Commit | Line | Data |
---|---|---|
a0d0e21e LW |
1 | =head1 NAME |
2 | ||
3 | perlre - Perl regular expressions | |
4 | ||
5 | =head1 DESCRIPTION | |
6 | ||
cb1a09d0 AD |
7 | This page describes the syntax of regular expressions in Perl. For a |
8 | description of how to actually I<use> regular expressions in matching | |
9 | operations, plus various examples of the same, see C<m//> and C<s///> in | |
10 | L<perlop>. | |
11 | ||
12 | The matching operations can | |
a0d0e21e LW |
13 | have various modifiers, some of which relate to the interpretation of |
14 | the regular expression inside. These are: | |
15 | ||
55497cff | 16 | =over 4 |
17 | ||
18 | =item i | |
19 | ||
20 | Do case-insensitive pattern matching. | |
21 | ||
22 | =item m | |
23 | ||
24 | Treat string as multiple lines. That is, change "^" and "$" from matching | |
25 | only at the very start or end of the string to the start or end of any | |
26 | line anywhere within the string, | |
27 | ||
28 | =item s | |
29 | ||
30 | Treat string as single line. That is, change "." to match any character | |
31 | whatsoever, even a newline, which it normally would not match. | |
32 | ||
33 | =item x | |
34 | ||
35 | Extend your pattern's legibility by permitting whitespace and comments. | |
36 | ||
37 | =back | |
a0d0e21e LW |
38 | |
39 | These are usually written as "the C</x> modifier", even though the delimiter | |
40 | in question might not actually be a slash. In fact, any of these | |
41 | modifiers may also be embedded within the regular expression itself using | |
42 | the new C<(?...)> construct. See below. | |
43 | ||
4633a7c4 | 44 | The C</x> modifier itself needs a little more explanation. It tells |
55497cff | 45 | the regular expression parser to ignore whitespace that is neither |
46 | backslashed nor within a character class. You can use this to break up | |
4633a7c4 LW |
47 | your regular expression into (slightly) more readable parts. The C<#> |
48 | character is also treated as a metacharacter introducing a comment, | |
55497cff | 49 | just as in ordinary Perl code. This also means that if you want real |
50 | whitespace or C<#> characters in the pattern that you'll have to either | |
51 | escape them or encode them using octal or hex escapes. Taken together, | |
52 | these features go a long way towards making Perl's regular expressions | |
53 | more readable. See the C comment deletion code in L<perlop>. | |
a0d0e21e LW |
54 | |
55 | =head2 Regular Expressions | |
56 | ||
57 | The patterns used in pattern matching are regular expressions such as | |
58 | those supplied in the Version 8 regexp routines. (In fact, the | |
59 | routines are derived (distantly) from Henry Spencer's freely | |
60 | redistributable reimplementation of the V8 routines.) | |
61 | See L<Version 8 Regular Expressions> for details. | |
62 | ||
63 | In particular the following metacharacters have their standard I<egrep>-ish | |
64 | meanings: | |
65 | ||
66 | \ Quote the next metacharacter | |
67 | ^ Match the beginning of the line | |
68 | . Match any character (except newline) | |
c07a80fd | 69 | $ Match the end of the line (or before newline at the end) |
a0d0e21e LW |
70 | | Alternation |
71 | () Grouping | |
72 | [] Character class | |
73 | ||
74 | By default, the "^" character is guaranteed to match only at the | |
75 | beginning of the string, the "$" character only at the end (or before the | |
76 | newline at the end) and Perl does certain optimizations with the | |
77 | assumption that the string contains only one line. Embedded newlines | |
78 | will not be matched by "^" or "$". You may, however, wish to treat a | |
79 | string as a multi-line buffer, such that the "^" will match after any | |
80 | newline within the string, and "$" will match before any newline. At the | |
81 | cost of a little more overhead, you can do this by using the /m modifier | |
82 | on the pattern match operator. (Older programs did this by setting C<$*>, | |
