Re: Another tokenize() question

aha sorry I picked {} as a random chice (1001 years of TeX usage
probably influenced me:-) Of course, you can use any two characters you
like so long as they don't otherwise occur in your data. However if you
do use { and } you have to double them up as {{ and }} as the regex
attribute is actually an attribute value template and so inside { ...}
it evaluates XPath expressions. 
this is usuful for building up large regex as a set of variables then
doing

regex="{$word-re}|{$special-hidden-elements-re}"

and have the xpath variables $word-re expanded before the regexp is
processed.

David



________________________________________________________________________
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
________________________________________________________________________