* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
*/
-#ifdef HAVE_CONFIG_H
-# include <config.h>
-#endif
-
-#include "gregex.h"
+#include "config.h"
-#include <glib.h>
-#include <glib/gi18n.h>
#include <string.h>
#ifdef USE_SYSTEM_PCRE
#include "pcre/pcre.h"
#endif
-#include "galias.h"
+#include "gtypes.h"
+#include "gregex.h"
+#include "glibintl.h"
+#include "glist.h"
+#include "gmessages.h"
+#include "gstrfuncs.h"
+#include "gatomic.h"
+#include "gthread.h"
+
+/**
+ * SECTION:gregex
+ * @title: Perl-compatible regular expressions
+ * @short_description: matches strings against regular expressions
+ * @see_also: <xref linkend="glib-regex-syntax"/>
+ *
+ * The <function>g_regex_*()</function> functions implement regular
+ * expression pattern matching using syntax and semantics similar to
+ * Perl regular expression.
+ *
+ * Some functions accept a @start_position argument, setting it differs
+ * from just passing over a shortened string and setting #G_REGEX_MATCH_NOTBOL
+ * in the case of a pattern that begins with any kind of lookbehind assertion.
+ * For example, consider the pattern "\Biss\B" which finds occurrences of "iss"
+ * in the middle of words. ("\B" matches only if the current position in the
+ * subject is not a word boundary.) When applied to the string "Mississipi"
+ * from the fourth byte, namely "issipi", it does not match, because "\B" is
+ * always false at the start of the subject, which is deemed to be a word
+ * boundary. However, if the entire string is passed , but with
+ * @start_position set to 4, it finds the second occurrence of "iss" because
+ * it is able to look behind the starting point to discover that it is
+ * preceded by a letter.
+ *
+ * Note that, unless you set the #G_REGEX_RAW flag, all the strings passed
+ * to these functions must be encoded in UTF-8. The lengths and the positions
+ * inside the strings are in bytes and not in characters, so, for instance,
+ * "\xc3\xa0" (i.e. "à") is two bytes long but it is treated as a
+ * single character. If you set #G_REGEX_RAW the strings can be non-valid
+ * UTF-8 strings and a byte is treated as a character, so "\xc3\xa0" is two
+ * bytes and two characters long.
+ *
+ * When matching a pattern, "\n" matches only against a "\n" character in
+ * the string, and "\r" matches only a "\r" character. To match any newline
+ * sequence use "\R". This particular group matches either the two-character
+ * sequence CR + LF ("\r\n"), or one of the single characters LF (linefeed,
+ * U+000A, "\n"), VT vertical tab, U+000B, "\v"), FF (formfeed, U+000C, "\f"),
+ * CR (carriage return, U+000D, "\r"), NEL (next line, U+0085), LS (line
+ * separator, U+2028), or PS (paragraph separator, U+2029).
+ *
+ * The behaviour of the dot, circumflex, and dollar metacharacters are
+ * affected by newline characters, the default is to recognize any newline
+ * character (the same characters recognized by "\R"). This can be changed
+ * with #G_REGEX_NEWLINE_CR, #G_REGEX_NEWLINE_LF and #G_REGEX_NEWLINE_CRLF
+ * compile options, and with #G_REGEX_MATCH_NEWLINE_ANY,
+ * #G_REGEX_MATCH_NEWLINE_CR, #G_REGEX_MATCH_NEWLINE_LF and
+ * #G_REGEX_MATCH_NEWLINE_CRLF match options. These settings are also
+ * relevant when compiling a pattern if #G_REGEX_EXTENDED is set, and an
+ * unescaped "#" outside a character class is encountered. This indicates
+ * a comment that lasts until after the next newline.
+ *
+ * Creating and manipulating the same #GRegex structure from different
+ * threads is not a problem as #GRegex does not modify its internal
+ * state between creation and destruction, on the other hand #GMatchInfo
+ * is not threadsafe.
+ *
+ * The regular expressions low-level functionalities are obtained through
+ * the excellent <ulink url="http://www.pcre.org/">PCRE</ulink> library
+ * written by Philip Hazel.
+ */
/* Mask of all the possible values for GRegexCompileFlags. */
-#define G_REGEX_COMPILE_MASK (G_REGEX_CASELESS | \
- G_REGEX_MULTILINE | \
- G_REGEX_DOTALL | \
- G_REGEX_EXTENDED | \
- G_REGEX_ANCHORED | \
- G_REGEX_DOLLAR_ENDONLY | \
- G_REGEX_UNGREEDY | \
- G_REGEX_RAW | \
- G_REGEX_NO_AUTO_CAPTURE | \
- G_REGEX_DUPNAMES | \
- G_REGEX_NEWLINE_CR | \
- G_REGEX_NEWLINE_LF | \
- G_REGEX_NEWLINE_CRLF)
+#define G_REGEX_COMPILE_MASK (G_REGEX_CASELESS | \
+ G_REGEX_MULTILINE | \
+ G_REGEX_DOTALL | \
+ G_REGEX_EXTENDED | \
+ G_REGEX_ANCHORED | \
+ G_REGEX_DOLLAR_ENDONLY | \
+ G_REGEX_UNGREEDY | \
+ G_REGEX_RAW | \
+ G_REGEX_NO_AUTO_CAPTURE | \
+ G_REGEX_OPTIMIZE | \
+ G_REGEX_DUPNAMES | \
+ G_REGEX_NEWLINE_CR | \
+ G_REGEX_NEWLINE_LF | \
+ G_REGEX_NEWLINE_CRLF)
/* Mask of all the possible values for GRegexMatchFlags. */
-#define G_REGEX_MATCH_MASK (G_REGEX_MATCH_ANCHORED | \
- G_REGEX_MATCH_NOTBOL | \
- G_REGEX_MATCH_NOTEOL | \
- G_REGEX_MATCH_NOTEMPTY | \
- G_REGEX_MATCH_PARTIAL | \
- G_REGEX_MATCH_NEWLINE_CR | \
- G_REGEX_MATCH_NEWLINE_LF | \
- G_REGEX_MATCH_NEWLINE_CRLF | \
- G_REGEX_MATCH_NEWLINE_ANY)
+#define G_REGEX_MATCH_MASK (G_REGEX_MATCH_ANCHORED | \
+ G_REGEX_MATCH_NOTBOL | \
+ G_REGEX_MATCH_NOTEOL | \
+ G_REGEX_MATCH_NOTEMPTY | \
+ G_REGEX_MATCH_PARTIAL | \
+ G_REGEX_MATCH_NEWLINE_CR | \
+ G_REGEX_MATCH_NEWLINE_LF | \
+ G_REGEX_MATCH_NEWLINE_CRLF | \
+ G_REGEX_MATCH_NEWLINE_ANY)
/* if the string is in UTF-8 use g_utf8_ functions, else use
* use just +/- 1. */
-#define NEXT_CHAR(re, s) (((re)->pattern->compile_opts & PCRE_UTF8) ? \
- g_utf8_next_char (s) : \
- ((s) + 1))
-#define PREV_CHAR(re, s) (((re)->pattern->compile_opts & PCRE_UTF8) ? \
- g_utf8_prev_char (s) : \
- ((s) - 1))
-
-#define WORKSPACE_INITIAL 1000
-#define OFFSETS_DFA_MIN_SIZE 21
-
-/* atomically returns the pcre_extra struct in the regex. */
-#define REGEX_GET_EXTRA(re) ((pcre_extra *)g_atomic_pointer_get (&(re)->pattern->extra))
-
-/* this struct can be shared by more regexes */
-typedef struct
+#define NEXT_CHAR(re, s) (((re)->compile_opts & PCRE_UTF8) ? \
+ g_utf8_next_char (s) : \
+ ((s) + 1))
+#define PREV_CHAR(re, s) (((re)->compile_opts & PCRE_UTF8) ? \
+ g_utf8_prev_char (s) : \
+ ((s) - 1))
+
+struct _GMatchInfo
{
- volatile guint ref_count; /* the ref count for the immutable part */
- gchar *pattern; /* the pattern */
- pcre *pcre_re; /* compiled form of the pattern */
- GRegexCompileFlags compile_opts; /* options used at compile time on the pattern */
- GRegexMatchFlags match_opts; /* options used at match time on the regex */
- pcre_extra *extra; /* data stored when g_regex_optimize() is used */
-} GRegexPattern;
-
-/* this struct is used only by a single regex */
-typedef struct
-{
- gint matches; /* number of matching sub patterns */
- gint pos; /* position in the string where last match left off */
- gint *offsets; /* array of offsets paired 0,1 ; 2,3 ; 3,4 etc */
- gint n_offsets; /* number of offsets */
- gint *workspace; /* workspace for pcre_dfa_exec() */
- gint n_workspace; /* number of workspace elements */
- gssize string_len; /* length of the string last used against */
- GSList *delims; /* delimiter sub strings from split next */
- gint last_separator_end; /* position of the last separator for split_next_full() */
- gboolean last_match_is_empty; /* was the last match in split_next_full() 0 bytes long? */
-} GRegexMatch;
+ volatile gint ref_count; /* the ref count */
+ GRegex *regex; /* the regex */
+ GRegexMatchFlags match_opts; /* options used at match time on the regex */
+ gint matches; /* number of matching sub patterns */
+ gint pos; /* position in the string where last match left off */
+ gint n_offsets; /* number of offsets */
+ gint *offsets; /* array of offsets paired 0,1 ; 2,3 ; 3,4 etc */
+ gint *workspace; /* workspace for pcre_dfa_exec() */
+ gint n_workspace; /* number of workspace elements */
+ const gchar *string; /* string passed to the match function */
+ gssize string_len; /* length of string */
+};
struct _GRegex
{
- GRegexPattern *pattern; /* immutable part, shared */
- GRegexMatch *match; /* mutable part, not shared */
+ volatile gint ref_count; /* the ref count for the immutable part */
+ gchar *pattern; /* the pattern */
+ pcre *pcre_re; /* compiled form of the pattern */
+ GRegexCompileFlags compile_opts; /* options used at compile time on the pattern */
+ GRegexMatchFlags match_opts; /* options used at match time on the regex */
+ pcre_extra *extra; /* data stored when G_REGEX_OPTIMIZE is used */
};
/* TRUE if ret is an error code, FALSE otherwise. */
#define IS_PCRE_ERROR(ret) ((ret) < PCRE_ERROR_NOMATCH && (ret) != PCRE_ERROR_PARTIAL)
+typedef struct _InterpolationData InterpolationData;
+static gboolean interpolation_list_needs_match (GList *list);
+static gboolean interpolate_replacement (const GMatchInfo *match_info,
+ GString *result,
+ gpointer data);
+static GList *split_replacement (const gchar *replacement,
+ GError **error);
+static void free_interpolation_data (InterpolationData *data);
+
+
static const gchar *
match_error (gint errcode)
{
switch (errcode)
- {
+ {
case PCRE_ERROR_NOMATCH:
/* not an error */
break;
case PCRE_ERROR_DFA_UCOND:
return _("back references as conditions are not supported for partial matching");
case PCRE_ERROR_DFA_UMLIMIT:
- /* the match_field field is not udes in GRegex */
+ /* the match_field field is not used in GRegex */
break;
case PCRE_ERROR_DFA_WSSIZE:
/* handled expanding the workspace */
return _("workspace limit for empty substrings reached");
case PCRE_ERROR_BADNEWLINE:
return _("invalid combination of newline flags");
+ case PCRE_ERROR_BADOFFSET:
+ return _("bad offset");
+ case PCRE_ERROR_SHORTUTF8:
+ return _("short utf8");
default:
break;
}
return _("unknown error");
}
-GQuark
-g_regex_error_quark (void)
-{
- static GQuark error_quark = 0;
-
- if (error_quark == 0)
- error_quark = g_quark_from_static_string ("g-regex-error-quark");
-
- return error_quark;
-}
-
-static GRegexPattern *
-regex_pattern_new (pcre *re,
- const gchar *pattern,
- GRegexCompileFlags compile_options,
- GRegexMatchFlags match_options)
-{
- GRegexPattern *rp = g_new0 (GRegexPattern, 1);
- rp->ref_count = 1;
- rp->pcre_re = re;
- rp->pattern = g_strdup (pattern);
- rp->compile_opts = compile_options;
- rp->match_opts = match_options;
- return rp;
-}
-
-static GRegexPattern *
-regex_pattern_ref (GRegexPattern *rp)
-{
- /* increases the ref count of the immutable part of the GRegex */
- g_atomic_int_inc ((gint*) &rp->ref_count);
- return rp;
-}
-
static void
-regex_pattern_unref (GRegexPattern *rp)
+translate_compile_error (gint *errcode, const gchar **errmsg)
{
- /* decreases the ref count of the immutable part of the GRegex
- * and deletes it if the ref count went to 0 */
- if (g_atomic_int_exchange_and_add ((gint *) &rp->ref_count, -1) - 1 == 0)
+ /* Compile errors are created adding 100 to the error code returned
+ * by PCRE.
+ * If errcode is known we put the translatable error message in
+ * erromsg. If errcode is unknown we put the generic
+ * G_REGEX_ERROR_COMPILE error code in errcode and keep the
+ * untranslated error message returned by PCRE.
+ * Note that there can be more PCRE errors with the same GRegexError
+ * and that some PCRE errors are useless for us.
+ */
+ *errcode += 100;
+
+ switch (*errcode)
{
- g_free (rp->pattern);
- if (rp->pcre_re != NULL)
- pcre_free (rp->pcre_re);
- if (rp->extra != NULL)
- pcre_free (rp->extra);
- g_free (rp);
+ case G_REGEX_ERROR_STRAY_BACKSLASH:
+ *errmsg = _("\\ at end of pattern");
+ break;
+ case G_REGEX_ERROR_MISSING_CONTROL_CHAR:
+ *errmsg = _("\\c at end of pattern");
+ break;
+ case G_REGEX_ERROR_UNRECOGNIZED_ESCAPE:
+ *errmsg = _("unrecognized character follows \\");
+ break;
+ case 137:
+ /* A number of Perl escapes are not handled by PCRE.
+ * Therefore it explicitly raises ERR37.
+ */
+ *errcode = G_REGEX_ERROR_UNRECOGNIZED_ESCAPE;
+ *errmsg = _("case-changing escapes (\\l, \\L, \\u, \\U) are not allowed here");
+ break;
+ case G_REGEX_ERROR_QUANTIFIERS_OUT_OF_ORDER:
+ *errmsg = _("numbers out of order in {} quantifier");
+ break;
+ case G_REGEX_ERROR_QUANTIFIER_TOO_BIG:
+ *errmsg = _("number too big in {} quantifier");
+ break;
+ case G_REGEX_ERROR_UNTERMINATED_CHARACTER_CLASS:
+ *errmsg = _("missing terminating ] for character class");
+ break;
+ case G_REGEX_ERROR_INVALID_ESCAPE_IN_CHARACTER_CLASS:
+ *errmsg = _("invalid escape sequence in character class");
+ break;
+ case G_REGEX_ERROR_RANGE_OUT_OF_ORDER:
+ *errmsg = _("range out of order in character class");
+ break;
+ case G_REGEX_ERROR_NOTHING_TO_REPEAT:
+ *errmsg = _("nothing to repeat");
+ break;
+ case G_REGEX_ERROR_UNRECOGNIZED_CHARACTER:
+ *errmsg = _("unrecognized character after (?");
+ break;
+ case 124:
+ *errcode = G_REGEX_ERROR_UNRECOGNIZED_CHARACTER;
+ *errmsg = _("unrecognized character after (?<");
+ break;
+ case 141:
+ *errcode = G_REGEX_ERROR_UNRECOGNIZED_CHARACTER;
+ *errmsg = _("unrecognized character after (?P");
+ break;
+ case G_REGEX_ERROR_POSIX_NAMED_CLASS_OUTSIDE_CLASS:
+ *errmsg = _("POSIX named classes are supported only within a class");
+ break;
+ case G_REGEX_ERROR_UNMATCHED_PARENTHESIS:
+ *errmsg = _("missing terminating )");
+ break;
+ case 122:
+ *errcode = G_REGEX_ERROR_UNMATCHED_PARENTHESIS;
+ *errmsg = _(") without opening (");
+ break;
+ case 129:
+ *errcode = G_REGEX_ERROR_UNMATCHED_PARENTHESIS;
+ /* translators: '(?R' and '(?[+-]digits' are both meant as (groups of)
+ * sequences here, '(?-54' would be an example for the second group.
