GRRDL & CSS

Hi folks,

I was just re-reading the GRDLL doc [1] when I realised it didn't say
anything about using arbitrary CSS to express semantics within XHTML,
for later extraction with XSLT.

This is a twist of functionality, but my own little experiments
suggest it can work well. For example, here's a snippet for the XOW
[2] stuff (using who,what,where...) I put together for Cannes:

XHTML:

<ul id="list-3" title="Scope">
   <li class="what">Wide!</li>
</ul>

post-XSLT:

<w6:Idea rdf:nodeID="list-3">
  <dc:title>Scope</dc:title>
  <w6:what rdf:parseType="Resource">
    <dc:title>Wide!</dc:title>
  </w6:what>
</w6:Idea>

Note that there's a level of transparency/dual purpose - the elements
and structures identified as being semantically interesting also
happen to be the ones which are likely to require CSS styling.

For the example above I've got the CSS displaying an icon
corresponding to class="what", and  id="list-3" is also used by
Javascript for dynamic manipulation of the XHTML.

This would lead me to think that in quite a few cases, where a company
has in-house styling conventions, useful RDF/XML could be extracted
from existing documents by leveraging the existing CSS.

Cheers,
Danny.

[1] http://www.w3.org/2004/01/rdxh/spec
[2] http://semtext.org/2004-02/index.htm

Received on Thursday, 26 August 2004 07:34:41 UTC