Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
The PR was accidentally merged into master instead of staging and thus reverted. Now, in staging, we can re-revert it.
This reverts commit 4df2f78ec72f6a8d2fe286cd34eb3acdfcac81f3.
|
|
|
|
Should go to staging instead.
This reverts commit f8d9f59abe02ce97f621da24723156934a54611e, reversing
changes made to b27a19d5bf799f581a8afc2b554f178e58c1f524.
|
|
libarchive: link to zstd (split zstd output)
|
|
Also split outputs and don't provide static lib by default.
|
|
|
|
The previous CMake version did not run any tests, the current version
tries to run all tests (I could not await their completion), and the
last GNU Make version ran `make -C tests shortest` which ran only
playTests.sh, so I've enabled only the playTests test.
|
|
|
|
|
|
CMake is used for static builds because the Makefile does not
offer a way to disable the shared library.
|
|
Can just use single commit from
https://github.com/facebook/rocksdb/pull/6076 for rocksdb!
|
|
|
|
|
|
https://github.com/facebook/zstd/releases/tag/v1.4.3
|
|
https://github.com/facebook/zstd/releases/tag/v1.4.2
|
|
|
|
https://github.com/facebook/zstd/releases/tag/v1.4.0
|
|
https://github.com/facebook/zstd/releases/tag/v1.3.8
|
|
|
|
|
|
|
|
|
|
|
|
It had an install name '/usr/local/lib/libzstd.1.dylib'.
|
|
Goodbye, and thanks for all the Nix...
|
|
|
|
|
|
This removes some skipping of e.g. fixupPhase and cleans up occurences
where this led to duplicating code
|
|
|
|
|
|
|
|
|
|
Also updated to use the new repository location under facebook organization.
|
|
|
|
New:
- compression format specification zstd_compression_format.md
- -- separator, stating that all following arguments are file names
- ZSTD_getDecompressedSize()
Fixes:
- dictBuilder using HC levels
- legacy support from ZSTD_decompress_usingDDict()
- multi-blocks decoding with intermediate uncompressed blocks
|
|
Fixes:
- ZSTD_decompressBlock() using multiple consecutive blocks.
- potential segfault on very large files (many gigabytes).
- CLI displays system error message when destination file
cannot be created.
- potential leak in zdict.
|
|
|
|
Fixes:
- ZBUFF_compressEnd() called multiple times with too small dst buffer
- dictBuilder fails if first sample is too small
- corruption issue
- cli breaks during destination file overwrite confirmation
Changes:
- frame checksum enabled by default in command line mode
|
|
Changes: https://github.com/Cyan4973/zstd/releases/tag/v0.7.0
|
|
|
|
Changes:
- Stronger high compression modes
- Changed : highest compression modes require --ultra command to
remove memory restrictions
- API: ZSTD_getFrameParams() provides size of decompressed content
- Fixed: zstd cli return error code > 0 and removes dst file artifact
when decompression fails
- Various fixes and small performance improvements
|
|
Changes:
- New: Optimal parsing => Very high compression modes, thanks to @inikep
- Changed: Dictionary builder integrated into libzstd and zstd cli
- Changed (!): zstd cli now uses "multiple input files" as default mode. See zstd -h.
- Fix: high compression modes for big-endian platforms (#123)
- New: zstd cli : -t | --test command
|
|
“New: dictionary builder tool
Changed: advanced API for streaming and dictionary compression
Improved: better compression of small data.”
|
|
|