SQLite

Check-in [111ca61671]
Login

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

Overview
Comment:When gathering a doclist for querying, don't discard empty position lists until the end; this allows empty position lists to override non-empty lists encountered later in the gathering process. This fixes #1982, which was caused by the fact that for all-column queries we weren't discarding empty position lists at all. (CVS 3433)
Downloads: Tarball | ZIP archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: 111ca616713dd89b5d1e114de29c83256731c482
User & Date: adamd 2006-09-21 20:56:52.000
Context
2006-09-22
00:06
Implemented UPDATE for full-text tables.

We handle an UPDATE to a row by performing an UPDATE on the content table and by building new position lists for each term which appears in either the old or new versions of the row. We write these position lists all at once; this is presumably more efficient than a delete followed by an insert (which would first write empty position lists, then new position lists). (CVS 3434) (check-in: 757fa22400 user: adamd tags: trunk)

2006-09-21
20:56
When gathering a doclist for querying, don't discard empty position lists until the end; this allows empty position lists to override non-empty lists encountered later in the gathering process. This fixes #1982, which was caused by the fact that for all-column queries we weren't discarding empty position lists at all. (CVS 3433) (check-in: 111ca61671 user: adamd tags: trunk)
11:02
Be more aggressive with the SQLITE_OMIT_VACUUM macro. Saves about 150 bytes of code space. (CVS 3432) (check-in: 7e618db457 user: drh tags: trunk)
Changes
Unified Diff Ignore Whitespace Patch
Changes to ext/fts1/fts1.c.
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
  sqlite_int64 ret;
  assert( !atEnd(pReader) );
  assert( pReader->iLastPos==-1 );
  getVarint(pReader->p, &ret);
  return ret;
}

/* Read the next docid.   See also nextValidDocid().
*/
static sqlite_int64 readDocid(DocListReader *pReader){
  sqlite_int64 ret;
  assert( !atEnd(pReader) );
  assert( pReader->iLastPos==-1 );
  pReader->p += getVarint(pReader->p, &ret);
  if( pReader->pDoclist->iType>=DL_POSITIONS ){







|







335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
  sqlite_int64 ret;
  assert( !atEnd(pReader) );
  assert( pReader->iLastPos==-1 );
  getVarint(pReader->p, &ret);
  return ret;
}

/* Read the next docid.   See also nextDocid().
*/
static sqlite_int64 readDocid(DocListReader *pReader){
  sqlite_int64 ret;
  assert( !atEnd(pReader) );
  assert( pReader->iLastPos==-1 );
  pReader->p += getVarint(pReader->p, &ret);
  if( pReader->pDoclist->iType>=DL_POSITIONS ){
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475


476
477


























478
479
480
481
482
483
484
485
486
487
488
489
490
    }
  }
  printf("\n");
  fflush(stdout);
}
#endif /* SQLITE_DEBUG */

/* Trim the given doclist to contain only positions in column [iRestrictColumn],
 * discarding any docids without any remaining positions. */
static void docListRestrictColumn(DocList *in, int iRestrictColumn){
  DocListReader r;
  DocList out;

  assert( in->iType>=DL_POSITIONS );
  readerInit(&r, in);
  docListInit(&out, DL_POSITIONS, NULL, 0);

  while( !atEnd(&r) ){
    sqlite_int64 iDocid = readDocid(&r);
    int match = 0;
    int iPos, iColumn;


    while( (iPos = readPosition(&r, &iColumn)) != -1 ){
      if( iColumn==iRestrictColumn ){


























        if( !match ){
          docListAddDocid(&out, iDocid);
          match = 1;
        }
        docListAddPos(&out, iColumn, iPos);
      }
    }
  }

  docListDestroy(in);
  *in = out;
}








|
|










<

>
>


>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
|
|
|
|
<







455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473

474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509

510
511
512
513
514
515
516
    }
  }
  printf("\n");
  fflush(stdout);
}
#endif /* SQLITE_DEBUG */

