zstd-ruby 0.1.0

Sign up to get free protection for your applications and to get access to all the features.
Files changed (71) hide show
  1. checksums.yaml +7 -0
  2. data/.gitignore +19 -0
  3. data/.gitmodules +3 -0
  4. data/.rspec +2 -0
  5. data/.travis.yml +11 -0
  6. data/CODE_OF_CONDUCT.md +74 -0
  7. data/Gemfile +4 -0
  8. data/LICENSE.txt +29 -0
  9. data/README.md +63 -0
  10. data/Rakefile +22 -0
  11. data/bin/console +14 -0
  12. data/bin/setup +8 -0
  13. data/ext/zstdruby/extconf.rb +23 -0
  14. data/ext/zstdruby/libzstd/.gitignore +2 -0
  15. data/ext/zstdruby/libzstd/Makefile +133 -0
  16. data/ext/zstdruby/libzstd/README.md +77 -0
  17. data/ext/zstdruby/libzstd/common/bitstream.h +414 -0
  18. data/ext/zstdruby/libzstd/common/entropy_common.c +227 -0
  19. data/ext/zstdruby/libzstd/common/error_private.c +43 -0
  20. data/ext/zstdruby/libzstd/common/error_private.h +76 -0
  21. data/ext/zstdruby/libzstd/common/fse.h +668 -0
  22. data/ext/zstdruby/libzstd/common/fse_decompress.c +329 -0
  23. data/ext/zstdruby/libzstd/common/huf.h +238 -0
  24. data/ext/zstdruby/libzstd/common/mem.h +372 -0
  25. data/ext/zstdruby/libzstd/common/xxhash.c +867 -0
  26. data/ext/zstdruby/libzstd/common/xxhash.h +309 -0
  27. data/ext/zstdruby/libzstd/common/zstd_common.c +77 -0
  28. data/ext/zstdruby/libzstd/common/zstd_errors.h +60 -0
  29. data/ext/zstdruby/libzstd/common/zstd_internal.h +270 -0
  30. data/ext/zstdruby/libzstd/compress/fse_compress.c +850 -0
  31. data/ext/zstdruby/libzstd/compress/huf_compress.c +609 -0
  32. data/ext/zstdruby/libzstd/compress/zstd_compress.c +3291 -0
  33. data/ext/zstdruby/libzstd/compress/zstd_opt.h +919 -0
  34. data/ext/zstdruby/libzstd/decompress/huf_decompress.c +885 -0
  35. data/ext/zstdruby/libzstd/decompress/zstd_decompress.c +2154 -0
  36. data/ext/zstdruby/libzstd/deprecated/zbuff.h +210 -0
  37. data/ext/zstdruby/libzstd/deprecated/zbuff_compress.c +145 -0
  38. data/ext/zstdruby/libzstd/deprecated/zbuff_decompress.c +74 -0
  39. data/ext/zstdruby/libzstd/dictBuilder/divsufsort.c +1913 -0
  40. data/ext/zstdruby/libzstd/dictBuilder/divsufsort.h +67 -0
  41. data/ext/zstdruby/libzstd/dictBuilder/zdict.c +1012 -0
  42. data/ext/zstdruby/libzstd/dictBuilder/zdict.h +111 -0
  43. data/ext/zstdruby/libzstd/dll/example/Makefile +47 -0
  44. data/ext/zstdruby/libzstd/dll/example/README.md +69 -0
  45. data/ext/zstdruby/libzstd/dll/example/build_package.bat +17 -0
  46. data/ext/zstdruby/libzstd/dll/example/fullbench-dll.sln +25 -0
  47. data/ext/zstdruby/libzstd/dll/example/fullbench-dll.vcxproj +179 -0
  48. data/ext/zstdruby/libzstd/dll/libzstd.def +86 -0
  49. data/ext/zstdruby/libzstd/legacy/zstd_legacy.h +259 -0
  50. data/ext/zstdruby/libzstd/legacy/zstd_v01.c +2095 -0
  51. data/ext/zstdruby/libzstd/legacy/zstd_v01.h +80 -0
  52. data/ext/zstdruby/libzstd/legacy/zstd_v02.c +3518 -0
  53. data/ext/zstdruby/libzstd/legacy/zstd_v02.h +79 -0
  54. data/ext/zstdruby/libzstd/legacy/zstd_v03.c +3159 -0
  55. data/ext/zstdruby/libzstd/legacy/zstd_v03.h +79 -0
  56. data/ext/zstdruby/libzstd/legacy/zstd_v04.c +3795 -0
  57. data/ext/zstdruby/libzstd/legacy/zstd_v04.h +128 -0
  58. data/ext/zstdruby/libzstd/legacy/zstd_v05.c +4056 -0
  59. data/ext/zstdruby/libzstd/legacy/zstd_v05.h +149 -0
  60. data/ext/zstdruby/libzstd/legacy/zstd_v06.c +4167 -0
  61. data/ext/zstdruby/libzstd/legacy/zstd_v06.h +159 -0
  62. data/ext/zstdruby/libzstd/legacy/zstd_v07.c +4540 -0
  63. data/ext/zstdruby/libzstd/legacy/zstd_v07.h +173 -0
  64. data/ext/zstdruby/libzstd/libzstd.pc.in +14 -0
  65. data/ext/zstdruby/libzstd/zstd.h +673 -0
  66. data/ext/zstdruby/zstdruby.c +117 -0
  67. data/ext/zstdruby/zstdruby.h +6 -0
  68. data/lib/zstd-ruby.rb +6 -0
  69. data/lib/zstd-ruby/version.rb +3 -0
  70. data/zstd-ruby.gemspec +37 -0
  71. metadata +170 -0
@@ -0,0 +1,43 @@
1
+ /**
2
+ * Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
3
+ * All rights reserved.
