Re: [xml-dev] Using Saxon 9 as a Schematron 1.5 back end

[Philippe Poulard <philippe.poulard@sophia.inria.fr>]

David Carlisle a écrit :
>> Surely. We can find cases when one is better that the other. Naming 
>> elements has the advantage that in many cases we can identify the 
>> element in fault at a glance.
> 
> yes, for humans names (including prefixes)  are usually better.
> In a schematron context though there's no standard way of expressing the
> required namespace bindings to a following process, so (for machine use)
> it's simpler to use the prefix free form which you can just expose as
> a string for some later Xpath parser to pick up.
> 
> In a system where you have full control of the pipeling mechanisms, the
> tradefoffs may well be different...

In fact, I did it for both usages : it's human readable AND it can be 
added in an XML report to be processed automatically
In this last case, if I append my canonical path to an element (as text 
context or as an attribute), the relevant namespace declarations will be 
added if necessary

-- 
Cordialement,

               ///
              (. .)
  --------ooO--(_)--Ooo--------
|      Philippe Poulard       |
  -----------------------------
  http://reflex.gforge.inria.fr/
        Have the RefleX !