=head1 NAME
-Regexp::Wildcards - Converts wildcards expressions to Perl regular expressions.
+Regexp::Wildcards - Converts wildcard expressions to Perl regular expressions.
=head1 VERSION
-Version 0.01
+Version 0.05
=cut
-our $VERSION = '0.01';
+our $VERSION = '0.05';
=head1 SYNOPSIS
use Regexp::Wildcards qw/wc2re/;
my $re;
- $re = wc2re 'a{b.,c}*' => 'unix'; # Do it Unix style.
- $re = wc2re 'a.,b*' => 'win32'; # Do it Windows style.
- $re = wc2re '*{x,y}.' => 'jokers'; # Process the jokers & escape the rest.
+ $re = wc2re 'a{b?,c}*' => 'unix'; # Do it Unix style.
+ $re = wc2re 'a?,b*' => 'win32'; # Do it Windows style.
+ $re = wc2re '*{x,y}?' => 'jokers'; # Process the jokers & escape the rest.
=head1 DESCRIPTION
-In many situations, users may want to specify patterns to match but don't need the full power of regexps. Wildcards make one of those sets of simplified rules. This module converts wildcards expressions to Perl regular expressions, so that you can use them for matching. It handles the C<*> and C<?> jokers, as well as Unix bracketed alternatives C<{,}>, and uses the backspace (C<\>) as an escape character. Wrappers are provided to mimic the behaviour of Windows and Unix shells.
+In many situations, users may want to specify patterns to match but don't need the full power of regexps. Wildcards make one of those sets of simplified rules. This module converts wildcard expressions to Perl regular expressions, so that you can use them for matching. It handles the C<*> and C<?> jokers, as well as Unix bracketed alternatives C<{,}>, and uses the backspace (C<\>) as an escape character. Wrappers are provided to mimic the behaviour of Windows and Unix shells.
-=head1 EXPORT
+=head1 VARIABLES
+
+These variables control if the wildcards jokers and brackets must capture their match. They can be globally set by writing in your program
+
+ $Regexp::Wildcards::CaptureSingle = 1;
+ # From then, the '?' joker is capturing
+
+or can be locally specified via C<local>
+
+ {
+ local $Regexp::Wildcards::CaptureAny = 1;
+ # In this block, the '?' joker is capturing.
+ ...
+ }
+ # Back to the situation from before the block
-Four functions are exported only on request : C<wc2re>, C<wc2re_unix>, C<wc2re_win32> and C<wc2re_jokers>.
+This section describes also how those elements are translated by the L<functions|/FUNCTIONS>.
+
+=head2 C<$CaptureSingle>
+
+When this variable is true, each occurence of the unescaped C<?> joker is made capturing in the resulting regexp (they are be replaced by C<(.)>). Otherwise, they are just replaced by C<.>. Default is the latter.
+
+ 'a???b\\??' is translated to 'a(.)(.)(.)b\\?(.)' if $CaptureSingle is true
+ 'a...b\\?.' otherwise (default)
=cut
-use base qw/Exporter/;
+our $CaptureSingle = 0;
-my %types = (
- 'jokers' => \&wc2re_jokers,
- 'unix' => \&wc2re_unix,
- 'win32' => \&wc2re_win32
-);
+=head2 C<$CaptureAny>
-our @EXPORT = ();
-our @EXPORT_OK = ('wc2re', map { 'wc2re_' . $_ } keys %types);
-our @EXPORT_FAIL = qw/extract do_jokers do_commas do_brackets do_bracketed/;
-our %EXPORT_TAGS = ( all => [ @EXPORT_OK ] );
+By default this variable is false, and successions of unescaped C<*> jokers are replaced by B<one> single C<.*>. When it evalutes to true, those sequences of C<*> are made into B<one> capture, which is greedy (C<(.*)>) for C<$CaptureAny E<gt> 0> and otherwise non-greedy (C<(.*?)>).
+
+ 'a***b\\**' is translated to 'a.*b\\*.*' if $CaptureAny is false (default)
+ 'a(.*)b\\*(.*)' if $CaptureAny > 0
+ 'a(.*?)b\\*(.*?)' otherwise
+
+=cut
+
+our $CaptureAny = 0;
+
+=head2 C<$CaptureBrackets>
+
+If this variable is set to true, valid brackets constructs are made into C<( | )> captures, and otherwise they are replaced by non-capturing alternations (C<(?: | >)), which is the default.
+
+ 'a{b\\},\\{c}' is translated to 'a(b\\}|\\{c)' if $CaptureBrackets is true
+ 'a(?:b\\}|\\{c)' otherwise (default)
+
+=cut
+
+our $CaptureBrackets = 0;
=head1 FUNCTIONS
+=head2 C<wc2re_jokers>
+
+This function takes as its only argument the wildcard string to process, and returns the corresponding regular expression where the jokers C<?> and C<*> have been translated into their regexp equivalents (see L</VARIABLES> for more details). All other unprotected regexp metacharacters are escaped.
