Closed jbsarrodie closed 3 years ago
I'll look at it...
Might be because of this?
private boolean isNestedTypeRelationship(IArchimateRelationship r) {
return r instanceof ICompositionRelationship
|| r instanceof ISpecializationRelationship
|| r instanceof IAggregationRelationship
|| r instanceof IAssignmentRelationship
|| r instanceof IRealizationRelationship
|| r instanceof IAccessRelationship;
}
No, this looks like if at some point all relations that exist between both elements (without even being used in the view) were taken into consideration and in this case the flow is returned and the warning happens.
I have to check the whole logic.
Maybe also time to take ARM settings into account (there's an old issue of mine on this topic).
No, this looks like if at some point all relations that exist between both elements (without even being used in the view) were taken into consideration and in this case the flow is returned and the warning happens.
Indeed. In the example case, it's reporting that there is a non-nesting relationship...because there is (in the model).
Depends whether this is the desired behaviour, or whether to only check only for diagram instances or to use ARM settings.
it's reporting that there is a non-nesting relationship...because there is (in the model).
Well, that's not the intended behavior, even with the current version (ie. not taking ARM into account).
This should go as follows for this rule: For each nesting:
Commit added to nested-model-checker
branch that should fix the original problem and uses the rules above.
Committed to master. Will be in 4.9.
See the forum post for more information
The validator raises a warning for the attached model while it shouldn't. If we remove the (not used) flow, then the model validates fine.
The model: test.zip