/ Changes On Branch fts3-matchinfo-y
Login

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

Changes In Branch fts3-matchinfo-y Excluding Merge-Ins

This is equivalent to a diff from 85a4a46c3b to 82e5a6e088

2015-05-11
19:01
Add new fts3 matchinfo option 'b'. Also optimize existing option 'y'. (check-in: 2e7679a1df user: dan tags: trunk)
18:48
Add missing "finish_test" commands to the end of the two new test scripts for sqlite3_analyzer and sqldiff. (check-in: 1d5e72b1c4 user: drh tags: trunk)
18:46
Merge latest trunk changes into this branch. (Closed-Leaf check-in: 82e5a6e088 user: dan tags: fts3-matchinfo-y)
17:46
Test cases for sqlite3_analyzer and sqldiff. Fix a problem with sqlite3_analyzer related to the renaming of the initialization routine. (check-in: 85a4a46c3b user: drh tags: trunk)
16:27
Fix minor Makefile typos. Improve consistency of MSVC makefile. Add new targets to clean. (check-in: f84fbe9899 user: mistachkin tags: trunk)
2015-05-06
18:15
Merge latest trunk changes with this branch. (check-in: 8a13e1fdbe user: dan tags: fts3-matchinfo-y)

Changes to ext/fts3/fts3.c.

1671
1672
1673
1674
1675
1676
1677
1678

1679
1680
1681
1682
1683
1684
1685
1671
1672
1673
1674
1675
1676
1677

1678
1679
1680
1681
1682
1683
1684
1685







-
+







static int fts3CloseMethod(sqlite3_vtab_cursor *pCursor){
  Fts3Cursor *pCsr = (Fts3Cursor *)pCursor;
  assert( ((Fts3Table *)pCsr->base.pVtab)->pSegments==0 );
  sqlite3_finalize(pCsr->pStmt);
  sqlite3Fts3ExprFree(pCsr->pExpr);
  sqlite3Fts3FreeDeferredTokens(pCsr);
  sqlite3_free(pCsr->aDoclist);
  sqlite3_free(pCsr->aMatchinfo);
  sqlite3Fts3MIBufferFree(pCsr->pMIBuffer);
  assert( ((Fts3Table *)pCsr->base.pVtab)->pSegments==0 );
  sqlite3_free(pCsr);
  return SQLITE_OK;
}

/*
** If pCsr->pStmt has not been prepared (i.e. if pCsr->pStmt==0), then
3172
3173
3174
3175
3176
3177
3178
3179

3180
3181
3182
3183
3184
3185
3186
3172
3173
3174
3175
3176
3177
3178

3179
3180
3181
3182
3183
3184
3185
3186







-
+







  if( idxNum & FTS3_HAVE_DOCID_GE ) pDocidGe = apVal[iIdx++];
  if( idxNum & FTS3_HAVE_DOCID_LE ) pDocidLe = apVal[iIdx++];
  assert( iIdx==nVal );

  /* In case the cursor has been used before, clear it now. */
  sqlite3_finalize(pCsr->pStmt);
  sqlite3_free(pCsr->aDoclist);
  sqlite3_free(pCsr->aMatchinfo);
  sqlite3Fts3MIBufferFree(pCsr->pMIBuffer);
  sqlite3Fts3ExprFree(pCsr->pExpr);
  memset(&pCursor[1], 0, sizeof(Fts3Cursor)-sizeof(sqlite3_vtab_cursor));

  /* Set the lower and upper bounds on docids to return */
  pCsr->iMinDocid = fts3DocidRange(pDocidGe, SMALLEST_INT64);
  pCsr->iMaxDocid = fts3DocidRange(pDocidLe, LARGEST_INT64);

Changes to ext/fts3/fts3Int.h.

193
194
195
196
197
198
199


200
201
202
203
204
205
206
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208







+
+








typedef struct Fts3Doclist Fts3Doclist;
typedef struct Fts3SegFilter Fts3SegFilter;
typedef struct Fts3DeferredToken Fts3DeferredToken;
typedef struct Fts3SegReader Fts3SegReader;
typedef struct Fts3MultiSegReader Fts3MultiSegReader;

typedef struct MatchinfoBuffer MatchinfoBuffer;

/*
** A connection to a fulltext index is an instance of the following
** structure. The xCreate and xConnect methods create an instance
** of this structure and xDestroy and xDisconnect free that instance.
** All other methods receive a pointer to the structure as one of their
** arguments.
*/
302
303
304
305
306
307
308
309
310
311

