In XPath 2.0 I think you can achieve this using

codepoints-to-string(string-to-codepoints(normalize-unicode($in, 'NFKD'))[.
lt 127])

This splits composite characters into the base character plus modifiers,
then strips off the modifiers.

Michael Kay
http://www.saxonica.com/


> -----Original Message-----
> From: Peter Hickman [mailto:peter@xxxxxxxxxxxxx]
> Sent: 15 February 2007 14:27
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject:  Flattening characters to plain latin
>
> Sorry if this has been answered a dozen times but I am
> looking for some xslt to take a string of latin script
> unicode characters and flatten them to their plain
> equivalents. For example the g would become c, q would become
> n, v would become o and so on.
>
> Or do I have a massive list of unicode pairs to enter?