[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: RDDL/RDF - no RDF tax?

• To: "Mike Champion" <mc@x...>, <xml-dev@l...>
• Subject: Re: RDDL/RDF - no RDF tax?
• From: "Jonathan Borden" <jonathan@o...>
• Date: Sun, 17 Nov 2002 15:39:36 -0500
• References: <00c801c28e52$7a011930$7c674544@n...> <oprf0lzak6ezizxn@localhost> <013701c28e73$0d648570$7c674544@n...> <oprf0wihdsezizxn@localhost>

Mike Champion wrote:


> On Sun, 17 Nov 2002 14:53:51 -0500, Jonathan Borden
>
> > that XML makes the following assertions:
> >
> > <#foo> rdf:type rddl:resource .
> > <#foo> rddl:nature <http://example.org/nature> .
> > <#foo> rddl:nature <http://example.org/purpose> .
> > <#foo> rddl:nature <http://example.org/L.dtd> .
> > <#foo> rddl:prose "<p>A description of the 'L'
language>"^^rdf:XMLLiteral
> > .
>
> How does the RDF processor know that the rddl:resource is an rdf:type?

It matches the "typed node element" production in the RDF syntax. Basically
whenever you expect to see:

<rdf:Description ...

and instead you see:

<foo:bar ...

it is "shorthand" for:

<rdf:Description rdf:ID="...">
    <rdf:type resource="&foo;bar"/>
...

where &foo; is the namespace URI bound to the "foo" prefix.

> or that the rddl:prose content is an rdf:XMLLiteral?

the DTD can have a default attribute rdf:parseType="Literal", but I forgot
to attach this attribute to the <rddl:prose> property, so it should be:

<foo:bar rdf:ID="baz">
    <rddl:prose rdf:parseType="Literal">
            <example>the value of the -rddl:prose- property is this piece of
XML</example>
    </rddl:prose>
</foo:bar>

> Is there an RDF schema somewhere that asserts this?
>
> More generally, can one use RDF schema to keep the author of an XML
> document only minimally aware of the RDF underpinnings?
>

Generally not, RDF schemas really don't do syntactic things like default
attributes. RDF Schemas -themselves- don't do much. Both RDF Schema and OWL
act on the parsed set of triples, not the original XML.

I've written a short OWL for RDDL at http://www.rddl.org/rddl2.owl .

In OWL you can limit the number of a particular property via the
<cardinality> restriction. Similarly you can say that owl:allValuesFrom a
certain property must be of a particular type, or that the rdfs:range of a
particular property is restricted to objects of a particular rdf:type. In
this way OWL allows you to say many of the things UML does -- i.e. OWL can
be viewed as a syntax for UML diagrams - FWIW

Jonathan

• References:
  • RDDL/RDF - no RDF tax?
    • From: "Jonathan Borden" <jonathan@o...>
  • Re: RDDL/RDF - no RDF tax?
    • From: Mike Champion <mc@x...>
  • Re: RDDL/RDF - no RDF tax?
    • From: "Jonathan Borden" <jonathan@o...>
  • Re: RDDL/RDF - no RDF tax?
    • From: Mike Champion <mc@x...>

• Prev by Date: Re: RDDL/RDF - no RDF tax?
• Next by Date: Re: RDDL/RDF - no RDF tax?
• Previous by thread: Re: RDDL/RDF - no RDF tax?
• Next by thread: Re: RDDL/RDF - no RDF tax?
• Index(es):
  • Date
  • Thread