1 /* gutf8.c - Operations on UTF-8 strings.

     2  *

     3  * Copyright (C) 1999 Tom Tromey

     4  * Copyright (C) 2000 Red Hat, Inc.

     5  * Portions copyright (c) 2006 Nokia Corporation.  All rights reserved.

     6  *

     7  * This library is free software; you can redistribute it and/or

     8  * modify it under the terms of the GNU Lesser General Public

     9  * License as published by the Free Software Foundation; either

    10  * version 2 of the License, or (at your option) any later version.

    11  *

    12  * This library is distributed in the hope that it will be useful,

    13  * but WITHOUT ANY WARRANTY; without even the implied warranty of

    14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU

    15  * Lesser General Public License for more details.

    16  *

    17  * You should have received a copy of the GNU Lesser General Public

    18  * License along with this library; if not, write to the

    19  * Free Software Foundation, Inc., 59 Temple Place - Suite 330,

    20  * Boston, MA 02111-1307, USA.

    21  */

    22 

    23 #include <config.h>

    24 

    25 #include <stdlib.h>

    26 #ifdef HAVE_CODESET

    27 #include <langinfo.h>

    28 #endif

    29 #include <string.h>

    30 

    31 #include "glib.h"

    32 #include "galias.h"

    33 

    34 #ifdef G_PLATFORM_WIN32

    35 #include <stdio.h>

    36 #define STRICT

    37 #include <windows.h>

    38 #undef STRICT

    39 #endif

    40 

    41 #include "libcharset/libcharset.h"

    42 

    43 #include "glibintl.h"

    44 

    45 #ifdef __SYMBIAN32__

    46 #include <glib_wsd.h>

    47 #include "glibbackend.h"

    48 #endif /* __SYMBIAN32__ */

    49 

    50 #if EMULATOR

    51 #define g_thread_functions_for_glib_use (*_g_thread_functions_for_glib_use())

    52 #define g_thread_use_default_impl (*_g_thread_use_default_impl())

    53 #endif /* EMULATOR */

    54 

    55 #define UTF8_COMPUTE(Char, Mask, Len)					      \

    56   if (Char < 128)							      \

    57     {									      \

    58       Len = 1;								      \

    59       Mask = 0x7f;							      \

    60     }									      \

    61   else if ((Char & 0xe0) == 0xc0)					      \

    62     {									      \

    63       Len = 2;								      \

    64       Mask = 0x1f;							      \

    65     }									      \

    66   else if ((Char & 0xf0) == 0xe0)					      \

    67     {									      \

    68       Len = 3;								      \

    69       Mask = 0x0f;							      \

    70     }									      \

    71   else if ((Char & 0xf8) == 0xf0)					      \

    72     {									      \

    73       Len = 4;								      \

    74       Mask = 0x07;							      \

    75     }									      \

    76   else if ((Char & 0xfc) == 0xf8)					      \

    77     {									      \

    78       Len = 5;								      \

    79       Mask = 0x03;							      \

    80     }									      \

    81   else if ((Char & 0xfe) == 0xfc)					      \

    82     {									      \

    83       Len = 6;								      \

    84       Mask = 0x01;							      \

    85     }									      \

    86   else									      \

    87     Len = -1;

    88 

    89 #define UTF8_LENGTH(Char)              \

    90   ((Char) < 0x80 ? 1 :                 \

    91    ((Char) < 0x800 ? 2 :               \

    92     ((Char) < 0x10000 ? 3 :            \

    93      ((Char) < 0x200000 ? 4 :          \

    94       ((Char) < 0x4000000 ? 5 : 6)))))

    95    

    96 

    97 #define UTF8_GET(Result, Chars, Count, Mask, Len)			      \

    98   (Result) = (Chars)[0] & (Mask);					      \

    99   for ((Count) = 1; (Count) < (Len); ++(Count))				      \

   100     {									      \

   101       if (((Chars)[(Count)] & 0xc0) != 0x80)				      \

   102 	{								      \

   103 	  (Result) = -1;						      \

   104 	  break;							      \

   105 	}								      \

   106       (Result) <<= 6;							      \

   107       (Result) |= ((Chars)[(Count)] & 0x3f);				      \

   108     }

   109 

   110 #define UNICODE_VALID(Char)                   \

   111     ((Char) < 0x110000 &&                     \

   112      (((Char) & 0xFFFFF800) != 0xD800) &&     \

   113      ((Char) < 0xFDD0 || (Char) > 0xFDEF) &&  \

   114      ((Char) & 0xFFFE) != 0xFFFE)

   115    

   116      

   117 static const gchar utf8_skip_data[256] = {

   118   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,

   119   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,

   120   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,

   121   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,

   122   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,

   123   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,

   124   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,

   125   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

   126 };

