Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
Overview
Comment: | Add experimental way to specify an alternative tokenizer when writing to or querying an fts5 table. |
---|---|
Downloads: | Tarball | ZIP archive | SQL archive |
Timelines: | family | ancestors | descendants | both | fts5-tokenize-blob |
Files: | files | file ages | folders |
SHA3-256: |
6c51c9c6a8a6a730c1d9e0119bc39ede |
User & Date: | dan 2024-04-15 20:24:50 |
Context
2024-04-16
| ||
14:23 | Add some tests for the fts5 tokenize-blob functionality on this branch. (check-in: c2f9d125 user: dan tags: fts5-tokenize-blob) | |
2024-04-15
| ||
20:24 | Add experimental way to specify an alternative tokenizer when writing to or querying an fts5 table. (check-in: 6c51c9c6 user: dan tags: fts5-tokenize-blob) | |
2024-04-12
| ||
18:46 | If a build fails in testrunner.tcl, do not attempt to run the jobs that depend on that build. Instead, report those jobs as having been skipped. (check-in: b40580be user: drh tags: trunk) | |
Changes
Changes to ext/fts5/fts5.h.
︙ | ︙ | |||
22 23 24 25 26 27 28 29 30 31 32 33 34 35 | #define _FTS5_H #include "sqlite3.h" #ifdef __cplusplus extern "C" { #endif /************************************************************************* ** CUSTOM AUXILIARY FUNCTIONS ** ** Virtual table implementations may overload SQL functions by implementing ** the sqlite3_module.xFindFunction() method. */ | > > | 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | #define _FTS5_H #include "sqlite3.h" #ifdef __cplusplus extern "C" { #endif #define SQLITE_FTS5_TOKENIZE_SUBTYPE ((unsigned int)'T') /************************************************************************* ** CUSTOM AUXILIARY FUNCTIONS ** ** Virtual table implementations may overload SQL functions by implementing ** the sqlite3_module.xFindFunction() method. */ |
︙ | ︙ |
Changes to ext/fts5/fts5Int.h.
︙ | ︙ | |||
139 140 141 142 143 144 145 146 147 148 149 150 151 152 | /************************************************************************** ** Interface to code in fts5_config.c. fts5_config.c contains contains code ** to parse the arguments passed to the CREATE VIRTUAL TABLE statement. */ typedef struct Fts5Config Fts5Config; /* ** An instance of the following structure encodes all information that can ** be gleaned from the CREATE VIRTUAL TABLE statement. ** ** And all information loaded from the %_config table. ** ** nAutomerge: | > > > > > > > > | 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 | /************************************************************************** ** Interface to code in fts5_config.c. fts5_config.c contains contains code ** to parse the arguments passed to the CREATE VIRTUAL TABLE statement. */ typedef struct Fts5Config Fts5Config; typedef struct Fts5TokenizerInst Fts5TokenizerInst; struct Fts5TokenizerInst { char *zSpec; Fts5Tokenizer *pTok; fts5_tokenizer *pTokApi; Fts5TokenizerInst *pNext; }; /* ** An instance of the following structure encodes all information that can ** be gleaned from the CREATE VIRTUAL TABLE statement. ** ** And all information loaded from the %_config table. ** ** nAutomerge: |
︙ | ︙ | |||
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 | ** are ignored. This value is configured using: ** ** INSERT INTO tbl(tbl, rank) VALUES('prefix-index', $bPrefixIndex); ** */ struct Fts5Config { sqlite3 *db; /* Database handle */ char *zDb; /* Database holding FTS index (e.g. "main") */ char *zName; /* Name of FTS index */ int nCol; /* Number of columns */ char **azCol; /* Column names */ u8 *abUnindexed; /* True for unindexed columns */ int nPrefix; /* Number of prefix indexes */ int *aPrefix; /* Sizes in bytes of nPrefix prefix indexes */ int eContent; /* An FTS5_CONTENT value */ int bContentlessDelete; /* "contentless_delete=" option (dflt==0) */ char *zContent; /* content table */ char *zContentRowid; /* "content_rowid=" option value */ int bColumnsize; /* "columnsize=" option value (dflt==1) */ int bTokendata; /* "tokendata=" option value (dflt==0) */ int eDetail; /* FTS5_DETAIL_XXX value */ char *zContentExprlist; | > | < | 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 | ** are ignored. This value is configured using: ** ** INSERT INTO tbl(tbl, rank) VALUES('prefix-index', $bPrefixIndex); ** */ struct Fts5Config { sqlite3 *db; /* Database handle */ Fts5Global *pGlobal; /* Database