Prevent redacted events from appearing in message search (#6377)

6 files changed, 78 insertions, 43 deletions

diff --git a/changelog.d/6377.bugfix b/changelog.d/6377.bugfix
new file mode 100644
index 0000000000..ccda96962f
--- /dev/null
+++ b/changelog.d/6377.bugfix
@@ -0,0 +1 @@
+Prevent redacted events from being returned during message search.
\ No newline at end of file
diff --git a/synapse/handlers/federation.py b/synapse/handlers/federation.py
index 13865c470c..8f3c9d7702 100644
--- a/synapse/handlers/federation.py
+++ b/synapse/handlers/federation.py
@@ -63,6 +63,7 @@ from synapse.replication.http.federation import (
 )
 from synapse.replication.http.membership import ReplicationUserJoinedLeftRoomRestServlet
 from synapse.state import StateResolutionStore, resolve_events_with_store
+from synapse.storage.data_stores.main.events_worker import EventRedactBehaviour
 from synapse.types import UserID, get_domain_from_id
 from synapse.util import batch_iter, unwrapFirstError
 from synapse.util.async_helpers import Linearizer
@@ -423,7 +424,7 @@ class FederationHandler(BaseHandler):
                     evs = yield self.store.get_events(
                         list(state_map.values()),
                         get_prev_content=False,
-                        check_redacted=False,
+                        redact_behaviour=EventRedactBehaviour.AS_IS,
                     )
                     event_map.update(evs)
 
@@ -1000,7 +1001,9 @@ class FederationHandler(BaseHandler):
         forward_events = yield self.store.get_successor_events(list(extremities))
 
         extremities_events = yield self.store.get_events(
-            forward_events, check_redacted=False, get_prev_content=False
+            forward_events,
+            redact_behaviour=EventRedactBehaviour.AS_IS,
+            get_prev_content=False,
         )
 
         # We set `check_history_visibility_only` as we might otherwise get false
diff --git a/synapse/handlers/message.py b/synapse/handlers/message.py
index 54fa216d83..bf9add7fe2 100644
--- a/synapse/handlers/message.py
+++ b/synapse/handlers/message.py
@@ -46,6 +46,7 @@ from synapse.events.validator import EventValidator
 from synapse.logging.context import run_in_background
 from synapse.metrics.background_process_metrics import run_as_background_process
 from synapse.replication.http.send_event import ReplicationSendEventRestServlet
+from synapse.storage.data_stores.main.events_worker import EventRedactBehaviour
 from synapse.storage.state import StateFilter
 from synapse.types import RoomAlias, UserID, create_requester
 from synapse.util.async_helpers import Linearizer
@@ -875,7 +876,7 @@ class EventCreationHandler(object):
             if event.type == EventTypes.Redaction:
                 original_event = yield self.store.get_event(
                     event.redacts,
-                    check_redacted=False,
+                    redact_behaviour=EventRedactBehaviour.AS_IS,
                     get_prev_content=False,
                     allow_rejected=False,
                     allow_none=True,
@@ -952,7 +953,7 @@ class EventCreationHandler(object):
         if event.type == EventTypes.Redaction:
             original_event = yield self.store.get_event(
                 event.redacts,
-                check_redacted=False,
+                redact_behaviour=EventRedactBehaviour.AS_IS,
                 get_prev_content=False,
                 allow_rejected=False,
                 allow_none=True,
diff --git a/synapse/state/__init__.py b/synapse/state/__init__.py
index 139beef8ed..3e6d62eef1 100644
--- a/synapse/state/__init__.py
+++ b/synapse/state/__init__.py
@@ -32,6 +32,7 @@ from synapse.events import EventBase
 from synapse.events.snapshot import EventContext
 from synapse.logging.utils import log_function
 from synapse.state import v1, v2
+from synapse.storage.data_stores.main.events_worker import EventRedactBehaviour
 from synapse.util.async_helpers import Linearizer
 from synapse.util.caches import get_cache_factor_for
 from synapse.util.caches.expiringcache import ExpiringCache
@@ -645,7 +646,7 @@ class StateResolutionStore(object):
 
