress files which don't have the +correct magic header bytes. If forced (@code{-f}), however, it will +pass such files through unmodified. This is how GNU @code{gzip} +behaves. +@item -k --keep +Keep (don't delete) input files during compression +or decompression. +@item -s --small +Reduce memory usage, for compression, decompression and testing. Files +are decompressed and tested using a modified algorithm which only +requires 2.5 bytes per block byte. This means any file can be +decompressed in 2300k of memory, albeit at about half the normal speed. + +During compression, @code{-s} selects a block size of 200k, which limits +memory use to around the same figure, at the expense of your compression +ratio. In short, if your machine is low on memory (8 megabytes or +less), use -s for everything. See MEMORY MANAGEMENT below. +@item -q --quiet +Suppress non-essential warning messages. Messages pertaining to +I/O errors and other critical events will not be suppressed. +@item -v --verbose +Verbose mode -- show the compression ratio for each file processed. +Further @code{-v}'s increase the verbosity level, spewing out lots of +information which is primarily of interest for diagnostic purposes. +@item -L --license -V --version +Display the software version, license terms and conditions. +@item -1 (or --fast) to -9 (or --best) +Set the block size to 100 k, 200 k .. 900 k when compressing. Has no +effect when decompressing. See MEMORY MANAGEMENT below. +The @code{--fast} and @code{--best} aliases are primarily for GNU +@code{gzip} compatibility. In particular, @code{--fast} doesn't make +things significantly faster. And @code{--best} merely selects the +default behaviour. +@item -- +Treats all subsequent arguments as file names, even if they start +with a dash. This is so you can handle files with names beginning +with a dash, for example: @code{bzip2 -- -myfilename}. +@item --repetitive-fast +@item --repetitive-best +These flags are redundant in versions 0.9.5 and above. They provided +some coarse control over the behaviour of the sorting algorithm in +earlier versions, which was sometimes useful. 0.9.5 and above have an +improved algorithm which renders these flags irrelevant. +@end table + + +@unnumberedsubsubsec MEMORY MANAGEMENT + +@code{bzip2} compresses large files in blocks. The block size affects +both the compression ratio achieved, and the amount of memory needed for +compression and decompression. The flags @code{-1} through @code{-9} +specify the block size to be 100,000 bytes through 900,000 bytes (the +default) respectively. At decompression time, the block size used for +compression is read from the header of the compressed file, and +@code{bunzip2} then allocates itself just enough memory to decompress +the file. Since block sizes are stored in compressed files, it follows +that the flags @code{-1} to @code{-9} are irrelevant to and so ignored +during decompression. + +Compression and decompression requirements, in bytes, can be estimated +as: +@example + Compression: 400k + ( 8 x block size ) + + Decompression: 100k + ( 4 x block size ), or + 100k + ( 2.5 x block size ) +@end example +Larger block sizes give rapidly diminishing marginal returns. Most of +the compression comes from the first two or three hundred k of block +size, a fact worth bearing in mind when using @code{bzip2} on small machines. +It is also important to appreciate that the decompression memory +requirement is set at compression time by the choice of block size. + +For files compressed with the default 900k block size, @code{bunzip2} +will require about 3700 kbytes to decompress. To support decompression +of any file on a 4 megabyte machine, @code{bunzip2} has an option to +decompress using approximately half this amount of memory, about 2300 +kbytes. Decompression speed is also halved, so you should use this +option only where necessary. The relevant flag is @code{-s}. + +In general, try and use the largest block size memory constraints allow, +since that maximises the compression achieved. Compression and +decompression speed are virtually unaffected by block size. + +Another significant point applies to files which fit in a single block +-- that means most files you'd encounter using a large block size. The +amount of real memory touched is proportional to the size of the file, +since the file is smaller than a block. For example, compressing a file +20,000 bytes long with the flag @code{-9} will cause the compressor to +allocate around 7600k of memory, but only touch 400k + 20000 * 8 = 560 +kbytes of it. Similarly, the decompressor will allocate 3700k but only +touch 100k + 20000 * 4 = 180 kbytes. + +Here is a table which summarises the maximum memory usage for different +block sizes. Also recorded is the total compressed size for 14 files of +the Calgary Text Compression Corpus totalling 3,141,622 bytes. This +column gives some feel for how compression varies with block size. +These figures tend to understate the advantage of larger block sizes for +larger files, since the Corpus is dominated by smaller files. +@example + Compress Decompress Decompress Corpus + Flag usage usage -s usage Size + + -1 1200k 500k 350k 914704 + -2 2000k 900k 600k 877703 + -3 2800k 1300k 850k 860338 + -4 3600k 1700k 1100k 846899 + -5 4400k 2100k 1350k 845160 + -6 5200k 2500k 1600k 838626 + -7 6100k 2900k 1850k 834096 + -8 6800k 3300k 2100k 828642 + -9 7600k 3700k 2350k 828642 +@end example + +@unnumberedsubsubsec RECOVERING DATA FROM DAMAGED FILES + +@code{bzip2} compresses files in blocks, usually 900kbytes long. Each +block is handled independently. If a media or transmission error causes +a multi-block @code{.bz2} file to become damaged, it may be possible to +recover data from the undamaged blocks in the file. + +The compressed representation of each block is delimited by a 48-bit +pattern, which makes it possible to find the block boundaries with +reasonable certainty. Each block also carries its own 32-bit CRC, so +damaged blocks can be distinguished from undamaged ones. + +@code{bzip2recover} is a simple program whose purpose is to search for +blocks in @code{.bz2} files, and write each block out into its own +@code{.bz2} file. You can then use @code{bzip2 -t} to test the +integrity of the resulting files, and decompress those which are +undamaged. + +@code{bzip2recover} +takes a single argument, the name of the damaged file, and writes a +number of files @code{rec00001file.bz2}, @code{rec00002file.bz2}, etc, +containing the extracted blocks. The output filenames are designed so +that the use of wildcards in subsequent processing -- for example, +@code{bzip2 -dc rec*file.bz2 > recovered_data} -- processes the files in +the correct order. + +@code{bzip2recover} should be of most use dealing with large @code{.bz2} +files, as these will contain many blocks. It is clearly futile to use +it on damaged single-block files, since a damaged block cannot be +recovered. If you wish to minimise any potential data loss through +media or transmission errors, you might consider compressing with a +smaller block size. + + +@unnumberedsubsubsec PERFORMANCE NOTES + +The sorting phase of compression gathers together similar strings in the +file. Because of this, files containing very long runs of repeated +symbols, like "aabaabaabaab ..." (repeated several hundred times) may +compress more slowly than normal. Versions 0.9.5 and above fare much +better than previous versions in this respect. The ratio between +worst-case and average-case compression time is in the region of 10:1. +For previous versions, this figure was more like 100:1. You can use the +@code{-vvvv} option to monitor progress in great detail, if you want. + +Decompression speed is unaffected by these phenomena. + +@code{bzip2} usually allocates several megabytes of memory to operate +in, and then charges all over it in a fairly random fashion. This means +that performance, both for compressing and decompressing, is largely +determined by the speed at which your machine can service cache misses. +Because of this, small changes to the code to reduce the miss rate have +been observed to give disproportionately large performance improvements. +I imagine @code{bzip2} will perform best on machines with very large +caches. + + +@unnumberedsubsubsec CAVEATS + +I/O error messages are not as helpful as they could be. @code{bzip2} +tries hard to detect I/O errors and exit cleanly, but the details of +what the problem is sometimes seem rather misleading. + +This manual page pertains to version 1.0.2 of @code{bzip2}. Compressed +data created by this version is entirely forwards and backwards +compatible with the previous public releases, versions 0.1pl2, 0.9.0, +0.9.5, 1.0.0 and 1.0.1, but with the following exception: 0.9.0 and +above can correctly decompress multiple concatenated compressed files. +0.1pl2 cannot do this; it will stop after decompressing just the first +file in the stream. + +@code{bzip2recover} versions prior to this one, 1.0.2, used 32-bit +integers to represent bit positions in compressed files, so it could not +handle compressed files more than 512 megabytes long. Version 1.0.2 and +above uses 64-bit ints on some platforms which support them (GNU +supported targets, and Windows). To establish whether or not +@code{bzip2recover} was built with such a limitation, run it without +arguments. In any event you can build yourself an unlimited version if +you can recompile it with @code{MaybeUInt64} set to be an unsigned +64-bit integer. + + + +@unnumberedsubsubsec AUTHOR +Julian Seward, @code{jseward@@acm.org}. + +@code{http://sources.redhat.com/bzip2} + +The ideas embodied in @code{bzip2} are due to (at least) the following +people: Michael Burrows and David Wheeler (for the block sorting +transformation), David Wheeler (again, for the Huffman coder), Peter +Fenwick (for the structured coding model in the original @code{bzip}, +and many refinements), and Alistair Moffat, Radford Neal and Ian Witten +(for the arithmetic coder in the original @code{bzip}). I am much +indebted for their help, support and advice. See the manual in the +source distribution for pointers to sources of documentation. Christian +von Roques encouraged me to look for faster sorting algorithms, so as to +speed up compression. Bela Lubkin encouraged me to improve the +worst-case compression performance. The @code{bz*} scripts are derived +from those of GNU @code{gzip}. Many people sent patches, helped with +portability problems, lent machines, gave advice and were generally +helpful. + +@end quotation + + + + +@chapter Programming with @code{libbzip2} + +This chapter describes the programming interface to @code{libbzip2}. + +For general background information, particularly about memory +use and performance aspects, you'd be well advised to read Chapter 2 +as well. + +@section Top-level structure + +@code{libbzip2} is a flexible library for compressing and decompressing +data in the @code{bzip2} data format. Although packaged as a single +entity, it helps to regard the library as three separate parts: the low +level interface, and the high level interface, and some utility +functions. + +The structure of @code{libbzip2}'s interfaces is similar to +that of Jean-loup Gailly's and Mark Adler's excellent @code{zlib} +library. + +All externally visible symbols have names beginning @code{BZ2_}. +This is new in version 1.0. The intention is to minimise pollution +of the namespaces of library clients. + +@subsection Low-level summary + +This interface provides services for compressing and decompressing +data in memory. There's no provision for dealing with files, streams +or any other I/O mechanisms, just straight memory-to-memory work. +In fact, this part of the library can be compiled without inclusion +of @code{stdio.h}, which may be helpful for embedded applications. + +The low-level part of the library has no global variables and +is therefore thread-safe. + +Six routines make up the low level interface: +@code{BZ2_bzCompressInit}, @code{BZ2_bzCompress}, and @* @code{BZ2_bzCompressEnd} +for compression, +and a corresponding trio @code{BZ2_bzDecompressInit}, @* @code{BZ2_bzDecompress} +and @code{BZ2_bzDecompressEnd} for decompression. +The @code{*Init} functions allocate +memory for compression/decompression and do other +initialisations, whilst the @code{*End} functions close down operations +and release memory. + +The real work is done by @code{BZ2_bzCompress} and @code{BZ2_bzDecompress}. +These compress and decompress data from a user-supplied input buffer +to a user-supplied output buffer. These buffers can be any size; +arbitrary quantities of data are handled by making repeated calls +to these functions. This is a flexible mechanism allowing a +consumer-pull style of activity, or producer-push, or a mixture of +both. + + + +@subsection High-level summary + +This interface provides some handy wrappers around the low-level +interface to facilitate reading and writing @code{bzip2} format +files (@code{.bz2} files). The routines provide hooks to facilitate +reading files in which the @code{bzip2} data stream is embedded +within some larger-scale file structure, or where there are +multiple @code{bzip2} data streams concatenated end-to-end. + +For reading files, @code{BZ2_bzReadOpen}, @code{BZ2_bzRead}, +@code{BZ2_bzReadClose} and @* @code{BZ2_bzReadGetUnused} are supplied. For +writing files, @code{BZ2_bzWriteOpen}, @code{BZ2_bzWrite} and +@code{BZ2_bzWriteFinish} are available. + +As with the low-level library, no global variables are used +so the library is per se thread-safe. However, if I/O errors +occur whilst reading or writing the underlying compressed files, +you may have to consult @code{errno} to determine the cause of +the error. In that case, you'd need a C library which correctly +supports @code{errno} in a multithreaded environment. + +To make the library a little simpler and more portable, +@code{BZ2_bzReadOpen} and @code{BZ2_bzWriteOpen} require you to pass them file +handles (@code{FILE*}s) which have previously been opened for reading or +writing respectively. That avoids portability problems associated with +file operations and file attributes, whilst not being much of an +imposition on the programmer. + + + +@subsection Utility functions summary +For very simple needs, @code{BZ2_bzBuffToBuffCompress} and +@code{BZ2_bzBuffToBuffDecompress} are provided. These compress +data in memory from one buffer to another buffer in a single +function call. You should assess whether these functions +fulfill your memory-to-memory compression/decompression +requirements before investing effort in understanding the more +general but more complex low-level interface. + +Yoshioka Tsuneo (@code{QWF00133@@niftyserve.or.jp} / +@code{tsuneo-y@@is.aist-nara.ac.jp}) has contributed some functions to +give better @code{zlib} compatibility. These functions are +@code{BZ2_bzopen}, @code{BZ2_bzread}, @code{BZ2_bzwrite}, @code{BZ2_bzflush}, +@code{BZ2_bzclose}, +@code{BZ2_bzerror} and @code{BZ2_bzlibVersion}. You may find these functions +more convenient for simple file reading and writing, than those in the +high-level interface. These functions are not (yet) officially part of +the library, and are minimally documented here. If they break, you +get to keep all the pieces. I hope to document them properly when time +permits. + +Yoshioka also contributed modifications to allow the library to be +built as a Windows DLL. + + +@section Error handling + +The library is designed to recover cleanly in all situations, including +the worst-case situation of decompressing random data. I'm not +100% sure that it can always do this, so you might want to add +a signal handler to catch segmentation violations during decompression +if you are feeling especially paranoid. I would be interested in +hearing more about the robustness of the library to corrupted +compressed data. + +Version 1.0 is much more robust in this respect than +0.9.0 or 0.9.5. Investigations with Checker (a tool for +detecting problems with memory management, similar to Purify) +indicate that, at least for the few files I tested, all single-bit +errors in the decompressed data are caught properly, with no +segmentation faults, no reads of uninitialised data and no +out of range reads or writes. So it's certainly much improved, +although I wouldn't claim it to be totally bombproof. + +The file @code{bzlib.h} contains all definitions needed to use +the library. In particular, you should definitely not include +@code{bzlib_private.h}. + +In @code{bzlib.h}, the various return values are defined. The following +list is not intended as an exhaustive description of the circumstances +in which a given value may be returned -- those descriptions are given +later. Rather, it is intended to convey the rough meaning of each +return value. The first five actions are normal and not intended to +denote an error situation. +@table @code +@item BZ_OK +The requested action was completed successfully. +@item BZ_RUN_OK +@itemx BZ_FLUSH_OK +@itemx BZ_FINISH_OK +In @code{BZ2_bzCompress}, the requested flush/finish/nothing-special action +was completed successfully. +@item BZ_STREAM_END +Compression of data was completed, or the logical stream end was +detected during decompression. +@end table + +The following return values indicate an error of some kind. +@table @code +@item BZ_CONFIG_ERROR +Indicates that the library has been improperly compiled on your +platform -- a major configuration error. Specifically, it means +that @code{sizeof(char)}, @code{sizeof(short)} and @code{sizeof(int)} +are not 1, 2 and 4 respectively, as they should be. Note that the +library should still work properly on 64-bit platforms which follow +the LP64 programming model -- that is, where @code{sizeof(long)} +and @code{sizeof(void*)} are 8. Under LP64, @code{sizeof(int)} is +still 4, so @code{libbzip2}, which doesn't use the @code{long} type, +is OK. +@item BZ_SEQUENCE_ERROR +When using the library, it is important to call the functions in the +correct sequence and with data structures (buffers etc) in the correct +states. @code{libbzip2} checks as much as it can to ensure this is +happening, and returns @code{BZ_SEQUENCE_ERROR} if not. Code which +complies precisely with the function semantics, as detailed below, +should never receive this value; such an event denotes buggy code +which you should investigate. +@item BZ_PARAM_ERROR +Returned when a parameter to a function call is out of range +or otherwise manifestly incorrect. As with @code{BZ_SEQUENCE_ERROR}, +this denotes a bug in the client code. The distinction between +@code{BZ_PARAM_ERROR} and @code{BZ_SEQUENCE_ERROR} is a bit hazy, but still worth +making. +@item BZ_MEM_ERROR +Returned when a request to allocate memory failed. Note that the +quantity of memory needed to decompress a stream cannot be determined +until the stream's header has been read. So @code{BZ2_bzDecompress} and +@code{BZ2_bzRead} may return @code{BZ_MEM_ERROR} even though some of +the compressed data has been read. The same is not true for +compression; once @code{BZ2_bzCompressInit} or @code{BZ2_bzWriteOpen} have +successfully completed, @code{BZ_MEM_ERROR} cannot occur. +@item BZ_DATA_ERROR +Returned when a data integrity error is detected during decompression. +Most importantly, this means when stored and computed CRCs for the +data do not match. This value is also returned upon detection of any +other anomaly in the compressed data. +@item BZ_DATA_ERROR_MAGIC +As a special case of @code{BZ_DATA_ERROR}, it is sometimes useful to +know when the compressed stream does not start with the correct +magic bytes (@code{'B' 'Z' 'h'}). +@item BZ_IO_ERROR +Returned by @code{BZ2_bzRead} and @code{BZ2_bzWrite} when there is an error +reading or writing in the compressed file, and by @code{BZ2_bzReadOpen} +and @code{BZ2_bzWriteOpen} for attempts to use a file for which the +error indicator (viz, @code{ferror(f)}) is set. +On receipt of @code{BZ_IO_ERROR}, the caller should consult +@code{errno} and/or @code{perror} to acquire operating-system +specific information about the problem. +@item BZ_UNEXPECTED_EOF +Returned by @code{BZ2_bzRead} when the compressed file finishes +before the logical end of stream is detected. +@item BZ_OUTBUFF_FULL +Returned by @code{BZ2_bzBuffToBuffCompress} and +@code{BZ2_bzBuffToBuffDecompress} to indicate that the output data +will not fit into the output buffer provided. +@end table + + + +@section Low-level interface + +@subsection @code{BZ2_bzCompressInit} +@example +typedef + struct @{ + char *next_in; + unsigned int avail_in; + unsigned int total_in_lo32; + unsigned int total_in_hi32; + + char *next_out; + unsigned int avail_out; + unsigned int total_out_lo32; + unsigned int total_out_hi32; + + void *state; + + void *(*bzalloc)(void *,int,int); + void (*bzfree)(void *,void *); + void *opaque; + @} + bz_stream; + +int BZ2_bzCompressInit ( bz_stream *strm, + int blockSize100k, + int verbosity, + int workFactor ); + +@end example + +Prepares for compression. The @code{bz_stream} structure +holds all data pertaining to the compression activity. +A @code{bz_stream} structure should be allocated and initialised +prior to the call. +The fields of @code{bz_stream} +comprise the entirety of the user-visible data. @code{state} +is a pointer to the private data structures required for compression. + +Custom memory allocators are supported, via fields @code{bzalloc}, +@code{bzfree}, +and @code{opaque}. The value +@code{opaque} is passed to as the first argument to +all calls to @code{bzalloc} and @code{bzfree}, but is +otherwise ignored by the library. +The call @code{bzalloc ( opaque, n, m )} is expected to return a +pointer @code{p} to +@code{n * m} bytes of memory, and @code{bzfree ( opaque, p )} +should free +that memory. + +If you don't want to use a custom memory allocator, set @code{bzalloc}, +@code{bzfree} and +@code{opaque} to @code{NULL}, +and the library will then use the standard @code{malloc}/@code{free} +routines. + +Before calling @code{BZ2_bzCompressInit}, fields @code{bzalloc}, +@code{bzfree} and @code{opaque} should +be filled appropriately, as just described. Upon return, the internal +state will have been allocated and initialised, and @code{total_in_lo32}, +@code{total_in_hi32}, @code{total_out_lo32} and +@code{total_out_hi32} will have been set to zero. +These four fields are used by the library +to inform the caller of the total amount of data passed into and out of +the library, respectively. You should not try to change them. +As of version 1.0, 64-bit counts are maintained, even on 32-bit +platforms, using the @code{_hi32} fields to store the upper 32 bits +of the count. So, for example, the total amount of data in +is @code{(total_in_hi32 << 32) + total_in_lo32}. + +Parameter @code{blockSize100k} specifies the block size to be used for +compression. It should be a value between 1 and 9 inclusive, and the +actual block size used is 100000 x this figure. 9 gives the best +compression but takes most memory. + +Parameter @code{verbosity} should be set to a number between 0 and 4 +inclusive. 0 is silent, and greater numbers give increasingly verbose +monitoring/debugging output. If the library has been compiled with +@code{-DBZ_NO_STDIO}, no such output will appear for any verbosity +setting. + +Parameter @code{workFactor} controls how the compression phase behaves +when presented with worst case, highly repetitive, input data. If +compression runs into difficulties caused by repetitive data, the +library switches from the standard sorting algorithm to a fallback +algorithm. The fallback is slower than the standard algorithm by +perhaps a factor of three, but always behaves reasonably, no matter how +bad the input. + +Lower values of @code{workFactor} reduce the amount of effort the +standard algorithm will expend before resorting to the fallback. You +should set this parameter carefully; too low, and many inputs will be +handled by the fallback algorithm and so compress rather slowly, too +high, and your average-to-worst case compression times can become very +large. The default value of 30 gives reasonable behaviour over a wide +range of circumstances. + +Allowable values range from 0 to 250 inclusive. 0 is a special case, +equivalent to using the default value of 30. + +Note that the compressed output generated is the same regardless of +whether or not the fallback algorithm is used. + +Be aware also that this parameter may disappear entirely in future +versions of the library. In principle it should be possible to devise a +good way to automatically choose which algorithm to use. Such a +mechanism would render the parameter obsolete. + +Possible return values: +@display + @code{BZ_CONFIG_ERROR} + if the library has been mis-compiled + @code{BZ_PARAM_ERROR} + if @code{strm} is @code{NULL} + or @code{blockSize} < 1 or @code{blockSize} > 9 + or @code{verbosity} < 0 or @code{verbosity} > 4 + or @code{workFactor} < 0 or @code{workFactor} > 250 + @code{BZ_MEM_ERROR} + if not enough memory is available + @code{BZ_OK} + otherwise +@end display +Allowable next actions: +@display + @code{BZ2_bzCompress} + if @code{BZ_OK} is returned + no specific action needed in case of error +@end display + +@subsection @code{BZ2_bzCompress} +@example + int BZ2_bzCompress ( bz_stream *strm, int action ); +@end example +Provides more input and/or output buffer space for the library. The +caller maintains input and output buffers, and calls @code{BZ2_bzCompress} to +transfer data between them. + +Before each call to @code{BZ2_bzCompress}, @code{next_in} should point at +the data to be compressed, and @code{avail_in} should indicate how many +bytes the library may read. @code{BZ2_bzCompress} updates @code{next_in}, +@code{avail_in} and @code{total_in} to reflect the number of bytes it +has read. + +Similarly, @code{next_out} should point to a buffer in which the +compressed data is to be placed, with @code{avail_out} indicating how +much output space is available. @code{BZ2_bzCompress} updates +@code{next_out}, @code{avail_out} and @code{total_out} to reflect the +number of bytes output. + +You may provide and remove as little or as much data as you like on each +call of @code{BZ2_bzCompress}. In the limit, it is acceptable to supply and +remove data one byte at a time, although this would be terribly +inefficient. You should always ensure that at least one byte of output +space is available at each call. + +A second purpose of @code{BZ2_bzCompress} is to request a change of mode of the +compressed stream. + +Conceptually, a compressed stream can be in one of four states: IDLE, +RUNNING, FLUSHING and FINISHING. Before initialisation +(@code{BZ2_bzCompressInit}) and after termination (@code{BZ2_bzCompressEnd}), a +stream is regarded as IDLE. + +Upon initialisation (@code{BZ2_bzCompressInit}), the stream is placed in the +RUNNING state. Subsequent calls to @code{BZ2_bzCompress} should pass +@code{BZ_RUN} as the requested action; other actions are illegal and +will result in @code{BZ_SEQUENCE_ERROR}. + +At some point, the calling program will have provided all the input data +it wants to. It will then want to finish up -- in effect, asking the +library to process any data it might have buffered internally. In this +state, @code{BZ2_bzCompress} will no longer attempt to read data from +@code{next_in}, but it will want to write data to @code{next_out}. +Because the output buffer supplied by the user can be arbitrarily small, +the finishing-up operation cannot necessarily be done with a single call +of @code{BZ2_bzCompress}. + +Instead, the calling program passes @code{BZ_FINISH} as an action to +@code{BZ2_bzCompress}. This changes the stream's state to FINISHING. Any +remaining input (ie, @code{next_in[0 .. avail_in-1]}) is compressed and +transferred to the output buffer. To do this, @code{BZ2_bzCompress} must be +called repeatedly until all the output has been consumed. At that +point, @code{BZ2_bzCompress} returns @code{BZ_STREAM_END}, and the stream's +state is set back to IDLE. @code{BZ2_bzCompressEnd} should then be +called. + +Just to make sure the calling program does not cheat, the library makes +a note of @code{avail_in} at the time of the first call to +@code{BZ2_bzCompress} which has @code{BZ_FINISH} as an action (ie, at the +time the program has announced its intention to not supply any more +input). By comparing this value with that of @code{avail_in} over +subsequent calls to @code{BZ2_bzCompress}, the library can detect any +attempts to slip in more data to compress. Any calls for which this is +detected will return @code{BZ_SEQUENCE_ERROR}. This indicates a +programming mistake which should be corrected. + +Instead of asking to finish, the calling program may ask +@code{BZ2_bzCompress} to take all the remaining input, compress it and +terminate the current (Burrows-Wheeler) compression block. This could +be useful for error control purposes. The mechanism is analogous to +that for finishing: call @code{BZ2_bzCompress} with an action of +@code{BZ_FLUSH}, remove output data, and persist with the +@code{BZ_FLUSH} action until the value @code{BZ_RUN} is returned. As +with finishing, @code{BZ2_bzCompress} detects any attempt to provide more +input data once the flush has begun. + +Once the flush is complete, the stream returns to the normal RUNNING +state. + +This all sounds pretty complex, but isn't really. Here's a table +which shows which actions are allowable in each state, what action +will be taken, what the next state is, and what the non-error return +values are. Note that you can't explicitly ask what state the +stream is in, but nor do you need to -- it can be inferred from the +values returned by @code{BZ2_bzCompress}. +@display +IDLE/@code{any} + Illegal. IDLE state only exists after @code{BZ2_bzCompressEnd} or + before @code{BZ2_bzCompressInit}. + Return value = @code{BZ_SEQUENCE_ERROR} + +RUNNING/@code{BZ_RUN} + Compress from @code{next_in} to @code{next_out} as much as possible. + Next state = RUNNING + Return value = @code{BZ_RUN_OK} + +RUNNING/@code{BZ_FLUSH} + Remember current value of @code{next_in}. Compress from @code{next_in} + to @code{next_out} as much as possible, but do not accept any more input. + Next state = FLUSHING + Return value = @code{BZ_FLUSH_OK} + +RUNNING/@code{BZ_FINISH} + Remember current value of @code{next_in}. Compress from @code{next_in} + to @code{next_out} as much as possible, but do not accept any more input. + Next state = FINISHING + Return value = @code{BZ_FINISH_OK} + +FLUSHING/@code{BZ_FLUSH} + Compress from @code{next_in} to @code{next_out} as much as possible, + but do not accept any more input. + If all the existing input has been used up and all compressed + output has been removed + Next state = RUNNING; Return value = @code{BZ_RUN_OK} + else + Next state = FLUSHING; Return value = @code{BZ_FLUSH_OK} + +FLUSHING/other + Illegal. + Return value = @code{BZ_SEQUENCE_ERROR} + +FINISHING/@code{BZ_FINISH} + Compress from @code{next_in} to @code{next_out} as much as possible, + but to not accept any more input. + If all the existing input has been used up and all compressed + output has been removed + Next state = IDLE; Return value = @code{BZ_STREAM_END} + else + Next state = FINISHING; Return value = @code{BZ_FINISHING} + +FINISHING/other + Illegal. + Return value = @code{BZ_SEQUENCE_ERROR} +@end display + +That still looks complicated? Well, fair enough. The usual sequence +of calls for compressing a load of data is: +@itemize @bullet +@item Get started with @code{BZ2_bzCompressInit}. +@item Shovel data in and shlurp out its compressed form using zero or more +calls of @code{BZ2_bzCompress} with action = @code{BZ_RUN}. +@item Finish up. +Repeatedly call @code{BZ2_bzCompress} with action = @code{BZ_FINISH}, +copying out the compressed output, until @code{BZ_STREAM_END} is returned. +@item Close up and go home. Call @code{BZ2_bzCompressEnd}. +@end itemize +If the data you want to compress fits into your input buffer all +at once, you can skip the calls of @code{BZ2_bzCompress ( ..., BZ_RUN )} and +just do the @code{BZ2_bzCompress ( ..., BZ_FINISH )} calls. + +All required memory is allocated by @code{BZ2_bzCompressInit}. The +compression library can accept any data at all (obviously). So you +shouldn't get any error return values from the @code{BZ2_bzCompress} calls. +If you do, they will be @code{BZ_SEQUENCE_ERROR}, and indicate a bug in +your programming. + +Trivial other possible return values: +@display + @code{BZ_PARAM_ERROR} + if @code{strm} is @code{NULL}, or @code{strm->s} is @code{NULL} +@end display + +@subsection @code{BZ2_bzCompressEnd} +@example +int BZ2_bzCompressEnd ( bz_stream *strm ); +@end example +Releases all memory associated with a compression stream. + +Possible return values: +@display + @code{BZ_PARAM_ERROR} if @code{strm} is @code{NULL} or @code{strm->s} is @code{NULL} + @code{BZ_OK} otherwise +@end display + + +@subsection @code{BZ2_bzDecompressInit} +@example +int BZ2_bzDecompressInit ( bz_stream *strm, int verbosity, int small ); +@end example +Prepares for decompression. As with @code{BZ2_bzCompressInit}, a +@code{bz_stream} record should be allocated and initialised before the +call. Fields @code{bzalloc}, @code{bzfree} and @code{opaque} should be +set if a custom memory allocator is required, or made @code{NULL} for +the normal @code{malloc}/@code{free} routines. Upon return, the internal +state will have been initialised, and @code{total_in} and +@code{total_out} will be zero. + +For the meaning of parameter @code{verbosity}, see @code{BZ2_bzCompressInit}. + +If @code{small} is nonzero, the library will use an alternative +decompression algorithm which uses less memory but at the cost of +decompressing more slowly (roughly speaking, half the speed, but the +maximum memory requirement drops to around 2300k). See Chapter 2 for +more information on memory management. + +Note that the amount of memory needed to decompress +a stream cannot be determined until the stream's header has been read, +so even if @code{BZ2_bzDecompressInit} succeeds, a subsequent +@code{BZ2_bzDecompress} could fail with @code{BZ_MEM_ERROR}. + +Possible return values: +@display + @code{BZ_CONFIG_ERROR} + if the library has been mis-compiled + @code{BZ_PARAM_ERROR} + if @code{(small != 0 && small != 1)} + or @code{(verbosity < 0 || verbosity > 4)} + @code{BZ_MEM_ERROR} + if insufficient memory is available +@end display + +Allowable next actions: +@display + @code{BZ2_bzDecompress} + if @code{BZ_OK} was returned + no specific action required in case of error +@end display + + + +@subsection @code{BZ2_bzDecompress} +@example +int BZ2_bzDecompress ( bz_stream *strm ); +@end example +Provides more input and/out output buffer space for the library. The +caller maintains input and output buffers, and uses @code{BZ2_bzDecompress} +to transfer data between them. + +Before each call to @code{BZ2_bzDecompress}, @code{next_in} +should point at the compressed data, +and @code{avail_in} should indicate how many bytes the library +may read. @code{BZ2_bzDecompress} updates @code{next_in}, @code{avail_in} +and @code{total_in} +to reflect the number of bytes it has read. + +Similarly, @code{next_out} should point to a buffer in which the uncompressed +output is to be placed, with @code{avail_out} indicating how much output space +is available. @code{BZ2_bzCompress} updates @code{next_out}, +@code{avail_out} and @code{total_out} to reflect +the number of bytes output. + +You may provide and remove as little or as much data as you like on +each call of @code{BZ2_bzDecompress}. +In the limit, it is acceptable to +supply and remove data one byte at a time, although this would be +terribly inefficient. You should always ensure that at least one +byte of output space is available at each call. + +Use of @code{BZ2_bzDecompress} is simpler than @code{BZ2_bzCompress}. + +You should provide input and remove output as described above, and +repeatedly call @code{BZ2_bzDecompress} until @code{BZ_STREAM_END} is +returned. Appearance of @code{BZ_STREAM_END} denotes that +@code{BZ2_bzDecompress} has detected the logical end of the compressed +stream. @code{BZ2_bzDecompress} will not produce @code{BZ_STREAM_END} until +all output data has been placed into the output buffer, so once +@code{BZ_STREAM_END} appears, you are guaranteed to have available all +the decompressed output, and @code{BZ2_bzDecompressEnd} can safely be +called. + +If case of an error return value, you should call @code{BZ2_bzDecompressEnd} +to clean up and release memory. + +Possible return values: +@display + @code{BZ_PARAM_ERROR} + if @code{strm} is @code{NULL} or @code{strm->s} is @code{NULL} + or @code{strm->avail_out < 1} + @code{BZ_DATA_ERROR} + if a data integrity error is detected in the compressed stream + @code{BZ_DATA_ERROR_MAGIC} + if the compressed stream doesn't begin with the right magic bytes + @code{BZ_MEM_ERROR} + if there wasn't enough memory available + @code{BZ_STREAM_END} + if the logical end of the data stream was detected and all + output in has been consumed, eg @code{s->avail_out > 0} + @code{BZ_OK} + otherwise +@end display +Allowable next actions: +@display + @code{BZ2_bzDecompress} + if @code{BZ_OK} was returned + @code{BZ2_bzDecompressEnd} + otherwise +@end display + + +@subsection @code{BZ2_bzDecompressEnd} +@example +int BZ2_bzDecompressEnd ( bz_stream *strm ); +@end example +Releases all memory associated with a decompression stream. + +Possible return values: +@display + @code{BZ_PARAM_ERROR} + if @code{strm} is @code{NULL} or @code{strm->s} is @code{NULL} + @code{BZ_OK} + otherwise +@end display + +Allowable next actions: +@display + None. +@end display + + +@section High-level interface + +This interface provides functions for reading and writing +@code{bzip2} format files. First, some general points. + +@itemize @bullet +@item All of the functions take an @code{int*} first argument, + @code{bzerror}. + After each call, @code{bzerror} should be consulted first to determine + the outcome of the call. If @code{bzerror} is @code{BZ_OK}, + the call completed + successfully, and only then should the return value of the function + (if any) be consulted. If @code{bzerror} is @code{BZ_IO_ERROR}, + there was an error + reading/writing the underlying compressed file, and you should + then consult @code{errno}/@code{perror} to determine the + cause of the difficulty. + @code{bzerror} may also be set to various other values; precise details are + given on a per-function basis below. +@item If @code{bzerror} indicates an error + (ie, anything except @code{BZ_OK} and @code{BZ_STREAM_END}), + you should immediately call @code{BZ2_bzReadClose} (or @code{BZ2_bzWriteClose}, + depending on whether you are attempting to read or to write) + to free up all resources associated + with the stream. Once an error has been indicated, behaviour of all calls + except @code{BZ2_bzReadClose} (@code{BZ2_bzWriteClose}) is undefined. + The implication is that (1) @code{bzerror} should + be checked after each call, and (2) if @code{bzerror} indicates an error, + @code{BZ2_bzReadClose} (@code{BZ2_bzWriteClose}) should then be called to clean up. +@item The @code{FILE*} arguments passed to + @code{BZ2_bzReadOpen}/@code{BZ2_bzWriteOpen} + should be set to binary mode. + Most Unix systems will do this by default, but other platforms, + including Windows and Mac, will not. If you omit this, you may + encounter problems when moving code to new platforms. +@item Memory allocation requests are handled by + @code{malloc}/@code{free}. + At present + there is no facility for user-defined memory allocators in the file I/O + functions (could easily be added, though). +@end itemize + + + +@subsection @code{BZ2_bzReadOpen} +@example + typedef void BZFILE; + + BZFILE *BZ2_bzReadOpen ( int *bzerror, FILE *f, + int small, int verbosity, + void *unused, int nUnused ); +@end example +Prepare to read compressed data from file handle @code{f}. @code{f} +should refer to a file which has been opened for reading, and for which +the error indicator (@code{ferror(f)})is not set. If @code{small} is 1, +the library will try to decompress using less memory, at the expense of +speed. + +For reasons explained below, @code{BZ2_bzRead} will decompress the +@code{nUnused} bytes starting at @code{unused}, before starting to read +from the file @code{f}. At most @code{BZ_MAX_UNUSED} bytes may be +supplied like this. If this facility is not required, you should pass +@code{NULL} and @code{0} for @code{unused} and n@code{Unused} +respectively. + +For the meaning of parameters @code{small} and @code{verbosity}, +see @code{BZ2_bzDecompressInit}. + +The amount of memory needed to decompress a file cannot be determined +until the file's header has been read. So it is possible that +@code{BZ2_bzReadOpen} returns @code{BZ_OK} but a subsequent call of +@code{BZ2_bzRead} will return @code{BZ_MEM_ERROR}. + +Possible assignments to @code{bzerror}: +@display + @code{BZ_CONFIG_ERROR} + if the library has been mis-compiled + @code{BZ_PARAM_ERROR} + if @code{f} is @code{NULL} + or @code{small} is neither @code{0} nor @code{1} + or @code{(unused == NULL && nUnused != 0)} + or @code{(unused != NULL && !(0 <= nUnused <= BZ_MAX_UNUSED))} + @code{BZ_IO_ERROR} + if @code{ferror(f)} is nonzero + @code{BZ_MEM_ERROR} + if insufficient memory is available + @code{BZ_OK} + otherwise. +@end display + +Possible return values: +@display + Pointer to an abstract @code{BZFILE} + if @code{bzerror} is @code{BZ_OK} + @code{NULL} + otherwise +@end display + +Allowable next actions: +@display + @code{BZ2_bzRead} + if @code{bzerror} is @code{BZ_OK} + @code{BZ2_bzClose} + otherwise +@end display + + +@subsection @code{BZ2_bzRead} +@example + int BZ2_bzRead ( int *bzerror, BZFILE *b, void *buf, int len ); +@end example +Reads up to @code{len} (uncompressed) bytes from the compressed file +@code{b} into +the buffer @code{buf}. If the read was successful, +@code{bzerror} is set to @code{BZ_OK} +and the number of bytes read is returned. If the logical end-of-stream +was detected, @code{bzerror} will be set to @code{BZ_STREAM_END}, +and the number +of bytes read is returned. All other @code{bzerror} values denote an error. + +@code{BZ2_bzRead} will supply @code{len} bytes, +unless the logical stream end is detected +or an error occurs. Because of this, it is possible to detect the +stream end by observing when the number of bytes returned is +less than the number +requested. Nevertheless, this is regarded as inadvisable; you should +instead check @code{bzerror} after every call and watch out for +@code{BZ_STREAM_END}. + +Internally, @code{BZ2_bzRead} copies data from the compressed file in chunks +of size @code{BZ_MAX_UNUSED} bytes +before decompressing it. If the file contains more bytes than strictly +needed to reach the logical end-of-stream, @code{BZ2_bzRead} will almost certainly +read some of the trailing data before signalling @code{BZ_SEQUENCE_END}. +To collect the read but unused data once @code{BZ_SEQUENCE_END} has +appeared, call @code{BZ2_bzReadGetUnused} immediately before @code{BZ2_bzReadClose}. + +Possible assignments to @code{bzerror}: +@display + @code{BZ_PARAM_ERROR} + if @code{b} is @code{NULL} or @code{buf} is @code{NULL} or @code{len < 0} + @code{BZ_SEQUENCE_ERROR} + if @code{b} was opened with @code{BZ2_bzWriteOpen} + @code{BZ_IO_ERROR} + if there is an error reading from the compressed file + @code{BZ_UNEXPECTED_EOF} + if the compressed file ended before the logical end-of-stream was detected + @code{BZ_DATA_ERROR} + if a data integrity error was detected in the compressed stream + @code{BZ_DATA_ERROR_MAGIC} + if the stream does not begin with the requisite header bytes (ie, is not + a @code{bzip2} data file). This is really a special case of @code{BZ_DATA_ERROR}. + @code{BZ_MEM_ERROR} + if insufficient memory was available + @code{BZ_STREAM_END} + if the logical end of stream was detected. + @code{BZ_OK} + otherwise. +@end display + +Possible return values: +@display + number of bytes read + if @code{bzerror} is @code{BZ_OK} or @code{BZ_STREAM_END} + undefined + otherwise +@end display + +Allowable next actions: +@display + collect data from @code{buf}, then @code{BZ2_bzRead} or @code{BZ2_bzReadClose} + if @code{bzerror} is @code{BZ_OK} + collect data from @code{buf}, then @code{BZ2_bzReadClose} or @code{BZ2_bzReadGetUnused} + if @code{bzerror} is @code{BZ_SEQUENCE_END} + @code{BZ2_bzReadClose} + otherwise +@end display + + + +@subsection @code{BZ2_bzReadGetUnused} +@example + void BZ2_bzReadGetUnused ( int* bzerror, BZFILE *b, + void** unused, int* nUnused ); +@end example +Returns data which was read from the compressed file but was not needed +to get to the logical end-of-stream. @code{*unused} is set to the address +of the data, and @code{*nUnused} to the number of bytes. @code{*nUnused} will +be set to a value between @code{0} and @code{BZ_MAX_UNUSED} inclusive. + +This function may only be called once @code{BZ2_bzRead} has signalled +@code{BZ_STREAM_END} but before @code{BZ2_bzReadClose}. + +Possible assignments to @code{bzerror}: +@display + @code{BZ_PARAM_ERROR} + if @code{b} is @code{NULL} + or @code{unused} is @code{NULL} or @code{nUnused} is @code{NULL} + @code{BZ_SEQUENCE_ERROR} + if @code{BZ_STREAM_END} has not been signalled + or if @code{b} was opened with @code{BZ2_bzWriteOpen} + @code{BZ_OK} + otherwise +@end display + +Allowable next actions: +@display + @code{BZ2_bzReadClose} +@end display + + +@subsection @code{BZ2_bzReadClose} +@example + void BZ2_bzReadClose ( int *bzerror, BZFILE *b ); +@end example +Releases all memory pertaining to the compressed file @code{b}. +@code{BZ2_bzReadClose} does not call @code{fclose} on the underlying file +handle, so you should do that yourself if appropriate. +@code{BZ2_bzReadClose} should be called to clean up after all error +situations. + +Possible assignments to @code{bzerror}: +@display + @code{BZ_SEQUENCE_ERROR} + if @code{b} was opened with @code{BZ2_bzOpenWrite} + @code{BZ_OK} + otherwise +@end display + +Allowable next actions: +@display + none +@end display + + + +@subsection @code{BZ2_bzWriteOpen} +@example + BZFILE *BZ2_bzWriteOpen ( int *bzerror, FILE *f, + int blockSize100k, int verbosity, + int workFactor ); +@end example +Prepare to write compressed data to file handle @code{f}. +@code{f} should refer to +a file which has been opened for writing, and for which the error +indicator (@code{ferror(f)})is not set. + +For the meaning of parameters @code{blockSize100k}, +@code{verbosity} and @code{workFactor}, see +@* @code{BZ2_bzCompressInit}. + +All required memory is allocated at this stage, so if the call +completes successfully, @code{BZ_MEM_ERROR} cannot be signalled by a +subsequent call to @code{BZ2_bzWrite}. + +Possible assignments to @code{bzerror}: +@display + @code{BZ_CONFIG_ERROR} + if the library has been mis-compiled + @code{BZ_PARAM_ERROR} + if @code{f} is @code{NULL} + or @code{blockSize100k < 1} or @code{blockSize100k > 9} + @code{BZ_IO_ERROR} + if @code{ferror(f)} is nonzero + @code{BZ_MEM_ERROR} + if insufficient memory is available + @code{BZ_OK} + otherwise +@end display + +Possible return values: +@display + Pointer to an abstract @code{BZFILE} + if @code{bzerror} is @code{BZ_OK} + @code{NULL} + otherwise +@end display + +Allowable next actions: +@display + @code{BZ2_bzWrite} + if @code{bzerror} is @code{BZ_OK} + (you could go directly to @code{BZ2_bzWriteClose}, but this would be pretty pointless) + @code{BZ2_bzWriteClose} + otherwise +@end display + + + +@subsection @code{BZ2_bzWrite} +@example + void BZ2_bzWrite ( int *bzerror, BZFILE *b, void *buf, int len ); +@end example +Absorbs @code{len} bytes from the buffer @code{buf}, eventually to be +compressed and written to the file. + +Possible assignments to @code{bzerror}: +@display + @code{BZ_PARAM_ERROR} + if @code{b} is @code{NULL} or @code{buf} is @code{NULL} or @code{len < 0} + @code{BZ_SEQUENCE_ERROR} + if b was opened with @code{BZ2_bzReadOpen} + @code{BZ_IO_ERROR} + if there is an error writing the compressed file. + @code{BZ_OK} + otherwise +@end display + + + + +@subsection @code{BZ2_bzWriteClose} +@example + void BZ2_bzWriteClose ( int *bzerror, BZFILE* f, + int abandon, + unsigned int* nbytes_in, + unsigned int* nbytes_out ); + + void BZ2_bzWriteClose64 ( int *bzerror, BZFILE* f, + int abandon, + unsigned int* nbytes_in_lo32, + unsigned int* nbytes_in_hi32, + unsigned int* nbytes_out_lo32, + unsigned int* nbytes_out_hi32 ); +@end example + +Compresses and flushes to the compressed file all data so far supplied +by @code{BZ2_bzWrite}. The logical end-of-stream markers are also written, so +subsequent calls to @code{BZ2_bzWrite} are illegal. All memory associated +with the compressed file @code{b} is released. +@code{fflush} is called on the +compressed file, but it is not @code{fclose}'d. + +If @code{BZ2_bzWriteClose} is called to clean up after an error, the only +action is to release the memory. The library records the error codes +issued by previous calls, so this situation will be detected +automatically. There is no attempt to complete the compression +operation, nor to @code{fflush} the compressed file. You can force this +behaviour to happen even in the case of no error, by passing a nonzero +value to @code{abandon}. + +If @code{nbytes_in} is non-null, @code{*nbytes_in} will be set to be the +total volume of uncompressed data handled. Similarly, @code{nbytes_out} +will be set to the total volume of compressed data written. For +compatibility with older versions of the library, @code{BZ2_bzWriteClose} +only yields the lower 32 bits of these counts. Use +@code{BZ2_bzWriteClose64} if you want the full 64 bit counts. These +two functions are otherwise absolutely identical. + + +Possible assignments to @code{bzerror}: +@display + @code{BZ_SEQUENCE_ERROR} + if @code{b} was opened with @code{BZ2_bzReadOpen} + @code{BZ_IO_ERROR} + if there is an error writing the compressed file + @code{BZ_OK} + otherwise +@end display + +@subsection Handling embedded compressed data streams + +The high-level library facilitates use of +@code{bzip2} data streams which form some part of a surrounding, larger +data stream. +@itemize @bullet +@item For writing, the library takes an open file handle, writes +compressed data to it, @code{fflush}es it but does not @code{fclose} it. +The calling application can write its own data before and after the +compressed data stream, using that same file handle. +@item Reading is more complex, and the facilities are not as general +as they could be since generality is hard to reconcile with efficiency. +@code{BZ2_bzRead} reads from the compressed file in blocks of size +@code{BZ_MAX_UNUSED} bytes, and in doing so probably will overshoot +the logical end of compressed stream. +To recover this data once decompression has +ended, call @code{BZ2_bzReadGetUnused} after the last call of @code{BZ2_bzRead} +(the one returning @code{BZ_STREAM_END}) but before calling +@code{BZ2_bzReadClose}. +@end itemize + +This mechanism makes it easy to decompress multiple @code{bzip2} +streams placed end-to-end. As the end of one stream, when @code{BZ2_bzRead} +returns @code{BZ_STREAM_END}, call @code{BZ2_bzReadGetUnused} to collect the +unused data (copy it into your own buffer somewhere). +That data forms the start of the next compressed stream. +To start uncompressing that next stream, call @code{BZ2_bzReadOpen} again, +feeding in the unused data via the @code{unused}/@code{nUnused} +parameters. +Keep doing this until @code{BZ_STREAM_END} return coincides with the +physical end of file (@code{feof(f)}). In this situation +@code{BZ2_bzReadGetUnused} +will of course return no data. + +This should give some feel for how the high-level interface can be used. +If you require extra flexibility, you'll have to bite the bullet and get +to grips with the low-level interface. + +@subsection Standard file-reading/writing code +Here's how you'd write data to a compressed file: +@example @code +FILE* f; +BZFILE* b; +int nBuf; +char buf[ /* whatever size you like */ ]; +int bzerror; +int nWritten; + +f = fopen ( "myfile.bz2", "w" ); +if (!f) @{ + /* handle error */ +@} +b = BZ2_bzWriteOpen ( &bzerror, f, 9 ); +if (bzerror != BZ_OK) @{ + BZ2_bzWriteClose ( b ); + /* handle error */ +@} + +while ( /* condition */ ) @{ + /* get data to write into buf, and set nBuf appropriately */ + nWritten = BZ2_bzWrite ( &bzerror, b, buf, nBuf ); + if (bzerror == BZ_IO_ERROR) @{ + BZ2_bzWriteClose ( &bzerror, b ); + /* handle error */ + @} +@} + +BZ2_bzWriteClose ( &bzerror, b ); +if (bzerror == BZ_IO_ERROR) @{ + /* handle error */ +@} +@end example +And to read from a compressed file: +@example +FILE* f; +BZFILE* b; +int nBuf; +char buf[ /* whatever size you like */ ]; +int bzerror; +int nWritten; + +f = fopen ( "myfile.bz2", "r" ); +if (!f) @{ + /* handle error */ +@} +b = BZ2_bzReadOpen ( &bzerror, f, 0, NULL, 0 ); +if (bzerror != BZ_OK) @{ + BZ2_bzReadClose ( &bzerror, b ); + /* handle error */ +@} + +bzerror = BZ_OK; +while (bzerror == BZ_OK && /* arbitrary other conditions */) @{ + nBuf = BZ2_bzRead ( &bzerror, b, buf, /* size of buf */ ); + if (bzerror == BZ_OK) @{ + /* do something with buf[0 .. nBuf-1] */ + @} +@} +if (bzerror != BZ_STREAM_END) @{ + BZ2_bzReadClose ( &bzerror, b ); + /* handle error */ +@} else @{ + BZ2_bzReadClose ( &bzerror ); +@} +@end example + + + +@section Utility functions +@subsection @code{BZ2_bzBuffToBuffCompress} +@example + int BZ2_bzBuffToBuffCompress( char* dest, + unsigned int* destLen, + char* source, + unsigned int sourceLen, + int blockSize100k, + int verbosity, + int workFactor ); +@end example +Attempts to compress the data in @code{source[0 .. sourceLen-1]} +into the destination buffer, @code{dest[0 .. *destLen-1]}. +If the destination buffer is big enough, @code{*destLen} is +set to the size of the compressed data, and @code{BZ_OK} is +returned. If the compressed data won't fit, @code{*destLen} +is unchanged, and @code{BZ_OUTBUFF_FULL} is returned. + +Compression in this manner is a one-shot event, done with a single call +to this function. The resulting compressed data is a complete +@code{bzip2} format data stream. There is no mechanism for making +additional calls to provide extra input data. If you want that kind of +mechanism, use the low-level interface. + +For the meaning of parameters @code{blockSize100k}, @code{verbosity} +and @code{workFactor}, @* see @code{BZ2_bzCompressInit}. + +To guarantee that the compressed data will fit in its buffer, allocate +an output buffer of size 1% larger than the uncompressed data, plus +six hundred extra bytes. + +@code{BZ2_bzBuffToBuffDecompress} will not write data at or +beyond @code{dest[*destLen]}, even in case of buffer overflow. + +Possible return values: +@display + @code{BZ_CONFIG_ERROR} + if the library has been mis-compiled + @code{BZ_PARAM_ERROR} + if @code{dest} is @code{NULL} or @code{destLen} is @code{NULL} + or @code{blockSize100k < 1} or @code{blockSize100k > 9} + or @code{verbosity < 0} or @code{verbosity > 4} + or @code{workFactor < 0} or @code{workFactor > 250} + @code{BZ_MEM_ERROR} + if insufficient memory is available + @code{BZ_OUTBUFF_FULL} + if the size of the compressed data exceeds @code{*destLen} + @code{BZ_OK} + otherwise +@end display + + + +@subsection @code{BZ2_bzBuffToBuffDecompress} +@example + int BZ2_bzBuffToBuffDecompress ( char* dest, + unsigned int* destLen, + char* source, + unsigned int sourceLen, + int small, + int verbosity ); +@end example +Attempts to decompress the data in @code{source[0 .. sourceLen-1]} +into the destination buffer, @code{dest[0 .. *destLen-1]}. +If the destination buffer is big enough, @code{*destLen} is +set to the size of the uncompressed data, and @code{BZ_OK} is +returned. If the compressed data won't fit, @code{*destLen} +is unchanged, and @code{BZ_OUTBUFF_FULL} is returned. + +@code{source} is assumed to hold a complete @code{bzip2} format +data stream. @* @code{BZ2_bzBuffToBuffDecompress} tries to decompress +the entirety of the stream into the output buffer. + +For the meaning of parameters @code{small} and @code{verbosity}, +see @code{BZ2_bzDecompressInit}. + +Because the compression ratio of the compressed data cannot be known in +advance, there is no easy way to guarantee that the output buffer will +be big enough. You may of course make arrangements in your code to +record the size of the uncompressed data, but such a mechanism is beyond +the scope of this library. + +@code{BZ2_bzBuffToBuffDecompress} will not write data at or +beyond @code{dest[*destLen]}, even in case of buffer overflow. + +Possible return values: +@display + @code{BZ_CONFIG_ERROR} + if the library has been mis-compiled + @code{BZ_PARAM_ERROR} + if @code{dest} is @code{NULL} or @code{destLen} is @code{NULL} + or @code{small != 0 && small != 1} + or @code{verbosity < 0} or @code{verbosity > 4} + @code{BZ_MEM_ERROR} + if insufficient memory is available + @code{BZ_OUTBUFF_FULL} + if the size of the compressed data exceeds @code{*destLen} + @code{BZ_DATA_ERROR} + if a data integrity error was detected in the compressed data + @code{BZ_DATA_ERROR_MAGIC} + if the compressed data doesn't begin with the right magic bytes + @code{BZ_UNEXPECTED_EOF} + if the compressed data ends unexpectedly + @code{BZ_OK} + otherwise +@end display + + + +@section @code{zlib} compatibility functions +Yoshioka Tsuneo has contributed some functions to +give better @code{zlib} compatibility. These functions are +@code{BZ2_bzopen}, @code{BZ2_bzread}, @code{BZ2_bzwrite}, @code{BZ2_bzflush}, +@code{BZ2_bzclose}, +@code{BZ2_bzerror} and @code{BZ2_bzlibVersion}. +These functions are not (yet) officially part of +the library. If they break, you get to keep all the pieces. +Nevertheless, I think they work ok. +@example +typedef void BZFILE; + +const char * BZ2_bzlibVersion ( void ); +@end example +Returns a string indicating the library version. +@example +BZFILE * BZ2_bzopen ( const char *path, const char *mode ); +BZFILE * BZ2_bzdopen ( int fd, const char *mode ); +@end example +Opens a @code{.bz2} file for reading or writing, using either its name +or a pre-existing file descriptor. +Analogous to @code{fopen} and @code{fdopen}. +@example +int BZ2_bzread ( BZFILE* b, void* buf, int len ); +int BZ2_bzwrite ( BZFILE* b, void* buf, int len ); +@end example +Reads/writes data from/to a previously opened @code{BZFILE}. +Analogous to @code{fread} and @code{fwrite}. +@example +int BZ2_bzflush ( BZFILE* b ); +void BZ2_bzclose ( BZFILE* b ); +@end example +Flushes/closes a @code{BZFILE}. @code{BZ2_bzflush} doesn't actually do +anything. Analogous to @code{fflush} and @code{fclose}. + +@example +const char * BZ2_bzerror ( BZFILE *b, int *errnum ) +@end example +Returns a string describing the more recent error status of +@code{b}, and also sets @code{*errnum} to its numerical value. + + +@section Using the library in a @code{stdio}-free environment + +@subsection Getting rid of @code{stdio} + +In a deeply embedded application, you might want to use just +the memory-to-memory functions. You can do this conveniently +by compiling the library with preprocessor symbol @code{BZ_NO_STDIO} +defined. Doing this gives you a library containing only the following +eight functions: + +@code{BZ2_bzCompressInit}, @code{BZ2_bzCompress}, @code{BZ2_bzCompressEnd} @* +@code{BZ2_bzDecompressInit}, @code{BZ2_bzDecompress}, @code{BZ2_bzDecompressEnd} @* +@code{BZ2_bzBuffToBuffCompress}, @code{BZ2_bzBuffToBuffDecompress} + +When compiled like this, all functions will ignore @code{verbosity} +settings. + +@subsection Critical error handling +@code{libbzip2} contains a number of internal assertion checks which +should, needless to say, never be activated. Nevertheless, if an +assertion should fail, behaviour depends on whether or not the library +was compiled with @code{BZ_NO_STDIO} set. + +For a normal compile, an assertion failure yields the message +@example + bzip2/libbzip2: internal error number N. + This is a bug in bzip2/libbzip2, 1.0.2, 30-Dec-2001. + Please report it to me at: jseward@@acm.org. If this happened + when you were using some program which uses libbzip2 as a + component, you should also report this bug to the author(s) + of that program. Please make an effort to report this bug; + timely and accurate bug reports eventually lead to higher + quality software. Thanks. Julian Seward, 30 December 2001. +@end example +where @code{N} is some error code number. If @code{N == 1007}, it also +prints some extra text advising the reader that unreliable memory is +often associated with internal error 1007. (This is a +frequently-observed-phenomenon with versions 1.0.0/1.0.1). + +@code{exit(3)} is then called. + +For a @code{stdio}-free library, assertion failures result +in a call to a function declared as: +@example + extern void bz_internal_error ( int errcode ); +@end example +The relevant code is passed as a parameter. You should supply +such a function. + +In either case, once an assertion failure has occurred, any +@code{bz_stream} records involved can be regarded as invalid. +You should not attempt to resume normal operation with them. + +You may, of course, change critical error handling to suit +your needs. As I said above, critical errors indicate bugs +in the library and should not occur. All "normal" error +situations are indicated via error return codes from functions, +and can be recovered from. + + +@section Making a Windows DLL +Everything related to Windows has been contributed by Yoshioka Tsuneo +@* (@code{QWF00133@@niftyserve.or.jp} / +@code{tsuneo-y@@is.aist-nara.ac.jp}), so you should send your queries to +him (but perhaps Cc: me, @code{jseward@@acm.org}). + +My vague understanding of what to do is: using Visual C++ 5.0, +open the project file @code{libbz2.dsp}, and build. That's all. + +If you can't +open the project file for some reason, make a new one, naming these files: +@code{blocksort.c}, @code{bzlib.c}, @code{compress.c}, +@code{crctable.c}, @code{decompress.c}, @code{huffman.c}, @* +@code{randtable.c} and @code{libbz2.def}. You will also need +to name the header files @code{bzlib.h} and @code{bzlib_private.h}. + +If you don't use VC++, you may need to define the proprocessor symbol +@code{_WIN32}. + +Finally, @code{dlltest.c} is a sample program using the DLL. It has a +project file, @code{dlltest.dsp}. + +If you just want a makefile for Visual C, have a look at +@code{makefile.msc}. + +Be aware that if you compile @code{bzip2} itself on Win32, you must set +@code{BZ_UNIX} to 0 and @code{BZ_LCCWIN32} to 1, in the file +@code{bzip2.c}, before compiling. Otherwise the resulting binary won't +work correctly. + +I haven't tried any of this stuff myself, but it all looks plausible. + + + +@chapter Miscellanea + +These are just some random thoughts of mine. Your mileage may +vary. + +@section Limitations of the compressed file format +@code{bzip2-1.0}, @code{0.9.5} and @code{0.9.0} +use exactly the same file format as the previous +version, @code{bzip2-0.1}. This decision was made in the interests of +stability. Creating yet another incompatible compressed file format +would create further confusion and disruption for users. + +Nevertheless, this is not a painless decision. Development +work since the release of @code{bzip2-0.1} in August 1997 +has shown complexities in the file format which slow down +decompression and, in retrospect, are unnecessary. These are: +@itemize @bullet +@item The run-length encoder, which is the first of the + compression transformations, is entirely irrelevant. + The original purpose was to protect the sorting algorithm + from the very worst case input: a string of repeated + symbols. But algorithm steps Q6a and Q6b in the original + Burrows-Wheeler technical report (SRC-124) show how + repeats can be handled without difficulty in block + sorting. +@item The randomisation mechanism doesn't really need to be + there. Udi Manber and Gene Myers published a suffix + array construction algorithm a few years back, which + can be employed to sort any block, no matter how + repetitive, in O(N log N) time. Subsequent work by + Kunihiko Sadakane has produced a derivative O(N (log N)^2) + algorithm which usually outperforms the Manber-Myers + algorithm. + + I could have changed to Sadakane's algorithm, but I find + it to be slower than @code{bzip2}'s existing algorithm for + most inputs, and the randomisation mechanism protects + adequately against bad cases. I didn't think it was + a good tradeoff to make. Partly this is due to the fact + that I was not flooded with email complaints about + @code{bzip2-0.1}'s performance on repetitive data, so + perhaps it isn't a problem for real inputs. + + Probably the best long-term solution, + and the one I have incorporated into 0.9.5 and above, + is to use the existing sorting + algorithm initially, and fall back to a O(N (log N)^2) + algorithm if the standard algorithm gets into difficulties. +@item The compressed file format was never designed to be + handled by a library, and I have had to jump though + some hoops to produce an efficient implementation of + decompression. It's a bit hairy. Try passing + @code{decompress.c} through the C preprocessor + and you'll see what I mean. Much of this complexity + could have been avoided if the compressed size of + each block of data was recorded in the data stream. +@item An Adler-32 checksum, rather than a CRC32 checksum, + would be faster to compute. +@end itemize +It would be fair to say that the @code{bzip2} format was frozen +before I properly and fully understood the performance +consequences of doing so. + +Improvements which I was able to incorporate into +0.9.0, despite using the same file format, are: +@itemize @bullet +@item Single array implementation of the inverse BWT. This + significantly speeds up decompression, presumably + because it reduces the number of cache misses. +@item Faster inverse MTF transform for large MTF values. The + new implementation is based on the notion of sliding blocks + of values. +@item @code{bzip2-0.9.0} now reads and writes files with @code{fread} + and @code{fwrite}; version 0.1 used @code{putc} and @code{getc}. + Duh! Well, you live and learn. + +@end itemize +Further ahead, it would be nice +to be able to do random access into files. This will +require some careful design of compressed file formats. + + + +@section Portability issues +After some consideration, I have decided not to use +GNU @code{autoconf} to configure 0.9.5 or 1.0. + +@code{autoconf}, admirable and wonderful though it is, +mainly assists with portability problems between Unix-like +platforms. But @code{bzip2} doesn't have much in the way +of portability problems on Unix; most of the difficulties appear +when porting to the Mac, or to Microsoft's operating systems. +@code{autoconf} doesn't help in those cases, and brings in a +whole load of new complexity. + +Most people should be able to compile the library and program +under Unix straight out-of-the-box, so to speak, especially +if you have a version of GNU C available. + +There are a couple of @code{__inline__} directives in the code. GNU C +(@code{gcc}) should be able to handle them. If you're not using +GNU C, your C compiler shouldn't see them at all. +If your compiler does, for some reason, see them and doesn't +like them, just @code{#define} @code{__inline__} to be @code{/* */}. One +easy way to do this is to compile with the flag @code{-D__inline__=}, +which should be understood by most Unix compilers. + +If you still have difficulties, try compiling with the macro +@code{BZ_STRICT_ANSI} defined. This should enable you to build the +library in a strictly ANSI compliant environment. Building the program +itself like this is dangerous and not supported, since you remove +@code{bzip2}'s checks against compressing directories, symbolic links, +devices, and other not-really-a-file entities. This could cause +filesystem corruption! + +One other thing: if you create a @code{bzip2} binary for public +distribution, please try and link it statically (@code{gcc -s}). This +avoids all sorts of library-version issues that others may encounter +later on. + +If you build @code{bzip2} on Win32, you must set @code{BZ_UNIX} to 0 and +@code{BZ_LCCWIN32} to 1, in the file @code{bzip2.c}, before compiling. +Otherwise the resulting binary won't work correctly. + + + +@section Reporting bugs +I tried pretty hard to make sure @code{bzip2} is +bug free, both by design and by testing. Hopefully +you'll never need to read this section for real. + +Nevertheless, if @code{bzip2} dies with a segmentation +fault, a bus error or an internal assertion failure, it +will ask you to email me a bug report. Experience with +version 0.1 shows that almost all these problems can +be traced to either compiler bugs or hardware problems. +@itemize @bullet +@item +Recompile the program with no optimisation, and see if it +works. And/or try a different compiler. +I heard all sorts of stories about various flavours +of GNU C (and other compilers) generating bad code for +@code{bzip2}, and I've run across two such examples myself. + +2.7.X versions of GNU C are known to generate bad code from +time to time, at high optimisation levels. +If you get problems, try using the flags +@code{-O2} @code{-fomit-frame-pointer} @code{-fno-strength-reduce}. +You should specifically @emph{not} use @code{-funroll-loops}. + +You may notice that the Makefile runs six tests as part of +the build process. If the program passes all of these, it's +a pretty good (but not 100%) indication that the compiler has +done its job correctly. +@item +If @code{bzip2} crashes randomly, and the crashes are not +repeatable, you may have a flaky memory subsystem. @code{bzip2} +really hammers your memory hierarchy, and if it's a bit marginal, +you may get these problems. Ditto if your disk or I/O subsystem +is slowly failing. Yup, this really does happen. + +Try using a different machine of the same type, and see if +you can repeat the problem. +@item This isn't really a bug, but ... If @code{bzip2} tells +you your file is corrupted on decompression, and you +obtained the file via FTP, there is a possibility that you +forgot to tell FTP to do a binary mode transfer. That absolutely +will cause the file to be non-decompressible. You'll have to transfer +it again. +@end itemize + +If you've incorporated @code{libbzip2} into your own program +and are getting problems, please, please, please, check that the +parameters you are passing in calls to the library, are +correct, and in accordance with what the documentation says +is allowable. I have tried to make the library robust against +such problems, but I'm sure I haven't succeeded. + +Finally, if the above comments don't help, you'll have to send +me a bug report. Now, it's just amazing how many people will +send me a bug report saying something like +@display + bzip2 crashed with segmentation fault on my machine +@end display +and absolutely nothing else. Needless to say, a such a report +is @emph{totally, utterly, completely and comprehensively 100% useless; +a waste of your time, my time, and net bandwidth}. +With no details at all, there's no way I can possibly begin +to figure out what the problem is. + +The rules of the game are: facts, facts, facts. Don't omit +them because "oh, they won't be relevant". At the bare +minimum: +@display + Machine type. Operating system version. + Exact version of @code{bzip2} (do @code{bzip2 -V}). + Exact version of the compiler used. + Flags passed to the compiler. +@end display +However, the most important single thing that will help me is +the file that you were trying to compress or decompress at the +time the problem happened. Without that, my ability to do anything +more than speculate about the cause, is limited. + +Please remember that I connect to the Internet with a modem, so +you should contact me before mailing me huge files. + + +@section Did you get the right package? + +@code{bzip2} is a resource hog. It soaks up large amounts of CPU cycles +and memory. Also, it gives very large latencies. In the worst case, you +can feed many megabytes of uncompressed data into the library before +getting any compressed output, so this probably rules out applications +requiring interactive behaviour. + +These aren't faults of my implementation, I hope, but more +an intrinsic property of the Burrows-Wheeler transform (unfortunately). +Maybe this isn't what you want. + +If you want a compressor and/or library which is faster, uses less +memory but gets pretty good compression, and has minimal latency, +consider Jean-loup +Gailly's and Mark Adler's work, @code{zlib-1.1.3} and +@code{gzip-1.2.4}. Look for them at + +@code{http://www.zlib.org} and +@code{http://www.gzip.org} respectively. + +For something faster and lighter still, you might try Markus F X J +Oberhumer's @code{LZO} real-time compression/decompression library, at +@* @code{http://wildsau.idv.uni-linz.ac.at/mfx/lzo.html}. + +If you want to use the @code{bzip2} algorithms to compress small blocks +of data, 64k bytes or smaller, for example on an on-the-fly disk +compressor, you'd be well advised not to use this library. Instead, +I've made a special library tuned for that kind of use. It's part of +@code{e2compr-0.40}, an on-the-fly disk compressor for the Linux +@code{ext2} filesystem. Look at +@code{http://www.netspace.net.au/~reiter/e2compr}. + + + +@section Testing + +A record of the tests I've done. + +First, some data sets: +@itemize @bullet +@item B: a directory containing 6001 files, one for every length in the + range 0 to 6000 bytes. The files contain random lowercase + letters. 