aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--CHANGES88
-rw-r--r--LICENSE4
-rw-r--r--Makefile81
-rw-r--r--Makefile-libbz2_so15
-rw-r--r--README89
-rw-r--r--README.COMPILATION.PROBLEMS4
-rw-r--r--blocksort.c11
-rw-r--r--bzdiff76
-rw-r--r--bzdiff.147
-rw-r--r--bzgrep71
-rw-r--r--bzgrep.156
-rw-r--r--bzip2.156
-rw-r--r--bzip2.1.preformatted226
-rw-r--r--bzip2.c533
-rw-r--r--bzip2.txt134
-rw-r--r--bzip2recover.c161
-rw-r--r--bzlib.c35
-rw-r--r--bzlib.h6
-rw-r--r--bzlib_private.h11
-rw-r--r--bzmore61
-rw-r--r--bzmore.1152
-rw-r--r--compress.c10
-rw-r--r--crctable.c2
-rw-r--r--decompress.c14
-rw-r--r--dlltest.c6
-rw-r--r--huffman.c2
-rw-r--r--makefile.msc2
-rw-r--r--manual.texi114
-rw-r--r--mk251.c16
-rw-r--r--randtable.c2
-rw-r--r--words34
31 files changed, 1464 insertions, 625 deletions
diff --git a/CHANGES b/CHANGES
index ecaf417..d984395 100644
--- a/CHANGES
+++ b/CHANGES
@@ -134,7 +134,7 @@ Several minor bugfixes and enhancements:
134 134
135* Advance the version number to 1.0, so as to counteract the 135* Advance the version number to 1.0, so as to counteract the
136 (false-in-this-case) impression some people have that programs 136 (false-in-this-case) impression some people have that programs
137 with version numbers less than 1.0 are in someway, experimental, 137 with version numbers less than 1.0 are in some way, experimental,
138 pre-release versions. 138 pre-release versions.
139 139
140* Create an initial Makefile-libbz2_so to build a shared library. 140* Create an initial Makefile-libbz2_so to build a shared library.
@@ -165,3 +165,89 @@ There are no functionality changes or bug fixes relative to version
1651.0.0. This is just a documentation update + a fix for minor Win32 1651.0.0. This is just a documentation update + a fix for minor Win32
166build problems. For almost everyone, upgrading from 1.0.0 to 1.0.1 is 166build problems. For almost everyone, upgrading from 1.0.0 to 1.0.1 is
167utterly pointless. Don't bother. 167utterly pointless. Don't bother.
168
169
1701.0.2
171~~~~~
172A bug fix release, addressing various minor issues which have appeared
173in the 18 or so months since 1.0.1 was released. Most of the fixes
174are to do with file-handling or documentation bugs. To the best of my
175knowledge, there have been no data-loss-causing bugs reported in the
176compression/decompression engine of 1.0.0 or 1.0.1.
177
178Note that this release does not improve the rather crude build system
179for Unix platforms. The general plan here is to autoconfiscate/
180libtoolise 1.0.2 soon after release, and release the result as 1.1.0
181or perhaps 1.2.0. That, however, is still just a plan at this point.
182
183Here are the changes in 1.0.2. Bug-reporters and/or patch-senders in
184parentheses.
185
186* Fix an infinite segfault loop in 1.0.1 when a directory is
187 encountered in -f (force) mode.
188 (Trond Eivind Glomsrod, Nicholas Nethercote, Volker Schmidt)
189
190* Avoid double fclose() of output file on certain I/O error paths.
191 (Solar Designer)
192
193* Don't fail with internal error 1007 when fed a long stream (> 48MB)
194 of byte 251. Also print useful message suggesting that 1007s may be
195 caused by bad memory.
196 (noticed by Juan Pedro Vallejo, fixed by me)
197
198* Fix uninitialised variable silly bug in demo prog dlltest.c.
199 (Jorj Bauer)
200
201* Remove 512-MB limitation on recovered file size for bzip2recover
202 on selected platforms which support 64-bit ints. At the moment
203 all GCC supported platforms, and Win32.
204 (me, Alson van der Meulen)
205
206* Hard-code header byte values, to give correct operation on platforms
207 using EBCDIC as their native character set (IBM's OS/390).
208 (Leland Lucius)
209
210* Copy file access times correctly.
211 (Marty Leisner)
212
213* Add distclean and check targets to Makefile.
214 (Michael Carmack)
215
216* Parameterise use of ar and ranlib in Makefile. Also add $(LDFLAGS).
217 (Rich Ireland, Bo Thorsen)
218
219* Pass -p (create parent dirs as needed) to mkdir during make install.
220 (Jeremy Fusco)
221
222* Dereference symlinks when copying file permissions in -f mode.
223 (Volker Schmidt)
224
225* Majorly simplify implementation of uInt64_qrm10.
226 (Bo Lindbergh)
227
228* Check the input file still exists before deleting the output one,
229 when aborting in cleanUpAndFail().
230 (Joerg Prante, Robert Linden, Matthias Krings)
231
232Also a bunch of patches courtesy of Philippe Troin, the Debian maintainer
233of bzip2:
234
235* Wrapper scripts (with manpages): bzdiff, bzgrep, bzmore.
236
237* Spelling changes and minor enhancements in bzip2.1.
238
239* Avoid race condition between creating the output file and setting its
240 interim permissions safely, by using fopen_output_safely().
241 No changes to bzip2recover since there is no issue with file
242 permissions there.
243
244* do not print senseless report with -v when compressing an empty
245 file.
246
247* bzcat -f works on non-bzip2 files.
248
249* do not try to escape shell meta-characters on unix (the shell takes
250 care of these).
251
252* added --fast and --best aliases for -1 -9 for gzip compatibility.
253
diff --git a/LICENSE b/LICENSE
index 88fa6d8..9d4fa43 100644
--- a/LICENSE
+++ b/LICENSE
@@ -1,6 +1,6 @@
1 1
2This program, "bzip2" and associated library "libbzip2", are 2This program, "bzip2" and associated library "libbzip2", are
3copyright (C) 1996-2000 Julian R Seward. All rights reserved. 3copyright (C) 1996-2002 Julian R Seward. All rights reserved.
4 4
5Redistribution and use in source and binary forms, with or without 5Redistribution and use in source and binary forms, with or without
6modification, are permitted provided that the following conditions 6modification, are permitted provided that the following conditions
@@ -35,5 +35,5 @@ SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
35 35
36Julian Seward, Cambridge, UK. 36Julian Seward, Cambridge, UK.
37jseward@acm.org 37jseward@acm.org
38bzip2/libbzip2 version 1.0 of 21 March 2000 38bzip2/libbzip2 version 1.0.2 of 30 December 2001
39 39
diff --git a/Makefile b/Makefile
index ab17f49..8305235 100644
--- a/Makefile
+++ b/Makefile
@@ -1,9 +1,20 @@
1 1
2SHELL=/bin/sh 2SHELL=/bin/sh
3
4# To assist in cross-compiling
3CC=gcc 5CC=gcc
6AR=ar
7RANLIB=ranlib
8LDFLAGS=
9
10# Suitably paranoid flags to avoid bugs in gcc-2.7
4BIGFILES=-D_FILE_OFFSET_BITS=64 11BIGFILES=-D_FILE_OFFSET_BITS=64
5CFLAGS=-Wall -Winline -O2 -fomit-frame-pointer -fno-strength-reduce $(BIGFILES) 12CFLAGS=-Wall -Winline -O2 -fomit-frame-pointer -fno-strength-reduce $(BIGFILES)
6 13
14# Where you want it installed when you do 'make install'
15PREFIX=/usr
16
17
7OBJS= blocksort.o \ 18OBJS= blocksort.o \
8 huffman.o \ 19 huffman.o \
9 crctable.o \ 20 crctable.o \
@@ -15,20 +26,21 @@ OBJS= blocksort.o \
15all: libbz2.a bzip2 bzip2recover test 26all: libbz2.a bzip2 bzip2recover test
16 27
17bzip2: libbz2.a bzip2.o 28bzip2: libbz2.a bzip2.o
18 $(CC) $(CFLAGS) -o bzip2 bzip2.o -L. -lbz2 29 $(CC) $(CFLAGS) $(LDFLAGS) -o bzip2 bzip2.o -L. -lbz2
19 30
20bzip2recover: bzip2recover.o 31bzip2recover: bzip2recover.o
21 $(CC) $(CFLAGS) -o bzip2recover bzip2recover.o 32 $(CC) $(CFLAGS) $(LDFLAGS) -o bzip2recover bzip2recover.o
22 33
23libbz2.a: $(OBJS) 34libbz2.a: $(OBJS)
24 rm -f libbz2.a 35 rm -f libbz2.a
25 ar cq libbz2.a $(OBJS) 36 $(AR) cq libbz2.a $(OBJS)
26 @if ( test -f /usr/bin/ranlib -o -f /bin/ranlib -o \ 37 @if ( test -f $(RANLIB) -o -f /usr/bin/ranlib -o \
27 -f /usr/ccs/bin/ranlib ) ; then \ 38 -f /bin/ranlib -o -f /usr/ccs/bin/ranlib ) ; then \
28 echo ranlib libbz2.a ; \ 39 echo $(RANLIB) libbz2.a ; \
29 ranlib libbz2.a ; \ 40 $(RANLIB) libbz2.a ; \
30 fi 41 fi
31 42
43check: test
32test: bzip2 44test: bzip2
33 @cat words1 45 @cat words1
34 ./bzip2 -1 < sample1.ref > sample1.rb2 46 ./bzip2 -1 < sample1.ref > sample1.rb2
@@ -45,14 +57,12 @@ test: bzip2
45 cmp sample3.tst sample3.ref 57 cmp sample3.tst sample3.ref
46 @cat words3 58 @cat words3
47 59
48PREFIX=/usr
49
50install: bzip2 bzip2recover 60install: bzip2 bzip2recover
51 if ( test ! -d $(PREFIX)/bin ) ; then mkdir $(PREFIX)/bin ; fi 61 if ( test ! -d $(PREFIX)/bin ) ; then mkdir -p $(PREFIX)/bin ; fi
52 if ( test ! -d $(PREFIX)/lib ) ; then mkdir $(PREFIX)/lib ; fi 62 if ( test ! -d $(PREFIX)/lib ) ; then mkdir -p $(PREFIX)/lib ; fi
53 if ( test ! -d $(PREFIX)/man ) ; then mkdir $(PREFIX)/man ; fi 63 if ( test ! -d $(PREFIX)/man ) ; then mkdir -p $(PREFIX)/man ; fi
54 if ( test ! -d $(PREFIX)/man/man1 ) ; then mkdir $(PREFIX)/man/man1 ; fi 64 if ( test ! -d $(PREFIX)/man/man1 ) ; then mkdir -p $(PREFIX)/man/man1 ; fi
55 if ( test ! -d $(PREFIX)/include ) ; then mkdir $(PREFIX)/include ; fi 65 if ( test ! -d $(PREFIX)/include ) ; then mkdir -p $(PREFIX)/include ; fi
56 cp -f bzip2 $(PREFIX)/bin/bzip2 66 cp -f bzip2 $(PREFIX)/bin/bzip2
57 cp -f bzip2 $(PREFIX)/bin/bunzip2 67 cp -f bzip2 $(PREFIX)/bin/bunzip2
58 cp -f bzip2 $(PREFIX)/bin/bzcat 68 cp -f bzip2 $(PREFIX)/bin/bzcat
@@ -67,7 +77,26 @@ install: bzip2 bzip2recover
67 chmod a+r $(PREFIX)/include/bzlib.h 77 chmod a+r $(PREFIX)/include/bzlib.h
68 cp -f libbz2.a $(PREFIX)/lib 78 cp -f libbz2.a $(PREFIX)/lib
69 chmod a+r $(PREFIX)/lib/libbz2.a 79 chmod a+r $(PREFIX)/lib/libbz2.a
80 cp -f bzgrep $(PREFIX)/bin/bzgrep
81 ln $(PREFIX)/bin/bzgrep $(PREFIX)/bin/bzegrep
82 ln $(PREFIX)/bin/bzgrep $(PREFIX)/bin/bzfgrep
83 chmod a+x $(PREFIX)/bin/bzgrep
84 cp -f bzmore $(PREFIX)/bin/bzmore
85 ln $(PREFIX)/bin/bzmore $(PREFIX)/bin/bzless
86 chmod a+x $(PREFIX)/bin/bzmore
87 cp -f bzdiff $(PREFIX)/bin/bzdiff
88 ln $(PREFIX)/bin/bzdiff $(PREFIX)/bin/bzcmp
89 chmod a+x $(PREFIX)/bin/bzdiff
90 cp -f bzgrep.1 bzmore.1 bzdiff.1 $(PREFIX)/man/man1
91 chmod a+r $(PREFIX)/man/man1/bzgrep.1
92 chmod a+r $(PREFIX)/man/man1/bzmore.1
93 chmod a+r $(PREFIX)/man/man1/bzdiff.1
94 echo ".so man1/bzgrep.1" > $(PREFIX)/man/man1/bzegrep.1
95 echo ".so man1/bzgrep.1" > $(PREFIX)/man/man1/bzfgrep.1
96 echo ".so man1/bzmore.1" > $(PREFIX)/man/man1/bzless.1
97 echo ".so man1/bzdiff.1" > $(PREFIX)/man/man1/bzcmp.1
70 98
99distclean: clean
71clean: 100clean:
72 rm -f *.o libbz2.a bzip2 bzip2recover \ 101 rm -f *.o libbz2.a bzip2 bzip2recover \
73 sample1.rb2 sample2.rb2 sample3.rb2 \ 102 sample1.rb2 sample2.rb2 sample3.rb2 \
@@ -93,7 +122,7 @@ bzip2.o: bzip2.c
93bzip2recover.o: bzip2recover.c 122bzip2recover.o: bzip2recover.c
94 $(CC) $(CFLAGS) -c bzip2recover.c 123 $(CC) $(CFLAGS) -c bzip2recover.c
95 124
96DISTNAME=bzip2-1.0.1 125DISTNAME=bzip2-1.0.2
97tarfile: 126tarfile:
98 rm -f $(DISTNAME) 127 rm -f $(DISTNAME)
99 ln -sf . $(DISTNAME) 128 ln -sf . $(DISTNAME)
@@ -112,6 +141,7 @@ tarfile:
112 $(DISTNAME)/Makefile \ 141 $(DISTNAME)/Makefile \
113 $(DISTNAME)/manual.texi \ 142 $(DISTNAME)/manual.texi \
114 $(DISTNAME)/manual.ps \ 143 $(DISTNAME)/manual.ps \
144 $(DISTNAME)/manual.pdf \
115 $(DISTNAME)/LICENSE \ 145 $(DISTNAME)/LICENSE \
116 $(DISTNAME)/bzip2.1 \ 146 $(DISTNAME)/bzip2.1 \
117 $(DISTNAME)/bzip2.1.preformatted \ 147 $(DISTNAME)/bzip2.1.preformatted \
@@ -138,4 +168,25 @@ tarfile:
138 $(DISTNAME)/Y2K_INFO \ 168 $(DISTNAME)/Y2K_INFO \
139 $(DISTNAME)/unzcrash.c \ 169 $(DISTNAME)/unzcrash.c \
140 $(DISTNAME)/spewG.c \ 170 $(DISTNAME)/spewG.c \
171 $(DISTNAME)/mk251.c \
172 $(DISTNAME)/bzdiff \
173 $(DISTNAME)/bzdiff.1 \
174 $(DISTNAME)/bzmore \
175 $(DISTNAME)/bzmore.1 \
176 $(DISTNAME)/bzgrep \
177 $(DISTNAME)/bzgrep.1 \
141 $(DISTNAME)/Makefile-libbz2_so 178 $(DISTNAME)/Makefile-libbz2_so
179 gzip -v $(DISTNAME).tar
180
181# For rebuilding the manual from sources on my RedHat 7.2 box
182manual: manual.ps manual.pdf manual.html
183
184manual.ps: manual.texi
185 tex manual.texi
186 dvips -o manual.ps manual.dvi
187
188manual.pdf: manual.ps
189 ps2pdf manual.ps
190
191manual.html: manual.texi
192 texi2html -split_chapter manual.texi
diff --git a/Makefile-libbz2_so b/Makefile-libbz2_so
index a347c50..4986fe2 100644
--- a/Makefile-libbz2_so
+++ b/Makefile-libbz2_so
@@ -1,8 +1,9 @@
1 1
2# This Makefile builds a shared version of the library, 2# This Makefile builds a shared version of the library,
3# libbz2.so.1.0.1, with soname libbz2.so.1.0, 3# libbz2.so.1.0.2, with soname libbz2.so.1.0,
4# at least on x86-Linux (RedHat 5.2), 4# at least on x86-Linux (RedHat 7.2),
5# with gcc-2.7.2.3. Please see the README file for some 5# with gcc-2.96 20000731 (Red Hat Linux 7.1 2.96-98).
6# Please see the README file for some
6# important info about building the library like this. 7# important info about building the library like this.
7 8
8SHELL=/bin/sh 9SHELL=/bin/sh
@@ -19,13 +20,13 @@ OBJS= blocksort.o \
19 bzlib.o 20 bzlib.o
20 21
21all: $(OBJS) 22all: $(OBJS)
22 $(CC) -shared -Wl,-soname -Wl,libbz2.so.1.0 -o libbz2.so.1.0.1 $(OBJS) 23 $(CC) -shared -Wl,-soname -Wl,libbz2.so.1.0 -o libbz2.so.1.0.2 $(OBJS)
23 $(CC) $(CFLAGS) -o bzip2-shared bzip2.c libbz2.so.1.0.1 24 $(CC) $(CFLAGS) -o bzip2-shared bzip2.c libbz2.so.1.0.2
24 rm -f libbz2.so.1.0 25 rm -f libbz2.so.1.0
25 ln -s libbz2.so.1.0.1 libbz2.so.1.0 26 ln -s libbz2.so.1.0.2 libbz2.so.1.0
26 27
27clean: 28clean:
28 rm -f $(OBJS) bzip2.o libbz2.so.1.0.1 libbz2.so.1.0 bzip2-shared 29 rm -f $(OBJS) bzip2.o libbz2.so.1.0.2 libbz2.so.1.0 bzip2-shared
29 30
30blocksort.o: blocksort.c 31blocksort.o: blocksort.c
31 $(CC) $(CFLAGS) -c blocksort.c 32 $(CC) $(CFLAGS) -c blocksort.c
diff --git a/README b/README
index 22945a2..07505d8 100644
--- a/README
+++ b/README
@@ -1,15 +1,15 @@
1 1
2This is the README for bzip2, a block-sorting file compressor, version 2This is the README for bzip2, a block-sorting file compressor, version
31.0. This version is fully compatible with the previous public 31.0.2. This version is fully compatible with the previous public
4releases, bzip2-0.1pl2, bzip2-0.9.0 and bzip2-0.9.5. 4releases, versions 0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1.
5 5
6bzip2-1.0 is distributed under a BSD-style license. For details, 6bzip2-1.0.2 is distributed under a BSD-style license. For details,
7see the file LICENSE. 7see the file LICENSE.
8 8
9Complete documentation is available in Postscript form (manual.ps) or 9Complete documentation is available in Postscript form (manual.ps),
10html (manual_toc.html). A plain-text version of the manual page is 10PDF (manual.pdf, amazingly enough) or html (manual_toc.html). A
11available as bzip2.txt. A statement about Y2K issues is now included 11plain-text version of the manual page is available as bzip2.txt.
12in the file Y2K_INFO. 12A statement about Y2K issues is now included in the file Y2K_INFO.
13 13
14 14
15HOW TO BUILD -- UNIX 15HOW TO BUILD -- UNIX
@@ -33,34 +33,41 @@ not actually execute them.
33HOW TO BUILD -- UNIX, shared library libbz2.so. 33HOW TO BUILD -- UNIX, shared library libbz2.so.
34 34
35Do 'make -f Makefile-libbz2_so'. This Makefile seems to work for 35Do 'make -f Makefile-libbz2_so'. This Makefile seems to work for
36Linux-ELF (RedHat 5.2 on an x86 box), with gcc. I make no claims 36Linux-ELF (RedHat 7.2 on an x86 box), with gcc. I make no claims
37that it works for any other platform, though I suspect it probably 37that it works for any other platform, though I suspect it probably
38will work for most platforms employing both ELF and gcc. 38will work for most platforms employing both ELF and gcc.
39 39
40bzip2-shared, a client of the shared library, is also build, but 40bzip2-shared, a client of the shared library, is also built, but not
41not self-tested. So I suggest you also build using the normal 41self-tested. So I suggest you also build using the normal Makefile,
42Makefile, since that conducts a self-test. 42since that conducts a self-test. A second reason to prefer the
43version statically linked to the library is that, on x86 platforms,
44building shared objects makes a valuable register (%ebx) unavailable
45to gcc, resulting in a slowdown of 10%-20%, at least for bzip2.
43 46
44Important note for people upgrading .so's from 0.9.0/0.9.5 to 47Important note for people upgrading .so's from 0.9.0/0.9.5 to version
45version 1.0. All the functions in the library have been renamed, 481.0.X. All the functions in the library have been renamed, from (eg)
46from (eg) bzCompress to BZ2_bzCompress, to avoid namespace pollution. 49bzCompress to BZ2_bzCompress, to avoid namespace pollution.
47Unfortunately this means that the libbz2.so created by 50Unfortunately this means that the libbz2.so created by
48Makefile-libbz2_so will not work with any program which used an 51Makefile-libbz2_so will not work with any program which used an older
49older version of the library. Sorry. I do encourage library 52version of the library. Sorry. I do encourage library clients to
50clients to make the effort to upgrade to use version 1.0, since 53make the effort to upgrade to use version 1.0, since it is both faster
51it is both faster and more robust than previous versions. 54and more robust than previous versions.
52 55
53 56
54HOW TO BUILD -- Windows 95, NT, DOS, Mac, etc. 57HOW TO BUILD -- Windows 95, NT, DOS, Mac, etc.
55 58
56It's difficult for me to support compilation on all these platforms. 59It's difficult for me to support compilation on all these platforms.
57My approach is to collect binaries for these platforms, and put them 60My approach is to collect binaries for these platforms, and put them
58on the master web page (http://sourceware.cygnus.com/bzip2). Look 61on the master web page (http://sources.redhat.com/bzip2). Look there.
