(?-xism:(?-xism:(?:\<(?:(?>[^\<\>]+)|(??{$Regexp::Common::balanced [0]}))*\>))) #### #!/usr/bin/perl -- use strict; use warnings; use Test::More qw' no_plan '; use Regexp::Common qw/ balanced /; use Data::Dump qw/ dd pp /; sub TRACE; sub DEBUG; *TRACE = *DEBUG = sub { print STDERR @_,"\n" }; #~ *TRACE = *DEBUG = sub { }; use vars qw/ $re_balanced_angles /; #~ our $re_balanced_angles = qr{(?^:(?^:(?:\<(?:(?>[^\<\>]+)|(??{ $re_balanced_angles }))*\>)))}; #~ our $re_balanced_angles = qr{(?:(?:(?:\<(?:(?>[^\<\>]+)|(??{ $re_balanced_angles }))*\>)))}; #~ our $re_balanced_angles = qr{ #~ (?: #~ (?: #~ (?: #~ \< #~ (?: #~ (?> #~ [^\<\>]+ #~ ) #~ | #~ (??{ $re_balanced_angles }) #~ )* #~ \> #~ ) #~ ) #~ )}x; ## still works the same #~ our $re_balanced_angles = qr{ #~ \< #~ (?: #~ (?> #~ [^\<\>]+ #~ ) #~ | #~ (??{ $re_balanced_angles }) #~ )* #~ \> #~ }x; #~ our $re_balanced_angles = qr{ #~ \< #~ (?: #~ (?> #~ [^\<\>]+ #~ ) #~ | #~ (??{ $re_balanced_angles }) #~ )* #~ \> #~ }x; our $re_balanced_angles = qr{ \< (?: (?> [^\<\>]+ ) | (??{ $re_balanced_angles }) )* \> }x; our $allowed = join '|', qw[ ABBR ACRONYM B BIG CITE CODE DFN EM I KBD SAMP SMALL SPAN STRONG SUB SUP TT VAR ]; { my $in = 'SPAN|class="span_class"> some more text B.'; my $out = shabba($in); my $wanted = 'text italic text some more text bold text.'; is( $out, $wanted , 'shabba'); } { my $in = 'SPAN text>|class="span_class"> some more text B.'; my $out = shabba($in); my $wanted = 'text italic and bold text some more text bold text.'; is( $out, $wanted , 'shabba'); } exit( 0 ); sub shabba { local $_ = $_[0]; my $dent = $_[1] || 0; pos = 0; my $ret = ""; SHABBALOOP: while( length > pos ){ m{\G(\s+)}gcsx and do { $ret .= $1; next SHABBALOOP; }; m{ #~ \G( $allowed )( $RE{balanced}{-parens=>'<>'} ) \G( $allowed )( $re_balanced_angles ) }gcsx and do { TRACE "# $dent allowed<> { $1 ( $2 ) }"; $ret .= shabba_allowed( "$1" , "$2" , $dent ); next SHABBALOOP; }; #~ confusion :) #~ \G(\w+\b) #~ fail #~ \G([^<]+)(?!:$allowed)\b #~ \w+\b #~ \G([^<]+?)(?!:$allowed\<)\b #~ fail #~ \G([^<]+?)(?!:$allowed\<) #~ fail #~ \G([^<]+)(?!:$allowed\<) #~ fail \G([^<]+)(?!:\<) #~ inch #~ \G([^<]+?)(?!:\<) #~ \G([^<]+?\b[^<]) #~ \G([^<]+?[^<]) #~ FAIL #~ \G([^<]+[^<]) #~ \G([^<]+[^<]\b) #~ 2same#~ \G( (?!:$allowed\<) .+ ) #~ 2same#~ \G( .+(?!:$allowed\<) ) m{ \G([^<]+\s) }gcmx and do { TRACE "# $dent text { $1 }"; $ret .= shabba_text( "$1" ); next SHABBALOOP; };;; m{ \G([\<\>]) }gcmx and do { TRACE "## $dent error-stray<> { $1 } at pos(@{[pos]})"; last SHABBALOOP; };;; m{ \G(\S) }gcmx and do { TRACE "# $dent inch-forward { $1 }"; $ret .= shabba_text( "$1" ); next SHABBALOOP; };;; } $ret; } sub shabba_allowed { my( $tag , $stuff, $dent ) = @_; $stuff = $1 if $stuff =~ m{^<(.*)>$}gs; my $ret = ""; $ret .= "<\L$tag\E" if $tag; $stuff =~ s{\|([^<>]+)$}{ $ret .= " $1"; ## shabba_allowed_atts($tag,$1); ""; }gsex if defined $stuff ; $ret .= ">" if $tag; if( defined $stuff and length $stuff and $stuff =~ m{[<>]}g ){ $ret .= shabba( $stuff , $dent+1) ; ## recurse } else { $ret .= $stuff; } $ret .= "" if $tag; $ret; } sub shabba_text { join'',@_ } __END__