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
|
/* dfa.c - deterministic extended regexp routines for GNU
Copyright 1988, 1998, 2000, 2002, 2004, 2005, 2007, 2009
Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3, or (at your option)
any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc.,
51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA */
/* Written June, 1988 by Mike Haertel
Modified July, 1988 by Arthur David Olson to assist BMG speedups */
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif
#include <assert.h>
#include <ctype.h>
#include <stdio.h>
#ifndef VMS
#include <sys/types.h>
#else
#include <stddef.h>
#endif
#ifdef STDC_HEADERS
#include <stdlib.h>
#else
extern char *calloc(), *malloc(), *realloc();
extern void free();
#endif
#if defined(HAVE_STRING_H) || defined(STDC_HEADERS)
#include <string.h>
#else
#include <strings.h>
#endif
#if HAVE_SETLOCALE
# include <locale.h>
#endif
#ifndef DEBUG /* use the same approach as regex.c */
#undef assert
#define assert(e)
#endif /* DEBUG */
#ifndef isgraph
#define isgraph(C) (isprint(C) && !isspace(C))
#endif
#if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII))
#define ISALPHA(C) isalpha(C)
#define ISUPPER(C) isupper(C)
#define ISLOWER(C) islower(C)
#define ISDIGIT(C) isdigit(C)
#define ISXDIGIT(C) isxdigit(C)
#define ISSPACE(C) isspace(C)
#define ISPUNCT(C) ispunct(C)
#define ISALNUM(C) isalnum(C)
#define ISPRINT(C) isprint(C)
#define ISGRAPH(C) isgraph(C)
#define ISCNTRL(C) iscntrl(C)
#else
#define ISALPHA(C) (isascii(C) && isalpha(C))
#define ISUPPER(C) (isascii(C) && isupper(C))
#define ISLOWER(C) (isascii(C) && islower(C))
#define ISDIGIT(C) (isascii(C) && isdigit(C))
#define ISXDIGIT(C) (isascii(C) && isxdigit(C))
#define ISSPACE(C) (isascii(C) && isspace(C))
#define ISPUNCT(C) (isascii(C) && ispunct(C))
#define ISALNUM(C) (isascii(C) && isalnum(C))
#define ISPRINT(C) (isascii(C) && isprint(C))
#define ISGRAPH(C) (isascii(C) && isgraph(C))
#define ISCNTRL(C) (isascii(C) && iscntrl(C))
#endif
/* ISASCIIDIGIT differs from ISDIGIT, as follows:
- Its arg may be any int or unsigned int; it need not be an unsigned char.
- It's guaranteed to evaluate its argument exactly once.
- It's typically faster.
Posix 1003.2-1992 section 2.5.2.1 page 50 lines 1556-1558 says that
only '0' through '9' are digits. Prefer ISASCIIDIGIT to ISDIGIT unless
it's important to use the locale's definition of `digit' even when the
host does not conform to Posix. */
#define ISASCIIDIGIT(c) ((unsigned) (c) - '0' <= 9)
/* gettext.h ensures that we don't use gettext if ENABLE_NLS is not defined */
#include "gettext.h"
#define _(str) gettext (str)
#ifndef NO_MBSUPPORT
#include "mbsupport.h" /* defines MBS_SUPPORT if appropriate */
#endif
#ifdef MBS_SUPPORT
/* We can handle multibyte strings. */
# include <wchar.h>
# include <wctype.h>
#endif
#include "regex.h"
#include "dfa.h"
#include "hard-locale.h"
/* HPUX, define those as macros in sys/param.h */
#ifdef setbit
# undef setbit
#endif
#ifdef clrbit
# undef clrbit
#endif
static void dfamust PARAMS ((struct dfa *dfa));
static ptr_t xcalloc PARAMS ((size_t n, size_t s));
static ptr_t xmalloc PARAMS ((size_t n));
static ptr_t xrealloc PARAMS ((ptr_t p, size_t n));
#ifdef DEBUG
static void prtok PARAMS ((token t));
#endif
static int tstbit PARAMS ((unsigned b, charclass c));
static void setbit PARAMS ((unsigned b, charclass c));
static void clrbit PARAMS ((unsigned b, charclass c));
static void copyset PARAMS ((charclass src, charclass dst));
static void zeroset PARAMS ((charclass s));
static void notset PARAMS ((charclass s));
static int equal PARAMS ((charclass s1, charclass s2));
static int charclass_index PARAMS ((charclass s));
static int looking_at PARAMS ((const char *s));
static token lex PARAMS ((void));
static void addtok PARAMS ((token t));
static void atom PARAMS ((void));
static int nsubtoks PARAMS ((int tindex));
static void copytoks PARAMS ((int tindex, int ntokens));
static void closure PARAMS ((void));
static void branch PARAMS ((void));
static void regexp PARAMS ((int toplevel));
static void copy PARAMS ((position_set const *src, position_set *dst));
static void insert PARAMS ((position p, position_set *s));
static void merge PARAMS ((position_set const *s1, position_set const *s2, position_set *m));
static void delete PARAMS ((position p, position_set *s));
static int state_index PARAMS ((struct dfa *d, position_set const *s,
int newline, int letter));
static void build_state PARAMS ((int s, struct dfa *d));
static void build_state_zero PARAMS ((struct dfa *d));
static char *icatalloc PARAMS ((char *old, char *new));
static char *icpyalloc PARAMS ((char *string));
static char *istrstr PARAMS ((char *lookin, char *lookfor));
static void ifree PARAMS ((char *cp));
static void freelist PARAMS ((char **cpp));
static char **enlist PARAMS ((char **cpp, char *new, size_t len));
static char **comsubs PARAMS ((char *left, char *right));
static char **addlists PARAMS ((char **old, char **new));
static char **inboth PARAMS ((char **left, char **right));
static ptr_t
xcalloc (size_t n, size_t s)
{
ptr_t r = calloc(n, s);
if (!r)
dfaerror(_("Memory exhausted"));
return r;
}
static ptr_t
xmalloc (size_t n)
{
ptr_t r = malloc(n);
assert(n != 0);
if (!r)
dfaerror(_("Memory exhausted"));
return r;
}
static ptr_t
xrealloc (ptr_t p, size_t n)
{
ptr_t r = realloc(p, n);
assert(n != 0);
if (!r)
dfaerror(_("Memory exhausted"));
return r;
}
#define CALLOC(p, t, n) ((p) = (t *) xcalloc((size_t)(n), sizeof (t)))
#define MALLOC(p, t, n) ((p) = (t *) xmalloc((n) * sizeof (t)))
#define REALLOC(p, t, n) ((p) = (t *) xrealloc((ptr_t) (p), (n) * sizeof (t)))
/* Reallocate an array of type t if nalloc is too small for index. */
#define REALLOC_IF_NECESSARY(p, t, nalloc, index) \
if ((index) >= (nalloc)) \
{ \
do \
(nalloc) *= 2; \
while ((index) >= (nalloc)); \
REALLOC(p, t, nalloc); \
}
#ifdef DEBUG
static void
prtok (token t)
{
char const *s;
if (t < 0)
fprintf(stderr, "END");
else if (t < NOTCHAR)
fprintf(stderr, "%c", t);
else
{
switch (t)
{
case EMPTY: s = "EMPTY"; break;
case BACKREF: s = "BACKREF"; break;
case BEGLINE: s = "BEGLINE"; break;
case ENDLINE: s = "ENDLINE"; break;
case BEGWORD: s = "BEGWORD"; break;
case ENDWORD: s = "ENDWORD"; break;
case LIMWORD: s = "LIMWORD"; break;
case NOTLIMWORD: s = "NOTLIMWORD"; break;
case QMARK: s = "QMARK"; break;
case STAR: s = "STAR"; break;
case PLUS: s = "PLUS"; break;
case CAT: s = "CAT"; break;
case OR: s = "OR"; break;
case ORTOP: s = "ORTOP"; break;
case LPAREN: s = "LPAREN"; break;
case RPAREN: s = "RPAREN"; break;
case CRANGE: s = "CRANGE"; break;
#ifdef MBS_SUPPORT
case ANYCHAR: s = "ANYCHAR"; break;
case MBCSET: s = "MBCSET"; break;
#endif /* MBS_SUPPORT */
default: s = "CSET"; break;
}
fprintf(stderr, "%s", s);
}
}
#endif /* DEBUG */
/* Stuff pertaining to charclasses. */
static int
tstbit (unsigned b, charclass c)
{
return c[b / INTBITS] & 1 << b % INTBITS;
}
static void
setbit (unsigned b, charclass c)
{
c[b / INTBITS] |= 1 << b % INTBITS;
}
static void
clrbit (unsigned b, charclass c)
{
c[b / INTBITS] &= ~(1 << b % INTBITS);
}
static void
copyset (charclass src, charclass dst)
{
memcpy (dst, src, sizeof (charclass));
}
static void
zeroset (charclass s)
{
memset (s, 0, sizeof (charclass));
}
static void
notset (charclass s)
{
int i;
for (i = 0; i < CHARCLASS_INTS; ++i)
s[i] = ~s[i];
}
static int
equal (charclass s1, charclass s2)
{
return memcmp (s1, s2, sizeof (charclass)) == 0;
}
/* A pointer to the current dfa is kept here during parsing. */
static struct dfa *dfa;
/* Find the index of charclass s in dfa->charclasses, or allocate a new charclass. */
static int
charclass_index (charclass s)
{
int i;
for (i = 0; i < dfa->cindex; ++i)
if (equal(s, dfa->charclasses[i]))
return i;
REALLOC_IF_NECESSARY(dfa->charclasses, charclass, dfa->calloc, dfa->cindex);
++dfa->cindex;
copyset(s, dfa->charclasses[i]);
return i;
}
/* Syntax bits controlling the behavior of the lexical analyzer. */
static reg_syntax_t syntax_bits, syntax_bits_set;
/* Flag for case-folding letters into sets. */
static int case_fold;
/* End-of-line byte in data. */
static unsigned char eolbyte;
/* Entry point to set syntax options. */
void
dfasyntax (reg_syntax_t bits, int fold, unsigned char eol)
{
syntax_bits_set = 1;
syntax_bits = bits;
case_fold = fold;
eolbyte = eol;
}
/* Like setbit, but if case is folded, set both cases of a letter. */
static void
setbit_case_fold (unsigned b, charclass c)
{
setbit (b, c);
if (case_fold)
{
if (ISUPPER (b))
setbit (tolower (b), c);
else if (ISLOWER (b))
setbit (toupper (b), c);
}
}
/* Lexical analyzer. All the dross that deals with the obnoxious
GNU Regex syntax bits is located here. The poor, suffering
reader is referred to the GNU Regex documentation for the
meaning of the @#%!@#%^!@ syntax bits. */
static char const *lexptr; /* Pointer to next input character. */
static int lexleft; /* Number of characters remaining. */
static token lasttok; /* Previous token returned; initially END. */
static int laststart; /* True if we're separated from beginning or (, |
only by zero-width characters. */
static int parens; /* Count of outstanding left parens. */
static int minrep, maxrep; /* Repeat counts for {m,n}. */
static int hard_LC_COLLATE; /* Nonzero if LC_COLLATE is hard. */
#ifdef MBS_SUPPORT
/* These variables are used only if (MB_CUR_MAX > 1). */
static mbstate_t mbs; /* Mbstate for mbrlen(). */
static int cur_mb_len; /* Byte length of the current scanning
multibyte character. */
static int cur_mb_index; /* Byte index of the current scanning multibyte
character.
single byte character : cur_mb_index = 0
multibyte character
1st byte : cur_mb_index = 1
2nd byte : cur_mb_index = 2
...
nth byte : cur_mb_index = n */
static unsigned char *mblen_buf;/* Correspond to the input buffer in dfaexec().
Each element store the amount of remain
byte of corresponding multibyte character
in the input string. A element's value
is 0 if corresponding character is a
single byte chracter.
e.g. input : 'a', <mb(0)>, <mb(1)>, <mb(2)>
mblen_buf : 0, 3, 2, 1
*/
static wchar_t *inputwcs; /* Wide character representation of input
string in dfaexec().
The length of this array is same as
the length of input string(char array).
inputstring[i] is a single-byte char,
or 1st byte of a multibyte char.
And inputwcs[i] is the codepoint. */
static unsigned char const *buf_begin; /* reference to begin in dfaexec(). */
static unsigned char const *buf_end; /* reference to end in dfaexec(). */
#endif /* MBS_SUPPORT */
#ifdef MBS_SUPPORT
/* This function update cur_mb_len, and cur_mb_index.
p points current lexptr, len is the remaining buffer length. */
static void
update_mb_len_index (unsigned char const *p, int len)
{
/* If last character is a part of a multibyte character,
we update cur_mb_index. */
if (cur_mb_index)
cur_mb_index = (cur_mb_index >= cur_mb_len)? 0
: cur_mb_index + 1;
/* If last character is a single byte character, or the
last portion of a multibyte character, we check whether
next character is a multibyte character or not. */
if (! cur_mb_index)
{
cur_mb_len = mbrlen(p, len, &mbs);
if (cur_mb_len > 1)
/* It is a multibyte character.
cur_mb_len was already set by mbrlen(). */
cur_mb_index = 1;
else if (cur_mb_len < 1)
/* Invalid sequence. We treat it as a single byte character.
cur_mb_index is aleady 0. */
cur_mb_len = 1;
/* Otherwise, cur_mb_len == 1, it is a single byte character.
cur_mb_index is aleady 0. */
}
}
#endif /* MBS_SUPPORT */
#ifdef MBS_SUPPORT
/* Note that characters become unsigned here. */
# define FETCH(c, eoferr) \
{ \
if (! lexleft) \
{ \
if (eoferr != 0) \
dfaerror (eoferr); \
else \
return lasttok = END; \
} \
if (MB_CUR_MAX > 1) \
update_mb_len_index(lexptr, lexleft); \
(c) = (unsigned char) *lexptr++; \
--lexleft; \
}
/* This function fetch a wide character, and update cur_mb_len,
used only if the current locale is a multibyte environment. */
static wint_t
fetch_wc (char const *eoferr)
{
wchar_t wc;
if (! lexleft)
{
if (eoferr != 0)
dfaerror (eoferr);
else
return WEOF;
}
cur_mb_len = mbrtowc(&wc, lexptr, lexleft, &mbs);
if (cur_mb_len <= 0)
{
cur_mb_len = 1;
wc = *lexptr;
}
lexptr += cur_mb_len;
lexleft -= cur_mb_len;
return wc;
}
#else
/* Note that characters become unsigned here. */
# define FETCH(c, eoferr) \
{ \
if (! lexleft) \
{ \
if (eoferr != 0) \
dfaerror (eoferr); \
else \
return lasttok = END; \
} \
(c) = (unsigned char) *lexptr++; \
--lexleft; \
}
#endif /* MBS_SUPPORT */
#ifdef MBS_SUPPORT
/* Multibyte character handling sub-routine for lex.