59there. However (FWIW), bzip2-1.0 is very standard ANSI C and should 62However (FWIW), bzip2-1.0.X is very standard ANSI C and should compile
60compile unmodified with MS Visual C. For Win32, there is one 63unmodified with MS Visual C. If you have difficulties building, you
61important caveat: in bzip2.c, you must set BZ_UNIX to 0 and 64might want to read README.COMPILATION.PROBLEMS.
62BZ_LCCWIN32 to 1 before building. If you have difficulties building, 65
63you might want to read README.COMPILATION.PROBLEMS. 66At least using MS Visual C++ 6, you can build from the unmodified
67sources by issuing, in a command shell:
68 nmake -f makefile.msc
69(you may need to first run the MSVC-provided script VCVARS32.BAT
70 so as to set up paths to the MSVC tools correctly).
64 71
65 72
66VALIDATION 73VALIDATION
@@ -138,29 +145,37 @@ WHAT'S NEW IN 0.9.5 ?
138 * Many small improvements in file and flag handling. 145 * Many small improvements in file and flag handling.
139 * A Y2K statement. 146 * A Y2K statement.
140 147
141WHAT'S NEW IN 1.0 148WHAT'S NEW IN 1.0.0 ?
142 149
143 See the CHANGES file. 150 See the CHANGES file.
144 151
152WHAT'S NEW IN 1.0.2 ?
153
154 See the CHANGES file.
155
156
145I hope you find bzip2 useful. Feel free to contact me at 157I hope you find bzip2 useful. Feel free to contact me at
146 jseward@acm.org 158 jseward@acm.org
147if you have any suggestions or queries. Many people mailed me with 159if you have any suggestions or queries. Many people mailed me with
148comments, suggestions and patches after the releases of bzip-0.15, 160comments, suggestions and patches after the releases of bzip-0.15,
149bzip-0.21, bzip2-0.1pl2 and bzip2-0.9.0, and the changes in bzip2 are 161bzip-0.21, and bzip2 versions 0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1,
150largely a result of this feedback. I thank you for your comments. 162and the changes in bzip2 are largely a result of this feedback.
163I thank you for your comments.
151 164
152At least for the time being, bzip2's "home" is (or can be reached via) 165At least for the time being, bzip2's "home" is (or can be reached via)
153http://www.muraroa.demon.co.uk. 166http://sources.redhat.com/bzip2.
154 167
155Julian Seward 168Julian Seward
156jseward@acm.org 169jseward@acm.org
157 170
158Cambridge, UK 171Cambridge, UK (and what a great town this is!)
15918 July 1996 (version 0.15) 172
16025 August 1996 (version 0.21) 17318 July 1996 (version 0.15)
161 7 August 1997 (bzip2, version 0.1) 17425 August 1996 (version 0.21)
16229 August 1997 (bzip2, version 0.1pl2) 175 7 August 1997 (bzip2, version 0.1)
16323 August 1998 (bzip2, version 0.9.0) 17629 August 1997 (bzip2, version 0.1pl2)
164 8 June 1999 (bzip2, version 0.9.5) 17723 August 1998 (bzip2, version 0.9.0)
165 4 Sept 1999 (bzip2, version 0.9.5d) 178 8 June 1999 (bzip2, version 0.9.5)
166 5 May 2000 (bzip2, version 1.0pre8) 179 4 Sept 1999 (bzip2, version 0.9.5d)
180 5 May 2000 (bzip2, version 1.0pre8)
18130 December 2001 (bzip2, version 1.0.2pre1) \ No newline at end of file
diff --git a/README.COMPILATION.PROBLEMS b/README.COMPILATION.PROBLEMS
index d621ad5..bd1822d 100644
--- a/README.COMPILATION.PROBLEMS
+++ b/README.COMPILATION.PROBLEMS
@@ -117,11 +117,11 @@ Known problems as of 1.0pre8:
117 All that said: you might be able to get somewhere 117 All that said: you might be able to get somewhere
118 by finding the line in Makefile-libbz2_so which says 118 by finding the line in Makefile-libbz2_so which says
119 119
120 $(CC) -shared -Wl,-soname -Wl,libbz2.so.1.0 -o libbz2.so.1.0.1 $(OBJS) 120 $(CC) -shared -Wl,-soname -Wl,libbz2.so.1.0 -o libbz2.so.1.0.2 $(OBJS)
121 121
122 and replacing with 122 and replacing with
123 123
124 ($CC) -G -shared -o libbz2.so.1.0.1 -h libbz2.so.1.0 $(OBJS) 124 $(CC) -G -shared -o libbz2.so.1.0.2 -h libbz2.so.1.0 $(OBJS)
125 125
126 If gcc objects to the combination -fpic -fPIC, get rid of 126 If gcc objects to the combination -fpic -fPIC, get rid of
127 the second one, leaving just "-fpic". 127 the second one, leaving just "-fpic".
diff --git a/blocksort.c b/blocksort.c
index ec42672..aba3efc 100644
--- a/blocksort.c
+++ b/blocksort.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
@@ -981,7 +981,14 @@ void mainSort ( UInt32* ptr,
981 } 981 }
982 } 982 }
983 983
984 AssertH ( copyStart[ss]-1 == copyEnd[ss], 1007 ); 984 AssertH ( (copyStart[ss]-1 == copyEnd[ss])
985 ||
986 /* Extremely rare case missing in bzip2-1.0.0 and 1.0.1.
987 Necessity for this case is demonstrated by compressing
988 a sequence of approximately 48.5 million of character
989 251; 1.0.0/1.0.1 will then die here. */
990 (copyStart[ss] == 0 && copyEnd[ss] == nblock-1),
991 1007 )
985 992
986 for (j = 0; j <= 255; j++) ftab[(j << 8) + ss] |= SETMASK; 993 for (j = 0; j <= 255; j++) ftab[(j << 8) + ss] |= SETMASK;
987 994
diff --git a/bzdiff b/bzdiff
new file mode 100644
index 0000000..3c2eb85
--- /dev/null
+++ b/bzdiff
@@ -0,0 +1,76 @@
1#!/bin/sh
2# sh is buggy on RS/6000 AIX 3.2. Replace above line with #!/bin/ksh
3
4# Bzcmp/diff wrapped for bzip2,
5# adapted from zdiff by Philippe Troin <phil@fifi.org> for Debian GNU/Linux.
6
7# Bzcmp and bzdiff are used to invoke the cmp or the diff pro-
8# gram on compressed files. All options specified are passed
9# directly to cmp or diff. If only 1 file is specified, then
10# the files compared are file1 and an uncompressed file1.gz.
11# If two files are specified, then they are uncompressed (if
12# necessary) and fed to cmp or diff. The exit status from cmp
13# or diff is preserved.
14
15PATH="/usr/bin:$PATH"; export PATH
16prog=`echo $0 | sed 's|.*/||'`
17case "$prog" in
18 *cmp) comp=${CMP-cmp} ;;
19 *) comp=${DIFF-diff} ;;
20esac
21
22OPTIONS=
23FILES=
24for ARG
25do
26 case "$ARG" in
27 -*) OPTIONS="$OPTIONS $ARG";;
28 *) if test -f "$ARG"; then
29 FILES="$FILES $ARG"
30 else
31 echo "${prog}: $ARG not found or not a regular file"
32 exit 1
33 fi ;;
34 esac
35done
36if test -z "$FILES"; then
37 echo "Usage: $prog [${comp}_options] file [file]"
38 exit 1
39fi
40tmp=`tempfile -d /tmp -p bz` || {
41 echo 'cannot create a temporary file' >&2
42 exit 1
43}
44set $FILES
45if test $# -eq 1; then
46 FILE=`echo "$1" | sed 's/.bz2$//'`
47 bzip2 -cd "$FILE.bz2" | $comp $OPTIONS - "$FILE"
48 STAT="$?"
49
50elif test $# -eq 2; then
51 case "$1" in
52 *.bz2)
53 case "$2" in
54 *.bz2)
55 F=`echo "$2" | sed 's|.*/||;s|.bz2$||'`
56 bzip2 -cdfq "$2" > $tmp
57 bzip2 -cdfq "$1" | $comp $OPTIONS - $tmp
58 STAT="$?"
59 /bin/rm -f $tmp;;
60
61 *) bzip2 -cdfq "$1" | $comp $OPTIONS - "$2"
62 STAT="$?";;
63 esac;;
64 *) case "$2" in
65 *.bz2)
66 bzip2 -cdfq "$2" | $comp $OPTIONS "$1" -
67 STAT="$?";;
68 *) $comp $OPTIONS "$1" "$2"
69 STAT="$?";;
70 esac;;
71 esac
72 exit "$STAT"
73else
74 echo "Usage: $prog [${comp}_options] file [file]"
75 exit 1
76fi
diff --git a/bzdiff.1 b/bzdiff.1
new file mode 100644
index 0000000..adb7a8e
--- /dev/null
+++ b/bzdiff.1
@@ -0,0 +1,47 @@
1\"Shamelessly copied from zmore.1 by Philippe Troin <phil@fifi.org>
2\"for Debian GNU/Linux
3.TH BZDIFF 1
4.SH NAME
5bzcmp, bzdiff \- compare bzip2 compressed files
6.SH SYNOPSIS
7.B bzcmp
8[ cmp_options ] file1
9[ file2 ]
10.br
11.B bzdiff
12[ diff_options ] file1
13[ file2 ]
14.SH DESCRIPTION
15.I Bzcmp
16and
17.I bzdiff
18are used to invoke the
19.I cmp
20or the
21.I diff
22program on bzip2 compressed files. All options specified are passed
23directly to
24.I cmp
25or
26.IR diff "."
27If only 1 file is specified, then the files compared are
28.I file1
29and an uncompressed
30.IR file1 ".bz2."
31If two files are specified, then they are uncompressed if necessary and fed to
32.I cmp
33or
34.IR diff "."
35The exit status from
36.I cmp
37or
38.I diff
39is preserved.
40.SH "SEE ALSO"
41cmp(1), diff(1), bzmore(1), bzless(1), bzgrep(1), bzip2(1)
42.SH BUGS
43Messages from the
44.I cmp
45or
46.I diff
47programs refer to temporary filenames instead of those specified.
diff --git a/bzgrep b/bzgrep
new file mode 100644
index 0000000..dbfc00e
--- /dev/null
+++ b/bzgrep
@@ -0,0 +1,71 @@
1#!/bin/sh
2
3# Bzgrep wrapped for bzip2,
4# adapted from zgrep by Philippe Troin <phil@fifi.org> for Debian GNU/Linux.
5## zgrep notice:
6## zgrep -- a wrapper around a grep program that decompresses files as needed
7## Adapted from a version sent by Charles Levert <charles@comm.polymtl.ca>
8
9PATH="/usr/bin:$PATH"; export PATH
10
11prog=`echo $0 | sed 's|.*/||'`
12case "$prog" in
13 *egrep) grep=${EGREP-egrep} ;;
14 *fgrep) grep=${FGREP-fgrep} ;;
15 *) grep=${GREP-grep} ;;
16esac
17pat=""
18while test $# -ne 0; do
19 case "$1" in
20 -e | -f) opt="$opt $1"; shift; pat="$1"
21 if test "$grep" = grep; then # grep is buggy with -e on SVR4
22 grep=egrep
23 fi;;
24 -A | -B) opt="$opt $1 $2"; shift;;
25 -*) opt="$opt $1";;
26 *) if test -z "$pat"; then
27 pat="$1"
28 else
29 break;
30 fi;;
31 esac
32 shift
33done
34
35if test -z "$pat"; then
36 echo "grep through bzip2 files"
37 echo "usage: $prog [grep_options] pattern [files]"
38 exit 1
39fi
40
41list=0
42silent=0
43op=`echo "$opt" | sed -e 's/ //g' -e 's/-//g'`
44case "$op" in
45 *l*) list=1
46esac
47case "$op" in
48 *h*) silent=1
49esac
50
51if test $# -eq 0; then
52 bzip2 -cdfq | $grep $opt "$pat"
53 exit $?
54fi
55
56res=0
57for i do
58 if test -f "$i"; then :; else if test -f "$i.bz2"; then i="$i.bz2"; fi; fi
59 if test $list -eq 1; then
60 bzip2 -cdfq "$i" | $grep $opt "$pat" 2>&1 > /dev/null && echo $i
61 r=$?
62 elif test $# -eq 1 -o $silent -eq 1; then
63 bzip2 -cdfq "$i" | $grep $opt "$pat"
64 r=$?
65 else
66 bzip2 -cdfq "$i" | $grep $opt "$pat" | sed "s|^|${i}:|"
67 r=$?
68 fi
69 test "$r" -ne 0 && res="$r"
70done
71exit $res
diff --git a/bzgrep.1 b/bzgrep.1
new file mode 100644
index 0000000..930af8c
--- /dev/null
+++ b/bzgrep.1
@@ -0,0 +1,56 @@
1\"Shamelessly copied from zmore.1 by Philippe Troin <phil@fifi.org>
2\"for Debian GNU/Linux
3.TH BZGREP 1
4.SH NAME
5bzgrep, bzfgrep, bzegrep \- search possibly bzip2 compressed files for a regular expression
6.SH SYNOPSIS
7.B bzgrep
8[ grep_options ]
9.BI [\ -e\ ] " pattern"
10.IR filename ".\|.\|."
11.br
12.B bzegrep
13[ egrep_options ]
14.BI [\ -e\ ] " pattern"
15.IR filename ".\|.\|."
16.br
17.B bzfgrep
18[ fgrep_options ]
19.BI [\ -e\ ] " pattern"
20.IR filename ".\|.\|."
21.SH DESCRIPTION
22.IR Bzgrep
23is used to invoke the
24.I grep
25on bzip2-compressed files. All options specified are passed directly to
26.I grep.
27If no file is specified, then the standard input is decompressed
28if necessary and fed to grep.
29Otherwise the given files are uncompressed if necessary and fed to
30.I grep.
31.PP
32If
33.I bzgrep
34is invoked as
35.I bzegrep
36or
37.I bzfgrep
38then
39.I egrep
40or
41.I fgrep
42is used instead of
43.I grep.
44If the GREP environment variable is set,
45.I bzgrep
46uses it as the
47.I grep
48program to be invoked. For example:
49
50 for sh: GREP=fgrep bzgrep string files
51 for csh: (setenv GREP fgrep; bzgrep string files)
52.SH AUTHOR
53Charles Levert (charles@comm.polymtl.ca). Adapted to bzip2 by Philippe
54Troin <phil@fifi.org> for Debian GNU/Linux.
55.SH "SEE ALSO"
56grep(1), egrep(1), fgrep(1), bzdiff(1), bzmore(1), bzless(1), bzip2(1)
diff --git a/bzip2.1 b/bzip2.1
index 7de54a0..623435c 100644
--- a/bzip2.1
+++ b/bzip2.1
@@ -1,7 +1,7 @@
1.PU 1.PU
2.TH bzip2 1 2.TH bzip2 1
3.SH NAME 3.SH NAME
4bzip2, bunzip2 \- a block-sorting file compressor, v1.0 4bzip2, bunzip2 \- a block-sorting file compressor, v1.0.2
5.br 5.br
6bzcat \- decompresses files to stdout 6bzcat \- decompresses files to stdout
7.br 7.br
@@ -197,7 +197,7 @@ to decompress.
197.TP 197.TP
198.B \-z --compress 198.B \-z --compress
199The complement to \-d: forces compression, regardless of the 199The complement to \-d: forces compression, regardless of the
200invokation name. 200invocation name.
201.TP 201.TP
202.B \-t --test 202.B \-t --test
203Check integrity of the specified file(s), but don't decompress them. 203Check integrity of the specified file(s), but don't decompress them.
@@ -211,6 +211,10 @@ existing output files. Also forces
211.I bzip2 211.I bzip2
212to break hard links 212to break hard links
213to files, which it otherwise wouldn't do. 213to files, which it otherwise wouldn't do.
214
215bzip2 normally declines to decompress files which don't have the
216correct magic header bytes. If forced (-f), however, it will pass
217such files through unmodified. This is how GNU gzip behaves.
214.TP 218.TP
215.B \-k --keep 219.B \-k --keep
216Keep (don't delete) input files during compression 220Keep (don't delete) input files during compression
@@ -239,9 +243,13 @@ information which is primarily of interest for diagnostic purposes.
239.B \-L --license -V --version 243.B \-L --license -V --version
240Display the software version, license terms and conditions. 244Display the software version, license terms and conditions.
241.TP 245.TP
242.B \-1 to \-9 246.B \-1 (or \-\-fast) to \-9 (or \-\-best)
243Set the block size to 100 k, 200 k .. 900 k when compressing. Has no 247Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
244effect when decompressing. See MEMORY MANAGEMENT below. 248effect when decompressing. See MEMORY MANAGEMENT below.
249The \-\-fast and \-\-best aliases are primarily for GNU gzip
250compatibility. In particular, \-\-fast doesn't make things
251significantly faster.
252And \-\-best merely selects the default behaviour.
245.TP 253.TP
246.B \-- 254.B \--
247Treats all subsequent arguments as file names, even if they start 255Treats all subsequent arguments as file names, even if they start
@@ -352,11 +360,11 @@ undamaged.
352 360
353.I bzip2recover 361.I bzip2recover
354takes a single argument, the name of the damaged file, 362takes a single argument, the name of the damaged file,
355and writes a number of files "rec0001file.bz2", 363and writes a number of files "rec00001file.bz2",
356"rec0002file.bz2", etc, containing the extracted blocks. 364"rec00002file.bz2", etc, containing the extracted blocks.
357The output filenames are designed so that the use of 365The output filenames are designed so that the use of
358wildcards in subsequent processing -- for example, 366wildcards in subsequent processing -- for example,
359"bzip2 -dc rec*file.bz2 > recovered_data" -- lists the files in 367"bzip2 -dc rec*file.bz2 > recovered_data" -- processes the files in
360the correct order. 368the correct order.
361 369
362.I bzip2recover 370.I bzip2recover
@@ -397,27 +405,31 @@ I/O error messages are not as helpful as they could be.
397tries hard to detect I/O errors and exit cleanly, but the details of 405tries hard to detect I/O errors and exit cleanly, but the details of
398what the problem is sometimes seem rather misleading. 406what the problem is sometimes seem rather misleading.
399 407
400This manual page pertains to version 1.0 of 408This manual page pertains to version 1.0.2 of
401.I bzip2. 409.I bzip2.
402Compressed 410Compressed data created by this version is entirely forwards and
403data created by this version is entirely forwards and backwards 411backwards compatible with the previous public releases, versions
404compatible with the previous public releases, versions 0.1pl2, 0.9.0 4120.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1, but with the following
405and 0.9.5, 413exception: 0.9.0 and above can correctly decompress multiple
406but with the following exception: 0.9.0 and above can correctly 414concatenated compressed files. 0.1pl2 cannot do this; it will stop
407decompress multiple concatenated compressed files. 0.1pl2 cannot do 415after decompressing just the first file in the stream.
408this; it will stop after decompressing just the first file in the
409stream.
410 416
411.I bzip2recover 417.I bzip2recover
412uses 32-bit integers to represent bit positions in 418versions prior to this one, 1.0.2, used 32-bit integers to represent
413compressed files, so it cannot handle compressed files more than 512 419bit positions in compressed files, so it could not handle compressed
414megabytes long. This could easily be fixed. 420files more than 512 megabytes long. Version 1.0.2 and above uses
42164-bit ints on some platforms which support them (GNU supported
422targets, and Windows). To establish whether or not bzip2recover was
423built with such a limitation, run it without arguments. In any event
424you can build yourself an unlimited version if you can recompile it
425with MaybeUInt64 set to be an unsigned 64-bit integer.
426
427
415 428
416.SH AUTHOR 429.SH AUTHOR
417Julian Seward, jseward@acm.org. 430Julian Seward, jseward@acm.org.
418 431
419http://sourceware.cygnus.com/bzip2 432http://sources.redhat.com/bzip2
420http://www.muraroa.demon.co.uk
421 433
422The ideas embodied in 434The ideas embodied in
423.I bzip2 435.I bzip2
@@ -434,6 +446,8 @@ indebted for their help, support and advice. See the manual in the
434source distribution for pointers to sources of documentation. Christian 446source distribution for pointers to sources of documentation. Christian
435von Roques encouraged me to look for faster sorting algorithms, so as to 447von Roques encouraged me to look for faster sorting algorithms, so as to
436speed up compression. Bela Lubkin encouraged me to improve the 448speed up compression. Bela Lubkin encouraged me to improve the
437worst-case compression performance. Many people sent patches, helped 449worst-case compression performance.
450The bz* scripts are derived from those of GNU gzip.
451Many people sent patches, helped
438with portability problems, lent machines, gave advice and were generally 452with portability problems, lent machines, gave advice and were generally
439helpful. 453helpful.
diff --git a/bzip2.1.preformatted b/bzip2.1.preformatted
index 9f18339..0f20cb5 100644
--- a/bzip2.1.preformatted
+++ b/bzip2.1.preformatted
@@ -1,11 +1,9 @@
1
2
3
4bzip2(1) bzip2(1) 1bzip2(1) bzip2(1)
5 2
6 3
4
7NNAAMMEE 5NNAAMMEE
8 bzip2, bunzip2 - a block-sorting file compressor, v1.0 6 bzip2, bunzip2 - a block-sorting file compressor, v1.0.2
9 bzcat - decompresses files to stdout 7 bzcat - decompresses files to stdout
10 bzip2recover - recovers data from damaged bzip2 files 8 bzip2recover - recovers data from damaged bzip2 files
11 9
@@ -22,20 +20,20 @@ DDEESSCCRRIIPPTTIIOONN
22 sorting text compression algorithm, and Huffman coding. 20 sorting text compression algorithm, and Huffman coding.
23 Compression is generally considerably better than that 21 Compression is generally considerably better than that
24 achieved by more conventional LZ77/LZ78-based compressors, 22 achieved by more conventional LZ77/LZ78-based compressors,
25 and approaches the performance of the PPM family of sta- 23 and approaches the performance of the PPM family of sta­
26 tistical compressors. 24 tistical compressors.
27 25
28 The command-line options are deliberately very similar to 26 The command-line options are deliberately very similar to
29 those of _G_N_U _g_z_i_p_, but they are not identical. 27 those of _G_N_U _g_z_i_p_, but they are not identical.
