--- /dev/null
+diff -Nru bzip2-1.0.1/AUTHORS bzip2-1.0.1.new/AUTHORS
+--- bzip2-1.0.1/AUTHORS Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/AUTHORS Sat Jun 24 20:13:05 2000
+@@ -0,0 +1 @@
++Julian Seward <jseward@acm.org>
+diff -Nru bzip2-1.0.1/CHANGES bzip2-1.0.1.new/CHANGES
+--- bzip2-1.0.1/CHANGES Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/CHANGES Thu Jan 1 01:00:00 1970
+@@ -1,167 +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 someway, 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.
+diff -Nru bzip2-1.0.1/COPYING bzip2-1.0.1.new/COPYING
+--- bzip2-1.0.1/COPYING Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/COPYING Sat Jun 24 20:13:05 2000
+@@ -0,0 +1,39 @@
++
++This program, "bzip2" and associated library "libbzip2", are
++copyright (C) 1996-2000 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 of 21 March 2000
++
+diff -Nru bzip2-1.0.1/ChangeLog bzip2-1.0.1.new/ChangeLog
+--- bzip2-1.0.1/ChangeLog Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/ChangeLog Sat Jun 24 20:13:05 2000
+@@ -0,0 +1 @@
++
+diff -Nru bzip2-1.0.1/INSTALL bzip2-1.0.1.new/INSTALL
+--- bzip2-1.0.1/INSTALL Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/INSTALL Sat Jun 24 20:13:06 2000
+@@ -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 bzip2-1.0.1/LICENSE bzip2-1.0.1.new/LICENSE
+--- bzip2-1.0.1/LICENSE Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/LICENSE Thu Jan 1 01:00:00 1970
+@@ -1,39 +0,0 @@
+-
+-This program, "bzip2" and associated library "libbzip2", are
+-copyright (C) 1996-2000 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 of 21 March 2000
+-
+diff -Nru bzip2-1.0.1/Makefile-libbz2_so bzip2-1.0.1.new/Makefile-libbz2_so
+--- bzip2-1.0.1/Makefile-libbz2_so Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/Makefile-libbz2_so Thu Jan 1 01:00:00 1970
+@@ -1,43 +0,0 @@
+-
+-# This Makefile builds a shared version of the library,
+-# libbz2.so.1.0.1, with soname libbz2.so.1.0,
+-# at least on x86-Linux (RedHat 5.2),
+-# with gcc-2.7.2.3. 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.1.0.1 $(OBJS)
+- $(CC) $(CFLAGS) -o bzip2-shared bzip2.c libbz2.so.1.0.1
+- rm -f libbz2.so.1.0
+- ln -s libbz2.so.1.0.1 libbz2.so.1.0
+-
+-clean:
+- rm -f $(OBJS) bzip2.o libbz2.so.1.0.1 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 bzip2-1.0.1/Makefile.am bzip2-1.0.1.new/Makefile.am
+--- bzip2-1.0.1/Makefile.am Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/Makefile.am Sat Jun 24 20:17:47 2000
+@@ -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
++
++bzip2SCRIPTS = bzless
++
++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 bzip2-1.0.1/NEWS bzip2-1.0.1.new/NEWS
+--- bzip2-1.0.1/NEWS Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/NEWS Sat Jun 24 20:13:06 2000
+@@ -0,0 +1,12 @@
++
++
++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.
+diff -Nru bzip2-1.0.1/acinclude.m4 bzip2-1.0.1.new/acinclude.m4
+--- bzip2-1.0.1/acinclude.m4 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/acinclude.m4 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1,129 @@
++#serial 7
++
++dnl By default, many hosts won't let programs access large files;
++dnl one must use special compiler options to get large-file access to work.
++dnl For more details about this brain damage please see:
++dnl http://www.sas.com/standards/large.file/x_open.20Mar96.html
++
++dnl Written by Paul Eggert <eggert@twinsun.com>.
++
++dnl Internal subroutine of AC_SYS_LARGEFILE.
++dnl AC_SYS_LARGEFILE_FLAGS(FLAGSNAME)
++AC_DEFUN(AC_SYS_LARGEFILE_FLAGS,
++ [AC_CACHE_CHECK([for $1 value to request large file support],
++ ac_cv_sys_largefile_$1,
++ [if ($GETCONF LFS_$1) >conftest.1 2>conftest.2 && test ! -s conftest.2
++ then
++ ac_cv_sys_largefile_$1=`cat conftest.1`
++ else
++ ac_cv_sys_largefile_$1=no
++ ifelse($1, CFLAGS,
++ [case "$host_os" in
++ # HP-UX 10.20 requires -D__STDC_EXT__ with gcc 2.95.1.
++changequote(, )dnl
++ hpux10.[2-9][0-9]* | hpux1[1-9]* | hpux[2-9][0-9]*)
++changequote([, ])dnl
++ if test "$GCC" = yes; then
++ ac_cv_sys_largefile_CFLAGS=-D__STDC_EXT__
++ fi
++ ;;
++ # IRIX 6.2 and later require cc -n32.
++changequote(, )dnl
++ irix6.[2-9]* | irix6.1[0-9]* | irix[7-9].* | irix[1-9][0-9]*)
++changequote([, ])dnl
++ if test "$GCC" != yes; then
++ ac_cv_sys_largefile_CFLAGS=-n32
++ fi
++ esac
++ if test "$ac_cv_sys_largefile_CFLAGS" != no; then
++ ac_save_CC="$CC"
++ CC="$CC $ac_cv_sys_largefile_CFLAGS"
++ AC_TRY_LINK(, , , ac_cv_sys_largefile_CFLAGS=no)
++ CC="$ac_save_CC"
++ fi])
++ fi
++ rm -f conftest*])])
++
++dnl Internal subroutine of AC_SYS_LARGEFILE.
++dnl AC_SYS_LARGEFILE_SPACE_APPEND(VAR, VAL)
++AC_DEFUN(AC_SYS_LARGEFILE_SPACE_APPEND,
++ [case $2 in
++ no) ;;
++ ?*)
++ case "[$]$1" in
++ '') $1=$2 ;;
++ *) $1=[$]$1' '$2 ;;
++ esac ;;
++ esac])
++
++dnl Internal subroutine of AC_SYS_LARGEFILE.
++dnl AC_SYS_LARGEFILE_MACRO_VALUE(C-MACRO, CACHE-VAR, COMMENT, CODE-TO-SET-DEFAULT)
++AC_DEFUN(AC_SYS_LARGEFILE_MACRO_VALUE,
++ [AC_CACHE_CHECK([for $1], $2,
++ [$2=no
++changequote(, )dnl
++ $4
++ for ac_flag in $ac_cv_sys_largefile_CFLAGS no; do
++ case "$ac_flag" in
++ -D$1)
++ $2=1 ;;
++ -D$1=*)
++ $2=`expr " $ac_flag" : '[^=]*=\(.*\)'` ;;
++ esac
++ done
++changequote([, ])dnl
++ ])
++ if test "[$]$2" != no; then
++ AC_DEFINE_UNQUOTED([$1], [$]$2, [$3])
++ fi])
++
++AC_DEFUN(AC_SYS_LARGEFILE,
++ [AC_REQUIRE([AC_CANONICAL_HOST])
++ AC_ARG_ENABLE(largefile,
++ [ --disable-largefile omit support for large files])
++ if test "$enable_largefile" != no; then
++ AC_CHECK_TOOL(GETCONF, getconf)
++ AC_SYS_LARGEFILE_FLAGS(CFLAGS)
++ AC_SYS_LARGEFILE_FLAGS(LDFLAGS)
++ AC_SYS_LARGEFILE_FLAGS(LIBS)
++
++ for ac_flag in $ac_cv_sys_largefile_CFLAGS no; do
++ case "$ac_flag" in
++ no) ;;
++ -D_FILE_OFFSET_BITS=*) ;;
++ -D_LARGEFILE_SOURCE | -D_LARGEFILE_SOURCE=*) ;;
++ -D_LARGE_FILES | -D_LARGE_FILES=*) ;;
++ -D?* | -I?*)
++ AC_SYS_LARGEFILE_SPACE_APPEND(CPPFLAGS, "$ac_flag") ;;
++ *)
++ AC_SYS_LARGEFILE_SPACE_APPEND(CFLAGS, "$ac_flag") ;;
++ esac
++ done
++ AC_SYS_LARGEFILE_SPACE_APPEND(LDFLAGS, "$ac_cv_sys_largefile_LDFLAGS")
++ AC_SYS_LARGEFILE_SPACE_APPEND(LIBS, "$ac_cv_sys_largefile_LIBS")
++ AC_SYS_LARGEFILE_MACRO_VALUE(_FILE_OFFSET_BITS,
++ ac_cv_sys_file_offset_bits,
++ [Number of bits in a file offset, on hosts where this is settable.],
++ [case "$host_os" in
++ # HP-UX 10.20 and later
++ hpux10.[2-9][0-9]* | hpux1[1-9]* | hpux[2-9][0-9]*)
++ ac_cv_sys_file_offset_bits=64 ;;
++ esac])
++ AC_SYS_LARGEFILE_MACRO_VALUE(_LARGEFILE_SOURCE,
++ ac_cv_sys_largefile_source,
++ [Define to make fseeko etc. visible, on some hosts.],
++ [case "$host_os" in
++ # HP-UX 10.20 and later
++ hpux10.[2-9][0-9]* | hpux1[1-9]* | hpux[2-9][0-9]*)
++ ac_cv_sys_largefile_source=1 ;;
++ esac])
++ AC_SYS_LARGEFILE_MACRO_VALUE(_LARGE_FILES,
++ ac_cv_sys_large_files,
++ [Define for large files, on AIX-style hosts.],
++ [case "$host_os" in
++ # AIX 4.2 and later
++ aix4.[2-9]* | aix4.1[0-9]* | aix[5-9].* | aix[1-9][0-9]*)
++ ac_cv_sys_large_files=1 ;;
++ esac])
++ fi
++ ])
+diff -Nru bzip2-1.0.1/bzip2.1 bzip2-1.0.1.new/bzip2.1
+--- bzip2-1.0.1/bzip2.1 Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/bzip2.1 Thu Jan 1 01:00:00 1970
+@@ -1,439 +0,0 @@
+-.PU
+-.TH bzip2 1
+-.SH NAME
+-bzip2, bunzip2 \- a block-sorting file compressor, v1.0
+-.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
+-invokation 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.
+-.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 to \-9
+-Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
+-effect when decompressing. See MEMORY MANAGEMENT below.
+-.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 "rec0001file.bz2",
+-"rec0002file.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" -- lists 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 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
+-and 0.9.5,
+-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
+-uses 32-bit integers to represent bit positions in
+-compressed files, so it cannot handle compressed files more than 512
+-megabytes long. This could easily be fixed.
+-
+-.SH AUTHOR
+-Julian Seward, jseward@acm.org.
+-
+-http://sourceware.cygnus.com/bzip2
+-http://www.muraroa.demon.co.uk
+-
+-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. Many people sent patches, helped
+-with portability problems, lent machines, gave advice and were generally
+-helpful.
+diff -Nru bzip2-1.0.1/bzip2.1.preformatted bzip2-1.0.1.new/bzip2.1.preformatted
+--- bzip2-1.0.1/bzip2.1.preformatted Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/bzip2.1.preformatted Thu Jan 1 01:00:00 1970
+@@ -1,462 +0,0 @@
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+-N\bNA\bAM\bME\bE
+- bzip2, bunzip2 - a block-sorting file compressor, v1.0
+- bzcat - decompresses files to stdout
+- bzip2recover - recovers data from damaged bzip2 files
+-
+-
+-S\bSY\bYN\bNO\bOP\bPS\bSI\bIS\bS
+- b\bbz\bzi\bip\bp2\b2 [ -\b-c\bcd\bdf\bfk\bkq\bqs\bst\btv\bvz\bzV\bVL\bL1\b12\b23\b34\b45\b56\b67\b78\b89\b9 ] [ _\bf_\bi_\bl_\be_\bn_\ba_\bm_\be_\bs _\b._\b._\b. ]
+- b\bbu\bun\bnz\bzi\bip\bp2\b2 [ -\b-f\bfk\bkv\bvs\bsV\bVL\bL ] [ _\bf_\bi_\bl_\be_\bn_\ba_\bm_\be_\bs _\b._\b._\b. ]
+- b\bbz\bzc\bca\bat\bt [ -\b-s\bs ] [ _\bf_\bi_\bl_\be_\bn_\ba_\bm_\be_\bs _\b._\b._\b. ]
+- b\bbz\bzi\bip\bp2\b2r\bre\bec\bco\bov\bve\ber\br _\bf_\bi_\bl_\be_\bn_\ba_\bm_\be
+-
+-
+-D\bDE\bES\bSC\bCR\bRI\bIP\bPT\bTI\bIO\bON\bN
+- _\bb_\bz_\bi_\bp_\b2 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 _\bG_\bN_\bU _\bg_\bz_\bi_\bp_\b, but they are not identical.
+-
+- _\bb_\bz_\bi_\bp_\b2 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.
+-
+- _\bb_\bz_\bi_\bp_\b2 and _\bb_\bu_\bn_\bz_\bi_\bp_\b2 will by default not overwrite existing
+- files. If you want this to happen, specify the -f flag.
+-
+- If no file names are specified, _\bb_\bz_\bi_\bp_\b2 compresses from
+- standard input to standard output. In this case, _\bb_\bz_\bi_\bp_\b2
+- will decline to write compressed output to a terminal, as
+- this would be entirely incomprehensible and therefore
+- pointless.
+-
+- _\bb_\bu_\bn_\bz_\bi_\bp_\b2 (or _\bb_\bz_\bi_\bp_\b2 _\b-_\bd_\b) decompresses all specified files.
+- Files which were not created by _\bb_\bz_\bi_\bp_\b2 will be detected and
+- ignored, and a warning issued. _\bb_\bz_\bi_\bp_\b2 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
+-
+-
+-
+- 1
+-
+-
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+- filename.tbz becomes filename.tar
+- anyothername becomes anyothername.out
+-
+- If the file does not end in one of the recognised endings,
+- _\b._\bb_\bz_\b2_\b, _\b._\bb_\bz_\b, _\b._\bt_\bb_\bz_\b2 or _\b._\bt_\bb_\bz_\b, _\bb_\bz_\bi_\bp_\b2 complains that it cannot
+- guess the name of the original file, and uses the original
+- name with _\b._\bo_\bu_\bt appended.
+-
+- As with compression, supplying no filenames causes decom-
+- pression from standard input to standard output.
+-
+- _\bb_\bu_\bn_\bz_\bi_\bp_\b2 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 _\bb_\bz_\bi_\bp_\b2 version 0.9.0 or
+- later. Earlier versions of _\bb_\bz_\bi_\bp_\b2 will stop after decom-
+- pressing the first file in the stream.
+-
+- _\bb_\bz_\bc_\ba_\bt (or _\bb_\bz_\bi_\bp_\b2 _\b-_\bd_\bc_\b) decompresses all specified files to
+- the standard output.
+-
+- _\bb_\bz_\bi_\bp_\b2 will read arguments from the environment variables
+- _\bB_\bZ_\bI_\bP_\b2 and _\bB_\bZ_\bI_\bP_\b, 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, _\bb_\bz_\bi_\bp_\b2 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 _\bb_\bz_\bi_\bp_\b2 (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
+- _\bb_\bz_\bi_\bp_\b2_\br_\be_\bc_\bo_\bv_\be_\br to try to recover data from damaged files.
+-
+-
+-
+- 2
+-
+-
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+- 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 _\bb_\bz_\bi_\bp_\b2 to panic.
+-
+-
+-O\bOP\bPT\bTI\bIO\bON\bNS\bS
+- -\b-c\bc -\b--\b-s\bst\btd\bdo\bou\but\bt
+- Compress or decompress to standard output.
+-
+- -\b-d\bd -\b--\b-d\bde\bec\bco\bom\bmp\bpr\bre\bes\bss\bs
+- Force decompression. _\bb_\bz_\bi_\bp_\b2_\b, _\bb_\bu_\bn_\bz_\bi_\bp_\b2 and _\bb_\bz_\bc_\ba_\bt 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 _\bb_\bz_\bi_\bp_\b2 to decompress.
+-
+- -\b-z\bz -\b--\b-c\bco\bom\bmp\bpr\bre\bes\bss\bs
+- The complement to -d: forces compression, regard-
+- less of the invokation name.
+-
+- -\b-t\bt -\b--\b-t\bte\bes\bst\bt
+- Check integrity of the specified file(s), but don't
+- decompress them. This really performs a trial
+- decompression and throws away the result.
+-
+- -\b-f\bf -\b--\b-f\bfo\bor\brc\bce\be
+- Force overwrite of output files. Normally, _\bb_\bz_\bi_\bp_\b2
+- will not overwrite existing output files. Also
+- forces _\bb_\bz_\bi_\bp_\b2 to break hard links to files, which it
+- otherwise wouldn't do.
+-
+- -\b-k\bk -\b--\b-k\bke\bee\bep\bp
+- Keep (don't delete) input files during compression
+- or decompression.
+-
+- -\b-s\bs -\b--\b-s\bsm\bma\bal\bll\bl
+- 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.
+-
+- -\b-q\bq -\b--\b-q\bqu\bui\bie\bet\bt
+- Suppress non-essential warning messages. Messages
+- pertaining to I/O errors and other critical events
+-
+-
+-
+- 3
+-
+-
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+- will not be suppressed.
+-
+- -\b-v\bv -\b--\b-v\bve\ber\brb\bbo\bos\bse\be
+- 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.
+-
+- -\b-L\bL -\b--\b-l\bli\bic\bce\ben\bns\bse\be -\b-V\bV -\b--\b-v\bve\ber\brs\bsi\bio\bon\bn
+- Display the software version, license terms and
+- conditions.
+-
+- -\b-1\b1 t\bto\bo -\b-9\b9
+- Set the block size to 100 k, 200 k .. 900 k when
+- compressing. Has no effect when decompressing.
+- See MEMORY MANAGEMENT below.
+-
+- -\b--\b- 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.
+-
+- -\b--\b-r\bre\bep\bpe\bet\bti\bit\bti\biv\bve\be-\b-f\bfa\bas\bst\bt -\b--\b-r\bre\bep\bpe\bet\bti\bit\bti\biv\bve\be-\b-b\bbe\bes\bst\bt
+- 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.
+-
+-
+-M\bME\bEM\bMO\bOR\bRY\bY M\bMA\bAN\bNA\bAG\bGE\bEM\bME\bEN\bNT\bT
+- _\bb_\bz_\bi_\bp_\b2 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 _\bb_\bu_\bn_\bz_\bi_\bp_\b2 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
+-
+-
+-
+- 4
+-
+-
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+- or three hundred k of block size, a fact worth bearing in
+- mind when using _\bb_\bz_\bi_\bp_\b2 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,
+- _\bb_\bu_\bn_\bz_\bi_\bp_\b2 will require about 3700 kbytes to decompress. To
+- support decompression of any file on a 4 megabyte machine,
+- _\bb_\bu_\bn_\bz_\bi_\bp_\b2 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
+-
+-
+-
+-
+-
+-
+- 5
+-
+-
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+-R\bRE\bEC\bCO\bOV\bVE\bER\bRI\bIN\bNG\bG D\bDA\bAT\bTA\bA F\bFR\bRO\bOM\bM D\bDA\bAM\bMA\bAG\bGE\bED\bD F\bFI\bIL\bLE\bES\bS
+- _\bb_\bz_\bi_\bp_\b2 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.
+-
+- _\bb_\bz_\bi_\bp_\b2_\br_\be_\bc_\bo_\bv_\be_\br 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 _\bb_\bz_\bi_\bp_\b2 -t to test
+- the integrity of the resulting files, and decompress those
+- which are undamaged.
+-
+- _\bb_\bz_\bi_\bp_\b2_\br_\be_\bc_\bo_\bv_\be_\br takes a single argument, the name of the dam-
+- aged file, and writes a number of files "rec0001file.bz2",
+- "rec0002file.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" -- lists the files in
+- the correct order.
+-
+- _\bb_\bz_\bi_\bp_\b2_\br_\be_\bc_\bo_\bv_\be_\br 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.
+-
+-
+-P\bPE\bER\bRF\bFO\bOR\bRM\bMA\bAN\bNC\bCE\bE N\bNO\bOT\bTE\bES\bS
+- 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.
+-
+- _\bb_\bz_\bi_\bp_\b2 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
+-
+-
+-
+- 6
+-
+-
+-
+-
+-
+-bzip2(1) bzip2(1)
+-
+-
+- 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 _\bb_\bz_\bi_\bp_\b2 will per-
+- form best on machines with very large caches.
+-
+-
+-C\bCA\bAV\bVE\bEA\bAT\bTS\bS
+- I/O error messages are not as helpful as they could be.
+- _\bb_\bz_\bi_\bp_\b2 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 of _\bb_\bz_\bi_\bp_\b2_\b. Com-
+- pressed data created by this version is entirely forwards
+- and backwards compatible with the previous public
+- releases, versions 0.1pl2, 0.9.0 and 0.9.5, but with the
+- following exception: 0.9.0 and above can correctly decom-
+- press multiple concatenated compressed files. 0.1pl2 can-
+- not do this; it will stop after decompressing just the
+- first file in the stream.
+-
+- _\bb_\bz_\bi_\bp_\b2_\br_\be_\bc_\bo_\bv_\be_\br uses 32-bit integers to represent bit posi-
+- tions in compressed files, so it cannot handle compressed
+- files more than 512 megabytes long. This could easily be
+- fixed.
+-
+-
+-A\bAU\bUT\bTH\bHO\bOR\bR
+- Julian Seward, jseward@acm.org.
+-
+- http://sourceware.cygnus.com/bzip2
+- http://www.muraroa.demon.co.uk
+-
+- The ideas embodied in _\bb_\bz_\bi_\bp_\b2 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 _\bb_\bz_\bi_\bp_\b, and many refinements), and
+- Alistair Moffat, Radford Neal and Ian Witten (for the
+- arithmetic coder in the original _\bb_\bz_\bi_\bp_\b)_\b. 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. Many people sent patches, helped
+- with portability problems, lent machines, gave advice and
+- were generally helpful.
+-
+-
+-
+-
+-
+-
+-
+-
+- 7
+-
+-
+diff -Nru bzip2-1.0.1/bzless bzip2-1.0.1.new/bzless
+--- bzip2-1.0.1/bzless Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/bzless Sat Jun 24 20:16:09 2000
+@@ -0,0 +1,2 @@
++#!/bin/sh
++%{_bindir}/bunzip2 -c "\$@" | /usr/bin/less
+diff -Nru bzip2-1.0.1/config.h.in bzip2-1.0.1.new/config.h.in
+--- bzip2-1.0.1/config.h.in Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/config.h.in Sat Jun 24 20:13:06 2000
+@@ -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 bzip2-1.0.1/configure.in bzip2-1.0.1.new/configure.in
+--- bzip2-1.0.1/configure.in Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/configure.in Sat Jun 24 20:13:06 2000
+@@ -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 bzip2-1.0.1/crctable.c bzip2-1.0.1.new/crctable.c
+--- bzip2-1.0.1/crctable.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/crctable.c Sat Jun 24 20:13:06 2000
+@@ -58,6 +58,10 @@
+ For more information on these sources, see the manual.
+ --*/
+
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
+
+ #include "bzlib_private.h"
+
+diff -Nru bzip2-1.0.1/decompress.c bzip2-1.0.1.new/decompress.c
+--- bzip2-1.0.1/decompress.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/decompress.c Sat Jun 24 20:13:06 2000
+@@ -58,6 +58,10 @@
+ For more information on these sources, see the manual.
+ --*/
+
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
+
+ #include "bzlib_private.h"
+
+diff -Nru bzip2-1.0.1/dlltest.c bzip2-1.0.1.new/dlltest.c
+--- bzip2-1.0.1/dlltest.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/dlltest.c Sat Jun 24 20:13:06 2000
+@@ -8,6 +8,10 @@
+ usage: minibz2 [-d] [-{1,2,..9}] [[srcfilename] destfilename]\r
+ */\r
+ \r
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
+ #define BZ_IMPORT\r
+ #include <stdio.h>\r
+ #include <stdlib.h>\r
+diff -Nru bzip2-1.0.1/doc/Makefile.am bzip2-1.0.1.new/doc/Makefile.am
+--- bzip2-1.0.1/doc/Makefile.am Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/Makefile.am Sat Jun 24 20:14:43 2000
+@@ -0,0 +1,5 @@
++
++SUBDIRS = pl
++
++man_MANS = bzip2.1 bunzip2.1 bzcat.1 bzip2recover.1
++#info_TEXINFOS = bzip2.texi
+diff -Nru bzip2-1.0.1/doc/bunzip2.1 bzip2-1.0.1.new/doc/bunzip2.1
+--- bzip2-1.0.1/doc/bunzip2.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/bunzip2.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++.so bzip2.1
+\ No newline at end of file
+diff -Nru bzip2-1.0.1/doc/bzcat.1 bzip2-1.0.1.new/doc/bzcat.1
+--- bzip2-1.0.1/doc/bzcat.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/bzcat.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++.so bzip2.1
+\ No newline at end of file
+diff -Nru bzip2-1.0.1/doc/bzip2.1 bzip2-1.0.1.new/doc/bzip2.1
+--- bzip2-1.0.1/doc/bzip2.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/bzip2.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1,439 @@
++.PU
++.TH bzip2 1
++.SH NAME
++bzip2, bunzip2 \- a block-sorting file compressor, v1.0
++.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
++invokation 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.
++.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 to \-9
++Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
++effect when decompressing. See MEMORY MANAGEMENT below.
++.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 "rec0001file.bz2",
++"rec0002file.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" -- lists 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 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
++and 0.9.5,
++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
++uses 32-bit integers to represent bit positions in
++compressed files, so it cannot handle compressed files more than 512
++megabytes long. This could easily be fixed.
++
++.SH AUTHOR
++Julian Seward, jseward@acm.org.
++
++http://sourceware.cygnus.com/bzip2
++http://www.muraroa.demon.co.uk
++
++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. Many people sent patches, helped
++with portability problems, lent machines, gave advice and were generally
++helpful.
+diff -Nru bzip2-1.0.1/doc/bzip2.texi bzip2-1.0.1.new/doc/bzip2.texi
+--- bzip2-1.0.1/doc/bzip2.texi Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/bzip2.texi Sat Jun 24 20:13:06 2000
+@@ -0,0 +1,2217 @@
++\input texinfo @c -*- Texinfo -*-
++@setfilename bzip2.info
++
++@ignore
++This file documents bzip2 version 1.0, and associated library
++libbzip2, written by Julian Seward (jseward@acm.org).
++
++Copyright (C) 1996-2000 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
++@dircategory File utilities:
++* Bzip2: (bzip2). A program and library for data
++ compression
++@end direntry
++@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) 1996-2000 Julian Seward
++@subtitle version 1.0 of 21 March 2000
++@author Julian Seward
++
++@end titlepage
++
++@parindent 0mm
++@parskip 2mm
++
++@end iftex
++@node Top, Overview, (dir), (dir)
++
++@top bzip2
++
++This program, @code{bzip2},
++and associated library @code{libbzip2}, are
++Copyright (C) 1996-2000 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{http://sourceware.cygnus.com/bzip2}
++
++@code{http://www.cacheprof.org}
++
++@code{http://www.muraroa.demon.co.uk}
++
++@code{bzip2}/@code{libbzip2} version 1.0 of 21 March 2000.
++
++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.
++
++
++
++
++
++
++
++@node Overview, Implementation, Top, Top
++@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
++@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.
++@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 to -9
++Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
++effect when decompressing. See MEMORY MANAGEMENT below.
++@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{rec0001file.bz2},
++ @code{rec0002file.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} -- lists 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 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 and
++0.9.5, 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} uses 32-bit integers to represent bit positions in
++compressed files, so it cannot handle compressed files more than 512
++megabytes long. This could easily be fixed.
++
++
++@unnumberedsubsubsec AUTHOR
++Julian Seward, @code{jseward@@acm.org}.
++
++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. 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 of 21-Mar-2000.
++ 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, 21 March 2000.
++@end example
++where @code{N} is some error code number. @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-1.1.2} and
++@code{gzip-1.2.4}. Look for them at
++
++@code{http://www.cdrom.com/pub/infozip/zlib} 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 bzip2-1.0.1/doc/bzip2recover.1 bzip2-1.0.1.new/doc/bzip2recover.1
+--- bzip2-1.0.1/doc/bzip2recover.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/bzip2recover.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++.so bzip2.1
+\ No newline at end of file
+diff -Nru bzip2-1.0.1/doc/pl/Makefile.am bzip2-1.0.1.new/doc/pl/Makefile.am
+--- bzip2-1.0.1/doc/pl/Makefile.am Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/pl/Makefile.am Sat Jun 24 20:13:06 2000
+@@ -0,0 +1,4 @@
++
++mandir = @mandir@/pl
++man_MANS = bzip2.1 bunzip2.1 bzcat.1 bzip2recover.1
++
+diff -Nru bzip2-1.0.1/doc/pl/bunzip2.1 bzip2-1.0.1.new/doc/pl/bunzip2.1
+--- bzip2-1.0.1/doc/pl/bunzip2.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/pl/bunzip2.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++.so bzip2.1
+\ No newline at end of file
+diff -Nru bzip2-1.0.1/doc/pl/bzcat.1 bzip2-1.0.1.new/doc/pl/bzcat.1
+--- bzip2-1.0.1/doc/pl/bzcat.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/pl/bzcat.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++.so bzip2.1
+\ No newline at end of file
+diff -Nru bzip2-1.0.1/doc/pl/bzip2.1 bzip2-1.0.1.new/doc/pl/bzip2.1
+--- bzip2-1.0.1/doc/pl/bzip2.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/pl/bzip2.1 Sat Jun 24 20:13:06 2000
+@@ -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 u¿ywaj±c algorytmu sortowania bloków Burrowsa-Wheelera i
++kodu Huffmana. Kompresja jest generalnie sporo lepsza od konwencjonalnych
++kompresorów opartych o metodê LZ77/LZ78, i jest porównywalna z
++osi±gniêciami statystycznych kompresorów z rodziny PPM.
++
++Opcje linii poleceñ s± w wiêkszo¶ci bardzo podobne do tych z
++.IR "GNU gzip" ,
++ale nie s± identyczne.
++
++.I bzip2
++oczekuje listy plików towarzysz±cych parametrom linii poleceñ. Ka¿dy plik jest
++zastêpowany przez swoj± skompresowan± wersjê, z nazw±
++"oryginalny_plik.bz2". Ka¿dy skompresowany plik ma ten sam czas modyfikacji,
++uprawnienia i, je¶li to mo¿liwe, w³a¶ciciela co orygina³, po to, aby te
++ustawienia mog³y zostaæ odtworzone podczas dekompresji. Utrzymywanie nazwy
++plików nie jest do koñca dok³adne w tym sensie, ¿e nie ma mo¿liwo¶ci
++przetrzymywania daty, uprawnieñ, w³a¶ciciela i nazw plików na systemach, na
++których brakuje tych mo¿liwo¶ci lub maj± ograniczenia co do d³ugo¶ci nazwy,
++tak np. jak MS-DOS.
++
++.I bzip2
++i
++.I bunzip2
++standardowo nie nadpisuj± istniej±cych ju¿ plików. Je¶li chcesz aby to
++robi³y, musisz u¿yæ parametru \-f.
++
++Je¶li nie podano ¿adnej nazwy pliku,
++.I bzip2
++kompresuje ze standardowego wej¶cia na standardowe wyj¶cie. Odmiawia wówczas
++wypisywania skompresowanego wyj¶cie na terminal, gdy¿ by³oby to
++ca³kiem niezrozumia³e i przez to bez wiêkszego 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 siê komunikat
++ostrzegawczy.
++.I bzip2
++próbuje zgadn±æ nazwê dla dekompresowanego pliku w nastêpuj±cy sposób:
++.nf
++ nazwa_pliku.bz2 staje siê nazwa_pliku
++ nazwa_pliku.bz staje siê nazwa_pliku
++ nazwa_pliku.tbz2 staje siê nazwa_pliku.tar
++ nazwa_pliku.tbz staje siê nazwa_pliku.tar
++ inna_nazwa staje siê inna_nazwa.out
++.fi
++Je¶li plik nie ma jednego z nastêpuj±cych rozpoznawalnych rozszerzeñ,
++.IR .bz2 ,
++.IR .bz ,
++.I .tbz2
++lub
++.IR .tbz ,
++to
++.I bzip2
++napisze, ¿e nie mo¿e zgadn±æ nazwy pierwotnego pliku, i u¿yje
++oryginalnej nazwy z dodanym rozszerzeniem
++.IR .out .
++
++Tak jak kompresja, nie posiadaj±ca ¿adnych plików, powoduje kompresjê ze
++standardowego wej¶cia na standardowe wyj¶cie.
++
++.I bunzip2
++poprawnie zdekompresuje plik, który jest po³aczeniem dwóch lub wiêcej
++skompresowanych plików. Rezultatem jest po³±czony odpowiedni
++nieskompresowany plik. Obs³ugiwane jest równie¿ sprawdzanie spójno¶ci
++(\-t) po³±czonych skompresowanych plików.
++
++Mo¿esz równie¿ kompresowaæ lub dekompresowaæ pliki na standardowe wyj¶cie
++u¿ywaj±c parametru \-c. W ten w³a¶nie sposób mo¿na przeprowadzaæ kompresjê
++wielu plików równocze¶nie.
++Powsta³e wyniki s± przesy³ane sekwencyjnie na standardowe wyj¶cie.
++W ten sposób kompresja wielu plików generuje strumieñ
++zawieraj±cy reprezentacje kilku skompresowanych plików. Taki strumieñ mo¿e
++byæ zdekompresowany poprawnie tylko przez
++.I bzip2
++w wersji 0.9.0 lub pó¼niejszej. Wcze¶niejsze wersje
++.I bzip2
++zatrzymaj± siê 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 poleceñ. To dobra metoda na specyfikowanie
++standardowych ustawieñ.
++
++Kompresja stosowana jest zawsze, nawet je¶li skompresowany plik jest
++nieznaczniej wiêkszy od pliku oryginalnego. Pliki mniejsze ni¿ mniej wiêcej
++sto bajtów staj± siê wiêksze, poniewa¿ mechanizm kompresji ma sta³y
++nag³ówek wynosz±cy oko³o 50 bajtów. Przypadkowe dane (w³±czaj±c wyj¶cie
++wiêkszo¶ci kompresorów plików) d± kodowane na mniej wiêcej 8.05 bitu na
++bajt, daj±c zysk oko³o 0.5%.
++
++Jako samosprawdzenie dla twojej ochrony
++.I bzip2
++u¿ywa 32-bitowego CRC aby upewniæ siê, ¿e zdekompresowana wersja pliku jest
++identyczna z oryginaln±. To strze¿e przed stratami w skompresowanych danych
++i przed niewykrytymi b³êdami w
++.I bzip2
++(na szczê¶cie bardzo rzadkich). Mo¿liwo¶æ niewykrycia utraty danych
++jest mikroskopijna, mniej wiêcej jedna szansa na cztery biliony dla ka¿dego
++pliku. Uwa¿aj jednak, gdy¿ sprawdzenie jest dokonywane przed dekompresj±,
++wiêc dowiesz siê tylko tego, ¿e co¶ jest nie w porz±dku. Nie pomo¿e ci to odzyskaæ
++oryginalnych nieskompresowanych danych. Mo¿esz u¿yæ
++.I bzip2recover
++aby spróbowaæ odzyskaæ 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 b³êdu skompresowanego pliku, 3 dla wewnêtrznego b³êdu (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 dekompresjê.
++.IR bzip2 ,
++.I bunzip2
++i
++.I bzcat
++s± tak naprawdê tymi samymi programami i decyzja jakie akcje bêd± wykonane
++jest wykonywana na podstawie nazwy jaka zosta³a u¿yta. Ten parametr ma wy¿szy
++priorytet i wymusza na \fIbzip2\fP dekompresjê.
++.TP
++.B \-z --compress
++Podobne do \-d: wymusza kompresjê, bez wzglêdu na sposób wywo³ania.
++.TP
++.B \-t --test
++Sprawdza integralno¶æ wybranego pliku(ów), ale nie dekompresuje ich. Wymusza
++to próbn± dekompresjê i mówi, jaki jest rezultat.
++.TP
++.B \-f --force
++Wymusza zastêpowanie plików wyj¶ciowych. Normalnie, \fIbzip2\fP nie
++zastêpuje istniej±cych plików wyj¶ciowych. Wymusza równie¿ na \fIbzip2\fP
++³amanie dowi±zañ twardych, czego normalnie nie robi.
++.TP
++.B \-k --keep
++Zatrzymaj (nie kasuj) pliki wej¶ciowe przy kompresji lub dekompresji.
++.TP
++.B \-s --small
++Zredukuj u¿ycie pamiêci na kompresjê, dekompresjê i testowanie. Pliki s±
++dekompresowane i testowane przy u¿yciu zmodyfikowanego algorytmu, który
++potrzebuje tylko 2.5 bajtu na blok bajtów. Oznacza to, ¿e ka¿dy plik mo¿e
++byæ zdekompresowany przy u¿yciu oko³o 2300k pamiêci, jednak trac±c oko³o po³owê
++normalnej szybko¶ci.
++
++Podczas kompresji, \-s wybiera bloki wielko¶ci 200k, których limity
++pamiêci wynosz± mniej wiêcej tyle samo, w zamian za jako¶æ kompresji. W
++skrócie, je¶li twój komputer ma ma³o pamiêci (8 megabajtów lub mniej),
++u¿ywaj opcji \-s do wszystkiego. Zobacz \fBzarz±dzanie pamiêci±\fP poni¿ej.
++.TP
++.B \-q --quiet
++Wy³±cza wszystkie nieistotne komunikaty ostrzegawcze.
++Nie s± eliminowane komunikaty dotycz±ce b³êdów wej¶cia/wyj¶cia i innych
++zdarzeñ krytycznych.
++.TP
++.B \-v --verbose
++Tryb gadatliwy -- pokazuje stopieñ kompresji dla ka¿dego pliku. Nastêpne
++\fB\-v\fP zwiêkszaj± stopieñ gadatliwo¶ci, powoduj±c wy¶wietlanie du¿ej
++ilo¶ci informacji, przydatnych g³ównie przy diagnostyce.
++.TP
++.B \-L --license -V --version
++Wy¶wietla wersjê programu i warunki licencji.
++.TP
++.B \-1 to \-9
++Ustawia wielko¶æ bloku na 100 k, 200 k .. 900 k przy kompresji. Nie ma
++¿adnego znaczenia przy dekompresji. Zobacz \fBzarz±dzanie pamiêci±\fP
++poni¿ej.
++.TP
++.B \--
++Traktuje wszystkie nastêpuj±ce po nim argumenty jako nazwy plików, nawet je¶li
++zaczynaj± siê one od my¶lnika. Mo¿esz wiêc kompresowaæ i dekompresowaæ
++pliki, których nazwa zaczyna siê 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 wy¿szych. Umo¿liwia³y one
++pewn± infantyln± kontrolê nad zachowaniem algorytmu sortuj±cego we
++wcze¶niejszych wersjach, co by³o czasami u¿yteczne. Wersje 0.9.5 i wy¿sze
++maj± usprawniony algorytm, który powoduje bezu¿yteczno¶æ tej funkcji.
++
++.SH ZARZ¡DZANIE PAMIÊCI¡
++.I bzip2
++kompresuje du¿e pliki w blokach. Rozmiar bloku ma wp³yw zarówno na stopieñ
++osi±ganej kompresji, jak równie¿ na ilo¶æ pamiêci potrzebnej do kompresji
++i dekompresji. Parametry od \-1 do \-9 wybieraj± rozmiar bloku odpowiednio
++od 100,000 bajtów a¿ do 900,000 bajtów (standardowo). W czasie dekompresji,
++rozmiar bloku u¿ytego do kompresji jest odczytywany z nag³ówku pliku
++skompresowanego i
++.I bunzip2
++sam zajmuje odpowiedni± do dekompresji ilo¶æ pamiêci. Poniewa¿ rozmiar
++bloków jest przetrzymywany w pliku skompresowanym, parametry od \-1 do \-9
++nie maj± przy dekompresji ¿adnego znaczenia.
++
++Wymagania kompresji i dekompresji w bajtach, mog± byæ wyliczone przez:
++
++ Kompresja : 400k + ( 8 x rozmiar bloku )
++
++ Dekompresja : 100k + ( 4 x rozmiar bloku ) lub
++ 100k + ( 2.5 x rozmiar bloku )
++
++Wiêksze bloki daj± du¿e zmniejszenie zwrotów marginalnych. Wiêkszo¶æ
++kompresji pochodzi z pierwszych stu lub dwustu kilobajtów rozmiaru bloku.
++Warto o tym pamiêtaæ u¿ywaj±c \fIbzip2\fP na wolnych
++komputerach. Warto równie¿ podkre¶liæ, ¿e rozmiar pamiêci potrzebnej do
++dekompresji jest wybierany poprzez ustawienie odpowiedniej
++wielko¶ci bloku przy kompresji.
++
++Dla plików skompresowanych standardowym blokiem wielko¶ci 900k,
++\fIbunzip2\fP bêdzie wymaga³ oko³o 3700 kilobajtów do dekompresji. Aby
++umo¿liwiæ dekompresjê na komputerze wyposa¿onym jedynie w 4 megabajty
++pamiêci, \fIbunzip2\fP ma opcjê, która mo¿e zmniejszyæ wymagania prawie do
++po³owy, tzn. oko³o 2300 kilobajtów. Prêdko¶æ dekompresji jest równie¿ bardzo
++zmiejszona, wiêc u¿ywaj tej opcji tylko wtedy, kiedy jest to konieczne. Tym
++parametrem jest -s.
++
++Generalnie, próbuj i u¿ywaj najwiêkszych rozmiarów bloków, je¶li ilo¶æ
++pamiêci ci na to pozwala. Prêdko¶æ kompresji i dekompresji w zasadzie nie
++zale¿y od wielko¶ci u¿ytego bloku.
++
++Inna wa¿na rzecz dotyczy plików, które mieszcz± siê w pojedyñczym bloku --
++oznacza to wiêkszo¶æ plików na które siê natkniesz u¿ywaj±c du¿ych bloków.
++Rozmiar realny pamiêci zabieranej jest proporcjonalny do wielko¶ci pliku,
++je¶li plik jest mniejszy ni¿ blok. Na przyk³ad, kompresja pliku o
++wielko¶ci 20,000 bajtów z parametrem -9 wymusi na kompresorze odnalezienie
++7600 k pamiêci, ale zajêcie 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 u¿ycie pamiêci dla ró¿nych
++rozmiarów bloków. Podano te¿ ca³kowity rozmiar skompresowanych 14
++plików tekstowych (Calgary Text Compressione Corpus) zajmuj±cych razem
++3,141,622 bajtów. Ta kolumna daje pewne pojêcie o tym, jaki wp³yw na
++kompresjê ma wielko¶æ bloków. Ta tabela uzmys³awia równie¿ przewagê u¿ycia
++wiêkszych bloków dla wiêkszych plików, poniewa¿ "Corpus" jest zdominowany
++przez mniejsze pliki.
++.nf
++ U¿ycie U¿ycie U¿ycie 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. Ka¿dy blok jest
++trzymany osobno. Je¶li b³êdy transmisji lub no¶nika uszkodz± plik
++wieloblokowy .bz2, mo¿liwe jest odtworzenie danych zawartych w
++niezniszczonych blokach pliku.
++
++Ka¿dy blok jest reprezentowany przez 48-bitowy wzorzec, który umo¿liwia
++znajdowanie przyporz±dkowañ bloków z rozs±dn± pewno¶ci±. Ka¿dy blok
++ma równie¿ swój 32-bitowy CRC, wiêc bloki uszkodzone mog± byæ ³atwo
++odseparowane od poprawnych.
++
++.I bzip2recover
++jest oddzielnym programem, którego zadaniem jest poszukiwanie bloków w
++plikach .bz2 i zapisywanie ich do w³asnego pliku .bz2. Mo¿esz potem u¿yæ
++\fIbzip2\fP \-t aby sprawdziæ spójno¶æ wyj¶ciowego pliku i zdekompresowaæ
++te, które nie s± uszkodzone.
++
++.I bzip2recover
++pobiera pojedynczy argument, nazwê uszkodzonego pliku, i tworzy pewn± liczbê
++plików "rec0001plik.bz2", "rec0002plik.bz2", itd., przetrzymuj±ce odzyskane
++bloki. Wyj¶ciowe nazwy plików s± tak tworzone, aby ³atwo by³o potem u¿ywaæ
++ich razem za pomoc± gwiazdek -- na przyk³ad, "bzip2 -dc rec*plik.bz2 >
++odzyskany_plik" -- wylistuje pliki we w³a¶ciwej kolejno¶ci.
++
++.I bzip2recover
++powinien byæ u¿ywany najczê¶ciej z du¿ymi plikami .bz2, jako i¿ one
++zawieraj± najczê¶ciej du¿o bloków. Jest czystym bezsensem u¿ywaæ go na
++uszkodzonym jedno-blokowym pliku, poniewa¿ uszkodzony blok nie mo¿e byæ
++odzyskany. Je¶li chcesz zminimalizowaæ jakiekolwiek mo¿liwe straty danych
++poprzez no¶nik lub transmisjê, powiniene¶ zastanowiæ siê nad u¿yciem
++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 siê symboli, jak
++"aabaabaabaab ..." (powtórzone kilkaset razy) mog± byæ kompresowane wolniej
++ni¿ normalnie. Wersje 0.9.5 i wy¿sze zachowuj± siê du¿o lepiej w tej
++sytuacji ni¿ wersje poprzednie. Ró¿nica stopnia kompresji pomiêdzy
++najgorszym i najlepszym przypadkiem kompresji wynosi oko³o 10:1. Dla
++wcze¶niejszych wersji by³o to nawet oko³o 100:1. Je¶li chcesz, mo¿esz u¿yæ
++parametru \-vvvv aby monitorowaæ postêpy bardzo szczegó³owo.
++
++Prêdko¶æ dekompresji nie jest zmieniana przez to zjawisko.
++
++.I bzip2
++zazwyczaj rezerwuje kilka megabajtów pamiêci do dzia³ania a
++potem wykorzystuje j± w sposób zupe³nie przypadkowy.
++Oznacza to, ¿e zarówno prêdko¶æ kompresji jak i dekompresji jest w
++du¿ej czê¶ci zale¿na od prêdko¶ci, z jak± twój komputer mo¿e naprawiaæ braki
++bufora podrêcznego. Z tego powodu, wprowadzone zosta³y ma³e zmiany kody aby
++zmniejszyæ straty, które da³y nieproporcjonalnie du¿y wzrost osi±gniêæ.
++My¶lê, ¿e
++.I bzip2
++bêdzie dzia³a³ najlepiej na komputerach z du¿ymi buforami podrêcznymi.
++
++.SH ZAKAMARKI
++Wiadomo¶ci o b³êdach wej¶cia/wyj¶cia nie s± a¿ tak pomocne, jak mog³yby byæ.
++.I bzip2
++stara siê wykryæ b³±d wej¶cia/wyj¶cia i wyj¶æ "czysto", ale
++szczegó³y tego, jaki to problem mog± byæ czasami bardzo myl±ce.
++
++Ta strona podrêcznika odnosi siê do wersji 1.0 programu \fIbzip2\fP.
++Skompresowane pliki utworzone przez tê wersjê 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 wy¿sze potrafi±
++poprawnie dekompresowaæ wiele skompresowanych plików z³±czonych w jeden.
++0.1pl2 nie potrafi tego; zatrzyma siê ju¿ po dekompresji pierwszego pliku w
++strumieniu.
++
++.I bzip2recover
++u¿ywa 32-bitowych liczb do reprezentacji pozycji bitu w skompresowanym
++pliku, wiêc nie mo¿e przetwarzaæ skompresowanych plików d³u¿szych ni¿ 512
++megabajtów. Mo¿na to ³atwo naprawiæ.
++
++.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) pomiêdzy
++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
++udoskonaleñ), i Alistair Moffar, Radford Neal i Ian Witten (arytmetyczny
++koder w oryginalnym \fIbzip2\fP). Jestem im bardzo wdziêczny za ich pomoc,
++wsparcie i porady. Zobacz stronê manuala w ¼ród³owej dystrybucji po
++wska¼niki do ¼róde³ dokumentacji. Christian von Roques zachêci³ mnie do
++wymy¶lenia szybszego algorytmu sortuj±cego, po to ¿eby przyspieszyæ
++kompresjê. Bela Lubkin zachêci³a mnie do polepszenia najgorszych wyników
++kompresji. Wiele ludzi przys³a³o ³atki, pomog³o w ró¿nych problemach,
++po¿yczy³o komputerów, da³o rady i by³o ogólnie pomocnych.
+diff -Nru bzip2-1.0.1/doc/pl/bzip2recover.1 bzip2-1.0.1.new/doc/pl/bzip2recover.1
+--- bzip2-1.0.1/doc/pl/bzip2recover.1 Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/doc/pl/bzip2recover.1 Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++.so bzip2.1
+\ No newline at end of file
+diff -Nru bzip2-1.0.1/huffman.c bzip2-1.0.1.new/huffman.c
+--- bzip2-1.0.1/huffman.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/huffman.c Sat Jun 24 20:13:06 2000
+@@ -58,6 +58,10 @@
+ For more information on these sources, see the manual.
+ --*/
+
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
+
+ #include "bzlib_private.h"
+
+diff -Nru bzip2-1.0.1/makefile.msc bzip2-1.0.1.new/makefile.msc
+--- bzip2-1.0.1/makefile.msc Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/makefile.msc Thu Jan 1 01:00:00 1970
+@@ -1,63 +0,0 @@
+-# Makefile for Microsoft Visual C++ 6.0\r
+-# usage: nmake -f makefile.msc\r
+-# K.M. Syring (syring@gsf.de)\r
+-# Fixed up by JRS for bzip2-0.9.5d release.\r
+-\r
+-CC=cl\r
+-CFLAGS= -DWIN32 -MD -Ox -D_FILE_OFFSET_BITS=64\r
+-\r
+-OBJS= blocksort.obj \\r
+- huffman.obj \\r
+- crctable.obj \\r
+- randtable.obj \\r
+- compress.obj \\r
+- decompress.obj \\r
+- bzlib.obj\r
+-\r
+-all: lib bzip2 test\r
+-\r
+-bzip2: lib\r
+- $(CC) $(CFLAGS) -o bzip2 bzip2.c libbz2.lib setargv.obj\r
+- $(CC) $(CFLAGS) -o bzip2recover bzip2recover.c\r
+-\r
+-lib: $(OBJS)\r
+- lib /out:libbz2.lib $(OBJS)\r
+-\r
+-test: bzip2\r
+- type words1\r
+- .\\bzip2 -1 < sample1.ref > sample1.rb2\r
+- .\\bzip2 -2 < sample2.ref > sample2.rb2\r
+- .\\bzip2 -3 < sample3.ref > sample3.rb2\r
+- .\\bzip2 -d < sample1.bz2 > sample1.tst\r
+- .\\bzip2 -d < sample2.bz2 > sample2.tst\r
+- .\\bzip2 -ds < sample3.bz2 > sample3.tst\r
+- @echo All six of the fc's should find no differences.\r
+- @echo If fc finds an error on sample3.bz2, this could be\r
+- @echo because WinZip's 'TAR file smart CR/LF conversion'\r
+- @echo is too clever for its own good. Disable this option.\r
+- @echo The correct size for sample3.ref is 120,244. If it\r
+- @echo is 150,251, WinZip has messed it up.\r
+- fc sample1.bz2 sample1.rb2 \r
+- fc sample2.bz2 sample2.rb2\r
+- fc sample3.bz2 sample3.rb2\r
+- fc sample1.tst sample1.ref\r
+- fc sample2.tst sample2.ref\r
+- fc sample3.tst sample3.ref\r
+-\r
+-\r
+-\r
+-clean: \r
+- del *.obj\r
+- del libbz2.lib \r
+- del bzip2.exe\r
+- del bzip2recover.exe\r
+- del sample1.rb2 \r
+- del sample2.rb2 \r
+- del sample3.rb2\r
+- del sample1.tst \r
+- del sample2.tst\r
+- del sample3.tst\r
+-\r
+-.c.obj: \r
+- $(CC) $(CFLAGS) -c $*.c -o $*.obj\r
+-\r
+diff -Nru bzip2-1.0.1/manual.ps bzip2-1.0.1.new/manual.ps
+--- bzip2-1.0.1/manual.ps Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual.ps Thu Jan 1 01:00:00 1970
+@@ -1,3808 +0,0 @@
+-%!PS-Adobe-2.0
+-%%Creator: dvips(k) 5.78 Copyright 1998 Radical Eye Software (www.radicaleye.com)
+-%%Title: manual.dvi
+-%%Pages: 39
+-%%PageOrder: Ascend
+-%%BoundingBox: 0 0 596 842
+-%%EndComments
+-%DVIPSCommandLine: dvips -o manual.ps manual.dvi
+-%DVIPSParameters: dpi=600, compressed
+-%DVIPSSource: TeX output 2000.03.23:2343
+-%%BeginProcSet: texc.pro
+-%!
