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

appengine/findit/waterfall/identify_try_job_culprit_pipeline.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 collections import defaultdict
 import logging
 
 from google.appengine.ext import ndb
 
 from common.git_repository import GitRepository
 from common.http_client_appengine import HttpClientAppengine as HttpClient
 from common.pipeline_wrapper import BasePipeline
 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_try_job import WfTryJob
 from model.wf_try_job_data import WfTryJobData
 from waterfall.send_notification_for_culprit_pipeline import (
     SendNotificationForCulpritPipeline)
+from waterfall import suspected_cl_util
 
 
 GIT_REPO = GitRepository(
     'https://chromium.googlesource.com/chromium/src.git', HttpClient())
 
 
 def _GetResultAnalysisStatus(analysis, result):
   """Returns the analysis status based on existing status and try job result.
 
   Args:
@@ skipping 13 matching lines @@
   if (try_job_found_culprit and
       (old_result_status is None or
        old_result_status == result_status.NOT_FOUND_UNTRIAGED or
        old_result_status == result_status.NOT_FOUND_INCORRECT or
        old_result_status == result_status.NOT_FOUND_CORRECT)):
     return result_status.FOUND_UNTRIAGED
 
   return old_result_status
 
 
-def _GetSuspectedCLs(analysis, result):
+def _GetTestTryJobSuspectedCLs(culprit):
+  # Suspected CLs are from test failures.
+  suspected_cl_revisions = []
+  suspected_cls = []
+  for results in culprit.itervalues():
+    for test_cl_info in results['tests'].itervalues():
+      revision = test_cl_info['revision']
+      if revision not in suspected_cl_revisions:
+        suspected_cl_revisions.append(revision)
+        suspected_cls.append(test_cl_info)
+
+  return suspected_cls
+
+
+def _GetTryJobSuspectedCLs(result):
+  """Returns all suspected CLs found by the try job.
+
+  Args:
+    result: A result dict containing the culprit from the results of
+      this try job.
+
+  Returns:
+    A list of suspected CLs found by this try job.
+  """
+  if not result or not result.get('culprit'):
+    return []
+
+  culprit = result.get('culprit')
+  compile_cl_info = culprit.get('compile')
+
+  if compile_cl_info:
+    # Suspected CL is from compile failure.
+    return [compile_cl_info]
+
+  return _GetTestTryJobSuspectedCLs(culprit)
+
+
+def _GetSuspectedCLs(analysis, try_job_suspected_cls):
   """Returns a list of suspected CLs.
 
   Args:
     analysis: The WfAnalysis entity corresponding to this try job.
-    result: A result dict containing the culprit from the results of
-      this try job.
+    is_compile_try_job: True if the try job is for compile failure,

[lijeffrey, 2016/08/17 19:31:11]

where is this arg?

[chanli, 2016/09/12 20:56:15]

Done.

+      otherwise False.
+    try_job_suspected_cls: A list of suspected CLs found by the try job.
 
   Returns:
     A combined list of suspected CLs from those already in analysis and those
     found by this try job.
   """
   suspected_cls = analysis.suspected_cls[:] if analysis.suspected_cls else []
   suspected_cl_revisions = [cl['revision'] for cl in suspected_cls]
-  culprit = result.get('culprit')
-  compile_cl_info = culprit.get('compile')
 
-  if compile_cl_info:
-    # Suspected CL is from compile failure.
-    revision = compile_cl_info.get('revision')
-    if revision not in suspected_cl_revisions:
-      suspected_cl_revisions.append(revision)
-      suspected_cls.append(compile_cl_info)
-    return suspected_cls
-
-  # Suspected CLs are from test failures.
-  for results in culprit.itervalues():
-    if results.get('revision'):
-      # Non swarming test failures, only have step level failure info.
-      revision = results.get('revision')
-      cl_info = {
-          'url': results.get('url'),
-          'repo_name': results.get('repo_name'),
-          'revision': results.get('revision'),
-          'commit_position': results.get('commit_position')
-      }
-      if revision not in suspected_cl_revisions:
-        suspected_cl_revisions.append(revision)
-        suspected_cls.append(cl_info)
-    else:
-      for test_cl_info in results['tests'].values():
-        revision = test_cl_info.get('revision')
-        if revision not in suspected_cl_revisions:
-          suspected_cl_revisions.append(revision)
-          suspected_cls.append(test_cl_info)
+  for try_job_suspected_cl in try_job_suspected_cls:
+    if try_job_suspected_cl['revision'] not in suspected_cl_revisions:
+      suspected_cl_revisions.append(try_job_suspected_cl['revision'])
+      suspected_cls.append(try_job_suspected_cl)
 
   return suspected_cls
 
 
 def _GetFailedRevisionFromResultsDict(results_dict):
   """Finds the failed revision from the given dict of revisions.
 
