1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477
|
/*************************************************
* Perl-Compatible Regular Expressions *
*************************************************/
/*
This is a library of functions to support regular expressions whose syntax
and semantics are as close as possible to those of the Perl 5 language. See
the file Tech.Notes for some information on the internals.
Written by: Philip Hazel <ph10@cam.ac.uk>
Copyright (c) 1997-1999 University of Cambridge
-----------------------------------------------------------------------------
Permission is granted to anyone to use this software for any purpose on any
computer system, and to redistribute it freely, subject to the following
restrictions:
1. This software 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.
2. The origin of this software must not be misrepresented, either by
explicit claim or by omission.
3. Altered versions must be plainly marked as such, and must not be
misrepresented as being the original software.
4. If PCRE is embedded in any software that is released under the GNU
General Purpose Licence (GPL), then the terms of that licence shall
supersede any condition above with which it is incompatible.
-----------------------------------------------------------------------------
*/
/* Define DEBUG to get debugging output on stdout. */
/* #define DEBUG */
/* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
inline, and there are *still* stupid compilers about that don't like indented
pre-processor statements. I suppose it's only been 10 years... */
#ifdef DEBUG
#define DPRINTF(p) printf p
#else
#define DPRINTF(p) /*nothing*/
#endif
/* Include the internals header, which itself includes Standard C headers plus
the external pcre header. */
#include "internal.h"
/* Allow compilation as C++ source code, should anybody want to do that. */
#ifdef __cplusplus
#define class pcre_class
#endif
/* Number of items on the nested bracket stacks at compile time. This should
not be set greater than 200. */
#define BRASTACK_SIZE 200
/* Min and max values for the common repeats; for the maxima, 0 => infinity */
static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };
static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };
/* Text forms of OP_ values and things, for debugging (not all used) */
#ifdef DEBUG
static const char *OP_names[] = {
"End", "\\A", "\\B", "\\b", "\\D", "\\d",
"\\S", "\\s", "\\W", "\\w", "\\Z", "\\z",
"Opt", "^", "$", "Any", "chars", "not",
"*", "*?", "+", "+?", "?", "??", "{", "{", "{",
"*", "*?", "+", "+?", "?", "??", "{", "{", "{",
"*", "*?", "+", "+?", "?", "??", "{", "{", "{",
"*", "*?", "+", "+?", "?", "??", "{", "{",
"class", "Ref",
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",
"AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref",
"Brazero", "Braminzero", "Bra"
};
#endif
/* Table for handling escaped characters in the range '0'-'z'. Positive returns
are simple data values; negative values are for special things like \d and so
on. Zero means further processing is needed (for things like \x), or the escape
is invalid. */
static const short int escapes[] = {
0, 0, 0, 0, 0, 0, 0, 0, /* 0 - 7 */
0, 0, ':', ';', '<', '=', '>', '?', /* 8 - ? */
'@', -ESC_A, -ESC_B, 0, -ESC_D, 0, 0, 0, /* @ - G */
0, 0, 0, 0, 0, 0, 0, 0, /* H - O */
0, 0, 0, -ESC_S, 0, 0, 0, -ESC_W, /* P - W */
0, 0, -ESC_Z, '[', '\\', ']', '^', '_', /* X - _ */
'`', 7, -ESC_b, 0, -ESC_d, 27, '\f', 0, /* ` - g */
0, 0, 0, 0, 0, 0, '\n', 0, /* h - o */
0, 0, '\r', -ESC_s, '\t', 0, 0, -ESC_w, /* p - w */
0, 0, -ESC_z /* x - z */
};
/* Definition to allow mutual recursion */
static BOOL
compile_regex(int, int, int *, uschar **, const uschar **, const char **,
BOOL, int, int *, int *, compile_data *);
/*************************************************
* Global variables *
*************************************************/
/* PCRE is thread-clean and doesn't use any global variables in the normal
sense. However, it calls memory allocation and free functions via the two
indirections below, which are can be changed by the caller, but are shared
between all threads. */
void *(*pcre_malloc)(size_t) = malloc;
void (*pcre_free)(void *) = free;
/*************************************************
* Default character tables *
*************************************************/
/* A default set of character tables is included in the PCRE binary. Its source
is built by the maketables auxiliary program, which uses the default C ctypes
functions, and put in the file chartables.c. These tables are used by PCRE
whenever the caller of pcre_compile() does not provide an alternate set of
tables. */
#include "chartables.c"
/*************************************************
* Return version string *
*************************************************/
#define STRING(a) # a
#define XSTRING(s) STRING(s)
const char *
pcre_version(void)
{
return XSTRING(PCRE_MAJOR) "." XSTRING(PCRE_MINOR) " " XSTRING(PCRE_DATE);
}
/*************************************************
* Return info about a compiled pattern *
*************************************************/
/* This function picks potentially useful data out of the private
structure. The public options are passed back in an int - though the
re->options field has been expanded to a long int, all the public options
at the low end of it, and so even on 16-bit systems this will still be OK.
Therefore, I haven't changed the API for pcre_info().
Arguments:
external_re points to compiled code
optptr where to pass back the options
first_char where to pass back the first character,
or -1 if multiline and all branches start ^,
or -2 otherwise
Returns: number of identifying extraction brackets
or negative values on error
*/
int
pcre_info(const pcre *external_re, int *optptr, int *first_char)
{
const real_pcre *re = (const real_pcre *)external_re;
if (re == NULL) return PCRE_ERROR_NULL;
if (re->magic_number != MAGIC_NUMBER) return PCRE_ERROR_BADMAGIC;
if (optptr != NULL) *optptr = (int)(re->options & PUBLIC_OPTIONS);
if (first_char != NULL)
*first_char = ((re->options & PCRE_FIRSTSET) != 0)? re->first_char :
((re->options & PCRE_STARTLINE) != 0)? -1 : -2;
return re->top_bracket;
}
#ifdef DEBUG
/*************************************************
* Debugging function to print chars *
*************************************************/
/* Print a sequence of chars in printable format, stopping at the end of the
subject if the requested.
Arguments:
p points to characters
length number to print
is_subject TRUE if printing from within md->start_subject
md pointer to matching data block, if is_subject is TRUE
Returns: nothing
*/
static void
pchars(const uschar *p, int length, BOOL is_subject, match_data *md)
{
int c;
if (is_subject && length > md->end_subject - p) length = md->end_subject - p;
while (length-- > 0)
if (isprint(c = *(p++))) printf("%c", c); else printf("\\x%02x", c);
}
#endif
/*************************************************
* Handle escapes *
*************************************************/
/* This function is called when a \ has been encountered. It either returns a
positive value for a simple escape such as \n, or a negative value which
encodes one of the more complicated things such as \d. On entry, ptr is
pointing at the \. On exit, it is on the final character of the escape
sequence.
Arguments:
ptrptr points to the pattern position pointer
errorptr points to the pointer to the error message
bracount number of previous extracting brackets
options the options bits
isclass TRUE if inside a character class
cd pointer to char tables block
Returns: zero or positive => a data character
negative => a special escape sequence
on error, errorptr is set
*/
static int
check_escape(const uschar **ptrptr, const char **errorptr, int bracount,
int options, BOOL isclass, compile_data *cd)
{
const uschar *ptr = *ptrptr;
int c = *(++ptr) & 255; /* Ensure > 0 on signed-char systems */
int i;
if (c == 0) *errorptr = ERR1;
/* Digits or letters may have special meaning; all others are literals. */
else if (c < '0' || c > 'z') {}
/* Do an initial lookup in a table. A non-zero result is something that can be
returned immediately. Otherwise further processing may be required. */
else if ((i = escapes[c - '0']) != 0) c = i;
/* Escapes that need further processing, or are illegal. */
else
{
const uschar *oldptr;
switch (c)
{
/* The handling of escape sequences consisting of a string of digits
starting with one that is not zero is not straightforward. By experiment,
the way Perl works seems to be as follows:
Outside a character class, the digits are read as a decimal number. If the
number is less than 10, or if there are that many previous extracting
left brackets, then it is a back reference. Otherwise, up to three octal
digits are read to form an escaped byte. Thus \123 is likely to be octal
123 (cf \0123, which is octal 012 followed by the literal 3). If the octal
value is greater than 377, the least significant 8 bits are taken. Inside a
character class, \ followed by a digit is always an octal number. */
case '1': case '2': case '3': case '4': case '5':
case '6': case '7': case '8': case '9':
if (!isclass)
{
oldptr = ptr;
c -= '0';
while ((cd->ctypes[ptr[1]] & ctype_digit) != 0)
c = c * 10 + *(++ptr) - '0';
if (c < 10 || c <= bracount)
{
c = -(ESC_REF + c);
break;
}
ptr = oldptr; /* Put the pointer back and fall through */
}
/* Handle an octal number following \. If the first digit is 8 or 9, Perl
generates a binary zero byte and treats the digit as a following literal.
Thus we have to pull back the pointer by one. */
if ((c = *ptr) >= '8')
{
ptr--;
c = 0;
break;
}
/* \0 always starts an octal number, but we may drop through to here with a
larger first octal digit */
case '0':
c -= '0';
while(i++ < 2 && (cd->ctypes[ptr[1]] & ctype_digit) != 0 &&
ptr[1] != '8' && ptr[1] != '9')
c = c * 8 + *(++ptr) - '0';
break;
/* Special escapes not starting with a digit are straightforward */
case 'x':
c = 0;
while (i++ < 2 && (cd->ctypes[ptr[1]] & ctype_xdigit) != 0)
{
ptr++;
c = c * 16 + cd->lcc[*ptr] -
(((cd->ctypes[*ptr] & ctype_digit) != 0)? '0' : 'W');
}
break;
case 'c':
c = *(++ptr);
if (c == 0)
{
*errorptr = ERR2;
return 0;
}
/* A letter is upper-cased; then the 0x40 bit is flipped */
if (c >= 'a' && c <= 'z') c = cd->fcc[c];
c ^= 0x40;
break;
/* PCRE_EXTRA enables extensions to Perl in the matter of escapes. Any
other alphameric following \ is an error if PCRE_EXTRA was set; otherwise,
for Perl compatibility, it is a literal. This code looks a bit odd, but
there used to be some cases other than the default, and there may be again
in future, so I haven't "optimized" it. */
default:
if ((options & PCRE_EXTRA) != 0) switch(c)
{
default:
*errorptr = ERR3;
break;
}
break;
}
}
*ptrptr = ptr;
return c;
}
/*************************************************
* Check for counted repeat *
*************************************************/
/* This function is called when a '{' is encountered in a place where it might
start a quantifier. It looks ahead to see if it really is a quantifier or not.
It is only a quantifier if it is one of the forms {ddd} {ddd,} or {ddd,ddd}
where the ddds are digits.
Arguments:
p pointer to the first char after '{'
cd pointer to char tables block
Returns: TRUE or FALSE
*/
static BOOL
is_counted_repeat(const uschar *p, compile_data *cd)
{
if ((cd->ctypes[*p++] & ctype_digit) == 0) return FALSE;
while ((cd->ctypes[*p] & ctype_digit) != 0) p++;
if (*p == '}') return TRUE;
if (*p++ != ',') return FALSE;
if (*p == '}') return TRUE;
if ((cd->ctypes[*p++] & ctype_digit) == 0) return FALSE;
while ((cd->ctypes[*p] & ctype_digit) != 0) p++;
return (*p == '}');
}
/*************************************************
* Read repeat counts *
*************************************************/
/* Read an item of the form {n,m} and return the values. This is called only
after is_counted_repeat() has confirmed that a repeat-count quantifier exists,
so the syntax is guaranteed to be correct, but we need to check the values.
Arguments:
p pointer to first char after '{'
minp pointer to int for min
maxp pointer to int for max
returned as -1 if no max
errorptr points to pointer to error message
cd pointer to character tables clock
Returns: pointer to '}' on success;
current ptr on error, with errorptr set
*/
static const uschar *
read_repeat_counts(const uschar *p, int *minp, int *maxp,
const char **errorptr, compile_data *cd)
{
int min = 0;
int max = -1;
while ((cd->ctypes[*p] & ctype_digit) != 0) min = min * 10 + *p++ - '0';
if (*p == '}') max = min; else
{
if (*(++p) != '}')
{
max = 0;
while((cd->ctypes[*p] & ctype_digit) != 0) max = max * 10 + *p++ - '0';
if (max < min)
{
*errorptr = ERR4;
return p;
}
}
}
/* Do paranoid checks, then fill in the required variables, and pass back the
pointer to the terminating '}'. */
if (min > 65535 || max > 65535)
*errorptr = ERR5;
else
{
*minp = min;
*maxp = max;
}
return p;
}
/*************************************************
* Find the fixed length of a pattern *
*************************************************/
/* Scan a pattern and compute the fixed length of subject that will match it,
if the length is fixed. This is needed for dealing with backward assertions.
Arguments:
code points to the start of the pattern (the bracket)
Returns: the fixed length, or -1 if there is no fixed length
*/
static int
find_fixedlength(uschar *code)
{
int length = -1;
register int branchlength = 0;
register uschar *cc = code + 3;
/* Scan along the opcodes for this branch. If we get to the end of the
branch, check the length against that of the other branches. */
for (;;)
{
int d;
register int op = *cc;
if (op >= OP_BRA) op = OP_BRA;
switch (op)
{
case OP_BRA:
case OP_ONCE:
case OP_COND:
d = find_fixedlength(cc);
if (d < 0) return -1;
branchlength += d;
do cc += (cc[1] << 8) + cc[2]; while (*cc == OP_ALT);
cc += 3;
break;
/* Reached end of a branch; if it's a ket it is the end of a nested
call. If it's ALT it is an alternation in a nested call. If it is
END it's the end of the outer call. All can be handled by the same code. */
case OP_ALT:
case OP_KET:
case OP_KETRMAX:
case OP_KETRMIN:
case OP_END:
if (length < 0) length = branchlength;
else if (length != branchlength) return -1;
if (*cc != OP_ALT) return length;
cc += 3;
branchlength = 0;
break;
/* Skip over assertive subpatterns */
case OP_ASSERT:
case OP_ASSERT_NOT:
case OP_ASSERTBACK:
case OP_ASSERTBACK_NOT:
do cc += (cc[1] << 8) + cc[2]; while (*cc == OP_ALT);
cc += 3;
break;
/* Skip over things that don't match chars */
case OP_REVERSE:
cc++;
/* Fall through */
case OP_CREF:
case OP_OPT:
cc++;
/* Fall through */
case OP_SOD:
case OP_EOD:
case OP_EODN:
case OP_CIRC:
case OP_DOLL:
case OP_NOT_WORD_BOUNDARY:
case OP_WORD_BOUNDARY:
cc++;
break;
/* Handle char strings */
case OP_CHARS:
branchlength += *(++cc);
cc += *cc + 1;
break;
/* Handle exact repetitions */
case OP_EXACT:
case OP_TYPEEXACT:
branchlength += (cc[1] << 8) + cc[2];
cc += 4;
break;
/* Handle single-char matchers */
case OP_NOT_DIGIT:
case OP_DIGIT:
case OP_NOT_WHITESPACE:
case OP_WHITESPACE:
case OP_NOT_WORDCHAR:
case OP_WORDCHAR:
case OP_ANY:
branchlength++;
cc++;
break;
/* Check a class for variable quantification */
case OP_CLASS:
cc += (*cc == OP_REF)? 2 : 33;
switch (*cc)
{
case OP_CRSTAR:
case OP_CRMINSTAR:
case OP_CRQUERY:
case OP_CRMINQUERY:
return -1;
case OP_CRRANGE:
case OP_CRMINRANGE:
if ((cc[1] << 8) + cc[2] != (cc[3] << 8) + cc[4]) return -1;
branchlength += (cc[1] << 8) + cc[2];
cc += 5;
break;
default:
branchlength++;
}
break;
/* Anything else is variable length */
default:
return -1;
}
}
/* Control never gets here */
}
/*************************************************
* Compile one branch *
*************************************************/
/* Scan the pattern, compiling it into the code vector.
