[telemetry] Automatic hash file creation and upload to Cloud Storage.

tools/perf/page_sets/PRESUBMIT.py

Index: tools/perf/page_sets/PRESUBMIT.py
diff --git a/tools/perf/page_sets/PRESUBMIT.py b/tools/perf/page_sets/PRESUBMIT.py
new file mode 100644
index 0000000000000000000000000000000000000000..f9eeac219371d654063e8272410fd280ba41b427
--- /dev/null
+++ b/tools/perf/page_sets/PRESUBMIT.py
@@ -0,0 +1,66 @@
+# Copyright (c) 2013 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 os
+import re
+import sys
+
+
+DEFAULT_BUCKET = 'chromium-wpr'
+
+
+def _SyncFilesToCloud(input_api, output_api):
+  """Searches for .sha1 files and uploads them to Cloud Storage.
+
+  It validates all the hashes and skips upload if not necessary.
+  """
+  # Because this script will be called from a magic PRESUBMIT demon,
+  # avoid angering it; don't pollute its sys.path.
+  old_sys_path = sys.path
+  try:
+    sys.path = [os.path.join(os.pardir, os.pardir, 'telemetry')] + sys.path
+    from telemetry.page import cloud_storage
+  finally:
+    sys.path = old_sys_path
+
+  hashes_in_cloud_storage = cloud_storage.List(DEFAULT_BUCKET)
+
+  results = []
+  for hash_path in input_api.AbsoluteLocalPaths():
+    file_path, extension = os.path.splitext(hash_path)
+    if extension != '.sha1':
+      continue
+
+    with open(hash_path, 'rb') as f:
+      file_hash = f.read(1024).rstrip()
+    if file_hash in hashes_in_cloud_storage:
+      results.append(output_api.PresubmitNotifyResult(
+          'File already in Cloud Storage, skipping upload: %s' % hash_path))
+      continue
+
+    if not re.match('^([A-Za-z0-9]{40})$', file_hash):
+      results.append(output_api.PresubmitError(
+          'Hash file does not contain a valid SHA-1 hash: %s' % hash_path))
+      continue
+    if not os.path.exists(file_path):
+      results.append(output_api.PresubmitError(
+          'Hash file exists, but file not found: %s' % hash_path))
+      continue
+    if cloud_storage.GetHash(file_path) != file_hash:
+      results.append(output_api.PresubmitError(
+          'Hash file does not match file\'s actual hash: %s' % hash_path))
+      continue
+
+    try:
+      cloud_storage.Insert(DEFAULT_BUCKET, file_hash, file_path)
+    except cloud_storage.CloudStorageError:
+      results.append(output_api.PresubmitError(
+          'Unable to upload to Cloud Storage: %s' % hash_path))
+
+  return results
+
+
+def CheckChangeOnCommit(input_api, output_api):
+  results = []
+  results += _SyncFilesToCloud(input_api, output_api)
+  return results