summary refs log tree commit diff
path: root/synapse
diff options
context:
space:
mode:
authorDaniel Wagner-Hall <daniel@matrix.org>2015-10-15 11:51:55 +0100
committerDaniel Wagner-Hall <daniel@matrix.org>2015-10-15 11:51:55 +0100
commitf38df51e8dd8e462871c9bfb8fb49a89d157bf2d (patch)
tree08736d0f751f3ee3ee8f0c8f7b1cbce5de5569a7 /synapse
parentUse more descriptive error code (diff)
parentMerge pull request #305 from matrix-org/markjh/v2_sync_api (diff)
downloadsynapse-f38df51e8dd8e462871c9bfb8fb49a89d157bf2d.tar.xz
Merge branch 'develop' into daniel/3pidinvites
Diffstat (limited to '')
-rw-r--r--synapse/api/filtering.py62
-rwxr-xr-xsynapse/app/homeserver.py6
-rw-r--r--synapse/config/logger.py8
-rw-r--r--synapse/handlers/federation.py2
-rw-r--r--synapse/handlers/message.py10
-rw-r--r--synapse/handlers/sync.py199
-rw-r--r--synapse/rest/client/v2_alpha/sync.py170
-rw-r--r--synapse/storage/__init__.py375
-rw-r--r--synapse/storage/engines/postgres.py2
-rw-r--r--synapse/storage/engines/sqlite3.py4
-rw-r--r--synapse/storage/prepare_database.py395
-rw-r--r--synapse/storage/roommember.py14
-rw-r--r--synapse/storage/state.py10
-rw-r--r--synapse/util/debug.py72
14 files changed, 748 insertions, 581 deletions
diff --git a/synapse/api/filtering.py b/synapse/api/filtering.py
index 4d570b74f8..e79e91e7eb 100644
--- a/synapse/api/filtering.py
+++ b/synapse/api/filtering.py
@@ -54,7 +54,7 @@ class Filtering(object):
         ]
 
         room_level_definitions = [
-            "state", "events", "ephemeral"
+            "state", "timeline", "ephemeral"
         ]
 
         for key in top_level_definitions:
@@ -135,17 +135,23 @@ class Filter(object):
     def __init__(self, filter_json):
         self.filter_json = filter_json
 
-    def filter_public_user_data(self, events):
-        return self._filter_on_key(events, ["public_user_data"])
+    def timeline_limit(self):
+        return self.filter_json.get("room", {}).get("timeline", {}).get("limit", 10)
 
-    def filter_private_user_data(self, events):
-        return self._filter_on_key(events, ["private_user_data"])
+    def presence_limit(self):
+        return self.filter_json.get("presence", {}).get("limit", 10)
+
+    def ephemeral_limit(self):
+        return self.filter_json.get("room", {}).get("ephemeral", {}).get("limit", 10)
+
+    def filter_presence(self, events):
+        return self._filter_on_key(events, ["presence"])
 
     def filter_room_state(self, events):
         return self._filter_on_key(events, ["room", "state"])
 
-    def filter_room_events(self, events):
-        return self._filter_on_key(events, ["room", "events"])
+    def filter_room_timeline(self, events):
+        return self._filter_on_key(events, ["room", "timeline"])
 
     def filter_room_ephemeral(self, events):
         return self._filter_on_key(events, ["room", "ephemeral"])
@@ -169,11 +175,34 @@ class Filter(object):
         return [e for e in events if self._passes_definition(definition, e)]
 
     def _passes_definition(self, definition, event):
+        """Check if the event passes the filter definition
+        Args:
+            definition(dict): The filter definition to check against
+            event(dict or Event): The event to check
+        Returns:
+            True if the event passes the filter in the definition
+        """
+        if type(event) is dict:
+            room_id = event.get("room_id")
+            sender = event.get("sender")
+            event_type = event["type"]
+        else:
+            room_id = getattr(event, "room_id", None)
+            sender = getattr(event, "sender", None)
+            event_type = event.type
+        return self._event_passes_definition(
+            definition, room_id, sender, event_type
+        )
+
+    def _event_passes_definition(self, definition, room_id, sender,
+                                 event_type):
         """Check if the event passes through the given definition.
 
         Args:
             definition(dict): The definition to check against.
-            event(Event): The event to check.
+            room_id(str): The id of the room this event is in or None.
+            sender(str): The sender of the event
+            event_type(str): The type of the event.
         Returns:
             True if the event passes through the filter.
         """
@@ -185,8 +214,7 @@ class Filter(object):
         #     and 'not_types' then it is treated as only being in 'not_types')
 
         # room checks
-        if hasattr(event, "room_id"):
-            room_id = event.room_id
+        if room_id is not None:
             allow_rooms = definition.get("rooms", None)
             reject_rooms = definition.get("not_rooms", None)
             if reject_rooms and room_id in reject_rooms:
@@ -195,9 +223,7 @@ class Filter(object):
                 return False
 
         # sender checks
-        if hasattr(event, "sender"):
-            # Should we be including event.state_key for some event types?
-            sender = event.sender
+        if sender is not None:
             allow_senders = definition.get("senders", None)
             reject_senders = definition.get("not_senders", None)
             if reject_senders and sender in reject_senders:
@@ -208,12 +234,12 @@ class Filter(object):
         # type checks
         if "not_types" in definition:
             for def_type in definition["not_types"]:
-                if self._event_matches_type(event, def_type):
+                if self._event_matches_type(event_type, def_type):
                     return False
         if "types" in definition:
             included = False
             for def_type in definition["types"]:
-                if self._event_matches_type(event, def_type):
+                if self._event_matches_type(event_type, def_type):
                     included = True
                     break
             if not included:
@@ -221,9 +247,9 @@ class Filter(object):
 
         return True
 
-    def _event_matches_type(self, event, def_type):
+    def _event_matches_type(self, event_type, def_type):
         if def_type.endswith("*"):
             type_prefix = def_type[:-1]
-            return event.type.startswith(type_prefix)
+            return event_type.startswith(type_prefix)
         else:
-            return event.type == def_type
+            return event_type == def_type
diff --git a/synapse/app/homeserver.py b/synapse/app/homeserver.py
index 190b03e2f7..cf2fa221dc 100755
--- a/synapse/app/homeserver.py
+++ b/synapse/app/homeserver.py
@@ -33,11 +33,9 @@ if __name__ == '__main__':
         sys.stderr.writelines(message)
         sys.exit(1)
 
-
 from synapse.storage.engines import create_engine, IncorrectDatabaseSetup
-from synapse.storage import (
-    are_all_users_on_domain, UpgradeDatabaseException,
-)
+from synapse.storage import are_all_users_on_domain
+from synapse.storage.prepare_database import UpgradeDatabaseException
 
 from synapse.server import HomeServer
 
diff --git a/synapse/config/logger.py b/synapse/config/logger.py
index bd0c17c861..a13dc170c4 100644
--- a/synapse/config/logger.py
+++ b/synapse/config/logger.py
@@ -22,6 +22,7 @@ import yaml
 from string import Template
 import os
 import signal
+from synapse.util.debug import debug_deferreds
 
 
 DEFAULT_LOG_CONFIG = Template("""
@@ -69,6 +70,8 @@ class LoggingConfig(Config):
         self.verbosity = config.get("verbose", 0)
         self.log_config = self.abspath(config.get("log_config"))
         self.log_file = self.abspath(config.get("log_file"))
+        if config.get("full_twisted_stacktraces"):
+            debug_deferreds()
 
     def default_config(self, config_dir_path, server_name, **kwargs):
         log_file = self.abspath("homeserver.log")
@@ -84,6 +87,11 @@ class LoggingConfig(Config):
 
