zstd-ruby 0.1.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +19 -0
- data/.gitmodules +3 -0
- data/.rspec +2 -0
- data/.travis.yml +11 -0
- data/CODE_OF_CONDUCT.md +74 -0
- data/Gemfile +4 -0
- data/LICENSE.txt +29 -0
- data/README.md +63 -0
- data/Rakefile +22 -0
- data/bin/console +14 -0
- data/bin/setup +8 -0
- data/ext/zstdruby/extconf.rb +23 -0
- data/ext/zstdruby/libzstd/.gitignore +2 -0
- data/ext/zstdruby/libzstd/Makefile +133 -0
- data/ext/zstdruby/libzstd/README.md +77 -0
- data/ext/zstdruby/libzstd/common/bitstream.h +414 -0
- data/ext/zstdruby/libzstd/common/entropy_common.c +227 -0
- data/ext/zstdruby/libzstd/common/error_private.c +43 -0
- data/ext/zstdruby/libzstd/common/error_private.h +76 -0
- data/ext/zstdruby/libzstd/common/fse.h +668 -0
- data/ext/zstdruby/libzstd/common/fse_decompress.c +329 -0
- data/ext/zstdruby/libzstd/common/huf.h +238 -0
- data/ext/zstdruby/libzstd/common/mem.h +372 -0
- data/ext/zstdruby/libzstd/common/xxhash.c +867 -0
- data/ext/zstdruby/libzstd/common/xxhash.h +309 -0
- data/ext/zstdruby/libzstd/common/zstd_common.c +77 -0
- data/ext/zstdruby/libzstd/common/zstd_errors.h +60 -0
- data/ext/zstdruby/libzstd/common/zstd_internal.h +270 -0
- data/ext/zstdruby/libzstd/compress/fse_compress.c +850 -0
- data/ext/zstdruby/libzstd/compress/huf_compress.c +609 -0
- data/ext/zstdruby/libzstd/compress/zstd_compress.c +3291 -0
- data/ext/zstdruby/libzstd/compress/zstd_opt.h +919 -0
- data/ext/zstdruby/libzstd/decompress/huf_decompress.c +885 -0
- data/ext/zstdruby/libzstd/decompress/zstd_decompress.c +2154 -0
- data/ext/zstdruby/libzstd/deprecated/zbuff.h +210 -0
- data/ext/zstdruby/libzstd/deprecated/zbuff_compress.c +145 -0
- data/ext/zstdruby/libzstd/deprecated/zbuff_decompress.c +74 -0
- data/ext/zstdruby/libzstd/dictBuilder/divsufsort.c +1913 -0
- data/ext/zstdruby/libzstd/dictBuilder/divsufsort.h +67 -0
- data/ext/zstdruby/libzstd/dictBuilder/zdict.c +1012 -0
- data/ext/zstdruby/libzstd/dictBuilder/zdict.h +111 -0
- data/ext/zstdruby/libzstd/dll/example/Makefile +47 -0
- data/ext/zstdruby/libzstd/dll/example/README.md +69 -0
- data/ext/zstdruby/libzstd/dll/example/build_package.bat +17 -0
- data/ext/zstdruby/libzstd/dll/example/fullbench-dll.sln +25 -0
- data/ext/zstdruby/libzstd/dll/example/fullbench-dll.vcxproj +179 -0
- data/ext/zstdruby/libzstd/dll/libzstd.def +86 -0
- data/ext/zstdruby/libzstd/legacy/zstd_legacy.h +259 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v01.c +2095 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v01.h +80 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v02.c +3518 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v02.h +79 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v03.c +3159 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v03.h +79 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v04.c +3795 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v04.h +128 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v05.c +4056 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v05.h +149 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v06.c +4167 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v06.h +159 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v07.c +4540 -0
- data/ext/zstdruby/libzstd/legacy/zstd_v07.h +173 -0
- data/ext/zstdruby/libzstd/libzstd.pc.in +14 -0
- data/ext/zstdruby/libzstd/zstd.h +673 -0
- data/ext/zstdruby/zstdruby.c +117 -0
- data/ext/zstdruby/zstdruby.h +6 -0
- data/lib/zstd-ruby.rb +6 -0
- data/lib/zstd-ruby/version.rb +3 -0
- data/zstd-ruby.gemspec +37 -0
- metadata +170 -0
@@ -0,0 +1,609 @@
|
|
1
|
+
/* ******************************************************************
|
2
|
+
Huffman encoder, part of New Generation Entropy library
|
3
|
+
Copyright (C) 2013-2016, Yann Collet.
|
4
|
+
|
5
|
+
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
|
6
|
+
|
7
|
+
Redistribution and use in source and binary forms, with or without
|
8
|
+
modification, are permitted provided that the following conditions are
|
9
|
+
met:
|
10
|
+
|
11
|
+
* Redistributions of source code must retain the above copyright
|
12
|
+
notice, this list of conditions and the following disclaimer.
|
13
|
+
* Redistributions in binary form must reproduce the above
|
14
|
+
copyright notice, this list of conditions and the following disclaimer
|
15
|
+
in the documentation and/or other materials provided with the
|
16
|
+
distribution.
