[Findit] Pipeline change to save suspected cls to data store.

appengine/findit/waterfall/test/identify_try_job_culprit_pipeline_test.py

 # 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.
 
 from testing_utils import testing
 
 from common.git_repository import GitRepository
 from common.waterfall import failure_type
+from model import analysis_approach_type
 from model import analysis_status
 from model import result_status
 from model.wf_analysis import WfAnalysis
+from model.wf_suspected_cl import WfSuspectedCL
 from model.wf_try_job import WfTryJob
 from model.wf_try_job_data import WfTryJobData
 from waterfall import identify_try_job_culprit_pipeline
 from waterfall.identify_try_job_culprit_pipeline import(
     IdentifyTryJobCulpritPipeline)
 
 
 class IdentifyTryJobCulpritPipelineTest(testing.AppengineTestCase):
 
   def _MockGetChangeLog(self, revision):
     class MockedChangeLog(object):
 
       def __init__(self, commit_position, code_review_url):
         self.commit_position = commit_position
         self.code_review_url = code_review_url
 
     mock_change_logs = {}
-    mock_change_logs['rev1'] = MockedChangeLog('1', 'url_1')
-    mock_change_logs['rev2'] = MockedChangeLog('2', 'url_2')
+    mock_change_logs['rev1'] = MockedChangeLog(1, 'url_1')
+    mock_change_logs['rev2'] = MockedChangeLog(2, 'url_2')
     return mock_change_logs.get(revision)
 
   def setUp(self):
     super(IdentifyTryJobCulpritPipelineTest, self).setUp()
 
     self.mock(GitRepository, 'GetChangeLog', self._MockGetChangeLog)
 
   def testGetFailedRevisionFromResultsDict(self):
     self.assertIsNone(
         identify_try_job_culprit_pipeline._GetFailedRevisionFromResultsDict({}))
@@ skipping 53 matching lines @@
   def testGetResultAnalysisStatusWithTryJobCulpritNotFoundUntriaged(self):
     # Heuristic analysis provided no results, but the try job found a culprit.
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.result_status = result_status.NOT_FOUND_UNTRIAGED
     analysis.put()
 
     result = {
         'culprit': {
             'compile': {
                 'revision': 'rev1',
-                'commit_position': '1',
+                'commit_position': 1,
                 'url': 'url_1',
                 'repo_name': 'chromium'
             }
         }
     }
 
     status = identify_try_job_culprit_pipeline._GetResultAnalysisStatus(
         analysis, result)
 
     self.assertEqual(status, result_status.FOUND_UNTRIAGED)
 
   def testGetResultAnalysisStatusWithTryJobCulpritNotFoundCorrect(self):
     # Heuristic analysis found no results, which was correct. In this case, the
     # try job result is actually a false positive.
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.result_status = result_status.NOT_FOUND_CORRECT
     analysis.put()
 
     result = {
         'culprit': {
             'compile': {
                 'revision': 'rev1',
-                'commit_position': '1',
+                'commit_position': 1,
                 'url': 'url_1',
                 'repo_name': 'chromium'
             }
         }
     }
 
     status = identify_try_job_culprit_pipeline._GetResultAnalysisStatus(
         analysis, result)
 
     self.assertEqual(status, result_status.FOUND_UNTRIAGED)
 
   def testGetResultanalysisStatusWithTryJobCulpritNotFoundIncorrect(self):
     # Heuristic analysis found no results and was triaged to incorrect before a
     # try job result was found. In this case the try job result should override
     # the heuristic result.
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.result_status = result_status.NOT_FOUND_INCORRECT
     analysis.put()
 
     result = {
         'culprit': {
             'compile': {
                 'revision': 'rev1',
-                'commit_position': '1',
+                'commit_position': 1,
                 'url': 'url_1',
                 'repo_name': 'chromium'
             }
         }
     }
 
     status = identify_try_job_culprit_pipeline._GetResultAnalysisStatus(
         analysis, result)
 
     self.assertEqual(status, result_status.FOUND_UNTRIAGED)
 
   def testGetResultanalysisStatusWithTryJobCulpritNoHeuristicResult(self):
     # In this case, the try job found a result before the heuristic result is
     # available. This case should generally never happen, as heuristic analysis
     # is usually much faster than try jobs.
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.put()
 
