diff options
author | Mark Haines <mark.haines@matrix.org> | 2015-11-02 10:57:00 +0000 |
---|---|---|
committer | Mark Haines <mark.haines@matrix.org> | 2015-11-02 10:57:00 +0000 |
commit | 0e367563838de76972bac1fd7bc3ab3943b9a57e (patch) | |
tree | f3de1ae204fa232033d8951c31a45a71a797d94c /synapse/storage | |
parent | Support clients supplying older tokens, fix short poll test (diff) | |
parent | Merge pull request #337 from matrix-org/markjh/v2_sync_joining (diff) | |
download | synapse-0e367563838de76972bac1fd7bc3ab3943b9a57e.tar.xz |
Merge branch 'develop' into markjh/room_tags
Diffstat (limited to 'synapse/storage')
-rw-r--r-- | synapse/storage/stream.py | 137 |
1 files changed, 136 insertions, 1 deletions
diff --git a/synapse/storage/stream.py b/synapse/storage/stream.py index 3cab06fdef..15d4c2bf68 100644 --- a/synapse/storage/stream.py +++ b/synapse/storage/stream.py @@ -23,7 +23,7 @@ paginate bacwards. This is implemented by keeping two ordering columns: stream_ordering and topological_ordering. Stream ordering is basically insertion/received order -(except for events from backfill requests). The topolgical_ordering is a +(except for events from backfill requests). The topological_ordering is a weak ordering of events based on the pdu graph. This means that we have to have two different types of tokens, depending on @@ -436,3 +436,138 @@ class StreamStore(SQLBaseStore): internal = event.internal_metadata internal.before = str(RoomStreamToken(topo, stream - 1)) internal.after = str(RoomStreamToken(topo, stream)) + + @defer.inlineCallbacks + def get_events_around(self, room_id, event_id, before_limit, after_limit): + """Retrieve events and pagination tokens around a given event in a + room. + + Args: + room_id (str) + event_id (str) + before_limit (int) + after_limit (int) + + Returns: + dict + """ + + results = yield self.runInteraction( + "get_events_around", self._get_events_around_txn, + room_id, event_id, before_limit, after_limit + ) + + events_before = yield self._get_events( + [e for e in results["before"]["event_ids"]], + get_prev_content=True + ) + + events_after = yield self._get_events( + [e for e in results["after"]["event_ids"]], + get_prev_content=True + ) + + defer.returnValue({ + "events_before": events_before, + "events_after": events_after, + "start": results["before"]["token"], + "end": results["after"]["token"], + }) + + def _get_events_around_txn(self, txn, room_id, event_id, before_limit, after_limit): + """Retrieves event_ids and pagination tokens around a given event in a + room. + + Args: + room_id (str) + event_id (str) + before_limit (int) + after_limit (int) + + Returns: + dict + """ + + results = self._simple_select_one_txn( + txn, + "events", + keyvalues={ + "event_id": event_id, + "room_id": room_id, + }, + retcols=["stream_ordering", "topological_ordering"], + ) + + stream_ordering = results["stream_ordering"] + topological_ordering = results["topological_ordering"] + + query_before = ( + "SELECT topological_ordering, stream_ordering, event_id FROM events" + " WHERE room_id = ? AND (topological_ordering < ?" + " OR (topological_ordering = ? AND stream_ordering < ?))" + " ORDER BY topological_ordering DESC, stream_ordering DESC" + " LIMIT ?" + ) + + query_after = ( + "SELECT topological_ordering, stream_ordering, event_id FROM events" + " WHERE room_id = ? AND (topological_ordering > ?" + " OR (topological_ordering = ? AND stream_ordering > ?))" + " ORDER BY topological_ordering ASC, stream_ordering ASC" + " LIMIT ?" + ) + + txn.execute( + query_before, + ( + room_id, topological_ordering, topological_ordering, + stream_ordering, before_limit, + ) + ) + + rows = self.cursor_to_dict(txn) + events_before = [r["event_id"] for r in rows] + + if rows: + start_token = str(RoomStreamToken( + rows[0]["topological_ordering"], + rows[0]["stream_ordering"] - 1, + )) + else: + start_token = str(RoomStreamToken( + topological_ordering, + stream_ordering - 1, + )) + + txn.execute( + query_after, + ( + room_id, topological_ordering, topological_ordering, + stream_ordering, after_limit, + ) + ) + + rows = self.cursor_to_dict(txn) + events_after = [r["event_id"] for r in rows] + + if rows: + end_token = str(RoomStreamToken( + rows[-1]["topological_ordering"], + rows[-1]["stream_ordering"], + )) + else: + end_token = str(RoomStreamToken( + topological_ordering, + stream_ordering, + )) + + return { + "before": { + "event_ids": events_before, + "token": start_token, + }, + "after": { + "event_ids": events_after, + "token": end_token, + }, + } |