extzstd 0.0.2.CONCEPT → 0.0.3.CONCEPT

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (68) hide show
  1. checksums.yaml +4 -4
  2. data/README.md +72 -22
  3. data/Rakefile +73 -19
  4. data/contrib/zstd/README.md +68 -0
  5. data/contrib/zstd/common/bitstream.h +414 -0
  6. data/contrib/zstd/common/entropy_common.c +231 -0
  7. data/contrib/zstd/common/error_private.h +125 -0
  8. data/contrib/zstd/common/error_public.h +77 -0
  9. data/contrib/zstd/common/fse.h +628 -0
  10. data/contrib/zstd/common/fse_decompress.c +331 -0
  11. data/contrib/zstd/common/huf.h +228 -0
  12. data/contrib/zstd/common/mem.h +377 -0
  13. data/contrib/zstd/common/xxhash.c +854 -0
  14. data/contrib/zstd/common/xxhash.h +273 -0
  15. data/contrib/zstd/common/zbuff.h +197 -0
  16. data/contrib/zstd/common/zstd.h +475 -0
  17. data/contrib/zstd/common/zstd_common.c +91 -0
  18. data/contrib/zstd/common/zstd_internal.h +238 -0
  19. data/contrib/zstd/compress/fse_compress.c +807 -0
  20. data/contrib/zstd/compress/huf_compress.c +577 -0
  21. data/contrib/zstd/compress/zbuff_compress.c +327 -0
  22. data/contrib/zstd/compress/zstd_compress.c +3074 -0
  23. data/contrib/zstd/compress/zstd_opt.h +1046 -0
  24. data/contrib/zstd/decompress/huf_decompress.c +894 -0
  25. data/contrib/zstd/decompress/zbuff_decompress.c +294 -0
  26. data/contrib/zstd/decompress/zstd_decompress.c +1362 -0
  27. data/contrib/zstd/dictBuilder/divsufsort.c +1913 -0
  28. data/contrib/zstd/dictBuilder/divsufsort.h +67 -0
  29. data/contrib/zstd/dictBuilder/zdict.c +1045 -0
  30. data/contrib/zstd/dictBuilder/zdict.h +113 -0
  31. data/contrib/zstd/legacy/zstd_legacy.h +140 -0
  32. data/contrib/zstd/legacy/zstd_v01.c +2178 -0
  33. data/contrib/zstd/{zstd.h → legacy/zstd_v01.h} +46 -39
  34. data/contrib/zstd/legacy/zstd_v02.c +3748 -0
  35. data/contrib/zstd/legacy/zstd_v02.h +99 -0
  36. data/contrib/zstd/legacy/zstd_v03.c +3389 -0
  37. data/contrib/zstd/legacy/zstd_v03.h +99 -0
  38. data/contrib/zstd/legacy/zstd_v04.c +4056 -0
  39. data/contrib/zstd/legacy/zstd_v04.h +148 -0
  40. data/contrib/zstd/legacy/zstd_v05.c +4325 -0
  41. data/contrib/zstd/legacy/zstd_v05.h +171 -0
  42. data/contrib/zstd/legacy/zstd_v06.c +4581 -0
  43. data/contrib/zstd/legacy/zstd_v06.h +185 -0
  44. data/ext/extconf.rb +10 -12
  45. data/ext/extzstd.c +497 -144
  46. data/ext/extzstd.h +127 -22
  47. data/ext/extzstd_buffered.c +265 -0
  48. data/ext/extzstd_nogvls.h +174 -0
  49. data/ext/zstd_amalgam.c +18 -0
  50. data/ext/zstd_legacy_v01.c +1 -0
  51. data/ext/zstd_legacy_v02.c +1 -0
  52. data/ext/zstd_legacy_v03.c +1 -0
  53. data/ext/zstd_legacy_v04.c +1 -0
  54. data/ext/zstd_legacy_v05.c +1 -0
  55. data/ext/zstd_legacy_v06.c +1 -0
  56. data/gemstub.rb +17 -1
  57. data/lib/extzstd.rb +197 -77
  58. data/lib/extzstd/version.rb +1 -1
  59. data/test/test_basic.rb +41 -0
  60. metadata +70 -20
  61. data/contrib/zstd/Makefile +0 -115
  62. data/contrib/zstd/fse.c +0 -2466
  63. data/contrib/zstd/fse.h +0 -320
  64. data/contrib/zstd/fse_static.h +0 -282
  65. data/contrib/zstd/libzstd.pc.in +0 -14
  66. data/contrib/zstd/zstd.c +0 -1768
  67. data/contrib/zstd/zstd_static.h +0 -89
  68. data/ext/extzstd-stream.c +0 -398
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA1:
3
- metadata.gz: e5563734d016a497235ac43fed1b14a41d32e1c3
4
- data.tar.gz: 63389454aa8f607e5f823b27314be3ae7efe45ed
3
+ metadata.gz: 2fe6b36e5eb4cffa0d077511dcbd70d869d499f1
4
+ data.tar.gz: 9acfaccd4a530eff2581bb407a2d221a1926e3e1
5
5
  SHA512:
6
- metadata.gz: a7dd05b9080beec1707db4e76948154309052fb85c615a211bcb4c82147fc84a666aadeddf670dbb1932a406ef88448e133ed33967cd06b27af2f1536c6ab042
7
- data.tar.gz: 53e088c35e916c5bb21e10fe849e57a20e1e933b9f97cc5ad43be6c0e8531fe3984682bbeb2fe7b542fadccffb07ac9e2b1a8a978696fd43ba80a6155d7bbc6c
6
+ metadata.gz: 23b1a8f69661f30443e2eeb0207dc4260b760e2cff9240fcc8e7f4f89656d12f3302f983c9101a4f5fb786304e61d854fa83f36199640a082950a292b6eeb47f
7
+ data.tar.gz: 0c666f160c2d2a53c52327bbc1e482152f712c11eaff45ae6e472467c84f6a8be96d2e716d5034cc9bf3ce32491ff4f045c96b897458f45653e9f99b5c68a1bc
data/README.md CHANGED
@@ -1,38 +1,53 @@
1
- # encoding:utf-8 ;
2
-
3
1
  # extzstd - ruby bindings for Zstd (Zstandard)
4
2
 
