[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..7ae04a8d5a5386f0d5e0cb7caf0ac03312596904
--- /dev/null
+++ b/appengine/findit/crash/project_classifier.py
@@ -0,0 +1,98 @@
+# 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.
+
+from crash import config
+from crash.classifier import Classifier
+from crash.type_enums import CallStackLanguageType
+from model.crash.crash_config import CrashConfig
+
+
+class ProjectClassifier(Classifier):
+  """Determines the project of a crash - (project_name, project_path).
+
+  For example: ('chromium', 'src/'), ('skia', 'src/skia/'), ...etc.
+  """
+
+  CONFIG = CrashConfig.Get().project_classifier
+
+  def _GetProjectNameFromDepPath(self, dep_path):
+    """Return the component name of a path."""
+    if dep_path == 'src/':
+      return 'chromium'
+
+    for host_directory in ProjectClassifier.CONFIG['host_directories']:
+      if dep_path.startswith(host_directory):
+        path = dep_path[len(host_directory):]
+        return 'chromium-%s' % path.split('/')[0].lower()
+
+  # Unknown path, return the whole path as component name.
+  return 'chromium-%s' % '_'.join(dep_path.split('/'))
+
+  def GetClassFromStackFrame(self, frame):
+    """Returns a tuple (project_name, project_path) of a StackFrame."""
+    if frame.dep_path:
+      # dep paths are parsed from chromium dependency, if a frame has non-empty

[Martin Barbella, 2016/05/05 18:33:41]

Nit: Dep

[Sharu Jiang, 2016/05/10 03:11:02]

Done.

+      # dep path, the project of this frame is chromium-*.
+      return self._GetProjectNameFromDepPath(frame.dep_path), frame.dep_path
+
+    if not ProjectClassifier.CONFIG:
+      return '', ''
+
+    for marker, name in ProjectClassifier.CONFIG[
+        'function_marker_to_project_name'].iteritems():
+      if frame.function.startswith(marker):
+        return name, ''
+
+    for marker, name in ProjectClassifier.CONFIG[
+        'file_path_to_project_name'].iteritems():
+      if marker in frame.file_path or marker in frame.raw_file_path:
+        return name, ''
+
+    return '', ''
+
+  def GetClassFromResult(self, result):
+    """Returns (project_name, project_path) of a Result."""
+    if result.dep_path:
+      # dep paths are parsed from chromium dependency, if a result has non-empty

[Martin Barbella, 2016/05/05 18:33:41]

Ditto.

[Sharu Jiang, 2016/05/10 03:11:03]

Done.

+      # dep path, the project of this culprit result is chromium-*.
+      return self._GetProjectNameFromDepPath(result.dep_path), result.dep_path
+
+    for frame, _ in result.file_path_to_project_name.itervalues():
+      return self.GetProjectFromStackFrame(frame)
+
+  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():
+      if (crash_stack.language_type == CallStackLanguageType.JAVA and
+          ProjectClassifier.CONFIG):
+        def _RankFunctionForJava(class_info):
+          project_name, _ = class_info.class_value
+          return (class_info.count,
+                  0 if 'chromium' in project_name else ProjectClassifier.CONFIG[
+                      'non_chromium_project_rank_priority'][project_name])
+
+        return _RankFunctionForJava
+
+      return self._DefaultRankFunction
+
+    # Set the max_classes to 1, so the returned projects only has one element.
+    projects = self._Classify(results, crash_stack,
+                              ProjectClassifier.CONFIG['top_n'], 1,
+                              rank_function=_GetRankFunction())
+
+    def _Format(project_name, project_path):
+      return '%s (%s)' % project_name, project_path
+
+    if projects:
+      return _Format(*projects[0])
+
+    return ''