/* Trim the given doclist to contain only positions in column
 * [iRestrictColumn]. */
static void docListRestrictColumn(DocList *in, int iRestrictColumn){
  DocListReader r;
  DocList out;

  assert( in->iType>=DL_POSITIONS );
  readerInit(&r, in);
  docListInit(&out, DL_POSITIONS, NULL, 0);

  while( !atEnd(&r) ){
    sqlite_int64 iDocid = readDocid(&r);

    int iPos, iColumn;

    docListAddDocid(&out, iDocid);
    while( (iPos = readPosition(&r, &iColumn)) != -1 ){
      if( iColumn==iRestrictColumn ){
        docListAddPos(&out, iColumn, iPos);
      }
    }
  }

  docListDestroy(in);
  *in = out;
}

/* Trim the given doclist by discarding any docids without any remaining
 * positions. */
static void docListDiscardEmpty(DocList *in) {
  DocListReader r;
  DocList out;

  /* TODO: It would be nice to implement this operation in place; that
   * could save a significant amount of memory in queries with long doclists. */
  assert( in->iType>=DL_POSITIONS );
  readerInit(&r, in);
  docListInit(&out, DL_POSITIONS, NULL, 0);

  while( !atEnd(&r) ){
    sqlite_int64 iDocid = readDocid(&r);
    int match = 0;
    int iPos, iColumn;
    while( (iPos = readPosition(&r, &iColumn)) != -1 ){
      if( !match ){
        docListAddDocid(&out, iDocid);
        match = 1;
      }
      docListAddPos(&out, iColumn, iPos);

    }
  }

  docListDestroy(in);
  *in = out;
}

567
568
569
570
571
572
573
574
575



576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
    sqlite_int64 iDocid = readDocid(&updateReader);
    skipPositionList(&updateReader);
    docListSpliceElement(&accReader, iDocid, pSource, updateReader.p-pSource);
  }
}

/*
** Read the next non-deleted docid off of pIn.  Return
** 0 if we reach the end of pDoclist.



*/
static sqlite_int64 nextValidDocid(DocListReader *pIn){
  sqlite_int64 docid = 0;
  skipPositionList(pIn);
  while( !atEnd(pIn) && (docid = readDocid(pIn))==0 ){
    skipPositionList(pIn);
  }
  return docid;
}

/*
** pLeft and pRight are two DocListReaders that are pointing to
** positions lists of the same document: iDocid. 
**
** If there are no instances in pLeft or pRight where the position







<
|
>
>
>

|
<

<
<
<
|







593
594
595
596
597
598
599

600
601
602
603
604
605

606



607
608
609
610
611
612
613
614
    sqlite_int64 iDocid = readDocid(&updateReader);
    skipPositionList(&updateReader);
    docListSpliceElement(&accReader, iDocid, pSource, updateReader.p-pSource);
  }
}

/*

** Read the next docid off of pIn.  Return 0 if we reach the end.
*
* TODO: This assumes that docids are never 0, but they may actually be 0 since
* users can choose docids when inserting into a full-text table.  Fix this.
*/
static sqlite_int64 nextDocid(DocListReader *pIn){

  skipPositionList(pIn);



  return atEnd(pIn) ? 0 : readDocid(pIn);
}

/*
** pLeft and pRight are two DocListReaders that are pointing to
** positions lists of the same document: iDocid. 
**
** If there are no instances in pLeft or pRight where the position
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
  DocList *pOut      /* Write the combined doclist here */
){
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight;

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextValidDocid(&left);
  docidRight = nextValidDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    if( docidLeft<docidRight ){
      docidLeft = nextValidDocid(&left);
    }else if( docidRight<docidLeft ){
      docidRight = nextValidDocid(&right);
    }else{
      mergePosList(&left, &right, docidLeft, pOut);
      docidLeft = nextValidDocid(&left);
      docidRight = nextValidDocid(&right);
    }
  }
}

/* We have two doclists:  pLeft and pRight.
** Write the intersection of these two doclists into pOut.
** Only docids are matched.  Position information is ignored.







|
|



|

|


|
|







671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
  DocList *pOut      /* Write the combined doclist here */
){
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight;

