Python event log analyzer tool: fix of indexing issue.

webrtc/tools/py_event_log_analyzer/rtp_analyzertest.py

Index: webrtc/tools/py_event_log_analyzer/rtp_analyzertest.py
diff --git a/webrtc/tools/py_event_log_analyzer/rtp_analyzertest.py b/webrtc/tools/py_event_log_analyzer/rtp_analyzertest.py
new file mode 100755
index 0000000000000000000000000000000000000000..9c3e2527366d6517ef84c124ab5d651dc2e75a9d
--- /dev/null
+++ b/webrtc/tools/py_event_log_analyzer/rtp_analyzertest.py
@@ -0,0 +1,55 @@
+#!/usr/bin/env python
+#  Copyright (c) 2016 The WebRTC project authors. All Rights Reserved.
+#
+#  Use of this source code is governed by a BSD-style license
+#  that can be found in the LICENSE file in the root of the source
+#  tree. An additional intellectual property rights grant can be found
+#  in the file PATENTS.  All contributing project authors may
+#  be found in the AUTHORS file in the root of the source tree.
+
+"""Run the tests with
+
+      python rtp_analyzer_test.py
+or
+      python3 rtp_analyzer_test.py
+"""
+
+import collections
+import unittest
+
+import numpy
+import rtp_analyzer
+
+FakePoint = collections.namedtuple("FakePoint",
+                                   ["real_send_time_ms", "absdelay"])
+
+
+class TestDelay(unittest.TestCase):
+

[phoglund, 2016/09/06 11:43:05]

Remove blank line.

[aleloi, 2016/09/06 12:15:37]

Ok! But gpylint complains if there is no blank line. git cl presubmit is fine
with it, though.

Is the rule mostly consistency with other code, or is there a policy somewhere?

+  def assertMaskEqual(self, masked_array, data, mask):
+    self.assertEqual(list(masked_array.data), data)
+
+    if isinstance(masked_array.mask, numpy.bool_):
+      array_mask = masked_array.mask
+    else:
+      array_mask = list(masked_array.mask)
+    self.assertEqual(array_mask, mask)
+
+  def testCalculateDelaySimple(self):
+    points = [FakePoint(0, 0), FakePoint(1, 0)]
+    mask = rtp_analyzer.calculate_delay(0, 1, 1, points)
+    self.assertMaskEqual(mask, [0, 0], False)
+
+  def testCalculateDelayMissing(self):
+    points = [FakePoint(0, 0), FakePoint(2, 0)]
+    mask = rtp_analyzer.calculate_delay(0, 2, 1, points)
+    self.assertMaskEqual(mask, [0, -1, 0], [False, True, False])
+
+  def testCalculateDelayBorders(self):
+    points = [FakePoint(0, 0), FakePoint(2, 0)]
+    mask = rtp_analyzer.calculate_delay(0, 3, 2, points)
+    self.assertMaskEqual(mask, [0, 0, -1], [False, False, True])
+
+
+if __name__ == "__main__":
+  unittest.main()

[aleloi, 2016/09/06 11:08:11]

If I rename this file to end with '_test', the presubmit test will pick it up
and run it every time. This is a problem, because this test depends on numpy,
which is not a webrtc  dependency. Most probably, we don't want to make it a
webrtc dependency either. I couldn't think of anything better than leaving it as
is.

[phoglund, 2016/09/06 11:43:04]

Umm, why not add numpy as a dependency? Your tool is using it, and your
production code can import it, so the test should be able to as well. How does
your production code gain access to numpy?