2 Regexp::Wildcards - Converts wildcard expressions to Perl regular
11 my $rw = Regexp::Wildcards->new(type => 'unix');
14 $re = $rw->convert('a{b?,c}*'); # Do it Unix shell style.
15 $re = $rw->convert('a?,b*', 'win32'); # Do it Windows shell style.
16 $re = $rw->convert('*{x,y}?', 'jokers'); # Process the jokers and escape the rest.
17 $re = $rw->convert('%a_c%', 'sql'); # Turn SQL wildcards into regexps.
19 $rw = Regexp::Wildcards->new(
20 do => [ qw/jokers brackets/ ], # Do jokers and brackets.
21 capture => [ qw/any greedy/ ], # Capture *'s greedily.
24 $rw->do(add => 'groups'); # Don't escape groups.
25 $rw->capture(rem => [ qw/greedy/ ]); # Actually we want non-greedy matches.
26 $re = $rw->convert('*a{,(b)?}?c*'); # '(.*?)a(?:|(b).).c(.*?)'
27 $rw->capture(); # No more captures.
30 In many situations, users may want to specify patterns to match but
31 don't need the full power of regexps. Wildcards make one of those sets
32 of simplified rules. This module converts wildcard expressions to Perl
33 regular expressions, so that you can use them for matching.
35 It handles the "*" and "?" jokers, as well as Unix bracketed
36 alternatives "{,}", but also "%" and "_" SQL wildcards. If required, it
37 can also keep original "(...)" groups or "^" and "$" anchors. Backspace
38 ("\") is used as an escape character.
40 Typesets that mimic the behaviour of Windows and Unix shells are also
44 "new [ do => $what | type => $type ], capture => $captures"
45 Constructs a new Regexp::Wildcard object.
47 "do" lists all features that should be enabled when converting wildcards
48 to regexps. Refer to "do" for details on what can be passed in $what.
50 The "type" specifies a predefined set of "do" features to use. See
51 "type" for details on which types are valid. The "do" option overrides
54 "capture" lists which atoms should be capturing. Refer to "capture" for
57 "do [ $what | set => $c1, add => $c2, rem => $c3 ]"
58 Specifies the list of metacharacters to convert or to prevent for
59 escaping. They fit into six classes :
61 * 'jokers' converts "?" to "." and "*" to ".*" ;
63 'a**\\*b??\\?c' ==> 'a.*\\*b..\\?c'
65 * 'sql' converts "_" to "." and "%" to ".*" ;
67 'a%%\\%b__\\_c' ==> 'a.*\\%b..\\_c'
69 * 'commas' converts all "," to "|" and puts the complete resulting
70 regular expression inside "(?: ... )" ;
72 'a,b{c,d},e' ==> '(?:a|b\\{c|d\\}|e)'
74 * 'brackets' converts all matching "{ ... , ... }" brackets to "(?:
75 ... | ... )" alternations. If some brackets are unbalanced, it tries
76 to substitute as many of them as possible, and then escape the
77 remaining unmatched "{" and "}". Commas outside of any
78 bracket-delimited block are also escaped ;
80 'a,b{c,d},e' ==> 'a\\,b(?:c|d)\\,e'
81 '{a\\{b,c}d,e}' ==> '(?:a\\{b|c)d\\,e\\}'
82 '{a{b,c\\}d,e}' ==> '\\{a\\{b\\,c\\}d\\,e\\}'
84 * 'groups' keeps the parenthesis "( ... )" of the original string
85 without escaping them. Currently, no check is done to ensure that
86 the parenthesis are matching.
88 'a(b(c))d\\(\\)' ==> (no change)
90 * 'anchors' prevents the *beginning-of-line* "^" and *end-of-line* "$"
91 anchors to be escaped. Since "[...]" character class are currently
92 escaped, a "^" will always be interpreted as *beginning-of-line*.
94 'a^b$c' ==> (no change)
96 Each $c can be any of :
98 * A hash reference, with wanted metacharacter group names (described
99 above) as keys and booleans as values ;
101 * An array reference containing the list of wanted metacharacter
104 * A plain scalar, when only one group is required.
106 When "set" is present, the classes given as its value replace the
107 current object options. Then the "add" classes are added, and the "rem"
110 Passing a sole scalar $what is equivalent as passing "set => $what". No
111 argument means "set => [ ]".
