OLD | NEW |
(Empty) | |
| 1 #!/usr/bin/env python |
| 2 # Copyright (c) 2012 The Chromium Authors. All rights reserved. |
| 3 # Use of this source code is governed by a BSD-style license that can be |
| 4 # found in the LICENSE file. |
| 5 |
| 6 import json |
| 7 import os |
| 8 import re |
| 9 import tempfile |
| 10 |
| 11 |
| 12 def CompressList(lines, max_length, middle_replacement): |
| 13 """Ensures that |lines| is no longer than |max_length|. If |lines| need to |
| 14 be compressed then the middle items are replaced by |middle_replacement|. |
| 15 """ |
| 16 if len(lines) <= max_length: |
| 17 return lines |
| 18 remove_from_start = max_length / 2 |
| 19 return (lines[:remove_from_start] + |
| 20 [middle_replacement] + |
| 21 lines[len(lines) - (max_length - remove_from_start):]) |
| 22 |
| 23 |
| 24 class GTestLogParser(object): |
| 25 """This helper class process GTest test output.""" |
| 26 |
| 27 def __init__(self): |
| 28 # State tracking for log parsing |
| 29 self.completed = False |
| 30 self._current_test = '' |
| 31 self._failure_description = [] |
| 32 self._current_suppression_hash = '' |
| 33 self._current_suppression = [] |
| 34 self._parsing_failures = False |
| 35 |
| 36 # Line number currently being processed. |
| 37 self._line_number = 0 |
| 38 |
| 39 # List of parsing errors, as human-readable strings. |
| 40 self._internal_error_lines = [] |
| 41 |
| 42 # Tests are stored here as 'test.name': (status, [description]). |
| 43 # The status should be one of ('started', 'OK', 'failed', 'timeout', |
| 44 # 'warning'). Warning indicates that a test did not pass when run in |
| 45 # parallel with other tests but passed when run alone. The description is |
| 46 # a list of lines detailing the test's error, as reported in the log. |
| 47 self._test_status = {} |
| 48 |
| 49 # Suppressions are stored here as 'hash': [suppression]. |
| 50 self._suppressions = {} |
| 51 |
| 52 # This may be either text or a number. It will be used in the phrase |
| 53 # '%s disabled' or '%s flaky' on the waterfall display. |
| 54 self._disabled_tests = 0 |
| 55 self._flaky_tests = 0 |
| 56 |
| 57 # Regular expressions for parsing GTest logs. Test names look like |
| 58 # SomeTestCase.SomeTest |
| 59 # SomeName/SomeTestCase.SomeTest/1 |
| 60 # This regexp also matches SomeName.SomeTest/1, which should be harmless. |
| 61 test_name_regexp = r'((\w+/)?\w+\.\w+(/\d+)?)' |
| 62 |
| 63 self._master_name_re = re.compile('\[Running for master: "([^"]*)"') |
| 64 self.master_name = '' |
| 65 |
| 66 self._test_name = re.compile(test_name_regexp) |
| 67 self._test_start = re.compile('\[\s+RUN\s+\] ' + test_name_regexp) |
| 68 self._test_ok = re.compile('\[\s+OK\s+\] ' + test_name_regexp) |
| 69 self._test_fail = re.compile('\[\s+FAILED\s+\] ' + test_name_regexp) |
| 70 self._test_passed = re.compile('\[\s+PASSED\s+\] \d+ tests?.') |
| 71 self._run_test_cases_line = re.compile( |
| 72 '\[\s*\d+\/\d+\]\s+[0-9\.]+s ' + test_name_regexp + ' .+') |
| 73 self._test_timeout = re.compile( |
| 74 'Test timeout \([0-9]+ ms\) exceeded for ' + test_name_regexp) |
| 75 self._disabled = re.compile('\s*YOU HAVE (\d+) DISABLED TEST') |
| 76 self._flaky = re.compile('\s*YOU HAVE (\d+) FLAKY TEST') |
| 77 |
| 78 self._suppression_start = re.compile( |
| 79 'Suppression \(error hash=#([0-9A-F]+)#\):') |
| 80 self._suppression_end = re.compile('^}\s*$') |
| 81 |
| 82 self._retry_message = re.compile('RETRYING FAILED TESTS:') |
| 83 self.retrying_failed = False |
| 84 |
| 85 def GetCurrentTest(self): |
| 86 return self._current_test |
| 87 |
| 88 def _StatusOfTest(self, test): |
| 89 """Returns the status code for the given test, or 'not known'.""" |
| 90 test_status = self._test_status.get(test, ('not known', [])) |
| 91 return test_status[0] |
| 92 |
| 93 def _TestsByStatus(self, status, include_fails, include_flaky): |
| 94 """Returns list of tests with the given status. |
| 95 |
| 96 Args: |
| 97 include_fails: If False, tests containing 'FAILS_' anywhere in their |
