Define measurements as a tuple of name and spec.

mojo/devtools/common/mojo_benchmark

 #!/usr/bin/env python
 # Copyright 2015 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.
 
 """Runner for Mojo application benchmarks."""
 
 import argparse
 import json
 import logging
@@ skipping 13 matching lines @@
 global variable, containing entries of the following form:
 
   {
     'name': '<name of the benchmark>',
     'app': '<url of the app to benchmark>',
     'shell-args': [],
     'duration': <duration in seconds>,
 
     # List of measurements to make.
     'measurements': [
-      '<measurement type>/<event category>/<event name>',
+      {
+        'name': my_measurement,
+        'spec': spec,
+      },
+      (...)
     ]
   }
 
+For each measurement, 'name' is a label used for presentation purposes. 'spec'
+defines the measurement.
+
 Available measurement types are:
 
-  - 'time_until' - time until the first occurence of the targeted event
-  - 'avg_duration' - average duration of the targeted event
-  - 'percentile_duration' - value at XXth percentile of the targeted event where
-    XX is from the measurement spec, i.e. .../<event name>/0.XX
+'time_until' - time until the first occurence of the targeted event. The spec
+takes the following format:
+
+  'time_until/category/event'
+
+'avg_duration' - average duration of the targeted event. The spec takes the
+following format:
+
+  'avg_duration/category/event'
+
+'percentile_duration' - value at the given percentile of the targeted event. The
+spec takes the following format:
+
+  'percentile_duration/category/event/percentile'
 
 |benchmark_list_file| may reference the |target_os| global that will be any of
 ['android', 'linux'], indicating the system on which the benchmarks are to be
 run.
 """
 
 _logger = logging.getLogger()
 
 _BENCHMARK_APP = 'https://core.mojoapps.io/benchmark.mojo'
 _CACHE_SERVICE_URL = 'mojo:url_response_disk_cache'
 _NETWORK_SERVICE_URL = 'mojo:network_service'
 
 _COLD_START_SHELL_ARGS = [
   '--args-for=%s %s' % (_CACHE_SERVICE_URL, '--clear'),
   '--args-for=%s %s' % (_NETWORK_SERVICE_URL, '--clear'),
 ]
 
 # Additional time in seconds allocated per shell run to accommodate start-up.
 # The shell should terminate before hitting this time out, it is an error if it
 # doesn't.
 _EXTRA_TIMEOUT = 20
 
 _MEASUREMENT_RESULT_FORMAT = r"""
 ^              # Beginning of the line.
 measurement:   # Hard-coded tag.
-\s+(\S+)       # Match measurement name.
+\s+(\S+)       # Match measurement spec.
 \s+(\S+)       # Match measurement result.
 $              # End of the line.
 """
 
 _MEASUREMENT_REGEX = re.compile(_MEASUREMENT_RESULT_FORMAT, re.VERBOSE)
 
 
 def _generate_benchmark_variants(benchmark_spec):
   """Generates benchmark specifications for individual variants of the given
   benchmark: cold start and warm start.
@@ skipping 34 matching lines @@
 
   device_output_file = None
   if output_file:
     if android:
       device_output_file = os.path.join(shell.get_tmp_dir_path(), output_file)
       benchmark_args.append('--trace-output=' + device_output_file)
     else:
       benchmark_args.append('--trace-output=' + output_file)
 
   for measurement in measurements:
-    benchmark_args.append(measurement)
+    benchmark_args.append(measurement['spec'])
 
   shell_args = list(shell_args)
   shell_args.append(_BENCHMARK_APP)
   shell_args.append('--force-offline-by-default')
   shell_args.append('--args-for=%s %s' % (_BENCHMARK_APP,
                                           ' '.join(benchmark_args)))
 
   if verbose:
     print 'shell arguments: ' + str(shell_args)
   return_code, output, did_time_out = shell.run_and_get_output(
@@ skipping 13 matching lines @@
 
 def _parse_measurement_results(output):
   """Parses the measurement results present in the benchmark output and returns
   the dictionary of correctly recognized and parsed results.
   """
   measurement_results = {}
   output_lines = [line.strip() for line in output.split('\n')]
   for line in output_lines:
     match = re.match(_MEASUREMENT_REGEX, line)
     if match:
-      measurement_name = match.group(1)
+      measurement_spec = match.group(1)
       measurement_result = match.group(2)
       try:
-        measurement_results[measurement_name] = float(measurement_result)
+        measurement_results[measurement_spec] = float(measurement_result)
       except ValueError:
         pass
   return measurement_results
 
 
 def main():
   parser = argparse.ArgumentParser(
       formatter_class=argparse.RawDescriptionHelpFormatter,
       description=_DESCRIPTION)
   parser.add_argument('benchmark_list_file', type=file,
@@ skipping 44 matching lines @@
           shell, shell_args, variant_name, app, duration, measurements,
           script_args.verbose, script_args.android, output_file)
 
       print '[ %s ] %s ' % (benchmark_name, variant_name)
 
       some_measurements_failed = False
       if benchmark_succeeded:
         measurement_results = _parse_measurement_results(output)
         # Iterate over the list of specs, not the dictionary, to detect missing
         # results and preserve the required order.
-        for measurement_spec in measurements:
-          if measurement_spec in measurement_results:
-            result = measurement_results[measurement_spec]
-            print '%s %s' % (measurement_spec, result)
+        for measurement in measurements:
+          if measurement['spec'] in measurement_results:
+            result = measurement_results[measurement['spec']]
+            print '%10.4f  %s' % (result, measurement['name'])
 
             if chart_data_recorder:
               chart_name = benchmark_name + '__' + variant_name
               chart_data_recorder.record_scalar(
                   perf_dashboard.normalize_label(chart_name),
-                  perf_dashboard.normalize_label(measurement_spec),
+                  perf_dashboard.normalize_label(measurement['name']),
                   'ms', result)
           else:
-            print '%s ?' % measurement_spec
+            print '?  %s' % measurement['name']
             some_measurements_failed = True
 
       if not benchmark_succeeded or some_measurements_failed:
         if not benchmark_succeeded:
           print 'benchmark failed: ' + benchmark_error
         if some_measurements_failed:
           print 'some measurements failed'
         print 'output: '
         print '-' * 72
         print output
         print '-' * 72
         exit_code = 1
 
       if script_args.upload:
         perf_dashboard.upload_chart_data(
             script_args.master_name, script_args.bot_name,
             script_args.test_name, script_args.builder_name,
             script_args.build_number, chart_data_recorder.get_chart_data(),
             script_args.server_url, script_args.dry_run)
 
   return exit_code
 
 if __name__ == '__main__':
   sys.exit(main())