4
+ *
5
+ * This source code is licensed under the BSD-style license found in the
6
+ * LICENSE file in the root directory of this source tree. An additional grant
7
+ * of patent rights can be found in the PATENTS file in the same directory.
8
+ */
9
+
10
+ /* The purpose of this file is to have a single list of error strings embedded in binary */
11
+
12
+ #include "error_private.h"
13
+
14
+ const char* ERR_getErrorString(ERR_enum code)
15
+ {
16
+ static const char* const notErrorCode = "Unspecified error code";
17
+ switch( code )
18
+ {
19
+ case PREFIX(no_error): return "No error detected";
20
+ case PREFIX(GENERIC): return "Error (generic)";
21
+ case PREFIX(prefix_unknown): return "Unknown frame descriptor";
22
+ case PREFIX(version_unsupported): return "Version not supported";
23
+ case PREFIX(parameter_unknown): return "Unknown parameter type";
24
+ case PREFIX(frameParameter_unsupported): return "Unsupported frame parameter";
25
+ case PREFIX(frameParameter_unsupportedBy32bits): return "Frame parameter unsupported in 32-bits mode";
26
+ case PREFIX(frameParameter_windowTooLarge): return "Frame requires too much memory for decoding";
27
+ case PREFIX(compressionParameter_unsupported): return "Compression parameter is out of bound";
28
+ case PREFIX(init_missing): return "Context should be init first";
29
+ case PREFIX(memory_allocation): return "Allocation error : not enough memory";
30
+ case PREFIX(stage_wrong): return "Operation not authorized at current processing stage";
31
+ case PREFIX(dstSize_tooSmall): return "Destination buffer is too small";
32
+ case PREFIX(srcSize_wrong): return "Src size incorrect";
33
+ case PREFIX(corruption_detected): return "Corrupted block detected";
34
+ case PREFIX(checksum_wrong): return "Restored data doesn't match checksum";
35
+ case PREFIX(tableLog_tooLarge): return "tableLog requires too much memory : unsupported";
36
+ case PREFIX(maxSymbolValue_tooLarge): return "Unsupported max Symbol Value : too large";
37
+ case PREFIX(maxSymbolValue_tooSmall): return "Specified maxSymbolValue is too small";
38
+ case PREFIX(dictionary_corrupted): return "Dictionary is corrupted";
39
+ case PREFIX(dictionary_wrong): return "Dictionary mismatch";
40
+ case PREFIX(maxCode):
41
+ default: return notErrorCode;
42
+ }
43
+ }
@@ -0,0 +1,76 @@
1
+ /**
2
+ * Copyright (c) 2016-present, Yann Collet, Facebook, Inc.
3
+ * All rights reserved.
4
+ *
5
+ * This source code is licensed under the BSD-style license found in the
6
+ * LICENSE file in the root directory of this source tree. An additional grant
7
+ * of patent rights can be found in the PATENTS file in the same directory.
8
+ */
9
+
10
+ /* Note : this module is expected to remain private, do not expose it */
11
+
12
+ #ifndef ERROR_H_MODULE
13
+ #define ERROR_H_MODULE
14
+
15
+ #if defined (__cplusplus)
16
+ extern "C" {
17
+ #endif
18
+
19
+
20
+ /* ****************************************
21
+ * Dependencies
22
+ ******************************************/
23
+ #include <stddef.h> /* size_t */
24
+ #include "zstd_errors.h" /* enum list */
25
+
26
+
27
+ /* ****************************************
28
+ * Compiler-specific
29
+ ******************************************/
30
+ #if defined(__GNUC__)
31
+ # define ERR_STATIC static __attribute__((unused))
32
+ #elif defined (__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */)
33
+ # define ERR_STATIC static inline
34
+ #elif defined(_MSC_VER)
35
+ # define ERR_STATIC static __inline
36
+ #else
37
+ # define ERR_STATIC static /* this version may generate warnings for unused static functions; disable the relevant warning */
38
+ #endif
39
+
40
+
41
+ /*-****************************************
42
+ * Customization (error_public.h)
43
+ ******************************************/
44
+ typedef ZSTD_ErrorCode ERR_enum;
45
+ #define PREFIX(name) ZSTD_error_##name
46
+
47
+
48
+ /*-****************************************
49
+ * Error codes handling
50
+ ******************************************/
51
+ #ifdef ERROR
52
+ # undef ERROR /* reported already defined on VS 2015 (Rich Geldreich) */
53
+ #endif
54
+ #define ERROR(name) ((size_t)-PREFIX(name))
55
+
56
+ ERR_STATIC unsigned ERR_isError(size_t code) { return (code > ERROR(maxCode)); }
57
+
58
+ ERR_STATIC ERR_enum ERR_getErrorCode(size_t code) { if (!ERR_isError(code)) return (ERR_enum)0; return (ERR_enum) (0-code); }
59
+
60
+
61
+ /*-****************************************
62
+ * Error Strings
63
+ ******************************************/
64
+
65
+ const char* ERR_getErrorString(ERR_enum code); /* error_private.c */
66
+
67
+ ERR_STATIC const char* ERR_getErrorName(size_t code)
68
+ {
69
+ return ERR_getErrorString(ERR_getErrorCode(code));
70
+ }
71
+
72
+ #if defined (__cplusplus)
73
+ }
74
+ #endif
75
+
76
+ #endif /* ERROR_H_MODULE */
@@ -0,0 +1,668 @@
1
+ /* ******************************************************************
2
+ FSE : Finite State Entropy codec
3
+ Public Prototypes declaration
4
+ Copyright (C) 2013-2016, Yann Collet.
5
+
6
+ BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
7
+
8
+ Redistribution and use in source and binary forms, with or without
9
+ modification, are permitted provided that the following conditions are
10
+ met:
11
+
12
+ * Redistributions of source code must retain the above copyright
13
+ notice, this list of conditions and the following disclaimer.