         # A yaml python logging config file
         log_config: "%(log_config)s"
+
+        # Stop twisted from discarding the stack traces of exceptions in
+        # deferreds by waiting a reactor tick before running a deferred's
+        # callbacks.
+        # full_twisted_stacktraces: true
         """ % locals()
 
     def read_arguments(self, args):
diff --git a/synapse/handlers/federation.py b/synapse/handlers/federation.py
index 2b3c4cec8e..946ff97c7d 100644
--- a/synapse/handlers/federation.py
+++ b/synapse/handlers/federation.py
@@ -242,7 +242,7 @@ class FederationHandler(BaseHandler):
     @defer.inlineCallbacks
     def _filter_events_for_server(self, server_name, room_id, events):
         event_to_state = yield self.store.get_state_for_events(
-            room_id, frozenset(e.event_id for e in events),
+            frozenset(e.event_id for e in events),
             types=(
                 (EventTypes.RoomHistoryVisibility, ""),
                 (EventTypes.Member, None),
diff --git a/synapse/handlers/message.py b/synapse/handlers/message.py
index b70258697b..dfeeae76db 100644
--- a/synapse/handlers/message.py
+++ b/synapse/handlers/message.py
@@ -164,7 +164,7 @@ class MessageHandler(BaseHandler):
     @defer.inlineCallbacks
     def _filter_events_for_client(self, user_id, room_id, events):
         event_id_to_state = yield self.store.get_state_for_events(
-            room_id, frozenset(e.event_id for e in events),
+            frozenset(e.event_id for e in events),
             types=(
                 (EventTypes.RoomHistoryVisibility, ""),
                 (EventTypes.Member, user_id),
@@ -290,7 +290,7 @@ class MessageHandler(BaseHandler):
         elif member_event.membership == Membership.LEAVE:
             key = (event_type, state_key)
             room_state = yield self.store.get_state_for_events(
-                room_id, [member_event.event_id], [key]
+                [member_event.event_id], [key]
             )
             data = room_state[member_event.event_id].get(key)
 
@@ -314,7 +314,7 @@ class MessageHandler(BaseHandler):
             room_state = yield self.state_handler.get_current_state(room_id)
         elif member_event.membership == Membership.LEAVE:
             room_state = yield self.store.get_state_for_events(
-                room_id, [member_event.event_id], None
+                [member_event.event_id], None
             )
             room_state = room_state[member_event.event_id]
 
@@ -406,7 +406,7 @@ class MessageHandler(BaseHandler):
                 elif event.membership == Membership.LEAVE:
                     room_end_token = "s%d" % (event.stream_ordering,)
                     deferred_room_state = self.store.get_state_for_events(
-                        event.room_id, [event.event_id], None
+                        [event.event_id], None
                     )
                     deferred_room_state.addCallback(
                         lambda states: states[event.event_id]
@@ -499,7 +499,7 @@ class MessageHandler(BaseHandler):
     def _room_initial_sync_parted(self, user_id, room_id, pagin_config,
                                   member_event):
         room_state = yield self.store.get_state_for_events(
-            member_event.room_id, [member_event.event_id], None
+            [member_event.event_id], None
         )
 
         room_state = room_state[member_event.event_id]
diff --git a/synapse/handlers/sync.py b/synapse/handlers/sync.py
index 9914ff6f9c..21cf50101a 100644
--- a/synapse/handlers/sync.py
+++ b/synapse/handlers/sync.py
@@ -28,21 +28,28 @@ logger = logging.getLogger(__name__)
 
 SyncConfig = collections.namedtuple("SyncConfig", [
     "user",
-    "limit",
-    "gap",
-    "sort",
-    "backfill",
     "filter",
 ])
 
 
-class RoomSyncResult(collections.namedtuple("RoomSyncResult", [
-    "room_id",
-    "limited",
-    "published",
+class TimelineBatch(collections.namedtuple("TimelineBatch", [
+    "prev_batch",
     "events",
+    "limited",
+])):
+    __slots__ = []
+
+    def __nonzero__(self):
+        """Make the result appear empty if there are no updates. This is used
+        to tell if room needs to be part of the sync result.
+        """
+        return bool(self.events)
+
+
+class JoinedSyncResult(collections.namedtuple("JoinedSyncResult", [
+    "room_id",
+    "timeline",
     "state",
-    "prev_batch",
     "ephemeral",
 ])):
     __slots__ = []
@@ -51,14 +58,21 @@ class RoomSyncResult(collections.namedtuple("RoomSyncResult", [
         """Make the result appear empty if there are no updates. This is used
         to tell if room needs to be part of the sync result.
         """
-        return bool(self.events or self.state or self.ephemeral)
+        return bool(self.timeline or self.state or self.ephemeral)
+
+
+class InvitedSyncResult(collections.namedtuple("InvitedSyncResult", [
+    "room_id",
+    "invite",
+])):
+    __slots__ = []
 
 
 class SyncResult(collections.namedtuple("SyncResult", [
     "next_batch",  # Token for the next sync
-    "private_user_data",  # List of private events for the user.
-    "public_user_data",  # List of public events for all users.
-    "rooms",  # RoomSyncResult for each room.
+    "presence",  # List of presence events for the user.
+    "joined",  # JoinedSyncResult for each joined room.
+    "invited",  # InvitedSyncResult for each invited room.
 ])):
     __slots__ = []
 
@@ -68,7 +82,7 @@ class SyncResult(collections.namedtuple("SyncResult", [
         events.
         """
         return bool(
-            self.private_user_data or self.public_user_data or self.rooms
+            self.presence or self.joined or self.invited
         )
 
 
@@ -108,8 +122,8 @@ class SyncHandler(BaseHandler):
                 )
 
             result = yield self.notifier.wait_for_events(
-                sync_config.user, room_ids,
-                sync_config.filter, timeout, current_sync_callback
+                sync_config.user, room_ids, timeout, current_sync_callback,
+                from_token=since_token
             )
             defer.returnValue(result)
 
@@ -121,11 +135,7 @@ class SyncHandler(BaseHandler):
         if since_token is None:
             return self.initial_sync(sync_config)
         else:
-            if sync_config.gap:
-                return self.incremental_sync_with_gap(sync_config, since_token)
-            else:
-                # TODO(mjark): Handle gapless sync
-                raise NotImplementedError()
+            return self.incremental_sync_with_gap(sync_config, since_token)
 
     @defer.inlineCallbacks
     def initial_sync(self, sync_config):
@@ -133,12 +143,6 @@ class SyncHandler(BaseHandler):
         Returns:
             A Deferred SyncResult.
         """
-        if sync_config.sort == "timeline,desc":
-            # TODO(mjark): Handle going through events in reverse order?.
-            # What does "most recent events" mean when applying the limits mean
-            # in this case?
-            raise NotImplementedError()
-
         now_token = yield self.event_sources.get_current_token()
 
         presence_stream = self.event_sources.sources["presence"]
@@ -155,33 +159,36 @@ class SyncHandler(BaseHandler):
             membership_list=[Membership.INVITE, Membership.JOIN]
         )
 
-        # TODO (mjark): Does public mean "published"?
-        published_rooms = yield self.store.get_rooms(is_public=True)
-        published_room_ids = set(r["room_id"] for r in published_rooms)
-
-        rooms = []
+        joined = []
+        invited = []
         for event in room_list:
-            room_sync = yield self.initial_sync_for_room(
-                event.room_id, sync_config, now_token, published_room_ids
-            )
-            rooms.append(room_sync)
+            if event.membership == Membership.JOIN:
+                room_sync = yield self.initial_sync_for_joined_room(
+                    event.room_id, sync_config, now_token,
+                )
+                joined.append(room_sync)
+            elif event.membership == Membership.INVITE:
+                invite = yield self.store.get_event(event.event_id)
+                invited.append(InvitedSyncResult(
+                    room_id=event.room_id,
+                    invite=invite,
+                ))
 
         defer.returnValue(SyncResult(
-            public_user_data=presence,
-            private_user_data=[],
-            rooms=rooms,
+            presence=presence,
+            joined=joined,
+            invited=invited,
             next_batch=now_token,
         ))
 
     @defer.inlineCallbacks
-    def initial_sync_for_room(self, room_id, sync_config, now_token,
-                              published_room_ids):
+    def initial_sync_for_joined_room(self, room_id, sync_config, now_token):
         """Sync a room for a client which is starting without any state
         Returns:
-            A Deferred RoomSyncResult.
+            A Deferred JoinedSyncResult.
         """
 
