[
Lists Home |
Date Index |
Thread Index
]
> am wondering if anyone can think of any
> other models for how xml applications handle
> namespaces that are not in the applications
> known namespaces than these:
>
> parse-tree: all elements in the unknown
> namespace is ignored, and children of the
> element are parsed.
>
> ignore-tree: as soon as an element in an
> unknown namespace is encountered that
> element and any children are ignored.
>
> fail: unknown namespaces are a cause for
> failure. Generally due to validation
> failure.
Yep - not ignoring the unknown namespaced elements/attributes. With RDF/XML
partial knowledge can be inferred from the structural relationships between
the XML entities, the clearest example being 'striping':
<x:a>
<y:b>
<z:c/>
</y:b>
</x:a>
Assuming x:a is an rdfs:Class, the application can tell that y:b is an
rdfs:Property and z:c is another rdfs:Class.
Additionally, if the application knew the namespace y: (i.e. it had the RDF
schema for it) there may be domain and range restrictions for the property
y:b, from which the application can infer rdf:type's for x:a and z:c,
without knowing anything else about x: and z:.
Cheers,
Danny.
|