14
+ * Redistributions in binary form must reproduce the above
15
+ copyright notice, this list of conditions and the following disclaimer
16
+ in the documentation and/or other materials provided with the
17
+ distribution.
18
+
19
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22
+ A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23
+ OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24
+ SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25
+ LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26
+ DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27
+ THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28
+ (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29
+ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30
+
31
+ You can contact the author at :
32
+ - Source repository : https://github.com/Cyan4973/FiniteStateEntropy
33
+ ****************************************************************** */
34
+ #ifndef FSE_H
35
+ #define FSE_H
36
+
37
+ #if defined (__cplusplus)
38
+ extern "C" {
39
+ #endif
40
+
41
+
42
+ /*-*****************************************
43
+ * Dependencies
44
+ ******************************************/
45
+ #include <stddef.h> /* size_t, ptrdiff_t */
46
+
47
+
48
+ /*-****************************************
49
+ * FSE simple functions
50
+ ******************************************/
51
+ /*! FSE_compress() :
52
+ Compress content of buffer 'src', of size 'srcSize', into destination buffer 'dst'.
53
+ 'dst' buffer must be already allocated. Compression runs faster is dstCapacity >= FSE_compressBound(srcSize).
54
+ @return : size of compressed data (<= dstCapacity).
55
+ Special values : if return == 0, srcData is not compressible => Nothing is stored within dst !!!
56
+ if return == 1, srcData is a single byte symbol * srcSize times. Use RLE compression instead.
57
+ if FSE_isError(return), compression failed (more details using FSE_getErrorName())
58
+ */
59
+ size_t FSE_compress(void* dst, size_t dstCapacity,
60
+ const void* src, size_t srcSize);
61
+
62
+ /*! FSE_decompress():
63
+ Decompress FSE data from buffer 'cSrc', of size 'cSrcSize',
64
+ into already allocated destination buffer 'dst', of size 'dstCapacity'.
65
+ @return : size of regenerated data (<= maxDstSize),
66
+ or an error code, which can be tested using FSE_isError() .
67
+
68
+ ** Important ** : FSE_decompress() does not decompress non-compressible nor RLE data !!!
69
+ Why ? : making this distinction requires a header.
70
+ Header management is intentionally delegated to the user layer, which can better manage special cases.
71
+ */
72
+ size_t FSE_decompress(void* dst, size_t dstCapacity,
73
+ const void* cSrc, size_t cSrcSize);
74
+
75
+
76
+ /*-*****************************************
77
+ * Tool functions
78
+ ******************************************/
79
+ size_t FSE_compressBound(size_t size); /* maximum compressed size */
80
+
81
+ /* Error Management */
82
+ unsigned FSE_isError(size_t code); /* tells if a return value is an error code */
83
+ const char* FSE_getErrorName(size_t code); /* provides error code string (useful for debugging) */
84
+
85
+
86
+ /*-*****************************************
87
+ * FSE advanced functions
88
+ ******************************************/
89
+ /*! FSE_compress2() :
90
+ Same as FSE_compress(), but allows the selection of 'maxSymbolValue' and 'tableLog'
91
+ Both parameters can be defined as '0' to mean : use default value
92
+ @return : size of compressed data
93
+ Special values : if return == 0, srcData is not compressible => Nothing is stored within cSrc !!!
94
+ if return == 1, srcData is a single byte symbol * srcSize times. Use RLE compression.
95
+ if FSE_isError(return), it's an error code.
96
+ */
97
+ size_t FSE_compress2 (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog);
98
+
99
+
100
+ /*-*****************************************
101
+ * FSE detailed API
102
+ ******************************************/
103
+ /*!
104
+ FSE_compress() does the following:
105
+ 1. count symbol occurrence from source[] into table count[]
106
+ 2. normalize counters so that sum(count[]) == Power_of_2 (2^tableLog)
107
+ 3. save normalized counters to memory buffer using writeNCount()
108
+ 4. build encoding table 'CTable' from normalized counters
109
+ 5. encode the data stream using encoding table 'CTable'
110
+
111
+ FSE_decompress() does the following:
112
+ 1. read normalized counters with readNCount()
113
+ 2. build decoding table 'DTable' from normalized counters
114
+ 3. decode the data stream using decoding table 'DTable'
115
+
116
+ The following API allows targeting specific sub-functions for advanced tasks.
117
+ For example, it's possible to compress several blocks using the same 'CTable',
118
+ or to save and provide normalized distribution using external method.
119
+ */
120
+
121
+ /* *** COMPRESSION *** */
122
+
123
+ /*! FSE_count():
124
+ Provides the precise count of each byte within a table 'count'.
125
+ 'count' is a table of unsigned int, of minimum size (*maxSymbolValuePtr+1).
126
+ *maxSymbolValuePtr will be updated if detected smaller than initial value.
127
+ @return : the count of the most frequent symbol (which is not identified).
128
+ if return == srcSize, there is only one symbol.
129
+ Can also return an error code, which can be tested with FSE_isError(). */
130
+ size_t FSE_count(unsigned* count, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
131
+
132
+ /*! FSE_optimalTableLog():
133
+ dynamically downsize 'tableLog' when conditions are met.
134
+ It saves CPU time, by using smaller tables, while preserving or even improving compression ratio.
135
+ @return : recommended tableLog (necessarily <= 'maxTableLog') */
136
+ unsigned FSE_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue);
137
+
138
+ /*! FSE_normalizeCount():
139
+ normalize counts so that sum(count[]) == Power_of_2 (2^tableLog)
140
+ 'normalizedCounter' is a table of short, of minimum size (maxSymbolValue+1).
