Hi Michael Kay and others,
I found this link:
www.xml.com/pub/a/2001/04/25/xsltuk.html#performance. One of
your advices is don't use xsl:number. Why not?
Because a naïve implementation will have O(n^2) performance to number n
nodes. Using position() is always likely to be more efficient if it
gives the right answer. Of course techniques like
count(preceding-sibling::*) are also likely to suffer the same O(n^2)
problem.
Michael Kay
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list