Commit | Line | Data |
---|---|---|
8a118206 RGS |
1 | =head1 NAME |
2 | ||
3 | perlrebackslash - Perl Regular Expression Backslash Sequences and Escapes | |
4 | ||
5 | =head1 DESCRIPTION | |
6 | ||
7 | The top level documentation about Perl regular expressions | |
8 | is found in L<perlre>. | |
9 | ||
10 | This document describes all backslash and escape sequences. After | |
11 | explaining the role of the backslash, it lists all the sequences that have | |
12 | a special meaning in Perl regular expressions (in alphabetical order), | |
13 | then describes each of them. | |
14 | ||
15 | Most sequences are described in detail in different documents; the primary | |
16 | purpose of this document is to have a quick reference guide describing all | |
17 | backslash and escape sequences. | |
18 | ||
8a118206 RGS |
19 | =head2 The backslash |
20 | ||
21 | In a regular expression, the backslash can perform one of two tasks: | |
22 | it either takes away the special meaning of the character following it | |
23 | (for instance, C<\|> matches a vertical bar, it's not an alternation), | |
24 | or it is the start of a backslash or escape sequence. | |
25 | ||
26 | The rules determining what it is are quite simple: if the character | |
df225385 | 27 | following the backslash is an ASCII punctuation (non-word) character (that is, |
b6538e4f TC |
28 | anything that is not a letter, digit, or underscore), then the backslash just |
29 | takes away any special meaning of the character following it. | |
df225385 KW |
30 | |
31 | If the character following the backslash is an ASCII letter or an ASCII digit, | |
32 | then the sequence may be special; if so, it's listed below. A few letters have | |
6b46370c KW |
33 | not been used yet, so escaping them with a backslash doesn't change them to be |
34 | special. A future version of Perl may assign a special meaning to them, so if | |
b6538e4f | 35 | you have warnings turned on, Perl issues a warning if you use such a |
6b46370c | 36 | sequence. [1]. |
8a118206 | 37 | |
e2cb52ee | 38 | It is however guaranteed that backslash or escape sequences never have a |
8a118206 RGS |
39 | punctuation character following the backslash, not now, and not in a future |
40 | version of Perl 5. So it is safe to put a backslash in front of a non-word | |
41 | character. | |
42 | ||
43 | Note that the backslash itself is special; if you want to match a backslash, | |
44 | you have to escape the backslash with a backslash: C</\\/> matches a single | |
45 | backslash. | |
46 | ||
47 | =over 4 | |
48 | ||
49 | =item [1] | |
50 | ||
b6538e4f | 51 | There is one exception. If you use an alphanumeric character as the |
8a118206 | 52 | delimiter of your pattern (which you probably shouldn't do for readability |
b6538e4f | 53 | reasons), you have to escape the delimiter if you want to match |
8a118206 RGS |
54 | it. Perl won't warn then. See also L<perlop/Gory details of parsing |
55 | quoted constructs>. | |
56 | ||
57 | =back | |
58 | ||
59 | ||
60 | =head2 All the sequences and escapes | |
61 | ||
df225385 KW |
62 | Those not usable within a bracketed character class (like C<[\da-z]>) are marked |
63 | as C<Not in [].> | |
64 | ||
f0a2b745 | 65 | \000 Octal escape sequence. See also \o{}. |
df225385 | 66 | \1 Absolute backreference. Not in []. |
8a118206 | 67 | \a Alarm or bell. |
df225385 | 68 | \A Beginning of string. Not in []. |
64935bc6 | 69 | \b{}, \b Boundary. (\b is a backspace in []). |
bc1d9728 | 70 | \B{}, \B Not a boundary. Not in []. |
f321be7e | 71 | \cX Control-X. |
05a0cace KW |
72 | \d Match any digit character. |
73 | \D Match any character that isn't a digit. | |
8a118206 | 74 | \e Escape character. |
df225385 | 75 | \E Turn off \Q, \L and \U processing. Not in []. |
8a118206 | 76 | \f Form feed. |
628253b8 | 77 | \F Foldcase till \E. Not in []. |
f321be7e SK |
78 | \g{}, \g1 Named, absolute or relative backreference. |
79 | Not in []. | |
df225385 | 80 | \G Pos assertion. Not in []. |
05a0cace KW |
81 | \h Match any horizontal whitespace character. |
82 | \H Match any character that isn't horizontal whitespace. | |
df225385 KW |
83 | \k{}, \k<>, \k'' Named backreference. Not in []. |
84 | \K Keep the stuff left of \K. Not in []. | |
85 | \l Lowercase next character. Not in []. | |
86 | \L Lowercase till \E. Not in []. | |
8a118206 | 87 | \n (Logical) newline character. |
05a0cace | 88 | \N Match any character but newline. Not in []. |
fb121860 | 89 | \N{} Named or numbered (Unicode) character or sequence. |
f0a2b745 | 90 | \o{} Octal escape sequence. |
05a0cace KW |
91 | \p{}, \pP Match any character with the given Unicode property. |
92 | \P{}, \PP Match any character without the given property. | |
736fe711 KW |
93 | \Q Quote (disable) pattern metacharacters till \E. Not |
94 | in []. | |
8a118206 | 95 | \r Return character. |
df225385 | 96 | \R Generic new line. Not in []. |
05a0cace KW |
97 | \s Match any whitespace character. |
98 | \S Match any character that isn't a whitespace. | |
8a118206 | 99 | \t Tab character. |
df225385 KW |
100 | \u Titlecase next character. Not in []. |
101 | \U Uppercase till \E. Not in []. | |
05a0cace KW |
102 | \v Match any vertical whitespace character. |
103 | \V Match any character that isn't vertical whitespace | |
104 | \w Match any word character. | |