18.7 megabytes. +@item H: my home directory tree. Documents, source code, mail files, + compressed data. H contains B, and also a directory of + files designed as boundary cases for the sorting; mostly very + repetitive, nasty files. 565 megabytes. +@item A: directory tree holding various applications built from source: + @code{egcs}, @code{gcc-2.8.1}, KDE, GTK, Octave, etc. + 2200 megabytes. +@end itemize +The tests conducted are as follows. Each test means compressing +(a copy of) each file in the data set, decompressing it and +comparing it against the original. + +First, a bunch of tests with block sizes and internal buffer +sizes set very small, +to detect any problems with the +blocking and buffering mechanisms. +This required modifying the source code so as to try to +break it. +@enumerate +@item Data set H, with + buffer size of 1 byte, and block size of 23 bytes. +@item Data set B, buffer sizes 1 byte, block size 1 byte. +@item As (2) but small-mode decompression. +@item As (2) with block size 2 bytes. +@item As (2) with block size 3 bytes. +@item As (2) with block size 4 bytes. +@item As (2) with block size 5 bytes. +@item As (2) with block size 6 bytes and small-mode decompression. +@item H with buffer size of 1 byte, but normal block + size (up to 900000 bytes). +@end enumerate +Then some tests with unmodified source code. +@enumerate +@item H, all settings normal. +@item As (1), with small-mode decompress. +@item H, compress with flag @code{-1}. +@item H, compress with flag @code{-s}, decompress with flag @code{-s}. +@item Forwards compatibility: H, @code{bzip2-0.1pl2} compressing, + @code{bzip2-0.9.5} decompressing, all settings normal. +@item Backwards compatibility: H, @code{bzip2-0.9.5} compressing, + @code{bzip2-0.1pl2} decompressing, all settings normal. +@item Bigger tests: A, all settings normal. +@item As (7), using the fallback (Sadakane-like) sorting algorithm. +@item As (8), compress with flag @code{-1}, decompress with flag + @code{-s}. +@item H, using the fallback sorting algorithm. +@item Forwards compatibility: A, @code{bzip2-0.1pl2} compressing, + @code{bzip2-0.9.5} decompressing, all settings normal. +@item Backwards compatibility: A, @code{bzip2-0.9.5} compressing, + @code{bzip2-0.1pl2} decompressing, all settings normal. +@item Misc test: about 400 megabytes of @code{.tar} files with + @code{bzip2} compiled with Checker (a memory access error + detector, like Purify). +@item Misc tests to make sure it builds and runs ok on non-Linux/x86 + platforms. +@end enumerate +These tests were conducted on a 225 MHz IDT WinChip machine, running +Linux 2.0.36. They represent nearly a week of continuous computation. +All tests completed successfully. + + +@section Further reading +@code{bzip2} is not research work, in the sense that it doesn't present +any new ideas. Rather, it's an engineering exercise based on existing +ideas. + +Four documents describe essentially all the ideas behind @code{bzip2}: +@example +Michael Burrows and D. J. Wheeler: + "A block-sorting lossless data compression algorithm" + 10th May 1994. + Digital SRC Research Report 124. + ftp://ftp.digital.com/pub/DEC/SRC/research-reports/SRC-124.ps.gz + If you have trouble finding it, try searching at the + New Zealand Digital Library, http://www.nzdl.org. + +Daniel S. Hirschberg and Debra A. LeLewer + "Efficient Decoding of Prefix Codes" + Communications of the ACM, April 1990, Vol 33, Number 4. + You might be able to get an electronic copy of this + from the ACM Digital Library. + +David J. Wheeler + Program bred3.c and accompanying document bred3.ps. + This contains the idea behind the multi-table Huffman + coding scheme. + ftp://ftp.cl.cam.ac.uk/users/djw3/ + +Jon L. Bentley and Robert Sedgewick + "Fast Algorithms for Sorting and Searching Strings" + Available from Sedgewick's web page, + www.cs.princeton.edu/~rs +@end example +The following paper gives valuable additional insights into the +algorithm, but is not immediately the basis of any code +used in bzip2. +@example +Peter Fenwick: + Block Sorting Text Compression + Proceedings of the 19th Australasian Computer Science Conference, + Melbourne, Australia. Jan 31 - Feb 2, 1996. + ftp://ftp.cs.auckland.ac.nz/pub/peter-f/ACSC96paper.ps +@end example +Kunihiko Sadakane's sorting algorithm, mentioned above, +is available from: +@example +http://naomi.is.s.u-tokyo.ac.jp/~sada/papers/Sada98b.ps.gz +@end example +The Manber-Myers suffix array construction +algorithm is described in a paper +available from: +@example +http://www.cs.arizona.edu/people/gene/PAPERS/suffix.ps +@end example +Finally, the following paper documents some recent investigations +I made into the performance of sorting algorithms: +@example +Julian Seward: + On the Performance of BWT Sorting Algorithms + Proceedings of the IEEE Data Compression Conference 2000 + Snowbird, Utah. 28-30 March 2000. +@end example + + +@contents + +@bye + diff -Nru bzip2-1.0.2/doc/bzip2recover.1 bzip2-1.0.2.new/doc/bzip2recover.1 --- bzip2-1.0.2/doc/bzip2recover.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzip2recover.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1 @@ +.so bzip2.1 \ No newline at end of file diff -Nru bzip2-1.0.2/doc/bzmore.1 bzip2-1.0.2.new/doc/bzmore.1 --- bzip2-1.0.2/doc/bzmore.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/bzmore.1 Sun Dec 30 03:12:35 2001 @@ -0,0 +1,152 @@ +.\"Shamelessly copied from zmore.1 by Philippe Troin +.\"for Debian GNU/Linux +.TH BZMORE 1 +.SH NAME +bzmore, bzless \- file perusal filter for crt viewing of bzip2 compressed text +.SH SYNOPSIS +.B bzmore +[ name ... ] +.br +.B bzless +[ name ... ] +.SH NOTE +In the following description, +.I bzless +and +.I less +can be used interchangeably with +.I bzmore +and +.I more. +.SH DESCRIPTION +.I Bzmore +is a filter which allows examination of compressed or plain text files +one screenful at a time on a soft-copy terminal. +.I bzmore +works on files compressed with +.I bzip2 +and also on uncompressed files. +If a file does not exist, +.I bzmore +looks for a file of the same name with the addition of a .bz2 suffix. +.PP +.I Bzmore +normally pauses after each screenful, printing --More-- +at the bottom of the screen. +If the user then types a carriage return, one more line is displayed. +If the user hits a space, +another screenful is displayed. Other possibilities are enumerated later. +.PP +.I Bzmore +looks in the file +.I /etc/termcap +to determine terminal characteristics, +and to determine the default window size. +On a terminal capable of displaying 24 lines, +the default window size is 22 lines. +Other sequences which may be typed when +.I bzmore +pauses, and their effects, are as follows (\fIi\fP is an optional integer +argument, defaulting to 1) : +.PP +.IP \fIi\|\fP +display +.I i +more lines, (or another screenful if no argument is given) +.PP +.IP ^D +display 11 more lines (a ``scroll''). +If +.I i +is given, then the scroll size is set to \fIi\|\fP. +.PP +.IP d +same as ^D (control-D) +.PP +.IP \fIi\|\fPz +same as typing a space except that \fIi\|\fP, if present, becomes the new +window size. Note that the window size reverts back to the default at the +end of the current file. +.PP +.IP \fIi\|\fPs +skip \fIi\|\fP lines and print a screenful of lines +.PP +.IP \fIi\|\fPf +skip \fIi\fP screenfuls and print a screenful of lines +.PP +.IP "q or Q" +quit reading the current file; go on to the next (if any) +.PP +.IP "e or q" +When the prompt --More--(Next file: +.IR file ) +is printed, this command causes bzmore to exit. +.PP +.IP s +When the prompt --More--(Next file: +.IR file ) +is printed, this command causes bzmore to skip the next file and continue. +.PP +.IP = +Display the current line number. +.PP +.IP \fIi\|\fP/expr +search for the \fIi\|\fP-th occurrence of the regular expression \fIexpr.\fP +If the pattern is not found, +.I bzmore +goes on to the next file (if any). +Otherwise, a screenful is displayed, starting two lines before the place +where the expression was found. +The user's erase and kill characters may be used to edit the regular +expression. +Erasing back past the first column cancels the search command. +.PP +.IP \fIi\|\fPn +search for the \fIi\|\fP-th occurrence of the last regular expression entered. +.PP +.IP !command +invoke a shell with \fIcommand\|\fP. +The character `!' in "command" are replaced with the +previous shell command. The sequence "\\!" is replaced by "!". +.PP +.IP ":q or :Q" +quit reading the current file; go on to the next (if any) +(same as q or Q). +.PP +.IP . +(dot) repeat the previous command. +.PP +The commands take effect immediately, i.e., it is not necessary to +type a carriage return. +Up to the time when the command character itself is given, +the user may hit the line kill character to cancel the numerical +argument being formed. +In addition, the user may hit the erase character to redisplay the +--More-- message. +.PP +At any time when output is being sent to the terminal, the user can +hit the quit key (normally control\-\\). +.I Bzmore +will stop sending output, and will display the usual --More-- +prompt. +The user may then enter one of the above commands in the normal manner. +Unfortunately, some output is lost when this is done, due to the +fact that any characters waiting in the terminal's output queue +are flushed when the quit signal occurs. +.PP +The terminal is set to +.I noecho +mode by this program so that the output can be continuous. +What you type will thus not show on your terminal, except for the / and ! +commands. +.PP +If the standard output is not a teletype, then +.I bzmore +acts just like +.I bzcat, +except that a header is printed before each file. +.SH FILES +.DT +/etc/termcap Terminal data base +.SH "SEE ALSO" +more(1), less(1), bzip2(1), bzdiff(1), bzgrep(1) diff -Nru bzip2-1.0.2/doc/pl/Makefile.am bzip2-1.0.2.new/doc/pl/Makefile.am --- bzip2-1.0.2/doc/pl/Makefile.am Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/pl/Makefile.am Fri Feb 1 04:19:11 2002 @@ -0,0 +1,4 @@ + +mandir = @mandir@/pl +man_MANS = bzip2.1 bunzip2.1 bzcat.1 bzip2recover.1 + diff -Nru bzip2-1.0.2/doc/pl/bunzip2.1 bzip2-1.0.2.new/doc/pl/bunzip2.1 --- bzip2-1.0.2/doc/pl/bunzip2.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/pl/bunzip2.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1 @@ +.so bzip2.1 \ No newline at end of file diff -Nru bzip2-1.0.2/doc/pl/bzcat.1 bzip2-1.0.2.new/doc/pl/bzcat.1 --- bzip2-1.0.2/doc/pl/bzcat.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/pl/bzcat.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1 @@ +.so bzip2.1 \ No newline at end of file diff -Nru bzip2-1.0.2/doc/pl/bzip2.1 bzip2-1.0.2.new/doc/pl/bzip2.1 --- bzip2-1.0.2/doc/pl/bzip2.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/pl/bzip2.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1,384 @@ +.\" T³umaczenie Maciej Wojciechowski wojciech@staszic.waw.pl +.PU +.TH bzip2 1 "" "" "wersja 1.0" +.SH NAZWA +bzip2, bunzip2 \- sortuj±cy bloki kompresor/dekompresor plików, v1.0 +.br +bzcat \- dekompresuje pliki na standardowe wyj¶cie +.br +bzip2recover \- odzyskuje dane ze zniszczonych archiwów bzip2 +.SH SK£ADNIA +.ll +8 +.B bzip2 +.RB [ \-cdfkqstvzVL123456789 ] +.RI [ nazwy_plików \&...] +.ll -8 +.br +.B bunzip2 +.RB [ \-fkvsVL ] +.RI [ nazwy_plików \&...] +.br +.B bzcat +.RB [ \-s ] +.RI [ nazwy_plików \&...] +.br +.B bzip2recover +.I nazwa_pliku +.SH OPIS +.I bzip2 +kompresuje pliki uæywaj±c algorytmu sortowania bloków Burrowsa-Wheelera i +kodu Huffmana. Kompresja jest generalnie sporo lepsza od konwencjonalnych +kompresorów opartych o metodź LZ77/LZ78, i jest porównywalna z +osi±gniźciami statystycznych kompresorów z rodziny PPM. + +Opcje linii poleceń s± w wiźkszo¶ci bardzo podobne do tych z +.IR "GNU gzip" , +ale nie s± identyczne. + +.I bzip2 +oczekuje listy plików towarzysz±cych parametrom linii poleceń. Kaædy plik jest +zastźpowany przez swoj± skompresowan± wersjź, z nazw± +"oryginalny_plik.bz2". Kaædy skompresowany plik ma ten sam czas modyfikacji, +uprawnienia i, je¶li to moæliwe, w³a¶ciciela co orygina³, po to, aby te +ustawienia mog³y zostaę odtworzone podczas dekompresji. Utrzymywanie nazwy +plików nie jest do końca dok³adne w tym sensie, æe nie ma moæliwo¶ci +przetrzymywania daty, uprawnień, w³a¶ciciela i nazw plików na systemach, na +których brakuje tych moæliwo¶ci lub maj± ograniczenia co do d³ugo¶ci nazwy, +tak np. jak MS-DOS. + +.I bzip2 +i +.I bunzip2 +standardowo nie nadpisuj± istniej±cych juæ plików. Je¶li chcesz aby to +robi³y, musisz uæyę parametru \-f. + +Je¶li nie podano æadnej nazwy pliku, +.I bzip2 +kompresuje ze standardowego wej¶cia na standardowe wyj¶cie. Odmiawia wówczas +wypisywania skompresowanego wyj¶cie na terminal, gdyæ by³oby to +ca³kiem niezrozumia³e i przez to bez wiźkszego sensu. + +.I bunzip2 +(lub +.IR bzip2 \-d ) +dekompresuje wszystkie podane pliki. Pliki, które nie by³y +utworzone przez +.I bzip2 +zostan± wykryte i zignorowane, a na ekranie pojawi siź komunikat +ostrzegawczy. +.I bzip2 +próbuje zgadn±ę nazwź dla dekompresowanego pliku w nastźpuj±cy sposób: +.nf + nazwa_pliku.bz2 staje siź nazwa_pliku + nazwa_pliku.bz staje siź nazwa_pliku + nazwa_pliku.tbz2 staje siź nazwa_pliku.tar + nazwa_pliku.tbz staje siź nazwa_pliku.tar + inna_nazwa staje siź inna_nazwa.out +.fi +Je¶li plik nie ma jednego z nastźpuj±cych rozpoznawalnych rozszerzeń, +.IR .bz2 , +.IR .bz , +.I .tbz2 +lub +.IR .tbz , +to +.I bzip2 +napisze, æe nie moæe zgadn±ę nazwy pierwotnego pliku, i uæyje +oryginalnej nazwy z dodanym rozszerzeniem +.IR .out . + +Tak jak kompresja, nie posiadaj±ca æadnych plików, powoduje kompresjź ze +standardowego wej¶cia na standardowe wyj¶cie. + +.I bunzip2 +poprawnie zdekompresuje plik, który jest po³aczeniem dwóch lub wiźcej +skompresowanych plików. Rezultatem jest po³±czony odpowiedni +nieskompresowany plik. Obs³ugiwane jest równieæ sprawdzanie spójno¶ci +(\-t) po³±czonych skompresowanych plików. + +Moæesz równieæ kompresowaę lub dekompresowaę pliki na standardowe wyj¶cie +uæywaj±c parametru \-c. W ten w³a¶nie sposób moæna przeprowadzaę kompresjź +wielu plików równocze¶nie. +Powsta³e wyniki s± przesy³ane sekwencyjnie na standardowe wyj¶cie. +W ten sposób kompresja wielu plików generuje strumień +zawieraj±cy reprezentacje kilku skompresowanych plików. Taki strumień moæe +byę zdekompresowany poprawnie tylko przez +.I bzip2 +w wersji 0.9.0 lub pó¼niejszej. Wcze¶niejsze wersje +.I bzip2 +zatrzymaj± siź po zdekmpresowaniu pierwszego pliku w strumieniu. + +.I bzcat +(lub +.I bzip2 -dc) +dekompresuje wszystkie wybrane pliki na standardowe wyj¶cie. + +.I bzip2 +czyta argumenty ze zmiennych ¶rodowiskowych +.I BZIP2 +i +.I BZIP, +w podanej kolejno¶ci, i przetwarza je przed jakimikolwiek argumentami +przeczytanymi z linii poleceń. To dobra metoda na specyfikowanie +standardowych ustawień. + +Kompresja stosowana jest zawsze, nawet je¶li skompresowany plik jest +nieznaczniej wiźkszy od pliku oryginalnego. Pliki mniejsze niæ mniej wiźcej +sto bajtów staj± siź wiźksze, poniewaæ mechanizm kompresji ma sta³y +nag³ówek wynosz±cy oko³o 50 bajtów. Przypadkowe dane (w³±czaj±c wyj¶cie +wiźkszo¶ci kompresorów plików) d± kodowane na mniej wiźcej 8.05 bitu na +bajt, daj±c zysk oko³o 0.5%. + +Jako samosprawdzenie dla twojej ochrony +.I bzip2 +uæywa 32-bitowego CRC aby upewnię siź, æe zdekompresowana wersja pliku jest +identyczna z oryginaln±. To strzeæe przed stratami w skompresowanych danych +i przed niewykrytymi b³źdami w +.I bzip2 +(na szczź¶cie bardzo rzadkich). Moæliwo¶ę niewykrycia utraty danych +jest mikroskopijna, mniej wiźcej jedna szansa na cztery biliony dla kaædego +pliku. Uwaæaj jednak, gdyæ sprawdzenie jest dokonywane przed dekompresj±, +wiźc dowiesz siź tylko tego, æe co¶ jest nie w porz±dku. Nie pomoæe ci to odzyskaę +oryginalnych nieskompresowanych danych. Moæesz uæyę +.I bzip2recover +aby spróbowaę odzyskaę dane z uszkodzonych plików. + +Zwracane warto¶ci: 0 dla normalnego wyj¶cia, 1 dla problemów technicznych +(plik nie znaleziony, niew³a¶ciwy parametr, b³±d wyj¶cia/wyj¶cia itp.), 2 dla +zasygnalizowania b³źdu skompresowanego pliku, 3 dla wewnźtrznego b³źdu (np. +bug), który zmusi³ \fIbzip2\fP do przerwania. + +.SH OPCJE +.TP +.B \-c --stdout +Kompresuje lub dekompresuje na standardowe wyj¶cie. +.TP +.B \-d --decompress +Wymusza dekompresjź. +.IR bzip2 , +.I bunzip2 +i +.I bzcat +s± tak naprawdź tymi samymi programami i decyzja jakie akcje bźd± wykonane +jest wykonywana na podstawie nazwy jaka zosta³a uæyta. Ten parametr ma wyæszy +priorytet i wymusza na \fIbzip2\fP dekompresjź. +.TP +.B \-z --compress +Podobne do \-d: wymusza kompresjź, bez wzglźdu na sposób wywo³ania. +.TP +.B \-t --test +Sprawdza integralno¶ę wybranego pliku(ów), ale nie dekompresuje ich. Wymusza +to próbn± dekompresjź i mówi, jaki jest rezultat. +.TP +.B \-f --force +Wymusza zastźpowanie plików wyj¶ciowych. Normalnie, \fIbzip2\fP nie +zastźpuje istniej±cych plików wyj¶ciowych. Wymusza równieæ na \fIbzip2\fP +³amanie dowi±zań twardych, czego normalnie nie robi. +.TP +.B \-k --keep +Zatrzymaj (nie kasuj) pliki wej¶ciowe przy kompresji lub dekompresji. +.TP +.B \-s --small +Zredukuj uæycie pamiźci na kompresjź, dekompresjź i testowanie. Pliki s± +dekompresowane i testowane przy uæyciu zmodyfikowanego algorytmu, który +potrzebuje tylko 2.5 bajtu na blok bajtów. Oznacza to, æe kaædy plik moæe +byę zdekompresowany przy uæyciu oko³o 2300k pamiźci, jednak trac±c oko³o po³owź +normalnej szybko¶ci. + +Podczas kompresji, \-s wybiera bloki wielko¶ci 200k, których limity +pamiźci wynosz± mniej wiźcej tyle samo, w zamian za jako¶ę kompresji. W +skrócie, je¶li twój komputer ma ma³o pamiźci (8 megabajtów lub mniej), +uæywaj opcji \-s do wszystkiego. Zobacz \fBzarz±dzanie pamiźci±\fP poniæej. +.TP +.B \-q --quiet +Wy³±cza wszystkie nieistotne komunikaty ostrzegawcze. +Nie s± eliminowane komunikaty dotycz±ce b³źdów wej¶cia/wyj¶cia i innych +zdarzeń krytycznych. +.TP +.B \-v --verbose +Tryb gadatliwy -- pokazuje stopień kompresji dla kaædego pliku. Nastźpne +\fB\-v\fP zwiźkszaj± stopień gadatliwo¶ci, powoduj±c wy¶wietlanie duæej +ilo¶ci informacji, przydatnych g³ównie przy diagnostyce. +.TP +.B \-L --license -V --version +Wy¶wietla wersjź programu i warunki licencji. +.TP +.B \-1 to \-9 +Ustawia wielko¶ę bloku na 100 k, 200 k .. 900 k przy kompresji. Nie ma +æadnego znaczenia przy dekompresji. Zobacz \fBzarz±dzanie pamiźci±\fP +poniæej. +.TP +.B \-- +Traktuje wszystkie nastźpuj±ce po nim argumenty jako nazwy plików, nawet je¶li +zaczynaj± siź one od my¶lnika. Moæesz wiźc kompresowaę i dekompresowaę +pliki, których nazwa zaczyna siź od my¶lnika, na przyk³ad: bzip2 \-- +\-mój_plik. +.TP +.B \--repetitive-fast --repetitive-best +Te parametry nie maj± znaczenia w wersjach 0.9.5 i wyæszych. Umoæliwia³y one +pewn± infantyln± kontrolź nad zachowaniem algorytmu sortuj±cego we +wcze¶niejszych wersjach, co by³o czasami uæyteczne. Wersje 0.9.5 i wyæsze +maj± usprawniony algorytm, który powoduje bezuæyteczno¶ę tej funkcji. + +.SH ZARZ”DZANIE PAMIŹCI” +.I bzip2 +kompresuje duæe pliki w blokach. Rozmiar bloku ma wp³yw zarówno na stopień +osi±ganej kompresji, jak równieæ na ilo¶ę pamiźci potrzebnej do kompresji +i dekompresji. Parametry od \-1 do \-9 wybieraj± rozmiar bloku odpowiednio +od 100,000 bajtów aæ do 900,000 bajtów (standardowo). W czasie dekompresji, +rozmiar bloku uæytego do kompresji jest odczytywany z nag³ówku pliku +skompresowanego i +.I bunzip2 +sam zajmuje odpowiedni± do dekompresji ilo¶ę pamiźci. Poniewaæ rozmiar +bloków jest przetrzymywany w pliku skompresowanym, parametry od \-1 do \-9 +nie maj± przy dekompresji æadnego znaczenia. + +Wymagania kompresji i dekompresji w bajtach, mog± byę wyliczone przez: + + Kompresja : 400k + ( 8 x rozmiar bloku ) + + Dekompresja : 100k + ( 4 x rozmiar bloku ) lub + 100k + ( 2.5 x rozmiar bloku ) + +Wiźksze bloki daj± duæe zmniejszenie zwrotów marginalnych. Wiźkszo¶ę +kompresji pochodzi z pierwszych stu lub dwustu kilobajtów rozmiaru bloku. +Warto o tym pamiźtaę uæywaj±c \fIbzip2\fP na wolnych +komputerach. Warto równieæ podkre¶lię, æe rozmiar pamiźci potrzebnej do +dekompresji jest wybierany poprzez ustawienie odpowiedniej +wielko¶ci bloku przy kompresji. + +Dla plików skompresowanych standardowym blokiem wielko¶ci 900k, +\fIbunzip2\fP bździe wymaga³ oko³o 3700 kilobajtów do dekompresji. Aby +umoæliwię dekompresjź na komputerze wyposaæonym jedynie w 4 megabajty +pamiźci, \fIbunzip2\fP ma opcjź, która moæe zmniejszyę wymagania prawie do +po³owy, tzn. oko³o 2300 kilobajtów. Prźdko¶ę dekompresji jest równieæ bardzo +zmiejszona, wiźc uæywaj tej opcji tylko wtedy, kiedy jest to konieczne. Tym +parametrem jest -s. + +Generalnie, próbuj i uæywaj najwiźkszych rozmiarów bloków, je¶li ilo¶ę +pamiźci ci na to pozwala. Prźdko¶ę kompresji i dekompresji w zasadzie nie +zaleæy od wielko¶ci uæytego bloku. + +Inna waæna rzecz dotyczy plików, które mieszcz± siź w pojedyńczym bloku -- +oznacza to wiźkszo¶ę plików na które siź natkniesz uæywaj±c duæych bloków. +Rozmiar realny pamiźci zabieranej jest proporcjonalny do wielko¶ci pliku, +je¶li plik jest mniejszy niæ blok. Na przyk³ad, kompresja pliku o +wielko¶ci 20,000 bajtów z parametrem -9 wymusi na kompresorze odnalezienie +7600 k pamiźci, ale zajźcie tylko 400k + 20000 * 8 = 560 kilobajtów z +tego. Podobnie, dekompresor odnajdzie 3700k, ale zajmie tylko 100k + 20000 +* 4 = 180 kilobajtów. + +Tu jest tabela, która podsumowuje maksymalne uæycie pamiźci dla róænych +rozmiarów bloków. Podano teæ ca³kowity rozmiar skompresowanych 14 +plików tekstowych (Calgary Text Compressione Corpus) zajmuj±cych razem +3,141,622 bajtów. Ta kolumna daje pewne pojźcie o tym, jaki wp³yw na +kompresjź ma wielko¶ę bloków. Ta tabela uzmys³awia równieæ przewagź uæycia +wiźkszych bloków dla wiźkszych plików, poniewaæ "Corpus" jest zdominowany +przez mniejsze pliki. +.nf + Uæycie Uæycie Uæycie Corpus + Parametr kompresji dekompresji dekompresji -s Size + + -1 1200k 500k 350k 914704 + -2 2000k 900k 600k 877703 + -3 2800k 1300k 850k 860338 + -4 3600k 1700k 1100k 846899 + -5 4400k 2100k 1350k 845160 + -6 5200k 2500k 1600k 838626 + -7 6100k 2900k 1850k 834096 + -8 6800k 3300k 2100k 828642 + -9 7600k 3700k 2350k 828642 +.fi +.SH ODZYSKIWANIE DANYCH ZE ZNISZCZONYCH PLIKÓW BZIP2 +.I bzip2 +kompresuje pliki w blokach, zazwyczaj 900 kilbajtowych. Kaædy blok jest +trzymany osobno. Je¶li b³źdy transmisji lub no¶nika uszkodz± plik +wieloblokowy .bz2, moæliwe jest odtworzenie danych zawartych w +niezniszczonych blokach pliku. + +Kaædy blok jest reprezentowany przez 48-bitowy wzorzec, który umoæliwia +znajdowanie przyporz±dkowań bloków z rozs±dn± pewno¶ci±. Kaædy blok +ma równieæ swój 32-bitowy CRC, wiźc bloki uszkodzone mog± byę ³atwo +odseparowane od poprawnych. + +.I bzip2recover +jest oddzielnym programem, którego zadaniem jest poszukiwanie bloków w +plikach .bz2 i zapisywanie ich do w³asnego pliku .bz2. Moæesz potem uæyę +\fIbzip2\fP \-t aby sprawdzię spójno¶ę wyj¶ciowego pliku i zdekompresowaę +te, które nie s± uszkodzone. + +.I bzip2recover +pobiera pojedynczy argument, nazwź uszkodzonego pliku, i tworzy pewn± liczbź +plików "rec0001plik.bz2", "rec0002plik.bz2", itd., przetrzymuj±ce odzyskane +bloki. Wyj¶ciowe nazwy plików s± tak tworzone, aby ³atwo by³o potem uæywaę +ich razem za pomoc± gwiazdek -- na przyk³ad, "bzip2 -dc rec*plik.bz2 > +odzyskany_plik" -- wylistuje pliki we w³a¶ciwej kolejno¶ci. + +.I bzip2recover +powinien byę uæywany najczź¶ciej z duæymi plikami .bz2, jako iæ one +zawieraj± najczź¶ciej duæo bloków. Jest czystym bezsensem uæywaę go na +uszkodzonym jedno-blokowym pliku, poniewaæ uszkodzony blok nie moæe byę +odzyskany. Je¶li chcesz zminimalizowaę jakiekolwiek moæliwe straty danych +poprzez no¶nik lub transmisjź, powiniene¶ zastanowię siź nad uæyciem +mniejszych bloków. + +.SH OPISY WYNIKÓW +Etap sortuj±cy kompresji ³±czy razem podobne ci±gi znaków w pliku. Przez +to, pliki zawieraj±ce bardzo d³ugie ci±gi powtarzaj±cych siź symboli, jak +"aabaabaabaab ..." (powtórzone kilkaset razy) mog± byę kompresowane wolniej +niæ normalnie. Wersje 0.9.5 i wyæsze zachowuj± siź duæo lepiej w tej +sytuacji niæ wersje poprzednie. Róænica stopnia kompresji pomiźdzy +najgorszym i najlepszym przypadkiem kompresji wynosi oko³o 10:1. Dla +wcze¶niejszych wersji by³o to nawet oko³o 100:1. Je¶li chcesz, moæesz uæyę +parametru \-vvvv aby monitorowaę postźpy bardzo szczegó³owo. + +Prźdko¶ę dekompresji nie jest zmieniana przez to zjawisko. + +.I bzip2 +zazwyczaj rezerwuje kilka megabajtów pamiźci do dzia³ania a +potem wykorzystuje j± w sposób zupe³nie przypadkowy. +Oznacza to, æe zarówno prźdko¶ę kompresji jak i dekompresji jest w +duæej czź¶ci zaleæna od prźdko¶ci, z jak± twój komputer moæe naprawiaę braki +bufora podrźcznego. Z tego powodu, wprowadzone zosta³y ma³e zmiany kody aby +zmniejszyę straty, które da³y nieproporcjonalnie duæy wzrost osi±gniźę. +My¶lź, æe +.I bzip2 +bździe dzia³a³ najlepiej na komputerach z duæymi buforami podrźcznymi. + +.SH ZAKAMARKI +Wiadomo¶ci o b³źdach wej¶cia/wyj¶cia nie s± aæ tak pomocne, jak mog³yby byę. +.I bzip2 +stara siź wykryę b³±d wej¶cia/wyj¶cia i wyj¶ę "czysto", ale +szczegó³y tego, jaki to problem mog± byę czasami bardzo myl±ce. + +Ta strona podrźcznika odnosi siź do wersji 1.0 programu \fIbzip2\fP. +Skompresowane pliki utworzone przez tź wersjź s± kompatybilne zarówno z +w przód jak i wstecznie z poprzednimi publicznymi wydaniami, +wersjami 0.1pl2, 0.9.0 i 0.9.5 ale z ma³ymi wyj±tkami: 0.9.0 i wyæsze potrafi± +poprawnie dekompresowaę wiele skompresowanych plików z³±czonych w jeden. +0.1pl2 nie potrafi tego; zatrzyma siź juæ po dekompresji pierwszego pliku w +strumieniu. + +.I bzip2recover +uæywa 32-bitowych liczb do reprezentacji pozycji bitu w skompresowanym +pliku, wiźc nie moæe przetwarzaę skompresowanych plików d³uæszych niæ 512 +megabajtów. Moæna to ³atwo naprawię. + +.SH AUTOR +Julian Seward, jseward@acm.org. + +http://www.muraroa.demon.co.uk +http://sourceware.cygnus.com/bzip2 + +Idee zawarte w \fIbzip2\fP s± podzielone (przynajmniej) pomiźdzy +nastepuj±cy ludzi: Michael Burrows i David Wheeler (transformacja +sortuj±c± bloki), David Wheeler (znów, koder Huffmana), Peter Fenwick +(struktura kodowania modelu w oryginalnym \fIbzip2\fP, i wiele +udoskonaleń), i Alistair Moffar, Radford Neal i Ian Witten (arytmetyczny +koder w oryginalnym \fIbzip2\fP). Jestem im bardzo wdziźczny za ich pomoc, +wsparcie i porady. Zobacz stronź manuala w ¼ród³owej dystrybucji po +wska¼niki do ¼róde³ dokumentacji. Christian von Roques zachźci³ mnie do +wymy¶lenia szybszego algorytmu sortuj±cego, po to æeby przyspieszyę +kompresjź. Bela Lubkin zachźci³a mnie do polepszenia najgorszych wyników +kompresji. Wiele ludzi przys³a³o ³atki, pomog³o w róænych problemach, +poæyczy³o komputerów, da³o rady i by³o ogólnie pomocnych. diff -Nru bzip2-1.0.2/doc/pl/bzip2recover.1 bzip2-1.0.2.new/doc/pl/bzip2recover.1 --- bzip2-1.0.2/doc/pl/bzip2recover.1 Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/doc/pl/bzip2recover.1 Fri Feb 1 04:19:11 2002 @@ -0,0 +1 @@ +.so bzip2.1 \ No newline at end of file diff -Nru bzip2-1.0.2/huffman.c bzip2-1.0.2.new/huffman.c --- bzip2-1.0.2/huffman.c Sun Dec 30 03:19:17 2001 +++ bzip2-1.0.2.new/huffman.c Fri Feb 1 04:19:11 2002 @@ -58,6 +58,10 @@ For more information on these sources, see the manual. --*/ +#ifdef HAVE_CONFIG_H +#include +#endif + #include "bzlib_private.h" diff -Nru bzip2-1.0.2/makefile.msc bzip2-1.0.2.new/makefile.msc --- bzip2-1.0.2/makefile.msc Wed Jan 2 05:02:33 2002 +++ bzip2-1.0.2.new/makefile.msc Thu Jan 1 01:00:00 1970 @@ -1,63 +0,0 @@ -# Makefile for Microsoft Visual C++ 6.0 -# usage: nmake -f makefile.msc -# K.M. Syring (syring@gsf.de) -# Fixed up by JRS for bzip2-0.9.5d release. - -CC=cl -CFLAGS= -DWIN32 -MD -Ox -D_FILE_OFFSET_BITS=64 -nologo - -OBJS= blocksort.obj \ - huffman.obj \ - crctable.obj \ - randtable.obj \ - compress.obj \ - decompress.obj \ - bzlib.obj - -all: lib bzip2 test - -bzip2: lib - $(CC) $(CFLAGS) -o bzip2 bzip2.c libbz2.lib setargv.obj - $(CC) $(CFLAGS) -o bzip2recover bzip2recover.c - -lib: $(OBJS) - lib /out:libbz2.lib $(OBJS) - -test: bzip2 - type words1 - .\\bzip2 -1 < sample1.ref > sample1.rb2 - .\\bzip2 -2 < sample2.ref > sample2.rb2 - .\\bzip2 -3 < sample3.ref > sample3.rb2 - .\\bzip2 -d < sample1.bz2 > sample1.tst - .\\bzip2 -d < sample2.bz2 > sample2.tst - .\\bzip2 -ds < sample3.bz2 > sample3.tst - @echo All six of the fc's should find no differences. - @echo If fc finds an error on sample3.bz2, this could be - @echo because WinZip's 'TAR file smart CR/LF conversion' - @echo is too clever for its own good. Disable this option. - @echo The correct size for sample3.ref is 120,244. If it - @echo is 150,251, WinZip has messed it up. - fc sample1.bz2 sample1.rb2 - fc sample2.bz2 sample2.rb2 - fc sample3.bz2 sample3.rb2 - fc sample1.tst sample1.ref - fc sample2.tst sample2.ref - fc sample3.tst sample3.ref - - - -clean: - del *.obj - del libbz2.lib - del bzip2.exe - del bzip2recover.exe - del sample1.rb2 - del sample2.rb2 - del sample3.rb2 - del sample1.tst - del sample2.tst - del sample3.tst - -.c.obj: - $(CC) $(CFLAGS) -c $*.c -o $*.obj - diff -Nru bzip2-1.0.2/manual.ps bzip2-1.0.2.new/manual.ps --- bzip2-1.0.2/manual.ps Sat Jan 5 01:52:57 2002 +++ bzip2-1.0.2.new/manual.ps Thu Jan 1 01:00:00 1970 @@ -1,3991 +0,0 @@ -%!PS-Adobe-2.0 -%%Creator: dvips(k) 5.86 Copyright 1999 Radical Eye Software -%%Title: manual.dvi -%%Pages: 40 -%%PageOrder: Ascend -%%BoundingBox: 0 0 596 842 -%%EndComments -%DVIPSWebPage: (www.radicaleye.com) -%DVIPSCommandLine: dvips -o manual.ps manual.dvi -%DVIPSParameters: dpi=600, compressed -%DVIPSSource: TeX output 2002.01.05:0052 -%%BeginProcSet: texc.pro -%! -/TeXDict 300 dict def TeXDict begin/N{def}def/B{bind def}N/S{exch}N/X{S -N}B/A{dup}B/TR{translate}N/isls false N/vsize 11 72 mul N/hsize 8.5 72 -mul N/landplus90{false}def/@rigin{isls{[0 landplus90{1 -1}{-1 1}ifelse 0 -0 0]concat}if 72 Resolution div 72 VResolution div neg scale isls{ -landplus90{VResolution 72 div vsize mul 0 exch}{Resolution -72 div hsize -mul 0}ifelse TR}if Resolution VResolution vsize -72 div 1 add mul TR[ -matrix currentmatrix{A A round sub abs 0.00001 lt{round}if}forall round -exch round exch]setmatrix}N/@landscape{/isls true N}B/@manualfeed{ -statusdict/manualfeed true put}B/@copies{/#copies X}B/FMat[1 0 0 -1 0 0] -N/FBB[0 0 0 0]N/nn 0 N/IEn 0 N/ctr 0 N/df-tail{/nn 8 dict N nn begin -/FontType 3 N/FontMatrix fntrx N/FontBBox FBB N string/base X array -/BitMaps X/BuildChar{CharBuilder}N/Encoding IEn N end A{/foo setfont}2 -array copy cvx N load 0 nn put/ctr 0 N[}B/sf 0 N/df{/sf 1 N/fntrx FMat N -df-tail}B/dfs{div/sf X/fntrx[sf 0 0 sf neg 0 0]N df-tail}B/E{pop nn A -definefont setfont}B/Cw{Cd A length 5 sub get}B/Ch{Cd A length 4 sub get -}B/Cx{128 Cd A length 3 sub get sub}B/Cy{Cd A length 2 sub get 127 sub} -B/Cdx{Cd A length 1 sub get}B/Ci{Cd A type/stringtype ne{ctr get/ctr ctr -1 add N}if}B/id 0 N/rw 0 N/rc 0 N/gp 0 N/cp 0 N/G 0 N/CharBuilder{save 3 -1 roll S A/base get 2 index get S/BitMaps get S get/Cd X pop/ctr 0 N Cdx -0 Cx Cy Ch sub Cx Cw add Cy setcachedevice Cw Ch true[1 0 0 -1 -.1 Cx -sub Cy .1 sub]/id Ci N/rw Cw 7 add 8 idiv string N/rc 0 N/gp 0 N/cp 0 N{ -rc 0 ne{rc 1 sub/rc X rw}{G}ifelse}imagemask restore}B/G{{id gp get/gp -gp 1 add N A 18 mod S 18 idiv pl S get exec}loop}B/adv{cp add/cp X}B -/chg{rw cp id gp 4 index getinterval putinterval A gp add/gp X adv}B/nd{ -/cp 0 N rw exit}B/lsh{rw cp 2 copy get A 0 eq{pop 1}{A 255 eq{pop 254}{ -A A add 255 and S 1 and or}ifelse}ifelse put 1 adv}B/rsh{rw cp 2 copy -get A 0 eq{pop 128}{A 255 eq{pop 127}{A 2 idiv S 128 and or}ifelse} -ifelse put 1 adv}B/clr{rw cp 2 index string putinterval adv}B/set{rw cp -fillstr 0 4 index getinterval putinterval adv}B/fillstr 18 string 0 1 17 -{2 copy 255 put pop}for N/pl[{adv 1 chg}{adv 1 chg nd}{1 add chg}{1 add -chg nd}{adv lsh}{adv lsh nd}{adv rsh}{adv rsh nd}{1 add adv}{/rc X nd}{ -1 add set}{1 add clr}{adv 2 chg}{adv 2 chg nd}{pop nd}]A{bind pop} -forall N/D{/cc X A type/stringtype ne{]}if nn/base get cc ctr put nn -/BitMaps get S ctr S sf 1 ne{A A length 1 sub A 2 index S get sf div put -}if put/ctr ctr 1 add N}B/I{cc 1 add D}B/bop{userdict/bop-hook known{ -bop-hook}if/SI save N @rigin 0 0 moveto/V matrix currentmatrix A 1 get A -mul exch 0 get A mul add .99 lt{/QV}{/RV}ifelse load def pop pop}N/eop{ -SI restore userdict/eop-hook known{eop-hook}if showpage}N/@start{ -userdict/start-hook known{start-hook}if pop/VResolution X/Resolution X -1000 div/DVImag X/IEn 256 array N 2 string 0 1 255{IEn S A 360 add 36 4 -index cvrs cvn put}for pop 65781.76 div/vsize X 65781.76 div/hsize X}N -/p{show}N/RMat[1 0 0 -1 0 0]N/BDot 260 string N/Rx 0 N/Ry 0 N/V{}B/RV/v{ -/Ry X/Rx X V}B statusdict begin/product where{pop false[(Display)(NeXT) -(LaserWriter 16/600)]{A length product length le{A length product exch 0 -exch getinterval eq{pop true exit}if}{pop}ifelse}forall}{false}ifelse -end{{gsave TR -.1 .1 TR 1 1 scale Rx Ry false RMat{BDot}imagemask -grestore}}{{gsave TR -.1 .1 TR Rx Ry scale 1 1 false RMat{BDot} -imagemask grestore}}ifelse B/QV{gsave newpath transform round exch round -exch itransform moveto Rx 0 rlineto 0 Ry neg rlineto Rx neg 0 rlineto -fill grestore}B/a{moveto}B/delta 0 N/tail{A/delta X 0 rmoveto}B/M{S p -delta add tail}B/b{S p tail}B/c{-4 M}B/d{-3 M}B/e{-2 M}B/f{-1 M}B/g{0 M} -B/h{1 M}B/i{2 M}B/j{3 M}B/k{4 M}B/w{0 rmoveto}B/l{p -4 w}B/m{p -3 w}B/n{ -p -2 w}B/o{p -1 w}B/q{p 1 w}B/r{p 2 w}B/s{p 3 w}B/t{p 4 w}B/x{0 S -rmoveto}B/y{3 2 roll p a}B/bos{/SS save N}B/eos{SS restore}B end - -%%EndProcSet -TeXDict begin 39158280 55380996 1000 600 600 (manual.dvi) -@start -%DVIPSBitmapFont: Fa cmbxti10 14.4 1 -/Fa 1 47 df<13FCEA03FF000F13804813C05AA25AA2B5FCA31480A214006C5A6C5A6C5A -EA0FE0121271912B>46 D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fb cmti10 10.95 26 -/Fb 26 122 df37 D44 D<120FEA3FC0127FA212FFA31380EA7F00123C0A0A77891C>46 -D<15FE913807FF8091381F07C091387C01F0ECF000494813F8494813780107147C495A49 -C7FC167E133E137EA25BA2485AA2000315FEA25B000715FCA2491301120FA34848EB03F8 -A44848EB07F0A448C7EA0FE0A316C0007E141F12FE1680153FA2481500A2157EA25DA25D -4813015D6C495A127C4A5A4A5A6C49C7FC143E6C5B380FC1F03803FFC0C648C8FC273F76 -BC2E>48 D<15031507150F151F151E153E157EEC01FEEC03FC1407141FEB01FF90380FFB -F8EB1FC3EB0E07130015F0A2140FA215E0A2141FA215C0A2143FA21580A2147FA21500A2 -5CA25CA21301A25CA21303A25CA21307A25CA2130FA25CA2131FA25CEB7FE0B612F0A215 -E0203D77BC2E>I59 D<147E49B47E903907C1C38090391F80EF -C090383F00FF017E137F4914804848133F485AA248481400120F5B001F5C157E485AA215 -FE007F5C90C7FCA21401485C5AA21403EDF0385AA21407EDE078020F1370127C021F13F0 -007E013F13E0003E137FECF3E1261F01E313C03A0F8781E3803A03FF00FF00D800FC133E -252977A72E>97 DII -II<167C4BB4FC923807C78092380F83C0ED1F87161FED3F3FA2157E -A21780EE0E004BC7FCA414015DA414035DA30103B512F8A390260007E0C7FCA3140F5DA5 -141F5DA4143F92C8FCA45C147EA414FE5CA413015CA4495AA4495AA4495A121E127F5C12 -FF49C9FCA2EAFE1EEAF83C1270EA7878EA3FE0EA0F802A5383BF1C>I -104 D<1478EB01FCA21303A314F8EB00E01400AD137C48B4FC38038F80EA0707000E13C0 -121E121CEA3C0F1238A2EA781F00701380A2EAF03F140012005B137E13FE5BA212015BA2 -12035B1438120713E0000F1378EBC070A214F0EB80E0A2EB81C01383148038078700EA03 -FEEA00F8163E79BC1C>I108 DIII<903903E001F890390FF807FE90 -3A1E7C1E0F80903A1C3E3C07C0013C137801389038E003E0EB783F017001C013F0ED8001 -9038F07F0001E015F8147E1603000113FEA2C75AA20101140717F05CA20103140F17E05C -A20107EC1FC0A24A1480163F010F15005E167E5E131F4B5A6E485A4B5A90393FB80F80DA -9C1FC7FCEC0FFCEC03E049C9FCA2137EA213FEA25BA21201A25BA21203A2387FFFE0B5FC -A22D3A80A72E>I -114 DII<137C48B4141C26 -038F80137EEA0707000E7F001E15FE121CD83C0F5C12381501EA781F007001805BA2D8F0 -3F1303140000005D5B017E1307A201FE5C5B150F1201495CA2151F0003EDC1C0491481A2 -153F1683EE0380A2ED7F07000102FF13005C01F8EBDF0F00009038079F0E90397C0F0F1C -90391FFC07F8903907F001F02A2979A731>I<017CEB01C048B4EB07F038038F80EA0707 -000E01C013F8121E001C1403EA3C0F0038EC01F0A2D8781F130000705BA2EAF03F91C712 -E012005B017E130116C013FE5B1503000115805BA2ED07001203495B150EA25DA25D1578 -000114706D5B0000495A6D485AD97E0FC7FCEB1FFEEB03F0252979A72A>I<017C167048 -B491387001FC3A038F8001F8EA0707000E01C015FE001E1403001CEDF000EA3C0F003817 -7C1507D8781F4A133C00701380A2D8F03F130F020049133812005B017E011F14784C1370 -13FE5B033F14F0000192C712E05BA2170100034A14C049137E17031880A2EF070015FE17 -0E00010101141E01F86D131C0000D9039F5BD9FC076D5A903A3E0F07C1E0903A1FFC03FF -C0902703F0007FC7FC372979A73C>I<137C48B4143826038F8013FCEA0707000E7F001E -1401001C15F8EA3C0F12381503D8781F14F000701380A2D8F03F1307020013E012005B01 -7E130F16C013FE5B151F1201491480A2153F000315005BA25D157EA315FE5D00011301EB -F8030000130790387C1FF8EB3FF9EB07E1EB00035DA21407000E5CEA3F80007F495AA24A -5AD8FF0090C7FC143E007C137E00705B387801F0383803E0381E0FC06CB4C8FCEA03F826 -3B79A72C>121 D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fc cmtt12 14.4 10 -/Fc 10 123 df50 D<383FFF805AB57EA37E7EEA003FAEED07FC92383FFF -8092B512E002C314F802CF8002DF8091B7FCDBF80F1380DBC00113C092C713E04A143F4A -EC1FF04A15F84A140F4AEC07FCA217034A15FE1701A318FF83A95F18FEA280170318FC6E -140718F86E140FEF1FF06E143F6EEC7FE06EECFFC0DBC0031380EDF01F92B6120002DF14 -FC02CF5C02C35CD91F8114C090260F807F90C7FC90C7EA0FF8384A7FC83E>98 -D<923803FFF85D4B7FA38181ED0003AEEC1FF0ECFFFE0103EBFF83010F14E34914F3017F -14FB90B7FC48EBF80F48EBC00191C7FC4848143F4848141F5B4848140F491407123F4914 -03127F5BA312FF90C8FCA97F127FA216077F123F6D140FA26C6C141F6D143F000F157F6C -6C14FF01FF5B6C6D5A6CD9F01FEBFFFC6C90B500FB13FE6D02F313FF6D14E3010F14C36D -020113FE010101FC14FC9026003FE0C8FC384A7CC83E>100 D<143E147F4A7E497FA56D -5B6EC8FC143E91C9FCAC003FB57E5A81A47EC7123FB3B3007FB71280B812C0A56C16802A -4A76C93E>105 D<007FB512C0B6FC81A47EC7121FB3B3B3A5007FB712F8B812FCA56C16 -F82E4978C83E>108 D111 -DI<903901FFF00F011F9038 -FE1F8090B612BF000315FF5A5A5A393FFE003F01F01307D87FC0130190C8FC5A48157FA4 -7EEE3F00D87FC091C7FC13F0EA3FFE381FFFF06CEBFFC06C14FE6C6E7EC615E0013F14F8 -010780D9003F7F02007F03071380030013C0003EED3FE0007F151F48150F17F06D1407A3 -7FA26D140F6D15E0161F01FCEC3FC06D14FF9026FFC00F138091B612005E485D013F5C6D -14E0D8FC0714802778007FF8C7FC2C3677B43E>115 D<147C14FC497EAD003FB712FC5A -B87EA36C5EA2260001FEC9FCB3A6173FA2EF7F80A76E14FF6D16006F5A9238C007FE9138 -7FF01F92B55A6E5C6E5C6E5C6E1480020149C7FC9138003FF031437DC13E>I<000FB812 -804817C04817E0A418C001C0C712014C13804C1300EE1FFE4C5AEE7FF06C484A5A4B5BC8 -485B4B90C7FC4B5A4B5A4B5A4B5A4B5A4A5B4A5B4A90C8FC4A5A4A5A4A5A4A5A4A5A495B -495B4990C9FC495A495A4948EC0FC0495A4948EC1FE0485B00075B4890C8FCEA1FFC485A -485A90B8FCB9FCA46C17C07E33337CB23E>122 D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fd cmtt12 13.14 31 -/Fd 31 123 df<903801FFC0010F13F8013F13FF4914C048B67E48814815FC4801007FD8 -1FF8130F01E0EB03FF48480100138049EC7FC0007F153F90C8121F4816E0160F17F07F16 -07A36CC8FC123EC9FCA2160F17E0A2161F17C0163FA2EE7F80EEFF00A24B5A4B5A15074B -5A4B5A4B5A4B5A4B5A020390C7FC4A5A4A5A4A5A4A5AEC7FC049485A4990C8FC495A495A -495AEB3FE0EBFFC04849EB03E04890C7EA07F0485A485AEA1FF048B7FC5AB8FCA37E6C16 -E02C447AC338>50 D<003FB6FC4815E0B712F882826C816C16802701FC000113C0923800 -7FE0161FEE0FF0A2160717F81603A6160717F0A2160FEE1FE0163FEE7FC0923801FF8003 -0F130090B65A5E16F08216FEEEFF8017C001FCC7EA7FE0EE1FF0EE07F8160317FCEE01FE -1600A217FF177FA717FF17FE16011603EE07FC160FEE3FF8EEFFF0003FB7FC4816E0B812 -C01780EEFE006C15F86C15C030437DC238>66 DI<007FB512F8B7FC16C08216F8826C813A03F8001FFF15070301 -13806F13C0167FEE3FE0161FEE0FF0A2EE07F8A2EE03FCA21601A217FE1600A4177FAD17 -FF17FEA4160117FCA2160317F81607A2EE0FF0161FEE3FE0167FEEFFC04B13805D031F13 -00007FB65AB75A5E5E16C093C7FC6C14F830437DC238>I<007FB712FCB87EA57ED801FC -C8FCA9177C94C7FCA6ED07C04B7EA590B6FCA79038FC000FA56F5A92C9FCA7EF0F80EF1F -C0AA007FB8FCB9FCA56C178032437DC238>I<91391FF003C091397FFC07E049B5FC0107 -14CF4914EF4914FF5B90387FF81F9038FFE00748EB800191C7FC4848147F485A49143F48 -5A161F485AA249140F123F5BA2127F90C8EA07C093C7FCA35A5AAA923807FFFC4B13FE4B -13FF7E7E6F13FE6F13FC9238000FE07F003F151FA27F121F7F163F6C7EA26C6C147F7F6C -6C14FF6C6C5B6E5A6C6D5A90387FF81F6DB6FC6D14EF6D14CF6D148F0101140F903A007F -FC07C0DA1FF0C7FC30457CC338>71 D<007FB612F0B712F8A56C15F0260001FCC7FCB3B3 -B1007FB612F0B712F8A56C15F0254377C238>73 D<90380FFFFE90B612E0000315F84881 -48814881A2273FFE000F138001F01301497F49147F4848EC3FC0A290C8121FA44816E0A2 -48150FB3AC6C151FA36C16C0A36D143FA36C6CEC7F806D14FF6D5B01FE130F6CB71200A2 -6C5D6C5D6C5DC615E0010F49C7FC2B457AC338>79 D<003FB512F04814FEB77E16E0826C -816C813A01FC003FFEED07FF03017F81707E163F161F83160FA7161F5F163F167F4C5A5D -030790C7FCED3FFE90B65A5E5E5EA282829038FC001FED07FC6F7E150115008282AA18E0 -EF01F0EF03F8A31783EE3F87263FFFE0ECC7F0486D14FFB56C7F18E07013C06C496D1380 -6C496D1300CA12FC35447EC238>82 D<003FB8FC481780B9FCA53BFE0007F0003FA9007C -EE1F00C792C7FCB3B3A70107B512F04980A56D5C31437DC238>84 -D<267FFFF090387FFFF0B56C90B512F8A56C496D13F0D801FCC73801FC00B3B3A66D1403 -00005EA36D14076D5D6E130F6D6C495A6E133F6D6C495A6D6C495AECFF076D90B5C7FC6D -5C6D5C6D5C023F13E0020F1380DA03FEC8FC35447FC238>I87 -D<001FB712F04816F85AA417F090C8121F17E0EE3FC0167F1780EEFF00A24B5A4B5A123E -C8485A4B5AA24B5A4B5AA24B5A4BC7FCA24A5A14035D4A5A140F5D4A5A143F5D4A5A14FF -92C8FC495A13035C495AA2495A495AA2495A495A17F849C7EA01FC485AA2485A485AA248 -5A121F5B485A127F90B7FCB8FCA56C16F82E437BC238>90 D<003FB712804816C0B812E0 -A46C16C06C16802B087A7D38>95 D97 DIIIII<14F0497E497E497EA4 -6D5A6D5A6D5A91C8FCAB383FFFFC487FB5FCA37E7EC7FCB3AF007FB612F0B712F816FCA3 -16F86C15F0264476C338>105 D<387FFFFEB6FCA57EC77EB3B3B1007FB7FCB81280A56C -1600294379C238>108 D<023FEB07E03B3FE0FFC01FF8D87FF39038E07FFCD8FFF76D48 -7E90B500F97F15FB6C91B612806C01C1EBF83F00030100EBE01F4902C013C0A24990387F -800FA2491400A349137EB3A73C3FFF03FFE07FFC4801879038F0FFFEB500C76D13FFA36C -01874913FE6C01039038E07FFC383080AF38>IIII114 D<903907FF80F0017FEBF1F848B5 -12FD000714FF5A5A5AEBFC00D87FE0131F0180130F48C71207481403A5007FEC01F001C0 -90C7FCEA3FF013FE381FFFF86CEBFFC0000314F8C614FF013F1480010714E0D9003F13F0 -020013F8ED0FFC1503003CEC01FE007E140000FE15FF167F7EA37F6D14FF16FE01F01303 -6DEB07FC01FF137F91B512F816F016E04815C0D8FC3F1400010F13FCD8780113E0283278 -B038>III<000FB712FC4816FE5AA417 -FC0180C7EA1FF8EE3FF0EE7FE0EEFFC04B13804B13006CC7485AC8485A4B5A4B5A4B5A4B -5A4A5B4A90C7FCEC07FC4A5A4A5A4A5A4A5A49485A4990C8FC495A495A495A495A494814 -7C494814FE485B4890C8FC485A485A485A485A48B7FCB8FCA56C16FC2F2F7DAE38>122 -D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fe cmbx12 13.14 53 -/Fe 53 122 df<923803FFE092B512FC020714FF021F81027F9038007FC0DAFFF0EB0FE0 -010301C0804990C7EA3FF84948147F4A81494814FF495AA2137F5CA2715A715A715AEF07 -8094C8FCA8EF07FCB9FCA526007FF0C7123F171FB3B3003FB5D8E00FB512F8A53D4D7ECC -44>12 D45 DI<177817F8EE01FCA21603A2EE07F8A217F0160FA217 -E0161FA2EE3FC0A21780167FA217005EA24B5AA25E1503A24B5AA25E150FA25E151FA24B -5AA25E157FA24BC7FCA25D1401A25D1403A24A5AA25D140FA24A5AA25D143FA25D147FA2 -4AC8FCA25C1301A25C1303A2495AA25C130FA2495AA25C133FA25C137FA249C9FCA25B12 -01A2485AA25B1207A25B120FA2485AA25B123FA25B127FA248CAFCA25AA2127CA22E6D79 -D13D>I<15F014011407141F147FEB03FF137FB6FCA313FC1380C7FCB3B3B2007FB712E0 -A52B4777C63D>49 DIIIII<121F7F7F13FE90B812E0A45A18C0188018005F5FA25F485E90C8EA07E0 -007E4B5A5F007C151F4CC7FC167E5E485D15014B5A4B5AC8485A4B5AA24BC8FC157EA25D -1401A24A5A1407A24A5AA2141FA24A5AA2147FA314FFA3495BA45BA55BAA6D5BA26D90C9 -FCEB007C334B79C93D>III65 -D<93261FFF80EB01C00307B500F81303033F02FE13074AB7EA800F0207EEE01F021F903A -FE007FF83F027F01E0903807FC7F91B5C73801FEFF010301FCEC007F4901F0814901C015 -0F4949814990C97E494882495A48498248197F5C48193F5C48191F5C48190FA2485BA21A -075AA391CDFCA2B5FCAD7EA280F207C0A27EA36C7F1A0F6C1A80806C191F6E18006C6180 -6C197E6C6D177C6D6C17FC6D6C4C5A6D6D4B5A6D6D4B5A6D01F0ED1FC06D01FC4B5A0100 -01FF03FFC7FC6E01E0EB07FE021F01FEEB3FFC020790B612F0020116C0DA003F92C8FC03 -0714F8DB001F13804A4D79CB59>67 DIII<93261FFF80EB01C00307B500F81303 -033F02FE13074AB7EA800F0207EEE01F021F903AFE007FF83F027F01E0903807FC7F91B5 -C73801FEFF010301FCEC007F4901F0814901C0150F4949814990C97E494882495A484982 -48197F5C48193F5C48191F5C48190FA2485BA21A075AA391CEFCA2B5FCAD7E050FB712C0 -80A37E94C7001FEBC000807EA27E807E807E806C7F7E6D7E6D7E6D7F6D01E05D6D6D5D6D -13FC010001FF4AB5FC6E01E0EB07F9021F01FFEB3FF0020791B5EAE07F0201EEC01FDA00 -3FED0007030702F81301DB001F018090C8FC524D79CB60>II< -B712FEA5D8000FEBE000B3B3B3ABB712FEA5274B7DCA2E>I76 DIIII82 DI<003FBB12C0A5DA80019038FC001FD9FC001601D87FF0943800 -7FE001C0183F49181F90C7170FA2007E1907A3007C1903A500FC1AF0481901A5C894C7FC -B3B3A749B812FCA54C4A7CC955>III89 -D97 DI<91380FFF8091B512F8010314FF010F158090263FFE0013C0 -D97FF8EB1FE0D9FFE0EB3FF04849EB7FF8484913FF4890C7FC5A5B121F5B003FED7FF0EE -3FE0007FED1FC093C7FC5BA212FFAC127F7FA2123FA26D153E121F6D157E6C167C6C6D14 -FC6C16F86C6D13036C01F0EB07F0D97FFCEB1FE06DB4EBFFC0010F90B5120001035C0100 -14F0020F13802F337CB137>IIIIII<13FCEA03FF487F487FA2487FA66C -5BA26C5B6C90C7FCEA00FC90C8FCACEB7FC0B5FCA512037EB3B3B61280A5194D7BCC22> -I108 D<90287FC001FFE0EC7F -F8B5010F01FC0103B5FC033F6D010F804B6D6C4814E0DBFE079026C03F817F9126C3F001 -9138FC007F0003D9C7E0DAE1F8806CDA8000D9F1E06D7E02CFC7EBF3C002DE91267FF780 -131F02FC03FFC77FA24A5DA24A5DA34A5DB3A9B6D8C03FB5D8F00FB512FCA55E317BB067 ->I<903A7FC001FFC0B5010F13F8033F13FE4B7F9126C1FE077F9126C3F0037F00039038 -C7C0016CD9CF8080150002DE7F02FC81A25CA25CA35CB3A9B6D8C07FEBFFE0A53B317BB0 -44>I<913807FF80027F13F80103B6FC010F15C090261FFE017F903A7FF0003FF8D9FFC0 -EB0FFC48496D7E4890C76C7E4817804980000F17C048486E13E0A2003F17F049157FA200 -7F17F8A400FF17FCAB007F17F8A36C6CEDFFF0A3001F17E06D5C000F17C06C6C4A13806C -17006C6D495A6C01E0EB1FFC6D6C495A903A3FFE01FFF0010FB612C0010392C7FCD9007F -13F80207138036337DB13D>I<90397FC00FFFB5017F13F002C1B512FC02C714FF9126CF -F80F7F9126FFC00313E0000391C77F6C01FC6E7E4A6E7E5C4A6E7E848319808319C0A471 -13E0AC4D13C0A319805FA219004D5A804D5A6E4A5A6E4A5A02FF495BDB80075B9126EFF0 -1F5B02E7B548C7FC02E314F802E014E0DB0FFEC8FC92CAFCAFB612C0A53B467CB044>I< -9039FF803FE0B5EBFFF8028113FE02837FDA87E11380EC8F830003D99F0713C06C139E14 -BCA214F8A24A6C13806F13004A6C5A93C7FCA45CB3A7B612E0A52A317CB032>114 -D<90390FFF8070017FEBF1F048B6FC1207380FFC01391FE0003F4848130F491307127F90 -C712035A1501A27FA213E06D90C7FC13FE387FFFF0ECFFC015F06C14FC6C14FF6C15806C -15C06C15E0C615F0013F14F8010714FCEB007F14019138003FFE150F0078140700F81403 -A26C1401A37E16FC6C14036D14F87F6DEB07F001F8EB1FE001FFEBFFC091B51280D8FC7F -1400D8F81F13FCD8E00313C027337CB130>I<14F8A61301A41303A21307A2130FA2131F -133F137F13FF1203000F90B512F0B7FCA426007FF8C7FCB3A6167CAA013F14F880A29039 -1FFE01F0010F1303903907FF87E06DEBFFC06D14806D6C1300EC0FFC26467EC430>I -III<007FB500C090387FFFE0A5C601F0C73803F8006E5D -017F5E6E1407013F5E6E140F011F5E6E141FA26D6D91C7FC5F6D153E6F137E6D157C6F13 -FC6D5DEDF0016D5DEDF803027F5C15FC1607DA3FFE5B160F021F5CEDFF1F6E91C8FC16BF -6E13BE16FE6E5BA36E5BA26E5BA26F5AA26F5AA26F5AA393C9FC5D153E157E157CD81F80 -13FC486C5B387FE001D8FFF05B14035D14074A5A49485A007F49CAFCEBC07E383F81FC6C -B45A6C5B6C13C0C648CBFC3B467EAF41>121 D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Ff cmtt12 17.28 6 -/Ff 6 123 df<913803FFC0023F13FC49B67E010715F04981013F15FE498190B812C048 -8348D9FC0180489026E0001F7F480180130391C87F48486F7E49153F4848ED0FFF834848 -178083496F13C012FF8319E07FA2187FA36C5A6C5A6C5ACBFCA218FFA219C05FA219805F -A24D13005F604D5A173F4D5A4D5AA24C5B4C5B4C5B041F90C7FC4C5A4C5A4C5A4B5B4B5B -4B5B031F5B4B90C8FC4B5AEDFFF84A5B4A5B4A5B021F5B4A90C9FCEC7FFC4A5A495B495B -010F5B495B4948CAFC4948ED1F804948ED3FC04849ED7FE0485B000F5B4890C9FC4890B8 -FC5ABAFCA56C18C06C18803B5A79D94A>50 D<383FFFF0487F80B5FCA37EA27EEA000FB0 -EE0FFC93B57E030714E0031F14F84B14FE92B7FC02FD8291B87E85DCE01F7FEE000703FC -01017F4B6D7F03E0143F4B6E7E4B140F8592C87E4A6F1380A34A6F13C0A284A21AE0A219 -7FAA19FFA21AC0A26E5DA24E138080606F1600606F4A5A6F143F6F4A5A6F4A5A6F130303 -FF010F5BDCC03F5B93B65A6102FD93C7FC02FC5D6F5C031F14F0902607F80714C0902603 -F00191C8FC90C8EA3FF043597FD74A>98 D105 D<003FB512FE4880B77EA57E7EC71201B3B3B3 -B0003FB812FC4817FEBAFCA56C17FE6C17FC385877D74A>108 D -112 D<000FB912E04818F04818F8A619F001F0C8000313E04D13C04D13804D13004D5A4D -5A4D5A6C484A5B6C484A5BC9000F5B4C5B4C90C7FC4C5A4C5A4B5B4B5B4B5B4B5B4B5B4B -90C8FC4B5A4B5A4A5B4A5B4A5B4A5B4A5B4A90C9FC4A5A4A5A495B495B495B4949EC07E0 -4949EC0FF04948C8EA1FF8495A495A485B485B485B485B4890C9FC485A48B9FCBAFCA66C -18F06C18E03D3E7BBD4A>122 D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fg cmbx12 17.28 28 -/Fg 28 120 df<16F04B7E1507151F153FEC01FF1407147F010FB5FCB7FCA41487EBF007 -C7FCB3B3B3B3007FB91280A6395E74DD51>49 D<913801FFF8021FEBFFC091B612F80103 -15FF010F16C0013F8290267FFC0114F89027FFE0003F7F4890C7000F7F48486E7FD807F8 -6E148048486E14C048486E14E048486F13F001FC17F8486C816D17FC6E80B56C16FE8380 -A219FFA283A36C5BA26C5B6C90C8FCD807FC5DEA01F0CA14FEA34D13FCA219F85F19F04D -13E0A294B512C019804C14004C5B604C5B4C5B604C13804C90C7FC4C5A4C5A4B13F05F4B -13804B90C8FC4B5AED1FF84B5A4B5A4B48143F4A5B4A48C8FC4A5A4A48157E4A5A4A5AEC -7F8092C9FC02FE16FE495A495A4948ED01FCD90FC0150749B8FC5B5B90B9FC5A4818F85A -5A5A5A5ABAFCA219F0A4405E78DD51>I<92B5FC020F14F8023F14FF49B712C04916F001 -0FD9C01F13FC90271FFC00077FD93FE001017F49486D8049C86C7F484883486C6F7F14C0 -486D826E806E82487FA4805CA36C5E4A5E6C5B6C5B6C495E011FC85A90C95CA294B55A61 -4C91C7FC604C5B4C5B4C5B4C5B047F138092260FFFFEC8FC020FB512F817E094C9FC17F8 -17FF91C7003F13E0040713F8040113FE707F717F7113E085717FA2717F85A285831A80A3 -1AC0EA03FCEA0FFF487F487F487FA2B57EA31A80A34D14005C7E4A5E5F6C495E49C8485B -D81FF85F000F5ED807FE92B55A6C6C6C4914806C01F0010791C7FC6C9026FF803F5B6D90 -B65A011F16F0010716C001014BC8FCD9001F14F0020149C9FC426079DD51>II<4DB5ED03C0057F02F0 -14070407B600FE140F047FDBFFC0131F4BB800F0133F030F05FC137F033F9127F8007FFE -13FF92B6C73807FF814A02F0020113C3020702C09138007FE74A91C9001FB5FC023F01FC -16074A01F08291B54882490280824991CB7E49498449498449498449865D49498490B5FC -484A84A2484A84A24891CD127FA25A4A1A3F5AA348491A1FA44899C7FCA25CA3B5FCB07E -A380A27EA2F50FC0A26C7FA37E6E1A1F6C1D80A26C801D3F6C6E1A00A26C6E616D1BFE6D -7F6F4E5A7F6D6D4E5A6D6D4E5A6D6D4E5A6D6E171F6D02E04D5A6E6DEFFF806E01FC4C90 -C7FC020F01FFEE07FE6E02C0ED1FF8020102F8ED7FF06E02FF913803FFE0033F02F8013F -1380030F91B648C8FC030117F86F6C16E004071680DC007F02F8C9FC050191CAFC626677 -E375>67 D72 DI77 -D80 D<001FBEFCA64849C79126E0000F148002E0180091 -C8171F498601F81A0349864986A2491B7FA2491B3F007F1DC090C9181FA4007E1C0FA600 -FE1DE0481C07A5CA95C7FCB3B3B3A3021FBAFCA663617AE070>84 -D<913803FFFE027FEBFFF00103B612FE010F6F7E4916E090273FFE001F7FD97FE001077F -D9FFF801017F486D6D7F717E486D6E7F85717FA2717FA36C496E7FA26C5B6D5AEB1FC090 -C9FCA74BB6FC157F0207B7FC147F49B61207010F14C0013FEBFE004913F048B512C04891 -C7FC485B4813F85A5C485B5A5CA2B55AA45FA25F806C5E806C047D7F6EEB01F96C6DD903 -F1EBFF806C01FED90FE114FF6C9027FFC07FC01580000191B5487E6C6C4B7E011F02FC13 -0F010302F001011400D9001F90CBFC49437CC14E>97 D<903807FF80B6FCA6C6FC7F7FB3 -A8EFFFF8040FEBFF80047F14F00381B612FC038715FF038F010014C0DBBFF0011F7FDBFF -C001077F93C76C7F4B02007F03F8824B6F7E4B6F13804B17C0851BE0A27313F0A21BF8A3 -7313FCA41BFEAE1BFCA44F13F8A31BF0A24F13E0A24F13C06F17804F1300816F4B5A6F4A -5B4AB402075B4A6C6C495B9126F83FE0013F13C09127F00FFC03B55A4A6CB648C7FCDAC0 -0115F84A6C15E091C7001F91C8FC90C8000313E04F657BE35A>I<92380FFFF04AB67E02 -0F15F0023F15FC91B77E01039039FE001FFF4901F8010113804901E0010713C049018049 -13E0017F90C7FC49484A13F0A2485B485B5A5C5A7113E0485B7113C048701380943800FE -0095C7FC485BA4B5FCAE7EA280A27EA2806C18FCA26C6D150119F87E6C6D15036EED07F0 -6C18E06C6D150F6D6DEC1FC06D01E0EC7F806D6DECFF00010701FCEB03FE6D9039FFC03F -FC010091B512F0023F5D020F1580020102FCC7FCDA000F13C03E437BC148>II<92380FFFC0 -4AB512FC020FECFF80023F15E091B712F80103D9FE037F499039F0007FFF011F01C0011F -7F49496D7F4990C76C7F49486E7F48498048844A804884485B727E5A5C48717EA35A5C72 -1380A2B5FCA391B9FCA41A0002C0CBFCA67EA380A27EA27E6E160FF11F806C183F6C7FF1 -7F006C7F6C6D16FE6C17016D6C4B5A6D6D4A5A6D01E04A5A6D6DEC3FE0010301FC49B45A -6D9026FFC01F90C7FC6D6C90B55A021F15F8020715E0020092C8FC030713F041437CC14A ->III<903807FF80B6FCA6C6FC7F7FB3A8EF1FFF94B512F0040714 -FC041F14FF4C8193267FE07F7F922781FE001F7FDB83F86D7FDB87F07FDB8FC0814C7F03 -9FC78015BE03BC8003FC825DA25DA25DA45DB3B2B7D8F007B71280A651647BE35A>II<903807FF80B6 -FCA6C6FC7F7FB3B3B3B3ADB712E0A623647BE32C>108 D<902607FF80D91FFFEEFFF8B6 -91B500F00207EBFF80040702FC023F14E0041F02FF91B612F84C6F488193267FE07F6D48 -01037F922781FE001F9027E00FF0007FC6DA83F86D9026F01FC06D7F6DD987F06D4A487F -6DD98FC0DBF87EC7804C6D027C80039FC76E488203BEEEFDF003BC6E4A8003FC04FF834B -5FA24B5FA24B94C8FCA44B5EB3B2B7D8F007B7D8803FB612FCA67E417BC087>I<902607 -FF80EB1FFFB691B512F0040714FC041F14FF4C8193267FE07F7F922781FE001F7FC6DA83 -F86D7F6DD987F07F6DD98FC0814C7F039FC78015BE03BC8003FC825DA25DA25DA45DB3B2 -B7D8F007B71280A651417BC05A>I<923807FFE092B6FC020715E0021F15F8027F15FE49 -4848C66C6C7E010701F0010F13E04901C001037F49496D7F4990C87F49486F7E49486F7E -48496F13804819C04A814819E048496F13F0A24819F8A348496F13FCA34819FEA4B518FF -AD6C19FEA46C6D4B13FCA36C19F8A26C6D4B13F0A26C19E06C6D4B13C0A26C6D4B13806C -6D4B13006D6C4B5A6D6D495B6D6D495B010701F0010F13E06D01FE017F5B010090B7C7FC -023F15FC020715E0020092C8FC030713E048437CC151>I114 D<913A3FFF8007800107B5EAF81F011FECFE7F017F91B5FC48B8FC48EBE0 -014890C7121FD80FFC1407D81FF0801600485A007F167F49153FA212FF171FA27F7F7F6D -92C7FC13FF14E014FF6C14F8EDFFC06C15FC16FF6C16C06C16F06C826C826C826C82013F -1680010F16C01303D9007F15E0020315F0EC001F1500041F13F81607007C150100FC8117 -7F6C163FA2171F7EA26D16F0A27F173F6D16E06D157F6D16C001FEEDFF806D0203130002 -C0EB0FFE02FCEB7FFC01DFB65A010F5DD8FE0315C026F8007F49C7FC48010F13E035437B -C140>II<90 -2607FFC0ED3FFEB60207B5FCA6C6EE00076D826D82B3B3A260A360A2607F60183E6D6D14 -7E4E7F6D6D4948806D6DD907F0ECFF806D01FFEB3FE06D91B55A6E1500021F5C020314F8 -DA003F018002F0C7FC51427BC05A>I119 D -E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fh cmsy10 10.95 1 -/Fh 1 16 df15 -D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fi cmtt10 10.95 89 -/Fi 89 127 df<121C127FEAFF80B3EA7F00B2123EC7FCA8121C127FA2EAFF80A3EA7F00 -A2121C09396DB830>33 D<00101304007C131F00FEEB3F80A26C137FA248133FB2007E14 -00007C7F003C131E00101304191C75B830>I<903907C007C0A2496C487EA8011F131FA2 -02C05BA3007FB7FCA2B81280A36C16006C5D3A007F807F80A2020090C7FCA9495BA2003F -90B512FE4881B81280A36C1600A22701FC01FCC7FCA300031303A201F85BA76C486C5AA2 -29387DB730>I38 DI<141E147F14FF5BEB03FEEB07FCEB0FF0EB1FE0EB3FC0EB7F80EBFF00 -485A5B12035B485A120F5BA2485AA2123F5BA2127F90C7FCA412FEAD127FA47F123FA27F -121FA26C7EA27F12076C7E7F12017F6C7EEB7F80EB3FC0EB1FE0EB0FF0EB07FCEB03FEEB -01FF7F147F141E184771BE30>I<127812FE7E7F6C7E6C7EEA0FF06C7E6C7E6C7E6C7EEB -7F80133F14C0131FEB0FE014F01307A2EB03F8A214FC1301A214FE1300A4147FAD14FEA4 -130114FCA2130314F8A2EB07F0A2130F14E0EB1FC0133F1480137FEBFF00485A485A485A -485AEA3FE0485A485A90C7FC5A1278184778BE30>I<14E0497E497EA60038EC0380007E -EC0FC0D8FF83EB3FE001C3137F9038F3F9FF267FFBFB13C06CB61280000FECFE00000314 -F86C5C6C6C13C0011F90C7FC017F13C048B512F04880000F14FE003FECFF80267FFBFB13 -C026FFF3F913E09038C3F87F0183133FD87E03EB0FC00038EC0380000091C7FCA66D5A6D -5A23277AAE30>I<143EA2147FAF007FB7FCA2B81280A36C1600A2C76CC8FCAF143EA229 -297DAF30>II<007FB612F0 -A2B712F8A36C15F0A225077B9E30>I<120FEA3FC0EA7FE0A2EAFFF0A4EA7FE0A2EA3FC0 -EA0F000C0C6E8B30>I<16F01501ED03F8A21507A2ED0FF0A2ED1FE0A2ED3FC0A2ED7F80 -A2EDFF00A24A5AA25D1403A24A5AA24A5AA24A5AA24A5AA24A5AA24AC7FCA2495AA25C13 -03A2495AA2495AA2495AA2495AA2495AA249C8FCA2485AA25B1203A2485AA2485AA2485A -A2485AA2485AA248C9FCA25AA2127CA225477BBE30>I<14FE903807FFC0497F013F13F8 -497F90B57E48EB83FF4848C6138049137F4848EB3FC04848EB1FE049130F001F15F04913 -07A24848EB03F8A290C712014815FCA400FEEC00FEAD6C14016C15FCA36D1303003F15F8 -A26D1307001F15F0A26D130F6C6CEB1FE0A26C6CEB3FC06C6CEB7F806D13FF2601FF8313 -006CEBFFFE6D5B6D5B010F13E06D5BD900FEC7FC273A7CB830>IIIII<000FB612804815C05AA316800180C8FCAEEB83FF019F13C0 -90B512F015FC8181D9FE0313809039F0007FC049133F0180EB1FE06CC7120F000E15F0C8 -1207A216F81503A31218127EA2B4FC150716F048140F6C15E06C141F6DEB3FC06D137F3A -3FE001FF80261FFC0F13006CB55A6C5C6C5C6C14E06C6C1380D90FFCC7FC25397BB730> -II<127CB712FC16FEA416FC48C7EA0FF816F0ED1FE0007CEC3FC0C8EA7F80EDFF -00A24A5A4A5A5D14075D140F5D4A5AA24A5AA24AC7FCA25C5C13015CA213035CA213075C -A4495AA6131F5CA96D5A6DC8FC273A7CB830>I<49B4FC011F13F0017F13FC90B57E0003 -ECFF804815C048010113E03A1FF8003FF049131FD83FC0EB07F8A24848EB03FC90C71201 -A56D1303003F15F86D13076C6CEB0FF06C6CEB1FE0D807FCEB7FC03A03FF83FF806C90B5 -12006C6C13FC011F13F0497F90B512FE48802607FE0013C0D80FF8EB3FE0D81FE0EB0FF0 -4848EB07F8491303007F15FC90C712014815FE481400A66C14016C15FC6D1303003F15F8 -6D1307D81FF0EB1FF06D133F3A0FFF01FFE06C90B512C06C1580C6ECFE006D5B011F13F0 -010190C7FC273A7CB830>I<49B4FC010F13E0013F13F890B57E4880488048010113803A -0FFC007FC0D81FF0EB3FE04848131F49EB0FF048481307A290C7EA03F85A4815FC1501A4 -16FEA37E7E6D1303A26C6C13076C6C130F6D133FD80FFC13FF6CB6FC7E6C14FE6C14F901 -3FEBE1FC010F138190380060011400ED03F8A2150716F0150F000F15E0486C131F486CEB -3FC0157FEDFF804A1300EC07FE391FF01FFC90B55A6C5C6C5C6C1480C649C7FCEB3FF027 -3A7CB830>I<120FEA3FC0EA7FE0A2EAFFF0A4EA7FE0A2EA3FC0EA0F00C7FCAF120FEA3F -C0EA7FE0A2EAFFF0A4EA7FE0A2EA3FC0EA0F000C276EA630>II<16F01503ED07F8151F157FEDFFF014034A13C0021F138091383FFE00ECFFF849 -5B010713C0495BD93FFEC7FC495A3801FFF0485B000F13804890C8FCEA7FFC5BEAFFE05B -7FEA7FF87FEA1FFF6C7F000313E06C7F38007FFC6D7E90380FFF806D7F010113F06D7FEC -3FFE91381FFF80020713C06E13F01400ED7FF8151F1507ED03F01500252F7BB230>I<00 -7FB7FCA2B81280A36C16006C5DCBFCA7003FB612FE4881B81280A36C1600A229157DA530 ->I<1278127EB4FC13C07FEA7FF813FEEA1FFF6C13C000037F6C13F86C6C7EEB1FFF6D7F -010313E06D7F9038007FFC6E7E91380FFF806E13C0020113F080ED3FF8151F153FEDFFF0 -5C020713C04A138091383FFE004A5A903801FFF0495B010F13804990C7FCEB7FFC48485A -4813E0000F5B4890C8FCEA7FFE13F8EAFFE05B90C9FC127E1278252F7BB230>I64 -D<147F4A7EA2497FA4497F14F7A401077F14E3A3010F7FA314C1A2011F7FA490383F80FE -A590387F007FA4498049133F90B6FCA34881A39038FC001F00038149130FA40007814913 -07A2D87FFFEB7FFFB56CB51280A46C496C130029397DB830>I<007FB512F0B612FE6F7E -82826C813A03F8001FF815076F7E1501A26F7EA615015EA24B5A1507ED1FF0ED7FE090B6 -5A5E4BC7FC6F7E16E0829039F8000FF8ED03FC6F7E1500167FA3EE3F80A6167F1700A25E -4B5A1503ED1FFC007FB6FCB75A5E16C05E6C02FCC7FC29387EB730>I<91387F803C9039 -03FFF03E49EBFC7E011F13FE49EBFFFE5B9038FFE07F48EB801F3903FE000F484813075B -48481303A2484813015B123F491300A2127F90C8FC167C16005A5AAC7E7EA2167C6D14FE -123FA27F121F6D13016C6C14FCA26C6CEB03F86D13076C6CEB0FF03901FF801F6C9038E0 -7FE06DB512C06D14806D1400010713FC6D13F09038007FC0273A7CB830>I<003FB512E0 -4814FCB67E6F7E6C816C813A03F8007FF0ED1FF8150F6F7E6F7E15016F7EA2EE7F80A216 -3F17C0161FA4EE0FE0AC161F17C0A3163F1780A2167F17005E4B5A15034B5A150F4B5AED -7FF0003FB65A485DB75A93C7FC6C14FC6C14E02B387FB730>I<007FB7FCB81280A47ED8 -03F8C7123FA8EE1F0093C7FCA4157C15FEA490B5FCA6EBF800A4157C92C8FCA5EE07C0EE -0FE0A9007FB7FCB8FCA46C16C02B387EB730>I<003FB712804816C0B8FCA27E7ED801FC -C7121FA8EE0F8093C7FCA5153E157FA490B6FCA69038FC007FA4153E92C8FCAE383FFFF8 -487FB5FCA27E6C5B2A387EB730>I<02FF13F00103EBC0F8010F13F1013F13FD4913FF90 -B6FC4813C1EC007F4848133F4848131F49130F485A491307121F5B123F491303A2127F90 -C7FC6F5A92C8FC5A5AA892B5FC4A14805CA26C7F6C6D1400ED03F8A27F003F1407A27F12 -1F6D130F120F7F6C6C131FA2D803FE133F6C6C137FECC1FF6C90B5FC7F6D13FB010F13F3 -0103EBC1F0010090C8FC293A7DB830>I<3B3FFF800FFFE0486D4813F0B56C4813F8A26C -496C13F06C496C13E0D803F8C7EAFE00B290B6FCA601F8C7FCB3A23B3FFF800FFFE0486D -4813F0B56C4813F8A26C496C13F06C496C13E02D387FB730>I<007FB6FCB71280A46C15 -00260007F0C7FCB3B3A8007FB6FCB71280A46C1500213879B730>I<49B512F04914F85B -A27F6D14F090C7EAFE00B3B3123C127EB4FCA24A5A1403EB8007397FF01FF86CB55A5D6C -5C00075C000149C7FC38003FF025397AB730>II<383FFFF8487FB57EA26C5B6C -5BD801FCC9FCB3B0EE0F80EE1FC0A9003FB7FC5AB8FCA27E6C16802A387EB730>III< -90383FFFE048B512FC000714FF4815804815C04815E0EBF80001E0133FD87F80EB0FF0A2 -90C71207A44815F8481403B3A96C1407A26C15F0A36D130FA26D131F6C6CEB3FE001F813 -FF90B6FC6C15C06C15806C1500000114FCD8003F13E0253A7BB830>I<007FB512F0B612 -FE6F7E16E0826C813903F8003FED0FFCED03FE15016F7EA2821780163FA6167F17005EA2 -4B5A1503ED0FFCED3FF890B6FC5E5E16804BC7FC15F001F8C9FCB0387FFFC0B57EA46C5B -29387EB730>I<90383FFFE048B512FC000714FF4815804815C04815E0EBF80001E0133F -4848EB1FF049130F90C71207A44815F8481403B3A8147E14FE6CEBFF076C15F0EC7F87A2 -EC3FC7018013CF9038C01FFFD83FE014E0EBF80F90B6FC6C15C06C15806C1500000114FC -D8003F7FEB00016E7EA21680157F16C0153F16E0151F16F0150FED07E025467BB830>I< -003FB57E4814F0B612FC15FF6C816C812603F8017F9138003FF0151F6F7E150715038215 -01A515035E1507150F4B5A153F4AB45A90B65A5E93C7FC5D8182D9F8007FED3FE0151F15 -0F821507A817F8EEF1FCA53A3FFF8003FB4801C0EBFFF8B56C7E17F06C496C13E06C49EB -7FC0C9EA1F002E397FB730>I<90390FF803C0D97FFF13E048B512C74814F74814FF5A38 -1FF80F383FE001497E4848137F90C7123F5A48141FA2150FA37EED07C06C91C7FC7F7FEA -3FF0EA1FFEEBFFF06C13FF6C14E0000114F86C80011F13FF01031480D9003F13C0140191 -38007FE0151FED0FF0A2ED07F8A2007C140312FEA56C140716F07F6DEB0FE06D131F01F8 -EB3FC001FF13FF91B51280160000FD5CD8FC7F13F8D8F81F5BD878011380253A7BB830> -I<003FB712C04816E0B8FCA43AFE003F800FA8007CED07C0C791C7FCB3B1011FB5FC4980 -A46D91C7FC2B387EB730>I<3B7FFFC007FFFCB56C4813FEA46C496C13FCD803F8C7EA3F -80B3B16D147F00011600A36C6C14FE6D13016D5CEC800390393FE00FF890391FF83FF06D -B55A6D5C6D5C6D91C7FC9038007FFCEC1FF02F3980B730>III<3A3FFF01FFF84801837F02C77FA202835B6C0101 -5B3A01FC007F806D91C7FC00005C6D5BEB7F01EC81FCEB3F8314C3011F5B14E7010F5B14 -FF6D5BA26D5BA26D5BA26D90C8FCA4497FA2497FA2815B81EB0FE781EB1FC381EB3F8181 -EB7F0081497F49800001143F49800003141F49800007140FD87FFEEB7FFFB590B5128080 -A25C6C486D130029387DB730>II<001FB612FC4815FE5AA490C7EA03FCED07F816 -F0150FED1FE016C0153FED7F80003E1500C85A4A5A5D14034A5A5D140F4A5A5D143F4A5A -92C7FC5C495A5C1303495A5C130F495A5C133F495A91C8FC5B4848147C4914FE1203485A -5B120F485A5B123F485A90B6FCB7FCA46C15FC27387CB730>I<007FB5FCB61280A41500 -48C8FCB3B3B3A5B6FC1580A46C140019476DBE30>I<007FB5FCB61280A47EC7123FB3B3 -B3A5007FB5FCB6FCA46C140019477DBE30>93 D<1307EB1FC0EB7FF0497E000313FE000F -EBFF80003F14E0D87FFD13F039FFF07FF8EBC01FEB800F38FE0003007CEB01F00010EB00 -401D0E77B730>I<007FB612F0A2B712F8A36C15F0A225077B7D30>I97 DII<913801FFE0 -4A7F5CA28080EC0007AAEB03FE90381FFF874913E790B6FC5A5A481303380FFC00D81FF0 -133F49131F485A150F4848130790C7FCA25AA25AA87E6C140FA27F003F141F6D133F6C7E -6D137F390FF801FF2607FE07EBFFC06CB712E06C16F06C14F76D01C713E0011F010313C0 -D907FCC8FC2C397DB730>I<49B4FC010713E0011F13F8017F7F90B57E48804801811380 -3A07FC007FC04848133FD81FE0EB1FE0150F484814F0491307127F90C7FCED03F85A5AB7 -FCA516F048C9FC7E7EA27F003FEC01F06DEB03F86C7E6C7E6D1307D807FEEB1FF03A03FF -C07FE06C90B5FC6C15C0013F14806DEBFE00010713F8010013C0252A7CA830>IIII<14E0EB03F8A2497EA36D5AA2EB00E091C8FCA9381FFFF8487F5AA27E7EEA0001B3 -A9003FB612C04815E0B7FCA27E6C15C023397AB830>III<387FFFF8B57EA47EEA0001B3B3A8007FB612F0B712F8A46C15F025 -387BB730>I<02FC137E3B7FC3FF01FF80D8FFEF01877F90B500CF7F15DF92B57E6C010F -13872607FE07EB03F801FC13FE9039F803FC01A201F013F8A301E013F0B3A23C7FFE0FFF -07FF80B548018F13C0A46C486C01071380322881A730>II<49B4FC010F13E0013F13F8497F90B57E0003ECFF8014013A07FC007FC04848EB3F -E0D81FE0EB0FF0A24848EB07F8491303007F15FC90C71201A300FEEC00FEA86C14016C15 -FCA26D1303003F15F86D13076D130F6C6CEB1FF06C6CEB3FE06D137F3A07FF01FFC06C90 -B512806C15006C6C13FC6D5B010F13E0010190C7FC272A7CA830>II<49B413F8010FEBC1FC013F13F14913FD48B6FC -5A481381390FFC007F49131F4848130F491307485A491303127F90C7FC15015A5AA77E7E -15037FA26C6C1307150F6C6C131F6C6C133F01FC137F3907FF01FF6C90B5FC6C14FD6C14 -F9013F13F1010F13C1903803FE0190C7FCAD92B512F84A14FCA46E14F82E3C7DA730>I< -ED07F83A3FFF803FFF486DB51280B512C302CF14C06C13DF6C9038FFFC3FD8001F13E092 -38801F809238000F004A90C7FC5C5C5CA25CA45CAF003FB512FC4880B7FCA26C5C6C5C2A -287EA730>I<90381FFC1E48B5129F000714FF5A5A5A387FF007EB800100FEC7FC4880A4 -6C143E007F91C7FC13E06CB4FC6C13FC6CEBFF806C14E0000114F86C6C7F01037F903800 -0FFF02001380007C147F00FEEC1FC0A2150F7EA27F151F6DEB3F806D137F9039FC03FF00 -90B6FC5D5D00FC14F0D8F83F13C026780FFEC7FC222A79A830>II -I<3B3FFFC07FFF80486DB512C0B515E0A26C16C06C496C13803B01F80003F000A26D1307 -00005DA26D130F017E5CA2017F131F6D5CA2EC803F011F91C7FCA26E5A010F137EA2ECE0 -FE01075BA214F101035BA3903801FBF0A314FF6D5BA36E5A6E5A2B277EA630>I<3B3FFF -C01FFFE0486D4813F0B515F8A26C16F06C496C13E0D807E0C7EA3F00A26D5C0003157EA5 -6D14FE00015DEC0F80EC1FC0EC3FE0A33A00FC7FF1F8A2147DA2ECFDF9017C5C14F8A301 -7E13FBA290393FF07FE0A3ECE03FA2011F5C90390F800F802D277FA630>I<3A3FFF81FF -FC4801C37FB580A26C5D6C01815BC648C66CC7FC137FEC80FE90383F81FC90381FC3F8EB -0FE3ECE7F06DB45A6D5B7F6D5B92C8FC147E147F5C497F81903803F7E0EB07E790380FE3 -F0ECC1F890381F81FC90383F80FE90387F007E017E137F01FE6D7E48486D7E267FFF80B5 -FCB500C1148014E3A214C16C0180140029277DA630>I<3B3FFFC07FFF80486DB512C0B5 -15E0A26C16C06C496C13803B01FC0003F000A2000014076D5C137E150F017F5C7F151FD9 -1F805BA214C0010F49C7FCA214E00107137EA2EB03F0157C15FCEB01F85DA2EB00F9ECFD -F0147D147FA26E5AA36E5AA35DA2143F92C8FCA25C147EA2000F13FE486C5AEA3FC1EBC3 -F81387EB8FF0EBFFE06C5B5C6C90C9FC6C5AEA01F02B3C7EA630>I<001FB612FC4815FE -5AA316FC90C7EA0FF8ED1FF0ED3FE0ED7FC0EDFF80003E491300C7485A4A5A4A5A4A5A4A -5A4A5A4A5A4990C7FC495A495A495A495A495A495A4948133E4890C7127F485A485A485A -485A485A48B7FCB8FCA46C15FE28277DA630>II<127CA212FEB3B3B3AD127CA207476CBE30>II<01 -7C133848B4137C48EB80FE4813C14813C348EBEFFC397FEFFFF0D8FF8713E0010713C048 -6C1380D87C0113003838007C1F0C78B730>I E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fj cmb10 10.95 87 -/Fj 87 125 df11 DII<913B07FF8003FF80027FD9E03F13E049B56CB512F8010702FB80 -011F0103B5EA01FE90263FF8019038F800FF90267FE003495A4948484948138002804A5A -48010014805D481600497FA26F6E13006F6E5AF0007896C7FCA6F17F80BBFCA50003D900 -01EB800384B3ABB5D8F83FD9FC3F13FEA547407EBF4C>I33 DI37 DII<147814F81301EB03F0EB07E0EB0FC0A2EB1F80133FEB7F00A213FE -A2485A1203A25B1207A2485AA3121F5BA2123FA4485AA612FFB3A2127FA66C7EA4121FA2 -7F120FA36C7EA212037FA212016C7EA2137FA2EB3F80131FEB0FC0A2EB07E0EB03F0EB01 -F813001478155A78C323>I<127012F87E127E7E6C7EA26C7E7F6C7EA26C7EA26C7E7FA2 -12007FA2EB7F80A314C0133FA214E0A4EB1FF0A614F8B3A214F0A6EB3FE0A414C0A2137F -1480A3EBFF00A25B1201A25B485AA2485AA2485A5B485AA248C7FC127E5A5A1270155A7B -C323>II44 -DIIII<143C147CEB01FC1307131FEA03FFB5FCA4EAFC -1F1200B3B3A8007FB6FCA5203C7ABB2D>IIII<000E1407D80F -80133F9038F801FF90B6FC5D5DA25D5D5D158092C7FC14FC14F00180C8FCA9EB87FE9038 -9FFFC090B512F0819038FC0FFC9038F007FE9038C003FF0180148016C0497EC714E0A416 -F0A21207EA1FC0EA3FE0EA7FF012FF13F8A316E013F0A26C484813C01380D87E0014806C -491300391FC00FFE390FF03FFC6CB55A6C5C6C14C06C91C7FCEB1FF8243D7CBB2D>II<121F7F13F090B612FCA45A16F816F016E0A216C016805A -007EC7EA3F00157E007C147C15FC4A5A4A5A485C14074A5AC75B141F4AC7FCA25C14FEA2 -1301A2495AA31307A2130FA25C131FA3133FA6137FAA6D5A6D5A6D5A263F7BBD2D>IIIII<007FB912E0BA12F0A4003F18E0CDFCB0003FB912E0BA12F0A46C -18E03C1C7BA447>61 D63 D65 -DIIIIIIII<010FB61280A5D90001 -EBE000B3B3A4EA1FE0EA3FF0EA7FF8A2EAFFFCA44A5BA26C48485B01F091C7FC6C48485A -391FF03FFC6CB55A6C5C000114C026001FFCC8FC293F81BD2F>IIIII<913801FFC0021F13FC91B67E01 -0315E04901807F903A1FFE003FFCD93FF8EB0FFE49486D7E49486D7F48496D7F48834A7F -48834890C86C7EA2488349153FA2003F83A249151F007F83A400FF1880AE007F1800A36D -5DA2003F5FA36C6C4B5AA26C5F6E14FF6C5F6C6D495B6E5B6C5F6C6D495B6D6C4990C7FC -D93FFEEB3FFE6D6C6CB45A010790B512F06D5D01001580021F01FCC8FC020113C039407B -BE44>II<913801FFC0021F13FC91B67E010315E04901807F903A1FFE003FFCD93FF8EB -0FFE49486D7E49486D7F48496D7F48496D7FA24890C86C7E488349153F001F83A249151F -003F83A3007F834981A300FF1880AE007F1800A46D5D003F5FA36C6C4B5AA2000F027F5C -903AFF01FF807F6C4901E05B6C01876D485A15C16C9026CF80795B6C9026EF007F5B6DB4 -6D90C7FC6D5D6D01805B01079038C0FFF06D90B55A0100ED8001021FED03800201EBCFC0 -91C7EA0FE01807EFF81F94B5FCA3701400A4705BA2705B60705B715AEF1FC039517BBE44 ->III<003FB812FCA5D9FC03EBE01FD87FE0ED07FE01 -C01501018015001300007E177EA3007C173EA400FC173F48171FA5C71600B3AF013FB612 -FEA5383D7DBC3F>IIII<003FB5D8FC0F -B512F0A5D8003F90C7380FC0006D6D495A4DC7FC6D7F6D6D137E5F6D7F4C5A6D6D485A6D -13FC4C5ADA7FFE5B160F6E6C485A6E139F04BFC8FC6E13FEA26E5B6E5BA28082806F7EA2 -6F7E83A24B7F92B5FC83DA01FB7FEC03F303F17FEC07E1DA0FC07F707EEC1F804B6C7E5C -027E6D7F707F5C49486D7FA249486D7F49486D7FA249486D7F49486E7EA2B6011FB6FCA5 -403E7EBD45>II<003FB712F8A5DAC00313F09026FC000713E05B01E04913C05B49491380127F90 -C74813004B5A127E4B5AA24A5B127C4A5B5C5E4A5BC7FC4A5BA24A90C7FC5C5D4A5AA249 -5BA2495B5B5D4949137CA2495BA24990C7FC4915FC4A14F8495AA2485B1601485BA24849 -1303485B16074849130F161F4890C7123FEEFFF04848130FB8FCA52E3E7BBD38>II<0160130301F0EB07800001140F0003 -141FD807C0EB3E00A248485B48C75A001E5C003E1301003C5C007C1303A200785CA200F8 -1307485CA2D8F7E013BFD8FFF0EBFF8001F814C001FC14E001FE14F0A3007F7FA36C486C -13E06C486C13C06C48EB7F806C48EB3F00242077BE31>II97 D<13FFB5FCA512077EAFEC03FE -91381FFFC0027F13F091B57E9138FE0FFE9138F003FF4A7E02C014804A6C13C017E0A3EE -7FF0A317F8AC17F0A3EEFFE0A217C05D02C014806E4813006E485A9138FC1FFC01FCB55A -496C13E0D9F01F1380C7D807FCC7FC2D407EBE33>III<49B47E010F13F0013F7F90B512FE48EBC3 -FF48010013804848EB7FC04848133F001F15E05B003FEC1FF0A2485A150F16F8A212FFA2 -90B6FCA401F0C8FCA5127FA37F003F15F8A26C6C1301000F15F06D13036C6CEB07E06C90 -38800FC06C9038F07F806C6CB512006D5B010F13F8010013C0252B7EA92A>III<13FFB5FCA512077EAFED7F -E0913801FFF802077F4A7F91381FC3FFDA3E031380147CEC780102F014C014E0A214C0A3 -1480B3A4B5D8FE1F13FFA5303F7EBE33>III<13FFB5FCA512077EB092B512E0A592380FE000 -4B5A4B5A4BC7FC15FE4A5A4A5A4A5A4A5A4A5A143FECFFE0A28181A2ECDFFCEC8FFE140F -6E7E6E7FA26E7F6E7FA26F7E6F7EA26F7E6F7EB539FC7FFFF8A52D3F7FBE30>I<13FFB5 -FCA512077EB3B3AFB512FCA5163F7EBE19>I<01FFD97FE0EB3FF0B52601FFF8EBFFFC02 -07D9FC037F4A6D487F91281FC3FF0FE17FDA3E03D99F017F0007017C14BE6CD97801EBBC -0002F002F88002E05CA202C05CA302805CB3A4B5D8FE1FD9FF0FEBFF80A549297EA84C> -I<01FFEB7FE0B53801FFF802077F4A7F91381FC3FFDA3E0313800007137C6CEB780102F0 -14C014E0A214C0A31480B3A4B5D8FE1F13FFA530297EA833>I<49B47E010F13F0013F13 -FC90B6FC48018113803A03FE007FC04848EB3FE0000F15F049131F001F15F8A24848EB0F -FCA2007F15FEA400FF15FFAB007F15FEA3003F15FC6D131F001F15F8A26C6CEB3FF00007 -15E06C6CEB7FC03A01FF81FF806C90B51200013F13FC010F13F001011380282B7EA92D> -I<9038FF03FEB5381FFFC0027F13F091B57E9138FE1FFE9138F007FF0007497E6C01C014 -804A6C13C017E081A217F0A2167F17F8ACEEFFF0A317E05D17C0A26E4813806E4813006E -485A9138FC1FFC91B55A16E0029F1380DA87FCC7FC0280C8FCACB512FEA52D3B7EA833> -I<49B4131E010FEBC03E013FEBE07E90B5EAF0FE48EBC1F948EB807F48EB003F485A4848 -131FA2003F140F5B127F1507A2485AAC127F7FA2123F150F6C7E151F6C7E0007143F6C6C -13FF6C13C36C90B5FC6D13EF011F138F903803FE0F90C7FCAC0203B512F8A52D3B7DA830 ->I<3901FE07F800FFEB0FFE91383FFF804A13C0EC7C7F9138F8FFE03807FFF06C5B5CA2 -ED7FC09138803F80ED1F0092C7FCA291C8FCB3A3B6FCA523297FA827>I<90387FE1E038 -03FFFB4813FF5A381FE07F383F801F387F000F007E130712FE1403A27EA26DC7FC13F013 -FF6C13F014FC6C13FF15806C14C07E6C14E0000114F06C7E010313F8EB003F140F007813 -0712F814037EA36C14F06C1307A29038800FE09038F03FC090B51280150000F813FC38E0 -3FF01D2B7DA924>I<131FA65BA55BA25BA25A5A5A001FEBFFC0B6FCA4000790C7FCB3EC -03E0A97EEC87C0A26CEBCF806C13FF6D1300EB1FFEEB07F81B3B7EB923>IIII<3B7FFFFC7FFFE0A5C69039E007E0004B5A -90387FF01FD93FF85B4BC7FC90381FFC7EEB0FFE5D6D6C5A7F5D6D5B7FA26E7E143F814A -7E14FF81497FEB03F301078002E17FEB0FC049487F6F7E90383F003F017E806F7EB500E0 -B512F8A52D287FA730>II<001FB61280A4D9F80F1300EBE01F01C05B49485A130048495A4A5AA2D83E01 -5B5B5D495BA2C64890C7FC5B5C495A017FEB0F8014F8EBFFF0A248EBE01F48150014C048 -1380A248495A485A5D48485B007F5B9038F80FFEB6FCA421287EA728>III E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fk cmbx12 14.4 49 -/Fk 49 122 df12 D45 DI<913803FFC0023F13FC91B6FC010315 -C0010F018113F0903A1FFC003FF849486D7E49486D7E49486D7E48496D138048496D13C0 -A24817E04890C813F0A34817F8A24817FC49157FA3007F17FEA600FF17FFB3A5007F17FE -A6003F17FCA26D15FFA26C17F8A36C17F0A26C6D4913E0A26C6D4913C06C17806E5B6C6D -4913006D6C495AD91FFCEB3FF8903A0FFF81FFF06D90B55A01011580D9003F01FCC7FC02 -0313C0384F7BCD43>48 D<157815FC14031407141F14FF130F0007B5FCB6FCA2147F13F0 -EAF800C7FCB3B3B3A6007FB712FEA52F4E76CD43>I -I<91380FFFC091B512FC0107ECFF80011F15E090263FF8077F9026FF800113FC4848C76C -7ED803F86E7E491680D807FC8048B416C080486D15E0A4805CA36C17C06C5B6C90C75AD8 -01FC1680C9FC4C13005FA24C5A4B5B4B5B4B13C04B5BDBFFFEC7FC91B512F816E016FCEE -FF80DA000713E0030113F89238007FFE707E7013807013C018E07013F0A218F8A27013FC -A218FEA2EA03E0EA0FF8487E487E487EB57EA318FCA25E18F891C7FC6C17F0495C6C4816 -E001F04A13C06C484A1380D80FF84A13006CB44A5A6CD9F0075BC690B612F06D5D011F15 -80010302FCC7FCD9001F1380374F7ACD43>I<177C17FEA2160116031607160FA2161F16 -3F167FA216FF5D5DA25D5DED1FBFED3F3F153E157C15FCEC01F815F0EC03E01407EC0FC0 -1580EC1F005C147E147C5C1301495A495A5C495A131F49C7FC133E5B13FC485A5B485A12 -07485A485A90C8FC123E127E5ABA12C0A5C96C48C7FCAF020FB712C0A53A4F7CCE43>I< -D80380150ED807E0157E01FEEC03FED9FFF0137F91B65A5F5F5F5F5F94C7FC5E5E16F016 -C093C8FC15F801E190C9FC01E0CAFCABEC0FFF027F13F001E3B512FE01E76E7E9026FFF8 -077FDAC0017F49C713F8496E7E49143F4981496E7E6C481680C9FC18C08218E0A418F0A3 -EA0FE0487E487E487E487EA418E0A35B6C484A13C05B491680003EC85A003F17006C6C4A -5A6D5D6C6C4A5AD807F8495BD803FE01075B2701FFC03F5B6C90B65A013F4AC7FC6D14F8 -010314C09026007FF8C8FC344F79CD43>II<121F7F7FEBFF8091B81280A45A1900606060A260 -6060485F0180C86CC7FC007EC95A4C5A007C4B5A5F4C5A160F4C5A484B5A4C5A94C8FC16 -FEC812014B5A5E4B5A150F4B5AA24B5AA24B5A15FFA24A90C9FCA25C5D1407A2140FA25D -141FA2143FA4147F5DA314FFA55BAC6D5BA2EC3FC06E5A395279D043>I<913807FFC002 -7F13FC0103B67E010F15E090261FFC0113F8903A3FE0003FFCD97F80EB0FFE49C76C7E48 -488048486E1380000717C04980120F18E0177FA2121F7FA27F7F6E14FF02E015C014F802 -FE4913806C7FDBC00313009238F007FE6C02F85B9238FE1FF86C9138FFBFF06CEDFFE017 -806C4BC7FC6D806D81010F15E06D81010115FC010781011F81491680EBFFE748018115C0 -48D9007F14E04848011F14F048487F48481303030014F8484880161F4848020713FC1601 -824848157F173FA2171FA2170FA218F8A27F007F17F06D151FA26C6CED3FE0001F17C06D -157F6C6CEDFF806C6C6C010313006C01E0EB0FFE6C01FCEBFFFC6C6CB612F06D5D010F15 -80010102FCC7FCD9000F13C0364F7ACD43>I<91B5FC010F14F8017F14FF90B712C00003 -D9C00F7F2707FC00017FD80FE06D7F48486E7E48C87FD87FE06E7E7F7F486C1680A66C5A -18006C485C6C5AC9485A5F4B5B4B5B4B5B4B5B4B90C7FC16FC4B5A4B5A16C04B5A93C8FC -4A5A5D14035D5D14075DA25D140FA25DAB91CAFCAAEC1FC04A7EECFFF8497FA2497FA76D -5BA26D5BEC3FE06E5A315479D340>63 D68 DII72 -DI<027FB71280A591C76C -90C7FCB3B3B3EA07F0EA1FFC487E487EA2B57EA44C5AA34A485B7E49495BD83FF8495BD8 -1FE05DD80FFC011F5B2707FF807F90C8FC000190B512FC6C6C14F0011F14C0010101F8C9 -FC39537DD145>I76 -DI80 D82 D<91260FFF80130791B500F85B010702FF5B011FEDC0 -3F49EDF07F9026FFFC006D5A4801E0EB0FFD4801800101B5FC4848C87E48488149150F00 -1F824981123F4981007F82A28412FF84A27FA26D82A27F7F6D93C7FC14C06C13F014FF15 -F86CECFF8016FC6CEDFFC017F06C16FC6C16FF6C17C06C836C836D826D82010F82130301 -0082021F16801400030F15C0ED007F040714E01600173F050F13F08383A200788200F882 -A3187FA27EA219E07EA26CEFFFC0A27F6D4B13806D17006D5D01FC4B5A01FF4B5A02C04A -5A02F8EC7FF0903B1FFFC003FFE0486C90B65AD8FC0393C7FC48C66C14FC48010F14F048 -D9007F90C8FC3C5479D24B>I<003FBC1280A59126C0003F9038C0007F49C71607D87FF8 -060113C001E08449197F49193F90C8171FA2007E1A0FA3007C1A07A500FC1BE0481A03A6 -C994C7FCB3B3AC91B912F0A553517BD05E>II87 D97 DI<913801FFF8021FEBFF8091B612F00103 -15FC010F9038C00FFE903A1FFE0001FFD97FFC491380D9FFF05B4817C048495B5C5A485B -A2486F138091C7FC486F1300705A4892C8FC5BA312FFAD127F7FA27EA2EF03E06C7F1707 -6C6D15C07E6E140F6CEE1F806C6DEC3F006C6D147ED97FFE5C6D6CEB03F8010F9038E01F -F0010390B55A01001580023F49C7FC020113E033387CB63C>I<4DB47E0407B5FCA5EE00 -1F1707B3A4913801FFE0021F13FC91B6FC010315C7010F9038E03FE74990380007F7D97F -FC0101B5FC49487F4849143F484980485B83485B5A91C8FC5AA3485AA412FFAC127FA36C -7EA37EA26C7F5F6C6D5C7E6C6D5C6C6D49B5FC6D6C4914E0D93FFED90FEFEBFF80903A0F -FFC07FCF6D90B5128F0101ECFE0FD9003F13F8020301C049C7FC41547CD24B>I<913803 -FFC0023F13FC49B6FC010715C04901817F903A3FFC007FF849486D7E49486D7E4849130F -48496D7E48178048497F18C0488191C7FC4817E0A248815B18F0A212FFA490B8FCA318E0 -49CAFCA6127FA27F7EA218E06CEE01F06E14037E6C6DEC07E0A26C6DEC0FC06C6D141F6C -6DEC3F806D6CECFF00D91FFEEB03FE903A0FFFC03FF8010390B55A010015C0021F49C7FC -020113F034387CB63D>II -II<137F497E000313E0487FA2487FA76C5BA26C5BC613806DC7FC -90C8FCADEB3FF0B5FCA512017EB3B3A6B612E0A51B547BD325>I107 DIII<913801FFE002 -1F13FE91B612C0010315F0010F9038807FFC903A1FFC000FFED97FF86D6C7E49486D7F48 -496D7F48496D7F4A147F48834890C86C7EA24883A248486F7EA3007F1880A400FF18C0AC -007F1880A3003F18006D5DA26C5FA26C5F6E147F6C5F6C6D4A5A6C6D495B6C6D495B6D6C -495BD93FFE011F90C7FC903A0FFF807FFC6D90B55A010015C0023F91C8FC020113E03A38 -7CB643>I<903A3FF001FFE0B5010F13FE033FEBFFC092B612F002F301017F913AF7F800 -7FFE0003D9FFE0EB1FFFC602806D7F92C76C7F4A824A6E7F4A6E7FA2717FA285187F85A4 -721380AC1A0060A36118FFA2615F616E4A5BA26E4A5B6E4A5B6F495B6F4990C7FC03F0EB -FFFC9126FBFE075B02F8B612E06F1480031F01FCC8FC030313C092CBFCB1B612F8A5414D -7BB54B>I<90397FE003FEB590380FFF80033F13E04B13F09238FE1FF89139E1F83FFC00 -03D9E3E013FEC6ECC07FECE78014EF150014EE02FEEB3FFC5CEE1FF8EE0FF04A90C7FCA5 -5CB3AAB612FCA52F367CB537>114 D<903903FFF00F013FEBFE1F90B7FC120348EB003F -D80FF81307D81FE0130148487F4980127F90C87EA24881A27FA27F01F091C7FC13FCEBFF -C06C13FF15F86C14FF16C06C15F06C816C816C81C681013F1580010F15C01300020714E0 -EC003F030713F015010078EC007F00F8153F161F7E160FA27E17E07E6D141F17C07F6DEC -3F8001F8EC7F0001FEEB01FE9039FFC00FFC6DB55AD8FC1F14E0D8F807148048C601F8C7 -FC2C387CB635>I<143EA6147EA414FEA21301A313031307A2130F131F133F13FF5A000F -90B6FCB8FCA426003FFEC8FCB3A9EE07C0AB011FEC0F8080A26DEC1F0015806DEBC03E6D -EBF0FC6DEBFFF86D6C5B021F5B020313802A4D7ECB34>IIII121 D E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fl cmr10 10.95 51 -/Fl 51 122 df12 D<1430147014E0EB01C0EB038013 -07EB0F00131E133E133C5B13F85B12015B1203A2485AA2120F5BA2121F90C7FCA25AA312 -3E127EA6127C12FCB2127C127EA6123E123FA37EA27F120FA27F1207A26C7EA212017F12 -007F13787F133E131E7FEB07801303EB01C0EB00E014701430145A77C323>40 -D<12C07E12707E7E121E7E6C7E7F12036C7E7F12007F1378137CA27FA2133F7FA2148013 -0FA214C0A3130714E0A6130314F0B214E01307A614C0130FA31480A2131F1400A25B133E -A25BA2137813F85B12015B485A12075B48C7FC121E121C5A5A5A5A145A7BC323>I45 D<121EEA7F80A2EAFFC0A4EA7F80A2EA1E000A0A798919>I48 DIII<150E151E153EA2157EA215FE1401A21403EC077E1406140E141C -A214381470A214E0EB01C0A2EB0380EB0700A2130E5BA25B5BA25B5B1201485A90C7FC5A -120E120C121C5AA25A5AB8FCA3C8EAFE00AC4A7E49B6FCA3283E7EBD2D>I<00061403D8 -0780131F01F813FE90B5FC5D5D5D15C092C7FC14FCEB3FE090C9FCACEB01FE90380FFF80 -90383E03E090387001F8496C7E49137E497F90C713800006141FC813C0A216E0150FA316 -F0A3120C127F7F12FFA416E090C7121F12FC007015C012780038EC3F80123C6CEC7F0000 -1F14FE6C6C485A6C6C485A3903F80FE0C6B55A013F90C7FCEB07F8243F7CBC2D>II<1238123C123F90B612FCA316F85A16F0 -16E00078C712010070EC03C0ED078016005D48141E151C153C5DC8127015F04A5A5D1403 -4A5A92C7FC5C141EA25CA2147C147814F8A213015C1303A31307A3130F5CA2131FA6133F -AA6D5A0107C8FC26407BBD2D>III<121EEA7F80A2EAFFC0 -A4EA7F80A2EA1E00C7FCB3121EEA7F80A2EAFFC0A4EA7F80A2EA1E000A2779A619>I67 DII72 DI<011FB512FCA3D9000713006E5A1401B3B3A6123FEA7F80EAFFC0A44A -5A1380D87F005B007C130700385C003C495A6C495A6C495A2603E07EC7FC3800FFF8EB3F -C026407CBD2F>I76 DII80 D83 D<003FB91280A3903AF0007FE001018090393FC0003F48C7ED1FC0007E1707127C -00781703A300701701A548EF00E0A5C81600B3B14B7E4B7E0107B612FEA33B3D7DBC42> -I89 D97 DI<49B4FC010F13E090383F00F8 -017C131E4848131F4848137F0007ECFF80485A5B121FA24848EB7F00151C007F91C7FCA2 -90C9FC5AAB6C7EA3003FEC01C07F001F140316806C6C13076C6C14000003140E6C6C131E -6C6C137890383F01F090380FFFC0D901FEC7FC222A7DA828>III -I<167C903903F801FF903A1FFF078F8090397E0FDE1F9038F803F83803F001A23B07E000 -FC0600000F6EC7FC49137E001F147FA8000F147E6D13FE00075C6C6C485AA23901F803E0 -3903FE0FC026071FFFC8FCEB03F80006CAFC120EA3120FA27F7F6CB512E015FE6C6E7E6C -15E06C810003813A0FC0001FFC48C7EA01FE003E140048157E825A82A46C5D007C153E00 -7E157E6C5D6C6C495A6C6C495AD803F0EB0FC0D800FE017FC7FC90383FFFFC010313C029 -3D7EA82D>III108 D<2701F801FE14FF00FF902707FFC00313E0913B1E07E00F03F0 -913B7803F03C01F80007903BE001F87000FC2603F9C06D487F000101805C01FBD900FF14 -7F91C75B13FF4992C7FCA2495CB3A6486C496CECFF80B5D8F87FD9FC3F13FEA347287DA7 -4C>I<3901F801FE00FF903807FFC091381E07E091387803F000079038E001F82603F9C0 -7F0001138001FB6D7E91C7FC13FF5BA25BB3A6486C497EB5D8F87F13FCA32E287DA733> -I<14FF010713E090381F81F890387E007E01F8131F4848EB0F804848EB07C04848EB03E0 -000F15F04848EB01F8A2003F15FCA248C812FEA44815FFA96C15FEA36C6CEB01FCA3001F -15F86C6CEB03F0A26C6CEB07E06C6CEB0FC06C6CEB1F80D8007EEB7E0090383F81FC9038 -0FFFF0010090C7FC282A7EA82D>I<3901FC03FC00FF90381FFF8091387C0FE09039FDE0 -03F03A07FFC001FC6C496C7E6C90C7127F49EC3F805BEE1FC017E0A2EE0FF0A3EE07F8AA -EE0FF0A4EE1FE0A2EE3FC06D1580EE7F007F6E13FE9138C001F89039FDE007F09039FC78 -0FC0DA3FFFC7FCEC07F891C9FCAD487EB512F8A32D3A7EA733>I<3901F807E000FFEB1F -F8EC787CECE1FE3807F9C100031381EA01FB1401EC00FC01FF1330491300A35BB3A5487E -B512FEA31F287EA724>114 D<90383FC0603901FFF8E03807C03F381F000F003E130700 -3C1303127C0078130112F81400A27E7E7E6D1300EA7FF8EBFFC06C13F86C13FE6C7F6C14 -80000114C0D8003F13E0010313F0EB001FEC0FF800E01303A214017E1400A27E15F07E14 -016C14E06CEB03C0903880078039F3E01F0038E0FFFC38C01FE01D2A7DA824>I<131CA6 -133CA4137CA213FCA2120112031207001FB512C0B6FCA2D801FCC7FCB3A215E0A9120090 -38FE01C0A2EB7F03013F138090381F8700EB07FEEB01F81B397EB723>IIIIII E -%EndDVIPSBitmapFont -%DVIPSBitmapFont: Fm cmbx12 20.736 9 -/Fm 9 123 df<92380FFFE04AB67E020F15F0027F15FE49B87E4917E0010F17F8013F83 -49D9C01F14FF9027FFFC0001814801E06D6C80480180021F804890C86C8048486F804848 -6F8001FF6F804801C06E8002F081486D18806E816E18C0B5821BE06E81A37214F0A56C5B -A36C5B6C5B6C5B000313C0C690C9FC90CA15E060A34E14C0A21B80601B0060626295B55A -5F624D5C624D5C4D91C7FC614D5B4D13F04D5B6194B55A4C49C8FC4C5B4C5B4C13E04C5B -604C90C9FCEE7FFC4C5A4B5B4B5B4B0180EC0FF04B90C8FC4B5A4B5A4B48ED1FE0EDFFE0 -4A5B4A5B4A90C9FC4A48163F4A5ADA3FF017C05D4A48167F4A5A4990CA12FFD903FC1607 -49BAFC5B4919805B5B90BBFC5A5A5A5A481A005A5ABCFCA462A44C7176F061>50 -D<92383FFFF80207B612E0027F15FC49B87E010717E0011F83499026F0007F13FC4948C7 -000F7F90B502036D7E486E6D806F6D80727F486E6E7F8486727FA28684A26C5C72806C5C -6D90C8FC6D5AEB0FF8EB03E090CAFCA70507B6FC041FB7FC0303B8FC157F0203B9FC021F -ECFE0391B612800103ECF800010F14C04991C7FC017F13FC90B512F04814C0485C4891C8 -FC485B5A485B5C5A5CA2B5FC5CA360A36E5DA26C5F6E5D187E6C6D846E4A48806C6D4A48 -14FC6C6ED90FF0ECFFFC6C02E090263FE07F14FE00019139FC03FFC06C91B6487E013F4B -487E010F4B1307010303F01301D9003F0280D9003F13FC020101F8CBFC57507ACE5E>97 -D<903801FFFCB6FCA8C67E131F7FB3ADF0FFFC050FEBFFE0057F14FE0403B77E040F16E0 -043F16F84CD9007F13FE9226FDFFF001077F92B500C001018094C86C13E004FC6F7F4C6F -7F04E06F7F4C6F7F5E747F93C915804B7014C0A27414E0A21DF087A21DF8A31DFC87A41D -FEAF1DFCA4631DF8A31DF098B5FC1DE0A25014C0A26F1980501400705D705F704B5B505B -704B5B04FC4B5BDBE7FE92B55A9226C3FF8001035C038101E0011F49C7FC9226807FFC90 -B55A4B6CB712F04A010F16C04A010393C8FC4A010015F84A023F14C090C9000301F0C9FC -5F797AF76C>I<97380FFFE00607B6FCA8F00003190086B3AD93383FFF800307B512F803 -3F14FF4AB712C0020716F0021F16FC027F9039FE007FFE91B500F0EB0FFF010302800101 -90B5FC4949C87E49498149498149498149498190B548814884484A8192CAFC5AA2485BA2 -5A5C5AA35A5CA4B5FCAF7EA4807EA37EA2807EA26C7F616C6E5D6C606C80616D6D5D6D6D -5D6D6D92B67E6D6D4A15FC010301FF0207EDFFFE6D02C0EB3FFE6D6C9039FC01FFF86E90 -B65A020F16C002031600DA007F14FC030F14E09226007FFEC749C7FC5F797AF76C>100 -D105 D<903801FFFCB6FCA8C67E131F7FB3B3B3B3B3ABB812C0A82A7879F7 -35>108 D<902601FFF891380FFFE0B692B512FE05036E7E050F15E0053F15F84D819327 -01FFF01F7F4CD900077FDC07FC6D80C66CDA0FF06D80011FDA1FC07F6D4A48824CC8FC04 -7E6F7F5EEDF9F85E03FB707F5E15FF5EA25EA293C9FCA45DB3B3A6B8D8E003B81280A861 -4E79CD6C>110 D<902601FFFCEC7FFEB6020FB512F0057F14FE4CB712C0040716F0041F -82047F16FE93B5C66C7F92B500F0010F14C0C66C0380010380011F4AC76C806D4A6E8004 -F06F7F4C6F7F4C6F7F4C8193C915804B7014C0861DE0A27414F0A27414F8A47513FCA575 -13FEAF5113FCA598B512F8A31DF0621DE0621DC0621D806F5E701800704B5B505B704B5B -7092B55A04FC4A5C704A5C706C010F5C05E0013F49C7FC9227FE7FFC01B55A70B712F004 -0F16C0040393C8FC040015F8053F14C0050301F0C9FC94CCFCB3A6B812E0A85F6F7ACD6C ->112 D<0007BA12FE1AFFA503E0C76C13FE4AC8B512FC4801F04A14F84A5C4A17F091C8 -4814E04D14C0495D4918804D1400494B5B94B5FC61494A5C4C5C5E61001F4B5C494A5C5E -96C7FC4C5B93B55A5DC85D4B5C4B5C5D604B5C4B91C8FC5D5F92B55A4A5C5C4D14FF4A5C -4A5C5C5F4A91C75A4A4915FE91B5FC5E495C495C4917035E495C495C49170793C8FC4949 -ED0FFC90B55A48181F5D484A153F484A157F4818FF4B1403484A140F4891C8123F480407 -B5FC92B8FCBB12F8A57E484D7BCC56>122 D E -%EndDVIPSBitmapFont -end -%%EndProlog -%%BeginSetup -%%Feature: *Resolution 600dpi -TeXDict begin -%%PaperSize: A4 - -%%EndSetup -%%Page: 1 1 -1 0 bop 150 1318 a Fm(bzip2)64 b(and)g(libbzip2)p 150 -1418 3600 34 v 2010 1515 a Fl(a)31 b(program)f(and)g(library)e(for)i -(data)h(compression)2198 1623 y(cop)m(yrigh)m(t)f(\(C\))h(1996-2002)j -(Julian)28 b(Sew)m(ard)2394 1731 y(v)m(ersion)i(1.0.2)i(of)f(30)g -(Decem)m(b)s(er)g(2001)150 5091 y Fk(Julian)46 b(Sew)l(ard)p -150 5141 3600 17 v eop -%%Page: 1 2 -1 1 bop 3705 -116 a Fl(1)150 299 y(The)40 b(follo)m(wing)f(text)i(is)f -(the)h(License)f(for)g(this)f(soft)m(w)m(are.)73 b(Y)-8 -b(ou)41 b(should)d(\014nd)h(it)h(iden)m(tical)f(to)i(that)150 -408 y(con)m(tained)31 b(in)e(the)h(\014le)g(LICENSE)f(in)g(the)h -(source)h(distribution.)150 565 y Fj(||||||)f(ST)-8 b(AR)g(T)30 -b(OF)h(THE)f(LICENSE)f(||||||)150 722 y(This)24 b(program,)i -Fi(bzip2)p Fj(,)f(and)f(asso)s(ciated)h(library)g Fi(libbzip2)p -Fj(,)e(are)i(Cop)m(yrigh)m(t)h(\(C\))f(1996-2002)j(Julian)150 -832 y(R)i(Sew)m(ard.)41 b(All)30 b(righ)m(ts)h(reserv)m(ed.)150 -989 y(Redistribution)46 b(and)f(use)g(in)g(source)h(and)f(binary)f -(forms,)50 b(with)45 b(or)h(without)g(mo)s(di\014cation,)j(are)150 -1098 y(p)s(ermitted)31 b(pro)m(vided)f(that)h(the)g(follo)m(wing)g -(conditions)g(are)g(met:)225 1255 y Fh(\017)60 b Fj(Redistributions)44 -b(of)g(source)g(co)s(de)f(m)m(ust)h(retain)g(the)g(ab)s(o)m(v)m(e)h -(cop)m(yrigh)m(t)g(notice,)k(this)43 b(list)i(of)330 -1365 y(conditions)31 b(and)e(the)i(follo)m(wing)h(disclaimer.)225 -1499 y Fh(\017)60 b Fj(The)32 b(origin)g(of)g(this)g(soft)m(w)m(are)h -(m)m(ust)f(not)g(b)s(e)f(misrepresen)m(ted;)j(y)m(ou)e(m)m(ust)g(not)g -(claim)g(that)h(y)m(ou)330 1609 y(wrote)d(the)f(original)h(soft)m(w)m -(are.)42 b(If)29 b(y)m(ou)g(use)g(this)g(soft)m(w)m(are)i(in)d(a)h(pro) -s(duct,)g(an)g(ac)m(kno)m(wledgmen)m(t)330 1718 y(in)h(the)h(pro)s -(duct)e(do)s(cumen)m(tation)j(w)m(ould)e(b)s(e)f(appreciated)j(but)d -(is)i(not)g(required.)225 1853 y Fh(\017)60 b Fj(Altered)36 -b(source)g(v)m(ersions)g(m)m(ust)f(b)s(e)f(plainly)i(mark)m(ed)f(as)g -(suc)m(h,)i(and)d(m)m(ust)h(not)h(b)s(e)f(misrepre-)330 -1962 y(sen)m(ted)c(as)f(b)s(eing)g(the)h(original)h(soft)m(w)m(are.)225 -2097 y Fh(\017)60 b Fj(The)25 b(name)g(of)g(the)g(author)g(ma)m(y)h -(not)f(b)s(e)f(used)h(to)g(endorse)g(or)h(promote)g(pro)s(ducts)e -(deriv)m(ed)h(from)330 2206 y(this)30 b(soft)m(w)m(are)i(without)f(sp)s -(eci\014c)f(prior)g(written)h(p)s(ermission.)150 2388 -y(THIS)36 b(SOFTW)-10 b(ARE)35 b(IS)h(PR)m(O)m(VIDED)g(BY)f(THE)h(A)m -(UTHOR)h(\\AS)f(IS")g(AND)f(ANY)h(EXPRESS)150 2498 y(OR)29 -b(IMPLIED)g(W)-10 b(ARRANTIES,)28 b(INCLUDING,)h(BUT)g(NOT)g(LIMITED)g -(TO,)h(THE)e(IMPLIED)150 2607 y(W)-10 b(ARRANTIES)60 -b(OF)h(MER)m(CHANT)-8 b(ABILITY)61 b(AND)g(FITNESS)f(F)m(OR)i(A)f(P)-8 -b(AR)g(TICULAR)150 2717 y(PURPOSE)40 b(ARE)g(DISCLAIMED.)f(IN)i(NO)g -(EVENT)f(SHALL)f(THE)i(A)m(UTHOR)g(BE)f(LIABLE)150 2827 -y(F)m(OR)d(ANY)f(DIRECT,)g(INDIRECT,)g(INCIDENT)-8 b(AL,)37 -b(SPECIAL,)d(EXEMPLAR)-8 b(Y,)36 b(OR)h(CON-)150 2936 -y(SEQUENTIAL)30 b(D)m(AMA)m(GES)h(\(INCLUDING,)h(BUT)e(NOT)i(LIMITED)f -(TO,)h(PR)m(OCUREMENT)150 3046 y(OF)55 b(SUBSTITUTE)e(GOODS)i(OR)f(SER) --10 b(VICES;)53 b(LOSS)g(OF)i(USE,)f(D)m(A)-8 b(T)g(A,)55 -b(OR)g(PR)m(OFITS;)150 3155 y(OR)e(BUSINESS)d(INTERR)m(UPTION\))k(HO)m -(WEVER)e(CA)m(USED)f(AND)h(ON)h(ANY)f(THEOR)-8 b(Y)150 -3265 y(OF)71 b(LIABILITY,)f(WHETHER)g(IN)h(CONTRA)m(CT,)g(STRICT)g -(LIABILITY,)e(OR)i(TOR)-8 b(T)150 3374 y(\(INCLUDING)45 -b(NEGLIGENCE)g(OR)g(OTHER)-10 b(WISE\))45 b(ARISING)g(IN)g(ANY)g(W)-10 -b(A)i(Y)44 b(OUT)i(OF)150 3484 y(THE)i(USE)f(OF)h(THIS)g(SOFTW)-10 -b(ARE,)47 b(EVEN)g(IF)h(AD)m(VISED)e(OF)i(THE)g(POSSIBILITY)f(OF)150 -3594 y(SUCH)30 b(D)m(AMA)m(GE.)150 3750 y(Julian)g(Sew)m(ard,)g(Cam)m -(bridge,)h(UK.)150 3907 y Fi(jseward@acm.org)150 4064 -y(bzip2)p Fj(/)p Fi(libbzip2)c Fj(v)m(ersion)k(1.0.2)i(of)d(30)h(Decem) -m(b)s(er)g(2001.)150 4221 y(||||||)f(END)g(OF)g(THE)g(LICENSE)f(||||||) -150 4378 y(W)-8 b(eb)31 b(sites:)150 4535 y Fi -(http://sources.redhat.co)o(m/bz)o(ip2)150 4691 y -(http://www.cacheprof.org)150 4848 y Fj(P)-8 b(A)g(TENTS:)40 -b(T)-8 b(o)40 b(the)f(b)s(est)g(of)g(m)m(y)h(kno)m(wledge,)i -Fi(bzip2)c Fj(do)s(es)h(not)g(use)g(an)m(y)h(paten)m(ted)g(algorithms.) -150 4958 y(Ho)m(w)m(ev)m(er,)31 b(I)c(do)h(not)g(ha)m(v)m(e)g(the)g -(resources)h(a)m(v)-5 b(ailable)29 b(to)f(carry)g(out)g(a)g(full)f -(paten)m(t)i(searc)m(h.)40 b(Therefore)150 5067 y(I)30 -b(cannot)h(giv)m(e)h(an)m(y)f(guaran)m(tee)g(of)g(the)g(ab)s(o)m(v)m(e) -g(statemen)m(t.)p eop -%%Page: 2 3 -2 2 bop 150 -116 a Fl(Chapter)30 b(1:)41 b(In)m(tro)s(duction)2591 -b(2)150 299 y Fg(1)80 b(In)l(tro)t(duction)150 555 y -Fi(bzip2)40 b Fj(compresses)h(\014les)g(using)g(the)g(Burro)m -(ws-Wheeler)h(blo)s(c)m(k-sorting)h(text)f(compression)f(algo-)150 -665 y(rithm,)54 b(and)49 b(Hu\013man)f(co)s(ding.)98 -b(Compression)49 b(is)h(generally)h(considerably)e(b)s(etter)h(than)f -(that)150 775 y(ac)m(hiev)m(ed)43 b(b)m(y)f(more)g(con)m(v)m(en)m -(tional)i(LZ77/LZ78-based)e(compressors,)j(and)40 b(approac)m(hes)i -(the)g(p)s(er-)150 884 y(formance)31 b(of)f(the)h(PPM)g(family)f(of)h -(statistical)i(compressors.)150 1041 y Fi(bzip2)g Fj(is)h(built)h(on)f -(top)h(of)f Fi(libbzip2)p Fj(,)f(a)i(\015exible)f(library)h(for)f -(handling)g(compressed)g(data)h(in)f(the)150 1151 y Fi(bzip2)29 -b Fj(format.)41 b(This)30 b(man)m(ual)g(describ)s(es)g(b)s(oth)f(ho)m -(w)i(to)g(use)f(the)g(program)h(and)e(ho)m(w)h(to)h(w)m(ork)g(with)150 -1260 y(the)26 b(library)f(in)m(terface.)41 b(Most)26 -b(of)f(the)h(man)m(ual)f(is)g(dev)m(oted)i(to)f(this)f(library)-8 -b(,)27 b(not)f(the)g(program,)g(whic)m(h)150 1370 y(is)k(go)s(o)s(d)h -(news)e(if)i(y)m(our)f(in)m(terest)i(is)f(only)f(in)g(the)h(program.) -150 1527 y(Chapter)45 b(2)h(describ)s(es)e(ho)m(w)i(to)g(use)f -Fi(bzip2)p Fj(;)51 b(this)46 b(is)f(the)h(only)f(part)g(y)m(ou)h(need)f -(to)h(read)f(if)h(y)m(ou)150 1636 y(just)39 b(w)m(an)m(t)g(to)h(kno)m -(w)f(ho)m(w)h(to)f(op)s(erate)h(the)g(program.)67 b(Chapter)38 -b(3)i(describ)s(es)e(the)i(programming)150 1746 y(in)m(terfaces)27 -b(in)f(detail,)i(and)d(Chapter)g(4)h(records)g(some)g(miscellaneous)h -(notes)f(whic)m(h)g(I)g(though)m(t)g(ough)m(t)150 1855 -y(to)31 b(b)s(e)f(recorded)h(somewhere.)p eop -%%Page: 3 4 -3 3 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(3)150 299 y Fg(2)80 b(Ho)l(w)53 -b(to)g(use)g Ff(bzip2)150 566 y Fj(This)30 b(c)m(hapter)h(con)m(tains)h -(a)e(cop)m(y)h(of)g(the)g Fi(bzip2)d Fj(man)i(page,)h(and)f(nothing)g -(else.)390 818 y Fe(NAME)570 1004 y Fi(bzip2)p Fj(,)f -Fi(bunzip2)g Fj(-)h(a)h(blo)s(c)m(k-sorting)h(\014le)e(compressor,)h -(v1.0.2)570 1136 y Fi(bzcat)e Fj(-)i(decompresses)f(\014les)h(to)g -(stdout)570 1267 y Fi(bzip2recover)c Fj(-)k(reco)m(v)m(ers)h(data)f -(from)f(damaged)g(bzip2)h(\014les)390 1519 y Fe(SYNOPSIS)570 -1706 y Fi(bzip2)e Fj([)i(-cdfkqstvzVL123456789)j(])d([)g(\014lenames)f -(...)41 b(])570 1837 y Fi(bunzip2)28 b Fj([)j(-fkvsVL)f(])g([)h -(\014lenames)f(...)41 b(])570 1968 y Fi(bzcat)29 b Fj([)i(-s)f(])h([)g -(\014lenames)f(...)41 b(])570 2100 y Fi(bzip2recover)27 -b Fj(\014lename)390 2352 y Fe(DESCRIPTION)390 2538 y -Fi(bzip2)e Fj(compresses)i(\014les)f(using)g(the)h(Burro)m(ws-Wheeler)g -(blo)s(c)m(k)f(sorting)i(text)f(compres-)390 2642 y(sion)39 -b(algorithm,)k(and)38 b(Hu\013man)h(co)s(ding.)67 b(Compression)39 -b(is)g(generally)i(considerably)390 2746 y(b)s(etter)23 -b(than)g(that)g(ac)m(hiev)m(ed)h(b)m(y)f(more)g(con)m(v)m(en)m(tional)i -(LZ77/LZ78-based)e(compressors,)390 2850 y(and)k(approac)m(hes)i(the)f -(p)s(erformance)g(of)h(the)f(PPM)h(family)f(of)h(statistical)h -(compressors.)390 3001 y(The)c(command-line)g(options)g(are)g(delib)s -(erately)h(v)m(ery)f(similar)h(to)f(those)h(of)f(GNU)g -Fi(gzip)p Fj(,)390 3104 y(but)k(they)g(are)h(not)g(iden)m(tical.)390 -3255 y Fi(bzip2)e Fj(exp)s(ects)j(a)f(list)h(of)f(\014le)g(names)f(to)i -(accompan)m(y)g(the)f(command-line)g(\015ags.)43 b(Eac)m(h)390 -3359 y(\014le)d(is)h(replaced)g(b)m(y)f(a)g(compressed)h(v)m(ersion)g -(of)f(itself,)k(with)d(the)f(name)g Fi(original_)390 -3463 y(name.bz2)p Fj(.)45 b(Eac)m(h)33 b(compressed)g(\014le)g(has)f -(the)h(same)g(mo)s(di\014cation)g(date,)h(p)s(ermissions,)390 -3567 y(and,)47 b(when)c(p)s(ossible,)k(o)m(wnership)c(as)h(the)g -(corresp)s(onding)g(original,)49 b(so)44 b(that)g(these)390 -3671 y(prop)s(erties)37 b(can)g(b)s(e)g(correctly)i(restored)f(at)g -(decompression)f(time.)62 b(File)39 b(name)d(han-)390 -3774 y(dling)e(is)g(naiv)m(e)h(in)f(the)g(sense)g(that)h(there)f(is)g -(no)g(mec)m(hanism)g(for)h(preserving)f(original)390 -3878 y(\014le)41 b(names,)i(p)s(ermissions,)g(o)m(wnerships)d(or)h -(dates)g(in)g(\014lesystems)g(whic)m(h)f(lac)m(k)i(these)390 -3982 y(concepts,)32 b(or)e(ha)m(v)m(e)i(serious)e(\014le)h(name)f -(length)h(restrictions,)h(suc)m(h)e(as)g(MS-DOS.)390 -4133 y Fi(bzip2)24 b Fj(and)i Fi(bunzip2)d Fj(will)k(b)m(y)f(default)f -(not)i(o)m(v)m(erwrite)h(existing)f(\014les.)39 b(If)26 -b(y)m(ou)g(w)m(an)m(t)h(this)390 4237 y(to)k(happ)s(en,)e(sp)s(ecify)h -(the)h Fi(-f)e Fj(\015ag.)390 4388 y(If)k(no)f(\014le)h(names)f(are)i -(sp)s(eci\014ed,)e Fi(bzip2)g Fj(compresses)h(from)f(standard)g(input)g -(to)i(stan-)390 4491 y(dard)d(output.)44 b(In)32 b(this)f(case,)i -Fi(bzip2)e Fj(will)h(decline)g(to)h(write)f(compressed)g(output)f(to)i -(a)390 4595 y(terminal,)e(as)g(this)f(w)m(ould)g(b)s(e)g(en)m(tirely)i -(incomprehensible)e(and)g(therefore)h(p)s(oin)m(tless.)390 -4746 y Fi(bunzip2)j Fj(\(or)j Fi(bzip2)29 b(-d)p Fj(\))36 -b(decompresses)g(all)h(sp)s(eci\014ed)e(\014les.)58 b(Files)37 -b(whic)m(h)f(w)m(ere)h(not)390 4850 y(created)f(b)m(y)e -Fi(bzip2)g Fj(will)g(b)s(e)g(detected)j(and)c(ignored,)k(and)c(a)i(w)m -(arning)g(issued.)52 b Fi(bzip2)390 4954 y Fj(attempts)29 -b(to)f(guess)g(the)g(\014lename)f(for)h(the)g(decompressed)f(\014le)h -(from)f(that)h(of)g(the)g(com-)390 5058 y(pressed)i(\014le)g(as)h -(follo)m(ws:)570 5209 y Fi(filename.bz2)57 b Fj(b)s(ecomes)31 -b Fi(filename)570 5340 y(filename.bz)58 b Fj(b)s(ecomes)30 -b Fi(filename)p eop -%%Page: 4 5 -4 4 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(4)570 299 y Fi(filename.tbz2)27 -b Fj(b)s(ecomes)j Fi(filename.tar)570 470 y(filename.tbz)57 -b Fj(b)s(ecomes)31 b Fi(filename.tar)570 641 y(anyothername)57 -b Fj(b)s(ecomes)31 b Fi(anyothername.out)390 859 y Fj(If)i(the)g -(\014le)f(do)s(es)h(not)g(end)f(in)g(one)i(of)e(the)h(recognised)h -(endings,)f Fi(.bz2)p Fj(,)g Fi(.bz)p Fj(,)f Fi(.tbz2)g -Fj(or)390 963 y Fi(.tbz)p Fj(,)h Fi(bzip2)g Fj(complains)g(that)h(it)h -(cannot)f(guess)f(the)h(name)g(of)f(the)h(original)h(\014le,)g(and)390 -1067 y(uses)30 b(the)h(original)g(name)f(with)h Fi(.out)e -Fj(app)s(ended.)390 1218 y(As)g(with)h(compression,)h(supplying)e(no)h -(\014lenames)f(causes)i(decompression)f(from)g(stan-)390 -1321 y(dard)f(input)h(to)h(standard)f(output.)390 1472 -y Fi(bunzip2)h Fj(will)j(correctly)i(decompress)d(a)h(\014le)f(whic)m -(h)g(is)h(the)g(concatenation)h(of)f(t)m(w)m(o)h(or)390 -1576 y(more)i(compressed)g(\014les.)61 b(The)37 b(result)g(is)g(the)g -(concatenation)i(of)e(the)g(corresp)s(onding)390 1680 -y(uncompressed)c(\014les.)52 b(In)m(tegrit)m(y)37 b(testing)e(\()p -Fi(-t)p Fj(\))f(of)h(concatenated)h(compressed)e(\014les)g(is)390 -1784 y(also)d(supp)s(orted.)390 1935 y(Y)-8 b(ou)38 b(can)g(also)g -(compress)g(or)g(decompress)g(\014les)f(to)i(the)f(standard)f(output)g -(b)m(y)h(giving)390 2039 y(the)44 b Fi(-c)e Fj(\015ag.)79 -b(Multiple)44 b(\014les)g(ma)m(y)f(b)s(e)g(compressed)g(and)f -(decompressed)h(lik)m(e)i(this.)390 2142 y(The)31 b(resulting)h -(outputs)e(are)i(fed)f(sequen)m(tially)h(to)g(stdout.)43 -b(Compression)31 b(of)g(m)m(ultiple)390 2246 y(\014les)43 -b(in)g(this)g(manner)g(generates)i(a)e(stream)h(con)m(taining)g(m)m -(ultiple)h(compressed)e(\014le)390 2350 y(represen)m(tations.)48 -b(Suc)m(h)32 b(a)g(stream)h(can)f(b)s(e)g(decompressed)g(correctly)j -(only)d(b)m(y)g Fi(bzip2)390 2454 y Fj(v)m(ersion)h(0.9.0)h(or)f -(later.)48 b(Earlier)33 b(v)m(ersions)g(of)g Fi(bzip2)e -Fj(will)i(stop)g(after)g(decompressing)390 2558 y(the)e(\014rst)f -(\014le)g(in)g(the)h(stream.)390 2709 y Fi(bzcat)e Fj(\(or)i -Fi(bzip2)e(-dc)p Fj(\))h(decompresses)g(all)h(sp)s(eci\014ed)f(\014les) -g(to)i(the)e(standard)g(output.)390 2860 y Fi(bzip2)c -Fj(will)j(read)e(argumen)m(ts)i(from)e(the)h(en)m(vironmen)m(t)h(v)-5 -b(ariables)28 b Fi(BZIP2)f Fj(and)g Fi(BZIP)p Fj(,)g(in)390 -2963 y(that)21 b(order,)i(and)d(will)h(pro)s(cess)f(them)h(b)s(efore)g -(an)m(y)f(argumen)m(ts)h(read)g(from)f(the)h(command)390 -3067 y(line.)41 b(This)30 b(giv)m(es)i(a)e(con)m(v)m(enien)m(t)j(w)m(a) -m(y)e(to)g(supply)e(default)i(argumen)m(ts.)390 3218 -y(Compression)e(is)h(alw)m(a)m(ys)h(p)s(erformed,)e(ev)m(en)h(if)g(the) -g(compressed)f(\014le)h(is)g(sligh)m(tly)h(larger)390 -3322 y(than)24 b(the)h(original.)40 b(Files)25 b(of)f(less)h(than)f(ab) -s(out)g(one)h(h)m(undred)d(b)m(ytes)j(tend)g(to)g(get)h(larger,)390 -3426 y(since)33 b(the)g(compression)f(mec)m(hanism)h(has)f(a)h(constan) -m(t)g(o)m(v)m(erhead)h(in)e(the)h(region)g(of)g(50)390 -3529 y(b)m(ytes.)50 b(Random)33 b(data)h(\(including)f(the)h(output)f -(of)g(most)h(\014le)g(compressors\))g(is)f(co)s(ded)390 -3633 y(at)e(ab)s(out)f(8.05)i(bits)e(p)s(er)g(b)m(yte,)h(giving)h(an)e -(expansion)g(of)g(around)f(0.5\045.)390 3784 y(As)f(a)h(self-c)m(hec)m -(k)i(for)e(y)m(our)g(protection,)i Fi(bzip2)d Fj(uses)g(32-bit)i(CR)m -(Cs)f(to)g(mak)m(e)h(sure)e(that)390 3888 y(the)43 b(decompressed)g(v)m -(ersion)h(of)f(a)g(\014le)f(is)h(iden)m(tical)i(to)f(the)f(original.)79 -b(This)43 b(guards)390 3992 y(against)h(corruption)f(of)h(the)f -(compressed)g(data,)k(and)42 b(against)i(undetected)g(bugs)e(in)390 -4096 y Fi(bzip2)33 b Fj(\(hop)s(efully)i(v)m(ery)g(unlik)m(ely\).)55 -b(The)35 b(c)m(hances)h(of)f(data)g(corruption)g(going)h(unde-)390 -4199 y(tected)e(is)f(microscopic,)h(ab)s(out)e(one)h(c)m(hance)g(in)g -(four)f(billion)g(for)h(eac)m(h)g(\014le)g(pro)s(cessed.)390 -4303 y(Be)j(a)m(w)m(are,)k(though,)d(that)g(the)g(c)m(hec)m(k)h(o)s -(ccurs)e(up)s(on)f(decompression,)j(so)f(it)f(can)h(only)390 -4407 y(tell)28 b(y)m(ou)e(that)h(something)g(is)g(wrong.)39 -b(It)27 b(can't)g(help)f(y)m(ou)h(reco)m(v)m(er)i(the)d(original)i -(uncom-)390 4511 y(pressed)f(data.)40 b(Y)-8 b(ou)27 -b(can)g(use)h Fi(bzip2recover)23 b Fj(to)29 b(try)e(to)i(reco)m(v)m(er) -g(data)f(from)f(damaged)390 4614 y(\014les.)390 4766 -y(Return)42 b(v)-5 b(alues:)65 b(0)42 b(for)h(a)f(normal)g(exit,)47 -b(1)c(for)f(en)m(vironmen)m(tal)i(problems)e(\(\014le)h(not)390 -4869 y(found,)29 b(in)m(v)-5 b(alid)30 b(\015ags,)g(I/O)h(errors,)g -(&c\),)f(2)h(to)f(indicate)h(a)f(corrupt)g(compressed)g(\014le,)h(3)390 -4973 y(for)f(an)g(in)m(ternal)i(consistency)f(error)g(\(eg,)h(bug\))e -(whic)m(h)g(caused)g Fi(bzip2)f Fj(to)i(panic.)390 5304 -y Fe(OPTIONS)p eop -%%Page: 5 6 -5 5 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(5)390 299 y Fi(-c)30 -b(--stdout)870 403 y Fj(Compress)g(or)g(decompress)h(to)g(standard)e -(output.)390 550 y Fi(-d)h(--decompress)870 653 y Fj(F)-8 -b(orce)42 b(decompression.)72 b Fi(bzip2)p Fj(,)43 b -Fi(bunzip2)38 b Fj(and)i Fi(bzcat)g Fj(are)h(really)h(the)870 -757 y(same)26 b(program,)i(and)d(the)h(decision)h(ab)s(out)f(what)g -(actions)h(to)g(tak)m(e)h(is)e(done)870 861 y(on)j(the)h(basis)f(of)g -(whic)m(h)h(name)f(is)g(used.)40 b(This)29 b(\015ag)g(o)m(v)m(errides)i -(that)f(mec)m(h-)870 965 y(anism,)g(and)g(forces)h(bzip2)f(to)h -(decompress.)390 1112 y Fi(-z)f(--compress)870 1215 y -Fj(The)37 b(complemen)m(t)h(to)g Fi(-d)p Fj(:)54 b(forces)37 -b(compression,)j(regardless)d(of)h(the)f(in-)870 1319 -y(v)m(ok)-5 b(ation)32 b(name.)390 1466 y Fi(-t)e(--test)66 -b Fj(Chec)m(k)26 b(in)m(tegrit)m(y)i(of)e(the)g(sp)s(eci\014ed)f -(\014le\(s\),)i(but)e(don't)h(decompress)g(them.)870 -1570 y(This)39 b(really)h(p)s(erforms)f(a)g(trial)h(decompression)g -(and)e(thro)m(ws)i(a)m(w)m(a)m(y)h(the)870 1674 y(result.)390 -1821 y Fi(-f)30 b(--force)870 1924 y Fj(F)-8 b(orce)27 -b(o)m(v)m(erwrite)h(of)e(output)g(\014les.)39 b(Normally)-8 -b(,)28 b Fi(bzip2)c Fj(will)j(not)f(o)m(v)m(erwrite)870 -2028 y(existing)i(output)e(\014les.)40 b(Also)27 b(forces)g -Fi(bzip2)f Fj(to)h(break)g(hard)f(links)g(to)i(\014les,)870 -2132 y(whic)m(h)i(it)h(otherwise)g(w)m(ouldn't)g(do.)870 -2257 y Fi(bzip2)h Fj(normally)h(declines)h(to)f(decompress)h(\014les)f -(whic)m(h)f(don't)i(ha)m(v)m(e)g(the)870 2361 y(correct)39 -b(magic)g(header)e(b)m(ytes.)63 b(If)38 b(forced)g(\()p -Fi(-f)p Fj(\),)h(ho)m(w)m(ev)m(er,)i(it)e(will)f(pass)870 -2465 y(suc)m(h)30 b(\014les)g(through)g(unmo)s(di\014ed.)39 -b(This)30 b(is)h(ho)m(w)f(GNU)h Fi(gzip)e Fj(b)s(eha)m(v)m(es.)390 -2612 y Fi(-k)h(--keep)66 b Fj(Keep)34 b(\(don't)g(delete\))h(input)d -(\014les)i(during)e(compression)i(or)f(decompres-)870 -2716 y(sion.)390 2863 y Fi(-s)d(--small)870 2966 y Fj(Reduce)35 -b(memory)g(usage,)i(for)e(compression,)i(decompression)e(and)f(test-) -870 3070 y(ing.)40 b(Files)29 b(are)f(decompressed)g(and)f(tested)i -(using)f(a)g(mo)s(di\014ed)f(algorithm)870 3174 y(whic)m(h)32 -b(only)g(requires)h(2.5)g(b)m(ytes)g(p)s(er)f(blo)s(c)m(k)g(b)m(yte.)48 -b(This)32 b(means)f(an)m(y)i(\014le)870 3278 y(can)d(b)s(e)f -(decompressed)g(in)h(2300k)h(of)f(memory)-8 b(,)31 b(alb)s(eit)f(at)g -(ab)s(out)f(half)h(the)870 3381 y(normal)g(sp)s(eed.)870 -3507 y(During)35 b(compression,)j Fi(-s)d Fj(selects)j(a)e(blo)s(c)m(k) -h(size)f(of)h(200k,)h(whic)m(h)e(limits)870 3611 y(memory)43 -b(use)f(to)h(around)e(the)i(same)f(\014gure,)k(at)d(the)f(exp)s(ense)g -(of)h(y)m(our)870 3714 y(compression)32 b(ratio.)47 b(In)31 -b(short,)i(if)e(y)m(our)i(mac)m(hine)f(is)g(lo)m(w)g(on)g(memory)g(\(8) -870 3818 y(megab)m(ytes)39 b(or)e(less\),)j(use)d(-s)g(for)g(ev)m -(erything.)62 b(See)38 b(MEMOR)-8 b(Y)37 b(MAN-)870 3922 -y(A)m(GEMENT)31 b(b)s(elo)m(w.)390 4069 y Fi(-q)f(--quiet)870 -4173 y Fj(Suppress)h(non-essen)m(tial)j(w)m(arning)f(messages.)49 -b(Messages)34 b(p)s(ertaining)f(to)870 4276 y(I/O)e(errors)g(and)e -(other)i(critical)i(ev)m(en)m(ts)f(will)e(not)h(b)s(e)f(suppressed.)390 -4423 y Fi(-v)g(--verbose)870 4527 y Fj(V)-8 b(erb)s(ose)25 -b(mo)s(de)f({)h(sho)m(w)f(the)h(compression)g(ratio)g(for)g(eac)m(h)g -(\014le)g(pro)s(cessed.)870 4631 y(F)-8 b(urther)27 b -Fi(-v)p Fj('s)g(increase)h(the)g(v)m(erb)s(osit)m(y)g(lev)m(el,)i(sp)s -(ewing)d(out)g(lots)h(of)g(infor-)870 4735 y(mation)j(whic)m(h)f(is)g -(primarily)h(of)f(in)m(terest)i(for)f(diagnostic)g(purp)s(oses.)390 -4882 y Fi(-L)f(--license)e(-V)h(--version)870 4985 y -Fj(Displa)m(y)i(the)f(soft)m(w)m(are)i(v)m(ersion,)g(license)f(terms)g -(and)e(conditions.)390 5132 y Fi(-1)h(\(or)f(--fast\))g(to)g(-9)h(\(or) -g(--best\))870 5236 y Fj(Set)h(the)g(blo)s(c)m(k)g(size)g(to)h(100)g -(k,)e(200)i(k)f(..)41 b(900)32 b(k)f(when)e(compressing.)42 -b(Has)870 5340 y(no)37 b(e\013ect)i(when)d(decompressing.)62 -b(See)37 b(MEMOR)-8 b(Y)38 b(MANA)m(GEMENT)p eop -%%Page: 6 7 -6 6 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(6)870 299 y Fj(b)s(elo)m(w.)83 -b(The)44 b Fi(--fast)f Fj(and)g Fi(--best)g Fj(aliases)i(are)g -(primarily)f(for)g(GNU)870 403 y Fi(gzip)27 b Fj(compatibilit)m(y)-8 -b(.)43 b(In)28 b(particular,)i Fi(--fast)d Fj(do)s(esn't)h(mak)m(e)i -(things)e(sig-)870 506 y(ni\014can)m(tly)d(faster.)40 -b(And)23 b Fi(--best)g Fj(merely)i(selects)h(the)f(default)g(b)s(eha)m -(viour.)390 674 y Fi(--)384 b Fj(T)-8 b(reats)38 b(all)f(subsequen)m(t) -f(argumen)m(ts)g(as)h(\014le)f(names,)i(ev)m(en)f(if)f(they)h(start)870 -778 y(with)31 b(a)h(dash.)42 b(This)31 b(is)h(so)f(y)m(ou)h(can)f -(handle)g(\014les)g(with)h(names)f(b)s(eginning)870 881 -y(with)f(a)h(dash,)e(for)i(example:)41 b Fi(bzip2)29 -b(--)h(-myfilename)p Fj(.)390 1049 y Fi(--repetitive-fast)390 -1185 y(--repetitive-best)870 1288 y Fj(These)j(\015ags)g(are)g -(redundan)m(t)f(in)h(v)m(ersions)g(0.9.5)i(and)d(ab)s(o)m(v)m(e.)49 -b(They)33 b(pro-)870 1392 y(vided)g(some)h(coarse)h(con)m(trol)g(o)m(v) -m(er)g(the)f(b)s(eha)m(viour)f(of)h(the)g(sorting)g(algo-)870 -1496 y(rithm)j(in)g(earlier)h(v)m(ersions,)i(whic)m(h)d(w)m(as)g -(sometimes)h(useful.)61 b(0.9.5)39 b(and)870 1600 y(ab)s(o)m(v)m(e)31 -b(ha)m(v)m(e)f(an)g(impro)m(v)m(ed)g(algorithm)h(whic)m(h)e(renders)g -(these)h(\015ags)g(irrel-)870 1704 y(ev)-5 b(an)m(t.)390 -1964 y Fe(MEMOR)-10 b(Y)40 b(MANA)m(GEMENT)390 2151 y -Fi(bzip2)23 b Fj(compresses)i(large)h(\014les)e(in)g(blo)s(c)m(ks.)40 -b(The)24 b(blo)s(c)m(k)h(size)h(a\013ects)g(b)s(oth)e(the)g(compres-) -390 2255 y(sion)38 b(ratio)g(ac)m(hiev)m(ed,)k(and)36 -b(the)i(amoun)m(t)g(of)g(memory)g(needed)f(for)h(compression)g(and)390 -2358 y(decompression.)57 b(The)35 b(\015ags)h Fi(-1)f -Fj(through)g Fi(-9)g Fj(sp)s(ecify)g(the)h(blo)s(c)m(k)g(size)h(to)f(b) -s(e)f(100,000)390 2462 y(b)m(ytes)25 b(through)f(900,000)j(b)m(ytes)e -(\(the)h(default\))e(resp)s(ectiv)m(ely)-8 b(.)41 b(A)m(t)25 -b(decompression)g(time,)390 2566 y(the)k(blo)s(c)m(k)h(size)g(used)f -(for)g(compression)g(is)g(read)h(from)f(the)g(header)g(of)g(the)h -(compressed)390 2670 y(\014le,)j(and)f Fi(bunzip2)f Fj(then)h(allo)s -(cates)i(itself)g(just)e(enough)g(memory)h(to)g(decompress)g(the)390 -2773 y(\014le.)40 b(Since)27 b(blo)s(c)m(k)h(sizes)h(are)f(stored)g(in) -f(compressed)g(\014les,)h(it)h(follo)m(ws)f(that)g(the)g(\015ags)g -Fi(-1)390 2877 y Fj(to)j Fi(-9)f Fj(are)h(irrelev)-5 -b(an)m(t)32 b(to)f(and)f(so)g(ignored)h(during)e(decompression.)390 -3028 y(Compression)h(and)f(decompression)i(requiremen)m(ts,)g(in)f(b)m -(ytes,)i(can)e(b)s(e)g(estimated)i(as:)869 3179 y Fi(Compression:)140 -b(400k)46 b(+)i(\()f(8)h(x)f(block)f(size)h(\))869 3387 -y(Decompression:)d(100k)i(+)i(\()f(4)h(x)f(block)f(size)h(\),)g(or)1585 -3491 y(100k)f(+)i(\()f(2.5)g(x)g(block)g(size)f(\))390 -3642 y Fj(Larger)27 b(blo)s(c)m(k)f(sizes)h(giv)m(e)h(rapidly)e -(diminishing)f(marginal)i(returns.)39 b(Most)27 b(of)f(the)h(com-)390 -3745 y(pression)e(comes)i(from)f(the)g(\014rst)f(t)m(w)m(o)i(or)f -(three)h(h)m(undred)d(k)h(of)h(blo)s(c)m(k)g(size,)i(a)e(fact)h(w)m -(orth)390 3849 y(b)s(earing)32 b(in)g(mind)f(when)g(using)h -Fi(bzip2)f Fj(on)h(small)g(mac)m(hines.)47 b(It)32 b(is)h(also)f(imp)s -(ortan)m(t)h(to)390 3953 y(appreciate)f(that)g(the)f(decompression)g -(memory)g(requiremen)m(t)h(is)f(set)g(at)h(compression)390 -4057 y(time)f(b)m(y)f(the)h(c)m(hoice)h(of)f(blo)s(c)m(k)g(size.)390 -4208 y(F)-8 b(or)43 b(\014les)f(compressed)h(with)f(the)g(default)h -(900k)g(blo)s(c)m(k)g(size,)k Fi(bunzip2)40 b Fj(will)j(require)390 -4311 y(ab)s(out)28 b(3700)i(kb)m(ytes)f(to)h(decompress.)40 -b(T)-8 b(o)29 b(supp)s(ort)e(decompression)i(of)g(an)m(y)f(\014le)h(on) -f(a)h(4)390 4415 y(megab)m(yte)h(mac)m(hine,)g Fi(bunzip2)c -Fj(has)i(an)h(option)g(to)g(decompress)g(using)f(appro)m(ximately)390 -4519 y(half)33 b(this)g(amoun)m(t)h(of)f(memory)-8 b(,)35 -b(ab)s(out)d(2300)j(kb)m(ytes.)50 b(Decompression)34 -b(sp)s(eed)e(is)h(also)390 4623 y(halv)m(ed,)h(so)f(y)m(ou)g(should)e -(use)i(this)f(option)h(only)g(where)g(necessary)-8 b(.)48 -b(The)33 b(relev)-5 b(an)m(t)34 b(\015ag)390 4727 y(is)c -Fi(-s)p Fj(.)390 4878 y(In)i(general,)i(try)f(and)e(use)h(the)h -(largest)h(blo)s(c)m(k)f(size)g(memory)g(constrain)m(ts)g(allo)m(w,)i -(since)390 4981 y(that)43 b(maximises)g(the)g(compression)f(ac)m(hiev)m -(ed.)79 b(Compression)42 b(and)g(decompression)390 5085 -y(sp)s(eed)30 b(are)g(virtually)h(una\013ected)g(b)m(y)f(blo)s(c)m(k)h -(size.)390 5236 y(Another)24 b(signi\014can)m(t)i(p)s(oin)m(t)e -(applies)g(to)i(\014les)e(whic)m(h)g(\014t)h(in)f(a)g(single)h(blo)s(c) -m(k)g({)g(that)g(means)390 5340 y(most)g(\014les)f(y)m(ou'd)h(encoun)m -(ter)g(using)f(a)g(large)i(blo)s(c)m(k)f(size.)39 b(The)25 -b(amoun)m(t)f(of)h(real)g(memory)p eop -%%Page: 7 8 -7 7 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(7)390 299 y Fj(touc)m(hed)36 -b(is)g(prop)s(ortional)g(to)g(the)g(size)h(of)f(the)g(\014le,)h(since)f -(the)g(\014le)g(is)f(smaller)i(than)e(a)390 403 y(blo)s(c)m(k.)48 -b(F)-8 b(or)33 b(example,)h(compressing)f(a)g(\014le)f(20,000)j(b)m -(ytes)f(long)f(with)f(the)h(\015ag)g Fi(-9)f Fj(will)390 -506 y(cause)26 b(the)g(compressor)f(to)i(allo)s(cate)g(around)e(7600k)i -(of)f(memory)-8 b(,)27 b(but)e(only)g(touc)m(h)i(400k)390 -610 y Fi(+)h Fj(20000)i(*)f(8)g(=)f(560)h(kb)m(ytes)g(of)g(it.)40 -b(Similarly)-8 b(,)30 b(the)f(decompressor)f(will)h(allo)s(cate)h -(3700k)390 714 y(but)g(only)g(touc)m(h)h(100k)h Fi(+)e -Fj(20000)j(*)d(4)h(=)f(180)i(kb)m(ytes.)390 865 y(Here)39 -b(is)f(a)g(table)h(whic)m(h)f(summarises)f(the)i(maxim)m(um)e(memory)h -(usage)h(for)f(di\013eren)m(t)390 969 y(blo)s(c)m(k)c(sizes.)50 -b(Also)33 b(recorded)g(is)h(the)f(total)i(compressed)e(size)h(for)g(14) -g(\014les)f(of)g(the)h(Cal-)390 1073 y(gary)39 b(T)-8 -b(ext)39 b(Compression)f(Corpus)g(totalling)i(3,141,622)i(b)m(ytes.)65 -b(This)38 b(column)h(giv)m(es)390 1176 y(some)g(feel)f(for)h(ho)m(w)f -(compression)g(v)-5 b(aries)39 b(with)f(blo)s(c)m(k)g(size.)65 -b(These)38 b(\014gures)g(tend)g(to)390 1280 y(understate)28 -b(the)g(adv)-5 b(an)m(tage)29 b(of)f(larger)h(blo)s(c)m(k)g(sizes)f -(for)g(larger)h(\014les,)g(since)f(the)g(Corpus)390 1384 -y(is)i(dominated)h(b)m(y)f(smaller)h(\014les.)1107 1535 -y Fi(Compress)141 b(Decompress)g(Decompress)f(Corpus)773 -1639 y(Flag)238 b(usage)285 b(usage)332 b(-s)48 b(usage)237 -b(Size)821 1846 y(-1)286 b(1200k)332 b(500k)429 b(350k)285 -b(914704)821 1950 y(-2)h(2000k)332 b(900k)429 b(600k)285 -b(877703)821 2054 y(-3)h(2800k)f(1300k)428 b(850k)285 -b(860338)821 2158 y(-4)h(3600k)f(1700k)380 b(1100k)285 -b(846899)821 2261 y(-5)h(4400k)f(2100k)380 b(1350k)285 -b(845160)821 2365 y(-6)h(5200k)f(2500k)380 b(1600k)285 -b(838626)821 2469 y(-7)h(6100k)f(2900k)380 b(1850k)285 -b(834096)821 2573 y(-8)h(6800k)f(3300k)380 b(2100k)285 -b(828642)821 2676 y(-9)h(7600k)f(3700k)380 b(2350k)285 -b(828642)390 2992 y Fe(RECO)m(VERING)38 b(D)m(A)-10 b(T)g(A)40 -b(FR)m(OM)h(D)m(AMA)m(GED)f(FILES)390 3179 y Fi(bzip2)24 -b Fj(compresses)i(\014les)g(in)f(blo)s(c)m(ks,)i(usually)e(900kb)m -(ytes)j(long.)40 b(Eac)m(h)25 b(blo)s(c)m(k)h(is)g(handled)390 -3283 y(indep)s(enden)m(tly)-8 b(.)42 b(If)31 b(a)g(media)g(or)h -(transmission)f(error)g(causes)g(a)g(m)m(ulti-blo)s(c)m(k)i -Fi(.bz2)d Fj(\014le)390 3387 y(to)35 b(b)s(ecome)f(damaged,)i(it)e(ma)m -(y)h(b)s(e)e(p)s(ossible)h(to)h(reco)m(v)m(er)h(data)f(from)f(the)g -(undamaged)390 3491 y(blo)s(c)m(ks)d(in)f(the)h(\014le.)390 -3642 y(The)g(compressed)g(represen)m(tation)i(of)f(eac)m(h)g(blo)s(c)m -(k)g(is)f(delimited)h(b)m(y)f(a)h(48-bit)g(pattern,)390 -3745 y(whic)m(h)27 b(mak)m(es)g(it)h(p)s(ossible)e(to)i(\014nd)d(the)j -(blo)s(c)m(k)f(b)s(oundaries)e(with)i(reasonable)h(certain)m(t)m(y)-8 -b(.)390 3849 y(Eac)m(h)33 b(blo)s(c)m(k)g(also)h(carries)g(its)f(o)m -(wn)g(32-bit)h(CR)m(C,)f(so)g(damaged)g(blo)s(c)m(ks)h(can)f(b)s(e)f -(distin-)390 3953 y(guished)e(from)g(undamaged)f(ones.)390 -4104 y Fi(bzip2recover)36 b Fj(is)j(a)g(simple)f(program)h(whose)g -(purp)s(ose)f(is)h(to)g(searc)m(h)h(for)f(blo)s(c)m(ks)g(in)390 -4208 y Fi(.bz2)33 b Fj(\014les,)j(and)e(write)h(eac)m(h)h(blo)s(c)m(k)f -(out)g(in)m(to)g(its)g(o)m(wn)g Fi(.bz2)e Fj(\014le.)54 -b(Y)-8 b(ou)34 b(can)h(then)f(use)390 4311 y Fi(bzip2)29 -b(-t)23 b Fj(to)h(test)h(the)f(in)m(tegrit)m(y)i(of)d(the)h(resulting)g -(\014les,)h(and)e(decompress)g(those)h(whic)m(h)390 4415 -y(are)31 b(undamaged.)390 4566 y Fi(bzip2recover)40 b -Fj(tak)m(es)45 b(a)f(single)g(argumen)m(t,)j(the)d(name)g(of)f(the)h -(damaged)g(\014le,)j(and)390 4670 y(writes)33 b(a)f(n)m(um)m(b)s(er)f -(of)h(\014les)h Fi(rec00001file.bz2)p Fj(,)28 b Fi(rec00002file.bz2)p -Fj(,)g(etc,)34 b(con)m(tain-)390 4774 y(ing)29 b(the)g(extracted)i(blo) -s(c)m(ks.)40 b(The)29 b(output)g(\014lenames)g(are)g(designed)g(so)g -(that)g(the)g(use)g(of)390 4878 y(wildcards)h(in)f(subsequen)m(t)h(pro) -s(cessing)g({)h(for)f(example,)h Fi(bzip2)e(-dc)g(rec*file.bz2)e(>)390 -4981 y(recovered_data)f Fj({)31 b(pro)s(cesses)g(the)f(\014les)h(in)f -(the)h(correct)h(order.)390 5132 y Fi(bzip2recover)37 -b Fj(should)i(b)s(e)h(of)h(most)f(use)g(dealing)h(with)f(large)i -Fi(.bz2)d Fj(\014les,)k(as)d(these)390 5236 y(will)31 -b(con)m(tain)i(man)m(y)e(blo)s(c)m(ks.)43 b(It)32 b(is)f(clearly)h -(futile)g(to)g(use)f(it)g(on)g(damaged)g(single-blo)s(c)m(k)390 -5340 y(\014les,)g(since)g(a)g(damaged)f(blo)s(c)m(k)h(cannot)h(b)s(e)e -(reco)m(v)m(ered.)43 b(If)31 b(y)m(ou)g(wish)f(to)h(minimise)g(an)m(y)p -eop -%%Page: 8 9 -8 8 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(8)390 299 y Fj(p)s(oten)m(tial)36 -b(data)f(loss)f(through)g(media)h(or)g(transmission)f(errors,)i(y)m(ou) -e(migh)m(t)i(consider)390 403 y(compressing)31 b(with)f(a)g(smaller)h -(blo)s(c)m(k)g(size.)390 711 y Fe(PERF)m(ORMANCE)39 b(NOTES)390 -897 y Fj(The)c(sorting)i(phase)e(of)g(compression)h(gathers)g(together) -i(similar)e(strings)g(in)f(the)h(\014le.)390 1001 y(Because)51 -b(of)g(this,)57 b(\014les)51 b(con)m(taining)h(v)m(ery)f(long)h(runs)d -(of)j(rep)s(eated)f(sym)m(b)s(ols,)56 b(lik)m(e)390 1105 -y Fi(")p Fj(aabaabaabaab)c(...)p Fi(")h Fj(\(rep)s(eated)g(sev)m(eral)h -(h)m(undred)c(times\))k(ma)m(y)f(compress)f(more)390 -1209 y(slo)m(wly)33 b(than)f(normal.)45 b(V)-8 b(ersions)33 -b(0.9.5)g(and)e(ab)s(o)m(v)m(e)i(fare)g(m)m(uc)m(h)f(b)s(etter)g(than)g -(previous)390 1312 y(v)m(ersions)54 b(in)g(this)g(resp)s(ect.)111 -b(The)54 b(ratio)h(b)s(et)m(w)m(een)g(w)m(orst-case)g(and)e(a)m(v)m -(erage-case)390 1416 y(compression)39 b(time)g(is)g(in)f(the)h(region)h -(of)e(10:1.)67 b(F)-8 b(or)40 b(previous)e(v)m(ersions,)k(this)c -(\014gure)390 1520 y(w)m(as)j(more)g(lik)m(e)g(100:1.)74 -b(Y)-8 b(ou)40 b(can)h(use)f(the)h Fi(-vvvv)e Fj(option)i(to)g(monitor) -g(progress)g(in)390 1624 y(great)32 b(detail,)f(if)g(y)m(ou)f(w)m(an)m -(t.)390 1775 y(Decompression)h(sp)s(eed)e(is)i(una\013ected)g(b)m(y)f -(these)h(phenomena.)390 1926 y Fi(bzip2)g Fj(usually)i(allo)s(cates)i -(sev)m(eral)f(megab)m(ytes)g(of)f(memory)g(to)h(op)s(erate)g(in,)f(and) -f(then)390 2030 y(c)m(harges)37 b(all)f(o)m(v)m(er)h(it)f(in)f(a)h -(fairly)g(random)f(fashion.)55 b(This)36 b(means)f(that)h(p)s -(erformance,)390 2133 y(b)s(oth)f(for)g(compressing)h(and)e -(decompressing,)j(is)f(largely)g(determined)g(b)m(y)f(the)h(sp)s(eed) -390 2237 y(at)27 b(whic)m(h)e(y)m(our)h(mac)m(hine)h(can)f(service)h -(cac)m(he)g(misses.)39 b(Because)27 b(of)f(this,)h(small)f(c)m(hanges) -390 2341 y(to)31 b(the)f(co)s(de)h(to)g(reduce)f(the)h(miss)e(rate)i -(ha)m(v)m(e)g(b)s(een)f(observ)m(ed)h(to)g(giv)m(e)g(disprop)s(ortion-) -390 2445 y(ately)i(large)h(p)s(erformance)e(impro)m(v)m(emen)m(ts.)47 -b(I)33 b(imagine)g Fi(bzip2)d Fj(will)j(p)s(erform)e(b)s(est)h(on)390 -2548 y(mac)m(hines)f(with)f(v)m(ery)h(large)g(cac)m(hes.)390 -2856 y Fe(CA)-14 b(VEA)k(TS)390 3043 y Fj(I/O)45 b(error)f(messages)h -(are)f(not)h(as)f(helpful)f(as)h(they)g(could)g(b)s(e.)81 -b Fi(bzip2)43 b Fj(tries)i(hard)390 3147 y(to)37 b(detect)g(I/O)f -(errors)h(and)d(exit)j(cleanly)-8 b(,)39 b(but)c(the)h(details)h(of)f -(what)f(the)h(problem)g(is)390 3251 y(sometimes)c(seem)e(rather)h -(misleading.)390 3402 y(This)c(man)m(ual)f(page)i(p)s(ertains)e(to)i(v) -m(ersion)g(1.0.2)h(of)e Fi(bzip2)p Fj(.)38 b(Compressed)26 -b(data)i(created)390 3505 y(b)m(y)22 b(this)h(v)m(ersion)g(is)f(en)m -(tirely)i(forw)m(ards)f(and)e(bac)m(kw)m(ards)i(compatible)g(with)f -(the)h(previous)390 3609 y(public)g(releases,)k(v)m(ersions)d(0.1pl2,)i -(0.9.0,)h(0.9.5,)g(1.0.0)f(and)d(1.0.1,)k(but)c(with)g(the)h(follo)m -(w-)390 3713 y(ing)h(exception:)39 b(0.9.0)27 b(and)c(ab)s(o)m(v)m(e)j -(can)f(correctly)i(decompress)d(m)m(ultiple)i(concatenated)390 -3817 y(compressed)36 b(\014les.)59 b(0.1pl2)37 b(cannot)g(do)f(this;)k -(it)d(will)f(stop)h(after)g(decompressing)f(just)390 -3921 y(the)31 b(\014rst)f(\014le)g(in)g(the)h(stream.)390 -4072 y Fi(bzip2recover)19 b Fj(v)m(ersions)j(prior)g(to)h(this)f(one,)i -(1.0.2,)h(used)c(32-bit)j(in)m(tegers)f(to)g(represen)m(t)390 -4175 y(bit)32 b(p)s(ositions)g(in)f(compressed)h(\014les,)g(so)g(it)g -(could)g(not)g(handle)f(compressed)h(\014les)f(more)390 -4279 y(than)39 b(512)h(megab)m(ytes)h(long.)68 b(V)-8 -b(ersion)39 b(1.0.2)i(and)d(ab)s(o)m(v)m(e)j(uses)d(64-bit)j(in)m(ts)e -(on)h(some)390 4383 y(platforms)e(whic)m(h)g(supp)s(ort)f(them)i(\(GNU) -g(supp)s(orted)e(targets,)42 b(and)c(Windo)m(ws\).)64 -b(T)-8 b(o)390 4487 y(establish)27 b(whether)g(or)f(not)i -Fi(bzip2recover)23 b Fj(w)m(as)k(built)f(with)h(suc)m(h)f(a)h -(limitation,)j(run)25 b(it)390 4591 y(without)33 b(argumen)m(ts.)50 -b(In)33 b(an)m(y)g(ev)m(en)m(t)i(y)m(ou)e(can)h(build)e(y)m(ourself)h -(an)g(unlimited)g(v)m(ersion)390 4694 y(if)28 b(y)m(ou)h(can)g -(recompile)h(it)f(with)f Fi(MaybeUInt64)e Fj(set)j(to)g(b)s(e)f(an)g -(unsigned)g(64-bit)i(in)m(teger.)390 5002 y Fe(A)m(UTHOR)390 -5189 y Fj(Julian)g(Sew)m(ard,)g Fi(jseward@acm.org)p -Fj(.)390 5340 y Fi(http://sources.redhat.co)o(m/bz)o(ip2)p -eop -%%Page: 9 10 -9 9 bop 150 -116 a Fl(Chapter)30 b(2:)41 b(Ho)m(w)31 -b(to)g(use)f Fi(bzip2)2375 b Fl(9)390 299 y Fj(The)22 -b(ideas)h(em)m(b)s(o)s(died)f(in)g Fi(bzip2)f Fj(are)i(due)e(to)j(\(at) -f(least\))h(the)f(follo)m(wing)g(p)s(eople:)37 b(Mic)m(hael)390 -403 y(Burro)m(ws)45 b(and)g(Da)m(vid)h(Wheeler)g(\(for)g(the)g(blo)s(c) -m(k)h(sorting)f(transformation\),)51 b(Da)m(vid)390 506 -y(Wheeler)42 b(\(again,)j(for)c(the)h(Hu\013man)e(co)s(der\),)45 -b(P)m(eter)e(F)-8 b(en)m(wic)m(k)43 b(\(for)e(the)h(structured)390 -610 y(co)s(ding)34 b(mo)s(del)h(in)f(the)g(original)i -Fi(bzip)p Fj(,)e(and)g(man)m(y)g(re\014nemen)m(ts\),)i(and)e(Alistair)h -(Mof-)390 714 y(fat,)45 b(Radford)d(Neal)g(and)f(Ian)h(Witten)h(\(for)f -(the)h(arithmetic)g(co)s(der)f(in)g(the)g(original)390 -818 y Fi(bzip)p Fj(\).)d(I)26 b(am)h(m)m(uc)m(h)g(indebted)f(for)h -(their)g(help,)g(supp)s(ort)e(and)h(advice.)40 b(See)27 -b(the)g(man)m(ual)390 922 y(in)k(the)h(source)f(distribution)h(for)f(p) -s(oin)m(ters)h(to)g(sources)f(of)h(do)s(cumen)m(tation.)44 -b(Christian)390 1025 y(v)m(on)27 b(Ro)s(ques)f(encouraged)g(me)h(to)g -(lo)s(ok)g(for)f(faster)h(sorting)g(algorithms,)h(so)e(as)h(to)g(sp)s -(eed)390 1129 y(up)32 b(compression.)50 b(Bela)34 b(Lubkin)d -(encouraged)j(me)g(to)g(impro)m(v)m(e)g(the)g(w)m(orst-case)h(com-)390 -1233 y(pression)e(p)s(erformance.)51 b(The)34 b Fi(bz*)f -Fj(scripts)h(are)g(deriv)m(ed)g(from)g(those)g(of)g(GNU)h -Fi(gzip)p Fj(.)390 1337 y(Man)m(y)40 b(p)s(eople)f(sen)m(t)h(patc)m -(hes,)i(help)s(ed)c(with)h(p)s(ortabilit)m(y)i(problems,)g(len)m(t)f -(mac)m(hines,)390 1440 y(ga)m(v)m(e)32 b(advice)f(and)f(w)m(ere)h -(generally)h(helpful.)p eop -%%Page: 10 11 -10 10 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(10)150 299 y Fg(3)80 b(Programming)53 -b(with)h Ff(libbzip2)150 568 y Fj(This)30 b(c)m(hapter)h(describ)s(es)f -(the)h(programming)f(in)m(terface)i(to)g Fi(libbzip2)p -Fj(.)150 725 y(F)-8 b(or)33 b(general)g(bac)m(kground)f(information,)h -(particularly)f(ab)s(out)g(memory)g(use)g(and)f(p)s(erformance)h(as-) -150 834 y(p)s(ects,)f(y)m(ou'd)f(b)s(e)g(w)m(ell)h(advised)f(to)i(read) -e(Chapter)g(2)h(as)f(w)m(ell.)150 1124 y Fk(3.1)68 b(T)-11 -b(op-lev)l(el)46 b(structure)150 1316 y Fi(libbzip2)32 -b Fj(is)i(a)g(\015exible)h(library)f(for)g(compressing)g(and)g -(decompressing)g(data)h(in)e(the)i Fi(bzip2)e Fj(data)150 -1426 y(format.)38 b(Although)22 b(pac)m(k)-5 b(aged)24 -b(as)e(a)g(single)h(en)m(tit)m(y)-8 b(,)26 b(it)d(helps)f(to)h(regard)f -(the)h(library)f(as)g(three)h(separate)150 1535 y(parts:)41 -b(the)30 b(lo)m(w)h(lev)m(el)i(in)m(terface,)f(and)d(the)i(high)f(lev)m -(el)i(in)m(terface,)h(and)c(some)i(utilit)m(y)h(functions.)150 -1692 y(The)37 b(structure)g(of)g Fi(libbzip2)p Fj('s)e(in)m(terfaces)j -(is)f(similar)g(to)h(that)f(of)g(Jean-loup)g(Gailly's)i(and)d(Mark)150 -1802 y(Adler's)30 b(excellen)m(t)j Fi(zlib)c Fj(library)-8 -b(.)150 1959 y(All)29 b(externally)h(visible)g(sym)m(b)s(ols)e(ha)m(v)m -(e)i(names)f(b)s(eginning)f Fi(BZ2_)p Fj(.)39 b(This)29 -b(is)g(new)f(in)h(v)m(ersion)h(1.0.)41 b(The)150 2068 -y(in)m(ten)m(tion)32 b(is)f(to)g(minimise)f(p)s(ollution)h(of)f(the)h -(namespaces)f(of)h(library)f(clien)m(ts.)150 2321 y Fe(3.1.1)63 -b(Lo)m(w-lev)m(el)39 b(summary)150 2514 y Fj(This)52 -b(in)m(terface)i(pro)m(vides)e(services)h(for)f(compressing)g(and)g -(decompressing)g(data)g(in)g(memory)-8 b(.)150 2623 y(There's)42 -b(no)g(pro)m(vision)g(for)g(dealing)g(with)f(\014les,)k(streams)d(or)g -(an)m(y)g(other)g(I/O)h(mec)m(hanisms,)h(just)150 2733 -y(straigh)m(t)51 b(memory-to-memory)g(w)m(ork.)98 b(In)48 -b(fact,)55 b(this)49 b(part)h(of)f(the)h(library)f(can)g(b)s(e)f -(compiled)150 2843 y(without)31 b(inclusion)f(of)g Fi(stdio.h)p -Fj(,)f(whic)m(h)h(ma)m(y)h(b)s(e)f(helpful)f(for)i(em)m(b)s(edded)e -(applications.)150 2999 y(The)h(lo)m(w-lev)m(el)k(part)c(of)h(the)f -(library)g(has)g(no)h(global)g(v)-5 b(ariables)31 b(and)e(is)i -(therefore)g(thread-safe.)150 3156 y(Six)26 b(routines)h(mak)m(e)h(up)e -(the)h(lo)m(w)g(lev)m(el)i(in)m(terface:)41 b Fi(BZ2_bzCompressInit)p -Fj(,)22 b Fi(BZ2_bzCompress)p Fj(,)i(and)150 3266 y Fi -(BZ2_bzCompressEnd)i Fj(for)k(compression,)h(and)f(a)g(corresp)s -(onding)g(trio)h Fi(BZ2_bzDecompressInit)p Fj(,)150 3375 -y Fi(BZ2_bzDecompress)36 b Fj(and)j Fi(BZ2_bzDecompressEnd)c -Fj(for)41 b(decompression.)70 b(The)40 b Fi(*Init)f Fj(functions)150 -3485 y(allo)s(cate)k(memory)e(for)g(compression/decompression)h(and)e -(do)h(other)h(initialisations,)k(whilst)41 b(the)150 -3595 y Fi(*End)29 b Fj(functions)h(close)i(do)m(wn)e(op)s(erations)h -(and)e(release)j(memory)-8 b(.)150 3751 y(The)35 b(real)h(w)m(ork)f(is) -g(done)g(b)m(y)g Fi(BZ2_bzCompress)d Fj(and)i Fi(BZ2_bzDecompress)p -Fj(.)50 b(These)36 b(compress)f(and)150 3861 y(decompress)28 -b(data)g(from)f(a)h(user-supplied)e(input)h(bu\013er)g(to)h(a)g -(user-supplied)e(output)i(bu\013er.)39 b(These)150 3971 -y(bu\013ers)31 b(can)g(b)s(e)g(an)m(y)h(size;)h(arbitrary)f(quan)m -(tities)h(of)f(data)g(are)g(handled)e(b)m(y)i(making)f(rep)s(eated)h -(calls)150 4080 y(to)g(these)g(functions.)43 b(This)31 -b(is)g(a)h(\015exible)f(mec)m(hanism)h(allo)m(wing)g(a)g(consumer-pull) -f(st)m(yle)h(of)g(activit)m(y)-8 b(,)150 4190 y(or)31 -b(pro)s(ducer-push,)d(or)j(a)f(mixture)h(of)f(b)s(oth.)150 -4443 y Fe(3.1.2)63 b(High-lev)m(el)41 b(summary)150 4635 -y Fj(This)c(in)m(terface)i(pro)m(vides)e(some)h(handy)e(wrapp)s(ers)f -(around)i(the)g(lo)m(w-lev)m(el)j(in)m(terface)f(to)f(facilitate)150 -4745 y(reading)25 b(and)f(writing)h Fi(bzip2)f Fj(format)h(\014les)g -(\()p Fi(.bz2)f Fj(\014les\).)39 b(The)25 b(routines)h(pro)m(vide)f(ho) -s(oks)g(to)g(facilitate)150 4854 y(reading)42 b(\014les)f(in)g(whic)m -(h)h(the)g Fi(bzip2)e Fj(data)i(stream)g(is)f(em)m(b)s(edded)g(within)g -(some)h(larger-scale)i(\014le)150 4964 y(structure,)31 -b(or)g(where)f(there)h(are)g(m)m(ultiple)g Fi(bzip2)e -Fj(data)h(streams)h(concatenated)i(end-to-end.)150 5121 -y(F)-8 b(or)31 b(reading)g(\014les,)f Fi(BZ2_bzReadOpen)p -Fj(,)d Fi(BZ2_bzRead)p Fj(,)h Fi(BZ2_bzReadClose)e Fj(and)150 -5230 y Fi(BZ2_bzReadGetUnused)18 b Fj(are)23 b(supplied.)37 -b(F)-8 b(or)24 b(writing)f(\014les,)i Fi(BZ2_bzWriteOpen)p -Fj(,)20 b Fi(BZ2_bzWrite)g Fj(and)150 5340 y Fi(BZ2_bzWriteFinish)26 -b Fj(are)k(a)m(v)-5 b(ailable.)p eop -%%Page: 11 12 -11 11 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(11)150 299 y Fj(As)32 -b(with)h(the)g(lo)m(w-lev)m(el)j(library)-8 b(,)33 b(no)g(global)h(v)-5 -b(ariables)33 b(are)g(used)f(so)h(the)g(library)g(is)g(p)s(er)f(se)h -(thread-)150 408 y(safe.)62 b(Ho)m(w)m(ev)m(er,)41 b(if)c(I/O)h(errors) -g(o)s(ccur)f(whilst)g(reading)h(or)f(writing)h(the)f(underlying)g -(compressed)150 518 y(\014les,)j(y)m(ou)e(ma)m(y)h(ha)m(v)m(e)f(to)h -(consult)g Fi(errno)d Fj(to)j(determine)f(the)h(cause)f(of)g(the)g -(error.)64 b(In)37 b(that)i(case,)150 628 y(y)m(ou'd)30 -b(need)h(a)f(C)g(library)g(whic)m(h)g(correctly)j(supp)s(orts)c -Fi(errno)g Fj(in)h(a)g(m)m(ultithreaded)h(en)m(vironmen)m(t.)150 -784 y(T)-8 b(o)77 b(mak)m(e)f(the)g(library)g(a)f(little)j(simpler)e -(and)e(more)j(p)s(ortable,)87 b Fi(BZ2_bzReadOpen)72 -b Fj(and)150 894 y Fi(BZ2_bzWriteOpen)34 b Fj(require)39 -b(y)m(ou)g(to)g(pass)f(them)h(\014le)f(handles)g(\()p -Fi(FILE*)p Fj(s\))f(whic)m(h)h(ha)m(v)m(e)i(previously)150 -1004 y(b)s(een)56 b(op)s(ened)g(for)h(reading)f(or)h(writing)g(resp)s -(ectiv)m(ely)-8 b(.)122 b(That)57 b(a)m(v)m(oids)h(p)s(ortabilit)m(y)f -(problems)150 1113 y(asso)s(ciated)33 b(with)g(\014le)f(op)s(erations)h -(and)e(\014le)i(attributes,)h(whilst)e(not)h(b)s(eing)f(m)m(uc)m(h)g -(of)h(an)f(imp)s(osition)150 1223 y(on)e(the)h(programmer.)150 -1474 y Fe(3.1.3)63 b(Utilit)m(y)40 b(functions)h(summary)150 -1666 y Fj(F)-8 b(or)43 b(v)m(ery)h(simple)e(needs,)k -Fi(BZ2_bzBuffToBuffCompres)o(s)36 b Fj(and)42 b Fi -(BZ2_bzBuffToBuffDecompres)o(s)150 1776 y Fj(are)26 b(pro)m(vided.)39 -b(These)26 b(compress)f(data)h(in)f(memory)h(from)f(one)h(bu\013er)f -(to)h(another)g(bu\013er)f(in)g(a)g(single)150 1885 y(function)37 -b(call.)62 b(Y)-8 b(ou)38 b(should)e(assess)h(whether)g(these)h -(functions)f(ful\014ll)f(y)m(our)i(memory-to-memory)150 -1995 y(compression/decompression)22 b(requiremen)m(ts)g(b)s(efore)f(in) -m(v)m(esting)i(e\013ort)f(in)f(understanding)f(the)h(more)150 -2105 y(general)31 b(but)f(more)h(complex)g(lo)m(w-lev)m(el)i(in)m -(terface.)150 2261 y(Y)-8 b(oshiok)j(a)47 b(Tsuneo)g(\()p -Fi(QWF00133@niftyserve.or.jp)40 b Fj(/)47 b Fi -(tsuneo-y@is.aist-nara.ac)o(.jp)p Fj(\))40 b(has)150 -2371 y(con)m(tributed)f(some)f(functions)g(to)h(giv)m(e)h(b)s(etter)e -Fi(zlib)f Fj(compatibilit)m(y)-8 b(.)67 b(These)38 b(functions)g(are)g -Fi(BZ2_)150 2481 y(bzopen)p Fj(,)g Fi(BZ2_bzread)p Fj(,)f -Fi(BZ2_bzwrite)p Fj(,)g Fi(BZ2_bzflush)p Fj(,)g Fi(BZ2_bzclose)p -Fj(,)g Fi(BZ2_bzerror)e Fj(and)i Fi(BZ2_)150 2590 y(bzlibVersion)p -Fj(.)44 b(Y)-8 b(ou)33 b(ma)m(y)g(\014nd)e(these)i(functions)g(more)g -(con)m(v)m(enien)m(t)i(for)e(simple)f(\014le)h(reading)g(and)150 -2700 y(writing,)d(than)g(those)g(in)f(the)h(high-lev)m(el)i(in)m -(terface.)42 b(These)30 b(functions)f(are)h(not)g(\(y)m(et\))i -(o\016cially)f(part)150 2809 y(of)h(the)g(library)-8 -b(,)33 b(and)e(are)h(minimally)g(do)s(cumen)m(ted)g(here.)45 -b(If)31 b(they)h(break,)h(y)m(ou)f(get)h(to)g(k)m(eep)f(all)h(the)150 -2919 y(pieces.)42 b(I)30 b(hop)s(e)g(to)h(do)s(cumen)m(t)f(them)h(prop) -s(erly)e(when)h(time)h(p)s(ermits.)150 3076 y(Y)-8 b(oshiok)j(a)27 -b(also)g(con)m(tributed)g(mo)s(di\014cations)g(to)g(allo)m(w)h(the)e -(library)h(to)g(b)s(e)f(built)g(as)g(a)h(Windo)m(ws)f(DLL.)150 -3362 y Fk(3.2)68 b(Error)45 b(handling)150 3554 y Fj(The)22 -b(library)g(is)g(designed)g(to)i(reco)m(v)m(er)g(cleanly)f(in)f(all)h -(situations,)i(including)c(the)i(w)m(orst-case)h(situation)150 -3664 y(of)k(decompressing)g(random)f(data.)40 b(I'm)27 -b(not)h(100\045)h(sure)e(that)i(it)f(can)g(alw)m(a)m(ys)g(do)g(this,)g -(so)g(y)m(ou)g(migh)m(t)150 3774 y(w)m(an)m(t)i(to)g(add)e(a)i(signal)f -(handler)g(to)h(catc)m(h)g(segmen)m(tation)i(violations)f(during)d -(decompression)h(if)h(y)m(ou)150 3883 y(are)c(feeling)g(esp)s(ecially)g -(paranoid.)39 b(I)25 b(w)m(ould)g(b)s(e)g(in)m(terested)i(in)e(hearing) -g(more)h(ab)s(out)f(the)h(robustness)150 3993 y(of)31 -b(the)f(library)h(to)g(corrupted)f(compressed)g(data.)150 -4150 y(V)-8 b(ersion)38 b(1.0)h(is)e(m)m(uc)m(h)h(more)g(robust)f(in)g -(this)h(resp)s(ect)g(than)f(0.9.0)i(or)f(0.9.5.)64 b(In)m(v)m -(estigations)40 b(with)150 4259 y(Chec)m(k)m(er)30 b(\(a)f(to)s(ol)h -(for)f(detecting)h(problems)f(with)f(memory)h(managemen)m(t,)i(similar) -e(to)h(Purify\))e(indi-)150 4369 y(cate)34 b(that,)g(at)f(least)g(for)g -(the)g(few)f(\014les)g(I)h(tested,)h(all)f(single-bit)h(errors)e(in)g -(the)h(decompressed)g(data)150 4478 y(are)28 b(caugh)m(t)h(prop)s(erly) --8 b(,)29 b(with)f(no)g(segmen)m(tation)i(faults,)f(no)f(reads)g(of)g -(uninitialised)g(data)g(and)g(no)g(out)150 4588 y(of)j(range)h(reads)f -(or)h(writes.)44 b(So)31 b(it's)h(certainly)h(m)m(uc)m(h)e(impro)m(v)m -(ed,)i(although)e(I)h(w)m(ouldn't)f(claim)h(it)g(to)150 -4698 y(b)s(e)e(totally)i(b)s(om)m(bpro)s(of.)150 4854 -y(The)24 b(\014le)g Fi(bzlib.h)e Fj(con)m(tains)j(all)f(de\014nitions)g -(needed)g(to)g(use)g(the)g(library)-8 b(.)39 b(In)24 -b(particular,)i(y)m(ou)e(should)150 4964 y(de\014nitely)31 -b(not)f(include)h Fi(bzlib_private.h)p Fj(.)150 5121 -y(In)38 b Fi(bzlib.h)p Fj(,)g(the)g(v)-5 b(arious)38 -b(return)g(v)-5 b(alues)38 b(are)g(de\014ned.)62 b(The)38 -b(follo)m(wing)i(list)f(is)f(not)g(in)m(tended)g(as)150 -5230 y(an)33 b(exhaustiv)m(e)h(description)g(of)g(the)g(circumstances)g -(in)f(whic)m(h)g(a)h(giv)m(en)g(v)-5 b(alue)34 b(ma)m(y)g(b)s(e)f -(returned)g({)150 5340 y(those)h(descriptions)f(are)h(giv)m(en)g -(later.)50 b(Rather,)34 b(it)g(is)f(in)m(tended)g(to)h(con)m(v)m(ey)h -(the)e(rough)g(meaning)g(of)p eop -%%Page: 12 13 -12 12 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(12)150 299 y Fj(eac)m(h)35 -b(return)f(v)-5 b(alue.)53 b(The)34 b(\014rst)g(\014v)m(e)h(actions)g -(are)g(normal)f(and)f(not)i(in)m(tended)f(to)i(denote)f(an)e(error)150 -408 y(situation.)150 592 y Fi(BZ_OK)240 b Fj(The)30 b(requested)h -(action)h(w)m(as)e(completed)i(successfully)-8 b(.)150 -756 y Fi(BZ_RUN_OK)150 866 y(BZ_FLUSH_OK)150 975 y(BZ_FINISH_OK)630 -1085 y Fj(In)22 b Fi(BZ2_bzCompress)p Fj(,)e(the)j(requested)g -(\015ush/\014nish/nothing-sp)s(ecial)e(action)j(w)m(as)e(com-)630 -1194 y(pleted)31 b(successfully)-8 b(.)150 1358 y Fi(BZ_STREAM_END)630 -1468 y Fj(Compression)38 b(of)g(data)h(w)m(as)f(completed,)j(or)e(the)f -(logical)j(stream)e(end)e(w)m(as)h(detected)630 1577 -y(during)29 b(decompression.)150 1761 y(The)h(follo)m(wing)i(return)e -(v)-5 b(alues)31 b(indicate)g(an)f(error)h(of)f(some)h(kind.)150 -1945 y Fi(BZ_CONFIG_ERROR)630 2055 y Fj(Indicates)46 -b(that)h(the)f(library)f(has)h(b)s(een)f(improp)s(erly)g(compiled)h(on) -g(y)m(our)g(platform)630 2164 y({)k(a)h(ma)5 b(jor)50 -b(con\014guration)h(error.)100 b(Sp)s(eci\014cally)-8 -b(,)56 b(it)51 b(means)e(that)i Fi(sizeof\(char\))p Fj(,)630 -2274 y Fi(sizeof\(short\))43 b Fj(and)i Fi(sizeof\(int\))e -Fj(are)k(not)g(1,)k(2)46 b(and)g(4)h(resp)s(ectiv)m(ely)-8 -b(,)52 b(as)47 b(they)630 2384 y(should)26 b(b)s(e.)38 -b(Note)29 b(that)e(the)g(library)f(should)g(still)h(w)m(ork)g(prop)s -(erly)f(on)h(64-bit)h(platforms)630 2493 y(whic)m(h)43 -b(follo)m(w)h(the)g(LP64)f(programming)h(mo)s(del)e({)i(that)g(is,)i -(where)d Fi(sizeof\(long\))630 2603 y Fj(and)32 b Fi(sizeof\(void*\))e -Fj(are)j(8.)50 b(Under)32 b(LP64,)j Fi(sizeof\(int\))30 -b Fj(is)j(still)h(4,)h(so)e Fi(libbzip2)p Fj(,)630 2712 -y(whic)m(h)d(do)s(esn't)g(use)h(the)f Fi(long)f Fj(t)m(yp)s(e,)i(is)g -(OK.)150 2876 y Fi(BZ_SEQUENCE_ERROR)630 2986 y Fj(When)41 -b(using)g(the)h(library)-8 b(,)45 b(it)d(is)g(imp)s(ortan)m(t)g(to)g -(call)h(the)f(functions)f(in)g(the)h(correct)630 3095 -y(sequence)25 b(and)f(with)g(data)h(structures)g(\(bu\013ers)f(etc\))i -(in)f(the)g(correct)h(states.)40 b Fi(libbzip2)630 3205 -y Fj(c)m(hec)m(ks)25 b(as)e(m)m(uc)m(h)g(as)g(it)h(can)f(to)h(ensure)f -(this)g(is)g(happ)s(ening,)g(and)g(returns)f Fi(BZ_SEQUENCE_)630 -3314 y(ERROR)35 b Fj(if)i(not.)59 b(Co)s(de)36 b(whic)m(h)h(complies)g -(precisely)h(with)e(the)h(function)f(seman)m(tics,)k(as)630 -3424 y(detailed)c(b)s(elo)m(w,)h(should)e(nev)m(er)g(receiv)m(e)j(this) -d(v)-5 b(alue;)38 b(suc)m(h)d(an)g(ev)m(en)m(t)i(denotes)f(buggy)630 -3534 y(co)s(de)31 b(whic)m(h)f(y)m(ou)g(should)g(in)m(v)m(estigate.)150 -3697 y Fi(BZ_PARAM_ERROR)630 3807 y Fj(Returned)41 b(when)f(a)h -(parameter)h(to)g(a)g(function)f(call)h(is)f(out)h(of)f(range)h(or)f -(otherwise)630 3917 y(manifestly)35 b(incorrect.)55 b(As)34 -b(with)g Fi(BZ_SEQUENCE_ERROR)p Fj(,)d(this)k(denotes)g(a)f(bug)g(in)h -(the)630 4026 y(clien)m(t)24 b(co)s(de.)38 b(The)22 b(distinction)h(b)s -(et)m(w)m(een)f Fi(BZ_PARAM_ERROR)d Fj(and)i Fi(BZ_SEQUENCE_ERROR)630 -4136 y Fj(is)30 b(a)h(bit)f(hazy)-8 b(,)32 b(but)d(still)j(w)m(orth)e -(making.)150 4300 y Fi(BZ_MEM_ERROR)630 4409 y Fj(Returned)44 -b(when)g(a)h(request)g(to)g(allo)s(cate)i(memory)e(failed.)83 -b(Note)46 b(that)g(the)f(quan-)630 4519 y(tit)m(y)d(of)f(memory)f -(needed)g(to)i(decompress)e(a)g(stream)h(cannot)g(b)s(e)f(determined)h -(un)m(til)630 4628 y(the)f(stream's)h(header)e(has)h(b)s(een)f(read.)69 -b(So)39 b Fi(BZ2_bzDecompress)d Fj(and)j Fi(BZ2_bzRead)630 -4738 y Fj(ma)m(y)29 b(return)f Fi(BZ_MEM_ERROR)d Fj(ev)m(en)30 -b(though)e(some)h(of)g(the)g(compressed)g(data)g(has)f(b)s(een)630 -4847 y(read.)76 b(The)42 b(same)g(is)g(not)g(true)g(for)h(compression;) -48 b(once)43 b Fi(BZ2_bzCompressInit)37 b Fj(or)630 4957 -y Fi(BZ2_bzWriteOpen)26 b Fj(ha)m(v)m(e)32 b(successfully)e(completed,) -i Fi(BZ_MEM_ERROR)27 b Fj(cannot)k(o)s(ccur.)150 5121 -y Fi(BZ_DATA_ERROR)630 5230 y Fj(Returned)d(when)g(a)h(data)g(in)m -(tegrit)m(y)i(error)e(is)g(detected)h(during)d(decompression.)41 -b(Most)630 5340 y(imp)s(ortan)m(tly)-8 b(,)32 b(this)e(means)g(when)f -(stored)i(and)e(computed)h(CR)m(Cs)g(for)g(the)h(data)f(do)h(not)p -eop -%%Page: 13 14 -13 13 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(13)630 299 y Fj(matc)m(h.)41 -b(This)28 b(v)-5 b(alue)28 b(is)h(also)g(returned)e(up)s(on)g -(detection)j(of)f(an)m(y)f(other)h(anomaly)g(in)f(the)630 -408 y(compressed)i(data.)150 560 y Fi(BZ_DATA_ERROR_MAGIC)630 -670 y Fj(As)j(a)g(sp)s(ecial)g(case)h(of)f Fi(BZ_DATA_ERROR)p -Fj(,)d(it)k(is)f(sometimes)h(useful)f(to)h(kno)m(w)f(when)f(the)630 -779 y(compressed)c(stream)g(do)s(es)g(not)g(start)g(with)g(the)g -(correct)i(magic)e(b)m(ytes)h(\()p Fi('B')h('Z')f('h')p -Fj(\).)150 931 y Fi(BZ_IO_ERROR)630 1040 y Fj(Returned)h(b)m(y)h -Fi(BZ2_bzRead)d Fj(and)h Fi(BZ2_bzWrite)f Fj(when)i(there)h(is)g(an)f -(error)h(reading)g(or)630 1150 y(writing)e(in)f(the)h(compressed)g -(\014le,)g(and)f(b)m(y)h Fi(BZ2_bzReadOpen)24 b Fj(and)k -Fi(BZ2_bzWriteOpen)630 1259 y Fj(for)h(attempts)i(to)f(use)f(a)h -(\014le)f(for)h(whic)m(h)f(the)h(error)f(indicator)i(\(viz,)f -Fi(ferror\(f\))p Fj(\))e(is)h(set.)630 1369 y(On)h(receipt)i(of)f -Fi(BZ_IO_ERROR)p Fj(,)c(the)k(caller)h(should)e(consult)h -Fi(errno)e Fj(and/or)h Fi(perror)f Fj(to)630 1479 y(acquire)i(op)s -(erating-system)h(sp)s(eci\014c)e(information)h(ab)s(out)f(the)g -(problem.)150 1630 y Fi(BZ_UNEXPECTED_EOF)630 1740 y -Fj(Returned)j(b)m(y)g Fi(BZ2_bzRead)e Fj(when)h(the)i(compressed)f -(\014le)h(\014nishes)e(b)s(efore)h(the)h(logical)630 -1849 y(end)c(of)g(stream)h(is)g(detected.)150 2001 y -Fi(BZ_OUTBUFF_FULL)630 2110 y Fj(Returned)d(b)m(y)g Fi -(BZ2_bzBuffToBuffCompress)22 b Fj(and)27 b Fi -(BZ2_bzBuffToBuffDecompres)o(s)630 2220 y Fj(to)k(indicate)h(that)f -(the)f(output)g(data)h(will)g(not)g(\014t)f(in)m(to)h(the)g(output)f -(bu\013er)g(pro)m(vided.)150 2492 y Fk(3.3)68 b(Lo)l(w-lev)l(el)47 -b(in)l(terface)150 2766 y Fe(3.3.1)63 b Fd(BZ2_bzCompressInit)390 -2953 y Fi(typedef)533 3057 y(struct)46 b({)676 3161 y(char)h(*next_in;) -676 3264 y(unsigned)f(int)h(avail_in;)676 3368 y(unsigned)f(int)h -(total_in_lo32;)676 3472 y(unsigned)f(int)h(total_in_hi32;)676 -3680 y(char)g(*next_out;)676 3783 y(unsigned)f(int)h(avail_out;)676 -3887 y(unsigned)f(int)h(total_out_lo32;)676 3991 y(unsigned)f(int)h -(total_out_hi32;)676 4198 y(void)g(*state;)676 4406 y(void)g -(*\(*bzalloc\)\(void)c(*,int,int\);)676 4510 y(void)k -(\(*bzfree\)\(void)d(*,void)i(*\);)676 4614 y(void)h(*opaque;)533 -4717 y(})533 4821 y(bz_stream;)390 5029 y(int)g(BZ2_bzCompressInit)c -(\()k(bz_stream)e(*strm,)1583 5132 y(int)i(blockSize100k,)1583 -5236 y(int)g(verbosity,)1583 5340 y(int)g(workFactor)e(\);)p -eop -%%Page: 14 15 -14 14 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(14)150 456 y Fj(Prepares)31 -b(for)g(compression.)42 b(The)30 b Fi(bz_stream)e Fj(structure)j(holds) -g(all)g(data)g(p)s(ertaining)g(to)g(the)g(com-)150 565 -y(pression)k(activit)m(y)-8 b(.)59 b(A)35 b Fi(bz_stream)d -Fj(structure)k(should)e(b)s(e)h(allo)s(cated)i(and)e(initialised)h -(prior)f(to)i(the)150 675 y(call.)65 b(The)38 b(\014elds)g(of)g -Fi(bz_stream)e Fj(comprise)i(the)h(en)m(tiret)m(y)h(of)f(the)f -(user-visible)h(data.)64 b Fi(state)37 b Fj(is)h(a)150 -784 y(p)s(oin)m(ter)31 b(to)g(the)g(priv)-5 b(ate)31 -b(data)f(structures)h(required)f(for)h(compression.)150 -941 y(Custom)36 b(memory)g(allo)s(cators)i(are)f(supp)s(orted,)f(via)h -(\014elds)e Fi(bzalloc)p Fj(,)h Fi(bzfree)p Fj(,)g(and)g -Fi(opaque)p Fj(.)56 b(The)150 1051 y(v)-5 b(alue)32 b -Fi(opaque)e Fj(is)i(passed)f(to)i(as)f(the)g(\014rst)g(argumen)m(t)g -(to)h(all)f(calls)h(to)g Fi(bzalloc)d Fj(and)h Fi(bzfree)p -Fj(,)g(but)g(is)150 1160 y(otherwise)c(ignored)g(b)m(y)g(the)g(library) --8 b(.)39 b(The)27 b(call)g Fi(bzalloc)i(\()h(opaque,)e(n,)i(m)g(\))c -Fj(is)h(exp)s(ected)g(to)g(return)150 1270 y(a)h(p)s(oin)m(ter)g -Fi(p)f Fj(to)i Fi(n)h(*)g(m)d Fj(b)m(ytes)h(of)g(memory)-8 -b(,)29 b(and)e Fi(bzfree)h(\()i(opaque,)f(p)h(\))d Fj(should)g(free)h -(that)g(memory)-8 b(.)150 1427 y(If)32 b(y)m(ou)h(don't)g(w)m(an)m(t)g -(to)h(use)e(a)h(custom)g(memory)g(allo)s(cator,)i(set)e -Fi(bzalloc)p Fj(,)f Fi(bzfree)e Fj(and)i Fi(opaque)f -Fj(to)150 1537 y Fi(NULL)p Fj(,)e(and)h(the)h(library)f(will)h(then)f -(use)g(the)h(standard)e Fi(malloc)p Fj(/)p Fi(free)f -Fj(routines.)150 1693 y(Before)37 b(calling)h Fi(BZ2_bzCompressInit)p -Fj(,)33 b(\014elds)k Fi(bzalloc)p Fj(,)f Fi(bzfree)f -Fj(and)h Fi(opaque)e Fj(should)i(b)s(e)g(\014lled)150 -1803 y(appropriately)-8 b(,)35 b(as)e(just)f(describ)s(ed.)48 -b(Up)s(on)31 b(return,)j(the)f(in)m(ternal)h(state)g(will)f(ha)m(v)m(e) -h(b)s(een)e(allo)s(cated)150 1913 y(and)43 b(initialised,)49 -b(and)43 b Fi(total_in_lo32)p Fj(,)h Fi(total_in_hi32)p -Fj(,)g Fi(total_out_lo32)c Fj(and)j Fi(total_out_)150 -2022 y(hi32)35 b Fj(will)i(ha)m(v)m(e)g(b)s(een)f(set)h(to)g(zero.)59 -b(These)37 b(four)f(\014elds)f(are)i(used)f(b)m(y)g(the)g(library)h(to) -g(inform)f(the)150 2132 y(caller)41 b(of)e(the)h(total)h(amoun)m(t)f -(of)f(data)h(passed)e(in)m(to)j(and)d(out)i(of)f(the)h(library)-8 -b(,)42 b(resp)s(ectiv)m(ely)-8 b(.)70 b(Y)-8 b(ou)150 -2241 y(should)34 b(not)h(try)h(to)g(c)m(hange)g(them.)55 -b(As)34 b(of)h(v)m(ersion)h(1.0,)i(64-bit)e(coun)m(ts)g(are)f(main)m -(tained,)i(ev)m(en)f(on)150 2351 y(32-bit)j(platforms,)g(using)e(the)h -Fi(_hi32)d Fj(\014elds)i(to)h(store)h(the)e(upp)s(er)f(32)i(bits)f(of)h -(the)f(coun)m(t.)63 b(So,)39 b(for)150 2460 y(example,)31 -b(the)g(total)h(amoun)m(t)f(of)f(data)h(in)f(is)g Fi(\(total_in_hi32)d -(<<)j(32\))f(+)h(total_in_lo32)p Fj(.)150 2617 y(P)m(arameter)37 -b Fi(blockSize100k)31 b Fj(sp)s(eci\014es)k(the)h(blo)s(c)m(k)g(size)g -(to)g(b)s(e)e(used)h(for)g(compression.)55 b(It)36 b(should)150 -2727 y(b)s(e)27 b(a)h(v)-5 b(alue)28 b(b)s(et)m(w)m(een)g(1)g(and)f(9)h -(inclusiv)m(e,)i(and)c(the)i(actual)h(blo)s(c)m(k)f(size)h(used)e(is)g -(100000)k(x)c(this)h(\014gure.)150 2836 y(9)j(giv)m(es)h(the)e(b)s(est) -g(compression)h(but)f(tak)m(es)h(most)g(memory)-8 b(.)150 -2993 y(P)m(arameter)28 b Fi(verbosity)c Fj(should)h(b)s(e)g(set)i(to)g -(a)g(n)m(um)m(b)s(er)e(b)s(et)m(w)m(een)i(0)f(and)g(4)g(inclusiv)m(e.) -41 b(0)26 b(is)g(silen)m(t,)j(and)150 3103 y(greater)d(n)m(um)m(b)s -(ers)e(giv)m(e)i(increasingly)g(v)m(erb)s(ose)f(monitoring/debugging)h -(output.)38 b(If)25 b(the)g(library)f(has)150 3212 y(b)s(een)30 -b(compiled)h(with)f Fi(-DBZ_NO_STDIO)p Fj(,)d(no)j(suc)m(h)g(output)g -(will)h(app)s(ear)f(for)g(an)m(y)g(v)m(erb)s(osit)m(y)i(setting.)150 -3369 y(P)m(arameter)g Fi(workFactor)c Fj(con)m(trols)k(ho)m(w)f(the)g -(compression)g(phase)f(b)s(eha)m(v)m(es)h(when)e(presen)m(ted)i(with) -150 3479 y(w)m(orst)38 b(case,)j(highly)c(rep)s(etitiv)m(e,)42 -b(input)37 b(data.)62 b(If)38 b(compression)g(runs)e(in)m(to)i -(di\016culties)h(caused)e(b)m(y)150 3588 y(rep)s(etitiv)m(e)46 -b(data,)i(the)c(library)f(switc)m(hes)i(from)f(the)g(standard)f -(sorting)h(algorithm)h(to)g(a)f(fallbac)m(k)150 3698 -y(algorithm.)c(The)27 b(fallbac)m(k)g(is)f(slo)m(w)m(er)i(than)d(the)i -(standard)e(algorithm)j(b)m(y)e(p)s(erhaps)e(a)i(factor)i(of)e(three,) -150 3808 y(but)k(alw)m(a)m(ys)h(b)s(eha)m(v)m(es)g(reasonably)-8 -b(,)31 b(no)f(matter)i(ho)m(w)e(bad)g(the)g(input.)150 -3965 y(Lo)m(w)m(er)23 b(v)-5 b(alues)23 b(of)g Fi(workFactor)c -Fj(reduce)k(the)g(amoun)m(t)g(of)f(e\013ort)i(the)f(standard)f -(algorithm)h(will)g(exp)s(end)150 4074 y(b)s(efore)h(resorting)g(to)h -(the)f(fallbac)m(k.)39 b(Y)-8 b(ou)24 b(should)f(set)h(this)g -(parameter)g(carefully;)j(to)s(o)d(lo)m(w,)i(and)d(man)m(y)150 -4184 y(inputs)32 b(will)h(b)s(e)f(handled)g(b)m(y)h(the)g(fallbac)m(k)g -(algorithm)h(and)e(so)h(compress)g(rather)g(slo)m(wly)-8 -b(,)35 b(to)s(o)f(high,)150 4293 y(and)41 b(y)m(our)g(a)m(v)m -(erage-to-w)m(orst)46 b(case)d(compression)e(times)h(can)g(b)s(ecome)g -(v)m(ery)g(large.)74 b(The)42 b(default)150 4403 y(v)-5 -b(alue)31 b(of)f(30)h(giv)m(es)h(reasonable)f(b)s(eha)m(viour)f(o)m(v)m -(er)i(a)f(wide)f(range)h(of)f(circumstances.)150 4560 -y(Allo)m(w)m(able)j(v)-5 b(alues)31 b(range)g(from)g(0)g(to)h(250)g -(inclusiv)m(e.)44 b(0)31 b(is)g(a)g(sp)s(ecial)h(case,)g(equiv)-5 -b(alen)m(t)32 b(to)g(using)f(the)150 4669 y(default)f(v)-5 -b(alue)31 b(of)g(30.)150 4826 y(Note)k(that)e(the)h(compressed)f -(output)g(generated)h(is)f(the)h(same)f(regardless)h(of)f(whether)g(or) -g(not)h(the)150 4936 y(fallbac)m(k)d(algorithm)h(is)e(used.)150 -5093 y(Be)f(a)m(w)m(are)g(also)h(that)f(this)f(parameter)i(ma)m(y)f -(disapp)s(ear)e(en)m(tirely)j(in)e(future)g(v)m(ersions)i(of)e(the)h -(library)-8 b(.)150 5202 y(In)43 b(principle)h(it)g(should)f(b)s(e)g(p) -s(ossible)h(to)g(devise)h(a)f(go)s(o)s(d)f(w)m(a)m(y)i(to)f -(automatically)i(c)m(ho)s(ose)f(whic)m(h)150 5312 y(algorithm)32 -b(to)f(use.)40 b(Suc)m(h)30 b(a)g(mec)m(hanism)h(w)m(ould)f(render)g -(the)h(parameter)g(obsolete.)p eop -%%Page: 15 16 -15 15 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(15)150 299 y Fj(P)m(ossible)32 -b(return)e(v)-5 b(alues:)572 450 y Fi(BZ_CONFIG_ERROR)663 -554 y Fj(if)30 b(the)h(library)f(has)g(b)s(een)g(mis-compiled)572 -657 y Fi(BZ_PARAM_ERROR)663 761 y Fj(if)g Fi(strm)f Fj(is)i -Fi(NULL)663 865 y Fj(or)f Fi(blockSize)e(<)i Fj(1)h(or)g -Fi(blockSize)c(>)k Fj(9)663 969 y(or)f Fi(verbosity)e(<)i -Fj(0)h(or)g Fi(verbosity)c(>)k Fj(4)663 1073 y(or)f Fi(workFactor)e(<)i -Fj(0)h(or)f Fi(workFactor)e(>)i Fj(250)572 1176 y Fi(BZ_MEM_ERROR)663 -1280 y Fj(if)g(not)h(enough)f(memory)h(is)f(a)m(v)-5 -b(ailable)572 1384 y Fi(BZ_OK)663 1488 y Fj(otherwise)150 -1645 y(Allo)m(w)m(able)32 b(next)f(actions:)572 1796 -y Fi(BZ2_bzCompress)663 1899 y Fj(if)f Fi(BZ_OK)f Fj(is)h(returned)572 -2003 y(no)g(sp)s(eci\014c)g(action)i(needed)e(in)g(case)h(of)g(error) -150 2255 y Fe(3.3.2)63 b Fd(BZ2_bzCompress)533 2441 y -Fi(int)47 b(BZ2_bzCompress)d(\()j(bz_stream)f(*strm,)g(int)h(action)f -(\);)150 2598 y Fj(Pro)m(vides)27 b(more)f(input)f(and/or)h(output)g -(bu\013er)f(space)h(for)g(the)h(library)-8 b(.)39 b(The)26 -b(caller)h(main)m(tains)g(input)150 2708 y(and)j(output)g(bu\013ers,)g -(and)f(calls)i Fi(BZ2_bzCompress)c Fj(to)k(transfer)f(data)h(b)s(et)m -(w)m(een)g(them.)150 2865 y(Before)h(eac)m(h)h(call)g(to)f -Fi(BZ2_bzCompress)p Fj(,)d Fi(next_in)h Fj(should)g(p)s(oin)m(t)i(at)h -(the)f(data)g(to)h(b)s(e)e(compressed,)150 2974 y(and)40 -b Fi(avail_in)f Fj(should)h(indicate)i(ho)m(w)f(man)m(y)g(b)m(ytes)h -(the)f(library)g(ma)m(y)g(read.)73 b Fi(BZ2_bzCompress)150 -3084 y Fj(up)s(dates)29 b Fi(next_in)p Fj(,)g Fi(avail_in)f -Fj(and)i Fi(total_in)e Fj(to)j(re\015ect)g(the)g(n)m(um)m(b)s(er)e(of)i -(b)m(ytes)g(it)g(has)f(read.)150 3241 y(Similarly)-8 -b(,)30 b Fi(next_out)d Fj(should)h(p)s(oin)m(t)h(to)h(a)f(bu\013er)f -(in)h(whic)m(h)g(the)g(compressed)g(data)g(is)g(to)h(b)s(e)e(placed,) -150 3350 y(with)k Fi(avail_out)f Fj(indicating)i(ho)m(w)g(m)m(uc)m(h)f -(output)h(space)g(is)f(a)m(v)-5 b(ailable.)49 b Fi(BZ2_bzCompress)29 -b Fj(up)s(dates)150 3460 y Fi(next_out)p Fj(,)f Fi(avail_out)g -Fj(and)i Fi(total_out)e Fj(to)j(re\015ect)g(the)g(n)m(um)m(b)s(er)e(of) -i(b)m(ytes)g(output.)150 3617 y(Y)-8 b(ou)39 b(ma)m(y)g(pro)m(vide)g -(and)f(remo)m(v)m(e)i(as)f(little)i(or)e(as)g(m)m(uc)m(h)f(data)h(as)g -(y)m(ou)g(lik)m(e)h(on)f(eac)m(h)g(call)h(of)f Fi(BZ2_)150 -3726 y(bzCompress)p Fj(.)46 b(In)32 b(the)i(limit,)h(it)e(is)g -(acceptable)i(to)f(supply)e(and)g(remo)m(v)m(e)j(data)e(one)g(b)m(yte)h -(at)g(a)f(time,)150 3836 y(although)27 b(this)g(w)m(ould)f(b)s(e)g -(terribly)i(ine\016cien)m(t.)40 b(Y)-8 b(ou)27 b(should)f(alw)m(a)m(ys) -i(ensure)e(that)h(at)g(least)h(one)f(b)m(yte)150 3946 -y(of)k(output)f(space)g(is)h(a)m(v)-5 b(ailable)32 b(at)f(eac)m(h)g -(call.)150 4102 y(A)36 b(second)g(purp)s(ose)f(of)i Fi(BZ2_bzCompress) -32 b Fj(is)37 b(to)g(request)g(a)f(c)m(hange)h(of)g(mo)s(de)f(of)g(the) -h(compressed)150 4212 y(stream.)150 4369 y(Conceptually)-8 -b(,)24 b(a)d(compressed)g(stream)g(can)g(b)s(e)f(in)h(one)g(of)g(four)f -(states:)38 b(IDLE,)20 b(R)m(UNNING,)i(FLUSH-)150 4478 -y(ING)34 b(and)d(FINISHING.)k(Before)e(initialisation)i(\()p -Fi(BZ2_bzCompressInit)p Fj(\))28 b(and)k(after)i(termination)150 -4588 y(\()p Fi(BZ2_bzCompressEnd)p Fj(\),)27 b(a)j(stream)h(is)f -(regarded)h(as)f(IDLE.)150 4745 y(Up)s(on)j(initialisation)i(\()p -Fi(BZ2_bzCompressInit)p Fj(\),)c(the)j(stream)g(is)g(placed)g(in)f(the) -h(R)m(UNNING)h(state.)150 4854 y(Subsequen)m(t)k(calls)h(to)h -Fi(BZ2_bzCompress)36 b Fj(should)i(pass)i Fi(BZ_RUN)e -Fj(as)h(the)h(requested)g(action;)46 b(other)150 4964 -y(actions)31 b(are)g(illegal)i(and)c(will)i(result)g(in)f -Fi(BZ_SEQUENCE_ERROR)p Fj(.)150 5121 y(A)m(t)37 b(some)f(p)s(oin)m(t,)i -(the)e(calling)h(program)f(will)h(ha)m(v)m(e)g(pro)m(vided)e(all)i(the) -f(input)g(data)g(it)g(w)m(an)m(ts)h(to.)58 b(It)150 5230 -y(will)30 b(then)f(w)m(an)m(t)h(to)g(\014nish)e(up)g({)i(in)f -(e\013ect,)j(asking)d(the)h(library)f(to)h(pro)s(cess)g(an)m(y)f(data)h -(it)g(migh)m(t)g(ha)m(v)m(e)150 5340 y(bu\013ered)23 -b(in)m(ternally)-8 b(.)40 b(In)24 b(this)g(state,)j Fi(BZ2_bzCompress) -20 b Fj(will)k(no)g(longer)h(attempt)g(to)g(read)f(data)g(from)p -eop -%%Page: 16 17 -16 16 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(16)150 299 y Fi(next_in)p -Fj(,)32 b(but)g(it)i(will)f(w)m(an)m(t)h(to)g(write)g(data)f(to)h -Fi(next_out)p Fj(.)46 b(Because)34 b(the)f(output)g(bu\013er)f -(supplied)150 408 y(b)m(y)d(the)g(user)g(can)g(b)s(e)f(arbitrarily)i -(small,)g(the)f(\014nishing-up)f(op)s(eration)h(cannot)h(necessarily)g -(b)s(e)e(done)150 518 y(with)i(a)h(single)g(call)g(of)g -Fi(BZ2_bzCompress)p Fj(.)150 675 y(Instead,)47 b(the)d(calling)g -(program)g(passes)f Fi(BZ_FINISH)e Fj(as)i(an)g(action)i(to)f -Fi(BZ2_bzCompress)p Fj(.)76 b(This)150 784 y(c)m(hanges)26 -b(the)g(stream's)g(state)h(to)f(FINISHING.)h(An)m(y)e(remaining)h -(input)e(\(ie,)k Fi(next_in[0)g(..)i(avail_)150 894 y(in-1])p -Fj(\))k(is)h(compressed)g(and)f(transferred)h(to)h(the)f(output)g -(bu\013er.)54 b(T)-8 b(o)36 b(do)f(this,)i Fi(BZ2_bzCompress)150 -1004 y Fj(m)m(ust)h(b)s(e)f(called)i(rep)s(eatedly)g(un)m(til)f(all)g -(the)h(output)f(has)f(b)s(een)g(consumed.)63 b(A)m(t)38 -b(that)h(p)s(oin)m(t,)h Fi(BZ2_)150 1113 y(bzCompress)g -Fj(returns)i Fi(BZ_STREAM_END)p Fj(,)g(and)f(the)i(stream's)g(state)h -(is)e(set)h(bac)m(k)g(to)h(IDLE.)d Fi(BZ2_)150 1223 y(bzCompressEnd)27 -b Fj(should)i(then)h(b)s(e)g(called.)150 1380 y(Just)g(to)i(mak)m(e)f -(sure)f(the)h(calling)g(program)g(do)s(es)f(not)h(c)m(heat,)h(the)f -(library)f(mak)m(es)h(a)g(note)g(of)g Fi(avail_)150 1489 -y(in)i Fj(at)h(the)g(time)h(of)e(the)h(\014rst)g(call)g(to)h -Fi(BZ2_bzCompress)30 b Fj(whic)m(h)j(has)g Fi(BZ_FINISH)e -Fj(as)j(an)f(action)i(\(ie,)150 1599 y(at)g(the)g(time)h(the)e(program) -h(has)f(announced)g(its)h(in)m(ten)m(tion)h(to)g(not)f(supply)e(an)m(y) -h(more)h(input\).)53 b(By)150 1708 y(comparing)30 b(this)f(v)-5 -b(alue)29 b(with)h(that)f(of)h Fi(avail_in)d Fj(o)m(v)m(er)k(subsequen) -m(t)d(calls)j(to)f Fi(BZ2_bzCompress)p Fj(,)c(the)150 -1818 y(library)k(can)f(detect)j(an)m(y)e(attempts)h(to)f(slip)g(in)f -(more)h(data)g(to)h(compress.)41 b(An)m(y)29 b(calls)i(for)e(whic)m(h)h -(this)150 1928 y(is)h(detected)i(will)f(return)f Fi(BZ_SEQUENCE_ERROR)p -Fj(.)39 b(This)31 b(indicates)i(a)e(programming)g(mistak)m(e)i(whic)m -(h)150 2037 y(should)c(b)s(e)h(corrected.)150 2194 y(Instead)36 -b(of)g(asking)h(to)g(\014nish,)f(the)h(calling)g(program)f(ma)m(y)h -(ask)f Fi(BZ2_bzCompress)c Fj(to)37 b(tak)m(e)h(all)f(the)150 -2304 y(remaining)f(input,)g(compress)f(it)h(and)f(terminate)h(the)g -(curren)m(t)g(\(Burro)m(ws-Wheeler\))h(compression)150 -2413 y(blo)s(c)m(k.)j(This)25 b(could)h(b)s(e)g(useful)f(for)h(error)g -(con)m(trol)i(purp)s(oses.)38 b(The)26 b(mec)m(hanism)g(is)g(analogous) -h(to)f(that)150 2523 y(for)35 b(\014nishing:)48 b(call)35 -b Fi(BZ2_bzCompress)c Fj(with)k(an)f(action)i(of)f Fi(BZ_FLUSH)p -Fj(,)f(remo)m(v)m(e)i(output)f(data,)h(and)150 2632 y(p)s(ersist)i -(with)g(the)h Fi(BZ_FLUSH)d Fj(action)k(un)m(til)e(the)h(v)-5 -b(alue)39 b Fi(BZ_RUN)d Fj(is)j(returned.)64 b(As)38 -b(with)g(\014nishing,)150 2742 y Fi(BZ2_bzCompress)21 -b Fj(detects)27 b(an)m(y)f(attempt)g(to)h(pro)m(vide)e(more)h(input)f -(data)g(once)i(the)e(\015ush)f(has)h(b)s(egun.)150 2899 -y(Once)31 b(the)g(\015ush)d(is)j(complete,)h(the)f(stream)f(returns)g -(to)i(the)e(normal)h(R)m(UNNING)h(state.)150 3056 y(This)f(all)g -(sounds)f(prett)m(y)i(complex,)g(but)f(isn't)g(really)-8 -b(.)44 b(Here's)32 b(a)f(table)h(whic)m(h)e(sho)m(ws)h(whic)m(h)g -(actions)150 3165 y(are)d(allo)m(w)m(able)i(in)e(eac)m(h)h(state,)h -(what)d(action)j(will)e(b)s(e)f(tak)m(en,)j(what)d(the)i(next)f(state)h -(is,)g(and)e(what)h(the)150 3275 y(non-error)g(return)f(v)-5 -b(alues)28 b(are.)40 b(Note)30 b(that)e(y)m(ou)g(can't)g(explicitly)i -(ask)d(what)h(state)h(the)f(stream)g(is)g(in,)150 3384 -y(but)h(nor)h(do)f(y)m(ou)h(need)g(to)g({)h(it)f(can)g(b)s(e)f -(inferred)g(from)h(the)g(v)-5 b(alues)30 b(returned)f(b)m(y)h -Fi(BZ2_bzCompress)p Fj(.)390 3535 y(IDLE/)p Fi(any)572 -3639 y Fj(Illegal.)63 b(IDLE)29 b(state)j(only)e(exists)h(after)g -Fi(BZ2_bzCompressEnd)26 b Fj(or)572 3743 y(b)s(efore)k -Fi(BZ2_bzCompressInit)p Fj(.)572 3847 y(Return)g(v)-5 -b(alue)31 b(=)f Fi(BZ_SEQUENCE_ERROR)390 4054 y Fj(R)m(UNNING/)p -Fi(BZ_RUN)572 4158 y Fj(Compress)g(from)g Fi(next_in)e -Fj(to)j Fi(next_out)d Fj(as)j(m)m(uc)m(h)f(as)h(p)s(ossible.)572 -4262 y(Next)g(state)h(=)e(R)m(UNNING)572 4366 y(Return)g(v)-5 -b(alue)31 b(=)f Fi(BZ_RUN_OK)390 4573 y Fj(R)m(UNNING/)p -Fi(BZ_FLUSH)572 4677 y Fj(Remem)m(b)s(er)g(curren)m(t)h(v)-5 -b(alue)31 b(of)f Fi(next_in)p Fj(.)59 b(Compress)30 b(from)g -Fi(next_in)572 4781 y Fj(to)h Fi(next_out)d Fj(as)j(m)m(uc)m(h)f(as)g -(p)s(ossible,)g(but)g(do)h(not)f(accept)i(an)m(y)e(more)h(input.)572 -4885 y(Next)g(state)h(=)e(FLUSHING)572 4988 y(Return)g(v)-5 -b(alue)31 b(=)f Fi(BZ_FLUSH_OK)390 5196 y Fj(R)m(UNNING/)p -Fi(BZ_FINISH)572 5300 y Fj(Remem)m(b)s(er)g(curren)m(t)h(v)-5 -b(alue)31 b(of)f Fi(next_in)p Fj(.)59 b(Compress)30 b(from)g -Fi(next_in)p eop -%%Page: 17 18 -17 17 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(17)572 299 y Fj(to)31 -b Fi(next_out)d Fj(as)j(m)m(uc)m(h)f(as)g(p)s(ossible,)g(but)g(do)h -(not)f(accept)i(an)m(y)e(more)h(input.)572 403 y(Next)g(state)h(=)e -(FINISHING)572 506 y(Return)g(v)-5 b(alue)31 b(=)f Fi(BZ_FINISH_OK)390 -714 y Fj(FLUSHING/)p Fi(BZ_FLUSH)572 818 y Fj(Compress)g(from)g -Fi(next_in)e Fj(to)j Fi(next_out)d Fj(as)j(m)m(uc)m(h)f(as)h(p)s -(ossible,)572 922 y(but)f(do)g(not)h(accept)g(an)m(y)g(more)g(input.) -572 1025 y(If)f(all)h(the)g(existing)g(input)f(has)g(b)s(een)f(used)h -(up)f(and)h(all)h(compressed)572 1129 y(output)f(has)g(b)s(een)g(remo)m -(v)m(ed)663 1233 y(Next)h(state)h(=)e(R)m(UNNING;)i(Return)e(v)-5 -b(alue)31 b(=)f Fi(BZ_RUN_OK)572 1337 y Fj(else)663 1440 -y(Next)h(state)h(=)e(FLUSHING;)g(Return)h(v)-5 b(alue)30 -b(=)g Fi(BZ_FLUSH_OK)390 1648 y Fj(FLUSHING/other)572 -1752 y(Illegal.)572 1856 y(Return)g(v)-5 b(alue)31 b(=)f -Fi(BZ_SEQUENCE_ERROR)390 2063 y Fj(FINISHING/)p Fi(BZ_FINISH)572 -2167 y Fj(Compress)g(from)g Fi(next_in)e Fj(to)j Fi(next_out)d -Fj(as)j(m)m(uc)m(h)f(as)h(p)s(ossible,)572 2271 y(but)f(to)h(not)g -(accept)g(an)m(y)g(more)g(input.)572 2374 y(If)f(all)h(the)g(existing)g -(input)f(has)g(b)s(een)f(used)h(up)f(and)h(all)h(compressed)572 -2478 y(output)f(has)g(b)s(een)g(remo)m(v)m(ed)663 2582 -y(Next)h(state)h(=)e(IDLE;)f(Return)i(v)-5 b(alue)30 -b(=)g Fi(BZ_STREAM_END)572 2686 y Fj(else)663 2790 y(Next)h(state)h(=)e -(FINISHING;)h(Return)g(v)-5 b(alue)30 b(=)g Fi(BZ_FINISHING)390 -2997 y Fj(FINISHING/other)572 3101 y(Illegal.)572 3205 -y(Return)g(v)-5 b(alue)31 b(=)f Fi(BZ_SEQUENCE_ERROR)150 -3361 y Fj(That)23 b(still)h(lo)s(oks)g(complicated?)39 -b(W)-8 b(ell,)26 b(fair)d(enough.)39 b(The)23 b(usual)f(sequence)h(of)h -(calls)g(for)f(compressing)150 3471 y(a)30 b(load)h(of)g(data)f(is:)225 -3628 y Fh(\017)60 b Fj(Get)32 b(started)f(with)f Fi(BZ2_bzCompressInit) -p Fj(.)225 3774 y Fh(\017)60 b Fj(Sho)m(v)m(el)37 b(data)g(in)f(and)g -(shlurp)f(out)i(its)f(compressed)h(form)f(using)g(zero)i(or)e(more)h -(calls)h(of)e Fi(BZ2_)330 3884 y(bzCompress)28 b Fj(with)i(action)h(=)f -Fi(BZ_RUN)p Fj(.)225 4030 y Fh(\017)60 b Fj(Finish)24 -b(up.)38 b(Rep)s(eatedly)25 b(call)h Fi(BZ2_bzCompress)21 -b Fj(with)k(action)h(=)e Fi(BZ_FINISH)p Fj(,)g(cop)m(ying)h(out)h(the) -330 4139 y(compressed)k(output,)h(un)m(til)g Fi(BZ_STREAM_END)26 -b Fj(is)31 b(returned.)225 4285 y Fh(\017)60 b Fj(Close)31 -b(up)e(and)h(go)h(home.)41 b(Call)30 b Fi(BZ2_bzCompressEnd)p -Fj(.)150 4478 y(If)23 b(the)g(data)g(y)m(ou)h(w)m(an)m(t)g(to)f -(compress)g(\014ts)g(in)m(to)h(y)m(our)g(input)e(bu\013er)g(all)i(at)f -(once,)j(y)m(ou)d(can)g(skip)g(the)g(calls)150 4588 y(of)37 -b Fi(BZ2_bzCompress)26 b(\()k(...,)f(BZ_RUN)g(\))36 b -Fj(and)g(just)g(do)g(the)h Fi(BZ2_bzCompress)26 b(\()k(...,)f -(BZ_FINISH)150 4698 y(\))h Fj(calls.)150 4854 y(All)35 -b(required)g(memory)h(is)f(allo)s(cated)i(b)m(y)e Fi -(BZ2_bzCompressInit)p Fj(.)51 b(The)35 b(compression)h(library)f(can) -150 4964 y(accept)f(an)m(y)e(data)h(at)g(all)h(\(ob)m(viously\).)48 -b(So)32 b(y)m(ou)h(shouldn't)f(get)i(an)m(y)f(error)g(return)f(v)-5 -b(alues)32 b(from)h(the)150 5074 y Fi(BZ2_bzCompress)28 -b Fj(calls.)48 b(If)32 b(y)m(ou)g(do,)h(they)g(will)g(b)s(e)e -Fi(BZ_SEQUENCE_ERROR)p Fj(,)e(and)i(indicate)i(a)g(bug)e(in)150 -5183 y(y)m(our)g(programming.)150 5340 y(T)-8 b(rivial)32 -b(other)f(p)s(ossible)f(return)g(v)-5 b(alues:)p eop -%%Page: 18 19 -18 18 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(18)572 299 y Fi(BZ_PARAM_ERROR)663 -403 y Fj(if)30 b Fi(strm)f Fj(is)i Fi(NULL)p Fj(,)e(or)i -Fi(strm->s)d Fj(is)j Fi(NULL)150 652 y Fe(3.3.3)63 b -Fd(BZ2_bzCompressEnd)390 839 y Fi(int)47 b(BZ2_bzCompressEnd)c(\()k -(bz_stream)f(*strm)g(\);)150 996 y Fj(Releases)32 b(all)f(memory)f -(asso)s(ciated)i(with)e(a)h(compression)f(stream.)150 -1153 y(P)m(ossible)i(return)e(v)-5 b(alues:)481 1304 -y Fi(BZ_PARAM_ERROR)117 b Fj(if)31 b Fi(strm)e Fj(is)h -Fi(NULL)f Fj(or)i Fi(strm->s)d Fj(is)j Fi(NULL)481 1408 -y(BZ_OK)120 b Fj(otherwise)150 1657 y Fe(3.3.4)63 b Fd -(BZ2_bzDecompressInit)390 1844 y Fi(int)47 b(BZ2_bzDecompressInit)42 -b(\()48 b(bz_stream)d(*strm,)h(int)h(verbosity,)e(int)i(small)f(\);)150 -2001 y Fj(Prepares)39 b(for)f(decompression.)65 b(As)38 -b(with)h Fi(BZ2_bzCompressInit)p Fj(,)c(a)k Fi(bz_stream)d -Fj(record)j(should)150 2110 y(b)s(e)30 b(allo)s(cated)j(and)d -(initialised)i(b)s(efore)f(the)g(call.)44 b(Fields)31 -b Fi(bzalloc)p Fj(,)e Fi(bzfree)g Fj(and)h Fi(opaque)g -Fj(should)g(b)s(e)150 2220 y(set)35 b(if)f(a)g(custom)h(memory)f(allo)s -(cator)i(is)e(required,)h(or)g(made)f Fi(NULL)f Fj(for)h(the)g(normal)g -Fi(malloc)p Fj(/)p Fi(free)150 2330 y Fj(routines.)72 -b(Up)s(on)40 b(return,)j(the)e(in)m(ternal)h(state)g(will)g(ha)m(v)m(e) -f(b)s(een)f(initialised,)45 b(and)40 b Fi(total_in)f -Fj(and)150 2439 y Fi(total_out)28 b Fj(will)j(b)s(e)e(zero.)150 -2596 y(F)-8 b(or)31 b(the)g(meaning)f(of)h(parameter)g -Fi(verbosity)p Fj(,)d(see)j Fi(BZ2_bzCompressInit)p Fj(.)150 -2753 y(If)26 b Fi(small)f Fj(is)i(nonzero,)h(the)f(library)f(will)h -(use)f(an)g(alternativ)m(e)j(decompression)e(algorithm)g(whic)m(h)f -(uses)150 2862 y(less)42 b(memory)h(but)e(at)i(the)f(cost)i(of)e -(decompressing)g(more)h(slo)m(wly)g(\(roughly)f(sp)s(eaking,)j(half)d -(the)150 2972 y(sp)s(eed,)29 b(but)f(the)h(maxim)m(um)f(memory)i -(requiremen)m(t)f(drops)f(to)i(around)e(2300k\).)42 b(See)29 -b(Chapter)g(2)g(for)150 3082 y(more)i(information)g(on)f(memory)g -(managemen)m(t.)150 3238 y(Note)37 b(that)f(the)g(amoun)m(t)g(of)g -(memory)g(needed)f(to)h(decompress)g(a)g(stream)g(cannot)g(b)s(e)f -(determined)150 3348 y(un)m(til)41 b(the)f(stream's)h(header)f(has)g(b) -s(een)f(read,)k(so)e(ev)m(en)g(if)f Fi(BZ2_bzDecompressInit)35 -b Fj(succeeds,)43 b(a)150 3458 y(subsequen)m(t)30 b Fi -(BZ2_bzDecompress)c Fj(could)k(fail)h(with)f Fi(BZ_MEM_ERROR)p -Fj(.)150 3614 y(P)m(ossible)i(return)e(v)-5 b(alues:)572 -3765 y Fi(BZ_CONFIG_ERROR)663 3869 y Fj(if)30 b(the)h(library)f(has)g -(b)s(een)g(mis-compiled)572 3973 y Fi(BZ_PARAM_ERROR)663 -4077 y Fj(if)g Fi(\(small)46 b(!=)h(0)h(&&)f(small)f(!=)h(1\))663 -4181 y Fj(or)30 b Fi(\(verbosity)45 b(<)j(0)f(||)g(verbosity)f(>)h(4\)) -572 4284 y(BZ_MEM_ERROR)663 4388 y Fj(if)30 b(insu\016cien)m(t)h -(memory)f(is)h(a)m(v)-5 b(ailable)150 4545 y(Allo)m(w)m(able)32 -b(next)f(actions:)572 4696 y Fi(BZ2_bzDecompress)663 -4800 y Fj(if)f Fi(BZ_OK)f Fj(w)m(as)h(returned)572 4904 -y(no)g(sp)s(eci\014c)g(action)i(required)e(in)g(case)h(of)g(error)150 -5153 y Fe(3.3.5)63 b Fd(BZ2_bzDecompress)390 5340 y Fi(int)47 -b(BZ2_bzDecompress)c(\()48 b(bz_stream)d(*strm)h(\);)p -eop -%%Page: 19 20 -19 19 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(19)150 299 y Fj(Pro)m(vides)41 -b(more)g(input)f(and/out)g(output)h(bu\013er)e(space)i(for)f(the)h -(library)-8 b(.)71 b(The)41 b(caller)h(main)m(tains)150 -408 y(input)30 b(and)f(output)h(bu\013ers,)g(and)g(uses)g -Fi(BZ2_bzDecompress)25 b Fj(to)32 b(transfer)e(data)h(b)s(et)m(w)m(een) -g(them.)150 565 y(Before)45 b(eac)m(h)h(call)f(to)h Fi -(BZ2_bzDecompress)p Fj(,)e Fi(next_in)e Fj(should)i(p)s(oin)m(t)h(at)g -(the)g(compressed)f(data,)150 675 y(and)33 b Fi(avail_in)e -Fj(should)i(indicate)i(ho)m(w)e(man)m(y)h(b)m(ytes)g(the)g(library)g -(ma)m(y)g(read.)50 b Fi(BZ2_bzDecompress)150 784 y Fj(up)s(dates)29 -b Fi(next_in)p Fj(,)g Fi(avail_in)f Fj(and)i Fi(total_in)e -Fj(to)j(re\015ect)g(the)g(n)m(um)m(b)s(er)e(of)i(b)m(ytes)g(it)g(has)f -(read.)150 941 y(Similarly)-8 b(,)40 b Fi(next_out)34 -b Fj(should)i(p)s(oin)m(t)h(to)h(a)f(bu\013er)f(in)h(whic)m(h)g(the)g -(uncompressed)f(output)h(is)g(to)h(b)s(e)150 1051 y(placed,)f(with)e -Fi(avail_out)e Fj(indicating)j(ho)m(w)f(m)m(uc)m(h)h(output)f(space)g -(is)h(a)m(v)-5 b(ailable.)57 b Fi(BZ2_bzCompress)150 -1160 y Fj(up)s(dates)29 b Fi(next_out)p Fj(,)g Fi(avail_out)f -Fj(and)h Fi(total_out)f Fj(to)j(re\015ect)h(the)e(n)m(um)m(b)s(er)g(of) -g(b)m(ytes)h(output.)150 1317 y(Y)-8 b(ou)39 b(ma)m(y)g(pro)m(vide)g -(and)f(remo)m(v)m(e)i(as)f(little)i(or)e(as)g(m)m(uc)m(h)f(data)h(as)g -(y)m(ou)g(lik)m(e)h(on)f(eac)m(h)g(call)h(of)f Fi(BZ2_)150 -1427 y(bzDecompress)p Fj(.)d(In)27 b(the)h(limit,)h(it)f(is)f -(acceptable)i(to)f(supply)e(and)g(remo)m(v)m(e)j(data)f(one)g(b)m(yte)g -(at)f(a)h(time,)150 1537 y(although)f(this)g(w)m(ould)f(b)s(e)g -(terribly)i(ine\016cien)m(t.)40 b(Y)-8 b(ou)27 b(should)f(alw)m(a)m(ys) -i(ensure)e(that)h(at)g(least)h(one)f(b)m(yte)150 1646 -y(of)k(output)f(space)g(is)h(a)m(v)-5 b(ailable)32 b(at)f(eac)m(h)g -(call.)150 1803 y(Use)g(of)f Fi(BZ2_bzDecompress)c Fj(is)k(simpler)h -(than)f Fi(BZ2_bzCompress)p Fj(.)150 1960 y(Y)-8 b(ou)28 -b(should)f(pro)m(vide)i(input)f(and)f(remo)m(v)m(e)j(output)e(as)g -(describ)s(ed)g(ab)s(o)m(v)m(e,)i(and)d(rep)s(eatedly)i(call)g -Fi(BZ2_)150 2069 y(bzDecompress)k Fj(un)m(til)k Fi(BZ_STREAM_END)c -Fj(is)k(returned.)59 b(App)s(earance)36 b(of)g Fi(BZ_STREAM_END)d -Fj(denotes)150 2179 y(that)45 b Fi(BZ2_bzDecompress)40 -b Fj(has)45 b(detected)h(the)f(logical)i(end)d(of)g(the)h(compressed)g -(stream.)84 b Fi(BZ2_)150 2289 y(bzDecompress)27 b Fj(will)k(not)g(pro) -s(duce)e Fi(BZ_STREAM_END)e Fj(un)m(til)k(all)g(output)g(data)f(has)g -(b)s(een)g(placed)h(in)m(to)150 2398 y(the)k(output)f(bu\013er,)i(so)e -(once)i Fi(BZ_STREAM_END)31 b Fj(app)s(ears,)k(y)m(ou)g(are)g(guaran)m -(teed)g(to)h(ha)m(v)m(e)f(a)m(v)-5 b(ailable)150 2508 -y(all)31 b(the)g(decompressed)f(output,)h(and)e Fi(BZ2_bzDecompressEnd) -c Fj(can)31 b(safely)f(b)s(e)g(called.)150 2665 y(If)38 -b(case)i(of)f(an)f(error)h(return)f(v)-5 b(alue,)42 b(y)m(ou)d(should)e -(call)j Fi(BZ2_bzDecompressEnd)33 b Fj(to)40 b(clean)f(up)f(and)150 -2774 y(release)32 b(memory)-8 b(.)150 2931 y(P)m(ossible)32 -b(return)e(v)-5 b(alues:)572 3082 y Fi(BZ_PARAM_ERROR)663 -3186 y Fj(if)30 b Fi(strm)f Fj(is)i Fi(NULL)e Fj(or)i -Fi(strm->s)d Fj(is)i Fi(NULL)663 3290 y Fj(or)g Fi(strm->avail_out)44 -b(<)j(1)572 3393 y(BZ_DATA_ERROR)663 3497 y Fj(if)30 -b(a)h(data)f(in)m(tegrit)m(y)j(error)e(is)f(detected)i(in)e(the)h -(compressed)g(stream)572 3601 y Fi(BZ_DATA_ERROR_MAGIC)663 -3705 y Fj(if)f(the)h(compressed)f(stream)h(do)s(esn't)f(b)s(egin)g -(with)h(the)f(righ)m(t)i(magic)f(b)m(ytes)572 3808 y -Fi(BZ_MEM_ERROR)663 3912 y Fj(if)f(there)h(w)m(asn't)g(enough)f(memory) -h(a)m(v)-5 b(ailable)572 4016 y Fi(BZ_STREAM_END)663 -4120 y Fj(if)30 b(the)h(logical)h(end)e(of)h(the)f(data)h(stream)g(w)m -(as)f(detected)i(and)e(all)663 4224 y(output)g(in)g(has)g(b)s(een)g -(consumed,)g(eg)h Fi(s->avail_out)44 b(>)k(0)572 4327 -y(BZ_OK)663 4431 y Fj(otherwise)150 4588 y(Allo)m(w)m(able)32 -b(next)f(actions:)572 4739 y Fi(BZ2_bzDecompress)663 -4843 y Fj(if)f Fi(BZ_OK)f Fj(w)m(as)h(returned)572 4946 -y Fi(BZ2_bzDecompressEnd)663 5050 y Fj(otherwise)p eop -%%Page: 20 21 -20 20 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(20)150 299 y Fe(3.3.6)63 -b Fd(BZ2_bzDecompressEnd)390 486 y Fi(int)47 b(BZ2_bzDecompressEnd)42 -b(\()48 b(bz_stream)d(*strm)i(\);)150 643 y Fj(Releases)32 -b(all)f(memory)f(asso)s(ciated)i(with)e(a)h(decompression)f(stream.)150 -799 y(P)m(ossible)i(return)e(v)-5 b(alues:)572 950 y -Fi(BZ_PARAM_ERROR)663 1054 y Fj(if)30 b Fi(strm)f Fj(is)i -Fi(NULL)e Fj(or)i Fi(strm->s)d Fj(is)i Fi(NULL)572 1158 -y(BZ_OK)663 1262 y Fj(otherwise)150 1419 y(Allo)m(w)m(able)i(next)f -(actions:)572 1570 y(None.)150 1857 y Fk(3.4)68 b(High-lev)l(el)47 -b(in)l(terface)150 2050 y Fj(This)35 b(in)m(terface)i(pro)m(vides)e -(functions)g(for)g(reading)g(and)f(writing)i Fi(bzip2)d -Fj(format)j(\014les.)55 b(First,)37 b(some)150 2159 y(general)31 -b(p)s(oin)m(ts.)225 2316 y Fh(\017)60 b Fj(All)35 b(of)g(the)h -(functions)e(tak)m(e)j(an)d Fi(int*)g Fj(\014rst)h(argumen)m(t,)i -Fi(bzerror)p Fj(.)52 b(After)36 b(eac)m(h)f(call,)j Fi(bzerror)330 -2426 y Fj(should)23 b(b)s(e)f(consulted)i(\014rst)f(to)i(determine)f -(the)g(outcome)h(of)e(the)h(call.)40 b(If)23 b Fi(bzerror)e -Fj(is)j Fi(BZ_OK)p Fj(,)g(the)330 2535 y(call)35 b(completed)g -(successfully)-8 b(,)36 b(and)d(only)i(then)e(should)h(the)g(return)g -(v)-5 b(alue)34 b(of)g(the)h(function)f(\(if)330 2645 -y(an)m(y\))27 b(b)s(e)e(consulted.)40 b(If)26 b Fi(bzerror)f -Fj(is)h Fi(BZ_IO_ERROR)p Fj(,)e(there)j(w)m(as)g(an)e(error)i -(reading/writing)h(the)330 2754 y(underlying)k(compressed)h(\014le,)h -(and)e(y)m(ou)h(should)f(then)h(consult)g Fi(errno)p -Fj(/)p Fi(perror)e Fj(to)i(determine)330 2864 y(the)k(cause)f(of)h(the) -f(di\016cult)m(y)-8 b(.)60 b Fi(bzerror)34 b Fj(ma)m(y)j(also)g(b)s(e)e -(set)i(to)g(v)-5 b(arious)37 b(other)g(v)-5 b(alues;)39 -b(precise)330 2974 y(details)31 b(are)g(giv)m(en)g(on)g(a)f(p)s -(er-function)g(basis)g(b)s(elo)m(w.)225 3111 y Fh(\017)60 -b Fj(If)39 b Fi(bzerror)e Fj(indicates)i(an)g(error)g(\(ie,)k(an)m -(ything)c(except)h Fi(BZ_OK)d Fj(and)i Fi(BZ_STREAM_END)p -Fj(\),)f(y)m(ou)330 3220 y(should)56 b(immediately)i(call)g -Fi(BZ2_bzReadClose)53 b Fj(\(or)58 b Fi(BZ2_bzWriteClose)p -Fj(,)h(dep)s(ending)d(on)330 3330 y(whether)37 b(y)m(ou)h(are)f -(attempting)i(to)f(read)f(or)h(to)g(write\))g(to)g(free)g(up)e(all)i -(resources)g(asso)s(ciated)330 3439 y(with)44 b(the)g(stream.)83 -b(Once)44 b(an)g(error)h(has)e(b)s(een)h(indicated,)k(b)s(eha)m(viour)c -(of)g(all)h(calls)g(except)330 3549 y Fi(BZ2_bzReadClose)h -Fj(\()p Fi(BZ2_bzWriteClose)p Fj(\))h(is)j(unde\014ned.)99 -b(The)51 b(implication)g(is)g(that)g(\(1\))330 3659 y -Fi(bzerror)42 b Fj(should)i(b)s(e)f(c)m(hec)m(k)m(ed)k(after)e(eac)m(h) -g(call,)k(and)43 b(\(2\))j(if)e Fi(bzerror)e Fj(indicates)j(an)f -(error,)330 3768 y Fi(BZ2_bzReadClose)26 b Fj(\()p Fi(BZ2_bzWriteClose) -p Fj(\))h(should)i(then)h(b)s(e)g(called)i(to)f(clean)g(up.)225 -3905 y Fh(\017)60 b Fj(The)32 b Fi(FILE*)f Fj(argumen)m(ts)i(passed)f -(to)h Fi(BZ2_bzReadOpen)p Fj(/)p Fi(BZ2_bzWriteO)o(pen)26 -b Fj(should)31 b(b)s(e)h(set)h(to)330 4015 y(binary)22 -b(mo)s(de.)38 b(Most)24 b(Unix)f(systems)g(will)g(do)g(this)g(b)m(y)g -(default,)i(but)d(other)i(platforms,)g(including)330 -4124 y(Windo)m(ws)41 b(and)g(Mac,)46 b(will)c(not.)76 -b(If)41 b(y)m(ou)i(omit)f(this,)j(y)m(ou)e(ma)m(y)f(encoun)m(ter)h -(problems)e(when)330 4234 y(mo)m(ving)31 b(co)s(de)g(to)g(new)f -(platforms.)225 4371 y Fh(\017)60 b Fj(Memory)45 b(allo)s(cation)h -(requests)f(are)g(handled)e(b)m(y)h Fi(malloc)p Fj(/)p -Fi(free)p Fj(.)80 b(A)m(t)45 b(presen)m(t)g(there)g(is)f(no)330 -4481 y(facilit)m(y)39 b(for)f(user-de\014ned)e(memory)h(allo)s(cators)j -(in)d(the)g(\014le)h(I/O)g(functions)f(\(could)g(easily)i(b)s(e)330 -4590 y(added,)30 b(though\).)150 4842 y Fe(3.4.1)63 b -Fd(BZ2_bzReadOpen)533 5029 y Fi(typedef)46 b(void)h(BZFILE;)533 -5236 y(BZFILE)f(*BZ2_bzReadOpen)e(\()j(int)g(*bzerror,)f(FILE)g(*f,) -1726 5340 y(int)h(small,)f(int)h(verbosity,)p eop -%%Page: 21 22 -21 21 bop 150 -116 a Fl(Chapter)30 b(3:)h(Programming)e(with)g -Fi(libbzip2)1891 b Fl(21)1726 299 y Fi(void)47 b(*unused,)f(int)g -(nUnused)g(\);)150 456 y Fj(Prepare)27 b(to)g(read)f(compressed)h(data) -g(from)f(\014le)g(handle)g Fi(f)p Fj(.)39 b Fi(f)26 b -Fj(should)g(refer)g(to)i(a)e(\014le)h(whic)m(h)f(has)g(b)s(een)150 -565 y(op)s(ened)i(for)g(reading,)i(and)d(for)i(whic)m(h)f(the)h(error)g -(indicator)h(\()p Fi(ferror\(f\))p Fj(\)is)c(not)j(set.)41 -b(If)29 b Fi(small)e Fj(is)h(1,)150 675 y(the)j(library)f(will)h(try)f -(to)i(decompress)e(using)g(less)h(memory)-8 b(,)31 b(at)g(the)f(exp)s -(ense)g(of)h(sp)s(eed.)150 832 y(F)-8 b(or)37 b(reasons)f(explained)h -(b)s(elo)m(w,)h Fi(BZ2_bzRead)33 b Fj(will)k(decompress)f(the)h -Fi(nUnused)d Fj(b)m(ytes)j(starting)g(at)150 941 y Fi(unused)p -Fj(,)j(b)s(efore)f(starting)h(to)g(read)f(from)g(the)h(\014le)f -Fi(f)p Fj(.)67 b(A)m(t)40 b(most)g Fi(BZ_MAX_UNUSED)35 -b Fj(b)m(ytes)40 b(ma)m(y)g(b)s(e)150 1051 y(supplied)34 -b(lik)m(e)i(this.)54 b(If)34 b(this)h(facilit)m(y)i(is)e(not)g -(required,)h(y)m(ou)f(should)f(pass)g Fi(NULL)f Fj(and)h -Fi(0)h Fj(for)g Fi(unused)150 1160 y Fj(and)30 b(n)p -Fi(Unused)e Fj(resp)s(ectiv)m(ely)-8 b(.)150 1317 y(F)g(or)31 -b(the)g(meaning)f(of)h(parameters)g Fi(small)e Fj(and)g -Fi(verbosity)p Fj(,)f(see)j Fi(BZ2_bzDecompressInit)p -Fj(.)150 1474 y(The)i(amoun)m(t)g(of)f(memory)h(needed)g(to)g -(decompress)g(a)g(\014le)g(cannot)g(b)s(e)f(determined)h(un)m(til)g -(the)g(\014le's)150 1584 y(header)h(has)f(b)s(een)g(read.)51 -b(So)33 b(it)i(is)e(p)s(ossible)h(that)g Fi(BZ2_bzReadOpen)c -Fj(returns)j Fi(BZ_OK)g Fj(but)g(a)h(subse-)150 1693 -y(quen)m(t)d(call)g(of)g Fi(BZ2_bzRead)c Fj(will)k(return)f -Fi(BZ_MEM_ERROR)p Fj(.)150 1850 y(P)m(ossible)i(assignmen)m(ts)e(to)i -Fi(bzerror)p Fj(:)572 2001 y Fi(BZ_CONFIG_ERROR)663 2105 -y Fj(if)e(the)h(library)f(has)g(b)s(een)g(mis-compiled)572 -2209 y Fi(BZ_PARAM_ERROR)663 2313 y Fj(if)g Fi(f)g Fj(is)h -Fi(NULL)663 2416 y Fj(or)f Fi(small)f Fj(is)i(neither)g -Fi(0)f Fj(nor)g Fi(1)663 2520 y Fj(or)g Fi(\(unused)46 -b(==)h(NULL)g(&&)g(nUnused)f(!=)h(0\))663 2624 y Fj(or)30 -b Fi(\(unused)46 b(!=)h(NULL)g(&&)g(!\(0)g(<=)g(nUnused)f(<=)h -(BZ_MAX_UNUSED\)\))572 2728 y(BZ_IO_ERROR)663 2831 y -Fj(if)30 b Fi(ferror\(f\))e Fj(is)i(nonzero)572 2935 -y Fi(BZ_MEM_ERROR)663 3039 y Fj(if)g(insu\016cien)m(t)h(memory)f(is)h -(a)m(v)-5 b(ailable)572 3143 y Fi(BZ_OK)663 3247 y Fj(otherwise.)150 -3403 y(P)m(ossible)32 b(return)e(v)-5 b(alues:)572 3554 -y(P)m(oin)m(ter)32 b(to)f(an)f(abstract)i Fi(BZFILE)663 -3658 y Fj(if)e Fi(bzerror)e Fj(is)j Fi(BZ_OK)572 3762 -y(NULL)663 3866 y Fj(otherwise)150 4023 y(Allo)m(w)m(able)h(next)f -(actions:)572 4174 y Fi(BZ2_bzRead)663 4277 y Fj(if)f -Fi(bzerror)e Fj(is)j Fi(BZ_OK)572 4381 y(BZ2_bzClose)663 -4485 y Fj(otherwise)150 4887 y Fe(3.4.2)63 b Fd(BZ2_bzRead)533 -5074 y Fi(int)47 b(BZ2_bzRead)e(\()j(int)e(*bzerror,)g(BZFILE)g(*b,)h -(void)f(*buf,)h(int)g(len)g(\);)150 5230 y Fj(Reads)33 -b(up)e(to)j Fi(len)d Fj(\(uncompressed\))i(b)m(ytes)g(from)f(the)h -(compressed)g(\014le)g Fi(b)f Fj(in)m(to)i(the)f(bu\013er)f -Fi(buf)p Fj(.)46 b(If)150 5340 y(the)28 b(read)f(w)m(as)h(successful,)g -Fi(bzerror)d Fj(is)j(set)g(to)g Fi(BZ_OK)f Fj(and)f(the)i(n)m(um)m(b)s -(er)e(of)i(b)m(ytes)g(read)g(is)f(returned.)p eop -%%Page: 22 23 -22 22 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(22)150 299 y Fj(If)32 -b(the)g(logical)i(end-of-stream)e(w)m(as)g(detected,)i -Fi(bzerror)c Fj(will)i(b)s(e)g(set)g(to)h Fi(BZ_STREAM_END)p -Fj(,)c(and)i(the)150 408 y(n)m(um)m(b)s(er)e(of)i(b)m(ytes)g(read)f(is) -h(returned.)40 b(All)30 b(other)h Fi(bzerror)e Fj(v)-5 -b(alues)30 b(denote)h(an)f(error.)150 565 y Fi(BZ2_bzRead)k -Fj(will)k(supply)e Fi(len)g Fj(b)m(ytes,)k(unless)d(the)g(logical)j -(stream)d(end)g(is)g(detected)i(or)f(an)e(error)150 675 -y(o)s(ccurs.)70 b(Because)40 b(of)g(this,)j(it)e(is)f(p)s(ossible)f(to) -i(detect)h(the)e(stream)h(end)e(b)m(y)h(observing)h(when)e(the)150 -784 y(n)m(um)m(b)s(er)24 b(of)h(b)m(ytes)h(returned)e(is)h(less)h(than) -e(the)i(n)m(um)m(b)s(er)e(requested.)39 b(Nev)m(ertheless,)29 -b(this)c(is)g(regarded)150 894 y(as)37 b(inadvisable;)i(y)m(ou)e -(should)f(instead)h(c)m(hec)m(k)h Fi(bzerror)c Fj(after)k(ev)m(ery)f -(call)h(and)e(w)m(atc)m(h)h(out)g(for)g Fi(BZ_)150 1004 -y(STREAM_END)p Fj(.)150 1160 y(In)m(ternally)-8 b(,)47 -b Fi(BZ2_bzRead)39 b Fj(copies)k(data)f(from)g(the)g(compressed)g -(\014le)g(in)g(c)m(h)m(unks)g(of)g(size)h Fi(BZ_MAX_)150 -1270 y(UNUSED)28 b Fj(b)m(ytes)j(b)s(efore)e(decompressing)h(it.)41 -b(If)30 b(the)g(\014le)g(con)m(tains)h(more)f(b)m(ytes)h(than)e -(strictly)i(needed)150 1380 y(to)46 b(reac)m(h)g(the)f(logical)i -(end-of-stream,)j Fi(BZ2_bzRead)42 b Fj(will)j(almost)h(certainly)g -(read)f(some)g(of)h(the)150 1489 y(trailing)f(data)g(b)s(efore)f -(signalling)h Fi(BZ_SEQUENCE_END)p Fj(.)78 b(T)-8 b(o)45 -b(collect)i(the)d(read)g(but)g(un)m(used)f(data)150 1599 -y(once)27 b Fi(BZ_SEQUENCE_END)22 b Fj(has)k(app)s(eared,)g(call)h -Fi(BZ2_bzReadGetUnused)21 b Fj(immediately)28 b(b)s(efore)e -Fi(BZ2_)150 1708 y(bzReadClose)p Fj(.)150 1865 y(P)m(ossible)32 -b(assignmen)m(ts)e(to)i Fi(bzerror)p Fj(:)572 2016 y -Fi(BZ_PARAM_ERROR)663 2120 y Fj(if)e Fi(b)g Fj(is)h Fi(NULL)e -Fj(or)h Fi(buf)g Fj(is)g Fi(NULL)f Fj(or)i Fi(len)47 -b(<)g(0)572 2224 y(BZ_SEQUENCE_ERROR)663 2328 y Fj(if)30 -b Fi(b)g Fj(w)m(as)h(op)s(ened)e(with)h Fi(BZ2_bzWriteOpen)572 -2431 y(BZ_IO_ERROR)663 2535 y Fj(if)g(there)h(is)f(an)g(error)h -(reading)g(from)f(the)h(compressed)f(\014le)572 2639 -y Fi(BZ_UNEXPECTED_EOF)663 2743 y Fj(if)g(the)h(compressed)f(\014le)h -(ended)e(b)s(efore)i(the)f(logical)j(end-of-stream)e(w)m(as)g(detected) -572 2847 y Fi(BZ_DATA_ERROR)663 2950 y Fj(if)f(a)h(data)f(in)m(tegrit)m -(y)j(error)e(w)m(as)g(detected)g(in)g(the)f(compressed)h(stream)572 -3054 y Fi(BZ_DATA_ERROR_MAGIC)663 3158 y Fj(if)f(the)h(stream)g(do)s -(es)f(not)g(b)s(egin)g(with)h(the)f(requisite)i(header)e(b)m(ytes)h -(\(ie,)h(is)e(not)663 3262 y(a)g Fi(bzip2)f Fj(data)i(\014le\).)61 -b(This)30 b(is)h(really)g(a)f(sp)s(ecial)h(case)g(of)g -Fi(BZ_DATA_ERROR)p Fj(.)572 3365 y Fi(BZ_MEM_ERROR)663 -3469 y Fj(if)f(insu\016cien)m(t)h(memory)f(w)m(as)h(a)m(v)-5 -b(ailable)572 3573 y Fi(BZ_STREAM_END)663 3677 y Fj(if)30 -b(the)h(logical)h(end)e(of)h(stream)g(w)m(as)f(detected.)572 -3781 y Fi(BZ_OK)663 3884 y Fj(otherwise.)150 4041 y(P)m(ossible)i -(return)e(v)-5 b(alues:)572 4192 y(n)m(um)m(b)s(er)29 -b(of)i(b)m(ytes)g(read)663 4296 y(if)f Fi(bzerror)e Fj(is)j -Fi(BZ_OK)e Fj(or)h Fi(BZ_STREAM_END)572 4400 y Fj(unde\014ned)663 -4503 y(otherwise)150 4660 y(Allo)m(w)m(able)i(next)f(actions:)572 -4811 y(collect)i(data)d(from)h Fi(buf)p Fj(,)e(then)i -Fi(BZ2_bzRead)c Fj(or)k Fi(BZ2_bzReadClose)663 4915 y -Fj(if)f Fi(bzerror)e Fj(is)j Fi(BZ_OK)572 5019 y Fj(collect)i(data)d -(from)h Fi(buf)p Fj(,)e(then)i Fi(BZ2_bzReadClose)26 -b Fj(or)k Fi(BZ2_bzReadGetUnused)663 5123 y Fj(if)g Fi(bzerror)e -Fj(is)j Fi(BZ_SEQUENCE_END)572 5226 y(BZ2_bzReadClose)663 -5330 y Fj(otherwise)p eop -%%Page: 23 24 -23 23 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(23)150 299 y Fe(3.4.3)63 -b Fd(BZ2_bzReadGetUnused)533 486 y Fi(void)47 b(BZ2_bzReadGetUnused)42 -b(\()48 b(int*)e(bzerror,)g(BZFILE)g(*b,)1822 589 y(void**)g(unused,)g -(int*)g(nUnused)g(\);)150 746 y Fj(Returns)35 b(data)g(whic)m(h)f(w)m -(as)i(read)e(from)h(the)g(compressed)g(\014le)g(but)g(w)m(as)g(not)g -(needed)g(to)h(get)g(to)g(the)150 856 y(logical)44 b(end-of-stream.)76 -b Fi(*unused)39 b Fj(is)j(set)h(to)f(the)g(address)f(of)h(the)g(data,)k -(and)40 b Fi(*nUnused)g Fj(to)j(the)150 965 y(n)m(um)m(b)s(er)28 -b(of)h(b)m(ytes.)41 b Fi(*nUnused)27 b Fj(will)i(b)s(e)g(set)g(to)h(a)f -(v)-5 b(alue)30 b(b)s(et)m(w)m(een)g Fi(0)e Fj(and)g -Fi(BZ_MAX_UNUSED)e Fj(inclusiv)m(e.)150 1122 y(This)f(function)g(ma)m -(y)h(only)f(b)s(e)g(called)h(once)g Fi(BZ2_bzRead)d Fj(has)i(signalled) -h Fi(BZ_STREAM_END)21 b Fj(but)k(b)s(efore)150 1232 y -Fi(BZ2_bzReadClose)p Fj(.)150 1389 y(P)m(ossible)32 b(assignmen)m(ts)e -(to)i Fi(bzerror)p Fj(:)572 1540 y Fi(BZ_PARAM_ERROR)663 -1644 y Fj(if)e Fi(b)g Fj(is)h Fi(NULL)663 1747 y Fj(or)f -Fi(unused)f Fj(is)h Fi(NULL)g Fj(or)g Fi(nUnused)f Fj(is)h -Fi(NULL)572 1851 y(BZ_SEQUENCE_ERROR)663 1955 y Fj(if)g -Fi(BZ_STREAM_END)d Fj(has)j(not)g(b)s(een)g(signalled)663 -2059 y(or)g(if)h Fi(b)f Fj(w)m(as)g(op)s(ened)g(with)g -Fi(BZ2_bzWriteOpen)542 2162 y(BZ_OK)663 2266 y Fj(otherwise)150 -2423 y(Allo)m(w)m(able)i(next)f(actions:)572 2574 y Fi(BZ2_bzReadClose) -150 2882 y Fe(3.4.4)63 b Fd(BZ2_bzReadClose)533 3068 -y Fi(void)47 b(BZ2_bzReadClose)c(\()48 b(int)f(*bzerror,)e(BZFILE)h(*b) -h(\);)150 3225 y Fj(Releases)35 b(all)e(memory)h(p)s(ertaining)f(to)h -(the)f(compressed)h(\014le)f Fi(b)p Fj(.)49 b Fi(BZ2_bzReadClose)29 -b Fj(do)s(es)j(not)i(call)150 3335 y Fi(fclose)d Fj(on)h(the)g -(underlying)g(\014le)g(handle,)h(so)f(y)m(ou)h(should)e(do)h(that)h(y)m -(ourself)g(if)f(appropriate.)47 b Fi(BZ2_)150 3445 y(bzReadClose)27 -b Fj(should)j(b)s(e)f(called)j(to)f(clean)g(up)e(after)i(all)g(error)g -(situations.)150 3601 y(P)m(ossible)h(assignmen)m(ts)e(to)i -Fi(bzerror)p Fj(:)572 3752 y Fi(BZ_SEQUENCE_ERROR)663 -3856 y Fj(if)e Fi(b)g Fj(w)m(as)h(op)s(ened)e(with)h -Fi(BZ2_bzOpenWrite)572 3960 y(BZ_OK)663 4064 y Fj(otherwise)150 -4221 y(Allo)m(w)m(able)i(next)f(actions:)572 4372 y(none)150 -4679 y Fe(3.4.5)63 b Fd(BZ2_bzWriteOpen)533 4866 y Fi(BZFILE)46 -b(*BZ2_bzWriteOpen)e(\()j(int)g(*bzerror,)e(FILE)i(*f,)1774 -4970 y(int)g(blockSize100k,)d(int)j(verbosity,)1774 5074 -y(int)g(workFactor)e(\);)150 5230 y Fj(Prepare)31 b(to)h(write)g -(compressed)f(data)g(to)h(\014le)f(handle)f Fi(f)p Fj(.)42 -b Fi(f)30 b Fj(should)g(refer)i(to)f(a)h(\014le)f(whic)m(h)f(has)h(b)s -(een)150 5340 y(op)s(ened)f(for)g(writing,)h(and)f(for)g(whic)m(h)g -(the)h(error)g(indicator)g(\()p Fi(ferror\(f\))p Fj(\)is)e(not)h(set.)p -eop -%%Page: 24 25 -24 24 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(24)150 299 y Fj(F)-8 b(or)31 -b(the)g(meaning)f(of)h(parameters)g Fi(blockSize100k)p -Fj(,)c Fi(verbosity)h Fj(and)h Fi(workFactor)p Fj(,)f(see)150 -408 y Fi(BZ2_bzCompressInit)p Fj(.)150 565 y(All)38 b(required)g -(memory)g(is)g(allo)s(cated)i(at)e(this)h(stage,)i(so)d(if)g(the)h -(call)g(completes)g(successfully)-8 b(,)41 b Fi(BZ_)150 -675 y(MEM_ERROR)28 b Fj(cannot)j(b)s(e)e(signalled)i(b)m(y)g(a)f -(subsequen)m(t)g(call)h(to)h Fi(BZ2_bzWrite)p Fj(.)150 -832 y(P)m(ossible)g(assignmen)m(ts)e(to)i Fi(bzerror)p -Fj(:)572 983 y Fi(BZ_CONFIG_ERROR)663 1087 y Fj(if)e(the)h(library)f -(has)g(b)s(een)g(mis-compiled)572 1190 y Fi(BZ_PARAM_ERROR)663 -1294 y Fj(if)g Fi(f)g Fj(is)h Fi(NULL)663 1398 y Fj(or)f -Fi(blockSize100k)45 b(<)i(1)30 b Fj(or)h Fi(blockSize100k)44 -b(>)j(9)572 1502 y(BZ_IO_ERROR)663 1605 y Fj(if)30 b -Fi(ferror\(f\))e Fj(is)i(nonzero)572 1709 y Fi(BZ_MEM_ERROR)663 -1813 y Fj(if)g(insu\016cien)m(t)h(memory)f(is)h(a)m(v)-5 -b(ailable)572 1917 y Fi(BZ_OK)663 2021 y Fj(otherwise)150 -2177 y(P)m(ossible)32 b(return)e(v)-5 b(alues:)572 2328 -y(P)m(oin)m(ter)32 b(to)f(an)f(abstract)i Fi(BZFILE)663 -2432 y Fj(if)e Fi(bzerror)e Fj(is)j Fi(BZ_OK)572 2536 -y(NULL)663 2640 y Fj(otherwise)150 2797 y(Allo)m(w)m(able)h(next)f -(actions:)572 2948 y Fi(BZ2_bzWrite)663 3051 y Fj(if)f -Fi(bzerror)e Fj(is)j Fi(BZ_OK)597 3155 y Fj(\(y)m(ou)24 -b(could)f(go)h(directly)g(to)g Fi(BZ2_bzWriteClose)p -Fj(,)19 b(but)j(this)h(w)m(ould)g(b)s(e)g(prett)m(y)h(p)s(oin)m -(tless\))572 3259 y Fi(BZ2_bzWriteClose)663 3363 y Fj(otherwise)150 -3852 y Fe(3.4.6)63 b Fd(BZ2_bzWrite)533 4039 y Fi(void)47 -b(BZ2_bzWrite)e(\()i(int)g(*bzerror,)e(BZFILE)h(*b,)h(void)g(*buf,)f -(int)h(len)g(\);)150 4196 y Fj(Absorbs)35 b Fi(len)h -Fj(b)m(ytes)h(from)f(the)h(bu\013er)f Fi(buf)p Fj(,)h(ev)m(en)m(tually) -i(to)e(b)s(e)f(compressed)g(and)g(written)h(to)h(the)150 -4306 y(\014le.)150 4463 y(P)m(ossible)32 b(assignmen)m(ts)e(to)i -Fi(bzerror)p Fj(:)572 4614 y Fi(BZ_PARAM_ERROR)663 4717 -y Fj(if)e Fi(b)g Fj(is)h Fi(NULL)e Fj(or)h Fi(buf)g Fj(is)g -Fi(NULL)f Fj(or)i Fi(len)47 b(<)g(0)572 4821 y(BZ_SEQUENCE_ERROR)663 -4925 y Fj(if)30 b(b)g(w)m(as)g(op)s(ened)g(with)g Fi(BZ2_bzReadOpen)572 -5029 y(BZ_IO_ERROR)663 5132 y Fj(if)g(there)h(is)f(an)g(error)h -(writing)g(the)g(compressed)f(\014le.)572 5236 y Fi(BZ_OK)663 -5340 y Fj(otherwise)p eop -%%Page: 25 26 -25 25 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(25)150 299 y Fe(3.4.7)63 -b Fd(BZ2_bzWriteClose)533 486 y Fi(void)47 b(BZ2_bzWriteClose)c(\()48 -b(int)f(*bzerror,)e(BZFILE*)h(f,)1679 589 y(int)h(abandon,)1679 -693 y(unsigned)e(int*)i(nbytes_in,)1679 797 y(unsigned)e(int*)i -(nbytes_out)e(\);)533 1005 y(void)i(BZ2_bzWriteClose64)c(\()k(int)g -(*bzerror,)e(BZFILE*)h(f,)1774 1108 y(int)h(abandon,)1774 -1212 y(unsigned)f(int*)g(nbytes_in_lo32,)1774 1316 y(unsigned)g(int*)g -(nbytes_in_hi32,)1774 1420 y(unsigned)g(int*)g(nbytes_out_lo32,)1774 -1523 y(unsigned)g(int*)g(nbytes_out_hi32)e(\);)150 1680 -y Fj(Compresses)38 b(and)f(\015ushes)g(to)i(the)g(compressed)f(\014le)h -(all)f(data)h(so)g(far)f(supplied)f(b)m(y)h Fi(BZ2_bzWrite)p -Fj(.)150 1790 y(The)h(logical)i(end-of-stream)e(mark)m(ers)h(are)f -(also)g(written,)j(so)d(subsequen)m(t)g(calls)g(to)h -Fi(BZ2_bzWrite)150 1899 y Fj(are)31 b(illegal.)45 b(All)31 -b(memory)g(asso)s(ciated)h(with)f(the)g(compressed)g(\014le)h -Fi(b)e Fj(is)h(released.)44 b Fi(fflush)29 b Fj(is)i(called)150 -2009 y(on)f(the)h(compressed)f(\014le,)h(but)f(it)h(is)f(not)h -Fi(fclose)p Fj('d.)150 2166 y(If)g Fi(BZ2_bzWriteClose)d -Fj(is)j(called)i(to)f(clean)g(up)f(after)h(an)f(error,)i(the)f(only)g -(action)h(is)e(to)i(release)g(the)150 2275 y(memory)-8 -b(.)40 b(The)28 b(library)g(records)g(the)g(error)g(co)s(des)g(issued)f -(b)m(y)g(previous)h(calls,)h(so)f(this)g(situation)h(will)150 -2385 y(b)s(e)37 b(detected)i(automatically)-8 b(.)64 -b(There)38 b(is)f(no)h(attempt)g(to)h(complete)g(the)e(compression)h -(op)s(eration,)150 2495 y(nor)f(to)h Fi(fflush)e Fj(the)h(compressed)g -(\014le.)62 b(Y)-8 b(ou)37 b(can)g(force)h(this)g(b)s(eha)m(viour)e(to) -i(happ)s(en)e(ev)m(en)i(in)f(the)150 2604 y(case)31 b(of)g(no)f(error,) -h(b)m(y)f(passing)g(a)h(nonzero)g(v)-5 b(alue)30 b(to)i -Fi(abandon)p Fj(.)150 2761 y(If)h Fi(nbytes_in)d Fj(is)j(non-n)m(ull,)g -Fi(*nbytes_in)d Fj(will)j(b)s(e)f(set)i(to)f(b)s(e)g(the)g(total)h(v)m -(olume)g(of)f(uncompressed)150 2871 y(data)k(handled.)58 -b(Similarly)-8 b(,)38 b Fi(nbytes_out)c Fj(will)j(b)s(e)f(set)h(to)g -(the)g(total)h(v)m(olume)g(of)e(compressed)h(data)150 -2980 y(written.)i(F)-8 b(or)24 b(compatibilit)m(y)i(with)d(older)h(v)m -(ersions)g(of)g(the)g(library)-8 b(,)25 b Fi(BZ2_bzWriteClose)19 -b Fj(only)24 b(yields)150 3090 y(the)39 b(lo)m(w)m(er)i(32)e(bits)g(of) -g(these)g(coun)m(ts.)67 b(Use)39 b Fi(BZ2_bzWriteClose64)34 -b Fj(if)39 b(y)m(ou)g(w)m(an)m(t)g(the)h(full)e(64)i(bit)150 -3199 y(coun)m(ts.)h(These)31 b(t)m(w)m(o)h(functions)e(are)h(otherwise) -g(absolutely)g(iden)m(tical.)150 3356 y(P)m(ossible)h(assignmen)m(ts)e -(to)i Fi(bzerror)p Fj(:)572 3507 y Fi(BZ_SEQUENCE_ERROR)663 -3611 y Fj(if)e Fi(b)g Fj(w)m(as)h(op)s(ened)e(with)h -Fi(BZ2_bzReadOpen)572 3715 y(BZ_IO_ERROR)663 3819 y Fj(if)g(there)h(is) -f(an)g(error)h(writing)g(the)g(compressed)f(\014le)572 -3922 y Fi(BZ_OK)663 4026 y Fj(otherwise)150 4296 y Fe(3.4.8)63 -b(Handling)41 b(em)m(b)s(edded)g(compressed)h(data)e(streams)150 -4489 y Fj(The)i(high-lev)m(el)h(library)e(facilitates)k(use)c(of)h -Fi(bzip2)e Fj(data)i(streams)g(whic)m(h)f(form)g(some)h(part)g(of)g(a) -150 4598 y(surrounding,)29 b(larger)i(data)g(stream.)225 -4755 y Fh(\017)60 b Fj(F)-8 b(or)49 b(writing,)k(the)c(library)f(tak)m -(es)i(an)e(op)s(en)f(\014le)i(handle,)j(writes)d(compressed)f(data)h -(to)g(it,)330 4865 y Fi(fflush)p Fj(es)37 b(it)h(but)f(do)s(es)h(not)g -Fi(fclose)f Fj(it.)64 b(The)38 b(calling)h(application)g(can)f(write)g -(its)h(o)m(wn)f(data)330 4974 y(b)s(efore)30 b(and)g(after)h(the)g -(compressed)f(data)h(stream,)g(using)f(that)h(same)f(\014le)h(handle.) -225 5121 y Fh(\017)60 b Fj(Reading)33 b(is)g(more)g(complex,)h(and)e -(the)h(facilities)i(are)e(not)g(as)g(general)h(as)f(they)g(could)f(b)s -(e)h(since)330 5230 y(generalit)m(y)e(is)d(hard)g(to)h(reconcile)i -(with)d(e\016ciency)-8 b(.)42 b Fi(BZ2_bzRead)25 b Fj(reads)k(from)f -(the)h(compressed)330 5340 y(\014le)39 b(in)g(blo)s(c)m(ks)g(of)h(size) -f Fi(BZ_MAX_UNUSED)d Fj(b)m(ytes,)42 b(and)c(in)h(doing)g(so)g -(probably)g(will)g(o)m(v)m(ersho)s(ot)p eop -%%Page: 26 27 -26 26 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(26)330 299 y Fj(the)37 -b(logical)i(end)e(of)g(compressed)g(stream.)60 b(T)-8 -b(o)38 b(reco)m(v)m(er)i(this)c(data)i(once)f(decompression)g(has)330 -408 y(ended,)27 b(call)h Fi(BZ2_bzReadGetUnused)21 b -Fj(after)27 b(the)g(last)h(call)f(of)g Fi(BZ2_bzRead)d -Fj(\(the)j(one)g(returning)330 518 y Fi(BZ_STREAM_END)p -Fj(\))g(but)j(b)s(efore)g(calling)i Fi(BZ2_bzReadClose)p -Fj(.)150 702 y(This)j(mec)m(hanism)h(mak)m(es)g(it)h(easy)f(to)g -(decompress)g(m)m(ultiple)g Fi(bzip2)e Fj(streams)i(placed)g -(end-to-end.)150 812 y(As)67 b(the)h(end)g(of)f(one)i(stream,)77 -b(when)67 b Fi(BZ2_bzRead)e Fj(returns)j Fi(BZ_STREAM_END)p -Fj(,)73 b(call)c Fi(BZ2_)150 921 y(bzReadGetUnused)34 -b Fj(to)40 b(collect)h(the)e(un)m(used)e(data)i(\(cop)m(y)g(it)g(in)m -(to)h(y)m(our)f(o)m(wn)f(bu\013er)g(somewhere\).)150 -1031 y(That)52 b(data)h(forms)f(the)g(start)h(of)f(the)h(next)f -(compressed)g(stream.)107 b(T)-8 b(o)53 b(start)g(uncompressing)150 -1140 y(that)61 b(next)f(stream,)68 b(call)62 b Fi(BZ2_bzReadOpen)56 -b Fj(again,)68 b(feeding)61 b(in)e(the)i(un)m(used)e(data)h(via)h(the) -150 1250 y Fi(unused)p Fj(/)p Fi(nUnused)28 b Fj(parameters.)45 -b(Keep)32 b(doing)f(this)h(un)m(til)g Fi(BZ_STREAM_END)c -Fj(return)j(coincides)h(with)150 1360 y(the)j(ph)m(ysical)g(end)f(of)h -(\014le)f(\()p Fi(feof\(f\))p Fj(\).)52 b(In)34 b(this)h(situation)g -Fi(BZ2_bzReadGetUnused)30 b Fj(will)35 b(of)f(course)150 -1469 y(return)c(no)g(data.)150 1626 y(This)24 b(should)g(giv)m(e)i -(some)f(feel)h(for)f(ho)m(w)g(the)g(high-lev)m(el)h(in)m(terface)h(can) -d(b)s(e)g(used.)38 b(If)25 b(y)m(ou)g(require)g(extra)150 -1736 y(\015exibilit)m(y)-8 b(,)32 b(y)m(ou'll)g(ha)m(v)m(e)f(to)g(bite) -g(the)g(bullet)g(and)e(get)j(to)f(grips)f(with)h(the)f(lo)m(w-lev)m(el) -k(in)m(terface.)150 1987 y Fe(3.4.9)63 b(Standard)40 -b(\014le-reading/writing)j(co)s(de)150 2179 y Fj(Here's)31 -b(ho)m(w)g(y)m(ou'd)f(write)h(data)g(to)g(a)g(compressed)f(\014le:)390 -2538 y Fi(FILE*)142 b(f;)390 2642 y(BZFILE*)46 b(b;)390 -2746 y(int)238 b(nBuf;)390 2849 y(char)190 b(buf[)46 -b(/*)i(whatever)d(size)i(you)g(like)f(*/)i(];)390 2953 -y(int)238 b(bzerror;)390 3057 y(int)g(nWritten;)390 3264 -y(f)47 b(=)h(fopen)e(\()i("myfile.bz2",)c("w")j(\);)390 -3368 y(if)g(\(!f\))g({)533 3472 y(/*)g(handle)f(error)h(*/)390 -3576 y(})390 3680 y(b)g(=)h(BZ2_bzWriteOpen)c(\()j(&bzerror,)e(f,)i(9)h -(\);)390 3783 y(if)f(\(bzerror)f(!=)h(BZ_OK\))f({)533 -3887 y(BZ2_bzWriteClose)e(\()j(b)g(\);)533 3991 y(/*)g(handle)f(error)h -(*/)390 4095 y(})390 4302 y(while)f(\()i(/*)f(condition)e(*/)i(\))h({) -533 4406 y(/*)f(get)g(data)g(to)g(write)f(into)h(buf,)g(and)g(set)g -(nBuf)f(appropriately)e(*/)533 4510 y(nWritten)i(=)h(BZ2_bzWrite)e(\()i -(&bzerror,)f(b,)h(buf,)f(nBuf)h(\);)533 4614 y(if)g(\(bzerror)f(==)h -(BZ_IO_ERROR\))e({)676 4717 y(BZ2_bzWriteClose)f(\()j(&bzerror,)e(b)j -(\);)676 4821 y(/*)g(handle)e(error)g(*/)533 4925 y(})390 -5029 y(})390 5236 y(BZ2_bzWriteClose)d(\()48 b(&bzerror,)d(b)j(\);)390 -5340 y(if)f(\(bzerror)f(==)h(BZ_IO_ERROR\))d({)p eop -%%Page: 27 28 -27 27 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(27)533 299 y Fi(/*)47 -b(handle)f(error)h(*/)390 403 y(})150 559 y Fj(And)29 -b(to)i(read)g(from)f(a)g(compressed)h(\014le:)390 711 -y Fi(FILE*)142 b(f;)390 814 y(BZFILE*)46 b(b;)390 918 -y(int)238 b(nBuf;)390 1022 y(char)190 b(buf[)46 b(/*)i(whatever)d(size) -i(you)g(like)f(*/)i(];)390 1126 y(int)238 b(bzerror;)390 -1229 y(int)g(nWritten;)390 1437 y(f)47 b(=)h(fopen)e(\()i -("myfile.bz2",)c("r")j(\);)390 1541 y(if)g(\(!f\))g({)533 -1645 y(/*)g(handle)f(error)h(*/)390 1748 y(})390 1852 -y(b)g(=)h(BZ2_bzReadOpen)c(\()j(&bzerror,)f(f,)h(0,)g(NULL,)f(0)i(\);) -390 1956 y(if)f(\(bzerror)f(!=)h(BZ_OK\))f({)533 2060 -y(BZ2_bzReadClose)e(\()j(&bzerror,)f(b)h(\);)533 2163 -y(/*)g(handle)f(error)h(*/)390 2267 y(})390 2475 y(bzerror)f(=)h -(BZ_OK;)390 2579 y(while)f(\(bzerror)g(==)h(BZ_OK)f(&&)i(/*)f -(arbitrary)e(other)h(conditions)f(*/\))i({)533 2682 y(nBuf)g(=)g -(BZ2_bzRead)e(\()j(&bzerror,)d(b,)i(buf,)g(/*)g(size)g(of)g(buf)g(*/)g -(\);)533 2786 y(if)g(\(bzerror)f(==)h(BZ_OK\))f({)676 -2890 y(/*)i(do)f(something)e(with)i(buf[0)f(..)h(nBuf-1])f(*/)533 -2994 y(})390 3097 y(})390 3201 y(if)h(\(bzerror)f(!=)h(BZ_STREAM_END\)) -d({)533 3305 y(BZ2_bzReadClose)g(\()j(&bzerror,)f(b)h(\);)533 -3409 y(/*)g(handle)f(error)h(*/)390 3513 y(})g(else)g({)533 -3616 y(BZ2_bzReadClose)d(\()j(&bzerror)f(\);)390 3720 -y(})150 3991 y Fk(3.5)68 b(Utilit)l(y)47 b(functions)150 -4264 y Fe(3.5.1)63 b Fd(BZ2_bzBuffToBuffCompress)533 -4451 y Fi(int)47 b(BZ2_bzBuffToBuffCompress\()41 b(char*)428 -b(dest,)1965 4555 y(unsigned)46 b(int*)g(destLen,)1965 -4658 y(char*)428 b(source,)1965 4762 y(unsigned)46 b(int)94 -b(sourceLen,)1965 4866 y(int)524 b(blockSize100k,)1965 -4970 y(int)g(verbosity,)1965 5074 y(int)g(workFactor)45 -b(\);)150 5230 y Fj(A)m(ttempts)32 b(to)f(compress)f(the)h(data)g(in)f -Fi(source[0)e(..)i(sourceLen-1])d Fj(in)m(to)k(the)g(destination)g -(bu\013er,)150 5340 y Fi(dest[0)e(..)g(*destLen-1])p -Fj(.)36 b(If)25 b(the)g(destination)h(bu\013er)e(is)h(big)f(enough,)i -Fi(*destLen)d Fj(is)i(set)g(to)h(the)f(size)p eop -%%Page: 28 29 -28 28 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(28)150 299 y Fj(of)27 -b(the)f(compressed)h(data,)h(and)d Fi(BZ_OK)g Fj(is)i(returned.)38 -b(If)27 b(the)f(compressed)h(data)g(w)m(on't)g(\014t,)g -Fi(*destLen)150 408 y Fj(is)j(unc)m(hanged,)g(and)g Fi(BZ_OUTBUFF_FULL) -c Fj(is)31 b(returned.)150 565 y(Compression)22 b(in)h(this)f(manner)g -(is)h(a)f(one-shot)i(ev)m(en)m(t,)i(done)d(with)f(a)h(single)g(call)h -(to)f(this)g(function.)38 b(The)150 675 y(resulting)26 -b(compressed)g(data)f(is)h(a)g(complete)h Fi(bzip2)d -Fj(format)i(data)g(stream.)39 b(There)26 b(is)f(no)h(mec)m(hanism)150 -784 y(for)d(making)g(additional)h(calls)f(to)h(pro)m(vide)g(extra)f -(input)g(data.)38 b(If)23 b(y)m(ou)g(w)m(an)m(t)h(that)f(kind)f(of)i -(mec)m(hanism,)150 894 y(use)30 b(the)h(lo)m(w-lev)m(el)i(in)m -(terface.)150 1051 y(F)-8 b(or)31 b(the)g(meaning)f(of)h(parameters)g -Fi(blockSize100k)p Fj(,)c Fi(verbosity)h Fj(and)h Fi(workFactor)p -Fj(,)150 1160 y(see)i Fi(BZ2_bzCompressInit)p Fj(.)150 -1317 y(T)-8 b(o)37 b(guaran)m(tee)h(that)f(the)f(compressed)h(data)f -(will)h(\014t)f(in)g(its)h(bu\013er,)g(allo)s(cate)h(an)e(output)g -(bu\013er)g(of)150 1427 y(size)31 b(1\045)g(larger)g(than)f(the)h -(uncompressed)e(data,)i(plus)f(six)g(h)m(undred)f(extra)i(b)m(ytes.)150 -1584 y Fi(BZ2_bzBuffToBuffDecompre)o(ss)24 b Fj(will)31 -b(not)g(write)g(data)f(at)h(or)g(b)s(ey)m(ond)e Fi(dest[*destLen])p -Fj(,)e(ev)m(en)k(in)150 1693 y(case)g(of)g(bu\013er)e(o)m(v)m(er\015o)m -(w.)150 1850 y(P)m(ossible)j(return)e(v)-5 b(alues:)572 -2001 y Fi(BZ_CONFIG_ERROR)663 2105 y Fj(if)30 b(the)h(library)f(has)g -(b)s(een)g(mis-compiled)572 2209 y Fi(BZ_PARAM_ERROR)663 -2313 y Fj(if)g Fi(dest)f Fj(is)i Fi(NULL)e Fj(or)i Fi(destLen)d -Fj(is)i Fi(NULL)663 2416 y Fj(or)g Fi(blockSize100k)45 -b(<)i(1)30 b Fj(or)h Fi(blockSize100k)44 b(>)j(9)663 -2520 y Fj(or)30 b Fi(verbosity)46 b(<)h(0)30 b Fj(or)h -Fi(verbosity)45 b(>)i(4)663 2624 y Fj(or)30 b Fi(workFactor)45 -b(<)j(0)30 b Fj(or)g Fi(workFactor)45 b(>)j(250)572 2728 -y(BZ_MEM_ERROR)663 2831 y Fj(if)30 b(insu\016cien)m(t)h(memory)f(is)h -(a)m(v)-5 b(ailable)572 2935 y Fi(BZ_OUTBUFF_FULL)663 -3039 y Fj(if)30 b(the)h(size)g(of)g(the)f(compressed)h(data)f(exceeds)i -Fi(*destLen)572 3143 y(BZ_OK)663 3247 y Fj(otherwise)150 -3616 y Fe(3.5.2)63 b Fd(BZ2_bzBuffToBuffDecompress)533 -3803 y Fi(int)47 b(BZ2_bzBuffToBuffDecompres)o(s)42 b(\()47 -b(char*)428 b(dest,)2108 3906 y(unsigned)46 b(int*)g(destLen,)2108 -4010 y(char*)428 b(source,)2108 4114 y(unsigned)46 b(int)94 -b(sourceLen,)2108 4218 y(int)524 b(small,)2108 4322 y(int)g(verbosity) -46 b(\);)150 4478 y Fj(A)m(ttempts)22 b(to)g(decompress)f(the)h(data)f -(in)g Fi(source[0)28 b(..)i(sourceLen-1])18 b Fj(in)m(to)k(the)f -(destination)h(bu\013er,)150 4588 y Fi(dest[0)29 b(..)g(*destLen-1])p -Fj(.)61 b(If)37 b(the)h(destination)h(bu\013er)e(is)h(big)g(enough,)i -Fi(*destLen)35 b Fj(is)j(set)h(to)g(the)150 4698 y(size)e(of)e(the)h -(uncompressed)f(data,)j(and)c Fi(BZ_OK)h Fj(is)g(returned.)56 -b(If)36 b(the)g(compressed)f(data)h(w)m(on't)h(\014t,)150 -4807 y Fi(*destLen)28 b Fj(is)j(unc)m(hanged,)f(and)f -Fi(BZ_OUTBUFF_FULL)d Fj(is)31 b(returned.)150 4964 y -Fi(source)e Fj(is)h(assumed)f(to)j(hold)e(a)g(complete)i -Fi(bzip2)d Fj(format)i(data)g(stream.)150 5074 y Fi -(BZ2_bzBuffToBuffDecompre)o(ss)38 b Fj(tries)46 b(to)f(decompress)f -(the)h(en)m(tiret)m(y)i(of)d(the)h(stream)g(in)m(to)h(the)150 -5183 y(output)30 b(bu\013er.)150 5340 y(F)-8 b(or)31 -b(the)g(meaning)f(of)h(parameters)g Fi(small)e Fj(and)g -Fi(verbosity)p Fj(,)f(see)j Fi(BZ2_bzDecompressInit)p -Fj(.)p eop -%%Page: 29 30 -29 29 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(29)150 299 y Fj(Because)31 -b(the)g(compression)g(ratio)g(of)g(the)g(compressed)f(data)h(cannot)g -(b)s(e)f(kno)m(wn)g(in)g(adv)-5 b(ance,)31 b(there)150 -408 y(is)g(no)f(easy)i(w)m(a)m(y)f(to)h(guaran)m(tee)g(that)f(the)g -(output)g(bu\013er)f(will)h(b)s(e)f(big)h(enough.)42 -b(Y)-8 b(ou)30 b(ma)m(y)i(of)e(course)150 518 y(mak)m(e)j(arrangemen)m -(ts)h(in)e(y)m(our)h(co)s(de)f(to)i(record)f(the)g(size)g(of)g(the)g -(uncompressed)e(data,)j(but)e(suc)m(h)g(a)150 628 y(mec)m(hanism)e(is)h -(b)s(ey)m(ond)f(the)g(scop)s(e)h(of)f(this)h(library)-8 -b(.)150 784 y Fi(BZ2_bzBuffToBuffDecompre)o(ss)24 b Fj(will)31 -b(not)g(write)g(data)f(at)h(or)g(b)s(ey)m(ond)e Fi(dest[*destLen])p -Fj(,)e(ev)m(en)k(in)150 894 y(case)g(of)g(bu\013er)e(o)m(v)m(er\015o)m -(w.)150 1051 y(P)m(ossible)j(return)e(v)-5 b(alues:)572 -1202 y Fi(BZ_CONFIG_ERROR)663 1306 y Fj(if)30 b(the)h(library)f(has)g -(b)s(een)g(mis-compiled)572 1409 y Fi(BZ_PARAM_ERROR)663 -1513 y Fj(if)g Fi(dest)f Fj(is)i Fi(NULL)e Fj(or)i Fi(destLen)d -Fj(is)i Fi(NULL)663 1617 y Fj(or)g Fi(small)47 b(!