Archiving and compression

From ArchWiki
(Redirected from Archiver)

The traditional Unix archiving and compression tools are separated according to the Unix philosophy:

  • A file archiver combines several files into one archive file, e.g. tar.
  • A compression tool compresses and decompresses data, e.g. gzip.

These tools are often used in sequence by firstly creating an archive file and then compressing it.

Of course there are also tools that do both, which tend to additionally offer encryption, error detection and recovery.

Archiving only

Name Package Manuals Description
GNU tar tar tar(1), info Core utility for manipulating the ubiquitous tar archives (tarballs), which are used by pacman and the AUR.
libarchive libarchive bsdtar(1)
Implementation of tar and cpio that also offers a library. Used by pacman and mkinitcpio.
ar binutils ar(1) Legacy Unix archiver before tar. Today only used for creating static library files.
cpio cpio cpio(1) File archiver via stdin/stdout, supports cpio and tar formats.
DAR darAUR dar(1) Archiver to backup large live filesystems, takes care of hard links, extended attributes, sparse files and inode types.
Tip: Both GNU and BSD tar automatically do decompression delegation for bzip2, compress, gzip, lzip, lzma, lzop, zstd, and xz compressed archives. Only BSD tar supports lz4 natively (but GNU tar can do an equivalent with --use-compress-program=lz4/-Ilz4). When creating archives both support the -a switch to automatically filter the created archive through the right compression program based on the file extension. While BSD tar recognizes compression formats based on the format, GNU tar only guesses based on the file extension.

See also #Archiving only usage.

Compression tools

Compression only

These compression programs implement their own file format.

Name Package Manual Ext Tar ext Description Parallel implementations
bzip2 bzip2 bzip2(1) .bz2, .bz .tbz2, .tbz Uses the Burrows–Wheeler algorithm. lbzip2, pbzip2
bzip3 bzip3 bzip3(1) .bz3 .tbz3 Uses the Burrows–Wheeler algorithm.
gzip gzip gzip(1) .gz, .z .tgz, .taz GNU zip, based on DEFLATE algorithm. pigz, bgzip(1) (part of htslibAUR), crabzAUR, python-rapidgzipAUR
lrzip lrzip lrzip(1) .lrz Improved version of rzip, uses multiple algorithms. is multithreaded
LZ4 lz4 lz4(1) .lz4 Written in C, focused on compression and decompression speed. Can be multithreaded, though the lz4 is not. See for alternatives.
lzip lzip lzip(1) .lz Uses LZMA. plzipAUR
lzop lzop lzop(1) .lzo .tzo Uses the LZO library (lzo).
xz xz xz(1) .xz, .lzma .txz, .tlz Uses LZMA, default for GNU coreutils and kernel archive files. is multithreaded, pixz, pxz-gitAUR
zstd zstd zstd(1) .zst Uses Zstandard algorithm. is multithreaded
  • Parallel implementations offer improved speeds by using multiple CPU cores.
  • Tar extensions refers to compressed archives where tar and the compression tool is used, e.g. .tzo is .tar.lzo.
  • See also #Compression only usage.

Archiving and compression

Name Packages Manuals Ext Description
p7zip p7zip 7z(1) .7z The third-party POSIX port of 7-zip's command-line.
7-Zip 7-zipAUR, 7-zip-fullAUR - The upstream Linux version of a file archiver with a high compression ratio.
RAR rarAUR, unrar rar(1)[dead link 2023-08-09] .rar Both the format and the rar utility are proprietary.
t2sz t2szAUR .tar.zst .tzst Tar archiving utility in C with member-aligned zstd-compression
tarlz tarlzAUR tarlz(1) .tar.lz .tlz Tar archiving utility in C++ with member-aligned lzip compression
ZIP zip, unzip zip(1), unzip(1) .zip Widely used outside of the Linux world.
Unarchiver unarchiver unar(1), lsar(1) many Command-line tool of a Mac application, supports over 40 archive formats.
ZPAQ zpaqAUR zpaq(1) .zpaq A high compression ratio archiver written in C++, uses several algorithms.
LHa lhasa, lhaAUR lha(1) .lzh (on Amiga: .lha) LZH/LHA archiver, supports the lh7-method.
WinAce unace unace(1)[dead link 2023-08-09] .ace Both the ACE file format and the archiving tool are proprietary.

See also #Archiving and compression usage.

Feature charts

Some of the tools above are capable of handling multiple formats, allowing for fewer installed packages.