83 | but this practice is deprecated in Perl 5.) | |
84 | ||
85 | To facilitate multi-line substitutions, the "." character never matches a | |
55497cff | 86 | newline unless you use the C</s> modifier, which in effect tells Perl to pretend |
a0d0e21e LW |
87 | the string is a single line--even if it isn't. The C</s> modifier also |
88 | overrides the setting of C<$*>, in case you have some (badly behaved) older | |
89 | code that sets it in another module. | |
90 | ||
91 | The following standard quantifiers are recognized: | |
92 | ||
93 | * Match 0 or more times | |
94 | + Match 1 or more times | |
95 | ? Match 1 or 0 times | |
96 | {n} Match exactly n times | |
97 | {n,} Match at least n times | |
98 | {n,m} Match at least n but not more than m times | |
99 | ||
100 | (If a curly bracket occurs in any other context, it is treated | |
101 | as a regular character.) The "*" modifier is equivalent to C<{0,}>, the "+" | |
25f94b33 | 102 | modifier to C<{1,}>, and the "?" modifier to C<{0,1}>. n and m are limited |
c07a80fd | 103 | to integral values less than 65536. |
a0d0e21e LW |
104 | |
105 | By default, a quantified subpattern is "greedy", that is, it will match as | |
32fd1c90 | 106 | many times as possible without causing the rest of the pattern not to match. |
107 | The standard quantifiers are all "greedy", in that they match as many | |
a0d0e21e LW |
108 | occurrences as possible (given a particular starting location) without |
109 | causing the pattern to fail. If you want it to match the minimum number | |
110 | of times possible, follow the quantifier with a "?" after any of them. | |
111 | Note that the meanings don't change, just the "gravity": | |
112 | ||
113 | *? Match 0 or more times | |
114 | +? Match 1 or more times | |
115 | ?? Match 0 or 1 time | |
116 | {n}? Match exactly n times | |
117 | {n,}? Match at least n times | |
118 | {n,m}? Match at least n but not more than m times | |
119 | ||
120 | Since patterns are processed as double quoted strings, the following | |
121 | also work: | |
122 | ||
0f36ee90 | 123 | \t tab (HT, TAB) |
124 | \n newline (LF, NL) | |
125 | \r return (CR) | |
126 | \f form feed (FF) | |
127 | \a alarm (bell) (BEL) | |
128 | \e escape (think troff) (ESC) | |
cb1a09d0 AD |
129 | \033 octal char (think of a PDP-11) |
130 | \x1B hex char | |
a0d0e21e | 131 | \c[ control char |
cb1a09d0 AD |
132 | \l lowercase next char (think vi) |
133 | \u uppercase next char (think vi) | |
134 | \L lowercase till \E (think vi) | |
135 | \U uppercase till \E (think vi) | |
136 | \E end case modification (think vi) | |
a0d0e21e LW |
137 | \Q quote regexp metacharacters till \E |
138 | ||
139 | In addition, Perl defines the following: | |
140 | ||
141 | \w Match a "word" character (alphanumeric plus "_") | |
142 | \W Match a non-word character | |
143 | \s Match a whitespace character | |
144 | \S Match a non-whitespace character | |
145 | \d Match a digit character | |
146 | \D Match a non-digit character | |
147 | ||
148 | Note that C<\w> matches a single alphanumeric character, not a whole | |
cb1a09d0 AD |
149 | word. To match a word you'd need to say C<\w+>. You may use C<\w>, |
150 | C<\W>, C<\s>, C<\S>, C<\d> and C<\D> within character classes (though not | |
151 | as either end of a range). | |
a0d0e21e LW |
152 | |
153 | Perl defines the following zero-width assertions: | |
154 | ||
155 | \b Match a word boundary | |
156 | \B Match a non-(word boundary) | |
157 | \A Match only at beginning of string | |
c07a80fd | 158 | \Z Match only at end of string (or before newline at the end) |
a0d0e21e LW |
159 | \G Match only where previous m//g left off |
160 | ||
161 | A word boundary (C<\b>) is defined as a spot between two characters that | |
162 | has a C<\w> on one side of it and and a C<\W> on the other side of it (in | |
163 | either order), counting the imaginary characters off the beginning and | |
