/ Check-in [df44308b]
Login

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

Overview
Comment:Update spellfix1_scriptcode() to output 215 (latin) for pure numeric text.
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: df44308b862e932ff0ad25b6328d13be22047b96
User & Date: drh 2016-03-11 14:30:18
Context
2016-03-11
15:25
Tweak the amalgamation autoconf files so that the --disable-static-shell option works again. check-in: 83efcdeb user: dan tags: trunk
14:30
Update spellfix1_scriptcode() to output 215 (latin) for pure numeric text. check-in: df44308b user: drh tags: trunk
2016-03-10
19:08
Check for PDB files prior to attempting to copy them in the MSVC batch build tool. check-in: d7c20f04 user: mistachkin tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to ext/misc/spellfix.c.

1730
1731
1732
1733
1734
1735
1736

1737
1738
1739
1740
1741
1742
1743
1744
1745
1746

1747
1748



1749
1750
1751
1752
1753
1754
1755
1756
1757
1758

1759
1760
1761
1762
1763
1764
1765
  sqlite3_value **argv
){
  const unsigned char *zIn = sqlite3_value_text(argv[0]);
  int nIn = sqlite3_value_bytes(argv[0]);
  int c, sz;
  int scriptMask = 0;
  int res;

# define SCRIPT_LATIN       0x0001
# define SCRIPT_CYRILLIC    0x0002
# define SCRIPT_GREEK       0x0004
# define SCRIPT_HEBREW      0x0008
# define SCRIPT_ARABIC      0x0010

  while( nIn>0 ){
    c = utf8Read(zIn, nIn, &sz);
    zIn += sz;
    nIn -= sz;

    if( c<0x02af && (c>=0x80 || midClass[c&0x7f]<CCLASS_DIGIT) ){
      scriptMask |= SCRIPT_LATIN;



    }else if( c>=0x0400 && c<=0x04ff ){
      scriptMask |= SCRIPT_CYRILLIC;
    }else if( c>=0x0386 && c<=0x03ce ){
      scriptMask |= SCRIPT_GREEK;
    }else if( c>=0x0590 && c<=0x05ff ){
      scriptMask |= SCRIPT_HEBREW;
    }else if( c>=0x0600 && c<=0x06ff ){
      scriptMask |= SCRIPT_ARABIC;
    }
  }

  switch( scriptMask ){
    case 0:                res = 999; break;
    case SCRIPT_LATIN:     res = 215; break;
    case SCRIPT_CYRILLIC:  res = 220; break;
    case SCRIPT_GREEK:     res = 200; break;
    case SCRIPT_HEBREW:    res = 125; break;
    case SCRIPT_ARABIC:    res = 160; break;







>










>
|
|
>
>
>










>







1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
  sqlite3_value **argv
){
  const unsigned char *zIn = sqlite3_value_text(argv[0]);
  int nIn = sqlite3_value_bytes(argv[0]);
  int c, sz;
  int scriptMask = 0;
  int res;
  int seenDigit = 0;
# define SCRIPT_LATIN       0x0001
# define SCRIPT_CYRILLIC    0x0002
# define SCRIPT_GREEK       0x0004
# define SCRIPT_HEBREW      0x0008
# define SCRIPT_ARABIC      0x0010

  while( nIn>0 ){
    c = utf8Read(zIn, nIn, &sz);
    zIn += sz;
    nIn -= sz;
    if( c<0x02af ){
      if( c>=0x80 || midClass[c&0x7f]<CCLASS_DIGIT ){
        scriptMask |= SCRIPT_LATIN;
      }else if( c>='0' && c<='9' ){
        seenDigit = 1;
      }
    }else if( c>=0x0400 && c<=0x04ff ){
      scriptMask |= SCRIPT_CYRILLIC;
    }else if( c>=0x0386 && c<=0x03ce ){
      scriptMask |= SCRIPT_GREEK;
    }else if( c>=0x0590 && c<=0x05ff ){
      scriptMask |= SCRIPT_HEBREW;
    }else if( c>=0x0600 && c<=0x06ff ){
      scriptMask |= SCRIPT_ARABIC;
    }
  }
  if( scriptMask==0 && seenDigit ) scriptMask = SCRIPT_LATIN;
  switch( scriptMask ){
    case 0:                res = 999; break;
    case SCRIPT_LATIN:     res = 215; break;
    case SCRIPT_CYRILLIC:  res = 220; break;
    case SCRIPT_GREEK:     res = 200; break;
    case SCRIPT_HEBREW:    res = 125; break;
    case SCRIPT_ARABIC:    res = 160; break;

Changes to test/spellfix3.test.

31
32
33
34
35
36
37
38
39
40
41









42
43
  SELECT spellfix1_scriptcode('וַיֹּ֥אמֶר אֱלֹהִ֖ים יְהִ֣י א֑וֹר וַֽיְהִי־אֽוֹר׃');
} {125}
do_execsql_test 140 {
  SELECT spellfix1_scriptcode('فِي ذَلِكَ الوَقتِ، قالَ اللهُ: لِيَكُنْ نُورٌ. فَصَارَ نُورٌ.');
} {160}
do_execsql_test 200 {
  SELECT spellfix1_scriptcode('+3.14159');
} {999}
do_execsql_test 210 {
  SELECT spellfix1_scriptcode('And God said: "Да будет свет"');
} {998}










finish_test







|



>
>
>
>
>
>
>
>
>


31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
  SELECT spellfix1_scriptcode('וַיֹּ֥אמֶר אֱלֹהִ֖ים יְהִ֣י א֑וֹר וַֽיְהִי־אֽוֹר׃');
} {125}
do_execsql_test 140 {
  SELECT spellfix1_scriptcode('فِي ذَلِكَ الوَقتِ، قالَ اللهُ: لِيَكُنْ نُورٌ. فَصَارَ نُورٌ.');
} {160}
do_execsql_test 200 {
  SELECT spellfix1_scriptcode('+3.14159');
} {215}
do_execsql_test 210 {
  SELECT spellfix1_scriptcode('And God said: "Да будет свет"');
} {998}
do_execsql_test 220 {
  SELECT spellfix1_scriptcode('+3.14159 light');
} {215}
do_execsql_test 230 {
  SELECT spellfix1_scriptcode('+3.14159 свет');
} {220}
do_execsql_test 240 {
  SELECT spellfix1_scriptcode('וַיֹּ֥אמֶר +3.14159');
} {125}

finish_test