XOM API for Processing XML with Java

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

At 2:14 PM -0500 2/2/04, jcowan AT reutershealth.com wrote:

> 2) Consider flushing the no-namespace constructors: they save only 3 
> characters
>    (plus optional whitespace) over the corresponding namespace constructors.
>    Trying to let people use a namespaceless subset of XOM really doesn't save
>    much mental space, I don't think; DOM has these things because Level 1
>    predates namespaces.

The sentiment (including mine) is running against this. I'm keeping these.


> 2) Unify all namespace checking on a single routine, checkNamespace(String
>    prefix, String URI), to be called in place of checkNamespacePrefix,
>    checkNamespaceURI, and checkAddNamespaceDeclaration.  In the first two
>    cases, the existing URI or prefix is passed.  This allows applications
>    to enforce rules about prefix mapping.  It also increases conceptual
>    compatibility with Attribute, which already has checkNamespace.


Checking the addition of a namespace declaration is potentially very 
different than checking the changing of a namespace for an element. 
checkAddNamespaceDeclaration will remain.

The reason Element and Attribute are asymmetrical here is that 
elements and attributes are not symmetrical. Because attributes are 
required to have prefixes to be in a namespace, the setNamespace 
method is different for the two. In Attribute, this means that you 
have to call setNamespace(prefix, URI) whereas for element you can 
call setNamespacePrefix(prefix) and setNamespaceURI(uri). Otherwise 
it would be possible to create an attribute in a namespace with no 
prefix, a very bad thing.(*) The check methods match the setter 
methods.

That said, I'm not sure they have to match. Element could fill in the 
prefix or URI as necessary from either setNamespacePrefix(prefix) or 
setNamespaceURI(uri) and then call checkNamespace(prefix, URI) or 
Attribute could call checkNamespacePrefix and checkNamespaceURI 
separately. Probably the latter makes more sense because otherwise 
the checks on prefix and namespace URI would each be made twice, once 
for the URI and once for the prefix. What do people think?
--

  Elliotte Rusty Harold
  elharo AT metalab.unc.edu
  Effective XML (Addison-Wesley, 2003)
  http://www.cafeconleche.org/books/effectivexml
  http://www.amazon.com/exec/obidos/ISBN%3D0321150406/ref%3Dnosim/cafeaulaitA

(*) In Element, calling setNamespacePrefix() before calling 
setNamespaceURI() throws an exception. However, the reverse order is 
allowed, because an element is still namespace well-formed if you 
call setNamespaceURI and then don't call setNamespacePrefix(). 
However, with attributes neither order could be allowed, because one 
without the other is always namespace malformed.