+-/TeXDict 300 dict def TeXDict begin /N{def}def /B{bind def}N /S{exch}N
+-/X{S N}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{dup dup 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 /IE 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 IE N
+-end dup{/foo setfont}2 array copy cvx N load 0 nn put /ctr 0 N[}B /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 dup definefont setfont}B /ch-width{ch-data dup
+-length 5 sub get}B /ch-height{ch-data dup length 4 sub get}B /ch-xoff{
+-128 ch-data dup length 3 sub get sub}B /ch-yoff{ch-data dup length 2 sub
+-get 127 sub}B /ch-dx{ch-data dup length 1 sub get}B /ch-image{ch-data
+-dup 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 /sf 0 N /CharBuilder{save 3 1 roll S dup
+-/base get 2 index get S /BitMaps get S get /ch-data X pop /ctr 0 N ch-dx
+-0 ch-xoff ch-yoff ch-height sub ch-xoff ch-width add ch-yoff
+-setcachedevice ch-width ch-height true[1 0 0 -1 -.1 ch-xoff sub ch-yoff
+-.1 sub]/id ch-image N /rw ch-width 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 dup 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 dup gp add
+-/gp X adv}B /nd{/cp 0 N rw exit}B /lsh{rw cp 2 copy get dup 0 eq{pop 1}{
+-dup 255 eq{pop 254}{dup dup add 255 and S 1 and or}ifelse}ifelse put 1
+-adv}B /rsh{rw cp 2 copy get dup 0 eq{pop 128}{dup 255 eq{pop 127}{dup 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}]dup{bind pop}forall N /D{/cc X dup type /stringtype ne{]
+-}if nn /base get cc ctr put nn /BitMaps get S ctr S sf 1 ne{dup dup
+-length 1 sub dup 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 dup 1 get dup mul exch 0 get dup 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
+-/IE 256 array N 2 string 0 1 255{IE S dup 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 /rulex 0 N /ruley 0 N /v{/ruley
+-X /rulex X V}B /V{}B /RV statusdict begin /product where{pop false[
+-(Display)(NeXT)(LaserWriter 16/600)]{dup length product length le{dup
+-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 rulex ruley false
+-RMat{BDot}imagemask grestore}}{{gsave TR -.1 .1 TR rulex ruley scale 1 1
+-false RMat{BDot}imagemask grestore}}ifelse B /QV{gsave newpath transform
+-round exch round exch itransform moveto rulex 0 rlineto 0 ruley neg
+-rlineto rulex neg 0 rlineto fill grestore}B /a{moveto}B /delta 0 N /tail
+-{dup /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 cmti10 10.95 1
+-/Fa 1 47 df<120FEA3FC0127FA212FFA31380EA7F00123C0A0A77891C>46
+-D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fb cmbxti10 14.4 1
+-/Fb 1 47 df<13FCEA03FF000F13804813C05AA25AA2B5FCA31480A214006C5A6C5A6C5A
+-EA0FE0121271912B>46 D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fc cmsl10 10.95 25
+-/Fc 25 122 df<D901F8160CD907FE161ED91F07163E90263E0380157E017C6D15FC496C
+-6CEC03F8D801F06DEC07F00003D900FCEC1FE04901EFEC7FC00007913AE3E007EF804901
+-E0B5129F000F913A701FF83F0093C7127E48485E604D5A03F01303003F4A495A90C74A5A
+-4D5A4DC7FC0201143E4B5B17FC4C5A003E4948485A4C5A003F01075C6C4A485A020E131F
+-021E49C8FC6C6C48137E02385B3A07C0F001F82603E1E05B3A00FF8003E090263E000714
+-7E90C73A0FC001FF80923A1F8007C1C0923A3F000F80E0033E90381F00F04B013E137003
+-FC4913784A4801FC13384A485B913807E001DA0FC05BDB8003141CEC1F004A495A147E5C
+-4948163C4948010F14384A5C495A130F4948167849C71570137E017C17F0494B13E01201
+-4848EDC0014848020714C04848EE03804848160790C83903E00F00003E170E007E923801
+-F03C486F6C5A48EE3FE00070EE0F803F4975C34C>37 D<EA0780EA0FE0EA1FF0123FA512
+-1FEA0F301200A213701360A213E013C0120113801203EA07001206120E5A5A5A5A5A0C1C
+-7A8919>44 D<EC07F8EC3FFF9138FC0FC0903903F003E0903907C001F0D90F8013F849C7
+-FC013E14FC017E147C017C147E13FC485AA20003157F5B1207A2120F5BA2121F16FF5BA2
+-123FA44848EB01FEA648C7EA03FCA5ED07F8A25A16F0A2150F16E0A3ED1FC0A21680007E
+-143F1600157E123E003F5C4A5AD81F805B000FEB07E06C6C485A2603F03FC7FC3800FFFC
+-EB1FE0283F79BC2D>48 D<157015F014011407143F903803FFE0137FEBFFCFEBF80F1300
+-141F15C0A5143F1580A5147F1500A55C5CA513015CA513035CA513075CA5130F5CA3131F
+-497EB612F8A31D3D78BC2D>I<133C137F5B481380A31400A26C5A137890C7FCB3EA0780
+-EA0FE0121F123FA5121FEA0F601200A213E05BA212015B120390C7FC5A1206120E5A5A12
+-3012705A5A11397AA619>59 D<EC7FC0903803FFF890380FC07E90383E003F496D7E01FF
+-6D7E82A248140782A26C5A137890C7120FA25EA2EC03FF147F903807FF1FEB1FE0D97F80
+-5B3801FE00EA03F8485A4848133F485A003F5D49EC81C048C7FCA2157F48ED03804814FF
+-A2007F5B913903BF0700903880073F3A3FC00E1F8E260FE03C13FC3A03FFF00FF83A007F
+-C003E02A2A7CA82D>97 D<EB3F80EA1FFFA3C6FC137FA291C9FCA55B5BA512015BA4EC07
+-F80003EB3FFF9039F8F80FC09039FBE003E09039FF8001F891C77E5B4848147E49147F5B
+-821780A2120F5B17C0A3167F001F16805BA4EEFF00123F5B4B5AA24B5A5E007F4A5AA24B
+-5A6D495A4BC7FCD87CE0137E39F87001F839F03C07E039E00FFF80260003FCC8FC2A4077
+-BE33>I<EC1FF0ECFFFE903903F01F8090390FC003C0D93F0013E0017E130F49131F0001
+-15F04848EB3FE0485AA24848EB1FC0001FEC0F004990C7FC123FA2485AA412FF90C9FCA9
+-6CEC0380150716006C6C5B151E001F5C6C6C5B6C6C5B6C6C485A3901F80F8026007FFEC7
+-FCEB0FF0242A7AA828>I<EE03F8ED01FFA3ED000F1607A217F0A4160FA217E0A4161FA2
+-17C0A491380FF03FECFFFC902603F81F138090390FC007BF90391F8003FF90387E000149
+-7F0001157F48481500485A120F5B001F5D485A5E5B127FA2150112FF90C75BA41503A25E
+-A37E1507A26C4A5A7F001F141F6C6C133F6C6CEBFFF83B03F001EFFFC03900F80F8F9038
+-3FFE0FD90FF0EBE0002D407ABE33>I<EC3FE0903801FFF8903807E07E90380F801F9039
+-3F000F80017E14C049EB07E0485A12034848EB03F0485AA2121F5B123FA248481307A290
+-B6FCA2D8FF80C8FC90C9FCA87EED01C015036C15806D1307001FEC0F006D131E000F5C6C
+-6C5B6C6C485A3900FC07C0D93FFFC7FCEB07F8242A7BA828>I<ED07F0ED3FFCEDFC1E91
+-3803F03F4A48B4FC4A481380141FEC3F81DA7F0113008102FE137C93C7FCA213015CA513
+-035CA50007B512F8A3260007F0C8FCA3130F5CA5131F5CA5133F5CA5137F91C9FCA55B5B
+-A4EA03FF007F13FEB5FCA229407DBF1C>I<147FEB3FFFA313017FA25CA513015CA51303
+-5CA4ED07F80107EB1FFF9139F0781FC09138F1E00F9139F38007E0ECF70002FE14F0495A
+-5CA25CA24A130F131F4A14E0A4161F133F4A14C0A4163F137F91C71380A4167F5B491500
+-A300015D486C491380B5D8F87F13FCA32E3F7DBE33>104 D<1478EB01FE130314FFA25B
+-14FE130314FCEB00F01400ACEB03F8EA01FF14F0A2EA001F130FA314E0A5131F14C0A513
+-3F1480A5137F1400A55B5BA4EA03FF007F13F0A2B5FC183E7DBD1A>I<143FEB1FFF5BA2
+-13017FA214FEA5130114FCA5130314F8A5130714F0A5130F14E0A5131F14C0A5133F1480
+-A5137F1400A55B5BA4EA03FF007F13F8A2B5FC183F7DBE1A>108
+-D<902707F007F8EB03FCD803FFD91FFF90380FFF80913CE0781FC03C0FE09126E1E00FEB
+-F0073E001FE38007E1C003F090260FE700EBE38002EEDAF70013F802FC14FE02D85C14F8
+-4A5CA24A5C011F020F14074A4A14F0A5013F021F140F4A4A14E0A5017F023F141F91C749
+-14C0A549027F143F4992C71380A300014B147F486C496DEBFFC0B5D8F87FD9FC3F13FEA3
+-47287DA74C>I<903907F007F8D803FFEB1FFF9139E0781FC09138E1E00F3B001FE38007
+-E090380FE70002EE14F014FC14D814F85CA24A130F131F4A14E0A4161F133F4A14C0A416
+-3F137F91C71380A4167F5B491500A300015D486C491380B5D8F87F13FCA32E287DA733>
+-I<EC0FF0ECFFFE903903F01F8090390FC007C049C66C7E013E6D7E01FC6D7E4848804914
+-7C0003157E485A000F157F5B121FA2485AA2007F1680A2170048C85AA54B5AA25E5A6C4A
+-5A7E4B5A5E6C140F6C6C5C4B5A6C6C013EC7FC6C6C5B6C6C485A3900FC0FE090383FFF80
+-D90FF8C8FC292A7BA82D>I<91387F01FE903A7FFF0FFFC09139FE3E03F09238F801F890
+-3A01FFE000FE4B137F6D497F4990C713804A15C04A141FA218E0A20103150F5C18F0A317
+-1F010716E05CA3173F18C0130F4A147F1880A2EFFF004C5A011F5D16034C5A6E495AEE1F
+-C06E495AD93FDC017EC7FC91388F01F8913883FFE0028090C8FC92C9FC137FA291CAFCA4
+-5BA25BA31201487EB512F8A3343A81A733>I<903907F01F80D803FFEB7FE09138E1E1F0
+-9138E387F839001FE707EB0FE614EE02FC13F002D813E09138F801804AC7FCA25C131FA2
+-5CA4133F5CA5137F91C8FCA55B5BA31201487EB512FEA325287EA724>114
+-D<9138FF81C0010713E390381F807F90397C003F8049131F4848130F5B00031407A24848
+-1400A27FA27F6D90C7FCEBFF8014FC6C13FF6C14C015F06C6C7F011F7F13079038007FFE
+-1403140100381300157EA2123C153E157E007C147CA2007E147815F8007F495A4A5A486C
+-485A26F9E01FC7FC38E0FFFC38C01FE0222A7DA824>I<EB0380A4130791C7FCA25BA25B
+-A2133EA2137E13FE12011207001FB512C0B6FCA2D801FCC7FCA312035BA512075BA5120F
+-5BA41407001F130E13C0A4141E141C1380A26D5AA2000F5B14F03807E1E03801FF80D800
+-7EC7FC1A3978B723>I<01FE147F00FFEC7FFF4914FEA20007140300031401A34914FCA4
+-150312074914F8A41507120F4914F0A4150F121F4914E0A2151FA3153F4914C0157F15FF
+-EC01DF3A0FC003BFE09138073FFF3803F01E3801FFF826003FE01380282977A733>I<B5
+-39E007FFF05D17E02707FE000313006C48EB01FC6F5A5E00014A5A5EA24B5A6D13070000
+-92C7FC5D150E6D5B7F5DA25D1480013F5B14815D14C3011F5B02C7C8FCA214CE14EEEB0F
+-FCA25CA26D5A5CA25CA26D5A2C2878A630>I<B500C3B53803FFFCA204FE14F8290FFE00
+-3FE00013C0D807F86D48EB7F000003173E183C150F18386D5E0001141F705B153F4D5A15
+-776D4B5A0000ECE7F04DC7FCEC01C3170E9038FF0383017F5D91380703F85FEC0E01021E
+-5CD93F9C14F002BC6D5A02B813FDDAF8005B4A13FF5F6D5A94C8FC5C4A137E167C6DC7FC
+-1678010E14383E2878A642>I<90B539E007FFF05E18E0902707FE000313006D48EB01FC
+-705A5F01014A5A5F16036E5C0100140794C7FC160E805E805E1678ED8070023F13F05EED
+-81C015C191381FC38015C793C8FC15EF15EEEC0FFCA25DA26E5AA25DA26E5A5DA24AC9FC
+-5C140E141E141C5C121C003F5B5A485B495A130300FE5B4848CAFCEA701EEA783CEA3FF0
+-EA0FC0343A80A630>121 D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fd cmtt12 14.4 10
+-/Fd 10 123 df<EC3FF80103B57E010F14E0013F14F84914FE48B7FC4816C048D9C01F7F
+-270FFE00017F49EB007FD81FF0EC1FF848486E7E160748486E7E4914018300FF81A26D16
+-80177FA36C5AA26CC9FCCAFCA217FF1800A25E5F16035F16075F160F4C5A163F4C5A5F4C
+-5A4B90C7FCED07FE4B5A4B5A4B5A4B5A4B5A4A5B4A90C8FCEC0FFE4A5AEC3FF04A5A4A5A
+-495B4990C9FCEB0FFC495A495A495A4948143F485B4890C8EA7F80EA0FFC485A48B8FC5A
+-B9FCA36C17007E314A7AC93E>50 D<383FFF805AB57EA37E7EEA003FAEED07FC92383FFF
+-8092B512E002C314F802CF8002DF8091B7FCDBF80F1380DBC00113C092C713E04A143F4A
+-EC1FF04A15F84A140F4AEC07FCA217034A15FE1701A318FF83A95F18FEA280170318FC6E
+-140718F86E140FEF1FF06E143F6EEC7FE06EECFFC0DBC0031380EDF01F92B6120002DF14
+-FC02CF5C02C35CD91F8114C090260F807F90C7FC90C7EA0FF8384A7FC83E>98
+-D<923803FFF85D4B7FA38181ED0003AEEC1FF0ECFFFE0103EBFF83010F14E34914F3017F
+-14FB90B7FC48EBF80F48EBC00191C7FC4848143F4848141F5B4848140F491407123F4914
+-03127F5BA312FF90C8FCA97F127FA216077F123F6D140FA26C6C141F6D143F000F157F6C
+-6C14FF01FF5B6C6D5A6CD9F01FEBFFFC6C90B500FB13FE6D02F313FF6D14E3010F14C36D
+-020113FE010101FC14FC9026003FE0C8FC384A7CC83E>100 D<143E147F4A7E497FA56D
+-5B6EC8FC143E91C9FCAC003FB57E5A81A47EC7123FB3B3007FB71280B812C0A56C16802A
+-4A76C93E>105 D<007FB512C0B6FC81A47EC7121FB3B3B3A5007FB712F8B812FCA56C16
+-F82E4978C83E>108 D<EC1FF0ECFFFE01036D7E010F14E04980017F14FC90B67E489038
+-F01FFFECC00748D90001138048486D13C04848EC7FE049143F4848EC1FF049140F003F16
+-F8491407A24848EC03FCA448C8EA01FEAA6D1403007F16FCA26D1407A2003F16F86D140F
+-6C6CEC1FF0A26C6CEC3FE06D147F6C6CECFFC06D5B6CD9C00713806CD9F01F13006C90B5
+-5A6D5C6D5C010F14E001031480010049C7FCEC1FF02F3679B43E>111
+-D<ED07FC3B3FFF803FFF804891B512E0B500C314F802CF8002DF806C90B7FC6CDAF80F13
+-80D8003FD9C00113C092C713E04A143F4AEC1FF04A15F84A140F4AEC07FCA217034A15FE
+-1701A318FF83A95F18FEA280170318FC6E140718F86E140FEF1FF06E143F6EEC7FE06EEC
+-FFC0DBC0031380EDF01F92B6120002DF14FC02CF5C02C35C02C114C0DAC07F90C7FCED0F
+-F892C9FCB3003FB512C04880B67EA36C5C6C5C384F7FB33E>I<903901FFF00F011F9038
+-FE1F8090B612BF000315FF5A5A5A393FFE003F01F01307D87FC0130190C8FC5A48157FA4
+-7EEE3F00D87FC091C7FC13F0EA3FFE381FFFF06CEBFFC06C14FE6C6E7EC615E0013F14F8
+-010780D9003F7F02007F03071380030013C0003EED3FE0007F151F48150F17F06D1407A3
+-7FA26D140F6D15E0161F01FCEC3FC06D14FF9026FFC00F138091B612005E485D013F5C6D
+-14E0D8FC0714802778007FF8C7FC2C3677B43E>115 D<147C14FC497EAD003FB712FC5A
+-B87EA36C5EA2260001FEC9FCB3A6173FA2EF7F80A76E14FF6D16006F5A9238C007FE9138
+-7FF01F92B55A6E5C6E5C6E5C6E1480020149C7FC9138003FF031437DC13E>I<000FB812
+-804817C04817E0A418C001C0C712014C13804C1300EE1FFE4C5AEE7FF06C484A5A4B5BC8
+-485B4B90C7FC4B5A4B5A4B5A4B5A4B5A4A5B4A5B4A90C8FC4A5A4A5A4A5A4A5A4A5A495B
+-495B4990C9FC495A495A4948EC0FC0495A4948EC1FE0485B00075B4890C8FCEA1FFC485A
+-485A90B8FCB9FCA46C17C07E33337CB23E>122 D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fe cmtt12 13.14 31
+-/Fe 31 123 df<ECFFE0010713FC011FEBFF80017F14E048B67E48814815FE489038007F
+-FF01F8130748480101138048489038007FC049EC3FE04848141F90C8120F17F048150717
+-F87F1603A36CC8FC123EC9FCA2160717F0A2160F17E0161F17C0163FEE7F80A2EEFF004B
+-5A15034B5A4B5A4B5AED7FE04B5A4A5B4A90C7FC4A5AEC1FF84A5A4A5A4A5A495BD907FE
+-C8FC495A495A495A495A2601FF80EB01F04890C7EA03F8485A485A485A48B7FC5AB8FCA3
+-7E6C16F02D447BC338>50 D<003FB6FC4815E0B712F882826C816C16802701FC000113C0
+-9238007FE0161FEE0FF0A2160717F81603A6160717F0A2160FEE1FE0163FEE7FC0923801
+-FF80030F130090B65A5E16F08216FEEEFF8017C001FCC7EA7FE0EE1FF0EE07F8160317FC
+-EE01FE1600A217FF177FA717FF17FE16011603EE07FC160FEE3FF8EEFFF0003FB7FC4816
+-E0B812C01780EEFE006C15F86C15C030437DC238>66 D<DA07FC133C91393FFF807E91B5
+-12E00103ECF8FE4914FC4914FF5B90383FFE0390387FF000D9FFC0133F4849131F91C712
+-0F48481407485A49140348481401A2485AA248481400A25BA2007F167C90CAFCA35A5AAD
+-7E7EA37F003F167C17FE7FA26C7EA26C6C140117FC6C7E6D14036C6C15F86C6C14076EEB
+-0FF06C6D131FD97FF0EB7FE0903A3FFE03FFC06DB612806D15006D5C6D5C010014F0023F
+-13C0DA07FCC7FC2F457CC338>I<007FB512F8B7FC16C08216F8826C813A03F8001FFF15
+-07030113806F13C0167FEE3FE0161FEE0FF0A2EE07F8A2EE03FCA21601A217FE1600A417
+-7FAD17FF17FEA4160117FCA2160317F81607A2EE0FF0161FEE3FE0167FEEFFC04B13805D
+-031F1300007FB65AB75A5E5E16C093C7FC6C14F830437DC238>I<007FB712FCB87EA57E
+-D801FCC8FCA9177C94C7FCA6ED07C04B7EA590B6FCA79038FC000FA56F5A92C9FCA7EF0F
+-80EF1FC0AA007FB8FCB9FCA56C178032437DC238>I<91391FF003C091397FFC07E049B5
+-FC010714CF4914EF4914FF5B90387FF81F9038FFE00748EB800191C7FC4848147F485A49
+-143F485A161F485AA249140F123F5BA2127F90C8EA07C093C7FCA35A5AAA923807FFFC4B
+-13FE4B13FF7E7E6F13FE6F13FC9238000FE07F003F151FA27F121F7F163F6C7EA26C6C14
+-7F7F6C6C14FF6C6C5B6E5A6C6D5A90387FF81F6DB6FC6D14EF6D14CF6D148F0101140F90
+-3A007FFC07C0DA1FF0C7FC30457CC338>71 D<007FB612F0B712F8A56C15F0260001FCC7
+-FCB3B3B1007FB612F0B712F8A56C15F0254377C238>73 D<90380FFFFE90B612E0000315
+-F8488148814881A2273FFE000F138001F01301497F49147F4848EC3FC0A290C8121FA448
+-16E0A248150FB3AC6C151FA36C16C0A36D143FA36C6CEC7F806D14FF6D5B01FE130F6CB7
+-1200A26C5D6C5D6C5DC615E0010F49C7FC2B457AC338>79 D<003FB512F04814FEB77E16
+-E0826C816C813A01FC003FFEED07FF03017F81707E163F161F83160FA7161F5F163F167F
+-4C5A5D030790C7FCED3FFE90B65A5E5E5EA282829038FC001FED07FC6F7E150115008282
+-AA18E0EF01F0EF03F8A31783EE3F87263FFFE0ECC7F0486D14FFB56C7F18E07013C06C49
+-6D13806C496D1300CA12FC35447EC238>82 D<003FB8FC481780B9FCA53BFE0007F0003F
+-A9007CEE1F00C792C7FCB3B3A70107B512F04980A56D5C31437DC238>84
+-D<267FFFF090387FFFF0B56C90B512F8A56C496D13F0D801FCC73801FC00B3B3A66D1403
+-00005EA36D14076D5D6E130F6D6C495A6E133F6D6C495A6D6C495AECFF076D90B5C7FC6D
+-5C6D5C6D5C023F13E0020F1380DA03FEC8FC35447FC238>I<D83FFE913803FFE0486C4A
+-13F0B56C4913F8A36C90C76C13F06C486E13E0D80FC09138001F80A36C6CED3F00A86C6C
+-157EA86C6C5DEC01FC4A7EA24A7EA4D800FC9038DF01F8020F1381A3158F021F13C1A201
+-7C5DA2017E14C31507023F13E3A490263E3E035BA2027E13F3A2EC7C01013F14F7A3011F
+-5DEC7800A202F813FFA34A137F010F5DA24A133F6D486DC7FC35447FC238>87
+-D<001FB712F04816F85AA417F090C8121F17E0EE3FC0167F1780EEFF00A24B5A4B5A123E
+-C8485A4B5AA24B5A4B5AA24B5A4BC7FCA24A5A14035D4A5A140F5D4A5A143F5D4A5A14FF
+-92C8FC495A13035C495AA2495A495AA2495A495A17F849C7EA01FC485AA2485A485AA248
+-5A121F5B485A127F90B7FCB8FCA56C16F82E437BC238>90 D<003FB712804816C0B812E0
+-A46C16C06C16802B087A7D38>95 D<EB1FFF90B512E0000314F84814FE4880824881D9F0
+-017F9138003FF0ED0FF815076C486D7E6C481301C87FA21500A4EC03FF147F0107B5FC13
+-1F90B6FC5A000714C048EBE0004890C7FCEA3FF813E0EA7F8090C8FC5A5AA415017E6C6C
+-13036D13076C6CEB3FFFD9FC01B6FC6CB812807E6C816C4A7E6CECF80F6C6CD9C0031300
+-D90FFCC9FC31327AB038>97 D<EA3FFE487EB5FCA37E7EC67EACED3FC0913801FFF80207
+-13FE021F7F4A804A8091B67E03C07F9139FE001FF802F86D7E4A13034A6D7E5C707E4A80
+-188091C8123FA318C0171FA9173F6E1580A2177FA26EECFF00A26E495A6E13034C5A6E49
+-5A02FE133F6E6CB45A92B55A6E5C6E5C6E91C7FC6D6C13FCD91E0313F090C7EA7F803244
+-7FC238>I<EC3FFF49B512E0010714F8011F8049805B90B7FC48EBE001481300485AEA0F
+-F8496D5A4848147C93C7FC485A5B127F90CAFCA35A5AA97E7EA27F123F6D141F6DEC3F80
+-121F6C7E01FC147F6C6CECFF006C6C6C5A6C9038F00FFE91B55A6C6C5C6D5C6D5C010714
+-80010149C7FC9038003FF0293278B038>I<ED1FFF4B7F5DA38181ED003FACEC7F809038
+-03FFF0010F13FC013F13FF4914BF90B7FC5A48EBC07F48EB000FD80FFC7F497F48487F49
+-7F485A167F485A163F90C8FCA25A5AA97E6C157FA36D14FF123F6D5B6C7E5D6C6C5B01FC
+-5B6C6C133F2603FFC0B6FC6C90B712806C03BF13C06D143F6D13FC6D01F814800107D9E0
+-1F1300010090C9FC32447CC238>I<EC3FE0903801FFFC010713FF011F14C04980498090
+-B67E489038E03FFC4890380007FE48481301D80FF86D7E49804848158049143F003F16C0
+-49141F127F90C8FC17E0160F5A90B7FCA617C048CAFC7E7EA27FA26C7EEE07C06C6CEC0F
+-E07F6C7ED807FE141F6DEC3FC06C01C013FF6CD9FC0713806C90B612006D5C6D5C010F5C
+-010314E001001480DA1FFCC7FC2B327AB038>I<ED1FF8ED7FFE4AB5FC020714804A14C0
+-5C5C91387FF07F15809139FF003F804AEB1F00010191C7FC5CA9003FB612FE4881B8FCA4
+-6C5D260001FCC8FCB3AF003FB612E04881A56C5D2A447CC338>I<14F0497E497E497EA4
+-6D5A6D5A6D5A91C8FCAB383FFFFC487FB5FCA37E7EC7FCB3AF007FB612F0B712F816FCA3
+-16F86C15F0264476C338>105 D<387FFFFEB6FCA57EC77EB3B3B1007FB7FCB81280A56C
+-1600294379C238>108 D<023FEB07E03B3FE0FFC01FF8D87FF39038E07FFCD8FFF76D48
+-7E90B500F97F15FB6C91B612806C01C1EBF83F00030100EBE01F4902C013C0A24990387F
+-800FA2491400A349137EB3A73C3FFF03FFE07FFC4801879038F0FFFEB500C76D13FFA36C
+-01874913FE6C01039038E07FFC383080AF38>I<ED7FC03A3FFE01FFF8267FFF0713FEB5
+-001F7F4A805C6C90B67E6CECC07F3B007FFE001FE002F8130F4A805C16075C5CA391C7FC
+-B3A6273FFFFE03B512E0486D4814F0B6008F14F8A36C020714F06C496C14E035307FAF38
+->I<EC7FC0903803FFF8010F13FE497F017F14C090B67E4881489038C07FF8489038001F
+-FC01FC130748486D7E49130148486D7E4980003F168049143F007F16C090C8121FA300FE
+-ED0FE0A96C151FA26C16C0A26D143FA26C6CEC7F80A26C6CECFF006D5B6C6C495A6D1307
+-6CB4EB1FFC6C9038C07FF86C90B55A6C5D6D5C6D5C010F49C7FC010313F89038007FC02B
+-327AB038>I<ED3FC03A3FFE01FFF8267FFF0713FEB5001F7F4A804A806C90B67E6C02C0
+-7F3B007FFE001FF802F86D7E4A13034A6D7E5C707E4A80188091C8123FA318C0171FA917
+-3F6E1580A2177FA26EECFF00A26E495A6E13034C5A6E495A02FE133F6E6CB45A92B55A6E
+-5C6E5C6E91C7FC6E13FC020313F09138007F8092C9FCB0383FFFFE487FB67EA36C91C9FC
+-6C5B32487FAF38>I<EEFFC0277FFFC00713F8B5D8E01F13FC037F13FE92B5FC02E114FF
+-14E36C01E713813A000FEFFC019139FFF000FE4B137C0380130092C8FCA25C5CA25CA25C
+-A45CB2007FB67EB77E82A35E6C5D30307CAF38>114 D<903907FF80F0017FEBF1F848B5
+-12FD000714FF5A5A5AEBFC00D87FE0131F0180130F48C71207481403A5007FEC01F001C0
+-90C7FCEA3FF013FE381FFFF86CEBFFC0000314F8C614FF013F1480010714E0D9003F13F0
+-020013F8ED0FFC1503003CEC01FE007E140000FE15FF167F7EA37F6D14FF16FE01F01303
+-6DEB07FC01FF137F91B512F816F016E04815C0D8FC3F1400010F13FCD8780113E0283278
+-B038>I<EB01E0497E1307AB003FB712804816C0B8FCA46C1680260007F0C8FCB3A4EE03
+-E0EE07F0A66E130F17E06D6C131F6EEB3FC0903901FF81FF92B512806D15006E5B6E5B6E
+-13F002075B020190C7FC2C3D7DBB38>I<D83FFE903803FFE0486C497FB55BA36C806C80
+-C66CEB0007B3A9160FA2161F163F6E137F6D6C13FFDAF007EBFFE091B712F06D16F87F6D
+-14E76D02C713F00100020313E0DA3FF8C8FC35307FAE38>I<000FB712FC4816FE5AA417
+-FC0180C7EA1FF8EE3FF0EE7FE0EEFFC04B13804B13006CC7485AC8485A4B5A4B5A4B5A4B
+-5A4A5B4A90C7FCEC07FC4A5A4A5A4A5A4A5A49485A4990C8FC495A495A495A495A494814
+-7C494814FE485B4890C8FC485A485A485A485A48B7FCB8FCA56C16FC2F2F7DAE38>122
+-D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Ff cmbx12 13.14 53
+-/Ff 53 122 df<923807FFE092B512FC020714FF021F81027F9038007FC0902601FFF0EB
+-0FE04901C0497E4990C7487ED90FFC147F011F824A14FF495AA2137F5CA2715A715A715A
+-EF078094C8FCA7EF07FCB9FCA526007FF0C7123F171FB3B3A2003FB5D8E00FB512F8A53D
+-4D7ECC44>12 D<B7FCAA200A7F9D29>45 D<EA07E0EA1FF8EA3FFCEA7FFEA2B5FCA6EA7F
+-FEA2EA3FFCEA1FF8EA07E01010778F22>I<177817F8EE01FCA21603A2EE07F8A217F016
+-0FA217E0161FA2EE3FC0A21780167FA217005EA24B5AA25E1503A24B5AA25E150FA25E15
+-1FA24B5AA25E157FA24BC7FCA25D1401A25D1403A24A5AA25D140FA24A5AA25D143FA25D
+-147FA24AC8FCA25C1301A25C1303A2495AA25C130FA2495AA25C133FA25C137FA249C9FC
+-A25B1201A2485AA25B1207A25B120FA2485AA25B123FA25B127FA248CAFCA25AA2127CA2
+-2E6D79D13D>I<15F014011407141F147FEB03FF137FB6FCA313FC1380C7FCB3B3B2007F
+-B712E0A52B4777C63D>49 D<ECFFF80107EBFF80013F14F090B612FC4881480101EBFF80
+-2707F8003F13C0D80FE0010F13E0D81F806D13F0003F80D87FF06D13F86D15FC6D7F00FF
+-16FE6D147FA217FF82A36C5A6C5A6C5A6C5AC95A17FEA3EEFFFCA24B13F817F05D17E04B
+-13C017804B13004B5A4B5A5EED7FE04B5A4A5B4A90C7FCEC07FC4A5A4A5A4B131FEC3F80
+-4AC7FC14FE4948143E495AEB07E0495A4948147E49C8FC017E15FE90B7FC4816FC5A5A5A
+-5A5A5AB8FC17F8A430477AC63D>I<EC3FFE0103B512E0010F14FC013F14FF90267FE01F
+-7F9026FF000713E0D801FC6D7FD803F07F486C6D7FD80FFE817F486D80167FA3805C16FF
+-7E91C75B6C5A6C5AD80020495B90C75C5D5F4B5B5F031F90C7FCED3FFC4AB45A49B512E0
+-168016E016FC90C7EA3FFF030713C06F7F6F7F6F7F83707E83A2701380A318C0EA07E0EA
+-1FF8487E487EA2B5FCA31880A25E491600127F494A5A6C485D01E05B001F4A5BD80FFC49
+-5B2707FFC03F13C06C90B65AC64BC7FC013F14F8010714E09026007FFEC8FC32487BC63D
+->I<EE07E0160FA2161F163F167F16FFA25D5D5DA25D5D5DA2157D15FDEC01F915F1EC03
+-E11407EC0FC1EC1F811501143F147E14FC14F8EB01F01303EB07E014C0EB0F80131FEB3F
+-00133E5B13FC485A485A5B1207485A485A90C7FC123E127E5AB912FCA5C80003EBE000AD
+-023FB612FCA536487DC73D>I<D8038015E001E0140301FC143F9039FFE003FF91B612C0
+-17801700A25E5E16F05E5E93C7FC15FC15F001E790C8FC01E0C9FCAAEC1FFC01E1B512C0
+-01E714F001EF14FC9039FFE01FFFDA0007138001FC6D13C001F06D13E04915F0497F17F8
+-C913FC167F17FEA417FFA3EA0FC0EA3FF0487EA2487EA317FEA34914FF6C4815FC5B0180
+-15F86CC74813F07F6C6C4913E0D80FF04913C0D807FC011F13806CB46CB512006C90B512
+-FC6C5D013F14C0010F91C7FC010113F030487AC63D>I<ED7FF8913807FFFE021F6D7E02
+-7F80903A01FFF01FE0010790388003F04948486C7E49486D7ED93FF013074948130F01FF
+-4A7E4849133F5C5A4890C7FCA25A705A48486E5A705A003F92C8FCA3485AA21520913807
+-FFE0021F13FC00FF497F4A6D7EDAFC017F9026FDF0007F4A6D7ED9FFC06D7E4A6D7E8391
+-C7FC8382491680A318C05BA3127FA6123FA27F001F1780A3000F4B1300A26C6C5DA26C6D
+-495A6C6D5C6C6D495A6D6C48485A90263FFC075B6DB65A6D4AC7FC01035C010014F0020F
+-90C8FC32487BC63D>I<121F7F7F13FE90B812E0A45A18C0188018005F5FA25F485E90C8
+-EA07E0007E4B5A5F007C151F4CC7FC167E5E485D15014B5A4B5AC8485A4B5AA24BC8FC15
+-7EA25D1401A24A5A1407A24A5AA2141FA24A5AA2147FA314FFA3495BA45BA55BAA6D5BA2
+-6D90C9FCEB007C334B79C93D>I<EC1FFF49B512F0010714FC011F14FF90263FF00713C0
+-49C77F01FCEC3FF04848EC0FF848481407000782491403000F821601A2121F7FA27F13FE
+-6D140302C05C14F002FC495A6C6D130FDAFF805B9238E01FE06C6E485A9238FCFF806C91
+-B5C7FC6C15FC6C5D7F6D14FE6D806D15C06D81011F81017F81D9FFDF804813072603FE01
+-8048486C804848133F4848010F1480003F8049130148486D6C13C0161F824848140382A2
+-82A2177FA218807F127FEFFF007F6C6C4A5AA2D81FFC4A5A6C6CEC0FF86C6C6CEB3FF06C
+-9039F003FFE06C90B612806C6C92C7FC011F14FC010714E09026003FFEC8FC32487BC63D
+->I<EC1FFE49B512C0010F14F04914FC90397FFC0FFE903AFFE003FF804849C67F48496D
+-7E4890C7FC486F7E484881161F003F825B007F82A28300FF81A31880A518C0A45E127FA3
+-003F5D7F121F5E120F6C6C91B5FC6C90388001EF6CEBC0036C9038E00FCF6DB5128F011F
+-140F010701FE1480010113F8903800010091C7FCA24C1300A3D803F85D487E487E486C4A
+-5AA25F4C5AA24C5A49495B6C485D49010790C7FC01E0495AD807F8EB3FFC6CB448B45A6C
+-90B55A6C15C06D91C8FC011F13FC010313C032487BC63D>I<EE01F8A24C7EA34C7EA24C
+-7EA34C7FA24C7FA34C7FA293B57EA34B8016F303038016E316E103078016C0030F805E83
+-031F814C7E4B81153E83037E81037C7F03FC815D830201824B7F0203825D830207824B7F
+-020F825D84021F8392B8FC4A83A34A83027CC8120F02FC835C840101844A810103845C84
+-0107844A81010F845C85011F85496C82B600C091B712F0A5544D7CCC5D>65
+-D<93261FFF80EB01C00307B500F81303033F02FE13074AB7EAC00F0207EEE03F021F903A
+-FE007FF87F027F01E0903807FCFF91B5C70001B5FC010301FC6E7E4901F0151F4901C081
+-4949814990C97E494882494882485B48197F4A173F5A4A171F5A5C48190FA2485B1A07A2
+-5AA297C7FC91CDFCA2B5FCAD7EA280A2F207C07EA36C7FA26C190F6E18807E6E171F6C1A
+-006E5F6C193E6C6D177E6D6C5F6D6C4C5A6D6D15036D6D4B5A6D01F04B5A6D01FCED3FC0
+-010001FFEDFF806E01E0D903FEC7FC021F01FEEB3FFC020790B612F002015EDA003F92C8
+-FC030714FCDB001F13804A4D79CB59>67 D<B912F0F0FF8019F019FC19FFD8001F902680
+-000114C0DD001F7F060713F806017F726C7E737E737F737F737F8587737F8587A2747EA3
+-8786A21C80A51CC0A586A462A51C80A51C00A26263A2631AFF636163614F5B634F5B073F
+-90C7FC4F5A4F5A06035B061F5B4DB512C0BBC8FC19FC19F0198006F0C9FC524B7CCA5E>
+-I<BB12C0A486D8000F01E0C77E18071801F0007F193F191F190F1907861903A31901A3EF
+-0F80A2737EA497C7FC171FA2173F177F17FF160392B6FCA5EDE0031600177F173F171FA2
+-050FEC0F80A3F21F00A494C8FC621A3EA21A7EA31AFE6219011903A21907190FF13FF819
+-FF1803183FBBFCA262A3494A7CC951>I<BBFCA41A80D8001F01C0C7FC181F1803848419
+-7F193F191F1AC0190FA31907A4171FF103E0A496C7FCA25FA25F5F5E160792B6FCA5EDC0
+-071601828383A283A794C9FCB1B8FCA5434A7CC94D>I<93261FFF80EB01C00307B500F8
+-1303033F02FE13074AB7EAC00F0207EEE03F021F903AFE007FF87F027F01E0903807FCFF
+-91B5C70001B5FC010301FC6E7E4901F0151F4901C0814949814990C97E49488249488248
+-5B48197F4A173F5A4A171F5A5C48190FA2485B1A07A25AA297C8FC91CEFCA2B5FCAD6C04
+-0FB712C0A280A36C93C7001FEBC000A2807EA27E807E807E806C7F7E6D7E6D7E6D7F6D01
+-E05D6D6D5D6D13FC010001FF4AB5FC6E01E0EB07F9021F01FFEB3FF0020791B5EAE07F02
+-01EEC01FDA003FED0007030702F81301DB001F018090C8FC524D79CB61>I<B7D8FC01B7
+-12F8A5D8001F01C0C8001FEBC000B3AA92B9FCA503C0C8121FB3AEB7D8FC01B712F8A555
+-4B7BCA60>I<B712FEA5D8000FEBE000B3B3B3ABB712FEA5274B7DCA2E>I<B8FCA5D8001F
+-01C0C9FCB3B3A4193EA4197E197CA519FCA31801A2F003F8A21807180F181F183F187FEF
+-01FF1707173FBA12F0A53F4B7BCA4A>76 D<B600E04DB612806F5FA26F5FA2D8001F09FC
+-C7FC6FEF0F7FA2DABFFE171EA2DA9FFF173CA3028F6D1678A202876D16F0A202836DED01
+-E0A302816DED03C0A202806DED0780A26F6CED0F00A36F6C151EA26F6C5DA26F6D5CA26F
+-6D5CA36F6D495AA26F6D495AA26F6D495AA3706C49C7FCA2706C131EA2706C5BA3706D5A
+-A2706D5AA270EBE1E0A370EBF3C0A270EBFF80A27190C8FCA2715AA3715AA2715A497EB6
+-00F06D480103B71280A3715A715A694B7BCA74>I<B600E092B612F88181A281D8001F6D
+-9239001FE0006F705A82A28202BF7F029F7FA2028F7F02877F02837F8214810280806F7F
+-6F7F83816F7F6F7F6F7F83816F80707F707F8482707F707F707FA2707F7014807113C019
+-E0837113F07113F87113FC19FE837113FF71148F7213CF1AEF847213FF8484A284848485
+-A2858585A285858585497EB600F8167F1A3F1A1F1A0FA2554B7BCA60>I<EEFFF8031FEB
+-FFC04AB612FC020715FF021FD9C01F13C091277FFE000313F0902601FFF09038007FFC49
+-496E7E490180EC0FFF4990C86C7F49486F7F49486F7F017F8449486F7F4849707EA24849
+-707E4885A24849701380A2481AC04A82A2481AE0A34890CA6C13F0A5B519F8AE6C1AF0A2
+-6E5EA36C1AE0A26E5E6C1AC0A26C1A806E5E6C1A006E5E6C616E16FF6C616C6D4B5B6D6C
+-4B5B6E5D6D6D4A5B6D6D4A5B01076D4A90C7FC6D01F8ECFFFE6D01FE01035B9028007FFF
+-C01F13F0021F90B612C0020793C8FC020115FCDA001F14C0030101FCC9FC4D4D79CB5C>
+-I<B912C018FCF0FF8019F085D8001F902680000713FE05007F063F1380060F13C07213E0
+-1AF0841AF8A27213FCA31AFEA81AFCA34E13F8A21AF0601AE04E13C0063F138095B51200
+-050713FC92B75A19E096C7FC18F803C0CAFCB3ABB712FCA5474B7BCA54>I<B9FC18F8F0
+-FF8019E019F8D8000F9026C0000713FE9439007FFF80061F7F727F727F727F84868684A2
+-86A862A24E5BA2624E5B4E5B4E5B4E5B95B5C8FC050713FC92B712F0198006FCC9FC18FF
+-9226C0003F13C0050713F0717F717F717F187F85727FA28486A786A71C3E86A28474137E
+-72157C726D13FCB700FC6D9038FE01F872EBFF8373EBFFF0071F14E007031480CD383FFE
+-00574C7CCA5C>82 D<DA7FFCEB01C00103B5EAC003011FECF00749ECFC0F90B7121F48D9
+-E00F13BF4890C713FFD807FC141F4848804848140382484880177F485A173F171F12FFA2
+-170F7FA217077F7F7F6D92C7FC6D7E6C13F014FF15F86CECFF8016F86C15FF6C16C0836C
+-826C826C826C82013F816D1680010716C01300020F15E01400030714F0ED007F160F1603
+-7013F882177F127800F8163FA3171FA27E18F0A27EA26CEE3FE07F18C001E0157F6DEDFF
+-8001FC160001FF140302E0EB0FFED97FFEEB3FFC486CB612F0D8FC0F5DD8F803158048C6
+-6C49C7FC48010313F0354D79CB44>I<003FBB12C0A5DA80019038FC001FD9FC001601D8
+-7FF09438007FE001C0183F49181F90C7170FA2007E1907A3007C1903A500FC1AF0481901
+-A5C894C7FCB3B3A749B812FCA54C4A7CC955>I<B700F8023FB512F8A5D8001F01C0C938
+-0FE000745AB3B3AD6D180F63811A1F6D96C7FC626D7F1A7E6D7F6D606E6C4B5A6E6CED07
+-F06E6C4B5A6E01C0EC3FC06E01F049B45A020101FF011F90C8FC6E91B55A033F15F80307
+-15E0030092C9FC040713F0554C7CCA5E>I<B700F00203B6FCA5D8001F01E0C93803FC00
+-745A6D61811A036D6D5F1A076D6D5F1A0F6D616F161FA26D6D94C7FC626E6D153E1A7E6E
+-177C7015FC6E5F8219016E6D5D19036E5F7014076E5F82190F6E6D5D191F6E6D92C8FC61
+-6F153E83197E6F6D137C19FC6F6D5B18016F5DEFF003A26F01F85B18076F01FC5B180F6F
+-5DEFFE1F6F92C9FC17FF607013BE18FE705BA2705BA3705BA2705BA2705BA3705BA27090
+-CAFCA2177EA2584C7ECA5D>I<B700F84AB6FCA5D8001F01F0C93803FC006F705A6D4E5A
+-6D6D4C5A816D4E5A6D6D4C5A826D4EC7FC6E6D5D70157E6E5F6E7F704A5A6E4C5A6E7F70
+-4A5A6E4C5A6E7F71495A6E4CC8FC6F7F71137E6F5D6F7F71485A6F4A5A6F13FC71485A6F
+-4A5A6F13FFF09F806F02BFC9FC7013FF60705B8260705B8260B3A7037FB612FEA5584B7E
+-CA5D>89 D<ECFFFC010FEBFFC0017F14F090B612FC489038803FFE3B03FC0007FF80486C
+-6D7F6D7F486D6C7F83167F83A2707E6C90C7FC6C5A6C5AC9FCA5ED1FFF021FB5FC49B6FC
+-130F013FEBC03F9038FFFE00000313F04813C04890C7FC485A485AA2485AA2485AA4167F
+-A26D14FF007F15EF6D01017F6C6C903907CFFF806C6CD90F8F13FE6C9038E07F076C9038
+-FFFE0300014A7ED8003F9038F0007F0103018090C7FC37347CB23C>97
+-D<EB7FC0B5FCA512037EB3ED07FF037F13E002C1B512FC02C714FF9126CFF80F7F9126DF
+-C0017F02FFC77F4AEC3FF002F8814A6E7E717E4A81831980A37113C0A319E0AC19C0A35F
+-1980A219005F606E141F6E5D4D5A6E4A5A02BF4A5A91261FC0035B9027FE0FF01F5B496C
+-B548C7FCD9F80114F8496C6C13E0C8D80FFEC8FC3B4D7CCB44>I<91380FFF8091B512F8
+-010314FF010F15804948C613C0D97FF8EB1FE0D9FFE0EB3FF04849137F4849EBFFF84890
+-C7FCA2485A121FA24848EC7FF0EE3FE0EE1FC0007F92C7FC5BA212FFAC127FA27FA2123F
+-A26C6C153EA26C6C157E177C6C6D14FC6C6D14F86C6D13036C6DEB07F0D97FFCEB1FE06D
+-B4EBFFC0010F90B5120001035C010014F0020F13802F347CB237>I<EF1FF0EE3FFFA516
+-00177FB3EC0FFF91B512E0010314F8010F14FC013FEB01FF903A7FF8003FFFD9FFE0130F
+-48497F48497F4890C77E48815B121F5B123FA2127F5BA312FFAC127FA36C7EA3121F6D5C
+-120F6C6C5C6C6D5B6C5D6C01E0013F7F6D6C49EBFFE090393FFE03FE6DB512F801075C01
+-00148091261FFC00EBC0003B4D7CCB44>I<EC0FFF91B512F0010314FC010F14FF90263F
+-FE077F90267FF0007F49486D7E4801806D7E486F7E4890C7120F484881707E121F491403
+-003F82A2127F5B701380A212FFA390B8FCA401F8CAFCA5127FA27FA2123FA26C6CED0F80
+-A26C6C151F6C17006E5C6C6D147E6C6D5C6C6D495AD93FFCEB07F0903A1FFF803FE00107
+-90B55A010192C7FC6D6C13FC020713C031347DB238>I<ED7FF0913807FFFC023F7F4A7F
+-902601FFE0138049018113C049484813E0EB0FFCEB1FF8A2EB3FF0A2017F6D13C04A6C13
+-80EE7F00163E93C7FCACB77EA526007FF0C8FCB3B3A3003FB512F8A52B4D7DCC26>I<DA
+-7FFCEB0FE00103B538807FF8010F9138E1FFFC013F02FB13FE903A7FF01FFFF39026FFC0
+-07138348D90001130348168348486DEBC1FCEFC0F8000FEEE00049147F001F82A9000F5E
+-6D14FF00075EA26C6C495B6C93C7FC6C9038C007FE9138F01FFC4890B55A01EF14E0D803
+-C314809026C07FFCC8FC000790CAFCA47FA27F13FC90B612FCEEFFC06C16F817FE6C8218
+-806C17C06D16E00003B812F0120FD81FFCC7000F13F8D83FF0140049153F4848ED1FFC00
+-FF160F491507A56D150F007F17F86D151F6C6CED3FF06C6CED7FE0D80FFE913801FFC06C
+-6C6C010713806C01F8017F1300C690B612FC013F15F001071580D9003F01F0C7FC37497D
+-B13D>I<EB7FC0B5FCA512037EB3923801FFC0030F13F8033F7F4B13FFDBFE077F9138C1
+-F0039126C3E0017FDAC78080ECCF0014DE02DC6D7F14FC5CA25CA35CB3AAB6D8C07FEBFF
+-E0A53B4C7BCB44>I<13FCEA03FF487F487FA2487FA66C5BA26C5B6C90C7FCEA00FC90C8
+-FCABEB7FC0B5FCA512037EB3B3A2B61280A5194D7BCC22>I<EB7FC0B5FCA512037EB3B3
+-B3AAB61280A5194C7BCB22>108 D<90287FC001FFC0EC7FF0B5010F01FC0103B5FC033F
+-6D010F804B6D4980DBFE079026803F817F9126C1F801903AC07E007FF00003D9C3E0DAE0
+-F8806C9026C78000D9F1E06D7E02CFC7EBF3C002DEEDF780DD7FFF6E7E02FC93C7FC4A5D
+-A24A5DA34A5DB3AAB6D8C03FB5D8F00FB512FCA55E327BB167>I<903A7FC001FFC0B501
+-0F13F8033F7F4B13FFDBFE077F9138C1F00300039026C3E0017F6CD9C78080ECCF0014DE
+-02DC6D7F14FC5CA25CA35CB3AAB6D8C07FEBFFE0A53B327BB144>I<913807FF80027F13
+-F80103B6FC010F15C090261FFE017F903A7FF0003FF849486D7E480180EB07FE4890C76C
+-7E4817804980000F17C048486E13E0A2003F17F0A249157F007F17F8A400FF17FCAB007F
+-17F8A46C6CEDFFF0A2001F17E0A26C6C4A13C0A26C6C4A13806C6D4913006C5E6C01E0EB
+-1FFC6D6C495A903A3FFE01FFF0010FB612C0010392C7FCD9007F13F80207138036347DB2
+-3D>I<90397FC007FFB5017F13E002C1B512FC02C714FF9126CFF80F7F9126DFC0037F00
+-0301FFC77F6C496E7E02F8814A6E7E717E4A81831980A28319C0A37113E0AC19C05FA319
+-805F19005F606E143F6E5D4D5A6E4A5A02FF495BDBC0075B9126EFF01F5B02E7B548C7FC
+-02E114F8DAE07F13E0DB0FFEC8FC92CAFCAFB612C0A53B477CB144>I<9039FF803FE0B5
+-EBFFF8028113FE02837FDA87E11380EC8F830003D99F0713C06C139E14BCA214F8A24A6C
+-13806F13006F5A4A90C7FCA45CB3A8B612E0A52A327CB132>114
+-D<903907FF8070017FEBF1F048B6FC1207380FFC01391FE0003F4848130F491307127F90
+-C71203A2481401A27FA27F01F090C7FC13FCEBFFC06C13FEECFFE06C14FC6C806CECFF80
+-6C15C06C15E06C15F06C7E011F14F8010114FCEB000FEC007FED1FFE0078140F00F81407
+-15037E1501A27E16FC7E15036D14F86D13076D14F001F8EB1FE001FFEBFFC04890B51280
+-486C1400D8F81F13FCD8E00313C027347CB230>I<14F8A51301A41303A21307A2130FA2
+-131F133F137F13FF1203000F90B512F0B7FCA426007FF8C7FCB3A7167CAA013F14F880A2
+-90391FFE01F0010F1303903907FF87E06DEBFFC06D14806D6C1300EC0FFC26467EC430>
+-I<D97FE0EC3FF0B5EC7FFFA5000315016C81B3AC5EA25EA25E7E6EEB0F7F017F021E7F6E
+-017CEBFFE090393FFE01F86DB512F0010714E0010114C09027003FFE00EBC0003B337BB1
+-44>I<B600C090387FFFF0A5C601F0C73803F8006E5D017F5E801707013F5E6E140F011F
+-5E6E141F6D93C7FC6F5B6D153E81177E6D157C6F13FC6D5DEDF0016D5DEDF803027F5CA2
+-EDFC07023F5CEDFE0F021F5CEDFF1F6E91C8FC169F16BF6E13BE16FE6E5BA26E5BA26E5B
+-A36F5AA26F5AA26F5AA23C327EB041>I<B60081B500FC90387FFFF0A500019026E00003
+-90C73801FC006E6F5C6C6F5E6E6F1303017F61A26E6F1307013F4A5E846E49150F011F61
+-6E6F131F6D4A93C7FC179F03806E5B6D021F153EEF0FFE6DD9C03F5DEE3E0703E06E13FC
+-6D027E5DEE7C0303F015816D4A6C5C03F815C3DA7FF95E4C7E03FD15E7DA3FFF5E4C137F
+-19FF6E496D5BA26E94C8FC4C7FA26E5E93C7120FA26E486E5AA202015E4B1403A26E486E
+-5A54327EB059>I<007FB500C090387FFFE0A5C601F0C73803F8006E5D017F5E6E140701
+-3F5E80170F011F5E6E141F6D93C7FC6F5B6D153E6F137E6D157C6F13FCA26D6D5B16016D
+-5DEDF803027F5CEDFC07023F5CEDFE0F021F5C15FF161F6E91C8FC16BF6E13BE16FE6E5B
+-A26E5BA36E5BA26F5AA26F5AA26F5AA393C9FC5D153E157E157CD81F8013FC486C5B387F
+-E001D8FFF05B14035D14074A5A49485A007F133F4948CAFC383F81FE381FFFF86C5B6C13
+-C0C648CBFC3B477EB041>121 D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fg cmtt12 17.28 6
+-/Fg 6 123 df<913803FFC0023F13FC49B67E010715F04981013F15FE498190B812C048
+-8348D9FC0180489026E0001F7F480180130391C87F48486F7E49153F4848ED0FFF834848
+-178083496F13C012FF8319E07FA2187FA36C5A6C5A6C5ACBFCA218FFA219C05FA219805F
+-A24D13005F604D5A173F4D5A4D5AA24C5B4C5B4C5B041F90C7FC4C5A4C5A4C5A4B5B4B5B
+-4B5B031F5B4B90C8FC4B5AEDFFF84A5B4A5B4A5B021F5B4A90C9FCEC7FFC4A5A495B495B
+-010F5B495B4948CAFC4948ED1F804948ED3FC04849ED7FE0485B000F5B4890C9FC4890B8
+-FC5ABAFCA56C18C06C18803B5A79D94A>50 D<383FFFF0487F80B5FCA37EA27EEA000FB0
+-EE0FFC93B57E030714E0031F14F84B14FE92B7FC02FD8291B87E85DCE01F7FEE000703FC
+-01017F4B6D7F03E0143F4B6E7E4B140F8592C87E4A6F1380A34A6F13C0A284A21AE0A219
+-7FAA19FFA21AC0A26E5DA24E138080606F1600606F4A5A6F143F6F4A5A6F4A5A6F130303
+-FF010F5BDCC03F5B93B65A6102FD93C7FC02FC5D6F5C031F14F0902607F80714C0902603
+-F00191C8FC90C8EA3FF043597FD74A>98 D<EC01F0EC07FC4A7EA24A7EA56E5AA26E5AEC
+-01F091CAFCAE001FB512FC4880815AA37EA27EC71203B3B3A8003FB812C04817E0A2B912
+-F0A36C17E0A26C17C0345974D84A>105 D<003FB512FE4880B77EA57E7EC71201B3B3B3
+-B0003FB812FC4817FEBAFCA56C17FE6C17FC385877D74A>108 D<EE0FFC263FFFF090B5
+-7E48D9F80714E0DAFC1F14F8B54914FE92B7FC02FD826C90B87E856CDBE01F7FD8000FEC
+-000703FC01017F4B6D7F03E0143F4B6E7E4B140F8592C87E4A6F1380A34A6F13C0A284A2
+-1AE0A2197FAA19FFA21AC0A26E5DA24E138080606F1600606F4A5A6F143F6F4A5A6F4A5A
+-6F130303FF010F5BDCC03F5B93B65A6102FD93C7FC02FC5D6F5C031F14F0030714C00301
+-91C8FC9238003FF093CAFCB3A4003FB6FC4881A2B77EA36C5DA26C92CAFC435F7FBE4A>
+-112 D<000FB912E04818F04818F8A619F001F0C8000313E04D13C04D13804D13004D5A4D
+-5A4D5A6C484A5B6C484A5BC9000F5B4C5B4C90C7FC4C5A4C5A4B5B4B5B4B5B4B5B4B5B4B
+-90C8FC4B5A4B5A4A5B4A5B4A5B4A5B4A5B4A90C9FC4A5A4A5A495B495B495B4949EC07E0
+-4949EC0FF04948C8EA1FF8495A495A485B485B485B485B4890C9FC485A48B9FCBAFCA66C
+-18F06C18E03D3E7BBD4A>122 D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fh cmbx12 17.28 28
+-/Fh 28 120 df<16F04B7E1507151F153FEC01FF1407147F010FB5FCB7FCA41487EBF007
+-C7FCB3B3B3B3007FB91280A6395E74DD51>49 D<913801FFF8021FEBFFC091B612F80103
+-15FF010F16C0013F8290267FFC0114F89027FFE0003F7F4890C7000F7F48486E7FD807F8
+-6E148048486E14C048486E14E048486F13F001FC17F8486C816D17FC6E80B56C16FE8380
+-A219FFA283A36C5BA26C5B6C90C8FCD807FC5DEA01F0CA14FEA34D13FCA219F85F19F04D
+-13E0A294B512C019804C14004C5B604C5B4C5B604C13804C90C7FC4C5A4C5A4B13F05F4B
+-13804B90C8FC4B5AED1FF84B5A4B5A4B48143F4A5B4A48C8FC4A5A4A48157E4A5A4A5AEC
+-7F8092C9FC02FE16FE495A495A4948ED01FCD90FC0150749B8FC5B5B90B9FC5A4818F85A
+-5A5A5A5ABAFCA219F0A4405E78DD51>I<92B5FC020F14F8023F14FF49B712C04916F001
+-0FD9C01F13FC90271FFC00077FD93FE001017F49486D8049C86C7F484883486C6F7F14C0
+-486D826E806E82487FA4805CA36C5E4A5E6C5B6C5B6C495E011FC85A90C95CA294B55A61
+-4C91C7FC604C5B4C5B4C5B4C5B047F138092260FFFFEC8FC020FB512F817E094C9FC17F8
+-17FF91C7003F13E0040713F8040113FE707F717F7113E085717FA2717F85A285831A80A3
+-1AC0EA03FCEA0FFF487F487F487FA2B57EA31A80A34D14005C7E4A5E5F6C495E49C8485B
+-D81FF85F000F5ED807FE92B55A6C6C6C4914806C01F0010791C7FC6C9026FF803F5B6D90
+-B65A011F16F0010716C001014BC8FCD9001F14F0020149C9FC426079DD51>I<F01F804E
+-7E187F18FFA25F5F5F5FA25F5F5FA294B5FC5E5E5EA25E5EEE3FBFEE7F3FA216FEED01FC
+-ED03F8ED07F0A2ED0FE0ED1FC0ED3F8016005D15FE4A5A4A5AA24A5A4A5A4A5A4A5AA24A
+-C7FC14FE495A5C1303495A495A495A5C133F49C8FC13FE485AA2485A485A485A5B121F48
+-5A48C9FC12FEBCFCA6CA6CEBC000B1037FB8FCA6485E7CDD51>I<4DB5ED03C0057F02F0
+-14070407B600FE140F047FDBFFC0131F4BB800F0133F030F05FC137F033F9127F8007FFE
+-13FF92B6C73807FF814A02F0020113C3020702C09138007FE74A91C9001FB5FC023F01FC
+-16074A01F08291B54882490280824991CB7E49498449498449498449865D49498490B5FC
+-484A84A2484A84A24891CD127FA25A4A1A3F5AA348491A1FA44899C7FCA25CA3B5FCB07E
+-A380A27EA2F50FC0A26C7FA37E6E1A1F6C1D80A26C801D3F6C6E1A00A26C6E616D1BFE6D
+-7F6F4E5A7F6D6D4E5A6D6D4E5A6D6D4E5A6D6E171F6D02E04D5A6E6DEFFF806E01FC4C90
+-C7FC020F01FFEE07FE6E02C0ED1FF8020102F8ED7FF06E02FF913803FFE0033F02F8013F
+-1380030F91B648C8FC030117F86F6C16E004071680DC007F02F8C9FC050191CAFC626677
+-E375>67 D<B96C90B91280A6D8000102C0C9000102C0C7FCB3B3A293BBFCA604C0C91201
+-B3B3A6B96C90B91280A671627AE17E>72 D<B912E0A6C702E0C7FCB3B3B3B3AEB912E0A6
+-33627CE13C>I<B700C0083FB612F070627097B7FCA37061D800010DF8C7FC70F103EFA2
+-02FD6DF107CFA202FC6DF10F8FA36F6DF01F0FA26F6D183EA26F6D187CA26F6D18F8A36F
+-6DEF01F0A26F6DEF03E0A26F6DEF07C0A26F6DEF0F80A3706DEE1F00A2706D163EA2706D
+-5EA2706D5EA3706D4B5AA2706D4B5AA2706D4B5AA2706D4B5AA3716D4AC7FCA2716D143E
+-A2716D5CA2716D5CA3716D495AA2716D495AA2716D495AA2716D495AA3726D48C8FCA272
+-EBC03EA2726D5AA2726D5AA372EBF9F0A272EBFFE0A2725CA2725CA37390C9FCA2735AA2
+-735A90381FFFC0B700F86E480207B812F0A3735AA2735A8C627AE199>77
+-D<BB7E1AFCF2FFC01BF81BFE757ED800010280C7001F80070114F0736C7F081F7F747F74
+-7F7414807414C0A27414E0A21DF0A27513F8A41DFCA91DF8A498B512F0A21DE0A25014C0
+-1D8062501400505B505B087F5B4FB512E0071F5C93B9C7FC1BFC1BF01B8008F0C8FC04C0
+-CCFCB3B3A2B97EA65E627AE16E>80 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>I<F17FF805
+-0FB5FCA6EF000F8484B3A892380FFF804AB512F8020F14FE023FECFF8391B712E3010391
+-38807FF3499039F8000FFB011F01E00103B5FC494913004990C87E49488148498148834A
+-815A485BA2485BA25AA3485BA4B5FCAE7EA46C7FA37EA26C7FA26C5F806C5F6C6D5D6C6D
+-5D017F93B5FC6D6C6C0103806D6D49806D01F0D91FF7EBFFFE6D9039FE01FFE7010190B6
+-12876D6CECFE07021F14F8020314E09127003FFE00ECC0004F657BE35A>I<92380FFFC0
+-4AB512FC020FECFF80023F15E091B712F80103D9FE037F499039F0007FFF011F01C0011F
+-7F49496D7F4990C76C7F49486E7F48498048844A804884485B727E5A5C48717EA35A5C72
+-1380A2B5FCA391B9FCA41A0002C0CBFCA67EA380A27EA27E6E160FF11F806C183F6C7FF1
+-7F006C7F6C6D16FE6C17016D6C4B5A6D6D4A5A6D01E04A5A6D6DEC3FE0010301FC49B45A
+-6D9026FFC01F90C7FC6D6C90B55A021F15F8020715E0020092C8FC030713F041437CC14A
+->I<EE3FFC0307B51280033F14C04AB612F0020715F84A9038F03FFC4AEB807F913A7FFE
+-00FFFE4A5A4B4813FF4913F05B4913E0A24913C0A27013FE4949EB7FFCEF3FF8EF1FF0EF
+-07C094C7FCB0B812C0A6D8001F01C0C8FCB3B3B0007FB612FCA638657CE431>I<F107F8
+-DB7FFEEC3FFE020FB5D8F001B5FC027FDAFE03148049B7128F49DCDFFD13C0010FD9F00F
+-EBFFC149D9800114014990C7EBFC0349486E6C7E4948EC3FFF48496E0181138007801300
+-48F0C03E97C7FC48496E7FA34884A96C60A36C6D4A5BA26C60A26C6D4A90C8FC6D6C4A5A
+-6D6C4A5A6D6D485BDBF00F5B4990B612C060D97C7F4AC9FCD9FC0F14F09126007FFECAFC
+-92CCFC1201A47FA27F8014F091B77E18FE6CEFFFC019F06D17FC19FF6D846D846D846D84
+-013F8490BAFC0003854801E0C712014890C9000F7F484816014848EE007F4848717E8512
+-FF5B85A56D5F007F616D173F003F616D177F6C6C4D5A6C01C003035B6C6D4B5B6C01F803
+-1F5BC601FF92B5C7FC6D01F8011F5B011F90B712F8010717E0010094C8FC020F15F0DA00
+-3F01FCC9FC4A607CC151>I<903807FF80B6FCA6C6FC7F7FB3A8EF1FFF94B512F0040714
+-FC041F14FF4C8193267FE07F7F922781FE001F7FDB83F86D7FDB87F07FDB8FC0814C7F03
+-9FC78015BE03BC8003FC825DA25DA25DA45DB3B2B7D8F007B71280A651647BE35A>I<EB
+-0FE0EB3FF8497E48B5FCA24880A24880A76C5CA26C91C7FCA238007FFC6D5AEB0FE090C9
+-FCAF903807FF80007FB5FCA6C6FC7F7FB3B3AEB712C0A622657BE42C>I<903807FF80B6
+-FCA6C6FC7F7FB3B3B3B3ADB712E0A623647BE32C>108 D<902607FF80D91FFFEEFFF8B6
+-91B500F00207EBFF80040702FC023F14E0041F02FF91B612F84C6F488193267FE07F6D48
+-01037F922781FE001F9027E00FF0007FC6DA83F86D9026F01FC06D7F6DD987F06D4A487F
+-6DD98FC0DBF87EC7804C6D027C80039FC76E488203BEEEFDF003BC6E4A8003FC04FF834B
+-5FA24B5FA24B94C8FCA44B5EB3B2B7D8F007B7D8803FB612FCA67E417BC087>I<902607
+-FF80EB1FFFB691B512F0040714FC041F14FF4C8193267FE07F7F922781FE001F7FC6DA83
+-F86D7F6DD987F07F6DD98FC0814C7F039FC78015BE03BC8003FC825DA25DA25DA45DB3B2
+-B7D8F007B71280A651417BC05A>I<923807FFE092B6FC020715E0021F15F8027F15FE49
+-4848C66C6C7E010701F0010F13E04901C001037F49496D7F4990C87F49486F7E49486F7E
+-48496F13804819C04A814819E048496F13F0A24819F8A348496F13FCA34819FEA4B518FF
+-AD6C19FEA46C6D4B13FCA36C19F8A26C6D4B13F0A26C19E06C6D4B13C0A26C6D4B13806C
+-6D4B13006D6C4B5A6D6D495B6D6D495B010701F0010F13E06D01FE017F5B010090B7C7FC
+-023F15FC020715E0020092C8FC030713E048437CC151>I<D90FFFEB0FFCB690383FFF80
+-93B512E04B14F04B14F8923907FC7FFC92390FE0FFFEC6EC1F806DD93F0113FF6D133E15
+-7E157C15F8A215F07013FEA24BEB7FFCEF3FF8EF0FE04B90C7FCA55DB3B0B712F8A63841
+-7BC042>114 D<913A3FFF8007800107B5EAF81F011FECFE7F017F91B5FC48B8FC48EBE0
+-014890C7121FD80FFC1407D81FF0801600485A007F167F49153FA212FF171FA27F7F7F6D
+-92C7FC13FF14E014FF6C14F8EDFFC06C15FC16FF6C16C06C16F06C826C826C826C82013F
+-1680010F16C01303D9007F15E0020315F0EC001F1500041F13F81607007C150100FC8117
+-7F6C163FA2171F7EA26D16F0A27F173F6D16E06D157F6D16C001FEEDFF806D0203130002
+-C0EB0FFE02FCEB7FFC01DFB65A010F5DD8FE0315C026F8007F49C7FC48010F13E035437B
+-C140>I<EC07E0A6140FA5141FA3143FA2147FA214FF5BA25B5B5B5B137F48B5FC000F91
+-B512FEB8FCA5D8001F01E0C8FCB3AFEF0FC0AC171F6D6D1480A2173F6D16006F5B6D6D13
+-7E6D6D5B6DEBFF836EEBFFF86E5C020F14C002035C9126003FFCC7FC325C7DDA3F>I<90
+-2607FFC0ED3FFEB60207B5FCA6C6EE00076D826D82B3B3A260A360A2607F60183E6D6D14
+-7E4E7F6D6D4948806D6DD907F0ECFF806D01FFEB3FE06D91B55A6E1500021F5C020314F8
+-DA003F018002F0C7FC51427BC05A>I<B70081B600FC0103B512FCA6C66C0180C701FCC8
+-381FFE006F6FED03F86D047F5F856F6E16076D646F70140F6D705F866F6E161F6D646F4A
+-6D143F6D99C7FC4E7F6F616D1B7E6F4A6D14FE6D6395B57E7001FC15016E62DCC0016E13
+-036EDBF87F5D05038004E0496C14076E62DCF007ED800F6E4B6C5D050F15C004F8496C14
+-1F6E62DCFC1FEDE03F6E4B6C92C8FC053F15F004FE496C5C6E197E7048EDF8FE6E027E6D
+-5C05FE15FC4D6D13FD6F601BFF6F496E5BA24D806F60A26F496E5BA24D806F60A26F496E
+-90C9FCA294C87E6F5FA26F486F5A047C6F5A6E417DBF75>119 D
+-E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fi cmsy10 10.95 1
+-/Fi 1 16 df<EB0FFCEB3FFF90B512C0000314F04880488048804880A2481580A3B712C0
+-AA6C1580A36C1500A26C5C6C5C6C5C6C5CC614C0013F90C7FCEB0FFC22227BA72D>15
+-D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fj cmtt10 10.95 89
+-/Fj 89 127 df<121C127FEAFF80B3EA7F00B2123EC7FCA8121C127FA2EAFF80A3EA7F00
+-A2121C09396DB830>33 D<00101304007C131F00FEEB3F80A26C137FA248133FB2007E14
+-00007C7F003C131E00101304191C75B830>I<903907C007C0A2496C487EA8011F131FA2
+-02C05BA3007FB7FCA2B81280A36C16006C5D3A007F807F80A2020090C7FCA9495BA2003F
+-90B512FE4881B81280A36C1600A22701FC01FCC7FCA300031303A201F85BA76C486C5AA2
+-29387DB730>I<EB07E0EB1FF8497E137F497E803801FC7F497E810003131F13F0A6143F
+-92C8FC91387F0FFF9026F87E1F1380000113FEEBF9FC13FB4A6C1300D9FFF013C06C13E0
+-151F02C05BEB7F809038FF003F4892C7FC485C48EB807E5A15FE391FDFC0FC383F8FE014
+-E1397F07F1F8EB03F300FEEBFBF0EB01FF5D7FEDC006027F130F91393F801F8015C06C13
+-7F6CEBFFE049EBF83F018701FC1300263FFFFBB5FC6C01F15B14E06C9038C03FFC000390
+-38001FF8D801FCEB07E0293A7DB830>38 D<EA07C0EA0FF0EA1FF8A213FCA213FE120F12
+-07EA007EA513FE13FCA2120113F81203EA07F0120FEA1FE0127FEAFFC013801300127C12
