=head1 VERSION
-Version 0.08
+Version 1.02
=cut
-our $VERSION = '0.08';
+use vars qw/$VERSION/;
+BEGIN {
+ $VERSION = '1.02';
+}
=head1 SYNOPSIS
$re = $rw->convert('a{b?,c}*'); # Do it Unix shell style.
$re = $rw->convert('a?,b*', 'win32'); # Do it Windows shell style.
$re = $rw->convert('*{x,y}?', 'jokers'); # Process the jokers and escape the rest.
- $re = $rw->convert('%a_c%', 'sql'); # Turn SQL wildcards into regexps.
+ $re = $rw->convert('%a_c%', 'sql'); # Turn SQL wildcards into regexps.
$rw = Regexp::Wildcards->new(
do => [ qw/jokers brackets/ ], # Do jokers and brackets.
=head1 DESCRIPTION
-In many situations, users may want to specify patterns to match but don't need the full power of regexps. Wildcards make one of those sets of simplified rules. This module converts wildcard expressions to Perl regular expressions, so that you can use them for matching.
+In many situations, users may want to specify patterns to match but don't need the full power of regexps.
+Wildcards make one of those sets of simplified rules.
+This module converts wildcard expressions to Perl regular expressions, so that you can use them for matching.
-It handles the C<*> and C<?> shell jokers, as well as Unix bracketed alternatives C<{,}>, but also C<%> and C<_> SQL wildcards. It can also keep original C<(...)> groups. Backspace (C<\>) is used as an escape character.
+It handles the C<*> and C<?> jokers, as well as Unix bracketed alternatives C<{,}>, but also C<%> and C<_> SQL wildcards.
+If required, it can also keep original C<(...)> groups or C<^> and C<$> anchors.
+Backspace (C<\>) is used as an escape character.
Typesets that mimic the behaviour of Windows and Unix shells are also provided.
win32 => [ qw/jokers commas/ ],
);
$types{$_} = $types{win32} for qw/dos os2 MSWin32 cygwin/;
+$types{$_} = $types{unix} for qw/linux
+ darwin machten next
+ aix irix hpux dgux dynixptx
+ bsdos freebsd openbsd
+ svr4 solaris sunos dec_osf
+ sco_sv unicos unicosmk/;
my %escapes = (
jokers => '?*',
- sql => '%_',
+ sql => '_%',
commas => ',',
brackets => '{},',
groups => '()',
+ anchors => '^$',
);
my %captures = (
my $valid = shift;
my $old = shift;
$old = { } unless defined $old;
- my $c;
+
+ my %opts;
if (@_ <= 1) {
- $c = { set => $_[0] };
+ $opts{set} = defined $_[0] ? $_[0] : { };
} elsif (@_ % 2) {
croak 'Arguments must be passed as an unique scalar or as key => value pairs';
} else {
- my %args = @_;
- $c = { map { (exists $args{$_}) ? ($_ => $args{$_}) : () } qw/set add rem/ };
+ %opts = @_;
}
+
+ my %checked;
for (qw/set add rem/) {
- my $v = $c->{$_};
- next unless defined $v;
+ my $opt = $opts{$_};
+ next unless defined $opt;
my $cb = {
'' => sub { +{ ($_[0] => 1) x (exists $valid->{$_[0]}) } },
'ARRAY' => sub { +{ map { ($_ => 1) x (exists $valid->{$_}) } @{$_[0]} } },
'HASH' => sub { +{ map { ($_ => $_[0]->{$_}) x (exists $valid->{$_}) }
keys %{$_[0]} } }
- }->{ ref $v };
+ }->{ ref $opt };
croak 'Wrong option set' unless $cb;
- $c->{$_} = $cb->($v);
+ $checked{$_} = $cb->($opt);
}
- my $config = (exists $c->{set}) ? $c->{set} : $old;
- $config->{$_} = $c->{add}->{$_} for grep $c->{add}->{$_},
- keys %{$c->{add} || {}};
- delete $config->{$_} for grep $c->{rem}->{$_}, keys %{$c->{rem} || {}};
+
+ my $config = (exists $checked{set}) ? $checked{set} : $old;
+ $config->{$_} = $checked{add}->{$_} for grep $checked{add}->{$_},
+ keys %{$checked{add} || {}};
+ delete $config->{$_} for grep $checked{rem}->{$_},
+ keys %{$checked{rem} || {}};
+
$config;
}
$self->capture($args{capture});
}
-=head2 C<< new [ do => $what | type => $type ], capture => $captures >>
+=head2 C<< new [ do => $what E<verbar> type => $type ], capture => $captures >>
Constructs a new L<Regexp::Wildcard> object.
