RE: (LINKS) One potential XML mapping for ebXML Core Components

[raymond.e.reeves@mail.sprint.com]

Hi all,
 
At the time, there is an effort within T1M1 to develop a
standard "Mapping of UML notation to XML Schemas (and vice versa)." This
is documented in T1M1/2001-100 and T1M1/2001-132R1.
 
These documents purport to define a mechanism for mapping UML model
management and static view to XML Schemas. In order to take advantage of
a critical set of features existing in XML Schemas (e.g. attributes,
elementGroup, etc.), these documents (implicitly) define a "UML profile
for XML Schemas".
 
These two documents are under review and may be merged together so that
there will be a section describing the UML profile for XML Schemas, a
set of mapping rules from UML notation to XML Schemas and an addendum
for describing additional extensions to support (initially) "Data
Dictionary" like capabilities (now rudimentary in 132) with profile
extensions to support the DD definition and mapping.
 
I believe the Core Components Data Dictionary could be modeled using UML
(and better yet, we could analyze the attributes used in it to make them
part of the DD profile properties -tagged values-) and use "UML profile
for Data Dictionary Definitions" stereotypes, properties (tag values),
etc. to model all what ebCC has defined. I believe we at T1M1 would
benefit from learning about ebXML's approach to define what a "Data
Dictionary" (or simply Data Dictionary Entry) looks like to them and
leverage on it.
 
It would be of great value to us all if a consistent "mapping mechanism"
is defined. As of now, there are several efforts within T1M1 (e.g., TA,
UOM, DSLsp, CLDR) where this mapping is crucial. Furthermore, ITU-T SG4
WP4 Question 9 has a work item for defining a "Mapping of UML to XML
Schemas" and T1M1/2001-100 (and its revisions) is the proposed standard
coming from T1M1.
 
Please consider these documents and provide comments, suggestions as
appropriate on style, approach, etc. It is important to us to have a
broad set of eyes to look into this effort on which many others are
anchored.
 
Thanks,
 
Raymond Reeves
Member of T1M1 tML/DSLsp/UOM AHG/WGs
913-534-3405
 
Links to the documents follow:

100: http://www.t1.org/FileMgr/GetOneFile.taf?FileName=1M10100*

132: http://www.t1.org/FileMgr/GetOneFile.taf?FileName=1M101321

Sample documents that use 100 to map UML to XML include:
http://www.t1.org/FileMgr/GetOneFile.taf?FileName=1M151185
http://www.t1.org/FileMgr/GetOneFile.taf?FileName=1M151643

   -----Original Message-----
   From: bhandspicker [mailto:bhandspicker@foliage.com]
   Sent: Wednesday, September 12, 2001 4:14 PM
   To: ebtwg-ccs; ebxml-dev
   Cc: bhandspicker; knaujok
   Subject: One potential XML mapping for ebXML Core Components
   
   

   
   ebXML Community,
    
   For one of our clients we developed XML Schemas for business
   documents based on the ebXML Core Components.  Of course, this
   required us to create a mapping of the Core Components Data
   Dictionary into a Core Components XML Schema.  With the encouragement
   of our client (who enthusiastically supports public standards
   efforts), we are contributing both "attribute rich" and "attribute
   free" variants of these Core Components XML Schemas for consideration
   by the ebXML and ebTWG community.
    
   We are looking forward to suggestions for improvements, debate over
   style of mapping, comments on our embedded comments, etc.
    
   Brian Handspicker
    
   N.B.  To satisfy UN/CEFACT's goal of openness, Foliage agrees to
   remove any IPR 
   constraints or restrictions associated with this contribution. 
   Foliage offers this contribution for evaluation and standards
   development purposes only, with no implied or explicit warrantee.
   
   --
   Brian D. Handspicker
   Engineering Director
   Foliage Software Systems, Inc.
    
   bhandspicker@foliage.com
   1-781-993-5500
   www.foliage.com