From: Kouhei Sutou <kou@...>
Date: 2006-07-20T00:01:11+09:00
Subject: Re: Another look at RSS...

Hi,

In <Pine.GSO.4.64.0607181759140.20906@brains.eng.cse.dmu.ac.uk>
  "Another look at RSS..." on Wed, 19 Jul 2006 02:04:43 +0900,
  Hugh Sasse <hgs@dmu.ac.uk> wrote:

> I've had another look at lib/rss and the patch below is mostly doc
> changes, but I have also changed the method to_class_name.

Thanks, Hugh. I applied your patch expect some changes. See
the below.

>                                                             My testing
> with underscored names shows that it behaves the same, so hopefully 
> I have not broken anything.  I think the code is clearer.  

You can check whether your changes breaks something or not
by "/installed/ruby test/runner test/rss".

> --- ./lib/rss/utils.rb.orig	2006-06-18 14:42:43.000000000 +0100
> +++ ./lib/rss/utils.rb	2006-07-18 17:32:24.047018000 +0100

>      def to_class_name(name)
> -      name.split(/_/).collect do |part|
> -        "#{part[0, 1].upcase}#{part[1..-1]}"
> -      end.join("")
> +      name.split(/_/).collect{|part| part.capitalize}.join("")
>      end

This changes was rejected because this changes breaks
current API. For example, 'skipDays' should be converted to
'SkipDays' not 'Skipdays'.

> --- ./lib/rss/parser.rb.orig	2006-07-13 12:35:17.000000000 +0100
> +++ ./lib/rss/parser.rb	2006-07-18 17:50:33.094732000 +0100

>      def normalize_rss(rss)
>        return rss if maybe_xml?(rss)
>  
>        uri = to_uri(rss)
>        
>        if uri.respond_to?(:read)
> -        uri.read
> +        uri.read        # FIXME? - do we care if the URI is tainted?

I can't judge whether this comment is needed or not. So, I
rejected.

>        elsif !rss.tainted? and File.readable?(rss)
>          File.open(rss) {|f| f.read}
>        else

Thanks,
--
kou