Commit | Line | Data |
---|---|---|
68dc0745 | 1 | =head1 NAME |
2 | ||
500071f4 | 3 | perlfaq6 - Regular Expressions ($Revision: 3606 $) |
68dc0745 | 4 | |
5 | =head1 DESCRIPTION | |
6 | ||
7 | This section is surprisingly small because the rest of the FAQ is | |
8 | littered with answers involving regular expressions. For example, | |
9 | decoding a URL and checking whether something is a number are handled | |
10 | with regular expressions, but those answers are found elsewhere in | |
b432a672 AL |
11 | this document (in L<perlfaq9>: "How do I decode or create those %-encodings |
12 | on the web" and L<perlfaq4>: "How do I determine whether a scalar is | |
13 | a number/whole/integer/float", to be precise). | |
68dc0745 | 14 | |
54310121 | 15 | =head2 How can I hope to use regular expressions without creating illegible and unmaintainable code? |
d74e8afc ITB |
16 | X<regex, legibility> X<regexp, legibility> |
17 | X<regular expression, legibility> X</x> | |
68dc0745 | 18 | |
19 | Three techniques can make regular expressions maintainable and | |
20 | understandable. | |
21 | ||
22 | =over 4 | |
23 | ||
d92eb7b0 | 24 | =item Comments Outside the Regex |
68dc0745 | 25 | |
26 | Describe what you're doing and how you're doing it, using normal Perl | |
27 | comments. | |
28 | ||
29 | # turn the line into the first word, a colon, and the | |
30 | # number of characters on the rest of the line | |
5a964f20 | 31 | s/^(\w+)(.*)/ lc($1) . ":" . length($2) /meg; |
68dc0745 | 32 | |
d92eb7b0 | 33 | =item Comments Inside the Regex |
68dc0745 | 34 | |
d92eb7b0 | 35 | The C</x> modifier causes whitespace to be ignored in a regex pattern |
68dc0745 | 36 | (except in a character class), and also allows you to use normal |
37 | comments there, too. As you can imagine, whitespace and comments help | |
38 | a lot. | |
39 | ||
40 | C</x> lets you turn this: | |
41 | ||
42 | s{<(?:[^>'"]*|".*?"|'.*?')+>}{}gs; | |
43 | ||
44 | into this: | |
45 | ||
46 | s{ < # opening angle bracket | |
47 | (?: # Non-backreffing grouping paren | |
48 | [^>'"] * # 0 or more things that are neither > nor ' nor " | |
49 | | # or else | |
50 | ".*?" # a section between double quotes (stingy match) | |
51 | | # or else | |
52 | '.*?' # a section between single quotes (stingy match) | |
53 | ) + # all occurring one or more times | |
54 | > # closing angle bracket | |
55 | }{}gsx; # replace with nothing, i.e. delete | |
56 | ||
57 | It's still not quite so clear as prose, but it is very useful for | |
58 | describing the meaning of each part of the pattern. | |
59 | ||
60 | =item Different Delimiters | |
61 | ||
62 | While we normally think of patterns as being delimited with C</> | |
63 | characters, they can be delimited by almost any character. L<perlre> | |
64 | describes this. For example, the C<s///> above uses braces as | |
65 | delimiters. Selecting another delimiter can avoid quoting the | |
66 | delimiter within the pattern: | |
67 | ||
68 | s/\/usr\/local/\/usr\/share/g; # bad delimiter choice | |
69 | s#/usr/local#/usr/share#g; # better | |
70 | ||
71 | =back | |
72 | ||
73 | =head2 I'm having trouble matching over more than one line. What's wrong? | |
d74e8afc | 74 | X<regex, multiline> X<regexp, multiline> X<regular expression, multiline> |
68dc0745 | 75 | |
3392b9ec JH |
76 | Either you don't have more than one line in the string you're looking |
77 | at (probably), or else you aren't using the correct modifier(s) on | |
78 | your pattern (possibly). | |
68dc0745 | 79 | |
80 | There are many ways to get multiline data into a string. If you want | |
81 | it to happen automatically while reading input, you'll want to set $/ | |
82 | (probably to '' for paragraphs or C<undef> for the whole file) to | |
83 | allow you to read more than one line at a time. | |
84 | ||
85 | Read L<perlre> to help you decide which of C</s> and C</m> (or both) | |
86 | you might want to use: C</s> allows dot to include newline, and C</m> | |
87 | allows caret and dollar to match next to a newline, not just at the | |
88 | end of the string. You do need to make sure that you've actually | |
89 | got a multiline string in there. | |
90 | ||
91 | For example, this program detects duplicate words, even when they span | |
92 | line breaks (but not paragraph ones). For this example, we don't need | |
93 | C</s> because we aren't using dot in a regular expression that we want | |
94 | to cross line boundaries. Neither do we need C</m> because we aren't | |
95 | wanting caret or dollar to match at any point inside the record next | |
96 | to newlines. But it's imperative that $/ be set to something other | |
97 | than the default, or else we won't actually ever have a multiline | |
98 | record read in. | |
99 | ||
100 | $/ = ''; # read in more whole paragraph, not just one line | |
101 | while ( <> ) { | |
5a964f20 | 102 | while ( /\b([\w'-]+)(\s+\1)+\b/gi ) { # word starts alpha |
68dc0745 | 103 | print "Duplicate $1 at paragraph $.\n"; |
54310121 | 104 | } |
105 | } | |
68dc0745 | 106 | |
107 | Here's code that finds sentences that begin with "From " (which would | |