164 | end of the string as matching a C<\W>. (Within character classes C<\b> | |
165 | represents backspace rather than a word boundary.) The C<\A> and C<\Z> are | |
166 | just like "^" and "$" except that they won't match multiple times when the | |
167 | C</m> modifier is used, while "^" and "$" will match at every internal line | |
c07a80fd | 168 | boundary. To match the actual end of the string, not ignoring newline, |
169 | you can use C<\Z(?!\n)>. | |
a0d0e21e | 170 | |
0f36ee90 | 171 | When the bracketing construct C<( ... )> is used, \E<lt>digitE<gt> matches the |
cb1a09d0 | 172 | digit'th substring. Outside of the pattern, always use "$" instead of "\" |
0f36ee90 | 173 | in front of the digit. (While the \E<lt>digitE<gt> notation can on rare occasion work |
cb1a09d0 | 174 | outside the current pattern, this should not be relied upon. See the |
0f36ee90 | 175 | WARNING below.) The scope of $E<lt>digitE<gt> (and C<$`>, C<$&>, and C<$'>) |
cb1a09d0 AD |
176 | extends to the end of the enclosing BLOCK or eval string, or to the next |
177 | successful pattern match, whichever comes first. If you want to use | |
32fd1c90 | 178 | parentheses to delimit a subpattern (e.g. a set of alternatives) without |
84dc3c4d | 179 | saving it as a subpattern, follow the ( with a ?:. |
cb1a09d0 AD |
180 | |
181 | You may have as many parentheses as you wish. If you have more | |
a0d0e21e LW |
182 | than 9 substrings, the variables $10, $11, ... refer to the |
183 | corresponding substring. Within the pattern, \10, \11, etc. refer back | |
184 | to substrings if there have been at least that many left parens before | |
c07a80fd | 185 | the backreference. Otherwise (for backward compatibility) \10 is the |
a0d0e21e LW |
186 | same as \010, a backspace, and \11 the same as \011, a tab. And so |
187 | on. (\1 through \9 are always backreferences.) | |
188 | ||
189 | C<$+> returns whatever the last bracket match matched. C<$&> returns the | |
0f36ee90 | 190 | entire matched string. (C<$0> used to return the same thing, but not any |
a0d0e21e LW |
191 | more.) C<$`> returns everything before the matched string. C<$'> returns |
192 | everything after the matched string. Examples: | |
193 | ||
194 | s/^([^ ]*) *([^ ]*)/$2 $1/; # swap first two words | |
195 | ||
196 | if (/Time: (..):(..):(..)/) { | |
197 | $hours = $1; | |
198 | $minutes = $2; | |
199 | $seconds = $3; | |
200 | } | |
201 | ||
202 | You will note that all backslashed metacharacters in Perl are | |
203 | alphanumeric, such as C<\b>, C<\w>, C<\n>. Unlike some other regular expression | |
204 | languages, there are no backslashed symbols that aren't alphanumeric. | |
0f36ee90 | 205 | So anything that looks like \\, \(, \), \E<lt>, \E<gt>, \{, or \} is always |
a0d0e21e LW |
206 | interpreted as a literal character, not a metacharacter. This makes it |
207 | simple to quote a string that you want to use for a pattern but that | |
208 | you are afraid might contain metacharacters. Simply quote all the | |
209 | non-alphanumeric characters: | |
210 | ||
211 | $pattern =~ s/(\W)/\\$1/g; | |
212 | ||
213 | You can also use the built-in quotemeta() function to do this. | |
214 | An even easier way to quote metacharacters right in the match operator | |
c07a80fd | 215 | is to say |
a0d0e21e LW |
216 | |
217 | /$unquoted\Q$quoted\E$unquoted/ | |
218 | ||
219 | Perl 5 defines a consistent extension syntax for regular expressions. | |
220 | The syntax is a pair of parens with a question mark as the first thing | |
221 | within the parens (this was a syntax error in Perl 4). The character | |
222 | after the question mark gives the function of the extension. Several | |
223 | extensions are already supported: | |
224 | ||
225 | =over 10 | |
226 | ||
227 | =item (?#text) | |
228 | ||
cb1a09d0 AD |
229 | A comment. The text is ignored. If the C</x> switch is used to enable |
230 | whitespace formatting, a simple C<#> will suffice. | |
a0d0e21e LW |
