(Off-topic) Re: Type-assignment (was Re: Are we losing out because ofgrammars?)

[Rick Jelliffe <ricko@allette.com.au>]

From: James Clark <jjc@jclark.com>

> I think I
> can live with fairly trivial type-assignment (as in W3C's XML Schemas),
> which allows type assignment to be done in a single, streaming pass, but

XML-DEVers may be interested that for XML Schemas there was a model
sometimes floating around in WG discussions that went something like this
psuedo-code

  schema-assess(element e) {
        if not(exists(e.namespace.schema))
e.schema=fetch-schema(e.namespace));
        e.psvi.typeInfo = assess-type(e);
        for-each(e.child  e') schema-assess(e');
        e.psvi.validationInfo = validate(e);
  }
in other words type assessment occurs on the way down (e.g., at the start
tag) and validity
assessment occurs on the way up (e.g., at the end-tag).  However, this is
not something that needs to be in the XML Schemas spec, because there is no
reason to constrain people to particular implementation techniques.

Cheers
Rick Jelliffe

n.b. psvi is "post schema-validation infoset"