Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
Overview
Comment: | Copy fts1/ to fts2/, changing reference from fts1 to fts2. For future
reference, the source versions copied were:
README.txt r1.1 fts1.c r1.37 fts1.h r1.2 fts1_hash.c r1.1 fts1_hash.h r1.1 fts1_porter.c r1.1 fts1_tokenizer.h r1.4 fts1_tokenizer1.c r1.6 (CVS 3471) |
---|---|
Downloads: | Tarball | ZIP archive |
Timelines: | family | ancestors | descendants | both | trunk |
Files: | files | file ages | folders |
SHA1: |
d0d1e7cdcc1dd085f1e359ce35c44169 |
User & Date: | shess 2006-10-10 17:37:14.000 |
Context
2006-10-10
| ||
23:22 | Fix leaky symbols. With this change, fts1 and fts2 can both be statically linked. (CVS 3472) (check-in: 5e8bbb85c1 user: shess tags: trunk) | |
17:37 |
Copy fts1/ to fts2/, changing reference from fts1 to fts2. For future
reference, the source versions copied were:
README.txt r1.1 fts1.c r1.37 fts1.h r1.2 fts1_hash.c r1.1 fts1_hash.h r1.1 fts1_porter.c r1.1 fts1_tokenizer.h r1.4 fts1_tokenizer1.c r1.6 (CVS 3471) (check-in: d0d1e7cdcc user: shess tags: trunk) | |
13:07 | VACUUM now uses a temporary file in the official TEMP folder instead of a file in the same directory as the original database. (CVS 3470) (check-in: b743429dd5 user: drh tags: trunk) | |
Changes
Added ext/fts2/README.txt.
> > > > | 1 2 3 4 | This folder contains source code to the second full-text search extension for SQLite. While the API is the same, this version uses a substantially different storage schema from fts1, so tables will need to be rebuilt. |
Added ext/fts2/fts2.c.
> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | /* The author disclaims copyright to this source code. * * This is an SQLite module implementing full-text search. */ /* ** The code in this file is only compiled if: ** ** * The FTS2 module is being built as an extension ** (in which case SQLITE_CORE is not defined), or ** ** * The FTS2 module is being built into the core of ** SQLite (in which case SQLITE_ENABLE_FTS2 is defined). */ #if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) #if defined(SQLITE_ENABLE_FTS2) && !defined(SQLITE_CORE) # define SQLITE_CORE 1 #endif #include <assert.h> #if !defined(__APPLE__) #include <malloc.h> #else #include <stdlib.h> #endif #include <stdio.h> #include <string.h> #include <ctype.h> #include "fts2.h" #include "fts2_hash.h" #include "fts2_tokenizer.h" #include "sqlite3.h" #include "sqlite3ext.h" SQLITE_EXTENSION_INIT1 #if 0 # define TRACE(A) printf A; fflush(stdout) #else # define TRACE(A) #endif /* utility functions */ typedef struct StringBuffer { int len; /* length, not including null terminator */ int alloced; /* Space allocated for s[] */ char *s; /* Content of the string */ } StringBuffer; void initStringBuffer(StringBuffer *sb){ sb->len = 0; sb->alloced = 100; sb->s = malloc(100); sb->s[0] = '\0'; } void nappend(StringBuffer *sb, const char *zFrom, int nFrom){ if( sb->len + nFrom >= sb->alloced ){ sb->alloced = sb->len + nFrom + 100; sb->s = realloc(sb->s, sb->alloced+1); if( sb->s==0 ){ initStringBuffer(sb); return; } } memcpy(sb->s + sb->len, zFrom, nFrom); sb->len += nFrom; sb->s[sb->len] = 0; } void append(StringBuffer *sb, const char *zFrom){ nappend(sb, zFrom, strlen(zFrom)); } /* We encode variable-length integers in little-endian order using seven bits * per byte as follows: ** ** KEY: ** A = 0xxxxxxx 7 bits of data and one flag bit ** B = 1xxxxxxx 7 bits of data and one flag bit ** ** 7 bits - A ** 14 bits - BA ** 21 bits - BBA ** and so on. */ /* We may need up to VARINT_MAX bytes to store an encoded 64-bit integer. */ #define VARINT_MAX 10 /* Write a 64-bit variable-length integer to memory starting at p[0]. * The length of data written will be between 1 and VARINT_MAX bytes. * The number of bytes written is returned. */ static int putVarint(char *p, sqlite_int64 v){ unsigned char *q = (unsigned char *) p; sqlite_uint64 vu = v; do{ *q++ = (unsigned char) ((vu & 0x7f) | 0x80); vu >>= 7; }while( vu!=0 ); q[-1] &= 0x7f; /* turn off high bit in final byte */ assert( q - (unsigned char *)p <= VARINT_MAX ); return (int) (q - (unsigned char *)p); } /* Read a 64-bit variable-length integer from memory starting at p[0]. * Return the number of bytes read, or 0 on error. * The value is stored in *v. */ static int getVarint(const char *p, sqlite_int64 *v){ const unsigned char *q = (const unsigned char *) p; sqlite_uint64 x = 0, y = 1; while( (*q & 0x80) == 0x80 ){ x += y * (*q++ & 0x7f); y <<= 7; if( q - (unsigned char *)p >= VARINT_MAX ){ /* bad data */ assert( 0 ); return 0; } } x += y * (*q++); *v = (sqlite_int64) x; return (int) (q - (unsigned char *)p); } static int getVarint32(const char *p, int *pi){ sqlite_int64 i; int ret = getVarint(p, &i); *pi = (int) i; assert( *pi==i ); return ret; } /*** Document lists *** * * A document list holds a sorted list of varint-encoded document IDs. * * A doclist with type DL_POSITIONS_OFFSETS is stored like this: * * array { * varint docid; * array { * varint position; (delta from previous position plus POS_BASE) * varint startOffset; (delta from previous startOffset) * varint endOffset; (delta from startOffset) * } * } * * Here, array { X } means zero or more occurrences of X, adjacent in memory. * * A position list may hold positions for text in multiple columns. A position * POS_COLUMN is followed by a varint containing the index of the column for * following positions in the list. Any positions appearing before any * occurrences of POS_COLUMN are for column 0. * * A doclist with type DL_POSITIONS is like the above, but holds only docids * and positions without offset information. * * A doclist with type DL_DOCIDS is like the above, but holds only docids * without positions or offset information. * * On disk, every document list has positions and offsets, so we don't bother * to serialize a doclist's type. * * We don't yet delta-encode document IDs; doing so will probably be a * modest win. * * NOTE(shess) I've thought of a slightly (1%) better offset encoding. * After the first offset, estimate the next offset by using the * current token position and the previous token position and offset, * offset to handle some variance. So the estimate would be * (iPosition*w->iStartOffset/w->iPosition-64), which is delta-encoded * as normal. Offsets more than 64 chars from the estimate are * encoded as the delta to the previous start offset + 128. An * additional tiny increment can be gained by using the end offset of * the previous token to make the estimate a tiny bit more precise. */ typedef enum DocListType { DL_DOCIDS, /* docids only */ DL_POSITIONS, /* docids + positions */ DL_POSITIONS_OFFSETS /* docids + positions + offsets */ } DocListType; /* ** By default, only positions and not offsets are stored in the doclists. ** To change this so that offsets are stored too, compile with ** ** -DDL_DEFAULT=DL_POSITIONS_OFFSETS ** */ #ifndef DL_DEFAULT # define DL_DEFAULT DL_POSITIONS #endif typedef struct DocList { char *pData; int nData; DocListType iType; int iLastColumn; /* the last column written */ int iLastPos; /* the last position written */ int iLastOffset; /* the last start offset written */ } DocList; enum { POS_END = 0, /* end of this position list */ POS_COLUMN, /* followed by new column number */ POS_BASE }; /* Initialize a new DocList to hold the given data. */ static void docListInit(DocList *d, DocListType iType, const char *pData, int nData){ d->nData = nData; if( nData>0 ){ d->pData = malloc(nData); memcpy(d->pData, pData, nData); } else { d->pData = NULL; } d->iType = iType; d->iLastColumn = 0; d->iLastPos = d->iLastOffset = 0; } /* Create a new dynamically-allocated DocList. */ static DocList *docListNew(DocListType iType){ DocList *d = (DocList *) malloc(sizeof(DocList)); docListInit(d, iType, 0, 0); return d; } static void docListDestroy(DocList *d){ free(d->pData); #ifndef NDEBUG memset(d, 0x55, sizeof(*d)); #endif } static void docListDelete(DocList *d){ docListDestroy(d); free(d); } static char *docListEnd(DocList *d){ return d->pData + d->nData; } /* Append a varint to a DocList's data. */ static void appendVarint(DocList *d, sqlite_int64 i){ char c[VARINT_MAX]; int n = putVarint(c, i); d->pData = realloc(d->pData, d->nData + n); memcpy(d->pData + d->nData, c, n); d->nData += n; } static void docListAddDocid(DocList *d, sqlite_int64 iDocid){ appendVarint(d, iDocid); if( d->iType>=DL_POSITIONS ){ appendVarint(d, POS_END); /* initially empty position list */ d->iLastColumn = 0; d->iLastPos = d->iLastOffset = 0; } } /* helper function for docListAddPos and docListAddPosOffset */ static void addPos(DocList *d, int iColumn, int iPos){ assert( d->nData>0 ); --d->nData; /* remove previous terminator */ if( iColumn!=d->iLastColumn ){ assert( iColumn>d->iLastColumn ); appendVarint(d, POS_COLUMN); appendVarint(d, iColumn); d->iLastColumn = iColumn; d->iLastPos = d->iLastOffset = 0; } assert( iPos>=d->iLastPos ); appendVarint(d, iPos-d->iLastPos+POS_BASE); d->iLastPos = iPos; } /* Add a position to the last position list in a doclist. */ static void docListAddPos(DocList *d, int iColumn, int iPos){ assert( d->iType==DL_POSITIONS ); addPos(d, iColumn, iPos); appendVarint(d, POS_END); /* add new terminator */ } /* ** Add a position and starting and ending offsets to a doclist. ** ** If the doclist is setup to handle only positions, then insert ** the position only and ignore the offsets. */ static void docListAddPosOffset( DocList *d, /* Doclist under construction */ int iColumn, /* Column the inserted term is part of */ int iPos, /* Position of the inserted term */ int iStartOffset, /* Starting offset of inserted term */ int iEndOffset /* Ending offset of inserted term */ ){ assert( d->iType>=DL_POSITIONS ); addPos(d, iColumn, iPos); if( d->iType==DL_POSITIONS_OFFSETS ){ assert( iStartOffset>=d->iLastOffset ); appendVarint(d, iStartOffset-d->iLastOffset); d->iLastOffset = iStartOffset; assert( iEndOffset>=iStartOffset ); appendVarint(d, iEndOffset-iStartOffset); } appendVarint(d, POS_END); /* add new terminator */ } /* ** A DocListReader object is a cursor into a doclist. Initialize ** the cursor to the beginning of the doclist by calling readerInit(). ** Then use routines ** ** peekDocid() ** readDocid() ** readPosition() ** skipPositionList() ** and so forth... ** ** to read information out of the doclist. When we reach the end ** of the doclist, atEnd() returns TRUE. */ typedef struct DocListReader { DocList *pDoclist; /* The document list we are stepping through */ char *p; /* Pointer to next unread byte in the doclist */ int iLastColumn; int iLastPos; /* the last position read, or -1 when not in a position list */ } DocListReader; /* ** Initialize the DocListReader r to point to the beginning of pDoclist. */ static void readerInit(DocListReader *r, DocList *pDoclist){ r->pDoclist = pDoclist; if( pDoclist!=NULL ){ r->p = pDoclist->pData; } r->iLastColumn = -1; r->iLastPos = -1; } /* ** Return TRUE if we have reached then end of pReader and there is ** nothing else left to read. */ static int atEnd(DocListReader *pReader){ return pReader->pDoclist==0 || (pReader->p >= docListEnd(pReader->pDoclist)); } /* Peek at the next docid without advancing the read pointer. */ static sqlite_int64 peekDocid(DocListReader *pReader){ sqlite_int64 ret; assert( !atEnd(pReader) ); assert( pReader->iLastPos==-1 ); getVarint(pReader->p, &ret); return ret; } /* Read the next docid. See also nextDocid(). */ static sqlite_int64 readDocid(DocListReader *pReader){ sqlite_int64 ret; assert( !atEnd(pReader) ); assert( pReader->iLastPos==-1 ); pReader->p += getVarint(pReader->p, &ret); if( pReader->pDoclist->iType>=DL_POSITIONS ){ pReader->iLastColumn = 0; pReader->iLastPos = 0; } return ret; } /* Read the next position and column index from a position list. * Returns the position, or -1 at the end of the list. */ static int readPosition(DocListReader *pReader, int *iColumn){ int i; int iType = pReader->pDoclist->iType; if( pReader->iLastPos==-1 ){ return -1; } assert( !atEnd(pReader) ); if( iType<DL_POSITIONS ){ return -1; } pReader->p += getVarint32(pReader->p, &i); if( i==POS_END ){ pReader->iLastColumn = pReader->iLastPos = -1; *iColumn = -1; return -1; } if( i==POS_COLUMN ){ pReader->p += getVarint32(pReader->p, &pReader->iLastColumn); pReader->iLastPos = 0; pReader->p += getVarint32(pReader->p, &i); assert( i>=POS_BASE ); } pReader->iLastPos += ((int) i)-POS_BASE; if( iType>=DL_POSITIONS_OFFSETS ){ /* Skip over offsets, ignoring them for now. */ int iStart, iEnd; pReader->p += getVarint32(pReader->p, &iStart); pReader->p += getVarint32(pReader->p, &iEnd); } *iColumn = pReader->iLastColumn; return pReader->iLastPos; } /* Skip past the end of a position list. */ static void skipPositionList(DocListReader *pReader){ DocList *p = pReader->pDoclist; if( p && p->iType>=DL_POSITIONS ){ int iColumn; while( readPosition(pReader, &iColumn)!=-1 ){} } } /* Skip over a docid, including its position list if the doclist has * positions. */ static void skipDocument(DocListReader *pReader){ readDocid(pReader); skipPositionList(pReader); } /* Skip past all docids which are less than [iDocid]. Returns 1 if a docid * matching [iDocid] was found. */ static int skipToDocid(DocListReader *pReader, sqlite_int64 iDocid){ sqlite_int64 d = 0; while( !atEnd(pReader) && (d=peekDocid(pReader))<iDocid ){ skipDocument(pReader); } return !atEnd(pReader) && d==iDocid; } /* Return the first document in a document list. */ static sqlite_int64 firstDocid(DocList *d){ DocListReader r; readerInit(&r, d); return readDocid(&r); } #ifdef SQLITE_DEBUG /* ** This routine is used for debugging purpose only. ** ** Write the content of a doclist to standard output. */ static void printDoclist(DocList *p){ DocListReader r; const char *zSep = ""; readerInit(&r, p); while( !atEnd(&r) ){ sqlite_int64 docid = readDocid(&r); if( docid==0 ){ skipPositionList(&r); continue; } printf("%s%lld", zSep, docid); zSep = ","; if( p->iType>=DL_POSITIONS ){ int iPos, iCol; const char *zDiv = ""; printf("("); while( (iPos = readPosition(&r, &iCol))>=0 ){ printf("%s%d:%d", zDiv, iCol, iPos); zDiv = ":"; } printf(")"); } } printf("\n"); fflush(stdout); } #endif /* SQLITE_DEBUG */ /* Trim the given doclist to contain only positions in column * [iRestrictColumn]. */ static void docListRestrictColumn(DocList *in, int iRestrictColumn){ DocListReader r; DocList out; assert( in->iType>=DL_POSITIONS ); readerInit(&r, in); docListInit(&out, DL_POSITIONS, NULL, 0); while( !atEnd(&r) ){ sqlite_int64 iDocid = readDocid(&r); int iPos, iColumn; docListAddDocid(&out, iDocid); while( (iPos = readPosition(&r, &iColumn)) != -1 ){ if( iColumn==iRestrictColumn ){ docListAddPos(&out, iColumn, iPos); } } } docListDestroy(in); *in = out; } /* Trim the given doclist by discarding any docids without any remaining * positions. */ static void docListDiscardEmpty(DocList *in) { DocListReader r; DocList out; /* TODO: It would be nice to implement this operation in place; that * could save a significant amount of memory in queries with long doclists. */ assert( in->iType>=DL_POSITIONS ); readerInit(&r, in); docListInit(&out, DL_POSITIONS, NULL, 0); while( !atEnd(&r) ){ sqlite_int64 iDocid = readDocid(&r); int match = 0; int iPos, iColumn; while( (iPos = readPosition(&r, &iColumn)) != -1 ){ if( !match ){ docListAddDocid(&out, iDocid); match = 1; } docListAddPos(&out, iColumn, iPos); } } docListDestroy(in); *in = out; } /* Helper function for docListUpdate() and docListAccumulate(). ** Splices a doclist element into the doclist represented by r, ** leaving r pointing after the newly spliced element. */ static void docListSpliceElement(DocListReader *r, sqlite_int64 iDocid, const char *pSource, int nSource){ DocList *d = r->pDoclist; char *pTarget; int nTarget, found; found = skipToDocid(r, iDocid); /* Describe slice in d to place pSource/nSource. */ pTarget = r->p; if( found ){ skipDocument(r); nTarget = r->p-pTarget; }else{ nTarget = 0; } /* The sense of the following is that there are three possibilities. ** If nTarget==nSource, we should not move any memory nor realloc. ** If nTarget>nSource, trim target and realloc. ** If nTarget<nSource, realloc then expand target. */ if( nTarget>nSource ){ memmove(pTarget+nSource, pTarget+nTarget, docListEnd(d)-(pTarget+nTarget)); } if( nTarget!=nSource ){ int iDoclist = pTarget-d->pData; d->pData = realloc(d->pData, d->nData+nSource-nTarget); pTarget = d->pData+iDoclist; } if( nTarget<nSource ){ memmove(pTarget+nSource, pTarget+nTarget, docListEnd(d)-(pTarget+nTarget)); } memcpy(pTarget, pSource, nSource); d->nData += nSource-nTarget; r->p = pTarget+nSource; } /* Insert/update pUpdate into the doclist. */ static void docListUpdate(DocList *d, DocList *pUpdate){ DocListReader reader; assert( d!=NULL && pUpdate!