108 | be mangled by many mailers): | |
109 | ||
110 | $/ = ''; # read in more whole paragraph, not just one line | |
111 | while ( <> ) { | |
112 | while ( /^From /gm ) { # /m makes ^ match next to \n | |
113 | print "leading from in paragraph $.\n"; | |
114 | } | |
115 | } | |
116 | ||
117 | Here's code that finds everything between START and END in a paragraph: | |
118 | ||
119 | undef $/; # read in whole file, not just one line or paragraph | |
120 | while ( <> ) { | |
fd89e497 | 121 | while ( /START(.*?)END/sgm ) { # /s makes . cross line boundaries |
68dc0745 | 122 | print "$1\n"; |
123 | } | |
124 | } | |
125 | ||
126 | =head2 How can I pull out lines between two patterns that are themselves on different lines? | |
d74e8afc | 127 | X<..> |
68dc0745 | 128 | |
129 | You can use Perl's somewhat exotic C<..> operator (documented in | |
130 | L<perlop>): | |
131 | ||
132 | perl -ne 'print if /START/ .. /END/' file1 file2 ... | |
133 | ||
134 | If you wanted text and not lines, you would use | |
135 | ||
65acb1b1 | 136 | perl -0777 -ne 'print "$1\n" while /START(.*?)END/gs' file1 file2 ... |
68dc0745 | 137 | |
138 | But if you want nested occurrences of C<START> through C<END>, you'll | |
139 | run up against the problem described in the question in this section | |
140 | on matching balanced text. | |
141 | ||
5a964f20 TC |
142 | Here's another example of using C<..>: |
143 | ||
144 | while (<>) { | |
145 | $in_header = 1 .. /^$/; | |
146 | $in_body = /^$/ .. eof(); | |
147 | # now choose between them | |
148 | } continue { | |
149 | reset if eof(); # fix $. | |
197aec24 | 150 | } |
5a964f20 | 151 | |
68dc0745 | 152 | =head2 I put a regular expression into $/ but it didn't work. What's wrong? |
d74e8afc ITB |
153 | X<$/, regexes in> X<$INPUT_RECORD_SEPARATOR, regexes in> |
154 | X<$RS, regexes in> | |
68dc0745 | 155 | |
197aec24 | 156 | Up to Perl 5.8.0, $/ has to be a string. This may change in 5.10, |
49d635f9 RGS |
157 | but don't get your hopes up. Until then, you can use these examples |
158 | if you really need to do this. | |
159 | ||
28b41a80 RGS |
160 | If you have File::Stream, this is easy. |
161 | ||
162 | use File::Stream; | |
163 | my $stream = File::Stream->new( | |
164 | $filehandle, | |
165 | separator => qr/\s*,\s*/, | |
166 | ); | |
167 | ||
168 | print "$_\n" while <$stream>; | |
169 | ||
170 | If you don't have File::Stream, you have to do a little more work. | |
171 | ||
172 | You can use the four argument form of sysread to continually add to | |
197aec24 | 173 | a buffer. After you add to the buffer, you check if you have a |
49d635f9 RGS |
174 | complete line (using your regular expression). |
175 | ||
176 | local $_ = ""; | |
177 | while( sysread FH, $_, 8192, length ) { | |
178 | while( s/^((?s).*?)your_pattern/ ) { | |
179 | my $record = $1; | |
180 | # do stuff here. | |
181 | } | |
182 | } | |
197aec24 | 183 | |
49d635f9 RGS |
184 | You can do the same thing with foreach and a match using the |
185 | c flag and the \G anchor, if you do not mind your entire file | |
186 | being in memory at the end. | |
197aec24 | 187 | |
49d635f9 RGS |
188 | local $_ = ""; |
189 | while( sysread FH, $_, 8192, length ) { | |
190 | foreach my $record ( m/\G((?s).*?)your_pattern/gc ) { | |
191 | # do stuff here. | |
192 | } | |
193 | substr( $_, 0, pos ) = "" if pos; | |
194 | } | |
68dc0745 | 195 | |
3fe9a6f1 | 196 | |
a6dd486b | 197 | =head2 How do I substitute case insensitively on the LHS while preserving case on the RHS? |
d74e8afc ITB |
198 | X<replace, case preserving> X<substitute, case preserving> |
199 | X<substitution, case preserving> X<s, case preserving> | |
68dc0745 | 200 | |
d92eb7b0 GS |
201 | Here's a lovely Perlish solution by Larry Rosler. It exploits |
202 | properties of bitwise xor on ASCII strings. | |
203 | ||
204 | $_= "this is a TEsT case"; | |
205 | ||
206 | $old = 'test'; | |
207 | $new = 'success'; | |
208 | ||
575cc754 | 209 | s{(\Q$old\E)} |
d92eb7b0 GS |
210 | { uc $new | (uc $1 ^ $1) . |
211 | (uc(substr $1, -1) ^ substr $1, -1) x | |
212 | (length($new) - length $1) | |
213 | }egi; | |
214 | ||
215 | print; | |
216 | ||
8305e449 | 217 | And here it is as a subroutine, modeled after the above: |
d92eb7b0 GS |
218 | |
219 | sub preserve_case($$) { | |
220 | my ($old, $new) = @_; | |
221 | my $mask = uc $old ^ $old; | |
222 | ||
223 | uc $new | $mask . | |
197aec24 | 224 | substr($mask, -1) x (length($new) - length($old)) |
d92eb7b0 GS |
225 | } |
226 | ||
227 | $a = "this is a TEsT case"; | |
228 | $a =~ s/(test)/preserve_case($1, "success")/egi; | |
229 | print "$a\n"; | |
230 | ||
231 | This prints: | |
232 | ||
233 | this is a SUcCESS case | |
234 | ||
74b9445a JP |
235 | As an alternative, to keep the case of the replacement word if it is |
236 | longer than the original, you can use this code, by Jeff Pinyan: | |
237 | ||
238 | sub preserve_case { | |
239 | my ($from, $to) = @_; | |
240 | my ($lf, $lt) = map length, @_; | |
7207e29d | 241 | |
74b9445a JP |
242 | if ($lt < $lf) { $from = substr $from, 0, $lt } |
243 | else { $from .= substr $to, $lf } | |
7207e29d | 244 | |
