/ Check-in [7f2bd4ff]
Login

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

Overview
Comment:Improved resolution of large integer values in "CAST(x AS NUMERIC)".
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA3-256: 7f2bd4ff45fba29528c18cac6da983bd9b164303525d3965056f5b40f85dc83f
User & Date: drh 2017-09-20 10:47:10
Context
2017-09-20
17:49
Small size and performance optimization in the bytecode engine. check-in: 3b3e32d4 user: drh tags: trunk
10:47
Improved resolution of large integer values in "CAST(x AS NUMERIC)". check-in: 7f2bd4ff user: drh tags: trunk
09:09
Updates to the "lemon.html" document received from Andy Goth. check-in: 5b2002f3 user: drh tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to src/expr.c.

3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
    sqlite3VdbeAddOp2(v, OP_Integer, i, iMem);
  }else{
    int c;
    i64 value;
    const char *z = pExpr->u.zToken;
    assert( z!=0 );
    c = sqlite3DecOrHexToI64(z, &value);
    if( c==1 || (c==2 && !negFlag) || (negFlag && value==SMALLEST_INT64)){
#ifdef SQLITE_OMIT_FLOATING_POINT
      sqlite3ErrorMsg(pParse, "oversized integer: %s%s", negFlag ? "-" : "", z);
#else
#ifndef SQLITE_OMIT_HEX_INTEGER
      if( sqlite3_strnicmp(z,"0x",2)==0 ){
        sqlite3ErrorMsg(pParse, "hex literal too big: %s%s", negFlag?"-":"",z);
      }else
#endif
      {
        codeReal(v, z, negFlag, iMem);
      }
#endif
    }else{
      if( negFlag ){ value = c==2 ? SMALLEST_INT64 : -value; }
      sqlite3VdbeAddOp4Dup8(v, OP_Int64, 0, iMem, 0, (u8*)&value, P4_INT64);
    }
  }
}

/*
** Erase column-cache entry number i







|













|







3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
    sqlite3VdbeAddOp2(v, OP_Integer, i, iMem);
  }else{
    int c;
    i64 value;
    const char *z = pExpr->u.zToken;
    assert( z!=0 );
    c = sqlite3DecOrHexToI64(z, &value);
    if( (c==3 && !negFlag) || (c==2) || (negFlag && value==SMALLEST_INT64)){
#ifdef SQLITE_OMIT_FLOATING_POINT
      sqlite3ErrorMsg(pParse, "oversized integer: %s%s", negFlag ? "-" : "", z);
#else
#ifndef SQLITE_OMIT_HEX_INTEGER
      if( sqlite3_strnicmp(z,"0x",2)==0 ){
        sqlite3ErrorMsg(pParse, "hex literal too big: %s%s", negFlag?"-":"",z);
      }else
#endif
      {
        codeReal(v, z, negFlag, iMem);
      }
#endif
    }else{
      if( negFlag ){ value = c==3 ? SMALLEST_INT64 : -value; }
      sqlite3VdbeAddOp4Dup8(v, OP_Int64, 0, iMem, 0, (u8*)&value, P4_INT64);
    }
  }
}

/*
** Erase column-cache entry number i

Changes to src/main.c.

3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
sqlite3_int64 sqlite3_uri_int64(
  const char *zFilename,    /* Filename as passed to xOpen */
  const char *zParam,       /* URI parameter sought */
  sqlite3_int64 bDflt       /* return if parameter is missing */
){
  const char *z = sqlite3_uri_parameter(zFilename, zParam);
  sqlite3_int64 v;
  if( z && sqlite3DecOrHexToI64(z, &v)==SQLITE_OK ){
    bDflt = v;
  }
  return bDflt;
}

/*
** Return the Btree pointer identified by zDbName.  Return NULL if not found.







|







3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
sqlite3_int64 sqlite3_uri_int64(
  const char *zFilename,    /* Filename as passed to xOpen */
  const char *zParam,       /* URI parameter sought */
  sqlite3_int64 bDflt       /* return if parameter is missing */
){
  const char *z = sqlite3_uri_parameter(zFilename, zParam);
  sqlite3_int64 v;
  if( z && sqlite3DecOrHexToI64(z, &v)==0 ){
    bDflt = v;
  }
  return bDflt;
}

/*
** Return the Btree pointer identified by zDbName.  Return NULL if not found.