Arguments:
options the option bits
brackets points to number of brackets used
code points to the pointer to the current code point
ptrptr points to the current pattern pointer
errorptr points to pointer to error message
optchanged set to the value of the last OP_OPT item compiled
reqchar set to the last literal character required, else -1
countlits set to count of mandatory literal characters
cd contains pointers to tables
Returns: TRUE on success
FALSE, with *errorptr set on error
*/
static BOOL
compile_branch(int options, int *brackets, uschar **codeptr,
const uschar **ptrptr, const char **errorptr, int *optchanged,
int *reqchar, int *countlits, compile_data *cd)
{
int repeat_type, op_type;
int repeat_min, repeat_max;
int bravalue, length;
int greedy_default, greedy_non_default;
int prevreqchar;
int condcount = 0;
int subcountlits = 0;
register int c;
register uschar *code = *codeptr;
uschar *tempcode;
const uschar *ptr = *ptrptr;
const uschar *tempptr;
uschar *previous = NULL;
uschar class[32];
/* Set up the default and non-default settings for greediness */
greedy_default = ((options & PCRE_UNGREEDY) != 0);
greedy_non_default = greedy_default ^ 1;
/* Initialize no required char, and count of literals */
*reqchar = prevreqchar = -1;
*countlits = 0;
/* Switch on next character until the end of the branch */
for (;; ptr++)
{
BOOL negate_class;
int class_charcount;
int class_lastchar;
int newoptions;
int condref;
int subreqchar;
c = *ptr;
if ((options & PCRE_EXTENDED) != 0)
{
if ((cd->ctypes[c] & ctype_space) != 0) continue;
if (c == '#')
{
while ((c = *(++ptr)) != 0 && c != '\n');
continue;
}
}
switch(c)
{
/* The branch terminates at end of string, |, or ). */
case 0:
case '|':
case ')':
*codeptr = code;
*ptrptr = ptr;
return TRUE;
/* Handle single-character metacharacters */
case '^':
previous = NULL;
*code++ = OP_CIRC;
break;
case '$':
previous = NULL;
*code++ = OP_DOLL;
break;
case '.':
previous = code;
*code++ = OP_ANY;
break;
/* Character classes. These always build a 32-byte bitmap of the permitted
characters, except in the special case where there is only one character.
For negated classes, we build the map as usual, then invert it at the end.
*/
case '[':
previous = code;
*code++ = OP_CLASS;
/* If the first character is '^', set the negation flag and skip it. */
if ((c = *(++ptr)) == '^')
{
negate_class = TRUE;
c = *(++ptr);
}
else negate_class = FALSE;
/* Keep a count of chars so that we can optimize the case of just a single
character. */
class_charcount = 0;
class_lastchar = -1;
/* Initialize the 32-char bit map to all zeros. We have to build the
map in a temporary bit of store, in case the class contains only 1
character, because in that case the compiled code doesn't use the
bit map. */
memset(class, 0, 32 * sizeof(uschar));
/* Process characters until ] is reached. By writing this as a "do" it
means that an initial ] is taken as a data character. */
do
{
if (c == 0)
{
*errorptr = ERR6;
goto FAILED;
}
/* Backslash may introduce a single character, or it may introduce one
of the specials, which just set a flag. Escaped items are checked for
validity in the pre-compiling pass. The sequence \b is a special case.
Inside a class (and only there) it is treated as backspace. Elsewhere
it marks a word boundary. Other escapes have preset maps ready to
or into the one we are building. We assume they have more than one
character in them, so set class_count bigger than one. */
if (c == '\\')
{
c = check_escape(&ptr, errorptr, *brackets, options, TRUE, cd);
if (-c == ESC_b) c = '\b';
else if (c < 0)
{
register const uschar *cbits = cd->cbits;
class_charcount = 10;
switch (-c)
{
case ESC_d:
for (c = 0; c < 32; c++) class[c] |= cbits[c+cbit_digit];
continue;
case ESC_D:
for (c = 0; c < 32; c++) class[c] |= ~cbits[c+cbit_digit];
continue;
case ESC_w:
for (c = 0; c < 32; c++)
class[c] |= (cbits[c+cbit_digit] | cbits[c+cbit_word]);
continue;
case ESC_W:
for (c = 0; c < 32; c++)
class[c] |= ~(cbits[c+cbit_digit] | cbits[c+cbit_word]);
continue;
case ESC_s:
for (c = 0; c < 32; c++) class[c] |= cbits[c+cbit_space];
continue;
case ESC_S:
for (c = 0; c < 32; c++) class[c] |= ~cbits[c+cbit_space];
continue;
default:
*errorptr = ERR7;
goto FAILED;
}
}
/* Fall through if single character */
}
/* A single character may be followed by '-' to form a range. However,
Perl does not permit ']' to be the end of the range. A '-' character
here is treated as a literal. */
if (ptr[1] == '-' && ptr[2] != ']')
{
int d;
ptr += 2;
d = *ptr;
if (d == 0)
{
*errorptr = ERR6;
goto FAILED;
}
/* The second part of a range can be a single-character escape, but
not any of the other escapes. */
if (d == '\\')
{
d = check_escape(&ptr, errorptr, *brackets, options, TRUE, cd);
if (d < 0)
{
if (d == -ESC_b) d = '\b'; else
{
*errorptr = ERR7;
goto FAILED;
}
}
}
if (d < c)
{
*errorptr = ERR8;
goto FAILED;
}
for (; c <= d; c++)
{
class[c/8] |= (1 << (c&7));
if ((options & PCRE_CASELESS) != 0)
{
int uc = cd->fcc[c]; /* flip case */
class[uc/8] |= (1 << (uc&7));
}
class_charcount++; /* in case a one-char range */
class_lastchar = c;
}
continue; /* Go get the next char in the class */
}
/* Handle a lone single character - we can get here for a normal
non-escape char, or after \ that introduces a single character. */
class [c/8] |= (1 << (c&7));
if ((options & PCRE_CASELESS) != 0)
{
c = cd->fcc[c]; /* flip case */
class[c/8] |= (1 << (c&7));
}
class_charcount++;
class_lastchar = c;
}
/* Loop until ']' reached; the check for end of string happens inside the
loop. This "while" is the end of the "do" above. */
while ((c = *(++ptr)) != ']');
/* If class_charcount is 1 and class_lastchar is not negative, we saw
precisely one character. This doesn't need the whole 32-byte bit map.
We turn it into a 1-character OP_CHAR if it's positive, or OP_NOT if
it's negative. */
if (class_charcount == 1 && class_lastchar >= 0)
{
if (negate_class)
{
code[-1] = OP_NOT;
}
else
{
code[-1] = OP_CHARS;
*code++ = 1;
}
*code++ = class_lastchar;
}
/* Otherwise, negate the 32-byte map if necessary, and copy it into
the code vector. */
else
{
if (negate_class)
for (c = 0; c < 32; c++) code[c] = ~class[c];
else
memcpy(code, class, 32);
code += 32;
}
break;
/* Various kinds of repeat */
case '{':
if (!is_counted_repeat(ptr+1, cd)) goto NORMAL_CHAR;
ptr = read_repeat_counts(ptr+1, &repeat_min, &repeat_max, errorptr, cd);
if (*errorptr != NULL) goto FAILED;
goto REPEAT;
case '*':
repeat_min = 0;
repeat_max = -1;
goto REPEAT;
case '+':
repeat_min = 1;
repeat_max = -1;
goto REPEAT;
case '?':
repeat_min = 0;
repeat_max = 1;
REPEAT:
if (previous == NULL)
{
*errorptr = ERR9;
goto FAILED;
}
/* If the next character is '?' this is a minimizing repeat, by default,
but if PCRE_UNGREEDY is set, it works the other way round. Advance to the
next character. */
if (ptr[1] == '?')
{ repeat_type = greedy_non_default; ptr++; }
else repeat_type = greedy_default;
/* If previous was a string of characters, chop off the last one and use it
as the subject of the repeat. If there was only one character, we can
abolish the previous item altogether. A repeat with a zero minimum wipes
out any reqchar setting, backing up to the previous value. We must also
adjust the countlits value. */
if (*previous == OP_CHARS)
{
int len = previous[1];
if (repeat_min == 0) *reqchar = prevreqchar;
*countlits += repeat_min - 1;
if (len == 1)
{
c = previous[2];
code = previous;
}
else
{
c = previous[len+1];
previous[1]--;
code--;
}
op_type = 0; /* Use single-char op codes */
goto OUTPUT_SINGLE_REPEAT; /* Code shared with single character types */
}
/* If previous was a single negated character ([^a] or similar), we use
one of the special opcodes, replacing it. The code is shared with single-
character repeats by adding a suitable offset into repeat_type. */
else if ((int)*previous == OP_NOT)
{
op_type = OP_NOTSTAR - OP_STAR; /* Use "not" opcodes */
c = previous[1];
code = previous;
goto OUTPUT_SINGLE_REPEAT;
}
/* If previous was a character type match (\d or similar), abolish it and
create a suitable repeat item. The code is shared with single-character
repeats by adding a suitable offset into repeat_type. */
else if ((int)*previous < OP_EODN || *previous == OP_ANY)
{
op_type = OP_TYPESTAR - OP_STAR; /* Use type opcodes */
c = *previous;
code = previous;
OUTPUT_SINGLE_REPEAT:
/* If the maximum is zero then the minimum must also be zero; Perl allows
this case, so we do too - by simply omitting the item altogether. */
if (repeat_max == 0) goto END_REPEAT;
/* Combine the op_type with the repeat_type */
repeat_type += op_type;
/* A minimum of zero is handled either as the special case * or ?, or as
an UPTO, with the maximum given. */
if (repeat_min == 0)
{
if (repeat_max == -1) *code++ = OP_STAR + repeat_type;
else if (repeat_max == 1) *code++ = OP_QUERY + repeat_type;
else
{
*code++ = OP_UPTO + repeat_type;
*code++ = repeat_max >> 8;
*code++ = (repeat_max & 255);
}
}
/* The case {1,} is handled as the special case + */
else if (repeat_min == 1 && repeat_max == -1)
*code++ = OP_PLUS + repeat_type;
/* The case {n,n} is just an EXACT, while the general case {n,m} is
handled as an EXACT followed by an UPTO. An EXACT of 1 is optimized. */
else
{
if (repeat_min != 1)
{
*code++ = OP_EXACT + op_type; /* NB EXACT doesn't have repeat_type */
*code++ = repeat_min >> 8;
*code++ = (repeat_min & 255);
}
/* If the mininum is 1 and the previous item was a character string,
we either have to put back the item that got cancelled if the string
length was 1, or add the character back onto the end of a longer
string. For a character type nothing need be done; it will just get
put back naturally. Note that the final character is always going to
get added below. */
else if (*previous == OP_CHARS)
{
if (code == previous) code += 2; else previous[1]++;
}
/* For a single negated character we also have to put back the
item that got cancelled. */
else if (*previous == OP_NOT) code++;
/* If the maximum is unlimited, insert an OP_STAR. */
if (repeat_max < 0)
{
*code++ = c;
*code++ = OP_STAR + repeat_type;
}
/* Else insert an UPTO if the max is greater than the min. */
else if (repeat_max != repeat_min)
{
*code++ = c;
repeat_max -= repeat_min;
*code++ = OP_UPTO + repeat_type;
*code++ = repeat_max >> 8;
*code++ = (repeat_max & 255);
}
}
/* The character or character type itself comes last in all cases. */
*code++ = c;
}
/* If previous was a character class or a back reference, we put the repeat
stuff after it, but just skip the item if the repeat was {0,0}. */
else if (*previous == OP_CLASS || *previous == OP_REF)
{
if (repeat_max == 0)
{
code = previous;
goto END_REPEAT;
}
if (repeat_min == 0 && repeat_max == -1)
*code++ = OP_CRSTAR + repeat_type;
else if (repeat_min == 1 && repeat_max == -1)
*code++ = OP_CRPLUS + repeat_type;
else if (repeat_min == 0 && repeat_max == 1)
*code++ = OP_CRQUERY + repeat_type;
else
{
*code++ = OP_CRRANGE + repeat_type;
*code++ = repeat_min >> 8;
*code++ = repeat_min & 255;
if (repeat_max == -1) repeat_max = 0; /* 2-byte encoding for max */
*code++ = repeat_max >> 8;
*code++ = repeat_max & 255;
}
}
/* If previous was a bracket group, we may have to replicate it in certain
cases. */
else if ((int)*previous >= OP_BRA || (int)*previous == OP_ONCE ||
(int)*previous == OP_COND)
{
register int i;
int ketoffset = 0;
int len = code - previous;
uschar *bralink = NULL;
/* If the maximum repeat count is unlimited, find the end of the bracket
by scanning through from the start, and compute the offset back to it
from the current code pointer. There may be an OP_OPT setting following
the final KET, so we can't find the end just by going back from the code
pointer. */
if (repeat_max == -1)
{
register uschar *ket = previous;
do ket += (ket[1] << 8) + ket[2]; while (*ket != OP_KET);
ketoffset = code - ket;
}
/* The case of a zero minimum is special because of the need to stick
OP_BRAZERO in front of it, and because the group appears once in the
data, whereas in other cases it appears the minimum number of times. For
this reason, it is simplest to treat this case separately, as otherwise
the code gets far too mess. There are several special subcases when the
minimum is zero. */
if (repeat_min == 0)
{
/* If we set up a required char from the bracket, we must back off
to the previous value and reset the countlits value too. */
if (subcountlits > 0)
{
*reqchar = prevreqchar;
*countlits -= subcountlits;
}
/* If the maximum is also zero, we just omit the group from the output
altogether. */
if (repeat_max == 0)
{
code = previous;
goto END_REPEAT;
}
/* If the maximum is 1 or unlimited, we just have to stick in the
BRAZERO and do no more at this point. */
if (repeat_max <= 1)
{
memmove(previous+1, previous, len);
code++;
*previous++ = OP_BRAZERO + repeat_type;
}
/* If the maximum is greater than 1 and limited, we have to replicate
in a nested fashion, sticking OP_BRAZERO before each set of brackets.
