1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
|
/*
LZ4 HC - High Compression Mode of LZ4
Copyright (C) 2011-2017, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- LZ4 source repository : https://github.com/lz4/lz4
- LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
*/
/* note : lz4hc is not an independent module, it requires lz4.h/lz4.c for proper compilation */
/* *************************************
* Tuning Parameter
***************************************/
/*! HEAPMODE :
* Select how default compression function will allocate workplace memory,
* in stack (0:fastest), or in heap (1:requires malloc()).
* Since workplace is rather large, heap mode is recommended.
*/
#ifndef LZ4HC_HEAPMODE
# define LZ4HC_HEAPMODE 1
#endif
/*=== Dependency ===*/
#define LZ4_HC_STATIC_LINKING_ONLY
#include "lz4hc.h"
/*=== Common LZ4 definitions ===*/
#if defined(__GNUC__)
# pragma GCC diagnostic ignored "-Wunused-function"
#endif
#if defined (__clang__)
# pragma clang diagnostic ignored "-Wunused-function"
#endif
#define LZ4_COMMONDEFS_ONLY
#include "lz4.c" /* LZ4_count, constants, mem */
/*=== Constants ===*/
#define OPTIMAL_ML (int)((ML_MASK-1)+MINMATCH)
#define LZ4_OPT_NUM (1<<12)
/*=== Macros ===*/
#define MIN(a,b) ( (a) < (b) ? (a) : (b) )
#define MAX(a,b) ( (a) > (b) ? (a) : (b) )
#define HASH_FUNCTION(i) (((i) * 2654435761U) >> ((MINMATCH*8)-LZ4HC_HASH_LOG))
#define DELTANEXTMAXD(p) chainTable[(p) & LZ4HC_MAXD_MASK] /* flexible, LZ4HC_MAXD dependent */
#define DELTANEXTU16(table, pos) table[(U16)(pos)] /* faster */
static U32 LZ4HC_hashPtr(const void* ptr) { return HASH_FUNCTION(LZ4_read32(ptr)); }
/*=== Enums ===*/
typedef enum { noDictCtx, usingDictCtx } dictCtx_directive;
/**************************************
* HC Compression
**************************************/
static void LZ4HC_clearTables (LZ4HC_CCtx_internal* hc4)
{
DEBUGLOG(4, "LZ4HC_clearTables(%p)", hc4);
MEM_INIT((void*)hc4->hashTable, 0, sizeof(hc4->hashTable));
MEM_INIT(hc4->chainTable, 0xFF, sizeof(hc4->chainTable));
}
static void LZ4HC_init (LZ4HC_CCtx_internal* hc4, const BYTE* start)
{
uptrval startingOffset = hc4->end - hc4->base;
DEBUGLOG(4, "LZ4HC_init(%p, %p)", hc4, start);
if (startingOffset > 1 GB) {
LZ4HC_clearTables(hc4);
startingOffset = 0;
}
startingOffset += 64 KB;
hc4->nextToUpdate = (U32) startingOffset;
hc4->base = start - startingOffset;
hc4->end = start;
hc4->dictBase = start - startingOffset;
hc4->dictLimit = (U32) startingOffset;
hc4->lowLimit = (U32) startingOffset;
}
/* Update chains up to ip (excluded) */
LZ4_FORCE_INLINE void LZ4HC_Insert (LZ4HC_CCtx_internal* hc4, const BYTE* ip)
{
U16* const chainTable = hc4->chainTable;
U32* const hashTable = hc4->hashTable;
const BYTE* const base = hc4->base;
U32 const target = (U32)(ip - base);
U32 idx = hc4->nextToUpdate;
while (idx < target) {
U32 const h = LZ4HC_hashPtr(base+idx);
size_t delta = idx - hashTable[h];
if (delta>MAX_DISTANCE) delta = MAX_DISTANCE;
DELTANEXTU16(chainTable, idx) = (U16)delta;
hashTable[h] = idx;
idx++;
}
hc4->nextToUpdate = target;
}
/** LZ4HC_countBack() :
* @return : negative value, nb of common bytes before ip/match */
LZ4_FORCE_INLINE
int LZ4HC_countBack(const BYTE* const ip, const BYTE* const match,
const BYTE* const iMin, const BYTE* const mMin)
{
int back = 0;
int const min = (int)MAX(iMin - ip, mMin - match);
assert(ip >= iMin); assert((size_t)(ip-iMin) < (1U<<31));
assert(match >= mMin); assert((size_t)(match - mMin) < (1U<<31));
while ( (back > min)
&& (ip[back-1] == match[back-1]) )
back--;
return back;
}
/* LZ4HC_countPattern() :
* pattern32 must be a sample of repetitive pattern of length 1, 2 or 4 (but not 3!) */
static unsigned
LZ4HC_countPattern(const BYTE* ip, const BYTE* const iEnd, U32 const pattern32)
{
const BYTE* const iStart = ip;
reg_t const pattern = (sizeof(pattern)==8) ? (reg_t)pattern32 + (((reg_t)pattern32) << 32) : pattern32;
while (likely(ip < iEnd-(sizeof(pattern)-1))) {
reg_t const diff = LZ4_read_ARCH(ip) ^ pattern;
if (!diff) { ip+=sizeof(pattern); continue; }
ip += LZ4_NbCommonBytes(diff);
return (unsigned)(ip - iStart);
}
if (LZ4_isLittleEndian()) {
reg_t patternByte = pattern;
while ((ip<iEnd) && (*ip == (BYTE)patternByte)) {
ip++; patternByte >>= 8;
}
} else { /* big endian */
U32 bitOffset = (sizeof(pattern)*8) - 8;
while (ip < iEnd) {
BYTE const byte = (BYTE)(pattern >> bitOffset);
if (*ip != byte) break;
ip ++; bitOffset -= 8;
}
}
return (unsigned)(ip - iStart);
}
/* LZ4HC_reverseCountPattern() :
* pattern must be a sample of repetitive pattern of length 1, 2 or 4 (but not 3!)