-        recents, prev_batch_token, limited = yield self.load_filtered_recents(
+        batch = yield self.load_filtered_recents(
             room_id, sync_config, now_token,
         )
 
@@ -190,13 +197,10 @@ class SyncHandler(BaseHandler):
         )
         current_state_events = current_state.values()
 
-        defer.returnValue(RoomSyncResult(
+        defer.returnValue(JoinedSyncResult(
             room_id=room_id,
-            published=room_id in published_room_ids,
-            events=recents,
-            prev_batch=prev_batch_token,
+            timeline=batch,
             state=current_state_events,
-            limited=limited,
             ephemeral=[],
         ))
 
@@ -207,19 +211,13 @@ class SyncHandler(BaseHandler):
         Returns:
             A Deferred SyncResult.
         """
-        if sync_config.sort == "timeline,desc":
-            # TODO(mjark): Handle going through events in reverse order?.
-            # What does "most recent events" mean when applying the limits mean
-            # in this case?
-            raise NotImplementedError()
-
         now_token = yield self.event_sources.get_current_token()
 
         presence_source = self.event_sources.sources["presence"]
         presence, presence_key = yield presence_source.get_new_events_for_user(
             user=sync_config.user,
             from_key=since_token.presence_key,
-            limit=sync_config.limit,
+            limit=sync_config.filter.presence_limit(),
         )
         now_token = now_token.copy_and_replace("presence_key", presence_key)
 
@@ -227,7 +225,7 @@ class SyncHandler(BaseHandler):
         typing, typing_key = yield typing_source.get_new_events_for_user(
             user=sync_config.user,
             from_key=since_token.typing_key,
-            limit=sync_config.limit,
+            limit=sync_config.filter.ephemeral_limit(),
         )
         now_token = now_token.copy_and_replace("typing_key", typing_key)
 
@@ -242,33 +240,37 @@ class SyncHandler(BaseHandler):
         )
         if app_service:
             rooms = yield self.store.get_app_service_rooms(app_service)
-            room_ids = set(r.room_id for r in rooms)
+            joined_room_ids = set(r.room_id for r in rooms)
         else:
-            room_ids = yield rm_handler.get_joined_rooms_for_user(
+            joined_room_ids = yield rm_handler.get_joined_rooms_for_user(
                 sync_config.user
             )
 
-        # TODO (mjark): Does public mean "published"?
-        published_rooms = yield self.store.get_rooms(is_public=True)
-        published_room_ids = set(r["room_id"] for r in published_rooms)
+        timeline_limit = sync_config.filter.timeline_limit()
 
         room_events, _ = yield self.store.get_room_events_stream(
             sync_config.user.to_string(),
             from_key=since_token.room_key,
             to_key=now_token.room_key,
             room_id=None,
-            limit=sync_config.limit + 1,
+            limit=timeline_limit + 1,
         )
 
-        rooms = []
-        if len(room_events) <= sync_config.limit:
+        joined = []
+        if len(room_events) <= timeline_limit:
             # There is no gap in any of the rooms. Therefore we can just
             # partition the new events by room and return them.
+            invite_events = []
             events_by_room_id = {}
             for event in room_events:
                 events_by_room_id.setdefault(event.room_id, []).append(event)
+                if event.room_id not in joined_room_ids:
+                    if (event.type == EventTypes.Member
+                            and event.membership == Membership.INVITE
+                            and event.state_key == sync_config.user.to_string()):
+                        invite_events.append(event)
 
-            for room_id in room_ids:
+            for room_id in joined_room_ids:
                 recents = events_by_room_id.get(room_id, [])
                 state = [event for event in recents if event.is_state()]
                 if recents:
@@ -282,37 +284,47 @@ class SyncHandler(BaseHandler):
                     sync_config, room_id, state
                 )
 
-                room_sync = RoomSyncResult(
+                room_sync = JoinedSyncResult(
                     room_id=room_id,
-                    published=room_id in published_room_ids,
-                    events=recents,
-                    prev_batch=prev_batch,
+                    timeline=TimelineBatch(
+                        events=recents,
+                        prev_batch=prev_batch,
+                        limited=False,
+                    ),
                     state=state,
-                    limited=False,
                     ephemeral=typing_by_room.get(room_id, [])
                 )
                 if room_sync:
-                    rooms.append(room_sync)
+                    joined.append(room_sync)
         else:
-            for room_id in room_ids:
+            invite_events = yield self.store.get_invites_for_user(
+                sync_config.user.to_string()
+            )
+
+            for room_id in joined_room_ids:
                 room_sync = yield self.incremental_sync_with_gap_for_room(
                     room_id, sync_config, since_token, now_token,
-                    published_room_ids, typing_by_room
+                    typing_by_room
                 )
                 if room_sync:
-                    rooms.append(room_sync)
+                    joined.append(room_sync)
+
+        invited = [
+            InvitedSyncResult(room_id=event.room_id, invite=event)
+            for event in invite_events
+        ]
 
         defer.returnValue(SyncResult(
-            public_user_data=presence,
-            private_user_data=[],
-            rooms=rooms,
+            presence=presence,
+            joined=joined,
+            invited=invited,
             next_batch=now_token,
         ))
 
     @defer.inlineCallbacks
     def _filter_events_for_client(self, user_id, room_id, events):
         event_id_to_state = yield self.store.get_state_for_events(
-            room_id, frozenset(e.event_id for e in events),
+            frozenset(e.event_id for e in events),
             types=(
                 (EventTypes.RoomHistoryVisibility, ""),
                 (EventTypes.Member, user_id),
@@ -361,12 +373,13 @@ class SyncHandler(BaseHandler):
         limited = True
         recents = []
         filtering_factor = 2
-        load_limit = max(sync_config.limit * filtering_factor, 100)
+        timeline_limit = sync_config.filter.timeline_limit()
+        load_limit = max(timeline_limit * filtering_factor, 100)
         max_repeat = 3  # Only try a few times per room, otherwise
         room_key = now_token.room_key
         end_key = room_key
 
-        while limited and len(recents) < sync_config.limit and max_repeat:
+        while limited and len(recents) < timeline_limit and max_repeat:
             events, keys = yield self.store.get_recent_events_for_room(
                 room_id,
                 limit=load_limit + 1,
@@ -375,7 +388,7 @@ class SyncHandler(BaseHandler):
             )
             (room_key, _) = keys
             end_key = "s" + room_key.split('-')[-1]
-            loaded_recents = sync_config.filter.filter_room_events(events)
+            loaded_recents = sync_config.filter.filter_room_timeline(events)
             loaded_recents = yield self._filter_events_for_client(
                 sync_config.user.to_string(), room_id, loaded_recents,
             )
@@ -385,34 +398,37 @@ class SyncHandler(BaseHandler):
                 limited = False
             max_repeat -= 1
 
-        if len(recents) > sync_config.limit:
-            recents = recents[-sync_config.limit:]
+        if len(recents) > timeline_limit:
+            limited = True
+            recents = recents[-timeline_limit:]
             room_key = recents[0].internal_metadata.before
 
         prev_batch_token = now_token.copy_and_replace(
             "room_key", room_key
         )
 
-        defer.returnValue((recents, prev_batch_token, limited))
+        defer.returnValue(TimelineBatch(
+            events=recents, prev_batch=prev_batch_token, limited=limited
+        ))
 
     @defer.inlineCallbacks
     def incremental_sync_with_gap_for_room(self, room_id, sync_config,
                                            since_token, now_token,
-                                           published_room_ids, typing_by_room):
+                                           typing_by_room):
         """ Get the incremental delta needed to bring the client up to date for
         the room. Gives the client the most recent events and the changes to
         state.
         Returns:
-            A Deferred RoomSyncResult
+            A Deferred JoinedSyncResult
         """
 
         # TODO(mjark): Check for redactions we might have missed.
 
-        recents, prev_batch_token, limited = yield self.load_filtered_recents(
+        batch = yield self.load_filtered_recents(
             room_id, sync_config, now_token, since_token,
         )
 
