summaryrefslogtreecommitdiff
path: root/src/search.c
blob: 3fc3faeda98a060e02f423188ce70b80d9bbd910 (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
/* String search routines for GNU Emacs.
   Copyright (C) 1985, 1986, 1987, 1993, 1994, 1997, 1998, 1999, 2002, 2003,
                 2004, 2005, 2006 Free Software Foundation, Inc.

This file is part of GNU Emacs.

GNU Emacs is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.

GNU Emacs is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with GNU Emacs; see the file COPYING.  If not, write to
the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
Boston, MA 02110-1301, USA.  */


#include <config.h>
#include "lisp.h"
#include "syntax.h"
#include "category.h"
#include "buffer.h"
#include "charset.h"
#include "region-cache.h"
#include "commands.h"
#include "blockinput.h"
#include "intervals.h"

#include <sys/types.h>
#include "regex.h"

#define REGEXP_CACHE_SIZE 20

/* If the regexp is non-nil, then the buffer contains the compiled form
   of that regexp, suitable for searching.  */
struct regexp_cache
{
  struct regexp_cache *next;
  Lisp_Object regexp, whitespace_regexp;
  /* Syntax table for which the regexp applies.  We need this because
     of character classes.  If this is t, then the compiled pattern is valid
     for any syntax-table.  */
  Lisp_Object syntax_table;
  struct re_pattern_buffer buf;
  char fastmap[0400];
  /* Nonzero means regexp was compiled to do full POSIX backtracking.  */
  char posix;
};

/* The instances of that struct.  */
struct regexp_cache searchbufs[REGEXP_CACHE_SIZE];

/* The head of the linked list; points to the most recently used buffer.  */
struct regexp_cache *searchbuf_head;


/* Every call to re_match, etc., must pass &search_regs as the regs
   argument unless you can show it is unnecessary (i.e., if re_match
   is certainly going to be called again before region-around-match
   can be called).

   Since the registers are now dynamically allocated, we need to make
   sure not to refer to the Nth register before checking that it has
   been allocated by checking search_regs.num_regs.

   The regex code keeps track of whether it has allocated the search
   buffer using bits in the re_pattern_buffer.  This means that whenever
   you compile a new pattern, it completely forgets whether it has
   allocated any registers, and will allocate new registers the next
   time you call a searching or matching function.  Therefore, we need
   to call re_set_registers after compiling a new pattern or after
   setting the match registers, so that the regex functions will be
   able to free or re-allocate it properly.  */
static struct re_registers search_regs;

/* The buffer in which the last search was performed, or
   Qt if the last search was done in a string;
   Qnil if no searching has been done yet.  */
static Lisp_Object last_thing_searched;

/* error condition signaled when regexp compile_pattern fails */

Lisp_Object Qinvalid_regexp;

/* Error condition used for failing searches */
Lisp_Object Qsearch_failed;

Lisp_Object Vsearch_spaces_regexp;

static void set_search_regs ();
static void save_search_regs ();
static int simple_search ();
static int boyer_moore ();
static int search_buffer ();
static void matcher_overflow () NO_RETURN;

static void
matcher_overflow ()
{
  error ("Stack overflow in regexp matcher");
}

/* Compile a regexp and signal a Lisp error if anything goes wrong.
   PATTERN is the pattern to compile.
   CP is the place to put the result.
   TRANSLATE is a translation table for ignoring case, or nil for none.
   REGP is the structure that says where to store the "register"
   values that will result from matching this pattern.
   If it is 0, we should compile the pattern not to record any
   subexpression bounds.
   POSIX is nonzero if we want full backtracking (POSIX style)
   for this pattern.  0 means backtrack only enough to get a valid match.
   MULTIBYTE is nonzero if we want to handle multibyte characters in
   PATTERN.  0 means all multibyte characters are recognized just as
   sequences of binary data.

   The behavior also depends on Vsearch_spaces_regexp.  */

static void
compile_pattern_1 (cp, pattern, translate, regp, posix, multibyte)
     struct regexp_cache *cp;
     Lisp_Object pattern;
     Lisp_Object translate;
     struct re_registers *regp;
     int posix;
     int multibyte;
{
  unsigned char *raw_pattern;
  int raw_pattern_size;
  char *val;
  reg_syntax_t old;

  /* MULTIBYTE says whether the text to be searched is multibyte.
     We must convert PATTERN to match that, or we will not really
     find things right.  */

  if (multibyte == STRING_MULTIBYTE (pattern))
    {
      raw_pattern = (unsigned char *) SDATA (pattern);
      raw_pattern_size = SBYTES (pattern);
    }
  else if (multibyte)
    {
      raw_pattern_size = count_size_as_multibyte (SDATA (pattern),
						  SCHARS (pattern));
      raw_pattern = (unsigned char *) alloca (raw_pattern_size + 1);
      copy_text (SDATA (pattern), raw_pattern,
		 SCHARS (pattern), 0, 1);
    }
  else
    {
      /* Converting multibyte to single-byte.

	 ??? Perhaps this conversion should be done in a special way
	 by subtracting nonascii-insert-offset from each non-ASCII char,
	 so that only the multibyte chars which really correspond to
	 the chosen single-byte character set can possibly match.  */
      raw_pattern_size = SCHARS (pattern);
      raw_pattern = (unsigned char *) alloca (raw_pattern_size + 1);
      copy_text (SDATA (pattern), raw_pattern,
		 SBYTES (pattern), 1, 0);
    }

  cp->regexp = Qnil;
  cp->buf.translate = (! NILP (translate) ? translate : make_number (0));
  cp->posix = posix;
  cp->buf.multibyte = multibyte;
  cp->whitespace_regexp = Vsearch_spaces_regexp;
  /* rms: I think BLOCK_INPUT is not needed here any more,
     because regex.c defines malloc to call xmalloc.
     Using BLOCK_INPUT here means the debugger won't run if an error occurs.
     So let's turn it off.  */
  /*  BLOCK_INPUT;  */
  old = re_set_syntax (RE_SYNTAX_EMACS
		       | (posix ? 0 : RE_NO_POSIX_BACKTRACKING));

  re_set_whitespace_regexp (NILP (Vsearch_spaces_regexp) ? NULL
			    : SDATA (Vsearch_spaces_regexp));

  val = (char *) re_compile_pattern ((char *)raw_pattern,
				     raw_pattern_size, &cp->buf);

  /* If the compiled pattern hard codes some of the contents of the
     syntax-table, it can only be reused with *this* syntax table.  */
  cp->syntax_table = cp->buf.used_syntax ? current_buffer->syntax_table : Qt;

  re_set_whitespace_regexp (NULL);

  re_set_syntax (old);
  /* UNBLOCK_INPUT;  */
  if (val)
    xsignal1 (Qinvalid_regexp, build_string (val));

  cp->regexp = Fcopy_sequence (pattern);
}

/* Shrink each compiled regexp buffer in the cache
   to the size actually used right now.
   This is called from garbage collection.  */

void
shrink_regexp_cache ()
{
  struct regexp_cache *cp;

  for (cp = searchbuf_head; cp != 0; cp = cp->next)
    {
      cp->buf.allocated = cp->buf.used;
      cp->buf.buffer
	= (unsigned char *) xrealloc (cp->buf.buffer, cp->buf.used);
    }
}

/* Clear the regexp cache w.r.t. a particular syntax table,
   because it was changed.
   There is no danger of memory leak here because re_compile_pattern
   automagically manages the memory in each re_pattern_buffer struct,
   based on its `allocated' and `buffer' values.  */
void
clear_regexp_cache ()
{
  int i;

  for (i = 0; i < REGEXP_CACHE_SIZE; ++i)
    /* It's tempting to compare with the syntax-table we've actually changd,
       but it's not sufficient because char-table inheritance mewans that
       modifying one syntax-table can change others at the same time.  */
    if (!EQ (searchbufs[i].syntax_table, Qt))
      searchbufs[i].regexp = Qnil;
}

/* Compile a regexp if necessary, but first check to see if there's one in
   the cache.
   PATTERN is the pattern to compile.
   TRANSLATE is a translation table for ignoring case, or nil for none.
   REGP is the structure that says where to store the "register"
   values that will result from matching this pattern.
   If it is 0, we should compile the pattern not to record any
   subexpression bounds.
   POSIX is nonzero if we want full backtracking (POSIX style)
   for this pattern.  0 means backtrack only enough to get a valid match.  */

struct re_pattern_buffer *
compile_pattern (pattern, regp, translate, posix, multibyte)
     Lisp_Object pattern;
     struct re_registers *regp;
     Lisp_Object translate;
     int posix, multibyte;
{
  struct regexp_cache *cp, **cpp;

  for (cpp = &searchbuf_head; ; cpp = &cp->next)
    {
      cp = *cpp;
      /* Entries are initialized to nil, and may be set to nil by
	 compile_pattern_1 if the pattern isn't valid.  Don't apply
	 string accessors in those cases.  However, compile_pattern_1
	 is only applied to the cache entry we pick here to reuse.  So
	 nil should never appear before a non-nil entry.  */
      if (NILP (cp->regexp))
	goto compile_it;
      if (SCHARS (cp->regexp) == SCHARS (pattern)
	  && STRING_MULTIBYTE (cp->regexp) == STRING_MULTIBYTE (pattern)
	  && !NILP (Fstring_equal (cp->regexp, pattern))
	  && EQ (cp->buf.translate, (! NILP (translate) ? translate : make_number (0)))
	  && cp->posix == posix
	  && cp->buf.multibyte == multibyte
	  && (EQ (cp->syntax_table, Qt)
	      || EQ (cp->syntax_table, current_buffer->syntax_table))
	  && !NILP (Fequal (cp->whitespace_regexp, Vsearch_spaces_regexp)))
	break;

      /* If we're at the end of the cache, compile into the nil cell
	 we found, or the last (least recently used) cell with a
	 string value.  */
      if (cp->next == 0)
	{
	compile_it:
	  compile_pattern_1 (cp, pattern, translate, regp, posix, multibyte);
	  break;
	}
    }

  /* When we get here, cp (aka *cpp) contains the compiled pattern,
     either because we found it in the cache or because we just compiled it.
     Move it to the front of the queue to mark it as most recently used.  */
  *cpp = cp->next;
  cp->next = searchbuf_head;
  searchbuf_head = cp;

  /* Advise the searching functions about the space we have allocated
     for register data.  */
  if (regp)
    re_set_registers (&cp->buf, regp, regp->num_regs, regp->start, regp->end);

  return &cp->buf;
}


static Lisp_Object
looking_at_1 (string, posix)
     Lisp_Object string;
     int posix;
{
  Lisp_Object val;
  unsigned char *p1, *p2;
  int s1, s2;
  register int i;
  struct re_pattern_buffer *bufp;

  if (running_asynch_code)
    save_search_regs ();

  /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
  XCHAR_TABLE (current_buffer->case_canon_table)->extras[2]
    = current_buffer->case_eqv_table;

  CHECK_STRING (string);
  bufp = compile_pattern (string, &search_regs,
			  (!NILP (current_buffer->case_fold_search)
			   ? current_buffer->case_canon_table : Qnil),
			  posix,
			  !NILP (current_buffer->enable_multibyte_characters));

  immediate_quit = 1;
  QUIT;			/* Do a pending quit right away, to avoid paradoxical behavior */

  /* Get pointers and sizes of the two strings
     that make up the visible portion of the buffer. */

  p1 = BEGV_ADDR;
  s1 = GPT_BYTE - BEGV_BYTE;
  p2 = GAP_END_ADDR;
  s2 = ZV_BYTE - GPT_BYTE;
  if (s1 < 0)
    {
      p2 = p1;
      s2 = ZV_BYTE - BEGV_BYTE;
      s1 = 0;
    }
  if (s2 < 0)
    {
      s1 = ZV_BYTE - BEGV_BYTE;
      s2 = 0;
    }

  re_match_object = Qnil;

  i = re_match_2 (bufp, (char *) p1, s1, (char *) p2, s2,
		  PT_BYTE - BEGV_BYTE, &search_regs,
		  ZV_BYTE - BEGV_BYTE);
  immediate_quit = 0;

  if (i == -2)
    matcher_overflow ();

  val = (0 <= i ? Qt : Qnil);
  if (i >= 0)
    for (i = 0; i < search_regs.num_regs; i++)
      if (search_regs.start[i] >= 0)
	{
	  search_regs.start[i]
	    = BYTE_TO_CHAR (search_regs.start[i] + BEGV_BYTE);
	  search_regs.end[i]
	    = BYTE_TO_CHAR (search_regs.end[i] + BEGV_BYTE);
	}
  XSETBUFFER (last_thing_searched, current_buffer);
  return val;
}

