1 /* gutf8.c - Operations on UTF-8 strings.
3 * Copyright (C) 1999 Tom Tromey
4 * Copyright (C) 2000 Red Hat, Inc.
6 * This library is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2 of the License, or (at your option) any later version.
11 * This library is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with this library; if not, write to the
18 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
19 * Boston, MA 02111-1307, USA.
32 #ifdef G_PLATFORM_WIN32
39 #include "libcharset/libcharset.h"
44 #define UTF8_COMPUTE(Char, Mask, Len) \
50 else if ((Char & 0xe0) == 0xc0) \
55 else if ((Char & 0xf0) == 0xe0) \
60 else if ((Char & 0xf8) == 0xf0) \
65 else if ((Char & 0xfc) == 0xf8) \
70 else if ((Char & 0xfe) == 0xfc) \
78 #define UTF8_LENGTH(Char) \
79 ((Char) < 0x80 ? 1 : \
80 ((Char) < 0x800 ? 2 : \
81 ((Char) < 0x10000 ? 3 : \
82 ((Char) < 0x200000 ? 4 : \
83 ((Char) < 0x4000000 ? 5 : 6)))))
86 #define UTF8_GET(Result, Chars, Count, Mask, Len) \
87 (Result) = (Chars)[0] & (Mask); \
88 for ((Count) = 1; (Count) < (Len); ++(Count)) \
90 if (((Chars)[(Count)] & 0xc0) != 0x80) \
96 (Result) |= ((Chars)[(Count)] & 0x3f); \
99 #define UNICODE_VALID(Char) \
100 ((Char) < 0x110000 && \
101 (((Char) & 0xFFFFF800) != 0xD800) && \
102 ((Char) < 0xFDD0 || (Char) > 0xFDEF) && \
103 ((Char) & 0xFFFE) != 0xFFFE)
106 static const gchar utf8_skip_data[256] = {
107 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
108 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
109 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
110 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
111 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
112 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
113 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
114 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6,1,1
117 const gchar * const g_utf8_skip = utf8_skip_data;
120 * g_utf8_find_prev_char:
121 * @str: pointer to the beginning of a UTF-8 encoded string
122 * @p: pointer to some position within @str
124 * Given a position @p with a UTF-8 encoded string @str, find the start
125 * of the previous UTF-8 character starting before @p. Returns %NULL if no
126 * UTF-8 characters are present in @str before @p.
128 * @p does not have to be at the beginning of a UTF-8 character. No check
129 * is made to see if the character found is actually valid other than
130 * it starts with an appropriate byte.
132 * Return value: a pointer to the found character or %NULL.
135 g_utf8_find_prev_char (const char *str,
138 for (--p; p >= str; --p)
140 if ((*p & 0xc0) != 0x80)
147 * g_utf8_find_next_char:
148 * @p: a pointer to a position within a UTF-8 encoded string
149 * @end: a pointer to the end of the string, or %NULL to indicate
150 * that the string is nul-terminated, in which case
151 * the returned value will be
153 * Finds the start of the next UTF-8 character in the string after @p.
155 * @p does not have to be at the beginning of a UTF-8 character. No check
156 * is made to see if the character found is actually valid other than
157 * it starts with an appropriate byte.
159 * Return value: a pointer to the found character or %NULL
162 g_utf8_find_next_char (const gchar *p,
168 for (++p; p < end && (*p & 0xc0) == 0x80; ++p)
171 for (++p; (*p & 0xc0) == 0x80; ++p)
174 return (p == end) ? NULL : (gchar *)p;
179 * @p: a pointer to a position within a UTF-8 encoded string
181 * Finds the previous UTF-8 character in the string before @p.
183 * @p does not have to be at the beginning of a UTF-8 character. No check
184 * is made to see if the character found is actually valid other than
185 * it starts with an appropriate byte. If @p might be the first
186 * character of the string, you must use g_utf8_find_prev_char() instead.
188 * Return value: a pointer to the found character.
191 g_utf8_prev_char (const gchar *p)
196 if ((*p & 0xc0) != 0x80)
203 * @p: pointer to the start of a UTF-8 encoded string.
204 * @max: the maximum number of bytes to examine. If @max
205 * is less than 0, then the string is assumed to be
206 * nul-terminated. If @max is 0, @p will not be examined and
209 * Returns the length of the string in characters.
