Script to load a list of URLs in a loop and save traces.

tools/android/loading/run_sandwich.py

Index: tools/android/loading/run_sandwich.py
diff --git a/tools/android/loading/run_sandwich.py b/tools/android/loading/run_sandwich.py
new file mode 100755
index 0000000000000000000000000000000000000000..93a227fe6f14ea6bdfed3b76fd8a52b6546175ab
--- /dev/null
+++ b/tools/android/loading/run_sandwich.py
@@ -0,0 +1,115 @@
+#! /usr/bin/env python
+# Copyright 2016 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.
+
+"""Instructs Chrome to load series of web pages and reports results.
+
+When running Chrome is sandwiched between preprocessed disk caches and
+WepPageReplay serving all connections.
+
+TODO(pasko): implement cache preparation and WPR.
+"""
+
+import argparse
+import logging
+import os
+import sys
+
+_SRC_DIR = os.path.abspath(os.path.join(
+    os.path.dirname(__file__), '..', '..', '..'))
+
+sys.path.append(os.path.join(_SRC_DIR, 'third_party', 'catapult', 'devil'))
+from devil.android import device_utils
+
+sys.path.append(os.path.join(_SRC_DIR, 'build', 'android'))
+import devil_chromium
+
+import device_setup
+import devtools_monitor
+import json
+import page_track
+import tracing
+
+
+_JOB_SEARCH_PATH = 'sandwich_jobs'
+
+
+def _ReadUrlsFromJobDescription(job_name):
+  """Retrieves the list of URLs associated with the job name."""
+  try:
+    # Extra sugar: attempt to load from a relative path.
+    json_file_name = os.path.join(os.path.dirname(__file__), _JOB_SEARCH_PATH,
+        job_name)
+    with open(json_file_name) as f:
+      json_data = json.load(f)
+  except IOError:
+    # Attempt to read by regular file name.
+    with open(job_name) as f:
+      json_data = json.load(f)
+
+  key = 'urls'
+  if json_data and key in json_data:
+    url_list = json_data[key]
+    if isinstance(url_list, list) and len(url_list) > 0:
+      return url_list
+  raise Exception('Job description does not define a list named "urls"')
+
+
+def _SaveChromeTrace(events, directory, subdirectory):
+  """Saves the trace events, ignores IO errors.
+
+  Args:
+    events: a dict as returned by TracingTrack.ToJsonDict()
+    directory: directory name contining all traces
+    subdirectory: directory name to create this particular trace in
+  """
+  target_directory = os.path.join(directory, subdirectory)
+  file_name = os.path.join(target_directory, 'trace.json')
+  try:
+    os.makedirs(target_directory)
+    with open(file_name, 'w') as f:

[mattcary, 2016/01/28 15:35:35]

Could use gzip.GzipFile if we get concerned about space ever.

[pasko, 2016/01/28 15:48:52]

Thanks, noticed that it is done in another script. The current size is about
5MiB per page, not concerned yet, will change as necessary.

+      json.dump({'traceEvents': events['events'], 'metadata': {}}, f)
+  except IOError:
+    logging.warning('Could not save a trace: %s' % file_name)
+    # Swallow the exception.
+
+
+def main():
+  logging.basicConfig(level=logging.INFO)
+  devil_chromium.Initialize()
+
+  parser = argparse.ArgumentParser()
+  parser.add_argument('--job', required=True,
+                      help='JSON file with job description.')
+  parser.add_argument('--output', required=True,
+                      help='Name of output directory to create.')
+  parser.add_argument('--repeat', default='1',

[mattcary, 2016/01/28 15:35:34]

Setting type=int avoids the int(args.repeat) conversion further down, and may
give better error messages if you pass in a non-int for the argument (eg, not
just an error is thrown, maybe floating point is handled correctly).

[pasko, 2016/01/28 15:48:52]

oh nice, done

+                      help='How many times to run the job')
+  args = parser.parse_args()
+
+  try:
+    os.makedirs(args.output)
+  except OSError:
+    logging.error('Cannot create directory for results: %s' % args.output)
+    raise
+
+  job_urls = _ReadUrlsFromJobDescription(args.job)
+  device = device_utils.DeviceUtils.HealthyDevices()[0]
+  pages_loaded = 0
+  for iteration in xrange(0, int(args.repeat)):

[mattcary, 2016/01/28 15:35:35]

0, unnecessary for xrange.

[pasko, 2016/01/28 15:48:52]

Done

+    for url in job_urls:
+      with device_setup.DeviceConnection(device) as connection:
+        page = page_track.PageTrack(connection)
+        tracing_track = tracing.TracingTrack(connection,
+            categories='blink,cc,netlog,renderer.scheduler,toplevel,v8')
+        connection.SetUpMonitoring()
+        connection.SendAndIgnoreResponse('Page.navigate', {'url': url})
+        connection.StartMonitoring()
+        pages_loaded += 1
+        _SaveChromeTrace(tracing_track.ToJsonDict(), args.output,
+            str(pages_loaded))
+
+
+if __name__ == '__main__':
+  sys.exit(main())