|
17
|
+
|
18
|
+
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
19
|
+
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
20
|
+
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
21
|
+
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
22
|
+
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
23
|
+
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
24
|
+
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
25
|
+
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
26
|
+
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
27
|
+
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
28
|
+
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
29
|
+
|
30
|
+
You can contact the author at :
|
31
|
+
- FSE+HUF source repository : https://github.com/Cyan4973/FiniteStateEntropy
|
32
|
+
- Public forum : https://groups.google.com/forum/#!forum/lz4c
|
33
|
+
****************************************************************** */
|
34
|
+
|
35
|
+
/* **************************************************************
|
36
|
+
* Compiler specifics
|
37
|
+
****************************************************************/
|
38
|
+
#ifdef _MSC_VER /* Visual Studio */
|
39
|
+
# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */
|
40
|
+
#endif
|
41
|
+
|
42
|
+
|
43
|
+
/* **************************************************************
|
44
|
+
* Includes
|
45
|
+
****************************************************************/
|
46
|
+
#include <string.h> /* memcpy, memset */
|
47
|
+
#include <stdio.h> /* printf (debug) */
|
48
|
+
#include "bitstream.h"
|
49
|
+
#define FSE_STATIC_LINKING_ONLY /* FSE_optimalTableLog_internal */
|
50
|
+
#include "fse.h" /* header compression */
|
51
|
+
#define HUF_STATIC_LINKING_ONLY
|
52
|
+
#include "huf.h"
|
53
|
+
|
54
|
+
|
55
|
+
/* **************************************************************
|
56
|
+
* Error Management
|
57
|
+
****************************************************************/
|
58
|
+
#define HUF_STATIC_ASSERT(c) { enum { HUF_static_assert = 1/(int)(!!(c)) }; } /* use only *after* variable declarations */
|
59
|
+
#define CHECK_V_F(e, f) size_t const e = f; if (ERR_isError(e)) return f
|
60
|
+
#define CHECK_F(f) { CHECK_V_F(_var_err__, f); }
|
61
|
+
|
62
|
+
|
63
|
+
/* **************************************************************
|
64
|
+
* Utils
|
65
|
+
****************************************************************/
|
66
|
+
unsigned HUF_optimalTableLog(unsigned maxTableLog, size_t srcSize, unsigned maxSymbolValue)
|
67
|
+
{
|
68
|
+
return FSE_optimalTableLog_internal(maxTableLog, srcSize, maxSymbolValue, 1);
|
69
|
+
}
|
70
|
+
|
71
|
+
|
72
|
+
/* *******************************************************
|
73
|
+
* HUF : Huffman block compression
|
74
|
+
*********************************************************/
|
75
|
+
/* HUF_compressWeights() :
|
76
|
+
* Same as FSE_compress(), but dedicated to huff0's weights compression.
|
77
|
+
* The use case needs much less stack memory.
|
78
|
+
* Note : all elements within weightTable are supposed to be <= HUF_TABLELOG_MAX.
|
79
|
+
*/
|
80
|
+
#define MAX_FSE_TABLELOG_FOR_HUFF_HEADER 6
|
81
|
+
size_t HUF_compressWeights (void* dst, size_t dstSize, const void* weightTable, size_t wtSize)
|
82
|
+
{
|
83
|
+
BYTE* const ostart = (BYTE*) dst;
|
84
|
+
BYTE* op = ostart;
|
85
|
+
BYTE* const oend = ostart + dstSize;
|
86
|
+
|
87
|
+
U32 maxSymbolValue = HUF_TABLELOG_MAX;
|
88
|
+
U32 tableLog = MAX_FSE_TABLELOG_FOR_HUFF_HEADER;
|
89
|
+
|
90
|
+
FSE_CTable CTable[FSE_CTABLE_SIZE_U32(MAX_FSE_TABLELOG_FOR_HUFF_HEADER, HUF_TABLELOG_MAX)];
|
91
|
+
BYTE scratchBuffer[1<<MAX_FSE_TABLELOG_FOR_HUFF_HEADER];
|
92
|
+
|
93
|
+
U32 count[HUF_TABLELOG_MAX+1];
|
94
|
+
S16 norm[HUF_TABLELOG_MAX+1];
|
95
|
+
|
96
|
+
/* init conditions */
|
97
|
+
if (wtSize <= 1) return 0; /* Not compressible */
|
98
|
+
|
99
|
+
/* Scan input and build symbol stats */
|
100
|
+
{ CHECK_V_F(maxCount, FSE_count_simple(count, &maxSymbolValue, weightTable, wtSize) );
|
101
|
+
if (maxCount == wtSize) return 1; /* only a single symbol in src : rle */
|
102
|
+
if (maxCount == 1) return 0; /* each symbol present maximum once => not compressible */
|
103
|
+
}
|
104
|
+
|
105
|
+
tableLog = FSE_optimalTableLog(tableLog, wtSize, maxSymbolValue);
|
106
|
+
CHECK_F( FSE_normalizeCount(norm, tableLog, count, wtSize, maxSymbolValue) );
|
107
|
+
|
108
|
+
/* Write table description header */
|
109
|
+
{ CHECK_V_F(hSize, FSE_writeNCount(op, oend-op, norm, maxSymbolValue, tableLog) );
|
110
|
+
op += hSize;
|
111
|
+
}
|
112
|
+
|
113
|
+
/* Compress */
|
114
|
+
CHECK_F( FSE_buildCTable_wksp(CTable, norm, maxSymbolValue, tableLog, scratchBuffer, sizeof(scratchBuffer)) );
|
115
|
+
{ CHECK_V_F(cSize, FSE_compress_usingCTable(op, oend - op, weightTable, wtSize, CTable) );
|
116
|
+
if (cSize == 0) return 0; /* not enough space for compressed data */
|
117
|
+
op += cSize;
|
118
|
+
}
|
119
|
+
|
120
|
+
return op-ostart;
|
121
|
+
}
|
122
|
+
|
123
|
+
|
124
|
+
struct HUF_CElt_s {
|
125
|
+
U16 val;
|
126
|
+
BYTE nbBits;
|
127
|
+
}; /* typedef'd to HUF_CElt within "huf.h" */
|
128
|
+
|
129
|
+
/*! HUF_writeCTable() :
|
130
|
+
`CTable` : huffman tree to save, using huf representation.
