1 /* GRegex -- regular expression API wrapper around PCRE.
3 * Copyright (C) 1999, 2000 Scott Wimer
4 * Copyright (C) 2004, Matthias Clasen <mclasen@redhat.com>
5 * Copyright (C) 2005 - 2007, Marco Barisione <marco@barisione.org>
7 * This library is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
12 * This library is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
17 * You should have received a copy of the GNU Lesser General Public License
18 * along with this library; if not, see <http://www.gnu.org/licenses/>.
25 #ifdef USE_SYSTEM_PCRE
28 #include "pcre/pcre.h"
35 #include "gmessages.h"
36 #include "gstrfuncs.h"
42 * @title: Perl-compatible regular expressions
43 * @short_description: matches strings against regular expressions
44 * @see_also: [Regular expression syntax][glib-regex-syntax]
46 * The g_regex_*() functions implement regular
47 * expression pattern matching using syntax and semantics similar to
48 * Perl regular expression.
50 * Some functions accept a @start_position argument, setting it differs
51 * from just passing over a shortened string and setting #G_REGEX_MATCH_NOTBOL
52 * in the case of a pattern that begins with any kind of lookbehind assertion.
53 * For example, consider the pattern "\Biss\B" which finds occurrences of "iss"
54 * in the middle of words. ("\B" matches only if the current position in the
55 * subject is not a word boundary.) When applied to the string "Mississipi"
56 * from the fourth byte, namely "issipi", it does not match, because "\B" is
57 * always false at the start of the subject, which is deemed to be a word
58 * boundary. However, if the entire string is passed , but with
59 * @start_position set to 4, it finds the second occurrence of "iss" because
60 * it is able to look behind the starting point to discover that it is
61 * preceded by a letter.
63 * Note that, unless you set the #G_REGEX_RAW flag, all the strings passed
64 * to these functions must be encoded in UTF-8. The lengths and the positions
65 * inside the strings are in bytes and not in characters, so, for instance,
66 * "\xc3\xa0" (i.e. "à") is two bytes long but it is treated as a
67 * single character. If you set #G_REGEX_RAW the strings can be non-valid
68 * UTF-8 strings and a byte is treated as a character, so "\xc3\xa0" is two
69 * bytes and two characters long.
71 * When matching a pattern, "\n" matches only against a "\n" character in
72 * the string, and "\r" matches only a "\r" character. To match any newline
73 * sequence use "\R". This particular group matches either the two-character
74 * sequence CR + LF ("\r\n"), or one of the single characters LF (linefeed,
75 * U+000A, "\n"), VT vertical tab, U+000B, "\v"), FF (formfeed, U+000C, "\f"),
76 * CR (carriage return, U+000D, "\r"), NEL (next line, U+0085), LS (line
77 * separator, U+2028), or PS (paragraph separator, U+2029).
79 * The behaviour of the dot, circumflex, and dollar metacharacters are
80 * affected by newline characters, the default is to recognize any newline
81 * character (the same characters recognized by "\R"). This can be changed
82 * with #G_REGEX_NEWLINE_CR, #G_REGEX_NEWLINE_LF and #G_REGEX_NEWLINE_CRLF
83 * compile options, and with #G_REGEX_MATCH_NEWLINE_ANY,
84 * #G_REGEX_MATCH_NEWLINE_CR, #G_REGEX_MATCH_NEWLINE_LF and
85 * #G_REGEX_MATCH_NEWLINE_CRLF match options. These settings are also
86 * relevant when compiling a pattern if #G_REGEX_EXTENDED is set, and an
87 * unescaped "#" outside a character class is encountered. This indicates
88 * a comment that lasts until after the next newline.
90 * When setting the %G_REGEX_JAVASCRIPT_COMPAT flag, pattern syntax and pattern
91 * matching is changed to be compatible with the way that regular expressions
92 * work in JavaScript. More precisely, a lonely ']' character in the pattern
93 * is a syntax error; the '\x' escape only allows 0 to 2 hexadecimal digits, and
94 * you must use the '\u' escape sequence with 4 hex digits to specify a unicode
95 * codepoint instead of '\x' or 'x{....}'. If '\x' or '\u' are not followed by
96 * the specified number of hex digits, they match 'x' and 'u' literally; also
97 * '\U' always matches 'U' instead of being an error in the pattern. Finally,
98 * pattern matching is modified so that back references to an unset subpattern
99 * group produces a match with the empty string instead of an error. See
100 * pcreapi(3) for more information.
102 * Creating and manipulating the same #GRegex structure from different
103 * threads is not a problem as #GRegex does not modify its internal
104 * state between creation and destruction, on the other hand #GMatchInfo
107 * The regular expressions low-level functionalities are obtained through
109 * [PCRE](http://www.pcre.org/)
110 * library written by Philip Hazel.
113 /* Mask of all the possible values for GRegexCompileFlags. */
114 #define G_REGEX_COMPILE_MASK (G_REGEX_CASELESS | \
115 G_REGEX_MULTILINE | \
119 G_REGEX_DOLLAR_ENDONLY | \
122 G_REGEX_NO_AUTO_CAPTURE | \
124 G_REGEX_FIRSTLINE | \
126 G_REGEX_NEWLINE_CR | \
127 G_REGEX_NEWLINE_LF | \
128 G_REGEX_NEWLINE_CRLF | \
129 G_REGEX_NEWLINE_ANYCRLF | \
130 G_REGEX_BSR_ANYCRLF | \
131 G_REGEX_JAVASCRIPT_COMPAT)
133 /* Mask of all GRegexCompileFlags values that are (not) passed trough to PCRE */
134 #define G_REGEX_COMPILE_PCRE_MASK (G_REGEX_COMPILE_MASK & ~G_REGEX_COMPILE_NONPCRE_MASK)
135 #define G_REGEX_COMPILE_NONPCRE_MASK (G_REGEX_RAW | \
138 /* Mask of all the possible values for GRegexMatchFlags. */
139 #define G_REGEX_MATCH_MASK (G_REGEX_MATCH_ANCHORED | \
140 G_REGEX_MATCH_NOTBOL | \
141 G_REGEX_MATCH_NOTEOL | \
142 G_REGEX_MATCH_NOTEMPTY | \
143 G_REGEX_MATCH_PARTIAL | \
144 G_REGEX_MATCH_NEWLINE_CR | \
145 G_REGEX_MATCH_NEWLINE_LF | \
146 G_REGEX_MATCH_NEWLINE_CRLF | \
147 G_REGEX_MATCH_NEWLINE_ANY | \
148 G_REGEX_MATCH_NEWLINE_ANYCRLF | \
149 G_REGEX_MATCH_BSR_ANYCRLF | \
150 G_REGEX_MATCH_BSR_ANY | \
151 G_REGEX_MATCH_PARTIAL_SOFT | \
152 G_REGEX_MATCH_PARTIAL_HARD | \
153 G_REGEX_MATCH_NOTEMPTY_ATSTART)
155 /* we rely on these flags having the same values */
156 G_STATIC_ASSERT (G_REGEX_CASELESS
== PCRE_CASELESS
);
157 G_STATIC_ASSERT (G_REGEX_MULTILINE
== PCRE_MULTILINE
);
158 G_STATIC_ASSERT (G_REGEX_DOTALL
== PCRE_DOTALL
);
159 G_STATIC_ASSERT (G_REGEX_EXTENDED
== PCRE_EXTENDED
);
160 G_STATIC_ASSERT (G_REGEX_ANCHORED
== PCRE_ANCHORED
);
161 G_STATIC_ASSERT (G_REGEX_DOLLAR_ENDONLY
== PCRE_DOLLAR_ENDONLY
);
162 G_STATIC_ASSERT (G_REGEX_UNGREEDY
== PCRE_UNGREEDY
);
163 G_STATIC_ASSERT (G_REGEX_NO_AUTO_CAPTURE
== PCRE_NO_AUTO_CAPTURE
);
164 G_STATIC_ASSERT (G_REGEX_FIRSTLINE
== PCRE_FIRSTLINE
);
165 G_STATIC_ASSERT (G_REGEX_DUPNAMES
== PCRE_DUPNAMES
);
166 G_STATIC_ASSERT (G_REGEX_NEWLINE_CR
== PCRE_NEWLINE_CR
);
167 G_STATIC_ASSERT (G_REGEX_NEWLINE_LF
== PCRE_NEWLINE_LF
);
168 G_STATIC_ASSERT (G_REGEX_NEWLINE_CRLF
== PCRE_NEWLINE_CRLF
);
169 G_STATIC_ASSERT (G_REGEX_NEWLINE_ANYCRLF
== PCRE_NEWLINE_ANYCRLF
);
170 G_STATIC_ASSERT (G_REGEX_BSR_ANYCRLF
== PCRE_BSR_ANYCRLF
);
171 G_STATIC_ASSERT (G_REGEX_JAVASCRIPT_COMPAT
== PCRE_JAVASCRIPT_COMPAT
);
173 G_STATIC_ASSERT (G_REGEX_MATCH_ANCHORED
== PCRE_ANCHORED
);
174 G_STATIC_ASSERT (G_REGEX_MATCH_NOTBOL
== PCRE_NOTBOL
);
175 G_STATIC_ASSERT (G_REGEX_MATCH_NOTEOL
== PCRE_NOTEOL
);
176 G_STATIC_ASSERT (G_REGEX_MATCH_NOTEMPTY
== PCRE_NOTEMPTY
);
177 G_STATIC_ASSERT (G_REGEX_MATCH_PARTIAL
== PCRE_PARTIAL
);
178 G_STATIC_ASSERT (G_REGEX_MATCH_NEWLINE_CR
== PCRE_NEWLINE_CR
);
179 G_STATIC_ASSERT (G_REGEX_MATCH_NEWLINE_LF
== PCRE_NEWLINE_LF
);
180 G_STATIC_ASSERT (G_REGEX_MATCH_NEWLINE_CRLF
== PCRE_NEWLINE_CRLF
);
181 G_STATIC_ASSERT (G_REGEX_MATCH_NEWLINE_ANY
== PCRE_NEWLINE_ANY
);
182 G_STATIC_ASSERT (G_REGEX_MATCH_NEWLINE_ANYCRLF
== PCRE_NEWLINE_ANYCRLF
);
183 G_STATIC_ASSERT (G_REGEX_MATCH_BSR_ANYCRLF
== PCRE_BSR_ANYCRLF
);
184 G_STATIC_ASSERT (G_REGEX_MATCH_BSR_ANY
== PCRE_BSR_UNICODE
);
185 G_STATIC_ASSERT (G_REGEX_MATCH_PARTIAL_SOFT
== PCRE_PARTIAL_SOFT
);
186 G_STATIC_ASSERT (G_REGEX_MATCH_PARTIAL_HARD
== PCRE_PARTIAL_HARD
);
187 G_STATIC_ASSERT (G_REGEX_MATCH_NOTEMPTY_ATSTART
== PCRE_NOTEMPTY_ATSTART
);
189 /* These PCRE flags are unused or not exposed publically in GRegexFlags, so
190 * it should be ok to reuse them for different things.