312
313
314
315
316
317
318
304
305
306
307
308
309
310



311
312
313
314
315
316
317
318







-
-
-
+







  u8 bDesc;                       /* True to sort in descending order */
  int eEvalmode;                  /* An FTS3_EVAL_XX constant */
  int nRowAvg;                    /* Average size of database rows, in pages */
  sqlite3_int64 nDoc;             /* Documents in table */
  i64 iMinDocid;                  /* Minimum docid to return */
  i64 iMaxDocid;                  /* Maximum docid to return */
  int isMatchinfoNeeded;          /* True when aMatchinfo[] needs filling in */
  u32 *aMatchinfo;                /* Information about most recent match */
  int nMatchinfo;                 /* Number of elements in aMatchinfo[] */
  char *zMatchinfo;               /* Matchinfo specification */
  MatchinfoBuffer *pMIBuffer;     /* Buffer for matchinfo data */
};

#define FTS3_EVAL_FILTER    0
#define FTS3_EVAL_NEXT      1
#define FTS3_EVAL_MATCHINFO 2

/*
424
425
426
427
428
429
430


431

432
433
434
435
436
437
438
424
425
426
427
428
429
430
431
432

433
434
435
436
437
438
439
440







+
+
-
+








  /* The following are used by the fts3_eval.c module. */
  sqlite3_int64 iDocid;      /* Current docid */
  u8 bEof;                   /* True this expression is at EOF already */
  u8 bStart;                 /* True if iDocid is valid */
  u8 bDeferred;              /* True if this expression is entirely deferred */

  /* The following are used by the fts3_snippet.c module. */
  int iPhrase;               /* Index of this phrase in matchinfo() results */
  u32 *aMI;
  u32 *aMI;                  /* See above */
};

/*
** Candidate values for Fts3Query.eType. Note that the order of the first
** four values is in order of precedence when parsing expressions. For 
** example, the following:
**
560
561
562
563
564
565
566

567
568
569
570
571
572
573
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576







+








/* fts3_snippet.c */
void sqlite3Fts3Offsets(sqlite3_context*, Fts3Cursor*);
void sqlite3Fts3Snippet(sqlite3_context *, Fts3Cursor *, const char *,
  const char *, const char *, int, int
);
void sqlite3Fts3Matchinfo(sqlite3_context *, Fts3Cursor *, const char *);
void sqlite3Fts3MIBufferFree(MatchinfoBuffer *p);

/* fts3_expr.c */
int sqlite3Fts3ExprParse(sqlite3_tokenizer *, int,
  char **, int, int, int, const char *, int, Fts3Expr **, char **
);
void sqlite3Fts3ExprFree(Fts3Expr *);
#ifdef SQLITE_TEST

Changes to ext/fts3/fts3_snippet.c.

24
25
26
27
28
29
30

31
32
33
34
35
36
37
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38







+







#define FTS3_MATCHINFO_NCOL      'c'        /* 1 value */
#define FTS3_MATCHINFO_NDOC      'n'        /* 1 value */
#define FTS3_MATCHINFO_AVGLENGTH 'a'        /* nCol values */
#define FTS3_MATCHINFO_LENGTH    'l'        /* nCol values */
#define FTS3_MATCHINFO_LCS       's'        /* nCol values */
#define FTS3_MATCHINFO_HITS      'x'        /* 3*nCol*nPhrase values */
#define FTS3_MATCHINFO_LHITS     'y'        /* nCol*nPhrase values */
#define FTS3_MATCHINFO_LHITS_BM  'b'        /* nCol*nPhrase values */

/*
** The default value for the second argument to matchinfo(). 
*/
#define FTS3_MATCHINFO_DEFAULT   "pcx"


85
86
87
88
89
90
91

92
93
94












95
96
97
98
99
100
101
102
103
104
105
106
107
108



























































































109
110
111
112
113
114
115
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220







+



+
+
+
+
+
+
+
+
+
+
+
+














+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+







*/
typedef struct MatchInfo MatchInfo;
struct MatchInfo {
  Fts3Cursor *pCursor;            /* FTS3 Cursor */
  int nCol;                       /* Number of columns in table */
  int nPhrase;                    /* Number of matchable phrases in query */
  sqlite3_int64 nDoc;             /* Number of docs in database */
  char flag;
  u32 *aMatchinfo;                /* Pre-allocated buffer */
};

/*
** An instance of this structure is used to manage a pair of buffers, each
** (nElem * sizeof(u32)) bytes in size. See the MatchinfoBuffer code below
** for details.
*/
struct MatchinfoBuffer {
  u8 aRef[3];
  int nElem;
  int bGlobal;                    /* Set if global data is loaded */
  char *zMatchinfo;
  u32 aMatchinfo[0];
};