+ */
+ *errmsg = _("(?R or (?[+-]digits must be followed by )");
+ break;
+ case G_REGEX_ERROR_INEXISTENT_SUBPATTERN_REFERENCE:
+ *errmsg = _("reference to non-existent subpattern");
+ break;
+ case G_REGEX_ERROR_UNTERMINATED_COMMENT:
+ *errmsg = _("missing ) after comment");
+ break;
+ case G_REGEX_ERROR_EXPRESSION_TOO_LARGE:
+ *errmsg = _("regular expression too large");
+ break;
+ case G_REGEX_ERROR_MEMORY_ERROR:
+ *errmsg = _("failed to get memory");
+ break;
+ case G_REGEX_ERROR_VARIABLE_LENGTH_LOOKBEHIND:
+ *errmsg = _("lookbehind assertion is not fixed length");
+ break;
+ case G_REGEX_ERROR_MALFORMED_CONDITION:
+ *errmsg = _("malformed number or name after (?(");
+ break;
+ case G_REGEX_ERROR_TOO_MANY_CONDITIONAL_BRANCHES:
+ *errmsg = _("conditional group contains more than two branches");
+ break;
+ case G_REGEX_ERROR_ASSERTION_EXPECTED:
+ *errmsg = _("assertion expected after (?(");
+ break;
+ case G_REGEX_ERROR_UNKNOWN_POSIX_CLASS_NAME:
+ *errmsg = _("unknown POSIX class name");
+ break;
+ case G_REGEX_ERROR_POSIX_COLLATING_ELEMENTS_NOT_SUPPORTED:
+ *errmsg = _("POSIX collating elements are not supported");
+ break;
+ case G_REGEX_ERROR_HEX_CODE_TOO_LARGE:
+ *errmsg = _("character value in \\x{...} sequence is too large");
+ break;
+ case G_REGEX_ERROR_INVALID_CONDITION:
+ *errmsg = _("invalid condition (?(0)");
+ break;
+ case G_REGEX_ERROR_SINGLE_BYTE_MATCH_IN_LOOKBEHIND:
+ *errmsg = _("\\C not allowed in lookbehind assertion");
+ break;
+ case G_REGEX_ERROR_INFINITE_LOOP:
+ *errmsg = _("recursive call could loop indefinitely");
+ break;
+ case G_REGEX_ERROR_MISSING_SUBPATTERN_NAME_TERMINATOR:
+ *errmsg = _("missing terminator in subpattern name");
+ break;
+ case G_REGEX_ERROR_DUPLICATE_SUBPATTERN_NAME:
+ *errmsg = _("two named subpatterns have the same name");
+ break;
+ case G_REGEX_ERROR_MALFORMED_PROPERTY:
+ *errmsg = _("malformed \\P or \\p sequence");
+ break;
+ case G_REGEX_ERROR_UNKNOWN_PROPERTY:
+ *errmsg = _("unknown property name after \\P or \\p");
+ break;
+ case G_REGEX_ERROR_SUBPATTERN_NAME_TOO_LONG:
+ *errmsg = _("subpattern name is too long (maximum 32 characters)");
+ break;
+ case G_REGEX_ERROR_TOO_MANY_SUBPATTERNS:
+ *errmsg = _("too many named subpatterns (maximum 10,000)");
+ break;
+ case G_REGEX_ERROR_INVALID_OCTAL_VALUE:
+ *errmsg = _("octal value is greater than \\377");
+ break;
+ case G_REGEX_ERROR_TOO_MANY_BRANCHES_IN_DEFINE:
+ *errmsg = _("DEFINE group contains more than one branch");
+ break;
+ case G_REGEX_ERROR_DEFINE_REPETION:
+ *errmsg = _("repeating a DEFINE group is not allowed");
+ break;
+ case G_REGEX_ERROR_INCONSISTENT_NEWLINE_OPTIONS:
+ *errmsg = _("inconsistent NEWLINE options");
+ break;
+ case G_REGEX_ERROR_MISSING_BACK_REFERENCE:
+ *errmsg = _("\\g is not followed by a braced name or an optionally "
+ "braced non-zero number");
+ break;
+ case 11:
+ *errcode = G_REGEX_ERROR_INTERNAL;
+ *errmsg = _("unexpected repeat");
+ break;
+ case 23:
+ *errcode = G_REGEX_ERROR_INTERNAL;
+ *errmsg = _("code overflow");
+ break;
+ case 52:
+ *errcode = G_REGEX_ERROR_INTERNAL;
+ *errmsg = _("overran compiling workspace");
+ break;
+ case 53:
+ *errcode = G_REGEX_ERROR_INTERNAL;
+ *errmsg = _("previously-checked referenced subpattern not found");
+ break;
+ case 16:
+ /* This should not happen as we never pass a NULL erroffset */
+ g_warning ("erroffset passed as NULL");
+ *errcode = G_REGEX_ERROR_COMPILE;
+ break;
+ case 17:
+ /* This should not happen as we check options before passing them
+ * to pcre_compile2() */
+ g_warning ("unknown option bit(s) set");
+ *errcode = G_REGEX_ERROR_COMPILE;
+ break;
+ case 32:
+ case 44:
+ case 45:
+ /* These errors should not happen as we are using an UTF8-enabled PCRE
+ * and we do not check if strings are valid */
+ g_warning ("%s", *errmsg);
+ *errcode = G_REGEX_ERROR_COMPILE;
+ break;
+ default:
+ *errcode = G_REGEX_ERROR_COMPILE;
}
}
-static void
-regex_match_free (GRegexMatch *rm)
-{
- if (rm == NULL)
- return;
-
- g_slist_foreach (rm->delims, (GFunc) g_free, NULL);
- g_slist_free (rm->delims);
- g_free (rm->offsets);
- g_free (rm->workspace);
- g_free (rm);
-}
-
-static void
-regex_lazy_init_match (GRegex *regex,
- gint min_offsets)
-{
- gint n_offsets;
-
- if (regex->match != NULL)
- return;
-
- pcre_fullinfo (regex->pattern->pcre_re,
- REGEX_GET_EXTRA (regex),
- PCRE_INFO_CAPTURECOUNT, &n_offsets);
- n_offsets = (MAX (n_offsets, min_offsets) + 1) * 3;
+/* GMatchInfo */
- regex->match = g_new0 (GRegexMatch, 1);
- regex->match->string_len = -1;
- regex->match->matches = -1000;
- regex->match->n_offsets = n_offsets;
- regex->match->offsets = g_new0 (gint, n_offsets);
-}
-
-/**
- * g_regex_new:
- * @pattern: the regular expression.
- * @compile_options: compile options for the regular expression.
- * @match_options: match options for the regular expression.
- * @error: return location for a #GError.
- *
- * Compiles the regular expression to an internal form, and does the initial
- * setup of the #GRegex structure.
- *
- * Returns: a #GRegex structure.
- *
- * Since: 2.14
- */
-GRegex *
-g_regex_new (const gchar *pattern,
- GRegexCompileFlags compile_options,
- GRegexMatchFlags match_options,
- GError **error)
+static GMatchInfo *
+match_info_new (const GRegex *regex,
+ const gchar *string,
+ gint string_len,
+ gint start_position,
+ gint match_options,
+ gboolean is_dfa)
{
- pcre *re;
- const gchar *errmsg;
- gint erroffset;
- static gboolean initialized = FALSE;
-
- g_return_val_if_fail (pattern != NULL, NULL);
- g_return_val_if_fail (error == NULL || *error == NULL, NULL);
- g_return_val_if_fail ((compile_options & ~G_REGEX_COMPILE_MASK) == 0, NULL);
- g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, NULL);
-
- if (!initialized)
- {
- gint support;
- const gchar *msg;
-
- pcre_config (PCRE_CONFIG_UTF8, &support);
- if (!support)
- {
- msg = N_("PCRE library is compiled without UTF8 support");
- g_critical (msg);
- g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_COMPILE, gettext (msg));
- return NULL;
- }
-
- pcre_config (PCRE_CONFIG_UNICODE_PROPERTIES, &support);
- if (!support)
- {
- msg = N_("PCRE library is compiled without UTF8 properties support");
- g_critical (msg);
- g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_COMPILE, gettext (msg));
- return NULL;
- }
-
- initialized = TRUE;
- }
+ GMatchInfo *match_info;
- /* In GRegex the string are, by default, UTF-8 encoded. PCRE
- * instead uses UTF-8 only if required with PCRE_UTF8. */
- if (compile_options & G_REGEX_RAW)
+ if (string_len < 0)
+ string_len = strlen (string);
+
+ match_info = g_new0 (GMatchInfo, 1);
+ match_info->ref_count = 1;
+ match_info->regex = g_regex_ref ((GRegex *)regex);
+ match_info->string = string;
+ match_info->string_len = string_len;
+ match_info->matches = PCRE_ERROR_NOMATCH;
+ match_info->pos = start_position;
+ match_info->match_opts = match_options;
+
+ if (is_dfa)
{
- /* disable utf-8 */
- compile_options &= ~G_REGEX_RAW;
+ /* These values should be enough for most cases, if they are not
+ * enough g_regex_match_all_full() will expand them. */
+ match_info->n_offsets = 24;
+ match_info->n_workspace = 100;
+ match_info->workspace = g_new (gint, match_info->n_workspace);
}
else
{
- /* enable utf-8 */
- compile_options |= PCRE_UTF8 | PCRE_NO_UTF8_CHECK;
- match_options |= PCRE_NO_UTF8_CHECK;
+ gint capture_count;
+ pcre_fullinfo (regex->pcre_re, regex->extra,
+ PCRE_INFO_CAPTURECOUNT, &capture_count);
+ match_info->n_offsets = (capture_count + 1) * 3;
}
- /* compile the pattern */
- re = pcre_compile (pattern, compile_options, &errmsg, &erroffset, NULL);
-
- /* if the compilation failed, set the error member and return
- * immediately */
- if (re == NULL)
- {
- GError *tmp_error = g_error_new (G_REGEX_ERROR,
- G_REGEX_ERROR_COMPILE,
- _("Error while compiling regular "
- "expression %s at char %d: %s"),
- pattern, erroffset, errmsg);
- g_propagate_error (error, tmp_error);
+ match_info->offsets = g_new0 (gint, match_info->n_offsets);
+ /* Set an invalid position for the previous match. */
+ match_info->offsets[0] = -1;
+ match_info->offsets[1] = -1;
- return NULL;
- }
- else
- {
- GRegex *regex = g_new0 (GRegex, 1);
- regex->pattern = regex_pattern_new (re, pattern,
- compile_options, match_options);
- return regex;
- }
+ return match_info;
}
/**
- * g_regex_free:
- * @regex: a #GRegex.
+ * g_match_info_get_regex:
+ * @match_info: a #GMatchInfo
+ *
+ * Returns #GRegex object used in @match_info. It belongs to Glib
+ * and must not be freed. Use g_regex_ref() if you need to keep it
+ * after you free @match_info object.
*
- * Frees all the memory associated with the regex structure.
+ * Returns: #GRegex object used in @match_info
*
* Since: 2.14
*/
-void
-g_regex_free (GRegex *regex)
+GRegex *
+g_match_info_get_regex (const GMatchInfo *match_info)
{
- if (regex == NULL)
- return;
-
- regex_pattern_unref (regex->pattern);
- regex_match_free (regex->match);
- g_free (regex);
+ g_return_val_if_fail (match_info != NULL, NULL);
+ return match_info->regex;
}
/**
- * g_regex_copy:
- * @regex: a #GRegex structure from g_regex_new().
+ * g_match_info_get_string:
+ * @match_info: a #GMatchInfo
*
- * Copies a #GRegex. The returned #Gregex is in the same state as after
- * a call to g_regex_clear(), so it does not contain information on the
- * last match. If @regex is %NULL it returns %NULL.
+ * Returns the string searched with @match_info. This is the
+ * string passed to g_regex_match() or g_regex_replace() so
+ * you may not free it before calling this function.
*
- * The returned copy shares some of its internal state with the original
- * @regex, and the other internal variables are created only when needed,
- * so the copy is a lightweight operation.
- *
- * Returns: a newly allocated copy of @regex, or %NULL if an error
- * occurred.
+ * Returns: the string searched with @match_info
*
* Since: 2.14
*/
-GRegex *
-g_regex_copy (const GRegex *regex)
+const gchar *
+g_match_info_get_string (const GMatchInfo *match_info)
{
- GRegex *copy;
-
- if (regex == NULL)
- return NULL;
-
- copy = g_new0 (GRegex, 1);
- copy->pattern = regex_pattern_ref (regex->pattern);
-
- return copy;
+ g_return_val_if_fail (match_info != NULL, NULL);
+ return match_info->string;
}
/**
- * g_regex_get_pattern:
- * @regex: a #GRegex structure.
+ * g_match_info_ref:
+ * @match_info: a #GMatchInfo
*
- * Gets the pattern string associated with @regex, i.e. a copy of the string passed
- * to g_regex_new().
+ * Increases reference count of @match_info by 1.
*
- * Returns: the pattern of @regex.
+ * Returns: @match_info
*
- * Since: 2.14
+ * Since: 2.30
*/
-const gchar *
-g_regex_get_pattern (const GRegex *regex)
+GMatchInfo *
+g_match_info_ref (GMatchInfo *match_info)
{
- g_return_val_if_fail (regex != NULL, NULL);
+ g_return_val_if_fail (match_info != NULL, NULL);
+ g_atomic_int_inc (&match_info->ref_count);
+ return match_info;
+}
- return regex->pattern->pattern;
+/**
+ * g_match_info_unref:
+ * @match_info: a #GMatchInfo
+ *
+ * Decreases reference count of @match_info by 1. When reference count drops
+ * to zero, it frees all the memory associated with the match_info structure.
+ *
+ * Since: 2.30
+ */
+void
+g_match_info_unref (GMatchInfo *match_info)
+{
+ if (g_atomic_int_dec_and_test (&match_info->ref_count))
+ {
+ g_regex_unref (match_info->regex);
+ g_free (match_info->offsets);
+ g_free (match_info->workspace);
+ g_free (match_info);
+ }
}
/**
- * g_regex_clear:
- * @regex: a #GRegex structure.
+ * g_match_info_free:
+ * @match_info: (allow-none): a #GMatchInfo, or %NULL
*
- * Clears out the members of @regex that are holding information about the
- * last set of matches for this pattern. g_regex_clear() needs to be
- * called between uses of g_regex_match_next() or g_regex_match_next_full()
- * against new target strings.
+ * If @match_info is not %NULL, calls g_match_info_unref(); otherwise does
+ * nothing.
*
* Since: 2.14
*/
void
-g_regex_clear (GRegex *regex)
+g_match_info_free (GMatchInfo *match_info)
{
- g_return_if_fail (regex != NULL);
-
- if (regex->match == NULL)
+ if (match_info == NULL)
return;
- regex->match->matches = -1000; /* an error code not used by PCRE */
- regex->match->string_len = -1;
- regex->match->pos = 0;
-
- /* if the pattern was used with g_regex_split_next(), it may have
- * delimiter offsets stored. Free up those guys as well. */
- if (regex->match->delims != NULL)
- {
- g_slist_foreach (regex->match->delims, (GFunc) g_free, NULL);
- g_slist_free (regex->match->delims);
- regex->match->delims = NULL;
- }
+ g_match_info_unref (match_info);
}
/**
- * g_regex_optimize:
- * @regex: a #GRegex structure.
- * @error: return location for a #GError.
+ * g_match_info_next:
+ * @match_info: a #GMatchInfo structure
+ * @error: location to store the error occurring, or %NULL to ignore errors
*
- * If the pattern will be used many times, then it may be worth the
- * effort to optimize it to improve the speed of matches.
+ * Scans for the next match using the same parameters of the previous
+ * call to g_regex_match_full() or g_regex_match() that returned
+ * @match_info.
*
- * Returns: %TRUE if @regex has been optimized or was already optimized,
- * %FALSE otherwise.
+ * The match is done on the string passed to the match function, so you
+ * cannot free it before calling this function.
+ *
+ * Returns: %TRUE is the string matched, %FALSE otherwise
*
* Since: 2.14
*/
gboolean
-g_regex_optimize (GRegex *regex,
- GError **error)
+g_match_info_next (GMatchInfo *match_info,
+ GError **error)
{
- const gchar *errmsg;
- pcre_extra *extra;
- pcre_extra G_GNUC_MAY_ALIAS **extra_p;
+ gint prev_match_start;
+ gint prev_match_end;
- g_return_val_if_fail (regex != NULL, FALSE);
+ g_return_val_if_fail (match_info != NULL, FALSE);
g_return_val_if_fail (error == NULL || *error == NULL, FALSE);
+ g_return_val_if_fail (match_info->pos >= 0, FALSE);
- if (REGEX_GET_EXTRA (regex) != NULL)
- /* already optimized. */
- return TRUE;
+ prev_match_start = match_info->offsets[0];
+ prev_match_end = match_info->offsets[1];
- extra = pcre_study (regex->pattern->pcre_re, 0, &errmsg);
+ if (match_info->pos > match_info->string_len)
+ {
+ /* we have reached the end of the string */
+ match_info->pos = -1;
+ match_info->matches = PCRE_ERROR_NOMATCH;
+ return FALSE;
+ }
- if (errmsg != NULL)
+ match_info->matches = pcre_exec (match_info->regex->pcre_re,
+ match_info->regex->extra,
+ match_info->string,
+ match_info->string_len,
+ match_info->pos,
+ match_info->regex->match_opts | match_info->match_opts,
+ match_info->offsets,
+ match_info->n_offsets);
+ if (IS_PCRE_ERROR (match_info->matches))
{
- GError *tmp_error = g_error_new (G_REGEX_ERROR,
- G_REGEX_ERROR_OPTIMIZE,
- _("Error while optimizing "
- "regular expression %s: %s"),
- regex->pattern->pattern,
- errmsg);
- g_propagate_error (error, tmp_error);
+ g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_MATCH,
+ _("Error while matching regular expression %s: %s"),
+ match_info->regex->pattern, match_error (match_info->matches));
return FALSE;
}
- if (extra == NULL)
- return TRUE;
+ /* avoid infinite loops if the pattern is an empty string or something
+ * equivalent */
+ if (match_info->pos == match_info->offsets[1])
+ {
+ if (match_info->pos > match_info->string_len)
+ {
+ /* we have reached the end of the string */
+ match_info->pos = -1;
+ match_info->matches = PCRE_ERROR_NOMATCH;
+ return FALSE;
+ }
+
+ match_info->pos = NEXT_CHAR (match_info->regex,
+ &match_info->string[match_info->pos]) -
+ match_info->string;
+ }
+ else
+ {
+ match_info->pos = match_info->offsets[1];
+ }
- extra_p = ®ex->pattern->extra;
- if (!g_atomic_pointer_compare_and_exchange ((gpointer *)extra_p, NULL, extra))
- /* someone else has optimized the regex while this function was running */
- pcre_free (extra);
+ /* it's possible to get two identical matches when we are matching
+ * empty strings, for instance if the pattern is "(?=[A-Z0-9])" and
+ * the string is "RegExTest" we have:
+ * - search at position 0: match from 0 to 0
+ * - search at position 1: match from 3 to 3
+ * - search at position 3: match from 3 to 3 (duplicate)
+ * - search at position 4: match from 5 to 5
+ * - search at position 5: match from 5 to 5 (duplicate)
+ * - search at position 6: no match -> stop
+ * so we have to ignore the duplicates.