This function parse a bracket expression and build a struct
mb_char_classes. */
static void
parse_bracket_exp_mb ()
{
wint_t wc, wc1, wc2;
/* Work area to build a mb_char_classes. */
struct mb_char_classes *work_mbc;
int chars_al, range_sts_al, range_ends_al, ch_classes_al,
equivs_al, coll_elems_al;
REALLOC_IF_NECESSARY(dfa->mbcsets, struct mb_char_classes,
dfa->mbcsets_alloc, dfa->nmbcsets + 1);
/* dfa->multibyte_prop[] hold the index of dfa->mbcsets.
We will update dfa->multibyte_prop[] in addtok(), because we can't
decide the index in dfa->tokens[]. */
/* Initialize work are */
work_mbc = &(dfa->mbcsets[dfa->nmbcsets++]);
chars_al = 1;
range_sts_al = range_ends_al = 0;
ch_classes_al = equivs_al = coll_elems_al = 0;
MALLOC(work_mbc->chars, wchar_t, chars_al);
work_mbc->nchars = work_mbc->nranges = work_mbc->nch_classes = 0;
work_mbc->nequivs = work_mbc->ncoll_elems = 0;
work_mbc->ch_classes = NULL;
work_mbc->range_sts = work_mbc->range_ends = NULL;
work_mbc->equivs = work_mbc->coll_elems = NULL;
wc = fetch_wc(_("Unbalanced ["));
if (wc == L'^')
{
wc = fetch_wc(_("Unbalanced ["));
work_mbc->invert = 1;
}
else
work_mbc->invert = 0;
do
{
wc1 = WEOF; /* mark wc1 is not initialized". */
/* Note that if we're looking at some other [:...:] construct,
we just treat it as a bunch of ordinary characters. We can do
this because we assume regex has checked for syntax errors before
dfa is ever called. */
if (wc == L'[' && (syntax_bits & RE_CHAR_CLASSES))
{
#define BRACKET_BUFFER_SIZE 128
char str[BRACKET_BUFFER_SIZE];
wc1 = wc;
wc = fetch_wc(_("Unbalanced ["));
/* If pattern contains `[[:', `[[.', or `[[='. */
if (cur_mb_len == 1 && (wc == L':' || wc == L'.' || wc == L'='))
{
unsigned char c;
unsigned char delim = (unsigned char)wc;
int len = 0;
for (;;)
{
if (! lexleft)
dfaerror (_("Unbalanced ["));
c = (unsigned char) *lexptr++;
--lexleft;
if ((c == delim && *lexptr == ']') || lexleft == 0)
break;
if (len < BRACKET_BUFFER_SIZE)
str[len++] = c;
else
/* This is in any case an invalid class name. */
str[0] = '\0';
}
str[len] = '\0';
if (lexleft == 0)
{
REALLOC_IF_NECESSARY(work_mbc->chars, wchar_t, chars_al,
work_mbc->nchars + 2);
work_mbc->chars[work_mbc->nchars++] = L'[';
work_mbc->chars[work_mbc->nchars++] = delim;
break;
}
if (--lexleft, *lexptr++ != ']')
dfaerror (_("Unbalanced ["));
if (delim == ':')
/* build character class. */
{
wctype_t wt;
/* Query the character class as wctype_t. */
if (case_fold && (strcmp(str, "upper") == 0 || strcmp(str, "lower") == 0))
strcpy(str, "alpha");
wt = wctype (str);
if (ch_classes_al == 0)
MALLOC(work_mbc->ch_classes, wctype_t, ++ch_classes_al);
REALLOC_IF_NECESSARY(work_mbc->ch_classes, wctype_t,
ch_classes_al,
work_mbc->nch_classes + 1);
work_mbc->ch_classes[work_mbc->nch_classes++] = wt;
}
else if (delim == '=' || delim == '.')
{
char *elem;
MALLOC(elem, char, len + 1);
strncpy(elem, str, len + 1);
if (delim == '=')
/* build equivalent class. */
{
if (equivs_al == 0)
MALLOC(work_mbc->equivs, char*, ++equivs_al);
REALLOC_IF_NECESSARY(work_mbc->equivs, char*,
equivs_al,
work_mbc->nequivs + 1);
work_mbc->equivs[work_mbc->nequivs++] = elem;
}
if (delim == '.')
/* build collating element. */
{
if (coll_elems_al == 0)
MALLOC(work_mbc->coll_elems, char*, ++coll_elems_al);
REALLOC_IF_NECESSARY(work_mbc->coll_elems, char*,
coll_elems_al,
work_mbc->ncoll_elems + 1);
work_mbc->coll_elems[work_mbc->ncoll_elems++] = elem;
}
}
wc1 = wc = WEOF;
}
else
/* We treat '[' as a normal character here. */
{
wc2 = wc1; wc1 = wc; wc = wc2; /* swap */
}
}
else
{
if (wc == L'\\' && (syntax_bits & RE_BACKSLASH_ESCAPE_IN_LISTS))
wc = fetch_wc(("Unbalanced ["));
}
if (wc1 == WEOF)
wc1 = fetch_wc(_("Unbalanced ["));
if (wc1 == L'-')
/* build range characters. */
{
wc2 = fetch_wc(_("Unbalanced ["));
if (wc2 == L']')
{
/* In the case [x-], the - is an ordinary hyphen,
which is left in c1, the lookahead character. */
lexptr -= cur_mb_len;
lexleft += cur_mb_len;
wc2 = wc;
}
else
{
if (wc2 == L'\\'
&& (syntax_bits & RE_BACKSLASH_ESCAPE_IN_LISTS))
wc2 = fetch_wc(_("Unbalanced ["));
wc1 = fetch_wc(_("Unbalanced ["));
}
if (range_sts_al == 0)
{
MALLOC(work_mbc->range_sts, wchar_t, ++range_sts_al);
MALLOC(work_mbc->range_ends, wchar_t, ++range_ends_al);
}
REALLOC_IF_NECESSARY(work_mbc->range_sts, wchar_t,
range_sts_al, work_mbc->nranges + 1);
work_mbc->range_sts[work_mbc->nranges] = (wchar_t)wc;
REALLOC_IF_NECESSARY(work_mbc->range_ends, wchar_t,
range_ends_al, work_mbc->nranges + 1);
work_mbc->range_ends[work_mbc->nranges++] = (wchar_t)wc2;
if (case_fold && (iswlower((wint_t)wc) || iswupper((wint_t)wc))
&& (iswlower((wint_t)wc2) || iswupper((wint_t)wc2)))
{
wint_t altcase;
altcase = wc;
if (iswlower((wint_t)wc))
altcase = towupper((wint_t)wc);
else
altcase = towlower((wint_t)wc);
REALLOC_IF_NECESSARY(work_mbc->range_sts, wchar_t,
range_sts_al, work_mbc->nranges + 1);
work_mbc->range_sts[work_mbc->nranges] = (wchar_t)altcase;
altcase = wc2;
if (iswlower((wint_t)wc2))
altcase = towupper((wint_t)wc2);
else
altcase = towlower((wint_t)wc2);
REALLOC_IF_NECESSARY(work_mbc->range_ends, wchar_t,
range_ends_al, work_mbc->nranges + 1);
work_mbc->range_ends[work_mbc->nranges++] = (wchar_t)altcase;
}
}
else if (wc != WEOF)
/* build normal characters. */
{
REALLOC_IF_NECESSARY(work_mbc->chars, wchar_t, chars_al,
work_mbc->nchars + 1);
work_mbc->chars[work_mbc->nchars++] = (wchar_t)wc;
if (case_fold && (iswlower(wc) || iswupper(wc)))
{
REALLOC_IF_NECESSARY(work_mbc->chars, wchar_t, chars_al,
work_mbc->nchars + 1);
work_mbc->chars[work_mbc->nchars++] =
(wchar_t) (iswlower(wc) ? towupper(wc) : towlower(wc));
}
}
}
while ((wc = wc1) != L']');
}
#endif /* MBS_SUPPORT */
#ifdef __STDC__
#define FUNC(F, P) static int F(int c) { return P(c); }
#else
#define FUNC(F, P) static int F(c) int c; { return P(c); }
#endif
FUNC(is_alpha, ISALPHA)
FUNC(is_upper, ISUPPER)
FUNC(is_lower, ISLOWER)
FUNC(is_digit, ISDIGIT)
FUNC(is_xdigit, ISXDIGIT)
FUNC(is_space, ISSPACE)
FUNC(is_punct, ISPUNCT)
FUNC(is_alnum, ISALNUM)
FUNC(is_print, ISPRINT)
FUNC(is_graph, ISGRAPH)
FUNC(is_cntrl, ISCNTRL)
static int
is_blank (int c)
{
return (c == ' ' || c == '\t');
}
/* The following list maps the names of the Posix named character classes
to predicate functions that determine whether a given character is in
the class. The leading [ has already been eaten by the lexical analyzer. */
static struct {
const char *name;
int (*pred) PARAMS ((int));
} const prednames[] = {
{ ":alpha:]", is_alpha },
{ ":upper:]", is_upper },
{ ":lower:]", is_lower },
{ ":digit:]", is_digit },
{ ":xdigit:]", is_xdigit },
{ ":space:]", is_space },
{ ":punct:]", is_punct },
{ ":alnum:]", is_alnum },
{ ":print:]", is_print },
{ ":graph:]", is_graph },
{ ":cntrl:]", is_cntrl },
{ ":blank:]", is_blank },
{ 0, 0 }
};
/* Return non-zero if C is a `word-constituent' byte; zero otherwise. */
#define IS_WORD_CONSTITUENT(C) (ISALNUM(C) || (C) == '_')
static int
looking_at (char const *s)
{
size_t len;
len = strlen(s);
if (lexleft < len)
return 0;
return strncmp(s, lexptr, len) == 0;
}
static token
lex (void)
{
unsigned c, c1, c2;
int backslash = 0, invert;
charclass ccl;
int i;
/* Basic plan: We fetch a character. If it's a backslash,
we set the backslash flag and go through the loop again.
On the plus side, this avoids having a duplicate of the
main switch inside the backslash case. On the minus side,
it means that just about every case begins with
"if (backslash) ...". */
for (i = 0; i < 2; ++i)
{
FETCH(c, 0);
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1 && cur_mb_index)
/* If this is a part of a multi-byte character, we must treat
this byte data as a normal character.