The first one has to be handled carefully because it's the original
copy, which has to be moved up. The remainder can be handled by code
that is common with the non-zero minimum case below. We just have to
adjust the value or repeat_max, since one less copy is required. */
else
{
int offset;
memmove(previous+4, previous, len);
code += 4;
*previous++ = OP_BRAZERO + repeat_type;
*previous++ = OP_BRA;
/* We chain together the bracket offset fields that have to be
filled in later when the ends of the brackets are reached. */
offset = (bralink == NULL)? 0 : previous - bralink;
bralink = previous;
*previous++ = offset >> 8;
*previous++ = offset & 255;
}
repeat_max--;
}
/* If the minimum is greater than zero, replicate the group as many
times as necessary, and adjust the maximum to the number of subsequent
copies that we need. */
else
{
for (i = 1; i < repeat_min; i++)
{
memcpy(code, previous, len);
code += len;
}
if (repeat_max > 0) repeat_max -= repeat_min;
}
/* This code is common to both the zero and non-zero minimum cases. If
the maximum is limited, it replicates the group in a nested fashion,
remembering the bracket starts on a stack. In the case of a zero minimum,
the first one was set up above. In all cases the repeat_max now specifies
the number of additional copies needed. */
if (repeat_max >= 0)
{
for (i = repeat_max - 1; i >= 0; i--)
{
*code++ = OP_BRAZERO + repeat_type;
/* All but the final copy start a new nesting, maintaining the
chain of brackets outstanding. */
if (i != 0)
{
int offset;
*code++ = OP_BRA;
offset = (bralink == NULL)? 0 : code - bralink;
bralink = code;
*code++ = offset >> 8;
*code++ = offset & 255;
}
memcpy(code, previous, len);
code += len;
}
/* Now chain through the pending brackets, and fill in their length
fields (which are holding the chain links pro tem). */
while (bralink != NULL)
{
int oldlinkoffset;
int offset = code - bralink + 1;
uschar *bra = code - offset;
oldlinkoffset = (bra[1] << 8) + bra[2];
bralink = (oldlinkoffset == 0)? NULL : bralink - oldlinkoffset;
*code++ = OP_KET;
*code++ = bra[1] = offset >> 8;
*code++ = bra[2] = (offset & 255);
}
}
/* If the maximum is unlimited, set a repeater in the final copy. We
can't just offset backwards from the current code point, because we
don't know if there's been an options resetting after the ket. The
correct offset was computed above. */
else code[-ketoffset] = OP_KETRMAX + repeat_type;
}
/* Else there's some kind of shambles */
else
{
*errorptr = ERR11;
goto FAILED;
}
/* In all case we no longer have a previous item. */
END_REPEAT:
previous = NULL;
break;
/* Start of nested bracket sub-expression, or comment or lookahead or
lookbehind or option setting or condition. First deal with special things
that can come after a bracket; all are introduced by ?, and the appearance
of any of them means that this is not a referencing group. They were
checked for validity in the first pass over the string, so we don't have to
check for syntax errors here. */
case '(':
newoptions = options;
condref = -1;
if (*(++ptr) == '?')
{
int set, unset;
int *optset;
switch (*(++ptr))
{
case '#': /* Comment; skip to ket */
ptr++;
while (*ptr != ')') ptr++;
continue;
case ':': /* Non-extracting bracket */
bravalue = OP_BRA;
ptr++;
break;
case '(':
bravalue = OP_COND; /* Conditional group */
if ((cd->ctypes[*(++ptr)] & ctype_digit) != 0)
{
condref = *ptr - '0';
while (*(++ptr) != ')') condref = condref*10 + *ptr - '0';
ptr++;
}
else ptr--;
break;
case '=': /* Positive lookahead */
bravalue = OP_ASSERT;
ptr++;
break;
case '!': /* Negative lookahead */
bravalue = OP_ASSERT_NOT;
ptr++;
break;
case '<': /* Lookbehinds */
switch (*(++ptr))
{
case '=': /* Positive lookbehind */
bravalue = OP_ASSERTBACK;
ptr++;
break;
case '!': /* Negative lookbehind */
bravalue = OP_ASSERTBACK_NOT;
ptr++;
break;
default: /* Syntax error */
*errorptr = ERR24;
goto FAILED;
}
break;
case '>': /* One-time brackets */
bravalue = OP_ONCE;
ptr++;
break;
default: /* Option setting */
set = unset = 0;
optset = &set;
while (*ptr != ')' && *ptr != ':')
{
switch (*ptr++)
{
case '-': optset = &unset; break;
case 'i': *optset |= PCRE_CASELESS; break;
case 'm': *optset |= PCRE_MULTILINE; break;
case 's': *optset |= PCRE_DOTALL; break;
case 'x': *optset |= PCRE_EXTENDED; break;
case 'U': *optset |= PCRE_UNGREEDY; break;
case 'X': *optset |= PCRE_EXTRA; break;
default:
*errorptr = ERR12;
goto FAILED;
}
}
/* Set up the changed option bits, but don't change anything yet. */
newoptions = (options | set) & (~unset);
/* If the options ended with ')' this is not the start of a nested
group with option changes, so the options change at this level. At top
level there is nothing else to be done (the options will in fact have
been set from the start of compiling as a result of the first pass) but
at an inner level we must compile code to change the ims options if
necessary, and pass the new setting back so that it can be put at the
start of any following branches, and when this group ends, a resetting
item can be compiled. */
if (*ptr == ')')
{
if ((options & PCRE_INGROUP) != 0 &&
(options & PCRE_IMS) != (newoptions & PCRE_IMS))
{
*code++ = OP_OPT;
*code++ = *optchanged = newoptions & PCRE_IMS;
}
options = newoptions; /* Change options at this level */
previous = NULL; /* This item can't be repeated */
continue; /* It is complete */
}
/* If the options ended with ':' we are heading into a nested group
with possible change of options. Such groups are non-capturing and are
not assertions of any kind. All we need to do is skip over the ':';
the newoptions value is handled below. */
bravalue = OP_BRA;
ptr++;
}
}
/* Else we have a referencing group; adjust the opcode. */
else
{
if (++(*brackets) > EXTRACT_MAX)
{
*errorptr = ERR13;
goto FAILED;
}
bravalue = OP_BRA + *brackets;
}
/* Process nested bracketed re. Assertions may not be repeated, but other
kinds can be. We copy code into a non-register variable in order to be able
to pass its address because some compilers complain otherwise. Pass in a
new setting for the ims options if they have changed. */
previous = (bravalue >= OP_ONCE)? code : NULL;
*code = bravalue;
tempcode = code;
if (!compile_regex(
options | PCRE_INGROUP, /* Set for all nested groups */
((options & PCRE_IMS) != (newoptions & PCRE_IMS))?
newoptions & PCRE_IMS : -1, /* Pass ims options if changed */
brackets, /* Bracket level */
&tempcode, /* Where to put code (updated) */
&ptr, /* Input pointer (updated) */
errorptr, /* Where to put an error message */
(bravalue == OP_ASSERTBACK ||
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */
condref, /* Condition reference number */
&subreqchar, /* For possible last char */
&subcountlits, /* For literal count */
cd)) /* Tables block */
goto FAILED;
/* At the end of compiling, code is still pointing to the start of the
group, while tempcode has been updated to point past the end of the group
and any option resetting that may follow it. The pattern pointer (ptr)
is on the bracket. */
/* If this is a conditional bracket, check that there are no more than
two branches in the group. */
if (bravalue == OP_COND)
{
uschar *tc = code;
condcount = 0;
do {
condcount++;
tc += (tc[1] << 8) | tc[2];
}
while (*tc != OP_KET);
if (condcount > 2)
{
*errorptr = ERR27;
goto FAILED;
}
}
/* Handle updating of the required character. If the subpattern didn't
set one, leave it as it was. Otherwise, update it for normal brackets of
all kinds, forward assertions, and conditions with two branches. Don't
update the literal count for forward assertions, however. If the bracket
is followed by a quantifier with zero repeat, we have to back off. Hence
the definition of prevreqchar and subcountlits outside the main loop so
that they can be accessed for the back off. */
if (subreqchar > 0 &&
(bravalue >= OP_BRA || bravalue == OP_ONCE || bravalue == OP_ASSERT ||
(bravalue == OP_COND && condcount == 2)))
{
prevreqchar = *reqchar;
*reqchar = subreqchar;
if (bravalue != OP_ASSERT) *countlits += subcountlits;
}
/* Now update the main code pointer to the end of the group. */
code = tempcode;
/* Error if hit end of pattern */
if (*ptr != ')')
{
*errorptr = ERR14;
goto FAILED;
}
break;
/* Check \ for being a real metacharacter; if not, fall through and handle
it as a data character at the start of a string. Escape items are checked
for validity in the pre-compiling pass. */
case '\\':
tempptr = ptr;
c = check_escape(&ptr, errorptr, *brackets, options, FALSE, cd);
/* Handle metacharacters introduced by \. For ones like \d, the ESC_ values
are arranged to be the negation of the corresponding OP_values. For the
back references, the values are ESC_REF plus the reference number. Only
back references and those types that consume a character may be repeated.
We can test for values between ESC_b and ESC_Z for the latter; this may
have to change if any new ones are ever created. */
if (c < 0)
{
if (-c >= ESC_REF)
{
previous = code;
*code++ = OP_REF;
*code++ = -c - ESC_REF;
}
else
{
previous = (-c > ESC_b && -c < ESC_Z)? code : NULL;
*code++ = -c;
}
continue;
}
/* Data character: reset and fall through */
ptr = tempptr;
c = '\\';
/* Handle a run of data characters until a metacharacter is encountered.
The first character is guaranteed not to be whitespace or # when the
extended flag is set. */
NORMAL_CHAR:
default:
previous = code;
*code = OP_CHARS;
code += 2;
length = 0;
do
{
if ((options & PCRE_EXTENDED) != 0)
{
if ((cd->ctypes[c] & ctype_space) != 0) continue;
if (c == '#')
{
while ((c = *(++ptr)) != 0 && c != '\n');
if (c == 0) break;
continue;
}
}
/* Backslash may introduce a data char or a metacharacter. Escaped items
are checked for validity in the pre-compiling pass. Stop the string
before a metaitem. */
if (c == '\\')
{
tempptr = ptr;
c = check_escape(&ptr, errorptr, *brackets, options, FALSE, cd);
if (c < 0) { ptr = tempptr; break; }
}
/* Ordinary character or single-char escape */
*code++ = c;
length++;
}
/* This "while" is the end of the "do" above. */
while (length < 255 && (cd->ctypes[c = *(++ptr)] & ctype_meta) == 0);
/* Update the last character and the count of literals */
prevreqchar = (length > 1)? code[-2] : *reqchar;
*reqchar = code[-1];
*countlits += length;
/* Compute the length and set it in the data vector, and advance to
the next state. */
previous[1] = length;
if (length < 255) ptr--;
break;
}
} /* end of big loop */
/* Control never reaches here by falling through, only by a goto for all the
error states. Pass back the position in the pattern so that it can be displayed
to the user for diagnosing the error. */
FAILED:
*ptrptr = ptr;
return FALSE;
}
/*************************************************
* Compile sequence of alternatives *
*************************************************/
/* On entry, ptr is pointing past the bracket character, but on return
it points to the closing bracket, or vertical bar, or end of string.
The code variable is pointing at the byte into which the BRA operator has been
stored. If the ims options are changed at the start (for a (?ims: group) or
during any branch, we need to insert an OP_OPT item at the start of every
following branch to ensure they get set correctly at run time, and also pass
the new options into every subsequent branch compile.