141
+ @return : tableLog,
142
+ or an errorCode, which can be tested using FSE_isError() */
143
+ size_t FSE_normalizeCount(short* normalizedCounter, unsigned tableLog, const unsigned* count, size_t srcSize, unsigned maxSymbolValue);
144
+
145
+ /*! FSE_NCountWriteBound():
146
+ Provides the maximum possible size of an FSE normalized table, given 'maxSymbolValue' and 'tableLog'.
147
+ Typically useful for allocation purpose. */
148
+ size_t FSE_NCountWriteBound(unsigned maxSymbolValue, unsigned tableLog);
149
+
150
+ /*! FSE_writeNCount():
151
+ Compactly save 'normalizedCounter' into 'buffer'.
152
+ @return : size of the compressed table,
153
+ or an errorCode, which can be tested using FSE_isError(). */
154
+ size_t FSE_writeNCount (void* buffer, size_t bufferSize, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog);
155
+
156
+
157
+ /*! Constructor and Destructor of FSE_CTable.
158
+ Note that FSE_CTable size depends on 'tableLog' and 'maxSymbolValue' */
159
+ typedef unsigned FSE_CTable; /* don't allocate that. It's only meant to be more restrictive than void* */
160
+ FSE_CTable* FSE_createCTable (unsigned tableLog, unsigned maxSymbolValue);
161
+ void FSE_freeCTable (FSE_CTable* ct);
162
+
163
+ /*! FSE_buildCTable():
164
+ Builds `ct`, which must be already allocated, using FSE_createCTable().
165
+ @return : 0, or an errorCode, which can be tested using FSE_isError() */
166
+ size_t FSE_buildCTable(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog);
167
+
168
+ /*! FSE_compress_usingCTable():
169
+ Compress `src` using `ct` into `dst` which must be already allocated.
170
+ @return : size of compressed data (<= `dstCapacity`),
171
+ or 0 if compressed data could not fit into `dst`,
172
+ or an errorCode, which can be tested using FSE_isError() */
173
+ size_t FSE_compress_usingCTable (void* dst, size_t dstCapacity, const void* src, size_t srcSize, const FSE_CTable* ct);
174
+
175
+ /*!
176
+ Tutorial :
177
+ ----------
178
+ The first step is to count all symbols. FSE_count() does this job very fast.
179
+ Result will be saved into 'count', a table of unsigned int, which must be already allocated, and have 'maxSymbolValuePtr[0]+1' cells.
180
+ 'src' is a table of bytes of size 'srcSize'. All values within 'src' MUST be <= maxSymbolValuePtr[0]
181
+ maxSymbolValuePtr[0] will be updated, with its real value (necessarily <= original value)
182
+ FSE_count() will return the number of occurrence of the most frequent symbol.
183
+ This can be used to know if there is a single symbol within 'src', and to quickly evaluate its compressibility.
184
+ If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError()).
185
+
186
+ The next step is to normalize the frequencies.
187
+ FSE_normalizeCount() will ensure that sum of frequencies is == 2 ^'tableLog'.
188
+ It also guarantees a minimum of 1 to any Symbol with frequency >= 1.
189
+ You can use 'tableLog'==0 to mean "use default tableLog value".
190
+ If you are unsure of which tableLog value to use, you can ask FSE_optimalTableLog(),
191
+ which will provide the optimal valid tableLog given sourceSize, maxSymbolValue, and a user-defined maximum (0 means "default").
192
+
193
+ The result of FSE_normalizeCount() will be saved into a table,
194
+ called 'normalizedCounter', which is a table of signed short.
195
+ 'normalizedCounter' must be already allocated, and have at least 'maxSymbolValue+1' cells.
196
+ The return value is tableLog if everything proceeded as expected.
197
+ It is 0 if there is a single symbol within distribution.
198
+ If there is an error (ex: invalid tableLog value), the function will return an ErrorCode (which can be tested using FSE_isError()).
199
+
200
+ 'normalizedCounter' can be saved in a compact manner to a memory area using FSE_writeNCount().
201
+ 'buffer' must be already allocated.
202
+ For guaranteed success, buffer size must be at least FSE_headerBound().
203
+ The result of the function is the number of bytes written into 'buffer'.
204
+ If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError(); ex : buffer size too small).
205
+
206
+ 'normalizedCounter' can then be used to create the compression table 'CTable'.
207
+ The space required by 'CTable' must be already allocated, using FSE_createCTable().
208
+ You can then use FSE_buildCTable() to fill 'CTable'.
209
+ If there is an error, both functions will return an ErrorCode (which can be tested using FSE_isError()).
210
+
211
+ 'CTable' can then be used to compress 'src', with FSE_compress_usingCTable().
212
+ Similar to FSE_count(), the convention is that 'src' is assumed to be a table of char of size 'srcSize'
213
+ The function returns the size of compressed data (without header), necessarily <= `dstCapacity`.
214
+ If it returns '0', compressed data could not fit into 'dst'.
215
+ If there is an error, the function will return an ErrorCode (which can be tested using FSE_isError()).
216
+ */
217
+
218
+
219
+ /* *** DECOMPRESSION *** */
220
+
221
+ /*! FSE_readNCount():
222
+ Read compactly saved 'normalizedCounter' from 'rBuffer'.
223
+ @return : size read from 'rBuffer',
224
+ or an errorCode, which can be tested using FSE_isError().
225
+ maxSymbolValuePtr[0] and tableLogPtr[0] will also be updated with their respective values */
226
+ size_t FSE_readNCount (short* normalizedCounter, unsigned* maxSymbolValuePtr, unsigned* tableLogPtr, const void* rBuffer, size_t rBuffSize);
227
+
228
+ /*! Constructor and Destructor of FSE_DTable.
