iconvdata/gb2312.h

   1 /* Access functions for GB2312 conversion.
   2    Copyright (C) 1998, 1999, 2003 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
   5
   6    The GNU C 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    The GNU C 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 the GNU C Library; if not, write to the Free
  18    Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
  19    02111-1307 USA.  */
  20
  21 #ifndef _GB2312_H
  22 #define _GB2312_H       1
  23
  24 #include <gconv.h>
  25 #include <stdint.h>
  26 #include <assert.h>
  27
  28 /* Conversion table.  */
  29 extern const uint16_t __gb2312_to_ucs[];
  30
  31
  32 static inline uint32_t
  33 __attribute ((always_inline))
  34 gb2312_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset)
  35 {
  36   unsigned char ch = *(*s);
  37   unsigned char ch2;
  38   int idx;
  39
  40   if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) > 0x77)
  41     return __UNKNOWN_10646_CHAR;
  42
  43   if (avail < 2)
  44     return 0;
  45
  46   ch2 = (*s)[1];
  47   if ((ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f)
  48     return __UNKNOWN_10646_CHAR;
  49
  50   idx = (ch - 0x21 - offset) * 94 + (ch2 - 0x21 - offset);
  51   if (idx > 0x1ff1)
  52     return __UNKNOWN_10646_CHAR;
  53
  54   (*s) += 2;
  55
  56   return __gb2312_to_ucs[idx] ?: ((*s) -= 2, __UNKNOWN_10646_CHAR);
  57 }
  58
  59
  60 extern const char __gb2312_from_ucs4_tab1[][2];
  61 extern const char __gb2312_from_ucs4_tab2[][2];
  62 extern const char __gb2312_from_ucs4_tab3[][2];
  63 extern const char __gb2312_from_ucs4_tab4[][2];
  64 extern const char __gb2312_from_ucs4_tab5[][2];
  65 extern const char __gb2312_from_ucs4_tab6[][2];
  66 extern const char __gb2312_from_ucs4_tab7[][2];
  67 extern const char __gb2312_from_ucs4_tab8[][2];
  68 extern const char __gb2312_from_ucs4_tab9[][2];
  69
  70 static inline size_t
  71 __attribute ((always_inline))
  72 ucs4_to_gb2312 (uint32_t wch, unsigned char *s, size_t avail)
  73 {
  74   unsigned int ch = (unsigned int) wch;
  75   char buf[2];
  76   const char *cp = buf;
  77
  78   switch (ch)
  79     {
  80     case 0xa4 ... 0x101:
  81       cp = __gb2312_from_ucs4_tab1[ch - 0xa4];
  82       break;
  83     case 0x113:
  84       cp = "\x28\x25";
  85       break;
  86     case 0x11b:
  87       cp = "\x28\x27";
  88       break;
  89     case 0x12b:
  90       cp = "\x28\x29";
  91       break;
  92     case 0x14d:
  93       cp = "\x28\x2d";
  94       break;
  95     case 0x16b:
  96       cp = "\x28\x31";
  97       break;
  98     case 0x1ce:
  99       cp = "\x28\x23";
 100       break;
 101     case 0x1d0:
 102       cp = "\x28\x2b";
 103       break;
 104     case 0x1d2:
 105       cp = "\x28\x2f";
 106       break;
 107     case 0x1d4:
 108       cp = "\x28\x33";
 109       break;
 110     case 0x1d6:
 111       cp = "\x28\x35";
 112       break;
 113     case 0x1d8:
 114       cp = "\x28\x36";
 115       break;
 116     case 0x1da:
 117       cp = "\x28\x37";
 118       break;
 119     case 0x1dc:
 120       cp = "\x28\x38";
 121       break;
 122     case 0x2c7:
 123       cp = "\x21\x26";
 124       break;
 125     case 0x2c9:
 126       cp = "\x21\x25";
 127       break;
 128     case 0x391 ... 0x3c9:
 129       cp = __gb2312_from_ucs4_tab2[ch - 0x391];
 130       break;
 131     case 0x401 ... 0x451:
 132       cp = __gb2312_from_ucs4_tab3[ch - 0x401];
 133       break;
 134     case 0x2015 ... 0x203b:
 135       cp = __gb2312_from_ucs4_tab4[ch - 0x2015];
 136       break;
 137     case 0x2103 ... 0x22a5:
 138       cp = __gb2312_from_ucs4_tab5[ch - 0x2103];
 139       break;
 140     case 0x2312:
 141       cp = "\x21\x50";
 142       break;
 143     case 0x2460 ... 0x249b:
 144       cp = __gb2312_from_ucs4_tab6[ch - 0x2460];
 145       break;
 146     case 0x2500 ... 0x254b:
 147       buf[0] = '\x29';
 148       buf[1] = '\x24' + (ch % 256);
 149       break;
 150     case 0x25a0:
 151       cp = "\x21\x76";
 152       break;
 153     case 0x25a1:
 154       cp = "\x21\x75";
 155       break;
 156     case 0x25b2:
 157       cp = "\x21\x78";
 158       break;
 159     case 0x25b3:
 160       cp = "\x21\x77";
 161       break;
 162     case 0x25c6:
 163       cp = "\x21\x74";
 164       break;
 165     case 0x25c7:
 166       cp = "\x21\x73";
 167       break;
 168     case 0x25cb:
 169       cp = "\x21\x70";
 170       break;
 171     case 0x25ce:
 172       cp = "\x21\x72";
 173       break;
 174     case 0x25cf:
 175       cp = "\x21\x71";
 176       break;
 177     case 0x2605:
 178       cp = "\x21\x6f";
 179       break;
 180     case 0x2606:
 181       cp = "\x21\x6e";
 182       break;
 183     case 0x2640:
 184       cp = "\x21\x62";
 185       break;
 186     case 0x2642:
 187       cp = "\x21\x61";
 188       break;
 189     case 0x3000 ... 0x3129:
 190       cp = __gb2312_from_ucs4_tab7[ch - 0x3000];
 191       break;
 192     case 0x3220 ... 0x3229:
 193       buf[0] = '\x22';
 194       buf[1] = '\x65' + (ch - 0x3220);
 195       break;
 196     case 0x4e00 ... 0x9fa0:
 197       cp = __gb2312_from_ucs4_tab8[ch - 0x4e00];
 198       break;
 199     case 0xff01 ... 0xff5e:
 200       cp = __gb2312_from_ucs4_tab9[ch - 0xff01];
 201       break;
 202     case 0xffe0:
 203       cp = "\x21\x69";
 204       break;
 205     case 0xffe1:
 206       cp = "\x21\x6a";
 207       break;
 208     case 0xffe3:
 209       cp = "\x23\x7e";
 210       break;
 211     case 0xffe5:
 212       cp = "\x23\x24";
 213       break;
 214     default:
 215       return __UNKNOWN_10646_CHAR;
 216     }
 217
 218   if (cp[0] == '\0')
 219     return __UNKNOWN_10646_CHAR;
 220
 221   assert (cp[1] != '\0');
 222
 223   if (avail < 2)
 224     return 0;
 225
 226   s[0] = cp[0];
 227   s[1] = cp[1];
 228
 229   return 2;
 230 }
 231
 232 #endif  /* gb2312.h */