DEFUN ("looking-at", Flooking_at, Slooking_at, 1, 1, 0,
       doc: /* Return t if text after point matches regular expression REGEXP.
This function modifies the match data that `match-beginning',
`match-end' and `match-data' access; save and restore the match
data if you want to preserve them.  */)
     (regexp)
     Lisp_Object regexp;
{
  return looking_at_1 (regexp, 0);
}

DEFUN ("posix-looking-at", Fposix_looking_at, Sposix_looking_at, 1, 1, 0,
       doc: /* Return t if text after point matches regular expression REGEXP.
Find the longest match, in accord with Posix regular expression rules.
This function modifies the match data that `match-beginning',
`match-end' and `match-data' access; save and restore the match
data if you want to preserve them.  */)
     (regexp)
     Lisp_Object regexp;
{
  return looking_at_1 (regexp, 1);
}

static Lisp_Object
string_match_1 (regexp, string, start, posix)
     Lisp_Object regexp, string, start;
     int posix;
{
  int val;
  struct re_pattern_buffer *bufp;
  int pos, pos_byte;
  int i;

  if (running_asynch_code)
    save_search_regs ();

  CHECK_STRING (regexp);
  CHECK_STRING (string);

  if (NILP (start))
    pos = 0, pos_byte = 0;
  else
    {
      int len = SCHARS (string);

      CHECK_NUMBER (start);
      pos = XINT (start);
      if (pos < 0 && -pos <= len)
	pos = len + pos;
      else if (0 > pos || pos > len)
	args_out_of_range (string, start);
      pos_byte = string_char_to_byte (string, pos);
    }

  /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
  XCHAR_TABLE (current_buffer->case_canon_table)->extras[2]
    = current_buffer->case_eqv_table;

  bufp = compile_pattern (regexp, &search_regs,
			  (!NILP (current_buffer->case_fold_search)
			   ? current_buffer->case_canon_table : Qnil),
			  posix,
			  STRING_MULTIBYTE (string));
  immediate_quit = 1;
  re_match_object = string;

  val = re_search (bufp, (char *) SDATA (string),
		   SBYTES (string), pos_byte,
		   SBYTES (string) - pos_byte,
		   &search_regs);
  immediate_quit = 0;
  last_thing_searched = Qt;
  if (val == -2)
    matcher_overflow ();
  if (val < 0) return Qnil;

  for (i = 0; i < search_regs.num_regs; i++)
    if (search_regs.start[i] >= 0)
      {
	search_regs.start[i]
	  = string_byte_to_char (string, search_regs.start[i]);
	search_regs.end[i]
	  = string_byte_to_char (string, search_regs.end[i]);
      }

  return make_number (string_byte_to_char (string, val));
}

DEFUN ("string-match", Fstring_match, Sstring_match, 2, 3, 0,
       doc: /* Return index of start of first match for REGEXP in STRING, or nil.
Matching ignores case if `case-fold-search' is non-nil.
If third arg START is non-nil, start search at that index in STRING.
For index of first char beyond the match, do (match-end 0).
`match-end' and `match-beginning' also give indices of substrings
matched by parenthesis constructs in the pattern.

You can use the function `match-string' to extract the substrings
matched by the parenthesis constructions in REGEXP. */)
     (regexp, string, start)
     Lisp_Object regexp, string, start;
{
  return string_match_1 (regexp, string, start, 0);
}

DEFUN ("posix-string-match", Fposix_string_match, Sposix_string_match, 2, 3, 0,
       doc: /* Return index of start of first match for REGEXP in STRING, or nil.
Find the longest match, in accord with Posix regular expression rules.
Case is ignored if `case-fold-search' is non-nil in the current buffer.
If third arg START is non-nil, start search at that index in STRING.
For index of first char beyond the match, do (match-end 0).
`match-end' and `match-beginning' also give indices of substrings
matched by parenthesis constructs in the pattern.  */)
     (regexp, string, start)
     Lisp_Object regexp, string, start;
{
  return string_match_1 (regexp, string, start, 1);
}

/* Match REGEXP against STRING, searching all of STRING,
   and return the index of the match, or negative on failure.
   This does not clobber the match data.  */

int
fast_string_match (regexp, string)
     Lisp_Object regexp, string;
{
  int val;
  struct re_pattern_buffer *bufp;

  bufp = compile_pattern (regexp, 0, Qnil,
			  0, STRING_MULTIBYTE (string));
  immediate_quit = 1;
  re_match_object = string;

  val = re_search (bufp, (char *) SDATA (string),
		   SBYTES (string), 0,
		   SBYTES (string), 0);
  immediate_quit = 0;
  return val;
}

/* Match REGEXP against STRING, searching all of STRING ignoring case,
   and return the index of the match, or negative on failure.
   This does not clobber the match data.
   We assume that STRING contains single-byte characters.  */

extern Lisp_Object Vascii_downcase_table;

int
fast_c_string_match_ignore_case (regexp, string)
     Lisp_Object regexp;
     const char *string;
{
  int val;
  struct re_pattern_buffer *bufp;
  int len = strlen (string);

  regexp = string_make_unibyte (regexp);
  re_match_object = Qt;
  bufp = compile_pattern (regexp, 0,
			  Vascii_canon_table, 0,
			  0);
  immediate_quit = 1;
  val = re_search (bufp, string, len, 0, len, 0);
  immediate_quit = 0;
  return val;
}

/* Like fast_string_match but ignore case.  */

int
fast_string_match_ignore_case (regexp, string)
     Lisp_Object regexp, string;
{
  int val;
  struct re_pattern_buffer *bufp;

  bufp = compile_pattern (regexp, 0, Vascii_canon_table,
			  0, STRING_MULTIBYTE (string));
  immediate_quit = 1;
  re_match_object = string;

  val = re_search (bufp, (char *) SDATA (string),
		   SBYTES (string), 0,
		   SBYTES (string), 0);
  immediate_quit = 0;
  return val;
}

/* The newline cache: remembering which sections of text have no newlines.  */

/* If the user has requested newline caching, make sure it's on.
   Otherwise, make sure it's off.
   This is our cheezy way of associating an action with the change of
   state of a buffer-local variable.  */
static void
newline_cache_on_off (buf)
     struct buffer *buf;
{
  if (NILP (buf->cache_long_line_scans))
    {
      /* It should be off.  */
      if (buf->newline_cache)
        {
          free_region_cache (buf->newline_cache);
          buf->newline_cache = 0;
        }
    }
  else
    {
      /* It should be on.  */
      if (buf->newline_cache == 0)
        buf->newline_cache = new_region_cache ();
    }
}


/* Search for COUNT instances of the character TARGET between START and END.

   If COUNT is positive, search forwards; END must be >= START.
   If COUNT is negative, search backwards for the -COUNTth instance;
      END must be <= START.
   If COUNT is zero, do anything you please; run rogue, for all I care.

   If END is zero, use BEGV or ZV instead, as appropriate for the
   direction indicated by COUNT.

   If we find COUNT instances, set *SHORTAGE to zero, and return the
   position past the COUNTth match.  Note that for reverse motion
   this is not the same as the usual convention for Emacs motion commands.

   If we don't find COUNT instances before reaching END, set *SHORTAGE
   to the number of TARGETs left unfound, and return END.

   If ALLOW_QUIT is non-zero, set immediate_quit.  That's good to do
   except when inside redisplay.  */

int
scan_buffer (target, start, end, count, shortage, allow_quit)
     register int target;
     int start, end;
     int count;
     int *shortage;
     int allow_quit;
{
  struct region_cache *newline_cache;
  int direction;

  if (count > 0)
    {
      direction = 1;
      if (! end) end = ZV;
    }
  else
    {
      direction = -1;
      if (! end) end = BEGV;
    }

  newline_cache_on_off (current_buffer);
  newline_cache = current_buffer->newline_cache;

  if (shortage != 0)
    *shortage = 0;

  immediate_quit = allow_quit;

  if (count > 0)
    while (start != end)
      {
        /* Our innermost scanning loop is very simple; it doesn't know
           about gaps, buffer ends, or the newline cache.  ceiling is
           the position of the last character before the next such
           obstacle --- the last character the dumb search loop should
           examine.  */
	int ceiling_byte = CHAR_TO_BYTE (end) - 1;
	int start_byte = CHAR_TO_BYTE (start);
	int tem;

        /* If we're looking for a newline, consult the newline cache
           to see where we can avoid some scanning.  */
        if (target == '\n' && newline_cache)
          {
            int next_change;
            immediate_quit = 0;
            while (region_cache_forward
                   (current_buffer, newline_cache, start_byte, &next_change))
              start_byte = next_change;
            immediate_quit = allow_quit;

            /* START should never be after END.  */
            if (start_byte > ceiling_byte)
              start_byte = ceiling_byte;

            /* Now the text after start is an unknown region, and
               next_change is the position of the next known region. */
            ceiling_byte = min (next_change - 1, ceiling_byte);
          }

        /* The dumb loop can only scan text stored in contiguous
           bytes. BUFFER_CEILING_OF returns the last character
           position that is contiguous, so the ceiling is the
           position after that.  */
	tem = BUFFER_CEILING_OF (start_byte);
	ceiling_byte = min (tem, ceiling_byte);

        {
          /* The termination address of the dumb loop.  */
          register unsigned char *ceiling_addr
	    = BYTE_POS_ADDR (ceiling_byte) + 1;
          register unsigned char *cursor
	    = BYTE_POS_ADDR (start_byte);
          unsigned char *base = cursor;

          while (cursor < ceiling_addr)
            {
              unsigned char *scan_start = cursor;

              /* The dumb loop.  */
              while (*cursor != target && ++cursor < ceiling_addr)
                ;

              /* If we're looking for newlines, cache the fact that
                 the region from start to cursor is free of them. */
              if (target == '\n' && newline_cache)
                know_region_cache (current_buffer, newline_cache,
                                   start_byte + scan_start - base,
                                   start_byte + cursor - base);

              /* Did we find the target character?  */
              if (cursor < ceiling_addr)
                {
                  if (--count == 0)
                    {
                      immediate_quit = 0;
                      return BYTE_TO_CHAR (start_byte + cursor - base + 1);
                    }
                  cursor++;
                }
            }

          start = BYTE_TO_CHAR (start_byte + cursor - base);
        }
      }
  else
    while (start > end)
      {
        /* The last character to check before the next obstacle.  */
	int ceiling_byte = CHAR_TO_BYTE (end);
	int start_byte = CHAR_TO_BYTE (start);
	int tem;

        /* Consult the newline cache, if appropriate.  */
        if (target == '\n' && newline_cache)
          {
            int next_change;
            immediate_quit = 0;
            while (region_cache_backward
                   (current_buffer, newline_cache, start_byte, &next_change))
              start_byte = next_change;
            immediate_quit = allow_quit;

            /* Start should never be at or before end.  */
            if (start_byte <= ceiling_byte)
              start_byte = ceiling_byte + 1;

            /* Now the text before start is an unknown region, and
               next_change is the position of the next known region. */
            ceiling_byte = max (next_change, ceiling_byte);
          }

        /* Stop scanning before the gap.  */
	tem = BUFFER_FLOOR_OF (start_byte - 1);
	ceiling_byte = max (tem, ceiling_byte);

        {
          /* The termination address of the dumb loop.  */
          register unsigned char *ceiling_addr = BYTE_POS_ADDR (ceiling_byte);
          register unsigned char *cursor = BYTE_POS_ADDR (start_byte - 1);
          unsigned char *base = cursor;

          while (cursor >= ceiling_addr)
            {
              unsigned char *scan_start = cursor;

              while (*cursor != target && --cursor >= ceiling_addr)
                ;

              /* If we're looking for newlines, cache the fact that
                 the region from after the cursor to start is free of them.  */
              if (target == '\n' && newline_cache)
                know_region_cache (current_buffer, newline_cache,
                                   start_byte + cursor - base,
                                   start_byte + scan_start - base);

              /* Did we find the target character?  */
              if (cursor >= ceiling_addr)
                {
                  if (++count >= 0)
                    {
                      immediate_quit = 0;
                      return BYTE_TO_CHAR (start_byte + cursor - base);
                    }
                  cursor--;
                }
            }

	  start = BYTE_TO_CHAR (start_byte + cursor - base);
        }
      }

  immediate_quit = 0;
  if (shortage != 0)
    *shortage = count * direction;
  return start;
}

/* Search for COUNT instances of a line boundary, which means either a
   newline or (if selective display enabled) a carriage return.
   Start at START.  If COUNT is negative, search backwards.