30 28
31 _b_z_i_p_2 expects a list of file names to accompany the com- 29 _b_z_i_p_2 expects a list of file names to accompany the com­
32 mand-line flags. Each file is replaced by a compressed 30 mand-line flags. Each file is replaced by a compressed
33 version of itself, with the name "original_name.bz2". 31 version of itself, with the name "original_name.bz2".
34 Each compressed file has the same modification date, per- 32 Each compressed file has the same modification date, per­
35 missions, and, when possible, ownership as the correspond- 33 missions, and, when possible, ownership as the correspond­
36 ing original, so that these properties can be correctly 34 ing original, so that these properties can be correctly
37 restored at decompression time. File name handling is 35 restored at decompression time. File name handling is
38 naive in the sense that there is no mechanism for preserv- 36 naive in the sense that there is no mechanism for preserv­
39 ing original file names, permissions, ownerships or dates 37 ing original file names, permissions, ownerships or dates
40 in filesystems which lack these concepts, or have serious 38 in filesystems which lack these concepts, or have serious
41 file name length restrictions, such as MS-DOS. 39 file name length restrictions, such as MS-DOS.
@@ -58,18 +56,6 @@ DDEESSCCRRIIPPTTIIOONN
58 filename.bz2 becomes filename 56 filename.bz2 becomes filename
59 filename.bz becomes filename 57 filename.bz becomes filename
60 filename.tbz2 becomes filename.tar 58 filename.tbz2 becomes filename.tar
61
62
63
64 1
65
66
67
68
69
70bzip2(1) bzip2(1)
71
72
73 filename.tbz becomes filename.tar 59 filename.tbz becomes filename.tar
74 anyothername becomes anyothername.out 60 anyothername becomes anyothername.out
75 61
@@ -78,23 +64,23 @@ bzip2(1) bzip2(1)
78 guess the name of the original file, and uses the original 64 guess the name of the original file, and uses the original
79 name with _._o_u_t appended. 65 name with _._o_u_t appended.
80 66
81 As with compression, supplying no filenames causes decom- 67 As with compression, supplying no filenames causes decom­
82 pression from standard input to standard output. 68 pression from standard input to standard output.
83 69
84 _b_u_n_z_i_p_2 will correctly decompress a file which is the con- 70 _b_u_n_z_i_p_2 will correctly decompress a file which is the con­
85 catenation of two or more compressed files. The result is 71 catenation of two or more compressed files. The result is
86 the concatenation of the corresponding uncompressed files. 72 the concatenation of the corresponding uncompressed files.
87 Integrity testing (-t) of concatenated compressed files is 73 Integrity testing (-t) of concatenated compressed files is
88 also supported. 74 also supported.
89 75
90 You can also compress or decompress files to the standard 76 You can also compress or decompress files to the standard
91 output by giving the -c flag. Multiple files may be com- 77 output by giving the -c flag. Multiple files may be com­
92 pressed and decompressed like this. The resulting outputs 78 pressed and decompressed like this. The resulting outputs
93 are fed sequentially to stdout. Compression of multiple 79 are fed sequentially to stdout. Compression of multiple
94 files in this manner generates a stream containing multi- 80 files in this manner generates a stream containing multi­
95 ple compressed file representations. Such a stream can be 81 ple compressed file representations. Such a stream can be
96 decompressed correctly only by _b_z_i_p_2 version 0.9.0 or 82 decompressed correctly only by _b_z_i_p_2 version 0.9.0 or
97 later. Earlier versions of _b_z_i_p_2 will stop after decom- 83 later. Earlier versions of _b_z_i_p_2 will stop after decom­
98 pressing the first file in the stream. 84 pressing the first file in the stream.
99 85
100 _b_z_c_a_t (or _b_z_i_p_2 _-_d_c_) decompresses all specified files to 86 _b_z_c_a_t (or _b_z_i_p_2 _-_d_c_) decompresses all specified files to
@@ -115,7 +101,7 @@ bzip2(1) bzip2(1)
115 101
116 As a self-check for your protection, _b_z_i_p_2 uses 32-bit 102 As a self-check for your protection, _b_z_i_p_2 uses 32-bit
117 CRCs to make sure that the decompressed version of a file 103 CRCs to make sure that the decompressed version of a file
118 is identical to the original. This guards against corrup- 104 is identical to the original. This guards against corrup­
119 tion of the compressed data, and against undetected bugs 105 tion of the compressed data, and against undetected bugs
120 in _b_z_i_p_2 (hopefully very unlikely). The chances of data 106 in _b_z_i_p_2 (hopefully very unlikely). The chances of data
121 corruption going undetected is microscopic, about one 107 corruption going undetected is microscopic, about one
@@ -125,17 +111,6 @@ bzip2(1) bzip2(1)
125 you recover the original uncompressed data. You can use 111 you recover the original uncompressed data. You can use
126 _b_z_i_p_2_r_e_c_o_v_e_r to try to recover data from damaged files. 112 _b_z_i_p_2_r_e_c_o_v_e_r to try to recover data from damaged files.
127 113
128
129
130 2
131
132
133
134
135
136bzip2(1) bzip2(1)
137
138
139 Return values: 0 for a normal exit, 1 for environmental 114 Return values: 0 for a normal exit, 1 for environmental
140 problems (file not found, invalid flags, I/O errors, &c), 115 problems (file not found, invalid flags, I/O errors, &c),
141 2 to indicate a corrupt compressed file, 3 for an internal 116 2 to indicate a corrupt compressed file, 3 for an internal
@@ -154,8 +129,8 @@ OOPPTTIIOONNSS
154 and forces _b_z_i_p_2 to decompress. 129 and forces _b_z_i_p_2 to decompress.
155 130
156 --zz ----ccoommpprreessss 131 --zz ----ccoommpprreessss
157 The complement to -d: forces compression, regard- 132 The complement to -d: forces compression,
158 less of the invokation name. 133 regardless of the invocation name.
159 134
160 --tt ----tteesstt 135 --tt ----tteesstt
161 Check integrity of the specified file(s), but don't 136 Check integrity of the specified file(s), but don't
@@ -168,6 +143,11 @@ OOPPTTIIOONNSS
168 forces _b_z_i_p_2 to break hard links to files, which it 143 forces _b_z_i_p_2 to break hard links to files, which it
169 otherwise wouldn't do. 144 otherwise wouldn't do.
170 145
146 bzip2 normally declines to decompress files which
147 don't have the correct magic header bytes. If
148 forced (-f), however, it will pass such files
149 through unmodified. This is how GNU gzip behaves.
150
171 --kk ----kkeeeepp 151 --kk ----kkeeeepp
172 Keep (don't delete) input files during compression 152 Keep (don't delete) input files during compression
173 or decompression. 153 or decompression.
@@ -190,23 +170,11 @@ OOPPTTIIOONNSS
190 --qq ----qquuiieett 170 --qq ----qquuiieett
191 Suppress non-essential warning messages. Messages 171 Suppress non-essential warning messages. Messages
192 pertaining to I/O errors and other critical events 172 pertaining to I/O errors and other critical events
193
194
195
196 3
197
198
199
200
201
202bzip2(1) bzip2(1)
203
204
205 will not be suppressed. 173 will not be suppressed.
206 174
207 --vv ----vveerrbboossee 175 --vv ----vveerrbboossee
208 Verbose mode -- show the compression ratio for each 176 Verbose mode -- show the compression ratio for each
209 file processed. Further -v's increase the ver- 177 file processed. Further -v's increase the ver­
210 bosity level, spewing out lots of information which 178 bosity level, spewing out lots of information which
211 is primarily of interest for diagnostic purposes. 179 is primarily of interest for diagnostic purposes.
212 180
@@ -214,20 +182,24 @@ bzip2(1) bzip2(1)
214 Display the software version, license terms and 182 Display the software version, license terms and
215 conditions. 183 conditions.
216 184
217 --11 ttoo --99 185 --11 ((oorr ----ffaasstt)) ttoo --99 ((oorr ----bbeesstt))
218 Set the block size to 100 k, 200 k .. 900 k when 186 Set the block size to 100 k, 200 k .. 900 k when
219 compressing. Has no effect when decompressing. 187 compressing. Has no effect when decompressing.
220 See MEMORY MANAGEMENT below. 188 See MEMORY MANAGEMENT below. The --fast and --best
189 aliases are primarily for GNU gzip compatibility.
190 In particular, --fast doesn't make things signifi­
191 cantly faster. And --best merely selects the
192 default behaviour.
221 193
222 ---- Treats all subsequent arguments as file names, even 194 ---- Treats all subsequent arguments as file names, even
223 if they start with a dash. This is so you can han- 195 if they start with a dash. This is so you can han­
224 dle files with names beginning with a dash, for 196 dle files with names beginning with a dash, for
225 example: bzip2 -- -myfilename. 197 example: bzip2 -- -myfilename.
226 198
227 ----rreeppeettiittiivvee--ffaasstt ----rreeppeettiittiivvee--bbeesstt 199 ----rreeppeettiittiivvee--ffaasstt ----rreeppeettiittiivvee--bbeesstt
228 These flags are redundant in versions 0.9.5 and 200 These flags are redundant in versions 0.9.5 and
229 above. They provided some coarse control over the 201 above. They provided some coarse control over the
230 behaviour of the sorting algorithm in earlier ver- 202 behaviour of the sorting algorithm in earlier ver­
231 sions, which was sometimes useful. 0.9.5 and above 203 sions, which was sometimes useful. 0.9.5 and above
232 have an improved algorithm which renders these 204 have an improved algorithm which renders these
233 flags irrelevant. 205 flags irrelevant.
@@ -238,7 +210,7 @@ MMEEMMOORRYY MMAANNAAGGEEMMEENNTT
238 affects both the compression ratio achieved, and the 210 affects both the compression ratio achieved, and the
239 amount of memory needed for compression and decompression. 211 amount of memory needed for compression and decompression.
240 The flags -1 through -9 specify the block size to be 212 The flags -1 through -9 specify the block size to be
241 100,000 bytes through 900,000 bytes (the default) respec- 213 100,000 bytes through 900,000 bytes (the default) respec­
242 tively. At decompression time, the block size used for 214 tively. At decompression time, the block size used for
243 compression is read from the header of the compressed 215 compression is read from the header of the compressed
244 file, and _b_u_n_z_i_p_2 then allocates itself just enough memory 216 file, and _b_u_n_z_i_p_2 then allocates itself just enough memory
@@ -256,18 +228,6 @@ MMEEMMOORRYY MMAANNAAGGEEMMEENNTT
256 228
257 Larger block sizes give rapidly diminishing marginal 229 Larger block sizes give rapidly diminishing marginal
258 returns. Most of the compression comes from the first two 230 returns. Most of the compression comes from the first two
259
260
261
262 4
263
264
265
266
267
268bzip2(1) bzip2(1)
269
270
271 or three hundred k of block size, a fact worth bearing in 231 or three hundred k of block size, a fact worth bearing in
272 mind when using _b_z_i_p_2 on small machines. It is also 232 mind when using _b_z_i_p_2 on small machines. It is also
273 important to appreciate that the decompression memory 233 important to appreciate that the decompression memory
@@ -278,13 +238,13 @@ bzip2(1) bzip2(1)
278 _b_u_n_z_i_p_2 will require about 3700 kbytes to decompress. To 238 _b_u_n_z_i_p_2 will require about 3700 kbytes to decompress. To
279 support decompression of any file on a 4 megabyte machine, 239 support decompression of any file on a 4 megabyte machine,
280 _b_u_n_z_i_p_2 has an option to decompress using approximately 240 _b_u_n_z_i_p_2 has an option to decompress using approximately
281 half this amount of memory, about 2300 kbytes. Decompres- 241 half this amount of memory, about 2300 kbytes. Decompres­
282 sion speed is also halved, so you should use this option 242 sion speed is also halved, so you should use this option
283 only where necessary. The relevant flag is -s. 243 only where necessary. The relevant flag is -s.
284 244
285 In general, try and use the largest block size memory con- 245 In general, try and use the largest block size memory con­
286 straints allow, since that maximises the compression 246 straints allow, since that maximises the compression
287 achieved. Compression and decompression speed are virtu- 247 achieved. Compression and decompression speed are virtu­
288 ally unaffected by block size. 248 ally unaffected by block size.
289 249
290 Another significant point applies to files which fit in a 250 Another significant point applies to files which fit in a
@@ -300,11 +260,11 @@ bzip2(1) bzip2(1)
300 260
301 Here is a table which summarises the maximum memory usage 261 Here is a table which summarises the maximum memory usage
302 for different block sizes. Also recorded is the total 262 for different block sizes. Also recorded is the total
303 compressed size for 14 files of the Calgary Text Compres- 263 compressed size for 14 files of the Calgary Text Compres­
304 sion Corpus totalling 3,141,622 bytes. This column gives 264 sion Corpus totalling 3,141,622 bytes. This column gives
305 some feel for how compression varies with block size. 265 some feel for how compression varies with block size.
306 These figures tend to understate the advantage of larger 266 These figures tend to understate the advantage of larger
307 block sizes for larger files, since the Corpus is domi- 267 block sizes for larger files, since the Corpus is domi­
308 nated by smaller files. 268 nated by smaller files.
309 269
310 Compress Decompress Decompress Corpus 270 Compress Decompress Decompress Corpus
@@ -321,22 +281,9 @@ bzip2(1) bzip2(1)
321 -9 7600k 3700k 2350k 828642 281 -9 7600k 3700k 2350k 828642
322 282
323 283
324
325
326
327
328 5
329
330
331
332
333
334bzip2(1) bzip2(1)
335
336
337RREECCOOVVEERRIINNGG DDAATTAA FFRROOMM DDAAMMAAGGEEDD FFIILLEESS 284RREECCOOVVEERRIINNGG DDAATTAA FFRROOMM DDAAMMAAGGEEDD FFIILLEESS
338 _b_z_i_p_2 compresses files in blocks, usually 900kbytes long. 285 _b_z_i_p_2 compresses files in blocks, usually 900kbytes long.
339 Each block is handled independently. If a media or trans- 286 Each block is handled independently. If a media or trans­
340 mission error causes a multi-block .bz2 file to become 287 mission error causes a multi-block .bz2 file to become
341 damaged, it may be possible to recover data from the 288 damaged, it may be possible to recover data from the
342 undamaged blocks in the file. 289 undamaged blocks in the file.
@@ -353,19 +300,19 @@ RREECCOOVVEERRIINNGG DDAATTAA FFRROOMM DDAAMMAAGGEEDD F
353 the integrity of the resulting files, and decompress those 300 the integrity of the resulting files, and decompress those
354 which are undamaged. 301 which are undamaged.
355 302
356 _b_z_i_p_2_r_e_c_o_v_e_r takes a single argument, the name of the dam- 303 _b_z_i_p_2_r_e_c_o_v_e_r takes a single argument, the name of the dam­
357 aged file, and writes a number of files "rec0001file.bz2", 304 aged file, and writes a number of files
358 "rec0002file.bz2", etc, containing the extracted blocks. 305 "rec00001file.bz2", "rec00002file.bz2", etc, containing
359 The output filenames are designed so that the use of 306 the extracted blocks. The output filenames are
360 wildcards in subsequent processing -- for example, "bzip2 307 designed so that the use of wildcards in subsequent pro­
361 -dc rec*file.bz2 > recovered_data" -- lists the files in 308 cessing -- for example, "bzip2 -dc rec*file.bz2 > recov­
362 the correct order. 309 ered_data" -- processes the files in the correct order.
363 310
364 _b_z_i_p_2_r_e_c_o_v_e_r should be of most use dealing with large .bz2 311 _b_z_i_p_2_r_e_c_o_v_e_r should be of most use dealing with large .bz2
365 files, as these will contain many blocks. It is clearly 312 files, as these will contain many blocks. It is clearly
366 futile to use it on damaged single-block files, since a 313 futile to use it on damaged single-block files, since a
367 damaged block cannot be recovered. If you wish to min- 314 damaged block cannot be recovered. If you wish to min­
368 imise any potential data loss through media or transmis- 315 imise any potential data loss through media or transmis­
369 sion errors, you might consider compressing with a smaller 316 sion errors, you might consider compressing with a smaller
370 block size. 317 block size.
371 318
@@ -379,31 +326,19 @@ PPEERRFFOORRMMAANNCCEE NNOOTTEESS
379 better than previous versions in this respect. The ratio 326 better than previous versions in this respect. The ratio
380 between worst-case and average-case compression time is in 327 between worst-case and average-case compression time is in
381 the region of 10:1. For previous versions, this figure 328 the region of 10:1. For previous versions, this figure
382 was more like 100:1. You can use the -vvvv option to mon- 329 was more like 100:1. You can use the -vvvv option to mon­
383 itor progress in great detail, if you want. 330 itor progress in great detail, if you want.
384 331
385 Decompression speed is unaffected by these phenomena. 332 Decompression speed is unaffected by these phenomena.
386 333
387 _b_z_i_p_2 usually allocates several megabytes of memory to 334 _b_z_i_p_2 usually allocates several megabytes of memory to
388 operate in, and then charges all over it in a fairly ran- 335 operate in, and then charges all over it in a fairly ran­
389 dom fashion. This means that performance, both for com- 336 dom fashion. This means that performance, both for com­
390 pressing and decompressing, is largely determined by the 337 pressing and decompressing, is largely determined by the
391
392
393
394 6
395
396
397
398
399
400bzip2(1) bzip2(1)
401
402
403 speed at which your machine can service cache misses. 338 speed at which your machine can service cache misses.
404 Because of this, small changes to the code to reduce the 339 Because of this, small changes to the code to reduce the
405 miss rate have been observed to give disproportionately 340 miss rate have been observed to give disproportionately
406 large performance improvements. I imagine _b_z_i_p_2 will per- 341 large performance improvements. I imagine _b_z_i_p_2 will per­
407 form best on machines with very large caches. 342 form best on machines with very large caches.
408 343
409 344
@@ -413,50 +348,51 @@ CCAAVVEEAATTSS
413 but the details of what the problem is sometimes seem 348 but the details of what the problem is sometimes seem
414 rather misleading. 349 rather misleading.
415 350
416 This manual page pertains to version 1.0 of _b_z_i_p_2_. Com- 351 This manual page pertains to version 1.0.2 of _b_z_i_p_2_. Com­
417 pressed data created by this version is entirely forwards 352 pressed data created by this version is entirely forwards
418 and backwards compatible with the previous public 353 and backwards compatible with the previous public
419 releases, versions 0.1pl2, 0.9.0 and 0.9.5, but with the 354 releases, versions 0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1,
420 following exception: 0.9.0 and above can correctly decom- 355 but with the following exception: 0.9.0 and above can cor­
421 press multiple concatenated compressed files. 0.1pl2 can- 356 rectly decompress multiple concatenated compressed files.
422 not do this; it will stop after decompressing just the 357 0.1pl2 cannot do this; it will stop after decompressing
423 first file in the stream. 358 just the first file in the stream.
359
360 _b_z_i_p_2_r_e_c_o_v_e_r versions prior to this one, 1.0.2, used
361 32-bit integers to represent bit positions in compressed
362 files, so it could not handle compressed files more than
363 512 megabytes long. Version 1.0.2 and above uses 64-bit
364 ints on some platforms which support them (GNU supported
365 targets, and Windows). To establish whether or not
366 bzip2recover was built with such a limitation, run it
367 without arguments. In any event you can build yourself an
368 unlimited version if you can recompile it with MaybeUInt64
369 set to be an unsigned 64-bit integer.
370
424 371
425 _b_z_i_p_2_r_e_c_o_v_e_r uses 32-bit integers to represent bit posi-
426 tions in compressed files, so it cannot handle compressed
427 files more than 512 megabytes long. This could easily be
428 fixed.
429 372
430 373
431AAUUTTHHOORR 374AAUUTTHHOORR
432 Julian Seward, jseward@acm.org. 375 Julian Seward, jseward@acm.org.
433 376
434 http://sourceware.cygnus.com/bzip2 377 http://sources.redhat.com/bzip2
435 http://www.muraroa.demon.co.uk
436 378
437 The ideas embodied in _b_z_i_p_2 are due to (at least) the fol- 379 The ideas embodied in _b_z_i_p_2 are due to (at least) the fol­
438 lowing people: Michael Burrows and David Wheeler (for the 380 lowing people: Michael Burrows and David Wheeler (for the
439 block sorting transformation), David Wheeler (again, for 381 block sorting transformation), David Wheeler (again, for
440 the Huffman coder), Peter Fenwick (for the structured cod- 382 the Huffman coder), Peter Fenwick (for the structured cod­
441 ing model in the original _b_z_i_p_, and many refinements), and 383 ing model in the original _b_z_i_p_, and many refinements), and
442 Alistair Moffat, Radford Neal and Ian Witten (for the 384 Alistair Moffat, Radford Neal and Ian Witten (for the
443 arithmetic coder in the original _b_z_i_p_)_. I am much 385 arithmetic coder in the original _b_z_i_p_)_. I am much
444 indebted for their help, support and advice. See the man- 386 indebted for their help, support and advice. See the man­
445 ual in the source distribution for pointers to sources of 387 ual in the source distribution for pointers to sources of
446 documentation. Christian von Roques encouraged me to look 388 documentation. Christian von Roques encouraged me to look
447 for faster sorting algorithms, so as to speed up compres- 389 for faster sorting algorithms, so as to speed up compres­
448 sion. Bela Lubkin encouraged me to improve the worst-case 390 sion. Bela Lubkin encouraged me to improve the worst-case
449 compression performance. Many people sent patches, helped 391 compression performance. The bz* scripts are derived from
450 with portability problems, lent machines, gave advice and 392 those of GNU gzip. Many people sent patches, helped with
451 were generally helpful. 393 portability problems, lent machines, gave advice and were
452 394 generally helpful.
453
454
455
456
457
458
459 395
460 7
461 396
462 397
398 bzip2(1)
diff --git a/bzip2.c b/bzip2.c
index 56adfdc..807f420 100644
--- a/bzip2.c
+++ b/bzip2.c
@@ -7,7 +7,7 @@
7 This file is a part of bzip2 and/or libbzip2, a program and 7 This file is a part of bzip2 and/or libbzip2, a program and
8 library for lossless, block-sorting data compression. 8 library for lossless, block-sorting data compression.
