summaryrefslogtreecommitdiffstats
path: root/src/3rdparty/pcre/pcre_dfa_exec.c
blob: adb1bbf3f5eff907d05b5072419b750b5ce357b1 (plain)
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
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
/*************************************************
*      Perl-Compatible Regular Expressions       *
*************************************************/

/* PCRE is a library of functions to support regular expressions whose syntax
and semantics are as close as possible to those of the Perl 5 language (but see
below for why this module is different).

                       Written by Philip Hazel
           Copyright (c) 1997-2012 University of Cambridge

-----------------------------------------------------------------------------
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.

    * Neither the name of the University of Cambridge nor the names of its
      contributors may be used to endorse or promote products derived from
      this software without specific prior written permission.

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.
-----------------------------------------------------------------------------
*/

/* This module contains the external function pcre_dfa_exec(), which is an
alternative matching function that uses a sort of DFA algorithm (not a true
FSM). This is NOT Perl-compatible, but it has advantages in certain
applications. */


/* NOTE ABOUT PERFORMANCE: A user of this function sent some code that improved
the performance of his patterns greatly. I could not use it as it stood, as it
was not thread safe, and made assumptions about pattern sizes. Also, it caused
test 7 to loop, and test 9 to crash with a segfault.

The issue is the check for duplicate states, which is done by a simple linear
search up the state list. (Grep for "duplicate" below to find the code.) For
many patterns, there will never be many states active at one time, so a simple
linear search is fine. In patterns that have many active states, it might be a
bottleneck. The suggested code used an indexing scheme to remember which states
had previously been used for each character, and avoided the linear search when
it knew there was no chance of a duplicate. This was implemented when adding
states to the state lists.

I wrote some thread-safe, not-limited code to try something similar at the time
of checking for duplicates (instead of when adding states), using index vectors
on the stack. It did give a 13% improvement with one specially constructed
pattern for certain subject strings, but on other strings and on many of the
simpler patterns in the test suite it did worse. The major problem, I think,
was the extra time to initialize the index. This had to be done for each call
of internal_dfa_exec(). (The supplied patch used a static vector, initialized
only once - I suspect this was the cause of the problems with the tests.)

Overall, I concluded that the gains in some cases did not outweigh the losses
in others, so I abandoned this code. */



#ifdef PCRE_HAVE_CONFIG_H
#include "config.h"
#endif

#define NLBLOCK md             /* Block containing newline information */
#define PSSTART start_subject  /* Field containing processed string start */
#define PSEND   end_subject    /* Field containing processed string end */

#include "pcre_internal.h"


/* For use to indent debugging output */

#define SP "                   "


/*************************************************
*      Code parameters and static tables         *
*************************************************/

/* These are offsets that are used to turn the OP_TYPESTAR and friends opcodes
into others, under special conditions. A gap of 20 between the blocks should be
enough. The resulting opcodes don't have to be less than 256 because they are
never stored, so we push them well clear of the normal opcodes. */

#define OP_PROP_EXTRA       300
#define OP_EXTUNI_EXTRA     320
#define OP_ANYNL_EXTRA      340
#define OP_HSPACE_EXTRA     360
#define OP_VSPACE_EXTRA     380


/* This table identifies those opcodes that are followed immediately by a
character that is to be tested in some way. This makes it possible to
centralize the loading of these characters. In the case of Type * etc, the
"character" is the opcode for \D, \d, \S, \s, \W, or \w, which will always be a
small value. Non-zero values in the table are the offsets from the opcode where
the character is to be found. ***NOTE*** If the start of this table is
modified, the three tables that follow must also be modified. */

static const pcre_uint8 coptable[] = {
  0,                             /* End                                    */
  0, 0, 0, 0, 0,                 /* \A, \G, \K, \B, \b                     */
  0, 0, 0, 0, 0, 0,              /* \D, \d, \S, \s, \W, \w                 */
  0, 0, 0,                       /* Any, AllAny, Anybyte                   */
  0, 0,                          /* \P, \p                                 */
  0, 0, 0, 0, 0,                 /* \R, \H, \h, \V, \v                     */
  0,                             /* \X                                     */
  0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
  1,                             /* Char                                   */
  1,                             /* Chari                                  */
  1,                             /* not                                    */
  1,                             /* noti                                   */
  /* Positive single-char repeats                                          */
  1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
  1+IMM2_SIZE, 1+IMM2_SIZE,      /* upto, minupto                          */
  1+IMM2_SIZE,                   /* exact                                  */
  1, 1, 1, 1+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */
  1, 1, 1, 1, 1, 1,              /* *I, *?I, +I, +?I, ?I, ??I              */
  1+IMM2_SIZE, 1+IMM2_SIZE,      /* upto I, minupto I                      */
  1+IMM2_SIZE,                   /* exact I                                */
  1, 1, 1, 1+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */
  /* Negative single-char repeats - only for chars < 256                   */
  1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
  1+IMM2_SIZE, 1+IMM2_SIZE,      /* NOT upto, minupto                      */
  1+IMM2_SIZE,                   /* NOT exact                              */
  1, 1, 1, 1+IMM2_SIZE,          /* NOT *+, ++, ?+, upto+                  */
  1, 1, 1, 1, 1, 1,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */
  1+IMM2_SIZE, 1+IMM2_SIZE,      /* NOT upto I, minupto I                  */
  1+IMM2_SIZE,                   /* NOT exact I                            */
  1, 1, 1, 1+IMM2_SIZE,          /* NOT *+I, ++I, ?+I, upto+I              */
  /* Positive type repeats                                                 */
  1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
  1+IMM2_SIZE, 1+IMM2_SIZE,      /* Type upto, minupto                     */
  1+IMM2_SIZE,                   /* Type exact                             */
  1, 1, 1, 1+IMM2_SIZE,          /* Type *+, ++, ?+, upto+                 */
  /* Character class & ref repeats                                         */
  0, 0, 0, 0, 0, 0,              /* *, *?, +, +?, ?, ??                    */
  0, 0,                          /* CRRANGE, CRMINRANGE                    */
  0,                             /* CLASS                                  */
  0,                             /* NCLASS                                 */
  0,                             /* XCLASS - variable length               */
  0,                             /* REF                                    */
  0,                             /* REFI                                   */
  0,                             /* RECURSE                                */
  0,                             /* CALLOUT                                */
  0,                             /* Alt                                    */
  0,                             /* Ket                                    */
  0,                             /* KetRmax                                */
  0,                             /* KetRmin                                */
  0,                             /* KetRpos                                */
  0,                             /* Reverse                                */
  0,                             /* Assert                                 */
  0,                             /* Assert not                             */
  0,                             /* Assert behind                          */
  0,                             /* Assert behind not                      */
  0, 0,                          /* ONCE, ONCE_NC                          */
  0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
  0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
  0, 0,                          /* CREF, NCREF                            */
  0, 0,                          /* RREF, NRREF                            */
  0,                             /* DEF                                    */
  0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
  0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
  0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG         */
  0, 0, 0, 0,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */
  0, 0                           /* CLOSE, SKIPZERO  */
};

/* This table identifies those opcodes that inspect a character. It is used to
remember the fact that a character could have been inspected when the end of
the subject is reached. ***NOTE*** If the start of this table is modified, the
two tables that follow must also be modified. */

static const pcre_uint8 poptable[] = {
  0,                             /* End                                    */
  0, 0, 0, 1, 1,                 /* \A, \G, \K, \B, \b                     */
  1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */
  1, 1, 1,                       /* Any, AllAny, Anybyte                   */
  1, 1,                          /* \P, \p                                 */
  1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */
  1,                             /* \X                                     */
  0, 0, 0, 0, 0, 0,              /* \Z, \z, ^, ^M, $, $M                   */
  1,                             /* Char                                   */
  1,                             /* Chari                                  */
  1,                             /* not                                    */
  1,                             /* noti                                   */
  /* Positive single-char repeats                                          */
  1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
  1, 1, 1,                       /* upto, minupto, exact                   */
  1, 1, 1, 1,                    /* *+, ++, ?+, upto+                      */
  1, 1, 1, 1, 1, 1,              /* *I, *?I, +I, +?I, ?I, ??I              */
  1, 1, 1,                       /* upto I, minupto I, exact I             */
  1, 1, 1, 1,                    /* *+I, ++I, ?+I, upto+I                  */
  /* Negative single-char repeats - only for chars < 256                   */
  1, 1, 1, 1, 1, 1,              /* NOT *, *?, +, +?, ?, ??                */
  1, 1, 1,                       /* NOT upto, minupto, exact               */
  1, 1, 1, 1,                    /* NOT *+, ++, ?+, upto+                  */
  1, 1, 1, 1, 1, 1,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */
  1, 1, 1,                       /* NOT upto I, minupto I, exact I         */
  1, 1, 1, 1,                    /* NOT *+I, ++I, ?+I, upto+I              */
  /* Positive type repeats                                                 */
  1, 1, 1, 1, 1, 1,              /* Type *, *?, +, +?, ?, ??               */
  1, 1, 1,                       /* Type upto, minupto, exact              */
  1, 1, 1, 1,                    /* Type *+, ++, ?+, upto+                 */
  /* Character class & ref repeats                                         */
  1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */
  1, 1,                          /* CRRANGE, CRMINRANGE                    */
  1,                             /* CLASS                                  */
  1,                             /* NCLASS                                 */
  1,                             /* XCLASS - variable length               */
  0,                             /* REF                                    */
  0,                             /* REFI                                   */
  0,                             /* RECURSE                                */
  0,                             /* CALLOUT                                */
  0,                             /* Alt                                    */
  0,                             /* Ket                                    */
  0,                             /* KetRmax                                */
  0,                             /* KetRmin                                */
  0,                             /* KetRpos                                */
  0,                             /* Reverse                                */
  0,                             /* Assert                                 */
  0,                             /* Assert not                             */
  0,                             /* Assert behind                          */
  0,                             /* Assert behind not                      */
  0, 0,                          /* ONCE, ONCE_NC                          */
  0, 0, 0, 0, 0,                 /* BRA, BRAPOS, CBRA, CBRAPOS, COND       */
  0, 0, 0, 0, 0,                 /* SBRA, SBRAPOS, SCBRA, SCBRAPOS, SCOND  */
  0, 0,                          /* CREF, NCREF                            */
  0, 0,                          /* RREF, NRREF                            */
  0,                             /* DEF                                    */
  0, 0, 0,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */
  0, 0, 0,                       /* MARK, PRUNE, PRUNE_ARG                 */
  0, 0, 0, 0,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG         */
  0, 0, 0, 0,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */
  0, 0                           /* CLOSE, SKIPZERO                        */
};

/* These 2 tables allow for compact code for testing for \D, \d, \S, \s, \W,
and \w */

static const pcre_uint8 toptable1[] = {
  0, 0, 0, 0, 0, 0,
  ctype_digit, ctype_digit,
  ctype_space, ctype_space,
  ctype_word,  ctype_word,
  0, 0                            /* OP_ANY, OP_ALLANY */
};

static const pcre_uint8 toptable2[] = {
  0, 0, 0, 0, 0, 0,
  ctype_digit, 0,
  ctype_space, 0,
  ctype_word,  0,
  1, 1                            /* OP_ANY, OP_ALLANY */
};


/* Structure for holding data about a particular state, which is in effect the
current data for an active path through the match tree. It must consist
entirely of ints because the working vector we are passed, and which we put
these structures in, is a vector of ints. */

typedef struct stateblock {
  int offset;                     /* Offset to opcode */
  int count;                      /* Count for repeats */
  int data;                       /* Some use extra data */
} stateblock;

#define INTS_PER_STATEBLOCK  (int)(sizeof(stateblock)/sizeof(int))


#ifdef PCRE_DEBUG
/*************************************************
*             Print character string             *
*************************************************/

/* Character string printing function for debugging.

Arguments:
  p            points to string
  length       number of bytes
  f            where to print

Returns:       nothing
*/

static void
pchars(const pcre_uchar *p, int length, FILE *f)
{
pcre_uint32 c;
while (length-- > 0)
  {
  if (isprint(c = *(p++)))
    fprintf(f, "%c", c);
  else
    fprintf(f, "\\x{%02x}", c);
  }
}
#endif



/*************************************************
*    Execute a Regular Expression - DFA engine   *
*************************************************/

/* This internal function applies a compiled pattern to a subject string,
starting at a given point, using a DFA engine. This function is called from the
external one, possibly multiple times if the pattern is not anchored. The
function calls itself recursively for some kinds of subpattern.

