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
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
|
/*
LZ4 HC - High Compression Mode of LZ4
Copyright (C) 2011-2020, 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 stateless HC compression functions like `LZ4_compress_HC()`
* allocate memory for their workspace:
* in stack (0:fastest), or in heap (1:default, requires malloc()).
* Since workspace 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"
#include <limits.h>
/*=== Shared lz4.c code ===*/
#ifndef LZ4_SRC_INCLUDED
# 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 */
#endif
/*=== Enums ===*/
typedef enum { noDictCtx, usingDictCtxHc } dictCtx_directive;
/*=== 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) )
/*=== Levels definition ===*/
typedef enum { lz4mid, lz4hc, lz4opt } lz4hc_strat_e;
typedef struct {
lz4hc_strat_e strat;
int nbSearches;
U32 targetLength;
} cParams_t;
static const cParams_t k_clTable[LZ4HC_CLEVEL_MAX+1] = {
{ lz4mid, 2, 16 }, /* 0, unused */
{ lz4mid, 2, 16 }, /* 1, unused */
{ lz4mid, 2, 16 }, /* 2 */
{ 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,16384,LZ4_OPT_NUM }, /* 12==LZ4HC_CLEVEL_MAX */
};
static cParams_t LZ4HC_getCLevelParams(int cLevel)
{
/* note : clevel convention is a bit different from lz4frame,
* possibly something worth revisiting for consistency */
if (cLevel < 1)
cLevel = LZ4HC_CLEVEL_DEFAULT;
cLevel = MIN(LZ4HC_CLEVEL_MAX, cLevel);
return k_clTable[cLevel];
}
/*=== Hashing ===*/
#define LZ4HC_HASHSIZE 4
#define HASH_FUNCTION(i) (((i) * 2654435761U) >> ((MINMATCH*8)-LZ4HC_HASH_LOG))
static U32 LZ4HC_hashPtr(const void* ptr) { return HASH_FUNCTION(LZ4_read32(ptr)); }
#if defined(LZ4_FORCE_MEMORY_ACCESS) && (LZ4_FORCE_MEMORY_ACCESS==2)
/* lie to the compiler about data alignment; use with caution */
static U64 LZ4_read64(const void* memPtr) { return *(const U64*) memPtr; }
#elif defined(LZ4_FORCE_MEMORY_ACCESS) && (LZ4_FORCE_MEMORY_ACCESS==1)
/* __pack instructions are safer, but compiler specific */
LZ4_PACK(typedef struct { U64 u64; }) LZ4_unalign64;
static U64 LZ4_read64(const void* ptr) { return ((const LZ4_unalign64*)ptr)->u64; }
#else /* safe and portable access using memcpy() */
static U64 LZ4_read64(const void* memPtr)
{
U64 val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val;
}
#endif /* LZ4_FORCE_MEMORY_ACCESS */
#define LZ4MID_HASHSIZE 8
#define LZ4MID_HASHLOG (LZ4HC_HASH_LOG-1)
#define LZ4MID_HASHTABLESIZE (1 << LZ4MID_HASHLOG)
static U32 LZ4MID_hash4(U32 v) { return (v * 2654435761U) >> (32-LZ4MID_HASHLOG); }
static U32 LZ4MID_hash4Ptr(const void* ptr) { return LZ4MID_hash4(LZ4_read32(ptr)); }
/* note: hash7 hashes the lower 56-bits.
* It presumes input was read using little endian.*/
static U32 LZ4MID_hash7(U64 v) { return (U32)(((v << (64-56)) * 58295818150454627ULL) >> (64-LZ4MID_HASHLOG)) ; }
static U64 LZ4_readLE64(const void* memPtr);
static U32 LZ4MID_hash8Ptr(const void* ptr) { return LZ4MID_hash7(LZ4_readLE64(ptr)); }
static U64 LZ4_readLE64(const void* memPtr)
{
if (LZ4_isLittleEndian()) {
return LZ4_read64(memPtr);
} else {
const BYTE* p = (const BYTE*)memPtr;
/* note: relies on the compiler to simplify this expression */
return (U64)p[0] | ((U64)p[1]<<8) | ((U64)p[2]<<16) | ((U64)p[3]<<24)
| ((U64)p[4]<<32) | ((U64)p[5]<<40) | ((U64)p[6]<<48) | ((U64)p[7]<<56);
}
}
/*=== Count match length ===*/
LZ4_FORCE_INLINE
unsigned LZ4HC_NbCommonBytes32(U32 val)
{
assert(val != 0);
if (LZ4_isLittleEndian()) {
# if defined(_MSC_VER) && (_MSC_VER >= 1400) && !defined(LZ4_FORCE_SW_BITCOUNT)
unsigned long r;
_BitScanReverse(&r, val);
return (unsigned)((31 - r) >> 3);
# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \
((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \
!defined(LZ4_FORCE_SW_BITCOUNT)
return (unsigned)__builtin_clz(val) >> 3;
# else
val >>= 8;
val = ((((val + 0x00FFFF00) | 0x00FFFFFF) + val) |
(val + 0x00FF0000)) >> 24;
return (unsigned)val ^ 3;
# endif
} else {
# if defined(_MSC_VER) && (_MSC_VER >= 1400) && !defined(LZ4_FORCE_SW_BITCOUNT)
unsigned long r;
_BitScanForward(&r, val);
return (unsigned)(r >> 3);
# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \
((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \
!defined(LZ4_FORCE_SW_BITCOUNT)
return (unsigned)__builtin_ctz(val) >> 3;
# else
const U32 m = 0x01010101;
return (unsigned)((((val - 1) ^ val) & (m - 1)) * m) >> 24;
# endif
}
}
/** 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(min <= 0);
assert(ip >= iMin); assert((size_t)(ip-iMin) < (1U<<31));
assert(match >= mMin); assert((size_t)(match - mMin) < (1U<<31));
while ((back - min) > 3) {
U32 const v = LZ4_read32(ip + back - 4) ^ LZ4_read32(match + back - 4);
if (v) {
return (back - (int)LZ4HC_NbCommonBytes32(v));
} else back -= 4; /* 4-byte step */
}
/* check remainder if any */
while ( (back > min)
&& (ip[back-1] == match[back-1]) )
back--;
return back;
}
/*=== Chain table updates ===*/
#define DELTANEXTU16(table, pos) table[(U16)(pos)] /* faster */
/* Make fields passed to, and updated by LZ4HC_encodeSequence explicit */
#define UPDATABLE(ip, op, anchor) &ip, &op, &anchor
/**************************************
* Init
**************************************/
static void LZ4HC_clearTables (LZ4HC_CCtx_internal* hc4)
{
MEM_INIT(hc4->hashTable, 0, sizeof(hc4->hashTable));
MEM_INIT(hc4->chainTable, 0xFF, sizeof(hc4->chainTable));
}
static void LZ4HC_init_internal (LZ4HC_CCtx_internal* hc4, const BYTE* start)
{
size_t const bufferSize = (size_t)(hc4->end - hc4->prefixStart);
size_t newStartingOffset = bufferSize + hc4->dictLimit;
DEBUGLOG(5, "LZ4HC_init_internal");
assert(newStartingOffset >= bufferSize); /* check overflow */
if (newStartingOffset > 1 GB) {
LZ4HC_clearTables(hc4);