192 G_STATIC_ASSERT (G_REGEX_OPTIMIZE
== PCRE_NO_UTF8_CHECK
);
193 G_STATIC_ASSERT (G_REGEX_RAW
== PCRE_UTF8
);
195 /* if the string is in UTF-8 use g_utf8_ functions, else use
197 #define NEXT_CHAR(re, s) (((re)->compile_opts & G_REGEX_RAW) ? \
199 g_utf8_next_char (s))
200 #define PREV_CHAR(re, s) (((re)->compile_opts & G_REGEX_RAW) ? \
202 g_utf8_prev_char (s))
206 volatile gint ref_count
; /* the ref count */
207 GRegex
*regex
; /* the regex */
208 GRegexMatchFlags match_opts
; /* options used at match time on the regex */
209 gint matches
; /* number of matching sub patterns */
210 gint pos
; /* position in the string where last match left off */
211 gint n_offsets
; /* number of offsets */
212 gint
*offsets
; /* array of offsets paired 0,1 ; 2,3 ; 3,4 etc */
213 gint
*workspace
; /* workspace for pcre_dfa_exec() */
214 gint n_workspace
; /* number of workspace elements */
215 const gchar
*string
; /* string passed to the match function */
216 gssize string_len
; /* length of string, in bytes */
221 volatile gint ref_count
; /* the ref count for the immutable part */
222 gchar
*pattern
; /* the pattern */
223 pcre
*pcre_re
; /* compiled form of the pattern */
224 GRegexCompileFlags compile_opts
; /* options used at compile time on the pattern */
225 GRegexMatchFlags match_opts
; /* options used at match time on the regex */
226 pcre_extra
*extra
; /* data stored when G_REGEX_OPTIMIZE is used */
229 /* TRUE if ret is an error code, FALSE otherwise. */
230 #define IS_PCRE_ERROR(ret) ((ret) < PCRE_ERROR_NOMATCH && (ret) != PCRE_ERROR_PARTIAL)
232 typedef struct _InterpolationData InterpolationData
;
233 static gboolean
interpolation_list_needs_match (GList
*list
);
234 static gboolean
interpolate_replacement (const GMatchInfo
*match_info
,
237 static GList
*split_replacement (const gchar
*replacement
,
239 static void free_interpolation_data (InterpolationData
*data
);
243 match_error (gint errcode
)
247 case PCRE_ERROR_NOMATCH
:
250 case PCRE_ERROR_NULL
:
251 /* NULL argument, this should not happen in GRegex */
252 g_warning ("A NULL argument was passed to PCRE");
254 case PCRE_ERROR_BADOPTION
:
255 return "bad options";
256 case PCRE_ERROR_BADMAGIC
:
257 return _("corrupted object");
258 case PCRE_ERROR_UNKNOWN_OPCODE
:
259 return N_("internal error or corrupted object");
260 case PCRE_ERROR_NOMEMORY
:
261 return _("out of memory");
262 case PCRE_ERROR_NOSUBSTRING
:
263 /* not used by pcre_exec() */
265 case PCRE_ERROR_MATCHLIMIT
:
266 return _("backtracking limit reached");
267 case PCRE_ERROR_CALLOUT
:
268 /* callouts are not implemented */
270 case PCRE_ERROR_BADUTF8
:
271 case PCRE_ERROR_BADUTF8_OFFSET
:
272 /* we do not check if strings are valid */
274 case PCRE_ERROR_PARTIAL
:
277 case PCRE_ERROR_BADPARTIAL
:
278 return _("the pattern contains items not supported for partial matching");
279 case PCRE_ERROR_INTERNAL
:
280 return _("internal error");
281 case PCRE_ERROR_BADCOUNT
:
282 /* negative ovecsize, this should not happen in GRegex */
283 g_warning ("A negative ovecsize was passed to PCRE");
285 case PCRE_ERROR_DFA_UITEM
:
286 return _("the pattern contains items not supported for partial matching");
287 case PCRE_ERROR_DFA_UCOND
:
288 return _("back references as conditions are not supported for partial matching");
289 case PCRE_ERROR_DFA_UMLIMIT
:
290 /* the match_field field is not used in GRegex */
292 case PCRE_ERROR_DFA_WSSIZE
:
293 /* handled expanding the workspace */
295 case PCRE_ERROR_DFA_RECURSE
:
296 case PCRE_ERROR_RECURSIONLIMIT
:
297 return _("recursion limit reached");
298 case PCRE_ERROR_BADNEWLINE
:
299 return _("invalid combination of newline flags");
300 case PCRE_ERROR_BADOFFSET
:
301 return _("bad offset");
302 case PCRE_ERROR_SHORTUTF8
:
303 return _("short utf8");
304 case PCRE_ERROR_RECURSELOOP
:
305 return _("recursion loop");
309 return _("unknown error");
313 translate_compile_error (gint
*errcode
, const gchar
**errmsg
)
315 /* Compile errors are created adding 100 to the error code returned
317 * If errcode is known we put the translatable error message in
318 * erromsg. If errcode is unknown we put the generic
319 * G_REGEX_ERROR_COMPILE error code in errcode and keep the
320 * untranslated error message returned by PCRE.
321 * Note that there can be more PCRE errors with the same GRegexError
322 * and that some PCRE errors are useless for us.
328 case G_REGEX_ERROR_STRAY_BACKSLASH
:
329 *errmsg
= _("\\ at end of pattern");
331 case G_REGEX_ERROR_MISSING_CONTROL_CHAR
:
332 *errmsg
= _("\\c at end of pattern");
334 case G_REGEX_ERROR_UNRECOGNIZED_ESCAPE
:
335 *errmsg
= _("unrecognized character following \\");
337 case G_REGEX_ERROR_QUANTIFIERS_OUT_OF_ORDER
:
338 *errmsg
= _("numbers out of order in {} quantifier");
340 case G_REGEX_ERROR_QUANTIFIER_TOO_BIG
:
341 *errmsg
= _("number too big in {} quantifier");
343 case G_REGEX_ERROR_UNTERMINATED_CHARACTER_CLASS
:
344 *errmsg
= _("missing terminating ] for character class");
346 case G_REGEX_ERROR_INVALID_ESCAPE_IN_CHARACTER_CLASS
:
347 *errmsg
= _("invalid escape sequence in character class");
349 case G_REGEX_ERROR_RANGE_OUT_OF_ORDER
:
350 *errmsg
= _("range out of order in character class");
352 case G_REGEX_ERROR_NOTHING_TO_REPEAT
:
353 *errmsg
= _("nothing to repeat");
355 case 111: /* internal error: unexpected repeat */
356 *errcode
= G_REGEX_ERROR_INTERNAL
;
357 *errmsg
= _("unexpected repeat");
359 case G_REGEX_ERROR_UNRECOGNIZED_CHARACTER
:
360 *errmsg
= _("unrecognized character after (? or (?-");
362 case G_REGEX_ERROR_POSIX_NAMED_CLASS_OUTSIDE_CLASS
:
363 *errmsg
= _("POSIX named classes are supported only within a class");
365 case G_REGEX_ERROR_UNMATCHED_PARENTHESIS
:
366 *errmsg
= _("missing terminating )");
368 case G_REGEX_ERROR_INEXISTENT_SUBPATTERN_REFERENCE
:
369 *errmsg
= _("reference to non-existent subpattern");
371 case G_REGEX_ERROR_UNTERMINATED_COMMENT
:
372 *errmsg
= _("missing ) after comment");
374 case G_REGEX_ERROR_EXPRESSION_TOO_LARGE
:
375 *errmsg
= _("regular expression is too large");
377 case G_REGEX_ERROR_MEMORY_ERROR
:
378 *errmsg
= _("failed to get memory");
380 case 122: /* unmatched parentheses */
381 *errcode
= G_REGEX_ERROR_UNMATCHED_PARENTHESIS
;
382 *errmsg
= _(") without opening (");
384 case 123: /* internal error: code overflow */
385 *errcode
= G_REGEX_ERROR_INTERNAL
;
386 *errmsg
= _("code overflow");
388 case 124: /* "unrecognized character after (?<\0 */
389 *errcode
= G_REGEX_ERROR_UNRECOGNIZED_CHARACTER
;
390 *errmsg
= _("unrecognized character after (?<");
392 case G_REGEX_ERROR_VARIABLE_LENGTH_LOOKBEHIND
:
393 *errmsg
= _("lookbehind assertion is not fixed length");
395 case G_REGEX_ERROR_MALFORMED_CONDITION
:
396 *errmsg
= _("malformed number or name after (?(");
398 case G_REGEX_ERROR_TOO_MANY_CONDITIONAL_BRANCHES
:
399 *errmsg
= _("conditional group contains more than two branches");
401 case G_REGEX_ERROR_ASSERTION_EXPECTED
:
402 *errmsg
= _("assertion expected after (?(");
405 *errcode
= G_REGEX_ERROR_UNMATCHED_PARENTHESIS
;
406 /* translators: '(?R' and '(?[+-]digits' are both meant as (groups of)
407 * sequences here, '(?-54' would be an example for the second group.
409 *errmsg
= _("(?R or (?[+-]digits must be followed by )");
411 case G_REGEX_ERROR_UNKNOWN_POSIX_CLASS_NAME
:
412 *errmsg
= _("unknown POSIX class name");
414 case G_REGEX_ERROR_POSIX_COLLATING_ELEMENTS_NOT_SUPPORTED
:
415 *errmsg
= _("POSIX collating elements are not supported");
417 case G_REGEX_ERROR_HEX_CODE_TOO_LARGE
:
418 *errmsg
= _("character value in \\x{...} sequence is too large");
420 case G_REGEX_ERROR_INVALID_CONDITION
:
421 *errmsg
= _("invalid condition (?(0)");
423 case G_REGEX_ERROR_SINGLE_BYTE_MATCH_IN_LOOKBEHIND
:
424 *errmsg
= _("\\C not allowed in lookbehind assertion");
426 case 137: /* PCRE does not support \\L, \\l, \\N{name}, \\U, or \\u\0 */
427 /* A number of Perl escapes are not handled by PCRE.
428 * Therefore it explicitly raises ERR37.
430 *errcode
= G_REGEX_ERROR_UNRECOGNIZED_ESCAPE
;
431 *errmsg
= _("escapes \\L, \\l, \\N{name}, \\U, and \\u are not supported");
433 case G_REGEX_ERROR_INFINITE_LOOP
:
434 *errmsg
= _("recursive call could loop indefinitely");
436 case 141: /* unrecognized character after (?P\0 */
437 *errcode
= G_REGEX_ERROR_UNRECOGNIZED_CHARACTER
;
438 *errmsg
= _("unrecognized character after (?P");
440 case G_REGEX_ERROR_MISSING_SUBPATTERN_NAME_TERMINATOR
:
441 *errmsg
= _("missing terminator in subpattern name");
443 case G_REGEX_ERROR_DUPLICATE_SUBPATTERN_NAME
:
444 *errmsg
= _("two named subpatterns have the same name");
446 case G_REGEX_ERROR_MALFORMED_PROPERTY
:
447 *errmsg
= _("malformed \\P or \\p sequence");
449 case G_REGEX_ERROR_UNKNOWN_PROPERTY
:
450 *errmsg
= _("unknown property name after \\P or \\p");
452 case G_REGEX_ERROR_SUBPATTERN_NAME_TOO_LONG
:
453 *errmsg
= _("subpattern name is too long (maximum 32 characters)");
455 case G_REGEX_ERROR_TOO_MANY_SUBPATTERNS
:
456 *errmsg
= _("too many named subpatterns (maximum 10,000)");
458 case G_REGEX_ERROR_INVALID_OCTAL_VALUE
:
459 *errmsg
= _("octal value is greater than \\377");
461 case 152: /* internal error: overran compiling workspace */
462 *errcode
= G_REGEX_ERROR_INTERNAL
;
463 *errmsg
= _("overran compiling workspace");
465 case 153: /* internal error: previously-checked referenced subpattern not found */
466 *errcode
= G_REGEX_ERROR_INTERNAL
;
467 *errmsg
= _("previously-checked referenced subpattern not found");
469 case G_REGEX_ERROR_TOO_MANY_BRANCHES_IN_DEFINE
:
470 *errmsg
= _("DEFINE group contains more than one branch");
472 case G_REGEX_ERROR_INCONSISTENT_NEWLINE_OPTIONS
:
473 *errmsg
= _("inconsistent NEWLINE options");
475 case G_REGEX_ERROR_MISSING_BACK_REFERENCE
:
476 *errmsg
= _("\\g is not followed by a braced, angle-bracketed, or quoted name or "
477 "number, or by a plain number");
479 case G_REGEX_ERROR_INVALID_RELATIVE_REFERENCE
:
480 *errmsg
= _("a numbered reference must not be zero");
482 case G_REGEX_ERROR_BACKTRACKING_CONTROL_VERB_ARGUMENT_FORBIDDEN
:
483 *errmsg
= _("an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)");
485 case G_REGEX_ERROR_UNKNOWN_BACKTRACKING_CONTROL_VERB
:
486 *errmsg
= _("(*VERB) not recognized");
488 case G_REGEX_ERROR_NUMBER_TOO_BIG
:
489 *errmsg
= _("number is too big");
491 case G_REGEX_ERROR_MISSING_SUBPATTERN_NAME
:
492 *errmsg
= _("missing subpattern name after (?&");
494 case G_REGEX_ERROR_MISSING_DIGIT
:
495 *errmsg
= _("digit expected after (?+");
497 case G_REGEX_ERROR_INVALID_DATA_CHARACTER
:
498 *errmsg
= _("] is an invalid data character in JavaScript compatibility mode");
500 case G_REGEX_ERROR_EXTRA_SUBPATTERN_NAME
:
501 *errmsg
= _("different names for subpatterns of the same number are not allowed");
503 case G_REGEX_ERROR_BACKTRACKING_CONTROL_VERB_ARGUMENT_REQUIRED
:
504 *errmsg
= _("(*MARK) must have an argument");
506 case G_REGEX_ERROR_INVALID_CONTROL_CHAR
:
507 *errmsg
= _( "\\c must be followed by an ASCII character");
509 case G_REGEX_ERROR_MISSING_NAME
:
510 *errmsg
= _("\\k is not followed by a braced, angle-bracketed, or quoted name");
512 case G_REGEX_ERROR_NOT_SUPPORTED_IN_CLASS
:
513 *errmsg
= _("\\N is not supported in a class");
515 case G_REGEX_ERROR_TOO_MANY_FORWARD_REFERENCES
:
516 *errmsg
= _("too many forward references");
518 case G_REGEX_ERROR_NAME_TOO_LONG
:
519 *errmsg
= _("name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)");
521 case G_REGEX_ERROR_CHARACTER_VALUE_TOO_LARGE
:
522 *errmsg
= _("character value in \\u.... sequence is too large");
525 case 116: /* erroffset passed as NULL */
526 /* This should not happen as we never pass a NULL erroffset */
527 g_warning ("erroffset passed as NULL");
528 *errcode
= G_REGEX_ERROR_COMPILE
;
530 case 117: /* unknown option bit(s) set */
531 /* This should not happen as we check options before passing them
532 * to pcre_compile2() */
533 g_warning ("unknown option bit(s) set");
534 *errcode
= G_REGEX_ERROR_COMPILE
;
536 case 132: /* this version of PCRE is compiled without UTF support */
537 case 144: /* invalid UTF-8 string */
538 case 145: /* support for \\P, \\p, and \\X has not been compiled */
539 case 167: /* this version of PCRE is not compiled with Unicode property support */
540 case 173: /* disallowed Unicode code point (>= 0xd800 && <= 0xdfff) */
541 case 174: /* invalid UTF-16 string */
542 /* These errors should not happen as we are using an UTF-8 and UCP-enabled PCRE
543 * and we do not check if strings are valid */
544 case 170: /* internal error: unknown opcode in find_fixedlength() */
545 *errcode
= G_REGEX_ERROR_INTERNAL
;
549 *errcode
= G_REGEX_ERROR_COMPILE
;
556 match_info_new (const GRegex
*regex
,
563 GMatchInfo
*match_info
;
566 string_len
= strlen (string
);
568 match_info
= g_new0 (GMatchInfo
, 1);
569 match_info
->ref_count
= 1;
570 match_info
->regex
= g_regex_ref ((GRegex
*)regex
);
571 match_info
->string
= string
;
572 match_info
->string_len
= string_len
;
573 match_info
->matches
= PCRE_ERROR_NOMATCH
;
574 match_info
->pos
= start_position
;
575 match_info
->match_opts
= match_options
;
579 /* These values should be enough for most cases, if they are not
580 * enough g_regex_match_all_full() will expand them. */
581 match_info
->n_offsets
= 24;
582 match_info
->n_workspace
= 100;
583 match_info
->workspace
= g_new (gint
, match_info
->n_workspace
);
588 pcre_fullinfo (regex
->pcre_re
, regex
->extra
,
589 PCRE_INFO_CAPTURECOUNT
, &capture_count
);
590 match_info
->n_offsets
= (capture_count
+ 1) * 3;
593 match_info
->offsets
= g_new0 (gint
, match_info
->n_offsets
);
594 /* Set an invalid position for the previous match. */
595 match_info
->offsets
[0] = -1;
596 match_info
->offsets
[1] = -1;
602 * g_match_info_get_regex:
603 * @match_info: a #GMatchInfo
605 * Returns #GRegex object used in @match_info. It belongs to Glib
606 * and must not be freed. Use g_regex_ref() if you need to keep it
607 * after you free @match_info object.