229
+ Note that its size depends on 'tableLog' */
230
+ typedef unsigned FSE_DTable; /* don't allocate that. It's just a way to be more restrictive than void* */
231
+ FSE_DTable* FSE_createDTable(unsigned tableLog);
232
+ void FSE_freeDTable(FSE_DTable* dt);
233
+
234
+ /*! FSE_buildDTable():
235
+ Builds 'dt', which must be already allocated, using FSE_createDTable().
236
+ return : 0, or an errorCode, which can be tested using FSE_isError() */
237
+ size_t FSE_buildDTable (FSE_DTable* dt, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog);
238
+
239
+ /*! FSE_decompress_usingDTable():
240
+ Decompress compressed source `cSrc` of size `cSrcSize` using `dt`
241
+ into `dst` which must be already allocated.
242
+ @return : size of regenerated data (necessarily <= `dstCapacity`),
243
+ or an errorCode, which can be tested using FSE_isError() */
244
+ size_t FSE_decompress_usingDTable(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, const FSE_DTable* dt);
245
+
246
+ /*!
247
+ Tutorial :
248
+ ----------
249
+ (Note : these functions only decompress FSE-compressed blocks.
250
+ If block is uncompressed, use memcpy() instead
251
+ If block is a single repeated byte, use memset() instead )
252
+
253
+ The first step is to obtain the normalized frequencies of symbols.
254
+ This can be performed by FSE_readNCount() if it was saved using FSE_writeNCount().
255
+ 'normalizedCounter' must be already allocated, and have at least 'maxSymbolValuePtr[0]+1' cells of signed short.
256
+ In practice, that means it's necessary to know 'maxSymbolValue' beforehand,
257
+ or size the table to handle worst case situations (typically 256).
258
+ FSE_readNCount() will provide 'tableLog' and 'maxSymbolValue'.
259
+ The result of FSE_readNCount() is the number of bytes read from 'rBuffer'.
260
+ Note that 'rBufferSize' must be at least 4 bytes, even if useful information is less than that.
261
+ If there is an error, the function will return an error code, which can be tested using FSE_isError().
262
+
263
+ The next step is to build the decompression tables 'FSE_DTable' from 'normalizedCounter'.
264
+ This is performed by the function FSE_buildDTable().
265
+ The space required by 'FSE_DTable' must be already allocated using FSE_createDTable().
266
+ If there is an error, the function will return an error code, which can be tested using FSE_isError().
267
+
268
+ `FSE_DTable` can then be used to decompress `cSrc`, with FSE_decompress_usingDTable().
269
+ `cSrcSize` must be strictly correct, otherwise decompression will fail.
270
+ FSE_decompress_usingDTable() result will tell how many bytes were regenerated (<=`dstCapacity`).
271
+ If there is an error, the function will return an error code, which can be tested using FSE_isError(). (ex: dst buffer too small)
272
+ */
273
+
274
+
275
+ #ifdef FSE_STATIC_LINKING_ONLY
276
+
277
+ /* *** Dependency *** */
278
+ #include "bitstream.h"
279
+
280
+
281
+ /* *****************************************
282
+ * Static allocation
283
+ *******************************************/
284
+ /* FSE buffer bounds */
285
+ #define FSE_NCOUNTBOUND 512
286
+ #define FSE_BLOCKBOUND(size) (size + (size>>7))
287
+ #define FSE_COMPRESSBOUND(size) (FSE_NCOUNTBOUND + FSE_BLOCKBOUND(size)) /* Macro version, useful for static allocation */
288
+
289
+ /* It is possible to statically allocate FSE CTable/DTable as a table of FSE_CTable/FSE_DTable using below macros */
290
+ #define FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) (1 + (1<<(maxTableLog-1)) + ((maxSymbolValue+1)*2))
291
+ #define FSE_DTABLE_SIZE_U32(maxTableLog) (1 + (1<<maxTableLog))
292
+
293
+
294
+ /* *****************************************
295
+ * FSE advanced API
296
+ *******************************************/
297
+ /* FSE_count_wksp() :
298
+ * Same as FSE_count(), but using an externally provided scratch buffer.
299
+ * `workSpace` size must be table of >= `1024` unsigned
300
+ */
301
+ size_t FSE_count_wksp(unsigned* count, unsigned* maxSymbolValuePtr,
302
+ const void* source, size_t sourceSize, unsigned* workSpace);
303
+
304
+ /** FSE_countFast() :
305
+ * same as FSE_count(), but blindly trusts that all byte values within src are <= *maxSymbolValuePtr
306
+ */
307
+ size_t FSE_countFast(unsigned* count, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
308
+
309
+ /* FSE_countFast_wksp() :
310
+ * Same as FSE_countFast(), but using an externally provided scratch buffer.
311
+ * `workSpace` must be a table of minimum `1024` unsigned
312
+ */
313
+ size_t FSE_countFast_wksp(unsigned* count, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize, unsigned* workSpace);
314
+
315
+ /*! FSE_count_simple
316
+ * Same as FSE_countFast(), but does not use any additional memory (not even on stack).
317
+ * This function is unsafe, and will segfault if any value within `src` is `> *maxSymbolValuePtr` (presuming it's also the size of `count`).
318
+ */
319
+ size_t FSE_count_simple(unsigned* count, unsigned* maxSymbolValuePtr, const void* src, size_t srcSize);
320
+
321
+
322
+
323
+ unsigned FSE_optimalTableLog_internal(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue, unsigned minus);
324
+ /**< same as FSE_optimalTableLog(), which used `minus==2` */
325
+
326
+ /* FSE_compress_wksp() :
327
+ * Same as FSE_compress2(), but using an externally allocated scratch buffer (`workSpace`).
328
+ * FSE_WKSP_SIZE_U32() provides the minimum size required for `workSpace` as a table of FSE_CTable.