=NULL ); assert( d->iType==pUpdate->iType); readerInit(&reader, d); docListSpliceElement(&reader, firstDocid(pUpdate), pUpdate->pData, pUpdate->nData); } /* Propagate elements from pUpdate to pAcc, overwriting elements with ** matching docids. */ static void docListAccumulate(DocList *pAcc, DocList *pUpdate){ DocListReader accReader, updateReader; /* Handle edge cases where one doclist is empty. */ assert( pAcc!=NULL ); if( pUpdate==NULL || pUpdate->nData==0 ) return; if( pAcc->nData==0 ){ pAcc->pData = malloc(pUpdate->nData); memcpy(pAcc->pData, pUpdate->pData, pUpdate->nData); pAcc->nData = pUpdate->nData; return; } readerInit(&accReader, pAcc); readerInit(&updateReader, pUpdate); while( !atEnd(&updateReader) ){ char *pSource = updateReader.p; sqlite_int64 iDocid = readDocid(&updateReader); skipPositionList(&updateReader); docListSpliceElement(&accReader, iDocid, pSource, updateReader.p-pSource); } } /* ** Read the next docid off of pIn. Return 0 if we reach the end. * * TODO: This assumes that docids are never 0, but they may actually be 0 since * users can choose docids when inserting into a full-text table. Fix this. */ static sqlite_int64 nextDocid(DocListReader *pIn){ skipPositionList(pIn); return atEnd(pIn) ? 0 : readDocid(pIn); } /* ** pLeft and pRight are two DocListReaders that are pointing to ** positions lists of the same document: iDocid. ** ** If there are no instances in pLeft or pRight where the position ** of pLeft is one less than the position of pRight, then this ** routine adds nothing to pOut. ** ** If there are one or more instances where positions from pLeft ** are exactly one less than positions from pRight, then add a new ** document record to pOut. If pOut wants to hold positions, then ** include the positions from pRight that are one more than a ** position in pLeft. In other words: pRight.iPos==pLeft.iPos+1. ** ** pLeft and pRight are left pointing at the next document record. */ static void mergePosList( DocListReader *pLeft, /* Left position list */ DocListReader *pRight, /* Right position list */ sqlite_int64 iDocid, /* The docid from pLeft and pRight */ DocList *pOut /* Write the merged document record here */ ){ int iLeftCol, iLeftPos = readPosition(pLeft, &iLeftCol); int iRightCol, iRightPos = readPosition(pRight, &iRightCol); int match = 0; /* Loop until we've reached the end of both position lists. */ while( iLeftPos!=-1 && iRightPos!=-1 ){ if( iLeftCol==iRightCol && iLeftPos+1==iRightPos ){ if( !match ){ docListAddDocid(pOut, iDocid); match = 1; } if( pOut->iType>=DL_POSITIONS ){ docListAddPos(pOut, iRightCol, iRightPos); } iLeftPos = readPosition(pLeft, &iLeftCol); iRightPos = readPosition(pRight, &iRightCol); }else if( iRightCol<iLeftCol || (iRightCol==iLeftCol && iRightPos<iLeftPos+1) ){ iRightPos = readPosition(pRight, &iRightCol); }else{ iLeftPos = readPosition(pLeft, &iLeftCol); } } if( iLeftPos>=0 ) skipPositionList(pLeft); if( iRightPos>=0 ) skipPositionList(pRight); } /* We have two doclists: pLeft and pRight. ** Write the phrase intersection of these two doclists into pOut. ** ** A phrase intersection means that two documents only match ** if pLeft.iPos+1==pRight.iPos. ** ** The output pOut may or may not contain positions. If pOut ** does contain positions, they are the positions of pRight. */ static void docListPhraseMerge( DocList *pLeft, /* Doclist resulting from the words on the left */ DocList *pRight, /* Doclist for the next word to the right */ DocList *pOut /* Write the combined doclist here */ ){ DocListReader left, right; sqlite_int64 docidLeft, docidRight; readerInit(&left, pLeft); readerInit(&right, pRight); docidLeft = nextDocid(&left); docidRight = nextDocid(&right); while( docidLeft>0 && docidRight>0 ){ if( docidLeft<docidRight ){ docidLeft = nextDocid(&left); }else if( docidRight<docidLeft ){ docidRight = nextDocid(&right); }else{ mergePosList(&left, &right, docidLeft, pOut); docidLeft = nextDocid(&left); docidRight = nextDocid(&right); } } } /* We have two doclists: pLeft and pRight. ** Write the intersection of these two doclists into pOut. ** Only docids are matched. Position information is ignored. ** ** The output pOut never holds positions. */ static void docListAndMerge( DocList *pLeft, /* Doclist resulting from the words on the left */ DocList *pRight, /* Doclist for the next word to the right */ DocList *pOut /* Write the combined doclist here */ ){ DocListReader left, right; sqlite_int64 docidLeft, docidRight; assert( pOut->iType<DL_POSITIONS ); readerInit(&left, pLeft); readerInit(&right, pRight); docidLeft = nextDocid(&left); docidRight = nextDocid(&right); while( docidLeft>0 && docidRight>0 ){ if( docidLeft<docidRight ){ docidLeft = nextDocid(&left); }else if( docidRight<docidLeft ){ docidRight = nextDocid(&right); }else{ docListAddDocid(pOut, docidLeft); docidLeft = nextDocid(&left); docidRight = nextDocid(&right); } } } /* We have two doclists: pLeft and pRight. ** Write the union of these two doclists into pOut. ** Only docids are matched. Position information is ignored. ** ** The output pOut never holds positions. */ static void docListOrMerge( DocList *pLeft, /* Doclist resulting from the words on the left */ DocList *pRight, /* Doclist for the next word to the right */ DocList *pOut /* Write the combined doclist here */ ){ DocListReader left, right; sqlite_int64 docidLeft, docidRight, priorLeft; readerInit(&left, pLeft); readerInit(&right, pRight); docidLeft = nextDocid(&left); docidRight = nextDocid(&right); while( docidLeft>0 && docidRight>0 ){ if( docidLeft<=docidRight ){ docListAddDocid(pOut, docidLeft); }else{ docListAddDocid(pOut, docidRight); } priorLeft = docidLeft; if( docidLeft<=docidRight ){ docidLeft = nextDocid(&left); } if( docidRight>0 && docidRight<=priorLeft ){ docidRight = nextDocid(&right); } } while( docidLeft>0 ){ docListAddDocid(pOut, docidLeft); docidLeft = nextDocid(&left); } while( docidRight>0 ){ docListAddDocid(pOut, docidRight); docidRight = nextDocid(&right); } } /* We have two doclists: pLeft and pRight. ** Write into pOut all documents that occur in pLeft but not ** in pRight. ** ** Only docids are matched. Position information is ignored. ** ** The output pOut never holds positions. */ static void docListExceptMerge( DocList *pLeft, /* Doclist resulting from the words on the left */ DocList *pRight, /* Doclist for the next word to the right */ DocList *pOut /* Write the combined doclist here */ ){ DocListReader left, right; sqlite_int64 docidLeft, docidRight, priorLeft; readerInit(&left, pLeft); readerInit(&right, pRight); docidLeft = nextDocid(&left); docidRight = nextDocid(&right); while( docidLeft>0 && docidRight>0 ){ priorLeft = docidLeft; if( docidLeft<docidRight ){ docListAddDocid(pOut, docidLeft); } if( docidLeft<=docidRight ){ docidLeft = nextDocid(&left); } if( docidRight>0 && docidRight<=priorLeft ){ docidRight = nextDocid(&right); } } while( docidLeft>0 ){ docListAddDocid(pOut, docidLeft); docidLeft = nextDocid(&left); } } static char *string_dup_n(const char *s, int n){ char *str = malloc(n + 1); memcpy(str, s, n); str[n] = '\0'; return str; } /* Duplicate a string; the caller must free() the returned string. * (We don't use strdup() since it's not part of the standard C library and * may not be available everywhere.) */ static char *string_dup(const char *s){ return string_dup_n(s, strlen(s)); } /* Format a string, replacing each occurrence of the % character with * zName. This may be more convenient than sqlite_mprintf() * when one string is used repeatedly in a format string. * The caller must free() the returned string. */ static char *string_format(const char *zFormat, const char *zName){ const char *p; size_t len = 0; size_t nName = strlen(zName); char *result; char *r; /* first compute length needed */ for(p = zFormat ; *p ; ++p){ len += (*p=='%' ? nName : 1); } len += 1; /* for null terminator */ r = result = malloc(len); for(p = zFormat; *p; ++p){ if( *p=='%' ){ memcpy(r, zName, nName); r += nName; } else { *r++ = *p; } } *r++ = '\0'; assert( r == result + len ); return result; } static int sql_exec(sqlite3 *db, const char *zName, const char *zFormat){ char *zCommand = string_format(zFormat, zName); int rc; TRACE(("FTS2 sql: %s\n", zCommand)); rc = sqlite3_exec(db, zCommand, NULL, 0, NULL); free(zCommand); return rc; } static int sql_prepare(sqlite3 *db, const char *zName, sqlite3_stmt **ppStmt, const char *zFormat){ char *zCommand = string_format(zFormat, zName); int rc; TRACE(("FTS2 prepare: %s\n", zCommand)); rc = sqlite3_prepare(db, zCommand, -1, ppStmt, NULL); free(zCommand); return rc; } /* end utility functions */ /* Forward reference */ typedef struct fulltext_vtab fulltext_vtab; /* A single term in a query is represented by an instances of ** the following structure. */ typedef struct QueryTerm { short int nPhrase; /* How many following terms are part of the same phrase */ short int iPhrase; /* This is the i-th term of a phrase. */ short int iColumn; /* Column of the index that must match this term */ signed char isOr; /* this term is preceded by "OR" */ signed char isNot; /* this term is preceded by "-" */ char *pTerm; /* text of the term. '\000' terminated. malloced */ int nTerm; /* Number of bytes in pTerm[] */ } QueryTerm; /* A query string is parsed into a Query structure. * * We could, in theory, allow query strings to be complicated * nested expressions with precedence determined by parentheses. * But none of the major search engines do this. (Perhaps the * feeling is that an parenthesized expression is two complex of * an idea for the average user to grasp.) Taking our lead from * the major search engines, we will allow queries to be a list * of terms (with an implied AND operator) or phrases in double-quotes, * with a single optional "-" before each non-phrase term to designate * negation and an optional OR connector. * * OR binds more tightly than the implied AND, which is what the * major search engines seem to do. So, for example: * * [one two OR three] ==> one AND (two OR three) * [one OR two three] ==> (one OR two) AND three * * A "-" before a term matches all entries that lack that term. * The "-" must occur immediately before the term with in intervening * space. This is how the search engines do it. * * A NOT term cannot be the right-hand operand of an OR. If this * occurs in the query string, the NOT is ignored: * * [one OR -two] ==> one OR two * */ typedef struct Query { fulltext_vtab *pFts; /* The full text index */ int nTerms; /* Number of terms in the query */ QueryTerm *pTerms; /* Array of terms. Space obtained from malloc() */ int nextIsOr; /* Set the isOr flag on the next inserted term */ int nextColumn; /* Next word parsed must be in this column */ int dfltColumn; /* The default column */ } Query; /* ** An instance of the following structure keeps track of generated ** matching-word offset information and snippets. */ typedef struct Snippet { int nMatch; /* Total number of matches */ int nAlloc; /* Space allocated for aMatch[] */ struct snippetMatch { /* One entry for each matching term */ char snStatus; /* Status flag for use while constructing snippets */ short int iCol; /* The column that contains the match */ short int iTerm; /* The index in Query.pTerms[] of the matching term */ short int nByte; /* Number of bytes in the term */ int iStart; /* The offset to the first character of the term */ } *aMatch; /* Points to space obtained from malloc */ char *zOffset; /* Text rendering of aMatch[] */ int nOffset; /* strlen(zOffset) */ char *zSnippet; /* Snippet text */ int nSnippet; /* strlen(zSnippet) */ } Snippet; typedef enum QueryType { QUERY_GENERIC, /* table scan */ QUERY_ROWID, /* lookup by rowid */ QUERY_FULLTEXT /* QUERY_FULLTEXT + [i] is a full-text search for column i*/ } QueryType; /* TODO(shess) CHUNK_MAX controls how much data we allow in segment 0 ** before we start aggregating into larger segments. Lower CHUNK_MAX ** means that for a given input we have more individual segments per ** term, which means more rows in the table and a bigger index (due to ** both more rows and bigger rowids). But it also reduces the average ** cost of adding new elements to the segment 0 doclist, and it seems ** to reduce the number of pages read and written during inserts. 256 ** was chosen by measuring insertion times for a certain input (first ** 10k documents of Enron corpus), though including query performance ** in the decision may argue for a larger value. */ #define CHUNK_MAX 256 typedef enum fulltext_statement { CONTENT_INSERT_STMT, CONTENT_SELECT_STMT, CONTENT_UPDATE_STMT, CONTENT_DELETE_STMT, TERM_SELECT_STMT, TERM_SELECT_ALL_STMT, TERM_INSERT_STMT, TERM_UPDATE_STMT, TERM_DELETE_STMT, MAX_STMT /* Always at end! */ } fulltext_statement; /* These must exactly match the enum above. */ /* TODO(adam): Is there some risk that a statement (in particular, ** pTermSelectStmt) will be used in two cursors at once, e.g. if a ** query joins a virtual table to itself? If so perhaps we should ** move some of these to the cursor object. */ static const char *const fulltext_zStatement[MAX_STMT] = { /* CONTENT_INSERT */ NULL, /* generated in contentInsertStatement() */ /* CONTENT_SELECT */ "select * from %_content where rowid = ?", /* CONTENT_UPDATE */ NULL, /* generated in contentUpdateStatement() */ /* CONTENT_DELETE */ "delete from %_content where rowid = ?", /* TERM_SELECT */ "select rowid, doclist from %_term where term = ? and segment = ?", /* TERM_SELECT_ALL */ "select doclist from %_term where term = ? order by segment", /* TERM_INSERT */ "insert into %_term (rowid, term, segment, doclist) values (?, ?, ?, ?)", /* TERM_UPDATE */ "update %_term set doclist = ? where rowid = ?", /* TERM_DELETE */ "delete from %_term where rowid = ?", }; /* ** A connection to a fulltext index is an instance of the following ** structure. The xCreate and xConnect methods create an instance ** of this structure and xDestroy and xDisconnect free that instance. ** All other methods receive a pointer to the structure as one of their ** arguments. */ struct fulltext_vtab { sqlite3_vtab base; /* Base class used by SQLite core */ sqlite3 *db; /* The database connection */ const char *zName; /* virtual table name */ int nColumn; /* number of columns in virtual table */ char **azColumn; /* column names. malloced */ char **azContentColumn; /* column names in content table; malloced */ sqlite3_tokenizer *pTokenizer; /* tokenizer for inserts and queries */ /* Precompiled statements which we keep as long as the table is ** open. */ sqlite3_stmt *pFulltextStatements[MAX_STMT]; }; /* ** When the core wants to do a query, it create a cursor using a ** call to xOpen. This structure is an instance of a cursor. It ** is destroyed by xClose. */ typedef struct fulltext_cursor { sqlite3_vtab_cursor base; /* Base class used by SQLite core */ QueryType iCursorType; /* Copy of sqlite3_index_info.idxNum */ sqlite3_stmt *pStmt; /* Prepared statement in use by the cursor */ int eof; /* True if at End Of Results */ Query q; /* Parsed query string */ Snippet snippet; /* Cached snippet for the current row */ int iColumn; /* Column being searched */ DocListReader result; /* used when iCursorType == QUERY_FULLTEXT */ } fulltext_cursor; static struct fulltext_vtab *cursor_vtab(fulltext_cursor *c){ return (fulltext_vtab *) c->base.pVtab; } static const sqlite3_module fulltextModule; /* forward declaration */ /* Append a list of strings separated by commas to a StringBuffer. */ static void appendList(StringBuffer *sb, int nString, char **azString){ int i; for(i=0; i<nString; ++i){ if( i>0 ) append(sb, ", "); append(sb, azString[i]); } } /* Return a dynamically generated statement of the form * insert into %_content (rowid, ...) values (?, ...) */ static const char *contentInsertStatement(fulltext_vtab *v){ StringBuffer sb; int i; initStringBuffer(&sb); append(&sb, "insert into %_content (rowid, "); appendList(&sb, v->nColumn, v->azContentColumn); append(&sb, ") values (?"); for(i=0; i<v->nColumn; ++i) append(&sb, ", ?"); append(&sb, ")"); return sb.s; } /* Return a dynamically generated statement of the form * update %_content set [col_0] = ?, [col_1] = ?, ... * where rowid = ? */ static const char *contentUpdateStatement(fulltext_vtab *v){ StringBuffer sb; int i; initStringBuffer(&sb); append(&sb, "update %_content set "); for(i=0; i<v->nColumn; ++i) { if( i>0 ){ append(&sb, ", "); } append(&sb, v->azContentColumn[i]); append(&sb, " = ?"); } append(&sb, " where rowid = ?"); return sb.s; } /* Puts a freshly-prepared statement determined by iStmt in *ppStmt. ** If the indicated statement has never been prepared, it is prepared ** and cached, otherwise the cached version is reset. */ static int sql_get_statement(fulltext_vtab *v, fulltext_statement iStmt, sqlite3_stmt **ppStmt){ assert( iStmt<MAX_STMT ); if( v->pFulltextStatements[iStmt]==NULL ){ const char *zStmt; int rc; switch( iStmt ){ case CONTENT_INSERT_STMT: zStmt = contentInsertStatement(v); break; case CONTENT_UPDATE_STMT: zStmt = contentUpdateStatement(v); break; default: zStmt = fulltext_zStatement[iStmt]; } rc = sql_prepare(v->db, v->zName, &v->pFulltextStatements[iStmt], zStmt); if( zStmt != fulltext_zStatement[iStmt]) free((void *) zStmt); if( rc!