         return self.store.get_events(
             event_ids,
-            check_redacted=False,
+            redact_behaviour=EventRedactBehaviour.AS_IS,
             get_prev_content=False,
             allow_rejected=allow_rejected,
         )
diff --git a/synapse/storage/data_stores/main/events_worker.py b/synapse/storage/data_stores/main/events_worker.py
index 9ee117ce0f..2c9142814c 100644
--- a/synapse/storage/data_stores/main/events_worker.py
+++ b/synapse/storage/data_stores/main/events_worker.py
@@ -19,8 +19,10 @@ import itertools
 import logging
 import threading
 from collections import namedtuple
+from typing import List, Optional
 
 from canonicaljson import json
+from constantly import NamedConstant, Names
 
 from twisted.internet import defer
 
@@ -55,6 +57,16 @@ EVENT_QUEUE_TIMEOUT_S = 0.1  # Timeout when waiting for requests for events
 _EventCacheEntry = namedtuple("_EventCacheEntry", ("event", "redacted_event"))
 
 
+class EventRedactBehaviour(Names):
+    """
+    What to do when retrieving a redacted event from the database.
+    """
+
+    AS_IS = NamedConstant()
+    REDACT = NamedConstant()
+    BLOCK = NamedConstant()
+
+
 class EventsWorkerStore(SQLBaseStore):
     def __init__(self, database: Database, db_conn, hs):
         super(EventsWorkerStore, self).__init__(database, db_conn, hs)
@@ -125,25 +137,27 @@ class EventsWorkerStore(SQLBaseStore):
     @defer.inlineCallbacks
     def get_event(
         self,
-        event_id,
-        check_redacted=True,
-        get_prev_content=False,
-        allow_rejected=False,
-        allow_none=False,
-        check_room_id=None,
+        event_id: List[str],
+        redact_behaviour: EventRedactBehaviour = EventRedactBehaviour.REDACT,
+        get_prev_content: bool = False,
+        allow_rejected: bool = False,
+        allow_none: bool = False,
+        check_room_id: Optional[str] = None,
     ):
         """Get an event from the database by event_id.
 
         Args:
-            event_id (str): The event_id of the event to fetch
-            check_redacted (bool): If True, check if event has been redacted
-                and redact it.
-            get_prev_content (bool): If True and event is a state event,
+            event_id: The event_id of the event to fetch
+            redact_behaviour: Determine what to do with a redacted event. Possible values:
+                * AS_IS - Return the full event body with no redacted content
+                * REDACT - Return the event but with a redacted body
+                * DISALLOW - Do not return redacted events
+            get_prev_content: If True and event is a state event,
                 include the previous states content in the unsigned field.
-            allow_rejected (bool): If True return rejected events.
-            allow_none (bool): If True, return None if no event found, if
+            allow_rejected: If True return rejected events.
+            allow_none: If True, return None if no event found, if
                 False throw a NotFoundError
-            check_room_id (str|None): if not None, check the room of the found event.
+            check_room_id: if not None, check the room of the found event.
                 If there is a mismatch, behave as per allow_none.
 
         Returns:
@@ -154,7 +168,7 @@ class EventsWorkerStore(SQLBaseStore):
 
         events = yield self.get_events_as_list(
             [event_id],
-            check_redacted=check_redacted,
+            redact_behaviour=redact_behaviour,
             get_prev_content=get_prev_content,
             allow_rejected=allow_rejected,
         )
@@ -173,27 +187,30 @@ class EventsWorkerStore(SQLBaseStore):
     @defer.inlineCallbacks
     def get_events(
         self,
-        event_ids,
-        check_redacted=True,
-        get_prev_content=False,
-        allow_rejected=False,
+        event_ids: List[str],
+        redact_behaviour: EventRedactBehaviour = EventRedactBehaviour.REDACT,
+        get_prev_content: bool = False,
+        allow_rejected: bool = False,
     ):
         """Get events from the database
 
         Args:
-            event_ids (list): The event_ids of the events to fetch
-            check_redacted (bool): If True, check if event has been redacted
-                and redact it.
-            get_prev_content (bool): If True and event is a state event,
+            event_ids: The event_ids of the events to fetch
+            redact_behaviour: Determine what to do with a redacted event. Possible
+                values:
+                * AS_IS - Return the full event body with no redacted content
+                * REDACT - Return the event but with a redacted body
+                * DISALLOW - Do not return redacted events
+            get_prev_content: If True and event is a state event,
                 include the previous states content in the unsigned field.
-            allow_rejected (bool): If True return rejected events.
+            allow_rejected: If True return rejected events.
 