   127 

   128 const gchar * const g_utf8_skip = utf8_skip_data;

   129 

   130 #ifdef __SYMBIAN32__

   131 EXPORT_C const gchar * const * _g_utf8_skip()

   132 {

   133 	return &g_utf8_skip;

   134 }

   135 #endif /* __SYMBIAN32__ */

   136 

   137 /**

   138  * g_utf8_find_prev_char:

   139  * @str: pointer to the beginning of a UTF-8 encoded string

   140  * @p: pointer to some position within @str

   141  * 

   142  * Given a position @p with a UTF-8 encoded string @str, find the start

   143  * of the previous UTF-8 character starting before @p. Returns %NULL if no

   144  * UTF-8 characters are present in @str before @p.

   145  *

   146  * @p does not have to be at the beginning of a UTF-8 character. No check

   147  * is made to see if the character found is actually valid other than

   148  * it starts with an appropriate byte.

   149  *

   150  * Return value: a pointer to the found character or %NULL.

   151  **/

   152 EXPORT_C gchar *

   153 g_utf8_find_prev_char (const char *str,

   154 		       const char *p)

   155 {

   156   for (--p; p >= str; --p)

   157     {

   158       if ((*p & 0xc0) != 0x80)

   159 	return (gchar *)p;

   160     }

   161   return NULL;

   162 }

   163 

   164 /**

   165  * g_utf8_find_next_char:

   166  * @p: a pointer to a position within a UTF-8 encoded string

   167  * @end: a pointer to the end of the string, or %NULL to indicate

   168  *        that the string is nul-terminated, in which case

   169  *        the returned value will be 

   170  *

   171  * Finds the start of the next UTF-8 character in the string after @p.

   172  *

   173  * @p does not have to be at the beginning of a UTF-8 character. No check

   174  * is made to see if the character found is actually valid other than

   175  * it starts with an appropriate byte.

   176  * 

   177  * Return value: a pointer to the found character or %NULL

   178  **/

   179 EXPORT_C gchar *

   180 g_utf8_find_next_char (const gchar *p,

   181 		       const gchar *end)

   182 {

   183   if (*p)

   184     {

   185       if (end)

   186 	for (++p; p < end && (*p & 0xc0) == 0x80; ++p)

   187 	  ;

   188       else

   189 	for (++p; (*p & 0xc0) == 0x80; ++p)

   190 	  ;

   191     }

   192   return (p == end) ? NULL : (gchar *)p;

   193 }

   194 

   195 /**

   196  * g_utf8_prev_char:

   197  * @p: a pointer to a position within a UTF-8 encoded string

   198  *

   199  * Finds the previous UTF-8 character in the string before @p.

   200  *

   201  * @p does not have to be at the beginning of a UTF-8 character. No check

   202  * is made to see if the character found is actually valid other than

   203  * it starts with an appropriate byte. If @p might be the first

   204  * character of the string, you must use g_utf8_find_prev_char() instead.

   205  * 

   206  * Return value: a pointer to the found character.

   207  **/

   208 EXPORT_C gchar *

   209 g_utf8_prev_char (const gchar *p)

   210 {

   211   while (TRUE)

   212     {

   213       p--;

   214       if ((*p & 0xc0) != 0x80)

   215 	return (gchar *)p;

   216     }

   217 }

   218 

   219 /**

   220  * g_utf8_strlen:

   221  * @p: pointer to the start of a UTF-8 encoded string.

   222  * @max: the maximum number of bytes to examine. If @max

   223  *       is less than 0, then the string is assumed to be

   224  *       nul-terminated. If @max is 0, @p will not be examined and 

   225  *       may be %NULL.

   226  * 

   227  * Returns the length of the string in characters.

   228  *

   229  * Return value: the length of the string in characters

   230  **/

   231 EXPORT_C glong

   232 g_utf8_strlen (const gchar *p,

   233                gssize       max)

   234 {

   235   glong len = 0;

   236   const gchar *start = p;

   237   g_return_val_if_fail (p != NULL || max == 0, 0);

   238 

   239   if (max < 0)

   240     {

   241       while (*p)

   242         {

   243           p = g_utf8_next_char (p);

   244           ++len;

   245         }

   246     }

   247   else

   248     {

   249       if (max == 0 || !*p)

   250         return 0;

   251       

   252       p = g_utf8_next_char (p);          

   253 

   254       while (p - start < max && *p)

   255         {

   256           ++len;

   257           p = g_utf8_next_char (p);          

   258         }

   259 

   260       /* only do the last len increment if we got a complete

   261        * char (don't count partial chars)

   262        */

   263       if (p - start <= max)

   264         ++len;

   265     }

   266 

   267   return len;

   268 }

   269 

   270 /**

   271  * g_utf8_get_char:

   272  * @p: a pointer to Unicode character encoded as UTF-8

   273  * 

   274  * Converts a sequence of bytes encoded as UTF-8 to a Unicode character.