5
- This is ruby bindings for compression library
3
+ This is unofficial ruby bindings for compression library
6
4
  [Zstd (https://github.com/Cyan4973/zstd)](https://github.com/Cyan4973/zstd).
7
5
 
8
6
  * package name: extzstd
9
- * author: dearblue (mailto:dearblue@users.osdn.me)
10
- * version: 0.0.2.CONCEPT
7
+ * version: 0.0.3.CONCEPT
11
8
  * software quality: EXPERIMENTAL
12
9
  * license: 2-clause BSD License
13
- * report issue to: https://osdn.jp/projects/rutsubo/ticket/
10
+ * author: dearblue <mailto:dearblue@users.osdn.me>
11
+ * report issue to: <https://osdn.jp/projects/rutsubo/ticket/>
14
12
  * dependency ruby: ruby-2.0+
15
13
  * dependency ruby gems: (none)
16
14
  * dependency library: (none)
17
- * bundled external libraries:
18
- * zstd-0.1.2 (https://github.com/Cyan4973/zstd/tree/zstd-0.1.2)
15
+ * bundled external C library:
16
+ * zstd-0.7.4 <https://github.com/Cyan4973/zstd/tree/v0.7.4>
17
+
18
+ "extzstd" is supported the legacy formats (zstd-0.1 - 0.6).
19
19
 
20
20
 
21
- ## ***WARNING***
21
+ ## Quick API Reference
22
22
 
23
- Zstd data format compatibility is not guaranteed in future versions
24
- (There is a possibility that it becomes impossible to future use).
23
+ * Encode
24
+ * ``Zstd.encode(buf, params = nil, dict = nil) -> encoded string``
25
+ * ``Zstd.encode(outport, params = nil, dict = nil) -> an instance of Zstd::Encoder``
26
+ * ``Zstd.encode(outport, params = nil, dict = nil) { |encoder| ... } -> block returned value``
27
+ * ``Zstd::Encoder#write(buf) -> this instance``
28
+ * ``Zstd::Encoder#close -> nil``
25
29
 
26
- Written in [zstd/README.md](https://github.com/Cyan4973/zstd/blob/zstd-0.1.2/README.md):
30
+ * Decode
31
+ * ``Zstd.decode(zstd_buf, dict = nil) -> decoded string``
32
+ * ``Zstd.decode(inport, dict = nil) -> an intance of Zstd::Decoder``
33
+ * ``Zstd.decode(inport, dict = nil) { |decoder| ... } -> block returned value``
34
+ * ``Zstd::Decoder#read(size = nil, buf = nil) -> buf``
35
+ * ``Zstd::Decoder#close -> nil``
27
36
 
28
- > Zstd has not yet reached "stable" status. Specifically, it doesn't
29
- > guarantee yet that its current compressed format will remain stable
30
- > and supported in future versions.
37
+ * buffered encoder
38
+ * ``Zstd::BufferedEncoder.new(params, dict)`` (``ZBUFF_createCCtx``, ``ZBUFF_compressInit_advanced``)
39
+ * ``Zstd::BufferedEncoder#continue(data, dataoff, dest, maxdest) -> integer as new data offset`` (``ZBUFF_compressContinue``)
40
+ * ``Zstd::BufferedEncoder#flush(dest, maxdest) -> dest`` (``ZBUFF_compressFlush``)
41
+ * ``Zstd::BufferedEncoder#end(dest, maxdest) -> dest`` (``ZBUFF_compressEnd``)
42
+
43
+ * Train dictionary
44
+ * ``Zstd.dict_train_from_buffer(buf, dict_capacity) -> dictionary'd string`` (``ZDICT_trainFromBuffer``)
45
+ * ``Zstd.dict_add_entropy_tables_from_buffer(dict, dict_capacity, sample) -> dict`` (``ZDICT_addEntropyTablesFromBuffer``)
31
46
 
32
47
 
33
48
  ## HOW TO USE
34
49
 
35
- ### basic usage (one pass encode/decode)
50
+ ### basic usage (simply encode/decode)
36
51
 
37
52
  ``` ruby:ruby
38
53
  # First, load library
@@ -41,19 +56,54 @@ require "extzstd"
41
56
  # Prepair data
42
57
  source = "sample data..." * 50
43
58
 
44
- # Directly compression
59
+ # simply compression
45
60
  encdata = Zstd.encode(source)
46
61
  puts "encdata.bytesize=#{encdata.bytesize}"
47
62
 
48
- # Directly decompression
49
- maxdestsize = source.bytesize
50
- decdata = Zstd.decode(encdata, maxdestsize)
63
+ # simply decompression
64
+ decdata = Zstd.decode(encdata)
51
65
  puts "decdata.bytesize=#{decdata.bytesize}"
52
66
 
53
67
  # Comparison source and decoded data
54
68
  p source == decdata # => true
55
69
  ```
56
70
 
57
- ----
71
+ ### Streaming compression (with block)
72
+
73
+ ``` ruby:ruby
74
+ outport = StringIO.new("")
75
+ Zstd.encode(outport) do |encoder|
76
+ encoder.write "abcdefg\n"
77
+ encoder << "hijklmn\n"
78
+ encoder.write "opqrstu\n"
79
+ encoder << "vwxyz\n"
80
+ end
81
+ ```
82
+
83
+ ### Streaming compression (without block and write to file)
58
84
 