newStartingOffset = 0;
}
newStartingOffset += 64 KB;
hc4->nextToUpdate = (U32)newStartingOffset;
hc4->prefixStart = start;
hc4->end = start;
hc4->dictStart = start;
hc4->dictLimit = (U32)newStartingOffset;
hc4->lowLimit = (U32)newStartingOffset;
}
/**************************************
* Encode
**************************************/
/* 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,
int offset,
limitedOutput_directive limit,
BYTE* oend)
{
#define ip (*_ip)
#define op (*_op)
#define anchor (*_anchor)
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:%4u, match:%4i, offset:%5i, cost:%4u + %5u",
pos,
(U32)(ip - anchor), matchLength, offset,
cost, totalCost);
totalCost += cost;
#endif
/* Encode Literal length */
length = (size_t)(ip - anchor);
LZ4_STATIC_ASSERT(notLimited == 0);
/* Check output limit */
if (limit && ((op + (length / 255) + length + (2 + 1 + LASTLITERALS)) > oend)) {
DEBUGLOG(6, "Not enough room to write %i literals (%i bytes remaining)",
(int)length, (int)(oend - op));
return 1;
}
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_wildCopy8(op, anchor, op + length);
op += length;
/* Encode Offset */
assert(offset <= LZ4_DISTANCE_MAX );
assert(offset > 0);
LZ4_writeLE16(op, (U16)(offset)); op += 2;
/* Encode MatchLength */
assert(matchLength >= MINMATCH);
length = (size_t)matchLength - MINMATCH;
if (limit && (op + (length / 255) + (1 + LASTLITERALS) > oend)) {
DEBUGLOG(6, "Not enough room to write match length");
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;
#undef ip
#undef op
#undef anchor
}
typedef struct {
int off;
int len;
int back; /* negative value */
} LZ4HC_match_t;
LZ4HC_match_t LZ4HC_searchExtDict(const BYTE* ip, U32 ipIndex,
const BYTE* const iLowLimit, const BYTE* const iHighLimit,
const LZ4HC_CCtx_internal* dictCtx, U32 gDictEndIndex,
int currentBestML, int nbAttempts)
{
size_t const lDictEndIndex = (size_t)(dictCtx->end - dictCtx->prefixStart) + dictCtx->dictLimit;
U32 lDictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)];
U32 matchIndex = lDictMatchIndex + gDictEndIndex - (U32)lDictEndIndex;
int offset = 0, sBack = 0;
assert(lDictEndIndex <= 1 GB);
if (lDictMatchIndex>0)
DEBUGLOG(7, "lDictEndIndex = %zu, lDictMatchIndex = %u", lDictEndIndex, lDictMatchIndex);
while (ipIndex - matchIndex <= LZ4_DISTANCE_MAX && nbAttempts--) {
const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + lDictMatchIndex;
if (LZ4_read32(matchPtr) == LZ4_read32(ip)) {
int mlt;
int back = 0;
const BYTE* vLimit = ip + (lDictEndIndex - lDictMatchIndex);
if (vLimit > iHighLimit) vLimit = iHighLimit;
mlt = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH;
back = (ip > iLowLimit) ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->prefixStart) : 0;
mlt -= back;
if (mlt > currentBestML) {
currentBestML = mlt;
offset = (int)(ipIndex - matchIndex);
sBack = back;
DEBUGLOG(7, "found match of length %i within extDictCtx", currentBestML);
} }
{ U32 const nextOffset = DELTANEXTU16(dictCtx->chainTable, lDictMatchIndex);
lDictMatchIndex -= nextOffset;
matchIndex -= nextOffset;
} }
{ LZ4HC_match_t md;
md.len = currentBestML;
md.off = offset;
md.back = sBack;
return md;
}
}
typedef LZ4HC_match_t (*LZ4MID_searchIntoDict_f)(const BYTE* ip, U32 ipIndex,
const BYTE* const iHighLimit,
const LZ4HC_CCtx_internal* dictCtx, U32 gDictEndIndex);
static LZ4HC_match_t LZ4MID_searchHCDict(const BYTE* ip, U32 ipIndex,
const BYTE* const iHighLimit,
const LZ4HC_CCtx_internal* dictCtx, U32 gDictEndIndex)
{
return LZ4HC_searchExtDict(ip,ipIndex,
ip, iHighLimit,
dictCtx, gDictEndIndex,
MINMATCH-1, 2);
}
static LZ4HC_match_t LZ4MID_searchExtDict(const BYTE* ip, U32 ipIndex,
const BYTE* const iHighLimit,
const LZ4HC_CCtx_internal* dictCtx, U32 gDictEndIndex)
{
size_t const lDictEndIndex = (size_t)(dictCtx->end - dictCtx->prefixStart) + dictCtx->dictLimit;
const U32* const hash4Table = dictCtx->hashTable;
const U32* const hash8Table = hash4Table + LZ4MID_HASHTABLESIZE;
DEBUGLOG(7, "LZ4MID_searchExtDict (ipIdx=%u)", ipIndex);
/* search long match first */
{ U32 l8DictMatchIndex = hash8Table[LZ4MID_hash8Ptr(ip)];
U32 m8Index = l8DictMatchIndex + gDictEndIndex - (U32)lDictEndIndex;
assert(lDictEndIndex <= 1 GB);
if (ipIndex - m8Index <= LZ4_DISTANCE_MAX) {
const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + l8DictMatchIndex;
const size_t safeLen = MIN(lDictEndIndex - l8DictMatchIndex, (size_t)(iHighLimit - ip));
int mlt = (int)LZ4_count(ip, matchPtr, ip + safeLen);
if (mlt >= MINMATCH) {
LZ4HC_match_t md;
DEBUGLOG(7, "Found long ExtDict match of len=%u", mlt);
md.len = mlt;
md.off = (int)(ipIndex - m8Index);
md.back = 0;
return md;
}
}
}
/* search for short match second */
{ U32 l4DictMatchIndex = hash4Table[LZ4MID_hash4Ptr(ip)];
U32 m4Index = l4DictMatchIndex + gDictEndIndex - (U32)lDictEndIndex;
if (ipIndex - m4Index <= LZ4_DISTANCE_MAX) {
const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + l4DictMatchIndex;
const size_t safeLen = MIN(lDictEndIndex - l4DictMatchIndex, (size_t)(iHighLimit - ip));
int mlt = (int)LZ4_count(ip, matchPtr, ip + safeLen);
if (mlt >= MINMATCH) {
LZ4HC_match_t md;
DEBUGLOG(7, "Found short ExtDict match of len=%u", mlt);
md.len = mlt;
md.off = (int)(ipIndex - m4Index);
md.back = 0;
return md;
}
}
}
/* nothing found */
{ LZ4HC_match_t const md = {0, 0, 0 };
return md;
}
}
/**************************************
* Mid Compression (level 2)
**************************************/
LZ4_FORCE_INLINE void
LZ4MID_addPosition(U32* hTable, U32 hValue, U32 index)
{
hTable[hValue] = index;
}
#define ADDPOS8(_p, _idx) LZ4MID_addPosition(hash8Table, LZ4MID_hash8Ptr(_p), _idx)
#define ADDPOS4(_p, _idx) LZ4MID_addPosition(hash4Table, LZ4MID_hash4Ptr(_p), _idx)
/* Fill hash tables with references into dictionary.
