Docserver: Persist stat cache for versioned file systems

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

Index: chrome/common/extensions/docs/server2/refresh_tracker.py
diff --git a/chrome/common/extensions/docs/server2/refresh_tracker.py b/chrome/common/extensions/docs/server2/refresh_tracker.py
new file mode 100644
index 0000000000000000000000000000000000000000..60bd598ec166fcd5989cc3fba06c2de18a191279
--- /dev/null
+++ b/chrome/common/extensions/docs/server2/refresh_tracker.py
@@ -0,0 +1,55 @@
+# Copyright 2014 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.
+
+from future import All, Future
+
+
+class _RefreshWorkOrder(object):
+  '''A set of tasks that must be completed in a refresh cycle.'''
+  def __init__(self, tasks):
+    self.tasks = set(tasks)
+
+
+class _RefreshTaskCompletion(object):
+  '''Marks the completion of a single named task for some refresh cycle.'''
+  def __init__(self, task):
+    self.task = task
+
+
+def _GetTaskCompletionKey(refresh_id, task):
+  return '%s@%s' % (refresh_id, task)
+
+
+class RefreshTracker(object):
+  '''Manages and tracks the progress of data refresh cycles.'''
+  def __init__(self, object_store_creator):
+    self._work_orders = object_store_creator.Create(_RefreshWorkOrder,
+        category='refresh_work_orders', start_empty=False)
+    self._task_completions = object_store_creator.Create(_RefreshTaskCompletion,
+        category='refresh_task_completions', start_empty=False)

[not at google - send to devlin, 2014/10/24 00:04:52]

It seems like you're double-namespacing these things?

_RefreshWorkOrder is just a set, so you only need to store sets in _work_orders.
The only advantage I could see in making it a class is to get it namespaced on
_RefreshWorkOrder, but, that's what category is for.

Likewise, isn't _RefreshTaskCompletion just an object?

[Ken Rockot(use gerrit already), 2014/10/24 21:26:54]

As noted on CommitTracker, yeah, I was using this incorrectly. Fixed.

+
+  def _GetWorkOrder(self, refresh_id):
+    '''Retrieves the work order for a refresh cycle identified by |id|.'''
+    return self._work_orders.Get(refresh_id)
+
+  def RegisterRefresh(self, refresh_id, tasks):
+    work_order = _RefreshWorkOrder(tasks)
+    return self._work_orders.Set(refresh_id, work_order)
+
+  def GetRefreshComplete(self, refresh_id):
+    '''Determines if a refresh cycle identified by |id| is complete. Returns
+    a |Future| which resolves to either |True| or |False|.'''
+    def is_work_order_complete(work_order):
+      # If the identified refresh cycle is not known, always return False.
+      if work_order is None:
+        return False
+      future = self._task_completions.GetMulti(
+          _GetTaskCompletionKey(refresh_id, task) for task in work_order.tasks)
+      return len(future.Get()) == len(work_order.tasks)

[not at google - send to devlin, 2014/10/24 00:04:52]

Can you turn this .Get() into a .Then()?

[Ken Rockot(use gerrit already), 2014/10/24 21:26:54]

Done.

+
+    return self._GetWorkOrder(refresh_id).Then(is_work_order_complete)
+
+  def MarkTaskComplete(self, refresh_id, task):
+    return self._task_completions.Set(_GetTaskCompletionKey(refresh_id, task),
+        _RefreshTaskCompletion(task))