609 * Returns: #GRegex object used in @match_info
614 g_match_info_get_regex (const GMatchInfo
*match_info
)
616 g_return_val_if_fail (match_info
!= NULL
, NULL
);
617 return match_info
->regex
;
621 * g_match_info_get_string:
622 * @match_info: a #GMatchInfo
624 * Returns the string searched with @match_info. This is the
625 * string passed to g_regex_match() or g_regex_replace() so
626 * you may not free it before calling this function.
628 * Returns: the string searched with @match_info
633 g_match_info_get_string (const GMatchInfo
*match_info
)
635 g_return_val_if_fail (match_info
!= NULL
, NULL
);
636 return match_info
->string
;
641 * @match_info: a #GMatchInfo
643 * Increases reference count of @match_info by 1.
645 * Returns: @match_info
650 g_match_info_ref (GMatchInfo
*match_info
)
652 g_return_val_if_fail (match_info
!= NULL
, NULL
);
653 g_atomic_int_inc (&match_info
->ref_count
);
658 * g_match_info_unref:
659 * @match_info: a #GMatchInfo
661 * Decreases reference count of @match_info by 1. When reference count drops
662 * to zero, it frees all the memory associated with the match_info structure.
667 g_match_info_unref (GMatchInfo
*match_info
)
669 if (g_atomic_int_dec_and_test (&match_info
->ref_count
))
671 g_regex_unref (match_info
->regex
);
672 g_free (match_info
->offsets
);
673 g_free (match_info
->workspace
);
680 * @match_info: (nullable): a #GMatchInfo, or %NULL
682 * If @match_info is not %NULL, calls g_match_info_unref(); otherwise does
688 g_match_info_free (GMatchInfo
*match_info
)
690 if (match_info
== NULL
)
693 g_match_info_unref (match_info
);
698 * @match_info: a #GMatchInfo structure
699 * @error: location to store the error occurring, or %NULL to ignore errors
701 * Scans for the next match using the same parameters of the previous
702 * call to g_regex_match_full() or g_regex_match() that returned
705 * The match is done on the string passed to the match function, so you
706 * cannot free it before calling this function.
708 * Returns: %TRUE is the string matched, %FALSE otherwise
713 g_match_info_next (GMatchInfo
*match_info
,
716 gint prev_match_start
;
719 g_return_val_if_fail (match_info
!= NULL
, FALSE
);
720 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, FALSE
);
721 g_return_val_if_fail (match_info
->pos
>= 0, FALSE
);
723 prev_match_start
= match_info
->offsets
[0];
724 prev_match_end
= match_info
->offsets
[1];
726 if (match_info
->pos
> match_info
->string_len
)
728 /* we have reached the end of the string */
729 match_info
->pos
= -1;
730 match_info
->matches
= PCRE_ERROR_NOMATCH
;
734 match_info
->matches
= pcre_exec (match_info
->regex
->pcre_re
,
735 match_info
->regex
->extra
,
737 match_info
->string_len
,
739 match_info
->regex
->match_opts
| match_info
->match_opts
,
741 match_info
->n_offsets
);
742 if (IS_PCRE_ERROR (match_info
->matches
))
744 g_set_error (error
, G_REGEX_ERROR
, G_REGEX_ERROR_MATCH
,
745 _("Error while matching regular expression %s: %s"),
746 match_info
->regex
->pattern
, match_error (match_info
->matches
));
750 /* avoid infinite loops if the pattern is an empty string or something
752 if (match_info
->pos
== match_info
->offsets
[1])
754 if (match_info
->pos
> match_info
->string_len
)
756 /* we have reached the end of the string */
757 match_info
->pos
= -1;
758 match_info
->matches
= PCRE_ERROR_NOMATCH
;
762 match_info
->pos
= NEXT_CHAR (match_info
->regex
,
763 &match_info
->string
[match_info
->pos
]) -
768 match_info
->pos
= match_info
->offsets
[1];
771 /* it's possible to get two identical matches when we are matching
772 * empty strings, for instance if the pattern is "(?=[A-Z0-9])" and
773 * the string is "RegExTest" we have:
774 * - search at position 0: match from 0 to 0
775 * - search at position 1: match from 3 to 3
776 * - search at position 3: match from 3 to 3 (duplicate)
777 * - search at position 4: match from 5 to 5
778 * - search at position 5: match from 5 to 5 (duplicate)
779 * - search at position 6: no match -> stop
780 * so we have to ignore the duplicates.
781 * see bug #515944: http://bugzilla.gnome.org/show_bug.cgi?id=515944 */
782 if (match_info
->matches
>= 0 &&
783 prev_match_start
== match_info
->offsets
[0] &&
784 prev_match_end
== match_info
->offsets
[1])
786 /* ignore this match and search the next one */
787 return g_match_info_next (match_info
, error
);
790 return match_info
->matches
>= 0;
794 * g_match_info_matches:
795 * @match_info: a #GMatchInfo structure
797 * Returns whether the previous match operation succeeded.
799 * Returns: %TRUE if the previous match operation succeeded,
805 g_match_info_matches (const GMatchInfo
*match_info
)
807 g_return_val_if_fail (match_info
!= NULL
, FALSE
);
809 return match_info
->matches
>= 0;
813 * g_match_info_get_match_count:
814 * @match_info: a #GMatchInfo structure
816 * Retrieves the number of matched substrings (including substring 0,
817 * that is the whole matched text), so 1 is returned if the pattern
818 * has no substrings in it and 0 is returned if the match failed.
820 * If the last match was obtained using the DFA algorithm, that is
821 * using g_regex_match_all() or g_regex_match_all_full(), the retrieved
822 * count is not that of the number of capturing parentheses but that of
823 * the number of matched substrings.
825 * Returns: Number of matched substrings, or -1 if an error occurred
830 g_match_info_get_match_count (const GMatchInfo
*match_info
)
832 g_return_val_if_fail (match_info
, -1);
834 if (match_info
->matches
== PCRE_ERROR_NOMATCH
)
837 else if (match_info
->matches
< PCRE_ERROR_NOMATCH
)
842 return match_info
->matches
;
846 * g_match_info_is_partial_match:
847 * @match_info: a #GMatchInfo structure
849 * Usually if the string passed to g_regex_match*() matches as far as
850 * it goes, but is too short to match the entire pattern, %FALSE is
851 * returned. There are circumstances where it might be helpful to
852 * distinguish this case from other cases in which there is no match.
854 * Consider, for example, an application where a human is required to
855 * type in data for a field with specific formatting requirements. An
856 * example might be a date in the form ddmmmyy, defined by the pattern
857 * "^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$".
858 * If the application sees the user’s keystrokes one by one, and can
859 * check that what has been typed so far is potentially valid, it is
860 * able to raise an error as soon as a mistake is made.
862 * GRegex supports the concept of partial matching by means of the
863 * #G_REGEX_MATCH_PARTIAL_SOFT and #G_REGEX_MATCH_PARTIAL_HARD flags.
864 * When they are used, the return code for
865 * g_regex_match() or g_regex_match_full() is, as usual, %TRUE
866 * for a complete match, %FALSE otherwise. But, when these functions
867 * return %FALSE, you can check if the match was partial calling
868 * g_match_info_is_partial_match().
870 * The difference between #G_REGEX_MATCH_PARTIAL_SOFT and
871 * #G_REGEX_MATCH_PARTIAL_HARD is that when a partial match is encountered
872 * with #G_REGEX_MATCH_PARTIAL_SOFT, matching continues to search for a
873 * possible complete match, while with #G_REGEX_MATCH_PARTIAL_HARD matching
874 * stops at the partial match.
875 * When both #G_REGEX_MATCH_PARTIAL_SOFT and #G_REGEX_MATCH_PARTIAL_HARD
876 * are set, the latter takes precedence.
878 * There were formerly some restrictions on the pattern for partial matching.
879 * The restrictions no longer apply.
881 * See pcrepartial(3) for more information on partial matching.
883 * Returns: %TRUE if the match was partial, %FALSE otherwise
888 g_match_info_is_partial_match (const GMatchInfo
*match_info
)
890 g_return_val_if_fail (match_info
!= NULL
, FALSE
);
892 return match_info
->matches
== PCRE_ERROR_PARTIAL
;
896 * g_match_info_expand_references:
897 * @match_info: (nullable): a #GMatchInfo or %NULL
898 * @string_to_expand: the string to expand
899 * @error: location to store the error occurring, or %NULL to ignore errors
901 * Returns a new string containing the text in @string_to_expand with
902 * references and escape sequences expanded. References refer to the last
903 * match done with @string against @regex and have the same syntax used by
906 * The @string_to_expand must be UTF-8 encoded even if #G_REGEX_RAW was
907 * passed to g_regex_new().
909 * The backreferences are extracted from the string passed to the match
910 * function, so you cannot call this function after freeing the string.
