| 1 | /* Access functions for GB2312 conversion. | 
|---|
| 2 | Copyright (C) 1998-2020 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, see | 
|---|
| 18 | <https://www.gnu.org/licenses/>.  */ | 
|---|
| 19 |  | 
|---|
| 20 | #ifndef _GB2312_H | 
|---|
| 21 | #define _GB2312_H	1 | 
|---|
| 22 |  | 
|---|
| 23 | #include <gconv.h> | 
|---|
| 24 | #include <stdint.h> | 
|---|
| 25 | #include <assert.h> | 
|---|
| 26 |  | 
|---|
| 27 | /* Conversion table.  */ | 
|---|
| 28 | extern const uint16_t __gb2312_to_ucs[]; | 
|---|
| 29 |  | 
|---|
| 30 |  | 
|---|
| 31 | static inline uint32_t | 
|---|
| 32 | __attribute ((always_inline)) | 
|---|
| 33 | gb2312_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset) | 
|---|
| 34 | { | 
|---|
| 35 | unsigned char ch = *(*s); | 
|---|
| 36 | unsigned char ch2; | 
|---|
| 37 | int idx; | 
|---|
| 38 |  | 
|---|
| 39 | if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) > 0x77) | 
|---|
| 40 | return __UNKNOWN_10646_CHAR; | 
|---|
| 41 |  | 
|---|
| 42 | if (avail < 2) | 
|---|
| 43 | return 0; | 
|---|
| 44 |  | 
|---|
| 45 | ch2 = (*s)[1]; | 
|---|
| 46 | if ((ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f) | 
|---|
| 47 | return __UNKNOWN_10646_CHAR; | 
|---|
| 48 |  | 
|---|
| 49 | idx = (ch - 0x21 - offset) * 94 + (ch2 - 0x21 - offset); | 
|---|
| 50 | if (idx > 0x1ff1) | 
|---|
| 51 | return __UNKNOWN_10646_CHAR; | 
|---|
| 52 |  | 
|---|
| 53 | (*s) += 2; | 
|---|
| 54 |  | 
|---|
| 55 | return __gb2312_to_ucs[idx] ?: ((*s) -= 2, __UNKNOWN_10646_CHAR); | 
|---|
| 56 | } | 
|---|
| 57 |  | 
|---|
| 58 |  | 
|---|
| 59 | extern const char __gb2312_from_ucs4_tab1[][2]; | 
|---|
| 60 | extern const char __gb2312_from_ucs4_tab2[][2]; | 
|---|
| 61 | extern const char __gb2312_from_ucs4_tab3[][2]; | 
|---|
| 62 | extern const char __gb2312_from_ucs4_tab4[][2]; | 
|---|
| 63 | extern const char __gb2312_from_ucs4_tab5[][2]; | 
|---|
| 64 | extern const char __gb2312_from_ucs4_tab6[][2]; | 
|---|
| 65 | extern const char __gb2312_from_ucs4_tab7[][2]; | 
|---|
| 66 | extern const char __gb2312_from_ucs4_tab8[][2]; | 
|---|
| 67 | extern const char __gb2312_from_ucs4_tab9[][2]; | 
|---|
| 68 |  | 
|---|
| 69 | static inline size_t | 
|---|
| 70 | __attribute ((always_inline)) | 
|---|
| 71 | ucs4_to_gb2312 (uint32_t wch, unsigned char *s, size_t avail) | 
|---|
| 72 | { | 
|---|
| 73 | unsigned int ch = (unsigned int) wch; | 
|---|
| 74 | char buf[2]; | 
|---|