/*
** The snippet() and offsets() functions both return text values. An instance
** of the following structure is used to accumulate those values while the
** functions are running. See fts3StringAppend() for details.
*/
typedef struct StrBuffer StrBuffer;
struct StrBuffer {
  char *z;                        /* Pointer to buffer containing string */
  int n;                          /* Length of z in bytes (excl. nul-term) */
  int nAlloc;                     /* Allocated size of buffer z in bytes */
};


/*************************************************************************
** Start of MatchinfoBuffer code.
*/

/*
** Allocate a two-slot MatchinfoBuffer object.
*/
static MatchinfoBuffer *fts3MIBufferNew(int nElem, const char *zMatchinfo){
  MatchinfoBuffer *pRet;
  int nByte = sizeof(u32) * (2*nElem + 2) + sizeof(MatchinfoBuffer);
  int nStr = strlen(zMatchinfo);

  pRet = sqlite3_malloc(nByte + nStr+1);
  if( pRet ){
    memset(pRet, 0, nByte);
    pRet->aMatchinfo[0] = (u8*)(&pRet->aMatchinfo[1]) - (u8*)pRet;
    pRet->aMatchinfo[1+nElem] = pRet->aMatchinfo[0] + sizeof(u32)*(nElem+1);
    pRet->nElem = nElem;
    pRet->zMatchinfo = ((char*)pRet) + nByte;
    memcpy(pRet->zMatchinfo, zMatchinfo, nStr+1);
    pRet->aRef[0] = 1;
  }

  return pRet;
}

static void fts3MIBufferFree(void *p){
  MatchinfoBuffer *pBuf = (MatchinfoBuffer*)((u8*)p - ((u32*)p)[-1]);

  assert( (u32*)p==&pBuf->aMatchinfo[1] 
       || (u32*)p==&pBuf->aMatchinfo[pBuf->nElem+2] 
  );
  if( (u32*)p==&pBuf->aMatchinfo[1] ){
    pBuf->aRef[1] = 0;
  }else{
    pBuf->aRef[2] = 0;
  }

  if( pBuf->aRef[0]==0 && pBuf->aRef[1]==0 && pBuf->aRef[2]==0 ){
    sqlite3_free(pBuf);
  }
}

static void (*fts3MIBufferAlloc(MatchinfoBuffer *p, u32 **paOut))(void*){
  void (*xRet)(void*) = 0;
  u32 *aOut = 0;

  if( p->aRef[1]==0 ){
    p->aRef[1] = 1;
    aOut = &p->aMatchinfo[1];
    xRet = fts3MIBufferFree;
  }
  else if( p->aRef[2]==0 ){
    p->aRef[2] = 1;
    aOut = &p->aMatchinfo[p->nElem+2];
    xRet = fts3MIBufferFree;
  }else{
    aOut = (u32*)sqlite3_malloc(p->nElem * sizeof(u32));
    if( aOut ){
      xRet = sqlite3_free;
      if( p->bGlobal ) memcpy(aOut, &p->aMatchinfo[1], p->nElem*sizeof(u32));
    }
  }

  *paOut = aOut;
  return xRet;
}

static void fts3MIBufferSetGlobal(MatchinfoBuffer *p){
  p->bGlobal = 1;
  memcpy(&p->aMatchinfo[2+p->nElem], &p->aMatchinfo[1], p->nElem*sizeof(u32));
}

/*
** Free a MatchinfoBuffer object allocated using fts3MIBufferNew()
*/
void sqlite3Fts3MIBufferFree(MatchinfoBuffer *p){
  if( p ){
    assert( p->aRef[0]==1 );
    p->aRef[0] = 0;
    if( p->aRef[0]==0 && p->aRef[1]==0 && p->aRef[2]==0 ){
      sqlite3_free(p);
    }
  }
}

/* 
** End of MatchinfoBuffer code.
*************************************************************************/


/*
** This function is used to help iterate through a position-list. A position
** list is a list of unique integers, sorted from smallest to largest. Each
** element of the list is represented by an FTS3 varint that takes the value
** of the difference between the current element and the previous one plus
** two. For example, to store the position-list:
139
140
141
142
143
144
145
146

147
148
149
150
151
152
153
244
245
246
247
248
249
250

251
252
253
254
255
256
257
258







-
+







static int fts3ExprIterate2(
  Fts3Expr *pExpr,                /* Expression to iterate phrases of */
  int *piPhrase,                  /* Pointer to phrase counter */
  int (*x)(Fts3Expr*,int,void*),  /* Callback function to invoke for phrases */
  void *pCtx                      /* Second argument to pass to callback */
){
  int rc;                         /* Return code */
  int eType = pExpr->eType;       /* Type of expression node pExpr */
  int eType = pExpr->eType;     /* Type of expression node pExpr */

  if( eType!=FTSQUERY_PHRASE ){
    assert( pExpr->pLeft && pExpr->pRight );
    rc = fts3ExprIterate2(pExpr->pLeft, piPhrase, x, pCtx);
    if( rc==SQLITE_OK && eType!=FTSQUERY_NOT ){
      rc = fts3ExprIterate2(pExpr->pRight, piPhrase, x, pCtx);
    }
172
173
174
175
176
177
178

