Sun XML early access

[Don Park]

>Has anyone had a chance to look at the early access Sun XML package yet?
>I'm just starting my explorations, and (so far, to me) it looks pretty
>intriguing.  The DOM interface is promising, and I'm very glad to see SAX
>support.

I have looked at the DOM implementation only and found that:

1. sibling-based navigation is considerablly slower than index-based
navigation.

Actually, this can be fixed rather easily by caching last-accessed index.

2. getElementsByTagName is not efficient.

getElementsByTagName is best implemented using lazy evaluation techniques
but Sun XML package does cache the evaluation result.  For example, the
getLength() method of the returned NodeList walks the whole subtree
everytime.  item() method also walks the subtree until the desired number of
nodes have been seen.

3. Output needs more design work.

One needs to subclass to override output and there is no built in support
for conversion to other formats.  For example, you can not write out a DOM
Document as HTML without some serious patching or overriding.

4. XML Bean concept is a little disappointing for what it does.

I am afraid I can not go into details about this due to Docuverse's own work
in this area but Sun needs to broaden their view of what a bean is.

I am quite sure that efficiency issues will all be resolved over time but
design issues are quite serious and deserves more attention.

Best,

Don Park
Docuverse




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)