e.g. In case of SJIS encoding, some character contains '\',
but they must not be backslash. */
goto normal_char;
#endif /* MBS_SUPPORT */
switch (c)
{
case '\\':
if (backslash)
goto normal_char;
if (lexleft == 0)
dfaerror(_("Unfinished \\ escape"));
backslash = 1;
break;
case '^':
if (backslash)
goto normal_char;
if (syntax_bits & RE_CONTEXT_INDEP_ANCHORS
|| lasttok == END
|| lasttok == LPAREN
|| lasttok == OR)
return lasttok = BEGLINE;
goto normal_char;
case '$':
if (backslash)
goto normal_char;
if (syntax_bits & RE_CONTEXT_INDEP_ANCHORS
|| lexleft == 0
|| (syntax_bits & RE_NO_BK_PARENS
? lexleft > 0 && *lexptr == ')'
: lexleft > 1 && lexptr[0] == '\\' && lexptr[1] == ')')
|| (syntax_bits & RE_NO_BK_VBAR
? lexleft > 0 && *lexptr == '|'
: lexleft > 1 && lexptr[0] == '\\' && lexptr[1] == '|')
|| ((syntax_bits & RE_NEWLINE_ALT)
&& lexleft > 0 && *lexptr == '\n'))
return lasttok = ENDLINE;
goto normal_char;
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
if (backslash && !(syntax_bits & RE_NO_BK_REFS))
{
laststart = 0;
return lasttok = BACKREF;
}
goto normal_char;
case '`':
if (backslash && !(syntax_bits & RE_NO_GNU_OPS))
return lasttok = BEGLINE; /* FIXME: should be beginning of string */
goto normal_char;
case '\'':
if (backslash && !(syntax_bits & RE_NO_GNU_OPS))
return lasttok = ENDLINE; /* FIXME: should be end of string */
goto normal_char;
case '<':
if (backslash && !(syntax_bits & RE_NO_GNU_OPS))
return lasttok = BEGWORD;
goto normal_char;
case '>':
if (backslash && !(syntax_bits & RE_NO_GNU_OPS))
return lasttok = ENDWORD;
goto normal_char;
case 'b':
if (backslash && !(syntax_bits & RE_NO_GNU_OPS))
return lasttok = LIMWORD;
goto normal_char;
case 'B':
if (backslash && !(syntax_bits & RE_NO_GNU_OPS))
return lasttok = NOTLIMWORD;
goto normal_char;
case '?':
if (syntax_bits & RE_LIMITED_OPS)
goto normal_char;
if (backslash != ((syntax_bits & RE_BK_PLUS_QM) != 0))
goto normal_char;
if (!(syntax_bits & RE_CONTEXT_INDEP_OPS) && laststart)
goto normal_char;
return lasttok = QMARK;
case '*':
if (backslash)
goto normal_char;
if (!(syntax_bits & RE_CONTEXT_INDEP_OPS) && laststart)
goto normal_char;
return lasttok = STAR;
case '+':
if (syntax_bits & RE_LIMITED_OPS)
goto normal_char;
if (backslash != ((syntax_bits & RE_BK_PLUS_QM) != 0))
goto normal_char;
if (!(syntax_bits & RE_CONTEXT_INDEP_OPS) && laststart)
goto normal_char;
return lasttok = PLUS;
case '{':
if (!(syntax_bits & RE_INTERVALS))
goto normal_char;
if (backslash != ((syntax_bits & RE_NO_BK_BRACES) == 0))
goto normal_char;
if (!(syntax_bits & RE_CONTEXT_INDEP_OPS) && laststart)
goto normal_char;
if (syntax_bits & RE_NO_BK_BRACES)
{
/* Scan ahead for a valid interval; if it's not valid,
treat it as a literal '{'. */
int lo = -1, hi = -1;
char const *p = lexptr;
char const *lim = p + lexleft;
for (; p != lim && ISASCIIDIGIT (*p); p++)
lo = (lo < 0 ? 0 : lo * 10) + *p - '0';
if (p != lim && *p == ',')
while (++p != lim && ISASCIIDIGIT (*p))
hi = (hi < 0 ? 0 : hi * 10) + *p - '0';
else
hi = lo;
if (p == lim || *p != '}'
|| lo < 0 || RE_DUP_MAX < hi || (0 <= hi && hi < lo))
goto normal_char;
}
minrep = 0;
/* Cases:
{M} - exact count
{M,} - minimum count, maximum is infinity
{M,N} - M through N */
FETCH(c, _("unfinished repeat count"));
if (ISASCIIDIGIT (c))
{
minrep = c - '0';
for (;;)
{
FETCH(c, _("unfinished repeat count"));
if (! ISASCIIDIGIT (c))
break;
minrep = 10 * minrep + c - '0';
}
}
else
dfaerror(_("malformed repeat count"));
if (c == ',')
{
FETCH (c, _("unfinished repeat count"));
if (! ISASCIIDIGIT (c))
maxrep = -1;
else
{
maxrep = c - '0';
for (;;)
{
FETCH (c, _("unfinished repeat count"));
if (! ISASCIIDIGIT (c))
break;
maxrep = 10 * maxrep + c - '0';
}
if (0 <= maxrep && maxrep < minrep)
dfaerror (_("malformed repeat count"));
}
}
else
maxrep = minrep;
if (!(syntax_bits & RE_NO_BK_BRACES))
{
if (c != '\\')
dfaerror(_("malformed repeat count"));
FETCH(c, _("unfinished repeat count"));
}
if (c != '}')
dfaerror(_("malformed repeat count"));
laststart = 0;
#ifdef GAWK
dfa->broken = (minrep == maxrep && minrep == 0);
#endif
return lasttok = REPMN;
case '|':
if (syntax_bits & RE_LIMITED_OPS)
goto normal_char;
if (backslash != ((syntax_bits & RE_NO_BK_VBAR) == 0))
goto normal_char;
laststart = 1;
return lasttok = OR;
case '\n':
if (syntax_bits & RE_LIMITED_OPS
|| backslash
|| !(syntax_bits & RE_NEWLINE_ALT))
goto normal_char;
laststart = 1;
return lasttok = OR;
case '(':
if (backslash != ((syntax_bits & RE_NO_BK_PARENS) == 0))
goto normal_char;
++parens;
laststart = 1;
return lasttok = LPAREN;
case ')':
if (backslash != ((syntax_bits & RE_NO_BK_PARENS) == 0))
goto normal_char;
if (parens == 0 && syntax_bits & RE_UNMATCHED_RIGHT_PAREN_ORD)
goto normal_char;
--parens;
laststart = 0;
return lasttok = RPAREN;
case '.':
if (backslash)
goto normal_char;
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
/* In multibyte environment period must match with a single
character not a byte. So we use ANYCHAR. */
laststart = 0;
return lasttok = ANYCHAR;
}
#endif /* MBS_SUPPORT */
zeroset(ccl);
notset(ccl);
if (!(syntax_bits & RE_DOT_NEWLINE))
clrbit(eolbyte, ccl);
if (syntax_bits & RE_DOT_NOT_NULL)
clrbit('\0', ccl);
laststart = 0;
return lasttok = CSET + charclass_index(ccl);
#ifndef GAWK
case 's':
case 'S':
if (!backslash || (syntax_bits & RE_NO_GNU_OPS))
goto normal_char;
zeroset(ccl);
for (c2 = 0; c2 < NOTCHAR; ++c2)
if (ISSPACE(c2))
setbit(c2, ccl);
if (c == 'S')
notset(ccl);
laststart = 0;
return lasttok = CSET + charclass_index(ccl);
#endif
case 'w':
case 'W':
if (!backslash || (syntax_bits & RE_NO_GNU_OPS))
goto normal_char;
zeroset(ccl);
for (c2 = 0; c2 < NOTCHAR; ++c2)
if (IS_WORD_CONSTITUENT(c2))
setbit(c2, ccl);
if (c == 'W')
notset(ccl);
laststart = 0;
return lasttok = CSET + charclass_index(ccl);
case '[':
if (backslash)
goto normal_char;
laststart = 0;
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
/* In multibyte environment a bracket expression may contain
multibyte characters, which must be treated as characters
(not bytes). So we parse it by parse_bracket_exp_mb(). */
parse_bracket_exp_mb();
return lasttok = MBCSET;
}
#endif
zeroset(ccl);
FETCH(c, _("Unbalanced ["));
if (c == '^')
{
FETCH(c, _("Unbalanced ["));
invert = 1;
}
else
invert = 0;
do
{
/* Nobody ever said this had to be fast. :-)
Note that if we're looking at some other [:...:]
construct, we just treat it as a bunch of ordinary
characters. We can do this because we assume
regex has checked for syntax errors before
dfa is ever called. */
if (c == '[' && (syntax_bits & RE_CHAR_CLASSES))
for (c1 = 0; prednames[c1].name; ++c1)
if (looking_at(prednames[c1].name))
{
int (*pred) PARAMS ((int)) = prednames[c1].pred;
for (c2 = 0; c2 < NOTCHAR; ++c2)
if ((*pred)(c2))
setbit_case_fold (c2, ccl);
lexptr += strlen(prednames[c1].name);
lexleft -= strlen(prednames[c1].name);
FETCH(c1, _("Unbalanced ["));
goto skip;
}
if (c == '\\' && (syntax_bits & RE_BACKSLASH_ESCAPE_IN_LISTS))
FETCH(c, _("Unbalanced ["));
FETCH(c1, _("Unbalanced ["));
if (c1 == '-')
{
FETCH(c2, _("Unbalanced ["));
if (c2 == ']')
{
/* In the case [x-], the - is an ordinary hyphen,
which is left in c1, the lookahead character. */
--lexptr;
++lexleft;
}
else
{
if (c2 == '\\'
&& (syntax_bits & RE_BACKSLASH_ESCAPE_IN_LISTS))
FETCH(c2, _("Unbalanced ["));
FETCH(c1, _("Unbalanced ["));
if (!hard_LC_COLLATE) {
for (; c <= c2; c++)
setbit_case_fold (c, ccl);
} else {
/* POSIX locales are painful - leave the decision to libc */
regex_t re;
char expr[6]; /* = { '[', c, '-', c2, ']', '\0' }; */
expr[0] = '['; expr[1] = c; expr[2] = '-';
expr[3] = c2; expr[4] = ']'; expr[5] = '\0';
if (regcomp (&re, expr, case_fold ? REG_ICASE : 0) == REG_NOERROR) {
for (c = 0; c < NOTCHAR; ++c) {
regmatch_t mat;
char buf[2]; /* = { c, '\0' }; */
buf[0] = c; buf[1] = '\0';
if (regexec (&re, buf, 1, &mat, 0) == REG_NOERROR
&& mat.rm_so == 0 && mat.rm_eo == 1)
setbit_case_fold (c, ccl);
}
regfree (&re);
}
}
continue;
}
}
setbit_case_fold (c, ccl);
skip:
;
}
while ((c = c1) != ']');
if (invert)
{
notset(ccl);
if (syntax_bits & RE_HAT_LISTS_NOT_NEWLINE)
clrbit(eolbyte, ccl);
}
return lasttok = CSET + charclass_index(ccl);
default:
normal_char:
laststart = 0;
if (case_fold && ISALPHA(c))
{
zeroset(ccl);
setbit_case_fold (c, ccl);
return lasttok = CSET + charclass_index(ccl);
}
return lasttok = c;
}
}
/* The above loop should consume at most a backslash
and some other character. */
abort();
return END; /* keeps pedantic compilers happy. */
}
/* Recursive descent parser for regular expressions. */
static token tok; /* Lookahead token. */
static int depth; /* Current depth of a hypothetical stack
holding deferred productions. This is
used to determine the depth that will be
required of the real stack later on in
dfaanalyze(). */
/* Add the given token to the parse tree, maintaining the depth count and
updating the maximum depth if necessary. */
static void
addtok (token t)
{
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
REALLOC_IF_NECESSARY(dfa->multibyte_prop, int, dfa->nmultibyte_prop,
dfa->tindex);
/* Set dfa->multibyte_prop. See struct dfa in dfa.h. */
if (t == MBCSET)
dfa->multibyte_prop[dfa->tindex] = ((dfa->nmbcsets - 1) << 2) + 3;
else if (t < NOTCHAR)
dfa->multibyte_prop[dfa->tindex]
= (cur_mb_len == 1)? 3 /* single-byte char */
: (((cur_mb_index == 1)? 1 : 0) /* 1st-byte of multibyte char */
+ ((cur_mb_index == cur_mb_len)? 2 : 0)); /* last-byte */
else
/* It may be unnecessary, but it is safer to treat other
symbols as single byte characters. */
dfa->multibyte_prop[dfa->tindex] = 3;
}
#endif
REALLOC_IF_NECESSARY(dfa->tokens, token, dfa->talloc, dfa->tindex);
dfa->tokens[dfa->tindex++] = t;
switch (t)
{
case QMARK:
case STAR:
case PLUS:
break;
case CAT:
case OR:
case ORTOP:
--depth;
break;
default:
++dfa->nleaves;
case EMPTY:
++depth;
break;
}
if (depth > dfa->depth)
dfa->depth = depth;
}
/* The grammar understood by the parser is as follows.
regexp:
regexp OR branch
branch
branch:
branch closure
closure
closure:
closure QMARK
closure STAR
closure PLUS
closure REPMN
atom
atom:
<normal character>
<multibyte character>
ANYCHAR
MBCSET
CSET
BACKREF
BEGLINE
ENDLINE
BEGWORD
ENDWORD
LIMWORD
NOTLIMWORD
CRANGE
LPAREN regexp RPAREN
<empty>
The parser builds a parse tree in postfix form in an array of tokens. */
static void
atom (void)
{
if ((tok >= 0 && tok < NOTCHAR) || tok >= CSET || tok == BACKREF
|| tok == BEGLINE || tok == ENDLINE || tok == BEGWORD
#ifdef MBS_SUPPORT
|| tok == ANYCHAR || tok == MBCSET /* MB_CUR_MAX > 1 */
#endif /* MBS_SUPPORT */
|| tok == ENDWORD || tok == LIMWORD || tok == NOTLIMWORD)
{
addtok(tok);
tok = lex();
#ifdef MBS_SUPPORT
/* We treat a multibyte character as a single atom, so that DFA
can treat a multibyte character as a single expression.
e.g. We construct following tree from "<mb1><mb2>".
