Re: Counting nodes efficiently

"Wendell Piez" <wapiez@xxxxxxxxxxxxxxxx> wrote in message
news:6.0.0.22.0.20040219114856.035626e0@xxxxxxxxxxxxxxxx
> Oops,
[snip]

> In my last e-mail I was reading too quickly, and neglected to observe,
> before answering, that you're stepping forward one node at a time along
the
> following-sibling axis.

Yes, but I always traverse first the first child node. This is called DFS
(depth-first search) if any searching was involved.

Node-by-node sounds well (thank you!) so probably we can speak about
node-by-node DFS?

>
> I'd probably distinguish this from the usual identity
> transform-by-tree-traversal by referring to its "stepping forward".
> "Stepping forward" or even "tiptoeing forward" (since you're often
> conducting a test as you go) is the way I usually describe the family of
> methods that traverse the following-sibling axis to allow this kind of
> fine-grained control.

Using only "forward" is insufficient -- a tree is 2-dimensional (this is why
there are many possible traversal orders) and we use two independent axes -- 
"ancestor" and "preceding".

Thanks again.

Cheers,

Dimitre Novatchev
FXSL developer,

http://fxsl.sourceforge.net/ -- the home of FXSL
Resume: http://fxsl.sf.net/DNovatchev/Resume/Res.html




 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list