74b9445a JP |
245 | return uc $to | ($from ^ uc $from); |
246 | } | |
247 | ||
248 | This changes the sentence to "this is a SUcCess case." | |
249 | ||
d92eb7b0 GS |
250 | Just to show that C programmers can write C in any programming language, |
251 | if you prefer a more C-like solution, the following script makes the | |
252 | substitution have the same case, letter by letter, as the original. | |
253 | (It also happens to run about 240% slower than the Perlish solution runs.) | |
254 | If the substitution has more characters than the string being substituted, | |
255 | the case of the last character is used for the rest of the substitution. | |
68dc0745 | 256 | |
257 | # Original by Nathan Torkington, massaged by Jeffrey Friedl | |
258 | # | |
259 | sub preserve_case($$) | |
260 | { | |
261 | my ($old, $new) = @_; | |
262 | my ($state) = 0; # 0 = no change; 1 = lc; 2 = uc | |
263 | my ($i, $oldlen, $newlen, $c) = (0, length($old), length($new)); | |
264 | my ($len) = $oldlen < $newlen ? $oldlen : $newlen; | |
265 | ||
266 | for ($i = 0; $i < $len; $i++) { | |
267 | if ($c = substr($old, $i, 1), $c =~ /[\W\d_]/) { | |
268 | $state = 0; | |
269 | } elsif (lc $c eq $c) { | |
270 | substr($new, $i, 1) = lc(substr($new, $i, 1)); | |
271 | $state = 1; | |
272 | } else { | |
273 | substr($new, $i, 1) = uc(substr($new, $i, 1)); | |
274 | $state = 2; | |
275 | } | |
276 | } | |
277 | # finish up with any remaining new (for when new is longer than old) | |
278 | if ($newlen > $oldlen) { | |
279 | if ($state == 1) { | |
280 | substr($new, $oldlen) = lc(substr($new, $oldlen)); | |
281 | } elsif ($state == 2) { | |
282 | substr($new, $oldlen) = uc(substr($new, $oldlen)); | |
283 | } | |
284 | } | |
285 | return $new; | |
286 | } | |
287 | ||
5a964f20 | 288 | =head2 How can I make C<\w> match national character sets? |
d74e8afc | 289 | X<\w> |
68dc0745 | 290 | |
49d635f9 RGS |
291 | Put C<use locale;> in your script. The \w character class is taken |
292 | from the current locale. | |
293 | ||
294 | See L<perllocale> for details. | |
68dc0745 | 295 | |
296 | =head2 How can I match a locale-smart version of C</[a-zA-Z]/>? | |
d74e8afc | 297 | X<alpha> |
68dc0745 | 298 | |
49d635f9 RGS |
299 | You can use the POSIX character class syntax C</[[:alpha:]]/> |
300 | documented in L<perlre>. | |
301 | ||
302 | No matter which locale you are in, the alphabetic characters are | |
303 | the characters in \w without the digits and the underscore. | |
304 | As a regex, that looks like C</[^\W\d_]/>. Its complement, | |
197aec24 RGS |
305 | the non-alphabetics, is then everything in \W along with |
306 | the digits and the underscore, or C</[\W\d_]/>. | |
68dc0745 | 307 | |
d92eb7b0 | 308 | =head2 How can I quote a variable to use in a regex? |
d74e8afc | 309 | X<regex, escaping> X<regexp, escaping> X<regular expression, escaping> |
68dc0745 | 310 | |
311 | The Perl parser will expand $variable and @variable references in | |
312 | regular expressions unless the delimiter is a single quote. Remember, | |
79a522f5 | 313 | too, that the right-hand side of a C<s///> substitution is considered |
68dc0745 | 314 | a double-quoted string (see L<perlop> for more details). Remember |
d92eb7b0 | 315 | also that any regex special characters will be acted on unless you |
68dc0745 | 316 | precede the substitution with \Q. Here's an example: |
317 | ||
c83084d1 MJD |
318 | $string = "Placido P. Octopus"; |
319 | $regex = "P."; | |
68dc0745 | 320 | |
c83084d1 MJD |
321 | $string =~ s/$regex/Polyp/; |
322 | # $string is now "Polypacido P. Octopus" | |
68dc0745 | 323 | |
c83084d1 MJD |
324 | Because C<.> is special in regular expressions, and can match any |
325 | single character, the regex C<P.> here has matched the <Pl> in the | |
326 | original string. | |
327 | ||
328 | To escape the special meaning of C<.>, we use C<\Q>: | |
329 | ||
330 | $string = "Placido P. Octopus"; | |
331 | $regex = "P."; | |
332 | ||
333 | $string =~ s/\Q$regex/Polyp/; | |
334 | # $string is now "Placido Polyp Octopus" | |
335 | ||
336 | The use of C<\Q> causes the <.> in the regex to be treated as a | |
337 | regular character, so that C<P.> matches a C<P> followed by a dot. | |
68dc0745 | 338 | |
339 | =head2 What is C</o> really for? | |
d74e8afc | 340 | X</o> |
68dc0745 | 341 | |
46fc3d4c | 342 | Using a variable in a regular expression match forces a re-evaluation |
a6dd486b JB |
343 | (and perhaps recompilation) each time the regular expression is |
344 | encountered. The C</o> modifier locks in the regex the first time | |
345 | it's used. This always happens in a constant regular expression, and | |
346 | in fact, the pattern was compiled into the internal format at the same | |
347 | time your entire program was. | |
68dc0745 | 348 | |
349 | Use of C</o> is irrelevant unless variable interpolation is used in | |
d92eb7b0 | 350 | the pattern, and if so, the regex engine will neither know nor care |
68dc0745 | 351 | whether the variables change after the pattern is evaluated the I<very |
352 | first> time. | |
353 | ||
354 | C</o> is often used to gain an extra measure of efficiency by not | |