   We report the resulting position by calling TEMP_SET_PT_BOTH.

   If we find COUNT instances. we position after (always after,
   even if scanning backwards) the COUNTth match, and return 0.

   If we don't find COUNT instances before reaching the end of the
   buffer (or the beginning, if scanning backwards), we return
   the number of line boundaries left unfound, and position at
   the limit we bumped up against.

   If ALLOW_QUIT is non-zero, set immediate_quit.  That's good to do
   except in special cases.  */

int
scan_newline (start, start_byte, limit, limit_byte, count, allow_quit)
     int start, start_byte;
     int limit, limit_byte;
     register int count;
     int allow_quit;
{
  int direction = ((count > 0) ? 1 : -1);

  register unsigned char *cursor;
  unsigned char *base;

  register int ceiling;
  register unsigned char *ceiling_addr;

  int old_immediate_quit = immediate_quit;

  /* The code that follows is like scan_buffer
     but checks for either newline or carriage return.  */

  if (allow_quit)
    immediate_quit++;

  start_byte = CHAR_TO_BYTE (start);

  if (count > 0)
    {
      while (start_byte < limit_byte)
	{
	  ceiling =  BUFFER_CEILING_OF (start_byte);
	  ceiling = min (limit_byte - 1, ceiling);
	  ceiling_addr = BYTE_POS_ADDR (ceiling) + 1;
	  base = (cursor = BYTE_POS_ADDR (start_byte));
	  while (1)
	    {
	      while (*cursor != '\n' && ++cursor != ceiling_addr)
		;

	      if (cursor != ceiling_addr)
		{
		  if (--count == 0)
		    {
		      immediate_quit = old_immediate_quit;
		      start_byte = start_byte + cursor - base + 1;
		      start = BYTE_TO_CHAR (start_byte);
		      TEMP_SET_PT_BOTH (start, start_byte);
		      return 0;
		    }
		  else
		    if (++cursor == ceiling_addr)
		      break;
		}
	      else
		break;
	    }
	  start_byte += cursor - base;
	}
    }
  else
    {
      while (start_byte > limit_byte)
	{
	  ceiling = BUFFER_FLOOR_OF (start_byte - 1);
	  ceiling = max (limit_byte, ceiling);
	  ceiling_addr = BYTE_POS_ADDR (ceiling) - 1;
	  base = (cursor = BYTE_POS_ADDR (start_byte - 1) + 1);
	  while (1)
	    {
	      while (--cursor != ceiling_addr && *cursor != '\n')
		;

	      if (cursor != ceiling_addr)
		{
		  if (++count == 0)
		    {
		      immediate_quit = old_immediate_quit;
		      /* Return the position AFTER the match we found.  */
		      start_byte = start_byte + cursor - base + 1;
		      start = BYTE_TO_CHAR (start_byte);
		      TEMP_SET_PT_BOTH (start, start_byte);
		      return 0;
		    }
		}
	      else
		break;
	    }
	  /* Here we add 1 to compensate for the last decrement
	     of CURSOR, which took it past the valid range.  */
	  start_byte += cursor - base + 1;
	}
    }

  TEMP_SET_PT_BOTH (limit, limit_byte);
  immediate_quit = old_immediate_quit;

  return count * direction;
}

int
find_next_newline_no_quit (from, cnt)
     register int from, cnt;
{
  return scan_buffer ('\n', from, 0, cnt, (int *) 0, 0);
}

/* Like find_next_newline, but returns position before the newline,
   not after, and only search up to TO.  This isn't just
   find_next_newline (...)-1, because you might hit TO.  */

int
find_before_next_newline (from, to, cnt)
     int from, to, cnt;
{
  int shortage;
  int pos = scan_buffer ('\n', from, to, cnt, &shortage, 1);

  if (shortage == 0)
    pos--;

  return pos;
}

/* Subroutines of Lisp buffer search functions. */

static Lisp_Object
search_command (string, bound, noerror, count, direction, RE, posix)
     Lisp_Object string, bound, noerror, count;
     int direction;
     int RE;
     int posix;
{
  register int np;
  int lim, lim_byte;
  int n = direction;

  if (!NILP (count))
    {
      CHECK_NUMBER (count);
      n *= XINT (count);
    }

  CHECK_STRING (string);
  if (NILP (bound))
    {
      if (n > 0)
	lim = ZV, lim_byte = ZV_BYTE;
      else
	lim = BEGV, lim_byte = BEGV_BYTE;
    }
  else
    {
      CHECK_NUMBER_COERCE_MARKER (bound);
      lim = XINT (bound);
      if (n > 0 ? lim < PT : lim > PT)
	error ("Invalid search bound (wrong side of point)");
      if (lim > ZV)
	lim = ZV, lim_byte = ZV_BYTE;
      else if (lim < BEGV)
	lim = BEGV, lim_byte = BEGV_BYTE;
      else
	lim_byte = CHAR_TO_BYTE (lim);
    }

  /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
  XCHAR_TABLE (current_buffer->case_canon_table)->extras[2]
    = current_buffer->case_eqv_table;

  np = search_buffer (string, PT, PT_BYTE, lim, lim_byte, n, RE,
		      (!NILP (current_buffer->case_fold_search)
		       ? current_buffer->case_canon_table
		       : Qnil),
		      (!NILP (current_buffer->case_fold_search)
		       ? current_buffer->case_eqv_table
		       : Qnil),
		      posix);
  if (np <= 0)
    {
      if (NILP (noerror))
	xsignal1 (Qsearch_failed, string);

      if (!EQ (noerror, Qt))
	{
	  if (lim < BEGV || lim > ZV)
	    abort ();
	  SET_PT_BOTH (lim, lim_byte);
	  return Qnil;
#if 0 /* This would be clean, but maybe programs depend on
	 a value of nil here.  */
	  np = lim;
#endif
	}
      else
	return Qnil;
    }

  if (np < BEGV || np > ZV)
    abort ();

  SET_PT (np);

  return make_number (np);
}

/* Return 1 if REGEXP it matches just one constant string.  */

static int
trivial_regexp_p (regexp)
     Lisp_Object regexp;
{
  int len = SBYTES (regexp);
  unsigned char *s = SDATA (regexp);
  while (--len >= 0)
    {
      switch (*s++)
	{
	case '.': case '*': case '+': case '?': case '[': case '^': case '$':
	  return 0;
	case '\\':
	  if (--len < 0)
	    return 0;
	  switch (*s++)
	    {
	    case '|': case '(': case ')': case '`': case '\'': case 'b':
	    case 'B': case '<': case '>': case 'w': case 'W': case 's':
	    case 'S': case '=': case '{': case '}': case '_':
	    case 'c': case 'C':	/* for categoryspec and notcategoryspec */
	    case '1': case '2': case '3': case '4': case '5':
	    case '6': case '7': case '8': case '9':
	      return 0;
	    }
	}
    }
  return 1;
}

/* Search for the n'th occurrence of STRING in the current buffer,
   starting at position POS and stopping at position LIM,
   treating STRING as a literal string if RE is false or as
   a regular expression if RE is true.

   If N is positive, searching is forward and LIM must be greater than POS.
   If N is negative, searching is backward and LIM must be less than POS.

   Returns -x if x occurrences remain to be found (x > 0),
   or else the position at the beginning of the Nth occurrence
   (if searching backward) or the end (if searching forward).

   POSIX is nonzero if we want full backtracking (POSIX style)
   for this pattern.  0 means backtrack only enough to get a valid match.  */

#define TRANSLATE(out, trt, d)			\
do						\
  {						\
    if (! NILP (trt))				\
      {						\
	Lisp_Object temp;			\
	temp = Faref (trt, make_number (d));	\
	if (INTEGERP (temp))			\
	  out = XINT (temp);			\
	else					\
	  out = d;				\
      }						\
    else					\
      out = d;					\
  }						\
while (0)

static int
search_buffer (string, pos, pos_byte, lim, lim_byte, n,
	       RE, trt, inverse_trt, posix)
     Lisp_Object string;
     int pos;
     int pos_byte;
     int lim;
     int lim_byte;
     int n;
     int RE;
     Lisp_Object trt;
     Lisp_Object inverse_trt;
     int posix;
{
  int len = SCHARS (string);
  int len_byte = SBYTES (string);
  register int i;

  if (running_asynch_code)
    save_search_regs ();

  /* Searching 0 times means don't move.  */
  /* Null string is found at starting position.  */
  if (len == 0 || n == 0)
    {
      set_search_regs (pos_byte, 0);
      return pos;
    }

  if (RE && !(trivial_regexp_p (string) && NILP (Vsearch_spaces_regexp)))
    {
      unsigned char *p1, *p2;
      int s1, s2;
      struct re_pattern_buffer *bufp;

      bufp = compile_pattern (string, &search_regs, trt, posix,
			      !NILP (current_buffer->enable_multibyte_characters));

      immediate_quit = 1;	/* Quit immediately if user types ^G,
				   because letting this function finish
				   can take too long. */
      QUIT;			/* Do a pending quit right away,
				   to avoid paradoxical behavior */
      /* Get pointers and sizes of the two strings
	 that make up the visible portion of the buffer. */

      p1 = BEGV_ADDR;
      s1 = GPT_BYTE - BEGV_BYTE;
      p2 = GAP_END_ADDR;
      s2 = ZV_BYTE - GPT_BYTE;
      if (s1 < 0)
	{
	  p2 = p1;
	  s2 = ZV_BYTE - BEGV_BYTE;
	  s1 = 0;
	}
      if (s2 < 0)
	{
	  s1 = ZV_BYTE - BEGV_BYTE;
	  s2 = 0;
	}
      re_match_object = Qnil;

      while (n < 0)
	{
	  int val;
	  val = re_search_2 (bufp, (char *) p1, s1, (char *) p2, s2,
			     pos_byte - BEGV_BYTE, lim_byte - pos_byte,
			     &search_regs,
			     /* Don't allow match past current point */
			     pos_byte - BEGV_BYTE);
	  if (val == -2)
	    {
	      matcher_overflow ();
	    }
	  if (val >= 0)
	    {
	      pos_byte = search_regs.start[0] + BEGV_BYTE;
	      for (i = 0; i < search_regs.num_regs; i++)
		if (search_regs.start[i] >= 0)
		  {
		    search_regs.start[i]
		      = BYTE_TO_CHAR (search_regs.start[i] + BEGV_BYTE);
		    search_regs.end[i]
		      = BYTE_TO_CHAR (search_regs.end[i] + BEGV_BYTE);
		  }
	      XSETBUFFER (last_thing_searched, current_buffer);
	      /* Set pos to the new position. */
	      pos = search_regs.start[0];
	    }
	  else
	    {
	      immediate_quit = 0;
	      return (n);
	    }
	  n++;
	}
      while (n > 0)
	{
	  int val;
	  val = re_search_2 (bufp, (char *) p1, s1, (char *) p2, s2,
			     pos_byte - BEGV_BYTE, lim_byte - pos_byte,
			     &search_regs,
			     lim_byte - BEGV_BYTE);
	  if (val == -2)
	    {
	      matcher_overflow ();
	    }
	  if (val >= 0)
	    {
	      pos_byte = search_regs.end[0] + BEGV_BYTE;
	      for (i = 0; i < search_regs.num_regs; i++)
		if (search_regs.start[i] >= 0)
		  {
		    search_regs.start[i]
		      = BYTE_TO_CHAR (search_regs.start[i] + BEGV_BYTE);
		    search_regs.end[i]
		      = BYTE_TO_CHAR (search_regs.end[i] + BEGV_BYTE);
		  }
	      XSETBUFFER (last_thing_searched, current_buffer);
	      pos = search_regs.end[0];
	    }
	  else
	    {
	      immediate_quit = 0;
	      return (0 - n);
	    }
	  n--;
	}
      immediate_quit = 0;
      return (pos);
    }
  else				/* non-RE case */
    {
      unsigned char *raw_pattern, *pat;
      int raw_pattern_size;
      int raw_pattern_size_byte;
      unsigned char *patbuf;
      int multibyte = !NILP (current_buffer->enable_multibyte_characters);
      unsigned char *base_pat;
      /* Set to positive if we find a non-ASCII char that need
	 translation.  Otherwise set to zero later.  */
      int charset_base = -1;
      int boyer_moore_ok = 1;

      /* MULTIBYTE says whether the text to be searched is multibyte.
	 We must convert PATTERN to match that, or we will not really
	 find things right.  */

      if (multibyte == STRING_MULTIBYTE (string))
	{
	  raw_pattern = (unsigned char *) SDATA (string);
	  raw_pattern_size = SCHARS (string);
	  raw_pattern_size_byte = SBYTES (string);
	}
      else if (multibyte)
	{
	  raw_pattern_size = SCHARS (string);
	  raw_pattern_size_byte
	    = count_size_as_multibyte (SDATA (string),
				       raw_pattern_size);
	  raw_pattern = (unsigned char *) alloca (raw_pattern_size_byte + 1);
	  copy_text (SDATA (string), raw_pattern,
		     SCHARS (string), 0, 1);
	}
      else
	{
	  /* Converting multibyte to single-byte.

