On Thu, Jan 12, 2006 at 05:35:13 +0800, Jeremy Wong ��?��?�?? wrote: > > Below is a SPARQL query to retrieve pairs of resource and resource's > property (when the resource acts as subject). It is to find out whether the > set of rdfs:domain of a rdf:Property is a subset of the set of rdfs:Class > of a resource. It looks so lengthy... Any way to improve the query? > > PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#> > SELECT DISTINCT ?r0 ?property{ > ?r0 a ?c0. > ?property rdfs:domain ?c0. > OPTIONAL{ > ?property rdfs:domain ?c1. > FILTER(?c1=?c0) > OPTIONAL{ > ?property rdfs:domain ?c2a. > FILTER(?c2a!=?c1) > } > FILTER(BOUND(?c2a)) I think all this line does is negate the OPTIONAL keyword obove, if c2a must be bound then the OPTIONAL block had to match, which has the same semantics as { ?property rdfs:domain ?c2a. FILTER(?c2a!=?c1) } - SteveReceived on Thursday, 12 January 2006 09:51:45 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:22:39 GMT