Argument:
options the option bits
optchanged new ims options to set as if (?ims) were at the start, or -1
for no change
brackets -> int containing the number of extracting brackets used
codeptr -> the address of the current code pointer
ptrptr -> the address of the current pattern pointer
errorptr -> pointer to error message
lookbehind TRUE if this is a lookbehind assertion
condref > 0 for OPT_CREF setting at start of conditional group
reqchar -> place to put the last required character, or a negative number
countlits -> place to put the shortest literal count of any branch
cd points to the data block with tables pointers
Returns: TRUE on success
*/
static BOOL
compile_regex(int options, int optchanged, int *brackets, uschar **codeptr,
const uschar **ptrptr, const char **errorptr, BOOL lookbehind, int condref,
int *reqchar, int *countlits, compile_data *cd)
{
const uschar *ptr = *ptrptr;
uschar *code = *codeptr;
uschar *last_branch = code;
uschar *start_bracket = code;
uschar *reverse_count = NULL;
int oldoptions = options & PCRE_IMS;
int branchreqchar, branchcountlits;
*reqchar = -1;
*countlits = INT_MAX;
code += 3;
/* At the start of a reference-based conditional group, insert the reference
number as an OP_CREF item. */
if (condref > 0)
{
*code++ = OP_CREF;
*code++ = condref;
}
/* Loop for each alternative branch */
for (;;)
{
int length;
/* Handle change of options */
if (optchanged >= 0)
{
*code++ = OP_OPT;
*code++ = optchanged;
options = (options & ~PCRE_IMS) | optchanged;
}
/* Set up dummy OP_REVERSE if lookbehind assertion */
if (lookbehind)
{
*code++ = OP_REVERSE;
reverse_count = code;
*code++ = 0;
*code++ = 0;
}
/* Now compile the branch */
if (!compile_branch(options, brackets, &code, &ptr, errorptr, &optchanged,
&branchreqchar, &branchcountlits, cd))
{
*ptrptr = ptr;
return FALSE;
}
/* Fill in the length of the last branch */
length = code - last_branch;
last_branch[1] = length >> 8;
last_branch[2] = length & 255;
/* Save the last required character if all branches have the same; a current
value of -1 means unset, while -2 means "previous branch had no last required
char". */
if (*reqchar != -2)
{
if (branchreqchar >= 0)
{
if (*reqchar == -1) *reqchar = branchreqchar;
else if (*reqchar != branchreqchar) *reqchar = -2;
}
else *reqchar = -2;
}
/* Keep the shortest literal count */
if (branchcountlits < *countlits) *countlits = branchcountlits;
DPRINTF(("literal count = %d min=%d\n", branchcountlits, *countlits));
/* If lookbehind, check that this branch matches a fixed-length string,
and put the length into the OP_REVERSE item. Temporarily mark the end of
the branch with OP_END. */
if (lookbehind)
{
*code = OP_END;
length = find_fixedlength(last_branch);
DPRINTF(("fixed length = %d\n", length));
if (length < 0)
{
*errorptr = ERR25;
*ptrptr = ptr;
return FALSE;
}
reverse_count[0] = (length >> 8);
reverse_count[1] = length & 255;
}
/* Reached end of expression, either ')' or end of pattern. Insert a
terminating ket and the length of the whole bracketed item, and return,
leaving the pointer at the terminating char. If any of the ims options
were changed inside the group, compile a resetting op-code following. */
if (*ptr != '|')
{
length = code - start_bracket;
*code++ = OP_KET;
*code++ = length >> 8;
*code++ = length & 255;
if (optchanged >= 0)
{
*code++ = OP_OPT;
*code++ = oldoptions;
}
*codeptr = code;
*ptrptr = ptr;
return TRUE;
}
/* Another branch follows; insert an "or" node and advance the pointer. */
*code = OP_ALT;
last_branch = code;
code += 3;
ptr++;
}
/* Control never reaches here */
}
/*************************************************
* Find first significant op code *
*************************************************/
/* This is called by several functions that scan a compiled expression looking
for a fixed first character, or an anchoring op code etc. It skips over things
that do not influence this. For one application, a change of caseless option is
important.
Arguments:
code pointer to the start of the group
options pointer to external options
optbit the option bit whose changing is significant, or
zero if none are
optstop TRUE to return on option change, otherwise change the options
value and continue
Returns: pointer to the first significant opcode
*/
static const uschar*
first_significant_code(const uschar *code, int *options, int optbit,
BOOL optstop)
{
for (;;)
{
switch ((int)*code)
{
case OP_OPT:
if (optbit > 0 && ((int)code[1] & optbit) != (*options & optbit))
{
if (optstop) return code;
*options = (int)code[1];
}
code += 2;
break;
case OP_CREF:
code += 2;
break;
case OP_WORD_BOUNDARY:
case OP_NOT_WORD_BOUNDARY:
code++;
break;
case OP_ASSERT_NOT:
case OP_ASSERTBACK:
case OP_ASSERTBACK_NOT:
do code += (code[1] << 8) + code[2]; while (*code == OP_ALT);
code += 3;
break;
default:
return code;
}
}
/* Control never reaches here */
}
/*************************************************
* Check for anchored expression *
*************************************************/
/* Try to find out if this is an anchored regular expression. Consider each
alternative branch. If they all start with OP_SOD or OP_CIRC, or with a bracket
all of whose alternatives start with OP_SOD or OP_CIRC (recurse ad lib), then
it's anchored. However, if this is a multiline pattern, then only OP_SOD
counts, since OP_CIRC can match in the middle.
A branch is also implicitly anchored if it starts with .* and DOTALL is set,
because that will try the rest of the pattern at all possible matching points,
so there is no point trying them again.
Arguments:
code points to start of expression (the bracket)
options points to the options setting
Returns: TRUE or FALSE
*/
static BOOL
is_anchored(register const uschar *code, int *options)
{
do {
const uschar *scode = first_significant_code(code + 3, options,
PCRE_MULTILINE, FALSE);
register int op = *scode;
if (op >= OP_BRA || op == OP_ASSERT || op == OP_ONCE || op == OP_COND)
{ if (!is_anchored(scode, options)) return FALSE; }
else if ((op == OP_TYPESTAR || op == OP_TYPEMINSTAR) &&
(*options & PCRE_DOTALL) != 0)
{ if (scode[1] != OP_ANY) return FALSE; }
else if (op != OP_SOD &&
((*options & PCRE_MULTILINE) != 0 || op != OP_CIRC))
return FALSE;
code += (code[1] << 8) + code[2];
}
while (*code == OP_ALT);
return TRUE;
}
/*************************************************
* Check for starting with ^ or .* *
*************************************************/
/* This is called to find out if every branch starts with ^ or .* so that
"first char" processing can be done to speed things up in multiline
matching and for non-DOTALL patterns that start with .* (which must start at
the beginning or after \n).
Argument: points to start of expression (the bracket)
Returns: TRUE or FALSE
*/
static BOOL
is_startline(const uschar *code)
{
do {
const uschar *scode = first_significant_code(code + 3, NULL, 0, FALSE);
register int op = *scode;
if (op >= OP_BRA || op == OP_ASSERT || op == OP_ONCE || op == OP_COND)
{ if (!is_startline(scode)) return FALSE; }
else if (op == OP_TYPESTAR || op == OP_TYPEMINSTAR)
{ if (scode[1] != OP_ANY) return FALSE; }
else if (op != OP_CIRC) return FALSE;
code += (code[1] << 8) + code[2];
}
while (*code == OP_ALT);
return TRUE;
}
/*************************************************
* Check for fixed first char *
*************************************************/
/* Try to find out if there is a fixed first character. This is called for
unanchored expressions, as it speeds up their processing quite considerably.
Consider each alternative branch. If they all start with the same char, or with
a bracket all of whose alternatives start with the same char (recurse ad lib),
then we return that char, otherwise -1.
Arguments:
code points to start of expression (the bracket)
options pointer to the options (used to check casing changes)
Returns: -1 or the fixed first char
*/
static int
find_firstchar(const uschar *code, int *options)
{
register int c = -1;
do {
int d;
const uschar *scode = first_significant_code(code + 3, options,
PCRE_CASELESS, TRUE);
register int op = *scode;
if (op >= OP_BRA) op = OP_BRA;
switch(op)
{
default:
return -1;
case OP_BRA:
case OP_ASSERT:
case OP_ONCE:
case OP_COND:
if ((d = find_firstchar(scode, options)) < 0) return -1;
if (c < 0) c = d; else if (c != d) return -1;
break;
case OP_EXACT: /* Fall through */
scode++;
case OP_CHARS: /* Fall through */
scode++;
case OP_PLUS:
case OP_MINPLUS:
if (c < 0) c = scode[1]; else if (c != scode[1]) return -1;
break;
}
code += (code[1] << 8) + code[2];
}
while (*code == OP_ALT);
return c;
}
/*************************************************
* Compile a Regular Expression *
*************************************************/
/* This function takes a string and returns a pointer to a block of store
holding a compiled version of the expression.
Arguments:
pattern the regular expression
options various option bits
errorptr pointer to pointer to error text
erroroffset ptr offset in pattern where error was detected
tables pointer to character tables or NULL
Returns: pointer to compiled data block, or NULL on error,
with errorptr and erroroffset set
*/
pcre *
pcre_compile(const char *pattern, int options, const char **errorptr,
int *erroroffset, const unsigned char *tables)
{
real_pcre *re;
int length = 3; /* For initial BRA plus length */
int runlength;
int c, size, reqchar, countlits;
int bracount = 0;
int top_backref = 0;
int branch_extra = 0;
int branch_newextra;
unsigned int brastackptr = 0;
uschar *code;
const uschar *ptr;
compile_data compile_block;
int brastack[BRASTACK_SIZE];
uschar bralenstack[BRASTACK_SIZE];
#ifdef DEBUG
uschar *code_base, *code_end;
#endif
/* We can't pass back an error message if errorptr is NULL; I guess the best we
can do is just return NULL. */
if (errorptr == NULL) return NULL;
*errorptr = NULL;
/* However, we can give a message for this error */
if (erroroffset == NULL)
{
*errorptr = ERR16;
return NULL;
}
*erroroffset = 0;
if ((options & ~PUBLIC_OPTIONS) != 0)
{
*errorptr = ERR17;
return NULL;
}
/* Set up pointers to the individual character tables */
if (tables == NULL) tables = pcre_default_tables;
compile_block.lcc = tables + lcc_offset;
compile_block.fcc = tables + fcc_offset;
compile_block.cbits = tables + cbits_offset;
compile_block.ctypes = tables + ctypes_offset;
/* Reflect pattern for debugging output */
DPRINTF(("------------------------------------------------------------------\n"));
DPRINTF(("%s\n", pattern));
/* The first thing to do is to make a pass over the pattern to compute the
amount of store required to hold the compiled code. This does not have to be
perfect as long as errors are overestimates. At the same time we can detect any
internal flag settings. Make an attempt to correct for any counted white space
if an "extended" flag setting appears late in the pattern. We can't be so
clever for #-comments. */
ptr = (const uschar *)(pattern - 1);
while ((c = *(++ptr)) != 0)
{
int min, max;
int class_charcount;
if ((options & PCRE_EXTENDED) != 0)
{
if ((compile_block.ctypes[c] & ctype_space) != 0) continue;
if (c == '#')
{
while ((c = *(++ptr)) != 0 && c != '\n');
continue;
}
}
switch(c)
{
/* A backslashed item may be an escaped "normal" character or a
character type. For a "normal" character, put the pointers and
character back so that tests for whitespace etc. in the input
are done correctly. */
case '\\':
{
const uschar *save_ptr = ptr;
c = check_escape(&ptr, errorptr, bracount, options, FALSE, &compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
if (c >= 0)
{
ptr = save_ptr;
c = '\\';
goto NORMAL_CHAR;
}
}
length++;
/* A back reference needs an additional char, plus either one or 5
bytes for a repeat. We also need to keep the value of the highest
back reference. */
if (c <= -ESC_REF)
{
int refnum = -c - ESC_REF;
if (refnum > top_backref) top_backref = refnum;
length++; /* For single back reference */
if (ptr[1] == '{' && is_counted_repeat(ptr+2, &compile_block))
{
ptr = read_repeat_counts(ptr+2, &min, &max, errorptr, &compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
if ((min == 0 && (max == 1 || max == -1)) ||
(min == 1 && max == -1))
length++;
else length += 5;
if (ptr[1] == '?') ptr++;
}
}
continue;
case '^':
case '.':
case '$':
case '*': /* These repeats won't be after brackets; */
case '+': /* those are handled separately */
case '?':
length++;
continue;
/* This covers the cases of repeats after a single char, metachar, class,
or back reference. */
case '{':
if (!is_counted_repeat(ptr+1, &compile_block)) goto NORMAL_CHAR;
ptr = read_repeat_counts(ptr+1, &min, &max, errorptr, &compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
if ((min == 0 && (max == 1 || max == -1)) ||
(min == 1 && max == -1))
length++;
else
{
length--; /* Uncount the original char or metachar */
if (min == 1) length++; else if (min > 0) length += 4;
if (max > 0) length += 4; else length += 2;
}
if (ptr[1] == '?') ptr++;
continue;
/* An alternation contains an offset to the next branch or ket. If any ims
options changed in the previous branch(es), and/or if we are in a
lookbehind assertion, extra space will be needed at the start of the
branch. This is handled by branch_extra. */
case '|':
length += 3 + branch_extra;
continue;
/* A character class uses 33 characters. Don't worry about character types
that aren't allowed in classes - they'll get picked up during the compile.
A character class that contains only one character uses 2 or 3 bytes,
depending on whether it is negated or not. Notice this where we can. */
case '[':
class_charcount = 0;
if (*(++ptr) == '^') ptr++;
do
{
if (*ptr == '\\')
{
int ch = check_escape(&ptr, errorptr, bracount, options, TRUE,
&compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
if (-ch == ESC_b) class_charcount++; else class_charcount = 10;
}
else class_charcount++;
ptr++;
}
while (*ptr != 0 && *ptr != ']');
/* Repeats for negated single chars are handled by the general code */
if (class_charcount == 1) length += 3; else
{
length += 33;
/* A repeat needs either 1 or 5 bytes. */
if (*ptr != 0 && ptr[1] == '{' && is_counted_repeat(ptr+2, &compile_block))
{
ptr = read_repeat_counts(ptr+2, &min, &max, errorptr, &compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
if ((min == 0 && (max == 1 || max == -1)) ||
(min == 1 && max == -1))
length++;
else length += 5;
if (ptr[1] == '?') ptr++;
}
}
continue;
/* Brackets may be genuine groups or special things */
case '(':
branch_newextra = 0;
/* Handle special forms of bracket, which all start (? */
if (ptr[1] == '?')
{
int set, unset;
int *optset;
switch (c = ptr[2])
{
/* Skip over comments entirely */
case '#':
ptr += 3;
while (*ptr != 0 && *ptr != ')') ptr++;
if (*ptr == 0)
{
*errorptr = ERR18;
goto PCRE_ERROR_RETURN;
}
continue;
/* Non-referencing groups and lookaheads just move the pointer on, and
then behave like a non-special bracket, except that they don't increment
the count of extracting brackets. Ditto for the "once only" bracket,
which is in Perl from version 5.005. */
case ':':
case '=':
case '!':
case '>':
ptr += 2;
break;
/* Lookbehinds are in Perl from version 5.005 */
case '<':
if (ptr[3] == '=' || ptr[3] == '!')
{
ptr += 3;
branch_newextra = 3;
length += 3; /* For the first branch */
break;
}
*errorptr = ERR24;
goto PCRE_ERROR_RETURN;
/* Conditionals are in Perl from version 5.005. The bracket must either
be followed by a number (for bracket reference) or by an assertion
group. */
case '(':
if ((compile_block.ctypes[ptr[3]] & ctype_digit) != 0)
{
ptr += 4;
length += 2;
while ((compile_block.ctypes[*ptr] & ctype_digit) != 0) ptr++;
if (*ptr != ')')
{
*errorptr = ERR26;
goto PCRE_ERROR_RETURN;
}
}
else /* An assertion must follow */
{
ptr++; /* Can treat like ':' as far as spacing is concerned */
if (ptr[2] != '?' || strchr("=!<", ptr[3]) == NULL)
{
ptr += 2; /* To get right offset in message */
*errorptr = ERR28;
goto PCRE_ERROR_RETURN;
}
}
break;
/* Else loop checking valid options until ) is met. Anything else is an
error. If we are without any brackets, i.e. at top level, the settings
act as if specified in the options, so massage the options immediately.