912 * @match_info may be %NULL in which case @string_to_expand must not
913 * contain references. For instance "foo\n" does not refer to an actual
914 * pattern and '\n' merely will be replaced with \n character,
915 * while to expand "\0" (whole match) one needs the result of a match.
916 * Use g_regex_check_replacement() to find out whether @string_to_expand
917 * contains references.
919 * Returns: (nullable): the expanded string, or %NULL if an error occurred
924 g_match_info_expand_references (const GMatchInfo
*match_info
,
925 const gchar
*string_to_expand
,
930 GError
*tmp_error
= NULL
;
932 g_return_val_if_fail (string_to_expand
!= NULL
, NULL
);
933 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, NULL
);
935 list
= split_replacement (string_to_expand
, &tmp_error
);
936 if (tmp_error
!= NULL
)
938 g_propagate_error (error
, tmp_error
);
942 if (!match_info
&& interpolation_list_needs_match (list
))
944 g_critical ("String '%s' contains references to the match, can't "
945 "expand references without GMatchInfo object",
950 result
= g_string_sized_new (strlen (string_to_expand
));
951 interpolate_replacement (match_info
, result
, list
);
953 g_list_free_full (list
, (GDestroyNotify
) free_interpolation_data
);
955 return g_string_free (result
, FALSE
);
959 * g_match_info_fetch:
960 * @match_info: #GMatchInfo structure
961 * @match_num: number of the sub expression
963 * Retrieves the text matching the @match_num'th capturing
964 * parentheses. 0 is the full text of the match, 1 is the first paren
965 * set, 2 the second, and so on.
967 * If @match_num is a valid sub pattern but it didn't match anything
968 * (e.g. sub pattern 1, matching "b" against "(a)?b") then an empty
969 * string is returned.
971 * If the match was obtained using the DFA algorithm, that is using
972 * g_regex_match_all() or g_regex_match_all_full(), the retrieved
973 * string is not that of a set of parentheses but that of a matched
974 * substring. Substrings are matched in reverse order of length, so
975 * 0 is the longest match.
977 * The string is fetched from the string passed to the match function,
978 * so you cannot call this function after freeing the string.
980 * Returns: (nullable): The matched substring, or %NULL if an error
981 * occurred. You have to free the string yourself
986 g_match_info_fetch (const GMatchInfo
*match_info
,
989 /* we cannot use pcre_get_substring() because it allocates the
990 * string using pcre_malloc(). */
994 g_return_val_if_fail (match_info
!= NULL
, NULL
);
995 g_return_val_if_fail (match_num
>= 0, NULL
);
997 /* match_num does not exist or it didn't matched, i.e. matching "b"
998 * against "(a)?b" then group 0 is empty. */
999 if (!g_match_info_fetch_pos (match_info
, match_num
, &start
, &end
))
1001 else if (start
== -1)
1002 match
= g_strdup ("");
1004 match
= g_strndup (&match_info
->string
[start
], end
- start
);
1010 * g_match_info_fetch_pos:
1011 * @match_info: #GMatchInfo structure
1012 * @match_num: number of the sub expression
1013 * @start_pos: (out) (optional): pointer to location where to store
1014 * the start position, or %NULL
1015 * @end_pos: (out) (optional): pointer to location where to store
1016 * the end position, or %NULL
1018 * Retrieves the position in bytes of the @match_num'th capturing
1019 * parentheses. 0 is the full text of the match, 1 is the first
1020 * paren set, 2 the second, and so on.
1022 * If @match_num is a valid sub pattern but it didn't match anything
1023 * (e.g. sub pattern 1, matching "b" against "(a)?b") then @start_pos
1024 * and @end_pos are set to -1 and %TRUE is returned.
1026 * If the match was obtained using the DFA algorithm, that is using
1027 * g_regex_match_all() or g_regex_match_all_full(), the retrieved
1028 * position is not that of a set of parentheses but that of a matched
1029 * substring. Substrings are matched in reverse order of length, so
1030 * 0 is the longest match.
1032 * Returns: %TRUE if the position was fetched, %FALSE otherwise. If
1033 * the position cannot be fetched, @start_pos and @end_pos are left
1039 g_match_info_fetch_pos (const GMatchInfo
*match_info
,
1044 g_return_val_if_fail (match_info
!= NULL
, FALSE
);
1045 g_return_val_if_fail (match_num
>= 0, FALSE
);
1047 /* make sure the sub expression number they're requesting is less than
1048 * the total number of sub expressions that were matched. */
1049 if (match_num
>= match_info
->matches
)
1052 if (start_pos
!= NULL
)
1053 *start_pos
= match_info
->offsets
[2 * match_num
];
1055 if (end_pos
!= NULL
)
1056 *end_pos
= match_info
->offsets
[2 * match_num
+ 1];
1062 * Returns number of first matched subpattern with name @name.
1063 * There may be more than one in case when DUPNAMES is used,
1064 * and not all subpatterns with that name match;
1065 * pcre_get_stringnumber() does not work in that case.
1068 get_matched_substring_number (const GMatchInfo
*match_info
,
1072 gchar
*first
, *last
;
1075 if (!(match_info
->regex
->compile_opts
& G_REGEX_DUPNAMES
))
1076 return pcre_get_stringnumber (match_info
->regex
->pcre_re
, name
);
1078 /* This code is copied from pcre_get.c: get_first_set() */
1079 entrysize
= pcre_get_stringtable_entries (match_info
->regex
->pcre_re
,
1087 for (entry
= (guchar
*) first
; entry
<= (guchar
*) last
; entry
+= entrysize
)
1089 gint n
= (entry
[0] << 8) + entry
[1];
1090 if (match_info
->offsets
[n
*2] >= 0)
1094 return (first
[0] << 8) + first
[1];
1098 * g_match_info_fetch_named:
1099 * @match_info: #GMatchInfo structure
1100 * @name: name of the subexpression
1102 * Retrieves the text matching the capturing parentheses named @name.
1104 * If @name is a valid sub pattern name but it didn't match anything
1105 * (e.g. sub pattern "X", matching "b" against "(?P<X>a)?b")
1106 * then an empty string is returned.
1108 * The string is fetched from the string passed to the match function,
1109 * so you cannot call this function after freeing the string.
1111 * Returns: (nullable): The matched substring, or %NULL if an error
1112 * occurred. You have to free the string yourself
1117 g_match_info_fetch_named (const GMatchInfo
*match_info
,
1120 /* we cannot use pcre_get_named_substring() because it allocates the
1121 * string using pcre_malloc(). */
1124 g_return_val_if_fail (match_info
!= NULL
, NULL
);
1125 g_return_val_if_fail (name
!= NULL
, NULL
);
1127 num
= get_matched_substring_number (match_info
, name
);
1131 return g_match_info_fetch (match_info
, num
);
1135 * g_match_info_fetch_named_pos:
1136 * @match_info: #GMatchInfo structure
1137 * @name: name of the subexpression
1138 * @start_pos: (out) (optional): pointer to location where to store
1139 * the start position, or %NULL
1140 * @end_pos: (out) (optional): pointer to location where to store
1141 * the end position, or %NULL
1143 * Retrieves the position in bytes of the capturing parentheses named @name.
1145 * If @name is a valid sub pattern name but it didn't match anything
1146 * (e.g. sub pattern "X", matching "b" against "(?P<X>a)?b")
1147 * then @start_pos and @end_pos are set to -1 and %TRUE is returned.
1149 * Returns: %TRUE if the position was fetched, %FALSE otherwise.
1150 * If the position cannot be fetched, @start_pos and @end_pos
1151 * are left unchanged.
1156 g_match_info_fetch_named_pos (const GMatchInfo
*match_info
,
1163 g_return_val_if_fail (match_info
!= NULL
, FALSE
);
1164 g_return_val_if_fail (name
!= NULL
, FALSE
);
1166 num
= get_matched_substring_number (match_info
, name
);
1170 return g_match_info_fetch_pos (match_info
, num
, start_pos
, end_pos
);
1174 * g_match_info_fetch_all:
1175 * @match_info: a #GMatchInfo structure
1177 * Bundles up pointers to each of the matching substrings from a match
1178 * and stores them in an array of gchar pointers. The first element in
1179 * the returned array is the match number 0, i.e. the entire matched
1182 * If a sub pattern didn't match anything (e.g. sub pattern 1, matching
1183 * "b" against "(a)?b") then an empty string is inserted.
1185 * If the last match was obtained using the DFA algorithm, that is using
1186 * g_regex_match_all() or g_regex_match_all_full(), the retrieved
1187 * strings are not that matched by sets of parentheses but that of the
1188 * matched substring. Substrings are matched in reverse order of length,
1189 * so the first one is the longest match.
1191 * The strings are fetched from the string passed to the match function,
1192 * so you cannot call this function after freeing the string.
1194 * Returns: (transfer full): a %NULL-terminated array of gchar *
1195 * pointers. It must be freed using g_strfreev(). If the previous
1196 * match failed %NULL is returned
1201 g_match_info_fetch_all (const GMatchInfo
*match_info
)
1203 /* we cannot use pcre_get_substring_list() because the returned value
1204 * isn't suitable for g_strfreev(). */
1208 g_return_val_if_fail (match_info
!= NULL
, NULL
);
1210 if (match_info
->matches
< 0)
1213 result
= g_new (gchar
*, match_info
->matches
+ 1);
1214 for (i
= 0; i
< match_info
->matches
; i
++)
1215 result
[i
] = g_match_info_fetch (match_info
, i
);
1224 G_DEFINE_QUARK (g
-regex
-error
-quark
, g_regex_error
)
1230 * Increases reference count of @regex by 1.
1237 g_regex_ref (GRegex
*regex
)
1239 g_return_val_if_fail (regex
!= NULL
, NULL
);
1240 g_atomic_int_inc (®ex
->ref_count
);
1248 * Decreases reference count of @regex by 1. When reference count drops
1249 * to zero, it frees all the memory associated with the regex structure.
1254 g_regex_unref (GRegex
*regex
)
1256 g_return_if_fail (regex
!= NULL
);
1258 if (g_atomic_int_dec_and_test (®ex
->ref_count
))
1260 g_free (regex
->pattern
);
1261 if (regex
->pcre_re
!= NULL
)
1262 pcre_free (regex
->pcre_re
);
1263 if (regex
->extra
!= NULL
)
1264 pcre_free (regex
->extra
);
1270 * @match_options: (inout) (optional):
1272 static pcre
*regex_compile (const gchar
*pattern
,
1273 GRegexCompileFlags compile_options
,
1274 GRegexCompileFlags
*compile_options_out
,
1275 GRegexMatchFlags
*match_options
,
1280 * @pattern: the regular expression
1281 * @compile_options: compile options for the regular expression, or 0
1282 * @match_options: match options for the regular expression, or 0
1283 * @error: return location for a #GError
1285 * Compiles the regular expression to an internal form, and does
1286 * the initial setup of the #GRegex structure.