	     ??? Perhaps this conversion should be done in a special way
	     by subtracting nonascii-insert-offset from each non-ASCII char,
	     so that only the multibyte chars which really correspond to
	     the chosen single-byte character set can possibly match.  */
	  raw_pattern_size = SCHARS (string);
	  raw_pattern_size_byte = SCHARS (string);
	  raw_pattern = (unsigned char *) alloca (raw_pattern_size + 1);
	  copy_text (SDATA (string), raw_pattern,
		     SBYTES (string), 1, 0);
	}

      /* Copy and optionally translate the pattern.  */
      len = raw_pattern_size;
      len_byte = raw_pattern_size_byte;
      patbuf = (unsigned char *) alloca (len_byte);
      pat = patbuf;
      base_pat = raw_pattern;
      if (multibyte)
	{
	  /* Fill patbuf by translated characters in STRING while
	     checking if we can use boyer-moore search.  If TRT is
	     non-nil, we can use boyer-moore search only if TRT can be
	     represented by the byte array of 256 elements.  For that,
	     all non-ASCII case-equivalents of all case-senstive
	     characters in STRING must belong to the same charset and
	     row.  */

	  while (--len >= 0)
	    {
	      unsigned char str_base[MAX_MULTIBYTE_LENGTH], *str;
	      int c, translated, inverse;
	      int in_charlen, charlen;

	      /* If we got here and the RE flag is set, it's because we're
		 dealing with a regexp known to be trivial, so the backslash
		 just quotes the next character.  */
	      if (RE && *base_pat == '\\')
		{
		  len--;
		  raw_pattern_size--;
		  len_byte--;
		  base_pat++;
		}

	      c = STRING_CHAR_AND_LENGTH (base_pat, len_byte, in_charlen);

	      if (NILP (trt))
		{
		  str = base_pat;
		  charlen = in_charlen;
		}
	      else
		{
		  /* Translate the character.  */
		  TRANSLATE (translated, trt, c);
		  charlen = CHAR_STRING (translated, str_base);
		  str = str_base;

		  /* Check if C has any other case-equivalents.  */
		  TRANSLATE (inverse, inverse_trt, c);
		  /* If so, check if we can use boyer-moore.  */
		  if (c != inverse && boyer_moore_ok)
		    {
		      /* Check if all equivalents belong to the same
			 charset & row.  Note that the check of C
			 itself is done by the last iteration.  Note
			 also that we don't have to check ASCII
			 characters because boyer-moore search can
			 always handle their translation.  */
		      while (1)
			{
			  if (ASCII_BYTE_P (inverse))
			    {
			      if (charset_base > 0)
				{
				  boyer_moore_ok = 0;
				  break;
				}
			      charset_base = 0;
			    }
			  else if (SINGLE_BYTE_CHAR_P (inverse))
			    {
			      /* Boyer-moore search can't handle a
				 translation of an eight-bit
				 character.  */
			      boyer_moore_ok = 0;
			      break;
			    }
			  else if (charset_base < 0)
			    charset_base = inverse & ~CHAR_FIELD3_MASK;
			  else if ((inverse & ~CHAR_FIELD3_MASK)
				   != charset_base)
			    {
			      boyer_moore_ok = 0;
			      break;
			    }
			  if (c == inverse)
			    break;
			  TRANSLATE (inverse, inverse_trt, inverse);
			}
		    }
		}
	      if (charset_base < 0)
		charset_base = 0;

	      /* Store this character into the translated pattern.  */
	      bcopy (str, pat, charlen);
	      pat += charlen;
	      base_pat += in_charlen;
	      len_byte -= in_charlen;
	    }
	}
      else
	{
	  /* Unibyte buffer.  */
	  charset_base = 0;
	  while (--len >= 0)
	    {
	      int c, translated;

	      /* If we got here and the RE flag is set, it's because we're
		 dealing with a regexp known to be trivial, so the backslash
		 just quotes the next character.  */
	      if (RE && *base_pat == '\\')
		{
		  len--;
		  raw_pattern_size--;
		  base_pat++;
		}
	      c = *base_pat++;
	      TRANSLATE (translated, trt, c);
	      *pat++ = translated;
	    }
	}

      len_byte = pat - patbuf;
      len = raw_pattern_size;
      pat = base_pat = patbuf;

      if (boyer_moore_ok)
	return boyer_moore (n, pat, len, len_byte, trt, inverse_trt,
			    pos, pos_byte, lim, lim_byte,
			    charset_base);
      else
	return simple_search (n, pat, len, len_byte, trt,
			      pos, pos_byte, lim, lim_byte);
    }
}

/* Do a simple string search N times for the string PAT,
   whose length is LEN/LEN_BYTE,
   from buffer position POS/POS_BYTE until LIM/LIM_BYTE.
   TRT is the translation table.

   Return the character position where the match is found.
   Otherwise, if M matches remained to be found, return -M.

   This kind of search works regardless of what is in PAT and
   regardless of what is in TRT.  It is used in cases where
   boyer_moore cannot work.  */

static int
simple_search (n, pat, len, len_byte, trt, pos, pos_byte, lim, lim_byte)
     int n;
     unsigned char *pat;
     int len, len_byte;
     Lisp_Object trt;
     int pos, pos_byte;
     int lim, lim_byte;
{
  int multibyte = ! NILP (current_buffer->enable_multibyte_characters);
  int forward = n > 0;

  if (lim > pos && multibyte)
    while (n > 0)
      {
	while (1)
	  {
	    /* Try matching at position POS.  */
	    int this_pos = pos;
	    int this_pos_byte = pos_byte;
	    int this_len = len;
	    int this_len_byte = len_byte;
	    unsigned char *p = pat;
	    if (pos + len > lim)
	      goto stop;

	    while (this_len > 0)
	      {
		int charlen, buf_charlen;
		int pat_ch, buf_ch;

		pat_ch = STRING_CHAR_AND_LENGTH (p, this_len_byte, charlen);
		buf_ch = STRING_CHAR_AND_LENGTH (BYTE_POS_ADDR (this_pos_byte),
						 ZV_BYTE - this_pos_byte,
						 buf_charlen);
		TRANSLATE (buf_ch, trt, buf_ch);

		if (buf_ch != pat_ch)
		  break;

		this_len_byte -= charlen;
		this_len--;
		p += charlen;

		this_pos_byte += buf_charlen;
		this_pos++;
	      }

	    if (this_len == 0)
	      {
		pos += len;
		pos_byte += len_byte;
		break;
	      }

	    INC_BOTH (pos, pos_byte);
	  }

	n--;
      }
  else if (lim > pos)
    while (n > 0)
      {
	while (1)
	  {
	    /* Try matching at position POS.  */
	    int this_pos = pos;
	    int this_len = len;
	    unsigned char *p = pat;

	    if (pos + len > lim)
	      goto stop;

	    while (this_len > 0)
	      {
		int pat_ch = *p++;
		int buf_ch = FETCH_BYTE (this_pos);
		TRANSLATE (buf_ch, trt, buf_ch);

		if (buf_ch != pat_ch)
		  break;

		this_len--;
		this_pos++;
	      }

	    if (this_len == 0)
	      {
		pos += len;
		break;
	      }

	    pos++;
	  }

	n--;
      }
  /* Backwards search.  */
  else if (lim < pos && multibyte)
    while (n < 0)
      {
	while (1)
	  {
	    /* Try matching at position POS.  */
	    int this_pos = pos - len;
	    int this_pos_byte = pos_byte - len_byte;
	    int this_len = len;
	    int this_len_byte = len_byte;
	    unsigned char *p = pat;

	    if (this_pos < lim || this_pos_byte < lim_byte)
	      goto stop;

	    while (this_len > 0)
	      {
		int charlen, buf_charlen;
		int pat_ch, buf_ch;

		pat_ch = STRING_CHAR_AND_LENGTH (p, this_len_byte, charlen);
		buf_ch = STRING_CHAR_AND_LENGTH (BYTE_POS_ADDR (this_pos_byte),
						 ZV_BYTE - this_pos_byte,
						 buf_charlen);
		TRANSLATE (buf_ch, trt, buf_ch);

		if (buf_ch != pat_ch)
		  break;

		this_len_byte -= charlen;
		this_len--;
		p += charlen;
		this_pos_byte += buf_charlen;
		this_pos++;
	      }

	    if (this_len == 0)
	      {
		pos -= len;
		pos_byte -= len_byte;
		break;
	      }

	    DEC_BOTH (pos, pos_byte);
	  }

	n++;
      }
  else if (lim < pos)
    while (n < 0)
      {
	while (1)
	  {
	    /* Try matching at position POS.  */
	    int this_pos = pos - len;
	    int this_len = len;
	    unsigned char *p = pat;

	    if (pos - len < lim)
	      goto stop;

	    while (this_len > 0)
	      {
		int pat_ch = *p++;
		int buf_ch = FETCH_BYTE (this_pos);
		TRANSLATE (buf_ch, trt, buf_ch);

		if (buf_ch != pat_ch)
		  break;
		this_len--;
		this_pos++;
	      }

	    if (this_len == 0)
	      {
		pos -= len;
		break;
	      }

	    pos--;
	  }

	n++;
      }

 stop:
  if (n == 0)
    {
      if (forward)
	set_search_regs ((multibyte ? pos_byte : pos) - len_byte, len_byte);
      else
	set_search_regs (multibyte ? pos_byte : pos, len_byte);

      return pos;
    }
  else if (n > 0)
    return -n;
  else
    return n;
}

/* Do Boyer-Moore search N times for the string BASE_PAT,
   whose length is LEN/LEN_BYTE,
   from buffer position POS/POS_BYTE until LIM/LIM_BYTE.
   DIRECTION says which direction we search in.
   TRT and INVERSE_TRT are translation tables.
   Characters in PAT are already translated by TRT.

   This kind of search works if all the characters in BASE_PAT that
   have nontrivial translation are the same aside from the last byte.
   This makes it possible to translate just the last byte of a
   character, and do so after just a simple test of the context.
   CHARSET_BASE is nonzero iff there is such a non-ASCII character.