Name gzip bzip2 ZIP LHa/LZH RAR compress CAB ARJ
gzip Yes No Partial1 No No Yes No No
p7zip Yes Yes Yes Yes Yes No Yes Yes
unarchiver Yes Yes Yes Yes Yes Yes Yes Partial
  1. gzip's gunzip can only decompress single member ZIP files.

Usage comparison

Archiving only usage

Name Create archive Extract archive List content
tar(1) tar cfv archive.tar file1 file2 tar xfv archive.tar tar -tvf archive.tar
cpio(1) ls file1 file2 | cpio -o > archive.cpio cpio -i -vd < archive.cpio cpio -t < archive.cpio

Compression only usage

Name Compress Decompress Decompress to stdout
bzip2(1) bzip2 file bzip2 -d file.bz2 bzcat file.bz2
gzip(1) gzip file gzip -d file.gz zcat file.gz
lrzip(1) lrzip file
lrztar folder
lrzip -d file.lrz
lrztar -d folder.tar.lrz
lrzcat file.lrz
xz(1) xz file xz -d file.xz xzcat file.xz

Archiving and compression usage

Name Compress Decompress Decompress to stdout List content
7z(1) 7z a archive.7z file1 file2 7z x archive.7z 7z e -so archive.7z file1 7z l archive.7z
rar(1)[dead link 2023-08-09] rar a archive.rar file1 file2 rar x archive.rar rar p -inul archive.rar file1 rar l archive.rar
zip(1), unzip(1) zip file1 file2 unzip unzip -p file1 unzip -l
lha(1) lha ao7 archive.lzh file1 file2 lha x archive.lzh minimal: lha l archive.lzh verbose: lha v archive.lzh

Convenience tools

  • atool — Script for managing file archives of various types. || atool
  • dtrx — An intelligent archive extraction tool. || dtrxAUR
  • J7Z — GUI for Linux in java which attempts to simplify data compression and backup. It can create 7z, BZip2, Zip, GZip, Tar archives. || j7zAUR
  • unp — Command line tool that can unpack archives easily. || python-unpAUR
  • unpack — Wrapper script for handling multiple archive formats. || unpack-gitAUR
  • zutils — A drop-in replacement of zcat(1)/zgrep(1)/zdiff(1) in C++ to decompress gz/bz2/xz/lzip/zstd file transparently like bsdcat(1) but using user-defined utilities, usually used in combination with parallel variants of common (de)compressor like pixz(1). || zutilsAUR

Determining archive format

To extract an archive, its file format needs to be determined. If the file is properly named you can deduce its format from the file extension.

Otherwise you can use the file tool, see file(1).

Esoteric, rare or deprecated tools

Name Packages Ext Description
ARC arcAUR .arc, .ark Was very popular during the early days of the dial-up BBS. Superseded by ZIP.
ARJ arj .arj An archiver used on DOS/Windows in mid-1990s. This is an open source clone.
compress ncompress .Z The de facto standard UNIX compression utility to success the Huffman-based pack(1) before gzip become a thing.
PAR2 par2cmdline .par2 Parity archiver for increased data integrity. See also Parchive.
shar sharutils .shar Creates self-extracting archives that are valid shell scripts.
Zoo zooAUR .zoo Was mostly popular on the OpenVMS operating system before PKZIP became popular.

File system compression

Some file systems support on-the-fly compression of file data:

  • Btrfs can be configured to compress individual files, directories, or entire volumes by default.
  • On ZFS, compression can be enabled on pools or file systems.

Device mapper compression

There is work being done to mainline (integrate into the Linux kernel project) the open-sourced VDO project, which provides a deduplication and compression device mapper layer in the interest of increasing storage efficiency. The following packages are available:

  • vdo — Userspace tools for managing VDO volumes || vdoAUR
  • kvdo — A pair of kernel modules which provide pools of deduplicated and/or compressed block storage || kvdo-dkmsAUR

Compression libraries

  • Brotli — Compression algorithm for data streams using the LZ77 algorithm, Huffman coding and 2nd order context modeling. || brotli
  • libzip — Provides creation and extraction of ZIP files. Used by KDE and Deepin in place of the zip/unzip tools. || libzip
  • zlib — Compression library implementing the deflate compression method found in gzip and PKZIP. || zlib
  • Zopfli — High compress ratio file compressor from Google, using a deflate-compatible algorithm called zopfli. || zopfli-gitAUR


Garbled Japanese Filenames

Japanese versions of Windows encode ZIP archives with Shift-JIS. By default, these archives will suffer from mojibake filenames when extracted. To extract properly, use unzip in the command-line using the shift-jis option.

$ unzip -O shift-jis

See also