59
- [a stub]
85
+ ``` ruby:ruby
86
+ file = File.open("sample.zst", "wb")
87
+ encoder = Zstd.encode(file)
88
+
89
+ encoder.write "abcdefg\n"
90
+ encoder << "hijklmn\n"
91
+ encoder.write "opqrstu\n"
92
+ encoder << "vwxyz\n"
93
+
94
+ encoder.close
95
+ file.close
96
+ ```
97
+
98
+ ### Streaming decompression (with block and read from file)
99
+
100
+ ``` ruby:ruby
101
+ File.open("sample.zst", "rb") do |file|
102
+ Zstd.decode(file) do |decoder|
103
+ p decoder.read(8) # => "abcdefg\n"
104
+ p decoder.read(1) # => "h"
105
+ p decoder.read(2) # => "ij"
106
+ p decoder.read # => "klmn\nopqrstu\nvwxyz\n"
107
+ end
108
+ end
109
+ ```
data/Rakefile CHANGED
@@ -1,11 +1,14 @@
1
1
 
2
+ require "pathname"
2
3
  require "rake/clean"
3
4
 
4
- DOC = FileList["{README,LICENSE,CHANGELOG,Changelog,HISTORY}{,.ja}{,.txt,.rd,.rdoc,.md,.markdown}"] +
5
- FileList["{contrib,ext}/**/{README,LICENSE,CHANGELOG,Changelog,HISTORY}{,.ja}{,.txt,.rd,.rdoc,.md,.markdown}"] +
6
- FileList["ext/**/*.{c,C,cc,cxx,cpp,h,H,hh}"]
7
- #EXT = FileList["ext/**/*.{h,hh,c,cc,cpp,cxx}"] +
8
- # FileList["ext/externals/**/*"]
5
+ docnames = "{README,LICENSE,CHANGELOG,Changelog,HISTORY}"
6
+ doctypes = "{,.txt,.rd,.rdoc,.md,.markdown}"
7
+ cexttypes = "{c,C,cc,cxx,cpp,h,H,hh}"
8
+
9
+ DOC = FileList["#{docnames}{,.ja}#{doctypes}"] +
10
+ FileList["{contrib,ext}/**/#{docnames}{,.ja}#{doctypes}"] +
11
+ FileList["ext/**/*.#{cexttypes}"]
9
12
  EXT = FileList["ext/**/*"]
10
13
  BIN = FileList["bin/*"]
11
14
  LIB = FileList["lib/**/*.rb"]
@@ -15,18 +18,23 @@ EXAMPLE = FileList["examples/**/*"]
15
18
  GEMSTUB_SRC = "gemstub.rb"
16
19
  RAKEFILE = [File.basename(__FILE__), GEMSTUB_SRC]
17
20
  EXTRA = []
21
+ EXTCONF = FileList["ext/**/extconf.rb"]
22
+ EXTCONF.reject! { |n| !File.file?(n) }
23
+ EXTMAP = {}
18
24
 
19
25
  load GEMSTUB_SRC
20
26
 
21
- EXTCONF = FileList["ext/extconf.rb"]
22
- EXTCONF.reject! { |n| !File.file?(n) }
27
+ EXTMAP.dup.each_pair do |dir, name|
28
+ EXTMAP[Pathname.new(dir).cleanpath.to_s] = Pathname.new(name).cleanpath.to_s
29
+ end
30
+
23
31
  GEMSTUB.extensions += EXTCONF
24
32
  GEMSTUB.executables += FileList["bin/*"].map { |n| File.basename n }
25
33
  GEMSTUB.executables.sort!
26
34
 
27
35
  PACKAGENAME = "#{GEMSTUB.name}-#{GEMSTUB.version}"
28
36
  GEMFILE = "#{PACKAGENAME}.gem"
29
- GEMSPEC = "#{GEMSTUB.name}.gemspec"
37
+ GEMSPEC = "#{PACKAGENAME}.gemspec"
30
38
 
31
39
  GEMSTUB.files += DOC + EXT + EXTCONF + BIN + LIB + SPEC + TEST + EXAMPLE + RAKEFILE + EXTRA
32
40
  GEMSTUB.files.sort!
@@ -37,11 +45,27 @@ end
37
45
  GEMSTUB.extra_rdoc_files += DOC + LIB + EXT.reject { |n| n.include?("/externals/") || !%w(.h .hh .c .cc .cpp .cxx).include?(File.extname(n)) }
38
46
  GEMSTUB.extra_rdoc_files.sort!
39
47
 
40
- CLEAN << GEMSPEC
48
+ GEMSTUB_TRYOUT = GEMSTUB.dup
49
+ GEMSTUB_TRYOUT.version = "#{GEMSTUB.version}#{Time.now.strftime(".TRYOUT.%Y%m%d.%H%M%S")}"
50
+ PACKAGENAME_TRYOUT = "#{GEMSTUB.name}-#{GEMSTUB_TRYOUT.version}"
51
+ GEMFILE_TRYOUT = "#{PACKAGENAME_TRYOUT}.gem"
52
+ GEMSPEC_TRYOUT = "#{PACKAGENAME_TRYOUT}.gemspec"
53
+
54
+ CLEAN << GEMSPEC << GEMSPEC_TRYOUT
41
55
  CLOBBER << GEMFILE
42
56
 
43
- task :default => :all
57
+ task :default => :tryout do
58
+ $stderr.puts <<-EOS
59
+ #{__FILE__}:#{__LINE__}:
60
+ \ttype ``rake release'' to build release package.
61
+ EOS
62
+ end
63
+
64
+ desc "build tryout package"
65
+ task :tryout
44
66
 
