2 * Copyright (c) 2004, 2006, 2007, 2008 Kungliga Tekniska Högskolan
3 * (Royal Institute of Technology, Stockholm, Sweden).
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
10 * 1. Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
17 * 3. Neither the name of the Institute nor the names of its contributors
18 * may be used to endorse or promote products derived from this software
19 * without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE INSTITUTE AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED. IN NO EVENT SHALL THE INSTITUTE OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
39 RCSID("$Id: utf8.c 23246 2008-06-01 22:29:04Z lha $");
42 utf8toutf32(const unsigned char **pp, uint32_t *out)
44 const unsigned char *p = *pp;
48 if ((c & 0xE0) == 0xC0) {
49 const unsigned c2 = *++p;
50 if ((c2 & 0xC0) == 0x80) {
51 *out = ((c & 0x1F) << 6)
54 return WIND_ERR_INVALID_UTF8;
56 } else if ((c & 0xF0) == 0xE0) {
57 const unsigned c2 = *++p;
58 if ((c2 & 0xC0) == 0x80) {
59 const unsigned c3 = *++p;
60 if ((c3 & 0xC0) == 0x80) {
61 *out = ((c & 0x0F) << 12)
65 return WIND_ERR_INVALID_UTF8;
68 return WIND_ERR_INVALID_UTF8;
70 } else if ((c & 0xF8) == 0xF0) {
71 const unsigned c2 = *++p;
72 if ((c2 & 0xC0) == 0x80) {
73 const unsigned c3 = *++p;
74 if ((c3 & 0xC0) == 0x80) {
75 const unsigned c4 = *++p;
76 if ((c4 & 0xC0) == 0x80) {
77 *out = ((c & 0x07) << 18)
82 return WIND_ERR_INVALID_UTF8;
85 return WIND_ERR_INVALID_UTF8;
88 return WIND_ERR_INVALID_UTF8;
91 return WIND_ERR_INVALID_UTF8;
103 * Convert an UTF-8 string to an UCS4 string.
105 * @param in an UTF-8 string to convert.
106 * @param out the resulting UCS4 strint, must be at least
107 * wind_utf8ucs4_length() long. If out is NULL, the function will
108 * calculate the needed space for the out variable (just like
109 * wind_utf8ucs4_length()).
110 * @param out_len before processing out_len should be the length of
111 * the out variable, after processing it will be the length of the out
114 * @return returns 0 on success, an wind error code otherwise
119 wind_utf8ucs4(const char *in, uint32_t *out, size_t *out_len)
121 const unsigned char *p;
125 for (p = (const unsigned char *)in; *p != '\0'; ++p) {
128 ret = utf8toutf32(&p, &u);
134 return WIND_ERR_OVERRUN;
144 * Calculate the length of from converting a UTF-8 string to a UCS4
147 * @param in an UTF-8 string to convert.
148 * @param out_len the length of the resulting UCS4 string.
150 * @return returns 0 on success, an wind error code otherwise
155 wind_utf8ucs4_length(const char *in, size_t *out_len)
157 return wind_utf8ucs4(in, NULL, out_len);
160 static const char first_char[4] =
161 { 0x00, 0xC0, 0xE0, 0xF0 };
164 * Convert an UCS4 string to a UTF-8 string.
166 * @param in an UCS4 string to convert.
167 * @param in_len the length input array.
169 * @param out the resulting UTF-8 strint, must be at least
170 * wind_ucs4utf8_length() + 1 long (the extra char for the NUL). If
171 * out is NULL, the function will calculate the needed space for the
172 * out variable (just like wind_ucs4utf8_length()).
174 * @param out_len before processing out_len should be the length of
175 * the out variable, after processing it will be the length of the out
178 * @return returns 0 on success, an wind error code otherwise
183 wind_ucs4utf8(const uint32_t *in, size_t in_len, char *out, size_t *out_len)
188 for (o = 0, i = 0; i < in_len; i++) {
193 } else if (ch < 0x800) {
195 } else if (ch < 0x10000) {
197 } else if (ch <= 0x10FFFF) {
200 return WIND_ERR_INVALID_UTF32;
206 return WIND_ERR_OVERRUN;
210 out[3] = (ch | 0x80) & 0xbf;
213 out[2] = (ch | 0x80) & 0xbf;
216 out[1] = (ch | 0x80) & 0xbf;
219 out[0] = ch | first_char[len - 1];
225 if (o + 1 >= *out_len)
226 return WIND_ERR_OVERRUN;
234 * Calculate the length of from converting a UCS4 string to an UTF-8 string.
236 * @param in an UCS4 string to convert.
237 * @param in_len the length of UCS4 string to convert.
238 * @param out_len the length of the resulting UTF-8 string.
240 * @return returns 0 on success, an wind error code otherwise
245 wind_ucs4utf8_length(const uint32_t *in, size_t in_len, size_t *out_len)
247 return wind_ucs4utf8(in, in_len, NULL, out_len);
251 * Read in an UCS2 from a buffer.
253 * @param ptr The input buffer to read from.
254 * @param len the length of the input buffer.
255 * @param flags Flags to control the behavior of the function.
256 * @param out the output UCS2, the array must be at least out/2 long.
257 * @param out_len the output length
259 * @return returns 0 on success, an wind error code otherwise.
