CHANGES.md

v1.5.3 2023-09-19 Paris (France)

• Delete libdecompress.a (@dinosaure, #152)

  This artifact is unused and complexify the distribution. We decided to delete it. If people want to use libdecompress.a, we can provide an other package which will implement the ctypes ceremony to safely produce a libdecompress.a.

• Lint dependencies (lower bounds) (@dinosaure, #153)

• Remove unsafe accesses into our decompress.lzo implementation (@dinosaure, #154)

• Improve and fix our decompress.lzo implementation (@dinosaure, #155, #158, #159)

v1.5.2 2023-01-23 Essaouira (Maroc)

• Remove remaining bigarray-compat dependencies (@copy, #147)

v1.5.1 2022-08-30 Paris (France)

• Fix the stream of gzip inflation. If the user wants to know how many bytes are available into the output buffer, he/she must be in the DEFLATE internal state. Otherwise, we raise an exception. However, such information is not available into the API so we decided to say that the full output buffer is free when we are into the GZip header state.

  It ensures a real full stream API. (@dinosaure, #144)

v1.5.0 2022-08-28 Paris (France)

• Update with ocamlformat.0.20.0 (@dinosaure, #133)

• Add lzo into the binary (@dinosaure, #140)

• Be able to deflate/inflate files (@dinosaure, #141)

• Implement the zero-compression into zlib/gzip (@dinosaure, #142) breaking change: The behavior of flat block changed. The user does not specify how many bytes he/she wants to give. He/She can just specify that he/she wants a Flat block. The Flat block will contains what the queue has and nothing more: if the queue has 4 elements, we will encode a Flat block with 4 elements, if the queue has more than 65535 elements, we will encore only 65535 elements into the Flat block.

  The user is not able to stream a Flat block: we can not fill one Flat block with multiple Await. This behavior is specific to the Flat block, the rest (Fixed and Dynamic blocks) did not change.

  For higher level API, the level = 0 informs the deflator to copy as is input - no compression was done for zlib and gzip and we emit only Flat blocks. By default, the level = 4 is given so you probably will not notice anything but be care that we shifted compression level and 0 becomes a level without compression.

v1.4.3 2022-04-08 Paris (France)

• Replace deprecated function of fmt (@dinosaure, #135)
• Remove bigarray-compat, stdlib-shims and support only OCaml >= 4.08.0 (@hannesm, #138)
• Update to cmdliner.1.1.0 (@hannesm, #138)

v1.4.2 2021-08-02 Paris (France)

• Fix lower bounds of cmdliner (@kit-ty-kate, #130)
• Fix big-endian support (@dinosaure, @talex5, #131)

v1.4.1 2021-05-11 Paris (France)

• Fix and of file and end of block op-code (@dinosaure, #123) breaking changes Semantically, the module De has another behavior about the inflation. Previously, the stream inflation was smart enough to recognize the end of the stream and the user did not need to really emit: De.Inf.src decoder empty 0 0 to say the end of the stream. Now, such call is required to notice to De.Inf the end of the stream. By this way, we are able to terminate the inflation correctly and we still continue to raise an error for unterminated stream (see tests/invalid_distance_code).

  For Zl/Gz users, this update does not imply anything when these implementations take care about such detail. Only De users should update their code to really emit the end of the stream with De.Inf.src decoder empty 0 0. In the PR, the diff show how to upgrade such code.

• Upgrade decompress to optint.0.1.0 (@samoht, @dinosaure, #124)

• Fix compilation of benchmarks (@dinosaure, #128)

• Fix out of bounds errors on the non-stream implementation (@clecat, @ewanmellor, @dinosaure, #126, #127)

• Optimize memcpy used on the non-stream implementation (@clecat, @dinosaure, #129)

v1.4.0 2021-04-22 Paris (France)

• Add a well-know limitation about the encoding on the documentation, the output buffer must be upper than 2 bytes in any cases. (@dinosaure, #114)

• Improve the documentation (@dinosaure, @brendanlong, #115)