| 98 names will be excluded from the list. |
| 99 include_flaky: If False, tests containing 'FLAKY_' anywhere in their |
| 100 names will be excluded from the list. |
| 101 """ |
| 102 test_list = [x[0] for x in self._test_status.items() |
| 103 if self._StatusOfTest(x[0]) == status] |
| 104 |
| 105 if not include_fails: |
| 106 test_list = [x for x in test_list if x.find('FAILS_') == -1] |
| 107 if not include_flaky: |
| 108 test_list = [x for x in test_list if x.find('FLAKY_') == -1] |
| 109 |
| 110 return test_list |
| 111 |
| 112 def _RecordError(self, line, reason): |
| 113 """Record a log line that produced a parsing error. |
| 114 |
| 115 Args: |
| 116 line: text of the line at which the error occurred |
| 117 reason: a string describing the error |
| 118 """ |
| 119 self._internal_error_lines.append('%s: %s [%s]' % |
| 120 (self._line_number, line.strip(), reason)) |
| 121 |
| 122 def RunningTests(self): |
| 123 """Returns list of tests that appear to be currently running.""" |
| 124 return self._TestsByStatus('started', True, True) |
| 125 |
| 126 def ParsingErrors(self): |
| 127 """Returns a list of lines that have caused parsing errors.""" |
| 128 return self._internal_error_lines |
| 129 |
| 130 def ClearParsingErrors(self): |
| 131 """Clears the currently stored parsing errors.""" |
| 132 self._internal_error_lines = ['Cleared.'] |
| 133 |
| 134 def PassedTests(self): |
| 135 """Returns list of tests that passed.""" |
| 136 return self._TestsByStatus('OK', False, False) |
| 137 |
| 138 def FailedTests(self, include_fails=False, include_flaky=False): |
| 139 """Returns list of tests that failed, timed out, or didn't finish |
| 140 (crashed). |
| 141 |
| 142 This list will be incorrect until the complete log has been processed, |
| 143 because it will show currently running tests as having failed. |
| 144 |
| 145 Args: |
| 146 include_fails: If true, all failing tests with FAILS_ in their names will |
| 147 be included. Otherwise, they will only be included if they crashed or |
| 148 timed out. |
| 149 include_flaky: If true, all failing tests with FLAKY_ in their names will |
| 150 be included. Otherwise, they will only be included if they crashed or |
| 151 timed out. |
| 152 |
| 153 """ |
| 154 return (self._TestsByStatus('failed', include_fails, include_flaky) + |
| 155 self._TestsByStatus('timeout', True, True) + |
| 156 self._TestsByStatus('warning', include_fails, include_flaky) + |
| 157 self.RunningTests()) |
| 158 |
| 159 def DisabledTests(self): |
| 160 """Returns the name of the disabled test (if there is only 1) or the number |
| 161 of disabled tests. |
| 162 """ |
| 163 return self._disabled_tests |
| 164 |
| 165 def FlakyTests(self): |
| 166 """Returns the name of the flaky test (if there is only 1) or the number |
| 167 of flaky tests. |
| 168 """ |
| 169 return self._flaky_tests |
| 170 |
| 171 def FailureDescription(self, test): |
| 172 """Returns a list containing the failure description for the given test. |
| 173 |
| 174 If the test didn't fail or timeout, returns []. |
| 175 """ |
| 176 test_status = self._test_status.get(test, ('', [])) |
| 177 return ['%s: ' % test] + test_status[1] |
| 178 |
| 179 def SuppressionHashes(self): |
| 180 """Returns list of suppression hashes found in the log.""" |
| 181 return self._suppressions.keys() |
| 182 |
| 183 def Suppression(self, suppression_hash): |
| 184 """Returns a list containing the suppression for a given hash. |
| 185 |
| 186 If the suppression hash doesn't exist, returns []. |
| 187 """ |
| 188 return self._suppressions.get(suppression_hash, []) |
| 189 |
| 190 def CompletedWithoutFailure(self): |
| 191 """Returns True if all tests completed and no tests failed unexpectedly.""" |
| 192 return self.completed and not self.FailedTests() |
| 193 |
| 194 def ProcessLine(self, line): |
| 195 """This is called once with each line of the test log.""" |
| 196 |
| 197 # Track line number for error messages. |
| 198 self._line_number += 1 |
| 199 |