355 | performing subsequent evaluations when you know it won't matter | |
356 | (because you know the variables won't change), or more rarely, when | |
d92eb7b0 | 357 | you don't want the regex to notice if they do. |
68dc0745 | 358 | |
359 | For example, here's a "paragrep" program: | |
360 | ||
361 | $/ = ''; # paragraph mode | |
362 | $pat = shift; | |
363 | while (<>) { | |
364 | print if /$pat/o; | |
365 | } | |
366 | ||
367 | =head2 How do I use a regular expression to strip C style comments from a file? | |
368 | ||
369 | While this actually can be done, it's much harder than you'd think. | |
370 | For example, this one-liner | |
371 | ||
372 | perl -0777 -pe 's{/\*.*?\*/}{}gs' foo.c | |
373 | ||
374 | will work in many but not all cases. You see, it's too simple-minded for | |
375 | certain kinds of C programs, in particular, those with what appear to be | |
376 | comments in quoted strings. For that, you'd need something like this, | |
d92eb7b0 | 377 | created by Jeffrey Friedl and later modified by Fred Curtis. |
68dc0745 | 378 | |
379 | $/ = undef; | |
380 | $_ = <>; | |
c98c5709 | 381 | s#/\*[^*]*\*+([^/*][^*]*\*+)*/|("(\\.|[^"\\])*"|'(\\.|[^'\\])*'|.[^/"'\\]*)#defined $2 ? $2 : ""#gse; |
68dc0745 | 382 | print; |
383 | ||
384 | This could, of course, be more legibly written with the C</x> modifier, adding | |
d92eb7b0 GS |
385 | whitespace and comments. Here it is expanded, courtesy of Fred Curtis. |
386 | ||
387 | s{ | |
388 | /\* ## Start of /* ... */ comment | |
389 | [^*]*\*+ ## Non-* followed by 1-or-more *'s | |
390 | ( | |
391 | [^/*][^*]*\*+ | |
392 | )* ## 0-or-more things which don't start with / | |
393 | ## but do end with '*' | |
394 | / ## End of /* ... */ comment | |
395 | ||
396 | | ## OR various things which aren't comments: | |
397 | ||
398 | ( | |
399 | " ## Start of " ... " string | |
400 | ( | |
401 | \\. ## Escaped char | |
402 | | ## OR | |
403 | [^"\\] ## Non "\ | |
404 | )* | |
405 | " ## End of " ... " string | |
406 | ||
407 | | ## OR | |
408 | ||
409 | ' ## Start of ' ... ' string | |
410 | ( | |
411 | \\. ## Escaped char | |
412 | | ## OR | |
413 | [^'\\] ## Non '\ | |
414 | )* | |
415 | ' ## End of ' ... ' string | |
416 | ||
417 | | ## OR | |
418 | ||
419 | . ## Anything other char | |
420 | [^/"'\\]* ## Chars which doesn't start a comment, string or escape | |
421 | ) | |
c98c5709 | 422 | }{defined $2 ? $2 : ""}gxse; |
d92eb7b0 GS |
423 | |
424 | A slight modification also removes C++ comments: | |
425 | ||
c98c5709 | 426 | s#/\*[^*]*\*+([^/*][^*]*\*+)*/|//[^\n]*|("(\\.|[^"\\])*"|'(\\.|[^'\\])*'|.[^/"'\\]*)#defined $2 ? $2 : ""#gse; |
68dc0745 | 427 | |
428 | =head2 Can I use Perl regular expressions to match balanced text? | |
d74e8afc ITB |
429 | X<regex, matching balanced test> X<regexp, matching balanced test> |
430 | X<regular expression, matching balanced test> | |
68dc0745 | 431 | |
8305e449 JH |
432 | Historically, Perl regular expressions were not capable of matching |
433 | balanced text. As of more recent versions of perl including 5.6.1 | |
434 | experimental features have been added that make it possible to do this. | |
435 | Look at the documentation for the (??{ }) construct in recent perlre manual | |
436 | pages to see an example of matching balanced parentheses. Be sure to take | |
437 | special notice of the warnings present in the manual before making use | |
438 | of this feature. | |
439 | ||
440 | CPAN contains many modules that can be useful for matching text | |
441 | depending on the context. Damian Conway provides some useful | |
442 | patterns in Regexp::Common. The module Text::Balanced provides a | |
443 | general solution to this problem. | |
444 | ||
445 | One of the common applications of balanced text matching is working | |
446 | with XML and HTML. There are many modules available that support | |
447 | these needs. Two examples are HTML::Parser and XML::Parser. There | |
448 | are many others. | |
68dc0745 | 449 | |
450 | An elaborate subroutine (for 7-bit ASCII only) to pull out balanced | |
451 | and possibly nested single chars, like C<`> and C<'>, C<{> and C<}>, | |
452 | or C<(> and C<)> can be found in | |
a93751fa | 453 | http://www.cpan.org/authors/id/TOMC/scripts/pull_quotes.gz . |
68dc0745 | 454 | |
8305e449 | 455 | The C::Scan module from CPAN also contains such subs for internal use, |
68dc0745 | 456 | but they are undocumented. |
457 | ||
d92eb7b0 | 458 | =head2 What does it mean that regexes are greedy? How can I get around it? |
d74e8afc | 459 | X<greedy> X<greediness> |
68dc0745 | 460 | |
d92eb7b0 | 461 | Most people mean that greedy regexes match as much as they can. |
68dc0745 | 462 | Technically speaking, it's actually the quantifiers (C<?>, C<*>, C<+>, |
463 | C<{}>) that are greedy rather than the whole pattern; Perl prefers local | |
464 | greed and immediate gratification to overall greed. To get non-greedy | |
465 | versions of the same quantifiers, use (C<??>, C<*?>, C<+?>, C<{}?>). | |
466 | ||
467 | An example: | |
468 | ||
469 | $s1 = $s2 = "I am very very cold"; | |
