> DeMorgan's theorem can be proved by complete induction over all
> possible inputs.  I don't think it makes any requirements other than
> that the functions behave like the boolean opeators they're meant to
> emulate.  For example, if there were also type conversions involved,
> one might need too be careful when applying DeMorgan's theorem to
> XPath expressions.  I've not actually analyzed whether there would be
> any screw-cases here or not.

I think de Morgan's law applies correctly to XPath conditional expressions,
even when evaluating one of the operands generates an error.

Mike Kay


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list