Made e2e audio quality test write its results to perf.
The https://chromereviews.googleplex.com/5573026/ patch will mark the test step as perf-printing - this cl will make the test actually print perf lines.
TEST=Ran locally.
Review URL: https://webrtc-codereview.appspot.com/933036
git-svn-id: http://webrtc.googlecode.com/svn/trunk@3252 4adac7df-926f-26a2-2b94-8c16560cd09d
diff --git a/tools/perf/perf_utils.py b/tools/perf/perf_utils.py
new file mode 100644
index 0000000..77eda1e
--- /dev/null
+++ b/tools/perf/perf_utils.py
@@ -0,0 +1,31 @@
+# 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.
+
+# Copied from /src/chrome/test/pyautolib/pyauto_utils.py in Chromium.
+
+import sys
+
+def PrintPerfResult(graph_name, series_name, data_point, units,
+ show_on_waterfall=False):
+ """Prints a line to stdout that is specially formatted for the perf bots.
+
+ Args:
+ graph_name: String name for the graph on which to plot the data.
+ series_name: String name for the series (line on the graph) associated with
+ the data. This is also the string displayed on the waterfall
+ if |show_on_waterfall| is True.
+ data_point: Numeric data value to plot on the graph for the current build.
+ This can be a single value or an array of values. If an array,
+ the graph will plot the average of the values, along with error
+ bars.
+ units: The string unit of measurement for the given |data_point|.
+ show_on_waterfall: Whether or not to display this result directly on the
+ buildbot waterfall itself (in the buildbot step running
+ this test on the waterfall page, not the stdio page).
+ """
+ waterfall_indicator = ['', '*'][show_on_waterfall]
+ print '%sRESULT %s: %s= %s %s' % (
+ waterfall_indicator, graph_name, series_name,
+ str(data_point).replace(' ', ''), units)
+ sys.stdout.flush()
\ No newline at end of file