<mb1(1st-byte)><mb1(2nd-byte)><CAT><mb1(3rd-byte)><CAT>
<mb2(1st-byte)><mb2(2nd-byte)><CAT><mb2(3rd-byte)><CAT><CAT>
*/
if (MB_CUR_MAX > 1)
{
while (cur_mb_index > 1 && tok >= 0 && tok < NOTCHAR)
{
addtok(tok);
addtok(CAT);
tok = lex();
}
}
#endif /* MBS_SUPPORT */
}
else if (tok == CRANGE)
{
/* A character range like "[a-z]" in a locale other than "C" or
"POSIX". This range might any sequence of one or more
characters. Unfortunately the POSIX locale primitives give
us no practical way to find what character sequences might be
matched. Treat this approximately like "(.\1)" -- i.e. match
one character, and then punt to the full matcher. */
charclass ccl;
zeroset (ccl);
notset (ccl);
addtok (CSET + charclass_index (ccl));
addtok (BACKREF);
addtok (CAT);
tok = lex ();
}
else if (tok == LPAREN)
{
tok = lex();
regexp(0);
if (tok != RPAREN)
dfaerror(_("Unbalanced ("));
tok = lex();
}
else
addtok(EMPTY);
}
/* Return the number of tokens in the given subexpression. */
static int
nsubtoks (int tindex)
{
int ntoks1;
switch (dfa->tokens[tindex - 1])
{
default:
return 1;
case QMARK:
case STAR:
case PLUS:
return 1 + nsubtoks(tindex - 1);
case CAT:
case OR:
case ORTOP:
ntoks1 = nsubtoks(tindex - 1);
return 1 + ntoks1 + nsubtoks(tindex - 1 - ntoks1);
}
}
/* Copy the given subexpression to the top of the tree. */
static void
copytoks (int tindex, int ntokens)
{
int i;
for (i = 0; i < ntokens; ++i)
{
addtok(dfa->tokens[tindex + i]);
#ifdef MBS_SUPPORT
/* Update index into multibyte csets. */
if (MB_CUR_MAX > 1 && dfa->tokens[tindex + i] == MBCSET)
dfa->multibyte_prop[dfa->tindex - 1] = dfa->multibyte_prop[tindex + i];
#endif
}
}
static void
closure (void)
{
int tindex, ntokens, i;
atom();
while (tok == QMARK || tok == STAR || tok == PLUS || tok == REPMN)
if (tok == REPMN)
{
ntokens = nsubtoks(dfa->tindex);
tindex = dfa->tindex - ntokens;
if (maxrep < 0)
addtok(PLUS);
if (minrep == 0)
addtok(QMARK);
for (i = 1; i < minrep; ++i)
{
copytoks(tindex, ntokens);
addtok(CAT);
}
for (; i < maxrep; ++i)
{
copytoks(tindex, ntokens);
addtok(QMARK);
addtok(CAT);
}
tok = lex();
}
else
{
addtok(tok);
tok = lex();
}
}
static void
branch (void)
{
closure();
while (tok != RPAREN && tok != OR && tok >= 0)
{
closure();
addtok(CAT);
}
}
static void
regexp (int toplevel)
{
branch();
while (tok == OR)
{
tok = lex();
branch();
if (toplevel)
addtok(ORTOP);
else
addtok(OR);
}
}
/* Main entry point for the parser. S is a string to be parsed, len is the
length of the string, so s can include NUL characters. D is a pointer to
the struct dfa to parse into. */
void
dfaparse (char const *s, size_t len, struct dfa *d)
{
dfa = d;
lexptr = s;
lexleft = len;
lasttok = END;
laststart = 1;
parens = 0;
#ifdef LC_COLLATE
hard_LC_COLLATE = hard_locale (LC_COLLATE);
#endif
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
cur_mb_index = 0;
cur_mb_len = 0;
memset(&mbs, 0, sizeof(mbstate_t));
}
#endif /* MBS_SUPPORT */
if (! syntax_bits_set)
dfaerror(_("No syntax specified"));
tok = lex();
depth = d->depth;
regexp(1);
if (tok != END)
dfaerror(_("Unbalanced )"));
addtok(END - d->nregexps);
addtok(CAT);
if (d->nregexps)
addtok(ORTOP);
++d->nregexps;
}
/* Some primitives for operating on sets of positions. */
/* Copy one set to another; the destination must be large enough. */
static void
copy (position_set const *src, position_set *dst)
{
int i;
for (i = 0; i < src->nelem; ++i)
dst->elems[i] = src->elems[i];
dst->nelem = src->nelem;
}
/* Insert a position in a set. Position sets are maintained in sorted
order according to index. If position already exists in the set with
the same index then their constraints are logically or'd together.
S->elems must point to an array large enough to hold the resulting set. */
static void
insert (position p, position_set *s)
{
int i;
int start = -1;
int end = s->nelem - 1;
while (start < end)
{
int midpoint = (start + end + 1) / 2;
if (p.index < s->elems[midpoint].index)
{
/* We can skip to the midpoint without missing our insert position */
start = midpoint;
}
else
{
/* The midpoint is after our insert position, go back */
end = midpoint - 1;
}
}
i = start + 1;
if (i < s->nelem && p.index == s->elems[i].index)
s->elems[i].constraint |= p.constraint;
else
{
int update_pos;
++s->nelem;
for (update_pos = s->nelem - 1; update_pos > i; update_pos--)
{
s->elems[update_pos] = s->elems[update_pos - 1];
}
s->elems[i] = p;
}
}
/* Merge two sets of positions into a third. The result is exactly as if
the positions of both sets were inserted into an initially empty set. */
static void
merge (position_set const *s1, position_set const *s2, position_set *m)
{
int i = 0, j = 0;
m->nelem = 0;
while (i < s1->nelem && j < s2->nelem)
if (s1->elems[i].index > s2->elems[j].index)
m->elems[m->nelem++] = s1->elems[i++];
else if (s1->elems[i].index < s2->elems[j].index)
m->elems[m->nelem++] = s2->elems[j++];
else
{
m->elems[m->nelem] = s1->elems[i++];
m->elems[m->nelem++].constraint |= s2->elems[j++].constraint;
}
while (i < s1->nelem)
m->elems[m->nelem++] = s1->elems[i++];
while (j < s2->nelem)
m->elems[m->nelem++] = s2->elems[j++];
}
/* Delete a position from a set. */
static void
delete (position p, position_set *s)
{
int i;
for (i = 0; i < s->nelem; ++i)
if (p.index == s->elems[i].index)
break;
if (i < s->nelem)
for (--s->nelem; i < s->nelem; ++i)
s->elems[i] = s->elems[i + 1];
}
/* Find the index of the state corresponding to the given position set with
the given preceding context, or create a new state if there is no such
state. Newline and letter tell whether we got here on a newline or
letter, respectively. */
static int
state_index (struct dfa *d, position_set const *s, int newline, int letter)
{
int hash = 0;
int constraint;
int i, j;
newline = newline ? 1 : 0;
letter = letter ? 1 : 0;
for (i = 0; i < s->nelem; ++i)
hash ^= s->elems[i].index + s->elems[i].constraint;
/* Try to find a state that exactly matches the proposed one. */
for (i = 0; i < d->sindex; ++i)
{
if (hash != d->states[i].hash || s->nelem != d->states[i].elems.nelem
|| newline != d->states[i].newline || letter != d->states[i].letter)
continue;
for (j = 0; j < s->nelem; ++j)
if (s->elems[j].constraint
!= d->states[i].elems.elems[j].constraint
|| s->elems[j].index != d->states[i].elems.elems[j].index)
break;
if (j == s->nelem)
return i;
}
/* We'll have to create a new state. */
REALLOC_IF_NECESSARY(d->states, dfa_state, d->salloc, d->sindex);
d->states[i].hash = hash;
MALLOC(d->states[i].elems.elems, position, s->nelem);
copy(s, &d->states[i].elems);
d->states[i].newline = newline;
d->states[i].letter = letter;
d->states[i].backref = 0;
d->states[i].constraint = 0;
d->states[i].first_end = 0;
#ifdef MBS_SUPPORT
d->states[i].mbps.nelem = 0;
d->states[i].mbps.elems = NULL;
#endif
for (j = 0; j < s->nelem; ++j)
if (d->tokens[s->elems[j].index] < 0)
{
constraint = s->elems[j].constraint;
if (SUCCEEDS_IN_CONTEXT(constraint, newline, 0, letter, 0)
|| SUCCEEDS_IN_CONTEXT(constraint, newline, 0, letter, 1)
|| SUCCEEDS_IN_CONTEXT(constraint, newline, 1, letter, 0)
|| SUCCEEDS_IN_CONTEXT(constraint, newline, 1, letter, 1))
d->states[i].constraint |= constraint;
if (! d->states[i].first_end)
d->states[i].first_end = d->tokens[s->elems[j].index];
}
else if (d->tokens[s->elems[j].index] == BACKREF)
{
d->states[i].constraint = NO_CONSTRAINT;
d->states[i].backref = 1;
}
++d->sindex;
return i;
}
/* Find the epsilon closure of a set of positions. If any position of the set
contains a symbol that matches the empty string in some context, replace
that position with the elements of its follow labeled with an appropriate
constraint. Repeat exhaustively until no funny positions are left.
S->elems must be large enough to hold the result. */
static void
epsclosure (position_set *s, struct dfa const *d)
{
int i, j;
char *visited; /* array of booleans, enough to use char, not int */
position p, old;
CALLOC(visited, char, d->tindex);
for (i = 0; i < s->nelem; ++i)
if (d->tokens[s->elems[i].index] >= NOTCHAR
&& d->tokens[s->elems[i].index] != BACKREF
#ifdef MBS_SUPPORT
&& d->tokens[s->elems[i].index] != ANYCHAR
&& d->tokens[s->elems[i].index] != MBCSET
#endif
&& d->tokens[s->elems[i].index] < CSET)
{
old = s->elems[i];
p.constraint = old.constraint;
delete(s->elems[i], s);
if (visited[old.index])
{
--i;
continue;
}
visited[old.index] = 1;
switch (d->tokens[old.index])
{
case BEGLINE:
p.constraint &= BEGLINE_CONSTRAINT;
break;
case ENDLINE:
p.constraint &= ENDLINE_CONSTRAINT;
break;
case BEGWORD:
p.constraint &= BEGWORD_CONSTRAINT;
break;
case ENDWORD:
p.constraint &= ENDWORD_CONSTRAINT;
break;
case LIMWORD:
p.constraint &= LIMWORD_CONSTRAINT;
break;
case NOTLIMWORD:
p.constraint &= NOTLIMWORD_CONSTRAINT;
break;
default:
break;
}
for (j = 0; j < d->follows[old.index].nelem; ++j)
{
p.index = d->follows[old.index].elems[j].index;
insert(p, s);
}
/* Force rescan to start at the beginning. */
i = -1;
}
free(visited);
}
/* Perform bottom-up analysis on the parse tree, computing various functions.
Note that at this point, we're pretending constructs like \< are real
characters rather than constraints on what can follow them.
Nullable: A node is nullable if it is at the root of a regexp that can
match the empty string.
* EMPTY leaves are nullable.
* No other leaf is nullable.
* A QMARK or STAR node is nullable.
* A PLUS node is nullable if its argument is nullable.
* A CAT node is nullable if both its arguments are nullable.
* An OR node is nullable if either argument is nullable.
Firstpos: The firstpos of a node is the set of positions (nonempty leaves)
that could correspond to the first character of a string matching the
regexp rooted at the given node.
* EMPTY leaves have empty firstpos.
* The firstpos of a nonempty leaf is that leaf itself.
* The firstpos of a QMARK, STAR, or PLUS node is the firstpos of its
argument.
* The firstpos of a CAT node is the firstpos of the left argument, union
the firstpos of the right if the left argument is nullable.
* The firstpos of an OR node is the union of firstpos of each argument.
Lastpos: The lastpos of a node is the set of positions that could
correspond to the last character of a string matching the regexp at
the given node.
* EMPTY leaves have empty lastpos.
* The lastpos of a nonempty leaf is that leaf itself.
* The lastpos of a QMARK, STAR, or PLUS node is the lastpos of its
argument.
* The lastpos of a CAT node is the lastpos of its right argument, union
the lastpos of the left if the right argument is nullable.
* The lastpos of an OR node is the union of the lastpos of each argument.
Follow: The follow of a position is the set of positions that could
correspond to the character following a character matching the node in
a string matching the regexp. At this point we consider special symbols
that match the empty string in some context to be just normal characters.
Later, if we find that a special symbol is in a follow set, we will
replace it with the elements of its follow, labeled with an appropriate
constraint.
* Every node in the firstpos of the argument of a STAR or PLUS node is in
the follow of every node in the lastpos.
* Every node in the firstpos of the second argument of a CAT node is in
the follow of every node in the lastpos of the first argument.
Because of the postfix representation of the parse tree, the depth-first
analysis is conveniently done by a linear scan with the aid of a stack.