Changes to src/util.c.

553
554
555
556
557
558
559
560
561
562

563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581

582
583
584
585
586
587
588
...
614
615
616
617
618
619
620
621
622
623





624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
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
673
674
675
676
677
678
679
680
  return c;
}

/*
** Convert zNum to a 64-bit signed integer.  zNum must be decimal. This
** routine does *not* accept hexadecimal notation.
**
** If the zNum value is representable as a 64-bit twos-complement 
** integer, then write that value into *pNum and return 0.
**

** If zNum is exactly 9223372036854775808, return 2.  This special
** case is broken out because while 9223372036854775808 cannot be a 
** signed 64-bit integer, its negative -9223372036854775808 can be.
**
** If zNum is too big for a 64-bit integer and is not
** 9223372036854775808  or if zNum contains any non-numeric text,
** then return 1.
**
** length is the number of bytes in the string (bytes, not characters).
** The string is not necessarily zero-terminated.  The encoding is
** given by enc.
*/
int sqlite3Atoi64(const char *zNum, i64 *pNum, int length, u8 enc){
  int incr;
  u64 u = 0;
  int neg = 0; /* assume positive */
  int i;
  int c = 0;
  int nonNum = 0;  /* True if input contains UTF16 with high byte non-zero */

  const char *zStart;
  const char *zEnd = zNum + length;
  assert( enc==SQLITE_UTF8 || enc==SQLITE_UTF16LE || enc==SQLITE_UTF16BE );
  if( enc==SQLITE_UTF8 ){
    incr = 1;
  }else{
    incr = 2;
................................................................................
    *pNum = (i64)u;
  }
  testcase( i==18 );
  testcase( i==19 );
  testcase( i==20 );
  if( &zNum[i]<zEnd              /* Extra bytes at the end */
   || (i==0 && zStart==zNum)     /* No digits */
   || i>19*incr                  /* Too many digits */
   || nonNum                     /* UTF16 with high-order bytes non-zero */
  ){





    /* zNum is empty or contains non-numeric text or is longer
    ** than 19 digits (thus guaranteeing that it is too large) */
    return 1;
  }else if( i<19*incr ){
    /* Less than 19 digits, so we know that it fits in 64 bits */
    assert( u<=LARGEST_INT64 );
    return 0;
  }else{
    /* zNum is a 19-digit numbers.  Compare it against 9223372036854775808. */
    c = compare2pow63(zNum, incr);
    if( c<0 ){
      /* zNum is less than 9223372036854775808 so it fits */
      assert( u<=LARGEST_INT64 );
      return 0;
    }else if( c>0 ){
      /* zNum is greater than 9223372036854775808 so it overflows */
      return 1;
    }else{
      /* zNum is exactly 9223372036854775808.  Fits if negative.  The
      ** special case 2 overflow if positive */
      assert( u-1==LARGEST_INT64 );
      return neg ? 0 : 2;
    }
  }
}

/*
** Transform a UTF-8 integer literal, in either decimal or hexadecimal,
** into a 64-bit signed integer.  This routine accepts hexadecimal literals,
** whereas sqlite3Atoi64() does not.
**
** Returns:
**
**     0    Successful transformation.  Fits in a 64-bit signed integer.

