blob: 2c424054e628480531f4a1162fcfb882380d1c29 [file] [log] [blame]
/*
* Copyright (C) 1999-2001, 2008, 2011-2012, 2016, 2018 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 Lesser General Public
* License as published by the Free Software Foundation; either version 2.1
* 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with the GNU LIBICONV Library; see the file COPYING.LIB.
* If not, see <https://www.gnu.org/licenses/>.
*/
/*
* ISO-2022-JP-MS
* alias CP50221
*
* This is an extension of ISO-2022-JP-1 with larger character sets.
* It uses ESC $ B and ESC $ ( D to denote *extensions* of JIS X 0208 and
* JIS X 0212, respectively. This violates the principles of ISO 2022,
* where
* 1. character sets to be used by ISO 2022 have to be registered at the
* ISO IR registry <https://www.itscj.ipsj.or.jp/ISO-IR/>,
* 2. different character sets are designated by different escape
* sequences.
* It's a typical instance of the "embrace and extend" strategy by Microsoft
* <https://en.wikipedia.org/wiki/Embrace,_extend_and_extinguish>.
*/
/*
* Windows has three encodings CP50220, CP50221, CP50222.
* The common parts are:
* - US-ASCII (0x00..0x7F)
* - JIS X 0208 extended by
* - one row (0x2D),
* - a private use area (rows 0x75..0x7E = U+E000..U+E3AB),
* enabled with ESC $ B, disabled with ESC ( B.
* - JIS X 0212 extended by
* - two rows (0x73..0x74),
* - a private use area (rows 0x75..0x7E = U+E3AC..U+E757),
* enabled with ESC $ ( D, disabled with ESC ( B.
* They differ in the handling of JIS X 0201 characters (halfwidth Katakana)
* in the conversion direction Unicode -> CP5022x:
* * CP50220 maps the halfwidth Katakana to fullwidth Katakana characters.
* * CP50221 contains the JIS X 0201 halfwidth Katakana characters,
* enabled with ESC ( I, disabled with ESC ( B.
* * CP50222 contains the JIS X 0201 halfwidth Katakana characters,
* enabled with ESC ( J 0x0E, disabled with ESC ( B.
* In the conversion direction CP5022x -> Unicode, all three operate the same:
* - ESC ( I is supported and understood.
* - ESC ( J 0x0E is not accepted. (Tested on Windows XP SP3.)
* Conclusion:
* - CP50222 should not be used, because the multibyte sequence that it
* produces cannot be parsed by either of the three encodings.
* - CP50221 is preferrable to CP50220, because it can faithfully represent
* the halfwidth Katakana characters.
* We therefore implement CP50221. As an extension, in the mbtowc conversion
* direction, we support also ESC ( J 0x0E, just in case.
*/
#include "cp50221_0208_ext.h"
#include "cp50221_0212_ext.h"
#define ESC 0x1b
#define SO 0x0e
#define SI 0x0f
/*
* The state can be one of the following values.
*/
#define STATE_ASCII 0 /* Esc ( B */
#define STATE_JISX0201ROMAN 1 /* Esc ( J */ /* only in mbtowc direction */
#define STATE_JISX0201KATAKANA 2 /* Esc ( I */
#define STATE_JISX0208MS 3 /* Esc $ @ or Esc $ B */
#define STATE_JISX0212MS 4 /* Esc $ ( D */
static int
iso2022_jpms_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)
{
state_t state = conv->istate;
int count = 0;
unsigned char c;
for (;;) {
c = *s;
if (c == ESC) {