67
+ desc "build release package"
68
+ task :release => :all
45
69
 
46
70
  unless EXTCONF.empty?
47
71
  RUBYSET ||= (ENV["RUBYSET"] || "").split(",")
@@ -54,24 +78,43 @@ unless EXTCONF.empty?
54
78
  | variable for set ruby interpreters by comma separated.
55
79
  |
56
80
  | e.g.) $ rake RUBYSET=ruby
57
- | or) $ rake RUBYSET=ruby20,ruby21,ruby22
81
+ | or) $ rake RUBYSET=ruby21,ruby22,ruby23
58
82
  |
59
83
  EOS
60
84
  else
61
- platforms = RUBYSET.map { |ruby| `#{ruby} --disable gems -rrbconfig -e "puts RbConfig::CONFIG['arch']"`.chomp }
85
+ platforms = RUBYSET.map { |ruby| `#{ruby} --disable-gems -e "puts RUBY_PLATFORM"`.chomp }
62
86
  platforms1 = platforms.uniq
63
87
  unless platforms1.size == 1 && !platforms1[0].empty?
64
- raise "different platforms (#{Hash[*RUBYSET.zip(platforms).flatten].inspect})"
88
+ abort <<-EOS
89
+ #{__FILE__}:#{__LINE__}: different platforms:
90
+ #{RUBYSET.zip(platforms).map { |ruby, platform| "%24s => %s" % [ruby, platform] }.join("\n")}
91
+ ABORTED.
92
+ EOS
65
93
  end
66
94
  PLATFORM = platforms1[0]
67
95
 
68
96
  RUBY_VERSIONS = RUBYSET.map do |ruby|
69
- ver = `#{ruby} --disable gem -rrbconfig -e "puts RbConfig::CONFIG['ruby_version']"`.slice(/\d+\.\d+/)
97
+ ver = `#{ruby} --disable-gems -e "puts RUBY_VERSION"`.slice(/\d+\.\d+/)
70
98
  raise "failed ruby checking - ``#{ruby}''" unless $?.success?
71
99
  [ver, ruby]
72
100
  end
73
- SOFILES_SET = RUBY_VERSIONS.map { |(ver, ruby)| ["lib/#{ver}/#{GEMSTUB.name}.so", ruby] }
74
- SOFILES = SOFILES_SET.map { |(lib, ruby)| lib }
101
+
102
+ SOFILES_SET = RUBY_VERSIONS.map { |(ver, ruby)|
103
+ EXTCONF.map { |extconf|
104
+ extdir = Pathname.new(extconf).cleanpath.dirname.to_s
105
+ case
106
+ when soname = EXTMAP[extdir.sub(/^ext\//i, "")]
107
+ soname = soname.sub(/\.so$/i, "")
108
+ when extdir == "ext" || extdir == "."
109
+ soname = GEMSTUB.name
110
+ else
111
+ soname = File.basename(extdir)
112
+ end
113
+
114
+ [ruby, File.join("lib", "#{soname.sub(/(?<=\/)|^(?!.*\/)/, "#{ver}/")}.so"), extconf]
115
+ }
116
+ }.flatten(1)
117
+ SOFILES = SOFILES_SET.map { |(ruby, sopath, extconf)| sopath }
75
118
 
76
119
  GEMSTUB_NATIVE = GEMSTUB.dup
77
120
  GEMSTUB_NATIVE.files += SOFILES
@@ -99,7 +142,7 @@ unless EXTCONF.empty?
99
142
  desc "build c-extension libraries"
100
143
  task "sofiles" => SOFILES
101
144
 
102
- SOFILES_SET.each do |(soname, ruby)|
145
+ SOFILES_SET.each do |(ruby, soname, extconf)|
103
146
  sodir = File.dirname(soname)
104
147
  makefile = File.join(sodir, "Makefile")
105
148
 
@@ -109,9 +152,10 @@ unless EXTCONF.empty?
109
152
  directory sodir
110
153
 
111
154
  desc "generate Makefile for binary extension library"
112
- file makefile => [sodir] + EXTCONF do
155
+ file makefile => [sodir, extconf] do
156
+ rel_extconf = Pathname.new(extconf).relative_path_from(Pathname.new(sodir)).to_s
113
157
  cd sodir do
114
- sh "#{ruby} ../../#{EXTCONF[0]} \"--ruby=#{ruby}\""
158
+ sh *%W"#{ruby} #{rel_extconf} --ruby=#{ruby} #{ENV["EXTCONF"]}"
115
159
  end
116
160
  end
117
161
 
@@ -127,6 +171,7 @@ end
127
171
 
128
172
 
129
173
  task :all => GEMFILE
174
+ task :tryout => GEMFILE_TRYOUT
130
175
 
131
176
  desc "generate local rdoc"
132
177
  task :rdoc => DOC + LIB do
@@ -153,6 +198,15 @@ file GEMFILE => DOC + EXT + EXTCONF + BIN + LIB + SPEC + TEST + EXAMPLE + RAKEFI
153
198
  sh "gem build #{GEMSPEC}"
154
199
  end
155
200
 
