Philippe Poulard wrote:

 > hi,
 >
 > i try to design a schema language, and i encoutered the famous 
well-known problem :
 > [snip]
 > to express the constraints, we could write :
 > (odd, even)*, odd?
 > but it is invalid in DTD because of the violation of the 
"non-deterministic content model" rule
 >
 > as you know, this problem is unresolvable because it is an 
unambiguous pattern which is not deterministic and can't be rewritten in 
a deterministic form (i picked it from the Eric's relax book)
 >
 > my schema language can express such a content model (4 lines) without 
looking forward, but i'd like to know if :
 > -it is possible with W3C XML Schemas ?


No.

 > -it is possible with Relax NG ?

Yes.

In fact, in the compact syntax, with suitable definitions of odd and 
even, it can be written exactly as you have it: (odd,even)*, odd?.

 > additionally, i'd like to know whether Schematron can be used for 
editing purposes or not (i don't think so), that is to say if one can 
predict if an element is available with a given context, for example.

Don't understand your question. If you're asking whether the above 
constraint could be enforced with Schematron, the answer is yes.

Bob Foster
http://xmlbuddy.com/

(Reply not cross-posted.)

 > thanks for your help