File: e-data-server-util.c

package info (click to toggle)
evolution-data-server 1.6.3-5etch3
  • links: PTS
  • area: main
  • in suites: etch
  • size: 59,384 kB
  • ctags: 43,218
  • sloc: ansic: 319,315; tcl: 30,499; xml: 19,166; sh: 18,776; perl: 11,529; cpp: 8,259; java: 7,653; makefile: 6,448; awk: 1,338; yacc: 1,103; sed: 772; cs: 505; lex: 134; asm: 14
file content (731 lines) | stat: -rw-r--r-- 21,376 bytes parent folder | download | duplicates (2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
/* -*- Mode: C; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
/*
 * Copyright (C) 2003 Novell Inc.
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of version 2 of the GNU Lesser General Public
 * License as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 02111-1307, USA.
 *
 * Authors: Rodrigo Moya <rodrigo@ximian.com>
 */

#include "config.h"

#include <string.h>
#include <sys/stat.h>
#include <time.h>
#include <unistd.h>

#include <glib.h>

#include "e-data-server-util.h"

/**
 * e_util_mkdir_hier:
 * @path: The directory hierarchy to create.
 * @mode: The permissions to use for the directories.
 *
 * Creates a directory hierarchy based on the string @path. If @path
 * is an absolute path, the directories will be created relative to
 * the root of the file system; otherwise, the directories will be
 * created relative to the current directory.
 *
 * Returns: 0 on success; -1 on failure.
 **/
int
e_util_mkdir_hier (const char *path, mode_t mode)
{
#if GLIB_CHECK_VERSION(2,8,0)
	return g_mkdir_with_parents (path, mode);
#else
        char *copy, *p;
                                                                                
        if (g_path_is_absolute (path)) {
                p = copy = g_strdup (path);
        } else {
                gchar *current_dir = g_get_current_dir();
                p = copy = g_build_filename (current_dir, path, NULL);
		g_free (current_dir);
        }
                                                                                
	p = (char *)g_path_skip_root (p);                                                                       
        do {
                char *p1 = strchr (p, '/');
#ifdef G_OS_WIN32
		{
			char *p2 = strchr (p, '\\');
			if (p1 == NULL ||
			    (p2 != NULL && p2 < p1))
				p1 = p2;
		}
#endif
		p = p1;
		if (p)
			*p = '\0';
		if (!g_file_test (copy, G_FILE_TEST_IS_DIR)) {
			if (g_mkdir (copy, mode) == -1) {
				g_free (copy);
				return -1;
			}
		}
		if (p) {
			*p++ = '/';
		}
        } while (p);
                                                                                
        g_free (copy);
        return 0;
#endif
}

/**
 * e_util_strstrcase:
 * @haystack: The string to search in.
 * @needle: The string to search for.
 *
 * Find the first instance of @needle in @haystack, ignoring case for
 * bytes that are ASCII characters.
 *
 * Returns: A pointer to the start of @needle in @haystack, or NULL if
 *          @needle is not found.
 **/
gchar *
e_util_strstrcase (const gchar *haystack, const gchar *needle)
{
        /* find the needle in the haystack neglecting case */
        const gchar *ptr;
        guint len;

        g_return_val_if_fail (haystack != NULL, NULL);
        g_return_val_if_fail (needle != NULL, NULL);

        len = strlen(needle);
        if (len > strlen(haystack))
                return NULL;

        if (len == 0)
                return (gchar *) haystack;

        for (ptr = haystack; *(ptr + len - 1) != '\0'; ptr++)
                if (!g_ascii_strncasecmp (ptr, needle, len))
                        return (gchar *) ptr;

        return NULL;
}

/** 
 * e_util_unicode_get_utf8:
 * @text: The string to take the UTF-8 character from.
 * @out: The location to store the UTF-8 character in.
 * 
 * Get a UTF-8 character from the beginning of @text.
 *
 * Returns: A pointer to the next character in @text after @out. 
 **/
gchar *
e_util_unicode_get_utf8 (const gchar *text, gunichar *out)
{
        *out = g_utf8_get_char (text);
        return (*out == (gunichar)-1) ? NULL : g_utf8_next_char (text);
}