201
+ file GEMFILE_TRYOUT => DOC + EXT + EXTCONF + BIN + LIB + SPEC + TEST + EXAMPLE + RAKEFILE + [GEMSPEC_TRYOUT] do
202
+ #file GEMFILE_TRYOUT do
203
+ sh "gem build #{GEMSPEC_TRYOUT}"
204
+ end
205
+
156
206
  file GEMSPEC => RAKEFILE do
157
207
  File.write(GEMSPEC, GEMSTUB.to_ruby, mode: "wb")
158
208
  end
209
+
210
+ file GEMSPEC_TRYOUT => RAKEFILE do
211
+ File.write(GEMSPEC_TRYOUT, GEMSTUB_TRYOUT.to_ruby, mode: "wb")
212
+ end
@@ -0,0 +1,68 @@
1
+ zstd - library files
2
+ ================================
3
+
4
+ The __lib__ directory contains several files, but depending on target use case, some of them may not be necessary.
5
+
6
+ #### Minimal library files
7
+
8
+ To build the zstd library the following files are required:
9
+
10
+ - [common/bitstream.h](common/bitstream.h)
11
+ - [common/error_private.h](common/error_private.h)
12
+ - [common/error_public.h](common/error_public.h)
13
+ - common/fse.h
14
+ - common/fse_decompress.c
15
+ - common/huf.h
16
+ - [common/mem.h](common/mem.h)
17
+ - [common/zstd.h]
18
+ - common/zstd_internal.h
19
+ - compress/fse_compress.c
20
+ - compress/huf_compress.c
21
+ - compress/zstd_compress.c
22
+ - compress/zstd_opt.h
23
+ - decompress/huf_decompress.c
24
+ - decompress/zstd_decompress.c
25
+
26
+ Stable API is exposed in [common/zstd.h].
27
+ Advanced and experimental API can be enabled by defining `ZSTD_STATIC_LINKING_ONLY`.
28
+ Never use them with a dynamic library, as their definition may change in future versions.
29
+
30
+ [common/zstd.h]: common/zstd.h
31
+
32
+
33
+ #### Separate compressor and decompressor
34
+
35
+ To build a separate zstd compressor all files from `common/` and `compressor/` directories are required.
36
+ In a similar way to build a separate zstd decompressor all files from `common/` and `decompressor/` directories are needed.
37
+
38
+
39
+ #### Buffered streaming
40
+
41
+ This complementary API makes streaming integration easier.
42
+ It is used by `zstd` command line utility, and [7zip plugin](http://mcmilk.de/projects/7-Zip-ZStd) :
43
+
44
+ - common/zbuff.h
45
+ - compress/zbuff_compress.c
46
+ - decompress/zbuff_decompress.c
47
+
48
+
49
+ #### Dictionary builder
50
+
51
+ In order to create dictionaries from some training sets,
52
+ it's needed to include all files from [dictBuilder directory](dictBuilder/)
53
+
54
+
55
+ #### Legacy support
56
+
57
+ Zstandard can decode previous formats, starting from v0.1.
58
+ Support for these format is provided in [folder legacy](legacy/).
59
+ It's also required to compile the library with `ZSTD_LEGACY_SUPPORT = 1`.
60
+
61
+
62
+ #### Miscellaneous
63
+
64
+ The other files are not source code. There are :
65
+
66
+ - LICENSE : contains the BSD license text
67
+ - Makefile : script to compile or install zstd library (static or dynamic)
68
+ - libzstd.pc.in : for pkg-config (make install)
@@ -0,0 +1,414 @@
1
+ /* ******************************************************************
2
+ bitstream
3
+ Part of FSE library
4
+ header file (to include)
5
+ Copyright (C) 2013-2016, Yann Collet.
6
+
7
+ BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
8
+
9
+ Redistribution and use in source and binary forms, with or without
10
+ modification, are permitted provided that the following conditions are
11
+ met:
12
+
13
+ * Redistributions of source code must retain the above copyright
14
+ notice, this list of conditions and the following disclaimer.
15
+ * Redistributions in binary form must reproduce the above
16
+ copyright notice, this list of conditions and the following disclaimer
17
+ in the documentation and/or other materials provided with the
18
+ distribution.
19
+
20
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
21
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
22
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
23
+ A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
24
+ OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
25
+ SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
26
+ LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
27
+ DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
28
+ THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
29
+ (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
30
+ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
31
+
32
+ You can contact the author at :
33
+ - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
34
+ ****************************************************************** */
35
+ #ifndef BITSTREAM_H_MODULE
36
+ #define BITSTREAM_H_MODULE
37
+
38
+ #if defined (__cplusplus)
39
+ extern "C" {
40
+ #endif
41
+
42
+
43
+ /*
44
+ * This API consists of small unitary functions, which must be inlined for best performance.
45
+ * Since link-time-optimization is not available for all compilers,
46
+ * these functions are defined into a .h to be included.
47
+ */
48
+
49
+ /*-****************************************
50
+ * Dependencies
51
+ ******************************************/
52
+ #include "mem.h" /* unaligned access routines */
53
+ #include "error_private.h" /* error codes and messages */
