XOM API for Processing XML with Java

[Elliotte Harold <elharo AT metalab.unc.edu>]

Michael Kay wrote:

> As far as I can tell, XOM is rejecting source documents that use namespace
> names where the namespace name is not a valid absolute URI.

You are correct.

> Assuming I'm right, I can't see the justification for this.

The theoretical justification is that a document that uses relative 
namespace URIs does not have an information set, according to the XML 
Infoset specification.

The philosophical justification is that XOM prevents users from doing 
things they shouldn't be doing, such as using relative namespace URIs.

The practical justification is that XOM tends to notice when users 
accidentally swap the namespace and name arguments to various methods, 
instead of silently producing bad data.

> Getting down to practicalities, I have hundreds of documents in a test suite
> that use relative URIs as namespace names, and I can't use XOM on these
> documents. I suspect they were automatically generated; they certainly
> predate W3C's decision to deprecate relative namespace URIs. With
> considerable effort I could convert the documents to use absolute URIs, but
> this would invalidate my reference results for the tests.

The OASIS XSLT test suite, right? I was bugging the XSLT team about this 
 a week and a half ago. They really need to fix those tests. If you'd 
care to add your voice to the clamor for the test suite to be updated, 
that would be appreciated. Up until about two months ago, their comments 
form was simply dumping comments into a black hole that no one ever saw, 
so they haven't seen any comments you sent until about August, 2004. I 
finally got in touch with a human to fix this, and the form should work 
now, but any other issues you've had with the suite have probably been 
unintentionally ignored unless you were in direct contact with the 
people on the working group.

In the meantime, if you look at the XSLTransformTest class in 
nu.xom.tests you'll notice that XOM essentially skips any tests that 
depend on relative namespace URIs.

Practically I've encountered about three cases where relative namespace 
URIs have been used. Two of them were test suites. I forget what the 
third was. Almost all the examples in the test suites were simply lazy 
typing. Nobody was actually trying to test whether relative namespace 
URIs were handled.

--
Elliotte Rusty Harold
elharo AT metalab.unc.edu