         Returns:
             Deferred : Dict from event_id to event.
         """
         events = yield self.get_events_as_list(
             event_ids,
-            check_redacted=check_redacted,
+            redact_behaviour=redact_behaviour,
             get_prev_content=get_prev_content,
             allow_rejected=allow_rejected,
         )
@@ -203,21 +220,23 @@ class EventsWorkerStore(SQLBaseStore):
     @defer.inlineCallbacks
     def get_events_as_list(
         self,
-        event_ids,
-        check_redacted=True,
-        get_prev_content=False,
-        allow_rejected=False,
+        event_ids: List[str],
+        redact_behaviour: EventRedactBehaviour = EventRedactBehaviour.REDACT,
+        get_prev_content: bool = False,
+        allow_rejected: bool = False,
     ):
         """Get events from the database and return in a list in the same order
         as given by `event_ids` arg.
 
         Args:
-            event_ids (list): The event_ids of the events to fetch
-            check_redacted (bool): If True, check if event has been redacted
-                and redact it.
-            get_prev_content (bool): If True and event is a state event,
+            event_ids: The event_ids of the events to fetch
+            redact_behaviour: Determine what to do with a redacted event. Possible values:
+                * AS_IS - Return the full event body with no redacted content
+                * REDACT - Return the event but with a redacted body
+                * DISALLOW - Do not return redacted events
+            get_prev_content: If True and event is a state event,
                 include the previous states content in the unsigned field.
-            allow_rejected (bool): If True return rejected events.
+            allow_rejected: If True, return rejected events.
 
         Returns:
             Deferred[list[EventBase]]: List of events fetched from the database. The
@@ -319,10 +338,14 @@ class EventsWorkerStore(SQLBaseStore):
                     # Update the cache to save doing the checks again.
                     entry.event.internal_metadata.recheck_redaction = False
 
-            if check_redacted and entry.redacted_event:
-                event = entry.redacted_event
-            else:
-                event = entry.event
+            event = entry.event
+
+            if entry.redacted_event:
+                if redact_behaviour == EventRedactBehaviour.BLOCK:
+                    # Skip this event
+                    continue
+                elif redact_behaviour == EventRedactBehaviour.REDACT:
+                    event = entry.redacted_event
 
             events.append(event)
 
diff --git a/synapse/storage/data_stores/main/search.py b/synapse/storage/data_stores/main/search.py
index 4eec2fae5e..dfb46ee0f8 100644
--- a/synapse/storage/data_stores/main/search.py
+++ b/synapse/storage/data_stores/main/search.py
@@ -25,6 +25,7 @@ from twisted.internet import defer
 
 from synapse.api.errors import SynapseError
 from synapse.storage._base import SQLBaseStore, make_in_list_sql_clause
+from synapse.storage.data_stores.main.events_worker import EventRedactBehaviour
 from synapse.storage.database import Database
 from synapse.storage.engines import PostgresEngine, Sqlite3Engine
 
@@ -453,7 +454,12 @@ class SearchStore(SearchBackgroundUpdateStore):
 
         results = list(filter(lambda row: row["room_id"] in room_ids, results))
 
-        events = yield self.get_events_as_list([r["event_id"] for r in results])
+        # We set redact_behaviour to BLOCK here to prevent redacted events being returned in
+        # search results (which is a data leak)
+        events = yield self.get_events_as_list(
+            [r["event_id"] for r in results],
+            redact_behaviour=EventRedactBehaviour.BLOCK,
+        )
 
         event_map = {ev.event_id: ev for ev in events}