| 200 # Some tests (net_unittests in particular) run subprocesses which can write |
| 201 # stuff to shared stdout buffer. Sometimes such output appears between new |
| 202 # line and gtest directives ('[ RUN ]', etc) which breaks the parser. |
| 203 # Code below tries to detect such cases and recognize a mixed line as two |
| 204 # separate lines. |
| 205 |
| 206 # List of regexps that parses expects to find at the start of a line but |
| 207 # which can be somewhere in the middle. |
| 208 gtest_regexps = [ |
| 209 self._test_start, |
| 210 self._test_ok, |
| 211 self._test_fail, |
| 212 self._test_passed, |
| 213 ] |
| 214 |
| 215 for regexp in gtest_regexps: |
| 216 match = regexp.search(line) |
| 217 if match: |
| 218 break |
| 219 |
| 220 if not match or match.start() == 0: |
| 221 self._ProcessLine(line) |
| 222 else: |
| 223 self._ProcessLine(line[:match.start()]) |
| 224 self._ProcessLine(line[match.start():]) |
| 225 |
| 226 def _ProcessLine(self, line): |
| 227 """Parses the line and changes the state of parsed tests accordingly. |
| 228 |
| 229 Will recognize newly started tests, OK or FAILED statuses, timeouts, etc. |
| 230 """ |
| 231 |
| 232 # Note: When sharding, the number of disabled and flaky tests will be read |
| 233 # multiple times, so this will only show the most recent values (but they |
| 234 # should all be the same anyway). |
| 235 |
| 236 # Is it a line listing the master name? |
| 237 if not self.master_name: |
| 238 results = self._master_name_re.match(line) |
| 239 if results: |
| 240 self.master_name = results.group(1) |
| 241 |
| 242 results = self._run_test_cases_line.match(line) |
| 243 if results: |
| 244 # A run_test_cases.py output. |
| 245 if self._current_test: |
| 246 if self._test_status[self._current_test][0] == 'started': |
| 247 self._test_status[self._current_test] = ( |
| 248 'timeout', self._failure_description) |
| 249 self._current_test = '' |
| 250 self._failure_description = [] |
| 251 return |
| 252 |
| 253 # Is it a line declaring all tests passed? |
| 254 results = self._test_passed.match(line) |
| 255 if results: |
| 256 self.completed = True |
| 257 self._current_test = '' |
| 258 return |
| 259 |
| 260 # Is it a line reporting disabled tests? |
| 261 results = self._disabled.match(line) |
| 262 if results: |
| 263 try: |
| 264 disabled = int(results.group(1)) |
| 265 except ValueError: |
| 266 disabled = 0 |
| 267 if disabled > 0 and isinstance(self._disabled_tests, int): |
| 268 self._disabled_tests = disabled |
| 269 else: |
| 270 # If we can't parse the line, at least give a heads-up. This is a |
| 271 # safety net for a case that shouldn't happen but isn't a fatal error. |
| 272 self._disabled_tests = 'some' |
| 273 return |
| 274 |
| 275 # Is it a line reporting flaky tests? |
| 276 results = self._flaky.match(line) |
| 277 if results: |
| 278 try: |
| 279 flaky = int(results.group(1)) |
| 280 except ValueError: |
| 281 flaky = 0 |
| 282 if flaky > 0 and isinstance(self._flaky_tests, int): |
| 283 self._flaky_tests = flaky |
| 284 else: |
| 285 # If we can't parse the line, at least give a heads-up. This is a |
| 286 # safety net for a case that shouldn't happen but isn't a fatal error. |
| 287 self._flaky_tests = 'some' |
| 288 return |
| 289 |
| 290 # Is it the start of a test? |
| 291 results = self._test_start.match(line) |
| 292 if results: |
| 293 if self._current_test: |
| 294 if self._test_status[self._current_test][0] == 'started': |
| 295 self._test_status[self._current_test] = ( |
| 296 'timeout', self._failure_description) |
| 297 test_name = results.group(1) |
| 298 self._test_status[test_name] = ('started', ['Did not complete.']) |
| 299 self._current_test = test_name |
| 300 if self.retrying_failed: |
| 301 self._failure_description = self._test_status[test_name][1] |
| 302 self._failure_description.extend(['', 'RETRY OUTPUT:', '']) |
| 303 else: |
| 304 self._failure_description = [] |
| 305 return |
| 306 |
| 307 # Is it a test success line? |
| 308 results = self._test_ok.match(line) |
| 309 if results: |