Arguments:
  md                the match_data block with fixed information
  this_start_code   the opening bracket of this subexpression's code
  current_subject   where we currently are in the subject string
  start_offset      start offset in the subject string
  offsets           vector to contain the matching string offsets
  offsetcount       size of same
  workspace         vector of workspace
  wscount           size of same
  rlevel            function call recursion level

Returns:            > 0 => number of match offset pairs placed in offsets
                    = 0 => offsets overflowed; longest matches are present
                     -1 => failed to match
                   < -1 => some kind of unexpected problem

The following macros are used for adding states to the two state vectors (one
for the current character, one for the following character). */

#define ADD_ACTIVE(x,y) \
  if (active_count++ < wscount) \
    { \
    next_active_state->offset = (x); \
    next_active_state->count  = (y); \
    next_active_state++; \
    DPRINTF(("%.*sADD_ACTIVE(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
    } \
  else return PCRE_ERROR_DFA_WSSIZE

#define ADD_ACTIVE_DATA(x,y,z) \
  if (active_count++ < wscount) \
    { \
    next_active_state->offset = (x); \
    next_active_state->count  = (y); \
    next_active_state->data   = (z); \
    next_active_state++; \
    DPRINTF(("%.*sADD_ACTIVE_DATA(%d,%d,%d)\n", rlevel*2-2, SP, (x), (y), (z))); \
    } \
  else return PCRE_ERROR_DFA_WSSIZE

#define ADD_NEW(x,y) \
  if (new_count++ < wscount) \
    { \
    next_new_state->offset = (x); \
    next_new_state->count  = (y); \
    next_new_state++; \
    DPRINTF(("%.*sADD_NEW(%d,%d)\n", rlevel*2-2, SP, (x), (y))); \
    } \
  else return PCRE_ERROR_DFA_WSSIZE

#define ADD_NEW_DATA(x,y,z) \
  if (new_count++ < wscount) \
    { \
    next_new_state->offset = (x); \
    next_new_state->count  = (y); \
    next_new_state->data   = (z); \
    next_new_state++; \
    DPRINTF(("%.*sADD_NEW_DATA(%d,%d,%d) line %d\n", rlevel*2-2, SP, \
      (x), (y), (z), __LINE__)); \
    } \
  else return PCRE_ERROR_DFA_WSSIZE

/* And now, here is the code */

static int
internal_dfa_exec(
  dfa_match_data *md,
  const pcre_uchar *this_start_code,
  const pcre_uchar *current_subject,
  int start_offset,
  int *offsets,
  int offsetcount,
  int *workspace,
  int wscount,
  int  rlevel)
{
stateblock *active_states, *new_states, *temp_states;
stateblock *next_active_state, *next_new_state;

const pcre_uint8 *ctypes, *lcc, *fcc;
const pcre_uchar *ptr;
const pcre_uchar *end_code, *first_op;

dfa_recursion_info new_recursive;

int active_count, new_count, match_count;

/* Some fields in the md block are frequently referenced, so we load them into
independent variables in the hope that this will perform better. */

const pcre_uchar *start_subject = md->start_subject;
const pcre_uchar *end_subject = md->end_subject;
const pcre_uchar *start_code = md->start_code;

#ifdef SUPPORT_UTF
BOOL utf = (md->poptions & PCRE_UTF8) != 0;
#else
BOOL utf = FALSE;
#endif

BOOL reset_could_continue = FALSE;

rlevel++;
offsetcount &= (-2);

wscount -= 2;
wscount = (wscount - (wscount % (INTS_PER_STATEBLOCK * 2))) /
          (2 * INTS_PER_STATEBLOCK);

DPRINTF(("\n%.*s---------------------\n"
  "%.*sCall to internal_dfa_exec f=%d\n",
  rlevel*2-2, SP, rlevel*2-2, SP, rlevel));

ctypes = md->tables + ctypes_offset;
lcc = md->tables + lcc_offset;
fcc = md->tables + fcc_offset;

match_count = PCRE_ERROR_NOMATCH;   /* A negative number */

active_states = (stateblock *)(workspace + 2);
next_new_state = new_states = active_states + wscount;
new_count = 0;

first_op = this_start_code + 1 + LINK_SIZE +
  ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
    *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
    ? IMM2_SIZE:0);

/* The first thing in any (sub) pattern is a bracket of some sort. Push all
the alternative states onto the list, and find out where the end is. This
makes is possible to use this function recursively, when we want to stop at a
matching internal ket rather than at the end.

If the first opcode in the first alternative is OP_REVERSE, we are dealing with
a backward assertion. In that case, we have to find out the maximum amount to
move back, and set up each alternative appropriately. */

if (*first_op == OP_REVERSE)
  {
  int max_back = 0;
  int gone_back;

  end_code = this_start_code;
  do
    {
    int back = GET(end_code, 2+LINK_SIZE);
    if (back > max_back) max_back = back;
    end_code += GET(end_code, 1);
    }
  while (*end_code == OP_ALT);

  /* If we can't go back the amount required for the longest lookbehind
  pattern, go back as far as we can; some alternatives may still be viable. */

#ifdef SUPPORT_UTF
  /* In character mode we have to step back character by character */

  if (utf)
    {
    for (gone_back = 0; gone_back < max_back; gone_back++)
      {
      if (current_subject <= start_subject) break;
      current_subject--;
      ACROSSCHAR(current_subject > start_subject, *current_subject, current_subject--);
      }
    }
  else
#endif

  /* In byte-mode we can do this quickly. */

    {
    gone_back = (current_subject - max_back < start_subject)?
      (int)(current_subject - start_subject) : max_back;
    current_subject -= gone_back;
    }

  /* Save the earliest consulted character */

  if (current_subject < md->start_used_ptr)
    md->start_used_ptr = current_subject;

  /* Now we can process the individual branches. */

  end_code = this_start_code;
  do
    {
    int back = GET(end_code, 2+LINK_SIZE);
    if (back <= gone_back)
      {
      int bstate = (int)(end_code - start_code + 2 + 2*LINK_SIZE);
      ADD_NEW_DATA(-bstate, 0, gone_back - back);
      }
    end_code += GET(end_code, 1);
    }
  while (*end_code == OP_ALT);
 }

/* This is the code for a "normal" subpattern (not a backward assertion). The
start of a whole pattern is always one of these. If we are at the top level,
we may be asked to restart matching from the same point that we reached for a
previous partial match. We still have to scan through the top-level branches to
find the end state. */

else
  {
  end_code = this_start_code;

  /* Restarting */

  if (rlevel == 1 && (md->moptions & PCRE_DFA_RESTART) != 0)
    {
    do { end_code += GET(end_code, 1); } while (*end_code == OP_ALT);
    new_count = workspace[1];
    if (!workspace[0])
      memcpy(new_states, active_states, new_count * sizeof(stateblock));
    }

  /* Not restarting */

  else
    {
    int length = 1 + LINK_SIZE +
      ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
        *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
        ? IMM2_SIZE:0);
    do
      {
      ADD_NEW((int)(end_code - start_code + length), 0);
      end_code += GET(end_code, 1);
      length = 1 + LINK_SIZE;
      }
    while (*end_code == OP_ALT);
    }
  }

workspace[0] = 0;    /* Bit indicating which vector is current */

DPRINTF(("%.*sEnd state = %d\n", rlevel*2-2, SP, (int)(end_code - start_code)));

/* Loop for scanning the subject */

ptr = current_subject;
for (;;)
  {
  int i, j;
  int clen, dlen;
  pcre_uint32 c, d;
  int forced_fail = 0;
  BOOL partial_newline = FALSE;
  BOOL could_continue = reset_could_continue;
  reset_could_continue = FALSE;

  /* Make the new state list into the active state list and empty the
  new state list. */

  temp_states = active_states;
  active_states = new_states;
  new_states = temp_states;
  active_count = new_count;
  new_count = 0;

  workspace[0] ^= 1;              /* Remember for the restarting feature */
  workspace[1] = active_count;

#ifdef PCRE_DEBUG
  printf("%.*sNext character: rest of subject = \"", rlevel*2-2, SP);
  pchars(ptr, STRLEN_UC(ptr), stdout);
  printf("\"\n");

  printf("%.*sActive states: ", rlevel*2-2, SP);
  for (i = 0; i < active_count; i++)
    printf("%d/%d ", active_states[i].offset, active_states[i].count);
  printf("\n");
#endif

  /* Set the pointers for adding new states */

  next_active_state = active_states + active_count;
  next_new_state = new_states;

  /* Load the current character from the subject outside the loop, as many
  different states may want to look at it, and we assume that at least one
  will. */

  if (ptr < end_subject)
    {
    clen = 1;        /* Number of data items in the character */
#ifdef SUPPORT_UTF
    GETCHARLENTEST(c, ptr, clen);
#else
    c = *ptr;
#endif  /* SUPPORT_UTF */
    }
  else
    {
    clen = 0;        /* This indicates the end of the subject */
    c = NOTACHAR;    /* This value should never actually be used */
    }

  /* Scan up the active states and act on each one. The result of an action
  may be to add more states to the currently active list (e.g. on hitting a
  parenthesis) or it may be to put states on the new list, for considering
  when we move the character pointer on. */

  for (i = 0; i < active_count; i++)
    {
    stateblock *current_state = active_states + i;
    BOOL caseless = FALSE;
    const pcre_uchar *code;
    int state_offset = current_state->offset;
    int codevalue, rrc;
    unsigned int count;

#ifdef PCRE_DEBUG
    printf ("%.*sProcessing state %d c=", rlevel*2-2, SP, state_offset);
    if (clen == 0) printf("EOL\n");
      else if (c > 32 && c < 127) printf("'%c'\n", c);
        else printf("0x%02x\n", c);
#endif

    /* A negative offset is a special case meaning "hold off going to this
    (negated) state until the number of characters in the data field have
    been skipped". If the could_continue flag was passed over from a previous
    state, arrange for it to passed on. */

    if (state_offset < 0)
      {
      if (current_state->data > 0)
        {
        DPRINTF(("%.*sSkipping this character\n", rlevel*2-2, SP));
        ADD_NEW_DATA(state_offset, current_state->count,
          current_state->data - 1);
        if (could_continue) reset_could_continue = TRUE;
        continue;
        }
      else
        {
        current_state->offset = state_offset = -state_offset;
        }
      }

    /* Check for a duplicate state with the same count, and skip if found.
    See the note at the head of this module about the possibility of improving
    performance here. */

    for (j = 0; j < i; j++)
      {
      if (active_states[j].offset == state_offset &&
          active_states[j].count == current_state->count)
        {
        DPRINTF(("%.*sDuplicate state: skipped\n", rlevel*2-2, SP));
        goto NEXT_ACTIVE_STATE;
        }
      }

    /* The state offset is the offset to the opcode */

    code = start_code + state_offset;
    codevalue = *code;

    /* If this opcode inspects a character, but we are at the end of the
    subject, remember the fact for use when testing for a partial match. */

    if (clen == 0 && poptable[codevalue] != 0)
      could_continue = TRUE;

    /* If this opcode is followed by an inline character, load it. It is
    tempting to test for the presence of a subject character here, but that
    is wrong, because sometimes zero repetitions of the subject are
    permitted.

    We also use this mechanism for opcodes such as OP_TYPEPLUS that take an
    argument that is not a data character - but is always one byte long because
    the values are small. We have to take special action to deal with  \P, \p,
    \H, \h, \V, \v and \X in this case. To keep the other cases fast, convert
    these ones to new opcodes. */

    if (coptable[codevalue] > 0)
      {
      dlen = 1;
#ifdef SUPPORT_UTF
      if (utf) { GETCHARLEN(d, (code + coptable[codevalue]), dlen); } else
#endif  /* SUPPORT_UTF */
      d = code[coptable[codevalue]];
      if (codevalue >= OP_TYPESTAR)
        {
        switch(d)
          {
          case OP_ANYBYTE: return PCRE_ERROR_DFA_UITEM;
          case OP_NOTPROP:
          case OP_PROP: codevalue += OP_PROP_EXTRA; break;
          case OP_ANYNL: codevalue += OP_ANYNL_EXTRA; break;
          case OP_EXTUNI: codevalue += OP_EXTUNI_EXTRA; break;
          case OP_NOT_HSPACE:
          case OP_HSPACE: codevalue += OP_HSPACE_EXTRA; break;
          case OP_NOT_VSPACE:
          case OP_VSPACE: codevalue += OP_VSPACE_EXTRA; break;
          default: break;
          }
        }
      }
    else
      {
      dlen = 0;         /* Not strictly necessary, but compilers moan */
      d = NOTACHAR;     /* if these variables are not set. */
      }


    /* Now process the individual opcodes */

    switch (codevalue)
      {
/* ========================================================================== */
      /* These cases are never obeyed. This is a fudge that causes a compile-
      time error if the vectors coptable or poptable, which are indexed by
      opcode, are not the correct length. It seems to be the only way to do
      such a check at compile time, as the sizeof() operator does not work
      in the C preprocessor. */

      case OP_TABLE_LENGTH:
      case OP_TABLE_LENGTH +
        ((sizeof(coptable) == OP_TABLE_LENGTH) &&
         (sizeof(poptable) == OP_TABLE_LENGTH)):
      break;

/* ========================================================================== */
      /* Reached a closing bracket. If not at the end of the pattern, carry
      on with the next opcode. For repeating opcodes, also add the repeat
      state. Note that KETRPOS will always be encountered at the end of the
      subpattern, because the possessive subpattern repeats are always handled
      using recursive calls. Thus, it never adds any new states.

