Sorry for the confusion. It should be "URI pairs" that need to be resolved,
not "all URI's". Such pairs can be resolved in one of 5 ways in 4.3.2,
where the namespace name is mentioned many times. In Xerces'
implementation, we first try to use the namespace name to look up for a
grammar. If this fails, we use the location hint to load the grammar. So
the namespace name is definitely useful.
Actually, I don't want to emphasize on whether the spec requires us to
resolve all location hints. My point was whether it's worth it. If there is
no clear benefit, why do we want to complicate the (already very
complicated) schema loading mechanism.
Cheers,
Sandy Gao
Software Developer, IBM Canada
(1-905) 413-3255
[EMAIL PROTECTED]
Curt Arnold
<carnold@houst To: [EMAIL PROTECTED]
on.rr.com> cc:
Subject: Re: Should we defer loading
schema grammars? (Bug
07/05/2002 10433)
10:24 AM
Please respond
to
xerces-j-dev
The section of the spec that you quoted is "as clear as is the summer's
sun" (http://classics.mit.edu/Shakespeare/henryv/henryv.1.2.html)
If the schema processor is supposed to attempt to load all of the URL's
mentioned in xsi:schemaLocation, then what is the value of the namespace
key? However, if you want a definitive interpretation, you'll should
ask on xmlschema-dev.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