9 9
10 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 10 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
11 11
12 Redistribution and use in source and binary forms, with or without 12 Redistribution and use in source and binary forms, with or without
13 modification, are permitted provided that the following conditions 13 modification, are permitted provided that the following conditions
@@ -113,13 +113,16 @@
113/*-- 113/*--
114 Generic 32-bit Unix. 114 Generic 32-bit Unix.
115 Also works on 64-bit Unix boxes. 115 Also works on 64-bit Unix boxes.
116 This is the default.
116--*/ 117--*/
117#define BZ_UNIX 1 118#define BZ_UNIX 1
118 119
119/*-- 120/*--
120 Win32, as seen by Jacob Navia's excellent 121 Win32, as seen by Jacob Navia's excellent
121 port of (Chris Fraser & David Hanson)'s excellent 122 port of (Chris Fraser & David Hanson)'s excellent
122 lcc compiler. 123 lcc compiler. Or with MS Visual C.
124 This is selected automatically if compiled by a compiler which
125 defines _WIN32, not including the Cygwin GCC.
123--*/ 126--*/
124#define BZ_LCCWIN32 0 127#define BZ_LCCWIN32 0
125 128
@@ -156,6 +159,7 @@
156--*/ 159--*/
157 160
158#if BZ_UNIX 161#if BZ_UNIX
162# include <fcntl.h>
159# include <sys/types.h> 163# include <sys/types.h>
160# include <utime.h> 164# include <utime.h>
161# include <unistd.h> 165# include <unistd.h>
@@ -164,8 +168,9 @@
164 168
165# define PATH_SEP '/' 169# define PATH_SEP '/'
166# define MY_LSTAT lstat 170# define MY_LSTAT lstat
167# define MY_S_IFREG S_ISREG
168# define MY_STAT stat 171# define MY_STAT stat
172# define MY_S_ISREG S_ISREG
173# define MY_S_ISDIR S_ISDIR
169 174
170# define APPEND_FILESPEC(root, name) \ 175# define APPEND_FILESPEC(root, name) \
171 root=snocString((root), (name)) 176 root=snocString((root), (name))
@@ -180,19 +185,23 @@
180# else 185# else
181# define NORETURN /**/ 186# define NORETURN /**/
182# endif 187# endif
188
183# ifdef __DJGPP__ 189# ifdef __DJGPP__
184# include <io.h> 190# include <io.h>
185# include <fcntl.h> 191# include <fcntl.h>
186# undef MY_LSTAT 192# undef MY_LSTAT
193# undef MY_STAT
187# define MY_LSTAT stat 194# define MY_LSTAT stat
195# define MY_STAT stat
188# undef SET_BINARY_MODE 196# undef SET_BINARY_MODE
189# define SET_BINARY_MODE(fd) \ 197# define SET_BINARY_MODE(fd) \
190 do { \ 198 do { \
191 int retVal = setmode ( fileno ( fd ), \ 199 int retVal = setmode ( fileno ( fd ), \
192 O_BINARY ); \ 200 O_BINARY ); \
193 ERROR_IF_MINUS_ONE ( retVal ); \ 201 ERROR_IF_MINUS_ONE ( retVal ); \
194 } while ( 0 ) 202 } while ( 0 )
195# endif 203# endif
204
196# ifdef __CYGWIN__ 205# ifdef __CYGWIN__
197# include <io.h> 206# include <io.h>
198# include <fcntl.h> 207# include <fcntl.h>
@@ -200,11 +209,11 @@
200# define SET_BINARY_MODE(fd) \ 209# define SET_BINARY_MODE(fd) \
201 do { \ 210 do { \
202 int retVal = setmode ( fileno ( fd ), \ 211 int retVal = setmode ( fileno ( fd ), \
203 O_BINARY ); \ 212 O_BINARY ); \
204 ERROR_IF_MINUS_ONE ( retVal ); \ 213 ERROR_IF_MINUS_ONE ( retVal ); \
205 } while ( 0 ) 214 } while ( 0 )
206# endif 215# endif
207#endif 216#endif /* BZ_UNIX */
208 217
209 218
210 219
@@ -217,46 +226,23 @@
217# define PATH_SEP '\\' 226# define PATH_SEP '\\'
218# define MY_LSTAT _stat 227# define MY_LSTAT _stat
219# define MY_STAT _stat 228# define MY_STAT _stat
220# define MY_S_IFREG(x) ((x) & _S_IFREG) 229# define MY_S_ISREG(x) ((x) & _S_IFREG)
230# define MY_S_ISDIR(x) ((x) & _S_IFDIR)
221 231
222# define APPEND_FLAG(root, name) \ 232# define APPEND_FLAG(root, name) \
223 root=snocString((root), (name)) 233 root=snocString((root), (name))
224 234
225# if 0
226 /*-- lcc-win32 seems to expand wildcards itself --*/
227# define APPEND_FILESPEC(root, spec) \
228 do { \
229 if ((spec)[0] == '-') { \
230 root = snocString((root), (spec)); \
231 } else { \
232 struct _finddata_t c_file; \
233 long hFile; \
234 hFile = _findfirst((spec), &c_file); \
235 if ( hFile == -1L ) { \
236 root = snocString ((root), (spec)); \
237 } else { \
238 int anInt = 0; \
239 while ( anInt == 0 ) { \
240 root = snocString((root), \
241 &c_file.name[0]); \
242 anInt = _findnext(hFile, &c_file); \
243 } \
244 } \
245 } \
246 } while ( 0 )
247# else
248# define APPEND_FILESPEC(root, name) \ 235# define APPEND_FILESPEC(root, name) \
249 root = snocString ((root), (name)) 236 root = snocString ((root), (name))
250# endif
251 237
252# define SET_BINARY_MODE(fd) \ 238# define SET_BINARY_MODE(fd) \
253 do { \ 239 do { \
254 int retVal = setmode ( fileno ( fd ), \ 240 int retVal = setmode ( fileno ( fd ), \
255 O_BINARY ); \ 241 O_BINARY ); \
256 ERROR_IF_MINUS_ONE ( retVal ); \ 242 ERROR_IF_MINUS_ONE ( retVal ); \
257 } while ( 0 ) 243 } while ( 0 )
258 244
259#endif 245#endif /* BZ_LCCWIN32 */
260 246
261 247
262/*---------------------------------------------*/ 248/*---------------------------------------------*/
@@ -338,6 +324,7 @@ typedef
338 struct { UChar b[8]; } 324 struct { UChar b[8]; }
339 UInt64; 325 UInt64;
340 326
327
341static 328static
342void uInt64_from_UInt32s ( UInt64* n, UInt32 lo32, UInt32 hi32 ) 329void uInt64_from_UInt32s ( UInt64* n, UInt32 lo32, UInt32 hi32 )
343{ 330{
@@ -351,6 +338,7 @@ void uInt64_from_UInt32s ( UInt64* n, UInt32 lo32, UInt32 hi32 )
351 n->b[0] = (UChar) (lo32 & 0xFF); 338 n->b[0] = (UChar) (lo32 & 0xFF);
352} 339}
353 340
341
354static 342static
355double uInt64_to_double ( UInt64* n ) 343double uInt64_to_double ( UInt64* n )
356{ 344{
@@ -364,77 +352,6 @@ double uInt64_to_double ( UInt64* n )
364 return sum; 352 return sum;
365} 353}
366 354
367static
368void uInt64_add ( UInt64* src, UInt64* dst )
369{
370 Int32 i;
371 Int32 carry = 0;
372 for (i = 0; i < 8; i++) {
373 carry += ( ((Int32)src->b[i]) + ((Int32)dst->b[i]) );
374 dst->b[i] = (UChar)(carry & 0xFF);
375 carry >>= 8;
376 }
377}
378
379static
380void uInt64_sub ( UInt64* src, UInt64* dst )
381{
382 Int32 t, i;
383 Int32 borrow = 0;
384 for (i = 0; i < 8; i++) {
385 t = ((Int32)dst->b[i]) - ((Int32)src->b[i]) - borrow;
386 if (t < 0) {
387 dst->b[i] = (UChar)(t + 256);
388 borrow = 1;
389 } else {
390 dst->b[i] = (UChar)t;
391 borrow = 0;
392 }
393 }
394}
395
396static
397void uInt64_mul ( UInt64* a, UInt64* b, UInt64* r_hi, UInt64* r_lo )
398{
399 UChar sum[16];
400 Int32 ia, ib, carry;
401 for (ia = 0; ia < 16; ia++) sum[ia] = 0;
402 for (ia = 0; ia < 8; ia++) {
403 carry = 0;
404 for (ib = 0; ib < 8; ib++) {
405 carry += ( ((Int32)sum[ia+ib])
406 + ((Int32)a->b[ia]) * ((Int32)b->b[ib]) );
407 sum[ia+ib] = (UChar)(carry & 0xFF);
408 carry >>= 8;
409 }
410 sum[ia+8] = (UChar)(carry & 0xFF);
411 if ((carry >>= 8) != 0) panic ( "uInt64_mul" );
412 }
413
414 for (ia = 0; ia < 8; ia++) r_hi->b[ia] = sum[ia+8];
415 for (ia = 0; ia < 8; ia++) r_lo->b[ia] = sum[ia];
416}
417
418
419static
420void uInt64_shr1 ( UInt64* n )
421{
422 Int32 i;
423 for (i = 0; i < 8; i++) {
424 n->b[i] >>= 1;
425 if (i < 7 && (n->b[i+1] & 1)) n->b[i] |= 0x80;
426 }
427}
428
429static
430void uInt64_shl1 ( UInt64* n )
431{
432 Int32 i;
433 for (i = 7; i >= 0; i--) {
434 n->b[i] <<= 1;
435 if (i > 0 && (n->b[i-1] & 0x80)) n->b[i]++;
436 }
437}
438 355
439static 356static
440Bool uInt64_isZero ( UInt64* n ) 357Bool uInt64_isZero ( UInt64* n )
@@ -445,49 +362,23 @@ Bool uInt64_isZero ( UInt64* n )
445 return 1; 362 return 1;
446} 363}
447 364
448static 365
366/* Divide *n by 10, and return the remainder. */
367static
449Int32 uInt64_qrm10 ( UInt64* n ) 368Int32 uInt64_qrm10 ( UInt64* n )
450{ 369{
451 /* Divide *n by 10, and return the remainder. Long division 370 UInt32 rem, tmp;
452 is difficult, so we cheat and instead multiply by
453 0xCCCC CCCC CCCC CCCD, which is 0.8 (viz, 0.1 << 3).
454 */
455 Int32 i; 371 Int32 i;
456 UInt64 tmp1, tmp2, n_orig, zero_point_eight; 372 rem = 0;
457 373 for (i = 7; i >= 0; i--) {
458 zero_point_eight.b[1] = zero_point_eight.b[2] = 374 tmp = rem * 256 + n->b[i];
459 zero_point_eight.b[3] = zero_point_eight.b[4] = 375 n->b[i] = tmp / 10;
460 zero_point_eight.b[5] = zero_point_eight.b[6] = 376 rem = tmp % 10;
461 zero_point_eight.b[7] = 0xCC; 377 }
462 zero_point_eight.b[0] = 0xCD; 378 return rem;
463
464 n_orig = *n;
465
466 /* divide n by 10,
467 by multiplying by 0.8 and then shifting right 3 times */
468 uInt64_mul ( n, &zero_point_eight, &tmp1, &tmp2 );
469 uInt64_shr1(&tmp1); uInt64_shr1(&tmp1); uInt64_shr1(&tmp1);
470 *n = tmp1;
471
472 /* tmp1 = 8*n, tmp2 = 2*n */
473 uInt64_shl1(&tmp1); uInt64_shl1(&tmp1); uInt64_shl1(&tmp1);
474 tmp2 = *n; uInt64_shl1(&tmp2);
475
476 /* tmp1 = 10*n */
477 uInt64_add ( &tmp2, &tmp1 );
478
479 /* n_orig = n_orig - 10*n */
480 uInt64_sub ( &tmp1, &n_orig );
481
482 /* n_orig should now hold quotient, in range 0 .. 9 */
483 for (i = 7; i >= 1; i--)
484 if (n_orig.b[i] != 0) panic ( "uInt64_qrm10(1)" );
485 if (n_orig.b[0] > 9)
486 panic ( "uInt64_qrm10(2)" );
487
488 return (int)n_orig.b[0];
489} 379}
490 380
381
491/* ... and the Whole Entire Point of all this UInt64 stuff is 382/* ... and the Whole Entire Point of all this UInt64 stuff is
492 so that we can supply the following function. 383 so that we can supply the following function.
493*/ 384*/
@@ -504,7 +395,8 @@ void uInt64_toAscii ( char* outbuf, UInt64* n )
504 nBuf++; 395 nBuf++;
505 } while (!uInt64_isZero(&n_copy)); 396 } while (!uInt64_isZero(&n_copy));
506 outbuf[nBuf] = 0; 397 outbuf[nBuf] = 0;
507 for (i = 0; i < nBuf; i++) outbuf[i] = buf[nBuf-i-1]; 398 for (i = 0; i < nBuf; i++)
399 outbuf[i] = buf[nBuf-i-1];
508} 400}
509 401
510 402
@@ -566,35 +458,38 @@ void compressStream ( FILE *stream, FILE *zStream )
566 if (ret == EOF) goto errhandler_io; 458 if (ret == EOF) goto errhandler_io;
567 if (zStream != stdout) { 459 if (zStream != stdout) {
568 ret = fclose ( zStream ); 460 ret = fclose ( zStream );
461 outputHandleJustInCase = NULL;
569 if (ret == EOF) goto errhandler_io; 462 if (ret == EOF) goto errhandler_io;
570 } 463 }
464 outputHandleJustInCase = NULL;
571 if (ferror(stream)) goto errhandler_io; 465 if (ferror(stream)) goto errhandler_io;
572 ret = fclose ( stream ); 466 ret = fclose ( stream );
573 if (ret == EOF) goto errhandler_io; 467 if (ret == EOF) goto errhandler_io;
574 468
575 if (nbytes_in_lo32 == 0 && nbytes_in_hi32 == 0)
576 nbytes_in_lo32 = 1;
577
578 if (verbosity >= 1) { 469 if (verbosity >= 1) {
579 Char buf_nin[32], buf_nout[32]; 470 if (nbytes_in_lo32 == 0 && nbytes_in_hi32 == 0) {
580 UInt64 nbytes_in, nbytes_out; 471 fprintf ( stderr, " no data compressed.\n");
581 double nbytes_in_d, nbytes_out_d; 472 } else {
582 uInt64_from_UInt32s ( &nbytes_in, 473 Char buf_nin[32], buf_nout[32];
583 nbytes_in_lo32, nbytes_in_hi32 ); 474 UInt64 nbytes_in, nbytes_out;
584 uInt64_from_UInt32s ( &nbytes_out, 475 double nbytes_in_d, nbytes_out_d;
585 nbytes_out_lo32, nbytes_out_hi32 ); 476 uInt64_from_UInt32s ( &nbytes_in,
586 nbytes_in_d = uInt64_to_double ( &nbytes_in ); 477 nbytes_in_lo32, nbytes_in_hi32 );
587 nbytes_out_d = uInt64_to_double ( &nbytes_out ); 478 uInt64_from_UInt32s ( &nbytes_out,
588 uInt64_toAscii ( buf_nin, &nbytes_in ); 479 nbytes_out_lo32, nbytes_out_hi32 );
589 uInt64_toAscii ( buf_nout, &nbytes_out ); 480 nbytes_in_d = uInt64_to_double ( &nbytes_in );
590 fprintf ( stderr, "%6.3f:1, %6.3f bits/byte, " 481 nbytes_out_d = uInt64_to_double ( &nbytes_out );
591 "%5.2f%% saved, %s in, %s out.\n", 482 uInt64_toAscii ( buf_nin, &nbytes_in );
592 nbytes_in_d / nbytes_out_d, 483 uInt64_toAscii ( buf_nout, &nbytes_out );
593 (8.0 * nbytes_out_d) / nbytes_in_d, 484 fprintf ( stderr, "%6.3f:1, %6.3f bits/byte, "
594 100.0 * (1.0 - nbytes_out_d / nbytes_in_d), 485 "%5.2f%% saved, %s in, %s out.\n",
595 buf_nin, 486 nbytes_in_d / nbytes_out_d,
596 buf_nout 487 (8.0 * nbytes_out_d) / nbytes_in_d,
597 ); 488 100.0 * (1.0 - nbytes_out_d / nbytes_in_d),
489 buf_nin,
490 buf_nout
491 );
492 }
598 } 493 }
599 494
600 return; 495 return;
@@ -652,7 +547,7 @@ Bool uncompressStream ( FILE *zStream, FILE *stream )
652 547
653 while (bzerr == BZ_OK) { 548 while (bzerr == BZ_OK) {
654 nread = BZ2_bzRead ( &bzerr, bzf, obuf, 5000 ); 549 nread = BZ2_bzRead ( &bzerr, bzf, obuf, 5000 );
655 if (bzerr == BZ_DATA_ERROR_MAGIC) goto errhandler; 550 if (bzerr == BZ_DATA_ERROR_MAGIC) goto trycat;
656 if ((bzerr == BZ_OK || bzerr == BZ_STREAM_END) && nread > 0) 551 if ((bzerr == BZ_OK || bzerr == BZ_STREAM_END) && nread > 0)
657 fwrite ( obuf, sizeof(UChar), nread, stream ); 552 fwrite ( obuf, sizeof(UChar), nread, stream );
658 if (ferror(stream)) goto errhandler_io; 553 if (ferror(stream)) goto errhandler_io;
@@ -668,9 +563,9 @@ Bool uncompressStream ( FILE *zStream, FILE *stream )
668 if (bzerr != BZ_OK) panic ( "decompress:bzReadGetUnused" ); 563 if (bzerr != BZ_OK) panic ( "decompress:bzReadGetUnused" );
669 564
670 if (nUnused == 0 && myfeof(zStream)) break; 565 if (nUnused == 0 && myfeof(zStream)) break;
671
672 } 566 }
673 567
568 closeok:
674 if (ferror(zStream)) goto errhandler_io; 569 if (ferror(zStream)) goto errhandler_io;
675 ret = fclose ( zStream ); 570 ret = fclose ( zStream );
676 if (ret == EOF) goto errhandler_io; 571 if (ret == EOF) goto errhandler_io;
@@ -680,11 +575,26 @@ Bool uncompressStream ( FILE *zStream, FILE *stream )
680 if (ret != 0) goto errhandler_io; 575 if (ret != 0) goto errhandler_io;
681 if (stream != stdout) { 576 if (stream != stdout) {
682 ret = fclose ( stream ); 577 ret = fclose ( stream );
578 outputHandleJustInCase = NULL;
683 if (ret == EOF) goto errhandler_io; 579 if (ret == EOF) goto errhandler_io;
684 } 580 }
581 outputHandleJustInCase = NULL;
685 if (verbosity >= 2) fprintf ( stderr, "\n " ); 582 if (verbosity >= 2) fprintf ( stderr, "\n " );
686 return True; 583 return True;
687 584
585 trycat:
586 if (forceOverwrite) {
587 rewind(zStream);
588 while (True) {
589 if (myfeof(zStream)) break;
590 nread = fread ( obuf, sizeof(UChar), 5000, zStream );
591 if (ferror(zStream)) goto errhandler_io;
592 if (nread > 0) fwrite ( obuf, sizeof(UChar), nread, stream );
593 if (ferror(stream)) goto errhandler_io;
594 }
595 goto closeok;
596 }
597
688 errhandler: 598 errhandler:
689 BZ2_bzReadClose ( &bzerr_dummy, bzf ); 599 BZ2_bzReadClose ( &bzerr_dummy, bzf );
690 switch (bzerr) { 600 switch (bzerr) {
@@ -832,7 +742,7 @@ void cadvise ( void )
832 stderr, 742 stderr,
833 "\nIt is possible that the compressed file(s) have become corrupted.\n" 743 "\nIt is possible that the compressed file(s) have become corrupted.\n"
834 "You can use the -tvv option to test integrity of such files.\n\n" 744 "You can use the -tvv option to test integrity of such files.\n\n"
835 "You can use the `bzip2recover' program to *attempt* to recover\n" 745 "You can use the `bzip2recover' program to attempt to recover\n"
836 "data from undamaged sections of corrupted files.\n\n" 746 "data from undamaged sections of corrupted files.\n\n"
837 ); 747 );
838} 748}
@@ -855,28 +765,55 @@ void showFileNames ( void )
855static 765static
856void cleanUpAndFail ( Int32 ec ) 766void cleanUpAndFail ( Int32 ec )
857{ 767{
858 IntNative retVal; 768 IntNative retVal;
769 struct MY_STAT statBuf;
859 770
860 if ( srcMode == SM_F2F 771 if ( srcMode == SM_F2F
861 && opMode != OM_TEST 772 && opMode != OM_TEST
862 && deleteOutputOnInterrupt ) { 773 && deleteOutputOnInterrupt ) {
863 if (noisy) 774
864 fprintf ( stderr, "%s: Deleting output file %s, if it exists.\n", 775 /* Check whether input file still exists. Delete output file
865 progName, outName ); 776 only if input exists to avoid loss of data. Joerg Prante, 5
866 if (outputHandleJustInCase != NULL) 777 January 2002. (JRS 06-Jan-2002: other changes in 1.0.2 mean
867 fclose ( outputHandleJustInCase ); 778 this is less likely to happen. But to be ultra-paranoid, we
868 retVal = remove ( outName ); 779 do the check anyway.) */
869 if (retVal != 0) 780 retVal = MY_STAT ( inName, &statBuf );
781 if (retVal == 0) {
782 if (noisy)
783 fprintf ( stderr,
784 "%s: Deleting output file %s, if it exists.\n",
785 progName, outName );
786 if (outputHandleJustInCase != NULL)
787 fclose ( outputHandleJustInCase );
788 retVal = remove ( outName );
789 if (retVal != 0)
790 fprintf ( stderr,
791 "%s: WARNING: deletion of output file "
792 "(apparently) failed.\n",
793 progName );
794 } else {
870 fprintf ( stderr, 795 fprintf ( stderr,
871 "%s: WARNING: deletion of output file (apparently) failed.\n", 796 "%s: WARNING: deletion of output file suppressed\n",
797 progName );
798 fprintf ( stderr,
799 "%s: since input file no longer exists. Output file\n",
872 progName ); 800 progName );
801 fprintf ( stderr,
802 "%s: `%s' may be incomplete.\n",
803 progName, outName );
804 fprintf ( stderr,
805 "%s: I suggest doing an integrity test (bzip2 -tv)"
806 " of it.\n",
807 progName );
808 }
873 } 809 }
810
874 if (noisy && numFileNames > 0 && numFilesProcessed < numFileNames) { 811 if (noisy && numFileNames > 0 && numFilesProcessed < numFileNames) {
875 fprintf ( stderr, 812 fprintf ( stderr,
876 "%s: WARNING: some files have not been processed:\n" 813 "%s: WARNING: some files have not been processed:\n"
877 "\t%d specified on command line, %d not processed yet.\n\n", 814 "%s: %d specified on command line, %d not processed yet.\n\n",
878 progName, numFileNames, 815 progName, progName,
879 numFileNames - numFilesProcessed ); 816 numFileNames, numFileNames - numFilesProcessed );
880 } 817 }
881 setExit(ec); 818 setExit(ec);
882 exit(exitValue); 819 exit(exitValue);
@@ -915,14 +852,16 @@ void crcError ( void )
915static 852static
916void compressedStreamEOF ( void ) 853void compressedStreamEOF ( void )
917{ 854{
918 fprintf ( stderr, 855 if (noisy) {
919 "\n%s: Compressed file ends unexpectedly;\n\t" 856 fprintf ( stderr,
920 "perhaps it is corrupted? *Possible* reason follows.\n", 857 "\n%s: Compressed file ends unexpectedly;\n\t"
921 progName ); 858 "perhaps it is corrupted? *Possible* reason follows.\n",
922 perror ( progName ); 859 progName );
923 showFileNames(); 860 perror ( progName );
924 cadvise(); 861 showFileNames();
925 cleanUpAndFail( 2 ); 862 cadvise();
863 }
864 cleanUpAndFail( 2 );
926} 865}
927 866
928 867
@@ -1038,6 +977,11 @@ void configError ( void )
1038/*--- The main driver machinery ---*/ 977/*--- The main driver machinery ---*/
1039/*---------------------------------------------------*/ 978/*---------------------------------------------------*/
1040 979
980/* All rather crufty. The main problem is that input files
981 are stat()d multiple times before use. This should be
982 cleaned up.