=SQLITE_OK ) return rc; } else { int rc = sqlite3_reset(v->pFulltextStatements[iStmt]); if( rc!=SQLITE_OK ) return rc; } *ppStmt = v->pFulltextStatements[iStmt]; return SQLITE_OK; } /* Step the indicated statement, handling errors SQLITE_BUSY (by ** retrying) and SQLITE_SCHEMA (by re-preparing and transferring ** bindings to the new statement). ** TODO(adam): We should extend this function so that it can work with ** statements declared locally, not only globally cached statements. */ static int sql_step_statement(fulltext_vtab *v, fulltext_statement iStmt, sqlite3_stmt **ppStmt){ int rc; sqlite3_stmt *s = *ppStmt; assert( iStmt<MAX_STMT ); assert( s==v->pFulltextStatements[iStmt] ); while( (rc=sqlite3_step(s))!=SQLITE_DONE && rc!=SQLITE_ROW ){ sqlite3_stmt *pNewStmt; if( rc==SQLITE_BUSY ) continue; if( rc!=SQLITE_ERROR ) return rc; rc = sqlite3_reset(s); if( rc!=SQLITE_SCHEMA ) return SQLITE_ERROR; v->pFulltextStatements[iStmt] = NULL; /* Still in s */ rc = sql_get_statement(v, iStmt, &pNewStmt); if( rc!=SQLITE_OK ) goto err; *ppStmt = pNewStmt; rc = sqlite3_transfer_bindings(s, pNewStmt); if( rc!=SQLITE_OK ) goto err; rc = sqlite3_finalize(s); if( rc!=SQLITE_OK ) return rc; s = pNewStmt; } return rc; err: sqlite3_finalize(s); return rc; } /* Like sql_step_statement(), but convert SQLITE_DONE to SQLITE_OK. ** Useful for statements like UPDATE, where we expect no results. */ static int sql_single_step_statement(fulltext_vtab *v, fulltext_statement iStmt, sqlite3_stmt **ppStmt){ int rc = sql_step_statement(v, iStmt, ppStmt); return (rc==SQLITE_DONE) ? SQLITE_OK : rc; } /* insert into %_content (rowid, ...) values ([rowid], [pValues]) */ static int content_insert(fulltext_vtab *v, sqlite3_value *rowid, sqlite3_value **pValues){ sqlite3_stmt *s; int i; int rc = sql_get_statement(v, CONTENT_INSERT_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_value(s, 1, rowid); if( rc!=SQLITE_OK ) return rc; for(i=0; i<v->nColumn; ++i){ rc = sqlite3_bind_value(s, 2+i, pValues[i]); if( rc!=SQLITE_OK ) return rc; } return sql_single_step_statement(v, CONTENT_INSERT_STMT, &s); } /* update %_content set col0 = pValues[0], col1 = pValues[1], ... * where rowid = [iRowid] */ static int content_update(fulltext_vtab *v, sqlite3_value **pValues, sqlite_int64 iRowid){ sqlite3_stmt *s; int i; int rc = sql_get_statement(v, CONTENT_UPDATE_STMT, &s); if( rc!=SQLITE_OK ) return rc; for(i=0; i<v->nColumn; ++i){ rc = sqlite3_bind_value(s, 1+i, pValues[i]); if( rc!=SQLITE_OK ) return rc; } rc = sqlite3_bind_int64(s, 1+v->nColumn, iRowid); if( rc!=SQLITE_OK ) return rc; return sql_single_step_statement(v, CONTENT_UPDATE_STMT, &s); } void freeStringArray(int nString, const char **pString){ int i; for (i=0 ; i < nString ; ++i) { free((void *) pString[i]); } free((void *) pString); } /* select * from %_content where rowid = [iRow] * The caller must delete the returned array and all strings in it. * * TODO: Perhaps we should return pointer/length strings here for consistency * with other code which uses pointer/length. */ static int content_select(fulltext_vtab *v, sqlite_int64 iRow, const char ***pValues){ sqlite3_stmt *s; const char **values; int i; int rc; *pValues = NULL; rc = sql_get_statement(v, CONTENT_SELECT_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_int64(s, 1, iRow); if( rc!=SQLITE_OK ) return rc; rc = sql_step_statement(v, CONTENT_SELECT_STMT, &s); if( rc!=SQLITE_ROW ) return rc; values = (const char **) malloc(v->nColumn * sizeof(const char *)); for(i=0; i<v->nColumn; ++i){ values[i] = string_dup((char*)sqlite3_column_text(s, i)); } /* We expect only one row. We must execute another sqlite3_step() * to complete the iteration; otherwise the table will remain locked. */ rc = sqlite3_step(s); if( rc==SQLITE_DONE ){ *pValues = values; return SQLITE_OK; } freeStringArray(v->nColumn, values); return rc; } /* delete from %_content where rowid = [iRow ] */ static int content_delete(fulltext_vtab *v, sqlite_int64 iRow){ sqlite3_stmt *s; int rc = sql_get_statement(v, CONTENT_DELETE_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_int64(s, 1, iRow); if( rc!=SQLITE_OK ) return rc; return sql_single_step_statement(v, CONTENT_DELETE_STMT, &s); } /* select rowid, doclist from %_term * where term = [pTerm] and segment = [iSegment] * If found, returns SQLITE_ROW; the caller must free the * returned doclist. If no rows found, returns SQLITE_DONE. */ static int term_select(fulltext_vtab *v, const char *pTerm, int nTerm, int iSegment, sqlite_int64 *rowid, DocList *out){ sqlite3_stmt *s; int rc = sql_get_statement(v, TERM_SELECT_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_text(s, 1, pTerm, nTerm, SQLITE_STATIC); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_int(s, 2, iSegment); if( rc!=SQLITE_OK ) return rc; rc = sql_step_statement(v, TERM_SELECT_STMT, &s); if( rc!=SQLITE_ROW ) return rc; *rowid = sqlite3_column_int64(s, 0); docListInit(out, DL_DEFAULT, sqlite3_column_blob(s, 1), sqlite3_column_bytes(s, 1)); /* We expect only one row. We must execute another sqlite3_step() * to complete the iteration; otherwise the table will remain locked. */ rc = sqlite3_step(s); return rc==SQLITE_DONE ? SQLITE_ROW : rc; } /* Load the segment doclists for term pTerm and merge them in ** appropriate order into out. Returns SQLITE_OK if successful. If ** there are no segments for pTerm, successfully returns an empty ** doclist in out. ** ** Each document consists of 1 or more "columns". The number of ** columns is v->nColumn. If iColumn==v->nColumn, then return ** position information about all columns. If iColumn<v->nColumn, ** then only return position information about the iColumn-th column ** (where the first column is 0). */ static int term_select_all( fulltext_vtab *v, /* The fulltext index we are querying against */ int iColumn, /* If <nColumn, only look at the iColumn-th column */ const char *pTerm, /* The term whose posting lists we want */ int nTerm, /* Number of bytes in pTerm */ DocList *out /* Write the resulting doclist here */ ){ DocList doclist; sqlite3_stmt *s; int rc = sql_get_statement(v, TERM_SELECT_ALL_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_text(s, 1, pTerm, nTerm, SQLITE_STATIC); if( rc!=SQLITE_OK ) return rc; docListInit(&doclist, DL_DEFAULT, 0, 0); /* TODO(shess) Handle schema and busy errors. */ while( (rc=sql_step_statement(v, TERM_SELECT_ALL_STMT, &s))==SQLITE_ROW ){ DocList old; /* TODO(shess) If we processed doclists from oldest to newest, we ** could skip the malloc() involved with the following call. For ** now, I'd rather keep this logic similar to index_insert_term(). ** We could additionally drop elements when we see deletes, but ** that would require a distinct version of docListAccumulate(). */ docListInit(&old, DL_DEFAULT, sqlite3_column_blob(s, 0), sqlite3_column_bytes(s, 0)); if( iColumn<v->nColumn ){ /* querying a single column */ docListRestrictColumn(&old, iColumn); } /* doclist contains the newer data, so write it over old. Then ** steal accumulated result for doclist. */ docListAccumulate(&old, &doclist); docListDestroy(&doclist); doclist = old; } if( rc!=SQLITE_DONE ){ docListDestroy(&doclist); return rc; } docListDiscardEmpty(&doclist); *out = doclist; return SQLITE_OK; } /* insert into %_term (rowid, term, segment, doclist) values ([piRowid], [pTerm], [iSegment], [doclist]) ** Lets sqlite select rowid if piRowid is NULL, else uses *piRowid. ** ** NOTE(shess) piRowid is IN, with values of "space of int64" plus ** null, it is not used to pass data back to the caller. */ static int term_insert(fulltext_vtab *v, sqlite_int64 *piRowid, const char *pTerm, int nTerm, int iSegment, DocList *doclist){ sqlite3_stmt *s; int rc = sql_get_statement(v, TERM_INSERT_STMT, &s); if( rc!=SQLITE_OK ) return rc; if( piRowid==NULL ){ rc = sqlite3_bind_null(s, 1); }else{ rc = sqlite3_bind_int64(s, 1, *piRowid); } if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_text(s, 2, pTerm, nTerm, SQLITE_STATIC); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_int(s, 3, iSegment); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_blob(s, 4, doclist->pData, doclist->nData, SQLITE_STATIC); if( rc!=SQLITE_OK ) return rc; return sql_single_step_statement(v, TERM_INSERT_STMT, &s); } /* update %_term set doclist = [doclist] where rowid = [rowid] */ static int term_update(fulltext_vtab *v, sqlite_int64 rowid, DocList *doclist){ sqlite3_stmt *s; int rc = sql_get_statement(v, TERM_UPDATE_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_blob(s, 1, doclist->pData, doclist->nData, SQLITE_STATIC); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_int64(s, 2, rowid); if( rc!=SQLITE_OK ) return rc; return sql_single_step_statement(v, TERM_UPDATE_STMT, &s); } static int term_delete(fulltext_vtab *v, sqlite_int64 rowid){ sqlite3_stmt *s; int rc = sql_get_statement(v, TERM_DELETE_STMT, &s); if( rc!=SQLITE_OK ) return rc; rc = sqlite3_bind_int64(s, 1, rowid); if( rc!=SQLITE_OK ) return rc; return sql_single_step_statement(v, TERM_DELETE_STMT, &s); } /* ** Free the memory used to contain a fulltext_vtab structure. */ static void fulltext_vtab_destroy(fulltext_vtab *v){ int iStmt, i; TRACE(("FTS2 Destroy %p\n", v)); for( iStmt=0; iStmt<MAX_STMT; iStmt++ ){ if( v->pFulltextStatements[iStmt]!=NULL ){ sqlite3_finalize(v->pFulltextStatements[iStmt]); v->pFulltextStatements[iStmt] = NULL; } } if( v->pTokenizer!=NULL ){ v->pTokenizer->pModule->xDestroy(v->pTokenizer); v->pTokenizer = NULL; } free(v->azColumn); for(i = 0; i < v->nColumn; ++i) { sqlite3_free(v->azContentColumn[i]); } free(v->azContentColumn); free(v); } /* ** Token types for parsing the arguments to xConnect or xCreate. */ #define TOKEN_EOF 0 /* End of file */ #define TOKEN_SPACE 1 /* Any kind of whitespace */ #define TOKEN_ID 2 /* An identifier */ #define TOKEN_STRING 3 /* A string literal */ #define TOKEN_PUNCT 4 /* A single punctuation character */ /* ** If X is a character that can be used in an identifier then ** IdChar(X) will be true. Otherwise it is false. ** ** For ASCII, any character with the high-order bit set is ** allowed in an identifier. For 7-bit characters, ** sqlite3IsIdChar[X] must be 1. ** ** Ticket #1066. the SQL standard does not allow '$' in the ** middle of identfiers. But many SQL implementations do. ** SQLite will allow '$' in identifiers for compatibility. ** But the feature is undocumented. */ static const char isIdChar[] = { /* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */ 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 2x */ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, /* 3x */ 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 4x */ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, /* 5x */ 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 6x */ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, /* 7x */ }; #define IdChar(C) (((c=C)&0x80)!=0 || (c>0x1f && isIdChar[c-0x20])) /* ** Return the length of the token that begins at z[0]. ** Store the token type in *tokenType before returning. */ static int getToken(const char *z, int *tokenType){ int i, c; switch( *z ){ case 0: { *tokenType = TOKEN_EOF; return 0; } case ' ': case '\t': case '\n': case '\f': case '\r': { for(i=1; isspace(z[i]); i++){} *tokenType = TOKEN_SPACE; return i; } case '\'': case '"': { int delim = z[0]; for(i=1; (c=z[i])!=0; i++){ if( c==delim ){ if( z[i+1]==delim ){ i++; }else{ break; } } } *tokenType = TOKEN_STRING; return i + (c!=0); } case '[': { for(i=1, c=z[0]; c!=']' && (c=z[i])!=0; i++){} *tokenType = TOKEN_ID; return i; } default: { if( !IdChar(*z) ){ break; } for(i=1; IdChar(z[i]); i++){} *tokenType = TOKEN_ID; return i; } } *tokenType = TOKEN_PUNCT; return 1; } /* ** A token extracted from a string is an instance of the following ** structure. */ typedef struct Token { const char *z; /* Pointer to token text. Not '\000' terminated */ short int n; /* Length of the token text in bytes. */ } Token; /* ** Given a input string (which is really one of the argv[] parameters ** passed into xConnect or xCreate) split the string up into tokens. ** Return an array of pointers to '\000' terminated strings, one string ** for each non-whitespace token. ** ** The returned array is terminated by a single NULL pointer. ** ** Space to hold the returned array is obtained from a single ** malloc and should be freed by passing the return value to free(). ** The individual strings within the token list are all a part of ** the single memory allocation and will all be freed at once. */ static char **tokenizeString(const char *z, int *pnToken){ int nToken = 0; Token *aToken = malloc( strlen(z) * sizeof(aToken[0]) ); int n = 1; int e, i; int totalSize = 0; char **azToken; char *zCopy; while( n>0 ){ n = getToken(z, &e); if( e!=TOKEN_SPACE ){ aToken[nToken].z = z; aToken[nToken].n = n; nToken++; totalSize += n+1; } z += n; } azToken = (char**)malloc( nToken*sizeof(char*) + totalSize ); zCopy = (char*)&azToken[nToken]; nToken--; for(i=0; i<nToken; i++){ azToken[i] = zCopy; n = aToken[i].n; memcpy(zCopy, aToken[i].z, n); zCopy[n] = 0; zCopy += n+1; } azToken[nToken] = 0; free(aToken); *pnToken = nToken; return azToken; } /* ** Convert an SQL-style quoted string into a normal string by removing ** the quote characters. The conversion is done in-place. If the ** input does not begin with a quote character, then this routine ** is a no-op. ** ** Examples: ** ** "abc" becomes abc ** 'xyz' becomes xyz ** [pqr] becomes pqr ** `mno` becomes mno */ void dequoteString(char *z){ int quote; int i, j; if( z==0 ) return; quote = z[0]; switch( quote ){ case '\'': break; case '"': break; case '`': break; /* For MySQL compatibility */ case '[': quote = ']'; break; /* For MS SqlServer compatibility */ default: return; } for(i=1, j=0; z[i]; i++){ if( z[i]==quote ){ if( z[i+1]==quote ){ z[j++] = quote; i++; }else{ z[j++] = 0; break; } }else{ z[j++] = z[i]; } } } /* ** The input azIn is a NULL-terminated list of tokens. Remove the first ** token and all punctuation tokens. Remove the quotes from ** around string literal tokens. ** ** Example: ** ** input: tokenize chinese ( 'simplifed' , 'mixed' ) ** output: chinese simplifed mixed ** ** Another example: ** ** input: delimiters ( '[' , ']' , '...' ) ** output: [ ] ... */ void tokenListToIdList(char **azIn){ int i, j; if( azIn ){ for(i=0, j=-1; azIn[i]; i++){ if( isalnum(azIn[i][0]) || azIn[i][1] ){ dequoteString(azIn[i]); if( j>=0 ){ azIn[j] = azIn[i]; } j++; } } azIn[j] = 0; } } /* ** Find the first alphanumeric token in the string zIn. Null-terminate ** this token. Remove any quotation marks. And return a pointer to ** the result. */ static char *firstToken(char *zIn, char **pzTail){ int i, n, ttype; i = 0; while(1){ n = getToken(zIn, &ttype); if( ttype==TOKEN_SPACE ){ zIn += n; }else if( ttype==TOKEN_EOF ){ *pzTail = zIn; return 0; }else{ zIn[n] = 0; *pzTail = &zIn[1]; dequoteString(zIn); return zIn; } } /*NOTREACHED*/ } /* Return true if... ** ** * s begins with the string t, ignoring case ** * s is longer than t ** * The first character of s beyond t is not a alphanumeric ** ** Ignore leading space in *s. ** ** To put it another way, return true if the first token of ** s[] is t[]. */ static int startsWith(const char *s, const char *t){ while( isspace(*s) ){ s++; } while( *t ){ if( tolower(*s++)!=tolower(*t++) ) return 0; } return *s!