+ * see bug #515944: http://bugzilla.gnome.org/show_bug.cgi?id=515944 */
+ if (match_info->matches >= 0 &&
+ prev_match_start == match_info->offsets[0] &&
+ prev_match_end == match_info->offsets[1])
+ {
+ /* ignore this match and search the next one */
+ return g_match_info_next (match_info, error);
+ }
- return TRUE;
+ return match_info->matches >= 0;
}
/**
- * g_regex_match_simple:
- * @pattern: the regular expression.
- * @string: the string to scan for matches.
- * @compile_options: compile options for the regular expression.
- * @match_options: match options.
+ * g_match_info_matches:
+ * @match_info: a #GMatchInfo structure
*
- * Scans for a match in @string for @pattern.
- *
- * This function is equivalent to g_regex_match() but it does not
- * require to compile the pattern with g_regex_new(), avoiding some
- * lines of code when you need just to do a match without extracting
- * substrings, capture counts, and so on.
- *
- * If this function is to be called on the same @pattern more than
- * once, it's more efficient to compile the pattern once with
- * g_regex_new() and then use g_regex_match().
+ * Returns whether the previous match operation succeeded.
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
+ * Returns: %TRUE if the previous match operation succeeded,
+ * %FALSE otherwise
*
* Since: 2.14
*/
gboolean
-g_regex_match_simple (const gchar *pattern,
- const gchar *string,
- GRegexCompileFlags compile_options,
- GRegexMatchFlags match_options)
+g_match_info_matches (const GMatchInfo *match_info)
{
- GRegex *regex;
- gboolean result;
+ g_return_val_if_fail (match_info != NULL, FALSE);
- regex = g_regex_new (pattern, compile_options, 0, NULL);
- if (!regex)
- return FALSE;
- result = g_regex_match_full (regex, string, -1, 0, match_options, NULL);
- g_regex_free (regex);
- return result;
+ return match_info->matches >= 0;
}
/**
- * g_regex_match:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: the string to scan for matches.
- * @match_options: match options.
+ * g_match_info_get_match_count:
+ * @match_info: a #GMatchInfo structure
+ *
+ * Retrieves the number of matched substrings (including substring 0,
+ * that is the whole matched text), so 1 is returned if the pattern
+ * has no substrings in it and 0 is returned if the match failed.
*
- * Scans for a match in string for the pattern in @regex. The @match_options
- * are combined with the match options specified when the @regex structure
- * was created, letting you have more flexibility in reusing #GRegex
- * structures.
+ * If the last match was obtained using the DFA algorithm, that is
+ * using g_regex_match_all() or g_regex_match_all_full(), the retrieved
+ * count is not that of the number of capturing parentheses but that of
+ * the number of matched substrings.
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
+ * Returns: Number of matched substrings, or -1 if an error occurred
*
* Since: 2.14
*/
-gboolean
-g_regex_match (GRegex *regex,
- const gchar *string,
- GRegexMatchFlags match_options)
+gint
+g_match_info_get_match_count (const GMatchInfo *match_info)
{
- return g_regex_match_full (regex, string, -1, 0,
- match_options, NULL);
+ g_return_val_if_fail (match_info, -1);
+
+ if (match_info->matches == PCRE_ERROR_NOMATCH)
+ /* no match */
+ return 0;
+ else if (match_info->matches < PCRE_ERROR_NOMATCH)
+ /* error */
+ return -1;
+ else
+ /* match */
+ return match_info->matches;
}
/**
- * g_regex_match_full:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: the string to scan for matches.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @match_options: match options.
- * @error: location to store the error occuring, or NULL to ignore errors.
+ * g_match_info_is_partial_match:
+ * @match_info: a #GMatchInfo structure
+ *
+ * Usually if the string passed to g_regex_match*() matches as far as
+ * it goes, but is too short to match the entire pattern, %FALSE is
+ * returned. There are circumstances where it might be helpful to
+ * distinguish this case from other cases in which there is no match.
+ *
+ * Consider, for example, an application where a human is required to
+ * type in data for a field with specific formatting requirements. An
+ * example might be a date in the form ddmmmyy, defined by the pattern
+ * "^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$".
+ * If the application sees the user’s keystrokes one by one, and can
+ * check that what has been typed so far is potentially valid, it is
+ * able to raise an error as soon as a mistake is made.
+ *
+ * GRegex supports the concept of partial matching by means of the
+ * #G_REGEX_MATCH_PARTIAL flag. When this is set the return code for
+ * g_regex_match() or g_regex_match_full() is, as usual, %TRUE
+ * for a complete match, %FALSE otherwise. But, when these functions
+ * return %FALSE, you can check if the match was partial calling
+ * g_match_info_is_partial_match().
*
- * Scans for a match in string for the pattern in @regex. The @match_options
- * are combined with the match options specified when the @regex structure
- * was created, letting you have more flexibility in reusing #GRegex
- * structures.
+ * When using partial matching you cannot use g_match_info_fetch*().
*
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
+ * Because of the way certain internal optimizations are implemented
+ * the partial matching algorithm cannot be used with all patterns.
+ * So repeated single characters such as "a{2,4}" and repeated single
+ * meta-sequences such as "\d+" are not permitted if the maximum number
+ * of occurrences is greater than one. Optional items such as "\d?"
+ * (where the maximum is one) are permitted. Quantifiers with any values
+ * are permitted after parentheses, so the invalid examples above can be
+ * coded thus "(a){2,4}" and "(\d)+". If #G_REGEX_MATCH_PARTIAL is set
+ * for a pattern that does not conform to the restrictions, matching
+ * functions return an error.
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
+ * Returns: %TRUE if the match was partial, %FALSE otherwise
*
* Since: 2.14
*/
gboolean
-g_regex_match_full (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- GRegexMatchFlags match_options,
- GError **error)
+g_match_info_is_partial_match (const GMatchInfo *match_info)
{
- g_return_val_if_fail (regex != NULL, FALSE);
- g_return_val_if_fail (string != NULL, FALSE);
- g_return_val_if_fail (start_position >= 0, FALSE);
- g_return_val_if_fail (error == NULL || *error == NULL, FALSE);
- g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, FALSE);
-
- regex_lazy_init_match (regex, 0);
-
- if (string_len < 0)
- string_len = strlen(string);
-
- regex->match->string_len = string_len;
+ g_return_val_if_fail (match_info != NULL, FALSE);
- /* create regex->match->offsets if it does not exist */
- regex_lazy_init_match (regex, 0);
-
- /* perform the match */
- regex->match->matches = pcre_exec (regex->pattern->pcre_re,
- REGEX_GET_EXTRA (regex),
- string, regex->match->string_len,
- start_position,
- regex->pattern->match_opts | match_options,
- regex->match->offsets, regex->match->n_offsets);
- if (IS_PCRE_ERROR (regex->match->matches))
- {
- g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_MATCH,
- _("Error while matching regular expression %s: %s"),
- regex->pattern->pattern, match_error (regex->match->matches));
- return FALSE;
- }
-
- /* set regex->match->pos to -1 so that a call to g_regex_match_next()
- * fails without a previous call to g_regex_clear(). */
- regex->match->pos = -1;
-
- return regex->match->matches >= 0;
+ return match_info->matches == PCRE_ERROR_PARTIAL;
}
/**
- * g_regex_match_next:
- * @regex: a #GRegex structure.
- * @string: the string to scan for matches.
- * @match_options: the match options.
+ * g_match_info_expand_references:
+ * @match_info: a #GMatchInfo or %NULL
+ * @string_to_expand: the string to expand
+ * @error: location to store the error occurring, or %NULL to ignore errors
*
- * Scans for the next match in @string of the pattern in @regex.
- * array. The match options are combined with the match options set when
- * the @regex was created.
- *
- * You have to call g_regex_clear() to reuse the same pattern on a new
- * string.
+ * Returns a new string containing the text in @string_to_expand with
+ * references and escape sequences expanded. References refer to the last
+ * match done with @string against @regex and have the same syntax used by
+ * g_regex_replace().
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
- *
- * Since: 2.14
- */
-gboolean
-g_regex_match_next (GRegex *regex,
- const gchar *string,
- GRegexMatchFlags match_options)
-{
- return g_regex_match_next_full (regex, string, -1, 0,
- match_options, NULL);
-}
-
-/**
- * g_regex_match_next_full:
- * @regex: a #GRegex structure.
- * @string: the string to scan for matches.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @match_options: the match options.
- * @error: location to store the error occuring, or NULL to ignore errors.
- *
- * Scans for the next match in @string of the pattern in @regex. Calling
- * g_regex_match_next_full() until it returns %FALSE, you can retrieve
- * all the non-overlapping matches of the pattern in @string. Empty matches
- * are included, so matching the string "ab" with the pattern "b*" will
- * find three matches: "" at position 0, "b" from position 1 to 2 and
- * "" at position 2.
- *
- * The match options are combined with the match options set when the
- * @regex was created.
+ * The @string_to_expand must be UTF-8 encoded even if #G_REGEX_RAW was
+ * passed to g_regex_new().
*
- * You have to call g_regex_clear() to reuse the same pattern on a new
- * string.
+ * The backreferences are extracted from the string passed to the match
+ * function, so you cannot call this function after freeing the string.
*
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
+ * @match_info may be %NULL in which case @string_to_expand must not
+ * contain references. For instance "foo\n" does not refer to an actual
+ * pattern and '\n' merely will be replaced with \n character,
+ * while to expand "\0" (whole match) one needs the result of a match.
+ * Use g_regex_check_replacement() to find out whether @string_to_expand
+ * contains references.
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
+ * Returns: (allow-none): the expanded string, or %NULL if an error occurred
*
* Since: 2.14
*/
-gboolean
-g_regex_match_next_full (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- GRegexMatchFlags match_options,
- GError **error)
+gchar *
+g_match_info_expand_references (const GMatchInfo *match_info,
+ const gchar *string_to_expand,
+ GError **error)
{
- g_return_val_if_fail (regex != NULL, FALSE);
- g_return_val_if_fail (string != NULL, FALSE);
- g_return_val_if_fail (start_position >= 0, FALSE);
- g_return_val_if_fail (error == NULL || *error == NULL, FALSE);
- g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, FALSE);
+ GString *result;
+ GList *list;
+ GError *tmp_error = NULL;
- regex_lazy_init_match (regex, 0);
+ g_return_val_if_fail (string_to_expand != NULL, NULL);
+ g_return_val_if_fail (error == NULL || *error == NULL, NULL);
- if (G_UNLIKELY (regex->match->pos < 0))
+ list = split_replacement (string_to_expand, &tmp_error);
+ if (tmp_error != NULL)
{
- const gchar *msg = _("g_regex_match_next_full: called without a "
- "previous call to g_regex_clear()");
- g_log (G_LOG_DOMAIN, G_LOG_LEVEL_CRITICAL, msg);
- g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_MATCH, msg);
- return FALSE;
+ g_propagate_error (error, tmp_error);
+ return NULL;
}
- /* if this regex hasn't been used on this string before, then we
- * need to calculate the length of the string, and set pos to the
- * start of it.
- * Knowing if this regex has been used on this string is a bit of
- * a challenge. For now, we require the user to call g_regex_clear()
- * in between usages on a new string. Not perfect, but not such a
- * bad solution either.
- */
- if (regex->match->string_len == -1)
+ if (!match_info && interpolation_list_needs_match (list))
{
- if (string_len < 0)
- string_len = strlen(string);
- regex->match->string_len = string_len;
-
- regex->match->pos = start_position;
+ g_critical ("String '%s' contains references to the match, can't "
+ "expand references without GMatchInfo object",
+ string_to_expand);
+ return NULL;
}
- /* create regex->match->offsets if it does not exist */
- regex_lazy_init_match (regex, 0);
-
- /* perform the match */
- regex->match->matches = pcre_exec (regex->pattern->pcre_re,
- REGEX_GET_EXTRA (regex),
- string, regex->match->string_len,
- regex->match->pos,
- regex->pattern->match_opts | match_options,
- regex->match->offsets, regex->match->n_offsets);
- if (IS_PCRE_ERROR (regex->match->matches))
- {
- g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_MATCH,
- _("Error while matching regular expression %s: %s"),
- regex->pattern->pattern, match_error (regex->match->matches));
- return FALSE;
- }
+ result = g_string_sized_new (strlen (string_to_expand));
+ interpolate_replacement (match_info, result, list);
- /* avoid infinite loops if regex is an empty string or something
- * equivalent */
- if (regex->match->pos == regex->match->offsets[1])
- {
- if (regex->match->pos > regex->match->string_len)
- {
- /* we have reached the end of the string */
- regex->match->pos = -1;
- return FALSE;
- }
- regex->match->pos = NEXT_CHAR (regex, &string[regex->match->pos]) - string;
- }
- else
- {
- regex->match->pos = regex->match->offsets[1];
- }
+ g_list_foreach (list, (GFunc)free_interpolation_data, NULL);
+ g_list_free (list);
- return regex->match->matches >= 0;
+ return g_string_free (result, FALSE);
}
/**
- * g_regex_match_all:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: the string to scan for matches.
- * @match_options: match options.
+ * g_match_info_fetch:
+ * @match_info: #GMatchInfo structure
+ * @match_num: number of the sub expression
*
- * Using the standard algorithm for regular expression matching only the
- * longest match in the string is retrieved. This function uses a
- * different algorithm so it can retrieve all the possible matches.
- * For more documentation see g_regex_match_all_full().
+ * Retrieves the text matching the @match_num<!-- -->'th capturing
+ * parentheses. 0 is the full text of the match, 1 is the first paren
+ * set, 2 the second, and so on.
+ *
+ * If @match_num is a valid sub pattern but it didn't match anything
+ * (e.g. sub pattern 1, matching "b" against "(a)?b") then an empty
+ * string is returned.
+ *
+ * If the match was obtained using the DFA algorithm, that is using
+ * g_regex_match_all() or g_regex_match_all_full(), the retrieved
+ * string is not that of a set of parentheses but that of a matched
+ * substring. Substrings are matched in reverse order of length, so
+ * 0 is the longest match.
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
+ * The string is fetched from the string passed to the match function,
+ * so you cannot call this function after freeing the string.
+ *
+ * Returns: (allow-none): The matched substring, or %NULL if an error
+ * occurred. You have to free the string yourself
*
* Since: 2.14
*/
-gboolean
-g_regex_match_all (GRegex *regex,
- const gchar *string,
- GRegexMatchFlags match_options)
+gchar *
+g_match_info_fetch (const GMatchInfo *match_info,
+ gint match_num)
{
- return g_regex_match_all_full (regex, string, -1, 0,
- match_options, NULL);
+ /* we cannot use pcre_get_substring() because it allocates the
+ * string using pcre_malloc(). */
+ gchar *match = NULL;
+ gint start, end;
+
+ g_return_val_if_fail (match_info != NULL, NULL);
+ g_return_val_if_fail (match_num >= 0, NULL);
+
+ /* match_num does not exist or it didn't matched, i.e. matching "b"
+ * against "(a)?b" then group 0 is empty. */
+ if (!g_match_info_fetch_pos (match_info, match_num, &start, &end))
+ match = NULL;
+ else if (start == -1)
+ match = g_strdup ("");
+ else
+ match = g_strndup (&match_info->string[start], end - start);
+
+ return match;
}
/**
- * g_regex_match_all_full:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: the string to scan for matches.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @match_options: match options.
- * @error: location to store the error occuring, or NULL to ignore errors.
- *
- * Using the standard algorithm for regular expression matching only the
- * longest match in the string is retrieved, it is not possibile to obtain
- * all the available matches. For instance matching
- * "<a> <b> <c>" against the pattern "<.*>" you get
- * "<a> <b> <c>".
- *
- * This function uses a different algorithm (called DFA, i.e. deterministic
- * finite automaton), so it can retrieve all the possible matches, all
- * starting at the same point in the string. For instance matching
- * "<a> <b> <c>" against the pattern "<.*>" you
- * would obtain three matches: "<a> <b> <c>",
- * "<a> <b>" and "<a>".
- *
- * The number of matched strings is retrieved using
- * g_regex_get_match_count().
- * To obtain the matched strings and their position you can use,
- * respectively, g_regex_fetch() and g_regex_fetch_pos(). Note that the
- * strings are returned in reverse order of length; that is, the longest
- * matching string is given first.
