Just an idea (not well-thought-out) high-level idea, but by customizing the
binding to produce javadoc and using the xjc:superclass extension, you could
generate a class which would extend a base CMP class with XDoclet javadoc.
With that, you could run the XDoclet Ant task, and there you go. CMP beans
from XML Schema files.
[I'm sure there are about 100 holes in this approach :) ]
-----Original Message-----
From: Hank Ratzesberger [mailto:hankr_at_crustal.ucsb.edu]
Sent: Thursday, June 26, 2003 4:06 PM
To: JAXB-INTEREST_at_JAVA.SUN.COM
Subject: Generate CMP Entity Beans
Thanks to everyone for the feedback or just
having to click <del>. I feel like I am
off the ground now and hope I can provide
some good feedback myself.
One question I would like to ask, one more
step in the model driven process, is anyone
experimenting with xml shredding and related
to this, has anyone tried to create CMP
entity beans to match schema ?
I realize that Oracle and DB2 have shredding
capability, but the license costs are high
for the particular project I have in mind.
It's complex, I realize, but automatically
generating a class model is a major step
(thank you JAXB team). The next is persistence.
I was envisioning an annotated schema file
or specific set of xsl transforms that
extract identifiers from the model and use
them in deployment descriptor for a CMP.
I think you could go very far using the
schema to create CMP entity beans.
Best regards,
Hank
============================================
Replying to a list? I'll get my copy there!
Hank Ratzesberger
http://crustal.ucsb.edu/ |
http://neesgrid.org/
Institute for Crustal Studies
University of California, Santa Barbara
============================================