+-380F1D70B730>I<141E147F14FF5BEB03FEEB07FCEB0FF0EB1FE0EB3FC0EB7F80EBFF00
+-485A5B12035B485A120F5BA2485AA2123F5BA2127F90C7FCA412FEAD127FA47F123FA27F
+-121FA26C7EA27F12076C7E7F12017F6C7EEB7F80EB3FC0EB1FE0EB0FF0EB07FCEB03FEEB
+-01FF7F147F141E184771BE30>I<127812FE7E7F6C7E6C7EEA0FF06C7E6C7E6C7E6C7EEB
+-7F80133F14C0131FEB0FE014F01307A2EB03F8A214FC1301A214FE1300A4147FAD14FEA4
+-130114FCA2130314F8A2EB07F0A2130F14E0EB1FC0133F1480137FEBFF00485A485A485A
+-485AEA3FE0485A485A90C7FC5A1278184778BE30>I<14E0497E497EA60038EC0380007E
+-EC0FC0D8FF83EB3FE001C3137F9038F3F9FF267FFBFB13C06CB61280000FECFE00000314
+-F86C5C6C6C13C0011F90C7FC017F13C048B512F04880000F14FE003FECFF80267FFBFB13
+-C026FFF3F913E09038C3F87F0183133FD87E03EB0FC00038EC0380000091C7FCA66D5A6D
+-5A23277AAE30>I<143EA2147FAF007FB7FCA2B81280A36C1600A2C76CC8FCAF143EA229
+-297DAF30>I<EA03E0EA0FF0EA1FF813FCEA3FFEA213FFA27EA27E1203EA007FA2137E13
+-FEEA01FC1203EA07F8EA3FF0127FEAFFE0EA7F801300123C1019708B30>I<007FB612F0
+-A2B712F8A36C15F0A225077B9E30>I<120FEA3FC0EA7FE0A2EAFFF0A4EA7FE0A2EA3FC0
+-EA0F000C0C6E8B30>I<16F01501ED03F8A21507A2ED0FF0A2ED1FE0A2ED3FC0A2ED7F80
+-A2EDFF00A24A5AA25D1403A24A5AA24A5AA24A5AA24A5AA24A5AA24AC7FCA2495AA25C13
+-03A2495AA2495AA2495AA2495AA2495AA249C8FCA2485AA25B1203A2485AA2485AA2485A
+-A2485AA2485AA248C9FCA25AA2127CA225477BBE30>I<14FE903807FFC0497F013F13F8
+-497F90B57E48EB83FF4848C6138049137F4848EB3FC04848EB1FE049130F001F15F04913
+-07A24848EB03F8A290C712014815FCA400FEEC00FEAD6C14016C15FCA36D1303003F15F8
+-A26D1307001F15F0A26D130F6C6CEB1FE0A26C6CEB3FC06C6CEB7F806D13FF2601FF8313
+-006CEBFFFE6D5B6D5B010F13E06D5BD900FEC7FC273A7CB830>I<EB03C0497EA2130FA2
+-131FA2133F137F13FF1203123FB5FCA213EF138FEA7E0F1200B3B0003FB512F84814FCB6
+-12FEA26C14FC6C14F81F3977B830>I<EB07FC90383FFFC090B512F00003804814FE4880
+-261FF80F1380263FE00113C09038C0007F4848EB3FE090C7121FED0FF04814075A6C15F8
+-1503A3127E1218C8FCA2150716F0150F16E0151F16C0153FED7F8015FF4A13005DEC07FC
+-4A5A4A5A4A5A4A5A4A5A4990C7FC495A495AEB0FF0EB3FE0495A495A4890C8FC4848EB01
+-F04848EB03F8485AEA1FE048B6FCB7FCA37E6C15F025397BB830>I<EB03FF013F13E090
+-B512F84814FE4880481580260FFE0113C09038F0007F4848EB1FE0150F16F01507A26C5A
+-6C5AC8FC150F16E0A2151FED3FC0157FEDFF8002071300903807FFFE495B5D8115FF6D14
+-80D9000113C09138003FE0ED1FF0ED07F8150316FC150116FE1500A21218127EB4FCA215
+-0116FC4814036C15F86C6C13076DEB1FF0D83FF0133F3A1FFE01FFE06CB612C06C15806C
+-ECFE00C65C013F13F001031380273A7CB830>I<EC03FC4A7E140F141FA2143F147F157E
+-14FEA2EB01FCEB03F8A2EB07F0A2EB0FE0EB1FC0A2EB3F80A2EB7F0013FEA2485A485AA2
+-485AA2485A485AA2485AA248C7FC12FEB8FC1780A46C1600C8007EC7FCAA91387FFFFE91
+-B6FCA46E5B29397DB830>I<000FB612804815C05AA316800180C8FCAEEB83FF019F13C0
+-90B512F015FC8181D9FE0313809039F0007FC049133F0180EB1FE06CC7120F000E15F0C8
+-1207A216F81503A31218127EA2B4FC150716F048140F6C15E06C141F6DEB3FC06D137F3A
+-3FE001FF80261FFC0F13006CB55A6C5C6C5C6C14E06C6C1380D90FFCC7FC25397BB730>
+-I<EC0FF8EC7FFF49B51280010714E0131F4914F090387FF80F9039FFC007F84813803803
+-FE005B485A4848EB03F0ED01E0484890C7FC5B123F5BA2127FEB000C903803FFE0010F13
+-F8D8FF3F13FE48B6FCB7128016C09039FE007FE001F8EB1FF001E0130F49EB07F8ED03FC
+-5B90C7120116FE1500A37EA46C7E15016D14FC121F6D1303000FEC07F86D130F6C6CEB1F
+-F06DEB3FE03A03FF81FFC06C90B512806C15006D5B011F13F8010713E001011380273A7C
+-B830>I<127CB712FC16FEA416FC48C7EA0FF816F0ED1FE0007CEC3FC0C8EA7F80EDFF00
+-A24A5A4A5A5D14075D140F5D4A5AA24A5AA24AC7FCA25C5C13015CA213035CA213075CA4
+-495AA6131F5CA96D5A6DC8FC273A7CB830>I<49B4FC011F13F0017F13FC90B57E0003EC
+-FF804815C048010113E03A1FF8003FF049131FD83FC0EB07F8A24848EB03FC90C71201A5
+-6D1303003F15F86D13076C6CEB0FF06C6CEB1FE0D807FCEB7FC03A03FF83FF806C90B512
+-006C6C13FC011F13F0497F90B512FE48802607FE0013C0D80FF8EB3FE0D81FE0EB0FF048
+-48EB07F8491303007F15FC90C712014815FE481400A66C14016C15FC6D1303003F15F86D
+-1307D81FF0EB1FF06D133F3A0FFF01FFE06C90B512C06C1580C6ECFE006D5B011F13F001
+-0190C7FC273A7CB830>I<49B4FC010F13E0013F13F890B57E4880488048010113803A0F
+-FC007FC0D81FF0EB3FE04848131F49EB0FF048481307A290C7EA03F85A4815FC1501A416
+-FEA37E7E6D130315076C7E6C6C130F6D133FD80FFC13FF6CB6FC7E6C14FE6C14F9013FEB
+-E1FC010F138190380060011400ED03F8A2150716F0150F000F15E0486C131F486CEB3FC0
+-157FEDFF804A1300EC07FE391FF01FFC90B55A6C5C6C5C6C1480C649C7FCEB3FF0273A7C
+-B830>I<120FEA3FC0EA7FE0A2EAFFF0A4EA7FE0A2EA3FC0EA0F00C7FCAF120FEA3FC0EA
+-7FE0A2EAFFF0A4EA7FE0A2EA3FC0EA0F000C276EA630>I<EA03C0EA0FF0EA1FF8A2EA3F
+-FCA4EA1FF8A2EA0FF0EA03C0C7FCAFEA03C0EA0FF0121F13F8123F13FCA3121FA2120F12
+-031200120113F8120313F01207EA1FE0123FEA7FC0EAFF80EA7F00127E12380E3470A630
+->I<16F01503ED07F8151F157FEDFFF014034A13C0021F138091383FFE00ECFFF8495B01
+-0713C0495BD93FFEC7FC495A3801FFF0485B000F13804890C8FCEA7FFC5BEAFFE05B7FEA
+-7FF87FEA1FFF6C7F000313E06C7F38007FFC6D7E90380FFF806D7F010113F06D7FEC3FFE
+-91381FFF80020713C06E13F01400ED7FF8151F1507ED03F01500252F7BB230>I<007FB7
+-FCA2B81280A36C16006C5DCBFCA7003FB612FE4881B81280A36C1600A229157DA530>I<
+-1278127EB4FC13C07FEA7FF813FEEA1FFF6C13C000037F6C13F86C6C7EEB1FFF6D7F0103
+-13E06D7F9038007FFC6E7E91380FFF806E13C0020113F080ED3FF8151F153FEDFFF05C02
+-0713C04A138091383FFE004A5A903801FFF0495B010F13804990C7FCEB7FFC48485A4813
+-E0000F5B4890C8FCEA7FFE13F8EAFFE05B90C9FC127E1278252F7BB230>I<EC1FE0ECFF
+-F8010313FE010F7F4914804914C090397FF03FE09038FF800F4890380007F0D803FC1303
+-3A07F801FBF89038F007FF380FE01F4A13FCEA1FC0495A003FEBFF0F903800FE07903901
+-FC03FE007FEBF801EA7E03ECF000A2EAFE0700FC49137EAA00FE6D13FED87E0314FCA2EC
+-F801D87F0114F8003FEBFC03903900FE07F0903880FF0F001F90387FFFE06D6C13C0EA0F
+-E06E13803A07F007FE009038F801F86C6CC7127C6CB414FE6CEB800390387FF01F6DB512
+-FC6D14F86D14E0010314C00100EBFE00EC1FF0273A7CB830>64 D<147F4A7EA2497FA449
+-7F14F7A401077F14E3A3010F7FA314C1A2011F7FA490383F80FEA590387F007FA4498049
+-133F90B6FCA34881A39038FC001F00038149130FA4000781491307A2D87FFFEB7FFFB56C
+-B51280A46C496C130029397DB830>I<007FB512F0B612FE6F7E82826C813A03F8001FF8
+-15076F7E1501A26F7EA615015EA24B5A1507ED1FF0ED7FE090B65A5E4BC7FC6F7E16E082
+-9039F8000FF8ED03FC6F7E1500167FA3EE3F80A6167F1700A25E4B5A1503ED1FFC007FB6
+-FCB75A5E16C05E6C02FCC7FC29387EB730>I<91387F803C903903FFF03E49EBFC7E011F
+-13FE49EBFFFE5B9038FFE07F48EB801F3903FE000F484813075B48481303A2484813015B
+-123F491300A2127F90C8FC167C16005A5AAC7E7EA2167C6D14FE123FA27F121F6D13016C
+-6C14FCA26C6CEB03F86D13076C6CEB0FF03901FF801F6C9038E07FE06DB512C06D14806D
+-1400010713FC6D13F09038007FC0273A7CB830>I<003FB512E04814FCB67E6F7E6C816C
+-813A03F8007FF0ED1FF8150F6F7E6F7E15016F7EA2EE7F80A2163F17C0161FA4EE0FE0AC
+-161F17C0A3163F1780A2167F17005E4B5A15034B5A150F4B5AED7FF0003FB65A485DB75A
+-93C7FC6C14FC6C14E02B387FB730>I<007FB7FCB81280A47ED803F8C7123FA8EE1F0093
+-C7FCA4157C15FEA490B5FCA6EBF800A4157C92C8FCA5EE07C0EE0FE0A9007FB7FCB8FCA4
+-6C16C02B387EB730>I<003FB712804816C0B8FCA27E7ED801FCC7121FA8EE0F8093C7FC
+-A5153E157FA490B6FCA69038FC007FA4153E92C8FCAE383FFFF8487FB5FCA27E6C5B2A38
+-7EB730>I<02FF13F00103EBC0F8010F13F1013F13FD4913FF90B6FC4813C1EC007F4848
+-133F4848131F49130F485A491307121F5B123F491303A2127F90C7FC6F5A92C8FC5A5AA8
+-92B5FC4A14805CA26C7F6C6D1400ED03F8A27F003F1407A27F121F6D130F120F7F6C6C13
+-1FA2D803FE133F6C6C137FECC1FF6C90B5FC7F6D13FB010F13F30103EBC1F0010090C8FC
+-293A7DB830>I<3B3FFF800FFFE0486D4813F0B56C4813F8A26C496C13F06C496C13E0D8
+-03F8C7EAFE00B290B6FCA601F8C7FCB3A23B3FFF800FFFE0486D4813F0B56C4813F8A26C
+-496C13F06C496C13E02D387FB730>I<007FB6FCB71280A46C1500260007F0C7FCB3B3A8
+-007FB6FCB71280A46C1500213879B730>I<49B512F04914F85BA27F6D14F090C7EAFE00
+-B3B3123C127EB4FCA24A5A1403EB8007397FF01FF86CB55A5D6C5C00075C000149C7FC38
+-003FF025397AB730>I<D83FFF90380FFF80486D4813C0B56C5AA26C497E6C496C1380D8
+-03F0903803F8004B5A4B5A151F4B5A5E4BC7FC15FE14014A5A5D4A5A4A5A141F5D4A5A4A
+-C8FC5C13F18101F37F13F790B57E14EFECC7F01483EC03F8140101FE7F496C7E5B157F49
+-7F82151F82150F826F7EA26F7E1501821500D83FFF903803FFC0486D4813E0B56C5AA26C
+-497E6C496C13C02B387FB730>I<383FFFF8487FB57EA26C5B6C5BD801FCC9FCB3B0EE0F
+-80EE1FC0A9003FB7FC5AB8FCA27E6C16802A387EB730>I<D83FF8ECFFE0486C4913F048
+-6C4913F8A2007F16F06C6C4913E00007160001EF14BFEC800FA39039E7C01F3FA4ECE03F
+-01E3133EA2ECF07EA201E1137CA2ECF8FCA201E013F8A214FDEC7DF0A3147FEC3FE0A3EC
+-1FC0A2EC070091C7FCADD83FFC903801FFE0486C4913F0B54913F8A26C486D13F06C486D
+-13E02D387FB730>I<D83FFC90381FFF80486C4913C0B54913E0A26C6D6C13C06C6E1380
+-0003913801F800EBF7C0A3EBF3E0A314F013F1A214F8A213F014FCA2147C147EA2143E14
+-3FA2141FA21581A2140F15C1A2140715E1A2140315F1A21401A215F91400A3157DA3153F
+-EA3FFF481380B5EAC01FA26CEB800F6C496C5A2B387EB730>I<90383FFFE048B512FC00
+-0714FF4815804815C04815E0EBF80001E0133FD87F80EB0FF0A290C71207A44815F84814
+-03B3A96C1407A26C15F0A36D130FA26D131F6C6CEB3FE001F813FF90B6FC6C15C06C1580
+-6C1500000114FCD8003F13E0253A7BB830>I<007FB512F0B612FE6F7E16E0826C813903
+-F8003FED0FFCED03FE15016F7EA2821780163FA6167F17005EA24B5A1503ED0FFCED3FF8
+-90B6FC5E5E16804BC7FC15F001F8C9FCB0387FFFC0B57EA46C5B29387EB730>I<90383F
+-FFE048B512FC000714FF4815804815C04815E0EBF80001E0133F4848EB1FF049130F90C7
+-1207A44815F8481403B3A8147E14FE6CEBFF076C15F0EC7F87A2EC3FC7018013CF9038C0
+-1FFFD83FE014E0EBF80F90B6FC6C15C06C15806C1500000114FCD8003F7FEB00016E7EA2
+-1680157F16C0153F16E0151F16F0150FED07E025467BB830>I<003FB57E4814F0B612FC
+-15FF6C816C812603F8017F9138003FF0151F6F7E15071503821501A515035E1507150F4B
+-5A153F4AB45A90B65A5E93C7FC5D8182D9F8007FED3FE0151F150F821507A817F8EEF1FC
+-A53A3FFF8003FB4801C0EBFFF8B56C7E17F06C496C13E06C49EB7FC0C9EA1F002E397FB7
+-30>I<90390FF803C0D97FFF13E048B512C74814F74814FF5A381FF80F383FE001497E48
+-48137F90C7123F5A48141FA2150FA37EED07C06C91C7FC7F7FEA3FF0EA1FFEEBFFF06C13
+-FF6C14E0000114F86C80011F13FF01031480D9003F13C014019138007FE0151FED0FF0A2
+-ED07F8A2007C140312FEA56C140716F07F6DEB0FE06D131F01F8EB3FC001FF13FF91B512
+-80160000FD5CD8FC7F13F8D8F81F5BD878011380253A7BB830>I<003FB712C04816E0B8
+-FCA43AFE003F800FA8007CED07C0C791C7FCB3B1011FB5FC4980A46D91C7FC2B387EB730
+->I<3B7FFFC007FFFCB56C4813FEA46C496C13FCD803F8C7EA3F80B3B16D147F00011600
+-A36C6C14FE6D13016D5CEC800390393FE00FF890391FF83FF06DB55A6D5C6D5C6D91C7FC
+-9038007FFCEC1FF02F3980B730>I<D87FFE90380FFFC0B54913E06E5AA24A7E6C486D13
+-C0D807F0903801FC00A26D130300035DA46C6C495AA46C6C495AA46D131F6D5CA3EC803F
+-013F5CA46D6C48C7FCA490380FE0FEA401075B14F1A301035BA314FB01015BA314FFA26D
+-5BA46E5A6E5A2B397EB730>I<D83FFC903801FFE0486C4913F000FF16F8A2007F16F06C
+-486D13E0D81FC09038001FC0000F1680A76D143F00071600A7000390380F803E9039F01F
+-C07EEC3FE0A3EC7FF0A2147D0001157CA29039F8FDF8FCA314F8A300005D01F913FCA2EC
+-F07CA201FD137DA2017D5CECE03DA3017F133FA2ECC01FA2013F5CA2EC800F6D486C5A2D
+-397FB730>I<3A3FFF01FFF84801837F02C77FA202835B6C01015B3A01FC007F806D91C7
+-FC00005C6D5BEB7F01EC81FCEB3F8314C3011F5B14E7010F5B14FF6D5BA26D5BA26D5BA2
+-6D90C8FCA4497FA2497FA2815B81EB0FE781EB1FC381EB3F8181EB7F0081497F49800001
+-143F49800003141F49800007140FD87FFEEB7FFFB590B5128080A25C6C486D130029387D
+-B730>I<D87FFF90381FFFC0B56C4813E0A46C496C13C0D803F8903803F8006D1307A26C
+-6C495AA26C6C5C151F6D5CEC803F013F5CECC07F011F91C7FCA290380FE0FEA214F10107
+-5BA2903803FBF8A201015B14FF6D5BA26E5AA36E5AB1903803FFF8497F497FA26D5B6D5B
+-2B387EB730>I<001FB612FC4815FE5AA490C7EA03FCED07F816F0150FED1FE016C0153F
+-ED7F80003E1500C85A4A5A5D14034A5A5D140F4A5A5D143F4A5A92C7FC5C495A5C130349
+-5A5C130F495A5C133F495A91C8FC5B4848147C4914FE1203485A5B120F485A5B123F485A
+-90B6FCB7FCA46C15FC27387CB730>I<007FB5FCB61280A4150048C8FCB3B3B3A5B6FC15
+-80A46C140019476DBE30>I<007FB5FCB61280A47EC7123FB3B3B3A5007FB5FCB6FCA46C
+-140019477DBE30>93 D<1307EB1FC0EB7FF0497E000313FE000FEBFF80003F14E0D87FFD
+-13F039FFF07FF8EBC01FEB800F38FE0003007CEB01F00010EB00401D0E77B730>I<007F
+-B612F0A2B712F8A36C15F0A225077B7D30>I<EB3FFC48B57E4814E04880488048809038
+-F00FFE9038E001FF806F7E6C48133F6C4880C8121FA491B5FC130F137F48B6FC12075A48
+-EBC01F383FFC00EA7FE0138048C7FC5AA46C143FA26C6C137F9038C001FF263FF80FEBFF
+-C06CB712E0A2000714F76C14C3C6020013C0D93FF090C7FC2B2A7CA830>97
+-D<EA3FFC487E12FFA2127F123F1200AAEC03FE91381FFF80027F13E091B57E90B612FC82
+-ECFE079138F001FF4A6C13804A137F4AEB3FC091C7121F4915E0160FA217F01607A8160F
+-A217E07F161F6EEB3FC0A26EEB7F806E13FFDAF00313009138FC0FFE91B55A5E495CD97E
+-7F13C0D93C1F90C7FC90380003FC2C3980B730>I<ECFFE0010713FC011F7F017F7F90B6
+-12804815C048EB807F3907FC003F485A485A49EB1F804848EB0F004990C7FC127F90C9FC
+-A25A5AA87E7EA27F003FEC07C06DEB0FE06C7E6D131F6C6C14C0D807FE133F9039FFC0FF
+-806C90B5FCC615006D5B011F13F801075B01011380232A7AA830>I<913801FFE04A7F5C
+-A28080EC0007AAEB03FE90381FFF874913E790B6FC5A5A481303380FFC00D81FF0133F49
+-131F485A150F4848130790C7FCA25AA25AA87E6C140FA27F003F141F6D133F6C7E6D137F
+-390FF801FF2607FE07EBFFC06CB712E06C16F06C14F76D01C713E0011F010313C0D907FC
+-C8FC2C397DB730>I<49B4FC010713E0011F13F8017F7F90B57E488048018113803A07FC
+-007FC04848133FD81FE0EB1FE0150F484814F0491307127F90C7FCED03F85A5AB7FCA516
+-F048C9FC7E7EA27F003FEC01F06DEB03F86C7E6C7E6D1307D807FEEB1FF03A03FFC07FE0
+-6C90B5FC6C15C0013F14806DEBFE00010713F8010013C0252A7CA830>I<EDFF80020713
+-E0021F13F05C4A13F891B5FC491387903803FE079138FC03F0903907F800C04A1300A800
+-3FB612C04815E0B7FCA36C15C0260007F0C7FCB3A9003FB512FE4880B71280A26C15006C
+-5C25397DB830>I<D903FC13FF90261FFF8713C04913DF90B712E05A5A2607FE07138F90
+-3AF801FE07C048486C6CC7FCA2497F001F8149133FA56D137F000F92C7FC6D5BA26C6C48
+-5AEBFE0790B55A5D485C15C001DF5BD9C3FCC8FC01C0C9FCA37F7F6CB512F015FF6C15C0
+-4815F0488148813A3FE0001FFE0180130148C8127F007E8100FE168048151FA56C153F00
+-7FED7F006D5C6C6C495A01F013076CB4EB7FFC6C90B55A6C5D000115C06C6C91C7FC011F
+-13FC010113C02B3E7DA730>I<EA3FFC487E12FFA2127F123F1200AAEC01FE91380FFF80
+-023F13E091B57E90B67EA29138FE07FCECF8039138E001FE14C0EC8000A291C7FCA25BB3
+-A23B3FFFF81FFFF8486D4813FCB500FE14FEA26C01FC14FC6C496C13F82F3880B730>I<
+-14E0EB03F8A2497EA36D5AA2EB00E091C8FCA9381FFFF8487F5AA27E7EEA0001B3A9003F
+-B612C04815E0B7FCA27E6C15C023397AB830>I<EC01C0EC07F0A2EC0FF8A3EC07F0A2EC
+-01C091C7FCA990B512F04814F8A47EEB0003B3B3A5EC07F0A2123C007EEB0FE0B4131FEC
+-3FC0147F90B512806C14005C6C5B000F13F0000313C01D4E7CB830>I<EA7FF8487EA412
+-7F1200AB0203B512804A14C017E0A217C06E14809139001FE0004B5A4B5A4BC7FC4A5A4A
+-5AEC0FF84A5A4A5A4A5A4A5A01FD7F90B57E8114F7ECE3F8ECC1FCEC81FEEC00FF497F49
+-6D7E6F7E826F7E15076F7E6F7E3B7FFFF81FFFE0B56C4813F017F8A217F06C496C13E02D
+-387FB730>I<387FFFF8B57EA47EEA0001B3B3A8007FB612F0B712F8A46C15F025387BB7
+-30>I<02FC137E3B7FC3FF01FF80D8FFEF01877F90B500CF7F15DF92B57E6C010F138726
+-07FE07EB03F801FC13FE9039F803FC01A201F013F8A301E013F0B3A23C7FFE0FFF07FF80
+-B548018F13C0A46C486C01071380322881A730>I<EC01FE3A3FFC0FFF80267FFE3F13E0
+-00FF90B57E90B67E7E6C9038FE07FCC6EBF8039138E001FE14C0EC8000A291C7FCA25BB3
+-A23B3FFFF81FFFF8486D4813FCB500FE14FEA26C01FC14FC6C496C13F82F2880A730>I<
+-49B4FC010F13E0013F13F8497F90B57E0003ECFF8014013A07FC007FC04848EB3FE0D81F
+-E0EB0FF0A24848EB07F8491303007F15FC90C71201A300FEEC00FEA86C14016C15FCA26D
+-1303003F15F86D13076D130F6C6CEB1FF06C6CEB3FE06D137F3A07FF01FFC06C90B51280
+-6C15006C6C13FC6D5B010F13E0010190C7FC272A7CA830>I<EC03FE3A3FFC1FFF80267F
+-FE7F13E000FF90B57E90B612FC6C816CEBFE07C69038F001FF4A6C13804A137F4AEB3FC0
+-91C7121F4915E0160FA217F01607A8160FA217E07F161F6EEB3FC0A26EEB7F806E13FFDA
+-F00313009138FC0FFE91B55A5E495C6E13C0021F90C7FCEC03FC91C9FCAD383FFFF8487F
+-B57EA26C5B6C5B2C3C80A730>I<49B413F8010FEBC1FC013F13F14913FD48B6FC5A4813
+-81390FFC007F49131F4848130F491307485A491303127F90C7FC15015A5AA77E7E15037F
+-A26C6C1307150F6C6C131F6C6C133F01FC137F3907FF01FF6C90B5FC6C14FD6C14F9013F
+-13F1010F13C1903803FE0190C7FCAD92B512F84A14FCA46E14F82E3C7DA730>I<ED07F8
+-3A3FFF803FFF486DB51280B512C302CF14C06C13DF6C9038FFFC3FD8001F13E09238801F
+-809238000F004A90C7FC5C5C5CA25CA45CAF003FB512FC4880B7FCA26C5C6C5C2A287EA7
+-30>I<90381FFC1E48B5129F000714FF5A5A5A387FF007EB800100FEC7FC4880A46C143E
+-007F91C7FC13E06CB4FC6C13FC6CEBFF806C14E0000114F86C6C7F01037F9038000FFF02
+-001380007C147F00FEEC1FC0A2150F7EA27F151F6DEB3F806D137F9039FC03FF0090B6FC
+-5D5D00FC14F0D8F83F13C026780FFEC7FC222A79A830>I<EB0780497E131FA9003FB612
+-E04815F0B7FCA36C15E026001FC0C7FCB216F8ED01FCA5ECE003010FEB07F814F09138FC
+-1FF06DB512E06D14C016806D14009038007FFCEC1FF026337EB130>I<D83FFCEB3FFC48
+-6C497E00FF14FFA2007F147F003F143F00001400B3A41501A2150315076D130F903A7FC0
+-7FFFF891B612FC6D15FE7F6D4913FC6D9038F87FF8010001C0C7FC2F2880A630>I<3B3F
+-FFC07FFF80486DB512C0B515E0A26C16C06C496C13803B01F80003F000A26D130700005D
+-A26D130F017E5CA2017F131F6D5CA2EC803F011F91C7FCA26E5A010F137EA2ECE0FE0107
+-5BA214F101035BA3903801FBF0A314FF6D5BA36E5A6E5A2B277EA630>I<3B3FFFC01FFF
+-E0486D4813F0B515F8A26C16F06C496C13E0D807E0C7EA3F00A26D5C0003157EA56D14FE
+-00015DEC0F80EC1FC0EC3FE0A33A00FC7FF1F8A2147DA2ECFDF9017C5C14F8A3017E13FB
+-A290393FF07FE0A3ECE03FA2011F5C90390F800F802D277FA630>I<3A3FFF81FFFC4801
+-C37FB580A26C5D6C01815BC648C66CC7FC137FEC80FE90383F81FC90381FC3F8EB0FE3EC
+-E7F06DB45A6D5B7F6D5B92C8FC147E147F5C497F81903803F7E0EB07E790380FE3F0ECC1
+-F890381F81FC90383F80FE90387F007E017E137F01FE6D7E48486D7E267FFF80B5FCB500
+-C1148014E3A214C16C0180140029277DA630>I<3B3FFFC07FFF80486DB512C0B515E0A2
+-6C16C06C496C13803B01FC0003F000A2000014076D5C137E150F017F5C7F151FD91F805B
+-A214C0010F49C7FCA214E00107137EA2EB03F0157C15FCEB01F85DA2EB00F9ECFDF0147D
+-147FA26E5AA36E5AA35DA2143F92C8FCA25C147EA2000F13FE486C5AEA3FC1EBC3F81387
+-EB8FF0EBFFE06C5B5C6C90C9FC6C5AEA01F02B3C7EA630>I<001FB612FC4815FE5AA316
+-FC90C7EA0FF8ED1FF0ED3FE0ED7FC0EDFF80003E491300C7485A4A5A4A5A4A5A4A5A4A5A
+-4A5A4990C7FC495A495A495A495A495A495A4948133E4890C7127F485A485A485A485A48
+-5A48B7FCB8FCA46C15FE28277DA630>I<ED3FF0913803FFF8140F5C147F16F09138FFF0
+-0092C7FC495A5CB3A21303495A133F383FFFF0B55A5C91C8FC14C080003F7F38003FF813
+-076D7E1301B3A2806D7E15F091387FFFF016F8141F8014039138003FF025477BBE30>I<
+-127CA212FEB3B3B3AD127CA207476CBE30>I<EA7FE0EAFFFE6D7E8014F07EC66C7E1307
+-6D7E1301B3A2806D7E15E091387FFFE06E13F8801407141F5C4A13E09138FFE00092C7FC
+-495A5CB3A21303495A137F387FFFF0B5FC14C05C49C8FCEA7FE025477BBE30>I<017C13
+-3848B4137C48EB80FE4813C14813C348EBEFFC397FEFFFF0D8FF8713E0010713C0486C13
+-80D87C0113003838007C1F0C78B730>I E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fk cmbx12 14.4 49
+-/Fk 49 122 df<EEFFFC031FEBFF804AB612E0020781021F9038C00FF8913A7FFE0003FC
+-DAFFF0EB00FE4949EB03FF4901805B4990C7487F49485CA2495A4D7F013F6F5B5CA37190
+-C7FC715AEF01F894C9FCA90403B512C0BAFCA526003FFCC7120783B3B3A6003FB5D8FC03
+-B612C0A542547DD34B>12 D<B712F0AB240B7F9F2D>45 D<EA07F0487E487E487E487EB5
+-1280A76C13006C5A6C5A6C5A6C5A1111769025>I<157815FC14031407141F14FF130F00
+-07B5FCB6FCA2147F13F0EAF800C7FCB3B3B3A6007FB712FEA52F4E76CD43>49
+-D<EC3FFE0103B512E0010F14FC013F14FF90B712C048D9C07F7F2703FE000F13F8D807F8
+-01037FD80FE06D7F48486D7F48488001F01680486C6E13C07F486C6E13E07FA27013F0A5
+-6C5AA26C5AEA0FF0EA03C0C914E05EA218C05E1880A24C13005F4C5A4B5B5F4B5B5F4B5B
+-4B90C7FC4B5A5E4B5AED7FE04B5A4A5B4A48C8FC4A5A5D4A48EB01F04A5AEC3F804AC7FC
+-02FEEC03E0495A495A495A495AD91F80140749C8FC013E150F017FB7FC90B812C05A5A5A
+-5A5A5A5AB9FC1880A4344E79CD43>I<91380FFFC091B512FC0107ECFF80011F15E09026
+-3FF8077F9026FF800113FC4848C76C7ED803F86E7E491680D807FC8048B416C080486D15
+-E0A4805CA36C17C06C5B6C90C75AD801FC1680C9FC4C13005FA24C5A4B5B4B5B4B13C04B
+-5BDBFFFEC7FC91B512F816E016FCEEFF80DA000713E0030113F89238007FFE707E701380
+-7013C018E07013F0A218F8A27013FCA218FEA2EA03E0EA0FF8487E487E487EB57EA318FC
+-A25E18F891C7FC6C17F0495C6C4816E001F04A13C06C484A1380D80FF84A13006CB44A5A
+-6CD9F0075BC690B612F06D5D011F1580010302FCC7FCD9001F1380374F7ACD43>I<177C
+-17FEA2160116031607160FA2161F163F167FA216FF5D5DA25D5DED1FBFED3F3F153E157C
+-15FCEC01F815F0EC03E01407EC0FC01580EC1F005C147E147C5C1301495A495A5C495A13
+-1F49C7FC133E5B13FC485A5B485A1207485A485A90C8FC123E127E5ABA12C0A5C96C48C7
+-FCAF020FB712C0A53A4F7CCE43>I<D80380150ED807E0157E01FEEC03FED9FFF0137F91
+-B65A5F5F5F5F5F94C7FC5E5E16F016C093C8FC15F801E190C9FC01E0CAFCABEC0FFF027F
+-13F001E3B512FE01E76E7E9026FFF8077FDAC0017F49C713F8496E7E49143F4981496E7E
+-6C481680C9FC18C08218E0A418F0A3EA0FE0487E487E487E487EA418E0A35B6C484A13C0
+-5B491680003EC85A003F17006C6C4A5A6D5D6C6C4A5AD807F8495BD803FE01075B2701FF
+-C03F5B6C90B65A013F4AC7FC6D14F8010314C09026007FF8C8FC344F79CD43>I<ED0FFF
+-92B512E0020780021F14FC91397FFE03FE903A01FFF0007F4901C0EB3F804990C7121F49
+-48EC7FC0494814FF49484913E049485B01FF5C485BA2485B5AA2486F13C04A6D1380486F
+-1300177E94C7FC5AA291CAFC5AA21508913801FFF8020713FFB54814C04A14F04AC66C7E
+-023C6D7E4A6D7E4A6D7E7013804A15C0A24A15E07013F05C18F8A491C714FCA37EA67EA4
+-6C17F880A27E18F06C5D18E06C6D15C07E6E4913806C6D15006D6C495A6D6CEB7FFC6DB4
+-48485A6D90B55A010315C0010092C7FC023F13FC020713C0364F7ACD43>I<121F7F7FEB
+-FF8091B81280A45A1900606060A2606060485F0180C86CC7FC007EC95A4C5A007C4B5A5F
+-4C5A160F4C5A484B5A4C5A94C8FC16FEC812014B5A5E4B5A150F4B5AA24B5AA24B5A15FF
+-A24A90C9FCA25C5D1407A2140FA25D141FA2143FA4147F5DA314FFA55BAC6D5BA2EC3FC0
+-6E5A395279D043>I<913807FFC0027F13FC0103B67E010F15E090261FFC0113F8903A3F
+-E0003FFCD97F80EB0FFE49C76C7E48488048486E1380000717C04980120F18E0177FA212
+-1F7FA27F7F6E14FF02E015C014F802FE4913806C7FDBC00313009238F007FE6C02F85B92
+-38FE1FF86C9138FFBFF06CEDFFE017806C4BC7FC6D806D81010F15E06D81010115FC0107
+-81011F81491680EBFFE748018115C048D9007F14E04848011F14F048487F484813030300
+-14F8484880161F4848020713FC1601824848157F173FA2171FA2170FA218F8A27F007F17
+-F06D151FA26C6CED3FE0001F17C06D157F6C6CEDFF806C6C6C010313006C01E0EB0FFE6C
+-01FCEBFFFC6C6CB612F06D5D010F1580010102FCC7FCD9000F13C0364F7ACD43>I<9138
+-0FFF8091B512F8010314FE010F6E7E4901037F90267FF8007F4948EB3FF048496D7E4849
+-80486F7E484980824817805A91C714C05A7013E0A218F0B5FCA318F8A618FCA46C5DA37E
+-A25E6C7F6C5DA26C5D6C7F6C6D137B6C6D13F390387FF803011FB512E36D14C301030283
+-13F89039007FFE03EC00401500A218F05EA3D801F816E0487E486C16C0487E486D491380
+-A218005E5F4C5A91C7FC6C484A5A494A5A49495B6C48495BD803FC010F5B9027FF807FFE
+-C7FC6C90B55A6C6C14F06D14C0010F49C8FC010013F0364F7ACD43>I<91B5FC010F14F8
+-017F14FF90B712C00003D9C00F7F2707FC00017FD80FE06D7F48486E7E48C87FD87FE06E
+-7E7F7F486C1680A66C5A18006C485C6C5AC9485A5F4B5B4B5B4B5B4B5B4B90C7FC16FC4B
+-5A4B5A16C04B5A93C8FC4A5A5D14035D5D14075DA25D140FA25DAB91CAFCAAEC1FC04A7E
+-ECFFF8497FA2497FA76D5BA26D5BEC3FE06E5A315479D340>63 D<BA7E19FCF1FF801AF0
+-1AFCD8000701F0C7000F13FF060014C0071F7F070713F807017F737F747E747F747F8674
+-7F747F8886888688A2757EA31D8087A21DC0A51DE0A387A963A31DC0A51D80A2631D00A3
+-515AA2646264505B6264505B505B5090C7FCF2FFFE4F5B07075B071F5B96B512C0060F91
+-C8FCBB5A1AF01AC007FCC9FC19805B527CD167>68 D<BC1280A5D8000701F8C7000114C0
+-F0001F19071901851A7F1A3F1A1FA2F20FE0A21A07A31A03A318F81BF01A01A497C7FC17
+-01A317031707170F177F92B6FCA59238F8007F170F170717031701A317001B3EA31B7CA3
+-95C8FCA21BFCA21BF8A21A01A31A031BF01A071A0FA21A1F1A3FF27FE0F101FF1907191F
+-0603B5FCBCFCA21BC0A34F517CD058>I<BB12FEA5D8000701F8C700077FF0007F191F19
+-0785858586861B80A21A1FA31A0FA41BC006F81307A497C7FCA31701A317031707170F17
+-7F92B6FCA59238F8007F170F170717031701A31700A795C9FCB3B812F8A54A517CD055>
+-I<B8D8C003B8FCA5D8000701F8C9001FEBE000B3AE92BAFCA503F8C9121FB3B1B8D8C003
+-B8FCA560527CD169>72 D<B812C0A5D8000701F8C7FCB3B3B3B2B812C0A52A527CD132>
+-I<027FB71280A591C76C90C7FCB3B3B3EA07F0EA1FFC487E487EA2B57EA44C5AA34A485B
+-7E49495BD83FF8495BD81FE05DD80FFC011F5B2707FF807F90C8FC000190B512FC6C6C14
+-F0011F14C0010101F8C9FC39537DD145>I<B812F8A5D8000701F8CAFCB3B3A91A7CA41A
+-FC1AF8A51901A31903A219071AF0190FA2191F193F197F19FF180360183F4DB5FCBB12E0
+-A546527CD151>76 D<B600FC073FB512FE6F61A26F96B6FCA2D80007F5C00070EF01EFA2
+-02EF6DEF03CFA202E76DEF078FA202E36DEF0F0FA202E16D171EA302E06D173CA26F6C17
+-78A26F6C17F0A26F6DED01E0A26F6DED03C0A36F6DED0780A26F6DED0F00A26F6D151EA2
+-6F6D5DA3706C5DA2706C5DA2706D495AA2706D495AA2706D495AA3706D49C7FCA2706D13
+-1EA2706D5BA2716C5BA3716C5BA271EB81E0A271EBC3C0A271EBE780A27101FFC8FCA371
+-5BA2715BA2725AA2725AA2D93FFC6F5AB74DB712FEA2725AA2725A77527CD180>I<BAFC
+-19F819FF1AE086D8000701F0C7001F13FC060113FF726C13807313C0070F13E01BF08573
+-13F81BFCA27313FEA41BFFA81BFEA31BFC61A21BF84F13F04F13E0614F13C04F13004E48
+-5A061F5B92B812F01AC04FC7FC19E003F8CBFCB3AEB812C0A550527CD15C>80
+-D<B912F0F0FF8019F819FF1AC0D8000701F0C714F0060F7F060113FE727F737F737F8573
+-7F87A2737FA387A863A2616363A24F5B4F5B4F90C8FC4F5A06035B060F13F095B512C092
+-B8C9FC19F819E019F89226F0000313FE9439007FFF80727F727F727F727F727F8684A286
+-84A787A71D1C75133EA38575137E73157C7513FC731401B86C6D9038F803F807039038FE
+-07F07390B512E0736C14C0080F1400CEEA7FFC5F537CD164>82 D<91260FFF80130791B5
+-00F85B010702FF5B011FEDC03F49EDF07F9026FFFC006D5A4801E0EB0FFD4801800101B5
+-FC4848C87E48488149150F001F824981123F4981007F82A28412FF84A27FA26D82A27F7F
+-6D93C7FC14C06C13F014FF15F86CECFF8016FC6CEDFFC017F06C16FC6C16FF6C17C06C83
+-6C836D826D82010F821303010082021F16801400030F15C0ED007F040714E01600173F05
+-0F13F08383A200788200F882A3187FA27EA219E07EA26CEFFFC0A27F6D4B13806D17006D
+-5D01FC4B5A01FF4B5A02C04A5A02F8EC7FF0903B1FFFC003FFE0486C90B65AD8FC0393C7
+-FC48C66C14FC48010F14F048D9007F90C8FC3C5479D24B>I<003FBC1280A59126C0003F
+-9038C0007F49C71607D87FF8060113C001E08449197F49193F90C8171FA2007E1A0FA300
+-7C1A07A500FC1BE0481A03A6C994C7FCB3B3AC91B912F0A553517BD05E>I<B800C00103
+-B612FCA5D8000701F8CAEBF000F31F80B3B3B11B3FA26D97C7FC81637F1B7E6D6D17FE50
+-5A6E7E505A6E6D15076E4D5A6E6D4B5A6E6D4B5A6E01F84B5A6E6DDA03FFC8FC6E6CB46C
+-EB0FFE6F9039F001FFF8030F90B65A030316C0DB007F92C9FC040F14F8DC007F13805E53
+-7CD167>I<B700FC017FB600FE91B612F0A5D8003F01C0C8001F01E0C9EBF8006F71EE0F
+-C06D7161876F1C1F6D7196C7FC6F8373606D1E3E6F836D7160876F1CFC6D666F4B801F01
+-6D66704A806E525A88704A17076E059F5F70021F80080F160F6E6570023F806EDC3E074C
+-C8FC8870027E5F6EDC7C03163E7002FC804F6C167E6E1C7C700101814F6C16FC6E745B70
+-010317016E4C6D5D060716C00580496D14036F63DDC00F16E04F6D14076F07F05BDDE01F
+-170F6F92C76C5D1DF8DDF03E6E141F6F98C9FCDDF87E16FC067C6E5C6FF1FE3EDDFCFC17
+-7E6F4A6E147C1DFFDDFFF06E14FC6F62A24E816F62A270496F5BA24E817061A295C97E70
+-61A270487090CAFCA37048705AA24D1601040360A27048705A84537DD18B>87
+-D<EC7FFF0107B512F0013F14FE90B77E48D9E00F7F2703FE000113F0486C6D7F6EEB3FFC
+-48826E131F83707FA36C496D7FA26C90C7FC6C5AC9FCA6037FB5FC020FB6FC91B7FC0107
+-1487013FEBF0074913803901FFFC004813F0485B485B485B4890C7FC5A5BA2485AA45EA2
+-6D5C007F151D163D6C6C02797F6C6D01F113F86C9026C003E1EBFFE06C9026F81FC014F0
+-6C90B5487EC6ED001F011F01FC010713E0010101E090C8FC3C387CB641>97
+-D<EB3FF0B5FCA51203C6FCB3A4923801FFE0030F13FE033FEBFFC092B612F002F301017F
+-913AF7F8003FFEDAFFE0EB0FFF03806D7F92C76C7F4A6E7F4A824A6E7FA2727EA285A285
+-84A31A80AC1A00A44E5AA36118FF616E4A5BA26E4A5B6E4A5B6F495BDACFC04990C7FCDA
+-87F0EB7FFC913A03FE03FFF849C6B612E0496D148049011F01FCC8FC90C7000313C04154
+-7BD24B>I<913801FFF8021FEBFF8091B612F0010315FC010F9038C00FFE903A1FFE0001
+-FFD97FFC491380D9FFF05B4817C048495B5C5A485BA2486F138091C7FC486F1300705A48
+-92C8FC5BA312FFAD127F7FA27EA2EF03E06C7F17076C6D15C07E6E140F6CEE1F806C6DEC
+-3F006C6D147ED97FFE5C6D6CEB03F8010F9038E01FF0010390B55A01001580023F49C7FC
+-020113E033387CB63C>I<4DB47E0407B5FCA5EE001F1707B3A4913801FFE0021F13FC91
+-B6FC010315C7010F9038E03FE74990380007F7D97FFC0101B5FC49487F4849143F484980
+-485B83485B5A91C8FC5AA3485AA412FFAC127FA36C7EA37EA26C7F5F6C6D5C7E6C6D5C6C
+-6D49B5FC6D6C4914E0D93FFED90FEFEBFF80903A0FFFC07FCF6D90B5128F0101ECFE0FD9
+-003F13F8020301C049C7FC41547CD24B>I<913803FFC0023F13FC49B6FC010715C04901
+-817F903A3FFC007FF849486D7E49486D7E4849130F48496D7E48178048497F18C0488191
+-C7FC4817E0A248815B18F0A212FFA490B8FCA318E049CAFCA6127FA27F7EA218E06CEE01
+-F06E14037E6C6DEC07E0A26C6DEC0FC06C6D141F6C6DEC3F806D6CECFF00D91FFEEB03FE
+-903A0FFFC03FF8010390B55A010015C0021F49C7FC020113F034387CB63D>I<ED3FFC02
+-03B5FC020F14C0023F14E09139FFF81FF0499038C03FF849EB807F49903800FFFC495A49
+-5AA2495AA2EE7FF8495AEE3FF0EE0FC093C7FCAEB712E0A526007FF8C8FCB3B3A7007FB5
+-12FEA52E547CD329>I<DA3FFF14FF0103B5D8F00713C0010FDAFC1F13E0013FECFF7F90
+-267FFC0F9038FF9FF09026FFE001EBF83F48496C13E0484990387FF01F4890C7D83FF813
+-E0489338FC0FC0F0078048486E6CC7FCA2003F82A9001F5EA26C6C4A5AA26C5E6C6D495A
+-6C6D495A6C6D485BDAFC0F5B4890B6C8FCD803EF14FC01C314F02607C03F90C9FC91CBFC
+-A2120FA37FA213F813FE90B7FC6C16F817FF18C06C836C836C836D828448B9FC12074848
+-C700031480D81FF8EC003F4848150748486F13C083485A83A56D5D007F18806D5D003F18
+-006C6C4B5AD80FFEED1FFC6C6C6CEC7FF86C01E049485A6C01FE011F5B6C6CB71280010F
+-03FCC7FC010115E0D9000F01FCC8FC3C4F7CB543>I<EB3FF0B5FCA51203C6FCB3A4EE1F
+-FC93B512C0030314F0030F8092391FE07FFC92393F001FFE037C8003F07FDAF1E081ECF3
+-C0DAF7807F8502FFC7FC5CA25CA45CB3ACB6D8F807B612C0A542537BD24B>I<137F497E
+-000313E0487FA2487FA76C5BA26C5BC613806DC7FC90C8FCADEB3FF0B5FCA512017EB3B3
+-A6B612E0A51B547BD325>I<EB3FF0B5FCA51203C6FCB3A54CB512F8A59339003FFE00EF
+-1FF0EF3FC04D5A4DC7FCEE03FEEE07F84C5A4C5AEE7FC04CC8FC4B5A4B5AED0FF8ED1FE0
+-4B7E4B7EECF1FF02F37F02F77F91B6FC83159F030F7F02FE80DAF8077F4A7E6F7F6F7F83
+-707E82707F84707F707F82707F84707F177F717E4D13C0B6D8F003B6FCA540537CD247>
+-107 D<EB3FF0B5FCA512017EB3B3B3B1B612F0A51C537BD225>I<D93FF0D91FFCEDFFE0
+-B591B500C0010713FE030302F0011F6D7E030F6E017F8092271FE07FFCD9FF037F922A3F
+-001FFE01F8007F0003027C9126FF03E080C602F06DD90780137FDAF1E0038FC77FDAF3C0
+-159EDAF7806D01BC143F07FC8102FFC75C4A5EA24A5EA44A5EB3ACB6D8F807B6D8C03FB5
+-12FEA567367BB570>I<D93FF0EB1FFCB591B512C0030314F0030F8092391FE07FFC9239
+-3F001FFE0003027C80C602F07FDAF1E081ECF3C0DAF7807F8502FFC7FC5CA25CA45CB3AC
+-B6D8F807B612C0A542367BB54B>I<913801FFE0021F13FE91B612C0010315F0010F9038
+-807FFC903A1FFC000FFED97FF86D6C7E49486D7F48496D7F48496D7F4A147F48834890C8
+-6C7EA24883A248486F7EA3007F1880A400FF18C0AC007F1880A3003F18006D5DA26C5FA2
+-6C5F6E147F6C5F6C6D4A5A6C6D495B6C6D495B6D6C495BD93FFE011F90C7FC903A0FFF80
+-7FFC6D90B55A010015C0023F91C8FC020113E03A387CB643>I<903A3FF001FFE0B5010F
+-13FE033FEBFFC092B612F002F301017F913AF7F8007FFE0003D9FFE0EB1FFFC602806D7F
+-92C76C7F4A824A6E7F4A6E7FA2717FA285187F85A4721380AC1A0060A36118FFA2615F61
+-6E4A5BA26E4A5B6E4A5B6F495B6F4990C7FC03F0EBFFFC9126FBFE075B02F8B612E06F14
+-80031F01FCC8FC030313C092CBFCB1B612F8A5414D7BB54B>I<90397FE003FEB590380F
+-FF80033F13E04B13F09238FE1FF89139E1F83FFC0003D9E3E013FEC6ECC07FECE78014EF
+-150014EE02FEEB3FFC5CEE1FF8EE0FF04A90C7FCA55CB3AAB612FCA52F367CB537>114
+-D<903903FFF00F013FEBFE1F90B7FC120348EB003FD80FF81307D81FE0130148487F4980
+-127F90C87EA24881A27FA27F01F091C7FC13FCEBFFC06C13FF15F86C14FF16C06C15F06C
+-816C816C81C681013F1580010F15C01300020714E0EC003F030713F015010078EC007F00
+-F8153F161F7E160FA27E17E07E6D141F17C07F6DEC3F8001F8EC7F0001FEEB01FE9039FF
+-C00FFC6DB55AD8FC1F14E0D8F807148048C601F8C7FC2C387CB635>I<143EA6147EA414
+-FEA21301A313031307A2130F131F133F13FF5A000F90B6FCB8FCA426003FFEC8FCB3A9EE
+-07C0AB011FEC0F8080A26DEC1F0015806DEBC03E6DEBF0FC6DEBFFF86D6C5B021F5B0203
+-13802A4D7ECB34>I<D93FF8913801FFC0B50207B5FCA50003ED001FC61607B3AE5FA35F
+-A2017F5D173B177B6D6C14F3DC01E313F06D6CD907C3EBFFC0903A0FFFC03F836D90B512
+-03010114FE6D6C13F8020701E091C7FC42377BB54B>I<B600F00107B5FCA5000101F8C8
+-EA7FE06C6DED3F00A2017F163E6E157E013F167C6E15FC6D5E6F13016D5E8117036D5E6F
+-13076D5E6F130F6D5E6F131F6D93C7FC815F6E6C133E177E023F147C6F13FC6E5C16816E
+-5C16C3A26EEBE3E016E76E5C16FF6E5CA26E91C8FCA26F5AA36F5AA26F5AA26F5AA26F5A
+-6F5A40367DB447>I<B6D8E07FB5D8C003B512C0A5000101F0C701F0C7381FF8006E027F
+-ED07E06C715DA26E023F150F017F705DA26E181F013F4B6C92C7FC6E606D70143E94B5FC
+-6F177E6D4A6E137C03C001F315FC6D715B160303E001E114016D020702E05B03F013C06D
+-71485A160F03F8D9807F13076D05F85B93381F003F03FC160F027F4902FC5BDBFE3E011F
+-131F023F04FE90C8FC167EDBFF7C010F5B6E01FCECFF3E4C6D137E6E5FA24C7F6E5F4C7F
+-6E5FA24C7F6E5F4C147FA26E5F93C8123F6F5EA2033E6FC9FC5A367DB461>I<B600F001
+-07B5FCA5C601F8C8EA7FE06EED3F00A26D6C153E187E013F167C6E15FC6D5E6F13016D5E
+-6F13036D5E8117076D6D5C170F6D6D5C171F6D93C7FC6F5B027F143E6F137E023F147C6F
+-13FCA26E6D5A16816EEBC1F016C36E5C16E76E5C16FF6E5CA26E91C8FCA36F5AA26F5AA2
+-6F5AA26F5AA26F5AA35E150F5E151F93C9FC5DD81FC0133E486C137E486C137C486C13FC
+-5D14015D14034A5A6C48485A49485A263FC07FCAFCEB81FE6CB45A6C13F000035BC690CB
+-FC404D7DB447>121 D E
+-%EndDVIPSBitmapFont
+-%DVIPSBitmapFont: Fl cmr10 10.95 86
+-/Fl 86 124 df<4AB4EB0FE0021F9038E03FFC913A7F00F8FC1ED901FC90383FF03FD907
+-F090397FE07F80494801FF13FF4948485BD93F805C137F0200ED7F00EF003E01FE6D91C7
+-FC82ADB97EA3C648C76CC8FCB3AE486C4A7E007FD9FC3FEBFF80A339407FBF35>11
+-D<4AB4FC021F13C091387F01F0903901FC0078D907F0131C4948133E494813FF49485A13
+-7F1400A213FE6F5A163893C7FCAA167FB8FCA33900FE00018182B3AC486CECFF80007FD9
+-FC3F13FEA32F407FBF33>I<4AB47E021F13F791387F00FFEB01F8903807F001EB0FE0EB
+-1FC0EB3F80137F14008101FE80AEB8FCA3C648C77EB3AE486CECFF80007FD9FC3F13FEA3
+-2F407FBF33>I<4AB4ECFF80021FD9C00F13E0913B7F01F03F80F8903C01F80078FE003C
+-D907F0D93FF8130E49484948131F49484948EB7F804948484913FF137F02005CA201FE92
+-C7FC6FED7F0070141C96C7FCAAF13F80BBFCA3C648C76CC7FC197F193FB3AC486C4A6CEB
+-7FC0007FD9FC3FD9FE1FB5FCA348407FBF4C>I<121EEA7F80EAFFC0A9EA7F80ACEA3F00
+-AC121EAB120CC7FCA8121EEA7F80A2EAFFC0A4EA7F80A2EA1E000A4179C019>33
+-D<001E130F397F803FC000FF137F01C013E0A201E013F0A3007F133F391E600F30000013
+-00A401E01370491360A3000114E04913C00003130101001380481303000EEB070048130E
+-0018130C0038131C003013181C1C7DBE2D>I<013F4C7ED9FFC04B7E2601E0E015072607
+-C070150F48486C4B5A023E4BC7FC48486C5D48D90FC0EB01FE003ED90EF0EB07FCDA0F3F
+-133E007E903A070FFFF8F8007C0200EBC1F0EE000300FC6D6C495A604D5A171F95C8FC17
+-3E177E177C5F16015F007C4948485A1607007E5E003E49495A020E131F003F93C9FC6C49
+-133E260F803C137E0238137C6C6C485B3901E0E0016CB448485AD93F0049133F90C74848
+-EBFFC0030F903801E0E093398007C0704B4848487E4B153C033E90381F001C4B497F03FC
+-133E4B150F4A48017E7F0203147C5D4A4801FCEB0380140F5D4AC7FC5C143E5C14FC5C49
+-5A13034948027CEB07005C4948147E011F033E5B91C8140E013E153F017E6F5B017C9238
+-0F803C4917380001706C5A49923801E0E0496FB45A6C48043FC7FC41497BC34C>37
+-D<EC0FC0EC3FF0ECF878903801F01CEB03E049487E130FEC800F011F7FA2EB3F00A5EC80
+-0EA25DA25DA25D6D6C5AECC1C0A2ECC38002E7C7387FFFFCEB0FEE14FC4A020713C06D48
+-913801FE006E5DEF00F06D7E4D5A496C5D010F1503D91DFF4A5A013893C7FC496C6C5B01
+-E0150E48486C6C131E00036E131C2607801F143C000F6E5B001F6D6C1370263F000714F0
+-6F485A48D903FE5B913801FF03486D495A0487C8FCED7FCFED3FFE6F4814386D6D5AA200
+-7F6E6C14786D6D6C14704B6C14F06C6C496C6C13E0001F91393E3FC0016C6C903AFC1FF0
+-03C03D07FC07F007FC1F800001B5D8C001B512006C6C90C7EA7FFCD90FF8EC0FF03E437C
+-C047>I<121EEA7F8012FF13C0A213E0A3127FEA1E601200A413E013C0A3120113801203
+-13005A120E5A1218123812300B1C79BE19>I<1430147014E0EB01C0EB03801307EB0F00
+-131E133E133C5B13F85B12015B1203A2485AA2120F5BA2121F90C7FCA25AA3123E127EA6
+-127C12FCB2127C127EA6123E123FA37EA27F120FA27F1207A26C7EA212017F12007F1378
+-7F133E131E7FEB07801303EB01C0EB00E014701430145A77C323>I<12C07E12707E7E12
+-1E7E6C7E7F12036C7E7F12007F1378137CA27FA2133F7FA21480130FA214C0A3130714E0
+-A6130314F0B214E01307A614C0130FA31480A2131F1400A25B133EA25BA2137813F85B12
+-015B485A12075B48C7FC121E121C5A5A5A5A145A7BC323>I<EB03C0A2805CA600F0140F
+-00FC143F00FE147F00FF14FF393FC3C3FC390FE187F03903F18FC03900FDBF00EB3FFCEB
+-0FF0EB03C0EB0FF0EB3FFCEBFDBF3903F18FC0390FE187F0393FC3C3FC39FF03C0FF00FE
+-147F00FC143F00F0140F00001400A6805CA220277AC32D>I<121EEA7F8012FF13C0A213
+-E0A3127FEA1E601200A413E013C0A312011380120313005A120E5A1218123812300B1C79
+-8919>44 D<B512FEA617067F961E>I<121EEA7F80A2EAFFC0A4EA7F80A2EA1E000A0A79
+-8919>I<ED0180ED03C01507A21680150FA216005DA2151E153EA2153C157CA2157815F8
+-A25D1401A25D1403A25D1407A25D140FA24AC7FCA2141E143EA2143C147CA2147814F8A2
+-5C1301A25C1303A25C1307A25C130FA291C8FC5BA2131E133EA25BA2137813F8A25B1201
+-A25B1203A25B1207A25B120FA290C9FC5AA2121E123EA2123C127CA2127812F8A25A1260
+-225B7BC32D>I<EB01FE90380FFFC090383F03F090387C00F849137C48487F48487F4848
+-EB0F80A2000F15C04848EB07E0A3003F15F0A290C712034815F8A64815FCB3A26C15F8A5
+-6C6CEB07F0A3001F15E0A36C6CEB0FC0A26C6CEB1F80000315006C6C133E6C6C5B017C5B
+-90383F03F090380FFFC0D901FEC7FC263F7DBC2D>I<EB01C013031307131F137FEA07FF
+-B5FC139FEAF81F1200B3B3ACEB7FF0B612F8A31D3D78BC2D>I<EB07FC90383FFF8090B5
+-12E03903F01FF83907C007FC390F0001FE001E6D7E001C1580003CEC7FC05AED3FE01270
+-B4FC6DEB1FF07FA56C5A6CC7FC120CC813E0153FA216C0157F168015FF16004A5A5D4A5A
+-4A5A5D4A5A4A5A4AC7FC147E147C5C495A495A495A495A49C71270133E133C5B4914E048
+-5A485A485A48C7120148B6FCA25A4815C0B7FCA3243D7CBC2D>I<EB07FC90383FFF8090
+-38F80FE03901E003F839078001FCD80F007F000E6D7E001E1580D81F80137F486C14C07F
+-A27F5BA2121F6C5AC8138015FF1600A24A5AA24A5A5DEC07E04A5A023FC7FCEB1FFCECFF
+-809038000FE0EC07F86E7E6E7E6E7E1680ED7FC0A216E0153FA216F0A2120C123F487E48
+-7EA316E0A249137F6CC713C01278EDFF807E6C4913006C495A3907C007FC3903F80FF0C6
+-B55A013F1380D907F8C7FC243F7CBC2D>I<150E151E153EA2157EA215FE1401A21403EC
+-077E1406140E141CA214381470A214E0EB01C0A2EB0380EB0700A2130E5BA25B5BA25B5B
+-1201485A90C7FC5A120E120C121C5AA25A5AB8FCA3C8EAFE00AC4A7E49B6FCA3283E7EBD
+-2D>I<00061403D80780131F01F813FE90B5FC5D5D5D15C092C7FC14FCEB3FE090C9FCAC
+-EB01FE90380FFF8090383E03E090387001F8496C7E49137E497F90C713800006141FC813
+-C0A216E0150FA316F0A3120C127F7F12FFA416E090C7121F12FC007015C012780038EC3F
+-80123C6CEC7F00001F14FE6C6C485A6C6C485A3903F80FE0C6B55A013F90C7FCEB07F824
+-3F7CBC2D>I<EC1FE0ECFFF8903803F03E90380FC00F90391F000780133E017EEB1FC049
+-133F4848137F12035B12074848EB3F80ED1F00001F91C7FC5BA2123FA3485AA214FE9038
+-87FF8039FF8F07E090389C01F09038B800FC01B0137E13F0497F16804914C0A2ED1FE0A3
+-4914F0A5127FA6123F6D14E0A2121FED3FC0A26C6C1480A20007EC7F006C6C137E6C6C5B
+-6C6C485A90387E07F06DB45A010F1380D903FCC7FC243F7CBC2D>I<1238123C123F90B6
+-12FCA316F85A16F016E00078C712010070EC03C0ED078016005D48141E151C153C5DC812
+-7015F04A5A5D14034A5A92C7FC5C141EA25CA2147C147814F8A213015C1303A31307A313
+-0F5CA2131FA6133FAA6D5A0107C8FC26407BBD2D>I<EB03FC90381FFF8090387C07E090
+-38F001F83901E0007C48487F48487F48C7FCED0F80121E16C0003E1407A4123FA26DEB0F
+-807F6C6C131F6D140001FC133E6C6C5B9038FF80786C6D5A6CEBF3E06CEBFF806C91C7FC
+-133F6D13C06D7F013F13F801787F48486C7E3903E01FFF48486C1380260F800313C04848
+-7E489038007FE0003E143F007E141F007CEC0FF01507481403A31501A46C15E0007C1403
+-A2007E15C06C14076CEC0F806DEB1F006C6C133ED807F05B3901FC03F86CB512E0011F13
+-80D903FCC7FC243F7CBC2D>I<EB03FCEB1FFF90387E07C09038FC03F048486C7E48486C
+-7E4848137C000F147E4848137F81003F15805B007F15C0A2151F12FF16E0A516F0A5127F
+-153FA36C7EA2001F147F120F6C6C13FF6D13DF000313013900F8039F90387E0F1FD91FFE
+-13E0EB07F090C7FCA2ED3FC0A41680157FD80F801400487E486C13FEA24A5A5D49485AEB
+-8007391E000FE0001F495A260FC07FC7FC3803FFFE6C13F838003FC0243F7CBC2D>I<12
+-1EEA7F80A2EAFFC0A4EA7F80A2EA1E00C7FCB3121EEA7F80A2EAFFC0A4EA7F80A2EA1E00
+-0A2779A619>I<121EEA7F80A2EAFFC0A4EA7F80A2EA1E00C7FCB3121E127FEAFF80A213
+-C0A4127F121E1200A412011380A3120313005A1206120E120C121C5A1230A20A3979A619
+->I<007FB912E0BA12F0A26C18E0CDFCAE007FB912E0BA12F0A26C18E03C167BA147>61
+-D<EB1FF890B5FC3903E01FC0390F0007F0001EEB03F848EB01FC4814FE140000FE14FF7E
+-7FA46CC7FC123EC7EA01FEA2EC03FCEC07F815F0EC0FC0EC1F80EC3F00143E5C147814F8
+-5C13015CA2495AA25CAB91C7FC90C8FCA8EB0780EB1FE0A2497EA46D5AA2EB078020407B
+-BF2B>63 D<15074B7EA34B7EA34B7EA34B7EA34B7E15E7A2913801C7FC15C3A291380381
+-FEA34AC67EA3020E6D7EA34A6D7EA34A6D7EA34A6D7EA34A6D7EA349486D7E91B6FCA249
+-819138800001A249C87EA24982010E157FA2011E82011C153FA2013C820138151FA20178
+-82170F13FC00034C7ED80FFF4B7EB500F0010FB512F8A33D417DC044>65
+-D<B712FCEEFF8017F00001903980000FF86C6CC7EA03FE707E701380EF7FC0EF3FE0A2EF
+-1FF0A218F8A3170F171FA318F0A2EF3FE0177F18C0EFFF804C1300EE03FCEE0FF8EE7FE0
+-91B6C7FC17E091C7EA07FCEE01FE933800FF80EF7FC0EF3FE0EF1FF018F8170F18FC1707
+-A218FEA718FC170FA2EF1FF818F0173FEF7FE0EFFFC00403138048486C90380FFE00B85A
+-17E094C7FC373E7DBD40>I<DB3FF01306912603FFFE130E020F9038FF801E913A3FF007
+-E03E9139FF8000F8D903FEC7EA7C7ED907F8EC1EFE4948140FD93FE0140749481403495A
+-91C812014848150012034848167E5B000F173EA24848161EA2123F5B180E127FA3491600
+-12FFAC127F7F180EA2123FA27F001F171E181C6C7EA20007173C6D16386C6C1678000117
+-706C6C16F06EEC01E06D6C15C06D6C1403D90FF0EC07806D6CEC1F00D903FE143E902600
+-FF8013F891393FF007F0020FB512C0020391C7FC9138003FF037427BBF42>I<B712FCEE
+-FF8017E000019039C0001FF86C6C48EB03FEEE00FF717E717EEF0FE084717E717E170184
+-717EA21980187F19C0A3F03FE0A519F0AB19E0A5F07FC0A21980A218FF19004D5AA24D5A
+-6017074D5A4D5AEF7FC04DC7FCEE03FE48486CEB1FF8B85A178004FCC8FC3C3E7DBD45>
+-I<B912E0A300019038C000016C6C48EB001FEF0FF01703A217011700A31870A41838161C
+-A41800A2163CA2167C16FC150391B5FCA3EC80031500167C163CA2161CA21807A3180E93
+-C7FCA4181E181CA2183CA2187CA218F8170117031707171F48486CEB01FFB912F0A3383E
+-7DBD3E>I<B91280A300019038C000036C6C48EB007FEF1FC0170F1707A21703A31701A4
+-EF00E0A21638A31800A31678A216F81501150791B5FCA3EC8007150115001678A21638A6
+-93C8FCAF3801FFE0B612F0A3333E7DBD3B>I<DB3FE0130C912603FFFE131C021F9038FF
+-803C913A7FF00FC07C9139FF0001F0D903FC90380078FC4948143DD91FE0141F4948140F
+-4948140701FF15034890C8FC491501485A000716005B000F177C5B001F173CA2485AA218
+-1C127FA25B95C7FC12FFAB041FB512F0127FA26D9139000FFE00EF03FC123FA27F121FA2
+-6C7EA212077F12036C7E7F6C7F6D6C14076D7E6D6C140FD907F8141ED903FEEC3C7C9026
+-00FF80EBF83C913A7FF007F01C021FB5EAC00C020391C8FC9138003FF03C427BBF47>I<
+-B6D8C01FB512F8A3000101E0C7383FFC0026007F80EC0FF0B3A691B7FCA30280C7120FB3
+-A92601FFE0EC3FFCB6D8C01FB512F8A33D3E7DBD44>I<B612F0A3C6EBF000EB3FC0B3B3
+-B2EBFFF0B612F0A31C3E7EBD21>I<011FB512FCA3D9000713006E5A1401B3B3A6123FEA
+-7F80EAFFC0A44A5A1380D87F005B007C130700385C003C495A6C495A6C495A2603E07EC7
+-FC3800FFF8EB3FC026407CBD2F>I<B600C090387FFFFCA3000101E0C7000F138026007F
+-80913807FE0018F818E0604D5A4DC7FC173E5F5F4C5A4C5A4C5A4C5A4CC8FC163E5E5E4B
+-5A4B5AED07804B7E151F4B7E4B7E15FF913881EFF8913883C7FCEC878791388F03FE9138
+-9E01FF14BCDAF8007F4A6D7E5C4A6D7E4A6D7EA2707E707EA2707E707EA2707F717E8417
+-3F717E717EA2717E848419802601FFE04A13C0B600C090B6FCA3403E7DBD47>I<B612F8
+-A3000101E0C9FC38007F80B3B0EF0380A517071800A45FA35FA25F5F5F4C5A160748486C
+-133FB8FCA3313E7DBD39>I<B500C093383FFFF0A300016D93387FF800D8007F18E0D977
+-F016EFA3D973F8ED01CFA2D971FCED038FA3D970FEED070FA26E150E80A26E6C141CA36E
+-6C1438A26E6C1470A36E6C14E0A26E6CEB01C0A36E6CEB0380A36E6CEB0700A2037F130E
+-A36F6C5AA26F6C5AA36F6C5AA25FED07F0A2923803F9C0A36FB45AA26F90C7FCA213F848
+-6C147ED807FFEF3FF8B500F8013C011FB512F0A34C3E7DBD53>I<B56C91B512F88080D8
+-007F030713006EEC01FC6E6E5A1870EB77FCEB73FEA2EB71FF01707FA26E7E6E7EA26E7E
+-6E7EA26E7E6E7EA26E7E6E7FA26F7E6F7EA26F7E6F7EA26F7E6F7EA26F7E6F1380A2EE7F
+-C0EE3FE0A2EE1FF0EE0FF8A2EE07FCEE03FEA2EE01FF7013F0A2177F173FA2171F170FA2
+-170701F81503487ED807FF1501B500F81400A218703D3E7DBD44>I<ED7FE0913807FFFE
+-91391FC03F8091397E0007E04948EB03F8D907F0EB00FE4948147F49486E7E49486E7E49
+-C86C7E01FE6F7E00018349150300038348486F7EA248486F7EA2001F188049167F003F18
+-C0A3007F18E049163FA300FF18F0AC007F18E06D167FA4003F18C0A26C6CEEFF80A36C6C
+-4B1300A26C6C4B5A00035F6D150700015F6C6C4B5A6D5E6D6C4A5A6D6C4A5A6D6C4AC7FC
+-6D6C14FED901FCEB03F8D9007FEB0FE091391FC03F80912607FFFEC8FC9138007FE03C42
+-7BBF47>I<B712F8EEFF8017E000019039C0003FF86C6C48EB07FCEE01FE707EEF7F80EF
+-3FC018E0A2EF1FF0A218F8A818F0A2EF3FE0A218C0EF7F80EFFF004C5AEE07FCEE3FF091
+-B612C04CC7FC0280C9FCB3A73801FFE0B612C0A3353E7DBD3E>I<ED7FE0913807FFFE91
+-391FC03F8091397F000FE0D901FCEB03F8D907F0EB00FE4948147F49486E7E49486E7E49
+-C86C7E498248486F7E49150300038348486F7EA2000F834981001F1880A24848EE7FC0A3
+-007F18E0A249163FA200FF18F0AC007F18E0A26D167FA3003F18C0A26C6CEEFF80A3000F
+-18006D5D0007DA0F805B6C6C90393FE003FCED70706C6C496C485A6C6C48486C485A017F
+-D9800E5BD93F819038061FC0D91FC19038073F80D90FE14AC7FCD907F1EB03FE902601FD
+-C013F8903A007EE007E091271FF03FC013180207B5FC9139007FE1E0DB00011438837113
+-78A2706C13F0EFFF0318FFA27113E0A37113C0711380711300715AEF01F83D527BBF47>
+-I<B712C016FCEEFF800001D9C00013E06C6C48EB1FF0EE07FCEE01FE707E84717EA2717E
+-A284A760177F606017FF95C7FCEE01FCEE07F8EE1FE0EEFF8091B500FCC8FC16F0913880
+-01FCED003FEE1FC0707E707E83160383160183A383A484A4F0C004190EA28218E0057F13
+-1E2601FFE0161CB600C0EB3FF094381FF83805071370CA3801FFE09438003F803F407DBD
+-43>I<D907FC130C90391FFF801C017FEBF03C3901FC03F83A03F0007E7CD807C0EB1FFC
+-4848130F001F140748C71203003E1401007E1400A2007C157C12FCA2163CA36C151CA27E
+-A26C6C14007F7FEA3FF8EBFF806C13F86CEBFF806C14F06C14FC6C14FF6C15C0013F14E0
+-010714F0EB007F020713F89138007FFC150FED07FE15031501ED00FFA200E0157FA3163F
+-A27EA3163E7E167E6C157C6C15FC6C15F86D13016DEB03F06DEB07E0D8F9FCEB0FC03AF0
+-7F803F8090391FFFFE00D8E00713F839C0007FC028427BBF33>I<003FB91280A3903AF0
+-007FE001018090393FC0003F48C7ED1FC0007E1707127C00781703A300701701A548EF00
+-E0A5C81600B3B14B7E4B7E0107B612FEA33B3D7DBC42>I<B600C090B512F8A3000101E0
+-C70007130026007F80EC01FC715A1870B3B3A4013F16F06E5DA21701011F5E80010F1503
+-6E4A5A010793C7FC6D6C5C6D6C141E6D6C5C027F14F86E6C485A91390FF00FE00203B512
+-80020049C8FCED1FF03D407DBD44>I<B691380FFFFEA3000301E0020113E06C01809138
+-007F806CEF3F00017F163E181C6E153C013F1638A26E1578011F1670A26D6C5DA26E1401