- *
- * Note that the DFA algorithm is slower than the standard one and it is not
- * able to capture substrings, so backreferences do not work.
- *
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
+ * g_match_info_fetch_pos:
+ * @match_info: #GMatchInfo structure
+ * @match_num: number of the sub expression
+ * @start_pos: (out) (allow-none): pointer to location where to store
+ * the start position, or %NULL
+ * @end_pos: (out) (allow-none): pointer to location where to store
+ * the end position, or %NULL
+ *
+ * Retrieves the position in bytes of the @match_num<!-- -->'th capturing
+ * parentheses. 0 is the full text of the match, 1 is the first
+ * paren set, 2 the second, and so on.
+ *
+ * If @match_num is a valid sub pattern but it didn't match anything
+ * (e.g. sub pattern 1, matching "b" against "(a)?b") then @start_pos
+ * and @end_pos are set to -1 and %TRUE is returned.
+ *
+ * If the match was obtained using the DFA algorithm, that is using
+ * g_regex_match_all() or g_regex_match_all_full(), the retrieved
+ * position is not that of a set of parentheses but that of a matched
+ * substring. Substrings are matched in reverse order of length, so
+ * 0 is the longest match.
*
- * Returns: %TRUE is the string matched, %FALSE otherwise.
+ * Returns: %TRUE if the position was fetched, %FALSE otherwise. If
+ * the position cannot be fetched, @start_pos and @end_pos are left
+ * unchanged
*
* Since: 2.14
*/
gboolean
-g_regex_match_all_full (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- GRegexMatchFlags match_options,
- GError **error)
+g_match_info_fetch_pos (const GMatchInfo *match_info,
+ gint match_num,
+ gint *start_pos,
+ gint *end_pos)
{
- g_return_val_if_fail (regex != NULL, FALSE);
- g_return_val_if_fail (string != NULL, FALSE);
- g_return_val_if_fail (start_position >= 0, FALSE);
- g_return_val_if_fail (error == NULL || *error == NULL, FALSE);
- g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, FALSE);
+ g_return_val_if_fail (match_info != NULL, FALSE);
+ g_return_val_if_fail (match_num >= 0, FALSE);
- regex_lazy_init_match (regex, 0);
+ /* make sure the sub expression number they're requesting is less than
+ * the total number of sub expressions that were matched. */
+ if (match_num >= match_info->matches)
+ return FALSE;
- if (string_len < 0)
- string_len = strlen(string);
+ if (start_pos != NULL)
+ *start_pos = match_info->offsets[2 * match_num];
- regex->match->string_len = string_len;
+ if (end_pos != NULL)
+ *end_pos = match_info->offsets[2 * match_num + 1];
- if (regex->match->workspace == NULL)
- {
- regex->match->n_workspace = WORKSPACE_INITIAL;
- regex->match->workspace = g_new (gint, regex->match->n_workspace);
- }
+ return TRUE;
+}
- if (regex->match->n_offsets < OFFSETS_DFA_MIN_SIZE)
- {
- regex->match->n_offsets = OFFSETS_DFA_MIN_SIZE;
- regex->match->offsets = g_realloc (regex->match->offsets,
- regex->match->n_offsets * sizeof(gint));
- }
+/*
+ * Returns number of first matched subpattern with name @name.
+ * There may be more than one in case when DUPNAMES is used,
+ * and not all subpatterns with that name match;
+ * pcre_get_stringnumber() does not work in that case.
+ */
+static gint
+get_matched_substring_number (const GMatchInfo *match_info,
+ const gchar *name)
+{
+ gint entrysize;
+ gchar *first, *last;
+ guchar *entry;
- /* perform the match */
- regex->match->matches = pcre_dfa_exec (regex->pattern->pcre_re,
- REGEX_GET_EXTRA (regex),
- string, regex->match->string_len,
- start_position,
- regex->pattern->match_opts | match_options,
- regex->match->offsets, regex->match->n_offsets,
- regex->match->workspace,
- regex->match->n_workspace);
- if (regex->match->matches == PCRE_ERROR_DFA_WSSIZE)
- {
- /* regex->match->workspace is too small. */
- regex->match->n_workspace *= 2;
- regex->match->workspace = g_realloc (regex->match->workspace,
- regex->match->n_workspace * sizeof(gint));
- return g_regex_match_all_full (regex, string, string_len,
- start_position, match_options, error);
- }
- else if (regex->match->matches == 0)
- {
- /* regex->match->offsets is too small. */
- regex->match->n_offsets *= 2;
- regex->match->offsets = g_realloc (regex->match->offsets,
- regex->match->n_offsets * sizeof(gint));
- return g_regex_match_all_full (regex, string, string_len,
- start_position, match_options, error);
- }
- else if (IS_PCRE_ERROR (regex->match->matches))
- {
- g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_MATCH,
- _("Error while matching regular expression %s: %s"),
- regex->pattern->pattern, match_error (regex->match->matches));
- return FALSE;
- }
+ if (!(match_info->regex->compile_opts & G_REGEX_DUPNAMES))
+ return pcre_get_stringnumber (match_info->regex->pcre_re, name);
- /* set regex->match->pos to -1 so that a call to g_regex_match_next()
- * fails without a previous call to g_regex_clear(). */
- regex->match->pos = -1;
+ /* This code is copied from pcre_get.c: get_first_set() */
+ entrysize = pcre_get_stringtable_entries (match_info->regex->pcre_re,
+ name,
+ &first,
+ &last);
+
+ if (entrysize <= 0)
+ return entrysize;
+
+ for (entry = (guchar*) first; entry <= (guchar*) last; entry += entrysize)
+ {
+ gint n = (entry[0] << 8) + entry[1];
+ if (match_info->offsets[n*2] >= 0)
+ return n;
+ }
- return regex->match->matches >= 0;
+ return (first[0] << 8) + first[1];
}
/**
- * g_regex_get_match_count:
- * @regex: a #GRegex structure.
+ * g_match_info_fetch_named:
+ * @match_info: #GMatchInfo structure
+ * @name: name of the subexpression
*
- * Retrieves the number of matched substrings (including substring 0, that
- * is the whole matched text) in the last call to g_regex_match*(), so 1
- * is returned if the pattern has no substrings in it and 0 is returned if
- * the match failed.
+ * Retrieves the text matching the capturing parentheses named @name.
*
- * If the last match was obtained using the DFA algorithm, that is using
- * g_regex_match_all() or g_regex_match_all_full(), the retrieved
- * count is not that of the number of capturing parentheses but that of
- * the number of matched substrings.
+ * If @name is a valid sub pattern name but it didn't match anything
+ * (e.g. sub pattern "X", matching "b" against "(?P<X>a)?b")
+ * then an empty string is returned.
+ *
+ * The string is fetched from the string passed to the match function,
+ * so you cannot call this function after freeing the string.
*
- * Returns: Number of matched substrings, or -1 if an error occurred.
+ * Returns: (allow-none): The matched substring, or %NULL if an error
+ * occurred. You have to free the string yourself
*
* Since: 2.14
*/
-gint
-g_regex_get_match_count (const GRegex *regex)
+gchar *
+g_match_info_fetch_named (const GMatchInfo *match_info,
+ const gchar *name)
{
- g_return_val_if_fail (regex != NULL, -1);
+ /* we cannot use pcre_get_named_substring() because it allocates the
+ * string using pcre_malloc(). */
+ gint num;
- if (regex->match == NULL)
- return -1;
+ g_return_val_if_fail (match_info != NULL, NULL);
+ g_return_val_if_fail (name != NULL, NULL);
- if (regex->match->matches == PCRE_ERROR_NOMATCH)
- /* no match */
- return 0;
- else if (regex->match->matches < PCRE_ERROR_NOMATCH)
- /* error */
- return -1;
+ num = get_matched_substring_number (match_info, name);
+ if (num < 0)
+ return NULL;
else
- /* match */
- return regex->match->matches;
+ return g_match_info_fetch (match_info, num);
}
/**
- * g_regex_is_partial_match:
- * @regex: a #GRegex structure.
- *
- * Usually if the string passed to g_regex_match*() matches as far as
- * it goes, but is too short to match the entire pattern, %FALSE is
- * returned. There are circumstances where it might be helpful to
- * distinguish this case from other cases in which there is no match.
- *
- * Consider, for example, an application where a human is required to
- * type in data for a field with specific formatting requirements. An
- * example might be a date in the form ddmmmyy, defined by the pattern
- * "^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$".
- * If the application sees the user’s keystrokes one by one, and can
- * check that what has been typed so far is potentially valid, it is
- * able to raise an error as soon as a mistake is made.
+ * g_match_info_fetch_named_pos:
+ * @match_info: #GMatchInfo structure
+ * @name: name of the subexpression
+ * @start_pos: (out) (allow-none): pointer to location where to store
+ * the start position, or %NULL
+ * @end_pos: (out) (allow-none): pointer to location where to store
+ * the end position, or %NULL
*
- * GRegex supports the concept of partial matching by means of the
- * #G_REGEX_MATCH_PARTIAL flag. When this is set the return code for
- * g_regex_match() or g_regex_match_full() is, as usual, %TRUE
- * for a complete match, %FALSE otherwise. But, when this functions
- * returns %FALSE, you can check if the match was partial calling
- * g_regex_is_partial_match().
+ * Retrieves the position in bytes of the capturing parentheses named @name.
*
- * When using partial matching you cannot use g_regex_fetch*().
+ * If @name is a valid sub pattern name but it didn't match anything
+ * (e.g. sub pattern "X", matching "b" against "(?P<X>a)?b")
+ * then @start_pos and @end_pos are set to -1 and %TRUE is returned.
*
- * Because of the way certain internal optimizations are implemented the
- * partial matching algorithm cannot be used with all patterns. So repeated
- * single characters such as "a{2,4}" and repeated single metasequences such
- * as "\d+" are not permitted if the maximum number of occurrences is
- * greater than one. Optional items such as "\d?" (where the maximum is one)
- * are permitted. Quantifiers with any values are permitted after
- * parentheses, so the invalid examples above can be coded thus "(a){2,4}"
- * and "(\d)+". If #G_REGEX_MATCH_PARTIAL is set for a pattern that does
- * not conform to the restrictions, matching functions return an error.
- *
- * Returns: %TRUE if the match was partial, %FALSE otherwise.
+ * Returns: %TRUE if the position was fetched, %FALSE otherwise.
+ * If the position cannot be fetched, @start_pos and @end_pos
+ * are left unchanged.
*
* Since: 2.14
*/
gboolean
-g_regex_is_partial_match (const GRegex *regex)
+g_match_info_fetch_named_pos (const GMatchInfo *match_info,
+ const gchar *name,
+ gint *start_pos,
+ gint *end_pos)
{
- g_return_val_if_fail (regex != NULL, FALSE);
+ gint num;
- if (regex->match == NULL)
+ g_return_val_if_fail (match_info != NULL, FALSE);
+ g_return_val_if_fail (name != NULL, FALSE);
+
+ num = get_matched_substring_number (match_info, name);
+ if (num < 0)
return FALSE;
- return regex->match->matches == PCRE_ERROR_PARTIAL;
+ return g_match_info_fetch_pos (match_info, num, start_pos, end_pos);
}
/**
- * g_regex_fetch:
- * @regex: #GRegex structure used in last match.
- * @match_num: number of the sub expression.
- * @string: the string on which the last match was made.
+ * g_match_info_fetch_all:
+ * @match_info: a #GMatchInfo structure
*
- * Retrieves the text matching the @match_num<!-- -->'th capturing parentheses.
- * 0 is the full text of the match, 1 is the first paren set, 2 the second,
- * and so on.
+ * Bundles up pointers to each of the matching substrings from a match
+ * and stores them in an array of gchar pointers. The first element in
+ * the returned array is the match number 0, i.e. the entire matched
+ * text.
*
- * If @match_num is a valid sub pattern but it didn't match anything (e.g.
- * sub pattern 1, matching "b" against "(a)?b") then an empty string is
- * returned.
+ * If a sub pattern didn't match anything (e.g. sub pattern 1, matching
+ * "b" against "(a)?b") then an empty string is inserted.
*
* If the last match was obtained using the DFA algorithm, that is using
* g_regex_match_all() or g_regex_match_all_full(), the retrieved
- * string is not that of a set of parentheses but that of a matched
- * substring. Substrings are matched in reverse order of length, so 0 is
- * the longest match.
+ * strings are not that matched by sets of parentheses but that of the
+ * matched substring. Substrings are matched in reverse order of length,
+ * so the first one is the longest match.
+ *
+ * The strings are fetched from the string passed to the match function,
+ * so you cannot call this function after freeing the string.
*
- * Returns: The matched substring, or %NULL if an error occurred.
- * You have to free the string yourself.
+ * Returns: (allow-none): a %NULL-terminated array of gchar * pointers.
+ * It must be freed using g_strfreev(). If the previous match failed
+ * %NULL is returned
*
* Since: 2.14
*/
-gchar *
-g_regex_fetch (const GRegex *regex,
- gint match_num,
- const gchar *string)
+gchar **
+g_match_info_fetch_all (const GMatchInfo *match_info)
{
- /* we cannot use pcre_get_substring() because it allocates the
- * string using pcre_malloc(). */
- gchar *match = NULL;
- gint start, end;
+ /* we cannot use pcre_get_substring_list() because the returned value
+ * isn't suitable for g_strfreev(). */
+ gchar **result;
+ gint i;
- g_return_val_if_fail (regex != NULL, NULL);
- g_return_val_if_fail (match_num >= 0, NULL);
+ g_return_val_if_fail (match_info != NULL, NULL);
- if (regex->match == NULL)
+ if (match_info->matches < 0)
return NULL;
- if (regex->match->string_len < 0)
- return NULL;
+ result = g_new (gchar *, match_info->matches + 1);
+ for (i = 0; i < match_info->matches; i++)
+ result[i] = g_match_info_fetch (match_info, i);
+ result[i] = NULL;
- /* match_num does not exist or it didn't matched, i.e. matching "b"
- * against "(a)?b" then group 0 is empty. */
- if (!g_regex_fetch_pos (regex, match_num, &start, &end))
- match = NULL;
- else if (start == -1)
- match = g_strdup ("");
- else
- match = g_strndup (&string[start], end - start);
+ return result;
+}
- return match;
+
+/* GRegex */
+
+GQuark
+g_regex_error_quark (void)
+{
+ static GQuark error_quark = 0;
+
+ if (error_quark == 0)
+ error_quark = g_quark_from_static_string ("g-regex-error-quark");
+
+ return error_quark;
}
/**
- * g_regex_fetch_pos:
- * @regex: #GRegex structure used in last match.
- * @match_num: number of the sub expression.
- * @start_pos: pointer to location where to store the start position.
- * @end_pos: pointer to location where to store the end position.
+ * g_regex_ref:
+ * @regex: a #GRegex
*
- * Retrieves the position of the @match_num<!-- -->'th capturing parentheses.
- * 0 is the full text of the match, 1 is the first paren set, 2 the second,
- * and so on.
+ * Increases reference count of @regex by 1.
*
- * If @match_num is a valid sub pattern but it didn't match anything (e.g.
- * sub pattern 1, matching "b" against "(a)?b") then @start_pos and @end_pos
- * are set to -1 and %TRUE is returned.
+ * Returns: @regex
*
- * If the last match was obtained using the DFA algorithm, that is using
- * g_regex_match_all() or g_regex_match_all_full(), the retrieved
- * position is not that of a set of parentheses but that of a matched
- * substring. Substrings are matched in reverse order of length, so 0 is
- * the longest match.
+ * Since: 2.14
+ */
+GRegex *
+g_regex_ref (GRegex *regex)
+{
+ g_return_val_if_fail (regex != NULL, NULL);
+ g_atomic_int_inc (®ex->ref_count);
+ return regex;
+}
+
+/**
+ * g_regex_unref:
+ * @regex: a #GRegex
*
- * Returns: %TRUE if the position was fetched, %FALSE otherwise. If the
- * position cannot be fetched, @start_pos and @end_pos are left
- * unchanged.
+ * Decreases reference count of @regex by 1. When reference count drops
+ * to zero, it frees all the memory associated with the regex structure.
*
* Since: 2.14
*/
-gboolean
-g_regex_fetch_pos (const GRegex *regex,
- gint match_num,
- gint *start_pos,
- gint *end_pos)
+void
+g_regex_unref (GRegex *regex)
{
- g_return_val_if_fail (regex != NULL, FALSE);
- g_return_val_if_fail (match_num >= 0, FALSE);
-
- if (regex->match == NULL)
- return FALSE;
+ g_return_if_fail (regex != NULL);
- /* make sure the sub expression number they're requesting is less than
- * the total number of sub expressions that were matched. */
- if (match_num >= regex->match->matches)
- return FALSE;
+ if (g_atomic_int_dec_and_test (®ex->ref_count))
+ {
+ g_free (regex->pattern);
+ if (regex->pcre_re != NULL)
+ pcre_free (regex->pcre_re);
+ if (regex->extra != NULL)
+ pcre_free (regex->extra);
+ g_free (regex);
+ }
+}
- if (start_pos != NULL)
+/**
+ * g_regex_new:
+ * @pattern: the regular expression
+ * @compile_options: compile options for the regular expression, or 0
+ * @match_options: match options for the regular expression, or 0
+ * @error: return location for a #GError
+ *
+ * Compiles the regular expression to an internal form, and does
+ * the initial setup of the #GRegex structure.
