SQLite

Check-in [adb780e0dc]
Login

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:Module spec parser enhancements for FTS1. Now able to cope with column names in the spec that are SQL keywords or have special characters, etc. Also added support for additional control lines. Column names can be followed by a type specifier (which is ignored.) (CVS 3410)
Downloads: Tarball | ZIP archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: adb780e0dc8bc7dcd1102efbfa4bc17eefdf968e
User & Date: drh 2006-09-13 15:20:13.000
Context
2006-09-13
16:02
Implementation of "column:" modifiers in FTS1 queries. (CVS 3411) (check-in: 820634f71e user: drh tags: trunk)
15:20
Module spec parser enhancements for FTS1. Now able to cope with column names in the spec that are SQL keywords or have special characters, etc. Also added support for additional control lines. Column names can be followed by a type specifier (which is ignored.) (CVS 3410) (check-in: adb780e0dc user: drh tags: trunk)
12:36
Fix the FTS1 test cases and add new tests. Comments added to the FTS1 code. (CVS 3409) (check-in: 528036c828 user: drh tags: trunk)
Changes
Side-by-Side Diff Ignore Whitespace Patch
Changes to ext/fts1/fts1.c.
891
892
893
894
895
896
897
898
899



900
901
902
903
904
905
906
891
892
893
894
895
896
897


898
899
900
901
902
903
904
905
906
907







-
-
+
+
+







  /* TERM_DELETE */ "delete from %_term where rowid = ?",
};

typedef struct fulltext_vtab {
  sqlite3_vtab base;
  sqlite3 *db;
  const char *zName;               /* virtual table name */
  int nColumns;                    /* number of columns in virtual table */
  const char *zColumnNames;        /* all column names, separated by commas */
  int nColumn;                     /* number of columns in virtual table */
  char **azColumn;                 /* column names.  malloced */
  char *zColumnList;               /* comma-separate list of column names */
  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];
} fulltext_vtab;
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
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







-
+

-
+













+
+
-
-
-
+
+
+







 */
static const char *contentInsertStatement(fulltext_vtab *v){
  StringBuffer sb;
  int i;

  initStringBuffer(&sb);
  append(&sb, "insert into %_content (rowid, ");
  append(&sb, v->zColumnNames);
  append(&sb, v->zColumnList);
  append(&sb, ") values (?");
  for(i=0; i<v->nColumns; ++i)
  for(i=0; i<v->nColumn; ++i)
    append(&sb, ", ?");
  append(&sb, ")");
  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;
    const char *zStmt = iStmt==CONTENT_INSERT_STMT ? contentInsertStatement(v) : 
                                                     fulltext_zStatement[iStmt];
    int rc = sql_prepare(v->db, v->zName, &v->pFulltextStatements[iStmt],
    zStmt = iStmt==CONTENT_INSERT_STMT ? contentInsertStatement(v) : 
                                         fulltext_zStatement[iStmt];
    rc = sql_prepare(v->db, v->zName, &v->pFulltextStatements[iStmt],
                         zStmt);
    if( iStmt==CONTENT_INSERT_STMT ) free((void *) zStmt);
    if( rc!=SQLITE_OK ) return rc;
  } else {
    int rc = sqlite3_reset(v->pFulltextStatements[iStmt]);
    if( rc!=SQLITE_OK ) return rc;
  }
1019
1020
1021
1022
1023
1024
1025
1026

1027
1028
1029
1030
1031
1032
1033
1022
1023
1024
1025
1026
1027
1028

1029
1030
1031
1032
1033
1034
1035
1036







-
+







  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->nColumns; ++i){
  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);
}

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
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







-
-
-
+
+
+










-
+








  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->nColumns * sizeof(const char *));
  for(i=0; i<v->nColumns; ++i){
    values[i] = string_dup(sqlite3_column_text(s, i));
  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->nColumns, values);
  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);
1159
1160
1161
1162
1163
1164
1165
1166