| 310 test_name = results.group(1) |
| 311 status = self._StatusOfTest(test_name) |
| 312 if status != 'started': |
| 313 self._RecordError(line, 'success while in status %s' % status) |
| 314 if self.retrying_failed: |
| 315 self._test_status[test_name] = ('warning', self._failure_description) |
| 316 else: |
| 317 self._test_status[test_name] = ('OK', []) |
| 318 self._failure_description = [] |
| 319 self._current_test = '' |
| 320 return |
| 321 |
| 322 # Is it a test failure line? |
| 323 results = self._test_fail.match(line) |
| 324 if results: |
| 325 test_name = results.group(1) |
| 326 status = self._StatusOfTest(test_name) |
| 327 if status not in ('started', 'failed', 'timeout'): |
| 328 self._RecordError(line, 'failure while in status %s' % status) |
| 329 # Don't overwrite the failure description when a failing test is listed a |
| 330 # second time in the summary, or if it was already recorded as timing |
| 331 # out. |
| 332 if status not in ('failed', 'timeout'): |
| 333 self._test_status[test_name] = ('failed', self._failure_description) |
| 334 self._failure_description = [] |
| 335 self._current_test = '' |
| 336 return |
| 337 |
| 338 # Is it a test timeout line? |
| 339 results = self._test_timeout.search(line) |
| 340 if results: |
| 341 test_name = results.group(1) |
| 342 status = self._StatusOfTest(test_name) |
| 343 if status not in ('started', 'failed'): |
| 344 self._RecordError(line, 'timeout while in status %s' % status) |
| 345 self._test_status[test_name] = ( |
| 346 'timeout', self._failure_description + ['Killed (timed out).']) |
| 347 self._failure_description = [] |
| 348 self._current_test = '' |
| 349 return |
| 350 |
| 351 # Is it the start of a new valgrind suppression? |
| 352 results = self._suppression_start.match(line) |
| 353 if results: |
| 354 suppression_hash = results.group(1) |
| 355 if suppression_hash in self._suppressions: |
| 356 self._RecordError(line, 'suppression reported more than once') |
| 357 self._suppressions[suppression_hash] = [] |
| 358 self._current_suppression_hash = suppression_hash |
| 359 self._current_suppression = [line] |
| 360 return |
| 361 |
| 362 # Is it the end of a valgrind suppression? |
| 363 results = self._suppression_end.match(line) |
| 364 if results and self._current_suppression_hash: |
| 365 self._current_suppression.append(line) |
| 366 self._suppressions[self._current_suppression_hash] = ( |
| 367 self._current_suppression) |
| 368 self._current_suppression_hash = '' |
| 369 self._current_suppression = [] |
| 370 return |
| 371 |
| 372 # Is it the start of the retry tests? |
| 373 results = self._retry_message.match(line) |
| 374 if results: |
| 375 self.retrying_failed = True |
| 376 return |
| 377 |
| 378 # Random line: if we're in a suppression, collect it. Suppressions are |
| 379 # generated after all tests are finished, so this should always belong to |
| 380 # the current suppression hash. |
| 381 if self._current_suppression_hash: |
| 382 self._current_suppression.append(line) |
| 383 return |
| 384 |
| 385 # Random line: if we're in a test, collect it for the failure description. |
| 386 # Tests may run simultaneously, so this might be off, but it's worth a try. |
| 387 # This also won't work if a test times out before it begins running. |
| 388 if self._current_test: |
| 389 self._failure_description.append(line) |
| 390 |
| 391 # Parse the "Failing tests:" list at the end of the output, and add any |
| 392 # additional failed tests to the list. For example, this includes tests |
| 393 # that crash after the OK line. |
| 394 if self._parsing_failures: |
| 395 results = self._test_name.match(line) |
| 396 if results: |
| 397 test_name = results.group(1) |
| 398 status = self._StatusOfTest(test_name) |
| 399 if status in ('not known', 'OK'): |
| 400 self._test_status[test_name] = ( |
| 401 'failed', ['Unknown error, see stdio log.']) |
| 402 else: |
| 403 self._parsing_failures = False |
| 404 elif line.startswith('Failing tests:'): |
| 405 self._parsing_failures = True |