211 * Return value: the length of the string in characters
214 g_utf8_strlen (const gchar *p,
218 const gchar *start = p;
219 g_return_val_if_fail (p != NULL || max == 0, 0);
225 p = g_utf8_next_char (p);
234 p = g_utf8_next_char (p);
236 while (p - start < max && *p)
239 p = g_utf8_next_char (p);
242 /* only do the last len increment if we got a complete
243 * char (don't count partial chars)
245 if (p - start <= max)
254 * @p: a pointer to Unicode character encoded as UTF-8
256 * Converts a sequence of bytes encoded as UTF-8 to a Unicode character.
257 * If @p does not point to a valid UTF-8 encoded character, results are
258 * undefined. If you are not sure that the bytes are complete
259 * valid Unicode characters, you should use g_utf8_get_char_validated()
262 * Return value: the resulting character
265 g_utf8_get_char (const gchar *p)
267 int i, mask = 0, len;
269 unsigned char c = (unsigned char) *p;
271 UTF8_COMPUTE (c, mask, len);
274 UTF8_GET (result, p, i, mask, len);
280 * g_utf8_offset_to_pointer:
281 * @str: a UTF-8 encoded string
282 * @offset: a character offset within @str
284 * Converts from an integer character offset to a pointer to a position
287 * Since 2.10, this function allows to pass a negative @offset to
288 * step backwards. It is usually worth stepping backwards from the end
289 * instead of forwards if @offset is in the last fourth of the string,
290 * since moving forward is about 3 times faster than moving backward.
292 * Return value: the resulting pointer
295 g_utf8_offset_to_pointer (const gchar *str,
298 const gchar *s = str;
302 s = g_utf8_next_char (s);
307 /* This nice technique for fast backwards stepping
308 * through a UTF-8 string was dubbed "stutter stepping"
309 * by its inventor, Larry Ewing.
315 while ((*s & 0xc0) == 0x80)
318 offset += g_utf8_pointer_to_offset (s, s1);
326 * g_utf8_pointer_to_offset:
327 * @str: a UTF-8 encoded string
328 * @pos: a pointer to a position within @str
330 * Converts from a pointer to position within a string to a integer
333 * Since 2.10, this function allows @pos to be before @str, and returns
334 * a negative offset in this case.
336 * Return value: the resulting character offset
339 g_utf8_pointer_to_offset (const gchar *str,
342 const gchar *s = str;
346 offset = - g_utf8_pointer_to_offset (pos, str);
350 s = g_utf8_next_char (s);
360 * @dest: buffer to fill with characters from @src
361 * @src: UTF-8 encoded string
362 * @n: character count
364 * Like the standard C strncpy() function, but
365 * copies a given number of characters instead of a given number of
366 * bytes. The @src string must be valid UTF-8 encoded text.
367 * (Use g_utf8_validate() on all text before trying to use UTF-8
368 * utility functions with it.)
370 * Return value: @dest
373 g_utf8_strncpy (gchar *dest,
377 const gchar *s = src;
380 s = g_utf8_next_char(s);
383 strncpy(dest, src, s - src);
388 G_LOCK_DEFINE_STATIC (aliases);
391 get_alias_hash (void)
393 static GHashTable *alias_hash = NULL;
400 alias_hash = g_hash_table_new (g_str_hash, g_str_equal);
402 aliases = _g_locale_get_charset_aliases ();
403 while (*aliases != '\0')
405 const char *canonical;
407 const char **alias_array;
411 aliases += strlen (aliases) + 1;
413 aliases += strlen (aliases) + 1;
415 alias_array = g_hash_table_lookup (alias_hash, canonical);
418 while (alias_array[count])
422 alias_array = g_renew (const char *, alias_array, count + 2);
423 alias_array[count] = alias;
424 alias_array[count + 1] = NULL;
426 g_hash_table_insert (alias_hash, (char *)canonical, alias_array);
435 /* As an abuse of the alias table, the following routines gets
436 * the charsets that are aliases for the canonical name.