+-01075EA26E140301035EA26D6C4AC7FCA2806D150EA26F131E027F141CA26F133C023F14
+-38A26E6C5BA26F13F0020F5CA2EDF80102075CA26E6C485AA2EDFE07020191C8FCA26F5A
+-6E130EA2ED7F9CA216DCED3FF8A36F5AA36F5AA26F5AA36F5A3F407EBD44>I<B500FE01
+-7FB5D88007B5FCA3000301C0010101E0C713F86C90C849EC3FE07148EC0F807E7215006E
+-143F017F190E84A26D6C60A24D7E6D6C60A2EFE7F86D6C60A2933801C3FC6E18F0010761
+-04037F6E0281140101036104077F17006D6C4D5AA2040EEB7F806D6C4DC7FCA24CEB3FC0
+-DA7F80160EA24CEB1FE003C0161E023F171C047814F0DBE070010F133C021F173804F014
+-F84C1307DA0FF05EA2DBF1C0EB03FCDA07F95EA2DBFB80EB01FEDA03FF6F5AA293C8FCA2
+-6E5FA24B157F020094C8FCA24B81037C153EA20378151E0338151C58407EBD5D>I<007F
+-B5D8C003B512E0A3C649C7EBFC00D93FF8EC3FE06D48EC1F806D6C92C7FC171E6D6C141C
+-6D6C143C5F6D6C14706D6D13F04C5ADA7FC05B023F13036F485ADA1FF090C8FC020F5BED
+-F81E913807FC1C163C6E6C5A913801FF7016F06E5B6F5AA26F7E6F7EA28282153FED3BFE
+-ED71FF15F103E07F913801C07F0203804B6C7EEC07004A6D7E020E6D7E5C023C6D7E0238
+-6D7E14784A6D7E4A6D7F130149486E7E4A6E7E130749C86C7E496F7E497ED9FFC04A7E00
+-076DEC7FFFB500FC0103B512FEA33F3E7EBD44>I<B66C0103B51280A3000101F0C8EBF8
+-006C6C48ED3FC0725A013F041EC7FC6D7E606D6C15386D6C1578606D6C5D6E14016D5E6D
+-6D1303606E6C49C8FC6E6C5B170E6E6C131E171C6E6C5B6E6C137817706E6C13F06F5B6E
+-13016EEB83C05FED7FC7DB3FE7C9FC16EFED1FFE5E150F6F5AB3A4ED1FFC020FB512FCA3
+-413E7FBD44>I<003FB712F8A391C7EA1FF013F801E0EC3FE00180EC7FC090C8FC003EED
+-FF80A2003C4A1300007C4A5A12784B5A4B5AA200704A5AA24B5A4B5AA2C8485A4A90C7FC
+-A24A5A4A5AA24A5AA24A5A4A5AA24A5A4A5AA24990C8FCA2495A4948141CA2495A495AA2
+-495A495A173C495AA24890C8FC485A1778485A484815F8A24848140116034848140F4848
+-143FED01FFB8FCA32E3E7BBD38>I<EAFFFCA4EAF000B3B3B3B3ABEAFFFCA40E5B77C319>
+-I<486C13C00003130101001380481303000EEB070048130E0018130C0038131C00301318
+-0070133800601330A300E01370481360A400CFEB678039FFC07FE001E013F0A3007F133F
+-A2003F131F01C013E0390F0007801C1C73BE2D>I<EAFFFCA4EA003CB3B3B3B3ABEAFFFC
+-A40E5B7FC319>I<EB0FF8EBFFFE3903F01F8039078007E0000F6D7E9038E001F8D81FF0
+-7F6E7EA3157F6C5AEA0380C8FCA4EC1FFF0103B5FC90381FF87FEB7F803801FC00EA07F8
+-EA0FE0485A485AA248C7FCEE038012FEA315FFA3007F5BEC03BF3B3F80071F8700261FC0
+-0E13CF3A07F03C0FFE3A01FFF807FC3A003FC001F0292A7DA82D>97
+-D<EA01FC12FFA3120712031201B1EC03FC91381FFF8091387C07E09039FDE001F09039FF
+-C000FC4A137E91C77E49158049141F17C0EE0FE0A217F0A2160717F8AA17F0A2160FA217
+-E0161F17C06D1580EE3F006D5C6E13FE9039F3C001F89039F1E003F09039E0780FC09026
+-C03FFFC7FCC7EA07F82D407EBE33>I<49B4FC010F13E090383F00F8017C131E4848131F
+-4848137F0007ECFF80485A5B121FA24848EB7F00151C007F91C7FCA290C9FC5AAB6C7EA3
+-003FEC01C07F001F140316806C6C13076C6C14000003140E6C6C131E6C6C137890383F01
+-F090380FFFC0D901FEC7FC222A7DA828>I<ED01FC15FFA3150715031501B114FF010713
+-E190381F80F990387E003D49131FD803F81307485A49130348481301121F123F5B127FA2
+-90C7FCA25AAA7E7FA2123FA26C7E000F14037F000714076C6C497E6C6C497ED8007C0179
+-13F890383F01F190380FFFC1903A01FE01FC002D407DBE33>I<EB01FE90380FFFC09038
+-3F03F09038FC01F848486C7E4848137E48487F000F158049131F001F15C04848130FA212
+-7F16E090C7FCA25AA290B6FCA290C9FCA67EA27F123F16E06C7E1501000F15C06C6C1303
+-6DEB07806C6C1400C66C131E017E5B90381F80F8903807FFE0010090C7FC232A7EA828>
+-I<EC1FC0EC7FF8903801F83C903807E07E90380FC0FFEB1FC1EB3F811401137FEC00FE01
+-FE137C1500AEB6FCA3C648C7FCB3AE487E007F13FFA320407EBF1C>I<167C903903F801
+-FF903A1FFF078F8090397E0FDE1F9038F803F83803F001A23B07E000FC0600000F6EC7FC
+-49137E001F147FA8000F147E6D13FE00075C6C6C485AA23901F803E03903FE0FC026071F
+-FFC8FCEB03F80006CAFC120EA3120FA27F7F6CB512E015FE6C6E7E6C15E06C810003813A
+-0FC0001FFC48C7EA01FE003E140048157E825A82A46C5D007C153E007E157E6C5D6C6C49
+-5A6C6C495AD803F0EB0FC0D800FE017FC7FC90383FFFFC010313C0293D7EA82D>I<EA01
+-FC12FFA3120712031201B1EC01FE913807FFC091381E07E091387803F09138E001F8D9FD
+-C07F148001FF6D7E91C7FCA25BA25BB3A6486C497EB5D8F87F13FCA32E3F7DBE33>I<EA
+-01E0EA07F8A2487EA46C5AA2EA01E0C8FCACEA01FC127FA3120712031201B3AC487EB512
+-F0A3143E7DBD1A>I<1478EB01FEA2EB03FFA4EB01FEA2EB00781400AC147FEB7FFFA313
+-017F147FB3B3A5123E127F38FF807E14FEA214FCEB81F8EA7F01387C03F0381E07C0380F
+-FF803801FC00185185BD1C>I<EA01FC12FFA3120712031201B292B51280A392383FFC00
+-16E0168093C7FC153C5D5D4A5AEC07C04A5A4AC8FC143E147F4A7E13FD9038FFDFC0EC9F
+-E0140F496C7E01FC7F496C7E1401816E7E81826F7E151F826F7EA282486C14FEB539F07F
+-FFE0A32B3F7EBE30>I<EA01FC12FFA3120712031201B3B3B1487EB512F8A3153F7DBE1A>
+-I<2701F801FE14FF00FF902707FFC00313E0913B1E07E00F03F0913B7803F03C01F80007
+-903BE001F87000FC2603F9C06D487F000101805C01FBD900FF147F91C75B13FF4992C7FC
+-A2495CB3A6486C496CECFF80B5D8F87FD9FC3F13FEA347287DA74C>I<3901F801FE00FF
+-903807FFC091381E07E091387803F000079038E001F82603F9C07F0001138001FB6D7E91
+-C7FC13FF5BA25BB3A6486C497EB5D8F87F13FCA32E287DA733>I<14FF010713E090381F
+-81F890387E007E01F8131F4848EB0F804848EB07C04848EB03E0000F15F04848EB01F8A2
+-003F15FCA248C812FEA44815FFA96C15FEA36C6CEB01FCA3001F15F86C6CEB03F0A26C6C
+-EB07E06C6CEB0FC06C6CEB1F80D8007EEB7E0090383F81FC90380FFFF0010090C7FC282A
+-7EA82D>I<3901FC03FC00FF90381FFF8091387C0FE09039FDE003F03A03FFC001FC6C49
+-6C7E91C7127F49EC3F805BEE1FC017E0A2EE0FF0A3EE07F8AAEE0FF0A4EE1FE0A2EE3FC0
+-6D1580EE7F007F6E13FE9138C001F89039FDE007F09039FC780FC0DA3FFFC7FCEC07F891
+-C9FCAD487EB512F8A32D3A7EA733>I<02FF131C0107EBC03C90381F80F090397F00387C
+-01FC131CD803F8130E4848EB0FFC150748481303121F485A1501485AA448C7FCAA6C7EA3
+-6C7EA2001F14036C7E15076C6C130F6C7E6C6C133DD8007E137990383F81F190380FFFC1
+-903801FE0190C7FCAD4B7E92B512F8A32D3A7DA730>I<3901F807E000FFEB1FF8EC787C
+-ECE1FE3807F9C100031381EA01FB1401EC00FC01FF1330491300A35BB3A5487EB512FEA3
+-1F287EA724>I<90383FC0603901FFF8E03807C03F381F000F003E1307003C1303127C00
+-78130112F81400A27E7E7E6D1300EA7FF8EBFFC06C13F86C13FE6C7F6C1480000114C0D8
+-003F13E0010313F0EB001FEC0FF800E01303A214017E1400A27E15F07E14016C14E06CEB
+-03C0903880078039F3E01F0038E0FFFC38C01FE01D2A7DA824>I<131CA6133CA4137CA2
+-13FCA2120112031207001FB512C0B6FCA2D801FCC7FCB3A215E0A912009038FE01C0A2EB
+-7F03013F138090381F8700EB07FEEB01F81B397EB723>I<D801FC14FE00FF147FA30007
+-14030003140100011400B3A51501A31503120015076DEB06FF017E010E13806D4913FC90
+-381FC078903807FFE00100903880FE002E297DA733>I<B539E00FFFE0A32707FE000313
+-006C48EB00FC5E00015D7F00005DA26D13016D5CA26D6C485AA2ECC007011F91C7FCA290
+-380FE00EA2ECF01E0107131CA26D6C5AA2ECFC7801011370A2ECFEF001005BA2EC7FC0A3
+-6E5AA26EC8FCA3140E2B287EA630>I<B53BC3FFFE03FFF8A3290FFE003FE00013C06C48
+-6D48EB3F806C4817006D010F141E00016F131C15076D163C00004A6C1338A2017F5E4B7E
+-151DD93F805DED3DFC1538D91FC04A5AED78FE9238707E03D90FE0017F5BEDE03F02F014
+-0701070387C7FC9138F1C01F02F9148F010315CE9138FB800F02FF14DE6D15FCED00076D
+-5DA24A1303027E5CA2027C1301023C5C023813003D287EA642>I<B539F01FFFE0A30003
+-D9C00F1300C690388007F8D97F0013E002805BD93FC05B011F49C7FC90380FE00EECF01E
+-6D6C5A01035B6D6C5A6E5AEB00FF6E5A6E5A81141F814A7E81147BECF1FC903801E1FEEC
+-C0FF01037F49486C7ED90F007F011E6D7E013E130F496D7E01FC80486C80000F4A7EB539
+-803FFFF8A32D277FA630>I<B539E00FFFE0A32707FE000313006C48EB01FC6F5A00015D
+-7F00005DA2017F495AA2EC8003013F5CA26D6C48C7FCA26E5A010F130EA26D6C5AA2ECF8
+-3C01031338A26D6C5AA2ECFEF001005BA2EC7FC0A36E5AA36EC8FCA2140EA2141E141C14
+-3C1438A2147800181370127EB45BA2495AA248485AD87E07C9FCEA780EEA3C3CEA1FF8EA
+-07E02B3A7EA630>I<001FB61280A2EBE0000180140049485A001E495A121C4A5A003C49
+-5A141F00385C4A5A147F5D4AC7FCC6485AA2495A495A130F5C495A90393FC00380A2EB7F
+-80EBFF005A5B484813071207491400485A48485BA248485B4848137F00FF495A90B6FCA2
+-21277EA628>I<B812F0A22C0280982D>I 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
+-D<EB01FCEB07FF011F13C0497F497F90B57EA24880A24880A76C5CA26C5CA26D5B6D5B6D
+-5B010790C8FCEB01FC90CAFCB2903801FFFC007FB5FCA8C67E131F7FB3B3B3A5B81280A8
+-297979F835>105 D<903801FFFCB6FCA8C67E131F7FB3B3B3B3B3ABB812C0A82A7879F7
+-35>108 D<902601FFF891380FFFE0B692B512FE05036E7E050F15E0053F15F84D819327
+-01FFF01F7F4CD900077FDC07FC6D80C66CDA0FF06D80011FDA1FC07F6D4A48824CC8FC04
+-7E6F7F5EEDF9F85E03FB707F5E15FF5EA25EA293C9FCA45DB3B3A6B8D8E003B81280A861
+-4E79CD6C>110 D<902601FFFCEC7FFEB6020FB512F0057F14FE4CB712C0040716F0041F
+-82047F16FE93B5C66C7F92B500F0010F14C0C66C0380010380011F4AC76C806D4A6E8004
+-F06F7F4C6F7F4C6F7F4C8193C915804B7014C0861DE0A27414F0A27414F8A47513FCA575
+-13FEAF5113FCA598B512F8A31DF0621DE0621DC0621D806F5E701800704B5B505B704B5B
+-7092B55A04FC4A5C704A5C706C010F5C05E0013F49C7FC9227FE7FFC01B55A70B712F004
+-0F16C0040393C8FC040015F8053F14C0050301F0C9FC94CCFCB3A6B812E0A85F6F7ACD6C
+->112 D<0007BA12FC1AFEA503E0C714FC4AC74814F84801F04A14F05C02804A14E091C8
+-4814C04D14805B494B14004D5B4992B55AA24C5C494A5C615E4C5C001F4B5C5B4C91C7FC
+-4C5B93B55AA24B5CC8485C4B5CA24B5C4B5C4B91C8FCA24B5B92B55AA24A5C4A5C4A4A14
+-FFA24A5C4A5C4A91C8FC614A4915FE91B55A495CA2495C494A14035E5B495C4991C81207
+-A24949ED0FFC90B55A484A151FA2484A153F484A157F484A15FF1803484A140F4891C812
+-3F48490207B5FC91B9FCBB12F8A57E484D7BCC56>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(1996-2000)j
+-(Julian)28 b(Sew)m(ard)2605 1731 y(v)m(ersion)i(1.0)h(of)g(21)g(Marc)m
+-(h)g(2000)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(This)24 b(program,)j
+-Fj(bzip2)p Fl(,)e(and)g(asso)s(ciated)i(library)c Fj(libbzip2)p
+-Fl(,)i(are)h(Cop)m(yrigh)m(t)g(\(C\))g(1996-2000)j(Julian)150
+-408 y(R)h(Sew)m(ard.)40 b(All)29 b(righ)m(ts)h(reserv)m(ed.)150
+-565 y(Redistribution)f(and)i(use)h(in)f(source)h(and)g(binary)e(forms,)
+-j(with)e(or)h(without)f(mo)s(di\014cation,)g(are)i(p)s(er-)150
+-675 y(mitted)d(pro)m(vided)f(that)i(the)f(follo)m(wing)f(conditions)g
+-(are)i(met:)225 832 y Fi(\017)60 b Fl(Redistributions)26
+-b(of)k(source)g(co)s(de)g(m)m(ust)g(retain)f(the)h(ab)s(o)m(v)m(e)h
+-(cop)m(yrigh)m(t)g(notice,)f(this)f(list)f(of)i(con-)330
+-941 y(ditions)e(and)i(the)h(follo)m(wing)e(disclaimer.)225
+-1076 y Fi(\017)60 b Fl(The)33 b(origin)f(of)h(this)f(soft)m(w)m(are)j
+-(m)m(ust)e(not)h(b)s(e)e(misrepresen)m(ted;)i(y)m(ou)g(m)m(ust)f(not)g
+-(claim)g(that)h(y)m(ou)330 1185 y(wrote)d(the)h(original)d(soft)m(w)m
+-(are.)44 b(If)31 b(y)m(ou)g(use)g(this)f(soft)m(w)m(are)i(in)e(a)h(pro)
+-s(duct,)g(an)f(ac)m(kno)m(wledgmen)m(t)330 1295 y(in)f(the)i(pro)s
+-(duct)e(do)s(cumen)m(tation)h(w)m(ould)f(b)s(e)h(appreciated)g(but)g
+-(is)f(not)i(required.)225 1429 y Fi(\017)60 b Fl(Altered)21
+-b(source)g(v)m(ersions)f(m)m(ust)h(b)s(e)f(plainly)e(mark)m(ed)j(as)g
+-(suc)m(h,)i(and)d(m)m(ust)h(not)g(b)s(e)f(misrepresen)m(ted)330
+-1539 y(as)31 b(b)s(eing)e(the)h(original)f(soft)m(w)m(are.)225
+-1674 y Fi(\017)60 b Fl(The)27 b(name)h(of)f(the)h(author)f(ma)m(y)h
+-(not)g(b)s(e)f(used)g(to)h(endorse)f(or)h(promote)g(pro)s(ducts)e
+-(deriv)m(ed)g(from)330 1783 y(this)j(soft)m(w)m(are)j(without)d(sp)s
+-(eci\014c)h(prior)e(written)i(p)s(ermission.)150 1965
+-y(THIS)37 b(SOFTW)-10 b(ARE)38 b(IS)f(PR)m(O)m(VIDED)i(BY)g(THE)f(A)m
+-(UTHOR)g(\\AS)g(IS")g(AND)h(ANY)f(EXPRESS)150 2074 y(OR)31
+-b(IMPLIED)h(W)-10 b(ARRANTIES,)31 b(INCLUDING,)i(BUT)f(NOT)f(LIMITED)g
+-(TO,)h(THE)f(IMPLIED)150 2184 y(W)-10 b(ARRANTIES)27
+-b(OF)h(MER)m(CHANT)-8 b(ABILITY)28 b(AND)g(FITNESS)f(F)m(OR)g(A)h(P)-8
+-b(AR)g(TICULAR)28 b(PUR-)150 2294 y(POSE)37 b(ARE)g(DISCLAIMED.)h(IN)f
+-(NO)h(EVENT)f(SHALL)g(THE)g(A)m(UTHOR)h(BE)g(LIABLE)g(F)m(OR)150
+-2403 y(ANY)56 b(DIRECT,)f(INDIRECT,)h(INCIDENT)-8 b(AL,)56
+-b(SPECIAL,)e(EXEMPLAR)-8 b(Y,)57 b(OR)e(CONSE-)150 2513
+-y(QUENTIAL)48 b(D)m(AMA)m(GES)i(\(INCLUDING,)g(BUT)f(NOT)f(LIMITED)g
+-(TO,)g(PR)m(OCUREMENT)150 2622 y(OF)35 b(SUBSTITUTE)e(GOODS)i(OR)f(SER)
+--10 b(VICES;)34 b(LOSS)f(OF)i(USE,)g(D)m(A)-8 b(T)g(A,)36
+-b(OR)f(PR)m(OFITS;)f(OR)150 2732 y(BUSINESS)28 b(INTERR)m(UPTION\))g
+-(HO)m(WEVER)i(CA)m(USED)f(AND)g(ON)g(ANY)g(THEOR)-8 b(Y)29
+-b(OF)g(LIA-)150 2842 y(BILITY,)36 b(WHETHER)g(IN)g(CONTRA)m(CT,)g
+-(STRICT)e(LIABILITY,)i(OR)g(TOR)-8 b(T)35 b(\(INCLUDING)150
+-2951 y(NEGLIGENCE)45 b(OR)g(OTHER)-10 b(WISE\))44 b(ARISING)h(IN)g(ANY)
+-h(W)-10 b(A)i(Y)46 b(OUT)e(OF)i(THE)e(USE)h(OF)150 3061
+-y(THIS)29 b(SOFTW)-10 b(ARE,)31 b(EVEN)f(IF)g(AD)m(VISED)i(OF)e(THE)g
+-(POSSIBILITY)e(OF)j(SUCH)f(D)m(AMA)m(GE.)150 3218 y(Julian)e(Sew)m
+-(ard,)i(Cam)m(bridge,)g(UK.)150 3374 y Fj(jseward@acm.org)150
+-3531 y(http://sourceware.cygnus)o(.com)o(/bzi)o(p2)150
+-3688 y(http://www.cacheprof.org)150 3845 y(http://www.muraroa.demon)o
+-(.co.)o(uk)150 4002 y(bzip2)p Fl(/)p Fj(libbzip2)d Fl(v)m(ersion)j(1.0)
+-i(of)e(21)h(Marc)m(h)g(2000.)150 4159 y(P)-8 b(A)g(TENTS:)40
+-b(T)-8 b(o)40 b(the)g(b)s(est)g(of)g(m)m(y)g(kno)m(wledge,)j
+-Fj(bzip2)38 b Fl(do)s(es)i(not)g(use)g(an)m(y)g(paten)m(ted)h
+-(algorithms.)150 4268 y(Ho)m(w)m(ev)m(er,)33 b(I)e(do)f(not)h(ha)m(v)m
+-(e)h(the)f(resources)g(a)m(v)-5 b(ailable)30 b(to)h(carry)g(out)g(a)g
+-(full)d(paten)m(t)k(searc)m(h.)42 b(Therefore)150 4378
+-y(I)30 b(cannot)h(giv)m(e)g(an)m(y)g(guaran)m(tee)h(of)e(the)h(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 Fh(1)80 b(In)l(tro)t(duction)150 555 y
+-Fj(bzip2)20 b Fl(compresses)h(\014les)f(using)g(the)h(Burro)m
+-(ws-Wheeler)g(blo)s(c)m(k-sorting)f(text)j(compression)d(algorithm,)150
+-665 y(and)33 b(Hu\013man)g(co)s(ding.)50 b(Compression)32
+-b(is)h(generally)g(considerably)f(b)s(etter)i(than)f(that)h(ac)m(hiev)m
+-(ed)h(b)m(y)150 775 y(more)f(con)m(v)m(en)m(tional)g(LZ77/LZ78-based)g
+-(compressors,)g(and)f(approac)m(hes)h(the)f(p)s(erformance)g(of)h(the)
+-150 884 y(PPM)c(family)f(of)i(statistical)f(compressors.)150
+-1041 y Fj(bzip2)k Fl(is)h(built)e(on)i(top)h(of)g Fj(libbzip2)p
+-Fl(,)e(a)i(\015exible)e(library)f(for)i(handling)e(compressed)i(data)i
+-(in)d(the)150 1151 y Fj(bzip2)c Fl(format.)43 b(This)30
+-b(man)m(ual)g(describ)s(es)g(b)s(oth)g(ho)m(w)i(to)g(use)f(the)g
+-(program)g(and)g(ho)m(w)g(to)h(w)m(ork)f(with)150 1260
+-y(the)d(library)d(in)m(terface.)41 b(Most)28 b(of)g(the)g(man)m(ual)f
+-(is)g(dev)m(oted)i(to)f(this)f(library)-8 b(,)26 b(not)i(the)g
+-(program,)g(whic)m(h)150 1370 y(is)h(go)s(o)s(d)i(news)e(if)h(y)m(our)g
+-(in)m(terest)h(is)e(only)g(in)h(the)g(program.)150 1527
+-y(Chapter)24 b(2)g(describ)s(es)f(ho)m(w)h(to)h(use)f
+-Fj(bzip2)p Fl(;)h(this)e(is)g(the)i(only)e(part)h(y)m(ou)h(need)f(to)h
+-(read)f(if)f(y)m(ou)h(just)g(w)m(an)m(t)150 1636 y(to)35
+-b(kno)m(w)f(ho)m(w)g(to)g(op)s(erate)h(the)f(program.)51
+-b(Chapter)34 b(3)g(describ)s(es)e(the)i(programming)f(in)m(terfaces)h
+-(in)150 1746 y(detail,)23 b(and)d(Chapter)h(4)h(records)f(some)h
+-(miscellaneous)e(notes)i(whic)m(h)e(I)h(though)m(t)h(ough)m(t)g(to)g(b)
+-s(e)f(recorded)150 1855 y(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 Fj(bzip2)2375 b Fl(3)150 299 y Fh(2)80 b(Ho)l(w)53
+-b(to)g(use)g Fg(bzip2)150 566 y Fl(This)29 b(c)m(hapter)i(con)m(tains)f
+-(a)h(cop)m(y)g(of)g(the)f Fj(bzip2)f Fl(man)h(page,)h(and)f(nothing)g
+-(else.)390 818 y Ff(NAME)570 1004 y Fj(bzip2)p Fl(,)f
+-Fj(bunzip2)g Fl(-)h(a)h(blo)s(c)m(k-sorting)f(\014le)f(compressor,)i
+-(v1.0)570 1136 y Fj(bzcat)e Fl(-)i(decompresses)f(\014les)f(to)i
+-(stdout)570 1267 y Fj(bzip2recover)c Fl(-)k(reco)m(v)m(ers)h(data)f
+-(from)f(damaged)g(bzip2)g(\014les)390 1519 y Ff(SYNOPSIS)570
+-1706 y Fj(bzip2)f Fl([)h(-cdfkqstvzVL123456789)35 b(])c([)g
+-(\014lenames)e(...)41 b(])570 1837 y Fj(bunzip2)28 b
+-Fl([)j(-fkvsVL)f(])h([)f(\014lenames)g(...)41 b(])570
+-1968 y Fj(bzcat)29 b Fl([)h(-s)h(])g([)f(\014lenames)g(...)41
+-b(])570 2100 y Fj(bzip2recover)27 b Fl(\014lename)390
+-2352 y Ff(DESCRIPTION)390 2538 y Fj(bzip2)i Fl(compresses)i(\014les)f
+-(using)f(the)i(Burro)m(ws-Wheeler)g(blo)s(c)m(k)f(sorting)g(text)i
+-(compres-)390 2642 y(sion)40 b(algorithm,)j(and)d(Hu\013man)h(co)s
+-(ding.)71 b(Compression)40 b(is)g(generally)g(considerably)390
+-2746 y(b)s(etter)25 b(than)g(that)h(ac)m(hiev)m(ed)g(b)m(y)f(more)g
+-(con)m(v)m(en)m(tional)h(LZ77/LZ78-based)g(compressors,)390
+-2850 y(and)k(approac)m(hes)h(the)f(p)s(erformance)g(of)h(the)f(PPM)g
+-(family)f(of)i(statistical)f(compressors.)390 3001 y(The)e
+-(command-line)e(options)i(are)h(delib)s(erately)d(v)m(ery)i(similar)e
+-(to)j(those)g(of)f(GNU)h Fj(gzip)p Fl(,)390 3104 y(but)h(they)g(are)h
+-(not)g(iden)m(tical.)390 3255 y Fj(bzip2)f Fl(exp)s(ects)h(a)g(list)f
+-(of)h(\014le)f(names)h(to)h(accompan)m(y)h(the)e(command-line)e
+-(\015ags.)43 b(Eac)m(h)390 3359 y(\014le)e(is)h(replaced)g(b)m(y)g(a)h
+-(compressed)f(v)m(ersion)g(of)g(itself,)i(with)e(the)g(name)g
+-Fj(original_)390 3463 y(name.bz2)p Fl(.)49 b(Eac)m(h)34
+-b(compressed)g(\014le)f(has)g(the)h(same)g(mo)s(di\014cation)e(date,)k
+-(p)s(ermissions,)390 3567 y(and,)24 b(when)f(p)s(ossible,)f(o)m
+-(wnership)f(as)j(the)f(corresp)s(onding)f(original,)h(so)g(that)h
+-(these)g(prop-)390 3671 y(erties)34 b(can)g(b)s(e)f(correctly)i
+-(restored)f(at)g(decompression)f(time.)51 b(File)34 b(name)g(handling)d
+-(is)390 3774 y(naiv)m(e)26 b(in)f(the)i(sense)f(that)h(there)f(is)f(no)
+-i(mec)m(hanism)e(for)h(preserving)f(original)f(\014le)i(names,)390
+-3878 y(p)s(ermissions,)37 b(o)m(wnerships)f(or)h(dates)i(in)d
+-(\014lesystems)h(whic)m(h)g(lac)m(k)h(these)g(concepts,)j(or)390
+-3982 y(ha)m(v)m(e)32 b(serious)d(\014le)g(name)i(length)f
+-(restrictions,)f(suc)m(h)h(as)h(MS-DOS.)390 4133 y Fj(bzip2)26
+-b Fl(and)h Fj(bunzip2)e Fl(will)f(b)m(y)k(default)e(not)i(o)m(v)m
+-(erwrite)g(existing)e(\014les.)38 b(If)27 b(y)m(ou)h(w)m(an)m(t)g(this)
+-390 4237 y(to)j(happ)s(en,)e(sp)s(ecify)g(the)i Fj(-f)e
+-Fl(\015ag.)390 4388 y(If)34 b(no)h(\014le)f(names)g(are)i(sp)s
+-(eci\014ed,)e Fj(bzip2)f Fl(compresses)i(from)f(standard)g(input)f(to)j
+-(stan-)390 4491 y(dard)c(output.)49 b(In)32 b(this)g(case,)k
+-Fj(bzip2)31 b Fl(will)g(decline)h(to)i(write)e(compressed)h(output)g
+-(to)h(a)390 4595 y(terminal,)29 b(as)i(this)e(w)m(ould)g(b)s(e)h(en)m
+-(tirely)f(incomprehensible)e(and)j(therefore)h(p)s(oin)m(tless.)390
+-4746 y Fj(bunzip2)36 b Fl(\(or)j Fj(bzip2)29 b(-d)p Fl(\))37
+-b(decompresses)i(all)e(sp)s(eci\014ed)f(\014les.)63 b(Files)37
+-b(whic)m(h)g(w)m(ere)i(not)390 4850 y(created)e(b)m(y)f
+-Fj(bzip2)f Fl(will)e(b)s(e)i(detected)j(and)d(ignored,)i(and)e(a)i(w)m
+-(arning)d(issued.)56 b Fj(bzip2)390 4954 y Fl(attempts)31
+-b(to)f(guess)g(the)g(\014lename)f(for)h(the)g(decompressed)f(\014le)g
+-(from)h(that)g(of)g(the)g(com-)390 5058 y(pressed)f(\014le)h(as)h
+-(follo)m(ws:)570 5209 y Fj(filename.bz2)57 b Fl(b)s(ecomes)31
+-b Fj(filename)570 5340 y(filename.bz)58 b Fl(b)s(ecomes)30
+-b Fj(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 Fj(bzip2)2375 b Fl(4)570 299 y Fj(filename.tbz2)27
+-b Fl(b)s(ecomes)j Fj(filename.tar)570 470 y(filename.tbz)57
+-b Fl(b)s(ecomes)31 b Fj(filename.tar)570 641 y(anyothername)57
+-b Fl(b)s(ecomes)31 b Fj(anyothername.out)390 859 y Fl(If)j(the)h
+-(\014le)e(do)s(es)i(not)f(end)g(in)f(one)i(of)g(the)g(recognised)f
+-(endings,)g Fj(.bz2)p Fl(,)h Fj(.bz)p Fl(,)g Fj(.tbz2)e
+-Fl(or)390 963 y Fj(.tbz)p Fl(,)h Fj(bzip2)f Fl(complains)f(that)j(it)e
+-(cannot)i(guess)f(the)g(name)h(of)f(the)g(original)e(\014le,)j(and)390
+-1067 y(uses)30 b(the)g(original)f(name)h(with)g Fj(.out)f
+-Fl(app)s(ended.)390 1218 y(As)j(with)f(compression,)h(supplying)c(no)k
+-(\014lenames)f(causes)i(decompression)e(from)h(stan-)390
+-1321 y(dard)d(input)g(to)i(standard)e(output.)390 1472
+-y Fj(bunzip2)k Fl(will)g(correctly)j(decompress)e(a)i(\014le)e(whic)m
+-(h)g(is)h(the)g(concatenation)i(of)e(t)m(w)m(o)i(or)390
+-1576 y(more)j(compressed)f(\014les.)67 b(The)39 b(result)g(is)g(the)g
+-(concatenation)i(of)f(the)g(corresp)s(onding)390 1680
+-y(uncompressed)c(\014les.)59 b(In)m(tegrit)m(y)38 b(testing)f(\()p
+-Fj(-t)p Fl(\))g(of)g(concatenated)i(compressed)e(\014les)f(is)390
+-1784 y(also)30 b(supp)s(orted.)390 1935 y(Y)-8 b(ou)40
+-b(can)g(also)f(compress)g(or)g(decompress)g(\014les)g(to)h(the)f
+-(standard)g(output)g(b)m(y)g(giving)390 2039 y(the)30
+-b Fj(-c)g Fl(\015ag.)40 b(Multiple)28 b(\014les)h(ma)m(y)i(b)s(e)e
+-(compressed)h(and)f(decompressed)h(lik)m(e)f(this.)39
+-b(The)390 2142 y(resulting)31 b(outputs)i(are)h(fed)f(sequen)m(tially)f
+-(to)i(stdout.)49 b(Compression)32 b(of)h(m)m(ultiple)e(\014les)390
+-2246 y(in)24 b(this)g(manner)h(generates)h(a)g(stream)f(con)m(taining)g
+-(m)m(ultiple)e(compressed)i(\014le)f(represen-)390 2350
+-y(tations.)58 b(Suc)m(h)36 b(a)g(stream)g(can)h(b)s(e)e(decompressed)h
+-(correctly)h(only)e(b)m(y)h Fj(bzip2)e Fl(v)m(ersion)390
+-2454 y(0.9.0)g(or)e(later.)47 b(Earlier)30 b(v)m(ersions)i(of)g
+-Fj(bzip2)f Fl(will)f(stop)i(after)h(decompressing)e(the)i(\014rst)390
+-2558 y(\014le)c(in)h(the)g(stream.)390 2709 y Fj(bzcat)f
+-Fl(\(or)i Fj(bzip2)e(-dc)p Fl(\))g(decompresses)i(all)e(sp)s(eci\014ed)
+-g(\014les)g(to)i(the)g(standard)e(output.)390 2860 y
+-Fj(bzip2)f Fl(will)g(read)i(argumen)m(ts)g(from)f(the)h(en)m(vironmen)m
+-(t)g(v)-5 b(ariables)28 b Fj(BZIP2)h Fl(and)g Fj(BZIP)p
+-Fl(,)g(in)390 2963 y(that)24 b(order,)g(and)f(will)e(pro)s(cess)i(them)
+-g(b)s(efore)g(an)m(y)h(argumen)m(ts)f(read)h(from)f(the)g(command)390
+-3067 y(line.)39 b(This)29 b(giv)m(es)h(a)h(con)m(v)m(enien)m(t)h(w)m(a)
+-m(y)f(to)g(supply)d(default)i(argumen)m(ts.)390 3218
+-y(Compression)h(is)h(alw)m(a)m(ys)i(p)s(erformed,)e(ev)m(en)h(if)f(the)
+-h(compressed)g(\014le)f(is)g(sligh)m(tly)f(larger)390
+-3322 y(than)26 b(the)g(original.)38 b(Files)25 b(of)h(less)g(than)g(ab)
+-s(out)g(one)g(h)m(undred)e(b)m(ytes)j(tend)f(to)h(get)g(larger,)390
+-3426 y(since)34 b(the)g(compression)f(mec)m(hanism)h(has)f(a)i(constan)
+-m(t)g(o)m(v)m(erhead)h(in)d(the)h(region)g(of)g(50)390
+-3529 y(b)m(ytes.)54 b(Random)34 b(data)h(\(including)d(the)i(output)h
+-(of)f(most)h(\014le)f(compressors\))h(is)e(co)s(ded)390
+-3633 y(at)e(ab)s(out)f(8.05)i(bits)d(p)s(er)h(b)m(yte,)h(giving)e(an)h
+-(expansion)g(of)g(around)g(0.5\045.)390 3784 y(As)h(a)g(self-c)m(hec)m
+-(k)h(for)e(y)m(our)h(protection,)g Fj(bzip2)f Fl(uses)g(32-bit)h(CR)m
+-(Cs)f(to)i(mak)m(e)f(sure)f(that)390 3888 y(the)45 b(decompressed)f(v)m
+-(ersion)g(of)g(a)h(\014le)e(is)h(iden)m(tical)f(to)i(the)g(original.)81
+-b(This)43 b(guards)390 3992 y(against)i(corruption)e(of)h(the)h
+-(compressed)f(data,)49 b(and)44 b(against)h(undetected)g(bugs)e(in)390
+-4096 y Fj(bzip2)35 b Fl(\(hop)s(efully)e(v)m(ery)k(unlik)m(ely\).)56
+-b(The)36 b(c)m(hances)h(of)f(data)h(corruption)e(going)h(unde-)390
+-4199 y(tected)g(is)e(microscopic,)h(ab)s(out)f(one)h(c)m(hance)g(in)f
+-(four)g(billion)d(for)j(eac)m(h)i(\014le)d(pro)s(cessed.)390
+-4303 y(Be)38 b(a)m(w)m(are,)k(though,)d(that)f(the)g(c)m(hec)m(k)i(o)s
+-(ccurs)d(up)s(on)f(decompression,)j(so)f(it)f(can)h(only)390
+-4407 y(tell)28 b(y)m(ou)g(that)i(something)d(is)h(wrong.)40
+-b(It)28 b(can't)i(help)d(y)m(ou)i(reco)m(v)m(er)h(the)e(original)f
+-(uncom-)390 4511 y(pressed)h(data.)41 b(Y)-8 b(ou)30
+-b(can)f(use)g Fj(bzip2recover)d Fl(to)k(try)f(to)h(reco)m(v)m(er)h
+-(data)f(from)e(damaged)390 4614 y(\014les.)390 4766 y(Return)22
+-b(v)-5 b(alues:)37 b(0)23 b(for)g(a)g(normal)f(exit,)j(1)e(for)g(en)m
+-(vironmen)m(tal)f(problems)f(\(\014le)i(not)g(found,)390
+-4869 y(in)m(v)-5 b(alid)30 b(\015ags,)k(I/O)f(errors,)g(&c\),)h(2)f(to)
+-g(indicate)f(a)h(corrupt)f(compressed)h(\014le,)f(3)i(for)e(an)390
+-4973 y(in)m(ternal)d(consistency)h(error)g(\(eg,)i(bug\))e(whic)m(h)f
+-(caused)i Fj(bzip2)e Fl(to)i(panic.)390 5304 y Ff(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 Fj(bzip2)2375 b Fl(5)390 299 y Fj(-c)30
+-b(--stdout)870 403 y Fl(Compress)f(or)i(decompress)f(to)h(standard)e
+-(output.)390 557 y Fj(-d)h(--decompress)870 661 y Fl(F)-8
+-b(orce)44 b(decompression.)77 b Fj(bzip2)p Fl(,)44 b
+-Fj(bunzip2)d Fl(and)h Fj(bzcat)f Fl(are)i(really)f(the)870
+-764 y(same)27 b(program,)h(and)e(the)i(decision)d(ab)s(out)i(what)g
+-(actions)g(to)h(tak)m(e)g(is)e(done)870 868 y(on)k(the)h(basis)e(of)i
+-(whic)m(h)e(name)h(is)g(used.)40 b(This)28 b(\015ag)j(o)m(v)m(errides)f
+-(that)h(mec)m(h-)870 972 y(anism,)e(and)h(forces)h(bzip2)e(to)i
+-(decompress.)390 1126 y Fj(-z)f(--compress)870 1230 y
+-Fl(The)39 b(complemen)m(t)h(to)g Fj(-d)p Fl(:)59 b(forces)40
+-b(compression,)h(regardless)d(of)i(the)g(in-)870 1334
+-y(v)m(ok)-5 b(ation)31 b(name.)390 1488 y Fj(-t)f(--test)8
+-b Fl(Chec)m(k)33 b(in)m(tegrit)m(y)j(of)f(the)g(sp)s(eci\014ed)e
+-(\014le\(s\),)k(but)d(don't)h(decompress)g(them.)870
+-1591 y(This)40 b(really)g(p)s(erforms)g(a)i(trial)e(decompression)h
+-(and)g(thro)m(ws)g(a)m(w)m(a)m(y)j(the)870 1695 y(result.)390
+-1849 y Fj(-f)30 b(--force)870 1953 y Fl(F)-8 b(orce)31
+-b(o)m(v)m(erwrite)f(of)g(output)f(\014les.)40 b(Normally)-8
+-b(,)29 b Fj(bzip2)f Fl(will)f(not)j(o)m(v)m(erwrite)870
+-2057 y(existing)e(output)g(\014les.)39 b(Also)28 b(forces)h
+-Fj(bzip2)e Fl(to)i(break)g(hard)e(links)f(to)k(\014les,)870
+-2161 y(whic)m(h)f(it)h(otherwise)g(w)m(ouldn't)f(do.)390
+-2315 y Fj(-k)h(--keep)8 b Fl(Keep)24 b(\(don't)i(delete\))h(input)d
+-(\014les)g(during)g(compression)h(or)h(decompression.)390
+-2469 y Fj(-s)k(--small)870 2573 y Fl(Reduce)23 b(memory)f(usage,)j(for)
+-d(compression,)h(decompression)f(and)g(testing.)870 2676
+-y(Files)f(are)i(decompressed)e(and)h(tested)h(using)e(a)h(mo)s
+-(di\014ed)e(algorithm)h(whic)m(h)870 2780 y(only)30 b(requires)g(2.5)j
+-(b)m(ytes)f(p)s(er)e(blo)s(c)m(k)h(b)m(yte.)44 b(This)30
+-b(means)h(an)m(y)h(\014le)e(can)i(b)s(e)870 2884 y(decompressed)d(in)f
+-(2300k)j(of)e(memory)-8 b(,)30 b(alb)s(eit)e(at)i(ab)s(out)f(half)g
+-(the)g(normal)870 2988 y(sp)s(eed.)870 3117 y(During)42
+-b(compression,)k Fj(-s)d Fl(selects)h(a)g(blo)s(c)m(k)g(size)f(of)h
+-(200k,)k(whic)m(h)42 b(lim-)870 3220 y(its)33 b(memory)g(use)g(to)h
+-(around)e(the)i(same)f(\014gure,)h(at)g(the)g(exp)s(ense)f(of)g(y)m
+-(our)870 3324 y(compression)g(ratio.)50 b(In)33 b(short,)i(if)d(y)m
+-(our)i(mac)m(hine)f(is)g(lo)m(w)g(on)h(memory)f(\(8)870
+-3428 y(megab)m(ytes)42 b(or)e(less\),)j(use)d(-s)g(for)g(ev)m
+-(erything.)71 b(See)40 b(MEMOR)-8 b(Y)41 b(MAN-)870 3532
+-y(A)m(GEMENT)31 b(b)s(elo)m(w.)390 3686 y Fj(-q)f(--quiet)870
+-3790 y Fl(Suppress)j(non-essen)m(tial)j(w)m(arning)e(messages.)58
+-b(Messages)38 b(p)s(ertaining)33 b(to)870 3893 y(I/O)d(errors)g(and)g
+-(other)h(critical)e(ev)m(en)m(ts)j(will)27 b(not)k(b)s(e)f(suppressed.)
+-390 4047 y Fj(-v)g(--verbose)870 4151 y Fl(V)-8 b(erb)s(ose)28
+-b(mo)s(de)f({)i(sho)m(w)e(the)h(compression)f(ratio)h(for)f(eac)m(h)i
+-(\014le)e(pro)s(cessed.)870 4255 y(F)-8 b(urther)30 b
+-Fj(-v)p Fl('s)g(increase)g(the)g(v)m(erb)s(osit)m(y)g(lev)m(el,)h(sp)s
+-(ewing)d(out)j(lots)f(of)g(infor-)870 4359 y(mation)g(whic)m(h)f(is)h
+-(primarily)d(of)j(in)m(terest)h(for)f(diagnostic)g(purp)s(oses.)390
+-4513 y Fj(-L)g(--license)e(-V)h(--version)870 4617 y
+-Fl(Displa)m(y)h(the)g(soft)m(w)m(are)i(v)m(ersion,)e(license)f(terms)i
+-(and)e(conditions.)390 4771 y Fj(-1)h(to)g(-9)72 b Fl(Set)35
+-b(the)g(blo)s(c)m(k)f(size)h(to)g(100)h(k,)g(200)g(k)f(..)53
+-b(900)36 b(k)f(when)f(compressing.)53 b(Has)870 4875
+-y(no)41 b(e\013ect)h(when)d(decompressing.)71 b(See)41
+-b(MEMOR)-8 b(Y)41 b(MANA)m(GEMENT)870 4978 y(b)s(elo)m(w.)390
+-5132 y Fj(--)324 b Fl(T)-8 b(reats)25 b(all)e(subsequen)m(t)g(argumen)m
+-(ts)i(as)f(\014le)g(names,)h(ev)m(en)g(if)e(they)i(start)f(with)870
+-5236 y(a)32 b(dash.)43 b(This)29 b(is)h(so)i(y)m(ou)g(can)f(handle)f
+-(\014les)g(with)g(names)i(b)s(eginning)c(with)870 5340
+-y(a)j(dash,)f(for)g(example:)40 b Fj(bzip2)29 b(--)h(-myfilename)p
+-Fl(.)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 Fj(bzip2)2375 b Fl(6)390 299 y Fj(--repetitive-fast)390
+-427 y(--repetitive-best)870 530 y Fl(These)34 b(\015ags)g(are)h
+-(redundan)m(t)e(in)g(v)m(ersions)g(0.9.5)j(and)e(ab)s(o)m(v)m(e.)53
+-b(They)34 b(pro-)870 634 y(vided)h(some)i(coarse)g(con)m(trol)g(o)m(v)m
+-(er)g(the)g(b)s(eha)m(viour)e(of)h(the)g(sorting)g(algo-)870
+-738 y(rithm)h(in)h(earlier)g(v)m(ersions,)j(whic)m(h)d(w)m(as)h
+-(sometimes)h(useful.)65 b(0.9.5)41 b(and)870 842 y(ab)s(o)m(v)m(e)34
+-b(ha)m(v)m(e)g(an)f(impro)m(v)m(ed)g(algorithm)f(whic)m(h)f(renders)h
+-(these)h(\015ags)h(irrel-)870 946 y(ev)-5 b(an)m(t.)390
+-1190 y Ff(MEMOR)-10 b(Y)40 b(MANA)m(GEMENT)390 1377 y
+-Fj(bzip2)25 b Fl(compresses)i(large)g(\014les)e(in)g(blo)s(c)m(ks.)39
+-b(The)26 b(blo)s(c)m(k)h(size)f(a\013ects)i(b)s(oth)e(the)h(compres-)
+-390 1481 y(sion)39 b(ratio)g(ac)m(hiev)m(ed,)k(and)d(the)f(amoun)m(t)i
+-(of)e(memory)h(needed)f(for)h(compression)f(and)390 1585
+-y(decompression.)59 b(The)36 b(\015ags)h Fj(-1)f Fl(through)h
+-Fj(-9)f Fl(sp)s(ecify)f(the)i(blo)s(c)m(k)g(size)f(to)i(b)s(e)e
+-(100,000)390 1688 y(b)m(ytes)29 b(through)e(900,000)k(b)m(ytes)d(\(the)
+-h(default\))e(resp)s(ectiv)m(ely)-8 b(.)40 b(A)m(t)29
+-b(decompression)e(time,)390 1792 y(the)32 b(blo)s(c)m(k)g(size)g(used)g
+-(for)g(compression)f(is)g(read)h(from)g(the)g(header)g(of)h(the)f
+-(compressed)390 1896 y(\014le,)j(and)f Fj(bunzip2)e Fl(then)i(allo)s
+-(cates)h(itself)e(just)h(enough)g(memory)g(to)i(decompress)e(the)390
+-2000 y(\014le.)39 b(Since)29 b(blo)s(c)m(k)g(sizes)g(are)h(stored)f(in)
+-f(compressed)h(\014les,)g(it)g(follo)m(ws)f(that)i(the)g(\015ags)g
+-Fj(-1)390 2103 y Fl(to)h Fj(-9)f Fl(are)h(irrelev)-5
+-b(an)m(t)29 b(to)i(and)f(so)h(ignored)e(during)f(decompression.)390
+-2255 y(Compression)h(and)g(decompression)h(requiremen)m(ts,)f(in)g(b)m
+-(ytes,)j(can)e(b)s(e)g(estimated)h(as:)869 2406 y Fj(Compression:)140
+-b(400k)46 b(+)i(\()f(8)h(x)f(block)f(size)h(\))869 2613
+-y(Decompression:)d(100k)i(+)i(\()f(4)h(x)f(block)f(size)h(\),)g(or)1585
+-2717 y(100k)f(+)i(\()f(2.5)g(x)g(block)g(size)f(\))390
+-2868 y Fl(Larger)29 b(blo)s(c)m(k)f(sizes)h(giv)m(e)g(rapidly)d
+-(diminishing)e(marginal)k(returns.)39 b(Most)29 b(of)g(the)g(com-)390
+-2972 y(pression)d(comes)j(from)f(the)g(\014rst)g(t)m(w)m(o)h(or)f
+-(three)h(h)m(undred)d(k)i(of)g(blo)s(c)m(k)g(size,)g(a)h(fact)g(w)m
+-(orth)390 3075 y(b)s(earing)j(in)f(mind)g(when)h(using)f
+-Fj(bzip2)h Fl(on)g(small)g(mac)m(hines.)47 b(It)33 b(is)f(also)h(imp)s
+-(ortan)m(t)f(to)390 3179 y(appreciate)j(that)h(the)f(decompression)f
+-(memory)h(requiremen)m(t)f(is)h(set)g(at)h(compression)390
+-3283 y(time)30 b(b)m(y)g(the)h(c)m(hoice)g(of)g(blo)s(c)m(k)f(size.)390
+-3434 y(F)-8 b(or)45 b(\014les)f(compressed)g(with)g(the)g(default)g
+-(900k)i(blo)s(c)m(k)e(size,)49 b Fj(bunzip2)42 b Fl(will)g(require)390
+-3538 y(ab)s(out)29 b(3700)j(kb)m(ytes)e(to)h(decompress.)40
+-b(T)-8 b(o)30 b(supp)s(ort)e(decompression)h(of)h(an)m(y)g(\014le)f(on)
+-g(a)i(4)390 3642 y(megab)m(yte)h(mac)m(hine,)d Fj(bunzip2)f
+-Fl(has)i(an)g(option)f(to)i(decompress)e(using)g(appro)m(ximately)390
+-3745 y(half)k(this)g(amoun)m(t)i(of)f(memory)-8 b(,)36
+-b(ab)s(out)e(2300)i(kb)m(ytes.)53 b(Decompression)34
+-b(sp)s(eed)g(is)f(also)390 3849 y(halv)m(ed,)i(so)f(y)m(ou)h(should)d
+-(use)h(this)g(option)h(only)f(where)h(necessary)-8 b(.)53
+-b(The)33 b(relev)-5 b(an)m(t)35 b(\015ag)390 3953 y(is)29
+-b Fj(-s)p Fl(.)390 4104 y(In)34 b(general,)i(try)f(and)f(use)g(the)h
+-(largest)h(blo)s(c)m(k)e(size)h(memory)f(constrain)m(ts)h(allo)m(w,)h
+-(since)390 4208 y(that)45 b(maximises)f(the)h(compression)f(ac)m(hiev)m
+-(ed.)85 b(Compression)43 b(and)h(decompression)390 4311
+-y(sp)s(eed)30 b(are)g(virtually)e(una\013ected)j(b)m(y)f(blo)s(c)m(k)g
+-(size.)390 4463 y(Another)25 b(signi\014can)m(t)f(p)s(oin)m(t)g
+-(applies)f(to)j(\014les)e(whic)m(h)g(\014t)h(in)e(a)j(single)d(blo)s(c)
+-m(k)i({)g(that)h(means)390 4566 y(most)g(\014les)g(y)m(ou'd)g(encoun)m
+-(ter)h(using)d(a)j(large)f(blo)s(c)m(k)g(size.)39 b(The)25
+-b(amoun)m(t)i(of)f(real)g(memory)390 4670 y(touc)m(hed)38
+-b(is)e(prop)s(ortional)f(to)j(the)f(size)g(of)h(the)f(\014le,)h(since)f
+-(the)g(\014le)g(is)f(smaller)g(than)h(a)390 4774 y(blo)s(c)m(k.)49
+-b(F)-8 b(or)35 b(example,)f(compressing)e(a)i(\014le)e(20,000)k(b)m
+-(ytes)e(long)f(with)f(the)i(\015ag)g Fj(-9)f Fl(will)390
+-4878 y(cause)28 b(the)f(compressor)g(to)h(allo)s(cate)f(around)f(7600k)
+-j(of)e(memory)-8 b(,)28 b(but)f(only)f(touc)m(h)i(400k)390
+-4981 y Fj(+)h Fl(20000)j(*)e(8)g(=)f(560)i(kb)m(ytes)f(of)g(it.)40
+-b(Similarly)-8 b(,)26 b(the)k(decompressor)f(will)e(allo)s(cate)j
+-(3700k)390 5085 y(but)g(only)f(touc)m(h)i(100k)h Fj(+)e
+-Fl(20000)i(*)f(4)g(=)f(180)i(kb)m(ytes.)390 5236 y(Here)41
+-b(is)f(a)i(table)f(whic)m(h)e(summarises)g(the)j(maxim)m(um)d(memory)i
+-(usage)h(for)e(di\013eren)m(t)390 5340 y(blo)s(c)m(k)25
+-b(sizes.)38 b(Also)25 b(recorded)g(is)f(the)i(total)g(compressed)e
+-(size)h(for)g(14)h(\014les)e(of)i(the)f(Calgary)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 Fj(bzip2)2375 b Fl(7)390 299 y(T)-8 b(ext)38
+-b(Compression)d(Corpus)h(totalling)h(3,141,622)k(b)m(ytes.)61
+-b(This)36 b(column)g(giv)m(es)i(some)390 403 y(feel)23
+-b(for)f(ho)m(w)h(compression)f(v)-5 b(aries)23 b(with)e(blo)s(c)m(k)i
+-(size.)38 b(These)23 b(\014gures)f(tend)g(to)i(understate)390
+-506 y(the)g(adv)-5 b(an)m(tage)26 b(of)e(larger)f(blo)s(c)m(k)h(sizes)f
+-(for)h(larger)f(\014les,)h(since)g(the)g(Corpus)e(is)h(dominated)390
+-610 y(b)m(y)30 b(smaller)f(\014les.)1107 761 y Fj(Compress)141
+-b(Decompress)g(Decompress)f(Corpus)773 865 y(Flag)238
+-b(usage)285 b(usage)332 b(-s)48 b(usage)237 b(Size)821
+-1073 y(-1)286 b(1200k)332 b(500k)429 b(350k)285 b(914704)821
+-1176 y(-2)h(2000k)332 b(900k)429 b(600k)285 b(877703)821
+-1280 y(-3)h(2800k)f(1300k)428 b(850k)285 b(860338)821
+-1384 y(-4)h(3600k)f(1700k)380 b(1100k)285 b(846899)821
+-1488 y(-5)h(4400k)f(2100k)380 b(1350k)285 b(845160)821
+-1591 y(-6)h(5200k)f(2500k)380 b(1600k)285 b(838626)821
+-1695 y(-7)h(6100k)f(2900k)380 b(1850k)285 b(834096)821
+-1799 y(-8)h(6800k)f(3300k)380 b(2100k)285 b(828642)821
+-1903 y(-9)h(7600k)f(3700k)380 b(2350k)285 b(828642)390
+-2147 y Ff(RECO)m(VERING)37 b(D)m(A)-10 b(T)g(A)40 b(FR)m(OM)h(D)m(AMA)m
+-(GED)e(FILES)390 2333 y Fj(bzip2)25 b Fl(compresses)h(\014les)g(in)f
+-(blo)s(c)m(ks,)h(usually)e(900kb)m(ytes)29 b(long.)39
+-b(Eac)m(h)27 b(blo)s(c)m(k)e(is)h(handled)390 2437 y(indep)s(enden)m
+-(tly)-8 b(.)47 b(If)32 b(a)i(media)e(or)h(transmission)e(error)i
+-(causes)h(a)f(m)m(ulti-blo)s(c)m(k)f Fj(.bz2)g Fl(\014le)390
+-2541 y(to)k(b)s(ecome)h(damaged,)g(it)e(ma)m(y)i(b)s(e)e(p)s(ossible)e
+-(to)k(reco)m(v)m(er)g(data)f(from)g(the)f(undamaged)390
+-2645 y(blo)s(c)m(ks)30 b(in)f(the)h(\014le.)390 2796
+-y(The)j(compressed)h(represen)m(tation)f(of)h(eac)m(h)h(blo)s(c)m(k)e
+-(is)g(delimited)e(b)m(y)j(a)g(48-bit)g(pattern,)390 2900
+-y(whic)m(h)27 b(mak)m(es)j(it)e(p)s(ossible)e(to)j(\014nd)e(the)i(blo)s
+-(c)m(k)f(b)s(oundaries)e(with)i(reasonable)g(certain)m(t)m(y)-8
+-b(.)390 3003 y(Eac)m(h)34 b(blo)s(c)m(k)f(also)g(carries)g(its)g(o)m
+-(wn)g(32-bit)g(CR)m(C,)h(so)f(damaged)h(blo)s(c)m(ks)f(can)g(b)s(e)g
+-(distin-)390 3107 y(guished)c(from)h(undamaged)g(ones.)390
+-3258 y Fj(bzip2recover)37 b Fl(is)j(a)h(simple)e(program)h(whose)g
+-(purp)s(ose)f(is)h(to)i(searc)m(h)f(for)f(blo)s(c)m(ks)g(in)390
+-3362 y Fj(.bz2)34 b Fl(\014les,)i(and)f(write)f(eac)m(h)j(blo)s(c)m(k)d
+-(out)i(in)m(to)f(its)g(o)m(wn)g Fj(.bz2)f Fl(\014le.)55
+-b(Y)-8 b(ou)36 b(can)f(then)g(use)390 3466 y Fj(bzip2)29
+-b(-t)c Fl(to)i(test)f(the)g(in)m(tegrit)m(y)g(of)g(the)g(resulting)e
+-(\014les,)i(and)f(decompress)h(those)g(whic)m(h)390 3569
+-y(are)31 b(undamaged.)390 3721 y Fj(bzip2recover)41 b
+-Fl(tak)m(es)46 b(a)f(single)e(argumen)m(t,)49 b(the)44
+-b(name)h(of)g(the)f(damaged)h(\014le,)j(and)390 3824
+-y(writes)33 b(a)i(n)m(um)m(b)s(er)d(of)j(\014les)e Fj(rec0001file.bz2)p
+-Fl(,)e Fj(rec0002file.bz2)p Fl(,)g(etc,)36 b(con)m(taining)390
+-3928 y(the)42 b(extracted)g(blo)s(c)m(ks.)74 b(The)41
+-b(output)g(\014lenames)f(are)i(designed)e(so)i(that)g(the)g(use)f(of)
+-390 4032 y(wildcards)30 b(in)h(subsequen)m(t)h(pro)s(cessing)f({)i(for)
+-g(example,)g Fj(bzip2)c(-dc)g(rec*file.bz2)e(>)390 4136
+-y(recovered_data)f Fl({)31 b(lists)e(the)i(\014les)e(in)g(the)i
+-(correct)g(order.)390 4287 y Fj(bzip2recover)38 b Fl(should)i(b)s(e)g
+-(of)i(most)g(use)f(dealing)f(with)g(large)i Fj(.bz2)e
+-Fl(\014les,)k(as)d(these)390 4390 y(will)29 b(con)m(tain)j(man)m(y)g
+-(blo)s(c)m(ks.)45 b(It)32 b(is)f(clearly)g(futile)f(to)i(use)g(it)f(on)
+-h(damaged)g(single-blo)s(c)m(k)390 4494 y(\014les,)g(since)f(a)h
+-(damaged)h(blo)s(c)m(k)e(cannot)i(b)s(e)e(reco)m(v)m(ered.)47
+-b(If)32 b(y)m(ou)g(wish)e(to)j(minimise)c(an)m(y)390
+-4598 y(p)s(oten)m(tial)36 b(data)i(loss)e(through)g(media)h(or)f
+-(transmission)f(errors,)j(y)m(ou)f(migh)m(t)g(consider)390
+-4702 y(compressing)29 b(with)g(a)i(smaller)e(blo)s(c)m(k)h(size.)390
+-4946 y Ff(PERF)m(ORMANCE)39 b(NOTES)390 5132 y Fl(The)f(sorting)f
+-(phase)h(of)h(compression)e(gathers)i(together)h(similar)35
+-b(strings)i(in)g(the)i(\014le.)390 5236 y(Because)54
+-b(of)f(this,)58 b(\014les)52 b(con)m(taining)g(v)m(ery)h(long)g(runs)e
+-(of)i(rep)s(eated)g(sym)m(b)s(ols,)58 b(lik)m(e)390 5340
+-y Fj(")p Fl(aabaabaabaab)e(...)p Fj(")g Fl(\(rep)s(eated)g(sev)m(eral)f
+-(h)m(undred)e(times\))i(ma)m(y)h(compress)f(more)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 Fj(bzip2)2375 b Fl(8)390 299 y(slo)m(wly)33
+-b(than)g(normal.)50 b(V)-8 b(ersions)33 b(0.9.5)i(and)f(ab)s(o)m(v)m(e)
+-h(fare)e(m)m(uc)m(h)h(b)s(etter)g(than)f(previous)390
+-403 y(v)m(ersions)i(in)g(this)f(resp)s(ect.)57 b(The)35
+-b(ratio)h(b)s(et)m(w)m(een)h(w)m(orst-case)g(and)e(a)m(v)m(erage-case)
+-40 b(com-)390 506 y(pression)e(time)h(is)f(in)g(the)h(region)g(of)h
+-(10:1.)69 b(F)-8 b(or)40 b(previous)e(v)m(ersions,)j(this)d(\014gure)h
+-(w)m(as)390 610 y(more)f(lik)m(e)g(100:1.)66 b(Y)-8 b(ou)38
+-b(can)h(use)e(the)i Fj(-vvvv)d Fl(option)i(to)h(monitor)e(progress)h
+-(in)f(great)390 714 y(detail,)30 b(if)f(y)m(ou)i(w)m(an)m(t.)390
+-865 y(Decompression)f(sp)s(eed)g(is)f(una\013ected)i(b)m(y)f(these)h
+-(phenomena.)390 1016 y Fj(bzip2)i Fl(usually)g(allo)s(cates)i(sev)m
+-(eral)f(megab)m(ytes)j(of)d(memory)h(to)g(op)s(erate)h(in,)e(and)g
+-(then)390 1120 y(c)m(harges)k(all)d(o)m(v)m(er)j(it)f(in)e(a)i(fairly)e
+-(random)h(fashion.)59 b(This)34 b(means)j(that)g(p)s(erformance,)390
+-1224 y(b)s(oth)27 b(for)h(compressing)f(and)g(decompressing,)h(is)f
+-(largely)g(determined)g(b)m(y)h(the)g(sp)s(eed)f(at)390
+-1327 y(whic)m(h)35 b(y)m(our)h(mac)m(hine)g(can)g(service)g(cac)m(he)i
+-(misses.)57 b(Because)37 b(of)g(this,)f(small)f(c)m(hanges)390
+-1431 y(to)f(the)f(co)s(de)h(to)f(reduce)g(the)h(miss)d(rate)j(ha)m(v)m
+-(e)h(b)s(een)d(observ)m(ed)h(to)h(giv)m(e)g(disprop)s(ortion-)390
+-1535 y(ately)i(large)f(p)s(erformance)f(impro)m(v)m(emen)m(ts.)56
+-b(I)35 b(imagine)f Fj(bzip2)g Fl(will)e(p)s(erform)i(b)s(est)h(on)390
+-1639 y(mac)m(hines)30 b(with)f(v)m(ery)i(large)f(cac)m(hes.)390
+-1885 y Ff(CA)-14 b(VEA)k(TS)390 2072 y Fl(I/O)38 b(error)g(messages)h
+-(are)f(not)h(as)f(helpful)e(as)i(they)g(could)f(b)s(e.)64
+-b Fj(bzip2)37 b Fl(tries)g(hard)g(to)390 2176 y(detect)29
+-b(I/O)e(errors)g(and)f(exit)i(cleanly)-8 b(,)27 b(but)g(the)h(details)e
+-(of)h(what)h(the)f(problem)f(is)g(some-)390 2280 y(times)k(seem)h
+-(rather)f(misleading.)390 2431 y(This)j(man)m(ual)g(page)i(p)s(ertains)
+-e(to)i(v)m(ersion)f(1.0)i(of)e Fj(bzip2)p Fl(.)51 b(Compressed)34
+-b(data)h(created)390 2534 y(b)m(y)25 b(this)e(v)m(ersion)i(is)e(en)m
+-(tirely)h(forw)m(ards)h(and)f(bac)m(kw)m(ards)h(compatible)f(with)f
+-(the)i(previous)390 2638 y(public)18 b(releases,)24 b(v)m(ersions)c
+-(0.1pl2,)k(0.9.0)e(and)f(0.9.5,)k(but)20 b(with)g(the)h(follo)m(wing)e
+-(exception:)390 2742 y(0.9.0)43 b(and)e(ab)s(o)m(v)m(e)h(can)g
+-(correctly)f(decompress)g(m)m(ultiple)e(concatenated)k(compressed)390
+-2846 y(\014les.)c(0.1pl2)30 b(cannot)g(do)f(this;)f(it)h(will)e(stop)i
+-(after)h(decompressing)e(just)g(the)i(\014rst)e(\014le)g(in)390
+-2949 y(the)j(stream.)390 3100 y Fj(bzip2recover)20 b
+-Fl(uses)k(32-bit)g(in)m(tegers)f(to)i(represen)m(t)f(bit)e(p)s
+-(ositions)g(in)g(compressed)i(\014les,)390 3204 y(so)j(it)f(cannot)i
+-(handle)d(compressed)i(\014les)f(more)h(than)f(512)i(megab)m(ytes)h
+-(long.)39 b(This)25 b(could)390 3308 y(easily)30 b(b)s(e)f(\014xed.)390
+-3555 y Ff(A)m(UTHOR)390 3741 y Fl(Julian)f(Sew)m(ard,)i
+-Fj(jseward@acm.org)p Fl(.)390 3892 y(The)24 b(ideas)f(em)m(b)s(o)s
+-(died)f(in)h Fj(bzip2)f Fl(are)j(due)e(to)i(\(at)g(least\))g(the)f
+-(follo)m(wing)e(p)s(eople:)37 b(Mic)m(hael)390 3996 y(Burro)m(ws)48
+-b(and)g(Da)m(vid)h(Wheeler)f(\(for)h(the)g(blo)s(c)m(k)f(sorting)g
+-(transformation\),)53 b(Da)m(vid)390 4100 y(Wheeler)45
+-b(\(again,)50 b(for)45 b(the)g(Hu\013man)g(co)s(der\),)k(P)m(eter)d(F)
+--8 b(en)m(wic)m(k)46 b(\(for)g(the)f(structured)390 4204
+-y(co)s(ding)26 b(mo)s(del)g(in)f(the)i(original)e Fj(bzip)p
+-Fl(,)i(and)f(man)m(y)h(re\014nemen)m(ts\),)h(and)e(Alistair)f
+-(Mo\013at,)390 4307 y(Radford)34 b(Neal)h(and)f(Ian)h(Witten)g(\(for)f
+-(the)h(arithmetic)g(co)s(der)f(in)g(the)h(original)d
+-Fj(bzip)p Fl(\).)390 4411 y(I)41 b(am)g(m)m(uc)m(h)h(indebted)e(for)h
+-(their)f(help,)j(supp)s(ort)c(and)i(advice.)74 b(See)41
+-b(the)h(man)m(ual)e(in)390 4515 y(the)28 b(source)g(distribution)23
+-b(for)28 b(p)s(oin)m(ters)e(to)j(sources)e(of)h(do)s(cumen)m(tation.)40
+-b(Christian)25 b(v)m(on)390 4619 y(Ro)s(ques)31 b(encouraged)h(me)g(to)
+-g(lo)s(ok)f(for)h(faster)g(sorting)f(algorithms,)f(so)i(as)g(to)g(sp)s
+-(eed)f(up)390 4723 y(compression.)47 b(Bela)34 b(Lubkin)c(encouraged)k
+-(me)f(to)g(impro)m(v)m(e)g(the)g(w)m(orst-case)i(compres-)390
+-4826 y(sion)25 b(p)s(erformance.)38 b(Man)m(y)26 b(p)s(eople)f(sen)m(t)
+-h(patc)m(hes,)h(help)s(ed)d(with)g(p)s(ortabilit)m(y)f(problems,)390
+-4930 y(len)m(t)30 b(mac)m(hines,)g(ga)m(v)m(e)j(advice)d(and)g(w)m(ere)
+-h(generally)f(helpful.)p eop
+-%%Page: 9 10
+-9 9 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1927 b Fl(9)150 299 y Fh(3)80 b(Programming)53
+-b(with)h Fg(libbzip2)150 568 y Fl(This)29 b(c)m(hapter)i(describ)s(es)d
+-(the)j(programming)e(in)m(terface)i(to)g Fj(libbzip2)p
+-Fl(.)150 725 y(F)-8 b(or)36 b(general)e(bac)m(kground)h(information,)f
+-(particularly)f(ab)s(out)h(memory)h(use)f(and)g(p)s(erformance)g(as-)
+-150 834 y(p)s(ects,)d(y)m(ou'd)f(b)s(e)g(w)m(ell)f(advised)g(to)j(read)
+-e(Chapter)g(2)g(as)h(w)m(ell.)150 1124 y Fk(3.1)68 b(T)-11
+-b(op-lev)l(el)46 b(structure)150 1316 y Fj(libbzip2)33
+-b Fl(is)i(a)h(\015exible)e(library)f(for)j(compressing)f(and)g
+-(decompressing)f(data)j(in)d(the)i Fj(bzip2)e Fl(data)150
+-1426 y(format.)39 b(Although)24 b(pac)m(k)-5 b(aged)26
+-b(as)e(a)h(single)e(en)m(tit)m(y)-8 b(,)27 b(it)d(helps)f(to)i(regard)g
+-(the)g(library)d(as)i(three)h(separate)150 1535 y(parts:)40
+-b(the)31 b(lo)m(w)f(lev)m(el)g(in)m(terface,)h(and)f(the)h(high)e(lev)m
+-(el)h(in)m(terface,)h(and)f(some)h(utilit)m(y)d(functions.)150
+-1692 y(The)38 b(structure)g(of)g Fj(libbzip2)p Fl('s)e(in)m(terfaces)j
+-(is)e(similar)f(to)j(that)g(of)g(Jean-loup)e(Gailly's)g(and)h(Mark)150
+-1802 y(Adler's)29 b(excellen)m(t)i Fj(zlib)e Fl(library)-8
+-b(.)150 1959 y(All)29 b(externally)g(visible)f(sym)m(b)s(ols)h(ha)m(v)m
+-(e)i(names)f(b)s(eginning)e Fj(BZ2_)p Fl(.)39 b(This)29
+-b(is)g(new)h(in)f(v)m(ersion)h(1.0.)41 b(The)150 2068
+-y(in)m(ten)m(tion)30 b(is)f(to)i(minimise)d(p)s(ollution)f(of)k(the)f
+-(namespaces)h(of)g(library)d(clien)m(ts.)150 2321 y Ff(3.1.1)63
+-b(Lo)m(w-lev)m(el)39 b(summary)150 2514 y Fl(This)21
+-b(in)m(terface)h(pro)m(vides)g(services)g(for)g(compressing)f(and)h
+-(decompressing)f(data)i(in)e(memory)-8 b(.)38 b(There's)150
+-2623 y(no)43 b(pro)m(vision)e(for)h(dealing)g(with)f(\014les,)k
+-(streams)e(or)g(an)m(y)g(other)g(I/O)g(mec)m(hanisms,)i(just)e(straigh)
+-m(t)150 2733 y(memory-to-memory)25 b(w)m(ork.)38 b(In)23
+-b(fact,)k(this)22 b(part)i(of)f(the)h(library)d(can)j(b)s(e)f(compiled)
+-f(without)h(inclusion)150 2843 y(of)31 b Fj(stdio.h)p
+-Fl(,)d(whic)m(h)h(ma)m(y)i(b)s(e)f(helpful)d(for)k(em)m(b)s(edded)e
+-(applications.)150 2999 y(The)h(lo)m(w-lev)m(el)g(part)g(of)h(the)f
+-(library)e(has)i(no)h(global)e(v)-5 b(ariables)29 b(and)h(is)g
+-(therefore)g(thread-safe.)150 3156 y(Six)d(routines)g(mak)m(e)j(up)d
+-(the)i(lo)m(w)f(lev)m(el)g(in)m(terface:)41 b Fj(BZ2_bzCompressInit)p
+-Fl(,)24 b Fj(BZ2_bzCompress)p Fl(,)h(and)150 3266 y Fj
+-(BZ2_bzCompressEnd)h Fl(for)k(compression,)f(and)h(a)h(corresp)s
+-(onding)d(trio)i Fj(BZ2_bzDecompressInit)p Fl(,)150 3375
+-y Fj(BZ2_bzDecompress)37 b Fl(and)j Fj(BZ2_bzDecompressEnd)c
+-Fl(for)42 b(decompression.)72 b(The)41 b Fj(*Init)e Fl(functions)150
+-3485 y(allo)s(cate)44 b(memory)g(for)f(compression/decompression)f(and)
+-h(do)h(other)g(initialisations,)f(whilst)f(the)150 3595
+-y Fj(*End)29 b Fl(functions)g(close)i(do)m(wn)f(op)s(erations)f(and)h
+-(release)h(memory)-8 b(.)150 3751 y(The)36 b(real)f(w)m(ork)i(is)e
+-(done)h(b)m(y)g Fj(BZ2_bzCompress)c Fl(and)j Fj(BZ2_bzDecompress)p
+-Fl(.)54 b(These)36 b(compress)g(and)150 3861 y(decompress)30
+-b(data)h(from)f(a)h(user-supplied)c(input)i(bu\013er)g(to)i(a)g
+-(user-supplied)c(output)j(bu\013er.)40 b(These)150 3971