|
131
|
+
@return : size of saved CTable */
|
132
|
+
size_t HUF_writeCTable (void* dst, size_t maxDstSize,
|
133
|
+
const HUF_CElt* CTable, U32 maxSymbolValue, U32 huffLog)
|
134
|
+
{
|
135
|
+
BYTE bitsToWeight[HUF_TABLELOG_MAX + 1]; /* precomputed conversion table */
|
136
|
+
BYTE huffWeight[HUF_SYMBOLVALUE_MAX];
|
137
|
+
BYTE* op = (BYTE*)dst;
|
138
|
+
U32 n;
|
139
|
+
|
140
|
+
/* check conditions */
|
141
|
+
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(maxSymbolValue_tooLarge);
|
142
|
+
|
143
|
+
/* convert to weight */
|
144
|
+
bitsToWeight[0] = 0;
|
145
|
+
for (n=1; n<huffLog+1; n++)
|
146
|
+
bitsToWeight[n] = (BYTE)(huffLog + 1 - n);
|
147
|
+
for (n=0; n<maxSymbolValue; n++)
|
148
|
+
huffWeight[n] = bitsToWeight[CTable[n].nbBits];
|
149
|
+
|
150
|
+
/* attempt weights compression by FSE */
|
151
|
+
{ CHECK_V_F(hSize, HUF_compressWeights(op+1, maxDstSize-1, huffWeight, maxSymbolValue) );
|
152
|
+
if ((hSize>1) & (hSize < maxSymbolValue/2)) { /* FSE compressed */
|
153
|
+
op[0] = (BYTE)hSize;
|
154
|
+
return hSize+1;
|
155
|
+
} }
|
156
|
+
|
157
|
+
/* write raw values as 4-bits (max : 15) */
|
158
|
+
if (maxSymbolValue > (256-128)) return ERROR(GENERIC); /* should not happen : likely means source cannot be compressed */
|
159
|
+
if (((maxSymbolValue+1)/2) + 1 > maxDstSize) return ERROR(dstSize_tooSmall); /* not enough space within dst buffer */
|
160
|
+
op[0] = (BYTE)(128 /*special case*/ + (maxSymbolValue-1));
|
161
|
+
huffWeight[maxSymbolValue] = 0; /* to be sure it doesn't cause msan issue in final combination */
|
162
|
+
for (n=0; n<maxSymbolValue; n+=2)
|
163
|
+
op[(n/2)+1] = (BYTE)((huffWeight[n] << 4) + huffWeight[n+1]);
|
164
|
+
return ((maxSymbolValue+1)/2) + 1;
|
165
|
+
}
|
166
|
+
|
167
|
+
|
168
|
+
size_t HUF_readCTable (HUF_CElt* CTable, U32 maxSymbolValue, const void* src, size_t srcSize)
|
169
|
+
{
|
170
|
+
BYTE huffWeight[HUF_SYMBOLVALUE_MAX + 1]; /* init not required, even though some static analyzer may complain */
|
171
|
+
U32 rankVal[HUF_TABLELOG_ABSOLUTEMAX + 1]; /* large enough for values from 0 to 16 */
|
172
|
+
U32 tableLog = 0;
|
173
|
+
U32 nbSymbols = 0;
|
174
|
+
|
175
|
+
/* get symbol weights */
|
176
|
+
CHECK_V_F(readSize, HUF_readStats(huffWeight, HUF_SYMBOLVALUE_MAX+1, rankVal, &nbSymbols, &tableLog, src, srcSize));
|
177
|
+
|
178
|
+
/* check result */
|
179
|
+
if (tableLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
|
180
|
+
if (nbSymbols > maxSymbolValue+1) return ERROR(maxSymbolValue_tooSmall);
|
181
|
+
|
182
|
+
/* Prepare base value per rank */
|
183
|
+
{ U32 n, nextRankStart = 0;
|
184
|
+
for (n=1; n<=tableLog; n++) {
|
185
|
+
U32 current = nextRankStart;
|
186
|
+
nextRankStart += (rankVal[n] << (n-1));
|
187
|
+
rankVal[n] = current;
|
188
|
+
} }
|
189
|
+
|
190
|
+
/* fill nbBits */
|
191
|
+
{ U32 n; for (n=0; n<nbSymbols; n++) {
|
192
|
+
const U32 w = huffWeight[n];
|
193
|
+
CTable[n].nbBits = (BYTE)(tableLog + 1 - w);
|
194
|
+
} }
|
195
|
+
|
196
|
+
/* fill val */
|
197
|
+
{ U16 nbPerRank[HUF_TABLELOG_MAX+2] = {0}; /* support w=0=>n=tableLog+1 */
|
198
|
+
U16 valPerRank[HUF_TABLELOG_MAX+2] = {0};
|
199
|
+
{ U32 n; for (n=0; n<nbSymbols; n++) nbPerRank[CTable[n].nbBits]++; }
|
200
|
+
/* determine stating value per rank */
|
201
|
+
valPerRank[tableLog+1] = 0; /* for w==0 */
|
202
|
+
{ U16 min = 0;
|
203
|
+
U32 n; for (n=tableLog; n>0; n--) { /* start at n=tablelog <-> w=1 */
|
204
|
+
