From 7c8c97e635811609c5a7ae4c0bb94e6573c30753 Mon Sep 17 00:00:00 2001
From: Erik Johnston <erik@matrix.org>
Date: Wed, 30 Oct 2019 15:12:49 +0000
Subject: Split purge API into events vs state

---
 tests/storage/test_purge.py | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

(limited to 'tests')

diff --git a/tests/storage/test_purge.py b/tests/storage/test_purge.py
index f671599cb8..b9fafaa1a6 100644
--- a/tests/storage/test_purge.py
+++ b/tests/storage/test_purge.py
@@ -40,23 +40,24 @@ class PurgeTests(HomeserverTestCase):
         third = self.helper.send(self.room_id, body="test3")
         last = self.helper.send(self.room_id, body="test4")
 
-        storage = self.hs.get_datastore()
+        store = self.hs.get_datastore()
+        storage = self.hs.get_storage()
 
         # Get the topological token
-        event = storage.get_topological_token_for_event(last["event_id"])
+        event = store.get_topological_token_for_event(last["event_id"])
         self.pump()
         event = self.successResultOf(event)
 
         # Purge everything before this topological token
-        purge = storage.purge_history(self.room_id, event, True)
+        purge = storage.purge_events.purge_history(self.room_id, event, True)
         self.pump()
         self.assertEqual(self.successResultOf(purge), None)
 
         # Try and get the events
-        get_first = storage.get_event(first["event_id"])
-        get_second = storage.get_event(second["event_id"])
-        get_third = storage.get_event(third["event_id"])
-        get_last = storage.get_event(last["event_id"])
+        get_first = store.get_event(first["event_id"])
+        get_second = store.get_event(second["event_id"])
+        get_third = store.get_event(third["event_id"])
+        get_last = store.get_event(last["event_id"])
         self.pump()
 
         # 1-3 should fail and last will succeed, meaning that 1-3 are deleted
-- 
cgit 1.5.1


From 5c3363233cca7044a333b7e19ba239eaf5587ff8 Mon Sep 17 00:00:00 2001
From: Erik Johnston <erik@matrix.org>
Date: Wed, 6 Nov 2019 17:02:05 +0000
Subject: Fix deleting state groups during room purge.

And fix the tests to actually test that things got deleted.
---
 synapse/storage/data_stores/main/events.py | 27 ++++++++++++++-------------
 tests/rest/admin/test_admin.py             |  4 +++-
 2 files changed, 17 insertions(+), 14 deletions(-)

(limited to 'tests')

diff --git a/synapse/storage/data_stores/main/events.py b/synapse/storage/data_stores/main/events.py
index d69c59f5a1..946823876a 100644
--- a/synapse/storage/data_stores/main/events.py
+++ b/synapse/storage/data_stores/main/events.py
@@ -1633,7 +1633,20 @@ class EventsStore(
         return self.runInteraction("purge_room", self._purge_room_txn, room_id)
 
     def _purge_room_txn(self, txn, room_id):
-        # First delete tables which lack an index on room_id but have one on event_id
+        # First we fetch all the state groups that should be deleted, before
+        # we delete that information.
+        txn.execute(
+            """
+                SELECT DISTINCT state_group FROM events
+                INNER JOIN event_to_state_groups USING(event_id)
+                WHERE events.room_id = ?
+            """,
+            (room_id,),
+        )
+
+        state_groups = [row[0] for row in txn]
+
+        # Now we delete tables which lack an index on room_id but have one on event_id
         for table in (
             "event_auth",
             "event_edges",
@@ -1717,18 +1730,6 @@ class EventsStore(
         #       index on them. In any case we should be clearing out 'stream' tables
         #       periodically anyway (#5888)
 
-        # Now we fetch all the state groups that should be deleted.
-        txn.execute(
-            """
-                SELECT DISTINCT state_group FROM events
-                INNER JOIN event_to_state_groups USING(event_id)
-                WHERE events.room_id = ?
-            """,
-            (room_id,),
-        )
-
-        state_groups = [row[0] for row in txn]
-
         # TODO: we could probably usefully do a bunch of cache invalidation here
 
         logger.info("[purge] done")
diff --git a/tests/rest/admin/test_admin.py b/tests/rest/admin/test_admin.py
index 8e1ca8b738..d9f1b95cb0 100644
--- a/tests/rest/admin/test_admin.py
+++ b/tests/rest/admin/test_admin.py
@@ -628,10 +628,12 @@ class PurgeRoomTestCase(unittest.HomeserverTestCase):
             "local_invites",
             "room_account_data",
             "room_tags",
+            "state_groups",
+            "state_groups_state",
         ):
             count = self.get_success(
                 self.store._simple_select_one_onecol(
-                    table="events",
+                    table=table,
                     keyvalues={"room_id": room_id},
                     retcol="COUNT(*)",
                     desc="test_purge_room",
-- 
cgit 1.5.1