[Findit] Fixed a bug where blame information is not correctly filtered

tools/findit/crash_utils.py

 # Copyright (c) 2014 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 cgi
 import ConfigParser
 import json
 import logging
 import os
 import time
 import urllib2
 
 from result import Result
 
 
 INFINITY = float('inf')
 
 
 def ParseURLsFromConfig(file_name):
   """Parses URLS from the config file.
 
   The file should be in python config format, where svn section is in the
-  format "svn:component_path", except for git URLs and codereview URL.
+  format "svn:component_path".
   Each of the section for svn should contain changelog_url, revision_url,
   diff_url and blame_url.
 
   Args:
     file_name: The name of the file that contains URL information.
 
   Returns:
     A dictionary that maps repository type to list of URLs. For svn, it maps
     key 'svn' to another dictionary, which maps component path to the URLs
     as explained above. For git, it maps to the URLs as explained above.
-    Codereview maps to codereview API url.
   """
   config = ConfigParser.ConfigParser()
 
   # Get the absolute path of the config file, and read the file. If it fails,
   # return none.
   config_file_path = os.path.join(os.path.abspath(os.path.dirname(__file__)),
                                   file_name)
   config.read(config_file_path)
   if not config:
     logging.error('Config file with URLs does not exist.')
     return None
 
   # Iterate through the config file, check for sections.
-  repository_type_to_url_map = {}
+  parsed_config = {}
   for section in config.sections():
     # These two do not need another layer of dictionary, so add it and go
     # to next section.
-    if section == 'git' or section == 'codereview':
+    if ':' not in section:
       for option in config.options(section):
-        if section not in repository_type_to_url_map:
-          repository_type_to_url_map[section] = {}
+        if section not in parsed_config:
+          parsed_config[section] = {}
 
         url = config.get(section, option)
-        repository_type_to_url_map[section][option] = url
+        parsed_config[section][option] = url
 
       continue
 
     # Get repository type and component name from the section name.
     repository_type_and_component = section.split(':')
     repository_type = repository_type_and_component[0]
     component_path = repository_type_and_component[1]
 
     # Add 'svn' as the key, if it is not already there.
-    if repository_type not in repository_type_to_url_map:
-      repository_type_to_url_map[repository_type] = {}
-    url_map_for_repository = repository_type_to_url_map[repository_type]
+    if repository_type not in parsed_config:
+      parsed_config[repository_type] = {}
+    url_map_for_repository = parsed_config[repository_type]
 
     # Add the path to the 'svn', if it is not already there.
     if component_path not in url_map_for_repository:
       url_map_for_repository[component_path] = {}
     type_to_url = url_map_for_repository[component_path]
 
     # Add all URLs to this map.
     for option in config.options(section):
       url = config.get(section, option)
       type_to_url[option] = url
 
-  return repository_type_to_url_map
+  return parsed_config
 
 
 def NormalizePathLinux(path, parsed_deps):
   """Normalizes linux path.
 
   Args:
     path: A string representing a path.
     parsed_deps: A map from component path to its component name, repository,
                  etc.
 
@@ skipping 23 matching lines @@
 
       # Normalize the path by stripping everything off the component's relative
       # path.
       normalized_path = normalized_path.split(new_path,1)[1]
 
       # Add 'src/' or 'Source/' at the front of the normalized path, depending
       # on what prefix the component path uses. For example, blink uses
       # 'Source' but chromium uses 'src/', and blink component path is
       # 'src/third_party/WebKit/Source', so add 'Source/' in front of the
       # normalized path.
-      if not normalized_path.startswith('src/') or \
-          normalized_path.startswith('Source/'):
+      if not (normalized_path.startswith('src/') or \

[stgao, 2014/08/15 19:42:38]

Ending \ could be removed.

+          normalized_path.startswith('Source/')):
 
         if (new_path.lower().endswith('src/') or
             new_path.lower().endswith('source/')):
           normalized_path = new_path.split('/')[-2] + '/' + normalized_path
 
         else:
           normalized_path = 'src/' + normalized_path
 
       component_name = parsed_deps[component_path]['name']
 
@@ skipping 39 matching lines @@
     JSON object if the string represents a JSON object, None otherwise.
   """
   try:
     data = json.loads(json_string)
   except ValueError:
     data = None
 
   return data
 
 
-def GetDataFromURL(url, retries=10, sleep_time=0.1, timeout=10):
+def GetDataFromURL(url, retries=10, sleep_time=0.1, timeout=5):
   """Retrieves raw data from URL, tries 10 times.
 
   Args:
     url: URL to get data from.
     retries: Number of times to retry connection.
     sleep_time: Time in seconds to wait before retrying connection.
     timeout: Time in seconds to wait before time out.
 
   Returns:
     None if the data retrieval fails, or the raw data.
@@ skipping 235 matching lines @@
     suspected_cl = blame.revision
     revision_url = blame.url
     component_name = blame.component_name
     author = blame.author
     reason = (
         'The CL changes line %s of file %s from stack %d.' %
         (blame.line_number, blame.file, blame.stack_frame_index))
     # Blame object does not have review url and reviewers.
     review_url = None
     reviewers = None
-    line_content = blame.content
+    line_content = blame.line_content
 
     result = Result(suspected_cl, revision_url, component_name, author, reason,
                     review_url, reviewers, line_content)
     result_list.append(result)
 
   return result_list
-
-
-def ResultListToJSON(result_list):
-  """Converts result list to JSON format.
-
-  Args:
-    result_list: A list of result objects
-
-  Returns:
-    A string, JSON format of the result_list.
-
-  """
-  return json.dumps([result.ToDictionary() for result in result_list])