[uf-discuss] ann: SCRuF

[Tom Morris]

On Dec 27, 2007 2:44 PM, André Luís <andr3.pt at gmail.com> wrote:
> Very interesting! Well done.
>
> One question, though. I know that, by definition, xsl works only for
> xml documents, but I was wondering if there was a way of using your
> work with 100% valid html strict documents? Apart from trying to
> render an xhtml version of the document (tidy?), there isn't, right?
>

Generally, GRDDL implementations will use Tidy to convert HTML 4
into XHTML 1.0 before applying the XSLT across them, although
you *can* use GRDDL on HTML in another way, which is you can
set up a pointer with GRDDL to a URL which returns an XSLT
document containing the converted data. This way you could quite
easily write a system for loading an HTML document into a DOM
tree, extracting the relevant data out of it and creating RDF/XML.

Danny has a blog post on this latter approach:
http://dannyayers.com/2007/10/29/non-xslt-grddl
(You can also use this for bridging to legacy systems, APIs, WS-*
stuff and for binary formats.)

-- 
Tom Morris
http://tommorris.org/