/ Check-in [ecf2dec6]
Login

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

Overview
Comment:Changes fts2 to use only sqlite3_malloc() and not system malloc. Backports (4554) and (4555) from fts3. (CVS 5454)
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1:ecf2dec66cb979cb7d8db3b7ce5c64cab57fe2bb
User & Date: shess 2008-07-22 22:57:54
Context
2008-07-22
23:08
Change prefix search from O(N*M) to O(NlogM). Backports (4599) from fts3. (CVS 5455) check-in: 3f614453 user: shess tags: trunk
22:57
Changes fts2 to use only sqlite3_malloc() and not system malloc. Backports (4554) and (4555) from fts3. (CVS 5454) check-in: ecf2dec6 user: shess tags: trunk
22:20
fts2.c buildTerms() passes -1 for nInput. Backports (4511) from fts3. (CVS 5453) check-in: d562515e user: shess tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to ext/fts2/fts2.c.

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
....
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
....
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
....
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
....
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
....
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
....
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
....
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
....
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
....
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
....
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
....
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
....
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
....
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
....
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
....
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
....
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
....
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
....
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
....
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965


3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
....
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
....
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
....
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
....
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
  int nData;            /* End of data loaded into pData. */
} DataBuffer;

static void dataBufferInit(DataBuffer *pBuffer, int nCapacity){
  assert( nCapacity>=0 );
  pBuffer->nData = 0;
  pBuffer->nCapacity = nCapacity;
  pBuffer->pData = nCapacity==0 ? NULL : malloc(nCapacity);
}
static void dataBufferReset(DataBuffer *pBuffer){
  pBuffer->nData = 0;
}
static void dataBufferDestroy(DataBuffer *pBuffer){
  if( pBuffer->pData!=NULL ) free(pBuffer->pData);
  SCRAMBLE(pBuffer);
}
static void dataBufferExpand(DataBuffer *pBuffer, int nAddCapacity){
  assert( nAddCapacity>0 );
  /* TODO(shess) Consider expanding more aggressively.  Note that the
  ** underlying malloc implementation may take care of such things for
  ** us already.
  */
  if( pBuffer->nData+nAddCapacity>pBuffer->nCapacity ){
    pBuffer->nCapacity = pBuffer->nData+nAddCapacity;
    pBuffer->pData = realloc(pBuffer->pData, pBuffer->nCapacity);
  }
}
static void dataBufferAppend(DataBuffer *pBuffer,
                             const char *pSource, int nSource){
  assert( nSource>0 && pSource!=NULL );
  dataBufferExpand(pBuffer, nSource);
  memcpy(pBuffer->pData+pBuffer->nData, pSource, nSource);
................................................................................
}
static void dlcAddPos(DLCollector *pCollector, int iColumn, int iPos,
                      int iStartOffset, int iEndOffset){
  plwAdd(&pCollector->plw, iColumn, iPos, iStartOffset, iEndOffset);
}

static DLCollector *dlcNew(sqlite_int64 iDocid, DocListType iType){
  DLCollector *pCollector = malloc(sizeof(DLCollector));
  dataBufferInit(&pCollector->b, 0);
  dlwInit(&pCollector->dlw, iType, &pCollector->b);
  plwInit(&pCollector->plw, &pCollector->dlw, iDocid);
  return pCollector;
}
static void dlcDelete(DLCollector *pCollector){
  plwDestroy(&pCollector->plw);
  dlwDestroy(&pCollector->dlw);
  dataBufferDestroy(&pCollector->b);
  SCRAMBLE(pCollector);
  free(pCollector);
}


/* Copy the doclist data of iType in pData/nData into *out, trimming
** unnecessary data as we go.  Only columns matching iColumn are
** copied, all columns copied if iColumn is -1.  Elements with no
** matching columns are dropped.  The output is an iOutType doclist.
................................................................................

  dlrDestroy(&left);
  dlrDestroy(&right);
  dlwDestroy(&writer);
}

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 is not part of the standard C library and
................................................................................

