def _gather_baselines(self, results_json): # Rebaseline server and it's associated JavaScript expected the tests subtree to # be key-value pairs instead of hierarchical. # FIXME: make the rebaseline server use the hierarchical tree. new_tests_subtree = {} def gather_baselines_for_test(test_name, result_dict): result = JSONTestResult(test_name, result_dict) if result.did_pass_or_run_as_expected(): return result_dict['state'] = STATE_NEEDS_REBASELINE result_dict['baselines'] = get_test_baselines(test_name, self._test_config) new_tests_subtree[test_name] = result_dict for_each_test(results_json['tests'], gather_baselines_for_test) results_json['tests'] = new_tests_subtree
def _gather_baselines(self, results_json): # Rebaseline server and it's associated JavaScript expected the tests subtree to # be key-value pairs instead of hierarchical. # FIXME: make the rebaseline server use the hierarchical tree. new_tests_subtree = {} def gather_baselines_for_test(test_name, result_dict): result = JSONTestResult(test_name, result_dict) if result.did_pass_or_run_as_expected(): return result_dict["state"] = STATE_NEEDS_REBASELINE result_dict["baselines"] = get_test_baselines(test_name, self._test_config) new_tests_subtree[test_name] = result_dict for_each_test(results_json["tests"], gather_baselines_for_test) results_json["tests"] = new_tests_subtree
def print_unexpected_results(self, summarized_results, enabled_pixel_tests_in_retry=False): passes = {} flaky = {} regressions = {} def add_to_dict_of_lists(dict, key, value): dict.setdefault(key, []).append(value) def add_result(test, results, passes=passes, flaky=flaky, regressions=regressions): actual = results['actual'].split(" ") expected = results['expected'].split(" ") if 'is_unexpected' not in results or not results['is_unexpected']: # Don't print anything for tests that ran as expected. return if actual == ['PASS']: if 'CRASH' in expected: add_to_dict_of_lists(passes, 'Expected to crash, but passed', test) elif 'TIMEOUT' in expected: add_to_dict_of_lists(passes, 'Expected to timeout, but passed', test) else: add_to_dict_of_lists(passes, 'Expected to fail, but passed', test) elif enabled_pixel_tests_in_retry and actual == ['TEXT', 'IMAGE+TEXT']: add_to_dict_of_lists(regressions, actual[0], test) elif len(actual) > 1 and actual[-1] in expected: # We group flaky tests by the first actual result we got. add_to_dict_of_lists(flaky, actual[0], test) else: add_to_dict_of_lists(regressions, actual[0], test) layouttestresults.for_each_test(summarized_results['tests'], add_result) if len(passes) or len(flaky) or len(regressions): self._print("") if len(passes): for key, tests in passes.iteritems(): self._print("%s: (%d)" % (key, len(tests))) tests.sort() for test in tests: self._print(" %s" % test) self._print("") self._print("") if len(flaky): descriptions = TestExpectations.EXPECTATION_DESCRIPTIONS for key, tests in flaky.iteritems(): result_type = TestExpectations.EXPECTATIONS[key.lower()] self._print("Unexpected flakiness: %s (%d)" % (descriptions[result_type], len(tests))) tests.sort() for test in tests: result = layouttestresults.result_for_test(summarized_results['tests'], test) actual = result['actual'].split(" ") expected = result['expected'].split(" ") # FIXME: clean this up once the old syntax is gone new_expectations_list = [TestExpectationParser._inverted_expectation_tokens[exp] for exp in list(set(actual) | set(expected))] self._print(" %s [ %s ]" % (test, " ".join(new_expectations_list))) self._print("") self._print("") if len(regressions): descriptions = TestExpectations.EXPECTATION_DESCRIPTIONS for key, tests in regressions.iteritems(): result_type = TestExpectations.EXPECTATIONS[key.lower()] self._print("Regressions: Unexpected %s (%d)" % (descriptions[result_type], len(tests))) tests.sort() for test in tests: result = layouttestresults.result_for_test(summarized_results['tests'], test) actual = result['actual'].split(" ") expected = result['expected'].split(" ") new_expectations_list = [TestExpectationParser._inverted_expectation_tokens[exp] for exp in actual] self._print(" %s [ %s ]" % (test, " ".join(new_expectations_list))) self._print("") if len(summarized_results['tests']) and self.debug_logging: self._print("%s" % ("-" * 78))
def print_unexpected_results(self, summarized_results, enabled_pixel_tests_in_retry=False): passes = {} flaky = {} regressions = {} def add_to_dict_of_lists(dict, key, value): dict.setdefault(key, []).append(value) def add_result(test, results, passes=passes, flaky=flaky, regressions=regressions): actual = results['actual'].split(" ") expected = results['expected'].split(" ") if 'is_unexpected' not in results or not results['is_unexpected']: # Don't print anything for tests that ran as expected. return if actual == ['PASS']: if 'CRASH' in expected: add_to_dict_of_lists(passes, 'Expected to crash, but passed', test) elif 'TIMEOUT' in expected: add_to_dict_of_lists(passes, 'Expected to timeout, but passed', test) else: add_to_dict_of_lists(passes, 'Expected to fail, but passed', test) elif enabled_pixel_tests_in_retry and actual == ['TEXT', 'IMAGE+TEXT']: add_to_dict_of_lists(regressions, actual[0], test) elif len(actual) > 1: # We group flaky tests by the first actual result we got. add_to_dict_of_lists(flaky, actual[0], test) else: add_to_dict_of_lists(regressions, results['actual'], test) layouttestresults.for_each_test(summarized_results['tests'], add_result) if len(passes) or len(flaky) or len(regressions): self._print("") if len(passes): for key, tests in passes.iteritems(): self._print("%s: (%d)" % (key, len(tests))) tests.sort() for test in tests: self._print(" %s" % test) self._print("") self._print("") if len(flaky): descriptions = TestExpectations.EXPECTATION_DESCRIPTIONS for key, tests in flaky.iteritems(): result = TestExpectations.EXPECTATIONS[key.lower()] self._print("Unexpected flakiness: %s (%d)" % (descriptions[result], len(tests))) tests.sort() for test in tests: result = layouttestresults.result_for_test(summarized_results['tests'], test) actual = result['actual'].split(" ") expected = result['expected'].split(" ") result = TestExpectations.EXPECTATIONS[key.lower()] # FIXME: clean this up once the old syntax is gone new_expectations_list = [TestExpectationParser._inverted_expectation_tokens[exp] for exp in list(set(actual) | set(expected))] self._print(" %s [ %s ]" % (test, " ".join(new_expectations_list))) self._print("") self._print("") if len(regressions): descriptions = TestExpectations.EXPECTATION_DESCRIPTIONS for key, tests in regressions.iteritems(): result = TestExpectations.EXPECTATIONS[key.lower()] self._print("Regressions: Unexpected %s (%d)" % (descriptions[result], len(tests))) tests.sort() for test in tests: self._print(" %s [ %s ]" % (test, TestExpectationParser._inverted_expectation_tokens[key])) self._print("") if len(summarized_results['tests']) and self.debug_logging: self._print("%s" % ("-" * 78))