XOM API for Processing XML with Java

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

Paul Duffin wrote:

> Basically, I have a number of documents that include boilerplate and
> common fragments. These fragments are stored in an Ivy / Maven
> repository and extracted automatically as part of the build. At the
> moment the xi:include href attribute has to point directly to the
> fragments and that makes the document dependent on the build and
> repository structure. What I would like to do is use some well known
> pseudo URLs, e.g. template:/copyright.xml and have the build use an
> XMLCatalog to map that to the actual artifact extracted from ivy.

Your use case is common and reasonable. However I strongly reocmmend 
using standard http or file URLs for this, not inventing a new scheme.

> I have had a look at the XIncluder source and can see that there is no
> way to do that, as it constructs and accesses the URL directly. So the
> questions I have are:
> 1) Is this something that other people would find useful ?

Yes.

> 2) Is this something you (ERH) would like to see in the XIncluder ?

Yes.

> 3) If so how do you think it should be done ?

XML Catalog support is on my TODO list but barring serious interest 
(that is, funding) it's liable to take some time (possibly years) till I 
get around to it.

I do think the XIncluder is the wrong place to put this though. It needs 
to slip in at a lower layer in the Builder class. Maybe a subclass of 
Builder or just an extra method on Builder to supply a catalog or list 
of catalogs?

Possibly you could hack something together quickly simply by passing in 
a properly configured XMLReader to the Builder. That would probably work 
for including XML documents. However text documents are loaded directly 
without passing through the builder so that would need some work inside 
XIncluder to truly enable.

--
Elliotte Rusty Harold  elharo AT metalab.unc.edu
Refactoring HTML Just Published!
http://www.amazon.com/exec/obidos/ISBN=0321503635/ref=nosim/cafeaulaitA