  /* first compute length needed */
  for(p = zFormat ; *p ; ++p){
    len += (*p=='%' ? nFullTableName : 1);
  }
  len += 1;  /* for null terminator */

  r = result = malloc(len);
  for(p = zFormat; *p; ++p){
    if( *p=='%' ){
      memcpy(r, zDb, nDb);
      r += nDb;
      *r++ = '.';
      memcpy(r, zName, nName);
      r += nName;
................................................................................

static int sql_exec(sqlite3 *db, const char *zDb, const char *zName,
                    const char *zFormat){
  char *zCommand = string_format(zFormat, zDb, 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 *zDb, const char *zName,
                       sqlite3_stmt **ppStmt, const char *zFormat){
  char *zCommand = string_format(zFormat, zDb, zName);
  int rc;
  TRACE(("FTS2 prepare: %s\n", zCommand));
  rc = sqlite3_prepare_v2(db, zCommand, -1, ppStmt, NULL);
  free(zCommand);
  return rc;
}

/* end utility functions */

/* Forward reference */
typedef struct fulltext_vtab fulltext_vtab;
................................................................................
      case CONTENT_UPDATE_STMT:
        zStmt = contentUpdateStatement(v); break;
      default:
        zStmt = fulltext_zStatement[iStmt];
    }
    rc = sql_prepare(v->db, v->zDb, 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 sql_single_step(s);
}

static void freeStringArray(int nString, const char **pString){
  int i;

  for (i=0 ; i < nString ; ++i) {
    if( pString[i]!=NULL ) 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.
 * null fields will be NULL in the returned array.
 *
 * TODO: Perhaps we should return pointer/length strings here for consistency
................................................................................

  rc = sqlite3_bind_int64(s, 1, iRow);
  if( rc!=SQLITE_OK ) return rc;

  rc = sqlite3_step(s);
  if( rc!=SQLITE_ROW ) return rc;

  values = (const char **) malloc(v->nColumn * sizeof(const char *));
  for(i=0; i<v->nColumn; ++i){
    if( sqlite3_column_type(s, i)==SQLITE_NULL ){
      values[i] = NULL;
    }else{
      values[i] = string_dup((char*)sqlite3_column_text(s, i));
    }
  }
................................................................................
  if( v->pTokenizer!=NULL ){
    v->pTokenizer->pModule->xDestroy(v->pTokenizer);
    v->pTokenizer = NULL;
  }

  clearPendingTerms(v);

  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 */
................................................................................
** 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);
................................................................................
      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
................................................................................
  char **azTokenizer;      /* Name of tokenizer and its arguments */
} TableSpec;

/*
** Reclaim all of the memory used by a TableSpec
*/
static 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))
 *
................................................................................
  ** The argv[][] array is read-only and transient.  We can write to the
  ** copy in order to modify things and the copy is persistent.
  */
  CLEAR(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]);
................................................................................
  ** 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]);
................................................................................
  fulltext_vtab *v = 0;
  const sqlite3_tokenizer_module *m = NULL;
  char *schema;

  char const *zTok;         /* Name of tokenizer to use for this fts table */
  int nTok;                 /* Length of zTok, including nul terminator */

  v = (fulltext_vtab *) malloc(sizeof(fulltext_vtab));
  if( v==0 ) return SQLITE_NOMEM;
  CLEAR(v);
  /* sqlite will initialize v->base */
  v->db = db;
  v->zDb = spec->zDb;       /* Freed when azColumn is freed */
  v->zName = spec->zName;   /* Freed when azColumn is freed */
  v->nColumn = spec->nColumn;
................................................................................
  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);
  CLEAR(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);
  CLEAR(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++;
................................................................................
  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;
................................................................................
  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.nData!=0 ) dlrDestroy(&c->reader);
  dataBufferDestroy(&c->result);
  free(c);
  return SQLITE_OK;
}

static int fulltextNext(sqlite3_vtab_cursor *pCursor){
  fulltext_cursor *c = (fulltext_cursor *) pCursor;
  int rc;

................................................................................
}

/* 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];
  CLEAR(t);
  t->pTerm = malloc(nTerm+1);
  memcpy(t->pTerm, pTerm, nTerm);
  t->pTerm[nTerm] = 0;
  t->nTerm = nTerm;
  t->isOr = q->nextIsOr;
  t->isPrefix = 0;
  q->nextIsOr = 0;
  t->iColumn = q->nextColumn;
................................................................................
  DataBuffer term;           /* Leftmost term in block's subtree. */
  DataBuffer data;           /* Accumulated data for the block. */
  struct InteriorBlock *next;
} InteriorBlock;

