RE: [ubl-dev] Duplicate Definition of Documentation Parameters

[Morgan Kathleen C <Kathleen.C.Morgan@irs.gov>]

I am actually using an unmodified version of the UBL schemas in the xsdrt
directory.  I have two tools - XMYSpy and Visual Studio.  The XMLSpy seems
to validate all the schemas with no errors.  But, my Visual Studio tool,
gives a long list of "Global element has already been defined" errors in the
CommonAggregateComponents schema - but no errors appear in the maindoc
schemas.  

I am actually writing custom schemas and trying to replicate the
structure/syntax of the UBL schemas.  I started getting these errors on my
custom schemas and then attempted to validate the UBL schemas to narrow down
the error to one that I caused or one that I replicated.  

The problem is that we are tyring to implement webMethods as our broker.
That tool seems to import the main schema as well as all of the imported
schema files.  This same type of error appears within webMethods when it
hits the CAC schema.  When we played around with webMethods, the only fix we
came up with was to remove the documentation - and that resolved the
problem.

I was interested in finding out if anyone knew if this was truly a toolset
problem or some sort of problem that I could correct from within the
schemas.  Since we wrote the schemas in XMLSpy, we never hit this problem
and I'm afraid of having to rewrite a bunch of stuff.  I'm thinking that
since XML Spy seems to be ok, that it is a toolset problem - but I just
wans't sure.

Kathleen C. Morgan
PRIME - IRS EDMO
(301) 731-6768
kathleen.c.morgan@irs.gov / kmorgan24@csc.com


-----Original Message-----
From: Chin Chee-Kai [mailto:cheekai@softml.net] 
Sent: Tuesday, December 13, 2005 11:29 PM
To: Morgan Kathleen C
Cc: 'ubl-dev@lists.oasis-open.org'
Subject: Re: [ubl-dev] Duplicate Definition of Documentation Parameters


I vaguely remember this being brought up sometime back about the duplicate
definitions here on ubl-dev, perhaps by XMLSpy?  Outcome of some further
discussion was a quick-fix to just use the set of equivalent schemas which
do not contain the CoreComponentParameters. Those are found in directory
"xsdrt" (stands for XSD Run-Time).

Setting the "Definition" component to minOccurs=0  seems to be an
independent solution from the duplicate definition issue.  The "Definition"
component appears to be required by NDR to be 1..1 actually.

Duplicate definitions are benign in most cases if the schema definitions
imported or included multiple times are exactly the same.  However, 
if such imported or included schema has structurally-modifying xsd 
commands, such as "xsd:redefine", then the importing order may be affected,
usually with more serious errors involved (syntactically or, worse, during
run-time).  As such, a duplicate-definition warning indicating such
potential is a good sign to let us know of such potential issues and set
straight the import sequence and/or remove duplicate
imports.    Hopefully, UBL2.0's xsdrt will be really a run-time version,
rather than a  run-time-cum-avoid-warning  set of schemas.


Best Regards,
Chin Chee-Kai
SoftML
Tel: +65-6820-2979
Fax: +65-6743-7875
Email: cheekai@SoftML.Net
http://SoftML.Net/


On Mon, 12 Dec 2005, Morgan Kathleen C wrote:

>>I am trying to load and validate the .xsd files for UBL 1.0 - however, 
>>I get numerous duplicate definition (global element already defined 
>>and complex type already defined) when I attempt to validate the UBL 
>>schemas in Visual Studio.  The errors are linking back to 
>>CoreComponentsParameters.  I made sure that I have the "Definition" 
>>component minOccurs=0.  But, I still get the errors.
>>
>>I noticed in the UBL 1.0 schemas, the ccts documentation parameters 
>>are used in numerous places -but in the UBL 2.0 schemas (xsdrt.zip) 
>>that were circulated, the ccts tags are no longer there.  Was that to 
>>avoid the error?
>>
>>Any help would be appreciated - I am trying to implement UBL-design 
>>guidance on our agency schemas and I get the same set of duplicate 
>>definition errors and can't load our schemas directly in webMethods 
>>without commenting out the import statements.
>>
>>Thank you,
>>Kathleen C. Morgan
>>
>>