This is for backward compatibility with Perl 5.004. */
default:
set = unset = 0;
optset = &set;
ptr += 2;
for (;; ptr++)
{
c = *ptr;
switch (c)
{
case 'i':
*optset |= PCRE_CASELESS;
continue;
case 'm':
*optset |= PCRE_MULTILINE;
continue;
case 's':
*optset |= PCRE_DOTALL;
continue;
case 'x':
*optset |= PCRE_EXTENDED;
continue;
case 'X':
*optset |= PCRE_EXTRA;
continue;
case 'U':
*optset |= PCRE_UNGREEDY;
continue;
case '-':
optset = &unset;
continue;
/* A termination by ')' indicates an options-setting-only item;
this is global at top level; otherwise nothing is done here and
it is handled during the compiling process on a per-bracket-group
basis. */
case ')':
if (brastackptr == 0)
{
options = (options | set) & (~unset);
set = unset = 0; /* To save length */
}
/* Fall through */
/* A termination by ':' indicates the start of a nested group with
the given options set. This is again handled at compile time, but
we must allow for compiled space if any of the ims options are
set. We also have to allow for resetting space at the end of
the group, which is why 4 is added to the length and not just 2.
If there are several changes of options within the same group, this
will lead to an over-estimate on the length, but this shouldn't
matter very much. We also have to allow for resetting options at
the start of any alternations, which we do by setting
branch_newextra to 2. Finally, we record whether the case-dependent
flag ever changes within the regex. This is used by the "required
character" code. */
case ':':
if (((set|unset) & PCRE_IMS) != 0)
{
length += 4;
branch_newextra = 2;
if (((set|unset) & PCRE_CASELESS) != 0) options |= PCRE_ICHANGED;
}
goto END_OPTIONS;
/* Unrecognized option character */
default:
*errorptr = ERR12;
goto PCRE_ERROR_RETURN;
}
}
/* If we hit a closing bracket, that's it - this is a freestanding
option-setting. We need to ensure that branch_extra is updated if
necessary. The only values branch_newextra can have here are 0 or 2.
If the value is 2, then branch_extra must either be 2 or 5, depending
on whether this is a lookbehind group or not. */
END_OPTIONS:
if (c == ')')
{
if (branch_newextra == 2 && (branch_extra == 0 || branch_extra == 3))
branch_extra += branch_newextra;
continue;
}
/* If options were terminated by ':' control comes here. Fall through
to handle the group below. */
}
}
/* Extracting brackets must be counted so we can process escapes in a
Perlish way. */
else bracount++;
/* Non-special forms of bracket. Save length for computing whole length
at end if there's a repeat that requires duplication of the group. Also
save the current value of branch_extra, and start the new group with
the new value. If non-zero, this will either be 2 for a (?imsx: group, or 3
for a lookbehind assertion. */
if (brastackptr >= sizeof(brastack)/sizeof(int))
{
*errorptr = ERR19;
goto PCRE_ERROR_RETURN;
}
bralenstack[brastackptr] = branch_extra;
branch_extra = branch_newextra;
brastack[brastackptr++] = length;
length += 3;
continue;
/* Handle ket. Look for subsequent max/min; for certain sets of values we
have to replicate this bracket up to that many times. If brastackptr is
0 this is an unmatched bracket which will generate an error, but take care
not to try to access brastack[-1] when computing the length and restoring
the branch_extra value. */
case ')':
length += 3;
{
int minval = 1;
int maxval = 1;
int duplength;
if (brastackptr > 0)
{
duplength = length - brastack[--brastackptr];
branch_extra = bralenstack[brastackptr];
}
else duplength = 0;
/* Leave ptr at the final char; for read_repeat_counts this happens
automatically; for the others we need an increment. */
if ((c = ptr[1]) == '{' && is_counted_repeat(ptr+2, &compile_block))
{
ptr = read_repeat_counts(ptr+2, &minval, &maxval, errorptr,
&compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
}
else if (c == '*') { minval = 0; maxval = -1; ptr++; }
else if (c == '+') { maxval = -1; ptr++; }
else if (c == '?') { minval = 0; ptr++; }
/* If the minimum is zero, we have to allow for an OP_BRAZERO before the
group, and if the maximum is greater than zero, we have to replicate
maxval-1 times; each replication acquires an OP_BRAZERO plus a nesting
bracket set - hence the 7. */
if (minval == 0)
{
length++;
if (maxval > 0) length += (maxval - 1) * (duplength + 7);
}
/* When the minimum is greater than zero, 1 we have to replicate up to
minval-1 times, with no additions required in the copies. Then, if
there is a limited maximum we have to replicate up to maxval-1 times
allowing for a BRAZERO item before each optional copy and nesting
brackets for all but one of the optional copies. */
else
{
length += (minval - 1) * duplength;
if (maxval > minval) /* Need this test as maxval=-1 means no limit */
length += (maxval - minval) * (duplength + 7) - 6;
}
}
continue;
/* Non-special character. For a run of such characters the length required
is the number of characters + 2, except that the maximum run length is 255.
We won't get a skipped space or a non-data escape or the start of a #
comment as the first character, so the length can't be zero. */
NORMAL_CHAR:
default:
length += 2;
runlength = 0;
do
{
if ((options & PCRE_EXTENDED) != 0)
{
if ((compile_block.ctypes[c] & ctype_space) != 0) continue;
if (c == '#')
{
while ((c = *(++ptr)) != 0 && c != '\n');
continue;
}
}
/* Backslash may introduce a data char or a metacharacter; stop the
string before the latter. */
if (c == '\\')
{
const uschar *saveptr = ptr;
c = check_escape(&ptr, errorptr, bracount, options, FALSE,
&compile_block);
if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
if (c < 0) { ptr = saveptr; break; }
}
/* Ordinary character or single-char escape */
runlength++;
}
/* This "while" is the end of the "do" above. */
while (runlength < 255 &&
(compile_block.ctypes[c = *(++ptr)] & ctype_meta) == 0);
ptr--;
length += runlength;
continue;
}
}
length += 4; /* For final KET and END */
if (length > 65539)
{
*errorptr = ERR20;
return NULL;
}
/* Compute the size of data block needed and get it, either from malloc or
externally provided function. We specify "code[0]" in the offsetof() expression
rather than just "code", because it has been reported that one broken compiler
fails on "code" because it is also an independent variable. It should make no
difference to the value of the offsetof(). */
size = length + offsetof(real_pcre, code[0]);
re = (real_pcre *)(pcre_malloc)(size);
if (re == NULL)
{
*errorptr = ERR21;
return NULL;
}
/* Put in the magic number and the options. */
re->magic_number = MAGIC_NUMBER;
re->options = options;
re->tables = tables;
/* Set up a starting, non-extracting bracket, then compile the expression. On
error, *errorptr will be set non-NULL, so we don't need to look at the result
of the function here. */
ptr = (const uschar *)pattern;
code = re->code;
*code = OP_BRA;
bracount = 0;
(void)compile_regex(options, -1, &bracount, &code, &ptr, errorptr, FALSE, -1,
&reqchar, &countlits, &compile_block);
re->top_bracket = bracount;
re->top_backref = top_backref;
/* If not reached end of pattern on success, there's an excess bracket. */
if (*errorptr == NULL && *ptr != 0) *errorptr = ERR22;
/* Fill in the terminating state and check for disastrous overflow, but
if debugging, leave the test till after things are printed out. */
*code++ = OP_END;
#ifndef DEBUG
if (code - re->code > length) *errorptr = ERR23;
#endif
/* Give an error if there's back reference to a non-existent capturing
subpattern. */
if (top_backref > re->top_bracket) *errorptr = ERR15;
/* Failed to compile */
if (*errorptr != NULL)
{
(pcre_free)(re);
PCRE_ERROR_RETURN:
*erroroffset = ptr - (const uschar *)pattern;
return NULL;
}
/* If the anchored option was not passed, set flag if we can determine that the
pattern is anchored by virtue of ^ characters or \A or anything else (such as
starting with .* when DOTALL is set).
Otherwise, see if we can determine what the first character has to be, because
that speeds up unanchored matches no end. If not, see if we can set the
PCRE_STARTLINE flag. This is helpful for multiline matches when all branches
start with ^. and also when all branches start with .* for non-DOTALL matches.
*/
if ((options & PCRE_ANCHORED) == 0)
{
int temp_options = options;
if (is_anchored(re->code, &temp_options))
re->options |= PCRE_ANCHORED;
else
{
int ch = find_firstchar(re->code, &temp_options);
if (ch >= 0)
{
re->first_char = ch;
re->options |= PCRE_FIRSTSET;
}
else if (is_startline(re->code))
re->options |= PCRE_STARTLINE;
}
}
/* Save the last required character if there are at least two literal
characters on all paths, or if there is no first character setting. */
if (reqchar >= 0 && (countlits > 1 || (re->options & PCRE_FIRSTSET) == 0))
{
re->req_char = reqchar;
re->options |= PCRE_REQCHSET;
}
/* Print out the compiled data for debugging */
#ifdef DEBUG
printf("Length = %d top_bracket = %d top_backref = %d\n",
length, re->top_bracket, re->top_backref);
if (re->options != 0)
{
printf("%s%s%s%s%s%s%s%s%s\n",
((re->options & PCRE_ANCHORED) != 0)? "anchored " : "",
((re->options & PCRE_CASELESS) != 0)? "caseless " : "",
((re->options & PCRE_ICHANGED) != 0)? "case state changed " : "",
((re->options & PCRE_EXTENDED) != 0)? "extended " : "",
((re->options & PCRE_MULTILINE) != 0)? "multiline " : "",
((re->options & PCRE_DOTALL) != 0)? "dotall " : "",
((re->options & PCRE_DOLLAR_ENDONLY) != 0)? "endonly " : "",
((re->options & PCRE_EXTRA) != 0)? "extra " : "",
((re->options & PCRE_UNGREEDY) != 0)? "ungreedy " : "");
}
if ((re->options & PCRE_FIRSTSET) != 0)
{
if (isprint(re->first_char)) printf("First char = %c\n", re->first_char);
else printf("First char = \\x%02x\n", re->first_char);
}
if ((re->options & PCRE_REQCHSET) != 0)
{
if (isprint(re->req_char)) printf("Req char = %c\n", re->req_char);
else printf("Req char = \\x%02x\n", re->req_char);
}
code_end = code;
code_base = code = re->code;
while (code < code_end)
{
int charlength;
printf("%3d ", code - code_base);
if (*code >= OP_BRA)
{
printf("%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA);
code += 2;
}
else switch(*code)
{
case OP_OPT:
printf(" %.2x %s", code[1], OP_names[*code]);
code++;
break;
case OP_COND:
printf("%3d Cond", (code[1] << 8) + code[2]);
code += 2;
break;
case OP_CREF:
printf(" %.2d %s", code[1], OP_names[*code]);
code++;
break;
case OP_CHARS:
charlength = *(++code);
printf("%3d ", charlength);
while (charlength-- > 0)
if (isprint(c = *(++code))) printf("%c", c); else printf("\\x%02x", c);
break;
case OP_KETRMAX:
case OP_KETRMIN:
case OP_ALT:
case OP_KET:
case OP_ASSERT:
case OP_ASSERT_NOT:
case OP_ASSERTBACK:
case OP_ASSERTBACK_NOT:
case OP_ONCE:
printf("%3d %s", (code[1] << 8) + code[2], OP_names[*code]);
code += 2;
break;
case OP_REVERSE:
printf("%3d %s", (code[1] << 8) + code[2], OP_names[*code]);
code += 2;
break;
case OP_STAR:
case OP_MINSTAR:
case OP_PLUS:
case OP_MINPLUS:
case OP_QUERY:
case OP_MINQUERY:
case OP_TYPESTAR:
case OP_TYPEMINSTAR:
case OP_TYPEPLUS:
case OP_TYPEMINPLUS:
case OP_TYPEQUERY:
case OP_TYPEMINQUERY:
if (*code >= OP_TYPESTAR)
printf(" %s", OP_names[code[1]]);
else if (isprint(c = code[1])) printf(" %c", c);
else printf(" \\x%02x", c);
printf("%s", OP_names[*code++]);
break;
case OP_EXACT:
case OP_UPTO:
case OP_MINUPTO:
if (isprint(c = code[3])) printf(" %c{", c);
else printf(" \\x%02x{", c);
if (*code != OP_EXACT) printf("0,");
printf("%d}", (code[1] << 8) + code[2]);
if (*code == OP_MINUPTO) printf("?");
code += 3;
break;
case OP_TYPEEXACT:
case OP_TYPEUPTO:
case OP_TYPEMINUPTO:
printf(" %s{", OP_names[code[3]]);
if (*code != OP_TYPEEXACT) printf(",");
printf("%d}", (code[1] << 8) + code[2]);
if (*code == OP_TYPEMINUPTO) printf("?");
code += 3;
break;
case OP_NOT:
if (isprint(c = *(++code))) printf(" [^%c]", c);
else printf(" [^\\x%02x]", c);
break;
case OP_NOTSTAR:
case OP_NOTMINSTAR:
case OP_NOTPLUS:
case OP_NOTMINPLUS:
case OP_NOTQUERY:
case OP_NOTMINQUERY:
if (isprint(c = code[1])) printf(" [^%c]", c);
else printf(" [^\\x%02x]", c);
printf("%s", OP_names[*code++]);
break;
case OP_NOTEXACT:
case OP_NOTUPTO:
case OP_NOTMINUPTO:
if (isprint(c = code[3])) printf(" [^%c]{", c);
else printf(" [^\\x%02x]{", c);
if (*code != OP_NOTEXACT) printf(",");
printf("%d}", (code[1] << 8) + code[2]);
if (*code == OP_NOTMINUPTO) printf("?");
code += 3;
break;
case OP_REF:
printf(" \\%d", *(++code));
code ++;
goto CLASS_REF_REPEAT;
case OP_CLASS:
{
int i, min, max;
code++;
printf(" [");
for (i = 0; i < 256; i++)
{
if ((code[i/8] & (1 << (i&7))) != 0)
{
int j;
for (j = i+1; j < 256; j++)
if ((code[j/8] & (1 << (j&7))) == 0) break;
if (i == '-' || i == ']') printf("\\");
if (isprint(i)) printf("%c", i); else printf("\\x%02x", i);
if (--j > i)
{
printf("-");
if (j == '-' || j == ']') printf("\\");
if (isprint(j)) printf("%c", j); else printf("\\x%02x", j);
}
i = j;
}
}
printf("]");
code += 32;
CLASS_REF_REPEAT:
switch(*code)
{
case OP_CRSTAR:
case OP_CRMINSTAR:
case OP_CRPLUS:
case OP_CRMINPLUS:
case OP_CRQUERY:
case OP_CRMINQUERY:
printf("%s", OP_names[*code]);
break;
case OP_CRRANGE:
case OP_CRMINRANGE:
min = (code[1] << 8) + code[2];
max = (code[3] << 8) + code[4];
if (max == 0) printf("{%d,}", min);
else printf("{%d,%d}", min, max);
if (*code == OP_CRMINRANGE) printf("?");
code += 4;
break;
default:
code--;
}
}
break;
/* Anything else is just a one-node item */
default:
printf(" %s", OP_names[*code]);
break;
}
code++;
printf("\n");
}
printf("------------------------------------------------------------------\n");
/* This check is done here in the debugging case so that the code that
was compiled can be seen. */
if (code - re->code > length)
{
*errorptr = ERR23;
(pcre_free)(re);
*erroroffset = ptr - (uschar *)pattern;
return NULL;
}
#endif
return (pcre *)re;
}
/*************************************************
* Match a back-reference *
*************************************************/
/* If a back reference hasn't been set, the length that is passed is greater
than the number of characters left in the string, so the match fails.