static InteriorBlock *interiorBlockNew(int iHeight, sqlite_int64 iChildBlock,
                                       const char *pTerm, int nTerm){
  InteriorBlock *block = calloc(1, sizeof(InteriorBlock));
  char c[VARINT_MAX+VARINT_MAX];
  int n;



  dataBufferInit(&block->term, 0);
  dataBufferReplace(&block->term, pTerm, nTerm);

  n = putVarint(c, iHeight);
  n += putVarint(c+n, iChildBlock);
  dataBufferInit(&block->data, INTERIOR_MAX);
  dataBufferReplace(&block->data, c, n);

  return block;
}

#ifndef NDEBUG
/* Verify that the data is readable as an interior node. */
static void interiorBlockValidate(InteriorBlock *pBlock){
  const char *pData = pBlock->data.pData;
................................................................................
  InteriorBlock *block = pWriter->first;

  while( block!=NULL ){
    InteriorBlock *b = block;
    block = block->next;
    dataBufferDestroy(&b->term);
    dataBufferDestroy(&b->data);
    free(b);
  }
  if( pWriter->parentWriter!=NULL ){
    interiorWriterDestroy(pWriter->parentWriter);
    free(pWriter->parentWriter);
  }
  dataBufferDestroy(&pWriter->term);
  SCRAMBLE(pWriter);
  return SQLITE_OK;
}

/* If pWriter can fit entirely in ROOT_MAX, return it as the root info
................................................................................
  ** interior node.
  */
  ASSERT_VALID_INTERIOR_BLOCK(block);
  rc = block_insert(v, block->data.pData, block->data.nData, &iBlockid);
  if( rc!=SQLITE_OK ) return rc;
  *piEndBlockid = iBlockid;

  pWriter->parentWriter = malloc(sizeof(*pWriter->parentWriter));
  interiorWriterInit(pWriter->iHeight+1,
                     block->term.pData, block->term.nData,
                     iBlockid, pWriter->parentWriter);

  /* Flush additional blocks and append to the higher interior
  ** node.
  */
................................................................................
  DataBuffer dl;

  /* Determine the next index at level 0, merging as necessary. */
  rc = segdirNextIndex(v, 0, &idx);
  if( rc!=SQLITE_OK ) return rc;

  n = fts2HashCount(pTerms);
  pData = malloc(n*sizeof(TermData));

  for(i = 0, e = fts2HashFirst(pTerms); e; i++, e = fts2HashNext(e)){
    assert( i<n );
    pData[i].pTerm = fts2HashKey(e);
    pData[i].nTerm = fts2HashKeysize(e);
    pData[i].pCollector = fts2HashData(e);
  }
................................................................................
                        pData[i].pTerm, pData[i].nTerm, dl.pData, dl.nData);
    if( rc!=SQLITE_OK ) goto err;
  }
  rc = leafWriterFinalize(v, &writer);

 err:
  dataBufferDestroy(&dl);
  free(pData);
  leafWriterDestroy(&writer);
  return rc;
}

/* If pendingTerms has data, free it. */
static int clearPendingTerms(fulltext_vtab *v){
  if( v->nPendingData>=0 ){







|





|










|







 







|










|







 







|







 







|







 







|









|







 







|







 







|

|







 







|







 







|



|
|







 







|







 







|










|







 







|
|
|







 







|







 







|







 







|







 







|
>
>
|
|
|
<
|
>
>
>








|

|







|
|
|







 







|







 







|







 







|







 







|






|







 







|



>
>
|
|

|
|
|
|
|







 







|



|







 







|







 







|







 







|







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
....
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
....
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
....
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
....
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
....
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
....
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
....
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
....
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
....
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
....
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
....
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
....
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
....
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
....
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
....
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
....
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
....
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
....
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
....
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
....
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
....
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
....
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
....
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
  int nData;            /* End of data loaded into pData. */
} DataBuffer;