1167
1168
1169
1170
1171
1172
1173
1162
1163
1164
1165
1166
1167
1168

1169
1170
1171
1172
1173
1174
1175
1176







-
+







    ** 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, doclist.iType,
                sqlite3_column_blob(s, 0), sqlite3_column_bytes(s, 0));

    if( iColumn<v->nColumns ){   /* querying a single column */
    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);
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
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







-
-
+
+



+
+
+
+
+
+
+
+
-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+



+
+
+
+
+
+
+
+
-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+


-
-
+
+
-
-
+
+
+

-
-
-
-
-
+
+
+
+
+
+


+
+
+
-
-
-
+
+
+
+
+










-
-
+
+
+
+
+
+
+






+

-
-
-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
-
+
-
-
-
+
+
-
-
-
+
+
-
-
-
-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
-
-
-
-
-
-
+
+
+
+
+
+
-
-


+
+
-
-
+
+

-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
-
-
+
+
+
+
-
-
-
+
+
+
+
-

+
+
+
+
-
-
+
+
+
+
+
+
+
+
+
+
+
+

-
+


-
-
+
+
+
+
+
-
-
-
-
+
+
+
+
+
+
+
+
+
+
-
-
-
-
+
+
+
+
-
-
-
-
-
+
+
+
-
-
-
+

-
-
+
+
+


+
+
+
+
-
-
+
+
+
+
+
+

+





-
+



-
-
-
+
+
+
+
+
+

-
+
-
-
-
-
-
-
-
-
-
-
-
-
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
-





-
+

-
+







  }

  if( v->pTokenizer!=NULL ){
    v->pTokenizer->pModule->xDestroy(v->pTokenizer);
    v->pTokenizer = NULL;
  }
  
  free((void *) v->zColumnNames);
  free((void *) v->zName);
  free(v->azColumn);
  free(v->zColumnList);
  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 */
/* Return true if the string s begins with the string t, ignoring case. */
static int startsWithIgnoreCase(const char *s, const char *t){
  while( *t )
    if( tolower(*s++)!=tolower(*t++) ) return 0;

/*
** 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;
const char *kTokenize = "tokenize";

static int isTokenize(const char *arg){
  return startsWithIgnoreCase(arg, kTokenize);

/*
** 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);
}

static const char *tokenizerSpec(const char *arg){
  return arg + strlen(kTokenize);
/*
** An instance of this structure defines the "spec" of a the
}

** full text index.  This structure is populated by parseSpec
** and use by fulltextConnect and fulltextCreate.
*/
typedef struct TableSpec {
  const char *zName;
  int nColumns;
  const char * const *zColumnNames;
  char *zTokenizer;
  char *zTokenizerArg;
  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 *zColumnList;       /* Comma-separated list of names for %_content */
  char **azTokenizer;      /* Name of tokenizer and its arguments */
  char **azDelimiter;      /* Delimiters used for snippets */
} TableSpec;

/*
** Reclaim all of the memory used by a TableSpec
*/
void destroyTableSpec(TableSpec *p) {
  free(p->zTokenizer);
  free(p->zTokenizerArg);
void clearTableSpec(TableSpec *p) {
  free(p->azColumn);
  free(p->zColumnList);
  free(p->azTokenizer);
  free(p->azDelimiter);
}

/* Parse a CREATE VIRTUAL TABLE statement, which looks like this:
 *
 * CREATE VIRTUAL TABLE email
 *        USING fts1(subject, body, tokenize mytokenizer(myarg))
 *
 * We return parsed information in a TableSpec structure.
 * 
 */
int parseSpec(TableSpec *pSpec, int argc, const char * const *argv){
  int i;
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 */
  const char *zDelimiter = 0;    /* argv[] entry describing the delimiters */

  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.
  */
  pSpec->zName = argv[2];
  for (i=3; i<argc && !isTokenize(argv[i]); ++i)
    ;
  pSpec->nColumns = i-3;
  if( pSpec->nColumns<1) return SQLITE_ERROR;
  pSpec->zColumnNames = &argv[3];