/** 
 * e_util_utf8_strstrcase:
 * @haystack: The string to search in.
 * @needle: The string to search for.
 * 
 * Find the first instance of @needle in @haystack, ignoring case. (No
 * proper case folding or decomposing is done.) Both @needle and
 * @haystack are UTF-8 strings.
 *
 * Returns: A pointer to the first instance of @needle in @haystack, or
 *          %NULL if no match is found, or if either of the strings are
 *          not legal UTF-8 strings.
 **/
const gchar *
e_util_utf8_strstrcase (const gchar *haystack, const gchar *needle)
{
        gunichar *nuni;
        gunichar unival;
        gint nlen;
        const guchar *o, *p;

        if (haystack == NULL) return NULL;
        if (needle == NULL) return NULL;
        if (strlen (needle) == 0) return haystack;
        if (strlen (haystack) == 0) return NULL;

        nuni = g_alloca (sizeof (gunichar) * strlen (needle));

        nlen = 0;
        for (p = e_util_unicode_get_utf8 (needle, &unival); p && unival; p = e_util_unicode_get_utf8 (p, &unival)) {
                nuni[nlen++] = g_unichar_tolower (unival);
        }
        /* NULL means there was illegal utf-8 sequence */
        if (!p) return NULL;

	o = haystack;
        for (p = e_util_unicode_get_utf8 (o, &unival); p && unival; p = e_util_unicode_get_utf8 (p, &unival)) {
                gint sc;
                sc = g_unichar_tolower (unival);
                /* We have valid stripped char */
                if (sc == nuni[0]) {
                        const gchar *q = p;
                        gint npos = 1;
                        while (npos < nlen) {
                                q = e_util_unicode_get_utf8 (q, &unival);
                                if (!q || !unival) return NULL;
                                sc = g_unichar_tolower (unival);
                                if (sc != nuni[npos]) break;
                                npos++;
                        }
                        if (npos == nlen) {
                                return o;
                        }
                }
                o = p;
        }

        return NULL;
}


static gunichar
stripped_char (gunichar ch)
{
        gunichar *decomp, retval;
        GUnicodeType utype;
        gint dlen;

        utype = g_unichar_type (ch);

        switch (utype) {
        case G_UNICODE_CONTROL:
        case G_UNICODE_FORMAT:
        case G_UNICODE_UNASSIGNED:
        case G_UNICODE_COMBINING_MARK:
                /* Ignore those */
                return 0;
               break;
        default:
                /* Convert to lowercase, fall through */
                ch = g_unichar_tolower (ch);
        case G_UNICODE_LOWERCASE_LETTER:
                if ((decomp = g_unicode_canonical_decomposition (ch, &dlen))) {
                        retval = decomp[0];
                        g_free (decomp);
                        return retval;
                }
                break;
        }

        return 0;
}

/** 
 * e_util_utf8_strstrcasedecomp:
 * @haystack: The string to search in.
 * @needle: The string to search for.
 * 
 * Find the first instance of @needle in @haystack, where both @needle
 * and @haystack are UTF-8 strings. Both strings are stripped and
 * decomposed for comparison, and case is ignored.
 *
 * Returns: A pointer to the first instance of @needle in @haystack, or
 *          %NULL if either of the strings are not legal UTF-8 strings.
 **/
const gchar *
e_util_utf8_strstrcasedecomp (const gchar *haystack, const gchar *needle)
{
        gunichar *nuni;
        gunichar unival;
        gint nlen;
        const guchar *o, *p;

        if (haystack == NULL) return NULL;
        if (needle == NULL) return NULL;
        if (strlen (needle) == 0) return haystack;
        if (strlen (haystack) == 0) return NULL;

        nuni = g_alloca (sizeof (gunichar) * strlen (needle));

        nlen = 0;
        for (p = e_util_unicode_get_utf8 (needle, &unival); p && unival; p = e_util_unicode_get_utf8 (p, &unival)) {
                gint sc;
                sc = stripped_char (unival);
                if (sc) {
                       nuni[nlen++] = sc;
                }
        }
        /* NULL means there was illegal utf-8 sequence */
        if (!p) return NULL;
        /* If everything is correct, we have decomposed, lowercase, stripped needle */
        if (nlen < 1) return haystack;

        o = haystack;
        for (p = e_util_unicode_get_utf8 (o, &unival); p && unival; p = e_util_unicode_get_utf8 (p, &unival)) {
                gint sc;
                sc = stripped_char (unival);
                if (sc) {
                        /* We have valid stripped char */
                        if (sc == nuni[0]) {
                                const gchar *q = p;
                                gint npos = 1;
                                while (npos < nlen) {
                                        q = e_util_unicode_get_utf8 (q, &unival);
                                        if (!q || !unival) return NULL;
                                        sc = stripped_char (unival);
                                        if ((!sc) || (sc != nuni[npos])) break;
                                        npos++;
                                }
                                if (npos == nlen) {
                                        return o;
                                }
                        }
                }
                o = p;
        }