+ *
+ * Returns: a #GRegex structure. Call g_regex_unref() when you
+ * are done with it
+ *
+ * Since: 2.14
+ */
+GRegex *
+g_regex_new (const gchar *pattern,
+ GRegexCompileFlags compile_options,
+ GRegexMatchFlags match_options,
+ GError **error)
+{
+ GRegex *regex;
+ pcre *re;
+ const gchar *errmsg;
+ gint erroffset;
+ gint errcode;
+ gboolean optimize = FALSE;
+ static gsize initialised;
+ unsigned long int pcre_compile_options;
+
+ g_return_val_if_fail (pattern != NULL, NULL);
+ g_return_val_if_fail (error == NULL || *error == NULL, NULL);
+ g_return_val_if_fail ((compile_options & ~G_REGEX_COMPILE_MASK) == 0, NULL);
+ g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, NULL);
+
+ if (g_once_init_enter (&initialised))
{
- *start_pos = regex->match->offsets[2 * match_num];
+ gint support;
+ const gchar *msg;
+
+ pcre_config (PCRE_CONFIG_UTF8, &support);
+ if (!support)
+ {
+ msg = N_("PCRE library is compiled without UTF8 support");
+ g_critical ("%s", msg);
+ g_set_error_literal (error, G_REGEX_ERROR, G_REGEX_ERROR_COMPILE, gettext (msg));
+ return NULL;
+ }
+
+ pcre_config (PCRE_CONFIG_UNICODE_PROPERTIES, &support);
+ if (!support)
+ {
+ msg = N_("PCRE library is compiled without UTF8 properties support");
+ g_critical ("%s", msg);
+ g_set_error_literal (error, G_REGEX_ERROR, G_REGEX_ERROR_COMPILE, gettext (msg));
+ return NULL;
+ }
+
+ g_once_init_leave (&initialised, TRUE);
}
- if (end_pos != NULL)
+ /* G_REGEX_OPTIMIZE has the same numeric value of PCRE_NO_UTF8_CHECK,
+ * as we do not need to wrap PCRE_NO_UTF8_CHECK. */
+ if (compile_options & G_REGEX_OPTIMIZE)
+ optimize = TRUE;
+
+ /* In GRegex the string are, by default, UTF-8 encoded. PCRE
+ * instead uses UTF-8 only if required with PCRE_UTF8. */
+ if (compile_options & G_REGEX_RAW)
+ {
+ /* disable utf-8 */
+ compile_options &= ~G_REGEX_RAW;
+ }
+ else
{
- *end_pos = regex->match->offsets[2 * match_num + 1];
+ /* enable utf-8 */
+ compile_options |= PCRE_UTF8 | PCRE_NO_UTF8_CHECK;
+ match_options |= PCRE_NO_UTF8_CHECK;
}
- return TRUE;
+ /* PCRE_NEWLINE_ANY is the default for the internal PCRE but
+ * not for the system one. */
+ if (!(compile_options & G_REGEX_NEWLINE_CR) &&
+ !(compile_options & G_REGEX_NEWLINE_LF))
+ {
+ compile_options |= PCRE_NEWLINE_ANY;
+ }
+
+ compile_options |= PCRE_UCP;
+
+ /* compile the pattern */
+ re = pcre_compile2 (pattern, compile_options, &errcode,
+ &errmsg, &erroffset, NULL);
+
+ /* if the compilation failed, set the error member and return
+ * immediately */
+ if (re == NULL)
+ {
+ GError *tmp_error;
+
+ /* Translate the PCRE error code to GRegexError and use a translated
+ * error message if possible */
+ translate_compile_error (&errcode, &errmsg);
+
+ /* PCRE uses byte offsets but we want to show character offsets */
+ erroffset = g_utf8_pointer_to_offset (pattern, &pattern[erroffset]);
+
+ tmp_error = g_error_new (G_REGEX_ERROR, errcode,
+ _("Error while compiling regular "
+ "expression %s at char %d: %s"),
+ pattern, erroffset, errmsg);
+ g_propagate_error (error, tmp_error);
+
+ return NULL;
+ }
+
+ /* For options set at the beginning of the pattern, pcre puts them into
+ * compile options, e.g. "(?i)foo" will make the pcre structure store
+ * PCRE_CASELESS even though it wasn't explicitly given for compilation. */
+ pcre_fullinfo (re, NULL, PCRE_INFO_OPTIONS, &pcre_compile_options);
+ compile_options = pcre_compile_options;
+
+ if (!(compile_options & G_REGEX_DUPNAMES))
+ {
+ gboolean jchanged = FALSE;
+ pcre_fullinfo (re, NULL, PCRE_INFO_JCHANGED, &jchanged);
+ if (jchanged)
+ compile_options |= G_REGEX_DUPNAMES;
+ }
+
+ regex = g_new0 (GRegex, 1);
+ regex->ref_count = 1;
+ regex->pattern = g_strdup (pattern);
+ regex->pcre_re = re;
+ regex->compile_opts = compile_options;
+ regex->match_opts = match_options;
+
+ if (optimize)
+ {
+ regex->extra = pcre_study (regex->pcre_re, 0, &errmsg);
+ if (errmsg != NULL)
+ {
+ GError *tmp_error = g_error_new (G_REGEX_ERROR,
+ G_REGEX_ERROR_OPTIMIZE,
+ _("Error while optimizing "
+ "regular expression %s: %s"),
+ regex->pattern,
+ errmsg);
+ g_propagate_error (error, tmp_error);
+
+ g_regex_unref (regex);
+ return NULL;
+ }
+ }
+
+ return regex;
}
/**
- * g_regex_fetch_named:
- * @regex: #GRegex structure used in last match.
- * @name: name of the subexpression.
- * @string: the string on which the last match was made.
+ * g_regex_get_pattern:
+ * @regex: a #GRegex structure
*
- * Retrieves the text matching the capturing parentheses named @name.
+ * Gets the pattern string associated with @regex, i.e. a copy of
+ * the string passed to g_regex_new().
*
- * If @name is a valid sub pattern name but it didn't match anything (e.g.
- * sub pattern "X", matching "b" against "(?P<X>a)?b") then an empty
- * string is returned.
+ * Returns: the pattern of @regex
+ *
+ * Since: 2.14
+ */
+const gchar *
+g_regex_get_pattern (const GRegex *regex)
+{
+ g_return_val_if_fail (regex != NULL, NULL);
+
+ return regex->pattern;
+}
+
+/**
+ * g_regex_get_max_backref:
+ * @regex: a #GRegex
+ *
+ * Returns the number of the highest back reference
+ * in the pattern, or 0 if the pattern does not contain
+ * back references.
*
- * Returns: The matched substring, or %NULL if an error occurred.
- * You have to free the string yourself.
+ * Returns: the number of the highest back reference
*
* Since: 2.14
*/
-gchar *
-g_regex_fetch_named (const GRegex *regex,
- const gchar *name,
- const gchar *string)
+gint
+g_regex_get_max_backref (const GRegex *regex)
{
- /* we cannot use pcre_get_named_substring() because it allocates the
- * string using pcre_malloc(). */
- gint num;
+ gint value;
- g_return_val_if_fail (regex != NULL, NULL);
- g_return_val_if_fail (string != NULL, NULL);
- g_return_val_if_fail (name != NULL, NULL);
+ pcre_fullinfo (regex->pcre_re, regex->extra,
+ PCRE_INFO_BACKREFMAX, &value);
- num = g_regex_get_string_number (regex, name);
- if (num == -1)
- return NULL;
- else
- return g_regex_fetch (regex, num, string);
+ return value;
}
/**
- * g_regex_fetch_named_pos:
- * @regex: #GRegex structure used in last match.
- * @name: name of the subexpression.
- * @start_pos: pointer to location where to store the start position.
- * @end_pos: pointer to location where to store the end position.
+ * g_regex_get_capture_count:
+ * @regex: a #GRegex
+ *
+ * Returns the number of capturing subpatterns in the pattern.
*
- * Retrieves the position of the capturing parentheses named @name.
+ * Returns: the number of capturing subpatterns
+ *
+ * Since: 2.14
+ */
+gint
+g_regex_get_capture_count (const GRegex *regex)
+{
+ gint value;
+
+ pcre_fullinfo (regex->pcre_re, regex->extra,
+ PCRE_INFO_CAPTURECOUNT, &value);
+
+ return value;
+}
+
+/**
+ * g_regex_get_compile_flags:
+ * @regex: a #GRegex
*
- * If @name is a valid sub pattern name but it didn't match anything (e.g.
- * sub pattern "X", matching "b" against "(?P<X>a)?b") then @start_pos and
- * @end_pos are set to -1 and %TRUE is returned.
+ * Returns the compile options that @regex was created with.
*
- * Returns: %TRUE if the position was fetched, %FALSE otherwise. If the
- * position cannot be fetched, @start_pos and @end_pos are left
- * unchanged.
+ * Returns: flags from #GRegexCompileFlags
+ *
+ * Since: 2.26
+ */
+GRegexCompileFlags
+g_regex_get_compile_flags (const GRegex *regex)
+{
+ g_return_val_if_fail (regex != NULL, 0);
+
+ return regex->compile_opts;
+}
+
+/**
+ * g_regex_get_match_flags:
+ * @regex: a #GRegex
+ *
+ * Returns the match options that @regex was created with.
+ *
+ * Returns: flags from #GRegexMatchFlags
+ *
+ * Since: 2.26
+ */
+GRegexMatchFlags
+g_regex_get_match_flags (const GRegex *regex)
+{
+ g_return_val_if_fail (regex != NULL, 0);
+
+ return regex->match_opts;
+}
+
+/**
+ * g_regex_match_simple:
+ * @pattern: the regular expression
+ * @string: the string to scan for matches
+ * @compile_options: compile options for the regular expression, or 0
+ * @match_options: match options, or 0
+ *
+ * Scans for a match in @string for @pattern.
+ *
+ * This function is equivalent to g_regex_match() but it does not
+ * require to compile the pattern with g_regex_new(), avoiding some
+ * lines of code when you need just to do a match without extracting
+ * substrings, capture counts, and so on.
+ *
+ * If this function is to be called on the same @pattern more than
+ * once, it's more efficient to compile the pattern once with
+ * g_regex_new() and then use g_regex_match().
+ *
+ * Returns: %TRUE if the string matched, %FALSE otherwise
*
* Since: 2.14
*/
gboolean
-g_regex_fetch_named_pos (const GRegex *regex,
- const gchar *name,
- gint *start_pos,
- gint *end_pos)
+g_regex_match_simple (const gchar *pattern,
+ const gchar *string,
+ GRegexCompileFlags compile_options,
+ GRegexMatchFlags match_options)
{
- gint num;
+ GRegex *regex;
+ gboolean result;
- num = g_regex_get_string_number (regex, name);
- if (num == -1)
+ regex = g_regex_new (pattern, compile_options, 0, NULL);
+ if (!regex)
return FALSE;
+ result = g_regex_match_full (regex, string, -1, 0, match_options, NULL, NULL);
+ g_regex_unref (regex);
+ return result;
+}
- return g_regex_fetch_pos (regex, num, start_pos, end_pos);
+/**
+ * g_regex_match:
+ * @regex: a #GRegex structure from g_regex_new()
+ * @string: the string to scan for matches
+ * @match_options: match options
+ * @match_info: (out) (allow-none): pointer to location where to store
+ * the #GMatchInfo, or %NULL if you do not need it
+ *
+ * Scans for a match in string for the pattern in @regex.
+ * The @match_options are combined with the match options specified
+ * when the @regex structure was created, letting you have more
+ * flexibility in reusing #GRegex structures.
+ *
+ * A #GMatchInfo structure, used to get information on the match,
+ * is stored in @match_info if not %NULL. Note that if @match_info
+ * is not %NULL then it is created even if the function returns %FALSE,
+ * i.e. you must free it regardless if regular expression actually matched.
+ *
+ * To retrieve all the non-overlapping matches of the pattern in
+ * string you can use g_match_info_next().
+ *
+ * |[
+ * static void
+ * print_uppercase_words (const gchar *string)
+ * {
+ * /* Print all uppercase-only words. */
+ * GRegex *regex;
+ * GMatchInfo *match_info;
+ *
+ * regex = g_regex_new ("[A-Z]+", 0, 0, NULL);
+ * g_regex_match (regex, string, 0, &match_info);
+ * while (g_match_info_matches (match_info))
+ * {
+ * gchar *word = g_match_info_fetch (match_info, 0);
+ * g_print ("Found: %s\n", word);
+ * g_free (word);
+ * g_match_info_next (match_info, NULL);
+ * }
+ * g_match_info_free (match_info);
+ * g_regex_unref (regex);
+ * }
+ * ]|
+ *
+ * @string is not copied and is used in #GMatchInfo internally. If
+ * you use any #GMatchInfo method (except g_match_info_free()) after
+ * freeing or modifying @string then the behaviour is undefined.
+ *
+ * Returns: %TRUE is the string matched, %FALSE otherwise
+ *
+ * Since: 2.14
+ */
+gboolean
+g_regex_match (const GRegex *regex,
+ const gchar *string,
+ GRegexMatchFlags match_options,
+ GMatchInfo **match_info)
+{
+ return g_regex_match_full (regex, string, -1, 0, match_options,
+ match_info, NULL);
}
/**
- * g_regex_fetch_all:
- * @regex: a #GRegex structure.
- * @string: the string on which the last match was made.
+ * g_regex_match_full:
+ * @regex: a #GRegex structure from g_regex_new()
+ * @string: (array length=string_len): the string to scan for matches
+ * @string_len: the length of @string, or -1 if @string is nul-terminated
+ * @start_position: starting index of the string to match
+ * @match_options: match options
+ * @match_info: (out) (allow-none): pointer to location where to store
+ * the #GMatchInfo, or %NULL if you do not need it
+ * @error: location to store the error occurring, or %NULL to ignore errors
+ *
+ * Scans for a match in string for the pattern in @regex.
+ * The @match_options are combined with the match options specified
+ * when the @regex structure was created, letting you have more
+ * flexibility in reusing #GRegex structures.
+ *
+ * Setting @start_position differs from just passing over a shortened
+ * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
+ * that begins with any kind of lookbehind assertion, such as "\b".
+ *
+ * A #GMatchInfo structure, used to get information on the match, is
+ * stored in @match_info if not %NULL. Note that if @match_info is
+ * not %NULL then it is created even if the function returns %FALSE,
+ * i.e. you must free it regardless if regular expression actually
+ * matched.
+ *
+ * @string is not copied and is used in #GMatchInfo internally. If
+ * you use any #GMatchInfo method (except g_match_info_free()) after
+ * freeing or modifying @string then the behaviour is undefined.
+ *
+ * To retrieve all the non-overlapping matches of the pattern in
+ * string you can use g_match_info_next().
+ *
+ * |[
+ * static void
+ * print_uppercase_words (const gchar *string)
+ * {
+ * /* Print all uppercase-only words. */
+ * GRegex *regex;
+ * GMatchInfo *match_info;
+ * GError *error = NULL;
+ *
+ * regex = g_regex_new ("[A-Z]+", 0, 0, NULL);
+ * g_regex_match_full (regex, string, -1, 0, 0, &match_info, &error);
+ * while (g_match_info_matches (match_info))
+ * {
+ * gchar *word = g_match_info_fetch (match_info, 0);
+ * g_print ("Found: %s\n", word);
+ * g_free (word);
+ * g_match_info_next (match_info, &error);
+ * }
+ * g_match_info_free (match_info);
+ * g_regex_unref (regex);
+ * if (error != NULL)
+ * {
+ * g_printerr ("Error while matching: %s\n", error->message);
+ * g_error_free (error);
+ * }
+ * }
+ * ]|
+ *
+ * Returns: %TRUE is the string matched, %FALSE otherwise
*
- * Bundles up pointers to each of the matching substrings from a match
- * and stores them in an array of gchar pointers. The first element in
- * the returned array is the match number 0, i.e. the entire matched
- * text.
+ * Since: 2.14
+ */
+gboolean
+g_regex_match_full (const GRegex *regex,
+ const gchar *string,
+ gssize string_len,
+ gint start_position,
+ GRegexMatchFlags match_options,
+ GMatchInfo **match_info,
+ GError **error)
+{
+ GMatchInfo *info;
+ gboolean match_ok;
+
+ g_return_val_if_fail (regex != NULL, FALSE);
+ g_return_val_if_fail (string != NULL, FALSE);
+ g_return_val_if_fail (start_position >= 0, FALSE);
+ g_return_val_if_fail (error == NULL || *error == NULL, FALSE);
+ g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, FALSE);
+
+ info = match_info_new (regex, string, string_len, start_position,
+ match_options, FALSE);
+ match_ok = g_match_info_next (info, error);
+ if (match_info != NULL)
+ *match_info = info;
+ else
+ g_match_info_free (info);
+
+ return match_ok;
+}
+
+/**
+ * g_regex_match_all:
+ * @regex: a #GRegex structure from g_regex_new()
+ * @string: the string to scan for matches
+ * @match_options: match options
+ * @match_info: (out) (allow-none): pointer to location where to store
+ * the #GMatchInfo, or %NULL if you do not need it
+ *
+ * Using the standard algorithm for regular expression matching only
+ * the longest match in the string is retrieved. This function uses
+ * a different algorithm so it can retrieve all the possible matches.
