Re: [xml-dev] xs:assert and Schematron

[Rick Jelliffe <rjelliffe@allette.com.au>]

Jesper Tverskov wrote:
> I still would like to know if we have important use cases where
> xs:assert can not do the job but Schematron. I would like a few
> examples of what Schematron can test and not XML Schema 1.1.
>   
Have a look at 
http://broadcast.oreilly.com/2008/11/validating-code-lists-with-sch.html
The "complex example" there is a *real* use case as part of the 
Australian Spatial Data
Infrastructure Directory, for the government organization Geoscience 
Australia, who do
resource mapping of different kinds (oil, minerals, ore, coastal, etc). 
See under Schematron
in http://asdd.ga.gov.au/asdd/profileinfo/   

Try doing anything like that with xs:assert! They are pitched at very 
different requirements.

Is that an important use case? Yes, and there is an increasing amount of 
data with similar
characteristics. Is it a common use case?  I think Schematron is not 
used enough for these
kind of issues. 

> Is it fair to say that probably more than 99% of what Schematron is
> actually used for today when used as a supplement to XML Schema, can
> now be done in XML Schema 1.1 directly
I hope it will be welcomed by many people.  99%? I tend to only see 
schemas with
harder constraints that would be supported, so I am probably not a good 
judge.
I don't know that anyone else has any facts either...

Another angle on it is that it would be perfectly possible to implement 
the xs:assert by
converting it to Schematron. (The code that I gave on my website last 
year or so
lets you build the context, and the assertion test would be directly 
copied.) So an
XSD 1.1 implementation could be kept streaming for the non-assertion 
constraints
and the assertions extracted out for in-memory validation.

Cheers
Rick