  /* 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];
  pSpec->zTokenizer = pSpec->zTokenizerArg = NULL;
  if( i<argc ){  /* we have a tokenizer */
  for(i=0; i<argc; i++){
    const char *start, *end;
    assert( isTokenize(argv[i]) );
    start = tokenizerSpec(argv[i]);
    azArg[i] = z;
    strcpy(z, argv[i]);
    while( isspace(*start) ){
      ++start;
    }
    z += strlen(z)+1;
  }
    end = start;
    while( isalnum(*end) ){
      ++end;
    }
    pSpec->zTokenizer = string_dup_n(start, end-start);

    start = end;

  /* 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";
  zDelimiter = "delimiters('[',']','...')";
  n = 0;
  for(i=3, j=0; i<argc; ++i){
    if( startsWith(azArg[i],"tokenize") ){
      zTokenizer = azArg[i];
    while( isspace(*start) ){
      ++start;
    }
    if( *start=='(' ){  /* tokenizer has an argument */
      ++start;
      end = strchr(start, ')');
    }else if( startsWith(azArg[i],"delimiters") ){
      zDelimiter = azArg[i];
    }else{
      z = azArg[pSpec->nColumn] = firstToken(azArg[i], &zDummy);
      pSpec->nColumn++;
      n += strlen(z) + 6;
      if( !end ) return SQLITE_ERROR;
      pSpec->zTokenizerArg = string_dup_n(start, end-start);
    }
  }
  if( pSpec->nColumn==0 ){
    azArg[0] = "content";
  return SQLITE_OK;
}
    pSpec->nColumn = 1;
  }

/* Concatenate an array of strings into a single string, separating with commas.
 * The caller must free the returned string. */
static char *commaConcatenate(int nColumns, const char * const *zColumns){
  /*
  ** Construct the comma-separated list of column names.
  **
  ** Each 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.
  */
  z = pSpec->zColumnList = malloc( n );
  StringBuffer buf;
  int i;

  if( z==0 ){
    clearTableSpec(pSpec);
    return SQLITE_NOMEM;
  }
  initStringBuffer(&buf);
  for(i=0; i<nColumns ; ++i){
    if( i>0 ){
  for(i=0; i<pSpec->nColumn; i++){
    sqlite3_snprintf(n, z, "c%d%s", i, azArg[i]);
    for(j=0; z[j]; j++){
      if( !isalnum(z[j]) ) z[j] = '_';
      append(&buf, ", ");
    }
    z[j] = ',';
    z += j+1;
  }
  z[-1] = 0;
    append(&buf, zColumns[i]);
  }

  /*
  ** Parse the tokenizer specification string.
  */
  pSpec->azTokenizer = tokenizeString(zTokenizer, &n);
  tokenListToIdList(pSpec->azTokenizer);

  /*
  ** Parse the delimiter specification string.
  */
  pSpec->azDelimiter = tokenizeString(zDelimiter, &n);
  tokenListToIdList(pSpec->azDelimiter);

  return buf.s;
  return SQLITE_OK;
}

static char *fulltextSchema(char *name, int nColumns,
                            const char * const *zColumns, int magic){
/*
** Generate a CREATE TABLE statement that describes the schema of
** the virtual table.  Return a pointer to this schema.  
**
** If the addAllColumn parameter is true, then add a column named
  StringBuffer buf;
  int i;

  initStringBuffer(&buf);
** "_all" to the end of the schema.
**
** 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 */
){
  int i;
  append(&buf, "create table ");
  append(&buf, name);
  append(&buf, "(");
  for(i=0; i<nColumns; ++i){
  char *zSchema, *zNext;
  const char *zSep = "(";
  zSchema = sqlite3_mprintf("CREATE TABLE x");
  for(i=0; i<nColumn; i++){
    if( i>0 ){
      append(&buf, ", ");
    }
    append(&buf, zColumns[i]);
    append(&buf, " text");
    zNext = sqlite3_mprintf("%s%s%Q", zSchema, zSep, azColumn[i]);
    sqlite3_free(zSchema);
    zSchema = zNext;
  }
  if( magic ){
    append(&buf, ", _all text");
    zSep = ",";
  }
  append(&buf, ")");
  return buf.s;
  zNext = sqlite3_mprintf("%s,_all)", zSchema);
  sqlite3_free(zSchema);
  return zNext;
}

