Hey,
I'm toying with the RDF output, to see how it can be improved. One thing
I want to do early is to add the possibility to specify the RDF names
for properties ad types. I've added a @URI annotation for this purposes
and here's an example:
@URI( "http://purl.org/dc/dcmitype/PhysicalObject";)
interface TestEntity
extends EntityComposite
{
@URI( DcRdf.DC+"title")
@NotEmpty Property<String> name();
Association<TestEntity> association();
ManyAssociation<TestEntity> manyAssoc();
@URI("http://www.w3.org/2001/vcard-rdf/3.0#GROUP";)
ListAssociation<TestEntity> listAssoc();
}
---
This sets the type to a known RDF-type, the "name" property becomes a
Dublin Core Title, and the listAssoc() is a vCard GROUP.
The resulting RDF (from EntitySerializer) becomes:
<?xml version="1.0" encoding="UTF-8"?>
<rdf:RDF
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#";
xmlns:dc="http://purl.org/dc/elements/1.1/";
xmlns:ns1="http://www.w3.org/2001/vcard-rdf/3.0#";
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#";>
<PhysicalObject xmlns="http://purl.org/dc/dcmitype/";
rdf:about="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity/test2">
<dc:title>Niclas</dc:title>
<identity
xmlns="urn:qi4j:entity:org.qi4j.entity.Identity:">test2</identity>
<association
xmlns="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity:"
rdf:resource="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity/test1"/>
<manyAssoc
xmlns="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity:">
<rdf:Bag>
<rdf:li
rdf:resource="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity/test1"/>
</rdf:Bag>
</manyAssoc>
<ns1:GROUP>
<rdf:Seq>
<rdf:li
rdf:resource="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity/test1"/>
<rdf:li
rdf:resource="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity/test1"/>
<rdf:li
rdf:resource="urn:qi4j:entity:org.qi4j.library.rdf.entity.TestEntity/test1"/>
</rdf:Seq>
</ns1:GROUP>
</PhysicalObject>
</rdf:RDF>
---
which is quite nice :-) This will make it easier for non-Java clients to
consume the RDF, and also helps versioning (as you can specify exact
schema versions using this approach).
What do you think? Any comments on how to improve, or any dangers with it?
/Rickard
_______________________________________________
qi4j-dev mailing list
[email protected]
http://lists.ops4j.org/mailman/listinfo/qi4j-dev