   If that criterion is not satisfied, do not call this function.  */

static int
boyer_moore (n, base_pat, len, len_byte, trt, inverse_trt,
	     pos, pos_byte, lim, lim_byte, charset_base)
     int n;
     unsigned char *base_pat;
     int len, len_byte;
     Lisp_Object trt;
     Lisp_Object inverse_trt;
     int pos, pos_byte;
     int lim, lim_byte;
     int charset_base;
{
  int direction = ((n > 0) ? 1 : -1);
  register int dirlen;
  int infinity, limit, stride_for_teases = 0;
  register int *BM_tab;
  int *BM_tab_base;
  register unsigned char *cursor, *p_limit;
  register int i, j;
  unsigned char *pat, *pat_end;
  int multibyte = ! NILP (current_buffer->enable_multibyte_characters);

  unsigned char simple_translate[0400];
  /* These are set to the preceding bytes of a byte to be translated
     if charset_base is nonzero.  As the maximum byte length of a
     multibyte character is 4, we have to check at most three previous
     bytes.  */
  int translate_prev_byte1 = 0;
  int translate_prev_byte2 = 0;
  int translate_prev_byte3 = 0;

#ifdef C_ALLOCA
  int BM_tab_space[0400];
  BM_tab = &BM_tab_space[0];
#else
  BM_tab = (int *) alloca (0400 * sizeof (int));
#endif
  /* The general approach is that we are going to maintain that we know */
  /* the first (closest to the present position, in whatever direction */
  /* we're searching) character that could possibly be the last */
  /* (furthest from present position) character of a valid match.  We */
  /* advance the state of our knowledge by looking at that character */
  /* and seeing whether it indeed matches the last character of the */
  /* pattern.  If it does, we take a closer look.  If it does not, we */
  /* move our pointer (to putative last characters) as far as is */
  /* logically possible.  This amount of movement, which I call a */
  /* stride, will be the length of the pattern if the actual character */
  /* appears nowhere in the pattern, otherwise it will be the distance */
  /* from the last occurrence of that character to the end of the */
  /* pattern. */
  /* As a coding trick, an enormous stride is coded into the table for */
  /* characters that match the last character.  This allows use of only */
  /* a single test, a test for having gone past the end of the */
  /* permissible match region, to test for both possible matches (when */
  /* the stride goes past the end immediately) and failure to */
  /* match (where you get nudged past the end one stride at a time). */

  /* Here we make a "mickey mouse" BM table.  The stride of the search */
  /* is determined only by the last character of the putative match. */
  /* If that character does not match, we will stride the proper */
  /* distance to propose a match that superimposes it on the last */
  /* instance of a character that matches it (per trt), or misses */
  /* it entirely if there is none. */

  dirlen = len_byte * direction;
  infinity = dirlen - (lim_byte + pos_byte + len_byte + len_byte) * direction;

  /* Record position after the end of the pattern.  */
  pat_end = base_pat + len_byte;
  /* BASE_PAT points to a character that we start scanning from.
     It is the first character in a forward search,
     the last character in a backward search.  */
  if (direction < 0)
    base_pat = pat_end - 1;

  BM_tab_base = BM_tab;
  BM_tab += 0400;
  j = dirlen;		/* to get it in a register */
  /* A character that does not appear in the pattern induces a */
  /* stride equal to the pattern length. */
  while (BM_tab_base != BM_tab)
    {
      *--BM_tab = j;
      *--BM_tab = j;
      *--BM_tab = j;
      *--BM_tab = j;
    }

  /* We use this for translation, instead of TRT itself.
     We fill this in to handle the characters that actually
     occur in the pattern.  Others don't matter anyway!  */
  bzero (simple_translate, sizeof simple_translate);
  for (i = 0; i < 0400; i++)
    simple_translate[i] = i;

  if (charset_base)
    {
      /* Setup translate_prev_byte1/2/3 from CHARSET_BASE.  Only a
	 byte following them are the target of translation.  */
      int sample_char = charset_base | 0x20;
      unsigned char str[MAX_MULTIBYTE_LENGTH];
      int len = CHAR_STRING (sample_char, str);

      translate_prev_byte1 = str[len - 2];
      if (len > 2)
	{
	  translate_prev_byte2 = str[len - 3];
	  if (len > 3)
	    translate_prev_byte3 = str[len - 4];
	}
    }

  i = 0;
  while (i != infinity)
    {
      unsigned char *ptr = base_pat + i;
      i += direction;
      if (i == dirlen)
	i = infinity;
      if (! NILP (trt))
	{
	  /* If the byte currently looking at is the last of a
	     character to check case-equivalents, set CH to that
	     character.  An ASCII character and a non-ASCII character
	     matching with CHARSET_BASE are to be checked.  */
	  int ch = -1;

	  if (ASCII_BYTE_P (*ptr) || ! multibyte)
	    ch = *ptr;
	  else if (charset_base
		   && ((pat_end - ptr) == 1 || CHAR_HEAD_P (ptr[1])))
	    {
	      unsigned char *charstart = ptr - 1;

	      while (! (CHAR_HEAD_P (*charstart)))
		charstart--;
	      ch = STRING_CHAR (charstart, ptr - charstart + 1);
	      if (charset_base != (ch & ~CHAR_FIELD3_MASK))
		ch = -1;
	    }

	  if (ch >= 0400)
	    j = ((unsigned char) ch) | 0200;
	  else
	    j = *ptr;

	  if (i == infinity)
	    stride_for_teases = BM_tab[j];

	  BM_tab[j] = dirlen - i;
	  /* A translation table is accompanied by its inverse -- see */
	  /* comment following downcase_table for details */
	  if (ch >= 0)
	    {
	      int starting_ch = ch;
	      int starting_j = j;

	      while (1)
		{
		  TRANSLATE (ch, inverse_trt, ch);
		  if (ch >= 0400)
		    j = ((unsigned char) ch) | 0200;
		  else
		    j = (unsigned char) ch;

		  /* For all the characters that map into CH,
		     set up simple_translate to map the last byte
		     into STARTING_J.  */
		  simple_translate[j] = starting_j;
		  if (ch == starting_ch)
		    break;
		  BM_tab[j] = dirlen - i;
		}
	    }
	}
      else
	{
	  j = *ptr;

	  if (i == infinity)
	    stride_for_teases = BM_tab[j];
	  BM_tab[j] = dirlen - i;
	}
      /* stride_for_teases tells how much to stride if we get a */
      /* match on the far character but are subsequently */
      /* disappointed, by recording what the stride would have been */
      /* for that character if the last character had been */
      /* different. */
    }
  infinity = dirlen - infinity;
  pos_byte += dirlen - ((direction > 0) ? direction : 0);
  /* loop invariant - POS_BYTE points at where last char (first
     char if reverse) of pattern would align in a possible match.  */
  while (n != 0)
    {
      int tail_end;
      unsigned char *tail_end_ptr;

      /* It's been reported that some (broken) compiler thinks that
	 Boolean expressions in an arithmetic context are unsigned.
	 Using an explicit ?1:0 prevents this.  */
      if ((lim_byte - pos_byte - ((direction > 0) ? 1 : 0)) * direction
	  < 0)
	return (n * (0 - direction));
      /* First we do the part we can by pointers (maybe nothing) */
      QUIT;
      pat = base_pat;
      limit = pos_byte - dirlen + direction;
      if (direction > 0)
	{
	  limit = BUFFER_CEILING_OF (limit);
	  /* LIMIT is now the last (not beyond-last!) value POS_BYTE
	     can take on without hitting edge of buffer or the gap.  */
	  limit = min (limit, pos_byte + 20000);
	  limit = min (limit, lim_byte - 1);
	}
      else
	{
	  limit = BUFFER_FLOOR_OF (limit);
	  /* LIMIT is now the last (not beyond-last!) value POS_BYTE
	     can take on without hitting edge of buffer or the gap.  */
	  limit = max (limit, pos_byte - 20000);
	  limit = max (limit, lim_byte);
	}
      tail_end = BUFFER_CEILING_OF (pos_byte) + 1;
      tail_end_ptr = BYTE_POS_ADDR (tail_end);

      if ((limit - pos_byte) * direction > 20)
	{
	  unsigned char *p2;

	  p_limit = BYTE_POS_ADDR (limit);
	  p2 = (cursor = BYTE_POS_ADDR (pos_byte));
	  /* In this loop, pos + cursor - p2 is the surrogate for pos */
	  while (1)		/* use one cursor setting as long as i can */
	    {
	      if (direction > 0) /* worth duplicating */
		{
		  /* Use signed comparison if appropriate
		     to make cursor+infinity sure to be > p_limit.
		     Assuming that the buffer lies in a range of addresses
		     that are all "positive" (as ints) or all "negative",
		     either kind of comparison will work as long
		     as we don't step by infinity.  So pick the kind
		     that works when we do step by infinity.  */
		  if ((EMACS_INT) (p_limit + infinity) > (EMACS_INT) p_limit)
		    while ((EMACS_INT) cursor <= (EMACS_INT) p_limit)
		      cursor += BM_tab[*cursor];
		  else
		    while ((EMACS_UINT) cursor <= (EMACS_UINT) p_limit)
		      cursor += BM_tab[*cursor];
		}
	      else
		{
		  if ((EMACS_INT) (p_limit + infinity) < (EMACS_INT) p_limit)
		    while ((EMACS_INT) cursor >= (EMACS_INT) p_limit)
		      cursor += BM_tab[*cursor];
		  else
		    while ((EMACS_UINT) cursor >= (EMACS_UINT) p_limit)
		      cursor += BM_tab[*cursor];
		}
/* If you are here, cursor is beyond the end of the searched region. */
/* This can happen if you match on the far character of the pattern, */
/* because the "stride" of that character is infinity, a number able */
/* to throw you well beyond the end of the search.  It can also */
/* happen if you fail to match within the permitted region and would */
/* otherwise try a character beyond that region */
	      if ((cursor - p_limit) * direction <= len_byte)
		break;	/* a small overrun is genuine */
	      cursor -= infinity; /* large overrun = hit */
	      i = dirlen - direction;
	      if (! NILP (trt))
		{
		  while ((i -= direction) + direction != 0)
		    {
		      int ch;
		      cursor -= direction;
		      /* Translate only the last byte of a character.  */
		      if (! multibyte
			  || ((cursor == tail_end_ptr
			       || CHAR_HEAD_P (cursor[1]))
			      && (CHAR_HEAD_P (cursor[0])
				  /* Check if this is the last byte of
				     a translable character.  */
				  || (translate_prev_byte1 == cursor[-1]
				      && (CHAR_HEAD_P (translate_prev_byte1)
					  || (translate_prev_byte2 == cursor[-2]
					      && (CHAR_HEAD_P (translate_prev_byte2)
						  || (translate_prev_byte3 == cursor[-3]))))))))
			ch = simple_translate[*cursor];
		      else
			ch = *cursor;
		      if (pat[i] != ch)
			break;
		    }
		}
	      else
		{
		  while ((i -= direction) + direction != 0)
		    {
		      cursor -= direction;
		      if (pat[i] != *cursor)
			break;
		    }
		}
	      cursor += dirlen - i - direction;	/* fix cursor */
	      if (i + direction == 0)
		{
		  int position;

		  cursor -= direction;

		  position = pos_byte + cursor - p2 + ((direction > 0)
						       ? 1 - len_byte : 0);
		  set_search_regs (position, len_byte);

		  if ((n -= direction) != 0)
		    cursor += dirlen; /* to resume search */
		  else
		    return ((direction > 0)
			    ? search_regs.end[0] : search_regs.start[0]);
		}
	      else
		cursor += stride_for_teases; /* <sigh> we lose -  */
	    }
	  pos_byte += cursor - p2;
	}
      else
	/* Now we'll pick up a clump that has to be done the hard */
	/* way because it covers a discontinuity */
	{
	  limit = ((direction > 0)
		   ? BUFFER_CEILING_OF (pos_byte - dirlen + 1)
		   : BUFFER_FLOOR_OF (pos_byte - dirlen - 1));
	  limit = ((direction > 0)
		   ? min (limit + len_byte, lim_byte - 1)
		   : max (limit - len_byte, lim_byte));
	  /* LIMIT is now the last value POS_BYTE can have
	     and still be valid for a possible match.  */
	  while (1)
	    {
	      /* This loop can be coded for space rather than */
	      /* speed because it will usually run only once. */
	      /* (the reach is at most len + 21, and typically */
	      /* does not exceed len) */
	      while ((limit - pos_byte) * direction >= 0)
		pos_byte += BM_tab[FETCH_BYTE (pos_byte)];
	      /* now run the same tests to distinguish going off the */
	      /* end, a match or a phony match. */
	      if ((pos_byte - limit) * direction <= len_byte)
		break;	/* ran off the end */
	      /* Found what might be a match.
		 Set POS_BYTE back to last (first if reverse) pos.  */
	      pos_byte -= infinity;
	      i = dirlen - direction;
	      while ((i -= direction) + direction != 0)
		{
		  int ch;
		  unsigned char *ptr;
		  pos_byte -= direction;
		  ptr = BYTE_POS_ADDR (pos_byte);
		  /* Translate only the last byte of a character.  */
		  if (! multibyte
		      || ((ptr == tail_end_ptr
			   || CHAR_HEAD_P (ptr[1]))
			  && (CHAR_HEAD_P (ptr[0])
			      /* Check if this is the last byte of a
				 translable character.  */
			      || (translate_prev_byte1 == ptr[-1]
				  && (CHAR_HEAD_P (translate_prev_byte1)
				      || (translate_prev_byte2 == ptr[-2]
					  && (CHAR_HEAD_P (translate_prev_byte2)
					      || translate_prev_byte3 == ptr[-3])))))))
		    ch = simple_translate[*ptr];
		  else
		    ch = *ptr;
		  if (pat[i] != ch)
		    break;
		}
	      /* Above loop has moved POS_BYTE part or all the way
		 back to the first pos (last pos if reverse).
		 Set it once again at the last (first if reverse) char.  */
	      pos_byte += dirlen - i- direction;
	      if (i + direction == 0)
		{
		  int position;
		  pos_byte -= direction;

		  position = pos_byte + ((direction > 0) ? 1 - len_byte : 0);

		  set_search_regs (position, len_byte);

		  if ((n -= direction) != 0)
		    pos_byte += dirlen; /* to resume search */
		  else
		    return ((direction > 0)
			    ? search_regs.end[0] : search_regs.start[0]);
		}
	      else
		pos_byte += stride_for_teases;
	    }
	  }
      /* We have done one clump.  Can we continue? */
      if ((lim_byte - pos_byte) * direction < 0)
	return ((0 - n) * direction);
    }
  return BYTE_TO_CHAR (pos_byte);
}