='_' && !isalnum(*s); } /* ** An instance of this structure defines the "spec" of a ** full text index. This structure is populated by parseSpec ** and use by fulltextConnect and fulltextCreate. */ typedef struct TableSpec { const char *zName; /* Name of the full-text index */ int nColumn; /* Number of columns to be indexed */ char **azColumn; /* Original names of columns to be indexed */ char **azContentColumn; /* Column names for %_content */ char **azTokenizer; /* Name of tokenizer and its arguments */ } TableSpec; /* ** Reclaim all of the memory used by a TableSpec */ void clearTableSpec(TableSpec *p) { free(p->azColumn); free(p->azContentColumn); free(p->azTokenizer); } /* Parse a CREATE VIRTUAL TABLE statement, which looks like this: * * CREATE VIRTUAL TABLE email * USING fts2(subject, body, tokenize mytokenizer(myarg)) * * We return parsed information in a TableSpec structure. * */ int parseSpec(TableSpec *pSpec, int argc, const char *const*argv, char**pzErr){ int i, j, n; char *z, *zDummy; char **azArg; const char *zTokenizer = 0; /* argv[] entry describing the tokenizer */ assert( argc>=3 ); /* Current interface: ** argv[0] - module name ** argv[1] - database name ** argv[2] - table name ** argv[3..] - columns, optionally followed by tokenizer specification ** and snippet delimiters specification. */ /* Make a copy of the complete argv[][] array in a single allocation. ** The argv[][] array is read-only and transient. We can write to the ** copy in order to modify things and the copy is persistent. */ memset(pSpec, 0, sizeof(*pSpec)); for(i=n=0; i<argc; i++){ n += strlen(argv[i]) + 1; } azArg = malloc( sizeof(char*)*argc + n ); if( azArg==0 ){ return SQLITE_NOMEM; } z = (char*)&azArg[argc]; for(i=0; i<argc; i++){ azArg[i] = z; strcpy(z, argv[i]); z += strlen(z)+1; } /* Identify the column names and the tokenizer and delimiter arguments ** in the argv[][] array. */ pSpec->zName = azArg[2]; pSpec->nColumn = 0; pSpec->azColumn = azArg; zTokenizer = "tokenize simple"; for(i=3, j=0; i<argc; ++i){ if( startsWith(azArg[i],"tokenize") ){ zTokenizer = azArg[i]; }else{ z = azArg[pSpec->nColumn] = firstToken(azArg[i], &zDummy); pSpec->nColumn++; } } if( pSpec->nColumn==0 ){ azArg[0] = "content"; pSpec->nColumn = 1; } /* ** Construct the list of content column names. ** ** Each content column name will be of the form cNNAAAA ** where NN is the column number and AAAA is the sanitized ** column name. "sanitized" means that special characters are ** converted to "_". The cNN prefix guarantees that all column ** names are unique. ** ** The AAAA suffix is not strictly necessary. It is included ** for the convenience of people who might examine the generated ** %_content table and wonder what the columns are used for. */ pSpec->azContentColumn = malloc( pSpec->nColumn * sizeof(char *) ); if( pSpec->azContentColumn==0 ){ clearTableSpec(pSpec); return SQLITE_NOMEM; } for(i=0; i<pSpec->nColumn; i++){ char *p; pSpec->azContentColumn[i] = sqlite3_mprintf("c%d%s", i, azArg[i]); for (p = pSpec->azContentColumn[i]; *p ; ++p) { if( !isalnum(*p) ) *p = '_'; } } /* ** Parse the tokenizer specification string. */ pSpec->azTokenizer = tokenizeString(zTokenizer, &n); tokenListToIdList(pSpec->azTokenizer); return SQLITE_OK; } /* ** Generate a CREATE TABLE statement that describes the schema of ** the virtual table. Return a pointer to this schema string. ** ** Space is obtained from sqlite3_mprintf() and should be freed ** using sqlite3_free(). */ static char *fulltextSchema( int nColumn, /* Number of columns */ const char *const* azColumn, /* List of columns */ const char *zTableName /* Name of the table */ ){ int i; char *zSchema, *zNext; const char *zSep = "("; zSchema = sqlite3_mprintf("CREATE TABLE x"); for(i=0; i<nColumn; i++){ zNext = sqlite3_mprintf("%s%s%Q", zSchema, zSep, azColumn[i]); sqlite3_free(zSchema); zSchema = zNext; zSep = ","; } zNext = sqlite3_mprintf("%s,%Q)", zSchema, zTableName); sqlite3_free(zSchema); return zNext; } /* ** Build a new sqlite3_vtab structure that will describe the ** fulltext index defined by spec. */ static int constructVtab( sqlite3 *db, /* The SQLite database connection */ TableSpec *spec, /* Parsed spec information from parseSpec() */ sqlite3_vtab **ppVTab, /* Write the resulting vtab structure here */ char **pzErr /* Write any error message here */ ){ int rc; int n; fulltext_vtab *v = 0; const sqlite3_tokenizer_module *m = NULL; char *schema; v = (fulltext_vtab *) malloc(sizeof(fulltext_vtab)); if( v==0 ) return SQLITE_NOMEM; memset(v, 0, sizeof(*v)); /* sqlite will initialize v->base */ v->db = db; v->zName = spec->zName; /* Freed when azColumn is freed */ v->nColumn = spec->nColumn; v->azContentColumn = spec->azContentColumn; spec->azContentColumn = 0; v->azColumn = spec->azColumn; spec->azColumn = 0; if( spec->azTokenizer==0 ){ return SQLITE_NOMEM; } /* TODO(shess) For now, add new tokenizers as else if clauses. */ if( spec->azTokenizer[0]==0 || startsWith(spec->azTokenizer[0], "simple") ){ sqlite3Fts2SimpleTokenizerModule(&m); }else if( startsWith(spec->azTokenizer[0], "porter") ){ sqlite3Fts2PorterTokenizerModule(&m); }else{ *pzErr = sqlite3_mprintf("unknown tokenizer: %s", spec->azTokenizer[0]); rc = SQLITE_ERROR; goto err; } for(n=0; spec->azTokenizer[n]; n++){} if( n ){ rc = m->xCreate(n-1, (const char*const*)&spec->azTokenizer[1], &v->pTokenizer); }else{ rc = m->xCreate(0, 0, &v->pTokenizer); } if( rc!=SQLITE_OK ) goto err; v->pTokenizer->pModule = m; /* TODO: verify the existence of backing tables foo_content, foo_term */ schema = fulltextSchema(v->nColumn, (const char*const*)v->azColumn, spec->zName); rc = sqlite3_declare_vtab(db, schema); sqlite3_free(schema); if( rc!=SQLITE_OK ) goto err; memset(v->pFulltextStatements, 0, sizeof(v->pFulltextStatements)); *ppVTab = &v->base; TRACE(("FTS2 Connect %p\n", v)); return rc; err: fulltext_vtab_destroy(v); return rc; } static int fulltextConnect( sqlite3 *db, void *pAux, int argc, const char *const*argv, sqlite3_vtab **ppVTab, char **pzErr ){ TableSpec spec; int rc = parseSpec(&spec, argc, argv, pzErr); if( rc!=SQLITE_OK ) return rc; rc = constructVtab(db, &spec, ppVTab, pzErr); clearTableSpec(&spec); return rc; } /* The %_content table holds the text of each document, with ** the rowid used as the docid. ** ** The %_term table maps each term to a document list blob ** containing elements sorted by ascending docid, each element ** encoded as: ** ** docid varint-encoded ** token elements: ** position+1 varint-encoded as delta from previous position ** start offset varint-encoded as delta from previous start offset ** end offset varint-encoded as delta from start offset ** ** The sentinel position of 0 indicates the end of the token list. ** ** Additionally, doclist blobs are chunked into multiple segments, ** using segment to order the segments. New elements are added to ** the segment at segment 0, until it exceeds CHUNK_MAX. Then ** segment 0 is deleted, and the doclist is inserted at segment 1. ** If there is already a doclist at segment 1, the segment 0 doclist ** is merged with it, the segment 1 doclist is deleted, and the ** merged doclist is inserted at segment 2, repeating those ** operations until an insert succeeds. ** ** Since this structure doesn't allow us to update elements in place ** in case of deletion or update, these are simply written to ** segment 0 (with an empty token list in case of deletion), with ** docListAccumulate() taking care to retain lower-segment ** information in preference to higher-segment information. */ /* TODO(shess) Provide a VACUUM type operation which both removes ** deleted elements which are no longer necessary, and duplicated ** elements. I suspect this will probably not be necessary in ** practice, though. */ static int fulltextCreate(sqlite3 *db, void *pAux, int argc, const char * const *argv, sqlite3_vtab **ppVTab, char **pzErr){ int rc; TableSpec spec; StringBuffer schema; TRACE(("FTS2 Create\n")); rc = parseSpec(&spec, argc, argv, pzErr); if( rc!=SQLITE_OK ) return rc; initStringBuffer(&schema); append(&schema, "CREATE TABLE %_content("); appendList(&schema, spec.nColumn, spec.azContentColumn); append(&schema, ")"); rc = sql_exec(db, spec.zName, schema.s); free(schema.s); if( rc!=SQLITE_OK ) goto out; rc = sql_exec(db, spec.zName, "create table %_term(term text, segment integer, doclist blob, " "primary key(term, segment));"); if( rc!=SQLITE_OK ) goto out; rc = constructVtab(db, &spec, ppVTab, pzErr); out: clearTableSpec(&spec); return rc; } /* Decide how to handle an SQL query. */ static int fulltextBestIndex(sqlite3_vtab *pVTab, sqlite3_index_info *pInfo){ int i; for(i=0; i<pInfo->nConstraint; ++i){ const struct sqlite3_index_constraint *pConstraint; pConstraint = &pInfo->aConstraint[i]; if( pConstraint->usable ) { if( pConstraint->iColumn==-1 && pConstraint->op==SQLITE_INDEX_CONSTRAINT_EQ ){ pInfo->idxNum = QUERY_ROWID; /* lookup by rowid */ } else if( pConstraint->iColumn>=0 && pConstraint->op==SQLITE_INDEX_CONSTRAINT_MATCH ){ /* full-text search */ pInfo->idxNum = QUERY_FULLTEXT + pConstraint->iColumn; } else continue; pInfo->aConstraintUsage[i].argvIndex = 1; pInfo->aConstraintUsage[i].omit = 1; /* An arbitrary value for now. * TODO: Perhaps rowid matches should be considered cheaper than * full-text searches. */ pInfo->estimatedCost = 1.0; return SQLITE_OK; } } pInfo->idxNum = QUERY_GENERIC; TRACE(("FTS2 BestIndex\n")); return SQLITE_OK; } static int fulltextDisconnect(sqlite3_vtab *pVTab){ TRACE(("FTS2 Disconnect %p\n", pVTab)); fulltext_vtab_destroy((fulltext_vtab *)pVTab); return SQLITE_OK; } static int fulltextDestroy(sqlite3_vtab *pVTab){ fulltext_vtab *v = (fulltext_vtab *)pVTab; int rc; TRACE(("FTS2 Destroy %p\n", pVTab)); rc = sql_exec(v->db, v->zName, "drop table %_content; drop table %_term"); if( rc!=SQLITE_OK ) return rc; fulltext_vtab_destroy((fulltext_vtab *)pVTab); return SQLITE_OK; } static int fulltextOpen(sqlite3_vtab *pVTab, sqlite3_vtab_cursor **ppCursor){ fulltext_cursor *c; c = (fulltext_cursor *) calloc(sizeof(fulltext_cursor), 1); /* sqlite will initialize c->base */ *ppCursor = &c->base; TRACE(("FTS2 Open %p: %p\n", pVTab, c)); return SQLITE_OK; } /* Free all of the dynamically allocated memory held by *q */ static void queryClear(Query *q){ int i; for(i = 0; i < q->nTerms; ++i){ free(q->pTerms[i].pTerm); } free(q->pTerms); memset(q, 0, sizeof(*q)); } /* Free all of the dynamically allocated memory held by the ** Snippet */ static void snippetClear(Snippet *p){ free(p->aMatch); free(p->zOffset); free(p->zSnippet); memset(p, 0, sizeof(*p)); } /* ** Append a single entry to the p->aMatch[] log. */ static void snippetAppendMatch( Snippet *p, /* Append the entry to this snippet */ int iCol, int iTerm, /* The column and query term */ int iStart, int nByte /* Offset and size of the match */ ){ int i; struct snippetMatch *pMatch; if( p->nMatch+1>=p->nAlloc ){ p->nAlloc = p->nAlloc*2 + 10; p->aMatch = realloc(p->aMatch, p->nAlloc*sizeof(p->aMatch[0]) ); if( p->aMatch==0 ){ p->nMatch = 0; p->nAlloc = 0; return; } } i = p->nMatch++; pMatch = &p->aMatch[i]; pMatch->iCol = iCol; pMatch->iTerm = iTerm; pMatch->iStart = iStart; pMatch->nByte = nByte; } /* ** Sizing information for the circular buffer used in snippetOffsetsOfColumn() */ #define FTS2_ROTOR_SZ (32) #define FTS2_ROTOR_MASK (FTS2_ROTOR_SZ-1) /* ** Add entries to pSnippet->aMatch[] for every match that occurs against ** document zDoc[0..nDoc-1] which is stored in column iColumn. */ static void snippetOffsetsOfColumn( Query *pQuery, Snippet *pSnippet, int iColumn, const char *zDoc, int nDoc ){ const sqlite3_tokenizer_module *pTModule; /* The tokenizer module */ sqlite3_tokenizer *pTokenizer; /* The specific tokenizer */ sqlite3_tokenizer_cursor *pTCursor; /* Tokenizer cursor */ fulltext_vtab *pVtab; /* The full text index */ int nColumn; /* Number of columns in the index */ const QueryTerm *aTerm; /* Query string terms */ int nTerm; /* Number of query string terms */ int i, j; /* Loop counters */ int rc; /* Return code */ unsigned int match, prevMatch; /* Phrase search bitmasks */ const char *zToken; /* Next token from the tokenizer */ int nToken; /* Size of zToken */ int iBegin, iEnd, iPos; /* Offsets of beginning and end */ /* The following variables keep a circular buffer of the last ** few tokens */ unsigned int iRotor = 0; /* Index of current token */ int iRotorBegin[FTS2_ROTOR_SZ]; /* Beginning offset of token */ int iRotorLen[FTS2_ROTOR_SZ]; /* Length of token */ pVtab = pQuery->pFts; nColumn = pVtab->nColumn; pTokenizer = pVtab->pTokenizer; pTModule = pTokenizer->pModule; rc = pTModule->xOpen(pTokenizer, zDoc, nDoc, &pTCursor); if( rc ) return; pTCursor->pTokenizer = pTokenizer; aTerm = pQuery->pTerms; nTerm = pQuery->nTerms; if( nTerm>=FTS2_ROTOR_SZ ){ nTerm = FTS2_ROTOR_SZ - 1; } prevMatch = 0; while(1){ rc = pTModule->xNext(pTCursor, &zToken, &nToken, &iBegin, &iEnd, &iPos); if( rc ) break; iRotorBegin[iRotor&FTS2_ROTOR_MASK] = iBegin; iRotorLen[iRotor&FTS2_ROTOR_MASK] = iEnd-iBegin; match = 0; for(i=0; i<nTerm; i++){ int iCol; iCol = aTerm[i].iColumn; if( iCol>=0 && iCol<nColumn && iCol!=iColumn ) continue; if( aTerm[i].nTerm!=nToken ) continue; if( memcmp(aTerm[i].pTerm, zToken, nToken) ) continue; if( aTerm[i].iPhrase>1 && (prevMatch & (1<<i))==0 ) continue; match |= 1<<i; if( i==nTerm-1 || aTerm[i+1].iPhrase==1 ){ for(j=aTerm[i].iPhrase-1; j>=0; j--){ int k = (iRotor-j) & FTS2_ROTOR_MASK; snippetAppendMatch(pSnippet, iColumn, i-j, iRotorBegin[k], iRotorLen[k]); } } } prevMatch = match<<1; iRotor++; } pTModule->xClose(pTCursor); } /* ** Compute all offsets for the current row of the query. ** If the offsets have already been computed, this routine is a no-op. */ static void snippetAllOffsets(fulltext_cursor *p){ int nColumn; int iColumn, i; int iFirst, iLast; fulltext_vtab *pFts; if( p->snippet.nMatch ) return; if( p->q.nTerms==0 ) return; pFts = p->q.pFts; nColumn = pFts->nColumn; iColumn = p->iCursorType; if( iColumn<0 || iColumn>=nColumn ){ iFirst = 0; iLast = nColumn-1; }else{ iFirst = iColumn; iLast = iColumn; } for(i=iFirst; i<=iLast; i++){ const char *zDoc; int nDoc; zDoc = (const char*)sqlite3_column_text(p->pStmt, i+1); nDoc = sqlite3_column_bytes(p->pStmt, i+1); snippetOffsetsOfColumn(&p->q, &p->snippet, i, zDoc, nDoc); } } /* ** Convert the information in the aMatch[] array of the snippet ** into the string zOffset[0..nOffset-1]. */ static void snippetOffsetText(Snippet *p){ int i; int cnt = 0; StringBuffer sb; char zBuf[200]; if( p->zOffset ) return; initStringBuffer(&sb); for(i=0; i<p->nMatch; i++){ struct snippetMatch *pMatch = &p->aMatch[i]; zBuf[0] = ' '; sprintf(&zBuf[cnt>0], "%d %d %d %d", pMatch->iCol, pMatch->iTerm, pMatch->iStart, pMatch->nByte); append(&sb, zBuf); cnt++; } p->zOffset = sb.s; p->nOffset = sb.len; } /* ** zDoc[0..nDoc-1] is phrase of text. aMatch[0..nMatch-1] are a set ** of matching words some of which might be in zDoc. zDoc is column ** number iCol. ** ** iBreak is suggested spot in zDoc where we could begin or end an ** excerpt. Return a value similar to iBreak but possibly adjusted ** to be a little left or right so that the break point is better. */ static int wordBoundary( int iBreak, /* The suggested break point */ const char *zDoc, /* Document text */ int nDoc, /* Number of bytes in zDoc[] */ struct snippetMatch *aMatch, /* Matching words */ int nMatch, /* Number of entries in aMatch[] */ int iCol /* The column number for zDoc[] */ ){ int i; if( iBreak<=10 ){ return 0; } if( iBreak>=nDoc-10 ){ return nDoc; } for(i=0; i<nMatch && aMatch[i].iCol<iCol; i++){} while( i<nMatch && aMatch[i].iStart+aMatch[i].nByte<iBreak ){ i++; } if( i<nMatch ){ if( aMatch[i].iStart<iBreak+10 ){ return aMatch[i].iStart; } if( i>0 && aMatch[i-1].iStart+aMatch[i-1].nByte>=iBreak ){ return aMatch[i-1].iStart; } } for(i=1; i<=10; i++){ if( isspace(zDoc[iBreak-i]) ){ return iBreak - i + 1; } if( isspace(zDoc[iBreak+i]) ){ return iBreak + i + 1; } } return iBreak; } /* ** If the StringBuffer does not end in white space, add a single ** space character to the end. */ static void appendWhiteSpace(StringBuffer *p){ if( p->len==0 ) return; if( isspace(p->s[p->len-1]) ) return; append(p, " "); } /* ** Remove white space from teh end of the StringBuffer */ static void trimWhiteSpace(StringBuffer *p){ while( p->len>0 && isspace(p->s[p->len-1]) ){ p->len--; } } /* ** Allowed values for Snippet.aMatch[].snStatus */ #define SNIPPET_IGNORE 0 /* It is ok to omit this match from the snippet */ #define SNIPPET_DESIRED 1 /* We want to include this match in the snippet */ /* ** Generate the text of a snippet. */ static void snippetText( fulltext_cursor *pCursor, /* The cursor we need the snippet for */ const char *zStartMark, /* Markup to appear before each match */ const char *zEndMark, /* Markup to appear after each match */ const char *zEllipsis /* Ellipsis mark */ ){ int i, j; struct snippetMatch *aMatch; int nMatch; int nDesired; StringBuffer sb; int tailCol; int tailOffset; int iCol; int nDoc; const char *zDoc; int iStart, iEnd; int tailEllipsis = 0; int iMatch; free(pCursor->snippet.zSnippet); pCursor->snippet.zSnippet = 0; aMatch = pCursor->snippet.aMatch; nMatch = pCursor->snippet.nMatch; initStringBuffer(&sb); for(i=0; i<nMatch; i++){ aMatch[i].snStatus = SNIPPET_IGNORE; } nDesired = 0; for(i=0; i<pCursor->q.nTerms; i++){ for(j=0; j<nMatch; j++){ if( aMatch[j].iTerm==i ){ aMatch[j].snStatus = SNIPPET_DESIRED; nDesired++; break; } } } iMatch = 0; tailCol = -1; tailOffset = 0; for(i=0; i<nMatch && nDesired>0; i++){ if( aMatch[i].snStatus!=SNIPPET_DESIRED ) continue; nDesired--; iCol = aMatch[i].iCol; zDoc = (const char*)sqlite3_column_text(pCursor->pStmt, iCol+1); nDoc = sqlite3_column_bytes(pCursor->pStmt, iCol+1); iStart = aMatch[i].iStart - 40; iStart = wordBoundary(iStart, zDoc, nDoc, aMatch, nMatch, iCol); if( iStart<=10 ){ iStart = 0; } if( iCol==tailCol && iStart<=tailOffset+20 ){ iStart = tailOffset; } if( (iCol!