static void dataBufferInit(DataBuffer *pBuffer, int nCapacity){
  assert( nCapacity>=0 );
  pBuffer->nData = 0;
  pBuffer->nCapacity = nCapacity;
  pBuffer->pData = nCapacity==0 ? NULL : sqlite3_malloc(nCapacity);
}
static void dataBufferReset(DataBuffer *pBuffer){
  pBuffer->nData = 0;
}
static void dataBufferDestroy(DataBuffer *pBuffer){
  if( pBuffer->pData!=NULL ) sqlite3_free(pBuffer->pData);
  SCRAMBLE(pBuffer);
}
static void dataBufferExpand(DataBuffer *pBuffer, int nAddCapacity){
  assert( nAddCapacity>0 );
  /* TODO(shess) Consider expanding more aggressively.  Note that the
  ** underlying malloc implementation may take care of such things for
  ** us already.
  */
  if( pBuffer->nData+nAddCapacity>pBuffer->nCapacity ){
    pBuffer->nCapacity = pBuffer->nData+nAddCapacity;
    pBuffer->pData = sqlite3_realloc(pBuffer->pData, pBuffer->nCapacity);
  }
}
static void dataBufferAppend(DataBuffer *pBuffer,
                             const char *pSource, int nSource){
  assert( nSource>0 && pSource!=NULL );
  dataBufferExpand(pBuffer, nSource);
  memcpy(pBuffer->pData+pBuffer->nData, pSource, nSource);
................................................................................
}
static void dlcAddPos(DLCollector *pCollector, int iColumn, int iPos,
                      int iStartOffset, int iEndOffset){
  plwAdd(&pCollector->plw, iColumn, iPos, iStartOffset, iEndOffset);
}

static DLCollector *dlcNew(sqlite_int64 iDocid, DocListType iType){
  DLCollector *pCollector = sqlite3_malloc(sizeof(DLCollector));
  dataBufferInit(&pCollector->b, 0);
  dlwInit(&pCollector->dlw, iType, &pCollector->b);
  plwInit(&pCollector->plw, &pCollector->dlw, iDocid);
  return pCollector;
}
static void dlcDelete(DLCollector *pCollector){
  plwDestroy(&pCollector->plw);
  dlwDestroy(&pCollector->dlw);
  dataBufferDestroy(&pCollector->b);
  SCRAMBLE(pCollector);
  sqlite3_free(pCollector);
}


/* Copy the doclist data of iType in pData/nData into *out, trimming
** unnecessary data as we go.  Only columns matching iColumn are
** copied, all columns copied if iColumn is -1.  Elements with no
** matching columns are dropped.  The output is an iOutType doclist.
................................................................................

  dlrDestroy(&left);
  dlrDestroy(&right);
  dlwDestroy(&writer);
}

static char *string_dup_n(const char *s, int n){
  char *str = sqlite3_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 is not part of the standard C library and
................................................................................

  /* first compute length needed */
  for(p = zFormat ; *p ; ++p){
    len += (*p=='%' ? nFullTableName : 1);
  }
  len += 1;  /* for null terminator */

  r = result = sqlite3_malloc(len);
  for(p = zFormat; *p; ++p){
    if( *p=='%' ){
      memcpy(r, zDb, nDb);
      r += nDb;
      *r++ = '.';
      memcpy(r, zName, nName);
      r += nName;
................................................................................

static int sql_exec(sqlite3 *db, const char *zDb, const char *zName,
                    const char *zFormat){
  char *zCommand = string_format(zFormat, zDb, zName);
  int rc;
  TRACE(("FTS2 sql: %s\n", zCommand));
  rc = sqlite3_exec(db, zCommand, NULL, 0, NULL);
  sqlite3_free(zCommand);
  return rc;
}

static int sql_prepare(sqlite3 *db, const char *zDb, const char *zName,
                       sqlite3_stmt **ppStmt, const char *zFormat){
  char *zCommand = string_format(zFormat, zDb, zName);
  int rc;
  TRACE(("FTS2 prepare: %s\n", zCommand));
  rc = sqlite3_prepare_v2(db, zCommand, -1, ppStmt, NULL);
  sqlite3_free(zCommand);
  return rc;
}

/* end utility functions */

/* Forward reference */
typedef struct fulltext_vtab fulltext_vtab;
................................................................................
      case CONTENT_UPDATE_STMT:
        zStmt = contentUpdateStatement(v); break;
      default:
        zStmt = fulltext_zStatement[iStmt];
    }
    rc = sql_prepare(v->db, v->zDb, v->zName, &v->pFulltextStatements[iStmt],
                         zStmt);
    if( zStmt != fulltext_zStatement[iStmt]) sqlite3_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 sql_single_step(s);
}

static void freeStringArray(int nString, const char **pString){
  int i;

  for (i=0 ; i < nString ; ++i) {
    if( pString[i]!=NULL ) sqlite3_free((void *) pString[i]);
  }
  sqlite3_free((void *) pString);
}