-        logging.debug("Recents %r", recents)
+        logging.debug("Recents %r", batch)
 
         # TODO(mjark): This seems racy since this isn't being passed a
         # token to indicate what point in the stream this is
@@ -435,13 +451,10 @@ class SyncHandler(BaseHandler):
             sync_config, room_id, state_events_delta
         )
 
-        room_sync = RoomSyncResult(
+        room_sync = JoinedSyncResult(
             room_id=room_id,
-            published=room_id in published_room_ids,
-            events=recents,
-            prev_batch=prev_batch_token,
+            timeline=batch,
             state=state_events_delta,
-            limited=limited,
             ephemeral=typing_by_room.get(room_id, [])
         )
 
diff --git a/synapse/rest/client/v2_alpha/sync.py b/synapse/rest/client/v2_alpha/sync.py
index cac28b47b6..fffecb24f5 100644
--- a/synapse/rest/client/v2_alpha/sync.py
+++ b/synapse/rest/client/v2_alpha/sync.py
@@ -16,7 +16,7 @@
 from twisted.internet import defer
 
 from synapse.http.servlet import (
-    RestServlet, parse_string, parse_integer, parse_boolean
+    RestServlet, parse_string, parse_integer
 )
 from synapse.handlers.sync import SyncConfig
 from synapse.types import StreamToken
@@ -26,6 +26,7 @@ from synapse.events.utils import (
 from synapse.api.filtering import Filter
 from ._base import client_v2_pattern
 
+import copy
 import logging
 
 logger = logging.getLogger(__name__)
@@ -36,51 +37,44 @@ class SyncRestServlet(RestServlet):
 
     GET parameters::
         timeout(int): How long to wait for new events in milliseconds.
-        limit(int): Maxiumum number of events per room to return.
-        gap(bool): Create gaps the message history if limit is exceeded to
-            ensure that the client has the most recent messages. Defaults to
-            "true".
-        sort(str,str): tuple of sort key (e.g. "timeline") and direction
-            (e.g. "asc", "desc"). Defaults to "timeline,asc".
         since(batch_token): Batch token when asking for incremental deltas.
         set_presence(str): What state the device presence should be set to.
             default is "online".
-        backfill(bool): Should the HS request message history from other
-            servers. This may take a long time making it unsuitable for clients
-            expecting a prompt response. Defaults to "true".
         filter(filter_id): A filter to apply to the events returned.
-        filter_*: Filter override parameters.
 
     Response JSON::
         {
-            "next_batch": // batch token for the next /sync
-            "private_user_data": // private events for this user.
-            "public_user_data": // public events for all users including the
-                                // public events for this user.
-            "rooms": [{ // List of rooms with updates.
-                "room_id": // Id of the room being updated
-                "limited": // Was the per-room event limit exceeded?
-                "published": // Is the room published by our HS?
+          "next_batch": // batch token for the next /sync
+          "presence": // presence data for the user.
+          "rooms": {
+            "joined": { // Joined rooms being updated.
+              "${room_id}": { // Id of the room being updated
                 "event_map": // Map of EventID -> event JSON.
-                "events": { // The recent events in the room if gap is "true"
-                            // otherwise the next events in the room.
-                    "batch": [] // list of EventIDs in the "event_map".
-                    "prev_batch": // back token for getting previous events.
+                "timeline": { // The recent events in the room if gap is "true"
+                  "limited": // Was the per-room event limit exceeded?
+                             // otherwise the next events in the room.
+                  "events": [] // list of EventIDs in the "event_map".
+                  "prev_batch": // back token for getting previous events.
                 }
-                "state": [] // list of EventIDs updating the current state to
-                            // be what it should be at the end of the batch.
-                "ephemeral": []
-            }]
+                "state": {"events": []} // list of EventIDs updating the
+                                        // current state to be what it should
+                                        // be at the end of the batch.
+                "ephemeral": {"events": []} // list of event objects
+              }
+            },
+            "invited": {}, // Invited rooms being updated.
+            "archived": {} // Archived rooms being updated.
+          }
         }
     """
 
     PATTERN = client_v2_pattern("/sync$")
-    ALLOWED_SORT = set(["timeline,asc", "timeline,desc"])
-    ALLOWED_PRESENCE = set(["online", "offline", "idle"])
+    ALLOWED_PRESENCE = set(["online", "offline"])
 
     def __init__(self, hs):
         super(SyncRestServlet, self).__init__()
         self.auth = hs.get_auth()
+        self.event_stream_handler = hs.get_handlers().event_stream_handler
         self.sync_handler = hs.get_handlers().sync_handler
         self.clock = hs.get_clock()
         self.filtering = hs.get_filtering()
@@ -90,45 +84,29 @@ class SyncRestServlet(RestServlet):
         user, token_id = yield self.auth.get_user_by_req(request)
 
         timeout = parse_integer(request, "timeout", default=0)
-        limit = parse_integer(request, "limit", required=True)
-        gap = parse_boolean(request, "gap", default=True)
-        sort = parse_string(
-            request, "sort", default="timeline,asc",
-            allowed_values=self.ALLOWED_SORT
-        )
         since = parse_string(request, "since")
         set_presence = parse_string(
             request, "set_presence", default="online",
             allowed_values=self.ALLOWED_PRESENCE
         )
-        backfill = parse_boolean(request, "backfill", default=False)
         filter_id = parse_string(request, "filter", default=None)
 
         logger.info(
-            "/sync: user=%r, timeout=%r, limit=%r, gap=%r, sort=%r, since=%r,"
-            " set_presence=%r, backfill=%r, filter_id=%r" % (
-                user, timeout, limit, gap, sort, since, set_presence,
-                backfill, filter_id
+            "/sync: user=%r, timeout=%r, since=%r,"
+            " set_presence=%r, filter_id=%r" % (
+                user, timeout, since, set_presence, filter_id
             )
         )
 
-        # TODO(mjark): Load filter and apply overrides.
         try:
             filter = yield self.filtering.get_user_filter(
                 user.localpart, filter_id
             )
         except:
             filter = Filter({})
-        # filter = filter.apply_overrides(http_request)
-        # if filter.matches(event):
-        #   # stuff
 
         sync_config = SyncConfig(
             user=user,
-            gap=gap,
-            limit=limit,
-            sort=sort,
-            backfill=backfill,
             filter=filter,
         )
 
@@ -137,43 +115,81 @@ class SyncRestServlet(RestServlet):
         else:
             since_token = None
 
-        sync_result = yield self.sync_handler.wait_for_sync_for_user(
-            sync_config, since_token=since_token, timeout=timeout
-        )
+        if set_presence == "online":
+            yield self.event_stream_handler.started_stream(user)
+
+        try:
+            sync_result = yield self.sync_handler.wait_for_sync_for_user(
+                sync_config, since_token=since_token, timeout=timeout
+            )
+        finally:
+            if set_presence == "online":
+                self.event_stream_handler.stopped_stream(user)
 
         time_now = self.clock.time_msec()
 
+        joined = self.encode_joined(
+            sync_result.joined, filter, time_now, token_id
+        )
+
+        invited = self.encode_invited(
+            sync_result.invited, filter, time_now, token_id
+        )
+
         response_content = {
-            "public_user_data": self.encode_user_data(
-                sync_result.public_user_data, filter, time_now
-            ),
-            "private_user_data": self.encode_user_data(
-                sync_result.private_user_data, filter, time_now
-            ),
-            "rooms": self.encode_rooms(
-                sync_result.rooms, filter, time_now, token_id
+            "presence": self.encode_presence(
+                sync_result.presence, filter, time_now
             ),
+            "rooms": {
+                "joined": joined,
+                "invited": invited,
+                "archived": {},
+            },
             "next_batch": sync_result.next_batch.to_string(),
         }
 
         defer.returnValue((200, response_content))
 