470 | $s1 =~ s/ve.*y //; # I am cold | |
471 | $s2 =~ s/ve.*?y //; # I am very cold | |
472 | ||
473 | Notice how the second substitution stopped matching as soon as it | |
474 | encountered "y ". The C<*?> quantifier effectively tells the regular | |
475 | expression engine to find a match as quickly as possible and pass | |
476 | control on to whatever is next in line, like you would if you were | |
477 | playing hot potato. | |
478 | ||
f9ac83b8 | 479 | =head2 How do I process each word on each line? |
d74e8afc | 480 | X<word> |
68dc0745 | 481 | |
482 | Use the split function: | |
483 | ||
484 | while (<>) { | |
197aec24 | 485 | foreach $word ( split ) { |
68dc0745 | 486 | # do something with $word here |
197aec24 | 487 | } |
54310121 | 488 | } |
68dc0745 | 489 | |
54310121 | 490 | Note that this isn't really a word in the English sense; it's just |
491 | chunks of consecutive non-whitespace characters. | |
68dc0745 | 492 | |
f1cbbd6e GS |
493 | To work with only alphanumeric sequences (including underscores), you |
494 | might consider | |
68dc0745 | 495 | |
496 | while (<>) { | |
497 | foreach $word (m/(\w+)/g) { | |
498 | # do something with $word here | |
499 | } | |
500 | } | |
501 | ||
502 | =head2 How can I print out a word-frequency or line-frequency summary? | |
503 | ||
504 | To do this, you have to parse out each word in the input stream. We'll | |
54310121 | 505 | pretend that by word you mean chunk of alphabetics, hyphens, or |
506 | apostrophes, rather than the non-whitespace chunk idea of a word given | |
68dc0745 | 507 | in the previous question: |
508 | ||
509 | while (<>) { | |
510 | while ( /(\b[^\W_\d][\w'-]+\b)/g ) { # misses "`sheep'" | |
511 | $seen{$1}++; | |
54310121 | 512 | } |
513 | } | |
68dc0745 | 514 | while ( ($word, $count) = each %seen ) { |
515 | print "$count $word\n"; | |
54310121 | 516 | } |
68dc0745 | 517 | |
518 | If you wanted to do the same thing for lines, you wouldn't need a | |
519 | regular expression: | |
520 | ||
197aec24 | 521 | while (<>) { |
68dc0745 | 522 | $seen{$_}++; |
54310121 | 523 | } |
68dc0745 | 524 | while ( ($line, $count) = each %seen ) { |
525 | print "$count $line"; | |
526 | } | |
527 | ||
b432a672 AL |
528 | If you want these output in a sorted order, see L<perlfaq4>: "How do I |
529 | sort a hash (optionally by value instead of key)?". | |
68dc0745 | 530 | |
531 | =head2 How can I do approximate matching? | |
d74e8afc | 532 | X<match, approximate> X<matching, approximate> |
68dc0745 | 533 | |
534 | See the module String::Approx available from CPAN. | |
535 | ||
536 | =head2 How do I efficiently match many regular expressions at once? | |
d74e8afc ITB |
537 | X<regex, efficiency> X<regexp, efficiency> |
538 | X<regular expression, efficiency> | |
68dc0745 | 539 | |
7678cced RGS |
540 | ( contributed by brian d foy ) |
541 | ||
6670e5e7 | 542 | Avoid asking Perl to compile a regular expression every time |
7678cced RGS |
543 | you want to match it. In this example, perl must recompile |
544 | the regular expression for every iteration of the foreach() | |
545 | loop since it has no way to know what $pattern will be. | |
546 | ||
547 | @patterns = qw( foo bar baz ); | |
6670e5e7 RGS |
548 | |
549 | LINE: while( <> ) | |
7678cced | 550 | { |
6670e5e7 | 551 | foreach $pattern ( @patterns ) |
7678cced RGS |
552 | { |
553 | print if /\b$pattern\b/i; | |
554 | next LINE; | |
555 | } | |
556 | } | |
68dc0745 | 557 | |
7678cced RGS |
558 | The qr// operator showed up in perl 5.005. It compiles a |
559 | regular expression, but doesn't apply it. When you use the | |
560 | pre-compiled version of the regex, perl does less work. In | |
561 | this example, I inserted a map() to turn each pattern into | |
562 | its pre-compiled form. The rest of the script is the same, | |
563 | but faster. | |
564 | ||
565 | @patterns = map { qr/\b$_\b/i } qw( foo bar baz ); | |
566 | ||
6670e5e7 | 567 | LINE: while( <> ) |
7678cced | 568 | { |
6670e5e7 | 569 | foreach $pattern ( @patterns ) |
7678cced RGS |
570 | { |
571 | print if /\b$pattern\b/i; | |
572 | next LINE; | |
573 | } | |
574 | } | |
6670e5e7 | 575 | |
7678cced RGS |
576 | In some cases, you may be able to make several patterns into |
577 | a single regular expression. Beware of situations that require | |
578 | backtracking though. | |
65acb1b1 | 579 | |
7678cced RGS |
580 | $regex = join '|', qw( foo bar baz ); |
581 | ||
6670e5e7 | 582 | LINE: while( <> ) |
7678cced RGS |
583 | { |
584 | print if /\b(?:$regex)\b/i; | |
585 | } | |
586 | ||
587 | For more details on regular expression efficiency, see Mastering | |
588 | Regular Expressions by Jeffrey Freidl. He explains how regular | |
589 | expressions engine work and why some patterns are surprisingly | |
6670e5e7 | 590 | inefficient. Once you understand how perl applies regular |
7678cced | 591 | expressions, you can tune them for individual situations. |
68dc0745 | 592 | |
593 | =head2 Why don't word-boundary searches with C<\b> work for me? | |
d74e8afc | 594 | X<\b> |
68dc0745 | 595 | |
7678cced RGS |
596 | (contributed by brian d foy) |
597 | ||
