RE: CDATA conversion issue

["Arnold, Curt" <Curt.Arnold@hyprotech.com>]

I also can't answer from the top of my head about the behavior of those two parsers.  Definitely Crimson's and Xerces's implementation of the DOM allow you to retain CDATA sections if you desire.  If
you set isCoalescing=true for Crimson (default is false), it will coalescing the CDATA sections with nearby text which is what most people (but not you) would desire.  (Xerces's doesn't appear to
respond to setting isCoalescing=true).

Possibly, the default value for Oracle and Forte's parser equivalent of isCoalescing is set to true.