Merge pull request #666 from matrix-org/erikj/intern

Intern lots of strings

8 files changed, 100 insertions, 31 deletions

diff --git a/synapse/events/__init__.py b/synapse/events/__init__.py
index abed6b5e6b..23f8b612ae 100644
--- a/synapse/events/__init__.py
+++ b/synapse/events/__init__.py
@@ -14,6 +14,7 @@
 # limitations under the License.
 
 from synapse.util.frozenutils import freeze
+from synapse.util.caches import intern_dict
 
 
 # Whether we should use frozen_dict in FrozenEvent. Using frozen_dicts prevents
@@ -140,6 +141,10 @@ class FrozenEvent(EventBase):
 
         unsigned = dict(event_dict.pop("unsigned", {}))
 
+        # We intern these strings because they turn up a lot (especially when
+        # caching).
+        event_dict = intern_dict(event_dict)
+
         if USE_FROZEN_DICTS:
             frozen_dict = freeze(event_dict)
         else:
diff --git a/synapse/federation/federation_client.py b/synapse/federation/federation_client.py
index 83c1f46586..37ee469fa2 100644
--- a/synapse/federation/federation_client.py
+++ b/synapse/federation/federation_client.py
@@ -418,6 +418,7 @@ class FederationClient(FederationBase):
                     "Failed to make_%s via %s: %s",
                     membership, destination, e.message
                 )
+                raise
 
         raise RuntimeError("Failed to send to any server.")
 
diff --git a/synapse/federation/transport/server.py b/synapse/federation/transport/server.py
index 208bff8d4f..d65a7893d8 100644
--- a/synapse/federation/transport/server.py
+++ b/synapse/federation/transport/server.py
@@ -175,7 +175,7 @@ class BaseFederationServlet(object):
 
 
 class FederationSendServlet(BaseFederationServlet):
-    PATH = "/send/([^/]*)/"
+    PATH = "/send/(?P<transaction_id>[^/]*)/"
 
     def __init__(self, handler, server_name, **kwargs):
         super(FederationSendServlet, self).__init__(
@@ -250,7 +250,7 @@ class FederationPullServlet(BaseFederationServlet):
 
 
 class FederationEventServlet(BaseFederationServlet):
-    PATH = "/event/([^/]*)/"
+    PATH = "/event/(?P<event_id>[^/]*)/"
 
     # This is when someone asks for a data item for a given server data_id pair.
     def on_GET(self, origin, content, query, event_id):
@@ -258,7 +258,7 @@ class FederationEventServlet(BaseFederationServlet):
 
 
 class FederationStateServlet(BaseFederationServlet):
-    PATH = "/state/([^/]*)/"
+    PATH = "/state/(?P<context>[^/]*)/"
 
     # This is when someone asks for all data for a given context.
     def on_GET(self, origin, content, query, context):
@@ -270,7 +270,7 @@ class FederationStateServlet(BaseFederationServlet):
 
 
 class FederationBackfillServlet(BaseFederationServlet):
-    PATH = "/backfill/([^/]*)/"
+    PATH = "/backfill/(?P<context>[^/]*)/"
 
     def on_GET(self, origin, content, query, context):
         versions = query["v"]
@@ -285,7 +285,7 @@ class FederationBackfillServlet(BaseFederationServlet):
 
 
 class FederationQueryServlet(BaseFederationServlet):
-    PATH = "/query/([^/]*)"
+    PATH = "/query/(?P<query_type>[^/]*)"
 
     # This is when we receive a server-server Query
     def on_GET(self, origin, content, query, query_type):
@@ -296,7 +296,7 @@ class FederationQueryServlet(BaseFederationServlet):
 
 
 class FederationMakeJoinServlet(BaseFederationServlet):
-    PATH = "/make_join/([^/]*)/([^/]*)"
+    PATH = "/make_join/(?P<context>[^/]*)/(?P<user_id>[^/]*)"
 
     @defer.inlineCallbacks
     def on_GET(self, origin, content, query, context, user_id):
@@ -305,7 +305,7 @@ class FederationMakeJoinServlet(BaseFederationServlet):
 
 
 class FederationMakeLeaveServlet(BaseFederationServlet):
-    PATH = "/make_leave/([^/]*)/([^/]*)"
+    PATH = "/make_leave/(?P<context>[^/]*)/(?P<user_id>[^/]*)"
 
     @defer.inlineCallbacks
     def on_GET(self, origin, content, query, context, user_id):
@@ -314,7 +314,7 @@ class FederationMakeLeaveServlet(BaseFederationServlet):
 
 
 class FederationSendLeaveServlet(BaseFederationServlet):
-    PATH = "/send_leave/([^/]*)/([^/]*)"
+    PATH = "/send_leave/(?P<room_id>[^/]*)/(?P<txid>[^/]*)"
 
