ref https://github.com/DrInfy/sharpy-sc2/issues/54

[eladyaniv01]

knowledge.py

• added attribute self.manager_dict

• modified set_managers to account for the dict assuming MyManager is a manager, it could be accessed like so knowledge.manager_dict[type(MyManager)]

• modified ( and simplified? ) configure_managers usage

 def configure_managers(self) -> Optional[List[ManagerBase]]:
        # override built in data manager
        self.data_manager = MyDataManager()
        self.knowledge.data_manager = self.data_manager

        # add our own custom manager
        self.larva_manager = LarvaManager()
        self.knowledge.larva_manager = self.larva_manager
        self.knowledge.set_managers(additional_managers=[self.larva_manager])

        return None  # Do not add overriding managers to additional managers

[eladyaniv01]

not sure what is the purpose of the return None after the change is accepted i will update the relevant docs on the wiki

[eladyaniv01]

Does not solve any of the requested changes in issue #54 .

can you please be a little more constructive?

However, there's no easy mechanism for accessing the added managers.

custom managers can be accessed via knowledge.manager_dict[type(manager)]

Using type as key has the benefit of some "compile-time" checks. It prevents adding multiple instances of the same manager (probably not needed).

this seems like one of the changes requested

It's important to preserve the order of sharpy's own managers, because the order affects how they are updated.

this was also done

[DrInfy]

With the current code knowledge.manager_dict[type(manager)] does not work. With the current code you need the instance of the manager to request the instance of manager. Which doesn't make any sense. The order of the managers is preserved with zero changes already.