+ * For more documentation see g_regex_match_all_full().
*
- * If a sub pattern didn't match anything (e.g. sub pattern 1, matching
- * "b" against "(a)?b") then an empty string is inserted.
+ * A #GMatchInfo structure, used to get information on the match, is
+ * stored in @match_info if not %NULL. Note that if @match_info is
+ * not %NULL then it is created even if the function returns %FALSE,
+ * i.e. you must free it regardless if regular expression actually
+ * matched.
*
- * If the last match was obtained using the DFA algorithm, that is using
- * g_regex_match_all() or g_regex_match_all_full(), the retrieved
- * strings are not that matched by sets of parentheses but that of the
- * matched substring. Substrings are matched in reverse order of length,
- * so the first one is the longest match.
+ * @string is not copied and is used in #GMatchInfo internally. If
+ * you use any #GMatchInfo method (except g_match_info_free()) after
+ * freeing or modifying @string then the behaviour is undefined.
+ *
+ * Returns: %TRUE is the string matched, %FALSE otherwise
+ *
+ * Since: 2.14
+ */
+gboolean
+g_regex_match_all (const GRegex *regex,
+ const gchar *string,
+ GRegexMatchFlags match_options,
+ GMatchInfo **match_info)
+{
+ return g_regex_match_all_full (regex, string, -1, 0, match_options,
+ match_info, NULL);
+}
+
+/**
+ * g_regex_match_all_full:
+ * @regex: a #GRegex structure from g_regex_new()
+ * @string: (array length=string_len): the string to scan for matches
+ * @string_len: the length of @string, or -1 if @string is nul-terminated
+ * @start_position: starting index of the string to match
+ * @match_options: match options
+ * @match_info: (out) (allow-none): pointer to location where to store
+ * the #GMatchInfo, or %NULL if you do not need it
+ * @error: location to store the error occurring, or %NULL to ignore errors
+ *
+ * Using the standard algorithm for regular expression matching only
+ * the longest match in the string is retrieved, it is not possible
+ * to obtain all the available matches. For instance matching
+ * "<a> <b> <c>" against the pattern "<.*>"
+ * you get "<a> <b> <c>".
+ *
+ * This function uses a different algorithm (called DFA, i.e. deterministic
+ * finite automaton), so it can retrieve all the possible matches, all
+ * starting at the same point in the string. For instance matching
+ * "<a> <b> <c>" against the pattern "<.*>"
+ * you would obtain three matches: "<a> <b> <c>",
+ * "<a> <b>" and "<a>".
+ *
+ * The number of matched strings is retrieved using
+ * g_match_info_get_match_count(). To obtain the matched strings and
+ * their position you can use, respectively, g_match_info_fetch() and
+ * g_match_info_fetch_pos(). Note that the strings are returned in
+ * reverse order of length; that is, the longest matching string is
+ * given first.
*
- * Returns: a %NULL-terminated array of gchar * pointers. It must be freed
- * using g_strfreev(). If the memory can't be allocated, returns
- * %NULL.
+ * Note that the DFA algorithm is slower than the standard one and it
+ * is not able to capture substrings, so backreferences do not work.
+ *
+ * Setting @start_position differs from just passing over a shortened
+ * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
+ * that begins with any kind of lookbehind assertion, such as "\b".
+ *
+ * A #GMatchInfo structure, used to get information on the match, is
+ * stored in @match_info if not %NULL. Note that if @match_info is
+ * not %NULL then it is created even if the function returns %FALSE,
+ * i.e. you must free it regardless if regular expression actually
+ * matched.
+ *
+ * @string is not copied and is used in #GMatchInfo internally. If
+ * you use any #GMatchInfo method (except g_match_info_free()) after
+ * freeing or modifying @string then the behaviour is undefined.
+ *
+ * Returns: %TRUE is the string matched, %FALSE otherwise
*
* Since: 2.14
*/
-gchar **
-g_regex_fetch_all (const GRegex *regex,
- const gchar *string)
+gboolean
+g_regex_match_all_full (const GRegex *regex,
+ const gchar *string,
+ gssize string_len,
+ gint start_position,
+ GRegexMatchFlags match_options,
+ GMatchInfo **match_info,
+ GError **error)
{
- /* we cannot use pcre_get_substring_list() because the returned value
- * isn't suitable for g_strfreev(). */
- gchar **result;
- gint i;
+ GMatchInfo *info;
+ gboolean done;
g_return_val_if_fail (regex != NULL, FALSE);
g_return_val_if_fail (string != NULL, FALSE);
+ g_return_val_if_fail (start_position >= 0, FALSE);
+ g_return_val_if_fail (error == NULL || *error == NULL, FALSE);
+ g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, FALSE);
- if (regex->match == NULL)
- return NULL;
+ info = match_info_new (regex, string, string_len, start_position,
+ match_options, TRUE);
- if (regex->match->matches < 0)
- return NULL;
+ done = FALSE;
+ while (!done)
+ {
+ done = TRUE;
+ info->matches = pcre_dfa_exec (regex->pcre_re, regex->extra,
+ info->string, info->string_len,
+ info->pos,
+ regex->match_opts | match_options,
+ info->offsets, info->n_offsets,
+ info->workspace, info->n_workspace);
+ if (info->matches == PCRE_ERROR_DFA_WSSIZE)
+ {
+ /* info->workspace is too small. */
+ info->n_workspace *= 2;
+ info->workspace = g_realloc (info->workspace,
+ info->n_workspace * sizeof (gint));
+ done = FALSE;
+ }
+ else if (info->matches == 0)
+ {
+ /* info->offsets is too small. */
+ info->n_offsets *= 2;
+ info->offsets = g_realloc (info->offsets,
+ info->n_offsets * sizeof (gint));
+ done = FALSE;
+ }
+ else if (IS_PCRE_ERROR (info->matches))
+ {
+ g_set_error (error, G_REGEX_ERROR, G_REGEX_ERROR_MATCH,
+ _("Error while matching regular expression %s: %s"),
+ regex->pattern, match_error (info->matches));
+ }
+ }
- result = g_new (gchar *, regex->match->matches + 1);
- for (i = 0; i < regex->match->matches; i++)
- result[i] = g_regex_fetch (regex, i, string);
- result[i] = NULL;
+ /* set info->pos to -1 so that a call to g_match_info_next() fails. */
+ info->pos = -1;
- return result;
+ if (match_info != NULL)
+ *match_info = info;
+ else
+ g_match_info_free (info);
+
+ return info->matches >= 0;
}
/**
* g_regex_get_string_number:
- * @regex: #GRegex structure.
- * @name: name of the subexpression.
+ * @regex: #GRegex structure
+ * @name: name of the subexpression
*
* Retrieves the number of the subexpression named @name.
*
- * Returns: The number of the subexpression or -1 if @name does not exists.
+ * Returns: The number of the subexpression or -1 if @name
+ * does not exists
*
* Since: 2.14
*/
gint
g_regex_get_string_number (const GRegex *regex,
- const gchar *name)
+ const gchar *name)
{
gint num;
g_return_val_if_fail (regex != NULL, -1);
g_return_val_if_fail (name != NULL, -1);
- num = pcre_get_stringnumber (regex->pattern->pcre_re, name);
+ num = pcre_get_stringnumber (regex->pcre_re, name);
if (num == PCRE_ERROR_NOSUBSTRING)
- num = -1;
+ num = -1;
return num;
}
/**
* g_regex_split_simple:
- * @pattern: the regular expression.
- * @string: the string to scan for matches.
- * @compile_options: compile options for the regular expression.
- * @match_options: match options.
+ * @pattern: the regular expression
+ * @string: the string to scan for matches
+ * @compile_options: compile options for the regular expression, or 0
+ * @match_options: match options, or 0
*
- * Breaks the string on the pattern, and returns an array of the tokens.
- * If the pattern contains capturing parentheses, then the text for each
- * of the substrings will also be returned. If the pattern does not match
- * anywhere in the string, then the whole string is returned as the first
- * token.
+ * Breaks the string on the pattern, and returns an array of
+ * the tokens. If the pattern contains capturing parentheses,
+ * then the text for each of the substrings will also be returned.
+ * If the pattern does not match anywhere in the string, then the
+ * whole string is returned as the first token.
*
- * This function is equivalent to g_regex_split() but it does not
- * require to compile the pattern with g_regex_new(), avoiding some
- * lines of code when you need just to do a split without extracting
- * substrings, capture counts, and so on.
+ * This function is equivalent to g_regex_split() but it does
+ * not require to compile the pattern with g_regex_new(), avoiding
+ * some lines of code when you need just to do a split without
+ * extracting substrings, capture counts, and so on.
*
* If this function is to be called on the same @pattern more than
* once, it's more efficient to compile the pattern once with
* g_regex_new() and then use g_regex_split().
*
- * As a special case, the result of splitting the empty string "" is an
- * empty vector, not a vector containing a single string. The reason for
- * this special case is that being able to represent a empty vector is
- * typically more useful than consistent handling of empty elements. If
- * you do need to represent empty elements, you'll need to check for the
- * empty string before calling this function.
+ * As a special case, the result of splitting the empty string ""
+ * is an empty vector, not a vector containing a single string.
+ * The reason for this special case is that being able to represent
+ * a empty vector is typically more useful than consistent handling
+ * of empty elements. If you do need to represent empty elements,
+ * you'll need to check for the empty string before calling this
+ * function.
*
- * A pattern that can match empty strings splits @string into separate
- * characters wherever it matches the empty string between characters.
- * For example splitting "ab c" using as a separator "\s*", you will get
- * "a", "b" and "c".
+ * A pattern that can match empty strings splits @string into
+ * separate characters wherever it matches the empty string between
+ * characters. For example splitting "ab c" using as a separator
+ * "\s*", you will get "a", "b" and "c".
*
- * Returns: a %NULL-terminated gchar ** array. Free it using g_strfreev().
+ * Returns: a %NULL-terminated array of strings. Free it using g_strfreev()
*
* Since: 2.14
**/
gchar **
g_regex_split_simple (const gchar *pattern,
- const gchar *string,
- GRegexCompileFlags compile_options,
- GRegexMatchFlags match_options)
+ const gchar *string,
+ GRegexCompileFlags compile_options,
+ GRegexMatchFlags match_options)
{
GRegex *regex;
gchar **result;
regex = g_regex_new (pattern, compile_options, 0, NULL);
if (!regex)
return NULL;
+
result = g_regex_split_full (regex, string, -1, 0, match_options, 0, NULL);
- g_regex_free (regex);
+ g_regex_unref (regex);
return result;
}
/**
* g_regex_split:
- * @regex: a #GRegex structure.
- * @string: the string to split with the pattern.
- * @match_options: match time option flags.
+ * @regex: a #GRegex structure
+ * @string: the string to split with the pattern
+ * @match_options: match time option flags
*
* Breaks the string on the pattern, and returns an array of the tokens.
* If the pattern contains capturing parentheses, then the text for each
* For example splitting "ab c" using as a separator "\s*", you will get
* "a", "b" and "c".
*
- * Returns: a %NULL-terminated gchar ** array. Free it using g_strfreev().
+ * Returns: a %NULL-terminated gchar ** array. Free it using g_strfreev()
*
* Since: 2.14
**/
gchar **
-g_regex_split (GRegex *regex,
- const gchar *string,
- GRegexMatchFlags match_options)
+g_regex_split (const GRegex *regex,
+ const gchar *string,
+ GRegexMatchFlags match_options)
{
return g_regex_split_full (regex, string, -1, 0,
match_options, 0, NULL);
/**
* g_regex_split_full:
- * @regex: a #GRegex structure.
- * @string: the string to split with the pattern.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @match_options: match time option flags.
- * @max_tokens: the maximum number of tokens to split @string into. If this
- * is less than 1, the string is split completely.
- * @error: return location for a #GError.
+ * @regex: a #GRegex structure
+ * @string: (array length=string_len): the string to split with the pattern
+ * @string_len: the length of @string, or -1 if @string is nul-terminated
+ * @start_position: starting index of the string to match
+ * @match_options: match time option flags
+ * @max_tokens: the maximum number of tokens to split @string into.
+ * If this is less than 1, the string is split completely
+ * @error: return location for a #GError
*
* Breaks the string on the pattern, and returns an array of the tokens.
* If the pattern contains capturing parentheses, then the text for each
* For example splitting "ab c" using as a separator "\s*", you will get
* "a", "b" and "c".
*
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
+ * Setting @start_position differs from just passing over a shortened
+ * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
+ * that begins with any kind of lookbehind assertion, such as "\b".
*
- * Returns: a %NULL-terminated gchar ** array. Free it using g_strfreev().
+ * Returns: a %NULL-terminated gchar ** array. Free it using g_strfreev()
*
* Since: 2.14
**/
gchar **
-g_regex_split_full (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- GRegexMatchFlags match_options,
- gint max_tokens,
- GError **error)
+g_regex_split_full (const GRegex *regex,
+ const gchar *string,
+ gssize string_len,
+ gint start_position,
+ GRegexMatchFlags match_options,
+ gint max_tokens,
+ GError **error)
{
- gchar **string_list; /* The array of char **s worked on */
- gint pos;
- gint tokens;
- GList *list, *last;
GError *tmp_error = NULL;
+ GMatchInfo *match_info;
+ GList *list, *last;
+ gint i;
+ gint token_count;
+ gboolean match_ok;
+ /* position of the last separator. */
+ gint last_separator_end;
+ /* was the last match 0 bytes long? */
+ gboolean last_match_is_empty;
+ /* the returned array of char **s */
+ gchar **string_list;
g_return_val_if_fail (regex != NULL, NULL);
g_return_val_if_fail (string != NULL, NULL);
g_return_val_if_fail (error == NULL || *error == NULL, NULL);
g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, NULL);
- regex_lazy_init_match (regex, 0);
-
if (max_tokens <= 0)
max_tokens = G_MAXINT;
if (string_len < 0)
- string_len = strlen(string);
+ string_len = strlen (string);
+ /* zero-length string */
if (string_len - start_position == 0)
return g_new0 (gchar *, 1);
- /* clear out the regex for reuse, just in case */
- g_regex_clear (regex);
-
- list = NULL;
- tokens = 0;
- while (TRUE)
+ if (max_tokens == 1)
{
- gchar *token;
-
- /* -1 to leave room for the last part. */
- if (tokens >= max_tokens - 1)
- {
- /* we have reached the maximum number of tokens, so we copy
- * the remaining part of the string. */
- if (regex->match->last_match_is_empty)
- {
- /* the last match was empty, so we have moved one char
- * after the real position to avoid empty matches at the
- * same position. */
- regex->match->pos = PREV_CHAR (regex, &string[regex->match->pos]) - string;
- }
- /* the if is needed in the case we have terminated the available
- * tokens, but we are at the end of the string, so there are no
- * characters left to copy. */
- if (string_len > regex->match->pos)
- {
- token = g_strndup (string + regex->match->pos,
- string_len - regex->match->pos);
- list = g_list_prepend (list, token);
- }
- /* end the loop. */
- break;
- }
-
- token = g_regex_split_next_full (regex, string, string_len, start_position,
- match_options, &tmp_error);
- if (tmp_error != NULL)
- {
- g_propagate_error (error, tmp_error);
- g_list_foreach (list, (GFunc)g_free, NULL);
- g_list_free (list);
- regex->match->pos = -1;
- return NULL;
- }
-
- if (token == NULL)
- /* no more tokens. */
- break;
-
- tokens++;
- list = g_list_prepend (list, token);
+ string_list = g_new0 (gchar *, 2);
+ string_list[0] = g_strndup (&string[start_position],
+ string_len - start_position);
+ return string_list;
}
- string_list = g_new (gchar *, g_list_length (list) + 1);
- pos = 0;
- for (last = g_list_last (list); last; last = g_list_previous (last))
- string_list[pos++] = last->data;
- string_list[pos] = 0;
-
- regex->match->pos = -1;
- g_list_free (list);
-
- return string_list;
-}
-
-/**
- * g_regex_split_next:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: the string to split on pattern.
- * @match_options: match time options for the regex.
- *
- * g_regex_split_next() breaks the string on pattern, and returns the
- * tokens, one per call. If the pattern contains capturing parentheses,
- * then the text for each of the substrings will also be returned.
- * If the pattern does not match anywhere in the string, then the whole
- * string is returned as the first token.
- *
- * A pattern that can match empty strings splits @string into separate
- * characters wherever it matches the empty string between characters.
- * For example splitting "ab c" using as a separator "\s*", you will get
- * "a", "b" and "c".
- *
- * You have to call g_regex_clear() to reuse the same pattern on a new
- * string.
- *
- * Returns: a gchar * to the next token of the string.
- *
- * Since: 2.14
- */
-gchar *
-g_regex_split_next (GRegex *regex,
- const gchar *string,
- GRegexMatchFlags match_options)
-{
- return g_regex_split_next_full (regex, string, -1, 0, match_options,
- NULL);
-}
-
-/**
- * g_regex_split_next_full:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: the string to split on pattern.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @match_options: match time options for the regex.
- * @error: return location for a #GError.
- *
- * g_regex_split_next_full() breaks the string on pattern, and returns
- * the tokens, one per call. If the pattern contains capturing parentheses,
- * then the text for each of the substrings will also be returned.
- * If the pattern does not match anywhere in the string, then the whole
- * string is returned as the first token.
- *
- * A pattern that can match empty strings splits @string into separate
- * characters wherever it matches the empty string between characters.