**     1    Integer too large for a 64-bit signed integer or is malformed
**     2    Special case of 9223372036854775808
*/
int sqlite3DecOrHexToI64(const char *z, i64 *pOut){
#ifndef SQLITE_OMIT_HEX_INTEGER
  if( z[0]=='0'
   && (z[1]=='x' || z[1]=='X')
  ){
    u64 u = 0;
    int i, k;
    for(i=2; z[i]=='0'; i++){}
    for(k=i; sqlite3Isxdigit(z[k]); k++){
      u = u*16 + sqlite3HexToInt(z[k]);
    }
    memcpy(pOut, &u, 8);
    return (z[k]==0 && k-i<=16) ? 0 : 1;
  }else
#endif /* SQLITE_OMIT_HEX_INTEGER */
  {
    return sqlite3Atoi64(z, pOut, sqlite3Strlen30(z), SQLITE_UTF8);
  }
}








|
<

>
|
|
|
<
<
<
<












>







 







<


>
>
>
>
>


|



|






|


|




|












>
|
|













|







553
554
555
556
557
558
559
560

561
562
563
564
565




566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
...
611
612
613
614
615
616
617

618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
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
673
674
675
676
677
678
679
680
681
682
  return c;
}

/*
** Convert zNum to a 64-bit signed integer.  zNum must be decimal. This
** routine does *not* accept hexadecimal notation.
**
** Returns:

**
**     0    Successful transformation.  Fits in a 64-bit signed integer.
**     1    Excess text after the integer value
**     2    Integer too large for a 64-bit signed integer or is malformed
**     3    Special case of 9223372036854775808




**
** length is the number of bytes in the string (bytes, not characters).
** The string is not necessarily zero-terminated.  The encoding is
** given by enc.
*/
int sqlite3Atoi64(const char *zNum, i64 *pNum, int length, u8 enc){
  int incr;
  u64 u = 0;
  int neg = 0; /* assume positive */
  int i;
  int c = 0;
  int nonNum = 0;  /* True if input contains UTF16 with high byte non-zero */
  int rc;          /* Baseline return code */
  const char *zStart;
  const char *zEnd = zNum + length;
  assert( enc==SQLITE_UTF8 || enc==SQLITE_UTF16LE || enc==SQLITE_UTF16BE );
  if( enc==SQLITE_UTF8 ){
    incr = 1;
  }else{
    incr = 2;
................................................................................
    *pNum = (i64)u;
  }
  testcase( i==18 );
  testcase( i==19 );
  testcase( i==20 );
  if( &zNum[i]<zEnd              /* Extra bytes at the end */
   || (i==0 && zStart==zNum)     /* No digits */

   || nonNum                     /* UTF16 with high-order bytes non-zero */
  ){
    rc = 1;
  }else{
    rc = 0;
  }
  if( i>19*incr ){                /* Too many digits */
    /* zNum is empty or contains non-numeric text or is longer
    ** than 19 digits (thus guaranteeing that it is too large) */
    return 2;
  }else if( i<19*incr ){
    /* Less than 19 digits, so we know that it fits in 64 bits */
    assert( u<=LARGEST_INT64 );
    return rc;
  }else{
    /* zNum is a 19-digit numbers.  Compare it against 9223372036854775808. */
    c = compare2pow63(zNum, incr);
    if( c<0 ){
      /* zNum is less than 9223372036854775808 so it fits */
      assert( u<=LARGEST_INT64 );
      return rc;
    }else if( c>0 ){
      /* zNum is greater than 9223372036854775808 so it overflows */
      return 2;
    }else{
      /* zNum is exactly 9223372036854775808.  Fits if negative.  The
      ** special case 2 overflow if positive */
      assert( u-1==LARGEST_INT64 );
      return neg ? rc : 3;
    }
  }
}

