Re: ANN: examplotron.

[Murata Makoto <mura034@attglobal.net>]

Eric van der Vlist wrote:
>
>John Cowan wrote:
>> Very nice!  The obvious limitation is that recursive models cannot
>> be supported, which should be mentioned in the doco.
>
>Right, I'll add it.


If you see my definition of hedge regular expressions, you will find 
two star operators.  One repeats concatenation of hedges, and you 
already have this.  The other star operator repeats embedding of 
hedges.

You only have to introduce this star operator.  Then, 
examplotron will become as expressive as RELAX and TREX, and more 
expressive than XML Schema and DTD.  It is also possible to convert 
examlotron (even with this extension) to RELAX and TREX.  

More about this, see my PODS (Principles of Database Systems) 2001 
paper at:

	http://www.geocities.com/murata_makoto/podsMURATA.pdf

>It's possible to extend it like I have shown for the definition of the
>occurrences, but then it would loose much of its simplicity and become a
>classical schema language.

Although it is possible to introduce the star operator for embedding, 
I am very sure that schemas containing many occurrences of this star 
operator will become extremely hard to understand.

Cheers,

Makoto



Cheers,

Makoto