Back to XObjects was XLink - where a

[David Brownell]

Responding to Graham's comment about binding elements to classes, I wrote:
> 
> I've got an API driven approach, which could be driven based on config
> files using an XML syntax like the one I suggested a while back.  But
> the API can also be driven in other ways, too.  ("Element Factory", an
> interface that can be implemented in many ways.  I'll post more later.)

It's "later", after a busy Thursday and a morning cuppa Java ...

	// to customize DOM tree building
	public interface ElementFactory {
		// like DOM createElement, used with namespaces disabled
		ElementEx createElementEx (String tag);

		// ... namespace aware version
		ElementEx createElementEx (String uri, String tag);
	}

	// implemented by elements and attributes for namespace support
	public interface NamespaceScoped extends Node {
		// returns "local part" of node name, no prefix
		String getLocalName ();

		// returns URI of namespace, null for default
		String getNamespace ();
	}

	// extended element functionality
	public interface ElementEx extends Element, NamespaceScoped {
		// namespace-aware
		String getAttribute (String uri, String name);

		// namespace-aware
		Attr getAttributeNode (String uri, String name);

		// basic delegation hook
		Object getUserObject ();

		// basic delegation hook
		void setUserObject (Object obj);

		// has more, for other purposes ...
		...
	}

That "user object" bit is a delegation hook; it's just stored for use by
applications.  Supports many-to-one association of elements to objects in
an application model, no backlink.  I don't know if it suffices, but I
suspect that any more would become specific to some problem domain.

For the moment, I'm not listing the DocumentBuilder bit here.  Basically,
you can feed an ElementFactory in to a DocumentBuilder and it'll use it
when building the DOM tree; if you don't, it has defaults.  And there's a
simple table-driven implementation that's easily customized.  (Maybe it'll
learn a standard "config file" syntax -- in XML of course!)

Other implementations are clearly possible, including ones using heuristics
like those like James Anderson posted.

- Dave

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)