-    def encode_user_data(self, events, filter, time_now):
-        return events
+    def encode_presence(self, events, filter, time_now):
+        formatted = []
+        for event in events:
+            event = copy.deepcopy(event)
+            event['sender'] = event['content'].pop('user_id')
+            formatted.append(event)
+        return {"events": filter.filter_presence(formatted)}
+
+    def encode_joined(self, rooms, filter, time_now, token_id):
+        joined = {}
+        for room in rooms:
+            joined[room.room_id] = self.encode_room(
+                room, filter, time_now, token_id
+            )
+
+        return joined
+
+    def encode_invited(self, rooms, filter, time_now, token_id):
+        invited = {}
+        for room in rooms:
+            invite = serialize_event(
+                room.invite, time_now, token_id=token_id,
+                event_format=format_event_for_client_v2_without_event_id,
+            )
+            invited_state = invite.get("unsigned", {}).pop("invite_room_state", [])
+            invited_state.append(invite)
+            invited[room.room_id] = {
+                "invite_state": {"events": invited_state}
+            }
 
-    def encode_rooms(self, rooms, filter, time_now, token_id):
-        return [
-            self.encode_room(room, filter, time_now, token_id)
-            for room in rooms
-        ]
+        return invited
 
     @staticmethod
     def encode_room(room, filter, time_now, token_id):
         event_map = {}
         state_events = filter.filter_room_state(room.state)
-        recent_events = filter.filter_room_events(room.events)
+        timeline_events = filter.filter_room_timeline(room.timeline.events)
+        ephemeral_events = filter.filter_room_ephemeral(room.ephemeral)
         state_event_ids = []
-        recent_event_ids = []
+        timeline_event_ids = []
         for event in state_events:
             # TODO(mjark): Respect formatting requirements in the filter.
             event_map[event.event_id] = serialize_event(
@@ -182,24 +198,22 @@ class SyncRestServlet(RestServlet):
             )
             state_event_ids.append(event.event_id)
 
-        for event in recent_events:
+        for event in timeline_events:
             # TODO(mjark): Respect formatting requirements in the filter.
             event_map[event.event_id] = serialize_event(
                 event, time_now, token_id=token_id,
                 event_format=format_event_for_client_v2_without_event_id,
             )
-            recent_event_ids.append(event.event_id)
+            timeline_event_ids.append(event.event_id)
         result = {
-            "room_id": room.room_id,
             "event_map": event_map,
-            "events": {
-                "batch": recent_event_ids,
-                "prev_batch": room.prev_batch.to_string(),
+            "timeline": {
+                "events": timeline_event_ids,
+                "prev_batch": room.timeline.prev_batch.to_string(),
+                "limited": room.timeline.limited,
             },
-            "state": state_event_ids,
-            "limited": room.limited,
-            "published": room.published,
-            "ephemeral": room.ephemeral,
+            "state": {"events": state_event_ids},
+            "ephemeral": {"events": ephemeral_events},
         }
         return result
 
diff --git a/synapse/storage/__init__.py b/synapse/storage/__init__.py
index 340e59afcb..48a0633746 100644
--- a/synapse/storage/__init__.py
+++ b/synapse/storage/__init__.py
@@ -42,22 +42,12 @@ from .end_to_end_keys import EndToEndKeyStore
 from .receipts import ReceiptsStore
 
 
-import fnmatch
-import imp
 import logging
-import os
-import re
 
 
 logger = logging.getLogger(__name__)
 
 
-# Remember to update this number every time a change is made to database
-# schema files, so the users will be informed on server restarts.
-SCHEMA_VERSION = 24
-
-dir_path = os.path.abspath(os.path.dirname(__file__))
-
 # Number of msec of granularity to store the user IP 'last seen' time. Smaller
 # times give more inserts into the database even for readonly API hits
 # 120 seconds == 2 minutes
@@ -158,371 +148,6 @@ class DataStore(RoomMemberStore, RoomStore,
         )
 
