[
Lists Home |
Date Index |
Thread Index
]
Jonathan Robie wrote:
> At 05:21 PM 9/26/2003, John Cowan wrote:
> >Norman Walsh scripsit:
> > > That is: RELAX NG has an XML syntax, the XML syntax was designed
> > > first, and the language works as an XML grammar. The compact syntax is
> > > a (carefully designed, thoughtfully conceived) add-on.
> >
> >I don't know that it's essential to your point that the XML syntax was
> >designed first (for all we know, James was fiddling with some sort of
> >compact syntax before he ever issued TREX, never mind RELAX NG).
> >I agree that it's important to have an XML syntax as well as the
> >compact syntax, though I have always authored in the compact syntax
> >even before it was official.
>
> FWIW, the first compact syntax I saw for XML Schema was in the XML Query
> Formal Semantics, and was designed by Phil Wadler, Mary Fernandez, and
> Jerome Simeon. It was kicked around in the Working Group for quite a while
> before being published here:
>
> http://www.w3.org/TR/2000/WD-query-algebra-20001204/
>
> I suspect that this compact notation may have influenced the others.
I'd venture a guess that Relax-NG, via TREX, was more
influenced by Brzozowski's 1964 paper "Derivatives of
regular expressions", which uses a purely syntactic
approach. (The derivative algorithm, for instance,
is what makes the interleave operator feasible.)
The algebraic approach to regular languages is familiar
to most mathematicians, but for some reason computer
scientists are, with a few notable exceptions, largely
unaware of it.
--Joe English
jenglish@flightlab.com
|