-C<do> lists all features that should be enabled when converting wildcards to regexps. Refer to L</do> for details on what can be passed in C<$what>.
+C<do> lists all features that should be enabled when converting wildcards to regexps.
+Refer to L</do> for details on what can be passed in C<$what>.
-The C<type> specifies a predefined set of C<do> features to use.
+The C<type> specifies a predefined set of C<do> features to use.
+See L</type> for details on which types are valid.
+The C<do> option overrides C<type>.
+C<capture> lists which atoms should be capturing.
+Refer to L</capture> for more details.
-C<$type> can be any of C<'jokers'>, C<'sql'>, C<'commas'>, C<'brackets'>, C<'win32'> or C<'unix'>. An unknown value defaults to C<'unix'>, except for C<'dos'>, C<'os2'>, C<'MSWin32'> and C<'cygwin'> that default to C<'win32'>. With this set of options, you can pass C<$^O> as the C<$type> so that you get the corresponding shell behaviour.
+=head2 C<< do [ $what E<verbar> set => $c1, add => $c2, rem => $c3 ] >>
+
+Specifies the list of metacharacters to convert or to prevent for escaping.
+They fit into six classes :
=over 4
-=item C<>
+=item *
-For the C<$capture> syntax, refer to the L</capture> method.
+C<'jokers'>
-=head3 C<type>
+Converts C<?> to C<.> and C<*> to C<.*>.
-=head3 C<capture>
+ 'a**\\*b??\\?c' ==> 'a.*\\*b..\\?c'
-=over 4
+=item *
+
+C<'sql'>
+
+Converts C<_> to C<.> and C<%> to C<.*>.
-=head2 C<$CaptureSingle>
+ 'a%%\\%b__\\_c' ==> 'a.*\\%b..\\_c'
-When this variable is true, each occurence of unescaped I<"exactly one"> wildcards (i.e. C<?> jokers or C<_> for SQL wildcards) are made capturing in the resulting regexp (they are be replaced by C<(.)>). Otherwise, they are just replaced by C<.>. Default is the latter.
+=item *
- For jokers :
- 'a???b\\??' is translated to 'a(.)(.)(.)b\\?(.)' if $CaptureSingle is true
- 'a...b\\?.' otherwise (default)
+C<'commas'>
- For SQL wildcards :
- 'a___b\\__' is translated to 'a(.)(.)(.)b\\_(.)' if $CaptureSingle is true
- 'a...b\\_.' otherwise (default)
+Converts all C<,> to C<|> and puts the complete resulting regular expression inside C<(?: ... )>.
+ 'a,b{c,d},e' ==> '(?:a|b\\{c|d\\}|e)'
-=item C<any>
+=item *
-By default this variable is false, and successions of unescaped I<"any"> wildcards (i.e. C<*> jokers or C<%> for SQL wildcards) are replaced by B<one> single C<.*>. When it evalutes to true, those sequences of I<"any"> wildcards are made into B<one> capture, which is greedy (C<(.*)>) for C<$CaptureAny E<gt> 0> and otherwise non-greedy (C<(.*?)>).