| 75 | const char *cp = buf; | 
|---|
| 76 |  | 
|---|
| 77 | switch (ch) | 
|---|
| 78 | { | 
|---|
| 79 | case 0xa4 ... 0x101: | 
|---|
| 80 | cp = __gb2312_from_ucs4_tab1[ch - 0xa4]; | 
|---|
| 81 | break; | 
|---|
| 82 | case 0x113: | 
|---|
| 83 | cp = "\x28\x25"; | 
|---|
| 84 | break; | 
|---|
| 85 | case 0x11b: | 
|---|
| 86 | cp = "\x28\x27"; | 
|---|
| 87 | break; | 
|---|
| 88 | case 0x12b: | 
|---|
| 89 | cp = "\x28\x29"; | 
|---|
| 90 | break; | 
|---|
| 91 | case 0x14d: | 
|---|
| 92 | cp = "\x28\x2d"; | 
|---|
| 93 | break; | 
|---|
| 94 | case 0x16b: | 
|---|
| 95 | cp = "\x28\x31"; | 
|---|
| 96 | break; | 
|---|
| 97 | case 0x1ce: | 
|---|
| 98 | cp = "\x28\x23"; | 
|---|
| 99 | break; | 
|---|
| 100 | case 0x1d0: | 
|---|
| 101 | cp = "\x28\x2b"; | 
|---|
| 102 | break; | 
|---|
| 103 | case 0x1d2: | 
|---|
| 104 | cp = "\x28\x2f"; | 
|---|
| 105 | break; | 
|---|
| 106 | case 0x1d4: | 
|---|
| 107 | cp = "\x28\x33"; | 
|---|
| 108 | break; | 
|---|
| 109 | case 0x1d6: | 
|---|
| 110 | cp = "\x28\x35"; | 
|---|
| 111 | break; | 
|---|
| 112 | case 0x1d8: | 
|---|
| 113 | cp = "\x28\x36"; | 
|---|
| 114 | break; | 
|---|
| 115 | case 0x1da: | 
|---|
| 116 | cp = "\x28\x37"; | 
|---|
| 117 | break; | 
|---|
| 118 | case 0x1dc: | 
|---|
| 119 | cp = "\x28\x38"; | 
|---|
| 120 | break; | 
|---|
| 121 | case 0x2c7: | 
|---|
| 122 | cp = "\x21\x26"; | 
|---|
| 123 | break; | 
|---|
| 124 | case 0x2c9: | 
|---|
| 125 | cp = "\x21\x25"; | 
|---|
| 126 | break; | 
|---|
| 127 | case 0x391 ... 0x3c9: | 
|---|
| 128 | cp = __gb2312_from_ucs4_tab2[ch - 0x391]; | 
|---|
| 129 | break; | 
|---|
| 130 | case 0x401 ... 0x451: | 
|---|
| 131 | cp = __gb2312_from_ucs4_tab3[ch - 0x401]; | 
|---|
| 132 | break; | 
|---|
| 133 | case 0x2015 ... 0x203b: | 
|---|
| 134 | cp = __gb2312_from_ucs4_tab4[ch - 0x2015]; | 
|---|
| 135 | break; | 
|---|
| 136 | case 0x2103 ... 0x22a5: | 
|---|
| 137 | cp = __gb2312_from_ucs4_tab5[ch - 0x2103]; | 
|---|
| 138 | break; | 
|---|
| 139 | case 0x2312: | 
|---|
| 140 | cp = "\x21\x50"; | 
|---|
| 141 | break; | 
|---|
| 142 | case 0x2460 ... 0x249b: | 
|---|
| 143 | cp = __gb2312_from_ucs4_tab6[ch - 0x2460]; | 
|---|
| 144 | break; | 
|---|
| 145 | case 0x2500 ... 0x254b: | 
|---|
| 146 | buf[0] = '\x29'; | 
|---|
| 147 | buf[1] = '\x24' + (ch % 256); | 
|---|
| 148 | break; | 
|---|
| 149 | case 0x25a0: | 
|---|
| 150 | cp = "\x21\x76"; | 
|---|
| 151 | break; | 
|---|
| 152 | case 0x25a1: | 