+
+ # Everything is escaped.
+ print 'ok' if wc2re_jokers('{a{b,c}d,e}') eq '\\{a\\{b\\,c\\}d\\,e\\}';
+
+=cut
+
+sub wc2re_jokers {
+ my ($wc) = @_;
+ $wc =~ s/(?<!\\)((?:\\\\)*[^\w\s?*\\])/\\$1/g;
+ return do_jokers($wc);
+}
+
=head2 C<wc2re_unix>
-This function takes as its only argument the wildcard string to process, and returns the corresponding regular expression (or C<undef> if the source is invalid) according to standard Unix wildcard rules. It successively escapes all unprotected regexp special characters that doesn't hold any meaning for wildcards, turns jokers into their regexp equivalents, and changes bracketed blocks into C<(?:|)> alternations. If brackets are unbalanced, it will try to substitute as many of them as possible, and then escape the remaining C<{> and C<}>.
+Similar to the precedent, but this one conforms to standard Unix shell wildcard rules. It successively escapes all unprotected regexp special characters that doesn't hold any meaning for wildcards, turns jokers into their regexp equivalents (see L</wc2re_jokers>), and changes bracketed blocks into (possibly capturing) alternations as described in L</VARIABLES>. If brackets are unbalanced, it tries to substitute as many of them as possible, and then escape the remaining C<{> and C<}>. Commas outside of any bracket-delimited block are also escaped.
+
+ # This is a valid bracket expression, and is completely translated.
+ print 'ok' if wc2re_unix('{a{b,c}d,e}') eq '(?:a(?:b|c)d|e)';
-Unbalanced bracket expressions can always be rescued, but it may change completely its meaning. As a side effect, commas that first appear to be between brackets can be taken at the uppermost level, which invalidates the pattern. For example :
+The function handles unbalanced bracket expressions, by escaping everything it can't recognize. For example :
- # The last orphaned } gets escaped, and the first comma is replaced.
- # We also need to escape the comma because unix doesn't allow them out
- # of brackets.
- print 'ok' if wc2re_unix('{a\\{b,c}d\\,e}') eq '(?:a\\{b|c)d\\,e\\}';
+ # The first comma is replaced, and the remaining brackets and comma are escaped.
+ print 'ok' if wc2re_unix('{a\\{b,c}d,e}') eq '(?:a\\{b|c)d\\,e\\}';
- # All of the unprotected brackets are escaped, which means that we must
- # escape all the commas.
- print 'ok' if wc2re_unix('{a{b\\,c\\}d\\,e}') eq '\\{a\\{b\\,c\\}d\\,e\\}';
+ # All the brackets and commas are escaped.
+ print 'ok' if wc2re_unix('{a{b,c\\}d,e}') eq '\\{a\\{b\\,c\\}d\\,e\\}';
=cut
=head2 C<wc2re_win32>
-Similar to the precedent, but for Windows wildcards. Bracketed blocks are no longer handled (which means that brackets will be escaped), but you can still provide a comma-separated list of items.
+This one works just like the two before, but for Windows wildcards. Bracketed blocks are no longer handled (which means that brackets are escaped), but you can provide a comma-separated list of items.
+
+ # All the brackets are escaped, and commas are seen as list delimiters.
+ print 'ok' if wc2re_win32('{a{b,c}d,e}') eq '(?:\\{a\\{b|c\\}d|e\\})';
=cut
$wc =~ s/(?<!\\)((?:\\\\)*[^\w\s?*\\,])/\\$1/g;
my $re = do_jokers($wc);
if ($re =~ /(?<!\\)(?:\\\\)*,/) { # win32 allows comma-separated lists
- $re = '(?:' . do_commas($re) . ')';
+ $re = ($CaptureBrackets ? '(' : '(?:') . do_commas($re) . ')';
}
return $re;
}
-=head2 C<wc2re_jokers>
+=head2 C<wc2re>
-This one only handles the C<?> and C<*> jokers. All other unquoted regexp metacharacters will be quoted.
+A generic function that wraps around all the different rules. The first argument is the wildcard expression, and the second one is the type of rules to apply which can be :
-=cut
+=over 4
-sub wc2re_jokers {
- my ($wc) = @_;
- $wc =~ s/(?<!\\)((?:\\\\)*[^\w\s?*\\])/\\$1/g;
- return do_jokers($wc);
-}
+=item C<'unix'>, C<'win32'>, C<'jokers'>
-=head2 C<wc2re>
+For one of those raw rule names, C<wc2re> simply maps to C<wc2re_unix>, C<wc2re_win32> and C<wc2re_jokers> respectively.
