[
Lists Home |
Date Index |
Thread Index
]
> I think confusion has come in due to an inconsistency between the notion of
> a node-set being unordered, and the fact that XSLT can process node-sets as
> if they are ordered. For instance, the "for-each" element will process the
This seems to be part of a wider problem. XPath implementations are
mostly tied to XSLT because XPath was defined with XSLT in mind. There
are two Java implementations that I know of, that provide some kind of
reasonable interface for querying a DOM tree using XPath in a standalone
application (Xalan and Jaxen), and only Jaxen provides a full, useable
interface.
I have to say, most XPath engines are horribly badly engineered for
standalone applications. Even Xalan makes it a nightmare to bind
user-defined variables for a query. The Xalan2 api docs state that
variable binding has been made "more low-level" still.. why??
I won't even start with my success of binding my own variables in the
.NET System.Xml classes. Their System.Xml.XPath classes do not even
envisage that someone might ever want to do this.. you have to pick out
Xsl classes, and then the exceptions start...
A couple of use cases for the XPath specifications could have prevented
this mess.
Christian
|