* The resulting table is only exploitable by LZ4MID (level 2) */
static void
LZ4MID_fillHTable (LZ4HC_CCtx_internal* cctx, const void* dict, size_t size)
{
U32* const hash4Table = cctx->hashTable;
U32* const hash8Table = hash4Table + LZ4MID_HASHTABLESIZE;
const BYTE* const prefixPtr = (const BYTE*)dict;
U32 const prefixIdx = cctx->dictLimit;
U32 const target = prefixIdx + (U32)size - LZ4MID_HASHSIZE;
U32 idx = cctx->nextToUpdate;
assert(dict == cctx->prefixStart);
DEBUGLOG(4, "LZ4MID_fillHTable (size:%zu)", size);
if (size <= LZ4MID_HASHSIZE)
return;
for (; idx < target; idx += 3) {
ADDPOS4(prefixPtr+idx-prefixIdx, idx);
ADDPOS8(prefixPtr+idx+1-prefixIdx, idx+1);
}
idx = (size > 32 KB + LZ4MID_HASHSIZE) ? target - 32 KB : cctx->nextToUpdate;
for (; idx < target; idx += 1) {
ADDPOS8(prefixPtr+idx-prefixIdx, idx);
}
cctx->nextToUpdate = target;
}
static LZ4MID_searchIntoDict_f select_searchDict_function(const LZ4HC_CCtx_internal* dictCtx)
{
if (dictCtx == NULL) return NULL;
if (LZ4HC_getCLevelParams(dictCtx->compressionLevel).strat == lz4mid)
return LZ4MID_searchExtDict;
return LZ4MID_searchHCDict;
}
static int LZ4MID_compress (
LZ4HC_CCtx_internal* const ctx,
const char* const src,
char* const dst,
int* srcSizePtr,
int const maxOutputSize,
const limitedOutput_directive limit,
const dictCtx_directive dict
)
{
U32* const hash4Table = ctx->hashTable;
U32* const hash8Table = hash4Table + LZ4MID_HASHTABLESIZE;
const BYTE* ip = (const BYTE*)src;
const BYTE* anchor = ip;
const BYTE* const iend = ip + *srcSizePtr;
const BYTE* const mflimit = iend - MFLIMIT;
const BYTE* const matchlimit = (iend - LASTLITERALS);
const BYTE* const ilimit = (iend - LZ4MID_HASHSIZE);
BYTE* op = (BYTE*)dst;
BYTE* oend = op + maxOutputSize;
const BYTE* const prefixPtr = ctx->prefixStart;
const U32 prefixIdx = ctx->dictLimit;
const U32 ilimitIdx = (U32)(ilimit - prefixPtr) + prefixIdx;
const BYTE* const dictStart = ctx->dictStart;
const U32 dictIdx = ctx->lowLimit;
const U32 gDictEndIndex = ctx->lowLimit;
const LZ4MID_searchIntoDict_f searchIntoDict = (dict == usingDictCtxHc) ? select_searchDict_function(ctx->dictCtx) : NULL;
unsigned matchLength;
unsigned matchDistance;
/* input sanitization */
DEBUGLOG(5, "LZ4MID_compress (%i bytes)", *srcSizePtr);
if (dict == usingDictCtxHc) DEBUGLOG(5, "usingDictCtxHc");
assert(*srcSizePtr >= 0);
if (*srcSizePtr) assert(src != NULL);
if (maxOutputSize) assert(dst != NULL);
if (*srcSizePtr < 0) return 0; /* invalid */
if (maxOutputSize < 0) return 0; /* invalid */
if (*srcSizePtr > LZ4_MAX_INPUT_SIZE) {
/* forbidden: no input is allowed to be that large */
return 0;
}
if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */
if (*srcSizePtr < LZ4_minLength)
goto _lz4mid_last_literals; /* Input too small, no compression (all literals) */
/* main loop */
while (ip <= mflimit) {
const U32 ipIndex = (U32)(ip - prefixPtr) + prefixIdx;
/* search long match */
{ U32 const h8 = LZ4MID_hash8Ptr(ip);
U32 const pos8 = hash8Table[h8];
assert(h8 < LZ4MID_HASHTABLESIZE);
assert(pos8 < ipIndex);
LZ4MID_addPosition(hash8Table, h8, ipIndex);
if (ipIndex - pos8 <= LZ4_DISTANCE_MAX) {
/* match candidate found */
if (pos8 >= prefixIdx) {
const BYTE* const matchPtr = prefixPtr + pos8 - prefixIdx;
assert(matchPtr < ip);
matchLength = LZ4_count(ip, matchPtr, matchlimit);
if (matchLength >= MINMATCH) {
DEBUGLOG(7, "found long match at pos %u (len=%u)", pos8, matchLength);
matchDistance = ipIndex - pos8;
goto _lz4mid_encode_sequence;
}
} else {
if (pos8 >= dictIdx) {
/* extDict match candidate */
const BYTE* const matchPtr = dictStart + (pos8 - dictIdx);
const size_t safeLen = MIN(prefixIdx - pos8, (size_t)(matchlimit - ip));
matchLength = LZ4_count(ip, matchPtr, ip + safeLen);
if (matchLength >= MINMATCH) {
DEBUGLOG(7, "found long match at ExtDict pos %u (len=%u)", pos8, matchLength);
matchDistance = ipIndex - pos8;
goto _lz4mid_encode_sequence;
}
}
}
} }
/* search short match */
{ U32 const h4 = LZ4MID_hash4Ptr(ip);
U32 const pos4 = hash4Table[h4];
assert(h4 < LZ4MID_HASHTABLESIZE);
assert(pos4 < ipIndex);
LZ4MID_addPosition(hash4Table, h4, ipIndex);
if (ipIndex - pos4 <= LZ4_DISTANCE_MAX) {
/* match candidate found */
if (pos4 >= prefixIdx) {
/* only search within prefix */
const BYTE* const matchPtr = prefixPtr + (pos4 - prefixIdx);
assert(matchPtr < ip);
assert(matchPtr >= prefixPtr);
matchLength = LZ4_count(ip, matchPtr, matchlimit);
if (matchLength >= MINMATCH) {
/* short match found, let's just check ip+1 for longer */
U32 const h8 = LZ4MID_hash8Ptr(ip+1);
U32 const pos8 = hash8Table[h8];
U32 const m2Distance = ipIndex + 1 - pos8;
matchDistance = ipIndex - pos4;
if ( m2Distance <= LZ4_DISTANCE_MAX
&& pos8 >= prefixIdx /* only search within prefix */
&& likely(ip < mflimit)
) {
const BYTE* const m2Ptr = prefixPtr + (pos8 - prefixIdx);
unsigned ml2 = LZ4_count(ip+1, m2Ptr, matchlimit);
if (ml2 > matchLength) {
LZ4MID_addPosition(hash8Table, h8, ipIndex+1);
ip++;
matchLength = ml2;
matchDistance = m2Distance;
} }
goto _lz4mid_encode_sequence;
}
} else {
if (pos4 >= dictIdx) {
/* extDict match candidate */
const BYTE* const matchPtr = dictStart + (pos4 - dictIdx);
const size_t safeLen = MIN(prefixIdx - pos4, (size_t)(matchlimit - ip));
matchLength = LZ4_count(ip, matchPtr, ip + safeLen);
if (matchLength >= MINMATCH) {
DEBUGLOG(7, "found match at ExtDict pos %u (len=%u)", pos4, matchLength);
matchDistance = ipIndex - pos4;
goto _lz4mid_encode_sequence;
}
}
}
} }
/* no match found in prefix */
if ( (dict == usingDictCtxHc)
&& (ipIndex - gDictEndIndex < LZ4_DISTANCE_MAX - 8) ) {
/* search a match into external dictionary */
LZ4HC_match_t dMatch = searchIntoDict(ip, ipIndex,
matchlimit,
ctx->dictCtx, gDictEndIndex);
if (dMatch.len >= MINMATCH) {
DEBUGLOG(7, "found Dictionary match (offset=%i)", dMatch.off);
assert(dMatch.back == 0);
matchLength = (unsigned)dMatch.len;
matchDistance = (unsigned)dMatch.off;
goto _lz4mid_encode_sequence;
}
}
/* no match found */
ip += 1 + ((ip-anchor) >> 9); /* skip faster over incompressible data */
continue;
_lz4mid_encode_sequence:
/* catch back */
while (((ip > anchor) & ((U32)(ip-prefixPtr) > matchDistance)) && (unlikely(ip[-1] == ip[-(int)matchDistance-1]))) {
ip--; matchLength++;
};
/* fill table with beginning of match */
ADDPOS8(ip+1, ipIndex+1);
ADDPOS8(ip+2, ipIndex+2);
ADDPOS4(ip+1, ipIndex+1);
/* encode */
{ BYTE* const saved_op = op;
/* LZ4HC_encodeSequence always updates @op; on success, it updates @ip and @anchor */
if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
(int)matchLength, (int)matchDistance,
limit, oend) ) {
op = saved_op; /* restore @op value before failed LZ4HC_encodeSequence */
goto _lz4mid_dest_overflow;
}
}
/* fill table with end of match */
{ U32 endMatchIdx = (U32)(ip-prefixPtr) + prefixIdx;
U32 pos_m2 = endMatchIdx - 2;
if (pos_m2 < ilimitIdx) {
if (likely(ip - prefixPtr > 5)) {
ADDPOS8(ip-5, endMatchIdx - 5);
}
ADDPOS8(ip-3, endMatchIdx - 3);
ADDPOS8(ip-2, endMatchIdx - 2);
ADDPOS4(ip-2, endMatchIdx - 2);
ADDPOS4(ip-1, endMatchIdx - 1);
}
}
}
_lz4mid_last_literals:
/* Encode Last Literals */
{ size_t lastRunSize = (size_t)(iend - anchor); /* literals */
size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255;
size_t const totalSize = 1 + llAdd + lastRunSize;
if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */
if (limit && (op + totalSize > oend)) {
