Re: [clr-dev] Question about the validation tools

["G. Ken Holman" <gkholman@CraneSoftwrights.com>]

At 2010-05-13 08:40 -0700, ericdes wrote:
> > Any XPath expression can be used.
>
>I feel a bit uncomfortable with that because:
>
>(1) I'm thinking of working with object instances generated from XML 
>files and I want to apply the validation rules directly to these 
>objects at the GUI level -- or displaying the filtered choices to the user.
>
>(2) My idea is to have the application read the XPath / Schematron 
>expressions and see how I can apply these rules directly on the objects.
>
>(3) It might not be an easy task because I see that the XPath 
>grammar -- http://www.antlr.org/grammar/1264460091565/XPath2.g -- is 
>somehow very elaborated.
>
>Would this be a possibility that OASIS defines a subset of the XPath 
>grammar authorized in a CVA file, with this in mind that the 
>expressions allowed in that subset could easily be parsed by most 
>languages, even those that do not understand natively XPath?

I doubt that because it would not be backward compatible.

>After all, there might be a chance that not all of XPath be useful 
>to grab a record...

Granted ... but which subset would not preclude something that 
someone really needs.

Between trading partners you could agree on a subset of the XPath 
vocabulary and just use that.

Actually, note that CVA files have a queryBinding= attribute.  When 
it is unspecified, there are no constraints on the expression of 
document context.  My cva2sch implementation assumes it is all of 
XPath.  Two trading partners could agree on a value for queryBinding= 
(there are no normative values) and just stick within that, and it 
would all be conforming.

But you wouldn't be able to use, say, cva2sch to confirm that only 
the subset of XPath was being used.

But if you wrote your own CVA implementation, then you could confirm 
that the queryBinding= attribute exists with the value you are 
expecting, and then validate that the document context expressions 
are all satisfying that constraint.

This conversation would have been very useful on the CLR-Dev list for 
others to see.  I think it is a meaningful discussion for other 
readers.  If you post your original question in its original form, I 
will respond with the response I wrote earlier and we can try to 
recreate the thread.

. . . . . . . . . . Ken

--
XSLT/XQuery training:   after http://XMLPrague.cz 2011-03-28/04-01
Vote for your XML training:   http://www.CraneSoftwrights.com/c/i/
Crane Softwrights Ltd.          http://www.CraneSoftwrights.com/c/
G. Ken Holman                 mailto:gkholman@CraneSoftwrights.com
Male Cancer Awareness Nov'07  http://www.CraneSoftwrights.com/c/bc
Legal business disclaimers:  http://www.CraneSoftwrights.com/legal