W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > April 2005

Re: GRDDL implementation in Raptor

From: Ralph R. Swick <swick@w3.org>
Date: Tue, 05 Apr 2005 09:55:50 -0400
Message-Id: <5.1.0.14.2.20050405094844.0226de90@127.0.0.1>
To: Dave Beckett <dave.beckett@bristol.ac.uk>
Cc: public-rdf-in-xhtml-tf@w3.org

At 11:38 AM 4/4/2005 +0100, Dave Beckett wrote:
...
>It works by reading the source document (assumes it is XHTML) into a
>libxml2 document, then applying an XPath query:
>//html:html/html:head[@profile="http://www.w3.org/2003/g/data-
>view"]/html:link[@rel="transformation"]/@href
>
>where the prefix html: is defined to URI http://www.w3.org/1999/xhtml
>
>It then uses the resulting entries in the nodeset as relative URIs to
>the document of XSLT sheets, which are then fetched and applied with
>libxslt to the source libxml2 document, giving a new RDF/XML document
>which is parsed with raptor's RDF/XML parser to give RDF triples.

nice, Dave.

How robust do you think this particular pipeline implementation is?
E.g. (as I'm not intimately familiar with libxslt), if the transformation
URIs do not resolve to documents whose root element is in the
http://www.w3.org/1999/XSL/Transform namespace -- or even worse,
do not resolve to XML documents -- what are the failure modes?

http://www.w3.org/2004/01/rdxh/spec says ($Date: 2005/03/25 23:03:14 $)
"Transformation algorithms should be represented in XSLT... Other
representations may be used by prior agreement of all concerned parties."
so implementations have to expect that random bags of octets might be
found at those transformation URIs.
Received on Tuesday, 5 April 2005 13:56:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:14:59 GMT