I really don't think for this usage NaN is a usable response, do you.
It is (or rather was) reasonable, and in particular consistent with
XPath1's string-to-number casting function number() which gives NaN for
such a string.
But as Mike just replied, it's changed in the current spec to be more
like you wanted, and more like XSD validation, so you should be happy
now.
'bout sums up some of my responses to XSLT 2.0
I was going to comment in my first reply that I was surprised to see
you (well known for being pleased about all the schema stuff added to
XSLT) using casting to XSD types rather than using regexp handling.
Mike just said regexp are likely to be more efficient as well...
David
--
http://www.dcarlisle.demon.co.uk/matthew
________________________________________________________________________
This e-mail has been scanned for all viruses by Star Internet. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list