   275  * If @p does not point to a valid UTF-8 encoded character, results are

   276  * undefined. If you are not sure that the bytes are complete

   277  * valid Unicode characters, you should use g_utf8_get_char_validated()

   278  * instead.

   279  * 

   280  * Return value: the resulting character

   281  **/

   282 EXPORT_C gunichar

   283 g_utf8_get_char (const gchar *p)

   284 {

   285   int i, mask = 0, len;

   286   gunichar result;

   287   unsigned char c = (unsigned char) *p;

   288 

   289   UTF8_COMPUTE (c, mask, len);

   290   if (len == -1)

   291     return (gunichar)-1;

   292   UTF8_GET (result, p, i, mask, len);

   293 

   294   return result;

   295 }

   296 

   297 /**

   298  * g_utf8_offset_to_pointer:

   299  * @str: a UTF-8 encoded string

   300  * @offset: a character offset within @str

   301  * 

   302  * Converts from an integer character offset to a pointer to a position

   303  * within the string.

   304  * 

   305  * Since 2.10, this function allows to pass a negative @offset to

   306  * step backwards. It is usually worth stepping backwards from the end

   307  * instead of forwards if @offset is in the last fourth of the string, 

   308  * since moving forward is about 3 times faster than moving backward.

   309  * 

   310  * Return value: the resulting pointer

   311  **/

   312 EXPORT_C gchar *

   313 g_utf8_offset_to_pointer  (const gchar *str,

   314 			   glong        offset)    

   315 {

   316   const gchar *s = str;

   317 

   318   if (offset > 0) 

   319     while (offset--)

   320       s = g_utf8_next_char (s);

   321   else

   322     {

   323       const char *s1;

   324 

   325       /* This nice technique for fast backwards stepping 

   326        * through a UTF-8 string was dubbed "stutter stepping" 

   327        * by its inventor, Larry Ewing.

   328        */

   329       while (offset)

   330 	{

   331 	  s1 = s;

   332 	  s += offset;

   333 	  while ((*s & 0xc0) == 0x80)

   334 	    s--;

   335 

   336 	  offset += g_utf8_pointer_to_offset (s, s1);

   337 	}

   338     }

   339 

   340   return (gchar *)s;

   341 }

   342 

   343 /**

   344  * g_utf8_pointer_to_offset:

   345  * @str: a UTF-8 encoded string

   346  * @pos: a pointer to a position within @str

   347  * 

   348  * Converts from a pointer to position within a string to a integer

   349  * character offset.

   350  *

   351  * Since 2.10, this function allows @pos to be before @str, and returns

   352  * a negative offset in this case.

   353  * 

   354  * Return value: the resulting character offset

   355  **/

   356 EXPORT_C glong    

   357 g_utf8_pointer_to_offset (const gchar *str,

   358 			  const gchar *pos)

   359 {

   360   const gchar *s = str;

   361   glong offset = 0;    

   362 

   363   if (pos < str) 

   364     offset = - g_utf8_pointer_to_offset (pos, str);

   365   else

   366     while (s < pos)

   367       {

   368 	s = g_utf8_next_char (s);

   369 	offset++;

   370       }

   371   

   372   return offset;

   373 }

   374 

   375 

   376 /**

   377  * g_utf8_strncpy:

   378  * @dest: buffer to fill with characters from @src

   379  * @src: UTF-8 encoded string

   380  * @n: character count

   381  * 

   382  * Like the standard C strncpy() function, but 

   383  * copies a given number of characters instead of a given number of 

   384  * bytes. The @src string must be valid UTF-8 encoded text. 

   385  * (Use g_utf8_validate() on all text before trying to use UTF-8 

   386  * utility functions with it.)

   387  * 

   388  * Return value: @dest

   389  **/

   390 EXPORT_C gchar *

   391 g_utf8_strncpy (gchar       *dest,

   392 		const gchar *src,

   393 		gsize        n)

   394 {

   395   const gchar *s = src;

   396   while (n && *s)

   397     {

   398       s = g_utf8_next_char(s);

   399       n--;

   400     }

   401   strncpy(dest, src, s - src);

   402   dest[s - src] = 0;

   403   return dest;

   404 }

   405 

   406 #if EMULATOR

   407 

   408 PLS_MACRO(aliases,gutf8,GStaticMutex)

   409 #define g__aliases_lock (*FUNCTION_NAME_MACRO(aliases,gutf8)())

   410 

   411 PLS(alias_hash,get_alias_hash,GHashTable *)

   412 #define alias_hash (*FUNCTION_NAME(alias_hash,get_alias_hash)())

