Merge pull request #966 from matrix-org/markjh/fix_push

Create separate methods for getting messages to push

1 files changed, 155 insertions, 46 deletions

diff --git a/synapse/storage/event_push_actions.py b/synapse/storage/event_push_actions.py
index 958dbcc22b..df4000d0da 100644
--- a/synapse/storage/event_push_actions.py
+++ b/synapse/storage/event_push_actions.py
@@ -117,40 +117,149 @@ class EventPushActionsStore(SQLBaseStore):
         defer.returnValue(ret)
 
     @defer.inlineCallbacks
-    def get_unread_push_actions_for_user_in_range(self, user_id,
-                                                  min_stream_ordering,
-                                                  max_stream_ordering,
-                                                  limit=20):
+    def get_unread_push_actions_for_user_in_range_for_http(
+        self, user_id, min_stream_ordering, max_stream_ordering, limit=20
+    ):
         """Get a list of the most recent unread push actions for a given user,
-        within the given stream ordering range.
+        within the given stream ordering range. Called by the httppusher.
 
         Args:
-            user_id (str)
-            min_stream_ordering
-            max_stream_ordering
-            limit (int)
+            user_id (str): The user to fetch push actions for.
+            min_stream_ordering(int): The exclusive lower bound on the
+                stream ordering of event push actions to fetch.
+            max_stream_ordering(int): The inclusive upper bound on the
+                stream ordering of event push actions to fetch.
+            limit (int): The maximum number of rows to return.
+        Returns:
+            A promise which resolves to a list of dicts with the keys "event_id",
+            "room_id", "stream_ordering", "actions".
+            The list will be ordered by ascending stream_ordering.
+            The list will have between 0~limit entries.
+        """
+        # find rooms that have a read receipt in them and return the next
+        # push actions
+        def get_after_receipt(txn):
+            # find rooms that have a read receipt in them and return the next
+            # push actions
+            sql = (
+                "SELECT ep.event_id, ep.room_id, ep.stream_ordering, ep.actions"
+                " FROM ("
+                "   SELECT room_id,"
+                "       MAX(topological_ordering) as topological_ordering,"
+                "       MAX(stream_ordering) as stream_ordering"
+                "   FROM events"
+                "   INNER JOIN receipts_linearized USING (room_id, event_id)"
+                "   WHERE receipt_type = 'm.read' AND user_id = ?"
+                "   GROUP BY room_id"
+                ") AS rl,"
+                " event_push_actions AS ep"
+                " WHERE"
+                "   ep.room_id = rl.room_id"
+                "   AND ("
+                "       ep.topological_ordering > rl.topological_ordering"
+                "       OR ("
+                "           ep.topological_ordering = rl.topological_ordering"
+                "           AND ep.stream_ordering > rl.stream_ordering"
+                "       )"
+                "   )"
+                "   AND ep.user_id = ?"
+                "   AND ep.stream_ordering > ?"
+                "   AND ep.stream_ordering <= ?"
+                " ORDER BY ep.stream_ordering ASC LIMIT ?"
+            )
+            args = [
+                user_id, user_id,
+                min_stream_ordering, max_stream_ordering, limit,
+            ]
+            txn.execute(sql, args)
+            return txn.fetchall()
+        after_read_receipt = yield self.runInteraction(
+            "get_unread_push_actions_for_user_in_range_http_arr", get_after_receipt
+        )
+
+        # There are rooms with push actions in them but you don't have a read receipt in
+        # them e.g. rooms you've been invited to, so get push actions for rooms which do
+        # not have read receipts in them too.
+        def get_no_receipt(txn):
+            sql = (
+                "SELECT ep.event_id, ep.room_id, ep.stream_ordering, ep.actions,"
+                " e.received_ts"
+                " FROM event_push_actions AS ep"
+                " INNER JOIN events AS e USING (room_id, event_id)"
+                " WHERE"
+                "   ep.room_id NOT IN ("
+                "     SELECT room_id FROM receipts_linearized"
+                "       WHERE receipt_type = 'm.read' AND user_id = ?"
+                "       GROUP BY room_id"
+                "   )"
+                "   AND ep.user_id = ?"
+                "   AND ep.stream_ordering > ?"
+                "   AND ep.stream_ordering <= ?"
+                " ORDER BY ep.stream_ordering ASC LIMIT ?"
+            )
+            args = [
+                user_id, user_id,
+                min_stream_ordering, max_stream_ordering, limit,
+            ]
+            txn.execute(sql, args)
+            return txn.fetchall()
+        no_read_receipt = yield self.runInteraction(
+            "get_unread_push_actions_for_user_in_range_http_nrr", get_no_receipt
+        )
+
+        notifs = [
+            {
+                "event_id": row[0],
+                "room_id": row[1],
+                "stream_ordering": row[2],
+                "actions": json.loads(row[3]),
+            } for row in after_read_receipt + no_read_receipt
+        ]
+
+        # Now sort it so it's ordered correctly, since currently it will
+        # contain results from the first query, correctly ordered, followed
+        # by results from the second query, but we want them all ordered
+        # by stream_ordering, oldest first.
+        notifs.sort(key=lambda r: r['stream_ordering'])
+
+        # Take only up to the limit. We have to stop at the limit because
+        # one of the subqueries may have hit the limit.
+        defer.returnValue(notifs[:limit])
+
+    @defer.inlineCallbacks
+    def get_unread_push_actions_for_user_in_range_for_email(
+        self, user_id, min_stream_ordering, max_stream_ordering, limit=20
+    ):
+        """Get a list of the most recent unread push actions for a given user,
+        within the given stream ordering range. Called by the emailpusher
+
+        Args:
+            user_id (str): The user to fetch push actions for.
+            min_stream_ordering(int): The exclusive lower bound on the
+                stream ordering of event push actions to fetch.
+            max_stream_ordering(int): The inclusive upper bound on the
+                stream ordering of event push actions to fetch.
+            limit (int): The maximum number of rows to return.
         Returns:
             A promise which resolves to a list of dicts with the keys "event_id",
             "room_id", "stream_ordering", "actions", "received_ts".
+            The list will be ordered by descending received_ts.
             The list will have between 0~limit entries.
         """
         # find rooms that have a read receipt in them and return the most recent
         # push actions
         def get_after_receipt(txn):
-            # XXX: Do we really need to GROUP BY user_id on the inner SELECT?
-            # XXX: NATURAL JOIN obfuscates which columns are being joined on the
-            #      inner SELECT (the room_id and event_id), can we
-            #      INNER JOIN ... USING instead?
             sql = (
-                "SELECT ep.event_id, ep.room_id, ep.stream_ordering, ep.actions, "
-                "e.received_ts "
-                "FROM ("
-                "   SELECT room_id, user_id, "
-                "       max(topological_ordering) as topological_ordering, "
-                "       max(stream_ordering) as stream_ordering "
-                "       FROM events"
-                "   NATURAL JOIN receipts_linearized WHERE receipt_type = 'm.read'"
-                "   GROUP BY room_id, user_id"
+                "SELECT ep.event_id, ep.room_id, ep.stream_ordering, ep.actions,"
+                "  e.received_ts"
+                " FROM ("
+                "   SELECT room_id,"
+                "       MAX(topological_ordering) as topological_ordering,"
+                "       MAX(stream_ordering) as stream_ordering"
+                "   FROM events"
+                "   INNER JOIN receipts_linearized USING (room_id, event_id)"
+                "   WHERE receipt_type = 'm.read' AND user_id = ?"
+                "   GROUP BY room_id"
                 ") AS rl,"
                 " event_push_actions AS ep"
                 " INNER JOIN events AS e USING (room_id, event_id)"
@@ -163,49 +272,49 @@ class EventPushActionsStore(SQLBaseStore):
                 "           AND ep.stream_ordering > rl.stream_ordering"
                 "       )"
                 "   )"
-                "   AND ep.stream_ordering > ?"
                 "   AND ep.user_id = ?"
-                "   AND ep.user_id = rl.user_id"
+                "   AND ep.stream_ordering > ?"
+                "   AND ep.stream_ordering <= ?"
+                " ORDER BY ep.stream_ordering DESC LIMIT ?"
             )
-            args = [min_stream_ordering, user_id]
-            if max_stream_ordering is not None:
-                sql += " AND ep.stream_ordering <= ?"
-                args.append(max_stream_ordering)
-            sql += " ORDER BY ep.stream_ordering DESC LIMIT ?"
-            args.append(limit)
+            args = [
+                user_id, user_id,
+                min_stream_ordering, max_stream_ordering, limit,
+            ]
             txn.execute(sql, args)
             return txn.fetchall()
         after_read_receipt = yield self.runInteraction(
-            "get_unread_push_actions_for_user_in_range", get_after_receipt
+            "get_unread_push_actions_for_user_in_range_email_arr", get_after_receipt
         )
 
