mirror of
https://git.lyx.org/repos/lyx.git
synced 2024-12-25 05:55:34 +00:00
222 lines
5.3 KiB
C
222 lines
5.3 KiB
C
/*
|
|
* Copyright (C) 1999-2001, 2008, 2016 Free Software Foundation, Inc.
|
|
* This file is part of the GNU LIBICONV Library.
|
|
*
|
|
* The GNU LIBICONV Library is free software; you can redistribute it
|
|
* and/or modify it under the terms of the GNU Library General Public
|
|
* License as published by the Free Software Foundation; either version 2
|
|
* of the License, or (at your option) any later version.
|
|
*
|
|
* The GNU LIBICONV Library is distributed in the hope that it will be
|
|
* useful, but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
* Library General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU Library General Public
|
|
* License along with the GNU LIBICONV Library; see the file COPYING.LIB.
|
|
* If not, see <http://www.gnu.org/licenses/>.
|
|
*/
|
|
|
|
/*
|
|
* ISO-2022-KR
|
|
*/
|
|
|
|
/* Specification: RFC 1557 */
|
|
|
|
/* Note: CJK.INF says the SO designator needs to appear only once at the
|
|
beginning of a text, but to decrease the risk of ambiguities, when
|
|
producing ISO-2022-KR, we repeat the designator in every line containing
|
|
SO characters. RFC 1557 does not mandate this. */
|
|
|
|
#define ESC 0x1b
|
|
#define SO 0x0e
|
|
#define SI 0x0f
|
|
|
|
/*
|
|
* The state is composed of one of the following values
|
|
*/
|
|
#define STATE_ASCII 0
|
|
#define STATE_TWOBYTE 1
|
|
/*
|
|
* and one of the following values, << 8
|
|
*/
|
|
#define STATE2_NONE 0
|
|
#define STATE2_DESIGNATED_KSC5601 1
|
|
|
|
#define SPLIT_STATE \
|
|
unsigned int state1 = state & 0xff, state2 = state >> 8
|
|
#define COMBINE_STATE \
|
|
state = (state2 << 8) | state1
|
|
|
|
static int
|
|
iso2022_kr_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
|
|
{
|
|
state_t state = conv->istate;
|
|
SPLIT_STATE;
|
|
int count = 0;
|
|
unsigned char c;
|
|
for (;;) {
|
|
c = *s;
|
|
if (c == ESC) {
|
|
if (n < count+4)
|
|
goto none;
|
|
if (s[1] == '$') {
|
|
if (s[2] == ')') {
|
|
if (s[3] == 'C') {
|
|
state2 = STATE2_DESIGNATED_KSC5601;
|
|
s += 4; count += 4;
|
|
if (n < count+1)
|
|
goto none;
|
|
continue;
|
|
}
|
|
}
|
|
}
|
|
goto ilseq;
|
|
}
|
|
if (c == SO) {
|
|
if (state2 != STATE2_DESIGNATED_KSC5601)
|
|
goto ilseq;
|
|
state1 = STATE_TWOBYTE;
|
|
s++; count++;
|
|
if (n < count+1)
|
|
goto none;
|
|
continue;
|
|
}
|
|
if (c == SI) {
|
|
state1 = STATE_ASCII;
|
|
s++; count++;
|
|
if (n < count+1)
|
|
goto none;
|
|
continue;
|
|
}
|
|
break;
|
|
}
|
|
switch (state1) {
|
|
case STATE_ASCII:
|
|
if (c < 0x80) {
|
|
int ret = ascii_mbtowc(conv,pwc,s,1);
|
|
if (ret == RET_ILSEQ)
|
|
goto ilseq;
|
|
if (ret != 1) abort();
