- From: Jos De_Roo <jos.deroo.jd@belgium.agfa.com>
- Date: Sat, 12 Oct 2002 22:28:56 +0200
- To: "w3c-rdfcore-wg" <w3c-rdfcore-wg@w3.org>
I think there is a way to do Mike's long range datatyping
in C-urrent datatyping while using Patrick's _:x"y" nodes
i.e.
_:a foo:bar _:x"10" .
foo:bar rdfs:range xsd:integer .
would entail
_:a foo:bar xsd:integer"10" .
and even
_:a foo:bar _:x"10" .
foo:bar rdfs:range xsd:integer .
xsd:integer rdfs:subClassOf xsd:decimal .
would entail
_:a foo:bar xsd:decimal"10" .
but
_:a foo:bar xsd:string"10" .
foo:bar rdfs:range xsd:integer .
would clash
I've tested those 3 testcases using a closure rule
| IF | THEN
------------------------------------------------
:rule10 | ?p rdfs:range ?A . |
| ?y str:pair (?B ?a) . |
| ?s ?p ?x . |
| ?x str:pair (?A ?a) . |
| ?A rdfs:subClassOf ?B . | ?s ?p ?y .
where str:pair connects a datyped-literal
with its pair datatype-uri + lexical-form
(and it does appropriate promotion of an
existential type _:d to a universal type ?d
when it occurs in a rule premise or query)
and it works
(it's hard to kick-off from datatyping...)
-- ,
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/
Received on Saturday, 12 October 2002 16:29:31 UTC