1 \input texinfo @c -*-texinfo-*-
10 This manual is for GNU Gzip
11 (version @value{VERSION}, @value{UPDATED}),
12 and documents commands for compressing and decompressing data.
14 Copyright @copyright{} 1998-1999, 2001-2002, 2006-2007, 2009-2013 Free Software
17 Copyright @copyright{} 1992, 1993 Jean-loup Gailly
20 Permission is granted to copy, distribute and/or modify this document
21 under the terms of the GNU Free Documentation License, Version 1.3 or
22 any later version published by the Free Software Foundation; with no
23 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
24 Texts. A copy of the license is included in the section entitled ``GNU
25 Free Documentation License''.
29 @dircategory Compression
31 * Gzip: (gzip). General (de)compression of files (lzw).
34 @dircategory Individual utilities
36 * gunzip: (gzip)Overview. Decompression.
37 * gzexe: (gzip)Overview. Compress executables.
38 * zcat: (gzip)Overview. Decompression to stdout.
39 * zdiff: (gzip)Overview. Compare compressed files.
40 * zforce: (gzip)Overview. Force .gz extension on files.
41 * zgrep: (gzip)Overview. Search compressed files.
42 * zmore: (gzip)Overview. Decompression output by pages.
47 @subtitle The data compression program
48 @subtitle for Gzip version @value{VERSION}
49 @subtitle @value{UPDATED}
50 @author by Jean-loup Gailly
53 @vskip 0pt plus 1filll
61 @top GNU Gzip: General file (de)compression
67 * Overview:: Preliminary information.
68 * Sample:: Sample output from @command{gzip}.
69 * Invoking gzip:: How to run @command{gzip}.
70 * Advanced usage:: Concatenated files.
71 * Environment:: The @env{GZIP} environment variable
72 * Tapes:: Using @command{gzip} on tapes.
73 * Problems:: Reporting bugs.
74 * GNU Free Documentation License:: Copying and sharing this manual.
75 * Concept index:: Index of concepts.
82 @command{gzip} reduces the size of the named files using Lempel-Ziv coding
83 (LZ77). Whenever possible, each file is replaced by one with the
84 extension @samp{.gz}, while keeping the same ownership modes, access and
85 modification times. (The default extension is @samp{-gz} for @abbr{VMS},
86 @samp{z} for @abbr{MSDOS}, @abbr{OS/2} @abbr{FAT} and Atari.)
87 If no files are specified or
88 if a file name is "-", the standard input is compressed to the standard
89 output. @command{gzip} will only attempt to compress regular files. In
90 particular, it will ignore symbolic links.
92 If the new file name is too long for its file system, @command{gzip}
93 truncates it. @command{gzip} attempts to truncate only the parts of the
94 file name longer than 3 characters. (A part is delimited by dots.) If
95 the name consists of small parts only, the longest parts are truncated.
96 For example, if file names are limited to 14 characters, gzip.msdos.exe
97 is compressed to gzi.msd.exe.gz. Names are not truncated on systems
98 which do not have a limit on file name length.
100 By default, @command{gzip} keeps the original file name and time stamp in
101 the compressed file. These are used when decompressing the file with the
102 @option{-N} option. This is useful when the compressed file name was
103 truncated or when the time stamp was not preserved after a file
104 transfer. However, due to limitations in the current @command{gzip} file
105 format, fractional seconds are discarded. Also, time stamps must fall
106 within the range 1970-01-01 00:00:00 through 2106-02-07 06:28:15
107 @abbr{UTC}, and hosts whose operating systems use 32-bit time
108 stamps are further restricted to time stamps no later than 2038-01-19
109 03:14:07 @abbr{UTC}. The upper bounds assume the typical case
110 where leap seconds are ignored.
112 Compressed files can be restored to their original form using @samp{gzip -d}
113 or @command{gunzip} or @command{zcat}. If the original name saved in the
114 compressed file is not suitable for its file system, a new name is
115 constructed from the original one to make it legal.
117 @command{gunzip} takes a list of files on its command line and replaces
118 each file whose name ends with @samp{.gz}, @samp{.z}
119 @samp{-gz}, @samp{-z}, or @samp{_z} (ignoring case)
120 and which begins with the correct
121 magic number with an uncompressed file without the original extension.