if (n < count+3)
goto none;
if (s[1] == '(') {
if (s[2] == 'B') {
state = STATE_ASCII;
s += 3; count += 3;
if (n < count+1)
goto none;
continue;
}
if (s[2] == 'I') {
state = STATE_JISX0201KATAKANA;
s += 3; count += 3;
if (n < count+1)
goto none;
continue;
}
if (s[2] == 'J') {
state = STATE_JISX0201ROMAN;
s += 3; count += 3;
if (n < count+1)
goto none;
continue;
}
goto ilseq;
}
if (s[1] == '$') {
if (s[2] == '@' || s[2] == 'B') {
/* We don't distinguish JIS X 0208-1978 and JIS X 0208-1983. */
state = STATE_JISX0208MS;
s += 3; count += 3;
if (n < count+1)
goto none;
continue;
}
if (s[2] == '(') {
if (n < count+4)
goto none;
if (s[3] == 'D') {
state = STATE_JISX0212MS;
s += 4; count += 4;
if (n < count+1)
goto none;
continue;
}
}
goto ilseq;
}
goto ilseq;
}
if (c == SO) {
if (state == STATE_JISX0201ROMAN)
state = STATE_JISX0201KATAKANA;
s += 1; count += 1;
if (n < count+1)
goto none;
continue;
}
if (c == SI) {
if (state == STATE_JISX0201KATAKANA)
state = STATE_JISX0201ROMAN;
s += 1; count += 1;
if (n < count+1)
goto none;
continue;
}
break;
}
switch (state) {
case STATE_ASCII:
if (c < 0x80) {
int ret = ascii_mbtowc(conv,pwc,s,1);
if (ret == RET_ILSEQ)
goto ilseq;
if (ret != 1) abort();
conv->istate = state;
return count+1;
} else
goto ilseq;
case STATE_JISX0201ROMAN:
if (c < 0x80) {
int ret = jisx0201_mbtowc(conv,pwc,s,1);
if (ret == RET_ILSEQ)
goto ilseq;
if (ret != 1) abort();
conv->istate = state;
return count+1;
} else
goto ilseq;
case STATE_JISX0201KATAKANA:
if (c < 0x80) {
unsigned char buf = c+0x80;
int ret = jisx0201_mbtowc(conv,pwc,&buf,1);
if (ret == RET_ILSEQ)
goto ilseq;
if (ret != 1) abort();
conv->istate = state;
return count+1;
} else
goto ilseq;
case STATE_JISX0208MS:
if (n < count+2)
goto none;
if (s[0] < 0x80 && s[1] < 0x80) {
int ret;
if (s[0] < 0x75) {
if (s[0] == 0x2d) {
/* Extension of JIS X 0208. */
if (s[1] >= 0x21 && s[1] <= 0x79) {
unsigned char i = (s[1] - 0x21) + 1;
ret = cp50221_0208_ext_mbtowc(conv,pwc,&i,1);
if (ret == 1)
ret = 2;
} else
ret = RET_ILSEQ;
} else {
/* JIS X 0208. */
ret = jisx0208_mbtowc(conv,pwc,s,2);
}
} else {
/* Extension of JIS X 0208.
0x{75..7E}{21..8E} maps to U+E000..U+E3AB.
But some rows maps to characters present in CP932. */
if (s[0] <= 0x7e && (s[1] >= 0x21 && s[1] <= 0x7e)) {
unsigned short wc = 0xfffd;
if (s[0] >= 0x79 && s[0] <= 0x7c)
wc = cp932ext_2uni_pageed[(s[0] - 0x79) * 94 + (s[1] - 0x21)];
if (wc == 0xfffd)
wc = (s[0] - 0x75) * 94 + (s[1] - 0x21) + 0xe000;
*pwc = wc;
ret = 2;
} else
ret = RET_ILSEQ;
}
if (ret == RET_ILSEQ)
goto ilseq;
if (ret != 2) abort();
conv->istate = state;
return count+2;
} else
goto ilseq;
case STATE_JISX0212MS:
if (n < count+2)
goto none;
if (s[0] < 0x80 && s[1] < 0x80) {
int ret;
if (s[0] < 0x73) {
/* JIS X 0212. */
ret = jisx0212_mbtowc(conv,pwc,s,2);
} else {
if (s[0] < 0x75) {
/* Extension of JIS X 0212. */
if (s[1] >= 0x21 && s[1] <= 0x7e) {
unsigned char i = (s[0] - 0x73) * 94 + (s[1] - 0x21) + 1;
ret = cp50221_0212_ext_mbtowc(conv,pwc,&i,1);
if (ret == 1)
ret = 2;
} else
ret = RET_ILSEQ;
} else {
/* Extension of JIS X 0208.