     result = {
         'culprit': {
             'compile': {
                 'revision': 'rev1',
-                'commit_position': '1',
+                'commit_position': 1,
                 'url': 'url_1',
                 'repo_name': 'chromium'
             }
         }
     }
 
     status = identify_try_job_culprit_pipeline._GetResultAnalysisStatus(
         analysis, result)
 
     self.assertEqual(status, result_status.FOUND_UNTRIAGED)
@@ skipping 14 matching lines @@
     # In this case, heuristic analysis found the correct culprit. The try job
     # result should not overwrite it.
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.result_status = result_status.FOUND_CORRECT
     analysis.put()
 
     result = {
         'culprit': {
             'compile': {
                 'revision': 'rev1',
-                'commit_position': '1',
+                'commit_position': 1,
                 'url': 'url_1',
                 'repo_name': 'chromium'
             }
         }
     }
 
     status = identify_try_job_culprit_pipeline._GetResultAnalysisStatus(
         analysis, result)
     self.assertEqual(status, result_status.FOUND_CORRECT)
 
@@ skipping 21 matching lines @@
 
     result = {}
 
     status = identify_try_job_culprit_pipeline._GetResultAnalysisStatus(
         analysis, result)
     self.assertEqual(status, result_status.NOT_FOUND_INCORRECT)
 
   def testGetSuspectedCLsForCompileTryJob(self):
     heuristic_suspected_cl = {
         'revision': 'rev1',
-        'commit_position': '1',
+        'commit_position': 1,
         'url': 'url_1',
         'repo_name': 'chromium'
     }
 
     compile_suspected_cl = {
         'revision': 'rev2',
-        'commit_position': '2',
+        'commit_position': 2,
         'url': 'url_2',
         'repo_name': 'chromium'
     }
 
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.suspected_cls = [heuristic_suspected_cl]
     analysis.put()
 
-    result = {
-        'culprit': {
-            'compile': compile_suspected_cl
-        }
+    try_job_suspected_cls = {
+        'rev2': compile_suspected_cl
     }
 
     self.assertEqual(
-        identify_try_job_culprit_pipeline._GetSuspectedCLs(analysis, result),
+        identify_try_job_culprit_pipeline._GetSuspectedCLs(
+            analysis, try_job_suspected_cls),
         [heuristic_suspected_cl, compile_suspected_cl])
 
   def testGetSuspectedCLsForTestTryJobAndHeuristicResultsSame(self):
     suspected_cl = {
         'revision': 'rev1',
-        'commit_position': '1',
+        'commit_position': 1,
         'url': 'url_1',
         'repo_name': 'chromium'
     }
 
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.suspected_cls = [suspected_cl]
     analysis.put()
 
-    result = {
-        'culprit': {
-            'compile': suspected_cl
-        }
+    try_job_suspected_cls = {
+        'rev1': suspected_cl
     }
 
-    self.assertEqual(
-        identify_try_job_culprit_pipeline._GetSuspectedCLs(analysis, result),
-        [suspected_cl])
+    updated_cls = identify_try_job_culprit_pipeline._GetSuspectedCLs(
+      analysis, try_job_suspected_cls)
+    self.assertEqual(updated_cls, [suspected_cl])
 
   def testGetSuspectedCLsForTestTryJob(self):
     suspected_cl1 = {
         'revision': 'rev1',
-        'commit_position': '1',
+        'commit_position': 1,
         'url': 'url_1',
         'repo_name': 'chromium'
     }
     suspected_cl2 = {
         'revision': 'rev2',
-        'commit_position': '2',
+        'commit_position': 2,
         'url': 'url_2',
         'repo_name': 'chromium'
     }
     suspected_cl3 = {
         'revision': 'rev3',
-        'commit_position': '3',
+        'commit_position': 3,
         'url': 'url_3',
         'repo_name': 'chromium'
     }
 
     analysis = WfAnalysis.Create('m', 'b', 1)
-    analysis.suspected_cls = []
+    analysis.suspected_cls = [suspected_cl3]
     analysis.put()
 