valPerRank[n] = min; /* get starting value within each rank */
|
205
|
+
min += nbPerRank[n];
|
206
|
+
min >>= 1;
|
207
|
+
} }
|
208
|
+
/* assign value within rank, symbol order */
|
209
|
+
{ U32 n; for (n=0; n<=maxSymbolValue; n++) CTable[n].val = valPerRank[CTable[n].nbBits]++; }
|
210
|
+
}
|
211
|
+
|
212
|
+
return readSize;
|
213
|
+
}
|
214
|
+
|
215
|
+
|
216
|
+
typedef struct nodeElt_s {
|
217
|
+
U32 count;
|
218
|
+
U16 parent;
|
219
|
+
BYTE byte;
|
220
|
+
BYTE nbBits;
|
221
|
+
} nodeElt;
|
222
|
+
|
223
|
+
static U32 HUF_setMaxHeight(nodeElt* huffNode, U32 lastNonNull, U32 maxNbBits)
|
224
|
+
{
|
225
|
+
const U32 largestBits = huffNode[lastNonNull].nbBits;
|
226
|
+
if (largestBits <= maxNbBits) return largestBits; /* early exit : no elt > maxNbBits */
|
227
|
+
|
228
|
+
/* there are several too large elements (at least >= 2) */
|
229
|
+
{ int totalCost = 0;
|
230
|
+
const U32 baseCost = 1 << (largestBits - maxNbBits);
|
231
|
+
U32 n = lastNonNull;
|
232
|
+
|
233
|
+
while (huffNode[n].nbBits > maxNbBits) {
|
234
|
+
totalCost += baseCost - (1 << (largestBits - huffNode[n].nbBits));
|
235
|
+
huffNode[n].nbBits = (BYTE)maxNbBits;
|
236
|
+
n --;
|
237
|
+
} /* n stops at huffNode[n].nbBits <= maxNbBits */
|
238
|
+
while (huffNode[n].nbBits == maxNbBits) n--; /* n end at index of smallest symbol using < maxNbBits */
|
239
|
+
|
240
|
+
/* renorm totalCost */
|
241
|
+
totalCost >>= (largestBits - maxNbBits); /* note : totalCost is necessarily a multiple of baseCost */
|
242
|
+
|
243
|
+
/* repay normalized cost */
|
244
|
+
{ U32 const noSymbol = 0xF0F0F0F0;
|
245
|
+
U32 rankLast[HUF_TABLELOG_MAX+2];
|
246
|
+
int pos;
|
247
|
+
|
248
|
+
/* Get pos of last (smallest) symbol per rank */
|
249
|
+
memset(rankLast, 0xF0, sizeof(rankLast));
|
250
|
+
{ U32 currentNbBits = maxNbBits;
|
251
|
+
for (pos=n ; pos >= 0; pos--) {
|
252
|
+
if (huffNode[pos].nbBits >= currentNbBits) continue;
|
253
|
+
currentNbBits = huffNode[pos].nbBits; /* < maxNbBits */
|
254
|
+
rankLast[maxNbBits-currentNbBits] = pos;
|
255
|
+
} }
|
256
|
+
|
257
|
+
while (totalCost > 0) {
|
258
|
+
U32 nBitsToDecrease = BIT_highbit32(totalCost) + 1;
|
259
|
+
for ( ; nBitsToDecrease > 1; nBitsToDecrease--) {
|
260
|
+
U32 highPos = rankLast[nBitsToDecrease];
|
261
|
+
U32 lowPos = rankLast[nBitsToDecrease-1];
|
262
|
+
if (highPos == noSymbol) continue;
|
263
|
+
if (lowPos == noSymbol) break;
|
264
|
+
{ U32 const highTotal = huffNode[highPos].count;
|
265
|
+
U32 const lowTotal = 2 * huffNode[lowPos].count;
|
266
|
+
if (highTotal <= lowTotal) break;
|
267
|
+
} }
|
268
|
+
/* only triggered when no more rank 1 symbol left => find closest one (note : there is necessarily at least one !) */
|
269
|
+
while ((nBitsToDecrease<=HUF_TABLELOG_MAX) && (rankLast[nBitsToDecrease] == noSymbol)) /* HUF_MAX_TABLELOG test just to please gcc 5+; but it should not be necessary */
|
270
|
+
nBitsToDecrease ++;
|
271
|
+
totalCost -= 1 << (nBitsToDecrease-1);
|
272
|
+
if (rankLast[nBitsToDecrease-1] == noSymbol)
|
273
|
+
rankLast[nBitsToDecrease-1] = rankLast[nBitsToDecrease]; /* this rank is no longer empty */
|
274
|
+
huffNode[rankLast[nBitsToDecrease]].nbBits ++;
|
275
|
+
if (rankLast[nBitsToDecrease] == 0) /* special case, reached largest symbol */
|
276
|
+
rankLast[nBitsToDecrease] = noSymbol;
|
277
|
+
else {
|
278
|
+
rankLast[nBitsToDecrease]--;
|
279
|
+
if (huffNode[rankLast[nBitsToDecrease]].nbBits != maxNbBits-nBitsToDecrease)
|
280
|
+