598 | Ensure that you know what \b really does: it's the boundary between a | |
599 | word character, \w, and something that isn't a word character. That | |
600 | thing that isn't a word character might be \W, but it can also be the | |
601 | start or end of the string. | |
602 | ||
603 | It's not (not!) the boundary between whitespace and non-whitespace, | |
604 | and it's not the stuff between words we use to create sentences. | |
605 | ||
606 | In regex speak, a word boundary (\b) is a "zero width assertion", | |
607 | meaning that it doesn't represent a character in the string, but a | |
608 | condition at a certain position. | |
609 | ||
610 | For the regular expression, /\bPerl\b/, there has to be a word | |
611 | boundary before the "P" and after the "l". As long as something other | |
612 | than a word character precedes the "P" and succeeds the "l", the | |
613 | pattern will match. These strings match /\bPerl\b/. | |
614 | ||
615 | "Perl" # no word char before P or after l | |
616 | "Perl " # same as previous (space is not a word char) | |
617 | "'Perl'" # the ' char is not a word char | |
618 | "Perl's" # no word char before P, non-word char after "l" | |
619 | ||
620 | These strings do not match /\bPerl\b/. | |
621 | ||
622 | "Perl_" # _ is a word char! | |
623 | "Perler" # no word char before P, but one after l | |
6670e5e7 | 624 | |
7678cced | 625 | You don't have to use \b to match words though. You can look for |
d7f8936a | 626 | non-word characters surrounded by word characters. These strings |
7678cced RGS |
627 | match the pattern /\b'\b/. |
628 | ||
629 | "don't" # the ' char is surrounded by "n" and "t" | |
630 | "qep'a'" # the ' char is surrounded by "p" and "a" | |
6670e5e7 | 631 | |
7678cced | 632 | These strings do not match /\b'\b/. |
68dc0745 | 633 | |
7678cced | 634 | "foo'" # there is no word char after non-word ' |
6670e5e7 | 635 | |
7678cced RGS |
636 | You can also use the complement of \b, \B, to specify that there |
637 | should not be a word boundary. | |
68dc0745 | 638 | |
7678cced RGS |
639 | In the pattern /\Bam\B/, there must be a word character before the "a" |
640 | and after the "m". These patterns match /\Bam\B/: | |
68dc0745 | 641 | |
7678cced RGS |
642 | "llama" # "am" surrounded by word chars |
643 | "Samuel" # same | |
6670e5e7 | 644 | |
7678cced | 645 | These strings do not match /\Bam\B/ |
68dc0745 | 646 | |
7678cced RGS |
647 | "Sam" # no word boundary before "a", but one after "m" |
648 | "I am Sam" # "am" surrounded by non-word chars | |
68dc0745 | 649 | |
68dc0745 | 650 | |
651 | =head2 Why does using $&, $`, or $' slow my program down? | |
d74e8afc | 652 | X<$MATCH> X<$&> X<$POSTMATCH> X<$'> X<$PREMATCH> X<$`> |
68dc0745 | 653 | |
571e049f | 654 | (contributed by Anno Siegel) |
68dc0745 | 655 | |
571e049f | 656 | Once Perl sees that you need one of these variables anywhere in the |
b68463f7 RGS |
657 | program, it provides them on each and every pattern match. That means |
658 | that on every pattern match the entire string will be copied, part of it | |
659 | to $`, part to $&, and part to $'. Thus the penalty is most severe with | |
660 | long strings and patterns that match often. Avoid $&, $', and $` if you | |
661 | can, but if you can't, once you've used them at all, use them at will | |
662 | because you've already paid the price. Remember that some algorithms | |
663 | really appreciate them. As of the 5.005 release, the $& variable is no | |
664 | longer "expensive" the way the other two are. | |
665 | ||
666 | Since Perl 5.6.1 the special variables @- and @+ can functionally replace | |
667 | $`, $& and $'. These arrays contain pointers to the beginning and end | |
668 | of each match (see perlvar for the full story), so they give you | |
669 | essentially the same information, but without the risk of excessive | |
670 | string copying. | |
6670e5e7 | 671 | |
68dc0745 | 672 | =head2 What good is C<\G> in a regular expression? |
d74e8afc | 673 | X<\G> |
68dc0745 | 674 | |
49d635f9 RGS |
675 | You use the C<\G> anchor to start the next match on the same |
676 | string where the last match left off. The regular | |
677 | expression engine cannot skip over any characters to find | |
678 | the next match with this anchor, so C<\G> is similar to the | |
679 | beginning of string anchor, C<^>. The C<\G> anchor is typically | |
680 | used with the C<g> flag. It uses the value of pos() | |
681 | as the position to start the next match. As the match | |
682 | operator makes successive matches, it updates pos() with the | |
683 | position of the next character past the last match (or the | |
684 | first character of the next match, depending on how you like | |
685 | to look at it). Each string has its own pos() value. | |
686 | ||
687 | Suppose you want to match all of consective pairs of digits | |
688 | in a string like "1122a44" and stop matching when you | |
689 | encounter non-digits. You want to match C<11> and C<22> but | |
690 | the letter <a> shows up between C<22> and C<44> and you want | |
691 | to stop at C<a>. Simply matching pairs of digits skips over | |