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextDocid(&left);
  docidRight = nextDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    if( docidLeft<docidRight ){
      docidLeft = nextDocid(&left);
    }else if( docidRight<docidLeft ){
      docidRight = nextDocid(&right);
    }else{
      mergePosList(&left, &right, docidLeft, pOut);
      docidLeft = nextDocid(&left);
      docidRight = nextDocid(&right);
    }
  }
}

/* We have two doclists:  pLeft and pRight.
** Write the intersection of these two doclists into pOut.
** Only docids are matched.  Position information is ignored.
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight;

  assert( pOut->iType<DL_POSITIONS );

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextValidDocid(&left);
  docidRight = nextValidDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    if( docidLeft<docidRight ){
      docidLeft = nextValidDocid(&left);
    }else if( docidRight<docidLeft ){
      docidRight = nextValidDocid(&right);
    }else{
      docListAddDocid(pOut, docidLeft);
      docidLeft = nextValidDocid(&left);
      docidRight = nextValidDocid(&right);
    }
  }
}

/* We have two doclists:  pLeft and pRight.
** Write the union of these two doclists into pOut.
** Only docids are matched.  Position information is ignored.
**
** The output pOut never holds positions.
*/
static void docListOrMerge(
  DocList *pLeft,    /* Doclist resulting from the words on the left */
  DocList *pRight,   /* Doclist for the next word to the right */
  DocList *pOut      /* Write the combined doclist here */
){
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight, priorLeft;

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextValidDocid(&left);
  docidRight = nextValidDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    if( docidLeft<=docidRight ){
      docListAddDocid(pOut, docidLeft);
    }else{
      docListAddDocid(pOut, docidRight);
    }
    priorLeft = docidLeft;
    if( docidLeft<=docidRight ){
      docidLeft = nextValidDocid(&left);
    }
    if( docidRight>0 && docidRight<=priorLeft ){
      docidRight = nextValidDocid(&right);
    }
  }
  while( docidLeft>0 ){
    docListAddDocid(pOut, docidLeft);
    docidLeft = nextValidDocid(&left);
  }
  while( docidRight>0 ){
    docListAddDocid(pOut, docidRight);
    docidRight = nextValidDocid(&right);
  }
}

/* We have two doclists:  pLeft and pRight.
** Write into pOut all documents that occur in pLeft but not
** in pRight.
**







|
|



|

|


|
|




















|
|









|


|




|



|







705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight;

  assert( pOut->iType<DL_POSITIONS );

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextDocid(&left);
  docidRight = nextDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    if( docidLeft<docidRight ){
      docidLeft = nextDocid(&left);
    }else if( docidRight<docidLeft ){
      docidRight = nextDocid(&right);
    }else{
      docListAddDocid(pOut, docidLeft);
      docidLeft = nextDocid(&left);
      docidRight = nextDocid(&right);
    }
  }
}

/* We have two doclists:  pLeft and pRight.
** Write the union of these two doclists into pOut.
** Only docids are matched.  Position information is ignored.
**
** The output pOut never holds positions.
*/
static void docListOrMerge(
  DocList *pLeft,    /* Doclist resulting from the words on the left */
  DocList *pRight,   /* Doclist for the next word to the right */
  DocList *pOut      /* Write the combined doclist here */
){
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight, priorLeft;

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextDocid(&left);
  docidRight = nextDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    if( docidLeft<=docidRight ){
      docListAddDocid(pOut, docidLeft);
    }else{
      docListAddDocid(pOut, docidRight);
    }
    priorLeft = docidLeft;
    if( docidLeft<=docidRight ){
      docidLeft = nextDocid(&left);
    }
    if( docidRight>0 && docidRight<=priorLeft ){
      docidRight = nextDocid(&right);
    }
  }
  while( docidLeft>0 ){
    docListAddDocid(pOut, docidLeft);
    docidLeft = nextDocid(&left);
  }
  while( docidRight>0 ){
    docListAddDocid(pOut, docidRight);
    docidRight = nextDocid(&right);
  }
}

