Extensions Docs Server: Large performance increase

chrome/common/extensions/docs/server2/in_memory_object_store.py

Index: chrome/common/extensions/docs/server2/in_memory_object_store.py
diff --git a/chrome/common/extensions/docs/server2/in_memory_object_store.py b/chrome/common/extensions/docs/server2/in_memory_object_store.py
new file mode 100644
index 0000000000000000000000000000000000000000..f2499725a76226136caed1c4c71ff4b495e93169
--- /dev/null
+++ b/chrome/common/extensions/docs/server2/in_memory_object_store.py
@@ -0,0 +1,117 @@
+# Copyright (c) 2012 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+import time
+
+from future import Future
+from object_store import ObjectStore, CACHE_TIMEOUT
+from memcache_object_store import MemcacheObjectStore
+
+class _CacheEntry(object):
+  def __init__(self, value, expire_time):
+    self.value = value
+    self._never_expires = (expire_time == 0)
+    self._expiry = time.time() + expire_time
+
+  def HasExpired(self):
+    if self._never_expires:
+      return False
+    return time.time() > self._expiry
+
+class _AsyncGetFuture(object):
+  """A future for memcache gets.
+
+  Properties:
+  - |cache| the in-memory cache used by InMemoryObjectStore
+  - |time| the cache timeout
+  - |namespace| the namespace of the cache items
+  - |future| the |Future| from the backing |ObjectStore|
+  - |initial_mapping| a mapping of cache items already in memory
+  """
+  def __init__(self, cache, time, namespace, future, initial_mapping):
+    self._cache = cache
+    self._time = time
+    self._namespace = namespace
+    self._future = future
+    self._mapping = initial_mapping
+
+  def Get(self):
+    if self._future is not None:
+      result = self._future.Get()
+      self._cache[self._namespace].update(
+          dict((k, _CacheEntry(v, self._time)) for k, v in result.iteritems()))
+      self._mapping.update(result)
+    return self._mapping
+
+class InMemoryObjectStore(ObjectStore):
+  def __init__(self, branch):
+    self._branch = branch
+    self._cache = {}
+    self._object_store = MemcacheObjectStore()
+
+  def _SetInMemory(self, key, value, namespace, time):
+    if namespace not in self._cache:
+      self._cache[namespace] = {}
+    self._cache[namespace][key] = _CacheEntry(value, time)
+
+  def _TryBackingObjectStore(self, key, namespace, time):
+    result = self._object_store.Get(key, namespace, time=time)
+    if result is not None:
+      self._SetInMemory(key, result, namespace, time)
+    return result
+
+  def _MakeNamespace(self, namespace):
+    return 'ObjectStore.' + self._branch + '.' + namespace

[not at google - send to devlin, 2012/08/21 00:30:11]

+'ing strings in python is inefficient, use '.'.join(['ObjectStore',
self._branch, namespace]) or %s

[cduvall, 2012/08/21 01:33:33]

Done.

+
+  def Set(self, key, value, namespace, time=CACHE_TIMEOUT):
+    namespace = self._MakeNamespace(namespace)
+    self._SetInMemory(key, value, namespace, time)
+    self._object_store.Set(key, value, namespace, time=time)
+
+  def SetMulti(self, mapping, namespace, time=CACHE_TIMEOUT):
+    namespace = self._MakeNamespace(namespace)
+    for k, v in mapping.iteritems():
+      self._SetInMemory(k, v, namespace, time)
+      # TODO(cduvall): Use a batch set? App Engine kept throwing:
+      # ValueError: Values may not be more than 1000000 bytes in length
+      # for the batch set.
+      self._object_store.Set(k, v, namespace, time=time)
+
+  def Get(self, key, namespace, time=CACHE_TIMEOUT):
+    namespace = self._MakeNamespace(namespace)
+    if namespace not in self._cache:
+      return self._TryBackingObjectStore(key, namespace, time)
+    cache_entry = self._cache[namespace].get(key, None)
+    if not cache_entry:
+      return self._TryBackingObjectStore(key, namespace, time)
+    if cache_entry.HasExpired():
+      self._cache[namespace].pop(key)
+      return self._object_store.Get(key, namespace, time=time)
+    return cache_entry.value
+
+  def GetMulti(self, keys, namespace, time=CACHE_TIMEOUT):
+    namespace = self._MakeNamespace(namespace)
+    keys = keys[:]
+    mapping = {}
+    if namespace not in self._cache:
+      self._cache[namespace] = {}
+    for key in keys:
+      cache_entry = self._cache[namespace].get(key, None)
+      if cache_entry is None or cache_entry.HasExpired():
+        mapping[key] = None
+      else:
+        mapping[key] = cache_entry.value
+        keys.remove(key)
+    future = self._object_store.GetMulti(keys, namespace, time=time)
+    return Future(delegate=_AsyncGetFuture(self._cache,
+                                           time,
+                                           namespace,
+                                           future,
+                                           mapping))
+

[not at google - send to devlin, 2012/08/21 00:30:11]

A lot of this code would be simplified by having Get call GetMulti and Set call
SetMulti.

[cduvall, 2012/08/21 01:33:33]

Done.

+  def Delete(self, key, namespace):
+    namespace = self._MakeNamespace(namespace)
+    if namespace in self._cache:
+      self._cache[namespace].pop(key)
+    self._object_store.Delete(key, namespace)