RE: [office] Checking validity of ODF 1.2 Schemas

[robert_weir@us.ibm.com]

"Dennis E. Hamilton" <dennis.hamilton@acm.org> wrote on 04/06/2011 
02:48:11 PM:

> 
> RE: [office] Checking validity of ODF 1.2 Schemas
> 
> Thanks for providing this, Rob.
> 
> I think the only other cases would be any XML fragments in the 
> specification documents themselves.
> 
> I think "expressions" was being used in the OASIS document to refer 
> to cases where there is source code or formulas of some sort.  This 
> might apply to some places in the specification as well, with 
> OpenFormula being a likely place to check for expressions.  Also, 
> the places where BNF or some form of regular expression is used 
> would also be candidates for review.
> 
> I'll see if I can duplicate your checks, just for fun.
> 

I suspect the RNG stuff is pretty clean now.  We've been using it for 
validating text files, code generation for the ODF Toolkit, etc. for some 
time now.  Of course, there is always the possibility that there was a 
last minute stray character dropped in while doing a cut & paste or 
something.  So always worth checking.

The OWL stuff has seen less use, as far as I can tell, and it is something 
the TC has less experience with. So that is certainly worth a look.

Without a definition of "expression" and what "validity" means with 
respect to an expression, I cannot make sense of that other requirement. 
Elsewhere in the same requirement "well-formed" and "valid" are used in 
their formal XML senses.  So it is odd to extend them to non-XML 
constructs and silently slip in the undefined concept of a "valid 
expression".  For example, my impression was that grammatically correct 
EBNF grammars were termed "well-formed", not "valid".

In any case, I think the intent is that formal notations we use be 
grammatically correct with respect to the formal grammars that we claim 
the expressions are governed by.

-Rob

>  - Dennis
> 
> -----Original Message-----
> From: robert_weir@us.ibm.com [mailto:robert_weir@us.ibm.com] 
> Sent: Wednesday, April 06, 2011 11:12
> To: office@lists.oasis-open.org
> Subject: [office] Checking validity of ODF 1.2 Schemas
> 
> I've seen a few comments on the ongoing ballot to certify that our 
> schemas are well-formed/valid, etc.  Before I voted I wanted to do 
> my due diligence.  I'd like to share what I did so anyone who wishes
> to confirm these tests or run a complimentary test can do so.
> 
> ODF 1.2 includes 3 RNG schema files and 2 OWL ontologies.
> 
> The ISO Relax NG standard comes with a normative schema (in Relax NG
> format) that can be used to validate RNG schemas.  You can get this 
> from the standard itself ( http://standards.iso.org/ittf/
> PubliclyAvailableStandards/c037605_ISO_IEC_19757-2_2003(E).zip
> ) or, as I did, accompanying an RNG validator like James Clark's jing.
> 
> I downloaded jing:  http://www.thaiopensource.com/relaxng/jing.html
> 
> Then I unzipped it into a directory, and validated the 3 ODF 1.2 
> schemas, with the following commands and echoed output:
> 
> C:\ODF12Schemas>java -jar ./bin/jing.jar -t lib/relaxng.rng odf12/
> OpenDocument-v1.2-cs01-schema.rng
> Elapsed time 766+141=907 milliseconds
> 
> C:\ODF12Schemas>java -jar ./bin/jing.jar -t lib/relaxng.rng odf12/
> OpenDocument-v1.2-cs01-dsig-schema.rng
> Elapsed time 265+63=328 milliseconds
> 
> C:\ODF12Schemas>java -jar ./bin/jing.jar -t lib/relaxng.rng odf12/
> OpenDocument-v1.2-cs01-manifest-schema.rng
> Elapsed time 265+63=328 milliseconds
> 
> No errors or warnings were reported.
> 
> For the 2 OWL ontologies, I used an online validator: 
> http://www.mygrid.org.uk/OWL/Validator
> 
> I pasted in each of our files, and validated it against "OWL Full". 
> Again, no errors or warnings.
> 
> Although one could imagine a fun rainy day where we then go on 
> recursively to validate the RelaxNG.rng against itself, and verify 
> the EBNF of the W3C's XML 1.0 Rec, I think this would be pointless. 
> So I'll stop here. 
> This is enough for me to cast my vote.
> 
> -Rob
> 
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail.  Follow this link to all your TCs in OASIS at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php 
>