105 | \W Match any character that isn't a word character. | |
8a118206 | 106 | \x{}, \x00 Hexadecimal escape sequence. |
df225385 KW |
107 | \X Unicode "extended grapheme cluster". Not in []. |
108 | \z End of string. Not in []. | |
109 | \Z End of string. Not in []. | |
8a118206 RGS |
110 | |
111 | =head2 Character Escapes | |
112 | ||
113 | =head3 Fixed characters | |
114 | ||
e2cb52ee | 115 | A handful of characters have a dedicated I<character escape>. The following |
58151fe4 | 116 | table shows them, along with their ASCII code points (in decimal and hex), |
4948b50f KW |
117 | their ASCII name, the control escape on ASCII platforms and a short |
118 | description. (For EBCDIC platforms, see L<perlebcdic/OPERATOR DIFFERENCES>.) | |
8a118206 | 119 | |
4948b50f | 120 | Seq. Code Point ASCII Cntrl Description. |
8a118206 RGS |
121 | Dec Hex |
122 | \a 7 07 BEL \cG alarm or bell | |
123 | \b 8 08 BS \cH backspace [1] | |
124 | \e 27 1B ESC \c[ escape character | |
125 | \f 12 0C FF \cL form feed | |
126 | \n 10 0A LF \cJ line feed [2] | |
127 | \r 13 0D CR \cM carriage return | |
128 | \t 9 09 TAB \cI tab | |
129 | ||
130 | =over 4 | |
131 | ||
132 | =item [1] | |
133 | ||
301ba1af | 134 | C<\b> is the backspace character only inside a character class. Outside a |
64935bc6 KW |
135 | character class, C<\b> alone is a word-character/non-word-character |
136 | boundary, and C<\b{}> is some other type of boundary. | |
8a118206 RGS |
137 | |
138 | =item [2] | |
139 | ||
b6538e4f | 140 | C<\n> matches a logical newline. Perl converts between C<\n> and your |
f6993e9e | 141 | OS's native newline character when reading from or writing to text files. |
8a118206 RGS |
142 | |
143 | =back | |
144 | ||
145 | =head4 Example | |
146 | ||
147 | $str =~ /\t/; # Matches if $str contains a (horizontal) tab. | |
148 | ||
149 | =head3 Control characters | |
150 | ||
151 | C<\c> is used to denote a control character; the character following C<\c> | |
4948b50f KW |
152 | determines the value of the construct. For example the value of C<\cA> is |
153 | C<chr(1)>, and the value of C<\cb> is C<chr(2)>, etc. | |
154 | The gory details are in L<perlop/"Regexp Quote-Like Operators">. A complete | |
155 | list of what C<chr(1)>, etc. means for ASCII and EBCDIC platforms is in | |
156 | L<perlebcdic/OPERATOR DIFFERENCES>. | |
157 | ||
158 | Note that C<\c\> alone at the end of a regular expression (or doubled-quoted | |
159 | string) is not valid. The backslash must be followed by another character. | |
160 | That is, C<\c\I<X>> means C<chr(28) . 'I<X>'> for all characters I<X>. | |
161 | ||
162 | To write platform-independent code, you must use C<\N{I<NAME>}> instead, like | |
163 | C<\N{ESCAPE}> or C<\N{U+001B}>, see L<charnames>. | |
8a118206 RGS |
164 | |
165 | Mnemonic: I<c>ontrol character. | |
166 | ||
167 | =head4 Example | |
168 | ||
169 | $str =~ /\cK/; # Matches if $str contains a vertical tab (control-K). | |
170 | ||
fb121860 | 171 | =head3 Named or numbered characters and character sequences |
8a118206 | 172 | |
17148a1a KW |
173 | Unicode characters have a Unicode name and numeric code point (ordinal) |
174 | value. Use the | |
e526e8bb | 175 | C<\N{}> construct to specify a character by either of these values. |
fb121860 | 176 | Certain sequences of characters also have names. |
e526e8bb | 177 | |
fb121860 | 178 | To specify by name, the name of the character or character sequence goes |
fbb93542 | 179 | between the curly braces. |
e526e8bb | 180 | |
b6538e4f TC |
181 | To specify a character by Unicode code point, use the form C<\N{U+I<code |
182 | point>}>, where I<code point> is a number in hexadecimal that gives the | |
17148a1a | 183 | code point that Unicode has assigned to the desired character. It is |
b6538e4f TC |
184 | customary but not required to use leading zeros to pad the number to 4 |
185 | digits. Thus C<\N{U+0041}> means C<LATIN CAPITAL LETTER A>, and you will | |
186 | rarely see it written without the two leading zeros. C<\N{U+0041}> means | |
187 | "A" even on EBCDIC machines (where the ordinal value of "A" is not 0x41). | |
e526e8bb | 188 | |
1b2f32d5 KW |
189 | Blanks may freely be inserted adjacent to but within the braces |
190 | enclosing the name or code point. So S<C<\N{ U+0041 }>> is perfectly | |
191 | legal. | |
192 | ||
fb121860 | 193 | It is even possible to give your own names to characters and character |
945961fd KW |
194 | sequences by using the L<charnames> module. These custom names are |
195 | lexically scoped, and so a given code point may have different names | |
196 | in different scopes. The name used is what is in effect at the time the | |
197 | C<\N{}> is expanded. For patterns in double-quotish context, that means | |
198 | at the time the pattern is parsed. But for patterns that are delimitted | |
199 | by single quotes, the expansion is deferred until pattern compilation | |
200 | time, which may very well have a different C<charnames> translator in | |
201 | effect. | |
8a118206 | 202 | |
8c37f1d0 | 203 | (There is an expanded internal form that you may see in debug output: |
b6538e4f TC |
204 | C<\N{U+I<code point>.I<code point>...}>. |
205 | The C<...> means any number of these I<code point>s separated by dots. | |
8c37f1d0 KW |
206 | This represents the sequence formed by the characters. This is an internal |