179
180
181
182
183
184
185
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291







+







  Fts3Expr *pExpr,                /* Expression to iterate phrases of */
  int (*x)(Fts3Expr*,int,void*),  /* Callback function to invoke for phrases */
  void *pCtx                      /* Second argument to pass to callback */
){
  int iPhrase = 0;                /* Variable used as the phrase counter */
  return fts3ExprIterate2(pExpr, &iPhrase, x, pCtx);
}


/*
** This is an fts3ExprIterate() callback used while loading the doclists
** for each phrase into Fts3Expr.aDoclist[]/nDoclist. See also
** fts3ExprLoadDoclists().
*/
static int fts3ExprLoadDoclistsCb(Fts3Expr *pExpr, int iPhrase, void *ctx){
217
218
219
220
221
222
223
224
225

226
227
228
229
230
231
232
323
324
325
326
327
328
329


330
331
332
333
334
335
336
337







-
-
+







  if( pnPhrase ) *pnPhrase = sCtx.nPhrase;
  if( pnToken ) *pnToken = sCtx.nToken;
  return rc;
}

static int fts3ExprPhraseCountCb(Fts3Expr *pExpr, int iPhrase, void *ctx){
  (*(int *)ctx)++;
  UNUSED_PARAMETER(pExpr);
  UNUSED_PARAMETER(iPhrase);
  pExpr->iPhrase = iPhrase;
  return SQLITE_OK;
}
static int fts3ExprPhraseCount(Fts3Expr *pExpr){
  int nPhrase = 0;
  (void)fts3ExprIterate(pExpr, fts3ExprPhraseCountCb, (void *)&nPhrase);
  return nPhrase;
}
439
440
441
442
443
444
445
446

447
448
449
450
451
452
453
544
545
546
547
548
549
550

551
552
553
554
555
556
557
558







-
+







  ** the set of phrases in the expression to populate the aPhrase[] array.
  */
  sIter.pCsr = pCsr;
  sIter.iCol = iCol;
  sIter.nSnippet = nSnippet;
  sIter.nPhrase = nList;
  sIter.iCurrent = -1;
  rc = fts3ExprIterate(pCsr->pExpr, fts3SnippetFindPositions, (void *)&sIter);
  rc = fts3ExprIterate(pCsr->pExpr, fts3SnippetFindPositions, (void*)&sIter);
  if( rc==SQLITE_OK ){

    /* Set the *pmSeen output variable. */
    for(i=0; i<nList; i++){
      if( sIter.aPhrase[i].pHead ){
        *pmSeen |= (u64)1 << i;
      }
739
740
741
742
743
744
745






















































746
747
748
749
750
751
752
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911







+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+







    c = *pEnd++ & 0x80;
    if( !c ) nEntry++;
  }

  *ppCollist = pEnd;
  return nEntry;
}

/*
** This function gathers 'y' or 'b' data for a single phrase.
*/
static void fts3ExprLHits(
  Fts3Expr *pExpr,                /* Phrase expression node */
  MatchInfo *p                    /* Matchinfo context */
){
  Fts3Table *pTab = (Fts3Table *)p->pCursor->base.pVtab;
  int iStart;
  Fts3Phrase *pPhrase = pExpr->pPhrase;
  char *pIter = pPhrase->doclist.pList;
  int iCol = 0;

  assert( p->flag==FTS3_MATCHINFO_LHITS_BM || p->flag==FTS3_MATCHINFO_LHITS );
  if( p->flag==FTS3_MATCHINFO_LHITS ){
    iStart = pExpr->iPhrase * p->nCol;
  }else{
    iStart = pExpr->iPhrase * ((p->nCol + 31) / 32);
  }

  while( 1 ){
    int nHit = fts3ColumnlistCount(&pIter);
    if( (pPhrase->iColumn>=pTab->nColumn || pPhrase->iColumn==iCol) ){
      if( p->flag==FTS3_MATCHINFO_LHITS ){
        p->aMatchinfo[iStart + iCol] = (u32)nHit;
      }else if( nHit ){
        p->aMatchinfo[iStart + (iCol+1)/32] |= (1 << (iCol&0x1F));
      }
    }
    assert( *pIter==0x00 || *pIter==0x01 );
    if( *pIter!=0x01 ) break;
    pIter++;
    pIter += fts3GetVarint32(pIter, &iCol);
  }
}

