XOM API for Processing XML with Java

["Bradley S. Huffman" <hip AT cs.okstate.edu>]

"Michael Kay" writes:

> In this case I think that Elliotte's solution of mapping one XPath text
> node
> in the XPath result to all four corresponding XOM text nodes in the XOM
> result is still probably the best option; at any rate, I haven't seen a
> better one.
 
I thought that also, but as Hallvard also point out, even if you're not
using XPath, for robustness your still going to have to take adjacent text
nodes into account. So I'd say include all text nodes in the XPath result
(unwrapped), document the pitfalls with adjacent text nodes (both in the
XPath javadoc and in ParentNode javadoc), and let the user get use to dealing
with it. They are going to have to anyway.

Brad