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

1 files changed, 5 insertions, 1 deletions

diff --git a/synapse/handlers/message.py b/synapse/handlers/message.py
index c28b792e6f..e47799e7f9 100644
--- a/synapse/handlers/message.py
+++ b/synapse/handlers/message.py
@@ -757,6 +757,10 @@ class EventCreationHandler:
             The previous version of the event is returned, if it is found in the
             event context. Otherwise, None is returned.
         """
+        if event.internal_metadata.is_outlier():
+            # This can happen due to out of band memberships
+            return None
+
         prev_state_ids = await context.get_prev_state_ids()
         prev_event_id = prev_state_ids.get((event.type, event.state_key))
         if not prev_event_id:
@@ -1001,7 +1005,7 @@ class EventCreationHandler:
         # after it is created
         if builder.internal_metadata.outlier:
             event.internal_metadata.outlier = True
-            context = EventContext.for_outlier()
+            context = EventContext.for_outlier(self.storage)
         elif (
             event.type == EventTypes.MSC2716_INSERTION
             and state_event_ids