[Findit] Enable project classifier and component classifier

appengine/findit/crash/project_classifier.py

Index: appengine/findit/crash/project_classifier.py
diff --git a/appengine/findit/crash/project_classifier.py b/appengine/findit/crash/project_classifier.py
new file mode 100644
index 0000000000000000000000000000000000000000..84f156dd6d7a419671639d63afbeaae75c74e943
--- /dev/null
+++ b/appengine/findit/crash/project_classifier.py
@@ -0,0 +1,105 @@
+# Copyright 2016 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.
+
+import copy
+
+from crash import classifier
+from crash.type_enums import CallStackLanguageType
+from model.crash.crash_config import CrashConfig
+
+
+class ProjectClassifier(classifier.Classifier):
+  """Determines the project of a crash - (project_name, project_path).
+
+  For example: ('chromium', 'src/'), ('skia', 'src/skia/'), ...etc.
+  """
+
+  def __init__(self):
+    super(ProjectClassifier, self).__init__()
+
+    def _GetCompiledConfig(config):
+      config = copy.copy(config)
+      config['top_n'] = int(config['top_n'])
+      return config
+
+    self.config = _GetCompiledConfig(CrashConfig.Get().project_classifier)
+
+  def _GetProjectFromDepPath(self, dep_path):
+    """Returns the component name of a path."""

[Martin Barbella, 2016/05/13 23:05:07]

Update this (if needed once other comments are addressed).

[Sharu Jiang, 2016/05/17 19:27:54]

Done.

+    if not dep_path:
+      return '', ''

[Martin Barbella, 2016/05/13 23:05:07]

I'm a bit confused by why we need to return dep_path here if the callers clearly
already know it. Is this necessary?

[Sharu Jiang, 2016/05/17 19:27:54]

I just want to offer more information before, but right, it's not that
necessary, deleted the project path.

+
+    if dep_path == 'src/':
+      return 'chromium', dep_path
+
+    for host_directory in self.config['host_directories']:
+      if dep_path.startswith(host_directory):
+        path = dep_path[len(host_directory):]
+        return 'chromium-%s' % path.split('/')[0].lower(), dep_path
+
+    # Unknown path, return the whole path as component name.
+    return 'chromium-%s' % '_'.join(dep_path.split('/')), dep_path
+
+  def GetClassFromStackFrame(self, frame):
+    """Returns a tuple (project_name, project_path) of a StackFrame."""
+    for marker, name in self.config[
+        'function_marker_to_project_name'].iteritems():
+      if frame.function.startswith(marker):
+        return name, ''
+
+    for marker, name in self.config[
+        'file_path_marker_to_project_name'].iteritems():
+      if marker in frame.file_path or marker in frame.raw_file_path:
+        return name, ''
+
+    return self._GetProjectFromDepPath(frame.dep_path)
+
+  def GetClassFromResult(self, result):
+    """Returns (project_name, project_path) of a Result."""
+    if result.file_to_stack_infos:
+      # A file in culprit result should always have its stack_info, namely a
+      # list of (frame, callstack_priority) pairs.
+      frame, _ = result.file_to_stack_infos.values()[0][0]
+      return self.GetClassFromStackFrame(frame)
+
+    return '', ''
+
+  def Classify(self, results, crash_stack):
+    """Classify project of a crash.
+
+    Args:
+      results (list of Result): culprit results.
+      crash_stack (CallStack): the callstack that caused the crash.
+
+    Returns:
+      A tuple, project of the crash - (project_name, project_path).
+    """
+    def _GetRankFunction(language_type):
+      if language_type == CallStackLanguageType.JAVA:
+        def _RankFunctionForJava(class_occurrences_info):
+          project_name, _ = class_occurrences_info.name
+          return (len(class_occurrences_info.occurrences),
+                  0 if 'chromium' in project_name else self.config[
+                      'non_chromium_project_rank_priority'][project_name])
+
+        return _RankFunctionForJava
+
+      return classifier.DefaultRankFunction
+
+    # Set the max_classes to 1, so the returned projects only has one element.
+    projects = self._Classify(
+        results, crash_stack,
+        self.config['top_n'], 1,
+        rank_function=_GetRankFunction(crash_stack.language_type))
+
+    def _Format(project_name, project_path):
+      if project_path:
+        return '%s (%s)' % (project_name, project_path)
+
+      return project_name
+
+    if projects:
+      return _Format(*projects[0])
+
+    return ''