        return NULL;
}

int
e_util_utf8_strcasecmp (const gchar *s1, const gchar *s2)
{
	gchar *folded_s1, *folded_s2;
	int retval;

	g_return_val_if_fail (s1 != NULL && s2 != NULL, -1);
	
	if (strcmp (s1, s2) == 0)
		return 0;

	folded_s1 = g_utf8_casefold (s1, -1);
	folded_s2 = g_utf8_casefold (s2, -1);

	retval = g_utf8_collate (folded_s1, folded_s2);

	g_free (folded_s2);
	g_free (folded_s1);

	return retval;
}

/** 
 * e_strftime:
 * @s: The string array to store the result in.
 * @max: The size of array @s.
 * @fmt: The formatting to use on @tm.
 * @tm: The time value to format.
 *
 * This function is a wrapper around the strftime(3) function, which
 * converts the &percnt;l and &percnt;k (12h and 24h) format variables if necessary.
 *
 * Returns: The number of characters placed in @s.
 **/
size_t e_strftime(char *s, size_t max, const char *fmt, const struct tm *tm)
{
#ifdef HAVE_LKSTRFTIME
	return strftime(s, max, fmt, tm);
#else
	char *c, *ffmt, *ff;
	size_t ret;

	ffmt = g_strdup(fmt);
	ff = ffmt;
	while ((c = strstr(ff, "%l")) != NULL) {
		c[1] = 'I';
		ff = c;
	}

	ff = ffmt;
	while ((c = strstr(ff, "%k")) != NULL) {
		c[1] = 'H';
		ff = c;
	}

#ifdef G_OS_WIN32
	/* The Microsoft strftime() doesn't have %e either */
	ff = ffmt;
	while ((c = strstr(ff, "%e")) != NULL) {
		c[1] = 'd';
		ff = c;
	}
#endif

	ret = strftime(s, max, ffmt, tm);
	g_free(ffmt);
	return ret;
#endif
}

/** 
 * e_utf8_strftime:
 * @s: The string array to store the result in.
 * @max: The size of array @s.
 * @fmt: The formatting to use on @tm.
 * @tm: The time value to format.
 *
 * The UTF-8 equivalent of e_strftime().
 *
 * Returns: The number of characters placed in @s.
 **/
size_t 
e_utf8_strftime(char *s, size_t max, const char *fmt, const struct tm *tm)
{
	size_t sz, ret;
	char *locale_fmt, *buf;

	locale_fmt = g_locale_from_utf8(fmt, -1, NULL, &sz, NULL);
	if (!locale_fmt)
		return 0;

	ret = e_strftime(s, max, locale_fmt, tm);
	if (!ret) {
		g_free (locale_fmt);
		return 0;
	}

	buf = g_locale_to_utf8(s, ret, NULL, &sz, NULL);
	if (!buf) {
		g_free (locale_fmt);
		return 0;
	}

	if (sz >= max) {
		char *tmp = buf + max - 1;
		tmp = g_utf8_find_prev_char(buf, tmp);
		if (tmp)
			sz = tmp - buf;
		else
			sz = 0;
	}
	memcpy(s, buf, sz);
	s[sz] = '\0';
	g_free(locale_fmt);
	g_free(buf);
	return sz;
}

/**
 * e_util_pthread_id:
 * @t: A pthread_t value
 *
 * Returns a 64-bit integer hopefully uniquely identifying the
 * thread. To be used in debugging output and logging only. To test
 * whether two pthread_t values refer to the same thread, use
 * pthread_equal().
 *
 * There is no guarantee that calling e_util_pthread_id() on one
 * thread first and later after that thread has dies on another won't
 * return the same integer.
 *
 * On some platforms it might even be that when called two times on
 * the same thread's pthread_t (with some pthread API calls inbetween)
 * we will return different values (this of course makes this function
 * rather useless on such platforms).
 *
 * On Linux and Win32, known to really return a unique id for each
 * thread existing at a certain time. No guarantee that ids won't be
 * reused after a thread has terminated, though.
 *
 * Returns: A 64-bit integer.
 */
guint64
e_util_pthread_id (pthread_t t)
{
#ifdef HAVE_GUINT64_CASTABLE_PTHREAD_T
	/* We know that pthread_t is an integral type, or at least
	 * castable to such without loss of precision.
	 */
	return (guint64) t;
#elif defined (PTW32_VERSION)
	/* pthreads-win32 implementation on Windows: Return the
	 * pointer to the "actual object" (see pthread.h)
	 */
#if GLIB_SIZEOF_VOID_P == 8
	/* 64-bit Windows */
	return (guint64) t.p;
#else
	return (int) t.p;
#endif
#else
	/* Just return a checksum of the contents of the pthread_t */
	{
		guint64 retval = 0;
		guchar *const tend = (guchar *) ((&t)+1);
		guchar *tp = (guchar *) &t;

		while (tp < tend)
			retval = (retval << 5) - retval * *tp++;

		return retval;
	}
#endif
}

/* This only makes a filename safe for usage as a filename.  It still may have shell meta-characters in it. */