1288 * Returns: (nullable): a #GRegex structure or %NULL if an error occured. Call
1289 * g_regex_unref() when you are done with it
1294 g_regex_new (const gchar
*pattern
,
1295 GRegexCompileFlags compile_options
,
1296 GRegexMatchFlags match_options
,
1301 const gchar
*errmsg
;
1302 gboolean optimize
= FALSE
;
1303 static volatile gsize initialised
= 0;
1305 g_return_val_if_fail (pattern
!= NULL
, NULL
);
1306 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, NULL
);
1307 g_return_val_if_fail ((compile_options
& ~G_REGEX_COMPILE_MASK
) == 0, NULL
);
1308 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, NULL
);
1310 if (g_once_init_enter (&initialised
))
1312 int supports_utf8
, supports_ucp
;
1314 pcre_config (PCRE_CONFIG_UTF8
, &supports_utf8
);
1316 g_critical (_("PCRE library is compiled without UTF8 support"));
1318 pcre_config (PCRE_CONFIG_UNICODE_PROPERTIES
, &supports_ucp
);
1320 g_critical (_("PCRE library is compiled without UTF8 properties support"));
1322 g_once_init_leave (&initialised
, supports_utf8
&& supports_ucp
? 1 : 2);
1325 if (G_UNLIKELY (initialised
!= 1))
1327 g_set_error_literal (error
, G_REGEX_ERROR
, G_REGEX_ERROR_COMPILE
,
1328 _("PCRE library is compiled with incompatible options"));
1332 /* G_REGEX_OPTIMIZE has the same numeric value of PCRE_NO_UTF8_CHECK,
1333 * as we do not need to wrap PCRE_NO_UTF8_CHECK. */
1334 if (compile_options
& G_REGEX_OPTIMIZE
)
1337 re
= regex_compile (pattern
, compile_options
, &compile_options
,
1338 &match_options
, error
);
1343 regex
= g_new0 (GRegex
, 1);
1344 regex
->ref_count
= 1;
1345 regex
->pattern
= g_strdup (pattern
);
1346 regex
->pcre_re
= re
;
1347 regex
->compile_opts
= compile_options
;
1348 regex
->match_opts
= match_options
;
1352 regex
->extra
= pcre_study (regex
->pcre_re
, 0, &errmsg
);
1355 GError
*tmp_error
= g_error_new (G_REGEX_ERROR
,
1356 G_REGEX_ERROR_OPTIMIZE
,
1357 _("Error while optimizing "
1358 "regular expression %s: %s"),
1361 g_propagate_error (error
, tmp_error
);
1363 g_regex_unref (regex
);
1372 regex_compile (const gchar
*pattern
,
1373 GRegexCompileFlags compile_options
,
1374 GRegexCompileFlags
*compile_options_out
,
1375 GRegexMatchFlags
*match_options
,
1379 const gchar
*errmsg
;
1382 GRegexCompileFlags nonpcre_compile_options
;
1383 unsigned long int pcre_compile_options
;
1385 nonpcre_compile_options
= compile_options
& G_REGEX_COMPILE_NONPCRE_MASK
;
1387 /* In GRegex the string are, by default, UTF-8 encoded. PCRE
1388 * instead uses UTF-8 only if required with PCRE_UTF8. */
1389 if (compile_options
& G_REGEX_RAW
)
1392 compile_options
&= ~G_REGEX_RAW
;
1397 compile_options
|= PCRE_UTF8
| PCRE_NO_UTF8_CHECK
;
1399 if (match_options
!= NULL
)
1400 *match_options
|= PCRE_NO_UTF8_CHECK
;
1403 /* PCRE_NEWLINE_ANY is the default for the internal PCRE but
1404 * not for the system one. */
1405 if (!(compile_options
& G_REGEX_NEWLINE_CR
) &&
1406 !(compile_options
& G_REGEX_NEWLINE_LF
))
1408 compile_options
|= PCRE_NEWLINE_ANY
;
1411 compile_options
|= PCRE_UCP
;
1413 /* PCRE_BSR_UNICODE is the default for the internal PCRE but
1414 * possibly not for the system one.
1416 if (~compile_options
& G_REGEX_BSR_ANYCRLF
)
1417 compile_options
|= PCRE_BSR_UNICODE
;
1419 /* compile the pattern */
1420 re
= pcre_compile2 (pattern
, compile_options
, &errcode
,
1421 &errmsg
, &erroffset
, NULL
);
1423 /* if the compilation failed, set the error member and return
1429 /* Translate the PCRE error code to GRegexError and use a translated
1430 * error message if possible */
1431 translate_compile_error (&errcode
, &errmsg
);
1433 /* PCRE uses byte offsets but we want to show character offsets */
1434 erroffset
= g_utf8_pointer_to_offset (pattern
, &pattern
[erroffset
]);
1436 tmp_error
= g_error_new (G_REGEX_ERROR
, errcode
,
1437 _("Error while compiling regular "
1438 "expression %s at char %d: %s"),
1439 pattern
, erroffset
, errmsg
);
1440 g_propagate_error (error
, tmp_error
);
1445 /* For options set at the beginning of the pattern, pcre puts them into
1446 * compile options, e.g. "(?i)foo" will make the pcre structure store
1447 * PCRE_CASELESS even though it wasn't explicitly given for compilation. */
1448 pcre_fullinfo (re
, NULL
, PCRE_INFO_OPTIONS
, &pcre_compile_options
);
1449 compile_options
= pcre_compile_options
& G_REGEX_COMPILE_PCRE_MASK
;
1451 /* Don't leak PCRE_NEWLINE_ANY, which is part of PCRE_NEWLINE_ANYCRLF */
1452 if ((pcre_compile_options
& PCRE_NEWLINE_ANYCRLF
) != PCRE_NEWLINE_ANYCRLF
)
1453 compile_options
&= ~PCRE_NEWLINE_ANY
;
1455 compile_options
|= nonpcre_compile_options
;
1457 if (!(compile_options
& G_REGEX_DUPNAMES
))
1459 gboolean jchanged
= FALSE
;
1460 pcre_fullinfo (re
, NULL
, PCRE_INFO_JCHANGED
, &jchanged
);
1462 compile_options
|= G_REGEX_DUPNAMES
;
1465 if (compile_options_out
!= 0)
1466 *compile_options_out
= compile_options
;
1472 * g_regex_get_pattern:
1473 * @regex: a #GRegex structure
1475 * Gets the pattern string associated with @regex, i.e. a copy of
1476 * the string passed to g_regex_new().
1478 * Returns: the pattern of @regex
1483 g_regex_get_pattern (const GRegex
*regex
)
1485 g_return_val_if_fail (regex
!= NULL
, NULL
);
1487 return regex
->pattern
;
1491 * g_regex_get_max_backref:
1494 * Returns the number of the highest back reference
1495 * in the pattern, or 0 if the pattern does not contain
1498 * Returns: the number of the highest back reference
1503 g_regex_get_max_backref (const GRegex
*regex
)
1507 pcre_fullinfo (regex
->pcre_re
, regex
->extra
,
1508 PCRE_INFO_BACKREFMAX
, &value
);
1514 * g_regex_get_capture_count:
1517 * Returns the number of capturing subpatterns in the pattern.
1519 * Returns: the number of capturing subpatterns
1524 g_regex_get_capture_count (const GRegex
*regex
)
1528 pcre_fullinfo (regex
->pcre_re
, regex
->extra
,
1529 PCRE_INFO_CAPTURECOUNT
, &value
);
1535 * g_regex_get_has_cr_or_lf:
1536 * @regex: a #GRegex structure
1538 * Checks whether the pattern contains explicit CR or LF references.
1540 * Returns: %TRUE if the pattern contains explicit CR or LF references
1545 g_regex_get_has_cr_or_lf (const GRegex
*regex
)
1549 pcre_fullinfo (regex
->pcre_re
, regex
->extra
,
1550 PCRE_INFO_HASCRORLF
, &value
);
1556 * g_regex_get_max_lookbehind:
1557 * @regex: a #GRegex structure
1559 * Gets the number of characters in the longest lookbehind assertion in the
1560 * pattern. This information is useful when doing multi-segment matching using
1561 * the partial matching facilities.
1563 * Returns: the number of characters in the longest lookbehind assertion.
1568 g_regex_get_max_lookbehind (const GRegex
*regex
)
1570 gint max_lookbehind
;
1572 pcre_fullinfo (regex
->pcre_re
, regex
->extra
,
1573 PCRE_INFO_MAXLOOKBEHIND
, &max_lookbehind
);
1575 return max_lookbehind
;
1579 * g_regex_get_compile_flags:
1582 * Returns the compile options that @regex was created with.
1584 * Depending on the version of PCRE that is used, this may or may not
1585 * include flags set by option expressions such as `(?i)` found at the
1586 * top-level within the compiled pattern.
1588 * Returns: flags from #GRegexCompileFlags
1593 g_regex_get_compile_flags (const GRegex
*regex
)
1595 g_return_val_if_fail (regex
!= NULL
, 0);
1597 return regex
->compile_opts
;
1601 * g_regex_get_match_flags:
1604 * Returns the match options that @regex was created with.
1606 * Returns: flags from #GRegexMatchFlags
1611 g_regex_get_match_flags (const GRegex
*regex
)
1613 g_return_val_if_fail (regex
!= NULL
, 0);
1615 return regex
->match_opts
& G_REGEX_MATCH_MASK
;
1619 * g_regex_match_simple:
1620 * @pattern: the regular expression
1621 * @string: the string to scan for matches
1622 * @compile_options: compile options for the regular expression, or 0
1623 * @match_options: match options, or 0
1625 * Scans for a match in @string for @pattern.
1627 * This function is equivalent to g_regex_match() but it does not
1628 * require to compile the pattern with g_regex_new(), avoiding some
1629 * lines of code when you need just to do a match without extracting
1630 * substrings, capture counts, and so on.
1632 * If this function is to be called on the same @pattern more than
1633 * once, it's more efficient to compile the pattern once with
1634 * g_regex_new() and then use g_regex_match().
1636 * Returns: %TRUE if the string matched, %FALSE otherwise
1641 g_regex_match_simple (const gchar
*pattern
,
1642 const gchar
*string
,
1643 GRegexCompileFlags compile_options
,
1644 GRegexMatchFlags match_options
)
1649 regex
= g_regex_new (pattern
, compile_options
, 0, NULL
);
1652 result
= g_regex_match_full (regex
, string
, -1, 0, match_options
, NULL
, NULL
);
1653 g_regex_unref (regex
);
1659 * @regex: a #GRegex structure from g_regex_new()
1660 * @string: the string to scan for matches
1661 * @match_options: match options
1662 * @match_info: (out) (optional): pointer to location where to store
1663 * the #GMatchInfo, or %NULL if you do not need it
1665 * Scans for a match in @string for the pattern in @regex.
1666 * The @match_options are combined with the match options specified
1667 * when the @regex structure was created, letting you have more
1668 * flexibility in reusing #GRegex structures.
1670 * Unless %G_REGEX_RAW is specified in the options, @string must be valid UTF-8.
1672 * A #GMatchInfo structure, used to get information on the match,
1673 * is stored in @match_info if not %NULL. Note that if @match_info
1674 * is not %NULL then it is created even if the function returns %FALSE,
1675 * i.e. you must free it regardless if regular expression actually matched.
1677 * To retrieve all the non-overlapping matches of the pattern in
1678 * string you can use g_match_info_next().
1680 * |[<!-- language="C" -->
1682 * print_uppercase_words (const gchar *string)
1684 * // Print all uppercase-only words.
1686 * GMatchInfo *match_info;
1688 * regex = g_regex_new ("[A-Z]+", 0, 0, NULL);
1689 * g_regex_match (regex, string, 0, &match_info);
1690 * while (g_match_info_matches (match_info))
1692 * gchar *word = g_match_info_fetch (match_info, 0);
1693 * g_print ("Found: %s\n", word);
1695 * g_match_info_next (match_info, NULL);
1697 * g_match_info_free (match_info);
1698 * g_regex_unref (regex);
1702 * @string is not copied and is used in #GMatchInfo internally. If
1703 * you use any #GMatchInfo method (except g_match_info_free()) after
1704 * freeing or modifying @string then the behaviour is undefined.
1706 * Returns: %TRUE is the string matched, %FALSE otherwise
1711 g_regex_match (const GRegex
*regex
,
1712 const gchar
*string
,
1713 GRegexMatchFlags match_options
,
1714 GMatchInfo
**match_info
)
1716 return g_regex_match_full (regex
, string
, -1, 0, match_options
,
1721 * g_regex_match_full:
1722 * @regex: a #GRegex structure from g_regex_new()
1723 * @string: (array length=string_len): the string to scan for matches
1724 * @string_len: the length of @string, in bytes, or -1 if @string is nul-terminated
1725 * @start_position: starting index of the string to match, in bytes
1726 * @match_options: match options
1727 * @match_info: (out) (optional): pointer to location where to store
1728 * the #GMatchInfo, or %NULL if you do not need it
1729 * @error: location to store the error occurring, or %NULL to ignore errors
1731 * Scans for a match in @string for the pattern in @regex.
1732 * The @match_options are combined with the match options specified
1733 * when the @regex structure was created, letting you have more
1734 * flexibility in reusing #GRegex structures.
1736 * Setting @start_position differs from just passing over a shortened
1737 * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
1738 * that begins with any kind of lookbehind assertion, such as "\b".
1740 * Unless %G_REGEX_RAW is specified in the options, @string must be valid UTF-8.
1742 * A #GMatchInfo structure, used to get information on the match, is
1743 * stored in @match_info if not %NULL. Note that if @match_info is
1744 * not %NULL then it is created even if the function returns %FALSE,
1745 * i.e. you must free it regardless if regular expression actually
1748 * @string is not copied and is used in #GMatchInfo internally. If
1749 * you use any #GMatchInfo method (except g_match_info_free()) after
1750 * freeing or modifying @string then the behaviour is undefined.
1752 * To retrieve all the non-overlapping matches of the pattern in
1753 * string you can use g_match_info_next().
1755 * |[<!-- language="C" -->
1757 * print_uppercase_words (const gchar *string)
1759 * // Print all uppercase-only words.