wide data */ char *zDb; /* Database holding FTS index (e.g. "main") */ char *zName; /* Name of FTS index */ int nCol; /* Number of columns */ char **azCol; /* Column names */ u8 *abUnindexed; /* True for unindexed columns */ int nPrefix; /* Number of prefix indexes */ int *aPrefix; /* Sizes in bytes of nPrefix prefix indexes */ int eContent; /* An FTS5_CONTENT value */ int bContentlessDelete; /* "contentless_delete=" option (dflt==0) */ char *zContent; /* content table */ char *zContentRowid; /* "content_rowid=" option value */ int bColumnsize; /* "columnsize=" option value (dflt==1) */ int bTokendata; /* "tokendata=" option value (dflt==0) */ int eDetail; /* FTS5_DETAIL_XXX value */ char *zContentExprlist; Fts5TokenizerInst *pTokList; int bLock; /* True when table is preparing statement */ int ePattern; /* FTS_PATTERN_XXX constant */ /* Values loaded from the %_config table */ int iVersion; /* fts5 file format 'version' */ int iCookie; /* Incremented when %_config is modified */ int pgsz; /* Approximate page size used in %_data */ |
︙ | ︙ | |||
249 250 251 252 253 254 255 256 257 258 259 260 261 262 | void sqlite3Fts5ConfigFree(Fts5Config*); int sqlite3Fts5ConfigDeclareVtab(Fts5Config *pConfig); int sqlite3Fts5Tokenize( Fts5Config *pConfig, /* FTS5 Configuration object */ int flags, /* FTS5_TOKENIZE_* flags */ const char *pText, int nText, /* Text to tokenize */ void *pCtx, /* Context passed to xToken() */ int (*xToken)(void*, int, const char*, int, int, int) /* Callback */ ); void sqlite3Fts5Dequote(char *z); | > > > > > > > > > > > > > > > | 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 | void sqlite3Fts5ConfigFree(Fts5Config*); int sqlite3Fts5ConfigDeclareVtab(Fts5Config *pConfig); int sqlite3Fts5Tokenize( Fts5Config *pConfig, /* FTS5 Configuration object */ int flags, /* FTS5_TOKENIZE_* flags */ const char *pText, int nText, /* Text to tokenize */ void *pCtx, /* Context passed to xToken() */ int (*xToken)(void*, int, const char*, int, int, int) /* Callback */ ); int sqlite3Fts5ConfigFindTokenizer( Fts5Config *pConfig, const char *z, Fts5TokenizerInst **ppOut ); int sqlite3Fts5SpecTokenize( Fts5Config *pConfig, /* FTS5 Configuration object */ const char *zSpec, /* Tokenizer specification */ int flags, /* FTS5_TOKENIZE_* flags */ const char *pText, int nText, /* Text to tokenize */ void *pCtx, /* Context passed to xToken() */ int (*xToken)(void*, int, const char*, int, int, int) /* Callback */ ); void sqlite3Fts5Dequote(char *z); |
︙ | ︙ | |||
594 595 596 597 598 599 600 | struct Fts5Table { sqlite3_vtab base; /* Base class used by SQLite core */ Fts5Config *pConfig; /* Virtual table configuration */ Fts5Index *pIndex; /* Full-text index */ }; int sqlite3Fts5GetTokenizer( | < < < | | 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 | struct Fts5Table { sqlite3_vtab base; /* Base class used by SQLite core */ Fts5Config *pConfig; /* Virtual table configuration */ Fts5Index *pIndex; /* Full-text index */ }; int sqlite3Fts5GetTokenizer( Fts5Config*, const char *zSpec ); Fts5Table *sqlite3Fts5TableFromCsrid(Fts5Global*, i64); int sqlite3Fts5FlushToDisk(Fts5Table*); /* |
︙ | ︙ | |||
709 710 711 712 713 714 715 716 717 718 719 720 721 722 | ); int sqlite3Fts5StorageDeleteAll(Fts5Storage *p); int sqlite3Fts5StorageRebuild(Fts5Storage *p); int sqlite3Fts5StorageOptimize(Fts5Storage *p); int sqlite3Fts5StorageMerge(Fts5Storage *p, int nMerge); int sqlite3Fts5StorageReset(Fts5Storage *p); /* ** End of interface to code in fts5_storage.c. **************************************************************************/ /************************************************************************** | > > > > > > > > | 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 | ); int sqlite3Fts5StorageDeleteAll(Fts5Storage *p); int sqlite3Fts5StorageRebuild(Fts5Storage *p); int sqlite3Fts5StorageOptimize(Fts5Storage *p); int sqlite3Fts5StorageMerge(Fts5Storage *p, int nMerge); int sqlite3Fts5StorageReset(Fts5Storage *p); int sqlite3Fts5UnpackTokenizeBlob( Fts5Config *pConfig, sqlite3_value *pVal, Fts5TokenizerInst **ppTok, char **pzText, int *pbDel ); /* ** End of interface to code in fts5_storage.c. **************************************************************************/ /************************************************************************** |
︙ | ︙ |
Changes to ext/fts5/fts5_config.c.