|
|
#if 0 /* Accept ISO-2022-KR according to CJK.INF. */
|
|
if (*pwc == 0x000a || *pwc == 0x000d)
|
|
state2 = STATE2_NONE;
|
|
#endif
|
|
COMBINE_STATE;
|
|
conv->istate = state;
|
|
return count+1;
|
|
} else
|
|
goto ilseq;
|
|
case STATE_TWOBYTE:
|
|
if (n < count+2)
|
|
goto none;
|
|
if (state2 != STATE2_DESIGNATED_KSC5601) abort();
|
|
if (s[0] < 0x80 && s[1] < 0x80) {
|
|
int ret = ksc5601_mbtowc(conv,pwc,s,2);
|
|
if (ret == RET_ILSEQ)
|
|
goto ilseq;
|
|
if (ret != 2) abort();
|
|
COMBINE_STATE;
|
|
conv->istate = state;
|
|
return count+2;
|
|
} else
|
|
goto ilseq;
|
|
default: abort();
|
|
}
|
|
|
|
none:
|
|
COMBINE_STATE;
|
|
conv->istate = state;
|
|
return RET_TOOFEW(count);
|
|
|
|
ilseq:
|
|
COMBINE_STATE;
|
|
conv->istate = state;
|
|
return RET_SHIFT_ILSEQ(count);
|
|
}
|
|
|
|
static int
|
|
iso2022_kr_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
|
|
{
|
|
state_t state = conv->ostate;
|
|
SPLIT_STATE;
|
|
unsigned char buf[2];
|
|
int ret;
|
|
|
|
/* Try ASCII. */
|
|
ret = ascii_wctomb(conv,buf,wc,1);
|
|
if (ret != RET_ILUNI) {
|
|
if (ret != 1) abort();
|
|
if (buf[0] < 0x80) {
|
|
int count = (state1 == STATE_ASCII ? 1 : 2);
|
|
if (n < count)
|
|
return RET_TOOSMALL;
|
|
if (state1 != STATE_ASCII) {
|
|
r[0] = SI;
|
|
r += 1;
|
|
state1 = STATE_ASCII;
|
|
}
|
|
r[0] = buf[0];
|
|
if (wc == 0x000a || wc == 0x000d)
|
|
state2 = STATE2_NONE;
|
|
COMBINE_STATE;
|
|
conv->ostate = state;
|
|
return count;
|
|
}
|
|
}
|
|
|
|
/* Try KS C 5601-1992. */
|
|
ret = ksc5601_wctomb(conv,buf,wc,2);
|
|
if (ret != RET_ILUNI) {
|
|
if (ret != 2) abort();
|
|
if (buf[0] < 0x80 && buf[1] < 0x80) {
|
|
int count = (state2 == STATE2_DESIGNATED_KSC5601 ? 0 : 4) + (state1 == STATE_TWOBYTE ? 0 : 1) + 2;
|
|
if (n < count)
|
|
return RET_TOOSMALL;
|
|
if (state2 != STATE2_DESIGNATED_KSC5601) {
|
|
r[0] = ESC;
|
|
r[1] = '$';
|
|
r[2] = ')';
|
|
r[3] = 'C';
|
|
r += 4;
|
|
state2 = STATE2_DESIGNATED_KSC5601;
|
|
}
|
|
if (state1 != STATE_TWOBYTE) {
|
|
r[0] = SO;
|
|
r += 1;
|
|
state1 = STATE_TWOBYTE;
|
|
}
|
|
r[0] = buf[0];
|
|
r[1] = buf[1];
|
|
COMBINE_STATE;
|
|
conv->ostate = state;
|
|
return count;
|
|
}
|
|
}
|
|
|
|
return RET_ILUNI;
|
|
}
|
|
|
|
static int
|
|
iso2022_kr_reset (conv_t conv, unsigned char *r, size_t n)
|
|
{
|
|
state_t state = conv->ostate;
|
|
SPLIT_STATE;
|
|
(void)state2;
|
|
if (state1 != STATE_ASCII) {
|
|
if (n < 1)
|
|
return RET_TOOSMALL;
|
|
r[0] = SI;
|
|
/* conv->ostate = 0; will be done by the caller */
|
|
return 1;
|
|
} else
|
|
return 0;
|
|
}
|
|
|
|
#undef COMBINE_STATE
|
|
#undef SPLIT_STATE
|
|
#undef STATE2_DESIGNATED_KSC5601
|
|
#undef STATE2_NONE
|
|
#undef STATE_TWOBYTE
|
|
#undef STATE_ASCII
|