207 | form only, subject to change, and you should not try to use it yourself.) | |
208 | ||
8a118206 RGS |
209 | Mnemonic: I<N>amed character. |
210 | ||
b6538e4f TC |
211 | Note that a character or character sequence expressed as a named |
212 | or numbered character is considered a character without special | |
fb121860 | 213 | meaning by the regex engine, and will match "as is". |
df225385 | 214 | |
8a118206 RGS |
215 | =head4 Example |
216 | ||
8a118206 RGS |
217 | $str =~ /\N{THAI CHARACTER SO SO}/; # Matches the Thai SO SO character |
218 | ||
219 | use charnames 'Cyrillic'; # Loads Cyrillic names. | |
220 | $str =~ /\N{ZHE}\N{KA}/; # Match "ZHE" followed by "KA". | |
221 | ||
222 | =head3 Octal escapes | |
223 | ||
f0a2b745 | 224 | There are two forms of octal escapes. Each is used to specify a character by |
012ac233 | 225 | its code point specified in base 8. |
f0a2b745 KW |
226 | |
227 | One form, available starting in Perl 5.14 looks like C<\o{...}>, where the dots | |
228 | represent one or more octal digits. It can be used for any Unicode character. | |
229 | ||
230 | It was introduced to avoid the potential problems with the other form, | |
231 | available in all Perls. That form consists of a backslash followed by three | |
232 | octal digits. One problem with this form is that it can look exactly like an | |
233 | old-style backreference (see | |
234 | L</Disambiguation rules between old-style octal escapes and backreferences> | |
235 | below.) You can avoid this by making the first of the three digits always a | |
9645299c | 236 | zero, but that makes \077 the largest code point specifiable. |
f0a2b745 KW |
237 | |
238 | In some contexts, a backslash followed by two or even one octal digits may be | |
239 | interpreted as an octal escape, sometimes with a warning, and because of some | |
240 | bugs, sometimes with surprising results. Also, if you are creating a regex | |
c69ca1d4 | 241 | out of smaller snippets concatenated together, and you use fewer than three |
f0a2b745 KW |
242 | digits, the beginning of one snippet may be interpreted as adding digits to the |
243 | ending of the snippet before it. See L</Absolute referencing> for more | |
244 | discussion and examples of the snippet problem. | |
8a118206 | 245 | |
b6538e4f TC |
246 | Note that a character expressed as an octal escape is considered |
247 | a character without special meaning by the regex engine, and will match | |
8a118206 RGS |
248 | "as is". |
249 | ||
f0a2b745 | 250 | To summarize, the C<\o{}> form is always safe to use, and the other form is |
17148a1a | 251 | safe to use for code points through \077 when you use exactly three digits to |
f0a2b745 | 252 | specify them. |
8a118206 | 253 | |
f0a2b745 | 254 | Mnemonic: I<0>ctal or I<o>ctal. |
8a118206 | 255 | |
f0a2b745 | 256 | =head4 Examples (assuming an ASCII platform) |
8a118206 | 257 | |
f0a2b745 KW |
258 | $str = "Perl"; |
259 | $str =~ /\o{120}/; # Match, "\120" is "P". | |
260 | $str =~ /\120/; # Same. | |
f321be7e SK |
261 | $str =~ /\o{120}+/; # Match, "\120" is "P", |
262 | # it's repeated at least once. | |
f0a2b745 KW |
263 | $str =~ /\120+/; # Same. |
264 | $str =~ /P\053/; # No match, "\053" is "+" and taken literally. | |
265 | /\o{23073}/ # Black foreground, white background smiling face. | |
f321be7e | 266 | /\o{4801234567}/ # Raises a warning, and yields chr(4). |
1b2f32d5 KW |
267 | /\o{ 400}/ # LATIN CAPITAL LETTER A WITH MACRON |
268 | /\o{ 400 }/ # Same. These show blanks are allowed adjacent to | |
269 | # the braces | |
f0a2b745 KW |
270 | |
271 | =head4 Disambiguation rules between old-style octal escapes and backreferences | |
272 | ||
273 | Octal escapes of the C<\000> form outside of bracketed character classes | |
f321be7e | 274 | potentially clash with old-style backreferences (see L</Absolute referencing> |
f0a2b745 KW |
275 | below). They both consist of a backslash followed by numbers. So Perl has to |
276 | use heuristics to determine whether it is a backreference or an octal escape. | |
277 | Perl uses the following rules to disambiguate: | |
8a118206 RGS |
278 | |
279 | =over 4 | |
280 | ||
281 | =item 1 | |
282 | ||
353c6505 | 283 | If the backslash is followed by a single digit, it's a backreference. |
8a118206 RGS |
284 | |
285 | =item 2 | |
286 | ||
287 | If the first digit following the backslash is a 0, it's an octal escape. | |
288 | ||
289 | =item 3 | |
290 | ||
b6538e4f TC |
291 | If the number following the backslash is N (in decimal), and Perl already |
292 | has seen N capture groups, Perl considers this a backreference. Otherwise, | |
293 | it considers it an octal escape. If N has more than three digits, Perl | |
294 | takes only the first three for the octal escape; the rest are matched as is. | |
8a118206 RGS |
295 | |
296 | my $pat = "(" x 999; | |
297 | $pat .= "a"; | |
298 | $pat .= ")" x 999; | |
299 | /^($pat)\1000$/; # Matches 'aa'; there are 1000 capture groups. | |
300 | /^$pat\1000$/; # Matches 'a@0'; there are 999 capture groups | |
f321be7e | 301 | # and \1000 is seen as \100 (a '@') and a '0'. |
8a118206 RGS |
302 | |
303 | =back | |
304 | ||
17148a1a | 305 | You can force a backreference interpretation always by using the C<\g{...}> |
f0a2b745 KW |
