[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: XSD spec question

• To: "'Gregor'" <iamgregor@g...>,"'Razvan MIHAIU'" <mihaiu@m...>
• Subject: RE: XSD spec question
• From: "Michael Kay" <mike@s...>
• Date: Thu, 28 Apr 2005 08:31:14 +0100
• Cc: "'Eric van der Vlist'" <vdv@d...>,<xml-dev@l...>
• In-reply-to: <642f07b6050427215328609939@m...>
• Thread-index: AcVLroIoPT3GEoBCS6K0nGY6oj3oFwAFO+8w

> 
> The fact that that the Schema for Schemas can not be used for
> validating schemas can be checked here.
> (A reply by Xerces Project Lead Gareth Reakes to my mailinglist post
> on xerces-user-j on apache.org:
> http://blog.parthenoncomputing.com/xerces/archives/2005/04/can
> _i_validate.html
> 

That reply is specific to Xerces, and isn't a general statement that applies
to all schema processors.

The Schema for Schemas does one or two things that user-written schemas
aren't allowed to do, like defining new primitive types. A schema processor
will therefore only be able to perform validation using the S4S if it
recognizes it as being privileged. I had to make a few tweaks to the Saxon
schema processor to ensure that this worked, but it works now.

However, you get much stronger validation of a schema by compiling it than
you get by validating it against the S4S, so the question is a little
academic.

Michael Kay
http://www.saxonica.com/

• References:
  • Re: XSD spec question
    • From: Gregor <iamgregor@g...>

• Prev by Date: Re: XSD spec question
• Next by Date: Re: XSD spec question
• Previous by thread: RE: Re: Language Theorie concerning XML Schema (heavy,at least to me)
• Next by thread: Re: XSD spec question
• Index(es):
  • Date
  • Thread