* read using natural platform endianess */
static unsigned
LZ4HC_reverseCountPattern(const BYTE* ip, const BYTE* const iLow, U32 pattern)
{
const BYTE* const iStart = ip;
while (likely(ip >= iLow+4)) {
if (LZ4_read32(ip-4) != pattern) break;
ip -= 4;
}
{ const BYTE* bytePtr = (const BYTE*)(&pattern) + 3; /* works for any endianess */
while (likely(ip>iLow)) {
if (ip[-1] != *bytePtr) break;
ip--; bytePtr--;
} }
return (unsigned)(iStart - ip);
}
typedef enum { rep_untested, rep_not, rep_confirmed } repeat_state_e;
LZ4_FORCE_INLINE int
LZ4HC_InsertAndGetWiderMatch (
LZ4HC_CCtx_internal* hc4,
const BYTE* const ip,
const BYTE* const iLowLimit,
const BYTE* const iHighLimit,
int longest,
const BYTE** matchpos,
const BYTE** startpos,
const int maxNbAttempts,
const int patternAnalysis,
const dictCtx_directive dict)
{
U16* const chainTable = hc4->chainTable;
U32* const HashTable = hc4->hashTable;
const LZ4HC_CCtx_internal * const dictCtx = hc4->dictCtx;
const BYTE* const base = hc4->base;
const U32 dictLimit = hc4->dictLimit;
const BYTE* const lowPrefixPtr = base + dictLimit;
const U32 ipIndex = (U32)(ip - base);
const U32 lowLimit = (hc4->lowLimit + 64 KB > ipIndex) ? hc4->lowLimit : ipIndex - MAX_DISTANCE;
const BYTE* const dictBase = hc4->dictBase;
int const delta = (int)(ip-iLowLimit);
int nbAttempts = maxNbAttempts;
U32 const pattern = LZ4_read32(ip);
U32 matchIndex;
U32 dictMatchIndex;
repeat_state_e repeat = rep_untested;
size_t srcPatternLength = 0;
DEBUGLOG(7, "LZ4HC_InsertAndGetWiderMatch");
/* First Match */
LZ4HC_Insert(hc4, ip);
matchIndex = HashTable[LZ4HC_hashPtr(ip)];
DEBUGLOG(7, "First match at index %u / %u (lowLimit)",
matchIndex, lowLimit);
while ((matchIndex>=lowLimit) && (nbAttempts)) {
DEBUGLOG(7, "remaining attempts : %i", nbAttempts);
nbAttempts--;
if (matchIndex >= dictLimit) {
const BYTE* const matchPtr = base + matchIndex;
assert(longest >= 1);
if (LZ4_read16(iLowLimit + longest - 1) == LZ4_read16(matchPtr - delta + longest - 1)) {
if (LZ4_read32(matchPtr) == pattern) {
int mlt = MINMATCH + LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, iHighLimit);
int const back = delta ? LZ4HC_countBack(ip, matchPtr, iLowLimit, lowPrefixPtr) : 0;
mlt -= back;
if (mlt > longest) {
longest = mlt;
*matchpos = matchPtr+back;
*startpos = ip+back;
} }
}
} else { /* matchIndex < dictLimit */
const BYTE* const matchPtr = dictBase + matchIndex;
if (LZ4_read32(matchPtr) == pattern) {
int mlt;
int back = 0;
const BYTE* vLimit = ip + (dictLimit - matchIndex);
if (vLimit > iHighLimit) vLimit = iHighLimit;
mlt = LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH;
if ((ip+mlt == vLimit) && (vLimit < iHighLimit))
mlt += LZ4_count(ip+mlt, base+dictLimit, iHighLimit);
back = delta ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictBase+lowLimit) : 0;
mlt -= back;
if (mlt > longest) {
longest = mlt;
*matchpos = base + matchIndex + back;
*startpos = ip + back;
} } }
{ U32 const nextOffset = DELTANEXTU16(chainTable, matchIndex);
matchIndex -= nextOffset;
if (patternAnalysis && nextOffset==1) {
/* may be a repeated pattern */
if (repeat == rep_untested) {
if ( ((pattern & 0xFFFF) == (pattern >> 16))
& ((pattern & 0xFF) == (pattern >> 24)) ) {
repeat = rep_confirmed;
srcPatternLength = LZ4HC_countPattern(ip+4, iHighLimit, pattern) + 4;
} else {
repeat = rep_not;
} }
if ( (repeat == rep_confirmed)
&& (matchIndex >= dictLimit) ) { /* same segment only */
const BYTE* const matchPtr = base + matchIndex;
if (LZ4_read32(matchPtr) == pattern) { /* good candidate */
size_t const forwardPatternLength = LZ4HC_countPattern(matchPtr+sizeof(pattern), iHighLimit, pattern) + sizeof(pattern);
const BYTE* const maxLowPtr = (lowPrefixPtr + MAX_DISTANCE >= ip) ? lowPrefixPtr : ip - MAX_DISTANCE;
size_t const backLength = LZ4HC_reverseCountPattern(matchPtr, maxLowPtr, pattern);
size_t const currentSegmentLength = backLength + forwardPatternLength;
if ( (currentSegmentLength >= srcPatternLength) /* current pattern segment large enough to contain full srcPatternLength */
&& (forwardPatternLength <= srcPatternLength) ) { /* haven't reached this position yet */
matchIndex += (U32)forwardPatternLength - (U32)srcPatternLength; /* best position, full pattern, might be followed by more match */
} else {
matchIndex -= (U32)backLength; /* let's go to farthest segment position, will find a match of length currentSegmentLength + maybe some back */
}
} } } }
} /* while ((matchIndex>=lowLimit) && (nbAttempts)) */
if (dict == usingDictCtx && nbAttempts && ipIndex - lowLimit < MAX_DISTANCE) {
size_t const dictEndOffset = dictCtx->end - dictCtx->base;
assert(dictEndOffset <= 1 GB);
dictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)];
matchIndex = dictMatchIndex + lowLimit - (U32)dictEndOffset;
while (ipIndex - matchIndex <= MAX_DISTANCE && nbAttempts--) {
const BYTE* const matchPtr = dictCtx->base + dictMatchIndex;
if (LZ4_read32(matchPtr) == pattern) {
int mlt;
int back = 0;
const BYTE* vLimit = ip + (dictEndOffset - dictMatchIndex);
if (vLimit > iHighLimit) vLimit = iHighLimit;
mlt = LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH;
back = delta ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->base + dictCtx->dictLimit) : 0;
mlt -= back;
if (mlt > longest) {
longest = mlt;
*matchpos = base + matchIndex + back;
*startpos = ip + back;
}
}
{
U32 const nextOffset = DELTANEXTU16(dictCtx->chainTable, dictMatchIndex);
dictMatchIndex -= nextOffset;
matchIndex -= nextOffset;
}
}
}
return longest;
}
LZ4_FORCE_INLINE
int LZ4HC_InsertAndFindBestMatch(LZ4HC_CCtx_internal* const hc4, /* Index table will be updated */
const BYTE* const ip, const BYTE* const iLimit,
const BYTE** matchpos,
const int maxNbAttempts,
const int patternAnalysis,
