1 package Regexp::Wildcards;
7 use Scalar::Util qw<blessed>;
8 use Text::Balanced qw<extract_bracketed>;
12 Regexp::Wildcards - Converts wildcard expressions to Perl regular expressions.
20 use vars qw<$VERSION>;
27 use Regexp::Wildcards;
29 my $rw = Regexp::Wildcards->new(type => 'unix');
32 $re = $rw->convert('a{b?,c}*'); # Do it Unix shell style.
33 $re = $rw->convert('a?,b*', 'win32'); # Do it Windows shell style.
34 $re = $rw->convert('*{x,y}?', 'jokers'); # Process the jokers and
36 $re = $rw->convert('%a_c%', 'sql'); # Turn SQL wildcards into
39 $rw = Regexp::Wildcards->new(
40 do => [ qw<jokers brackets> ], # Do jokers and brackets.
41 capture => [ qw<any greedy> ], # Capture *'s greedily.
44 $rw->do(add => 'groups'); # Don't escape groups.
45 $rw->capture(rem => [ qw<greedy> ]); # Actually we want non-greedy
47 $re = $rw->convert('*a{,(b)?}?c*'); # '(.*?)a(?:|(b).).c(.*?)'
48 $rw->capture(); # No more captures.
52 In many situations, users may want to specify patterns to match but don't need the full power of regexps.
53 Wildcards make one of those sets of simplified rules.
54 This module converts wildcard expressions to Perl regular expressions, so that you can use them for matching.
56 It handles the C<*> and C<?> jokers, as well as Unix bracketed alternatives C<{,}>, but also C<%> and C<_> SQL wildcards.
57 If required, it can also keep original C<(...)> groups or C<^> and C<$> anchors.
58 Backspace (C<\>) is used as an escape character.
60 Typesets that mimic the behaviour of Windows and Unix shells are also provided.
67 croak 'First argument isn\'t a valid ' . __PACKAGE__ . ' object'
68 unless blessed $_[0] and $_[0]->isa(__PACKAGE__);
72 jokers => [ qw<jokers> ],
74 commas => [ qw<commas> ],
75 brackets => [ qw<brackets> ],
76 unix => [ qw<jokers brackets> ],
77 win32 => [ qw<jokers commas> ],
79 $types{$_} = $types{win32} for qw<dos os2 MSWin32 cygwin>;
80 $types{$_} = $types{unix} for qw<linux
82 aix irix hpux dgux dynixptx
84 svr4 solaris sunos dec_osf
85 sco_sv unicos unicosmk>;
97 single => sub { $_[1] ? '(.)' : '.' },
98 any => sub { $_[1] ? ($_[0]->{greedy} ? '(.*)'
101 brackets => sub { $_[1] ? '(' : '(?:'; },
110 $old = { } unless defined $old;
114 $opts{set} = defined $_[0] ? $_[0] : { };
116 croak 'Arguments must be passed as an unique scalar or as key => value pairs';
122 for (qw<set add rem>) {
124 next unless defined $opt;
127 '' => sub { +{ ($_[0] => 1) x (exists $valid->{$_[0]}) } },
128 'ARRAY' => sub { +{ map { ($_ => 1) x (exists $valid->{$_}) } @{$_[0]} } },
129 'HASH' => sub { +{ map { ($_ => $_[0]->{$_}) x (exists $valid->{$_}) }
132 croak 'Wrong option set' unless $cb;
133 $checked{$_} = $cb->($opt);
136 my $config = (exists $checked{set}) ? $checked{set} : $old;
137 $config->{$_} = $checked{add}->{$_} for grep $checked{add}->{$_},
138 keys %{$checked{add} || {}};
139 delete $config->{$_} for grep $checked{rem}->{$_},
140 keys %{$checked{rem} || {}};
149 $config->{do} = $self->_validate(\%escapes, $self->{do}, @_);
150 $config->{escape} = '';
151 $config->{escape} .= $escapes{$_} for keys %{$config->{do}};
152 $config->{escape} = quotemeta $config->{escape};
161 my $config = $self->_do(@_);
162 $self->{$_} = $config->{$_} for keys %$config;
171 $config->{capture} = $self->_validate(\%captures, $self->{capture}, @_);
172 $config->{greedy} = delete $config->{capture}->{greedy};
173 for (keys %captures) {
174 $config->{'c_' . $_} = $captures{$_}->($config, $config->{capture}->{$_})
175 if $captures{$_}; # Skip 'greedy'
185 my $config = $self->_capture(@_);
186 $self->{$_} = $config->{$_} for keys %$config;
192 my ($self, $type) = @_;
193 $type = 'unix' unless defined $type;
194 croak 'Wrong type' unless exists $types{$type};
196 my $config = $self->_do($types{$type});
197 $config->{type} = $type;
206 my $config = $self->_type(@_);
207 $self->{$_} = $config->{$_} for keys %$config;
214 $class = blessed($class) || $class || __PACKAGE__;
216 croak 'Optional arguments must be passed as key => value pairs' if @_ % 2;
219 my $self = bless { }, $class;
221 if (defined $args{do}) {
222 $self->do($args{do});
224 $self->type($args{type});
227 $self->capture($args{capture});
232 my $rw = Regexp::Wildcards->new(do => $what, capture => $capture);
233 my $rw = Regexp::Wildcards->new(type => $type, capture => $capture);