983*/
984
1041/*---------------------------------------------*/ 985/*---------------------------------------------*/
1042static 986static
1043void pad ( Char *s ) 987void pad ( Char *s )
@@ -1082,6 +1026,32 @@ Bool fileExists ( Char* name )
1082 1026
1083 1027
1084/*---------------------------------------------*/ 1028/*---------------------------------------------*/
1029/* Open an output file safely with O_EXCL and good permissions.
1030 This avoids a race condition in versions < 1.0.2, in which
1031 the file was first opened and then had its interim permissions
1032 set safely. We instead use open() to create the file with
1033 the interim permissions required. (--- --- rw-).
1034
1035 For non-Unix platforms, if we are not worrying about
1036 security issues, simple this simply behaves like fopen.
1037*/
1038FILE* fopen_output_safely ( Char* name, const char* mode )
1039{
1040# if BZ_UNIX
1041 FILE* fp;
1042 IntNative fh;
1043 fh = open(name, O_WRONLY|O_CREAT|O_EXCL, S_IWUSR|S_IRUSR);
1044 if (fh == -1) return NULL;
1045 fp = fdopen(fh, mode);
1046 if (fp == NULL) close(fh);
1047 return fp;
1048# else
1049 return fopen(name, mode);
1050# endif
1051}
1052
1053
1054/*---------------------------------------------*/
1085/*-- 1055/*--
1086 if in doubt, return True 1056 if in doubt, return True
1087--*/ 1057--*/
@@ -1093,7 +1063,7 @@ Bool notAStandardFile ( Char* name )
1093 1063
1094 i = MY_LSTAT ( name, &statBuf ); 1064 i = MY_LSTAT ( name, &statBuf );
1095 if (i != 0) return True; 1065 if (i != 0) return True;
1096 if (MY_S_IFREG(statBuf.st_mode)) return False; 1066 if (MY_S_ISREG(statBuf.st_mode)) return False;
1097 return True; 1067 return True;
1098} 1068}
1099 1069
@@ -1115,42 +1085,66 @@ Int32 countHardLinks ( Char* name )
1115 1085
1116 1086
1117/*---------------------------------------------*/ 1087/*---------------------------------------------*/
1088/* Copy modification date, access date, permissions and owner from the
1089 source to destination file. We have to copy this meta-info off
1090 into fileMetaInfo before starting to compress / decompress it,
1091 because doing it afterwards means we get the wrong access time.
1092
1093 To complicate matters, in compress() and decompress() below, the
1094 sequence of tests preceding the call to saveInputFileMetaInfo()
1095 involves calling fileExists(), which in turn establishes its result
1096 by attempting to fopen() the file, and if successful, immediately
1097 fclose()ing it again. So we have to assume that the fopen() call
1098 does not cause the access time field to be updated.
1099
1100 Reading of the man page for stat() (man 2 stat) on RedHat 7.2 seems
1101 to imply that merely doing open() will not affect the access time.
1102 Therefore we merely need to hope that the C library only does
1103 open() as a result of fopen(), and not any kind of read()-ahead
1104 cleverness.
1105
1106 It sounds pretty fragile to me. Whether this carries across
1107 robustly to arbitrary Unix-like platforms (or even works robustly
1108 on this one, RedHat 7.2) is unknown to me. Nevertheless ...
1109*/
1110#if BZ_UNIX
1111static
1112struct MY_STAT fileMetaInfo;
1113#endif
1114
1118static 1115static
1119void copyDatePermissionsAndOwner ( Char *srcName, Char *dstName ) 1116void saveInputFileMetaInfo ( Char *srcName )
1120{ 1117{
1121#if BZ_UNIX 1118# if BZ_UNIX
1119 IntNative retVal;
1120 /* Note use of stat here, not lstat. */
1121 retVal = MY_STAT( srcName, &fileMetaInfo );
1122 ERROR_IF_NOT_ZERO ( retVal );
1123# endif
1124}
1125
1126
1127static
1128void applySavedMetaInfoToOutputFile ( Char *dstName )
1129{
1130# if BZ_UNIX
1122 IntNative retVal; 1131 IntNative retVal;
1123 struct MY_STAT statBuf;
1124 struct utimbuf uTimBuf; 1132 struct utimbuf uTimBuf;
1125 1133
1126 retVal = MY_LSTAT ( srcName, &statBuf ); 1134 uTimBuf.actime = fileMetaInfo.st_atime;
1127 ERROR_IF_NOT_ZERO ( retVal ); 1135 uTimBuf.modtime = fileMetaInfo.st_mtime;
1128 uTimBuf.actime = statBuf.st_atime;
1129 uTimBuf.modtime = statBuf.st_mtime;
1130 1136
1131 retVal = chmod ( dstName, statBuf.st_mode ); 1137 retVal = chmod ( dstName, fileMetaInfo.st_mode );
1132 ERROR_IF_NOT_ZERO ( retVal ); 1138 ERROR_IF_NOT_ZERO ( retVal );
1133 1139
1134 retVal = utime ( dstName, &uTimBuf ); 1140 retVal = utime ( dstName, &uTimBuf );
1135 ERROR_IF_NOT_ZERO ( retVal ); 1141 ERROR_IF_NOT_ZERO ( retVal );
1136 1142
1137 retVal = chown ( dstName, statBuf.st_uid, statBuf.st_gid ); 1143 retVal = chown ( dstName, fileMetaInfo.st_uid, fileMetaInfo.st_gid );
1138 /* chown() will in many cases return with EPERM, which can 1144 /* chown() will in many cases return with EPERM, which can
1139 be safely ignored. 1145 be safely ignored.
1140 */ 1146 */
1141#endif 1147# endif
1142}
1143
1144
1145/*---------------------------------------------*/
1146static
1147void setInterimPermissions ( Char *dstName )
1148{
1149#if BZ_UNIX
1150 IntNative retVal;
1151 retVal = chmod ( dstName, S_IRUSR | S_IWUSR );
1152 ERROR_IF_NOT_ZERO ( retVal );
1153#endif
1154} 1148}
1155 1149
1156 1150
@@ -1158,10 +1152,19 @@ void setInterimPermissions ( Char *dstName )
1158static 1152static
1159Bool containsDubiousChars ( Char* name ) 1153Bool containsDubiousChars ( Char* name )
1160{ 1154{
1161 Bool cdc = False; 1155# if BZ_UNIX
1156 /* On unix, files can contain any characters and the file expansion
1157 * is performed by the shell.
1158 */
1159 return False;
1160# else /* ! BZ_UNIX */
1161 /* On non-unix (Win* platforms), wildcard characters are not allowed in
1162 * filenames.
1163 */
1162 for (; *name != '\0'; name++) 1164 for (; *name != '\0'; name++)
1163 if (*name == '?' || *name == '*') cdc = True; 1165 if (*name == '?' || *name == '*') return True;
1164 return cdc; 1166 return False;
1167# endif /* BZ_UNIX */
1165} 1168}
1166 1169
1167 1170
@@ -1201,6 +1204,7 @@ void compress ( Char *name )
1201 FILE *inStr; 1204 FILE *inStr;
1202 FILE *outStr; 1205 FILE *outStr;
1203 Int32 n, i; 1206 Int32 n, i;
1207 struct MY_STAT statBuf;
1204 1208
1205 deleteOutputOnInterrupt = False; 1209 deleteOutputOnInterrupt = False;
1206 1210
@@ -1246,6 +1250,16 @@ void compress ( Char *name )
1246 return; 1250 return;
1247 } 1251 }
1248 } 1252 }
1253 if ( srcMode == SM_F2F || srcMode == SM_F2O ) {
1254 MY_STAT(inName, &statBuf);
1255 if ( MY_S_ISDIR(statBuf.st_mode) ) {
1256 fprintf( stderr,
1257 "%s: Input file %s is a directory.\n",
1258 progName,inName);
1259 setExit(1);
1260 return;
1261 }
1262 }
1249 if ( srcMode == SM_F2F && !forceOverwrite && notAStandardFile ( inName )) { 1263 if ( srcMode == SM_F2F && !forceOverwrite && notAStandardFile ( inName )) {
1250 if (noisy) 1264 if (noisy)
1251 fprintf ( stderr, "%s: Input file %s is not a normal file.\n", 1265 fprintf ( stderr, "%s: Input file %s is not a normal file.\n",
@@ -1253,11 +1267,15 @@ void compress ( Char *name )
1253 setExit(1); 1267 setExit(1);
1254 return; 1268 return;
1255 } 1269 }
1256 if ( srcMode == SM_F2F && !forceOverwrite && fileExists ( outName ) ) { 1270 if ( srcMode == SM_F2F && fileExists ( outName ) ) {
1257 fprintf ( stderr, "%s: Output file %s already exists.\n", 1271 if (forceOverwrite) {
1258 progName, outName ); 1272 remove(outName);
1259 setExit(1); 1273 } else {
1260 return; 1274 fprintf ( stderr, "%s: Output file %s already exists.\n",
1275 progName, outName );
1276 setExit(1);
1277 return;
1278 }
1261 } 1279 }
1262 if ( srcMode == SM_F2F && !forceOverwrite && 1280 if ( srcMode == SM_F2F && !forceOverwrite &&
1263 (n=countHardLinks ( inName )) > 0) { 1281 (n=countHardLinks ( inName )) > 0) {
@@ -1267,6 +1285,12 @@ void compress ( Char *name )
1267 return; 1285 return;
1268 } 1286 }
1269 1287
1288 if ( srcMode == SM_F2F ) {
1289 /* Save the file's meta-info before we open it. Doing it later
1290 means we mess up the access times. */
1291 saveInputFileMetaInfo ( inName );
1292 }
1293
1270 switch ( srcMode ) { 1294 switch ( srcMode ) {
1271 1295
1272 case SM_I2O: 1296 case SM_I2O:
@@ -1306,7 +1330,7 @@ void compress ( Char *name )
1306 1330
1307 case SM_F2F: 1331 case SM_F2F:
1308 inStr = fopen ( inName, "rb" ); 1332 inStr = fopen ( inName, "rb" );
1309 outStr = fopen ( outName, "wb" ); 1333 outStr = fopen_output_safely ( outName, "wb" );
1310 if ( outStr == NULL) { 1334 if ( outStr == NULL) {
1311 fprintf ( stderr, "%s: Can't create output file %s: %s.\n", 1335 fprintf ( stderr, "%s: Can't create output file %s: %s.\n",
1312 progName, outName, strerror(errno) ); 1336 progName, outName, strerror(errno) );
@@ -1321,7 +1345,6 @@ void compress ( Char *name )
1321 setExit(1); 1345 setExit(1);
1322 return; 1346 return;
1323 }; 1347 };
1324 setInterimPermissions ( outName );
1325 break; 1348 break;
1326 1349
1327 default: 1350 default:
@@ -1343,7 +1366,7 @@ void compress ( Char *name )
1343 1366
1344 /*--- If there was an I/O error, we won't get here. ---*/ 1367 /*--- If there was an I/O error, we won't get here. ---*/
1345 if ( srcMode == SM_F2F ) { 1368 if ( srcMode == SM_F2F ) {
1346 copyDatePermissionsAndOwner ( inName, outName ); 1369 applySavedMetaInfoToOutputFile ( outName );
1347 deleteOutputOnInterrupt = False; 1370 deleteOutputOnInterrupt = False;
1348 if ( !keepInputFiles ) { 1371 if ( !keepInputFiles ) {
1349 IntNative retVal = remove ( inName ); 1372 IntNative retVal = remove ( inName );
@@ -1364,6 +1387,7 @@ void uncompress ( Char *name )
1364 Int32 n, i; 1387 Int32 n, i;
1365 Bool magicNumberOK; 1388 Bool magicNumberOK;
1366 Bool cantGuess; 1389 Bool cantGuess;
1390 struct MY_STAT statBuf;
1367 1391
1368 deleteOutputOnInterrupt = False; 1392 deleteOutputOnInterrupt = False;
1369 1393
@@ -1405,6 +1429,16 @@ void uncompress ( Char *name )
1405 setExit(1); 1429 setExit(1);
1406 return; 1430 return;
1407 } 1431 }
1432 if ( srcMode == SM_F2F || srcMode == SM_F2O ) {
1433 MY_STAT(inName, &statBuf);
1434 if ( MY_S_ISDIR(statBuf.st_mode) ) {
1435 fprintf( stderr,
1436 "%s: Input file %s is a directory.\n",
1437 progName,inName);
1438 setExit(1);
1439 return;
1440 }
1441 }
1408 if ( srcMode == SM_F2F && !forceOverwrite && notAStandardFile ( inName )) { 1442 if ( srcMode == SM_F2F && !forceOverwrite && notAStandardFile ( inName )) {
1409 if (noisy) 1443 if (noisy)
1410 fprintf ( stderr, "%s: Input file %s is not a normal file.\n", 1444 fprintf ( stderr, "%s: Input file %s is not a normal file.\n",
@@ -1419,11 +1453,15 @@ void uncompress ( Char *name )
1419 progName, inName, outName ); 1453 progName, inName, outName );
1420 /* just a warning, no return */ 1454 /* just a warning, no return */
1421 } 1455 }
1422 if ( srcMode == SM_F2F && !forceOverwrite && fileExists ( outName ) ) { 1456 if ( srcMode == SM_F2F && fileExists ( outName ) ) {
1423 fprintf ( stderr, "%s: Output file %s already exists.\n", 1457 if (forceOverwrite) {
1424 progName, outName ); 1458 remove(outName);
1425 setExit(1); 1459 } else {
1426 return; 1460 fprintf ( stderr, "%s: Output file %s already exists.\n",
1461 progName, outName );
1462 setExit(1);
1463 return;
1464 }
1427 } 1465 }
1428 if ( srcMode == SM_F2F && !forceOverwrite && 1466 if ( srcMode == SM_F2F && !forceOverwrite &&
1429 (n=countHardLinks ( inName ) ) > 0) { 1467 (n=countHardLinks ( inName ) ) > 0) {
@@ -1433,6 +1471,12 @@ void uncompress ( Char *name )
1433 return; 1471 return;
1434 } 1472 }
1435 1473
1474 if ( srcMode == SM_F2F ) {
1475 /* Save the file's meta-info before we open it. Doing it later
1476 means we mess up the access times. */
1477 saveInputFileMetaInfo ( inName );
1478 }
1479
1436 switch ( srcMode ) { 1480 switch ( srcMode ) {
1437 1481
1438 case SM_I2O: 1482 case SM_I2O:
@@ -1463,7 +1507,7 @@ void uncompress ( Char *name )
1463 1507
1464 case SM_F2F: 1508 case SM_F2F:
1465 inStr = fopen ( inName, "rb" ); 1509 inStr = fopen ( inName, "rb" );
1466 outStr = fopen ( outName, "wb" ); 1510 outStr = fopen_output_safely ( outName, "wb" );
1467 if ( outStr == NULL) { 1511 if ( outStr == NULL) {
1468 fprintf ( stderr, "%s: Can't create output file %s: %s.\n", 1512 fprintf ( stderr, "%s: Can't create output file %s: %s.\n",
1469 progName, outName, strerror(errno) ); 1513 progName, outName, strerror(errno) );
@@ -1478,7 +1522,6 @@ void uncompress ( Char *name )
1478 setExit(1); 1522 setExit(1);
1479 return; 1523 return;
1480 }; 1524 };
1481 setInterimPermissions ( outName );
1482 break; 1525 break;
1483 1526
1484 default: 1527 default:
@@ -1501,7 +1544,7 @@ void uncompress ( Char *name )
1501 /*--- If there was an I/O error, we won't get here. ---*/ 1544 /*--- If there was an I/O error, we won't get here. ---*/
1502 if ( magicNumberOK ) { 1545 if ( magicNumberOK ) {
1503 if ( srcMode == SM_F2F ) { 1546 if ( srcMode == SM_F2F ) {
1504 copyDatePermissionsAndOwner ( inName, outName ); 1547 applySavedMetaInfoToOutputFile ( outName );
1505 deleteOutputOnInterrupt = False; 1548 deleteOutputOnInterrupt = False;
1506 if ( !keepInputFiles ) { 1549 if ( !keepInputFiles ) {
1507 IntNative retVal = remove ( inName ); 1550 IntNative retVal = remove ( inName );
@@ -1539,6 +1582,7 @@ void testf ( Char *name )
1539{ 1582{
1540 FILE *inStr; 1583 FILE *inStr;
1541 Bool allOK; 1584 Bool allOK;
1585 struct MY_STAT statBuf;
1542 1586
1543 deleteOutputOnInterrupt = False; 1587 deleteOutputOnInterrupt = False;
1544 1588
@@ -1565,6 +1609,16 @@ void testf ( Char *name )
1565 setExit(1); 1609 setExit(1);
1566 return; 1610 return;
1567 } 1611 }
1612 if ( srcMode != SM_I2O ) {
1613 MY_STAT(inName, &statBuf);
1614 if ( MY_S_ISDIR(statBuf.st_mode) ) {
1615 fprintf( stderr,
1616 "%s: Input file %s is a directory.\n",
1617 progName,inName);
1618 setExit(1);
1619 return;
1620 }
1621 }
1568 1622
1569 switch ( srcMode ) { 1623 switch ( srcMode ) {
1570 1624
@@ -1603,6 +1657,7 @@ void testf ( Char *name )
1603 } 1657 }
1604 1658
1605 /*--- Now the input handle is sane. Do the Biz. ---*/ 1659 /*--- Now the input handle is sane. Do the Biz. ---*/
1660 outputHandleJustInCase = NULL;
1606 allOK = testStream ( inStr ); 1661 allOK = testStream ( inStr );
1607 1662
1608 if (allOK && verbosity >= 1) fprintf ( stderr, "ok\n" ); 1663 if (allOK && verbosity >= 1) fprintf ( stderr, "ok\n" );
@@ -1619,7 +1674,7 @@ void license ( void )
1619 "bzip2, a block-sorting file compressor. " 1674 "bzip2, a block-sorting file compressor. "
1620 "Version %s.\n" 1675 "Version %s.\n"
1621 " \n" 1676 " \n"
1622 " Copyright (C) 1996-2000 by Julian Seward.\n" 1677 " Copyright (C) 1996-2002 by Julian Seward.\n"
1623 " \n" 1678 " \n"
1624 " This program is free software; you can redistribute it and/or modify\n" 1679 " This program is free software; you can redistribute it and/or modify\n"
1625 " it under the terms set out in the LICENSE file, which is included\n" 1680 " it under the terms set out in the LICENSE file, which is included\n"
@@ -1658,6 +1713,8 @@ void usage ( Char *fullProgName )
1658 " -V --version display software version & license\n" 1713 " -V --version display software version & license\n"
1659 " -s --small use less memory (at most 2500k)\n" 1714 " -s --small use less memory (at most 2500k)\n"
1660 " -1 .. -9 set block size to 100k .. 900k\n" 1715 " -1 .. -9 set block size to 100k .. 900k\n"
1716 " --fast alias for -1\n"
1717 " --best alias for -9\n"
1661 "\n" 1718 "\n"
1662 " If invoked as `bzip2', default action is to compress.\n" 1719 " If invoked as `bzip2', default action is to compress.\n"
1663 " as `bunzip2', default action is to decompress.\n" 1720 " as `bunzip2', default action is to decompress.\n"
@@ -1666,9 +1723,9 @@ void usage ( Char *fullProgName )
1666 " If no file names are given, bzip2 compresses or decompresses\n" 1723 " If no file names are given, bzip2 compresses or decompresses\n"
1667 " from standard input to standard output. You can combine\n" 1724 " from standard input to standard output. You can combine\n"
1668 " short flags, so `-v -4' means the same as -v4 or -4v, &c.\n" 1725 " short flags, so `-v -4' means the same as -v4 or -4v, &c.\n"
1669#if BZ_UNIX 1726# if BZ_UNIX
1670 "\n" 1727 "\n"
1671#endif 1728# endif
1672 , 1729 ,
1673 1730
1674 BZ2_bzlibVersion(), 1731 BZ2_bzlibVersion(),
@@ -1818,11 +1875,11 @@ IntNative main ( IntNative argc, Char *argv[] )
1818 1875
1819 /*-- Set up signal handlers for mem access errors --*/ 1876 /*-- Set up signal handlers for mem access errors --*/
1820 signal (SIGSEGV, mySIGSEGVorSIGBUScatcher); 1877 signal (SIGSEGV, mySIGSEGVorSIGBUScatcher);
1821#if BZ_UNIX 1878# if BZ_UNIX
1822#ifndef __DJGPP__ 1879# ifndef __DJGPP__
1823 signal (SIGBUS, mySIGSEGVorSIGBUScatcher); 1880 signal (SIGBUS, mySIGSEGVorSIGBUScatcher);
1824#endif 1881# endif
1825#endif 1882# endif
1826 1883
1827 copyFileName ( inName, "(none)" ); 1884 copyFileName ( inName, "(none)" );
1828 copyFileName ( outName, "(none)" ); 1885 copyFileName ( outName, "(none)" );
@@ -1933,6 +1990,8 @@ IntNative main ( IntNative argc, Char *argv[] )
1933 if (ISFLAG("--exponential")) workFactor = 1; else 1990 if (ISFLAG("--exponential")) workFactor = 1; else
1934 if (ISFLAG("--repetitive-best")) redundant(aa->name); else 1991 if (ISFLAG("--repetitive-best")) redundant(aa->name); else
1935 if (ISFLAG("--repetitive-fast")) redundant(aa->name); else 1992 if (ISFLAG("--repetitive-fast")) redundant(aa->name); else
1993 if (ISFLAG("--fast")) blockSize100k = 1; else
1994 if (ISFLAG("--best")) blockSize100k = 9; else
1936 if (ISFLAG("--verbose")) verbosity++; else 1995 if (ISFLAG("--verbose")) verbosity++; else
1937 if (ISFLAG("--help")) { usage ( progName ); exit ( 0 ); } 1996 if (ISFLAG("--help")) { usage ( progName ); exit ( 0 ); }
1938 else 1997 else
diff --git a/bzip2.txt b/bzip2.txt
index 4f1ae86..6afe358 100644
--- a/bzip2.txt
+++ b/bzip2.txt
@@ -1,7 +1,6 @@
1 1
2
3NAME 2NAME
4 bzip2, bunzip2 - a block-sorting file compressor, v1.0 3 bzip2, bunzip2 - a block-sorting file compressor, v1.0.2
5 bzcat - decompresses files to stdout 4 bzcat - decompresses files to stdout
6 bzip2recover - recovers data from damaged bzip2 files 5 bzip2recover - recovers data from damaged bzip2 files
7 6
@@ -18,20 +17,20 @@ DESCRIPTION
18 sorting text compression algorithm, and Huffman coding. 17 sorting text compression algorithm, and Huffman coding.