/*
** Build a new sqlite3_vtab structure that will describe the
** fulltext index defined by spec.
*/
static int connect(sqlite3 *db, TableSpec *spec,
                   sqlite3_vtab **ppVTab, char **pzErr){
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_ERROR;
  if( v==0 ) return SQLITE_NOMEM;
  memset(v, 0, sizeof(*v));
  /* sqlite will initialize v->base */
  v->db = db;
  v->zName = string_dup(spec->zName);
  v->nColumns = spec->nColumns;
  v->zColumnNames = commaConcatenate(spec->nColumns, spec->zColumnNames);
  v->zName = spec->zName;   /* Freed when azColumn is freed */
  v->nColumn = spec->nColumn;
  v->zColumnList = spec->zColumnList;
  spec->zColumnList = 0;
  v->azColumn = spec->azColumn;
  spec->azColumn = 0;

  if( spec->zTokenizer == NULL ){
  if( spec->azTokenizer==0 ){
    sqlite3Fts1SimpleTokenizerModule(&m);
  } else {
    /* TODO(shess) For now, add new tokenizers as else if clauses. */
    if( !strcmp(spec->zTokenizer, "simple") ){
      sqlite3Fts1SimpleTokenizerModule(&m);
    } else {
      *pzErr = sqlite3_mprintf("unknown tokenizer: %s", spec->zTokenizer);
      rc = SQLITE_ERROR;
      goto err;
    }
  }

    return SQLITE_NOMEM;
  }
  /* TODO(shess) For now, add new tokenizers as else if clauses. */
  if( spec->azTokenizer[0]==0 || !strcmp(spec->azTokenizer[0], "simple") ){
    sqlite3Fts1SimpleTokenizerModule(&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);
  }
  /* TODO: Support multiple arguments to tokenizers. */
  rc = m->xCreate(1, &spec->zTokenizerArg, &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("x", spec->nColumns, spec->zColumnNames, 1);
  schema = fulltextSchema(v->nColumn, (const char*const*)v->azColumn);
  rc = sqlite3_declare_vtab(db, schema);
  free(schema);
  sqlite3_free(schema);
  if( rc!=SQLITE_OK ) goto err;

  memset(v->pFulltextStatements, 0, sizeof(v->pFulltextStatements));

  *ppVTab = &v->base;
  TRACE(("FTS1 Connect %p\n", v));

1445
1446
1447
1448
1449
1450
1451
1452

1453
1454
1455
1456


1457
1458
1459
1460
1461
1462
1463
1761
1762
1763
1764
1765
1766
1767

1768
1769
1770


1771
1772
1773
1774
1775
1776
1777
1778
1779







-
+


-
-
+
+







  sqlite3 *db,
  void *pAux,
  int argc, const char *const*argv,
  sqlite3_vtab **ppVTab,
  char **pzErr
){
  TableSpec spec;
  int rc = parseSpec(&spec, argc, argv);
  int rc = parseSpec(&spec, argc, argv, pzErr);
  if( rc!=SQLITE_OK ) return rc;

  rc = connect(db, &spec, ppVTab, pzErr);
  destroyTableSpec(&spec);
  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
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
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