54
+
55
+
56
+ /*=========================================
57
+ * Target specific
58
+ =========================================*/
59
+ #if defined(__BMI__) && defined(__GNUC__)
60
+ # include <immintrin.h> /* support for bextr (experimental) */
61
+ #endif
62
+
63
+
64
+ /*-******************************************
65
+ * bitStream encoding API (write forward)
66
+ ********************************************/
67
+ /* bitStream can mix input from multiple sources.
68
+ * A critical property of these streams is that they encode and decode in **reverse** direction.
69
+ * So the first bit sequence you add will be the last to be read, like a LIFO stack.
70
+ */
71
+ typedef struct
72
+ {
73
+ size_t bitContainer;
74
+ int bitPos;
75
+ char* startPtr;
76
+ char* ptr;
77
+ char* endPtr;
78
+ } BIT_CStream_t;
79
+
80
+ MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, void* dstBuffer, size_t dstCapacity);
81
+ MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC, size_t value, unsigned nbBits);
82
+ MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC);
83
+ MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC);
84
+
85
+ /* Start with initCStream, providing the size of buffer to write into.
86
+ * bitStream will never write outside of this buffer.
87
+ * `dstCapacity` must be >= sizeof(bitD->bitContainer), otherwise @return will be an error code.
88
+ *
89
+ * bits are first added to a local register.
90
+ * Local register is size_t, hence 64-bits on 64-bits systems, or 32-bits on 32-bits systems.
91
+ * Writing data into memory is an explicit operation, performed by the flushBits function.
92
+ * Hence keep track how many bits are potentially stored into local register to avoid register overflow.
93
+ * After a flushBits, a maximum of 7 bits might still be stored into local register.
94
+ *
95
+ * Avoid storing elements of more than 24 bits if you want compatibility with 32-bits bitstream readers.
96
+ *
97
+ * Last operation is to close the bitStream.
98
+ * The function returns the final size of CStream in bytes.
99
+ * If data couldn't fit into `dstBuffer`, it will return a 0 ( == not storable)
100
+ */
101
+
102
+
103
+ /*-********************************************
104
+ * bitStream decoding API (read backward)
105
+ **********************************************/
106
+ typedef struct
107
+ {
108
+ size_t bitContainer;
109
+ unsigned bitsConsumed;
110
+ const char* ptr;
111
+ const char* start;
112
+ } BIT_DStream_t;
113
+
114
+ typedef enum { BIT_DStream_unfinished = 0,
115
+ BIT_DStream_endOfBuffer = 1,
116
+ BIT_DStream_completed = 2,
117
+ BIT_DStream_overflow = 3 } BIT_DStream_status; /* result of BIT_reloadDStream() */
118
+ /* 1,2,4,8 would be better for bitmap combinations, but slows down performance a bit ... :( */
119
+
120
+ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize);
121
+ MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, unsigned nbBits);
122
+ MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD);
123
+ MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* bitD);
124
+
125
+
126
+ /* Start by invoking BIT_initDStream().
127
+ * A chunk of the bitStream is then stored into a local register.
128
+ * Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t).
129
+ * You can then retrieve bitFields stored into the local register, **in reverse order**.
130
+ * Local register is explicitly reloaded from memory by the BIT_reloadDStream() method.
131
+ * A reload guarantee a minimum of ((8*sizeof(bitD->bitContainer))-7) bits when its result is BIT_DStream_unfinished.
132
+ * Otherwise, it can be less than that, so proceed accordingly.
133
+ * Checking if DStream has reached its end can be performed with BIT_endOfDStream().
134
+ */
135
+
136
+
137
+ /*-****************************************
138
+ * unsafe API
139
+ ******************************************/
140
+ MEM_STATIC void BIT_addBitsFast(BIT_CStream_t* bitC, size_t value, unsigned nbBits);
141
+ /* faster, but works only if value is "clean", meaning all high bits above nbBits are 0 */
142
+
143
+ MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC);
144
+ /* unsafe version; does not check buffer overflow */
145
+
146
+ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, unsigned nbBits);
147
+ /* faster, but works only if nbBits >= 1 */
148
+
149
+
150
+
151
+ /*-**************************************************************
152
+ * Internal functions
153
+ ****************************************************************/
154
+ MEM_STATIC unsigned BIT_highbit32 (register U32 val)
155
+ {
156
+ # if defined(_MSC_VER) /* Visual */
157
+ unsigned long r=0;
158
+ _BitScanReverse ( &r, val );
159
+ return (unsigned) r;
160
+ # elif defined(__GNUC__) && (__GNUC__ >= 3) /* Use GCC Intrinsic */
161
+ return 31 - __builtin_clz (val);
162
+ # else /* Software version */
163
+ static const unsigned DeBruijnClz[32] = { 0, 9, 1, 10, 13, 21, 2, 29, 11, 14, 16, 18, 22, 25, 3, 30, 8, 12, 20, 28, 15, 17, 24, 7, 19, 27, 23, 6, 26, 5, 4, 31 };
164
+ U32 v = val;
165
+ v |= v >> 1;
166
+ v |= v >> 2;
167
+ v |= v >> 4;
