diff options
Diffstat (limited to 'nixpkgs/pkgs/tools/compression/lz4/default.nix')
-rw-r--r-- | nixpkgs/pkgs/tools/compression/lz4/default.nix | 68 |
1 files changed, 68 insertions, 0 deletions
diff --git a/nixpkgs/pkgs/tools/compression/lz4/default.nix b/nixpkgs/pkgs/tools/compression/lz4/default.nix new file mode 100644 index 000000000000..fd967fce9c6c --- /dev/null +++ b/nixpkgs/pkgs/tools/compression/lz4/default.nix @@ -0,0 +1,68 @@ +{ stdenv, fetchFromGitHub, valgrind, fetchpatch +, enableStatic ? false, enableShared ? true +}: + +stdenv.mkDerivation rec { + pname = "lz4"; + version = "1.9.1"; + + src = fetchFromGitHub { + sha256 = "1l1caxrik1hqs40vj3bpv1pikw6b74cfazv5c0v6g48zpcbmshl0"; + rev = "v${version}"; + repo = pname; + owner = pname; + }; + + patches = [ + # Fix detection of Darwin + (fetchpatch { + url = "https://github.com/lz4/lz4/commit/024216ef7394b6411eeaa5b52d0cec9953a44249.patch"; + sha256 = "0j0j2pr6pkplxf083hlwl5q4cfp86q3wd8mc64bcfcr7ysc5pzl3"; + }) + ]; + + # TODO(@Ericson2314): Separate binaries and libraries + outputs = [ "bin" "out" "dev" ]; + + buildInputs = stdenv.lib.optional doCheck valgrind; + + enableParallelBuilding = true; + + makeFlags = [ + "PREFIX=$(out)" + "INCLUDEDIR=$(dev)/include" + "BUILD_STATIC=${if enableStatic then "yes" else "no"}" + "BUILD_SHARED=${if enableShared then "yes" else "no"}" + "WINDRES:=${stdenv.cc.bintools.targetPrefix}windres" + ] + # TODO make full dictionary + ++ stdenv.lib.optional stdenv.hostPlatform.isMinGW "TARGET_OS=MINGW" + ; + + doCheck = false; # tests take a very long time + checkTarget = "test"; + + # TODO(@Ericson2314): Make resusable setup hook for this issue on Windows. + postInstall = + stdenv.lib.optionalString stdenv.hostPlatform.isWindows '' + mv $out/bin/*.dll $out/lib + ln -s $out/lib/*.dll + '' + + '' + moveToOutput bin "$bin" + ''; + + meta = with stdenv.lib; { + description = "Extremely fast compression algorithm"; + longDescription = '' + Very fast lossless compression algorithm, providing compression speed + at 400 MB/s per core, with near-linear scalability for multi-threaded + applications. It also features an extremely fast decoder, with speed in + multiple GB/s per core, typically reaching RAM speed limits on + multi-core systems. + ''; + homepage = https://lz4.github.io/lz4/; + license = with licenses; [ bsd2 gpl2Plus ]; + platforms = platforms.all; + }; +} |