-    result = {
-        'culprit': {
-            'a_test': {
-                'tests': {
-                    'a_test1': suspected_cl1,
-                    'a_test2': suspected_cl1
-                }
-            },
-            'b_test': {
-                'tests': {
-                    'b_test1': suspected_cl2
-                }
-            },
-            'c_test': {
-                'revision': 'rev3',
-                'commit_position': '3',
-                'url': 'url_3',
-                'repo_name': 'chromium',
-                'tests': {}
-            }
-        }
+    try_job_suspected_cls = {
+        'rev1': suspected_cl1,
+        'rev2': suspected_cl2
     }
 
+    result = identify_try_job_culprit_pipeline._GetSuspectedCLs(
+        analysis, try_job_suspected_cls)
     self.assertEqual(
-        identify_try_job_culprit_pipeline._GetSuspectedCLs(analysis, result),
-        [suspected_cl3, suspected_cl2, suspected_cl1])
+        result, [suspected_cl3, suspected_cl1, suspected_cl2])
 
   def testGetSuspectedCLsForTestTryJobWithHeuristicResult(self):
     suspected_cl = {
         'revision': 'rev1',
-        'commit_position': '1',
+        'commit_position': 1,
         'url': 'url_1',
         'repo_name': 'chromium'
     }
 
     analysis = WfAnalysis.Create('m', 'b', 1)
     analysis.suspected_cls = [suspected_cl]
     analysis.put()
 
-    result = {
-        'culprit': {
-            'a_test': {
-                'revision': 'rev1',
-                'commit_position': '1',
-                'url': 'url_1',
-                'repo_name': 'chromium',
-                'tests': {}
-            }
-        }
-    }
     self.assertEqual(
-        identify_try_job_culprit_pipeline._GetSuspectedCLs(analysis, result),
+        identify_try_job_culprit_pipeline._GetSuspectedCLs(analysis, {}),
         [suspected_cl])
 
   def testIdentifyCulpritForCompileTryJobNoCulprit(self):
     master_name = 'm'
     builder_name = 'b'
     build_number = 1
     try_job_id = '1'
 
     try_job = WfTryJob.Create(master_name, builder_name, build_number)
     try_job.put()
@@ skipping 50 matching lines @@
     analysis.put()
 
     pipeline = IdentifyTryJobCulpritPipeline()
     culprit = pipeline.run(
         master_name, builder_name, build_number, ['rev1'],
         failure_type.COMPILE, '1', compile_result)
 
     expected_culprit = 'rev2'
     expected_suspected_cl = {
         'revision': 'rev2',
-        'commit_position': '2',
+        'commit_position': 2,
         'url': 'url_2',
         'repo_name': 'chromium'
     }
     expected_compile_result = {
         'report': {
             'result': {
                 'rev1': 'passed',
                 'rev2': 'failed'
             }
         },
@@ skipping 80 matching lines @@
     self.assertIsNone(analysis.suspected_cls)
 
   def testIdentifyCulpritForTestTryJobNoTryJobResultWithHeuristicResult(self):
     master_name = 'm'
     builder_name = 'b'
     build_number = 1
     try_job_id = '1'
 
     suspected_cl = {
         'revision': 'rev1',
-        'commit_position': '1',
+        'commit_position': 1,
         'url': 'url_1',
         'repo_name': 'chromium'
     }
 
     WfTryJobData.Create(try_job_id).put()
     try_job = WfTryJob.Create(master_name, builder_name, build_number)
     try_job.status = analysis_status.RUNNING
     try_job.put()
 
     # Heuristic analysis already provided some results.
@@ skipping 130 matching lines @@
                 'rev1': {
                     'a_test': {
                         'status': 'failed',
                         'valid': True,
                         'failures': ['a_test1']
                     },
                     'b_test': {
                         'status': 'failed',
                         'valid': True,
                         'failures': ['b_test1']
-                    },
-                    'c_test': {
-                        'status': 'passed',
-                        'valid': True
                     }
                 },
                 'rev2': {
                     'a_test': {
                         'status': 'failed',
                         'valid': True,
                         'failures': ['a_test1', 'a_test2']
                     },
                     'b_test': {
                         'status': 'passed',
                         'valid': True
-                    },
-                    'c_test': {
-                        'status': 'failed',
-                        'valid': True,
-                        'failures': []
                     }
                 }
             }
         },
         'url': 'url',
         'try_job_id': try_job_id
     }
 
