| LZFSE | |
|---|---|
| Developer | Apple |
| Initial release | 2015 |
| Stable release | lzfse-1.0 / 8 May 2017; 8 years ago (2017-05-08) |
| Written in | C |
| Operating system | macOS,iOS,Linux[1] |
| Available in | C |
| Type | Data compression |
| License | 3-clause NewBSD License |
| Repository | lzfse onGitHub |
LZFSE (Lempel–Ziv Finite State Entropy) is anopen sourcelossless data compressionalgorithm created byApple Inc. It was released with a simpler algorithm calledLZVN.[2]
The name is an acronym forLempel–Ziv andfinite-state entropy[3] (implementation ofasymmetric numeral systems).LZFSE was introduced by Apple at itsWorldwide Developer Conference 2015. It shipped with that year'siOS 9 andOS X 10.11 releases.
Apple claims that LZFSE compresses with a ratio comparable to that ofzlib (DEFLATE) and decompresses two to three times faster while using fewer resources, therefore offering higherenergy efficiency than zlib. It was aimed for scenarios where decompression speed and rate should be prioritised equally.[3] Part of this energy efficiency was achieved by optimising the algorithm for modern micro-architectures, specifically focusing onarm64.[4] Third-party benchmarking confirms that LZFSE decompresses faster than zlib, but also suggests that many other modern compression algorithms may have more favorable compression algorithm performance characteristics such as density, compression speed and decompression speed by a significant margin.[5]
According to the Squash Benchmark, LZFSE is similar in speed tozstd (level 6), but has a slightly worse ratio. LZVN is similar in speed to LZ4 level 4, with a slightly worse ratio as well.[6] Neither LZFSE nor LZVN is tunable at runtime, although a few constants can be tweaked at compile time for the usual speed-ratio trade-off.[7]
A referenceClibrary written by Eric Bainville was made available under the3-clause BSD License afterWWDC 2016. It includes an executable to compress and decompress LZFSE streams as well. There are no plans to expose an LZVN API.[1]
Apple's LZFSE implementation uses a simpler algorithm called LZVN when the input is smaller thanLZFSE_ENCODE_LZVN_THRESHOLD (4096 bytes). This is aLZSS-type algorithm without entropy encoding but with three widths of REP (L,M,D) packets. In the open source reference implementation, Apple explains that LZFSE does not perform as well for small sizes, so LZVN is used instead.[7] This algorithm in libfastCompression.a was discovered earlier as the default kernelcache compression method in Mac OS X Yosemite Developer Preview 1 (2014), replacing the legacylzss compression fromHaruhiko Okumura.[8]
AppleFSCompression.framework (AFSC), the mechanism for quasi-transparent compression inHFS Plus andApple File System, supports LZFSE and LZVN since OS X 10.9.
Apple'sDisk Images framework has offered an LZFSE-based encoding calledULFO sinceMac OS X 10.11,[9] accessible viahdiutil(1)[10] and some third-party image utilities.
Apple introduced theApple Archive format and its associated API inmacOS High Sierra in 2017.[11] The extension name is.aar (sincemacOS Big Sur, used to be.yaa). Encryption was introduced inmacOS Monterey, when AA became the defaultArchive Utility format. Three command-line utilities are available in macOS to handle AA files.[12][13] Of third-party programs, Keka is able to use the system APIs to handle AA files, but no independent implementations exist on other systems.[14]