=tailCol && tailCol>=0) || iStart!=tailOffset ){ trimWhiteSpace(&sb); appendWhiteSpace(&sb); append(&sb, zEllipsis); appendWhiteSpace(&sb); } iEnd = aMatch[i].iStart + aMatch[i].nByte + 40; iEnd = wordBoundary(iEnd, zDoc, nDoc, aMatch, nMatch, iCol); if( iEnd>=nDoc-10 ){ iEnd = nDoc; tailEllipsis = 0; }else{ tailEllipsis = 1; } while( iMatch<nMatch && aMatch[iMatch].iCol<iCol ){ iMatch++; } while( iStart<iEnd ){ while( iMatch<nMatch && aMatch[iMatch].iStart<iStart && aMatch[iMatch].iCol<=iCol ){ iMatch++; } if( iMatch<nMatch && aMatch[iMatch].iStart<iEnd && aMatch[iMatch].iCol==iCol ){ nappend(&sb, &zDoc[iStart], aMatch[iMatch].iStart - iStart); iStart = aMatch[iMatch].iStart; append(&sb, zStartMark); nappend(&sb, &zDoc[iStart], aMatch[iMatch].nByte); append(&sb, zEndMark); iStart += aMatch[iMatch].nByte; for(j=iMatch+1; j<nMatch; j++){ if( aMatch[j].iTerm==aMatch[iMatch].iTerm && aMatch[j].snStatus==SNIPPET_DESIRED ){ nDesired--; aMatch[j].snStatus = SNIPPET_IGNORE; } } }else{ nappend(&sb, &zDoc[iStart], iEnd - iStart); iStart = iEnd; } } tailCol = iCol; tailOffset = iEnd; } trimWhiteSpace(&sb); if( tailEllipsis ){ appendWhiteSpace(&sb); append(&sb, zEllipsis); } pCursor->snippet.zSnippet = sb.s; pCursor->snippet.nSnippet = sb.len; } /* ** Close the cursor. For additional information see the documentation ** on the xClose method of the virtual table interface. */ static int fulltextClose(sqlite3_vtab_cursor *pCursor){ fulltext_cursor *c = (fulltext_cursor *) pCursor; TRACE(("FTS2 Close %p\n", c)); sqlite3_finalize(c->pStmt); queryClear(&c->q); snippetClear(&c->snippet); if( c->result.pDoclist!=NULL ){ docListDelete(c->result.pDoclist); } free(c); return SQLITE_OK; } static int fulltextNext(sqlite3_vtab_cursor *pCursor){ fulltext_cursor *c = (fulltext_cursor *) pCursor; sqlite_int64 iDocid; int rc; TRACE(("FTS2 Next %p\n", pCursor)); snippetClear(&c->snippet); if( c->iCursorType < QUERY_FULLTEXT ){ /* TODO(shess) Handle SQLITE_SCHEMA AND SQLITE_BUSY. */ rc = sqlite3_step(c->pStmt); switch( rc ){ case SQLITE_ROW: c->eof = 0; return SQLITE_OK; case SQLITE_DONE: c->eof = 1; return SQLITE_OK; default: c->eof = 1; return rc; } } else { /* full-text query */ rc = sqlite3_reset(c->pStmt); if( rc!=SQLITE_OK ) return rc; iDocid = nextDocid(&c->result); if( iDocid==0 ){ c->eof = 1; return SQLITE_OK; } rc = sqlite3_bind_int64(c->pStmt, 1, iDocid); if( rc!=SQLITE_OK ) return rc; /* TODO(shess) Handle SQLITE_SCHEMA AND SQLITE_BUSY. */ rc = sqlite3_step(c->pStmt); if( rc==SQLITE_ROW ){ /* the case we expect */ c->eof = 0; return SQLITE_OK; } /* an error occurred; abort */ return rc==SQLITE_DONE ? SQLITE_ERROR : rc; } } /* Return a DocList corresponding to the query term *pTerm. If *pTerm ** is the first term of a phrase query, go ahead and evaluate the phrase ** query and return the doclist for the entire phrase query. ** ** The result is stored in pTerm->doclist. */ static int docListOfTerm( fulltext_vtab *v, /* The full text index */ int iColumn, /* column to restrict to. No restrition if >=nColumn */ QueryTerm *pQTerm, /* Term we are looking for, or 1st term of a phrase */ DocList **ppResult /* Write the result here */ ){ DocList *pLeft, *pRight, *pNew; int i, rc; pLeft = docListNew(DL_POSITIONS); rc = term_select_all(v, iColumn, pQTerm->pTerm, pQTerm->nTerm, pLeft); if( rc ) return rc; for(i=1; i<=pQTerm->nPhrase; i++){ pRight = docListNew(DL_POSITIONS); rc = term_select_all(v, iColumn, pQTerm[i].pTerm, pQTerm[i].nTerm, pRight); if( rc ){ docListDelete(pLeft); return rc; } pNew = docListNew(i<pQTerm->nPhrase ? DL_POSITIONS : DL_DOCIDS); docListPhraseMerge(pLeft, pRight, pNew); docListDelete(pLeft); docListDelete(pRight); pLeft = pNew; } *ppResult = pLeft; return SQLITE_OK; } /* Add a new term pTerm[0..nTerm-1] to the query *q. */ static void queryAdd(Query *q, const char *pTerm, int nTerm){ QueryTerm *t; ++q->nTerms; q->pTerms = realloc(q->pTerms, q->nTerms * sizeof(q->pTerms[0])); if( q->pTerms==0 ){ q->nTerms = 0; return; } t = &q->pTerms[q->nTerms - 1]; memset(t, 0, sizeof(*t)); t->pTerm = malloc(nTerm+1); memcpy(t->pTerm, pTerm, nTerm); t->pTerm[nTerm] = 0; t->nTerm = nTerm; t->isOr = q->nextIsOr; q->nextIsOr = 0; t->iColumn = q->nextColumn; q->nextColumn = q->dfltColumn; } /* ** Check to see if the string zToken[0...nToken-1] matches any ** column name in the virtual table. If it does, ** return the zero-indexed column number. If not, return -1. */ static int checkColumnSpecifier( fulltext_vtab *pVtab, /* The virtual table */ const char *zToken, /* Text of the token */ int nToken /* Number of characters in the token */ ){ int i; for(i=0; i<pVtab->nColumn; i++){ if( memcmp(pVtab->azColumn[i], zToken, nToken)==0 && pVtab->azColumn[i][nToken]==0 ){ return i; } } return -1; } /* ** Parse the text at pSegment[0..nSegment-1]. Add additional terms ** to the query being assemblied in pQuery. ** ** inPhrase is true if pSegment[0..nSegement-1] is contained within ** double-quotes. If inPhrase is true, then the first term ** is marked with the number of terms in the phrase less one and ** OR and "-" syntax is ignored. If inPhrase is false, then every ** term found is marked with nPhrase=0 and OR and "-" syntax is significant. */ static int tokenizeSegment( sqlite3_tokenizer *pTokenizer, /* The tokenizer to use */ const char *pSegment, int nSegment, /* Query expression being parsed */ int inPhrase, /* True if within "..." */ Query *pQuery /* Append results here */ ){ const sqlite3_tokenizer_module *pModule = pTokenizer->pModule; sqlite3_tokenizer_cursor *pCursor; int firstIndex = pQuery->nTerms; int iCol; int nTerm = 1; int rc = pModule->xOpen(pTokenizer, pSegment, nSegment, &pCursor); if( rc!=SQLITE_OK ) return rc; pCursor->pTokenizer = pTokenizer; while( 1 ){ const char *pToken; int nToken, iBegin, iEnd, iPos; rc = pModule->xNext(pCursor, &pToken, &nToken, &iBegin, &iEnd, &iPos); if( rc!=SQLITE_OK ) break; if( !inPhrase && pSegment[iEnd]==':' && (iCol = checkColumnSpecifier(pQuery->pFts, pToken, nToken))>=0 ){ pQuery->nextColumn = iCol; continue; } if( !inPhrase && pQuery->nTerms>0 && nToken==2 && pSegment[iBegin]=='O' && pSegment[iBegin+1]=='R' ){ pQuery->nextIsOr = 1; continue; } queryAdd(pQuery, pToken, nToken); if( !inPhrase && iBegin>0 && pSegment[iBegin-1]=='-' ){ pQuery->pTerms[pQuery->nTerms-1].isNot = 1; } pQuery->pTerms[pQuery->nTerms-1].iPhrase = nTerm; if( inPhrase ){ nTerm++; } } if( inPhrase && pQuery->nTerms>firstIndex ){ pQuery->pTerms[firstIndex].nPhrase = pQuery->nTerms - firstIndex - 1; } return pModule->xClose(pCursor); } /* Parse a query string, yielding a Query object pQuery. ** ** The calling function will need to queryClear() to clean up ** the dynamically allocated memory held by pQuery. */ static int parseQuery( fulltext_vtab *v, /* The fulltext index */ const char *zInput, /* Input text of the query string */ int nInput, /* Size of the input text */ int dfltColumn, /* Default column of the index to match against */ Query *pQuery /* Write the parse results here. */ ){ int iInput, inPhrase = 0; if( zInput==0 ) nInput = 0; if( nInput<0 ) nInput = strlen(zInput); pQuery->nTerms = 0; pQuery->pTerms = NULL; pQuery->nextIsOr = 0; pQuery->nextColumn = dfltColumn; pQuery->dfltColumn = dfltColumn; pQuery->pFts = v; for(iInput=0; iInput<nInput; ++iInput){ int i; for(i=iInput; i<nInput && zInput[i]!='"'; ++i){} if( i>iInput ){ tokenizeSegment(v->pTokenizer, zInput+iInput, i-iInput, inPhrase, pQuery); } iInput = i; if( i<nInput ){ assert( zInput[i]=='"' ); inPhrase = !inPhrase; } } if( inPhrase ){ /* unmatched quote */ queryClear(pQuery); return SQLITE_ERROR; } return SQLITE_OK; } /* Perform a full-text query using the search expression in ** zInput[0..nInput-1]. Return a list of matching documents ** in pResult. ** ** Queries must match column iColumn. Or if iColumn>=nColumn ** they are allowed to match against any column. */ static int fulltextQuery( fulltext_vtab *v, /* The full text index */ int iColumn, /* Match against this column by default */ const char *zInput, /* The query string */ int nInput, /* Number of bytes in zInput[] */ DocList **pResult, /* Write the result doclist here */ Query *pQuery /* Put parsed query string here */ ){ int i, iNext, rc; DocList *pLeft = NULL; DocList *pRight, *pNew, *pOr; int nNot = 0; QueryTerm *aTerm; rc = parseQuery(v, zInput, nInput, iColumn, pQuery); if( rc!=SQLITE_OK ) return rc; /* Merge AND terms. */ aTerm = pQuery->pTerms; for(i = 0; i<pQuery->nTerms; i=iNext){ if( aTerm[i].isNot ){ /* Handle all NOT terms in a separate pass */ nNot++; iNext = i + aTerm[i].nPhrase+1; continue; } iNext = i + aTerm[i].nPhrase + 1; rc = docListOfTerm(v, aTerm[i].iColumn, &aTerm[i], &pRight); if( rc ){ queryClear(pQuery); return rc; } while( iNext<pQuery->nTerms && aTerm[iNext].isOr ){ rc = docListOfTerm(v, aTerm[iNext].iColumn, &aTerm[iNext], &pOr); iNext += aTerm[iNext].nPhrase + 1; if( rc ){ queryClear(pQuery); return rc; } pNew = docListNew(DL_DOCIDS); docListOrMerge(pRight, pOr, pNew); docListDelete(pRight); docListDelete(pOr); pRight = pNew; } if( pLeft==0 ){ pLeft = pRight; }else{ pNew = docListNew(DL_DOCIDS); docListAndMerge(pLeft, pRight, pNew); docListDelete(pRight); docListDelete(pLeft); pLeft = pNew; } } if( nNot && pLeft==0 ){ /* We do not yet know how to handle a query of only NOT terms */ return SQLITE_ERROR; } /* Do the EXCEPT terms */ for(i=0; i<pQuery->nTerms; i += aTerm[i].nPhrase + 1){ if( !aTerm[i].isNot ) continue; rc = docListOfTerm(v, aTerm[i].iColumn, &aTerm[i], &pRight); if( rc ){ queryClear(pQuery); docListDelete(pLeft); return rc; } pNew = docListNew(DL_DOCIDS); docListExceptMerge(pLeft, pRight, pNew); docListDelete(pRight); docListDelete(pLeft); pLeft = pNew; } *pResult = pLeft; return rc; } /* ** This is the xFilter interface for the virtual table. See ** the virtual table xFilter method documentation for additional ** information. ** ** If idxNum==QUERY_GENERIC then do a full table scan against ** the %_content table. ** ** If idxNum==QUERY_ROWID then do a rowid lookup for a single entry ** in the %_content table. ** ** If idxNum>=QUERY_FULLTEXT then use the full text index. The ** column on the left-hand side of the MATCH operator is column ** number idxNum-QUERY_FULLTEXT, 0 indexed. argv[0] is the right-hand ** side of the MATCH operator. */ static int fulltextFilter( sqlite3_vtab_cursor *pCursor, /* The cursor used for this query */ int idxNum, const char *idxStr, /* Which indexing scheme to use */ int argc, sqlite3_value **argv /* Arguments for the indexing scheme */ ){ fulltext_cursor *c = (fulltext_cursor *) pCursor; fulltext_vtab *v = cursor_vtab(c); int rc; char *zSql; TRACE(("FTS2 Filter %p\n",pCursor)); zSql = sqlite3_mprintf("select rowid, * from %%_content %s", idxNum==QUERY_GENERIC ? "" : "where rowid=?"); rc = sql_prepare(v->db, v->zName, &c->pStmt, zSql); sqlite3_free(zSql); if( rc!=SQLITE_OK ) goto out; c->iCursorType = idxNum; switch( idxNum ){ case QUERY_GENERIC: break; case QUERY_ROWID: rc = sqlite3_bind_int64(c->pStmt, 1, sqlite3_value_int64(argv[0])); if( rc!=SQLITE_OK ) goto out; break; default: /* full-text search */ { const char *zQuery = (const char *)sqlite3_value_text(argv[0]); DocList *pResult; assert( idxNum<=QUERY_FULLTEXT+v->nColumn); assert( argc==1 ); queryClear(&c->q); rc = fulltextQuery(v, idxNum-QUERY_FULLTEXT, zQuery, -1, &pResult, &c->q); if( rc!=SQLITE_OK ) goto out; readerInit(&c->result, pResult); break; } } rc = fulltextNext(pCursor); out: return rc; } /* This is the xEof method of the virtual table. The SQLite core ** calls this routine to find out if it has reached the end of ** a query's results set. */ static int fulltextEof(sqlite3_vtab_cursor *pCursor){ fulltext_cursor *c = (fulltext_cursor *) pCursor; return c->eof; } /* This is the xColumn method of the virtual table. The SQLite ** core calls this method during a query when it needs the value ** of a column from the virtual table. This method needs to use ** one of the sqlite3_result_*() routines to store the requested ** value back in the pContext. */ static int fulltextColumn(sqlite3_vtab_cursor *pCursor, sqlite3_context *pContext, int idxCol){ fulltext_cursor *c = (fulltext_cursor *) pCursor; fulltext_vtab *v = cursor_vtab(c); if( idxCol<v->nColumn ){ sqlite3_value *pVal = sqlite3_column_value(c->pStmt, idxCol+1); sqlite3_result_value(pContext, pVal); }else if( idxCol==v->nColumn ){ /* The extra column whose name is the same as the table. ** Return a blob which is a pointer to the cursor */ sqlite3_result_blob(pContext, &c, sizeof(c), SQLITE_TRANSIENT); } return SQLITE_OK; } /* This is the xRowid method. The SQLite core calls this routine to ** retrive the rowid for the current row of the result set. The ** rowid should be written to *pRowid. */ static int fulltextRowid(sqlite3_vtab_cursor *pCursor, sqlite_int64 *pRowid){ fulltext_cursor *c = (fulltext_cursor *) pCursor; *pRowid = sqlite3_column_int64(c->pStmt, 0); return SQLITE_OK; } /* Add all terms in [zText] to the given hash table. If [iColumn] > 0, * we also store positions and offsets in the hash table using the given * column number. */ static int buildTerms(fulltext_vtab *v, fts2Hash *terms, sqlite_int64 iDocid, const char *zText, int iColumn){ sqlite3_tokenizer *pTokenizer = v->pTokenizer; sqlite3_tokenizer_cursor *pCursor; const char *pToken; int nTokenBytes; int iStartOffset, iEndOffset, iPosition; int rc; rc = pTokenizer->pModule->xOpen(pTokenizer, zText, -1, &pCursor); if( rc!=SQLITE_OK ) return rc; pCursor->pTokenizer = pTokenizer; while( SQLITE_OK==pTokenizer->pModule->xNext(pCursor, &pToken, &nTokenBytes, &iStartOffset, &iEndOffset, &iPosition) ){ DocList *p; /* Positions can't be negative; we use -1 as a terminator internally. */ if( iPosition<0 ){ pTokenizer->pModule->xClose(pCursor); return SQLITE_ERROR; } p = fts2HashFind(terms, pToken, nTokenBytes); if( p==NULL ){ p = docListNew(DL_DEFAULT); docListAddDocid(p, iDocid); fts2HashInsert(terms, pToken, nTokenBytes, p); } if( iColumn>=0 ){ docListAddPosOffset(p, iColumn, iPosition, iStartOffset, iEndOffset); } } /* TODO(shess) Check return? Should this be able to cause errors at ** this point? Actually, same question about sqlite3_finalize(), ** though one could argue that failure there means that the data is ** not durable. *ponder* */ pTokenizer->pModule->xClose(pCursor); return rc; } /* Update the %_terms table to map the term [pTerm] to the given rowid. */ static int index_insert_term(fulltext_vtab *v, const char *pTerm, int nTerm, DocList *d){ sqlite_int64 iIndexRow; DocList doclist; int iSegment = 0, rc; rc = term_select(v, pTerm, nTerm, iSegment, &iIndexRow, &doclist); if( rc==SQLITE_DONE ){ docListInit(&doclist, DL_DEFAULT, 0, 0); docListUpdate(&doclist, d); /* TODO(shess) Consider length(doclist)>CHUNK_MAX? */ rc = term_insert(v, NULL, pTerm, nTerm, iSegment, &doclist); goto err; } if( rc!=SQLITE_ROW ) return SQLITE_ERROR; docListUpdate(&doclist, d); if( doclist.nData<=CHUNK_MAX ){ rc = term_update(v, iIndexRow, &doclist); goto err; } /* Doclist doesn't fit, delete what's there, and accumulate ** forward. */ rc = term_delete(v, iIndexRow); if( rc!=SQLITE_OK ) goto err; /* Try to insert the doclist into a higher segment bucket. On ** failure, accumulate existing doclist with the doclist from that ** bucket, and put results in the next bucket. */ iSegment++; while( (rc=term_insert(v, &iIndexRow, pTerm, nTerm, iSegment, &doclist))!=SQLITE_OK ){ sqlite_int64 iSegmentRow; DocList old; int rc2; /* Retain old error in case the term_insert() error was really an ** error rather than a bounced insert. */ rc2 = term_select(v, pTerm, nTerm, iSegment, &iSegmentRow, &old); if( rc2!=SQLITE_ROW ) goto err; rc = term_delete(v, iSegmentRow); if( rc!=SQLITE_OK ) goto err; /* Reusing lowest-number deleted row keeps the index smaller. */ if( iSegmentRow<iIndexRow ) iIndexRow = iSegmentRow; /* doclist contains the newer data, so accumulate it over old. ** Then steal accumulated data for doclist. */ docListAccumulate(&old, &doclist); docListDestroy(&doclist); doclist = old; iSegment++; } err: docListDestroy(&doclist); return rc; } /* Add doclists for all terms in [pValues] to the hash table [terms]. */ static int insertTerms(fulltext_vtab *v, fts2Hash *terms, sqlite_int64 iRowid, sqlite3_value **pValues){ int i; for(i = 0; i < v->nColumn ; ++i){ char *zText = (char*)sqlite3_value_text(pValues[i]); int rc = buildTerms(v, terms, iRowid, zText, i); if( rc!=SQLITE_OK ) return rc; } return SQLITE_OK; } /* Add empty doclists for all terms in the given row's content to the hash * table [pTerms]. */ static int deleteTerms(fulltext_vtab *v, fts2Hash *pTerms, sqlite_int64 iRowid){ const char **pValues; int i; int rc = content_select(v, iRowid, &pValues); if( rc!