Re: [tag] TA Model

["Dave Pawson" <dave.pawson@gmail.com>]

On 23/09/2007, stephen.green@systml.co.uk <stephen.green@systml.co.uk> wrote:
> So, Dave, would you not want the TA to put the Schematron
> validation after the XML Schema validation and to link the
> two together somehow?

Common sense would say use Schematron after xsd validation.
That's the seperable task for the test flow.
Link them together?
Only via the flow.
If xsdVal==pass:
  Schematron.test.
else
   X.fail



>
> Do you accept that there is no point validating with
> Schematron until first it is established with XML Schema
> that the instance has syntax which adheres strictly to
> the XML Schema (in a typical case, such as for codelists)?

Yes, but I would hate to have to test based on that.

I want to write independent tests.




>
> Does it not make sense to cater for a series of tests in
> a single test assertion

No, not in my ... years experience.


> when there are clearly cases where
> this is an essential part of making test assertions for a
> specification?

Your logic is flawed at that point Steve.
  The then clause isn't dependent on the boolean :-)
The essence is:

xsd validation must pass.
Schematron validation must pass.

The two are *only* linked by the test sequence.





>
> Isn't that why we are calling it a test assertion 'flow'?

Only if you separate it into a thread.
Flow has no connection with a test other than by test Identity.


>
> Isn't it likely that many TAs will have to say that one
> test must take place after another

Yes. But it is easy to isolate sequencing from tests.

Do that and testing becomes lower cost.
Honest!


regards

-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk