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 #GPatternSpec-struct is the 'compiled' form of a pattern. This
60 * structure is opaque and its fields cannot be accessed directly.
63 /* keep enum and structure of gpattern.c and patterntest.c in sync */
66 G_MATCH_ALL, /* "*A?A*" */
67 G_MATCH_ALL_TAIL, /* "*A?AA" */
68 G_MATCH_HEAD, /* "AAAA*" */
69 G_MATCH_TAIL, /* "*AAAA" */
70 G_MATCH_EXACT, /* "AAAAA" */
76 GMatchType match_type;
84 /* --- functions --- */
85 static inline gboolean
86 g_pattern_ph_match (const gchar *match_pattern,
87 const gchar *match_string,
88 gboolean *wildcard_reached_p)
90 register const gchar *pattern, *string;
93 pattern = match_pattern;
94 string = match_string;
105 string = g_utf8_next_char (string);
109 *wildcard_reached_p = TRUE;
118 string = g_utf8_next_char (string);
121 while (ch == '*' || ch == '?');
126 gboolean next_wildcard_reached = FALSE;
127 while (ch != *string)
131 string = g_utf8_next_char (string);
134 if (g_pattern_ph_match (pattern, string, &next_wildcard_reached))
136 if (next_wildcard_reached)
137 /* the forthcoming pattern substring up to the next wildcard has
138 * been matched, but a mismatch occoured for the rest of the
139 * pattern, following the next wildcard.
140 * there's no need to advance the current match position any
141 * further if the rest pattern will not match.
165 * @pspec: a #GPatternSpec
166 * @string_length: the length of @string (in bytes, i.e. strlen(),
167 * <emphasis>not</emphasis> g_utf8_strlen())
168 * @string: the UTF-8 encoded string to match
169 * @string_reversed: (allow-none): the reverse of @string or %NULL
171 * Matches a string against a compiled pattern. Passing the correct
172 * length of the string given is mandatory. The reversed string can be
173 * omitted by passing %NULL, this is more efficient if the reversed
174 * version of the string to be matched is not at hand, as
175 * g_pattern_match() will only construct it if the compiled pattern
176 * requires reverse matches.
178 * Note that, if the user code will (possibly) match a string against a
179 * multitude of patterns containing wildcards, chances are high that
180 * some patterns will require a reversed string. In this case, it's
181 * more efficient to provide the reversed string to avoid multiple
182 * constructions thereof in the various calls to g_pattern_match().
184 * Note also that the reverse of a UTF-8 encoded string can in general
185 * <emphasis>not</emphasis> be obtained by g_strreverse(). This works
186 * only if the string doesn't contain any multibyte characters. GLib
187 * offers the g_utf8_strreverse() function to reverse UTF-8 encoded
190 * Returns: %TRUE if @string matches @pspec
193 g_pattern_match (GPatternSpec *pspec,
196 const gchar *string_reversed)
198 g_return_val_if_fail (pspec != NULL, FALSE);
199 g_return_val_if_fail (string != NULL, FALSE);
201 if (string_length < pspec->min_length ||
202 string_length > pspec->max_length)
205 switch (pspec->match_type)
209 return g_pattern_ph_match (pspec->pattern, string, &dummy);
210 case G_MATCH_ALL_TAIL:
212 return g_pattern_ph_match (pspec->pattern, string_reversed, &dummy);
217 tmp = g_utf8_strreverse (string, string_length);
218 result = g_pattern_ph_match (pspec->pattern, tmp, &dummy);
223 if (pspec->pattern_length == string_length)
224 return strcmp (pspec->pattern, string) == 0;
225 else if (pspec->pattern_length)
226 return strncmp (pspec->pattern, string, pspec->pattern_length) == 0;
230 if (pspec->pattern_length)
231 return strcmp (pspec->pattern, string + (string_length - pspec->pattern_length)) == 0;
235 if (pspec->pattern_length != string_length)
238 return strcmp (pspec->pattern, string) == 0;
240 g_return_val_if_fail (pspec->match_type < G_MATCH_LAST, FALSE);
246 * g_pattern_spec_new:
247 * @pattern: a zero-terminated UTF-8 encoded string
249 * Compiles a pattern to a #GPatternSpec.