+
+=item C<$^O>
+
+If you supply the Perl operating system name, the call is deferred to C<wc2re_win32> for C< $^O> equal to C<'dos'>, C<'os2'>, C<'MSWin32'> or C<'cygwin'>, and to C<wc2re_unix> in all the other cases.
+
+=back
+
+If the type is undefined or not supported, it defaults to C<'unix'>.
+
+ # Wraps to wc2re_jokers ($re eq 'a\\{b\\,c\\}.*').
+ $re = wc2re 'a{b,c}*' => 'jokers';
-A generic function that wraps around all the different rules. The first argument is the wildcard expression, and the second one is the type of rules to apply, currently either C<unix>, C<win32> or C<jokers>. If the type is undefined, it defaults to C<unix>.
+ # Wraps to wc2re_win32 ($re eq '(?:a\\{b|c\\}.*)')
+ # or wc2re_unix ($re eq 'a(?:b|c).*') depending on $^O.
+ $re = wc2re 'a{b,c}*' => $^O;
=cut
+my %types = (
+ 'jokers' => \&wc2re_jokers,
+ 'unix' => \&wc2re_unix,
+ map { lc $_ => \&wc2re_win32 } qw/win32 dos os2 MSWin32 cygwin/
+);
+
sub wc2re {
my ($wc, $type) = @_;
return unless defined $wc;
- $type ||= 'unix';
- return $types{lc $type}($wc);
+ $type = $type ? lc $type : 'unix';
+ $type = 'unix' unless exists $types{$type};
+ return $types{$type}($wc);
}
+=head1 EXPORT
+
+These four functions are exported only on request : C<wc2re>, C<wc2re_unix>, C<wc2re_win32> and C<wc2re_jokers>. The variables are not exported.
+
+=cut
+
+use base qw/Exporter/;
+
+our @EXPORT = ();
+our @EXPORT_OK = ('wc2re', map { 'wc2re_' . $_ } keys %types);
+our @EXPORT_FAIL = qw/extract do_jokers do_commas do_brackets do_bracketed/;
+our %EXPORT_TAGS = ( all => [ @EXPORT_OK ] );
+
=head1 DEPENDENCIES
L<Text::Balanced>, which is bundled with perl since version 5.7.3
+=head1 CAVEATS
+
+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). 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.
+
=head1 SEE ALSO
Some modules provide incomplete alternatives as helper functions :
L<Net::FTPServer> has a method for that. Only jokers are translated, and escaping won't preserve them.
-L<File::Find::Match::Util> has a C<wildcar> function that compiles a matcher. Only handles C<*>.
+L<File::Find::Match::Util> has a C<wildcard> function that compiles a matcher. It only handles C<*>.
L<Text::Buffer> has the C<convertWildcardToRegex> class method that handles jokers.
=cut
-sub extract { extract_bracketed shift, '{', qr/.*?(?:(?<!\\)(?:\\\\)*)(?={)/; }
+sub extract { extract_bracketed shift, '{', qr/.*?(?<!\\)(?:\\\\)*(?={)/; }
sub do_jokers {
local $_ = shift;
# escape an odd number of \ that doesn't protect a regexp/wildcard special char
s/(?<!\\)((?:\\\\)*\\(?:[\w\s]|$))/\\$1/g;
# substitute ? preceded by an even number of \
- s/(?<!\\)((?:\\\\)*)\?/$1./g;
+ my $s = $CaptureSingle ? '(.)' : '.';
+ s/(?<!\\)((?:\\\\)*)\?/$1$s/g;
# substitute * preceded by an even number of \
- s/(?<!\\)((?:\\\\)*)\*+/$1.*/g;
+ $s = $CaptureAny ? (($CaptureAny > 0) ? '(.*)' : '(.*?)')
+ : '.*';
+ s/(?<!\\)((?:\\\\)*)\*+/$1$s/g;
return $_;
}
$re .= do_commas($prefix) . do_brackets($bracket);
}
$re .= do_commas($rest);
- return '(?:' . $re . ')';
+ return ($CaptureBrackets ? '(' : '(?:') . $re . ')';
}
sub do_bracketed {
my $rest = shift;
my ($re, $bracket, $prefix) = ('');
while (($bracket, $rest, $prefix) = extract $rest and $bracket) {
- return undef if $prefix =~ /(?<!\\)((?:\\\\)*),/;
$re .= $prefix . do_brackets($bracket);
}
- return undef if $rest =~ /(?<!\\)((?:\\\\)*),/;
$re .= $rest;
- $re =~ s/(?<!\\)((?:\\\\)*[\{\}])/\\$1/g;
+ $re =~ s/(?<!\\)((?:\\\\)*[\{\},])/\\$1/g;
return $re;
}
projects / perl / modules / Regexp-Wildcards.git / blobdiff