Tuesday, 07 February 2006

As I mentioned yesterday, I noted that Blosxom was misbehaving with RSS feeds where a flavour was used to customise the character set of the feed.

Tonight I made time to look under the hood, and found:

      if ($content_type =~ m{\Wxml$}) {
        # Escape <, >, and &, and to produce valid RSS
        my %escape = ('<'=>'&lt;', '>'=>'&gt;', '&'=>'&amp;', '"'=>'&quot;');  
        my $escape_re  = join '|' => keys %escape;
        $title =~ s/($escape_re)/$escape{$1}/g;
        $body =~ s/($escape_re)/$escape{$1}/g;
      }

Which I promptly changed to

      if ($content_type =~ m{\Wxml(; charset=.*)?$}) {
        # Escape <, >, and &, and to produce valid RSS
        my %escape = ('<'=>'&lt;', '>'=>'&gt;', '&'=>'&amp;', '"'=>'&quot;');  
        my $escape_re  = join '|' => keys %escape;
        $title =~ s/($escape_re)/$escape{$1}/g;
        $body =~ s/($escape_re)/$escape{$1}/g;
      }

and now I can have my cake and eat it: UTF-8 encoded RSS feeds.