|---|
| 153 | cp = "\x21\x75"; | 
|---|
| 154 | break; | 
|---|
| 155 | case 0x25b2: | 
|---|
| 156 | cp = "\x21\x78"; | 
|---|
| 157 | break; | 
|---|
| 158 | case 0x25b3: | 
|---|
| 159 | cp = "\x21\x77"; | 
|---|
| 160 | break; | 
|---|
| 161 | case 0x25c6: | 
|---|
| 162 | cp = "\x21\x74"; | 
|---|
| 163 | break; | 
|---|
| 164 | case 0x25c7: | 
|---|
| 165 | cp = "\x21\x73"; | 
|---|
| 166 | break; | 
|---|
| 167 | case 0x25cb: | 
|---|
| 168 | cp = "\x21\x70"; | 
|---|
| 169 | break; | 
|---|
| 170 | case 0x25ce: | 
|---|
| 171 | cp = "\x21\x72"; | 
|---|
| 172 | break; | 
|---|
| 173 | case 0x25cf: | 
|---|
| 174 | cp = "\x21\x71"; | 
|---|
| 175 | break; | 
|---|
| 176 | case 0x2605: | 
|---|
| 177 | cp = "\x21\x6f"; | 
|---|
| 178 | break; | 
|---|
| 179 | case 0x2606: | 
|---|
| 180 | cp = "\x21\x6e"; | 
|---|
| 181 | break; | 
|---|
| 182 | case 0x2640: | 
|---|
| 183 | cp = "\x21\x62"; | 
|---|
| 184 | break; | 
|---|
| 185 | case 0x2642: | 
|---|
| 186 | cp = "\x21\x61"; | 
|---|
| 187 | break; | 
|---|
| 188 | case 0x3000 ... 0x3129: | 
|---|
| 189 | cp = __gb2312_from_ucs4_tab7[ch - 0x3000]; | 
|---|
| 190 | break; | 
|---|
| 191 | case 0x3220 ... 0x3229: | 
|---|
| 192 | buf[0] = '\x22'; | 
|---|
| 193 | buf[1] = '\x65' + (ch - 0x3220); | 
|---|
| 194 | break; | 
|---|
| 195 | case 0x4e00 ... 0x9fa0: | 
|---|
| 196 | cp = __gb2312_from_ucs4_tab8[ch - 0x4e00]; | 
|---|
| 197 | break; | 
|---|
| 198 | case 0xff01 ... 0xff5e: | 
|---|
| 199 | cp = __gb2312_from_ucs4_tab9[ch - 0xff01]; | 
|---|
| 200 | break; | 
|---|
| 201 | case 0xffe0: | 
|---|
| 202 | cp = "\x21\x69"; | 
|---|
| 203 | break; | 
|---|
| 204 | case 0xffe1: | 
|---|
| 205 | cp = "\x21\x6a"; | 
|---|
| 206 | break; | 
|---|
| 207 | case 0xffe3: | 
|---|
| 208 | cp = "\x23\x7e"; | 
|---|
| 209 | break; | 
|---|
| 210 | case 0xffe5: | 
|---|
| 211 | cp = "\x23\x24"; | 
|---|
| 212 | break; | 
|---|
| 213 | default: | 
|---|
| 214 | return __UNKNOWN_10646_CHAR; | 
|---|
| 215 | } | 
|---|
| 216 |  | 
|---|
| 217 | if (cp[0] == '\0') | 
|---|
| 218 | return __UNKNOWN_10646_CHAR; | 
|---|
| 219 |  | 
|---|
| 220 | assert (cp[1] != '\0'); | 
|---|
| 221 |  | 
|---|
| 222 | if (avail < 2) | 
|---|
| 223 | return 0; | 
|---|
| 224 |  | 
|---|
| 225 | s[0] = cp[0]; | 
|---|
| 226 | s[1] = cp[1]; | 
|---|
| 227 |  | 
|---|
| 228 | return 2; | 
|---|
| 229 | } | 
|---|
| 230 |  | 
|---|
| 231 | #endif	/* gb2312.h */ | 
|---|
| 232 |  | 
|---|