| 1 |
/* |
| 2 |
* Copyright (C) 1999-2001, 2005, 2012, 2016 Free Software Foundation, Inc. |
| 3 |
* This file is part of the GNU LIBICONV Library. |
| 4 |
* |
| 5 |
* The GNU LIBICONV Library is free software; you can redistribute it |
| 6 |
* and/or modify it under the terms of the GNU Library General Public |
| 7 |
* License as published by the Free Software Foundation; either version 2 |
| 8 |
* of the License, or (at your option) any later version. |
| 9 |
* |
| 10 |
* The GNU LIBICONV Library is distributed in the hope that it will be |
| 11 |
* useful, but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 12 |
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 13 |
* Library General Public License for more details. |
| 14 |
* |
| 15 |
* You should have received a copy of the GNU Library General Public |
| 16 |
* License along with the GNU LIBICONV Library; see the file COPYING.LIB. |
| 17 |
* If not, see <http://www.gnu.org/licenses/>. |
| 18 |
*/ |
| 19 |
|
| 20 |
/* |
| 21 |
* ISO-IR-165 |
| 22 |
*/ |
| 23 |
|
| 24 |
/* |
| 25 |
* ISO-IR-165 is an extension of GB 2312, consisting of: |
| 26 |
* 1. GB 6345.1-86 corrections: |
| 27 |
* Two corrections to GB 2312, at 0x2367 and 0x6F71. |
| 28 |
* 2. GB 6345.1-86 additions: |
| 29 |
* - 6 new full-width pinyin characters in row 0x28. |
| 30 |
* - ISO646-CN in row 0x2A. |
| 31 |
* - 32 half-width pinyin characters in row 0x2B. |
| 32 |
* 3. GB 8565.2-88 additions: |
| 33 |
* - 50 characters in row 0x2D. |
| 34 |
* - 92 characters in row 0x2E. |
| 35 |
* - 93 characters in row 0x2F. |
| 36 |
* - 470 characters in rows 0x7A-0x7E. |
| 37 |
* 4. ISO-IR-165 additions: |
| 38 |
* - 22 characters in row 0x26. |
| 39 |
* - 94 characters in row 0x2C. |
| 40 |
* - 44 new characters in row 0x2D. |
| 41 |
* - 1 new character in row 0x2F. |
| 42 |
* |
| 43 |
* The conversion table was created from the following sources: |
| 44 |
* Ad 1. The 0x2367 correction is already integrated in the unicode.org |
| 45 |
* GB2312.TXT table. The 0x6F71 mapping is the same in the unicode.org |
| 46 |
* GB2312.TXT and UNIHAN.TXT table and in Koichi Yasuoka's Uni2GB table, |
| 47 |
* so we assume it's correct. |
| 48 |
* The unicode.org UNIHAN.TXT table about GB 8565 is not usable: it has |
| 49 |
* extraneous code points at rows 0x28, 0x2C, 0x2D. Note also that it does |
| 50 |
* not list the 69 non-hanzi in row 0x2F. Moreover, it has the characters |
| 51 |
* 0x2F7A-0x2F7D shifted down by one to 0x2F79-0x2F7C. |
| 52 |
* Therefore we take the GB8565 and ISO-IR-165 data from Koichi Yasuoka's |
| 53 |
* Uni2GB table. |
| 54 |
* Ad 1. Yasuoka maps 0x2367 to U+0261 (small script g) and 0x2840 to U+FF47 |
| 55 |
* (full-width small normal g). While coherent with ISO-IR's 165.pdf, |
| 56 |
* this disagrees with Ken Lunde's book: He says that ISO-IR-165 |
| 57 |
* includes the GB6345 correction, i.e. maps 0x2367 to U+FF47 or U+0067 |
| 58 |
* and _not_ to U+0261 (small script g). |
| 59 |
* To overcome the confusion, we just map both 0x2367 and 0x2840 to |
| 60 |
* U+FF47. |
| 61 |
* Ad 2. Row 0x28: Add a mapping from 0x283F to U+01F9. |
| 62 |
* Row 0x2A: Mapping is well-known, also present in Koichi Yasuoka's |
| 63 |
* table. |
| 64 |
