Re: Mixing-up Semantic Data with HTML Rendering

On Mon, 28 Nov 2011 13:18:10 +0000
landong zuo <landong.zuo@googlemail.com> wrote:

> This would probably suggest creating triples in RDFa extractor like
> below,
> 
> <http://blah.co.uk/... >
> <http://www.w3.org/1999/xhtml/vocab#stylesheet>
> <http://blah.co.uk/Styles/theme.css>  .
> 
> Although this triple does not exist in the original dataset.

Depends how you define the "original dataset". If you count the page
itself as the "original dataset", then the triple plainly does exist in
the original dataset.

If there's some underlying process that takes some other dataset,
converts it into RDFa, and then injects it into an HTML template, then
yes, if that HTML template already contains an xhv:stylesheet triple,
this might be seen as an additional triple which didn't exist in the
source data.

For stylesheets this is easily worked around:

 <style type="text/css">
   @import url(/Styles/theme.css);
   /* this creates no triples */
 </style>

As a more general argument though, the information that document X uses
stylesheet Y is perhaps of potential interest to some consumers, so if
you're publishing RDFa and RDF/XML and worried about consumers spotting
differences between the two formats, why not just add the
xhv:stylesheet triple to the RDF/XML too?

-- 
Toby A Inkster
<mailto:mail@tobyinkster.co.uk>
<http://tobyinkster.co.uk>

Received on Saturday, 3 December 2011 19:53:45 UTC