Rolf Kemper wrote:

> However, can you drop me some lines on xmlgawk ?

xmlgawk currently has the status of an experimental
extension of GNU Awk. You can find more information
about it in my posting in the newsgroup comp.lang.awk:

  http://groups.google.de/groups?hl=de&lr=&ie=UTF-8&frame=right&th=c016911846a52fdd&seekm=2o4bv6F6u6h2U1%40uni-berlin.de#link1

I have just started writing some GNU-like doc about it
but the doc is not ready yet. Until then, Stefan Tramm's
description may be helpful:

  http://homepage.mac.com/stefan.tramm/iWiki/XmlGawkMacOSX.html

> Can I use this within MSXML4 processor ?

I don't know too much about XML tools in the Microsoft Universe.

> How about schema check ?

xmlgawk is not meant to validate XML files.
xmlgawk does not use a DOM representation (although
Manuel Collado has written one in xmlgawk). Are you
sure you need a DOM-representation to do your work ?
My experience is that users often over-estimate the
need for a DOM. But I am ready to be convinced.

> As you stated, YES , the example is like counting some beans, but just
> for the sake of demonstration. It addresses a very tiny problem of an
> 5M mixed data set. So my real world is a farm of beans , fruits ,
> nuts, vegitables and many other things.

XML files with several MB of data, sounds interesting.
How long does it take your XSL processor to do a run
over these files ? Several seconds or more ? I am asking
because I am interested in finding out typical uses of
XML files (outside web design), including turnaround times.

> Myself is just creating different dishes out of it. XLST in
> combination with XPATH is a great help, as I have to present the same
> dish in differrent restaurants (HTML , Text , Exel , Spice ) too.

Spice ! Is this data-flow typical for industrial environments ?
If so, I would take this as an indication that XML has definitely
broken out of the web design niche.

Thanks for posting such an interesting use-case.