"FINLEY, Mike" <Mike.FINLEY@xxxxxxxxxx> writes:

Hello Mike,

> Why do you think your first offering is not "efficient"?  I doubt
> you'll find anything more concise (one interpretation of efficient)

If I say

> <xsl:template match="a[not(.//b)]"/>

then all the descendants of "a" are checked for "b" elements. Let's
have a closer look at the input structure.

> <a> 

<!-- match, all descendants of "a" are checked -->

>    <b/>
>    <c/>
>    <a>

<!-- match, all descendants of "a" are checked, again -->

>       <a>
<!-- match, all descendants of "a" are checked, again -->

>          <a>
<!-- match, all descendants of "a" are checked, again -->
>             <c>
>                <b/>
>             </c>
>          </a>
>          <a>

<!-- match, all descendants of "a" are checked, again -->

and so on...

Instead of this, I would like to go straight to the far-most child of
the first "a" element and flag the sub nodes that have "b"
descendants, moving upwards in the tree. (Sorry, my algorythms are a
bit rusty right now, but I think there is a name for this kind of tree
traversing...) This way I need to examine each node only once.

Or maybe I am misunderstanding how XLST implements this? (Anyway, it
doesn't seem very speedy to me ;-).

> mike

Thanks for your comments,

Teppo

-- 
Teppo Peltonen <mailto:teppo.peltonen@xxxxxx>     phone 09 4566080
VTT Information Technology                        mobile 040 5651878
Tekniikantie 4 B, P.O.Box 1201, Espoo 02044 VTT   telefax 09 4567052


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