Make EventStream rows have a type

... as a precursor to combining it with the CurrentStateDelta stream.

3 files changed, 94 insertions, 16 deletions

diff --git a/synapse/replication/slave/storage/events.py b/synapse/replication/slave/storage/events.py
index 4830c68f35..c57385d92f 100644
--- a/synapse/replication/slave/storage/events.py
+++ b/synapse/replication/slave/storage/events.py
@@ -16,6 +16,7 @@
 import logging
 
 from synapse.api.constants import EventTypes
+from synapse.replication.tcp.streams.events import EventsStreamEventRow
 from synapse.storage.event_federation import EventFederationWorkerStore
 from synapse.storage.event_push_actions import EventPushActionsWorkerStore
 from synapse.storage.events_worker import EventsWorkerStore
@@ -79,9 +80,12 @@ class SlavedEventStore(EventFederationWorkerStore,
         if stream_name == "events":
             self._stream_id_gen.advance(token)
             for row in rows:
+                if row.type != EventsStreamEventRow.TypeId:
+                    continue
+                data = row.data
                 self.invalidate_caches_for_event(
-                    token, row.event_id, row.room_id, row.type, row.state_key,
-                    row.redacts,
+                    token, data.event_id, data.room_id, data.type, data.state_key,
+                    data.redacts,
                     backfilled=False,
                 )
         elif stream_name == "backfill":
diff --git a/synapse/replication/tcp/protocol.py b/synapse/replication/tcp/protocol.py
index 9daec2c995..b51590cf8f 100644
--- a/synapse/replication/tcp/protocol.py
+++ b/synapse/replication/tcp/protocol.py
@@ -42,8 +42,8 @@ indicate which side is sending, these are *not* included on the wire::
     > POSITION backfill 1
     > POSITION caches 1
     > RDATA caches 2 ["get_user_by_id",["@01register-user:localhost:8823"],1490197670513]
-    > RDATA events 14 ["$149019767112vOHxz:localhost:8823",
-        "!AFDCvgApUmpdfVjIXm:localhost:8823","m.room.guest_access","",null]
+    > RDATA events 14 ["ev", ["$149019767112vOHxz:localhost:8823",
+        "!AFDCvgApUmpdfVjIXm:localhost:8823","m.room.guest_access","",null]]
     < PING 1490197675618
     > ERROR server stopping
     * connection closed by server *
diff --git a/synapse/replication/tcp/streams/events.py b/synapse/replication/tcp/streams/events.py
index 511dd6bcc7..928028e893 100644
--- a/synapse/replication/tcp/streams/events.py
+++ b/synapse/replication/tcp/streams/events.py
@@ -13,28 +13,102 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from collections import namedtuple
+
+import attr
+
+from twisted.internet import defer
 
 from ._base import Stream
 
-EventStreamRow = namedtuple("EventStreamRow", (
-    "event_id",  # str
-    "room_id",  # str
-    "type",  # str
-    "state_key",  # str, optional
-    "redacts",  # str, optional
-))
+
+"""Handling of the 'events' replication stream
+
+This stream contains rows of various types. Each row therefore contains a 'type'
+identifier before the real data. For example::
+
+    RDATA events 12345 ["ev", ["$event:id", "!room:id", "m.type", null, null]]
+
+An "ev" row is sent for each new event. The fields in the data part are:
+
+ * The new event id
+ * The room id for the event
+ * The type of the new event
+ * The state key of the event, for state events
+ * The event id of an event which is redacted by this event.
+
+"""
+
+
+@attr.s(slots=True, frozen=True)
+class EventsStreamRow(object):
+    """A parsed row from the events replication stream"""
+    type = attr.ib()  # str: the TypeId of one of the *EventsStreamRows
+    data = attr.ib()  # BaseEventsStreamRow
+
+
+class BaseEventsStreamRow(object):
+    """Base class for rows to be sent in the events stream.
+
+    Specifies how to identify, serialize and deserialize the different types.
+    """
+
+    TypeId = None  # Unique string that ids the type. Must be overriden in sub classes.
+
+    @classmethod
+    def from_data(cls, data):
+        """Parse the data from the replication stream into a row.
+
+        By default we just call the constructor with the data list as arguments
+
+        Args:
+            data: The value of the data object from the replication stream
+        """
+        return cls(*data)
+
+
+@attr.s(slots=True, frozen=True)
+class EventsStreamEventRow(BaseEventsStreamRow):
+    TypeId = "ev"
+
+    event_id = attr.ib()   # str
+    room_id = attr.ib()    # str
+    type = attr.ib()       # str
+    state_key = attr.ib()  # str, optional
+    redacts = attr.ib()    # str, optional
+
+
+TypeToRow = {
+    Row.TypeId: Row
+    for Row in (
+        EventsStreamEventRow,
+    )
+}
 
 
 class EventsStream(Stream):
     """We received a new event, or an event went from being an outlier to not
     """
     NAME = "events"
-    ROW_TYPE = EventStreamRow
 
     def __init__(self, hs):
-        store = hs.get_datastore()
-        self.current_token = store.get_current_events_token
-        self.update_function = store.get_all_new_forward_event_rows
+        self._store = hs.get_datastore()
+        self.current_token = self._store.get_current_events_token
 
         super(EventsStream, self).__init__(hs)
+
+    @defer.inlineCallbacks
+    def update_function(self, from_token, current_token, limit=None):
+        event_rows = yield self._store.get_all_new_forward_event_rows(
+            from_token, current_token, limit,
+        )
+        event_updates = (
+            (row[0], EventsStreamEventRow.TypeId, row[1:])
+            for row in event_rows
+        )
+        defer.returnValue(event_updates)
+
+    @classmethod
+    def parse_row(cls, row):
+        (typ, data) = row
+        data = TypeToRow[typ].from_data(data)
+        return EventsStreamRow(typ, data)