251 * Returns: a newly-allocated #GPatternSpec
254 g_pattern_spec_new (const gchar *pattern)
257 gboolean seen_joker = FALSE, seen_wildcard = FALSE, more_wildcards = FALSE;
258 gint hw_pos = -1, tw_pos = -1, hj_pos = -1, tj_pos = -1;
259 gboolean follows_wildcard = FALSE;
260 guint pending_jokers = 0;
265 g_return_val_if_fail (pattern != NULL, NULL);
267 /* canonicalize pattern and collect necessary stats */
268 pspec = g_new (GPatternSpec, 1);
269 pspec->pattern_length = strlen (pattern);
270 pspec->min_length = 0;
271 pspec->max_length = 0;
272 pspec->pattern = g_new (gchar, pspec->pattern_length + 1);
274 for (i = 0, s = pattern; *s != 0; s++)
279 if (follows_wildcard) /* compress multiple wildcards */
281 pspec->pattern_length--;
284 follows_wildcard = TRUE;
292 pspec->max_length += 4; /* maximum UTF-8 character length */
295 for (; pending_jokers; pending_jokers--, i++) {
301 follows_wildcard = FALSE;
309 for (; pending_jokers; pending_jokers--) {
316 seen_joker = hj_pos >= 0;
317 seen_wildcard = hw_pos >= 0;
318 more_wildcards = seen_wildcard && hw_pos != tw_pos;
320 pspec->max_length = G_MAXUINT;
322 /* special case sole head/tail wildcard or exact matches */
323 if (!seen_joker && !more_wildcards)
325 if (pspec->pattern[0] == '*')
327 pspec->match_type = G_MATCH_TAIL;
328 memmove (pspec->pattern, pspec->pattern + 1, --pspec->pattern_length);
329 pspec->pattern[pspec->pattern_length] = 0;
332 if (pspec->pattern_length > 0 &&
333 pspec->pattern[pspec->pattern_length - 1] == '*')
335 pspec->match_type = G_MATCH_HEAD;
336 pspec->pattern[--pspec->pattern_length] = 0;
341 pspec->match_type = G_MATCH_EXACT;
346 /* now just need to distinguish between head or tail match start */
347 tw_pos = pspec->pattern_length - 1 - tw_pos; /* last pos to tail distance */
348 tj_pos = pspec->pattern_length - 1 - tj_pos; /* last pos to tail distance */
350 pspec->match_type = tw_pos > hw_pos ? G_MATCH_ALL_TAIL : G_MATCH_ALL;
351 else /* seen_joker */
352 pspec->match_type = tj_pos > hj_pos ? G_MATCH_ALL_TAIL : G_MATCH_ALL;
353 if (pspec->match_type == G_MATCH_ALL_TAIL) {
354 gchar *tmp = pspec->pattern;
355 pspec->pattern = g_utf8_strreverse (pspec->pattern, pspec->pattern_length);
362 * g_pattern_spec_free:
363 * @pspec: a #GPatternSpec
365 * Frees the memory allocated for the #GPatternSpec.
368 g_pattern_spec_free (GPatternSpec *pspec)
370 g_return_if_fail (pspec != NULL);
372 g_free (pspec->pattern);
377 * g_pattern_spec_equal:
378 * @pspec1: a #GPatternSpec
379 * @pspec2: another #GPatternSpec
381 * Compares two compiled pattern specs and returns whether they will
382 * match the same set of strings.
384 * Returns: Whether the compiled patterns are equal
387 g_pattern_spec_equal (GPatternSpec *pspec1,
388 GPatternSpec *pspec2)
390 g_return_val_if_fail (pspec1 != NULL, FALSE);
391 g_return_val_if_fail (pspec2 != NULL, FALSE);
393 return (pspec1->pattern_length == pspec2->pattern_length &&
394 pspec1->match_type == pspec2->match_type &&
395 strcmp (pspec1->pattern, pspec2->pattern) == 0);
399 * g_pattern_match_string:
400 * @pspec: a #GPatternSpec
401 * @string: the UTF-8 encoded string to match
403 * Matches a string against a compiled pattern. If the string is to be
404 * matched against more than one pattern, consider using
405 * g_pattern_match() instead while supplying the reversed string.
407 * Returns: %TRUE if @string matches @pspec
410 g_pattern_match_string (GPatternSpec *pspec,
413 g_return_val_if_fail (pspec != NULL, FALSE);
414 g_return_val_if_fail (string != NULL, FALSE);
416 return g_pattern_match (pspec, strlen (string), string, NULL);
420 * g_pattern_match_simple:
421 * @pattern: the UTF-8 encoded pattern
422 * @string: the UTF-8 encoded string to match
424 * Matches a string against a pattern given as a string. If this
425 * function is to be called in a loop, it's more efficient to compile
426 * the pattern once with g_pattern_spec_new() and call
427 * g_pattern_match_string() repeatedly.
429 * Returns: %TRUE if @string matches @pspec
432 g_pattern_match_simple (const gchar *pattern,
438 g_return_val_if_fail (pattern != NULL, FALSE);
439 g_return_val_if_fail (string != NULL, FALSE);
441 pspec = g_pattern_spec_new (pattern);
442 ergo = g_pattern_match (pspec, strlen (string), string, NULL);
443 g_pattern_spec_free (pspec);