 
-def read_schema(path):
-    """ Read the named database schema.
-
-    Args:
-        path: Path of the database schema.
-    Returns:
-        A string containing the database schema.
-    """
-    with open(path) as schema_file:
-        return schema_file.read()
-
-
-class PrepareDatabaseException(Exception):
-    pass
-
-
-class UpgradeDatabaseException(PrepareDatabaseException):
-    pass
-
-
-def prepare_database(db_conn, database_engine):
-    """Prepares a database for usage. Will either create all necessary tables
-    or upgrade from an older schema version.
-    """
-    try:
-        cur = db_conn.cursor()
-        version_info = _get_or_create_schema_state(cur, database_engine)
-
-        if version_info:
-            user_version, delta_files, upgraded = version_info
-            _upgrade_existing_database(
-                cur, user_version, delta_files, upgraded, database_engine
-            )
-        else:
-            _setup_new_database(cur, database_engine)
-
-        # cur.execute("PRAGMA user_version = %d" % (SCHEMA_VERSION,))
-
-        cur.close()
-        db_conn.commit()
-    except:
-        db_conn.rollback()
-        raise
-
-
-def _setup_new_database(cur, database_engine):
-    """Sets up the database by finding a base set of "full schemas" and then
-    applying any necessary deltas.
-
-    The "full_schemas" directory has subdirectories named after versions. This
-    function searches for the highest version less than or equal to
-    `SCHEMA_VERSION` and executes all .sql files in that directory.
-
-    The function will then apply all deltas for all versions after the base
-    version.
-
-    Example directory structure:
-
-        schema/
-            delta/
-                ...
-            full_schemas/
-                3/
-                    test.sql
-                    ...
-                11/
-                    foo.sql
-                    bar.sql
-                ...
-
-    In the example foo.sql and bar.sql would be run, and then any delta files
-    for versions strictly greater than 11.
-    """
-    current_dir = os.path.join(dir_path, "schema", "full_schemas")
-    directory_entries = os.listdir(current_dir)
-
-    valid_dirs = []
-    pattern = re.compile(r"^\d+(\.sql)?$")
-    for filename in directory_entries:
-        match = pattern.match(filename)
-        abs_path = os.path.join(current_dir, filename)
-        if match and os.path.isdir(abs_path):
-            ver = int(match.group(0))
-            if ver <= SCHEMA_VERSION:
-                valid_dirs.append((ver, abs_path))
-        else:
-            logger.warn("Unexpected entry in 'full_schemas': %s", filename)
-
-    if not valid_dirs:
-        raise PrepareDatabaseException(
-            "Could not find a suitable base set of full schemas"
-        )
-
-    max_current_ver, sql_dir = max(valid_dirs, key=lambda x: x[0])
-
-    logger.debug("Initialising schema v%d", max_current_ver)
-
-    directory_entries = os.listdir(sql_dir)
-
-    for filename in fnmatch.filter(directory_entries, "*.sql"):
-        sql_loc = os.path.join(sql_dir, filename)
-        logger.debug("Applying schema %s", sql_loc)
-        executescript(cur, sql_loc)
-
-    cur.execute(
-        database_engine.convert_param_style(
-            "INSERT INTO schema_version (version, upgraded)"
-            " VALUES (?,?)"
-        ),
-        (max_current_ver, False,)
-    )
-
-    _upgrade_existing_database(
-        cur,
-        current_version=max_current_ver,
-        applied_delta_files=[],
-        upgraded=False,
-        database_engine=database_engine,
-    )
-
-
-def _upgrade_existing_database(cur, current_version, applied_delta_files,
-                               upgraded, database_engine):
-    """Upgrades an existing database.
-
-    Delta files can either be SQL stored in *.sql files, or python modules
-    in *.py.
-
-    There can be multiple delta files per version. Synapse will keep track of
-    which delta files have been applied, and will apply any that haven't been
-    even if there has been no version bump. This is useful for development
-    where orthogonal schema changes may happen on separate branches.
-
-    Different delta files for the same version *must* be orthogonal and give
-    the same result when applied in any order. No guarantees are made on the
-    order of execution of these scripts.
-
-    This is a no-op of current_version == SCHEMA_VERSION.
-
-    Example directory structure:
-
-        schema/
-            delta/
-                11/
-                    foo.sql
-                    ...
-                12/
-                    foo.sql
-                    bar.py
-                ...
-            full_schemas/
-                ...
-
-    In the example, if current_version is 11, then foo.sql will be run if and
-    only if `upgraded` is True. Then `foo.sql` and `bar.py` would be run in
-    some arbitrary order.
-
-    Args:
-        cur (Cursor)
-        current_version (int): The current version of the schema.
-        applied_delta_files (list): A list of deltas that have already been
-            applied.
-        upgraded (bool): Whether the current version was generated by having
-            applied deltas or from full schema file. If `True` the function
-            will never apply delta files for the given `current_version`, since
-            the current_version wasn't generated by applying those delta files.
-    """
-
-    if current_version > SCHEMA_VERSION:
-        raise ValueError(
-            "Cannot use this database as it is too " +
-            "new for the server to understand"
-        )
-
-    start_ver = current_version
-    if not upgraded:
-        start_ver += 1
-
-    logger.debug("applied_delta_files: %s", applied_delta_files)
-
-    for v in range(start_ver, SCHEMA_VERSION + 1):
-        logger.debug("Upgrading schema to v%d", v)
-
-        delta_dir = os.path.join(dir_path, "schema", "delta", str(v))
-
-        try:
-            directory_entries = os.listdir(delta_dir)
-        except OSError:
-            logger.exception("Could not open delta dir for version %d", v)
-            raise UpgradeDatabaseException(
-                "Could not open delta dir for version %d" % (v,)
-            )
-
-        directory_entries.sort()
-        for file_name in directory_entries:
-            relative_path = os.path.join(str(v), file_name)
-            logger.debug("Found file: %s", relative_path)
-            if relative_path in applied_delta_files:
-                continue
-
-            absolute_path = os.path.join(
-                dir_path, "schema", "delta", relative_path,
-            )
-            root_name, ext = os.path.splitext(file_name)
-            if ext == ".py":
-                # This is a python upgrade module. We need to import into some
-                # package and then execute its `run_upgrade` function.
-                module_name = "synapse.storage.v%d_%s" % (
-                    v, root_name
-                )
-                with open(absolute_path) as python_file:
-                    module = imp.load_source(
-                        module_name, absolute_path, python_file
-                    )
-                logger.debug("Running script %s", relative_path)
-                module.run_upgrade(cur, database_engine)
-            elif ext == ".pyc":
-                # Sometimes .pyc files turn up anyway even though we've
-                # disabled their generation; e.g. from distribution package
-                # installers. Silently skip it
-                pass
-            elif ext == ".sql":
-                # A plain old .sql file, just read and execute it
-                logger.debug("Applying schema %s", relative_path)
-                executescript(cur, absolute_path)
-            else:
-                # Not a valid delta file.
-                logger.warn(
-                    "Found directory entry that did not end in .py or"
-                    " .sql: %s",
-                    relative_path,
-                )
-                continue
-
-            # Mark as done.
-            cur.execute(
-                database_engine.convert_param_style(
-                    "INSERT INTO applied_schema_deltas (version, file)"
-                    " VALUES (?,?)",
-                ),
-                (v, relative_path)
-            )
-
-            cur.execute("DELETE FROM schema_version")
-            cur.execute(
-                database_engine.convert_param_style(
-                    "INSERT INTO schema_version (version, upgraded)"
-                    " VALUES (?,?)",
-                ),
-                (v, True)
-            )
-
-
-def get_statements(f):
-    statement_buffer = ""
-    in_comment = False  # If we're in a /* ... */ style comment
-
-    for line in f:
-        line = line.strip()
-
-        if in_comment:
-            # Check if this line contains an end to the comment
-            comments = line.split("*/", 1)
-            if len(comments) == 1:
-                continue
-            line = comments[1]
-            in_comment = False
-
-        # Remove inline block comments
-        line = re.sub(r"/\*.*\*/", " ", line)
-
-        # Does this line start a comment?
-        comments = line.split("/*", 1)
-        if len(comments) > 1:
-            line = comments[0]
-            in_comment = True
-
-        # Deal with line comments
-        line = line.split("--", 1)[0]
-        line = line.split("//", 1)[0]
-
-        # Find *all* semicolons. We need to treat first and last entry
-        # specially.
-        statements = line.split(";")
-
-        # We must prepend statement_buffer to the first statement
-        first_statement = "%s %s" % (
-            statement_buffer.strip(),
-            statements[0].strip()
-        )
-        statements[0] = first_statement
-
-        # Every entry, except the last, is a full statement
-        for statement in statements[:-1]:
-            yield statement.strip()
-
-        # The last entry did *not* end in a semicolon, so we store it for the
-        # next semicolon we find
-        statement_buffer = statements[-1].strip()
-
-
-def executescript(txn, schema_path):
-    with open(schema_path, 'r') as f:
-        for statement in get_statements(f):
-            txn.execute(statement)
-
-
-def _get_or_create_schema_state(txn, database_engine):
-    # Bluntly try creating the schema_version tables.
-    schema_path = os.path.join(
-        dir_path, "schema", "schema_version.sql",
-    )
-    executescript(txn, schema_path)
-
-    txn.execute("SELECT version, upgraded FROM schema_version")
-    row = txn.fetchone()
-    current_version = int(row[0]) if row else None
-    upgraded = bool(row[1]) if row else None
-
-    if current_version:
-        txn.execute(
-            database_engine.convert_param_style(
-                "SELECT file FROM applied_schema_deltas WHERE version >= ?"
-            ),
-            (current_version,)
-        )
-        applied_deltas = [d for d, in txn.fetchall()]
-        return current_version, applied_deltas, upgraded
-
-    return None
-
-
-def prepare_sqlite3_database(db_conn):
-    """This function should be called before `prepare_database` on sqlite3
-    databases.
-
-    Since we changed the way we store the current schema version and handle
-    updates to schemas, we need a way to upgrade from the old method to the
-    new. This only affects sqlite databases since they were the only ones
-    supported at the time.
-    """
-    with db_conn:
-        schema_path = os.path.join(
-            dir_path, "schema", "schema_version.sql",
-        )
-        create_schema = read_schema(schema_path)
-        db_conn.executescript(create_schema)
-
-        c = db_conn.execute("SELECT * FROM schema_version")
-        rows = c.fetchall()
-        c.close()
-
-        if not rows:
-            c = db_conn.execute("PRAGMA user_version")
-            row = c.fetchone()
-            c.close()
-
-            if row and row[0]:
-                db_conn.execute(
-                    "REPLACE INTO schema_version (version, upgraded)"
-                    " VALUES (?,?)",
-                    (row[0], False)
-                )
-
-
 def are_all_users_on_domain(txn, database_engine, domain):
     sql = database_engine.convert_param_style(
         "SELECT COUNT(*) FROM users WHERE name NOT LIKE ?"
diff --git a/synapse/storage/engines/postgres.py b/synapse/storage/engines/postgres.py
index 4a855ffd56..98d66e0a86 100644
--- a/synapse/storage/engines/postgres.py
+++ b/synapse/storage/engines/postgres.py
@@ -13,7 +13,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from synapse.storage import prepare_database
+from synapse.storage.prepare_database import prepare_database
 
 from ._base import IncorrectDatabaseSetup
 
diff --git a/synapse/storage/engines/sqlite3.py b/synapse/storage/engines/sqlite3.py
index d18e2808d1..bad3b5c5ac 100644
--- a/synapse/storage/engines/sqlite3.py
+++ b/synapse/storage/engines/sqlite3.py
@@ -13,7 +13,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from synapse.storage import prepare_database, prepare_sqlite3_database
+from synapse.storage.prepare_database import (
+    prepare_database, prepare_sqlite3_database
+)
 