=)g(0)g(&&)h(small)e -(!=)h(1)663 1721 y Fj(or)30 b Fi(verbosity)46 b(<)h(0)30 -b Fj(or)h Fi(verbosity)45 b(>)i(4)572 1825 y(BZ_MEM_ERROR)663 -1928 y Fj(if)30 b(insu\016cien)m(t)h(memory)f(is)h(a)m(v)-5 -b(ailable)572 2032 y Fi(BZ_OUTBUFF_FULL)663 2136 y Fj(if)30 -b(the)h(size)g(of)g(the)f(compressed)h(data)f(exceeds)i -Fi(*destLen)572 2240 y(BZ_DATA_ERROR)663 2343 y Fj(if)e(a)h(data)f(in)m -(tegrit)m(y)j(error)e(w)m(as)g(detected)g(in)g(the)f(compressed)h(data) -572 2447 y Fi(BZ_DATA_ERROR_MAGIC)663 2551 y Fj(if)f(the)h(compressed)f -(data)h(do)s(esn't)f(b)s(egin)g(with)g(the)h(righ)m(t)g(magic)h(b)m -(ytes)572 2655 y Fi(BZ_UNEXPECTED_EOF)663 2759 y Fj(if)e(the)h -(compressed)f(data)h(ends)e(unexp)s(ectedly)572 2862 -y Fi(BZ_OK)663 2966 y Fj(otherwise)150 3370 y Fk(3.6)68 -b Fc(zlib)43 b Fk(compatibilit)l(y)k(functions)150 3563 -y Fj(Y)-8 b(oshiok)j(a)32 b(Tsuneo)f(has)f(con)m(tributed)i(some)f -(functions)g(to)h(giv)m(e)h(b)s(etter)e Fi(zlib)f Fj(compatibilit)m(y) --8 b(.)46 b(These)150 3673 y(functions)37 b(are)g Fi(BZ2_bzopen)p -Fj(,)f Fi(BZ2_bzread)p Fj(,)g Fi(BZ2_bzwrite)p Fj(,)g -Fi(BZ2_bzflush)p Fj(,)f Fi(BZ2_bzclose)p Fj(,)h Fi(BZ2_)150 -3782 y(bzerror)22 b Fj(and)g Fi(BZ2_bzlibVersion)p Fj(.)34 -b(These)24 b(functions)f(are)h(not)g(\(y)m(et\))h(o\016cially)g(part)f -(of)g(the)f(library)-8 b(.)150 3892 y(If)30 b(they)h(break,)f(y)m(ou)h -(get)h(to)f(k)m(eep)g(all)g(the)g(pieces.)41 b(Nev)m(ertheless,)33 -b(I)e(think)f(they)g(w)m(ork)h(ok.)390 4043 y Fi(typedef)46 -b(void)g(BZFILE;)390 4250 y(const)g(char)h(*)g(BZ2_bzlibVersion)d(\()j -(void)g(\);)150 4407 y Fj(Returns)30 b(a)h(string)f(indicating)h(the)g -(library)f(v)m(ersion.)390 4558 y Fi(BZFILE)46 b(*)i(BZ2_bzopen)92 -b(\()48 b(const)e(char)h(*path,)f(const)g(char)h(*mode)f(\);)390 -4662 y(BZFILE)g(*)i(BZ2_bzdopen)c(\()k(int)381 b(fd,)190 -b(const)46 b(char)h(*mode)f(\);)150 4819 y Fj(Op)s(ens)36 -b(a)i Fi(.bz2)e Fj(\014le)i(for)f(reading)h(or)f(writing,)j(using)d -(either)h(its)g(name)f(or)h(a)g(pre-existing)g(\014le)g(de-)150 -4928 y(scriptor.)j(Analogous)31 b(to)g Fi(fopen)e Fj(and)h -Fi(fdopen)p Fj(.)390 5079 y Fi(int)47 b(BZ2_bzread)93 -b(\()47 b(BZFILE*)f(b,)h(void*)f(buf,)h(int)g(len)g(\);)390 -5183 y(int)g(BZ2_bzwrite)e(\()i(BZFILE*)f(b,)h(void*)f(buf,)h(int)g -(len)g(\);)150 5340 y Fj(Reads/writes)31 b(data)g(from/to)g(a)g -(previously)f(op)s(ened)f Fi(BZFILE)p Fj(.)39 b(Analogous)31 -b(to)g Fi(fread)e Fj(and)g Fi(fwrite)p Fj(.)p eop -%%Page: 30 31 -30 30 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(30)390 299 y Fi(int)95 -b(BZ2_bzflush)44 b(\()k(BZFILE*)e(b)h(\);)390 403 y(void)g(BZ2_bzclose) -d(\()k(BZFILE*)e(b)h(\);)150 559 y Fj(Flushes/closes)29 -b(a)f Fi(BZFILE)p Fj(.)38 b Fi(BZ2_bzflush)25 b Fj(do)s(esn't)j -(actually)h(do)f(an)m(ything.)41 b(Analogous)28 b(to)h -Fi(fflush)150 669 y Fj(and)h Fi(fclose)p Fj(.)390 820 -y Fi(const)46 b(char)h(*)g(BZ2_bzerror)e(\()j(BZFILE)e(*b,)h(int)g -(*errnum)e(\))150 977 y Fj(Returns)29 b(a)h(string)g(describing)g(the)f -(more)i(recen)m(t)g(error)f(status)g(of)f Fi(b)p Fj(,)h(and)f(also)h -(sets)g Fi(*errnum)e Fj(to)i(its)150 1087 y(n)m(umerical)h(v)-5 -b(alue.)150 1411 y Fk(3.7)68 b(Using)46 b(the)f(library)g(in)g(a)g -Fc(stdio)p Fk(-free)f(en)l(vironmen)l(t)150 1723 y Fe(3.7.1)63 -b(Getting)41 b(rid)g(of)g Fd(stdio)150 1915 y Fj(In)g(a)g(deeply)h(em)m -(b)s(edded)e(application,)46 b(y)m(ou)41 b(migh)m(t)i(w)m(an)m(t)f(to)g -(use)f(just)g(the)h(memory-to-memory)150 2025 y(functions.)d(Y)-8 -b(ou)27 b(can)g(do)g(this)g(con)m(v)m(enien)m(tly)j(b)m(y)d(compiling)h -(the)f(library)g(with)g(prepro)s(cessor)g(sym)m(b)s(ol)150 -2135 y Fi(BZ_NO_STDIO)34 b Fj(de\014ned.)60 b(Doing)38 -b(this)f(giv)m(es)i(y)m(ou)e(a)h(library)f(con)m(taining)i(only)e(the)h -(follo)m(wing)g(eigh)m(t)150 2244 y(functions:)150 2401 -y Fi(BZ2_bzCompressInit)p Fj(,)26 b Fi(BZ2_bzCompress)p -Fj(,)g Fi(BZ2_bzCompressEnd)150 2511 y(BZ2_bzDecompressInit)p -Fj(,)f Fi(BZ2_bzDecompress)p Fj(,)h Fi(BZ2_bzDecompressEnd)150 -2620 y(BZ2_bzBuffToBuffCompress)o Fj(,)f Fi(BZ2_bzBuffToBuffDecompre)o -(ss)150 2777 y Fj(When)30 b(compiled)h(lik)m(e)g(this,)g(all)g -(functions)f(will)h(ignore)g Fi(verbosity)d Fj(settings.)150 -3053 y Fe(3.7.2)63 b(Critical)40 b(error)h(handling)150 -3246 y Fi(libbzip2)18 b Fj(con)m(tains)k(a)f(n)m(um)m(b)s(er)e(of)i(in) -m(ternal)h(assertion)f(c)m(hec)m(ks)h(whic)m(h)f(should,)g(needless)g -(to)h(sa)m(y)-8 b(,)23 b(nev)m(er)150 3355 y(b)s(e)35 -b(activ)-5 b(ated.)58 b(Nev)m(ertheless,)40 b(if)35 b(an)h(assertion)g -(should)f(fail,)i(b)s(eha)m(viour)f(dep)s(ends)e(on)h(whether)h(or)150 -3465 y(not)31 b(the)f(library)h(w)m(as)f(compiled)h(with)f -Fi(BZ_NO_STDIO)e Fj(set.)150 3622 y(F)-8 b(or)31 b(a)g(normal)f -(compile,)i(an)e(assertion)h(failure)f(yields)h(the)g(message)533 -3773 y Fi(bzip2/libbzip2:)44 b(internal)h(error)i(number)f(N.)533 -3877 y(This)h(is)g(a)g(bug)g(in)h(bzip2/libbzip2,)43 -b(1.0.2,)j(30-Dec-2001.)533 3980 y(Please)g(report)g(it)i(to)f(me)g -(at:)g(jseward@acm.org.)91 b(If)47 b(this)g(happened)533 -4084 y(when)g(you)g(were)f(using)h(some)f(program)g(which)h(uses)f -(libbzip2)g(as)h(a)533 4188 y(component,)e(you)i(should)f(also)h -(report)f(this)h(bug)f(to)i(the)f(author\(s\))533 4292 -y(of)g(that)g(program.)93 b(Please)46 b(make)h(an)g(effort)f(to)h -(report)g(this)f(bug;)533 4395 y(timely)g(and)h(accurate)f(bug)h -(reports)e(eventually)g(lead)i(to)g(higher)533 4499 y(quality)f -(software.)93 b(Thanks.)h(Julian)46 b(Seward,)f(30)j(December)d(2001.) -150 4656 y Fj(where)33 b Fi(N)g Fj(is)g(some)g(error)h(co)s(de)f(n)m -(um)m(b)s(er.)48 b(If)32 b Fi(N)e(==)g(1007)p Fj(,)j(it)h(also)f(prin)m -(ts)g(some)h(extra)g(text)g(advising)150 4766 y(the)e(reader)f(that)h -(unreliable)f(memory)h(is)f(often)h(asso)s(ciated)g(with)f(in)m(ternal) -h(error)g(1007.)45 b(\(This)32 b(is)f(a)150 4875 y(frequen)m -(tly-observ)m(ed-phenomenon)g(with)f(v)m(ersions)h(1.0.0/1.0.1\).)150 -5032 y Fi(exit\(3\))d Fj(is)j(then)f(called.)150 5189 -y(F)-8 b(or)31 b(a)g Fi(stdio)p Fj(-free)e(library)-8 -b(,)31 b(assertion)h(failures)e(result)h(in)f(a)g(call)i(to)f(a)f -(function)h(declared)f(as:)533 5340 y Fi(extern)46 b(void)h -(bz_internal_error)c(\()k(int)g(errcode)f(\);)p eop -%%Page: 31 32 -31 31 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29 -b(with)g Fi(libbzip2)1881 b Fl(31)150 299 y Fj(The)30 -b(relev)-5 b(an)m(t)32 b(co)s(de)f(is)f(passed)g(as)g(a)h(parameter.)41 -b(Y)-8 b(ou)31 b(should)e(supply)g(suc)m(h)h(a)g(function.)150 -456 y(In)f(either)h(case,)g(once)g(an)f(assertion)h(failure)f(has)f(o)s -(ccurred,)i(an)m(y)f Fi(bz_stream)e Fj(records)i(in)m(v)m(olv)m(ed)i -(can)150 565 y(b)s(e)f(regarded)g(as)h(in)m(v)-5 b(alid.)41 -b(Y)-8 b(ou)30 b(should)f(not)i(attempt)h(to)f(resume)f(normal)h(op)s -(eration)g(with)f(them.)150 722 y(Y)-8 b(ou)27 b(ma)m(y)-8 -b(,)28 b(of)f(course,)h(c)m(hange)f(critical)i(error)e(handling)f(to)h -(suit)g(y)m(our)g(needs.)39 b(As)26 b(I)h(said)f(ab)s(o)m(v)m(e,)j -(crit-)150 832 y(ical)f(errors)g(indicate)g(bugs)e(in)h(the)h(library)f -(and)f(should)g(not)i(o)s(ccur.)40 b(All)27 b Fi(")p -Fj(normal)p Fi(")g Fj(error)g(situations)150 941 y(are)k(indicated)g -(via)f(error)h(return)f(co)s(des)h(from)f(functions,)g(and)g(can)g(b)s -(e)g(reco)m(v)m(ered)i(from.)150 1221 y Fk(3.8)68 b(Making)45 -b(a)g(Windo)l(ws)h(DLL)150 1414 y Fj(Ev)m(erything)31 -b(related)g(to)h(Windo)m(ws)e(has)f(b)s(een)h(con)m(tributed)h(b)m(y)f -(Y)-8 b(oshiok)j(a)32 b(Tsuneo)150 1523 y(\()p Fi -(QWF00133@niftyserve.or.jp)45 b Fj(/)53 b Fi(tsuneo-y@is.aist-nara.a)o -(c.jp)o Fj(\),)f(so)g(y)m(ou)h(should)e(send)150 1633 -y(y)m(our)31 b(queries)f(to)h(him)f(\(but)g(p)s(erhaps)f(Cc:)41 -b(me,)31 b Fi(jseward@acm.org)p Fj(\).)150 1790 y(My)42 -b(v)-5 b(ague)43 b(understanding)d(of)i(what)g(to)h(do)f(is:)64 -b(using)41 b(Visual)h(C)p Fi(++)f Fj(5.0,)46 b(op)s(en)c(the)g(pro)5 -b(ject)43 b(\014le)150 1899 y Fi(libbz2.dsp)p Fj(,)28 -b(and)h(build.)40 b(That's)31 b(all.)150 2056 y(If)39 -b(y)m(ou)g(can't)h(op)s(en)f(the)g(pro)5 b(ject)40 b(\014le)f(for)g -(some)h(reason,)i(mak)m(e)e(a)f(new)f(one,)k(naming)d(these)h(\014les:) -150 2166 y Fi(blocksort.c)p Fj(,)28 b Fi(bzlib.c)p Fj(,)g -Fi(compress.c)p Fj(,)g Fi(crctable.c)p Fj(,)g Fi(decompress.c)p -Fj(,)f Fi(huffman.c)p Fj(,)150 2275 y Fi(randtable.c)32 -b Fj(and)h Fi(libbz2.def)p Fj(.)51 b(Y)-8 b(ou)35 b(will)g(also)g(need) -f(to)i(name)e(the)h(header)g(\014les)f Fi(bzlib.h)f Fj(and)150 -2385 y Fi(bzlib_private.h)p Fj(.)150 2542 y(If)d(y)m(ou)h(don't)f(use)h -(V)m(C)p Fi(++)p Fj(,)e(y)m(ou)i(ma)m(y)g(need)f(to)h(de\014ne)f(the)h -(propro)s(cessor)f(sym)m(b)s(ol)g Fi(_WIN32)p Fj(.)150 -2698 y(Finally)-8 b(,)30 b Fi(dlltest.c)25 b Fj(is)j(a)g(sample)g -(program)h(using)e(the)i(DLL.)e(It)h(has)g(a)g(pro)5 -b(ject)29 b(\014le,)g Fi(dlltest.dsp)p Fj(.)150 2855 -y(If)h(y)m(ou)h(just)f(w)m(an)m(t)h(a)f(mak)m(e\014le)i(for)e(Visual)g -(C,)g(ha)m(v)m(e)i(a)e(lo)s(ok)h(at)g Fi(makefile.msc)p -Fj(.)150 3012 y(Be)j(a)m(w)m(are)i(that)f(if)g(y)m(ou)f(compile)i -Fi(bzip2)d Fj(itself)i(on)g(Win32,)h(y)m(ou)f(m)m(ust)f(set)i -Fi(BZ_UNIX)c Fj(to)j(0)g(and)f Fi(BZ_)150 3122 y(LCCWIN32)25 -b Fj(to)j(1,)g(in)f(the)h(\014le)f Fi(bzip2.c)p Fj(,)f(b)s(efore)h -(compiling.)41 b(Otherwise)27 b(the)h(resulting)g(binary)e(w)m(on't)150 -3231 y(w)m(ork)31 b(correctly)-8 b(.)150 3388 y(I)30 -b(ha)m(v)m(en't)i(tried)f(an)m(y)f(of)h(this)f(stu\013)g(m)m(yself,)i -(but)d(it)i(all)g(lo)s(oks)g(plausible.)p eop -%%Page: 32 33 -32 32 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586 -b(32)150 299 y Fg(4)80 b(Miscellanea)150 549 y Fj(These)31 -b(are)f(just)g(some)h(random)f(though)m(ts)h(of)f(mine.)41 -b(Y)-8 b(our)30 b(mileage)i(ma)m(y)f(v)-5 b(ary)d(.)150 -826 y Fk(4.1)68 b(Limitations)47 b(of)e(the)g(compressed)g(\014le)h -(format)150 1018 y Fi(bzip2-1.0)p Fj(,)c Fi(0.9.5)e Fj(and)h -Fi(0.9.0)f Fj(use)i(exactly)h(the)f(same)g(\014le)g(format)g(as)f(the)h -(previous)g(v)m(ersion,)150 1128 y Fi(bzip2-0.1)p Fj(.)72 -b(This)41 b(decision)h(w)m(as)g(made)f(in)g(the)h(in)m(terests)h(of)f -(stabilit)m(y)-8 b(.)76 b(Creating)43 b(y)m(et)g(another)150 -1238 y(incompatible)i(compressed)g(\014le)g(format)g(w)m(ould)f(create) -i(further)e(confusion)h(and)e(disruption)h(for)150 1347 -y(users.)150 1504 y(Nev)m(ertheless,)30 b(this)d(is)f(not)h(a)g -(painless)g(decision.)40 b(Dev)m(elopmen)m(t)28 b(w)m(ork)f(since)h -(the)f(release)h(of)f Fi(bzip2-)150 1614 y(0.1)g Fj(in)h(August)g(1997) -i(has)d(sho)m(wn)h(complexities)i(in)e(the)g(\014le)g(format)h(whic)m -(h)f(slo)m(w)h(do)m(wn)e(decompres-)150 1723 y(sion)j(and,)g(in)g -(retrosp)s(ect,)i(are)f(unnecessary)-8 b(.)41 b(These)30 -b(are:)225 1880 y Fh(\017)60 b Fj(The)33 b(run-length)g(enco)s(der,)h -(whic)m(h)f(is)g(the)h(\014rst)f(of)g(the)h(compression)f -(transformations,)i(is)e(en-)330 1990 y(tirely)h(irrelev)-5 -b(an)m(t.)48 b(The)32 b(original)i(purp)s(ose)d(w)m(as)i(to)g(protect)h -(the)f(sorting)g(algorithm)g(from)g(the)330 2099 y(v)m(ery)26 -b(w)m(orst)g(case)h(input:)37 b(a)26 b(string)g(of)f(rep)s(eated)h(sym) -m(b)s(ols.)39 b(But)25 b(algorithm)h(steps)g(Q6a)g(and)f(Q6b)330 -2209 y(in)j(the)g(original)i(Burro)m(ws-Wheeler)e(tec)m(hnical)i(rep)s -(ort)f(\(SR)m(C-124\))h(sho)m(w)e(ho)m(w)g(rep)s(eats)g(can)h(b)s(e)330 -2318 y(handled)g(without)i(di\016cult)m(y)g(in)f(blo)s(c)m(k)h -(sorting.)225 2451 y Fh(\017)60 b Fj(The)27 b(randomisation)h(mec)m -(hanism)f(do)s(esn't)g(really)i(need)e(to)h(b)s(e)e(there.)41 -b(Udi)27 b(Man)m(b)s(er)g(and)f(Gene)330 2561 y(My)m(ers)31 -b(published)d(a)j(su\016x)e(arra)m(y)h(construction)i(algorithm)f(a)f -(few)g(y)m(ears)h(bac)m(k,)g(whic)m(h)e(can)i(b)s(e)330 -2670 y(emplo)m(y)m(ed)26 b(to)g(sort)f(an)m(y)g(blo)s(c)m(k,)i(no)e -(matter)h(ho)m(w)f(rep)s(etitiv)m(e,)j(in)c(O\(N)i(log)g(N\))f(time.)40 -b(Subsequen)m(t)330 2780 y(w)m(ork)24 b(b)m(y)f(Kunihik)m(o)g(Sadak)-5 -b(ane)23 b(has)g(pro)s(duced)e(a)j(deriv)-5 b(ativ)m(e)25 -b(O\(N)f(\(log)g(N\))p Fi(^)p Fj(2\))h(algorithm)f(whic)m(h)330 -2890 y(usually)30 b(outp)s(erforms)g(the)g(Man)m(b)s(er-My)m(ers)i -(algorithm.)330 3023 y(I)f(could)f(ha)m(v)m(e)i(c)m(hanged)f(to)g -(Sadak)-5 b(ane's)30 b(algorithm,)i(but)e(I)h(\014nd)e(it)i(to)g(b)s(e) -f(slo)m(w)m(er)i(than)e Fi(bzip2)p Fj('s)330 3132 y(existing)39 -b(algorithm)f(for)g(most)g(inputs,)h(and)d(the)i(randomisation)g(mec)m -(hanism)g(protects)h(ade-)330 3242 y(quately)34 b(against)g(bad)f -(cases.)50 b(I)33 b(didn't)g(think)g(it)h(w)m(as)f(a)h(go)s(o)s(d)f -(tradeo\013)h(to)g(mak)m(e.)50 b(P)m(artly)35 b(this)330 -3351 y(is)40 b(due)f(to)i(the)f(fact)h(that)f(I)g(w)m(as)g(not)g(\015o) -s(o)s(ded)f(with)g(email)i(complain)m(ts)g(ab)s(out)e -Fi(bzip2-0.1)p Fj('s)330 3461 y(p)s(erformance)30 b(on)g(rep)s(etitiv)m -(e)j(data,)e(so)f(p)s(erhaps)f(it)i(isn't)g(a)f(problem)g(for)h(real)g -(inputs.)330 3594 y(Probably)g(the)h(b)s(est)f(long-term)i(solution,)f -(and)f(the)g(one)h(I)f(ha)m(v)m(e)i(incorp)s(orated)e(in)m(to)i(0.9.5)g -(and)330 3704 y(ab)s(o)m(v)m(e,)41 b(is)d(to)h(use)f(the)h(existing)g -(sorting)g(algorithm)g(initially)-8 b(,)42 b(and)37 b(fall)i(bac)m(k)g -(to)f(a)h(O\(N)f(\(log)330 3813 y(N\))p Fi(^)p Fj(2\))31 -b(algorithm)h(if)e(the)h(standard)e(algorithm)j(gets)f(in)m(to)h -(di\016culties.)225 3946 y Fh(\017)60 b Fj(The)29 b(compressed)g -(\014le)g(format)h(w)m(as)f(nev)m(er)g(designed)g(to)h(b)s(e)e(handled) -g(b)m(y)h(a)g(library)-8 b(,)30 b(and)e(I)h(ha)m(v)m(e)330 -4056 y(had)e(to)h(jump)e(though)h(some)h(ho)s(ops)e(to)i(pro)s(duce)f -(an)g(e\016cien)m(t)i(implemen)m(tation)g(of)e(decompres-)330 -4165 y(sion.)39 b(It's)24 b(a)h(bit)f(hairy)-8 b(.)39 -b(T)-8 b(ry)24 b(passing)g Fi(decompress.c)d Fj(through)i(the)i(C)e -(prepro)s(cessor)i(and)e(y)m(ou'll)330 4275 y(see)31 -b(what)e(I)h(mean.)41 b(Muc)m(h)30 b(of)g(this)g(complexit)m(y)i(could) -e(ha)m(v)m(e)g(b)s(een)g(a)m(v)m(oided)h(if)f(the)g(compressed)330 -4384 y(size)h(of)g(eac)m(h)g(blo)s(c)m(k)g(of)g(data)f(w)m(as)h -(recorded)f(in)h(the)f(data)h(stream.)225 4517 y Fh(\017)60 -b Fj(An)29 b(Adler-32)j(c)m(hec)m(ksum,)f(rather)g(than)f(a)g(CR)m(C32) -h(c)m(hec)m(ksum,)h(w)m(ould)e(b)s(e)g(faster)h(to)g(compute.)150 -4698 y(It)45 b(w)m(ould)g(b)s(e)f(fair)h(to)h(sa)m(y)f(that)g(the)h -Fi(bzip2)d Fj(format)i(w)m(as)g(frozen)g(b)s(efore)g(I)g(prop)s(erly)f -(and)g(fully)150 4807 y(understo)s(o)s(d)29 b(the)i(p)s(erformance)f -(consequences)h(of)f(doing)h(so.)150 4964 y(Impro)m(v)m(emen)m(ts)25 -b(whic)m(h)d(I)i(w)m(as)f(able)g(to)h(incorp)s(orate)g(in)m(to)g -(0.9.0,)j(despite)d(using)e(the)i(same)f(\014le)g(format,)150 -5074 y(are:)225 5230 y Fh(\017)60 b Fj(Single)30 b(arra)m(y)g(implemen) -m(tation)i(of)d(the)h(in)m(v)m(erse)h(BWT.)f(This)f(signi\014can)m(tly) -i(sp)s(eeds)e(up)f(decom-)330 5340 y(pression,)i(presumably)f(b)s -(ecause)i(it)g(reduces)f(the)h(n)m(um)m(b)s(er)e(of)i(cac)m(he)g -(misses.)p eop -%%Page: 33 34 -33 33 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586 -b(33)225 299 y Fh(\017)60 b Fj(F)-8 b(aster)25 b(in)m(v)m(erse)f(MTF)g -(transform)f(for)g(large)i(MTF)f(v)-5 b(alues.)38 b(The)23 -b(new)g(implemen)m(tation)i(is)f(based)330 408 y(on)30 -b(the)h(notion)g(of)f(sliding)h(blo)s(c)m(ks)g(of)f(v)-5 -b(alues.)225 544 y Fh(\017)60 b Fi(bzip2-0.9.0)23 b Fj(no)m(w)j(reads)g -(and)f(writes)h(\014les)g(with)g Fi(fread)f Fj(and)g -Fi(fwrite)p Fj(;)h(v)m(ersion)h(0.1)g(used)e Fi(putc)330 -653 y Fj(and)30 b Fi(getc)p Fj(.)39 b(Duh!)g(W)-8 b(ell,)32 -b(y)m(ou)f(liv)m(e)g(and)f(learn.)150 836 y(F)-8 b(urther)28 -b(ahead,)h(it)g(w)m(ould)f(b)s(e)g(nice)g(to)i(b)s(e)d(able)i(to)g(do)f -(random)f(access)j(in)m(to)f(\014les.)40 b(This)28 b(will)h(require)150 -945 y(some)i(careful)g(design)f(of)g(compressed)h(\014le)f(formats.)150 -1227 y Fk(4.2)68 b(P)l(ortabilit)l(y)47 b(issues)150 -1419 y Fj(After)33 b(some)h(consideration,)h(I)e(ha)m(v)m(e)h(decided)f -(not)g(to)h(use)f(GNU)g Fi(autoconf)e Fj(to)j(con\014gure)f(0.9.5)i(or) -150 1529 y(1.0.)150 1686 y Fi(autoconf)p Fj(,)30 b(admirable)i(and)e(w) -m(onderful)h(though)h(it)g(is,)h(mainly)e(assists)h(with)g(p)s -(ortabilit)m(y)g(problems)150 1795 y(b)s(et)m(w)m(een)46 -b(Unix-lik)m(e)h(platforms.)85 b(But)45 b Fi(bzip2)f -Fj(do)s(esn't)h(ha)m(v)m(e)h(m)m(uc)m(h)g(in)f(the)g(w)m(a)m(y)h(of)g -(p)s(ortabilit)m(y)150 1905 y(problems)21 b(on)h(Unix;)j(most)d(of)g -(the)g(di\016culties)g(app)s(ear)f(when)g(p)s(orting)h(to)g(the)h(Mac,) -h(or)e(to)h(Microsoft's)150 2015 y(op)s(erating)j(systems.)40 -b Fi(autoconf)23 b Fj(do)s(esn't)j(help)f(in)g(those)i(cases,)g(and)e -(brings)g(in)h(a)g(whole)f(load)i(of)e(new)150 2124 y(complexit)m(y)-8 -b(.)150 2281 y(Most)34 b(p)s(eople)e(should)g(b)s(e)g(able)h(to)g -(compile)h(the)f(library)f(and)g(program)h(under)e(Unix)h(straigh)m(t)i -(out-)150 2391 y(of-the-b)s(o)m(x,)e(so)e(to)i(sp)s(eak,)e(esp)s -(ecially)h(if)f(y)m(ou)h(ha)m(v)m(e)g(a)g(v)m(ersion)g(of)g(GNU)g(C)e -(a)m(v)-5 b(ailable.)150 2547 y(There)31 b(are)h(a)f(couple)g(of)g -Fi(__inline__)e Fj(directiv)m(es)k(in)d(the)i(co)s(de.)42 -b(GNU)32 b(C)f(\()p Fi(gcc)p Fj(\))f(should)g(b)s(e)h(able)g(to)150 -2657 y(handle)23 b(them.)38 b(If)23 b(y)m(ou're)h(not)g(using)f(GNU)h -(C,)f(y)m(our)h(C)e(compiler)j(shouldn't)d(see)i(them)g(at)g(all.)39 -b(If)23 b(y)m(our)150 2767 y(compiler)30 b(do)s(es,)f(for)f(some)h -(reason,)h(see)f(them)g(and)f(do)s(esn't)g(lik)m(e)i(them,)g(just)e -Fi(#define)f(__inline__)150 2876 y Fj(to)37 b(b)s(e)f -Fi(/*)29 b(*/)p Fj(.)58 b(One)36 b(easy)h(w)m(a)m(y)f(to)i(do)e(this)g -(is)g(to)h(compile)g(with)f(the)h(\015ag)f Fi(-D__inline__=)p -Fj(,)e(whic)m(h)150 2986 y(should)29 b(b)s(e)h(understo)s(o)s(d)f(b)m -(y)h(most)h(Unix)f(compilers.)150 3143 y(If)k(y)m(ou)g(still)h(ha)m(v)m -(e)g(di\016culties,)h(try)e(compiling)g(with)g(the)g(macro)h -Fi(BZ_STRICT_ANSI)30 b Fj(de\014ned.)50 b(This)150 3252 -y(should)27 b(enable)i(y)m(ou)g(to)g(build)e(the)i(library)f(in)g(a)g -(strictly)i(ANSI)e(complian)m(t)i(en)m(vironmen)m(t.)41 -b(Building)150 3362 y(the)22 b(program)h(itself)g(lik)m(e)g(this)f(is)g -(dangerous)g(and)f(not)h(supp)s(orted,)h(since)f(y)m(ou)h(remo)m(v)m(e) -h Fi(bzip2)p Fj('s)c(c)m(hec)m(ks)150 3471 y(against)27 -b(compressing)g(directories,)i(sym)m(b)s(olic)d(links,)h(devices,)h -(and)e(other)h(not-really-a-\014le)i(en)m(tities.)150 -3581 y(This)h(could)g(cause)h(\014lesystem)g(corruption!)150 -3738 y(One)c(other)h(thing:)40 b(if)27 b(y)m(ou)h(create)h(a)e -Fi(bzip2)f Fj(binary)h(for)g(public)g(distribution,)h(please)g(try)g -(and)e(link)h(it)150 3847 y(statically)g(\()p Fi(gcc)j(-s)p -Fj(\).)39 b(This)24 b(a)m(v)m(oids)j(all)e(sorts)h(of)f(library-v)m -(ersion)h(issues)e(that)i(others)g(ma)m(y)f(encoun)m(ter)150 -3957 y(later)32 b(on.)150 4114 y(If)d(y)m(ou)g(build)e -Fi(bzip2)h Fj(on)g(Win32,)i(y)m(ou)f(m)m(ust)g(set)h -Fi(BZ_UNIX)c Fj(to)k(0)f(and)f Fi(BZ_LCCWIN32)e Fj(to)k(1,)f(in)g(the)g -(\014le)150 4223 y Fi(bzip2.c)p Fj(,)g(b)s(efore)h(compiling.)41 -b(Otherwise)31 b(the)g(resulting)g(binary)f(w)m(on't)h(w)m(ork)g -(correctly)-8 b(.)150 4505 y Fk(4.3)68 b(Rep)t(orting)46 -b(bugs)150 4698 y Fj(I)23 b(tried)g(prett)m(y)h(hard)e(to)i(mak)m(e)f -(sure)g Fi(bzip2)e Fj(is)i(bug)f(free,)j(b)s(oth)d(b)m(y)h(design)g -(and)f(b)m(y)h(testing.)39 b(Hop)s(efully)150 4807 y(y)m(ou'll)31 -b(nev)m(er)g(need)g(to)g(read)f(this)h(section)g(for)g(real.)150 -4964 y(Nev)m(ertheless,)j(if)d Fi(bzip2)f Fj(dies)h(with)g(a)g(segmen)m -(tation)j(fault,)e(a)f(bus)f(error)i(or)f(an)g(in)m(ternal)h(assertion) -150 5074 y(failure,)j(it)f(will)f(ask)h(y)m(ou)g(to)g(email)g(me)g(a)f -(bug)g(rep)s(ort.)50 b(Exp)s(erience)33 b(with)g(v)m(ersion)i(0.1)f -(sho)m(ws)f(that)150 5183 y(almost)e(all)g(these)g(problems)f(can)h(b)s -(e)e(traced)j(to)f(either)g(compiler)g(bugs)f(or)h(hardw)m(are)f -(problems.)225 5340 y Fh(\017)60 b Fj(Recompile)22 b(the)f(program)f -(with)h(no)f(optimisation,)k(and)c(see)h(if)g(it)g(w)m(orks.)38 -b(And/or)20 b(try)g(a)h(di\013eren)m(t)p eop -%%Page: 34 35 -34 34 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586 -b(34)330 299 y Fj(compiler.)74 b(I)41 b(heard)f(all)i(sorts)f(of)g -(stories)i(ab)s(out)d(v)-5 b(arious)41 b(\015a)m(v)m(ours)g(of)h(GNU)f -(C)g(\(and)f(other)330 408 y(compilers\))45 b(generating)g(bad)e(co)s -(de)g(for)h Fi(bzip2)p Fj(,)i(and)d(I'v)m(e)i(run)d(across)i(t)m(w)m(o) -h(suc)m(h)f(examples)330 518 y(m)m(yself.)330 716 y(2.7.X)i(v)m -(ersions)f(of)g(GNU)h(C)e(are)h(kno)m(wn)g(to)g(generate)i(bad)d(co)s -(de)h(from)f(time)i(to)g(time,)j(at)330 825 y(high)32 -b(optimisation)h(lev)m(els.)47 b(If)31 b(y)m(ou)i(get)g(problems,)f -(try)g(using)f(the)i(\015ags)f Fi(-O2)f(-fomit-frame-)330 -935 y(pointer)d(-fno-strength-reduce)p Fj(.)36 b(Y)-8 -b(ou)30 b(should)f(sp)s(eci\014cally)i Fb(not)40 b Fj(use)30 -b Fi(-funroll-loops)p Fj(.)330 1132 y(Y)-8 b(ou)36 b(ma)m(y)g(notice)i -(that)e(the)g(Mak)m(e\014le)i(runs)d(six)g(tests)i(as)f(part)g(of)g -(the)h(build)e(pro)s(cess.)57 b(If)36 b(the)330 1242 -y(program)42 b(passes)g(all)h(of)f(these,)k(it's)d(a)f(prett)m(y)i(go)s -(o)s(d)e(\(but)g(not)g(100\045\))i(indication)f(that)g(the)330 -1352 y(compiler)31 b(has)f(done)g(its)h(job)f(correctly)-8 -b(.)225 1549 y Fh(\017)60 b Fj(If)32 b Fi(bzip2)e Fj(crashes)j -(randomly)-8 b(,)32 b(and)f(the)i(crashes)f(are)g(not)h(rep)s(eatable,) -g(y)m(ou)g(ma)m(y)f(ha)m(v)m(e)h(a)f(\015aky)330 1659 -y(memory)37 b(subsystem.)57 b Fi(bzip2)35 b Fj(really)i(hammers)f(y)m -(our)h(memory)f(hierarc)m(h)m(y)-8 b(,)39 b(and)d(if)g(it's)h(a)g(bit) -330 1768 y(marginal,)c(y)m(ou)g(ma)m(y)g(get)g(these)g(problems.)46 -b(Ditto)34 b(if)e(y)m(our)h(disk)f(or)g(I/O)h(subsystem)f(is)g(slo)m -(wly)330 1878 y(failing.)41 b(Y)-8 b(up,)30 b(this)g(really)i(do)s(es)e -(happ)s(en.)330 2075 y(T)-8 b(ry)27 b(using)f(a)h(di\013eren)m(t)h(mac) -m(hine)f(of)g(the)g(same)g(t)m(yp)s(e,)h(and)e(see)h(if)g(y)m(ou)g(can) -g(rep)s(eat)g(the)g(problem.)225 2273 y Fh(\017)60 b -Fj(This)21 b(isn't)h(really)g(a)g(bug,)h(but)d(...)39 -b(If)21 b Fi(bzip2)f Fj(tells)i(y)m(ou)g(y)m(our)g(\014le)f(is)h -(corrupted)f(on)g(decompression,)330 2383 y(and)g(y)m(ou)h(obtained)g -(the)h(\014le)f(via)g(FTP)-8 b(,)23 b(there)f(is)g(a)g(p)s(ossibilit)m -(y)h(that)f(y)m(ou)h(forgot)g(to)g(tell)g(FTP)f(to)h(do)330 -2492 y(a)31 b(binary)g(mo)s(de)g(transfer.)43 b(That)31 -b(absolutely)h(will)g(cause)f(the)h(\014le)f(to)h(b)s(e)e -(non-decompressible.)330 2602 y(Y)-8 b(ou'll)31 b(ha)m(v)m(e)g(to)h -(transfer)e(it)h(again.)150 2847 y(If)e(y)m(ou'v)m(e)i(incorp)s(orated) -e Fi(libbzip2)e Fj(in)m(to)k(y)m(our)e(o)m(wn)g(program)h(and)e(are)i -(getting)h(problems,)e(please,)150 2956 y(please,)e(please,)g(c)m(hec)m -(k)f(that)g(the)f(parameters)h(y)m(ou)f(are)h(passing)e(in)h(calls)h -(to)f(the)h(library)-8 b(,)26 b(are)g(correct,)150 3066 -y(and)g(in)g(accordance)i(with)f(what)f(the)h(do)s(cumen)m(tation)h(sa) -m(ys)f(is)f(allo)m(w)m(able.)42 b(I)27 b(ha)m(v)m(e)g(tried)h(to)f(mak) -m(e)h(the)150 3175 y(library)i(robust)g(against)h(suc)m(h)g(problems,)f -(but)g(I'm)g(sure)g(I)g(ha)m(v)m(en't)i(succeeded.)150 -3332 y(Finally)-8 b(,)33 b(if)f(the)h(ab)s(o)m(v)m(e)g(commen)m(ts)g -(don't)f(help,)g(y)m(ou'll)h(ha)m(v)m(e)g(to)g(send)e(me)h(a)g(bug)f -(rep)s(ort.)46 b(No)m(w,)33 b(it's)150 3442 y(just)d(amazing)h(ho)m(w)f -(man)m(y)h(p)s(eople)f(will)h(send)e(me)i(a)f(bug)g(rep)s(ort)h(sa)m -(ying)g(something)g(lik)m(e)481 3593 y(bzip2)f(crashed)g(with)h(segmen) -m(tation)h(fault)f(on)f(m)m(y)g(mac)m(hine)150 3750 y(and)h(absolutely) -h(nothing)g(else.)45 b(Needless)33 b(to)f(sa)m(y)-8 b(,)33 -b(a)f(suc)m(h)f(a)h(rep)s(ort)g(is)f Fb(total)5 b(ly,)36 -b(utterly,)e(c)-5 b(ompletely)150 3859 y(and)27 b(c)-5 -b(ompr)g(ehensively)29 b(100\045)f(useless;)g(a)f(waste)g(of)g(your)g -(time,)h(my)e(time,)i(and)f(net)g(b)-5 b(andwidth)p Fj(.)41 -b(With)150 3969 y(no)30 b(details)i(at)f(all,)g(there's)g(no)f(w)m(a)m -(y)h(I)g(can)f(p)s(ossibly)g(b)s(egin)g(to)h(\014gure)f(out)h(what)f -(the)g(problem)h(is.)150 4126 y(The)e(rules)f(of)h(the)g(game)h(are:)40 -b(facts,)30 b(facts,)g(facts.)41 b(Don't)29 b(omit)h(them)f(b)s(ecause) -f Fi(")p Fj(oh,)h(they)g(w)m(on't)h(b)s(e)150 4235 y(relev)-5 -b(an)m(t)p Fi(")p Fj(.)42 b(A)m(t)31 b(the)g(bare)f(minim)m(um:)481 -4386 y(Mac)m(hine)h(t)m(yp)s(e.)62 b(Op)s(erating)30 -b(system)h(v)m(ersion.)481 4490 y(Exact)g(v)m(ersion)g(of)g -Fi(bzip2)d Fj(\(do)j Fi(bzip2)46 b(-V)p Fj(\).)481 4594 -y(Exact)31 b(v)m(ersion)g(of)g(the)f(compiler)h(used.)481 -4698 y(Flags)g(passed)f(to)h(the)g(compiler.)150 4854 -y(Ho)m(w)m(ev)m(er,)f(the)e(most)g(imp)s(ortan)m(t)f(single)h(thing)g -(that)f(will)h(help)f(me)g(is)g(the)h(\014le)f(that)h(y)m(ou)f(w)m(ere) -h(trying)150 4964 y(to)35 b(compress)f(or)g(decompress)g(at)g(the)h -(time)f(the)h(problem)e(happ)s(ened.)50 b(Without)34 -b(that,)i(m)m(y)e(abilit)m(y)150 5074 y(to)d(do)g(an)m(ything)f(more)h -(than)f(sp)s(eculate)h(ab)s(out)f(the)h(cause,)f(is)h(limited.)150 -5230 y(Please)h(remem)m(b)s(er)g(that)g(I)f(connect)h(to)g(the)g(In)m -(ternet)g(with)f(a)h(mo)s(dem,)f(so)g(y)m(ou)h(should)e(con)m(tact)k -(me)150 5340 y(b)s(efore)c(mailing)h(me)g(h)m(uge)f(\014les.)p -eop -%%Page: 35 36 -35 35 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586 -b(35)150 299 y Fk(4.4)68 b(Did)45 b(y)l(ou)g(get)h(the)f(righ)l(t)h -(pac)l(k)-7 b(age?)150 491 y Fi(bzip2)33 b Fj(is)h(a)h(resource)g(hog.) -54 b(It)35 b(soaks)f(up)g(large)h(amoun)m(ts)g(of)f(CPU)h(cycles)g(and) -f(memory)-8 b(.)53 b(Also,)36 b(it)150 601 y(giv)m(es)25 -b(v)m(ery)e(large)i(latencies.)40 b(In)23 b(the)g(w)m(orst)h(case,)i(y) -m(ou)d(can)h(feed)f(man)m(y)g(megab)m(ytes)i(of)e(uncompressed)150 -711 y(data)42 b(in)m(to)h(the)f(library)g(b)s(efore)f(getting)j(an)m(y) -e(compressed)g(output,)i(so)e(this)g(probably)f(rules)h(out)150 -820 y(applications)31 b(requiring)g(in)m(teractiv)m(e)i(b)s(eha)m -(viour.)150 977 y(These)j(aren't)h(faults)f(of)g(m)m(y)g(implemen)m -(tation,)j(I)d(hop)s(e,)h(but)f(more)g(an)g(in)m(trinsic)g(prop)s(ert)m -(y)g(of)h(the)150 1087 y(Burro)m(ws-Wheeler)31 b(transform)f -(\(unfortunately\).)41 b(Ma)m(yb)s(e)31 b(this)f(isn't)h(what)f(y)m(ou) -h(w)m(an)m(t.)150 1244 y(If)f(y)m(ou)g(w)m(an)m(t)h(a)f(compressor)g -(and/or)g(library)g(whic)m(h)g(is)g(faster,)h(uses)f(less)g(memory)g -(but)g(gets)h(prett)m(y)150 1353 y(go)s(o)s(d)f(compression,)g(and)g -(has)f(minimal)h(latency)-8 b(,)32 b(consider)e(Jean-loup)g(Gailly's)i -(and)d(Mark)h(Adler's)150 1463 y(w)m(ork,)h Fi(zlib-1.1.3)d -Fj(and)h Fi(gzip-1.2.4)p Fj(.)38 b(Lo)s(ok)30 b(for)h(them)f(at)150 -1620 y Fi(http://www.zlib.org)25 b Fj(and)30 b Fi(http://www.gzip.org) -25 b Fj(resp)s(ectiv)m(ely)-8 b(.)150 1776 y(F)g(or)30 -b(something)h(faster)f(and)e(ligh)m(ter)k(still,)f(y)m(ou)f(migh)m(t)g -(try)g(Markus)f(F)h(X)f(J)h(Ob)s(erh)m(umer's)e Fi(LZO)h -Fj(real-)150 1886 y(time)i(compression/decompression)h(library)-8 -b(,)31 b(at)150 1996 y Fi(http://wildsau.idv.uni-l)o(inz.)o(ac.a)o(t/m) -o(fx/l)o(zo.h)o(tml)o Fj(.)150 2152 y(If)37 b(y)m(ou)h(w)m(an)m(t)h(to) -f(use)f(the)h Fi(bzip2)e Fj(algorithms)j(to)f(compress)g(small)g(blo)s -(c)m(ks)f(of)h(data,)i(64k)e(b)m(ytes)h(or)150 2262 y(smaller,)h(for)e -(example)g(on)g(an)g(on-the-\015y)g(disk)f(compressor,)j(y)m(ou'd)e(b)s -(e)f(w)m(ell)i(advised)e(not)i(to)f(use)150 2372 y(this)k(library)-8 -b(.)74 b(Instead,)45 b(I'v)m(e)d(made)g(a)f(sp)s(ecial)h(library)g -(tuned)f(for)g(that)h(kind)f(of)h(use.)74 b(It's)42 b(part)150 -2481 y(of)d Fi(e2compr-0.40)p Fj(,)e(an)i(on-the-\015y)g(disk)f -(compressor)h(for)g(the)g(Lin)m(ux)e Fi(ext2)h Fj(\014lesystem.)66 -b(Lo)s(ok)38 b(at)150 2591 y Fi(http://www.netspace.net.)o(au/~)o(reit) -o(er/)o(e2co)o(mpr)p Fj(.)150 2880 y Fk(4.5)68 b(T)-11 -b(esting)150 3072 y Fj(A)30 b(record)h(of)f(the)h(tests)h(I'v)m(e)f -(done.)150 3229 y(First,)g(some)g(data)g(sets:)225 3386 -y Fh(\017)60 b Fj(B:)29 b(a)g(directory)h(con)m(taining)h(6001)f -(\014les,)g(one)f(for)g(ev)m(ery)h(length)g(in)e(the)i(range)f(0)g(to)h -(6000)h(b)m(ytes.)330 3496 y(The)f(\014les)h(con)m(tain)g(random)f(lo)m -(w)m(ercase)j(letters.)42 b(18.7)32 b(megab)m(ytes.)225 -3633 y Fh(\017)60 b Fj(H:)33 b(m)m(y)h(home)f(directory)h(tree.)50 -b(Do)s(cumen)m(ts,)34 b(source)f(co)s(de,)h(mail)g(\014les,)g -(compressed)f(data.)49 b(H)330 3743 y(con)m(tains)39 -b(B,)e(and)g(also)i(a)e(directory)i(of)f(\014les)g(designed)f(as)h(b)s -(oundary)e(cases)i(for)g(the)g(sorting;)330 3853 y(mostly)31 -b(v)m(ery)g(rep)s(etitiv)m(e,)i(nast)m(y)e(\014les.)40 -b(565)32 b(megab)m(ytes.)225 3990 y Fh(\017)60 b Fj(A:)40 -b(directory)i(tree)g(holding)e(v)-5 b(arious)41 b(applications)g(built) -g(from)f(source:)62 b Fi(egcs)p Fj(,)42 b Fi(gcc-2.8.1)p -Fj(,)330 4100 y(KDE,)30 b(GTK,)h(Octa)m(v)m(e,)j(etc.)42 -b(2200)32 b(megab)m(ytes.)150 4285 y(The)i(tests)h(conducted)f(are)h -(as)f(follo)m(ws.)53 b(Eac)m(h)34 b(test)i(means)d(compressing)i(\(a)f -(cop)m(y)h(of)10 b(\))35 b(eac)m(h)g(\014le)f(in)150 -4394 y(the)d(data)f(set,)i(decompressing)e(it)h(and)f(comparing)h(it)g -(against)g(the)g(original.)150 4551 y(First,)36 b(a)f(bunc)m(h)f(of)h -(tests)g(with)g(blo)s(c)m(k)g(sizes)g(and)f(in)m(ternal)i(bu\013er)e -(sizes)h(set)g(v)m(ery)h(small,)g(to)f(detect)150 4661 -y(an)m(y)h(problems)g(with)h(the)f(blo)s(c)m(king)h(and)f(bu\013ering)g -(mec)m(hanisms.)59 b(This)36 b(required)g(mo)s(difying)g(the)150 -4770 y(source)31 b(co)s(de)f(so)h(as)f(to)i(try)e(to)i(break)e(it.)199 -4927 y(1.)61 b(Data)31 b(set)g(H,)g(with)f(bu\013er)g(size)h(of)f(1)h -(b)m(yte,)h(and)d(blo)s(c)m(k)i(size)g(of)g(23)g(b)m(ytes.)199 -5065 y(2.)61 b(Data)31 b(set)g(B,)f(bu\013er)g(sizes)h(1)g(b)m(yte,)g -(blo)s(c)m(k)g(size)g(1)g(b)m(yte.)199 5202 y(3.)61 b(As)30 -b(\(2\))h(but)f(small-mo)s(de)h(decompression.)199 5340 -y(4.)61 b(As)30 b(\(2\))h(with)g(blo)s(c)m(k)f(size)i(2)e(b)m(ytes.)p -eop -%%Page: 36 37 -36 36 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586 -b(36)199 299 y Fj(5.)61 b(As)30 b(\(2\))h(with)g(blo)s(c)m(k)f(size)i -(3)e(b)m(ytes.)199 431 y(6.)61 b(As)30 b(\(2\))h(with)g(blo)s(c)m(k)f -(size)i(4)e(b)m(ytes.)199 564 y(7.)61 b(As)30 b(\(2\))h(with)g(blo)s(c) -m(k)f(size)i(5)e(b)m(ytes.)199 697 y(8.)61 b(As)30 b(\(2\))h(with)g -(blo)s(c)m(k)f(size)i(6)e(b)m(ytes)h(and)f(small-mo)s(de)h -(decompression.)199 829 y(9.)61 b(H)31 b(with)f(bu\013er)f(size)j(of)e -(1)h(b)m(yte,)g(but)f(normal)g(blo)s(c)m(k)h(size)h(\(up)d(to)j(900000) -h(b)m(ytes\).)150 1009 y(Then)d(some)h(tests)g(with)f(unmo)s(di\014ed)f -(source)h(co)s(de.)199 1166 y(1.)61 b(H,)31 b(all)g(settings)g(normal.) -199 1299 y(2.)61 b(As)30 b(\(1\),)i(with)e(small-mo)s(de)g(decompress.) -199 1431 y(3.)61 b(H,)31 b(compress)f(with)g(\015ag)h -Fi(-1)p Fj(.)199 1564 y(4.)61 b(H,)31 b(compress)f(with)g(\015ag)h -Fi(-s)p Fj(,)f(decompress)g(with)g(\015ag)h Fi(-s)p Fj(.)199 -1697 y(5.)61 b(F)-8 b(orw)m(ards)31 b(compatibilit)m(y:)43 -b(H,)31 b Fi(bzip2-0.1pl2)d Fj(compressing,)j Fi(bzip2-0.9.5)c -Fj(decompressing,)330 1806 y(all)k(settings)h(normal.)199 -1939 y(6.)61 b(Bac)m(kw)m(ards)38 b(compatibilit)m(y:)59 -b(H,)39 b Fi(bzip2-0.9.5)c Fj(compressing,)41 b Fi(bzip2-0.1pl2)35 -b Fj(decompress-)330 2048 y(ing,)c(all)g(settings)g(normal.)199 -2181 y(7.)61 b(Bigger)31 b(tests:)42 b(A,)30 b(all)h(settings)h -(normal.)199 2314 y(8.)61 b(As)30 b(\(7\),)i(using)d(the)i(fallbac)m(k) -g(\(Sadak)-5 b(ane-lik)m(e\))33 b(sorting)e(algorithm.)199 -2446 y(9.)61 b(As)30 b(\(8\),)i(compress)e(with)g(\015ag)h -Fi(-1)p Fj(,)f(decompress)g(with)g(\015ag)h Fi(-s)p Fj(.)154 -2579 y(10.)61 b(H,)31 b(using)f(the)g(fallbac)m(k)i(sorting)f -(algorithm.)154 2711 y(11.)61 b(F)-8 b(orw)m(ards)31 -b(compatibilit)m(y:)44 b(A,)31 b Fi(bzip2-0.1pl2)d Fj(compressing,)k -Fi(bzip2-0.9.5)27 b Fj(decompressing,)330 2821 y(all)k(settings)h -(normal.)154 2954 y(12.)61 b(Bac)m(kw)m(ards)39 b(compatibilit)m(y:)59 -b(A,)39 b Fi(bzip2-0.9.5)c Fj(compressing,)41 b Fi(bzip2-0.1pl2)36 -b Fj(decompress-)330 3063 y(ing,)31 b(all)g(settings)g(normal.)154 -3196 y(13.)61 b(Misc)38 b(test:)57 b(ab)s(out)37 b(400)i(megab)m(ytes)g -(of)f Fi(.tar)f Fj(\014les)g(with)h Fi(bzip2)e Fj(compiled)i(with)g -(Chec)m(k)m(er)g(\(a)330 3305 y(memory)31 b(access)g(error)g(detector,) -h(lik)m(e)g(Purify\).)154 3438 y(14.)61 b(Misc)31 b(tests)h(to)f(mak)m -(e)g(sure)f(it)h(builds)e(and)h(runs)f(ok)h(on)h(non-Lin)m(ux/x86)f -(platforms.)150 3618 y(These)k(tests)g(w)m(ere)h(conducted)e(on)h(a)f -(225)i(MHz)f(IDT)g(WinChip)e(mac)m(hine,)j(running)d(Lin)m(ux)g -(2.0.36.)150 3728 y(They)c(represen)m(t)h(nearly)g(a)f(w)m(eek)h(of)f -(con)m(tin)m(uous)h(computation.)41 b(All)29 b(tests)g(completed)g -(successfully)-8 b(.)150 4003 y Fk(4.6)68 b(F)-11 b(urther)44 -b(reading)150 4196 y Fi(bzip2)26 b Fj(is)h(not)h(researc)m(h)g(w)m -(ork,)g(in)f(the)h(sense)f(that)h(it)g(do)s(esn't)f(presen)m(t)h(an)m -(y)f(new)g(ideas.)40 b(Rather,)28 b(it's)150 4306 y(an)i(engineering)h -(exercise)h(based)e(on)g(existing)i(ideas.)150 4463 y(F)-8 -b(our)31 b(do)s(cumen)m(ts)f(describ)s(e)g(essen)m(tially)i(all)f(the)g -(ideas)f(b)s(ehind)f Fi(bzip2)p Fj(:)390 4614 y Fi(Michael)46 -b(Burrows)g(and)h(D.)g(J.)g(Wheeler:)485 4717 y("A)h(block-sorting)c -(lossless)h(data)i(compression)e(algorithm")533 4821 -y(10th)i(May)g(1994.)533 4925 y(Digital)f(SRC)h(Research)e(Report)i -(124.)533 5029 y(ftp://ftp.digital.com/pub)o(/DEC)o(/SR)o(C/re)o(sear)o -(ch-)o(repo)o(rts/)o(SRC)o(-124)o(.ps.)o(gz)533 5132 -y(If)g(you)g(have)g(trouble)f(finding)g(it,)g(try)h(searching)f(at)h -(the)533 5236 y(New)g(Zealand)f(Digital)g(Library,)f -(http://www.nzdl.org.)p eop -%%Page: 37 38 -37 37 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586 -b(37)390 299 y Fi(Daniel)46 b(S.)h(Hirschberg)e(and)i(Debra)g(A.)g -(LeLewer)485 403 y("Efficient)e(Decoding)h(of)h(Prefix)f(Codes")533 -506 y(Communications)e(of)j(the)g(ACM,)g(April)f(1990,)h(Vol)f(33,)h -(Number)f(4.)533 610 y(You)h(might)f(be)i(able)e(to)h(get)g(an)h -(electronic)d(copy)h(of)h(this)676 714 y(from)g(the)g(ACM)g(Digital)f -(Library.)390 922 y(David)g(J.)i(Wheeler)533 1025 y(Program)e(bred3.c)g -(and)h(accompanying)d(document)i(bred3.ps.)533 1129 y(This)h(contains)e -(the)i(idea)g(behind)f(the)h(multi-table)e(Huffman)533 -1233 y(coding)h(scheme.)533 1337 y(ftp://ftp.cl.cam.ac.uk/us)o(ers/)o -(djw)o(3/)390 1544 y(Jon)h(L.)g(Bentley)f(and)h(Robert)f(Sedgewick)485 -1648 y("Fast)h(Algorithms)e(for)i(Sorting)f(and)g(Searching)g(Strings") -533 1752 y(Available)f(from)i(Sedgewick's)e(web)i(page,)533 -1856 y(www.cs.princeton.edu/~rs)150 2012 y Fj(The)29 -b(follo)m(wing)h(pap)s(er)d(giv)m(es)j(v)-5 b(aluable)29 -b(additional)g(insigh)m(ts)g(in)m(to)h(the)f(algorithm,)h(but)e(is)h -(not)g(imme-)150 2122 y(diately)i(the)g(basis)f(of)h(an)m(y)f(co)s(de)h -(used)e(in)h(bzip2.)390 2273 y Fi(Peter)46 b(Fenwick:)533 -2377 y(Block)h(Sorting)e(Text)i(Compression)533 2481 -y(Proceedings)e(of)i(the)g(19th)g(Australasian)d(Computer)i(Science)f -(Conference,)629 2584 y(Melbourne,)g(Australia.)92 b(Jan)47 -b(31)g(-)h(Feb)f(2,)g(1996.)533 2688 y(ftp://ftp.cs.auckland.ac.)o -(nz/p)o(ub/)o(pete)o(r-f/)o(ACS)o(C96p)o(aper)o(.ps)150 -2845 y Fj(Kunihik)m(o)30 b(Sadak)-5 b(ane's)30 b(sorting)i(algorithm,)f -(men)m(tioned)g(ab)s(o)m(v)m(e,)h(is)e(a)m(v)-5 b(ailable)32 -b(from:)390 2996 y Fi(http://naomi.is.s.u-toky)o(o.ac)o(.jp/)o(~sa)o -(da/p)o(aper)o(s/S)o(ada9)o(8b.p)o(s.g)o(z)150 3153 y -Fj(The)38 b(Man)m(b)s(er-My)m(ers)h(su\016x)e(arra)m(y)i(construction)g -(algorithm)h(is)e(describ)s(ed)f(in)h(a)g(pap)s(er)f(a)m(v)-5 -b(ailable)150 3262 y(from:)390 3413 y Fi(http://www.cs.arizona.ed)o -(u/pe)o(ople)o(/ge)o(ne/P)o(APER)o(S/s)o(uffi)o(x.ps)150 -3570 y Fj(Finally)d(,)33 b(the)e(follo)m(wing)i(pap)s(er)d(do)s(cumen)m -(ts)h(some)g(recen)m(t)i(in)m(v)m(estigations)h(I)d(made)g(in)m(to)i -(the)e(p)s(erfor-)150 3680 y(mance)g(of)f(sorting)h(algorithms:)390 -3831 y Fi(Julian)46 b(Seward:)533 3935 y(On)h(the)g(Performance)e(of)i -(BWT)g(Sorting)f(Algorithms)533 4038 y(Proceedings)f(of)i(the)g(IEEE)g -(Data)f(Compression)f(Conference)g(2000)629 4142 y(Snowbird,)g(Utah.)94 -b(28-30)46 b(March)h(2000.)p eop -%%Page: -1 39 --1 38 bop 3725 -116 a Fl(i)150 299 y Fg(T)-13 b(able)54 -b(of)g(Con)l(ten)l(ts)150 641 y Fk(1)135 b(In)l(tro)t(duction)15 -b Fa(.)20 b(.)f(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)h(.)f -(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)60 -b Fk(2)150 911 y(2)135 b(Ho)l(w)45 b(to)h(use)f Fc(bzip2)31 -b Fa(.)19 b(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)g -(.)h(.)f(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)78 b Fk(3)1047 -1048 y Fj(NAME)18 b Fb(.)d(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)48 b Fl(3)1047 -1157 y Fj(SYNOPSIS)18 b Fb(.)c(.)h(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)48 b Fl(3)1047 1267 y Fj(DESCRIPTION)28 -b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)59 -b Fl(3)1047 1377 y Fj(OPTIONS)16 b Fb(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)46 b Fl(4)1047 -1486 y Fj(MEMOR)-8 b(Y)31 b(MANA)m(GEMENT)9 b Fb(.)16 -b(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)39 b Fl(6)1047 1596 y Fj(RECO)m(VERING)30 b(D)m(A)-8 -b(T)g(A)31 b(FR)m(OM)g(D)m(AMA)m(GED)g(FILES)1256 1705 -y Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)57 b Fl(7)1047 1815 y -Fj(PERF)m(ORMANCE)30 b(NOTES)10 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)40 -b Fl(8)1047 1924 y Fj(CA)-10 b(VEA)i(TS)12 b Fb(.)h(.)i(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)42 -b Fl(8)1047 2034 y Fj(A)m(UTHOR)22 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)52 b -Fl(8)150 2276 y Fk(3)135 b(Programming)46 b(with)f Fc(libbzip2)27 -b Fa(.)16 b(.)j(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)72 -b Fk(10)449 2413 y Fj(3.1)92 b(T)-8 b(op-lev)m(el)33 -b(structure)18 b Fb(.)d(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)47 b Fl(10)748 2523 y Fj(3.1.1)93 -b(Lo)m(w-lev)m(el)32 b(summary)20 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)50 b Fl(10)748 2633 y Fj(3.1.2)93 b(High-lev)m(el)32 -b(summary)27 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)57 -b Fl(10)748 2742 y Fj(3.1.3)93 b(Utilit)m(y)32 b(functions)e(summary)10 -b Fb(.)k(.)h(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)39 b Fl(11)449 2852 y Fj(3.2)92 b(Error)30 -b(handling)15 b Fb(.)f(.)h(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)44 b Fl(11)449 -2961 y Fj(3.3)92 b(Lo)m(w-lev)m(el)32 b(in)m(terface)26 -b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)53 b Fl(13)748 3071 y Fj(3.3.1)93 b Fi(BZ2_bzCompressInit) -21 b Fb(.)9 b(.)15 b(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)50 b Fl(13)748 -3181 y Fj(3.3.2)93 b Fi(BZ2_bzCompress)9 b Fb(.)h(.)15 -b(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)38 b Fl(15)748 -3290 y Fj(3.3.3)93 b Fi(BZ2_bzCompressEnd)23 b Fb(.)10 -b(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)52 b Fl(18)748 3400 -y Fj(3.3.4)93 b Fi(BZ2_bzDecompressInit)16 b Fb(.)9 b(.)15 -b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)45 b Fl(18)748 3509 y Fj(3.3.5)93 -b Fi(BZ2_bzDecompress)21 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)55 b Fl(18)748 3619 y Fj(3.3.6)93 b Fi(BZ2_bzDecompressEnd)18 -b Fb(.)10 b(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)48 b Fl(20)449 -3729 y Fj(3.4)92 b(High-lev)m(el)33 b(in)m(terface)9 -b Fb(.)17 b(.)e(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.) -g(.)g(.)g(.)38 b Fl(20)748 3838 y Fj(3.4.1)93 b Fi(BZ2_bzReadOpen)9 -b Fb(.)h(.)15 b(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)38 -b Fl(20)748 3948 y Fj(3.4.2)93 b Fi(BZ2_bzRead)18 b Fb(.)12 -b(.)j(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)47 -b Fl(21)748 4057 y Fj(3.4.3)93 b Fi(BZ2_bzReadGetUnused)18 -b Fb(.)10 b(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)48 b Fl(23)748 -4167 y Fj(3.4.4)93 b Fi(BZ2_bzReadClose)23 b Fb(.)15 -b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)57 b Fl(23)748 -4276 y Fj(3.4.5)93 b Fi(BZ2_bzWriteOpen)23 b Fb(.)15 -b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)57 b Fl(23)748 -4386 y Fj(3.4.6)93 b Fi(BZ2_bzWrite)16 b Fb(.)11 b(.)k(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)45 b Fl(24)748 -4496 y Fj(3.4.7)93 b Fi(BZ2_bzWriteClose)21 b Fb(.)15 -b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)55 b Fl(25)748 -4605 y Fj(3.4.8)93 b(Handling)30 b(em)m(b)s(edded)f(compressed)h(data)h -(streams)9 b Fb(.)15 b(.)g(.)g(.)38 b Fl(25)748 4715 -y Fj(3.4.9)93 b(Standard)29 b(\014le-reading/writing)j(co)s(de)16 -b Fb(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)45 -b Fl(26)449 4824 y Fj(3.5)92 b(Utilit)m(y)32 b(functions)c -Fb(.)15 b(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)h(.)57 b Fl(27)748 4934 y Fj(3.5.1)93 -b Fi(BZ2_bzBuffToBuffCompres)o(s)22 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)57 b Fl(27)748 -5044 y Fj(3.5.2)93 b Fi(BZ2_bzBuffToBuffDecompr)o(ess)17 -b Fb(.)e(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -53 b Fl(28)449 5153 y Fj(3.6)92 b Fi(zlib)29 b Fj(compatibilit)m(y)k -(functions)23 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)52 -b Fl(29)449 5263 y Fj(3.7)92 b(Using)30 b(the)h(library)f(in)g(a)h -Fi(stdio)p Fj(-free)f(en)m(vironmen)m(t)12 b Fb(.)k(.)f(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)41 b Fl(30)p eop -%%Page: -2 40 --2 39 bop 3699 -116 a Fl(ii)748 83 y Fj(3.7.1)93 b(Getting)32 -b(rid)e(of)g Fi(stdio)17 b Fb(.)d(.)h(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)46 -b Fl(30)748 193 y Fj(3.7.2)93 b(Critical)31 b(error)g(handling)18 -b Fb(.)c(.)h(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)47 b Fl(30)449 302 y -Fj(3.8)92 b(Making)31 b(a)f(Windo)m(ws)g(DLL)17 b Fb(.)d(.)h(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)46 b Fl(31)150 -545 y Fk(4)135 b(Miscellanea)11 b Fa(.)21 b(.)f(.)f(.)h(.)f(.)g(.)h(.)f -(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)h(.) -f(.)h(.)f(.)g(.)h(.)56 b Fk(32)449 682 y Fj(4.1)92 b(Limitations)31 -b(of)g(the)g(compressed)f(\014le)g(format)c Fb(.)15 b(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)54 b Fl(32)449 -791 y Fj(4.2)92 b(P)m(ortabilit)m(y)33 b(issues)12 b -Fb(.)j(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)41 b Fl(33)449 901 y Fj(4.3)92 b(Rep)s(orting)31 -b(bugs)24 b Fb(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)54 b Fl(33)449 1010 -y Fj(4.4)92 b(Did)30 b(y)m(ou)g(get)i(the)f(righ)m(t)g(pac)m(k)-5 -b(age?)18 b Fb(.)f(.)e(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)47 b Fl(35)449 -1120 y Fj(4.5)92 b(T)-8 b(esting)16 b Fb(.)h(.)e(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)g(.)g(.)g(.)46 b Fl(35)449 1230 y Fj(4.6)92 b(F)-8 -b(urther)30 b(reading)17 b Fb(.)f(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g -(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.) -g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)46 b -Fl(36)p eop -%%Trailer -end -userdict /end-hook known{end-hook}if -%%EOF diff -Nru bzip2-1.0.2/manual.texi bzip2-1.0.2.new/manual.texi --- bzip2-1.0.2/manual.texi Sat Jan 5 01:52:50 2002 +++ bzip2-1.0.2.new/manual.texi Thu Jan 1 01:00:00 1970 @@ -1,2243 +0,0 @@ -\input texinfo @c -*- Texinfo -*- -@setfilename bzip2.info - -@ignore -This file documents bzip2 version 1.0.2, and associated library -libbzip2, written by Julian Seward (jseward@acm.org). - -Copyright (C) 1996-2002 Julian R Seward - -Permission is granted to make and distribute verbatim copies of -this manual provided the copyright notice and this permission notice -are preserved on all copies. - -Permission is granted to copy and distribute translations of this manual -into another language, under the above conditions for verbatim copies. -@end ignore - -@ifinfo -@format -START-INFO-DIR-ENTRY -* Bzip2: (bzip2). A program and library for data compression. -END-INFO-DIR-ENTRY -@end format - -@end ifinfo - -@iftex -@c @finalout -@settitle bzip2 and libbzip2 -@titlepage -@title bzip2 and libbzip2 -@subtitle a program and library for data compression -@subtitle copyright (C) 1996-2002 Julian Seward -@subtitle version 1.0.2 of 30 December 2001 -@author Julian Seward - -@end titlepage - -@parindent 0mm -@parskip 2mm - -@end iftex -@node Top,,, (dir) - -The following text is the License for this software. You should -find it identical to that contained in the file LICENSE in the -source distribution. - -@bf{------------------ START OF THE LICENSE ------------------} - -This program, @code{bzip2}, -and associated library @code{libbzip2}, are -Copyright (C) 1996-2002 Julian R Seward. All rights reserved. - -Redistribution and use in source and binary forms, with or without -modification, are permitted provided that the following conditions -are met: -@itemize @bullet -@item - Redistributions of source code must retain the above copyright - notice, this list of conditions and the following disclaimer. -@item - The origin of this software must not be misrepresented; you must - not claim that you wrote the original software. If you use this - software in a product, an acknowledgment in the product - documentation would be appreciated but is not required. -@item - Altered source versions must be plainly marked as such, and must - not be misrepresented as being the original software. -@item - The name of the author may not be used to endorse or promote - products derived from this software without specific prior written - permission. -@end itemize -THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS -OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED -WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE -ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY -DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL -DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE -GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS -INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, -WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING -NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS -SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - -Julian Seward, Cambridge, UK. - -@code{jseward@@acm.org} - -@code{bzip2}/@code{libbzip2} version 1.0.2 of 30 December 2001. - -@bf{------------------ END OF THE LICENSE ------------------} - -Web sites: - -@code{http://sources.redhat.com/bzip2} - -@code{http://www.cacheprof.org} - -PATENTS: To the best of my knowledge, @code{bzip2} does not use any patented -algorithms. However, I do not have the resources available to carry out -a full patent search. Therefore I cannot give any guarantee of the -above statement. - - - - - - - -@chapter Introduction - -@code{bzip2} compresses files using the Burrows-Wheeler -block-sorting text compression algorithm, and Huffman coding. -Compression is generally considerably better than that -achieved by more conventional LZ77/LZ78-based compressors, -and approaches the performance of the PPM family of statistical compressors. - -@code{bzip2} is built on top of @code{libbzip2}, a flexible library -for handling compressed data in the @code{bzip2} format. This manual -describes both how to use the program and -how to work with the library interface. Most of the -manual is devoted to this library, not the program, -which is good news if your interest is only in the program. - -Chapter 2 describes how to use @code{bzip2}; this is the only part -you need to read if you just want to know how to operate the program. -Chapter 3 describes the programming interfaces in detail, and -Chapter 4 records some miscellaneous notes which I thought -ought to be recorded somewhere. - - -@chapter How to use @code{bzip2} - -This chapter contains a copy of the @code{bzip2} man page, -and nothing else. - -@quotation - -@unnumberedsubsubsec NAME -@itemize -@item @code{bzip2}, @code{bunzip2} -- a block-sorting file compressor, v1.0.2 -@item @code{bzcat} -- decompresses files to stdout -@item @code{bzip2recover} -- recovers data from damaged bzip2 files -@end itemize - -@unnumberedsubsubsec SYNOPSIS -@itemize -@item @code{bzip2} [ -cdfkqstvzVL123456789 ] [ filenames ... ] -@item @code{bunzip2} [ -fkvsVL ] [ filenames ... ] -@item @code{bzcat} [ -s ] [ filenames ... ] -@item @code{bzip2recover} filename -@end itemize - -@unnumberedsubsubsec DESCRIPTION - -@code{bzip2} compresses files using the Burrows-Wheeler block sorting -text compression algorithm, and Huffman coding. Compression is -generally considerably better than that achieved by more conventional -LZ77/LZ78-based compressors, and approaches the performance of the PPM -family of statistical compressors. - -The command-line options are deliberately very similar to those of GNU -@code{gzip}, but they are not identical. - -@code{bzip2} expects a list of file names to accompany the command-line -flags. Each file is replaced by a compressed version of itself, with -the name @code{original_name.bz2}. Each compressed file has the same -modification date, permissions, and, when possible, ownership as the -corresponding original, so that these properties can be correctly -restored at decompression time. File name handling is naive in the -sense that there is no mechanism for preserving original file names, -permissions, ownerships or dates in filesystems which lack these -concepts, or have serious file name length restrictions, such as MS-DOS. - -@code{bzip2} and @code{bunzip2} will by default not overwrite existing -files. If you want this to happen, specify the @code{-f} flag. - -If no file names are specified, @code{bzip2} compresses from standard -input to standard output. In this case, @code{bzip2} will decline to -write compressed output to a terminal, as this would be entirely -incomprehensible and therefore pointless. - -@code{bunzip2} (or @code{bzip2 -d}) decompresses all -specified files. Files which were not created by @code{bzip2} -will be detected and ignored, and a warning issued. -@code{bzip2} attempts to guess the filename for the decompressed file -from that of the compressed file as follows: -@itemize -@item @code{filename.bz2 } becomes @code{filename} -@item @code{filename.bz } becomes @code{filename} -@item @code{filename.tbz2} becomes @code{filename.tar} -@item @code{filename.tbz } becomes @code{filename.tar} -@item @code{anyothername } becomes @code{anyothername.out} -@end itemize -If the file does not end in one of the recognised endings, -@code{.bz2}, @code{.bz}, -@code{.tbz2} or @code{.tbz}, @code{bzip2} complains that it cannot -guess the name of the original file, and uses the original name -with @code{.out} appended. - -As with compression, supplying no -filenames causes decompression from standard input to standard output. - -@code{bunzip2} will correctly decompress a file which is the -concatenation of two or more compressed files. The result is the -concatenation of the corresponding uncompressed files. Integrity -testing (@code{-t}) of concatenated compressed files is also supported. - -You can also compress or decompress files to the standard output by -giving the @code{-c} flag. Multiple files may be compressed and -decompressed like this. The resulting outputs are fed sequentially to -stdout. Compression of multiple files in this manner generates a stream -containing multiple compressed file representations. Such a stream -can be decompressed correctly only by @code{bzip2} version 0.9.0 or -later. Earlier versions of @code{bzip2} will stop after decompressing -the first file in the stream. - -@code{bzcat} (or @code{bzip2 -dc}) decompresses all specified files to -the standard output. - -@code{bzip2} will read arguments from the environment variables -@code{BZIP2} and @code{BZIP}, in that order, and will process them -before any arguments read from the command line. This gives a -convenient way to supply default arguments. - -Compression is always performed, even if the compressed file is slightly -larger than the original. Files of less than about one hundred bytes -tend to get larger, since the compression mechanism has a constant -overhead in the region of 50 bytes. Random data (including the output -of most file compressors) is coded at about 8.05 bits per byte, giving -an expansion of around 0.5%. - -As a self-check for your protection, @code{bzip2} uses 32-bit CRCs to -make sure that the decompressed version of a file is identical to the -original. This guards against corruption of the compressed data, and -against undetected bugs in @code{bzip2} (hopefully very unlikely). The -chances of data corruption going undetected is microscopic, about one -chance in four billion for each file processed. Be aware, though, that -the check occurs upon decompression, so it can only tell you that -something is wrong. It can't help you recover the original uncompressed -data. You can use @code{bzip2recover} to try to recover data from -damaged files. - -Return values: 0 for a normal exit, 1 for environmental problems (file -not found, invalid flags, I/O errors, &c), 2 to indicate a corrupt -compressed file, 3 for an internal consistency error (eg, bug) which -caused @code{bzip2} to panic. - - -@unnumberedsubsubsec OPTIONS -@table @code -@item -c --stdout -Compress or decompress to standard output. -@item -d --decompress -Force decompression. @code{bzip2}, @code{bunzip2} and @code{bzcat} are -really the same program, and the decision about what actions to take is -done on the basis of which name is used. This flag overrides that -mechanism, and forces bzip2 to decompress. -@item -z --compress -The complement to @code{-d}: forces compression, regardless of the -invokation name. -@item -t --test -Check integrity of the specified file(s), but don't decompress them. -This really performs a trial decompression and throws away the result. -@item -f --force -Force overwrite of output files. Normally, @code{bzip2} will not overwrite -existing output files. Also forces @code{bzip2} to break hard links -to files, which it otherwise wouldn't do. - -@code{bzip2} normally declines to decompress files which don't have the -correct magic header bytes. If forced (@code{-f}), however, it will -pass such files through unmodified. This is how GNU @code{gzip} -behaves. -@item -k --keep -Keep (don't delete) input files during compression -or decompression. -@item -s --small -Reduce memory usage, for compression, decompression and testing. Files -are decompressed and tested using a modified algorithm which only -requires 2.5 bytes per block byte. This means any file can be -decompressed in 2300k of memory, albeit at about half the normal speed. - -During compression, @code{-s} selects a block size of 200k, which limits -memory use to around the same figure, at the expense of your compression -ratio. In short, if your machine is low on memory (8 megabytes or -less), use -s for everything. See MEMORY MANAGEMENT below. -@item -q --quiet -Suppress non-essential warning messages. Messages pertaining to -I/O errors and other critical events will not be suppressed. -@item -v --verbose -Verbose mode -- show the compression ratio for each file processed. -Further @code{-v}'s increase the verbosity level, spewing out lots of -information which is primarily of interest for diagnostic purposes. -@item -L --license -V --version -Display the software version, license terms and conditions. -@item -1 (or --fast) to -9 (or --best) -Set the block size to 100 k, 200 k .. 900 k when compressing. Has no -effect when decompressing. See MEMORY MANAGEMENT below. -The @code{--fast} and @code{--best} aliases are primarily for GNU -@code{gzip} compatibility. In particular, @code{--fast} doesn't make -things significantly faster. And @code{--best} merely selects the -default behaviour. -@item -- -Treats all subsequent arguments as file names, even if they start -with a dash. This is so you can handle files with names beginning -with a dash, for example: @code{bzip2 -- -myfilename}. -@item --repetitive-fast -@item --repetitive-best -These flags are redundant in versions 0.9.5 and above. They provided -some coarse control over the behaviour of the sorting algorithm in -earlier versions, which was sometimes useful. 0.9.5 and above have an -improved algorithm which renders these flags irrelevant. -@end table - - -@unnumberedsubsubsec MEMORY MANAGEMENT - -@code{bzip2} compresses large files in blocks. The block size affects -both the compression ratio achieved, and the amount of memory needed for -compression and decompression. The flags @code{-1} through @code{-9} -specify the block size to be 100,000 bytes through 900,000 bytes (the -default) respectively. At decompression time, the block size used for -compression is read from the header of the compressed file, and -@code{bunzip2} then allocates itself just enough memory to decompress -the file. Since block sizes are stored in compressed files, it follows -that the flags @code{-1} to @code{-9} are irrelevant to and so ignored -during decompression. - -Compression and decompression requirements, in bytes, can be estimated -as: -@example - Compression: 400k + ( 8 x block size ) - - Decompression: 100k + ( 4 x block size ), or - 100k + ( 2.5 x block size ) -@end example -Larger block sizes give rapidly diminishing marginal returns. Most of -the compression comes from the first two or three hundred k of block -size, a fact worth bearing in mind when using @code{bzip2} on small machines. -It is also important to appreciate that the decompression memory -requirement is set at compression time by the choice of block size. - -For files compressed with the default 900k block size, @code{bunzip2} -will require about 3700 kbytes to decompress. To support decompression -of any file on a 4 megabyte machine, @code{bunzip2} has an option to -decompress using approximately half this amount of memory, about 2300 -kbytes. Decompression speed is also halved, so you should use this -option only where necessary. The relevant flag is @code{-s}. - -In general, try and use the largest block size memory constraints allow, -since that maximises the compression achieved. Compression and -decompression speed are virtually unaffected by block size. - -Another significant point applies to files which fit in a single block --- that means most files you'd encounter using a large block size. The -amount of real memory touched is proportional to the size of the file, -since the file is smaller than a block. For example, compressing a file -20,000 bytes long with the flag @code{-9} will cause the compressor to -allocate around 7600k of memory, but only touch 400k + 20000 * 8 = 560 -kbytes of it. Similarly, the decompressor will allocate 3700k but only -touch 100k + 20000 * 4 = 180 kbytes. - -Here is a table which summarises the maximum memory usage for different -block sizes. Also recorded is the total compressed size for 14 files of -the Calgary Text Compression Corpus totalling 3,141,622 bytes. This -column gives some feel for how compression varies with block size. -These figures tend to understate the advantage of larger block sizes for -larger files, since the Corpus is dominated by smaller files. -@example - Compress Decompress Decompress Corpus - Flag usage usage -s usage Size - - -1 1200k 500k 350k 914704 - -2 2000k 900k 600k 877703 - -3 2800k 1300k 850k 860338 - -4 3600k 1700k 1100k 846899 - -5 4400k 2100k 1350k 845160 - -6 5200k 2500k 1600k 838626 - -7 6100k 2900k 1850k 834096 - -8 6800k 3300k 2100k 828642 - -9 7600k 3700k 2350k 828642 -@end example - -@unnumberedsubsubsec RECOVERING DATA FROM DAMAGED FILES - -@code{bzip2} compresses files in blocks, usually 900kbytes long. Each -block is handled independently. If a media or transmission error causes -a multi-block @code{.bz2} file to become damaged, it may be possible to -recover data from the undamaged blocks in the file. - -The compressed representation of each block is delimited by a 48-bit -pattern, which makes it possible to find the block boundaries with -reasonable certainty. Each block also carries its own 32-bit CRC, so -damaged blocks can be distinguished from undamaged ones. - -@code{bzip2recover} is a simple program whose purpose is to search for -blocks in @code{.bz2} files, and write each block out into its own -@code{.bz2} file. You can then use @code{bzip2 -t} to test the -integrity of the resulting files, and decompress those which are -undamaged. - -@code{bzip2recover} -takes a single argument, the name of the damaged file, and writes a -number of files @code{rec00001file.bz2}, @code{rec00002file.bz2}, etc, -containing the extracted blocks. The output filenames are designed so -that the use of wildcards in subsequent processing -- for example, -@code{bzip2 -dc rec*file.bz2 > recovered_data} -- processes the files in -the correct order. - -@code{bzip2recover} should be of most use dealing with large @code{.bz2} -files, as these will contain many blocks. It is clearly futile to use -it on damaged single-block files, since a damaged block cannot be -recovered. If you wish to minimise any potential data loss through -media or transmission errors, you might consider compressing with a -smaller block size. - - -@unnumberedsubsubsec PERFORMANCE NOTES - -The sorting phase of compression gathers together similar strings in the -file. Because of this, files containing very long runs of repeated -symbols, like "aabaabaabaab ..." (repeated several hundred times) may -compress more slowly than normal. Versions 0.9.5 and above fare much -better than previous versions in this respect. The ratio between -worst-case and average-case compression time is in the region of 10:1. -For previous versions, this figure was more like 100:1. You can use the -@code{-vvvv} option to monitor progress in great detail, if you want. - -Decompression speed is unaffected by these phenomena. - -@code{bzip2} usually allocates several megabytes of memory to operate -in, and then charges all over it in a fairly random fashion. This means -that performance, both for compressing and decompressing, is largely -determined by the speed at which your machine can service cache misses. -Because of this, small changes to the code to reduce the miss rate have -been observed to give disproportionately large performance improvements. -I imagine @code{bzip2} will perform best on machines with very large -caches. - - -@unnumberedsubsubsec CAVEATS - -I/O error messages are not as helpful as they could be. @code{bzip2} -tries hard to detect I/O errors and exit cleanly, but the details of -what the problem is sometimes seem rather misleading. - -This manual page pertains to version 1.0.2 of @code{bzip2}. Compressed -data created by this version is entirely forwards and backwards -compatible with the previous public releases, versions 0.1pl2, 0.9.0, -0.9.5, 1.0.0 and 1.0.1, but with the following exception: 0.9.0 and -above can correctly decompress multiple concatenated compressed files. -0.1pl2 cannot do this; it will stop after decompressing just the first -file in the stream. - -@code{bzip2recover} versions prior to this one, 1.0.2, used 32-bit -integers to represent bit positions in compressed files, so it could not -handle compressed files more than 512 megabytes long. Version 1.0.2 and -above uses 64-bit ints on some platforms which support them (GNU -supported targets, and Windows). To establish whether or not -@code{bzip2recover} was built with such a limitation, run it without -arguments. In any event you can build yourself an unlimited version if -you can recompile it with @code{MaybeUInt64} set to be an unsigned -64-bit integer. - - - -@unnumberedsubsubsec AUTHOR -Julian Seward, @code{jseward@@acm.org}. - -@code{http://sources.redhat.com/bzip2} - -The ideas embodied in @code{bzip2} are due to (at least) the following -people: Michael Burrows and David Wheeler (for the block sorting -transformation), David Wheeler (again, for the Huffman coder), Peter -Fenwick (for the structured coding model in the original @code{bzip}, -and many refinements), and Alistair Moffat, Radford Neal and Ian Witten -(for the arithmetic coder in the original @code{bzip}). I am much -indebted for their help, support and advice. See the manual in the -source distribution for pointers to sources of documentation. Christian -von Roques encouraged me to look for faster sorting algorithms, so as to -speed up compression. Bela Lubkin encouraged me to improve the -worst-case compression performance. The @code{bz*} scripts are derived -from those of GNU @code{gzip}. Many people sent patches, helped with -portability problems, lent machines, gave advice and were generally -helpful. - -@end quotation - - - - -@chapter Programming with @code{libbzip2} - -This chapter describes the programming interface to @code{libbzip2}. - -For general background information, particularly about memory -use and performance aspects, you'd be well advised to read Chapter 2 -as well. - -@section Top-level structure - -@code{libbzip2} is a flexible library for compressing and decompressing -data in the @code{bzip2} data format. Although packaged as a single -entity, it helps to regard the library as three separate parts: the low -level interface, and the high level interface, and some utility -functions. - -The structure of @code{libbzip2}'s interfaces is similar to -that of Jean-loup Gailly's and Mark Adler's excellent @code{zlib} -library. - -All externally visible symbols have names beginning @code{BZ2_}. -This is new in version 1.0. The intention is to minimise pollution -of the namespaces of library clients. - -@subsection Low-level summary - -This interface provides services for compressing and decompressing -data in memory. There's no provision for dealing with files, streams -or any other I/O mechanisms, just straight memory-to-memory work. -In fact, this part of the library can be compiled without inclusion -of @code{stdio.h}, which may be helpful for embedded applications. - -The low-level part of the library has no global variables and -is therefore thread-safe. - -Six routines make up the low level interface: -@code{BZ2_bzCompressInit}, @code{BZ2_bzCompress}, and @* @code{BZ2_bzCompressEnd} -for compression, -and a corresponding trio @code{BZ2_bzDecompressInit}, @* @code{BZ2_bzDecompress} -and @code{BZ2_bzDecompressEnd} for decompression. -The @code{*Init} functions allocate -memory for compression/decompression and do other -initialisations, whilst the @code{*End} functions close down operations -and release memory. - -The real work is done by @code{BZ2_bzCompress} and @code{BZ2_bzDecompress}. -These compress and decompress data from a user-supplied input buffer -to a user-supplied output buffer. These buffers can be any size; -arbitrary quantities of data are handled by making repeated calls -to