264 wind_ucs2read(const void *ptr, size_t len, unsigned int *flags,
265 uint16_t *out, size_t *out_len)
267 const unsigned char *p = ptr;
268 int little = ((*flags) & WIND_RW_LE);
269 size_t olen = *out_len;
271 /** if len is zero, flags are unchanged */
277 /** if len is odd, WIND_ERR_LENGTH_NOT_MOD2 is returned */
279 return WIND_ERR_LENGTH_NOT_MOD2;
282 * If the flags WIND_RW_BOM is set, check for BOM. If not BOM is
283 * found, check is LE/BE flag is already and use that otherwise
284 * fail with WIND_ERR_NO_BOM. When done, clear WIND_RW_BOM and
285 * the LE/BE flag and set the resulting LE/BE flag.
287 if ((*flags) & WIND_RW_BOM) {
288 uint16_t bom = (p[0] << 8) + p[1];
289 if (bom == 0xfffe || bom == 0xfeff) {
290 little = (bom == 0xfffe);
293 } else if (((*flags) & (WIND_RW_LE|WIND_RW_BE)) != 0) {
294 /* little already set */
296 return WIND_ERR_NO_BOM;
297 *flags = ((*flags) & ~(WIND_RW_BOM|WIND_RW_LE|WIND_RW_BE));
298 *flags |= little ? WIND_RW_LE : WIND_RW_BE;
303 return WIND_ERR_OVERRUN;
305 *out = (p[1] << 8) + p[0];
307 *out = (p[0] << 8) + p[1];
308 out++; p += 2; len -= 2; olen--;
315 * Write an UCS2 string to a buffer.
317 * @param in The input UCS2 string.
318 * @param in_len the length of the input buffer.
319 * @param flags Flags to control the behavior of the function.
320 * @param ptr The input buffer to write to, the array must be at least
321 * (in + 1) * 2 bytes long.
322 * @param out_len the output length
324 * @return returns 0 on success, an wind error code otherwise.
329 wind_ucs2write(const uint16_t *in, size_t in_len, unsigned int *flags,
330 void *ptr, size_t *out_len)
332 unsigned char *p = ptr;
333 size_t len = *out_len;
335 /** If in buffer is not of length be mod 2, WIND_ERR_LENGTH_NOT_MOD2 is returned*/
337 return WIND_ERR_LENGTH_NOT_MOD2;
339 /** On zero input length, flags are preserved */
344 /** If flags have WIND_RW_BOM set, the byte order mark is written
345 * first to the output data */
346 if ((*flags) & WIND_RW_BOM) {
347 uint16_t bom = 0xfffe;
350 return WIND_ERR_OVERRUN;
352 if ((*flags) & WIND_RW_LE) {
353 p[0] = (bom >> 8) & 0xff;
354 p[1] = (bom ) & 0xff;
356 p[1] = (bom ) & 0xff;
357 p[0] = (bom >> 8) & 0xff;
363 /** If the output wont fit into out_len, WIND_ERR_OVERRUN is returned */
365 return WIND_ERR_OVERRUN;
366 if ((*flags) & WIND_RW_LE) {
367 p[0] = (in[0] >> 8) & 0xff;
368 p[1] = (in[0] ) & 0xff;
370 p[1] = (in[0] ) & 0xff;
371 p[0] = (in[0] >> 8) & 0xff;
384 * Convert an UTF-8 string to an UCS2 string.
386 * @param in an UTF-8 string to convert.
387 * @param out the resulting UCS2 strint, must be at least
388 * wind_utf8ucs2_length() long. If out is NULL, the function will
389 * calculate the needed space for the out variable (just like
390 * wind_utf8ucs2_length()).
391 * @param out_len before processing out_len should be the length of
392 * the out variable, after processing it will be the length of the out
395 * @return returns 0 on success, an wind error code otherwise
400 wind_utf8ucs2(const char *in, uint16_t *out, size_t *out_len)
402 const unsigned char *p;
406 for (p = (const unsigned char *)in; *p != '\0'; ++p) {
409 ret = utf8toutf32(&p, &u);
414 return WIND_ERR_NOT_UTF16;
418 return WIND_ERR_OVERRUN;
428 * Calculate the length of from converting a UTF-8 string to a UCS2
431 * @param in an UTF-8 string to convert.
432 * @param out_len the length of the resulting UCS4 string.
434 * @return returns 0 on success, an wind error code otherwise
439 wind_utf8ucs2_length(const char *in, size_t *out_len)
441 return wind_utf8ucs2(in, NULL, out_len);
445 * Convert an UCS2 string to a UTF-8 string.
447 * @param in an UCS2 string to convert.
448 * @param in_len the length of the in UCS2 string.
449 * @param out the resulting UTF-8 strint, must be at least
450 * wind_ucs2utf8_length() long. If out is NULL, the function will
451 * calculate the needed space for the out variable (just like
452 * wind_ucs2utf8_length()).
453 * @param out_len before processing out_len should be the length of
454 * the out variable, after processing it will be the length of the out
457 * @return returns 0 on success, an wind error code otherwise
462 wind_ucs2utf8(const uint16_t *in, size_t in_len, char *out, size_t *out_len)
467 for (o = 0, i = 0; i < in_len; i++) {
472 } else if (ch < 0x800) {
481 return WIND_ERR_OVERRUN;
485 out[2] = (ch | 0x80) & 0xbf;
488 out[1] = (ch | 0x80) & 0xbf;
491 out[0] = ch | first_char[len - 1];
498 return WIND_ERR_OVERRUN;
506 * Calculate the length of from converting a UCS2 string to an UTF-8 string.
508 * @param in an UCS2 string to convert.
509 * @param in_len an UCS2 string length to convert.
510 * @param out_len the length of the resulting UTF-8 string.
512 * @return returns 0 on success, an wind error code otherwise
517 wind_ucs2utf8_length(const uint16_t *in, size_t in_len, size_t *out_len)
519 return wind_ucs2utf8(in, in_len, NULL, out_len);