+-y(bu\013ers)32 b(can)i(b)s(e)e(an)m(y)i(size;)g(arbitrary)e(quan)m
+-(tities)h(of)g(data)h(are)g(handled)d(b)m(y)i(making)f(rep)s(eated)i
+-(calls)150 4080 y(to)f(these)f(functions.)44 b(This)30
+-b(is)h(a)h(\015exible)e(mec)m(hanism)i(allo)m(wing)e(a)i(consumer-pull)
+-e(st)m(yle)i(of)g(activit)m(y)-8 b(,)150 4190 y(or)30
+-b(pro)s(ducer-push,)e(or)i(a)h(mixture)e(of)i(b)s(oth.)150
+-4443 y Ff(3.1.2)63 b(High-lev)m(el)41 b(summary)150 4635
+-y Fl(This)d(in)m(terface)j(pro)m(vides)e(some)h(handy)f(wrapp)s(ers)f
+-(around)h(the)i(lo)m(w-lev)m(el)f(in)m(terface)g(to)h(facilitate)150
+-4745 y(reading)26 b(and)g(writing)f Fj(bzip2)g Fl(format)i(\014les)f
+-(\()p Fj(.bz2)g Fl(\014les\).)38 b(The)27 b(routines)e(pro)m(vide)h(ho)
+-s(oks)h(to)g(facilitate)150 4854 y(reading)43 b(\014les)f(in)h(whic)m
+-(h)f(the)i Fj(bzip2)f Fl(data)h(stream)g(is)f(em)m(b)s(edded)f(within)g
+-(some)i(larger-scale)g(\014le)150 4964 y(structure,)30
+-b(or)h(where)e(there)i(are)g(m)m(ultiple)d Fj(bzip2)h
+-Fl(data)i(streams)f(concatenated)j(end-to-end.)150 5121
+-y(F)-8 b(or)31 b(reading)f(\014les,)f Fj(BZ2_bzReadOpen)p
+-Fl(,)e Fj(BZ2_bzRead)p Fl(,)h Fj(BZ2_bzReadClose)e Fl(and)150
+-5230 y Fj(BZ2_bzReadGetUnused)19 b Fl(are)25 b(supplied.)36
+-b(F)-8 b(or)25 b(writing)d(\014les,)j Fj(BZ2_bzWriteOpen)p
+-Fl(,)d Fj(BZ2_bzWrite)g Fl(and)150 5340 y Fj(BZ2_bzWriteFinish)k
+-Fl(are)k(a)m(v)-5 b(ailable.)p eop
+-%%Page: 10 11
+-10 10 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(10)150 299 y(As)24 b(with)f(the)h(lo)m
+-(w-lev)m(el)h(library)-8 b(,)23 b(no)h(global)g(v)-5
+-b(ariables)23 b(are)h(used)g(so)g(the)h(library)c(is)j(p)s(er)f(se)h
+-(thread-safe.)150 408 y(Ho)m(w)m(ev)m(er,)32 b(if)c(I/O)h(errors)g(o)s
+-(ccur)g(whilst)e(reading)i(or)g(writing)e(the)j(underlying)c
+-(compressed)j(\014les,)g(y)m(ou)150 518 y(ma)m(y)j(ha)m(v)m(e)g(to)g
+-(consult)e Fj(errno)g Fl(to)h(determine)g(the)g(cause)g(of)h(the)f
+-(error.)42 b(In)30 b(that)i(case,)h(y)m(ou'd)e(need)g(a)150
+-628 y(C)f(library)e(whic)m(h)h(correctly)i(supp)s(orts)d
+-Fj(errno)h Fl(in)g(a)i(m)m(ultithreaded)e(en)m(vironmen)m(t.)150
+-784 y(T)-8 b(o)56 b(mak)m(e)g(the)g(library)d(a)j(little)e(simpler)f
+-(and)i(more)h(p)s(ortable,)61 b Fj(BZ2_bzReadOpen)51
+-b Fl(and)k Fj(BZ2_)150 894 y(bzWriteOpen)34 b Fl(require)j(y)m(ou)g(to)
+-i(pass)e(them)g(\014le)g(handles)f(\()p Fj(FILE*)p Fl(s\))g(whic)m(h)h
+-(ha)m(v)m(e)h(previously)e(b)s(een)150 1004 y(op)s(ened)41
+-b(for)g(reading)f(or)h(writing)f(resp)s(ectiv)m(ely)-8
+-b(.)73 b(That)41 b(a)m(v)m(oids)h(p)s(ortabilit)m(y)d(problems)g(asso)s
+-(ciated)150 1113 y(with)j(\014le)h(op)s(erations)g(and)g(\014le)g
+-(attributes,)j(whilst)c(not)i(b)s(eing)e(m)m(uc)m(h)h(of)h(an)g(imp)s
+-(osition)c(on)k(the)150 1223 y(programmer.)150 1474 y
+-Ff(3.1.3)63 b(Utilit)m(y)40 b(functions)h(summary)150
+-1666 y Fl(F)-8 b(or)45 b(v)m(ery)g(simple)d(needs,)48
+-b Fj(BZ2_bzBuffToBuffCompres)o(s)38 b Fl(and)44 b Fj
+-(BZ2_bzBuffToBuffDecompres)o(s)150 1776 y Fl(are)29 b(pro)m(vided.)38
+-b(These)28 b(compress)g(data)h(in)e(memory)h(from)g(one)h(bu\013er)e
+-(to)i(another)f(bu\013er)g(in)f(a)h(single)150 1885 y(function)38
+-b(call.)67 b(Y)-8 b(ou)40 b(should)d(assess)j(whether)f(these)h
+-(functions)d(ful\014ll)f(y)m(our)k(memory-to-memory)150
+-1995 y(compression/decompression)26 b(requiremen)m(ts)h(b)s(efore)g(in)
+-m(v)m(esting)g(e\013ort)i(in)d(understanding)f(the)j(more)150
+-2105 y(general)i(but)g(more)h(complex)f(lo)m(w-lev)m(el)g(in)m
+-(terface.)150 2261 y(Y)-8 b(oshiok)j(a)47 b(Tsuneo)e(\()p
+-Fj(QWF00133@niftyserve.or.jp)40 b Fl(/)46 b Fj
+-(tsuneo-y@is.aist-nara.ac.)o(jp)p Fl(\))40 b(has)150
+-2371 y(con)m(tributed)f(some)h(functions)e(to)j(giv)m(e)f(b)s(etter)g
+-Fj(zlib)f Fl(compatibilit)m(y)-8 b(.)67 b(These)40 b(functions)e(are)i
+-Fj(BZ2_)150 2481 y(bzopen)p Fl(,)e Fj(BZ2_bzread)p Fl(,)f
+-Fj(BZ2_bzwrite)p Fl(,)g Fj(BZ2_bzflush)p Fl(,)g Fj(BZ2_bzclose)p
+-Fl(,)f Fj(BZ2_bzerror)f Fl(and)i Fj(BZ2_)150 2590 y(bzlibVersion)p
+-Fl(.)49 b(Y)-8 b(ou)35 b(ma)m(y)g(\014nd)e(these)i(functions)d(more)j
+-(con)m(v)m(enien)m(t)g(for)f(simple)f(\014le)g(reading)h(and)150
+-2700 y(writing,)c(than)h(those)h(in)e(the)i(high-lev)m(el)e(in)m
+-(terface.)45 b(These)31 b(functions)f(are)i(not)g(\(y)m(et\))h
+-(o\016cially)d(part)150 2809 y(of)k(the)g(library)-8
+-b(,)33 b(and)g(are)h(minimally)c(do)s(cumen)m(ted)k(here.)51
+-b(If)33 b(they)h(break,)h(y)m(ou)f(get)h(to)g(k)m(eep)f(all)f(the)150
+-2919 y(pieces.)40 b(I)31 b(hop)s(e)e(to)i(do)s(cumen)m(t)g(them)f(prop)
+-s(erly)e(when)h(time)i(p)s(ermits.)150 3076 y(Y)-8 b(oshiok)j(a)27
+-b(also)g(con)m(tributed)f(mo)s(di\014cations)f(to)i(allo)m(w)f(the)h
+-(library)e(to)i(b)s(e)f(built)f(as)i(a)g(Windo)m(ws)f(DLL.)150
+-3362 y Fk(3.2)68 b(Error)45 b(handling)150 3554 y Fl(The)23
+-b(library)f(is)h(designed)g(to)i(reco)m(v)m(er)g(cleanly)f(in)e(all)h
+-(situations,)h(including)d(the)j(w)m(orst-case)i(situation)150
+-3664 y(of)j(decompressing)e(random)g(data.)41 b(I'm)28
+-b(not)h(100\045)g(sure)f(that)h(it)f(can)h(alw)m(a)m(ys)g(do)f(this,)g
+-(so)g(y)m(ou)h(migh)m(t)150 3774 y(w)m(an)m(t)i(to)g(add)e(a)i(signal)d
+-(handler)g(to)j(catc)m(h)h(segmen)m(tation)f(violations)e(during)f
+-(decompression)h(if)g(y)m(ou)150 3883 y(are)g(feeling)f(esp)s(ecially)f
+-(paranoid.)39 b(I)28 b(w)m(ould)g(b)s(e)g(in)m(terested)h(in)e(hearing)
+-h(more)h(ab)s(out)f(the)h(robustness)150 3993 y(of)i(the)f(library)e
+-(to)j(corrupted)f(compressed)g(data.)150 4150 y(V)-8
+-b(ersion)39 b(1.0)h(is)f(m)m(uc)m(h)g(more)h(robust)e(in)g(this)g(resp)
+-s(ect)i(than)f(0.9.0)i(or)e(0.9.5.)70 b(In)m(v)m(estigations)39
+-b(with)150 4259 y(Chec)m(k)m(er)21 b(\(a)g(to)s(ol)g(for)f(detecting)h
+-(problems)d(with)h(memory)h(managemen)m(t,)k(similar)18
+-b(to)j(Purify\))e(indicate)150 4369 y(that,)40 b(at)e(least)f(for)g
+-(the)h(few)e(\014les)h(I)g(tested,)j(all)c(single-bit)f(errors)i(in)e
+-(the)j(decompressed)f(data)h(are)150 4478 y(caugh)m(t)c(prop)s(erly)-8
+-b(,)31 b(with)g(no)i(segmen)m(tation)h(faults,)e(no)g(reads)h(of)g
+-(uninitialised)27 b(data)34 b(and)e(no)g(out)h(of)150
+-4588 y(range)f(reads)g(or)f(writes.)44 b(So)32 b(it's)f(certainly)g(m)m
+-(uc)m(h)h(impro)m(v)m(ed,)g(although)f(I)g(w)m(ouldn't)g(claim)g(it)g
+-(to)i(b)s(e)150 4698 y(totally)d(b)s(om)m(bpro)s(of.)150
+-4854 y(The)25 b(\014le)g Fj(bzlib.h)f Fl(con)m(tains)i(all)f
+-(de\014nitions)e(needed)i(to)i(use)e(the)h(library)-8
+-b(.)37 b(In)26 b(particular,)f(y)m(ou)h(should)150 4964
+-y(de\014nitely)i(not)j(include)d Fj(bzlib_private.h)p
+-Fl(.)150 5121 y(In)39 b Fj(bzlib.h)p Fl(,)h(the)g(v)-5
+-b(arious)39 b(return)f(v)-5 b(alues)39 b(are)h(de\014ned.)68
+-b(The)39 b(follo)m(wing)f(list)h(is)f(not)i(in)m(tended)f(as)150
+-5230 y(an)c(exhaustiv)m(e)h(description)d(of)i(the)h(circumstances)f
+-(in)f(whic)m(h)g(a)i(giv)m(en)f(v)-5 b(alue)35 b(ma)m(y)h(b)s(e)e
+-(returned)h({)150 5340 y(those)h(descriptions)d(are)j(giv)m(en)f
+-(later.)56 b(Rather,)37 b(it)d(is)h(in)m(tended)f(to)i(con)m(v)m(ey)h
+-(the)e(rough)g(meaning)g(of)p eop
+-%%Page: 11 12
+-11 11 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(11)150 299 y(eac)m(h)38
+-b(return)d(v)-5 b(alue.)59 b(The)36 b(\014rst)g(\014v)m(e)g(actions)h
+-(are)g(normal)f(and)f(not)i(in)m(tended)f(to)h(denote)g(an)f(error)150
+-408 y(situation.)150 592 y Fj(BZ_OK)180 b Fl(The)30 b(requested)g
+-(action)h(w)m(as)g(completed)f(successfully)-8 b(.)150
+-756 y Fj(BZ_RUN_OK)150 866 y(BZ_FLUSH_OK)150 975 y(BZ_FINISH_OK)630
+-1085 y Fl(In)24 b Fj(BZ2_bzCompress)p Fl(,)e(the)i(requested)g
+-(\015ush/\014nish/nothing-sp)s(ecial)c(action)k(w)m(as)h(com-)630
+-1194 y(pleted)30 b(successfully)-8 b(.)150 1358 y Fj(BZ_STREAM_END)630
+-1468 y Fl(Compression)38 b(of)j(data)f(w)m(as)h(completed,)h(or)f(the)f
+-(logical)f(stream)i(end)e(w)m(as)i(detected)630 1577
+-y(during)28 b(decompression.)150 1761 y(The)i(follo)m(wing)f(return)g
+-(v)-5 b(alues)30 b(indicate)f(an)h(error)g(of)h(some)g(kind.)150
+-1945 y Fj(BZ_CONFIG_ERROR)630 2055 y Fl(Indicates)48
+-b(that)h(the)g(library)e(has)h(b)s(een)g(improp)s(erly)d(compiled)j(on)
+-g(y)m(our)h(platform)630 2164 y({)j(a)g(ma)5 b(jor)51
+-b(con\014guration)g(error.)104 b(Sp)s(eci\014cally)-8
+-b(,)55 b(it)c(means)g(that)h Fj(sizeof\(char\))p Fl(,)630
+-2274 y Fj(sizeof\(short\))44 b Fl(and)i Fj(sizeof\(int\))f
+-Fl(are)j(not)f(1,)52 b(2)c(and)f(4)h(resp)s(ectiv)m(ely)-8
+-b(,)51 b(as)d(they)630 2384 y(should)27 b(b)s(e.)40 b(Note)30
+-b(that)g(the)f(library)e(should)g(still)g(w)m(ork)i(prop)s(erly)e(on)i
+-(64-bit)g(platforms)630 2493 y(whic)m(h)d(follo)m(w)h(the)g(LP64)h
+-(programming)e(mo)s(del)h({)g(that)h(is,)g(where)e Fj(sizeof\(long\))f
+-Fl(and)630 2603 y Fj(sizeof\(void*\))e Fl(are)k(8.)40
+-b(Under)25 b(LP64,)j Fj(sizeof\(int\))c Fl(is)h(still)f(4,)k(so)f
+-Fj(libbzip2)p Fl(,)e(whic)m(h)630 2712 y(do)s(esn't)30
+-b(use)g(the)h Fj(long)e Fl(t)m(yp)s(e,)i(is)e(OK.)150
+-2876 y Fj(BZ_SEQUENCE_ERROR)630 2986 y Fl(When)43 b(using)f(the)i
+-(library)-8 b(,)45 b(it)e(is)f(imp)s(ortan)m(t)h(to)h(call)e(the)i
+-(functions)e(in)g(the)i(correct)630 3095 y(sequence)28
+-b(and)f(with)f(data)j(structures)e(\(bu\013ers)f(etc\))j(in)e(the)g
+-(correct)i(states.)41 b Fj(libbzip2)630 3205 y Fl(c)m(hec)m(ks)26
+-b(as)e(m)m(uc)m(h)h(as)f(it)g(can)g(to)h(ensure)f(this)f(is)g(happ)s
+-(ening,)h(and)f(returns)g Fj(BZ_SEQUENCE_)630 3314 y(ERROR)36
+-b Fl(if)h(not.)62 b(Co)s(de)37 b(whic)m(h)g(complies)f(precisely)g
+-(with)h(the)g(function)g(seman)m(tics,)j(as)630 3424
+-y(detailed)d(b)s(elo)m(w,)i(should)d(nev)m(er)i(receiv)m(e)h(this)d(v)
+--5 b(alue;)41 b(suc)m(h)d(an)g(ev)m(en)m(t)h(denotes)f(buggy)630
+-3534 y(co)s(de)31 b(whic)m(h)e(y)m(ou)h(should)f(in)m(v)m(estigate.)150
+-3697 y Fj(BZ_PARAM_ERROR)630 3807 y Fl(Returned)43 b(when)f(a)i
+-(parameter)g(to)h(a)f(function)e(call)h(is)f(out)i(of)g(range)g(or)g
+-(otherwise)630 3917 y(manifestly)34 b(incorrect.)57 b(As)36
+-b(with)e Fj(BZ_SEQUENCE_ERROR)p Fl(,)f(this)i(denotes)h(a)g(bug)f(in)g
+-(the)630 4026 y(clien)m(t)23 b(co)s(de.)39 b(The)22 b(distinction)f(b)s
+-(et)m(w)m(een)j Fj(BZ_PARAM_ERROR)c Fl(and)j Fj(BZ_SEQUENCE_ERROR)630
+-4136 y Fl(is)29 b(a)i(bit)f(hazy)-8 b(,)31 b(but)f(still)e(w)m(orth)i
+-(making.)150 4300 y Fj(BZ_MEM_ERROR)630 4409 y Fl(Returned)g(when)f(a)i
+-(request)f(to)i(allo)s(cate)f(memory)f(failed.)40 b(Note)31
+-b(that)g(the)g(quan)m(tit)m(y)g(of)630 4519 y(memory)21
+-b(needed)g(to)i(decompress)e(a)g(stream)h(cannot)g(b)s(e)f(determined)f
+-(un)m(til)g(the)h(stream's)630 4628 y(header)29 b(has)g(b)s(een)g
+-(read.)40 b(So)29 b Fj(BZ2_bzDecompress)c Fl(and)j Fj(BZ2_bzRead)f
+-Fl(ma)m(y)j(return)e Fj(BZ_)630 4738 y(MEM_ERROR)d Fl(ev)m(en)k(though)
+-e(some)h(of)g(the)g(compressed)g(data)g(has)g(b)s(een)f(read.)39
+-b(The)28 b(same)630 4847 y(is)38 b(not)i(true)f(for)g(compression;)k
+-(once)d Fj(BZ2_bzCompressInit)34 b Fl(or)39 b Fj(BZ2_bzWriteOpen)630
+-4957 y Fl(ha)m(v)m(e)32 b(successfully)c(completed,)j
+-Fj(BZ_MEM_ERROR)c Fl(cannot)k(o)s(ccur.)150 5121 y Fj(BZ_DATA_ERROR)630
+-5230 y Fl(Returned)h(when)g(a)h(data)g(in)m(tegrit)m(y)g(error)g(is)e
+-(detected)k(during)30 b(decompression.)47 b(Most)630
+-5340 y(imp)s(ortan)m(tly)-8 b(,)31 b(this)f(means)i(when)f(stored)g
+-(and)g(computed)h(CR)m(Cs)f(for)g(the)h(data)g(do)g(not)p
+-eop
+-%%Page: 12 13
+-12 12 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(12)630 299 y(matc)m(h.)41
+-b(This)28 b(v)-5 b(alue)29 b(is)f(also)i(returned)e(up)s(on)g
+-(detection)i(of)g(an)m(y)g(other)f(anomaly)h(in)e(the)630
+-408 y(compressed)i(data.)150 560 y Fj(BZ_DATA_ERROR_MAGIC)630
+-670 y Fl(As)k(a)g(sp)s(ecial)f(case)i(of)f Fj(BZ_DATA_ERROR)p
+-Fl(,)d(it)i(is)g(sometimes)h(useful)e(to)j(kno)m(w)f(when)f(the)630
+-779 y(compressed)d(stream)h(do)s(es)f(not)g(start)h(with)e(the)i
+-(correct)h(magic)e(b)m(ytes)h(\()p Fj('B')f('Z')f('h')p
+-Fl(\).)150 931 y Fj(BZ_IO_ERROR)630 1040 y Fl(Returned)k(b)m(y)h
+-Fj(BZ2_bzRead)d Fl(and)i Fj(BZ2_bzWrite)e Fl(when)i(there)h(is)f(an)g
+-(error)h(reading)f(or)630 1150 y(writing)28 b(in)h(the)h(compressed)g
+-(\014le,)f(and)h(b)m(y)g Fj(BZ2_bzReadOpen)c Fl(and)j
+-Fj(BZ2_bzWriteOpen)630 1259 y Fl(for)i(attempts)i(to)f(use)f(a)h
+-(\014le)e(for)i(whic)m(h)e(the)h(error)g(indicator)g(\(viz,)g
+-Fj(ferror\(f\))p Fl(\))f(is)g(set.)630 1369 y(On)h(receipt)g(of)h
+-Fj(BZ_IO_ERROR)p Fl(,)e(the)h(caller)h(should)d(consult)i
+-Fj(errno)g Fl(and/or)g Fj(perror)f Fl(to)630 1479 y(acquire)g(op)s
+-(erating-system)g(sp)s(eci\014c)f(information)g(ab)s(out)h(the)h
+-(problem.)150 1630 y Fj(BZ_UNEXPECTED_EOF)630 1740 y
+-Fl(Returned)36 b(b)m(y)g Fj(BZ2_bzRead)e Fl(when)i(the)h(compressed)f
+-(\014le)g(\014nishes)e(b)s(efore)j(the)f(logical)630
+-1849 y(end)30 b(of)g(stream)h(is)e(detected.)150 2001
+-y Fj(BZ_OUTBUFF_FULL)630 2110 y Fl(Returned)g(b)m(y)i
+-Fj(BZ2_bzBuffToBuffCompres)o(s)24 b Fl(and)30 b Fj
+-(BZ2_bzBuffToBuffDecompres)o(s)630 2220 y Fl(to)h(indicate)f(that)h
+-(the)f(output)g(data)h(will)d(not)i(\014t)h(in)m(to)f(the)h(output)f
+-(bu\013er)f(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 Ff(3.3.1)63 b Fe(BZ2_bzCompressInit)390
+-2953 y Fj(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: 13 14
+-13 13 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(13)150 456 y(Prepares)32
+-b(for)h(compression.)47 b(The)32 b Fj(bz_stream)e Fl(structure)j(holds)
+-e(all)h(data)h(p)s(ertaining)e(to)i(the)g(com-)150 565
+-y(pression)i(activit)m(y)-8 b(.)62 b(A)37 b Fj(bz_stream)e
+-Fl(structure)h(should)f(b)s(e)i(allo)s(cated)g(and)f(initialised)e
+-(prior)h(to)j(the)150 675 y(call.)67 b(The)39 b(\014elds)e(of)j
+-Fj(bz_stream)d Fl(comprise)h(the)i(en)m(tiret)m(y)g(of)f(the)h
+-(user-visible)c(data.)68 b Fj(state)38 b Fl(is)h(a)150
+-784 y(p)s(oin)m(ter)29 b(to)i(the)g(priv)-5 b(ate)30
+-b(data)h(structures)f(required)e(for)i(compression.)150
+-941 y(Custom)37 b(memory)g(allo)s(cators)g(are)h(supp)s(orted,)f(via)g
+-(\014elds)f Fj(bzalloc)p Fl(,)h Fj(bzfree)p Fl(,)g(and)g
+-Fj(opaque)p Fl(.)59 b(The)150 1051 y(v)-5 b(alue)32 b
+-Fj(opaque)e Fl(is)i(passed)f(to)i(as)g(the)f(\014rst)g(argumen)m(t)h
+-(to)g(all)e(calls)g(to)i Fj(bzalloc)d Fl(and)i Fj(bzfree)p
+-Fl(,)f(but)h(is)150 1160 y(otherwise)d(ignored)g(b)m(y)h(the)g(library)
+--8 b(.)38 b(The)29 b(call)h Fj(bzalloc)e(\()i(opaque,)e(n,)i(m)g(\))g
+-Fl(is)e(exp)s(ected)j(to)f(return)150 1270 y(a)g(p)s(oin)m(ter)e
+-Fj(p)h Fl(to)h Fj(n)g(*)g(m)f Fl(b)m(ytes)h(of)g(memory)-8
+-b(,)30 b(and)e Fj(bzfree)h(\()h(opaque,)f(p)h(\))f Fl(should)e(free)i
+-(that)h(memory)-8 b(.)150 1427 y(If)33 b(y)m(ou)g(don't)h(w)m(an)m(t)g
+-(to)g(use)f(a)g(custom)h(memory)f(allo)s(cator,)h(set)g
+-Fj(bzalloc)p Fl(,)e Fj(bzfree)g Fl(and)h Fj(opaque)e
+-Fl(to)150 1537 y Fj(NULL)p Fl(,)e(and)h(the)h(library)d(will)f(then)k
+-(use)f(the)g(standard)g Fj(malloc)p Fl(/)p Fj(free)e
+-Fl(routines.)150 1693 y(Before)39 b(calling)d Fj(BZ2_bzCompressInit)p
+-Fl(,)f(\014elds)h Fj(bzalloc)p Fl(,)h Fj(bzfree)f Fl(and)h
+-Fj(opaque)f Fl(should)g(b)s(e)h(\014lled)150 1803 y(appropriately)-8
+-b(,)35 b(as)h(just)f(describ)s(ed.)53 b(Up)s(on)34 b(return,)i(the)g
+-(in)m(ternal)e(state)i(will)d(ha)m(v)m(e)j(b)s(een)f(allo)s(cated)150
+-1913 y(and)43 b(initialised,)g(and)g Fj(total_in_lo32)p
+-Fl(,)h Fj(total_in_hi32)p Fl(,)f Fj(total_out_lo32)d
+-Fl(and)j Fj(total_out_)150 2022 y(hi32)37 b Fl(will)f(ha)m(v)m(e)j(b)s
+-(een)f(set)h(to)g(zero.)65 b(These)38 b(four)g(\014elds)e(are)j(used)f
+-(b)m(y)g(the)g(library)e(to)j(inform)e(the)150 2132 y(caller)j(of)g
+-(the)h(total)g(amoun)m(t)g(of)g(data)g(passed)f(in)m(to)g(and)g(out)g
+-(of)h(the)g(library)-8 b(,)41 b(resp)s(ectiv)m(ely)-8
+-b(.)70 b(Y)-8 b(ou)150 2241 y(should)34 b(not)j(try)f(to)h(c)m(hange)g
+-(them.)58 b(As)36 b(of)h(v)m(ersion)f(1.0,)j(64-bit)d(coun)m(ts)h(are)f
+-(main)m(tained,)h(ev)m(en)g(on)150 2351 y(32-bit)i(platforms,)h(using)d
+-(the)i Fj(_hi32)e Fl(\014elds)g(to)j(store)f(the)g(upp)s(er)d(32)k
+-(bits)d(of)i(the)g(coun)m(t.)66 b(So,)41 b(for)150 2460
+-y(example,)30 b(the)h(total)g(amoun)m(t)g(of)f(data)h(in)f(is)f
+-Fj(\(total_in_hi32)d(<<)k(32\))g(+)g(total_in_lo32)p
+-Fl(.)150 2617 y(P)m(arameter)g Fj(blockSize100k)25 b
+-Fl(sp)s(eci\014es)i(the)h(blo)s(c)m(k)g(size)h(to)g(b)s(e)f(used)f(for)
+-h(compression.)40 b(It)28 b(should)f(b)s(e)150 2727 y(a)k(v)-5
+-b(alue)30 b(b)s(et)m(w)m(een)i(1)f(and)f(9)h(inclusiv)m(e,)e(and)h(the)
+-h(actual)g(blo)s(c)m(k)f(size)g(used)g(is)g(100000)j(x)e(this)e
+-(\014gure.)42 b(9)150 2836 y(giv)m(es)31 b(the)f(b)s(est)g(compression)
+-g(but)f(tak)m(es)j(most)f(memory)-8 b(.)150 2993 y(P)m(arameter)29
+-b Fj(verbosity)c Fl(should)h(b)s(e)h(set)i(to)f(a)h(n)m(um)m(b)s(er)d
+-(b)s(et)m(w)m(een)j(0)f(and)f(4)h(inclusiv)m(e.)38 b(0)28
+-b(is)f(silen)m(t,)h(and)150 3103 y(greater)j(n)m(um)m(b)s(ers)c(giv)m
+-(e)j(increasingly)d(v)m(erb)s(ose)j(monitoring/debugging)d(output.)40
+-b(If)29 b(the)g(library)e(has)150 3212 y(b)s(een)j(compiled)e(with)i
+-Fj(-DBZ_NO_STDIO)p Fl(,)d(no)j(suc)m(h)g(output)g(will)e(app)s(ear)h
+-(for)h(an)m(y)h(v)m(erb)s(osit)m(y)f(setting.)150 3369
+-y(P)m(arameter)35 b Fj(workFactor)d Fl(con)m(trols)i(ho)m(w)g(the)g
+-(compression)f(phase)h(b)s(eha)m(v)m(es)g(when)f(presen)m(ted)h(with)
+-150 3479 y(w)m(orst)40 b(case,)j(highly)37 b(rep)s(etitiv)m(e,)k(input)
+-d(data.)68 b(If)39 b(compression)g(runs)e(in)m(to)j(di\016culties)d
+-(caused)i(b)m(y)150 3588 y(rep)s(etitiv)m(e)34 b(data,)j(the)e(library)
+-d(switc)m(hes)j(from)f(the)h(standard)f(sorting)g(algorithm)g(to)i(a)f
+-(fallbac)m(k)f(al-)150 3698 y(gorithm.)47 b(The)32 b(fallbac)m(k)g(is)g
+-(slo)m(w)m(er)g(than)h(the)f(standard)g(algorithm)g(b)m(y)g(p)s(erhaps)
+-f(a)i(factor)h(of)e(three,)150 3808 y(but)e(alw)m(a)m(ys)h(b)s(eha)m(v)
+-m(es)f(reasonably)-8 b(,)31 b(no)f(matter)h(ho)m(w)g(bad)f(the)g
+-(input.)150 3965 y(Lo)m(w)m(er)25 b(v)-5 b(alues)24 b(of)h
+-Fj(workFactor)d Fl(reduce)i(the)h(amoun)m(t)g(of)g(e\013ort)g(the)g
+-(standard)f(algorithm)f(will)f(exp)s(end)150 4074 y(b)s(efore)j
+-(resorting)h(to)g(the)g(fallbac)m(k.)39 b(Y)-8 b(ou)27
+-b(should)c(set)k(this)e(parameter)h(carefully;)g(to)s(o)h(lo)m(w,)g
+-(and)e(man)m(y)150 4184 y(inputs)32 b(will)f(b)s(e)i(handled)f(b)m(y)i
+-(the)g(fallbac)m(k)g(algorithm)f(and)g(so)h(compress)g(rather)g(slo)m
+-(wly)-8 b(,)34 b(to)s(o)h(high,)150 4293 y(and)24 b(y)m(our)h(a)m(v)m
+-(erage-to-w)m(orst)30 b(case)c(compression)e(times)h(can)g(b)s(ecome)g
+-(v)m(ery)h(large.)39 b(The)24 b(default)g(v)-5 b(alue)150
+-4403 y(of)31 b(30)g(giv)m(es)f(reasonable)h(b)s(eha)m(viour)e(o)m(v)m
+-(er)i(a)g(wide)e(range)i(of)f(circumstances.)150 4560
+-y(Allo)m(w)m(able)h(v)-5 b(alues)31 b(range)i(from)e(0)i(to)f(250)h
+-(inclusiv)m(e.)44 b(0)32 b(is)f(a)h(sp)s(ecial)f(case,)i(equiv)-5
+-b(alen)m(t)32 b(to)g(using)f(the)150 4669 y(default)f(v)-5
+-b(alue)29 b(of)i(30.)150 4826 y(Note)38 b(that)f(the)g(compressed)f
+-(output)g(generated)h(is)f(the)g(same)h(regardless)f(of)h(whether)f(or)
+-g(not)h(the)150 4936 y(fallbac)m(k)30 b(algorithm)f(is)h(used.)150
+-5093 y(Be)23 b(a)m(w)m(are)h(also)f(that)g(this)f(parameter)h(ma)m(y)g
+-(disapp)s(ear)e(en)m(tirely)h(in)f(future)h(v)m(ersions)g(of)h(the)g
+-(library)-8 b(.)36 b(In)150 5202 y(principle)20 b(it)j(should)e(b)s(e)h
+-(p)s(ossible)f(to)j(devise)f(a)g(go)s(o)s(d)g(w)m(a)m(y)i(to)f
+-(automatically)f(c)m(ho)s(ose)h(whic)m(h)e(algorithm)150
+-5312 y(to)31 b(use.)41 b(Suc)m(h)29 b(a)i(mec)m(hanism)f(w)m(ould)f
+-(render)g(the)i(parameter)g(obsolete.)p eop
+-%%Page: 14 15
+-14 14 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(14)150 299 y(P)m(ossible)29
+-b(return)h(v)-5 b(alues:)572 450 y Fj(BZ_CONFIG_ERROR)663
+-554 y Fl(if)29 b(the)i(library)d(has)i(b)s(een)f(mis-compiled)572
+-657 y Fj(BZ_PARAM_ERROR)663 761 y Fl(if)g Fj(strm)g Fl(is)h
+-Fj(NULL)663 865 y Fl(or)g Fj(blockSize)e(<)i Fl(1)h(or)f
+-Fj(blockSize)e(>)i Fl(9)663 969 y(or)g Fj(verbosity)e(<)i
+-Fl(0)h(or)f Fj(verbosity)e(>)i Fl(4)663 1073 y(or)g Fj(workFactor)e(<)i
+-Fl(0)g(or)h Fj(workFactor)c(>)j Fl(250)572 1176 y Fj(BZ_MEM_ERROR)663
+-1280 y Fl(if)f(not)i(enough)f(memory)g(is)f(a)m(v)-5
+-b(ailable)572 1384 y Fj(BZ_OK)663 1488 y Fl(otherwise)150
+-1645 y(Allo)m(w)m(able)30 b(next)g(actions:)572 1796
+-y Fj(BZ2_bzCompress)663 1899 y Fl(if)f Fj(BZ_OK)g Fl(is)g(returned)572
+-2003 y(no)h(sp)s(eci\014c)f(action)i(needed)f(in)f(case)i(of)g(error)
+-150 2255 y Ff(3.3.2)63 b Fe(BZ2_bzCompress)533 2441 y
+-Fj(int)47 b(BZ2_bzCompress)d(\()j(bz_stream)f(*strm,)g(int)h(action)f
+-(\);)150 2598 y Fl(Pro)m(vides)28 b(more)g(input)f(and/or)h(output)g
+-(bu\013er)g(space)h(for)f(the)h(library)-8 b(.)38 b(The)28
+-b(caller)g(main)m(tains)f(input)150 2708 y(and)j(output)g(bu\013ers,)f
+-(and)h(calls)g Fj(BZ2_bzCompress)c Fl(to)31 b(transfer)f(data)h(b)s(et)
+-m(w)m(een)g(them.)150 2865 y(Before)j(eac)m(h)g(call)e(to)i
+-Fj(BZ2_bzCompress)p Fl(,)c Fj(next_in)h Fl(should)g(p)s(oin)m(t)h(at)h
+-(the)g(data)h(to)g(b)s(e)e(compressed,)150 2974 y(and)41
+-b Fj(avail_in)f Fl(should)g(indicate)h(ho)m(w)h(man)m(y)f(b)m(ytes)i
+-(the)f(library)d(ma)m(y)k(read.)75 b Fj(BZ2_bzCompress)150
+-3084 y Fl(up)s(dates)29 b Fj(next_in)p Fl(,)g Fj(avail_in)f
+-Fl(and)i Fj(total_in)e Fl(to)j(re\015ect)g(the)g(n)m(um)m(b)s(er)e(of)h
+-(b)m(ytes)h(it)f(has)g(read.)150 3241 y(Similarly)-8
+-b(,)27 b Fj(next_out)h Fl(should)g(p)s(oin)m(t)h(to)i(a)f(bu\013er)f
+-(in)g(whic)m(h)g(the)h(compressed)g(data)h(is)e(to)i(b)s(e)e(placed,)
+-150 3350 y(with)i Fj(avail_out)f Fl(indicating)h(ho)m(w)h(m)m(uc)m(h)h
+-(output)f(space)h(is)f(a)m(v)-5 b(ailable.)46 b Fj(BZ2_bzCompress)29
+-b Fl(up)s(dates)150 3460 y Fj(next_out)p Fl(,)f Fj(avail_out)g
+-Fl(and)i Fj(total_out)e Fl(to)j(re\015ect)g(the)f(n)m(um)m(b)s(er)g(of)
+-g(b)m(ytes)h(output.)150 3617 y(Y)-8 b(ou)40 b(ma)m(y)g(pro)m(vide)e
+-(and)h(remo)m(v)m(e)i(as)f(little)e(or)h(as)h(m)m(uc)m(h)f(data)h(as)g
+-(y)m(ou)f(lik)m(e)g(on)g(eac)m(h)i(call)e(of)g Fj(BZ2_)150
+-3726 y(bzCompress)p Fl(.)48 b(In)33 b(the)h(limit,)f(it)h(is)f
+-(acceptable)h(to)h(supply)c(and)j(remo)m(v)m(e)h(data)g(one)f(b)m(yte)g
+-(at)h(a)f(time,)150 3836 y(although)28 b(this)f(w)m(ould)g(b)s(e)h
+-(terribly)e(ine\016cien)m(t.)39 b(Y)-8 b(ou)29 b(should)e(alw)m(a)m(ys)
+-h(ensure)g(that)h(at)g(least)g(one)f(b)m(yte)150 3946
+-y(of)j(output)f(space)g(is)g(a)m(v)-5 b(ailable)30 b(at)h(eac)m(h)g
+-(call.)150 4102 y(A)38 b(second)h(purp)s(ose)d(of)j Fj(BZ2_bzCompress)
+-34 b Fl(is)j(to)i(request)f(a)h(c)m(hange)g(of)g(mo)s(de)e(of)i(the)f
+-(compressed)150 4212 y(stream.)150 4369 y(Conceptually)-8
+-b(,)24 b(a)g(compressed)g(stream)g(can)f(b)s(e)g(in)g(one)h(of)f(four)g
+-(states:)39 b(IDLE,)24 b(R)m(UNNING,)h(FLUSH-)150 4478
+-y(ING)37 b(and)g(FINISHING.)g(Before)i(initialisation)33
+-b(\()p Fj(BZ2_bzCompressInit)p Fl(\))g(and)j(after)i(termination)150
+-4588 y(\()p Fj(BZ2_bzCompressEnd)p Fl(\),)27 b(a)j(stream)h(is)f
+-(regarded)g(as)g(IDLE.)150 4745 y(Up)s(on)35 b(initialisation)e(\()p
+-Fj(BZ2_bzCompressInit)p Fl(\),)h(the)i(stream)h(is)e(placed)h(in)e(the)
+-j(R)m(UNNING)g(state.)150 4854 y(Subsequen)m(t)j(calls)g(to)i
+-Fj(BZ2_bzCompress)37 b Fl(should)j(pass)g Fj(BZ_RUN)g
+-Fl(as)h(the)g(requested)h(action;)47 b(other)150 4964
+-y(actions)31 b(are)f(illegal)f(and)h(will)d(result)j(in)f
+-Fj(BZ_SEQUENCE_ERROR)p Fl(.)150 5121 y(A)m(t)38 b(some)f(p)s(oin)m(t,)h
+-(the)f(calling)e(program)i(will)d(ha)m(v)m(e)k(pro)m(vided)e(all)f(the)
+-i(input)e(data)j(it)e(w)m(an)m(ts)i(to.)61 b(It)150 5230
+-y(will)28 b(then)h(w)m(an)m(t)i(to)g(\014nish)d(up)h({)i(in)d
+-(e\013ect,)k(asking)e(the)g(library)e(to)j(pro)s(cess)f(an)m(y)g(data)h
+-(it)f(migh)m(t)g(ha)m(v)m(e)150 5340 y(bu\013ered)25
+-b(in)m(ternally)-8 b(.)38 b(In)25 b(this)g(state,)k Fj(BZ2_bzCompress)
+-22 b Fl(will)i(no)i(longer)g(attempt)h(to)g(read)f(data)h(from)p
+-eop
+-%%Page: 15 16
+-15 15 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(15)150 299 y Fj(next_in)p
+-Fl(,)33 b(but)g(it)h(will)d(w)m(an)m(t)k(to)g(write)e(data)h(to)h
+-Fj(next_out)p Fl(.)49 b(Because)36 b(the)e(output)f(bu\013er)g
+-(supplied)150 408 y(b)m(y)e(the)h(user)e(can)i(b)s(e)f(arbitrarily)d
+-(small,)j(the)g(\014nishing-up)d(op)s(eration)i(cannot)i(necessarily)e
+-(b)s(e)h(done)150 518 y(with)e(a)i(single)e(call)h(of)g
+-Fj(BZ2_bzCompress)p Fl(.)150 675 y(Instead,)47 b(the)d(calling)f
+-(program)g(passes)h Fj(BZ_FINISH)d Fl(as)j(an)g(action)g(to)h
+-Fj(BZ2_bzCompress)p Fl(.)77 b(This)150 784 y(c)m(hanges)30
+-b(the)f(stream's)g(state)h(to)f(FINISHING.)g(An)m(y)g(remaining)e
+-(input)g(\(ie,)i Fj(next_in[0)f(..)i(avail_)150 894 y(in-1])p
+-Fl(\))36 b(is)f(compressed)i(and)f(transferred)g(to)h(the)g(output)g
+-(bu\013er.)58 b(T)-8 b(o)38 b(do)e(this,)i Fj(BZ2_bzCompress)150
+-1004 y Fl(m)m(ust)h(b)s(e)f(called)g(rep)s(eatedly)h(un)m(til)e(all)h
+-(the)h(output)f(has)h(b)s(een)f(consumed.)66 b(A)m(t)40
+-b(that)g(p)s(oin)m(t,)g Fj(BZ2_)150 1113 y(bzCompress)h
+-Fl(returns)h Fj(BZ_STREAM_END)p Fl(,)i(and)f(the)h(stream's)g(state)h
+-(is)d(set)j(bac)m(k)f(to)g(IDLE.)g Fj(BZ2_)150 1223 y(bzCompressEnd)27
+-b Fl(should)h(then)i(b)s(e)g(called.)150 1380 y(Just)25
+-b(to)i(mak)m(e)g(sure)e(the)i(calling)d(program)i(do)s(es)g(not)g(c)m
+-(heat,)i(the)f(library)c(mak)m(es)k(a)f(note)h(of)f Fj(avail_in)150
+-1489 y Fl(at)g(the)g(time)f(of)g(the)g(\014rst)g(call)g(to)h
+-Fj(BZ2_bzCompress)21 b Fl(whic)m(h)j(has)h Fj(BZ_FINISH)e
+-Fl(as)i(an)h(action)f(\(ie,)i(at)f(the)150 1599 y(time)d(the)h(program)
+-g(has)f(announced)g(its)h(in)m(ten)m(tion)f(to)h(not)g(supply)e(an)m(y)
+-i(more)g(input\).)37 b(By)24 b(comparing)150 1708 y(this)k(v)-5
+-b(alue)28 b(with)g(that)h(of)h Fj(avail_in)c Fl(o)m(v)m(er)k(subsequen)
+-m(t)f(calls)f(to)h Fj(BZ2_bzCompress)p Fl(,)d(the)j(library)e(can)150
+-1818 y(detect)33 b(an)m(y)e(attempts)i(to)f(slip)d(in)h(more)h(data)h
+-(to)h(compress.)43 b(An)m(y)31 b(calls)g(for)g(whic)m(h)f(this)g(is)h
+-(detected)150 1928 y(will)j(return)h Fj(BZ_SEQUENCE_ERROR)p
+-Fl(.)55 b(This)34 b(indicates)i(a)h(programming)e(mistak)m(e)i(whic)m
+-(h)e(should)g(b)s(e)150 2037 y(corrected.)150 2194 y(Instead)i(of)g
+-(asking)f(to)h(\014nish,)f(the)h(calling)f(program)g(ma)m(y)h(ask)g
+-Fj(BZ2_bzCompress)c Fl(to)38 b(tak)m(e)g(all)e(the)150
+-2304 y(remaining)j(input,)i(compress)f(it)g(and)g(terminate)h(the)g
+-(curren)m(t)f(\(Burro)m(ws-Wheeler\))h(compression)150
+-2413 y(blo)s(c)m(k.)e(This)26 b(could)h(b)s(e)g(useful)f(for)h(error)h
+-(con)m(trol)g(purp)s(oses.)38 b(The)27 b(mec)m(hanism)g(is)g(analogous)
+-h(to)g(that)150 2523 y(for)35 b(\014nishing:)46 b(call)35
+-b Fj(BZ2_bzCompress)c Fl(with)i(an)i(action)g(of)g Fj(BZ_FLUSH)p
+-Fl(,)g(remo)m(v)m(e)h(output)f(data,)i(and)150 2632 y(p)s(ersist)h
+-(with)g(the)i Fj(BZ_FLUSH)e Fl(action)i(un)m(til)e(the)i(v)-5
+-b(alue)39 b Fj(BZ_RUN)f Fl(is)h(returned.)68 b(As)39
+-b(with)g(\014nishing,)150 2742 y Fj(BZ2_bzCompress)23
+-b Fl(detects)28 b(an)m(y)f(attempt)h(to)f(pro)m(vide)f(more)h(input)e
+-(data)i(once)g(the)g(\015ush)e(has)i(b)s(egun.)150 2899
+-y(Once)j(the)h(\015ush)e(is)g(complete,)i(the)g(stream)f(returns)g(to)h
+-(the)f(normal)g(R)m(UNNING)h(state.)150 3056 y(This)f(all)h(sounds)g
+-(prett)m(y)h(complex,)h(but)e(isn't)g(really)-8 b(.)45
+-b(Here's)33 b(a)f(table)g(whic)m(h)f(sho)m(ws)h(whic)m(h)f(actions)150
+-3165 y(are)e(allo)m(w)m(able)f(in)f(eac)m(h)j(state,)g(what)f(action)g
+-(will)c(b)s(e)j(tak)m(en,)j(what)d(the)h(next)f(state)i(is,)e(and)g
+-(what)h(the)150 3275 y(non-error)h(return)f(v)-5 b(alues)29
+-b(are.)41 b(Note)32 b(that)e(y)m(ou)h(can't)g(explicitly)d(ask)i(what)g
+-(state)i(the)e(stream)h(is)e(in,)150 3384 y(but)h(nor)g(do)g(y)m(ou)h
+-(need)f(to)h({)g(it)e(can)i(b)s(e)f(inferred)e(from)i(the)h(v)-5
+-b(alues)29 b(returned)h(b)m(y)g Fj(BZ2_bzCompress)p Fl(.)390
+-3535 y(IDLE/)p Fj(any)572 3639 y Fl(Illegal.)60 b(IDLE)30
+-b(state)i(only)d(exists)h(after)h Fj(BZ2_bzCompressEnd)26
+-b Fl(or)572 3743 y(b)s(efore)k Fj(BZ2_bzCompressInit)p
+-Fl(.)572 3847 y(Return)f(v)-5 b(alue)30 b(=)g Fj(BZ_SEQUENCE_ERROR)390
+-4054 y Fl(R)m(UNNING/)p Fj(BZ_RUN)572 4158 y Fl(Compress)f(from)h
+-Fj(next_in)f Fl(to)i Fj(next_out)d Fl(as)i(m)m(uc)m(h)h(as)f(p)s
+-(ossible.)572 4262 y(Next)h(state)h(=)e(R)m(UNNING)572
+-4366 y(Return)f(v)-5 b(alue)30 b(=)g Fj(BZ_RUN_OK)390
+-4573 y Fl(R)m(UNNING/)p Fj(BZ_FLUSH)572 4677 y Fl(Remem)m(b)s(er)g
+-(curren)m(t)g(v)-5 b(alue)30 b(of)g Fj(next_in)p Fl(.)59
+-b(Compress)30 b(from)g Fj(next_in)572 4781 y Fl(to)h
+-Fj(next_out)d Fl(as)j(m)m(uc)m(h)f(as)h(p)s(ossible,)d(but)i(do)g(not)g
+-(accept)i(an)m(y)f(more)f(input.)572 4885 y(Next)h(state)h(=)e
+-(FLUSHING)572 4988 y(Return)f(v)-5 b(alue)30 b(=)g Fj(BZ_FLUSH_OK)390
+-5196 y Fl(R)m(UNNING/)p Fj(BZ_FINISH)572 5300 y Fl(Remem)m(b)s(er)g
+-(curren)m(t)g(v)-5 b(alue)30 b(of)g Fj(next_in)p Fl(.)59
+-b(Compress)30 b(from)g Fj(next_in)p eop
+-%%Page: 16 17
+-16 16 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(16)572 299 y(to)31 b Fj(next_out)d
+-Fl(as)j(m)m(uc)m(h)f(as)h(p)s(ossible,)d(but)i(do)g(not)g(accept)i(an)m
+-(y)f(more)f(input.)572 403 y(Next)h(state)h(=)e(FINISHING)572
+-506 y(Return)f(v)-5 b(alue)30 b(=)g Fj(BZ_FINISH_OK)390
+-714 y Fl(FLUSHING/)p Fj(BZ_FLUSH)572 818 y Fl(Compress)f(from)h
+-Fj(next_in)f Fl(to)i Fj(next_out)d Fl(as)i(m)m(uc)m(h)h(as)f(p)s
+-(ossible,)572 922 y(but)f(do)i(not)f(accept)i(an)m(y)f(more)f(input.)
+-572 1025 y(If)g(all)f(the)i(existing)e(input)f(has)i(b)s(een)g(used)g
+-(up)f(and)h(all)f(compressed)572 1129 y(output)h(has)g(b)s(een)g(remo)m
+-(v)m(ed)663 1233 y(Next)h(state)h(=)e(R)m(UNNING;)i(Return)d(v)-5
+-b(alue)30 b(=)g Fj(BZ_RUN_OK)572 1337 y Fl(else)663 1440
+-y(Next)h(state)h(=)e(FLUSHING;)h(Return)e(v)-5 b(alue)30
+-b(=)g Fj(BZ_FLUSH_OK)390 1648 y Fl(FLUSHING/other)572
+-1752 y(Illegal.)572 1856 y(Return)f(v)-5 b(alue)30 b(=)g
+-Fj(BZ_SEQUENCE_ERROR)390 2063 y Fl(FINISHING/)p Fj(BZ_FINISH)572
+-2167 y Fl(Compress)f(from)h Fj(next_in)f Fl(to)i Fj(next_out)d
+-Fl(as)i(m)m(uc)m(h)h(as)f(p)s(ossible,)572 2271 y(but)f(to)j(not)e
+-(accept)i(an)m(y)f(more)f(input.)572 2374 y(If)g(all)f(the)i(existing)e
+-(input)f(has)i(b)s(een)g(used)g(up)f(and)h(all)f(compressed)572
+-2478 y(output)h(has)g(b)s(een)g(remo)m(v)m(ed)663 2582
+-y(Next)h(state)h(=)e(IDLE;)g(Return)g(v)-5 b(alue)30
+-b(=)g Fj(BZ_STREAM_END)572 2686 y Fl(else)663 2790 y(Next)h(state)h(=)e
+-(FINISHING;)g(Return)g(v)-5 b(alue)30 b(=)g Fj(BZ_FINISHING)390
+-2997 y Fl(FINISHING/other)572 3101 y(Illegal.)572 3205
+-y(Return)f(v)-5 b(alue)30 b(=)g Fj(BZ_SEQUENCE_ERROR)150
+-3361 y Fl(That)24 b(still)f(lo)s(oks)g(complicated?)39
+-b(W)-8 b(ell,)25 b(fair)f(enough.)38 b(The)24 b(usual)f(sequence)i(of)f
+-(calls)g(for)g(compressing)150 3471 y(a)31 b(load)f(of)g(data)h(is:)225
+-3628 y Fi(\017)60 b Fl(Get)31 b(started)g(with)e Fj(BZ2_bzCompressInit)
+-p Fl(.)225 3774 y Fi(\017)60 b Fl(Sho)m(v)m(el)38 b(data)h(in)e(and)g
+-(shlurp)e(out)k(its)e(compressed)h(form)g(using)e(zero)j(or)f(more)h
+-(calls)e(of)h Fj(BZ2_)330 3884 y(bzCompress)28 b Fl(with)h(action)h(=)g
+-Fj(BZ_RUN)p Fl(.)225 4030 y Fi(\017)60 b Fl(Finish)23
+-b(up.)38 b(Rep)s(eatedly)25 b(call)f Fj(BZ2_bzCompress)e
+-Fl(with)i(action)h(=)g Fj(BZ_FINISH)p Fl(,)f(cop)m(ying)h(out)h(the)330
+-4139 y(compressed)k(output,)g(un)m(til)f Fj(BZ_STREAM_END)e
+-Fl(is)i(returned.)225 4285 y Fi(\017)60 b Fl(Close)30
+-b(up)f(and)h(go)h(home.)41 b(Call)29 b Fj(BZ2_bzCompressEnd)p
+-Fl(.)150 4478 y(If)23 b(the)h(data)h(y)m(ou)f(w)m(an)m(t)h(to)f
+-(compress)g(\014ts)f(in)m(to)h(y)m(our)g(input)e(bu\013er)h(all)f(at)j
+-(once,)h(y)m(ou)e(can)g(skip)f(the)h(calls)150 4588 y(of)37
+-b Fj(BZ2_bzCompress)26 b(\()k(...,)f(BZ_RUN)g(\))36 b
+-Fl(and)g(just)g(do)h(the)g Fj(BZ2_bzCompress)26 b(\()k(...,)f
+-(BZ_FINISH)150 4698 y(\))h Fl(calls.)150 4854 y(All)36
+-b(required)g(memory)h(is)f(allo)s(cated)i(b)m(y)f Fj
+-(BZ2_bzCompressInit)p Fl(.)56 b(The)37 b(compression)g(library)e(can)
+-150 4964 y(accept)g(an)m(y)f(data)h(at)g(all)d(\(ob)m(viously\).)51
+-b(So)34 b(y)m(ou)g(shouldn't)e(get)j(an)m(y)f(error)f(return)g(v)-5
+-b(alues)33 b(from)h(the)150 5074 y Fj(BZ2_bzCompress)29
+-b Fl(calls.)46 b(If)32 b(y)m(ou)h(do,)g(they)g(will)d(b)s(e)i
+-Fj(BZ_SEQUENCE_ERROR)p Fl(,)d(and)j(indicate)f(a)i(bug)f(in)150
+-5183 y(y)m(our)e(programming.)150 5340 y(T)-8 b(rivial)28
+-b(other)j(p)s(ossible)d(return)h(v)-5 b(alues:)p eop
+-%%Page: 17 18
+-17 17 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(17)572 299 y Fj(BZ_PARAM_ERROR)663
+-403 y Fl(if)29 b Fj(strm)g Fl(is)h Fj(NULL)p Fl(,)f(or)i
+-Fj(strm->s)d Fl(is)h Fj(NULL)150 652 y Ff(3.3.3)63 b
+-Fe(BZ2_bzCompressEnd)390 839 y Fj(int)47 b(BZ2_bzCompressEnd)c(\()k
+-(bz_stream)f(*strm)g(\);)150 996 y Fl(Releases)31 b(all)e(memory)h
+-(asso)s(ciated)h(with)e(a)i(compression)e(stream.)150
+-1153 y(P)m(ossible)g(return)h(v)-5 b(alues:)481 1304
+-y Fj(BZ_PARAM_ERROR)117 b Fl(if)30 b Fj(strm)f Fl(is)g
+-Fj(NULL)g Fl(or)i Fj(strm->s)d Fl(is)i Fj(NULL)481 1408
+-y(BZ_OK)120 b Fl(otherwise)150 1657 y Ff(3.3.4)63 b Fe
+-(BZ2_bzDecompressInit)390 1844 y Fj(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 Fl(Prepares)30 b(for)f(decompression.)40 b(As)29
+-b(with)g Fj(BZ2_bzCompressInit)p Fl(,)c(a)31 b Fj(bz_stream)c
+-Fl(record)j(should)e(b)s(e)150 2110 y(allo)s(cated)c(and)f(initialised)
+-e(b)s(efore)i(the)i(call.)38 b(Fields)22 b Fj(bzalloc)p
+-Fl(,)i Fj(bzfree)e Fl(and)i Fj(opaque)e Fl(should)g(b)s(e)h(set)i(if)
+-150 2220 y(a)h(custom)f(memory)g(allo)s(cator)g(is)g(required,)f(or)h
+-(made)h Fj(NULL)e Fl(for)h(the)g(normal)f Fj(malloc)p
+-Fl(/)p Fj(free)f Fl(routines.)150 2330 y(Up)s(on)h(return,)h(the)g(in)m
+-(ternal)f(state)i(will)c(ha)m(v)m(e)k(b)s(een)f(initialised,)d(and)i
+-Fj(total_in)f Fl(and)h Fj(total_out)f Fl(will)150 2439
+-y(b)s(e)30 b(zero.)150 2596 y(F)-8 b(or)31 b(the)g(meaning)e(of)i
+-(parameter)g Fj(verbosity)p Fl(,)d(see)j Fj(BZ2_bzCompressInit)p
+-Fl(.)150 2753 y(If)e Fj(small)e Fl(is)h(nonzero,)i(the)f(library)e
+-(will)f(use)j(an)g(alternativ)m(e)h(decompression)e(algorithm)g(whic)m
+-(h)f(uses)150 2862 y(less)c(memory)g(but)g(at)h(the)g(cost)h(of)e
+-(decompressing)g(more)g(slo)m(wly)g(\(roughly)f(sp)s(eaking,)i(half)f
+-(the)h(sp)s(eed,)150 2972 y(but)34 b(the)i(maxim)m(um)d(memory)i
+-(requiremen)m(t)g(drops)e(to)j(around)e(2300k\).)57 b(See)35
+-b(Chapter)g(2)g(for)g(more)150 3082 y(information)29
+-b(on)h(memory)g(managemen)m(t.)150 3238 y(Note)40 b(that)f(the)f(amoun)
+-m(t)h(of)g(memory)f(needed)g(to)i(decompress)e(a)h(stream)f(cannot)h(b)
+-s(e)f(determined)150 3348 y(un)m(til)j(the)h(stream's)h(header)f(has)g
+-(b)s(een)g(read,)j(so)e(ev)m(en)g(if)e Fj(BZ2_bzDecompressInit)c
+-Fl(succeeds,)46 b(a)150 3458 y(subsequen)m(t)30 b Fj(BZ2_bzDecompress)c
+-Fl(could)j(fail)g(with)g Fj(BZ_MEM_ERROR)p Fl(.)150 3614
+-y(P)m(ossible)g(return)h(v)-5 b(alues:)572 3765 y Fj(BZ_CONFIG_ERROR)
+-663 3869 y Fl(if)29 b(the)i(library)d(has)i(b)s(een)f(mis-compiled)572
+-3973 y Fj(BZ_PARAM_ERROR)663 4077 y Fl(if)g Fj(\(small)46
+-b(!=)h(0)h(&&)f(small)f(!=)h(1\))663 4181 y Fl(or)30
+-b Fj(\(verbosity)45 b(<)j(0)f(||)g(verbosity)e(>)j(4\))572
+-4284 y(BZ_MEM_ERROR)663 4388 y Fl(if)29 b(insu\016cien)m(t)g(memory)h
+-(is)f(a)m(v)-5 b(ailable)150 4545 y(Allo)m(w)m(able)30
+-b(next)g(actions:)572 4696 y Fj(BZ2_bzDecompress)663
+-4800 y Fl(if)f Fj(BZ_OK)g Fl(w)m(as)i(returned)572 4904
+-y(no)f(sp)s(eci\014c)f(action)i(required)e(in)g(case)i(of)g(error)150
+-5153 y Ff(3.3.5)63 b Fe(BZ2_bzDecompress)390 5340 y Fj(int)47
+-b(BZ2_bzDecompress)c(\()48 b(bz_stream)d(*strm)h(\);)p
+-eop
+-%%Page: 18 19
+-18 18 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(18)150 299 y(Pro)m(vides)24
+-b(more)g(input)f(and/out)h(output)g(bu\013er)g(space)h(for)f(the)g
+-(library)-8 b(.)37 b(The)24 b(caller)g(main)m(tains)f(input)150
+-408 y(and)30 b(output)g(bu\013ers,)f(and)h(uses)g Fj(BZ2_bzDecompress)c
+-Fl(to)31 b(transfer)f(data)h(b)s(et)m(w)m(een)g(them.)150
+-565 y(Before)g(eac)m(h)g(call)f(to)g Fj(BZ2_bzDecompress)p
+-Fl(,)c Fj(next_in)i Fl(should)h(p)s(oin)m(t)g(at)h(the)h(compressed)e
+-(data,)j(and)150 675 y Fj(avail_in)h Fl(should)h(indicate)h(ho)m(w)h
+-(man)m(y)f(b)m(ytes)i(the)e(library)f(ma)m(y)i(read.)56
+-b Fj(BZ2_bzDecompress)32 b Fl(up-)150 784 y(dates)f Fj(next_in)p
+-Fl(,)e Fj(avail_in)f Fl(and)h Fj(total_in)g Fl(to)i(re\015ect)g(the)f
+-(n)m(um)m(b)s(er)f(of)i(b)m(ytes)g(it)f(has)g(read.)150
+-941 y(Similarly)-8 b(,)37 b Fj(next_out)f Fl(should)g(p)s(oin)m(t)i(to)
+-g(a)h(bu\013er)e(in)g(whic)m(h)g(the)i(uncompressed)e(output)g(is)h(to)
+-h(b)s(e)150 1051 y(placed,)d(with)e Fj(avail_out)f Fl(indicating)g(ho)m
+-(w)i(m)m(uc)m(h)g(output)g(space)h(is)e(a)m(v)-5 b(ailable.)55
+-b Fj(BZ2_bzCompress)150 1160 y Fl(up)s(dates)29 b Fj(next_out)p
+-Fl(,)g Fj(avail_out)f Fl(and)h Fj(total_out)f Fl(to)j(re\015ect)g(the)g
+-(n)m(um)m(b)s(er)e(of)h(b)m(ytes)h(output.)150 1317 y(Y)-8
+-b(ou)40 b(ma)m(y)g(pro)m(vide)e(and)h(remo)m(v)m(e)i(as)f(little)e(or)h
+-(as)h(m)m(uc)m(h)f(data)h(as)g(y)m(ou)f(lik)m(e)g(on)g(eac)m(h)i(call)e
+-(of)g Fj(BZ2_)150 1427 y(bzDecompress)p Fl(.)e(In)27
+-b(the)i(limit,)d(it)i(is)f(acceptable)j(to)f(supply)d(and)h(remo)m(v)m
+-(e)j(data)f(one)f(b)m(yte)h(at)g(a)g(time,)150 1537 y(although)f(this)f
+-(w)m(ould)g(b)s(e)h(terribly)e(ine\016cien)m(t.)39 b(Y)-8
+-b(ou)29 b(should)e(alw)m(a)m(ys)h(ensure)g(that)h(at)g(least)g(one)f(b)
+-m(yte)150 1646 y(of)j(output)f(space)g(is)g(a)m(v)-5
+-b(ailable)30 b(at)h(eac)m(h)g(call.)150 1803 y(Use)g(of)f
+-Fj(BZ2_bzDecompress)c Fl(is)k(simpler)e(than)i Fj(BZ2_bzCompress)p
+-Fl(.)150 1960 y(Y)-8 b(ou)31 b(should)d(pro)m(vide)h(input)f(and)i
+-(remo)m(v)m(e)i(output)d(as)i(describ)s(ed)d(ab)s(o)m(v)m(e,)k(and)d
+-(rep)s(eatedly)h(call)f Fj(BZ2_)150 2069 y(bzDecompress)35
+-b Fl(un)m(til)i Fj(BZ_STREAM_END)e Fl(is)j(returned.)64
+-b(App)s(earance)39 b(of)g Fj(BZ_STREAM_END)c Fl(denotes)150
+-2179 y(that)47 b Fj(BZ2_bzDecompress)42 b Fl(has)k(detected)h(the)f
+-(logical)g(end)g(of)g(the)h(compressed)e(stream.)89 b
+-Fj(BZ2_)150 2289 y(bzDecompress)28 b Fl(will)g(not)j(pro)s(duce)f
+-Fj(BZ_STREAM_END)d Fl(un)m(til)j(all)f(output)i(data)h(has)e(b)s(een)h
+-(placed)f(in)m(to)150 2398 y(the)36 b(output)g(bu\013er,)h(so)g(once)g
+-Fj(BZ_STREAM_END)32 b Fl(app)s(ears,)38 b(y)m(ou)e(are)h(guaran)m(teed)
+-g(to)g(ha)m(v)m(e)h(a)m(v)-5 b(ailable)150 2508 y(all)29
+-b(the)i(decompressed)f(output,)g(and)g Fj(BZ2_bzDecompressEnd)25
+-b Fl(can)31 b(safely)f(b)s(e)f(called.)150 2665 y(If)40
+-b(case)h(of)f(an)h(error)e(return)h(v)-5 b(alue,)42 b(y)m(ou)f(should)d
+-(call)h Fj(BZ2_bzDecompressEnd)c Fl(to)41 b(clean)f(up)g(and)150
+-2774 y(release)31 b(memory)-8 b(.)150 2931 y(P)m(ossible)29
+-b(return)h(v)-5 b(alues:)572 3082 y Fj(BZ_PARAM_ERROR)663
+-3186 y Fl(if)29 b Fj(strm)g Fl(is)h Fj(NULL)f Fl(or)h
+-Fj(strm->s)f Fl(is)g Fj(NULL)663 3290 y Fl(or)h Fj(strm->avail_out)44
+-b(<)j(1)572 3393 y(BZ_DATA_ERROR)663 3497 y Fl(if)29
+-b(a)i(data)g(in)m(tegrit)m(y)f(error)g(is)g(detected)h(in)e(the)i
+-(compressed)f(stream)572 3601 y Fj(BZ_DATA_ERROR_MAGIC)663
+-3705 y Fl(if)f(the)i(compressed)f(stream)g(do)s(esn't)h(b)s(egin)e
+-(with)g(the)h(righ)m(t)g(magic)h(b)m(ytes)572 3808 y
+-Fj(BZ_MEM_ERROR)663 3912 y Fl(if)e(there)i(w)m(asn't)f(enough)h(memory)
+-f(a)m(v)-5 b(ailable)572 4016 y Fj(BZ_STREAM_END)663
+-4120 y Fl(if)29 b(the)i(logical)e(end)h(of)h(the)f(data)h(stream)g(w)m
+-(as)g(detected)g(and)f(all)663 4224 y(output)g(in)f(has)h(b)s(een)g
+-(consumed,)f(eg)j Fj(s->avail_out)44 b(>)k(0)572 4327
+-y(BZ_OK)663 4431 y Fl(otherwise)150 4588 y(Allo)m(w)m(able)30
+-b(next)g(actions:)572 4739 y Fj(BZ2_bzDecompress)663
+-4843 y Fl(if)f Fj(BZ_OK)g Fl(w)m(as)i(returned)572 4946
+-y Fj(BZ2_bzDecompressEnd)663 5050 y Fl(otherwise)p eop
+-%%Page: 19 20
+-19 19 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(19)150 299 y Ff(3.3.6)63
+-b Fe(BZ2_bzDecompressEnd)390 486 y Fj(int)47 b(BZ2_bzDecompressEnd)42
+-b(\()48 b(bz_stream)d(*strm)i(\);)150 643 y Fl(Releases)31
+-b(all)e(memory)h(asso)s(ciated)h(with)e(a)i(decompression)e(stream.)150
+-799 y(P)m(ossible)g(return)h(v)-5 b(alues:)572 950 y
+-Fj(BZ_PARAM_ERROR)663 1054 y Fl(if)29 b Fj(strm)g Fl(is)h
+-Fj(NULL)f Fl(or)h Fj(strm->s)f Fl(is)g Fj(NULL)572 1158
+-y(BZ_OK)663 1262 y Fl(otherwise)150 1419 y(Allo)m(w)m(able)h(next)g
+-(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 Fl(This)35 b(in)m(terface)j(pro)m(vides)d
+-(functions)h(for)g(reading)g(and)h(writing)e Fj(bzip2)g
+-Fl(format)i(\014les.)59 b(First,)39 b(some)150 2159 y(general)30
+-b(p)s(oin)m(ts.)225 2316 y Fi(\017)60 b Fl(All)35 b(of)h(the)g
+-(functions)e(tak)m(e)k(an)e Fj(int*)f Fl(\014rst)g(argumen)m(t,)j
+-Fj(bzerror)p Fl(.)56 b(After)36 b(eac)m(h)h(call,)g Fj(bzerror)330
+-2426 y Fl(should)23 b(b)s(e)i(consulted)g(\014rst)g(to)h(determine)e
+-(the)i(outcome)h(of)e(the)h(call.)38 b(If)25 b Fj(bzerror)f
+-Fl(is)g Fj(BZ_OK)p Fl(,)i(the)330 2535 y(call)35 b(completed)g
+-(successfully)-8 b(,)36 b(and)f(only)g(then)g(should)f(the)h(return)g
+-(v)-5 b(alue)35 b(of)h(the)f(function)g(\(if)330 2645
+-y(an)m(y\))30 b(b)s(e)f(consulted.)39 b(If)29 b Fj(bzerror)e
+-Fl(is)h Fj(BZ_IO_ERROR)p Fl(,)f(there)i(w)m(as)h(an)f(error)g
+-(reading/writing)e(the)330 2754 y(underlying)32 b(compressed)j(\014le,)
+-h(and)f(y)m(ou)h(should)d(then)i(consult)g Fj(errno)p
+-Fl(/)p Fj(perror)e Fl(to)j(determine)330 2864 y(the)i(cause)g(of)g(the)
+-g(di\016cult)m(y)-8 b(.)61 b Fj(bzerror)36 b Fl(ma)m(y)i(also)g(b)s(e)f
+-(set)h(to)g(v)-5 b(arious)37 b(other)h(v)-5 b(alues;)41
+-b(precise)330 2974 y(details)29 b(are)i(giv)m(en)g(on)f(a)h(p)s
+-(er-function)d(basis)h(b)s(elo)m(w.)225 3111 y Fi(\017)60
+-b Fl(If)40 b Fj(bzerror)f Fl(indicates)g(an)i(error)f(\(ie,)j(an)m
+-(ything)d(except)h Fj(BZ_OK)f Fl(and)g Fj(BZ_STREAM_END)p
+-Fl(\),)g(y)m(ou)330 3220 y(should)56 b(immediately)h(call)g
+-Fj(BZ2_bzReadClose)e Fl(\(or)j Fj(BZ2_bzWriteClose)p
+-Fl(,)j(dep)s(ending)56 b(on)330 3330 y(whether)50 b(y)m(ou)g(are)h
+-(attempting)g(to)g(read)f(or)g(to)i(write\))d(to)j(free)e(up)f(all)h
+-(resources)g(asso)s(ci-)330 3439 y(ated)33 b(with)e(the)i(stream.)47
+-b(Once)32 b(an)h(error)f(has)g(b)s(een)g(indicated,)f(b)s(eha)m(viour)g
+-(of)i(all)e(calls)h(except)330 3549 y Fj(BZ2_bzReadClose)46
+-b Fl(\()p Fj(BZ2_bzWriteClose)p Fl(\))h(is)j(unde\014ned.)99
+-b(The)50 b(implication)e(is)i(that)h(\(1\))330 3659 y
+-Fj(bzerror)44 b Fl(should)g(b)s(e)h(c)m(hec)m(k)m(ed)j(after)e(eac)m(h)
+-h(call,)i(and)c(\(2\))i(if)e Fj(bzerror)f Fl(indicates)g(an)i(error,)
+-330 3768 y Fj(BZ2_bzReadClose)26 b Fl(\()p Fj(BZ2_bzWriteClose)p
+-Fl(\))h(should)h(then)i(b)s(e)g(called)g(to)h(clean)f(up.)225
+-3905 y Fi(\017)60 b Fl(The)33 b Fj(FILE*)f Fl(argumen)m(ts)h(passed)g
+-(to)h Fj(BZ2_bzReadOpen)p Fl(/)p Fj(BZ2_bzWriteOp)o(en)27
+-b Fl(should)32 b(b)s(e)g(set)i(to)330 4015 y(binary)23
+-b(mo)s(de.)38 b(Most)26 b(Unix)d(systems)i(will)d(do)i(this)g(b)m(y)g
+-(default,)i(but)e(other)g(platforms,)h(including)330
+-4124 y(Windo)m(ws)20 b(and)g(Mac,)k(will)19 b(not.)38
+-b(If)20 b(y)m(ou)h(omit)g(this,)h(y)m(ou)f(ma)m(y)h(encoun)m(ter)f
+-(problems)e(when)h(mo)m(ving)330 4234 y(co)s(de)31 b(to)g(new)f
+-(platforms.)225 4371 y Fi(\017)60 b Fl(Memory)23 b(allo)s(cation)f
+-(requests)h(are)g(handled)e(b)m(y)i Fj(malloc)p Fl(/)p
+-Fj(free)p Fl(.)36 b(A)m(t)23 b(presen)m(t)g(there)g(is)f(no)h(facilit)m
+-(y)330 4481 y(for)40 b(user-de\014ned)e(memory)i(allo)s(cators)g(in)f
+-(the)h(\014le)g(I/O)g(functions)e(\(could)i(easily)f(b)s(e)g(added,)330
+-4590 y(though\).)150 4842 y Ff(3.4.1)63 b Fe(BZ2_bzReadOpen)533
+-5029 y Fj(typedef)46 b(void)h(BZFILE;)533 5236 y(BZFILE)f
+-(*BZ2_bzReadOpen)e(\()j(int)g(*bzerror,)f(FILE)g(*f,)1726
+-5340 y(int)h(small,)f(int)h(verbosity,)p eop
+-%%Page: 20 21
+-20 20 bop 150 -116 a Fl(Chapter)30 b(3:)h(Programming)e(with)g
+-Fj(libbzip2)1891 b Fl(20)1726 299 y Fj(void)47 b(*unused,)f(int)g
+-(nUnused)g(\);)150 456 y Fl(Prepare)29 b(to)g(read)g(compressed)f(data)
+-i(from)e(\014le)g(handle)f Fj(f)p Fl(.)40 b Fj(f)29 b
+-Fl(should)d(refer)j(to)h(a)f(\014le)f(whic)m(h)f(has)i(b)s(een)150
+-565 y(op)s(ened)h(for)h(reading,)f(and)h(for)f(whic)m(h)g(the)h(error)g
+-(indicator)e(\()p Fj(ferror\(f\))p Fl(\)is)f(not)k(set.)42
+-b(If)31 b Fj(small)e Fl(is)h(1,)150 675 y(the)h(library)d(will)f(try)j
+-(to)i(decompress)e(using)f(less)g(memory)-8 b(,)31 b(at)g(the)g(exp)s
+-(ense)f(of)g(sp)s(eed.)150 832 y(F)-8 b(or)39 b(reasons)f(explained)f
+-(b)s(elo)m(w,)j Fj(BZ2_bzRead)35 b Fl(will)h(decompress)i(the)g
+-Fj(nUnused)e Fl(b)m(ytes)j(starting)f(at)150 941 y Fj(unused)p
+-Fl(,)k(b)s(efore)e(starting)h(to)g(read)g(from)f(the)h(\014le)f
+-Fj(f)p Fl(.)71 b(A)m(t)42 b(most)f Fj(BZ_MAX_UNUSED)c
+-Fl(b)m(ytes)k(ma)m(y)h(b)s(e)150 1051 y(supplied)32 b(lik)m(e)k(this.)
