1: /*
2: * Copyright (C) 1999-2002, 2006, 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 <https://www.gnu.org/licenses/>.
18: */
19:
20: /*
21: * BIG5-HKSCS:2001
22: */
23:
24: /*
25: * BIG5-HKSCS:2001 can be downloaded from
26: * http://www.info.gov.hk/digital21/eng/hkscs/download.html
27: * http://www.info.gov.hk/digital21/eng/hkscs/index.html
28: *
29: * It extends BIG5-HKSCS:1999 through 116 characters.
30: *
31: * It extends BIG5 (without the rows 0xC6..0xC7) through the ranges
32: *
33: * 0x{88..8D}{40..7E,A1..FE} 757 characters
34: * 0x{8E..A0}{40..7E,A1..FE} 2898 characters
35: * 0x{C6..C8}{40..7E,A1..FE} 359 characters
36: * 0xF9{D6..FE} 41 characters
37: * 0x{FA..FE}{40..7E,A1..FE} 763 characters
38: *
39: * Note that some HKSCS characters are not contained in Unicode 3.2
40: * and are therefore best represented as sequences of Unicode characters:
41: * 0x8862 U+00CA U+0304 LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND MACRON
42: * 0x8864 U+00CA U+030C LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND CARON
43: * 0x88A3 U+00EA U+0304 LATIN SMALL LETTER E WITH CIRCUMFLEX AND MACRON
44: * 0x88A5 U+00EA U+030C LATIN SMALL LETTER E WITH CIRCUMFLEX AND CARON
45: */
46:
47: #include "hkscs2001.h"
48: #include "flushwc.h"
49:
50: static int
51: big5hkscs2001_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
52: {
53: ucs4_t last_wc = conv->istate;
54: if (last_wc) {
55: /* Output the buffered character. */
56: conv->istate = 0;
57: *pwc = last_wc;
58: return 0; /* Don't advance the input pointer. */
59: } else {
60: unsigned char c = *s;
61: /* Code set 0 (ASCII) */
62: if (c < 0x80)
63: return ascii_mbtowc(conv,pwc,s,n);
64: /* Code set 1 (BIG5 extended) */
65: if (c >= 0xa1 && c < 0xff) {
66: if (n < 2)
67: return RET_TOOFEW(0);
68: {
69: unsigned char c2 = s[1];
70: if ((c2 >= 0x40 && c2 < 0x7f) || (c2 >= 0xa1 && c2 < 0xff)) {
71: if (!((c == 0xc6 && c2 >= 0xa1) || c == 0xc7)) {
72: int ret = big5_mbtowc(conv,pwc,s,2);
73: if (ret != RET_ILSEQ)
74: return ret;
75: }
76: }
77: }
78: }
79: {
80: int ret = hkscs1999_mbtowc(conv,pwc,s,n);
81: if (ret != RET_ILSEQ)
82: return ret;
83: }
84: {
85: int ret = hkscs2001_mbtowc(conv,pwc,s,n);
86: if (ret != RET_ILSEQ)
87: return ret;
88: }
89: if (c == 0x88) {
90: if (n < 2)
91: return RET_TOOFEW(0);
92: {
93: unsigned char c2 = s[1];
94: if (c2 == 0x62 || c2 == 0x64 || c2 == 0xa3 || c2 == 0xa5) {
95: /* It's a composed character. */
96: ucs4_t wc1 = ((c2 >> 3) << 2) + 0x009a; /* = 0x00ca or 0x00ea */
97: ucs4_t wc2 = ((c2 & 6) << 2) + 0x02fc; /* = 0x0304 or 0x030c */
98: /* We cannot output two Unicode characters at once. So,
99: output the first character and buffer the second one. */
100: *pwc = wc1;
101: conv->istate = wc2;
102: return 2;
103: }
104: }
105: }
106: return RET_ILSEQ;
107: }
108: }
109:
110: #define big5hkscs2001_flushwc normal_flushwc
111:
112: static int
113: big5hkscs2001_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
114: {
115: int count = 0;
116: unsigned char last = conv->ostate;
117:
118: if (last) {
119: /* last is = 0x66 or = 0xa7. */
120: if (wc == 0x0304 || wc == 0x030c) {
121: /* Output the combined character. */
122: if (n >= 2) {
123: r[0] = 0x88;
124: r[1] = last + ((wc & 24) >> 2) - 4; /* = 0x62 or 0x64 or 0xa3 or 0xa5 */
125: conv->ostate = 0;
126: return 2;
127: } else
128: return RET_TOOSMALL;
129: }
130:
131: /* Output the buffered character. */
132: if (n < 2)
133: return RET_TOOSMALL;
134: r[0] = 0x88;
135: r[1] = last;
136: r += 2;
137: count = 2;
138: }
139:
140: /* Code set 0 (ASCII) */
141: if (wc < 0x0080) {
142: /* Plain ASCII character. */
143: if (n > count) {
144: r[0] = (unsigned char) wc;
145: conv->ostate = 0;
146: return count+1;
147: } else
148: return RET_TOOSMALL;
149: } else {
150: unsigned char buf[2];
151: int ret;
152:
153: /* Code set 1 (BIG5 extended) */
154: ret = big5_wctomb(conv,buf,wc,2);
155: if (ret != RET_ILUNI) {
156: if (ret != 2) abort();
157: if (!((buf[0] == 0xc6 && buf[1] >= 0xa1) || buf[0] == 0xc7)) {
158: if (n >= count+2) {
159: r[0] = buf[0];
160: r[1] = buf[1];
161: conv->ostate = 0;
162: return count+2;
163: } else
164: return RET_TOOSMALL;
165: }
166: }
167: ret = hkscs1999_wctomb(conv,buf,wc,2);
168: if (ret != RET_ILUNI) {
169: if (ret != 2) abort();
170: if ((wc & ~0x0020) == 0x00ca) {
171: /* A possible first character of a multi-character sequence. We have to
172: buffer it. */
173: if (!(buf[0] == 0x88 && (buf[1] == 0x66 || buf[1] == 0xa7))) abort();
174: conv->ostate = buf[1]; /* = 0x66 or = 0xa7 */
175: return count+0;
176: }
177: if (n >= count+2) {
178: r[0] = buf[0];
179: r[1] = buf[1];
180: conv->ostate = 0;
181: return count+2;
182: } else
183: return RET_TOOSMALL;
184: }
185: ret = hkscs2001_wctomb(conv,buf,wc,2);
186: if (ret != RET_ILUNI) {
187: if (ret != 2) abort();
188: if (n >= count+2) {
189: r[0] = buf[0];
190: r[1] = buf[1];
191: conv->ostate = 0;
192: return count+2;
193: } else
194: return RET_TOOSMALL;
195: }
196: return RET_ILUNI;
197: }
198: }
199:
200: static int
201: big5hkscs2001_reset (conv_t conv, unsigned char *r, size_t n)
202: {
203: unsigned char last = conv->ostate;
204:
205: if (last) {
206: if (n < 2)
207: return RET_TOOSMALL;
208: r[0] = 0x88;
209: r[1] = last;
210: /* conv->ostate = 0; will be done by the caller */
211: return 2;
212: } else
213: return 0;
214: }
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to big5hkscs2001.h CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / libiconv / lib