   413 

   414 #else

   415 

   416 G_LOCK_DEFINE_STATIC (aliases);

   417 

   418 #endif /* EMULATOR */

   419 

   420 static GHashTable *

   421 get_alias_hash (void)

   422 {

   423   #if !(EMULATOR)

   424   static GHashTable *alias_hash = NULL;

   425   #endif /* EMULATOR */

   426   const char *aliases;

   427 

   428   G_LOCK (aliases);

   429 

   430   if (!alias_hash)

   431     {

   432       alias_hash = g_hash_table_new (g_str_hash, g_str_equal);

   433       

   434       aliases = _g_locale_get_charset_aliases ();

   435       while (*aliases != '\0')

   436 	{

   437 	  const char *canonical;

   438 	  const char *alias;

   439 	  const char **alias_array;

   440 	  int count = 0;

   441 	  

   442 	  alias = aliases;

   443 	  aliases += strlen (aliases) + 1;

   444 	  canonical = aliases;

   445 	  aliases += strlen (aliases) + 1;

   446 	  

   447 	  alias_array = g_hash_table_lookup (alias_hash, canonical);

   448 	  if (alias_array)

   449 	    {

   450 	      while (alias_array[count])

   451 		count++;

   452 	    }

   453 	  alias_array = g_renew (const char *, alias_array, count + 2);

   454 	  alias_array[count] = alias;

   455 	  alias_array[count + 1] = NULL;

   456 	  

   457 	  g_hash_table_insert (alias_hash, (char *)canonical, alias_array);

   458 	}

   459     }

   460 

   461   G_UNLOCK (aliases);

   462 

   463   return alias_hash;

   464 }

   465 

   466 #if EMULATOR

   467 #undef alias_hash

   468 #endif /* EMULATOR */

   469 

   470 /* As an abuse of the alias table, the following routines gets

   471  * the charsets that are aliases for the canonical name.

   472  */

   473 const char ** G_GNUC_INTERNAL

   474 _g_charset_get_aliases (const char *canonical_name)

   475 {

   476   GHashTable *alias_hash = get_alias_hash ();

   477 

   478   return g_hash_table_lookup (alias_hash, canonical_name);

   479 }

   480 

   481 static gboolean

   482 g_utf8_get_charset_internal (const char  *raw_data,

   483 			     const char **a)

   484 {

   485   const char *charset = getenv("CHARSET");

   486 

   487   if (charset && *charset)

   488     {

   489       *a = charset;

   490 

   491       if (charset && strstr (charset, "UTF-8"))

   492 	return TRUE;

   493       else

   494 	return FALSE;

   495     }

   496 

   497   /* The libcharset code tries to be thread-safe without

   498    * a lock, but has a memory leak and a missing memory

   499    * barrier, so we lock for it

   500    */

   501   G_LOCK (aliases);

   502   charset = _g_locale_charset_unalias (raw_data);

   503   G_UNLOCK (aliases);

   504   

   505   if (charset && *charset)

   506     {

   507       *a = charset;

   508       

   509       if (charset && strstr (charset, "UTF-8"))

   510 	return TRUE;

   511       else

   512 	return FALSE;

   513     }

   514 

   515   /* Assume this for compatibility at present.  */

   516   *a = "US-ASCII";

   517   

   518   return FALSE;

   519 }

   520 

   521 typedef struct _GCharsetCache GCharsetCache;

   522 

   523 struct _GCharsetCache {

   524   gboolean is_utf8;

   525   gchar *raw;

   526   gchar *charset;

   527 };

   528 

   529 static void

   530 charset_cache_free (gpointer data)

   531 {

   532   GCharsetCache *cache = data;

   533 #ifndef __SYMBIAN32__

   534   g_free (cache->raw);

   535   g_free (cache->charset);

   536   g_free (cache);

   537 #else

   538   pFree(cache->raw);

   539   pFree(cache->charset);

   540   pFree(cache);

   541 #endif  

   542 }

   543 

   544 /**

   545  * g_get_charset:

   546  * @charset: return location for character set name

   547  * 

   548  * Obtains the character set for the current locale; you might use

   549  * this character set as an argument to g_convert(), to convert from

   550  * the current locale's encoding to some other encoding. (Frequently

   551  * g_locale_to_utf8() and g_locale_from_utf8() are nice shortcuts,

   552  * though.)

   553  *

   554  * The return value is %TRUE if the locale's encoding is UTF-8, in that

   555  * case you can perhaps avoid calling g_convert().

   556  *

   557  * The string returned in @charset is not allocated, and should not be

   558  * freed.

