/ Check-in [86fa0087]
Login

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

Overview
Comment:Fix the implementation of the built-in RTRIM collating sequence so that it works for control characters at the end of the string. Ticket [f1580ba1b574e9e9]
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA3-256: 86fa0087cd1f5c79ed51b99a226ec2eef5b0fa0a23981d469bec0e02fadf2a68
User & Date: drh 2019-06-14 13:24:46
References
2019-06-14
13:25 Fixed ticket [f1580ba1]: Built-in RTRIM collating sequence yields incorrect comparisons plus 8 other changes artifact: 71c96e66 user: drh
Context
2019-06-14
17:37
Deprecate the SQLITE_ALLOW_COVERING_INDEX_SCAN compile-time option when it is turned off. This probably will impact no-one. If it does, they will get an error message instructing them to contact developers, so that we can get feedback that somebody is actually using this feature. Start-time disabling of covering index scan is not affected. check-in: 23e49f99 user: drh tags: trunk
13:24
Fix the implementation of the built-in RTRIM collating sequence so that it works for control characters at the end of the string. Ticket [f1580ba1b574e9e9] check-in: 86fa0087 user: drh tags: trunk
12:28
Refactor the LIKE optimization decision logic so that it uses sqlite3AtoF() on both boundary keys to determine if the optimization can be used when the LHS is something that might not have TEXT affinity. Ticket [ce8717f0885af975]. See also [c94369cae9b561b1], [b043a54c3de54b28], [fd76310a5e843e07], and [158290c0abafde67]. check-in: b4a9e09e user: drh tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to src/main.c.

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
912
913
914
915
916
917
918
919
920
921
















922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
....
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
      break;
    }
  }
  va_end(ap);
  return rc;
}


/*
** Return true if the buffer z[0..n-1] contains all spaces.
*/
static int allSpaces(const char *z, int n){
  while( n>0 && z[n-1]==' ' ){ n--; }
  return n==0;
}

/*
** This is the default collating function named "BINARY" which is always
** available.
**
** If the padFlag argument is not NULL then space padding at the end
** of strings is ignored.  This implements the RTRIM collation.
*/
static int binCollFunc(
  void *padFlag,
  int nKey1, const void *pKey1,
  int nKey2, const void *pKey2
){
  int rc, n;

  n = nKey1<nKey2 ? nKey1 : nKey2;
  /* EVIDENCE-OF: R-65033-28449 The built-in BINARY collation compares
  ** strings byte by byte using the memcmp() function from the standard C
  ** library. */
  assert( pKey1 && pKey2 );
  rc = memcmp(pKey1, pKey2, n);
  if( rc==0 ){
    if( padFlag
     && allSpaces(((char*)pKey1)+n, nKey1-n)
     && allSpaces(((char*)pKey2)+n, nKey2-n)
    ){
      /* EVIDENCE-OF: R-31624-24737 RTRIM is like BINARY except that extra
      ** spaces at the end of either string do not change the result. In other
      ** words, strings will compare equal to one another as long as they
      ** differ only in the number of spaces at the end.
      */
    }else{
      rc = nKey1 - nKey2;
    }
  }
  return rc;
}

















/*
** Return true if CollSeq is the default built-in BINARY.
*/
int sqlite3IsBinary(const CollSeq *p){
  assert( p==0 || p->xCmp!=binCollFunc || p->pUser!=0
            || strcmp(p->zName,"BINARY")==0 );
  return p==0 || (p->xCmp==binCollFunc && p->pUser==0);
}

/*
** Another built-in collating sequence: NOCASE. 
**
** This collating sequence is intended to be used for "case independent
** comparison". SQLite's knowledge of upper and lower case equivalents
................................................................................
  ** EVIDENCE-OF: R-52786-44878 SQLite defines three built-in collating
  ** functions:
  */
  createCollation(db, sqlite3StrBINARY, SQLITE_UTF8, 0, binCollFunc, 0);
  createCollation(db, sqlite3StrBINARY, SQLITE_UTF16BE, 0, binCollFunc, 0);
  createCollation(db, sqlite3StrBINARY, SQLITE_UTF16LE, 0, binCollFunc, 0);
  createCollation(db, "NOCASE", SQLITE_UTF8, 0, nocaseCollatingFunc, 0);
  createCollation(db, "RTRIM", SQLITE_UTF8, (void*)1, binCollFunc, 0);
  if( db->mallocFailed ){
    goto opendb_out;
  }
  /* EVIDENCE-OF: R-08308-17224 The default collating function for all
  ** strings is BINARY. 
  */
  db->pDfltColl = sqlite3FindCollSeq(db, SQLITE_UTF8, sqlite3StrBINARY, 0);







<
<
<
<
<
<
<
<
<



<
<
<


|




>







<
<
<
<
<
<
<
<
<
<
|
|
<


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





|
<
|







 







|







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
912
913
914
915
916
917
918
919
920
921

922
923
924
925
926
927
928
929
....
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
      break;
    }
  }
  va_end(ap);
  return rc;
}










/*
** This is the default collating function named "BINARY" which is always
** available.