-
+


-
+

-
+







-
+


-
+







                          int argc, const char * const *argv,
                          sqlite3_vtab **ppVTab, char **pzErr){
  int rc;
  TableSpec spec;
  char *schema;
  TRACE(("FTS1 Create\n"));

  rc = parseSpec(&spec, argc, argv);
  rc = parseSpec(&spec, argc, argv, pzErr);
  if( rc!=SQLITE_OK ) return rc;

  schema = fulltextSchema("%_content", spec.nColumns, spec.zColumnNames, 0);
  schema = sqlite3_mprintf("CREATE TABLE %%_content(%s)", spec.zColumnList);
  rc = sql_exec(db, spec.zName, schema);
  free(schema);
  sqlite3_free(schema);
  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 = connect(db, &spec, ppVTab, pzErr);
  rc = constructVtab(db, &spec, ppVTab, pzErr);

out:
  destroyTableSpec(&spec);
  clearTableSpec(&spec);
  return rc;
}

/* Decide how to handle an SQL query. */
static int fulltextBestIndex(sqlite3_vtab *pVTab, sqlite3_index_info *pInfo){
  int i;

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
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







-
+



-
-
+
-
-
-
+
-
-
-
+
+
















-
+











-














-
+


-
+







  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;
  StringBuffer sb;
  char *zSql;

  TRACE(("FTS1 Filter %p\n",pCursor));

  initStringBuffer(&sb);
  append(&sb, "select rowid, ");
  zSql = sqlite3_mprintf("select rowid, * from %%_content %s",
  append(&sb, v->zColumnNames);
  append(&sb, " from %_content");
  if( idxNum != QUERY_GENERIC) {
                          idxNum==QUERY_GENERIC ? "" : "where rowid=?");
    append(&sb, " where rowid = ?"); 
  }
  rc = sql_prepare(v->db, v->zName, &c->pStmt, sb.s);
  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->nColumns);
      assert( idxNum<=QUERY_FULLTEXT+v->nColumn);
      assert( argc==1 );
      rc = fulltextQuery(v, idxNum-QUERY_FULLTEXT, zQuery, -1, &pResult);
      if( rc!=SQLITE_OK ) goto out;
      readerInit(&c->result, pResult);
      break;
    }
  }

  rc = fulltextNext(pCursor);

out:
  free(sb.s);
  return rc;
}

static int fulltextEof(sqlite3_vtab_cursor *pCursor){
  fulltext_cursor *c = (fulltext_cursor *) pCursor;
  return c->eof;
}

static int fulltextColumn(sqlite3_vtab_cursor *pCursor,
                          sqlite3_context *pContext, int idxCol){
  fulltext_cursor *c = (fulltext_cursor *) pCursor;
  fulltext_vtab *v = cursor_vtab(c);
  const char *s;

  if( idxCol==v->nColumns ){  /* a request for _all */
  if( idxCol==v->nColumn ){  /* a request for _all */
    sqlite3_result_null(pContext);
  } else {
    assert( idxCol<v->nColumns );
    assert( idxCol<v->nColumn );
    s = (const char *) sqlite3_column_text(c->pStmt, idxCol+1);
    sqlite3_result_text(pContext, s, -1, SQLITE_TRANSIENT);
  }

  return SQLITE_OK;
}

2119
2120
2121
2122
2123
2124
2125
2126
2127



2128
2129
2130
2131
2132
2133
2134
2430
2431
2432
2433
2434
2435
2436


2437
2438
2439
2440
2441
2442
2443
2444
2445
2446







-
-
+
+
+







  int rc;

  rc = content_insert(v, pRequestRowid, pValues);
  if( rc!=SQLITE_OK ) return rc;
  *piRowid = sqlite3_last_insert_rowid(v->db);

  fts1HashInit(&terms, FTS1_HASH_STRING, 1);
  for(i = 0; i < v->nColumns ; ++i){
    rc = buildTerms(v, &terms, i, sqlite3_value_text(pValues[i]), -1, *piRowid);
  for(i = 0; i < v->nColumn ; ++i){
    rc = buildTerms(v, &terms, i, (char*)sqlite3_value_text(pValues[i]), -1,
                    *piRowid);
    if( rc!=SQLITE_OK ) goto out;
  }

  for(e=fts1HashFirst(&terms); e; e=fts1HashNext(e)){
    DocList *p = fts1HashData(e);
    rc = index_insert_term(v, fts1HashKey(e), fts1HashKeysize(e), p);
    if( rc!=SQLITE_OK ) break;
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
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