     @defer.inlineCallbacks
     def on_PUT(self, origin, content, query, room_id, txid):
@@ -323,14 +323,14 @@ class FederationSendLeaveServlet(BaseFederationServlet):
 
 
 class FederationEventAuthServlet(BaseFederationServlet):
-    PATH = "/event_auth/([^/]*)/([^/]*)"
+    PATH = "/event_auth(?P<context>[^/]*)/(?P<event_id>[^/]*)"
 
     def on_GET(self, origin, content, query, context, event_id):
         return self.handler.on_event_auth(origin, context, event_id)
 
 
 class FederationSendJoinServlet(BaseFederationServlet):
-    PATH = "/send_join/([^/]*)/([^/]*)"
+    PATH = "/send_join/(?P<context>[^/]*)/(?P<event_id>[^/]*)"
 
     @defer.inlineCallbacks
     def on_PUT(self, origin, content, query, context, event_id):
@@ -341,7 +341,7 @@ class FederationSendJoinServlet(BaseFederationServlet):
 
 
 class FederationInviteServlet(BaseFederationServlet):
-    PATH = "/invite/([^/]*)/([^/]*)"
+    PATH = "/invite/(?P<context>[^/]*)/(?P<event_id>[^/]*)"
 
     @defer.inlineCallbacks
     def on_PUT(self, origin, content, query, context, event_id):
@@ -352,7 +352,7 @@ class FederationInviteServlet(BaseFederationServlet):
 
 
 class FederationThirdPartyInviteExchangeServlet(BaseFederationServlet):
-    PATH = "/exchange_third_party_invite/([^/]*)"
+    PATH = "/exchange_third_party_invite/(?P<room_id>[^/]*)"
 
     @defer.inlineCallbacks
     def on_PUT(self, origin, content, query, room_id):
@@ -381,7 +381,7 @@ class FederationClientKeysClaimServlet(BaseFederationServlet):
 
 
 class FederationQueryAuthServlet(BaseFederationServlet):
-    PATH = "/query_auth/([^/]*)/([^/]*)"
+    PATH = "/query_auth/(?P<context>[^/]*)/(?P<event_id>[^/]*)"
 
     @defer.inlineCallbacks
     def on_POST(self, origin, content, query, context, event_id):
@@ -394,7 +394,7 @@ class FederationQueryAuthServlet(BaseFederationServlet):
 
 class FederationGetMissingEventsServlet(BaseFederationServlet):
     # TODO(paul): Why does this path alone end with "/?" optional?
-    PATH = "/get_missing_events/([^/]*)/?"
+    PATH = "/get_missing_events/(?P<room_id>[^/]*)/?"
 
     @defer.inlineCallbacks
     def on_POST(self, origin, content, query, room_id):
diff --git a/synapse/http/server.py b/synapse/http/server.py
index b17b190ee5..b82196fd5e 100644
--- a/synapse/http/server.py
+++ b/synapse/http/server.py
@@ -18,6 +18,7 @@ from synapse.api.errors import (
     cs_exception, SynapseError, CodeMessageException, UnrecognizedRequestError, Codes
 )
 from synapse.util.logcontext import LoggingContext, PreserveLoggingContext
+from synapse.util.caches import intern_dict
 import synapse.metrics
 import synapse.events
 
@@ -229,11 +230,12 @@ class JsonResource(HttpServer, resource.Resource):
             else:
                 servlet_classname = "%r" % callback
 
-            args = [
-                urllib.unquote(u).decode("UTF-8") if u else u for u in m.groups()
-            ]
+            kwargs = intern_dict({
+                name: urllib.unquote(value).decode("UTF-8") if value else value
+                for name, value in m.groupdict().items()
+            })
 
-            callback_return = yield callback(request, *args)
+            callback_return = yield callback(request, **kwargs)
             if callback_return is not None:
                 code, response = callback_return
                 self._send_response(request, code, response)
diff --git a/synapse/storage/_base.py b/synapse/storage/_base.py
index 583b77a835..b75b79df36 100644
--- a/synapse/storage/_base.py
+++ b/synapse/storage/_base.py
@@ -18,6 +18,7 @@ from synapse.api.errors import StoreError
 from synapse.util.logcontext import LoggingContext, PreserveLoggingContext
 from synapse.util.caches.dictionary_cache import DictionaryCache
 from synapse.util.caches.descriptors import Cache
+from synapse.util.caches import intern_dict
 import synapse.metrics
 
