I'm using the following strategy for my domain model.
- One file per class, interface, enum, or annotation. Each class in a file only contains basic attributes (string, datetime, integer, decimal, etc).
- One file storing all relations, including inheritance, interface implementation, association, aggregation, and composition.
- Separate class diagram files (each representing a bounded context, AKA a subset of the larger domain model) with @startuml & @enduml directives that !include class/interface/enum/annotation files and the relations file.
The result of this strategy is that I have a high number of hide statements in all class diagram files so that only the classes that I intend to show on the diagram are visible. This gets tedious. Worse, if I add a new relation to the relations file, I have to go check all of the other class diagram files to ensure that I didn't just inadvertently add artifacts to the other class diagram files that include the same relations file.
In what way can I add one or more tags to classes/interfaces/enums/annotations, then, in the class diagram file, say something like show only myFirstTag myOtherTag to only show those classes with tags myFirstTag or myOtherTag so that I don't have to have so many hide statements?
This would be a really useful feature.