/*
** Transform a UTF-8 integer literal, in either decimal or hexadecimal,
** into a 64-bit signed integer.  This routine accepts hexadecimal literals,
** whereas sqlite3Atoi64() does not.
**
** Returns:
**
**     0    Successful transformation.  Fits in a 64-bit signed integer.
**     1    Excess text after the integer value
**     2    Integer too large for a 64-bit signed integer or is malformed
**     3    Special case of 9223372036854775808
*/
int sqlite3DecOrHexToI64(const char *z, i64 *pOut){
#ifndef SQLITE_OMIT_HEX_INTEGER
  if( z[0]=='0'
   && (z[1]=='x' || z[1]=='X')
  ){
    u64 u = 0;
    int i, k;
    for(i=2; z[i]=='0'; i++){}
    for(k=i; sqlite3Isxdigit(z[k]); k++){
      u = u*16 + sqlite3HexToInt(z[k]);
    }
    memcpy(pOut, &u, 8);
    return (z[k]==0 && k-i<=16) ? 0 : 2;
  }else
#endif /* SQLITE_OMIT_HEX_INTEGER */
  {
    return sqlite3Atoi64(z, pOut, sqlite3Strlen30(z), SQLITE_UTF8);
  }
}

Changes to src/vdbe.c.

350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
*/
static u16 SQLITE_NOINLINE computeNumericType(Mem *pMem){
  assert( (pMem->flags & (MEM_Int|MEM_Real))==0 );
  assert( (pMem->flags & (MEM_Str|MEM_Blob))!=0 );
  if( sqlite3AtoF(pMem->z, &pMem->u.r, pMem->n, pMem->enc)==0 ){
    return 0;
  }
  if( sqlite3Atoi64(pMem->z, &pMem->u.i, pMem->n, pMem->enc)==SQLITE_OK ){
    return MEM_Int;
  }
  return MEM_Real;
}

/*
** Return the numeric type for pMem, either MEM_Int or MEM_Real or both or







|







350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
*/
static u16 SQLITE_NOINLINE computeNumericType(Mem *pMem){
  assert( (pMem->flags & (MEM_Int|MEM_Real))==0 );
  assert( (pMem->flags & (MEM_Str|MEM_Blob))!=0 );
  if( sqlite3AtoF(pMem->z, &pMem->u.r, pMem->n, pMem->enc)==0 ){
    return 0;
  }
  if( sqlite3Atoi64(pMem->z, &pMem->u.i, pMem->n, pMem->enc)==0 ){
    return MEM_Int;
  }
  return MEM_Real;
}

/*
** Return the numeric type for pMem, either MEM_Int or MEM_Real or both or

Changes to src/vdbemem.c.

593
594
595
596
597
598
599

600
601
602

603
604
605






606
607

608
609
610
611
612
613
614
**
** Every effort is made to force the conversion, even if the input
** is a string that does not look completely like a number.  Convert
** as much of the string as we can and ignore the rest.
*/
int sqlite3VdbeMemNumerify(Mem *pMem){
  if( (pMem->flags & (MEM_Int|MEM_Real|MEM_Null))==0 ){

    assert( (pMem->flags & (MEM_Blob|MEM_Str))!=0 );
    assert( pMem->db==0 || sqlite3_mutex_held(pMem->db->mutex) );
    if( 0==sqlite3Atoi64(pMem->z, &pMem->u.i, pMem->n, pMem->enc) ){

      MemSetTypeFlag(pMem, MEM_Int);
    }else{
      pMem->u.r = sqlite3VdbeRealValue(pMem);






      MemSetTypeFlag(pMem, MEM_Real);
      sqlite3VdbeIntegerAffinity(pMem);

    }
  }
  assert( (pMem->flags & (MEM_Int|MEM_Real|MEM_Null))!=0 );
  pMem->flags &= ~(MEM_Str|MEM_Blob|MEM_Zero);
  return SQLITE_OK;
}