692 | the C<a> and still matches C<44>. | |
693 | ||
694 | $_ = "1122a44"; | |
695 | my @pairs = m/(\d\d)/g; # qw( 11 22 44 ) | |
696 | ||
697 | If you use the \G anchor, you force the match after C<22> to | |
698 | start with the C<a>. The regular expression cannot match | |
699 | there since it does not find a digit, so the next match | |
700 | fails and the match operator returns the pairs it already | |
701 | found. | |
702 | ||
703 | $_ = "1122a44"; | |
704 | my @pairs = m/\G(\d\d)/g; # qw( 11 22 ) | |
705 | ||
706 | You can also use the C<\G> anchor in scalar context. You | |
707 | still need the C<g> flag. | |
708 | ||
709 | $_ = "1122a44"; | |
710 | while( m/\G(\d\d)/g ) | |
711 | { | |
712 | print "Found $1\n"; | |
713 | } | |
197aec24 | 714 | |
49d635f9 RGS |
715 | After the match fails at the letter C<a>, perl resets pos() |
716 | and the next match on the same string starts at the beginning. | |
717 | ||
718 | $_ = "1122a44"; | |
719 | while( m/\G(\d\d)/g ) | |
720 | { | |
721 | print "Found $1\n"; | |
722 | } | |
723 | ||
724 | print "Found $1 after while" if m/(\d\d)/g; # finds "11" | |
725 | ||
726 | You can disable pos() resets on fail with the C<c> flag. | |
727 | Subsequent matches start where the last successful match | |
728 | ended (the value of pos()) even if a match on the same | |
729 | string as failed in the meantime. In this case, the match | |
730 | after the while() loop starts at the C<a> (where the last | |
731 | match stopped), and since it does not use any anchor it can | |
732 | skip over the C<a> to find "44". | |
733 | ||
734 | $_ = "1122a44"; | |
735 | while( m/\G(\d\d)/gc ) | |
736 | { | |
737 | print "Found $1\n"; | |
738 | } | |
739 | ||
740 | print "Found $1 after while" if m/(\d\d)/g; # finds "44" | |
741 | ||
742 | Typically you use the C<\G> anchor with the C<c> flag | |
743 | when you want to try a different match if one fails, | |
744 | such as in a tokenizer. Jeffrey Friedl offers this example | |
745 | which works in 5.004 or later. | |
68dc0745 | 746 | |
747 | while (<>) { | |
748 | chomp; | |
749 | PARSER: { | |
49d635f9 RGS |
750 | m/ \G( \d+\b )/gcx && do { print "number: $1\n"; redo; }; |
751 | m/ \G( \w+ )/gcx && do { print "word: $1\n"; redo; }; | |
752 | m/ \G( \s+ )/gcx && do { print "space: $1\n"; redo; }; | |
753 | m/ \G( [^\w\d]+ )/gcx && do { print "other: $1\n"; redo; }; | |
68dc0745 | 754 | } |
755 | } | |
756 | ||
49d635f9 RGS |
757 | For each line, the PARSER loop first tries to match a series |
758 | of digits followed by a word boundary. This match has to | |
759 | start at the place the last match left off (or the beginning | |
197aec24 | 760 | of the string on the first match). Since C<m/ \G( \d+\b |
49d635f9 RGS |
761 | )/gcx> uses the C<c> flag, if the string does not match that |
762 | regular expression, perl does not reset pos() and the next | |
763 | match starts at the same position to try a different | |
764 | pattern. | |
68dc0745 | 765 | |
d92eb7b0 | 766 | =head2 Are Perl regexes DFAs or NFAs? Are they POSIX compliant? |
d74e8afc | 767 | X<DFA> X<NFA> X<POSIX> |
68dc0745 | 768 | |
769 | While it's true that Perl's regular expressions resemble the DFAs | |
770 | (deterministic finite automata) of the egrep(1) program, they are in | |
46fc3d4c | 771 | fact implemented as NFAs (non-deterministic finite automata) to allow |
68dc0745 | 772 | backtracking and backreferencing. And they aren't POSIX-style either, |
773 | because those guarantee worst-case behavior for all cases. (It seems | |
774 | that some people prefer guarantees of consistency, even when what's | |
775 | guaranteed is slowness.) See the book "Mastering Regular Expressions" | |
776 | (from O'Reilly) by Jeffrey Friedl for all the details you could ever | |
777 | hope to know on these matters (a full citation appears in | |
778 | L<perlfaq2>). | |
779 | ||
788611b6 | 780 | =head2 What's wrong with using grep in a void context? |
d74e8afc | 781 | X<grep> |
68dc0745 | 782 | |
788611b6 A |
783 | The problem is that grep builds a return list, regardless of the context. |
784 | This means you're making Perl go to the trouble of building a list that | |
785 | you then just throw away. If the list is large, you waste both time and space. | |
786 | If your intent is to iterate over the list, then use a for loop for this | |
f05bbc40 | 787 | purpose. |
68dc0745 | 788 | |
788611b6 A |
789 | In perls older than 5.8.1, map suffers from this problem as well. |
790 | But since 5.8.1, this has been fixed, and map is context aware - in void | |
791 | context, no lists are constructed. | |
792 | ||
54310121 | 793 | =head2 How can I match strings with multibyte characters? |
d74e8afc ITB |
794 | X<regex, and multibyte characters> X<regexp, and multibyte characters> |
795 | X<regular expression, and multibyte characters> | |
68dc0745 | 796 | |
d9d154f2 JH |
797 | Starting from Perl 5.6 Perl has had some level of multibyte character |
798 | support. Perl 5.8 or later is recommended. Supported multibyte | |
fe854a6f | 799 | character repertoires include Unicode, and legacy encodings |
d9d154f2 JH |