if (limit == limitedOutput) return 0; /* not enough space in @dst */
/* adapt lastRunSize to fill 'dest' */
lastRunSize = (size_t)(oend - op) - 1 /*token*/;
llAdd = (lastRunSize + 256 - RUN_MASK) / 256;
lastRunSize -= llAdd;
}
DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize);
ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */
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);
}
assert(lastRunSize <= (size_t)(oend - op));
LZ4_memcpy(op, anchor, lastRunSize);
op += lastRunSize;
}
/* End */
DEBUGLOG(5, "compressed %i bytes into %i bytes", *srcSizePtr, (int)((char*)op - dst));
assert(ip >= (const BYTE*)src);
assert(ip <= iend);
*srcSizePtr = (int)(ip - (const BYTE*)src);
assert((char*)op >= dst);
assert(op <= oend);
assert((char*)op - dst < INT_MAX);
return (int)((char*)op - dst);
_lz4mid_dest_overflow:
if (limit == fillOutput) {
/* Assumption : @ip, @anchor, @optr and @matchLength must be set correctly */
size_t const ll = (size_t)(ip - anchor);
size_t const ll_addbytes = (ll + 240) / 255;
size_t const ll_totalCost = 1 + ll_addbytes + ll;
BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */
DEBUGLOG(6, "Last sequence is overflowing : %u literals, %u remaining space",
(unsigned)ll, (unsigned)(oend-op));
if (op + ll_totalCost <= maxLitPos) {
/* ll validated; now adjust match length */
size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost));
size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255);
assert(maxMlSize < INT_MAX);
if ((size_t)matchLength > maxMlSize) matchLength= (unsigned)maxMlSize;
if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + matchLength >= MFLIMIT) {
DEBUGLOG(6, "Let's encode a last sequence (ll=%u, ml=%u)", (unsigned)ll, matchLength);
LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
(int)matchLength, (int)matchDistance,
notLimited, oend);
} }
DEBUGLOG(6, "Let's finish with a run of literals (%u bytes left)", (unsigned)(oend-op));
goto _lz4mid_last_literals;
}
/* compression failed */
return 0;
}
/**************************************
* HC Compression - Search
**************************************/
/* 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 prefixPtr = hc4->prefixStart;
U32 const prefixIdx = hc4->dictLimit;
U32 const target = (U32)(ip - prefixPtr) + prefixIdx;
U32 idx = hc4->nextToUpdate;
assert(ip >= prefixPtr);
assert(target >= prefixIdx);
while (idx < target) {
U32 const h = LZ4HC_hashPtr(prefixPtr+idx-prefixIdx);
size_t delta = idx - hashTable[h];
if (delta>LZ4_DISTANCE_MAX) delta = LZ4_DISTANCE_MAX;
DELTANEXTU16(chainTable, idx) = (U16)delta;
hashTable[h] = idx;
idx++;
}
hc4->nextToUpdate = target;
}
#if defined(_MSC_VER)
# define LZ4HC_rotl32(x,r) _rotl(x,r)
#else
# define LZ4HC_rotl32(x,r) ((x << r) | (x >> (32 - r)))
#endif
static U32 LZ4HC_rotatePattern(size_t const rotate, U32 const pattern)
{
size_t const bitsToRotate = (rotate & (sizeof(pattern) - 1)) << 3;
if (bitsToRotate == 0) return pattern;
return LZ4HC_rotl32(pattern, (int)bitsToRotate);
}
/* 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) << (sizeof(pattern)*4)) : 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 endianness */
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 endianness */
while (likely(ip>iLow)) {
if (ip[-1] != *bytePtr) break;
ip--; bytePtr--;
} }
return (unsigned)(iStart - ip);
}
/* LZ4HC_protectDictEnd() :
* Checks if the match is in the last 3 bytes of the dictionary, so reading the
* 4 byte MINMATCH would overflow.
* @returns true if the match index is okay.
*/
static int LZ4HC_protectDictEnd(U32 const dictLimit, U32 const matchIndex)
{
return ((U32)((dictLimit - 1) - matchIndex) >= 3);
}
typedef enum { rep_untested, rep_not, rep_confirmed } repeat_state_e;
typedef enum { favorCompressionRatio=0, favorDecompressionSpeed } HCfavor_e;
LZ4_FORCE_INLINE LZ4HC_match_t
LZ4HC_InsertAndGetWiderMatch (
LZ4HC_CCtx_internal* const hc4,
const BYTE* const ip,
const BYTE* const iLowLimit, const BYTE* const iHighLimit,
int longest,
const int maxNbAttempts,
const int patternAnalysis, const int chainSwap,
const dictCtx_directive dict,
const HCfavor_e favorDecSpeed)
{
U16* const chainTable = hc4->chainTable;
U32* const hashTable = hc4->hashTable;
const LZ4HC_CCtx_internal* const dictCtx = hc4->dictCtx;
const BYTE* const prefixPtr = hc4->prefixStart;
const U32 prefixIdx = hc4->dictLimit;
const U32 ipIndex = (U32)(ip - prefixPtr) + prefixIdx;
const int withinStartDistance = (hc4->lowLimit + (LZ4_DISTANCE_MAX + 1) > ipIndex);
const U32 lowestMatchIndex = (withinStartDistance) ? hc4->lowLimit : ipIndex - LZ4_DISTANCE_MAX;
const BYTE* const dictStart = hc4->dictStart;
const U32 dictIdx = hc4->lowLimit;
const BYTE* const dictEnd = dictStart + prefixIdx - dictIdx;
int const lookBackLength = (int)(ip-iLowLimit);
int nbAttempts = maxNbAttempts;
U32 matchChainPos = 0;
U32 const pattern = LZ4_read32(ip);
U32 matchIndex;
repeat_state_e repeat = rep_untested;
size_t srcPatternLength = 0;
int offset = 0, sBack = 0;
DEBUGLOG(7, "LZ4HC_InsertAndGetWiderMatch");
/* First Match */
LZ4HC_Insert(hc4, ip); /* insert all prior positions up to ip (excluded) */
matchIndex = hashTable[LZ4HC_hashPtr(ip)];
DEBUGLOG(7, "First candidate match for pos %u found at index %u / %u (lowestMatchIndex)",
ipIndex, matchIndex, lowestMatchIndex);
while ((matchIndex>=lowestMatchIndex) && (nbAttempts>0)) {
int matchLength=0;
nbAttempts--;
assert(matchIndex < ipIndex);
if (favorDecSpeed && (ipIndex - matchIndex < 8)) {
/* do nothing:
* favorDecSpeed intentionally skips matches with offset < 8 */
} else if (matchIndex >= prefixIdx) { /* within current Prefix */
const BYTE* const matchPtr = prefixPtr + (matchIndex - prefixIdx);
assert(matchPtr < ip);
assert(longest >= 1);
if (LZ4_read16(iLowLimit + longest - 1) == LZ4_read16(matchPtr - lookBackLength + longest - 1)) {
if (LZ4_read32(matchPtr) == pattern) {
int const back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, prefixPtr) : 0;
matchLength = MINMATCH + (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, iHighLimit);
matchLength -= back;
if (matchLength > longest) {
longest = matchLength;
offset = (int)(ipIndex - matchIndex);
sBack = back;
DEBUGLOG(7, "Found match of len=%i within prefix, offset=%i, back=%i", longest, offset, -back);
} } }
} else { /* lowestMatchIndex <= matchIndex < dictLimit : within Ext Dict */
const BYTE* const matchPtr = dictStart + (matchIndex - dictIdx);
assert(matchIndex >= dictIdx);
if ( likely(matchIndex <= prefixIdx - 4)
&& (LZ4_read32(matchPtr) == pattern) ) {
int back = 0;
const BYTE* vLimit = ip + (prefixIdx - matchIndex);
if (vLimit > iHighLimit) vLimit = iHighLimit;
matchLength = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH;
if ((ip+matchLength == vLimit) && (vLimit < iHighLimit))
matchLength += LZ4_count(ip+matchLength, prefixPtr, iHighLimit);
back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictStart) : 0;
matchLength -= back;
if (matchLength > longest) {
longest = matchLength;
offset = (int)(ipIndex - matchIndex);
sBack = back;
DEBUGLOG(7, "Found match of len=%i within dict, offset=%i, back=%i", longest, offset, -back);
} } }
if (chainSwap && matchLength==longest) { /* better match => select a better chain */
assert(lookBackLength==0); /* search forward only */
if (matchIndex + (U32)longest <= ipIndex) {
int const kTrigger = 4;