      At the end of the (sub)pattern, unless we have an empty string and
      PCRE_NOTEMPTY is set, or PCRE_NOTEMPTY_ATSTART is set and we are at the
      start of the subject, save the match data, shifting up all previous
      matches so we always have the longest first. */

      case OP_KET:
      case OP_KETRMIN:
      case OP_KETRMAX:
      case OP_KETRPOS:
      if (code != end_code)
        {
        ADD_ACTIVE(state_offset + 1 + LINK_SIZE, 0);
        if (codevalue != OP_KET)
          {
          ADD_ACTIVE(state_offset - GET(code, 1), 0);
          }
        }
      else
        {
        if (ptr > current_subject ||
            ((md->moptions & PCRE_NOTEMPTY) == 0 &&
              ((md->moptions & PCRE_NOTEMPTY_ATSTART) == 0 ||
                current_subject > start_subject + md->start_offset)))
          {
          if (match_count < 0) match_count = (offsetcount >= 2)? 1 : 0;
            else if (match_count > 0 && ++match_count * 2 > offsetcount)
              match_count = 0;
          count = ((match_count == 0)? offsetcount : match_count * 2) - 2;
          if (count > 0) memmove(offsets + 2, offsets, count * sizeof(int));
          if (offsetcount >= 2)
            {
            offsets[0] = (int)(current_subject - start_subject);
            offsets[1] = (int)(ptr - start_subject);
            DPRINTF(("%.*sSet matched string = \"%.*s\"\n", rlevel*2-2, SP,
              offsets[1] - offsets[0], (char *)current_subject));
            }
          if ((md->moptions & PCRE_DFA_SHORTEST) != 0)
            {
            DPRINTF(("%.*sEnd of internal_dfa_exec %d: returning %d\n"
              "%.*s---------------------\n\n", rlevel*2-2, SP, rlevel,
              match_count, rlevel*2-2, SP));
            return match_count;
            }
          }
        }
      break;

/* ========================================================================== */
      /* These opcodes add to the current list of states without looking
      at the current character. */

      /*-----------------------------------------------------------------*/
      case OP_ALT:
      do { code += GET(code, 1); } while (*code == OP_ALT);
      ADD_ACTIVE((int)(code - start_code), 0);
      break;

      /*-----------------------------------------------------------------*/
      case OP_BRA:
      case OP_SBRA:
      do
        {
        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
        code += GET(code, 1);
        }
      while (*code == OP_ALT);
      break;

      /*-----------------------------------------------------------------*/
      case OP_CBRA:
      case OP_SCBRA:
      ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE + IMM2_SIZE),  0);
      code += GET(code, 1);
      while (*code == OP_ALT)
        {
        ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE),  0);
        code += GET(code, 1);
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_BRAZERO:
      case OP_BRAMINZERO:
      ADD_ACTIVE(state_offset + 1, 0);
      code += 1 + GET(code, 2);
      while (*code == OP_ALT) code += GET(code, 1);
      ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
      break;

      /*-----------------------------------------------------------------*/
      case OP_SKIPZERO:
      code += 1 + GET(code, 2);
      while (*code == OP_ALT) code += GET(code, 1);
      ADD_ACTIVE((int)(code - start_code + 1 + LINK_SIZE), 0);
      break;

      /*-----------------------------------------------------------------*/
      case OP_CIRC:
      if (ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0)
        { ADD_ACTIVE(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_CIRCM:
      if ((ptr == start_subject && (md->moptions & PCRE_NOTBOL) == 0) ||
          (ptr != end_subject && WAS_NEWLINE(ptr)))
        { ADD_ACTIVE(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_EOD:
      if (ptr >= end_subject)
        {
        if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
          could_continue = TRUE;
        else { ADD_ACTIVE(state_offset + 1, 0); }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_SOD:
      if (ptr == start_subject) { ADD_ACTIVE(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_SOM:
      if (ptr == start_subject + start_offset) { ADD_ACTIVE(state_offset + 1, 0); }
      break;


/* ========================================================================== */
      /* These opcodes inspect the next subject character, and sometimes
      the previous one as well, but do not have an argument. The variable
      clen contains the length of the current character and is zero if we are
      at the end of the subject. */

      /*-----------------------------------------------------------------*/
      case OP_ANY:
      if (clen > 0 && !IS_NEWLINE(ptr))
        {
        if (ptr + 1 >= md->end_subject &&
            (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
            NLBLOCK->nltype == NLTYPE_FIXED &&
            NLBLOCK->nllen == 2 &&
            c == NLBLOCK->nl[0])
          {
          could_continue = partial_newline = TRUE;
          }
        else
          {
          ADD_NEW(state_offset + 1, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_ALLANY:
      if (clen > 0)
        { ADD_NEW(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_EODN:
      if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
        could_continue = TRUE;
      else if (clen == 0 || (IS_NEWLINE(ptr) && ptr == end_subject - md->nllen))
        { ADD_ACTIVE(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_DOLL:
      if ((md->moptions & PCRE_NOTEOL) == 0)
        {
        if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
          could_continue = TRUE;
        else if (clen == 0 ||
            ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&
               (ptr == end_subject - md->nllen)
            ))
          { ADD_ACTIVE(state_offset + 1, 0); }
        else if (ptr + 1 >= md->end_subject &&
                 (md->moptions & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) != 0 &&
                 NLBLOCK->nltype == NLTYPE_FIXED &&
                 NLBLOCK->nllen == 2 &&
                 c == NLBLOCK->nl[0])
          {
          if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
            {
            reset_could_continue = TRUE;
            ADD_NEW_DATA(-(state_offset + 1), 0, 1);
            }
          else could_continue = partial_newline = TRUE;
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_DOLLM:
      if ((md->moptions & PCRE_NOTEOL) == 0)
        {
        if (clen == 0 && (md->moptions & PCRE_PARTIAL_HARD) != 0)
          could_continue = TRUE;
        else if (clen == 0 ||
            ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr)))
          { ADD_ACTIVE(state_offset + 1, 0); }
        else if (ptr + 1 >= md->end_subject &&
                 (md->moptions & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) != 0 &&
                 NLBLOCK->nltype == NLTYPE_FIXED &&
                 NLBLOCK->nllen == 2 &&
                 c == NLBLOCK->nl[0])
          {
          if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
            {
            reset_could_continue = TRUE;
            ADD_NEW_DATA(-(state_offset + 1), 0, 1);
            }
          else could_continue = partial_newline = TRUE;
          }
        }
      else if (IS_NEWLINE(ptr))
        { ADD_ACTIVE(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/

      case OP_DIGIT:
      case OP_WHITESPACE:
      case OP_WORDCHAR:
      if (clen > 0 && c < 256 &&
            ((ctypes[c] & toptable1[codevalue]) ^ toptable2[codevalue]) != 0)
        { ADD_NEW(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_NOT_DIGIT:
      case OP_NOT_WHITESPACE:
      case OP_NOT_WORDCHAR:
      if (clen > 0 && (c >= 256 ||
            ((ctypes[c] & toptable1[codevalue]) ^ toptable2[codevalue]) != 0))
        { ADD_NEW(state_offset + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_WORD_BOUNDARY:
      case OP_NOT_WORD_BOUNDARY:
        {
        int left_word, right_word;

        if (ptr > start_subject)
          {
          const pcre_uchar *temp = ptr - 1;
          if (temp < md->start_used_ptr) md->start_used_ptr = temp;
#if defined SUPPORT_UTF && !defined COMPILE_PCRE32
          if (utf) { BACKCHAR(temp); }
#endif
          GETCHARTEST(d, temp);
#ifdef SUPPORT_UCP
          if ((md->poptions & PCRE_UCP) != 0)
            {
            if (d == '_') left_word = TRUE; else
              {
              int cat = UCD_CATEGORY(d);
              left_word = (cat == ucp_L || cat == ucp_N);
              }
            }
          else
#endif
          left_word = d < 256 && (ctypes[d] & ctype_word) != 0;
          }
        else left_word = FALSE;

        if (clen > 0)
          {
#ifdef SUPPORT_UCP
          if ((md->poptions & PCRE_UCP) != 0)
            {
            if (c == '_') right_word = TRUE; else
              {
              int cat = UCD_CATEGORY(c);
              right_word = (cat == ucp_L || cat == ucp_N);
              }
            }
          else
#endif
          right_word = c < 256 && (ctypes[c] & ctype_word) != 0;
          }
        else right_word = FALSE;

        if ((left_word == right_word) == (codevalue == OP_NOT_WORD_BOUNDARY))
          { ADD_ACTIVE(state_offset + 1, 0); }
        }
      break;


      /*-----------------------------------------------------------------*/
      /* Check the next character by Unicode property. We will get here only
      if the support is in the binary; otherwise a compile-time error occurs.
      */

#ifdef SUPPORT_UCP
      case OP_PROP:
      case OP_NOTPROP:
      if (clen > 0)
        {
        BOOL OK;
        const pcre_uint32 *cp;
        const ucd_record * prop = GET_UCD(c);
        switch(code[1])
          {
          case PT_ANY:
          OK = TRUE;
          break;

          case PT_LAMP:
          OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
               prop->chartype == ucp_Lt;
          break;

          case PT_GC:
          OK = PRIV(ucp_gentype)[prop->chartype] == code[2];
          break;

          case PT_PC:
          OK = prop->chartype == code[2];
          break;

          case PT_SC:
          OK = prop->script == code[2];
          break;

          /* These are specials for combination cases. */

          case PT_ALNUM:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N;
          break;

          case PT_SPACE:    /* Perl space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_PXSPACE:  /* POSIX space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
               c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_WORD:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
               c == CHAR_UNDERSCORE;
          break;

          case PT_CLIST:
          cp = PRIV(ucd_caseless_sets) + code[2];
          for (;;)
            {
            if (c < *cp) { OK = FALSE; break; }
            if (c == *cp++) { OK = TRUE; break; }
            }
          break;

          /* Should never occur, but keep compilers from grumbling. */

          default:
          OK = codevalue != OP_PROP;
          break;
          }

        if (OK == (codevalue == OP_PROP)) { ADD_NEW(state_offset + 3, 0); }
        }
      break;
#endif



/* ========================================================================== */
      /* These opcodes likewise inspect the subject character, but have an
      argument that is not a data character. It is one of these opcodes:
      OP_ANY, OP_ALLANY, OP_DIGIT, OP_NOT_DIGIT, OP_WHITESPACE, OP_NOT_SPACE,
      OP_WORDCHAR, OP_NOT_WORDCHAR. The value is loaded into d. */

      case OP_TYPEPLUS:
      case OP_TYPEMINPLUS:
      case OP_TYPEPOSPLUS:
      count = current_state->count;  /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
      if (clen > 0)
        {
        if (d == OP_ANY && ptr + 1 >= md->end_subject &&
            (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
            NLBLOCK->nltype == NLTYPE_FIXED &&
            NLBLOCK->nllen == 2 &&
            c == NLBLOCK->nl[0])
          {
          could_continue = partial_newline = TRUE;
          }
        else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
            (c < 256 &&
              (d != OP_ANY || !IS_NEWLINE(ptr)) &&
              ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
          {
          if (count > 0 && codevalue == OP_TYPEPOSPLUS)
            {
            active_count--;            /* Remove non-match possibility */
            next_active_state--;
            }
          count++;
          ADD_NEW(state_offset, count);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_TYPEQUERY:
      case OP_TYPEMINQUERY:
      case OP_TYPEPOSQUERY:
      ADD_ACTIVE(state_offset + 2, 0);
      if (clen > 0)
        {
        if (d == OP_ANY && ptr + 1 >= md->end_subject &&
            (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
            NLBLOCK->nltype == NLTYPE_FIXED &&
            NLBLOCK->nllen == 2 &&
            c == NLBLOCK->nl[0])
          {
          could_continue = partial_newline = TRUE;
          }
        else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
            (c < 256 &&
              (d != OP_ANY || !IS_NEWLINE(ptr)) &&
              ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
          {
          if (codevalue == OP_TYPEPOSQUERY)
            {
            active_count--;            /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW(state_offset + 2, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_TYPESTAR:
      case OP_TYPEMINSTAR:
      case OP_TYPEPOSSTAR:
      ADD_ACTIVE(state_offset + 2, 0);
      if (clen > 0)
        {
        if (d == OP_ANY && ptr + 1 >= md->end_subject &&
            (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
            NLBLOCK->nltype == NLTYPE_FIXED &&
            NLBLOCK->nllen == 2 &&
            c == NLBLOCK->nl[0])
          {
          could_continue = partial_newline = TRUE;
          }
        else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
            (c < 256 &&
              (d != OP_ANY || !IS_NEWLINE(ptr)) &&
              ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
          {
          if (codevalue == OP_TYPEPOSSTAR)
            {
            active_count--;            /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW(state_offset, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_TYPEEXACT:
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        if (d == OP_ANY && ptr + 1 >= md->end_subject &&
            (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
            NLBLOCK->nltype == NLTYPE_FIXED &&
            NLBLOCK->nllen == 2 &&
            c == NLBLOCK->nl[0])
          {
          could_continue = partial_newline = TRUE;
          }
        else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
            (c < 256 &&
              (d != OP_ANY || !IS_NEWLINE(ptr)) &&
              ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
          {
          if (++count >= GET2(code, 1))
            { ADD_NEW(state_offset + 1 + IMM2_SIZE + 1, 0); }
          else
            { ADD_NEW(state_offset, count); }
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_TYPEUPTO:
      case OP_TYPEMINUPTO:
      case OP_TYPEPOSUPTO:
      ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0);
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        if (d == OP_ANY && ptr + 1 >= md->end_subject &&
            (md->moptions & (PCRE_PARTIAL_HARD)) != 0 &&
            NLBLOCK->nltype == NLTYPE_FIXED &&
            NLBLOCK->nllen == 2 &&
            c == NLBLOCK->nl[0])
          {
          could_continue = partial_newline = TRUE;
          }
        else if ((c >= 256 && d != OP_DIGIT && d != OP_WHITESPACE && d != OP_WORDCHAR) ||
            (c < 256 &&
              (d != OP_ANY || !IS_NEWLINE(ptr)) &&
              ((ctypes[c] & toptable1[d]) ^ toptable2[d]) != 0))
          {
          if (codevalue == OP_TYPEPOSUPTO)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          if (++count >= GET2(code, 1))
            { ADD_NEW(state_offset + 2 + IMM2_SIZE, 0); }
          else
            { ADD_NEW(state_offset, count); }
          }
        }
      break;