rankLast[nBitsToDecrease] = noSymbol; /* this rank is now empty */
|
281
|
+
} } /* while (totalCost > 0) */
|
282
|
+
|
283
|
+
while (totalCost < 0) { /* Sometimes, cost correction overshoot */
|
284
|
+
if (rankLast[1] == noSymbol) { /* special case : no rank 1 symbol (using maxNbBits-1); let's create one from largest rank 0 (using maxNbBits) */
|
285
|
+
while (huffNode[n].nbBits == maxNbBits) n--;
|
286
|
+
huffNode[n+1].nbBits--;
|
287
|
+
rankLast[1] = n+1;
|
288
|
+
totalCost++;
|
289
|
+
continue;
|
290
|
+
}
|
291
|
+
huffNode[ rankLast[1] + 1 ].nbBits--;
|
292
|
+
rankLast[1]++;
|
293
|
+
totalCost ++;
|
294
|
+
} } } /* there are several too large elements (at least >= 2) */
|
295
|
+
|
296
|
+
return maxNbBits;
|
297
|
+
}
|
298
|
+
|
299
|
+
|
300
|
+
typedef struct {
|
301
|
+
U32 base;
|
302
|
+
U32 current;
|
303
|
+
} rankPos;
|
304
|
+
|
305
|
+
static void HUF_sort(nodeElt* huffNode, const U32* count, U32 maxSymbolValue)
|
306
|
+
{
|
307
|
+
rankPos rank[32];
|
308
|
+
U32 n;
|
309
|
+
|
310
|
+
memset(rank, 0, sizeof(rank));
|
311
|
+
for (n=0; n<=maxSymbolValue; n++) {
|
312
|
+
U32 r = BIT_highbit32(count[n] + 1);
|
313
|
+
rank[r].base ++;
|
314
|
+
}
|
315
|
+
for (n=30; n>0; n--) rank[n-1].base += rank[n].base;
|
316
|
+
for (n=0; n<32; n++) rank[n].current = rank[n].base;
|
317
|
+
for (n=0; n<=maxSymbolValue; n++) {
|
318
|
+
U32 const c = count[n];
|
319
|
+
U32 const r = BIT_highbit32(c+1) + 1;
|
320
|
+
U32 pos = rank[r].current++;
|
321
|
+
while ((pos > rank[r].base) && (c > huffNode[pos-1].count)) huffNode[pos]=huffNode[pos-1], pos--;
|
322
|
+
huffNode[pos].count = c;
|
323
|
+
huffNode[pos].byte = (BYTE)n;
|
324
|
+
}
|
325
|
+
}
|
326
|
+
|
327
|
+
|
328
|
+
/** HUF_buildCTable_wksp() :
|
329
|
+
* Same as HUF_buildCTable(), but using externally allocated scratch buffer.
|
330
|
+
* `workSpace` must be aligned on 4-bytes boundaries, and be at least as large as a table of 1024 unsigned.
|
331
|
+
*/
|
332
|
+
#define STARTNODE (HUF_SYMBOLVALUE_MAX+1)
|
333
|
+
typedef nodeElt huffNodeTable[2*HUF_SYMBOLVALUE_MAX+1 +1];
|
334
|
+
size_t HUF_buildCTable_wksp (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits, void* workSpace, size_t wkspSize)
|
335
|
+
{
|
336
|
+
nodeElt* const huffNode0 = (nodeElt*)workSpace;
|
337
|
+
nodeElt* const huffNode = huffNode0+1;
|
338
|
+
U32 n, nonNullRank;
|
339
|
+
int lowS, lowN;
|
340
|
+
U16 nodeNb = STARTNODE;
|
341
|
+
U32 nodeRoot;
|
342
|
+
|
343
|
+
/* safety checks */
|
344
|
+
if (wkspSize < sizeof(huffNodeTable)) return ERROR(GENERIC); /* workSpace is not large enough */
|
345
|
+
if (maxNbBits == 0) maxNbBits = HUF_TABLELOG_DEFAULT;
|
346
|
+
if (maxSymbolValue > HUF_SYMBOLVALUE_MAX) return ERROR(GENERIC);
|
347
|
+
memset(huffNode0, 0, sizeof(huffNodeTable));
|
348
|
+
|
349
|
+
/* sort, decreasing order */
|
350
|
+
HUF_sort(huffNode, count, maxSymbolValue);
|
351
|
+
|
352
|
+
/* init for parents */
|
353
|
+
nonNullRank = maxSymbolValue;
|
354
|
+
while(huffNode[nonNullRank].count == 0) nonNullRank--;
|
355
|
+
lowS = nonNullRank; nodeRoot = nodeNb + lowS - 1; lowN = nodeNb;
|
356
|
+
huffNode[nodeNb].count = huffNode[lowS].count + huffNode[lowS-1].count;
|
357
|
+
huffNode[lowS].parent = huffNode[lowS-1].parent = nodeNb;
|
358
|
+
nodeNb++; lowS-=2;
|
359
|
+
for (n=nodeNb; n<=nodeRoot; n++) huffNode[n].count = (U32)(1U<<30);
|
360
|
+
huffNode0[0].count = (U32)(1U<<31); /* fake entry, strong barrier */
|
361
|
+
|
362
|
+
/* create parents */
|