︙ | ︙ | |||
219 220 221 222 223 224 225 226 227 228 229 230 231 232 | iVal = aEnum[i].eVal; } } *peVal = iVal; return iVal<0 ? SQLITE_ERROR : SQLITE_OK; } /* ** Parse a "special" CREATE VIRTUAL TABLE directive and update ** configuration object pConfig as appropriate. ** ** If successful, object pConfig is updated and SQLITE_OK returned. If ** an error occurs, an SQLite error code is returned and an error message | > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | iVal = aEnum[i].eVal; } } *peVal = iVal; return iVal<0 ? SQLITE_ERROR : SQLITE_OK; } int sqlite3Fts5ConfigFindTokenizer( Fts5Config *pConfig, const char *z, Fts5TokenizerInst **ppOut ){ Fts5TokenizerInst *pRet = 0; int rc = SQLITE_OK; assert( pConfig->pzErrmsg ); /* Search for an existing tokenizer that matches this spec */ for(pRet=pConfig->pTokList; pRet; pRet=pRet->pNext){ if( strcmp(pRet->zSpec, z)==0 ) break; } if( pRet==0 ){ /* No tokenizer found - create one. */ rc = sqlite3Fts5GetTokenizer(pConfig, z); if( rc==SQLITE_OK ){ pRet = pConfig->pTokList->pNext; } } if( ppOut ) *ppOut = pRet; return rc; } /* ** Parse a "special" CREATE VIRTUAL TABLE directive and update ** configuration object pConfig as appropriate. ** ** If successful, object pConfig is updated and SQLITE_OK returned. If ** an error occurs, an SQLite error code is returned and an error message |
︙ | ︙ | |||
292 293 294 295 296 297 298 | bFirst = 0; } assert( pConfig->nPrefix<=FTS5_MAX_PREFIX_INDEXES ); return rc; } if( sqlite3_strnicmp("tokenize", zCmd, nCmd)==0 ){ | < < < < < < < | | | | < < < | < | < < < < < < < < < < < < < | < | < | < < < < < | 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 | bFirst = 0; } assert( pConfig->nPrefix<=FTS5_MAX_PREFIX_INDEXES ); return rc; } if( sqlite3_strnicmp("tokenize", zCmd, nCmd)==0 ){ if( pConfig->pTokList ){ *pzErr = sqlite3_mprintf("multiple tokenize=... directives"); rc = SQLITE_ERROR; }else{ assert( pConfig->pzErrmsg==0 ); pConfig->pzErrmsg = pzErr; rc = sqlite3Fts5GetTokenizer(pConfig, zArg); pConfig->pzErrmsg = 0; } return rc; } if( sqlite3_strnicmp("content", zCmd, nCmd)==0 ){ if( pConfig->eContent!=FTS5_CONTENT_NORMAL ){ *pzErr = sqlite3_mprintf("multiple content=... directives"); rc = SQLITE_ERROR; |
︙ | ︙ | |||
408 409 410 411 412 413 414 | return rc; } *pzErr = sqlite3_mprintf("unrecognized option: \"%.*s\"", nCmd, zCmd); return SQLITE_ERROR; } | < < < < < < < < < < | 404 405 406 407 408 409 410 411 412 413 414 415 416 417 | return rc; } *pzErr = sqlite3_mprintf("unrecognized option: \"%.*s\"", nCmd, zCmd); return SQLITE_ERROR; } /* ** Gobble up the first bareword or quoted word from the input buffer zIn. ** Return a pointer to the character immediately following the last in ** the gobbled word if successful, or a NULL pointer otherwise (failed ** to find close-quote character). ** ** Before returning, set pzOut to point to a new buffer containing a |
︙ | ︙ | |||
551 552 553 554 555 556 557 558 559 560 561 562 563 564 | int i; sqlite3_int64 nByte; *ppOut = pRet = (Fts5Config*)sqlite3_malloc(sizeof(Fts5Config)); if( pRet==0 ) return SQLITE_NOMEM; memset(pRet, 0, sizeof(Fts5Config)); pRet->db = db; pRet->iCookie = -1; nByte = nArg * (sizeof(char*) + sizeof(u8)); pRet->azCol = (char**)sqlite3Fts5MallocZero(&rc, nByte); pRet->abUnindexed = pRet->azCol ? (u8*)&pRet->azCol[nArg] : 0; pRet->zDb = sqlite3Fts5Strndup(&rc, azArg[1], -1); pRet->zName = sqlite3Fts5Strndup(&rc, azArg[2], -1); | > | 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 | int i; sqlite3_int64 nByte; *ppOut = pRet = (Fts5Config*)sqlite3_malloc(sizeof(Fts5Config)); if( pRet==0 ) return SQLITE_NOMEM; memset(pRet, 0, sizeof(Fts5Config)); pRet->db = db; pRet->pGlobal = pGlobal; pRet->iCookie = -1; nByte = nArg * (sizeof(char*) + sizeof(u8)); pRet->azCol = (char**)sqlite3Fts5MallocZero(&rc, nByte); pRet->abUnindexed = pRet->azCol ? (u8*)&pRet->azCol[nArg] : 0; pRet->zDb = sqlite3Fts5Strndup(&rc, azArg[1], -1); pRet->zName = sqlite3Fts5Strndup(&rc, azArg[2], -1); |
︙ | ︙ | |||