306 | form. You can the force an octal interpretation always by using the C<\o{...}> |
307 | form, or for numbers up through \077 (= 63 decimal), by using three digits, | |
308 | beginning with a "0". | |
309 | ||
8a118206 RGS |
310 | =head3 Hexadecimal escapes |
311 | ||
f0a2b745 | 312 | Like octal escapes, there are two forms of hexadecimal escapes, but both start |
febd1aee | 313 | with the sequence C<\x>. This is followed by either exactly two hexadecimal |
f0a2b745 KW |
314 | digits forming a number, or a hexadecimal number of arbitrary length surrounded |
315 | by curly braces. The hexadecimal number is the code point of the character you | |
316 | want to express. | |
8a118206 | 317 | |
b6538e4f TC |
318 | Note that a character expressed as one of these escapes is considered a |
319 | character without special meaning by the regex engine, and will match | |
8a118206 RGS |
320 | "as is". |
321 | ||
322 | Mnemonic: heI<x>adecimal. | |
323 | ||
9f5650a8 | 324 | =head4 Examples (assuming an ASCII platform) |
8a118206 RGS |
325 | |
326 | $str = "Perl"; | |
327 | $str =~ /\x50/; # Match, "\x50" is "P". | |
f822d0dd | 328 | $str =~ /\x50+/; # Match, "\x50" is "P", it is repeated at least once |
8a118206 RGS |
329 | $str =~ /P\x2B/; # No match, "\x2B" is "+" and taken literally. |
330 | ||
331 | /\x{2603}\x{2602}/ # Snowman with an umbrella. | |
332 | # The Unicode character 2603 is a snowman, | |
333 | # the Unicode character 2602 is an umbrella. | |
334 | /\x{263B}/ # Black smiling face. | |
335 | /\x{263b}/ # Same, the hex digits A - F are case insensitive. | |
1b2f32d5 KW |
336 | /\x{ 263b }/ # Same, showing optional blanks adjacent to the |
337 | # braces | |
8a118206 RGS |
338 | |
339 | =head2 Modifiers | |
340 | ||
341 | A number of backslash sequences have to do with changing the character, | |
342 | or characters following them. C<\l> will lowercase the character following | |
5f2b17ca | 343 | it, while C<\u> will uppercase (or, more accurately, titlecase) the |
b6538e4f TC |
344 | character following it. They provide functionality similar to the |
345 | functions C<lcfirst> and C<ucfirst>. | |
8a118206 RGS |
346 | |
347 | To uppercase or lowercase several characters, one might want to use | |
348 | C<\L> or C<\U>, which will lowercase/uppercase all characters following | |
b6538e4f | 349 | them, until either the end of the pattern or the next occurrence of |
17148a1a | 350 | C<\E>, whichever comes first. They provide functionality similar to what |
b6538e4f | 351 | the functions C<lc> and C<uc> provide. |
8a118206 | 352 | |
736fe711 KW |
353 | C<\Q> is used to quote (disable) pattern metacharacters, up to the next |
354 | C<\E> or the end of the pattern. C<\Q> adds a backslash to any character | |
355 | that could have special meaning to Perl. In the ASCII range, it quotes | |
356 | every character that isn't a letter, digit, or underscore. See | |
357 | L<perlfunc/quotemeta> for details on what gets quoted for non-ASCII | |
358 | code points. Using this ensures that any character between C<\Q> and | |
359 | C<\E> will be matched literally, not interpreted as a metacharacter by | |
360 | the regex engine. | |
8a118206 | 361 | |
628253b8 BF |
362 | C<\F> can be used to casefold all characters following, up to the next C<\E> |
363 | or the end of the pattern. It provides the functionality similar to | |
364 | the C<fc> function. | |
365 | ||
366 | Mnemonic: I<L>owercase, I<U>ppercase, I<F>old-case, I<Q>uotemeta, I<E>nd. | |
8a118206 RGS |
367 | |
368 | =head4 Examples | |
369 | ||
370 | $sid = "sid"; | |
371 | $greg = "GrEg"; | |
372 | $miranda = "(Miranda)"; | |
373 | $str =~ /\u$sid/; # Matches 'Sid' | |
374 | $str =~ /\L$greg/; # Matches 'greg' | |
375 | $str =~ /\Q$miranda\E/; # Matches '(Miranda)', as if the pattern | |
376 | # had been written as /\(Miranda\)/ | |
377 | ||
378 | =head2 Character classes | |
379 | ||
380 | Perl regular expressions have a large range of character classes. Some of | |
381 | the character classes are written as a backslash sequence. We will briefly | |
382 | discuss those here; full details of character classes can be found in | |
383 | L<perlrecharclass>. | |
384 | ||
d35dd6c6 KW |
385 | C<\w> is a character class that matches any single I<word> character |
386 | (letters, digits, Unicode marks, and connector punctuation (like the | |
387 | underscore)). C<\d> is a character class that matches any decimal | |
388 | digit, while the character class C<\s> matches any whitespace character. | |
99d59c4d | 389 | New in perl 5.10.0 are the classes C<\h> and C<\v> which match horizontal |
418e7b04 | 390 | and vertical whitespace characters. |
cfaf538b KW |
391 | |
392 | The exact set of characters matched by C<\d>, C<\s>, and C<\w> varies | |
9645299c KW |
393 | depending on various pragma and regular expression modifiers. It is |
394 | possible to restrict the match to the ASCII range by using the C</a> | |
395 | regular expression modifier. See L<perlrecharclass>. | |
8a118206 RGS |
396 | |
397 | The uppercase variants (C<\W>, C<\D>, C<\S>, C<\H>, and C<\V>) are | |
e486b3cc KW |
398 | character classes that match, respectively, any character that isn't a |
399 | word character, digit, whitespace, horizontal whitespace, or vertical | |
400 | whitespace. | |
8a118206 RGS |
401 | |