     WfTryJobData.Create(try_job_id).put()
     try_job = WfTryJob.Create(master_name, builder_name, build_number)
     try_job.status = analysis_status.RUNNING
     try_job.test_results = [test_result]
     try_job.put()
     analysis = WfAnalysis.Create(master_name, builder_name, build_number)
     analysis.put()
 
     pipeline = IdentifyTryJobCulpritPipeline()
     culprit = pipeline.run(
         master_name, builder_name, build_number, ['rev1', 'rev2'],
         failure_type.TEST, '1', test_result)
 
     a_test1_suspected_cl = {
         'revision': 'rev1',
-        'commit_position': '1',
+        'commit_position': 1,
         'url': 'url_1',
         'repo_name': 'chromium'
     }
     a_test2_suspected_cl = {
         'revision': 'rev2',
-        'commit_position': '2',
+        'commit_position': 2,
         'url': 'url_2',
         'repo_name': 'chromium'
     }
 
     b_test1_suspected_cl = a_test1_suspected_cl
 
     expected_test_result = {
         'report': {
             'result': {
                 'rev1': {
                     'a_test': {
                         'status': 'failed',
                         'valid': True,
                         'failures': ['a_test1']
                     },
                     'b_test': {
                         'status': 'failed',
                         'valid': True,
                         'failures': ['b_test1']
-                    },
-                    'c_test': {
-                        'status': 'passed',
-                        'valid': True
                     }
                 },
                 'rev2': {
                     'a_test': {
                         'status': 'failed',
                         'valid': True,
                         'failures': ['a_test1', 'a_test2']
                     },
                     'b_test': {
                         'status': 'passed',
                         'valid': True
-                    },
-                    'c_test': {
-                        'status': 'failed',
-                        'valid': True,
-                        'failures': []
                     }
                 }
             }
         },
         'url': 'url',
         'try_job_id': try_job_id,
         'culprit': {
             'a_test': {
                 'tests': {
                     'a_test1': a_test1_suspected_cl,
                     'a_test2': a_test2_suspected_cl
                 }
             },
             'b_test': {
                 'tests': {
                     'b_test1': b_test1_suspected_cl
                 }
-            },
-            'c_test': {
-                'revision': 'rev2',
-                'commit_position': '2',
-                'url': 'url_2',
-                'repo_name': 'chromium',
-                'tests': {}
             }
         }
     }
 
     self.assertEqual(expected_test_result['culprit'], culprit)
 
     try_job = WfTryJob.Get(master_name, builder_name, build_number)
     self.assertEqual(expected_test_result, try_job.test_results[-1])
     self.assertEqual(analysis_status.COMPLETED, try_job.status)
 
     try_job_data = WfTryJobData.Get(try_job_id)
     analysis = WfAnalysis.Get(master_name, builder_name, build_number)
     expected_culprit_data = {
         'a_test': {
             'a_test1': 'rev1',
             'a_test2': 'rev2',
         },
         'b_test': {
             'b_test1': 'rev1',
-        },
-        'c_test': 'rev2'
         }
+    }
     self.assertEqual(expected_culprit_data, try_job_data.culprits)
     self.assertEqual(analysis.result_status,
                      result_status.FOUND_UNTRIAGED)
     self.assertEqual(analysis.suspected_cls,
-                     [a_test2_suspected_cl, a_test1_suspected_cl])
+                     [a_test1_suspected_cl, a_test2_suspected_cl])
 
   def testAnalysisIsUpdatedOnlyIfStatusOrSuspectedCLsChanged(self):
     master_name = 'm'
     builder_name = 'b'
     build_number = 1
     try_job_id = '1'
+    repo_name = 'chromium'
+    revision = 'rev1'
+    commit_position = 1
 