438 G_GNUC_INTERNAL const char **
439 _g_charset_get_aliases (const char *canonical_name)
441 GHashTable *alias_hash = get_alias_hash ();
443 return g_hash_table_lookup (alias_hash, canonical_name);
447 g_utf8_get_charset_internal (const char *raw_data,
450 const char *charset = getenv("CHARSET");
452 if (charset && *charset)
456 if (charset && strstr (charset, "UTF-8"))
462 /* The libcharset code tries to be thread-safe without
463 * a lock, but has a memory leak and a missing memory
464 * barrier, so we lock for it
467 charset = _g_locale_charset_unalias (raw_data);
470 if (charset && *charset)
474 if (charset && strstr (charset, "UTF-8"))
480 /* Assume this for compatibility at present. */
486 typedef struct _GCharsetCache GCharsetCache;
488 struct _GCharsetCache {
495 charset_cache_free (gpointer data)
497 GCharsetCache *cache = data;
499 g_free (cache->charset);
505 * @charset: return location for character set name
507 * Obtains the character set for the <link linkend="setlocale">current
508 * locale</link>; you might use this character set as an argument to
509 * g_convert(), to convert from the current locale's encoding to some
510 * other encoding. (Frequently g_locale_to_utf8() and g_locale_from_utf8()
511 * are nice shortcuts, though.)
513 * The return value is %TRUE if the locale's encoding is UTF-8, in that
514 * case you can perhaps avoid calling g_convert().
516 * The string returned in @charset is not allocated, and should not be
519 * Return value: %TRUE if the returned charset is UTF-8
522 g_get_charset (G_CONST_RETURN char **charset)
524 static GStaticPrivate cache_private = G_STATIC_PRIVATE_INIT;
525 GCharsetCache *cache = g_static_private_get (&cache_private);
530 cache = g_new0 (GCharsetCache, 1);
531 g_static_private_set (&cache_private, cache, charset_cache_free);
534 raw = _g_locale_charset_raw ();
536 if (!(cache->raw && strcmp (cache->raw, raw) == 0))
538 const gchar *new_charset;
541 g_free (cache->charset);
542 cache->raw = g_strdup (raw);
543 cache->is_utf8 = g_utf8_get_charset_internal (raw, &new_charset);
544 cache->charset = g_strdup (new_charset);
548 *charset = cache->charset;
550 return cache->is_utf8;
557 * @c: a Unicode character code
558 * @outbuf: output buffer, must have at least 6 bytes of space.
559 * If %NULL, the length will be computed and returned
560 * and nothing will be written to @outbuf.
562 * Converts a single character to UTF-8.
564 * Return value: number of bytes written
567 g_unichar_to_utf8 (gunichar c,
570 /* If this gets modified, also update the copy in g_string_insert_unichar() */
585 else if (c < 0x10000)
590 else if (c < 0x200000)
595 else if (c < 0x4000000)
608 for (i = len - 1; i > 0; --i)
610 outbuf[i] = (c & 0x3f) | 0x80;
613 outbuf[0] = c | first;
621 * @p: a nul-terminated UTF-8 encoded string
622 * @len: the maximum length of @p
623 * @c: a Unicode character
625 * Finds the leftmost occurrence of the given Unicode character
626 * in a UTF-8 encoded string, while limiting the search to @len bytes.
627 * If @len is -1, allow unbounded search.
629 * Return value: %NULL if the string does not contain the character,
630 * otherwise, a pointer to the start of the leftmost occurrence of
631 * the character in the string.
634 g_utf8_strchr (const char *p,
640 gint charlen = g_unichar_to_utf8 (c, ch);
643 return g_strstr_len (p, len, ch);
649 * @p: a nul-terminated UTF-8 encoded string
650 * @len: the maximum length of @p
651 * @c: a Unicode character
653 * Find the rightmost occurrence of the given Unicode character
654 * in a UTF-8 encoded string, while limiting the search to @len bytes.
655 * If @len is -1, allow unbounded search.
657 * Return value: %NULL if the string does not contain the character,
658 * otherwise, a pointer to the start of the rightmost occurrence of the
659 * character in the string.