+-55 b(If)36 b(this)e(facilit)m(y)h(is)g(not)h(required,)g(y)m(ou)g
+-(should)e(pass)h Fj(NULL)g Fl(and)g Fj(0)g Fl(for)h Fj(unused)150
+-1160 y Fl(and)30 b(n)p Fj(Unused)e Fl(resp)s(ectiv)m(ely)-8
+-b(.)150 1317 y(F)g(or)31 b(the)g(meaning)e(of)i(parameters)g
+-Fj(small)e Fl(and)g Fj(verbosity)p Fl(,)f(see)j Fj
+-(BZ2_bzDecompressInit)p Fl(.)150 1474 y(The)k(amoun)m(t)g(of)g(memory)g
+-(needed)g(to)g(decompress)g(a)h(\014le)e(cannot)h(b)s(e)g(determined)e
+-(un)m(til)h(the)h(\014le's)150 1584 y(header)22 b(has)f(b)s(een)g
+-(read.)38 b(So)22 b(it)f(is)g(p)s(ossible)e(that)k Fj(BZ2_bzReadOpen)17
+-b Fl(returns)k Fj(BZ_OK)f Fl(but)h(a)i(subsequen)m(t)150
+-1693 y(call)30 b(of)g Fj(BZ2_bzRead)e Fl(will)f(return)j
+-Fj(BZ_MEM_ERROR)p Fl(.)150 1850 y(P)m(ossible)f(assignmen)m(ts)h(to)h
+-Fj(bzerror)p Fl(:)572 2001 y Fj(BZ_CONFIG_ERROR)663 2105
+-y Fl(if)e(the)i(library)d(has)i(b)s(een)f(mis-compiled)572
+-2209 y Fj(BZ_PARAM_ERROR)663 2313 y Fl(if)g Fj(f)h Fl(is)g
+-Fj(NULL)663 2416 y Fl(or)g Fj(small)f Fl(is)g(neither)h
+-Fj(0)g Fl(nor)g Fj(1)663 2520 y Fl(or)g Fj(\(unused)46
+-b(==)h(NULL)g(&&)g(nUnused)f(!=)h(0\))663 2624 y Fl(or)30
+-b Fj(\(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
+-Fl(if)29 b Fj(ferror\(f\))f Fl(is)h(nonzero)572 2935
+-y Fj(BZ_MEM_ERROR)663 3039 y Fl(if)g(insu\016cien)m(t)g(memory)h(is)f
+-(a)m(v)-5 b(ailable)572 3143 y Fj(BZ_OK)663 3247 y Fl(otherwise.)150
+-3403 y(P)m(ossible)29 b(return)h(v)-5 b(alues:)572 3554
+-y(P)m(oin)m(ter)31 b(to)g(an)f(abstract)h Fj(BZFILE)663
+-3658 y Fl(if)e Fj(bzerror)f Fl(is)i Fj(BZ_OK)572 3762
+-y(NULL)663 3866 y Fl(otherwise)150 4023 y(Allo)m(w)m(able)g(next)g
+-(actions:)572 4174 y Fj(BZ2_bzRead)663 4277 y Fl(if)f
+-Fj(bzerror)f Fl(is)i Fj(BZ_OK)572 4381 y(BZ2_bzClose)663
+-4485 y Fl(otherwise)150 4887 y Ff(3.4.2)63 b Fe(BZ2_bzRead)533
+-5074 y Fj(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 Fl(Reads)35
+-b(up)f(to)h Fj(len)f Fl(\(uncompressed\))h(b)m(ytes)g(from)f(the)h
+-(compressed)g(\014le)f Fj(b)g Fl(in)m(to)h(the)g(bu\013er)f
+-Fj(buf)p Fl(.)53 b(If)150 5340 y(the)30 b(read)f(w)m(as)h(successful,)f
+-Fj(bzerror)e Fl(is)i(set)h(to)g Fj(BZ_OK)e Fl(and)h(the)h(n)m(um)m(b)s
+-(er)e(of)i(b)m(ytes)g(read)f(is)g(returned.)p eop
+-%%Page: 21 22
+-21 21 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(21)150 299 y(If)33 b(the)g(logical)g
+-(end-of-stream)h(w)m(as)g(detected,)i Fj(bzerror)31 b
+-Fl(will)g(b)s(e)h(set)i(to)g Fj(BZ_STREAM_END)p Fl(,)d(and)i(the)150
+-408 y(n)m(um)m(b)s(er)c(of)i(b)m(ytes)f(read)h(is)e(returned.)40
+-b(All)29 b(other)h Fj(bzerror)f Fl(v)-5 b(alues)29 b(denote)i(an)g
+-(error.)150 565 y Fj(BZ2_bzRead)37 b Fl(will)f(supply)h
+-Fj(len)i Fl(b)m(ytes,)j(unless)c(the)i(logical)f(stream)h(end)e(is)h
+-(detected)i(or)e(an)g(error)150 675 y(o)s(ccurs.)75 b(Because)43
+-b(of)f(this,)i(it)d(is)g(p)s(ossible)e(to)k(detect)g(the)f(stream)g
+-(end)f(b)m(y)h(observing)f(when)g(the)150 784 y(n)m(um)m(b)s(er)29
+-b(of)h(b)m(ytes)g(returned)f(is)g(less)g(than)h(the)g(n)m(um)m(b)s(er)f
+-(requested.)40 b(Nev)m(ertheless,)31 b(this)e(is)g(regarded)150
+-894 y(as)38 b(inadvisable;)g(y)m(ou)g(should)d(instead)i(c)m(hec)m(k)i
+-Fj(bzerror)d Fl(after)i(ev)m(ery)g(call)e(and)h(w)m(atc)m(h)i(out)f
+-(for)f Fj(BZ_)150 1004 y(STREAM_END)p Fl(.)150 1160 y(In)m(ternally)-8
+-b(,)47 b Fj(BZ2_bzRead)41 b Fl(copies)j(data)g(from)g(the)g(compressed)
+-g(\014le)f(in)f(c)m(h)m(unks)i(of)g(size)g Fj(BZ_MAX_)150
+-1270 y(UNUSED)31 b Fl(b)m(ytes)i(b)s(efore)f(decompressing)f(it.)47
+-b(If)32 b(the)h(\014le)e(con)m(tains)i(more)g(b)m(ytes)g(than)f
+-(strictly)f(needed)150 1380 y(to)48 b(reac)m(h)f(the)g(logical)f
+-(end-of-stream,)52 b Fj(BZ2_bzRead)44 b Fl(will)g(almost)j(certainly)f
+-(read)h(some)g(of)g(the)150 1489 y(trailing)c(data)j(b)s(efore)e
+-(signalling)f Fj(BZ_SEQUENCE_END)p Fl(.)80 b(T)-8 b(o)46
+-b(collect)f(the)g(read)g(but)g(un)m(used)e(data)150 1599
+-y(once)29 b Fj(BZ_SEQUENCE_END)24 b Fl(has)k(app)s(eared,)g(call)f
+-Fj(BZ2_bzReadGetUnused)c Fl(immediately)j(b)s(efore)i
+-Fj(BZ2_)150 1708 y(bzReadClose)p Fl(.)150 1865 y(P)m(ossible)h
+-(assignmen)m(ts)h(to)h Fj(bzerror)p Fl(:)572 2016 y Fj(BZ_PARAM_ERROR)
+-663 2120 y Fl(if)e Fj(b)h Fl(is)g Fj(NULL)f Fl(or)h Fj(buf)g
+-Fl(is)f Fj(NULL)g Fl(or)i Fj(len)46 b(<)i(0)572 2224
+-y(BZ_SEQUENCE_ERROR)663 2328 y Fl(if)29 b Fj(b)h Fl(w)m(as)h(op)s(ened)
+-e(with)h Fj(BZ2_bzWriteOpen)572 2431 y(BZ_IO_ERROR)663
+-2535 y Fl(if)f(there)i(is)e(an)h(error)g(reading)g(from)g(the)g
+-(compressed)g(\014le)572 2639 y Fj(BZ_UNEXPECTED_EOF)663
+-2743 y Fl(if)f(the)i(compressed)f(\014le)f(ended)h(b)s(efore)g(the)g
+-(logical)g(end-of-stream)h(w)m(as)g(detected)572 2847
+-y Fj(BZ_DATA_ERROR)663 2950 y Fl(if)e(a)i(data)g(in)m(tegrit)m(y)f
+-(error)g(w)m(as)h(detected)h(in)d(the)h(compressed)g(stream)572
+-3054 y Fj(BZ_DATA_ERROR_MAGIC)663 3158 y Fl(if)f(the)i(stream)f(do)s
+-(es)g(not)h(b)s(egin)e(with)g(the)i(requisite)e(header)h(b)m(ytes)h
+-(\(ie,)f(is)g(not)663 3262 y(a)g Fj(bzip2)f Fl(data)i(\014le\).)61
+-b(This)28 b(is)i(really)f(a)i(sp)s(ecial)e(case)i(of)g
+-Fj(BZ_DATA_ERROR)p Fl(.)572 3365 y Fj(BZ_MEM_ERROR)663
+-3469 y Fl(if)e(insu\016cien)m(t)g(memory)h(w)m(as)h(a)m(v)-5
+-b(ailable)572 3573 y Fj(BZ_STREAM_END)663 3677 y Fl(if)29
+-b(the)i(logical)e(end)h(of)h(stream)f(w)m(as)h(detected.)572
+-3781 y Fj(BZ_OK)663 3884 y Fl(otherwise.)150 4041 y(P)m(ossible)e
+-(return)h(v)-5 b(alues:)572 4192 y(n)m(um)m(b)s(er)29
+-b(of)h(b)m(ytes)h(read)663 4296 y(if)e Fj(bzerror)f Fl(is)i
+-Fj(BZ_OK)f Fl(or)h Fj(BZ_STREAM_END)572 4400 y Fl(unde\014ned)663
+-4503 y(otherwise)150 4660 y(Allo)m(w)m(able)g(next)g(actions:)572
+-4811 y(collect)h(data)g(from)f Fj(buf)p Fl(,)f(then)h
+-Fj(BZ2_bzRead)e Fl(or)i Fj(BZ2_bzReadClose)663 4915 y
+-Fl(if)f Fj(bzerror)f Fl(is)i Fj(BZ_OK)572 5019 y Fl(collect)h(data)g
+-(from)f Fj(buf)p Fl(,)f(then)h Fj(BZ2_bzReadClose)d Fl(or)j
+-Fj(BZ2_bzReadGetUnused)663 5123 y Fl(if)f Fj(bzerror)f
+-Fl(is)i Fj(BZ_SEQUENCE_END)572 5226 y(BZ2_bzReadClose)663
+-5330 y Fl(otherwise)p eop
+-%%Page: 22 23
+-22 22 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(22)150 299 y Ff(3.4.3)63
+-b Fe(BZ2_bzReadGetUnused)533 486 y Fj(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 Fl(Returns)36 b(data)i(whic)m(h)d(w)m
+-(as)j(read)f(from)f(the)h(compressed)g(\014le)f(but)g(w)m(as)h(not)h
+-(needed)e(to)i(get)g(to)g(the)150 856 y(logical)k(end-of-stream.)78
+-b Fj(*unused)41 b Fl(is)h(set)h(to)g(the)g(address)f(of)g(the)h(data,)k
+-(and)42 b Fj(*nUnused)e Fl(to)k(the)150 965 y(n)m(um)m(b)s(er)29
+-b(of)i(b)m(ytes.)41 b Fj(*nUnused)28 b Fl(will)g(b)s(e)h(set)i(to)g(a)g
+-(v)-5 b(alue)30 b(b)s(et)m(w)m(een)h Fj(0)f Fl(and)g
+-Fj(BZ_MAX_UNUSED)d Fl(inclusiv)m(e.)150 1122 y(This)d(function)h(ma)m
+-(y)h(only)g(b)s(e)f(called)g(once)i Fj(BZ2_bzRead)c Fl(has)j(signalled)
+-e Fj(BZ_STREAM_END)e Fl(but)j(b)s(efore)150 1232 y Fj(BZ2_bzReadClose)p
+-Fl(.)150 1389 y(P)m(ossible)k(assignmen)m(ts)h(to)h Fj(bzerror)p
+-Fl(:)572 1540 y Fj(BZ_PARAM_ERROR)663 1644 y Fl(if)e
+-Fj(b)h Fl(is)g Fj(NULL)663 1747 y Fl(or)g Fj(unused)f
+-Fl(is)g Fj(NULL)g Fl(or)i Fj(nUnused)d Fl(is)i Fj(NULL)572
+-1851 y(BZ_SEQUENCE_ERROR)663 1955 y Fl(if)f Fj(BZ_STREAM_END)e
+-Fl(has)j(not)h(b)s(een)e(signalled)663 2059 y(or)h(if)f
+-Fj(b)h Fl(w)m(as)h(op)s(ened)f(with)f Fj(BZ2_bzWriteOpen)542
+-2162 y(BZ_OK)663 2266 y Fl(otherwise)150 2423 y(Allo)m(w)m(able)h(next)
+-g(actions:)572 2574 y Fj(BZ2_bzReadClose)150 2882 y Ff(3.4.4)63
+-b Fe(BZ2_bzReadClose)533 3068 y Fj(void)47 b(BZ2_bzReadClose)c(\()48
+-b(int)f(*bzerror,)e(BZFILE)h(*b)h(\);)150 3225 y Fl(Releases)36
+-b(all)e(memory)h(p)s(ertaining)e(to)i(the)h(compressed)f(\014le)f
+-Fj(b)p Fl(.)54 b Fj(BZ2_bzReadClose)31 b Fl(do)s(es)k(not)h(call)150
+-3335 y Fj(fclose)c Fl(on)h(the)h(underlying)d(\014le)h(handle,)h(so)h
+-(y)m(ou)g(should)e(do)h(that)h(y)m(ourself)f(if)g(appropriate.)49
+-b Fj(BZ2_)150 3445 y(bzReadClose)27 b Fl(should)i(b)s(e)g(called)h(to)h
+-(clean)f(up)g(after)h(all)e(error)h(situations.)150 3601
+-y(P)m(ossible)f(assignmen)m(ts)h(to)h Fj(bzerror)p Fl(:)572
+-3752 y Fj(BZ_SEQUENCE_ERROR)663 3856 y Fl(if)e Fj(b)h
+-Fl(w)m(as)h(op)s(ened)e(with)h Fj(BZ2_bzOpenWrite)572
+-3960 y(BZ_OK)663 4064 y Fl(otherwise)150 4221 y(Allo)m(w)m(able)g(next)
+-g(actions:)572 4372 y(none)150 4679 y Ff(3.4.5)63 b Fe(BZ2_bzWriteOpen)
+-533 4866 y Fj(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 Fl(Prepare)33
+-b(to)g(write)f(compressed)h(data)h(to)f(\014le)f(handle)g
+-Fj(f)p Fl(.)47 b Fj(f)33 b Fl(should)e(refer)i(to)g(a)g(\014le)f(whic)m
+-(h)g(has)h(b)s(een)150 5340 y(op)s(ened)d(for)g(writing,)e(and)i(for)g
+-(whic)m(h)f(the)i(error)f(indicator)f(\()p Fj(ferror\(f\))p
+-Fl(\)is)f(not)i(set.)p eop
+-%%Page: 23 24
+-23 23 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(23)150 299 y(F)-8 b(or)31
+-b(the)g(meaning)e(of)i(parameters)g Fj(blockSize100k)p
+-Fl(,)c Fj(verbosity)g Fl(and)j Fj(workFactor)p Fl(,)e(see)150
+-408 y Fj(BZ2_bzCompressInit)p Fl(.)150 565 y(All)d(required)f(memory)i
+-(is)f(allo)s(cated)i(at)g(this)e(stage,)j(so)f(if)e(the)h(call)g
+-(completes)g(successfully)-8 b(,)26 b Fj(BZ_MEM_)150
+-675 y(ERROR)j Fl(cannot)i(b)s(e)f(signalled)e(b)m(y)i(a)h(subsequen)m
+-(t)f(call)f(to)i Fj(BZ2_bzWrite)p Fl(.)150 832 y(P)m(ossible)e
+-(assignmen)m(ts)h(to)h Fj(bzerror)p Fl(:)572 983 y Fj(BZ_CONFIG_ERROR)
+-663 1087 y Fl(if)e(the)i(library)d(has)i(b)s(een)f(mis-compiled)572
+-1190 y Fj(BZ_PARAM_ERROR)663 1294 y Fl(if)g Fj(f)h Fl(is)g
+-Fj(NULL)663 1398 y Fl(or)g Fj(blockSize100k)44 b(<)k(1)30
+-b Fl(or)g Fj(blockSize100k)44 b(>)k(9)572 1502 y(BZ_IO_ERROR)663
+-1605 y Fl(if)29 b Fj(ferror\(f\))f Fl(is)h(nonzero)572
+-1709 y Fj(BZ_MEM_ERROR)663 1813 y Fl(if)g(insu\016cien)m(t)g(memory)h
+-(is)f(a)m(v)-5 b(ailable)572 1917 y Fj(BZ_OK)663 2021
+-y Fl(otherwise)150 2177 y(P)m(ossible)29 b(return)h(v)-5
+-b(alues:)572 2328 y(P)m(oin)m(ter)31 b(to)g(an)f(abstract)h
+-Fj(BZFILE)663 2432 y Fl(if)e Fj(bzerror)f Fl(is)i Fj(BZ_OK)572
+-2536 y(NULL)663 2640 y Fl(otherwise)150 2797 y(Allo)m(w)m(able)g(next)g
+-(actions:)572 2948 y Fj(BZ2_bzWrite)663 3051 y Fl(if)f
+-Fj(bzerror)f Fl(is)i Fj(BZ_OK)604 3155 y Fl(\(y)m(ou)25
+-b(could)e(go)h(directly)f(to)h Fj(BZ2_bzWriteClose)p
+-Fl(,)c(but)j(this)g(w)m(ould)g(b)s(e)g(prett)m(y)h(p)s(oin)m(tless\))
+-572 3259 y Fj(BZ2_bzWriteClose)663 3363 y Fl(otherwise)150
+-3639 y Ff(3.4.6)63 b Fe(BZ2_bzWrite)533 3826 y Fj(void)47
+-b(BZ2_bzWrite)e(\()i(int)g(*bzerror,)e(BZFILE)h(*b,)h(void)g(*buf,)f
+-(int)h(len)g(\);)150 3983 y Fl(Absorbs)26 b Fj(len)g
+-Fl(b)m(ytes)i(from)e(the)i(bu\013er)e Fj(buf)p Fl(,)h(ev)m(en)m(tually)
+-g(to)h(b)s(e)e(compressed)h(and)f(written)g(to)i(the)g(\014le.)150
+-4140 y(P)m(ossible)h(assignmen)m(ts)h(to)h Fj(bzerror)p
+-Fl(:)572 4291 y Fj(BZ_PARAM_ERROR)663 4395 y Fl(if)e
+-Fj(b)h Fl(is)g Fj(NULL)f Fl(or)h Fj(buf)g Fl(is)f Fj(NULL)g
+-Fl(or)i Fj(len)46 b(<)i(0)572 4498 y(BZ_SEQUENCE_ERROR)663
+-4602 y Fl(if)29 b(b)h(w)m(as)h(op)s(ened)e(with)g Fj(BZ2_bzReadOpen)572
+-4706 y(BZ_IO_ERROR)663 4810 y Fl(if)g(there)i(is)e(an)h(error)g
+-(writing)f(the)h(compressed)g(\014le.)572 4914 y Fj(BZ_OK)663
+-5017 y Fl(otherwise)150 5294 y Ff(3.4.7)63 b Fe(BZ2_bzWriteClose)p
+-eop
+-%%Page: 24 25
+-24 24 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(24)533 299 y Fj(void)47
+-b(BZ2_bzWriteClose)c(\()48 b(int)f(*bzerror,)e(BZFILE*)h(f,)1679
+-403 y(int)h(abandon,)1679 506 y(unsigned)e(int*)i(nbytes_in,)1679
+-610 y(unsigned)e(int*)i(nbytes_out)e(\);)533 818 y(void)i
+-(BZ2_bzWriteClose64)c(\()k(int)g(*bzerror,)e(BZFILE*)h(f,)1774
+-922 y(int)h(abandon,)1774 1025 y(unsigned)f(int*)g(nbytes_in_lo32,)1774
+-1129 y(unsigned)g(int*)g(nbytes_in_hi32,)1774 1233 y(unsigned)g(int*)g
+-(nbytes_out_lo32,)1774 1337 y(unsigned)g(int*)g(nbytes_out_hi32)e(\);)
+-150 1493 y Fl(Compresses)39 b(and)g(\015ushes)g(to)h(the)g(compressed)g
+-(\014le)f(all)f(data)j(so)f(far)g(supplied)c(b)m(y)k
+-Fj(BZ2_bzWrite)p Fl(.)150 1603 y(The)27 b(logical)g(end-of-stream)h
+-(mark)m(ers)g(are)g(also)f(written,)h(so)f(subsequen)m(t)g(calls)g(to)h
+-Fj(BZ2_bzWrite)d Fl(are)150 1713 y(illegal.)50 b(All)33
+-b(memory)h(asso)s(ciated)g(with)f(the)i(compressed)e(\014le)h
+-Fj(b)f Fl(is)g(released.)52 b Fj(fflush)33 b Fl(is)g(called)g(on)150
+-1822 y(the)e(compressed)f(\014le,)f(but)h(it)g(is)f(not)i
+-Fj(fclose)p Fl('d.)150 1979 y(If)i Fj(BZ2_bzWriteClose)c
+-Fl(is)k(called)f(to)j(clean)e(up)f(after)i(an)g(error,)g(the)g(only)e
+-(action)i(is)f(to)h(release)g(the)150 2089 y(memory)-8
+-b(.)42 b(The)30 b(library)e(records)j(the)g(error)f(co)s(des)h(issued)e
+-(b)m(y)h(previous)f(calls,)i(so)f(this)g(situation)g(will)150
+-2198 y(b)s(e)c(detected)h(automatically)-8 b(.)40 b(There)26
+-b(is)g(no)g(attempt)h(to)h(complete)e(the)h(compression)f(op)s
+-(eration,)g(nor)150 2308 y(to)32 b Fj(fflush)d Fl(the)i(compressed)g
+-(\014le.)42 b(Y)-8 b(ou)32 b(can)f(force)h(this)e(b)s(eha)m(viour)g(to)
+-h(happ)s(en)f(ev)m(en)i(in)d(the)j(case)g(of)150 2417
+-y(no)e(error,)g(b)m(y)h(passing)e(a)i(nonzero)f(v)-5
+-b(alue)30 b(to)h Fj(abandon)p Fl(.)150 2574 y(If)j Fj(nbytes_in)d
+-Fl(is)j(non-n)m(ull,)f Fj(*nbytes_in)e Fl(will)h(b)s(e)h(set)i(to)g(b)s
+-(e)f(the)g(total)h(v)m(olume)f(of)g(uncompressed)150
+-2684 y(data)k(handled.)60 b(Similarly)-8 b(,)35 b Fj(nbytes_out)g
+-Fl(will)g(b)s(e)h(set)i(to)g(the)g(total)g(v)m(olume)f(of)g(compressed)
+-g(data)150 2793 y(written.)h(F)-8 b(or)27 b(compatibilit)m(y)d(with)h
+-(older)g(v)m(ersions)h(of)g(the)g(library)-8 b(,)25 b
+-Fj(BZ2_bzWriteClose)d Fl(only)j(yields)150 2903 y(the)40
+-b(lo)m(w)m(er)g(32)h(bits)d(of)i(these)h(coun)m(ts.)69
+-b(Use)40 b Fj(BZ2_bzWriteClose64)35 b Fl(if)k(y)m(ou)h(w)m(an)m(t)h
+-(the)f(full)d(64)k(bit)150 3013 y(coun)m(ts.)g(These)30
+-b(t)m(w)m(o)i(functions)d(are)i(otherwise)f(absolutely)f(iden)m(tical.)
+-150 3169 y(P)m(ossible)g(assignmen)m(ts)h(to)h Fj(bzerror)p
+-Fl(:)572 3320 y Fj(BZ_SEQUENCE_ERROR)663 3424 y Fl(if)e
+-Fj(b)h Fl(w)m(as)h(op)s(ened)e(with)h Fj(BZ2_bzReadOpen)572
+-3528 y(BZ_IO_ERROR)663 3632 y Fl(if)f(there)i(is)e(an)h(error)g
+-(writing)f(the)h(compressed)g(\014le)572 3736 y Fj(BZ_OK)663
+-3839 y Fl(otherwise)150 4161 y Ff(3.4.8)63 b(Handling)41
+-b(em)m(b)s(edded)g(compressed)h(data)e(streams)150 4354
+-y Fl(The)i(high-lev)m(el)g(library)f(facilitates)h(use)h(of)g
+-Fj(bzip2)e Fl(data)j(streams)f(whic)m(h)f(form)g(some)i(part)e(of)i(a)
+-150 4463 y(surrounding,)27 b(larger)j(data)h(stream.)225
+-4620 y Fi(\017)60 b Fl(F)-8 b(or)22 b(writing,)f(the)g(library)e(tak)m
+-(es)k(an)e(op)s(en)f(\014le)g(handle,)i(writes)e(compressed)h(data)h
+-(to)g(it,)g Fj(fflush)p Fl(es)330 4730 y(it)34 b(but)f(do)s(es)h(not)h
+-Fj(fclose)d Fl(it.)52 b(The)34 b(calling)f(application)g(can)h(write)g
+-(its)f(o)m(wn)i(data)g(b)s(efore)f(and)330 4839 y(after)d(the)f
+-(compressed)h(data)g(stream,)g(using)d(that)j(same)g(\014le)f(handle.)
+-225 5011 y Fi(\017)60 b Fl(Reading)34 b(is)f(more)i(complex,)g(and)f
+-(the)h(facilities)d(are)j(not)g(as)g(general)f(as)h(they)f(could)g(b)s
+-(e)g(since)330 5121 y(generalit)m(y)e(is)f(hard)f(to)j(reconcile)e
+-(with)f(e\016ciency)-8 b(.)46 b Fj(BZ2_bzRead)29 b Fl(reads)i(from)g
+-(the)h(compressed)330 5230 y(\014le)39 b(in)g(blo)s(c)m(ks)g(of)h(size)
+-g Fj(BZ_MAX_UNUSED)c Fl(b)m(ytes,)44 b(and)39 b(in)g(doing)g(so)h
+-(probably)e(will)f(o)m(v)m(ersho)s(ot)330 5340 y(the)i(logical)g(end)f
+-(of)h(compressed)f(stream.)67 b(T)-8 b(o)40 b(reco)m(v)m(er)g(this)e
+-(data)i(once)f(decompression)f(has)p eop
+-%%Page: 25 26
+-25 25 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(25)330 299 y(ended,)28
+-b(call)g Fj(BZ2_bzReadGetUnused)23 b Fl(after)29 b(the)g(last)f(call)g
+-(of)g Fj(BZ2_bzRead)e Fl(\(the)j(one)g(returning)330
+-408 y Fj(BZ_STREAM_END)p Fl(\))e(but)j(b)s(efore)g(calling)f
+-Fj(BZ2_bzReadClose)p Fl(.)150 596 y(This)51 b(mec)m(hanism)g(mak)m(es)j
+-(it)e(easy)h(to)g(decompress)f(m)m(ultiple)e Fj(bzip2)i
+-Fl(streams)g(placed)g(end-to-)150 706 y(end.)90 b(As)48
+-b(the)f(end)f(of)i(one)f(stream,)52 b(when)46 b Fj(BZ2_bzRead)f
+-Fl(returns)h Fj(BZ_STREAM_END)p Fl(,)i(call)e Fj(BZ2_)150
+-816 y(bzReadGetUnused)36 b Fl(to)41 b(collect)g(the)g(un)m(used)e(data)
+-i(\(cop)m(y)g(it)f(in)m(to)g(y)m(our)h(o)m(wn)f(bu\013er)f
+-(somewhere\).)150 925 y(That)25 b(data)g(forms)f(the)h(start)h(of)e
+-(the)h(next)g(compressed)g(stream.)39 b(T)-8 b(o)25 b(start)h
+-(uncompressing)c(that)k(next)150 1035 y(stream,)40 b(call)d
+-Fj(BZ2_bzReadOpen)d Fl(again,)40 b(feeding)d(in)g(the)h(un)m(used)e
+-(data)j(via)e(the)h Fj(unused)p Fl(/)p Fj(nUnused)150
+-1144 y Fl(parameters.)54 b(Keep)34 b(doing)g(this)f(un)m(til)g
+-Fj(BZ_STREAM_END)e Fl(return)j(coincides)f(with)h(the)g(ph)m(ysical)g
+-(end)150 1254 y(of)d(\014le)e(\()p Fj(feof\(f\))p Fl(\).)39
+-b(In)30 b(this)f(situation)h Fj(BZ2_bzReadGetUnused)25
+-b Fl(will)i(of)k(course)g(return)e(no)h(data.)150 1411
+-y(This)c(should)f(giv)m(e)j(some)g(feel)f(for)g(ho)m(w)h(the)g
+-(high-lev)m(el)e(in)m(terface)i(can)f(b)s(e)g(used.)39
+-b(If)27 b(y)m(ou)h(require)e(extra)150 1520 y(\015exibilit)m(y)-8
+-b(,)28 b(y)m(ou'll)i(ha)m(v)m(e)h(to)g(bite)f(the)h(bullet)d(and)i(get)
+-i(to)f(grips)e(with)g(the)h(lo)m(w-lev)m(el)h(in)m(terface.)150
+-1779 y Ff(3.4.9)63 b(Standard)40 b(\014le-reading/writing)j(co)s(de)150
+-1972 y Fl(Here's)31 b(ho)m(w)f(y)m(ou'd)h(write)e(data)j(to)f(a)f
+-(compressed)g(\014le:)390 2330 y Fj(FILE*)142 b(f;)390
+-2434 y(BZFILE*)46 b(b;)390 2538 y(int)238 b(nBuf;)390
+-2642 y(char)190 b(buf[)46 b(/*)i(whatever)d(size)i(you)g(like)f(*/)i
+-(];)390 2746 y(int)238 b(bzerror;)390 2849 y(int)g(nWritten;)390
+-3057 y(f)47 b(=)h(fopen)e(\()i("myfile.bz2",)c("w")j(\);)390
+-3161 y(if)g(\(!f\))g({)533 3264 y(/*)g(handle)f(error)h(*/)390
+-3368 y(})390 3472 y(b)g(=)h(BZ2_bzWriteOpen)c(\()j(&bzerror,)e(f,)i(9)h
+-(\);)390 3576 y(if)f(\(bzerror)f(!=)h(BZ_OK\))f({)533
+-3680 y(BZ2_bzWriteClose)e(\()j(b)g(\);)533 3783 y(/*)g(handle)f(error)h
+-(*/)390 3887 y(})390 4095 y(while)f(\()i(/*)f(condition)e(*/)i(\))h({)
+-533 4198 y(/*)f(get)g(data)g(to)g(write)f(into)h(buf,)g(and)g(set)g
+-(nBuf)f(appropriately)e(*/)533 4302 y(nWritten)i(=)h(BZ2_bzWrite)e(\()i
+-(&bzerror,)f(b,)h(buf,)f(nBuf)h(\);)533 4406 y(if)g(\(bzerror)f(==)h
+-(BZ_IO_ERROR\))e({)676 4510 y(BZ2_bzWriteClose)f(\()j(&bzerror,)e(b)j
+-(\);)676 4614 y(/*)g(handle)e(error)g(*/)533 4717 y(})390
+-4821 y(})390 5029 y(BZ2_bzWriteClose)d(\()48 b(&bzerror,)d(b)j(\);)390
+-5132 y(if)f(\(bzerror)f(==)h(BZ_IO_ERROR\))d({)533 5236
+-y(/*)j(handle)f(error)h(*/)390 5340 y(})p eop
+-%%Page: 26 27
+-26 26 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(26)150 299 y(And)29 b(to)j(read)e(from)g
+-(a)h(compressed)f(\014le:)390 450 y Fj(FILE*)142 b(f;)390
+-554 y(BZFILE*)46 b(b;)390 657 y(int)238 b(nBuf;)390 761
+-y(char)190 b(buf[)46 b(/*)i(whatever)d(size)i(you)g(like)f(*/)i(];)390
+-865 y(int)238 b(bzerror;)390 969 y(int)g(nWritten;)390
+-1176 y(f)47 b(=)h(fopen)e(\()i("myfile.bz2",)c("r")j(\);)390
+-1280 y(if)g(\(!f\))g({)533 1384 y(/*)g(handle)f(error)h(*/)390
+-1488 y(})390 1591 y(b)g(=)h(BZ2_bzReadOpen)c(\()j(&bzerror,)f(f,)h(0,)g
+-(NULL,)f(0)i(\);)390 1695 y(if)f(\(bzerror)f(!=)h(BZ_OK\))f({)533
+-1799 y(BZ2_bzReadClose)e(\()j(&bzerror,)f(b)h(\);)533
+-1903 y(/*)g(handle)f(error)h(*/)390 2007 y(})390 2214
+-y(bzerror)f(=)h(BZ_OK;)390 2318 y(while)f(\(bzerror)g(==)h(BZ_OK)f(&&)i
+-(/*)f(arbitrary)e(other)h(conditions)f(*/\))i({)533 2422
+-y(nBuf)g(=)g(BZ2_bzRead)e(\()j(&bzerror,)d(b,)i(buf,)g(/*)g(size)g(of)g
+-(buf)g(*/)g(\);)533 2525 y(if)g(\(bzerror)f(==)h(BZ_OK\))f({)676
+-2629 y(/*)i(do)f(something)e(with)i(buf[0)f(..)h(nBuf-1])f(*/)533
+-2733 y(})390 2837 y(})390 2941 y(if)h(\(bzerror)f(!=)h(BZ_STREAM_END\))
+-d({)533 3044 y(BZ2_bzReadClose)g(\()j(&bzerror,)f(b)h(\);)533
+-3148 y(/*)g(handle)f(error)h(*/)390 3252 y(})g(else)g({)533
+-3356 y(BZ2_bzReadClose)d(\()j(&bzerror)f(\);)390 3459
+-y(})150 3753 y Fk(3.5)68 b(Utilit)l(y)47 b(functions)150
+-4045 y Ff(3.5.1)63 b Fe(BZ2_bzBuffToBuffCompress)533
+-4232 y Fj(int)47 b(BZ2_bzBuffToBuffCompress\()41 b(char*)428
+-b(dest,)1965 4335 y(unsigned)46 b(int*)g(destLen,)1965
+-4439 y(char*)428 b(source,)1965 4543 y(unsigned)46 b(int)94
+-b(sourceLen,)1965 4647 y(int)524 b(blockSize100k,)1965
+-4751 y(int)g(verbosity,)1965 4854 y(int)g(workFactor)45
+-b(\);)150 5011 y Fl(A)m(ttempts)33 b(to)g(compress)f(the)g(data)h(in)e
+-Fj(source[0)d(..)i(sourceLen-1])e Fl(in)m(to)k(the)h(destination)e
+-(bu\013er,)150 5121 y Fj(dest[0)e(..)g(*destLen-1])p
+-Fl(.)37 b(If)26 b(the)g(destination)g(bu\013er)f(is)h(big)f(enough,)j
+-Fj(*destLen)c Fl(is)h(set)i(to)g(the)g(size)150 5230
+-y(of)i(the)f(compressed)h(data,)g(and)f Fj(BZ_OK)f Fl(is)h(returned.)39
+-b(If)28 b(the)h(compressed)f(data)h(w)m(on't)g(\014t,)g
+-Fj(*destLen)150 5340 y Fl(is)g(unc)m(hanged,)i(and)e
+-Fj(BZ_OUTBUFF_FULL)e Fl(is)i(returned.)p eop
+-%%Page: 27 28
+-27 27 bop 150 -116 a Fl(Chapter)30 b(3:)h(Programming)e(with)g
+-Fj(libbzip2)1891 b Fl(27)150 299 y(Compression)22 b(in)g(this)h(manner)
+-g(is)g(a)h(one-shot)g(ev)m(en)m(t,)j(done)c(with)g(a)h(single)e(call)h
+-(to)i(this)d(function.)37 b(The)150 408 y(resulting)25
+-b(compressed)i(data)i(is)d(a)i(complete)f Fj(bzip2)f
+-Fl(format)i(data)g(stream.)40 b(There)27 b(is)f(no)i(mec)m(hanism)150
+-518 y(for)23 b(making)g(additional)e(calls)i(to)h(pro)m(vide)f(extra)h
+-(input)e(data.)39 b(If)23 b(y)m(ou)h(w)m(an)m(t)g(that)g(kind)e(of)h
+-(mec)m(hanism,)150 628 y(use)30 b(the)h(lo)m(w-lev)m(el)f(in)m
+-(terface.)150 784 y(F)-8 b(or)31 b(the)g(meaning)e(of)i(parameters)g
+-Fj(blockSize100k)p Fl(,)c Fj(verbosity)g Fl(and)j Fj(workFactor)p
+-Fl(,)150 894 y(see)h Fj(BZ2_bzCompressInit)p Fl(.)150
+-1051 y(T)-8 b(o)27 b(guaran)m(tee)h(that)e(the)h(compressed)f(data)h
+-(will)d(\014t)i(in)f(its)g(bu\013er,)i(allo)s(cate)f(an)g(output)g
+-(bu\013er)g(of)g(size)150 1160 y(1\045)31 b(larger)f(than)g(the)g
+-(uncompressed)f(data,)j(plus)c(six)h(h)m(undred)g(extra)i(b)m(ytes.)150
+-1317 y Fj(BZ2_bzBuffToBuffDecompre)o(ss)25 b Fl(will)k(not)j(write)e
+-(data)j(at)f(or)f(b)s(ey)m(ond)g Fj(dest[*destLen])p
+-Fl(,)d(ev)m(en)k(in)150 1427 y(case)f(of)g(bu\013er)e(o)m(v)m(er\015o)m
+-(w.)150 1584 y(P)m(ossible)g(return)h(v)-5 b(alues:)572
+-1735 y Fj(BZ_CONFIG_ERROR)663 1839 y Fl(if)29 b(the)i(library)d(has)i
+-(b)s(een)f(mis-compiled)572 1942 y Fj(BZ_PARAM_ERROR)663
+-2046 y Fl(if)g Fj(dest)g Fl(is)h Fj(NULL)f Fl(or)h Fj(destLen)f
+-Fl(is)g Fj(NULL)663 2150 y Fl(or)h Fj(blockSize100k)44
+-b(<)k(1)30 b Fl(or)g Fj(blockSize100k)44 b(>)k(9)663
+-2254 y Fl(or)30 b Fj(verbosity)45 b(<)j(0)30 b Fl(or)g
+-Fj(verbosity)45 b(>)j(4)663 2357 y Fl(or)30 b Fj(workFactor)45
+-b(<)j(0)30 b Fl(or)g Fj(workFactor)45 b(>)i(250)572 2461
+-y(BZ_MEM_ERROR)663 2565 y Fl(if)29 b(insu\016cien)m(t)g(memory)h(is)f
+-(a)m(v)-5 b(ailable)572 2669 y Fj(BZ_OUTBUFF_FULL)663
+-2773 y Fl(if)29 b(the)i(size)f(of)g(the)h(compressed)f(data)h(exceeds)g
+-Fj(*destLen)572 2876 y(BZ_OK)663 2980 y Fl(otherwise)150
+-3349 y Ff(3.5.2)63 b Fe(BZ2_bzBuffToBuffDecompress)533
+-3536 y Fj(int)47 b(BZ2_bzBuffToBuffDecompres)o(s)42 b(\()47
+-b(char*)428 b(dest,)2108 3640 y(unsigned)46 b(int*)g(destLen,)2108
+-3744 y(char*)428 b(source,)2108 3848 y(unsigned)46 b(int)94
+-b(sourceLen,)2108 3951 y(int)524 b(small,)2108 4055 y(int)g(verbosity)
+-46 b(\);)150 4212 y Fl(A)m(ttempts)24 b(to)g(decompress)f(the)g(data)g
+-(in)f Fj(source[0)28 b(..)i(sourceLen-1])20 b Fl(in)m(to)j(the)g
+-(destination)f(bu\013er,)150 4322 y Fj(dest[0)29 b(..)g(*destLen-1])p
+-Fl(.)37 b(If)26 b(the)g(destination)g(bu\013er)f(is)h(big)f(enough,)j
+-Fj(*destLen)c Fl(is)h(set)i(to)g(the)g(size)150 4431
+-y(of)21 b(the)g(uncompressed)e(data,)24 b(and)c Fj(BZ_OK)f
+-Fl(is)h(returned.)36 b(If)20 b(the)h(compressed)g(data)g(w)m(on't)h
+-(\014t,)g Fj(*destLen)150 4541 y Fl(is)29 b(unc)m(hanged,)i(and)e
+-Fj(BZ_OUTBUFF_FULL)e Fl(is)i(returned.)150 4698 y Fj(source)g
+-Fl(is)g(assumed)h(to)h(hold)e(a)i(complete)f Fj(bzip2)f
+-Fl(format)i(data)g(stream.)150 4807 y Fj(BZ2_bzBuffToBuffDecompre)o(ss)
+-22 b Fl(tries)28 b(to)i(decompress)e(the)h(en)m(tiret)m(y)g(of)g(the)f
+-(stream)h(in)m(to)g(the)f(out-)150 4917 y(put)i(bu\013er.)150
+-5074 y(F)-8 b(or)31 b(the)g(meaning)e(of)i(parameters)g
+-Fj(small)e Fl(and)g Fj(verbosity)p Fl(,)f(see)j Fj
+-(BZ2_bzDecompressInit)p Fl(.)150 5230 y(Because)j(the)f(compression)e
+-(ratio)i(of)g(the)g(compressed)f(data)h(cannot)g(b)s(e)f(kno)m(wn)g(in)
+-g(adv)-5 b(ance,)34 b(there)150 5340 y(is)d(no)h(easy)g(w)m(a)m(y)h(to)
+-f(guaran)m(tee)i(that)e(the)g(output)f(bu\013er)g(will)e(b)s(e)i(big)g
+-(enough.)45 b(Y)-8 b(ou)32 b(ma)m(y)h(of)f(course)p eop
+-%%Page: 28 29
+-28 28 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(28)150 299 y(mak)m(e)36
+-b(arrangemen)m(ts)f(in)e(y)m(our)i(co)s(de)g(to)g(record)g(the)g(size)f
+-(of)h(the)g(uncompressed)f(data,)i(but)e(suc)m(h)h(a)150
+-408 y(mec)m(hanism)30 b(is)f(b)s(ey)m(ond)h(the)g(scop)s(e)h(of)f(this)
+-g(library)-8 b(.)150 565 y Fj(BZ2_bzBuffToBuffDecompre)o(ss)25
+-b Fl(will)k(not)j(write)e(data)j(at)f(or)f(b)s(ey)m(ond)g
+-Fj(dest[*destLen])p Fl(,)d(ev)m(en)k(in)150 675 y(case)f(of)g(bu\013er)
+-e(o)m(v)m(er\015o)m(w.)150 832 y(P)m(ossible)g(return)h(v)-5
+-b(alues:)572 983 y Fj(BZ_CONFIG_ERROR)663 1087 y Fl(if)29
+-b(the)i(library)d(has)i(b)s(een)f(mis-compiled)572 1190
+-y Fj(BZ_PARAM_ERROR)663 1294 y Fl(if)g Fj(dest)g Fl(is)h
+-Fj(NULL)f Fl(or)h Fj(destLen)f Fl(is)g Fj(NULL)663 1398
+-y Fl(or)h Fj(small)46 b(!=)i(0)f(&&)g(small)g(!=)g(1)663
+-1502 y Fl(or)30 b Fj(verbosity)45 b(<)j(0)30 b Fl(or)g
+-Fj(verbosity)45 b(>)j(4)572 1605 y(BZ_MEM_ERROR)663 1709
+-y Fl(if)29 b(insu\016cien)m(t)g(memory)h(is)f(a)m(v)-5
+-b(ailable)572 1813 y Fj(BZ_OUTBUFF_FULL)663 1917 y Fl(if)29
+-b(the)i(size)f(of)g(the)h(compressed)f(data)h(exceeds)g
+-Fj(*destLen)572 2021 y(BZ_DATA_ERROR)663 2124 y Fl(if)e(a)i(data)g(in)m
+-(tegrit)m(y)f(error)g(w)m(as)h(detected)h(in)d(the)h(compressed)g(data)
+-572 2228 y Fj(BZ_DATA_ERROR_MAGIC)663 2332 y Fl(if)f(the)i(compressed)f
+-(data)h(do)s(esn't)f(b)s(egin)f(with)g(the)i(righ)m(t)e(magic)i(b)m
+-(ytes)572 2436 y Fj(BZ_UNEXPECTED_EOF)663 2539 y Fl(if)e(the)i
+-(compressed)f(data)h(ends)e(unexp)s(ectedly)572 2643
+-y Fj(BZ_OK)663 2747 y Fl(otherwise)150 3116 y Fk(3.6)68
+-b Fd(zlib)43 b Fk(compatibilit)l(y)k(functions)150 3308
+-y Fl(Y)-8 b(oshiok)j(a)33 b(Tsuneo)e(has)h(con)m(tributed)g(some)g
+-(functions)f(to)i(giv)m(e)g(b)s(etter)f Fj(zlib)f Fl(compatibilit)m(y)
+--8 b(.)45 b(These)150 3418 y(functions)36 b(are)i Fj(BZ2_bzopen)p
+-Fl(,)e Fj(BZ2_bzread)p Fl(,)h Fj(BZ2_bzwrite)p Fl(,)f
+-Fj(BZ2_bzflush)p Fl(,)h Fj(BZ2_bzclose)p Fl(,)f Fj(BZ2_)150
+-3527 y(bzerror)23 b Fl(and)h Fj(BZ2_bzlibVersion)p Fl(.)34
+-b(These)25 b(functions)e(are)j(not)f(\(y)m(et\))h(o\016cially)e(part)h
+-(of)g(the)g(library)-8 b(.)150 3637 y(If)30 b(they)g(break,)h(y)m(ou)g
+-(get)g(to)g(k)m(eep)g(all)f(the)g(pieces.)41 b(Nev)m(ertheless,)31
+-b(I)f(think)f(they)i(w)m(ork)f(ok.)390 3788 y Fj(typedef)46
+-b(void)g(BZFILE;)390 3995 y(const)g(char)h(*)g(BZ2_bzlibVersion)d(\()j
+-(void)g(\);)150 4152 y Fl(Returns)29 b(a)i(string)f(indicating)e(the)i
+-(library)e(v)m(ersion.)390 4303 y Fj(BZFILE)46 b(*)i(BZ2_bzopen)92
+-b(\()48 b(const)e(char)h(*path,)f(const)g(char)h(*mode)f(\);)390
+-4407 y(BZFILE)g(*)i(BZ2_bzdopen)c(\()k(int)381 b(fd,)190
+-b(const)46 b(char)h(*mode)f(\);)150 4564 y Fl(Op)s(ens)19
+-b(a)j Fj(.bz2)e Fl(\014le)g(for)g(reading)g(or)h(writing,)g(using)f
+-(either)g(its)h(name)g(or)g(a)g(pre-existing)f(\014le)g(descriptor.)150
+-4674 y(Analogous)30 b(to)i Fj(fopen)c Fl(and)i Fj(fdopen)p
+-Fl(.)390 4825 y Fj(int)47 b(BZ2_bzread)93 b(\()47 b(BZFILE*)f(b,)h
+-(void*)f(buf,)h(int)g(len)g(\);)390 4928 y(int)g(BZ2_bzwrite)e(\()i
+-(BZFILE*)f(b,)h(void*)f(buf,)h(int)g(len)g(\);)150 5085
+-y Fl(Reads/writes)30 b(data)h(from/to)g(a)g(previously)d(op)s(ened)i
+-Fj(BZFILE)p Fl(.)39 b(Analogous)30 b(to)h Fj(fread)e
+-Fl(and)h Fj(fwrite)p Fl(.)390 5236 y Fj(int)95 b(BZ2_bzflush)44
+-b(\()k(BZFILE*)e(b)h(\);)390 5340 y(void)g(BZ2_bzclose)d(\()k(BZFILE*)e
+-(b)h(\);)p eop
+-%%Page: 29 30
+-29 29 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(29)150 299 y(Flushes/closes)27
+-b(a)h Fj(BZFILE)p Fl(.)39 b Fj(BZ2_bzflush)24 b Fl(do)s(esn't)k
+-(actually)f(do)h(an)m(ything.)39 b(Analogous)28 b(to)h
+-Fj(fflush)150 408 y Fl(and)h Fj(fclose)p Fl(.)390 559
+-y Fj(const)46 b(char)h(*)g(BZ2_bzerror)e(\()j(BZFILE)e(*b,)h(int)g
+-(*errnum)e(\))150 716 y Fl(Returns)31 b(a)i(string)e(describing)f(the)i
+-(more)g(recen)m(t)h(error)f(status)h(of)f Fj(b)p Fl(,)g(and)g(also)g
+-(sets)h Fj(*errnum)d Fl(to)j(its)150 826 y(n)m(umerical)c(v)-5
+-b(alue.)150 1242 y Fk(3.7)68 b(Using)46 b(the)f(library)g(in)g(a)g
+-Fd(stdio)p Fk(-free)f(en)l(vironmen)l(t)150 1615 y Ff(3.7.1)63
+-b(Getting)40 b(rid)h(of)g Fe(stdio)150 1807 y Fl(In)i(a)g(deeply)g(em)m
+-(b)s(edded)f(application,)j(y)m(ou)f(migh)m(t)f(w)m(an)m(t)h(to)g(use)f
+-(just)g(the)h(memory-to-memory)150 1917 y(functions.)39
+-b(Y)-8 b(ou)30 b(can)f(do)g(this)g(con)m(v)m(enien)m(tly)g(b)m(y)g
+-(compiling)e(the)j(library)d(with)h(prepro)s(cessor)g(sym)m(b)s(ol)150
+-2026 y Fj(BZ_NO_STDIO)35 b Fl(de\014ned.)63 b(Doing)39
+-b(this)e(giv)m(es)h(y)m(ou)h(a)f(library)e(con)m(taining)i(only)f(the)i
+-(follo)m(wing)e(eigh)m(t)150 2136 y(functions:)150 2293
+-y Fj(BZ2_bzCompressInit)p Fl(,)26 b Fj(BZ2_bzCompress)p
+-Fl(,)g Fj(BZ2_bzCompressEnd)150 2402 y(BZ2_bzDecompressInit)p
+-Fl(,)f Fj(BZ2_bzDecompress)p Fl(,)h Fj(BZ2_bzDecompressEnd)150
+-2512 y(BZ2_bzBuffToBuffCompress)o Fl(,)f Fj(BZ2_bzBuffToBuffDecompre)o
+-(ss)150 2669 y Fl(When)30 b(compiled)f(lik)m(e)h(this,)f(all)g
+-(functions)g(will)f(ignore)i Fj(verbosity)e Fl(settings.)150
+-3006 y Ff(3.7.2)63 b(Critical)40 b(error)h(handling)150
+-3199 y Fj(libbzip2)20 b Fl(con)m(tains)j(a)g(n)m(um)m(b)s(er)f(of)g(in)
+-m(ternal)g(assertion)g(c)m(hec)m(ks)i(whic)m(h)d(should,)i(needless)f
+-(to)h(sa)m(y)-8 b(,)26 b(nev)m(er)150 3308 y(b)s(e)g(activ)-5
+-b(ated.)40 b(Nev)m(ertheless,)28 b(if)d(an)i(assertion)f(should)e
+-(fail,)i(b)s(eha)m(viour)f(dep)s(ends)f(on)j(whether)e(or)i(not)150
+-3418 y(the)k(library)d(w)m(as)i(compiled)f(with)g Fj(BZ_NO_STDIO)e
+-Fl(set.)150 3575 y(F)-8 b(or)31 b(a)g(normal)e(compile,)h(an)g
+-(assertion)g(failure)f(yields)f(the)j(message)533 3726
+-y Fj(bzip2/libbzip2:)44 b(internal)h(error)i(number)f(N.)533
+-3829 y(This)h(is)g(a)g(bug)g(in)h(bzip2/libbzip2,)43
+-b(1.0)k(of)g(21-Mar-2000.)533 3933 y(Please)f(report)g(it)i(to)f(me)g
+-(at:)g(jseward@acm.org.)91 b(If)47 b(this)g(happened)533
+-4037 y(when)g(you)g(were)f(using)h(some)f(program)g(which)h(uses)f
+-(libbzip2)g(as)h(a)533 4141 y(component,)e(you)i(should)f(also)h
+-(report)f(this)h(bug)f(to)i(the)f(author\(s\))533 4244
+-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 4348 y(timely)g(and)h(accurate)f(bug)h
+-(reports)e(eventually)g(lead)i(to)g(higher)533 4452 y(quality)f
+-(software.)93 b(Thanks.)h(Julian)46 b(Seward,)f(21)j(March)e(2000.)150
+-4609 y Fl(where)30 b Fj(N)g Fl(is)f(some)i(error)f(co)s(de)h(n)m(um)m
+-(b)s(er.)39 b Fj(exit\(3\))28 b Fl(is)i(then)g(called.)150
+-4766 y(F)-8 b(or)31 b(a)g Fj(stdio)p Fl(-free)e(library)-8
+-b(,)29 b(assertion)h(failures)e(result)i(in)f(a)i(call)e(to)i(a)g
+-(function)e(declared)h(as:)533 4917 y Fj(extern)46 b(void)h
+-(bz_internal_error)c(\()k(int)g(errcode)f(\);)150 5074
+-y Fl(The)30 b(relev)-5 b(an)m(t)31 b(co)s(de)f(is)g(passed)f(as)i(a)g
+-(parameter.)41 b(Y)-8 b(ou)31 b(should)d(supply)g(suc)m(h)i(a)h
+-(function.)150 5230 y(In)g(either)g(case,)j(once)e(an)g(assertion)g
+-(failure)e(has)h(o)s(ccurred,)h(an)m(y)g Fj(bz_stream)e
+-Fl(records)h(in)m(v)m(olv)m(ed)h(can)150 5340 y(b)s(e)e(regarded)g(as)h
+-(in)m(v)-5 b(alid.)38 b(Y)-8 b(ou)31 b(should)d(not)j(attempt)g(to)g
+-(resume)f(normal)g(op)s(eration)f(with)g(them.)p eop
+-%%Page: 30 31
+-30 30 bop 150 -116 a Fl(Chapter)30 b(3:)41 b(Programming)29
+-b(with)g Fj(libbzip2)1881 b Fl(30)150 299 y(Y)-8 b(ou)22
+-b(ma)m(y)-8 b(,)25 b(of)d(course,)h(c)m(hange)g(critical)e(error)g
+-(handling)e(to)j(suit)f(y)m(our)g(needs.)38 b(As)21 b(I)h(said)e(ab)s
+-(o)m(v)m(e,)25 b(critical)150 408 y(errors)30 b(indicate)g(bugs)g(in)g
+-(the)h(library)d(and)i(should)f(not)i(o)s(ccur.)42 b(All)29
+-b Fj(")p Fl(normal)p Fj(")h Fl(error)g(situations)g(are)150
+-518 y(indicated)f(via)h(error)g(return)f(co)s(des)i(from)f(functions,)f
+-(and)g(can)i(b)s(e)f(reco)m(v)m(ered)i(from.)150 798
+-y Fk(3.8)68 b(Making)45 b(a)g(Windo)l(ws)h(DLL)150 990
+-y Fl(Ev)m(erything)30 b(related)g(to)h(Windo)m(ws)f(has)g(b)s(een)f
+-(con)m(tributed)h(b)m(y)g(Y)-8 b(oshiok)j(a)31 b(Tsuneo)150
+-1100 y(\()p Fj(QWF00133@niftyserve.or.jp)46 b Fl(/)52
+-b Fj(tsuneo-y@is.aist-nara.ac.j)o(p)p Fl(\),)g(so)h(y)m(ou)f(should)f
+-(send)150 1210 y(y)m(our)30 b(queries)g(to)h(him)e(\(but)h(p)s(erhaps)e
+-(Cc:)41 b(me,)31 b Fj(jseward@acm.org)p Fl(\).)150 1366
+-y(My)43 b(v)-5 b(ague)44 b(understanding)d(of)i(what)g(to)h(do)f(is:)65
+-b(using)41 b(Visual)h(C)p Fj(++)g Fl(5.0,)48 b(op)s(en)42
+-b(the)h(pro)5 b(ject)44 b(\014le)150 1476 y Fj(libbz2.dsp)p
+-Fl(,)28 b(and)i(build.)37 b(That's)31 b(all.)150 1633
+-y(If)41 b(y)m(ou)g(can't)h(op)s(en)e(the)h(pro)5 b(ject)42
+-b(\014le)e(for)h(some)g(reason,)j(mak)m(e)e(a)g(new)e(one,)k(naming)c
+-(these)i(\014les:)150 1742 y Fj(blocksort.c)p Fl(,)28
+-b Fj(bzlib.c)p Fl(,)g Fj(compress.c)p Fl(,)g Fj(crctable.c)p
+-Fl(,)g Fj(decompress.c)p Fl(,)f Fj(huffman.c)p Fl(,)150
+-1852 y Fj(randtable.c)32 b Fl(and)j Fj(libbz2.def)p Fl(.)53
+-b(Y)-8 b(ou)36 b(will)d(also)i(need)g(to)h(name)g(the)g(header)f
+-(\014les)f Fj(bzlib.h)g Fl(and)150 1962 y Fj(bzlib_private.h)p
+-Fl(.)150 2118 y(If)c(y)m(ou)h(don't)f(use)g(V)m(C)p Fj(++)p
+-Fl(,)g(y)m(ou)h(ma)m(y)g(need)f(to)h(de\014ne)f(the)h(propro)s(cessor)e
+-(sym)m(b)s(ol)g Fj(_WIN32)p Fl(.)150 2275 y(Finally)-8
+-b(,)28 b Fj(dlltest.c)e Fl(is)h(a)i(sample)f(program)g(using)g(the)g
+-(DLL.)h(It)g(has)f(a)h(pro)5 b(ject)29 b(\014le,)g Fj(dlltest.dsp)p
+-Fl(.)150 2432 y(If)h(y)m(ou)h(just)e(w)m(an)m(t)j(a)e(mak)m(e\014le)h
+-(for)f(Visual)f(C,)h(ha)m(v)m(e)i(a)e(lo)s(ok)g(at)i
+-Fj(makefile.msc)p Fl(.)150 2589 y(Be)k(a)m(w)m(are)g(that)g(if)e(y)m
+-(ou)h(compile)f Fj(bzip2)g Fl(itself)g(on)h(Win32,)h(y)m(ou)g(m)m(ust)f
+-(set)g Fj(BZ_UNIX)e Fl(to)j(0)f(and)g Fj(BZ_)150 2698
+-y(LCCWIN32)27 b Fl(to)j(1,)g(in)f(the)g(\014le)g Fj(bzip2.c)p
+-Fl(,)e(b)s(efore)i(compiling.)39 b(Otherwise)28 b(the)h(resulting)f
+-(binary)f(w)m(on't)150 2808 y(w)m(ork)j(correctly)-8
+-b(.)150 2965 y(I)30 b(ha)m(v)m(en't)i(tried)d(an)m(y)i(of)g(this)e
+-(stu\013)h(m)m(yself,)g(but)g(it)f(all)h(lo)s(oks)g(plausible.)p
+-eop
+-%%Page: 31 32
+-31 31 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586
+-b(31)150 299 y Fh(4)80 b(Miscellanea)150 583 y Fl(These)30
+-b(are)h(just)f(some)g(random)g(though)m(ts)h(of)f(mine.)40
+-b(Y)-8 b(our)30 b(mileage)h(ma)m(y)g(v)-5 b(ary)d(.)150
+-884 y Fk(4.1)68 b(Limitations)47 b(of)e(the)g(compressed)g(\014le)h
+-(format)150 1077 y Fj(bzip2-1.0)p Fl(,)e Fj(0.9.5)e Fl(and)g
+-Fj(0.9.0)g Fl(use)h(exactly)h(the)f(same)h(\014le)e(format)i(as)f(the)h
+-(previous)d(v)m(ersion,)150 1186 y Fj(bzip2-0.1)p Fl(.)75
+-b(This)41 b(decision)g(w)m(as)i(made)g(in)e(the)i(in)m(terests)g(of)g
+-(stabilit)m(y)-8 b(.)77 b(Creating)42 b(y)m(et)i(another)150
+-1296 y(incompatible)21 b(compressed)i(\014le)f(format)i(w)m(ould)e
+-(create)i(further)e(confusion)g(and)h(disruption)d(for)j(users.)150
+-1453 y(Nev)m(ertheless,)31 b(this)e(is)g(not)h(a)g(painless)e
+-(decision.)39 b(Dev)m(elopmen)m(t)31 b(w)m(ork)f(since)f(the)h(release)
+-h(of)f Fj(bzip2-)150 1562 y(0.1)19 b Fl(in)g(August)i(1997)h(has)e(sho)
+-m(wn)f(complexities)h(in)f(the)h(\014le)g(format)g(whic)m(h)f(slo)m(w)h
+-(do)m(wn)g(decompression)150 1672 y(and,)30 b(in)f(retrosp)s(ect,)i
+-(are)g(unnecessary)-8 b(.)40 b(These)31 b(are:)225 1829
+-y Fi(\017)60 b Fl(The)20 b(run-length)g(enco)s(der,)i(whic)m(h)e(is)g
+-(the)h(\014rst)f(of)h(the)g(compression)f(transformations,)i(is)e(en)m
+-(tirely)330 1938 y(irrelev)-5 b(an)m(t.)63 b(The)38 b(original)e(purp)s
+-(ose)g(w)m(as)j(to)g(protect)g(the)f(sorting)g(algorithm)f(from)g(the)i
+-(v)m(ery)330 2048 y(w)m(orst)h(case)h(input:)58 b(a)41
+-b(string)e(of)h(rep)s(eated)g(sym)m(b)s(ols.)68 b(But)40
+-b(algorithm)f(steps)h(Q6a)h(and)e(Q6b)330 2157 y(in)30
+-b(the)i(original)e(Burro)m(ws-Wheeler)i(tec)m(hnical)g(rep)s(ort)f
+-(\(SR)m(C-124\))i(sho)m(w)f(ho)m(w)g(rep)s(eats)g(can)g(b)s(e)330
+-2267 y(handled)c(without)i(di\016cult)m(y)f(in)g(blo)s(c)m(k)h
+-(sorting.)225 2409 y Fi(\017)60 b Fl(The)30 b(randomisation)e(mec)m
+-(hanism)i(do)s(esn't)g(really)f(need)h(to)g(b)s(e)g(there.)41
+-b(Udi)29 b(Man)m(b)s(er)h(and)f(Gene)330 2518 y(My)m(ers)j(published)c
+-(a)33 b(su\016x)e(arra)m(y)h(construction)f(algorithm)g(a)h(few)g(y)m
+-(ears)h(bac)m(k,)g(whic)m(h)d(can)j(b)s(e)330 2628 y(emplo)m(y)m(ed)27
+-b(to)h(sort)g(an)m(y)f(blo)s(c)m(k,)h(no)f(matter)h(ho)m(w)f(rep)s
+-(etitiv)m(e,)h(in)d(O\(N)j(log)f(N\))h(time.)39 b(Subsequen)m(t)330
+-2737 y(w)m(ork)25 b(b)m(y)f(Kunihik)m(o)f(Sadak)-5 b(ane)24
+-b(has)h(pro)s(duced)e(a)i(deriv)-5 b(ativ)m(e)24 b(O\(N)h(\(log)g(N\))p
+-Fj(^)p Fl(2\))h(algorithm)d(whic)m(h)330 2847 y(usually)28
+-b(outp)s(erforms)h(the)i(Man)m(b)s(er-My)m(ers)g(algorithm.)330
+-2988 y(I)g(could)g(ha)m(v)m(e)i(c)m(hanged)f(to)g(Sadak)-5
+-b(ane's)32 b(algorithm,)f(but)g(I)g(\014nd)f(it)h(to)h(b)s(e)f(slo)m(w)
+-m(er)h(than)f Fj(bzip2)p Fl('s)330 3098 y(existing)38
+-b(algorithm)g(for)h(most)h(inputs,)f(and)g(the)g(randomisation)f(mec)m
+-(hanism)g(protects)i(ade-)330 3208 y(quately)34 b(against)f(bad)g
+-(cases.)52 b(I)33 b(didn't)f(think)g(it)i(w)m(as)g(a)g(go)s(o)s(d)f
+-(tradeo\013)i(to)f(mak)m(e.)51 b(P)m(artly)34 b(this)330
+-3317 y(is)39 b(due)h(to)h(the)f(fact)h(that)g(I)f(w)m(as)g(not)h(\015o)
+-s(o)s(ded)e(with)g(email)g(complain)m(ts)g(ab)s(out)h
+-Fj(bzip2-0.1)p Fl('s)330 3427 y(p)s(erformance)30 b(on)g(rep)s(etitiv)m
+-(e)g(data,)h(so)g(p)s(erhaps)d(it)i(isn't)g(a)h(problem)d(for)j(real)f
+-(inputs.)330 3568 y(Probably)i(the)h(b)s(est)g(long-term)g(solution,)g
+-(and)g(the)g(one)h(I)f(ha)m(v)m(e)h(incorp)s(orated)e(in)m(to)i(0.9.5)h
+-(and)330 3678 y(ab)s(o)m(v)m(e,)42 b(is)c(to)h(use)f(the)h(existing)f
+-(sorting)g(algorithm)f(initially)-8 b(,)38 b(and)g(fall)f(bac)m(k)i(to)
+-h(a)f(O\(N)f(\(log)330 3787 y(N\))p Fj(^)p Fl(2\))31
+-b(algorithm)f(if)f(the)i(standard)e(algorithm)h(gets)h(in)m(to)f
+-(di\016culties.)225 3929 y Fi(\017)60 b Fl(The)31 b(compressed)f
+-(\014le)g(format)i(w)m(as)f(nev)m(er)h(designed)d(to)j(b)s(e)f(handled)
+-e(b)m(y)i(a)g(library)-8 b(,)29 b(and)i(I)g(ha)m(v)m(e)330
+-4039 y(had)d(to)i(jump)e(though)g(some)i(ho)s(ops)e(to)i(pro)s(duce)e
+-(an)h(e\016cien)m(t)g(implemen)m(tation)f(of)h(decompres-)330
+-4148 y(sion.)38 b(It's)26 b(a)h(bit)e(hairy)-8 b(.)38
+-b(T)-8 b(ry)26 b(passing)f Fj(decompress.c)d Fl(through)k(the)g(C)f
+-(prepro)s(cessor)g(and)h(y)m(ou'll)330 4258 y(see)32
+-b(what)g(I)f(mean.)45 b(Muc)m(h)32 b(of)g(this)e(complexit)m(y)i(could)
+-f(ha)m(v)m(e)i(b)s(een)e(a)m(v)m(oided)h(if)e(the)i(compressed)330
+-4367 y(size)e(of)h(eac)m(h)g(blo)s(c)m(k)f(of)h(data)g(w)m(as)g
+-(recorded)f(in)f(the)h(data)h(stream.)225 4509 y Fi(\017)60
+-b Fl(An)30 b(Adler-32)g(c)m(hec)m(ksum,)i(rather)e(than)g(a)h(CR)m(C32)
+-g(c)m(hec)m(ksum,)g(w)m(ould)e(b)s(e)h(faster)h(to)g(compute.)150
+-4698 y(It)e(w)m(ould)f(b)s(e)g(fair)g(to)h(sa)m(y)h(that)g(the)f
+-Fj(bzip2)e Fl(format)i(w)m(as)h(frozen)f(b)s(efore)f(I)h(prop)s(erly)d
+-(and)j(fully)d(under-)150 4807 y(sto)s(o)s(d)k(the)h(p)s(erformance)e
+-(consequences)i(of)g(doing)e(so.)150 4964 y(Impro)m(v)m(emen)m(ts)d
+-(whic)m(h)e(I)i(w)m(as)g(able)f(to)h(incorp)s(orate)f(in)m(to)g(0.9.0,)
+-k(despite)24 b(using)g(the)i(same)g(\014le)e(format,)150
+-5074 y(are:)225 5230 y Fi(\017)60 b Fl(Single)30 b(arra)m(y)i(implemen)
+-m(tation)e(of)h(the)h(in)m(v)m(erse)f(BWT.)h(This)e(signi\014can)m(tly)
+-f(sp)s(eeds)i(up)f(decom-)330 5340 y(pression,)f(presumably)f(b)s
+-(ecause)i(it)g(reduces)g(the)h(n)m(um)m(b)s(er)e(of)i(cac)m(he)h
+-(misses.)p eop
+-%%Page: 32 33
+-32 32 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586
+-b(32)225 299 y Fi(\017)60 b Fl(F)-8 b(aster)27 b(in)m(v)m(erse)e(MTF)h
+-(transform)f(for)g(large)h(MTF)f(v)-5 b(alues.)39 b(The)25
+-b(new)g(implemen)m(tation)f(is)g(based)330 408 y(on)30
+-b(the)h(notion)f(of)g(sliding)e(blo)s(c)m(ks)h(of)i(v)-5
+-b(alues.)225 544 y Fi(\017)60 b Fj(bzip2-0.9.0)24 b Fl(no)m(w)k(reads)f
+-(and)f(writes)h(\014les)f(with)g Fj(fread)g Fl(and)h
+-Fj(fwrite)p Fl(;)f(v)m(ersion)h(0.1)i(used)d Fj(putc)330
+-653 y Fl(and)k Fj(getc)p Fl(.)39 b(Duh!)h(W)-8 b(ell,)31
+-b(y)m(ou)f(liv)m(e)g(and)g(learn.)150 836 y(F)-8 b(urther)30
+-b(ahead,)g(it)f(w)m(ould)g(b)s(e)g(nice)h(to)g(b)s(e)g(able)f(to)i(do)e
+-(random)g(access)j(in)m(to)d(\014les.)40 b(This)28 b(will)f(require)150
+-945 y(some)k(careful)e(design)h(of)g(compressed)g(\014le)g(formats.)150
+-1227 y Fk(4.2)68 b(P)l(ortabilit)l(y)47 b(issues)150
+-1419 y Fl(After)36 b(some)g(consideration,)g(I)f(ha)m(v)m(e)i(decided)d
+-(not)i(to)g(use)g(GNU)g Fj(autoconf)d Fl(to)j(con\014gure)g(0.9.5)h(or)
+-150 1529 y(1.0.)150 1686 y Fj(autoconf)p Fl(,)31 b(admirable)g(and)h(w)
+-m(onderful)f(though)i(it)f(is,)h(mainly)d(assists)j(with)e(p)s
+-(ortabilit)m(y)g(problems)150 1795 y(b)s(et)m(w)m(een)f(Unix-lik)m(e)d
+-(platforms.)40 b(But)29 b Fj(bzip2)f Fl(do)s(esn't)h(ha)m(v)m(e)h(m)m
+-(uc)m(h)f(in)f(the)h(w)m(a)m(y)h(of)g(p)s(ortabilit)m(y)d(prob-)150
+-1905 y(lems)35 b(on)h(Unix;)j(most)d(of)g(the)h(di\016culties)d(app)s
+-(ear)h(when)g(p)s(orting)g(to)i(the)f(Mac,)j(or)d(to)h(Microsoft's)150
+-2015 y(op)s(erating)26 b(systems.)40 b Fj(autoconf)25
+-b Fl(do)s(esn't)h(help)g(in)f(those)j(cases,)h(and)d(brings)f(in)g(a)j
+-(whole)e(load)g(of)h(new)150 2124 y(complexit)m(y)-8
+-b(.)150 2281 y(Most)28 b(p)s(eople)f(should)f(b)s(e)h(able)g(to)h
+-(compile)e(the)i(library)d(and)i(program)h(under)e(Unix)g(straigh)m(t)i
+-(out-of-)150 2391 y(the-b)s(o)m(x,)j(so)g(to)g(sp)s(eak,)f(esp)s
+-(ecially)f(if)g(y)m(ou)i(ha)m(v)m(e)g(a)g(v)m(ersion)f(of)g(GNU)h(C)f
+-(a)m(v)-5 b(ailable.)150 2547 y(There)32 b(are)h(a)g(couple)f(of)h
+-Fj(__inline__)d Fl(directiv)m(es)i(in)f(the)i(co)s(de.)48
+-b(GNU)33 b(C)f(\()p Fj(gcc)p Fl(\))g(should)f(b)s(e)h(able)g(to)150
+-2657 y(handle)24 b(them.)39 b(If)25 b(y)m(ou're)i(not)e(using)g(GNU)h
+-(C,)f(y)m(our)h(C)f(compiler)f(shouldn't)g(see)i(them)f(at)i(all.)38
+-b(If)25 b(y)m(our)150 2767 y(compiler)k(do)s(es,)i(for)g(some)g
+-(reason,)h(see)f(them)g(and)f(do)s(esn't)h(lik)m(e)f(them,)i(just)e
+-Fj(#define)f(__inline__)150 2876 y Fl(to)37 b(b)s(e)f
+-Fj(/*)30 b(*/)p Fl(.)58 b(One)36 b(easy)h(w)m(a)m(y)g(to)h(do)e(this)f
+-(is)h(to)h(compile)e(with)g(the)i(\015ag)g Fj(-D__inline__=)p
+-Fl(,)d(whic)m(h)150 2986 y(should)28 b(b)s(e)i(understo)s(o)s(d)f(b)m
+-(y)h(most)h(Unix)e(compilers.)150 3143 y(If)35 b(y)m(ou)g(still)e(ha)m
+-(v)m(e)j(di\016culties,)e(try)h(compiling)e(with)g(the)j(macro)f
+-Fj(BZ_STRICT_ANSI)c Fl(de\014ned.)54 b(This)150 3252
+-y(should)28 b(enable)i(y)m(ou)h(to)g(build)d(the)i(library)e(in)h(a)i
+-(strictly)f(ANSI)g(complian)m(t)f(en)m(vironmen)m(t.)41
+-b(Building)150 3362 y(the)25 b(program)f(itself)f(lik)m(e)g(this)h(is)f
+-(dangerous)h(and)g(not)g(supp)s(orted,)g(since)g(y)m(ou)h(remo)m(v)m(e)
+-g Fj(bzip2)p Fl('s)e(c)m(hec)m(ks)150 3471 y(against)30
+-b(compressing)f(directories,)g(sym)m(b)s(olic)g(links,)f(devices,)i
+-(and)f(other)h(not-really-a-\014le)g(en)m(tities.)150
+-3581 y(This)f(could)g(cause)i(\014lesystem)f(corruption!)150
+-3738 y(One)e(other)i(thing:)39 b(if)27 b(y)m(ou)j(create)g(a)f
+-Fj(bzip2)f Fl(binary)f(for)i(public)d(distribution,)g(please)i(try)h
+-(and)g(link)d(it)150 3847 y(statically)g(\()p Fj(gcc)k(-s)p
+-Fl(\).)39 b(This)25 b(a)m(v)m(oids)i(all)f(sorts)h(of)g(library-v)m
+-(ersion)d(issues)h(that)i(others)g(ma)m(y)g(encoun)m(ter)150
+-3957 y(later)j(on.)150 4114 y(If)f(y)m(ou)g(build)e Fj(bzip2)h
+-Fl(on)h(Win32,)h(y)m(ou)f(m)m(ust)g(set)h Fj(BZ_UNIX)e
+-Fl(to)i(0)f(and)g Fj(BZ_LCCWIN32)d Fl(to)k(1,)g(in)e(the)i(\014le)150
+-4223 y Fj(bzip2.c)p Fl(,)f(b)s(efore)h(compiling.)38
+-b(Otherwise)29 b(the)i(resulting)d(binary)h(w)m(on't)i(w)m(ork)f
+-(correctly)-8 b(.)150 4505 y Fk(4.3)68 b(Rep)t(orting)46
+-b(bugs)150 4698 y Fl(I)25 b(tried)f(prett)m(y)i(hard)e(to)i(mak)m(e)g
+-(sure)f Fj(bzip2)e Fl(is)i(bug)f(free,)j(b)s(oth)d(b)m(y)h(design)f
+-(and)h(b)m(y)g(testing.)39 b(Hop)s(efully)150 4807 y(y)m(ou'll)29
+-b(nev)m(er)i(need)f(to)h(read)g(this)e(section)h(for)h(real.)150
+-4964 y(Nev)m(ertheless,)36 b(if)c Fj(bzip2)h Fl(dies)g(with)f(a)i
+-(segmen)m(tation)h(fault,)g(a)f(bus)f(error)g(or)h(an)g(in)m(ternal)e
+-(assertion)150 5074 y(failure,)i(it)h(will)d(ask)j(y)m(ou)g(to)g(email)
+-f(me)h(a)g(bug)f(rep)s(ort.)54 b(Exp)s(erience)33 b(with)h(v)m(ersion)g
+-(0.1)i(sho)m(ws)e(that)150 5183 y(almost)c(all)g(these)h(problems)d
+-(can)j(b)s(e)f(traced)h(to)g(either)f(compiler)e(bugs)i(or)g(hardw)m
+-(are)g(problems.)225 5340 y Fi(\017)60 b Fl(Recompile)22
+-b(the)h(program)g(with)f(no)h(optimisation,)g(and)f(see)i(if)e(it)g(w)m
+-(orks.)39 b(And/or)22 b(try)h(a)g(di\013eren)m(t)p eop
+-%%Page: 33 34
+-33 33 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586
+-b(33)330 299 y(compiler.)77 b(I)43 b(heard)f(all)g(sorts)h(of)h
+-(stories)e(ab)s(out)h(v)-5 b(arious)42 b(\015a)m(v)m(ours)h(of)h(GNU)f
+-(C)g(\(and)g(other)330 408 y(compilers\))20 b(generating)i(bad)e(co)s
+-(de)i(for)f Fj(bzip2)p Fl(,)h(and)f(I'v)m(e)h(run)e(across)i(t)m(w)m(o)
+-g(suc)m(h)f(examples)g(m)m(yself.)330 606 y(2.7.X)35
+-b(v)m(ersions)e(of)g(GNU)h(C)f(are)h(kno)m(wn)f(to)h(generate)h(bad)d
+-(co)s(de)i(from)f(time)g(to)h(time,)g(at)g(high)330 716
+-y(optimisation)20 b(lev)m(els.)37 b(If)21 b(y)m(ou)g(get)i(problems,)e
+-(try)g(using)f(the)i(\015ags)f Fj(-O2)f(-fomit-frame-pointer)330
+-825 y(-fno-strength-reduce)p Fl(.)35 b(Y)-8 b(ou)31 b(should)d(sp)s
+-(eci\014cally)h Fc(not)j Fl(use)e Fj(-funroll-loops)p
+-Fl(.)330 1023 y(Y)-8 b(ou)38 b(ma)m(y)g(notice)g(that)g(the)g(Mak)m
+-(e\014le)g(runs)e(six)g(tests)i(as)g(part)f(of)h(the)g(build)c(pro)s
+-(cess.)62 b(If)37 b(the)330 1132 y(program)43 b(passes)g(all)f(of)h
+-(these,)k(it's)c(a)h(prett)m(y)f(go)s(o)s(d)g(\(but)g(not)g(100\045\))i
+-(indication)c(that)j(the)330 1242 y(compiler)29 b(has)h(done)g(its)g
+-(job)g(correctly)-8 b(.)225 1440 y Fi(\017)60 b Fl(If)33
+-b Fj(bzip2)f Fl(crashes)i(randomly)-8 b(,)33 b(and)g(the)h(crashes)g
+-(are)g(not)g(rep)s(eatable,)g(y)m(ou)g(ma)m(y)g(ha)m(v)m(e)h(a)f
+-(\015aky)330 1549 y(memory)k(subsystem.)64 b Fj(bzip2)37
+-b Fl(really)g(hammers)h(y)m(our)g(memory)g(hierarc)m(h)m(y)-8
+-b(,)41 b(and)d(if)f(it's)h(a)h(bit)330 1659 y(marginal,)33
+-b(y)m(ou)h(ma)m(y)g(get)h(these)f(problems.)49 b(Ditto)34
+-b(if)f(y)m(our)h(disk)e(or)h(I/O)h(subsystem)e(is)h(slo)m(wly)330
+-1768 y(failing.)39 b(Y)-8 b(up,)30 b(this)f(really)g(do)s(es)h(happ)s
+-(en.)330 1966 y(T)-8 b(ry)28 b(using)f(a)i(di\013eren)m(t)f(mac)m(hine)
+-g(of)h(the)g(same)f(t)m(yp)s(e,)i(and)e(see)h(if)e(y)m(ou)i(can)g(rep)s
+-(eat)g(the)f(problem.)225 2163 y Fi(\017)60 b Fl(This)21
+-b(isn't)i(really)f(a)h(bug,)i(but)d(...)39 b(If)23 b
+-Fj(bzip2)f Fl(tells)g(y)m(ou)h(y)m(our)h(\014le)e(is)g(corrupted)h(on)g
+-(decompression,)330 2273 y(and)29 b(y)m(ou)g(obtained)f(the)i(\014le)e
+-(via)h(FTP)-8 b(,)29 b(there)h(is)e(a)h(p)s(ossibilit)m(y)d(that)k(y)m
+-(ou)f(forgot)h(to)g(tell)e(FTP)h(to)330 2383 y(do)23
+-b(a)g(binary)e(mo)s(de)i(transfer.)38 b(That)23 b(absolutely)f(will)e
+-(cause)j(the)h(\014le)e(to)h(b)s(e)g(non-decompressible.)330
+-2492 y(Y)-8 b(ou'll)30 b(ha)m(v)m(e)h(to)g(transfer)f(it)g(again.)150
+-2737 y(If)i(y)m(ou'v)m(e)h(incorp)s(orated)e Fj(libbzip2)f
+-Fl(in)m(to)i(y)m(our)g(o)m(wn)g(program)g(and)g(are)g(getting)h
+-(problems,)e(please,)150 2847 y(please,)d(please,)h(c)m(hec)m(k)g(that)
+-f(the)g(parameters)g(y)m(ou)g(are)g(passing)f(in)f(calls)h(to)h(the)g
+-(library)-8 b(,)26 b(are)j(correct,)150 2956 y(and)e(in)f(accordance)k
+-(with)c(what)i(the)g(do)s(cumen)m(tation)f(sa)m(ys)h(is)f(allo)m(w)m
+-(able.)39 b(I)28 b(ha)m(v)m(e)h(tried)e(to)h(mak)m(e)h(the)150
+-3066 y(library)f(robust)i(against)g(suc)m(h)g(problems,)f(but)h(I'm)g
+-(sure)g(I)g(ha)m(v)m(en't)h(succeeded.)150 3223 y(Finally)-8
+-b(,)32 b(if)g(the)h(ab)s(o)m(v)m(e)i(commen)m(ts)e(don't)g(help,)g(y)m
+-(ou'll)f(ha)m(v)m(e)i(to)g(send)e(me)h(a)g(bug)g(rep)s(ort.)48
+-b(No)m(w,)34 b(it's)150 3332 y(just)c(amazing)g(ho)m(w)h(man)m(y)f(p)s
+-(eople)g(will)d(send)j(me)g(a)h(bug)f(rep)s(ort)g(sa)m(ying)g
+-(something)g(lik)m(e)481 3483 y(bzip2)f(crashed)h(with)f(segmen)m
+-(tation)j(fault)e(on)g(m)m(y)g(mac)m(hine)150 3640 y(and)h(absolutely)f
+-(nothing)h(else.)44 b(Needless)32 b(to)g(sa)m(y)-8 b(,)33
+-b(a)f(suc)m(h)f(a)h(rep)s(ort)f(is)g Fc(totally)-8 b(,)32
+-b(utterly)-8 b(,)32 b(completely)150 3750 y(and)40 b(comprehensiv)m
+-(ely)g(100\045)h(useless;)46 b(a)41 b(w)m(aste)g(of)g(y)m(our)g(time,)i
+-(m)m(y)e(time,)i(and)e(net)g(bandwidth)p Fl(.)150 3859
+-y(With)31 b(no)h(details)f(at)i(all,)e(there's)h(no)g(w)m(a)m(y)h(I)f
+-(can)g(p)s(ossibly)d(b)s(egin)h(to)j(\014gure)e(out)i(what)e(the)i
+-(problem)150 3969 y(is.)150 4126 y(The)d(rules)e(of)i(the)g(game)h
+-(are:)41 b(facts,)32 b(facts,)f(facts.)41 b(Don't)31
+-b(omit)f(them)g(b)s(ecause)g Fj(")p Fl(oh,)g(they)g(w)m(on't)h(b)s(e)
+-150 4235 y(relev)-5 b(an)m(t)p Fj(")p Fl(.)41 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.)61
+-b(Op)s(erating)29 b(system)h(v)m(ersion.)481 4490 y(Exact)h(v)m(ersion)
+-f(of)h Fj(bzip2)e Fl(\(do)h Fj(bzip2)47 b(-V)p Fl(\).)481
+-4594 y(Exact)31 b(v)m(ersion)f(of)h(the)f(compiler)f(used.)481
+-4698 y(Flags)i(passed)e(to)j(the)e(compiler.)150 4854
+-y(Ho)m(w)m(ev)m(er,)i(the)d(most)h(imp)s(ortan)m(t)f(single)f(thing)g
+-(that)i(will)d(help)h(me)h(is)f(the)i(\014le)e(that)i(y)m(ou)g(w)m(ere)
+-g(trying)150 4964 y(to)f(compress)f(or)g(decompress)g(at)h(the)f(time)g
+-(the)g(problem)f(happ)s(ened.)38 b(Without)28 b(that,)h(m)m(y)g(abilit)
+-m(y)d(to)150 5074 y(do)k(an)m(ything)g(more)h(than)f(sp)s(eculate)g(ab)
+-s(out)g(the)g(cause,)i(is)d(limited.)150 5230 y(Please)34
+-b(remem)m(b)s(er)f(that)h(I)f(connect)i(to)f(the)g(In)m(ternet)g(with)e
+-(a)i(mo)s(dem,)g(so)f(y)m(ou)h(should)e(con)m(tact)k(me)150
+-5340 y(b)s(efore)30 b(mailing)e(me)j(h)m(uge)f(\014les.)p
+-eop
+-%%Page: 34 35
+-34 34 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586
+-b(34)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 Fj(bzip2)34 b Fl(is)h(a)h(resource)g(hog.)