-    suspected_cl = {
-        'revision': 'rev1',
-        'commit_position': '1',
+    heuristic_suspected_cl = {
+        'revision': revision,
+        'commit_position': commit_position,
         'url': 'url_1',
-        'repo_name': 'chromium'
+        'repo_name': repo_name
     }
 
     analysis = WfAnalysis.Create(master_name, builder_name, build_number)
-    analysis.suspected_cls = [suspected_cl]
+    analysis.suspected_cls = [heuristic_suspected_cl]
     analysis.result_status = result_status.FOUND_UNTRIAGED
     analysis.put()
     version = analysis.version
+
+    build_key = '%s/%s/%d' % (master_name, builder_name, build_number)
+    suspected_cl = WfSuspectedCL.Create(repo_name, revision, commit_position)
+    suspected_cl.approach = analysis_approach_type.HEURISTIC
+    suspected_cl.builds = {
+        build_key: {
+          'approach': analysis_approach_type.HEURISTIC,
+          'failure_type':failure_type.COMPILE,
+          'failures': {'compile':[]},
+          'status': None,
+          'top_score': 4
+        }
+    }
+    suspected_cl.put()
+
     compile_result = {
         'report': {
             'result': {
-                'rev1': 'failed',
+                revision: 'failed',
             },
         },
     }
 
     try_job_data = WfTryJobData.Create(try_job_id)
     try_job_data.put()
 
     try_job = WfTryJob.Create(master_name, builder_name, build_number)
     try_job.status = analysis_status.RUNNING
     try_job.compile_results = [{
         'report': {
             'result': {
-                'rev1': 'failed',
+                revision: 'failed',
             },
         },
         'try_job_id': try_job_id,
     }]
 
     try_job.put()
 
     pipeline = IdentifyTryJobCulpritPipeline()
-    pipeline.run(master_name, builder_name, build_number, ['rev1'],
+    pipeline.run(master_name, builder_name, build_number, [revision],
                  failure_type.COMPILE, '1', compile_result)
 
     self.assertEqual(analysis.result_status,
                      result_status.FOUND_UNTRIAGED)
-    self.assertEqual(analysis.suspected_cls, [suspected_cl])
+    self.assertEqual(analysis.suspected_cls, [heuristic_suspected_cl])
     self.assertEqual(version, analysis.version)  # No update to analysis.
 
+    expected_builds = {
+        build_key: {
+          'approach': analysis_approach_type.BOTH,
+          'failure_type':failure_type.COMPILE,
+          'failures': {'compile':[]},
+          'status': None,
+          'top_score': 4
+        }
+    }
+    suspected_cl = WfSuspectedCL.Get(repo_name, revision)
+    self.assertEqual(analysis_approach_type.BOTH, suspected_cl.approach)
+    self.assertEqual(expected_builds, suspected_cl.builds)
+
   def testFindCulpritForEachTestFailureRevisionNotRun(self):
     blame_list = ['rev1']
     result = {
         'report': {
             'result': {
                 'rev2': 'passed'
             }
         }
     }
 
@@ skipping 40 matching lines @@
     try_job = WfTryJob.Create(master_name, builder_name, build_number).put()
     pipeline = IdentifyTryJobCulpritPipeline()
     culprit = pipeline.run(master_name, builder_name, build_number, ['rev1'],
                  failure_type.TEST, None, None)
     self.assertIsNone(culprit)
 
     try_job = WfTryJob.Get(master_name, builder_name, build_number)
     self.assertEqual(try_job.test_results, [])
     self.assertEqual(try_job.status, analysis_status.COMPLETED)
 
-
   def testNotifyCulprits(self):
     instances = []
     class Mocked_SendNotificationForCulpritPipeline(object):
       def __init__(self, *args):
         self.args = args
         self.started = False
         instances.append(self)
 
       def start(self):
         self.started = True
@@ skipping 102 matching lines @@
         self.started = True
 
     self.mock(
         identify_try_job_culprit_pipeline, 'SendNotificationForCulpritPipeline',
         Mocked_SendNotificationForCulpritPipeline)
 
     identify_try_job_culprit_pipeline._NotifyCulprits(
         master_name, builder_name, build_number, None,
         heuristic_cls, compile_suspected_cl)
     self.assertEqual(1, len(instances))
-    self.assertTrue(instances[0].started)
+    self.assertTrue(instances[0].started)
+
+  def testGetTestFailureCausedByCLResultNone(self):
+    self.assertIsNone(
+        identify_try_job_culprit_pipeline._GetTestFailureCausedByCL(None))