363
|
+
while (nodeNb <= nodeRoot) {
|
364
|
+
U32 n1 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
|
365
|
+
U32 n2 = (huffNode[lowS].count < huffNode[lowN].count) ? lowS-- : lowN++;
|
366
|
+
huffNode[nodeNb].count = huffNode[n1].count + huffNode[n2].count;
|
367
|
+
huffNode[n1].parent = huffNode[n2].parent = nodeNb;
|
368
|
+
nodeNb++;
|
369
|
+
}
|
370
|
+
|
371
|
+
/* distribute weights (unlimited tree height) */
|
372
|
+
huffNode[nodeRoot].nbBits = 0;
|
373
|
+
for (n=nodeRoot-1; n>=STARTNODE; n--)
|
374
|
+
huffNode[n].nbBits = huffNode[ huffNode[n].parent ].nbBits + 1;
|
375
|
+
for (n=0; n<=nonNullRank; n++)
|
376
|
+
huffNode[n].nbBits = huffNode[ huffNode[n].parent ].nbBits + 1;
|
377
|
+
|
378
|
+
/* enforce maxTableLog */
|
379
|
+
maxNbBits = HUF_setMaxHeight(huffNode, nonNullRank, maxNbBits);
|
380
|
+
|
381
|
+
/* fill result into tree (val, nbBits) */
|
382
|
+
{ U16 nbPerRank[HUF_TABLELOG_MAX+1] = {0};
|
383
|
+
U16 valPerRank[HUF_TABLELOG_MAX+1] = {0};
|
384
|
+
if (maxNbBits > HUF_TABLELOG_MAX) return ERROR(GENERIC); /* check fit into table */
|
385
|
+
for (n=0; n<=nonNullRank; n++)
|
386
|
+
nbPerRank[huffNode[n].nbBits]++;
|
387
|
+
/* determine stating value per rank */
|
388
|
+
{ U16 min = 0;
|
389
|
+
for (n=maxNbBits; n>0; n--) {
|
390
|
+
valPerRank[n] = min; /* get starting value within each rank */
|
391
|
+
min += nbPerRank[n];
|
392
|
+
min >>= 1;
|
393
|
+
} }
|
394
|
+
for (n=0; n<=maxSymbolValue; n++)
|
395
|
+
tree[huffNode[n].byte].nbBits = huffNode[n].nbBits; /* push nbBits per symbol, symbol order */
|
396
|
+
for (n=0; n<=maxSymbolValue; n++)
|
397
|
+
tree[n].val = valPerRank[tree[n].nbBits]++; /* assign value within rank, symbol order */
|
398
|
+
}
|
399
|
+
|
400
|
+
return maxNbBits;
|
401
|
+
}
|
402
|
+
|
403
|
+
/** HUF_buildCTable() :
|
404
|
+
* Note : count is used before tree is written, so they can safely overlap
|
405
|
+
*/
|
406
|
+
size_t HUF_buildCTable (HUF_CElt* tree, const U32* count, U32 maxSymbolValue, U32 maxNbBits)
|
407
|
+
{
|
408
|
+
huffNodeTable nodeTable;
|
409
|
+
return HUF_buildCTable_wksp(tree, count, maxSymbolValue, maxNbBits, nodeTable, sizeof(nodeTable));
|
410
|
+
}
|
411
|
+
|
412
|
+
static void HUF_encodeSymbol(BIT_CStream_t* bitCPtr, U32 symbol, const HUF_CElt* CTable)
|
413
|
+
{
|
414
|
+
BIT_addBitsFast(bitCPtr, CTable[symbol].val, CTable[symbol].nbBits);
|
415
|
+
}
|
416
|
+
|
417
|
+
size_t HUF_compressBound(size_t size) { return HUF_COMPRESSBOUND(size); }
|
418
|
+
|
419
|
+
#define HUF_FLUSHBITS(s) (fast ? BIT_flushBitsFast(s) : BIT_flushBits(s))
|
420
|
+
|
421
|
+
#define HUF_FLUSHBITS_1(stream) \
|
422
|
+
if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*2+7) HUF_FLUSHBITS(stream)
|
423
|
+
|
424
|
+
#define HUF_FLUSHBITS_2(stream) \
|
425
|
+
if (sizeof((stream)->bitContainer)*8 < HUF_TABLELOG_MAX*4+7) HUF_FLUSHBITS(stream)
|
426
|
+
|
427
|
+
size_t HUF_compress1X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
|
428
|
+
{
|
429
|
+
const BYTE* ip = (const BYTE*) src;
|
430
|
+
BYTE* const ostart = (BYTE*)dst;
|
431
|
+
BYTE* const oend = ostart + dstSize;
|
432
|
+
BYTE* op = ostart;
|
433
|
+
size_t n;
|
434
|
+
const unsigned fast = (dstSize >= HUF_BLOCKBOUND(srcSize));
|
435
|
+
BIT_CStream_t bitC;
|
436
|
+
|
437
|
+
/* init */
|
438
|
+
if (dstSize < 8) return 0; /* not enough space to compress */
|
439
|
+
{ size_t const initErr = BIT_initCStream(&bitC, op, oend-op);
|
440
|
+
if (HUF_isError(initErr)) return 0; }