235 Constructs a new L<Regexp::Wildcard> object.
237 C<do> lists all features that should be enabled when converting wildcards to regexps.
238 Refer to L</do> for details on what can be passed in C<$what>.
240 The C<type> specifies a predefined set of C<do> features to use.
241 See L</type> for details on which types are valid.
242 The C<do> option overrides C<type>.
244 C<capture> lists which atoms should be capturing.
245 Refer to L</capture> for more details.
254 Specifies the list of metacharacters to convert or to prevent for escaping.
255 They fit into six classes :
263 Converts C<?> to C<.> and C<*> to C<.*>.
265 'a**\\*b??\\?c' ==> 'a.*\\*b..\\?c'
271 Converts C<_> to C<.> and C<%> to C<.*>.
273 'a%%\\%b__\\_c' ==> 'a.*\\%b..\\_c'
279 Converts all C<,> to C<|> and puts the complete resulting regular expression inside C<(?: ... )>.
281 'a,b{c,d},e' ==> '(?:a|b\\{c|d\\}|e)'
287 Converts all matching C<{ ... , ... }> brackets to C<(?: ... | ... )> alternations.
288 If some brackets are unbalanced, it tries to substitute as many of them as possible, and then escape the remaining unmatched C<{> and C<}>.
289 Commas outside of any bracket-delimited block are also escaped.
291 'a,b{c,d},e' ==> 'a\\,b(?:c|d)\\,e'
292 '{a\\{b,c}d,e}' ==> '(?:a\\{b|c)d\\,e\\}'
293 '{a{b,c\\}d,e}' ==> '\\{a\\{b\\,c\\}d\\,e\\}'
299 Keeps the parenthesis C<( ... )> of the original string without escaping them.
300 Currently, no check is done to ensure that the parenthesis are matching.
302 'a(b(c))d\\(\\)' ==> (no change)
308 Prevents the I<beginning-of-line> C<^> and I<end-of-line> C<$> anchors to be escaped.
309 Since C<[...]> character class are currently escaped, a C<^> will always be interpreted as I<beginning-of-line>.
311 'a^b$c' ==> (no change)
315 Each C<$c> can be any of :
321 A hash reference, with wanted metacharacter group names (described above) as keys and booleans as values ;
325 An array reference containing the list of wanted metacharacter classes ;
329 A plain scalar, when only one group is required.
333 When C<set> is present, the classes given as its value replace the current object options.
334 Then the C<add> classes are added, and the C<rem> classes removed.
336 Passing a sole scalar C<$what> is equivalent as passing C<< set => $what >>.
337 No argument means C<< set => [ ] >>.
339 $rw->do(set => 'jokers'); # Only translate jokers.
340 $rw->do('jokers'); # Same.
341 $rw->do(add => [ qw<sql commas> ]); # Translate also SQL and commas.
342 $rw->do(rem => 'jokers'); # Specifying both 'sql' and
343 # 'jokers' is useless.
344 $rw->do(); # Translate nothing.
346 The C<do> method returns the L<Regexp::Wildcards> object.
352 Notifies to convert the metacharacters that corresponds to the predefined type C<$type>.
353 C<$type> can be any of :
359 C<'jokers'>, C<'sql'>, C<'commas'>, C<'brackets'>
361 Singleton types that enable the corresponding C<do> classes.
367 Covers typical Unix shell globbing features (effectively C<'jokers'> and C<'brackets'>).
371 C<$^O> values for common Unix systems
373 Wrap to C<'unix'> (see L<perlport> for the list).
379 Defaults to C<'unix'>.
385 Covers typical Windows shell globbing features (effectively C<'jokers'> and C<'commas'>).
389 C<'dos'>, C<'os2'>, C<'MSWin32'>, C<'cygwin'>
395 In particular, you can usually pass C<$^O> as the C<$type> and get the corresponding shell behaviour.
397 $rw->type('win32'); # Set type to win32.
398 $rw->type($^O); # Set type to unix on Unices and win32 on Windows
399 $rw->type(); # Set type to unix.
401 The C<type> method returns the L<Regexp::Wildcards> object.
405 $rw->capture($captures);
406 $rw->capture(set => $c1);
407 $rw->capture(add => $c2);
408 $rw->capture(rem => $c3);
410 Specifies the list of atoms to capture.
411 This method works like L</do>, except that the classes are different :
419 Captures all unescaped I<"exactly one"> metacharacters, i.e. C<?> for wildcards or C<_> for SQL.
421 'a???b\\??' ==> 'a(.)(.)(.)b\\?(.)'
422 'a___b\\__' ==> 'a(.)(.)(.)b\\_(.)'