/*
** Gather the results for matchinfo directives 'y' and 'b'.
*/
static void fts3ExprLHitGather(
  Fts3Expr *pExpr,
  MatchInfo *p
){
  assert( (pExpr->pLeft==0)==(pExpr->pRight==0) );
  if( pExpr->bEof==0 && pExpr->iDocid==p->pCursor->iPrevId ){
    if( pExpr->pLeft ){
      fts3ExprLHitGather(pExpr->pLeft, p);
      fts3ExprLHitGather(pExpr->pRight, p);
    }else{
      fts3ExprLHits(pExpr, p);
    }
  }
}

/*
** fts3ExprIterate() callback used to collect the "global" matchinfo stats
** for a single query. 
**
** fts3ExprIterate() callback to load the 'global' elements of a
** FTS3_MATCHINFO_HITS matchinfo array. The global stats are those elements 
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870

871
872
873
874
875
876
877
965
966
967
968
969
970
971













































972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992







-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-













+







      p->aMatchinfo[iStart+i*3] = 0;
    }
  }

  return rc;
}

/*
** fts3ExprIterate() callback used to gather information for the matchinfo
** directive 'y'.
*/
static int fts3ExprLHitsCb(
  Fts3Expr *pExpr,                /* Phrase expression node */
  int iPhrase,                    /* Phrase number */
  void *pCtx                      /* Pointer to MatchInfo structure */
){
  MatchInfo *p = (MatchInfo *)pCtx;
  Fts3Table *pTab = (Fts3Table *)p->pCursor->base.pVtab;
  int rc = SQLITE_OK;
  int iStart = iPhrase * p->nCol;
  Fts3Expr *pEof;                 /* Ancestor node already at EOF */
  
  /* This must be a phrase */
  assert( pExpr->pPhrase );

  /* Initialize all output integers to zero. */
  memset(&p->aMatchinfo[iStart], 0, sizeof(u32) * p->nCol);

  /* Check if this or any parent node is at EOF. If so, then all output
  ** values are zero.  */
  for(pEof=pExpr; pEof && pEof->bEof==0; pEof=pEof->pParent);

  if( pEof==0 && pExpr->iDocid==p->pCursor->iPrevId ){
    Fts3Phrase *pPhrase = pExpr->pPhrase;
    char *pIter = pPhrase->doclist.pList;
    int iCol = 0;

    while( 1 ){
      int nHit = fts3ColumnlistCount(&pIter);
      if( (pPhrase->iColumn>=pTab->nColumn || pPhrase->iColumn==iCol) ){
        p->aMatchinfo[iStart + iCol] = (u32)nHit;
      }
      assert( *pIter==0x00 || *pIter==0x01 );
      if( *pIter!=0x01 ) break;
      pIter++;
      pIter += fts3GetVarint32(pIter, &iCol);
    }
  }

  return rc;
}

static int fts3MatchinfoCheck(
  Fts3Table *pTab, 
  char cArg,
  char **pzErr
){
  if( (cArg==FTS3_MATCHINFO_NPHRASE)
   || (cArg==FTS3_MATCHINFO_NCOL)
   || (cArg==FTS3_MATCHINFO_NDOC && pTab->bFts4)
   || (cArg==FTS3_MATCHINFO_AVGLENGTH && pTab->bFts4)
   || (cArg==FTS3_MATCHINFO_LENGTH && pTab->bHasDocsize)
   || (cArg==FTS3_MATCHINFO_LCS)
   || (cArg==FTS3_MATCHINFO_HITS)
   || (cArg==FTS3_MATCHINFO_LHITS)
   || (cArg==FTS3_MATCHINFO_LHITS_BM)
  ){
    return SQLITE_OK;
  }
  sqlite3Fts3ErrMsg(pzErr, "unrecognized matchinfo request: %c", cArg);
  return SQLITE_ERROR;
}

890
891
892
893
894
895
896




897
898
899
900
901
902
903
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022







+
+
+
+







    case FTS3_MATCHINFO_LCS:
      nVal = pInfo->nCol;
      break;

    case FTS3_MATCHINFO_LHITS:
      nVal = pInfo->nCol * pInfo->nPhrase;
      break;

    case FTS3_MATCHINFO_LHITS_BM:
      nVal = pInfo->nPhrase * ((pInfo->nCol + 31) / 32);
      break;

    default:
      assert( cArg==FTS3_MATCHINFO_HITS );
      nVal = pInfo->nCol * pInfo->nPhrase * 3;
      break;
  }

