1 package Regexp::Wildcards;
7 use Text::Balanced qw<extract_bracketed>;
11 Regexp::Wildcards - Converts wildcard expressions to Perl regular expressions.
19 use vars qw<$VERSION>;
26 use Regexp::Wildcards;
28 my $rw = Regexp::Wildcards->new(type => 'unix');
31 $re = $rw->convert('a{b?,c}*'); # Do it Unix shell style.
32 $re = $rw->convert('a?,b*', 'win32'); # Do it Windows shell style.
33 $re = $rw->convert('*{x,y}?', 'jokers'); # Process the jokers and escape the rest.
34 $re = $rw->convert('%a_c%', 'sql'); # Turn SQL wildcards into regexps.
36 $rw = Regexp::Wildcards->new(
37 do => [ qw<jokers brackets> ], # Do jokers and brackets.
38 capture => [ qw<any greedy> ], # Capture *'s greedily.
41 $rw->do(add => 'groups'); # Don't escape groups.
42 $rw->capture(rem => [ qw<greedy> ]); # Actually we want non-greedy matches.
43 $re = $rw->convert('*a{,(b)?}?c*'); # '(.*?)a(?:|(b).).c(.*?)'
44 $rw->capture(); # No more captures.
48 In many situations, users may want to specify patterns to match but don't need the full power of regexps.
49 Wildcards make one of those sets of simplified rules.
50 This module converts wildcard expressions to Perl regular expressions, so that you can use them for matching.
52 It handles the C<*> and C<?> jokers, as well as Unix bracketed alternatives C<{,}>, but also C<%> and C<_> SQL wildcards.
53 If required, it can also keep original C<(...)> groups or C<^> and C<$> anchors.
54 Backspace (C<\>) is used as an escape character.
56 Typesets that mimic the behaviour of Windows and Unix shells are also provided.
63 croak 'First argument isn\'t a valid ' . __PACKAGE__ . ' object'
64 unless ref $_[0] and $_[0]->isa(__PACKAGE__);
68 jokers => [ qw<jokers> ],
70 commas => [ qw<commas> ],
71 brackets => [ qw<brackets> ],
72 unix => [ qw<jokers brackets> ],
73 win32 => [ qw<jokers commas> ],
75 $types{$_} = $types{win32} for qw<dos os2 MSWin32 cygwin>;
76 $types{$_} = $types{unix} for qw<linux
78 aix irix hpux dgux dynixptx
80 svr4 solaris sunos dec_osf
81 sco_sv unicos unicosmk>;
93 single => sub { $_[1] ? '(.)' : '.' },
94 any => sub { $_[1] ? ($_[0]->{greedy} ? '(.*)'
97 brackets => sub { $_[1] ? '(' : '(?:'; },
106 $old = { } unless defined $old;
110 $opts{set} = defined $_[0] ? $_[0] : { };
112 croak 'Arguments must be passed as an unique scalar or as key => value pairs';
118 for (qw<set add rem>) {
120 next unless defined $opt;
122 '' => sub { +{ ($_[0] => 1) x (exists $valid->{$_[0]}) } },
123 'ARRAY' => sub { +{ map { ($_ => 1) x (exists $valid->{$_}) } @{$_[0]} } },
124 'HASH' => sub { +{ map { ($_ => $_[0]->{$_}) x (exists $valid->{$_}) }
127 croak 'Wrong option set' unless $cb;
128 $checked{$_} = $cb->($opt);
131 my $config = (exists $checked{set}) ? $checked{set} : $old;
132 $config->{$_} = $checked{add}->{$_} for grep $checked{add}->{$_},
133 keys %{$checked{add} || {}};
134 delete $config->{$_} for grep $checked{rem}->{$_},
135 keys %{$checked{rem} || {}};
143 $config->{do} = $self->_validate(\%escapes, $self->{do}, @_);
144 $config->{escape} = '';
145 $config->{escape} .= $escapes{$_} for keys %{$config->{do}};
146 $config->{escape} = quotemeta $config->{escape};
153 my $config = $self->_do(@_);
154 $self->{$_} = $config->{$_} for keys %$config;
161 $config->{capture} = $self->_validate(\%captures, $self->{capture}, @_);
162 $config->{greedy} = delete $config->{capture}->{greedy};
163 for (keys %captures) {
164 $config->{'c_' . $_} = $captures{$_}->($config, $config->{capture}->{$_})
165 if $captures{$_}; # Skip 'greedy'
173 my $config = $self->_capture(@_);
174 $self->{$_} = $config->{$_} for keys %$config;
179 my ($self, $type) = @_;
180 $type = 'unix' unless defined $type;
181 croak 'Wrong type' unless exists $types{$type};
182 my $config = $self->_do($types{$type});
183 $config->{type} = $type;
190 my $config = $self->_type(@_);
191 $self->{$_} = $config->{$_} for keys %$config;
197 $class = ref($class) || $class || __PACKAGE__;
198 croak 'Optional arguments must be passed as key => value pairs' if @_ % 2;
202 if (defined $args{do}) {
203 $self->do($args{do});
205 $self->type($args{type});
207 $self->capture($args{capture});
210 =head2 C<< new [ do => $what E<verbar> type => $type ], capture => $captures >>