-
+

















-
+







  fts1HashElem *e;
  DocList doclist;

  int rc = content_select(v, iRow, &pValues);
  if( rc!=SQLITE_OK ) return rc;

  fts1HashInit(&terms, FTS1_HASH_STRING, 1);
  for(i = 0 ; i < v->nColumns; ++i) {
  for(i = 0 ; i < v->nColumn; ++i) {
    rc = buildTerms(v, &terms, i, pValues[i], -1, iRow);
    if( rc!=SQLITE_OK ) goto out;
  }

  /* Delete by inserting a doclist with no positions.  This will
  ** overwrite existing data as it is merged forward by
  ** index_insert_term().
  */
  docListInit(&doclist, DL_POSITIONS_OFFSETS, 0, 0);
  docListAddDocid(&doclist, iRow);

  for(e=fts1HashFirst(&terms); e; e=fts1HashNext(e)){
    rc = index_insert_term(v, fts1HashKey(e), fts1HashKeysize(e), &doclist);
    if( rc!=SQLITE_OK ) break;
  }

out:
  freeStringArray(v->nColumns, pValues);
  freeStringArray(v->nColumn, pValues);
  for(e=fts1HashFirst(&terms); e; e=fts1HashNext(e)){
    DocList *p = fts1HashData(e);
    docListDelete(p);
  }
  fts1HashClear(&terms);
  docListDestroy(&doclist);

2195
2196
2197
2198
2199
2200
2201
2202
2203
2204



2205
2206
2207
2208
2209
2210
2211
2507
2508
2509
2510
2511
2512
2513



2514
2515
2516
2517
2518
2519
2520
2521
2522
2523







-
-
-
+
+
+







  }

  if( sqlite3_value_type(ppArg[0]) != SQLITE_NULL ){
    return SQLITE_ERROR;   /* an update; not yet supported */
  }

  /* ppArg[1] = rowid
   * ppArg[2..2+v->nColumns-1] = values
   * ppArg[2+v->nColumns] = value for _all (we ignore this) */
  assert( nArg==2+v->nColumns+1);    
   * ppArg[2..2+v->nColumn-1] = values
   * ppArg[2+v->nColumn] = value for _all (we ignore this) */
  assert( nArg==2+v->nColumn+1);    

  return index_insert(v, ppArg[1], &ppArg[2], pRowid);
}

static const sqlite3_module fulltextModule = {
  0,
  fulltextCreate,
Changes to test/fts1b.test.
1
2
3
4
5
6
7
8
9
10
11
12
13
14

15
16
17
18
19
20
21
1
2
3
4
5
6
7
8
9
10
11
12
13

14
15
16
17
18
19
20
21













-
+







# 2006 September 13
#
# 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 file implements regression tests for SQLite library.  The
# focus of this script is testing the FTS1 module.
#
# $Id: fts1b.test,v 1.1 2006/09/13 12:36:09 drh Exp $
# $Id: fts1b.test,v 1.2 2006/09/13 15:20:13 drh Exp $
#

set testdir [file dirname $argv0]
source $testdir/tester.tcl

# If SQLITE_ENABLE_FTS1 is defined, omit this file.
ifcapable !fts1 {
77
78
79
80
81
82
83








84
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92







+
+
+
+
+
+
+
+

do_test fts1b-1.6 {
  execsql {SELECT english, spanish, german FROM t1 WHERE rowid=1}
} {one un eine}
do_test fts1b-1.7 {
  execsql {SELECT rowid FROM t1 WHERE _all MATCH '"one un"'}
} {}

do_test fts1b-2.1 {
  execsql {
    CREATE VIRTUAL TABLE t2 USING fts1(from,to);
    INSERT INTO t2([from],[to]) VALUES ('one two three', 'four five six');
    SELECT [from], [to] FROM t2
  }
} {{one two three} {four five six}}

finish_test