231 | |
232 | =item (?:regexp) | |
233 | ||
0f36ee90 | 234 | This groups things like "()" but doesn't make backreferences like "()" does. So |
a0d0e21e LW |
235 | |
236 | split(/\b(?:a|b|c)\b/) | |
237 | ||
238 | is like | |
239 | ||
240 | split(/\b(a|b|c)\b/) | |
241 | ||
242 | but doesn't spit out extra fields. | |
243 | ||
244 | =item (?=regexp) | |
245 | ||
246 | A zero-width positive lookahead assertion. For example, C</\w+(?=\t)/> | |
247 | matches a word followed by a tab, without including the tab in C<$&>. | |
248 | ||
249 | =item (?!regexp) | |
250 | ||
251 | A zero-width negative lookahead assertion. For example C</foo(?!bar)/> | |
252 | matches any occurrence of "foo" that isn't followed by "bar". Note | |
253 | however that lookahead and lookbehind are NOT the same thing. You cannot | |
254 | use this for lookbehind: C</(?!foo)bar/> will not find an occurrence of | |
255 | "bar" that is preceded by something which is not "foo". That's because | |
256 | the C<(?!foo)> is just saying that the next thing cannot be "foo"--and | |
257 | it's not, it's a "bar", so "foobar" will match. You would have to do | |
0f36ee90 | 258 | something like C</(?!foo)...bar/> for that. We say "like" because there's |
a0d0e21e | 259 | the case of your "bar" not having three characters before it. You could |
c07a80fd | 260 | cover that this way: C</(?:(?!foo)...|^..?)bar/>. Sometimes it's still |
a0d0e21e LW |
261 | easier just to say: |
262 | ||
c07a80fd | 263 | if (/foo/ && $` =~ /bar$/) |
a0d0e21e LW |
264 | |
265 | ||
266 | =item (?imsx) | |
267 | ||
268 | One or more embedded pattern-match modifiers. This is particularly | |
269 | useful for patterns that are specified in a table somewhere, some of | |
270 | which want to be case sensitive, and some of which don't. The case | |
271 | insensitive ones merely need to include C<(?i)> at the front of the | |
272 | pattern. For example: | |
273 | ||
274 | $pattern = "foobar"; | |
c07a80fd | 275 | if ( /$pattern/i ) |
a0d0e21e LW |
276 | |
277 | # more flexible: | |
278 | ||
279 | $pattern = "(?i)foobar"; | |
c07a80fd | 280 | if ( /$pattern/ ) |
a0d0e21e LW |
281 | |
282 | =back | |
283 | ||
284 | The specific choice of question mark for this and the new minimal | |
285 | matching construct was because 1) question mark is pretty rare in older | |
286 | regular expressions, and 2) whenever you see one, you should stop | |
287 | and "question" exactly what is going on. That's psychology... | |
288 | ||
c07a80fd | 289 | =head2 Backtracking |
290 | ||
291 | A fundamental feature of regular expression matching involves the notion | |
292 | called I<backtracking>. which is used (when needed) by all regular | |
293 | expression quantifiers, namely C<*>, C<*?>, C<+>, C<+?>, C<{n,m}>, and | |
294 | C<{n,m}?>. | |
295 | ||
296 | For a regular expression to match, the I<entire> regular expression must | |
297 | match, not just part of it. So if the beginning of a pattern containing a | |
298 | quantifier succeeds in a way that causes later parts in the pattern to | |
299 | fail, the matching engine backs up and recalculates the beginning | |
300 | part--that's why it's called backtracking. | |
301 | ||
302 | Here is an example of backtracking: Let's say you want to find the | |
303 | word following "foo" in the string "Food is on the foo table.": | |
304 | ||
305 | $_ = "Food is on the foo table."; | |
306 | if ( /\b(foo)\s+(\w+)/i ) { | |
307 | print "$2 follows $1.\n"; | |
308 | } | |
309 | ||
310 | When the match runs, the first part of the regular expression (C<\b(foo)>) | |
311 | finds a possible match right at the beginning of the string, and loads up | |
312 | $1 with "Foo". However, as soon as the matching engine sees that there's | |
313 | no whitespace following the "Foo" that it had saved in $1, it realizes its | |
314 | mistake and starts over again one character after where it had had the | |