Arguments:
offset index into the offset vector
eptr points into the subject
length length to be matched
md points to match data block
ims the ims flags
Returns: TRUE if matched
*/
static BOOL
match_ref(int offset, register const uschar *eptr, int length, match_data *md,
unsigned long int ims)
{
const uschar *p = md->start_subject + md->offset_vector[offset];
#ifdef DEBUG
if (eptr >= md->end_subject)
printf("matching subject <null>");
else
{
printf("matching subject ");
pchars(eptr, length, TRUE, md);
}
printf(" against backref ");
pchars(p, length, FALSE, md);
printf("\n");
#endif
/* Always fail if not enough characters left */
if (length > md->end_subject - eptr) return FALSE;
/* Separate the caselesss case for speed */
if ((ims & PCRE_CASELESS) != 0)
{
while (length-- > 0)
if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE;
}
else
{ while (length-- > 0) if (*p++ != *eptr++) return FALSE; }
return TRUE;
}
/*************************************************
* Match from current position *
*************************************************/
/* On entry ecode points to the first opcode, and eptr to the first character
in the subject string, while eptrb holds the value of eptr at the start of the
last bracketed group - used for breaking infinite loops matching zero-length
strings.
Arguments:
eptr pointer in subject
ecode position in code
offset_top current top pointer
md pointer to "static" info for the match
ims current /i, /m, and /s options
condassert TRUE if called to check a condition assertion
eptrb eptr at start of last bracket
Returns: TRUE if matched
*/
static BOOL
match(register const uschar *eptr, register const uschar *ecode,
int offset_top, match_data *md, unsigned long int ims, BOOL condassert,
const uschar *eptrb)
{
unsigned long int original_ims = ims; /* Save for resetting on ')' */
for (;;)
{
int op = (int)*ecode;
int min, max, ctype;
register int i;
register int c;
BOOL minimize = FALSE;
/* Opening capturing bracket. If there is space in the offset vector, save
the current subject position in the working slot at the top of the vector. We
mustn't change the current values of the data slot, because they may be set
from a previous iteration of this group, and be referred to by a reference
inside the group.
If the bracket fails to match, we need to restore this value and also the
values of the final offsets, in case they were set by a previous iteration of
the same bracket.
If there isn't enough space in the offset vector, treat this as if it were a
non-capturing bracket. Don't worry about setting the flag for the error case
here; that is handled in the code for KET. */
if (op > OP_BRA)
{
int number = op - OP_BRA;
int offset = number << 1;
#ifdef DEBUG
printf("start bracket %d subject=", number);
pchars(eptr, 16, TRUE, md);
printf("\n");
#endif
if (offset < md->offset_max)
{
int save_offset1 = md->offset_vector[offset];
int save_offset2 = md->offset_vector[offset+1];
int save_offset3 = md->offset_vector[md->offset_end - number];
DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
md->offset_vector[md->offset_end - number] = eptr - md->start_subject;
do
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr)) return TRUE;
ecode += (ecode[1] << 8) + ecode[2];
}
while (*ecode == OP_ALT);
DPRINTF(("bracket %d failed\n", number));
md->offset_vector[offset] = save_offset1;
md->offset_vector[offset+1] = save_offset2;
md->offset_vector[md->offset_end - number] = save_offset3;
return FALSE;
}
/* Insufficient room for saving captured contents */
else op = OP_BRA;
}
/* Other types of node can be handled by a switch */
switch(op)
{
case OP_BRA: /* Non-capturing bracket: optimized */
DPRINTF(("start bracket 0\n"));
do
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr)) return TRUE;
ecode += (ecode[1] << 8) + ecode[2];
}
while (*ecode == OP_ALT);
DPRINTF(("bracket 0 failed\n"));
return FALSE;
/* Conditional group: compilation checked that there are no more than
two branches. If the condition is false, skipping the first branch takes us
past the end if there is only one branch, but that's OK because that is
exactly what going to the ket would do. */
case OP_COND:
if (ecode[3] == OP_CREF) /* Condition is extraction test */
{
int offset = ecode[4] << 1; /* Doubled reference number */
return match(eptr,
ecode + ((offset < offset_top && md->offset_vector[offset] >= 0)?
5 : 3 + (ecode[1] << 8) + ecode[2]),
offset_top, md, ims, FALSE, eptr);
}
/* The condition is an assertion. Call match() to evaluate it - setting
the final argument TRUE causes it to stop at the end of an assertion. */
else
{
if (match(eptr, ecode+3, offset_top, md, ims, TRUE, NULL))
{
ecode += 3 + (ecode[4] << 8) + ecode[5];
while (*ecode == OP_ALT) ecode += (ecode[1] << 8) + ecode[2];
}
else ecode += (ecode[1] << 8) + ecode[2];
return match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr);
}
/* Control never reaches here */
/* Skip over conditional reference data if encountered (should not be) */
case OP_CREF:
ecode += 2;
break;
/* End of the pattern. If PCRE_NOTEMPTY is set, fail if we have matched
an empty string - recursion will then try other alternatives, if any. */
case OP_END:
if (md->notempty && eptr == md->start_match) return FALSE;
md->end_match_ptr = eptr; /* Record where we ended */
md->end_offset_top = offset_top; /* and how many extracts were taken */
return TRUE;
/* Change option settings */
case OP_OPT:
ims = ecode[1];
ecode += 2;
DPRINTF(("ims set to %02lx\n", ims));
break;
/* Assertion brackets. Check the alternative branches in turn - the
matching won't pass the KET for an assertion. If any one branch matches,
the assertion is true. Lookbehind assertions have an OP_REVERSE item at the
start of each branch to move the current point backwards, so the code at
this level is identical to the lookahead case. */
case OP_ASSERT:
case OP_ASSERTBACK:
do
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, NULL)) break;
ecode += (ecode[1] << 8) + ecode[2];
}
while (*ecode == OP_ALT);
if (*ecode == OP_KET) return FALSE;
/* If checking an assertion for a condition, return TRUE. */
if (condassert) return TRUE;
/* Continue from after the assertion, updating the offsets high water
mark, since extracts may have been taken during the assertion. */
do ecode += (ecode[1] << 8) + ecode[2]; while (*ecode == OP_ALT);
ecode += 3;
offset_top = md->end_offset_top;
continue;
/* Negative assertion: all branches must fail to match */
case OP_ASSERT_NOT:
case OP_ASSERTBACK_NOT:
do
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, NULL)) return FALSE;
ecode += (ecode[1] << 8) + ecode[2];
}
while (*ecode == OP_ALT);
if (condassert) return TRUE;
ecode += 3;
continue;
/* Move the subject pointer back. This occurs only at the start of
each branch of a lookbehind assertion. If we are too close to the start to
move back, this match function fails. */
case OP_REVERSE:
eptr -= (ecode[1] << 8) + ecode[2];
if (eptr < md->start_subject) return FALSE;
ecode += 3;
break;
/* "Once" brackets are like assertion brackets except that after a match,
the point in the subject string is not moved back. Thus there can never be
a move back into the brackets. Check the alternative branches in turn - the
matching won't pass the KET for this kind of subpattern. If any one branch
matches, we carry on as at the end of a normal bracket, leaving the subject
pointer. */
case OP_ONCE:
{
const uschar *prev = ecode;
do
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr)) break;
ecode += (ecode[1] << 8) + ecode[2];
}
while (*ecode == OP_ALT);
/* If hit the end of the group (which could be repeated), fail */
if (*ecode != OP_ONCE && *ecode != OP_ALT) return FALSE;
/* Continue as from after the assertion, updating the offsets high water
mark, since extracts may have been taken. */
do ecode += (ecode[1] << 8) + ecode[2]; while (*ecode == OP_ALT);
offset_top = md->end_offset_top;
eptr = md->end_match_ptr;
/* For a non-repeating ket, just continue at this level. This also
happens for a repeating ket if no characters were matched in the group.
This is the forcible breaking of infinite loops as implemented in Perl
5.005. If there is an options reset, it will get obeyed in the normal
course of events. */
if (*ecode == OP_KET || eptr == eptrb)
{
ecode += 3;
break;
}
/* The repeating kets try the rest of the pattern or restart from the
preceding bracket, in the appropriate order. We need to reset any options
that changed within the bracket before re-running it, so check the next
opcode. */
if (ecode[3] == OP_OPT)
{
ims = (ims & ~PCRE_IMS) | ecode[4];
DPRINTF(("ims set to %02lx at group repeat\n", ims));
}
if (*ecode == OP_KETRMIN)
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr) ||
match(eptr, prev, offset_top, md, ims, FALSE, eptr)) return TRUE;
}
else /* OP_KETRMAX */
{
if (match(eptr, prev, offset_top, md, ims, FALSE, eptr) ||
match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr)) return TRUE;
}
}
return FALSE;
/* An alternation is the end of a branch; scan along to find the end of the
bracketed group and go to there. */
case OP_ALT:
do ecode += (ecode[1] << 8) + ecode[2]; while (*ecode == OP_ALT);
break;
/* BRAZERO and BRAMINZERO occur just before a bracket group, indicating
that it may occur zero times. It may repeat infinitely, or not at all -
i.e. it could be ()* or ()? in the pattern. Brackets with fixed upper
repeat limits are compiled as a number of copies, with the optional ones
preceded by BRAZERO or BRAMINZERO. */
case OP_BRAZERO:
{
const uschar *next = ecode+1;
if (match(eptr, next, offset_top, md, ims, FALSE, eptr)) return TRUE;
do next += (next[1] << 8) + next[2]; while (*next == OP_ALT);
ecode = next + 3;
}
break;
case OP_BRAMINZERO:
{
const uschar *next = ecode+1;
do next += (next[1] << 8) + next[2]; while (*next == OP_ALT);
if (match(eptr, next+3, offset_top, md, ims, FALSE, eptr)) return TRUE;
ecode++;
}
break;
/* End of a group, repeated or non-repeating. If we are at the end of
an assertion "group", stop matching and return TRUE, but record the
current high water mark for use by positive assertions. Do this also
for the "once" (not-backup up) groups. */
case OP_KET:
case OP_KETRMIN:
case OP_KETRMAX:
{
const uschar *prev = ecode - (ecode[1] << 8) - ecode[2];
if (*prev == OP_ASSERT || *prev == OP_ASSERT_NOT ||
*prev == OP_ASSERTBACK || *prev == OP_ASSERTBACK_NOT ||
*prev == OP_ONCE)
{
md->end_match_ptr = eptr; /* For ONCE */
md->end_offset_top = offset_top;
return TRUE;
}
/* In all other cases except a conditional group we have to check the
group number back at the start and if necessary complete handling an
extraction by setting the offsets and bumping the high water mark. */
if (*prev != OP_COND)
{
int number = *prev - OP_BRA;
int offset = number << 1;
DPRINTF(("end bracket %d\n", number));
if (number > 0)
{
if (offset >= md->offset_max) md->offset_overflow = TRUE; else
{
md->offset_vector[offset] =
md->offset_vector[md->offset_end - number];
md->offset_vector[offset+1] = eptr - md->start_subject;
if (offset_top <= offset) offset_top = offset + 2;
}
}
}
/* Reset the value of the ims flags, in case they got changed during
the group. */
ims = original_ims;
DPRINTF(("ims reset to %02lx\n", ims));
/* For a non-repeating ket, just continue at this level. This also
happens for a repeating ket if no characters were matched in the group.