>


|
>


<
>
>
>
>
>
>
|
<
>







593
594
595
596
597
598
599
600
601
602
603
604
605
606

607
608
609
610
611
612
613

614
615
616
617
618
619
620
621
**
** Every effort is made to force the conversion, even if the input
** is a string that does not look completely like a number.  Convert
** as much of the string as we can and ignore the rest.
*/
int sqlite3VdbeMemNumerify(Mem *pMem){
  if( (pMem->flags & (MEM_Int|MEM_Real|MEM_Null))==0 ){
    int rc;
    assert( (pMem->flags & (MEM_Blob|MEM_Str))!=0 );
    assert( pMem->db==0 || sqlite3_mutex_held(pMem->db->mutex) );
    rc = sqlite3Atoi64(pMem->z, &pMem->u.i, pMem->n, pMem->enc);
    if( rc==0 ){
      MemSetTypeFlag(pMem, MEM_Int);
    }else{

      i64 i = pMem->u.i;
      sqlite3AtoF(pMem->z, &pMem->u.r, pMem->n, pMem->enc);
      if( rc==1 && pMem->u.r==(double)i ){
        pMem->u.i = i;
        MemSetTypeFlag(pMem, MEM_Int);
      }else{
        MemSetTypeFlag(pMem, MEM_Real);

      }
    }
  }
  assert( (pMem->flags & (MEM_Int|MEM_Real|MEM_Null))!=0 );
  pMem->flags &= ~(MEM_Str|MEM_Blob|MEM_Zero);
  return SQLITE_OK;
}

Changes to test/e_expr.test.

1659
1660
1661
1662
1663
1664
1665












































1666
1667
1668
1669
1670
1671
1672

do_expr_test e_expr-32.2.3 { 
  CAST(-9223372036854775808 AS NUMERIC)
} integer -9223372036854775808
do_expr_test e_expr-32.2.4 { 
  CAST(9223372036854775807 AS NUMERIC)
} integer 9223372036854775807













































# EVIDENCE-OF: R-64550-29191 Note that the result from casting any
# non-BLOB value into a BLOB and the result from casting any BLOB value
# into a non-BLOB value may be different depending on whether the
# database encoding is UTF-8, UTF-16be, or UTF-16le.
#
ifcapable {utf16} {







>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>







1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716

do_expr_test e_expr-32.2.3 { 
  CAST(-9223372036854775808 AS NUMERIC)
} integer -9223372036854775808
do_expr_test e_expr-32.2.4 { 
  CAST(9223372036854775807 AS NUMERIC)
} integer 9223372036854775807
do_expr_test e_expr-32.2.5 { 
  CAST('9223372036854775807 ' AS NUMERIC)
} integer 9223372036854775807
do_expr_test e_expr-32.2.6 { 
  CAST('   9223372036854775807   ' AS NUMERIC)
} integer 9223372036854775807
do_expr_test e_expr-32.2.7 { 
  CAST('  ' AS NUMERIC)
} integer 0
do_execsql_test e_expr-32.2.8 {
  WITH t1(x) AS (VALUES
     ('9000000000000000001'),
     ('9000000000000000001x'),
     ('9000000000000000001 '),
     (' 9000000000000000001 '),
     (' 9000000000000000001'),
     (' 9000000000000000001.'),
     ('9223372036854775807'),
     ('9223372036854775807 '),
     ('   9223372036854775807   '),
     ('9223372036854775808'),
     ('   9223372036854775808   '),
     ('9223372036854775807.0'),
     ('9223372036854775807e+0'),
     ('-5.0'),
     ('-5e+0'))
  SELECT typeof(CAST(x AS NUMERIC)), CAST(x AS NUMERIC)||'' FROM t1;
} [list \
 integer 9000000000000000001 \
 integer 9000000000000000001 \
 integer 9000000000000000001 \
 integer 9000000000000000001 \
 integer 9000000000000000001 \
 integer 9000000000000000001 \
 integer 9223372036854775807 \
 integer 9223372036854775807 \
 integer 9223372036854775807 \
 real 9.22337203685478e+18 \
 real 9.22337203685478e+18 \
 integer 9223372036854775807 \
 integer 9223372036854775807 \
 integer -5 \
 integer -5 \
]

# EVIDENCE-OF: R-64550-29191 Note that the result from casting any
# non-BLOB value into a BLOB and the result from casting any BLOB value
# into a non-BLOB value may be different depending on whether the
# database encoding is UTF-8, UTF-16be, or UTF-16le.
#
ifcapable {utf16} {