/* select * from %_content where rowid = [iRow]
 * The caller must delete the returned array and all strings in it.
 * null fields will be NULL in the returned array.
 *
 * TODO: Perhaps we should return pointer/length strings here for consistency
................................................................................

  rc = sqlite3_bind_int64(s, 1, iRow);
  if( rc!=SQLITE_OK ) return rc;

  rc = sqlite3_step(s);
  if( rc!=SQLITE_ROW ) return rc;

  values = (const char **) sqlite3_malloc(v->nColumn * sizeof(const char *));
  for(i=0; i<v->nColumn; ++i){
    if( sqlite3_column_type(s, i)==SQLITE_NULL ){
      values[i] = NULL;
    }else{
      values[i] = string_dup((char*)sqlite3_column_text(s, i));
    }
  }
................................................................................
  if( v->pTokenizer!=NULL ){
    v->pTokenizer->pModule->xDestroy(v->pTokenizer);
    v->pTokenizer = NULL;
  }

  clearPendingTerms(v);

  sqlite3_free(v->azColumn);
  for(i = 0; i < v->nColumn; ++i) {
    sqlite3_free(v->azContentColumn[i]);
  }
  sqlite3_free(v->azContentColumn);
  sqlite3_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 */
................................................................................
** 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 = sqlite3_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);
................................................................................
      aToken[nToken].z = z;
      aToken[nToken].n = n;
      nToken++;
      totalSize += n+1;
    }
    z += n;
  }
  azToken = (char**)sqlite3_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;
  sqlite3_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
................................................................................
  char **azTokenizer;      /* Name of tokenizer and its arguments */
} TableSpec;

/*
** Reclaim all of the memory used by a TableSpec
*/
static void clearTableSpec(TableSpec *p) {
  sqlite3_free(p->azColumn);
  sqlite3_free(p->azContentColumn);
  sqlite3_free(p->azTokenizer);
}

/* Parse a CREATE VIRTUAL TABLE statement, which looks like this:
 *
 * CREATE VIRTUAL TABLE email
 *        USING fts2(subject, body, tokenize mytokenizer(myarg))
 *
................................................................................
  ** The argv[][] array is read-only and transient.  We can write to the
  ** copy in order to modify things and the copy is persistent.
  */
  CLEAR(pSpec);
  for(i=n=0; i<argc; i++){
    n += strlen(argv[i]) + 1;
  }
  azArg = sqlite3_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]);
................................................................................
  ** 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 = sqlite3_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]);
................................................................................
  fulltext_vtab *v = 0;
  const sqlite3_tokenizer_module *m = NULL;
  char *schema;

  char const *zTok;         /* Name of tokenizer to use for this fts table */
  int nTok;                 /* Length of zTok, including nul terminator */

  v = (fulltext_vtab *) sqlite3_malloc(sizeof(fulltext_vtab));
  if( v==0 ) return SQLITE_NOMEM;
  CLEAR(v);
  /* sqlite will initialize v->base */
  v->db = db;
  v->zDb = spec->zDb;       /* Freed when azColumn is freed */
  v->zName = spec->zName;   /* Freed when azColumn is freed */
  v->nColumn = spec->nColumn;
................................................................................
  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 *) sqlite3_malloc(sizeof(fulltext_cursor));
  if( c ){
    memset(c, 0, sizeof(fulltext_cursor));
    /* sqlite will initialize c->base */
    *ppCursor = &c->base;
    TRACE(("FTS2 Open %p: %p\n", pVTab, c));

    return SQLITE_OK;
  }else{
    return SQLITE_NOMEM;
  }
}


/* Free all of the dynamically allocated memory held by *q
*/
static void queryClear(Query *q){
  int i;
  for(i = 0; i < q->nTerms; ++i){
    sqlite3_free(q->pTerms[i].pTerm);
  }
  sqlite3_free(q->pTerms);
  CLEAR(q);
}