const dictCtx_directive dict)
{
const BYTE* uselessPtr = ip;
/* note : LZ4HC_InsertAndGetWiderMatch() is able to modify the starting position of a match (*startpos),
* but this won't be the case here, as we define iLowLimit==ip,
* so LZ4HC_InsertAndGetWiderMatch() won't be allowed to search past ip */
return LZ4HC_InsertAndGetWiderMatch(hc4, ip, ip, iLimit, MINMATCH-1, matchpos, &uselessPtr, maxNbAttempts, patternAnalysis, dict);
}
typedef enum {
noLimit = 0,
limitedOutput = 1,
limitedDestSize = 2,
} limitedOutput_directive;
/* LZ4HC_encodeSequence() :
* @return : 0 if ok,
* 1 if buffer issue detected */
LZ4_FORCE_INLINE int LZ4HC_encodeSequence (
const BYTE** ip,
BYTE** op,
const BYTE** anchor,
int matchLength,
const BYTE* const match,
limitedOutput_directive limit,
BYTE* oend)
{
size_t length;
BYTE* const token = (*op)++;
#if defined(LZ4_DEBUG) && (LZ4_DEBUG >= 6)
static const BYTE* start = NULL;
static U32 totalCost = 0;
U32 const pos = (start==NULL) ? 0 : (U32)(*anchor - start);
U32 const ll = (U32)(*ip - *anchor);
U32 const llAdd = (ll>=15) ? ((ll-15) / 255) + 1 : 0;
U32 const mlAdd = (matchLength>=19) ? ((matchLength-19) / 255) + 1 : 0;
U32 const cost = 1 + llAdd + ll + 2 + mlAdd;
if (start==NULL) start = *anchor; /* only works for single segment */
/* g_debuglog_enable = (pos >= 2228) & (pos <= 2262); */
DEBUGLOG(6, "pos:%7u -- literals:%3u, match:%4i, offset:%5u, cost:%3u + %u",
pos,
(U32)(*ip - *anchor), matchLength, (U32)(*ip-match),
cost, totalCost);
totalCost += cost;
#endif
/* Encode Literal length */
length = (size_t)(*ip - *anchor);
if ((limit) && ((*op + (length >> 8) + length + (2 + 1 + LASTLITERALS)) > oend)) return 1; /* Check output limit */
if (length >= RUN_MASK) {
size_t len = length - RUN_MASK;
*token = (RUN_MASK << ML_BITS);
for(; len >= 255 ; len -= 255) *(*op)++ = 255;
*(*op)++ = (BYTE)len;
} else {
*token = (BYTE)(length << ML_BITS);
}
/* Copy Literals */
LZ4_wildCopy(*op, *anchor, (*op) + length);
*op += length;
/* Encode Offset */
LZ4_writeLE16(*op, (U16)(*ip-match)); *op += 2;
/* Encode MatchLength */
assert(matchLength >= MINMATCH);
length = (size_t)(matchLength - MINMATCH);
if ((limit) && (*op + (length >> 8) + (1 + LASTLITERALS) > oend)) return 1; /* Check output limit */
if (length >= ML_MASK) {
*token += ML_MASK;
length -= ML_MASK;
for(; length >= 510 ; length -= 510) { *(*op)++ = 255; *(*op)++ = 255; }
if (length >= 255) { length -= 255; *(*op)++ = 255; }
*(*op)++ = (BYTE)length;
} else {
*token += (BYTE)(length);
}
/* Prepare next loop */
*ip += matchLength;
*anchor = *ip;
return 0;
}
LZ4_FORCE_INLINE int LZ4HC_compress_hashChain (
LZ4HC_CCtx_internal* const ctx,
const char* const source,
char* const dest,
int* srcSizePtr,
int const maxOutputSize,
unsigned maxNbAttempts,
const limitedOutput_directive limit,
const dictCtx_directive dict
)
{
const int inputSize = *srcSizePtr;
const int patternAnalysis = (maxNbAttempts > 64); /* levels 8+ */
const BYTE* ip = (const BYTE*) source;
const BYTE* anchor = ip;
const BYTE* const iend = ip + inputSize;
const BYTE* const mflimit = iend - MFLIMIT;
const BYTE* const matchlimit = (iend - LASTLITERALS);
BYTE* optr = (BYTE*) dest;
BYTE* op = (BYTE*) dest;
BYTE* oend = op + maxOutputSize;
int ml, ml2, ml3, ml0;
const BYTE* ref = NULL;
const BYTE* start2 = NULL;
const BYTE* ref2 = NULL;
const BYTE* start3 = NULL;
const BYTE* ref3 = NULL;
const BYTE* start0;
const BYTE* ref0;
/* init */
*srcSizePtr = 0;
if (limit == limitedDestSize) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */
if (inputSize < LZ4_minLength) goto _last_literals; /* Input too small, no compression (all literals) */
/* Main Loop */
while (ip <= mflimit) {
ml = LZ4HC_InsertAndFindBestMatch (ctx, ip, matchlimit, &ref, maxNbAttempts, patternAnalysis, dict);
if (ml<MINMATCH) { ip++; continue; }
/* saved, in case we would skip too much */
start0 = ip;
ref0 = ref;
ml0 = ml;
_Search2:
if (ip+ml <= mflimit)
ml2 = LZ4HC_InsertAndGetWiderMatch(ctx,
ip + ml - 2, ip + 0, matchlimit, ml, &ref2, &start2,
maxNbAttempts, patternAnalysis, dict);
else
ml2 = ml;
if (ml2 == ml) { /* No better match */
optr = op;
if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) goto _dest_overflow;
continue;
}
if (start0 < ip) {
if (start2 < ip + ml0) { /* empirical */
ip = start0;
ref = ref0;
ml = ml0;
}
}
/* Here, start0==ip */
if ((start2 - ip) < 3) { /* First Match too small : removed */
ml = ml2;
ip = start2;
ref =ref2;
goto _Search2;
}
_Search3:
/* At this stage, we have :
* ml2 > ml1, and
* ip1+3 <= ip2 (usually < ip1+ml1) */
if ((start2 - ip) < OPTIMAL_ML) {
int correction;
int new_ml = ml;
if (new_ml > OPTIMAL_ML) new_ml = OPTIMAL_ML;
if (ip+new_ml > start2 + ml2 - MINMATCH) new_ml = (int)(start2 - ip) + ml2 - MINMATCH;
correction = new_ml - (int)(start2 - ip);
if (correction > 0) {
start2 += correction;
ref2 += correction;
ml2 -= correction;
}
}
/* Now, we have start2 = ip+new_ml, with new_ml = min(ml, OPTIMAL_ML=18) */
if (start2 + ml2 <= mflimit)
ml3 = LZ4HC_InsertAndGetWiderMatch(ctx,
start2 + ml2 - 3, start2, matchlimit, ml2, &ref3, &start3,
maxNbAttempts, patternAnalysis, dict);
else
ml3 = ml2;
if (ml3 == ml2) { /* No better match : 2 sequences to encode */
/* ip & ref are known; Now for ml */
if (start2 < ip+ml) ml = (int)(start2 - ip);
/* Now, encode 2 sequences */
optr = op;
if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) goto _dest_overflow;
ip = start2;
optr = op;
if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml2, ref2, limit, oend)) goto _dest_overflow;
continue;
}
if (start3 < ip+ml+3) { /* Not enough space for match 2 : remove it */
if (start3 >= (ip+ml)) { /* can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1 */
if (start2 < ip+ml) {