| 406 |
| 407 |
| 408 class GTestJSONParser(object): |
| 409 # Limit of output snippet lines. Avoids flooding the logs with amount |
| 410 # of output that gums up the infrastructure. |
| 411 OUTPUT_SNIPPET_LINES_LIMIT = 5000 |
| 412 |
| 413 def __init__(self): |
| 414 self.json_file_path = None |
| 415 self.delete_json_file = False |
| 416 |
| 417 self.disabled_tests = set() |
| 418 self.passed_tests = set() |
| 419 self.failed_tests = set() |
| 420 self.flaky_tests = set() |
| 421 self.test_logs = {} |
| 422 |
| 423 self.parsing_errors = [] |
| 424 |
| 425 self.master_name = None |
| 426 |
| 427 def ProcessLine(self, line): |
| 428 # Deliberately do nothing - we parse out-of-band JSON summary |
| 429 # instead of in-band stdout. |
| 430 pass |
| 431 |
| 432 def PassedTests(self): |
| 433 return sorted(self.passed_tests) |
| 434 |
| 435 def FailedTests(self, include_fails=False, include_flaky=False): |
| 436 return sorted(self.failed_tests) |
| 437 |
| 438 def FailureDescription(self, test): |
| 439 return self.test_logs.get(test, []) |
| 440 |
| 441 @staticmethod |
| 442 def SuppressionHashes(): |
| 443 return [] |
| 444 |
| 445 def ParsingErrors(self): |
| 446 return self.parsing_errors |
| 447 |
| 448 def ClearParsingErrors(self): |
| 449 self.parsing_errors = ['Cleared.'] |
| 450 |
| 451 def DisabledTests(self): |
| 452 return len(self.disabled_tests) |
| 453 |
| 454 def FlakyTests(self): |
| 455 return len(self.flaky_tests) |
| 456 |
| 457 @staticmethod |
| 458 def RunningTests(): |
| 459 return [] |
| 460 |
| 461 def PrepareJSONFile(self, cmdline_path): |
| 462 if cmdline_path: |
| 463 self.json_file_path = cmdline_path |
| 464 # If the caller requested JSON summary, do not delete it. |
| 465 self.delete_json_file = False |
| 466 else: |
| 467 fd, self.json_file_path = tempfile.mkstemp() |
| 468 os.close(fd) |
| 469 # When we create the file ourselves, delete it to avoid littering. |
| 470 self.delete_json_file = True |
| 471 return self.json_file_path |
| 472 |
| 473 def ProcessJSONFile(self): |
| 474 if not self.json_file_path: |
| 475 return |
| 476 |
| 477 with open(self.json_file_path) as json_file: |
| 478 try: |
| 479 json_output = json_file.read() |
| 480 json_data = json.loads(json_output) |
| 481 except ValueError: |
| 482 # Only signal parsing error if the file is non-empty. Empty file |
| 483 # most likely means the binary doesn't support JSON output. |
| 484 if json_output: |
| 485 self.parsing_errors = json_output.split('\n') |
| 486 else: |
| 487 self.ProcessJSONData(json_data) |
| 488 |
| 489 if self.delete_json_file: |
| 490 os.remove(self.json_file_path) |
| 491 |
| 492 def ProcessJSONData(self, json_data): |
| 493 # TODO(phajdan.jr): Require disabled_tests to be present (May 2014). |
| 494 self.disabled_tests.update(json_data.get('disabled_tests', [])) |
| 495 |
| 496 for iteration_data in json_data['per_iteration_data']: |
| 497 for test_name, test_runs in iteration_data.iteritems(): |
| 498 if test_runs[-1]['status'] == 'SUCCESS': |
| 499 self.passed_tests.add(test_name) |
| 500 else: |
| 501 self.failed_tests.add(test_name) |
| 502 |
| 503 if len(test_runs) > 1: |
| 504 self.flaky_tests.add(test_name) |
| 505 |
| 506 self.test_logs.setdefault(test_name, []) |
| 507 for run_index, run_data in enumerate(test_runs, start=1): |
| 508 run_lines = ['%s (run #%d):' % (test_name, run_index)] |
| 509 # Make sure the annotations are ASCII to avoid character set related |
| 510 # errors. They are mostly informational anyway, and more detailed |
| 511 # info can be obtained from the original JSON output. |
| 512 ascii_lines = run_data['output_snippet'].encode('ascii', |
| 513 errors='replace') |
| 514 decoded_lines = CompressList( |
| 515 ascii_lines.decode('string_escape').split('\n'), |
| 516 self.OUTPUT_SNIPPET_LINES_LIMIT, |
| 517 '<truncated, full output is in gzipped JSON ' |
| 518 'output at end of step>') |
| 519 run_lines.extend(decoded_lines) |
| 520 self.test_logs[test_name].extend(run_lines) |
OLD | NEW |