> Thanks again to Uche for the clarification (I already thanked him
> privately) -- I either never knew or have completely forgotten the
> term "XPattern" for an XSLT pattern, but I wasn't much involved with
> the XSL people in the late 1990s.
>
> Perhaps the term 'XPattern' would be safe to reuse for a
> SAX/STAX-friendly XPath subset, then? That's what I had originally
> (and mistakenly) thought Uche was writing about.
Uche: are your XPatterns full XSLT patterns? That is, can they include any
XPath expression within a predicate? Or are they more like the subset of
XPath used in XML Schema uniqueness constraints, which is designed for
serial on-the-fly evaluation and doesn't allow predicates at all?
Michael Kay
http://www.saxonica.com/