1085
1086
1087
1088
1089
1090
1091
1092

1093
1094
1095
1096
1097
1098
1099
1204
1205
1206
1207
1208
1209
1210

1211
1212
1213
1214
1215
1216
1217
1218







-
+







){
  int rc = SQLITE_OK;
  int i;
  Fts3Table *pTab = (Fts3Table *)pCsr->base.pVtab;
  sqlite3_stmt *pSelect = 0;

  for(i=0; rc==SQLITE_OK && zArg[i]; i++){

    pInfo->flag = zArg[i];
    switch( zArg[i] ){
      case FTS3_MATCHINFO_NPHRASE:
        if( bGlobal ) pInfo->aMatchinfo[0] = pInfo->nPhrase;
        break;

      case FTS3_MATCHINFO_NCOL:
        if( bGlobal ) pInfo->aMatchinfo[0] = pInfo->nCol;
1145
1146
1147
1148
1149
1150
1151

1152
1153




1154

1155
1156
1157
1158
1159
1160
1161
1264
1265
1266
1267
1268
1269
1270
1271


1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284







+
-
-
+
+
+
+

+







      case FTS3_MATCHINFO_LCS:
        rc = fts3ExprLoadDoclists(pCsr, 0, 0);
        if( rc==SQLITE_OK ){
          rc = fts3MatchinfoLcs(pCsr, pInfo);
        }
        break;

      case FTS3_MATCHINFO_LHITS_BM:
      case FTS3_MATCHINFO_LHITS:
        (void)fts3ExprIterate(pCsr->pExpr, fts3ExprLHitsCb, (void*)pInfo);
      case FTS3_MATCHINFO_LHITS: {
        int nZero = fts3MatchinfoSize(pInfo, zArg[i]) * sizeof(u32);
        memset(pInfo->aMatchinfo, 0, nZero);
        fts3ExprLHitGather(pCsr->pExpr, pInfo);
        break;
      }

      default: {
        Fts3Expr *pExpr;
        assert( zArg[i]==FTS3_MATCHINFO_HITS );
        pExpr = pCsr->pExpr;
        rc = fts3ExprLoadDoclists(pCsr, 0, 0);
        if( rc!=SQLITE_OK ) break;
1180
1181
1182
1183
1184
1185
1186
1187


1188
1189
1190
1191
1192
1193
1194
1195



1196
1197
1198
1199
1200
1201
1202
1203

1204
1205
1206


1207
1208
1209
1210

1211
1212
1213
1214
1215

1216
1217
1218
1219
1220
1221
1222
1223
1224
1225






1226
1227
1228
1229
1230

1231
1232

1233
1234
1235
1236
1237
1238
1239
1240
1241








1242
1243


1244
1245

1246
1247
1248
1249











1250
1251
1252
1253
1254
1255
1256
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







-
+
+








+
+
+







-
+
-
-
-
+
+
-


-
+




-
+

-








+
+
+
+
+
+




-
+
-
-
+

-
-
-
-




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

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







}


/*
** Populate pCsr->aMatchinfo[] with data for the current row. The 
** 'matchinfo' data is an array of 32-bit unsigned integers (C type u32).
*/
static int fts3GetMatchinfo(
static void fts3GetMatchinfo(
  sqlite3_context *pCtx,        /* Return results here */
  Fts3Cursor *pCsr,               /* FTS3 Cursor object */
  const char *zArg                /* Second argument to matchinfo() function */
){
  MatchInfo sInfo;
  Fts3Table *pTab = (Fts3Table *)pCsr->base.pVtab;
  int rc = SQLITE_OK;
  int bGlobal = 0;                /* Collect 'global' stats as well as local */

  u32 *aOut = 0;
  void (*xDestroyOut)(void*) = 0;

  memset(&sInfo, 0, sizeof(MatchInfo));
  sInfo.pCursor = pCsr;
  sInfo.nCol = pTab->nColumn;

  /* If there is cached matchinfo() data, but the format string for the 
  ** cache does not match the format string for this request, discard 
  ** the cached data. */
  if( pCsr->zMatchinfo && strcmp(pCsr->zMatchinfo, zArg) ){
  if( pCsr->pMIBuffer && strcmp(pCsr->pMIBuffer->zMatchinfo, zArg) ){
    assert( pCsr->aMatchinfo );
    sqlite3_free(pCsr->aMatchinfo);
    pCsr->zMatchinfo = 0;
    sqlite3Fts3MIBufferFree(pCsr->pMIBuffer);
    pCsr->pMIBuffer = 0;
    pCsr->aMatchinfo = 0;
  }

  /* If Fts3Cursor.aMatchinfo[] is NULL, then this is the first time the
  /* If Fts3Cursor.pMIBuffer is NULL, then this is the first time the
  ** matchinfo function has been called for this query. In this case 
  ** allocate the array used to accumulate the matchinfo data and
  ** initialize those elements that are constant for every row.
  */
  if( pCsr->aMatchinfo==0 ){
  if( pCsr->pMIBuffer==0 ){
    int nMatchinfo = 0;           /* Number of u32 elements in match-info */
    int nArg;                     /* Bytes in zArg */
    int i;                        /* Used to iterate through zArg */