/* Free all of the dynamically allocated memory held by the
** Snippet
*/
static void snippetClear(Snippet *p){
  sqlite3_free(p->aMatch);
  sqlite3_free(p->zOffset);
  sqlite3_free(p->zSnippet);
  CLEAR(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 = sqlite3_realloc(p->aMatch, p->nAlloc*sizeof(p->aMatch[0]) );
    if( p->aMatch==0 ){
      p->nMatch = 0;
      p->nAlloc = 0;
      return;
    }
  }
  i = p->nMatch++;
................................................................................
  int nDoc;
  const char *zDoc;
  int iStart, iEnd;
  int tailEllipsis = 0;
  int iMatch;
  

  sqlite3_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;
................................................................................
  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.nData!=0 ) dlrDestroy(&c->reader);
  dataBufferDestroy(&c->result);
  sqlite3_free(c);
  return SQLITE_OK;
}

static int fulltextNext(sqlite3_vtab_cursor *pCursor){
  fulltext_cursor *c = (fulltext_cursor *) pCursor;
  int rc;

................................................................................
}

/* 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 = sqlite3_realloc(q->pTerms, q->nTerms * sizeof(q->pTerms[0]));
  if( q->pTerms==0 ){
    q->nTerms = 0;
    return;
  }
  t = &q->pTerms[q->nTerms - 1];
  CLEAR(t);
  t->pTerm = sqlite3_malloc(nTerm+1);
  memcpy(t->pTerm, pTerm, nTerm);
  t->pTerm[nTerm] = 0;
  t->nTerm = nTerm;
  t->isOr = q->nextIsOr;
  t->isPrefix = 0;
  q->nextIsOr = 0;
  t->iColumn = q->nextColumn;
................................................................................
  DataBuffer term;           /* Leftmost term in block's subtree. */
  DataBuffer data;           /* Accumulated data for the block. */
  struct InteriorBlock *next;
} InteriorBlock;

static InteriorBlock *interiorBlockNew(int iHeight, sqlite_int64 iChildBlock,
                                       const char *pTerm, int nTerm){
  InteriorBlock *block = sqlite3_malloc(sizeof(InteriorBlock));
  char c[VARINT_MAX+VARINT_MAX];
  int n;

  if( block ){
    memset(block, 0, sizeof(*block));
    dataBufferInit(&block->term, 0);
    dataBufferReplace(&block->term, pTerm, nTerm);

    n = putVarint(c, iHeight);
    n += putVarint(c+n, iChildBlock);
    dataBufferInit(&block->data, INTERIOR_MAX);
    dataBufferReplace(&block->data, c, n);
  }
  return block;
}

#ifndef NDEBUG
/* Verify that the data is readable as an interior node. */
static void interiorBlockValidate(InteriorBlock *pBlock){
  const char *pData = pBlock->data.pData;
................................................................................
  InteriorBlock *block = pWriter->first;

  while( block!=NULL ){
    InteriorBlock *b = block;
    block = block->next;
    dataBufferDestroy(&b->term);
    dataBufferDestroy(&b->data);
    sqlite3_free(b);
  }
  if( pWriter->parentWriter!=NULL ){
    interiorWriterDestroy(pWriter->parentWriter);
    sqlite3_free(pWriter->parentWriter);
  }
  dataBufferDestroy(&pWriter->term);
  SCRAMBLE(pWriter);
  return SQLITE_OK;
}

/* If pWriter can fit entirely in ROOT_MAX, return it as the root info
................................................................................
  ** interior node.
  */
  ASSERT_VALID_INTERIOR_BLOCK(block);
  rc = block_insert(v, block->data.pData, block->data.nData, &iBlockid);
  if( rc!=SQLITE_OK ) return rc;
  *piEndBlockid = iBlockid;

  pWriter->parentWriter = sqlite3_malloc(sizeof(*pWriter->parentWriter));
  interiorWriterInit(pWriter->iHeight+1,
                     block->term.pData, block->term.nData,
                     iBlockid, pWriter->parentWriter);

  /* Flush additional blocks and append to the higher interior
  ** node.
  */
................................................................................
  DataBuffer dl;

  /* Determine the next index at level 0, merging as necessary. */
  rc = segdirNextIndex(v, 0, &idx);
  if( rc!=SQLITE_OK ) return rc;

  n = fts2HashCount(pTerms);
  pData = sqlite3_malloc(n*sizeof(TermData));

  for(i = 0, e = fts2HashFirst(pTerms); e; i++, e = fts2HashNext(e)){
    assert( i<n );
    pData[i].pTerm = fts2HashKey(e);
    pData[i].nTerm = fts2HashKeysize(e);
    pData[i].pCollector = fts2HashData(e);
  }
................................................................................
                        pData[i].pTerm, pData[i].nTerm, dl.pData, dl.nData);
    if( rc!=SQLITE_OK ) goto err;
  }
  rc = leafWriterFinalize(v, &writer);

 err:
  dataBufferDestroy(&dl);
  sqlite3_free(pData);
  leafWriterDestroy(&writer);
  return rc;
}

/* If pendingTerms has data, free it. */
static int clearPendingTerms(fulltext_vtab *v){
  if( v->nPendingData>=0 ){

Changes to ext/fts2/fts2_hash.c.

