Re: (Correction) Re: Are we losing out because of grammars?

[Francis Norton <francis@redrice.com>]

James Clark wrote:
> 
...
> 
> But I've never argued that path-based rule systems are inferior to
> grammars.  As I said at the beginning of my last message I think they're
> a valuable complement to grammars, and for *some* problems they're a
> better solution than grammars.  All I'm claiming is that there are also
> some problems for which grammars are a better solution than path-based
> rules.
> 
"*some*" - hmmm... which? 

I've been looking at the Object Constraint Language for UML, used for
expressing features in UML that cannot easily be expressed in the
diagram set.

Page 1 of the book [1] defines constraint as a synonym for assertion.
This makes me think, naturally, of the schematron. Some of OCL could be
expressed in XPath, though to do the whole thing would require switch
and iterator constructs, which I'm idly wondering if could be done in
XSLT.

I'm just noodling around here, but is it possible that there is a
natural mapping between grammars and diagrams which mean that they can
solve the same kind of problems, and that they can both be complemented
in a similar way by assertion-based rule systems?

Francis.

[1]
http://www.amazon.com/exec/obidos/ASIN/0201379406/qid%3D981374393/107-0392824-5712560