639 640 641 642 643 644 645 | ); rc = SQLITE_ERROR; } /* If a tokenizer= option was successfully parsed, the tokenizer has ** already been allocated. Otherwise, allocate an instance of the default ** tokenizer (unicode61) now. */ | | | | 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 | ); rc = SQLITE_ERROR; } /* If a tokenizer= option was successfully parsed, the tokenizer has ** already been allocated. Otherwise, allocate an instance of the default ** tokenizer (unicode61) now. */ if( rc==SQLITE_OK && pRet->pTokList==0 ){ rc = sqlite3Fts5GetTokenizer(pRet, 0); } /* If no zContent option was specified, fill in the default values. */ if( rc==SQLITE_OK && pRet->zContent==0 ){ const char *zTail = 0; assert( pRet->eContent==FTS5_CONTENT_NORMAL || pRet->eContent==FTS5_CONTENT_NONE |
︙ | ︙ | |||
677 678 679 680 681 682 683 684 685 686 687 688 689 690 | if( rc!=SQLITE_OK ){ sqlite3Fts5ConfigFree(pRet); *ppOut = 0; } return rc; } /* ** Free the configuration object passed as the only argument. */ void sqlite3Fts5ConfigFree(Fts5Config *pConfig){ if( pConfig ){ int i; | > > > > > > > > > > > | < < > | 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 | if( rc!=SQLITE_OK ){ sqlite3Fts5ConfigFree(pRet); *ppOut = 0; } return rc; } static void fts5ConfigFreeTokenizers(Fts5Config *pConfig){ Fts5TokenizerInst *p = pConfig->pTokList; while( p ){ Fts5TokenizerInst *pNext = p->pNext; p->pTokApi->xDelete(p->pTok); sqlite3_free(p); p = pNext; } pConfig->pTokList = 0; } /* ** Free the configuration object passed as the only argument. */ void sqlite3Fts5ConfigFree(Fts5Config *pConfig){ if( pConfig ){ int i; fts5ConfigFreeTokenizers(pConfig); sqlite3_free(pConfig->zDb); sqlite3_free(pConfig->zName); fts5ConfigFreeTokenizers(pConfig); for(i=0; i<pConfig->nCol; i++){ sqlite3_free(pConfig->azCol[i]); } sqlite3_free(pConfig->azCol); sqlite3_free(pConfig->aPrefix); sqlite3_free(pConfig->zRank); sqlite3_free(pConfig->zRankArgs); |
︙ | ︙ | |||
761 762 763 764 765 766 767 | int sqlite3Fts5Tokenize( Fts5Config *pConfig, /* FTS5 Configuration object */ int flags, /* FTS5_TOKENIZE_* flags */ const char *pText, int nText, /* Text to tokenize */ void *pCtx, /* Context passed to xToken() */ int (*xToken)(void*, int, const char*, int, int, int) /* Callback */ ){ | | | | > > > > > > > > > > > | > > > > > > > > > | 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 | int sqlite3Fts5Tokenize( Fts5Config *pConfig, /* FTS5 Configuration object */ int flags, /* FTS5_TOKENIZE_* flags */ const char *pText, int nText, /* Text to tokenize */ void *pCtx, /* Context passed to xToken() */ int (*xToken)(void*, int, const char*, int, int, int) /* Callback */ ){ if( pText ){ Fts5TokenizerInst *p = pConfig->pTokList; return p->pTokApi->xTokenize(p->pTok, pCtx, flags, pText, nText, xToken); } return SQLITE_OK; } int sqlite3Fts5SpecTokenize( Fts5Config *pConfig, /* FTS5 Configuration object */ const char *zSpec, /* Tokenizer specification */ int flags, /* FTS5_TOKENIZE_* flags */ const char *pText, int nText, /* Text to tokenize */ void *pCtx, /* Context passed to xToken() */ int (*xToken)(void*, int, const char*, int, int, int) /* Callback */ ){ if( pText ){ Fts5TokenizerInst *p = pConfig->pTokList; if( zSpec ){ int rc = sqlite3Fts5ConfigFindTokenizer(pConfig, zSpec, &p); if( rc!=SQLITE_OK ) return rc; } return p->pTokApi->xTokenize(p->pTok, pCtx, flags, pText, nText, xToken); } return SQLITE_OK; } /* ** Argument pIn points to the first character in what is expected to be ** a comma-separated list of SQL literals followed by a ')' character. ** If it actually is this, return a pointer to the ')'. Otherwise, return ** NULL to indicate a parse error. |
︙ | ︙ |
Changes to ext/fts5/fts5_main.c.
︙ | ︙ | |||
111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 | void (*xDestroy)(void*); /* Destructor function */ Fts5TokenizerModule *pNext; /* Next registered tokenizer module */ }; struct Fts5FullTable { Fts5Table p; /* Public class members from fts5Int.h */ Fts5Storage *pStorage; /* Document store */ Fts5Global *pGlobal; /* Global (connection wide) data */ Fts5Cursor *pSortCsr; /* Sort data from this cursor */ int iSavepoint; /* Successful xSavepoint()+1 */ #ifdef SQLITE_DEBUG struct Fts5TransactionState ts; #endif }; | > > | 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 | void (*xDestroy)(void*); /* Destructor function */ Fts5TokenizerModule *pNext; /* Next registered tokenizer module */ }; struct Fts5FullTable { Fts5Table p; /* Public class members from fts5Int.h */ Fts5Storage *pStorage; /* Document store */ #if 0 Fts5Global *pGlobal; /* Global (connection wide) data */ #endif Fts5Cursor *pSortCsr; /* Sort data from this cursor */ int iSavepoint; /* Successful xSavepoint()+1 */ #ifdef SQLITE_DEBUG struct Fts5TransactionState ts; #endif }; |