212 Constructs a new L<Regexp::Wildcard> object.
214 C<do> lists all features that should be enabled when converting wildcards to regexps.
215 Refer to L</do> for details on what can be passed in C<$what>.
217 The C<type> specifies a predefined set of C<do> features to use.
218 See L</type> for details on which types are valid.
219 The C<do> option overrides C<type>.
221 C<capture> lists which atoms should be capturing.
222 Refer to L</capture> for more details.
224 =head2 C<< do [ $what E<verbar> set => $c1, add => $c2, rem => $c3 ] >>
226 Specifies the list of metacharacters to convert or to prevent for escaping.
227 They fit into six classes :
235 Converts C<?> to C<.> and C<*> to C<.*>.
237 'a**\\*b??\\?c' ==> 'a.*\\*b..\\?c'
243 Converts C<_> to C<.> and C<%> to C<.*>.
245 'a%%\\%b__\\_c' ==> 'a.*\\%b..\\_c'
251 Converts all C<,> to C<|> and puts the complete resulting regular expression inside C<(?: ... )>.
253 'a,b{c,d},e' ==> '(?:a|b\\{c|d\\}|e)'
259 Converts all matching C<{ ... , ... }> brackets to C<(?: ... | ... )> alternations.
260 If some brackets are unbalanced, it tries to substitute as many of them as possible, and then escape the remaining unmatched C<{> and C<}>.
261 Commas outside of any bracket-delimited block are also escaped.
263 'a,b{c,d},e' ==> 'a\\,b(?:c|d)\\,e'
264 '{a\\{b,c}d,e}' ==> '(?:a\\{b|c)d\\,e\\}'
265 '{a{b,c\\}d,e}' ==> '\\{a\\{b\\,c\\}d\\,e\\}'
271 Keeps the parenthesis C<( ... )> of the original string without escaping them.
272 Currently, no check is done to ensure that the parenthesis are matching.
274 'a(b(c))d\\(\\)' ==> (no change)
280 Prevents the I<beginning-of-line> C<^> and I<end-of-line> C<$> anchors to be escaped.
281 Since C<[...]> character class are currently escaped, a C<^> will always be interpreted as I<beginning-of-line>.
283 'a^b$c' ==> (no change)
287 Each C<$c> can be any of :
293 A hash reference, with wanted metacharacter group names (described above) as keys and booleans as values ;
297 An array reference containing the list of wanted metacharacter classes ;
301 A plain scalar, when only one group is required.
305 When C<set> is present, the classes given as its value replace the current object options.
306 Then the C<add> classes are added, and the C<rem> classes removed.
308 Passing a sole scalar C<$what> is equivalent as passing C<< set => $what >>.
309 No argument means C<< set => [ ] >>.
311 $rw->do(set => 'jokers'); # Only translate jokers.
312 $rw->do('jokers'); # Same.
313 $rw->do(add => [ qw<sql commas> ]); # Translate also SQL and commas.
314 $rw->do(rem => 'jokers'); # Specifying both 'sql' and 'jokers' is useless.
315 $rw->do(); # Translate nothing.
317 The C<do> method returns the L<Regexp::Wildcards> object.
321 Notifies to convert the metacharacters that corresponds to the predefined type C<$type>.
322 C<$type> can be any of :
328 C<'jokers'>, C<'sql'>, C<'commas'>, C<'brackets'>
330 Singleton types that enable the corresponding C<do> classes.
336 Covers typical Unix shell globbing features (effectively C<'jokers'> and C<'brackets'>).
340 C<$^O> values for common Unix systems
342 Wrap to C<'unix'> (see L<perlport> for the list).
348 Defaults to C<'unix'>.
354 Covers typical Windows shell globbing features (effectively C<'jokers'> and C<'commas'>).
358 C<'dos'>, C<'os2'>, C<'MSWin32'>, C<'cygwin'>
364 In particular, you can usually pass C<$^O> as the C<$type> and get the corresponding shell behaviour.
366 $rw->type('win32'); # Set type to win32.
367 $rw->type($^O); # Set type to unix on Unices and win32 on Windows
368 $rw->type(); # Set type to unix.
370 The C<type> method returns the L<Regexp::Wildcards> object.
372 =head2 C<< capture [ $captures E<verbar> set => $c1, add => $c2, rem => $c3 ] >>
374 Specifies the list of atoms to capture.
375 This method works like L</do>, except that the classes are different :
383 Captures all unescaped I<"exactly one"> metacharacters, i.e. C<?> for wildcards or C<_> for SQL.
385 'a???b\\??' ==> 'a(.)(.)(.)b\\?(.)'
386 'a___b\\__' ==> 'a(.)(.)(.)b\\_(.)'
392 Captures all unescaped I<"any"> metacharacters, i.e. C<*> for wildcards or C<%> for SQL.
394 'a***b\\**' ==> 'a(.*)b\\*(.*)'
395 'a%%%b\\%%' ==> 'a(.*)b\\%(.*)'
401 When used in conjunction with C<'any'>, it makes the C<'any'> captures greedy (by default they are not).
403 'a***b\\**' ==> 'a(.*?)b\\*(.*?)'
