UBL and OAG Common Core Component schemas

[Tim McGrath <tmcgrath@portcomm.com.au>]

It has been a long standing principle that UBL and OAG would try to 
align their implementations of schemas for Core Component Types and Data 
Types. The intention is that this will provide input into the work of a 
mutually agreed upon standards organization such as UN/CEFACT ATG2 or ISO.

In August 2003 both groups started with a common initial set of schemas. 
Since that time these have evolved separately to accommodate design and 
implementation issues both within OAGIS 9.0 and UBL 1.0.

As UBL is about to finalize its 1.0 package it is useful to review these 
differences and attempt to synchronise our developments.

To this end, Garrett Minakawa (representing OAG) and Tim McGrath 
(representing UBL) have reviewed the current OAGIS 9.0 and proposed UBL 
1.0 schemas.

We have identified five areas of misalignment:
1. Naming of Supplementary Components as attributes.
2. Use of XSD normalizedString for code, identifier and text components.
3. Use of XSD built-in dataypes requiring format Supplementary Component 
(Date Time, Indicator and Numeric).
4. Restrictions on Binary Object for Graphic, Picture, Sound and Video 
data type.
5. Patterns for Indicator data type.

We would like to propose the following immediate course of action to 
align these schemas.

Proposed Action Items
------------------------

1. Naming of Supplementary Components as attributes.
* Analysis:
UBL have adopted a naming convention for Supplementary Components based 
on the ObjectClass + PropertyTerm + RepresentationTerm rule that applies 
to BIEs.
OAG have informal naming rules inherited from the initial schemas.
* Proposal:
OAG consider adopting the same naming rules as UBL.

2. Use of XSD normalizedString for code, identifier and text components.
* Analysis:
OAG use the built-in XSD type,"token", for all code, identifier and text 
components (where there is no specific built-in type, such as "language").
UBL uses the built-in XSD type,"normalizedString", for all code and 
identifier components and the built-in XSD type,"string", for all text 
components (where there is no specific built-in type, such as "language").
* Proposal:
OAG consider the built-in XSD type,"normalizedString", for all code, 
identifier and text components (where there is no specific built-in 
type, such as "language").
UBL consider the built-in XSD type,"normalizedString", for all text 
components (where there is no specific built-in type, such as "language").

3. Use of XSD built-in dataypes requiring format Supplementary 
Component(Date Time, Indicator and Numeric).
* Analysis:
OAG explictly define an attribute for "format" in the Core Component 
Type schema. This is then restricted(prohibited) in the data type schema.
UBL do not define an attribute for "format" in the Core Component Type 
schema. This follows UBL Naming and Design rule [STD1]:
"For every ccts:CCT whose supplementary components map directly onto the 
properties of a built-in xsd:datatype, the ccts:CCT MUST be defined as a 
named xsd:simpleType in the ccts:CCT schema module."
* Proposal:
UBL consider relaxing NDR rule STD1 to allow adoption of the OAG approach.

4. Restrictions on Binary Object for Graphic, Picture, Sound and Video 
data type.
* Analysis:
OAG define different attributes for use in data types derived from 
Binary Object (Graphic, Picture, Sound and Video). For example, in OAG a 
Graphic type has characterSetCode,encodingCode,URI and filename whereas 
in UBL, a Graphic type has only mimeCode. (NB this is actually a UBL 
modeling error, it was supposed to have all Supplementary Components 
except the mimeCode).
* Proposal:
UBL consider adopting OAG restrictions for Graphic, Picture, Sound and 
Video data type.

5. Patterns for Indicator data type.
* Analysis:
OAG define a pattern of "true" or "false" for their Indicator data type.
UBL has no pattern.
* Proposal:
UBL consider adopting OAG pattern of "true" and "false" for the 
Indicator data type.


Open Work Items
---------------
We also identified some work areas that both OAG and UBL could jointly 
pursue. These are:

6. Namespace. OAGIS and UBL use different notation and naming in their 
namespace declarations. This should not be a major issue since it is 
expected that OAGIS and UBL will eventually use the same set of common 
core component schema files once they are officially approved and hosted 
by a mutually agreed upon international standards organization such as 
UN/CEFACT ATG2 or ISO. Once this occurs, OAGIS and UBL will simply 
reference the namespace names adopted by the mutually agreed upon 
standards organization. As long as the content of the common core 
component schema files remain unchanged, there should be no visible 
impact to end users.

7. Annotation. OAGIS and UBL have different documentation/annotation 
standards but again, this is not expected to be an issue once the common 
core component schema files are implemented by a mutually agreed upon 
international standards organization.

8.XML Schema Namespace Prefix. OAGIS uses xs: as the namespace prefix 
for http://www.w3.org/2001/XMLSchema. UBL uses the prefix xsd:. As with 
namespaces and annotations, this is not expected to be an issue once the 
common core component schema files are implemented by a mutually agreed 
upon international standards organization.

9. complexType Naming Convention of Representation Terms. UBL has 
appended the term “Type” to the name of all representation terms in 
“UnspecialisedDataTypes.xsd” (e.g. “AmountType” vs. “Amount”).

10. Name of RepresentationTerms schema (vs. UnspecialisedDataTypes)

11. Abbreviation for Identifier (ID vs. Id)

12. Develop a consistent method for representing prohibited attributes 
(and attributes with no changes from the base type) when using 
derivation by restriction.

-- 
regards
tim mcgrath
phone: +618 93352228  
postal: po box 1289   fremantle    western australia 6160