122 @command{gunzip} also recognizes the special extensions @samp{.tgz} and
123 @samp{.taz} as shorthands for @samp{.tar.gz} and @samp{.tar.Z}
124 respectively. When compressing, @command{gzip} uses the @samp{.tgz}
125 extension if necessary instead of truncating a file with a @samp{.tar}
128 @command{gunzip} can currently decompress files created by @command{gzip},
129 @command{zip}, @command{compress} or @command{pack}. The detection of the input
130 format is automatic. When using the first two formats, @command{gunzip}
131 checks a 32 bit @abbr{CRC} (cyclic redundancy check). For @command{pack},
132 @command{gunzip} checks the uncompressed length. The @command{compress} format
133 was not designed to allow consistency checks. However @command{gunzip} is
134 sometimes able to detect a bad @samp{.Z} file. If you get an error when
135 uncompressing a @samp{.Z} file, do not assume that the @samp{.Z} file is
136 correct simply because the standard @command{uncompress} does not complain.
137 This generally means that the standard @command{uncompress} does not check
138 its input, and happily generates garbage output. The @abbr{SCO} @samp{compress
139 -H} format (@abbr{LZH} compression method) does not include a @abbr{CRC} but
140 also allows some consistency checks.
142 Files created by @command{zip} can be uncompressed by @command{gzip} only if
143 they have a single member compressed with the 'deflation' method. This
144 feature is only intended to help conversion of @file{tar.zip} files to
145 the @file{tar.gz} format. To extract a @command{zip} file with a single
146 member, use a command like @samp{gunzip <foo.zip} or @samp{gunzip -S
147 .zip foo.zip}. To extract @command{zip} files with several
148 members, use @command{unzip} instead of @command{gunzip}.
150 @command{zcat} is identical to @samp{gunzip -c}. @command{zcat}
151 uncompresses either a list of files on the command line or its standard
152 input and writes the uncompressed data on standard output. @command{zcat}
153 will uncompress files that have the correct magic number whether they
154 have a @samp{.gz} suffix or not.
156 @command{gzip} uses the Lempel-Ziv algorithm used in @command{zip} and
158 The amount of compression obtained depends on the size of the input and
159 the distribution of common substrings. Typically, text such as source
160 code or English is reduced by 60-70%. Compression is generally much
161 better than that achieved by @abbr{LZW} (as used in @command{compress}), Huffman
162 coding (as used in @command{pack}), or adaptive Huffman coding
165 Compression is always performed, even if the compressed file is slightly
166 larger than the original. The worst case expansion is a few bytes for
167 the @command{gzip} file header, plus 5 bytes every 32K block, or an expansion
168 ratio of 0.015% for large files. Note that the actual number of used
169 disk blocks almost never increases. @command{gzip} normally preserves the mode,
170 ownership and time stamps of files when compressing or decompressing.