19 Compression is generally considerably better than that 18 Compression is generally considerably better than that
20 achieved by more conventional LZ77/LZ78-based compressors, 19 achieved by more conventional LZ77/LZ78-based compressors,
21 and approaches the performance of the PPM family of sta- 20 and approaches the performance of the PPM family of sta­
22 tistical compressors. 21 tistical compressors.
23 22
24 The command-line options are deliberately very similar to 23 The command-line options are deliberately very similar to
25 those of GNU gzip, but they are not identical. 24 those of GNU gzip, but they are not identical.
26 25
27 bzip2 expects a list of file names to accompany the com- 26 bzip2 expects a list of file names to accompany the com­
28 mand-line flags. Each file is replaced by a compressed 27 mand-line flags. Each file is replaced by a compressed
29 version of itself, with the name "original_name.bz2". 28 version of itself, with the name "original_name.bz2".
30 Each compressed file has the same modification date, per- 29 Each compressed file has the same modification date, per­
31 missions, and, when possible, ownership as the correspond- 30 missions, and, when possible, ownership as the correspond­
32 ing original, so that these properties can be correctly 31 ing original, so that these properties can be correctly
33 restored at decompression time. File name handling is 32 restored at decompression time. File name handling is
34 naive in the sense that there is no mechanism for preserv- 33 naive in the sense that there is no mechanism for preserv­
35 ing original file names, permissions, ownerships or dates 34 ing original file names, permissions, ownerships or dates
36 in filesystems which lack these concepts, or have serious 35 in filesystems which lack these concepts, or have serious
37 file name length restrictions, such as MS-DOS. 36 file name length restrictions, such as MS-DOS.
@@ -62,23 +61,23 @@ DESCRIPTION
62 guess the name of the original file, and uses the original 61 guess the name of the original file, and uses the original
63 name with .out appended. 62 name with .out appended.
64 63
65 As with compression, supplying no filenames causes decom- 64 As with compression, supplying no filenames causes decom­
66 pression from standard input to standard output. 65 pression from standard input to standard output.
67 66
68 bunzip2 will correctly decompress a file which is the con- 67 bunzip2 will correctly decompress a file which is the con­
69 catenation of two or more compressed files. The result is 68 catenation of two or more compressed files. The result is
70 the concatenation of the corresponding uncompressed files. 69 the concatenation of the corresponding uncompressed files.
71 Integrity testing (-t) of concatenated compressed files is 70 Integrity testing (-t) of concatenated compressed files is
72 also supported. 71 also supported.
73 72
74 You can also compress or decompress files to the standard 73 You can also compress or decompress files to the standard
75 output by giving the -c flag. Multiple files may be com- 74 output by giving the -c flag. Multiple files may be com­
76 pressed and decompressed like this. The resulting outputs 75 pressed and decompressed like this. The resulting outputs
77 are fed sequentially to stdout. Compression of multiple 76 are fed sequentially to stdout. Compression of multiple
78 files in this manner generates a stream containing multi- 77 files in this manner generates a stream containing multi­
79 ple compressed file representations. Such a stream can be 78 ple compressed file representations. Such a stream can be
80 decompressed correctly only by bzip2 version 0.9.0 or 79 decompressed correctly only by bzip2 version 0.9.0 or
81 later. Earlier versions of bzip2 will stop after decom- 80 later. Earlier versions of bzip2 will stop after decom­
82 pressing the first file in the stream. 81 pressing the first file in the stream.
83 82
84 bzcat (or bzip2 -dc) decompresses all specified files to 83 bzcat (or bzip2 -dc) decompresses all specified files to
@@ -99,7 +98,7 @@ DESCRIPTION
99 98
100 As a self-check for your protection, bzip2 uses 32-bit 99 As a self-check for your protection, bzip2 uses 32-bit
101 CRCs to make sure that the decompressed version of a file 100 CRCs to make sure that the decompressed version of a file
102 is identical to the original. This guards against corrup- 101 is identical to the original. This guards against corrup­
103 tion of the compressed data, and against undetected bugs 102 tion of the compressed data, and against undetected bugs
104 in bzip2 (hopefully very unlikely). The chances of data 103 in bzip2 (hopefully very unlikely). The chances of data
105 corruption going undetected is microscopic, about one 104 corruption going undetected is microscopic, about one
@@ -127,8 +126,8 @@ OPTIONS
127 and forces bzip2 to decompress. 126 and forces bzip2 to decompress.
128 127
129 -z --compress 128 -z --compress
130 The complement to -d: forces compression, regard- 129 The complement to -d: forces compression,
131 less of the invokation name. 130 regardless of the invocation name.
132 131
133 -t --test 132 -t --test
134 Check integrity of the specified file(s), but don't 133 Check integrity of the specified file(s), but don't
@@ -141,6 +140,11 @@ OPTIONS
141 forces bzip2 to break hard links to files, which it 140 forces bzip2 to break hard links to files, which it
142 otherwise wouldn't do. 141 otherwise wouldn't do.
143 142
143 bzip2 normally declines to decompress files which
144 don't have the correct magic header bytes. If
145 forced (-f), however, it will pass such files
146 through unmodified. This is how GNU gzip behaves.
147
144 -k --keep 148 -k --keep
145 Keep (don't delete) input files during compression 149 Keep (don't delete) input files during compression
146 or decompression. 150 or decompression.
@@ -167,7 +171,7 @@ OPTIONS
167 171
168 -v --verbose 172 -v --verbose
169 Verbose mode -- show the compression ratio for each 173 Verbose mode -- show the compression ratio for each
170 file processed. Further -v's increase the ver- 174 file processed. Further -v's increase the ver­
171 bosity level, spewing out lots of information which 175 bosity level, spewing out lots of information which
172 is primarily of interest for diagnostic purposes. 176 is primarily of interest for diagnostic purposes.
173 177
@@ -175,20 +179,24 @@ OPTIONS
175 Display the software version, license terms and 179 Display the software version, license terms and
176 conditions. 180 conditions.
177 181
178 -1 to -9 182 -1 (or --fast) to -9 (or --best)
179 Set the block size to 100 k, 200 k .. 900 k when 183 Set the block size to 100 k, 200 k .. 900 k when
180 compressing. Has no effect when decompressing. 184 compressing. Has no effect when decompressing.
181 See MEMORY MANAGEMENT below. 185 See MEMORY MANAGEMENT below. The --fast and --best
186 aliases are primarily for GNU gzip compatibility.
187 In particular, --fast doesn't make things signifi­
188 cantly faster. And --best merely selects the
189 default behaviour.
182 190
183 -- Treats all subsequent arguments as file names, even 191 -- Treats all subsequent arguments as file names, even
184 if they start with a dash. This is so you can han- 192 if they start with a dash. This is so you can han­
185 dle files with names beginning with a dash, for 193 dle files with names beginning with a dash, for
186 example: bzip2 -- -myfilename. 194 example: bzip2 -- -myfilename.
187 195
188 --repetitive-fast --repetitive-best 196 --repetitive-fast --repetitive-best
189 These flags are redundant in versions 0.9.5 and 197 These flags are redundant in versions 0.9.5 and
190 above. They provided some coarse control over the 198 above. They provided some coarse control over the
191 behaviour of the sorting algorithm in earlier ver- 199 behaviour of the sorting algorithm in earlier ver­
192 sions, which was sometimes useful. 0.9.5 and above 200 sions, which was sometimes useful. 0.9.5 and above
193 have an improved algorithm which renders these 201 have an improved algorithm which renders these
194 flags irrelevant. 202 flags irrelevant.
@@ -199,7 +207,7 @@ MEMORY MANAGEMENT
199 affects both the compression ratio achieved, and the 207 affects both the compression ratio achieved, and the
200 amount of memory needed for compression and decompression. 208 amount of memory needed for compression and decompression.
201 The flags -1 through -9 specify the block size to be 209 The flags -1 through -9 specify the block size to be
202 100,000 bytes through 900,000 bytes (the default) respec- 210 100,000 bytes through 900,000 bytes (the default) respec­
203 tively. At decompression time, the block size used for 211 tively. At decompression time, the block size used for
204 compression is read from the header of the compressed 212 compression is read from the header of the compressed
205 file, and bunzip2 then allocates itself just enough memory 213 file, and bunzip2 then allocates itself just enough memory
@@ -227,13 +235,13 @@ MEMORY MANAGEMENT
227 bunzip2 will require about 3700 kbytes to decompress. To 235 bunzip2 will require about 3700 kbytes to decompress. To
228 support decompression of any file on a 4 megabyte machine, 236 support decompression of any file on a 4 megabyte machine,
229 bunzip2 has an option to decompress using approximately 237 bunzip2 has an option to decompress using approximately
230 half this amount of memory, about 2300 kbytes. Decompres- 238 half this amount of memory, about 2300 kbytes. Decompres­
231 sion speed is also halved, so you should use this option 239 sion speed is also halved, so you should use this option
232 only where necessary. The relevant flag is -s. 240 only where necessary. The relevant flag is -s.
233 241
234 In general, try and use the largest block size memory con- 242 In general, try and use the largest block size memory con­
235 straints allow, since that maximises the compression 243 straints allow, since that maximises the compression
236 achieved. Compression and decompression speed are virtu- 244 achieved. Compression and decompression speed are virtu­
237 ally unaffected by block size. 245 ally unaffected by block size.
238 246
239 Another significant point applies to files which fit in a 247 Another significant point applies to files which fit in a
@@ -249,11 +257,11 @@ MEMORY MANAGEMENT
249 257
250 Here is a table which summarises the maximum memory usage 258 Here is a table which summarises the maximum memory usage
251 for different block sizes. Also recorded is the total 259 for different block sizes. Also recorded is the total
252 compressed size for 14 files of the Calgary Text Compres- 260 compressed size for 14 files of the Calgary Text Compres­
253 sion Corpus totalling 3,141,622 bytes. This column gives 261 sion Corpus totalling 3,141,622 bytes. This column gives
254 some feel for how compression varies with block size. 262 some feel for how compression varies with block size.
255 These figures tend to understate the advantage of larger 263 These figures tend to understate the advantage of larger
256 block sizes for larger files, since the Corpus is domi- 264 block sizes for larger files, since the Corpus is domi­
257 nated by smaller files. 265 nated by smaller files.
258 266
259 Compress Decompress Decompress Corpus 267 Compress Decompress Decompress Corpus
@@ -272,7 +280,7 @@ MEMORY MANAGEMENT
272 280
273RECOVERING DATA FROM DAMAGED FILES 281RECOVERING DATA FROM DAMAGED FILES
274 bzip2 compresses files in blocks, usually 900kbytes long. 282 bzip2 compresses files in blocks, usually 900kbytes long.
275 Each block is handled independently. If a media or trans- 283 Each block is handled independently. If a media or trans­
276 mission error causes a multi-block .bz2 file to become 284 mission error causes a multi-block .bz2 file to become
277 damaged, it may be possible to recover data from the 285 damaged, it may be possible to recover data from the
278 undamaged blocks in the file. 286 undamaged blocks in the file.
@@ -289,19 +297,19 @@ RECOVERING DATA FROM DAMAGED FILES
289 the integrity of the resulting files, and decompress those 297 the integrity of the resulting files, and decompress those
290 which are undamaged. 298 which are undamaged.
291 299
292 bzip2recover takes a single argument, the name of the dam- 300 bzip2recover takes a single argument, the name of the dam­
293 aged file, and writes a number of files "rec0001file.bz2", 301 aged file, and writes a number of files
294 "rec0002file.bz2", etc, containing the extracted blocks. 302 "rec00001file.bz2", "rec00002file.bz2", etc, containing
295 The output filenames are designed so that the use of 303 the extracted blocks. The output filenames are
296 wildcards in subsequent processing -- for example, "bzip2 304 designed so that the use of wildcards in subsequent pro­
297 -dc rec*file.bz2 > recovered_data" -- lists the files in 305 cessing -- for example, "bzip2 -dc rec*file.bz2 > recov­
298 the correct order. 306 ered_data" -- processes the files in the correct order.
299 307
300 bzip2recover should be of most use dealing with large .bz2 308 bzip2recover should be of most use dealing with large .bz2
301 files, as these will contain many blocks. It is clearly 309 files, as these will contain many blocks. It is clearly
302 futile to use it on damaged single-block files, since a 310 futile to use it on damaged single-block files, since a
303 damaged block cannot be recovered. If you wish to min- 311 damaged block cannot be recovered. If you wish to min­
304 imise any potential data loss through media or transmis- 312 imise any potential data loss through media or transmis­
305 sion errors, you might consider compressing with a smaller 313 sion errors, you might consider compressing with a smaller
306 block size. 314 block size.
307 315
@@ -315,19 +323,19 @@ PERFORMANCE NOTES
315 better than previous versions in this respect. The ratio 323 better than previous versions in this respect. The ratio
316 between worst-case and average-case compression time is in 324 between worst-case and average-case compression time is in
317 the region of 10:1. For previous versions, this figure 325 the region of 10:1. For previous versions, this figure
318 was more like 100:1. You can use the -vvvv option to mon- 326 was more like 100:1. You can use the -vvvv option to mon­
319 itor progress in great detail, if you want. 327 itor progress in great detail, if you want.
320 328
321 Decompression speed is unaffected by these phenomena. 329 Decompression speed is unaffected by these phenomena.
322 330
323 bzip2 usually allocates several megabytes of memory to 331 bzip2 usually allocates several megabytes of memory to
324 operate in, and then charges all over it in a fairly ran- 332 operate in, and then charges all over it in a fairly ran­
325 dom fashion. This means that performance, both for com- 333 dom fashion. This means that performance, both for com­
326 pressing and decompressing, is largely determined by the 334 pressing and decompressing, is largely determined by the
327 speed at which your machine can service cache misses. 335 speed at which your machine can service cache misses.
328 Because of this, small changes to the code to reduce the 336 Because of this, small changes to the code to reduce the
329 miss rate have been observed to give disproportionately 337 miss rate have been observed to give disproportionately
330 large performance improvements. I imagine bzip2 will per- 338 large performance improvements. I imagine bzip2 will per­
331 form best on machines with very large caches. 339 form best on machines with very large caches.
332 340
333 341
@@ -337,40 +345,46 @@ CAVEATS
337 but the details of what the problem is sometimes seem 345 but the details of what the problem is sometimes seem
338 rather misleading. 346 rather misleading.
339 347
340 This manual page pertains to version 1.0 of bzip2. Com- 348 This manual page pertains to version 1.0.2 of bzip2. Com­
341 pressed data created by this version is entirely forwards 349 pressed data created by this version is entirely forwards
342 and backwards compatible with the previous public 350 and backwards compatible with the previous public
343 releases, versions 0.1pl2, 0.9.0 and 0.9.5, but with the 351 releases, versions 0.1pl2, 0.9.0, 0.9.5, 1.0.0 and 1.0.1,
344 following exception: 0.9.0 and above can correctly decom- 352 but with the following exception: 0.9.0 and above can cor­
345 press multiple concatenated compressed files. 0.1pl2 can- 353 rectly decompress multiple concatenated compressed files.
346 not do this; it will stop after decompressing just the 354 0.1pl2 cannot do this; it will stop after decompressing
347 first file in the stream. 355 just the first file in the stream.
348 356
349 bzip2recover uses 32-bit integers to represent bit posi- 357 bzip2recover versions prior to this one, 1.0.2, used
350 tions in compressed files, so it cannot handle compressed 358 32-bit integers to represent bit positions in compressed
351 files more than 512 megabytes long. This could easily be 359 files, so it could not handle compressed files more than
352 fixed. 360 512 megabytes long. Version 1.0.2 and above uses 64-bit
361 ints on some platforms which support them (GNU supported
362 targets, and Windows). To establish whether or not
363 bzip2recover was built with such a limitation, run it
364 without arguments. In any event you can build yourself an
365 unlimited version if you can recompile it with MaybeUInt64
366 set to be an unsigned 64-bit integer.
353 367
354 368
355AUTHOR 369AUTHOR
356 Julian Seward, jseward@acm.org. 370 Julian Seward, jseward@acm.org.
357 371
358 http://sourceware.cygnus.com/bzip2 372 http://sources.redhat.com/bzip2
359 http://www.muraroa.demon.co.uk
360 373
361 The ideas embodied in bzip2 are due to (at least) the fol- 374 The ideas embodied in bzip2 are due to (at least) the fol­
362 lowing people: Michael Burrows and David Wheeler (for the 375 lowing people: Michael Burrows and David Wheeler (for the
363 block sorting transformation), David Wheeler (again, for 376 block sorting transformation), David Wheeler (again, for
364 the Huffman coder), Peter Fenwick (for the structured cod- 377 the Huffman coder), Peter Fenwick (for the structured cod­
365 ing model in the original bzip, and many refinements), and 378 ing model in the original bzip, and many refinements), and
366 Alistair Moffat, Radford Neal and Ian Witten (for the 379 Alistair Moffat, Radford Neal and Ian Witten (for the
367 arithmetic coder in the original bzip). I am much 380 arithmetic coder in the original bzip). I am much
368 indebted for their help, support and advice. See the man- 381 indebted for their help, support and advice. See the man­
369 ual in the source distribution for pointers to sources of 382 ual in the source distribution for pointers to sources of
370 documentation. Christian von Roques encouraged me to look 383 documentation. Christian von Roques encouraged me to look
371 for faster sorting algorithms, so as to speed up compres- 384 for faster sorting algorithms, so as to speed up compres­
372 sion. Bela Lubkin encouraged me to improve the worst-case 385 sion. Bela Lubkin encouraged me to improve the worst-case
373 compression performance. Many people sent patches, helped 386 compression performance. The bz* scripts are derived from
374 with portability problems, lent machines, gave advice and 387 those of GNU gzip. Many people sent patches, helped with
375 were generally helpful. 388 portability problems, lent machines, gave advice and were
389 generally helpful.
376 390
diff --git a/bzip2recover.c b/bzip2recover.c
index ba3d175..286873b 100644
--- a/bzip2recover.c
+++ b/bzip2recover.c
@@ -9,7 +9,7 @@
9 salvage from damaged files created by the accompanying 9 salvage from damaged files created by the accompanying
10 bzip2-1.0 program. 10 bzip2-1.0 program.
11 11
12 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 12 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
13 13
14 Redistribution and use in source and binary forms, with or without 14 Redistribution and use in source and binary forms, with or without
15 modification, are permitted provided that the following conditions 15 modification, are permitted provided that the following conditions
@@ -57,6 +57,29 @@
57#include <stdlib.h> 57#include <stdlib.h>
58#include <string.h> 58#include <string.h>
59 59
60
61/* This program records bit locations in the file to be recovered.
62 That means that if 64-bit ints are not supported, we will not
63 be able to recover .bz2 files over 512MB (2^32 bits) long.
64 On GNU supported platforms, we take advantage of the 64-bit
65 int support to circumvent this problem. Ditto MSVC.
66
67 This change occurred in version 1.0.2; all prior versions have
68 the 512MB limitation.