/* ========================================================================== */
      /* These are virtual opcodes that are used when something like
      OP_TYPEPLUS has OP_PROP, OP_NOTPROP, OP_ANYNL, or OP_EXTUNI as its
      argument. It keeps the code above fast for the other cases. The argument
      is in the d variable. */

#ifdef SUPPORT_UCP
      case OP_PROP_EXTRA + OP_TYPEPLUS:
      case OP_PROP_EXTRA + OP_TYPEMINPLUS:
      case OP_PROP_EXTRA + OP_TYPEPOSPLUS:
      count = current_state->count;           /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + 4, 0); }
      if (clen > 0)
        {
        BOOL OK;
        const pcre_uint32 *cp;
        const ucd_record * prop = GET_UCD(c);
        switch(code[2])
          {
          case PT_ANY:
          OK = TRUE;
          break;

          case PT_LAMP:
          OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
            prop->chartype == ucp_Lt;
          break;

          case PT_GC:
          OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
          break;

          case PT_PC:
          OK = prop->chartype == code[3];
          break;

          case PT_SC:
          OK = prop->script == code[3];
          break;

          /* These are specials for combination cases. */

          case PT_ALNUM:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N;
          break;

          case PT_SPACE:    /* Perl space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_PXSPACE:  /* POSIX space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
               c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_WORD:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
               c == CHAR_UNDERSCORE;
          break;

          case PT_CLIST:
          cp = PRIV(ucd_caseless_sets) + code[3];
          for (;;)
            {
            if (c < *cp) { OK = FALSE; break; }
            if (c == *cp++) { OK = TRUE; break; }
            }
          break;

          /* Should never occur, but keep compilers from grumbling. */

          default:
          OK = codevalue != OP_PROP;
          break;
          }

        if (OK == (d == OP_PROP))
          {
          if (count > 0 && codevalue == OP_PROP_EXTRA + OP_TYPEPOSPLUS)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          count++;
          ADD_NEW(state_offset, count);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_EXTUNI_EXTRA + OP_TYPEPLUS:
      case OP_EXTUNI_EXTRA + OP_TYPEMINPLUS:
      case OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS:
      count = current_state->count;  /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
      if (clen > 0)
        {
        int lgb, rgb;
        const pcre_uchar *nptr = ptr + clen;
        int ncount = 0;
        if (count > 0 && codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS)
          {
          active_count--;           /* Remove non-match possibility */
          next_active_state--;
          }
        lgb = UCD_GRAPHBREAK(c);
        while (nptr < end_subject)
          {
          dlen = 1;
          if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
          rgb = UCD_GRAPHBREAK(d);
          if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
          ncount++;
          lgb = rgb;
          nptr += dlen;
          }
        count++;
        ADD_NEW_DATA(-state_offset, count, ncount);
        }
      break;
#endif

      /*-----------------------------------------------------------------*/
      case OP_ANYNL_EXTRA + OP_TYPEPLUS:
      case OP_ANYNL_EXTRA + OP_TYPEMINPLUS:
      case OP_ANYNL_EXTRA + OP_TYPEPOSPLUS:
      count = current_state->count;  /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
      if (clen > 0)
        {
        int ncount = 0;
        switch (c)
          {
          case CHAR_VT:
          case CHAR_FF:
          case CHAR_NEL:
#ifndef EBCDIC
          case 0x2028:
          case 0x2029:
#endif  /* Not EBCDIC */
          if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
          goto ANYNL01;

          case CHAR_CR:
          if (ptr + 1 < end_subject && RAWUCHARTEST(ptr + 1) == CHAR_LF) ncount = 1;
          /* Fall through */

          ANYNL01:
          case CHAR_LF:
          if (count > 0 && codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSPLUS)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          count++;
          ADD_NEW_DATA(-state_offset, count, ncount);
          break;

          default:
          break;
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_VSPACE_EXTRA + OP_TYPEPLUS:
      case OP_VSPACE_EXTRA + OP_TYPEMINPLUS:
      case OP_VSPACE_EXTRA + OP_TYPEPOSPLUS:
      count = current_state->count;  /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
      if (clen > 0)
        {
        BOOL OK;
        switch (c)
          {
          VSPACE_CASES:
          OK = TRUE;
          break;

          default:
          OK = FALSE;
          break;
          }

        if (OK == (d == OP_VSPACE))
          {
          if (count > 0 && codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSPLUS)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          count++;
          ADD_NEW_DATA(-state_offset, count, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_HSPACE_EXTRA + OP_TYPEPLUS:
      case OP_HSPACE_EXTRA + OP_TYPEMINPLUS:
      case OP_HSPACE_EXTRA + OP_TYPEPOSPLUS:
      count = current_state->count;  /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
      if (clen > 0)
        {
        BOOL OK;
        switch (c)
          {
          HSPACE_CASES:
          OK = TRUE;
          break;

          default:
          OK = FALSE;
          break;
          }

        if (OK == (d == OP_HSPACE))
          {
          if (count > 0 && codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSPLUS)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          count++;
          ADD_NEW_DATA(-state_offset, count, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
#ifdef SUPPORT_UCP
      case OP_PROP_EXTRA + OP_TYPEQUERY:
      case OP_PROP_EXTRA + OP_TYPEMINQUERY:
      case OP_PROP_EXTRA + OP_TYPEPOSQUERY:
      count = 4;
      goto QS1;

      case OP_PROP_EXTRA + OP_TYPESTAR:
      case OP_PROP_EXTRA + OP_TYPEMINSTAR:
      case OP_PROP_EXTRA + OP_TYPEPOSSTAR:
      count = 0;

      QS1:

      ADD_ACTIVE(state_offset + 4, 0);
      if (clen > 0)
        {
        BOOL OK;
        const pcre_uint32 *cp;
        const ucd_record * prop = GET_UCD(c);
        switch(code[2])
          {
          case PT_ANY:
          OK = TRUE;
          break;

          case PT_LAMP:
          OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
            prop->chartype == ucp_Lt;
          break;

          case PT_GC:
          OK = PRIV(ucp_gentype)[prop->chartype] == code[3];
          break;

          case PT_PC:
          OK = prop->chartype == code[3];
          break;

          case PT_SC:
          OK = prop->script == code[3];
          break;

          /* These are specials for combination cases. */

          case PT_ALNUM:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N;
          break;

          case PT_SPACE:    /* Perl space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_PXSPACE:  /* POSIX space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
               c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_WORD:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
               c == CHAR_UNDERSCORE;
          break;

          case PT_CLIST:
          cp = PRIV(ucd_caseless_sets) + code[3];
          for (;;)
            {
            if (c < *cp) { OK = FALSE; break; }
            if (c == *cp++) { OK = TRUE; break; }
            }
          break;

          /* Should never occur, but keep compilers from grumbling. */

          default:
          OK = codevalue != OP_PROP;
          break;
          }

        if (OK == (d == OP_PROP))
          {
          if (codevalue == OP_PROP_EXTRA + OP_TYPEPOSSTAR ||
              codevalue == OP_PROP_EXTRA + OP_TYPEPOSQUERY)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW(state_offset + count, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_EXTUNI_EXTRA + OP_TYPEQUERY:
      case OP_EXTUNI_EXTRA + OP_TYPEMINQUERY:
      case OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY:
      count = 2;
      goto QS2;

      case OP_EXTUNI_EXTRA + OP_TYPESTAR:
      case OP_EXTUNI_EXTRA + OP_TYPEMINSTAR:
      case OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR:
      count = 0;

      QS2:

      ADD_ACTIVE(state_offset + 2, 0);
      if (clen > 0)
        {
        int lgb, rgb;
        const pcre_uchar *nptr = ptr + clen;
        int ncount = 0;
        if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSSTAR ||
            codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSQUERY)
          {
          active_count--;           /* Remove non-match possibility */
          next_active_state--;
          }
        lgb = UCD_GRAPHBREAK(c);
        while (nptr < end_subject)
          {
          dlen = 1;
          if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
          rgb = UCD_GRAPHBREAK(d);
          if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
          ncount++;
          lgb = rgb;
          nptr += dlen;
          }
        ADD_NEW_DATA(-(state_offset + count), 0, ncount);
        }
      break;
#endif

      /*-----------------------------------------------------------------*/
      case OP_ANYNL_EXTRA + OP_TYPEQUERY:
      case OP_ANYNL_EXTRA + OP_TYPEMINQUERY:
      case OP_ANYNL_EXTRA + OP_TYPEPOSQUERY:
      count = 2;
      goto QS3;

      case OP_ANYNL_EXTRA + OP_TYPESTAR:
      case OP_ANYNL_EXTRA + OP_TYPEMINSTAR:
      case OP_ANYNL_EXTRA + OP_TYPEPOSSTAR:
      count = 0;

      QS3:
      ADD_ACTIVE(state_offset + 2, 0);
      if (clen > 0)
        {
        int ncount = 0;
        switch (c)
          {
          case CHAR_VT:
          case CHAR_FF:
          case CHAR_NEL:
#ifndef EBCDIC
          case 0x2028:
          case 0x2029:
#endif  /* Not EBCDIC */
          if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
          goto ANYNL02;

          case CHAR_CR:
          if (ptr + 1 < end_subject && RAWUCHARTEST(ptr + 1) == CHAR_LF) ncount = 1;
          /* Fall through */

          ANYNL02:
          case CHAR_LF:
          if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSSTAR ||
              codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSQUERY)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW_DATA(-(state_offset + count), 0, ncount);
          break;

          default:
          break;
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_VSPACE_EXTRA + OP_TYPEQUERY:
      case OP_VSPACE_EXTRA + OP_TYPEMINQUERY:
      case OP_VSPACE_EXTRA + OP_TYPEPOSQUERY:
      count = 2;
      goto QS4;

      case OP_VSPACE_EXTRA + OP_TYPESTAR:
      case OP_VSPACE_EXTRA + OP_TYPEMINSTAR:
      case OP_VSPACE_EXTRA + OP_TYPEPOSSTAR:
      count = 0;

      QS4:
      ADD_ACTIVE(state_offset + 2, 0);
      if (clen > 0)
        {
        BOOL OK;
        switch (c)
          {
          VSPACE_CASES:
          OK = TRUE;
          break;

          default:
          OK = FALSE;
          break;
          }
        if (OK == (d == OP_VSPACE))
          {
          if (codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSSTAR ||
              codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSQUERY)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW_DATA(-(state_offset + count), 0, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_HSPACE_EXTRA + OP_TYPEQUERY:
      case OP_HSPACE_EXTRA + OP_TYPEMINQUERY:
      case OP_HSPACE_EXTRA + OP_TYPEPOSQUERY:
      count = 2;
      goto QS5;

      case OP_HSPACE_EXTRA + OP_TYPESTAR:
      case OP_HSPACE_EXTRA + OP_TYPEMINSTAR:
      case OP_HSPACE_EXTRA + OP_TYPEPOSSTAR:
      count = 0;

      QS5:
      ADD_ACTIVE(state_offset + 2, 0);
      if (clen > 0)
        {
        BOOL OK;
        switch (c)
          {
          HSPACE_CASES:
          OK = TRUE;
          break;

          default:
          OK = FALSE;
          break;
          }

        if (OK == (d == OP_HSPACE))
          {
          if (codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSSTAR ||
              codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSQUERY)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW_DATA(-(state_offset + count), 0, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
#ifdef SUPPORT_UCP
      case OP_PROP_EXTRA + OP_TYPEEXACT:
      case OP_PROP_EXTRA + OP_TYPEUPTO:
      case OP_PROP_EXTRA + OP_TYPEMINUPTO:
      case OP_PROP_EXTRA + OP_TYPEPOSUPTO:
      if (codevalue != OP_PROP_EXTRA + OP_TYPEEXACT)
        { ADD_ACTIVE(state_offset + 1 + IMM2_SIZE + 3, 0); }
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        BOOL OK;
        const pcre_uint32 *cp;
        const ucd_record * prop = GET_UCD(c);
        switch(code[1 + IMM2_SIZE + 1])
          {
          case PT_ANY:
          OK = TRUE;
          break;

          case PT_LAMP:
          OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll ||
            prop->chartype == ucp_Lt;
          break;

          case PT_GC:
          OK = PRIV(ucp_gentype)[prop->chartype] == code[1 + IMM2_SIZE + 2];
          break;

          case PT_PC:
          OK = prop->chartype == code[1 + IMM2_SIZE + 2];
          break;

          case PT_SC:
          OK = prop->script == code[1 + IMM2_SIZE + 2];
          break;

