Looking good voyager. Here's a small patch that will convert any UTF-8 extended ascii from XML::Parser to latin1 so odd nodetitles or usernames look right.

Example nodetitle from today's new nodes:

Before: Reactionary Coding—One-Shot Programs
After : Reactionary Coding—One-Shot Programs

Add this conversion subroutine:

sub UTF8_latin1
{ # UTF-8 to latin1 regex from XML::TiePYX (thanks to mirod)
my($text) = @_;
$text =~ s{([\xc0-\xc3])(.)}{
  my $hi = ord($1);
  my $lo = ord($2);
  chr((($hi & 0x03) <<6) | ($lo & 0x3F))
 }ge;
return $text;
}
[download]

Add two lines to the foreach loop in sub html_by_type:

foreach my $node (@$nodes) {
        $node->{cdata} = UTF8_latin1($node->{cdata}); # ADDED
        my $link = $q->a({-href => "$url_base$node->{node_id}"}, $node
+->{cdata});
        my $node_text = $q->li($link);
        my $author = $authors->{$node->{author_user}};
        $author = UTF8_latin1($author); # ADDED
        $node_text .= (' ' x 2) . $q->small(" by $author") if $SH
+OW_AUTHOR;
        print $node_text;
}
[download]

Enjoy.

--
Check out my Perlmonks Related Scripts like framechat, reputer, and xNN.