Implement filtering

synapse/api/filtering.py

@@ -20,6 +20,7 @@ from jsonschema import FormatChecker
 
 from twisted.internet import defer
 
+from synapse.api.constants import LabelsField
 from synapse.api.errors import SynapseError
 from synapse.storage.presence import UserPresenceState
 from synapse.types import RoomID, UserID
@@ -66,6 +67,8 @@ ROOM_EVENT_FILTER_SCHEMA = {
         "contains_url": {"type": "boolean"},
         "lazy_load_members": {"type": "boolean"},
         "include_redundant_members": {"type": "boolean"},
+        "org.matrix.labels": {"type": "array", "items": {"type": "string"}},
+        "org.matrix.not_labels": {"type": "array", "items": {"type": "string"}},
     },
 }
 
@@ -259,6 +262,9 @@ class Filter(object):
 
         self.contains_url = self.filter_json.get("contains_url", None)
 
+        self.labels = self.filter_json.get("org.matrix.labels", None)
+        self.not_labels = self.filter_json.get("org.matrix.not_labels", [])
+
     def filters_all_types(self):
         return "*" in self.not_types
 
@@ -282,6 +288,7 @@ class Filter(object):
             room_id = None
             ev_type = "m.presence"
             contains_url = False
+            labels = []
         else:
             sender = event.get("sender", None)
             if not sender:
@@ -300,10 +307,11 @@ class Filter(object):
             content = event.get("content", {})
             # check if there is a string url field in the content for filtering purposes
             contains_url = isinstance(content.get("url"), text_type)
+            labels = content.get(LabelsField)
 
-        return self.check_fields(room_id, sender, ev_type, contains_url)
+        return self.check_fields(room_id, sender, ev_type, labels, contains_url)
 
-    def check_fields(self, room_id, sender, event_type, contains_url):
+    def check_fields(self, room_id, sender, event_type, labels, contains_url):
         """Checks whether the filter matches the given event fields.
 
         Returns:
@@ -313,6 +321,7 @@ class Filter(object):
             "rooms": lambda v: room_id == v,
             "senders": lambda v: sender == v,
             "types": lambda v: _matches_wildcard(event_type, v),
+            "labels": lambda v: v in labels,
         }
 
         for name, match_func in literal_keys.items():

synapse/storage/data_stores/main/stream.py

@@ -229,6 +229,14 @@ def filter_to_clause(event_filter):
         clauses.append("contains_url = ?")
         args.append(event_filter.contains_url)
 
+    # We're only applying the "labels" filter on the database query, because applying the
+    # "not_labels" filter via a SQL query is non-trivial. Instead, we let
+    # event_filter.check_fields apply it, which is not as efficient but makes the
+    # implementation simpler.
+    if event_filter.labels:
+        clauses.append("(%s)" % " OR ".join("label = ?" for _ in event_filter.labels))
+        args.extend(event_filter.labels)
+
     return " AND ".join(clauses), args
 
 
@@ -866,6 +874,7 @@ class StreamWorkerStore(EventsWorkerStore, SQLBaseStore):
         sql = (
             "SELECT event_id, topological_ordering, stream_ordering"
             " FROM events"
+            " LEFT JOIN event_labels USING (event_id)"
             " WHERE outlier = ? AND room_id = ? AND %(bounds)s"
             " ORDER BY topological_ordering %(order)s,"
             " stream_ordering %(order)s LIMIT ?"