Sets are stored as arrays of the elements, obeying a stack-like allocation
scheme; the number of elements in each set deeper in the stack can be
used to determine the address of a particular set's array. */
void
dfaanalyze (struct dfa *d, int searchflag)
{
int *nullable; /* Nullable stack. */
int *nfirstpos; /* Element count stack for firstpos sets. */
position *firstpos; /* Array where firstpos elements are stored. */
int *nlastpos; /* Element count stack for lastpos sets. */
position *lastpos; /* Array where lastpos elements are stored. */
int *nalloc; /* Sizes of arrays allocated to follow sets. */
position_set tmp; /* Temporary set for merging sets. */
position_set merged; /* Result of merging sets. */
int wants_newline; /* True if some position wants newline info. */
int *o_nullable;
int *o_nfirst, *o_nlast;
position *o_firstpos, *o_lastpos;
int i, j;
position *pos;
#ifdef DEBUG
fprintf(stderr, "dfaanalyze:\n");
for (i = 0; i < d->tindex; ++i)
{
fprintf(stderr, " %d:", i);
prtok(d->tokens[i]);
}
putc('\n', stderr);
#endif
d->searchflag = searchflag;
MALLOC(nullable, int, d->depth);
o_nullable = nullable;
MALLOC(nfirstpos, int, d->depth);
o_nfirst = nfirstpos;
MALLOC(firstpos, position, d->nleaves);
o_firstpos = firstpos, firstpos += d->nleaves;
MALLOC(nlastpos, int, d->depth);
o_nlast = nlastpos;
MALLOC(lastpos, position, d->nleaves);
o_lastpos = lastpos, lastpos += d->nleaves;
CALLOC(nalloc, int, d->tindex);
MALLOC(merged.elems, position, d->nleaves);
CALLOC(d->follows, position_set, d->tindex);
for (i = 0; i < d->tindex; ++i)
#ifdef DEBUG
{ /* Nonsyntactic #ifdef goo... */
#endif
switch (d->tokens[i])
{
case EMPTY:
/* The empty set is nullable. */
*nullable++ = 1;
/* The firstpos and lastpos of the empty leaf are both empty. */
*nfirstpos++ = *nlastpos++ = 0;
break;
case STAR:
case PLUS:
/* Every element in the firstpos of the argument is in the follow
of every element in the lastpos. */
tmp.nelem = nfirstpos[-1];
tmp.elems = firstpos;
pos = lastpos;
for (j = 0; j < nlastpos[-1]; ++j)
{
merge(&tmp, &d->follows[pos[j].index], &merged);
REALLOC_IF_NECESSARY(d->follows[pos[j].index].elems, position,
nalloc[pos[j].index], merged.nelem - 1);
copy(&merged, &d->follows[pos[j].index]);
}
case QMARK:
/* A QMARK or STAR node is automatically nullable. */
if (d->tokens[i] != PLUS)
nullable[-1] = 1;
break;
case CAT:
/* Every element in the firstpos of the second argument is in the
follow of every element in the lastpos of the first argument. */
tmp.nelem = nfirstpos[-1];
tmp.elems = firstpos;
pos = lastpos + nlastpos[-1];
for (j = 0; j < nlastpos[-2]; ++j)
{
merge(&tmp, &d->follows[pos[j].index], &merged);
REALLOC_IF_NECESSARY(d->follows[pos[j].index].elems, position,
nalloc[pos[j].index], merged.nelem - 1);
copy(&merged, &d->follows[pos[j].index]);
}
/* The firstpos of a CAT node is the firstpos of the first argument,
union that of the second argument if the first is nullable. */
if (nullable[-2])
nfirstpos[-2] += nfirstpos[-1];
else
firstpos += nfirstpos[-1];
--nfirstpos;
/* The lastpos of a CAT node is the lastpos of the second argument,
union that of the first argument if the second is nullable. */
if (nullable[-1])
nlastpos[-2] += nlastpos[-1];
else
{
pos = lastpos + nlastpos[-2];
for (j = nlastpos[-1] - 1; j >= 0; --j)
pos[j] = lastpos[j];
lastpos += nlastpos[-2];
nlastpos[-2] = nlastpos[-1];
}
--nlastpos;
/* A CAT node is nullable if both arguments are nullable. */
nullable[-2] = nullable[-1] && nullable[-2];
--nullable;
break;
case OR:
case ORTOP:
/* The firstpos is the union of the firstpos of each argument. */
nfirstpos[-2] += nfirstpos[-1];
--nfirstpos;
/* The lastpos is the union of the lastpos of each argument. */
nlastpos[-2] += nlastpos[-1];
--nlastpos;
/* An OR node is nullable if either argument is nullable. */
nullable[-2] = nullable[-1] || nullable[-2];
--nullable;
break;
default:
/* Anything else is a nonempty position. (Note that special
constructs like \< are treated as nonempty strings here;
an "epsilon closure" effectively makes them nullable later.
Backreferences have to get a real position so we can detect
transitions on them later. But they are nullable. */
*nullable++ = d->tokens[i] == BACKREF;
/* This position is in its own firstpos and lastpos. */
*nfirstpos++ = *nlastpos++ = 1;
--firstpos, --lastpos;
firstpos->index = lastpos->index = i;
firstpos->constraint = lastpos->constraint = NO_CONSTRAINT;
/* Allocate the follow set for this position. */
nalloc[i] = 1;
MALLOC(d->follows[i].elems, position, nalloc[i]);
break;
}
#ifdef DEBUG
/* ... balance the above nonsyntactic #ifdef goo... */
fprintf(stderr, "node %d:", i);
prtok(d->tokens[i]);
putc('\n', stderr);
fprintf(stderr, nullable[-1] ? " nullable: yes\n" : " nullable: no\n");
fprintf(stderr, " firstpos:");
for (j = nfirstpos[-1] - 1; j >= 0; --j)
{
fprintf(stderr, " %d:", firstpos[j].index);
prtok(d->tokens[firstpos[j].index]);
}
fprintf(stderr, "\n lastpos:");
for (j = nlastpos[-1] - 1; j >= 0; --j)
{
fprintf(stderr, " %d:", lastpos[j].index);
prtok(d->tokens[lastpos[j].index]);
}
putc('\n', stderr);
}
#endif
/* For each follow set that is the follow set of a real position, replace
it with its epsilon closure. */
for (i = 0; i < d->tindex; ++i)
if (d->tokens[i] < NOTCHAR || d->tokens[i] == BACKREF
#ifdef MBS_SUPPORT
|| d->tokens[i] == ANYCHAR
|| d->tokens[i] == MBCSET
#endif
|| d->tokens[i] >= CSET)
{
#ifdef DEBUG
fprintf(stderr, "follows(%d:", i);
prtok(d->tokens[i]);
fprintf(stderr, "):");
for (j = d->follows[i].nelem - 1; j >= 0; --j)
{
fprintf(stderr, " %d:", d->follows[i].elems[j].index);
prtok(d->tokens[d->follows[i].elems[j].index]);
}
putc('\n', stderr);
#endif
copy(&d->follows[i], &merged);
epsclosure(&merged, d);
if (d->follows[i].nelem < merged.nelem)
REALLOC(d->follows[i].elems, position, merged.nelem);
copy(&merged, &d->follows[i]);
}
/* Get the epsilon closure of the firstpos of the regexp. The result will
be the set of positions of state 0. */
merged.nelem = 0;
for (i = 0; i < nfirstpos[-1]; ++i)
insert(firstpos[i], &merged);
epsclosure(&merged, d);
/* Check if any of the positions of state 0 will want newline context. */
wants_newline = 0;
for (i = 0; i < merged.nelem; ++i)
if (PREV_NEWLINE_DEPENDENT(merged.elems[i].constraint))
wants_newline = 1;
/* Build the initial state. */
d->salloc = 1;
d->sindex = 0;
MALLOC(d->states, dfa_state, d->salloc);
state_index(d, &merged, wants_newline, 0);
free(o_nullable);
free(o_nfirst);
free(o_firstpos);
free(o_nlast);
free(o_lastpos);
free(nalloc);
free(merged.elems);
}
/* Find, for each character, the transition out of state s of d, and store
it in the appropriate slot of trans.
We divide the positions of s into groups (positions can appear in more
than one group). Each group is labeled with a set of characters that
every position in the group matches (taking into account, if necessary,
preceding context information of s). For each group, find the union
of the its elements' follows. This set is the set of positions of the
new state. For each character in the group's label, set the transition
on this character to be to a state corresponding to the set's positions,
and its associated backward context information, if necessary.
If we are building a searching matcher, we include the positions of state
0 in every state.
The collection of groups is constructed by building an equivalence-class
partition of the positions of s.
For each position, find the set of characters C that it matches. Eliminate
any characters from C that fail on grounds of backward context.
Search through the groups, looking for a group whose label L has nonempty
intersection with C. If L - C is nonempty, create a new group labeled
L - C and having the same positions as the current group, and set L to
the intersection of L and C. Insert the position in this group, set
C = C - L, and resume scanning.
If after comparing with every group there are characters remaining in C,
create a new group labeled with the characters of C and insert this
position in that group. */
void
dfastate (int s, struct dfa *d, int trans[])
{
position_set grps[NOTCHAR]; /* As many as will ever be needed. */
charclass labels[NOTCHAR]; /* Labels corresponding to the groups. */
int ngrps = 0; /* Number of groups actually used. */
position pos; /* Current position being considered. */
charclass matches; /* Set of matching characters. */
int matchesf; /* True if matches is nonempty. */
charclass intersect; /* Intersection with some label set. */
int intersectf; /* True if intersect is nonempty. */
charclass leftovers; /* Stuff in the label that didn't match. */
int leftoversf; /* True if leftovers is nonempty. */
static charclass letters; /* Set of characters considered letters. */
static charclass newline; /* Set of characters that aren't newline. */
position_set follows; /* Union of the follows of some group. */
position_set tmp; /* Temporary space for merging sets. */
int state; /* New state. */
int wants_newline; /* New state wants to know newline context. */
int state_newline; /* New state on a newline transition. */
int wants_letter; /* New state wants to know letter context. */
int state_letter; /* New state on a letter transition. */
static int initialized; /* Flag for static initialization. */
#ifdef MBS_SUPPORT
int next_isnt_1st_byte = 0; /* Flag if we can't add state0. */
#endif
int i, j, k;
/* Initialize the set of letters, if necessary. */
if (! initialized)
{
initialized = 1;
for (i = 0; i < NOTCHAR; ++i)
if (IS_WORD_CONSTITUENT(i))
setbit(i, letters);
setbit(eolbyte, newline);
}
zeroset(matches);
for (i = 0; i < d->states[s].elems.nelem; ++i)
{
pos = d->states[s].elems.elems[i];
if (d->tokens[pos.index] >= 0 && d->tokens[pos.index] < NOTCHAR)
setbit(d->tokens[pos.index], matches);
else if (d->tokens[pos.index] >= CSET)
copyset(d->charclasses[d->tokens[pos.index] - CSET], matches);
#ifdef MBS_SUPPORT
else if (d->tokens[pos.index] == ANYCHAR
|| d->tokens[pos.index] == MBCSET)
/* MB_CUR_MAX > 1 */
{
/* ANYCHAR and MBCSET must match with a single character, so we
must put it to d->states[s].mbps, which contains the positions
which can match with a single character not a byte. */
if (d->states[s].mbps.nelem == 0)
{
MALLOC(d->states[s].mbps.elems, position,
d->states[s].elems.nelem);
}
insert(pos, &(d->states[s].mbps));
continue;
}
#endif /* MBS_SUPPORT */
else
continue;
/* Some characters may need to be eliminated from matches because
they fail in the current context. */
if (pos.constraint != 0xFF)
{
if (! MATCHES_NEWLINE_CONTEXT(pos.constraint,
d->states[s].newline, 1))
clrbit(eolbyte, matches);
if (! MATCHES_NEWLINE_CONTEXT(pos.constraint,
d->states[s].newline, 0))
for (j = 0; j < CHARCLASS_INTS; ++j)
matches[j] &= newline[j];
if (! MATCHES_LETTER_CONTEXT(pos.constraint,
d->states[s].letter, 1))
for (j = 0; j < CHARCLASS_INTS; ++j)
matches[j] &= ~letters[j];
if (! MATCHES_LETTER_CONTEXT(pos.constraint,
d->states[s].letter, 0))
for (j = 0; j < CHARCLASS_INTS; ++j)
matches[j] &= letters[j];
/* If there are no characters left, there's no point in going on. */
for (j = 0; j < CHARCLASS_INTS && !matches[j]; ++j)
continue;
if (j == CHARCLASS_INTS)
continue;
}
for (j = 0; j < ngrps; ++j)
{
/* If matches contains a single character only, and the current
group's label doesn't contain that character, go on to the
next group. */
if (d->tokens[pos.index] >= 0 && d->tokens[pos.index] < NOTCHAR
&& !tstbit(d->tokens[pos.index], labels[j]))
continue;
/* Check if this group's label has a nonempty intersection with
matches. */
intersectf = 0;
for (k = 0; k < CHARCLASS_INTS; ++k)
(intersect[k] = matches[k] & labels[j][k]) ? (intersectf = 1) : 0;
if (! intersectf)
continue;
/* It does; now find the set differences both ways. */
leftoversf = matchesf = 0;
for (k = 0; k < CHARCLASS_INTS; ++k)
{
/* Even an optimizing compiler can't know this for sure. */
int match = matches[k], label = labels[j][k];
(leftovers[k] = ~match & label) ? (leftoversf = 1) : 0;
(matches[k] = match & ~label) ? (matchesf = 1) : 0;
}
/* If there were leftovers, create a new group labeled with them. */
if (leftoversf)
{
copyset(leftovers, labels[ngrps]);
copyset(intersect, labels[j]);
MALLOC(grps[ngrps].elems, position, d->nleaves);
copy(&grps[j], &grps[ngrps]);
++ngrps;
}
/* Put the position in the current group. Note that there is no
reason to call insert() here. */
grps[j].elems[grps[j].nelem++] = pos;
/* If every character matching the current position has been
accounted for, we're done. */
if (! matchesf)
break;
}
/* If we've passed the last group, and there are still characters
unaccounted for, then we'll have to create a new group. */
if (j == ngrps)
{
copyset(matches, labels[ngrps]);
zeroset(matches);
MALLOC(grps[ngrps].elems, position, d->nleaves);
grps[ngrps].nelem = 1;
grps[ngrps].elems[0] = pos;
++ngrps;
}
}
MALLOC(follows.elems, position, d->nleaves);
MALLOC(tmp.elems, position, d->nleaves);
/* If we are a searching matcher, the default transition is to a state
containing the positions of state 0, otherwise the default transition
is to fail miserably. */
if (d->searchflag)
{
wants_newline = 0;
wants_letter = 0;
for (i = 0; i < d->states[0].elems.nelem; ++i)
{
if (PREV_NEWLINE_DEPENDENT(d->states[0].elems.elems[i].constraint))
wants_newline = 1;
if (PREV_LETTER_DEPENDENT(d->states[0].elems.elems[i].constraint))
wants_letter = 1;
}
copy(&d->states[0].elems, &follows);
state = state_index(d, &follows, 0, 0);
if (wants_newline)
state_newline = state_index(d, &follows, 1, 0);
else
state_newline = state;
if (wants_letter)
state_letter = state_index(d, &follows, 0, 1);
else
state_letter = state;
for (i = 0; i < NOTCHAR; ++i)
trans[i] = (IS_WORD_CONSTITUENT(i)) ? state_letter : state;
trans[eolbyte] = state_newline;
}
else
for (i = 0; i < NOTCHAR; ++i)
trans[i] = -1;
for (i = 0; i < ngrps; ++i)
{
follows.nelem = 0;
/* Find the union of the follows of the positions of the group.