- * For example splitting "ab c" using as a separator "\s*", you will get
- * "a", "b" and "c".
- *
- * You have to call g_regex_clear() to reuse the same pattern on a new
- * string.
- *
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
- *
- * Returns: a gchar * to the next token of the string.
- *
- * Since: 2.14
- */
-gchar *
-g_regex_split_next_full (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- GRegexMatchFlags match_options,
- GError **error)
-{
- gint new_pos;
- gchar *token = NULL;
- gboolean match_ok;
- gint match_count;
- GError *tmp_error = NULL;
-
- g_return_val_if_fail (regex != NULL, NULL);
- g_return_val_if_fail (string != NULL, NULL);
- g_return_val_if_fail (error == NULL || *error == NULL, NULL);
- g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, NULL);
-
- regex_lazy_init_match (regex, 0);
+ list = NULL;
+ token_count = 0;
+ last_separator_end = start_position;
+ last_match_is_empty = FALSE;
- new_pos = MAX (regex->match->pos, start_position);
- if (regex->match->last_match_is_empty)
- /* if the last match was empty, g_regex_match_next_full() has moved
- * forward to avoid infinite loops, but we still need to copy that
- * character. */
- new_pos = PREV_CHAR(regex, &string[new_pos]) - string;
+ match_ok = g_regex_match_full (regex, string, string_len, start_position,
+ match_options, &match_info, &tmp_error);
- /* if there are delimiter substrings stored, return those one at a
- * time.
- */
- if (regex->match->delims != NULL)
+ while (tmp_error == NULL)
{
- token = regex->match->delims->data;
- regex->match->delims = g_slist_remove (regex->match->delims, token);
- return token;
- }
+ if (match_ok)
+ {
+ last_match_is_empty =
+ (match_info->offsets[0] == match_info->offsets[1]);
+
+ /* we need to skip empty separators at the same position of the end
+ * of another separator. e.g. the string is "a b" and the separator
+ * is " *", so from 1 to 2 we have a match and at position 2 we have
+ * an empty match. */
+ if (last_separator_end != match_info->offsets[1])
+ {
+ gchar *token;
+ gint match_count;
+
+ token = g_strndup (string + last_separator_end,
+ match_info->offsets[0] - last_separator_end);
+ list = g_list_prepend (list, token);
+ token_count++;
+
+ /* if there were substrings, these need to be added to
+ * the list. */
+ match_count = g_match_info_get_match_count (match_info);
+ if (match_count > 1)
+ {
+ for (i = 1; i < match_count; i++)
+ list = g_list_prepend (list, g_match_info_fetch (match_info, i));
+ }
+ }
+ }
+ else
+ {
+ /* if there was no match, copy to end of string. */
+ if (!last_match_is_empty)
+ {
+ gchar *token = g_strndup (string + last_separator_end,
+ match_info->string_len - last_separator_end);
+ list = g_list_prepend (list, token);
+ }
+ /* no more tokens, end the loop. */
+ break;
+ }
- if (regex->match->pos == -1)
- /* the last call to g_regex_match_next_full() returned NULL. */
- return NULL;
+ /* -1 to leave room for the last part. */
+ if (token_count >= max_tokens - 1)
+ {
+ /* we have reached the maximum number of tokens, so we copy
+ * the remaining part of the string. */
+ if (last_match_is_empty)
+ {
+ /* the last match was empty, so we have moved one char
+ * after the real position to avoid empty matches at the
+ * same position. */
+ match_info->pos = PREV_CHAR (regex, &string[match_info->pos]) - string;
+ }
+ /* the if is needed in the case we have terminated the available
+ * tokens, but we are at the end of the string, so there are no
+ * characters left to copy. */
+ if (string_len > match_info->pos)
+ {
+ gchar *token = g_strndup (string + match_info->pos,
+ string_len - match_info->pos);
+ list = g_list_prepend (list, token);
+ }
+ /* end the loop. */
+ break;
+ }
- if (regex->match->string_len < 0)
- {
- regex->match->last_match_is_empty = FALSE;
- /* initialize last_separator_end to start_position to skip the
- * empty token at the beginning of the string. */
- regex->match->last_separator_end = start_position;
- }
+ last_separator_end = match_info->pos;
+ if (last_match_is_empty)
+ /* if the last match was empty, g_match_info_next() has moved
+ * forward to avoid infinite loops, but we still need to copy that
+ * character. */
+ last_separator_end = PREV_CHAR (regex, &string[last_separator_end]) - string;
- /* use g_regex_match_next() to find the next occurance of the pattern
- * in the string. We use new_pos to keep track of where the stuff
- * up to the current match starts. Copy that token of the string off
- * and append it to the buffer using g_strndup. */
- match_ok = g_regex_match_next_full (regex, string, string_len,
- start_position, match_options,
- &tmp_error);
+ match_ok = g_match_info_next (match_info, &tmp_error);
+ }
+ g_match_info_free (match_info);
if (tmp_error != NULL)
{
g_propagate_error (error, tmp_error);
+ g_list_foreach (list, (GFunc)g_free, NULL);
+ g_list_free (list);
+ match_info->pos = -1;
return NULL;
}
- if (match_ok)
- {
- regex->match->last_match_is_empty =
- (regex->match->offsets[0] == regex->match->offsets[1]);
-
- /* we need to skip empty separators at the same position of the end
- * of another separator. e.g. the string is "a b" and the separator
- * is "*", so from 1 to 2 we have a match and at position 2 we have
- * an empty match. */
- if (regex->match->last_separator_end != regex->match->offsets[1])
- {
- token = g_strndup (string + new_pos, regex->match->offsets[0] - new_pos);
-
- /* if there were substrings, these need to get added to the
- * list of delims */
- match_count = g_regex_get_match_count (regex);
- if (match_count > 1)
- {
- gint i;
- for (i = 1; i < match_count; i++)
- regex->match->delims = g_slist_append (regex->match->delims,
- g_regex_fetch (regex, i, string));
- }
-
- regex->match->last_separator_end = regex->match->offsets[1];
- }
- else
- {
- /* we have skipped an empty separator so we need to find the
- * next match. */
- return g_regex_split_next_full (regex, string, string_len,
- start_position, match_options,
- error);
- }
- }
- else
- {
- /* if there was no match, copy to end of string. */
- if (!regex->match->last_match_is_empty)
- token = g_strndup (string + new_pos, regex->match->string_len - new_pos);
- else
- token = NULL;
- }
+ string_list = g_new (gchar *, g_list_length (list) + 1);
+ i = 0;
+ for (last = g_list_last (list); last; last = g_list_previous (last))
+ string_list[i++] = last->data;
+ string_list[i] = NULL;
+ g_list_free (list);
- return token;
+ return string_list;
}
enum
REPL_TYPE_SYMBOLIC_REFERENCE,
REPL_TYPE_NUMERIC_REFERENCE,
REPL_TYPE_CHANGE_CASE
-};
+};
typedef enum
{
CHANGE_CASE_UPPER_MASK = CHANGE_CASE_UPPER | CHANGE_CASE_UPPER_SINGLE
} ChangeCase;
-typedef struct
+struct _InterpolationData
{
- gchar *text;
- gint type;
+ gchar *text;
+ gint type;
gint num;
gchar c;
ChangeCase change_case;
-} InterpolationData;
+};
static void
free_interpolation_data (InterpolationData *data)
static const gchar *
expand_escape (const gchar *replacement,
- const gchar *p,
- InterpolationData *data,
- GError **error)
+ const gchar *p,
+ InterpolationData *data,
+ GError **error)
{
const gchar *q, *r;
gint x, d, h, i;
p++;
x = 0;
if (*p == '{')
- {
- p++;
- do
- {
- h = g_ascii_xdigit_value (*p);
- if (h < 0)
- {
- error_detail = _("hexadecimal digit or '}' expected");
- goto error;
- }
- x = x * 16 + h;
- p++;
- }
- while (*p != '}');
- p++;
- }
+ {
+ p++;
+ do
+ {
+ h = g_ascii_xdigit_value (*p);
+ if (h < 0)
+ {
+ error_detail = _("hexadecimal digit or '}' expected");
+ goto error;
+ }
+ x = x * 16 + h;
+ p++;
+ }
+ while (*p != '}');
+ p++;
+ }
else
- {
- for (i = 0; i < 2; i++)
- {
- h = g_ascii_xdigit_value (*p);
- if (h < 0)
- {
- error_detail = _("hexadecimal digit expected");
- goto error;
- }
- x = x * 16 + h;
- p++;
- }
- }
+ {
+ for (i = 0; i < 2; i++)
+ {
+ h = g_ascii_xdigit_value (*p);
+ if (h < 0)
+ {
+ error_detail = _("hexadecimal digit expected");
+ goto error;
+ }
+ x = x * 16 + h;
+ p++;
+ }
+ }
data->type = REPL_TYPE_STRING;
data->text = g_new0 (gchar, 8);
g_unichar_to_utf8 (x, data->text);
case 'g':
p++;
if (*p != '<')
- {
- error_detail = _("missing '<' in symbolic reference");
- goto error;
- }
+ {
+ error_detail = _("missing '<' in symbolic reference");
+ goto error;
+ }
q = p + 1;
- do
- {
- p++;
- if (!*p)
- {
- error_detail = _("unfinished symbolic reference");
- goto error;
- }
- }
+ do
+ {
+ p++;
+ if (!*p)
+ {
+ error_detail = _("unfinished symbolic reference");
+ goto error;
+ }
+ }
while (*p != '>');
if (p - q == 0)
- {
- error_detail = _("zero-length symbolic reference");
- goto error;
- }
+ {
+ error_detail = _("zero-length symbolic reference");
+ goto error;
+ }
if (g_ascii_isdigit (*q))
- {
- x = 0;
- do
- {
- h = g_ascii_digit_value (*q);
- if (h < 0)
- {
- error_detail = _("digit expected");
- p = q;
- goto error;
- }
- x = x * 10 + h;
- q++;
- }
- while (q != p);
- data->num = x;
- data->type = REPL_TYPE_NUMERIC_REFERENCE;
- }
+ {
+ x = 0;
+ do
+ {
+ h = g_ascii_digit_value (*q);
+ if (h < 0)
+ {
+ error_detail = _("digit expected");
+ p = q;
+ goto error;
+ }
+ x = x * 10 + h;
+ q++;
+ }
+ while (q != p);
+ data->num = x;
+ data->type = REPL_TYPE_NUMERIC_REFERENCE;
+ }
else
- {
- r = q;
- do
- {
- if (!g_ascii_isalnum (*r))
- {
- error_detail = _("illegal symbolic reference");
- p = r;
- goto error;
- }
- r++;
- }
- while (r != p);
- data->text = g_strndup (q, p - q);
- data->type = REPL_TYPE_SYMBOLIC_REFERENCE;
- }
+ {
+ r = q;
+ do
+ {
+ if (!g_ascii_isalnum (*r))
+ {
+ error_detail = _("illegal symbolic reference");
+ p = r;
+ goto error;
+ }
+ r++;
+ }
+ while (r != p);
+ data->text = g_strndup (q, p - q);
+ data->type = REPL_TYPE_SYMBOLIC_REFERENCE;
+ }
p++;
break;
case '0':
x = 0;
d = 0;
for (i = 0; i < 3; i++)
- {
- h = g_ascii_digit_value (*p);
- if (h < 0)
- break;
- if (h > 7)
- {
- if (base == 8)
- break;
- else
- base = 10;
- }
- if (i == 2 && base == 10)
- break;
- x = x * 8 + h;
- d = d * 10 + h;
- p++;
- }
+ {
+ h = g_ascii_digit_value (*p);
+ if (h < 0)
+ break;
+ if (h > 7)
+ {
+ if (base == 8)
+ break;
+ else
+ base = 10;
+ }
+ if (i == 2 && base == 10)
+ break;
+ x = x * 8 + h;
+ d = d * 10 + h;
+ p++;
+ }
if (base == 8 || i == 3)
- {
- data->type = REPL_TYPE_STRING;
- data->text = g_new0 (gchar, 8);
- g_unichar_to_utf8 (x, data->text);
- }
+ {
+ data->type = REPL_TYPE_STRING;
+ data->text = g_new0 (gchar, 8);
+ g_unichar_to_utf8 (x, data->text);
+ }
else
- {
- data->type = REPL_TYPE_NUMERIC_REFERENCE;
- data->num = d;
- }
+ {
+ data->type = REPL_TYPE_NUMERIC_REFERENCE;
+ data->num = d;
+ }
break;
case 0:
error_detail = _("stray final '\\'");
error:
/* G_GSSIZE_FORMAT doesn't work with gettext, so we use %lu */
- tmp_error = g_error_new (G_REGEX_ERROR,
- G_REGEX_ERROR_REPLACE,
- _("Error while parsing replacement "
- "text \"%s\" at char %lu: %s"),
- replacement,
- (gulong)(p - replacement),
- error_detail);
+ tmp_error = g_error_new (G_REGEX_ERROR,
+ G_REGEX_ERROR_REPLACE,
+ _("Error while parsing replacement "
+ "text \"%s\" at char %lu: %s"),
+ replacement,
+ (gulong)(p - replacement),
+ error_detail);
g_propagate_error (error, tmp_error);
return NULL;
static GList *
split_replacement (const gchar *replacement,
- GError **error)
+ GError **error)
{
GList *list = NULL;
InterpolationData *data;
const gchar *p, *start;
-
- start = p = replacement;
+
+ start = p = replacement;
while (*p)
{
if (*p == '\\')
- {
- data = g_new0 (InterpolationData, 1);
- start = p = expand_escape (replacement, p, data, error);
- if (p == NULL)
- {
- g_list_foreach (list, (GFunc)free_interpolation_data, NULL);
- g_list_free (list);
- free_interpolation_data (data);
-
- return NULL;
- }
- list = g_list_prepend (list, data);
- }
+ {
+ data = g_new0 (InterpolationData, 1);
+ start = p = expand_escape (replacement, p, data, error);
+ if (p == NULL)
+ {
+ g_list_foreach (list, (GFunc)free_interpolation_data, NULL);
+ g_list_free (list);
+ free_interpolation_data (data);
+
+ return NULL;
+ }
+ list = g_list_prepend (list, data);
+ }
else
- {
- p++;
- if (*p == '\\' || *p == '\0')
- {
- if (p - start > 0)
- {
- data = g_new0 (InterpolationData, 1);
- data->text = g_strndup (start, p - start);
- data->type = REPL_TYPE_STRING;
- list = g_list_prepend (list, data);
- }
- }
- }
+ {
+ p++;
+ if (*p == '\\' || *p == '\0')
+ {
+ if (p - start > 0)
+ {
+ data = g_new0 (InterpolationData, 1);
+ data->text = g_strndup (start, p - start);
+ data->type = REPL_TYPE_STRING;
+ list = g_list_prepend (list, data);
+ }
+ }
+ }
}
return g_list_reverse (list);
/* Change the case of c based on change_case. */
#define CHANGE_CASE(c, change_case) \
- (((change_case) & CHANGE_CASE_LOWER_MASK) ? \
- g_unichar_tolower (c) : \
- g_unichar_toupper (c))
+ (((change_case) & CHANGE_CASE_LOWER_MASK) ? \
+ g_unichar_tolower (c) : \
+ g_unichar_toupper (c))
static void
string_append (GString *string,
- const gchar *text,
+ const gchar *text,
ChangeCase *change_case)
{
gunichar c;
}
static gboolean
-interpolate_replacement (const GRegex *regex,
- const gchar *string,
- GString *result,
- gpointer data)
+interpolate_replacement (const GMatchInfo *match_info,
+ GString *result,
+ gpointer data)
{
GList *list;
InterpolationData *idata;
{
idata = list->data;
switch (idata->type)
- {
- case REPL_TYPE_STRING:
- string_append (result, idata->text, &change_case);
- break;
- case REPL_TYPE_CHARACTER:
- g_string_append_c (result, CHANGE_CASE (idata->c, change_case));
+ {
+ case REPL_TYPE_STRING:
+ string_append (result, idata->text, &change_case);
+ break;
+ case REPL_TYPE_CHARACTER:
+ g_string_append_c (result, CHANGE_CASE (idata->c, change_case));
if (change_case & CHANGE_CASE_SINGLE_MASK)
change_case = CHANGE_CASE_NONE;
- break;
- case REPL_TYPE_NUMERIC_REFERENCE:
- match = g_regex_fetch (regex, idata->num, string);
- if (match)
- {
- string_append (result, match, &change_case);
- g_free (match);
- }
- break;
- case REPL_TYPE_SYMBOLIC_REFERENCE:
- match = g_regex_fetch_named (regex, idata->text, string);
- if (match)
- {
- string_append (result, match, &change_case);
- g_free (match);
- }
- break;
- case REPL_TYPE_CHANGE_CASE:
- change_case = idata->change_case;
- break;
- }
+ break;
+ case REPL_TYPE_NUMERIC_REFERENCE:
+ match = g_match_info_fetch (match_info, idata->num);
+ if (match)
+ {
+ string_append (result, match, &change_case);
+ g_free (match);
+ }
+ break;
+ case REPL_TYPE_SYMBOLIC_REFERENCE:
+ match = g_match_info_fetch_named (match_info, idata->text);
+ if (match)
+ {
+ string_append (result, match, &change_case);
+ g_free (match);
+ }
+ break;
+ case REPL_TYPE_CHANGE_CASE:
+ change_case = idata->change_case;
+ break;
+ }
}
- return FALSE;
+ return FALSE;
}
-/**
- * g_regex_expand_references:
- * @regex: #GRegex structure used in last match.
