Why XML Schema enforces UPA (was Re: a or b or both - mystery..)

[ht@cogsci.ed.ac.uk (Henry S. Thompson)]

Joe English <jenglish@flightlab.com> writes:

> Marcus Carr wrote:
> 
> > My understanding was that the requirement for deterministic
> > models was a concession made to those who were writing parsers
> > at the expense of those who were creating DTDs.
> 
> A more important reason is that, in SGML, the rules for 
> when start-tags may be omitted don't make any sense unless 
> content models are deterministic.
> 
> (There is a way to rephrase the rules for start-tag
> omissibility that avoids this, but I don't think it will
> ever make it into a revision of the SGML Standard.)
> 
> At any rate, there's really no need to retain the restriction
> in XML Schemas.  The argument that it makes it easier
> on parser authors is IMO bogus; there are *plenty*
> of well-known, simple algorithms for RE matching that
> have no problem with nondeterministic expressions.
> 
> A somewhat more compelling argument is that it's easier
> for *human beings* to understand deterministic content models,
> but that's better left to the discretion of Schema authors.

Neither of these is the reason it's there in XML Schema.  It's there,
as its name in XML Schema, namely Unique Particle Attribution,
suggests, so _other_ aspects of the particles besides those involved
directly in validation can be relied on, e.g. annotations and
key/keyref/unique declarations.

ht
-- 
  Henry S. Thompson, HCRC Language Technology Group, University of Edinburgh
          W3C Fellow 1999--2001, part-time member of W3C Team
     2 Buccleuch Place, Edinburgh EH8 9LW, SCOTLAND -- (44) 131 650-4440
	    Fax: (44) 131 650-4587, e-mail: ht@cogsci.ed.ac.uk
		     URL: http://www.ltg.ed.ac.uk/~ht/