iconvdata/cns11643.h

   1 /* Access functions for CNS 11643 handling.
   2    Copyright (C) 1998,1999,2000,2001,2002,2003,2007
   3    Free Software Foundation, Inc.
   4    This file is part of the GNU C Library.
   5    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
   6
   7    The GNU C Library is free software; you can redistribute it and/or
   8    modify it under the terms of the GNU Lesser General Public
   9    License as published by the Free Software Foundation; either
  10    version 2.1 of the License, or (at your option) any later version.
  11
  12    The GNU C Library is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15    Lesser General Public License for more details.
  16
  17    You should have received a copy of the GNU Lesser General Public
  18    License along with the GNU C Library; if not, see
  19    <http://www.gnu.org/licenses/>.  */
  20
  21 #include <stdint.h>
  22
  23
  24 /* Table for CNS 11643, plane 1 to UCS4 conversion.  */
  25 extern const uint16_t __cns11643l1_to_ucs4_tab[];
  26 /* Table for CNS 11643, plane 2 to UCS4 conversion.  */
  27 extern const uint16_t __cns11643l2_to_ucs4_tab[];
  28 /* Table for CNS 11643, plane 3 to UCS4 conversion.  */
  29 extern const uint32_t __cns11643l3_to_ucs4_tab[];
  30 /* Table for CNS 11643, plane 4 to UCS4 conversion.  */
  31 extern const uint32_t __cns11643l4_to_ucs4_tab[];
  32 /* Table for CNS 11643, plane 5 to UCS4 conversion.  */
  33 extern const uint32_t __cns11643l5_to_ucs4_tab[];
  34 /* Table for CNS 11643, plane 6 to UCS4 conversion.  */
  35 extern const uint32_t __cns11643l6_to_ucs4_tab[];
  36 /* Table for CNS 11643, plane 7 to UCS4 conversion.  */
  37 extern const uint32_t __cns11643l7_to_ucs4_tab[];
  38 /* Table for CNS 11643, plane 15 (old) to UCS4 conversion.  */
  39 extern const uint32_t __cns11643l15_to_ucs4_tab[];
  40
  41
  42 static inline uint32_t
  43 __attribute ((always_inline))
  44 cns11643_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset)
  45 {
  46   unsigned char ch = *(*s);
  47   unsigned char ch2;
  48   unsigned char ch3;
  49   uint32_t result;
  50   int idx;
  51
  52   if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) > 0x30)
  53     return __UNKNOWN_10646_CHAR;
  54
  55   if (avail < 3)
  56     return 0;
  57
  58   ch2 = (*s)[1];
  59   if ((ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f)
  60     return __UNKNOWN_10646_CHAR;
  61
  62   ch3 = (*s)[2];
  63   if ((ch3 - offset) <= 0x20 || (ch3 - offset) >= 0x7f)
  64     return __UNKNOWN_10646_CHAR;
  65
  66   idx = (ch2 - 0x21 - offset) * 94 + (ch3 - 0x21 - offset);
  67
  68   switch (ch - 0x20 - offset)
  69     {
  70     case 1:
  71       if (idx > 0x21f2)
  72         return __UNKNOWN_10646_CHAR;
  73       result = __cns11643l1_to_ucs4_tab[idx];
  74       break;
  75     case 2:
  76       if (idx > 0x1de1)
  77         return __UNKNOWN_10646_CHAR;
  78       result = __cns11643l2_to_ucs4_tab[idx];
  79       break;
  80     case 3:
  81       if (idx > 0x19bd)
  82         return __UNKNOWN_10646_CHAR;
  83       result = __cns11643l3_to_ucs4_tab[idx];
  84       break;
  85     case 4:
  86       if (idx > 0x1c81)
  87         return __UNKNOWN_10646_CHAR;
  88       result = __cns11643l4_to_ucs4_tab[idx];
  89       break;
  90     case 5:
  91       if (idx > 0x219a)
  92         return __UNKNOWN_10646_CHAR;
  93       result = __cns11643l5_to_ucs4_tab[idx];
  94       break;
  95     case 6:
  96       if (idx > 0x18f3)
  97         return __UNKNOWN_10646_CHAR;
  98       result = __cns11643l6_to_ucs4_tab[idx];
  99       break;
 100     case 7:
 101       if (idx > 0x198a)
 102         return __UNKNOWN_10646_CHAR;
 103       result = __cns11643l7_to_ucs4_tab[idx];
 104       break;
 105     case 15:
 106       if (idx > 0x1c00)
 107         return __UNKNOWN_10646_CHAR;
 108       result = __cns11643l15_to_ucs4_tab[idx];
 109       break;
 110     default:
 111       return __UNKNOWN_10646_CHAR;
 112     }
 113
 114   if (result != L'\0')
 115     (*s) += 3;
 116   else
 117     result = __UNKNOWN_10646_CHAR;
 118
 119   return result;
 120 }
 121
 122
 123 /* Tables for the UCS4 -> CNS conversion.  */
 124 extern const char __cns11643l1_from_ucs4_tab1[][2];
 125 extern const char __cns11643l1_from_ucs4_tab2[][2];
 126 extern const char __cns11643l1_from_ucs4_tab3[][2];
 127 extern const char __cns11643l1_from_ucs4_tab4[][2];
 128 extern const char __cns11643l1_from_ucs4_tab5[][2];