662 g_utf8_strrchr (const char *p,
668 gint charlen = g_unichar_to_utf8 (c, ch);
671 return g_strrstr_len (p, len, ch);
675 /* Like g_utf8_get_char, but take a maximum length
676 * and return (gunichar)-2 on incomplete trailing character
678 static inline gunichar
679 g_utf8_get_char_extended (const gchar *p,
683 gunichar wc = (guchar) *p;
723 if (max_len >= 0 && len > max_len)
725 for (i = 1; i < max_len; i++)
727 if ((((guchar *)p)[i] & 0xc0) != 0x80)
733 for (i = 1; i < len; ++i)
735 gunichar ch = ((guchar *)p)[i];
737 if ((ch & 0xc0) != 0x80)
749 if (UTF8_LENGTH(wc) != len)
756 * g_utf8_get_char_validated:
757 * @p: a pointer to Unicode character encoded as UTF-8
758 * @max_len: the maximum number of bytes to read, or -1, for no maximum or
759 * if @p is nul-terminated
761 * Convert a sequence of bytes encoded as UTF-8 to a Unicode character.
762 * This function checks for incomplete characters, for invalid characters
763 * such as characters that are out of the range of Unicode, and for
764 * overlong encodings of valid characters.
766 * Return value: the resulting character. If @p points to a partial
767 * sequence at the end of a string that could begin a valid
768 * character (or if @max_len is zero), returns (gunichar)-2;
769 * otherwise, if @p does not point to a valid UTF-8 encoded
770 * Unicode character, returns (gunichar)-1.
773 g_utf8_get_char_validated (const gchar *p,
781 result = g_utf8_get_char_extended (p, max_len);
783 if (result & 0x80000000)
785 else if (!UNICODE_VALID (result))
792 * g_utf8_to_ucs4_fast:
793 * @str: a UTF-8 encoded string
794 * @len: the maximum length of @str to use. If @len < 0, then
795 * the string is nul-terminated.
796 * @items_written: location to store the number of characters in the
799 * Convert a string from UTF-8 to a 32-bit fixed width
800 * representation as UCS-4, assuming valid UTF-8 input.
801 * This function is roughly twice as fast as g_utf8_to_ucs4()
802 * but does no error checking on the input.
804 * Return value: a pointer to a newly allocated UCS-4 string.
805 * This value must be freed with g_free().
808 g_utf8_to_ucs4_fast (const gchar *str,
810 glong *items_written)
817 g_return_val_if_fail (str != NULL, NULL);
825 p = g_utf8_next_char (p);
831 while (p < str + len && *p)
833 p = g_utf8_next_char (p);
838 result = g_new (gunichar, n_chars + 1);
841 for (i=0; i < n_chars; i++)
843 gunichar wc = ((unsigned char *)p)[0];
878 for (j = 1; j < charlen; j++)
881 wc |= ((unsigned char *)p)[j] & 0x3f;
898 * @str: a UTF-8 encoded string
899 * @len: the maximum length of @str to use. If @len < 0, then
900 * the string is nul-terminated.
901 * @items_read: location to store number of bytes read, or %NULL.
902 * If %NULL, then %G_CONVERT_ERROR_PARTIAL_INPUT will be
903 * returned in case @str contains a trailing partial
904 * character. If an error occurs then the index of the
905 * invalid input is stored here.
906 * @items_written: location to store number of characters written or %NULL.
907 * The value here stored does not include the trailing 0
909 * @error: location to store the error occuring, or %NULL to ignore
910 * errors. Any of the errors in #GConvertError other than
911 * %G_CONVERT_ERROR_NO_CONVERSION may occur.
913 * Convert a string from UTF-8 to a 32-bit fixed width
914 * representation as UCS-4. A trailing 0 will be added to the
915 * string after the converted text.
917 * Return value: a pointer to a newly allocated UCS-4 string.
918 * This value must be freed with g_free(). If an
919 * error occurs, %NULL will be returned and
923 g_utf8_to_ucs4 (const gchar *str,
926 glong *items_written,
929 gunichar *result = NULL;
935 while ((len < 0 || str + len - in > 0) && *in)
937 gunichar wc = g_utf8_get_char_extended (in, len < 0 ? 6 : str + len - in);
940 if (wc == (gunichar)-2)
945 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
946 _("Partial character sequence at end of input"));
949 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
950 _("Invalid byte sequence in conversion input"));
957 in = g_utf8_next_char (in);
960 result = g_new (gunichar, n_chars + 1);
963 for (i=0; i < n_chars; i++)
965 result[i] = g_utf8_get_char (in);
966 in = g_utf8_next_char (in);
971 *items_written = n_chars;
975 *items_read = in - str;
982 * @str: a UCS-4 encoded string
983 * @len: the maximum length (number of characters) of @str to use.
984 * If @len < 0, then the string is terminated with a 0 character.