172 The @command{gzip} file format is specified in P. Deutsch, @sc{gzip} file
173 format specification version 4.3,
174 @uref{http://www.ietf.org/rfc/rfc1952.txt, Internet @abbr{RFC} 1952} (May
175 1996). The @command{zip} deflation format is specified in P. Deutsch,
176 @sc{deflate} Compressed Data Format Specification version 1.3,
177 @uref{http://www.ietf.org/rfc/rfc1951.txt, Internet @abbr{RFC} 1951} (May
181 @chapter Sample output
184 Here are some realistic examples of running @command{gzip}.
186 This is the output of the command @samp{gzip -h}:
189 Usage: gzip [OPTION]... [FILE]...
190 Compress or uncompress FILEs (by default, compress FILES in-place).
192 Mandatory arguments to long options are mandatory for short options too.
194 -c, --stdout write on standard output, keep original files unchanged
195 -d, --decompress decompress
196 -f, --force force overwrite of output file and compress links
197 -h, --help give this help
198 -k, --keep keep (don't delete) input files
199 -l, --list list compressed file contents
200 -L, --license display software license
201 -n, --no-name do not save or restore the original name and time stamp
202 -N, --name save or restore the original name and time stamp
203 -q, --quiet suppress all warnings
204 -r, --recursive operate recursively on directories
205 -S, --suffix=SUF use suffix SUF on compressed files
206 -t, --test test compressed file integrity
207 -v, --verbose verbose mode
208 -V, --version display version number
209 -1, --fast compress faster
210 -9, --best compress better
211 --rsyncable Make rsync-friendly archive
213 With no FILE, or when FILE is -, read standard input.
215 Report bugs to <bug-gzip@@gnu.org>.
218 This is the output of the command @samp{gzip -v texinfo.tex}:
221 texinfo.tex: 69.3% -- replaced with texinfo.tex.gz
224 The following command will find all regular @samp{.gz} files in the
225 current directory and subdirectories (skipping file names that contain
226 newlines), and extract them in place without destroying the original,
227 stopping on the first failure:
231 *' -prune -o -name '*.gz' -type f -print |
234 s/^\\(.*\\)\\.gz$/gunzip <'\\1.gz' >'\\1'/
240 @chapter Invoking @command{gzip}
244 The format for running the @command{gzip} program is:
247 gzip @var{option} @dots{}
250 @command{gzip} supports the following options:
256 Write output on standard output; keep original files unchanged.
257 If there are several input files, the output consists of a sequence of
258 independently compressed members. To obtain better compression,
259 concatenate all input files before compressing them.
268 Force compression or decompression even if the file has multiple links
269 or the corresponding file already exists, or if the compressed data
270 is read from or written to a terminal. If the input data is not in
271 a format recognized by @command{gzip}, and if the option @option{--stdout} is also
272 given, copy the input data without change to the standard output: let
273 @command{zcat} behave as @command{cat}. If @option{-f} is not given, and
274 when not running in the background, @command{gzip} prompts to verify
275 whether an existing file should be overwritten.
279 Print an informative help message describing the options then quit.
283 Keep (don't delete) input files during compression or decompression.
287 For each compressed file, list the following fields:
290 compressed size: size of the compressed file
291 uncompressed size: size of the uncompressed file
292 ratio: compression ratio (0.0% if unknown)
293 uncompressed_name: name of the uncompressed file
296 The uncompressed size is given as @minus{}1 for files not in @command{gzip}
297 format, such as compressed @samp{.Z} files. To get the uncompressed size for
298 such a file, you can use:
304 In combination with the @option{--verbose} option, the following fields are also
308 method: compression method (deflate,compress,lzh,pack)
309 crc: the 32-bit CRC of the uncompressed data
310 date & time: time stamp for the uncompressed file
313 The @abbr{CRC} is given as ffffffff for a file not in gzip format.
315 With @option{--verbose}, the size totals and compression ratio for all files
316 is also displayed, unless some sizes are unknown. With @option{--quiet},
317 the title and totals lines are not displayed.
319 The @command{gzip} format represents the input size modulo
320 @math{2^32}, so the uncompressed size and compression ratio are listed
321 incorrectly for uncompressed files 4 GiB and larger. To work around
322 this problem, you can use the following command to discover a large
323 uncompressed file's true size:
331 Display the @command{gzip} license then quit.
335 When compressing, do not save the original file name and time stamp by
336 default. (The original name is always saved if the name had to be
337 truncated.) When decompressing, do not restore the original file name
338 if present (remove only the @command{gzip}
339 suffix from the compressed file name) and do not restore the original
340 time stamp if present (copy it from the compressed file). This option
341 is the default when decompressing.
345 When compressing, always save the original file name and time stamp; this
346 is the default. When decompressing, restore the original file name and
347 time stamp if present. This option is useful on systems which have
348 a limit on file name length or when the time stamp has been lost after
353 Suppress all warning messages.
357 Travel the directory structure recursively. If any of the file names
358 specified on the command line are directories, @command{gzip} will descend
359 into the directory and compress all the files it finds there (or
360 decompress them in the case of @command{gunzip}).
363 While compressing, synchronize the output occasionally based on the
364 input. This can reduce the compression slightly in some cases, but
365 means that the @code{rsync} program can take advantage of similarities
366 in the uncompressed input when syncronizing two files compressed with
367 this flag. @code{gunzip} cannot tell the difference between a
368 compressed file created with this option, and one created without it.
371 While compressing, synchronize the output occasionally based on
372 the input. This increases size by less than 1 percent most
373 cases, but means that the @command{rsync} program can much more efficiently
374 synchronize files compressed with this flag. @command{gunzip}
375 cannot tell the difference between a compressed file created
376 with this option, and one created without it.
379 While compressing, synchronize the output occasionally based on the
380 input. This can reduce the compression slightly in some cases, but
381 means that the @code{rsync} program can take advantage of similarities
382 in the uncompressed input when syncronizing two files compressed with
383 this flag. @code{gunzip} cannot tell the difference between a
384 compressed file created with this option, and one created without it.