25
26
27
28
29
30
31

32
33
34
35
36
37
38
*/
#if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2)

#include <assert.h>
#include <stdlib.h>
#include <string.h>


#include "fts2_hash.h"

/*
** Malloc and Free functions
*/
static void *fts2HashMalloc(int n){
  void *p = sqlite3_malloc(n);







>







25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
*/
#if !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_FTS2)

#include <assert.h>
#include <stdlib.h>
#include <string.h>

#include "sqlite3.h"
#include "fts2_hash.h"

/*
** Malloc and Free functions
*/
static void *fts2HashMalloc(int n){
  void *p = sqlite3_malloc(n);

Changes to ext/fts2/fts2_porter.c.

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
..
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
...
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
...
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
** Create a new tokenizer instance.
*/
static int porterCreate(
  int argc, const char * const *argv,
  sqlite3_tokenizer **ppTokenizer
){
  porter_tokenizer *t;
  t = (porter_tokenizer *) calloc(sizeof(*t), 1);
  if( t==NULL ) return SQLITE_NOMEM;

  *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 
................................................................................
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(*c));
  if( c==NULL ) return SQLITE_NOMEM;

  c->zInput = zInput;
  if( zInput==0 ){
    c->nInput = 0;
  }else if( nInput<0 ){
    c->nInput = (int)strlen(zInput);
................................................................................

/*
** 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,
................................................................................
      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);
        if( c->zToken==NULL ) return SQLITE_NOMEM;
      }
      porter_stemmer(&z[iStartOffset], n, c->zToken, pnBytes);
      *pzToken = c->zToken;
      *piStartOffset = iStartOffset;
      *piEndOffset = c->iOffset;
      *piPosition = c->iToken++;







|
|
|








|







 







|







 







|
|







 







|







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
..
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
...
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
...
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
** Create a new tokenizer instance.
*/
static int porterCreate(
  int argc, const char * const *argv,
  sqlite3_tokenizer **ppTokenizer
){
  porter_tokenizer *t;
  t = (porter_tokenizer *) sqlite3_malloc(sizeof(*t));
  if( t==NULL ) return SQLITE_NOMEM;
  memset(t, 0, sizeof(*t));
  *ppTokenizer = &t->base;
  return SQLITE_OK;
}

/*
** Destroy a tokenizer
*/
static int porterDestroy(sqlite3_tokenizer *pTokenizer){
  sqlite3_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 
................................................................................
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 *) sqlite3_malloc(sizeof(*c));
  if( c==NULL ) return SQLITE_NOMEM;

  c->zInput = zInput;
  if( zInput==0 ){
    c->nInput = 0;
  }else if( nInput<0 ){
    c->nInput = (int)strlen(zInput);
................................................................................

/*
** 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;
  sqlite3_free(c->zToken);
  sqlite3_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,
................................................................................
      c->iOffset++;
    }

    if( c->iOffset>iStartOffset ){
      int n = c->iOffset-iStartOffset;
      if( n>c->nAllocated ){
        c->nAllocated = n+20;
        c->zToken = sqlite3_realloc(c->zToken, c->nAllocated);
        if( c->zToken==NULL ) return SQLITE_NOMEM;
      }
      porter_stemmer(&z[iStartOffset], n, c->zToken, pnBytes);
      *pzToken = c->zToken;
      *piStartOffset = iStartOffset;
      *piEndOffset = c->iOffset;
      *piPosition = c->iToken++;

Changes to ext/fts2/fts2_tokenizer1.c.

