Support for new specifications in OWL 2

[DhanshreeA]

Does rdflib currently support adding data ranges as per OWL2 specifications? For Example, can you create a custom datatype like:

g.add((URIRef(ns+"Answer"), RDF.type, RDFS.Datatype))
g.add((URIRef(ns+"Answer"), OWL.unionOf, (XSD.string, XSD.boolean, XSD.positiveInteger))) 
g.add((URIRef(ns+"answer_value"), RDF.type, RDF.Property))
g.add((URIRef(ns+"answer_value"), RDFS.domain, URIRef(ns+"Question")))
g.add((URIRef(ns+"answer_value"), RDFS.range, URIRef(ns+"Answer")))

[MuraliKrishna26]

Hey DhansheeA, I have worked on the issue and fixed it. Now you can add more than one data-type(or any property) at once as you mentioned in the code.
The below line will work fine
g.add((URIRef(ns+"Answer"), OWL.unionOf, (XSD.string, XSD.boolean, XSD.positiveInteger))

I have observed there may be a typo in the line :
g.add((URIRef(ns+"answer_value"), RDF.type, RDFS.Property))
Property term is defined in RDF class but not in RDFS.

[nikokaoja]

@MuraliKrishna26 , I tried:
g.add((URIRef(ns+"Answer"), OWL.unionOf, (XSD.string, XSD.boolean, XSD.positiveInteger))

and it does not work. I am getting:

AssertionError: Object (rdflib.term.URIRef('http://www.w3.org/2001/XMLSchema#string'), rdflib.term.URIRef('http://www.w3.org/2001/XMLSchema#boolean'), rdflib.term.URIRef('http://www.w3.org/2001/XMLSchema#positiveInteger')) must be an rdflib term

[ashleysommer]

@MuraliKrishna26

Property term is defined in RDF class but not in RDFS.

RDF.type is correct. Even though it is specified in the RDFS spec, the type predicate intentionally uses rdf:type.
See https://en.wikipedia.org/wiki/RDF_Schema

[DhanshreeA]

@ashleysommer I believe @MuraliKrishna26 is talking about RDF.Property and is correct in pointing out that Property is in RDF Class and not RDFS. :) I am a bit rusty on the spec itself, but at least within this Python implementation RDFS.Property does not seem to be the case.

[ashleysommer]

@MuraliKrishna26 is talking about RDF.Property and is correct in pointing out that Property is in RDF Class and not RDFS

Yes, you're right. I misread what @MuraliKrishna26 had written.