these functions. This is a flexible mechanism allowing a -consumer-pull style of activity, or producer-push, or a mixture of -both. - - - -@subsection High-level summary - -This interface provides some handy wrappers around the low-level -interface to facilitate reading and writing @code{bzip2} format -files (@code{.bz2} files). The routines provide hooks to facilitate -reading files in which the @code{bzip2} data stream is embedded -within some larger-scale file structure, or where there are -multiple @code{bzip2} data streams concatenated end-to-end. - -For reading files, @code{BZ2_bzReadOpen}, @code{BZ2_bzRead}, -@code{BZ2_bzReadClose} and @* @code{BZ2_bzReadGetUnused} are supplied. For -writing files, @code{BZ2_bzWriteOpen}, @code{BZ2_bzWrite} and -@code{BZ2_bzWriteFinish} are available. - -As with the low-level library, no global variables are used -so the library is per se thread-safe. However, if I/O errors -occur whilst reading or writing the underlying compressed files, -you may have to consult @code{errno} to determine the cause of -the error. In that case, you'd need a C library which correctly -supports @code{errno} in a multithreaded environment. - -To make the library a little simpler and more portable, -@code{BZ2_bzReadOpen} and @code{BZ2_bzWriteOpen} require you to pass them file -handles (@code{FILE*}s) which have previously been opened for reading or -writing respectively. That avoids portability problems associated with -file operations and file attributes, whilst not being much of an -imposition on the programmer. - - - -@subsection Utility functions summary -For very simple needs, @code{BZ2_bzBuffToBuffCompress} and -@code{BZ2_bzBuffToBuffDecompress} are provided. These compress -data in memory from one buffer to another buffer in a single -function call. You should assess whether these functions -fulfill your memory-to-memory compression/decompression -requirements before investing effort in understanding the more -general but more complex low-level interface. - -Yoshioka Tsuneo (@code{QWF00133@@niftyserve.or.jp} / -@code{tsuneo-y@@is.aist-nara.ac.jp}) has contributed some functions to -give better @code{zlib} compatibility. These functions are -@code{BZ2_bzopen}, @code{BZ2_bzread}, @code{BZ2_bzwrite}, @code{BZ2_bzflush}, -@code{BZ2_bzclose}, -@code{BZ2_bzerror} and @code{BZ2_bzlibVersion}. You may find these functions -more convenient for simple file reading and writing, than those in the -high-level interface. These functions are not (yet) officially part of -the library, and are minimally documented here. If they break, you -get to keep all the pieces. I hope to document them properly when time -permits. - -Yoshioka also contributed modifications to allow the library to be -built as a Windows DLL. - - -@section Error handling - -The library is designed to recover cleanly in all situations, including -the worst-case situation of decompressing random data. I'm not -100% sure that it can always do this, so you might want to add -a signal handler to catch segmentation violations during decompression -if you are feeling especially paranoid. I would be interested in -hearing more about the robustness of the library to corrupted -compressed data. - -Version 1.0 is much more robust in this respect than -0.9.0 or 0.9.5. Investigations with Checker (a tool for -detecting problems with memory management, similar to Purify) -indicate that, at least for the few files I tested, all single-bit -errors in the decompressed data are caught properly, with no -segmentation faults, no reads of uninitialised data and no -out of range reads or writes. So it's certainly much improved, -although I wouldn't claim it to be totally bombproof. - -The file @code{bzlib.h} contains all definitions needed to use -the library. In particular, you should definitely not include -@code{bzlib_private.h}. - -In @code{bzlib.h}, the various return values are defined. The following -list is not intended as an exhaustive description of the circumstances -in which a given value may be returned -- those descriptions are given -later. Rather, it is intended to convey the rough meaning of each -return value. The first five actions are normal and not intended to -denote an error situation. -@table @code -@item BZ_OK -The requested action was completed successfully. -@item BZ_RUN_OK -@itemx BZ_FLUSH_OK -@itemx BZ_FINISH_OK -In @code{BZ2_bzCompress}, the requested flush/finish/nothing-special action -was completed successfully. -@item BZ_STREAM_END -Compression of data was completed, or the logical stream end was -detected during decompression. -@end table - -The following return values indicate an error of some kind. -@table @code -@item BZ_CONFIG_ERROR -Indicates that the library has been improperly compiled on your -platform -- a major configuration error. Specifically, it means -that @code{sizeof(char)}, @code{sizeof(short)} and @code{sizeof(int)} -are not 1, 2 and 4 respectively, as they should be. Note that the -library should still work properly on 64-bit platforms which follow -the LP64 programming model -- that is, where @code{sizeof(long)} -and @code{sizeof(void*)} are 8. Under LP64, @code{sizeof(int)} is -still 4, so @code{libbzip2}, which doesn't use the @code{long} type, -is OK. -@item BZ_SEQUENCE_ERROR -When using the library, it is important to call the functions in the -correct sequence and with data structures (buffers etc) in the correct -states. @code{libbzip2} checks as much as it can to ensure this is -happening, and returns @code{BZ_SEQUENCE_ERROR} if not. Code which -complies precisely with the function semantics, as detailed below, -should never receive this value; such an event denotes buggy code -which you should investigate. -@item BZ_PARAM_ERROR -Returned when a parameter to a function call is out of range -or otherwise manifestly incorrect. As with @code{BZ_SEQUENCE_ERROR}, -this denotes a bug in the client code. The distinction between -@code{BZ_PARAM_ERROR} and @code{BZ_SEQUENCE_ERROR} is a bit hazy, but still worth -making. -@item BZ_MEM_ERROR -Returned when a request to allocate memory failed. Note that the -quantity of memory needed to decompress a stream cannot be determined -until the stream's header has been read. So @code{BZ2_bzDecompress} and -@code{BZ2_bzRead} may return @code{BZ_MEM_ERROR} even though some of -the compressed data has been read. The same is not true for -compression; once @code{BZ2_bzCompressInit} or @code{BZ2_bzWriteOpen} have -successfully completed, @code{BZ_MEM_ERROR} cannot occur. -@item BZ_DATA_ERROR -Returned when a data integrity error is detected during decompression. -Most importantly, this means when stored and computed CRCs for the -data do not match. This value is also returned upon detection of any -other anomaly in the compressed data. -@item BZ_DATA_ERROR_MAGIC -As a special case of @code{BZ_DATA_ERROR}, it is sometimes useful to -know when the compressed stream does not start with the correct -magic bytes (@code{'B' 'Z' 'h'}). -@item BZ_IO_ERROR -Returned by @code{BZ2_bzRead} and @code{BZ2_bzWrite} when there is an error -reading or writing in the compressed file, and by @code{BZ2_bzReadOpen} -and @code{BZ2_bzWriteOpen} for attempts to use a file for which the -error indicator (viz, @code{ferror(f)}) is set. -On receipt of @code{BZ_IO_ERROR}, the caller should consult -@code{errno} and/or @code{perror} to acquire operating-system -specific information about the problem. -@item BZ_UNEXPECTED_EOF -Returned by @code{BZ2_bzRead} when the compressed file finishes -before the logical end of stream is detected. -@item BZ_OUTBUFF_FULL -Returned by @code{BZ2_bzBuffToBuffCompress} and -@code{BZ2_bzBuffToBuffDecompress} to indicate that the output data -will not fit into the output buffer provided. -@end table - - - -@section Low-level interface - -@subsection @code{BZ2_bzCompressInit} -@example -typedef - struct @{ - char *next_in; - unsigned int avail_in; - unsigned int total_in_lo32; - unsigned int total_in_hi32; - - char *next_out; - unsigned int avail_out; - unsigned int total_out_lo32; - unsigned int total_out_hi32; - - void *state; - - void *(*bzalloc)(void *,int,int); - void (*bzfree)(void *,void *); - void *opaque; - @} - bz_stream; - -int BZ2_bzCompressInit ( bz_stream *strm, - int blockSize100k, - int verbosity, - int workFactor ); - -@end example - -Prepares for compression. The @code{bz_stream} structure -holds all data pertaining to the compression activity. -A @code{bz_stream} structure should be allocated and initialised -prior to the call. -The fields of @code{bz_stream} -comprise the entirety of the user-visible data. @code{state} -is a pointer to the private data structures required for compression. - -Custom memory allocators are supported, via fields @code{bzalloc}, -@code{bzfree}, -and @code{opaque}. The value -@code{opaque} is passed to as the first argument to -all calls to @code{bzalloc} and @code{bzfree}, but is -otherwise ignored by the library. -The call @code{bzalloc ( opaque, n, m )} is expected to return a -pointer @code{p} to -@code{n * m} bytes of memory, and @code{bzfree ( opaque, p )} -should free -that memory. - -If you don't want to use a custom memory allocator, set @code{bzalloc}, -@code{bzfree} and -@code{opaque} to @code{NULL}, -and the library will then use the standard @code{malloc}/@code{free} -routines. - -Before calling @code{BZ2_bzCompressInit}, fields @code{bzalloc}, -@code{bzfree} and @code{opaque} should -be filled appropriately, as just described. Upon return, the internal -state will have been allocated and initialised, and @code{total_in_lo32}, -@code{total_in_hi32}, @code{total_out_lo32} and -@code{total_out_hi32} will have been set to zero. -These four fields are used by the library -to inform the caller of the total amount of data passed into and out of -the library, respectively. You should not try to change them. -As of version 1.0, 64-bit counts are maintained, even on 32-bit -platforms, using the @code{_hi32} fields to store the upper 32 bits -of the count. So, for example, the total amount of data in -is @code{(total_in_hi32 << 32) + total_in_lo32}. - -Parameter @code{blockSize100k} specifies the block size to be used for -compression. It should be a value between 1 and 9 inclusive, and the -actual block size used is 100000 x this figure. 9 gives the best -compression but takes most memory. - -Parameter @code{verbosity} should be set to a number between 0 and 4 -inclusive. 0 is silent, and greater numbers give increasingly verbose -monitoring/debugging output. If the library has been compiled with -@code{-DBZ_NO_STDIO}, no such output will appear for any verbosity -setting. - -Parameter @code{workFactor} controls how the compression phase behaves -when presented with worst case, highly repetitive, input data. If -compression runs into difficulties caused by repetitive data, the -library switches from the standard sorting algorithm to a fallback -algorithm. The fallback is slower than the standard algorithm by -perhaps a factor of three, but always behaves reasonably, no matter how -bad the input. - -Lower values of @code{workFactor} reduce the amount of effort the -standard algorithm will expend before resorting to the fallback. You -should set this parameter carefully; too low, and many inputs will be -handled by the fallback algorithm and so compress rather slowly, too -high, and your average-to-worst case compression times can become very -large. The default value of 30 gives reasonable behaviour over a wide -range of circumstances. - -Allowable values range from 0 to 250 inclusive. 0 is a special case, -equivalent to using the default value of 30. - -Note that the compressed output generated is the same regardless of -whether or not the fallback algorithm is used. - -Be aware also that this parameter may disappear entirely in future -versions of the library. In principle it should be possible to devise a -good way to automatically choose which algorithm to use. Such a -mechanism would render the parameter obsolete. - -Possible return values: -@display - @code{BZ_CONFIG_ERROR} - if the library has been mis-compiled - @code{BZ_PARAM_ERROR} - if @code{strm} is @code{NULL} - or @code{blockSize} < 1 or @code{blockSize} > 9 - or @code{verbosity} < 0 or @code{verbosity} > 4 - or @code{workFactor} < 0 or @code{workFactor} > 250 - @code{BZ_MEM_ERROR} - if not enough memory is available - @code{BZ_OK} - otherwise -@end display -Allowable next actions: -@display - @code{BZ2_bzCompress} - if @code{BZ_OK} is returned - no specific action needed in case of error -@end display - -@subsection @code{BZ2_bzCompress} -@example - int BZ2_bzCompress ( bz_stream *strm, int action ); -@end example -Provides more input and/or output buffer space for the library. The -caller maintains input and output buffers, and calls @code{BZ2_bzCompress} to -transfer data between them. - -Before each call to @code{BZ2_bzCompress}, @code{next_in} should point at -the data to be compressed, and @code{avail_in} should indicate how many -bytes the library may read. @code{BZ2_bzCompress} updates @code{next_in}, -@code{avail_in} and @code{total_in} to reflect the number of bytes it -has read. - -Similarly, @code{next_out} should point to a buffer in which the -compressed data is to be placed, with @code{avail_out} indicating how -much output space is available. @code{BZ2_bzCompress} updates -@code{next_out}, @code{avail_out} and @code{total_out} to reflect the -number of bytes output. - -You may provide and remove as little or as much data as you like on each -call of @code{BZ2_bzCompress}. In the limit, it is acceptable to supply and -remove data one byte at a time, although this would be terribly -inefficient. You should always ensure that at least one byte of output -space is available at each call. - -A second purpose of @code{BZ2_bzCompress} is to request a change of mode of the -compressed stream. - -Conceptually, a compressed stream can be in one of four states: IDLE, -RUNNING, FLUSHING and FINISHING. Before initialisation -(@code{BZ2_bzCompressInit}) and after termination (@code{BZ2_bzCompressEnd}), a -stream is regarded as IDLE. - -Upon initialisation (@code{BZ2_bzCompressInit}), the stream is placed in the -RUNNING state. Subsequent calls to @code{BZ2_bzCompress} should pass -@code{BZ_RUN} as the requested action; other actions are illegal and -will result in @code{BZ_SEQUENCE_ERROR}. - -At some point, the calling program will have provided all the input data -it wants to. It will then want to finish up -- in effect, asking the -library to process any data it might have buffered internally. In this -state, @code{BZ2_bzCompress} will no longer attempt to read data from -@code{next_in}, but it will want to write data to @code{next_out}. -Because the output buffer supplied by the user can be arbitrarily small, -the finishing-up operation cannot necessarily be done with a single call -of @code{BZ2_bzCompress}. - -Instead, the calling program passes @code{BZ_FINISH} as an action to -@code{BZ2_bzCompress}. This changes the stream's state to FINISHING. Any -remaining input (ie, @code{next_in[0 .. avail_in-1]}) is compressed and -transferred to the output buffer. To do this, @code{BZ2_bzCompress} must be -called repeatedly until all the output has been consumed. At that -point, @code{BZ2_bzCompress} returns @code{BZ_STREAM_END}, and the stream's -state is set back to IDLE. @code{BZ2_bzCompressEnd} should then be -called. - -Just to make sure the calling program does not cheat, the library makes -a note of @code{avail_in} at the time of the first call to -@code{BZ2_bzCompress} which has @code{BZ_FINISH} as an action (ie, at the -time the program has announced its intention to not supply any more -input). By comparing this value with that of @code{avail_in} over -subsequent calls to @code{BZ2_bzCompress}, the library can detect any -attempts to slip in more data to compress. Any calls for which this is -detected will return @code{BZ_SEQUENCE_ERROR}. This indicates a -programming mistake which should be corrected. - -Instead of asking to finish, the calling program may ask -@code{BZ2_bzCompress} to take all the remaining input, compress it and -terminate the current (Burrows-Wheeler) compression block. This could -be useful for error control purposes. The mechanism is analogous to -that for finishing: call @code{BZ2_bzCompress} with an action of -@code{BZ_FLUSH}, remove output data, and persist with the -@code{BZ_FLUSH} action until the value @code{BZ_RUN} is returned. As -with finishing, @code{BZ2_bzCompress} detects any attempt to provide more -input data once the flush has begun. - -Once the flush is complete, the stream returns to the normal RUNNING -state. - -This all sounds pretty complex, but isn't really. Here's a table -which shows which actions are allowable in each state, what action -will be taken, what the next state is, and what the non-error return -values are. Note that you can't explicitly ask what state the -stream is in, but nor do you need to -- it can be inferred from the -values returned by @code{BZ2_bzCompress}. -@display -IDLE/@code{any} - Illegal. IDLE state only exists after @code{BZ2_bzCompressEnd} or - before @code{BZ2_bzCompressInit}. - Return value = @code{BZ_SEQUENCE_ERROR} - -RUNNING/@code{BZ_RUN} - Compress from @code{next_in} to @code{next_out} as much as possible. - Next state = RUNNING - Return value = @code{BZ_RUN_OK} - -RUNNING/@code{BZ_FLUSH} - Remember current value of @code{next_in}. Compress from @code{next_in} - to @code{next_out} as much as possible, but do not accept any more input. - Next state = FLUSHING - Return value = @code{BZ_FLUSH_OK} - -RUNNING/@code{BZ_FINISH} - Remember current value of @code{next_in}. Compress from @code{next_in} - to @code{next_out} as much as possible, but do not accept any more input. - Next state = FINISHING - Return value = @code{BZ_FINISH_OK} - -FLUSHING/@code{BZ_FLUSH} - Compress from @code{next_in} to @code{next_out} as much as possible, - but do not accept any more input. - If all the existing input has been used up and all compressed - output has been removed - Next state = RUNNING; Return value = @code{BZ_RUN_OK} - else - Next state = FLUSHING; Return value = @code{BZ_FLUSH_OK} - -FLUSHING/other - Illegal. - Return value = @code{BZ_SEQUENCE_ERROR} - -FINISHING/@code{BZ_FINISH} - Compress from @code{next_in} to @code{next_out} as much as possible, - but to not accept any more input. - If all the existing input has been used up and all compressed - output has been removed - Next state = IDLE; Return value = @code{BZ_STREAM_END} - else - Next state = FINISHING; Return value = @code{BZ_FINISHING} - -FINISHING/other - Illegal. - Return value = @code{BZ_SEQUENCE_ERROR} -@end display - -That still looks complicated? Well, fair enough. The usual sequence -of calls for compressing a load of data is: -@itemize @bullet -@item Get started with @code{BZ2_bzCompressInit}. -@item Shovel data in and shlurp out its compressed form using zero or more -calls of @code{BZ2_bzCompress} with action = @code{BZ_RUN}. -@item Finish up. -Repeatedly call @code{BZ2_bzCompress} with action = @code{BZ_FINISH}, -copying out the compressed output, until @code{BZ_STREAM_END} is returned. -@item Close up and go home. Call @code{BZ2_bzCompressEnd}. -@end itemize -If the data you want to compress fits into your input buffer all -at once, you can skip the calls of @code{BZ2_bzCompress ( ..., BZ_RUN )} and -just do the @code{BZ2_bzCompress ( ..., BZ_FINISH )} calls. - -All required memory is allocated by @code{BZ2_bzCompressInit}. The -compression library can accept any data at all (obviously). So you -shouldn't get any error return values from the @code{BZ2_bzCompress} calls. -If you do, they will be @code{BZ_SEQUENCE_ERROR}, and indicate a bug in -your programming. - -Trivial other possible return values: -@display - @code{BZ_PARAM_ERROR} - if @code{strm} is @code{NULL}, or @code{strm->s} is @code{NULL} -@end display - -@subsection @code{BZ2_bzCompressEnd} -@example -int BZ2_bzCompressEnd ( bz_stream *strm ); -@end example -Releases all memory associated with a compression stream. - -Possible return values: -@display - @code{BZ_PARAM_ERROR} if @code{strm} is @code{NULL} or @code{strm->s} is @code{NULL} - @code{BZ_OK} otherwise -@end display - - -@subsection @code{BZ2_bzDecompressInit} -@example -int BZ2_bzDecompressInit ( bz_stream *strm, int verbosity, int small ); -@end example -Prepares for decompression. As with @code{BZ2_bzCompressInit}, a -@code{bz_stream} record should be allocated and initialised before the -call. Fields @code{bzalloc}, @code{bzfree} and @code{opaque} should be -set if a custom memory allocator is required, or made @code{NULL} for -the normal @code{malloc}/@code{free} routines. Upon return, the internal -state will have been initialised, and @code{total_in} and -@code{total_out} will be zero. - -For the meaning of parameter @code{verbosity}, see @code{BZ2_bzCompressInit}. - -If @code{small} is nonzero, the library will use an alternative -decompression algorithm which uses less memory but at the cost of -decompressing more slowly (roughly speaking, half the speed, but the -maximum memory requirement drops to around 2300k). See Chapter 2 for -more information on memory management. - -Note that the amount of memory needed to decompress -a stream cannot be determined until the stream's header has been read, -so even if @code{BZ2_bzDecompressInit} succeeds, a subsequent -@code{BZ2_bzDecompress} could fail with @code{BZ_MEM_ERROR}. - -Possible return values: -@display - @code{BZ_CONFIG_ERROR} - if the library has been mis-compiled - @code{BZ_PARAM_ERROR} - if @code{(small != 0 && small != 1)} - or @code{(verbosity < 0 || verbosity > 4)} - @code{BZ_MEM_ERROR} - if insufficient memory is available -@end display - -Allowable next actions: -@display - @code{BZ2_bzDecompress} - if @code{BZ_OK} was returned - no specific action required in case of error -@end display - - - -@subsection @code{BZ2_bzDecompress} -@example -int BZ2_bzDecompress ( bz_stream *strm ); -@end example -Provides more input and/out output buffer space for the library. The -caller maintains input and output buffers, and uses @code{BZ2_bzDecompress} -to transfer data between them. - -Before each call to @code{BZ2_bzDecompress}, @code{next_in} -should point at the compressed data, -and @code{avail_in} should indicate how many bytes the library -may read. @code{BZ2_bzDecompress} updates @code{next_in}, @code{avail_in} -and @code{total_in} -to reflect the number of bytes it has read. - -Similarly, @code{next_out} should point to a buffer in which the uncompressed -output is to be placed, with @code{avail_out} indicating how much output space -is available. @code{BZ2_bzCompress} updates @code{next_out}, -@code{avail_out} and @code{total_out} to reflect -the number of bytes output. - -You may provide and remove as little or as much data as you like on -each call of @code{BZ2_bzDecompress}. -In the limit, it is acceptable to -supply and remove data one byte at a time, although this would be -terribly inefficient. You should always ensure that at least one -byte of output space is available at each call. - -Use of @code{BZ2_bzDecompress} is simpler than @code{BZ2_bzCompress}. - -You should provide input and remove output as described above, and -repeatedly call @code{BZ2_bzDecompress} until @code{BZ_STREAM_END} is -returned. Appearance of @code{BZ_STREAM_END} denotes that -@code{BZ2_bzDecompress} has detected the logical end of the compressed -stream. @code{BZ2_bzDecompress} will not produce @code{BZ_STREAM_END} until -all output data has been placed into the output buffer, so once -@code{BZ_STREAM_END} appears, you are guaranteed to have available all -the decompressed output, and @code{BZ2_bzDecompressEnd} can safely be -called. - -If case of an error return value, you should call @code{BZ2_bzDecompressEnd} -to clean up and release memory. - -Possible return values: -@display - @code{BZ_PARAM_ERROR} - if @code{strm} is @code{NULL} or @code{strm->s} is @code{NULL} - or @code{strm->avail_out < 1} - @code{BZ_DATA_ERROR} - if a data integrity error is detected in the compressed stream - @code{BZ_DATA_ERROR_MAGIC} - if the compressed stream doesn't begin with the right magic bytes - @code{BZ_MEM_ERROR} - if there wasn't enough memory available - @code{BZ_STREAM_END} - if the logical end of the data stream was detected and all - output in has been consumed, eg @code{s->avail_out > 0} - @code{BZ_OK} - otherwise -@end display -Allowable next actions: -@display - @code{BZ2_bzDecompress} - if @code{BZ_OK} was returned - @code{BZ2_bzDecompressEnd} - otherwise -@end display - - -@subsection @code{BZ2_bzDecompressEnd} -@example -int BZ2_bzDecompressEnd ( bz_stream *strm ); -@end example -Releases all memory associated with a decompression stream. - -Possible return values: -@display - @code{BZ_PARAM_ERROR} - if @code{strm} is @code{NULL} or @code{strm->s} is @code{NULL} - @code{BZ_OK} - otherwise -@end display - -Allowable next actions: -@display - None. -@end display - - -@section High-level interface - -This interface provides functions for reading and writing -@code{bzip2} format files. First, some general points. - -@itemize @bullet -@item All of the functions take an @code{int*} first argument, - @code{bzerror}. - After each call, @code{bzerror} should be consulted first to determine - the outcome of the call. If @code{bzerror} is @code{BZ_OK}, - the call completed - successfully, and only then should the return value of the function - (if any) be consulted. If @code{bzerror} is @code{BZ_IO_ERROR}, - there was an error - reading/writing the underlying compressed file, and you should - then consult @code{errno}/@code{perror} to determine the - cause of the difficulty. - @code{bzerror} may also be set to various other values; precise details are - given on a per-function basis below. -@item If @code{bzerror} indicates an error - (ie, anything except @code{BZ_OK} and @code{BZ_STREAM_END}), - you should immediately call @code{BZ2_bzReadClose} (or @code{BZ2_bzWriteClose}, - depending on whether you are attempting to read or to write) - to free up all resources associated - with the stream. Once an error has been indicated, behaviour of all calls - except @code{BZ2_bzReadClose} (@code{BZ2_bzWriteClose}) is undefined. - The implication is that (1) @code{bzerror} should - be checked after each call, and (2) if @code{bzerror} indicates an error, - @code{BZ2_bzReadClose} (@code{BZ2_bzWriteClose}) should then be called to clean up. -@item The @code{FILE*} arguments passed to - @code{BZ2_bzReadOpen}/@code{BZ2_bzWriteOpen} - should be set to binary mode. - Most Unix systems will do this by default, but other platforms, - including Windows and Mac, will not. If you omit this, you may - encounter problems when moving code to new platforms. -@item Memory allocation requests are handled by - @code{malloc}/@code{free}. - At present - there is no facility for user-defined memory allocators in the file I/O - functions (could easily be added, though). -@end itemize - - - -@subsection @code{BZ2_bzReadOpen} -@example - typedef void BZFILE; - - BZFILE *BZ2_bzReadOpen ( int *bzerror, FILE *f, - int small, int verbosity, - void *unused, int nUnused ); -@end example -Prepare to read compressed data from file handle @code{f}. @code{f} -should refer to a file which has been opened for reading, and for which -the error indicator (@code{ferror(f)})is not set. If @code{small} is 1, -the library will try to decompress using less memory, at the expense of -speed. - -For reasons explained below, @code{BZ2_bzRead} will decompress the -@code{nUnused} bytes starting at @code{unused}, before starting to read -from the file @code{f}. At most @code{BZ_MAX_UNUSED} bytes may be -supplied like this. If this facility is not required, you should pass -@code{NULL} and @code{0} for @code{unused} and n@code{Unused} -respectively. - -For the meaning of parameters @code{small} and @code{verbosity}, -see @code{BZ2_bzDecompressInit}. - -The amount of memory needed to decompress a file cannot be determined -until the file's header has been read. So it is possible that -@code{BZ2_bzReadOpen} returns @code{BZ_OK} but a subsequent call of -@code{BZ2_bzRead} will return @code{BZ_MEM_ERROR}. - -Possible assignments to @code{bzerror}: -@display - @code{BZ_CONFIG_ERROR} - if the library has been mis-compiled - @code{BZ_PARAM_ERROR} - if @code{f} is @code{NULL} - or @code{small} is neither @code{0} nor @code{1} - or @code{(unused == NULL && nUnused != 0)} - or @code{(unused != NULL && !(0 <= nUnused <= BZ_MAX_UNUSED))} - @code{BZ_IO_ERROR} - if @code{ferror(f)} is nonzero - @code{BZ_MEM_ERROR} - if insufficient memory is available - @code{BZ_OK} - otherwise. -@end display - -Possible return values: -@display - Pointer to an abstract @code{BZFILE} - if @code{bzerror} is @code{BZ_OK} - @code{NULL} - otherwise -@end display - -Allowable next actions: -@display - @code{BZ2_bzRead} - if @code{bzerror} is @code{BZ_OK} - @code{BZ2_bzClose} - otherwise -@end display - - -@subsection @code{BZ2_bzRead} -@example - int BZ2_bzRead ( int *bzerror, BZFILE *b, void *buf, int len ); -@end example -Reads up to @code{len} (uncompressed) bytes from the compressed file -@code{b} into -the buffer @code{buf}. If the read was successful, -@code{bzerror} is set to @code{BZ_OK} -and the number of bytes read is returned. If the logical end-of-stream -was detected, @code{bzerror} will be set to @code{BZ_STREAM_END}, -and the number -of bytes read is returned. All other @code{bzerror} values denote an error. - -@code{BZ2_bzRead} will supply @code{len} bytes, -unless the logical stream end is detected -or an error occurs. Because of this, it is possible to detect the -stream end by observing when the number of bytes returned is -less than the number -requested. Nevertheless, this is regarded as inadvisable; you should -instead check @code{bzerror} after every call and watch out for -@code{BZ_STREAM_END}. - -Internally, @code{BZ2_bzRead} copies data from the compressed file in chunks -of size @code{BZ_MAX_UNUSED} bytes -before decompressing it. If the file contains more bytes than strictly -needed to reach the logical end-of-stream, @code{BZ2_bzRead} will almost certainly -read some of the trailing data before signalling @code{BZ_SEQUENCE_END}. -To collect the read but unused data once @code{BZ_SEQUENCE_END} has -appeared, call @code{BZ2_bzReadGetUnused} immediately before @code{BZ2_bzReadClose}. - -Possible assignments to @code{bzerror}: -@display - @code{BZ_PARAM_ERROR} - if @code{b} is @code{NULL} or @code{buf} is @code{NULL} or @code{len < 0} - @code{BZ_SEQUENCE_ERROR} - if @code{b} was opened with @code{BZ2_bzWriteOpen} - @code{BZ_IO_ERROR} - if there is an error reading from the compressed file - @code{BZ_UNEXPECTED_EOF} - if the compressed file ended before the logical end-of-stream was detected - @code{BZ_DATA_ERROR} - if a data integrity error was detected in the compressed stream - @code{BZ_DATA_ERROR_MAGIC} - if the stream does not begin with the requisite header bytes (ie, is not - a @code{bzip2} data file). This is really a special case of @code{BZ_DATA_ERROR}. - @code{BZ_MEM_ERROR} - if insufficient memory was available - @code{BZ_STREAM_END} - if the logical end of stream was detected. - @code{BZ_OK} - otherwise. -@end display - -Possible return values: -@display - number of bytes read - if @code{bzerror} is @code{BZ_OK} or @code{BZ_STREAM_END} - undefined - otherwise -@end display - -Allowable next actions: -@display - collect data from @code{buf}, then @code{BZ2_bzRead} or @code{BZ2_bzReadClose} - if @code{bzerror} is @code{BZ_OK} - collect data from @code{buf}, then @code{BZ2_bzReadClose} or @code{BZ2_bzReadGetUnused} - if @code{bzerror} is @code{BZ_SEQUENCE_END} - @code{BZ2_bzReadClose} - otherwise -@end display - - - -@subsection @code{BZ2_bzReadGetUnused} -@example - void BZ2_bzReadGetUnused ( int* bzerror, BZFILE *b, - void** unused, int* nUnused ); -@end example -Returns data which was read from the compressed file but was not needed -to get to the logical end-of-stream. @code{*unused} is set to the address -of the data, and @code{*nUnused} to the number of bytes. @code{*nUnused} will -be set to a value between @code{0} and @code{BZ_MAX_UNUSED} inclusive. - -This function may only be called once @code{BZ2_bzRead} has signalled -@code{BZ_STREAM_END} but before @code{BZ2_bzReadClose}. - -Possible assignments to @code{bzerror}: -@display - @code{BZ_PARAM_ERROR} - if @code{b} is @code{NULL} - or @code{unused} is @code{NULL} or @code{nUnused} is @code{NULL} - @code{BZ_SEQUENCE_ERROR} - if @code{BZ_STREAM_END} has not been signalled - or if @code{b} was opened with @code{BZ2_bzWriteOpen} - @code{BZ_OK} - otherwise -@end display - -Allowable next actions: -@display - @code{BZ2_bzReadClose} -@end display - - -@subsection @code{BZ2_bzReadClose} -@example - void BZ2_bzReadClose ( int *bzerror, BZFILE *b ); -@end example -Releases all memory pertaining to the compressed file @code{b}. -@code{BZ2_bzReadClose} does not call @code{fclose} on the underlying file -handle, so you should do that yourself if appropriate. -@code{BZ2_bzReadClose} should be called to clean up after all error -situations. - -Possible assignments to @code{bzerror}: -@display - @code{BZ_SEQUENCE_ERROR} - if @code{b} was opened with @code{BZ2_bzOpenWrite} - @code{BZ_OK} - otherwise -@end display - -Allowable next actions: -@display - none -@end display - - - -@subsection @code{BZ2_bzWriteOpen} -@example - BZFILE *BZ2_bzWriteOpen ( int *bzerror, FILE *f, - int blockSize100k, int verbosity, - int workFactor ); -@end example -Prepare to write compressed data to file handle @code{f}. -@code{f} should refer to -a file which has been opened for writing, and for which the error -indicator (@code{ferror(f)})is not set. - -For the meaning of parameters @code{blockSize100k}, -@code{verbosity} and @code{workFactor}, see -@* @code{BZ2_bzCompressInit}. - -All required memory is allocated at this stage, so if the call -completes successfully, @code{BZ_MEM_ERROR} cannot be signalled by a -subsequent call to @code{BZ2_bzWrite}. - -Possible assignments to @code{bzerror}: -@display - @code{BZ_CONFIG_ERROR} - if the library has been mis-compiled - @code{BZ_PARAM_ERROR} - if @code{f} is @code{NULL} - or @code{blockSize100k < 1} or @code{blockSize100k > 9} - @code{BZ_IO_ERROR} - if @code{ferror(f)} is nonzero - @code{BZ_MEM_ERROR} - if insufficient memory is available - @code{BZ_OK} - otherwise -@end display - -Possible return values: -@display - Pointer to an abstract @code{BZFILE} - if @code{bzerror} is @code{BZ_OK} - @code{NULL} - otherwise -@end display - -Allowable next actions: -@display - @code{BZ2_bzWrite} - if @code{bzerror} is @code{BZ_OK} - (you could go directly to @code{BZ2_bzWriteClose}, but this would be pretty pointless) - @code{BZ2_bzWriteClose} - otherwise -@end display - - - -@subsection @code{BZ2_bzWrite} -@example - void BZ2_bzWrite ( int *bzerror, BZFILE *b, void *buf, int len ); -@end example -Absorbs @code{len} bytes from the buffer @code{buf}, eventually to be -compressed and written to the file. - -Possible assignments to @code{bzerror}: -@display - @code{BZ_PARAM_ERROR} - if @code{b} is @code{NULL} or @code{buf} is @code{NULL} or @code{len < 0} - @code{BZ_SEQUENCE_ERROR} - if b was opened with @code{BZ2_bzReadOpen} - @code{BZ_IO_ERROR} - if there is an error writing the compressed file. - @code{BZ_OK} - otherwise -@end display - - - - -@subsection @code{BZ2_bzWriteClose} -@example - void BZ2_bzWriteClose ( int *bzerror, BZFILE* f, - int abandon, - unsigned int* nbytes_in, - unsigned int* nbytes_out ); - - void BZ2_bzWriteClose64 ( int *bzerror, BZFILE* f, - int abandon, - unsigned int* nbytes_in_lo32, - unsigned int* nbytes_in_hi32, - unsigned int* nbytes_out_lo32, - unsigned int* nbytes_out_hi32 ); -@end example - -Compresses and flushes to the compressed file all data so far supplied -by @code{BZ2_bzWrite}. The logical end-of-stream markers are also written, so -subsequent calls to @code{BZ2_bzWrite} are illegal. All memory associated -with the compressed file @code{b} is released. -@code{fflush} is called on the -compressed file, but it is not @code{fclose}'d. - -If @code{BZ2_bzWriteClose} is called to clean up after an error, the only -action is to release the memory. The library records the error codes -issued by previous calls, so this situation will be detected -automatically. There is no attempt to complete the compression -operation, nor to @code{fflush} the compressed file. You can force this -behaviour to happen even in the case of no error, by passing a nonzero -value to @code{abandon}. - -If @code{nbytes_in} is non-null, @code{*nbytes_in} will be set to be the -total volume of uncompressed data handled. Similarly, @code{nbytes_out} -will be set to the total volume of compressed data written. For -compatibility with older versions of the library, @code{BZ2_bzWriteClose} -only yields the lower 32 bits of these counts. Use -@code{BZ2_bzWriteClose64} if you want the full 64 bit counts. These -two functions are otherwise absolutely identical. - - -Possible assignments to @code{bzerror}: -@display - @code{BZ_SEQUENCE_ERROR} - if @code{b} was opened with @code{BZ2_bzReadOpen} - @code{BZ_IO_ERROR} - if there is an error writing the compressed file - @code{BZ_OK} - otherwise -@end display - -@subsection Handling embedded compressed data streams - -The high-level library facilitates use of -@code{bzip2} data streams which form some part of a surrounding, larger -data stream. -@itemize @bullet -@item For writing, the library takes an open file handle, writes -compressed data to it, @code{fflush}es it but does not @code{fclose} it. -The calling application can write its own data before and after the -compressed data stream, using that same file handle. -@item Reading is more complex, and the facilities are not as general -as they could be since generality is hard to reconcile with efficiency. -@code{BZ2_bzRead} reads from the compressed file in blocks of size -@code{BZ_MAX_UNUSED} bytes, and in doing so probably will overshoot -the logical end of compressed stream. -To recover this data once decompression has -ended, call @code{BZ2_bzReadGetUnused} after the last call of @code{BZ2_bzRead} -(the one returning @code{BZ_STREAM_END}) but before calling -@code{BZ2_bzReadClose}. -@end itemize - -This mechanism makes it easy to decompress multiple @code{bzip2} -streams placed end-to-end. As the end of one stream, when @code{BZ2_bzRead} -returns @code{BZ_STREAM_END}, call @code{BZ2_bzReadGetUnused} to collect the -unused data (copy it into your own buffer somewhere). -That data forms the start of the next compressed stream. -To start uncompressing that next stream, call @code{BZ2_bzReadOpen} again, -feeding in the unused data via the @code{unused}/@code{nUnused} -parameters. -Keep doing this until @code{BZ_STREAM_END} return coincides with the -physical end of file (@code{feof(f)}). In this situation -@code{BZ2_bzReadGetUnused} -will of course return no data. - -This should give some feel for how the high-level interface can be used. -If you require extra flexibility, you'll have to bite the bullet and get -to grips with the low-level interface. - -@subsection Standard file-reading/writing code -Here's how you'd write data to a compressed file: -@example @code -FILE* f; -BZFILE* b; -int nBuf; -char buf[ /* whatever size you like */ ]; -int bzerror; -int nWritten; - -f = fopen ( "myfile.bz2", "w" ); -if (!f) @{ - /* handle error */ -@} -b = BZ2_bzWriteOpen ( &bzerror, f, 9 ); -if (bzerror != BZ_OK) @{ - BZ2_bzWriteClose ( b ); - /* handle error */ -@} - -while ( /* condition */ ) @{ - /* get data to write into buf, and set nBuf appropriately */ - nWritten = BZ2_bzWrite ( &bzerror, b, buf, nBuf ); - if (bzerror == BZ_IO_ERROR) @{ - BZ2_bzWriteClose ( &bzerror, b ); - /* handle error */ - @} -@} - -BZ2_bzWriteClose ( &bzerror, b ); -if (bzerror == BZ_IO_ERROR) @{ - /* handle error */ -@} -@end example -And to read from a compressed file: -@example -FILE* f; -BZFILE* b; -int nBuf; -char buf[ /* whatever size you like */ ]; -int bzerror; -int nWritten; - -f = fopen ( "myfile.bz2", "r" ); -if (!f) @{ - /* handle error */ -@} -b = BZ2_bzReadOpen ( &bzerror, f, 0, NULL, 0 ); -if (bzerror != BZ_OK) @{ - BZ2_bzReadClose ( &bzerror, b ); - /* handle error */ -@} - -bzerror = BZ_OK; -while (bzerror == BZ_OK && /* arbitrary other conditions */) @{ - nBuf = BZ2_bzRead ( &bzerror, b, buf, /* size of buf */ ); - if (bzerror == BZ_OK) @{ - /* do something with buf[0 .. nBuf-1] */ - @} -@} -if (bzerror != BZ_STREAM_END) @{ - BZ2_bzReadClose ( &bzerror, b ); - /* handle error */ -@} else @{ - BZ2_bzReadClose ( &bzerror ); -@} -@end example - - - -@section Utility functions -@subsection @code{BZ2_bzBuffToBuffCompress} -@example - int BZ2_bzBuffToBuffCompress( char* dest, - unsigned int* destLen, - char* source, - unsigned int sourceLen, - int blockSize100k, - int verbosity, - int workFactor ); -@end example -Attempts to compress the data in @code{source[0 .. sourceLen-1]} -into the destination buffer, @code{dest[0 .. *destLen-1]}. -If the destination buffer is big enough, @code{*destLen} is -set to the size of the compressed data, and @code{BZ_OK} is -returned. If the compressed data won't fit, @code{*destLen} -is unchanged, and @code{BZ_OUTBUFF_FULL} is returned. - -Compression in this manner is a one-shot event, done with a single call -to this function. The resulting compressed data is a complete -@code{bzip2} format data stream. There is no mechanism for making -additional calls to provide extra input data. If you want that kind of -mechanism, use the low-level interface. - -For the meaning of parameters @code{blockSize100k}, @code{verbosity} -and @code{workFactor}, @* see @code{BZ2_bzCompressInit}. - -To guarantee that the compressed data will fit in its buffer, allocate -an output buffer of size 1% larger than the uncompressed data, plus -six hundred extra bytes. - -@code{BZ2_bzBuffToBuffDecompress} will not write data at or -beyond @code{dest[*destLen]}, even in case of buffer overflow. - -Possible return values: -@display - @code{BZ_CONFIG_ERROR} - if the library has been mis-compiled - @code{BZ_PARAM_ERROR} - if @code{dest} is @code{NULL} or @code{destLen} is @code{NULL} - or @code{blockSize100k < 1} or @code{blockSize100k > 9} - or @code{verbosity < 0} or @code{verbosity > 4} - or @code{workFactor < 0} or @code{workFactor > 250} - @code{BZ_MEM_ERROR} - if insufficient memory is available - @code{BZ_OUTBUFF_FULL} - if the size of the compressed data exceeds @code{*destLen} - @code{BZ_OK} - otherwise -@end display - - - -@subsection @code{BZ2_bzBuffToBuffDecompress} -@example - int BZ2_bzBuffToBuffDecompress ( char* dest, - unsigned int* destLen, - char* source, - unsigned int sourceLen, - int small, - int verbosity ); -@end example -Attempts to decompress the data in @code{source[0 .. sourceLen-1]} -into the destination buffer, @code{dest[0 .. *destLen-1]}. -If the destination buffer is big enough, @code{*destLen} is -set to the size of the uncompressed data, and @code{BZ_OK} is -returned. If the compressed data won't fit, @code{*destLen} -is unchanged, and @code{BZ_OUTBUFF_FULL} is returned. - -@code{source} is assumed to hold a complete @code{bzip2} format -data stream. @* @code{BZ2_bzBuffToBuffDecompress} tries to decompress -the entirety of the stream into the output buffer. - -For the meaning of parameters @code{small} and @code{verbosity}, -see @code{BZ2_bzDecompressInit}. - -Because the compression ratio of the compressed data cannot be known in -advance, there is no easy way to guarantee that the output buffer will -be big enough. You may of course make arrangements in your code to -record the size of the uncompressed data, but such a mechanism is beyond -the scope of this library. - -@code{BZ2_bzBuffToBuffDecompress} will not write data at or -beyond @code{dest[*destLen]}, even in case of buffer overflow. - -Possible return values: -@display - @code{BZ_CONFIG_ERROR} - if the library has been mis-compiled - @code{BZ_PARAM_ERROR} - if @code{dest} is @code{NULL} or @code{destLen} is @code{NULL} - or @code{small != 0 && small != 1} - or @code{verbosity < 0} or @code{verbosity > 4} - @code{BZ_MEM_ERROR} - if insufficient memory is available - @code{BZ_OUTBUFF_FULL} - if the size of the compressed data exceeds @code{*destLen} - @code{BZ_DATA_ERROR} - if a data integrity error was detected in the compressed data - @code{BZ_DATA_ERROR_MAGIC} - if the compressed data doesn't begin with the right magic bytes - @code{BZ_UNEXPECTED_EOF} - if the compressed data ends unexpectedly - @code{BZ_OK} - otherwise -@end display - - - -@section @code{zlib} compatibility functions -Yoshioka Tsuneo has contributed some functions to -give better @code{zlib} compatibility. These functions are -@code{BZ2_bzopen}, @code{BZ2_bzread}, @code{BZ2_bzwrite}, @code{BZ2_bzflush}, -@code{BZ2_bzclose}, -@code{BZ2_bzerror} and @code{BZ2_bzlibVersion}. -These functions are not (yet) officially part of -the library. If they break, you get to keep all the pieces. -Nevertheless, I think they work ok. -@example -typedef void BZFILE; - -const char * BZ2_bzlibVersion ( void ); -@end example -Returns a string indicating the library version. -@example -BZFILE * BZ2_bzopen ( const char *path, const char *mode ); -BZFILE * BZ2_bzdopen ( int fd, const char *mode ); -@end example -Opens a @code{.bz2} file for reading or writing, using either its name -or a pre-existing file descriptor. -Analogous to @code{fopen} and @code{fdopen}. -@example -int BZ2_bzread ( BZFILE* b, void* buf, int len ); -int BZ2_bzwrite ( BZFILE* b, void* buf, int len ); -@end example -Reads/writes data from/to a previously opened @code{BZFILE}. -Analogous to @code{fread} and @code{fwrite}. -@example -int BZ2_bzflush ( BZFILE* b ); -void BZ2_bzclose ( BZFILE* b ); -@end example -Flushes/closes a @code{BZFILE}. @code{BZ2_bzflush} doesn't actually do -anything. Analogous to @code{fflush} and @code{fclose}. - -@example -const char * BZ2_bzerror ( BZFILE *b, int *errnum ) -@end example -Returns a string describing the more recent error status of -@code{b}, and also sets @code{*errnum} to its numerical value. - - -@section Using the library in a @code{stdio}-free environment - -@subsection Getting rid of @code{stdio} - -In a deeply embedded application, you might want to use just -the memory-to-memory functions. You can do this conveniently -by compiling the library with preprocessor symbol @code{BZ_NO_STDIO} -defined. Doing this gives you a library containing only the following -eight functions: - -@code{BZ2_bzCompressInit}, @code{BZ2_bzCompress}, @code{BZ2_bzCompressEnd} @* -@code{BZ2_bzDecompressInit}, @code{BZ2_bzDecompress}, @code{BZ2_bzDecompressEnd} @* -@code{BZ2_bzBuffToBuffCompress}, @code{BZ2_bzBuffToBuffDecompress} - -When compiled like this, all functions will ignore @code{verbosity} -settings. - -@subsection Critical error handling -@code{libbzip2} contains a number of internal assertion checks which -should, needless to say, never be activated. Nevertheless, if an -assertion should fail, behaviour depends on whether or not the library -was compiled with @code{BZ_NO_STDIO} set. - -For a normal compile, an assertion failure yields the message -@example - bzip2/libbzip2: internal error number N. - This is a bug in bzip2/libbzip2, 1.0.2, 30-Dec-2001. - Please report it to me at: jseward@@acm.org. If this happened - when you were using some program which uses libbzip2 as a - component, you should also report this bug to the author(s) - of that program. Please make an effort to report this bug; - timely and accurate bug reports eventually lead to higher - quality software. Thanks. Julian Seward, 30 December 2001. -@end example -where @code{N} is some error code number. If @code{N == 1007}, it also -prints some extra text advising the reader that unreliable memory is -often associated with internal error 1007. (This is a -frequently-observed-phenomenon with versions 1.0.0/1.0.1). - -@code{exit(3)} is then called. - -For a @code{stdio}-free library, assertion failures result -in a call to a function declared as: -@example - extern void bz_internal_error ( int errcode ); -@end example -The relevant code is passed as a parameter. You should supply -such a function. - -In either case, once an assertion failure has occurred, any -@code{bz_stream} records involved can be regarded as invalid. -You should not attempt to resume normal operation with them. - -You may, of course, change critical error handling to suit -your needs. As I said above, critical errors indicate bugs -in the library and should not occur. All "normal" error -situations are indicated via error return codes from functions, -and can be recovered from. - - -@section Making a Windows DLL -Everything related to Windows has been contributed by Yoshioka Tsuneo -@* (@code{QWF00133@@niftyserve.or.jp} / -@code{tsuneo-y@@is.aist-nara.ac.jp}), so you should send your queries to -him (but perhaps Cc: me, @code{jseward@@acm.org}). - -My vague understanding of what to do is: using Visual C++ 5.0, -open the project file @code{libbz2.dsp}, and build. That's all. - -If you can't -open the project file for some reason, make a new one, naming these files: -@code{blocksort.c}, @code{bzlib.c}, @code{compress.c}, -@code{crctable.c}, @code{decompress.c}, @code{huffman.c}, @* -@code{randtable.c} and @code{libbz2.def}. You will also need -to name the header files @code{bzlib.h} and @code{bzlib_private.h}. - -If you don't use VC++, you may need to define the proprocessor symbol -@code{_WIN32}. - -Finally, @code{dlltest.c} is a sample program using the DLL. It has a -project file, @code{dlltest.dsp}. - -If you just want a makefile for Visual C, have a look at -@code{makefile.msc}. - -Be aware that if you compile @code{bzip2} itself on Win32, you must set -@code{BZ_UNIX} to 0 and @code{BZ_LCCWIN32} to 1, in the file -@code{bzip2.c}, before compiling. Otherwise the resulting binary won't -work correctly. - -I haven't tried any of this stuff myself, but it all looks plausible. - - - -@chapter Miscellanea - -These are just some random thoughts of mine. Your mileage may -vary. - -@section Limitations of the compressed file format -@code{bzip2-1.0}, @code{0.9.5} and @code{0.9.0} -use exactly the same file format as the previous -version, @code{bzip2-0.1}. This decision was made in the interests of -stability. Creating yet another incompatible compressed file format -would create further confusion and disruption for users. - -Nevertheless, this is not a painless decision. Development -work since the release of @code{bzip2-0.1} in August 1997 -has shown complexities in the file format which slow down -decompression and, in retrospect, are unnecessary. These are: -@itemize @bullet -@item The run-length encoder, which is the first of the - compression transformations, is entirely irrelevant. - The original purpose was to protect the sorting algorithm - from the very worst case input: a string of repeated - symbols. But algorithm steps Q6a and Q6b in the original - Burrows-Wheeler technical report (SRC-124) show how - repeats can be handled without difficulty in block - sorting. -@item The randomisation mechanism doesn't really need to be - there. Udi Manber and Gene Myers published a suffix - array construction algorithm a few years back, which - can be employed to sort any block, no matter how - repetitive, in O(N log N) time. Subsequent work by - Kunihiko Sadakane has produced a derivative O(N (log N)^2) - algorithm which usually outperforms the Manber-Myers - algorithm. - - I could have changed to Sadakane's algorithm, but I find - it to be slower than @code{bzip2}'s existing algorithm for - most inputs, and the randomisation mechanism protects - adequately against bad cases. I didn't think it was - a good tradeoff to make. Partly this is due to the fact - that I was not flooded with email complaints about - @code{bzip2-0.1}'s performance on repetitive data, so - perhaps it isn't a problem for real inputs. - - Probably the best long-term solution, - and the one I have incorporated into 0.9.5 and above, - is to use the existing sorting - algorithm initially, and fall back to a O(N (log N)^2) - algorithm if the standard algorithm gets into difficulties. -@item The compressed file format was never designed to be - handled by a library, and I have had to jump though - some hoops to produce an efficient implementation of - decompression. It's a bit hairy. Try passing - @code{decompress.c} through the C preprocessor - and you'll see what I mean. Much of this complexity - could have been avoided if the compressed size of - each block of data was recorded in the data stream. -@item An Adler-32 checksum, rather than a CRC32 checksum, - would be faster to compute. -@end itemize -It would be fair to say that the @code{bzip2} format was frozen -before I properly and fully understood the performance -consequences of doing so. - -Improvements which I was able to incorporate into -0.9.0, despite using the same file format, are: -@itemize @bullet -@item Single array implementation of the inverse BWT. This - significantly speeds up decompression, presumably - because it reduces the number of cache misses. -@item Faster inverse MTF transform for large MTF values. The - new implementation is based on the notion of sliding blocks - of values. -@item @code{bzip2-0.9.0} now reads and writes files with @code{fread} - and @code{fwrite}; version 0.1 used @code{putc} and @code{getc}. - Duh! Well, you live and learn. - -@end itemize -Further ahead, it would be nice -to be able to do random access into files. This will -require some careful design of compressed file formats. - - - -@section Portability issues -After some consideration, I have decided not to use -GNU @code{autoconf} to configure 0.9.5 or 1.0. - -@code{autoconf}, admirable and wonderful though it is, -mainly assists with portability problems between Unix-like -platforms. But @code{bzip2} doesn't have much in the way -of portability problems on Unix; most of the difficulties appear -when porting to the Mac, or to Microsoft's operating systems. -@code{autoconf} doesn't help in those cases, and brings in a -whole load of new complexity. - -Most people should be able to compile the library and program -under Unix straight out-of-the-box, so to speak, especially -if you have a version of GNU C available. - -There are a couple of @code{__inline__} directives in the code. GNU C -(@code{gcc}) should be able to handle them. If you're not using -GNU C, your C compiler shouldn't see them at all. -If your compiler does, for some reason, see them and doesn't -like them, just @code{#define} @code{__inline__} to be @code{/* */}. One -easy way to do this is to compile with the flag @code{-D__inline__=}, -which should be understood by most Unix compilers. - -If you still have difficulties, try compiling with the macro -@code{BZ_STRICT_ANSI} defined. This should enable you to build the -library in a strictly ANSI compliant environment. Building the program -itself like this is dangerous and not supported, since you remove -@code{bzip2}'s checks against compressing directories, symbolic links, -devices, and other not-really-a-file entities. This could cause -filesystem corruption! - -One other thing: if you create a @code{bzip2} binary for public -distribution, please try and link it statically (@code{gcc -s}). This -avoids all sorts of library-version issues that others may encounter -later on. - -If you build @code{bzip2} on Win32, you must set @code{BZ_UNIX} to 0 and -@code{BZ_LCCWIN32} to 1, in the file @code{bzip2.c}, before compiling. -Otherwise the resulting binary won't work correctly. - - - -@section Reporting bugs -I tried pretty hard to make sure @code{bzip2} is -bug free, both by design and by testing. Hopefully -you'll never need to read this section for real. - -Nevertheless, if @code{bzip2} dies with a segmentation -fault, a bus error or an internal assertion failure, it -will ask you to email me a bug report. Experience with -version 0.1 shows that almost all these problems can -be traced to either compiler bugs or hardware problems. -@itemize @bullet -@item -Recompile the program with no optimisation, and see if it -works. And/or try a different compiler. -I heard all sorts of stories about various flavours -of GNU C (and other compilers) generating bad code for -@code{bzip2}, and I've run across two such examples myself. - -2.7.X versions of GNU C are known to generate bad code from -time to time, at high optimisation levels. -If you get problems, try using the flags -@code{-O2} @code{-fomit-frame-pointer} @code{-fno-strength-reduce}. -You should specifically @emph{not} use @code{-funroll-loops}. - -You may notice that the Makefile runs six tests as part of -the build process. If the program passes all of these, it's -a pretty good (but not 100%) indication that the compiler has -done its job correctly. -@item -If @code{bzip2} crashes randomly, and the crashes are not -repeatable, you may have a flaky memory subsystem. @code{bzip2} -really hammers your memory hierarchy, and if it's a bit marginal, -you may get these problems. Ditto if your disk or I/O subsystem -is slowly failing. Yup, this really does happen. - -Try using a different machine of the same type, and see if -you can repeat the problem. -@item This isn't really a bug, but ... If @code{bzip2} tells -you your file is corrupted on decompression, and you -obtained the file via FTP, there is a possibility that you -forgot to tell FTP to do a binary mode transfer. That absolutely -will cause the file to be non-decompressible. You'll have to transfer -it again. -@end itemize - -If you've incorporated @code{libbzip2} into your own program -and are getting problems, please, please, please, check that the -parameters you are passing in calls to the library, are -correct, and in accordance with what the documentation says -is allowable. I have tried to make the library robust against -such problems, but I'm sure I haven't succeeded. - -Finally, if the above comments don't help, you'll have to send -me a bug report. Now, it's just amazing how many people will -send me a bug report saying something like -@display - bzip2 crashed with segmentation fault on my machine -@end display -and absolutely nothing else. Needless to say, a such a report -is @emph{totally, utterly, completely and comprehensively 100% useless; -a waste of your time, my time, and net bandwidth}. -With no details at all, there's no way I can possibly begin -to figure out what the problem is. - -The rules of the game are: facts, facts, facts. Don't omit -them because "oh, they won't be relevant". At the bare -minimum: -@display - Machine type. Operating system version. - Exact version of @code{bzip2} (do @code{bzip2 -V}). - Exact version of the compiler used. - Flags passed to the compiler. -@end display -However, the most important single thing that will help me is -the file that you were trying to compress or decompress at the -time the problem happened. Without that, my ability to do anything -more than speculate about the cause, is limited. - -Please remember that I connect to the Internet with a modem, so -you should contact me before mailing me huge files. - - -@section Did you get the right package? - -@code{bzip2} is a resource hog. It soaks up large amounts of CPU cycles -and memory. Also, it gives very large latencies. In the worst case, you -can feed many megabytes of uncompressed data into the library before -getting any compressed output, so this probably rules out applications -requiring interactive behaviour. - -These aren't faults of my implementation, I hope, but more -an intrinsic property of the Burrows-Wheeler transform (unfortunately). -Maybe this isn't what you want. - -If you want a compressor and/or library which is faster, uses less -memory but gets pretty good compression, and has minimal latency, -consider Jean-loup -Gailly's and Mark Adler's work, @code{zlib-1.1.3} and -@code{gzip-1.2.4}. Look for them at - -@code{http://www.zlib.org} and -@code{http://www.gzip.org} respectively. - -For something faster and lighter still, you might try Markus F X J -Oberhumer's @code{LZO} real-time compression/decompression library, at -@* @code{http://wildsau.idv.uni-linz.ac.at/mfx/lzo.html}. - -If you want to use the @code{bzip2} algorithms to compress small blocks -of data, 64k bytes or smaller, for example on an on-the-fly disk -compressor, you'd be well advised not to use this library. Instead, -I've made a special library tuned for that kind of use. It's part of -@code{e2compr-0.40}, an on-the-fly disk compressor for the Linux -@code{ext2} filesystem. Look at -@code{http://www.netspace.net.au/~reiter/e2compr}. - - - -@section Testing - -A record of the tests I've done. - -First, some data sets: -@itemize @bullet -@item B: a directory containing 6001 files, one for every length in the - range 0 to 6000 bytes. The files contain random lowercase - letters. 18.7 megabytes. -@item H: my home directory tree. Documents, source code, mail files, - compressed data. H contains B, and also a directory of - files designed as boundary cases for the sorting; mostly very - repetitive, nasty files. 565 megabytes. -@item A: directory tree holding various applications built from source: - @code{egcs}, @code{gcc-2.8.1}, KDE, GTK, Octave, etc. - 2200 megabytes. -@end itemize -The tests conducted are as follows. Each test means compressing -(a copy of) each file in the data set, decompressing it and -comparing it against the original. - -First, a bunch of tests with block sizes and internal buffer -sizes set very small, -to detect any problems with the -blocking and buffering mechanisms. -This required modifying the source code so as to try to -break it. -@enumerate -@item Data set H, with - buffer size of 1 byte, and block size of 23 bytes. -@item Data set B, buffer sizes 1 byte, block size 1 byte. -@item As (2) but small-mode decompression. -@item As (2) with block size 2 bytes. -@item As (2) with block size 3 bytes. -@item As (2) with block size 4 bytes. -@item As (2) with block size 5 bytes. -@item As (2) with block size 6 bytes and small-mode decompression. -@item H with buffer size of 1 byte, but normal block - size (up to 900000 bytes). -@end enumerate -Then some tests with unmodified source code. -@enumerate -@item H, all settings normal. -@item As (1), with small-mode decompress. -@item H, compress with flag @code{-1}. -@item H, compress with flag @code{-s}, decompress with flag @code{-s}. -@item Forwards compatibility: H, @code{bzip2-0.1pl2} compressing, - @code{bzip2-0.9.5} decompressing, all settings normal. -@item Backwards compatibility: H, @code{bzip2-0.9.5} compressing, - @code{bzip2-0.1pl2} decompressing, all settings normal. -@item Bigger tests: A, all settings normal. -@item As (7), using the fallback (Sadakane-like) sorting algorithm. -@item As (8), compress with flag @code{-1}, decompress with flag - @code{-s}. -@item H, using the fallback sorting algorithm. -@item Forwards compatibility: A, @code{bzip2-0.1pl2} compressing, - @code{bzip2-0.9.5} decompressing, all settings normal. -@item Backwards compatibility: A, @code{bzip2-0.9.5} compressing, - @code{bzip2-0.1pl2} decompressing, all settings normal. -@item Misc test: about 400 megabytes of @code{.tar} files with - @code{bzip2} compiled with Checker (a memory access error - detector, like Purify). -@item Misc tests to make sure it builds and runs ok on non-Linux/x86 - platforms. -@end enumerate -These tests were conducted on a 225 MHz IDT WinChip machine, running -Linux 2.0.36. They represent nearly a week of continuous computation. -All tests completed successfully. - - -@section Further reading -@code{bzip2} is not research work, in the sense that it doesn't present -any new ideas. Rather, it's an engineering exercise based on existing -ideas. - -Four documents describe essentially all the ideas behind @code{bzip2}: -@example -Michael Burrows and D. J. Wheeler: - "A block-sorting lossless data compression algorithm" - 10th May 1994. - Digital SRC Research Report 124. - ftp://ftp.digital.com/pub/DEC/SRC/research-reports/SRC-124.ps.gz - If you have trouble finding it, try searching at the - New Zealand Digital Library, http://www.nzdl.org. - -Daniel S. Hirschberg and Debra A. LeLewer - "Efficient Decoding of Prefix Codes" - Communications of the ACM, April 1990, Vol 33, Number 4. - You might be able to get an electronic copy of this - from the ACM Digital Library. - -David J. Wheeler - Program bred3.c and accompanying document bred3.ps. - This contains the idea behind the multi-table Huffman - coding scheme. - ftp://ftp.cl.cam.ac.uk/users/djw3/ - -Jon L. Bentley and Robert Sedgewick - "Fast Algorithms for Sorting and Searching Strings" - Available from Sedgewick's web page, - www.cs.princeton.edu/~rs -@end example -The following paper gives valuable additional insights into the -algorithm, but is not immediately the basis of any code -used in bzip2. -@example -Peter Fenwick: - Block Sorting Text Compression - Proceedings of the 19th Australasian Computer Science Conference, - Melbourne, Australia. Jan 31 - Feb 2, 1996. - ftp://ftp.cs.auckland.ac.nz/pub/peter-f/ACSC96paper.ps -@end example -Kunihiko Sadakane's sorting algorithm, mentioned above, -is available from: -@example -http://naomi.is.s.u-tokyo.ac.jp/~sada/papers/Sada98b.ps.gz -@end example -The Manber-Myers suffix array construction -algorithm is described in a paper -available from: -@example -http://www.cs.arizona.edu/people/gene/PAPERS/suffix.ps -@end example -Finally, the following paper documents some recent investigations -I made into the performance of sorting algorithms: -@example -Julian Seward: - On the Performance of BWT Sorting Algorithms - Proceedings of the IEEE Data Compression Conference 2000 - Snowbird, Utah. 28-30 March 2000. -@end example - - -@contents - -@bye - diff -Nru bzip2-1.0.2/manual_1.html bzip2-1.0.2.new/manual_1.html --- bzip2-1.0.2/manual_1.html Sat Jan 5 01:53:05 2002 +++ bzip2-1.0.2.new/manual_1.html Thu Jan 1 01:00:00 1970 @@ -1,81 +0,0 @@ - - - - - -Untitled Document: 1. Introduction - - - - - - - - - - - - - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