315 | tentative match. This time it goes all the way until the next occurrence | |
316 | of "foo". The complete regular expression matches this time, and you get | |
317 | the expected output of "table follows foo." | |
318 | ||
319 | Sometimes minimal matching can help a lot. Imagine you'd like to match | |
320 | everything between "foo" and "bar". Initially, you write something | |
321 | like this: | |
322 | ||
323 | $_ = "The food is under the bar in the barn."; | |
324 | if ( /foo(.*)bar/ ) { | |
325 | print "got <$1>\n"; | |
326 | } | |
327 | ||
328 | Which perhaps unexpectedly yields: | |
329 | ||
330 | got <d is under the bar in the > | |
331 | ||
332 | That's because C<.*> was greedy, so you get everything between the | |
333 | I<first> "foo" and the I<last> "bar". In this case, it's more effective | |
334 | to use minimal matching to make sure you get the text between a "foo" | |
335 | and the first "bar" thereafter. | |
336 | ||
337 | if ( /foo(.*?)bar/ ) { print "got <$1>\n" } | |
338 | got <d is under the > | |
339 | ||
340 | Here's another example: let's say you'd like to match a number at the end | |
341 | of a string, and you also want to keep the preceding part the match. | |
342 | So you write this: | |
343 | ||
344 | $_ = "I have 2 numbers: 53147"; | |
345 | if ( /(.*)(\d*)/ ) { # Wrong! | |
346 | print "Beginning is <$1>, number is <$2>.\n"; | |
347 | } | |
348 | ||
349 | That won't work at all, because C<.*> was greedy and gobbled up the | |
350 | whole string. As C<\d*> can match on an empty string the complete | |
351 | regular expression matched successfully. | |
352 | ||
8e1088bc | 353 | Beginning is <I have 2 numbers: 53147>, number is <>. |
c07a80fd | 354 | |
355 | Here are some variants, most of which don't work: | |
356 | ||
357 | $_ = "I have 2 numbers: 53147"; | |
358 | @pats = qw{ | |
359 | (.*)(\d*) | |
360 | (.*)(\d+) | |
361 | (.*?)(\d*) | |
362 | (.*?)(\d+) | |
363 | (.*)(\d+)$ | |
364 | (.*?)(\d+)$ | |
365 | (.*)\b(\d+)$ | |
366 | (.*\D)(\d+)$ | |
367 | }; | |
368 | ||
369 | for $pat (@pats) { | |
370 | printf "%-12s ", $pat; | |
371 | if ( /$pat/ ) { | |
372 | print "<$1> <$2>\n"; | |
373 | } else { | |
374 | print "FAIL\n"; | |
375 | } | |
376 | } | |
377 | ||
378 | That will print out: | |
379 | ||
380 | (.*)(\d*) <I have 2 numbers: 53147> <> | |
381 | (.*)(\d+) <I have 2 numbers: 5314> <7> | |
382 | (.*?)(\d*) <> <> | |
383 | (.*?)(\d+) <I have > <2> | |
384 | (.*)(\d+)$ <I have 2 numbers: 5314> <7> | |
385 | (.*?)(\d+)$ <I have 2 numbers: > <53147> | |
386 | (.*)\b(\d+)$ <I have 2 numbers: > <53147> | |
387 | (.*\D)(\d+)$ <I have 2 numbers: > <53147> | |
388 | ||
389 | As you see, this can be a bit tricky. It's important to realize that a | |
390 | regular expression is merely a set of assertions that gives a definition | |
391 | of success. There may be 0, 1, or several different ways that the | |
392 | definition might succeed against a particular string. And if there are | |
393 | multiple ways it might succeed, you need to understand backtracking in | |
394 | order to know which variety of success you will achieve. | |
395 | ||
396 | When using lookahead assertions and negations, this can all get even | |
397 | tricker. Imagine you'd like to find a sequence of nondigits not | |
398 | followed by "123". You might try to write that as | |
399 | ||
400 | $_ = "ABC123"; | |
401 | if ( /^\D*(?!123)/ ) { # Wrong! | |
402 | print "Yup, no 123 in $_\n"; | |
403 | } | |
404 | ||
405 | But that isn't going to match; at least, not the way you're hoping. It | |
406 | claims that there is no 123 in the string. Here's a clearer picture of | |
407 | why it that pattern matches, contrary to popular expectations: | |
408 | ||
409 | $x = 'ABC123' ; | |
410 | $y = 'ABC445' ; | |
411 | ||
412 | print "1: got $1\n" if $x =~ /^(ABC)(?!123)/ ; | |
413 | print "2: got $1\n" if $y =~ /^(ABC)(?!123)/ ; | |
414 | ||
415 | print "3: got $1\n" if $x =~ /^(\D*)(?!123)/ ; | |
