#!perl -l
# ex: set ts=4:
use strict;
use warnings;
use HTML::Parser;
use URI;
my (@tagstack, $BUF);
sub start { # enter into tags
my ($tag, $attr, $text) = @_;
$tag .= " href" if ($tag eq "a" && defined $attr->{"href"});
push @tagstack, $tag;
output($text);
}
sub end { # escape out of tags
my ($tag, $text) = @_;
shift @tagstack while (scalar @tagstack && $tagstack[0] ne $tag);
shift @tagstack if scalar @tagstack; # actually nuke element we're looking for
output($text);
}
sub text { # handle everything inside and around tags
my ($text) = @_;
if (unlinked()) {
# replace URLs with their linked equivalent if we're not within a link
$text =~ s{ \b(http://\S+) }{ "new($1)->canonical . "\">$1" }gex;
}
output($text);
}
# are we inside a link right now?
sub unlinked { return not scalar grep { /^a href$/ } @tagstack; }
# add to output buffer
sub output { $BUF .= shift @_; }
# start code
my $p = HTML::Parser->new(
"start_h" => [ \&start, "tagname, attr, text" ]
,"end_h" => [ \&end, "tagname, text" ]
,"text_h" => [ \&text, "dtext" ]
);
$p->parse(do{ local $/; });
print $BUF;
__DATA__
http://linked1.com
http://linked2.com
http://linked3.com
http://linked4.com
http://unlinked1
http://unlinked2.com