69*/
70#ifdef __GNUC__
71 typedef unsigned long long int MaybeUInt64;
72# define MaybeUInt64_FMT "%Lu"
73#else
74#ifdef _MSC_VER
75 typedef unsigned __int64 MaybeUInt64;
76# define MaybeUInt64_FMT "%I64u"
77#else
78 typedef unsigned int MaybeUInt64;
79# define MaybeUInt64_FMT "%u"
80#endif
81#endif
82
60typedef unsigned int UInt32; 83typedef unsigned int UInt32;
61typedef int Int32; 84typedef int Int32;
62typedef unsigned char UChar; 85typedef unsigned char UChar;
@@ -66,13 +89,25 @@ typedef unsigned char Bool;
66#define False ((Bool)0) 89#define False ((Bool)0)
67 90
68 91
69Char inFileName[2000]; 92#define BZ_MAX_FILENAME 2000
70Char outFileName[2000]; 93
71Char progName[2000]; 94Char inFileName[BZ_MAX_FILENAME];
95Char outFileName[BZ_MAX_FILENAME];
96Char progName[BZ_MAX_FILENAME];
97
98MaybeUInt64 bytesOut = 0;
99MaybeUInt64 bytesIn = 0;
72 100
73UInt32 bytesOut = 0;
74UInt32 bytesIn = 0;
75 101
102/*---------------------------------------------------*/
103/*--- Header bytes ---*/
104/*---------------------------------------------------*/
105
106#define BZ_HDR_B 0x42 /* 'B' */
107#define BZ_HDR_Z 0x5a /* 'Z' */
108#define BZ_HDR_h 0x68 /* 'h' */
109#define BZ_HDR_0 0x30 /* '0' */
110
76 111
77/*---------------------------------------------------*/ 112/*---------------------------------------------------*/
78/*--- I/O errors ---*/ 113/*--- I/O errors ---*/
@@ -116,6 +151,23 @@ void mallocFail ( Int32 n )
116} 151}
117 152
118 153
154/*---------------------------------------------*/
155void tooManyBlocks ( Int32 max_handled_blocks )
156{
157 fprintf ( stderr,
158 "%s: `%s' appears to contain more than %d blocks\n",
159 progName, inFileName, max_handled_blocks );
160 fprintf ( stderr,
161 "%s: and cannot be handled. To fix, increase\n",
162 progName );
163 fprintf ( stderr,
164 "%s: BZ_MAX_HANDLED_BLOCKS in bzip2recover.c, and recompile.\n",
165 progName );
166 exit ( 1 );
167}
168
169
170
119/*---------------------------------------------------*/ 171/*---------------------------------------------------*/
120/*--- Bit stream I/O ---*/ 172/*--- Bit stream I/O ---*/
121/*---------------------------------------------------*/ 173/*---------------------------------------------------*/
@@ -254,27 +306,37 @@ Bool endsInBz2 ( Char* name )
254/*--- ---*/ 306/*--- ---*/
255/*---------------------------------------------------*/ 307/*---------------------------------------------------*/
256 308
309/* This logic isn't really right when it comes to Cygwin. */
310#ifdef _WIN32
311# define BZ_SPLIT_SYM '\\' /* path splitter on Windows platform */
312#else
313# define BZ_SPLIT_SYM '/' /* path splitter on Unix platform */
314#endif
315
257#define BLOCK_HEADER_HI 0x00003141UL 316#define BLOCK_HEADER_HI 0x00003141UL
258#define BLOCK_HEADER_LO 0x59265359UL 317#define BLOCK_HEADER_LO 0x59265359UL
259 318
260#define BLOCK_ENDMARK_HI 0x00001772UL 319#define BLOCK_ENDMARK_HI 0x00001772UL
261#define BLOCK_ENDMARK_LO 0x45385090UL 320#define BLOCK_ENDMARK_LO 0x45385090UL
262 321
322/* Increase if necessary. However, a .bz2 file with > 50000 blocks
323 would have an uncompressed size of at least 40GB, so the chances
324 are low you'll need to up this.
325*/
326#define BZ_MAX_HANDLED_BLOCKS 50000
263 327
264UInt32 bStart[20000]; 328MaybeUInt64 bStart [BZ_MAX_HANDLED_BLOCKS];
265UInt32 bEnd[20000]; 329MaybeUInt64 bEnd [BZ_MAX_HANDLED_BLOCKS];
266UInt32 rbStart[20000]; 330MaybeUInt64 rbStart[BZ_MAX_HANDLED_BLOCKS];
267UInt32 rbEnd[20000]; 331MaybeUInt64 rbEnd [BZ_MAX_HANDLED_BLOCKS];
268 332
269Int32 main ( Int32 argc, Char** argv ) 333Int32 main ( Int32 argc, Char** argv )
270{ 334{
271 FILE* inFile; 335 FILE* inFile;
272 FILE* outFile; 336 FILE* outFile;
273 BitStream* bsIn, *bsWr; 337 BitStream* bsIn, *bsWr;
274 Int32 currBlock, b, wrBlock; 338 Int32 b, wrBlock, currBlock, rbCtr;
275 UInt32 bitsRead; 339 MaybeUInt64 bitsRead;
276 Int32 rbCtr;
277
278 340
279 UInt32 buffHi, buffLo, blockCRC; 341 UInt32 buffHi, buffLo, blockCRC;
280 Char* p; 342 Char* p;
@@ -282,11 +344,37 @@ Int32 main ( Int32 argc, Char** argv )
282 strcpy ( progName, argv[0] ); 344 strcpy ( progName, argv[0] );
283 inFileName[0] = outFileName[0] = 0; 345 inFileName[0] = outFileName[0] = 0;
284 346
285 fprintf ( stderr, "bzip2recover 1.0: extracts blocks from damaged .bz2 files.\n" ); 347 fprintf ( stderr,
348 "bzip2recover 1.0.2: extracts blocks from damaged .bz2 files.\n" );
286 349
287 if (argc != 2) { 350 if (argc != 2) {
288 fprintf ( stderr, "%s: usage is `%s damaged_file_name'.\n", 351 fprintf ( stderr, "%s: usage is `%s damaged_file_name'.\n",
289 progName, progName ); 352 progName, progName );
353 switch (sizeof(MaybeUInt64)) {
354 case 8:
355 fprintf(stderr,
356 "\trestrictions on size of recovered file: None\n");
357 break;
358 case 4:
359 fprintf(stderr,
360 "\trestrictions on size of recovered file: 512 MB\n");
361 fprintf(stderr,
362 "\tto circumvent, recompile with MaybeUInt64 as an\n"
363 "\tunsigned 64-bit int.\n");
364 break;
365 default:
366 fprintf(stderr,
367 "\tsizeof(MaybeUInt64) is not 4 or 8 -- "
368 "configuration error.\n");
369 break;
370 }
371 exit(1);
372 }
373
374 if (strlen(argv[1]) >= BZ_MAX_FILENAME-20) {
375 fprintf ( stderr,
376 "%s: supplied filename is suspiciously (>= %d chars) long. Bye!\n",
377 progName, strlen(argv[1]) );
290 exit(1); 378 exit(1);
291 } 379 }
292 380
@@ -316,7 +404,8 @@ Int32 main ( Int32 argc, Char** argv )
316 (bitsRead - bStart[currBlock]) >= 40) { 404 (bitsRead - bStart[currBlock]) >= 40) {
317 bEnd[currBlock] = bitsRead-1; 405 bEnd[currBlock] = bitsRead-1;
318 if (currBlock > 0) 406 if (currBlock > 0)
319 fprintf ( stderr, " block %d runs from %d to %d (incomplete)\n", 407 fprintf ( stderr, " block %d runs from " MaybeUInt64_FMT
408 " to " MaybeUInt64_FMT " (incomplete)\n",
320 currBlock, bStart[currBlock], bEnd[currBlock] ); 409 currBlock, bStart[currBlock], bEnd[currBlock] );
321 } else 410 } else
322 currBlock--; 411 currBlock--;
@@ -330,17 +419,22 @@ Int32 main ( Int32 argc, Char** argv )
330 ( (buffHi & 0x0000ffff) == BLOCK_ENDMARK_HI 419 ( (buffHi & 0x0000ffff) == BLOCK_ENDMARK_HI
331 && buffLo == BLOCK_ENDMARK_LO) 420 && buffLo == BLOCK_ENDMARK_LO)
332 ) { 421 ) {
333 if (bitsRead > 49) 422 if (bitsRead > 49) {
334 bEnd[currBlock] = bitsRead-49; else 423 bEnd[currBlock] = bitsRead-49;
424 } else {
335 bEnd[currBlock] = 0; 425 bEnd[currBlock] = 0;
426 }
336 if (currBlock > 0 && 427 if (currBlock > 0 &&
337 (bEnd[currBlock] - bStart[currBlock]) >= 130) { 428 (bEnd[currBlock] - bStart[currBlock]) >= 130) {
338 fprintf ( stderr, " block %d runs from %d to %d\n", 429 fprintf ( stderr, " block %d runs from " MaybeUInt64_FMT
430 " to " MaybeUInt64_FMT "\n",
339 rbCtr+1, bStart[currBlock], bEnd[currBlock] ); 431 rbCtr+1, bStart[currBlock], bEnd[currBlock] );
340 rbStart[rbCtr] = bStart[currBlock]; 432 rbStart[rbCtr] = bStart[currBlock];
341 rbEnd[rbCtr] = bEnd[currBlock]; 433 rbEnd[rbCtr] = bEnd[currBlock];
342 rbCtr++; 434 rbCtr++;
343 } 435 }
436 if (currBlock >= BZ_MAX_HANDLED_BLOCKS)
437 tooManyBlocks(BZ_MAX_HANDLED_BLOCKS);
344 currBlock++; 438 currBlock++;
345 439
346 bStart[currBlock] = bitsRead; 440 bStart[currBlock] = bitsRead;
@@ -400,10 +494,25 @@ Int32 main ( Int32 argc, Char** argv )
400 wrBlock++; 494 wrBlock++;
401 } else 495 } else
402 if (bitsRead == rbStart[wrBlock]) { 496 if (bitsRead == rbStart[wrBlock]) {
403 outFileName[0] = 0; 497 /* Create the output file name, correctly handling leading paths.
404 sprintf ( outFileName, "rec%4d", wrBlock+1 ); 498 (31.10.2001 by Sergey E. Kusikov) */
405 for (p = outFileName; *p != 0; p++) if (*p == ' ') *p = '0'; 499 Char* split;
406 strcat ( outFileName, inFileName ); 500 Int32 ofs, k;
501 for (k = 0; k < BZ_MAX_FILENAME; k++)
502 outFileName[k] = 0;
503 strcpy (outFileName, inFileName);
504 split = strrchr (outFileName, BZ_SPLIT_SYM);
505 if (split == NULL) {
506 split = outFileName;
507 } else {
508 ++split;
509 }
510 /* Now split points to the start of the basename. */
511 ofs = split - outFileName;
512 sprintf (split, "rec%5d", wrBlock+1);
513 for (p = split; *p != 0; p++) if (*p == ' ') *p = '0';
514 strcat (outFileName, inFileName + ofs);
515
407 if ( !endsInBz2(outFileName)) strcat ( outFileName, ".bz2" ); 516 if ( !endsInBz2(outFileName)) strcat ( outFileName, ".bz2" );
408 517
409 fprintf ( stderr, " writing block %d to `%s' ...\n", 518 fprintf ( stderr, " writing block %d to `%s' ...\n",
@@ -416,8 +525,10 @@ Int32 main ( Int32 argc, Char** argv )
416 exit(1); 525 exit(1);
417 } 526 }
418 bsWr = bsOpenWriteStream ( outFile ); 527 bsWr = bsOpenWriteStream ( outFile );
419 bsPutUChar ( bsWr, 'B' ); bsPutUChar ( bsWr, 'Z' ); 528 bsPutUChar ( bsWr, BZ_HDR_B );
420 bsPutUChar ( bsWr, 'h' ); bsPutUChar ( bsWr, '9' ); 529 bsPutUChar ( bsWr, BZ_HDR_Z );
530 bsPutUChar ( bsWr, BZ_HDR_h );
531 bsPutUChar ( bsWr, BZ_HDR_0 + 9 );
421 bsPutUChar ( bsWr, 0x31 ); bsPutUChar ( bsWr, 0x41 ); 532 bsPutUChar ( bsWr, 0x31 ); bsPutUChar ( bsWr, 0x41 );
422 bsPutUChar ( bsWr, 0x59 ); bsPutUChar ( bsWr, 0x26 ); 533 bsPutUChar ( bsWr, 0x59 ); bsPutUChar ( bsWr, 0x26 );
423 bsPutUChar ( bsWr, 0x53 ); bsPutUChar ( bsWr, 0x59 ); 534 bsPutUChar ( bsWr, 0x53 ); bsPutUChar ( bsWr, 0x59 );
diff --git a/bzlib.c b/bzlib.c
index 4a06d9f..7d1cb27 100644
--- a/bzlib.c
+++ b/bzlib.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
@@ -93,10 +93,39 @@ void BZ2_bz__AssertH__fail ( int errcode )
93 "component, you should also report this bug to the author(s)\n" 93 "component, you should also report this bug to the author(s)\n"
94 "of that program. Please make an effort to report this bug;\n" 94 "of that program. Please make an effort to report this bug;\n"
95 "timely and accurate bug reports eventually lead to higher\n" 95 "timely and accurate bug reports eventually lead to higher\n"
96 "quality software. Thanks. Julian Seward, 21 March 2000.\n\n", 96 "quality software. Thanks. Julian Seward, 30 December 2001.\n\n",
97 errcode, 97 errcode,
98 BZ2_bzlibVersion() 98 BZ2_bzlibVersion()
99 ); 99 );
100
101 if (errcode == 1007) {
102 fprintf(stderr,
103 "\n*** A special note about internal error number 1007 ***\n"
104 "\n"
105 "Experience suggests that a common cause of i.e. 1007\n"
106 "is unreliable memory or other hardware. The 1007 assertion\n"
107 "just happens to cross-check the results of huge numbers of\n"
108 "memory reads/writes, and so acts (unintendedly) as a stress\n"
109 "test of your memory system.\n"
110 "\n"
111 "I suggest the following: try compressing the file again,\n"
112 "possibly monitoring progress in detail with the -vv flag.\n"
113 "\n"
114 "* If the error cannot be reproduced, and/or happens at different\n"
115 " points in compression, you may have a flaky memory system.\n"
116 " Try a memory-test program. I have used Memtest86\n"
117 " (www.memtest86.com). At the time of writing it is free (GPLd).\n"
118 " Memtest86 tests memory much more thorougly than your BIOSs\n"
119 " power-on test, and may find failures that the BIOS doesn't.\n"
120 "\n"
121 "* If the error can be repeatably reproduced, this is a bug in\n"
122 " bzip2, and I would very much like to hear about it. Please\n"
123 " let me know, and, ideally, save a copy of the file causing the\n"
124 " problem -- without which I will be unable to investigate it.\n"
125 "\n"
126 );
127 }
128
100 exit(3); 129 exit(3);
101} 130}
102#endif 131#endif
@@ -1402,7 +1431,7 @@ BZFILE * bzopen_or_bzdopen
1402 smallMode = 1; break; 1431 smallMode = 1; break;
1403 default: 1432 default:
1404 if (isdigit((int)(*mode))) { 1433 if (isdigit((int)(*mode))) {
1405 blockSize100k = *mode-'0'; 1434 blockSize100k = *mode-BZ_HDR_0;
1406 } 1435 }
1407 } 1436 }
1408 mode++; 1437 mode++;
diff --git a/bzlib.h b/bzlib.h
index c9447a2..9ac43a1 100644
--- a/bzlib.h
+++ b/bzlib.h
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
@@ -110,8 +110,10 @@ typedef
110#define BZ_EXPORT 110#define BZ_EXPORT
111#endif 111#endif
112 112
113/* Need a definitition for FILE */
114#include <stdio.h>
115
113#ifdef _WIN32 116#ifdef _WIN32
114# include <stdio.h>
115# include <windows.h> 117# include <windows.h>
116# ifdef small 118# ifdef small
117 /* windows.h define small to char */ 119 /* windows.h define small to char */
diff --git a/bzlib_private.h b/bzlib_private.h
index fb51c7a..ff973c3 100644
--- a/bzlib_private.h
+++ b/bzlib_private.h
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
@@ -76,7 +76,7 @@
76 76
77/*-- General stuff. --*/ 77/*-- General stuff. --*/
78 78
79#define BZ_VERSION "1.0.1, 23-June-2000" 79#define BZ_VERSION "1.0.2, 30-Dec-2001"
80 80
81typedef char Char; 81typedef char Char;
82typedef unsigned char Bool; 82typedef unsigned char Bool;
@@ -137,6 +137,13 @@ extern void bz_internal_error ( int errcode );
137#define BZFREE(ppp) (strm->bzfree)(strm->opaque,(ppp)) 137#define BZFREE(ppp) (strm->bzfree)(strm->opaque,(ppp))
138 138
139 139
140/*-- Header bytes. --*/
141
142#define BZ_HDR_B 0x42 /* 'B' */
143#define BZ_HDR_Z 0x5a /* 'Z' */
144#define BZ_HDR_h 0x68 /* 'h' */
145#define BZ_HDR_0 0x30 /* '0' */
146
140/*-- Constants for the back end. --*/ 147/*-- Constants for the back end. --*/
141 148
142#define BZ_MAX_ALPHA_SIZE 258 149#define BZ_MAX_ALPHA_SIZE 258
diff --git a/bzmore b/bzmore
new file mode 100644
index 0000000..d314043
--- /dev/null
+++ b/bzmore
@@ -0,0 +1,61 @@
1#!/bin/sh
2
3# Bzmore wrapped for bzip2,
4# adapted from zmore by Philippe Troin <phil@fifi.org> for Debian GNU/Linux.
5
6PATH="/usr/bin:$PATH"; export PATH
7
8prog=`echo $0 | sed 's|.*/||'`
9case "$prog" in
10 *less) more=less ;;
11 *) more=more ;;
12esac
13
14if test "`echo -n a`" = "-n a"; then
15 # looks like a SysV system:
16 n1=''; n2='\c'
17else
18 n1='-n'; n2=''
19fi
20oldtty=`stty -g 2>/dev/null`
21if stty -cbreak 2>/dev/null; then
22 cb='cbreak'; ncb='-cbreak'
23else
24 # 'stty min 1' resets eof to ^a on both SunOS and SysV!
25 cb='min 1 -icanon'; ncb='icanon eof ^d'
26fi
27if test $? -eq 0 -a -n "$oldtty"; then
28 trap 'stty $oldtty 2>/dev/null; exit' 0 2 3 5 10 13 15
29else
30 trap 'stty $ncb echo 2>/dev/null; exit' 0 2 3 5 10 13 15
31fi
32
33if test $# = 0; then
34 if test -t 0; then
35 echo usage: $prog files...
36 else
37 bzip2 -cdfq | eval $more
38 fi
39else
40 FIRST=1
41 for FILE
42 do
43 if test $FIRST -eq 0; then
44 echo $n1 "--More--(Next file: $FILE)$n2"
45 stty $cb -echo 2>/dev/null
46 ANS=`dd bs=1 count=1 2>/dev/null`
47 stty $ncb echo 2>/dev/null
48 echo " "
49 if test "$ANS" = 'e' -o "$ANS" = 'q'; then
50 exit
51 fi
52 fi
53 if test "$ANS" != 's'; then
54 echo "------> $FILE <------"
55 bzip2 -cdfq "$FILE" | eval $more
56 fi
57 if test -t; then
58 FIRST=0
59 fi
60 done
61fi
diff --git a/bzmore.1 b/bzmore.1
new file mode 100644
index 0000000..b437d3b
--- /dev/null
+++ b/bzmore.1
@@ -0,0 +1,152 @@
1.\"Shamelessly copied from zmore.1 by Philippe Troin <phil@fifi.org>
2.\"for Debian GNU/Linux
3.TH BZMORE 1
4.SH NAME
5bzmore, bzless \- file perusal filter for crt viewing of bzip2 compressed text
6.SH SYNOPSIS
7.B bzmore
8[ name ... ]
9.br
10.B bzless
11[ name ... ]
12.SH NOTE
13In the following description,
14.I bzless
15and
16.I less
17can be used interchangeably with
18.I bzmore
19and
20.I more.
21.SH DESCRIPTION
22.I Bzmore
23is a filter which allows examination of compressed or plain text files
24one screenful at a time on a soft-copy terminal.
25.I bzmore
26works on files compressed with
27.I bzip2
28and also on uncompressed files.
29If a file does not exist,
30.I bzmore
31looks for a file of the same name with the addition of a .bz2 suffix.
32.PP
33.I Bzmore
34normally pauses after each screenful, printing --More--
35at the bottom of the screen.
36If the user then types a carriage return, one more line is displayed.
37If the user hits a space,
38another screenful is displayed. Other possibilities are enumerated later.
39.PP
40.I Bzmore
41looks in the file
42.I /etc/termcap
43to determine terminal characteristics,
44and to determine the default window size.
45On a terminal capable of displaying 24 lines,
46the default window size is 22 lines.
47Other sequences which may be typed when
48.I bzmore
49pauses, and their effects, are as follows (\fIi\fP is an optional integer
50argument, defaulting to 1) :
51.PP
52.IP \fIi\|\fP<space>
53display
54.I i
55more lines, (or another screenful if no argument is given)
56.PP
57.IP ^D
58display 11 more lines (a ``scroll'').
59If
60.I i
61is given, then the scroll size is set to \fIi\|\fP.
62.PP
63.IP d
64same as ^D (control-D)
65.PP
66.IP \fIi\|\fPz
67same as typing a space except that \fIi\|\fP, if present, becomes the new
68window size. Note that the window size reverts back to the default at the
69end of the current file.
70.PP
71.IP \fIi\|\fPs
72skip \fIi\|\fP lines and print a screenful of lines
73.PP
74.IP \fIi\|\fPf
75skip \fIi\fP screenfuls and print a screenful of lines
76.PP
77.IP "q or Q"
78quit reading the current file; go on to the next (if any)
79.PP
80.IP "e or q"
81When the prompt --More--(Next file:
82.IR file )
83is printed, this command causes bzmore to exit.
84.PP
85.IP s
86When the prompt --More--(Next file:
87.IR file )
88is printed, this command causes bzmore to skip the next file and continue.
89.PP
90.IP =
91Display the current line number.
92.PP
93.IP \fIi\|\fP/expr
94search for the \fIi\|\fP-th occurrence of the regular expression \fIexpr.\fP
95If the pattern is not found,
96.I bzmore
97goes on to the next file (if any).
98Otherwise, a screenful is displayed, starting two lines before the place
99where the expression was found.
100The user's erase and kill characters may be used to edit the regular
101expression.
102Erasing back past the first column cancels the search command.
103.PP
104.IP \fIi\|\fPn
105search for the \fIi\|\fP-th occurrence of the last regular expression entered.
106.PP
107.IP !command
108invoke a shell with \fIcommand\|\fP.
109The character `!' in "command" are replaced with the
110previous shell command. The sequence "\\!" is replaced by "!".
111.PP
112.IP ":q or :Q"
113quit reading the current file; go on to the next (if any)
114(same as q or Q).
115.PP
116.IP .
117(dot) repeat the previous command.
118.PP
119The commands take effect immediately, i.e., it is not necessary to
120type a carriage return.
121Up to the time when the command character itself is given,
122the user may hit the line kill character to cancel the numerical
123argument being formed.
124In addition, the user may hit the erase character to redisplay the
125--More-- message.
