Re: [xsl] Running through a subset of elements based on index

[David Carlisle <davidc@xxxxxxxxx>]

are you using xslt 2 or 1

In xslt 1 if you want to to much "processing" at the same time as
grouping it's almost always easier to do it in two passes, first use the
muenchian grouping, as you have it, to remove duplicates and sort, but
don't process each group, then (by using a second stylesheet or a
x:node-set() extension process the sorted nodes. As described the
processing os probably simple enough to do it in one pass, but it's on
the edge and if it gets any more complicated definitely doing it in two
passes is easier and more maintanable.

If on the other  hand you're using xslt2 this kind of processing is made
much easier with xsl:for-each-group.

David