[jdom-interest] Element, Attribute, Document: Factory

[bob]

I know that it's been discussed before about JDOM
probably not following the DOM route a using factories
for Element, Attribute and Document.  I think the
original context of the question was about specializing
Element into lots of different FooElement, based
upon the tag name.  ie:  HREFElement, TableElement.

What are your thoughts on going 1/2-way that far?
I'd like to create a single subclass of Elemet, and
have it used for all Element.  Basically, I just want
to add extra helper methods to the normal Element 
class.  Nothing schema/DTD specific.  Perhaps I just
want some slightly different signatures.

I'd like the FooBuilder to use my single subclass of
Element(Document/Attribute) instead of the default.
*Not* specialized by Tag name or anything.

On a particular builder, I'd like to do this before
parsing my document.

	builder.setElementClass( MyElement.class );
	builder.setAttributeClass( MyAttribute.class );
	builder.setDocumentClass( MyDocument.class );

And then parse my document, giving me a tree of my subclasses.

If those methods aren't used, then normal default org.jdom.*
classes are used.

The rationale:

I'm using WebMacro (www.webmacro.org), and it allows special
introspect to get to an object, if it's a JavaBean.  It does
special stuff with assuming 'get' prefixes, to produce pretty
HTML templates.

	$myDoc.Child.TagName.Attribute.AttrName.Value

translates into

	myDoc.getChild("TagName").getAttribute("AttrName").getValue();

Purely aesthetical reaons makes me want

	$myDoc.TagName.Attribute.AttrName

Which would be

	myDoc.get("TagName").getAttribute("AttrName").toString();

That'd require me adding Element.get() and overriding Attribute.toString().

How's that for Too Much Information?

	-bob

	(yes, someone will argue I should use XSLT or similar,
	but this is legacy project just now adopting XML.  No
	hope of migration at this point).