W3C home > Mailing lists > Public > xmlschema-dev@w3.org > March 2001

RE: Canonical XML Schema

From: Robin LaFontaine <robin@monsell.co.uk>
Date: Mon, 5 Mar 2001 14:09:57 +0000
Message-Id: <v04210109b6c94d7de35a@[158.152.20.130]>
To: "Arnold, Curt" <Curt.Arnold@hyprotech.com>
Cc: "'xmlschema-dev@w3.org'" <xmlschema-dev@w3.org>
Curt,

At 10:38 AM -0700 3/2/01, Arnold, Curt wrote:
> > 1. It looks like you were stretching XSL to its limits! Would you
> > advise that XSL is suitable for this type of work or would Java be
> > better? Multi-pass seems an admission of defeat in some senses.
>
>Well, a lot of C++ compilers are multi-pass under the hood and I'd 
>guess that Henry's XSV goes through multiple passes in interpreting 
>a schema.  Just in this case, all the passes are visible which I
>thought was a good thing.  Using XSLT for this basically allowed me 
>to think out loud.  The transforms definitely found bugs in the 
>various XSLT processors that I tried.  While I'm at it, I should
>suggest that serial transformations might be a good thing to address 
>in XSLT 2.0.

Yes, it's often difficult to do a job in one transform, and often XSL 
lends itself either to multi-pass with 'fixed' transforms or pass n 
generates an XSL transform for pass n+1.

>
> > 2. What was the driving purpose of this work? Note clear to me the
> > purpose of the compilation - was it to do instance data validation?
>
>There were multiple intended uses.  Once you had resolved 
>everything, you could generate documentation, schema-specific 
>validators or DOM's.

Our application is in making intelligent comparisons between schema 
structures - we need to know the 'canonical' structure in order to do 
this, much as in validation. I guess there are a lot of us doing very 
similar things here, inevitably!

>
> > 3. Do you have any intention of updating to October CR for XML Schema?
>
>Good intentions, but those haven't been worth much recently.  Rick 
>Jelliffe was curious about using it as part of a XML Schema to 
>Schematron translator.  I could see updating to the next draft with
>the current functionality, however I think it would be unlikely to 
>try to take on redefine or complex content restriction without some 
>external driving force.

I think redefine or complex content restriction should be fun [read: 
hard work!] to process in this way!

Robin

-- -----------------------------------------------------------------
Robin La Fontaine, Monsell EDM Ltd
(XML file comparison, Engineering data exchange and management using 
XML, R&D Project Management)
Tel: +44 1684 592 144 Fax: +44 1684 594 504
Email: robin@monsell.co.uk      http://www.deltaxml.com
Received on Monday, 5 March 2001 11:29:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 11 January 2011 00:14:20 GMT