=SQLITE_OK ) return rc; for(i = 0 ; i < v->nColumn; ++i) { rc = buildTerms(v, pTerms, iRowid, pValues[i], -1); if( rc!=SQLITE_OK ) break; } freeStringArray(v->nColumn, pValues); return SQLITE_OK; } /* Insert a row into the %_content table; set *piRowid to be the ID of the * new row. Fill [pTerms] with new doclists for the %_term table. */ static int index_insert(fulltext_vtab *v, sqlite3_value *pRequestRowid, sqlite3_value **pValues, sqlite_int64 *piRowid, fts2Hash *pTerms){ int rc; rc = content_insert(v, pRequestRowid, pValues); /* execute an SQL INSERT */ if( rc!=SQLITE_OK ) return rc; *piRowid = sqlite3_last_insert_rowid(v->db); return insertTerms(v, pTerms, *piRowid, pValues); } /* Delete a row from the %_content table; fill [pTerms] with empty doclists * to be written to the %_term table. */ static int index_delete(fulltext_vtab *v, sqlite_int64 iRow, fts2Hash *pTerms){ int rc = deleteTerms(v, pTerms, iRow); if( rc!=SQLITE_OK ) return rc; return content_delete(v, iRow); /* execute an SQL DELETE */ } /* Update a row in the %_content table; fill [pTerms] with new doclists for the * %_term table. */ static int index_update(fulltext_vtab *v, sqlite_int64 iRow, sqlite3_value **pValues, fts2Hash *pTerms){ /* Generate an empty doclist for each term that previously appeared in this * row. */ int rc = deleteTerms(v, pTerms, iRow); if( rc!=SQLITE_OK ) return rc; /* Now add positions for terms which appear in the updated row. */ rc = insertTerms(v, pTerms, iRow, pValues); if( rc!=SQLITE_OK ) return rc; return content_update(v, pValues, iRow); /* execute an SQL UPDATE */ } /* This function implements the xUpdate callback; it's the top-level entry * point for inserting, deleting or updating a row in a full-text table. */ static int fulltextUpdate(sqlite3_vtab *pVtab, int nArg, sqlite3_value **ppArg, sqlite_int64 *pRowid){ fulltext_vtab *v = (fulltext_vtab *) pVtab; fts2Hash terms; /* maps term string -> PosList */ int rc; fts2HashElem *e; TRACE(("FTS2 Update %p\n", pVtab)); fts2HashInit(&terms, FTS2_HASH_STRING, 1); if( nArg<2 ){ rc = index_delete(v, sqlite3_value_int64(ppArg[0]), &terms); } else if( sqlite3_value_type(ppArg[0]) != SQLITE_NULL ){ /* An update: * ppArg[0] = old rowid * ppArg[1] = new rowid * ppArg[2..2+v->nColumn-1] = values * ppArg[2+v->nColumn] = value for magic column (we ignore this) */ sqlite_int64 rowid = sqlite3_value_int64(ppArg[0]); if( sqlite3_value_type(ppArg[1]) != SQLITE_INTEGER || sqlite3_value_int64(ppArg[1]) != rowid ){ rc = SQLITE_ERROR; /* we don't allow changing the rowid */ } else { assert( nArg==2+v->nColumn+1); rc = index_update(v, rowid, &ppArg[2], &terms); } } else { /* An insert: * ppArg[1] = requested rowid * ppArg[2..2+v->nColumn-1] = values * ppArg[2+v->nColumn] = value for magic column (we ignore this) */ assert( nArg==2+v->nColumn+1); rc = index_insert(v, ppArg[1], &ppArg[2], pRowid, &terms); } if( rc==SQLITE_OK ){ /* Write updated doclists to disk. */ for(e=fts2HashFirst(&terms); e; e=fts2HashNext(e)){ DocList *p = fts2HashData(e); rc = index_insert_term(v, fts2HashKey(e), fts2HashKeysize(e), p); if( rc!=SQLITE_OK ) break; } } /* clean up */ for(e=fts2HashFirst(&terms); e; e=fts2HashNext(e)){ DocList *p = fts2HashData(e); docListDelete(p); } fts2HashClear(&terms); return rc; } /* ** Implementation of the snippet() function for FTS2 */ static void snippetFunc( sqlite3_context *pContext, int argc, sqlite3_value **argv ){ fulltext_cursor *pCursor; if( argc<1 ) return; if( sqlite3_value_type(argv[0])!=SQLITE_BLOB || sqlite3_value_bytes(argv[0])!=sizeof(pCursor) ){ sqlite3_result_error(pContext, "illegal first argument to html_snippet",-1); }else{ const char *zStart = "<b>"; const char *zEnd = "</b>"; const char *zEllipsis = "<b>...</b>"; memcpy(&pCursor, sqlite3_value_blob(argv[0]), sizeof(pCursor)); if( argc>=2 ){ zStart = (const char*)sqlite3_value_text(argv[1]); if( argc>=3 ){ zEnd = (const char*)sqlite3_value_text(argv[2]); if( argc>=4 ){ zEllipsis = (const char*)sqlite3_value_text(argv[3]); } } } snippetAllOffsets(pCursor); snippetText(pCursor, zStart, zEnd, zEllipsis); sqlite3_result_text(pContext, pCursor->snippet.zSnippet, pCursor->snippet.nSnippet, SQLITE_STATIC); } } /* ** Implementation of the offsets() function for FTS2 */ static void snippetOffsetsFunc( sqlite3_context *pContext, int argc, sqlite3_value **argv ){ fulltext_cursor *pCursor; if( argc<1 ) return; if( sqlite3_value_type(argv[0])!=SQLITE_BLOB || sqlite3_value_bytes(argv[0])!=sizeof(pCursor) ){ sqlite3_result_error(pContext, "illegal first argument to offsets",-1); }else{ memcpy(&pCursor, sqlite3_value_blob(argv[0]), sizeof(pCursor)); snippetAllOffsets(pCursor); snippetOffsetText(&pCursor->snippet); sqlite3_result_text(pContext, pCursor->snippet.zOffset, pCursor->snippet.nOffset, SQLITE_STATIC); } } /* ** This routine implements the xFindFunction method for the FTS2 ** virtual table. */ static int fulltextFindFunction( sqlite3_vtab *pVtab, int nArg, const char *zName, void (**pxFunc)(sqlite3_context*,int,sqlite3_value**), void **ppArg ){ if( strcmp(zName,"snippet")==0 ){ *pxFunc = snippetFunc; return 1; }else if( strcmp(zName,"offsets")==0 ){ *pxFunc = snippetOffsetsFunc; return 1; } return 0; } static const sqlite3_module fulltextModule = { /* iVersion */ 0, /* xCreate */ fulltextCreate, /* xConnect */ fulltextConnect, /* xBestIndex */ fulltextBestIndex, /* xDisconnect */ fulltextDisconnect, /* xDestroy */ fulltextDestroy, /* xOpen */ fulltextOpen, /* xClose */ fulltextClose, /* xFilter */ fulltextFilter, /* xNext */ fulltextNext, /* xEof */ fulltextEof, /* xColumn */ fulltextColumn, /* xRowid */ fulltextRowid, /* xUpdate */ fulltextUpdate, /* xBegin */ 0, /* xSync */ 0, /* xCommit */ 0, /* xRollback */ 0, /* xFindFunction */ fulltextFindFunction, }; int sqlite3Fts2Init(sqlite3 *db){ sqlite3_overload_function(db, "snippet", -1); sqlite3_overload_function(db, "offsets", -1); return sqlite3_create_module(db, "fts2", &fulltextModule, 0); } #if !SQLITE_CORE int sqlite3_extension_init(sqlite3 *db, char **pzErrMsg, const sqlite3_api_routines *pApi){ SQLITE_EXTENSION_INIT2(pApi) return sqlite3Fts2Init(db); } #endif #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) */ |
Added ext/fts2/fts2.h.
> > > > > > > > > > > | 1 2 3 4 5 6 7 8 9 10 11 | #include "sqlite3.h" #ifdef __cplusplus extern "C" { #endif /* __cplusplus */ int sqlite3Fts2Init(sqlite3 *db); #ifdef __cplusplus } /* extern "C" */ #endif /* __cplusplus */ |
Added ext/fts2/fts2_hash.c.
> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | /* ** 2001 September 22 ** ** The author disclaims copyright to this source code. In place of ** a legal notice, here is a blessing: ** ** May you do good and not evil. ** May you find forgiveness for yourself and forgive others. ** May you share freely, never taking more than you give. ** ************************************************************************* ** This is the implementation of generic hash-tables used in SQLite. ** We've modified it slightly to serve as a standalone hash table ** implementation for the full-text indexing module. */ #include <assert.h> #include <stdlib.h> #include <string.h> /* ** The code in this file is only compiled if: ** ** * The FTS2 module is being built as an extension ** (in which case SQLITE_CORE is not defined), or ** ** * The FTS2 module is being built into the core of ** SQLite (in which case SQLITE_ENABLE_FTS2 is defined). */ #if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) #include "fts2_hash.h" static void *malloc_and_zero(int n){ void *p = malloc(n); if( p ){ memset(p, 0, n); } return p; } /* Turn bulk memory into a hash table object by initializing the ** fields of the Hash structure. ** ** "pNew" is a pointer to the hash table that is to be initialized. ** keyClass is one of the constants ** FTS2_HASH_BINARY or FTS2_HASH_STRING. The value of keyClass ** determines what kind of key the hash table will use. "copyKey" is ** true if the hash table should make its own private copy of keys and ** false if it should just use the supplied pointer. */ void sqlite3Fts2HashInit(fts2Hash *pNew, int keyClass, int copyKey){ assert( pNew!=0 ); assert( keyClass>=FTS2_HASH_STRING && keyClass<=FTS2_HASH_BINARY ); pNew->keyClass = keyClass; pNew->copyKey = copyKey; pNew->first = 0; pNew->count = 0; pNew->htsize = 0; pNew->ht = 0; pNew->xMalloc = malloc_and_zero; pNew->xFree = free; } /* Remove all entries from a hash table. Reclaim all memory. ** Call this routine to delete a hash table or to reset a hash table ** to the empty state. */ void sqlite3Fts2HashClear(fts2Hash *pH){ fts2HashElem *elem; /* For looping over all elements of the table */ assert( pH!=0 ); elem = pH->first; pH->first = 0; if( pH->ht ) pH->xFree(pH->ht); pH->ht = 0; pH->htsize = 0; while( elem ){ fts2HashElem *next_elem = elem->next; if( pH->copyKey && elem->pKey ){ pH->xFree(elem->pKey); } pH->xFree(elem); elem = next_elem; } pH->count = 0; } /* ** Hash and comparison functions when the mode is FTS2_HASH_STRING */ static int strHash(const void *pKey, int nKey){ const char *z = (const char *)pKey; int h = 0; if( nKey<=0 ) nKey = (int) strlen(z); while( nKey > 0 ){ h = (h<<3) ^ h ^ *z++; nKey--; } return h & 0x7fffffff; } static int strCompare(const void *pKey1, int n1, const void *pKey2, int n2){ if( n1!=n2 ) return 1; return strncmp((const char*)pKey1,(const char*)pKey2,n1); } /* ** Hash and comparison functions when the mode is FTS2_HASH_BINARY */ static int binHash(const void *pKey, int nKey){ int h = 0; const char *z = (const char *)pKey; while( nKey-- > 0 ){ h = (h<<3) ^ h ^ *(z++); } return h & 0x7fffffff; } static int binCompare(const void *pKey1, int n1, const void *pKey2, int n2){ if( n1!=n2 ) return 1; return memcmp(pKey1,pKey2,n1); } /* ** Return a pointer to the appropriate hash function given the key class. ** ** The C syntax in this function definition may be unfamilar to some ** programmers, so we provide the following additional explanation: ** ** The name of the function is "hashFunction". The function takes a ** single parameter "keyClass". The return value of hashFunction() ** is a pointer to another function. Specifically, the return value ** of hashFunction() is a pointer to a function that takes two parameters ** with types "const void*" and "int" and returns an "int". */ static int (*hashFunction(int keyClass))(const void*,int){ if( keyClass==FTS2_HASH_STRING ){ return &strHash; }else{ assert( keyClass==FTS2_HASH_BINARY ); return &binHash; } } /* ** Return a pointer to the appropriate hash function given the key class. ** ** For help in interpreted the obscure C code in the function definition, ** see the header comment on the previous function. */ static int (*compareFunction(int keyClass))(const void*,int,const void*,int){ if( keyClass==FTS2_HASH_STRING ){ return &strCompare; }else{ assert( keyClass==FTS2_HASH_BINARY ); return &binCompare; } } /* Link an element into the hash table */ static void insertElement( fts2Hash *pH, /* The complete hash table */ struct _fts2ht *pEntry, /* The entry into which pNew is inserted */ fts2HashElem *pNew /* The element to be inserted */ ){ fts2HashElem *pHead; /* First element already in pEntry */ pHead = pEntry->chain; if( pHead ){ pNew->next = pHead; pNew->prev = pHead->prev; if( pHead->prev ){ pHead->prev->next = pNew; } else { pH->first = pNew; } pHead->prev = pNew; }else{ pNew->next = pH->first; if( pH->first ){ pH->first->prev = pNew; } pNew->prev = 0; pH->first = pNew; } pEntry->count++; pEntry->chain = pNew; } /* Resize the hash table so that it cantains "new_size" buckets. ** "new_size" must be a power of 2. The hash table might fail ** to resize if sqliteMalloc() fails. */ static void rehash(fts2Hash *pH, int new_size){ struct _fts2ht *new_ht; /* The new hash table */ fts2HashElem *elem, *next_elem; /* For looping over existing elements */ int (*xHash)(const void*,int); /* The hash function */ assert( (new_size & (new_size-1))==0 ); new_ht = (struct _fts2ht *)pH->xMalloc( new_size*sizeof(struct _fts2ht) ); if( new_ht==0 ) return; if( pH->ht ) pH->xFree(pH->ht); pH->ht = new_ht; pH->htsize = new_size; xHash = hashFunction(pH->keyClass); for(elem=pH->first, pH->first=0; elem; elem = next_elem){ int h = (*xHash)(elem->pKey, elem->nKey) & (new_size-1); next_elem = elem->next; insertElement(pH, &new_ht[h], elem); } } /* This function (for internal use only) locates an element in an ** hash table that matches the given key. The hash for this key has ** already been computed and is passed as the 4th parameter. */ static fts2HashElem *findElementGivenHash( const fts2Hash *pH, /* The pH to be searched */ const void *pKey, /* The key we are searching for */ int nKey, int h /* The hash for this key. */ ){ fts2HashElem *elem; /* Used to loop thru the element list */ int count; /* Number of elements left to test */ int (*xCompare)(const void*,int,const void*,int); /* comparison function */ if( pH->ht ){ struct _fts2ht *pEntry = &pH->ht[h]; elem = pEntry->chain; count = pEntry->count; xCompare = compareFunction(pH->keyClass); while( count-- && elem ){ if( (*xCompare)(elem->pKey,elem->nKey,pKey,nKey)==0 ){ return elem; } elem = elem->next; } } return 0; } /* Remove a single entry from the hash table given a pointer to that ** element and a hash on the element's key. */ static void removeElementGivenHash( fts2Hash *pH, /* The pH containing "elem" */ fts2HashElem* elem, /* The element to be removed from the pH */ int h /* Hash value for the element */ ){ struct _fts2ht *pEntry; if( elem->prev ){ elem->prev->next = elem->next; }else{ pH->first = elem->next; } if( elem->next ){ elem->next->prev = elem->prev; } pEntry = &pH->ht[h]; if( pEntry->chain==elem ){ pEntry->chain = elem->next; } pEntry->count--; if( pEntry->count<=0 ){ pEntry->chain = 0; } if( pH->copyKey && elem->pKey ){ pH->xFree(elem->pKey); } pH->xFree( elem ); pH->count--; if( pH->count<=0 ){ assert( pH->first==0 ); assert( pH->count==0 ); fts2HashClear(pH); } } /* Attempt to locate an element of the hash table pH with a key ** that matches pKey,nKey. Return the data for this element if it is ** found, or NULL if there is no match. */ void *sqlite3Fts2HashFind(const fts2Hash *pH, const void *pKey, int nKey){ int h; /* A hash on key */ fts2HashElem *elem; /* The element that matches key */ int (*xHash)(const void*,int); /* The hash function */ if( pH==0 || pH->ht==0 ) return 0; xHash = hashFunction(pH->keyClass); assert( xHash!=0 ); h = (*xHash)(pKey,nKey); assert( (pH->htsize & (pH->htsize-1))==0 ); elem = findElementGivenHash(pH,pKey,nKey, h & (pH->htsize-1)); return elem ? elem->data : 0; } /* Insert an element into the hash table pH. The key is pKey,nKey ** and the data is "data". ** ** If no element exists with a matching key, then a new ** element is created. A copy of the key is made if the copyKey ** flag is set. NULL is returned. ** ** If another element already exists with the same key, then the ** new data replaces the old data and the old data is returned. ** The key is not copied in this instance. If a malloc fails, then ** the new data is returned and the hash table is unchanged. ** ** If the "data" parameter to this function is NULL, then the ** element corresponding to "key" is removed from the hash table. */ void *sqlite3Fts2HashInsert( fts2Hash *pH, /* The hash table to insert into */ const void *pKey, /* The key */ int nKey, /* Number of bytes in the key */ void *data /* The data */ ){ int hraw; /* Raw hash value of the key */ int h; /* the hash of the key modulo hash table size */ fts2HashElem *elem; /* Used to loop thru the element list */ fts2HashElem *new_elem; /* New element added to the pH */ int (*xHash)(const void*,int); /* The hash function */ assert( pH!=0 ); xHash = hashFunction(pH->keyClass); assert( xHash!=0 ); hraw = (*xHash)(pKey, nKey); assert( (pH->htsize & (pH->htsize-1))==0 ); h = hraw & (pH->htsize-1); elem = findElementGivenHash(pH,pKey,nKey,h); if( elem ){ void *old_data = elem->data; if( data==0 ){ removeElementGivenHash(pH,elem,h); }else{ elem->data = data; } return old_data; } if( data==0 ) return 0; new_elem = (fts2HashElem*)pH->xMalloc( sizeof(fts2HashElem) ); if( new_elem==0 ) return data; if( pH->copyKey && pKey!=0 ){ new_elem->pKey = pH->xMalloc( nKey ); if( new_elem->pKey==0 ){ pH->xFree(new_elem); return data; } memcpy((void*)new_elem->pKey, pKey, nKey); }else{ new_elem->pKey = (void*)pKey; } new_elem->nKey = nKey; pH->count++; if( pH->htsize==0 ){ rehash(pH,8); if( pH->htsize==0 ){ pH->count = 0; pH->xFree(new_elem); return data; } } if( pH->count > pH->htsize ){ rehash(pH,pH->htsize*2); } assert( pH->htsize>0 ); assert( (pH->htsize & (pH->htsize-1))==0 ); h = hraw & (pH->htsize-1); insertElement(pH, &pH->ht[h], new_elem); new_elem->data = data; return 0; } #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) */ |
Added ext/fts2/fts2_hash.h.
> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | /* ** 2001 September 22 ** ** The author disclaims copyright to this source code. In place of ** a legal notice, here is a blessing: ** ** May you do good and not evil. ** May you find forgiveness for yourself and forgive others. ** May you share freely, never taking more than you give. ** ************************************************************************* ** This is the header file for the generic hash-table implemenation ** used in SQLite. We've modified it slightly to serve as a standalone ** hash table implementation for the full-text indexing module. ** */ #ifndef _FTS2_HASH_H_ #define _FTS2_HASH_H_ /* Forward declarations of structures. */ typedef struct fts2Hash fts2Hash; typedef struct fts2HashElem fts2HashElem; /* A complete hash table is an instance of the following structure. ** The internals of this structure are intended to be opaque -- client ** code should not attempt to access or modify the fields of this structure ** directly. Change this structure only by using the routines below. ** However, many of the "procedures" and "functions" for modifying and ** accessing this structure are really macros, so we can't really make ** this structure opaque. */ struct fts2Hash { char keyClass; /* HASH_INT, _POINTER, _STRING, _BINARY */ char copyKey; /* True if copy of key made on insert */ int count; /* Number of entries in this table */ fts2HashElem *first; /* The first element of the array */ void *(*xMalloc)(int); /* malloc() function to use */ void (*xFree)(void *); /* free() function to use */ int htsize; /* Number of buckets in the hash table */ struct _fts2ht { /* the hash table */ int count; /* Number of entries with this hash */ fts2HashElem *chain; /* Pointer to first entry with this hash */ } *ht; }; /* Each element in the hash table is an instance of the following ** structure. All elements are stored on a single doubly-linked list. ** ** Again, this structure is intended to be opaque, but it can't really ** be opaque because it is used by macros. */ struct fts2HashElem { fts2HashElem *next, *prev; /* Next and previous elements in the table */ void *data; /* Data associated with this element */ void *pKey; int nKey; /* Key associated with this element */ }; /* ** There are 2 different modes of operation for a hash table: ** ** FTS2_HASH_STRING pKey points to a string that is nKey bytes long ** (including the null-terminator, if any). Case ** is respected in comparisons. ** ** FTS2_HASH_BINARY pKey points to binary data nKey bytes long. ** memcmp() is used to compare keys. ** ** A copy of the key is made if the copyKey parameter to fts2HashInit is 1. */ #define FTS2_HASH_STRING 1 #define FTS2_HASH_BINARY 2 /* ** Access routines. To delete, insert a NULL pointer. */ void sqlite3Fts2HashInit(fts2Hash*, int keytype, int copyKey); void *sqlite3Fts2HashInsert(fts2Hash*, const void *pKey, int nKey, void *pData); void *sqlite3Fts2HashFind(const fts2Hash*, const void *pKey, int nKey); void sqlite3Fts2HashClear(fts2Hash*); /* ** Shorthand for the functions above */ #define fts2HashInit sqlite3Fts2HashInit #define fts2HashInsert sqlite3Fts2HashInsert #define fts2HashFind sqlite3Fts2HashFind #define fts2HashClear sqlite3Fts2HashClear /* ** Macros for looping over all elements of a hash table. The idiom is ** like this: ** ** fts2Hash h; ** fts2HashElem *p; ** ... ** for(p=fts2HashFirst(&h); p; p=fts2HashNext(p)){ ** SomeStructure *pData = fts2HashData(p); ** // do something with pData ** } */ #define fts2HashFirst(H) ((H)->first) #define fts2HashNext(E) ((E)->next) #define fts2HashData(E) ((E)->data) #define fts2HashKey(E) ((E)->pKey) #define fts2HashKeysize(E) ((E)->nKey) /* ** Number of entries in a hash table */ #define fts2HashCount(H) ((H)->count) #endif /* _FTS2_HASH_H_ */ |
Added ext/fts2/fts2_porter.c.
> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | /* ** 2006 September 30 ** ** The author disclaims copyright to this source code. In place of ** a legal notice, here is a blessing: ** ** May you do good and not evil. ** May you find forgiveness for yourself and forgive others. ** May you share freely, never taking more than you give. ** ************************************************************************* ** Implementation of the full-text-search tokenizer that implements ** a Porter stemmer. */ /* ** The code in this file is only compiled if: ** ** * The FTS2 module is being built as an extension ** (in which case SQLITE_CORE is not defined), or ** ** * The FTS2 module is being built into the core of ** SQLite (in which case SQLITE_ENABLE_FTS2 is defined). */ #if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) #include <assert.h> #if !defined(__APPLE__) #include <malloc.h> #else #include <stdlib.h> #endif #include <stdio.h> #include <string.h> #include <ctype.h> #include "fts2_tokenizer.h" /* ** Class derived from sqlite3_tokenizer */ typedef struct porter_tokenizer { sqlite3_tokenizer base; /* Base class */ } porter_tokenizer; /* ** Class derived from sqlit3_tokenizer_cursor */ typedef struct porter_tokenizer_cursor { sqlite3_tokenizer_cursor base; const char *zInput; /* input we are tokenizing */ int nInput; /* size of the input */ int iOffset; /* current position in zInput */ int iToken; /* index of next token to be returned */ char *zToken; /* storage for current token */ int nAllocated; /* space allocated to zToken buffer */ } porter_tokenizer_cursor; /* Forward declaration */ static const sqlite3_tokenizer_module porterTokenizerModule; /* ** Create a new tokenizer instance. */ static int porterCreate( int argc, const char * const *argv, sqlite3_tokenizer **ppTokenizer ){ porter_tokenizer *t; int i; for(i=0; i<argc; i++) printf("argv[%d] = %s\n", i, argv[i]); t = (porter_tokenizer *) calloc(sizeof(porter_tokenizer), 1); *ppTokenizer = &t->base; return SQLITE_OK; } /* ** Destroy a tokenizer */ static int porterDestroy(sqlite3_tokenizer *pTokenizer){ free(pTokenizer); return SQLITE_OK; } /* ** Prepare to begin tokenizing a particular string. The input ** string to be tokenized is zInput[0..nInput-1]. A cursor ** used to incrementally tokenize this string is returned in ** *ppCursor. */ static int porterOpen( sqlite3_tokenizer *pTokenizer, /* The tokenizer */ const char *zInput, int nInput, /* String to be tokenized */ sqlite3_tokenizer_cursor **ppCursor /* OUT: Tokenization cursor */ ){ porter_tokenizer_cursor *c; c = (porter_tokenizer_cursor *) malloc(sizeof(porter_tokenizer_cursor)); c->zInput = zInput; if( zInput==0 ){ c->nInput = 0; }else if( nInput<0 ){ c->nInput = (int)strlen(zInput); }else{ c->nInput = nInput; } c->iOffset = 0; /* start tokenizing at the beginning */ c->iToken = 0; c->zToken = NULL; /* no space allocated, yet. */ c->nAllocated = 0; *ppCursor = &c->base; return SQLITE_OK; } /* ** Close a tokenization cursor previously opened by a call to ** porterOpen() above. */ static int porterClose(sqlite3_tokenizer_cursor *pCursor){ porter_tokenizer_cursor *c = (porter_tokenizer_cursor *) pCursor; free(c->zToken); free(c); return SQLITE_OK; } /* ** Vowel or consonant */ static const char cType[] = { 0, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 2, 1 }; /* ** isConsonant() and isVowel() determine if their first character in ** the string they point to is a consonant or a vowel, according ** to Porter ruls. ** ** A consonate is any letter other than 'a', 'e', 'i', 'o', or 'u'. ** 'Y' is a consonant unless it follows another consonant, ** in which case it is a vowel. ** ** In these routine, the letters are in reverse order. So the 'y' rule ** is that 'y' is a consonant unless it is followed by another ** consonent. */ static int isVowel(const char*); static int isConsonant(const char *z){ int j; char x = *z; if( x==0 ) return 0; assert( x>='a' && x<='z' ); j = cType[x-'a']; if( j<2 ) return j; return z[1]==0 || isVowel(z + 1); } static int isVowel(const char *z){ int j; char x = *z; if( x==0 ) return 0; assert( x>='a' && x<='z' ); j = cType[x-'a']; if( j<2 ) return 1-j; return isConsonant(z + 1); } /* ** Let any sequence of one or more vowels be represented by V and let ** C be sequence of one or more consonants. Then every word can be ** represented as: ** ** [C] (VC){m} [V] ** ** In prose: A word is an optional consonant followed by zero or ** vowel-consonant pairs followed by an optional vowel. "m" is the ** number of vowel consonant pairs. This routine computes the value ** of m for the first i bytes of a word. ** ** Return true if the m-value for z is 1 or more. In other words, ** return true if z contains at least one vowel that is followed ** by a consonant. ** ** In this routine z[] is in reverse order. So we are really looking ** for an instance of of a consonant followed by a vowel. */ static int m_gt_0(const char *z){ while( isVowel(z) ){ z++; } if( *z==0 ) return 0; while( isConsonant(z) ){ z++; } return *z!=0; } /* Like mgt0 above except we are looking for a value of m which is ** exactly 1 */ static int m_eq_1(const char *z){ while( isVowel(z) ){ z++; } if( *z==0 ) return 0; while( isConsonant(z) ){ z++; } if( *z==0 ) return 0; while( isVowel(z) ){ z++; } if( *z==0 ) return 1; while( isConsonant(z) ){ z++; } return *z==0; } /* Like mgt0 above except we are looking for a value of m>1 instead ** or m>0 */ static int m_gt_1(const char *z){ while( isVowel(z) ){ z++; } if( *z==0 ) return 0; while( isConsonant(z) ){ z++; } if( *z==0 ) return 0; while( isVowel(z) ){ z++; } if( *z==0 ) return 0; while( isConsonant(z) ){ z++; } return *z!=0; } /* ** Return TRUE if there is a vowel anywhere within z[0..n-1] */ static int hasVowel(const char *z){ while( isConsonant(z) ){ z++; } return *z!=0; } /* ** Return TRUE if the word ends in a double consonant. ** ** The text is reversed here. So we are really looking at ** the first two characters of z[]. */ static int doubleConsonant(const char *z){ return isConsonant(z) && z[0]==z[1] && isConsonant(z+1); } /* ** Return TRUE if the word ends with three letters which ** are consonant-vowel-consonent and where the final consonant ** is not 'w', 'x', or 'y'. ** ** The word is reversed here. So we are really checking the ** first three letters and the first one cannot be in [wxy]. */ static int star_oh(const char *z){ return z[0]!=0 && isConsonant(z) && z[0]!='w' && z[0]!='x' && z[0]!='y' && z[1]!=0 && isVowel(z+1) && z[2]!=0 && isConsonant(z+2); } /* ** If the word ends with zFrom and xCond() is true for the stem ** of the word that preceeds the zFrom ending, then change the ** ending to zTo. ** ** The input word *pz and zFrom are both in reverse order. zTo ** is in normal order. ** ** Return TRUE if zFrom matches. Return FALSE if zFrom does not ** match. Not that TRUE is returned even if xCond() fails and ** no substitution occurs. */ static int stem( char **pz, /* The word being stemmed (Reversed) */ const char *zFrom, /* If the ending matches this... (Reversed) */ const char *zTo, /* ... change the ending to this (not reversed) */ int (*xCond)(const char*) /* Condition that must be true */ ){ char *z = *pz; while( *zFrom && *zFrom==*z ){ z++; zFrom++; } if( *zFrom!=0 ) return 0; if( xCond && !xCond(z) ) return 1; while( *zTo ){ *(--z) = *(zTo++); } *pz = z; return 1; } /* ** This is the fallback stemmer used when the porter stemmer is ** inappropriate. The input word is copied into the output with ** US-ASCII case folding. If the input word is too long (more ** than 20 bytes if it contains no digits or more than 6 bytes if ** it contains digits) then word is truncated to 20 or 6 bytes ** by taking 10 or 3 bytes from the beginning and end. */ static void copy_stemmer(const char *zIn, int nIn, char *zOut, int *pnOut){ int i, mx, j; int hasDigit = 0; for(i=0; i<nIn; i++){ int c = zIn[i]; if( c>='A' && c<='Z' ){ zOut[i] = c - 'A' + 'a'; }else{ if( c>='0' && c<='9' ) hasDigit = 1; zOut[i] = c; } } mx = hasDigit ? 3 : 10; if( nIn>mx*2 ){ for(j=mx, i=nIn-mx; i<nIn; i++, j++){ zOut[j] = zOut[i]; } i = j; } zOut[i] = 0; *pnOut = i; } /* ** Stem the input word zIn[0..nIn-1]. Store the output in zOut. ** zOut is at least big enough to hold nIn bytes. Write the actual ** size of the output word (exclusive of the '\0' terminator) into *pnOut. ** ** Any upper-case characters in the US-ASCII character set ([A-Z]) ** are converted to lower case. Upper-case UTF characters are ** unchanged. ** ** Words that are longer than about 20 bytes are stemmed by retaining ** a few bytes from the beginning and the end of the word. If the ** word contains digits, 3 bytes are taken from the beginning and ** 3 bytes from the end. For long words without digits, 10 bytes ** are taken from each end. US-ASCII case folding still applies. ** ** If the input word contains not digits but does characters not ** in [a-zA-Z] then no stemming is attempted and this routine just ** copies the input into the input into the output with US-ASCII ** case folding. ** ** Stemming never increases the length of the word. So there is ** no chance of overflowing the zOut buffer. */ static void porter_stemmer(const char *zIn, int nIn, char *zOut, int *pnOut){ int i, j, c; char zReverse[28]; char *z, *z2; if( nIn<3 || nIn>=sizeof(zReverse)-7 ){ /* The word is too big or too small for the porter stemmer. ** Fallback to the copy stemmer */ copy_stemmer(zIn, nIn, zOut, pnOut); return; } for(i=0, j=sizeof(zReverse)-6; i<nIn; i++, j--){ c = zIn[i]; if( c>='A' && c<='Z' ){ zReverse[j] = c + 'a' - 'A'; }else if( c>='a' && c<='z' ){ zReverse[j] = c; }else{ /* The use of a character not in [a-zA-Z] means that we fallback ** to the copy stemmer */ copy_stemmer(zIn, nIn, zOut, pnOut); return; } } memset(&zReverse[sizeof(zReverse)-5], 0, 5); z = &zReverse[j+1]; /* Step 1a */ if( z[0]=='s' ){ if( !stem(&z, "sess", "ss", 0) && !stem(&z, "sei", "i", 0) && !stem(&z, "ss", "ss", 0) ){ z++; } } /* Step 1b */ z2 = z; if( stem(&z, "dee", "ee", m_gt_0) ){ /* Do nothing. The work was all in the test */ }else if( (stem(&z, "gni", "", hasVowel) || stem(&z, "de", "", hasVowel)) && z!=z2 ){ if( stem(&z, "ta", "ate", 0) || stem(&z, "lb", "ble", 0) || stem(&z, "zi", "ize", 0) ){ /* Do nothing. The work was all in the test */ }else if( doubleConsonant(z) && (*z!