libs/unicode/fold.c

   1 /*
   2  * String folding
   3  *
   4  * Copyright 2003 Jon Griffiths
   5  *
   6  * This library is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * This library is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with this library; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  19  */
  20
  21 #include "wine/unicode.h"
  22
  23 static inline WCHAR to_unicode_digit( WCHAR ch )
  24 {
  25     extern const WCHAR wine_digitmap[];
  26     return ch + wine_digitmap[wine_digitmap[ch >> 8] + (ch & 0xff)];
  27 }
  28
  29 static inline WCHAR to_unicode_native( WCHAR ch )
  30 {
  31     extern const WCHAR wine_compatmap[];
  32     return ch + wine_compatmap[wine_compatmap[ch >> 8] + (ch & 0xff)];
  33 }
  34
  35 static const WCHAR wine_ligatures[] =
  36 {
  37     0x00c6, 0x00de, 0x00df, 0x00e6, 0x00fe, 0x0132, 0x0133, 0x0152,
  38     0x0153, 0x01c4, 0x01c5, 0x01c6, 0x01c7, 0x01c8, 0x01c9, 0x01ca,
  39     0x01cb, 0x01cc, 0x01e2, 0x01e3, 0x01f1, 0x01f2, 0x01f3, 0x01fc,
  40     0x01fd, 0x05f0, 0x05f1, 0x05f2, 0xfb00, 0xfb01, 0xfb02, 0xfb03,
  41     0xfb04, 0xfb05, 0xfb06
  42 };
  43
  44 /* Unicode expanded ligatures */
  45 static const WCHAR wine_expanded_ligatures[][4] =
  46 {
  47     { 'A','E','\0',1 },
  48     { 'T','H','\0',1 },
  49     { 's','s','\0',1 },
  50     { 'a','e','\0',1 },
  51     { 't','h','\0',1 },
  52     { 'I','J','\0',1 },
  53     { 'i','j','\0',1 },
  54     { 'O','E','\0',1 },
  55     { 'o','e','\0',1 },
  56     { 'D',0x017d,'\0',1 },
  57     { 'D',0x017e,'\0',1 },
  58     { 'd',0x017e,'\0',1 },
  59     { 'L','J','\0',1 },
  60     { 'L','j','\0',1 },
  61     { 'l','j','\0',1 },
  62     { 'N','J','\0',1 },
  63     { 'N','j','\0',1 },
  64     { 'n','j','\0',1 },
  65     { 0x0100,0x0112,'\0',1 },
  66     { 0x0101,0x0113,'\0',1 },
  67     { 'D','Z','\0',1 },
  68     { 'D','z','\0',1 },
  69     { 'd','z','\0',1 },
  70     { 0x00c1,0x00c9,'\0',1 },
  71     { 0x00e1,0x00e9,'\0',1 },
  72     { 0x05d5,0x05d5,'\0',1 },
  73     { 0x05d5,0x05d9,'\0',1 },
  74     { 0x05d9,0x05d9,'\0',1 },
  75     { 'f','f','\0',1 },
  76     { 'f','i','\0',1 },
  77     { 'f','l','\0',1 },
  78     { 'f','f','i',2 },
  79     { 'f','f','l',2 },
  80     { 0x017f,'t','\0',1 },
  81     { 's','t','\0',1 }
  82 };
  83
  84 static inline int get_ligature_len( WCHAR wc )
  85 {
  86     int low = 0, high = sizeof(wine_ligatures)/sizeof(WCHAR) -1;
  87     while (low <= high)
  88     {
  89         int pos = (low + high) / 2;
  90         if (wine_ligatures[pos] < wc)
  91             low = pos + 1;
  92         else if (wine_ligatures[pos] > wc)
  93             high = pos - 1;
  94         else
  95             return wine_expanded_ligatures[pos][3];
  96     }
  97     return 0;
  98 }
  99
 100 static inline const WCHAR* get_ligature( WCHAR wc )
 101 {
 102     static const WCHAR empty_ligature[] = { '\0','\0','\0', 0 };
 103     int low = 0, high = sizeof(wine_ligatures)/sizeof(WCHAR) -1;
 104     while (low <= high)
 105     {
 106         int pos = (low + high) / 2;
 107         if (wine_ligatures[pos] < wc)
 108             low = pos + 1;
 109         else if (wine_ligatures[pos] > wc)
 110             high = pos - 1;
 111         else
 112             return wine_expanded_ligatures[pos];
 113     }
 114     return empty_ligature;
 115 }
 116
 117 /* fold a unicode string */
 118 int wine_fold_string( int flags, const WCHAR *src, int srclen, WCHAR *dst, int dstlen )
 119 {
 120     WCHAR *dstbase = dst;
 121     const WCHAR *expand;
 122     int i;
 123
 124     if (srclen == -1)
 125         srclen = strlenW(src) + 1; /* Include terminating NUL in count */
 126
 127     if (!dstlen)
 128     {
 129         /* Calculate the required size for dst */
 130         dstlen = srclen;
 131
 132         if (flags & MAP_EXPAND_LIGATURES)
 133         {
 134             while (srclen--)
 135             {
 136                 dstlen += get_ligature_len(*src);
 137                 src++;
 138             }
 139         }
 140         else if (flags & MAP_COMPOSITE)
 141         {
 142             /* FIXME */
 143         }
 144         else if (flags & MAP_PRECOMPOSED)
 145         {
 146             /* FIXME */
 147         }
 148         return dstlen;
 149     }
 150
 151     if (srclen > dstlen)
 152         return 0;
 153
 154     dstlen -= srclen;
 155
 156     /* Actually perform the mapping(s) specified */
 157     for (i = 0; i < srclen; i++)
 158     {
 159         WCHAR ch = *src;
 160
 161         if (flags & MAP_EXPAND_LIGATURES)
 162         {
 163             expand = get_ligature(ch);
 164             if (expand[0])
 165             {
 166                 if (!dstlen--)
 167                     return 0;
 168                 dst[0] = expand[0];
 169                 if (expand[2])
 170                 {
 171                     if (!dstlen--)
 172                         return 0;
 173                     *++dst = expand[1];
 174                     ch = expand[2];
 175                 }
 176                 else
 177                     ch = expand[1];
 178                 dst++;
 179             }
 180         }
 181         else if (flags & MAP_COMPOSITE)
 182         {
 183             /* FIXME */
 184         }
 185         else if (flags & MAP_PRECOMPOSED)
 186         {
 187             /* FIXME */
 188         }
 189         if (flags & MAP_FOLDDIGITS)
 190             ch = to_unicode_digit(ch);
 191         if (flags & MAP_FOLDCZONE)
 192             ch = to_unicode_native(ch);
 193
 194         *dst++ = ch;
 195         src++;
 196     }
 197     return dst - dstbase;
 198 }