402 | Mnemonics: I<w>ord, I<d>igit, I<s>pace, I<h>orizontal, I<v>ertical. | |
403 | ||
404 | =head3 Unicode classes | |
405 | ||
406 | C<\pP> (where C<P> is a single letter) and C<\p{Property}> are used to | |
407 | match a character that matches the given Unicode property; properties | |
408 | include things like "letter", or "thai character". Capitalizing the | |
409 | sequence to C<\PP> and C<\P{Property}> make the sequence match a character | |
410 | that doesn't match the given Unicode property. For more details, see | |
4948b50f | 411 | L<perlrecharclass/Backslash sequences> and |
8a118206 RGS |
412 | L<perlunicode/Unicode Character Properties>. |
413 | ||
414 | Mnemonic: I<p>roperty. | |
415 | ||
8a118206 RGS |
416 | =head2 Referencing |
417 | ||
418 | If capturing parenthesis are used in a regular expression, we can refer | |
419 | to the part of the source string that was matched, and match exactly the | |
1843fd28 RGS |
420 | same thing. There are three ways of referring to such I<backreference>: |
421 | absolutely, relatively, and by name. | |
422 | ||
423 | =for later add link to perlrecapture | |
8a118206 RGS |
424 | |
425 | =head3 Absolute referencing | |
426 | ||
c27a5cfe | 427 | Either C<\gI<N>> (starting in Perl 5.10.0), or C<\I<N>> (old-style) where I<N> |
d8b950dc | 428 | is a positive (unsigned) decimal number of any length is an absolute reference |
c27a5cfe KW |
429 | to a capturing group. |
430 | ||
8e4698ef KW |
431 | I<N> refers to the Nth set of parentheses, so C<\gI<N>> refers to whatever has |
432 | been matched by that set of parentheses. Thus C<\g1> refers to the first | |
c27a5cfe KW |
433 | capture group in the regex. |
434 | ||
435 | The C<\gI<N>> form can be equivalently written as C<\g{I<N>}> | |
436 | which avoids ambiguity when building a regex by concatenating shorter | |
e13bc2af DIM |
437 | strings. Otherwise if you had a regex C<qr/$x$y/>, and C<$x> contained |
438 | C<"\g1">, and C<$y> contained C<"37">, you would get C</\g137/> which is | |
d8b950dc | 439 | probably not what you intended. |
c27a5cfe KW |
440 | |
441 | In the C<\I<N>> form, I<N> must not begin with a "0", and there must be at | |
b6538e4f TC |
442 | least I<N> capturing groups, or else I<N> is considered an octal escape |
443 | (but something like C<\18> is the same as C<\0018>; that is, the octal escape | |
c27a5cfe KW |
444 | C<"\001"> followed by a literal digit C<"8">). |
445 | ||
446 | Mnemonic: I<g>roup. | |
8a118206 RGS |
447 | |
448 | =head4 Examples | |
449 | ||
c27a5cfe | 450 | /(\w+) \g1/; # Finds a duplicated word, (e.g. "cat cat"). |
f321be7e | 451 | /(\w+) \1/; # Same thing; written old-style. |
012ac233 | 452 | /(\w+) \g{1}/; # Same, using the safer braced notation |
1b2f32d5 | 453 | /(\w+) \g{ 1 }/;# Same, showing optional blanks adjacent to the braces |
012ac233 | 454 | /(.)(.)\g2\g1/; # Match a four letter palindrome (e.g. "ABBA"). |
8a118206 RGS |
455 | |
456 | ||
457 | =head3 Relative referencing | |
458 | ||
c27a5cfe | 459 | C<\g-I<N>> (starting in Perl 5.10.0) is used for relative addressing. (It can |
7ea7c4bb | 460 | be written as C<\g{-I<N>}>.) It refers to the I<N>th group before the |
c27a5cfe | 461 | C<\g{-I<N>}>. |
8a118206 | 462 | |
c27a5cfe | 463 | The big advantage of this form is that it makes it much easier to write |
8a118206 RGS |
464 | patterns with references that can be interpolated in larger patterns, |
465 | even if the larger pattern also contains capture groups. | |
466 | ||
8a118206 RGS |
467 | =head4 Examples |
468 | ||
c27a5cfe KW |
469 | /(A) # Group 1 |
470 | ( # Group 2 | |
471 | (B) # Group 3 | |
472 | \g{-1} # Refers to group 3 (B) | |
473 | \g{-3} # Refers to group 1 (A) | |
1b2f32d5 | 474 | \g{ -3 } # Same, showing optional blanks adjacent to the braces |
8a118206 RGS |
475 | ) |
476 | /x; # Matches "ABBA". | |
477 | ||
478 | my $qr = qr /(.)(.)\g{-2}\g{-1}/; # Matches 'abab', 'cdcd', etc. | |
479 | /$qr$qr/ # Matches 'ababcdcd'. | |
480 | ||
481 | =head3 Named referencing | |
482 | ||
d8b950dc KW |
483 | C<\g{I<name>}> (starting in Perl 5.10.0) can be used to back refer to a |
484 | named capture group, dispensing completely with having to think about capture | |
485 | buffer positions. | |
8a118206 RGS |
486 | |
487 | To be compatible with .Net regular expressions, C<\g{name}> may also be | |
488 | written as C<\k{name}>, C<< \k<name> >> or C<\k'name'>. | |
489 | ||
d8b950dc KW |
490 | To prevent any ambiguity, I<name> must not start with a digit nor contain a |
491 | hyphen. | |
8a118206 RGS |
492 | |
493 | =head4 Examples | |
494 | ||
012ac233 KW |
495 | /(?<word>\w+) \g{word}/ # Finds duplicated word, (e.g. "cat cat") |
496 | /(?<word>\w+) \k{word}/ # Same. | |
1b2f32d5 KW |
497 | /(?<word>\w+) \g{ word }/ # Same, showing optional blanks adjacent to |
498 | # the braces | |
499 | /(?<word>\w+) \k{ word }/ # Same. | |
500 | /(?<word>\w+) \k<word>/ # Same. There are no braces, so no blanks | |
501 | # are permitted | |
8a118206 | 502 | /(?<letter1>.)(?<letter2>.)\g{letter2}\g{letter1}/ |
012ac233 KW |
503 | # Match a four letter palindrome (e.g. |
504 | # "ABBA") | |
8a118206 RGS |
505 | |
506 | =head2 Assertions | |
507 | ||
ac036724 | 508 | Assertions are conditions that have to be true; they don't actually |
8a118206 RGS |
509 | match parts of the substring. There are six assertions that are written as |
