Lists Home |
Date Index |
Validation in the middle of a query is a non-scenario as far as I'm concerned. If not that type assignment occurred as a side effect of validation I'd see no purpose for its existence in XQuery. Being able to swap out validation languages isn't that interesting, being able to swap out type systems is. However since you (as well as the XML Schema working group) has tied them together you conflate issues.
XML Schema already causes problems with regards to type assignment (union types) that are much worse in RELAX NG. However RELAX NG doesn't even do infoset augmentation which is required for the XQuery type system to do anything interesting with it.
Simply being able to swap out schema languages used by a black box to tell you if an XML fragment is valid or not doesn't seem useful in the general case although I'm sure there'll be some gung ho XML gurus who'll love it.
From: Jonathan Robie [mailto:email@example.com]
Sent: Tue 9/30/2003 7:51 AM
To: Dare Obasanjo; Rick Jelliffe; firstname.lastname@example.org
Subject: RE: [xml-dev] FW: [ANN]: XQuery: A Guided Tour
I think that the data types of XML Schema Part 2 would have to be a given.
But this is the most common set of types used in RELAX-NG as well.
On the other hand, validation in XQuery is pretty much a black box.
Currently, we import W3C XML Schemas, and validation is defined by that
spec. But there's nothing to say we couldn't import a different kind of
schema and use that for validation, as long as the named types are properly
assigned, and the built-in atomic types of W3C XML Schema are the ones used
for simple types.