int correction = (int)(ip+ml - start2);
start2 += correction;
ref2 += correction;
ml2 -= correction;
if (ml2 < MINMATCH) {
start2 = start3;
ref2 = ref3;
ml2 = ml3;
}
}
optr = op;
if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) goto _dest_overflow;
ip = start3;
ref = ref3;
ml = ml3;
start0 = start2;
ref0 = ref2;
ml0 = ml2;
goto _Search2;
}
start2 = start3;
ref2 = ref3;
ml2 = ml3;
goto _Search3;
}
/*
* OK, now we have 3 ascending matches; let's write at least the first one
* ip & ref are known; Now for ml
*/
if (start2 < ip+ml) {
if ((start2 - ip) < (int)ML_MASK) {
int correction;
if (ml > OPTIMAL_ML) ml = OPTIMAL_ML;
if (ip + ml > start2 + ml2 - MINMATCH) ml = (int)(start2 - ip) + ml2 - MINMATCH;
correction = ml - (int)(start2 - ip);
if (correction > 0) {
start2 += correction;
ref2 += correction;
ml2 -= correction;
}
} else {
ml = (int)(start2 - ip);
}
}
optr = op;
if (LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ref, limit, oend)) goto _dest_overflow;
ip = start2;
ref = ref2;
ml = ml2;
start2 = start3;
ref2 = ref3;
ml2 = ml3;
goto _Search3;
}
_last_literals:
/* Encode Last Literals */
{ size_t lastRunSize = (size_t)(iend - anchor); /* literals */
size_t litLength = (lastRunSize + 255 - RUN_MASK) / 255;
size_t const totalSize = 1 + litLength + lastRunSize;
if (limit == limitedDestSize) oend += LASTLITERALS; /* restore correct value */
if (limit && (op + totalSize > oend)) {
if (limit == limitedOutput) return 0; /* Check output limit */
/* adapt lastRunSize to fill 'dest' */
lastRunSize = (size_t)(oend - op) - 1;
litLength = (lastRunSize + 255 - RUN_MASK) / 255;
lastRunSize -= litLength;
}
ip = anchor + lastRunSize;
if (lastRunSize >= RUN_MASK) {
size_t accumulator = lastRunSize - RUN_MASK;
*op++ = (RUN_MASK << ML_BITS);
for(; accumulator >= 255 ; accumulator -= 255) *op++ = 255;
*op++ = (BYTE) accumulator;
} else {
*op++ = (BYTE)(lastRunSize << ML_BITS);
}
memcpy(op, anchor, lastRunSize);
op += lastRunSize;
}
/* End */
*srcSizePtr = (int) (((const char*)ip) - source);
return (int) (((char*)op)-dest);
_dest_overflow:
if (limit == limitedDestSize) {
op = optr; /* restore correct out pointer */
goto _last_literals;
}
return 0;
}
static int LZ4HC_compress_optimal( LZ4HC_CCtx_internal* ctx,
const char* const source, char* dst,
int* srcSizePtr, int dstCapacity,
int const nbSearches, size_t sufficient_len,
const limitedOutput_directive limit, int const fullUpdate,
const dictCtx_directive dict);
LZ4_FORCE_INLINE int LZ4HC_compress_generic_internal (
LZ4HC_CCtx_internal* const ctx,
const char* const src,
char* const dst,
int* const srcSizePtr,
int const dstCapacity,
int cLevel,
const limitedOutput_directive limit,
const dictCtx_directive dict
)
{
typedef enum { lz4hc, lz4opt } lz4hc_strat_e;
typedef struct {
lz4hc_strat_e strat;
U32 nbSearches;
U32 targetLength;
} cParams_t;
static const cParams_t clTable[LZ4HC_CLEVEL_MAX+1] = {
{ lz4hc, 2, 16 }, /* 0, unused */
{ lz4hc, 2, 16 }, /* 1, unused */
{ lz4hc, 2, 16 }, /* 2, unused */
{ lz4hc, 4, 16 }, /* 3 */
{ lz4hc, 8, 16 }, /* 4 */
{ lz4hc, 16, 16 }, /* 5 */
{ lz4hc, 32, 16 }, /* 6 */
{ lz4hc, 64, 16 }, /* 7 */
{ lz4hc, 128, 16 }, /* 8 */
{ lz4hc, 256, 16 }, /* 9 */
{ lz4opt, 96, 64 }, /*10==LZ4HC_CLEVEL_OPT_MIN*/
{ lz4opt, 512,128 }, /*11 */
{ lz4opt,8192, LZ4_OPT_NUM }, /* 12==LZ4HC_CLEVEL_MAX */
};
DEBUGLOG(4, "LZ4HC_compress_generic(%p, %p, %d)", ctx, src, *srcSizePtr);
if (limit == limitedDestSize && dstCapacity < 1) return 0; /* Impossible to store anything */
if ((U32)*srcSizePtr > (U32)LZ4_MAX_INPUT_SIZE) return 0; /* Unsupported input size (too large or negative) */
ctx->end += *srcSizePtr;
if (cLevel < 1) cLevel = LZ4HC_CLEVEL_DEFAULT; /* note : convention is different from lz4frame, maybe something to review */
cLevel = MIN(LZ4HC_CLEVEL_MAX, cLevel);
assert(cLevel >= 0);
assert(cLevel <= LZ4HC_CLEVEL_MAX);
{ cParams_t const cParam = clTable[cLevel];
if (cParam.strat == lz4hc)
return LZ4HC_compress_hashChain(ctx,
src, dst, srcSizePtr, dstCapacity,
cParam.nbSearches, limit, dict);
assert(cParam.strat == lz4opt);
return LZ4HC_compress_optimal(ctx,
src, dst, srcSizePtr, dstCapacity,
cParam.nbSearches, cParam.targetLength, limit,
cLevel == LZ4HC_CLEVEL_MAX, dict); /* ultra mode */
}
}
static void LZ4HC_setExternalDict(LZ4HC_CCtx_internal* ctxPtr, const BYTE* newBlock);
static int LZ4HC_compress_generic_noDictCtx (
LZ4HC_CCtx_internal* const ctx,
const char* const src,
char* const dst,
int* const srcSizePtr,
int const dstCapacity,
int cLevel,
limitedOutput_directive limit
)
{
assert(ctx->dictCtx == NULL);
return LZ4HC_compress_generic_internal(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit, noDictCtx);
}
static int LZ4HC_compress_generic_dictCtx (
LZ4HC_CCtx_internal* const ctx,
const char* const src,
char* const dst,
int* const srcSizePtr,
int const dstCapacity,
int cLevel,
limitedOutput_directive limit
)
{
const size_t position = ctx->end - ctx->base - ctx->lowLimit;
assert(ctx->dictCtx != NULL);
if (position >= 64 KB) {
ctx->dictCtx = NULL;
return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit);
} else if (position == 0 && *srcSizePtr > 4 KB) {
memcpy(ctx, ctx->dictCtx, sizeof(LZ4HC_CCtx_internal));
LZ4HC_setExternalDict(ctx, (const BYTE *)src);
ctx->compressionLevel = cLevel;
return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit);
} else {
return LZ4HC_compress_generic_internal(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit, usingDictCtx);
}
}
static int LZ4HC_compress_generic (
LZ4HC_CCtx_internal* const ctx,
const char* const src,
char* const dst,
int* const srcSizePtr,
int const dstCapacity,
int cLevel,
limitedOutput_directive limit
)
{
if (ctx->dictCtx == NULL) {
return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit);
