File: standard_gtest_merge.py

package info (click to toggle)
chromium 139.0.7258.127-1
  • links: PTS, VCS
  • area: main
  • in suites:
  • size: 6,122,068 kB
  • sloc: cpp: 35,100,771; ansic: 7,163,530; javascript: 4,103,002; python: 1,436,920; asm: 946,517; xml: 746,709; pascal: 187,653; perl: 88,691; sh: 88,436; objc: 79,953; sql: 51,488; cs: 44,583; fortran: 24,137; makefile: 22,147; tcl: 15,277; php: 13,980; yacc: 8,984; ruby: 7,485; awk: 3,720; lisp: 3,096; lex: 1,327; ada: 727; jsp: 228; sed: 36
file content (175 lines) | stat: -rwxr-xr-x 6,136 bytes parent folder | download | duplicates (9)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
#! /usr/bin/env python3
#
# Copyright 2017 The Chromium Authors
# Use of this source code is governed by a BSD-style license that can be
# found in the LICENSE file.


import argparse
import json
import os
import sys


def merge_shard_results(summary_json, jsons_to_merge):
  """Reads JSON test output from all shards and combines them into one.

  Returns dict with merged test output on success or None on failure. Emits
  annotations.
  """
  try:
    with open(summary_json) as f:
      summary = json.load(f)
  except (IOError, ValueError):
    # TODO(crbug.com/40196155):Re-enable this check after the recipe module
    # chromium_swarming can run it with py3
    # pylint: disable=raise-missing-from
    raise Exception('Summary json cannot be loaded.')

  # Merge all JSON files together. Keep track of missing shards.
  merged = {
    'all_tests': set(),
    'disabled_tests': set(),
    'global_tags': set(),
    'missing_shards': [],
    'per_iteration_data': [],
    'swarming_summary': summary,
    'links': set()
  }
  for index, result in enumerate(summary['shards']):
    if result is None:
      merged['missing_shards'].append(index)
      continue

    # Author note: this code path doesn't trigger convert_to_old_format() in
    # client/swarming.py, which means the state enum is saved in its string
    # name form, not in the number form.
    state = result.get('state')
    if state == 'BOT_DIED':
      print(
          'Shard #%d had a Swarming internal failure' % index, file=sys.stderr)
    elif state == 'EXPIRED':
      print('There wasn\'t enough capacity to run your test', file=sys.stderr)
    elif state == 'TIMED_OUT':
      print('Test runtime exceeded allocated time'
            'Either it ran for too long (hard timeout) or it didn\'t produce '
            'I/O for an extended period of time (I/O timeout)',
            file=sys.stderr)
    elif state != 'COMPLETED':
      print('Invalid Swarming task state: %s' % state, file=sys.stderr)

    json_data, err_msg = load_shard_json(index, result.get('task_id'),
                                         jsons_to_merge)
    if json_data:
      # Set-like fields.
      for key in ('all_tests', 'disabled_tests', 'global_tags', 'links'):
        merged[key].update(json_data.get(key), [])

      # 'per_iteration_data' is a list of dicts. Dicts should be merged
      # together, not the 'per_iteration_data' list itself.
      merged['per_iteration_data'] = merge_list_of_dicts(
          merged['per_iteration_data'], json_data.get('per_iteration_data', []))
    else:
      merged['missing_shards'].append(index)
      print('No result was found: %s' % err_msg, file=sys.stderr)

  # If some shards are missing, make it known. Continue parsing anyway. Step
  # should be red anyway, since swarming.py return non-zero exit code in that
  # case.
  if merged['missing_shards']:
    as_str = ', '.join([str(shard) for shard in merged['missing_shards']])
    print('some shards did not complete: %s' % as_str, file=sys.stderr)
    # Not all tests run, combined JSON summary can not be trusted.
    merged['global_tags'].add('UNRELIABLE_RESULTS')

  # Convert to jsonish dict.
  for key in ('all_tests', 'disabled_tests', 'global_tags', 'links'):
    merged[key] = sorted(merged[key])
  return merged


OUTPUT_JSON_SIZE_LIMIT = 100 * 1024 * 1024  # 100 MB


def load_shard_json(index, task_id, jsons_to_merge):
  """Reads JSON output of the specified shard.

  Args:
    output_dir: The directory in which to look for the JSON output to load.
    index: The index of the shard to load data for, this is for old api.
    task_id: The directory of the shard to load data for, this is for new api.

  Returns: A tuple containing:
    * The contents of path, deserialized into a python object.
    * An error string.
    (exactly one of the tuple elements will be non-None).
  """
  matching_json_files = [
      j for j in jsons_to_merge
      if (os.path.basename(j) == 'output.json' and
          (os.path.basename(os.path.dirname(j)) == str(index) or
           os.path.basename(os.path.dirname(j)) == task_id))]

  if not matching_json_files:
    print('shard %s test output missing' % index, file=sys.stderr)
    return (None, 'shard %s test output was missing' % index)
  if len(matching_json_files) > 1:
    print('duplicate test output for shard %s' % index, file=sys.stderr)
    return (None, 'shard %s test output was duplicated' % index)

  path = matching_json_files[0]

  try:
    filesize = os.stat(path).st_size
    if filesize > OUTPUT_JSON_SIZE_LIMIT:
      print(
          'output.json is %d bytes. Max size is %d' % (filesize,
                                                       OUTPUT_JSON_SIZE_LIMIT),
          file=sys.stderr)
      return (None, 'shard %s test output exceeded the size limit' % index)

    with open(path) as f:
      return (json.load(f), None)
  except (IOError, ValueError, OSError) as e:
    print('Missing or invalid gtest JSON file: %s' % path, file=sys.stderr)
    print('%s: %s' % (type(e).__name__, e), file=sys.stderr)

    return (None, 'shard %s test output was missing or invalid' % index)


def merge_list_of_dicts(left, right):
  """Merges dicts left[0] with right[0], left[1] with right[1], etc."""
  output = []
  for i in range(max(len(left), len(right))):
    left_dict = left[i] if i < len(left) else {}
    right_dict = right[i] if i < len(right) else {}
    merged_dict = left_dict.copy()
    merged_dict.update(right_dict)
    output.append(merged_dict)
  return output


def standard_gtest_merge(
    output_json, summary_json, jsons_to_merge):

  output = merge_shard_results(summary_json, jsons_to_merge)
  with open(output_json, 'w') as f:
    json.dump(output, f)

  return 0


def main(raw_args):
  parser = argparse.ArgumentParser()
  parser.add_argument('--summary-json')
  parser.add_argument('-o', '--output-json', required=True)
  parser.add_argument('jsons_to_merge', nargs='*')

  args = parser.parse_args(raw_args)

  return standard_gtest_merge(
      args.output_json, args.summary_json, args.jsons_to_merge)


if __name__ == '__main__':
  sys.exit(main(sys.argv[1:]))