delete_local_events for purge_history

Add a flag which makes the purger delete local events

5 files changed, 67 insertions, 15 deletions

diff --git a/docs/admin_api/purge_history_api.rst b/docs/admin_api/purge_history_api.rst
index 08b3306366..b4e5bd9d75 100644
--- a/docs/admin_api/purge_history_api.rst
+++ b/docs/admin_api/purge_history_api.rst
@@ -4,8 +4,6 @@ Purge History API
 The purge history API allows server admins to purge historic events from their
 database, reclaiming disk space.
 
-**NB!** This will not delete local events (locally sent messages content etc) from the database, but will remove lots of the metadata about them and does dramatically reduce the on disk space usage
-
 Depending on the amount of history being purged a call to the API may take
 several minutes or longer. During this period users will not be able to
 paginate further back in the room from the point being purged from.
@@ -15,3 +13,15 @@ The API is simply:
 ``POST /_matrix/client/r0/admin/purge_history/<room_id>/<event_id>``
 
 including an ``access_token`` of a server admin.
+
+By default, events sent by local users are not deleted, as they may represent
+the only copies of this content in existence. (Events sent by remote users are
+deleted, and room state data before the cutoff is always removed).
+
+To delete local events as well, set ``delete_local_events`` in the body:
+
+.. code:: json
+
+   {
+       "delete_local_events": True,
+   }
diff --git a/synapse/handlers/message.py b/synapse/handlers/message.py
index 1c7860bb05..276d1a7722 100644
--- a/synapse/handlers/message.py
+++ b/synapse/handlers/message.py
@@ -63,7 +63,7 @@ class MessageHandler(BaseHandler):
         self.spam_checker = hs.get_spam_checker()
 
     @defer.inlineCallbacks
-    def purge_history(self, room_id, event_id):
+    def purge_history(self, room_id, event_id, delete_local_events=False):
         event = yield self.store.get_event(event_id)
 
         if event.room_id != room_id:
@@ -72,7 +72,7 @@ class MessageHandler(BaseHandler):
         depth = event.depth
 
         with (yield self.pagination_lock.write(room_id)):
-            yield self.store.purge_history(room_id, depth)
+            yield self.store.purge_history(room_id, depth, delete_local_events)
 
     @defer.inlineCallbacks
     def get_messages(self, requester, room_id=None, pagin_config=None,
diff --git a/synapse/http/servlet.py b/synapse/http/servlet.py
index 71420e54db..ef8e62901b 100644
--- a/synapse/http/servlet.py
+++ b/synapse/http/servlet.py
@@ -148,11 +148,13 @@ def parse_string_from_args(args, name, default=None, required=False,
             return default
 
 
-def parse_json_value_from_request(request):
+def parse_json_value_from_request(request, allow_empty_body=False):
     """Parse a JSON value from the body of a twisted HTTP request.
 
     Args:
         request: the twisted HTTP request.
+        allow_empty_body (bool): if True, an empty body will be accepted and
+            turned into None
 
     Returns:
         The JSON value.
@@ -165,6 +167,9 @@ def parse_json_value_from_request(request):
     except Exception:
         raise SynapseError(400, "Error reading JSON content.")
 
+    if not content_bytes and allow_empty_body:
+        return None
+
     try:
         content = simplejson.loads(content_bytes)
     except Exception as e:
@@ -174,17 +179,24 @@ def parse_json_value_from_request(request):
     return content
 
 
-def parse_json_object_from_request(request):
+def parse_json_object_from_request(request, allow_empty_body=False):
     """Parse a JSON object from the body of a twisted HTTP request.
 
     Args:
         request: the twisted HTTP request.
+        allow_empty_body (bool): if True, an empty body will be accepted and
+            turned into an empty dict.
 
     Raises:
         SynapseError if the request body couldn't be decoded as JSON or
             if it wasn't a JSON object.
     """
-    content = parse_json_value_from_request(request)
+    content = parse_json_value_from_request(
+        request, allow_empty_body=allow_empty_body,
+    )
+
+    if allow_empty_body and content is None:
+        return {}
 
     if type(content) != dict:
         message = "Content must be a JSON object."
diff --git a/synapse/rest/client/v1/admin.py b/synapse/rest/client/v1/admin.py
index 5022808ea9..f954d2ea65 100644
--- a/synapse/rest/client/v1/admin.py
+++ b/synapse/rest/client/v1/admin.py
@@ -128,7 +128,16 @@ class PurgeHistoryRestServlet(ClientV1RestServlet):
         if not is_admin:
             raise AuthError(403, "You are not a server admin")
 
-        yield self.handlers.message_handler.purge_history(room_id, event_id)
+        body = parse_json_object_from_request(request, allow_empty_body=True)
+
+        delete_local_events = bool(
+            body.get("delete_local_history", False)
+        )
+
+        yield self.handlers.message_handler.purge_history(
+            room_id, event_id,
+            delete_local_events=delete_local_events,
+        )
 
         defer.returnValue((200, {}))
 
diff --git a/synapse/storage/events.py b/synapse/storage/events.py
index 24d9978304..11a2ff2d8a 100644
--- a/synapse/storage/events.py
+++ b/synapse/storage/events.py
@@ -2031,16 +2031,32 @@ class EventsStore(SQLBaseStore):
             )
         return self.runInteraction("get_all_new_events", get_all_new_events_txn)
 
-    def purge_history(self, room_id, topological_ordering):
+    def purge_history(
+        self, room_id, topological_ordering, delete_local_events,
+    ):
         """Deletes room history before a certain point
+
+        Args:
+            room_id (str):
+
+            topological_ordering (int):
+                minimum topo ordering to preserve
+
+            delete_local_events (bool):
+                if True, we will delete local events as well as remote ones
+                (instead of just marking them as outliers and deleting their
+                state groups).
         """
 
         return self.runInteraction(
             "purge_history",
-            self._purge_history_txn, room_id, topological_ordering
+            self._purge_history_txn, room_id, topological_ordering,
+            delete_local_events,
         )
 
-    def _purge_history_txn(self, txn, room_id, topological_ordering):
+    def _purge_history_txn(
+        self, txn, room_id, topological_ordering, delete_local_events,
+    ):
         # Tables that should be pruned:
         #     event_auth
         #     event_backward_extremities
@@ -2093,11 +2109,14 @@ class EventsStore(SQLBaseStore):
 
         to_delete = [
             (event_id,) for event_id, state_key in event_rows
-            if state_key is None and not self.hs.is_mine_id(event_id)
+            if state_key is None and (
+                delete_local_events or not self.hs.is_mine_id(event_id)
+            )
         ]
         logger.info(
-            "[purge] found %i events before cutoff, of which %i are remote"
-            " non-state events to delete", len(event_rows), len(to_delete))
+            "[purge] found %i events before cutoff, of which %i can be deleted",
+            len(event_rows), len(to_delete),
+        )
 
         logger.info("[purge] Finding new backward extremities")
 
@@ -2273,7 +2292,9 @@ class EventsStore(SQLBaseStore):
             " WHERE event_id = ?",
             [
                 (True, event_id,) for event_id, state_key in event_rows
-                if state_key is not None or self.hs.is_mine_id(event_id)
+                if state_key is not None or (
+                    not delete_local_events and self.hs.is_mine_id(event_id)
+                )
             ]
         )