} else {
return LZ4HC_compress_generic_dictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit);
}
}
int LZ4_sizeofStateHC(void) { return sizeof(LZ4_streamHC_t); }
int LZ4_compress_HC_extStateHC_fastReset (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel)
{
LZ4HC_CCtx_internal* const ctx = &((LZ4_streamHC_t*)state)->internal_donotuse;
if (((size_t)(state)&(sizeof(void*)-1)) != 0) return 0; /* Error : state is not aligned for pointers (32 or 64 bits) */
LZ4_resetStreamHC_fast((LZ4_streamHC_t*)state, compressionLevel);
LZ4HC_init (ctx, (const BYTE*)src);
if (dstCapacity < LZ4_compressBound(srcSize))
return LZ4HC_compress_generic (ctx, src, dst, &srcSize, dstCapacity, compressionLevel, limitedOutput);
else
return LZ4HC_compress_generic (ctx, src, dst, &srcSize, dstCapacity, compressionLevel, noLimit);
}
int LZ4_compress_HC_extStateHC (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel)
{
if (((size_t)(state)&(sizeof(void*)-1)) != 0) return 0; /* Error : state is not aligned for pointers (32 or 64 bits) */
LZ4_resetStreamHC ((LZ4_streamHC_t*)state, compressionLevel);
return LZ4_compress_HC_extStateHC_fastReset(state, src, dst, srcSize, dstCapacity, compressionLevel);
}
int LZ4_compress_HC(const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel)
{
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
LZ4_streamHC_t* const statePtr = (LZ4_streamHC_t*)malloc(sizeof(LZ4_streamHC_t));
#else
LZ4_streamHC_t state;
LZ4_streamHC_t* const statePtr = &state;
#endif
int const cSize = LZ4_compress_HC_extStateHC(statePtr, src, dst, srcSize, dstCapacity, compressionLevel);
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
free(statePtr);
#endif
return cSize;
}
/* LZ4_compress_HC_destSize() :
* only compatible with regular HC parser */
int LZ4_compress_HC_destSize(void* LZ4HC_Data, const char* source, char* dest, int* sourceSizePtr, int targetDestSize, int cLevel)
{
LZ4HC_CCtx_internal* const ctx = &((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse;
LZ4_resetStreamHC((LZ4_streamHC_t*)LZ4HC_Data, cLevel);
LZ4HC_init(ctx, (const BYTE*) source);
return LZ4HC_compress_generic(ctx, source, dest, sourceSizePtr, targetDestSize, cLevel, limitedDestSize);
}
/**************************************
* Streaming Functions
**************************************/
/* allocation */
LZ4_streamHC_t* LZ4_createStreamHC(void) {
LZ4_streamHC_t* LZ4_streamHCPtr = (LZ4_streamHC_t*)malloc(sizeof(LZ4_streamHC_t));
if (LZ4_streamHCPtr != NULL) {
LZ4_streamHCPtr->internal_donotuse.end = (const BYTE *)(ptrdiff_t)-1;
LZ4_streamHCPtr->internal_donotuse.base = NULL;
LZ4_streamHCPtr->internal_donotuse.dictCtx = NULL;
}
return LZ4_streamHCPtr;
}
int LZ4_freeStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr) {
DEBUGLOG(4, "LZ4_freeStreamHC(%p)", LZ4_streamHCPtr);
if (!LZ4_streamHCPtr) return 0; /* support free on NULL */
free(LZ4_streamHCPtr);
return 0;
}
/* initialization */
void LZ4_resetStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel)
{
LZ4_STATIC_ASSERT(sizeof(LZ4HC_CCtx_internal) <= sizeof(size_t) * LZ4_STREAMHCSIZE_SIZET); /* if compilation fails here, LZ4_STREAMHCSIZE must be increased */
DEBUGLOG(4, "LZ4_resetStreamHC(%p, %d)", LZ4_streamHCPtr, compressionLevel);
LZ4_streamHCPtr->internal_donotuse.end = (const BYTE *)(ptrdiff_t)-1;
LZ4_streamHCPtr->internal_donotuse.base = NULL;
LZ4_streamHCPtr->internal_donotuse.dictCtx = NULL;
LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel);
}
void LZ4_resetStreamHC_fast (LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel)
{
DEBUGLOG(4, "LZ4_resetStreamHC_fast(%p, %d)", LZ4_streamHCPtr, compressionLevel);
LZ4_streamHCPtr->internal_donotuse.end -= (uptrval)LZ4_streamHCPtr->internal_donotuse.base;
LZ4_streamHCPtr->internal_donotuse.base = NULL;
LZ4_streamHCPtr->internal_donotuse.dictCtx = NULL;
LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel);
}
void LZ4_setCompressionLevel(LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel)
{
if (compressionLevel < 1) compressionLevel = LZ4HC_CLEVEL_DEFAULT;
if (compressionLevel > LZ4HC_CLEVEL_MAX) compressionLevel = LZ4HC_CLEVEL_MAX;
LZ4_streamHCPtr->internal_donotuse.compressionLevel = compressionLevel;
}
int LZ4_loadDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, const char* dictionary, int dictSize)
{
LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse;
DEBUGLOG(4, "LZ4_loadDictHC(%p, %p, %d)", LZ4_streamHCPtr, dictionary, dictSize);
if (dictSize > 64 KB) {
dictionary += dictSize - 64 KB;
dictSize = 64 KB;
}
LZ4_resetStreamHC(LZ4_streamHCPtr, LZ4_streamHCPtr->internal_donotuse.compressionLevel);
LZ4HC_init (ctxPtr, (const BYTE*)dictionary);
ctxPtr->end = (const BYTE*)dictionary + dictSize;
if (dictSize >= 4) LZ4HC_Insert (ctxPtr, ctxPtr->end-3);
return dictSize;
}
void LZ4_attach_HC_dictionary(LZ4_streamHC_t *working_stream, const LZ4_streamHC_t *dictionary_stream) {
working_stream->internal_donotuse.dictCtx = dictionary_stream != NULL ? &(dictionary_stream->internal_donotuse) : NULL;
}
/* compression */
static void LZ4HC_setExternalDict(LZ4HC_CCtx_internal* ctxPtr, const BYTE* newBlock)
{
DEBUGLOG(4, "LZ4HC_setExternalDict(%p, %p)", ctxPtr, newBlock);
if (ctxPtr->end - ctxPtr->base - ctxPtr->nextToUpdate >= 4) LZ4HC_Insert (ctxPtr, ctxPtr->end-3); /* Referencing remaining dictionary content */
/* Only one memory segment for extDict, so any previous extDict is lost at this stage */
ctxPtr->lowLimit = ctxPtr->dictLimit;
ctxPtr->dictLimit = (U32)(ctxPtr->end - ctxPtr->base);
ctxPtr->dictBase = ctxPtr->base;
ctxPtr->base = newBlock - ctxPtr->dictLimit;
ctxPtr->end = newBlock;
ctxPtr->nextToUpdate = ctxPtr->dictLimit; /* match referencing will resume from there */