This is the forcible breaking of infinite loops as implemented in Perl
5.005. If there is an options reset, it will get obeyed in the normal
course of events. */
if (*ecode == OP_KET || eptr == eptrb)
{
ecode += 3;
break;
}
/* The repeating kets try the rest of the pattern or restart from the
preceding bracket, in the appropriate order. */
if (*ecode == OP_KETRMIN)
{
if (match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr) ||
match(eptr, prev, offset_top, md, ims, FALSE, eptr)) return TRUE;
}
else /* OP_KETRMAX */
{
if (match(eptr, prev, offset_top, md, ims, FALSE, eptr) ||
match(eptr, ecode+3, offset_top, md, ims, FALSE, eptr)) return TRUE;
}
}
return FALSE;
/* Start of subject unless notbol, or after internal newline if multiline */
case OP_CIRC:
if (md->notbol && eptr == md->start_subject) return FALSE;
if ((ims & PCRE_MULTILINE) != 0)
{
if (eptr != md->start_subject && eptr[-1] != '\n') return FALSE;
ecode++;
break;
}
/* ... else fall through */
/* Start of subject assertion */
case OP_SOD:
if (eptr != md->start_subject) return FALSE;
ecode++;
break;
/* Assert before internal newline if multiline, or before a terminating
newline unless endonly is set, else end of subject unless noteol is set. */
case OP_DOLL:
if ((ims & PCRE_MULTILINE) != 0)
{
if (eptr < md->end_subject) { if (*eptr != '\n') return FALSE; }
else { if (md->noteol) return FALSE; }
ecode++;
break;
}
else
{
if (md->noteol) return FALSE;
if (!md->endonly)
{
if (eptr < md->end_subject - 1 ||
(eptr == md->end_subject - 1 && *eptr != '\n')) return FALSE;
ecode++;
break;
}
}
/* ... else fall through */
/* End of subject assertion (\z) */
case OP_EOD:
if (eptr < md->end_subject) return FALSE;
ecode++;
break;
/* End of subject or ending \n assertion (\Z) */
case OP_EODN:
if (eptr < md->end_subject - 1 ||
(eptr == md->end_subject - 1 && *eptr != '\n')) return FALSE;
ecode++;
break;
/* Word boundary assertions */
case OP_NOT_WORD_BOUNDARY:
case OP_WORD_BOUNDARY:
{
BOOL prev_is_word = (eptr != md->start_subject) &&
((md->ctypes[eptr[-1]] & ctype_word) != 0);
BOOL cur_is_word = (eptr < md->end_subject) &&
((md->ctypes[*eptr] & ctype_word) != 0);
if ((*ecode++ == OP_WORD_BOUNDARY)?
cur_is_word == prev_is_word : cur_is_word != prev_is_word)
return FALSE;
}
break;
/* Match a single character type; inline for speed */
case OP_ANY:
if ((ims & PCRE_DOTALL) == 0 && eptr < md->end_subject && *eptr == '\n')
return FALSE;
if (eptr++ >= md->end_subject) return FALSE;
ecode++;
break;
case OP_NOT_DIGIT:
if (eptr >= md->end_subject ||
(md->ctypes[*eptr++] & ctype_digit) != 0)
return FALSE;
ecode++;
break;
case OP_DIGIT:
if (eptr >= md->end_subject ||
(md->ctypes[*eptr++] & ctype_digit) == 0)
return FALSE;
ecode++;
break;
case OP_NOT_WHITESPACE:
if (eptr >= md->end_subject ||
(md->ctypes[*eptr++] & ctype_space) != 0)
return FALSE;
ecode++;
break;
case OP_WHITESPACE:
if (eptr >= md->end_subject ||
(md->ctypes[*eptr++] & ctype_space) == 0)
return FALSE;
ecode++;
break;
case OP_NOT_WORDCHAR:
if (eptr >= md->end_subject ||
(md->ctypes[*eptr++] & ctype_word) != 0)
return FALSE;
ecode++;
break;
case OP_WORDCHAR:
if (eptr >= md->end_subject ||
(md->ctypes[*eptr++] & ctype_word) == 0)
return FALSE;
ecode++;
break;
/* Match a back reference, possibly repeatedly. Look past the end of the
item to see if there is repeat information following. The code is similar
to that for character classes, but repeated for efficiency. Then obey
similar code to character type repeats - written out again for speed.
However, if the referenced string is the empty string, always treat
it as matched, any number of times (otherwise there could be infinite
loops). */
case OP_REF:
{
int length;
int offset = ecode[1] << 1; /* Doubled reference number */
ecode += 2; /* Advance past the item */
/* If the reference is unset, set the length to be longer than the amount
of subject left; this ensures that every attempt at a match fails. We
can't just fail here, because of the possibility of quantifiers with zero
minima. */
length = (offset >= offset_top || md->offset_vector[offset] < 0)?
md->end_subject - eptr + 1 :
md->offset_vector[offset+1] - md->offset_vector[offset];
/* Set up for repetition, or handle the non-repeated case */
switch (*ecode)
{
case OP_CRSTAR:
case OP_CRMINSTAR:
case OP_CRPLUS:
case OP_CRMINPLUS:
case OP_CRQUERY:
case OP_CRMINQUERY:
c = *ecode++ - OP_CRSTAR;
minimize = (c & 1) != 0;
min = rep_min[c]; /* Pick up values from tables; */
max = rep_max[c]; /* zero for max => infinity */
if (max == 0) max = INT_MAX;
break;
case OP_CRRANGE:
case OP_CRMINRANGE:
minimize = (*ecode == OP_CRMINRANGE);
min = (ecode[1] << 8) + ecode[2];
max = (ecode[3] << 8) + ecode[4];
if (max == 0) max = INT_MAX;
ecode += 5;
break;
default: /* No repeat follows */
if (!match_ref(offset, eptr, length, md, ims)) return FALSE;
eptr += length;
continue; /* With the main loop */
}
/* If the length of the reference is zero, just continue with the
main loop. */
if (length == 0) continue;
/* First, ensure the minimum number of matches are present. We get back
the length of the reference string explicitly rather than passing the
address of eptr, so that eptr can be a register variable. */
for (i = 1; i <= min; i++)
{
if (!match_ref(offset, eptr, length, md, ims)) return FALSE;
eptr += length;
}
/* If min = max, continue at the same level without recursion.
They are not both allowed to be zero. */
if (min == max) continue;
/* If minimizing, keep trying and advancing the pointer */
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
if (i >= max || !match_ref(offset, eptr, length, md, ims))
return FALSE;
eptr += length;
}
/* Control never gets here */
}
/* If maximizing, find the longest string and work backwards */
else
{
const uschar *pp = eptr;
for (i = min; i < max; i++)
{
if (!match_ref(offset, eptr, length, md, ims)) break;
eptr += length;
}
while (eptr >= pp)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
eptr -= length;
}
return FALSE;
}
}
/* Control never gets here */
/* Match a character class, possibly repeatedly. Look past the end of the
item to see if there is repeat information following. Then obey similar
code to character type repeats - written out again for speed. */
case OP_CLASS:
{
const uschar *data = ecode + 1; /* Save for matching */
ecode += 33; /* Advance past the item */
switch (*ecode)
{
case OP_CRSTAR:
case OP_CRMINSTAR:
case OP_CRPLUS:
case OP_CRMINPLUS:
case OP_CRQUERY:
case OP_CRMINQUERY:
c = *ecode++ - OP_CRSTAR;
minimize = (c & 1) != 0;
min = rep_min[c]; /* Pick up values from tables; */
max = rep_max[c]; /* zero for max => infinity */
if (max == 0) max = INT_MAX;
break;
case OP_CRRANGE:
case OP_CRMINRANGE:
minimize = (*ecode == OP_CRMINRANGE);
min = (ecode[1] << 8) + ecode[2];
max = (ecode[3] << 8) + ecode[4];
if (max == 0) max = INT_MAX;
ecode += 5;
break;
default: /* No repeat follows */
min = max = 1;
break;
}
/* First, ensure the minimum number of matches are present. */
for (i = 1; i <= min; i++)
{
if (eptr >= md->end_subject) return FALSE;
c = *eptr++;
if ((data[c/8] & (1 << (c&7))) != 0) continue;
return FALSE;
}
/* If max == min we can continue with the main loop without the
need to recurse. */
if (min == max) continue;
/* If minimizing, keep testing the rest of the expression and advancing
the pointer while it matches the class. */
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
if (i >= max || eptr >= md->end_subject) return FALSE;
c = *eptr++;
if ((data[c/8] & (1 << (c&7))) != 0) continue;
return FALSE;
}
/* Control never gets here */
}
/* If maximizing, find the longest possible run, then work backwards. */
else
{
const uschar *pp = eptr;
for (i = min; i < max; eptr++, i++)
{
if (eptr >= md->end_subject) break;
c = *eptr;
if ((data[c/8] & (1 << (c&7))) != 0) continue;
break;
}
while (eptr >= pp)
if (match(eptr--, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
return FALSE;
}
}
/* Control never gets here */
/* Match a run of characters */
case OP_CHARS:
{
register int length = ecode[1];
ecode += 2;
#ifdef DEBUG /* Sigh. Some compilers never learn. */
if (eptr >= md->end_subject)
printf("matching subject <null> against pattern ");
else
{
printf("matching subject ");
pchars(eptr, length, TRUE, md);
printf(" against pattern ");
}
pchars(ecode, length, FALSE, md);
printf("\n");
#endif
if (length > md->end_subject - eptr) return FALSE;
if ((ims & PCRE_CASELESS) != 0)
{
while (length-- > 0)
if (md->lcc[*ecode++] != md->lcc[*eptr++])
return FALSE;
}
else
{
while (length-- > 0) if (*ecode++ != *eptr++) return FALSE;
}
}
break;
/* Match a single character repeatedly; different opcodes share code. */
case OP_EXACT:
min = max = (ecode[1] << 8) + ecode[2];
ecode += 3;
goto REPEATCHAR;
case OP_UPTO:
case OP_MINUPTO:
min = 0;
max = (ecode[1] << 8) + ecode[2];
minimize = *ecode == OP_MINUPTO;
ecode += 3;
goto REPEATCHAR;
case OP_STAR:
case OP_MINSTAR:
case OP_PLUS:
case OP_MINPLUS:
case OP_QUERY:
case OP_MINQUERY:
c = *ecode++ - OP_STAR;
minimize = (c & 1) != 0;
min = rep_min[c]; /* Pick up values from tables; */
max = rep_max[c]; /* zero for max => infinity */
if (max == 0) max = INT_MAX;
/* Common code for all repeated single-character matches. We can give
up quickly if there are fewer than the minimum number of characters left in
the subject. */
REPEATCHAR:
if (min > md->end_subject - eptr) return FALSE;
c = *ecode++;
/* The code is duplicated for the caseless and caseful cases, for speed,
since matching characters is likely to be quite common. First, ensure the
minimum number of matches are present. If min = max, continue at the same
level without recursing. Otherwise, if minimizing, keep trying the rest of
the expression and advancing one matching character if failing, up to the
maximum. Alternatively, if maximizing, find the maximum number of
characters and work backwards. */
DPRINTF(("matching %c{%d,%d} against subject %.*s\n", c, min, max,
max, eptr));
if ((ims & PCRE_CASELESS) != 0)
{
c = md->lcc[c];
for (i = 1; i <= min; i++)
if (c != md->lcc[*eptr++]) return FALSE;
if (min == max) continue;
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
if (i >= max || eptr >= md->end_subject ||
c != md->lcc[*eptr++])
return FALSE;
}
/* Control never gets here */
}
else
{
const uschar *pp = eptr;
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || c != md->lcc[*eptr]) break;
eptr++;
}
while (eptr >= pp)
if (match(eptr--, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
return FALSE;
}
/* Control never gets here */
}
/* Caseful comparisons */
else
{
for (i = 1; i <= min; i++) if (c != *eptr++) return FALSE;
if (min == max) continue;
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
if (i >= max || eptr >= md->end_subject || c != *eptr++) return FALSE;
}
/* Control never gets here */
}
else
{
const uschar *pp = eptr;
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || c != *eptr) break;
eptr++;
}
while (eptr >= pp)
if (match(eptr--, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
return FALSE;
}
}
/* Control never gets here */
/* Match a negated single character */
case OP_NOT:
if (eptr >= md->end_subject) return FALSE;
ecode++;
if ((ims & PCRE_CASELESS) != 0)
{
if (md->lcc[*ecode++] == md->lcc[*eptr++]) return FALSE;
}
else
{
if (*ecode++ == *eptr++) return FALSE;
}
break;
/* Match a negated single character repeatedly. This is almost a repeat of
the code for a repeated single character, but I haven't found a nice way of
commoning these up that doesn't require a test of the positive/negative
option for each character match. Maybe that wouldn't add very much to the
time taken, but character matching *is* what this is all about... */
case OP_NOTEXACT:
min = max = (ecode[1] << 8) + ecode[2];
ecode += 3;
goto REPEATNOTCHAR;
case OP_NOTUPTO:
case OP_NOTMINUPTO:
min = 0;
max = (ecode[1] << 8) + ecode[2];
minimize = *ecode == OP_NOTMINUPTO;
ecode += 3;
goto REPEATNOTCHAR;
case OP_NOTSTAR:
case OP_NOTMINSTAR:
case OP_NOTPLUS:
case OP_NOTMINPLUS:
case OP_NOTQUERY:
case OP_NOTMINQUERY:
c = *ecode++ - OP_NOTSTAR;
minimize = (c & 1) != 0;
min = rep_min[c]; /* Pick up values from tables; */
max = rep_max[c]; /* zero for max => infinity */
if (max == 0) max = INT_MAX;
/* Common code for all repeated single-character matches. We can give
up quickly if there are fewer than the minimum number of characters left in
the subject. */
REPEATNOTCHAR:
if (min > md->end_subject - eptr) return FALSE;
c = *ecode++;
/* The code is duplicated for the caseless and caseful cases, for speed,
since matching characters is likely to be quite common. First, ensure the
minimum number of matches are present. If min = max, continue at the same
level without recursing. Otherwise, if minimizing, keep trying the rest of
the expression and advancing one matching character if failing, up to the
maximum. Alternatively, if maximizing, find the maximum number of
characters and work backwards. */
DPRINTF(("negative matching %c{%d,%d} against subject %.*s\n", c, min, max,
max, eptr));
if ((ims & PCRE_CASELESS) != 0)
{
c = md->lcc[c];
for (i = 1; i <= min; i++)
if (c == md->lcc[*eptr++]) return FALSE;
if (min == max) continue;
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
if (i >= max || eptr >= md->end_subject ||
c == md->lcc[*eptr++])
return FALSE;
}
/* Control never gets here */
}
else
{
const uschar *pp = eptr;
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || c == md->lcc[*eptr]) break;
eptr++;
}
while (eptr >= pp)
if (match(eptr--, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
return FALSE;
}
/* Control never gets here */
}
/* Caseful comparisons */
else
{
for (i = 1; i <= min; i++) if (c == *eptr++) return FALSE;
if (min == max) continue;
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
if (i >= max || eptr >= md->end_subject || c == *eptr++) return FALSE;
}
/* Control never gets here */
}
else
{
const uschar *pp = eptr;
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || c == *eptr) break;
eptr++;
}
while (eptr >= pp)
if (match(eptr--, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
return FALSE;
}
}
/* Control never gets here */
/* Match a single character type repeatedly; several different opcodes
share code. This is very similar to the code for single characters, but we
repeat it in the interests of efficiency. */
case OP_TYPEEXACT:
min = max = (ecode[1] << 8) + ecode[2];
minimize = TRUE;
ecode += 3;
goto REPEATTYPE;