/* We have two doclists:  pLeft and pRight.
** Write into pOut all documents that occur in pLeft but not
** in pRight.
**
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
  DocList *pOut      /* Write the combined doclist here */
){
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight, priorLeft;

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextValidDocid(&left);
  docidRight = nextValidDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    priorLeft = docidLeft;
    if( docidLeft<docidRight ){
      docListAddDocid(pOut, docidLeft);
    }
    if( docidLeft<=docidRight ){
      docidLeft = nextValidDocid(&left);
    }
    if( docidRight>0 && docidRight<=priorLeft ){
      docidRight = nextValidDocid(&right);
    }
  }
  while( docidLeft>0 ){
    docListAddDocid(pOut, docidLeft);
    docidLeft = nextValidDocid(&left);
  }
}

static char *string_dup_n(const char *s, int n){
  char *str = malloc(n + 1);
  memcpy(str, s, n);
  str[n] = '\0';







|
|







|


|




|







782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
  DocList *pOut      /* Write the combined doclist here */
){
  DocListReader left, right;
  sqlite_int64 docidLeft, docidRight, priorLeft;

  readerInit(&left, pLeft);
  readerInit(&right, pRight);
  docidLeft = nextDocid(&left);
  docidRight = nextDocid(&right);

  while( docidLeft>0 && docidRight>0 ){
    priorLeft = docidLeft;
    if( docidLeft<docidRight ){
      docListAddDocid(pOut, docidLeft);
    }
    if( docidLeft<=docidRight ){
      docidLeft = nextDocid(&left);
    }
    if( docidRight>0 && docidRight<=priorLeft ){
      docidRight = nextDocid(&right);
    }
  }
  while( docidLeft>0 ){
    docListAddDocid(pOut, docidLeft);
    docidLeft = nextDocid(&left);
  }
}

static char *string_dup_n(const char *s, int n){
  char *str = malloc(n + 1);
  memcpy(str, s, n);
  str[n] = '\0';
1279
1280
1281
1282
1283
1284
1285

1286
1287
1288
1289
1290
1291
1292
    doclist = old;
  }
  if( rc!=SQLITE_DONE ){
    docListDestroy(&doclist);
    return rc;
  }


  *out = doclist;
  return SQLITE_OK;
}

/* insert into %_term (rowid, term, segment, doclist)
               values ([piRowid], [pTerm], [iSegment], [doclist])
** Lets sqlite select rowid if piRowid is NULL, else uses *piRowid.







>







1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
    doclist = old;
  }
  if( rc!=SQLITE_DONE ){
    docListDestroy(&doclist);
    return rc;
  }

  docListDiscardEmpty(&doclist);
  *out = doclist;
  return SQLITE_OK;
}

/* insert into %_term (rowid, term, segment, doclist)
               values ([piRowid], [pTerm], [iSegment], [doclist])
** Lets sqlite select rowid if piRowid is NULL, else uses *piRowid.
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
        c->eof = 1;
        return rc;
    }
  } else {  /* full-text query */
    rc = sqlite3_reset(c->pStmt);
    if( rc!=SQLITE_OK ) return rc;

    iDocid = nextValidDocid(&c->result);
    if( iDocid==0 ){
      c->eof = 1;
      return SQLITE_OK;
    }
    rc = sqlite3_bind_int64(c->pStmt, 1, iDocid);
    if( rc!=SQLITE_OK ) return rc;
    /* TODO(shess) Handle SQLITE_SCHEMA AND SQLITE_BUSY. */







|







2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
        c->eof = 1;
        return rc;
    }
  } else {  /* full-text query */
    rc = sqlite3_reset(c->pStmt);
    if( rc!=SQLITE_OK ) return rc;

    iDocid = nextDocid(&c->result);
    if( iDocid==0 ){
      c->eof = 1;
      return SQLITE_OK;
    }
    rc = sqlite3_bind_int64(c->pStmt, 1, iDocid);
    if( rc!=SQLITE_OK ) return rc;
    /* TODO(shess) Handle SQLITE_SCHEMA AND SQLITE_BUSY. */