1 /* GLIB - Library of useful routines for C programming
2 * Copyright (C) 1995-1997, 1999 Peter Mattis, Red Hat, Inc.
4 * This library is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU Lesser General Public
6 * License as published by the Free Software Foundation; either
7 * version 2 of the License, or (at your option) any later version.
9 * This library is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 * Lesser General Public License for more details.
14 * You should have received a copy of the GNU Lesser General Public
15 * License along with this library; if not, write to the
16 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
17 * Boston, MA 02111-1307, USA.
27 #include "gmessages.h"
34 * @title: Glob-style pattern matching
35 * @short_description: matches strings against patterns containing '*'
36 * (wildcard) and '?' (joker)
38 * The <function>g_pattern_match*</function> functions match a string
39 * against a pattern containing '*' and '?' wildcards with similar
40 * semantics as the standard glob() function: '*' matches an arbitrary,
41 * possibly empty, string, '?' matches an arbitrary character.
43 * Note that in contrast to glob(), the '/' character
44 * <emphasis>can</emphasis> be matched by the wildcards, there are no
45 * '[...]' character ranges and '*' and '?' can
46 * <emphasis>not</emphasis> be escaped to include them literally in a
49 * When multiple strings must be matched against the same pattern, it
50 * is better to compile the pattern to a #GPatternSpec using
51 * g_pattern_spec_new() and use g_pattern_match_string() instead of
52 * g_pattern_match_simple(). This avoids the overhead of repeated
53 * pattern compilation.
59 * A <structname>GPatternSpec</structname> is the 'compiled' form of a
60 * pattern. This structure is opaque and its fields cannot be accessed
64 /* keep enum and structure of gpattern.c and patterntest.c in sync */
67 G_MATCH_ALL, /* "*A?A*" */
68 G_MATCH_ALL_TAIL, /* "*A?AA" */
69 G_MATCH_HEAD, /* "AAAA*" */
70 G_MATCH_TAIL, /* "*AAAA" */
71 G_MATCH_EXACT, /* "AAAAA" */
77 GMatchType match_type;
85 /* --- functions --- */
86 static inline gboolean
87 g_pattern_ph_match (const gchar *match_pattern,
88 const gchar *match_string,
89 gboolean *wildcard_reached_p)
91 register const gchar *pattern, *string;
94 pattern = match_pattern;
95 string = match_string;
106 string = g_utf8_next_char (string);
110 *wildcard_reached_p = TRUE;
119 string = g_utf8_next_char (string);
122 while (ch == '*' || ch == '?');
127 gboolean next_wildcard_reached = FALSE;
128 while (ch != *string)
132 string = g_utf8_next_char (string);
135 if (g_pattern_ph_match (pattern, string, &next_wildcard_reached))
137 if (next_wildcard_reached)
138 /* the forthcoming pattern substring up to the next wildcard has
139 * been matched, but a mismatch occoured for the rest of the
140 * pattern, following the next wildcard.
141 * there's no need to advance the current match position any
142 * further if the rest pattern will not match.
166 * @pspec: a #GPatternSpec
167 * @string_length: the length of @string (in bytes, i.e. strlen(),
168 * <emphasis>not</emphasis> g_utf8_strlen())
169 * @string: the UTF-8 encoded string to match
170 * @string_reversed: (allow-none): the reverse of @string or %NULL
172 * Matches a string against a compiled pattern. Passing the correct
173 * length of the string given is mandatory. The reversed string can be
174 * omitted by passing %NULL, this is more efficient if the reversed
175 * version of the string to be matched is not at hand, as
176 * g_pattern_match() will only construct it if the compiled pattern
177 * requires reverse matches.
179 * Note that, if the user code will (possibly) match a string against a
180 * multitude of patterns containing wildcards, chances are high that
181 * some patterns will require a reversed string. In this case, it's
182 * more efficient to provide the reversed string to avoid multiple
183 * constructions thereof in the various calls to g_pattern_match().
185 * Note also that the reverse of a UTF-8 encoded string can in general
186 * <emphasis>not</emphasis> be obtained by g_strreverse(). This works
187 * only if the string doesn't contain any multibyte characters. GLib
188 * offers the g_utf8_strreverse() function to reverse UTF-8 encoded
191 * Returns: %TRUE if @string matches @pspec
194 g_pattern_match (GPatternSpec *pspec,
197 const gchar *string_reversed)
199 g_return_val_if_fail (pspec != NULL, FALSE);
200 g_return_val_if_fail (string != NULL, FALSE);
202 if (string_length < pspec->min_length ||
203 string_length > pspec->max_length)
206 switch (pspec->match_type)
210 return g_pattern_ph_match (pspec->pattern, string, &dummy);
211 case G_MATCH_ALL_TAIL:
213 return g_pattern_ph_match (pspec->pattern, string_reversed, &dummy);
218 tmp = g_utf8_strreverse (string, string_length);
219 result = g_pattern_ph_match (pspec->pattern, tmp, &dummy);
224 if (pspec->pattern_length == string_length)
225 return strcmp (pspec->pattern, string) == 0;
226 else if (pspec->pattern_length)
227 return strncmp (pspec->pattern, string, pspec->pattern_length) == 0;
231 if (pspec->pattern_length)
232 return strcmp (pspec->pattern, string + (string_length - pspec->pattern_length)) == 0;
236 if (pspec->pattern_length != string_length)
239 return strcmp (pspec->pattern, string) == 0;
241 g_return_val_if_fail (pspec->match_type < G_MATCH_LAST, FALSE);
247 * g_pattern_spec_new:
248 * @pattern: a zero-terminated UTF-8 encoded string
250 * Compiles a pattern to a #GPatternSpec.
