[Findit] Use step level analysis to exclude flaky test failures.

appengine/findit/waterfall/buildbot.py

Index: appengine/findit/waterfall/buildbot.py
diff --git a/appengine/findit/waterfall/buildbot.py b/appengine/findit/waterfall/buildbot.py
index 28d39db4ab0f9ee9d9348adb02abe0da31398105..1894220da03c255eebabc086aa281c3bc3e94702 100644
--- a/appengine/findit/waterfall/buildbot.py
+++ b/appengine/findit/waterfall/buildbot.py
@@ -2,11 +2,15 @@
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
+import contextlib
 from datetime import datetime
+import gzip
 import json
 import re
 import urllib
 
+import cloudstorage as gcs
+
 from waterfall.build_info import BuildInfo
 
 _MASTER_URL_PATTERN = re.compile(r'^https?://build\.chromium\.org/p/([^/]+)'
@@ -97,7 +101,12 @@ def CreateStdioLogUrl(master_name, builder_name, build_number, step_name):
               master_name, builder_name, build_number, step_name)
 
 
-def GetBuildDataFromBuildMaster(master_name, 
+def CreateGtestResultPath(master_name, builder_name, build_number, step_name):
+  return ('/chrome-gtest-results/buildbot/%s/%s/%s/%s.json.gz') % (
+              master_name, builder_name, build_number, step_name)
+
+
+def GetBuildDataFromBuildMaster(master_name,
                                 builder_name, build_number, http_client):
   """Returns the json-format data of the build from buildbot json API."""
   status_code, data = http_client.Get(
@@ -108,7 +117,7 @@ def GetBuildDataFromBuildMaster(master_name,
     return data
 
 
-def GetBuildDataFromArchive(master_name, 
+def GetBuildDataFromArchive(master_name,
                             builder_name, build_number, http_client):
   """Returns the json-format data of the build from build archive."""
   status_code, data = http_client.Get(
@@ -130,6 +139,20 @@ def GetStepStdio(master_name, builder_name, build_number,
     return data
 
 
+def GetGtestResultLog(master_name,
+                 builder_name, build_number, step_name):  # pragma: no cover
+  """Returns the archived Step Log file."""

[stgao, 2015/05/22 01:30:36]

This function is to return the content of the gtest json results for the
gtest-based step. So technically it is not step log file.

+  try:
+    with contextlib.closing(gcs.open(CreateGtestResultPath(
+        master_name, builder_name,
+        build_number, step_name))) as gtest_result_file:
+      with contextlib.closing(gzip.open(
+          gtest_result_file)) as unzipped_gtest_result_file:
+        return unzipped_gtest_result_file.read()
+  except gcs.NotFoundError:
+    return None
+
+
 def GetStepResult(step_data_json):
   """Returns the result of a step."""
   result = step_data_json.get('results')
@@ -184,8 +207,8 @@ def ExtractBuildInfo(master_name, builder_name, build_number, build_data):
   for step_data in steps:
     step_name = step_data['name']
 
-    step_logs = step_data.get('logs')
-    if step_logs and 'preamble' == step_logs[0][0]:
+    gtest_results = step_data.get('logs')

[stgao, 2015/05/22 01:30:36]

This is actually 'step_logs'. Let's keep it as it was.

+    if gtest_results and 'preamble' == gtest_results[0][0]:
       # Skip a annotating step like "steps" or "slave_steps", which wraps other
       # steps. A failed annotated step like "content_browsertests" will make
       # the annotating step like "steps" fail too. Such annotating steps have a