U32 distanceToNextMatch = 1;
int const end = longest - MINMATCH + 1;
int step = 1;
int accel = 1 << kTrigger;
int pos;
for (pos = 0; pos < end; pos += step) {
U32 const candidateDist = DELTANEXTU16(chainTable, matchIndex + (U32)pos);
step = (accel++ >> kTrigger);
if (candidateDist > distanceToNextMatch) {
distanceToNextMatch = candidateDist;
matchChainPos = (U32)pos;
accel = 1 << kTrigger;
} }
if (distanceToNextMatch > 1) {
if (distanceToNextMatch > matchIndex) break; /* avoid overflow */
matchIndex -= distanceToNextMatch;
continue;
} } }
{ U32 const distNextMatch = DELTANEXTU16(chainTable, matchIndex);
if (patternAnalysis && distNextMatch==1 && matchChainPos==0) {
U32 const matchCandidateIdx = matchIndex-1;
/* may be a repeated pattern */
if (repeat == rep_untested) {
if ( ((pattern & 0xFFFF) == (pattern >> 16))
& ((pattern & 0xFF) == (pattern >> 24)) ) {
DEBUGLOG(7, "Repeat pattern detected, char %02X", pattern >> 24);
repeat = rep_confirmed;
srcPatternLength = LZ4HC_countPattern(ip+sizeof(pattern), iHighLimit, pattern) + sizeof(pattern);
} else {
repeat = rep_not;
} }
if ( (repeat == rep_confirmed) && (matchCandidateIdx >= lowestMatchIndex)
&& LZ4HC_protectDictEnd(prefixIdx, matchCandidateIdx) ) {
const int extDict = matchCandidateIdx < prefixIdx;
const BYTE* const matchPtr = extDict ? dictStart + (matchCandidateIdx - dictIdx) : prefixPtr + (matchCandidateIdx - prefixIdx);
if (LZ4_read32(matchPtr) == pattern) { /* good candidate */
const BYTE* const iLimit = extDict ? dictEnd : iHighLimit;
size_t forwardPatternLength = LZ4HC_countPattern(matchPtr+sizeof(pattern), iLimit, pattern) + sizeof(pattern);
if (extDict && matchPtr + forwardPatternLength == iLimit) {
U32 const rotatedPattern = LZ4HC_rotatePattern(forwardPatternLength, pattern);
forwardPatternLength += LZ4HC_countPattern(prefixPtr, iHighLimit, rotatedPattern);
}
{ const BYTE* const lowestMatchPtr = extDict ? dictStart : prefixPtr;
size_t backLength = LZ4HC_reverseCountPattern(matchPtr, lowestMatchPtr, pattern);
size_t currentSegmentLength;
if (!extDict
&& matchPtr - backLength == prefixPtr
&& dictIdx < prefixIdx) {
U32 const rotatedPattern = LZ4HC_rotatePattern((U32)(-(int)backLength), pattern);
backLength += LZ4HC_reverseCountPattern(dictEnd, dictStart, rotatedPattern);
}
/* Limit backLength not go further than lowestMatchIndex */
backLength = matchCandidateIdx - MAX(matchCandidateIdx - (U32)backLength, lowestMatchIndex);
assert(matchCandidateIdx - backLength >= lowestMatchIndex);
currentSegmentLength = backLength + forwardPatternLength;
/* Adjust to end of pattern if the source pattern fits, otherwise the beginning of the pattern */
if ( (currentSegmentLength >= srcPatternLength) /* current pattern segment large enough to contain full srcPatternLength */
&& (forwardPatternLength <= srcPatternLength) ) { /* haven't reached this position yet */
U32 const newMatchIndex = matchCandidateIdx + (U32)forwardPatternLength - (U32)srcPatternLength; /* best position, full pattern, might be followed by more match */
if (LZ4HC_protectDictEnd(prefixIdx, newMatchIndex))
matchIndex = newMatchIndex;
else {
/* Can only happen if started in the prefix */
assert(newMatchIndex >= prefixIdx - 3 && newMatchIndex < prefixIdx && !extDict);
matchIndex = prefixIdx;
}
} else {
U32 const newMatchIndex = matchCandidateIdx - (U32)backLength; /* farthest position in current segment, will find a match of length currentSegmentLength + maybe some back */
if (!LZ4HC_protectDictEnd(prefixIdx, newMatchIndex)) {
assert(newMatchIndex >= prefixIdx - 3 && newMatchIndex < prefixIdx && !extDict);
matchIndex = prefixIdx;
} else {
matchIndex = newMatchIndex;
if (lookBackLength==0) { /* no back possible */
size_t const maxML = MIN(currentSegmentLength, srcPatternLength);
if ((size_t)longest < maxML) {
assert(prefixPtr - prefixIdx + matchIndex != ip);
if ((size_t)(ip - prefixPtr) + prefixIdx - matchIndex > LZ4_DISTANCE_MAX) break;
assert(maxML < 2 GB);
longest = (int)maxML;
offset = (int)(ipIndex - matchIndex);
assert(sBack == 0);
DEBUGLOG(7, "Found repeat pattern match of len=%i, offset=%i", longest, offset);
}
{ U32 const distToNextPattern = DELTANEXTU16(chainTable, matchIndex);
if (distToNextPattern > matchIndex) break; /* avoid overflow */
matchIndex -= distToNextPattern;
} } } } }
continue;
} }
} } /* PA optimization */
/* follow current chain */
matchIndex -= DELTANEXTU16(chainTable, matchIndex + matchChainPos);
} /* while ((matchIndex>=lowestMatchIndex) && (nbAttempts)) */
if ( dict == usingDictCtxHc
&& nbAttempts > 0
&& withinStartDistance) {
size_t const dictEndOffset = (size_t)(dictCtx->end - dictCtx->prefixStart) + dictCtx->dictLimit;
U32 dictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)];
assert(dictEndOffset <= 1 GB);
matchIndex = dictMatchIndex + lowestMatchIndex - (U32)dictEndOffset;
if (dictMatchIndex>0) DEBUGLOG(7, "dictEndOffset = %zu, dictMatchIndex = %u => relative matchIndex = %i", dictEndOffset, dictMatchIndex, (int)dictMatchIndex - (int)dictEndOffset);
while (ipIndex - matchIndex <= LZ4_DISTANCE_MAX && nbAttempts--) {
const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + dictMatchIndex;
if (LZ4_read32(matchPtr) == pattern) {
int mlt;
int back = 0;
const BYTE* vLimit = ip + (dictEndOffset - dictMatchIndex);
if (vLimit > iHighLimit) vLimit = iHighLimit;
mlt = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH;
back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->prefixStart) : 0;
mlt -= back;
if (mlt > longest) {
longest = mlt;
offset = (int)(ipIndex - matchIndex);
sBack = back;
DEBUGLOG(7, "found match of length %i within extDictCtx", longest);
} }
{ U32 const nextOffset = DELTANEXTU16(dictCtx->chainTable, dictMatchIndex);
dictMatchIndex -= nextOffset;
matchIndex -= nextOffset;
} } }
{ LZ4HC_match_t md;
assert(longest >= 0);
md.len = longest;
md.off = offset;
md.back = sBack;
return md;
}
}
LZ4_FORCE_INLINE LZ4HC_match_t
LZ4HC_InsertAndFindBestMatch(LZ4HC_CCtx_internal* const hc4, /* Index table will be updated */
const BYTE* const ip, const BYTE* const iLimit,
const int maxNbAttempts,
const int patternAnalysis,
const dictCtx_directive dict)
{
DEBUGLOG(7, "LZ4HC_InsertAndFindBestMatch");
/* 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, maxNbAttempts, patternAnalysis, 0 /*chainSwap*/, dict, favorCompressionRatio);
}
LZ4_FORCE_INLINE int LZ4HC_compress_hashChain (
LZ4HC_CCtx_internal* const ctx,
const char* const source,
char* const dest,
int* srcSizePtr,
int const maxOutputSize,
int maxNbAttempts,
const limitedOutput_directive limit,
const dictCtx_directive dict
)
{
const int inputSize = *srcSizePtr;
const int patternAnalysis = (maxNbAttempts > 128); /* levels 9+ */
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;
const BYTE* start0;
const BYTE* start2 = NULL;
const BYTE* start3 = NULL;
LZ4HC_match_t m0, m1, m2, m3;
const LZ4HC_match_t nomatch = {0, 0, 0};
/* init */
DEBUGLOG(5, "LZ4HC_compress_hashChain (dict?=>%i)", dict);
*srcSizePtr = 0;
if (limit == fillOutput) 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) {
m1 = LZ4HC_InsertAndFindBestMatch(ctx, ip, matchlimit, maxNbAttempts, patternAnalysis, dict);
if (m1.len<MINMATCH) { ip++; continue; }
/* saved, in case we would skip too much */
start0 = ip; m0 = m1;
_Search2:
DEBUGLOG(7, "_Search2 (currently found match of size %i)", m1.len);
if (ip+m1.len <= mflimit) {
