Re: [odata] RE: Topic: Schema - language

[Farrukh Najmi <farrukh@wellfleetsoftware.com>]

Please see inline below...

On 09/12/2012 09:23 AM, Stefan Drees wrote:

Dear Farukh,

On 12.09.12 14:56, Farrukh Najmi wrote:

On 09/12/2012 08:31 AM, Handl, Ralf wrote:

... The main reason why the XSD is "not normative" seems to be
thatit only describes the "minimum requirements" for a CSDL document
to be well-formed, and that additional restrictions apply to make
it correct.
These additional constraints are expressed in the prose
specification document.

That is quite a normal situation which can be met with a normative XSD +
prose in spec that specifies the additional constraints. Ido not see it
as a case for not having a normative XSD.

I don't know whether RELAX NG or schematron can help here;
AtomPub and AtomSyn use RELAX NG to describe the shape of
Atom documents and still need the prose RFC to define what
a correct document is, so RELAX NG seems to have the same
problem as XML Schema.

Schematron and OASIS Content Assembly Model (CAM)
<https://www.oasis-open.org/committees/tc_home.php?wg_abbrev=cam> can
define additional constraints but even they cannot express every
possible constraint.

same with prose ;-)

lol! True. but if you cannot express it in prose then its probably too complex a spec and ought to be simplified.

Ok, I should have suggested two different things (1 and 2 below), so they would not have been partly misunderstood, as they apparantly have been.

0.
DocBook also switched to relax + schematron so this is not new, and the clients may use whatever format they like/need: dtd, xsd, rng or rnc for validation purposes.

1.
May we switch to RELAX NG Compact [B] and derive RELAX NG XML [A] and W3C Xml Schema (XSD)?

Benefit would be to have a by far more readable and easily editable master schema 'csdl.rnc' paired with automatic derivation into the other two formats 'csdl.rng' and 'csdl.xsd'.

I think that having a normative schema (RelaxNG or XSD) is very important. Whether it is RELAX NG Compact [B] or RELAX NG XML [A] and W3C Xml Schema (XSD) seems less important and I can see the benefits of your suggestion above.

2.
SHOULD we use additionally a rule-based validation entity possibly
expressed in the schematron language [C] where possible, to reduce the
additional mandatory "human understandable only" prose to a minimum?

+1 on a three tier schema spec:

• Most of the schema is expressed as a normative RELAX NG Compact [B] , RELAX NG XML [A] or W3C Xml Schema (XSD)
• Additional contraints are specified in a normative Scxhematron file
• Remaining constraints are specified in prose within spec

If we find that the schema and prose are sufficient then we can eliminate schematron and simplify.

Again the part I feel strongly about is that the bulk of the syntax should be specified in a normative schema language such as RELAX NG Compact [B] , RELAX NG XML [A] or W3C Xml Schema (XSD).

Benefit should be clear (apart from the additional short term work of factoring out human prose into schematron etc.)


I hope this puts my proposal into the clear light of already ongoing fruitful discussion ...

Publicly available at ISO i.e. URL
http://standards.iso.org/ittf/PubliclyAvailableStandards/index.html


A) ISO/IEC 19757-2:2008 2nd (Committee: JTC1/SC34)

Information technology
-- Document Schema Definition Language (DSDL)
-- Part 2: Regular-grammar-based validation
-- RELAX NG

B) ISO/IEC 19757-2:2003/Amd.1:2006 1st (Committee: JTC1/SC34)

Information technology
-- Document Schema Definition Language (DSDL)
-- Part 2: Regular-grammar-based validation
-- RELAX NG -- Amendment 1: Compact Syntax

C) ISO/IEC 19757-3:2006 1st (Committee: JTC1/SC34)

Information technology
-- Document Schema Definition Language (DSDL)
-- Part 3: Rule-based validation
-- Schematron     JTC1/SC34

D) ISO/IEC 19757-4:2006 1st (Committee: JTC1/SC34)

Information technology
-- Document Schema Definition Languages (DSDL)
-- Part 4: Namespace-based Validation Dispatching Language (NVDL)



References:

[A]: http://standards.iso.org/ittf/PubliclyAvailableStandards/c052348_ISO_IEC_19757-2_2008(E).zip

[B]: http://standards.iso.org/ittf/PubliclyAvailableStandards/c040774_ISO_IEC_19757-2_2003_Amd_1_2006(E).zip

[C]: http://standards.iso.org/ittf/PubliclyAvailableStandards/c040833_ISO_IEC_19757-3_2006(E).zip

[D]: http://standards.iso.org/ittf/PubliclyAvailableStandards/c038615_ISO_IEC_19757-4_2006(E).zip

<snip/> ongoing discussion ...

All the best,
Stefan.

-- 
Regards,
Farrukh Najmi

Web: http://www.wellfleetsoftware.com