Presumably a parallel validation is just when each event on the SAX stream gets
sent as input to more than one validator. For example, James' RELAX NG schemas
for XHTML required a separate schema for testing the exclusion exceptions
(i.e. that there should be no html:a//html:a ).
The idea for Schemachine was that you would order schemas that could be
validated together (whether in series or in parallel) into the same <pass>.
Schemas where there was a dependency (such that you only want to validate
with one after you know the document passes another) would go into separate
<pass>es. (Why would you want to have a dependency? Because you can write
simpler rules in a subsequent validation if you know that certain cases won't
crop up. )
Parallel validation does not necessarily mean using SAX, of course.
For example, if you built a DOM, then you could also traverse over
it validing Schematron (and any other schema languages). Indeed,
you can validate two Schematron schemas in parallel just by concatenating
the active patterns from one into the other.
By the way, I have just been talking to some people from the local branch
of one of the big three US-based publishing houses, and it turns out that
they use Schematron a lot. They have an inhouse format for sequencing
validation which they probably will be sending me as an input to DSDL
(presumably for part 0). Interestingly, one of things they regard as a
problem is that they use multiple schemas (for different stages and for
different sources) as the rule rather than the exception. For example,
strict schemas for configuring syntax-directed editors, slack schemas
for validating legacy material, and ultra-slack ones for production
(to get ID/IDREF etc.)
Cheers
Rick Jelliffe
----- Original Message -----
From: "Martin Bryan" <martin@is-thought.co.uk>
To: <dsdl-discuss@dsdl.org>
Sent: Thursday, May 22, 2003 5:00 PM
Subject: [dsdl-discuss] Re: Semantic framework for namespace processing and MNS2
> James
>
> I not clear on your statement:
>
> > - there will always be an explicit <sequence> when you have two
> > validations happening in parallel
>
> What do you mean by "validations happening in parallel"? How can this happen
> in a streaming environment? Are we talking about partially overlapping
> validations (e.g. all elements of namespace x or y), or only about nested
> ones (e.g. attribute validation within elements, child elements within
> elements, etc)?
>
> Martin Bryan
> IS-Thought: Thinkers for the Information Society
> 29 Oldbury Orchard, Churchdown, Glos. GL3 2PU, UK
> Phone: +44 1452 714029 Fax: +44 1452 859991
> E-mail: martin@is-thought.co.uk
>
> Thought of the week:
> I am what I think: I think what I am
>
>
> --
> DSDL members discussion list
>
> To unsubscribe, please send a message with the
> command "unsubscribe" to dsdl-discuss-request@dsdl.org
> (mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)
>
-- DSDL members discussion list To unsubscribe, please send a message with the command "unsubscribe" to dsdl-discuss-request@dsdl.org (mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)Received on Thu May 22 14:45:32 2003
This archive was generated by hypermail 2.1.8 : Fri Dec 03 2004 - 14:00:27 UTC