As I said: on second thought this might make sense. In any case: you should clearly document this!
Ivan Mitko Iliev wrote: > Hi Ivan, > > Yes, the link with rel=meta will be retrieved regardless of GRDDL > presence . > > Best Regards, > Mitko > > Ivan Herman wrote: >> Hm. Your mail made me think about this... And I wonder what the right >> behaviour is. >> >> Let me ask: is that so that for *any* XHTML file you encounter you will >> retrieve the rel=meta element, regardless of GRDDL? I presume the answer >> is 'yes'. It may be the right behaviour after all... >> >> Ivan >> >> Mitko Iliev wrote: >> >>> Hi Ivan, >>> >>> The current implementation of the SPARQL IRI dereferencing (retrieve >>> missing graph pull-down option) will ask for RDF the target server, in >>> this case the source URL returns xHTML. >>> So when xHTML is retrieved , the SPARQL will extract links with rel=meta >>> , then GRDDL and finally will extract meta about HTML document itself. >>> This may be is not quite correct, so if target server do not return RDF >>> , but it return HTML having embedded RDF, GRDDL and links with rel=meta, >>> which to be preferred i'm not quite sure. >>> >>> Best Regards, >>> Mitko >>> >>> Ivan Herman wrote: >>> >>>> Kingsley et al, >>>> >>>> I am not sure this is a GRDDL/RDFa issue. However... >>>> >>>> When I run the query: >>>> >>>> SELECT * >>>> FROM <http://www.w3.org/People/Ivan/> >>>> WHERE { >>>> ?s ?p ?o. >>>> } >>>> >>>> It actually display _more_ than I expected. Indeed, it also displays >>>> triples that stem from my foaf file: http://www.ivan-herman.net/foaf.rdf >>>> >>>> There are two references to this foaf file in the GRDDL-d file, namely: >>>> >>>> <div class="right foaf:Person" id="me" about="#me" > >>>> <a rel="owl:sameAs" >>>> href="http://www.ivan-herman.net/foaf.rdf#me"; style="display:none"> </a> >>>> >>>> and >>>> >>>> <link rel="meta" type="application/rdf+xml" title="FOAF" >>>> href="http://www.ivan-herman.net/foaf.rdf"; /> >>>> >>>> >>>> I suspect the first statement is the one that triggered virtuoso to get >>>> all that information. >>>> >>>> It is an interesting issue whether this is o.k. or not. From a strict >>>> SPARQL recommendation point of view I think it is *not*. There should be >>>> some extension functions or extra, out-of-band statement in the SPARQL >>>> query that would instruct the processor what to do with owl:sameAs, as >>>> well as the rdfs:seeAlso. I realize that latter appears in one of the >>>> pull down menu items when invoking the query, but *I did not use any of >>>> those*, I simply 'Retrieve missing graphs' entry (second from the top). >>>> >>>> Ivan >>>> >>>> >>>> >>>> >>>> ------------------------------------------------------------------------ >>>> >>>> ------------------------------------------------------------------------- >>>> This SF.net email is sponsored by DB2 Express >>>> Download DB2 Express C - the FREE version of DB2 express and take >>>> control of your XML. No limits. Just data. Click to get it now. >>>> http://sourceforge.net/powerbar/db2/ >>>> ------------------------------------------------------------------------ >>>> >>>> _______________________________________________ >>>> Virtuoso-devel mailing list >>>> [email protected] >>>> https://lists.sourceforge.net/lists/listinfo/virtuoso-devel >>>> >>>> >>> >> > > -- Ivan Herman, W3C Semantic Web Activity Lead URL: http://www.w3.org/People/Ivan/ PGP Key: http://www.cwi.nl/%7Eivan/AboutMe/pgpkey.html FOAF: http://www.ivan-herman.net/foaf.rdf
smime.p7s
Description: S/MIME Cryptographic Signature