         # There are rooms with push actions in them but you don't have a read receipt in
         # them e.g. rooms you've been invited to, so get push actions for rooms which do
         # not have read receipts in them too.
         def get_no_receipt(txn):
-            # XXX: Does the inner SELECT really need to select from the events table?
-            #      We're just extracting the room_id, so isn't receipts_linearized enough?
             sql = (
                 "SELECT ep.event_id, ep.room_id, ep.stream_ordering, ep.actions,"
                 " e.received_ts"
                 " FROM event_push_actions AS ep"
-                " JOIN events e ON ep.room_id = e.room_id AND ep.event_id = e.event_id"
-                " WHERE ep.room_id not in ("
-                "   SELECT room_id FROM events NATURAL JOIN receipts_linearized"
-                "   WHERE receipt_type = 'm.read' AND user_id = ?"
-                "   GROUP BY room_id"
-                ") AND ep.user_id = ? AND ep.stream_ordering > ?"
+                " INNER JOIN events AS e USING (room_id, event_id)"
+                " WHERE"
+                "   ep.room_id NOT IN ("
+                "     SELECT room_id FROM receipts_linearized"
+                "       WHERE receipt_type = 'm.read' AND user_id = ?"
+                "       GROUP BY room_id"
+                "   )"
+                "   AND ep.user_id = ?"
+                "   AND ep.stream_ordering > ?"
+                "   AND ep.stream_ordering <= ?"
+                " ORDER BY ep.stream_ordering DESC LIMIT ?"
             )
-            args = [user_id, user_id, min_stream_ordering]
-            if max_stream_ordering is not None:
-                sql += " AND ep.stream_ordering <= ?"
-                args.append(max_stream_ordering)
-            sql += " ORDER BY ep.stream_ordering DESC LIMIT ?"
-            args.append(limit)
+            args = [
+                user_id, user_id,
+                min_stream_ordering, max_stream_ordering, limit,
+            ]
             txn.execute(sql, args)
             return txn.fetchall()
         no_read_receipt = yield self.runInteraction(
-            "get_unread_push_actions_for_user_in_range", get_no_receipt
+            "get_unread_push_actions_for_user_in_range_email_nrr", get_no_receipt
         )
 
         # Make a list of dicts from the two sets of results.