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
..
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
...
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
...
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
...
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
*/
static int simpleCreate(
  int argc, const char * const *argv,
  sqlite3_tokenizer **ppTokenizer
){
  simple_tokenizer *t;

  t = (simple_tokenizer *) calloc(sizeof(*t), 1);
  if( t==NULL ) return SQLITE_NOMEM;


  /* 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;
................................................................................
  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 
................................................................................
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(*c));
  if( c==NULL ) return SQLITE_NOMEM;

  c->pInput = pInput;
  if( pInput==0 ){
    c->nBytes = 0;
  }else if( nBytes<0 ){
    c->nBytes = (int)strlen(pInput);
................................................................................

/*
** 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().
*/
................................................................................
      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);
        if( c->pToken==NULL ) return SQLITE_NOMEM;
      }
      for(i=0; i<n; i++){
        /* TODO(shess) This needs expansion to handle UTF-8
        ** case-insensitivity.
        */
        unsigned char ch = p[iStartOffset+i];







|

>












|







 







|







 







|







 







|
|







 







|







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
..
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
...
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
...
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
...
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
*/
static int simpleCreate(
  int argc, const char * const *argv,
  sqlite3_tokenizer **ppTokenizer
){
  simple_tokenizer *t;

  t = (simple_tokenizer *) sqlite3_malloc(sizeof(*t));
  if( t==NULL ) return SQLITE_NOMEM;
  memset(t, 0, sizeof(*t));

  /* 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 ){
        sqlite3_free(t);
        return SQLITE_ERROR;
      }
      t->delim[ch] = 1;
    }
  } else {
    /* Mark non-alphanumeric ASCII characters as delimiters */
    int i;
................................................................................
  return SQLITE_OK;
}

/*
** Destroy a tokenizer
*/
static int simpleDestroy(sqlite3_tokenizer *pTokenizer){
  sqlite3_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 
................................................................................
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 *) sqlite3_malloc(sizeof(*c));
  if( c==NULL ) return SQLITE_NOMEM;

  c->pInput = pInput;
  if( pInput==0 ){
    c->nBytes = 0;
  }else if( nBytes<0 ){
    c->nBytes = (int)strlen(pInput);
................................................................................

/*
** 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;
  sqlite3_free(c->pToken);
  sqlite3_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().
*/
................................................................................
      c->iOffset++;
    }

    if( c->iOffset>iStartOffset ){
      int i, n = c->iOffset-iStartOffset;
      if( n>c->nTokenAllocated ){
        c->nTokenAllocated = n+20;
        c->pToken = sqlite3_realloc(c->pToken, c->nTokenAllocated);
        if( c->pToken==NULL ) return SQLITE_NOMEM;
      }
      for(i=0; i<n; i++){
        /* TODO(shess) This needs expansion to handle UTF-8
        ** case-insensitivity.
        */
        unsigned char ch = p[iStartOffset+i];

Added test/fts2.test.









































































































































>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
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
#
#    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 runs all tests.
#
# $Id: fts2.test,v 1.1 2008/07/22 22:57:54 shess Exp $

proc lshift {lvar} {
  upvar $lvar l
  set ret [lindex $l 0]
  set l [lrange $l 1 end]
  return $ret
}
while {[set arg [lshift argv]] != ""} {
  switch -- $arg {
    -sharedpagercache {
      sqlite3_enable_shared_cache 1
    }
    -soak {
       set SOAKTEST 1
    }
    default {
      set argv [linsert $argv 0 $arg]
      break
    }
  }
}

set testdir [file dirname $argv0]
source $testdir/tester.tcl
# If SQLITE_ENABLE_FTS2 is defined, omit this file.
ifcapable !fts2 {
  puts stderr "this build does not include FTS2 capability"
  exit 1
}
rename finish_test really_finish_test
proc finish_test {} {}
set ISQUICK 1

set EXCLUDE {
  fts2.test
}

# Files to include in the test.  If this list is empty then everything
# that is not in the EXCLUDE list is run.
#
set INCLUDE {
}

foreach testfile [lsort -dictionary [glob $testdir/fts2*.test]] {
  set tail [file tail $testfile]
  if {[lsearch -exact $EXCLUDE $tail]>=0} continue
  if {[llength $INCLUDE]>0 && [lsearch -exact $INCLUDE $tail]<0} continue
  source $testfile
  catch {db close}
  if {$sqlite_open_file_count>0} {
    puts "$tail did not close all files: $sqlite_open_file_count"
    incr nErr
    lappend ::failList $tail
    set sqlite_open_file_count 0
  }
}

set sqlite_open_file_count 0
really_finish_test