*/
static int binCollFunc(
  void *NotUsed,
  int nKey1, const void *pKey1,
  int nKey2, const void *pKey2
){
  int rc, n;
  UNUSED_PARAMETER(NotUsed);
  n = nKey1<nKey2 ? nKey1 : nKey2;
  /* EVIDENCE-OF: R-65033-28449 The built-in BINARY collation compares
  ** strings byte by byte using the memcmp() function from the standard C
  ** library. */
  assert( pKey1 && pKey2 );
  rc = memcmp(pKey1, pKey2, n);
  if( rc==0 ){










    rc = nKey1 - nKey2;
  }

  return rc;
}

/*
** This is the collating function named "RTRIM" which is always
** available.  Ignore trailing spaces.
*/
static int rtrimCollFunc(
  void *pUser,
  int nKey1, const void *pKey1,
  int nKey2, const void *pKey2
){
  const u8 *pK1 = (const u8*)pKey1;
  const u8 *pK2 = (const u8*)pKey2;
  while( nKey1 && pK1[nKey1-1]==' ' ) nKey1--;
  while( nKey2 && pK2[nKey2-1]==' ' ) nKey2--;
  return binCollFunc(pUser, nKey1, pKey1, nKey2, pKey2);
}

/*
** Return true if CollSeq is the default built-in BINARY.
*/
int sqlite3IsBinary(const CollSeq *p){
  assert( p==0 || p->xCmp!=binCollFunc || strcmp(p->zName,"BINARY")==0 );

  return p==0 || p->xCmp==binCollFunc;
}

/*
** Another built-in collating sequence: NOCASE. 
**
** This collating sequence is intended to be used for "case independent
** comparison". SQLite's knowledge of upper and lower case equivalents
................................................................................
  ** EVIDENCE-OF: R-52786-44878 SQLite defines three built-in collating
  ** functions:
  */
  createCollation(db, sqlite3StrBINARY, SQLITE_UTF8, 0, binCollFunc, 0);
  createCollation(db, sqlite3StrBINARY, SQLITE_UTF16BE, 0, binCollFunc, 0);
  createCollation(db, sqlite3StrBINARY, SQLITE_UTF16LE, 0, binCollFunc, 0);
  createCollation(db, "NOCASE", SQLITE_UTF8, 0, nocaseCollatingFunc, 0);
  createCollation(db, "RTRIM", SQLITE_UTF8, 0, rtrimCollFunc, 0);
  if( db->mallocFailed ){
    goto opendb_out;
  }
  /* EVIDENCE-OF: R-08308-17224 The default collating function for all
  ** strings is BINARY. 
  */
  db->pDfltColl = sqlite3FindCollSeq(db, SQLITE_UTF8, sqlite3StrBINARY, 0);

Changes to test/collate1.test.

396
397
398
399
400
401
402


403














404
    ORDER BY 1 COLLATE nocase COLLATE nocase COLLATE nocase COLLATE binary;
} {DEF abc}
do_execsql_test 7.2 {
   SELECT 'abc' UNION ALL SELECT 'DEF'
    ORDER BY 1 COLLATE binary COLLATE binary COLLATE binary COLLATE nocase;
} {abc DEF}


















finish_test







>
>
|
>
>
>
>
>
>
>
>
>
>
>
>
>
>

396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
    ORDER BY 1 COLLATE nocase COLLATE nocase COLLATE nocase COLLATE binary;
} {DEF abc}
do_execsql_test 7.2 {
   SELECT 'abc' UNION ALL SELECT 'DEF'
    ORDER BY 1 COLLATE binary COLLATE binary COLLATE binary COLLATE nocase;
} {abc DEF}

# 2019-06-14
# https://sqlite.org/src/info/f1580ba1b574e9e9
#
do_execsql_test 8.0 {
  SELECT ' ' > char(20) COLLATE rtrim;
} 0
do_execsql_test 8.1 {
  SELECT '' < char(20) COLLATE rtrim;
} 1
do_execsql_test 8.2 {
  DROP TABLE IF EXISTS t0;
  CREATE TABLE t0(c0 COLLATE RTRIM, c1 BLOB UNIQUE,
                  PRIMARY KEY (c0, c1)) WITHOUT ROWID;
  INSERT INTO t0 VALUES (123, 3), (' ', 1), ('	', 2), ('', 4);
  SELECT * FROM t0 WHERE c1 = 1;
} {{ } 1}

finish_test