Specific focus of use of "when" in the context of the <<specify>> stereotype usage for Augment
This is needed for TAPI 2.2 and WT
From last week: AI: Scott Mansfield: Use the "when" statement in Yang to qualify which augment is made during the generation of an instance of a thing (and hence what properties will be present in the instance when exposed over an interface)... is the mechanism in the following description valid. If not how would such a structure be defined in Yang. Need to bring Rob Wilton into the discussion.
Context:
Properties defined in a group to be used to augment specific structure
Several groups defined
A structure that has alternative augments
Mechanism
Use of "when" statement in Yang to influence which specific augment chosen
Use of "when" statement during formation of an instance of the structure
Use of a value of a property in the "create data" of the instance to be created to drive the when statement.
Example
A general class, "Profile", that has a number of different allowed sets of properties where each set is defined as an augment and where only one set is allowed in any one profile instance and where the profile has a property which is write/create (i.e., invariant through the life of the profile) that indicates its purpose and hence which augment to apply.
Clearly, considering the data (in XML etc.), there is no issue with a structure with variable content, the key is to state what is expected
New notes:
Karthik highlighted that the method for coding the rules against a <<specify>> had been covered in the UML-Yang guidelines and that this explained how XPath statements could be generated from simple statements in UML.
There was some confusion as to which was the latest UML-Yang guidelines (the version on the wiki (IISOMI Deliverables) is identified as v1.1.02 (Dec 2018) whereas Kam had an apparently more advanced version from an earlier date (v1.1.09 (July6)).
AI: Nigel Davis : compare the two UML-Yang document versions.
It was agreed that the document should be transformed into a wiki.
AI: Karthik Sethuraman: Transform the UML-Yang document into a wiki form.
It was agreed that the WT model work should follow the approach defined (which was to use a constraint rather than the <<Cond>> stereotype (that had been proposed/sketched by Nigel)
There was some discussion on the actual structure of the constraint and the model form. The team walked through the mechanism for construction using Thorsten's model environment.to t
It was agreed that the context link of the constraint should be to the class that is intended to augment (i.e., the class at the tail end of the <<specify>> relationship.
AI: Karthik Sethuraman: Add context link to the description of the constraint when used in conjunction with <<specify>> and update the associated diagram to illustrate this.
AI: Nigel Davis: Update the Core model to align with the usage identified agreed in the meeting.
It was noted that the tooling does not support coding of the when