- * @string: the string on which the last match was made.
- * @string_to_expand: the string to expand.
- * @error: location to store the error occuring, or NULL to ignore errors.
- *
- * Returns a new string containing the text in @string_to_expand with
- * references expanded. References refer to the last match done with
- * @string against @regex and have the same syntax used by g_regex_replace().
- *
- * The @string_to_expand must be UTF-8 encoded even if #G_REGEX_RAW was
- * passed to g_regex_new().
- *
- * Returns: the expanded string, or %NULL if an error occurred.
- *
- * Since: 2.14
+/* whether actual match_info is needed for replacement, i.e.
+ * whether there are references
*/
-gchar *
-g_regex_expand_references (GRegex *regex,
- const gchar *string,
- const gchar *string_to_expand,
- GError **error)
+static gboolean
+interpolation_list_needs_match (GList *list)
{
- GString *result;
- GList *list;
- GError *tmp_error = NULL;
-
- g_return_val_if_fail (regex != NULL, NULL);
- g_return_val_if_fail (string != NULL, NULL);
- g_return_val_if_fail (string_to_expand != NULL, NULL);
- g_return_val_if_fail (error == NULL || *error == NULL, NULL);
-
- list = split_replacement (string_to_expand, &tmp_error);
- if (tmp_error != NULL)
+ while (list != NULL)
{
- g_propagate_error (error, tmp_error);
- return NULL;
- }
+ InterpolationData *data = list->data;
- result = g_string_sized_new (strlen (string_to_expand));
- interpolate_replacement (regex, string, result, list);
+ if (data->type == REPL_TYPE_SYMBOLIC_REFERENCE ||
+ data->type == REPL_TYPE_NUMERIC_REFERENCE)
+ {
+ return TRUE;
+ }
- g_list_foreach (list, (GFunc)free_interpolation_data, NULL);
- g_list_free (list);
+ list = list->next;
+ }
- return g_string_free (result, FALSE);
+ return FALSE;
}
/**
* g_regex_replace:
- * @regex: a #GRegex structure.
- * @string: the string to perform matches against.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @replacement: text to replace each match with.
- * @match_options: options for the match.
- * @error: location to store the error occuring, or NULL to ignore errors.
- *
- * Replaces all occurances of the pattern in @regex with the
- * replacement text. Backreferences of the form '\number' or '\g<number>'
- * in the replacement text are interpolated by the number-th captured
- * subexpression of the match, '\g<name>' refers to the captured subexpression
- * with the given name. '\0' refers to the complete match, but '\0' followed
- * by a number is the octal representation of a character. To include a
- * literal '\' in the replacement, write '\\'.
+ * @regex: a #GRegex structure
+ * @string: (array length=string_len): the string to perform matches against
+ * @string_len: the length of @string, or -1 if @string is nul-terminated
+ * @start_position: starting index of the string to match
+ * @replacement: text to replace each match with
+ * @match_options: options for the match
+ * @error: location to store the error occurring, or %NULL to ignore errors
+ *
+ * Replaces all occurrences of the pattern in @regex with the
+ * replacement text. Backreferences of the form '\number' or
+ * '\g<number>' in the replacement text are interpolated by the
+ * number-th captured subexpression of the match, '\g<name>' refers
+ * to the captured subexpression with the given name. '\0' refers to the
+ * complete match, but '\0' followed by a number is the octal representation
+ * of a character. To include a literal '\' in the replacement, write '\\'.
* There are also escapes that changes the case of the following text:
*
* <variablelist>
* passed to g_regex_new(). If you want to use not UTF-8 encoded stings
* you can use g_regex_replace_literal().
*
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
+ * Setting @start_position differs from just passing over a shortened
+ * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that
+ * begins with any kind of lookbehind assertion, such as "\b".
*
- * Returns: a newly allocated string containing the replacements.
+ * Returns: a newly allocated string containing the replacements
*
* Since: 2.14
*/
gchar *
-g_regex_replace (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- const gchar *replacement,
- GRegexMatchFlags match_options,
- GError **error)
+g_regex_replace (const GRegex *regex,
+ const gchar *string,
+ gssize string_len,
+ gint start_position,
+ const gchar *replacement,
+ GRegexMatchFlags match_options,
+ GError **error)
{
gchar *result;
GList *list;
return NULL;
}
- result = g_regex_replace_eval (regex,
- string, string_len, start_position,
- match_options,
- interpolate_replacement,
- (gpointer)list,
+ result = g_regex_replace_eval (regex,
+ string, string_len, start_position,
+ match_options,
+ interpolate_replacement,
+ (gpointer)list,
&tmp_error);
if (tmp_error != NULL)
g_propagate_error (error, tmp_error);
}
static gboolean
-literal_replacement (const GRegex *regex,
- const gchar *string,
- GString *result,
- gpointer data)
+literal_replacement (const GMatchInfo *match_info,
+ GString *result,
+ gpointer data)
{
g_string_append (result, data);
return FALSE;
/**
* g_regex_replace_literal:
- * @regex: a #GRegex structure.
- * @string: the string to perform matches against.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @replacement: text to replace each match with.
- * @match_options: options for the match.
- * @error: location to store the error occuring, or NULL to ignore errors.
- *
- * Replaces all occurances of the pattern in @regex with the
+ * @regex: a #GRegex structure
+ * @string: (array length=string_len): the string to perform matches against
+ * @string_len: the length of @string, or -1 if @string is nul-terminated
+ * @start_position: starting index of the string to match
+ * @replacement: text to replace each match with
+ * @match_options: options for the match
+ * @error: location to store the error occurring, or %NULL to ignore errors
+ *
+ * Replaces all occurrences of the pattern in @regex with the
* replacement text. @replacement is replaced literally, to
* include backreferences use g_regex_replace().
*
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
+ * Setting @start_position differs from just passing over a
+ * shortened string and setting #G_REGEX_MATCH_NOTBOL in the
+ * case of a pattern that begins with any kind of lookbehind
+ * assertion, such as "\b".
*
- * Returns: a newly allocated string containing the replacements.
+ * Returns: a newly allocated string containing the replacements
*
* Since: 2.14
*/
gchar *
-g_regex_replace_literal (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- const gchar *replacement,
- GRegexMatchFlags match_options,
- GError **error)
+g_regex_replace_literal (const GRegex *regex,
+ const gchar *string,
+ gssize string_len,
+ gint start_position,
+ const gchar *replacement,
+ GRegexMatchFlags match_options,
+ GError **error)
{
g_return_val_if_fail (replacement != NULL, NULL);
g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, NULL);
return g_regex_replace_eval (regex,
- string, string_len, start_position,
- match_options,
- literal_replacement,
- (gpointer)replacement,
- error);
+ string, string_len, start_position,
+ match_options,
+ literal_replacement,
+ (gpointer)replacement,
+ error);
}
/**
* g_regex_replace_eval:
- * @regex: a #GRegex structure from g_regex_new().
- * @string: string to perform matches against.
- * @string_len: the length of @string, or -1 if @string is nul-terminated.
- * @start_position: starting index of the string to match.
- * @match_options: Options for the match.
- * @eval: a function to call for each match.
- * @user_data: user data to pass to the function.
- * @error: location to store the error occuring, or NULL to ignore errors.
- *
- * Replaces occurances of the pattern in regex with the output of @eval
- * for that occurance.
- *
- * Setting @start_position differs from just passing over a shortened string
- * and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that begins
- * with any kind of lookbehind assertion, such as "\b".
- *
- * Returns: a newly allocated string containing the replacements.
+ * @regex: a #GRegex structure from g_regex_new()
+ * @string: (array length=string_len): string to perform matches against
+ * @string_len: the length of @string, or -1 if @string is nul-terminated
+ * @start_position: starting index of the string to match
+ * @match_options: options for the match
+ * @eval: a function to call for each match
+ * @user_data: user data to pass to the function
+ * @error: location to store the error occurring, or %NULL to ignore errors
+ *
+ * Replaces occurrences of the pattern in regex with the output of
+ * @eval for that occurrence.
+ *
+ * Setting @start_position differs from just passing over a shortened
+ * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
+ * that begins with any kind of lookbehind assertion, such as "\b".
+ *
+ * The following example uses g_regex_replace_eval() to replace multiple
+ * strings at once:
+ * |[
+ * static gboolean
+ * eval_cb (const GMatchInfo *info,
+ * GString *res,
+ * gpointer data)
+ * {
+ * gchar *match;
+ * gchar *r;
+ *
+ * match = g_match_info_fetch (info, 0);
+ * r = g_hash_table_lookup ((GHashTable *)data, match);
+ * g_string_append (res, r);
+ * g_free (match);
+ *
+ * return FALSE;
+ * }
+ *
+ * /* ... */
+ *
+ * GRegex *reg;
+ * GHashTable *h;
+ * gchar *res;
+ *
+ * h = g_hash_table_new (g_str_hash, g_str_equal);
+ *
+ * g_hash_table_insert (h, "1", "ONE");
+ * g_hash_table_insert (h, "2", "TWO");
+ * g_hash_table_insert (h, "3", "THREE");
+ * g_hash_table_insert (h, "4", "FOUR");
+ *
+ * reg = g_regex_new ("1|2|3|4", 0, 0, NULL);
+ * res = g_regex_replace_eval (reg, text, -1, 0, 0, eval_cb, h, NULL);
+ * g_hash_table_destroy (h);
+ *
+ * /* ... */
+ * ]|
+ *
+ * Returns: a newly allocated string containing the replacements
*
* Since: 2.14
*/
gchar *
-g_regex_replace_eval (GRegex *regex,
- const gchar *string,
- gssize string_len,
- gint start_position,
- GRegexMatchFlags match_options,
- GRegexEvalCallback eval,
- gpointer user_data,
- GError **error)
+g_regex_replace_eval (const GRegex *regex,
+ const gchar *string,
+ gssize string_len,
+ gint start_position,
+ GRegexMatchFlags match_options,
+ GRegexEvalCallback eval,
+ gpointer user_data,
+ GError **error)
{
+ GMatchInfo *match_info;
GString *result;
gint str_pos = 0;
gboolean done = FALSE;
g_return_val_if_fail (eval != NULL, NULL);
g_return_val_if_fail ((match_options & ~G_REGEX_MATCH_MASK) == 0, NULL);
- regex_lazy_init_match (regex, 0);
-
if (string_len < 0)
- string_len = strlen(string);
-
- /* clear out the regex for reuse, just in case */
- g_regex_clear (regex);
+ string_len = strlen (string);
result = g_string_sized_new (string_len);
/* run down the string making matches. */
- while (!done &&
- g_regex_match_next_full (regex, string, string_len,
- start_position, match_options, &tmp_error))
+ g_regex_match_full (regex, string, string_len, start_position,
+ match_options, &match_info, &tmp_error);
+ while (!done && g_match_info_matches (match_info))
{
g_string_append_len (result,
- string + str_pos,
- regex->match->offsets[0] - str_pos);
- done = (*eval) (regex, string, result, user_data);
- str_pos = regex->match->offsets[1];
+ string + str_pos,
+ match_info->offsets[0] - str_pos);
+ done = (*eval) (match_info, result, user_data);
+ str_pos = match_info->offsets[1];
+ g_match_info_next (match_info, &tmp_error);
}
-
+ g_match_info_free (match_info);
if (tmp_error != NULL)
{
g_propagate_error (error, tmp_error);
}
g_string_append_len (result, string + str_pos, string_len - str_pos);
-
return g_string_free (result, FALSE);
}
/**
+ * g_regex_check_replacement:
+ * @replacement: the replacement string
+ * @has_references: (out) (allow-none): location to store information about
+ * references in @replacement or %NULL
+ * @error: location to store error
+ *
+ * Checks whether @replacement is a valid replacement string
+ * (see g_regex_replace()), i.e. that all escape sequences in
+ * it are valid.
+ *
+ * If @has_references is not %NULL then @replacement is checked
+ * for pattern references. For instance, replacement text 'foo\n'
+ * does not contain references and may be evaluated without information
+ * about actual match, but '\0\1' (whole match followed by first
+ * subpattern) requires valid #GMatchInfo object.
+ *
+ * Returns: whether @replacement is a valid replacement string
+ *
+ * Since: 2.14
+ */
+gboolean
+g_regex_check_replacement (const gchar *replacement,
+ gboolean *has_references,
+ GError **error)
+{
+ GList *list;
+ GError *tmp = NULL;
+
+ list = split_replacement (replacement, &tmp);
+
+ if (tmp)
+ {
+ g_propagate_error (error, tmp);
+ return FALSE;
+ }
+
+ if (has_references)
+ *has_references = interpolation_list_needs_match (list);
+
+ g_list_foreach (list, (GFunc) free_interpolation_data, NULL);
+ g_list_free (list);
+
+ return TRUE;
+}
+
+/**
+ * g_regex_escape_nul:
+ * @string: the string to escape
+ * @length: the length of @string
+ *
+ * Escapes the nul characters in @string to "\x00". It can be used
+ * to compile a regex with embedded nul characters.
+ *
+ * For completeness, @length can be -1 for a nul-terminated string.
+ * In this case the output string will be of course equal to @string.
+ *
+ * Returns: a newly-allocated escaped string
+ *
+ * Since: 2.30
+ */
+gchar *
+g_regex_escape_nul (const gchar *string,
+ gint length)
+{
+ GString *escaped;
+ const gchar *p, *piece_start, *end;
+ gint backslashes;
+
+ g_return_val_if_fail (string != NULL, NULL);
+
+ if (length < 0)
+ return g_strdup (string);
+
+ end = string + length;
+ p = piece_start = string;
+ escaped = g_string_sized_new (length + 1);
+
+ backslashes = 0;
+ while (p < end)
+ {
+ switch (*p)
+ {
+ case '\0':
+ if (p != piece_start)
+ {
+ /* copy the previous piece. */
+ g_string_append_len (escaped, piece_start, p - piece_start);
+ }
+ if ((backslashes & 1) == 0)
+ g_string_append_c (escaped, '\\');
+ g_string_append_c (escaped, 'x');
+ g_string_append_c (escaped, '0');
+ g_string_append_c (escaped, '0');
+ piece_start = ++p;
+ backslashes = 0;
+ break;
+ case '\\':
+ backslashes++;
+ ++p;
+ break;
+ default:
+ backslashes = 0;
+ p = g_utf8_next_char (p);
+ break;
+ }
+ }
+
+ if (piece_start < end)
+ g_string_append_len (escaped, piece_start, end - piece_start);
+
+ return g_string_free (escaped, FALSE);
+}
+
+/**
* g_regex_escape_string:
- * @string: the string to escape.
- * @length: the length of @string, or -1 if @string is nul-terminated.
+ * @string: (array length=length): the string to escape
+ * @length: the length of @string, or -1 if @string is nul-terminated
*
- * Escapes the special characters used for regular expressions in @string,
- * for instance "a.b*c" becomes "a\.b\*c". This function is useful to
- * dynamically generate regular expressions.
+ * Escapes the special characters used for regular expressions
+ * in @string, for instance "a.b*c" becomes "a\.b\*c". This
+ * function is useful to dynamically generate regular expressions.
*
- * @string can contain NULL characters that are replaced with "\0", in this
- * case remember to specify the correct length of @string in @length.
+ * @string can contain nul characters that are replaced with "\0",
+ * in this case remember to specify the correct length of @string
+ * in @length.
*
- * Returns: a newly allocated escaped string.
+ * Returns: a newly-allocated escaped string
*
* Since: 2.14
*/
gchar *
g_regex_escape_string (const gchar *string,
- gint length)
+ gint length)
{
GString *escaped;
const char *p, *piece_start, *end;
while (p < end)
{
switch (*p)
- {
- case '\0':
- case '\\':
- case '|':
- case '(':
- case ')':
- case '[':
- case ']':
- case '{':
- case '}':
- case '^':
- case '$':
- case '*':
- case '+':
- case '?':
- case '.':
- if (p != piece_start)
- /* copy the previous piece. */
- g_string_append_len (escaped, piece_start, p - piece_start);
- g_string_append_c (escaped, '\\');
- if (*p == '\0')
- g_string_append_c (escaped, '0');
- else
- g_string_append_c (escaped, *p);
- piece_start = ++p;
- break;
- default:
- p = g_utf8_next_char (p);
- }
+ {
+ case '\0':
+ case '\\':
+ case '|':
+ case '(':
+ case ')':
+ case '[':
+ case ']':
+ case '{':
+ case '}':
+ case '^':
+ case '$':
+ case '*':
+ case '+':
+ case '?':
+ case '.':
+ if (p != piece_start)
+ /* copy the previous piece. */
+ g_string_append_len (escaped, piece_start, p - piece_start);
+ g_string_append_c (escaped, '\\');
+ if (*p == '\0')
+ g_string_append_c (escaped, '0');
+ else
+ g_string_append_c (escaped, *p);
+ piece_start = ++p;
+ break;
+ default:
+ p = g_utf8_next_char (p);
+ break;
+ }
}
if (piece_start < end)
- g_string_append_len (escaped, piece_start, end - piece_start);
+ g_string_append_len (escaped, piece_start, end - piece_start);
return g_string_free (escaped, FALSE);
}
-
-#define __G_REGEX_C__
-#include "galiasdef.c"