This is a hideously inefficient loop. Fix it someday. */
for (j = 0; j < grps[i].nelem; ++j)
for (k = 0; k < d->follows[grps[i].elems[j].index].nelem; ++k)
insert(d->follows[grps[i].elems[j].index].elems[k], &follows);
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
/* If a token in follows.elems is not 1st byte of a multibyte
character, or the states of follows must accept the bytes
which are not 1st byte of the multibyte character.
Then, if a state of follows encounter a byte, it must not be
a 1st byte of a multibyte character nor single byte character.
We cansel to add state[0].follows to next state, because
state[0] must accept 1st-byte
For example, we assume <sb a> is a certain single byte
character, <mb A> is a certain multibyte character, and the
codepoint of <sb a> equals the 2nd byte of the codepoint of
<mb A>.
When state[0] accepts <sb a>, state[i] transit to state[i+1]
by accepting accepts 1st byte of <mb A>, and state[i+1]
accepts 2nd byte of <mb A>, if state[i+1] encounter the
codepoint of <sb a>, it must not be <sb a> but 2nd byte of
<mb A>, so we can not add state[0]. */
next_isnt_1st_byte = 0;
for (j = 0; j < follows.nelem; ++j)
{
if (!(d->multibyte_prop[follows.elems[j].index] & 1))
{
next_isnt_1st_byte = 1;
break;
}
}
}
#endif
/* If we are building a searching matcher, throw in the positions
of state 0 as well. */
#ifdef MBS_SUPPORT
if (d->searchflag && (MB_CUR_MAX == 1 || !next_isnt_1st_byte))
#else
if (d->searchflag)
#endif
for (j = 0; j < d->states[0].elems.nelem; ++j)
insert(d->states[0].elems.elems[j], &follows);
/* Find out if the new state will want any context information. */
wants_newline = 0;
if (tstbit(eolbyte, labels[i]))
for (j = 0; j < follows.nelem; ++j)
if (PREV_NEWLINE_DEPENDENT(follows.elems[j].constraint))
wants_newline = 1;
wants_letter = 0;
for (j = 0; j < CHARCLASS_INTS; ++j)
if (labels[i][j] & letters[j])
break;
if (j < CHARCLASS_INTS)
for (j = 0; j < follows.nelem; ++j)
if (PREV_LETTER_DEPENDENT(follows.elems[j].constraint))
wants_letter = 1;
/* Find the state(s) corresponding to the union of the follows. */
state = state_index(d, &follows, 0, 0);
if (wants_newline)
state_newline = state_index(d, &follows, 1, 0);
else
state_newline = state;
if (wants_letter)
state_letter = state_index(d, &follows, 0, 1);
else
state_letter = state;
/* Set the transitions for each character in the current label. */
for (j = 0; j < CHARCLASS_INTS; ++j)
for (k = 0; k < INTBITS; ++k)
if (labels[i][j] & 1 << k)
{
int c = j * INTBITS + k;
if (c == eolbyte)
trans[c] = state_newline;
else if (IS_WORD_CONSTITUENT(c))
trans[c] = state_letter;
else if (c < NOTCHAR)
trans[c] = state;
}
}
for (i = 0; i < ngrps; ++i)
free(grps[i].elems);
free(follows.elems);
free(tmp.elems);
}
/* Some routines for manipulating a compiled dfa's transition tables.
Each state may or may not have a transition table; if it does, and it
is a non-accepting state, then d->trans[state] points to its table.
If it is an accepting state then d->fails[state] points to its table.
If it has no table at all, then d->trans[state] is NULL.
TODO: Improve this comment, get rid of the unnecessary redundancy. */
static void
build_state (int s, struct dfa *d)
{
int *trans; /* The new transition table. */
int i;
/* Set an upper limit on the number of transition tables that will ever
exist at once. 1024 is arbitrary. The idea is that the frequently
used transition tables will be quickly rebuilt, whereas the ones that
were only needed once or twice will be cleared away. */
if (d->trcount >= 1024)
{
for (i = 0; i < d->tralloc; ++i)
if (d->trans[i])
{
free((ptr_t) d->trans[i]);
d->trans[i] = NULL;
}
else if (d->fails[i])
{
free((ptr_t) d->fails[i]);
d->fails[i] = NULL;
}
d->trcount = 0;
}
++d->trcount;
/* Set up the success bits for this state. */
d->success[s] = 0;
if (ACCEPTS_IN_CONTEXT(d->states[s].newline, 1, d->states[s].letter, 0,
s, *d))
d->success[s] |= 4;
if (ACCEPTS_IN_CONTEXT(d->states[s].newline, 0, d->states[s].letter, 1,
s, *d))
d->success[s] |= 2;
if (ACCEPTS_IN_CONTEXT(d->states[s].newline, 0, d->states[s].letter, 0,
s, *d))
d->success[s] |= 1;
MALLOC(trans, int, NOTCHAR);
dfastate(s, d, trans);
/* Now go through the new transition table, and make sure that the trans
and fail arrays are allocated large enough to hold a pointer for the
largest state mentioned in the table. */
for (i = 0; i < NOTCHAR; ++i)
if (trans[i] >= d->tralloc)
{
int oldalloc = d->tralloc;
while (trans[i] >= d->tralloc)
d->tralloc *= 2;
REALLOC(d->realtrans, int *, d->tralloc + 1);
d->trans = d->realtrans + 1;
REALLOC(d->fails, int *, d->tralloc);
REALLOC(d->success, int, d->tralloc);
REALLOC(d->newlines, int, d->tralloc);
while (oldalloc < d->tralloc)
{
d->trans[oldalloc] = NULL;
d->fails[oldalloc++] = NULL;
}
}
/* Keep the newline transition in a special place so we can use it as
a sentinel. */
d->newlines[s] = trans[eolbyte];
trans[eolbyte] = -1;
if (ACCEPTING(s, *d))
d->fails[s] = trans;
else
d->trans[s] = trans;
}
static void
build_state_zero (struct dfa *d)
{
d->tralloc = 1;
d->trcount = 0;
CALLOC(d->realtrans, int *, d->tralloc + 1);
d->trans = d->realtrans + 1;
CALLOC(d->fails, int *, d->tralloc);
MALLOC(d->success, int, d->tralloc);
MALLOC(d->newlines, int, d->tralloc);
build_state(0, d);
}
#ifdef MBS_SUPPORT
/* Multibyte character handling sub-routines for dfaexec. */
/* Initial state may encounter the byte which is not a single byte character
nor 1st byte of a multibyte character. But it is incorrect for initial
state to accept such a byte.
For example, in sjis encoding the regular expression like "\\" accepts
the codepoint 0x5c, but should not accept the 2nd byte of the codepoint
0x815c. Then Initial state must skip the bytes which are not a single byte
character nor 1st byte of a multibyte character. */
#define SKIP_REMAINS_MB_IF_INITIAL_STATE(s, p) \
if (s == 0) \
{ \
while (inputwcs[p - buf_begin] == 0 \
&& mblen_buf[p - buf_begin] > 0 \
&& (unsigned char const *)p < buf_end) \
++p; \
if ((char *)p >= end) \
{ \
free(mblen_buf); \
free(inputwcs); \
return NULL; \
} \
}
static void
realloc_trans_if_necessary(struct dfa *d, int new_state)
{
/* Make sure that the trans and fail arrays are allocated large enough
to hold a pointer for the new state. */
if (new_state >= d->tralloc)
{
int oldalloc = d->tralloc;
while (new_state >= d->tralloc)
d->tralloc *= 2;
REALLOC(d->realtrans, int *, d->tralloc + 1);
d->trans = d->realtrans + 1;
REALLOC(d->fails, int *, d->tralloc);
REALLOC(d->success, int, d->tralloc);
REALLOC(d->newlines, int, d->tralloc);
while (oldalloc < d->tralloc)
{
d->trans[oldalloc] = NULL;
d->fails[oldalloc++] = NULL;
}
}
}
/* Return values of transit_state_singlebyte(), and
transit_state_consume_1char. */
typedef enum
{
TRANSIT_STATE_IN_PROGRESS, /* State transition has not finished. */
TRANSIT_STATE_DONE, /* State transition has finished. */
TRANSIT_STATE_END_BUFFER /* Reach the end of the buffer. */
} status_transit_state;
/* Consume a single byte and transit state from 's' to '*next_state'.
This function is almost same as the state transition routin in dfaexec().
But state transition is done just once, otherwise matching succeed or
reach the end of the buffer. */
static status_transit_state
transit_state_singlebyte (struct dfa *d, int s, unsigned char const *p,
int *next_state)
{
int *t;
int works = s;
status_transit_state rval = TRANSIT_STATE_IN_PROGRESS;
while (rval == TRANSIT_STATE_IN_PROGRESS)
{
if ((t = d->trans[works]) != NULL)
{
works = t[*p];
rval = TRANSIT_STATE_DONE;
if (works < 0)
works = 0;
}
else if (works < 0)
{
if (p == buf_end)
/* At the moment, it must not happen. */
return TRANSIT_STATE_END_BUFFER;
works = 0;
}
else if (d->fails[works])
{
works = d->fails[works][*p];
rval = TRANSIT_STATE_DONE;
}
else
{
build_state(works, d);
}
}
*next_state = works;
return rval;
}
/* Check whether period can match or not in the current context. If it can,
return the amount of the bytes with which period can match, otherwise
return 0.
`pos' is the position of the period. `index' is the index from the
buf_begin, and it is the current position in the buffer. */
static int
match_anychar (struct dfa *d, int s, position pos, int index)
{
int newline = 0;
int letter = 0;
wchar_t wc;
int mbclen;
wc = inputwcs[index];
mbclen = (mblen_buf[index] == 0)? 1 : mblen_buf[index];
/* Check context. */
if (wc == (wchar_t)eolbyte)
{
if (!(syntax_bits & RE_DOT_NEWLINE))
return 0;
newline = 1;
}
else if (wc == (wchar_t)'\0')
{
if (syntax_bits & RE_DOT_NOT_NULL)
return 0;
newline = 1;
}
if (iswalnum(wc) || wc == L'_')
letter = 1;
if (!SUCCEEDS_IN_CONTEXT(pos.constraint, d->states[s].newline,
newline, d->states[s].letter, letter))
return 0;
return mbclen;
}
/* Check whether bracket expression can match or not in the current context.
If it can, return the amount of the bytes with which expression can match,
otherwise return 0.
`pos' is the position of the bracket expression. `index' is the index
from the buf_begin, and it is the current position in the buffer. */
int
match_mb_charset (struct dfa *d, int s, position pos, int index)
{
int i;
int match; /* Flag which represent that matching succeed. */
int match_len; /* Length of the character (or collating element)
with which this operator match. */
int op_len; /* Length of the operator. */
char buffer[128];
wchar_t wcbuf[6];
/* Pointer to the structure to which we are currently refering. */
struct mb_char_classes *work_mbc;
int newline = 0;
int letter = 0;
wchar_t wc; /* Current refering character. */
wc = inputwcs[index];
/* Check context. */
if (wc == (wchar_t)eolbyte)
{
if (!(syntax_bits & RE_DOT_NEWLINE))
return 0;
newline = 1;
}
else if (wc == (wchar_t)'\0')
{
if (syntax_bits & RE_DOT_NOT_NULL)
return 0;
newline = 1;
}
if (iswalnum(wc) || wc == L'_')
letter = 1;
if (!SUCCEEDS_IN_CONTEXT(pos.constraint, d->states[s].newline,
newline, d->states[s].letter, letter))
return 0;
/* Assign the current refering operator to work_mbc. */
work_mbc = &(d->mbcsets[(d->multibyte_prop[pos.index]) >> 2]);
match = !work_mbc->invert;
match_len = (mblen_buf[index] == 0)? 1 : mblen_buf[index];
/* match with a character class? */
for (i = 0; i<work_mbc->nch_classes; i++)
{
if (iswctype((wint_t)wc, work_mbc->ch_classes[i]))
goto charset_matched;
}
strncpy(buffer, buf_begin + index, match_len);
buffer[match_len] = '\0';
/* match with an equivalent class? */
for (i = 0; i<work_mbc->nequivs; i++)
{
op_len = strlen(work_mbc->equivs[i]);
strncpy(buffer, buf_begin + index, op_len);
buffer[op_len] = '\0';
if (strcoll(work_mbc->equivs[i], buffer) == 0)
{
match_len = op_len;
goto charset_matched;
}
}
/* match with a collating element? */
for (i = 0; i<work_mbc->ncoll_elems; i++)
{
op_len = strlen(work_mbc->coll_elems[i]);
strncpy(buffer, buf_begin + index, op_len);
buffer[op_len] = '\0';
if (strcoll(work_mbc->coll_elems[i], buffer) == 0)
{
match_len = op_len;
goto charset_matched;
}
}
wcbuf[0] = wc;
wcbuf[1] = wcbuf[3] = wcbuf[5] = '\0';
/* match with a range? */
for (i = 0; i<work_mbc->nranges; i++)
{
wcbuf[2] = work_mbc->range_sts[i];
wcbuf[4] = work_mbc->range_ends[i];
if (wcscoll(wcbuf, wcbuf+2) >= 0 &&
wcscoll(wcbuf+4, wcbuf) >= 0)
goto charset_matched;
}
/* match with a character? */
for (i = 0; i<work_mbc->nchars; i++)
{
if (wc == work_mbc->chars[i])
goto charset_matched;
}
match = !match;
charset_matched:
return match ? match_len : 0;
}
/* Check each of `d->states[s].mbps.elem' can match or not. Then return the
array which corresponds to `d->states[s].mbps.elem' and each element of
the array contains the amount of the bytes with which the element can
match.
`index' is the index from the buf_begin, and it is the current position
in the buffer.