0x{75..7E}{21..8E} maps to U+E3AC..U+E757. */
if (s[0] <= 0x7e && (s[1] >= 0x21 && s[1] <= 0x7e)) {
*pwc = (s[0] - 0x75) * 94 + (s[1] - 0x21) + 0xe3ac;
ret = 2;
} else
ret = RET_ILSEQ;
}
}
if (ret == RET_ILSEQ)
goto ilseq;
if (ret != 2) abort();
conv->istate = state;
return count+2;
} else
goto ilseq;
default: abort();
}
none:
conv->istate = state;
return RET_TOOFEW(count);
ilseq:
conv->istate = state;
return RET_SHIFT_ILSEQ(count);
}
static int
iso2022_jpms_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)
{
state_t state = conv->ostate;
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 = (state == STATE_ASCII ? 1 : 4);
if (n < count)
return RET_TOOSMALL;
if (state != STATE_ASCII) {
r[0] = ESC;
r[1] = '(';
r[2] = 'B';
r += 3;
state = STATE_ASCII;
}
r[0] = buf[0];
conv->ostate = state;
return count;
}
}
/* Try JIS X 0201-1976 Katakana. */
ret = jisx0201_wctomb(conv,buf,wc,1);
if (ret != RET_ILUNI) {
if (ret != 1) abort();
if (buf[0] >= 0x80) {
int count = (state == STATE_JISX0201KATAKANA ? 1 : 4);
if (n < count)
return RET_TOOSMALL;
if (state != STATE_JISX0201KATAKANA) {
r[0] = ESC;
r[1] = '(';
r[2] = 'I';
r += 3;
state = STATE_JISX0201KATAKANA;
}
r[0] = buf[0]-0x80;
conv->ostate = state;
return count;
}
}
/* Try JIS X 0208-1990, in place of JIS X 0208-1978 and JIS X 0208-1983,
and the extensions mentioned above. */
if (wc >= 0xe000 && wc < 0xe3ac) {
unsigned short i = wc - 0xe000;
buf[0] = (i / 94) + 0x75;
buf[1] = (i % 94) + 0x21;
ret = 2;
} else {
ret = jisx0208_wctomb(conv,buf,wc,2);
if (ret == RET_ILUNI) {
/* Extension of JIS X 0208. */
unsigned char i;
ret = cp50221_0208_ext_wctomb(conv,&i,wc,1);
if (ret == 1) {
buf[0] = 0x2d;
buf[1] = i-1 + 0x21;
ret = 2;
} else if (wc == 0x663B) {
buf[0] = 0x7a;
buf[1] = 0x36;
ret = 2;
} else if (wc == 0xffe2) {
buf[0] = 0x7c;
buf[1] = 0x7b;
ret = 2;
} else if (wc == 0xffe4) {
buf[0] = 0x7c;
buf[1] = 0x7c;
ret = 2;
}
}
}
if (ret != RET_ILUNI) {
if (ret != 2) abort();
if (buf[0] < 0x80 && buf[1] < 0x80) {
int count = (state == STATE_JISX0208MS ? 2 : 5);
if (n < count)
return RET_TOOSMALL;
if (state != STATE_JISX0208MS) {
r[0] = ESC;
r[1] = '$';
r[2] = 'B';
r += 3;
state = STATE_JISX0208MS;
}
r[0] = buf[0];
r[1] = buf[1];
conv->ostate = state;
return count;
}
}
/* Try JIS X 0212-1990 and the extensions mentioned above. */
if (wc >= 0xe3ac && wc < 0xe758) {
unsigned short i = wc - 0xe3ac;
buf[0] = (i / 94) + 0x75;
buf[1] = (i % 94) + 0x21;
ret = 2;
} else {
ret = jisx0212_wctomb(conv,buf,wc,2);
if (ret == RET_ILUNI) {
/* Extension of JIS X 0212. */
unsigned char i;
ret = cp50221_0212_ext_wctomb(conv,&i,wc,1);
if (ret == 1) {
i -= 1;
buf[0] = (i / 94) + 0x73;
buf[1] = (i % 94) + 0x21;
ret = 2;
}
}
}
if (ret != RET_ILUNI) {
if (ret != 2) abort();
if (buf[0] < 0x80 && buf[1] < 0x80) {
int count = (state == STATE_JISX0212MS ? 2 : 6);
if (n < count)
return RET_TOOSMALL;
if (state != STATE_JISX0212MS) {
r[0] = ESC;
r[1] = '$';
r[2] = '(';
r[3] = 'D';
r += 4;
state = STATE_JISX0212MS;
}
r[0] = buf[0];
r[1] = buf[1];
conv->ostate = state;
return count;
}
}
return RET_ILUNI;
}
static int
iso2022_jpms_reset (conv_t conv, unsigned char *r, size_t n)
{
state_t state = conv->ostate;
if (state != STATE_ASCII) {
if (n < 3)
return RET_TOOSMALL;
r[0] = ESC;
r[1] = '(';
r[2] = 'B';
/* conv->ostate = 0; will be done by the caller */
return 3;
} else
return 0;
}
#undef STATE_JISX0212MS
#undef STATE_JISX0208MS
#undef STATE_JISX0201KATAKANA
#undef STATE_JISX0201ROMAN
#undef STATE_ASCII