1 /* Character set conversion.
2 Copyright (C) 1999-2001, 2007, 2009-2014 Free Software Foundation, Inc.
4 This program is free software; you can redistribute it and/or modify
5 it under the terms of the GNU Lesser General Public License as published by
6 the Free Software Foundation; either version 2, or (at your option)
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU Lesser General Public License for more details.
14 You should have received a copy of the GNU Lesser General Public License along
15 with this program; if not, see <http://www.gnu.org/licenses/>. */
30 # define uintptr_t unsigned long
36 /* UTF-{16,32}{BE,LE} converters taken from GNU libiconv 1.11. */
38 /* Return code if invalid. (xxx_mbtowc) */
40 /* Return code if no bytes were read. (xxx_mbtowc) */
41 # define RET_TOOFEW -2
43 /* Return code if invalid. (xxx_wctomb) */
45 /* Return code if output buffer is too small. (xxx_wctomb, xxx_reset) */
46 # define RET_TOOSMALL -2
52 /* Specification: RFC 2781 */
55 utf16be_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
59 ucs4_t wc = (s[0] << 8) + s[1];
60 if (wc >= 0xd800 && wc < 0xdc00)
64 ucs4_t wc2 = (s[2] << 8) + s[3];
65 if (!(wc2 >= 0xdc00 && wc2 < 0xe000))
67 *pwc = 0x10000 + ((wc - 0xd800) << 10) + (wc2 - 0xdc00);
71 else if (wc >= 0xdc00 && wc < 0xe000)
85 utf16be_wctomb (unsigned char *r, ucs4_t wc, size_t n)
87 if (!(wc >= 0xd800 && wc < 0xe000))
93 r[0] = (unsigned char) (wc >> 8);
94 r[1] = (unsigned char) wc;
100 else if (wc < 0x110000)
104 ucs4_t wc1 = 0xd800 + ((wc - 0x10000) >> 10);
105 ucs4_t wc2 = 0xdc00 + ((wc - 0x10000) & 0x3ff);
106 r[0] = (unsigned char) (wc1 >> 8);
107 r[1] = (unsigned char) wc1;
108 r[2] = (unsigned char) (wc2 >> 8);
109 r[3] = (unsigned char) wc2;
123 /* Specification: RFC 2781 */
126 utf16le_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
130 ucs4_t wc = s[0] + (s[1] << 8);
131 if (wc >= 0xd800 && wc < 0xdc00)
135 ucs4_t wc2 = s[2] + (s[3] << 8);
136 if (!(wc2 >= 0xdc00 && wc2 < 0xe000))
138 *pwc = 0x10000 + ((wc - 0xd800) << 10) + (wc2 - 0xdc00);
142 else if (wc >= 0xdc00 && wc < 0xe000)
156 utf16le_wctomb (unsigned char *r, ucs4_t wc, size_t n)
158 if (!(wc >= 0xd800 && wc < 0xe000))
164 r[0] = (unsigned char) wc;
165 r[1] = (unsigned char) (wc >> 8);
171 else if (wc < 0x110000)
175 ucs4_t wc1 = 0xd800 + ((wc - 0x10000) >> 10);
176 ucs4_t wc2 = 0xdc00 + ((wc - 0x10000) & 0x3ff);
177 r[0] = (unsigned char) wc1;
178 r[1] = (unsigned char) (wc1 >> 8);
179 r[2] = (unsigned char) wc2;
180 r[3] = (unsigned char) (wc2 >> 8);
194 /* Specification: Unicode 3.1 Standard Annex #19 */
197 utf32be_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
201 ucs4_t wc = (s[0] << 24) + (s[1] << 16) + (s[2] << 8) + s[3];
202 if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
214 utf32be_wctomb (unsigned char *r, ucs4_t wc, size_t n)
216 if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
221 r[1] = (unsigned char) (wc >> 16);
222 r[2] = (unsigned char) (wc >> 8);
223 r[3] = (unsigned char) wc;
236 /* Specification: Unicode 3.1 Standard Annex #19 */
239 utf32le_mbtowc (ucs4_t *pwc, const unsigned char *s, size_t n)
243 ucs4_t wc = s[0] + (s[1] << 8) + (s[2] << 16) + (s[3] << 24);
244 if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
256 utf32le_wctomb (unsigned char *r, ucs4_t wc, size_t n)
258 if (wc < 0x110000 && !(wc >= 0xd800 && wc < 0xe000))
262 r[0] = (unsigned char) wc;
263 r[1] = (unsigned char) (wc >> 8);
264 r[2] = (unsigned char) (wc >> 16);
277 rpl_iconv (iconv_t cd,
278 ICONV_CONST char **inbuf, size_t *inbytesleft,
279 char **outbuf, size_t *outbytesleft)
282 #if REPLACE_ICONV_UTF
283 switch ((uintptr_t) cd)
286 int (*xxx_wctomb) (unsigned char *, ucs4_t, size_t);
288 case (uintptr_t) _ICONV_UTF8_UTF16BE:
289 xxx_wctomb = utf16be_wctomb;
291 case (uintptr_t) _ICONV_UTF8_UTF16LE:
292 xxx_wctomb = utf16le_wctomb;
294 case (uintptr_t) _ICONV_UTF8_UTF32BE:
295 xxx_wctomb = utf32be_wctomb;
297 case (uintptr_t) _ICONV_UTF8_UTF32LE:
298 xxx_wctomb = utf32le_wctomb;
302 if (inbuf == NULL || *inbuf == NULL)
305 ICONV_CONST char *inptr = *inbuf;
306 size_t inleft = *inbytesleft;
307 char *outptr = *outbuf;
308 size_t outleft = *outbytesleft;
313 int m = u8_mbtoucr (&uc, (const uint8_t *) inptr, inleft);
332 int n = xxx_wctomb ((uint8_t *) outptr, uc, outleft);
341 if (n == RET_TOOSMALL)
359 *inbytesleft = inleft;
361 *outbytesleft = outleft;
367 int (*xxx_mbtowc) (ucs4_t *, const unsigned char *, size_t);
369 case (uintptr_t) _ICONV_UTF16BE_UTF8:
370 xxx_mbtowc = utf16be_mbtowc;
372 case (uintptr_t) _ICONV_UTF16LE_UTF8:
373 xxx_mbtowc = utf16le_mbtowc;
375 case (uintptr_t) _ICONV_UTF32BE_UTF8:
376 xxx_mbtowc = utf32be_mbtowc;
378 case (uintptr_t) _ICONV_UTF32LE_UTF8:
379 xxx_mbtowc = utf32le_mbtowc;
383 if (inbuf == NULL || *inbuf == NULL)
386 ICONV_CONST char *inptr = *inbuf;
387 size_t inleft = *inbytesleft;
388 char *outptr = *outbuf;
389 size_t outleft = *outbytesleft;
394 int m = xxx_mbtowc (&uc, (const uint8_t *) inptr, inleft);
413 int n = u8_uctomb ((uint8_t *) outptr, uc, outleft);
440 *inbytesleft = inleft;
442 *outbytesleft = outleft;
448 return iconv (cd, inbuf, inbytesleft, outbuf, outbytesleft);