• breaking changes, the type of the window used to deflate according RFC 1951 was updated to De.Lz77.window (@dinosaure, #116 & #115)

• Fix a bug when we want to add the EOB op-code into the queue. The deflation takes care about that. Note that the queue must be larger than 2. (@dinosaure, @kluvin, #117)

• Improve the documentation a bit (@mseri, @dinosaure, #119)

• Fix the use of optint and how we handle large files with Gz. Fix an error when we encode the isize into the deflated stream. (@dinosaure, @igarnier, #121 & #120)

• Add a non-stream implementation (@clecat, @dinosaure, #102 & #92)

  The non-blocking stream API has a cost to maintain a state across syscall such as read and write. It's useful when we want to plug decompress behind something like a socket and care about memory-consumption but it has a big cost when we want to compress/decompress an object saved into one and unique buffer.

  The non-stream API gives an opportunity to inflate/deflate one and unique buffer without the usual plumbing required by the non-blocking stream API. However, we are limited to compute only objects which can fit into a bigarray.

  About performance, the non-stream API is better than the non-blocking stream API. See the PR for more details about performances. On the book2 (from the Calgary corpus) file:

  • decompress (stream): 15 Mb/s (deflation), 76 Mb/s (inflation), ratio: 42.46 %
  • decompress (non-stream): 17 Mb/s (deflation), 105 Mb/s (inflation), ratio: 34.66 %

  Even if we checked the implementation with our tests (we ran ocaml-git and irmin with this path), the implementation is young and we probably miss some details/bugs. So we advise the user to compare, at least, the non-stream implementation with the non-blocking stream implementation if something is wrong.

v1.3.0 2020-03-03 Paris (France)

• Add a little executable to benchmark inflation into the distribution (@dinosaure, #93)

• Add instructions for running benchmark (@gs0510, #94)

• Clarify the description (@XVilka, #96)

• Improve the benchmark and outputs (@dinosaure, @gs0510, #95)

• Avoid allocation of distance table (@Engil, @dinosaure, #97)

• Swapping from arithmetic to logical bitshifts on d.hold (@clecat, #99)

• Make the use of all Higher.compress arguments (@vect0r-vicall, #103)

• Apply ocamlformat.0.16.0 (@dinosaure, #105, #107)

• Improve Lz77 algorithms (@dinosaure, #108) breaking changes the deflation expects a new window: De.Lz77.make_window instead of De.make_window (which is twice larger to improve the compression algorithm)

  Depending on the level and your corpus, we did not observe performance regression on deflation (and #97 improves a lot performances). An higher level is slower (but the compression ratio is better). We advise, by default, to use the level 6.

  Note that the user is able to make its own compression algorithm according to his corpus. An example of such implementation is available on the new decompress.lz libraries which fills a queue and compress the input.

  breaking changes decompress expects a level between 0 and 9 (inclusive) (instead of 0 and 3).

• Add tests about level compression (@dinosaure, #109)

• Add level on GZip layer (@dinosaure, #110)

• Provide a ctypes reverse binding (@dinosaure, #98)

• Provide a binary decompress.pipe which can compress/uncompress with deflate, zlib or gzip format.

v1.2.0 2020-07-07 Paris (France)

• add LZO support (@dinosaure, @cfcs, @XVilka, #82)
• update binaries (@dinosaure, @XVilka, #89)
• fix an exception leak (@dinosaure, @b1gtang, #88)
• update README.md (@dinosaure, @XVilka, #87)
• fix a mis-use of Zl API (@dinosaure, #85)
• add dune as a dependency of rfc1951 (@kit-ty-kate)
• real non-blocking state with Zl (@dinosaure, #84)

v1.1.0 2019-03-10 Paris (France)

• add GZip support (@dinosaure, @copy, @hcarty, #79)
• breaking changes, Higher returns a result value instead to raise an exception (@dinosaure, @copy, #80)
• protect Zlib decoder on multiple no-op calls of decode
• test when we generate an empty zlib flow
• fix a bug on the DEFLATE layer when we must flush bits to avoid integer overflow
• really use the internal continuation of the Zlib state
• delete fmt depedency
• update fuzzer
• fix default level on Zl.Higher

v1.0.0 2019-08-30 Paris (France)

** breaking changes **

decompress.1.0.0 is 3 times faster about decompression than before. A huge amount of work was done to improve performance and coverage.