510 | backslash sequences. | |
511 | ||
512 | =over 4 | |
513 | ||
514 | =item \A | |
515 | ||
516 | C<\A> only matches at the beginning of the string. If the C</m> modifier | |
1726f7e8 | 517 | isn't used, then C</\A/> is equivalent to C</^/>. However, if the C</m> |
8a118206 RGS |
518 | modifier is used, then C</^/> matches internal newlines, but the meaning |
519 | of C</\A/> isn't changed by the C</m> modifier. C<\A> matches at the beginning | |
520 | of the string regardless whether the C</m> modifier is used. | |
521 | ||
522 | =item \z, \Z | |
523 | ||
524 | C<\z> and C<\Z> match at the end of the string. If the C</m> modifier isn't | |
b6538e4f TC |
525 | used, then C</\Z/> is equivalent to C</$/>; that is, it matches at the |
526 | end of the string, or one before the newline at the end of the string. If the | |
8a118206 RGS |
527 | C</m> modifier is used, then C</$/> matches at internal newlines, but the |
528 | meaning of C</\Z/> isn't changed by the C</m> modifier. C<\Z> matches at | |
529 | the end of the string (or just before a trailing newline) regardless whether | |
530 | the C</m> modifier is used. | |
531 | ||
b6538e4f TC |
532 | C<\z> is just like C<\Z>, except that it does not match before a trailing |
533 | newline. C<\z> matches at the end of the string only, regardless of the | |
534 | modifiers used, and not just before a newline. It is how to anchor the | |
535 | match to the true end of the string under all conditions. | |
8a118206 RGS |
536 | |
537 | =item \G | |
538 | ||
b6538e4f | 539 | C<\G> is usually used only in combination with the C</g> modifier. If the |
30659cfd | 540 | C</g> modifier is used and the match is done in scalar context, Perl |
b6538e4f | 541 | remembers where in the source string the last match ended, and the next time, |
8a118206 RGS |
542 | it will start the match from where it ended the previous time. |
543 | ||
30659cfd | 544 | C<\G> matches the point where the previous match on that string ended, |
b6538e4f | 545 | or the beginning of that string if there was no previous match. |
1843fd28 RGS |
546 | |
547 | =for later add link to perlremodifiers | |
8a118206 RGS |
548 | |
549 | Mnemonic: I<G>lobal. | |
550 | ||
64935bc6 | 551 | =item \b{}, \b, \B{}, \B |
8a118206 | 552 | |
64935bc6 KW |
553 | C<\b{...}>, available starting in v5.22, matches a boundary (between two |
554 | characters, or before the first character of the string, or after the | |
555 | final character of the string) based on the Unicode rules for the | |
6b659339 | 556 | boundary type specified inside the braces. The boundary |
64935bc6 KW |
557 | types are given a few paragraphs below. C<\B{...}> matches at any place |
558 | between characters where C<\b{...}> of the same type doesn't match. | |
559 | ||
012ac233 KW |
560 | C<\b> when not immediately followed by a C<"{"> is available in all |
561 | Perls. It matches at any place | |
64935bc6 KW |
562 | between a word (something matched by C<\w>) and a non-word character |
563 | (C<\W>); C<\B> when not immediately followed by a C<"{"> matches at any | |
ae3bb8ea | 564 | place between characters where C<\b> doesn't match. To get better |
a95b3d6a | 565 | word matching of natural language text, see L</\b{wb}> below. |
64935bc6 KW |
566 | |
567 | C<\b> | |
8a118206 RGS |
568 | and C<\B> assume there's a non-word character before the beginning and after |
569 | the end of the source string; so C<\b> will match at the beginning (or end) | |
570 | of the source string if the source string begins (or ends) with a word | |
30659cfd | 571 | character. Otherwise, C<\B> will match. |
b6538e4f TC |
572 | |
573 | Do not use something like C<\b=head\d\b> and expect it to match the | |
574 | beginning of a line. It can't, because for there to be a boundary before | |
30659cfd | 575 | the non-word "=", there must be a word character immediately previous. |
64935bc6 KW |
576 | All plain C<\b> and C<\B> boundary determinations look for word |
577 | characters alone, not for | |
578 | non-word characters nor for string ends. It may help to understand how | |
6b659339 | 579 | C<\b> and C<\B> work by equating them as follows: |
b6538e4f TC |
580 | |
581 | \b really means (?:(?<=\w)(?!\w)|(?<!\w)(?=\w)) | |
582 | \B really means (?:(?<=\w)(?=\w)|(?<!\w)(?!\w)) | |
8a118206 | 583 | |
54bdcd8e KW |
584 | In contrast, C<\b{...}> and C<\B{...}> may or may not match at the |
585 | beginning and end of the line, depending on the boundary type. These | |
586 | implement the Unicode default boundaries, specified in | |
30659cfd MM |
587 | L<https://www.unicode.org/reports/tr14/> and |
588 | L<https://www.unicode.org/reports/tr29/>. | |
6b659339 | 589 | The boundary types are: |
ae3bb8ea KW |
590 | |
591 | =over | |
592 | ||
593 | =item C<\b{gcb}> or C<\b{g}> | |
594 | ||
595 | This matches a Unicode "Grapheme Cluster Boundary". (Actually Perl | |
596 | always uses the improved "extended" grapheme cluster"). These are | |
eb992c6f | 597 | explained below under C<L</\X>>. In fact, C<\X> is another way to get |
ae3bb8ea KW |
598 | the same functionality. It is equivalent to C</.+?\b{gcb}/>. Use |
599 | whichever is most convenient for your situation. | |
600 | ||
6b659339 KW |
601 | =item C<\b{lb}> |
602 | ||