case OP_TYPEUPTO:
case OP_TYPEMINUPTO:
min = 0;
max = (ecode[1] << 8) + ecode[2];
minimize = *ecode == OP_TYPEMINUPTO;
ecode += 3;
goto REPEATTYPE;
case OP_TYPESTAR:
case OP_TYPEMINSTAR:
case OP_TYPEPLUS:
case OP_TYPEMINPLUS:
case OP_TYPEQUERY:
case OP_TYPEMINQUERY:
c = *ecode++ - OP_TYPESTAR;
minimize = (c & 1) != 0;
min = rep_min[c]; /* Pick up values from tables; */
max = rep_max[c]; /* zero for max => infinity */
if (max == 0) max = INT_MAX;
/* Common code for all repeated single character type matches */
REPEATTYPE:
ctype = *ecode++; /* Code for the character type */
/* First, ensure the minimum number of matches are present. Use inline
code for maximizing the speed, and do the type test once at the start
(i.e. keep it out of the loop). Also test that there are at least the
minimum number of characters before we start. */
if (min > md->end_subject - eptr) return FALSE;
if (min > 0) switch(ctype)
{
case OP_ANY:
if ((ims & PCRE_DOTALL) == 0)
{ for (i = 1; i <= min; i++) if (*eptr++ == '\n') return FALSE; }
else eptr += min;
break;
case OP_NOT_DIGIT:
for (i = 1; i <= min; i++)
if ((md->ctypes[*eptr++] & ctype_digit) != 0) return FALSE;
break;
case OP_DIGIT:
for (i = 1; i <= min; i++)
if ((md->ctypes[*eptr++] & ctype_digit) == 0) return FALSE;
break;
case OP_NOT_WHITESPACE:
for (i = 1; i <= min; i++)
if ((md->ctypes[*eptr++] & ctype_space) != 0) return FALSE;
break;
case OP_WHITESPACE:
for (i = 1; i <= min; i++)
if ((md->ctypes[*eptr++] & ctype_space) == 0) return FALSE;
break;
case OP_NOT_WORDCHAR:
for (i = 1; i <= min; i++)
if ((md->ctypes[*eptr++] & ctype_word) != 0)
return FALSE;
break;
case OP_WORDCHAR:
for (i = 1; i <= min; i++)
if ((md->ctypes[*eptr++] & ctype_word) == 0)
return FALSE;
break;
}
/* If min = max, continue at the same level without recursing */
if (min == max) continue;
/* If minimizing, we have to test the rest of the pattern before each
subsequent match. */
if (minimize)
{
for (i = min;; i++)
{
if (match(eptr, ecode, offset_top, md, ims, FALSE, eptrb)) return TRUE;
if (i >= max || eptr >= md->end_subject) return FALSE;
c = *eptr++;
switch(ctype)
{
case OP_ANY:
if ((ims & PCRE_DOTALL) == 0 && c == '\n') return FALSE;
break;
case OP_NOT_DIGIT:
if ((md->ctypes[c] & ctype_digit) != 0) return FALSE;
break;
case OP_DIGIT:
if ((md->ctypes[c] & ctype_digit) == 0) return FALSE;
break;
case OP_NOT_WHITESPACE:
if ((md->ctypes[c] & ctype_space) != 0) return FALSE;
break;
case OP_WHITESPACE:
if ((md->ctypes[c] & ctype_space) == 0) return FALSE;
break;
case OP_NOT_WORDCHAR:
if ((md->ctypes[c] & ctype_word) != 0) return FALSE;
break;
case OP_WORDCHAR:
if ((md->ctypes[c] & ctype_word) == 0) return FALSE;
break;
}
}
/* Control never gets here */
}
/* If maximizing it is worth using inline code for speed, doing the type
test once at the start (i.e. keep it out of the loop). */
else
{
const uschar *pp = eptr;
switch(ctype)
{
case OP_ANY:
if ((ims & PCRE_DOTALL) == 0)
{
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || *eptr == '\n') break;
eptr++;
}
}
else
{
c = max - min;
if (c > md->end_subject - eptr) c = md->end_subject - eptr;
eptr += c;
}
break;
case OP_NOT_DIGIT:
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || (md->ctypes[*eptr] & ctype_digit) != 0)
break;
eptr++;
}
break;
case OP_DIGIT:
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || (md->ctypes[*eptr] & ctype_digit) == 0)
break;
eptr++;
}
break;
case OP_NOT_WHITESPACE:
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || (md->ctypes[*eptr] & ctype_space) != 0)
break;
eptr++;
}
break;
case OP_WHITESPACE:
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || (md->ctypes[*eptr] & ctype_space) == 0)
break;
eptr++;
}
break;
case OP_NOT_WORDCHAR:
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || (md->ctypes[*eptr] & ctype_word) != 0)
break;
eptr++;
}
break;
case OP_WORDCHAR:
for (i = min; i < max; i++)
{
if (eptr >= md->end_subject || (md->ctypes[*eptr] & ctype_word) == 0)
break;
eptr++;
}
break;
}
while (eptr >= pp)
if (match(eptr--, ecode, offset_top, md, ims, FALSE, eptrb))
return TRUE;
return FALSE;
}
/* Control never gets here */
/* There's been some horrible disaster. */
default:
DPRINTF(("Unknown opcode %d\n", *ecode));
md->errorcode = PCRE_ERROR_UNKNOWN_NODE;
return FALSE;
}
/* Do not stick any code in here without much thought; it is assumed
that "continue" in the code above comes out to here to repeat the main
loop. */
} /* End of main loop */
/* Control never reaches here */
}
/*************************************************
* Execute a Regular Expression *
*************************************************/
/* This function applies a compiled re to a subject string and picks out
portions of the string if it matches. Two elements in the vector are set for
each substring: the offsets to the start and end of the substring.
Arguments:
external_re points to the compiled expression
external_extra points to "hints" from pcre_study() or is NULL
subject points to the subject string
length length of subject string (may contain binary zeros)
start_offset where to start in the subject string
options option bits
offsets points to a vector of ints to be filled in with offsets
offsetcount the number of elements in the vector
Returns: > 0 => success; value is the number of elements filled in
= 0 => success, but offsets is not big enough
-1 => failed to match
< -1 => some kind of unexpected problem
*/
int
pcre_exec(const pcre *external_re, const pcre_extra *external_extra,
const char *subject, int length, int start_offset, int options, int *offsets,
int offsetcount)
{
int resetcount, ocount;
int first_char = -1;
int req_char = -1;
int req_char2 = -1;
unsigned long int ims = 0;
match_data match_block;
const uschar *start_bits = NULL;
const uschar *start_match = (const uschar *)subject + start_offset;
const uschar *end_subject;
const uschar *req_char_ptr = start_match - 1;
const real_pcre *re = (const real_pcre *)external_re;
const real_pcre_extra *extra = (const real_pcre_extra *)external_extra;
BOOL using_temporary_offsets = FALSE;
BOOL anchored = ((re->options | options) & PCRE_ANCHORED) != 0;
BOOL startline = (re->options & PCRE_STARTLINE) != 0;
if ((options & ~PUBLIC_EXEC_OPTIONS) != 0) return PCRE_ERROR_BADOPTION;
if (re == NULL || subject == NULL ||
(offsets == NULL && offsetcount > 0)) return PCRE_ERROR_NULL;
if (re->magic_number != MAGIC_NUMBER) return PCRE_ERROR_BADMAGIC;
match_block.start_subject = (const uschar *)subject;
match_block.end_subject = match_block.start_subject + length;
end_subject = match_block.end_subject;
match_block.endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
match_block.notbol = (options & PCRE_NOTBOL) != 0;
match_block.noteol = (options & PCRE_NOTEOL) != 0;
match_block.notempty = (options & PCRE_NOTEMPTY) != 0;
match_block.errorcode = PCRE_ERROR_NOMATCH; /* Default error */
match_block.lcc = re->tables + lcc_offset;
match_block.ctypes = re->tables + ctypes_offset;
/* The ims options can vary during the matching as a result of the presence
of (?ims) items in the pattern. They are kept in a local variable so that
restoring at the exit of a group is easy. */
ims = re->options & (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL);
/* If the expression has got more back references than the offsets supplied can
hold, we get a temporary bit of working store to use during the matching.
Otherwise, we can use the vector supplied, rounding down its size to a multiple
of 3. */
ocount = offsetcount - (offsetcount % 3);
if (re->top_backref > 0 && re->top_backref >= ocount/3)
{
ocount = re->top_backref * 3 + 3;
match_block.offset_vector = (int *)(pcre_malloc)(ocount * sizeof(int));
if (match_block.offset_vector == NULL) return PCRE_ERROR_NOMEMORY;
using_temporary_offsets = TRUE;
DPRINTF(("Got memory to hold back references\n"));
}
else match_block.offset_vector = offsets;
match_block.offset_end = ocount;
match_block.offset_max = (2*ocount)/3;
match_block.offset_overflow = FALSE;
/* Compute the minimum number of offsets that we need to reset each time. Doing
this makes a huge difference to execution time when there aren't many brackets
in the pattern. */
resetcount = 2 + re->top_bracket * 2;
if (resetcount > offsetcount) resetcount = ocount;
/* Reset the working variable associated with each extraction. These should
never be used unless previously set, but they get saved and restored, and so we
initialize them to avoid reading uninitialized locations. */
if (match_block.offset_vector != NULL)
{
register int *iptr = match_block.offset_vector + ocount;
register int *iend = iptr - resetcount/2 + 1;
while (--iptr >= iend) *iptr = -1;
}
/* Set up the first character to match, if available. The first_char value is
never set for an anchored regular expression, but the anchoring may be forced
at run time, so we have to test for anchoring. The first char may be unset for
an unanchored pattern, of course. If there's no first char and the pattern was
studied, there may be a bitmap of possible first characters. */
if (!anchored)
{
if ((re->options & PCRE_FIRSTSET) != 0)
{
first_char = re->first_char;
if ((ims & PCRE_CASELESS) != 0) first_char = match_block.lcc[first_char];
}
else
if (!startline && extra != NULL &&
(extra->options & PCRE_STUDY_MAPPED) != 0)
start_bits = extra->start_bits;
}
/* For anchored or unanchored matches, there may be a "last known required
character" set. If the PCRE_CASELESS is set, implying that the match starts
caselessly, or if there are any changes of this flag within the regex, set up
both cases of the character. Otherwise set the two values the same, which will
avoid duplicate testing (which takes significant time). This covers the vast
majority of cases. It will be suboptimal when the case flag changes in a regex
and the required character in fact is caseful. */
if ((re->options & PCRE_REQCHSET) != 0)
{
req_char = re->req_char;
req_char2 = ((re->options & (PCRE_CASELESS | PCRE_ICHANGED)) != 0)?
(re->tables + fcc_offset)[req_char] : req_char;
}
/* Loop for handling unanchored repeated matching attempts; for anchored regexs
the loop runs just once. */
do
{
int rc;
register int *iptr = match_block.offset_vector;
register int *iend = iptr + resetcount;
/* Reset the maximum number of extractions we might see. */
while (iptr < iend) *iptr++ = -1;
/* Advance to a unique first char if possible */
if (first_char >= 0)
{
if ((ims & PCRE_CASELESS) != 0)
while (start_match < end_subject &&
match_block.lcc[*start_match] != first_char)
start_match++;
else
while (start_match < end_subject && *start_match != first_char)
start_match++;
}
/* Or to just after \n for a multiline match if possible */
else if (startline)
{
if (start_match > match_block.start_subject + start_offset)
{
while (start_match < end_subject && start_match[-1] != '\n')
start_match++;
}
}
/* Or to a non-unique first char after study */
else if (start_bits != NULL)
{
while (start_match < end_subject)
{
register int c = *start_match;
if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++; else break;
}
}
#ifdef DEBUG /* Sigh. Some compilers never learn. */
printf(">>>> Match against: ");
pchars(start_match, end_subject - start_match, TRUE, &match_block);
printf("\n");
#endif
/* If req_char is set, we know that that character must appear in the subject
for the match to succeed. If the first character is set, req_char must be
later in the subject; otherwise the test starts at the match point. This
optimization can save a huge amount of backtracking in patterns with nested
unlimited repeats that aren't going to match. We don't know what the state of
case matching may be when this character is hit, so test for it in both its
cases if necessary. However, the different cased versions will not be set up
unless PCRE_CASELESS was given or the casing state changes within the regex.
Writing separate code makes it go faster, as does using an autoincrement and
backing off on a match. */
if (req_char >= 0)
{
register const uschar *p = start_match + ((first_char >= 0)? 1 : 0);
/* We don't need to repeat the search if we haven't yet reached the
place we found it at last time. */
if (p > req_char_ptr)
{
/* Do a single test if no case difference is set up */
if (req_char == req_char2)
{
while (p < end_subject)
{
if (*p++ == req_char) { p--; break; }
}
}
/* Otherwise test for either case */
else
{
while (p < end_subject)
{
register int pp = *p++;
if (pp == req_char || pp == req_char2) { p--; break; }
}
}
/* If we can't find the required character, break the matching loop */
if (p >= end_subject) break;
/* If we have found the required character, save the point where we
found it, so that we don't search again next time round the loop if
the start hasn't passed this character yet. */
req_char_ptr = p;
}
}
/* When a match occurs, substrings will be set for all internal extractions;
we just need to set up the whole thing as substring 0 before returning. If
there were too many extractions, set the return code to zero. In the case
where we had to get some local store to hold offsets for backreferences, copy
those back references that we can. In this case there need not be overflow
if certain parts of the pattern were not used. */
match_block.start_match = start_match;
if (!match(start_match, re->code, 2, &match_block, ims, FALSE, start_match))
continue;
/* Copy the offset information from temporary store if necessary */
if (using_temporary_offsets)
{
if (offsetcount >= 4)
{
memcpy(offsets + 2, match_block.offset_vector + 2,
(offsetcount - 2) * sizeof(int));
DPRINTF(("Copied offsets from temporary memory\n"));
}
if (match_block.end_offset_top > offsetcount)
match_block.offset_overflow = TRUE;
DPRINTF(("Freeing temporary memory\n"));
(pcre_free)(match_block.offset_vector);
}
rc = match_block.offset_overflow? 0 : match_block.end_offset_top/2;
if (match_block.offset_end < 2) rc = 0; else
{
offsets[0] = start_match - match_block.start_subject;
offsets[1] = match_block.end_match_ptr - match_block.start_subject;
}
DPRINTF((">>>> returning %d\n", rc));
return rc;
}
/* This "while" is the end of the "do" above */
while (!anchored &&
match_block.errorcode == PCRE_ERROR_NOMATCH &&
start_match++ < end_subject);
if (using_temporary_offsets)
{
DPRINTF(("Freeing temporary memory\n"));
(pcre_free)(match_block.offset_vector);
}
DPRINTF((">>>> returning %d\n", match_block.errorcode));
return match_block.errorcode;
}
/* End of pcre.c */
|