404 'a%%%b\\%%' ==> 'a(.*?)b\\%(.*?)'
410 Capture matching C<{ ... , ... }> alternations.
412 'a{b\\},\\{c}' ==> 'a(b\\}|\\{c)'
416 $rw->capture(set => 'single'); # Only capture "exactly one" metacharacters.
417 $rw->capture('single'); # Same.
418 $rw->capture(add => [ qw<any greedy> ]); # Also greedily capture "any" metacharacters.
419 $rw->capture(rem => 'greedy'); # No more greed please.
420 $rw->capture(); # Capture nothing.
422 The C<capture> method returns the L<Regexp::Wildcards> object.
424 =head2 C<convert $wc [ , $type ]>
426 Converts the wildcard expression C<$wc> into a regular expression according to the options stored into the L<Regexp::Wildcards> object, or to C<$type> if it's supplied.
427 It successively escapes all unprotected regexp special characters that doesn't hold any meaning for wildcards, then replace C<'jokers'>, C<'sql'> and C<'commas'> or C<'brackets'> (depending on the L</do> or L</type> options), all of this by applying the C<'capture'> rules specified in the constructor or by L</capture>.
432 my ($self, $wc, $type) = @_;
434 my $config = (defined $type) ? $self->_type($type) : $self;
435 return unless defined $wc;
437 my $e = $config->{escape};
439 # - an even number of \ that doesn't protect a regexp/wildcard metachar
440 # - an odd number of \ that doesn't protect a wildcard metachar
448 (?: [^\W$e] | \s | $ )
453 my $do = $config->{do};
454 $wc = $self->_jokers($wc) if $do->{jokers};
455 $wc = $self->_sql($wc) if $do->{sql};
456 if ($do->{brackets}) {
457 $wc = $self->_bracketed($wc);
458 } elsif ($do->{commas} and $wc =~ /(?<!\\)(?:\\\\)*,/) {
459 $wc = $self->{'c_brackets'} . $self->_commas($wc) . ')';
467 An object module shouldn't export any function, and so does this one.
471 L<Carp> (core module since perl 5), L<Text::Balanced> (since 5.7.3).
475 This module does not implement the strange behaviours of Windows shell that result from the special handling of the three last characters (for the file extension).
476 For example, Windows XP shell matches C<*a> like C<.*a>, C<*a?> like C<.*a.?>, C<*a??> like C<.*a.{0,2}> and so on.
484 Vincent Pit, C<< <perl at profvince.com> >>, L<http://www.profvince.com>.
486 You can contact me by mail or on C<irc.perl.org> (vincent).
490 Please report any bugs or feature requests to C<bug-regexp-wildcards at rt.cpan.org>, or through the web interface at L<http://rt.cpan.org/NoAuth/ReportBug.html?Queue=Regexp-Wildcards>. I will be notified, and then you'll automatically be notified of progress on your bug as I make changes.
494 You can find documentation for this module with the perldoc command.
496 perldoc Regexp::Wildcards
498 Tests code coverage report is available at L<http://www.profvince.com/perl/cover/Regexp-Wildcards>.
500 =head1 COPYRIGHT & LICENSE
502 Copyright 2007-2009 Vincent Pit, all rights reserved.
504 This program is free software; you can redistribute it and/or modify it
505 under the same terms as Perl itself.
509 sub _extract ($) { extract_bracketed $_[0], '{', qr/.*?(?<!\\)(?:\\\\)*(?={)/ }
514 # substitute ? preceded by an even number of \
515 my $s = $self->{c_single};
516 s/(?<!\\)((?:\\\\)*)\?/$1$s/g;
517 # substitute * preceded by an even number of \
519 s/(?<!\\)((?:\\\\)*)\*+/$1$s/g;
526 # substitute _ preceded by an even number of \
527 my $s = $self->{c_single};
528 s/(?<!\\)((?:\\\\)*)_/$1$s/g;
529 # substitute % preceded by an even number of \
531 s/(?<!\\)((?:\\\\)*)%+/$1$s/g;
537 # substitute , preceded by an even number of \
538 s/(?<!\\)((?:\\\\)*),/$1|/g;
543 my ($self, $rest) = @_;
544 substr $rest, 0, 1, '';
546 my ($re, $bracket, $prefix) = ('');
547 while (do { ($bracket, $rest, $prefix) = _extract $rest; $bracket }) {
548 $re .= $self->_commas($prefix) . $self->_brackets($bracket);
550 $re .= $self->_commas($rest);
551 return $self->{c_brackets} . $re . ')';
555 my ($self, $rest) = @_;
556 my ($re, $bracket, $prefix) = ('');
557 while (do { ($bracket, $rest, $prefix) = _extract $rest; $bracket }) {
558 $re .= $prefix . $self->_brackets($bracket);
561 $re =~ s/(?<!\\)((?:\\\\)*[\{\},])/\\$1/g;
565 1; # End of Regexp::Wildcards