          /* These are specials for combination cases. */

          case PT_ALNUM:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N;
          break;

          case PT_SPACE:    /* Perl space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_PXSPACE:  /* POSIX space */
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
               c == CHAR_FF || c == CHAR_CR;
          break;

          case PT_WORD:
          OK = PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
               c == CHAR_UNDERSCORE;
          break;

          case PT_CLIST:
          cp = PRIV(ucd_caseless_sets) + code[1 + IMM2_SIZE + 2];
          for (;;)
            {
            if (c < *cp) { OK = FALSE; break; }
            if (c == *cp++) { OK = TRUE; break; }
            }
          break;

          /* Should never occur, but keep compilers from grumbling. */

          default:
          OK = codevalue != OP_PROP;
          break;
          }

        if (OK == (d == OP_PROP))
          {
          if (codevalue == OP_PROP_EXTRA + OP_TYPEPOSUPTO)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          if (++count >= GET2(code, 1))
            { ADD_NEW(state_offset + 1 + IMM2_SIZE + 3, 0); }
          else
            { ADD_NEW(state_offset, count); }
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_EXTUNI_EXTRA + OP_TYPEEXACT:
      case OP_EXTUNI_EXTRA + OP_TYPEUPTO:
      case OP_EXTUNI_EXTRA + OP_TYPEMINUPTO:
      case OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO:
      if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)
        { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        int lgb, rgb;
        const pcre_uchar *nptr = ptr + clen;
        int ncount = 0;
        if (codevalue == OP_EXTUNI_EXTRA + OP_TYPEPOSUPTO)
          {
          active_count--;           /* Remove non-match possibility */
          next_active_state--;
          }
        lgb = UCD_GRAPHBREAK(c);
        while (nptr < end_subject)
          {
          dlen = 1;
          if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
          rgb = UCD_GRAPHBREAK(d);
          if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
          ncount++;
          lgb = rgb;
          nptr += dlen;
          }
        if (nptr >= end_subject && (md->moptions & PCRE_PARTIAL_HARD) != 0)
            reset_could_continue = TRUE;
        if (++count >= GET2(code, 1))
          { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
        else
          { ADD_NEW_DATA(-state_offset, count, ncount); }
        }
      break;
#endif

      /*-----------------------------------------------------------------*/
      case OP_ANYNL_EXTRA + OP_TYPEEXACT:
      case OP_ANYNL_EXTRA + OP_TYPEUPTO:
      case OP_ANYNL_EXTRA + OP_TYPEMINUPTO:
      case OP_ANYNL_EXTRA + OP_TYPEPOSUPTO:
      if (codevalue != OP_ANYNL_EXTRA + OP_TYPEEXACT)
        { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        int ncount = 0;
        switch (c)
          {
          case CHAR_VT:
          case CHAR_FF:
          case CHAR_NEL:
#ifndef EBCDIC
          case 0x2028:
          case 0x2029:
#endif  /* Not EBCDIC */
          if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;
          goto ANYNL03;

          case CHAR_CR:
          if (ptr + 1 < end_subject && RAWUCHARTEST(ptr + 1) == CHAR_LF) ncount = 1;
          /* Fall through */

          ANYNL03:
          case CHAR_LF:
          if (codevalue == OP_ANYNL_EXTRA + OP_TYPEPOSUPTO)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          if (++count >= GET2(code, 1))
            { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, ncount); }
          else
            { ADD_NEW_DATA(-state_offset, count, ncount); }
          break;

          default:
          break;
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_VSPACE_EXTRA + OP_TYPEEXACT:
      case OP_VSPACE_EXTRA + OP_TYPEUPTO:
      case OP_VSPACE_EXTRA + OP_TYPEMINUPTO:
      case OP_VSPACE_EXTRA + OP_TYPEPOSUPTO:
      if (codevalue != OP_VSPACE_EXTRA + OP_TYPEEXACT)
        { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        BOOL OK;
        switch (c)
          {
          VSPACE_CASES:
          OK = TRUE;
          break;

          default:
          OK = FALSE;
          }

        if (OK == (d == OP_VSPACE))
          {
          if (codevalue == OP_VSPACE_EXTRA + OP_TYPEPOSUPTO)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          if (++count >= GET2(code, 1))
            { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
          else
            { ADD_NEW_DATA(-state_offset, count, 0); }
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_HSPACE_EXTRA + OP_TYPEEXACT:
      case OP_HSPACE_EXTRA + OP_TYPEUPTO:
      case OP_HSPACE_EXTRA + OP_TYPEMINUPTO:
      case OP_HSPACE_EXTRA + OP_TYPEPOSUPTO:
      if (codevalue != OP_HSPACE_EXTRA + OP_TYPEEXACT)
        { ADD_ACTIVE(state_offset + 2 + IMM2_SIZE, 0); }
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        BOOL OK;
        switch (c)
          {
          HSPACE_CASES:
          OK = TRUE;
          break;

          default:
          OK = FALSE;
          break;
          }

        if (OK == (d == OP_HSPACE))
          {
          if (codevalue == OP_HSPACE_EXTRA + OP_TYPEPOSUPTO)
            {
            active_count--;           /* Remove non-match possibility */
            next_active_state--;
            }
          if (++count >= GET2(code, 1))
            { ADD_NEW_DATA(-(state_offset + 2 + IMM2_SIZE), 0, 0); }
          else
            { ADD_NEW_DATA(-state_offset, count, 0); }
          }
        }
      break;

/* ========================================================================== */
      /* These opcodes are followed by a character that is usually compared
      to the current subject character; it is loaded into d. We still get
      here even if there is no subject character, because in some cases zero
      repetitions are permitted. */

      /*-----------------------------------------------------------------*/
      case OP_CHAR:
      if (clen > 0 && c == d) { ADD_NEW(state_offset + dlen + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      case OP_CHARI:
      if (clen == 0) break;

#ifdef SUPPORT_UTF
      if (utf)
        {
        if (c == d) { ADD_NEW(state_offset + dlen + 1, 0); } else
          {
          unsigned int othercase;
          if (c < 128)
            othercase = fcc[c];
          else
            /* If we have Unicode property support, we can use it to test the
            other case of the character. */
#ifdef SUPPORT_UCP
            othercase = UCD_OTHERCASE(c);
#else
            othercase = NOTACHAR;
#endif

          if (d == othercase) { ADD_NEW(state_offset + dlen + 1, 0); }
          }
        }
      else
#endif  /* SUPPORT_UTF */
      /* Not UTF mode */
        {
        if (TABLE_GET(c, lcc, c) == TABLE_GET(d, lcc, d))
          { ADD_NEW(state_offset + 2, 0); }
        }
      break;


#ifdef SUPPORT_UCP
      /*-----------------------------------------------------------------*/
      /* This is a tricky one because it can match more than one character.
      Find out how many characters to skip, and then set up a negative state
      to wait for them to pass before continuing. */

      case OP_EXTUNI:
      if (clen > 0)
        {
        int lgb, rgb;
        const pcre_uchar *nptr = ptr + clen;
        int ncount = 0;
        lgb = UCD_GRAPHBREAK(c);
        while (nptr < end_subject)
          {
          dlen = 1;
          if (!utf) d = *nptr; else { GETCHARLEN(d, nptr, dlen); }
          rgb = UCD_GRAPHBREAK(d);
          if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
          ncount++;
          lgb = rgb;
          nptr += dlen;
          }
        if (nptr >= end_subject && (md->moptions & PCRE_PARTIAL_HARD) != 0)
            reset_could_continue = TRUE;
        ADD_NEW_DATA(-(state_offset + 1), 0, ncount);
        }
      break;
#endif

      /*-----------------------------------------------------------------*/
      /* This is a tricky like EXTUNI because it too can match more than one
      character (when CR is followed by LF). In this case, set up a negative
      state to wait for one character to pass before continuing. */

      case OP_ANYNL:
      if (clen > 0) switch(c)
        {
        case CHAR_VT:
        case CHAR_FF:
        case CHAR_NEL:
#ifndef EBCDIC
        case 0x2028:
        case 0x2029:
#endif  /* Not EBCDIC */
        if ((md->moptions & PCRE_BSR_ANYCRLF) != 0) break;

        case CHAR_LF:
        ADD_NEW(state_offset + 1, 0);
        break;

        case CHAR_CR:
        if (ptr + 1 >= end_subject)
          {
          ADD_NEW(state_offset + 1, 0);
          if ((md->moptions & PCRE_PARTIAL_HARD) != 0)
            reset_could_continue = TRUE;
          }
        else if (RAWUCHARTEST(ptr + 1) == CHAR_LF)
          {
          ADD_NEW_DATA(-(state_offset + 1), 0, 1);
          }
        else
          {
          ADD_NEW(state_offset + 1, 0);
          }
        break;
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_NOT_VSPACE:
      if (clen > 0) switch(c)
        {
        VSPACE_CASES:
        break;

        default:
        ADD_NEW(state_offset + 1, 0);
        break;
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_VSPACE:
      if (clen > 0) switch(c)
        {
        VSPACE_CASES:
        ADD_NEW(state_offset + 1, 0);
        break;

        default:
        break;
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_NOT_HSPACE:
      if (clen > 0) switch(c)
        {
        HSPACE_CASES:
        break;

        default:
        ADD_NEW(state_offset + 1, 0);
        break;
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_HSPACE:
      if (clen > 0) switch(c)
        {
        HSPACE_CASES:
        ADD_NEW(state_offset + 1, 0);
        break;

        default:
        break;
        }
      break;

      /*-----------------------------------------------------------------*/
      /* Match a negated single character casefully. */

      case OP_NOT:
      if (clen > 0 && c != d) { ADD_NEW(state_offset + dlen + 1, 0); }
      break;

      /*-----------------------------------------------------------------*/
      /* Match a negated single character caselessly. */

      case OP_NOTI:
      if (clen > 0)
        {
        unsigned int otherd;
#ifdef SUPPORT_UTF
        if (utf && d >= 128)
          {
#ifdef SUPPORT_UCP
          otherd = UCD_OTHERCASE(d);
#endif  /* SUPPORT_UCP */
          }
        else
#endif  /* SUPPORT_UTF */
        otherd = TABLE_GET(d, fcc, d);
        if (c != d && c != otherd)
          { ADD_NEW(state_offset + dlen + 1, 0); }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_PLUSI:
      case OP_MINPLUSI:
      case OP_POSPLUSI:
      case OP_NOTPLUSI:
      case OP_NOTMINPLUSI:
      case OP_NOTPOSPLUSI:
      caseless = TRUE;
      codevalue -= OP_STARI - OP_STAR;