/* This code is rather misguided and mostly pointless, but can't be
 * changed because of backward compatibility, I guess.
 *
 * It replaces some perfectly safe characters like '%' with an
 * underscore. (Recall that on Unix, the only bytes not allowed in a
 * file name component are '\0' and '/'.) On the other hand, the UTF-8
 * for a printable non-ASCII Unicode character (that thus consists of
 * several very nonprintable non-ASCII bytes) is let through as
 * such. But those bytes are of course also allowed in filenames, so
 * it doesn't matter as such...
 */
void
e_filename_make_safe (gchar *string)
{
	gchar *p, *ts;
	gunichar c;
#ifdef G_OS_WIN32
	const char *unsafe_chars = " /'\"`&();|<>$%{}!\\:*?";
#else
	const char *unsafe_chars = " /'\"`&();|<>$%{}!";
#endif	
	
	g_return_if_fail (string != NULL);
	p = string;

	while(p && *p) {
		c = g_utf8_get_char (p);
		ts = p;
		p = g_utf8_next_char (p);
		/* I wonder what this code is supposed to actually
		 * achieve, and whether it does that as currently
		 * written?
		 */
		if (!g_unichar_isprint(c) || ( c < 0xff && strchr (unsafe_chars, c&0xff ))) {
			while (ts<p) 	
				*ts++ = '_';
		}
	}
}

#ifdef G_OS_WIN32

#include <windows.h>

/* The following function is lifted from libgnome. We don't want
 * libedataserver to depend on libgnome, especially as libgnome is
 * being deprecated. This function will move to GLib, presumably, but
 * isn't there yet.
 */

/**
 * gnome_win32_get_prefixes:
 * @hmodule: The handle to a DLL (a HMODULE).
 * @full_prefix: Where the full UTF-8 path to the DLL's installation folder
 *               will be returned.
 * @cp_prefix: Where a system codepage version of
 *             the installation folder will be returned.
 *
 * This function looks up the installation prefix of the DLL (or EXE)
 * with handle @hmodule. The prefix using long filenames and in UTF-8
 * form is returned in @full_prefix. The prefix using short file names
 * (if present in the file system) and in the system codepage is
 * returned in @cp_prefix. To determine the installation prefix, the
 * full path to the DLL or EXE is first fetched. If the last folder
 * component in that path is called "bin", its parent folder is used,
 * otherwise the folder itself.
 *
 * If either can't be obtained, %NULL is stored. The caller should be
 * prepared to handle that.
 *
 * The returned character pointers are newly allocated and should be
 * freed with g_free when not longer needed.
 */
static void
get_prefixes (gpointer  hmodule,
	      char    **full_prefix,
	      char    **cp_prefix)
{
        wchar_t wcbfr[1000];
        char cpbfr[1000];

        g_return_if_fail (full_prefix != NULL);
        g_return_if_fail (cp_prefix != NULL);

        *full_prefix = NULL;
        *cp_prefix = NULL;

        if (G_WIN32_HAVE_WIDECHAR_API ()) {
                /* NT-based Windows has wide char API */
                if (GetModuleFileNameW ((HMODULE) hmodule, wcbfr, G_N_ELEMENTS (wcbfr))) {
                        *full_prefix = g_utf16_to_utf8 (wcbfr, -1,
                                                        NULL, NULL, NULL);
                        if (GetShortPathNameW (wcbfr, wcbfr, G_N_ELEMENTS (wcbfr)) &&
                            /* Short pathnames always contain only
                             * ASCII, I think, but just in case, be
                             * prepared.
                             */
                            WideCharToMultiByte (CP_ACP, 0, wcbfr, -1,
                                                 cpbfr, G_N_ELEMENTS (cpbfr),
                                                 NULL, NULL))
                                *cp_prefix = g_strdup (cpbfr);
                        else if (*full_prefix)
                                *cp_prefix = g_locale_from_utf8 (*full_prefix, -1,
                                                                 NULL, NULL, NULL);
                }
        } else {
                /* Win9x */
                if (GetModuleFileNameA ((HMODULE) hmodule, cpbfr, G_N_ELEMENTS (cpbfr))) {
                        *full_prefix = g_locale_to_utf8 (cpbfr, -1,
                                                         NULL, NULL, NULL);
                        *cp_prefix = g_strdup (cpbfr);
                }
        }