1. Introduction

- -

- -bzip2 compresses files using the Burrows-Wheeler -block-sorting text compression algorithm, and Huffman coding. -Compression is generally considerably better than that -achieved by more conventional LZ77/LZ78-based compressors, -and approaches the performance of the PPM family of statistical compressors. -

- -bzip2 is built on top of libbzip2, a flexible library -for handling compressed data in the bzip2 format. This manual -describes both how to use the program and -how to work with the library interface. Most of the -manual is devoted to this library, not the program, -which is good news if your interest is only in the program. -

- -Chapter 2 describes how to use bzip2; this is the only part -you need to read if you just want to know how to operate the program. -Chapter 3 describes the programming interfaces in detail, and -Chapter 4 records some miscellaneous notes which I thought -ought to be recorded somewhere. -

- -

- - - - - - - -
[ << ][ >> ]           [Top][Contents][Index][ ? ]
-
- -This document was generated -by Julian Seward on January, 5 2002 -using texi2html - - - diff -Nru bzip2-1.0.2/manual_2.html bzip2-1.0.2.new/manual_2.html --- bzip2-1.0.2/manual_2.html Sat Jan 5 01:53:05 2002 +++ bzip2-1.0.2.new/manual_2.html Thu Jan 1 01:00:00 1970 @@ -1,579 +0,0 @@ - - - - - -Untitled Document: 2. How to use <CODE>bzip2</CODE> - - - - - - - - - - - - - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

2. How to use bzip2

- -

- -This chapter contains a copy of the bzip2 man page, -and nothing else. -

- -

-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

NAME

- -
    -
  • bzip2, bunzip2 -- a block-sorting file compressor, v1.0.2 -
  • bzcat -- decompresses files to stdout -
  • bzip2recover -- recovers data from damaged bzip2 files -
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

SYNOPSIS

- -
    -
  • bzip2 [ -cdfkqstvzVL123456789 ] [ filenames ... ] -
  • bunzip2 [ -fkvsVL ] [ filenames ... ] -
  • bzcat [ -s ] [ filenames ... ] -
  • bzip2recover filename -
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

DESCRIPTION

- -

- -bzip2 compresses files using the Burrows-Wheeler block sorting -text compression algorithm, and Huffman coding. Compression is -generally considerably better than that achieved by more conventional -LZ77/LZ78-based compressors, and approaches the performance of the PPM -family of statistical compressors. -

- -The command-line options are deliberately very similar to those of GNU -gzip, but they are not identical. -

- -bzip2 expects a list of file names to accompany the command-line -flags. Each file is replaced by a compressed version of itself, with -the name original_name.bz2. Each compressed file has the same -modification date, permissions, and, when possible, ownership as the -corresponding original, so that these properties can be correctly -restored at decompression time. File name handling is naive in the -sense that there is no mechanism for preserving original file names, -permissions, ownerships or dates in filesystems which lack these -concepts, or have serious file name length restrictions, such as MS-DOS. -

- -bzip2 and bunzip2 will by default not overwrite existing -files. If you want this to happen, specify the -f flag. -

- -If no file names are specified, bzip2 compresses from standard -input to standard output. In this case, bzip2 will decline to -write compressed output to a terminal, as this would be entirely -incomprehensible and therefore pointless. -

- -bunzip2 (or bzip2 -d) decompresses all -specified files. Files which were not created by bzip2 -will be detected and ignored, and a warning issued. -bzip2 attempts to guess the filename for the decompressed file -from that of the compressed file as follows: -