1761 * GMatchInfo *match_info;
1762 * GError *error = NULL;
1764 * regex = g_regex_new ("[A-Z]+", 0, 0, NULL);
1765 * g_regex_match_full (regex, string, -1, 0, 0, &match_info, &error);
1766 * while (g_match_info_matches (match_info))
1768 * gchar *word = g_match_info_fetch (match_info, 0);
1769 * g_print ("Found: %s\n", word);
1771 * g_match_info_next (match_info, &error);
1773 * g_match_info_free (match_info);
1774 * g_regex_unref (regex);
1775 * if (error != NULL)
1777 * g_printerr ("Error while matching: %s\n", error->message);
1778 * g_error_free (error);
1783 * Returns: %TRUE is the string matched, %FALSE otherwise
1788 g_regex_match_full (const GRegex
*regex
,
1789 const gchar
*string
,
1791 gint start_position
,
1792 GRegexMatchFlags match_options
,
1793 GMatchInfo
**match_info
,
1799 g_return_val_if_fail (regex
!= NULL
, FALSE
);
1800 g_return_val_if_fail (string
!= NULL
, FALSE
);
1801 g_return_val_if_fail (start_position
>= 0, FALSE
);
1802 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, FALSE
);
1803 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, FALSE
);
1805 info
= match_info_new (regex
, string
, string_len
, start_position
,
1806 match_options
, FALSE
);
1807 match_ok
= g_match_info_next (info
, error
);
1808 if (match_info
!= NULL
)
1811 g_match_info_free (info
);
1817 * g_regex_match_all:
1818 * @regex: a #GRegex structure from g_regex_new()
1819 * @string: the string to scan for matches
1820 * @match_options: match options
1821 * @match_info: (out) (optional): pointer to location where to store
1822 * the #GMatchInfo, or %NULL if you do not need it
1824 * Using the standard algorithm for regular expression matching only
1825 * the longest match in the string is retrieved. This function uses
1826 * a different algorithm so it can retrieve all the possible matches.
1827 * For more documentation see g_regex_match_all_full().
1829 * A #GMatchInfo structure, used to get information on the match, is
1830 * stored in @match_info if not %NULL. Note that if @match_info is
1831 * not %NULL then it is created even if the function returns %FALSE,
1832 * i.e. you must free it regardless if regular expression actually
1835 * @string is not copied and is used in #GMatchInfo internally. If
1836 * you use any #GMatchInfo method (except g_match_info_free()) after
1837 * freeing or modifying @string then the behaviour is undefined.
1839 * Returns: %TRUE is the string matched, %FALSE otherwise
1844 g_regex_match_all (const GRegex
*regex
,
1845 const gchar
*string
,
1846 GRegexMatchFlags match_options
,
1847 GMatchInfo
**match_info
)
1849 return g_regex_match_all_full (regex
, string
, -1, 0, match_options
,
1854 * g_regex_match_all_full:
1855 * @regex: a #GRegex structure from g_regex_new()
1856 * @string: (array length=string_len): the string to scan for matches
1857 * @string_len: the length of @string, in bytes, or -1 if @string is nul-terminated
1858 * @start_position: starting index of the string to match, in bytes
1859 * @match_options: match options
1860 * @match_info: (out) (optional): pointer to location where to store
1861 * the #GMatchInfo, or %NULL if you do not need it
1862 * @error: location to store the error occurring, or %NULL to ignore errors
1864 * Using the standard algorithm for regular expression matching only
1865 * the longest match in the @string is retrieved, it is not possible
1866 * to obtain all the available matches. For instance matching
1867 * "<a> <b> <c>" against the pattern "<.*>"
1868 * you get "<a> <b> <c>".
1870 * This function uses a different algorithm (called DFA, i.e. deterministic
1871 * finite automaton), so it can retrieve all the possible matches, all
1872 * starting at the same point in the string. For instance matching
1873 * "<a> <b> <c>" against the pattern "<.*>;"
1874 * you would obtain three matches: "<a> <b> <c>",
1875 * "<a> <b>" and "<a>".
1877 * The number of matched strings is retrieved using
1878 * g_match_info_get_match_count(). To obtain the matched strings and
1879 * their position you can use, respectively, g_match_info_fetch() and
1880 * g_match_info_fetch_pos(). Note that the strings are returned in
1881 * reverse order of length; that is, the longest matching string is
1884 * Note that the DFA algorithm is slower than the standard one and it
1885 * is not able to capture substrings, so backreferences do not work.
1887 * Setting @start_position differs from just passing over a shortened
1888 * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
1889 * that begins with any kind of lookbehind assertion, such as "\b".
1891 * Unless %G_REGEX_RAW is specified in the options, @string must be valid UTF-8.
1893 * A #GMatchInfo structure, used to get information on the match, is
1894 * stored in @match_info if not %NULL. Note that if @match_info is
1895 * not %NULL then it is created even if the function returns %FALSE,
1896 * i.e. you must free it regardless if regular expression actually
1899 * @string is not copied and is used in #GMatchInfo internally. If
1900 * you use any #GMatchInfo method (except g_match_info_free()) after
1901 * freeing or modifying @string then the behaviour is undefined.
1903 * Returns: %TRUE is the string matched, %FALSE otherwise
1908 g_regex_match_all_full (const GRegex
*regex
,
1909 const gchar
*string
,
1911 gint start_position
,
1912 GRegexMatchFlags match_options
,
1913 GMatchInfo
**match_info
,
1922 g_return_val_if_fail (regex
!= NULL
, FALSE
);
1923 g_return_val_if_fail (string
!= NULL
, FALSE
);
1924 g_return_val_if_fail (start_position
>= 0, FALSE
);
1925 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, FALSE
);
1926 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, FALSE
);
1928 #ifdef PCRE_NO_AUTO_POSSESS
1929 /* For PCRE >= 8.34 we need to turn off PCRE_NO_AUTO_POSSESS, which
1930 * is an optimization for normal regex matching, but results in omitting
1931 * some shorter matches here, and an observable behaviour change.
1933 * DFA matching is rather niche, and very rarely used according to
1934 * codesearch.debian.net, so don't bother caching the recompiled RE. */
1935 pcre_re
= regex_compile (regex
->pattern
,
1936 regex
->compile_opts
| PCRE_NO_AUTO_POSSESS
,
1939 if (pcre_re
== NULL
)
1942 /* Not bothering to cache the optimization data either, with similar
1946 /* For PCRE < 8.33 the precompiled regex is fine. */
1947 pcre_re
= regex
->pcre_re
;
1948 extra
= regex
->extra
;
1951 info
= match_info_new (regex
, string
, string_len
, start_position
,
1952 match_options
, TRUE
);
1958 info
->matches
= pcre_dfa_exec (pcre_re
, extra
,
1959 info
->string
, info
->string_len
,
1961 regex
->match_opts
| match_options
,
1962 info
->offsets
, info
->n_offsets
,
1963 info
->workspace
, info
->n_workspace
);
1964 if (info
->matches
== PCRE_ERROR_DFA_WSSIZE
)
1966 /* info->workspace is too small. */
1967 info
->n_workspace
*= 2;
1968 info
->workspace
= g_realloc (info
->workspace
,
1969 info
->n_workspace
* sizeof (gint
));
1972 else if (info
->matches
== 0)
1974 /* info->offsets is too small. */
1975 info
->n_offsets
*= 2;
1976 info
->offsets
= g_realloc (info
->offsets
,
1977 info
->n_offsets
* sizeof (gint
));
1980 else if (IS_PCRE_ERROR (info
->matches
))
1982 g_set_error (error
, G_REGEX_ERROR
, G_REGEX_ERROR_MATCH
,
1983 _("Error while matching regular expression %s: %s"),
1984 regex
->pattern
, match_error (info
->matches
));
1988 #ifdef PCRE_NO_AUTO_POSSESS
1989 pcre_free (pcre_re
);
1992 /* set info->pos to -1 so that a call to g_match_info_next() fails. */
1994 retval
= info
->matches
>= 0;
1996 if (match_info
!= NULL
)
1999 g_match_info_free (info
);
2005 * g_regex_get_string_number:
2006 * @regex: #GRegex structure
2007 * @name: name of the subexpression
2009 * Retrieves the number of the subexpression named @name.
2011 * Returns: The number of the subexpression or -1 if @name
2017 g_regex_get_string_number (const GRegex
*regex
,
2022 g_return_val_if_fail (regex
!= NULL
, -1);
2023 g_return_val_if_fail (name
!= NULL
, -1);
2025 num
= pcre_get_stringnumber (regex
->pcre_re
, name
);
2026 if (num
== PCRE_ERROR_NOSUBSTRING
)
2033 * g_regex_split_simple:
2034 * @pattern: the regular expression
2035 * @string: the string to scan for matches
2036 * @compile_options: compile options for the regular expression, or 0
2037 * @match_options: match options, or 0
2039 * Breaks the string on the pattern, and returns an array of
2040 * the tokens. If the pattern contains capturing parentheses,
2041 * then the text for each of the substrings will also be returned.
2042 * If the pattern does not match anywhere in the string, then the
2043 * whole string is returned as the first token.
2045 * This function is equivalent to g_regex_split() but it does
2046 * not require to compile the pattern with g_regex_new(), avoiding
2047 * some lines of code when you need just to do a split without
2048 * extracting substrings, capture counts, and so on.
2050 * If this function is to be called on the same @pattern more than
2051 * once, it's more efficient to compile the pattern once with
2052 * g_regex_new() and then use g_regex_split().
2054 * As a special case, the result of splitting the empty string ""
2055 * is an empty vector, not a vector containing a single string.
2056 * The reason for this special case is that being able to represent
2057 * a empty vector is typically more useful than consistent handling
2058 * of empty elements. If you do need to represent empty elements,
2059 * you'll need to check for the empty string before calling this
2062 * A pattern that can match empty strings splits @string into
2063 * separate characters wherever it matches the empty string between
2064 * characters. For example splitting "ab c" using as a separator
2065 * "\s*", you will get "a", "b" and "c".
2067 * Returns: (transfer full): a %NULL-terminated array of strings. Free
2068 * it using g_strfreev()
2073 g_regex_split_simple (const gchar
*pattern
,
2074 const gchar
*string
,
2075 GRegexCompileFlags compile_options
,
2076 GRegexMatchFlags match_options
)
2081 regex
= g_regex_new (pattern
, compile_options
, 0, NULL
);
2085 result
= g_regex_split_full (regex
, string
, -1, 0, match_options
, 0, NULL
);
2086 g_regex_unref (regex
);
2092 * @regex: a #GRegex structure
2093 * @string: the string to split with the pattern
2094 * @match_options: match time option flags
2096 * Breaks the string on the pattern, and returns an array of the tokens.
2097 * If the pattern contains capturing parentheses, then the text for each
2098 * of the substrings will also be returned. If the pattern does not match
2099 * anywhere in the string, then the whole string is returned as the first
2102 * As a special case, the result of splitting the empty string "" is an
2103 * empty vector, not a vector containing a single string. The reason for
2104 * this special case is that being able to represent a empty vector is
2105 * typically more useful than consistent handling of empty elements. If
2106 * you do need to represent empty elements, you'll need to check for the
2107 * empty string before calling this function.
2109 * A pattern that can match empty strings splits @string into separate
2110 * characters wherever it matches the empty string between characters.
2111 * For example splitting "ab c" using as a separator "\s*", you will get
2114 * Returns: (transfer full): a %NULL-terminated gchar ** array. Free
2115 * it using g_strfreev()
2120 g_regex_split (const GRegex
*regex
,
2121 const gchar
*string
,
2122 GRegexMatchFlags match_options
)
2124 return g_regex_split_full (regex
, string
, -1, 0,
2125 match_options
, 0, NULL
);
2129 * g_regex_split_full:
2130 * @regex: a #GRegex structure
2131 * @string: (array length=string_len): the string to split with the pattern
2132 * @string_len: the length of @string, in bytes, or -1 if @string is nul-terminated
2133 * @start_position: starting index of the string to match, in bytes
2134 * @match_options: match time option flags
2135 * @max_tokens: the maximum number of tokens to split @string into.
2136 * If this is less than 1, the string is split completely
2137 * @error: return location for a #GError
2139 * Breaks the string on the pattern, and returns an array of the tokens.