985 * @items_read: location to store number of characters read, or %NULL.
986 * @items_written: location to store number of bytes written or %NULL.
987 * The value here stored does not include the trailing 0
989 * @error: location to store the error occuring, or %NULL to ignore
990 * errors. Any of the errors in #GConvertError other than
991 * %G_CONVERT_ERROR_NO_CONVERSION may occur.
993 * Convert a string from a 32-bit fixed width representation as UCS-4.
994 * to UTF-8. The result will be terminated with a 0 byte.
996 * Return value: a pointer to a newly allocated UTF-8 string.
997 * This value must be freed with g_free(). If an
998 * error occurs, %NULL will be returned and
999 * @error set. In that case, @items_read will be
1000 * set to the position of the first invalid input
1004 g_ucs4_to_utf8 (const gunichar *str,
1007 glong *items_written,
1011 gchar *result = NULL;
1016 for (i = 0; len < 0 || i < len ; i++)
1021 if (str[i] >= 0x80000000)
1023 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1024 _("Character out of range for UTF-8"));
1028 result_length += UTF8_LENGTH (str[i]);
1031 result = g_malloc (result_length + 1);
1035 while (p < result + result_length)
1036 p += g_unichar_to_utf8 (str[i++], p);
1041 *items_written = p - result;
1050 #define SURROGATE_VALUE(h,l) (((h) - 0xd800) * 0x400 + (l) - 0xdc00 + 0x10000)
1054 * @str: a UTF-16 encoded string
1055 * @len: the maximum length (number of <type>gunichar2</type>) of @str to use.
1056 * If @len < 0, then the string is terminated with a 0 character.
1057 * @items_read: location to store number of words read, or %NULL.
1058 * If %NULL, then %G_CONVERT_ERROR_PARTIAL_INPUT will be
1059 * returned in case @str contains a trailing partial
1060 * character. If an error occurs then the index of the
1061 * invalid input is stored here.
1062 * @items_written: location to store number of bytes written, or %NULL.
1063 * The value stored here does not include the trailing
1065 * @error: location to store the error occuring, or %NULL to ignore
1066 * errors. Any of the errors in #GConvertError other than
1067 * %G_CONVERT_ERROR_NO_CONVERSION may occur.
1069 * Convert a string from UTF-16 to UTF-8. The result will be
1070 * terminated with a 0 byte.
1072 * Note that the input is expected to be already in native endianness,
1073 * an initial byte-order-mark character is not handled specially.
1074 * g_convert() can be used to convert a byte buffer of UTF-16 data of
1075 * ambiguous endianess.
1077 * Return value: a pointer to a newly allocated UTF-8 string.
1078 * This value must be freed with g_free(). If an
1079 * error occurs, %NULL will be returned and
1083 g_utf16_to_utf8 (const gunichar2 *str,
1086 glong *items_written,
1089 /* This function and g_utf16_to_ucs4 are almost exactly identical - The lines that differ
1092 const gunichar2 *in;
1094 gchar *result = NULL;
1096 gunichar high_surrogate;
1098 g_return_val_if_fail (str != 0, NULL);
1103 while ((len < 0 || in - str < len) && *in)
1108 if (c >= 0xdc00 && c < 0xe000) /* low surrogate */
1112 wc = SURROGATE_VALUE (high_surrogate, c);
1117 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1118 _("Invalid sequence in conversion input"));
1126 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1127 _("Invalid sequence in conversion input"));
1131 if (c >= 0xd800 && c < 0xdc00) /* high surrogate */
1140 /********** DIFFERENT for UTF8/UCS4 **********/
1141 n_bytes += UTF8_LENGTH (wc);
1147 if (high_surrogate && !items_read)
1149 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
1150 _("Partial character sequence at end of input"));
1154 /* At this point, everything is valid, and we just need to convert
1156 /********** DIFFERENT for UTF8/UCS4 **********/
1157 result = g_malloc (n_bytes + 1);
1162 while (out < result + n_bytes)
1167 if (c >= 0xdc00 && c < 0xe000) /* low surrogate */
1169 wc = SURROGATE_VALUE (high_surrogate, c);
1172 else if (c >= 0xd800 && c < 0xdc00) /* high surrogate */
1180 /********** DIFFERENT for UTF8/UCS4 **********/
1181 out += g_unichar_to_utf8 (wc, out);
1187 /********** DIFFERENT for UTF8/UCS4 **********/
1191 /********** DIFFERENT for UTF8/UCS4 **********/
1192 *items_written = out - result;
1196 *items_read = in - str;
1203 * @str: a UTF-16 encoded string
1204 * @len: the maximum length (number of <type>gunichar2</type>) of @str to use.