/* Record beginning BEG_BYTE and end BEG_BYTE + NBYTES
   for the overall match just found in the current buffer.
   Also clear out the match data for registers 1 and up.  */

static void
set_search_regs (beg_byte, nbytes)
     int beg_byte, nbytes;
{
  int i;

  /* Make sure we have registers in which to store
     the match position.  */
  if (search_regs.num_regs == 0)
    {
      search_regs.start = (regoff_t *) xmalloc (2 * sizeof (regoff_t));
      search_regs.end = (regoff_t *) xmalloc (2 * sizeof (regoff_t));
      search_regs.num_regs = 2;
    }

  /* Clear out the other registers.  */
  for (i = 1; i < search_regs.num_regs; i++)
    {
      search_regs.start[i] = -1;
      search_regs.end[i] = -1;
    }

  search_regs.start[0] = BYTE_TO_CHAR (beg_byte);
  search_regs.end[0] = BYTE_TO_CHAR (beg_byte + nbytes);
  XSETBUFFER (last_thing_searched, current_buffer);
}

/* Given a string of words separated by word delimiters,
  compute a regexp that matches those exact words
  separated by arbitrary punctuation.  */

static Lisp_Object
wordify (string)
     Lisp_Object string;
{
  register unsigned char *p, *o;
  register int i, i_byte, len, punct_count = 0, word_count = 0;
  Lisp_Object val;
  int prev_c = 0;
  int adjust;

  CHECK_STRING (string);
  p = SDATA (string);
  len = SCHARS (string);

  for (i = 0, i_byte = 0; i < len; )
    {
      int c;

      FETCH_STRING_CHAR_ADVANCE (c, string, i, i_byte);

      if (SYNTAX (c) != Sword)
	{
	  punct_count++;
	  if (i > 0 && SYNTAX (prev_c) == Sword)
	    word_count++;
	}

      prev_c = c;
    }

  if (SYNTAX (prev_c) == Sword)
    word_count++;
  if (!word_count)
    return empty_string;

  adjust = - punct_count + 5 * (word_count - 1) + 4;
  if (STRING_MULTIBYTE (string))
    val = make_uninit_multibyte_string (len + adjust,
					SBYTES (string)
					+ adjust);
  else
    val = make_uninit_string (len + adjust);

  o = SDATA (val);
  *o++ = '\\';
  *o++ = 'b';
  prev_c = 0;

  for (i = 0, i_byte = 0; i < len; )
    {
      int c;
      int i_byte_orig = i_byte;

      FETCH_STRING_CHAR_ADVANCE (c, string, i, i_byte);

      if (SYNTAX (c) == Sword)
	{
	  bcopy (SDATA (string) + i_byte_orig, o,
		 i_byte - i_byte_orig);
	  o += i_byte - i_byte_orig;
	}
      else if (i > 0 && SYNTAX (prev_c) == Sword && --word_count)
	{
	  *o++ = '\\';
	  *o++ = 'W';
	  *o++ = '\\';
	  *o++ = 'W';
	  *o++ = '*';
	}

      prev_c = c;
    }

  *o++ = '\\';
  *o++ = 'b';

  return val;
}

DEFUN ("search-backward", Fsearch_backward, Ssearch_backward, 1, 4,
       "MSearch backward: ",
       doc: /* Search backward from point for STRING.
Set point to the beginning of the occurrence found, and return point.
An optional second argument bounds the search; it is a buffer position.
The match found must not extend before that position.
Optional third argument, if t, means if fail just return nil (no error).
 If not nil and not t, position at limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.

Search case-sensitivity is determined by the value of the variable
`case-fold-search', which see.

See also the functions `match-beginning', `match-end' and `replace-match'.  */)
     (string, bound, noerror, count)
     Lisp_Object string, bound, noerror, count;
{
  return search_command (string, bound, noerror, count, -1, 0, 0);
}

DEFUN ("search-forward", Fsearch_forward, Ssearch_forward, 1, 4, "MSearch: ",
       doc: /* Search forward from point for STRING.
Set point to the end of the occurrence found, and return point.
An optional second argument bounds the search; it is a buffer position.
The match found must not extend after that position.  A value of nil is
  equivalent to (point-max).
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.

Search case-sensitivity is determined by the value of the variable
`case-fold-search', which see.

See also the functions `match-beginning', `match-end' and `replace-match'.  */)
     (string, bound, noerror, count)
     Lisp_Object string, bound, noerror, count;
{
  return search_command (string, bound, noerror, count, 1, 0, 0);
}

DEFUN ("word-search-backward", Fword_search_backward, Sword_search_backward, 1, 4,
       "sWord search backward: ",
       doc: /* Search backward from point for STRING, ignoring differences in punctuation.
Set point to the beginning of the occurrence found, and return point.
An optional second argument bounds the search; it is a buffer position.
The match found must not extend before that position.
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.  */)
     (string, bound, noerror, count)
     Lisp_Object string, bound, noerror, count;
{
  return search_command (wordify (string), bound, noerror, count, -1, 1, 0);
}

DEFUN ("word-search-forward", Fword_search_forward, Sword_search_forward, 1, 4,
       "sWord search: ",
       doc: /* Search forward from point for STRING, ignoring differences in punctuation.
Set point to the end of the occurrence found, and return point.
An optional second argument bounds the search; it is a buffer position.
The match found must not extend after that position.
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.  */)
     (string, bound, noerror, count)
     Lisp_Object string, bound, noerror, count;
{
  return search_command (wordify (string), bound, noerror, count, 1, 1, 0);
}

DEFUN ("re-search-backward", Fre_search_backward, Sre_search_backward, 1, 4,
       "sRE search backward: ",
       doc: /* Search backward from point for match for regular expression REGEXP.
Set point to the beginning of the match, and return point.
The match found is the one starting last in the buffer
and yet ending before the origin of the search.
An optional second argument bounds the search; it is a buffer position.
The match found must start at or after that position.
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.
See also the functions `match-beginning', `match-end', `match-string',
and `replace-match'.  */)
     (regexp, bound, noerror, count)
     Lisp_Object regexp, bound, noerror, count;
{
  return search_command (regexp, bound, noerror, count, -1, 1, 0);
}

DEFUN ("re-search-forward", Fre_search_forward, Sre_search_forward, 1, 4,
       "sRE search: ",
       doc: /* Search forward from point for regular expression REGEXP.
Set point to the end of the occurrence found, and return point.
An optional second argument bounds the search; it is a buffer position.
The match found must not extend after that position.
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.
See also the functions `match-beginning', `match-end', `match-string',
and `replace-match'.  */)
     (regexp, bound, noerror, count)
     Lisp_Object regexp, bound, noerror, count;
{
  return search_command (regexp, bound, noerror, count, 1, 1, 0);
}

DEFUN ("posix-search-backward", Fposix_search_backward, Sposix_search_backward, 1, 4,
       "sPosix search backward: ",
       doc: /* Search backward from point for match for regular expression REGEXP.
Find the longest match in accord with Posix regular expression rules.
Set point to the beginning of the match, and return point.
The match found is the one starting last in the buffer
and yet ending before the origin of the search.
An optional second argument bounds the search; it is a buffer position.
The match found must start at or after that position.
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.
See also the functions `match-beginning', `match-end', `match-string',
and `replace-match'.  */)
     (regexp, bound, noerror, count)
     Lisp_Object regexp, bound, noerror, count;
{
  return search_command (regexp, bound, noerror, count, -1, 1, 1);
}

DEFUN ("posix-search-forward", Fposix_search_forward, Sposix_search_forward, 1, 4,
       "sPosix search: ",
       doc: /* Search forward from point for regular expression REGEXP.
Find the longest match in accord with Posix regular expression rules.
Set point to the end of the occurrence found, and return point.
An optional second argument bounds the search; it is a buffer position.
The match found must not extend after that position.
Optional third argument, if t, means if fail just return nil (no error).
  If not nil and not t, move to limit of search and return nil.
Optional fourth argument is repeat count--search for successive occurrences.
See also the functions `match-beginning', `match-end', `match-string',
and `replace-match'.  */)
     (regexp, bound, noerror, count)
     Lisp_Object regexp, bound, noerror, count;
{
  return search_command (regexp, bound, noerror, count, 1, 1, 1);
}

DEFUN ("replace-match", Freplace_match, Sreplace_match, 1, 5, 0,
       doc: /* Replace text matched by last search with NEWTEXT.
Leave point at the end of the replacement text.

If second arg FIXEDCASE is non-nil, do not alter case of replacement text.
Otherwise maybe capitalize the whole text, or maybe just word initials,
based on the replaced text.
If the replaced text has only capital letters
and has at least one multiletter word, convert NEWTEXT to all caps.
Otherwise if all words are capitalized in the replaced text,
capitalize each word in NEWTEXT.

If third arg LITERAL is non-nil, insert NEWTEXT literally.
Otherwise treat `\\' as special:
  `\\&' in NEWTEXT means substitute original matched text.
  `\\N' means substitute what matched the Nth `\\(...\\)'.
       If Nth parens didn't match, substitute nothing.
  `\\\\' means insert one `\\'.
Case conversion does not apply to these substitutions.

FIXEDCASE and LITERAL are optional arguments.

The optional fourth argument STRING can be a string to modify.
This is meaningful when the previous match was done against STRING,
using `string-match'.  When used this way, `replace-match'
creates and returns a new string made by copying STRING and replacing
the part of STRING that was matched.

The optional fifth argument SUBEXP specifies a subexpression;
it says to replace just that subexpression with NEWTEXT,
rather than replacing the entire matched text.
This is, in a vague sense, the inverse of using `\\N' in NEWTEXT;
`\\N' copies subexp N into NEWTEXT, but using N as SUBEXP puts
NEWTEXT in place of subexp N.
This is useful only after a regular expression search or match,
since only regular expressions have distinguished subexpressions.  */)
     (newtext, fixedcase, literal, string, subexp)
     Lisp_Object newtext, fixedcase, literal, string, subexp;
{
  enum { nochange, all_caps, cap_initial } case_action;
  register int pos, pos_byte;
  int some_multiletter_word;
  int some_lowercase;
  int some_uppercase;
  int some_nonuppercase_initial;
  register int c, prevc;
  int sub;
  int opoint, newpoint;

  CHECK_STRING (newtext);

  if (! NILP (string))
    CHECK_STRING (string);

  case_action = nochange;	/* We tried an initialization */
				/* but some C compilers blew it */

  if (search_regs.num_regs <= 0)
    error ("`replace-match' called before any match found");

  if (NILP (subexp))
    sub = 0;
  else
    {
      CHECK_NUMBER (subexp);
      sub = XINT (subexp);
      if (sub < 0 || sub >= search_regs.num_regs)
	args_out_of_range (subexp, make_number (search_regs.num_regs));
    }

  if (NILP (string))
    {
      if (search_regs.start[sub] < BEGV
	  || search_regs.start[sub] > search_regs.end[sub]
	  || search_regs.end[sub] > ZV)
	args_out_of_range (make_number (search_regs.start[sub]),
			   make_number (search_regs.end[sub]));
    }
  else
    {
      if (search_regs.start[sub] < 0
	  || search_regs.start[sub] > search_regs.end[sub]
	  || search_regs.end[sub] > SCHARS (string))
	args_out_of_range (make_number (search_regs.start[sub]),
			   make_number (search_regs.end[sub]));
    }

  if (NILP (fixedcase))
    {
      /* Decide how to casify by examining the matched text. */
      int last;

      pos = search_regs.start[sub];
      last = search_regs.end[sub];

      if (NILP (string))
	pos_byte = CHAR_TO_BYTE (pos);
      else
	pos_byte = string_char_to_byte (string, pos);

      prevc = '\n';
      case_action = all_caps;