   559  * 

   560  * Return value: %TRUE if the returned charset is UTF-8

   561  **/

   562 

   563 #if EMULATOR

   564 

   565 PLS(cache_private,g_get_charset,GStaticPrivate)

   566 #define cache_private (*FUNCTION_NAME(cache_private,g_get_charset)())

   567 

   568 #endif /* EMULATOR */

   569 

   570 /*

   571 #ifdef __SYMBIAN32__

   572 IMPORT_C void *pAlloc(size_t nBytes);

   573 #endif //__SYMBIAN32__

   574 */

   575 

   576 EXPORT_C gboolean

   577 g_get_charset (G_CONST_RETURN char **charset) 

   578 {

   579   #if !(EMULATOR)

   580   static GStaticPrivate cache_private = G_STATIC_PRIVATE_INIT;

   581   #endif /* EMULATOR */

   582   GCharsetCache *cache = g_static_private_get (&cache_private);

   583   const gchar *raw;

   584 

   585   if (!cache)

   586     {

   587 #ifndef __SYMBIAN32__    

   588       cache = g_new0 (GCharsetCache, 1);

   589 #else      

   590       cache = (GCharsetCache *)pAlloc(sizeof(GCharsetCache));

   591       memset(cache,'\0',sizeof(GCharsetCache));

   592 #endif //__SYMBIAN32__      

   593       g_static_private_set (&cache_private, cache, charset_cache_free);

   594     }

   595 

   596   raw = _g_locale_charset_raw ();

   597   

   598   if (!(cache->raw && strcmp (cache->raw, raw) == 0))

   599     {

   600       const gchar *new_charset;

   601 	    

   602 #ifndef __SYMBIAN32__	    

   603       g_free (cache->raw);

   604       g_free (cache->charset);

   605       cache->raw = g_strdup (raw);

   606       cache->is_utf8 = g_utf8_get_charset_internal (raw, &new_charset);

   607       cache->charset = g_strdup (new_charset);

   608 #else

   609       gchar *temp,*temp1;

   610       

   611       pFree(cache->raw);

   612       pFree(cache->charset);

   613       

   614       temp = (gchar *)pAlloc(strlen(raw) * sizeof(gchar) + 1);

   615       strncpy(temp,raw,strlen(raw));

   616       temp[strlen(raw)] = '\0';

   617       

   618       //cache->raw = g_strdup (raw);

   619       cache->raw = temp;

   620       cache->is_utf8 = g_utf8_get_charset_internal (raw, &new_charset);

   621       

   622       temp1 = (gchar *)pAlloc(strlen(new_charset) * sizeof(gchar) + 1);

   623       strncpy(temp1,new_charset,strlen(new_charset));

   624       temp1[strlen(new_charset)] = '\0';

   625       

   626       //cache->charset = g_strdup (new_charset);

   627       cache->charset = temp1;

   628 #endif //__SYMBIAN32__      

   629     }

   630 

   631   if (charset)

   632     *charset = cache->charset;

   633   

   634   return cache->is_utf8;

   635 }

   636 

   637 #if EMULATOR

   638 #undef cache_private 

   639 #endif /* EMULATOR */

   640 

   641 /* unicode_strchr */

   642 

   643 /**

   644  * g_unichar_to_utf8:

   645  * @c: a Unicode character code

   646  * @outbuf: output buffer, must have at least 6 bytes of space.

   647  *       If %NULL, the length will be computed and returned

   648  *       and nothing will be written to @outbuf.

   649  * 

   650  * Converts a single character to UTF-8.

   651  * 

   652  * Return value: number of bytes written

   653  **/

   654 EXPORT_C int

   655 g_unichar_to_utf8 (gunichar c,

   656 		   gchar   *outbuf)

   657 {

   658   /* If this gets modified, also update the copy in g_string_insert_unichar() */

   659   guint len = 0;    

   660   int first;

   661   int i;

   662 

   663   if (c < 0x80)

   664     {

   665       first = 0;

   666       len = 1;

   667     }

   668   else if (c < 0x800)

   669     {

   670       first = 0xc0;

   671       len = 2;

   672     }

   673   else if (c < 0x10000)

   674     {

   675       first = 0xe0;

   676       len = 3;

   677     }

   678    else if (c < 0x200000)

   679     {

   680       first = 0xf0;

   681       len = 4;

   682     }

   683   else if (c < 0x4000000)

   684     {

   685       first = 0xf8;

   686       len = 5;

   687     }

   688   else

   689     {

   690       first = 0xfc;

   691       len = 6;

   692     }

   693 

   694   if (outbuf)

   695     {

   696       for (i = len - 1; i > 0; --i)

   697 	{

   698 	  outbuf[i] = (c & 0x3f) | 0x80;

   699 	  c >>= 6;

   700 	}

   701       outbuf[0] = c | first;

   702     }

   703 

   704   return len;

   705 }

   706 

   707 /**

   708  * g_utf8_strchr:

   709  * @p: a nul-terminated UTF-8 encoded string

   710  * @len: the maximum length of @p

   711  * @c: a Unicode character

   712  * 

   713  * Finds the leftmost occurrence of the given Unicode character

   714  * in a UTF-8 encoded string, while limiting the search to @len bytes.