 129 extern const char __cns11643l1_from_ucs4_tab6[][2];
 130 extern const char __cns11643l1_from_ucs4_tab7[][2];
 131 extern const char __cns11643l1_from_ucs4_tab8[][2];
 132 extern const char __cns11643l1_from_ucs4_tab9[][2];
 133 extern const char __cns11643l1_from_ucs4_tab10[][2];
 134 extern const char __cns11643l1_from_ucs4_tab11[][2];
 135 extern const char __cns11643l1_from_ucs4_tab12[][2];
 136 extern const char __cns11643l1_from_ucs4_tab13[][2];
 137 extern const char __cns11643l1_from_ucs4_tab14[][2];
 138 extern const char __cns11643_from_ucs4p0_tab[][3];
 139 extern const char __cns11643_from_ucs4p2_tab[][3];
 140 extern const char __cns11643_from_ucs4p2c_tab[][3];
 141
 142
 143 static inline size_t
 144 __attribute ((always_inline))
 145 ucs4_to_cns11643 (uint32_t wch, unsigned char *s, size_t avail)
 146 {
 147   unsigned int ch = (unsigned int) wch;
 148   char buf[2];
 149   const char *cp = buf;
 150   size_t needed = 2;
 151
 152   switch (ch)
 153     {
 154     case 0xa7 ... 0xf7:
 155       cp = __cns11643l1_from_ucs4_tab1[ch - 0xa7];
 156       break;
 157     case 0x2c7 ... 0x2d9:
 158       cp = __cns11643l1_from_ucs4_tab2[ch - 0x2c7];
 159       break;
 160     case 0x391 ... 0x3c9:
 161       cp = __cns11643l1_from_ucs4_tab3[ch - 0x391];
 162       break;
 163     case 0x2013 ... 0x203e:
 164       cp = __cns11643l1_from_ucs4_tab4[ch - 0x2013];
 165       break;
 166     case 0x2103:
 167       cp = "\x22\x6a";
 168       break;
 169     case 0x2105:
 170       cp = "\x22\x22";
 171       break;
 172     case 0x2109:
 173       cp = "\x22\x6b";
 174       break;
 175     case 0x2160 ... 0x2169:
 176       buf[0] = '\x24';
 177       buf[1] = '\x2b' + (ch - 0x2160);
 178       break;
 179     case 0x2170 ... 0x2179:
 180       buf[0] = '\x26';
 181       buf[1] = '\x35' + (ch - 0x2170);
 182       break;
 183     case 0x2190 ... 0x2199:
 184       cp = __cns11643l1_from_ucs4_tab5[ch - 0x2190];
 185       break;
 186     case 0x2215 ... 0x2267:
 187       cp = __cns11643l1_from_ucs4_tab6[ch - 0x2215];
 188       break;
 189     case 0x22a5:
 190       cp = "\x22\x47";
 191       break;
 192     case 0x22bf:
 193       cp = "\x22\x4a";
 194       break;
 195     case 0x2400 ... 0x2421:
 196       cp = __cns11643l1_from_ucs4_tab7[ch - 0x2400];
 197       break;
 198     case 0x2460 ... 0x247d:
 199       cp = __cns11643l1_from_ucs4_tab8[ch - 0x2460];
 200       break;
 201     case 0x2500 ... 0x2642:
 202       cp = __cns11643l1_from_ucs4_tab9[ch - 0x2500];
 203       break;
 204     case 0x3000 ... 0x3029:
 205       cp = __cns11643l1_from_ucs4_tab10[ch - 0x3000];
 206       break;
 207     case 0x30fb:
 208       cp = "\x21\x26";
 209       break;
 210     case 0x3105 ... 0x3129:
 211       buf[0] = '\x25';
 212       buf[1] = '\x47' + (ch - 0x3105);
 213       break;
 214     case 0x32a3:
 215       cp = "\x22\x21";
 216       break;
 217     case 0x338e ... 0x33d5:
 218       cp = __cns11643l1_from_ucs4_tab11[ch - 0x338e];
 219       break;
 220     case 0x4e00 ... 0x9f9c:
 221       cp = __cns11643l1_from_ucs4_tab12[ch - 0x4e00];
 222       if (cp[0] != '\0')
 223         break;
 224       /* FALLTHROUGH.  Let's try the other planes.  */
 225     case 0x3400 ... 0x4dff:
 226     case 0x9f9d ... 0x9fa5:
 227       /* Let's try the other planes.  */
 228       needed = 3;
 229       cp = __cns11643_from_ucs4p0_tab[ch - 0x3400];
 230       break;
 231     case 0xfa28:
 232       needed = 3;
 233       cp = "\x0f\x58\x4c";
 234       break;
 235     case 0xfe30 ... 0xfe6b:
 236       cp = __cns11643l1_from_ucs4_tab13[ch - 0xfe30];
 237       break;
 238     case 0xff01 ... 0xff5d:
 239       cp = __cns11643l1_from_ucs4_tab14[ch - 0xff01];
 240       break;
 241     case 0xffe0:
 242       cp = "\x22\x66";
 243       break;
 244     case 0xffe1:
 245       cp = "\x22\x67";
 246       break;
 247     case 0xffe5:
 248       cp = "\x22\x64";
 249       break;
 250     case 0x20000 ... 0x2a6d6:
 251       needed = 3;
 252       cp = __cns11643_from_ucs4p2_tab[ch - 0x20000];
 253       break;
 254     case 0x2f800 ... 0x2fa1d:
 255       needed = 3;
 256       cp = __cns11643_from_ucs4p2c_tab[ch - 0x2f800];
 257       break;
 258     default:
 259       return __UNKNOWN_10646_CHAR;
 260     }
 261
 262   if (cp[0] == '\0')
 263     return __UNKNOWN_10646_CHAR;
 264
 265   if (avail < needed)
 266     return 0;
 267
 268   s[0] = cp[0];
 269   s[1] = cp[1];
 270   if (needed == 3)
 271     s[2] = cp[2];
 272
 273   return needed;
 274 }