    /* Determine the number of phrases in the query */
    pCsr->nPhrase = fts3ExprPhraseCount(pCsr->pExpr);
    sInfo.nPhrase = pCsr->nPhrase;

    /* Determine the number of integers in the buffer returned by this call. */
    for(i=0; zArg[i]; i++){
      char *zErr = 0;
      if( fts3MatchinfoCheck(pTab, zArg[i], &zErr) ){
        sqlite3_result_error(pCtx, zErr, -1);
        sqlite3_free(zErr);
        return;
      }
      nMatchinfo += fts3MatchinfoSize(&sInfo, zArg[i]);
    }

    /* Allocate space for Fts3Cursor.aMatchinfo[] and Fts3Cursor.zMatchinfo. */
    nArg = (int)strlen(zArg);
    pCsr->pMIBuffer = fts3MIBufferNew(nMatchinfo, zArg);
    pCsr->aMatchinfo = (u32 *)sqlite3_malloc(sizeof(u32)*nMatchinfo + nArg + 1);
    if( !pCsr->aMatchinfo ) return SQLITE_NOMEM;
    if( !pCsr->pMIBuffer ) rc = SQLITE_NOMEM;

    pCsr->zMatchinfo = (char *)&pCsr->aMatchinfo[nMatchinfo];
    pCsr->nMatchinfo = nMatchinfo;
    memcpy(pCsr->zMatchinfo, zArg, nArg+1);
    memset(pCsr->aMatchinfo, 0, sizeof(u32)*nMatchinfo);
    pCsr->isMatchinfoNeeded = 1;
    bGlobal = 1;
  }

  if( rc==SQLITE_OK ){
    xDestroyOut = fts3MIBufferAlloc(pCsr->pMIBuffer, &aOut);
    if( xDestroyOut==0 ){
      rc = SQLITE_NOMEM;
    }
  }

  if( rc==SQLITE_OK ){
  sInfo.aMatchinfo = pCsr->aMatchinfo;
  sInfo.nPhrase = pCsr->nPhrase;
    sInfo.aMatchinfo = aOut;
    sInfo.nPhrase = pCsr->nPhrase;
  if( pCsr->isMatchinfoNeeded ){
    rc = fts3MatchinfoValues(pCsr, bGlobal, &sInfo, zArg);
    if( bGlobal ){
    pCsr->isMatchinfoNeeded = 0;
  }

  return rc;
      fts3MIBufferSetGlobal(pCsr->pMIBuffer);
    }
  }

  if( rc!=SQLITE_OK ){
    sqlite3_result_error_code(pCtx, rc);
    if( xDestroyOut ) xDestroyOut(aOut);
  }else{
    int n = pCsr->pMIBuffer->nElem * sizeof(u32);
    sqlite3_result_blob(pCtx, aOut, n, xDestroyOut);
  }
}

/*
** Implementation of snippet() function.
*/
void sqlite3Fts3Snippet(
  sqlite3_context *pCtx,          /* SQLite function call context */
1448
1449
1450
1451
1452
1453
1454
1455

1456
1457
1458
1459
1460
1461
1462
1588
1589
1590
1591
1592
1593
1594

1595
1596
1597
1598
1599
1600
1601
1602







-
+








    /* Initialize the contents of sCtx.aTerm[] for column iCol. There is 
    ** no way that this operation can fail, so the return code from
    ** fts3ExprIterate() can be discarded.
    */
    sCtx.iCol = iCol;
    sCtx.iTerm = 0;
    (void)fts3ExprIterate(pCsr->pExpr, fts3ExprTermOffsetInit, (void *)&sCtx);
    (void)fts3ExprIterate(pCsr->pExpr, fts3ExprTermOffsetInit, (void*)&sCtx);

    /* Retreive the text stored in column iCol. If an SQL NULL is stored 
    ** in column iCol, jump immediately to the next iteration of the loop.
    ** If an OOM occurs while retrieving the data (this can happen if SQLite
    ** needs to transform the data from utf-16 to utf-8), return SQLITE_NOMEM 
    ** to the caller. 
    */
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
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







