Re: Problem with iterating through a sequence of integ

Michael Kay wrote:

> >   On the other hand, maybe a way to tell the processor to 
> > deactivate static error checks at one place (because we know 
> > the generated code there is correct even if it can look 
> > suspicious) would be interesting.

> This is what "treat as" is designed for.

  Yes, it can help to solve the generation of errors while static
analysis.  I have a simple example it doesn't solve (I know this is a
sample different from the initial context of this thread), inspired
from Schematron implementation:

    (@* treat as node())/*

  This leads into a warning with Saxon telling that "The child axis
starting at an attribute node will never select anything."  That is
true.  But when the stylesheet is generated from a meta-stylesheet
(that contains the XPath expressions.)  On the other hand I wouldn't
want to introduce 'treat as' here to detect wrong expressions like '42'
(and in this case there are other techniques.)

  Anyway, the hint is interesting, thanks.

  Regards,

--drkm
























__________________________________________________
Do You Yahoo!?
En finir avec le spam? Yahoo! Mail vous offre la meilleure protection possible contre les messages non sollicitis 
http://mail.yahoo.fr Yahoo! Mail