      /* Fall through */
      case OP_PLUS:
      case OP_MINPLUS:
      case OP_POSPLUS:
      case OP_NOTPLUS:
      case OP_NOTMINPLUS:
      case OP_NOTPOSPLUS:
      count = current_state->count;  /* Already matched */
      if (count > 0) { ADD_ACTIVE(state_offset + dlen + 1, 0); }
      if (clen > 0)
        {
        pcre_uint32 otherd = NOTACHAR;
        if (caseless)
          {
#ifdef SUPPORT_UTF
          if (utf && d >= 128)
            {
#ifdef SUPPORT_UCP
            otherd = UCD_OTHERCASE(d);
#endif  /* SUPPORT_UCP */
            }
          else
#endif  /* SUPPORT_UTF */
          otherd = TABLE_GET(d, fcc, d);
          }
        if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
          {
          if (count > 0 &&
              (codevalue == OP_POSPLUS || codevalue == OP_NOTPOSPLUS))
            {
            active_count--;             /* Remove non-match possibility */
            next_active_state--;
            }
          count++;
          ADD_NEW(state_offset, count);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_QUERYI:
      case OP_MINQUERYI:
      case OP_POSQUERYI:
      case OP_NOTQUERYI:
      case OP_NOTMINQUERYI:
      case OP_NOTPOSQUERYI:
      caseless = TRUE;
      codevalue -= OP_STARI - OP_STAR;
      /* Fall through */
      case OP_QUERY:
      case OP_MINQUERY:
      case OP_POSQUERY:
      case OP_NOTQUERY:
      case OP_NOTMINQUERY:
      case OP_NOTPOSQUERY:
      ADD_ACTIVE(state_offset + dlen + 1, 0);
      if (clen > 0)
        {
        pcre_uint32 otherd = NOTACHAR;
        if (caseless)
          {
#ifdef SUPPORT_UTF
          if (utf && d >= 128)
            {
#ifdef SUPPORT_UCP
            otherd = UCD_OTHERCASE(d);
#endif  /* SUPPORT_UCP */
            }
          else
#endif  /* SUPPORT_UTF */
          otherd = TABLE_GET(d, fcc, d);
          }
        if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
          {
          if (codevalue == OP_POSQUERY || codevalue == OP_NOTPOSQUERY)
            {
            active_count--;            /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW(state_offset + dlen + 1, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_STARI:
      case OP_MINSTARI:
      case OP_POSSTARI:
      case OP_NOTSTARI:
      case OP_NOTMINSTARI:
      case OP_NOTPOSSTARI:
      caseless = TRUE;
      codevalue -= OP_STARI - OP_STAR;
      /* Fall through */
      case OP_STAR:
      case OP_MINSTAR:
      case OP_POSSTAR:
      case OP_NOTSTAR:
      case OP_NOTMINSTAR:
      case OP_NOTPOSSTAR:
      ADD_ACTIVE(state_offset + dlen + 1, 0);
      if (clen > 0)
        {
        pcre_uint32 otherd = NOTACHAR;
        if (caseless)
          {
#ifdef SUPPORT_UTF
          if (utf && d >= 128)
            {
#ifdef SUPPORT_UCP
            otherd = UCD_OTHERCASE(d);
#endif  /* SUPPORT_UCP */
            }
          else
#endif  /* SUPPORT_UTF */
          otherd = TABLE_GET(d, fcc, d);
          }
        if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
          {
          if (codevalue == OP_POSSTAR || codevalue == OP_NOTPOSSTAR)
            {
            active_count--;            /* Remove non-match possibility */
            next_active_state--;
            }
          ADD_NEW(state_offset, 0);
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_EXACTI:
      case OP_NOTEXACTI:
      caseless = TRUE;
      codevalue -= OP_STARI - OP_STAR;
      /* Fall through */
      case OP_EXACT:
      case OP_NOTEXACT:
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        pcre_uint32 otherd = NOTACHAR;
        if (caseless)
          {
#ifdef SUPPORT_UTF
          if (utf && d >= 128)
            {
#ifdef SUPPORT_UCP
            otherd = UCD_OTHERCASE(d);
#endif  /* SUPPORT_UCP */
            }
          else
#endif  /* SUPPORT_UTF */
          otherd = TABLE_GET(d, fcc, d);
          }
        if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
          {
          if (++count >= GET2(code, 1))
            { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
          else
            { ADD_NEW(state_offset, count); }
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_UPTOI:
      case OP_MINUPTOI:
      case OP_POSUPTOI:
      case OP_NOTUPTOI:
      case OP_NOTMINUPTOI:
      case OP_NOTPOSUPTOI:
      caseless = TRUE;
      codevalue -= OP_STARI - OP_STAR;
      /* Fall through */
      case OP_UPTO:
      case OP_MINUPTO:
      case OP_POSUPTO:
      case OP_NOTUPTO:
      case OP_NOTMINUPTO:
      case OP_NOTPOSUPTO:
      ADD_ACTIVE(state_offset + dlen + 1 + IMM2_SIZE, 0);
      count = current_state->count;  /* Number already matched */
      if (clen > 0)
        {
        pcre_uint32 otherd = NOTACHAR;
        if (caseless)
          {
#ifdef SUPPORT_UTF
          if (utf && d >= 128)
            {
#ifdef SUPPORT_UCP
            otherd = UCD_OTHERCASE(d);
#endif  /* SUPPORT_UCP */
            }
          else
#endif  /* SUPPORT_UTF */
          otherd = TABLE_GET(d, fcc, d);
          }
        if ((c == d || c == otherd) == (codevalue < OP_NOTSTAR))
          {
          if (codevalue == OP_POSUPTO || codevalue == OP_NOTPOSUPTO)
            {
            active_count--;             /* Remove non-match possibility */
            next_active_state--;
            }
          if (++count >= GET2(code, 1))
            { ADD_NEW(state_offset + dlen + 1 + IMM2_SIZE, 0); }
          else
            { ADD_NEW(state_offset, count); }
          }
        }
      break;


/* ========================================================================== */
      /* These are the class-handling opcodes */

      case OP_CLASS:
      case OP_NCLASS:
      case OP_XCLASS:
        {
        BOOL isinclass = FALSE;
        int next_state_offset;
        const pcre_uchar *ecode;

        /* For a simple class, there is always just a 32-byte table, and we
        can set isinclass from it. */

        if (codevalue != OP_XCLASS)
          {
          ecode = code + 1 + (32 / sizeof(pcre_uchar));
          if (clen > 0)
            {
            isinclass = (c > 255)? (codevalue == OP_NCLASS) :
              ((((pcre_uint8 *)(code + 1))[c/8] & (1 << (c&7))) != 0);
            }
          }

        /* An extended class may have a table or a list of single characters,
        ranges, or both, and it may be positive or negative. There's a
        function that sorts all this out. */

        else
         {
         ecode = code + GET(code, 1);
         if (clen > 0) isinclass = PRIV(xclass)(c, code + 1 + LINK_SIZE, utf);
         }

        /* At this point, isinclass is set for all kinds of class, and ecode
        points to the byte after the end of the class. If there is a
        quantifier, this is where it will be. */

        next_state_offset = (int)(ecode - start_code);

        switch (*ecode)
          {
          case OP_CRSTAR:
          case OP_CRMINSTAR:
          ADD_ACTIVE(next_state_offset + 1, 0);
          if (isinclass) { ADD_NEW(state_offset, 0); }
          break;

          case OP_CRPLUS:
          case OP_CRMINPLUS:
          count = current_state->count;  /* Already matched */
          if (count > 0) { ADD_ACTIVE(next_state_offset + 1, 0); }
          if (isinclass) { count++; ADD_NEW(state_offset, count); }
          break;

          case OP_CRQUERY:
          case OP_CRMINQUERY:
          ADD_ACTIVE(next_state_offset + 1, 0);
          if (isinclass) { ADD_NEW(next_state_offset + 1, 0); }
          break;

          case OP_CRRANGE:
          case OP_CRMINRANGE:
          count = current_state->count;  /* Already matched */
          if (count >= GET2(ecode, 1))
            { ADD_ACTIVE(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
          if (isinclass)
            {
            unsigned int max = GET2(ecode, 1 + IMM2_SIZE);
            if (++count >= max && max != 0)   /* Max 0 => no limit */
              { ADD_NEW(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
            else
              { ADD_NEW(state_offset, count); }
            }
          break;

          default:
          if (isinclass) { ADD_NEW(next_state_offset, 0); }
          break;
          }
        }
      break;

/* ========================================================================== */
      /* These are the opcodes for fancy brackets of various kinds. We have
      to use recursion in order to handle them. The "always failing" assertion
      (?!) is optimised to OP_FAIL when compiling, so we have to support that,
      though the other "backtracking verbs" are not supported. */

      case OP_FAIL:
      forced_fail++;    /* Count FAILs for multiple states */
      break;

      case OP_ASSERT:
      case OP_ASSERT_NOT:
      case OP_ASSERTBACK:
      case OP_ASSERTBACK_NOT:
        {
        int rc;
        int local_offsets[2];
        int local_workspace[1000];
        const pcre_uchar *endasscode = code + GET(code, 1);

        while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);

        rc = internal_dfa_exec(
          md,                                   /* static match data */
          code,                                 /* this subexpression's code */
          ptr,                                  /* where we currently are */
          (int)(ptr - start_subject),           /* start offset */
          local_offsets,                        /* offset vector */
          sizeof(local_offsets)/sizeof(int),    /* size of same */
          local_workspace,                      /* workspace vector */
          sizeof(local_workspace)/sizeof(int),  /* size of same */
          rlevel);                              /* function recursion level */

        if (rc == PCRE_ERROR_DFA_UITEM) return rc;
        if ((rc >= 0) == (codevalue == OP_ASSERT || codevalue == OP_ASSERTBACK))
            { ADD_ACTIVE((int)(endasscode + LINK_SIZE + 1 - start_code), 0); }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_COND:
      case OP_SCOND:
        {
        int local_offsets[1000];
        int local_workspace[1000];
        int codelink = GET(code, 1);
        int condcode;

        /* Because of the way auto-callout works during compile, a callout item
        is inserted between OP_COND and an assertion condition. This does not
        happen for the other conditions. */

        if (code[LINK_SIZE+1] == OP_CALLOUT)
          {
          rrc = 0;
          if (PUBL(callout) != NULL)
            {
            PUBL(callout_block) cb;
            cb.version          = 1;   /* Version 1 of the callout block */
            cb.callout_number   = code[LINK_SIZE+2];
            cb.offset_vector    = offsets;
#if defined COMPILE_PCRE8
            cb.subject          = (PCRE_SPTR)start_subject;
#elif defined COMPILE_PCRE16
            cb.subject          = (PCRE_SPTR16)start_subject;
#elif defined COMPILE_PCRE32
            cb.subject          = (PCRE_SPTR32)start_subject;
#endif
            cb.subject_length   = (int)(end_subject - start_subject);
            cb.start_match      = (int)(current_subject - start_subject);
            cb.current_position = (int)(ptr - start_subject);
            cb.pattern_position = GET(code, LINK_SIZE + 3);
            cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);
            cb.capture_top      = 1;
            cb.capture_last     = -1;
            cb.callout_data     = md->callout_data;
            cb.mark             = NULL;   /* No (*MARK) support */
            if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
            }
          if (rrc > 0) break;                      /* Fail this thread */
          code += PRIV(OP_lengths)[OP_CALLOUT];    /* Skip callout data */
          }

        condcode = code[LINK_SIZE+1];

        /* Back reference conditions are not supported */

        if (condcode == OP_CREF || condcode == OP_NCREF)
          return PCRE_ERROR_DFA_UCOND;

        /* The DEFINE condition is always false */

        if (condcode == OP_DEF)
          { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }

        /* The only supported version of OP_RREF is for the value RREF_ANY,
        which means "test if in any recursion". We can't test for specifically
        recursed groups. */

        else if (condcode == OP_RREF || condcode == OP_NRREF)
          {
          int value = GET2(code, LINK_SIZE + 2);
          if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;
          if (md->recursive != NULL)
            { ADD_ACTIVE(state_offset + LINK_SIZE + 2 + IMM2_SIZE, 0); }
          else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
          }

        /* Otherwise, the condition is an assertion */

        else
          {
          int rc;
          const pcre_uchar *asscode = code + LINK_SIZE + 1;
          const pcre_uchar *endasscode = asscode + GET(asscode, 1);

          while (*endasscode == OP_ALT) endasscode += GET(endasscode, 1);

          rc = internal_dfa_exec(
            md,                                   /* fixed match data */
            asscode,                              /* this subexpression's code */
            ptr,                                  /* where we currently are */
            (int)(ptr - start_subject),           /* start offset */
            local_offsets,                        /* offset vector */
            sizeof(local_offsets)/sizeof(int),    /* size of same */
            local_workspace,                      /* workspace vector */
            sizeof(local_workspace)/sizeof(int),  /* size of same */
            rlevel);                              /* function recursion level */

          if (rc == PCRE_ERROR_DFA_UITEM) return rc;
          if ((rc >= 0) ==
                (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))
            { ADD_ACTIVE((int)(endasscode + LINK_SIZE + 1 - start_code), 0); }
          else
            { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_RECURSE:
        {
        dfa_recursion_info *ri;
        int local_offsets[1000];
        int local_workspace[1000];
        const pcre_uchar *callpat = start_code + GET(code, 1);
        int recno = (callpat == md->start_code)? 0 :
          GET2(callpat, 1 + LINK_SIZE);
        int rc;

        DPRINTF(("%.*sStarting regex recursion\n", rlevel*2-2, SP));

        /* Check for repeating a recursion without advancing the subject
        pointer. This should catch convoluted mutual recursions. (Some simple
        cases are caught at compile time.) */

        for (ri = md->recursive; ri != NULL; ri = ri->prevrec)
          if (recno == ri->group_num && ptr == ri->subject_position)
            return PCRE_ERROR_RECURSELOOP;

        /* Remember this recursion and where we started it so as to
        catch infinite loops. */

        new_recursive.group_num = recno;
        new_recursive.subject_position = ptr;
        new_recursive.prevrec = md->recursive;
        md->recursive = &new_recursive;

        rc = internal_dfa_exec(
          md,                                   /* fixed match data */
          callpat,                              /* this subexpression's code */
          ptr,                                  /* where we currently are */
          (int)(ptr - start_subject),           /* start offset */
          local_offsets,                        /* offset vector */
          sizeof(local_offsets)/sizeof(int),    /* size of same */
          local_workspace,                      /* workspace vector */
          sizeof(local_workspace)/sizeof(int),  /* size of same */
          rlevel);                              /* function recursion level */

        md->recursive = new_recursive.prevrec;  /* Done this recursion */

        DPRINTF(("%.*sReturn from regex recursion: rc=%d\n", rlevel*2-2, SP,
          rc));

        /* Ran out of internal offsets */

        if (rc == 0) return PCRE_ERROR_DFA_RECURSE;

        /* For each successful matched substring, set up the next state with a
        count of characters to skip before trying it. Note that the count is in
        characters, not bytes. */

        if (rc > 0)
          {
          for (rc = rc*2 - 2; rc >= 0; rc -= 2)
            {
            int charcount = local_offsets[rc+1] - local_offsets[rc];
#if defined SUPPORT_UTF && !defined COMPILE_PCRE32
            if (utf)
              {
              const pcre_uchar *p = start_subject + local_offsets[rc];
              const pcre_uchar *pp = start_subject + local_offsets[rc+1];
              while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
              }
#endif
            if (charcount > 0)
              {
              ADD_NEW_DATA(-(state_offset + LINK_SIZE + 1), 0, (charcount - 1));
              }
            else
              {
              ADD_ACTIVE(state_offset + LINK_SIZE + 1, 0);
              }
            }
          }
        else if (rc != PCRE_ERROR_NOMATCH) return rc;
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_BRAPOS:
      case OP_SBRAPOS:
      case OP_CBRAPOS:
      case OP_SCBRAPOS:
      case OP_BRAPOSZERO:
        {
        int charcount, matched_count;
        const pcre_uchar *local_ptr = ptr;
        BOOL allow_zero;

        if (codevalue == OP_BRAPOSZERO)
          {
          allow_zero = TRUE;
          codevalue = *(++code);  /* Codevalue will be one of above BRAs */
          }
        else allow_zero = FALSE;