}
static int LZ4_compressHC_continue_generic (LZ4_streamHC_t* LZ4_streamHCPtr,
const char* src, char* dst,
int* srcSizePtr, int dstCapacity,
limitedOutput_directive limit)
{
LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse;
DEBUGLOG(4, "LZ4_compressHC_continue_generic(%p, %p, %d)", LZ4_streamHCPtr, src, *srcSizePtr);
/* auto-init if forgotten */
if (ctxPtr->base == NULL) LZ4HC_init (ctxPtr, (const BYTE*) src);
/* Check overflow */
if ((size_t)(ctxPtr->end - ctxPtr->base) > 2 GB) {
size_t dictSize = (size_t)(ctxPtr->end - ctxPtr->base) - ctxPtr->dictLimit;
if (dictSize > 64 KB) dictSize = 64 KB;
LZ4_loadDictHC(LZ4_streamHCPtr, (const char*)(ctxPtr->end) - dictSize, (int)dictSize);
}
/* Check if blocks follow each other */
if ((const BYTE*)src != ctxPtr->end) LZ4HC_setExternalDict(ctxPtr, (const BYTE*)src);
/* Check overlapping input/dictionary space */
{ const BYTE* sourceEnd = (const BYTE*) src + *srcSizePtr;
const BYTE* const dictBegin = ctxPtr->dictBase + ctxPtr->lowLimit;
const BYTE* const dictEnd = ctxPtr->dictBase + ctxPtr->dictLimit;
if ((sourceEnd > dictBegin) && ((const BYTE*)src < dictEnd)) {
if (sourceEnd > dictEnd) sourceEnd = dictEnd;
ctxPtr->lowLimit = (U32)(sourceEnd - ctxPtr->dictBase);
if (ctxPtr->dictLimit - ctxPtr->lowLimit < 4) ctxPtr->lowLimit = ctxPtr->dictLimit;
}
}
return LZ4HC_compress_generic (ctxPtr, src, dst, srcSizePtr, dstCapacity, ctxPtr->compressionLevel, limit);
}
int LZ4_compress_HC_continue (LZ4_streamHC_t* LZ4_streamHCPtr, const char* src, char* dst, int srcSize, int dstCapacity)
{
if (dstCapacity < LZ4_compressBound(srcSize))
return LZ4_compressHC_continue_generic (LZ4_streamHCPtr, src, dst, &srcSize, dstCapacity, limitedOutput);
else
return LZ4_compressHC_continue_generic (LZ4_streamHCPtr, src, dst, &srcSize, dstCapacity, noLimit);
}
int LZ4_compress_HC_continue_destSize (LZ4_streamHC_t* LZ4_streamHCPtr, const char* src, char* dst, int* srcSizePtr, int targetDestSize)
{
return LZ4_compressHC_continue_generic(LZ4_streamHCPtr, src, dst, srcSizePtr, targetDestSize, limitedDestSize);
}
/* dictionary saving */
int LZ4_saveDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, char* safeBuffer, int dictSize)
{
LZ4HC_CCtx_internal* const streamPtr = &LZ4_streamHCPtr->internal_donotuse;
int const prefixSize = (int)(streamPtr->end - (streamPtr->base + streamPtr->dictLimit));
DEBUGLOG(4, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize);
if (dictSize > 64 KB) dictSize = 64 KB;
if (dictSize < 4) dictSize = 0;
if (dictSize > prefixSize) dictSize = prefixSize;
memmove(safeBuffer, streamPtr->end - dictSize, dictSize);
{ U32 const endIndex = (U32)(streamPtr->end - streamPtr->base);
streamPtr->end = (const BYTE*)safeBuffer + dictSize;
streamPtr->base = streamPtr->end - endIndex;
streamPtr->dictLimit = endIndex - dictSize;
streamPtr->lowLimit = endIndex - dictSize;
if (streamPtr->nextToUpdate < streamPtr->dictLimit) streamPtr->nextToUpdate = streamPtr->dictLimit;
}
return dictSize;
}
/***********************************
* Deprecated Functions
***********************************/
/* These functions currently generate deprecation warnings */
/* Deprecated compression functions */
int LZ4_compressHC(const char* src, char* dst, int srcSize) { return LZ4_compress_HC (src, dst, srcSize, LZ4_compressBound(srcSize), 0); }
int LZ4_compressHC_limitedOutput(const char* src, char* dst, int srcSize, int maxDstSize) { return LZ4_compress_HC(src, dst, srcSize, maxDstSize, 0); }
int LZ4_compressHC2(const char* src, char* dst, int srcSize, int cLevel) { return LZ4_compress_HC (src, dst, srcSize, LZ4_compressBound(srcSize), cLevel); }
int LZ4_compressHC2_limitedOutput(const char* src, char* dst, int srcSize, int maxDstSize, int cLevel) { return LZ4_compress_HC(src, dst, srcSize, maxDstSize, cLevel); }
int LZ4_compressHC_withStateHC (void* state, const char* src, char* dst, int srcSize) { return LZ4_compress_HC_extStateHC (state, src, dst, srcSize, LZ4_compressBound(srcSize), 0); }
int LZ4_compressHC_limitedOutput_withStateHC (void* state, const char* src, char* dst, int srcSize, int maxDstSize) { return LZ4_compress_HC_extStateHC (state, src, dst, srcSize, maxDstSize, 0); }
int LZ4_compressHC2_withStateHC (void* state, const char* src, char* dst, int srcSize, int cLevel) { return LZ4_compress_HC_extStateHC(state, src, dst, srcSize, LZ4_compressBound(srcSize), cLevel); }
int LZ4_compressHC2_limitedOutput_withStateHC (void* state, const char* src, char* dst, int srcSize, int maxDstSize, int cLevel) { return LZ4_compress_HC_extStateHC(state, src, dst, srcSize, maxDstSize, cLevel); }
int LZ4_compressHC_continue (LZ4_streamHC_t* ctx, const char* src, char* dst, int srcSize) { return LZ4_compress_HC_continue (ctx, src, dst, srcSize, LZ4_compressBound(srcSize)); }
int LZ4_compressHC_limitedOutput_continue (LZ4_streamHC_t* ctx, const char* src, char* dst, int srcSize, int maxDstSize) { return LZ4_compress_HC_continue (ctx, src, dst, srcSize, maxDstSize); }
/* Deprecated streaming functions */
int LZ4_sizeofStreamStateHC(void) { return LZ4_STREAMHCSIZE; }
int LZ4_resetStreamStateHC(void* state, char* inputBuffer)
{
LZ4HC_CCtx_internal *ctx = &((LZ4_streamHC_t*)state)->internal_donotuse;
if ((((size_t)state) & (sizeof(void*)-1)) != 0) return 1; /* Error : pointer is not aligned for pointer (32 or 64 bits) */
LZ4_resetStreamHC((LZ4_streamHC_t*)state, ((LZ4_streamHC_t*)state)->internal_donotuse.compressionLevel);
LZ4HC_init(ctx, (const BYTE*)inputBuffer);
return 0;
}
void* LZ4_createHC (const char* inputBuffer)
{
LZ4_streamHC_t* hc4 = (LZ4_streamHC_t*)ALLOC(sizeof(LZ4_streamHC_t));
if (hc4 == NULL) return NULL; /* not enough memory */
LZ4_resetStreamHC(hc4, 0 /* compressionLevel */);