︙ | ︙ | |||
374 375 376 377 378 379 380 | pTab = (Fts5FullTable*)sqlite3Fts5MallocZero(&rc, sizeof(Fts5FullTable)); if( rc==SQLITE_OK ){ rc = sqlite3Fts5ConfigParse(pGlobal, db, argc, azConfig, &pConfig, pzErr); assert( (rc==SQLITE_OK && *pzErr==0) || pConfig==0 ); } if( rc==SQLITE_OK ){ pTab->p.pConfig = pConfig; | < | 376 377 378 379 380 381 382 383 384 385 386 387 388 389 | pTab = (Fts5FullTable*)sqlite3Fts5MallocZero(&rc, sizeof(Fts5FullTable)); if( rc==SQLITE_OK ){ rc = sqlite3Fts5ConfigParse(pGlobal, db, argc, azConfig, &pConfig, pzErr); assert( (rc==SQLITE_OK && *pzErr==0) || pConfig==0 ); } if( rc==SQLITE_OK ){ pTab->p.pConfig = pConfig; } /* Open the index sub-system */ if( rc==SQLITE_OK ){ rc = sqlite3Fts5IndexOpen(pConfig, bCreate, &pTab->p.pIndex, pzErr); } |
︙ | ︙ | |||
689 690 691 692 693 694 695 | pInfo->idxNum = idxFlags; return SQLITE_OK; } static int fts5NewTransaction(Fts5FullTable *pTab){ Fts5Cursor *pCsr; | | | | 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 | pInfo->idxNum = idxFlags; return SQLITE_OK; } static int fts5NewTransaction(Fts5FullTable *pTab){ Fts5Cursor *pCsr; for(pCsr=pTab->p.pConfig->pGlobal->pCsr; pCsr; pCsr=pCsr->pNext){ if( pCsr->base.pVtab==(sqlite3_vtab*)pTab ) return SQLITE_OK; } return sqlite3Fts5StorageReset(pTab->pStorage); } /* ** Implementation of xOpen method. */ static int fts5OpenMethod(sqlite3_vtab *pVTab, sqlite3_vtab_cursor **ppCsr){ Fts5FullTable *pTab = (Fts5FullTable*)pVTab; Fts5Config *pConfig = pTab->p.pConfig; Fts5Cursor *pCsr = 0; /* New cursor object */ sqlite3_int64 nByte; /* Bytes of space to allocate */ int rc; /* Return code */ rc = fts5NewTransaction(pTab); if( rc==SQLITE_OK ){ nByte = sizeof(Fts5Cursor) + pConfig->nCol * sizeof(int); pCsr = (Fts5Cursor*)sqlite3_malloc64(nByte); if( pCsr ){ Fts5Global *pGlobal = pConfig->pGlobal; memset(pCsr, 0, (size_t)nByte); pCsr->aColumnSize = (int*)&pCsr[1]; pCsr->pNext = pGlobal->pCsr; pGlobal->pCsr = pCsr; pCsr->iCsrId = ++pGlobal->iNextId; }else{ rc = SQLITE_NOMEM; |
︙ | ︙ | |||
797 798 799 800 801 802 803 | if( pCursor ){ Fts5FullTable *pTab = (Fts5FullTable*)(pCursor->pVtab); Fts5Cursor *pCsr = (Fts5Cursor*)pCursor; Fts5Cursor **pp; fts5FreeCursorComponents(pCsr); /* Remove the cursor from the Fts5Global.pCsr list */ | | | 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 | if( pCursor ){ Fts5FullTable *pTab = (Fts5FullTable*)(pCursor->pVtab); Fts5Cursor *pCsr = (Fts5Cursor*)pCursor; Fts5Cursor **pp; fts5FreeCursorComponents(pCsr); /* Remove the cursor from the Fts5Global.pCsr list */ for(pp=&pTab->p.pConfig->pGlobal->pCsr; (*pp)!=pCsr; pp=&(*pp)->pNext); *pp = pCsr->pNext; sqlite3_free(pCsr); } return SQLITE_OK; } |
︙ | ︙ | |||
850 851 852 853 854 855 856 | /* ** Set the FTS5CSR_REQUIRE_RESEEK flag on all FTS5_PLAN_MATCH cursors ** open on table pTab. */ static void fts5TripCursors(Fts5FullTable *pTab){ Fts5Cursor *pCsr; | | | 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 | /* ** Set the FTS5CSR_REQUIRE_RESEEK flag on all FTS5_PLAN_MATCH cursors ** open on table pTab. */ static void fts5TripCursors(Fts5FullTable *pTab){ Fts5Cursor *pCsr; for(pCsr=pTab->p.pConfig->pGlobal->pCsr; pCsr; pCsr=pCsr->pNext){ if( pCsr->ePlan==FTS5_PLAN_MATCH && pCsr->base.pVtab==(sqlite3_vtab*)pTab ){ CsrFlagSet(pCsr, FTS5CSR_REQUIRE_RESEEK); } } } |
︙ | ︙ | |||
1102 1103 1104 1105 1106 1107 1108 | ** Search for an auxiliary function named zName that can be used with table ** pTab. If one is found, return a pointer to the corresponding Fts5Auxiliary ** structure. Otherwise, if no such function exists, return NULL. */ static Fts5Auxiliary *fts5FindAuxiliary(Fts5FullTable *pTab, const char *zName){ Fts5Auxiliary *pAux; | | | 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 | ** Search for an auxiliary function named zName that can be used with table ** pTab. If one is found, return a pointer to the corresponding Fts5Auxiliary ** structure. Otherwise, if no such function exists, return NULL. */ static Fts5Auxiliary *fts5FindAuxiliary(Fts5FullTable *pTab, const char *zName){ Fts5Auxiliary *pAux; for(pAux=pTab->p.pConfig->pGlobal->pAux; pAux; pAux=pAux->pNext){ if( sqlite3_stricmp(zName, pAux->zFunc)==0 ) return pAux; } /* No function of the specified name was found. Return 0. */ return 0; } |