|
441
|
+
|
442
|
+
n = srcSize & ~3; /* join to mod 4 */
|
443
|
+
switch (srcSize & 3)
|
444
|
+
{
|
445
|
+
case 3 : HUF_encodeSymbol(&bitC, ip[n+ 2], CTable);
|
446
|
+
HUF_FLUSHBITS_2(&bitC);
|
447
|
+
case 2 : HUF_encodeSymbol(&bitC, ip[n+ 1], CTable);
|
448
|
+
HUF_FLUSHBITS_1(&bitC);
|
449
|
+
case 1 : HUF_encodeSymbol(&bitC, ip[n+ 0], CTable);
|
450
|
+
HUF_FLUSHBITS(&bitC);
|
451
|
+
case 0 :
|
452
|
+
default: ;
|
453
|
+
}
|
454
|
+
|
455
|
+
for (; n>0; n-=4) { /* note : n&3==0 at this stage */
|
456
|
+
HUF_encodeSymbol(&bitC, ip[n- 1], CTable);
|
457
|
+
HUF_FLUSHBITS_1(&bitC);
|
458
|
+
HUF_encodeSymbol(&bitC, ip[n- 2], CTable);
|
459
|
+
HUF_FLUSHBITS_2(&bitC);
|
460
|
+
HUF_encodeSymbol(&bitC, ip[n- 3], CTable);
|
461
|
+
HUF_FLUSHBITS_1(&bitC);
|
462
|
+
HUF_encodeSymbol(&bitC, ip[n- 4], CTable);
|
463
|
+
HUF_FLUSHBITS(&bitC);
|
464
|
+
}
|
465
|
+
|
466
|
+
return BIT_closeCStream(&bitC);
|
467
|
+
}
|
468
|
+
|
469
|
+
|
470
|
+
size_t HUF_compress4X_usingCTable(void* dst, size_t dstSize, const void* src, size_t srcSize, const HUF_CElt* CTable)
|
471
|
+
{
|
472
|
+
size_t const segmentSize = (srcSize+3)/4; /* first 3 segments */
|
473
|
+
const BYTE* ip = (const BYTE*) src;
|
474
|
+
const BYTE* const iend = ip + srcSize;
|
475
|
+
BYTE* const ostart = (BYTE*) dst;
|
476
|
+
BYTE* const oend = ostart + dstSize;
|
477
|
+
BYTE* op = ostart;
|
478
|
+
|
479
|
+
if (dstSize < 6 + 1 + 1 + 1 + 8) return 0; /* minimum space to compress successfully */
|
480
|
+
if (srcSize < 12) return 0; /* no saving possible : too small input */
|
481
|
+
op += 6; /* jumpTable */
|
482
|
+
|
483
|
+
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable) );
|
484
|
+
if (cSize==0) return 0;
|
485
|
+
MEM_writeLE16(ostart, (U16)cSize);
|
486
|
+
op += cSize;
|
487
|
+
}
|
488
|
+
|
489
|
+
ip += segmentSize;
|
490
|
+
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable) );
|
491
|
+
if (cSize==0) return 0;
|
492
|
+
MEM_writeLE16(ostart+2, (U16)cSize);
|
493
|
+
op += cSize;
|
494
|
+
}
|
495
|
+
|
496
|
+
ip += segmentSize;
|
497
|
+
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend-op, ip, segmentSize, CTable) );
|
498
|
+
if (cSize==0) return 0;
|
499
|
+
MEM_writeLE16(ostart+4, (U16)cSize);
|
500
|
+
op += cSize;
|
501
|
+
}
|
502
|
+
|
503
|
+
ip += segmentSize;
|
504
|
+
{ CHECK_V_F(cSize, HUF_compress1X_usingCTable(op, oend-op, ip, iend-ip, CTable) );
|
505
|
+
if (cSize==0) return 0;
|
506
|
+
op += cSize;
|
507
|
+
}
|
508
|
+
|
509
|
+
return op-ostart;
|
510
|
+
}
|
511
|
+
|
512
|
+
|
513
|
+
/* `workSpace` must a table of at least 1024 unsigned */
|
514
|
+
static size_t HUF_compress_internal (
|
515
|
+
void* dst, size_t dstSize,
|
516
|
+
const void* src, size_t srcSize,
|
517
|
+
unsigned maxSymbolValue, unsigned huffLog,
|
518
|
+
unsigned singleStream,
|
519
|
+
void* workSpace, size_t wkspSize)
|
520
|
+
{
|
521
|
+
BYTE* const ostart = (BYTE*)dst;
|
522
|
+
BYTE* const oend = ostart + dstSize;
|
523
|
+
BYTE* op = ostart;
|
524
|
+
|
525
|
+
union {
|
526
|
+
U32 count[HUF_SYMBOLVALUE_MAX+1];
|
527
|
+
HUF_CElt CTable[HUF_SYMBOLVALUE_MAX+1];
|
528
|
+
} table; /* `count` can overlap with `CTable`; saves 1 KB */
|
529
|
+
|
530
|
+
/* checks & inits */
|
531
|
+
if (wkspSize < sizeof(huffNodeTable)) return ERROR(GENERIC);
|
532
|
+
if (!srcSize) return 0; /* Uncompressed (note : 1 means rle, so first byte must be correct) */
|
533
|
+
if (!dstSize) return 0; /* cannot fit within dst budget */