   715  * If @len is -1, allow unbounded search.

   716  * 

   717  * Return value: %NULL if the string does not contain the character, 

   718  *   otherwise, a pointer to the start of the leftmost occurrence of 

   719  *   the character in the string.

   720  **/

   721 EXPORT_C gchar *

   722 g_utf8_strchr (const char *p,

   723 	       gssize      len,

   724 	       gunichar    c)

   725 {

   726   gchar ch[10];

   727 

   728   gint charlen = g_unichar_to_utf8 (c, ch);

   729   ch[charlen] = '\0';

   730   

   731   return g_strstr_len (p, len, ch);

   732 }

   733 

   734 

   735 /**

   736  * g_utf8_strrchr:

   737  * @p: a nul-terminated UTF-8 encoded string

   738  * @len: the maximum length of @p

   739  * @c: a Unicode character

   740  * 

   741  * Find the rightmost occurrence of the given Unicode character

   742  * in a UTF-8 encoded string, while limiting the search to @len bytes.

   743  * If @len is -1, allow unbounded search.

   744  * 

   745  * Return value: %NULL if the string does not contain the character, 

   746  *   otherwise, a pointer to the start of the rightmost occurrence of the 

   747  *   character in the string.

   748  **/

   749 EXPORT_C gchar *

   750 g_utf8_strrchr (const char *p,

   751 		gssize      len,

   752 		gunichar    c)

   753 {

   754   gchar ch[10];

   755 

   756   gint charlen = g_unichar_to_utf8 (c, ch);

   757   ch[charlen] = '\0';

   758   

   759   return g_strrstr_len (p, len, ch);

   760 }

   761 

   762 

   763 /* Like g_utf8_get_char, but take a maximum length

   764  * and return (gunichar)-2 on incomplete trailing character

   765  */

   766 static inline gunichar

   767 g_utf8_get_char_extended (const  gchar *p,

   768 			  gssize max_len)  

   769 {

   770   guint i, len;

   771   gunichar wc = (guchar) *p;

   772 

   773   if (wc < 0x80)

   774     {

   775       return wc;

   776     }

   777   else if (wc < 0xc0)

   778     {

   779       return (gunichar)-1;

   780     }

   781   else if (wc < 0xe0)

   782     {

   783       len = 2;

   784       wc &= 0x1f;

   785     }

   786   else if (wc < 0xf0)

   787     {

   788       len = 3;

   789       wc &= 0x0f;

   790     }

   791   else if (wc < 0xf8)

   792     {

   793       len = 4;

   794       wc &= 0x07;

   795     }

   796   else if (wc < 0xfc)

   797     {

   798       len = 5;

   799       wc &= 0x03;

   800     }

   801   else if (wc < 0xfe)

   802     {

   803       len = 6;

   804       wc &= 0x01;

   805     }

   806   else

   807     {

   808       return (gunichar)-1;

   809     }

   810   

   811   if (max_len >= 0 && len > max_len)

   812     {

   813       for (i = 1; i < max_len; i++)

   814 	{

   815 	  if ((((guchar *)p)[i] & 0xc0) != 0x80)

   816 	    return (gunichar)-1;

   817 	}

   818       return (gunichar)-2;

   819     }

   820 

   821   for (i = 1; i < len; ++i)

   822     {

   823       gunichar ch = ((guchar *)p)[i];

   824       

   825       if ((ch & 0xc0) != 0x80)

   826 	{

   827 	  if (ch)

   828 	    return (gunichar)-1;

   829 	  else

   830 	    return (gunichar)-2;

   831 	}

   832 

   833       wc <<= 6;

   834       wc |= (ch & 0x3f);

   835     }

   836 

   837   if (UTF8_LENGTH(wc) != len)

   838     return (gunichar)-1;

   839   

   840   return wc;

   841 }

   842 

   843 /**

   844  * g_utf8_get_char_validated:

   845  * @p: a pointer to Unicode character encoded as UTF-8

   846  * @max_len: the maximum number of bytes to read, or -1, for no maximum.

   847  * 

   848  * Convert a sequence of bytes encoded as UTF-8 to a Unicode character.

   849  * This function checks for incomplete characters, for invalid characters

   850  * such as characters that are out of the range of Unicode, and for

   851  * overlong encodings of valid characters.

   852  * 

   853  * Return value: the resulting character. If @p points to a partial

   854  *    sequence at the end of a string that could begin a valid 

   855  *    character, returns (gunichar)-2; otherwise, if @p does not point 

   856  *    to a valid UTF-8 encoded Unicode character, returns (gunichar)-1.