      /* some_multiletter_word is set nonzero if any original word
	 is more than one letter long. */
      some_multiletter_word = 0;
      some_lowercase = 0;
      some_nonuppercase_initial = 0;
      some_uppercase = 0;

      while (pos < last)
	{
	  if (NILP (string))
	    {
	      c = FETCH_CHAR (pos_byte);
	      INC_BOTH (pos, pos_byte);
	    }
	  else
	    FETCH_STRING_CHAR_ADVANCE (c, string, pos, pos_byte);

	  if (LOWERCASEP (c))
	    {
	      /* Cannot be all caps if any original char is lower case */

	      some_lowercase = 1;
	      if (SYNTAX (prevc) != Sword)
		some_nonuppercase_initial = 1;
	      else
		some_multiletter_word = 1;
	    }
	  else if (UPPERCASEP (c))
	    {
	      some_uppercase = 1;
	      if (SYNTAX (prevc) != Sword)
		;
	      else
		some_multiletter_word = 1;
	    }
	  else
	    {
	      /* If the initial is a caseless word constituent,
		 treat that like a lowercase initial.  */
	      if (SYNTAX (prevc) != Sword)
		some_nonuppercase_initial = 1;
	    }

	  prevc = c;
	}

      /* Convert to all caps if the old text is all caps
	 and has at least one multiletter word.  */
      if (! some_lowercase && some_multiletter_word)
	case_action = all_caps;
      /* Capitalize each word, if the old text has all capitalized words.  */
      else if (!some_nonuppercase_initial && some_multiletter_word)
	case_action = cap_initial;
      else if (!some_nonuppercase_initial && some_uppercase)
	/* Should x -> yz, operating on X, give Yz or YZ?
	   We'll assume the latter.  */
	case_action = all_caps;
      else
	case_action = nochange;
    }

  /* Do replacement in a string.  */
  if (!NILP (string))
    {
      Lisp_Object before, after;

      before = Fsubstring (string, make_number (0),
			   make_number (search_regs.start[sub]));
      after = Fsubstring (string, make_number (search_regs.end[sub]), Qnil);

      /* Substitute parts of the match into NEWTEXT
	 if desired.  */
      if (NILP (literal))
	{
	  int lastpos = 0;
	  int lastpos_byte = 0;
	  /* We build up the substituted string in ACCUM.  */
	  Lisp_Object accum;
	  Lisp_Object middle;
	  int length = SBYTES (newtext);

	  accum = Qnil;

	  for (pos_byte = 0, pos = 0; pos_byte < length;)
	    {
	      int substart = -1;
	      int subend = 0;
	      int delbackslash = 0;

	      FETCH_STRING_CHAR_ADVANCE (c, newtext, pos, pos_byte);

	      if (c == '\\')
		{
		  FETCH_STRING_CHAR_ADVANCE (c, newtext, pos, pos_byte);

		  if (c == '&')
		    {
		      substart = search_regs.start[sub];
		      subend = search_regs.end[sub];
		    }
		  else if (c >= '1' && c <= '9')
		    {
		      if (search_regs.start[c - '0'] >= 0
			  && c <= search_regs.num_regs + '0')
			{
			  substart = search_regs.start[c - '0'];
			  subend = search_regs.end[c - '0'];
			}
		      else
			{
			  /* If that subexp did not match,
			     replace \\N with nothing.  */
			  substart = 0;
			  subend = 0;
			}
		    }
		  else if (c == '\\')
		    delbackslash = 1;
		  else
		    error ("Invalid use of `\\' in replacement text");
		}
	      if (substart >= 0)
		{
		  if (pos - 2 != lastpos)
		    middle = substring_both (newtext, lastpos,
					     lastpos_byte,
					     pos - 2, pos_byte - 2);
		  else
		    middle = Qnil;
		  accum = concat3 (accum, middle,
				   Fsubstring (string,
					       make_number (substart),
					       make_number (subend)));
		  lastpos = pos;
		  lastpos_byte = pos_byte;
		}
	      else if (delbackslash)
		{
		  middle = substring_both (newtext, lastpos,
					   lastpos_byte,
					   pos - 1, pos_byte - 1);

		  accum = concat2 (accum, middle);
		  lastpos = pos;
		  lastpos_byte = pos_byte;
		}
	    }

	  if (pos != lastpos)
	    middle = substring_both (newtext, lastpos,
				     lastpos_byte,
				     pos, pos_byte);
	  else
	    middle = Qnil;

	  newtext = concat2 (accum, middle);
	}

      /* Do case substitution in NEWTEXT if desired.  */
      if (case_action == all_caps)
	newtext = Fupcase (newtext);
      else if (case_action == cap_initial)
	newtext = Fupcase_initials (newtext);

      return concat3 (before, newtext, after);
    }

  /* Record point, then move (quietly) to the start of the match.  */
  if (PT >= search_regs.end[sub])
    opoint = PT - ZV;
  else if (PT > search_regs.start[sub])
    opoint = search_regs.end[sub] - ZV;
  else
    opoint = PT;

  /* If we want non-literal replacement,
     perform substitution on the replacement string.  */
  if (NILP (literal))
    {
      int length = SBYTES (newtext);
      unsigned char *substed;
      int substed_alloc_size, substed_len;
      int buf_multibyte = !NILP (current_buffer->enable_multibyte_characters);
      int str_multibyte = STRING_MULTIBYTE (newtext);
      Lisp_Object rev_tbl;
      int really_changed = 0;

      rev_tbl= (!buf_multibyte && CHAR_TABLE_P (Vnonascii_translation_table)
		? Fchar_table_extra_slot (Vnonascii_translation_table,
					  make_number (0))
		: Qnil);

      substed_alloc_size = length * 2 + 100;
      substed = (unsigned char *) xmalloc (substed_alloc_size + 1);
      substed_len = 0;

      /* Go thru NEWTEXT, producing the actual text to insert in
	 SUBSTED while adjusting multibyteness to that of the current
	 buffer.  */

      for (pos_byte = 0, pos = 0; pos_byte < length;)
	{
	  unsigned char str[MAX_MULTIBYTE_LENGTH];
	  unsigned char *add_stuff = NULL;
	  int add_len = 0;
	  int idx = -1;

	  if (str_multibyte)
	    {
	      FETCH_STRING_CHAR_ADVANCE_NO_CHECK (c, newtext, pos, pos_byte);
	      if (!buf_multibyte)
		c = multibyte_char_to_unibyte (c, rev_tbl);
	    }
	  else
	    {
	      /* Note that we don't have to increment POS.  */
	      c = SREF (newtext, pos_byte++);
	      if (buf_multibyte)
		c = unibyte_char_to_multibyte (c);
	    }

	  /* Either set ADD_STUFF and ADD_LEN to the text to put in SUBSTED,
	     or set IDX to a match index, which means put that part
	     of the buffer text into SUBSTED.  */

	  if (c == '\\')
	    {
	      really_changed = 1;

	      if (str_multibyte)
		{
		  FETCH_STRING_CHAR_ADVANCE_NO_CHECK (c, newtext,
						      pos, pos_byte);
		  if (!buf_multibyte && !SINGLE_BYTE_CHAR_P (c))
		    c = multibyte_char_to_unibyte (c, rev_tbl);
		}
	      else
		{
		  c = SREF (newtext, pos_byte++);
		  if (buf_multibyte)
		    c = unibyte_char_to_multibyte (c);
		}

	      if (c == '&')
		idx = sub;
	      else if (c >= '1' && c <= '9' && c <= search_regs.num_regs + '0')
		{
		  if (search_regs.start[c - '0'] >= 1)
		    idx = c - '0';
		}
	      else if (c == '\\')
		add_len = 1, add_stuff = "\\";
	      else
		{
		  xfree (substed);
		  error ("Invalid use of `\\' in replacement text");
		}
	    }
	  else
	    {
	      add_len = CHAR_STRING (c, str);
	      add_stuff = str;
	    }

	  /* If we want to copy part of a previous match,
	     set up ADD_STUFF and ADD_LEN to point to it.  */
	  if (idx >= 0)
	    {
	      int begbyte = CHAR_TO_BYTE (search_regs.start[idx]);
	      add_len = CHAR_TO_BYTE (search_regs.end[idx]) - begbyte;
	      if (search_regs.start[idx] < GPT && GPT < search_regs.end[idx])
		move_gap (search_regs.start[idx]);
	      add_stuff = BYTE_POS_ADDR (begbyte);
	    }

	  /* Now the stuff we want to add to SUBSTED
	     is invariably ADD_LEN bytes starting at ADD_STUFF.  */

	  /* Make sure SUBSTED is big enough.  */
	  if (substed_len + add_len >= substed_alloc_size)
	    {
	      substed_alloc_size = substed_len + add_len + 500;
	      substed = (unsigned char *) xrealloc (substed,
						    substed_alloc_size + 1);
	    }

	  /* Now add to the end of SUBSTED.  */
	  if (add_stuff)
	    {
	      bcopy (add_stuff, substed + substed_len, add_len);
	      substed_len += add_len;
	    }
	}

      if (really_changed)
	{
	  if (buf_multibyte)
	    {
	      int nchars = multibyte_chars_in_text (substed, substed_len);

	      newtext = make_multibyte_string (substed, nchars, substed_len);
	    }
	  else
	    newtext = make_unibyte_string (substed, substed_len);
	}
      xfree (substed);
    }

  /* Replace the old text with the new in the cleanest possible way.  */
  replace_range (search_regs.start[sub], search_regs.end[sub],
		 newtext, 1, 0, 1);
  newpoint = search_regs.start[sub] + SCHARS (newtext);

  if (case_action == all_caps)
    Fupcase_region (make_number (search_regs.start[sub]),
		    make_number (newpoint));
  else if (case_action == cap_initial)
    Fupcase_initials_region (make_number (search_regs.start[sub]),
			     make_number (newpoint));

  /* Adjust search data for this change.  */
  {
    int oldend = search_regs.end[sub];
    int oldstart = search_regs.start[sub];
    int change = newpoint - search_regs.end[sub];
    int i;

    for (i = 0; i < search_regs.num_regs; i++)
      {
	if (search_regs.start[i] >= oldend)
	  search_regs.start[i] += change;
	else if (search_regs.start[i] > oldstart)
	  search_regs.start[i] = oldstart;
	if (search_regs.end[i] >= oldend)
	  search_regs.end[i] += change;
	else if (search_regs.end[i] > oldstart)
	  search_regs.end[i] = oldstart;
      }
  }

  /* Put point back where it was in the text.  */
  if (opoint <= 0)
    TEMP_SET_PT (opoint + ZV);
  else
    TEMP_SET_PT (opoint);

  /* Now move point "officially" to the start of the inserted replacement.  */
  move_if_not_intangible (newpoint);

  return Qnil;
}

static Lisp_Object
match_limit (num, beginningp)
     Lisp_Object num;
     int beginningp;
{
  register int n;

  CHECK_NUMBER (num);
  n = XINT (num);
  if (n < 0)
    args_out_of_range (num, make_number (0));
  if (search_regs.num_regs <= 0)
    error ("No match data, because no search succeeded");
  if (n >= search_regs.num_regs
      || search_regs.start[n] < 0)
    return Qnil;
  return (make_number ((beginningp) ? search_regs.start[n]
		                    : search_regs.end[n]));
}

DEFUN ("match-beginning", Fmatch_beginning, Smatch_beginning, 1, 1, 0,
       doc: /* Return position of start of text matched by last search.
SUBEXP, a number, specifies which parenthesized expression in the last
  regexp.
Value is nil if SUBEXPth pair didn't match, or there were less than
  SUBEXP pairs.
Zero means the entire text matched by the whole regexp or whole string.  */)
     (subexp)
     Lisp_Object subexp;
{
  return match_limit (subexp, 1);
}

DEFUN ("match-end", Fmatch_end, Smatch_end, 1, 1, 0,
       doc: /* Return position of end of text matched by last search.
SUBEXP, a number, specifies which parenthesized expression in the last
  regexp.
Value is nil if SUBEXPth pair didn't match, or there were less than
  SUBEXP pairs.
Zero means the entire text matched by the whole regexp or whole string.  */)
     (subexp)
     Lisp_Object subexp;
{
  return match_limit (subexp, 0);
}

DEFUN ("match-data", Fmatch_data, Smatch_data, 0, 3, 0,
       doc: /* Return a list containing all info on what the last search matched.
Element 2N is `(match-beginning N)'; element 2N + 1 is `(match-end N)'.
All the elements are markers or nil (nil if the Nth pair didn't match)
if the last match was on a buffer; integers or nil if a string was matched.
Use `store-match-data' to reinstate the data in this list.