+C<'brackets'>
- For jokers :
- 'a***b\\**' is translated to 'a.*b\\*.*' if $CaptureAny is false (default)
- 'a(.*)b\\*(.*)' if $CaptureAny > 0
- 'a(.*?)b\\*(.*?)' otherwise
+Converts all matching C<{ ... , ... }> brackets to C<(?: ... | ... )> alternations.
+If some brackets are unbalanced, it tries to substitute as many of them as possible, and then escape the remaining unmatched C<{> and C<}>.
+Commas outside of any bracket-delimited block are also escaped.
- For SQL wildcards :
- 'a%%%b\\%%' is translated to 'a.*b\\%.*' if $CaptureAny is false (default)
- 'a(.*)b\\%(.*)' if $CaptureAny > 0
- 'a(.*?)b\\%(.*?)' otherwise
+ 'a,b{c,d},e' ==> 'a\\,b(?:c|d)\\,e'
+ '{a\\{b,c}d,e}' ==> '(?:a\\{b|c)d\\,e\\}'
+ '{a{b,c\\}d,e}' ==> '\\{a\\{b\\,c\\}d\\,e\\}'
-=item C<brackets>
+=item *
-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.
+C<'groups'>
- 'a{b\\},\\{c}' is translated to 'a(b\\}|\\{c)' if $CaptureBrackets is true
- 'a(?:b\\}|\\{c)' otherwise (default)
+Keeps the parenthesis C<( ... )> of the original string without escaping them.
+Currently, no check is done to ensure that the parenthesis are matching.
+
+ 'a(b(c))d\\(\\)' ==> (no change)
+
+=item *
+
+C<'anchors'>
+
+Prevents the I<beginning-of-line> C<^> and I<end-of-line> C<$> anchors to be escaped.
+Since C<[...]> character class are currently escaped, a C<^> will always be interpreted as I<beginning-of-line>.
+
+ 'a^b$c' ==> (no change)
=back
-=head2 C<jokers>
+Each C<$c> can be any of :
-This function takes as its only argument the wildcard string to process, and returns the corresponding regular expression where the jokers C<?> (I<"exactly one">) and C<*> (I<"any">) have been translated into their regexp equivalents (see L</VARIABLES> for more details). All other unprotected regexp metacharacters are escaped.
+=over 4
- # Everything is escaped.
- print 'ok' if wc2re_jokers('{a{b,c}d,e}') eq '\\{a\\{b\\,c\\}d\\,e\\}';
+=item *
-=cut
+A hash reference, with wanted metacharacter group names (described above) as keys and booleans as values ;
-=head2 C<sql>
+=item *
-Similar to the precedent, but for the SQL wildcards C<_> (I<"exactly one">) and C<%> (I<"any">). All other unprotected regexp metacharacters are escaped.
-
-=cut
-
-=head2 C<shell>
+An array reference containing the list of wanted metacharacter classes ;
-This function conforms to standard Unix shell wildcard rules. It successively escapes all unprotected regexp special characters that doesn't hold any meaning for wildcards, turns C<?> and C<*> 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.
+=item *
- # 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)';
+A plain scalar, when only one group is required.
-The function handles unbalanced bracket expressions, by escaping everything it can't recognize. For example :
+=back
- # 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\\}';
+When C<set> is present, the classes given as its value replace the current object options.
+Then the C<add> classes are added, and the C<rem> classes removed.
- # All the brackets and commas are escaped.
- print 'ok' if wc2re_unix('{a{b,c\\}d,e}') eq '\\{a\\{b\\,c\\}d\\,e\\}';
+Passing a sole scalar C<$what> is equivalent as passing C<< set => $what >>.
+No argument means C<< set => [ ] >>.
-This one works just like the one 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.
+ $rw->do(set => 'jokers'); # Only translate jokers.
+ $rw->do('jokers'); # Same.
+ $rw->do(add => [ qw/sql commas/ ]); # Translate also SQL and commas.
+ $rw->do(rem => 'jokers'); # Specifying both 'sql' and 'jokers' is useless.
+ $rw->do(); # Translate nothing.