︙ | ︙ | |||
1273 1274 1275 1276 1277 1278 1279 | /* Decode the arguments passed through to this function. */ for(i=0; i<nVal; i++){ switch( idxStr[iIdxStr++] ){ case 'r': pRank = apVal[i]; break; case 'M': { | > | > > > > > > > < > > | 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 | /* Decode the arguments passed through to this function. */ for(i=0; i<nVal; i++){ switch( idxStr[iIdxStr++] ){ case 'r': pRank = apVal[i]; break; case 'M': { Fts5TokenizerInst *pInst = 0; char *zText = 0; int bDel = 0; rc = sqlite3Fts5UnpackTokenizeBlob( pConfig, apVal[i], &pInst, &zText, &bDel ); if( zText==0 ) zText = ""; iCol = 0; do{ iCol = iCol*10 + (idxStr[iIdxStr]-'0'); iIdxStr++; }while( idxStr[iIdxStr]>='0' && idxStr[iIdxStr]<='9' ); if( zText[0]=='*' ){ /* The user has issued a query of the form "MATCH '*...'". This ** indicates that the MATCH expression is not a full text query, ** but a request for an internal parameter. */ rc = fts5SpecialMatch(pTab, pCsr, &zText[1]); if( bDel ) sqlite3_free(zText); goto filter_out; }else{ char **pzErr = &pTab->p.base.zErrMsg; rc = sqlite3Fts5ExprNew(pConfig, 0, iCol, zText, &pExpr, pzErr); if( rc==SQLITE_OK ){ rc = sqlite3Fts5ExprAnd(&pCsr->pExpr, pExpr); pExpr = 0; } } if( bDel ) sqlite3_free(zText); if( rc!=SQLITE_OK ) goto filter_out; break; } case 'L': case 'G': { int bGlob = (idxStr[iIdxStr-1]=='G'); const char *zText = (const char*)sqlite3_value_text(apVal[i]); iCol = 0; |
︙ | ︙ | |||
2857 2858 2859 2860 2861 2862 2863 | rc = SQLITE_ERROR; } return rc; } int sqlite3Fts5GetTokenizer( | < < < | < > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | | | | | | > > > > > > > > | | | | | > | > | | | | | | < | > > | > | | > > | > > > > > | 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 | rc = SQLITE_ERROR; } return rc; } int sqlite3Fts5GetTokenizer( Fts5Config *pConfig, const char *zSpec ){ int rc = SQLITE_OK; char **pzErr = pConfig->pzErrmsg; const char **azArg = 0; char *pDel = 0; sqlite3_int64 nArg = 0; assert( pzErr || (zSpec==0 && pConfig->pTokList==0) ); if( zSpec ){ const char *p = (const char*)zSpec; char *pSpace = 0; nArg = strlen(zSpec) + 1; pDel = sqlite3Fts5MallocZero(&rc, nArg * 2); pSpace = pDel; azArg = (const char**)sqlite3Fts5MallocZero(&rc, sizeof(char*) * nArg); if( azArg && pSpace ){ for(nArg=0; p && *p; nArg++){ const char *p2 = fts5ConfigSkipWhitespace(p); if( *p2=='\'' ){ p = fts5ConfigSkipLiteral(p2); }else{ p = fts5ConfigSkipBareword(p2); } if( p ){ memcpy(pSpace, p2, p-p2); azArg[nArg] = pSpace; sqlite3Fts5Dequote(pSpace); pSpace += (p - p2) + 1; p = fts5ConfigSkipWhitespace(p); } } if( p==0 ){ *pzErr= sqlite3_mprintf("parse error in tokenize directive"); rc = SQLITE_ERROR; } } } if( rc==SQLITE_OK ){ Fts5TokenizerModule *pMod; pMod = fts5LocateTokenizer(pConfig->pGlobal, nArg==0 ? 0 : azArg[0]); if( pMod==0 ){ assert( nArg>0 ); rc = SQLITE_ERROR; *pzErr = sqlite3_mprintf("no such tokenizer: %s", azArg[0]); }else{ int nSpec = zSpec ? strlen(zSpec) + 1 : 0; int nByte = sizeof(Fts5TokenizerInst) + nSpec; Fts5TokenizerInst *pNew = sqlite3Fts5MallocZero(&rc, nByte); if( pNew ){ if( zSpec ){ pNew->zSpec = (char*)&pNew[1]; memcpy(pNew->zSpec, zSpec, nSpec); } rc = pMod->x.xCreate( pMod->pUserData, (azArg?&azArg[1]:0), (nArg?nArg-1:0), &pNew->pTok ); pNew->pTokApi = &pMod->x; if( rc!=SQLITE_OK ){ if( pzErr ){ *pzErr = sqlite3_mprintf("error in tokenizer constructor"); } }else if( pConfig->pTokList==0 ){ pConfig->ePattern = sqlite3Fts5TokenizerPattern( pMod->x.xCreate, pNew->pTok ); } } if( rc==SQLITE_OK ){ if( pConfig->pTokList ){ pNew->pNext = pConfig->pTokList->pNext; pConfig->pTokList->pNext = pNew; }else{ pConfig->pTokList = pNew; } }else{ sqlite3_free(pNew); } } } sqlite3_free(azArg); sqlite3_free(pDel); return rc; } static void fts5ModuleDestroy(void *pCtx){ Fts5TokenizerModule *pTok, *pNextTok; Fts5Auxiliary *pAux, *pNextAux; Fts5Global *pGlobal = (Fts5Global*)pCtx; |
︙ | ︙ |
Changes to ext/fts5/fts5_storage.c.
