#!perl -wl
# vim: set ts=8:
# how to replace certain text (your $name) with a link (to $url)
# based on what tags it is in inside the url
# the one problem i see is that if text is already in then we
# embed an into it, the text after the href but
# before the end of the original will lose its properties.
# this can be accounted for, but i'm lazy
use strict;
use HTML::Parser;
# global variables
our ($name, $url, @tagstack, $out) =
qw{pizza http://www.parseerror.com/};
# sub defs
sub start {
my ($tag, $attr, $text) = @_;
$tag .= " href" if defined $attr->{"href"};
push @tagstack, $tag;
output($text);
}
sub end {
my ($tag, $text) = @_;
shift @tagstack
while (@tagstack && $tagstack[0] !~ /^$tag$/);
shift @tagstack if @tagstack;
output($text);
}
sub text {
my ($text) = @_;
if ($text =~ /\b$name\b/ && canreplace() && unlinked()) {
$text =~ s#\b$name\b#$name#g;
}
output($text);
}
sub output {
my ($txt) = @_;
$out .= $txt if defined $txt;
}
# are we inside a tag we don't want to link?
sub canreplace {
return ! grep {m/^(head|title)$/} @tagstack;
}
# are we inside a link right now?
sub unlinked {
return ! grep {m/^a href$/} @tagstack;
}
# start code
my $p = HTML::Parser->new(
"start_h" => [ \&start, "tagname, attr, text" ]
,"end_h" => [ \&end, "tagname, text" ]
,"text_h" => [ \&text, "text" ]
);
$p->parse(q{
pizza is delicious!
pizza already linked
some more pizza
pizza in a but not linked
});
print $out;