XOM API for Processing XML with Java

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

> Any thoughts on getting rid of getAttributeCount(), getAttribute(index),
> getNamespaceDeclarationCount(), getNamespaceDeclaration(index) in favor
> of two methods, getAttributes(), getNamespaceDeclarations(), that return a
> iterator. I'm guessing you don't like java iterators cause they're not type
> safe, but index operations on attributes and declarations just seems weird.

I think we had something close to that at one point. I'm absolutely 
not going to uuse a type-unsafe iterator. And I will use indexed 
access, as SAX does.

However, I could see exposing the existing package protected 
Attributes and Namespaces classes publicly; not all of them but 
enough of them to allow read-only access. However, this would add at 
least two classes and four methods, a net gain of two classes.  On 
the other hand, perhaps more smaller classes is better, I admit that 
I'm uncomfortable with how big Element is. But I'm not convinced yet.

What does everyone else think?
--
Elliotte Rusty Harold
elharo AT metalab.unc.edu
Processing XML with Java (Addison-Wesley, 2002)
  http://www.cafeconleche.org/books/xmljava
  http://www.amazon.com/exec/obidos/ISBN%3D0201771861/cafeaulaitA