Caller MUST free the array which this function return. */
static int*
check_matching_with_multibyte_ops (struct dfa *d, int s, int index)
{
int i;
int* rarray;
MALLOC(rarray, int, d->states[s].mbps.nelem);
for (i = 0; i < d->states[s].mbps.nelem; ++i)
{
position pos = d->states[s].mbps.elems[i];
switch(d->tokens[pos.index])
{
case ANYCHAR:
rarray[i] = match_anychar(d, s, pos, index);
break;
case MBCSET:
rarray[i] = match_mb_charset(d, s, pos, index);
break;
default:
break; /* can not happen. */
}
}
return rarray;
}
/* Consume a single character and enumerate all of the positions which can
be next position from the state `s'.
`match_lens' is the input. It can be NULL, but it can also be the output
of check_matching_with_multibyte_ops() for optimization.
`mbclen' and `pps' are the output. `mbclen' is the length of the
character consumed, and `pps' is the set this function enumerate. */
static status_transit_state
transit_state_consume_1char (struct dfa *d, int s, unsigned char const **pp,
int *match_lens, int *mbclen, position_set *pps)
{
int i, j;
int s1, s2;
int* work_mbls;
status_transit_state rs = TRANSIT_STATE_DONE;
/* Calculate the length of the (single/multi byte) character
to which p points. */
*mbclen = (mblen_buf[*pp - buf_begin] == 0)? 1
: mblen_buf[*pp - buf_begin];
/* Calculate the state which can be reached from the state `s' by
consuming `*mbclen' single bytes from the buffer. */
s1 = s;
for (i = 0; i < *mbclen; i++)
{
s2 = s1;
rs = transit_state_singlebyte(d, s2, (*pp)++, &s1);
}
/* Copy the positions contained by `s1' to the set `pps'. */
copy(&(d->states[s1].elems), pps);
/* Check (inputed)match_lens, and initialize if it is NULL. */
if (match_lens == NULL && d->states[s].mbps.nelem != 0)
work_mbls = check_matching_with_multibyte_ops(d, s, *pp - buf_begin);
else
work_mbls = match_lens;
/* Add all of the positions which can be reached from `s' by consuming
a single character. */
for (i = 0; i < d->states[s].mbps.nelem ; i++)
{
if (work_mbls[i] == *mbclen)
for (j = 0; j < d->follows[d->states[s].mbps.elems[i].index].nelem;
j++)
insert(d->follows[d->states[s].mbps.elems[i].index].elems[j],
pps);
}
if (match_lens == NULL && work_mbls != NULL)
free(work_mbls);
return rs;
}
/* Transit state from s, then return new state and update the pointer of the
buffer. This function is for some operator which can match with a multi-
byte character or a collating element (which may be multi characters). */
static int
transit_state (struct dfa *d, int s, unsigned char const **pp)
{
int s1;
int mbclen; /* The length of current input multibyte character. */
int maxlen = 0;
int i, j;
int *match_lens = NULL;
int nelem = d->states[s].mbps.nelem; /* Just a alias. */
position_set follows;
unsigned char const *p1 = *pp;
status_transit_state rs;
wchar_t wc;
if (nelem > 0)
/* This state has (a) multibyte operator(s).
We check whether each of them can match or not. */
{
/* Note: caller must free the return value of this function. */
match_lens = check_matching_with_multibyte_ops(d, s, *pp - buf_begin);
for (i = 0; i < nelem; i++)
/* Search the operator which match the longest string,
in this state. */
{
if (match_lens[i] > maxlen)
maxlen = match_lens[i];
}
}
if (nelem == 0 || maxlen == 0)
/* This state has no multibyte operator which can match.
We need to check only one single byte character. */
{
status_transit_state rs;
rs = transit_state_singlebyte(d, s, *pp, &s1);
/* We must update the pointer if state transition succeeded. */
if (rs == TRANSIT_STATE_DONE)
++*pp;
if (match_lens != NULL)
free(match_lens);
return s1;
}
/* This state has some operators which can match a multibyte character. */
follows.nelem = 0;
MALLOC(follows.elems, position, d->nleaves);
/* `maxlen' may be longer than the length of a character, because it may
not be a character but a (multi character) collating element.
We enumerate all of the positions which `s' can reach by consuming
`maxlen' bytes. */
rs = transit_state_consume_1char(d, s, pp, match_lens, &mbclen, &follows);
wc = inputwcs[*pp - mbclen - buf_begin];
s1 = state_index(d, &follows, wc == L'\n', iswalnum(wc));
realloc_trans_if_necessary(d, s1);
while (*pp - p1 < maxlen)
{
follows.nelem = 0;
rs = transit_state_consume_1char(d, s1, pp, NULL, &mbclen, &follows);
for (i = 0; i < nelem ; i++)
{
if (match_lens[i] == *pp - p1)
for (j = 0;
j < d->follows[d->states[s1].mbps.elems[i].index].nelem; j++)
insert(d->follows[d->states[s1].mbps.elems[i].index].elems[j],
&follows);
}
wc = inputwcs[*pp - mbclen - buf_begin];
s1 = state_index(d, &follows, wc == L'\n', iswalnum(wc));
realloc_trans_if_necessary(d, s1);
}
free(match_lens);
free(follows.elems);
return s1;
}
#endif /* MBS_SUPPORT */
/* Search through a buffer looking for a match to the given struct dfa.
Find the first occurrence of a string matching the regexp in the buffer,
and the shortest possible version thereof. Return a pointer to the first
character after the match, or NULL if none is found. Begin points to
the beginning of the buffer, and end points to the first character after
its end. We store a newline in *end to act as a sentinel, so end had
better point somewhere valid. Newline is a flag indicating whether to
allow newlines to be in the matching string. If count is non-
NULL it points to a place we're supposed to increment every time we
see a newline. Finally, if backref is non-NULL it points to a place
where we're supposed to store a 1 if backreferencing happened and the
match needs to be verified by a backtracking matcher. Otherwise
we store a 0 in *backref. */
char *
dfaexec (struct dfa *d, char const *begin, char *end,
int newline, int *count, int *backref)
{
register int s, s1, tmp; /* Current state. */
register unsigned char const *p; /* Current input character. */
register int **trans, *t; /* Copy of d->trans so it can be optimized
into a register. */
register unsigned char eol = eolbyte; /* Likewise for eolbyte. */
static int sbit[NOTCHAR]; /* Table for anding with d->success. */
static int sbit_init;
if (! sbit_init)
{
int i;
sbit_init = 1;
for (i = 0; i < NOTCHAR; ++i)
sbit[i] = (IS_WORD_CONSTITUENT(i)) ? 2 : 1;
sbit[eol] = 4;
}
if (! d->tralloc)
build_state_zero(d);
s = s1 = 0;
p = (unsigned char const *) begin;
trans = d->trans;
*end = eol;
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
int remain_bytes, i;
buf_begin = begin;
buf_end = end;
/* initialize mblen_buf, and inputwcs. */
MALLOC(mblen_buf, unsigned char, end - begin + 2);
MALLOC(inputwcs, wchar_t, end - begin + 2);
memset(&mbs, 0, sizeof(mbstate_t));
remain_bytes = 0;
for (i = 0; i < end - begin + 1; i++)
{
if (remain_bytes == 0)
{
remain_bytes
= mbrtowc(inputwcs + i, begin + i, end - begin - i + 1, &mbs);
if (remain_bytes < 1
|| (remain_bytes == 1 && inputwcs[i] == (wchar_t)begin[i]))
{
remain_bytes = 0;
inputwcs[i] = (wchar_t)begin[i];
mblen_buf[i] = 0;
}
else
{
mblen_buf[i] = remain_bytes;
remain_bytes--;
}
}
else
{
mblen_buf[i] = remain_bytes;
inputwcs[i] = 0;
remain_bytes--;
}
}
mblen_buf[i] = 0;
inputwcs[i] = 0; /* sentinel */
}
#endif /* MBS_SUPPORT */
for (;;)
{
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
while ((t = trans[s]))
{
if ((char *) p > end)
break;
s1 = s;
if (d->states[s].mbps.nelem != 0)
{
/* Can match with a multibyte character (and multi character
collating element). */
unsigned char const *nextp;
SKIP_REMAINS_MB_IF_INITIAL_STATE(s, p);
nextp = p;
s = transit_state(d, s, &nextp);
p = (unsigned char *)nextp;
/* Trans table might be updated. */
trans = d->trans;
}
else
{
SKIP_REMAINS_MB_IF_INITIAL_STATE(s, p);
s = t[*p++];
}
}
else
#endif /* MBS_SUPPORT */
while ((t = trans[s]) != 0) { /* hand-optimized loop */
s1 = t[*p++];
if ((t = trans[s1]) == 0) {
tmp = s ; s = s1 ; s1 = tmp ; /* swap */
break;
}
s = t[*p++];
}
if (s >= 0 && p <= (unsigned char *) end && d->fails[s])
{
if (d->success[s] & sbit[*p])
{
if (backref)
*backref = (d->states[s].backref != 0);
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
free(mblen_buf);
free(inputwcs);
}
#endif /* MBS_SUPPORT */
return (char *) p;
}
s1 = s;
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
unsigned char const *nextp;
nextp = p;
s = transit_state(d, s, &nextp);
p = (unsigned char *)nextp;
/* Trans table might be updated. */
trans = d->trans;
}
else
#endif /* MBS_SUPPORT */
s = d->fails[s][*p++];
continue;
}
/* If the previous character was a newline, count it. */
if (count && (char *) p <= end && p[-1] == eol)
++*count;
/* Check if we've run off the end of the buffer. */
if ((char *) p > end)
{
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
free(mblen_buf);
free(inputwcs);
}
#endif /* MBS_SUPPORT */
return NULL;
}
if (s >= 0)
{
build_state(s, d);
trans = d->trans;
continue;
}
if (p[-1] == eol && newline)
{
s = d->newlines[s1];
continue;
}
s = 0;
}
}
/* Initialize the components of a dfa that the other routines don't
initialize for themselves. */
void
dfainit (struct dfa *d)
{
d->calloc = 1;
MALLOC(d->charclasses, charclass, d->calloc);
d->cindex = 0;
d->talloc = 1;
MALLOC(d->tokens, token, d->talloc);
d->tindex = d->depth = d->nleaves = d->nregexps = 0;
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
d->nmultibyte_prop = 1;
MALLOC(d->multibyte_prop, int, d->nmultibyte_prop);
d->nmbcsets = 0;
d->mbcsets_alloc = 1;
MALLOC(d->mbcsets, struct mb_char_classes, d->mbcsets_alloc);
}
#endif
d->searchflag = 0;
d->tralloc = 0;
d->musts = 0;
d->realtrans = 0;
d->fails = 0;
d->newlines = 0;
d->success = 0;
#ifdef GAWK
d->broken = 0;
#endif
}
/* Parse and analyze a single string of the given length. */
void
dfacomp (char const *s, size_t len, struct dfa *d, int searchflag)
{
if (case_fold && len) /* dummy folding in service of dfamust() */
{
char *lcopy;
int i;
lcopy = malloc(len);
if (!lcopy)
dfaerror(_("memory exhausted"));
/* This is a kludge. */
case_fold = 0;
for (i = 0; i < len; ++i)
if (ISUPPER ((unsigned char) s[i]))
lcopy[i] = tolower ((unsigned char) s[i]);
else
lcopy[i] = s[i];
dfainit(d);
dfaparse(lcopy, len, d);
free(lcopy);
dfamust(d);
d->cindex = d->tindex = d->depth = d->nleaves = d->nregexps = 0;
case_fold = 1;
dfaparse(s, len, d);
dfaanalyze(d, searchflag);
}
else
{
dfainit(d);
dfaparse(s, len, d);
dfamust(d);
dfaanalyze(d, searchflag);
}
}
/* Free the storage held by the components of a dfa. */
void
dfafree (struct dfa *d)
{
int i;
struct dfamust *dm, *ndm;
free((ptr_t) d->charclasses);
free((ptr_t) d->tokens);
#ifdef MBS_SUPPORT
if (MB_CUR_MAX > 1)
{
free((ptr_t) d->multibyte_prop);
for (i = 0; i < d->nmbcsets; ++i)
{
int j;
struct mb_char_classes *p = &(d->mbcsets[i]);
if (p->chars != NULL)
free(p->chars);
if (p->ch_classes != NULL)
free(p->ch_classes);
if (p->range_sts != NULL)
free(p->range_sts);
if (p->range_ends != NULL)
free(p->range_ends);
for (j = 0; j < p->nequivs; ++j)
free(p->equivs[j]);
if (p->equivs != NULL)
free(p->equivs);
for (j = 0; j < p->ncoll_elems; ++j)
free(p->coll_elems[j]);
if (p->coll_elems != NULL)
free(p->coll_elems);
}
free((ptr_t) d->mbcsets);
}
#endif /* MBS_SUPPORT */
for (i = 0; i < d->sindex; ++i) {
free((ptr_t) d->states[i].elems.elems);
#ifdef MBS_SUPPORT
if (d->states[i].mbps.nelem > 0)
free((ptr_t) d->states[i].mbps.elems);
#endif /* MBS_SUPPORT */
}
free((ptr_t) d->states);
for (i = 0; i < d->tindex; ++i)
if (d->follows[i].elems)
free((ptr_t) d->follows[i].elems);
free((ptr_t) d->follows);
for (i = 0; i < d->tralloc; ++i)
if (d->trans[i])
free((ptr_t) d->trans[i]);
else if (d->fails[i])
free((ptr_t) d->fails[i]);
if (d->realtrans) free((ptr_t) d->realtrans);
if (d->fails) free((ptr_t) d->fails);
if (d->newlines) free((ptr_t) d->newlines);
if (d->success) free((ptr_t) d->success);
for (dm = d->musts; dm; dm = ndm)
{
ndm = dm->next;
free(dm->must);
free((ptr_t) dm);
}
}
/* Having found the postfix representation of the regular expression,
try to find a long sequence of characters that must appear in any line
containing the r.e.
