Merge branch 'develop' into send_sni_for_federation_requests

# Conflicts:
#	synapse/http/endpoint.py

8 files changed, 151 insertions, 48 deletions

diff --git a/synapse/handlers/auth.py b/synapse/handlers/auth.py
index a131b7f73f..cbef1f2770 100644
--- a/synapse/handlers/auth.py
+++ b/synapse/handlers/auth.py
@@ -16,6 +16,8 @@
 
 from twisted.internet import defer, threads
 
+from canonicaljson import json
+
 from ._base import BaseHandler
 from synapse.api.constants import LoginType
 from synapse.api.errors import (
@@ -32,7 +34,6 @@ from twisted.web.client import PartialDownloadError
 import logging
 import bcrypt
 import pymacaroons
-import simplejson
 import attr
 
 import synapse.util.stringutils as stringutils
@@ -403,7 +404,7 @@ class AuthHandler(BaseHandler):
         except PartialDownloadError as pde:
             # Twisted is silly
             data = pde.response
-            resp_body = simplejson.loads(data)
+            resp_body = json.loads(data)
 
         if 'success' in resp_body:
             # Note that we do NOT check the hostname here: we explicitly
diff --git a/synapse/handlers/deactivate_account.py b/synapse/handlers/deactivate_account.py
index 8ec5ba2012..a84b7b8b80 100644
--- a/synapse/handlers/deactivate_account.py
+++ b/synapse/handlers/deactivate_account.py
@@ -12,7 +12,7 @@
 # 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 twisted.internet import defer
 
 from ._base import BaseHandler
 from synapse.types import UserID, create_requester
@@ -39,14 +39,15 @@ class DeactivateAccountHandler(BaseHandler):
 
         # Start the user parter loop so it can resume parting users from rooms where
         # it left off (if it has work left to do).
-        reactor.callWhenRunning(self._start_user_parting)
+        hs.get_reactor().callWhenRunning(self._start_user_parting)
 
     @defer.inlineCallbacks
-    def deactivate_account(self, user_id):
+    def deactivate_account(self, user_id, erase_data):
         """Deactivate a user's account
 
         Args:
             user_id (str): ID of user to be deactivated
+            erase_data (bool): whether to GDPR-erase the user's data
 
         Returns:
             Deferred
@@ -92,6 +93,11 @@ class DeactivateAccountHandler(BaseHandler):
         # delete from user directory
         yield self.user_directory_handler.handle_user_deactivated(user_id)
 
+        # Mark the user as erased, if they asked for that
+        if erase_data:
+            logger.info("Marking %s as erased", user_id)
+            yield self.store.mark_user_erased(user_id)
+
         # Now start the process that goes through that list and
         # parts users from rooms (if it isn't already running)
         self._start_user_parting()
diff --git a/synapse/handlers/e2e_keys.py b/synapse/handlers/e2e_keys.py
index 8a2d177539..62b4892a4e 100644
--- a/synapse/handlers/e2e_keys.py
+++ b/synapse/handlers/e2e_keys.py
@@ -14,10 +14,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import simplejson as json
 import logging
 
-from canonicaljson import encode_canonical_json
+from canonicaljson import encode_canonical_json, json
 from twisted.internet import defer
 from six import iteritems
 
@@ -80,7 +79,7 @@ class E2eKeysHandler(object):
             else:
                 remote_queries[user_id] = device_ids
 
-        # Firt get local devices.
+        # First get local devices.
         failures = {}
         results = {}
         if local_query:
@@ -357,7 +356,7 @@ def _exception_to_failure(e):
     # include ConnectionRefused and other errors
     #
     # Note that some Exceptions (notably twisted's ResponseFailed etc) don't
-    # give a string for e.message, which simplejson then fails to serialize.
+    # give a string for e.message, which json then fails to serialize.
     return {
         "status": 503, "message": str(e.message),
     }
diff --git a/synapse/handlers/federation.py b/synapse/handlers/federation.py
index 2571758284..13117d70fe 100644
--- a/synapse/handlers/federation.py
+++ b/synapse/handlers/federation.py
@@ -44,6 +44,7 @@ from synapse.util.frozenutils import unfreeze
 from synapse.crypto.event_signing import (
     compute_event_signature, add_hashes_and_signatures,
 )
+from synapse.state import resolve_events_with_factory
 from synapse.types import UserID, get_domain_from_id
 
 from synapse.events.utils import prune_event
@@ -89,7 +90,9 @@ class FederationHandler(BaseHandler):
 
     @defer.inlineCallbacks
     @log_function
-    def on_receive_pdu(self, origin, pdu, get_missing=True):
+    def on_receive_pdu(
+            self, origin, pdu, get_missing=True, sent_to_us_directly=False,
+    ):
         """ Process a PDU received via a federation /send/ transaction, or
         via backfill of missing prev_events
 
@@ -163,14 +166,11 @@ class FederationHandler(BaseHandler):
                     "Ignoring PDU %s for room %s from %s as we've left the room!",
                     pdu.event_id, pdu.room_id, origin,
                 )
-                return
+                defer.returnValue(None)
 
         state = None
-
         auth_chain = []
 