126.PP
127At any time when output is being sent to the terminal, the user can
128hit the quit key (normally control\-\\).
129.I Bzmore
130will stop sending output, and will display the usual --More--
131prompt.
132The user may then enter one of the above commands in the normal manner.
133Unfortunately, some output is lost when this is done, due to the
134fact that any characters waiting in the terminal's output queue
135are flushed when the quit signal occurs.
136.PP
137The terminal is set to
138.I noecho
139mode by this program so that the output can be continuous.
140What you type will thus not show on your terminal, except for the / and !
141commands.
142.PP
143If the standard output is not a teletype, then
144.I bzmore
145acts just like
146.I bzcat,
147except that a header is printed before each file.
148.SH FILES
149.DT
150/etc/termcap Terminal data base
151.SH "SEE ALSO"
152more(1), less(1), bzip2(1), bzdiff(1), bzgrep(1)
diff --git a/compress.c b/compress.c
index cc5e31d..56501c1 100644
--- a/compress.c
+++ b/compress.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
@@ -663,10 +663,10 @@ void BZ2_compressBlock ( EState* s, Bool is_last_block )
663 /*-- If this is the first block, create the stream header. --*/ 663 /*-- If this is the first block, create the stream header. --*/
664 if (s->blockNo == 1) { 664 if (s->blockNo == 1) {
665 BZ2_bsInitWrite ( s ); 665 BZ2_bsInitWrite ( s );
666 bsPutUChar ( s, 'B' ); 666 bsPutUChar ( s, BZ_HDR_B );
667 bsPutUChar ( s, 'Z' ); 667 bsPutUChar ( s, BZ_HDR_Z );
668 bsPutUChar ( s, 'h' ); 668 bsPutUChar ( s, BZ_HDR_h );
669 bsPutUChar ( s, (UChar)('0' + s->blockSize100k) ); 669 bsPutUChar ( s, (UChar)(BZ_HDR_0 + s->blockSize100k) );
670 } 670 }
671 671
672 if (s->nblock > 0) { 672 if (s->nblock > 0) {
diff --git a/crctable.c b/crctable.c
index 61c040c..b16746a 100644
--- a/crctable.c
+++ b/crctable.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
diff --git a/decompress.c b/decompress.c
index cdced18..e921347 100644
--- a/decompress.c
+++ b/decompress.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
@@ -235,18 +235,18 @@ Int32 BZ2_decompress ( DState* s )
235 switch (s->state) { 235 switch (s->state) {
236 236
237 GET_UCHAR(BZ_X_MAGIC_1, uc); 237 GET_UCHAR(BZ_X_MAGIC_1, uc);
238 if (uc != 'B') RETURN(BZ_DATA_ERROR_MAGIC); 238 if (uc != BZ_HDR_B) RETURN(BZ_DATA_ERROR_MAGIC);
239 239
240 GET_UCHAR(BZ_X_MAGIC_2, uc); 240 GET_UCHAR(BZ_X_MAGIC_2, uc);
241 if (uc != 'Z') RETURN(BZ_DATA_ERROR_MAGIC); 241 if (uc != BZ_HDR_Z) RETURN(BZ_DATA_ERROR_MAGIC);
242 242
243 GET_UCHAR(BZ_X_MAGIC_3, uc) 243 GET_UCHAR(BZ_X_MAGIC_3, uc)
244 if (uc != 'h') RETURN(BZ_DATA_ERROR_MAGIC); 244 if (uc != BZ_HDR_h) RETURN(BZ_DATA_ERROR_MAGIC);
245 245
246 GET_BITS(BZ_X_MAGIC_4, s->blockSize100k, 8) 246 GET_BITS(BZ_X_MAGIC_4, s->blockSize100k, 8)
247 if (s->blockSize100k < '1' || 247 if (s->blockSize100k < (BZ_HDR_0 + 1) ||
248 s->blockSize100k > '9') RETURN(BZ_DATA_ERROR_MAGIC); 248 s->blockSize100k > (BZ_HDR_0 + 9)) RETURN(BZ_DATA_ERROR_MAGIC);
249 s->blockSize100k -= '0'; 249 s->blockSize100k -= BZ_HDR_0;
250 250
251 if (s->smallDecompress) { 251 if (s->smallDecompress) {
252 s->ll16 = BZALLOC( s->blockSize100k * 100000 * sizeof(UInt16) ); 252 s->ll16 = BZALLOC( s->blockSize100k * 100000 * sizeof(UInt16) );
diff --git a/dlltest.c b/dlltest.c
index f79279c..2d7dcca 100644
--- a/dlltest.c
+++ b/dlltest.c
@@ -19,7 +19,7 @@
19 19
20#ifdef _WIN32 20#ifdef _WIN32
21 21
22#define BZ2_LIBNAME "libbz2-1.0.0.DLL" 22#define BZ2_LIBNAME "libbz2-1.0.2.DLL"
23 23
24#include <windows.h> 24#include <windows.h>
25static int BZ2DLLLoaded = 0; 25static int BZ2DLLLoaded = 0;
@@ -130,8 +130,8 @@ int main(int argc,char *argv[])
130 }else{ 130 }else{
131 fp_w = stdout; 131 fp_w = stdout;
132 } 132 }
133 if((BZ2fp_r == NULL && (BZ2fp_r = BZ2_bzdopen(fileno(stdin),"rb"))==NULL) 133 if((fn_r == NULL && (BZ2fp_r = BZ2_bzdopen(fileno(stdin),"rb"))==NULL)
134 || (BZ2fp_r != NULL && (BZ2fp_r = BZ2_bzopen(fn_r,"rb"))==NULL)){ 134 || (fn_r != NULL && (BZ2fp_r = BZ2_bzopen(fn_r,"rb"))==NULL)){
135 printf("can't bz2openstream\n"); 135 printf("can't bz2openstream\n");
136 exit(1); 136 exit(1);
137 } 137 }
diff --git a/huffman.c b/huffman.c
index 9b446c4..293095c 100644
--- a/huffman.c
+++ b/huffman.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
diff --git a/makefile.msc b/makefile.msc
index 3fe4232..799a18a 100644
--- a/makefile.msc
+++ b/makefile.msc
@@ -4,7 +4,7 @@
4# Fixed up by JRS for bzip2-0.9.5d release. 4# Fixed up by JRS for bzip2-0.9.5d release.
5 5
6CC=cl 6CC=cl
7CFLAGS= -DWIN32 -MD -Ox -D_FILE_OFFSET_BITS=64 7CFLAGS= -DWIN32 -MD -Ox -D_FILE_OFFSET_BITS=64 -nologo
8 8
9OBJS= blocksort.obj \ 9OBJS= blocksort.obj \
10 huffman.obj \ 10 huffman.obj \
diff --git a/manual.texi b/manual.texi
index 336776a..5bc27d5 100644
--- a/manual.texi
+++ b/manual.texi
@@ -2,10 +2,10 @@
2@setfilename bzip2.info 2@setfilename bzip2.info
3 3
4@ignore 4@ignore
5This file documents bzip2 version 1.0, and associated library 5This file documents bzip2 version 1.0.2, and associated library
6libbzip2, written by Julian Seward (jseward@acm.org). 6libbzip2, written by Julian Seward (jseward@acm.org).
7 7
8Copyright (C) 1996-2000 Julian R Seward 8Copyright (C) 1996-2002 Julian R Seward
9 9
10Permission is granted to make and distribute verbatim copies of 10Permission is granted to make and distribute verbatim copies of
11this manual provided the copyright notice and this permission notice 11this manual provided the copyright notice and this permission notice
@@ -30,8 +30,8 @@ END-INFO-DIR-ENTRY
30@titlepage 30@titlepage
31@title bzip2 and libbzip2 31@title bzip2 and libbzip2
32@subtitle a program and library for data compression 32@subtitle a program and library for data compression
33@subtitle copyright (C) 1996-2000 Julian Seward 33@subtitle copyright (C) 1996-2002 Julian Seward
34@subtitle version 1.0 of 21 March 2000 34@subtitle version 1.0.2 of 30 December 2001
35@author Julian Seward 35@author Julian Seward
36 36
37@end titlepage 37@end titlepage
@@ -40,11 +40,17 @@ END-INFO-DIR-ENTRY
40@parskip 2mm 40@parskip 2mm
41 41
42@end iftex 42@end iftex
43@node Top, Overview, (dir), (dir) 43@node Top,,, (dir)
44
45The following text is the License for this software. You should
46find it identical to that contained in the file LICENSE in the
47source distribution.
48
49@bf{------------------ START OF THE LICENSE ------------------}
44 50
45This program, @code{bzip2}, 51This program, @code{bzip2},
46and associated library @code{libbzip2}, are 52and associated library @code{libbzip2}, are
47Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 53Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
48 54
49Redistribution and use in source and binary forms, with or without 55Redistribution and use in source and binary forms, with or without
50modification, are permitted provided that the following conditions 56modification, are permitted provided that the following conditions
@@ -82,13 +88,15 @@ Julian Seward, Cambridge, UK.
82 88
83@code{jseward@@acm.org} 89@code{jseward@@acm.org}
84 90
85@code{http://sourceware.cygnus.com/bzip2} 91@code{bzip2}/@code{libbzip2} version 1.0.2 of 30 December 2001.
86 92
87@code{http://www.cacheprof.org} 93@bf{------------------ END OF THE LICENSE ------------------}
88 94
89@code{http://www.muraroa.demon.co.uk} 95Web sites:
90 96
91@code{bzip2}/@code{libbzip2} version 1.0 of 21 March 2000. 97@code{http://sources.redhat.com/bzip2}
98
99@code{http://www.cacheprof.org}
92 100
93PATENTS: To the best of my knowledge, @code{bzip2} does not use any patented 101PATENTS: To the best of my knowledge, @code{bzip2} does not use any patented
94algorithms. However, I do not have the resources available to carry out 102algorithms. However, I do not have the resources available to carry out
@@ -101,7 +109,6 @@ above statement.
101 109
102 110
103 111
104@node Overview, Implementation, Top, Top
105@chapter Introduction 112@chapter Introduction
106 113
107@code{bzip2} compresses files using the Burrows-Wheeler 114@code{bzip2} compresses files using the Burrows-Wheeler
@@ -134,7 +141,7 @@ and nothing else.
134@unnumberedsubsubsec NAME 141@unnumberedsubsubsec NAME
135@itemize 142@itemize
136@item @code{bzip2}, @code{bunzip2} 143@item @code{bzip2}, @code{bunzip2}
137- a block-sorting file compressor, v1.0 144- a block-sorting file compressor, v1.0.2
138@item @code{bzcat} 145@item @code{bzcat}
139- decompresses files to stdout 146- decompresses files to stdout
140@item @code{bzip2recover} 147@item @code{bzip2recover}
@@ -264,6 +271,11 @@ This really performs a trial decompression and throws away the result.
264Force overwrite of output files. Normally, @code{bzip2} will not overwrite 271Force overwrite of output files. Normally, @code{bzip2} will not overwrite
265existing output files. Also forces @code{bzip2} to break hard links 272existing output files. Also forces @code{bzip2} to break hard links
266to files, which it otherwise wouldn't do. 273to files, which it otherwise wouldn't do.
274
275@code{bzip2} normally declines to decompress files which don't have the
276correct magic header bytes. If forced (@code{-f}), however, it will
277pass such files through unmodified. This is how GNU @code{gzip}
278behaves.
267@item -k --keep 279@item -k --keep
268Keep (don't delete) input files during compression 280Keep (don't delete) input files during compression
269or decompression. 281or decompression.
@@ -286,9 +298,13 @@ Further @code{-v}'s increase the verbosity level, spewing out lots of
286information which is primarily of interest for diagnostic purposes. 298information which is primarily of interest for diagnostic purposes.
287@item -L --license -V --version 299@item -L --license -V --version
288Display the software version, license terms and conditions. 300Display the software version, license terms and conditions.
289@item -1 to -9 301@item -1 (or --fast) to -9 (or --best)
290Set the block size to 100 k, 200 k .. 900 k when compressing. Has no 302Set the block size to 100 k, 200 k .. 900 k when compressing. Has no
291effect when decompressing. See MEMORY MANAGEMENT below. 303effect when decompressing. See MEMORY MANAGEMENT below.
304The @code{--fast} and @code{--best} aliases are primarily for GNU
305@code{gzip} compatibility. In particular, @code{--fast} doesn't make
306things significantly faster. And @code{--best} merely selects the
307default behaviour.
292@item -- 308@item --
293Treats all subsequent arguments as file names, even if they start 309Treats all subsequent arguments as file names, even if they start
294with a dash. This is so you can handle files with names beginning 310with a dash. This is so you can handle files with names beginning
@@ -389,21 +405,19 @@ integrity of the resulting files, and decompress those which are
389undamaged. 405undamaged.
390 406
391@code{bzip2recover} 407@code{bzip2recover}
392takes a single argument, the name of the damaged file, 408takes a single argument, the name of the damaged file, and writes a
393and writes a number of files @code{rec0001file.bz2}, 409number of files @code{rec00001file.bz2}, @code{rec00002file.bz2}, etc,
394 @code{rec0002file.bz2}, etc, containing the extracted blocks. 410containing the extracted blocks. The output filenames are designed so
395 The output filenames are designed so that the use of 411that the use of wildcards in subsequent processing -- for example,
396 wildcards in subsequent processing -- for example, 412@code{bzip2 -dc rec*file.bz2 > recovered_data} -- processes the files in
397@code{bzip2 -dc rec*file.bz2 > recovered_data} -- lists the files in 413the correct order.
398 the correct order.
399 414
400@code{bzip2recover} should be of most use dealing with large @code{.bz2} 415@code{bzip2recover} should be of most use dealing with large @code{.bz2}
401 files, as these will contain many blocks. It is clearly 416files, as these will contain many blocks. It is clearly futile to use
402 futile to use it on damaged single-block files, since a 417it on damaged single-block files, since a damaged block cannot be
403 damaged block cannot be recovered. If you wish to minimise 418recovered. If you wish to minimise any potential data loss through
404any potential data loss through media or transmission errors, 419media or transmission errors, you might consider compressing with a
405you might consider compressing with a smaller 420smaller block size.
406 block size.
407 421
408 422
409@unnumberedsubsubsec PERFORMANCE NOTES 423@unnumberedsubsubsec PERFORMANCE NOTES
@@ -435,22 +449,31 @@ I/O error messages are not as helpful as they could be. @code{bzip2}
435tries hard to detect I/O errors and exit cleanly, but the details of 449tries hard to detect I/O errors and exit cleanly, but the details of
436what the problem is sometimes seem rather misleading. 450what the problem is sometimes seem rather misleading.
437 451
438This manual page pertains to version 1.0 of @code{bzip2}. Compressed 452This manual page pertains to version 1.0.2 of @code{bzip2}. Compressed
439data created by this version is entirely forwards and backwards 453data created by this version is entirely forwards and backwards
440compatible with the previous public releases, versions 0.1pl2, 0.9.0 and 454compatible with the previous public releases, versions 0.1pl2, 0.9.0,
4410.9.5, but with the following exception: 0.9.0 and above can correctly 4550.9.5, 1.0.0 and 1.0.1, but with the following exception: 0.9.0 and
442decompress multiple concatenated compressed files. 0.1pl2 cannot do 456above can correctly decompress multiple concatenated compressed files.
443this; it will stop after decompressing just the first file in the 4570.1pl2 cannot do this; it will stop after decompressing just the first
444stream. 458file in the stream.
459
460@code{bzip2recover} versions prior to this one, 1.0.2, used 32-bit
461integers to represent bit positions in compressed files, so it could not
462handle compressed files more than 512 megabytes long. Version 1.0.2 and
463above uses 64-bit ints on some platforms which support them (GNU
464supported targets, and Windows). To establish whether or not
465@code{bzip2recover} was built with such a limitation, run it without
466arguments. In any event you can build yourself an unlimited version if
467you can recompile it with @code{MaybeUInt64} set to be an unsigned
46864-bit integer.
445 469
446@code{bzip2recover} uses 32-bit integers to represent bit positions in
447compressed files, so it cannot handle compressed files more than 512
448megabytes long. This could easily be fixed.
449 470
450 471
451@unnumberedsubsubsec AUTHOR 472@unnumberedsubsubsec AUTHOR
452Julian Seward, @code{jseward@@acm.org}. 473Julian Seward, @code{jseward@@acm.org}.
453 474
475@code{http://sources.redhat.com/bzip2}
476
454The ideas embodied in @code{bzip2} are due to (at least) the following 477The ideas embodied in @code{bzip2} are due to (at least) the following
455people: Michael Burrows and David Wheeler (for the block sorting 478people: Michael Burrows and David Wheeler (for the block sorting
456transformation), David Wheeler (again, for the Huffman coder), Peter 479transformation), David Wheeler (again, for the Huffman coder), Peter
@@ -461,8 +484,9 @@ indebted for their help, support and advice. See the manual in the
461source distribution for pointers to sources of documentation. Christian 484source distribution for pointers to sources of documentation. Christian
462von Roques encouraged me to look for faster sorting algorithms, so as to 485von Roques encouraged me to look for faster sorting algorithms, so as to
463speed up compression. Bela Lubkin encouraged me to improve the 486speed up compression. Bela Lubkin encouraged me to improve the
464worst-case compression performance. Many people sent patches, helped 487worst-case compression performance. The @code{bz*} scripts are derived
465with portability problems, lent machines, gave advice and were generally 488from those of GNU @code{gzip}. Many people sent patches, helped with
489portability problems, lent machines, gave advice and were generally
466helpful. 490helpful.
467 491
468@end quotation 492@end quotation
@@ -1769,16 +1793,20 @@ was compiled with @code{BZ_NO_STDIO} set.
1769For a normal compile, an assertion failure yields the message 1793For a normal compile, an assertion failure yields the message
1770@example 1794@example
1771 bzip2/libbzip2: internal error number N. 1795 bzip2/libbzip2: internal error number N.
1772 This is a bug in bzip2/libbzip2, 1.0 of 21-Mar-2000. 1796 This is a bug in bzip2/libbzip2, 1.0.2, 30-Dec-2001.
1773 Please report it to me at: jseward@@acm.org. If this happened 1797 Please report it to me at: jseward@@acm.org. If this happened
1774 when you were using some program which uses libbzip2 as a 1798 when you were using some program which uses libbzip2 as a
1775 component, you should also report this bug to the author(s) 1799 component, you should also report this bug to the author(s)
1776 of that program. Please make an effort to report this bug; 1800 of that program. Please make an effort to report this bug;
1777 timely and accurate bug reports eventually lead to higher 1801 timely and accurate bug reports eventually lead to higher
1778 quality software. Thanks. Julian Seward, 21 March 2000. 1802 quality software. Thanks. Julian Seward, 30 December 2001.
1779@end example 1803@end example
1780where @code{N} is some error code number. @code{exit(3)} 1804where @code{N} is some error code number. If @code{N == 1007}, it also
1781is then called. 1805prints some extra text advising the reader that unreliable memory is
1806often associated with internal error 1007. (This is a
1807frequently-observed-phenomenon with versions 1.0.0/1.0.1).
1808
1809@code{exit(3)} is then called.
1782 1810
1783For a @code{stdio}-free library, assertion failures result 1811For a @code{stdio}-free library, assertion failures result
1784in a call to a function declared as: 1812in a call to a function declared as:
@@ -2056,10 +2084,10 @@ Maybe this isn't what you want.
2056If you want a compressor and/or library which is faster, uses less 2084If you want a compressor and/or library which is faster, uses less
2057memory but gets pretty good compression, and has minimal latency, 2085memory but gets pretty good compression, and has minimal latency,
2058consider Jean-loup 2086consider Jean-loup
2059Gailly's and Mark Adler's work, @code{zlib-1.1.2} and 2087Gailly's and Mark Adler's work, @code{zlib-1.1.3} and
2060@code{gzip-1.2.4}. Look for them at 2088@code{gzip-1.2.4}. Look for them at
2061 2089
2062@code{http://www.cdrom.com/pub/infozip/zlib} and 2090@code{http://www.zlib.org} and
2063@code{http://www.gzip.org} respectively. 2091@code{http://www.gzip.org} respectively.
2064 2092
2065For something faster and lighter still, you might try Markus F X J 2093For something faster and lighter still, you might try Markus F X J
diff --git a/mk251.c b/mk251.c
new file mode 100644
index 0000000..205778a
--- /dev/null
+++ b/mk251.c
@@ -0,0 +1,16 @@
1
2/* Spew out a long sequence of the byte 251. When fed to bzip2
3 versions 1.0.0 or 1.0.1, causes it to die with internal error
4 1007 in blocksort.c. This assertion misses an extremely rare
5 case, which is fixed in this version (1.0.2) and above.
6*/
7
8#include <stdio.h>
9
10int main ()
11{
12 int i;
13 for (i = 0; i < 48500000 ; i++)
14 putchar(251);
15 return 0;
16}
diff --git a/randtable.c b/randtable.c
index 983089d..5c922e9 100644
--- a/randtable.c
+++ b/randtable.c
@@ -8,7 +8,7 @@
8 This file is a part of bzip2 and/or libbzip2, a program and 8 This file is a part of bzip2 and/or libbzip2, a program and
9 library for lossless, block-sorting data compression. 9 library for lossless, block-sorting data compression.
10 10
11 Copyright (C) 1996-2000 Julian R Seward. All rights reserved. 11 Copyright (C) 1996-2002 Julian R Seward. All rights reserved.
12 12
13 Redistribution and use in source and binary forms, with or without 13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions 14 modification, are permitted provided that the following conditions
diff --git a/words3 b/words3
index 8486a84..7a6b462 100644
--- a/words3
+++ b/words3
@@ -15,8 +15,8 @@ not actually execute them.
15 15
16Instructions for use are in the preformatted manual page, in the file 16Instructions for use are in the preformatted manual page, in the file
17bzip2.txt. For more detailed documentation, read the full manual. 17bzip2.txt. For more detailed documentation, read the full manual.
18It is available in Postscript form (manual.ps) and HTML form 18It is available in Postscript form (manual.ps), PDF form (manual.pdf),
19(manual_toc.html). 19and HTML form (manual_toc.html).
20 20
21You can also do "bzip2 --help" to see some helpful information. 21You can also do "bzip2 --help" to see some helpful information.
22"bzip2 -L" displays the software license. 22"bzip2 -L" displays the software license.
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-03-15 20:24:18 +0000