 
 class Sqlite3Engine(object):
diff --git a/synapse/storage/prepare_database.py b/synapse/storage/prepare_database.py
new file mode 100644
index 0000000000..1ddf55be4d
--- /dev/null
+++ b/synapse/storage/prepare_database.py
@@ -0,0 +1,395 @@
+# -*- coding: utf-8 -*-
+# Copyright 2014, 2015 OpenMarket Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# 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.
+
+import fnmatch
+import imp
+import logging
+import os
+import re
+
+
+logger = logging.getLogger(__name__)
+
+
+# Remember to update this number every time a change is made to database
+# schema files, so the users will be informed on server restarts.
+SCHEMA_VERSION = 24
+
+dir_path = os.path.abspath(os.path.dirname(__file__))
+
+
+def read_schema(path):
+    """ Read the named database schema.
+
+    Args:
+        path: Path of the database schema.
+    Returns:
+        A string containing the database schema.
+    """
+    with open(path) as schema_file:
+        return schema_file.read()
+
+
+class PrepareDatabaseException(Exception):
+    pass
+
+
+class UpgradeDatabaseException(PrepareDatabaseException):
+    pass
+
+
+def prepare_database(db_conn, database_engine):
+    """Prepares a database for usage. Will either create all necessary tables
+    or upgrade from an older schema version.
+    """
+    try:
+        cur = db_conn.cursor()
+        version_info = _get_or_create_schema_state(cur, database_engine)
+
+        if version_info:
+            user_version, delta_files, upgraded = version_info
+            _upgrade_existing_database(
+                cur, user_version, delta_files, upgraded, database_engine
+            )
+        else:
+            _setup_new_database(cur, database_engine)
+
+        # cur.execute("PRAGMA user_version = %d" % (SCHEMA_VERSION,))
+
+        cur.close()
+        db_conn.commit()
+    except:
+        db_conn.rollback()
+        raise
+
+
+def _setup_new_database(cur, database_engine):
+    """Sets up the database by finding a base set of "full schemas" and then
+    applying any necessary deltas.
+
+    The "full_schemas" directory has subdirectories named after versions. This
+    function searches for the highest version less than or equal to
+    `SCHEMA_VERSION` and executes all .sql files in that directory.
+
+    The function will then apply all deltas for all versions after the base
+    version.
+
+    Example directory structure:
+
+        schema/
+            delta/
+                ...
+            full_schemas/
+                3/
+                    test.sql
+                    ...
+                11/
+                    foo.sql
+                    bar.sql
+                ...
+
+    In the example foo.sql and bar.sql would be run, and then any delta files
+    for versions strictly greater than 11.
+    """
+    current_dir = os.path.join(dir_path, "schema", "full_schemas")
+    directory_entries = os.listdir(current_dir)
+
+    valid_dirs = []
+    pattern = re.compile(r"^\d+(\.sql)?$")
+    for filename in directory_entries:
+        match = pattern.match(filename)
+        abs_path = os.path.join(current_dir, filename)
+        if match and os.path.isdir(abs_path):
+            ver = int(match.group(0))
+            if ver <= SCHEMA_VERSION:
+                valid_dirs.append((ver, abs_path))
+        else:
+            logger.warn("Unexpected entry in 'full_schemas': %s", filename)
+
+    if not valid_dirs:
+        raise PrepareDatabaseException(
+            "Could not find a suitable base set of full schemas"
+        )
+
+    max_current_ver, sql_dir = max(valid_dirs, key=lambda x: x[0])
+
+    logger.debug("Initialising schema v%d", max_current_ver)
+
+    directory_entries = os.listdir(sql_dir)
+
+    for filename in fnmatch.filter(directory_entries, "*.sql"):
+        sql_loc = os.path.join(sql_dir, filename)
+        logger.debug("Applying schema %s", sql_loc)
+        executescript(cur, sql_loc)
+
+    cur.execute(
+        database_engine.convert_param_style(
+            "INSERT INTO schema_version (version, upgraded)"
+            " VALUES (?,?)"
+        ),
+        (max_current_ver, False,)
+    )
+
+    _upgrade_existing_database(
+        cur,
+        current_version=max_current_ver,
+        applied_delta_files=[],
+        upgraded=False,
+        database_engine=database_engine,
+    )
+
+
+def _upgrade_existing_database(cur, current_version, applied_delta_files,
+                               upgraded, database_engine):
+    """Upgrades an existing database.
+
+    Delta files can either be SQL stored in *.sql files, or python modules
+    in *.py.
+
+    There can be multiple delta files per version. Synapse will keep track of
+    which delta files have been applied, and will apply any that haven't been
+    even if there has been no version bump. This is useful for development
+    where orthogonal schema changes may happen on separate branches.
+
+    Different delta files for the same version *must* be orthogonal and give
+    the same result when applied in any order. No guarantees are made on the
+    order of execution of these scripts.
+
+    This is a no-op of current_version == SCHEMA_VERSION.
+
+    Example directory structure:
+
+        schema/
+            delta/
+                11/
+                    foo.sql
+                    ...
+                12/
+                    foo.sql
+                    bar.py
+                ...
+            full_schemas/
+                ...
+
+    In the example, if current_version is 11, then foo.sql will be run if and
+    only if `upgraded` is True. Then `foo.sql` and `bar.py` would be run in
+    some arbitrary order.
+
+    Args:
+        cur (Cursor)
+        current_version (int): The current version of the schema.
+        applied_delta_files (list): A list of deltas that have already been
+            applied.
+        upgraded (bool): Whether the current version was generated by having
+            applied deltas or from full schema file. If `True` the function
+            will never apply delta files for the given `current_version`, since
+            the current_version wasn't generated by applying those delta files.
+    """
+
+    if current_version > SCHEMA_VERSION:
+        raise ValueError(
+            "Cannot use this database as it is too " +
+            "new for the server to understand"
+        )
+
+    start_ver = current_version
+    if not upgraded:
+        start_ver += 1
+
+    logger.debug("applied_delta_files: %s", applied_delta_files)
+
+    for v in range(start_ver, SCHEMA_VERSION + 1):
+        logger.debug("Upgrading schema to v%d", v)
+
+        delta_dir = os.path.join(dir_path, "schema", "delta", str(v))
+
+        try:
+            directory_entries = os.listdir(delta_dir)
+        except OSError:
+            logger.exception("Could not open delta dir for version %d", v)
+            raise UpgradeDatabaseException(
+                "Could not open delta dir for version %d" % (v,)
+            )
+
+        directory_entries.sort()
+        for file_name in directory_entries:
+            relative_path = os.path.join(str(v), file_name)
+            logger.debug("Found file: %s", relative_path)
+            if relative_path in applied_delta_files:
+                continue
+
+            absolute_path = os.path.join(
+                dir_path, "schema", "delta", relative_path,
+            )
+            root_name, ext = os.path.splitext(file_name)
+            if ext == ".py":
+                # This is a python upgrade module. We need to import into some
+                # package and then execute its `run_upgrade` function.
+                module_name = "synapse.storage.v%d_%s" % (
+                    v, root_name
+                )
+                with open(absolute_path) as python_file:
+                    module = imp.load_source(
+                        module_name, absolute_path, python_file
+                    )
+                logger.debug("Running script %s", relative_path)
+                module.run_upgrade(cur, database_engine)
+            elif ext == ".pyc":
+                # Sometimes .pyc files turn up anyway even though we've
+                # disabled their generation; e.g. from distribution package
+                # installers. Silently skip it
+                pass
+            elif ext == ".sql":
+                # A plain old .sql file, just read and execute it
+                logger.debug("Applying schema %s", relative_path)
+                executescript(cur, absolute_path)
+            else:
+                # Not a valid delta file.
+                logger.warn(
+                    "Found directory entry that did not end in .py or"
+                    " .sql: %s",
+                    relative_path,
+                )
+                continue
+
+            # Mark as done.
+            cur.execute(
+                database_engine.convert_param_style(
+                    "INSERT INTO applied_schema_deltas (version, file)"
+                    " VALUES (?,?)",
+                ),
+                (v, relative_path)
+            )
+
+            cur.execute("DELETE FROM schema_version")
+            cur.execute(
+                database_engine.convert_param_style(
+                    "INSERT INTO schema_version (version, upgraded)"
+                    " VALUES (?,?)",
+                ),
+                (v, True)
+            )
+
+
+def get_statements(f):
+    statement_buffer = ""
+    in_comment = False  # If we're in a /* ... */ style comment
+
+    for line in f:
+        line = line.strip()
+
+        if in_comment:
+            # Check if this line contains an end to the comment
+            comments = line.split("*/", 1)
+            if len(comments) == 1:
+                continue
+            line = comments[1]
+            in_comment = False
+
+        # Remove inline block comments
+        line = re.sub(r"/\*.*\*/", " ", line)
+
+        # Does this line start a comment?
+        comments = line.split("/*", 1)
+        if len(comments) > 1:
+            line = comments[0]
+            in_comment = True
+
+        # Deal with line comments
+        line = line.split("--", 1)[0]
+        line = line.split("//", 1)[0]
+
+        # Find *all* semicolons. We need to treat first and last entry
+        # specially.
+        statements = line.split(";")
+
+        # We must prepend statement_buffer to the first statement
+        first_statement = "%s %s" % (
+            statement_buffer.strip(),
+            statements[0].strip()
+        )
+        statements[0] = first_statement
+
+        # Every entry, except the last, is a full statement
+        for statement in statements[:-1]:
+            yield statement.strip()
+
+        # The last entry did *not* end in a semicolon, so we store it for the
+        # next semicolon we find
+        statement_buffer = statements[-1].strip()
+
+
+def executescript(txn, schema_path):
+    with open(schema_path, 'r') as f:
+        for statement in get_statements(f):
+            txn.execute(statement)
+
+
+def _get_or_create_schema_state(txn, database_engine):
+    # Bluntly try creating the schema_version tables.
+    schema_path = os.path.join(
+        dir_path, "schema", "schema_version.sql",
+    )
+    executescript(txn, schema_path)
+
+    txn.execute("SELECT version, upgraded FROM schema_version")
+    row = txn.fetchone()
+    current_version = int(row[0]) if row else None
+    upgraded = bool(row[1]) if row else None
+
+    if current_version:
+        txn.execute(
+            database_engine.convert_param_style(
+                "SELECT file FROM applied_schema_deltas WHERE version >= ?"
+            ),
+            (current_version,)
+        )
+        applied_deltas = [d for d, in txn.fetchall()]
+        return current_version, applied_deltas, upgraded
+
+    return None
+
+
+def prepare_sqlite3_database(db_conn):
+    """This function should be called before `prepare_database` on sqlite3
+    databases.
+
+    Since we changed the way we store the current schema version and handle
+    updates to schemas, we need a way to upgrade from the old method to the
+    new. This only affects sqlite databases since they were the only ones
+    supported at the time.
+    """
+    with db_conn:
+        schema_path = os.path.join(
+            dir_path, "schema", "schema_version.sql",
+        )
+        create_schema = read_schema(schema_path)
+        db_conn.executescript(create_schema)
+
+        c = db_conn.execute("SELECT * FROM schema_version")
+        rows = c.fetchall()
+        c.close()
+
+        if not rows:
+            c = db_conn.execute("PRAGMA user_version")
+            row = c.fetchone()
+            c.close()
+
+            if row and row[0]:
+                db_conn.execute(
+                    "REPLACE INTO schema_version (version, upgraded)"
+                    " VALUES (?,?)",
+                    (row[0], False)
+                )
diff --git a/synapse/storage/roommember.py b/synapse/storage/roommember.py
index 8c40d9a8a6..dd98dcfda8 100644
--- a/synapse/storage/roommember.py
+++ b/synapse/storage/roommember.py
@@ -110,6 +110,20 @@ class RoomMemberStore(SQLBaseStore):
             membership=membership,
         ).addCallback(self._get_events)
 
