HTML DOM

[Curt Arnold <ca...@houston.rr.com>]

The current Xerces-J implementation of HTML DOM requires the source document to not be namespace qualified, use upper case tag and attribute names, and normalizes parameters and return values to upper case.  This makes the implementation unusable with XHTML documents which are namespace qualified, use lowercase tag and attribute names and normalizes to lowercase.

I've make some quick hacks to get the implementation to work with XHTML documents for use in the development of the W3C HTML DOM test suite, but at the cost of breaking compatibility with the previous behavior.

Having an HTML DOM implementation that works with XHTML would seem to be much more useful than the current implementation.  It may be possible to support the old behavior with non-namespace-qualified documents and standard behavior with namespace-qualified but it could be ugly.

I could potentially provide patches that either implement XHTML behavior or a hybrid behavior or could just drop it.  Would be interested to get any feedback.