* Row 0x2B: Typed in by hand from appendix E in Ken Lunde's book. |
| 65 |
* When converting from Unicode to ISO-IR-165, prefer the half-width |
| 66 |
* range 0x2B{21..40} to the full-width range 0x28{21..40}. |
| 67 |
* Ad 3. Rows 0x2D, 0x2E: Both Koichi Yasuoka's Uni2GB table and the UNIHAN.TXT |
| 68 |
* data for GB 8565 agree here. |
| 69 |
* Row 0x2F: Taken from Koichi Yasuoka's Uni2GB table. |
| 70 |
* Rows 0x7A-0x7E: Koichi Yasuoka's Uni2GB table and the UNIHAN.TXT |
| 71 |
* data for GB 8565 agree here mostly. Differences: |
| 72 |
* 0x7C38 -> U+6F26 or U+527A ? We choose U+6F26. |
| 73 |
* 0x7C5A -> U+7A40 or U+6996 ? We choose U+6996. |
| 74 |
* Ad 4. Row 0x26: Mapping unknown. |
| 75 |
* Rows 0x2C, 0x2D: Both Koichi Yasuoka's Uni2GB table and the UNIHAN.TXT |
| 76 |
* data for GB 8565 (!) agree here. |
| 77 |
* Row 0x2F: Taken from Koichi Yasuoka's Uni2GB table. |
| 78 |
*/ |
| 79 |
|
| 80 |
#include "isoir165ext.h" |
| 81 |
|
| 82 |
static int |
| 83 |
isoir165_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n) |
| 84 |
{ |
| 85 |
int ret; |
| 86 |
|
| 87 |
/* Map full-width pinyin (row 0x28) like half-width pinyin (row 0x2B). */ |
| 88 |
if (s[0] == 0x28) { |
| 89 |
if (n >= 2) { |
| 90 |
unsigned char c2 = s[1]; |
| 91 |
if (c2 >= 0x21 && c2 <= 0x40) { |
| 92 |
unsigned char buf[2]; |
| 93 |
buf[0] = 0x2b; |
| 94 |
buf[1] = c2; |
| 95 |
ret = isoir165ext_mbtowc(conv,pwc,buf,2); |
| 96 |
if (ret != RET_ILSEQ) |
| 97 |
return ret; |
| 98 |
} |
| 99 |
} |
| 100 |
} |
| 101 |
/* Try the GB2312 -> Unicode table. */ |
| 102 |
ret = gb2312_mbtowc(conv,pwc,s,n); |
| 103 |
if (ret != RET_ILSEQ) |
| 104 |
return ret; |
| 105 |
/* Row 0x2A is GB_1988-80. */ |
| 106 |
if (s[0] == 0x2a) { |
| 107 |
if (n >= 2) { |
| 108 |
unsigned char c2 = s[1]; |
| 109 |
if (c2 >= 0x21 && c2 < 0x7f) { |
| 110 |
ret = iso646_cn_mbtowc(conv,pwc,s+1,1); |
| 111 |
if (ret != 1) abort(); |
| 112 |
return 2; |
| 113 |
} |
| 114 |
return RET_ILSEQ; |
| 115 |
} |
| 116 |
return RET_TOOFEW(0); |
| 117 |
} |
| 118 |
/* Try the ISO-IR-165 extensions -> Unicode table. */ |
| 119 |
ret = isoir165ext_mbtowc(conv,pwc,s,n); |
| 120 |
return ret; |
| 121 |
} |
| 122 |
|
| 123 |
static int |
| 124 |
isoir165_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n) |
| 125 |
{ |
| 126 |
unsigned char buf[2]; |
| 127 |
int ret; |
| 128 |
|
| 129 |
/* Try the Unicode -> GB2312 table. */ |
| 130 |
ret = gb2312_wctomb(conv,buf,wc,2); |
| 131 |
if (ret != RET_ILUNI) { |
| 132 |
if (ret != 2) abort(); |
| 133 |
if (!(buf[0] == 0x28 && buf[1] >= 0x21 && buf[1] <= 0x40)) { |
| 134 |
if (n >= 2) { |
| 135 |
r[0] = buf[0]; |
| 136 |
r[1] = buf[1]; |
| 137 |
return 2; |
| 138 |
} |
| 139 |
return RET_TOOSMALL; |
| 140 |
} |
| 141 |
} |
| 142 |
/* Row 0x2A is GB_1988-80. */ |
| 143 |
ret = iso646_cn_wctomb(conv,buf,wc,1); |
| 144 |
if (ret != RET_ILUNI) { |
| 145 |
if (ret != 1) abort(); |
| 146 |
if (buf[0] >= 0x21 && buf[0] < 0x7f) { |
| 147 |
if (n >= 2) { |
| 148 |
r[0] = 0x2a; |
| 149 |
r[1] = buf[0]; |
| 150 |
return 2; |
| 151 |
} |
| 152 |
return RET_TOOSMALL; |
| 153 |
} |
| 154 |
} |
| 155 |
/* Try the Unicode -> ISO-IR-165 extensions table. */ |
| 156 |
ret = isoir165ext_wctomb(conv,r,wc,n); |
| 157 |
return ret; |
| 158 |
} |