1205 * If @len < 0, then the string is terminated with a 0 character.
1206 * @items_read: location to store number of words read, or %NULL.
1207 * If %NULL, then %G_CONVERT_ERROR_PARTIAL_INPUT will be
1208 * returned in case @str contains a trailing partial
1209 * character. If an error occurs then the index of the
1210 * invalid input is stored here.
1211 * @items_written: location to store number of characters written, or %NULL.
1212 * The value stored here does not include the trailing
1214 * @error: location to store the error occuring, or %NULL to ignore
1215 * errors. Any of the errors in #GConvertError other than
1216 * %G_CONVERT_ERROR_NO_CONVERSION may occur.
1218 * Convert a string from UTF-16 to UCS-4. The result will be
1219 * terminated with a 0 character.
1221 * Return value: a pointer to a newly allocated UCS-4 string.
1222 * This value must be freed with g_free(). If an
1223 * error occurs, %NULL will be returned and
1227 g_utf16_to_ucs4 (const gunichar2 *str,
1230 glong *items_written,
1233 const gunichar2 *in;
1235 gchar *result = NULL;
1237 gunichar high_surrogate;
1239 g_return_val_if_fail (str != 0, NULL);
1244 while ((len < 0 || in - str < len) && *in)
1249 if (c >= 0xdc00 && c < 0xe000) /* low surrogate */
1253 wc = SURROGATE_VALUE (high_surrogate, c);
1258 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1259 _("Invalid sequence in conversion input"));
1267 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1268 _("Invalid sequence in conversion input"));
1272 if (c >= 0xd800 && c < 0xdc00) /* high surrogate */
1281 /********** DIFFERENT for UTF8/UCS4 **********/
1282 n_bytes += sizeof (gunichar);
1288 if (high_surrogate && !items_read)
1290 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
1291 _("Partial character sequence at end of input"));
1295 /* At this point, everything is valid, and we just need to convert
1297 /********** DIFFERENT for UTF8/UCS4 **********/
1298 result = g_malloc (n_bytes + 4);
1303 while (out < result + n_bytes)
1308 if (c >= 0xdc00 && c < 0xe000) /* low surrogate */
1310 wc = SURROGATE_VALUE (high_surrogate, c);
1313 else if (c >= 0xd800 && c < 0xdc00) /* high surrogate */
1321 /********** DIFFERENT for UTF8/UCS4 **********/
1322 *(gunichar *)out = wc;
1323 out += sizeof (gunichar);
1329 /********** DIFFERENT for UTF8/UCS4 **********/
1330 *(gunichar *)out = 0;
1333 /********** DIFFERENT for UTF8/UCS4 **********/
1334 *items_written = (out - result) / sizeof (gunichar);
1338 *items_read = in - str;
1340 return (gunichar *)result;
1345 * @str: a UTF-8 encoded string
1346 * @len: the maximum length (number of characters) of @str to use.
1347 * If @len < 0, then the string is nul-terminated.
1348 * @items_read: location to store number of bytes read, or %NULL.
1349 * If %NULL, then %G_CONVERT_ERROR_PARTIAL_INPUT will be
1350 * returned in case @str contains a trailing partial
1351 * character. If an error occurs then the index of the
1352 * invalid input is stored here.
1353 * @items_written: location to store number of <type>gunichar2</type> written,
1355 * The value stored here does not include the trailing 0.
1356 * @error: location to store the error occuring, or %NULL to ignore
1357 * errors. Any of the errors in #GConvertError other than
1358 * %G_CONVERT_ERROR_NO_CONVERSION may occur.
1360 * Convert a string from UTF-8 to UTF-16. A 0 character will be
1361 * added to the result after the converted text.
1363 * Return value: a pointer to a newly allocated UTF-16 string.