2140 * If the pattern contains capturing parentheses, then the text for each
2141 * of the substrings will also be returned. If the pattern does not match
2142 * anywhere in the string, then the whole string is returned as the first
2145 * As a special case, the result of splitting the empty string "" is an
2146 * empty vector, not a vector containing a single string. The reason for
2147 * this special case is that being able to represent a empty vector is
2148 * typically more useful than consistent handling of empty elements. If
2149 * you do need to represent empty elements, you'll need to check for the
2150 * empty string before calling this function.
2152 * A pattern that can match empty strings splits @string into separate
2153 * characters wherever it matches the empty string between characters.
2154 * For example splitting "ab c" using as a separator "\s*", you will get
2157 * Setting @start_position differs from just passing over a shortened
2158 * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
2159 * that begins with any kind of lookbehind assertion, such as "\b".
2161 * Returns: (transfer full): a %NULL-terminated gchar ** array. Free
2162 * it using g_strfreev()
2167 g_regex_split_full (const GRegex
*regex
,
2168 const gchar
*string
,
2170 gint start_position
,
2171 GRegexMatchFlags match_options
,
2175 GError
*tmp_error
= NULL
;
2176 GMatchInfo
*match_info
;
2181 /* position of the last separator. */
2182 gint last_separator_end
;
2183 /* was the last match 0 bytes long? */
2184 gboolean last_match_is_empty
;
2185 /* the returned array of char **s */
2186 gchar
**string_list
;
2188 g_return_val_if_fail (regex
!= NULL
, NULL
);
2189 g_return_val_if_fail (string
!= NULL
, NULL
);
2190 g_return_val_if_fail (start_position
>= 0, NULL
);
2191 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, NULL
);
2192 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, NULL
);
2194 if (max_tokens
<= 0)
2195 max_tokens
= G_MAXINT
;
2198 string_len
= strlen (string
);
2200 /* zero-length string */
2201 if (string_len
- start_position
== 0)
2202 return g_new0 (gchar
*, 1);
2204 if (max_tokens
== 1)
2206 string_list
= g_new0 (gchar
*, 2);
2207 string_list
[0] = g_strndup (&string
[start_position
],
2208 string_len
- start_position
);
2214 last_separator_end
= start_position
;
2215 last_match_is_empty
= FALSE
;
2217 match_ok
= g_regex_match_full (regex
, string
, string_len
, start_position
,
2218 match_options
, &match_info
, &tmp_error
);
2220 while (tmp_error
== NULL
)
2224 last_match_is_empty
=
2225 (match_info
->offsets
[0] == match_info
->offsets
[1]);
2227 /* we need to skip empty separators at the same position of the end
2228 * of another separator. e.g. the string is "a b" and the separator
2229 * is " *", so from 1 to 2 we have a match and at position 2 we have
2230 * an empty match. */
2231 if (last_separator_end
!= match_info
->offsets
[1])
2236 token
= g_strndup (string
+ last_separator_end
,
2237 match_info
->offsets
[0] - last_separator_end
);
2238 list
= g_list_prepend (list
, token
);
2241 /* if there were substrings, these need to be added to
2243 match_count
= g_match_info_get_match_count (match_info
);
2244 if (match_count
> 1)
2246 for (i
= 1; i
< match_count
; i
++)
2247 list
= g_list_prepend (list
, g_match_info_fetch (match_info
, i
));
2253 /* if there was no match, copy to end of string. */
2254 if (!last_match_is_empty
)
2256 gchar
*token
= g_strndup (string
+ last_separator_end
,
2257 match_info
->string_len
- last_separator_end
);
2258 list
= g_list_prepend (list
, token
);
2260 /* no more tokens, end the loop. */
2264 /* -1 to leave room for the last part. */
2265 if (token_count
>= max_tokens
- 1)
2267 /* we have reached the maximum number of tokens, so we copy
2268 * the remaining part of the string. */
2269 if (last_match_is_empty
)
2271 /* the last match was empty, so we have moved one char
2272 * after the real position to avoid empty matches at the
2274 match_info
->pos
= PREV_CHAR (regex
, &string
[match_info
->pos
]) - string
;
2276 /* the if is needed in the case we have terminated the available
2277 * tokens, but we are at the end of the string, so there are no
2278 * characters left to copy. */
2279 if (string_len
> match_info
->pos
)
2281 gchar
*token
= g_strndup (string
+ match_info
->pos
,
2282 string_len
- match_info
->pos
);
2283 list
= g_list_prepend (list
, token
);
2289 last_separator_end
= match_info
->pos
;
2290 if (last_match_is_empty
)
2291 /* if the last match was empty, g_match_info_next() has moved
2292 * forward to avoid infinite loops, but we still need to copy that
2294 last_separator_end
= PREV_CHAR (regex
, &string
[last_separator_end
]) - string
;
2296 match_ok
= g_match_info_next (match_info
, &tmp_error
);
2298 g_match_info_free (match_info
);
2299 if (tmp_error
!= NULL
)
2301 g_propagate_error (error
, tmp_error
);
2302 g_list_free_full (list
, g_free
);
2306 string_list
= g_new (gchar
*, g_list_length (list
) + 1);
2308 for (last
= g_list_last (list
); last
; last
= g_list_previous (last
))
2309 string_list
[i
++] = last
->data
;
2310 string_list
[i
] = NULL
;
2319 REPL_TYPE_CHARACTER
,
2320 REPL_TYPE_SYMBOLIC_REFERENCE
,
2321 REPL_TYPE_NUMERIC_REFERENCE
,
2322 REPL_TYPE_CHANGE_CASE
2327 CHANGE_CASE_NONE
= 1 << 0,
2328 CHANGE_CASE_UPPER
= 1 << 1,
2329 CHANGE_CASE_LOWER
= 1 << 2,
2330 CHANGE_CASE_UPPER_SINGLE
= 1 << 3,
2331 CHANGE_CASE_LOWER_SINGLE
= 1 << 4,
2332 CHANGE_CASE_SINGLE_MASK
= CHANGE_CASE_UPPER_SINGLE
| CHANGE_CASE_LOWER_SINGLE
,
2333 CHANGE_CASE_LOWER_MASK
= CHANGE_CASE_LOWER
| CHANGE_CASE_LOWER_SINGLE
,
2334 CHANGE_CASE_UPPER_MASK
= CHANGE_CASE_UPPER
| CHANGE_CASE_UPPER_SINGLE
2337 struct _InterpolationData
2343 ChangeCase change_case
;
2347 free_interpolation_data (InterpolationData
*data
)
2349 g_free (data
->text
);
2353 static const gchar
*
2354 expand_escape (const gchar
*replacement
,
2356 InterpolationData
*data
,
2361 const gchar
*error_detail
;
2363 GError
*tmp_error
= NULL
;
2371 data
->type
= REPL_TYPE_CHARACTER
;
2376 data
->type
= REPL_TYPE_CHARACTER
;
2381 data
->type
= REPL_TYPE_CHARACTER
;
2386 data
->type
= REPL_TYPE_CHARACTER
;
2391 data
->type
= REPL_TYPE_CHARACTER
;
2396 data
->type
= REPL_TYPE_CHARACTER
;
2401 data
->type
= REPL_TYPE_CHARACTER
;
2406 data
->type
= REPL_TYPE_CHARACTER
;
2416 h
= g_ascii_xdigit_value (*p
);
2419 error_detail
= _("hexadecimal digit or “}” expected");
2430 for (i
= 0; i
< 2; i
++)
2432 h
= g_ascii_xdigit_value (*p
);
2435 error_detail
= _("hexadecimal digit expected");
2442 data
->type
= REPL_TYPE_STRING
;
2443 data
->text
= g_new0 (gchar
, 8);
2444 g_unichar_to_utf8 (x
, data
->text
);
2448 data
->type
= REPL_TYPE_CHANGE_CASE
;
2449 data
->change_case
= CHANGE_CASE_LOWER_SINGLE
;
2453 data
->type
= REPL_TYPE_CHANGE_CASE
;
2454 data
->change_case
= CHANGE_CASE_UPPER_SINGLE
;
2458 data
->type
= REPL_TYPE_CHANGE_CASE
;
2459 data
->change_case
= CHANGE_CASE_LOWER
;
2463 data
->type
= REPL_TYPE_CHANGE_CASE
;
2464 data
->change_case
= CHANGE_CASE_UPPER
;
2468 data
->type
= REPL_TYPE_CHANGE_CASE
;
2469 data
->change_case
= CHANGE_CASE_NONE
;
2475 error_detail
= _("missing “<” in symbolic reference");
2484 error_detail
= _("unfinished symbolic reference");
2491 error_detail
= _("zero-length symbolic reference");
2494 if (g_ascii_isdigit (*q
))
2499 h
= g_ascii_digit_value (*q
);
2502 error_detail
= _("digit expected");
2511 data
->type
= REPL_TYPE_NUMERIC_REFERENCE
;
2518 if (!g_ascii_isalnum (*r
))
2520 error_detail
= _("illegal symbolic reference");
2527 data
->text
= g_strndup (q
, p
- q
);
2528 data
->type
= REPL_TYPE_SYMBOLIC_REFERENCE
;
2533 /* if \0 is followed by a number is an octal number representing a
2534 * character, else it is a numeric reference. */
2535 if (g_ascii_digit_value (*g_utf8_next_char (p
)) >= 0)
2538 p
= g_utf8_next_char (p
);
2551 for (i
= 0; i
< 3; i
++)
2553 h
= g_ascii_digit_value (*p
);
2563 if (i
== 2 && base
== 10)
2569 if (base
== 8 || i
== 3)
2571 data
->type
= REPL_TYPE_STRING
;
2572 data
->text
= g_new0 (gchar
, 8);
2573 g_unichar_to_utf8 (x
, data
->text
);
2577 data
->type
= REPL_TYPE_NUMERIC_REFERENCE
;
2582 error_detail
= _("stray final “\\”");
2586 error_detail
= _("unknown escape sequence");
2593 /* G_GSSIZE_FORMAT doesn't work with gettext, so we use %lu */
2594 tmp_error
= g_error_new (G_REGEX_ERROR
,
2595 G_REGEX_ERROR_REPLACE
,
2596 _("Error while parsing replacement "
2597 "text “%s” at char %lu: %s"),
2599 (gulong
)(p
- replacement
),
2601 g_propagate_error (error
, tmp_error
);
2607 split_replacement (const gchar
*replacement
,
2611 InterpolationData
*data
;
2612 const gchar
*p
, *start
;
2614 start
= p
= replacement
;
2619 data
= g_new0 (InterpolationData
, 1);
2620 start
= p
= expand_escape (replacement
, p
, data
, error
);
2623 g_list_free_full (list
, (GDestroyNotify
) free_interpolation_data
);
2624 free_interpolation_data (data
);
2628 list
= g_list_prepend (list
, data
);
2633 if (*p
== '\\' || *p
== '\0')
2637 data
= g_new0 (InterpolationData
, 1);
2638 data
->text
= g_strndup (start
, p
- start
);
2639 data
->type
= REPL_TYPE_STRING
;
2640 list
= g_list_prepend (list
, data
);
2646 return g_list_reverse (list
);
2649 /* Change the case of c based on change_case. */
2650 #define CHANGE_CASE(c, change_case) \
2651 (((change_case) & CHANGE_CASE_LOWER_MASK) ? \
2652 g_unichar_tolower (c) : \
2653 g_unichar_toupper (c))
2656 string_append (GString
*string
,
2658 ChangeCase
*change_case
)
2662 if (text
[0] == '\0')
2665 if (*change_case
== CHANGE_CASE_NONE
)
2667 g_string_append (string
, text
);
2669 else if (*change_case
& CHANGE_CASE_SINGLE_MASK
)
2671 c
= g_utf8_get_char (text
);
2672 g_string_append_unichar (string
, CHANGE_CASE (c
, *change_case
));
2673 g_string_append (string
, g_utf8_next_char (text
));
2674 *change_case
= CHANGE_CASE_NONE
;
2678 while (*text
!= '\0')
2680 c
= g_utf8_get_char (text
);
2681 g_string_append_unichar (string
, CHANGE_CASE (c
, *change_case
));
2682 text
= g_utf8_next_char (text
);
2688 interpolate_replacement (const GMatchInfo
*match_info
,
2693 InterpolationData
*idata
;
2695 ChangeCase change_case
= CHANGE_CASE_NONE
;
2697 for (list
= data
; list
; list
= list
->next
)
2700 switch (idata
->type
)
2702 case REPL_TYPE_STRING
:
2703 string_append (result
, idata
->text
, &change_case
);
2705 case REPL_TYPE_CHARACTER
:
2706 g_string_append_c (result
, CHANGE_CASE (idata
->c
, change_case
));
2707 if (change_case
& CHANGE_CASE_SINGLE_MASK
)
2708 change_case
= CHANGE_CASE_NONE
;
2710 case REPL_TYPE_NUMERIC_REFERENCE
:
2711 match
= g_match_info_fetch (match_info
, idata
->num
);
2714 string_append (result
, match
, &change_case
);
2718 case REPL_TYPE_SYMBOLIC_REFERENCE
:
2719 match
= g_match_info_fetch_named (match_info
, idata
->text
);
2722 string_append (result
, match
, &change_case
);
2726 case REPL_TYPE_CHANGE_CASE
:
2727 change_case
= idata
->change_case
;
2735 /* whether actual match_info is needed for replacement, i.e.