-        fetch_state = False
-
         # Get missing pdus if necessary.
         if not pdu.internal_metadata.is_outlier():
             # We only backfill backwards to the min depth.
@@ -225,26 +225,60 @@ class FederationHandler(BaseHandler):
                         list(prevs - seen)[:5],
                     )
 
-            if prevs - seen:
-                logger.info(
-                    "Still missing %d events for room %r: %r...",
-                    len(prevs - seen), pdu.room_id, list(prevs - seen)[:5]
+            if sent_to_us_directly and prevs - seen:
+                # If they have sent it to us directly, and the server
+                # isn't telling us about the auth events that it's
+                # made a message referencing, we explode
+                raise FederationError(
+                    "ERROR",
+                    403,
+                    (
+                        "Your server isn't divulging details about prev_events "
+                        "referenced in this event."
+                    ),
+                    affected=pdu.event_id,
                 )
-                fetch_state = True
+            elif prevs - seen:
+                # Calculate the state of the previous events, and
+                # de-conflict them to find the current state.
+                state_groups = []
+                auth_chains = set()
+                try:
+                    # Get the state of the events we know about
+                    ours = yield self.store.get_state_groups(pdu.room_id, list(seen))
+                    state_groups.append(ours)
+
+                    # Ask the remote server for the states we don't
+                    # know about
+                    for p in prevs - seen:
+                        state, got_auth_chain = (
+                            yield self.replication_layer.get_state_for_room(
+                                origin, pdu.room_id, p
+                            )
+                        )
+                        auth_chains.update(got_auth_chain)
+                        state_group = {(x.type, x.state_key): x.event_id for x in state}
+                        state_groups.append(state_group)
+
+                    # Resolve any conflicting state
+                    def fetch(ev_ids):
+                        return self.store.get_events(
+                            ev_ids, get_prev_content=False, check_redacted=False
+                        )
 
-        if fetch_state:
-            # We need to get the state at this event, since we haven't
-            # processed all the prev events.
-            logger.debug(
-                "_handle_new_pdu getting state for %s",
-                pdu.room_id
-            )
-            try:
-                state, auth_chain = yield self.replication_layer.get_state_for_room(
-                    origin, pdu.room_id, pdu.event_id,
-                )
-            except Exception:
-                logger.exception("Failed to get state for event: %s", pdu.event_id)
+                    state_map = yield resolve_events_with_factory(
+                        state_groups, {pdu.event_id: pdu}, fetch
+                    )
+
+                    state = (yield self.store.get_events(state_map.values())).values()
+                    auth_chain = list(auth_chains)
+                except Exception:
+                    raise FederationError(
+                        "ERROR",
+                        403,
+                        "We can't get valid state history.",
+                        affected=pdu.event_id,
+                    )
 
         yield self._process_received_pdu(
             origin,
@@ -322,11 +356,17 @@ class FederationHandler(BaseHandler):
 
         for e in missing_events:
             logger.info("Handling found event %s", e.event_id)
-            yield self.on_receive_pdu(
-                origin,
-                e,
-                get_missing=False
-            )
+            try:
+                yield self.on_receive_pdu(
+                    origin,
+                    e,
+                    get_missing=False
+                )
+            except FederationError as e:
+                if e.code == 403:
+                    logger.warn("Event %s failed history check.")
+                else:
+                    raise
 
     @log_function
     @defer.inlineCallbacks
@@ -460,6 +500,47 @@ class FederationHandler(BaseHandler):
     @measure_func("_filter_events_for_server")
     @defer.inlineCallbacks
     def _filter_events_for_server(self, server_name, room_id, events):
+        """Filter the given events for the given server, redacting those the
+        server can't see.
+
+        Assumes the server is currently in the room.
+
+        Returns
+            list[FrozenEvent]
+        """
+        # First lets check to see if all the events have a history visibility
+        # of "shared" or "world_readable". If thats the case then we don't
+        # need to check membership (as we know the server is in the room).
+        event_to_state_ids = yield self.store.get_state_ids_for_events(
+            frozenset(e.event_id for e in events),
+            types=(
+                (EventTypes.RoomHistoryVisibility, ""),
+            )
+        )
+
+        visibility_ids = set()
+        for sids in event_to_state_ids.itervalues():
+            hist = sids.get((EventTypes.RoomHistoryVisibility, ""))
+            if hist:
+                visibility_ids.add(hist)
+
+        # If we failed to find any history visibility events then the default
+        # is "shared" visiblity.
+        if not visibility_ids:
+            defer.returnValue(events)
+
+        event_map = yield self.store.get_events(visibility_ids)
+        all_open = all(
+            e.content.get("history_visibility") in (None, "shared", "world_readable")
+            for e in event_map.itervalues()
+        )
+
+        if all_open:
+            defer.returnValue(events)
+
+        # Ok, so we're dealing with events that have non-trivial visibility
+        # rules, so we need to also get the memberships of the room.
+
         event_to_state_ids = yield self.store.get_state_ids_for_events(
             frozenset(e.event_id for e in events),
             types=(
@@ -495,7 +576,20 @@ class FederationHandler(BaseHandler):
             for e_id, key_to_eid in event_to_state_ids.iteritems()
         }
 