Finding a "longest" sequence is beyond the scope here;
we take an easy way out and hope for the best.
(Take "(ab|a)b"--please.)
We do a bottom-up calculation of sequences of characters that must appear
in matches of r.e.'s represented by trees rooted at the nodes of the postfix
representation:
sequences that must appear at the left of the match ("left")
sequences that must appear at the right of the match ("right")
lists of sequences that must appear somewhere in the match ("in")
sequences that must constitute the match ("is")
When we get to the root of the tree, we use one of the longest of its
calculated "in" sequences as our answer. The sequence we find is returned in
d->must (where "d" is the single argument passed to "dfamust");
the length of the sequence is returned in d->mustn.
The sequences calculated for the various types of node (in pseudo ANSI c)
are shown below. "p" is the operand of unary operators (and the left-hand
operand of binary operators); "q" is the right-hand operand of binary
operators.
"ZERO" means "a zero-length sequence" below.
Type left right is in
---- ---- ----- -- --
char c # c # c # c # c
ANYCHAR ZERO ZERO ZERO ZERO
MBCSET ZERO ZERO ZERO ZERO
CSET ZERO ZERO ZERO ZERO
STAR ZERO ZERO ZERO ZERO
QMARK ZERO ZERO ZERO ZERO
PLUS p->left p->right ZERO p->in
CAT (p->is==ZERO)? (q->is==ZERO)? (p->is!=ZERO && p->in plus
p->left : q->right : q->is!=ZERO) ? q->in plus
p->is##q->left p->right##q->is p->is##q->is : p->right##q->left
ZERO
OR longest common longest common (do p->is and substrings common to
leading trailing q->is have same p->in and q->in
(sub)sequence (sub)sequence length and
of p->left of p->right content) ?
and q->left and q->right p->is : NULL
If there's anything else we recognize in the tree, all four sequences get set
to zero-length sequences. If there's something we don't recognize in the tree,
we just return a zero-length sequence.
Break ties in favor of infrequent letters (choosing 'zzz' in preference to
'aaa')?
And. . .is it here or someplace that we might ponder "optimizations" such as
egrep 'psi|epsilon' -> egrep 'psi'
egrep 'pepsi|epsilon' -> egrep 'epsi'
(Yes, we now find "epsi" as a "string
that must occur", but we might also
simplify the *entire* r.e. being sought)
grep '[c]' -> grep 'c'
grep '(ab|a)b' -> grep 'ab'
grep 'ab*' -> grep 'a'
grep 'a*b' -> grep 'b'
There are several issues:
Is optimization easy (enough)?
Does optimization actually accomplish anything,
or is the automaton you get from "psi|epsilon" (for example)
the same as the one you get from "psi" (for example)?
Are optimizable r.e.'s likely to be used in real-life situations
(something like 'ab*' is probably unlikely; something like is
'psi|epsilon' is likelier)? */
static char *
icatalloc (char *old, char *new)
{
char *result;
size_t oldsize, newsize;
newsize = (new == NULL) ? 0 : strlen(new);
if (old == NULL)
oldsize = 0;
else if (newsize == 0)
return old;
else oldsize = strlen(old);
if (old == NULL)
result = (char *) malloc(newsize + 1);
else
result = (char *) realloc((void *) old, oldsize + newsize + 1);
if (result != NULL && new != NULL)
(void) strcpy(result + oldsize, new);
return result;
}
static char *
icpyalloc (char *string)
{
return icatalloc((char *) NULL, string);
}
static char *
istrstr (char *lookin, char *lookfor)
{
char *cp;
size_t len;
len = strlen(lookfor);
for (cp = lookin; *cp != '\0'; ++cp)
if (strncmp(cp, lookfor, len) == 0)
return cp;
return NULL;
}
static void
ifree (char *cp)
{
if (cp != NULL)
free(cp);
}
static void
freelist (char **cpp)
{
int i;
if (cpp == NULL)
return;
for (i = 0; cpp[i] != NULL; ++i)
{
free(cpp[i]);
cpp[i] = NULL;
}
}
static char **
enlist (char **cpp, char *new, size_t len)
{
int i, j;
if (cpp == NULL)
return NULL;
if ((new = icpyalloc(new)) == NULL)
{
freelist(cpp);
return NULL;
}
new[len] = '\0';
/* Is there already something in the list that's new (or longer)? */
for (i = 0; cpp[i] != NULL; ++i)
if (istrstr(cpp[i], new) != NULL)
{
free(new);
return cpp;
}
/* Eliminate any obsoleted strings. */
j = 0;
while (cpp[j] != NULL)
if (istrstr(new, cpp[j]) == NULL)
++j;
else
{
free(cpp[j]);
if (--i == j)
break;
cpp[j] = cpp[i];
cpp[i] = NULL;
}
/* Add the new string. */
cpp = (char **) realloc((char *) cpp, (i + 2) * sizeof *cpp);
if (cpp == NULL)
return NULL;
cpp[i] = new;
cpp[i + 1] = NULL;
return cpp;
}
/* Given pointers to two strings, return a pointer to an allocated
list of their distinct common substrings. Return NULL if something
seems wild. */
static char **
comsubs (char *left, char *right)
{
char **cpp;
char *lcp;
char *rcp;
size_t i, len;
if (left == NULL || right == NULL)
return NULL;
cpp = (char **) malloc(sizeof *cpp);
if (cpp == NULL)
return NULL;
cpp[0] = NULL;
for (lcp = left; *lcp != '\0'; ++lcp)
{
len = 0;
rcp = strchr (right, *lcp);
while (rcp != NULL)
{
for (i = 1; lcp[i] != '\0' && lcp[i] == rcp[i]; ++i)
continue;
if (i > len)
len = i;
rcp = strchr (rcp + 1, *lcp);
}
if (len == 0)
continue;
if ((cpp = enlist(cpp, lcp, len)) == NULL)
break;
}
return cpp;
}
static char **
addlists (char **old, char **new)
{
int i;
if (old == NULL || new == NULL)
return NULL;
for (i = 0; new[i] != NULL; ++i)
{
old = enlist(old, new[i], strlen(new[i]));
if (old == NULL)
break;
}
return old;
}
/* Given two lists of substrings, return a new list giving substrings
common to both. */
static char **
inboth (char **left, char **right)
{
char **both;
char **temp;
int lnum, rnum;
if (left == NULL || right == NULL)
return NULL;
both = (char **) malloc(sizeof *both);
if (both == NULL)
return NULL;
both[0] = NULL;
for (lnum = 0; left[lnum] != NULL; ++lnum)
{
for (rnum = 0; right[rnum] != NULL; ++rnum)
{
temp = comsubs(left[lnum], right[rnum]);
if (temp == NULL)
{
freelist(both);
return NULL;
}
both = addlists(both, temp);
freelist(temp);
free(temp);
if (both == NULL)
return NULL;
}
}
return both;
}
typedef struct
{
char **in;
char *left;
char *right;
char *is;
} must;
static void
resetmust (must *mp)
{
mp->left[0] = mp->right[0] = mp->is[0] = '\0';
freelist(mp->in);
}
static void
dfamust (struct dfa *dfa)
{
must *musts;
must *mp;
char *result;
int ri;
int i;
int exact;
token t;
static must must0;
struct dfamust *dm;
static char empty_string[] = "";
result = empty_string;
exact = 0;
musts = (must *) malloc((dfa->tindex + 1) * sizeof *musts);
if (musts == NULL)
return;
mp = musts;
for (i = 0; i <= dfa->tindex; ++i)
mp[i] = must0;
for (i = 0; i <= dfa->tindex; ++i)
{
mp[i].in = (char **) malloc(sizeof *mp[i].in);
mp[i].left = malloc(2);
mp[i].right = malloc(2);
mp[i].is = malloc(2);
if (mp[i].in == NULL || mp[i].left == NULL ||
mp[i].right == NULL || mp[i].is == NULL)
goto done;
mp[i].left[0] = mp[i].right[0] = mp[i].is[0] = '\0';
mp[i].in[0] = NULL;
}
#ifdef DEBUG
fprintf(stderr, "dfamust:\n");
for (i = 0; i < dfa->tindex; ++i)
{
fprintf(stderr, " %d:", i);
prtok(dfa->tokens[i]);
}
putc('\n', stderr);
#endif
for (ri = 0; ri < dfa->tindex; ++ri)
{
switch (t = dfa->tokens[ri])
{
case LPAREN:
case RPAREN:
goto done; /* "cannot happen" */
case EMPTY:
case BEGLINE:
case ENDLINE:
case BEGWORD:
case ENDWORD:
case LIMWORD:
case NOTLIMWORD:
case BACKREF:
resetmust(mp);
break;
case STAR:
case QMARK:
if (mp <= musts)
goto done; /* "cannot happen" */
--mp;
resetmust(mp);
break;
case OR:
case ORTOP:
if (mp < &musts[2])
goto done; /* "cannot happen" */
{
char **new;
must *lmp;
must *rmp;
int j, ln, rn, n;
rmp = --mp;
lmp = --mp;
/* Guaranteed to be. Unlikely, but. . . */
if (strcmp(lmp->is, rmp->is) != 0)
lmp->is[0] = '\0';
/* Left side--easy */
i = 0;
while (lmp->left[i] != '\0' && lmp->left[i] == rmp->left[i])
++i;
lmp->left[i] = '\0';
/* Right side */
ln = strlen(lmp->right);
rn = strlen(rmp->right);
n = ln;
if (n > rn)
n = rn;
for (i = 0; i < n; ++i)
if (lmp->right[ln - i - 1] != rmp->right[rn - i - 1])
break;
for (j = 0; j < i; ++j)
lmp->right[j] = lmp->right[(ln - i) + j];
lmp->right[j] = '\0';
new = inboth(lmp->in, rmp->in);
if (new == NULL)
goto done;
freelist(lmp->in);
free((char *) lmp->in);
lmp->in = new;
}
break;
case PLUS:
if (mp <= musts)
goto done; /* "cannot happen" */
--mp;
mp->is[0] = '\0';
break;
case END:
if (mp != &musts[1])
goto done; /* "cannot happen" */
for (i = 0; musts[0].in[i] != NULL; ++i)
if (strlen(musts[0].in[i]) > strlen(result))
result = musts[0].in[i];
if (strcmp(result, musts[0].is) == 0)
exact = 1;
goto done;
case CAT:
if (mp < &musts[2])
goto done; /* "cannot happen" */
{
must *lmp;
must *rmp;
rmp = --mp;
lmp = --mp;
/* In. Everything in left, plus everything in
right, plus catenation of
left's right and right's left. */
lmp->in = addlists(lmp->in, rmp->in);
if (lmp->in == NULL)
goto done;
if (lmp->right[0] != '\0' &&
rmp->left[0] != '\0')
{
char *tp;
tp = icpyalloc(lmp->right);
if (tp == NULL)
goto done;
tp = icatalloc(tp, rmp->left);
if (tp == NULL)
goto done;
lmp->in = enlist(lmp->in, tp,
strlen(tp));
free(tp);
if (lmp->in == NULL)
goto done;
}
/* Left-hand */
if (lmp->is[0] != '\0')
{
lmp->left = icatalloc(lmp->left,
rmp->left);
if (lmp->left == NULL)
goto done;
}
/* Right-hand */
if (rmp->is[0] == '\0')
lmp->right[0] = '\0';
lmp->right = icatalloc(lmp->right, rmp->right);
if (lmp->right == NULL)
goto done;
/* Guaranteed to be */
if (lmp->is[0] != '\0' && rmp->is[0] != '\0')
{
lmp->is = icatalloc(lmp->is, rmp->is);
if (lmp->is == NULL)
goto done;
}
else
lmp->is[0] = '\0';
}
break;
default:
if (t < END)
{
/* "cannot happen" */
goto done;
}
else if (t == '\0')
{
/* not on *my* shift */
goto done;
}
else if (t >= CSET
#ifdef MBS_SUPPORT
|| t == ANYCHAR
|| t == MBCSET
#endif /* MBS_SUPPORT */
)
{
/* easy enough */
resetmust(mp);
}
else
{
/* plain character */
resetmust(mp);
mp->is[0] = mp->left[0] = mp->right[0] = t;
mp->is[1] = mp->left[1] = mp->right[1] = '\0';
mp->in = enlist(mp->in, mp->is, (size_t)1);
if (mp->in == NULL)
goto done;
}
break;
}
#ifdef DEBUG
fprintf(stderr, " node: %d:", ri);
prtok(dfa->tokens[ri]);
fprintf(stderr, "\n in:");
for (i = 0; mp->in[i]; ++i)
fprintf(stderr, " \"%s\"", mp->in[i]);
fprintf(stderr, "\n is: \"%s\"\n", mp->is);
fprintf(stderr, " left: \"%s\"\n", mp->left);
fprintf(stderr, " right: \"%s\"\n", mp->right);
#endif
++mp;
}
done:
if (strlen(result))
{
MALLOC(dm, struct dfamust, 1);
dm->exact = exact;
MALLOC(dm->must, char, strlen(result) + 1);
strcpy(dm->must, result);
dm->next = dfa->musts;
dfa->musts = dm;
}
mp = musts;
for (i = 0; i <= dfa->tindex; ++i)
{
freelist(mp[i].in);
ifree((char *) mp[i].in);
ifree(mp[i].left);
ifree(mp[i].right);
ifree(mp[i].is);
}
free((char *) mp);
}
/* vim:set shiftwidth=2: */
|