|
534
|
+
if (srcSize > HUF_BLOCKSIZE_MAX) return ERROR(srcSize_wrong); /* current block size limit */
|
535
|
+
if (huffLog > HUF_TABLELOG_MAX) return ERROR(tableLog_tooLarge);
|
536
|
+
if (!maxSymbolValue) maxSymbolValue = HUF_SYMBOLVALUE_MAX;
|
537
|
+
if (!huffLog) huffLog = HUF_TABLELOG_DEFAULT;
|
538
|
+
|
539
|
+
/* Scan input and build symbol stats */
|
540
|
+
{ CHECK_V_F(largest, FSE_count_wksp (table.count, &maxSymbolValue, (const BYTE*)src, srcSize, (U32*)workSpace) );
|
541
|
+
if (largest == srcSize) { *ostart = ((const BYTE*)src)[0]; return 1; } /* single symbol, rle */
|
542
|
+
if (largest <= (srcSize >> 7)+1) return 0; /* Fast heuristic : not compressible enough */
|
543
|
+
}
|
544
|
+
|
545
|
+
/* Build Huffman Tree */
|
546
|
+
huffLog = HUF_optimalTableLog(huffLog, srcSize, maxSymbolValue);
|
547
|
+
{ CHECK_V_F(maxBits, HUF_buildCTable_wksp (table.CTable, table.count, maxSymbolValue, huffLog, workSpace, wkspSize) );
|
548
|
+
huffLog = (U32)maxBits;
|
549
|
+
}
|
550
|
+
|
551
|
+
/* Write table description header */
|
552
|
+
{ CHECK_V_F(hSize, HUF_writeCTable (op, dstSize, table.CTable, maxSymbolValue, huffLog) );
|
553
|
+
if (hSize + 12 >= srcSize) return 0; /* not useful to try compression */
|
554
|
+
op += hSize;
|
555
|
+
}
|
556
|
+
|
557
|
+
/* Compress */
|
558
|
+
{ size_t const cSize = (singleStream) ?
|
559
|
+
HUF_compress1X_usingCTable(op, oend - op, src, srcSize, table.CTable) : /* single segment */
|
560
|
+
HUF_compress4X_usingCTable(op, oend - op, src, srcSize, table.CTable);
|
561
|
+
if (HUF_isError(cSize)) return cSize;
|
562
|
+
if (cSize==0) return 0; /* uncompressible */
|
563
|
+
op += cSize;
|
564
|
+
}
|
565
|
+
|
566
|
+
/* check compressibility */
|
567
|
+
if ((size_t)(op-ostart) >= srcSize-1)
|
568
|
+
return 0;
|
569
|
+
|
570
|
+
return op-ostart;
|
571
|
+
}
|
572
|
+
|
573
|
+
|
574
|
+
size_t HUF_compress1X_wksp (void* dst, size_t dstSize,
|
575
|
+
const void* src, size_t srcSize,
|
576
|
+
unsigned maxSymbolValue, unsigned huffLog,
|
577
|
+
void* workSpace, size_t wkspSize)
|
578
|
+
{
|
579
|
+
return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 1 /* single stream */, workSpace, wkspSize);
|
580
|
+
}
|
581
|
+
|
582
|
+
size_t HUF_compress1X (void* dst, size_t dstSize,
|
583
|
+
const void* src, size_t srcSize,
|
584
|
+
unsigned maxSymbolValue, unsigned huffLog)
|
585
|
+
{
|
586
|
+
unsigned workSpace[1024];
|
587
|
+
return HUF_compress1X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
|
588
|
+
}
|
589
|
+
|
590
|
+
size_t HUF_compress4X_wksp (void* dst, size_t dstSize,
|
591
|
+
const void* src, size_t srcSize,
|
592
|
+
unsigned maxSymbolValue, unsigned huffLog,
|
593
|
+
void* workSpace, size_t wkspSize)
|
594
|
+
{
|
595
|
+
return HUF_compress_internal(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, 0 /* 4 streams */, workSpace, wkspSize);
|
596
|
+
}
|
597
|
+
|
598
|
+
size_t HUF_compress2 (void* dst, size_t dstSize,
|
599
|
+
const void* src, size_t srcSize,
|
600
|
+
unsigned maxSymbolValue, unsigned huffLog)
|
601
|
+
{
|
602
|
+
unsigned workSpace[1024];
|
603
|
+
return HUF_compress4X_wksp(dst, dstSize, src, srcSize, maxSymbolValue, huffLog, workSpace, sizeof(workSpace));
|
604
|
+
}
|
605
|
+
|
606
|
+
size_t HUF_compress (void* dst, size_t maxDstSize, const void* src, size_t srcSize)
|
607
|
+
{
|
608
|
+
return HUF_compress2(dst, maxDstSize, src, (U32)srcSize, 255, HUF_TABLELOG_DEFAULT);
|
609
|
+
}
|