- # 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\\})';
+The C<do> method returns the L<Regexp::Wildcards> object.
-=cut
+=head2 C<type $type>
+
+Notifies to convert the metacharacters that corresponds to the predefined type C<$type>.
+C<$type> can be any of :
+
+=over 4
+
+=item *
+
+C<'jokers'>, C<'sql'>, C<'commas'>, C<'brackets'>
+
+Singleton types that enable the corresponding C<do> classes.
+
+=item *
+
+C<'unix'>
+
+Covers typical Unix shell globbing features (effectively C<'jokers'> and C<'brackets'>).
+
+=item *
+
+C<$^O> values for common Unix systems
+
+Wrap to C<'unix'> (see L<perlport> for the list).
+
+=item *
+
+C<undef>
+
+Defaults to C<'unix'>.
-=head2 C<convert>
+=item *
-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 :
+C<'win32'>
+
+Covers typical Windows shell globbing features (effectively C<'jokers'> and C<'commas'>).
+
+=item *
+
+C<'dos'>, C<'os2'>, C<'MSWin32'>, C<'cygwin'>
+
+Wrap to C<'win32'>.
+
+=back
+
+In particular, you can usually pass C<$^O> as the C<$type> and get the corresponding shell behaviour.
+
+ $rw->type('win32'); # Set type to win32.
+ $rw->type($^O); # Set type to unix on Unices and win32 on Windows
+ $rw->type(); # Set type to unix.
+
+The C<type> method returns the L<Regexp::Wildcards> object.
+
+=head2 C<< capture [ $captures E<verbar> set => $c1, add => $c2, rem => $c3 ] >>
+
+Specifies the list of atoms to capture.
+This method works like L</do>, except that the classes are different :
=over 4
-=item C<'unix'>, C<'win32'>, C<'jokers'>, C<'sql'>
+=item *
+
+C<'single'>
+
+Captures all unescaped I<"exactly one"> metacharacters, i.e. C<?> for wildcards or C<_> for SQL.
+
+ 'a???b\\??' ==> 'a(.)(.)(.)b\\?(.)'
+ 'a___b\\__' ==> 'a(.)(.)(.)b\\_(.)'
+
+=item *
+
+C<'any'>
+
+Captures all unescaped I<"any"> metacharacters, i.e. C<*> for wildcards or C<%> for SQL.
-For one of those raw rule names, C<wc2re> simply maps to C<wc2re_unix>, C<wc2re_win32>, C<wc2re_jokers> and C<wc2re_sql> respectively.
+ 'a***b\\**' ==> 'a(.*)b\\*(.*)'
+ 'a%%%b\\%%' ==> 'a(.*)b\\%(.*)'
-=item C<$^O>
+=item *
-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.
+C<'greedy'>
+
+When used in conjunction with C<'any'>, it makes the C<'any'> captures greedy (by default they are not).
+
+ 'a***b\\**' ==> 'a(.*?)b\\*(.*?)'
+ 'a%%%b\\%%' ==> 'a(.*?)b\\%(.*?)'
+
+=item *
+
+C<'brackets'>
+
+Capture matching C<{ ... , ... }> alternations.
+
+ 'a{b\\},\\{c}' ==> 'a(b\\}|\\{c)'
=back
-If the type is undefined or not supported, it defaults to C<'unix'>.
+ $rw->capture(set => 'single'); # Only capture "exactly one" metacharacters.
+ $rw->capture('single'); # Same.
+ $rw->capture(add => [ qw/any greedy/ ]); # Also greedily capture "any" metacharacters.
+ $rw->capture(rem => 'greedy'); # No more greed please.
+ $rw->capture(); # Capture nothing.
- # Wraps to wc2re_jokers ($re eq 'a\\{b\\,c\\}.*').
- $re = wc2re 'a{b,c}*' => 'jokers';
+The C<capture> method returns the L<Regexp::Wildcards> object.