︙ | ︙ | |||
394 395 396 397 398 399 400 401 402 403 404 405 406 407 | UNUSED_PARAM2(iUnused1, iUnused2); if( nToken>FTS5_MAX_TOKEN_SIZE ) nToken = FTS5_MAX_TOKEN_SIZE; if( (tflags & FTS5_TOKEN_COLOCATED)==0 || pCtx->szCol==0 ){ pCtx->szCol++; } return sqlite3Fts5IndexWrite(pIdx, pCtx->iCol, pCtx->szCol-1, pToken, nToken); } /* ** If a row with rowid iDel is present in the %_content table, add the ** delete-markers to the FTS index necessary to delete it. Do not actually ** remove the %_content row at this time though. */ static int fts5StorageDeleteFromIndex( | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | UNUSED_PARAM2(iUnused1, iUnused2); if( nToken>FTS5_MAX_TOKEN_SIZE ) nToken = FTS5_MAX_TOKEN_SIZE; if( (tflags & FTS5_TOKEN_COLOCATED)==0 || pCtx->szCol==0 ){ pCtx->szCol++; } return sqlite3Fts5IndexWrite(pIdx, pCtx->iCol, pCtx->szCol-1, pToken, nToken); } #define IS_TOKENIZE_BLOB(pVal) ( \ sqlite3_value_subtype(pVal)==SQLITE_FTS5_TOKENIZE_SUBTYPE \ && sqlite3_value_type(pVal)==SQLITE_BLOB \ ) static const char *fts5UnpackTokenizeBlob( sqlite3_value *pVal, const char **pzT, int *pnT ){ const u8 *pBlob = sqlite3_value_blob(pVal); int nBlob = sqlite3_value_bytes(pVal); int ii; assert( sqlite3_value_subtype(pVal)==SQLITE_FTS5_TOKENIZE_SUBTYPE ); assert( sqlite3_value_type(pVal)==SQLITE_BLOB ); for(ii=0; pBlob[ii]; ii++){ if( ii==nBlob ){ *pzT = 0; *pnT = 0; return 0; } } *pzT = (const char*)&pBlob[ii+1]; *pnT = nBlob - ii - 1; return (const char*)pBlob; } /* ** If a row with rowid iDel is present in the %_content table, add the ** delete-markers to the FTS index necessary to delete it. Do not actually ** remove the %_content row at this time though. */ static int fts5StorageDeleteFromIndex( |
︙ | ︙ | |||
425 426 427 428 429 430 431 | } } ctx.pStorage = p; ctx.iCol = -1; for(iCol=1; rc==SQLITE_OK && iCol<=pConfig->nCol; iCol++){ if( pConfig->abUnindexed[iCol-1]==0 ){ | | > | > > > > | | > | | 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 | } } ctx.pStorage = p; ctx.iCol = -1; for(iCol=1; rc==SQLITE_OK && iCol<=pConfig->nCol; iCol++){ if( pConfig->abUnindexed[iCol-1]==0 ){ const char *zText = 0; const char *zTok = 0; int nText = 0; assert( pSeek==0 || apVal==0 ); assert( pSeek!=0 || apVal!=0 ); if( pSeek ){ zText = (const char*)sqlite3_column_text(pSeek, iCol); nText = sqlite3_column_bytes(pSeek, iCol); }else if( ALWAYS(apVal) ){ sqlite3_value *pVal = apVal[iCol-1]; if( IS_TOKENIZE_BLOB(pVal) ){ zTok = fts5UnpackTokenizeBlob(pVal, &zText, &nText); }else{ zText = (const char*)sqlite3_value_text(apVal[iCol-1]); nText = sqlite3_value_bytes(apVal[iCol-1]); } }else{ continue; } ctx.szCol = 0; rc = sqlite3Fts5SpecTokenize(pConfig, zTok, FTS5_TOKENIZE_DOCUMENT, zText, nText, (void*)&ctx, fts5StorageInsertCallback ); p->aTotalSize[iCol-1] -= (i64)ctx.szCol; if( p->aTotalSize[iCol-1]<0 ){ rc = FTS5_CORRUPT; } } |
︙ | ︙ | |||
747 748 749 750 751 752 753 754 755 756 757 758 759 760 | } if( rc==SQLITE_OK ){ *piRowid = sqlite3_last_insert_rowid(p->pConfig->db); } } return rc; } /* ** Insert a new row into the FTS content table. */ int sqlite3Fts5StorageContentInsert( Fts5Storage *p, sqlite3_value **apVal, | > > > > > > > > > > > > > > > > > > > > > > > > | 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 | } if( rc==SQLITE_OK ){ *piRowid = sqlite3_last_insert_rowid(p->pConfig->db); } } return rc; } int sqlite3Fts5UnpackTokenizeBlob( Fts5Config *pConfig, sqlite3_value *pVal, Fts5TokenizerInst **ppTok, char **pzText, int *pbDel ){ int rc = SQLITE_OK; if( IS_TOKENIZE_BLOB(pVal) ){ const char *zTok = 0; const char *zText = 0; int nText = 0; zTok = fts5UnpackTokenizeBlob(pVal, &zText, &nText); rc = sqlite3Fts5ConfigFindTokenizer(pConfig, zTok, ppTok); *pzText = sqlite3Fts5Mprintf(&rc, "%.*s", nText, zText); *pbDel = 1; }else{ *pzText = (char*)sqlite3_value_text(pVal); *pbDel = 0; *ppTok = pConfig->pTokList; } return rc; } /* ** Insert a new row into the FTS content table. */ int sqlite3Fts5StorageContentInsert( Fts5Storage *p, sqlite3_value **apVal, |
︙ | ︙ | |||