    -
  • filename.bz2 becomes filename -
  • filename.bz becomes filename -
  • filename.tbz2 becomes filename.tar -
  • filename.tbz becomes filename.tar -
  • anyothername becomes anyothername.out -
-If the file does not end in one of the recognised endings, -.bz2, .bz, -.tbz2 or .tbz, bzip2 complains that it cannot -guess the name of the original file, and uses the original name -with .out appended. -

- -As with compression, supplying no -filenames causes decompression from standard input to standard output. -

- -bunzip2 will correctly decompress a file which is the -concatenation of two or more compressed files. The result is the -concatenation of the corresponding uncompressed files. Integrity -testing (-t) of concatenated compressed files is also supported. -

- -You can also compress or decompress files to the standard output by -giving the -c flag. Multiple files may be compressed and -decompressed like this. The resulting outputs are fed sequentially to -stdout. Compression of multiple files in this manner generates a stream -containing multiple compressed file representations. Such a stream -can be decompressed correctly only by bzip2 version 0.9.0 or -later. Earlier versions of bzip2 will stop after decompressing -the first file in the stream. -

- -bzcat (or bzip2 -dc) decompresses all specified files to -the standard output. -

- -bzip2 will read arguments from the environment variables -BZIP2 and BZIP, in that order, and will process them -before any arguments read from the command line. This gives a -convenient way to supply default arguments. -

- -Compression is always performed, even if the compressed file is slightly -larger than the original. Files of less than about one hundred bytes -tend to get larger, since the compression mechanism has a constant -overhead in the region of 50 bytes. Random data (including the output -of most file compressors) is coded at about 8.05 bits per byte, giving -an expansion of around 0.5%. -

- -As a self-check for your protection, bzip2 uses 32-bit CRCs to -make sure that the decompressed version of a file is identical to the -original. This guards against corruption of the compressed data, and -against undetected bugs in bzip2 (hopefully very unlikely). The -chances of data corruption going undetected is microscopic, about one -chance in four billion for each file processed. Be aware, though, that -the check occurs upon decompression, so it can only tell you that -something is wrong. It can't help you recover the original uncompressed -data. You can use bzip2recover to try to recover data from -damaged files. -

- -Return values: 0 for a normal exit, 1 for environmental problems (file -not found, invalid flags, I/O errors, &c), 2 to indicate a corrupt -compressed file, 3 for an internal consistency error (eg, bug) which -caused bzip2 to panic. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

OPTIONS

- -
-
-c --stdout -
Compress or decompress to standard output. -
-d --decompress -
Force decompression. bzip2, bunzip2 and bzcat are -really the same program, and the decision about what actions to take is -done on the basis of which name is used. This flag overrides that -mechanism, and forces bzip2 to decompress. -
-z --compress -
The complement to -d: forces compression, regardless of the -invokation name. -
-t --test -
Check integrity of the specified file(s), but don't decompress them. -This really performs a trial decompression and throws away the result. -
-f --force -
Force overwrite of output files. Normally, bzip2 will not overwrite -existing output files. Also forces bzip2 to break hard links -to files, which it otherwise wouldn't do. -

- -bzip2 normally declines to decompress files which don't have the -correct magic header bytes. If forced (-f), however, it will -pass such files through unmodified. This is how GNU gzip -behaves. -

-k --keep -
Keep (don't delete) input files during compression -or decompression. -
-s --small -
Reduce memory usage, for compression, decompression and testing. Files -are decompressed and tested using a modified algorithm which only -requires 2.5 bytes per block byte. This means any file can be -decompressed in 2300k of memory, albeit at about half the normal speed. -

- -During compression, -s selects a block size of 200k, which limits -memory use to around the same figure, at the expense of your compression -ratio. In short, if your machine is low on memory (8 megabytes or -less), use -s for everything. See MEMORY MANAGEMENT below. -

-q --quiet -
Suppress non-essential warning messages. Messages pertaining to -I/O errors and other critical events will not be suppressed. -
-v --verbose -
Verbose mode -- show the compression ratio for each file processed. -Further -v's increase the verbosity level, spewing out lots of -information which is primarily of interest for diagnostic purposes. -
-L --license -V --version -
Display the software version, license terms and conditions. -
-1 (or --fast) to -9 (or --best) -
Set the block size to 100 k, 200 k .. 900 k when compressing. Has no -effect when decompressing. See MEMORY MANAGEMENT below. -The --fast and --best aliases are primarily for GNU -gzip compatibility. In particular, --fast doesn't make -things significantly faster. And --best merely selects the -default behaviour. -
-- -
Treats all subsequent arguments as file names, even if they start -with a dash. This is so you can handle files with names beginning -with a dash, for example: bzip2 -- -myfilename. -
--repetitive-fast -
--repetitive-best -
These flags are redundant in versions 0.9.5 and above. They provided -some coarse control over the behaviour of the sorting algorithm in -earlier versions, which was sometimes useful. 0.9.5 and above have an -improved algorithm which renders these flags irrelevant. -
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

MEMORY MANAGEMENT

- -

- -bzip2 compresses large files in blocks. The block size affects -both the compression ratio achieved, and the amount of memory needed for -compression and decompression. The flags -1 through -9 -specify the block size to be 100,000 bytes through 900,000 bytes (the -default) respectively. At decompression time, the block size used for -compression is read from the header of the compressed file, and -bunzip2 then allocates itself just enough memory to decompress -the file. Since block sizes are stored in compressed files, it follows -that the flags -1 to -9 are irrelevant to and so ignored -during decompression. -

- -Compression and decompression requirements, in bytes, can be estimated -as: -
 
     Compression:   400k + ( 8 x block size )
-
-     Decompression: 100k + ( 4 x block size ), or
-                    100k + ( 2.5 x block size )
-
Larger block sizes give rapidly diminishing marginal returns. Most of -the compression comes from the first two or three hundred k of block -size, a fact worth bearing in mind when using bzip2 on small machines. -It is also important to appreciate that the decompression memory -requirement is set at compression time by the choice of block size. -

- -For files compressed with the default 900k block size, bunzip2 -will require about 3700 kbytes to decompress. To support decompression -of any file on a 4 megabyte machine, bunzip2 has an option to -decompress using approximately half this amount of memory, about 2300 -kbytes. Decompression speed is also halved, so you should use this -option only where necessary. The relevant flag is -s. -

- -In general, try and use the largest block size memory constraints allow, -since that maximises the compression achieved. Compression and -decompression speed are virtually unaffected by block size. -

- -Another significant point applies to files which fit in a single block --- that means most files you'd encounter using a large block size. The -amount of real memory touched is proportional to the size of the file, -since the file is smaller than a block. For example, compressing a file -20,000 bytes long with the flag -9 will cause the compressor to -allocate around 7600k of memory, but only touch 400k + 20000 * 8 = 560 -kbytes of it. Similarly, the decompressor will allocate 3700k but only -touch 100k + 20000 * 4 = 180 kbytes. -

- -Here is a table which summarises the maximum memory usage for different -block sizes. Also recorded is the total compressed size for 14 files of -the Calgary Text Compression Corpus totalling 3,141,622 bytes. This -column gives some feel for how compression varies with block size. -These figures tend to understate the advantage of larger block sizes for -larger files, since the Corpus is dominated by smaller files. -
 
          Compress   Decompress   Decompress   Corpus
-   Flag     usage      usage       -s usage     Size
-
-    -1      1200k       500k         350k      914704
-    -2      2000k       900k         600k      877703
-    -3      2800k      1300k         850k      860338
-    -4      3600k      1700k        1100k      846899
-    -5      4400k      2100k        1350k      845160
-    -6      5200k      2500k        1600k      838626
-    -7      6100k      2900k        1850k      834096
-    -8      6800k      3300k        2100k      828642
-    -9      7600k      3700k        2350k      828642
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

RECOVERING DATA FROM DAMAGED FILES

- -

- -bzip2 compresses files in blocks, usually 900kbytes long. Each -block is handled independently. If a media or transmission error causes -a multi-block .bz2 file to become damaged, it may be possible to -recover data from the undamaged blocks in the file. -

- -The compressed representation of each block is delimited by a 48-bit -pattern, which makes it possible to find the block boundaries with -reasonable certainty. Each block also carries its own 32-bit CRC, so -damaged blocks can be distinguished from undamaged ones. -

- -bzip2recover is a simple program whose purpose is to search for -blocks in .bz2 files, and write each block out into its own -.bz2 file. You can then use bzip2 -t to test the -integrity of the resulting files, and decompress those which are -undamaged. -

- -bzip2recover -takes a single argument, the name of the damaged file, and writes a -number of files rec00001file.bz2, rec00002file.bz2, etc, -containing the extracted blocks. The output filenames are designed so -that the use of wildcards in subsequent processing -- for example, -bzip2 -dc rec*file.bz2 > recovered_data -- processes the files in -the correct order. -

- -bzip2recover should be of most use dealing with large .bz2 -files, as these will contain many blocks. It is clearly futile to use -it on damaged single-block files, since a damaged block cannot be -recovered. If you wish to minimise any potential data loss through -media or transmission errors, you might consider compressing with a -smaller block size. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

PERFORMANCE NOTES

- -

- -The sorting phase of compression gathers together similar strings in the -file. Because of this, files containing very long runs of repeated -symbols, like "aabaabaabaab ..." (repeated several hundred times) may -compress more slowly than normal. Versions 0.9.5 and above fare much -better than previous versions in this respect. The ratio between -worst-case and average-case compression time is in the region of 10:1. -For previous versions, this figure was more like 100:1. You can use the --vvvv option to monitor progress in great detail, if you want. -

- -Decompression speed is unaffected by these phenomena. -

- -bzip2 usually allocates several megabytes of memory to operate -in, and then charges all over it in a fairly random fashion. This means -that performance, both for compressing and decompressing, is largely -determined by the speed at which your machine can service cache misses. -Because of this, small changes to the code to reduce the miss rate have -been observed to give disproportionately large performance improvements. -I imagine bzip2 will perform best on machines with very large -caches. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

CAVEATS

- -

- -I/O error messages are not as helpful as they could be. bzip2 -tries hard to detect I/O errors and exit cleanly, but the details of -what the problem is sometimes seem rather misleading. -

- -This manual page pertains to version 1.0.2 of bzip2. Compressed -data created by this version is entirely forwards and backwards -compatible with the previous public releases, versions 0.1pl2, 0.9.0, -0.9.5, 1.0.0 and 1.0.1, but with the following exception: 0.9.0 and -above can correctly decompress multiple concatenated compressed files. -0.1pl2 cannot do this; it will stop after decompressing just the first -file in the stream. -

- -bzip2recover versions prior to this one, 1.0.2, used 32-bit -integers to represent bit positions in compressed files, so it could not -handle compressed files more than 512 megabytes long. Version 1.0.2 and -above uses 64-bit ints on some platforms which support them (GNU -supported targets, and Windows). To establish whether or not -bzip2recover was built with such a limitation, run it without -arguments. In any event you can build yourself an unlimited version if -you can recompile it with MaybeUInt64 set to be an unsigned -64-bit integer. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

AUTHOR

- -Julian Seward, jseward@acm.org. -

- -http://sources.redhat.com/bzip2 -

- -The ideas embodied in bzip2 are due to (at least) the following -people: Michael Burrows and David Wheeler (for the block sorting -transformation), David Wheeler (again, for the Huffman coder), Peter -Fenwick (for the structured coding model in the original bzip, -and many refinements), and Alistair Moffat, Radford Neal and Ian Witten -(for the arithmetic coder in the original bzip). I am much -indebted for their help, support and advice. See the manual in the -source distribution for pointers to sources of documentation. Christian -von Roques encouraged me to look for faster sorting algorithms, so as to -speed up compression. Bela Lubkin encouraged me to improve the -worst-case compression performance. The bz* scripts are derived -from those of GNU gzip. Many people sent patches, helped with -portability problems, lent machines, gave advice and were generally -helpful. -

- -

- -
- - - - - - - -
[ << ][ >> ]           [Top][Contents][Index][ ? ]
-
- -This document was generated -by Julian Seward on January, 5 2002 -using texi2html - - - diff -Nru bzip2-1.0.2/manual_3.html bzip2-1.0.2.new/manual_3.html --- bzip2-1.0.2/manual_3.html Sat Jan 5 01:53:05 2002 +++ bzip2-1.0.2.new/manual_3.html Thu Jan 1 01:00:00 1970 @@ -1,1855 +0,0 @@ - - - - - -Untitled Document: 3. Programming with <CODE>libbzip2</CODE> - - - - - - - - - - - - - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3. Programming with libbzip2

- -

- -This chapter describes the programming interface to libbzip2. -

- -For general background information, particularly about memory -use and performance aspects, you'd be well advised to read Chapter 2 -as well. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.1 Top-level structure

- -

- -libbzip2 is a flexible library for compressing and decompressing -data in the bzip2 data format. Although packaged as a single -entity, it helps to regard the library as three separate parts: the low -level interface, and the high level interface, and some utility -functions. -

- -The structure of libbzip2's interfaces is similar to -that of Jean-loup Gailly's and Mark Adler's excellent zlib -library. -

- -All externally visible symbols have names beginning BZ2_. -This is new in version 1.0. The intention is to minimise pollution -of the namespaces of library clients. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.1.1 Low-level summary

- -

- -This interface provides services for compressing and decompressing -data in memory. There's no provision for dealing with files, streams -or any other I/O mechanisms, just straight memory-to-memory work. -In fact, this part of the library can be compiled without inclusion -of stdio.h, which may be helpful for embedded applications. -

- -The low-level part of the library has no global variables and -is therefore thread-safe. -

- -Six routines make up the low level interface: -BZ2_bzCompressInit, BZ2_bzCompress, and
BZ2_bzCompressEnd -for compression, -and a corresponding trio BZ2_bzDecompressInit,
BZ2_bzDecompress -and BZ2_bzDecompressEnd for decompression. -The *Init functions allocate -memory for compression/decompression and do other -initialisations, whilst the *End functions close down operations -and release memory. -

- -The real work is done by BZ2_bzCompress and BZ2_bzDecompress. -These compress and decompress data from a user-supplied input buffer -to a user-supplied output buffer. These buffers can be any size; -arbitrary quantities of data are handled by making repeated calls -to these functions. This is a flexible mechanism allowing a -consumer-pull style of activity, or producer-push, or a mixture of -both. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.1.2 High-level summary

- -

- -This interface provides some handy wrappers around the low-level -interface to facilitate reading and writing bzip2 format -files (.bz2 files). The routines provide hooks to facilitate -reading files in which the bzip2 data stream is embedded -within some larger-scale file structure, or where there are -multiple bzip2 data streams concatenated end-to-end. -

- -For reading files, BZ2_bzReadOpen, BZ2_bzRead, -BZ2_bzReadClose and
BZ2_bzReadGetUnused are supplied. For -writing files, BZ2_bzWriteOpen, BZ2_bzWrite and -BZ2_bzWriteFinish are available. -

- -As with the low-level library, no global variables are used -so the library is per se thread-safe. However, if I/O errors -occur whilst reading or writing the underlying compressed files, -you may have to consult errno to determine the cause of -the error. In that case, you'd need a C library which correctly -supports errno in a multithreaded environment. -

- -To make the library a little simpler and more portable, -BZ2_bzReadOpen and BZ2_bzWriteOpen require you to pass them file -handles (FILE*s) which have previously been opened for reading or -writing respectively. That avoids portability problems associated with -file operations and file attributes, whilst not being much of an -imposition on the programmer. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.1.3 Utility functions summary

- -For very simple needs, BZ2_bzBuffToBuffCompress and -BZ2_bzBuffToBuffDecompress are provided. These compress -data in memory from one buffer to another buffer in a single -function call. You should assess whether these functions -fulfill your memory-to-memory compression/decompression -requirements before investing effort in understanding the more -general but more complex low-level interface. -

- -Yoshioka Tsuneo (QWF00133@niftyserve.or.jp / -tsuneo-y@is.aist-nara.ac.jp) has contributed some functions to -give better zlib compatibility. These functions are -BZ2_bzopen, BZ2_bzread, BZ2_bzwrite, BZ2_bzflush, -BZ2_bzclose, -BZ2_bzerror and BZ2_bzlibVersion. You may find these functions -more convenient for simple file reading and writing, than those in the -high-level interface. These functions are not (yet) officially part of -the library, and are minimally documented here. If they break, you -get to keep all the pieces. I hope to document them properly when time -permits. -

- -Yoshioka also contributed modifications to allow the library to be -built as a Windows DLL. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.2 Error handling

- -

- -The library is designed to recover cleanly in all situations, including -the worst-case situation of decompressing random data. I'm not -100% sure that it can always do this, so you might want to add -a signal handler to catch segmentation violations during decompression -if you are feeling especially paranoid. I would be interested in -hearing more about the robustness of the library to corrupted -compressed data. -

- -Version 1.0 is much more robust in this respect than -0.9.0 or 0.9.5. Investigations with Checker (a tool for -detecting problems with memory management, similar to Purify) -indicate that, at least for the few files I tested, all single-bit -errors in the decompressed data are caught properly, with no -segmentation faults, no reads of uninitialised data and no -out of range reads or writes. So it's certainly much improved, -although I wouldn't claim it to be totally bombproof. -

- -The file bzlib.h contains all definitions needed to use -the library. In particular, you should definitely not include -bzlib_private.h. -

- -In bzlib.h, the various return values are defined. The following -list is not intended as an exhaustive description of the circumstances -in which a given value may be returned -- those descriptions are given -later. Rather, it is intended to convey the rough meaning of each -return value. The first five actions are normal and not intended to -denote an error situation. -

-
BZ_OK -
The requested action was completed successfully. -
BZ_RUN_OK -
BZ_FLUSH_OK -
BZ_FINISH_OK -
In BZ2_bzCompress, the requested flush/finish/nothing-special action -was completed successfully. -
BZ_STREAM_END -
Compression of data was completed, or the logical stream end was -detected during decompression. -
-

- -The following return values indicate an error of some kind. -

-
BZ_CONFIG_ERROR -
Indicates that the library has been improperly compiled on your -platform -- a major configuration error. Specifically, it means -that sizeof(char), sizeof(short) and sizeof(int) -are not 1, 2 and 4 respectively, as they should be. Note that the -library should still work properly on 64-bit platforms which follow -the LP64 programming model -- that is, where sizeof(long) -and sizeof(void*) are 8. Under LP64, sizeof(int) is -still 4, so libbzip2, which doesn't use the long type, -is OK. -
BZ_SEQUENCE_ERROR -
When using the library, it is important to call the functions in the -correct sequence and with data structures (buffers etc) in the correct -states. libbzip2 checks as much as it can to ensure this is -happening, and returns BZ_SEQUENCE_ERROR if not. Code which -complies precisely with the function semantics, as detailed below, -should never receive this value; such an event denotes buggy code -which you should investigate. -
BZ_PARAM_ERROR -
Returned when a parameter to a function call is out of range -or otherwise manifestly incorrect. As with BZ_SEQUENCE_ERROR, -this denotes a bug in the client code. The distinction between -BZ_PARAM_ERROR and BZ_SEQUENCE_ERROR is a bit hazy, but still worth -making. -
BZ_MEM_ERROR -
Returned when a request to allocate memory failed. Note that the -quantity of memory needed to decompress a stream cannot be determined -until the stream's header has been read. So BZ2_bzDecompress and -BZ2_bzRead may return BZ_MEM_ERROR even though some of -the compressed data has been read. The same is not true for -compression; once BZ2_bzCompressInit or BZ2_bzWriteOpen have -successfully completed, BZ_MEM_ERROR cannot occur. -
BZ_DATA_ERROR -
Returned when a data integrity error is detected during decompression. -Most importantly, this means when stored and computed CRCs for the -data do not match. This value is also returned upon detection of any -other anomaly in the compressed data. -
BZ_DATA_ERROR_MAGIC -
As a special case of BZ_DATA_ERROR, it is sometimes useful to -know when the compressed stream does not start with the correct -magic bytes ('B' 'Z' 'h'). -
BZ_IO_ERROR -
Returned by BZ2_bzRead and BZ2_bzWrite when there is an error -reading or writing in the compressed file, and by BZ2_bzReadOpen -and BZ2_bzWriteOpen for attempts to use a file for which the -error indicator (viz, ferror(f)) is set. -On receipt of BZ_IO_ERROR, the caller should consult -errno and/or perror to acquire operating-system -specific information about the problem. -
BZ_UNEXPECTED_EOF -
Returned by BZ2_bzRead when the compressed file finishes -before the logical end of stream is detected. -
BZ_OUTBUFF_FULL -
Returned by BZ2_bzBuffToBuffCompress and -BZ2_bzBuffToBuffDecompress to indicate that the output data -will not fit into the output buffer provided. -
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3 Low-level interface

- -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3.1 BZ2_bzCompressInit

- -
 
typedef 
-   struct {
-      char *next_in;
-      unsigned int avail_in;
-      unsigned int total_in_lo32;
-      unsigned int total_in_hi32;
-
-      char *next_out;
-      unsigned int avail_out;
-      unsigned int total_out_lo32;
-      unsigned int total_out_hi32;
-
-      void *state;
-
-      void *(*bzalloc)(void *,int,int);
-      void (*bzfree)(void *,void *);
-      void *opaque;
-   } 
-   bz_stream;
-
-int BZ2_bzCompressInit ( bz_stream *strm, 
-                         int blockSize100k, 
-                         int verbosity,
-                         int workFactor );
-
-

- -Prepares for compression. The bz_stream structure -holds all data pertaining to the compression activity. -A bz_stream structure should be allocated and initialised -prior to the call. -The fields of bz_stream -comprise the entirety of the user-visible data. state -is a pointer to the private data structures required for compression. -

- -Custom memory allocators are supported, via fields bzalloc, -bzfree, -and opaque. The value -opaque is passed to as the first argument to -all calls to bzalloc and bzfree, but is -otherwise ignored by the library. -The call bzalloc ( opaque, n, m ) is expected to return a -pointer p to -n * m bytes of memory, and bzfree ( opaque, p ) -should free -that memory. -

- -If you don't want to use a custom memory allocator, set bzalloc, -bzfree and -opaque to NULL, -and the library will then use the standard malloc/free -routines. -

- -Before calling BZ2_bzCompressInit, fields bzalloc, -bzfree and opaque should -be filled appropriately, as just described. Upon return, the internal -state will have been allocated and initialised, and total_in_lo32, -total_in_hi32, total_out_lo32 and -total_out_hi32 will have been set to zero. -These four fields are used by the library -to inform the caller of the total amount of data passed into and out of -the library, respectively. You should not try to change them. -As of version 1.0, 64-bit counts are maintained, even on 32-bit -platforms, using the _hi32 fields to store the upper 32 bits -of the count. So, for example, the total amount of data in -is (total_in_hi32 << 32) + total_in_lo32. -

- -Parameter blockSize100k specifies the block size to be used for -compression. It should be a value between 1 and 9 inclusive, and the -actual block size used is 100000 x this figure. 9 gives the best -compression but takes most memory. -

- -Parameter verbosity should be set to a number between 0 and 4 -inclusive. 0 is silent, and greater numbers give increasingly verbose -monitoring/debugging output. If the library has been compiled with --DBZ_NO_STDIO, no such output will appear for any verbosity -setting. -

- -Parameter workFactor controls how the compression phase behaves -when presented with worst case, highly repetitive, input data. If -compression runs into difficulties caused by repetitive data, the -library switches from the standard sorting algorithm to a fallback -algorithm. The fallback is slower than the standard algorithm by -perhaps a factor of three, but always behaves reasonably, no matter how -bad the input. -

- -Lower values of workFactor reduce the amount of effort the -standard algorithm will expend before resorting to the fallback. You -should set this parameter carefully; too low, and many inputs will be -handled by the fallback algorithm and so compress rather slowly, too -high, and your average-to-worst case compression times can become very -large. The default value of 30 gives reasonable behaviour over a wide -range of circumstances. -

- -Allowable values range from 0 to 250 inclusive. 0 is a special case, -equivalent to using the default value of 30. -

- -Note that the compressed output generated is the same regardless of -whether or not the fallback algorithm is used. -

- -Be aware also that this parameter may disappear entirely in future -versions of the library. In principle it should be possible to devise a -good way to automatically choose which algorithm to use. Such a -mechanism would render the parameter obsolete. -

- -Possible return values: -
 
      BZ_CONFIG_ERROR
-         if the library has been mis-compiled
-      BZ_PARAM_ERROR 
-         if strm is NULL 
-         or blockSize < 1 or blockSize > 9
-         or verbosity < 0 or verbosity > 4
-         or workFactor < 0 or workFactor > 250
-      BZ_MEM_ERROR 
-         if not enough memory is available
-      BZ_OK 
-         otherwise
-
Allowable next actions: -
 
      BZ2_bzCompress 
-         if BZ_OK is returned
-      no specific action needed in case of error
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3.2 BZ2_bzCompress

- -
 
   int BZ2_bzCompress ( bz_stream *strm, int action );
-
Provides more input and/or output buffer space for the library. The -caller maintains input and output buffers, and calls BZ2_bzCompress to -transfer data between them. -

- -Before each call to BZ2_bzCompress, next_in should point at -the data to be compressed, and avail_in should indicate how many -bytes the library may read. BZ2_bzCompress updates next_in, -avail_in and total_in to reflect the number of bytes it -has read. -

- -Similarly, next_out should point to a buffer in which the -compressed data is to be placed, with avail_out indicating how -much output space is available. BZ2_bzCompress updates -next_out, avail_out and total_out to reflect the -number of bytes output. -

- -You may provide and remove as little or as much data as you like on each -call of BZ2_bzCompress. In the limit, it is acceptable to supply and -remove data one byte at a time, although this would be terribly -inefficient. You should always ensure that at least one byte of output -space is available at each call. -

- -A second purpose of BZ2_bzCompress is to request a change of mode of the -compressed stream. -

- -Conceptually, a compressed stream can be in one of four states: IDLE, -RUNNING, FLUSHING and FINISHING. Before initialisation -(BZ2_bzCompressInit) and after termination (BZ2_bzCompressEnd), a -stream is regarded as IDLE. -

- -Upon initialisation (BZ2_bzCompressInit), the stream is placed in the -RUNNING state. Subsequent calls to BZ2_bzCompress should pass -BZ_RUN as the requested action; other actions are illegal and -will result in BZ_SEQUENCE_ERROR. -

- -At some point, the calling program will have provided all the input data -it wants to. It will then want to finish up -- in effect, asking the -library to process any data it might have buffered internally. In this -state, BZ2_bzCompress will no longer attempt to read data from -next_in, but it will want to write data to next_out. -Because the output buffer supplied by the user can be arbitrarily small, -the finishing-up operation cannot necessarily be done with a single call -of BZ2_bzCompress. -

- -Instead, the calling program passes BZ_FINISH as an action to -BZ2_bzCompress. This changes the stream's state to FINISHING. Any -remaining input (ie, next_in[0 .. avail_in-1]) is compressed and -transferred to the output buffer. To do this, BZ2_bzCompress must be -called repeatedly until all the output has been consumed. At that -point, BZ2_bzCompress returns BZ_STREAM_END, and the stream's -state is set back to IDLE. BZ2_bzCompressEnd should then be -called. -

- -Just to make sure the calling program does not cheat, the library makes -a note of avail_in at the time of the first call to -BZ2_bzCompress which has BZ_FINISH as an action (ie, at the -time the program has announced its intention to not supply any more -input). By comparing this value with that of avail_in over -subsequent calls to BZ2_bzCompress, the library can detect any -attempts to slip in more data to compress. Any calls for which this is -detected will return BZ_SEQUENCE_ERROR. This indicates a -programming mistake which should be corrected. -

- -Instead of asking to finish, the calling program may ask -BZ2_bzCompress to take all the remaining input, compress it and -terminate the current (Burrows-Wheeler) compression block. This could -be useful for error control purposes. The mechanism is analogous to -that for finishing: call BZ2_bzCompress with an action of -BZ_FLUSH, remove output data, and persist with the -BZ_FLUSH action until the value BZ_RUN is returned. As -with finishing, BZ2_bzCompress detects any attempt to provide more -input data once the flush has begun. -

- -Once the flush is complete, the stream returns to the normal RUNNING -state. -

- -This all sounds pretty complex, but isn't really. Here's a table -which shows which actions are allowable in each state, what action -will be taken, what the next state is, and what the non-error return -values are. Note that you can't explicitly ask what state the -stream is in, but nor do you need to -- it can be inferred from the -values returned by BZ2_bzCompress. -
 
IDLE/any           
-      Illegal.  IDLE state only exists after BZ2_bzCompressEnd or
-      before BZ2_bzCompressInit.
-      Return value = BZ_SEQUENCE_ERROR
-
-RUNNING/BZ_RUN     
-      Compress from next_in to next_out as much as possible.
-      Next state = RUNNING
-      Return value = BZ_RUN_OK
-
-RUNNING/BZ_FLUSH   
-      Remember current value of next_in.  Compress from next_in
-      to next_out as much as possible, but do not accept any more input.  
-      Next state = FLUSHING
-      Return value = BZ_FLUSH_OK
-
-RUNNING/BZ_FINISH  
-      Remember current value of next_in.  Compress from next_in
-      to next_out as much as possible, but do not accept any more input.
-      Next state = FINISHING
-      Return value = BZ_FINISH_OK
-
-FLUSHING/BZ_FLUSH  
-      Compress from next_in to next_out as much as possible, 
-      but do not accept any more input.  
-      If all the existing input has been used up and all compressed
-      output has been removed
-         Next state = RUNNING; Return value = BZ_RUN_OK
-      else
-         Next state = FLUSHING; Return value = BZ_FLUSH_OK
-
-FLUSHING/other     
-      Illegal.
-      Return value = BZ_SEQUENCE_ERROR
-
-FINISHING/BZ_FINISH  
-      Compress from next_in to next_out as much as possible,
-      but to not accept any more input.  
-      If all the existing input has been used up and all compressed
-      output has been removed
-         Next state = IDLE; Return value = BZ_STREAM_END
-      else
-         Next state = FINISHING; Return value = BZ_FINISHING
-
-FINISHING/other
-      Illegal.
-      Return value = BZ_SEQUENCE_ERROR
-

- -That still looks complicated? Well, fair enough. The usual sequence -of calls for compressing a load of data is: -

    -
  • Get started with BZ2_bzCompressInit. -
  • Shovel data in and shlurp out its compressed form using zero or more -calls of BZ2_bzCompress with action = BZ_RUN. -
  • Finish up. -Repeatedly call BZ2_bzCompress with action = BZ_FINISH, -copying out the compressed output, until BZ_STREAM_END is returned. -
  • Close up and go home. Call BZ2_bzCompressEnd. -
-If the data you want to compress fits into your input buffer all -at once, you can skip the calls of BZ2_bzCompress ( ..., BZ_RUN ) and -just do the BZ2_bzCompress ( ..., BZ_FINISH ) calls. -

- -All required memory is allocated by BZ2_bzCompressInit. The -compression library can accept any data at all (obviously). So you -shouldn't get any error return values from the BZ2_bzCompress calls. -If you do, they will be BZ_SEQUENCE_ERROR, and indicate a bug in -your programming. -

- -Trivial other possible return values: -
 
      BZ_PARAM_ERROR   
-         if strm is NULL, or strm->s is NULL
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3.3 BZ2_bzCompressEnd

- -
 
int BZ2_bzCompressEnd ( bz_stream *strm );
-
Releases all memory associated with a compression stream. -

- -Possible return values: -
 
   BZ_PARAM_ERROR    if strm is NULL or strm->s is NULL
-   BZ_OK    otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3.4 BZ2_bzDecompressInit

- -
 
int BZ2_bzDecompressInit ( bz_stream *strm, int verbosity, int small );
-
Prepares for decompression. As with BZ2_bzCompressInit, a -bz_stream record should be allocated and initialised before the -call. Fields bzalloc, bzfree and opaque should be -set if a custom memory allocator is required, or made NULL for -the normal malloc/free routines. Upon return, the internal -state will have been initialised, and total_in and -total_out will be zero. -

- -For the meaning of parameter verbosity, see BZ2_bzCompressInit. -

- -If small is nonzero, the library will use an alternative -decompression algorithm which uses less memory but at the cost of -decompressing more slowly (roughly speaking, half the speed, but the -maximum memory requirement drops to around 2300k). See Chapter 2 for -more information on memory management. -

- -Note that the amount of memory needed to decompress -a stream cannot be determined until the stream's header has been read, -so even if BZ2_bzDecompressInit succeeds, a subsequent -BZ2_bzDecompress could fail with BZ_MEM_ERROR. -

- -Possible return values: -
 
      BZ_CONFIG_ERROR
-         if the library has been mis-compiled
-      BZ_PARAM_ERROR
-         if (small != 0 && small != 1)
-         or (verbosity < 0 || verbosity > 4)
-      BZ_MEM_ERROR
-         if insufficient memory is available
-

- -Allowable next actions: -
 
      BZ2_bzDecompress
-         if BZ_OK was returned
-      no specific action required in case of error
-

- - -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3.5 BZ2_bzDecompress

- -
 
int BZ2_bzDecompress ( bz_stream *strm );
-
Provides more input and/out output buffer space for the library. The -caller maintains input and output buffers, and uses BZ2_bzDecompress -to transfer data between them. -

- -Before each call to BZ2_bzDecompress, next_in -should point at the compressed data, -and avail_in should indicate how many bytes the library -may read. BZ2_bzDecompress updates next_in, avail_in -and total_in -to reflect the number of bytes it has read. -

- -Similarly, next_out should point to a buffer in which the uncompressed -output is to be placed, with avail_out indicating how much output space -is available. BZ2_bzCompress updates next_out, -avail_out and total_out to reflect -the number of bytes output. -

- -You may provide and remove as little or as much data as you like on -each call of BZ2_bzDecompress. -In the limit, it is acceptable to -supply and remove data one byte at a time, although this would be -terribly inefficient. You should always ensure that at least one -byte of output space is available at each call. -

- -Use of BZ2_bzDecompress is simpler than BZ2_bzCompress. -

- -You should provide input and remove output as described above, and -repeatedly call BZ2_bzDecompress until BZ_STREAM_END is -returned. Appearance of BZ_STREAM_END denotes that -BZ2_bzDecompress has detected the logical end of the compressed -stream. BZ2_bzDecompress will not produce BZ_STREAM_END until -all output data has been placed into the output buffer, so once -BZ_STREAM_END appears, you are guaranteed to have available all -the decompressed output, and BZ2_bzDecompressEnd can safely be -called. -

- -If case of an error return value, you should call BZ2_bzDecompressEnd -to clean up and release memory. -

- -Possible return values: -
 
      BZ_PARAM_ERROR
-         if strm is NULL or strm->s is NULL
-         or strm->avail_out < 1
-      BZ_DATA_ERROR
-         if a data integrity error is detected in the compressed stream
-      BZ_DATA_ERROR_MAGIC
-         if the compressed stream doesn't begin with the right magic bytes
-      BZ_MEM_ERROR
-         if there wasn't enough memory available
-      BZ_STREAM_END
-         if the logical end of the data stream was detected and all
-         output in has been consumed, eg s->avail_out > 0
-      BZ_OK
-         otherwise
-
Allowable next actions: -
 
      BZ2_bzDecompress
-         if BZ_OK was returned
-      BZ2_bzDecompressEnd
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.3.6 BZ2_bzDecompressEnd

- -
 
int BZ2_bzDecompressEnd ( bz_stream *strm );
-
Releases all memory associated with a decompression stream. -

- -Possible return values: -
 
      BZ_PARAM_ERROR
-         if strm is NULL or strm->s is NULL
-      BZ_OK
-         otherwise
-

- -Allowable next actions: -
 
      None.
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4 High-level interface

- -

- -This interface provides functions for reading and writing -bzip2 format files. First, some general points. -

- -

    -
  • All of the functions take an int* first argument, - bzerror. - After each call, bzerror should be consulted first to determine - the outcome of the call. If bzerror is BZ_OK, - the call completed - successfully, and only then should the return value of the function - (if any) be consulted. If bzerror is BZ_IO_ERROR, - there was an error - reading/writing the underlying compressed file, and you should - then consult errno/perror to determine the - cause of the difficulty. - bzerror may also be set to various other values; precise details are - given on a per-function basis below. -
  • If bzerror indicates an error - (ie, anything except BZ_OK and BZ_STREAM_END), - you should immediately call BZ2_bzReadClose (or BZ2_bzWriteClose, - depending on whether you are attempting to read or to write) - to free up all resources associated - with the stream. Once an error has been indicated, behaviour of all calls - except BZ2_bzReadClose (BZ2_bzWriteClose) is undefined. - The implication is that (1) bzerror should - be checked after each call, and (2) if bzerror indicates an error, - BZ2_bzReadClose (BZ2_bzWriteClose) should then be called to clean up. -
  • The FILE* arguments passed to - BZ2_bzReadOpen/BZ2_bzWriteOpen - should be set to binary mode. - Most Unix systems will do this by default, but other platforms, - including Windows and Mac, will not. If you omit this, you may - encounter problems when moving code to new platforms. -
  • Memory allocation requests are handled by - malloc/free. - At present - there is no facility for user-defined memory allocators in the file I/O - functions (could easily be added, though). -
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.1 BZ2_bzReadOpen

- -
 
   typedef void BZFILE;
-
-   BZFILE *BZ2_bzReadOpen ( int *bzerror, FILE *f, 
-                            int small, int verbosity,
-                            void *unused, int nUnused );
-
Prepare to read compressed data from file handle f. f -should refer to a file which has been opened for reading, and for which -the error indicator (ferror(f))is not set. If small is 1, -the library will try to decompress using less memory, at the expense of -speed. -

- -For reasons explained below, BZ2_bzRead will decompress the -nUnused bytes starting at unused, before starting to read -from the file f. At most BZ_MAX_UNUSED bytes may be -supplied like this. If this facility is not required, you should pass -NULL and 0 for unused and nUnused -respectively. -

- -For the meaning of parameters small and verbosity, -see BZ2_bzDecompressInit. -

- -The amount of memory needed to decompress a file cannot be determined -until the file's header has been read. So it is possible that -BZ2_bzReadOpen returns BZ_OK but a subsequent call of -BZ2_bzRead will return BZ_MEM_ERROR. -

- -Possible assignments to bzerror: -
 
      BZ_CONFIG_ERROR
-         if the library has been mis-compiled
-      BZ_PARAM_ERROR
-         if f is NULL 
-         or small is neither 0 nor 1                 
-         or (unused == NULL && nUnused != 0)
-         or (unused != NULL && !(0 <= nUnused <= BZ_MAX_UNUSED))
-      BZ_IO_ERROR    
-         if ferror(f) is nonzero
-      BZ_MEM_ERROR   
-         if insufficient memory is available
-      BZ_OK
-         otherwise.
-

- -Possible return values: -
 
      Pointer to an abstract BZFILE        
-         if bzerror is BZ_OK   
-      NULL
-         otherwise
-

- -Allowable next actions: -
 
      BZ2_bzRead
-         if bzerror is BZ_OK   
-      BZ2_bzClose 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.2 BZ2_bzRead

- -
 
   int BZ2_bzRead ( int *bzerror, BZFILE *b, void *buf, int len );
-
Reads up to len (uncompressed) bytes from the compressed file -b into -the buffer buf. If the read was successful, -bzerror is set to BZ_OK -and the number of bytes read is returned. If the logical end-of-stream -was detected, bzerror will be set to BZ_STREAM_END, -and the number -of bytes read is returned. All other bzerror values denote an error. -

- -BZ2_bzRead will supply len bytes, -unless the logical stream end is detected -or an error occurs. Because of this, it is possible to detect the -stream end by observing when the number of bytes returned is -less than the number -requested. Nevertheless, this is regarded as inadvisable; you should -instead check bzerror after every call and watch out for -BZ_STREAM_END. -

- -Internally, BZ2_bzRead copies data from the compressed file in chunks -of size BZ_MAX_UNUSED bytes -before decompressing it. If the file contains more bytes than strictly -needed to reach the logical end-of-stream, BZ2_bzRead will almost certainly -read some of the trailing data before signalling BZ_SEQUENCE_END. -To collect the read but unused data once BZ_SEQUENCE_END has -appeared, call BZ2_bzReadGetUnused immediately before BZ2_bzReadClose. -

- -Possible assignments to bzerror: -
 
      BZ_PARAM_ERROR
-         if b is NULL or buf is NULL or len < 0
-      BZ_SEQUENCE_ERROR 
-         if b was opened with BZ2_bzWriteOpen
-      BZ_IO_ERROR 
-         if there is an error reading from the compressed file
-      BZ_UNEXPECTED_EOF 
-         if the compressed file ended before the logical end-of-stream was detected
-      BZ_DATA_ERROR 
-         if a data integrity error was detected in the compressed stream
-      BZ_DATA_ERROR_MAGIC
-         if the stream does not begin with the requisite header bytes (ie, is not 
-         a bzip2 data file).  This is really a special case of BZ_DATA_ERROR.
-      BZ_MEM_ERROR 
-         if insufficient memory was available
-      BZ_STREAM_END 
-         if the logical end of stream was detected.
-      BZ_OK
-         otherwise.
-

- -Possible return values: -
 
      number of bytes read
-         if bzerror is BZ_OK or BZ_STREAM_END
-      undefined
-         otherwise
-

- -Allowable next actions: -
 
      collect data from buf, then BZ2_bzRead or BZ2_bzReadClose
-         if bzerror is BZ_OK 
-      collect data from buf, then BZ2_bzReadClose or BZ2_bzReadGetUnused 
-         if bzerror is BZ_SEQUENCE_END   
-      BZ2_bzReadClose 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.3 BZ2_bzReadGetUnused

- -
 
   void BZ2_bzReadGetUnused ( int* bzerror, BZFILE *b, 
-                              void** unused, int* nUnused );
-
Returns data which was read from the compressed file but was not needed -to get to the logical end-of-stream. *unused is set to the address -of the data, and *nUnused to the number of bytes. *nUnused will -be set to a value between 0 and BZ_MAX_UNUSED inclusive. -

- -This function may only be called once BZ2_bzRead has signalled -BZ_STREAM_END but before BZ2_bzReadClose. -

- -Possible assignments to bzerror: -
 
      BZ_PARAM_ERROR 
-         if b is NULL 
-         or unused is NULL or nUnused is NULL
-      BZ_SEQUENCE_ERROR 
-         if BZ_STREAM_END has not been signalled
-         or if b was opened with BZ2_bzWriteOpen
-     BZ_OK
-         otherwise
-

- -Allowable next actions: -
 
      BZ2_bzReadClose
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.4 BZ2_bzReadClose

- -
 
   void BZ2_bzReadClose ( int *bzerror, BZFILE *b );
-
Releases all memory pertaining to the compressed file b. -BZ2_bzReadClose does not call fclose on the underlying file -handle, so you should do that yourself if appropriate. -BZ2_bzReadClose should be called to clean up after all error -situations. -

- -Possible assignments to bzerror: -
 
      BZ_SEQUENCE_ERROR 
-         if b was opened with BZ2_bzOpenWrite 
-      BZ_OK 
-         otherwise
-

- -Allowable next actions: -
 
      none
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.5 BZ2_bzWriteOpen

- -
 
   BZFILE *BZ2_bzWriteOpen ( int *bzerror, FILE *f, 
-                             int blockSize100k, int verbosity,
-                             int workFactor );
-
Prepare to write compressed data to file handle f. -f should refer to -a file which has been opened for writing, and for which the error -indicator (ferror(f))is not set. -

- -For the meaning of parameters blockSize100k, -verbosity and workFactor, see -
BZ2_bzCompressInit. -

- -All required memory is allocated at this stage, so if the call -completes successfully, BZ_MEM_ERROR cannot be signalled by a -subsequent call to BZ2_bzWrite. -

- -Possible assignments to bzerror: -
 
      BZ_CONFIG_ERROR
-         if the library has been mis-compiled
-      BZ_PARAM_ERROR 
-         if f is NULL 
-         or blockSize100k < 1 or blockSize100k > 9
-      BZ_IO_ERROR 
-         if ferror(f) is nonzero
-      BZ_MEM_ERROR 
-         if insufficient memory is available
-      BZ_OK 
-         otherwise
-

- -Possible return values: -
 
      Pointer to an abstract BZFILE  
-         if bzerror is BZ_OK   
-      NULL 
-         otherwise
-

- -Allowable next actions: -
 
      BZ2_bzWrite 
-         if bzerror is BZ_OK 
-         (you could go directly to BZ2_bzWriteClose, but this would be pretty pointless)
-      BZ2_bzWriteClose 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.6 BZ2_bzWrite

- -
 
   void BZ2_bzWrite ( int *bzerror, BZFILE *b, void *buf, int len );
-
Absorbs len bytes from the buffer buf, eventually to be -compressed and written to the file. -

- -Possible assignments to bzerror: -
 
      BZ_PARAM_ERROR 
-         if b is NULL or buf is NULL or len < 0
-      BZ_SEQUENCE_ERROR 
-         if b was opened with BZ2_bzReadOpen
-      BZ_IO_ERROR 
-         if there is an error writing the compressed file.
-      BZ_OK 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.7 BZ2_bzWriteClose

- -
 
   void BZ2_bzWriteClose ( int *bzerror, BZFILE* f,
-                           int abandon,
-                           unsigned int* nbytes_in,
-                           unsigned int* nbytes_out );
-
-   void BZ2_bzWriteClose64 ( int *bzerror, BZFILE* f,
-                             int abandon,
-                             unsigned int* nbytes_in_lo32,
-                             unsigned int* nbytes_in_hi32,
-                             unsigned int* nbytes_out_lo32,
-                             unsigned int* nbytes_out_hi32 );
-

- -Compresses and flushes to the compressed file all data so far supplied -by BZ2_bzWrite. The logical end-of-stream markers are also written, so -subsequent calls to BZ2_bzWrite are illegal. All memory associated -with the compressed file b is released. -fflush is called on the -compressed file, but it is not fclose'd. -

- -If BZ2_bzWriteClose is called to clean up after an error, the only -action is to release the memory. The library records the error codes -issued by previous calls, so this situation will be detected -automatically. There is no attempt to complete the compression -operation, nor to fflush the compressed file. You can force this -behaviour to happen even in the case of no error, by passing a nonzero -value to abandon. -

- -If nbytes_in is non-null, *nbytes_in will be set to be the -total volume of uncompressed data handled. Similarly, nbytes_out -will be set to the total volume of compressed data written. For -compatibility with older versions of the library, BZ2_bzWriteClose -only yields the lower 32 bits of these counts. Use -BZ2_bzWriteClose64 if you want the full 64 bit counts. These -two functions are otherwise absolutely identical. -

- -Possible assignments to bzerror: -
 
      BZ_SEQUENCE_ERROR 
-         if b was opened with BZ2_bzReadOpen
-      BZ_IO_ERROR 
-         if there is an error writing the compressed file
-      BZ_OK 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.8 Handling embedded compressed data streams

- -

- -The high-level library facilitates use of -bzip2 data streams which form some part of a surrounding, larger -data stream. -

    -
  • For writing, the library takes an open file handle, writes -compressed data to it, fflushes it but does not fclose it. -The calling application can write its own data before and after the -compressed data stream, using that same file handle. -
  • Reading is more complex, and the facilities are not as general -as they could be since generality is hard to reconcile with efficiency. -BZ2_bzRead reads from the compressed file in blocks of size -BZ_MAX_UNUSED bytes, and in doing so probably will overshoot -the logical end of compressed stream. -To recover this data once decompression has -ended, call BZ2_bzReadGetUnused after the last call of BZ2_bzRead -(the one returning BZ_STREAM_END) but before calling -BZ2_bzReadClose. -
-

- -This mechanism makes it easy to decompress multiple bzip2 -streams placed end-to-end. As the end of one stream, when BZ2_bzRead -returns BZ_STREAM_END, call BZ2_bzReadGetUnused to collect the -unused data (copy it into your own buffer somewhere). -That data forms the start of the next compressed stream. -To start uncompressing that next stream, call BZ2_bzReadOpen again, -feeding in the unused data via the unused/nUnused -parameters. -Keep doing this until BZ_STREAM_END return coincides with the -physical end of file (feof(f)). In this situation -BZ2_bzReadGetUnused -will of course return no data. -

- -This should give some feel for how the high-level interface can be used. -If you require extra flexibility, you'll have to bite the bullet and get -to grips with the low-level interface. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.4.9 Standard file-reading/writing code

- -Here's how you'd write data to a compressed file: -
 
FILE*   f;
-BZFILE* b;
-int     nBuf;
-char    buf[ /* whatever size you like */ ];
-int     bzerror;
-int     nWritten;
-
-f = fopen ( "myfile.bz2", "w" );
-if (!f) {
-   /* handle error */
-}
-b = BZ2_bzWriteOpen ( &bzerror, f, 9 );
-if (bzerror != BZ_OK) {
-   BZ2_bzWriteClose ( b );
-   /* handle error */
-}
-
-while ( /* condition */ ) {
-   /* get data to write into buf, and set nBuf appropriately */
-   nWritten = BZ2_bzWrite ( &bzerror, b, buf, nBuf );
-   if (bzerror == BZ_IO_ERROR) { 
-      BZ2_bzWriteClose ( &bzerror, b );
-      /* handle error */
-   }
-}
-
-BZ2_bzWriteClose ( &bzerror, b );
-if (bzerror == BZ_IO_ERROR) {
-   /* handle error */
-}
-
And to read from a compressed file: -
 
FILE*   f;
-BZFILE* b;
-int     nBuf;
-char    buf[ /* whatever size you like */ ];
-int     bzerror;
-int     nWritten;
-
-f = fopen ( "myfile.bz2", "r" );
-if (!f) {
-   /* handle error */
-}
-b = BZ2_bzReadOpen ( &bzerror, f, 0, NULL, 0 );
-if (bzerror != BZ_OK) {
-   BZ2_bzReadClose ( &bzerror, b );
-   /* handle error */
-}
-
-bzerror = BZ_OK;
-while (bzerror == BZ_OK && /* arbitrary other conditions */) {
-   nBuf = BZ2_bzRead ( &bzerror, b, buf, /* size of buf */ );
-   if (bzerror == BZ_OK) {
-      /* do something with buf[0 .. nBuf-1] */
-   }
-}
-if (bzerror != BZ_STREAM_END) {
-   BZ2_bzReadClose ( &bzerror, b );
-   /* handle error */
-} else {
-   BZ2_bzReadClose ( &bzerror );
-}
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.5 Utility functions

- -
- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.5.1 BZ2_bzBuffToBuffCompress

- -
 
   int BZ2_bzBuffToBuffCompress( char*         dest,
-                                 unsigned int* destLen,
-                                 char*         source,
-                                 unsigned int  sourceLen,
-                                 int           blockSize100k,
-                                 int           verbosity,
-                                 int           workFactor );
-
Attempts to compress the data in source[0 .. sourceLen-1] -into the destination buffer, dest[0 .. *destLen-1]. -If the destination buffer is big enough, *destLen is -set to the size of the compressed data, and BZ_OK is -returned. If the compressed data won't fit, *destLen -is unchanged, and BZ_OUTBUFF_FULL is returned. -

- -Compression in this manner is a one-shot event, done with a single call -to this function. The resulting compressed data is a complete -bzip2 format data stream. There is no mechanism for making -additional calls to provide extra input data. If you want that kind of -mechanism, use the low-level interface. -

- -For the meaning of parameters blockSize100k, verbosity -and workFactor,
see BZ2_bzCompressInit. -

- -To guarantee that the compressed data will fit in its buffer, allocate -an output buffer of size 1% larger than the uncompressed data, plus -six hundred extra bytes. -

- -BZ2_bzBuffToBuffDecompress will not write data at or -beyond dest[*destLen], even in case of buffer overflow. -

- -Possible return values: -
 
      BZ_CONFIG_ERROR
-         if the library has been mis-compiled
-      BZ_PARAM_ERROR 
-         if dest is NULL or destLen is NULL
-         or blockSize100k < 1 or blockSize100k > 9
-         or verbosity < 0 or verbosity > 4 
-         or workFactor < 0 or workFactor > 250
-      BZ_MEM_ERROR
-         if insufficient memory is available 
-      BZ_OUTBUFF_FULL
-         if the size of the compressed data exceeds *destLen
-      BZ_OK 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.5.2 BZ2_bzBuffToBuffDecompress

- -
 
   int BZ2_bzBuffToBuffDecompress ( char*         dest,
-                                    unsigned int* destLen,
-                                    char*         source,
-                                    unsigned int  sourceLen,
-                                    int           small,
-                                    int           verbosity );
-
Attempts to decompress the data in source[0 .. sourceLen-1] -into the destination buffer, dest[0 .. *destLen-1]. -If the destination buffer is big enough, *destLen is -set to the size of the uncompressed data, and BZ_OK is -returned. If the compressed data won't fit, *destLen -is unchanged, and BZ_OUTBUFF_FULL is returned. -

- -source is assumed to hold a complete bzip2 format -data stream.
BZ2_bzBuffToBuffDecompress tries to decompress -the entirety of the stream into the output buffer. -

- -For the meaning of parameters small and verbosity, -see BZ2_bzDecompressInit. -

- -Because the compression ratio of the compressed data cannot be known in -advance, there is no easy way to guarantee that the output buffer will -be big enough. You may of course make arrangements in your code to -record the size of the uncompressed data, but such a mechanism is beyond -the scope of this library. -

- -BZ2_bzBuffToBuffDecompress will not write data at or -beyond dest[*destLen], even in case of buffer overflow. -

- -Possible return values: -
 
      BZ_CONFIG_ERROR
-         if the library has been mis-compiled
-      BZ_PARAM_ERROR 
-         if dest is NULL or destLen is NULL
-         or small != 0 && small != 1
-         or verbosity < 0 or verbosity > 4 
-      BZ_MEM_ERROR
-         if insufficient memory is available 
-      BZ_OUTBUFF_FULL
-         if the size of the compressed data exceeds *destLen
-      BZ_DATA_ERROR
-         if a data integrity error was detected in the compressed data
-      BZ_DATA_ERROR_MAGIC
-         if the compressed data doesn't begin with the right magic bytes
-      BZ_UNEXPECTED_EOF
-         if the compressed data ends unexpectedly
-      BZ_OK 
-         otherwise
-

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.6 zlib compatibility functions

- -Yoshioka Tsuneo has contributed some functions to -give better zlib compatibility. These functions are -BZ2_bzopen, BZ2_bzread, BZ2_bzwrite, BZ2_bzflush, -BZ2_bzclose, -BZ2_bzerror and BZ2_bzlibVersion. -These functions are not (yet) officially part of -the library. If they break, you get to keep all the pieces. -Nevertheless, I think they work ok. -
 
typedef void BZFILE;
-
-const char * BZ2_bzlibVersion ( void );
-
Returns a string indicating the library version. -
 
BZFILE * BZ2_bzopen  ( const char *path, const char *mode );
-BZFILE * BZ2_bzdopen ( int        fd,    const char *mode );
-
Opens a .bz2 file for reading or writing, using either its name -or a pre-existing file descriptor. -Analogous to fopen and fdopen. -
 
int BZ2_bzread  ( BZFILE* b, void* buf, int len );
-int BZ2_bzwrite ( BZFILE* b, void* buf, int len );
-
Reads/writes data from/to a previously opened BZFILE. -Analogous to fread and fwrite. -
 
int  BZ2_bzflush ( BZFILE* b );
-void BZ2_bzclose ( BZFILE* b );
-
Flushes/closes a BZFILE. BZ2_bzflush doesn't actually do -anything. Analogous to fflush and fclose. -

- -
 
const char * BZ2_bzerror ( BZFILE *b, int *errnum )
-
Returns a string describing the more recent error status of -b, and also sets *errnum to its numerical value. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.7 Using the library in a stdio-free environment

- -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.7.1 Getting rid of stdio

- -

- -In a deeply embedded application, you might want to use just -the memory-to-memory functions. You can do this conveniently -by compiling the library with preprocessor symbol BZ_NO_STDIO -defined. Doing this gives you a library containing only the following -eight functions: -

- -BZ2_bzCompressInit, BZ2_bzCompress, BZ2_bzCompressEnd
-BZ2_bzDecompressInit, BZ2_bzDecompress, BZ2_bzDecompressEnd
-BZ2_bzBuffToBuffCompress, BZ2_bzBuffToBuffDecompress -

- -When compiled like this, all functions will ignore verbosity -settings. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.7.2 Critical error handling

- -libbzip2 contains a number of internal assertion checks which -should, needless to say, never be activated. Nevertheless, if an -assertion should fail, behaviour depends on whether or not the library -was compiled with BZ_NO_STDIO set. -

- -For a normal compile, an assertion failure yields the message -
 
   bzip2/libbzip2: internal error number N.
-   This is a bug in bzip2/libbzip2, 1.0.2, 30-Dec-2001.
-   Please report it to me at: jseward@acm.org.  If this happened
-   when you were using some program which uses libbzip2 as a
-   component, you should also report this bug to the author(s)
-   of that program.  Please make an effort to report this bug;
-   timely and accurate bug reports eventually lead to higher
-   quality software.  Thanks.  Julian Seward, 30 December 2001.
-
where N is some error code number. If N == 1007, it also -prints some extra text advising the reader that unreliable memory is -often associated with internal error 1007. (This is a -frequently-observed-phenomenon with versions 1.0.0/1.0.1). -

- -exit(3) is then called. -

- -For a stdio-free library, assertion failures result -in a call to a function declared as: -
 
   extern void bz_internal_error ( int errcode );
-
The relevant code is passed as a parameter. You should supply -such a function. -

- -In either case, once an assertion failure has occurred, any -bz_stream records involved can be regarded as invalid. -You should not attempt to resume normal operation with them. -

- -You may, of course, change critical error handling to suit -your needs. As I said above, critical errors indicate bugs -in the library and should not occur. All "normal" error -situations are indicated via error return codes from functions, -and can be recovered from. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

3.8 Making a Windows DLL

- -Everything related to Windows has been contributed by Yoshioka Tsuneo -
(QWF00133@niftyserve.or.jp / -tsuneo-y@is.aist-nara.ac.jp), so you should send your queries to -him (but perhaps Cc: me, jseward@acm.org). -

- -My vague understanding of what to do is: using Visual C++ 5.0, -open the project file libbz2.dsp, and build. That's all. -

- -If you can't -open the project file for some reason, make a new one, naming these files: -blocksort.c, bzlib.c, compress.c, -crctable.c, decompress.c, huffman.c,
-randtable.c and libbz2.def. You will also need -to name the header files bzlib.h and bzlib_private.h. -

- -If you don't use VC++, you may need to define the proprocessor symbol -_WIN32. -

- -Finally, dlltest.c is a sample program using the DLL. It has a -project file, dlltest.dsp. -

- -If you just want a makefile for Visual C, have a look at -makefile.msc. -

- -Be aware that if you compile bzip2 itself on Win32, you must set -BZ_UNIX to 0 and BZ_LCCWIN32 to 1, in the file -bzip2.c, before compiling. Otherwise the resulting binary won't -work correctly. -

- -I haven't tried any of this stuff myself, but it all looks plausible. -

- -

- - - - - - - -
[ << ][ >> ]           [Top][Contents][Index][ ? ]
-
- -This document was generated -by Julian Seward on January, 5 2002 -using texi2html - - - diff -Nru bzip2-1.0.2/manual_4.html bzip2-1.0.2.new/manual_4.html --- bzip2-1.0.2/manual_4.html Sat Jan 5 01:53:05 2002 +++ bzip2-1.0.2.new/manual_4.html Thu Jan 1 01:00:00 1970 @@ -1,530 +0,0 @@ - - - - - -Untitled Document: 4. Miscellanea - - - - - - - - - - - - - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4. Miscellanea

- -

- -These are just some random thoughts of mine. Your mileage may -vary. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4.1 Limitations of the compressed file format

- -bzip2-1.0, 0.9.5 and 0.9.0 -use exactly the same file format as the previous -version, bzip2-0.1. This decision was made in the interests of -stability. Creating yet another incompatible compressed file format -would create further confusion and disruption for users. -

- -Nevertheless, this is not a painless decision. Development -work since the release of bzip2-0.1 in August 1997 -has shown complexities in the file format which slow down -decompression and, in retrospect, are unnecessary. These are: -

    -
  • The run-length encoder, which is the first of the - compression transformations, is entirely irrelevant. - The original purpose was to protect the sorting algorithm - from the very worst case input: a string of repeated - symbols. But algorithm steps Q6a and Q6b in the original - Burrows-Wheeler technical report (SRC-124) show how - repeats can be handled without difficulty in block - sorting. -
  • The randomisation mechanism doesn't really need to be - there. Udi Manber and Gene Myers published a suffix - array construction algorithm a few years back, which - can be employed to sort any block, no matter how - repetitive, in O(N log N) time. Subsequent work by - Kunihiko Sadakane has produced a derivative O(N (log N)^2) - algorithm which usually outperforms the Manber-Myers - algorithm. -

    - - I could have changed to Sadakane's algorithm, but I find - it to be slower than bzip2's existing algorithm for - most inputs, and the randomisation mechanism protects - adequately against bad cases. I didn't think it was - a good tradeoff to make. Partly this is due to the fact - that I was not flooded with email complaints about - bzip2-0.1's performance on repetitive data, so - perhaps it isn't a problem for real inputs. -

    - - Probably the best long-term solution, - and the one I have incorporated into 0.9.5 and above, - is to use the existing sorting - algorithm initially, and fall back to a O(N (log N)^2) - algorithm if the standard algorithm gets into difficulties. -

  • The compressed file format was never designed to be - handled by a library, and I have had to jump though - some hoops to produce an efficient implementation of - decompression. It's a bit hairy. Try passing - decompress.c through the C preprocessor - and you'll see what I mean. Much of this complexity - could have been avoided if the compressed size of - each block of data was recorded in the data stream. -
  • An Adler-32 checksum, rather than a CRC32 checksum, - would be faster to compute. -
-It would be fair to say that the bzip2 format was frozen -before I properly and fully understood the performance -consequences of doing so. -

- -Improvements which I was able to incorporate into -0.9.0, despite using the same file format, are: -

    -
  • Single array implementation of the inverse BWT. This - significantly speeds up decompression, presumably - because it reduces the number of cache misses. -
  • Faster inverse MTF transform for large MTF values. The - new implementation is based on the notion of sliding blocks - of values. -
  • bzip2-0.9.0 now reads and writes files with fread - and fwrite; version 0.1 used putc and getc. - Duh! Well, you live and learn. -

    - -

-Further ahead, it would be nice -to be able to do random access into files. This will -require some careful design of compressed file formats. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4.2 Portability issues

- -After some consideration, I have decided not to use -GNU autoconf to configure 0.9.5 or 1.0. -

- -autoconf, admirable and wonderful though it is, -mainly assists with portability problems between Unix-like -platforms. But bzip2 doesn't have much in the way -of portability problems on Unix; most of the difficulties appear -when porting to the Mac, or to Microsoft's operating systems. -autoconf doesn't help in those cases, and brings in a -whole load of new complexity. -

- -Most people should be able to compile the library and program -under Unix straight out-of-the-box, so to speak, especially -if you have a version of GNU C available. -

- -There are a couple of __inline__ directives in the code. GNU C -(gcc) should be able to handle them. If you're not using -GNU C, your C compiler shouldn't see them at all. -If your compiler does, for some reason, see them and doesn't -like them, just #define __inline__ to be /* */. One -easy way to do this is to compile with the flag -D__inline__=, -which should be understood by most Unix compilers. -

- -If you still have difficulties, try compiling with the macro -BZ_STRICT_ANSI defined. This should enable you to build the -library in a strictly ANSI compliant environment. Building the program -itself like this is dangerous and not supported, since you remove -bzip2's checks against compressing directories, symbolic links, -devices, and other not-really-a-file entities. This could cause -filesystem corruption! -

- -One other thing: if you create a bzip2 binary for public -distribution, please try and link it statically (gcc -s). This -avoids all sorts of library-version issues that others may encounter -later on. -

- -If you build bzip2 on Win32, you must set BZ_UNIX to 0 and -BZ_LCCWIN32 to 1, in the file bzip2.c, before compiling. -Otherwise the resulting binary won't work correctly. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4.3 Reporting bugs

- -I tried pretty hard to make sure bzip2 is -bug free, both by design and by testing. Hopefully -you'll never need to read this section for real. -

- -Nevertheless, if bzip2 dies with a segmentation -fault, a bus error or an internal assertion failure, it -will ask you to email me a bug report. Experience with -version 0.1 shows that almost all these problems can -be traced to either compiler bugs or hardware problems. -

    -
  • -Recompile the program with no optimisation, and see if it -works. And/or try a different compiler. -I heard all sorts of stories about various flavours -of GNU C (and other compilers) generating bad code for -bzip2, and I've run across two such examples myself. -

    - -2.7.X versions of GNU C are known to generate bad code from -time to time, at high optimisation levels. -If you get problems, try using the flags --O2 -fomit-frame-pointer -fno-strength-reduce. -You should specifically not use -funroll-loops. -

    - -You may notice that the Makefile runs six tests as part of -the build process. If the program passes all of these, it's -a pretty good (but not 100%) indication that the compiler has -done its job correctly. -

  • -If bzip2 crashes randomly, and the crashes are not -repeatable, you may have a flaky memory subsystem. bzip2 -really hammers your memory hierarchy, and if it's a bit marginal, -you may get these problems. Ditto if your disk or I/O subsystem -is slowly failing. Yup, this really does happen. -

    - -Try using a different machine of the same type, and see if -you can repeat the problem. -

  • This isn't really a bug, but ... If bzip2 tells -you your file is corrupted on decompression, and you -obtained the file via FTP, there is a possibility that you -forgot to tell FTP to do a binary mode transfer. That absolutely -will cause the file to be non-decompressible. You'll have to transfer -it again. -
-

- -If you've incorporated libbzip2 into your own program -and are getting problems, please, please, please, check that the -parameters you are passing in calls to the library, are -correct, and in accordance with what the documentation says -is allowable. I have tried to make the library robust against -such problems, but I'm sure I haven't succeeded. -

- -Finally, if the above comments don't help, you'll have to send -me a bug report. Now, it's just amazing how many people will -send me a bug report saying something like -
 
   bzip2 crashed with segmentation fault on my machine
-
and absolutely nothing else. Needless to say, a such a report -is totally, utterly, completely and comprehensively 100% useless; -a waste of your time, my time, and net bandwidth. -With no details at all, there's no way I can possibly begin -to figure out what the problem is. -

- -The rules of the game are: facts, facts, facts. Don't omit -them because "oh, they won't be relevant". At the bare -minimum: -
 
   Machine type.  Operating system version.  
-   Exact version of bzip2 (do bzip2 -V).  
-   Exact version of the compiler used.  
-   Flags passed to the compiler.
-
However, the most important single thing that will help me is -the file that you were trying to compress or decompress at the -time the problem happened. Without that, my ability to do anything -more than speculate about the cause, is limited. -

- -Please remember that I connect to the Internet with a modem, so -you should contact me before mailing me huge files. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4.4 Did you get the right package?

- -

- -bzip2 is a resource hog. It soaks up large amounts of CPU cycles -and memory. Also, it gives very large latencies. In the worst case, you -can feed many megabytes of uncompressed data into the library before -getting any compressed output, so this probably rules out applications -requiring interactive behaviour. -

- -These aren't faults of my implementation, I hope, but more -an intrinsic property of the Burrows-Wheeler transform (unfortunately). -Maybe this isn't what you want. -

- -If you want a compressor and/or library which is faster, uses less -memory but gets pretty good compression, and has minimal latency, -consider Jean-loup -Gailly's and Mark Adler's work, zlib-1.1.3 and -gzip-1.2.4. Look for them at -

- -http://www.zlib.org and -http://www.gzip.org respectively. -

- -For something faster and lighter still, you might try Markus F X J -Oberhumer's LZO real-time compression/decompression library, at -
http://wildsau.idv.uni-linz.ac.at/mfx/lzo.html. -

- -If you want to use the bzip2 algorithms to compress small blocks -of data, 64k bytes or smaller, for example on an on-the-fly disk -compressor, you'd be well advised not to use this library. Instead, -I've made a special library tuned for that kind of use. It's part of -e2compr-0.40, an on-the-fly disk compressor for the Linux -ext2 filesystem. Look at -http://www.netspace.net.au/~reiter/e2compr. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4.5 Testing

- -

- -A record of the tests I've done. -

- -First, some data sets: -

    -
  • B: a directory containing 6001 files, one for every length in the - range 0 to 6000 bytes. The files contain random lowercase - letters. 18.7 megabytes. -
  • H: my home directory tree. Documents, source code, mail files, - compressed data. H contains B, and also a directory of - files designed as boundary cases for the sorting; mostly very - repetitive, nasty files. 565 megabytes. -
  • A: directory tree holding various applications built from source: - egcs, gcc-2.8.1, KDE, GTK, Octave, etc. - 2200 megabytes. -
-The tests conducted are as follows. Each test means compressing -(a copy of) each file in the data set, decompressing it and -comparing it against the original. -

- -First, a bunch of tests with block sizes and internal buffer -sizes set very small, -to detect any problems with the -blocking and buffering mechanisms. -This required modifying the source code so as to try to -break it. -

    -
  1. Data set H, with - buffer size of 1 byte, and block size of 23 bytes. -
  2. Data set B, buffer sizes 1 byte, block size 1 byte. -
  3. As (2) but small-mode decompression. -
  4. As (2) with block size 2 bytes. -
  5. As (2) with block size 3 bytes. -
  6. As (2) with block size 4 bytes. -
  7. As (2) with block size 5 bytes. -
  8. As (2) with block size 6 bytes and small-mode decompression. -
  9. H with buffer size of 1 byte, but normal block - size (up to 900000 bytes). -
-Then some tests with unmodified source code. -
    -
  1. H, all settings normal. -
  2. As (1), with small-mode decompress. -
  3. H, compress with flag -1. -
  4. H, compress with flag -s, decompress with flag -s. -
  5. Forwards compatibility: H, bzip2-0.1pl2 compressing, - bzip2-0.9.5 decompressing, all settings normal. -
  6. Backwards compatibility: H, bzip2-0.9.5 compressing, - bzip2-0.1pl2 decompressing, all settings normal. -
  7. Bigger tests: A, all settings normal. -
  8. As (7), using the fallback (Sadakane-like) sorting algorithm. -
  9. As (8), compress with flag -1, decompress with flag - -s. -
  10. H, using the fallback sorting algorithm. -
  11. Forwards compatibility: A, bzip2-0.1pl2 compressing, - bzip2-0.9.5 decompressing, all settings normal. -
  12. Backwards compatibility: A, bzip2-0.9.5 compressing, - bzip2-0.1pl2 decompressing, all settings normal. -
  13. Misc test: about 400 megabytes of .tar files with - bzip2 compiled with Checker (a memory access error - detector, like Purify). -
  14. Misc tests to make sure it builds and runs ok on non-Linux/x86 - platforms. -
-These tests were conducted on a 225 MHz IDT WinChip machine, running -Linux 2.0.36. They represent nearly a week of continuous computation. -All tests completed successfully. -

- -

- - - - - - - - - - - -
[ < ][ > ]   [ << ][ Up ][ >> ]         [Top][Contents][Index][ ? ]
-

4.6 Further reading

- -bzip2 is not research work, in the sense that it doesn't present -any new ideas. Rather, it's an engineering exercise based on existing -ideas. -

- -Four documents describe essentially all the ideas behind bzip2: -
 
Michael Burrows and D. J. Wheeler:
-  "A block-sorting lossless data compression algorithm"
-   10th May 1994. 
-   Digital SRC Research Report 124.
-   ftp://ftp.digital.com/pub/DEC/SRC/research-reports/SRC-124.ps.gz
-   If you have trouble finding it, try searching at the
-   New Zealand Digital Library, http://www.nzdl.org.
-
-Daniel S. Hirschberg and Debra A. LeLewer
-  "Efficient Decoding of Prefix Codes"
-   Communications of the ACM, April 1990, Vol 33, Number 4.
-   You might be able to get an electronic copy of this
-      from the ACM Digital Library.
-
-David J. Wheeler
-   Program bred3.c and accompanying document bred3.ps.
-   This contains the idea behind the multi-table Huffman
-   coding scheme.
-   ftp://ftp.cl.cam.ac.uk/users/djw3/
-
-Jon L. Bentley and Robert Sedgewick
-  "Fast Algorithms for Sorting and Searching Strings"
-   Available from Sedgewick's web page,
-   www.cs.princeton.edu/~rs
-
The following paper gives valuable additional insights into the -algorithm, but is not immediately the basis of any code -used in bzip2. -
 
Peter Fenwick:
-   Block Sorting Text Compression
-   Proceedings of the 19th Australasian Computer Science Conference,
-     Melbourne, Australia.  Jan 31 - Feb 2, 1996.
-   ftp://ftp.cs.auckland.ac.nz/pub/peter-f/ACSC96paper.ps
-
Kunihiko Sadakane's sorting algorithm, mentioned above, -is available from: -
 
http://naomi.is.s.u-tokyo.ac.jp/~sada/papers/Sada98b.ps.gz
-
The Manber-Myers suffix array construction -algorithm is described in a paper -available from: -
 
http://www.cs.arizona.edu/people/gene/PAPERS/suffix.ps
-
Finally, the following paper documents some recent investigations -I made into the performance of sorting algorithms: -
 
Julian Seward:
-   On the Performance of BWT Sorting Algorithms
-   Proceedings of the IEEE Data Compression Conference 2000
-     Snowbird, Utah.  28-30 March 2000.
-

- -

- - - - - - - -
[ << ][ >> ]           [Top][Contents][Index][ ? ]
-
- -This document was generated -by Julian Seward on January, 5 2002 -using texi2html - - - diff -Nru bzip2-1.0.2/manual_toc.html bzip2-1.0.2.new/manual_toc.html --- bzip2-1.0.2/manual_toc.html Sat Jan 5 01:53:05 2002 +++ bzip2-1.0.2.new/manual_toc.html Thu Jan 1 01:00:00 1970 @@ -1,163 +0,0 @@ - - - - - -Untitled Document: Table of Contents - - - - - - - - - - - - - - - - - -
[Top][Contents][Index][ ? ]
-

Table of Contents

- -
-
- -This document was generated -by Julian Seward on January, 5 2002 -using texi2html - - - diff -Nru bzip2-1.0.2/randtable.c bzip2-1.0.2.new/randtable.c --- bzip2-1.0.2/randtable.c Sun Dec 30 03:19:04 2001 +++ bzip2-1.0.2.new/randtable.c Fri Feb 1 04:19:18 2002 @@ -58,6 +58,10 @@ For more information on these sources, see the manual. --*/ +#ifdef HAVE_CONFIG_H +#include +#endif + #include "bzlib_private.h" diff -Nru bzip2-1.0.2/spewG.c bzip2-1.0.2.new/spewG.c --- bzip2-1.0.2/spewG.c Thu Apr 6 01:24:45 2000 +++ bzip2-1.0.2.new/spewG.c Fri Feb 1 04:19:18 2002 @@ -9,7 +9,10 @@ (but is otherwise harmless). */ -#define _FILE_OFFSET_BITS 64 +#ifdef HAVE_CONFIG_H +#include +#endif + #include #include diff -Nru bzip2-1.0.2/stamp-h.in bzip2-1.0.2.new/stamp-h.in --- bzip2-1.0.2/stamp-h.in Thu Jan 1 01:00:00 1970 +++ bzip2-1.0.2.new/stamp-h.in Fri Feb 1 04:19:18 2002 @@ -0,0 +1 @@ +timestamp diff -Nru bzip2-1.0.2/unzcrash.c bzip2-1.0.2.new/unzcrash.c --- bzip2-1.0.2/unzcrash.c Thu Apr 6 01:24:46 2000 +++ bzip2-1.0.2.new/unzcrash.c Fri Feb 1 04:19:18 2002 @@ -13,6 +13,12 @@ many hours. */ +#ifdef HAVE_CONFIG_H +#include +#endif + + + #include #include #include "bzlib.h"