   857  **/

   858 EXPORT_C gunichar

   859 g_utf8_get_char_validated (const  gchar *p,

   860 			   gssize max_len)

   861 {

   862   gunichar result = g_utf8_get_char_extended (p, max_len);

   863 

   864   if (result & 0x80000000)

   865     return result;

   866   else if (!UNICODE_VALID (result))

   867     return (gunichar)-1;

   868   else

   869     return result;

   870 }

   871 

   872 /**

   873  * g_utf8_to_ucs4_fast:

   874  * @str: a UTF-8 encoded string

   875  * @len: the maximum length of @str to use. If @len < 0, then

   876  *       the string is nul-terminated.

   877  * @items_written: location to store the number of characters in the

   878  *                 result, or %NULL.

   879  *

   880  * Convert a string from UTF-8 to a 32-bit fixed width

   881  * representation as UCS-4, assuming valid UTF-8 input.

   882  * This function is roughly twice as fast as g_utf8_to_ucs4()

   883  * but does no error checking on the input.

   884  * 

   885  * Return value: a pointer to a newly allocated UCS-4 string.

   886  *               This value must be freed with g_free().

   887  **/

   888 EXPORT_C gunichar *

   889 g_utf8_to_ucs4_fast (const gchar *str,

   890 		     glong        len,              

   891 		     glong       *items_written)    

   892 {

   893   gint j, charlen;

   894   gunichar *result;

   895   gint n_chars, i;

   896   const gchar *p;

   897 

   898   g_return_val_if_fail (str != NULL, NULL);

   899 

   900   p = str;

   901   n_chars = 0;

   902   if (len < 0)

   903     {

   904       while (*p)

   905 	{

   906 	  p = g_utf8_next_char (p);

   907 	  ++n_chars;

   908 	}

   909     }

   910   else

   911     {

   912       while (p < str + len && *p)

   913 	{

   914 	  p = g_utf8_next_char (p);

   915 	  ++n_chars;

   916 	}

   917     }

   918   result = g_new (gunichar, n_chars + 1);

   919   

   920   p = str;

   921   for (i=0; i < n_chars; i++)

   922     {

   923       gunichar wc = ((unsigned char *)p)[0];

   924 

   925       if (wc < 0x80)

   926 	{

   927 	  result[i] = wc;

   928 	  p++;

   929 	}

   930       else

   931 	{ 

   932 	  if (wc < 0xe0)

   933 	    {

   934 	      charlen = 2;

   935 	      wc &= 0x1f;

   936 	    }

   937 	  else if (wc < 0xf0)

   938 	    {

   939 	      charlen = 3;

   940 	      wc &= 0x0f;

   941 	    }

   942 	  else if (wc < 0xf8)

   943 	    {

   944 	      charlen = 4;

   945 	      wc &= 0x07;

   946 	    }

   947 	  else if (wc < 0xfc)

   948 	    {

   949 	      charlen = 5;

   950 	      wc &= 0x03;

   951 	    }

   952 	  else

   953 	    {

   954 	      charlen = 6;

   955 	      wc &= 0x01;

   956 	    }

   957 

   958 	  for (j = 1; j < charlen; j++)

   959 	    {

   960 	      wc <<= 6;

   961 	      wc |= ((unsigned char *)p)[j] & 0x3f;

   962 	    }

   963 

   964 	  result[i] = wc;

   965 	  p += charlen;

   966 	}

   967     }

   968   result[i] = 0;

   969 

   970   if (items_written)

   971     *items_written = i;

   972 

   973   return result;

   974 }

   975 

   976 /**

   977  * g_utf8_to_ucs4:

   978  * @str: a UTF-8 encoded string

   979  * @len: the maximum length of @str to use. If @len < 0, then

   980  *       the string is nul-terminated.

   981  * @items_read: location to store number of bytes read, or %NULL.

   982  *              If %NULL, then %G_CONVERT_ERROR_PARTIAL_INPUT will be

   983  *              returned in case @str contains a trailing partial

   984  *              character. If an error occurs then the index of the

   985  *              invalid input is stored here.

   986  * @items_written: location to store number of characters written or %NULL.

   987  *                 The value here stored does not include the trailing 0

   988  *                 character. 

   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.

   992  *

   993  * Convert a string from UTF-8 to a 32-bit fixed width

   994  * representation as UCS-4. A trailing 0 will be added to the

   995  * string after the converted text.

   996  * 

   997  * Return value: a pointer to a newly allocated UCS-4 string.

   998  *               This value must be freed with g_free(). If an

   999  *               error occurs, %NULL will be returned and

  1000  *               @error set.