168
+ v |= v >> 8;
169
+ v |= v >> 16;
170
+ return DeBruijnClz[ (U32) (v * 0x07C4ACDDU) >> 27];
171
+ # endif
172
+ }
173
+
174
+ /*===== Local Constants =====*/
175
+ static const unsigned BIT_mask[] = { 0, 1, 3, 7, 0xF, 0x1F, 0x3F, 0x7F, 0xFF, 0x1FF, 0x3FF, 0x7FF, 0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0x1FFFF, 0x3FFFF, 0x7FFFF, 0xFFFFF, 0x1FFFFF, 0x3FFFFF, 0x7FFFFF, 0xFFFFFF, 0x1FFFFFF, 0x3FFFFFF }; /* up to 26 bits */
176
+
177
+
178
+ /*-**************************************************************
179
+ * bitStream encoding
180
+ ****************************************************************/
181
+ /*! BIT_initCStream() :
182
+ * `dstCapacity` must be > sizeof(void*)
183
+ * @return : 0 if success,
184
+ otherwise an error code (can be tested using ERR_isError() ) */
185
+ MEM_STATIC size_t BIT_initCStream(BIT_CStream_t* bitC, void* startPtr, size_t dstCapacity)
186
+ {
187
+ bitC->bitContainer = 0;
188
+ bitC->bitPos = 0;
189
+ bitC->startPtr = (char*)startPtr;
190
+ bitC->ptr = bitC->startPtr;
191
+ bitC->endPtr = bitC->startPtr + dstCapacity - sizeof(bitC->ptr);
192
+ if (dstCapacity <= sizeof(bitC->ptr)) return ERROR(dstSize_tooSmall);
193
+ return 0;
194
+ }
195
+
196
+ /*! BIT_addBits() :
197
+ can add up to 26 bits into `bitC`.
198
+ Does not check for register overflow ! */
199
+ MEM_STATIC void BIT_addBits(BIT_CStream_t* bitC, size_t value, unsigned nbBits)
200
+ {
201
+ bitC->bitContainer |= (value & BIT_mask[nbBits]) << bitC->bitPos;
202
+ bitC->bitPos += nbBits;
203
+ }
204
+
205
+ /*! BIT_addBitsFast() :
206
+ * works only if `value` is _clean_, meaning all high bits above nbBits are 0 */
207
+ MEM_STATIC void BIT_addBitsFast(BIT_CStream_t* bitC, size_t value, unsigned nbBits)
208
+ {
209
+ bitC->bitContainer |= value << bitC->bitPos;
210
+ bitC->bitPos += nbBits;
211
+ }
212
+
213
+ /*! BIT_flushBitsFast() :
214
+ * unsafe version; does not check buffer overflow */
215
+ MEM_STATIC void BIT_flushBitsFast(BIT_CStream_t* bitC)
216
+ {
217
+ size_t const nbBytes = bitC->bitPos >> 3;
218
+ MEM_writeLEST(bitC->ptr, bitC->bitContainer);
219
+ bitC->ptr += nbBytes;
220
+ bitC->bitPos &= 7;
221
+ bitC->bitContainer >>= nbBytes*8; /* if bitPos >= sizeof(bitContainer)*8 --> undefined behavior */
222
+ }
223
+
224
+ /*! BIT_flushBits() :
225
+ * safe version; check for buffer overflow, and prevents it.
226
+ * note : does not signal buffer overflow. This will be revealed later on using BIT_closeCStream() */
227
+ MEM_STATIC void BIT_flushBits(BIT_CStream_t* bitC)
228
+ {
229
+ size_t const nbBytes = bitC->bitPos >> 3;
230
+ MEM_writeLEST(bitC->ptr, bitC->bitContainer);
231
+ bitC->ptr += nbBytes;
232
+ if (bitC->ptr > bitC->endPtr) bitC->ptr = bitC->endPtr;
233
+ bitC->bitPos &= 7;
234
+ bitC->bitContainer >>= nbBytes*8; /* if bitPos >= sizeof(bitContainer)*8 --> undefined behavior */
235
+ }
236
+
237
+ /*! BIT_closeCStream() :
238
+ * @return : size of CStream, in bytes,
239
+ or 0 if it could not fit into dstBuffer */
240
+ MEM_STATIC size_t BIT_closeCStream(BIT_CStream_t* bitC)
241
+ {
242
+ BIT_addBitsFast(bitC, 1, 1); /* endMark */
243
+ BIT_flushBits(bitC);
244
+
245
+ if (bitC->ptr >= bitC->endPtr) return 0; /* doesn't fit within authorized budget : cancel */
246
+
247
+ return (bitC->ptr - bitC->startPtr) + (bitC->bitPos > 0);
248
+ }
249
+
250
+
251
+ /*-********************************************************
252
+ * bitStream decoding
253
+ **********************************************************/
254
+ /*! BIT_initDStream() :
255
+ * Initialize a BIT_DStream_t.
256
+ * `bitD` : a pointer to an already allocated BIT_DStream_t structure.
257
+ * `srcSize` must be the *exact* size of the bitStream, in bytes.
258
+ * @return : size of stream (== srcSize) or an errorCode if a problem is detected
259
+ */
260
+ MEM_STATIC size_t BIT_initDStream(BIT_DStream_t* bitD, const void* srcBuffer, size_t srcSize)
261
+ {
262
+ if (srcSize < 1) { memset(bitD, 0, sizeof(*bitD)); return ERROR(srcSize_wrong); }
263
+
264
+ if (srcSize >= sizeof(bitD->bitContainer)) { /* normal case */
265
+ bitD->start = (const char*)srcBuffer;
266
+ bitD->ptr = (const char*)srcBuffer + srcSize - sizeof(bitD->bitContainer);
267
+ bitD->bitContainer = MEM_readLEST(bitD->ptr);
268
+ { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1];
269
+ bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0;
270
+ if (lastByte == 0) return ERROR(GENERIC); /* endMark not present */ }
271
+ } else {
272
+ bitD->start = (const char*)srcBuffer;
273
+ bitD->ptr = bitD->start;
274
+ bitD->bitContainer = *(const BYTE*)(bitD->start);
275
+ switch(srcSize)
276
+ {
277
+ case 7: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[6]) << (sizeof(bitD->bitContainer)*8 - 16);
278
+ case 6: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[5]) << (sizeof(bitD->bitContainer)*8 - 24);
279
+ case 5: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[4]) << (sizeof(bitD->bitContainer)*8 - 32);
280
+ case 4: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[3]) << 24;
281
+ case 3: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[2]) << 16;
