Elliotte Harold wrote:

> There are many different possible tree models for XML. The deeper you 
> look the more differences you find between DOM, the XML Infoset, XPath 
> 1.0, XPath 2.0, and other tree-based data models. But it doesn't stop 
> there. There are non-tree-based models of XML documents, and these may 
> also be profitably used in the right circumstances. XML is defined 
> such that it is possible to represent a well-formed document as a 
> tree. It is by no means necessary to do so.
>
Well, the thread is on XML-aware programming languages, and every such 
thing needs to define a data model.

No language designer can seriously claim that his language is the right 
one for all circumstances.

But, to process XML in a program, you can never do without the tree view.

> SAX exposes probably the most popular non-tree model, but it's hardly 
> the only one. Indeed there are non-event, non-tree models as well. Some 

Aha, event-based and in which order are the events called?

> people have shredded XML documents into relational tables. That works 
> too. None of these models are more correct than any of the others. 
> It's simply a question of which is most useful for the problem at hand 
> given the available tools.

But tools bring you only this far, and a programming language is 
providing more coherence than using a set of ill-matching tools. At 
least that is the assumption under which to design an XML aware 
programming language, and I find it valid.

regards,
Burak Emir

http://lamp.epfl.ch/~buraq