416 | print "4: got $1\n" if $y =~ /^(\D*)(?!123)/ ; | |
417 | ||
418 | This prints | |
419 | ||
420 | 2: got ABC | |
421 | 3: got AB | |
422 | 4: got ABC | |
423 | ||
424 | You might have expected test 3 to fail because it just seems to a more | |
425 | general purpose version of test 1. The important difference between | |
426 | them is that test 3 contains a quantifier (C<\D*>) and so can use | |
427 | backtracking, whereas test 1 will not. What's happening is | |
428 | that you've asked "Is it true that at the start of $x, following 0 or more | |
429 | nondigits, you have something that's not 123?" If the pattern matcher had | |
430 | let C<\D*> expand to "ABC", this would have caused the whole pattern to | |
431 | fail. | |
432 | The search engine will initially match C<\D*> with "ABC". Then it will | |
433 | try to match C<(?!123> with "123" which, of course, fails. But because | |
434 | a quantifier (C<\D*>) has been used in the regular expression, the | |
435 | search engine can backtrack and retry the match differently | |
436 | in the hope of matching the complete regular expression. | |
437 | ||
438 | Well now, | |
439 | the pattern really, I<really> wants to succeed, so it uses the | |
440 | standard regexp backoff-and-retry and lets C<\D*> expand to just "AB" this | |
441 | time. Now there's indeed something following "AB" that is not | |
442 | "123". It's in fact "C123", which suffices. | |
443 | ||
444 | We can deal with this by using both an assertion and a negation. We'll | |
445 | say that the first part in $1 must be followed by a digit, and in fact, it | |
446 | must also be followed by something that's not "123". Remember that the | |
447 | lookaheads are zero-width expressions--they only look, but don't consume | |
448 | any of the string in their match. So rewriting this way produces what | |
449 | you'd expect; that is, case 5 will fail, but case 6 succeeds: | |
450 | ||
451 | print "5: got $1\n" if $x =~ /^(\D*)(?=\d)(?!123)/ ; | |
452 | print "6: got $1\n" if $y =~ /^(\D*)(?=\d)(?!123)/ ; | |
453 | ||
454 | 6: got ABC | |
455 | ||
456 | In other words, the two zero-width assertions next to each other work like | |
457 | they're ANDed together, just as you'd use any builtin assertions: C</^$/> | |
458 | matches only if you're at the beginning of the line AND the end of the | |
459 | line simultaneously. The deeper underlying truth is that juxtaposition in | |
460 | regular expressions always means AND, except when you write an explicit OR | |
461 | using the vertical bar. C</ab/> means match "a" AND (then) match "b", | |
462 | although the attempted matches are made at different positions because "a" | |
463 | is not a zero-width assertion, but a one-width assertion. | |
464 | ||
465 | One warning: particularly complicated regular expressions can take | |
466 | exponential time to solve due to the immense number of possible ways they | |
467 | can use backtracking to try match. For example this will take a very long | |
468 | time to run | |
469 | ||
470 | /((a{0,5}){0,5}){0,5}/ | |
471 | ||
472 | And if you used C<*>'s instead of limiting it to 0 through 5 matches, then | |
473 | it would take literally forever--or until you ran out of stack space. | |
474 | ||
a0d0e21e LW |
475 | =head2 Version 8 Regular Expressions |
476 | ||
477 | In case you're not familiar with the "regular" Version 8 regexp | |
478 | routines, here are the pattern-matching rules not described above. | |
479 | ||
480 | Any single character matches itself, unless it is a I<metacharacter> | |
481 | with a special meaning described here or above. You can cause | |
482 | characters which normally function as metacharacters to be interpreted | |
483 | literally by prefixing them with a "\" (e.g. "\." matches a ".", not any | |
484 | character; "\\" matches a "\"). A series of characters matches that | |
485 | series of characters in the target string, so the pattern C<blurfl> | |