386 @item --suffix @var{suf}
388 Use suffix @var{suf} instead of @samp{.gz}. Any suffix can be
389 given, but suffixes other than @samp{.z} and @samp{.gz} should be
390 avoided to avoid confusion when files are transferred to other systems.
391 A null suffix forces gunzip to try decompression on all given files
392 regardless of suffix, as in:
395 gunzip -S "" * (*.* for MSDOS)
398 Previous versions of gzip used the @samp{.z} suffix. This was changed to
399 avoid a conflict with @command{pack}.
403 Test. Check the compressed file integrity.
407 Verbose. Display the name and percentage reduction for each file compressed.
411 Version. Display the version number and compilation options, then quit.
416 Regulate the speed of compression using the specified digit @var{n},
417 where @option{-1} or @option{--fast} indicates the fastest compression
418 method (less compression) and @option{--best} or @option{-9} indicates the
419 slowest compression method (optimal compression). The default
420 compression level is @option{-6} (that is, biased towards high compression at
425 @chapter Advanced usage
426 @cindex concatenated files
428 Multiple compressed files can be concatenated. In this case,
429 @command{gunzip} will extract all members at once. If one member is
430 damaged, other members might still be recovered after removal of the
431 damaged member. Better compression can be usually obtained if all
432 members are decompressed and then recompressed in a single step.
434 This is an example of concatenating @command{gzip} files:
437 gzip -c file1 > foo.gz
438 gzip -c file2 >> foo.gz
455 In case of damage to one member of a @samp{.gz} file, other members can
456 still be recovered (if the damaged member is removed). However,
457 you can get better compression by compressing all members at once:
460 cat file1 file2 | gzip > foo.gz
464 compresses better than
467 gzip -c file1 file2 > foo.gz
470 If you want to recompress concatenated files to get better compression, do:
473 zcat old.gz | gzip > new.gz
476 If a compressed file consists of several members, the uncompressed
477 size and @abbr{CRC} reported by the @option{--list} option applies to
479 only. If you need the uncompressed size for all members, you can use:
485 If you wish to create a single archive file with multiple members so
486 that members can later be extracted independently, use an archiver such
487 as @command{tar} or @command{zip}. @acronym{GNU} @command{tar}
488 supports the @option{-z}
489 option to invoke @command{gzip} transparently. @command{gzip} is designed as a
490 complement to @command{tar}, not as a replacement.
496 The environment variable @env{GZIP} can hold a set of default options for
497 @command{gzip}. These options are interpreted first and can be overwritten by
498 explicit command line parameters. For example:
501 for sh: GZIP="-8v --name"; export GZIP
502 for csh: setenv GZIP "-8v --name"
503 for MSDOS: set GZIP=-8v --name
506 On @abbr{VMS}, the name of the environment variable is @env{GZIP_OPT}, to
507 avoid a conflict with the symbol set for invocation of the program.
510 @chapter Using @command{gzip} on tapes
513 When writing compressed data to a tape, it is generally necessary to pad
514 the output with zeroes up to a block boundary. When the data is read and
515 the whole block is passed to @command{gunzip} for decompression,
516 @command{gunzip} detects that there is extra trailing garbage after the
517 compressed data and emits a warning by default if the garbage contains
518 nonzero bytes. You have to use the
519 @option{--quiet} option to suppress the warning. This option can be set in the
520 @env{GZIP} environment variable, as in:
523 for sh: GZIP="-q" tar -xfz --block-compress /dev/rst0
524 for csh: (setenv GZIP "-q"; tar -xfz --block-compress /dev/rst0)
527 In the above example, @command{gzip} is invoked implicitly by the @option{-z}
528 option of @acronym{GNU} @command{tar}. Make sure that the same block
530 option of @command{tar}) is used for reading and writing compressed data on
531 tapes. (This example assumes you are using the @acronym{GNU} version of
535 @chapter Reporting Bugs
538 If you find a bug in @command{gzip}, please send electronic mail to
539 @email{bug-gzip@@gnu.org}. Include the version number,
540 which you can find by running @w{@samp{gzip -V}}. Also include in your
541 message the hardware and operating system, the compiler used to compile
543 a description of the bug behavior, and the input to @command{gzip}
547 @node GNU Free Documentation License
548 @appendix GNU Free Documentation License
553 @appendix Concept index