='l' && *z!='s' && *z!='z') ){ z++; }else if( m_eq_1(z) && star_oh(z) ){ *(--z) = 'e'; } } /* Step 1c */ if( z[0]=='y' && hasVowel(z+1) ){ z[0] = 'i'; } /* Step 2 */ switch( z[1] ){ case 'a': stem(&z, "lanoita", "ate", m_gt_0) || stem(&z, "lanoit", "tion", m_gt_0); break; case 'c': stem(&z, "icne", "ence", m_gt_0) || stem(&z, "icna", "ance", m_gt_0); break; case 'e': stem(&z, "rezi", "ize", m_gt_0); break; case 'g': stem(&z, "igol", "log", m_gt_0); break; case 'l': stem(&z, "ilb", "ble", m_gt_0) || stem(&z, "illa", "al", m_gt_0) || stem(&z, "iltne", "ent", m_gt_0) || stem(&z, "ile", "e", m_gt_0) || stem(&z, "ilsuo", "ous", m_gt_0); break; case 'o': stem(&z, "noitazi", "ize", m_gt_0) || stem(&z, "noita", "ate", m_gt_0) || stem(&z, "rota", "ate", m_gt_0); break; case 's': stem(&z, "msila", "al", m_gt_0) || stem(&z, "ssenevi", "ive", m_gt_0) || stem(&z, "ssenluf", "ful", m_gt_0) || stem(&z, "ssensuo", "ous", m_gt_0); break; case 't': stem(&z, "itila", "al", m_gt_0) || stem(&z, "itivi", "ive", m_gt_0) || stem(&z, "itilib", "ble", m_gt_0); break; } /* Step 3 */ switch( z[0] ){ case 'e': stem(&z, "etaci", "ic", m_gt_0) || stem(&z, "evita", "", m_gt_0) || stem(&z, "ezila", "al", m_gt_0); break; case 'i': stem(&z, "itici", "ic", m_gt_0); break; case 'l': stem(&z, "laci", "ic", m_gt_0) || stem(&z, "luf", "", m_gt_0); break; case 's': stem(&z, "ssen", "", m_gt_0); break; } /* Step 4 */ switch( z[1] ){ case 'a': if( z[0]=='l' && m_gt_1(z+2) ){ z += 2; } break; case 'c': if( z[0]=='e' && z[2]=='n' && (z[3]=='a' || z[3]=='e') && m_gt_1(z+4) ){ z += 4; } break; case 'e': if( z[0]=='r' && m_gt_1(z+2) ){ z += 2; } break; case 'i': if( z[0]=='c' && m_gt_1(z+2) ){ z += 2; } break; case 'l': if( z[0]=='e' && z[2]=='b' && (z[3]=='a' || z[3]=='i') && m_gt_1(z+4) ){ z += 4; } break; case 'n': if( z[0]=='t' ){ if( z[2]=='a' ){ if( m_gt_1(z+3) ){ z += 3; } }else if( z[2]=='e' ){ stem(&z, "tneme", "", m_gt_1) || stem(&z, "tnem", "", m_gt_1) || stem(&z, "tne", "", m_gt_1); } } break; case 'o': if( z[0]=='u' ){ if( m_gt_1(z+2) ){ z += 2; } }else if( z[3]=='s' || z[3]=='t' ){ stem(&z, "noi", "", m_gt_1); } break; case 's': if( z[0]=='m' && z[2]=='i' && m_gt_1(z+3) ){ z += 3; } break; case 't': stem(&z, "eta", "", m_gt_1) || stem(&z, "iti", "", m_gt_1); break; case 'u': if( z[0]=='s' && z[2]=='o' && m_gt_1(z+3) ){ z += 3; } break; case 'v': case 'z': if( z[0]=='e' && z[2]=='i' && m_gt_1(z+3) ){ z += 3; } break; } /* Step 5a */ if( z[0]=='e' ){ if( m_gt_1(z+1) ){ z++; }else if( m_eq_1(z+1) && !star_oh(z+1) ){ z++; } } /* Step 5b */ if( m_gt_1(z) && z[0]=='l' && z[1]=='l' ){ z++; } /* z[] is now the stemmed word in reverse order. Flip it back ** around into forward order and return. */ *pnOut = i = strlen(z); zOut[i] = 0; while( *z ){ zOut[--i] = *(z++); } } /* ** Characters that can be part of a token. We assume any character ** whose value is greater than 0x80 (any UTF character) can be ** part of a token. In other words, delimiters all must have ** values of 0x7f or lower. */ const char isIdChar[] = { /* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, /* 3x */ 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 4x */ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, /* 5x */ 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 6x */ 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, /* 7x */ }; #define idChar(C) (((ch=C)&0x80)!=0 || (ch>0x2f && isIdChar[ch-0x30])) #define isDelim(C) (((ch=C)&0x80)==0 && (ch<0x30 || !isIdChar[ch-0x30])) /* ** Extract the next token from a tokenization cursor. The cursor must ** have been opened by a prior call to porterOpen(). */ static int porterNext( sqlite3_tokenizer_cursor *pCursor, /* Cursor returned by porterOpen */ const char **pzToken, /* OUT: *pzToken is the token text */ int *pnBytes, /* OUT: Number of bytes in token */ int *piStartOffset, /* OUT: Starting offset of token */ int *piEndOffset, /* OUT: Ending offset of token */ int *piPosition /* OUT: Position integer of token */ ){ porter_tokenizer_cursor *c = (porter_tokenizer_cursor *) pCursor; const char *z = c->zInput; while( c->iOffset<c->nInput ){ int iStartOffset, ch; /* Scan past delimiter characters */ while( c->iOffset<c->nInput && isDelim(z[c->iOffset]) ){ c->iOffset++; } /* Count non-delimiter characters. */ iStartOffset = c->iOffset; while( c->iOffset<c->nInput && !isDelim(z[c->iOffset]) ){ c->iOffset++; } if( c->iOffset>iStartOffset ){ int n = c->iOffset-iStartOffset; if( n>c->nAllocated ){ c->nAllocated = n+20; c->zToken = realloc(c->zToken, c->nAllocated); } porter_stemmer(&z[iStartOffset], n, c->zToken, pnBytes); *pzToken = c->zToken; *piStartOffset = iStartOffset; *piEndOffset = c->iOffset; *piPosition = c->iToken++; return SQLITE_OK; } } return SQLITE_DONE; } /* ** The set of routines that implement the porter-stemmer tokenizer */ static const sqlite3_tokenizer_module porterTokenizerModule = { 0, porterCreate, porterDestroy, porterOpen, porterClose, porterNext, }; /* ** Allocate a new porter tokenizer. Return a pointer to the new ** tokenizer in *ppModule */ void sqlite3Fts2PorterTokenizerModule( sqlite3_tokenizer_module const**ppModule ){ *ppModule = &porterTokenizerModule; } #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) */ |
Added ext/fts2/fts2_tokenizer.h.
> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | /* ** 2006 July 10 ** ** The author disclaims copyright to this source code. ** ************************************************************************* ** Defines the interface to tokenizers used by fulltext-search. There ** are three basic components: ** ** sqlite3_tokenizer_module is a singleton defining the tokenizer ** interface functions. This is essentially the class structure for ** tokenizers. ** ** sqlite3_tokenizer is used to define a particular tokenizer, perhaps ** including customization information defined at creation time. ** ** sqlite3_tokenizer_cursor is generated by a tokenizer to generate ** tokens from a particular input. */ #ifndef _FTS2_TOKENIZER_H_ #define _FTS2_TOKENIZER_H_ /* TODO(shess) Only used for SQLITE_OK and SQLITE_DONE at this time. ** If tokenizers are to be allowed to call sqlite3_*() functions, then ** we will need a way to register the API consistently. */ #include "sqlite3.h" /* ** Structures used by the tokenizer interface. */ typedef struct sqlite3_tokenizer sqlite3_tokenizer; typedef struct sqlite3_tokenizer_cursor sqlite3_tokenizer_cursor; typedef struct sqlite3_tokenizer_module sqlite3_tokenizer_module; struct sqlite3_tokenizer_module { int iVersion; /* currently 0 */ /* ** Create and destroy a tokenizer. argc/argv are passed down from ** the fulltext virtual table creation to allow customization. */ int (*xCreate)(int argc, const char *const*argv, sqlite3_tokenizer **ppTokenizer); int (*xDestroy)(sqlite3_tokenizer *pTokenizer); /* ** Tokenize a particular input. Call xOpen() to prepare to ** tokenize, xNext() repeatedly until it returns SQLITE_DONE, then ** xClose() to free any internal state. The pInput passed to ** xOpen() must exist until the cursor is closed. The ppToken ** result from xNext() is only valid until the next call to xNext() ** or until xClose() is called. */ /* TODO(shess) current implementation requires pInput to be ** nul-terminated. This should either be fixed, or pInput/nBytes ** should be converted to zInput. */ int (*xOpen)(sqlite3_tokenizer *pTokenizer, const char *pInput, int nBytes, sqlite3_tokenizer_cursor **ppCursor); int (*xClose)(sqlite3_tokenizer_cursor *pCursor); int (*xNext)(sqlite3_tokenizer_cursor *pCursor, const char **ppToken, int *pnBytes, int *piStartOffset, int *piEndOffset, int *piPosition); }; struct sqlite3_tokenizer { const sqlite3_tokenizer_module *pModule; /* The module for this tokenizer */ /* Tokenizer implementations will typically add additional fields */ }; struct sqlite3_tokenizer_cursor { sqlite3_tokenizer *pTokenizer; /* Tokenizer for this cursor. */ /* Tokenizer implementations will typically add additional fields */ }; /* ** Get the module for a tokenizer which generates tokens based on a ** set of non-token characters. The default is to break tokens at any ** non-alnum character, though the set of delimiters can also be ** specified by the first argv argument to xCreate(). */ /* TODO(shess) This doesn't belong here. Need some sort of ** registration process. */ void sqlite3Fts2SimpleTokenizerModule(sqlite3_tokenizer_module const**ppModule); void sqlite3Fts2PorterTokenizerModule(sqlite3_tokenizer_module const**ppModule); #endif /* _FTS2_TOKENIZER_H_ */ |
Added ext/fts2/fts2_tokenizer1.c.
> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | /* ** The author disclaims copyright to this source code. ** ************************************************************************* ** Implementation of the "simple" full-text-search tokenizer. */ /* ** The code in this file is only compiled if: ** ** * The FTS2 module is being built as an extension ** (in which case SQLITE_CORE is not defined), or ** ** * The FTS2 module is being built into the core of ** SQLite (in which case SQLITE_ENABLE_FTS2 is defined). */ #if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) #include <assert.h> #if !defined(__APPLE__) #include <malloc.h> #else #include <stdlib.h> #endif #include <stdio.h> #include <string.h> #include <ctype.h> #include "fts2_tokenizer.h" typedef struct simple_tokenizer { sqlite3_tokenizer base; char delim[128]; /* flag ASCII delimiters */ } simple_tokenizer; typedef struct simple_tokenizer_cursor { sqlite3_tokenizer_cursor base; const char *pInput; /* input we are tokenizing */ int nBytes; /* size of the input */ int iOffset; /* current position in pInput */ int iToken; /* index of next token to be returned */ char *pToken; /* storage for current token */ int nTokenAllocated; /* space allocated to zToken buffer */ } simple_tokenizer_cursor; /* Forward declaration */ static const sqlite3_tokenizer_module simpleTokenizerModule; static int isDelim(simple_tokenizer *t, unsigned char c){ return c<0x80 && t->delim[c]; } /* ** Create a new tokenizer instance. */ static int simpleCreate( int argc, const char * const *argv, sqlite3_tokenizer **ppTokenizer ){ simple_tokenizer *t; t = (simple_tokenizer *) calloc(sizeof(simple_tokenizer), 1); /* TODO(shess) Delimiters need to remain the same from run to run, ** else we need to reindex. One solution would be a meta-table to ** track such information in the database, then we'd only want this ** information on the initial create. */ if( argc>1 ){ int i, n = strlen(argv[1]); for(i=0; i<n; i++){ unsigned char ch = argv[1][i]; /* We explicitly don't support UTF-8 delimiters for now. */ if( ch>=0x80 ){ free(t); return SQLITE_ERROR; } t->delim[ch] = 1; } } else { /* Mark non-alphanumeric ASCII characters as delimiters */ int i; for(i=1; i<0x80; i++){ t->delim[i] = !isalnum(i); } } *ppTokenizer = &t->base; return SQLITE_OK; } /* ** Destroy a tokenizer */ static int simpleDestroy(sqlite3_tokenizer *pTokenizer){ free(pTokenizer); return SQLITE_OK; } /* ** Prepare to begin tokenizing a particular string. The input ** string to be tokenized is pInput[0..nBytes-1]. A cursor ** used to incrementally tokenize this string is returned in ** *ppCursor. */ static int simpleOpen( sqlite3_tokenizer *pTokenizer, /* The tokenizer */ const char *pInput, int nBytes, /* String to be tokenized */ sqlite3_tokenizer_cursor **ppCursor /* OUT: Tokenization cursor */ ){ simple_tokenizer_cursor *c; c = (simple_tokenizer_cursor *) malloc(sizeof(simple_tokenizer_cursor)); c->pInput = pInput; if( pInput==0 ){ c->nBytes = 0; }else if( nBytes<0 ){ c->nBytes = (int)strlen(pInput); }else{ c->nBytes = nBytes; } c->iOffset = 0; /* start tokenizing at the beginning */ c->iToken = 0; c->pToken = NULL; /* no space allocated, yet. */ c->nTokenAllocated = 0; *ppCursor = &c->base; return SQLITE_OK; } /* ** Close a tokenization cursor previously opened by a call to ** simpleOpen() above. */ static int simpleClose(sqlite3_tokenizer_cursor *pCursor){ simple_tokenizer_cursor *c = (simple_tokenizer_cursor *) pCursor; free(c->pToken); free(c); return SQLITE_OK; } /* ** Extract the next token from a tokenization cursor. The cursor must ** have been opened by a prior call to simpleOpen(). */ static int simpleNext( sqlite3_tokenizer_cursor *pCursor, /* Cursor returned by simpleOpen */ const char **ppToken, /* OUT: *ppToken is the token text */ int *pnBytes, /* OUT: Number of bytes in token */ int *piStartOffset, /* OUT: Starting offset of token */ int *piEndOffset, /* OUT: Ending offset of token */ int *piPosition /* OUT: Position integer of token */ ){ simple_tokenizer_cursor *c = (simple_tokenizer_cursor *) pCursor; simple_tokenizer *t = (simple_tokenizer *) pCursor->pTokenizer; unsigned char *p = (unsigned char *)c->pInput; while( c->iOffset<c->nBytes ){ int iStartOffset; /* Scan past delimiter characters */ while( c->iOffset<c->nBytes && isDelim(t, p[c->iOffset]) ){ c->iOffset++; } /* Count non-delimiter characters. */ iStartOffset = c->iOffset; while( c->iOffset<c->nBytes && !isDelim(t, p[c->iOffset]) ){ c->iOffset++; } if( c->iOffset>iStartOffset ){ int i, n = c->iOffset-iStartOffset; if( n>c->nTokenAllocated ){ c->nTokenAllocated = n+20; c->pToken = realloc(c->pToken, c->nTokenAllocated); } for(i=0; i<n; i++){ /* TODO(shess) This needs expansion to handle UTF-8 ** case-insensitivity. */ unsigned char ch = p[iStartOffset+i]; c->pToken[i] = ch<0x80 ? tolower(ch) : ch; } *ppToken = c->pToken; *pnBytes = n; *piStartOffset = iStartOffset; *piEndOffset = c->iOffset; *piPosition = c->iToken++; return SQLITE_OK; } } return SQLITE_DONE; } /* ** The set of routines that implement the simple tokenizer */ static const sqlite3_tokenizer_module simpleTokenizerModule = { 0, simpleCreate, simpleDestroy, simpleOpen, simpleClose, simpleNext, }; /* ** Allocate a new simple tokenizer. Return a pointer to the new ** tokenizer in *ppModule */ void sqlite3Fts2SimpleTokenizerModule( sqlite3_tokenizer_module const**ppModule ){ *ppModule = &simpleTokenizerModule; } #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2) */ |