+    def get_invites_for_user(self, user_id):
+        """ Get all the invite events for a user
+        Args:
+            user_id (str): The user ID.
+        Returns:
+            A deferred list of event objects.
+        """
+
+        return self.get_rooms_for_user_where_membership_is(
+            user_id, [Membership.INVITE]
+        ).addCallback(lambda invites: self._get_events([
+            invites.event_id for invite in invites
+        ]))
+
     def get_rooms_for_user_where_membership_is(self, user_id, membership_list):
         """ Get all the rooms for this user where the membership for this user
         matches one in the membership list.
diff --git a/synapse/storage/state.py b/synapse/storage/state.py
index e935b9443b..6f2a50d585 100644
--- a/synapse/storage/state.py
+++ b/synapse/storage/state.py
@@ -54,7 +54,7 @@ class StateStore(SQLBaseStore):
             defer.returnValue({})
 
         event_to_groups = yield self._get_state_group_for_events(
-            room_id, event_ids,
+            event_ids,
         )
 
         groups = set(event_to_groups.values())
@@ -208,7 +208,7 @@ class StateStore(SQLBaseStore):
         )
 
     @defer.inlineCallbacks
-    def get_state_for_events(self, room_id, event_ids, types):
+    def get_state_for_events(self, event_ids, types):
         """Given a list of event_ids and type tuples, return a list of state
         dicts for each event. The state dicts will only have the type/state_keys
         that are in the `types` list.
@@ -225,7 +225,7 @@ class StateStore(SQLBaseStore):
             The dicts are mappings from (type, state_key) -> state_events
         """
         event_to_groups = yield self._get_state_group_for_events(
-            room_id, event_ids,
+            event_ids,
         )
 
         groups = set(event_to_groups.values())
@@ -251,8 +251,8 @@ class StateStore(SQLBaseStore):
         )
 
     @cachedList(cache=_get_state_group_for_event.cache, list_name="event_ids",
-                num_args=2)
-    def _get_state_group_for_events(self, room_id, event_ids):
+                num_args=1)
+    def _get_state_group_for_events(self, event_ids):
         """Returns mapping event_id -> state_group
         """
         def f(txn):
diff --git a/synapse/util/debug.py b/synapse/util/debug.py
new file mode 100644
index 0000000000..f6a5a841a4
--- /dev/null
+++ b/synapse/util/debug.py
@@ -0,0 +1,72 @@
+# -*- coding: utf-8 -*-
+# Copyright 2015 OpenMarket Ltd
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# 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 twisted.internet import defer, reactor
+from functools import wraps
+from synapse.util.logcontext import LoggingContext, PreserveLoggingContext
+
+
+def debug_deferreds():
+    """Cause all deferreds to wait for a reactor tick before running their
+    callbacks. This increases the chance of getting a stack trace out of
+    a defer.inlineCallback since the code waiting on the deferred will get
+    a chance to add an errback before the deferred runs."""
+
+    # Helper method for retrieving and restoring the current logging context
+    # around a callback.
+    def with_logging_context(fn):
+        context = LoggingContext.current_context()
+
+        def restore_context_callback(x):
+            with PreserveLoggingContext():
+                LoggingContext.thread_local.current_context = context
+                return fn(x)
+
+        return restore_context_callback
+
+    # We are going to modify the __init__ method of defer.Deferred so we
+    # need to get a copy of the old method so we can still call it.
+    old__init__ = defer.Deferred.__init__
+
+    # We need to create a deferred to bounce the callbacks through the reactor
+    # but we don't want to add a callback when we create that deferred so we
+    # we create a new type of deferred that uses the old __init__ method.
+    # This is safe as long as the old __init__ method doesn't invoke an
+    # __init__ using super.
+    class Bouncer(defer.Deferred):
+        __init__ = old__init__
+
+    # We'll add this as a callback to all Deferreds. Twisted will wait until
+    # the bouncer deferred resolves before calling the callbacks of the
+    # original deferred.
+    def bounce_callback(x):
+        bouncer = Bouncer()
+        reactor.callLater(0, with_logging_context(bouncer.callback), x)
+        return bouncer
+
+    # We'll add this as an errback to all Deferreds. Twisted will wait until
+    # the bouncer deferred resolves before calling the errbacks of the
+    # original deferred.
+    def bounce_errback(x):
+        bouncer = Bouncer()
+        reactor.callLater(0, with_logging_context(bouncer.errback), x)
+        return bouncer
+
+    @wraps(old__init__)
+    def new__init__(self, *args, **kargs):
+        old__init__(self, *args, **kargs)
+        self.addCallbacks(bounce_callback, bounce_errback)
+
+    defer.Deferred.__init__ = new__init__