+        erased_senders = yield self.store.are_users_erased(
+            e.sender for e in events,
+        )
+
         def redact_disallowed(event, state):
+            # if the sender has been gdpr17ed, always return a redacted
+            # copy of the event.
+            if erased_senders[event.sender]:
+                logger.info(
+                    "Sender of %s has been erased, redacting",
+                    event.event_id,
+                )
+                return prune_event(event)
+
             if not state:
                 return event
 
diff --git a/synapse/handlers/identity.py b/synapse/handlers/identity.py
index f00dfe1d3e..277c2b7760 100644
--- a/synapse/handlers/identity.py
+++ b/synapse/handlers/identity.py
@@ -19,7 +19,7 @@
 
 import logging
 
-import simplejson as json
+from canonicaljson import json
 
 from twisted.internet import defer
 
diff --git a/synapse/handlers/message.py b/synapse/handlers/message.py
index 7b9946ab91..cbadf3c88e 100644
--- a/synapse/handlers/message.py
+++ b/synapse/handlers/message.py
@@ -14,13 +14,12 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 import logging
-import simplejson
 import sys
 
-from canonicaljson import encode_canonical_json
+from canonicaljson import encode_canonical_json, json
 import six
 from six import string_types, itervalues, iteritems
-from twisted.internet import defer, reactor
+from twisted.internet import defer
 from twisted.internet.defer import succeed
 from twisted.python.failure import Failure
 
@@ -157,7 +156,7 @@ class MessageHandler(BaseHandler):
             # remove the purge from the list 24 hours after it completes
             def clear_purge():
                 del self._purges_by_id[purge_id]
-            reactor.callLater(24 * 3600, clear_purge)
+            self.hs.get_reactor().callLater(24 * 3600, clear_purge)
 
     def get_purge_status(self, purge_id):
         """Get the current status of an active purge
@@ -491,7 +490,7 @@ class EventCreationHandler(object):
                         target, e
                     )
 
-        is_exempt = yield self._is_exempt_from_privacy_policy(builder)
+        is_exempt = yield self._is_exempt_from_privacy_policy(builder, requester)
         if not is_exempt:
             yield self.assert_accepted_privacy_policy(requester)
 
@@ -509,12 +508,13 @@ class EventCreationHandler(object):
 
         defer.returnValue((event, context))
 
-    def _is_exempt_from_privacy_policy(self, builder):
+    def _is_exempt_from_privacy_policy(self, builder, requester):
         """"Determine if an event to be sent is exempt from having to consent
         to the privacy policy
 
         Args:
             builder (synapse.events.builder.EventBuilder): event being created
+            requester (Requster): user requesting this event
 
         Returns:
             Deferred[bool]: true if the event can be sent without the user
@@ -525,6 +525,9 @@ class EventCreationHandler(object):
             membership = builder.content.get("membership", None)
             if membership == Membership.JOIN:
                 return self._is_server_notices_room(builder.room_id)
+            elif membership == Membership.LEAVE:
+                # the user is always allowed to leave (but not kick people)
+                return builder.state_key == requester.user.to_string()
         return succeed(False)
 
     @defer.inlineCallbacks
@@ -793,7 +796,7 @@ class EventCreationHandler(object):
         # Ensure that we can round trip before trying to persist in db
         try:
             dump = frozendict_json_encoder.encode(event.content)
-            simplejson.loads(dump)
+            json.loads(dump)
         except Exception:
             logger.exception("Failed to encode content: %r", event.content)
             raise
diff --git a/synapse/handlers/presence.py b/synapse/handlers/presence.py
index 7fe568132f..7db59fba00 100644
--- a/synapse/handlers/presence.py
+++ b/synapse/handlers/presence.py
@@ -22,7 +22,7 @@ The methods that define policy are:
     - should_notify
 """
 
-from twisted.internet import defer, reactor
+from twisted.internet import defer
 from contextlib import contextmanager
 
 from six import itervalues, iteritems
@@ -179,7 +179,7 @@ class PresenceHandler(object):
         # have not yet been persisted
         self.unpersisted_users_changes = set()
 
-        reactor.addSystemEventTrigger("before", "shutdown", self._on_shutdown)
+        hs.get_reactor().addSystemEventTrigger("before", "shutdown", self._on_shutdown)
 
         self.serial_to_user = {}
         self._next_serial = 1
diff --git a/synapse/handlers/sync.py b/synapse/handlers/sync.py
index 51ec727df0..7f486e48e5 100644
--- a/synapse/handlers/sync.py
+++ b/synapse/handlers/sync.py
@@ -145,7 +145,7 @@ class SyncResult(collections.namedtuple("SyncResult", [
     "invited",  # InvitedSyncResult for each invited room.
     "archived",  # ArchivedSyncResult for each archived room.
     "to_device",  # List of direct messages for the device.
-    "device_lists",  # List of user_ids whose devices have chanegd
+    "device_lists",  # List of user_ids whose devices have changed
     "device_one_time_keys_count",  # Dict of algorithm to count for one time keys
                                    # for this device
     "groups",