The main reason to update the API is to fix a bad design decision regarding split compression and encoding. User is able to implement a new compression algorithm and use it.

Release comes with regressions:

• decompress only supports Bigarray now, not Bytes
• GZIP layer does not exist anymore
• state of RFC1951 encoder/decoder is not referentially transparent anymore

Of course, v1.0.0 comes with fixes and improvements:

• decompress is able to compress/uncompress Calgary corpus
• tests are improved and they include all coverage tests from zlib
• compression algorithm has a fuzzer
• encoder has a fuzzer
• performance about decoder is 3 times better than decompress.v0.9.0 and 3 times slower than zlib

decompress is split into 2 main modules:

• dd which implements RFC1951
• zz which implements ZLIB

API of them are pretty-close to what decompress.v0.9.0 does with some advantages on dd:

• User can use their own compression algorithm instead of Dd.L
• encoder exposes more granular control over what it emits (which block, when, where)
• Huffman tree generation is out of dd

As a response to #25, dd provides a higher level API resembling camlzip.

v0.9.0 2019-07-10 Paris (France)

• Add support of 4.07 and 4.08 in Travis (@XVilka, @dinosaure, #70, #71)
• Use mmap (@XVilka, @dinosaure, @hannesm, #68, #69, #71)
• Update documentation (@yurug, @dinosaure, #65, #66)
• Micro-optimization about specialization (@dinosaure, #64)
• Re-organize internals of decompress (@dinosaure, #63)
• GZIP support (@clecat, review by @dinosaure, @cfcs, @hannesm, #60)

• fix #58 (@dinosaure)

v0.8.1 2018-10-16 Paris (France)

• Dunify project (@dinosaure)
• breaking-change Unbox Bytes.t and Bigstring.t as I/O buffer (@dinosaure)
• Add foreign tests vectors (@cfcs, @dinosaure)
• Catch invalid distance (@XVilka, @dinosaure)
• Better check on dictionaries (@XVilka, @dinosaure)
• breaking-change Add [wbits] argument to check Window size on RFC1951 (@XVilka, @dinosaure)

v0.8 2018-07-09 Paris (France)

• Implementation of RFC1951 (task from @cfcs)

• breaking change New interface of decompress

  We wrap API in Zlib_{inflate/deflate} and add RFC1951_{inflate/deflate}.

• Move to jbuilder/dune (task from @samoht)

• Better check on zlib header

• Fixed infinite loop (task fron @cfcs)

  See 2e3af68, decompress has an infinite loop when the inflated dictionary does not provide any bindings (and length of opcode is <= 0). In this case, decompress expects an empty input and provide an empty output in any case.

• Use re.1.7.2 on tests

• Use camlzip.1.07 on tests

v0.7 2017-10-18 Paris (France)

• Fixed Inflate.write function
• Fixed internal state to stick in a internal final state
• Fixed compilation with js_of_ocaml (use trampoline function to avoid stack-overflow)
• Fixed clash of name about state variable in the Inflate module
• Add afl program
• Export Adler-32 modules
• Add -i and -o option in the dpipe binary to inform the size of the internal chunk
• Change the value of -mode in the dpipe binary

v0.6 2017-05-11 Cao Lãnh (Vietnam)

• Fixed bug #29
• Produce far pattern (Lz77 compression)
• Optimize memory consumption of the Inflate module
• Move repository from oklm-wsh to mirage
• Learn topkg release

v0.5 2017-02-17 Essaouira (Maroc)

• Stabilize the interface (@dbuenzli's interface)
• Merge optimization from @yallop
• Add sync_flush, partial_flush, full_flush (experimental)
• Move the build system to topkg