282
+ case 2: bitD->bitContainer += (size_t)(((const BYTE*)(srcBuffer))[1]) << 8;
283
+ default:;
284
+ }
285
+ { BYTE const lastByte = ((const BYTE*)srcBuffer)[srcSize-1];
286
+ bitD->bitsConsumed = lastByte ? 8 - BIT_highbit32(lastByte) : 0;
287
+ if (lastByte == 0) return ERROR(GENERIC); /* endMark not present */ }
288
+ bitD->bitsConsumed += (U32)(sizeof(bitD->bitContainer) - srcSize)*8;
289
+ }
290
+
291
+ return srcSize;
292
+ }
293
+
294
+ MEM_STATIC size_t BIT_getUpperBits(size_t bitContainer, U32 const start)
295
+ {
296
+ return bitContainer >> start;
297
+ }
298
+
299
+ MEM_STATIC size_t BIT_getMiddleBits(size_t bitContainer, U32 const start, U32 const nbBits)
300
+ {
301
+ #if defined(__BMI__) && defined(__GNUC__) /* experimental */
302
+ # if defined(__x86_64__)
303
+ if (sizeof(bitContainer)==8)
304
+ return _bextr_u64(bitContainer, start, nbBits);
305
+ else
306
+ # endif
307
+ return _bextr_u32(bitContainer, start, nbBits);
308
+ #else
309
+ return (bitContainer >> start) & BIT_mask[nbBits];
310
+ #endif
311
+ }
312
+
313
+ MEM_STATIC size_t BIT_getLowerBits(size_t bitContainer, U32 const nbBits)
314
+ {
315
+ return bitContainer & BIT_mask[nbBits];
316
+ }
317
+
318
+ /*! BIT_lookBits() :
319
+ * Provides next n bits from local register.
320
+ * local register is not modified.
321
+ * On 32-bits, maxNbBits==24.
322
+ * On 64-bits, maxNbBits==56.
323
+ * @return : value extracted
324
+ */
325
+ MEM_STATIC size_t BIT_lookBits(const BIT_DStream_t* bitD, U32 nbBits)
326
+ {
327
+ #if defined(__BMI__) && defined(__GNUC__) /* experimental; fails if bitD->bitsConsumed + nbBits > sizeof(bitD->bitContainer)*8 */
328
+ return BIT_getMiddleBits(bitD->bitContainer, (sizeof(bitD->bitContainer)*8) - bitD->bitsConsumed - nbBits, nbBits);
329
+ #else
330
+ U32 const bitMask = sizeof(bitD->bitContainer)*8 - 1;
331
+ return ((bitD->bitContainer << (bitD->bitsConsumed & bitMask)) >> 1) >> ((bitMask-nbBits) & bitMask);
332
+ #endif
333
+ }
334
+
335
+ /*! BIT_lookBitsFast() :
336
+ * unsafe version; only works only if nbBits >= 1 */
337
+ MEM_STATIC size_t BIT_lookBitsFast(const BIT_DStream_t* bitD, U32 nbBits)
338
+ {
339
+ U32 const bitMask = sizeof(bitD->bitContainer)*8 - 1;
340
+ return (bitD->bitContainer << (bitD->bitsConsumed & bitMask)) >> (((bitMask+1)-nbBits) & bitMask);
341
+ }
342
+
343
+ MEM_STATIC void BIT_skipBits(BIT_DStream_t* bitD, U32 nbBits)
344
+ {
345
+ bitD->bitsConsumed += nbBits;
346
+ }
347
+
348
+ /*! BIT_readBits() :
349
+ * Read (consume) next n bits from local register and update.
350
+ * Pay attention to not read more than nbBits contained into local register.
351
+ * @return : extracted value.
352
+ */
353
+ MEM_STATIC size_t BIT_readBits(BIT_DStream_t* bitD, U32 nbBits)
354
+ {
355
+ size_t const value = BIT_lookBits(bitD, nbBits);
356
+ BIT_skipBits(bitD, nbBits);
357
+ return value;
358
+ }
359
+
360
+ /*! BIT_readBitsFast() :
361
+ * unsafe version; only works only if nbBits >= 1 */
362
+ MEM_STATIC size_t BIT_readBitsFast(BIT_DStream_t* bitD, U32 nbBits)
363
+ {
364
+ size_t const value = BIT_lookBitsFast(bitD, nbBits);
365
+ BIT_skipBits(bitD, nbBits);
366
+ return value;
367
+ }
368
+
369
+ /*! BIT_reloadDStream() :
370
+ * Refill `BIT_DStream_t` from src buffer previously defined (see BIT_initDStream() ).
371
+ * This function is safe, it guarantees it will not read beyond src buffer.
372
+ * @return : status of `BIT_DStream_t` internal register.
373
+ if status == unfinished, internal register is filled with >= (sizeof(bitD->bitContainer)*8 - 7) bits */
374
+ MEM_STATIC BIT_DStream_status BIT_reloadDStream(BIT_DStream_t* bitD)
375
+ {
376
+ if (bitD->bitsConsumed > (sizeof(bitD->bitContainer)*8)) /* should not happen => corruption detected */
377
+ return BIT_DStream_overflow;
378
+
379
+ if (bitD->ptr >= bitD->start + sizeof(bitD->bitContainer)) {
380
+ bitD->ptr -= bitD->bitsConsumed >> 3;
381
+ bitD->bitsConsumed &= 7;
382
+ bitD->bitContainer = MEM_readLEST(bitD->ptr);
383
+ return BIT_DStream_unfinished;
384
+ }
385
+ if (bitD->ptr == bitD->start) {
386
+ if (bitD->bitsConsumed < sizeof(bitD->bitContainer)*8) return BIT_DStream_endOfBuffer;
387
+ return BIT_DStream_completed;
388
+ }
389
+ { U32 nbBytes = bitD->bitsConsumed >> 3;
390
+ BIT_DStream_status result = BIT_DStream_unfinished;
391
+ if (bitD->ptr - nbBytes < bitD->start) {
392
+ nbBytes = (U32)(bitD->ptr - bitD->start); /* ptr > start */
393
+ result = BIT_DStream_endOfBuffer;
394
+ }
395
+ bitD->ptr -= nbBytes;
396
+ bitD->bitsConsumed -= nbBytes*8;
397
+ bitD->bitContainer = MEM_readLEST(bitD->ptr); /* reminder : srcSize > sizeof(bitD) */
398
+ return result;
399
+ }
400
+ }
401
+
402
+ /*! BIT_endOfDStream() :
403
+ * @return Tells if DStream has exactly reached its end (all bits consumed).
404
+ */
405
+ MEM_STATIC unsigned BIT_endOfDStream(const BIT_DStream_t* DStream)
406
+ {
407
+ return ((DStream->ptr == DStream->start) && (DStream->bitsConsumed == sizeof(DStream->bitContainer)*8));
408
+ }
409
+
410
+ #if defined (__cplusplus)
411
+ }
412
+ #endif
413
+
414
+ #endif /* BITSTREAM_H_MODULE */