- From: Dan Connolly <connolly@w3.org>
- Date: Thu, 04 Oct 2007 08:10:42 -0500
- To: Ben Adida <ben@adida.net>
- Cc: RDFa <public-rdf-in-xhtml-tf@w3.org>
On Wed, 2007-10-03 at 21:03 -0700, Ben Adida wrote: > > Hi all, > > I've updated the RDFa JavaScript implementation to support xml:lang, and > I've completely ripped out the way I was storing triples, using instead > the RDF Triple Store from Tabulator [1]. nifty. > And, best of all, the triple store supports provenance tagging. So you > can add other triples to the store, but look at only the RDFa-generated > triples by limiting yourself to the "rdfa" bucket as the fourth > parameter to .the(), .each(), and .statementsMatching(). That 4th parameter is usually the URI of the document that the triples came from, I think. Actually, the design is that in general, it's a justification for the triple; e.g. "I read <doc1> and <doc2> and did some OWL inference on them." So you might store something like ["rdfa", src_doc] meaning "extracted from src_doc using rdfa". But since RDFa is pretty much just another RDF syntax, and in the RDF/XML and N3 cases we don't bother to note the syntax*, you might not bother. * actually, the tabulator keeps HTTP data such as content-type around in a/the triple store. I'm not sure if that's part of the library or layered on top. > [1] http://www.w3.org/2005/ajar/tab -- Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Thursday, 4 October 2007 13:10:21 UTC