603 | This matches according to the default Unicode Line Breaking Algorithm | |
30659cfd | 604 | (L<https://www.unicode.org/reports/tr14/>), as customized in that |
6b659339 | 605 | document |
30659cfd | 606 | (L<Example 7 of revision 35|https://www.unicode.org/reports/tr14/tr14-35.html#Example7>) |
6b659339 KW |
607 | for better handling of numeric expressions. |
608 | ||
609 | This is suitable for many purposes, but the L<Unicode::LineBreak> module | |
610 | is available on CPAN that provides many more features, including | |
611 | customization. | |
612 | ||
06ae2722 KW |
613 | =item C<\b{sb}> |
614 | ||
615 | This matches a Unicode "Sentence Boundary". This is an aid to parsing | |
616 | natural language sentences. It gives good, but imperfect results. For | |
617 | example, it thinks that "Mr. Smith" is two sentences. More details are | |
30659cfd | 618 | at L<https://www.unicode.org/reports/tr29/>. Note also that it thinks |
d90f6819 | 619 | that anything matching L</\R> (except form feed and vertical tab) is a |
54bdcd8e KW |
620 | sentence boundary. C<\b{sb}> works with text designed for |
621 | word-processors which wrap lines | |
d90f6819 KW |
622 | automatically for display, but hard-coded line boundaries are considered |
623 | to be essentially the ends of text blocks (paragraphs really), and hence | |
dabde021 | 624 | the ends of sentences. C<\b{sb}> doesn't do well with text containing |
54bdcd8e KW |
625 | embedded newlines, like the source text of the document you are reading. |
626 | Such text needs to be preprocessed to get rid of the line separators | |
627 | before looking for sentence boundaries. Some people view this as a bug | |
268e6905 KW |
628 | in the Unicode standard, and this behavior is quite subject to change in |
629 | future Perl versions. | |
06ae2722 | 630 | |
ae3bb8ea KW |
631 | =item C<\b{wb}> |
632 | ||
f1f6961f KW |
633 | This matches a Unicode "Word Boundary", but tailored to Perl |
634 | expectations. This gives better (though not | |
ae3bb8ea KW |
635 | perfect) results for natural language processing than plain C<\b> |
636 | (without braces) does. For example, it understands that apostrophes can | |
54bdcd8e | 637 | be in the middle of words and that parentheses aren't (see the examples |
30659cfd | 638 | below). More details are at L<https://www.unicode.org/reports/tr29/>. |
ae3bb8ea | 639 | |
f1f6961f KW |
640 | The current Unicode definition of a Word Boundary matches between every |
641 | white space character. Perl tailors this, starting in version 5.24, to | |
642 | generally not break up spans of white space, just as plain C<\b> has | |
643 | always functioned. This allows C<\b{wb}> to be a drop-in replacement for | |
644 | C<\b>, but with generally better results for natural language | |
645 | processing. (The exception to this tailoring is when a span of white | |
646 | space is immediately followed by something like U+0303, COMBINING TILDE. | |
647 | If the final space character in the span is a horizontal white space, it | |
648 | is broken out so that it attaches instead to the combining character. | |
649 | To be precise, if a span of white space that ends in a horizontal space | |
be8bb1ac | 650 | has the character immediately following it have any of the Word |
b0e24409 | 651 | Boundary property values "Extend", "Format" or "ZWJ", the boundary between the |
f1f6961f KW |
652 | final horizontal space character and the rest of the span matches |
653 | C<\b{wb}>. In all other cases the boundary between two white space | |
654 | characters matches C<\B{wb}>.) | |
655 | ||
ae3bb8ea | 656 | =back |
64935bc6 | 657 | |
1ce7b77d KW |
658 | It is important to realize when you use these Unicode boundaries, |
659 | that you are taking a risk that a future version of Perl which contains | |
660 | a later version of the Unicode Standard will not work precisely the same | |
661 | way as it did when your code was written. These rules are not | |
662 | considered stable and have been somewhat more subject to change than the | |
663 | rest of the Standard. Unicode reserves the right to change them at | |
664 | will, and Perl reserves the right to update its implementation to | |
665 | Unicode's new rules. In the past, some changes have been because new | |
666 | characters have been added to the Standard which have different | |
667 | characteristics than all previous characters, so new rules are | |
668 | formulated for handling them. These should not cause any backward | |
669 | compatibility issues. But some changes have changed the treatment of | |
670 | existing characters because the Unicode Technical Committee has decided | |
671 | that the change is warranted for whatever reason. This could be to fix | |
672 | a bug, or because they think better results are obtained with the new | |
673 | rule. | |
674 | ||
675 | It is also important to realize that these are default boundary | |
676 | definitions, and that implementations may wish to tailor the results for | |
39a8449e | 677 | particular purposes and locales. For example, some languages, such as |
be8bb1ac KW |
678 | Japanese and Thai, require dictionary lookup to accurately determine |
679 | word boundaries. | |
54bdcd8e | 680 | |
8a118206 RGS |
681 | Mnemonic: I<b>oundary. |
682 | ||
683 | =back | |
684 | ||
685 | =head4 Examples | |
686 | ||
