Annotation of embedaddon/libiconv/lib/iso2022_kr.h, revision 1.1.1.2
1.1 misho 1: /*
1.1.1.2 ! misho 2: * Copyright (C) 1999-2001, 2008, 2016 Free Software Foundation, Inc.
1.1 misho 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.
1.1.1.2 ! misho 17: * If not, see <https://www.gnu.org/licenses/>.
1.1 misho 18: */
19:
20: /*
21: * ISO-2022-KR
22: */
23:
24: /* Specification: RFC 1557 */
25:
26: /* Note: CJK.INF says the SO designator needs to appear only once at the
27: beginning of a text, but to decrease the risk of ambiguities, when
28: producing ISO-2022-KR, we repeat the designator in every line containing
29: SO characters. RFC 1557 does not mandate this. */
30:
31: #define ESC 0x1b
32: #define SO 0x0e
33: #define SI 0x0f
34:
35: /*
36: * The state is composed of one of the following values
37: */
38: #define STATE_ASCII 0
39: #define STATE_TWOBYTE 1
40: /*
41: * and one of the following values, << 8
42: */
43: #define STATE2_NONE 0
44: #define STATE2_DESIGNATED_KSC5601 1
45:
46: #define SPLIT_STATE \
47: unsigned int state1 = state & 0xff, state2 = state >> 8
48: #define COMBINE_STATE \
49: state = (state2 << 8) | state1
50:
51: static int
1.1.1.2 ! misho 52: iso2022_kr_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
1.1 misho 53: {
54: state_t state = conv->istate;
55: SPLIT_STATE;
56: int count = 0;
57: unsigned char c;
58: for (;;) {
59: c = *s;
60: if (c == ESC) {
61: if (n < count+4)
62: goto none;
63: if (s[1] == '$') {
64: if (s[2] == ')') {
65: if (s[3] == 'C') {
66: state2 = STATE2_DESIGNATED_KSC5601;
67: s += 4; count += 4;
68: if (n < count+1)
69: goto none;
70: continue;
71: }
72: }
73: }
74: goto ilseq;
75: }
76: if (c == SO) {
77: if (state2 != STATE2_DESIGNATED_KSC5601)
78: goto ilseq;
79: state1 = STATE_TWOBYTE;
80: s++; count++;
81: if (n < count+1)
82: goto none;
83: continue;
84: }
85: if (c == SI) {
86: state1 = STATE_ASCII;
87: s++; count++;
88: if (n < count+1)
89: goto none;
90: continue;
91: }
92: break;
93: }
94: switch (state1) {
95: case STATE_ASCII:
96: if (c < 0x80) {
97: int ret = ascii_mbtowc(conv,pwc,s,1);
98: if (ret == RET_ILSEQ)
99: goto ilseq;
100: if (ret != 1) abort();
101: #if 0 /* Accept ISO-2022-KR according to CJK.INF. */
102: if (*pwc == 0x000a || *pwc == 0x000d)
103: state2 = STATE2_NONE;
104: #endif
105: COMBINE_STATE;
106: conv->istate = state;
107: return count+1;
108: } else
109: goto ilseq;
110: case STATE_TWOBYTE:
111: if (n < count+2)
112: goto none;
113: if (state2 != STATE2_DESIGNATED_KSC5601) abort();
114: if (s[0] < 0x80 && s[1] < 0x80) {
115: int ret = ksc5601_mbtowc(conv,pwc,s,2);
116: if (ret == RET_ILSEQ)
117: goto ilseq;
118: if (ret != 2) abort();
119: COMBINE_STATE;
120: conv->istate = state;
121: return count+2;
122: } else
123: goto ilseq;
124: default: abort();
125: }
126:
127: none:
128: COMBINE_STATE;
129: conv->istate = state;
130: return RET_TOOFEW(count);
131:
132: ilseq:
133: COMBINE_STATE;
134: conv->istate = state;
135: return RET_SHIFT_ILSEQ(count);
136: }
137:
138: static int
1.1.1.2 ! misho 139: iso2022_kr_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
1.1 misho 140: {
141: state_t state = conv->ostate;
142: SPLIT_STATE;
143: unsigned char buf[2];
144: int ret;
145:
146: /* Try ASCII. */
147: ret = ascii_wctomb(conv,buf,wc,1);
148: if (ret != RET_ILUNI) {
149: if (ret != 1) abort();
150: if (buf[0] < 0x80) {
151: int count = (state1 == STATE_ASCII ? 1 : 2);
152: if (n < count)
153: return RET_TOOSMALL;
154: if (state1 != STATE_ASCII) {
155: r[0] = SI;
156: r += 1;
157: state1 = STATE_ASCII;
158: }
159: r[0] = buf[0];
160: if (wc == 0x000a || wc == 0x000d)
161: state2 = STATE2_NONE;
162: COMBINE_STATE;
163: conv->ostate = state;
164: return count;
165: }
166: }
167:
168: /* Try KS C 5601-1992. */
169: ret = ksc5601_wctomb(conv,buf,wc,2);
170: if (ret != RET_ILUNI) {
171: if (ret != 2) abort();
172: if (buf[0] < 0x80 && buf[1] < 0x80) {
173: int count = (state2 == STATE2_DESIGNATED_KSC5601 ? 0 : 4) + (state1 == STATE_TWOBYTE ? 0 : 1) + 2;
174: if (n < count)
175: return RET_TOOSMALL;
176: if (state2 != STATE2_DESIGNATED_KSC5601) {
177: r[0] = ESC;
178: r[1] = '$';
179: r[2] = ')';
180: r[3] = 'C';
181: r += 4;
182: state2 = STATE2_DESIGNATED_KSC5601;
183: }
184: if (state1 != STATE_TWOBYTE) {
185: r[0] = SO;
186: r += 1;
187: state1 = STATE_TWOBYTE;
188: }
189: r[0] = buf[0];
190: r[1] = buf[1];
191: COMBINE_STATE;
192: conv->ostate = state;
193: return count;
194: }
195: }
196:
197: return RET_ILUNI;
198: }
199:
200: static int
1.1.1.2 ! misho 201: iso2022_kr_reset (conv_t conv, unsigned char *r, size_t n)
1.1 misho 202: {
203: state_t state = conv->ostate;
204: SPLIT_STATE;
205: (void)state2;
206: if (state1 != STATE_ASCII) {
207: if (n < 1)
208: return RET_TOOSMALL;
209: r[0] = SI;
210: /* conv->ostate = 0; will be done by the caller */
211: return 1;
212: } else
213: return 0;
214: }
215:
216: #undef COMBINE_STATE
217: #undef SPLIT_STATE
218: #undef STATE2_DESIGNATED_KSC5601
219: #undef STATE2_NONE
220: #undef STATE_TWOBYTE
221: #undef STATE_ASCII
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to iso2022_kr.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