1364 * This value must be freed with g_free(). If an
1365 * error occurs, %NULL will be returned and
1369 g_utf8_to_utf16 (const gchar *str,
1372 glong *items_written,
1375 gunichar2 *result = NULL;
1380 g_return_val_if_fail (str != NULL, NULL);
1384 while ((len < 0 || str + len - in > 0) && *in)
1386 gunichar wc = g_utf8_get_char_extended (in, len < 0 ? 6 : str + len - in);
1387 if (wc & 0x80000000)
1389 if (wc == (gunichar)-2)
1394 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
1395 _("Partial character sequence at end of input"));
1398 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1399 _("Invalid byte sequence in conversion input"));
1406 else if (wc < 0xe000)
1408 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1409 _("Invalid sequence in conversion input"));
1413 else if (wc < 0x10000)
1415 else if (wc < 0x110000)
1419 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1420 _("Character out of range for UTF-16"));
1425 in = g_utf8_next_char (in);
1428 result = g_new (gunichar2, n16 + 1);
1431 for (i = 0; i < n16;)
1433 gunichar wc = g_utf8_get_char (in);
1441 result[i++] = (wc - 0x10000) / 0x400 + 0xd800;
1442 result[i++] = (wc - 0x10000) % 0x400 + 0xdc00;
1445 in = g_utf8_next_char (in);
1451 *items_written = n16;
1455 *items_read = in - str;
1462 * @str: a UCS-4 encoded string
1463 * @len: the maximum length (number of characters) of @str to use.
1464 * If @len < 0, then the string is terminated with a 0 character.
1465 * @items_read: location to store number of bytes read, or %NULL.
1466 * If an error occurs then the index of the invalid input
1468 * @items_written: location to store number of <type>gunichar2</type>
1469 * written, or %NULL. The value stored here does not
1470 * include the trailing 0.
1471 * @error: location to store the error occuring, or %NULL to ignore
1472 * errors. Any of the errors in #GConvertError other than
1473 * %G_CONVERT_ERROR_NO_CONVERSION may occur.
1475 * Convert a string from UCS-4 to UTF-16. A 0 character will be
1476 * added to the result after the converted text.
1478 * Return value: a pointer to a newly allocated UTF-16 string.
1479 * This value must be freed with g_free(). If an
1480 * error occurs, %NULL will be returned and
1484 g_ucs4_to_utf16 (const gunichar *str,
1487 glong *items_written,
1490 gunichar2 *result = NULL;
1496 while ((len < 0 || i < len) && str[i])
1498 gunichar wc = str[i];
1502 else if (wc < 0xe000)
1504 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1505 _("Invalid sequence in conversion input"));
1509 else if (wc < 0x10000)
1511 else if (wc < 0x110000)
1515 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1516 _("Character out of range for UTF-16"));
1524 result = g_new (gunichar2, n16 + 1);
1526 for (i = 0, j = 0; j < n16; i++)
1528 gunichar wc = str[i];
1536 result[j++] = (wc - 0x10000) / 0x400 + 0xd800;
1537 result[j++] = (wc - 0x10000) % 0x400 + 0xdc00;
1543 *items_written = n16;
1552 #define CONTINUATION_CHAR \
1554 if ((*(guchar *)p & 0xc0) != 0x80) /* 10xxxxxx */ \
1557 val |= (*(guchar *)p) & 0x3f; \
1560 static const gchar *
1561 fast_validate (const char *str)
1568 for (p = str; *p; p++)
1570 if (*(guchar *)p < 128)
1577 if ((*(guchar *)p & 0xe0) == 0xc0) /* 110xxxxx */
1579 if (G_UNLIKELY ((*(guchar *)p & 0x1e) == 0))
1582 if (G_UNLIKELY ((*(guchar *)p & 0xc0) != 0x80)) /* 10xxxxxx */
1587 if ((*(guchar *)p & 0xf0) == 0xe0) /* 1110xxxx */
1590 val = *(guchar *)p & 0x0f;
1593 else if ((*(guchar *)p & 0xf8) == 0xf0) /* 11110xxx */
1596 val = *(guchar *)p & 0x07;
1609 if (G_UNLIKELY (val < min))
1612 if (G_UNLIKELY (!UNICODE_VALID(val)))
1626 static const gchar *
1627 fast_validate_len (const char *str,
1635 g_assert (max_len >= 0);
1637 for (p = str; ((p - str) < max_len) && *p; p++)
1639 if (*(guchar *)p < 128)
1646 if ((*(guchar *)p & 0xe0) == 0xc0) /* 110xxxxx */
1648 if (G_UNLIKELY (max_len - (p - str) < 2))
1651 if (G_UNLIKELY ((*(guchar *)p & 0x1e) == 0))
1654 if (G_UNLIKELY ((*(guchar *)p & 0xc0) != 0x80)) /* 10xxxxxx */
1659 if ((*(guchar *)p & 0xf0) == 0xe0) /* 1110xxxx */
1661 if (G_UNLIKELY (max_len - (p - str) < 3))
1665 val = *(guchar *)p & 0x0f;
1668 else if ((*(guchar *)p & 0xf8) == 0xf0) /* 11110xxx */
1670 if (G_UNLIKELY (max_len - (p - str) < 4))
1674 val = *(guchar *)p & 0x07;
1687 if (G_UNLIKELY (val < min))
1689 if (G_UNLIKELY (!UNICODE_VALID(val)))
1705 * @str: a pointer to character data
1706 * @max_len: max bytes to validate, or -1 to go until NUL
1707 * @end: return location for end of valid data
1709 * Validates UTF-8 encoded text. @str is the text to validate;
1710 * if @str is nul-terminated, then @max_len can be -1, otherwise
1711 * @max_len should be the number of bytes to validate.