428 Captures all unescaped I<"any"> metacharacters, i.e. C<*> for wildcards or C<%> for SQL.
430 'a***b\\**' ==> 'a(.*)b\\*(.*)'
431 'a%%%b\\%%' ==> 'a(.*)b\\%(.*)'
437 When used in conjunction with C<'any'>, it makes the C<'any'> captures greedy (by default they are not).
439 'a***b\\**' ==> 'a(.*?)b\\*(.*?)'
440 'a%%%b\\%%' ==> 'a(.*?)b\\%(.*?)'
446 Capture matching C<{ ... , ... }> alternations.
448 'a{b\\},\\{c}' ==> 'a(b\\}|\\{c)'
452 $rw->capture(set => 'single'); # Only capture "exactly one"
454 $rw->capture('single'); # Same.
455 $rw->capture(add => [ qw<any greedy> ]); # Also greedily capture
456 # "any" metacharacters.
457 $rw->capture(rem => 'greedy'); # No more greed please.
458 $rw->capture(); # Capture nothing.
460 The C<capture> method returns the L<Regexp::Wildcards> object.
464 my $rx = $rw->convert($wc);
465 my $rx = $rw->convert($wc, $type);
467 Converts the wildcard expression C<$wc> into a regular expression according to the options stored into the L<Regexp::Wildcards> object, or to C<$type> if it's supplied.
468 It successively escapes all unprotected regexp special characters that doesn't hold any meaning for wildcards, then replace C<'jokers'>, C<'sql'> and C<'commas'> or C<'brackets'> (depending on the L</do> or L</type> options), all of this by applying the C<'capture'> rules specified in the constructor or by L</capture>.
473 my ($self, $wc, $type) = @_;
476 my $config = (defined $type) ? $self->_type($type) : $self;
477 return unless defined $wc;
479 my $e = $config->{escape};
481 # - an even number of \ that doesn't protect a regexp/wildcard metachar
482 # - an odd number of \ that doesn't protect a wildcard metachar
490 (?: [^\W$e] | \s | $ )
495 my $do = $config->{do};
496 $wc = $self->_jokers($wc) if $do->{jokers};
497 $wc = $self->_sql($wc) if $do->{sql};
498 if ($do->{brackets}) {
499 $wc = $self->_bracketed($wc);
500 } elsif ($do->{commas} and $wc =~ /(?<!\\)(?:\\\\)*,/) {
501 $wc = $self->{'c_brackets'} . $self->_commas($wc) . ')';
509 An object module shouldn't export any function, and so does this one.
513 L<Carp> (core module since perl 5), L<Scalar::Util>, L<Text::Balanced> (since 5.7.3).
517 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).
518 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.
526 Vincent Pit, C<< <perl at profvince.com> >>, L<http://www.profvince.com>.
528 You can contact me by mail or on C<irc.perl.org> (vincent).
532 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>.
533 I will be notified, and then you'll automatically be notified of progress on your bug as I make changes.
537 You can find documentation for this module with the perldoc command.
539 perldoc Regexp::Wildcards
541 Tests code coverage report is available at L<http://www.profvince.com/perl/cover/Regexp-Wildcards>.
543 =head1 COPYRIGHT & LICENSE
545 Copyright 2007,2008,2009,2013 Vincent Pit, all rights reserved.
547 This program is free software; you can redistribute it and/or modify it
548 under the same terms as Perl itself.
552 sub _extract ($) { extract_bracketed $_[0], '{', qr/.*?(?<!\\)(?:\\\\)*(?={)/ }
558 # substitute ? preceded by an even number of \
559 my $s = $self->{c_single};
560 s/(?<!\\)((?:\\\\)*)\?/$1$s/g;
561 # substitute * preceded by an even number of \
563 s/(?<!\\)((?:\\\\)*)\*+/$1$s/g;
572 # substitute _ preceded by an even number of \
573 my $s = $self->{c_single};
574 s/(?<!\\)((?:\\\\)*)_/$1$s/g;
575 # substitute % preceded by an even number of \
577 s/(?<!\\)((?:\\\\)*)%+/$1$s/g;
585 # substitute , preceded by an even number of \
586 s/(?<!\\)((?:\\\\)*),/$1|/g;
592 my ($self, $rest) = @_;
594 substr $rest, 0, 1, '';
597 my ($re, $bracket, $prefix) = ('');
598 while (do { ($bracket, $rest, $prefix) = _extract $rest; $bracket }) {
599 $re .= $self->_commas($prefix) . $self->_brackets($bracket);
601 $re .= $self->_commas($rest);
603 $self->{c_brackets} . $re . ')';
607 my ($self, $rest) = @_;
609 my ($re, $bracket, $prefix) = ('');
610 while (do { ($bracket, $rest, $prefix) = _extract $rest; $bracket }) {
611 $re .= $prefix . $self->_brackets($bracket);
615 $re =~ s/(?<!\\)((?:\\\\)*[\{\},])/\\$1/g;
620 1; # End of Regexp::Wildcards