LZ4HC_init (&hc4->internal_donotuse, (const BYTE*)inputBuffer);
return hc4;
}
int LZ4_freeHC (void* LZ4HC_Data) {
if (!LZ4HC_Data) return 0; /* support free on NULL */
FREEMEM(LZ4HC_Data);
return 0;
}
int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* src, char* dst, int srcSize, int cLevel)
{
return LZ4HC_compress_generic (&((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse, src, dst, &srcSize, 0, cLevel, noLimit);
}
int LZ4_compressHC2_limitedOutput_continue (void* LZ4HC_Data, const char* src, char* dst, int srcSize, int dstCapacity, int cLevel)
{
return LZ4HC_compress_generic (&((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse, src, dst, &srcSize, dstCapacity, cLevel, limitedOutput);
}
char* LZ4_slideInputBufferHC(void* LZ4HC_Data)
{
LZ4_streamHC_t *ctx = (LZ4_streamHC_t*)LZ4HC_Data;
const BYTE *bufferStart = ctx->internal_donotuse.base + ctx->internal_donotuse.lowLimit;
LZ4_resetStreamHC_fast(ctx, ctx->internal_donotuse.compressionLevel);
/* avoid const char * -> char * conversion warning :( */
return (char *)(uptrval)bufferStart;
}
/* ================================================
* LZ4 Optimal parser (levels 10-12)
* ===============================================*/
typedef struct {
int price;
int off;
int mlen;
int litlen;
} LZ4HC_optimal_t;
/* price in bytes */
LZ4_FORCE_INLINE int LZ4HC_literalsPrice(int const litlen)
{
int price = litlen;
if (litlen >= (int)RUN_MASK)
price += 1 + (litlen-RUN_MASK)/255;
return price;
}
/* requires mlen >= MINMATCH */
LZ4_FORCE_INLINE int LZ4HC_sequencePrice(int litlen, int mlen)
{
int price = 1 + 2 ; /* token + 16-bit offset */
price += LZ4HC_literalsPrice(litlen);
if (mlen >= (int)(ML_MASK+MINMATCH))
price += 1 + (mlen-(ML_MASK+MINMATCH))/255;
return price;
}
typedef struct {
int off;
int len;
} LZ4HC_match_t;
LZ4_FORCE_INLINE LZ4HC_match_t
LZ4HC_FindLongerMatch(LZ4HC_CCtx_internal* const ctx,
const BYTE* ip, const BYTE* const iHighLimit,
int minLen, int nbSearches,
const dictCtx_directive dict)
{
LZ4HC_match_t match = { 0 , 0 };
const BYTE* matchPtr = NULL;
/* note : LZ4HC_InsertAndGetWiderMatch() is able to modify the starting position of a match (*startpos),
* but this won't be the case here, as we define iLowLimit==ip,
* so LZ4HC_InsertAndGetWiderMatch() won't be allowed to search past ip */
int const matchLength = LZ4HC_InsertAndGetWiderMatch(ctx,
ip, ip, iHighLimit, minLen, &matchPtr, &ip,
nbSearches, 1 /* patternAnalysis */, dict);
if (matchLength <= minLen) return match;
match.len = matchLength;
match.off = (int)(ip-matchPtr);
return match;
}
static int LZ4HC_compress_optimal (
LZ4HC_CCtx_internal* ctx,
const char* const source,
char* dst,
int* srcSizePtr,
int dstCapacity,
int const nbSearches,
size_t sufficient_len,
const limitedOutput_directive limit,
int const fullUpdate,
const dictCtx_directive dict
)
{
#define TRAILING_LITERALS 3
LZ4HC_optimal_t opt[LZ4_OPT_NUM + TRAILING_LITERALS]; /* ~64 KB, which is a bit large for stack... */
const BYTE* ip = (const BYTE*) source;
const BYTE* anchor = ip;
const BYTE* const iend = ip + *srcSizePtr;
const BYTE* const mflimit = iend - MFLIMIT;
const BYTE* const matchlimit = iend - LASTLITERALS;
BYTE* op = (BYTE*) dst;
BYTE* opSaved = (BYTE*) dst;
BYTE* oend = op + dstCapacity;
/* init */
DEBUGLOG(5, "LZ4HC_compress_optimal");
*srcSizePtr = 0;
if (limit == limitedDestSize) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */
if (sufficient_len >= LZ4_OPT_NUM) sufficient_len = LZ4_OPT_NUM-1;
/* Main Loop */
assert(ip - anchor < LZ4_MAX_INPUT_SIZE);
while (ip <= mflimit) {
int const llen = (int)(ip - anchor);
int best_mlen, best_off;
int cur, last_match_pos = 0;
LZ4HC_match_t const firstMatch = LZ4HC_FindLongerMatch(ctx, ip, matchlimit, MINMATCH-1, nbSearches, dict);
if (firstMatch.len==0) { ip++; continue; }
if ((size_t)firstMatch.len > sufficient_len) {
/* good enough solution : immediate encoding */
int const firstML = firstMatch.len;
const BYTE* const matchPos = ip - firstMatch.off;
opSaved = op;
if ( LZ4HC_encodeSequence(&ip, &op, &anchor, firstML, matchPos, limit, oend) ) /* updates ip, op and anchor */
goto _dest_overflow;
continue;
}
/* set prices for first positions (literals) */
{ int rPos;
for (rPos = 0 ; rPos < MINMATCH ; rPos++) {
int const cost = LZ4HC_literalsPrice(llen + rPos);
opt[rPos].mlen = 1;
opt[rPos].off = 0;
opt[rPos].litlen = llen + rPos;
opt[rPos].price = cost;
DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i) -- initial setup",
rPos, cost, opt[rPos].litlen);
} }
/* set prices using initial match */
{ int mlen = MINMATCH;
int const matchML = firstMatch.len; /* necessarily < sufficient_len < LZ4_OPT_NUM */
int const offset = firstMatch.off;
assert(matchML < LZ4_OPT_NUM);
for ( ; mlen <= matchML ; mlen++) {
int const cost = LZ4HC_sequencePrice(llen, mlen);
opt[mlen].mlen = mlen;
opt[mlen].off = offset;
opt[mlen].litlen = llen;
opt[mlen].price = cost;
DEBUGLOG(7, "rPos:%3i => price:%3i (matchlen=%i) -- initial setup",
mlen, cost, mlen);
} }
last_match_pos = firstMatch.len;
{ int addLit;
for (addLit = 1; addLit <= TRAILING_LITERALS; addLit ++) {
opt[last_match_pos+addLit].mlen = 1; /* literal */
opt[last_match_pos+addLit].off = 0;
opt[last_match_pos+addLit].litlen = addLit;
opt[last_match_pos+addLit].price = opt[last_match_pos].price + LZ4HC_literalsPrice(addLit);
DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i) -- initial setup",
last_match_pos+addLit, opt[last_match_pos+addLit].price, addLit);
} }
/* check further positions */
for (cur = 1; cur < last_match_pos; cur++) {
const BYTE* const curPtr = ip + cur;
LZ4HC_match_t newMatch;
if (curPtr > mflimit) break;
DEBUGLOG(7, "rPos:%u[%u] vs [%u]%u",
cur, opt[cur].price, opt[cur+1].price, cur+1);
if (fullUpdate) {
