libs/unicode/sortkey.c

   1 /*
   2  * Unicode sort key generation
   3  *
   4  * Copyright 2003 Dmitry Timoshkov
   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 #include "wine/unicode.h"
  21
  22 extern int get_decomposition(WCHAR src, WCHAR *dst, unsigned int dstlen);
  23
  24 /*
  25  * flags - normalization NORM_* flags
  26  *
  27  * FIXME: 'variable' flag not handled
  28  */
  29 int wine_get_sortkey(int flags, const WCHAR *src, int srclen, char *dst, int dstlen)
  30 {
  31     extern const unsigned int collation_table[];
  32     WCHAR dummy[4]; /* no decomposition is larger than 4 chars */
  33     int key_len[4];
  34     char *key_ptr[4];
  35     const WCHAR *src_save = src;
  36     int srclen_save = srclen;
  37
  38     key_len[0] = key_len[1] = key_len[2] = key_len[3] = 0;
  39     for (; srclen; srclen--, src++)
  40     {
  41         int decomposed_len = get_decomposition(*src, dummy, 4);
  42         if (decomposed_len)
  43         {
  44             int i;
  45             for (i = 0; i < decomposed_len; i++)
  46             {
  47                 WCHAR wch = dummy[i];
  48                 unsigned int ce;
  49
  50                 /* tests show that win2k just ignores NORM_IGNORENONSPACE,
  51                  * and skips white space and punctuation characters for
  52                  * NORM_IGNORESYMBOLS.
  53                  */
  54                 if ((flags & NORM_IGNORESYMBOLS) && (get_char_typeW(wch) & (C1_PUNCT | C1_SPACE)))
  55                     continue;
  56
  57                 if (flags & NORM_IGNORECASE) wch = tolowerW(wch);
  58
  59                 ce = collation_table[collation_table[wch >> 8] + (wch & 0xff)];
  60                 if (ce != (unsigned int)-1)
  61                 {
  62                     if (ce >> 16) key_len[0] += 2;
  63                     if ((ce >> 8) & 0xff) key_len[1]++;
  64                     if ((ce >> 4) & 0x0f) key_len[2]++;
  65                     /*if (ce & 1)
  66                     {
  67                         if (wch >> 8) key_len[3]++;
  68                         key_len[3]++;
  69                     }*/
  70                 }
  71                 /*else
  72                 {
  73                     key_len[0] += 2;
  74                     if (wch >> 8) key_len[0]++;
  75                     if (wch & 0xff) key_len[0]++;
  76                 }*/
  77             }
  78         }
  79     }
  80
  81     if (!dstlen) /* compute length */
  82         /* 4 * '\1' + 1 * '\0' + key length */
  83         return key_len[0] + key_len[1] + key_len[2] + key_len[3] + 4 + 1;
  84
  85     if (dstlen < key_len[0] + key_len[1] + key_len[2] + key_len[3] + 4 + 1)
  86         return 0; /* overflow */
  87
  88     src = src_save;
  89     srclen = srclen_save;
  90
  91     key_ptr[0] = dst;
  92     key_ptr[1] = key_ptr[0] + key_len[0] + 1;
  93     key_ptr[2] = key_ptr[1] + key_len[1] + 1;
  94     key_ptr[3] = key_ptr[2] + key_len[2] + 1;
  95
  96     for (; srclen; srclen--, src++)
  97     {
  98         int decomposed_len = get_decomposition(*src, dummy, 4);
  99         if (decomposed_len)
 100         {
 101             int i;
 102             for (i = 0; i < decomposed_len; i++)
 103             {
 104                 WCHAR wch = dummy[i];
 105                 unsigned int ce;
 106
 107                 /* tests show that win2k just ignores NORM_IGNORENONSPACE,
 108                  * and skips white space and punctuation characters for
 109                  * NORM_IGNORESYMBOLS.
 110                  */
 111                 if ((flags & NORM_IGNORESYMBOLS) && (get_char_typeW(wch) & (C1_PUNCT | C1_SPACE)))
 112                     continue;
 113
 114                 if (flags & NORM_IGNORECASE) wch = tolowerW(wch);
 115
 116                 ce = collation_table[collation_table[wch >> 8] + (wch & 0xff)];
 117                 if (ce != (unsigned int)-1)
 118                 {
 119                     WCHAR key;
 120                     if ((key = ce >> 16))
 121                     {
 122                         *key_ptr[0]++ = key >> 8;
 123                         *key_ptr[0]++ = key & 0xff;
 124                     }
 125                     /* make key 1 start from 2 */
 126                     if ((key = (ce >> 8) & 0xff)) *key_ptr[1]++ = key + 1;
 127                     /* make key 2 start from 2 */
 128                     if ((key = (ce >> 4) & 0x0f)) *key_ptr[2]++ = key + 1;
 129                     /* key 3 is always a character code */
 130                     /*if (ce & 1)
 131                     {
 132                         if (wch >> 8) *key_ptr[3]++ = wch >> 8;
 133                         if (wch & 0xff) *key_ptr[3]++ = wch & 0xff;
 134                     }*/
 135                 }
 136                 /*else
 137                 {
 138                     *key_ptr[0]++ = 0xff;
 139                     *key_ptr[0]++ = 0xfe;
 140                     if (wch >> 8) *key_ptr[0]++ = wch >> 8;
 141                     if (wch & 0xff) *key_ptr[0]++ = wch & 0xff;
 142                 }*/
 143             }
 144         }
 145     }
 146
 147     *key_ptr[0] = '\1';
 148     *key_ptr[1] = '\1';
 149     *key_ptr[2] = '\1';
 150     *key_ptr[3]++ = '\1';
 151     *key_ptr[3] = 0;
 152
 153     return key_ptr[3] - dst;
 154 }