        /* Loop to match the subpattern as many times as possible as if it were
        a complete pattern. */

        for (matched_count = 0;; matched_count++)
          {
          int local_offsets[2];
          int local_workspace[1000];

          int rc = internal_dfa_exec(
            md,                                   /* fixed match data */
            code,                                 /* this subexpression's code */
            local_ptr,                            /* where we currently are */
            (int)(ptr - start_subject),           /* start offset */
            local_offsets,                        /* offset vector */
            sizeof(local_offsets)/sizeof(int),    /* size of same */
            local_workspace,                      /* workspace vector */
            sizeof(local_workspace)/sizeof(int),  /* size of same */
            rlevel);                              /* function recursion level */

          /* Failed to match */

          if (rc < 0)
            {
            if (rc != PCRE_ERROR_NOMATCH) return rc;
            break;
            }

          /* Matched: break the loop if zero characters matched. */

          charcount = local_offsets[1] - local_offsets[0];
          if (charcount == 0) break;
          local_ptr += charcount;    /* Advance temporary position ptr */
          }

        /* At this point we have matched the subpattern matched_count
        times, and local_ptr is pointing to the character after the end of the
        last match. */

        if (matched_count > 0 || allow_zero)
          {
          const pcre_uchar *end_subpattern = code;
          int next_state_offset;

          do { end_subpattern += GET(end_subpattern, 1); }
            while (*end_subpattern == OP_ALT);
          next_state_offset =
            (int)(end_subpattern - start_code + LINK_SIZE + 1);

          /* Optimization: if there are no more active states, and there
          are no new states yet set up, then skip over the subject string
          right here, to save looping. Otherwise, set up the new state to swing
          into action when the end of the matched substring is reached. */

          if (i + 1 >= active_count && new_count == 0)
            {
            ptr = local_ptr;
            clen = 0;
            ADD_NEW(next_state_offset, 0);
            }
          else
            {
            const pcre_uchar *p = ptr;
            const pcre_uchar *pp = local_ptr;
            charcount = (int)(pp - p);
#if defined SUPPORT_UTF && !defined COMPILE_PCRE32
            if (utf) while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
#endif
            ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
            }
          }
        }
      break;

      /*-----------------------------------------------------------------*/
      case OP_ONCE:
      case OP_ONCE_NC:
        {
        int local_offsets[2];
        int local_workspace[1000];

        int rc = internal_dfa_exec(
          md,                                   /* fixed match data */
          code,                                 /* this subexpression's code */
          ptr,                                  /* where we currently are */
          (int)(ptr - start_subject),           /* start offset */
          local_offsets,                        /* offset vector */
          sizeof(local_offsets)/sizeof(int),    /* size of same */
          local_workspace,                      /* workspace vector */
          sizeof(local_workspace)/sizeof(int),  /* size of same */
          rlevel);                              /* function recursion level */

        if (rc >= 0)
          {
          const pcre_uchar *end_subpattern = code;
          int charcount = local_offsets[1] - local_offsets[0];
          int next_state_offset, repeat_state_offset;

          do { end_subpattern += GET(end_subpattern, 1); }
            while (*end_subpattern == OP_ALT);
          next_state_offset =
            (int)(end_subpattern - start_code + LINK_SIZE + 1);

          /* If the end of this subpattern is KETRMAX or KETRMIN, we must
          arrange for the repeat state also to be added to the relevant list.
          Calculate the offset, or set -1 for no repeat. */

          repeat_state_offset = (*end_subpattern == OP_KETRMAX ||
                                 *end_subpattern == OP_KETRMIN)?
            (int)(end_subpattern - start_code - GET(end_subpattern, 1)) : -1;

          /* If we have matched an empty string, add the next state at the
          current character pointer. This is important so that the duplicate
          checking kicks in, which is what breaks infinite loops that match an
          empty string. */

          if (charcount == 0)
            {
            ADD_ACTIVE(next_state_offset, 0);
            }

          /* Optimization: if there are no more active states, and there
          are no new states yet set up, then skip over the subject string
          right here, to save looping. Otherwise, set up the new state to swing
          into action when the end of the matched substring is reached. */

          else if (i + 1 >= active_count && new_count == 0)
            {
            ptr += charcount;
            clen = 0;
            ADD_NEW(next_state_offset, 0);

            /* If we are adding a repeat state at the new character position,
            we must fudge things so that it is the only current state.
            Otherwise, it might be a duplicate of one we processed before, and
            that would cause it to be skipped. */

            if (repeat_state_offset >= 0)
              {
              next_active_state = active_states;
              active_count = 0;
              i = -1;
              ADD_ACTIVE(repeat_state_offset, 0);
              }
            }
          else
            {
#if defined SUPPORT_UTF && !defined COMPILE_PCRE32
            if (utf)
              {
              const pcre_uchar *p = start_subject + local_offsets[0];
              const pcre_uchar *pp = start_subject + local_offsets[1];
              while (p < pp) if (NOT_FIRSTCHAR(*p++)) charcount--;
              }
#endif
            ADD_NEW_DATA(-next_state_offset, 0, (charcount - 1));
            if (repeat_state_offset >= 0)
              { ADD_NEW_DATA(-repeat_state_offset, 0, (charcount - 1)); }
            }
          }
        else if (rc != PCRE_ERROR_NOMATCH) return rc;
        }
      break;


/* ========================================================================== */
      /* Handle callouts */

      case OP_CALLOUT:
      rrc = 0;
      if (PUBL(callout) != NULL)
        {
        PUBL(callout_block) cb;
        cb.version          = 1;   /* Version 1 of the callout block */
        cb.callout_number   = code[1];
        cb.offset_vector    = offsets;
#if defined COMPILE_PCRE8
        cb.subject          = (PCRE_SPTR)start_subject;
#elif defined COMPILE_PCRE16
        cb.subject          = (PCRE_SPTR16)start_subject;
#elif defined COMPILE_PCRE32
        cb.subject          = (PCRE_SPTR32)start_subject;
#endif
        cb.subject_length   = (int)(end_subject - start_subject);
        cb.start_match      = (int)(current_subject - start_subject);
        cb.current_position = (int)(ptr - start_subject);
        cb.pattern_position = GET(code, 2);
        cb.next_item_length = GET(code, 2 + LINK_SIZE);
        cb.capture_top      = 1;
        cb.capture_last     = -1;
        cb.callout_data     = md->callout_data;
        cb.mark             = NULL;   /* No (*MARK) support */
        if ((rrc = (*PUBL(callout))(&cb)) < 0) return rrc;   /* Abandon */
        }
      if (rrc == 0)
        { ADD_ACTIVE(state_offset + PRIV(OP_lengths)[OP_CALLOUT], 0); }
      break;


/* ========================================================================== */
      default:        /* Unsupported opcode */
      return PCRE_ERROR_DFA_UITEM;
      }

    NEXT_ACTIVE_STATE: continue;

    }      /* End of loop scanning active states */

  /* We have finished the processing at the current subject character. If no
  new states have been set for the next character, we have found all the
  matches that we are going to find. If we are at the top level and partial
  matching has been requested, check for appropriate conditions.

  The "forced_ fail" variable counts the number of (*F) encountered for the
  character. If it is equal to the original active_count (saved in
  workspace[1]) it means that (*F) was found on every active state. In this
  case we don't want to give a partial match.

  The "could_continue" variable is true if a state could have continued but
  for the fact that the end of the subject was reached. */

  if (new_count <= 0)
    {
    if (rlevel == 1 &&                               /* Top level, and */
        could_continue &&                            /* Some could go on, and */
        forced_fail != workspace[1] &&               /* Not all forced fail & */
        (                                            /* either... */
        (md->moptions & PCRE_PARTIAL_HARD) != 0      /* Hard partial */
        ||                                           /* or... */
        ((md->moptions & PCRE_PARTIAL_SOFT) != 0 &&  /* Soft partial and */
         match_count < 0)                            /* no matches */
        ) &&                                         /* And... */
        (
        partial_newline ||                           /* Either partial NL */
          (                                          /* or ... */
          ptr >= end_subject &&                /* End of subject and */
          ptr > md->start_used_ptr)            /* Inspected non-empty string */
          )
        )
      {
      if (offsetcount >= 2)
        {
        offsets[0] = (int)(md->start_used_ptr - start_subject);
        offsets[1] = (int)(end_subject - start_subject);
        }
      match_count = PCRE_ERROR_PARTIAL;
      }

    DPRINTF(("%.*sEnd of internal_dfa_exec %d: returning %d\n"
      "%.*s---------------------\n\n", rlevel*2-2, SP, rlevel, match_count,
      rlevel*2-2, SP));
    break;        /* In effect, "return", but see the comment below */
    }

  /* One or more states are active for the next character. */

  ptr += clen;    /* Advance to next subject character */
  }               /* Loop to move along the subject string */

/* Control gets here from "break" a few lines above. We do it this way because
if we use "return" above, we have compiler trouble. Some compilers warn if
there's nothing here because they think the function doesn't return a value. On
the other hand, if we put a dummy statement here, some more clever compilers
complain that it can't be reached. Sigh. */

return match_count;
}




/*************************************************
*    Execute a Regular Expression - DFA engine   *
*************************************************/

/* This external function applies a compiled re to a subject string using a DFA
engine. This function calls the internal function multiple times if the pattern
is not anchored.

Arguments:
  argument_re     points to the compiled expression
  extra_data      points to extra data or is NULL
  subject         points to the subject string
  length          length of subject string (may contain binary zeros)
  start_offset    where to start in the subject string
  options         option bits
  offsets         vector of match offsets
  offsetcount     size of same
  workspace       workspace vector
  wscount         size of same

Returns:          > 0 => number of match offset pairs placed in offsets
                  = 0 => offsets overflowed; longest matches are present
                   -1 => failed to match
                 < -1 => some kind of unexpected problem
*/

#if defined COMPILE_PCRE8
PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,
  const char *subject, int length, int start_offset, int options, int *offsets,
  int offsetcount, int *workspace, int wscount)
#elif defined COMPILE_PCRE16
PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
pcre16_dfa_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
  PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
  int offsetcount, int *workspace, int wscount)
#elif defined COMPILE_PCRE32
PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
pcre32_dfa_exec(const pcre32 *argument_re, const pcre32_extra *extra_data,
  PCRE_SPTR32 subject, int length, int start_offset, int options, int *offsets,
  int offsetcount, int *workspace, int wscount)
#endif
{
REAL_PCRE *re = (REAL_PCRE *)argument_re;
dfa_match_data match_block;
dfa_match_data *md = &match_block;
BOOL utf, anchored, startline, firstline;
const pcre_uchar *current_subject, *end_subject;
const pcre_study_data *study = NULL;

const pcre_uchar *req_char_ptr;
const pcre_uint8 *start_bits = NULL;
BOOL has_first_char = FALSE;
BOOL has_req_char = FALSE;
pcre_uchar first_char = 0;
pcre_uchar first_char2 = 0;
pcre_uchar req_char = 0;
pcre_uchar req_char2 = 0;
int newline;

/* Plausibility checks */

if ((options & ~PUBLIC_DFA_EXEC_OPTIONS) != 0) return PCRE_ERROR_BADOPTION;
if (re == NULL || subject == NULL || workspace == NULL ||
   (offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;
if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
if (wscount < 20) return PCRE_ERROR_DFA_WSSIZE;
if (length < 0) return PCRE_ERROR_BADLENGTH;
if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;

/* Check that the first field in the block is the magic number. If it is not,
return with PCRE_ERROR_BADMAGIC. However, if the magic number is equal to
REVERSED_MAGIC_NUMBER we return with PCRE_ERROR_BADENDIANNESS, which
means that the pattern is likely compiled with different endianness. */

if (re->magic_number != MAGIC_NUMBER)
  return re->magic_number == REVERSED_MAGIC_NUMBER?
    PCRE_ERROR_BADENDIANNESS:PCRE_ERROR_BADMAGIC;
if ((re->flags & PCRE_MODE) == 0) return PCRE_ERROR_BADMODE;

/* If restarting after a partial match, do some sanity checks on the contents
of the workspace. */

if ((options & PCRE_DFA_RESTART) != 0)
  {
  if ((workspace[0] & (-2)) != 0 || workspace[1] < 1 ||
    workspace[1] > (wscount - 2)/INTS_PER_STATEBLOCK)
      return PCRE_ERROR_DFA_BADRESTART;
  }

/* Set up study, callout, and table data */

md->tables = re->tables;
md->callout_data = NULL;

if (extra_data != NULL)
  {
  unsigned int flags = extra_data->flags;
  if ((flags & PCRE_EXTRA_STUDY_DATA) != 0)
    study = (const pcre_study_data *)extra_data->study_data;
  if ((flags & PCRE_EXTRA_MATCH_LIMIT) != 0) return PCRE_ERROR_DFA_UMLIMIT;
  if ((flags & PCRE_EXTRA_MATCH_LIMIT_RECURSION) != 0)
    return PCRE_ERROR_DFA_UMLIMIT;
  if ((flags & PCRE_EXTRA_CALLOUT_DATA) != 0)
    md->callout_data = extra_data->callout_data;
  if ((flags & PCRE_EXTRA_TABLES) != 0)
    md->tables = extra_data->tables;
  }