If INTEGERS (the optional first argument) is non-nil, always use
integers \(rather than markers) to represent buffer positions.  In
this case, and if the last match was in a buffer, the buffer will get
stored as one additional element at the end of the list.

If REUSE is a list, reuse it as part of the value.  If REUSE is long
enough to hold all the values, and if INTEGERS is non-nil, no consing
is done.

If optional third arg RESEAT is non-nil, any previous markers on the
REUSE list will be modified to point to nowhere.

Return value is undefined if the last search failed.  */)
  (integers, reuse, reseat)
     Lisp_Object integers, reuse, reseat;
{
  Lisp_Object tail, prev;
  Lisp_Object *data;
  int i, len;

  if (!NILP (reseat))
    for (tail = reuse; CONSP (tail); tail = XCDR (tail))
      if (MARKERP (XCAR (tail)))
	{
	  unchain_marker (XMARKER (XCAR (tail)));
	  XSETCAR (tail, Qnil);
	}

  if (NILP (last_thing_searched))
    return Qnil;

  prev = Qnil;

  data = (Lisp_Object *) alloca ((2 * search_regs.num_regs + 1)
				 * sizeof (Lisp_Object));

  len = 0;
  for (i = 0; i < search_regs.num_regs; i++)
    {
      int start = search_regs.start[i];
      if (start >= 0)
	{
	  if (EQ (last_thing_searched, Qt)
	      || ! NILP (integers))
	    {
	      XSETFASTINT (data[2 * i], start);
	      XSETFASTINT (data[2 * i + 1], search_regs.end[i]);
	    }
	  else if (BUFFERP (last_thing_searched))
	    {
	      data[2 * i] = Fmake_marker ();
	      Fset_marker (data[2 * i],
			   make_number (start),
			   last_thing_searched);
	      data[2 * i + 1] = Fmake_marker ();
	      Fset_marker (data[2 * i + 1],
			   make_number (search_regs.end[i]),
			   last_thing_searched);
	    }
	  else
	    /* last_thing_searched must always be Qt, a buffer, or Qnil.  */
	    abort ();

	  len = 2 * i + 2;
	}
      else
	data[2 * i] = data[2 * i + 1] = Qnil;
    }

  if (BUFFERP (last_thing_searched) && !NILP (integers))
    {
      data[len] = last_thing_searched;
      len++;
    }

  /* If REUSE is not usable, cons up the values and return them.  */
  if (! CONSP (reuse))
    return Flist (len, data);

  /* If REUSE is a list, store as many value elements as will fit
     into the elements of REUSE.  */
  for (i = 0, tail = reuse; CONSP (tail);
       i++, tail = XCDR (tail))
    {
      if (i < len)
	XSETCAR (tail, data[i]);
      else
	XSETCAR (tail, Qnil);
      prev = tail;
    }

  /* If we couldn't fit all value elements into REUSE,
     cons up the rest of them and add them to the end of REUSE.  */
  if (i < len)
    XSETCDR (prev, Flist (len - i, data + i));

  return reuse;
}

/* Internal usage only:
   If RESEAT is `evaporate', put the markers back on the free list
   immediately.  No other references to the markers must exist in this case,
   so it is used only internally on the unwind stack and save-match-data from
   Lisp.  */

DEFUN ("set-match-data", Fset_match_data, Sset_match_data, 1, 2, 0,
       doc: /* Set internal data on last search match from elements of LIST.
LIST should have been created by calling `match-data' previously.

If optional arg RESEAT is non-nil, make markers on LIST point nowhere.  */)
    (list, reseat)
     register Lisp_Object list, reseat;
{
  register int i;
  register Lisp_Object marker;

  if (running_asynch_code)
    save_search_regs ();

  CHECK_LIST (list);

  /* Unless we find a marker with a buffer or an explicit buffer
     in LIST, assume that this match data came from a string.  */
  last_thing_searched = Qt;

  /* Allocate registers if they don't already exist.  */
  {
    int length = XFASTINT (Flength (list)) / 2;

    if (length > search_regs.num_regs)
      {
	if (search_regs.num_regs == 0)
	  {
	    search_regs.start
	      = (regoff_t *) xmalloc (length * sizeof (regoff_t));
	    search_regs.end
	      = (regoff_t *) xmalloc (length * sizeof (regoff_t));
	  }
	else
	  {
	    search_regs.start
	      = (regoff_t *) xrealloc (search_regs.start,
				       length * sizeof (regoff_t));
	    search_regs.end
	      = (regoff_t *) xrealloc (search_regs.end,
				       length * sizeof (regoff_t));
	  }

	for (i = search_regs.num_regs; i < length; i++)
	  search_regs.start[i] = -1;

	search_regs.num_regs = length;
      }

    for (i = 0; CONSP (list); i++)
      {
	marker = XCAR (list);
	if (BUFFERP (marker))
	  {
	    last_thing_searched = marker;
	    break;
	  }
	if (i >= length)
	  break;
	if (NILP (marker))
	  {
	    search_regs.start[i] = -1;
	    list = XCDR (list);
	  }
	else
	  {
	    int from;
	    Lisp_Object m;

	    m = marker;
	    if (MARKERP (marker))
	      {
		if (XMARKER (marker)->buffer == 0)
		  XSETFASTINT (marker, 0);
		else
		  XSETBUFFER (last_thing_searched, XMARKER (marker)->buffer);
	      }

	    CHECK_NUMBER_COERCE_MARKER (marker);
	    from = XINT (marker);

	    if (!NILP (reseat) && MARKERP (m))
	      {
		if (EQ (reseat, Qevaporate))
		  free_marker (m);
		else
		  unchain_marker (XMARKER (m));
		XSETCAR (list, Qnil);
	      }

	    if ((list = XCDR (list), !CONSP (list)))
	      break;

	    m = marker = XCAR (list);

	    if (MARKERP (marker) && XMARKER (marker)->buffer == 0)
	      XSETFASTINT (marker, 0);

	    CHECK_NUMBER_COERCE_MARKER (marker);
	    search_regs.start[i] = from;
	    search_regs.end[i] = XINT (marker);

	    if (!NILP (reseat) && MARKERP (m))
	      {
		if (EQ (reseat, Qevaporate))
		  free_marker (m);
		else
		  unchain_marker (XMARKER (m));
		XSETCAR (list, Qnil);
	      }
	  }
	list = XCDR (list);
      }

    for (; i < search_regs.num_regs; i++)
      search_regs.start[i] = -1;
  }

  return Qnil;
}

/* If non-zero the match data have been saved in saved_search_regs
   during the execution of a sentinel or filter. */
static int search_regs_saved;
static struct re_registers saved_search_regs;
static Lisp_Object saved_last_thing_searched;

/* Called from Flooking_at, Fstring_match, search_buffer, Fstore_match_data
   if asynchronous code (filter or sentinel) is running. */
static void
save_search_regs ()
{
  if (!search_regs_saved)
    {
      saved_search_regs.num_regs = search_regs.num_regs;
      saved_search_regs.start = search_regs.start;
      saved_search_regs.end = search_regs.end;
      saved_last_thing_searched = last_thing_searched;
      last_thing_searched = Qnil;
      search_regs.num_regs = 0;
      search_regs.start = 0;
      search_regs.end = 0;

      search_regs_saved = 1;
    }
}

/* Called upon exit from filters and sentinels. */
void
restore_search_regs ()
{
  if (search_regs_saved)
    {
      if (search_regs.num_regs > 0)
	{
	  xfree (search_regs.start);
	  xfree (search_regs.end);
	}
      search_regs.num_regs = saved_search_regs.num_regs;
      search_regs.start = saved_search_regs.start;
      search_regs.end = saved_search_regs.end;
      last_thing_searched = saved_last_thing_searched;
      saved_last_thing_searched = Qnil;
      search_regs_saved = 0;
    }
}

static Lisp_Object
unwind_set_match_data (list)
     Lisp_Object list;
{
  /* It is safe to free (evaporate) the markers immediately.  */
  return Fset_match_data (list, Qevaporate);
}

/* Called to unwind protect the match data.  */
void
record_unwind_save_match_data ()
{
  record_unwind_protect (unwind_set_match_data,
			 Fmatch_data (Qnil, Qnil, Qnil));
}

/* Quote a string to inactivate reg-expr chars */

DEFUN ("regexp-quote", Fregexp_quote, Sregexp_quote, 1, 1, 0,
       doc: /* Return a regexp string which matches exactly STRING and nothing else.  */)
     (string)
     Lisp_Object string;
{
  register unsigned char *in, *out, *end;
  register unsigned char *temp;
  int backslashes_added = 0;

  CHECK_STRING (string);

  temp = (unsigned char *) alloca (SBYTES (string) * 2);

  /* Now copy the data into the new string, inserting escapes. */

  in = SDATA (string);
  end = in + SBYTES (string);
  out = temp;

  for (; in != end; in++)
    {
      if (*in == '['
	  || *in == '*' || *in == '.' || *in == '\\'
	  || *in == '?' || *in == '+'
	  || *in == '^' || *in == '$')
	*out++ = '\\', backslashes_added++;
      *out++ = *in;
    }

  return make_specified_string (temp,
				SCHARS (string) + backslashes_added,
				out - temp,
				STRING_MULTIBYTE (string));
}

void
syms_of_search ()
{
  register int i;

  for (i = 0; i < REGEXP_CACHE_SIZE; ++i)
    {
      searchbufs[i].buf.allocated = 100;
      searchbufs[i].buf.buffer = (unsigned char *) xmalloc (100);
      searchbufs[i].buf.fastmap = searchbufs[i].fastmap;
      searchbufs[i].regexp = Qnil;
      searchbufs[i].whitespace_regexp = Qnil;
      searchbufs[i].syntax_table = Qnil;
      staticpro (&searchbufs[i].regexp);
      staticpro (&searchbufs[i].whitespace_regexp);
      staticpro (&searchbufs[i].syntax_table);
      searchbufs[i].next = (i == REGEXP_CACHE_SIZE-1 ? 0 : &searchbufs[i+1]);
    }
  searchbuf_head = &searchbufs[0];

  Qsearch_failed = intern ("search-failed");
  staticpro (&Qsearch_failed);
  Qinvalid_regexp = intern ("invalid-regexp");
  staticpro (&Qinvalid_regexp);

  Fput (Qsearch_failed, Qerror_conditions,
	Fcons (Qsearch_failed, Fcons (Qerror, Qnil)));
  Fput (Qsearch_failed, Qerror_message,
	build_string ("Search failed"));

  Fput (Qinvalid_regexp, Qerror_conditions,
	Fcons (Qinvalid_regexp, Fcons (Qerror, Qnil)));
  Fput (Qinvalid_regexp, Qerror_message,
	build_string ("Invalid regexp"));

  last_thing_searched = Qnil;
  staticpro (&last_thing_searched);

  saved_last_thing_searched = Qnil;
  staticpro (&saved_last_thing_searched);

  DEFVAR_LISP ("search-spaces-regexp", &Vsearch_spaces_regexp,
      doc: /* Regexp to substitute for bunches of spaces in regexp search.
Some commands use this for user-specified regexps.
Spaces that occur inside character classes or repetition operators
or other such regexp constructs are not replaced with this.
A value of nil (which is the normal value) means treat spaces literally.  */);
  Vsearch_spaces_regexp = Qnil;

  defsubr (&Slooking_at);
  defsubr (&Sposix_looking_at);
  defsubr (&Sstring_match);
  defsubr (&Sposix_string_match);
  defsubr (&Ssearch_forward);
  defsubr (&Ssearch_backward);
  defsubr (&Sword_search_forward);
  defsubr (&Sword_search_backward);
  defsubr (&Sre_search_forward);
  defsubr (&Sre_search_backward);
  defsubr (&Sposix_search_forward);
  defsubr (&Sposix_search_backward);
  defsubr (&Sreplace_match);
  defsubr (&Smatch_beginning);
  defsubr (&Smatch_end);
  defsubr (&Smatch_data);
  defsubr (&Sset_match_data);
  defsubr (&Sregexp_quote);
}

/* arch-tag: a6059d79-0552-4f14-a2cb-d379a4e3c78f
   (do not change this comment) */