* Lesser 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.
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
*/
+
+#include "config.h"
+
+#include <string.h>
+
#include "gpattern.h"
#include "gmacros.h"
#include "gmem.h"
#include "gunicode.h"
#include "gutils.h"
-#include <string.h>
+
+/**
+ * SECTION:patterns
+ * @title: Glob-style pattern matching
+ * @short_description: matches strings against patterns containing '*'
+ * (wildcard) and '?' (joker)
+ *
+ * The g_pattern_match* functions match a string
+ * against a pattern containing '*' and '?' wildcards with similar
+ * semantics as the standard glob() function: '*' matches an arbitrary,
+ * possibly empty, string, '?' matches an arbitrary character.
+ *
+ * Note that in contrast to glob(), the '/' character can be matched by
+ * the wildcards, there are no '[...]' character ranges and '*' and '?'
+ * can not be escaped to include them literally in a pattern.
+ *
+ * When multiple strings must be matched against the same pattern, it
+ * is better to compile the pattern to a #GPatternSpec using
+ * g_pattern_spec_new() and use g_pattern_match_string() instead of
+ * g_pattern_match_simple(). This avoids the overhead of repeated
+ * pattern compilation.
+ **/
+
+/**
+ * GPatternSpec:
+ *
+ * A GPatternSpec struct is the 'compiled' form of a pattern. This
+ * structure is opaque and its fields cannot be accessed directly.
+ */
/* keep enum and structure of gpattern.c and patterntest.c in sync */
typedef enum
GMatchType match_type;
guint pattern_length;
guint min_length;
+ guint max_length;
gchar *pattern;
};
/* --- functions --- */
-/**
- * g_utf8_reverse:
- * @string: a UTF-8 string.
- *
- * Reverses a UTF-8 string. The @string must be valid UTF-8 encoded text.
- * (Use g_utf8_validate() on all text before trying to use UTF-8
- * utility functions with it.)
- *
- * Note that unlike g_strreverse(), this function returns
- * newly-allocated memory, which should be freed with g_free() when
- * no longer needed.
- *
- * Returns: a newly-allocated string which is the reverse of @string.
- */
-static gchar *
-g_utf8_reverse (guint len, const gchar *string)
-{
- gchar *result;
- const gchar *p;
- gchar *m, *r, skip;
-
- result = g_new (gchar, len + 1);
- r = result + len;
- p = string;
- while (*p)
- {
- skip = g_utf8_skip[*(guchar*)p];
- r -= skip;
- for (m = r; skip; skip--)
- *m++ = *p++;
- }
- result[len] = 0;
-
- return result;
-}
-
static inline gboolean
g_pattern_ph_match (const gchar *match_pattern,
- const gchar *match_string)
+ const gchar *match_string,
+ gboolean *wildcard_reached_p)
{
- register const gchar *pattern, *string;
- register gchar ch;
+ const gchar *pattern, *string;
+ gchar ch;
pattern = match_pattern;
string = match_string;
break;
case '*':
+ *wildcard_reached_p = TRUE;
do
{
ch = *pattern;
return TRUE;
do
{
+ gboolean next_wildcard_reached = FALSE;
while (ch != *string)
{
if (!*string)
string = g_utf8_next_char (string);
}
string++;
- if (g_pattern_ph_match (pattern, string))
+ if (g_pattern_ph_match (pattern, string, &next_wildcard_reached))
return TRUE;
+ if (next_wildcard_reached)
+ /* the forthcoming pattern substring up to the next wildcard has
+ * been matched, but a mismatch occoured for the rest of the
+ * pattern, following the next wildcard.
+ * there's no need to advance the current match position any
+ * further if the rest pattern will not match.
+ */
+ return FALSE;
}
while (*string);
break;
return *string == 0;
}
+/**
+ * g_pattern_match:
+ * @pspec: a #GPatternSpec
+ * @string_length: the length of @string (in bytes, i.e. strlen(),
+ * not g_utf8_strlen())
+ * @string: the UTF-8 encoded string to match
+ * @string_reversed: (allow-none): the reverse of @string or %NULL
+ *
+ * Matches a string against a compiled pattern. Passing the correct
+ * length of the string given is mandatory. The reversed string can be
+ * omitted by passing %NULL, this is more efficient if the reversed
+ * version of the string to be matched is not at hand, as
+ * g_pattern_match() will only construct it if the compiled pattern
+ * requires reverse matches.
+ *
+ * Note that, if the user code will (possibly) match a string against a
+ * multitude of patterns containing wildcards, chances are high that
+ * some patterns will require a reversed string. In this case, it's
+ * more efficient to provide the reversed string to avoid multiple
+ * constructions thereof in the various calls to g_pattern_match().
+ *
+ * Note also that the reverse of a UTF-8 encoded string can in general
+ * not be obtained by g_strreverse(). This works only if the string
+ * does not contain any multibyte characters. GLib offers the
+ * g_utf8_strreverse() function to reverse UTF-8 encoded strings.
+ *
+ * Returns: %TRUE if @string matches @pspec
+ **/
gboolean
g_pattern_match (GPatternSpec *pspec,
guint string_length,
g_return_val_if_fail (pspec != NULL, FALSE);
g_return_val_if_fail (string != NULL, FALSE);
- if (pspec->min_length > string_length)
+ if (string_length < pspec->min_length ||
+ string_length > pspec->max_length)
return FALSE;
switch (pspec->match_type)
{
+ gboolean dummy;
case G_MATCH_ALL:
- return g_pattern_ph_match (pspec->pattern, string);
+ return g_pattern_ph_match (pspec->pattern, string, &dummy);
case G_MATCH_ALL_TAIL:
if (string_reversed)
- return g_pattern_ph_match (pspec->pattern, string_reversed);
+ return g_pattern_ph_match (pspec->pattern, string_reversed, &dummy);
else
{
gboolean result;
gchar *tmp;
- tmp = g_utf8_reverse (string_length, string);
- result = g_pattern_ph_match (pspec->pattern, tmp);
+ tmp = g_utf8_strreverse (string, string_length);
+ result = g_pattern_ph_match (pspec->pattern, tmp, &dummy);
g_free (tmp);
return result;
}
return TRUE;
case G_MATCH_EXACT:
if (pspec->pattern_length != string_length)
- return FALSE;
+ return FALSE;
else
- return strcmp (pspec->pattern, string) == 0;
+ return strcmp (pspec->pattern, string) == 0;
default:
g_return_val_if_fail (pspec->match_type < G_MATCH_LAST, FALSE);
return FALSE;
}
}
+/**
+ * g_pattern_spec_new:
+ * @pattern: a zero-terminated UTF-8 encoded string
+ *
+ * Compiles a pattern to a #GPatternSpec.
+ *
+ * Returns: a newly-allocated #GPatternSpec
+ **/
GPatternSpec*
g_pattern_spec_new (const gchar *pattern)
{
pspec = g_new (GPatternSpec, 1);
pspec->pattern_length = strlen (pattern);
pspec->min_length = 0;
+ pspec->max_length = 0;
pspec->pattern = g_new (gchar, pspec->pattern_length + 1);
d = pspec->pattern;
for (i = 0, s = pattern; *s != 0; s++)
case '?':
pending_jokers++;
pspec->min_length++;
+ pspec->max_length += 4; /* maximum UTF-8 character length */
continue;
default:
for (; pending_jokers; pending_jokers--, i++) {
}
follows_wildcard = FALSE;
pspec->min_length++;
+ pspec->max_length++;
break;
}
*d++ = *s;
seen_joker = hj_pos >= 0;
seen_wildcard = hw_pos >= 0;
more_wildcards = seen_wildcard && hw_pos != tw_pos;
+ if (seen_wildcard)
+ pspec->max_length = G_MAXUINT;
/* special case sole head/tail wildcard or exact matches */
if (!seen_joker && !more_wildcards)
pspec->pattern[pspec->pattern_length] = 0;
return pspec;
}
- if (pspec->pattern[pspec->pattern_length - 1] == '*')
+ if (pspec->pattern_length > 0 &&
+ pspec->pattern[pspec->pattern_length - 1] == '*')
{
pspec->match_type = G_MATCH_HEAD;
pspec->pattern[--pspec->pattern_length] = 0;
pspec->match_type = tj_pos > hj_pos ? G_MATCH_ALL_TAIL : G_MATCH_ALL;
if (pspec->match_type == G_MATCH_ALL_TAIL) {
gchar *tmp = pspec->pattern;
- pspec->pattern = g_utf8_reverse (pspec->pattern_length, pspec->pattern);
+ pspec->pattern = g_utf8_strreverse (pspec->pattern, pspec->pattern_length);
g_free (tmp);
}
return pspec;
}
+/**
+ * g_pattern_spec_free:
+ * @pspec: a #GPatternSpec
+ *
+ * Frees the memory allocated for the #GPatternSpec.
+ **/
void
g_pattern_spec_free (GPatternSpec *pspec)
{
g_free (pspec);
}
+/**
+ * g_pattern_spec_equal:
+ * @pspec1: a #GPatternSpec
+ * @pspec2: another #GPatternSpec
+ *
+ * Compares two compiled pattern specs and returns whether they will
+ * match the same set of strings.
+ *
+ * Returns: Whether the compiled patterns are equal
+ **/
gboolean
g_pattern_spec_equal (GPatternSpec *pspec1,
GPatternSpec *pspec2)
strcmp (pspec1->pattern, pspec2->pattern) == 0);
}
+/**
+ * g_pattern_match_string:
+ * @pspec: a #GPatternSpec
+ * @string: the UTF-8 encoded string to match
+ *
+ * Matches a string against a compiled pattern. If the string is to be
+ * matched against more than one pattern, consider using
+ * g_pattern_match() instead while supplying the reversed string.
+ *
+ * Returns: %TRUE if @string matches @pspec
+ **/
gboolean
g_pattern_match_string (GPatternSpec *pspec,
const gchar *string)
return g_pattern_match (pspec, strlen (string), string, NULL);
}
+/**
+ * g_pattern_match_simple:
+ * @pattern: the UTF-8 encoded pattern
+ * @string: the UTF-8 encoded string to match
+ *
+ * Matches a string against a pattern given as a string. If this
+ * function is to be called in a loop, it's more efficient to compile
+ * the pattern once with g_pattern_spec_new() and call
+ * g_pattern_match_string() repeatedly.
+ *
+ * Returns: %TRUE if @string matches @pspec
+ **/
gboolean
g_pattern_match_simple (const gchar *pattern,
const gchar *string)
projects / platform / upstream / glib.git / blobdiff