2736 * whether there are references
2739 interpolation_list_needs_match (GList
*list
)
2741 while (list
!= NULL
)
2743 InterpolationData
*data
= list
->data
;
2745 if (data
->type
== REPL_TYPE_SYMBOLIC_REFERENCE
||
2746 data
->type
== REPL_TYPE_NUMERIC_REFERENCE
)
2759 * @regex: a #GRegex structure
2760 * @string: (array length=string_len): the string to perform matches against
2761 * @string_len: the length of @string, in bytes, or -1 if @string is nul-terminated
2762 * @start_position: starting index of the string to match, in bytes
2763 * @replacement: text to replace each match with
2764 * @match_options: options for the match
2765 * @error: location to store the error occurring, or %NULL to ignore errors
2767 * Replaces all occurrences of the pattern in @regex with the
2768 * replacement text. Backreferences of the form '\number' or
2769 * '\g<number>' in the replacement text are interpolated by the
2770 * number-th captured subexpression of the match, '\g<name>' refers
2771 * to the captured subexpression with the given name. '\0' refers
2772 * to the complete match, but '\0' followed by a number is the octal
2773 * representation of a character. To include a literal '\' in the
2774 * replacement, write '\\\\'.
2776 * There are also escapes that changes the case of the following text:
2778 * - \l: Convert to lower case the next character
2779 * - \u: Convert to upper case the next character
2780 * - \L: Convert to lower case till \E
2781 * - \U: Convert to upper case till \E
2782 * - \E: End case modification
2784 * If you do not need to use backreferences use g_regex_replace_literal().
2786 * The @replacement string must be UTF-8 encoded even if #G_REGEX_RAW was
2787 * passed to g_regex_new(). If you want to use not UTF-8 encoded stings
2788 * you can use g_regex_replace_literal().
2790 * Setting @start_position differs from just passing over a shortened
2791 * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern that
2792 * begins with any kind of lookbehind assertion, such as "\b".
2794 * Returns: a newly allocated string containing the replacements
2799 g_regex_replace (const GRegex
*regex
,
2800 const gchar
*string
,
2802 gint start_position
,
2803 const gchar
*replacement
,
2804 GRegexMatchFlags match_options
,
2809 GError
*tmp_error
= NULL
;
2811 g_return_val_if_fail (regex
!= NULL
, NULL
);
2812 g_return_val_if_fail (string
!= NULL
, NULL
);
2813 g_return_val_if_fail (start_position
>= 0, NULL
);
2814 g_return_val_if_fail (replacement
!= NULL
, NULL
);
2815 g_return_val_if_fail (error
== NULL
|| *error
== NULL
, NULL
);
2816 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, NULL
);
2818 list
= split_replacement (replacement
, &tmp_error
);
2819 if (tmp_error
!= NULL
)
2821 g_propagate_error (error
, tmp_error
);
2825 result
= g_regex_replace_eval (regex
,
2826 string
, string_len
, start_position
,
2828 interpolate_replacement
,
2831 if (tmp_error
!= NULL
)
2832 g_propagate_error (error
, tmp_error
);
2834 g_list_free_full (list
, (GDestroyNotify
) free_interpolation_data
);
2840 literal_replacement (const GMatchInfo
*match_info
,
2844 g_string_append (result
, data
);
2849 * g_regex_replace_literal:
2850 * @regex: a #GRegex structure
2851 * @string: (array length=string_len): the string to perform matches against
2852 * @string_len: the length of @string, in bytes, or -1 if @string is nul-terminated
2853 * @start_position: starting index of the string to match, in bytes
2854 * @replacement: text to replace each match with
2855 * @match_options: options for the match
2856 * @error: location to store the error occurring, or %NULL to ignore errors
2858 * Replaces all occurrences of the pattern in @regex with the
2859 * replacement text. @replacement is replaced literally, to
2860 * include backreferences use g_regex_replace().
2862 * Setting @start_position differs from just passing over a
2863 * shortened string and setting #G_REGEX_MATCH_NOTBOL in the
2864 * case of a pattern that begins with any kind of lookbehind
2865 * assertion, such as "\b".
2867 * Returns: a newly allocated string containing the replacements
2872 g_regex_replace_literal (const GRegex
*regex
,
2873 const gchar
*string
,
2875 gint start_position
,
2876 const gchar
*replacement
,
2877 GRegexMatchFlags match_options
,
2880 g_return_val_if_fail (replacement
!= NULL
, NULL
);
2881 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, NULL
);
2883 return g_regex_replace_eval (regex
,
2884 string
, string_len
, start_position
,
2886 literal_replacement
,
2887 (gpointer
)replacement
,
2892 * g_regex_replace_eval:
2893 * @regex: a #GRegex structure from g_regex_new()
2894 * @string: (array length=string_len): string to perform matches against
2895 * @string_len: the length of @string, in bytes, or -1 if @string is nul-terminated
2896 * @start_position: starting index of the string to match, in bytes
2897 * @match_options: options for the match
2898 * @eval: a function to call for each match
2899 * @user_data: user data to pass to the function
2900 * @error: location to store the error occurring, or %NULL to ignore errors
2902 * Replaces occurrences of the pattern in regex with the output of
2903 * @eval for that occurrence.
2905 * Setting @start_position differs from just passing over a shortened
2906 * string and setting #G_REGEX_MATCH_NOTBOL in the case of a pattern
2907 * that begins with any kind of lookbehind assertion, such as "\b".
2909 * The following example uses g_regex_replace_eval() to replace multiple
2911 * |[<!-- language="C" -->
2913 * eval_cb (const GMatchInfo *info,
2920 * match = g_match_info_fetch (info, 0);
2921 * r = g_hash_table_lookup ((GHashTable *)data, match);
2922 * g_string_append (res, r);
2934 * h = g_hash_table_new (g_str_hash, g_str_equal);
2936 * g_hash_table_insert (h, "1", "ONE");
2937 * g_hash_table_insert (h, "2", "TWO");
2938 * g_hash_table_insert (h, "3", "THREE");
2939 * g_hash_table_insert (h, "4", "FOUR");
2941 * reg = g_regex_new ("1|2|3|4", 0, 0, NULL);
2942 * res = g_regex_replace_eval (reg, text, -1, 0, 0, eval_cb, h, NULL);
2943 * g_hash_table_destroy (h);
2948 * Returns: a newly allocated string containing the replacements
2953 g_regex_replace_eval (const GRegex
*regex
,
2954 const gchar
*string
,
2956 gint start_position
,
2957 GRegexMatchFlags match_options
,
2958 GRegexEvalCallback eval
,
2962 GMatchInfo
*match_info
;
2965 gboolean done
= FALSE
;
2966 GError
*tmp_error
= NULL
;
2968 g_return_val_if_fail (regex
!= NULL
, NULL
);
2969 g_return_val_if_fail (string
!= NULL
, NULL
);
2970 g_return_val_if_fail (start_position
>= 0, NULL
);
2971 g_return_val_if_fail (eval
!= NULL
, NULL
);
2972 g_return_val_if_fail ((match_options
& ~G_REGEX_MATCH_MASK
) == 0, NULL
);
2975 string_len
= strlen (string
);
2977 result
= g_string_sized_new (string_len
);
2979 /* run down the string making matches. */
2980 g_regex_match_full (regex
, string
, string_len
, start_position
,
2981 match_options
, &match_info
, &tmp_error
);
2982 while (!done
&& g_match_info_matches (match_info
))
2984 g_string_append_len (result
,
2986 match_info
->offsets
[0] - str_pos
);
2987 done
= (*eval
) (match_info
, result
, user_data
);
2988 str_pos
= match_info
->offsets
[1];
2989 g_match_info_next (match_info
, &tmp_error
);
2991 g_match_info_free (match_info
);
2992 if (tmp_error
!= NULL
)
2994 g_propagate_error (error
, tmp_error
);
2995 g_string_free (result
, TRUE
);
2999 g_string_append_len (result
, string
+ str_pos
, string_len
- str_pos
);
3000 return g_string_free (result
, FALSE
);
3004 * g_regex_check_replacement:
3005 * @replacement: the replacement string
3006 * @has_references: (out) (optional): location to store information about
3007 * references in @replacement or %NULL
3008 * @error: location to store error
3010 * Checks whether @replacement is a valid replacement string
3011 * (see g_regex_replace()), i.e. that all escape sequences in
3014 * If @has_references is not %NULL then @replacement is checked
3015 * for pattern references. For instance, replacement text 'foo\n'
3016 * does not contain references and may be evaluated without information
3017 * about actual match, but '\0\1' (whole match followed by first
3018 * subpattern) requires valid #GMatchInfo object.
3020 * Returns: whether @replacement is a valid replacement string
3025 g_regex_check_replacement (const gchar
*replacement
,
3026 gboolean
*has_references
,
3032 list
= split_replacement (replacement
, &tmp
);
3036 g_propagate_error (error
, tmp
);
3041 *has_references
= interpolation_list_needs_match (list
);
3043 g_list_free_full (list
, (GDestroyNotify
) free_interpolation_data
);
3049 * g_regex_escape_nul:
3050 * @string: the string to escape
3051 * @length: the length of @string
3053 * Escapes the nul characters in @string to "\x00". It can be used
3054 * to compile a regex with embedded nul characters.
3056 * For completeness, @length can be -1 for a nul-terminated string.
3057 * In this case the output string will be of course equal to @string.
3059 * Returns: a newly-allocated escaped string
3064 g_regex_escape_nul (const gchar
*string
,
3068 const gchar
*p
, *piece_start
, *end
;
3071 g_return_val_if_fail (string
!= NULL
, NULL
);
3074 return g_strdup (string
);
3076 end
= string
+ length
;
3077 p
= piece_start
= string
;
3078 escaped
= g_string_sized_new (length
+ 1);
3086 if (p
!= piece_start
)
3088 /* copy the previous piece. */
3089 g_string_append_len (escaped
, piece_start
, p
- piece_start
);
3091 if ((backslashes
& 1) == 0)
3092 g_string_append_c (escaped
, '\\');
3093 g_string_append_c (escaped
, 'x');
3094 g_string_append_c (escaped
, '0');
3095 g_string_append_c (escaped
, '0');
3105 p
= g_utf8_next_char (p
);
3110 if (piece_start
< end
)
3111 g_string_append_len (escaped
, piece_start
, end
- piece_start
);
3113 return g_string_free (escaped
, FALSE
);
3117 * g_regex_escape_string:
3118 * @string: (array length=length): the string to escape
3119 * @length: the length of @string, in bytes, or -1 if @string is nul-terminated
3121 * Escapes the special characters used for regular expressions
3122 * in @string, for instance "a.b*c" becomes "a\.b\*c". This
3123 * function is useful to dynamically generate regular expressions.
3125 * @string can contain nul characters that are replaced with "\0",
3126 * in this case remember to specify the correct length of @string
3129 * Returns: a newly-allocated escaped string
3134 g_regex_escape_string (const gchar
*string
,
3138 const char *p
, *piece_start
, *end
;
3140 g_return_val_if_fail (string
!= NULL
, NULL
);
3143 length
= strlen (string
);
3145 end
= string
+ length
;
3146 p
= piece_start
= string
;
3147 escaped
= g_string_sized_new (length
+ 1);
3168 if (p
!= piece_start
)
3169 /* copy the previous piece. */
3170 g_string_append_len (escaped
, piece_start
, p
- piece_start
);
3171 g_string_append_c (escaped
, '\\');
3173 g_string_append_c (escaped
, '0');
3175 g_string_append_c (escaped
, *p
);
3179 p
= g_utf8_next_char (p
);
3184 if (piece_start
< end
)
3185 g_string_append_len (escaped
, piece_start
, end
- piece_start
);
3187 return g_string_free (escaped
, FALSE
);