/* Set some local values */

current_subject = (const pcre_uchar *)subject + start_offset;
end_subject = (const pcre_uchar *)subject + length;
req_char_ptr = current_subject - 1;

#ifdef SUPPORT_UTF
/* PCRE_UTF(16|32) have the same value as PCRE_UTF8. */
utf = (re->options & PCRE_UTF8) != 0;
#else
utf = FALSE;
#endif

anchored = (options & (PCRE_ANCHORED|PCRE_DFA_RESTART)) != 0 ||
  (re->options & PCRE_ANCHORED) != 0;

/* The remaining fixed data for passing around. */

md->start_code = (const pcre_uchar *)argument_re +
    re->name_table_offset + re->name_count * re->name_entry_size;
md->start_subject = (const pcre_uchar *)subject;
md->end_subject = end_subject;
md->start_offset = start_offset;
md->moptions = options;
md->poptions = re->options;

/* If the BSR option is not set at match time, copy what was set
at compile time. */

if ((md->moptions & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) == 0)
  {
  if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
    md->moptions |= re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE);
#ifdef BSR_ANYCRLF
  else md->moptions |= PCRE_BSR_ANYCRLF;
#endif
  }

/* Handle different types of newline. The three bits give eight cases. If
nothing is set at run time, whatever was used at compile time applies. */

switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : (pcre_uint32)options) &
         PCRE_NEWLINE_BITS)
  {
  case 0: newline = NEWLINE; break;   /* Compile-time default */
  case PCRE_NEWLINE_CR: newline = CHAR_CR; break;
  case PCRE_NEWLINE_LF: newline = CHAR_NL; break;
  case PCRE_NEWLINE_CR+
       PCRE_NEWLINE_LF: newline = (CHAR_CR << 8) | CHAR_NL; break;
  case PCRE_NEWLINE_ANY: newline = -1; break;
  case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
  default: return PCRE_ERROR_BADNEWLINE;
  }

if (newline == -2)
  {
  md->nltype = NLTYPE_ANYCRLF;
  }
else if (newline < 0)
  {
  md->nltype = NLTYPE_ANY;
  }
else
  {
  md->nltype = NLTYPE_FIXED;
  if (newline > 255)
    {
    md->nllen = 2;
    md->nl[0] = (newline >> 8) & 255;
    md->nl[1] = newline & 255;
    }
  else
    {
    md->nllen = 1;
    md->nl[0] = newline;
    }
  }

/* Check a UTF-8 string if required. Unfortunately there's no way of passing
back the character offset. */

#ifdef SUPPORT_UTF
if (utf && (options & PCRE_NO_UTF8_CHECK) == 0)
  {
  int erroroffset;
  int errorcode = PRIV(valid_utf)((pcre_uchar *)subject, length, &erroroffset);
  if (errorcode != 0)
    {
    if (offsetcount >= 2)
      {
      offsets[0] = erroroffset;
      offsets[1] = errorcode;
      }
#if defined COMPILE_PCRE8
    return (errorcode <= PCRE_UTF8_ERR5 && (options & PCRE_PARTIAL_HARD) != 0) ?
      PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
#elif defined COMPILE_PCRE16
    return (errorcode <= PCRE_UTF16_ERR1 && (options & PCRE_PARTIAL_HARD) != 0) ?
      PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;
#elif defined COMPILE_PCRE32
    return PCRE_ERROR_BADUTF32;
#endif
    }
#if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
  if (start_offset > 0 && start_offset < length &&
        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
    return PCRE_ERROR_BADUTF8_OFFSET;
#endif
  }
#endif

/* If the exec call supplied NULL for tables, use the inbuilt ones. This
is a feature that makes it possible to save compiled regex and re-use them
in other programs later. */

if (md->tables == NULL) md->tables = PRIV(default_tables);

/* The "must be at the start of a line" flags are used in a loop when finding
where to start. */

startline = (re->flags & PCRE_STARTLINE) != 0;
firstline = (re->options & PCRE_FIRSTLINE) != 0;

/* Set up the first character to match, if available. The first_byte value is
never set for an anchored regular expression, but the anchoring may be forced
at run time, so we have to test for anchoring. The first char may be unset for
an unanchored pattern, of course. If there's no first char and the pattern was
studied, there may be a bitmap of possible first characters. */

if (!anchored)
  {
  if ((re->flags & PCRE_FIRSTSET) != 0)
    {
    has_first_char = TRUE;
    first_char = first_char2 = (pcre_uchar)(re->first_char);
    if ((re->flags & PCRE_FCH_CASELESS) != 0)
      {
      first_char2 = TABLE_GET(first_char, md->tables + fcc_offset, first_char);
#if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
      if (utf && first_char > 127)
        first_char2 = UCD_OTHERCASE(first_char);
#endif
      }
    }
  else
    {
    if (!startline && study != NULL &&
         (study->flags & PCRE_STUDY_MAPPED) != 0)
      start_bits = study->start_bits;
    }
  }

/* For anchored or unanchored matches, there may be a "last known required
character" set. */

if ((re->flags & PCRE_REQCHSET) != 0)
  {
  has_req_char = TRUE;
  req_char = req_char2 = (pcre_uchar)(re->req_char);
  if ((re->flags & PCRE_RCH_CASELESS) != 0)
    {
    req_char2 = TABLE_GET(req_char, md->tables + fcc_offset, req_char);
#if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
    if (utf && req_char > 127)
      req_char2 = UCD_OTHERCASE(req_char);
#endif
    }
  }

/* Call the main matching function, looping for a non-anchored regex after a
failed match. If not restarting, perform certain optimizations at the start of
a match. */

for (;;)
  {
  int rc;

  if ((options & PCRE_DFA_RESTART) == 0)
    {
    const pcre_uchar *save_end_subject = end_subject;

    /* If firstline is TRUE, the start of the match is constrained to the first
    line of a multiline string. Implement this by temporarily adjusting
    end_subject so that we stop scanning at a newline. If the match fails at
    the newline, later code breaks this loop. */

    if (firstline)
      {
      PCRE_PUCHAR t = current_subject;
#ifdef SUPPORT_UTF
      if (utf)
        {
        while (t < md->end_subject && !IS_NEWLINE(t))
          {
          t++;
          ACROSSCHAR(t < end_subject, *t, t++);
          }
        }
      else
#endif
      while (t < md->end_subject && !IS_NEWLINE(t)) t++;
      end_subject = t;
      }

    /* There are some optimizations that avoid running the match if a known
    starting point is not found. However, there is an option that disables
    these, for testing and for ensuring that all callouts do actually occur.
    The option can be set in the regex by (*NO_START_OPT) or passed in
    match-time options. */

    if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0)
      {
      /* Advance to a known first char. */

      if (has_first_char)
        {
        if (first_char != first_char2)
          {
          pcre_uchar csc;
          while (current_subject < end_subject &&
                 (csc = RAWUCHARTEST(current_subject)) != first_char && csc != first_char2)
            current_subject++;
          }
        else
          while (current_subject < end_subject &&
                 RAWUCHARTEST(current_subject) != first_char)
            current_subject++;
        }

      /* Or to just after a linebreak for a multiline match if possible */

      else if (startline)
        {
        if (current_subject > md->start_subject + start_offset)
          {
#ifdef SUPPORT_UTF
          if (utf)
            {
            while (current_subject < end_subject &&
                   !WAS_NEWLINE(current_subject))
              {
              current_subject++;
              ACROSSCHAR(current_subject < end_subject, *current_subject,
                current_subject++);
              }
            }
          else
#endif
          while (current_subject < end_subject && !WAS_NEWLINE(current_subject))
            current_subject++;

          /* If we have just passed a CR and the newline option is ANY or
          ANYCRLF, and we are now at a LF, advance the match position by one
          more character. */

          if (RAWUCHARTEST(current_subject - 1) == CHAR_CR &&
               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
               current_subject < end_subject &&
               RAWUCHARTEST(current_subject) == CHAR_NL)
            current_subject++;
          }
        }

      /* Or to a non-unique first char after study */

      else if (start_bits != NULL)
        {
        while (current_subject < end_subject)
          {
          register pcre_uint32 c = RAWUCHARTEST(current_subject);
#ifndef COMPILE_PCRE8
          if (c > 255) c = 255;
#endif
          if ((start_bits[c/8] & (1 << (c&7))) == 0)
            {
            current_subject++;
#if defined SUPPORT_UTF && defined COMPILE_PCRE8
            /* In non 8-bit mode, the iteration will stop for
            characters > 255 at the beginning or not stop at all. */
            if (utf)
              ACROSSCHAR(current_subject < end_subject, *current_subject,
                current_subject++);
#endif
            }
          else break;
          }
        }
      }

    /* Restore fudged end_subject */

    end_subject = save_end_subject;

    /* The following two optimizations are disabled for partial matching or if
    disabling is explicitly requested (and of course, by the test above, this
    code is not obeyed when restarting after a partial match). */

    if (((options | re->options) & PCRE_NO_START_OPTIMIZE) == 0 &&
        (options & (PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT)) == 0)
      {
      /* If the pattern was studied, a minimum subject length may be set. This
      is a lower bound; no actual string of that length may actually match the
      pattern. Although the value is, strictly, in characters, we treat it as
      bytes to avoid spending too much time in this optimization. */

      if (study != NULL && (study->flags & PCRE_STUDY_MINLEN) != 0 &&
          (pcre_uint32)(end_subject - current_subject) < study->minlength)
        return PCRE_ERROR_NOMATCH;

      /* If req_char is set, we know that that character must appear in the
      subject for the match to succeed. If the first character is set, req_char
      must be later in the subject; otherwise the test starts at the match
      point. This optimization can save a huge amount of work in patterns with
      nested unlimited repeats that aren't going to match. Writing separate
      code for cased/caseless versions makes it go faster, as does using an
      autoincrement and backing off on a match.

      HOWEVER: when the subject string is very, very long, searching to its end
      can take a long time, and give bad performance on quite ordinary
      patterns. This showed up when somebody was matching /^C/ on a 32-megabyte
      string... so we don't do this when the string is sufficiently long. */

      if (has_req_char && end_subject - current_subject < REQ_BYTE_MAX)
        {
        register PCRE_PUCHAR p = current_subject + (has_first_char? 1:0);

        /* We don't need to repeat the search if we haven't yet reached the
        place we found it at last time. */

        if (p > req_char_ptr)
          {
          if (req_char != req_char2)
            {
            while (p < end_subject)
              {
              register pcre_uint32 pp = RAWUCHARINCTEST(p);
              if (pp == req_char || pp == req_char2) { p--; break; }
              }
            }
          else
            {
            while (p < end_subject)
              {
              if (RAWUCHARINCTEST(p) == req_char) { p--; break; }
              }
            }

          /* If we can't find the required character, break the matching loop,
          which will cause a return or PCRE_ERROR_NOMATCH. */

          if (p >= end_subject) break;

          /* If we have found the required character, save the point where we
          found it, so that we don't search again next time round the loop if
          the start hasn't passed this character yet. */

          req_char_ptr = p;
          }
        }
      }
    }   /* End of optimizations that are done when not restarting */

  /* OK, now we can do the business */

  md->start_used_ptr = current_subject;
  md->recursive = NULL;

  rc = internal_dfa_exec(
    md,                                /* fixed match data */
    md->start_code,                    /* this subexpression's code */
    current_subject,                   /* where we currently are */
    start_offset,                      /* start offset in subject */
    offsets,                           /* offset vector */
    offsetcount,                       /* size of same */
    workspace,                         /* workspace vector */
    wscount,                           /* size of same */
    0);                                /* function recurse level */

  /* Anything other than "no match" means we are done, always; otherwise, carry
  on only if not anchored. */

  if (rc != PCRE_ERROR_NOMATCH || anchored) return rc;

  /* Advance to the next subject character unless we are at the end of a line
  and firstline is set. */

  if (firstline && IS_NEWLINE(current_subject)) break;
  current_subject++;
#ifdef SUPPORT_UTF
  if (utf)
    {
    ACROSSCHAR(current_subject < end_subject, *current_subject,
      current_subject++);
    }
#endif
  if (current_subject > end_subject) break;

  /* If we have just passed a CR and we are now at a LF, and the pattern does
  not contain any explicit matches for \r or \n, and the newline option is CRLF
  or ANY or ANYCRLF, advance the match position by one more character. */

  if (RAWUCHARTEST(current_subject - 1) == CHAR_CR &&
      current_subject < end_subject &&
      RAWUCHARTEST(current_subject) == CHAR_NL &&
      (re->flags & PCRE_HASCRORLF) == 0 &&
        (md->nltype == NLTYPE_ANY ||
         md->nltype == NLTYPE_ANYCRLF ||
         md->nllen == 2))
    current_subject++;

  }   /* "Bumpalong" loop */

return PCRE_ERROR_NOMATCH;
}

/* End of pcre_dfa_exec.c */