329
+ */
330
+ #define FSE_WKSP_SIZE_U32(maxTableLog, maxSymbolValue) ( FSE_CTABLE_SIZE_U32(maxTableLog, maxSymbolValue) + (1<<((maxTableLog>2)?(maxTableLog-2):0)) )
331
+ size_t FSE_compress_wksp (void* dst, size_t dstSize, const void* src, size_t srcSize, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
332
+
333
+ size_t FSE_buildCTable_raw (FSE_CTable* ct, unsigned nbBits);
334
+ /**< build a fake FSE_CTable, designed for a flat distribution, where each symbol uses nbBits */
335
+
336
+ size_t FSE_buildCTable_rle (FSE_CTable* ct, unsigned char symbolValue);
337
+ /**< build a fake FSE_CTable, designed to compress always the same symbolValue */
338
+
339
+ /* FSE_buildCTable_wksp() :
340
+ * Same as FSE_buildCTable(), but using an externally allocated scratch buffer (`workSpace`).
341
+ * `wkspSize` must be >= `(1<<tableLog)`.
342
+ */
343
+ size_t FSE_buildCTable_wksp(FSE_CTable* ct, const short* normalizedCounter, unsigned maxSymbolValue, unsigned tableLog, void* workSpace, size_t wkspSize);
344
+
345
+ size_t FSE_buildDTable_raw (FSE_DTable* dt, unsigned nbBits);
346
+ /**< build a fake FSE_DTable, designed to read a flat distribution where each symbol uses nbBits */
347
+
348
+ size_t FSE_buildDTable_rle (FSE_DTable* dt, unsigned char symbolValue);
349
+ /**< build a fake FSE_DTable, designed to always generate the same symbolValue */
350
+
351
+ size_t FSE_decompress_wksp(void* dst, size_t dstCapacity, const void* cSrc, size_t cSrcSize, FSE_DTable* workSpace, unsigned maxLog);
352
+ /**< same as FSE_decompress(), using an externally allocated `workSpace` produced with `FSE_DTABLE_SIZE_U32(maxLog)` */
353
+
354
+
355
+ /* *****************************************
356
+ * FSE symbol compression API
357
+ *******************************************/
358
+ /*!
359
+ This API consists of small unitary functions, which highly benefit from being inlined.
360
+ Hence their body are included in next section.
361
+ */
362
+ typedef struct {
363
+ ptrdiff_t value;
364
+ const void* stateTable;
365
+ const void* symbolTT;
366
+ unsigned stateLog;
367
+ } FSE_CState_t;
368
+
369
+ static void FSE_initCState(FSE_CState_t* CStatePtr, const FSE_CTable* ct);
370
+
371
+ static void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* CStatePtr, unsigned symbol);
372
+
373
+ static void FSE_flushCState(BIT_CStream_t* bitC, const FSE_CState_t* CStatePtr);
374
+
375
+ /**<
376
+ These functions are inner components of FSE_compress_usingCTable().
377
+ They allow the creation of custom streams, mixing multiple tables and bit sources.
378
+
379
+ A key property to keep in mind is that encoding and decoding are done **in reverse direction**.
380
+ So the first symbol you will encode is the last you will decode, like a LIFO stack.
381
+
382
+ You will need a few variables to track your CStream. They are :
383
+
384
+ FSE_CTable ct; // Provided by FSE_buildCTable()
385
+ BIT_CStream_t bitStream; // bitStream tracking structure
386
+ FSE_CState_t state; // State tracking structure (can have several)
387
+
388
+
389
+ The first thing to do is to init bitStream and state.
390
+ size_t errorCode = BIT_initCStream(&bitStream, dstBuffer, maxDstSize);
391
+ FSE_initCState(&state, ct);
392
+
393
+ Note that BIT_initCStream() can produce an error code, so its result should be tested, using FSE_isError();
394
+ You can then encode your input data, byte after byte.
395
+ FSE_encodeSymbol() outputs a maximum of 'tableLog' bits at a time.
396
+ Remember decoding will be done in reverse direction.
397
+ FSE_encodeByte(&bitStream, &state, symbol);
398
+
399
+ At any time, you can also add any bit sequence.
400
+ Note : maximum allowed nbBits is 25, for compatibility with 32-bits decoders
401
+ BIT_addBits(&bitStream, bitField, nbBits);
402
+
403
+ The above methods don't commit data to memory, they just store it into local register, for speed.
404
+ Local register size is 64-bits on 64-bits systems, 32-bits on 32-bits systems (size_t).
405
+ Writing data to memory is a manual operation, performed by the flushBits function.
406
+ BIT_flushBits(&bitStream);
407
+
408
+ Your last FSE encoding operation shall be to flush your last state value(s).
409
+ FSE_flushState(&bitStream, &state);
410
+
411
+ Finally, you must close the bitStream.
412
+ The function returns the size of CStream in bytes.
413
+ If data couldn't fit into dstBuffer, it will return a 0 ( == not compressible)
414
+ If there is an error, it returns an errorCode (which can be tested using FSE_isError()).
415
+ size_t size = BIT_closeCStream(&bitStream);
416
+ */
417
+
418
+
419
+ /* *****************************************
420
+ * FSE symbol decompression API
421
+ *******************************************/
422
+ typedef struct {
423
+ size_t state;
424
+ const void* table; /* precise table may vary, depending on U16 */
425
+ } FSE_DState_t;
426
+
427
+
428
+ static void FSE_initDState(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD, const FSE_DTable* dt);
429
+
430
+ static unsigned char FSE_decodeSymbol(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD);
431
+
432
+ static unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr);
433
+
434
+ /**<
435
+ Let's now decompose FSE_decompress_usingDTable() into its unitary components.
436
+ You will decode FSE-encoded symbols from the bitStream,
437
+ and also any other bitFields you put in, **in reverse order**.