  1001  **/

  1002 EXPORT_C gunichar *

  1003 g_utf8_to_ucs4 (const gchar *str,

  1004 		glong        len,             

  1005 		glong       *items_read,      

  1006 		glong       *items_written,   

  1007 		GError     **error)

  1008 {

  1009   gunichar *result = NULL;

  1010   gint n_chars, i;

  1011   const gchar *in;

  1012   

  1013   in = str;

  1014   n_chars = 0;

  1015   while ((len < 0 || str + len - in > 0) && *in)

  1016     {

  1017       gunichar wc = g_utf8_get_char_extended (in, len < 0 ? 6 : str + len - in);

  1018       if (wc & 0x80000000)

  1019 	{

  1020 	  if (wc == (gunichar)-2)

  1021 	    {

  1022 	      if (items_read)

  1023 		break;

  1024 	      else

  1025 		g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,

  1026 			     _("Partial character sequence at end of input"));

  1027 	    }

  1028 	  else

  1029 	    g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,

  1030 			 _("Invalid byte sequence in conversion input"));

  1031 

  1032 	  goto err_out;

  1033 	}

  1034 

  1035       n_chars++;

  1036 

  1037       in = g_utf8_next_char (in);

  1038     }

  1039   result = g_new (gunichar, n_chars + 1);

  1040   

  1041   in = str;

  1042   for (i=0; i < n_chars; i++)

  1043     {

  1044       result[i] = g_utf8_get_char (in);

  1045       in = g_utf8_next_char (in);

  1046     }

  1047   result[i] = 0;

  1048 

  1049   if (items_written)

  1050     *items_written = n_chars;

  1051 

  1052  err_out:

  1053   if (items_read)

  1054     *items_read = in - str;

  1055 

  1056   return result;

  1057 }

  1058 

  1059 /**

  1060  * g_ucs4_to_utf8:

  1061  * @str: a UCS-4 encoded string

  1062  * @len: the maximum length (number of characters) of @str to use. 

  1063  *       If @len < 0, then the string is terminated with a 0 character.

  1064  * @items_read: location to store number of characters read, or %NULL.

  1065  * @items_written: location to store number of bytes written or %NULL.

  1066  *                 The value here stored does not include the trailing 0

  1067  *                 byte. 

  1068  * @error: location to store the error occuring, or %NULL to ignore

  1069  *         errors. Any of the errors in #GConvertError other than

  1070  *         %G_CONVERT_ERROR_NO_CONVERSION may occur.

  1071  *

  1072  * Convert a string from a 32-bit fixed width representation as UCS-4.

  1073  * to UTF-8. The result will be terminated with a 0 byte.

  1074  * 

  1075  * Return value: a pointer to a newly allocated UTF-8 string.

  1076  *               This value must be freed with g_free(). If an

  1077  *               error occurs, %NULL will be returned and

  1078  *               @error set. In that case, @items_read will be

  1079  *               set to the position of the first invalid input 

  1080  *               character.

  1081  **/

  1082 EXPORT_C gchar *

  1083 g_ucs4_to_utf8 (const gunichar *str,

  1084 		glong           len,              

  1085 		glong          *items_read,       

  1086 		glong          *items_written,    

  1087 		GError        **error)

  1088 {

  1089   gint result_length;

  1090   gchar *result = NULL;

  1091   gchar *p;

  1092   gint i;

  1093 

  1094   result_length = 0;

  1095   for (i = 0; len < 0 || i < len ; i++)

  1096     {

  1097       if (!str[i])

  1098 	break;

  1099 

  1100       if (str[i] >= 0x80000000)

  1101 	{

  1102 	  g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,

  1103 		       _("Character out of range for UTF-8"));

  1104 	  goto err_out;

  1105 	}

  1106       

  1107       result_length += UTF8_LENGTH (str[i]);

  1108     }

  1109   result = g_malloc (result_length + 1);

  1110   p = result;

  1111 

  1112   i = 0;

  1113   while (p < result + result_length)

  1114     p += g_unichar_to_utf8 (str[i++], p);

  1115   

  1116   *p = '\0';

  1117 

  1118   if (items_written)

  1119     *items_written = p - result;

  1120 

  1121  err_out:

  1122   if (items_read)

  1123     *items_read = i;

  1124 

  1125   return result;

  1126 }

  1127 

  1128 #define SURROGATE_VALUE(h,l) (((h) - 0xd800) * 0x400 + (l) - 0xdc00 + 0x10000)

  1129 

  1130 /**

  1131  * g_utf16_to_utf8:

  1132  * @str: a UTF-16 encoded string

  1133  * @len: the maximum length (number of <type>gunichar2</type>) of @str to use. 

  1134  *       If @len < 0, then the string is terminated with a 0 character.