 
@@ -350,7 +351,7 @@ class SQLBaseStore(object):
         """
         col_headers = list(column[0] for column in cursor.description)
         results = list(
-            dict(zip(col_headers, row)) for row in cursor.fetchall()
+            intern_dict(dict(zip(col_headers, row))) for row in cursor.fetchall()
         )
         return results
 
diff --git a/synapse/storage/receipts.py b/synapse/storage/receipts.py
index dbc074d6b5..6b9d848eaa 100644
--- a/synapse/storage/receipts.py
+++ b/synapse/storage/receipts.py
@@ -62,18 +62,17 @@ class ReceiptsStore(SQLBaseStore):
 
     @cachedInlineCallbacks(num_args=2)
     def get_receipts_for_user(self, user_id, receipt_type):
-        def f(txn):
-            sql = (
-                "SELECT room_id,event_id "
-                "FROM receipts_linearized "
-                "WHERE user_id = ? AND receipt_type = ? "
-            )
-            txn.execute(sql, (user_id, receipt_type))
-            return txn.fetchall()
+        rows = yield self._simple_select_list(
+            table="receipts_linearized",
+            keyvalues={
+                "user_id": user_id,
+                "receipt_type": receipt_type,
+            },
+            retcols=("room_id", "event_id"),
+            desc="get_receipts_for_user",
+        )
 
-        defer.returnValue(dict(
-            (yield self.runInteraction("get_receipts_for_user", f))
-        ))
+        defer.returnValue({row["room_id"]: row["event_id"] for row in rows})
 
     @defer.inlineCallbacks
     def get_linearized_receipts_for_rooms(self, room_ids, to_key, from_key=None):
diff --git a/synapse/storage/state.py b/synapse/storage/state.py
index eab2c5a8ce..02cefdff26 100644
--- a/synapse/storage/state.py
+++ b/synapse/storage/state.py
@@ -15,6 +15,7 @@
 
 from ._base import SQLBaseStore
 from synapse.util.caches.descriptors import cached, cachedList
+from synapse.util.caches import intern_string
 
 from twisted.internet import defer
 
@@ -393,7 +394,10 @@ class StateStore(SQLBaseStore):
                     # cache absence of the key, on the assumption that if we've
                     # explicitly asked for some types then we will probably ask
                     # for them again.
-                    state_dict = {key: None for key in types}
+                    state_dict = {
+                        (intern_string(etype), intern_string(state_key)): None
+                        for (etype, state_key) in types
+                    }
                     state_dict.update(results[group])
                     results[group] = state_dict
                 else:
diff --git a/synapse/util/caches/__init__.py b/synapse/util/caches/__init__.py
index 1a14904194..d53569ca49 100644
--- a/synapse/util/caches/__init__.py
+++ b/synapse/util/caches/__init__.py
@@ -14,6 +14,10 @@
 # limitations under the License.
 
 import synapse.metrics
+from lrucache import LruCache
+import os
+
+CACHE_SIZE_FACTOR = float(os.environ.get("SYNAPSE_CACHE_FACTOR", 0.1))
 
 DEBUG_CACHES = False
 
@@ -25,3 +29,56 @@ cache_counter = metrics.register_cache(
     lambda: {(name,): len(caches_by_name[name]) for name in caches_by_name.keys()},
     labels=["name"],
 )
+
+_string_cache = LruCache(int(5000 * CACHE_SIZE_FACTOR))
+caches_by_name["string_cache"] = _string_cache
+
+
+KNOWN_KEYS = {
+    key: key for key in
+    (
+        "auth_events",
+        "content",
+        "depth",
+        "event_id",
+        "hashes",
+        "origin",
+        "origin_server_ts",
+        "prev_events",
+        "room_id",
+        "sender",
+        "signatures",
+        "state_key",
+        "type",
+        "unsigned",
+        "user_id",
+    )
+}
+
+
+def intern_string(string):
+    """Takes a (potentially) unicode string and interns using custom cache
+    """
+    return _string_cache.setdefault(string, string)
+
+
+def intern_dict(dictionary):
+    """Takes a dictionary and interns well known keys and their values
+    """
+    return {
+        KNOWN_KEYS.get(key, key): _intern_known_values(key, value)
+        for key, value in dictionary.items()
+    }
+
+
+def _intern_known_values(key, value):
+    intern_str_keys = ("event_id", "room_id")
+    intern_unicode_keys = ("sender", "user_id", "type", "state_key")
+
+    if key in intern_str_keys:
+        return intern(value.encode('ascii'))
+
+    if key in intern_unicode_keys:
+        return intern_string(value)
+
+    return value