-
-



-
-
-
-
-
-
-
-








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




*/
void sqlite3Fts3Matchinfo(
  sqlite3_context *pContext,      /* Function call context */
  Fts3Cursor *pCsr,               /* FTS3 table cursor */
  const char *zArg                /* Second arg to matchinfo() function */
){
  Fts3Table *pTab = (Fts3Table *)pCsr->base.pVtab;
  int rc;
  int i;
  const char *zFormat;

  if( zArg ){
    for(i=0; zArg[i]; i++){
      char *zErr = 0;
      if( fts3MatchinfoCheck(pTab, zArg[i], &zErr) ){
        sqlite3_result_error(pContext, zErr, -1);
        sqlite3_free(zErr);
        return;
      }
    }
    zFormat = zArg;
  }else{
    zFormat = FTS3_MATCHINFO_DEFAULT;
  }

  if( !pCsr->pExpr ){
    sqlite3_result_blob(pContext, "", 0, SQLITE_STATIC);
    return;
  }
  }else{

  /* Retrieve matchinfo() data. */
  rc = fts3GetMatchinfo(pCsr, zFormat);
  sqlite3Fts3SegmentsClose(pTab);
    /* Retrieve matchinfo() data. */
    fts3GetMatchinfo(pContext, pCsr, zFormat);
    sqlite3Fts3SegmentsClose(pTab);

  if( rc!=SQLITE_OK ){
    sqlite3_result_error_code(pContext, rc);
  }else{
    int n = pCsr->nMatchinfo * sizeof(u32);
    sqlite3_result_blob(pContext, pCsr->aMatchinfo, n, SQLITE_TRANSIENT);
  }
}

#endif

Changes to test/fts3matchinfo.test.

503
504
505
506
507
508
509
510

511
512
513




















514
515



















516

503
504
505
506
507
508
509

510
511
512

513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555







-
+


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


+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+

+
  7 "a OR (a AND b)" {
      1 {1 2 1 2 0 1}   2 {1 0 1 0 1 0}   3 {0 1 0 1 1 2}   4 {1 0 1 0 0 1}   
      5 {1 0 1 0 0 1}   6 {1 0 1 0 2 2}   7 {2 1 0 0 0 0}   8 {1 2 1 2 2 1}   
      9 {1 1 1 1 1 3}  10 {1 3 0 0 0 0}
  }

} {
  do_execsql_test 11.1.$tn  {
  do_execsql_test 11.1.$tn.1  {
    SELECT rowid, mit(matchinfo(tt, 'y')) FROM tt WHERE tt MATCH $expr
  } $res
}

  set r2 [list]
  foreach {rowid L} $res {
    lappend r2 $rowid
    set M [list]
    foreach {a b} $L {
      lappend M [expr ($a ? 1 : 0) + ($b ? 2 : 0)]
    }
    lappend r2 $M
  }

  do_execsql_test 11.1.$tn.2  {
    SELECT rowid, mit(matchinfo(tt, 'b')) FROM tt WHERE tt MATCH $expr
  } $r2
  breakpoint

  do_execsql_test 11.1.$tn.2  {
    SELECT rowid, mit(matchinfo(tt, 'b')) FROM tt WHERE tt MATCH $expr
  } $r2
}
set sqlite_fts3_enable_parentheses 0

#---------------------------------------------------------------------------
# Test the 'b' matchinfo flag
#
set sqlite_fts3_enable_parentheses 1
reset_db
db func mit mit

do_test 12.0 {
  set cols [list]
  for {set i 0} {$i < 50} {incr i} { lappend cols "c$i" }
  execsql "CREATE VIRTUAL TABLE tt USING fts3([join $cols ,])"
} {}

do_execsql_test 12.1 {
  INSERT INTO tt (rowid, c4, c45) VALUES(1, 'abc', 'abc');
  SELECT mit(matchinfo(tt, 'b')) FROM tt WHERE tt MATCH 'abc';
} [list [list [expr 1<<4] [expr 1<<(45-32)]]]

set sqlite_fts3_enable_parentheses 0
finish_test

Changes to test/fts3query.test.

169
170
171
172
173
174
175
176
177


178
179
180
181
182
183
184
169
170
171
172
173
174
175


176
177
178
179
180
181
182
183
184







-
-
+
+







} {
  1 "SELECT matchinfo(content) FROM t2 WHERE t2 MATCH 'history'" matchinfo
  2 "SELECT offsets(content) FROM t2 WHERE t2 MATCH 'history'"   offsets
  3 "SELECT snippet(content) FROM t2 WHERE t2 MATCH 'history'"   snippet
  4 "SELECT optimize(content) FROM t2 WHERE t2 MATCH 'history'"  optimize
}
do_catchsql_test 5.5.1 {
  SELECT matchinfo(t2, 'abc') FROM t2 WHERE t2 MATCH 'history'
} {1 {unrecognized matchinfo request: b}}
  SELECT matchinfo(t2, 'abcd') FROM t2 WHERE t2 MATCH 'history'
} {1 {unrecognized matchinfo request: d}}

do_execsql_test 5.5 { DROP TABLE t2 }


# Test the snippet() function with 1 to 6 arguments.
# 
do_execsql_test 6.1 {