1 /* gunicollate.c - Collation
3 * Copyright 2001 Red Hat, Inc.
5 * The Gnome Library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Lesser General Public License as
7 * published by the Free Software Foundation; either version 2 of the
8 * License, or (at your option) any later version.
10 * The Gnome Library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Lesser General Public License for more details.
15 * You should have received a copy of the GNU Lesser General Public
16 * License along with the Gnome Library; see the file COPYING.LIB. If not,
17 * write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
18 * Boston, MA 02111-1307, USA.
25 #ifdef __STDC_ISO_10646__
31 #include "gunicodeprivate.h"
35 * @str1: a UTF-8 encoded string
36 * @str2: a UTF-8 encoded string
38 * Compares two strings for ordering using the linguistically
39 * correct rules for the current locale. When sorting a large
40 * number of strings, it will be significantly faster to
41 * obtain collation keys with g_utf8_collate_key() and
42 * compare the keys with strcmp() when
43 * sorting instead of sorting the original strings.
45 * Return value: < 0 if @str1 compares before @str2,
46 * 0 if they compare equal, > 0 if @str1 compares after @str2.
49 g_utf8_collate (const gchar *str1,
54 #ifdef __STDC_ISO_10646__
59 g_return_val_if_fail (str1 != NULL, 0);
60 g_return_val_if_fail (str2 != NULL, 0);
62 str1_norm = _g_utf8_normalize_wc (str1, -1, G_NORMALIZE_ALL_COMPOSE);
63 str2_norm = _g_utf8_normalize_wc (str2, -1, G_NORMALIZE_ALL_COMPOSE);
65 result = wcscoll ((wchar_t *)str1_norm, (wchar_t *)str2_norm);
70 #else /* !__STDC_ISO_10646__ */
76 g_return_val_if_fail (str1 != NULL, 0);
77 g_return_val_if_fail (str2 != NULL, 0);
79 str1_norm = g_utf8_normalize (str1, -1, G_NORMALIZE_ALL_COMPOSE);
80 str2_norm = g_utf8_normalize (str2, -1, G_NORMALIZE_ALL_COMPOSE);
82 if (g_get_charset (&charset))
84 result = strcoll (str1_norm, str2_norm);
88 gchar *str1_locale = g_convert (str1_norm, -1, charset, "UTF-8", NULL, NULL, NULL);
89 gchar *str2_locale = g_convert (str2_norm, -1, charset, "UTF-8", NULL, NULL, NULL);
91 if (str1_locale && str2_locale)
92 result = strcoll (str1_locale, str2_locale);
98 result = strcmp (str1_norm, str2_norm);
100 g_free (str1_locale);
101 g_free (str2_locale);
107 #endif /* __STDC_ISO_10646__ */
112 #ifdef __STDC_ISO_10646__
113 /* We need UTF-8 encoding of numbers to encode the weights if
114 * we are using wcsxfrm. However, we aren't encoding Unicode
115 * characters, so we can't simply use g_unichar_to_utf8.
117 * The following routine is taken (with modification) from GNU
118 * libc's strxfrm routine:
120 * Copyright (C) 1995-1999,2000,2001 Free Software Foundation, Inc.
121 * Written by Ulrich Drepper <drepper@cygnus.com>, 1995.
124 utf8_encode (char *buf, wchar_t val)
138 for (step = 2; step < 6; ++step)
139 if ((val & (~(guint32)0 << (5 * step + 1))) == 0)
145 *buf = (unsigned char) (~0xff >> step);
149 buf[step] = 0x80 | (val & 0x3f);
159 #endif /* __STDC_ISO_10646__ */
162 * g_utf8_collate_key:
163 * @str: a UTF-8 encoded string.
164 * @len: length of @str, in bytes, or -1 if @str is nul-terminated.
166 * Converts a string into a collation key that can be compared
167 * with other collation keys using strcmp().
168 * The results of comparing the collation keys of two strings
169 * with strcmp() will always be the same as
170 * comparing the two original keys with g_utf8_collate().
172 * Return value: a newly allocated string. This string should
173 * be freed with g_free() when you are done with it.
176 g_utf8_collate_key (const gchar *str,
182 #ifdef __STDC_ISO_10646__
187 size_t result_len = 0;
189 g_return_val_if_fail (str != NULL, NULL);
191 str_norm = _g_utf8_normalize_wc (str, len, G_NORMALIZE_ALL_COMPOSE);
193 setlocale (LC_COLLATE, "");
195 xfrm_len = wcsxfrm (NULL, (wchar_t *)str_norm, 0);
196 result_wc = g_new (wchar_t, xfrm_len + 1);
197 wcsxfrm (result_wc, (wchar_t *)str_norm, xfrm_len + 1);
199 for (i=0; i < xfrm_len; i++)
200 result_len += utf8_encode (NULL, result_wc[i]);
202 result = g_malloc (result_len + 1);
204 for (i=0; i < xfrm_len; i++)
205 result_len += utf8_encode (result + result_len, result_wc[i]);
207 result[result_len] = '\0';
213 #else /* !__STDC_ISO_10646__ */
215 const gchar *charset;
218 g_return_val_if_fail (str != NULL, NULL);
220 str_norm = g_utf8_normalize (str, len, G_NORMALIZE_ALL_COMPOSE);
222 if (g_get_charset (&charset))
224 xfrm_len = strxfrm (NULL, str_norm, 0);
225 result = g_malloc (xfrm_len + 1);
226 strxfrm (result, str_norm, xfrm_len + 1);
230 gchar *str_locale = g_convert (str_norm, -1, charset, "UTF-8", NULL, NULL, NULL);
234 xfrm_len = strxfrm (NULL, str_locale, 0);
235 if (xfrm_len < 0 || xfrm_len >= G_MAXINT - 2)
243 result = g_malloc (xfrm_len + 2);
245 strxfrm (result + 1, str_locale, xfrm_len + 1);
251 xfrm_len = strlen (str_norm);
252 result = g_malloc (xfrm_len + 2);
254 memcpy (result + 1, str_norm, xfrm_len);
255 result[xfrm_len+1] = '\0';
260 #endif /* __STDC_ISO_10646__ */