start2 = ip + m1.len - 2;
m2 = LZ4HC_InsertAndGetWiderMatch(ctx,
start2, ip + 0, matchlimit, m1.len,
maxNbAttempts, patternAnalysis, 0, dict, favorCompressionRatio);
start2 += m2.back;
} else {
m2 = nomatch; /* do not search further */
}
if (m2.len <= m1.len) { /* No better match => encode ML1 immediately */
optr = op;
if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
m1.len, m1.off,
limit, oend) )
goto _dest_overflow;
continue;
}
if (start0 < ip) { /* first match was skipped at least once */
if (start2 < ip + m0.len) { /* squeezing ML1 between ML0(original ML1) and ML2 */
ip = start0; m1 = m0; /* restore initial Match1 */
} }
/* Here, start0==ip */
if ((start2 - ip) < 3) { /* First Match too small : removed */
ip = start2;
m1 = m2;
goto _Search2;
}
_Search3:
if ((start2 - ip) < OPTIMAL_ML) {
int correction;
int new_ml = m1.len;
if (new_ml > OPTIMAL_ML) new_ml = OPTIMAL_ML;
if (ip+new_ml > start2 + m2.len - MINMATCH)
new_ml = (int)(start2 - ip) + m2.len - MINMATCH;
correction = new_ml - (int)(start2 - ip);
if (correction > 0) {
start2 += correction;
m2.len -= correction;
}
}
if (start2 + m2.len <= mflimit) {
start3 = start2 + m2.len - 3;
m3 = LZ4HC_InsertAndGetWiderMatch(ctx,
start3, start2, matchlimit, m2.len,
maxNbAttempts, patternAnalysis, 0, dict, favorCompressionRatio);
start3 += m3.back;
} else {
m3 = nomatch; /* do not search further */
}
if (m3.len <= m2.len) { /* No better match => encode ML1 and ML2 */
/* ip & ref are known; Now for ml */
if (start2 < ip+m1.len) m1.len = (int)(start2 - ip);
/* Now, encode 2 sequences */
optr = op;
if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
m1.len, m1.off,
limit, oend) )
goto _dest_overflow;
ip = start2;
optr = op;
if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
m2.len, m2.off,
limit, oend) ) {
m1 = m2;
goto _dest_overflow;
}
continue;
}
if (start3 < ip+m1.len+3) { /* Not enough space for match 2 : remove it */
if (start3 >= (ip+m1.len)) { /* can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1 */
if (start2 < ip+m1.len) {
int correction = (int)(ip+m1.len - start2);
start2 += correction;
m2.len -= correction;
if (m2.len < MINMATCH) {
start2 = start3;
m2 = m3;
}
}
optr = op;
if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
m1.len, m1.off,
limit, oend) )
goto _dest_overflow;
ip = start3;
m1 = m3;
start0 = start2;
m0 = m2;
goto _Search2;
}
start2 = start3;
m2 = m3;
goto _Search3;
}
/*
* OK, now we have 3 ascending matches;
* let's write the first one ML1.
* ip & ref are known; Now decide ml.
*/
if (start2 < ip+m1.len) {
if ((start2 - ip) < OPTIMAL_ML) {
int correction;
if (m1.len > OPTIMAL_ML) m1.len = OPTIMAL_ML;
if (ip + m1.len > start2 + m2.len - MINMATCH)
m1.len = (int)(start2 - ip) + m2.len - MINMATCH;
correction = m1.len - (int)(start2 - ip);
if (correction > 0) {
start2 += correction;
m2.len -= correction;
}
} else {
m1.len = (int)(start2 - ip);
}
}
optr = op;
if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor),
m1.len, m1.off,
limit, oend) )
goto _dest_overflow;
/* ML2 becomes ML1 */
ip = start2; m1 = m2;
/* ML3 becomes ML2 */
start2 = start3; m2 = m3;
/* let's find a new ML3 */
goto _Search3;
}
_last_literals:
/* Encode Last Literals */
{ size_t lastRunSize = (size_t)(iend - anchor); /* literals */
size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255;
size_t const totalSize = 1 + llAdd + lastRunSize;
if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */
if (limit && (op + totalSize > oend)) {
if (limit == limitedOutput) return 0;
/* adapt lastRunSize to fill 'dest' */
lastRunSize = (size_t)(oend - op) - 1 /*token*/;
llAdd = (lastRunSize + 256 - RUN_MASK) / 256;
lastRunSize -= llAdd;
}
DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize);
ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */
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);
}
LZ4_memcpy(op, anchor, lastRunSize);
op += lastRunSize;
}
/* End */
*srcSizePtr = (int) (((const char*)ip) - source);
return (int) (((char*)op)-dest);
_dest_overflow:
if (limit == fillOutput) {
/* Assumption : @ip, @anchor, @optr and @m1 must be set correctly */
size_t const ll = (size_t)(ip - anchor);
size_t const ll_addbytes = (ll + 240) / 255;
size_t const ll_totalCost = 1 + ll_addbytes + ll;
BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */
DEBUGLOG(6, "Last sequence overflowing");
op = optr; /* restore correct out pointer */
if (op + ll_totalCost <= maxLitPos) {
/* ll validated; now adjust match length */
size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost));
size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255);
assert(maxMlSize < INT_MAX); assert(m1.len >= 0);
if ((size_t)m1.len > maxMlSize) m1.len = (int)maxMlSize;
if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + m1.len >= MFLIMIT) {
LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), m1.len, m1.off, notLimited, oend);
} }
goto _last_literals;
}
/* compression failed */
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,
const HCfavor_e favorDecSpeed);
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
)
{
DEBUGLOG(5, "LZ4HC_compress_generic_internal(src=%p, srcSize=%d)",
src, *srcSizePtr);
if (limit == fillOutput && 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;
{ cParams_t const cParam = LZ4HC_getCLevelParams(cLevel);
HCfavor_e const favor = ctx->favorDecSpeed ? favorDecompressionSpeed : favorCompressionRatio;
int result;
if (cParam.strat == lz4mid) {
result = LZ4MID_compress(ctx,
src, dst, srcSizePtr, dstCapacity,
limit, dict);
} else if (cParam.strat == lz4hc) {
result = LZ4HC_compress_hashChain(ctx,
src, dst, srcSizePtr, dstCapacity,
cParam.nbSearches, limit, dict);
} else {
assert(cParam.strat == lz4opt);
result = LZ4HC_compress_optimal(ctx,
src, dst, srcSizePtr, dstCapacity,
cParam.nbSearches, cParam.targetLength, limit,
cLevel >= LZ4HC_CLEVEL_MAX, /* ultra mode */
dict, favor);
}
if (result <= 0) ctx->dirty = 1;
return result;
}
}
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 isStateCompatible(const LZ4HC_CCtx_internal* ctx1, const LZ4HC_CCtx_internal* ctx2)
{
int const isMid1 = LZ4HC_getCLevelParams(ctx1->compressionLevel).strat == lz4mid;
int const isMid2 = LZ4HC_getCLevelParams(ctx2->compressionLevel).strat == lz4mid;
return !(isMid1 ^ isMid2);
}
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 = (size_t)(ctx->end - ctx->prefixStart) + (ctx->dictLimit - 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 && isStateCompatible(ctx, ctx->dictCtx)) {
LZ4_memcpy(ctx, ctx->dictCtx, sizeof(LZ4HC_CCtx_internal));
LZ4HC_setExternalDict(ctx, (const BYTE *)src);
ctx->compressionLevel = (short)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, usingDictCtxHc);
}
}
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 (int)sizeof(LZ4_streamHC_t); }
static size_t LZ4_streamHC_t_alignment(void)
{
#if LZ4_ALIGN_TEST
typedef struct { char c; LZ4_streamHC_t t; } t_a;
return sizeof(t_a) - sizeof(LZ4_streamHC_t);
#else
return 1; /* effectively disabled */
#endif
}
/* state is presumed correctly initialized,
* in which case its size and alignment have already been validate */
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 (!LZ4_isAligned(state, LZ4_streamHC_t_alignment())) return 0;
LZ4_resetStreamHC_fast((LZ4_streamHC_t*)state, compressionLevel);
LZ4HC_init_internal (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, notLimited);