- # Wraps to wc2re_win32 ($re eq '(?:a\\{b|c\\}.*)')
- # or wc2re_unix ($re eq 'a(?:b|c).*') depending on $^O.
- $re = wc2re 'a{b,c}*' => $^O;
+=head2 C<convert $wc [ , $type ]>
+
+Converts the wildcard expression C<$wc> into a regular expression according to the options stored into the L<Regexp::Wildcards> object, or to C<$type> if it's supplied.
+It successively escapes all unprotected regexp special characters that doesn't hold any meaning for wildcards, then replace C<'jokers'>, C<'sql'> and C<'commas'> or C<'brackets'> (depending on the L</do> or L</type> options), all of this by applying the C<'capture'> rules specified in the constructor or by L</capture>.
=cut
sub convert {
my ($self, $wc, $type) = @_;
_check_self $self;
- my $config;
- if (defined $type) {
- $config = $self->_type($type);
- } else {
- $config = $self;
- }
+ my $config = (defined $type) ? $self->_type($type) : $self;
return unless defined $wc;
+
+ my $e = $config->{escape};
+ # Escape :
+ # - an even number of \ that doesn't protect a regexp/wildcard metachar
+ # - an odd number of \ that doesn't protect a wildcard metachar
+ $wc =~ s/
+ (?<!\\)(
+ (?:\\\\)*
+ (?:
+ [^\w\s\\$e]
+ |
+ \\
+ (?: [^\W$e] | \s | $ )
+ )
+ )
+ /\\$1/gx;
+
my $do = $config->{do};
- my $e = $config->{escape};
- $wc =~ s/(?<!\\)((?:\\\\)*[^\w\s\\$e])/\\$1/g;
- return $self->_sql($wc) if $do->{sql};
$wc = $self->_jokers($wc) if $do->{jokers};
+ $wc = $self->_sql($wc) if $do->{sql};
if ($do->{brackets}) {
$wc = $self->_bracketed($wc);
- } elsif ($do->{commas}) {
- if ($wc =~ /(?<!\\)(?:\\\\)*,/) { # win32 allows comma-separated lists
- $wc = $self->{'c_brackets'} . $self->_commas($wc) . ')';
- }
+ } elsif ($do->{commas} and $wc =~ /(?<!\\)(?:\\\\)*,/) {
+ $wc = $self->{'c_brackets'} . $self->_commas($wc) . ')';
}
+
return $wc;
}
=head1 DEPENDENCIES
-L<Text::Balanced>, which is bundled with perl since version 5.7.3
+L<Carp> (core module since perl 5), L<Text::Balanced> (since 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.
+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<wildcard> function that compiles a matcher. It only handles C<*>.
-
-L<Text::Buffer> has the C<convertWildcardToRegex> class method that handles jokers.
+L<Text::Glob>.
=head1 AUTHOR
Vincent Pit, C<< <perl at profvince.com> >>, L<http://www.profvince.com>.
-You can contact me by mail or on #perl @ FreeNode (vincent or Prof_Vince).
+You can contact me by mail or on C<irc.perl.org> (vincent).
=head1 BUGS
sub _jokers {
my $self = shift;
local $_ = $_[0];
- # escape an odd number of \ that doesn't protect a regexp/wildcard special char
- s/(?<!\\)((?:\\\\)*\\(?:[\w\s]|$))/\\$1/g;
# substitute ? preceded by an even number of \
my $s = $self->{c_single};
s/(?<!\\)((?:\\\\)*)\?/$1$s/g;
sub _sql {
my $self = shift;
local $_ = $_[0];
- # escape an odd number of \ that doesn't protect a regexp/wildcard special char
- s/(?<!\\)((?:\\\\)*\\(?:[^\W_]|\s|$))/\\$1/g;
# substitute _ preceded by an even number of \
my $s = $self->{c_single};
s/(?<!\\)((?:\\\\)*)_/$1$s/g;
projects / perl / modules / Regexp-Wildcards.git / blobdiff