Refactor EventContext (#12689)

Refactor how the `EventContext` class works, with the intention of reducing the amount of state we fetch from the DB during event processing.

The idea here is to get rid of the cached `current_state_ids` and `prev_state_ids` that live in the `EventContext`, and instead defer straight to the database (and its caching). 

One change that may have a noticeable effect is that we now no longer prefill the `get_current_state_ids` cache on a state change. However, that query is relatively light, since its just a case of reading a table from the DB (unlike fetching state at an event which is more heavyweight). For deployments with workers this cache isn't even used.


Part of #12684

synapse/handlers/federation_event.py

@@ -1423,7 +1423,7 @@ class FederationEventHandler:
                 # we're not bothering about room state, so flag the event as an outlier.
                 event.internal_metadata.outlier = True
 
-                context = EventContext.for_outlier()
+                context = EventContext.for_outlier(self._storage)
                 try:
                     validate_event_for_room_version(room_version_obj, event)
                     check_auth_rules_for_event(room_version_obj, event, auth)
@@ -1874,10 +1874,10 @@ class FederationEventHandler:
         )
 
         return EventContext.with_state(
+            storage=self._storage,
             state_group=state_group,
             state_group_before_event=context.state_group_before_event,
-            current_state_ids=current_state_ids,
-            prev_state_ids=prev_state_ids,
+            state_delta_due_to_event=state_updates,
             prev_group=prev_group,
             delta_ids=state_updates,
             partial_state=context.partial_state,