800 | through the Encode module. See L<perluniintro>, L<perlunicode>, |
801 | and L<Encode>. | |
802 | ||
803 | If you are stuck with older Perls, you can do Unicode with the | |
804 | C<Unicode::String> module, and character conversions using the | |
805 | C<Unicode::Map8> and C<Unicode::Map> modules. If you are using | |
806 | Japanese encodings, you might try using the jperl 5.005_03. | |
807 | ||
808 | Finally, the following set of approaches was offered by Jeffrey | |
809 | Friedl, whose article in issue #5 of The Perl Journal talks about | |
810 | this very matter. | |
68dc0745 | 811 | |
fc36a67e | 812 | Let's suppose you have some weird Martian encoding where pairs of |
813 | ASCII uppercase letters encode single Martian letters (i.e. the two | |
814 | bytes "CV" make a single Martian letter, as do the two bytes "SG", | |
815 | "VS", "XX", etc.). Other bytes represent single characters, just like | |
816 | ASCII. | |
68dc0745 | 817 | |
fc36a67e | 818 | So, the string of Martian "I am CVSGXX!" uses 12 bytes to encode the |
819 | nine characters 'I', ' ', 'a', 'm', ' ', 'CV', 'SG', 'XX', '!'. | |
68dc0745 | 820 | |
821 | Now, say you want to search for the single character C</GX/>. Perl | |
fc36a67e | 822 | doesn't know about Martian, so it'll find the two bytes "GX" in the "I |
823 | am CVSGXX!" string, even though that character isn't there: it just | |
824 | looks like it is because "SG" is next to "XX", but there's no real | |
825 | "GX". This is a big problem. | |
68dc0745 | 826 | |
827 | Here are a few ways, all painful, to deal with it: | |
828 | ||
b432a672 | 829 | $martian =~ s/([A-Z][A-Z])/ $1 /g; # Make sure adjacent "martian" |
49d635f9 | 830 | # bytes are no longer adjacent. |
68dc0745 | 831 | print "found GX!\n" if $martian =~ /GX/; |
832 | ||
833 | Or like this: | |
834 | ||
835 | @chars = $martian =~ m/([A-Z][A-Z]|[^A-Z])/g; | |
836 | # above is conceptually similar to: @chars = $text =~ m/(.)/g; | |
837 | # | |
838 | foreach $char (@chars) { | |
839 | print "found GX!\n", last if $char eq 'GX'; | |
840 | } | |
841 | ||
842 | Or like this: | |
843 | ||
844 | while ($martian =~ m/\G([A-Z][A-Z]|.)/gs) { # \G probably unneeded | |
54310121 | 845 | print "found GX!\n", last if $1 eq 'GX'; |
68dc0745 | 846 | } |
847 | ||
49d635f9 | 848 | Here's another, slightly less painful, way to do it from Benjamin |
c98c5709 | 849 | Goldberg, who uses a zero-width negative look-behind assertion. |
49d635f9 | 850 | |
c98c5709 RGS |
851 | print "found GX!\n" if $martian =~ m/ |
852 | (?<![A-Z]) | |
853 | (?:[A-Z][A-Z])*? | |
854 | GX | |
855 | /x; | |
197aec24 | 856 | |
49d635f9 | 857 | This succeeds if the "martian" character GX is in the string, and fails |
c98c5709 RGS |
858 | otherwise. If you don't like using (?<!), a zero-width negative |
859 | look-behind assertion, you can replace (?<![A-Z]) with (?:^|[^A-Z]). | |
49d635f9 RGS |
860 | |
861 | It does have the drawback of putting the wrong thing in $-[0] and $+[0], | |
862 | but this usually can be worked around. | |
68dc0745 | 863 | |
65acb1b1 TC |
864 | =head2 How do I match a pattern that is supplied by the user? |
865 | ||
866 | Well, if it's really a pattern, then just use | |
867 | ||
868 | chomp($pattern = <STDIN>); | |
869 | if ($line =~ /$pattern/) { } | |
870 | ||
a6dd486b | 871 | Alternatively, since you have no guarantee that your user entered |
65acb1b1 TC |
872 | a valid regular expression, trap the exception this way: |
873 | ||
874 | if (eval { $line =~ /$pattern/ }) { } | |
875 | ||
58103a2e | 876 | If all you really want is to search for a string, not a pattern, |
65acb1b1 | 877 | then you should either use the index() function, which is made for |
58103a2e | 878 | string searching, or, if you can't be disabused of using a pattern |
65acb1b1 TC |
879 | match on a non-pattern, then be sure to use C<\Q>...C<\E>, documented |
880 | in L<perlre>. | |
881 | ||
882 | $pattern = <STDIN>; | |
883 | ||
884 | open (FILE, $input) or die "Couldn't open input $input: $!; aborting"; | |
885 | while (<FILE>) { | |
886 | print if /\Q$pattern\E/; | |
887 | } | |
888 | close FILE; | |
889 | ||
500071f4 RGS |
890 | =head1 REVISION |
891 | ||
892 | Revision: $Revision: 3606 $ | |
893 | ||
894 | Date: $Date: 2006-03-06 12:05:47 +0100 (lun, 06 mar 2006) $ | |
895 | ||
896 | See L<perlfaq> for source control details and availability. | |
897 | ||
68dc0745 | 898 | =head1 AUTHOR AND COPYRIGHT |
899 | ||
58103a2e | 900 | Copyright (c) 1997-2006 Tom Christiansen, Nathan Torkington, and |
7678cced | 901 | other authors as noted. All rights reserved. |
5a964f20 | 902 | |
5a7beb56 JH |
903 | This documentation is free; you can redistribute it and/or modify it |
904 | under the same terms as Perl itself. | |
5a964f20 TC |
905 | |
906 | Irrespective of its distribution, all code examples in this file | |
907 | are hereby placed into the public domain. You are permitted and | |
908 | encouraged to use this code in your own programs for fun | |
909 | or for profit as you see fit. A simple comment in the code giving | |
910 | credit would be courteous but is not required. |