438
+
439
+ You will need a few variables to track your bitStream. They are :
440
+
441
+ BIT_DStream_t DStream; // Stream context
442
+ FSE_DState_t DState; // State context. Multiple ones are possible
443
+ FSE_DTable* DTablePtr; // Decoding table, provided by FSE_buildDTable()
444
+
445
+ The first thing to do is to init the bitStream.
446
+ errorCode = BIT_initDStream(&DStream, srcBuffer, srcSize);
447
+
448
+ You should then retrieve your initial state(s)
449
+ (in reverse flushing order if you have several ones) :
450
+ errorCode = FSE_initDState(&DState, &DStream, DTablePtr);
451
+
452
+ You can then decode your data, symbol after symbol.
453
+ For information the maximum number of bits read by FSE_decodeSymbol() is 'tableLog'.
454
+ Keep in mind that symbols are decoded in reverse order, like a LIFO stack (last in, first out).
455
+ unsigned char symbol = FSE_decodeSymbol(&DState, &DStream);
456
+
457
+ You can retrieve any bitfield you eventually stored into the bitStream (in reverse order)
458
+ Note : maximum allowed nbBits is 25, for 32-bits compatibility
459
+ size_t bitField = BIT_readBits(&DStream, nbBits);
460
+
461
+ All above operations only read from local register (which size depends on size_t).
462
+ Refueling the register from memory is manually performed by the reload method.
463
+ endSignal = FSE_reloadDStream(&DStream);
464
+
465
+ BIT_reloadDStream() result tells if there is still some more data to read from DStream.
466
+ BIT_DStream_unfinished : there is still some data left into the DStream.
467
+ BIT_DStream_endOfBuffer : Dstream reached end of buffer. Its container may no longer be completely filled.
468
+ BIT_DStream_completed : Dstream reached its exact end, corresponding in general to decompression completed.
469
+ BIT_DStream_tooFar : Dstream went too far. Decompression result is corrupted.
470
+
471
+ When reaching end of buffer (BIT_DStream_endOfBuffer), progress slowly, notably if you decode multiple symbols per loop,
472
+ to properly detect the exact end of stream.
473
+ After each decoded symbol, check if DStream is fully consumed using this simple test :
474
+ BIT_reloadDStream(&DStream) >= BIT_DStream_completed
475
+
476
+ When it's done, verify decompression is fully completed, by checking both DStream and the relevant states.
477
+ Checking if DStream has reached its end is performed by :
478
+ BIT_endOfDStream(&DStream);
479
+ Check also the states. There might be some symbols left there, if some high probability ones (>50%) are possible.
480
+ FSE_endOfDState(&DState);
481
+ */
482
+
483
+
484
+ /* *****************************************
485
+ * FSE unsafe API
486
+ *******************************************/
487
+ static unsigned char FSE_decodeSymbolFast(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD);
488
+ /* faster, but works only if nbBits is always >= 1 (otherwise, result will be corrupted) */
489
+
490
+
491
+ /* *****************************************
492
+ * Implementation of inlined functions
493
+ *******************************************/
494
+ typedef struct {
495
+ int deltaFindState;
496
+ U32 deltaNbBits;
497
+ } FSE_symbolCompressionTransform; /* total 8 bytes */
498
+
499
+ MEM_STATIC void FSE_initCState(FSE_CState_t* statePtr, const FSE_CTable* ct)
500
+ {
501
+ const void* ptr = ct;
502
+ const U16* u16ptr = (const U16*) ptr;
503
+ const U32 tableLog = MEM_read16(ptr);
504
+ statePtr->value = (ptrdiff_t)1<<tableLog;
505
+ statePtr->stateTable = u16ptr+2;
506
+ statePtr->symbolTT = ((const U32*)ct + 1 + (tableLog ? (1<<(tableLog-1)) : 1));
507
+ statePtr->stateLog = tableLog;
508
+ }
509
+
510
+
511
+ /*! FSE_initCState2() :
512
+ * Same as FSE_initCState(), but the first symbol to include (which will be the last to be read)
513
+ * uses the smallest state value possible, saving the cost of this symbol */
514
+ MEM_STATIC void FSE_initCState2(FSE_CState_t* statePtr, const FSE_CTable* ct, U32 symbol)
515
+ {
516
+ FSE_initCState(statePtr, ct);
517
+ { const FSE_symbolCompressionTransform symbolTT = ((const FSE_symbolCompressionTransform*)(statePtr->symbolTT))[symbol];
518
+ const U16* stateTable = (const U16*)(statePtr->stateTable);
519
+ U32 nbBitsOut = (U32)((symbolTT.deltaNbBits + (1<<15)) >> 16);
520
+ statePtr->value = (nbBitsOut << 16) - symbolTT.deltaNbBits;
521
+ statePtr->value = stateTable[(statePtr->value >> nbBitsOut) + symbolTT.deltaFindState];
522
+ }
523
+ }
524
+
525
+ MEM_STATIC void FSE_encodeSymbol(BIT_CStream_t* bitC, FSE_CState_t* statePtr, U32 symbol)
526
+ {
527
+ const FSE_symbolCompressionTransform symbolTT = ((const FSE_symbolCompressionTransform*)(statePtr->symbolTT))[symbol];
528
+ const U16* const stateTable = (const U16*)(statePtr->stateTable);
529
+ U32 nbBitsOut = (U32)((statePtr->value + symbolTT.deltaNbBits) >> 16);