771 772 773 774 775 776 777 | rc = fts5StorageNewRowid(p, piRowid); } }else{ sqlite3_stmt *pInsert = 0; /* Statement to write %_content table */ int i; /* Counter variable */ rc = fts5StorageGetStmt(p, FTS5_STMT_INSERT_CONTENT, &pInsert, 0); for(i=1; rc==SQLITE_OK && i<=pConfig->nCol+1; i++){ | > > > > > > > | > | 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 | rc = fts5StorageNewRowid(p, piRowid); } }else{ sqlite3_stmt *pInsert = 0; /* Statement to write %_content table */ int i; /* Counter variable */ rc = fts5StorageGetStmt(p, FTS5_STMT_INSERT_CONTENT, &pInsert, 0); for(i=1; rc==SQLITE_OK && i<=pConfig->nCol+1; i++){ sqlite3_value *pVal = apVal[i]; if( IS_TOKENIZE_BLOB(pVal) ){ const char *zT = 0; int nT = 0; fts5UnpackTokenizeBlob(pVal, &zT, &nT); rc = sqlite3_bind_text(pInsert, i, zT, nT, SQLITE_STATIC); }else{ rc = sqlite3_bind_value(pInsert, i, apVal[i]); } } if( rc==SQLITE_OK ){ sqlite3_step(pInsert); rc = sqlite3_reset(pInsert); } *piRowid = sqlite3_last_insert_rowid(pConfig->db); } |
︙ | ︙ | |||
806 807 808 809 810 811 812 | if( rc==SQLITE_OK ){ rc = sqlite3Fts5IndexBeginWrite(p->pIndex, 0, iRowid); } for(ctx.iCol=0; rc==SQLITE_OK && ctx.iCol<pConfig->nCol; ctx.iCol++){ ctx.szCol = 0; if( pConfig->abUnindexed[ctx.iCol]==0 ){ | > > > > > > > > | | > | | < < | | 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 | if( rc==SQLITE_OK ){ rc = sqlite3Fts5IndexBeginWrite(p->pIndex, 0, iRowid); } for(ctx.iCol=0; rc==SQLITE_OK && ctx.iCol<pConfig->nCol; ctx.iCol++){ ctx.szCol = 0; if( pConfig->abUnindexed[ctx.iCol]==0 ){ sqlite3_value *pVal = apVal[ctx.iCol+2]; const char *zText = 0; const char *zTok = 0; int nText = 0; if( IS_TOKENIZE_BLOB(pVal) ){ zTok = fts5UnpackTokenizeBlob(pVal, &zText, &nText); }else{ zText = (const char*)sqlite3_value_text(apVal[ctx.iCol+2]); nText = sqlite3_value_bytes(apVal[ctx.iCol+2]); } rc = sqlite3Fts5SpecTokenize(pConfig, zTok, FTS5_TOKENIZE_DOCUMENT, zText, nText, (void*)&ctx, fts5StorageInsertCallback ); } sqlite3Fts5BufferAppendVarint(&rc, &buf, ctx.szCol); p->aTotalSize[ctx.iCol] += (i64)ctx.szCol; } p->nTotalRow++; |
︙ | ︙ |
Changes to ext/fts5/fts5_tcl.c.
︙ | ︙ | |||
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 | rc = sqlite3Fts5TestRegisterMatchinfo(db); if( rc!=SQLITE_OK ){ Tcl_SetResult(interp, (char*)sqlite3ErrName(rc), TCL_VOLATILE); return TCL_ERROR; } return TCL_OK; } static int SQLITE_TCLAPI f5tRegisterTok( void * clientData, Tcl_Interp *interp, int objc, Tcl_Obj *CONST objv[] ){ int rc; sqlite3 *db = 0; fts5_api *pApi = 0; if( objc!=2 ){ Tcl_WrongNumArgs(interp, 1, objv, "DB"); return TCL_ERROR; } if( f5tDbAndApi(interp, objv[1], &db, &pApi) ){ return TCL_ERROR; } rc = sqlite3Fts5TestRegisterTok(db, pApi); if( rc!=SQLITE_OK ){ Tcl_SetResult(interp, (char*)sqlite3ErrName(rc), TCL_VOLATILE); return TCL_ERROR; } return TCL_OK; } typedef struct OriginTextCtx OriginTextCtx; struct OriginTextCtx { sqlite3 *db; fts5_api *pApi; | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 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 | rc = sqlite3Fts5TestRegisterMatchinfo(db); if( rc!=SQLITE_OK ){ Tcl_SetResult(interp, (char*)sqlite3ErrName(rc), TCL_VOLATILE); return TCL_ERROR; } return TCL_OK; } static void f5tFree(void *p) { sqlite3_free(p); } static void f5tScalarFunc( sqlite3_context *ctx, int nArg, sqlite3_value **apArg ){ const char *zText = (const char*)sqlite3_value_text(apArg[0]); int nText = sqlite3_value_bytes(apArg[0]); const char *zTok = (const char*)sqlite3_value_text(apArg[1]); int nTok = sqlite3_value_bytes(apArg[1]); unsigned char *aBuf = 0; int nBuf = 0; assert( nArg==2 ); if( zTok==0 ) zTok = ""; nBuf = nTok + 1 + nText; aBuf = (unsigned char*)sqlite3_malloc(nBuf); if( aBuf==0 ){ sqlite3_result_error_nomem(ctx); }else{ memcpy(aBuf, zTok, nTok+1); memcpy(&aBuf[nTok+1], zText, nText); sqlite3_result_blob(ctx, aBuf, nBuf, f5tFree); sqlite3_result_subtype(ctx, SQLITE_FTS5_TOKENIZE_SUBTYPE); } } static int SQLITE_TCLAPI f5tRegisterTok( void * clientData, Tcl_Interp *interp, int objc, Tcl_Obj *CONST objv[] ){ int rc; sqlite3 *db = 0; fts5_api *pApi = 0; if( objc!=2 ){ Tcl_WrongNumArgs(interp, 1, objv, "DB"); return TCL_ERROR; } if( f5tDbAndApi(interp, objv[1], &db, &pApi) ){ return TCL_ERROR; } rc = sqlite3Fts5TestRegisterTok(db, pApi); if( rc==SQLITE_OK ){ rc = sqlite3_create_function(db, "fts5tokenize", 2, SQLITE_UTF8, 0, f5tScalarFunc, 0, 0 ); } if( rc!=SQLITE_OK ){ Tcl_SetResult(interp, (char*)sqlite3ErrName(rc), TCL_VOLATILE); return TCL_ERROR; } return TCL_OK; } typedef struct OriginTextCtx OriginTextCtx; struct OriginTextCtx { sqlite3 *db; fts5_api *pApi; |
︙ | ︙ |