252 * Returns: a newly-allocated #GPatternSpec
255 g_pattern_spec_new (const gchar *pattern)
258 gboolean seen_joker = FALSE, seen_wildcard = FALSE, more_wildcards = FALSE;
259 gint hw_pos = -1, tw_pos = -1, hj_pos = -1, tj_pos = -1;
260 gboolean follows_wildcard = FALSE;
261 guint pending_jokers = 0;
266 g_return_val_if_fail (pattern != NULL, NULL);
268 /* canonicalize pattern and collect necessary stats */
269 pspec = g_new (GPatternSpec, 1);
270 pspec->pattern_length = strlen (pattern);
271 pspec->min_length = 0;
272 pspec->max_length = 0;
273 pspec->pattern = g_new (gchar, pspec->pattern_length + 1);
275 for (i = 0, s = pattern; *s != 0; s++)
280 if (follows_wildcard) /* compress multiple wildcards */
282 pspec->pattern_length--;
285 follows_wildcard = TRUE;
293 pspec->max_length += 4; /* maximum UTF-8 character length */
296 for (; pending_jokers; pending_jokers--, i++) {
302 follows_wildcard = FALSE;
310 for (; pending_jokers; pending_jokers--) {
317 seen_joker = hj_pos >= 0;
318 seen_wildcard = hw_pos >= 0;
319 more_wildcards = seen_wildcard && hw_pos != tw_pos;
321 pspec->max_length = G_MAXUINT;
323 /* special case sole head/tail wildcard or exact matches */
324 if (!seen_joker && !more_wildcards)
326 if (pspec->pattern[0] == '*')
328 pspec->match_type = G_MATCH_TAIL;
329 memmove (pspec->pattern, pspec->pattern + 1, --pspec->pattern_length);
330 pspec->pattern[pspec->pattern_length] = 0;
333 if (pspec->pattern_length > 0 &&
334 pspec->pattern[pspec->pattern_length - 1] == '*')
336 pspec->match_type = G_MATCH_HEAD;
337 pspec->pattern[--pspec->pattern_length] = 0;
342 pspec->match_type = G_MATCH_EXACT;
347 /* now just need to distinguish between head or tail match start */
348 tw_pos = pspec->pattern_length - 1 - tw_pos; /* last pos to tail distance */
349 tj_pos = pspec->pattern_length - 1 - tj_pos; /* last pos to tail distance */
351 pspec->match_type = tw_pos > hw_pos ? G_MATCH_ALL_TAIL : G_MATCH_ALL;
352 else /* seen_joker */
353 pspec->match_type = tj_pos > hj_pos ? G_MATCH_ALL_TAIL : G_MATCH_ALL;
354 if (pspec->match_type == G_MATCH_ALL_TAIL) {
355 gchar *tmp = pspec->pattern;
356 pspec->pattern = g_utf8_strreverse (pspec->pattern, pspec->pattern_length);
363 * g_pattern_spec_free:
364 * @pspec: a #GPatternSpec
366 * Frees the memory allocated for the #GPatternSpec.
369 g_pattern_spec_free (GPatternSpec *pspec)
371 g_return_if_fail (pspec != NULL);
373 g_free (pspec->pattern);
378 * g_pattern_spec_equal:
379 * @pspec1: a #GPatternSpec
380 * @pspec2: another #GPatternSpec
382 * Compares two compiled pattern specs and returns whether they will
383 * match the same set of strings.
385 * Returns: Whether the compiled patterns are equal
388 g_pattern_spec_equal (GPatternSpec *pspec1,
389 GPatternSpec *pspec2)
391 g_return_val_if_fail (pspec1 != NULL, FALSE);
392 g_return_val_if_fail (pspec2 != NULL, FALSE);
394 return (pspec1->pattern_length == pspec2->pattern_length &&
395 pspec1->match_type == pspec2->match_type &&
396 strcmp (pspec1->pattern, pspec2->pattern) == 0);
400 * g_pattern_match_string:
401 * @pspec: a #GPatternSpec
402 * @string: the UTF-8 encoded string to match
404 * Matches a string against a compiled pattern. If the string is to be
405 * matched against more than one pattern, consider using
406 * g_pattern_match() instead while supplying the reversed string.
408 * Returns: %TRUE if @string matches @pspec
411 g_pattern_match_string (GPatternSpec *pspec,
414 g_return_val_if_fail (pspec != NULL, FALSE);
415 g_return_val_if_fail (string != NULL, FALSE);
417 return g_pattern_match (pspec, strlen (string), string, NULL);
421 * g_pattern_match_simple:
422 * @pattern: the UTF-8 encoded pattern
423 * @string: the UTF-8 encoded string to match
425 * Matches a string against a pattern given as a string. If this
426 * function is to be called in a loop, it's more efficient to compile
427 * the pattern once with g_pattern_spec_new() and call
428 * g_pattern_match_string() repeatedly.
430 * Returns: %TRUE if @string matches @pspec
433 g_pattern_match_simple (const gchar *pattern,
439 g_return_val_if_fail (pattern != NULL, FALSE);
440 g_return_val_if_fail (string != NULL, FALSE);
442 pspec = g_pattern_spec_new (pattern);
443 ergo = g_pattern_match (pspec, strlen (string), string, NULL);
444 g_pattern_spec_free (pspec);