+-56 b(It)36 b(soaks)g(up)f(large)g(amoun)m(ts)h(of)g(CPU)f(cycles)h(and)
+-f(memory)-8 b(.)57 b(Also,)36 b(it)150 601 y(giv)m(es)26
+-b(v)m(ery)h(large)f(latencies.)39 b(In)25 b(the)h(w)m(orst)g(case,)i(y)
+-m(ou)f(can)f(feed)g(man)m(y)g(megab)m(ytes)h(of)f(uncompressed)150
+-711 y(data)45 b(in)m(to)e(the)i(library)c(b)s(efore)j(getting)g(an)m(y)
+-g(compressed)g(output,)j(so)d(this)f(probably)f(rules)h(out)150
+-820 y(applications)29 b(requiring)e(in)m(teractiv)m(e)32
+-b(b)s(eha)m(viour.)150 977 y(These)38 b(aren't)h(faults)e(of)h(m)m(y)g
+-(implemen)m(tation,)h(I)f(hop)s(e,)i(but)d(more)h(an)g(in)m(trinsic)e
+-(prop)s(ert)m(y)h(of)i(the)150 1087 y(Burro)m(ws-Wheeler)30
+-b(transform)g(\(unfortunately\).)40 b(Ma)m(yb)s(e)31
+-b(this)e(isn't)h(what)g(y)m(ou)h(w)m(an)m(t.)150 1244
+-y(If)h(y)m(ou)h(w)m(an)m(t)g(a)g(compressor)g(and/or)f(library)e(whic)m
+-(h)h(is)h(faster,)i(uses)e(less)g(memory)g(but)g(gets)h(prett)m(y)150
+-1353 y(go)s(o)s(d)e(compression,)g(and)g(has)h(minimal)c(latency)-8
+-b(,)33 b(consider)e(Jean-loup)f(Gailly's)g(and)h(Mark)h(Adler's)150
+-1463 y(w)m(ork,)f Fj(zlib-1.1.2)c Fl(and)j Fj(gzip-1.2.4)p
+-Fl(.)38 b(Lo)s(ok)31 b(for)f(them)g(at)150 1620 y Fj
+-(http://www.cdrom.com/pub)o(/inf)o(ozip)o(/zl)o(ib)24
+-b Fl(and)30 b Fj(http://www.gzip.org)25 b Fl(resp)s(ectiv)m(ely)-8
+-b(.)150 1776 y(F)g(or)32 b(something)f(faster)i(and)e(ligh)m(ter)f
+-(still,)h(y)m(ou)g(migh)m(t)h(try)f(Markus)h(F)g(X)f(J)h(Ob)s(erh)m
+-(umer's)d Fj(LZO)i Fl(real-)150 1886 y(time)f
+-(compression/decompression)f(library)-8 b(,)28 b(at)150
+-1996 y Fj(http://wildsau.idv.uni-l)o(inz.)o(ac.a)o(t/m)o(fx/l)o(zo.h)o
+-(tml)o Fl(.)150 2152 y(If)38 b(y)m(ou)h(w)m(an)m(t)g(to)h(use)e(the)g
+-Fj(bzip2)g Fl(algorithms)f(to)i(compress)f(small)g(blo)s(c)m(ks)f(of)i
+-(data,)j(64k)d(b)m(ytes)g(or)150 2262 y(smaller,)i(for)e(example)g(on)h
+-(an)f(on-the-\015y)h(disk)e(compressor,)k(y)m(ou'd)e(b)s(e)f(w)m(ell)g
+-(advised)f(not)i(to)g(use)150 2372 y(this)i(library)-8
+-b(.)77 b(Instead,)47 b(I'v)m(e)d(made)f(a)h(sp)s(ecial)e(library)f
+-(tuned)h(for)h(that)h(kind)d(of)j(use.)79 b(It's)43 b(part)150
+-2481 y(of)d Fj(e2compr-0.40)p Fl(,)f(an)g(on-the-\015y)h(disk)e
+-(compressor)h(for)h(the)f(Lin)m(ux)f Fj(ext2)h Fl(\014lesystem.)67
+-b(Lo)s(ok)40 b(at)150 2591 y Fj(http://www.netspace.net.)o(au/~)o(reit)
+-o(er/)o(e2co)o(mpr)p Fl(.)150 2880 y Fk(4.5)68 b(T)-11
+-b(esting)150 3072 y Fl(A)30 b(record)h(of)f(the)h(tests)g(I'v)m(e)g
+-(done.)150 3229 y(First,)f(some)h(data)g(sets:)225 3386
+-y Fi(\017)60 b Fl(B:)32 b(a)f(directory)f(con)m(taining)h(6001)i
+-(\014les,)d(one)h(for)g(ev)m(ery)h(length)e(in)g(the)h(range)g(0)h(to)f
+-(6000)i(b)m(ytes.)330 3496 y(The)d(\014les)f(con)m(tain)i(random)e(lo)m
+-(w)m(ercase)j(letters.)41 b(18.7)32 b(megab)m(ytes.)225
+-3633 y Fi(\017)60 b Fl(H:)36 b(m)m(y)f(home)h(directory)f(tree.)56
+-b(Do)s(cumen)m(ts,)38 b(source)d(co)s(de,)i(mail)d(\014les,)i
+-(compressed)f(data.)57 b(H)330 3743 y(con)m(tains)39
+-b(B,)h(and)f(also)g(a)g(directory)g(of)g(\014les)f(designed)g(as)i(b)s
+-(oundary)d(cases)j(for)f(the)g(sorting;)330 3853 y(mostly)30
+-b(v)m(ery)h(rep)s(etitiv)m(e,)f(nast)m(y)h(\014les.)39
+-b(565)32 b(megab)m(ytes.)225 3990 y Fi(\017)60 b Fl(A:)43
+-b(directory)f(tree)i(holding)d(v)-5 b(arious)41 b(applications)g(built)
+-g(from)h(source:)66 b Fj(egcs)p Fl(,)45 b Fj(gcc-2.8.1)p
+-Fl(,)330 4100 y(KDE,)31 b(GTK,)f(Octa)m(v)m(e,)j(etc.)41
+-b(2200)33 b(megab)m(ytes.)150 4285 y(The)i(tests)g(conducted)g(are)h
+-(as)f(follo)m(ws.)54 b(Eac)m(h)36 b(test)g(means)f(compressing)f(\(a)h
+-(cop)m(y)h(of)7 b(\))36 b(eac)m(h)g(\014le)e(in)150 4394
+-y(the)d(data)g(set,)g(decompressing)e(it)h(and)g(comparing)f(it)h
+-(against)h(the)g(original.)150 4551 y(First,)26 b(a)g(bunc)m(h)f(of)h
+-(tests)h(with)d(blo)s(c)m(k)h(sizes)h(and)f(in)m(ternal)g(bu\013er)f
+-(sizes)i(set)g(v)m(ery)g(small,)g(to)g(detect)i(an)m(y)150
+-4661 y(problems)g(with)g(the)i(blo)s(c)m(king)f(and)g(bu\013ering)e
+-(mec)m(hanisms.)40 b(This)28 b(required)g(mo)s(difying)f(the)j(source)
+-150 4770 y(co)s(de)h(so)f(as)h(to)g(try)f(to)h(break)g(it.)199
+-4927 y(1.)61 b(Data)32 b(set)f(H,)g(with)e(bu\013er)g(size)h(of)h(1)g
+-(b)m(yte,)g(and)f(blo)s(c)m(k)g(size)g(of)g(23)i(b)m(ytes.)199
+-5065 y(2.)61 b(Data)32 b(set)f(B,)g(bu\013er)e(sizes)h(1)h(b)m(yte,)g
+-(blo)s(c)m(k)f(size)g(1)h(b)m(yte.)199 5202 y(3.)61 b(As)30
+-b(\(2\))i(but)d(small-mo)s(de)g(decompression.)199 5340
+-y(4.)61 b(As)30 b(\(2\))i(with)d(blo)s(c)m(k)h(size)g(2)h(b)m(ytes.)p
+-eop
+-%%Page: 35 36
+-35 35 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586
+-b(35)199 299 y(5.)61 b(As)30 b(\(2\))i(with)d(blo)s(c)m(k)h(size)g(3)h
+-(b)m(ytes.)199 431 y(6.)61 b(As)30 b(\(2\))i(with)d(blo)s(c)m(k)h(size)
+-g(4)h(b)m(ytes.)199 564 y(7.)61 b(As)30 b(\(2\))i(with)d(blo)s(c)m(k)h
+-(size)g(5)h(b)m(ytes.)199 697 y(8.)61 b(As)30 b(\(2\))i(with)d(blo)s(c)
+-m(k)h(size)g(6)h(b)m(ytes)g(and)e(small-mo)s(de)g(decompression.)199
+-829 y(9.)61 b(H)30 b(with)g(bu\013er)f(size)h(of)h(1)f(b)m(yte,)i(but)d
+-(normal)h(blo)s(c)m(k)g(size)g(\(up)f(to)j(900000)h(b)m(ytes\).)150
+-1009 y(Then)c(some)i(tests)g(with)e(unmo)s(di\014ed)f(source)i(co)s
+-(de.)199 1166 y(1.)61 b(H,)31 b(all)e(settings)h(normal.)199
+-1299 y(2.)61 b(As)30 b(\(1\),)i(with)d(small-mo)s(de)g(decompress.)199
+-1431 y(3.)61 b(H,)31 b(compress)f(with)f(\015ag)i Fj(-1)p
+-Fl(.)199 1564 y(4.)61 b(H,)31 b(compress)f(with)f(\015ag)i
+-Fj(-s)p Fl(,)f(decompress)g(with)f(\015ag)i Fj(-s)p Fl(.)199
+-1697 y(5.)61 b(F)-8 b(orw)m(ards)33 b(compatibilit)m(y:)45
+-b(H,)33 b Fj(bzip2-0.1pl2)d Fl(compressing,)j Fj(bzip2-0.9.5)d
+-Fl(decompressing,)330 1806 y(all)f(settings)i(normal.)199
+-1939 y(6.)61 b(Bac)m(kw)m(ards)23 b(compatibilit)m(y:)35
+-b(H,)23 b Fj(bzip2-0.9.5)c Fl(compressing,)k Fj(bzip2-0.1pl2)c
+-Fl(decompressing,)330 2048 y(all)29 b(settings)i(normal.)199
+-2181 y(7.)61 b(Bigger)31 b(tests:)41 b(A,)31 b(all)e(settings)i
+-(normal.)199 2314 y(8.)61 b(As)30 b(\(7\),)i(using)d(the)i(fallbac)m(k)
+-e(\(Sadak)-5 b(ane-lik)m(e\))31 b(sorting)f(algorithm.)199
+-2446 y(9.)61 b(As)30 b(\(8\),)i(compress)e(with)f(\015ag)i
+-Fj(-1)p Fl(,)f(decompress)g(with)f(\015ag)i Fj(-s)p Fl(.)154
+-2579 y(10.)61 b(H,)31 b(using)e(the)h(fallbac)m(k)g(sorting)g
+-(algorithm.)154 2711 y(11.)61 b(F)-8 b(orw)m(ards)33
+-b(compatibilit)m(y:)45 b(A,)33 b Fj(bzip2-0.1pl2)d Fl(compressing,)j
+-Fj(bzip2-0.9.5)d Fl(decompressing,)330 2821 y(all)f(settings)i(normal.)
+-154 2954 y(12.)61 b(Bac)m(kw)m(ards)23 b(compatibilit)m(y:)35
+-b(A,)23 b Fj(bzip2-0.9.5)c Fl(compressing,)k Fj(bzip2-0.1pl2)c
+-Fl(decompressing,)330 3063 y(all)29 b(settings)i(normal.)154
+-3196 y(13.)61 b(Misc)39 b(test:)58 b(ab)s(out)39 b(400)h(megab)m(ytes)h
+-(of)e Fj(.tar)f Fl(\014les)f(with)h Fj(bzip2)f Fl(compiled)h(with)f
+-(Chec)m(k)m(er)j(\(a)330 3305 y(memory)30 b(access)i(error)e(detector,)
+-i(lik)m(e)e(Purify\).)154 3438 y(14.)61 b(Misc)30 b(tests)h(to)g(mak)m
+-(e)h(sure)d(it)h(builds)e(and)h(runs)g(ok)i(on)f(non-Lin)m(ux/x86)g
+-(platforms.)150 3618 y(These)35 b(tests)h(w)m(ere)f(conducted)g(on)g(a)
+-h(225)g(MHz)g(IDT)f(WinChip)d(mac)m(hine,)k(running)d(Lin)m(ux)g
+-(2.0.36.)150 3728 y(They)d(represen)m(t)g(nearly)g(a)h(w)m(eek)g(of)f
+-(con)m(tin)m(uous)g(computation.)41 b(All)29 b(tests)i(completed)f
+-(successfully)-8 b(.)150 4003 y Fk(4.6)68 b(F)-11 b(urther)44
+-b(reading)150 4196 y Fj(bzip2)28 b Fl(is)h(not)h(researc)m(h)g(w)m
+-(ork,)g(in)e(the)i(sense)g(that)g(it)f(do)s(esn't)g(presen)m(t)h(an)m
+-(y)g(new)f(ideas.)40 b(Rather,)30 b(it's)150 4306 y(an)g(engineering)f
+-(exercise)i(based)f(on)g(existing)g(ideas.)150 4463 y(F)-8
+-b(our)31 b(do)s(cumen)m(ts)f(describ)s(e)e(essen)m(tially)i(all)f(the)i
+-(ideas)e(b)s(ehind)f Fj(bzip2)p Fl(:)390 4614 y Fj(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: 36 37
+-36 36 bop 150 -116 a Fl(Chapter)30 b(4:)41 b(Miscellanea)2586
+-b(36)390 299 y Fj(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 Fl(The)29
+-b(follo)m(wing)f(pap)s(er)g(giv)m(es)h(v)-5 b(aluable)28
+-b(additional)g(insigh)m(ts)f(in)m(to)j(the)f(algorithm,)g(but)g(is)f
+-(not)i(imme-)150 2122 y(diately)g(the)g(basis)f(of)i(an)m(y)g(co)s(de)f
+-(used)g(in)f(bzip2.)390 2273 y Fj(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 Fl(Kunihik)m(o)28 b(Sadak)-5 b(ane's)31 b(sorting)e(algorithm,)h
+-(men)m(tioned)g(ab)s(o)m(v)m(e,)i(is)d(a)m(v)-5 b(ailable)30
+-b(from:)390 2996 y Fj(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
+-Fl(The)41 b(Man)m(b)s(er-My)m(ers)g(su\016x)g(arra)m(y)g(construction)g
+-(algorithm)f(is)g(describ)s(ed)f(in)h(a)i(pap)s(er)e(a)m(v)-5
+-b(ailable)150 3262 y(from:)390 3413 y Fj(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 Fl(Finally)d(,)33 b(the)h(follo)m(wing)e(pap)s(er)h(do)s(cumen)m
+-(ts)g(some)h(recen)m(t)h(in)m(v)m(estigations)e(I)h(made)f(in)m(to)h
+-(the)g(p)s(erfor-)150 3680 y(mance)d(of)f(sorting)g(algorithms:)390
+-3831 y Fj(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 38
+--1 37 bop 3725 -116 a Fl(i)150 299 y Fh(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 Fb(.)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 Fd(bzip2)31
+-b Fb(.)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 Fl(NAME)20 b Fa(.)c(.)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(.)50 b Fl(3)1047
+-1157 y(SYNOPSIS)21 b Fa(.)13 b(.)i(.)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(.)g(.)50 b Fl(3)1047 1267 y(DESCRIPTION)10
+-b Fa(.)j(.)i(.)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(.)40
+-b Fl(3)1047 1377 y(OPTIONS)16 b Fa(.)d(.)i(.)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(.)45 b Fl(4)1047
+-1486 y(MEMOR)-8 b(Y)31 b(MANA)m(GEMENT)14 b Fa(.)j(.)e(.)g(.)g(.)g(.)g
+-(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)44
+-b Fl(6)1047 1596 y(RECO)m(VERING)30 b(D)m(A)-8 b(T)g(A)32
+-b(FR)m(OM)f(D)m(AMA)m(GED)i(FILES)1256 1705 y Fa(.)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(PERF)m(ORMANCE)30
+-b(NOTES)9 b Fa(.)14 b(.)h(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g
+-(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)38 b Fl(7)1047 1924
+-y(CA)-10 b(VEA)i(TS)10 b Fa(.)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(.)
+-h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)40 b Fl(8)1047 2034
+-y(A)m(UTHOR)23 b Fa(.)15 b(.)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(.)53 b Fl(8)150 2276 y Fk(3)135
+-b(Programming)46 b(with)f Fd(libbzip2)29 b Fb(.)16 b(.)j(.)h(.)f(.)h(.)
+-f(.)h(.)f(.)g(.)h(.)f(.)h(.)f(.)h(.)f(.)74 b Fk(9)449
+-2413 y Fl(3.1)92 b(T)-8 b(op-lev)m(el)30 b(structure)24
+-b Fa(.)15 b(.)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(.)54 b Fl(9)748 2523 y(3.1.1)93 b(Lo)m(w-lev)m(el)30
+-b(summary)23 b Fa(.)15 b(.)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(.)53
+-b Fl(9)748 2633 y(3.1.2)93 b(High-lev)m(el)29 b(summary)12
+-b Fa(.)i(.)h(.)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(.)41 b
+-Fl(9)748 2742 y(3.1.3)93 b(Utilit)m(y)29 b(functions)g(summary)12
+-b Fa(.)h(.)j(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)
+-g(.)g(.)g(.)g(.)g(.)41 b Fl(10)449 2852 y(3.2)92 b(Error)29
+-b(handling)18 b Fa(.)13 b(.)i(.)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(.)g(.)g(.)48 b Fl(10)449
+-2961 y(3.3)92 b(Lo)m(w-lev)m(el)31 b(in)m(terface)d Fa(.)15
+-b(.)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(.)57 b Fl(12)748 3071 y(3.3.1)93 b Fj(BZ2_bzCompressInit)21
+-b Fa(.)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(12)748
+-3181 y(3.3.2)93 b Fj(BZ2_bzCompress)9 b Fa(.)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(14)748 3290 y(3.3.3)93
+-b Fj(BZ2_bzCompressEnd)23 b Fa(.)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(17)748 3400 y(3.3.4)93 b Fj(BZ2_bzDecompressInit)16
+-b Fa(.)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(17)748 3509
+-y(3.3.5)93 b Fj(BZ2_bzDecompress)21 b Fa(.)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(17)748 3619 y(3.3.6)93 b Fj(BZ2_bzDecompressEnd)18
+-b Fa(.)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(19)449
+-3729 y(3.4)92 b(High-lev)m(el)30 b(in)m(terface)16 b
+-Fa(.)f(.)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(.)45 b Fl(19)748 3838 y(3.4.1)93 b Fj(BZ2_bzReadOpen)9
+-b Fa(.)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(19)748 3948 y(3.4.2)93 b Fj(BZ2_bzRead)18 b Fa(.)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(20)748 4057 y(3.4.3)93 b Fj(BZ2_bzReadGetUnused)18
+-b Fa(.)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(22)748
+-4167 y(3.4.4)93 b Fj(BZ2_bzReadClose)23 b Fa(.)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(22)748 4276 y(3.4.5)93 b
+-Fj(BZ2_bzWriteOpen)23 b Fa(.)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(22)748 4386 y(3.4.6)93 b Fj(BZ2_bzWrite)16
+-b Fa(.)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(23)748 4496 y(3.4.7)93 b Fj(BZ2_bzWriteClose)21
+-b Fa(.)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(23)748
+-4605 y(3.4.8)93 b(Handling)28 b(em)m(b)s(edded)h(compressed)h(data)h
+-(streams)17 b Fa(.)f(.)f(.)g(.)46 b Fl(24)748 4715 y(3.4.9)93
+-b(Standard)29 b(\014le-reading/writing)e(co)s(de)22 b
+-Fa(.)16 b(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)52
+-b Fl(25)449 4824 y(3.5)92 b(Utilit)m(y)29 b(functions)f
+-Fa(.)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(.)g(.)g(.)g(.)g(.)
+-g(.)g(.)g(.)g(.)g(.)59 b Fl(26)748 4934 y(3.5.1)93 b
+-Fj(BZ2_bzBuffToBuffCompres)o(s)22 b Fa(.)15 b(.)g(.)g(.)g(.)g(.)g(.)g
+-(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)f(.)g(.)g(.)g(.)57 b Fl(26)748
+-5044 y(3.5.2)93 b Fj(BZ2_bzBuffToBuffDecompr)o(ess)17
+-b Fa(.)e(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)
+-53 b Fl(27)449 5153 y(3.6)92 b Fj(zlib)29 b Fl(compatibilit)m(y)g
+-(functions)23 b Fa(.)13 b(.)i(.)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(.)52
+-b Fl(28)449 5263 y(3.7)92 b(Using)30 b(the)g(library)e(in)h(a)i
+-Fj(stdio)p Fl(-free)e(en)m(vironmen)m(t)23 b Fa(.)15
+-b(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h(.)52 b Fl(29)p
+-eop
+-%%Page: -2 39
+--2 38 bop 3699 -116 a Fl(ii)748 83 y(3.7.1)93 b(Getting)31
+-b(rid)d(of)j Fj(stdio)20 b Fa(.)13 b(.)i(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)
+-g(.)g(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)49
+-b Fl(29)748 193 y(3.7.2)93 b(Critical)28 b(error)i(handling)22
+-b Fa(.)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(.)54 b Fl(29)449 302
+-y(3.8)92 b(Making)30 b(a)h(Windo)m(ws)e(DLL)15 b Fa(.)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(.)h(.)f(.)g(.)g(.)44 b Fl(30)150 545
+-y Fk(4)135 b(Miscellanea)11 b Fb(.)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(31)449 682 y Fl(4.1)92 b(Limitations)29
+-b(of)h(the)h(compressed)f(\014le)f(format)9 b Fa(.)15
+-b(.)h(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)38
+-b Fl(31)449 791 y(4.2)92 b(P)m(ortabilit)m(y)30 b(issues)14
+-b Fa(.)f(.)j(.)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(.)g(.)43 b Fl(32)449 901 y(4.3)92 b(Rep)s(orting)29
+-b(bugs)f Fa(.)15 b(.)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(.)58 b Fl(32)449 1010 y(4.4)92
+-b(Did)29 b(y)m(ou)i(get)h(the)e(righ)m(t)g(pac)m(k)-5
+-b(age?)22 b Fa(.)17 b(.)e(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)h
+-(.)f(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)g(.)51
+-b Fl(34)449 1120 y(4.5)92 b(T)-8 b(esting)16 b Fa(.)f(.)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(.)g(.)g(.)g(.)g(.)g(.)h(.)45 b Fl(34)449 1230 y(4.6)92
+-b(F)-8 b(urther)30 b(reading)22 b Fa(.)14 b(.)h(.)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(.)51
+-b Fl(35)p eop
+-%%Trailer
+-end
+-userdict /end-hook known{end-hook}if
+-%%EOF
+diff -Nru bzip2-1.0.1/manual.texi bzip2-1.0.1.new/manual.texi
+--- bzip2-1.0.1/manual.texi Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual.texi Thu Jan 1 01:00:00 1970
+@@ -1,2215 +0,0 @@
+-\input texinfo @c -*- Texinfo -*-
+-@setfilename bzip2.info
+-
+-@ignore
+-This file documents bzip2 version 1.0, and associated library
+-libbzip2, written by Julian Seward (jseward@acm.org).
+-
+-Copyright (C) 1996-2000 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) 1996-2000 Julian Seward
+-@subtitle version 1.0 of 21 March 2000
+-@author Julian Seward
+-
+-@end titlepage
+-
+-@parindent 0mm
+-@parskip 2mm
+-
+-@end iftex
+-@node Top, Overview, (dir), (dir)
+-
+-This program, @code{bzip2},
+-and associated library @code{libbzip2}, are
+-Copyright (C) 1996-2000 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{http://sourceware.cygnus.com/bzip2}
+-
+-@code{http://www.cacheprof.org}
+-
+-@code{http://www.muraroa.demon.co.uk}
+-
+-@code{bzip2}/@code{libbzip2} version 1.0 of 21 March 2000.
+-
+-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.
+-
+-
+-
+-
+-
+-
+-
+-@node Overview, Implementation, Top, Top
+-@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
+-@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.
+-@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 to -9
+-Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
+-effect when decompressing. See MEMORY MANAGEMENT below.
+-@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{rec0001file.bz2},
+- @code{rec0002file.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} -- lists 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 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 and
+-0.9.5, 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} uses 32-bit integers to represent bit positions in
+-compressed files, so it cannot handle compressed files more than 512
+-megabytes long. This could easily be fixed.
+-
+-
+-@unnumberedsubsubsec AUTHOR
+-Julian Seward, @code{jseward@@acm.org}.
+-
+-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. 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 of 21-Mar-2000.
+- 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, 21 March 2000.
+-@end example
+-where @code{N} is some error code number. @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-1.1.2} and
+-@code{gzip-1.2.4}. Look for them at
+-
+-@code{http://www.cdrom.com/pub/infozip/zlib} 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 bzip2-1.0.1/manual_1.html bzip2-1.0.1.new/manual_1.html
+--- bzip2-1.0.1/manual_1.html Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual_1.html Thu Jan 1 01:00:00 1970
+@@ -1,47 +0,0 @@
+-<HTML>
+-<HEAD>
+-<!-- This HTML file has been created by texi2html 1.54
+- from manual.texi on 23 March 2000 -->
+-
+-<TITLE>bzip2 and libbzip2 - Introduction</TITLE>
+-<link href="manual_2.html" rel=Next>
+-<link href="manual_toc.html" rel=ToC>
+-
+-</HEAD>
+-<BODY>
+-<p>Go to the first, previous, <A HREF="manual_2.html">next</A>, <A HREF="manual_4.html">last</A> section, <A HREF="manual_toc.html">table of contents</A>.
+-<P><HR><P>
+-
+-
+-<H1><A NAME="SEC1" HREF="manual_toc.html#TOC1">Introduction</A></H1>
+-
+-<P>
+-<CODE>bzip2</CODE> 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.
+-
+-</P>
+-<P>
+-<CODE>bzip2</CODE> is built on top of <CODE>libbzip2</CODE>, a flexible library
+-for handling compressed data in the <CODE>bzip2</CODE> 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.
+-
+-</P>
+-<P>
+-Chapter 2 describes how to use <CODE>bzip2</CODE>; 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.
+-
+-</P>
+-
+-<P><HR><P>
+-<p>Go to the first, previous, <A HREF="manual_2.html">next</A>, <A HREF="manual_4.html">last</A> section, <A HREF="manual_toc.html">table of contents</A>.
+-</BODY>
+-</HTML>
+diff -Nru bzip2-1.0.1/manual_2.html bzip2-1.0.1.new/manual_2.html
+--- bzip2-1.0.1/manual_2.html Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual_2.html Thu Jan 1 01:00:00 1970
+@@ -1,484 +0,0 @@
+-<HTML>
+-<HEAD>
+-<!-- This HTML file has been created by texi2html 1.54
+- from manual.texi on 23 March 2000 -->
+-
+-<TITLE>bzip2 and libbzip2 - How to use bzip2</TITLE>
+-<link href="manual_3.html" rel=Next>
+-<link href="manual_1.html" rel=Previous>
+-<link href="manual_toc.html" rel=ToC>
+-
+-</HEAD>
+-<BODY>
+-<p>Go to the <A HREF="manual_1.html">first</A>, <A HREF="manual_1.html">previous</A>, <A HREF="manual_3.html">next</A>, <A HREF="manual_4.html">last</A> section, <A HREF="manual_toc.html">table of contents</A>.
+-<P><HR><P>
+-
+-
+-<H1><A NAME="SEC2" HREF="manual_toc.html#TOC2">How to use <CODE>bzip2</CODE></A></H1>
+-
+-<P>
+-This chapter contains a copy of the <CODE>bzip2</CODE> man page,
+-and nothing else.
+-
+-</P>
+-
+-<BLOCKQUOTE>
+-
+-
+-
+-<H4><A NAME="SEC3" HREF="manual_toc.html#TOC3">NAME</A></H4>
+-
+-<UL>
+-<LI><CODE>bzip2</CODE>, <CODE>bunzip2</CODE>
+-
+-- a block-sorting file compressor, v1.0
+-<LI><CODE>bzcat</CODE>
+-
+-- decompresses files to stdout
+-<LI><CODE>bzip2recover</CODE>
+-
+-- recovers data from damaged bzip2 files
+-</UL>
+-
+-
+-
+-<H4><A NAME="SEC4" HREF="manual_toc.html#TOC4">SYNOPSIS</A></H4>
+-
+-<UL>
+-<LI><CODE>bzip2</CODE> [ -cdfkqstvzVL123456789 ] [ filenames ... ]
+-
+-<LI><CODE>bunzip2</CODE> [ -fkvsVL ] [ filenames ... ]
+-
+-<LI><CODE>bzcat</CODE> [ -s ] [ filenames ... ]
+-
+-<LI><CODE>bzip2recover</CODE> filename
+-
+-</UL>
+-
+-
+-
+-<H4><A NAME="SEC5" HREF="manual_toc.html#TOC5">DESCRIPTION</A></H4>
+-
+-<P>
+-<CODE>bzip2</CODE> 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.
+-
+-</P>
+-<P>
+-The command-line options are deliberately very similar to those of GNU
+-<CODE>gzip</CODE>, but they are not identical.
+-
+-</P>
+-<P>
+-<CODE>bzip2</CODE> 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</CODE>. 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.
+-
+-</P>
+-<P>
+-<CODE>bzip2</CODE> and <CODE>bunzip2</CODE> will by default not overwrite existing
+-files. If you want this to happen, specify the <CODE>-f</CODE> flag.
+-
+-</P>
+-<P>
+-If no file names are specified, <CODE>bzip2</CODE> compresses from standard
+-input to standard output. In this case, <CODE>bzip2</CODE> will decline to
+-write compressed output to a terminal, as this would be entirely
+-incomprehensible and therefore pointless.
+-
+-</P>
+-<P>
+-<CODE>bunzip2</CODE> (or <CODE>bzip2 -d</CODE>) decompresses all
+-specified files. Files which were not created by <CODE>bzip2</CODE>
+-will be detected and ignored, and a warning issued.
+-<CODE>bzip2</CODE> attempts to guess the filename for the decompressed file
+-from that of the compressed file as follows:
+-
+-<UL>
+-<LI><CODE>filename.bz2 </CODE> becomes <CODE>filename</CODE>
+-
+-<LI><CODE>filename.bz </CODE> becomes <CODE>filename</CODE>
+-
+-<LI><CODE>filename.tbz2</CODE> becomes <CODE>filename.tar</CODE>
+-
+-<LI><CODE>filename.tbz </CODE> becomes <CODE>filename.tar</CODE>
+-
+-<LI><CODE>anyothername </CODE> becomes <CODE>anyothername.out</CODE>
+-
+-</UL>
+-
+-<P>
+-If the file does not end in one of the recognised endings,
+-<CODE>.bz2</CODE>, <CODE>.bz</CODE>,
+-<CODE>.tbz2</CODE> or <CODE>.tbz</CODE>, <CODE>bzip2</CODE> complains that it cannot
+-guess the name of the original file, and uses the original name
+-with <CODE>.out</CODE> appended.
+-
+-</P>
+-<P>
+-As with compression, supplying no
+-filenames causes decompression from standard input to standard output.
+-
+-</P>
+-<P>
+-<CODE>bunzip2</CODE> 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</CODE>) of concatenated compressed files is also supported.
+-
+-</P>
+-<P>
+-You can also compress or decompress files to the standard output by
+-giving the <CODE>-c</CODE> 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</CODE> version 0.9.0 or
+-later. Earlier versions of <CODE>bzip2</CODE> will stop after decompressing
+-the first file in the stream.
+-
+-</P>
+-<P>
+-<CODE>bzcat</CODE> (or <CODE>bzip2 -dc</CODE>) decompresses all specified files to
+-the standard output.
+-
+-</P>
+-<P>
+-<CODE>bzip2</CODE> will read arguments from the environment variables
+-<CODE>BZIP2</CODE> and <CODE>BZIP</CODE>, in that order, and will process them
+-before any arguments read from the command line. This gives a
+-convenient way to supply default arguments.
+-
+-</P>
+-<P>
+-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%.
+-
+-</P>
+-<P>
+-As a self-check for your protection, <CODE>bzip2</CODE> 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</CODE> (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</CODE> to try to recover data from
+-damaged files.
+-
+-</P>
+-<P>
+-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</CODE> to panic.
+-
+-</P>
+-
+-
+-
+-<H4><A NAME="SEC6" HREF="manual_toc.html#TOC6">OPTIONS</A></H4>
+-<DL COMPACT>
+-
+-<DT><CODE>-c --stdout</CODE>
+-<DD>
+-Compress or decompress to standard output.
+-<DT><CODE>-d --decompress</CODE>
+-<DD>
+-Force decompression. <CODE>bzip2</CODE>, <CODE>bunzip2</CODE> and <CODE>bzcat</CODE> 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.
+-<DT><CODE>-z --compress</CODE>
+-<DD>
+-The complement to <CODE>-d</CODE>: forces compression, regardless of the
+-invokation name.
+-<DT><CODE>-t --test</CODE>
+-<DD>
+-Check integrity of the specified file(s), but don't decompress them.
+-This really performs a trial decompression and throws away the result.
+-<DT><CODE>-f --force</CODE>
+-<DD>
+-Force overwrite of output files. Normally, <CODE>bzip2</CODE> will not overwrite
+-existing output files. Also forces <CODE>bzip2</CODE> to break hard links
+-to files, which it otherwise wouldn't do.
+-<DT><CODE>-k --keep</CODE>
+-<DD>
+-Keep (don't delete) input files during compression
+-or decompression.
+-<DT><CODE>-s --small</CODE>
+-<DD>
+-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</CODE> 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.
+-<DT><CODE>-q --quiet</CODE>
+-<DD>
+-Suppress non-essential warning messages. Messages pertaining to
+-I/O errors and other critical events will not be suppressed.
+-<DT><CODE>-v --verbose</CODE>
+-<DD>
+-Verbose mode -- show the compression ratio for each file processed.
+-Further <CODE>-v</CODE>'s increase the verbosity level, spewing out lots of
+-information which is primarily of interest for diagnostic purposes.
+-<DT><CODE>-L --license -V --version</CODE>
+-<DD>
+-Display the software version, license terms and conditions.
+-<DT><CODE>-1 to -9</CODE>
+-<DD>
+-Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
+-effect when decompressing. See MEMORY MANAGEMENT below.
+-<DT><CODE>--</CODE>
+-<DD>
+-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</CODE>.
+-<DT><CODE>--repetitive-fast</CODE>
+-<DD>
+-<DT><CODE>--repetitive-best</CODE>
+-<DD>
+-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.
+-</DL>
+-
+-
+-
+-<H4><A NAME="SEC7" HREF="manual_toc.html#TOC7">MEMORY MANAGEMENT</A></H4>
+-
+-<P>
+-<CODE>bzip2</CODE> 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</CODE> through <CODE>-9</CODE>
+-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</CODE> 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</CODE> to <CODE>-9</CODE> are irrelevant to and so ignored
+-during decompression.
+-
+-</P>
+-<P>
+-Compression and decompression requirements, in bytes, can be estimated
+-as:
+-
+-<PRE>
+- Compression: 400k + ( 8 x block size )
+-
+- Decompression: 100k + ( 4 x block size ), or
+- 100k + ( 2.5 x block size )
+-</PRE>
+-
+-<P>
+-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</CODE> 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.
+-
+-</P>
+-<P>
+-For files compressed with the default 900k block size, <CODE>bunzip2</CODE>
+-will require about 3700 kbytes to decompress. To support decompression
+-of any file on a 4 megabyte machine, <CODE>bunzip2</CODE> 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</CODE>.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-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</CODE> 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.
+-
+-</P>
+-<P>
+-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.
+-
+-<PRE>
+- 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
+-</PRE>
+-
+-
+-
+-<H4><A NAME="SEC8" HREF="manual_toc.html#TOC8">RECOVERING DATA FROM DAMAGED FILES</A></H4>
+-
+-<P>
+-<CODE>bzip2</CODE> compresses files in blocks, usually 900kbytes long. Each
+-block is handled independently. If a media or transmission error causes
+-a multi-block <CODE>.bz2</CODE> file to become damaged, it may be possible to
+-recover data from the undamaged blocks in the file.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-<CODE>bzip2recover</CODE> is a simple program whose purpose is to search for
+-blocks in <CODE>.bz2</CODE> files, and write each block out into its own
+-<CODE>.bz2</CODE> file. You can then use <CODE>bzip2 -t</CODE> to test the
+-integrity of the resulting files, and decompress those which are
+-undamaged.
+-
+-</P>
+-<P>
+-<CODE>bzip2recover</CODE>
+-takes a single argument, the name of the damaged file,
+-and writes a number of files <CODE>rec0001file.bz2</CODE>,
+- <CODE>rec0002file.bz2</CODE>, 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</CODE> -- lists the files in
+- the correct order.
+-
+-</P>
+-<P>
+-<CODE>bzip2recover</CODE> should be of most use dealing with large <CODE>.bz2</CODE>
+- 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.
+-
+-</P>
+-
+-
+-
+-<H4><A NAME="SEC9" HREF="manual_toc.html#TOC9">PERFORMANCE NOTES</A></H4>
+-
+-<P>
+-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</CODE> option to monitor progress in great detail, if you want.
+-
+-</P>
+-<P>
+-Decompression speed is unaffected by these phenomena.
+-
+-</P>
+-<P>
+-<CODE>bzip2</CODE> 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</CODE> will perform best on machines with very large
+-caches.
+-
+-</P>
+-
+-
+-
+-<H4><A NAME="SEC10" HREF="manual_toc.html#TOC10">CAVEATS</A></H4>
+-
+-<P>
+-I/O error messages are not as helpful as they could be. <CODE>bzip2</CODE>
+-tries hard to detect I/O errors and exit cleanly, but the details of
+-what the problem is sometimes seem rather misleading.
+-
+-</P>
+-<P>
+-This manual page pertains to version 1.0 of <CODE>bzip2</CODE>. Compressed
+-data created by this version is entirely forwards and backwards
+-compatible with the previous public releases, versions 0.1pl2, 0.9.0 and
+-0.9.5, 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.
+-
+-</P>
+-<P>
+-<CODE>bzip2recover</CODE> uses 32-bit integers to represent bit positions in
+-compressed files, so it cannot handle compressed files more than 512
+-megabytes long. This could easily be fixed.
+-
+-</P>
+-
+-
+-
+-<H4><A NAME="SEC11" HREF="manual_toc.html#TOC11">AUTHOR</A></H4>
+-<P>
+-Julian Seward, <CODE>jseward@acm.org</CODE>.
+-
+-</P>
+-<P>
+-The ideas embodied in <CODE>bzip2</CODE> 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</CODE>,
+-and many refinements), and Alistair Moffat, Radford Neal and Ian Witten
+-(for the arithmetic coder in the original <CODE>bzip</CODE>). 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. Many people sent patches, helped
+-with portability problems, lent machines, gave advice and were generally
+-helpful.
+-
+-</P>
+-</BLOCKQUOTE>
+-
+-<P><HR><P>
+-<p>Go to the <A HREF="manual_1.html">first</A>, <A HREF="manual_1.html">previous</A>, <A HREF="manual_3.html">next</A>, <A HREF="manual_4.html">last</A> section, <A HREF="manual_toc.html">table of contents</A>.
+-</BODY>
+-</HTML>
+diff -Nru bzip2-1.0.1/manual_3.html bzip2-1.0.1.new/manual_3.html
+--- bzip2-1.0.1/manual_3.html Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual_3.html Thu Jan 1 01:00:00 1970
+@@ -1,1773 +0,0 @@
+-<HTML>
+-<HEAD>
+-<!-- This HTML file has been created by texi2html 1.54
+- from manual.texi on 23 March 2000 -->
+-
+-<TITLE>bzip2 and libbzip2 - Programming with libbzip2</TITLE>
+-<link href="manual_4.html" rel=Next>
+-<link href="manual_2.html" rel=Previous>
+-<link href="manual_toc.html" rel=ToC>
+-
+-</HEAD>
+-<BODY>
+-<p>Go to the <A HREF="manual_1.html">first</A>, <A HREF="manual_2.html">previous</A>, <A HREF="manual_4.html">next</A>, <A HREF="manual_4.html">last</A> section, <A HREF="manual_toc.html">table of contents</A>.
+-<P><HR><P>
+-
+-
+-<H1><A NAME="SEC12" HREF="manual_toc.html#TOC12">Programming with <CODE>libbzip2</CODE></A></H1>
+-
+-<P>
+-This chapter describes the programming interface to <CODE>libbzip2</CODE>.
+-
+-</P>
+-<P>
+-For general background information, particularly about memory
+-use and performance aspects, you'd be well advised to read Chapter 2
+-as well.
+-
+-</P>
+-
+-
+-<H2><A NAME="SEC13" HREF="manual_toc.html#TOC13">Top-level structure</A></H2>
+-
+-<P>
+-<CODE>libbzip2</CODE> is a flexible library for compressing and decompressing
+-data in the <CODE>bzip2</CODE> 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.
+-
+-</P>
+-<P>
+-The structure of <CODE>libbzip2</CODE>'s interfaces is similar to
+-that of Jean-loup Gailly's and Mark Adler's excellent <CODE>zlib</CODE>
+-library.
+-
+-</P>
+-<P>
+-All externally visible symbols have names beginning <CODE>BZ2_</CODE>.
+-This is new in version 1.0. The intention is to minimise pollution
+-of the namespaces of library clients.
+-
+-</P>
+-
+-
+-<H3><A NAME="SEC14" HREF="manual_toc.html#TOC14">Low-level summary</A></H3>
+-
+-<P>
+-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</CODE>, which may be helpful for embedded applications.
+-
+-</P>
+-<P>
+-The low-level part of the library has no global variables and
+-is therefore thread-safe.
+-
+-</P>
+-<P>
+-Six routines make up the low level interface:
+-<CODE>BZ2_bzCompressInit</CODE>, <CODE>BZ2_bzCompress</CODE>, and <BR> <CODE>BZ2_bzCompressEnd</CODE>
+-for compression,
+-and a corresponding trio <CODE>BZ2_bzDecompressInit</CODE>, <BR> <CODE>BZ2_bzDecompress</CODE>
+-and <CODE>BZ2_bzDecompressEnd</CODE> for decompression.
+-The <CODE>*Init</CODE> functions allocate
+-memory for compression/decompression and do other
+-initialisations, whilst the <CODE>*End</CODE> functions close down operations
+-and release memory.
+-
+-</P>
+-<P>
+-The real work is done by <CODE>BZ2_bzCompress</CODE> and <CODE>BZ2_bzDecompress</CODE>.
+-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.
+-
+-</P>
+-
+-
+-
+-<H3><A NAME="SEC15" HREF="manual_toc.html#TOC15">High-level summary</A></H3>
+-
+-<P>
+-This interface provides some handy wrappers around the low-level
+-interface to facilitate reading and writing <CODE>bzip2</CODE> format
+-files (<CODE>.bz2</CODE> files). The routines provide hooks to facilitate
+-reading files in which the <CODE>bzip2</CODE> data stream is embedded
+-within some larger-scale file structure, or where there are
+-multiple <CODE>bzip2</CODE> data streams concatenated end-to-end.
+-
+-</P>
+-<P>
+-For reading files, <CODE>BZ2_bzReadOpen</CODE>, <CODE>BZ2_bzRead</CODE>,
+-<CODE>BZ2_bzReadClose</CODE> and <BR> <CODE>BZ2_bzReadGetUnused</CODE> are supplied. For
+-writing files, <CODE>BZ2_bzWriteOpen</CODE>, <CODE>BZ2_bzWrite</CODE> and
+-<CODE>BZ2_bzWriteFinish</CODE> are available.
+-
+-</P>
+-<P>
+-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</CODE> to determine the cause of
+-the error. In that case, you'd need a C library which correctly
+-supports <CODE>errno</CODE> in a multithreaded environment.
+-
+-</P>
+-<P>
+-To make the library a little simpler and more portable,
+-<CODE>BZ2_bzReadOpen</CODE> and <CODE>BZ2_bzWriteOpen</CODE> require you to pass them file
+-handles (<CODE>FILE*</CODE>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.
+-
+-</P>
+-
+-
+-
+-<H3><A NAME="SEC16" HREF="manual_toc.html#TOC16">Utility functions summary</A></H3>
+-<P>
+-For very simple needs, <CODE>BZ2_bzBuffToBuffCompress</CODE> and
+-<CODE>BZ2_bzBuffToBuffDecompress</CODE> 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.
+-
+-</P>
+-<P>
+-Yoshioka Tsuneo (<CODE>QWF00133@niftyserve.or.jp</CODE> /
+-<CODE>tsuneo-y@is.aist-nara.ac.jp</CODE>) has contributed some functions to
+-give better <CODE>zlib</CODE> compatibility. These functions are
+-<CODE>BZ2_bzopen</CODE>, <CODE>BZ2_bzread</CODE>, <CODE>BZ2_bzwrite</CODE>, <CODE>BZ2_bzflush</CODE>,
+-<CODE>BZ2_bzclose</CODE>,
+-<CODE>BZ2_bzerror</CODE> and <CODE>BZ2_bzlibVersion</CODE>. 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.
+-
+-</P>
+-<P>
+-Yoshioka also contributed modifications to allow the library to be
+-built as a Windows DLL.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC17" HREF="manual_toc.html#TOC17">Error handling</A></H2>
+-
+-<P>
+-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.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-The file <CODE>bzlib.h</CODE> contains all definitions needed to use
+-the library. In particular, you should definitely not include
+-<CODE>bzlib_private.h</CODE>.
+-
+-</P>
+-<P>
+-In <CODE>bzlib.h</CODE>, 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.
+-<DL COMPACT>
+-
+-<DT><CODE>BZ_OK</CODE>
+-<DD>
+-The requested action was completed successfully.
+-<DT><CODE>BZ_RUN_OK</CODE>
+-<DD>
+-<DT><CODE>BZ_FLUSH_OK</CODE>
+-<DD>
+-<DT><CODE>BZ_FINISH_OK</CODE>
+-<DD>
+-In <CODE>BZ2_bzCompress</CODE>, the requested flush/finish/nothing-special action
+-was completed successfully.
+-<DT><CODE>BZ_STREAM_END</CODE>
+-<DD>
+-Compression of data was completed, or the logical stream end was
+-detected during decompression.
+-</DL>
+-
+-<P>
+-The following return values indicate an error of some kind.
+-<DL COMPACT>
+-
+-<DT><CODE>BZ_CONFIG_ERROR</CODE>
+-<DD>
+-Indicates that the library has been improperly compiled on your
+-platform -- a major configuration error. Specifically, it means
+-that <CODE>sizeof(char)</CODE>, <CODE>sizeof(short)</CODE> and <CODE>sizeof(int)</CODE>
+-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)</CODE>
+-and <CODE>sizeof(void*)</CODE> are 8. Under LP64, <CODE>sizeof(int)</CODE> is
+-still 4, so <CODE>libbzip2</CODE>, which doesn't use the <CODE>long</CODE> type,
+-is OK.
+-<DT><CODE>BZ_SEQUENCE_ERROR</CODE>
+-<DD>
+-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</CODE> checks as much as it can to ensure this is
+-happening, and returns <CODE>BZ_SEQUENCE_ERROR</CODE> 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.
+-<DT><CODE>BZ_PARAM_ERROR</CODE>
+-<DD>
+-Returned when a parameter to a function call is out of range
+-or otherwise manifestly incorrect. As with <CODE>BZ_SEQUENCE_ERROR</CODE>,
+-this denotes a bug in the client code. The distinction between
+-<CODE>BZ_PARAM_ERROR</CODE> and <CODE>BZ_SEQUENCE_ERROR</CODE> is a bit hazy, but still worth
+-making.
+-<DT><CODE>BZ_MEM_ERROR</CODE>
+-<DD>
+-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</CODE> and
+-<CODE>BZ2_bzRead</CODE> may return <CODE>BZ_MEM_ERROR</CODE> even though some of
+-the compressed data has been read. The same is not true for
+-compression; once <CODE>BZ2_bzCompressInit</CODE> or <CODE>BZ2_bzWriteOpen</CODE> have
+-successfully completed, <CODE>BZ_MEM_ERROR</CODE> cannot occur.
+-<DT><CODE>BZ_DATA_ERROR</CODE>
+-<DD>
+-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.
+-<DT><CODE>BZ_DATA_ERROR_MAGIC</CODE>
+-<DD>
+-As a special case of <CODE>BZ_DATA_ERROR</CODE>, it is sometimes useful to
+-know when the compressed stream does not start with the correct
+-magic bytes (<CODE>'B' 'Z' 'h'</CODE>).
+-<DT><CODE>BZ_IO_ERROR</CODE>
+-<DD>
+-Returned by <CODE>BZ2_bzRead</CODE> and <CODE>BZ2_bzWrite</CODE> when there is an error
+-reading or writing in the compressed file, and by <CODE>BZ2_bzReadOpen</CODE>
+-and <CODE>BZ2_bzWriteOpen</CODE> for attempts to use a file for which the
+-error indicator (viz, <CODE>ferror(f)</CODE>) is set.
+-On receipt of <CODE>BZ_IO_ERROR</CODE>, the caller should consult
+-<CODE>errno</CODE> and/or <CODE>perror</CODE> to acquire operating-system
+-specific information about the problem.
+-<DT><CODE>BZ_UNEXPECTED_EOF</CODE>
+-<DD>
+-Returned by <CODE>BZ2_bzRead</CODE> when the compressed file finishes
+-before the logical end of stream is detected.
+-<DT><CODE>BZ_OUTBUFF_FULL</CODE>
+-<DD>
+-Returned by <CODE>BZ2_bzBuffToBuffCompress</CODE> and
+-<CODE>BZ2_bzBuffToBuffDecompress</CODE> to indicate that the output data
+-will not fit into the output buffer provided.
+-</DL>
+-
+-
+-
+-<H2><A NAME="SEC18" HREF="manual_toc.html#TOC18">Low-level interface</A></H2>
+-
+-
+-
+-<H3><A NAME="SEC19" HREF="manual_toc.html#TOC19"><CODE>BZ2_bzCompressInit</CODE></A></H3>
+-
+-<PRE>
+-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 );
+-
+-</PRE>
+-
+-<P>
+-Prepares for compression. The <CODE>bz_stream</CODE> structure
+-holds all data pertaining to the compression activity.
+-A <CODE>bz_stream</CODE> structure should be allocated and initialised
+-prior to the call.
+-The fields of <CODE>bz_stream</CODE>
+-comprise the entirety of the user-visible data. <CODE>state</CODE>
+-is a pointer to the private data structures required for compression.