        if (*full_prefix != NULL) {
                gchar *p = strrchr (*full_prefix, '\\');
                if (p != NULL)
                        *p = '\0';
      
                p = strrchr (*full_prefix, '\\');
                if (p && (g_ascii_strcasecmp (p + 1, "bin") == 0))
                        *p = '\0';
        }

        /* cp_prefix is in system codepage */
        if (*cp_prefix != NULL) {
                gchar *p = _mbsrchr (*cp_prefix, '\\');
                if (p != NULL)
                        *p = '\0';
      
                p = _mbsrchr (*cp_prefix, '\\');
                if (p && (g_ascii_strcasecmp (p + 1, "bin") == 0))
                        *p = '\0';
        }
}

static const char *prefix = NULL;
static const char *cp_prefix;

static const char *localedir;
static const char *extensiondir;
static const char *imagesdir;
static const char *ui_gladedir;

static HMODULE hmodule;
G_LOCK_DEFINE_STATIC (mutex);

/* Silence gcc with a prototype. Yes, this is silly. */
BOOL WINAPI DllMain (HINSTANCE hinstDLL,
		     DWORD     fdwReason,
		     LPVOID    lpvReserved);

/* Minimal DllMain that just tucks away the DLL's HMODULE */
BOOL WINAPI
DllMain (HINSTANCE hinstDLL,
	 DWORD     fdwReason,
	 LPVOID    lpvReserved)
{
        switch (fdwReason) {
        case DLL_PROCESS_ATTACH:
                hmodule = hinstDLL;
                break;
        }
        return TRUE;
}

char *
e_util_replace_prefix (const char *configure_time_prefix,
		       const char *runtime_prefix,
		       const char *configure_time_path)
{
	char *c_t_prefix_slash = g_strconcat (configure_time_prefix, "/",
					      NULL);
	char *retval;

        if (runtime_prefix &&
            g_str_has_prefix (configure_time_path, c_t_prefix_slash)) {
                retval = g_strconcat (runtime_prefix,
				      configure_time_path + strlen (configure_time_prefix),
				      NULL);
        } else
                retval = g_strdup (configure_time_path);

	g_free (c_t_prefix_slash);

	return retval;
}

static char *
replace_prefix (const char *runtime_prefix,
		const char *configure_time_path)
{
	return e_util_replace_prefix (E_DATA_SERVER_PREFIX,
				      runtime_prefix,
				      configure_time_path);
}

static void
setup (void)
{
	char *full_pfx;  
	char *cp_pfx; 

        G_LOCK (mutex);
        if (prefix != NULL) {
                G_UNLOCK (mutex);
                return;
        }

	/* This requires that the libedataserver DLL is installed in $bindir */
        get_prefixes (hmodule, &full_pfx, &cp_pfx);

	prefix = g_strdup (full_pfx);
	cp_prefix = g_strdup (cp_pfx);

	g_free (full_pfx);
	g_free (cp_pfx);

	localedir = replace_prefix (cp_prefix, EVOLUTION_LOCALEDIR);
	extensiondir = replace_prefix (prefix, E_DATA_SERVER_EXTENSIONDIR);
	imagesdir = replace_prefix (prefix, E_DATA_SERVER_IMAGESDIR);
	ui_gladedir = replace_prefix (prefix, E_DATA_SERVER_UI_GLADEDIR);

	G_UNLOCK (mutex);
}

#include "libedataserver-private.h" /* For prototypes */

#define GETTER_IMPL(varbl)			\
{						\
        setup ();				\
        return varbl;				\
}

#define PRIVATE_GETTER(varbl)			\
const char *					\
_libedataserver_get_##varbl (void)		\
	GETTER_IMPL(varbl)

#define PUBLIC_GETTER(varbl)			\
const char *					\
e_util_get_##varbl (void)			\
	GETTER_IMPL(varbl)

PRIVATE_GETTER(extensiondir)
PRIVATE_GETTER(imagesdir)
PRIVATE_GETTER(ui_gladedir)

PUBLIC_GETTER(prefix)
PUBLIC_GETTER(cp_prefix)
PUBLIC_GETTER(localedir)

#endif	/* G_OS_WIN32 */