}
int LZ4_compress_HC_extStateHC (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel)
{
LZ4_streamHC_t* const ctx = LZ4_initStreamHC(state, sizeof(*ctx));
if (ctx==NULL) return 0; /* init failure */
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)
{
int cSize;
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
LZ4_streamHC_t* const statePtr = (LZ4_streamHC_t*)ALLOC(sizeof(LZ4_streamHC_t));
if (statePtr==NULL) return 0;
#else
LZ4_streamHC_t state;
LZ4_streamHC_t* const statePtr = &state;
#endif
DEBUGLOG(5, "LZ4_compress_HC")
cSize = LZ4_compress_HC_extStateHC(statePtr, src, dst, srcSize, dstCapacity, compressionLevel);
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
FREEMEM(statePtr);
#endif
return cSize;
}
/* state is presumed sized correctly (>= sizeof(LZ4_streamHC_t)) */
int LZ4_compress_HC_destSize(void* state, const char* source, char* dest, int* sourceSizePtr, int targetDestSize, int cLevel)
{
LZ4_streamHC_t* const ctx = LZ4_initStreamHC(state, sizeof(*ctx));
if (ctx==NULL) return 0; /* init failure */
LZ4HC_init_internal(&ctx->internal_donotuse, (const BYTE*) source);
LZ4_setCompressionLevel(ctx, cLevel);
return LZ4HC_compress_generic(&ctx->internal_donotuse, source, dest, sourceSizePtr, targetDestSize, cLevel, fillOutput);
}
/**************************************
* Streaming Functions
**************************************/
/* allocation */
#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION)
LZ4_streamHC_t* LZ4_createStreamHC(void)
{
LZ4_streamHC_t* const state =
(LZ4_streamHC_t*)ALLOC_AND_ZERO(sizeof(LZ4_streamHC_t));
if (state == NULL) return NULL;
LZ4_setCompressionLevel(state, LZ4HC_CLEVEL_DEFAULT);
return state;
}
int LZ4_freeStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr)
{
DEBUGLOG(4, "LZ4_freeStreamHC(%p)", LZ4_streamHCPtr);
if (!LZ4_streamHCPtr) return 0; /* support free on NULL */
FREEMEM(LZ4_streamHCPtr);
return 0;
}
#endif
LZ4_streamHC_t* LZ4_initStreamHC (void* buffer, size_t size)
{
LZ4_streamHC_t* const LZ4_streamHCPtr = (LZ4_streamHC_t*)buffer;
DEBUGLOG(4, "LZ4_initStreamHC(%p, %u)", buffer, (unsigned)size);
/* check conditions */
if (buffer == NULL) return NULL;
if (size < sizeof(LZ4_streamHC_t)) return NULL;
if (!LZ4_isAligned(buffer, LZ4_streamHC_t_alignment())) return NULL;
/* init */
{ LZ4HC_CCtx_internal* const hcstate = &(LZ4_streamHCPtr->internal_donotuse);
MEM_INIT(hcstate, 0, sizeof(*hcstate)); }
LZ4_setCompressionLevel(LZ4_streamHCPtr, LZ4HC_CLEVEL_DEFAULT);
return LZ4_streamHCPtr;
}
/* just a stub */
void LZ4_resetStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel)
{
LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr));
LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel);
}
void LZ4_resetStreamHC_fast (LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel)
{
LZ4HC_CCtx_internal* const s = &LZ4_streamHCPtr->internal_donotuse;
DEBUGLOG(5, "LZ4_resetStreamHC_fast(%p, %d)", LZ4_streamHCPtr, compressionLevel);
if (s->dirty) {
LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr));
} else {
assert(s->end >= s->prefixStart);
s->dictLimit += (U32)(s->end - s->prefixStart);
s->prefixStart = NULL;
s->end = NULL;
s->dictCtx = NULL;
}
LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel);
}
void LZ4_setCompressionLevel(LZ4_streamHC_t* LZ4_streamHCPtr, int compressionLevel)
{
DEBUGLOG(5, "LZ4_setCompressionLevel(%p, %d)", LZ4_streamHCPtr, compressionLevel);
if (compressionLevel < 1) compressionLevel = LZ4HC_CLEVEL_DEFAULT;
if (compressionLevel > LZ4HC_CLEVEL_MAX) compressionLevel = LZ4HC_CLEVEL_MAX;
LZ4_streamHCPtr->internal_donotuse.compressionLevel = (short)compressionLevel;
}
void LZ4_favorDecompressionSpeed(LZ4_streamHC_t* LZ4_streamHCPtr, int favor)
{
LZ4_streamHCPtr->internal_donotuse.favorDecSpeed = (favor!=0);
}
/* LZ4_loadDictHC() :
* LZ4_streamHCPtr is presumed properly initialized */
int LZ4_loadDictHC (LZ4_streamHC_t* LZ4_streamHCPtr,
const char* dictionary, int dictSize)
{
LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse;
cParams_t cp;
DEBUGLOG(4, "LZ4_loadDictHC(ctx:%p, dict:%p, dictSize:%d, clevel=%d)", LZ4_streamHCPtr, dictionary, dictSize, ctxPtr->compressionLevel);
assert(dictSize >= 0);
assert(LZ4_streamHCPtr != NULL);
if (dictSize > 64 KB) {
dictionary += (size_t)dictSize - 64 KB;
dictSize = 64 KB;
}
/* need a full initialization, there are bad side-effects when using resetFast() */
{ int const cLevel = ctxPtr->compressionLevel;
LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr));
LZ4_setCompressionLevel(LZ4_streamHCPtr, cLevel);
cp = LZ4HC_getCLevelParams(cLevel);
}
LZ4HC_init_internal (ctxPtr, (const BYTE*)dictionary);
ctxPtr->end = (const BYTE*)dictionary + dictSize;
if (cp.strat == lz4mid) {
LZ4MID_fillHTable (ctxPtr, dictionary, (size_t)dictSize);
} else {
if (dictSize >= LZ4HC_HASHSIZE) 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->prefixStart + 4)
&& (LZ4HC_getCLevelParams(ctxPtr->compressionLevel).strat != lz4mid) ) {
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->dictStart = ctxPtr->prefixStart;
ctxPtr->dictLimit += (U32)(ctxPtr->end - ctxPtr->prefixStart);
ctxPtr->prefixStart = newBlock;
ctxPtr->end = newBlock;
ctxPtr->nextToUpdate = ctxPtr->dictLimit; /* match referencing will resume from there */
/* cannot reference an extDict and a dictCtx at the same time */
ctxPtr->dictCtx = NULL;
}
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(5, "LZ4_compressHC_continue_generic(ctx=%p, src=%p, srcSize=%d, limit=%d)",
LZ4_streamHCPtr, src, *srcSizePtr, limit);
assert(ctxPtr != NULL);
/* auto-init if forgotten */
if (ctxPtr->prefixStart == NULL)
LZ4HC_init_internal (ctxPtr, (const BYTE*) src);
/* Check overflow */
if ((size_t)(ctxPtr->end - ctxPtr->prefixStart) + ctxPtr->dictLimit > 2 GB) {
size_t dictSize = (size_t)(ctxPtr->end - ctxPtr->prefixStart);
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->dictStart;
const BYTE* const dictEnd = ctxPtr->dictStart + (ctxPtr->dictLimit - ctxPtr->lowLimit);
if ((sourceEnd > dictBegin) && ((const BYTE*)src < dictEnd)) {
if (sourceEnd > dictEnd) sourceEnd = dictEnd;
ctxPtr->lowLimit += (U32)(sourceEnd - ctxPtr->dictStart);
ctxPtr->dictStart += (U32)(sourceEnd - ctxPtr->dictStart);
/* invalidate dictionary is it's too small */
if (ctxPtr->dictLimit - ctxPtr->lowLimit < LZ4HC_HASHSIZE) {
ctxPtr->lowLimit = ctxPtr->dictLimit;
ctxPtr->dictStart = ctxPtr->prefixStart;
} } }
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)
{
DEBUGLOG(5, "LZ4_compress_HC_continue");
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, notLimited);
}
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, fillOutput);
}
/* LZ4_saveDictHC :
* save history content
* into a user-provided buffer
* which is then used to continue compression
*/
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->prefixStart);
DEBUGLOG(5, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize);
assert(prefixSize >= 0);
if (dictSize > 64 KB) dictSize = 64 KB;
if (dictSize < 4) dictSize = 0;
if (dictSize > prefixSize) dictSize = prefixSize;
if (safeBuffer == NULL) assert(dictSize == 0);
if (dictSize > 0)
LZ4_memmove(safeBuffer, streamPtr->end - dictSize, (size_t)dictSize);
{ U32 const endIndex = (U32)(streamPtr->end - streamPtr->prefixStart) + streamPtr->dictLimit;
streamPtr->end = (safeBuffer == NULL) ? NULL : (const BYTE*)safeBuffer + dictSize;