+-
+-</P>
+-<P>
+-Custom memory allocators are supported, via fields <CODE>bzalloc</CODE>,
+-<CODE>bzfree</CODE>,
+-and <CODE>opaque</CODE>. The value
+-<CODE>opaque</CODE> is passed to as the first argument to
+-all calls to <CODE>bzalloc</CODE> and <CODE>bzfree</CODE>, but is
+-otherwise ignored by the library.
+-The call <CODE>bzalloc ( opaque, n, m )</CODE> is expected to return a
+-pointer <CODE>p</CODE> to
+-<CODE>n * m</CODE> bytes of memory, and <CODE>bzfree ( opaque, p )</CODE>
+-should free
+-that memory.
+-
+-</P>
+-<P>
+-If you don't want to use a custom memory allocator, set <CODE>bzalloc</CODE>,
+-<CODE>bzfree</CODE> and
+-<CODE>opaque</CODE> to <CODE>NULL</CODE>,
+-and the library will then use the standard <CODE>malloc</CODE>/<CODE>free</CODE>
+-routines.
+-
+-</P>
+-<P>
+-Before calling <CODE>BZ2_bzCompressInit</CODE>, fields <CODE>bzalloc</CODE>,
+-<CODE>bzfree</CODE> and <CODE>opaque</CODE> should
+-be filled appropriately, as just described. Upon return, the internal
+-state will have been allocated and initialised, and <CODE>total_in_lo32</CODE>,
+-<CODE>total_in_hi32</CODE>, <CODE>total_out_lo32</CODE> and
+-<CODE>total_out_hi32</CODE> 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</CODE> 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</CODE>.
+-
+-</P>
+-<P>
+-Parameter <CODE>blockSize100k</CODE> 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.
+-
+-</P>
+-<P>
+-Parameter <CODE>verbosity</CODE> 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</CODE>, no such output will appear for any verbosity
+-setting.
+-
+-</P>
+-<P>
+-Parameter <CODE>workFactor</CODE> 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.
+-
+-</P>
+-<P>
+-Lower values of <CODE>workFactor</CODE> 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.
+-
+-</P>
+-<P>
+-Allowable values range from 0 to 250 inclusive. 0 is a special case,
+-equivalent to using the default value of 30.
+-
+-</P>
+-<P>
+-Note that the compressed output generated is the same regardless of
+-whether or not the fallback algorithm is used.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_CONFIG_ERROR</CODE>
+- if the library has been mis-compiled
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>strm</CODE> is <CODE>NULL</CODE>
+- or <CODE>blockSize</CODE> < 1 or <CODE>blockSize</CODE> > 9
+- or <CODE>verbosity</CODE> < 0 or <CODE>verbosity</CODE> > 4
+- or <CODE>workFactor</CODE> < 0 or <CODE>workFactor</CODE> > 250
+- <CODE>BZ_MEM_ERROR</CODE>
+- if not enough memory is available
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- <CODE>BZ2_bzCompress</CODE>
+- if <CODE>BZ_OK</CODE> is returned
+- no specific action needed in case of error
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC20" HREF="manual_toc.html#TOC20"><CODE>BZ2_bzCompress</CODE></A></H3>
+-
+-<PRE>
+- int BZ2_bzCompress ( bz_stream *strm, int action );
+-</PRE>
+-
+-<P>
+-Provides more input and/or output buffer space for the library. The
+-caller maintains input and output buffers, and calls <CODE>BZ2_bzCompress</CODE> to
+-transfer data between them.
+-
+-</P>
+-<P>
+-Before each call to <CODE>BZ2_bzCompress</CODE>, <CODE>next_in</CODE> should point at
+-the data to be compressed, and <CODE>avail_in</CODE> should indicate how many
+-bytes the library may read. <CODE>BZ2_bzCompress</CODE> updates <CODE>next_in</CODE>,
+-<CODE>avail_in</CODE> and <CODE>total_in</CODE> to reflect the number of bytes it
+-has read.
+-
+-</P>
+-<P>
+-Similarly, <CODE>next_out</CODE> should point to a buffer in which the
+-compressed data is to be placed, with <CODE>avail_out</CODE> indicating how
+-much output space is available. <CODE>BZ2_bzCompress</CODE> updates
+-<CODE>next_out</CODE>, <CODE>avail_out</CODE> and <CODE>total_out</CODE> to reflect the
+-number of bytes output.
+-
+-</P>
+-<P>
+-You may provide and remove as little or as much data as you like on each
+-call of <CODE>BZ2_bzCompress</CODE>. 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.
+-
+-</P>
+-<P>
+-A second purpose of <CODE>BZ2_bzCompress</CODE> is to request a change of mode of the
+-compressed stream.
+-
+-</P>
+-<P>
+-Conceptually, a compressed stream can be in one of four states: IDLE,
+-RUNNING, FLUSHING and FINISHING. Before initialisation
+-(<CODE>BZ2_bzCompressInit</CODE>) and after termination (<CODE>BZ2_bzCompressEnd</CODE>), a
+-stream is regarded as IDLE.
+-
+-</P>
+-<P>
+-Upon initialisation (<CODE>BZ2_bzCompressInit</CODE>), the stream is placed in the
+-RUNNING state. Subsequent calls to <CODE>BZ2_bzCompress</CODE> should pass
+-<CODE>BZ_RUN</CODE> as the requested action; other actions are illegal and
+-will result in <CODE>BZ_SEQUENCE_ERROR</CODE>.
+-
+-</P>
+-<P>
+-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</CODE> will no longer attempt to read data from
+-<CODE>next_in</CODE>, but it will want to write data to <CODE>next_out</CODE>.
+-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</CODE>.
+-
+-</P>
+-<P>
+-Instead, the calling program passes <CODE>BZ_FINISH</CODE> as an action to
+-<CODE>BZ2_bzCompress</CODE>. This changes the stream's state to FINISHING. Any
+-remaining input (ie, <CODE>next_in[0 .. avail_in-1]</CODE>) is compressed and
+-transferred to the output buffer. To do this, <CODE>BZ2_bzCompress</CODE> must be
+-called repeatedly until all the output has been consumed. At that
+-point, <CODE>BZ2_bzCompress</CODE> returns <CODE>BZ_STREAM_END</CODE>, and the stream's
+-state is set back to IDLE. <CODE>BZ2_bzCompressEnd</CODE> should then be
+-called.
+-
+-</P>
+-<P>
+-Just to make sure the calling program does not cheat, the library makes
+-a note of <CODE>avail_in</CODE> at the time of the first call to
+-<CODE>BZ2_bzCompress</CODE> which has <CODE>BZ_FINISH</CODE> 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</CODE> over
+-subsequent calls to <CODE>BZ2_bzCompress</CODE>, 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</CODE>. This indicates a
+-programming mistake which should be corrected.
+-
+-</P>
+-<P>
+-Instead of asking to finish, the calling program may ask
+-<CODE>BZ2_bzCompress</CODE> 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</CODE> with an action of
+-<CODE>BZ_FLUSH</CODE>, remove output data, and persist with the
+-<CODE>BZ_FLUSH</CODE> action until the value <CODE>BZ_RUN</CODE> is returned. As
+-with finishing, <CODE>BZ2_bzCompress</CODE> detects any attempt to provide more
+-input data once the flush has begun.
+-
+-</P>
+-<P>
+-Once the flush is complete, the stream returns to the normal RUNNING
+-state.
+-
+-</P>
+-<P>
+-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</CODE>.
+-
+-<PRE>
+-IDLE/<CODE>any</CODE>
+- Illegal. IDLE state only exists after <CODE>BZ2_bzCompressEnd</CODE> or
+- before <CODE>BZ2_bzCompressInit</CODE>.
+- Return value = <CODE>BZ_SEQUENCE_ERROR</CODE>
+-
+-RUNNING/<CODE>BZ_RUN</CODE>
+- Compress from <CODE>next_in</CODE> to <CODE>next_out</CODE> as much as possible.
+- Next state = RUNNING
+- Return value = <CODE>BZ_RUN_OK</CODE>
+-
+-RUNNING/<CODE>BZ_FLUSH</CODE>
+- Remember current value of <CODE>next_in</CODE>. Compress from <CODE>next_in</CODE>
+- to <CODE>next_out</CODE> as much as possible, but do not accept any more input.
+- Next state = FLUSHING
+- Return value = <CODE>BZ_FLUSH_OK</CODE>
+-
+-RUNNING/<CODE>BZ_FINISH</CODE>
+- Remember current value of <CODE>next_in</CODE>. Compress from <CODE>next_in</CODE>
+- to <CODE>next_out</CODE> as much as possible, but do not accept any more input.
+- Next state = FINISHING
+- Return value = <CODE>BZ_FINISH_OK</CODE>
+-
+-FLUSHING/<CODE>BZ_FLUSH</CODE>
+- Compress from <CODE>next_in</CODE> to <CODE>next_out</CODE> 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</CODE>
+- else
+- Next state = FLUSHING; Return value = <CODE>BZ_FLUSH_OK</CODE>
+-
+-FLUSHING/other
+- Illegal.
+- Return value = <CODE>BZ_SEQUENCE_ERROR</CODE>
+-
+-FINISHING/<CODE>BZ_FINISH</CODE>
+- Compress from <CODE>next_in</CODE> to <CODE>next_out</CODE> 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</CODE>
+- else
+- Next state = FINISHING; Return value = <CODE>BZ_FINISHING</CODE>
+-
+-FINISHING/other
+- Illegal.
+- Return value = <CODE>BZ_SEQUENCE_ERROR</CODE>
+-</PRE>
+-
+-<P>
+-That still looks complicated? Well, fair enough. The usual sequence
+-of calls for compressing a load of data is:
+-
+-<UL>
+-<LI>Get started with <CODE>BZ2_bzCompressInit</CODE>.
+-
+-<LI>Shovel data in and shlurp out its compressed form using zero or more
+-
+-calls of <CODE>BZ2_bzCompress</CODE> with action = <CODE>BZ_RUN</CODE>.
+-<LI>Finish up.
+-
+-Repeatedly call <CODE>BZ2_bzCompress</CODE> with action = <CODE>BZ_FINISH</CODE>,
+-copying out the compressed output, until <CODE>BZ_STREAM_END</CODE> is returned.
+-<LI>Close up and go home. Call <CODE>BZ2_bzCompressEnd</CODE>.
+-
+-</UL>
+-
+-<P>
+-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 )</CODE> and
+-just do the <CODE>BZ2_bzCompress ( ..., BZ_FINISH )</CODE> calls.
+-
+-</P>
+-<P>
+-All required memory is allocated by <CODE>BZ2_bzCompressInit</CODE>. The
+-compression library can accept any data at all (obviously). So you
+-shouldn't get any error return values from the <CODE>BZ2_bzCompress</CODE> calls.
+-If you do, they will be <CODE>BZ_SEQUENCE_ERROR</CODE>, and indicate a bug in
+-your programming.
+-
+-</P>
+-<P>
+-Trivial other possible return values:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>strm</CODE> is <CODE>NULL</CODE>, or <CODE>strm->s</CODE> is <CODE>NULL</CODE>
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC21" HREF="manual_toc.html#TOC21"><CODE>BZ2_bzCompressEnd</CODE></A></H3>
+-
+-<PRE>
+-int BZ2_bzCompressEnd ( bz_stream *strm );
+-</PRE>
+-
+-<P>
+-Releases all memory associated with a compression stream.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE> if <CODE>strm</CODE> is <CODE>NULL</CODE> or <CODE>strm->s</CODE> is <CODE>NULL</CODE>
+- <CODE>BZ_OK</CODE> otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC22" HREF="manual_toc.html#TOC22"><CODE>BZ2_bzDecompressInit</CODE></A></H3>
+-
+-<PRE>
+-int BZ2_bzDecompressInit ( bz_stream *strm, int verbosity, int small );
+-</PRE>
+-
+-<P>
+-Prepares for decompression. As with <CODE>BZ2_bzCompressInit</CODE>, a
+-<CODE>bz_stream</CODE> record should be allocated and initialised before the
+-call. Fields <CODE>bzalloc</CODE>, <CODE>bzfree</CODE> and <CODE>opaque</CODE> should be
+-set if a custom memory allocator is required, or made <CODE>NULL</CODE> for
+-the normal <CODE>malloc</CODE>/<CODE>free</CODE> routines. Upon return, the internal
+-state will have been initialised, and <CODE>total_in</CODE> and
+-<CODE>total_out</CODE> will be zero.
+-
+-</P>
+-<P>
+-For the meaning of parameter <CODE>verbosity</CODE>, see <CODE>BZ2_bzCompressInit</CODE>.
+-
+-</P>
+-<P>
+-If <CODE>small</CODE> 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.
+-
+-</P>
+-<P>
+-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</CODE> succeeds, a subsequent
+-<CODE>BZ2_bzDecompress</CODE> could fail with <CODE>BZ_MEM_ERROR</CODE>.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_CONFIG_ERROR</CODE>
+- if the library has been mis-compiled
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>(small != 0 && small != 1)</CODE>
+- or <CODE>(verbosity < 0 || verbosity > 4)</CODE>
+- <CODE>BZ_MEM_ERROR</CODE>
+- if insufficient memory is available
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- <CODE>BZ2_bzDecompress</CODE>
+- if <CODE>BZ_OK</CODE> was returned
+- no specific action required in case of error
+-</PRE>
+-
+-<P>
+-
+-
+-</P>
+-
+-
+-<H3><A NAME="SEC23" HREF="manual_toc.html#TOC23"><CODE>BZ2_bzDecompress</CODE></A></H3>
+-
+-<PRE>
+-int BZ2_bzDecompress ( bz_stream *strm );
+-</PRE>
+-
+-<P>
+-Provides more input and/out output buffer space for the library. The
+-caller maintains input and output buffers, and uses <CODE>BZ2_bzDecompress</CODE>
+-to transfer data between them.
+-
+-</P>
+-<P>
+-Before each call to <CODE>BZ2_bzDecompress</CODE>, <CODE>next_in</CODE>
+-should point at the compressed data,
+-and <CODE>avail_in</CODE> should indicate how many bytes the library
+-may read. <CODE>BZ2_bzDecompress</CODE> updates <CODE>next_in</CODE>, <CODE>avail_in</CODE>
+-and <CODE>total_in</CODE>
+-to reflect the number of bytes it has read.
+-
+-</P>
+-<P>
+-Similarly, <CODE>next_out</CODE> should point to a buffer in which the uncompressed
+-output is to be placed, with <CODE>avail_out</CODE> indicating how much output space
+-is available. <CODE>BZ2_bzCompress</CODE> updates <CODE>next_out</CODE>,
+-<CODE>avail_out</CODE> and <CODE>total_out</CODE> to reflect
+-the number of bytes output.
+-
+-</P>
+-<P>
+-You may provide and remove as little or as much data as you like on
+-each call of <CODE>BZ2_bzDecompress</CODE>.
+-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.
+-
+-</P>
+-<P>
+-Use of <CODE>BZ2_bzDecompress</CODE> is simpler than <CODE>BZ2_bzCompress</CODE>.
+-
+-</P>
+-<P>
+-You should provide input and remove output as described above, and
+-repeatedly call <CODE>BZ2_bzDecompress</CODE> until <CODE>BZ_STREAM_END</CODE> is
+-returned. Appearance of <CODE>BZ_STREAM_END</CODE> denotes that
+-<CODE>BZ2_bzDecompress</CODE> has detected the logical end of the compressed
+-stream. <CODE>BZ2_bzDecompress</CODE> will not produce <CODE>BZ_STREAM_END</CODE> until
+-all output data has been placed into the output buffer, so once
+-<CODE>BZ_STREAM_END</CODE> appears, you are guaranteed to have available all
+-the decompressed output, and <CODE>BZ2_bzDecompressEnd</CODE> can safely be
+-called.
+-
+-</P>
+-<P>
+-If case of an error return value, you should call <CODE>BZ2_bzDecompressEnd</CODE>
+-to clean up and release memory.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>strm</CODE> is <CODE>NULL</CODE> or <CODE>strm->s</CODE> is <CODE>NULL</CODE>
+- or <CODE>strm->avail_out < 1</CODE>
+- <CODE>BZ_DATA_ERROR</CODE>
+- if a data integrity error is detected in the compressed stream
+- <CODE>BZ_DATA_ERROR_MAGIC</CODE>
+- if the compressed stream doesn't begin with the right magic bytes
+- <CODE>BZ_MEM_ERROR</CODE>
+- if there wasn't enough memory available
+- <CODE>BZ_STREAM_END</CODE>
+- if the logical end of the data stream was detected and all
+- output in has been consumed, eg <CODE>s->avail_out > 0</CODE>
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- <CODE>BZ2_bzDecompress</CODE>
+- if <CODE>BZ_OK</CODE> was returned
+- <CODE>BZ2_bzDecompressEnd</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC24" HREF="manual_toc.html#TOC24"><CODE>BZ2_bzDecompressEnd</CODE></A></H3>
+-
+-<PRE>
+-int BZ2_bzDecompressEnd ( bz_stream *strm );
+-</PRE>
+-
+-<P>
+-Releases all memory associated with a decompression stream.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>strm</CODE> is <CODE>NULL</CODE> or <CODE>strm->s</CODE> is <CODE>NULL</CODE>
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- None.
+-</PRE>
+-
+-
+-
+-<H2><A NAME="SEC25" HREF="manual_toc.html#TOC25">High-level interface</A></H2>
+-
+-<P>
+-This interface provides functions for reading and writing
+-<CODE>bzip2</CODE> format files. First, some general points.
+-
+-</P>
+-
+-<UL>
+-<LI>All of the functions take an <CODE>int*</CODE> first argument,
+-
+- <CODE>bzerror</CODE>.
+- After each call, <CODE>bzerror</CODE> should be consulted first to determine
+- the outcome of the call. If <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE>,
+- the call completed
+- successfully, and only then should the return value of the function
+- (if any) be consulted. If <CODE>bzerror</CODE> is <CODE>BZ_IO_ERROR</CODE>,
+- there was an error
+- reading/writing the underlying compressed file, and you should
+- then consult <CODE>errno</CODE>/<CODE>perror</CODE> to determine the
+- cause of the difficulty.
+- <CODE>bzerror</CODE> may also be set to various other values; precise details are
+- given on a per-function basis below.
+-<LI>If <CODE>bzerror</CODE> indicates an error
+-
+- (ie, anything except <CODE>BZ_OK</CODE> and <CODE>BZ_STREAM_END</CODE>),
+- you should immediately call <CODE>BZ2_bzReadClose</CODE> (or <CODE>BZ2_bzWriteClose</CODE>,
+- 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> (<CODE>BZ2_bzWriteClose</CODE>) is undefined.
+- The implication is that (1) <CODE>bzerror</CODE> should
+- be checked after each call, and (2) if <CODE>bzerror</CODE> indicates an error,
+- <CODE>BZ2_bzReadClose</CODE> (<CODE>BZ2_bzWriteClose</CODE>) should then be called to clean up.
+-<LI>The <CODE>FILE*</CODE> arguments passed to
+-
+- <CODE>BZ2_bzReadOpen</CODE>/<CODE>BZ2_bzWriteOpen</CODE>
+- 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.
+-<LI>Memory allocation requests are handled by
+-
+- <CODE>malloc</CODE>/<CODE>free</CODE>.
+- At present
+- there is no facility for user-defined memory allocators in the file I/O
+- functions (could easily be added, though).
+-</UL>
+-
+-
+-
+-<H3><A NAME="SEC26" HREF="manual_toc.html#TOC26"><CODE>BZ2_bzReadOpen</CODE></A></H3>
+-
+-<PRE>
+- typedef void BZFILE;
+-
+- BZFILE *BZ2_bzReadOpen ( int *bzerror, FILE *f,
+- int small, int verbosity,
+- void *unused, int nUnused );
+-</PRE>
+-
+-<P>
+-Prepare to read compressed data from file handle <CODE>f</CODE>. <CODE>f</CODE>
+-should refer to a file which has been opened for reading, and for which
+-the error indicator (<CODE>ferror(f)</CODE>)is not set. If <CODE>small</CODE> is 1,
+-the library will try to decompress using less memory, at the expense of
+-speed.
+-
+-</P>
+-<P>
+-For reasons explained below, <CODE>BZ2_bzRead</CODE> will decompress the
+-<CODE>nUnused</CODE> bytes starting at <CODE>unused</CODE>, before starting to read
+-from the file <CODE>f</CODE>. At most <CODE>BZ_MAX_UNUSED</CODE> bytes may be
+-supplied like this. If this facility is not required, you should pass
+-<CODE>NULL</CODE> and <CODE>0</CODE> for <CODE>unused</CODE> and n<CODE>Unused</CODE>
+-respectively.
+-
+-</P>
+-<P>
+-For the meaning of parameters <CODE>small</CODE> and <CODE>verbosity</CODE>,
+-see <CODE>BZ2_bzDecompressInit</CODE>.
+-
+-</P>
+-<P>
+-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</CODE> returns <CODE>BZ_OK</CODE> but a subsequent call of
+-<CODE>BZ2_bzRead</CODE> will return <CODE>BZ_MEM_ERROR</CODE>.
+-
+-</P>
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_CONFIG_ERROR</CODE>
+- if the library has been mis-compiled
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>f</CODE> is <CODE>NULL</CODE>
+- or <CODE>small</CODE> is neither <CODE>0</CODE> nor <CODE>1</CODE>
+- or <CODE>(unused == NULL && nUnused != 0)</CODE>
+- or <CODE>(unused != NULL && !(0 <= nUnused <= BZ_MAX_UNUSED))</CODE>
+- <CODE>BZ_IO_ERROR</CODE>
+- if <CODE>ferror(f)</CODE> is nonzero
+- <CODE>BZ_MEM_ERROR</CODE>
+- if insufficient memory is available
+- <CODE>BZ_OK</CODE>
+- otherwise.
+-</PRE>
+-
+-<P>
+-Possible return values:
+-
+-<PRE>
+- Pointer to an abstract <CODE>BZFILE</CODE>
+- if <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE>
+- <CODE>NULL</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- <CODE>BZ2_bzRead</CODE>
+- if <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE>
+- <CODE>BZ2_bzClose</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC27" HREF="manual_toc.html#TOC27"><CODE>BZ2_bzRead</CODE></A></H3>
+-
+-<PRE>
+- int BZ2_bzRead ( int *bzerror, BZFILE *b, void *buf, int len );
+-</PRE>
+-
+-<P>
+-Reads up to <CODE>len</CODE> (uncompressed) bytes from the compressed file
+-<CODE>b</CODE> into
+-the buffer <CODE>buf</CODE>. If the read was successful,
+-<CODE>bzerror</CODE> is set to <CODE>BZ_OK</CODE>
+-and the number of bytes read is returned. If the logical end-of-stream
+-was detected, <CODE>bzerror</CODE> will be set to <CODE>BZ_STREAM_END</CODE>,
+-and the number
+-of bytes read is returned. All other <CODE>bzerror</CODE> values denote an error.
+-
+-</P>
+-<P>
+-<CODE>BZ2_bzRead</CODE> will supply <CODE>len</CODE> 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</CODE> after every call and watch out for
+-<CODE>BZ_STREAM_END</CODE>.
+-
+-</P>
+-<P>
+-Internally, <CODE>BZ2_bzRead</CODE> copies data from the compressed file in chunks
+-of size <CODE>BZ_MAX_UNUSED</CODE> bytes
+-before decompressing it. If the file contains more bytes than strictly
+-needed to reach the logical end-of-stream, <CODE>BZ2_bzRead</CODE> will almost certainly
+-read some of the trailing data before signalling <CODE>BZ_SEQUENCE_END</CODE>.
+-To collect the read but unused data once <CODE>BZ_SEQUENCE_END</CODE> has
+-appeared, call <CODE>BZ2_bzReadGetUnused</CODE> immediately before <CODE>BZ2_bzReadClose</CODE>.
+-
+-</P>
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>b</CODE> is <CODE>NULL</CODE> or <CODE>buf</CODE> is <CODE>NULL</CODE> or <CODE>len < 0</CODE>
+- <CODE>BZ_SEQUENCE_ERROR</CODE>
+- if <CODE>b</CODE> was opened with <CODE>BZ2_bzWriteOpen</CODE>
+- <CODE>BZ_IO_ERROR</CODE>
+- if there is an error reading from the compressed file
+- <CODE>BZ_UNEXPECTED_EOF</CODE>
+- if the compressed file ended before the logical end-of-stream was detected
+- <CODE>BZ_DATA_ERROR</CODE>
+- if a data integrity error was detected in the compressed stream
+- <CODE>BZ_DATA_ERROR_MAGIC</CODE>
+- if the stream does not begin with the requisite header bytes (ie, is not
+- a <CODE>bzip2</CODE> data file). This is really a special case of <CODE>BZ_DATA_ERROR</CODE>.
+- <CODE>BZ_MEM_ERROR</CODE>
+- if insufficient memory was available
+- <CODE>BZ_STREAM_END</CODE>
+- if the logical end of stream was detected.
+- <CODE>BZ_OK</CODE>
+- otherwise.
+-</PRE>
+-
+-<P>
+-Possible return values:
+-
+-<PRE>
+- number of bytes read
+- if <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE> or <CODE>BZ_STREAM_END</CODE>
+- undefined
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- collect data from <CODE>buf</CODE>, then <CODE>BZ2_bzRead</CODE> or <CODE>BZ2_bzReadClose</CODE>
+- if <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE>
+- collect data from <CODE>buf</CODE>, then <CODE>BZ2_bzReadClose</CODE> or <CODE>BZ2_bzReadGetUnused</CODE>
+- if <CODE>bzerror</CODE> is <CODE>BZ_SEQUENCE_END</CODE>
+- <CODE>BZ2_bzReadClose</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC28" HREF="manual_toc.html#TOC28"><CODE>BZ2_bzReadGetUnused</CODE></A></H3>
+-
+-<PRE>
+- void BZ2_bzReadGetUnused ( int* bzerror, BZFILE *b,
+- void** unused, int* nUnused );
+-</PRE>
+-
+-<P>
+-Returns data which was read from the compressed file but was not needed
+-to get to the logical end-of-stream. <CODE>*unused</CODE> is set to the address
+-of the data, and <CODE>*nUnused</CODE> to the number of bytes. <CODE>*nUnused</CODE> will
+-be set to a value between <CODE>0</CODE> and <CODE>BZ_MAX_UNUSED</CODE> inclusive.
+-
+-</P>
+-<P>
+-This function may only be called once <CODE>BZ2_bzRead</CODE> has signalled
+-<CODE>BZ_STREAM_END</CODE> but before <CODE>BZ2_bzReadClose</CODE>.
+-
+-</P>
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>b</CODE> is <CODE>NULL</CODE>
+- or <CODE>unused</CODE> is <CODE>NULL</CODE> or <CODE>nUnused</CODE> is <CODE>NULL</CODE>
+- <CODE>BZ_SEQUENCE_ERROR</CODE>
+- if <CODE>BZ_STREAM_END</CODE> has not been signalled
+- or if <CODE>b</CODE> was opened with <CODE>BZ2_bzWriteOpen</CODE>
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- <CODE>BZ2_bzReadClose</CODE>
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC29" HREF="manual_toc.html#TOC29"><CODE>BZ2_bzReadClose</CODE></A></H3>
+-
+-<PRE>
+- void BZ2_bzReadClose ( int *bzerror, BZFILE *b );
+-</PRE>
+-
+-<P>
+-Releases all memory pertaining to the compressed file <CODE>b</CODE>.
+-<CODE>BZ2_bzReadClose</CODE> does not call <CODE>fclose</CODE> on the underlying file
+-handle, so you should do that yourself if appropriate.
+-<CODE>BZ2_bzReadClose</CODE> should be called to clean up after all error
+-situations.
+-
+-</P>
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_SEQUENCE_ERROR</CODE>
+- if <CODE>b</CODE> was opened with <CODE>BZ2_bzOpenWrite</CODE>
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- none
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC30" HREF="manual_toc.html#TOC30"><CODE>BZ2_bzWriteOpen</CODE></A></H3>
+-
+-<PRE>
+- BZFILE *BZ2_bzWriteOpen ( int *bzerror, FILE *f,
+- int blockSize100k, int verbosity,
+- int workFactor );
+-</PRE>
+-
+-<P>
+-Prepare to write compressed data to file handle <CODE>f</CODE>.
+-<CODE>f</CODE> should refer to
+-a file which has been opened for writing, and for which the error
+-indicator (<CODE>ferror(f)</CODE>)is not set.
+-
+-</P>
+-<P>
+-For the meaning of parameters <CODE>blockSize100k</CODE>,
+-<CODE>verbosity</CODE> and <CODE>workFactor</CODE>, see
+-<BR> <CODE>BZ2_bzCompressInit</CODE>.
+-
+-</P>
+-<P>
+-All required memory is allocated at this stage, so if the call
+-completes successfully, <CODE>BZ_MEM_ERROR</CODE> cannot be signalled by a
+-subsequent call to <CODE>BZ2_bzWrite</CODE>.
+-
+-</P>
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_CONFIG_ERROR</CODE>
+- if the library has been mis-compiled
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>f</CODE> is <CODE>NULL</CODE>
+- or <CODE>blockSize100k < 1</CODE> or <CODE>blockSize100k > 9</CODE>
+- <CODE>BZ_IO_ERROR</CODE>
+- if <CODE>ferror(f)</CODE> is nonzero
+- <CODE>BZ_MEM_ERROR</CODE>
+- if insufficient memory is available
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Possible return values:
+-
+-<PRE>
+- Pointer to an abstract <CODE>BZFILE</CODE>
+- if <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE>
+- <CODE>NULL</CODE>
+- otherwise
+-</PRE>
+-
+-<P>
+-Allowable next actions:
+-
+-<PRE>
+- <CODE>BZ2_bzWrite</CODE>
+- if <CODE>bzerror</CODE> is <CODE>BZ_OK</CODE>
+- (you could go directly to <CODE>BZ2_bzWriteClose</CODE>, but this would be pretty pointless)
+- <CODE>BZ2_bzWriteClose</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC31" HREF="manual_toc.html#TOC31"><CODE>BZ2_bzWrite</CODE></A></H3>
+-
+-<PRE>
+- void BZ2_bzWrite ( int *bzerror, BZFILE *b, void *buf, int len );
+-</PRE>
+-
+-<P>
+-Absorbs <CODE>len</CODE> bytes from the buffer <CODE>buf</CODE>, eventually to be
+-compressed and written to the file.
+-
+-</P>
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>b</CODE> is <CODE>NULL</CODE> or <CODE>buf</CODE> is <CODE>NULL</CODE> or <CODE>len < 0</CODE>
+- <CODE>BZ_SEQUENCE_ERROR</CODE>
+- if b was opened with <CODE>BZ2_bzReadOpen</CODE>
+- <CODE>BZ_IO_ERROR</CODE>
+- if there is an error writing the compressed file.
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC32" HREF="manual_toc.html#TOC32"><CODE>BZ2_bzWriteClose</CODE></A></H3>
+-
+-<PRE>
+- 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 );
+-</PRE>
+-
+-<P>
+-Compresses and flushes to the compressed file all data so far supplied
+-by <CODE>BZ2_bzWrite</CODE>. The logical end-of-stream markers are also written, so
+-subsequent calls to <CODE>BZ2_bzWrite</CODE> are illegal. All memory associated
+-with the compressed file <CODE>b</CODE> is released.
+-<CODE>fflush</CODE> is called on the
+-compressed file, but it is not <CODE>fclose</CODE>'d.
+-
+-</P>
+-<P>
+-If <CODE>BZ2_bzWriteClose</CODE> 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</CODE> 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</CODE>.
+-
+-</P>
+-<P>
+-If <CODE>nbytes_in</CODE> is non-null, <CODE>*nbytes_in</CODE> will be set to be the
+-total volume of uncompressed data handled. Similarly, <CODE>nbytes_out</CODE>
+-will be set to the total volume of compressed data written. For
+-compatibility with older versions of the library, <CODE>BZ2_bzWriteClose</CODE>
+-only yields the lower 32 bits of these counts. Use
+-<CODE>BZ2_bzWriteClose64</CODE> if you want the full 64 bit counts. These
+-two functions are otherwise absolutely identical.
+-
+-</P>
+-
+-<P>
+-Possible assignments to <CODE>bzerror</CODE>:
+-
+-<PRE>
+- <CODE>BZ_SEQUENCE_ERROR</CODE>
+- if <CODE>b</CODE> was opened with <CODE>BZ2_bzReadOpen</CODE>
+- <CODE>BZ_IO_ERROR</CODE>
+- if there is an error writing the compressed file
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC33" HREF="manual_toc.html#TOC33">Handling embedded compressed data streams</A></H3>
+-
+-<P>
+-The high-level library facilitates use of
+-<CODE>bzip2</CODE> data streams which form some part of a surrounding, larger
+-data stream.
+-
+-<UL>
+-<LI>For writing, the library takes an open file handle, writes
+-
+-compressed data to it, <CODE>fflush</CODE>es it but does not <CODE>fclose</CODE> it.
+-The calling application can write its own data before and after the
+-compressed data stream, using that same file handle.
+-<LI>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</CODE> reads from the compressed file in blocks of size
+-<CODE>BZ_MAX_UNUSED</CODE> 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</CODE> after the last call of <CODE>BZ2_bzRead</CODE>
+-(the one returning <CODE>BZ_STREAM_END</CODE>) but before calling
+-<CODE>BZ2_bzReadClose</CODE>.
+-</UL>
+-
+-<P>
+-This mechanism makes it easy to decompress multiple <CODE>bzip2</CODE>
+-streams placed end-to-end. As the end of one stream, when <CODE>BZ2_bzRead</CODE>
+-returns <CODE>BZ_STREAM_END</CODE>, call <CODE>BZ2_bzReadGetUnused</CODE> 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</CODE> again,
+-feeding in the unused data via the <CODE>unused</CODE>/<CODE>nUnused</CODE>
+-parameters.
+-Keep doing this until <CODE>BZ_STREAM_END</CODE> return coincides with the
+-physical end of file (<CODE>feof(f)</CODE>). In this situation
+-<CODE>BZ2_bzReadGetUnused</CODE>
+-will of course return no data.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-
+-
+-<H3><A NAME="SEC34" HREF="manual_toc.html#TOC34">Standard file-reading/writing code</A></H3>
+-<P>
+-Here's how you'd write data to a compressed file:
+-
+-<PRE>
+-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 */
+-}
+-</PRE>
+-
+-<P>
+-And to read from a compressed file:
+-
+-<PRE>
+-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 );
+-}
+-</PRE>
+-
+-
+-
+-<H2><A NAME="SEC35" HREF="manual_toc.html#TOC35">Utility functions</A></H2>
+-
+-
+-<H3><A NAME="SEC36" HREF="manual_toc.html#TOC36"><CODE>BZ2_bzBuffToBuffCompress</CODE></A></H3>
+-
+-<PRE>
+- int BZ2_bzBuffToBuffCompress( char* dest,
+- unsigned int* destLen,
+- char* source,
+- unsigned int sourceLen,
+- int blockSize100k,
+- int verbosity,
+- int workFactor );
+-</PRE>
+-
+-<P>
+-Attempts to compress the data in <CODE>source[0 .. sourceLen-1]</CODE>
+-into the destination buffer, <CODE>dest[0 .. *destLen-1]</CODE>.
+-If the destination buffer is big enough, <CODE>*destLen</CODE> is
+-set to the size of the compressed data, and <CODE>BZ_OK</CODE> is
+-returned. If the compressed data won't fit, <CODE>*destLen</CODE>
+-is unchanged, and <CODE>BZ_OUTBUFF_FULL</CODE> is returned.
+-
+-</P>
+-<P>
+-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</CODE> 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.
+-
+-</P>
+-<P>
+-For the meaning of parameters <CODE>blockSize100k</CODE>, <CODE>verbosity</CODE>
+-and <CODE>workFactor</CODE>, <BR> see <CODE>BZ2_bzCompressInit</CODE>.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-<CODE>BZ2_bzBuffToBuffDecompress</CODE> will not write data at or
+-beyond <CODE>dest[*destLen]</CODE>, even in case of buffer overflow.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_CONFIG_ERROR</CODE>
+- if the library has been mis-compiled
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>dest</CODE> is <CODE>NULL</CODE> or <CODE>destLen</CODE> is <CODE>NULL</CODE>
+- or <CODE>blockSize100k < 1</CODE> or <CODE>blockSize100k > 9</CODE>
+- or <CODE>verbosity < 0</CODE> or <CODE>verbosity > 4</CODE>
+- or <CODE>workFactor < 0</CODE> or <CODE>workFactor > 250</CODE>
+- <CODE>BZ_MEM_ERROR</CODE>
+- if insufficient memory is available
+- <CODE>BZ_OUTBUFF_FULL</CODE>
+- if the size of the compressed data exceeds <CODE>*destLen</CODE>
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H3><A NAME="SEC37" HREF="manual_toc.html#TOC37"><CODE>BZ2_bzBuffToBuffDecompress</CODE></A></H3>
+-
+-<PRE>
+- int BZ2_bzBuffToBuffDecompress ( char* dest,
+- unsigned int* destLen,
+- char* source,
+- unsigned int sourceLen,
+- int small,
+- int verbosity );
+-</PRE>
+-
+-<P>
+-Attempts to decompress the data in <CODE>source[0 .. sourceLen-1]</CODE>
+-into the destination buffer, <CODE>dest[0 .. *destLen-1]</CODE>.
+-If the destination buffer is big enough, <CODE>*destLen</CODE> is
+-set to the size of the uncompressed data, and <CODE>BZ_OK</CODE> is
+-returned. If the compressed data won't fit, <CODE>*destLen</CODE>
+-is unchanged, and <CODE>BZ_OUTBUFF_FULL</CODE> is returned.
+-
+-</P>
+-<P>
+-<CODE>source</CODE> is assumed to hold a complete <CODE>bzip2</CODE> format
+-data stream. <BR> <CODE>BZ2_bzBuffToBuffDecompress</CODE> tries to decompress
+-the entirety of the stream into the output buffer.
+-
+-</P>
+-<P>
+-For the meaning of parameters <CODE>small</CODE> and <CODE>verbosity</CODE>,
+-see <CODE>BZ2_bzDecompressInit</CODE>.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-<CODE>BZ2_bzBuffToBuffDecompress</CODE> will not write data at or
+-beyond <CODE>dest[*destLen]</CODE>, even in case of buffer overflow.
+-
+-</P>
+-<P>
+-Possible return values:
+-
+-<PRE>
+- <CODE>BZ_CONFIG_ERROR</CODE>
+- if the library has been mis-compiled
+- <CODE>BZ_PARAM_ERROR</CODE>
+- if <CODE>dest</CODE> is <CODE>NULL</CODE> or <CODE>destLen</CODE> is <CODE>NULL</CODE>
+- or <CODE>small != 0 && small != 1</CODE>
+- or <CODE>verbosity < 0</CODE> or <CODE>verbosity > 4</CODE>
+- <CODE>BZ_MEM_ERROR</CODE>
+- if insufficient memory is available
+- <CODE>BZ_OUTBUFF_FULL</CODE>
+- if the size of the compressed data exceeds <CODE>*destLen</CODE>
+- <CODE>BZ_DATA_ERROR</CODE>
+- if a data integrity error was detected in the compressed data
+- <CODE>BZ_DATA_ERROR_MAGIC</CODE>
+- if the compressed data doesn't begin with the right magic bytes
+- <CODE>BZ_UNEXPECTED_EOF</CODE>
+- if the compressed data ends unexpectedly
+- <CODE>BZ_OK</CODE>
+- otherwise
+-</PRE>
+-
+-
+-
+-<H2><A NAME="SEC38" HREF="manual_toc.html#TOC38"><CODE>zlib</CODE> compatibility functions</A></H2>
+-<P>
+-Yoshioka Tsuneo has contributed some functions to
+-give better <CODE>zlib</CODE> compatibility. These functions are
+-<CODE>BZ2_bzopen</CODE>, <CODE>BZ2_bzread</CODE>, <CODE>BZ2_bzwrite</CODE>, <CODE>BZ2_bzflush</CODE>,
+-<CODE>BZ2_bzclose</CODE>,
+-<CODE>BZ2_bzerror</CODE> and <CODE>BZ2_bzlibVersion</CODE>.
+-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.
+-
+-<PRE>
+-typedef void BZFILE;
+-
+-const char * BZ2_bzlibVersion ( void );
+-</PRE>
+-
+-<P>
+-Returns a string indicating the library version.
+-
+-<PRE>
+-BZFILE * BZ2_bzopen ( const char *path, const char *mode );
+-BZFILE * BZ2_bzdopen ( int fd, const char *mode );
+-</PRE>
+-
+-<P>
+-Opens a <CODE>.bz2</CODE> file for reading or writing, using either its name
+-or a pre-existing file descriptor.
+-Analogous to <CODE>fopen</CODE> and <CODE>fdopen</CODE>.
+-
+-<PRE>
+-int BZ2_bzread ( BZFILE* b, void* buf, int len );
+-int BZ2_bzwrite ( BZFILE* b, void* buf, int len );
+-</PRE>
+-
+-<P>
+-Reads/writes data from/to a previously opened <CODE>BZFILE</CODE>.
+-Analogous to <CODE>fread</CODE> and <CODE>fwrite</CODE>.
+-
+-<PRE>
+-int BZ2_bzflush ( BZFILE* b );
+-void BZ2_bzclose ( BZFILE* b );
+-</PRE>
+-
+-<P>
+-Flushes/closes a <CODE>BZFILE</CODE>. <CODE>BZ2_bzflush</CODE> doesn't actually do
+-anything. Analogous to <CODE>fflush</CODE> and <CODE>fclose</CODE>.
+-
+-</P>
+-
+-<PRE>
+-const char * BZ2_bzerror ( BZFILE *b, int *errnum )
+-</PRE>
+-
+-<P>
+-Returns a string describing the more recent error status of
+-<CODE>b</CODE>, and also sets <CODE>*errnum</CODE> to its numerical value.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC39" HREF="manual_toc.html#TOC39">Using the library in a <CODE>stdio</CODE>-free environment</A></H2>
+-
+-
+-
+-<H3><A NAME="SEC40" HREF="manual_toc.html#TOC40">Getting rid of <CODE>stdio</CODE></A></H3>
+-
+-<P>
+-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</CODE>
+-defined. Doing this gives you a library containing only the following
+-eight functions:
+-
+-</P>
+-<P>
+-<CODE>BZ2_bzCompressInit</CODE>, <CODE>BZ2_bzCompress</CODE>, <CODE>BZ2_bzCompressEnd</CODE> <BR>
+-<CODE>BZ2_bzDecompressInit</CODE>, <CODE>BZ2_bzDecompress</CODE>, <CODE>BZ2_bzDecompressEnd</CODE> <BR>
+-<CODE>BZ2_bzBuffToBuffCompress</CODE>, <CODE>BZ2_bzBuffToBuffDecompress</CODE>
+-
+-</P>
+-<P>
+-When compiled like this, all functions will ignore <CODE>verbosity</CODE>
+-settings.
+-
+-</P>
+-
+-
+-<H3><A NAME="SEC41" HREF="manual_toc.html#TOC41">Critical error handling</A></H3>
+-<P>
+-<CODE>libbzip2</CODE> 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</CODE> set.
+-
+-</P>
+-<P>
+-For a normal compile, an assertion failure yields the message
+-
+-<PRE>
+- bzip2/libbzip2: internal error number N.
+- This is a bug in bzip2/libbzip2, 1.0 of 21-Mar-2000.
+- 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, 21 March 2000.
+-</PRE>
+-
+-<P>
+-where <CODE>N</CODE> is some error code number. <CODE>exit(3)</CODE>
+-is then called.
+-
+-</P>
+-<P>
+-For a <CODE>stdio</CODE>-free library, assertion failures result
+-in a call to a function declared as:
+-
+-<PRE>
+- extern void bz_internal_error ( int errcode );
+-</PRE>
+-
+-<P>
+-The relevant code is passed as a parameter. You should supply
+-such a function.
+-
+-</P>
+-<P>
+-In either case, once an assertion failure has occurred, any
+-<CODE>bz_stream</CODE> records involved can be regarded as invalid.
+-You should not attempt to resume normal operation with them.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC42" HREF="manual_toc.html#TOC42">Making a Windows DLL</A></H2>
+-<P>
+-Everything related to Windows has been contributed by Yoshioka Tsuneo
+-<BR> (<CODE>QWF00133@niftyserve.or.jp</CODE> /
+-<CODE>tsuneo-y@is.aist-nara.ac.jp</CODE>), so you should send your queries to
+-him (but perhaps Cc: me, <CODE>jseward@acm.org</CODE>).
+-
+-</P>
+-<P>
+-My vague understanding of what to do is: using Visual C++ 5.0,
+-open the project file <CODE>libbz2.dsp</CODE>, and build. That's all.
+-
+-</P>
+-<P>
+-If you can't
+-open the project file for some reason, make a new one, naming these files:
+-<CODE>blocksort.c</CODE>, <CODE>bzlib.c</CODE>, <CODE>compress.c</CODE>,
+-<CODE>crctable.c</CODE>, <CODE>decompress.c</CODE>, <CODE>huffman.c</CODE>, <BR>
+-<CODE>randtable.c</CODE> and <CODE>libbz2.def</CODE>. You will also need
+-to name the header files <CODE>bzlib.h</CODE> and <CODE>bzlib_private.h</CODE>.
+-
+-</P>
+-<P>
+-If you don't use VC++, you may need to define the proprocessor symbol
+-<CODE>_WIN32</CODE>.
+-
+-</P>
+-<P>
+-Finally, <CODE>dlltest.c</CODE> is a sample program using the DLL. It has a
+-project file, <CODE>dlltest.dsp</CODE>.
+-
+-</P>
+-<P>
+-If you just want a makefile for Visual C, have a look at
+-<CODE>makefile.msc</CODE>.
+-
+-</P>
+-<P>
+-Be aware that if you compile <CODE>bzip2</CODE> itself on Win32, you must set
+-<CODE>BZ_UNIX</CODE> to 0 and <CODE>BZ_LCCWIN32</CODE> to 1, in the file
+-<CODE>bzip2.c</CODE>, before compiling. Otherwise the resulting binary won't
+-work correctly.
+-
+-</P>
+-<P>
+-I haven't tried any of this stuff myself, but it all looks plausible.
+-
+-</P>
+-
+-<P><HR><P>
+-<p>Go to the <A HREF="manual_1.html">first</A>, <A HREF="manual_2.html">previous</A>, <A HREF="manual_4.html">next</A>, <A HREF="manual_4.html">last</A> section, <A HREF="manual_toc.html">table of contents</A>.
+-</BODY>
+-</HTML>
+diff -Nru bzip2-1.0.1/manual_4.html bzip2-1.0.1.new/manual_4.html
+--- bzip2-1.0.1/manual_4.html Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual_4.html Thu Jan 1 01:00:00 1970
+@@ -1,528 +0,0 @@
+-<HTML>
+-<HEAD>
+-<!-- This HTML file has been created by texi2html 1.54
+- from manual.texi on 23 March 2000 -->
+-
+-<TITLE>bzip2 and libbzip2 - Miscellanea</TITLE>
+-<link href="manual_3.html" rel=Previous>
+-<link href="manual_toc.html" rel=ToC>
+-
+-</HEAD>
+-<BODY>
+-<p>Go to the <A HREF="manual_1.html">first</A>, <A HREF="manual_3.html">previous</A>, next, last section, <A HREF="manual_toc.html">table of contents</A>.
+-<P><HR><P>
+-
+-
+-<H1><A NAME="SEC43" HREF="manual_toc.html#TOC43">Miscellanea</A></H1>
+-
+-<P>
+-These are just some random thoughts of mine. Your mileage may
+-vary.
+-
+-</P>
+-
+-
+-<H2><A NAME="SEC44" HREF="manual_toc.html#TOC44">Limitations of the compressed file format</A></H2>
+-<P>
+-<CODE>bzip2-1.0</CODE>, <CODE>0.9.5</CODE> and <CODE>0.9.0</CODE>
+-use exactly the same file format as the previous
+-version, <CODE>bzip2-0.1</CODE>. This decision was made in the interests of
+-stability. Creating yet another incompatible compressed file format
+-would create further confusion and disruption for users.
+-
+-</P>
+-<P>
+-Nevertheless, this is not a painless decision. Development
+-work since the release of <CODE>bzip2-0.1</CODE> in August 1997
+-has shown complexities in the file format which slow down
+-decompression and, in retrospect, are unnecessary. These are:
+-
+-<UL>
+-<LI>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.
+-<LI>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</CODE>'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</CODE>'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.
+-<LI>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</CODE> 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.
+-<LI>An Adler-32 checksum, rather than a CRC32 checksum,
+-
+- would be faster to compute.
+-</UL>
+-
+-<P>
+-It would be fair to say that the <CODE>bzip2</CODE> format was frozen
+-before I properly and fully understood the performance
+-consequences of doing so.
+-
+-</P>
+-<P>
+-Improvements which I was able to incorporate into
+-0.9.0, despite using the same file format, are:
+-
+-<UL>
+-<LI>Single array implementation of the inverse BWT. This
+-
+- significantly speeds up decompression, presumably
+- because it reduces the number of cache misses.
+-<LI>Faster inverse MTF transform for large MTF values. The
+-
+- new implementation is based on the notion of sliding blocks
+- of values.
+-<LI><CODE>bzip2-0.9.0</CODE> now reads and writes files with <CODE>fread</CODE>
+-
+- and <CODE>fwrite</CODE>; version 0.1 used <CODE>putc</CODE> and <CODE>getc</CODE>.
+- Duh! Well, you live and learn.
+-
+-</UL>
+-
+-<P>
+-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.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC45" HREF="manual_toc.html#TOC45">Portability issues</A></H2>
+-<P>
+-After some consideration, I have decided not to use
+-GNU <CODE>autoconf</CODE> to configure 0.9.5 or 1.0.
+-
+-</P>
+-<P>
+-<CODE>autoconf</CODE>, admirable and wonderful though it is,
+-mainly assists with portability problems between Unix-like
+-platforms. But <CODE>bzip2</CODE> 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</CODE> doesn't help in those cases, and brings in a
+-whole load of new complexity.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-There are a couple of <CODE>__inline__</CODE> directives in the code. GNU C
+-(<CODE>gcc</CODE>) 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> <CODE>__inline__</CODE> to be <CODE>/* */</CODE>. One
+-easy way to do this is to compile with the flag <CODE>-D__inline__=</CODE>,
+-which should be understood by most Unix compilers.
+-
+-</P>
+-<P>
+-If you still have difficulties, try compiling with the macro
+-<CODE>BZ_STRICT_ANSI</CODE> 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</CODE>'s checks against compressing directories, symbolic links,
+-devices, and other not-really-a-file entities. This could cause
+-filesystem corruption!
+-
+-</P>
+-<P>
+-One other thing: if you create a <CODE>bzip2</CODE> binary for public
+-distribution, please try and link it statically (<CODE>gcc -s</CODE>). This
+-avoids all sorts of library-version issues that others may encounter
+-later on.
+-
+-</P>
+-<P>
+-If you build <CODE>bzip2</CODE> on Win32, you must set <CODE>BZ_UNIX</CODE> to 0 and
+-<CODE>BZ_LCCWIN32</CODE> to 1, in the file <CODE>bzip2.c</CODE>, before compiling.
+-Otherwise the resulting binary won't work correctly.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC46" HREF="manual_toc.html#TOC46">Reporting bugs</A></H2>
+-<P>
+-I tried pretty hard to make sure <CODE>bzip2</CODE> is
+-bug free, both by design and by testing. Hopefully
+-you'll never need to read this section for real.
+-
+-</P>
+-<P>
+-Nevertheless, if <CODE>bzip2</CODE> 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.
+-
+-<UL>
+-<LI>
+-
+-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</CODE>, 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> <CODE>-fomit-frame-pointer</CODE> <CODE>-fno-strength-reduce</CODE>.
+-You should specifically <EM>not</EM> use <CODE>-funroll-loops</CODE>.
+-
+-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.
+-<LI>
+-
+-If <CODE>bzip2</CODE> crashes randomly, and the crashes are not
+-repeatable, you may have a flaky memory subsystem. <CODE>bzip2</CODE>
+-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.
+-<LI>This isn't really a bug, but ... If <CODE>bzip2</CODE> 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.
+-</UL>
+-
+-<P>
+-If you've incorporated <CODE>libbzip2</CODE> 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.
+-
+-</P>
+-<P>
+-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
+-
+-<PRE>
+- bzip2 crashed with segmentation fault on my machine
+-</PRE>
+-
+-<P>
+-and absolutely nothing else. Needless to say, a such a report
+-is <EM>totally, utterly, completely and comprehensively 100% useless;
+-a waste of your time, my time, and net bandwidth</EM>.
+-With no details at all, there's no way I can possibly begin
+-to figure out what the problem is.
+-
+-</P>
+-<P>
+-The rules of the game are: facts, facts, facts. Don't omit
+-them because "oh, they won't be relevant". At the bare
+-minimum:
+-
+-<PRE>
+- Machine type. Operating system version.
+- Exact version of <CODE>bzip2</CODE> (do <CODE>bzip2 -V</CODE>).
+- Exact version of the compiler used.
+- Flags passed to the compiler.
+-</PRE>
+-
+-<P>
+-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.
+-
+-</P>
+-<P>
+-Please remember that I connect to the Internet with a modem, so
+-you should contact me before mailing me huge files.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC47" HREF="manual_toc.html#TOC47">Did you get the right package?</A></H2>
+-
+-<P>
+-<CODE>bzip2</CODE> 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.
+-
+-</P>
+-<P>
+-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.
+-
+-</P>
+-<P>
+-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-1.1.2</CODE> and
+-<CODE>gzip-1.2.4</CODE>. Look for them at
+-
+-</P>
+-<P>
+-<CODE>http://www.cdrom.com/pub/infozip/zlib</CODE> and
+-<CODE>http://www.gzip.org</CODE> respectively.
+-
+-</P>
+-<P>
+-For something faster and lighter still, you might try Markus F X J
+-Oberhumer's <CODE>LZO</CODE> real-time compression/decompression library, at
+-<BR> <CODE>http://wildsau.idv.uni-linz.ac.at/mfx/lzo.html</CODE>.
+-
+-</P>
+-<P>
+-If you want to use the <CODE>bzip2</CODE> 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</CODE>, an on-the-fly disk compressor for the Linux
+-<CODE>ext2</CODE> filesystem. Look at
+-<CODE>http://www.netspace.net.au/~reiter/e2compr</CODE>.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC48" HREF="manual_toc.html#TOC48">Testing</A></H2>
+-
+-<P>
+-A record of the tests I've done.
+-
+-</P>
+-<P>
+-First, some data sets:
+-
+-<UL>
+-<LI>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.
+-<LI>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.
+-<LI>A: directory tree holding various applications built from source:
+-
+- <CODE>egcs</CODE>, <CODE>gcc-2.8.1</CODE>, KDE, GTK, Octave, etc.
+- 2200 megabytes.
+-</UL>
+-
+-<P>
+-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.
+-
+-</P>
+-<P>
+-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.
+-
+-<OL>
+-<LI>Data set H, with
+-
+- buffer size of 1 byte, and block size of 23 bytes.
+-<LI>Data set B, buffer sizes 1 byte, block size 1 byte.
+-
+-<LI>As (2) but small-mode decompression.
+-
+-<LI>As (2) with block size 2 bytes.
+-
+-<LI>As (2) with block size 3 bytes.
+-
+-<LI>As (2) with block size 4 bytes.
+-
+-<LI>As (2) with block size 5 bytes.
+-
+-<LI>As (2) with block size 6 bytes and small-mode decompression.
+-
+-<LI>H with buffer size of 1 byte, but normal block
+-
+- size (up to 900000 bytes).
+-</OL>
+-
+-<P>
+-Then some tests with unmodified source code.
+-
+-<OL>
+-<LI>H, all settings normal.
+-
+-<LI>As (1), with small-mode decompress.
+-
+-<LI>H, compress with flag <CODE>-1</CODE>.
+-
+-<LI>H, compress with flag <CODE>-s</CODE>, decompress with flag <CODE>-s</CODE>.
+-
+-<LI>Forwards compatibility: H, <CODE>bzip2-0.1pl2</CODE> compressing,
+-
+- <CODE>bzip2-0.9.5</CODE> decompressing, all settings normal.
+-<LI>Backwards compatibility: H, <CODE>bzip2-0.9.5</CODE> compressing,
+-
+- <CODE>bzip2-0.1pl2</CODE> decompressing, all settings normal.
+-<LI>Bigger tests: A, all settings normal.
+-
+-<LI>As (7), using the fallback (Sadakane-like) sorting algorithm.
+-
+-<LI>As (8), compress with flag <CODE>-1</CODE>, decompress with flag
+-
+- <CODE>-s</CODE>.
+-<LI>H, using the fallback sorting algorithm.
+-
+-<LI>Forwards compatibility: A, <CODE>bzip2-0.1pl2</CODE> compressing,
+-
+- <CODE>bzip2-0.9.5</CODE> decompressing, all settings normal.
+-<LI>Backwards compatibility: A, <CODE>bzip2-0.9.5</CODE> compressing,
+-
+- <CODE>bzip2-0.1pl2</CODE> decompressing, all settings normal.
+-<LI>Misc test: about 400 megabytes of <CODE>.tar</CODE> files with
+-
+- <CODE>bzip2</CODE> compiled with Checker (a memory access error
+- detector, like Purify).
+-<LI>Misc tests to make sure it builds and runs ok on non-Linux/x86
+-
+- platforms.
+-</OL>
+-
+-<P>
+-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.
+-
+-</P>
+-
+-
+-
+-<H2><A NAME="SEC49" HREF="manual_toc.html#TOC49">Further reading</A></H2>
+-<P>
+-<CODE>bzip2</CODE> 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.
+-
+-</P>
+-<P>
+-Four documents describe essentially all the ideas behind <CODE>bzip2</CODE>:
+-
+-<PRE>
+-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
+-</PRE>
+-
+-<P>
+-The following paper gives valuable additional insights into the
+-algorithm, but is not immediately the basis of any code
+-used in bzip2.
+-
+-<PRE>
+-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
+-</PRE>
+-
+-<P>
+-Kunihiko Sadakane's sorting algorithm, mentioned above,
+-is available from:
+-
+-<PRE>
+-http://naomi.is.s.u-tokyo.ac.jp/~sada/papers/Sada98b.ps.gz
+-</PRE>
+-
+-<P>
+-The Manber-Myers suffix array construction
+-algorithm is described in a paper
+-available from:
+-
+-<PRE>
+-http://www.cs.arizona.edu/people/gene/PAPERS/suffix.ps
+-</PRE>
+-
+-<P>
+-Finally, the following paper documents some recent investigations
+-I made into the performance of sorting algorithms:
+-
+-<PRE>
+-Julian Seward:
+- On the Performance of BWT Sorting Algorithms
+- Proceedings of the IEEE Data Compression Conference 2000
+- Snowbird, Utah. 28-30 March 2000.
+-</PRE>
+-
+-<P><HR><P>
+-<p>Go to the <A HREF="manual_1.html">first</A>, <A HREF="manual_3.html">previous</A>, next, last section, <A HREF="manual_toc.html">table of contents</A>.
+-</BODY>
+-</HTML>
+diff -Nru bzip2-1.0.1/manual_toc.html bzip2-1.0.1.new/manual_toc.html
+--- bzip2-1.0.1/manual_toc.html Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/manual_toc.html Thu Jan 1 01:00:00 1970
+@@ -1,173 +0,0 @@
+-<HTML>
+-<HEAD>
+-<!-- This HTML file has been created by texi2html 1.54
+- from manual.texi on 23 March 2000 -->
+-
+-<TITLE>bzip2 and libbzip2 - Table of Contents</TITLE>
+-
+-</HEAD>
+-<BODY>
+-<H1>bzip2 and libbzip2</H1>
+-<H2>a program and library for data compression</H2>
+-<H2>copyright (C) 1996-2000 Julian Seward</H2>
+-<H2>version 1.0 of 21 March 2000</H2>
+-<ADDRESS>Julian Seward</ADDRESS>
+-<P>
+-<P><HR><P>
+-
+-<P>
+-This program, <CODE>bzip2</CODE>,
+-and associated library <CODE>libbzip2</CODE>, are
+-Copyright (C) 1996-2000 Julian R Seward. All rights reserved.
+-
+-</P>
+-<P>
+-Redistribution and use in source and binary forms, with or without
+-modification, are permitted provided that the following conditions
+-are met:
+-
+-<UL>
+-<LI>
+-
+- Redistributions of source code must retain the above copyright
+- notice, this list of conditions and the following disclaimer.
+-<LI>
+-
+- 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.
+-<LI>
+-
+- Altered source versions must be plainly marked as such, and must
+- not be misrepresented as being the original software.
+-<LI>
+-
+- The name of the author may not be used to endorse or promote
+- products derived from this software without specific prior written
+- permission.
+-</UL>
+-
+-<P>
+-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.
+-
+-</P>
+-<P>
+-Julian Seward, Cambridge, UK.
+-
+-</P>
+-<P>
+-<CODE>jseward@acm.org</CODE>
+-
+-</P>
+-<P>
+-<CODE>http://sourceware.cygnus.com/bzip2</CODE>
+-
+-</P>
+-<P>
+-<CODE>http://www.cacheprof.org</CODE>
+-
+-</P>
+-<P>
+-<CODE>http://www.muraroa.demon.co.uk</CODE>
+-
+-</P>
+-<P>
+-<CODE>bzip2</CODE>/<CODE>libbzip2</CODE> version 1.0 of 21 March 2000.
+-
+-</P>
+-<P>
+-PATENTS: To the best of my knowledge, <CODE>bzip2</CODE> 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.
+-
+-</P>
+-
+-<UL>
+-<LI><A NAME="TOC1" HREF="manual_1.html#SEC1">Introduction</A>
+-<LI><A NAME="TOC2" HREF="manual_2.html#SEC2">How to use <CODE>bzip2</CODE></A>
+-<UL>
+-<UL>
+-<UL>
+-<LI><A NAME="TOC3" HREF="manual_2.html#SEC3">NAME</A>
+-<LI><A NAME="TOC4" HREF="manual_2.html#SEC4">SYNOPSIS</A>
+-<LI><A NAME="TOC5" HREF="manual_2.html#SEC5">DESCRIPTION</A>
+-<LI><A NAME="TOC6" HREF="manual_2.html#SEC6">OPTIONS</A>
+-<LI><A NAME="TOC7" HREF="manual_2.html#SEC7">MEMORY MANAGEMENT</A>
+-<LI><A NAME="TOC8" HREF="manual_2.html#SEC8">RECOVERING DATA FROM DAMAGED FILES</A>
+-<LI><A NAME="TOC9" HREF="manual_2.html#SEC9">PERFORMANCE NOTES</A>
+-<LI><A NAME="TOC10" HREF="manual_2.html#SEC10">CAVEATS</A>
+-<LI><A NAME="TOC11" HREF="manual_2.html#SEC11">AUTHOR</A>
+-</UL>
+-</UL>
+-</UL>
+-<LI><A NAME="TOC12" HREF="manual_3.html#SEC12">Programming with <CODE>libbzip2</CODE></A>
+-<UL>
+-<LI><A NAME="TOC13" HREF="manual_3.html#SEC13">Top-level structure</A>
+-<UL>
+-<LI><A NAME="TOC14" HREF="manual_3.html#SEC14">Low-level summary</A>
+-<LI><A NAME="TOC15" HREF="manual_3.html#SEC15">High-level summary</A>
+-<LI><A NAME="TOC16" HREF="manual_3.html#SEC16">Utility functions summary</A>
+-</UL>
+-<LI><A NAME="TOC17" HREF="manual_3.html#SEC17">Error handling</A>
+-<LI><A NAME="TOC18" HREF="manual_3.html#SEC18">Low-level interface</A>
+-<UL>
+-<LI><A NAME="TOC19" HREF="manual_3.html#SEC19"><CODE>BZ2_bzCompressInit</CODE></A>
+-<LI><A NAME="TOC20" HREF="manual_3.html#SEC20"><CODE>BZ2_bzCompress</CODE></A>
+-<LI><A NAME="TOC21" HREF="manual_3.html#SEC21"><CODE>BZ2_bzCompressEnd</CODE></A>
+-<LI><A NAME="TOC22" HREF="manual_3.html#SEC22"><CODE>BZ2_bzDecompressInit</CODE></A>
+-<LI><A NAME="TOC23" HREF="manual_3.html#SEC23"><CODE>BZ2_bzDecompress</CODE></A>
+-<LI><A NAME="TOC24" HREF="manual_3.html#SEC24"><CODE>BZ2_bzDecompressEnd</CODE></A>
+-</UL>
+-<LI><A NAME="TOC25" HREF="manual_3.html#SEC25">High-level interface</A>
+-<UL>
+-<LI><A NAME="TOC26" HREF="manual_3.html#SEC26"><CODE>BZ2_bzReadOpen</CODE></A>
+-<LI><A NAME="TOC27" HREF="manual_3.html#SEC27"><CODE>BZ2_bzRead</CODE></A>
+-<LI><A NAME="TOC28" HREF="manual_3.html#SEC28"><CODE>BZ2_bzReadGetUnused</CODE></A>
+-<LI><A NAME="TOC29" HREF="manual_3.html#SEC29"><CODE>BZ2_bzReadClose</CODE></A>
+-<LI><A NAME="TOC30" HREF="manual_3.html#SEC30"><CODE>BZ2_bzWriteOpen</CODE></A>
+-<LI><A NAME="TOC31" HREF="manual_3.html#SEC31"><CODE>BZ2_bzWrite</CODE></A>
+-<LI><A NAME="TOC32" HREF="manual_3.html#SEC32"><CODE>BZ2_bzWriteClose</CODE></A>
+-<LI><A NAME="TOC33" HREF="manual_3.html#SEC33">Handling embedded compressed data streams</A>
+-<LI><A NAME="TOC34" HREF="manual_3.html#SEC34">Standard file-reading/writing code</A>
+-</UL>
+-<LI><A NAME="TOC35" HREF="manual_3.html#SEC35">Utility functions</A>
+-<UL>
+-<LI><A NAME="TOC36" HREF="manual_3.html#SEC36"><CODE>BZ2_bzBuffToBuffCompress</CODE></A>
+-<LI><A NAME="TOC37" HREF="manual_3.html#SEC37"><CODE>BZ2_bzBuffToBuffDecompress</CODE></A>
+-</UL>
+-<LI><A NAME="TOC38" HREF="manual_3.html#SEC38"><CODE>zlib</CODE> compatibility functions</A>
+-<LI><A NAME="TOC39" HREF="manual_3.html#SEC39">Using the library in a <CODE>stdio</CODE>-free environment</A>
+-<UL>
+-<LI><A NAME="TOC40" HREF="manual_3.html#SEC40">Getting rid of <CODE>stdio</CODE></A>
+-<LI><A NAME="TOC41" HREF="manual_3.html#SEC41">Critical error handling</A>
+-</UL>
+-<LI><A NAME="TOC42" HREF="manual_3.html#SEC42">Making a Windows DLL</A>
+-</UL>
+-<LI><A NAME="TOC43" HREF="manual_4.html#SEC43">Miscellanea</A>
+-<UL>
+-<LI><A NAME="TOC44" HREF="manual_4.html#SEC44">Limitations of the compressed file format</A>
+-<LI><A NAME="TOC45" HREF="manual_4.html#SEC45">Portability issues</A>
+-<LI><A NAME="TOC46" HREF="manual_4.html#SEC46">Reporting bugs</A>
+-<LI><A NAME="TOC47" HREF="manual_4.html#SEC47">Did you get the right package?</A>
+-<LI><A NAME="TOC48" HREF="manual_4.html#SEC48">Testing</A>
+-<LI><A NAME="TOC49" HREF="manual_4.html#SEC49">Further reading</A>
+-</UL>
+-</UL>
+-<P><HR><P>
+-This document was generated on 23 March 2000 using the
+-<A HREF="http://wwwcn.cern.ch/dci/texi2html/">texi2html</A>
+-translator version 1.51a.</P>
+-</BODY>
+-</HTML>
+diff -Nru bzip2-1.0.1/randtable.c bzip2-1.0.1.new/randtable.c
+--- bzip2-1.0.1/randtable.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/randtable.c Sat Jun 24 20:13:06 2000
+@@ -58,6 +58,10 @@
+ For more information on these sources, see the manual.
+ --*/
+
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
+
+ #include "bzlib_private.h"
+
+diff -Nru bzip2-1.0.1/spewG.c bzip2-1.0.1.new/spewG.c
+--- bzip2-1.0.1/spewG.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/spewG.c Sat Jun 24 20:13:06 2000
+@@ -9,7 +9,10 @@
+ (but is otherwise harmless).
+ */
+
+-#define _FILE_OFFSET_BITS 64
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
+
+ #include <stdio.h>
+ #include <stdlib.h>
+diff -Nru bzip2-1.0.1/stamp-h.in bzip2-1.0.1.new/stamp-h.in
+--- bzip2-1.0.1/stamp-h.in Thu Jan 1 01:00:00 1970
++++ bzip2-1.0.1.new/stamp-h.in Sat Jun 24 20:13:06 2000
+@@ -0,0 +1 @@
++timestamp
+diff -Nru bzip2-1.0.1/unzcrash.c bzip2-1.0.1.new/unzcrash.c
+--- bzip2-1.0.1/unzcrash.c Sat Jun 24 20:13:27 2000
++++ bzip2-1.0.1.new/unzcrash.c Sat Jun 24 20:13:06 2000
+@@ -13,6 +13,12 @@
+ many hours.
+ */
+
++#ifdef HAVE_CONFIG_H
++#include <config.h>
++#endif
++
++
++
+ #include <stdio.h>
+ #include <assert.h>
+ #include "bzlib.h"
projects / packages / bzip2.git / commitdiff