687 | "cat" =~ /\Acat/; # Match. | |
688 | "cat" =~ /cat\Z/; # Match. | |
689 | "cat\n" =~ /cat\Z/; # Match. | |
690 | "cat\n" =~ /cat\z/; # No match. | |
691 | ||
692 | "cat" =~ /\bcat\b/; # Matches. | |
693 | "cats" =~ /\bcat\b/; # No match. | |
694 | "cat" =~ /\bcat\B/; # No match. | |
695 | "cats" =~ /\bcat\B/; # Match. | |
696 | ||
697 | while ("cat dog" =~ /(\w+)/g) { | |
698 | print $1; # Prints 'catdog' | |
699 | } | |
700 | while ("cat dog" =~ /\G(\w+)/g) { | |
701 | print $1; # Prints 'cat' | |
702 | } | |
703 | ||
54bdcd8e KW |
704 | my $s = "He said, \"Is pi 3.14? (I'm not sure).\""; |
705 | print join("|", $s =~ m/ ( .+? \b ) /xg), "\n"; | |
706 | print join("|", $s =~ m/ ( .+? \b{wb} ) /xg), "\n"; | |
ae3bb8ea | 707 | prints |
54bdcd8e KW |
708 | He| |said|, "|Is| |pi| |3|.|14|? (|I|'|m| |not| |sure |
709 | He| |said|,| |"|Is| |pi| |3.14|?| |(|I'm| |not| |sure|)|.|" | |
ae3bb8ea | 710 | |
8a118206 RGS |
711 | =head2 Misc |
712 | ||
713 | Here we document the backslash sequences that don't fall in one of the | |
b6538e4f | 714 | categories above. These are: |
8a118206 RGS |
715 | |
716 | =over 4 | |
717 | ||
8a118206 RGS |
718 | =item \K |
719 | ||
b6538e4f TC |
720 | This appeared in perl 5.10.0. Anything matched left of C<\K> is |
721 | not included in C<$&>, and will not be replaced if the pattern is | |
722 | used in a substitution. This lets you write C<s/PAT1 \K PAT2/REPL/x> | |
8a118206 RGS |
723 | instead of C<s/(PAT1) PAT2/${1}REPL/x> or C<s/(?<=PAT1) PAT2/REPL/x>. |
724 | ||
725 | Mnemonic: I<K>eep. | |
726 | ||
df225385 KW |
727 | =item \N |
728 | ||
2171640d | 729 | This feature, available starting in v5.12, matches any character |
b6538e4f | 730 | that is B<not> a newline. It is a short-hand for writing C<[^\n]>, and is |
b3b85878 KW |
731 | identical to the C<.> metasymbol, except under the C</s> flag, which changes |
732 | the meaning of C<.>, but not C<\N>. | |
df225385 | 733 | |
e526e8bb | 734 | Note that C<\N{...}> can mean a |
fb121860 KW |
735 | L<named or numbered character |
736 | |/Named or numbered characters and character sequences>. | |
df225385 KW |
737 | |
738 | Mnemonic: Complement of I<\n>. | |
739 | ||
8a118206 | 740 | =item \R |
6b46370c | 741 | X<\R> |
8a118206 | 742 | |
b6538e4f TC |
743 | C<\R> matches a I<generic newline>; that is, anything considered a |
744 | linebreak sequence by Unicode. This includes all characters matched by | |
745 | C<\v> (vertical whitespace), and the multi character sequence C<"\x0D\x0A"> | |
746 | (carriage return followed by a line feed, sometimes called the network | |
747 | newline; it's the end of line sequence used in Microsoft text files opened | |
1978b668 | 748 | in binary mode). C<\R> is equivalent to C<< (?>\x0D\x0A|\v) >>. (The |
040ac264 FC |
749 | reason it doesn't backtrack is that the sequence is considered |
750 | inseparable. That means that | |
1978b668 KW |
751 | |
752 | "\x0D\x0A" =~ /^\R\x0A$/ # No match | |
753 | ||
754 | fails, because the C<\R> matches the entire string, and won't backtrack | |
755 | to match just the C<"\x0D">.) Since | |
b6538e4f TC |
756 | C<\R> can match a sequence of more than one character, it cannot be put |
757 | inside a bracketed character class; C</[\R]/> is an error; use C<\v> | |
758 | instead. C<\R> was introduced in perl 5.10.0. | |
8a118206 | 759 | |
8129baca KW |
760 | Note that this does not respect any locale that might be in effect; it |
761 | matches according to the platform's native character set. | |
762 | ||
10fdd326 JH |
763 | Mnemonic: none really. C<\R> was picked because PCRE already uses C<\R>, |
764 | and more importantly because Unicode recommends such a regular expression | |
b6538e4f | 765 | metacharacter, and suggests C<\R> as its notation. |
8a118206 RGS |
766 | |
767 | =item \X | |
6b46370c | 768 | X<\X> |
8a118206 | 769 | |
0111a78f | 770 | This matches a Unicode I<extended grapheme cluster>. |
8a118206 | 771 | |
10fdd326 | 772 | C<\X> matches quite well what normal (non-Unicode-programmer) usage |
0111a78f | 773 | would consider a single character. As an example, consider a G with some sort |
c670e63a | 774 | of diacritic mark, such as an arrow. There is no such single character in |
df225385 | 775 | Unicode, but one can be composed by using a G followed by a Unicode "COMBINING |
c670e63a KW |
776 | UPWARDS ARROW BELOW", and would be displayed by Unicode-aware software as if it |
777 | were a single character. | |
10fdd326 | 778 | |
aa9e685b KW |
779 | The match is greedy and non-backtracking, so that the cluster is never |
780 | broken up into smaller components. | |
781 | ||
64935bc6 KW |
782 | See also L<C<\b{gcb}>|/\b{}, \b, \B{}, \B>. |
783 | ||
8a118206 RGS |
784 | Mnemonic: eI<X>tended Unicode character. |
785 | ||
786 | =back | |
787 | ||
788 | =head4 Examples | |
789 | ||
f822d0dd | 790 | $str =~ s/foo\Kbar/baz/g; # Change any 'bar' following a 'foo' to 'baz' |
d8b950dc | 791 | $str =~ s/(.)\K\g1//g; # Delete duplicated characters. |
8a118206 RGS |
792 | |
793 | "\n" =~ /^\R$/; # Match, \n is a generic newline. | |
794 | "\r" =~ /^\R$/; # Match, \r is a generic newline. | |
795 | "\r\n" =~ /^\R$/; # Match, \r\n is a generic newline. | |
796 | ||
b6538e4f | 797 | "P\x{307}" =~ /^\X$/ # \X matches a P with a dot above. |
8a118206 RGS |
798 | |
799 | =cut |