/* not useful to search here if next position has same (or lower) cost */
if ( (opt[cur+1].price <= opt[cur].price)
/* in some cases, next position has same cost, but cost rises sharply after, so a small match would still be beneficial */
&& (opt[cur+MINMATCH].price < opt[cur].price + 3/*min seq price*/) )
continue;
} else {
/* not useful to search here if next position has same (or lower) cost */
if (opt[cur+1].price <= opt[cur].price) continue;
}
DEBUGLOG(7, "search at rPos:%u", cur);
if (fullUpdate)
newMatch = LZ4HC_FindLongerMatch(ctx, curPtr, matchlimit, MINMATCH-1, nbSearches, dict);
else
/* only test matches of minimum length; slightly faster, but misses a few bytes */
newMatch = LZ4HC_FindLongerMatch(ctx, curPtr, matchlimit, last_match_pos - cur, nbSearches, dict);
if (!newMatch.len) continue;
if ( ((size_t)newMatch.len > sufficient_len)
|| (newMatch.len + cur >= LZ4_OPT_NUM) ) {
/* immediate encoding */
best_mlen = newMatch.len;
best_off = newMatch.off;
last_match_pos = cur + 1;
goto encode;
}
/* before match : set price with literals at beginning */
{ int const baseLitlen = opt[cur].litlen;
int litlen;
for (litlen = 1; litlen < MINMATCH; litlen++) {
int const price = opt[cur].price - LZ4HC_literalsPrice(baseLitlen) + LZ4HC_literalsPrice(baseLitlen+litlen);
int const pos = cur + litlen;
if (price < opt[pos].price) {
opt[pos].mlen = 1; /* literal */
opt[pos].off = 0;
opt[pos].litlen = baseLitlen+litlen;
opt[pos].price = price;
DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i)",
pos, price, opt[pos].litlen);
} } }
/* set prices using match at position = cur */
{ int const matchML = newMatch.len;
int ml = MINMATCH;
assert(cur + newMatch.len < LZ4_OPT_NUM);
for ( ; ml <= matchML ; ml++) {
int const pos = cur + ml;
int const offset = newMatch.off;
int price;
int ll;
DEBUGLOG(7, "testing price rPos %i (last_match_pos=%i)",
pos, last_match_pos);
if (opt[cur].mlen == 1) {
ll = opt[cur].litlen;
price = ((cur > ll) ? opt[cur - ll].price : 0)
+ LZ4HC_sequencePrice(ll, ml);
} else {
ll = 0;
price = opt[cur].price + LZ4HC_sequencePrice(0, ml);
}
if (pos > last_match_pos+TRAILING_LITERALS || price <= opt[pos].price) {
DEBUGLOG(7, "rPos:%3i => price:%3i (matchlen=%i)",
pos, price, ml);
assert(pos < LZ4_OPT_NUM);
if ( (ml == matchML) /* last pos of last match */
&& (last_match_pos < pos) )
last_match_pos = pos;
opt[pos].mlen = ml;
opt[pos].off = offset;
opt[pos].litlen = ll;
opt[pos].price = price;
} } }
/* complete following positions with literals */
{ int addLit;
for (addLit = 1; addLit <= TRAILING_LITERALS; addLit ++) {
opt[last_match_pos+addLit].mlen = 1; /* literal */
opt[last_match_pos+addLit].off = 0;
opt[last_match_pos+addLit].litlen = addLit;
opt[last_match_pos+addLit].price = opt[last_match_pos].price + LZ4HC_literalsPrice(addLit);
DEBUGLOG(7, "rPos:%3i => price:%3i (litlen=%i)", last_match_pos+addLit, opt[last_match_pos+addLit].price, addLit);
} }
} /* for (cur = 1; cur <= last_match_pos; cur++) */
best_mlen = opt[last_match_pos].mlen;
best_off = opt[last_match_pos].off;
cur = last_match_pos - best_mlen;
encode: /* cur, last_match_pos, best_mlen, best_off must be set */
assert(cur < LZ4_OPT_NUM);
assert(last_match_pos >= 1); /* == 1 when only one candidate */
DEBUGLOG(6, "reverse traversal, looking for shortest path (last_match_pos=%i)", last_match_pos);
{ int candidate_pos = cur;
int selected_matchLength = best_mlen;
int selected_offset = best_off;
while (1) { /* from end to beginning */
int const next_matchLength = opt[candidate_pos].mlen; /* can be 1, means literal */
int const next_offset = opt[candidate_pos].off;
DEBUGLOG(7, "pos %i: sequence length %i", candidate_pos, selected_matchLength);
opt[candidate_pos].mlen = selected_matchLength;
opt[candidate_pos].off = selected_offset;
selected_matchLength = next_matchLength;
selected_offset = next_offset;
if (next_matchLength > candidate_pos) break; /* last match elected, first match to encode */
assert(next_matchLength > 0); /* can be 1, means literal */
candidate_pos -= next_matchLength;
} }
/* encode all recorded sequences in order */
{ int rPos = 0; /* relative position (to ip) */
while (rPos < last_match_pos) {
int const ml = opt[rPos].mlen;
int const offset = opt[rPos].off;
if (ml == 1) { ip++; rPos++; continue; } /* literal; note: can end up with several literals, in which case, skip them */
rPos += ml;
assert(ml >= MINMATCH);
assert((offset >= 1) && (offset <= MAX_DISTANCE));
opSaved = op;
if ( LZ4HC_encodeSequence(&ip, &op, &anchor, ml, ip - offset, limit, oend) ) /* updates ip, op and anchor */
goto _dest_overflow;
} }
} /* while (ip <= mflimit) */
_last_literals:
/* Encode Last Literals */
{ size_t lastRunSize = (size_t)(iend - anchor); /* literals */
size_t litLength = (lastRunSize + 255 - RUN_MASK) / 255;
size_t const totalSize = 1 + litLength + lastRunSize;
if (limit == limitedDestSize) oend += LASTLITERALS; /* restore correct value */
if (limit && (op + totalSize > oend)) {
if (limit == limitedOutput) return 0; /* Check output limit */
/* adapt lastRunSize to fill 'dst' */
lastRunSize = (size_t)(oend - op) - 1;
litLength = (lastRunSize + 255 - RUN_MASK) / 255;
lastRunSize -= litLength;
}
ip = anchor + lastRunSize;
if (lastRunSize >= RUN_MASK) {
size_t accumulator = lastRunSize - RUN_MASK;
*op++ = (RUN_MASK << ML_BITS);
for(; accumulator >= 255 ; accumulator -= 255) *op++ = 255;
*op++ = (BYTE) accumulator;
} else {
*op++ = (BYTE)(lastRunSize << ML_BITS);
}
memcpy(op, anchor, lastRunSize);
op += lastRunSize;
}
/* End */
*srcSizePtr = (int) (((const char*)ip) - source);
return (int) ((char*)op-dst);
_dest_overflow:
if (limit == limitedDestSize) {
op = opSaved; /* restore correct out pointer */
goto _last_literals;
}
return 0;
}
|