Jeremy Carroll wrote: > > Bijan Parsia wrote: >> On 27 Feb 2008, at 14:43, Jeremy Carroll wrote: >> >>> Inter-op would argue for an XSLT 1.0 transform that probably can be >>> done, but it may be easier to have an XSLT 2.0 transform. >> >> Thinking about it, the trickiest bit will be dealing with URIs and >> CURIEs. Having to write a relative URI resolver will be annoying (but >> it's been done...there are RDF/XML parsers in XSLT) > > It is easiest if we can leave relative URIs as relative. If we have to > do major amounts of URI processing then I would advocate XSLT2 which has > some support. > > CURIEs could be an issue :( > > I think I did write some code for them, once upon a time; and presumably > Fabien's RDFa code does something with them too. > It does: look at the "expand-curie-or-uri" template in: http://ns.inria.fr/grddl/rdfa/2007/09/19/RDFa2RDFXML.xsl the script is not 100% up-to-date with the latest RDFa spec, but he said he would update it. Ivan >> >>> I tend to agree with Bijan's judgement that in principle this is >>> possible (although I would expect corner cases which don't work, e.g. >>> a property http://example.org/000) >> [snip] >> >> Well, this is a case where the RDF/XML can't rep it anyway, so that's >> fine (or, rather, an instance of a known limitation of the target >> format). >> > > Known limitations strike me as OK. We would be doing very well if the > only limitations are imposed by RDF/XML. > > Jeremy > > -- Ivan Herman, W3C Semantic Web Activity Lead Home: http://www.w3.org/People/Ivan/ PGP Key: http://www.ivan-herman.net/pgpkey.html FOAF: http://www.ivan-herman.net/foaf.rdf
This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 27 February 2008 15:04:40 GMT