streamPtr->prefixStart = (const BYTE*)safeBuffer;
streamPtr->dictLimit = endIndex - (U32)dictSize;
streamPtr->lowLimit = endIndex - (U32)dictSize;
streamPtr->dictStart = streamPtr->prefixStart;
if (streamPtr->nextToUpdate < streamPtr->dictLimit)
streamPtr->nextToUpdate = streamPtr->dictLimit;
}
return dictSize;
}
/* ================================================
* LZ4 Optimal parser (levels [LZ4HC_CLEVEL_OPT_MIN - LZ4HC_CLEVEL_MAX])
* ===============================================*/
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;
assert(litlen >= 0);
if (litlen >= (int)RUN_MASK)
price += 1 + ((litlen-(int)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 */
assert(litlen >= 0);
assert(mlen >= MINMATCH);
price += LZ4HC_literalsPrice(litlen);
if (mlen >= (int)(ML_MASK+MINMATCH))
price += 1 + ((mlen-(int)(ML_MASK+MINMATCH)) / 255);
return price;
}
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,
const HCfavor_e favorDecSpeed)
{
LZ4HC_match_t const match0 = { 0 , 0, 0 };
/* 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 */
LZ4HC_match_t md = LZ4HC_InsertAndGetWiderMatch(ctx, ip, ip, iHighLimit, minLen, nbSearches, 1 /*patternAnalysis*/, 1 /*chainSwap*/, dict, favorDecSpeed);
assert(md.back == 0);
if (md.len <= minLen) return match0;
if (favorDecSpeed) {
if ((md.len>18) & (md.len<=36)) md.len=18; /* favor dec.speed (shortcut) */
}
return md;
}
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,
const HCfavor_e favorDecSpeed)
{
int retval = 0;
#define TRAILING_LITERALS 3
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
LZ4HC_optimal_t* const opt = (LZ4HC_optimal_t*)ALLOC(sizeof(LZ4HC_optimal_t) * (LZ4_OPT_NUM + TRAILING_LITERALS));
#else
LZ4HC_optimal_t opt[LZ4_OPT_NUM + TRAILING_LITERALS]; /* ~64 KB, which is a bit large for stack... */
#endif
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;
int ovml = MINMATCH; /* overflow - last sequence */
int ovoff = 0;
/* init */
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
if (opt == NULL) goto _return_label;
#endif
DEBUGLOG(5, "LZ4HC_compress_optimal(dst=%p, dstCapa=%u)", dst, (unsigned)dstCapacity);
*srcSizePtr = 0;
if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */
if (sufficient_len >= LZ4_OPT_NUM) sufficient_len = LZ4_OPT_NUM-1;
/* Main Loop */
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, favorDecSpeed);
if (firstMatch.len==0) { ip++; continue; }
if ((size_t)firstMatch.len > sufficient_len) {
/* good enough solution : immediate encoding */
int const firstML = firstMatch.len;
opSaved = op;
if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), firstML, firstMatch.off, limit, oend) ) { /* updates ip, op and anchor */
ovml = firstML;
ovoff = firstMatch.off;
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 const matchML = firstMatch.len; /* necessarily < sufficient_len < LZ4_OPT_NUM */
int const offset = firstMatch.off;
int mlen;
assert(matchML < LZ4_OPT_NUM);
for (mlen = MINMATCH ; 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, favorDecSpeed);
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, favorDecSpeed);
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);
}
assert((U32)favorDecSpeed <= 1);
if (pos > last_match_pos+TRAILING_LITERALS
|| price <= opt[pos].price - (int)favorDecSpeed) {
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++) */
assert(last_match_pos < LZ4_OPT_NUM + TRAILING_LITERALS);
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 <= LZ4_DISTANCE_MAX));
opSaved = op;
if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml, offset, limit, oend) ) { /* updates ip, op and anchor */
ovml = ml;
ovoff = offset;
goto _dest_overflow;
} } }
} /* while (ip <= mflimit) */
_last_literals:
/* Encode Last Literals */
{ size_t lastRunSize = (size_t)(iend - anchor); /* literals */
size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255;
size_t const totalSize = 1 + llAdd + lastRunSize;
if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */
if (limit && (op + totalSize > oend)) {
if (limit == limitedOutput) { /* Check output limit */
retval = 0;
goto _return_label;
}
/* adapt lastRunSize to fill 'dst' */
lastRunSize = (size_t)(oend - op) - 1 /*token*/;
llAdd = (lastRunSize + 256 - RUN_MASK) / 256;
lastRunSize -= llAdd;
}
DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize);
ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */
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);
}
LZ4_memcpy(op, anchor, lastRunSize);
op += lastRunSize;
}
/* End */
*srcSizePtr = (int) (((const char*)ip) - source);
retval = (int) ((char*)op-dst);
goto _return_label;
_dest_overflow:
if (limit == fillOutput) {
/* Assumption : ip, anchor, ovml and ovref must be set correctly */
size_t const ll = (size_t)(ip - anchor);
size_t const ll_addbytes = (ll + 240) / 255;
size_t const ll_totalCost = 1 + ll_addbytes + ll;
BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */
DEBUGLOG(6, "Last sequence overflowing (only %i bytes remaining)", (int)(oend-1-opSaved));
op = opSaved; /* restore correct out pointer */
if (op + ll_totalCost <= maxLitPos) {
/* ll validated; now adjust match length */
size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost));
size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255);
assert(maxMlSize < INT_MAX); assert(ovml >= 0);
if ((size_t)ovml > maxMlSize) ovml = (int)maxMlSize;
if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + ovml >= MFLIMIT) {
DEBUGLOG(6, "Space to end : %i + ml (%i)", (int)((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1), ovml);
DEBUGLOG(6, "Before : ip = %p, anchor = %p", ip, anchor);
LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ovml, ovoff, notLimited, oend);
DEBUGLOG(6, "After : ip = %p, anchor = %p", ip, anchor);
} }
goto _last_literals;
}
_return_label:
#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1
if (opt) FREEMEM(opt);
#endif
return retval;
}
/***************************************************
* Deprecated Functions
***************************************************/
/* These functions currently generate deprecation warnings */
/* Wrappers for 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 sizeof(LZ4_streamHC_t); }
/* state is presumed correctly sized, aka >= sizeof(LZ4_streamHC_t)
* @return : 0 on success, !=0 if error */
int LZ4_resetStreamStateHC(void* state, char* inputBuffer)
{
LZ4_streamHC_t* const hc4 = LZ4_initStreamHC(state, sizeof(*hc4));
if (hc4 == NULL) return 1; /* init failed */
LZ4HC_init_internal (&hc4->internal_donotuse, (const BYTE*)inputBuffer);
return 0;
}
#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION)
void* LZ4_createHC (const char* inputBuffer)
{
LZ4_streamHC_t* const hc4 = LZ4_createStreamHC();
if (hc4 == NULL) return NULL; /* not enough memory */
LZ4HC_init_internal (&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;
}
#endif
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, notLimited);
}
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)
{
LZ4HC_CCtx_internal* const s = &((LZ4_streamHC_t*)LZ4HC_Data)->internal_donotuse;
const BYTE* const bufferStart = s->prefixStart - s->dictLimit + s->lowLimit;
LZ4_resetStreamHC_fast((LZ4_streamHC_t*)LZ4HC_Data, s->compressionLevel);
/* ugly conversion trick, required to evade (const char*) -> (char*) cast-qual warning :( */
return (char*)(uptrval)bufferStart;
}
|