   Args:
     results_dict: (dict) A dict that maps revisions to their results. For
     example:
@@ skipping 52 matching lines @@
       if (not test_result['valid'] or
           test_result['status'] != 'failed'):  # pragma: no cover
         continue
 
       failed_revisions.add(revision)
 
       if step not in culprit_map:
         culprit_map[step] = {
             'tests': {}
         }
-      if (not test_result['failures'] and
-          not culprit_map[step].get('revision')):
-        # Non swarming test failures, only have step level failure info.
-        culprit_map[step]['revision'] = revision
       for failed_test in test_result['failures']:
         # Swarming tests, gets first failed revision for each test.
         if failed_test not in culprit_map[step]['tests']:
           culprit_map[step]['tests'][failed_test] = {
               'revision': revision
           }
 
   return culprit_map, list(failed_revisions)
 
 
@@ skipping 56 matching lines @@
           }
           failed_revisions.add(revision)
       return culprit_map, list(failed_revisions)
 
     return _GetCulpritsForTestsFromResultsDict(
         blame_list, result['report'].get('result'))
 
   def _UpdateCulpritMapWithCulpritInfo(self, culprit_map, culprits):
     """Fills in commit_position and review url for each failed rev in map."""
     for step_culprit in culprit_map.values():
-      if step_culprit.get('revision'):
-        culprit = culprits[step_culprit['revision']]
-        step_culprit['commit_position'] = culprit['commit_position']
-        step_culprit['url'] = culprit['url']
-        step_culprit['repo_name'] = culprit['repo_name']
       for test_culprit in step_culprit.get('tests', {}).values():
         test_revision = test_culprit['revision']
         test_culprit.update(culprits[test_revision])
 
   def _GetCulpritDataForTest(self, culprit_map):
     """Gets culprit revision for each failure for try job metadata."""
     culprit_data = {}
     for step, step_culprit in culprit_map.iteritems():
-      if step_culprit['tests']:
-        culprit_data[step] = {}
-        for test, test_culprit in step_culprit['tests'].iteritems():
-          culprit_data[step][test] = test_culprit['revision']
-      else:
-        culprit_data[step] = step_culprit['revision']
+      culprit_data[step] = {}
+      for test, test_culprit in step_culprit['tests'].iteritems():
+        culprit_data[step][test] = test_culprit['revision']
     return culprit_data
 
   # Arguments number differs from overridden method - pylint: disable=W0221
   def run(
       self, master_name, builder_name, build_number, blame_list, try_job_type,
       try_job_id, result):
     """Identifies the information for failed revisions.
 
     Please refer to try_job_result_format.md for format check.
     """
@@ skipping 31 matching lines @@
             try_job_result.test_results)
         if (result_to_update and
             result_to_update[-1]['try_job_id'] == try_job_id):
           result_to_update[-1].update(result)
         else:  # pragma: no cover
           result_to_update.append(result)
       try_job_result.status = analysis_status.COMPLETED
       try_job_result.put()
 
     @ndb.transactional
-    def UpdateWfAnalysisWithTryJobResult():
+    def UpdateWfAnalysisWithTryJobResult(try_job_suspected_cls):
       if not culprits:
         return
-
       # Update analysis result and suspected CLs with results of this try job if
       # culprits were found.
       analysis = WfAnalysis.Get(master_name, builder_name, build_number)
       updated_result_status = _GetResultAnalysisStatus(analysis, result)
-      updated_suspected_cls = _GetSuspectedCLs(analysis, result)
-
+      updated_suspected_cls = _GetSuspectedCLs(analysis, try_job_suspected_cls)
       if (analysis.result_status != updated_result_status or
           analysis.suspected_cls != updated_suspected_cls):
         analysis.result_status = updated_result_status
         analysis.suspected_cls = updated_suspected_cls
         analysis.put()
 
+    def UpdateSuspectedCLs(try_job_suspected_cls):
+      if not culprits:
+        return
+
+      # Creates or updates each suspected_cl.
+      for culprit in try_job_suspected_cls:
+        suspected_cl_util.UpdateSuspectedCL(
+            analysis_approach_type.TRY_JOB, master_name, builder_name,
+            build_number, try_job_type,
+            culprit['repo_name'], culprit['revision'],
+            culprit.get('commit_position'))
+
+    try_job_suspected_cls = _GetTryJobSuspectedCLs(result)
     # Store try-job results.
     UpdateTryJobResult()
     # Add try-job results to WfAnalysis.
-    UpdateWfAnalysisWithTryJobResult()
+    UpdateWfAnalysisWithTryJobResult(try_job_suspected_cls)
+    # Updates suspected_cl

[lijeffrey, 2016/08/17 19:31:11]

nit: comment ends with .

[chanli, 2016/09/12 20:56:15]

Done.

+    UpdateSuspectedCLs(try_job_suspected_cls)
 
     _NotifyCulprits(master_name, builder_name, build_number, culprits)
-    return result.get('culprit') if result else None
+    return result.get('culprit') if result else None