XOM API for Processing XML with Java

[peter murray-rust <pm286 AT cam.ac.uk>]

At 11:23 07/02/2006, Michael Kay wrote:
> > Recent discussion suggests that namespaced attribute and content
> > values are considered harmful. It is possible to construct a system
> > which does the same independently of XML machinery but it is more
> > work and I thought I was gaining something from XML that helped.
>
> Using namespace prefixes in content is a very natural thing to do: if they
> make sense in names, then they make sense in content, because if objects are
> identified by QNames then you want to use those QNames within your content
> to refer to those objects. You can't keep data and metadata separate. That's
> why XSLT and XML Schema use this device so extensively. But it's true that
> it gives you problems, because it means that your content is dependent on
> namespace prefixes in a way that isn't apparent to the XML parser (even in
> the case where you have a schema that declares it as an xs:QName).
>
> So it's simplistic to say "don't do it". Just be aware of the consequences.

Many thanks to everyone. Michael's post echoes my current feelings; 
having raised this issue and (partially) understood it I am happy to 
follow and accept what XOM comes up with. "foo:bar" in content is 
extremely useful, memorable (unlike a UUID or a namespaceURI+id) and 
is robust enough for our applications. I am not at this stage worried 
about performance - Jumbo is a reference implementation and validation system.

P.

Peter Murray-Rust
Unilever Centre for Molecular Sciences Informatics
University of Cambridge,
Lensfield Road,  Cambridge CB2 1EW, UK
+44-1223-763069