530
+ BIT_addBits(bitC, statePtr->value, nbBitsOut);
531
+ statePtr->value = stateTable[ (statePtr->value >> nbBitsOut) + symbolTT.deltaFindState];
532
+ }
533
+
534
+ MEM_STATIC void FSE_flushCState(BIT_CStream_t* bitC, const FSE_CState_t* statePtr)
535
+ {
536
+ BIT_addBits(bitC, statePtr->value, statePtr->stateLog);
537
+ BIT_flushBits(bitC);
538
+ }
539
+
540
+
541
+ /* ====== Decompression ====== */
542
+
543
+ typedef struct {
544
+ U16 tableLog;
545
+ U16 fastMode;
546
+ } FSE_DTableHeader; /* sizeof U32 */
547
+
548
+ typedef struct
549
+ {
550
+ unsigned short newState;
551
+ unsigned char symbol;
552
+ unsigned char nbBits;
553
+ } FSE_decode_t; /* size == U32 */
554
+
555
+ MEM_STATIC void FSE_initDState(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD, const FSE_DTable* dt)
556
+ {
557
+ const void* ptr = dt;
558
+ const FSE_DTableHeader* const DTableH = (const FSE_DTableHeader*)ptr;
559
+ DStatePtr->state = BIT_readBits(bitD, DTableH->tableLog);
560
+ BIT_reloadDStream(bitD);
561
+ DStatePtr->table = dt + 1;
562
+ }
563
+
564
+ MEM_STATIC BYTE FSE_peekSymbol(const FSE_DState_t* DStatePtr)
565
+ {
566
+ FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state];
567
+ return DInfo.symbol;
568
+ }
569
+
570
+ MEM_STATIC void FSE_updateState(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD)
571
+ {
572
+ FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state];
573
+ U32 const nbBits = DInfo.nbBits;
574
+ size_t const lowBits = BIT_readBits(bitD, nbBits);
575
+ DStatePtr->state = DInfo.newState + lowBits;
576
+ }
577
+
578
+ MEM_STATIC BYTE FSE_decodeSymbol(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD)
579
+ {
580
+ FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state];
581
+ U32 const nbBits = DInfo.nbBits;
582
+ BYTE const symbol = DInfo.symbol;
583
+ size_t const lowBits = BIT_readBits(bitD, nbBits);
584
+
585
+ DStatePtr->state = DInfo.newState + lowBits;
586
+ return symbol;
587
+ }
588
+
589
+ /*! FSE_decodeSymbolFast() :
590
+ unsafe, only works if no symbol has a probability > 50% */
591
+ MEM_STATIC BYTE FSE_decodeSymbolFast(FSE_DState_t* DStatePtr, BIT_DStream_t* bitD)
592
+ {
593
+ FSE_decode_t const DInfo = ((const FSE_decode_t*)(DStatePtr->table))[DStatePtr->state];
594
+ U32 const nbBits = DInfo.nbBits;
595
+ BYTE const symbol = DInfo.symbol;
596
+ size_t const lowBits = BIT_readBitsFast(bitD, nbBits);
597
+
598
+ DStatePtr->state = DInfo.newState + lowBits;
599
+ return symbol;
600
+ }
601
+
602
+ MEM_STATIC unsigned FSE_endOfDState(const FSE_DState_t* DStatePtr)
603
+ {
604
+ return DStatePtr->state == 0;
605
+ }
606
+
607
+
608
+
609
+ #ifndef FSE_COMMONDEFS_ONLY
610
+
611
+ /* **************************************************************
612
+ * Tuning parameters
613
+ ****************************************************************/
614
+ /*!MEMORY_USAGE :
615
+ * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.)
616
+ * Increasing memory usage improves compression ratio
617
+ * Reduced memory usage can improve speed, due to cache effect
618
+ * Recommended max value is 14, for 16KB, which nicely fits into Intel x86 L1 cache */
619
+ #ifndef FSE_MAX_MEMORY_USAGE
620
+ # define FSE_MAX_MEMORY_USAGE 14
621
+ #endif
622
+ #ifndef FSE_DEFAULT_MEMORY_USAGE
623
+ # define FSE_DEFAULT_MEMORY_USAGE 13
624
+ #endif
625
+
626
+ /*!FSE_MAX_SYMBOL_VALUE :
627
+ * Maximum symbol value authorized.
628
+ * Required for proper stack allocation */
629
+ #ifndef FSE_MAX_SYMBOL_VALUE
630
+ # define FSE_MAX_SYMBOL_VALUE 255
631
+ #endif
632
+
633
+ /* **************************************************************
634
+ * template functions type & suffix
635
+ ****************************************************************/
636
+ #define FSE_FUNCTION_TYPE BYTE
637
+ #define FSE_FUNCTION_EXTENSION
638
+ #define FSE_DECODE_TYPE FSE_decode_t
639
+
640
+
641
+ #endif /* !FSE_COMMONDEFS_ONLY */
642
+
643
+
644
+ /* ***************************************************************
645
+ * Constants
646
+ *****************************************************************/
647
+ #define FSE_MAX_TABLELOG (FSE_MAX_MEMORY_USAGE-2)
648
+ #define FSE_MAX_TABLESIZE (1U<<FSE_MAX_TABLELOG)
649
+ #define FSE_MAXTABLESIZE_MASK (FSE_MAX_TABLESIZE-1)
650
+ #define FSE_DEFAULT_TABLELOG (FSE_DEFAULT_MEMORY_USAGE-2)
651
+ #define FSE_MIN_TABLELOG 5
652
+
653
+ #define FSE_TABLELOG_ABSOLUTE_MAX 15
654
+ #if FSE_MAX_TABLELOG > FSE_TABLELOG_ABSOLUTE_MAX
655
+ # error "FSE_MAX_TABLELOG > FSE_TABLELOG_ABSOLUTE_MAX is not supported"
656
+ #endif
657
+
658
+ #define FSE_TABLESTEP(tableSize) ((tableSize>>1) + (tableSize>>3) + 3)
659
+
660
+
661
+ #endif /* FSE_STATIC_LINKING_ONLY */
662
+
663
+
664
+ #if defined (__cplusplus)
665
+ }
666
+ #endif
667
+
668
+ #endif /* FSE_H */