486 | would match "blurfl" in the target string. | |
487 | ||
488 | You can specify a character class, by enclosing a list of characters | |
489 | in C<[]>, which will match any one of the characters in the list. If the | |
490 | first character after the "[" is "^", the class matches any character not | |
491 | in the list. Within a list, the "-" character is used to specify a | |
492 | range, so that C<a-z> represents all the characters between "a" and "z", | |
493 | inclusive. | |
494 | ||
495 | Characters may be specified using a metacharacter syntax much like that | |
496 | used in C: "\n" matches a newline, "\t" a tab, "\r" a carriage return, | |
497 | "\f" a form feed, etc. More generally, \I<nnn>, where I<nnn> is a string | |
498 | of octal digits, matches the character whose ASCII value is I<nnn>. | |
0f36ee90 | 499 | Similarly, \xI<nn>, where I<nn> are hexadecimal digits, matches the |
a0d0e21e LW |
500 | character whose ASCII value is I<nn>. The expression \cI<x> matches the |
501 | ASCII character control-I<x>. Finally, the "." metacharacter matches any | |
502 | character except "\n" (unless you use C</s>). | |
503 | ||
504 | You can specify a series of alternatives for a pattern using "|" to | |
505 | separate them, so that C<fee|fie|foe> will match any of "fee", "fie", | |
506 | or "foe" in the target string (as would C<f(e|i|o)e>). Note that the | |
507 | first alternative includes everything from the last pattern delimiter | |
508 | ("(", "[", or the beginning of the pattern) up to the first "|", and | |
509 | the last alternative contains everything from the last "|" to the next | |
510 | pattern delimiter. For this reason, it's common practice to include | |
511 | alternatives in parentheses, to minimize confusion about where they | |
748a9306 LW |
512 | start and end. Note however that "|" is interpreted as a literal with |
513 | square brackets, so if you write C<[fee|fie|foe]> you're really only | |
514 | matching C<[feio|]>. | |
a0d0e21e LW |
515 | |
516 | Within a pattern, you may designate subpatterns for later reference by | |
517 | enclosing them in parentheses, and you may refer back to the I<n>th | |
c07a80fd | 518 | subpattern later in the pattern using the metacharacter \I<n>. |
a0d0e21e LW |
519 | Subpatterns are numbered based on the left to right order of their |
520 | opening parenthesis. Note that a backreference matches whatever | |
521 | actually matched the subpattern in the string being examined, not the | |
748a9306 | 522 | rules for that subpattern. Therefore, C<(0|0x)\d*\s\1\d*> will |
a0d0e21e | 523 | match "0x1234 0x4321",but not "0x1234 01234", since subpattern 1 |
748a9306 | 524 | actually matched "0x", even though the rule C<0|0x> could |
a0d0e21e | 525 | potentially match the leading 0 in the second number. |
cb1a09d0 AD |
526 | |
527 | =head2 WARNING on \1 vs $1 | |
528 | ||
529 | Some people get too used to writing things like | |
530 | ||
531 | $pattern =~ s/(\W)/\\\1/g; | |
532 | ||
533 | This is grandfathered for the RHS of a substitute to avoid shocking the | |
534 | B<sed> addicts, but it's a dirty habit to get into. That's because in | |
535 | PerlThink, the right-hand side of a C<s///> is a double-quoted string. C<\1> in | |
536 | the usual double-quoted string means a control-A. The customary Unix | |
537 | meaning of C<\1> is kludged in for C<s///>. However, if you get into the habit | |
538 | of doing that, you get yourself into trouble if you then add an C</e> | |
539 | modifier. | |
540 | ||
541 | s/(\d+)/ \1 + 1 /eg; | |
542 | ||
543 | Or if you try to do | |
544 | ||
545 | s/(\d+)/\1000/; | |
546 | ||
547 | You can't disambiguate that by saying C<\{1}000>, whereas you can fix it with | |
548 | C<${1}000>. Basically, the operation of interpolation should not be confused | |
549 | with the operation of matching a backreference. Certainly they mean two | |
550 | different things on the I<left> side of the C<s///>. |