1712 * If @end is non-%NULL, then the end of the valid range
1713 * will be stored there (i.e. the start of the first invalid
1714 * character if some bytes were invalid, or the end of the text
1715 * being validated otherwise).
1717 * Note that g_utf8_validate() returns %FALSE if @max_len is
1718 * positive and NUL is met before @max_len bytes have been read.
1720 * Returns %TRUE if all of @str was valid. Many GLib and GTK+
1721 * routines <emphasis>require</emphasis> valid UTF-8 as input;
1722 * so data read from a file or the network should be checked
1723 * with g_utf8_validate() before doing anything else with it.
1725 * Return value: %TRUE if the text was valid UTF-8
1728 g_utf8_validate (const char *str,
1736 p = fast_validate (str);
1738 p = fast_validate_len (str, max_len);
1743 if ((max_len >= 0 && p != str + max_len) ||
1744 (max_len < 0 && *p != '\0'))
1751 * g_unichar_validate:
1752 * @ch: a Unicode character
1754 * Checks whether @ch is a valid Unicode character. Some possible
1755 * integer values of @ch will not be valid. 0 is considered a valid
1756 * character, though it's normally a string terminator.
1758 * Return value: %TRUE if @ch is a valid Unicode character
1761 g_unichar_validate (gunichar ch)
1763 return UNICODE_VALID (ch);
1767 * g_utf8_strreverse:
1768 * @str: a UTF-8 encoded string
1769 * @len: the maximum length of @str to use. If @len < 0, then
1770 * the string is nul-terminated.
1772 * Reverses a UTF-8 string. @str must be valid UTF-8 encoded text.
1773 * (Use g_utf8_validate() on all text before trying to use UTF-8
1774 * utility functions with it.)
1776 * Note that unlike g_strreverse(), this function returns
1777 * newly-allocated memory, which should be freed with g_free() when
1780 * Returns: a newly-allocated string which is the reverse of @str.
1785 g_utf8_strreverse (const gchar *str,
1795 result = g_new (gchar, len + 1);
1800 skip = g_utf8_skip[*(guchar*)p];
1802 for (m = r; skip; skip--)
1812 _g_utf8_make_valid (const gchar *name)
1815 const gchar *remainder, *invalid;
1816 gint remaining_bytes, valid_bytes;
1820 remaining_bytes = strlen (name);
1822 while (remaining_bytes != 0)
1824 if (g_utf8_validate (remainder, remaining_bytes, &invalid))
1826 valid_bytes = invalid - remainder;
1829 string = g_string_sized_new (remaining_bytes);
1831 g_string_append_len (string, remainder, valid_bytes);
1832 /* append U+FFFD REPLACEMENT CHARACTER */
1833 g_string_append (string, "\357\277\275");
1835 remaining_bytes -= valid_bytes + 1;
1836 remainder = invalid + 1;
1840 return g_strdup (name);
1842 g_string_append (string, remainder);
1844 g_assert (g_utf8_validate (string->str, -1, NULL));
1846 return g_string_free (string, FALSE);
1850 #define __G_UTF8_C__
1851 #include "galiasdef.c"