113 $rw->do(set => 'jokers'); # Only translate jokers.
114 $rw->do('jokers'); # Same.
115 $rw->do(add => [ qw/sql commas/ ]); # Translate also SQL and commas.
116 $rw->do(rem => 'jokers'); # Specifying both 'sql' and 'jokers' is useless.
117 $rw->do(); # Translate nothing.
120 Notifies to convert the metacharacters that corresponds to the
121 predefined type $type. $type can be any of 'jokers', 'sql', 'commas',
122 'brackets', 'win32' or 'unix'. An unknown or undefined value defaults to
123 'unix', except for 'dos', 'os2', 'MSWin32' and 'cygwin' that default to
124 'win32'. This means that you can pass $^O as the $type and get the
125 corresponding shell behaviour. Returns the object.
127 $rw->type('win32'); # Set type to win32.
128 $rw->type(); # Set type to unix.
130 "capture [ $captures | set => $c1, add => $c2, rem => $c3 ]"
131 Specifies the list of atoms to capture. This method works like "do",
132 except that the classes are different :
134 * 'single' will capture all unescaped *"exactly one"* metacharacters,
135 i.e. "?" for wildcards or "_" for SQL ;
137 'a???b\\??' ==> 'a(.)(.)(.)b\\?(.)'
138 'a___b\\__' ==> 'a(.)(.)(.)b\\_(.)'
140 * 'any' will capture all unescaped *"any"* metacharacters, i.e. "*"
141 for wildcards or "%" for SQL ;
143 'a***b\\**' ==> 'a(.*)b\\*(.*)'
144 'a%%%b\\%%' ==> 'a(.*)b\\%(.*)'
146 * 'greedy', when used in conjunction with 'any', will make the 'any'
147 captures greedy (by default they are not) ;
149 'a***b\\**' ==> 'a(.*?)b\\*(.*?)'
150 'a%%%b\\%%' ==> 'a(.*?)b\\%(.*?)'
152 * 'brackets' will capture matching "{ ... , ... }" alternations.
154 'a{b\\},\\{c}' ==> 'a(b\\}|\\{c)'
156 $rw->capture(set => 'single'); # Only capture "exactly one" metacharacters.
157 $rw->capture('single'); # Same.
158 $rw->capture(add => [ qw/any greedy/ ]); # Also greedily capture "any" metacharacters.
159 $rw->capture(rem => 'greedy'); # No more greed please.
160 $rw->capture(); # Capture nothing.
162 "convert $wc [ , $type ]"
163 Converts the wildcard expression $wc into a regular expression according
164 to the options stored into the Regexp::Wildcards object, or to $type if
165 it's supplied. It successively escapes all unprotected regexp special
166 characters that doesn't hold any meaning for wildcards, then replace
167 'jokers' or 'sql' and 'commas' or 'brackets' (depending on the "do" or
168 "type" options), all of this by applying the 'capture' rules specified
169 in the constructor or by "capture".
172 An object module shouldn't export any function, and so does this one.
175 Carp (core module since perl 5), Text::Balanced (since 5.7.3).
178 This module does not implement the strange behaviours of Windows shell
179 that result from the special handling of the three last characters (for
180 the file extension). For example, Windows XP shell matches *a like
181 ".*a", "*a?" like ".*a.?", "*a??" like ".*a.{0,2}" and so on.
184 Vincent Pit, "<perl at profvince.com>", <http://www.profvince.com>.
186 You can contact me by mail or on #perl @ FreeNode (vincent or
190 Please report any bugs or feature requests to "bug-regexp-wildcards at
191 rt.cpan.org", or through the web interface at
192 <http://rt.cpan.org/NoAuth/ReportBug.html?Queue=Regexp-Wildcards>. I
193 will be notified, and then you'll automatically be notified of progress
194 on your bug as I make changes.
197 You can find documentation for this module with the perldoc command.
199 perldoc Regexp::Wildcards
201 Tests code coverage report is available at
202 <http://www.profvince.com/perl/cover/Regexp-Wildcards>.
205 Copyright 2007-2008 Vincent Pit, all rights reserved.
207 This program is free software; you can redistribute it and/or modify it
208 under the same terms as Perl itself.