[Findit] Bug fixing and implemented some feature requests.

tools/findit/svn_repository_parser.py

-# Copyright 2014 The Chromium Authors. All rights reserved.
+# 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 logging
-import os
 import xml.dom.minidom as minidom
 from xml.parsers.expat import ExpatError
 
 import crash_utils
 from repository_parser_interface import ParserInterface
 
 
 # This number is 6 because each linediff page in src.chromium.org should
 # contain the following tables: table with revision number, table with actual
 # diff, table with dropdown menu, table with legend, a border table and a table
@@ skipping 54 matching lines @@
       revision_object['author'] = revision.getElementsByTagName(
           'author')[0].firstChild.nodeValue
 
       # Get the revision number from xml.
       revision_number = int(revision.getAttribute('revision'))
 
       # Iterate through the changed paths in the CL.
       paths = revision.getElementsByTagName('paths')
       if paths:
         for changed_path in paths[0].getElementsByTagName('path'):
           # Get path, file action and file name from the xml.

[stgao, 2014/08/22 06:50:54]

comment is not updated.

[jeun, 2014/08/22 22:58:44]

Done.

           file_path = changed_path.firstChild.nodeValue
-          file_action = changed_path.getAttribute('action')
-          changed_file = os.path.basename(file_path)
+          file_change_type = changed_path.getAttribute('action')
+
+          if file_path.startswith('/trunk/'):
+            file_path = file_path[len('/trunk/'):]
 
           # Add file to the map.
-          if changed_file not in file_to_revision_map:
-            file_to_revision_map[changed_file] = []
-          file_to_revision_map[changed_file].append(
-              (revision_number, file_action, file_path))
+          if file_path not in file_to_revision_map:
+            file_to_revision_map[file_path] = []
+          file_to_revision_map[file_path].append(
+              (revision_number, file_change_type))
 
       # Set commit message of the CL.
       revision_object['message'] = revision.getElementsByTagName('msg')[
           0].firstChild.nodeValue
 
       # Set url of this CL.
       revision_url = url_map['revision_url'] % revision_number
       revision_object['url'] = revision_url
 
       # Add this CL to the revision map.
       revision_map[revision_number] = revision_object
 
     return (revision_map, file_to_revision_map)
 
-  def ParseLineDiff(self, path, component, file_action, revision_number):
+  def ParseLineDiff(self, path, component, file_change_type, revision_number):
     changed_line_numbers = []
     changed_line_contents = []
 
     url_map = self.component_to_urls_map.get(component)
     if not url_map:
       logging.error('Component %s is not currently supported.', component)
       return (None, None, None)
 
     # If the file is added (not modified), treat it as if it is not changed.
     backup_url = url_map['revision_url'] % revision_number
-    if file_action == 'A':
+    if file_change_type == 'A':
       return (backup_url, changed_line_numbers, changed_line_contents)
 
     # Retrieve data from the url. If no data is retrieved, return empty lists.
     url = url_map['diff_url'] % (path, revision_number - 1,
                                  revision_number, revision_number)
     data = crash_utils.GetDataFromURL(url)
     if not data:
       logging.error('Failed to get line changes from %s.', url)
       return (backup_url, changed_line_numbers, changed_line_contents)
 
@@ skipping 85 matching lines @@
     blame_html = minidom.parseString(data)
 
     title = blame_html.getElementsByTagName('title')
     # If the returned html page is an exception page, return None.
     if title[0].firstChild.nodeValue == 'ViewVC Exception':
       logging.error('Failed to retrieve blame information from %s.', url)
       return None
 
     # Each of the blame result is in <tr>.
     blame_results = blame_html.getElementsByTagName('tr')
-    blame_result = blame_results[line]
+    try:
+      blame_result = blame_results[line]
+    except IndexError:
+      logging.error('Failed to retrieve blame information from %s.', url)
+      return None
 
     # There must be 4 <td> for each <tr>. If not, this page is wrong.
     tds = blame_result.getElementsByTagName('td')
     if len(tds) != 4:
       logging.error('Failed to retrieve blame information from %s.', url)
       return None
 
     # The third <td> has the line content, separated by <span>s. Combine
     # those to get a string of changed line. If it has nothing, the line
     # is empty.
@@ skipping 17 matching lines @@
       blame_result = blame_results[line]
       tds = blame_result.getElementsByTagName('td')
     author = tds[1].firstChild.nodeValue
 
     # Revision can either be in hyperlink or plain text.
     try:
       revision = tds[2].getElementsByTagName('a')[0].firstChild.nodeValue
     except IndexError:
       revision = tds[2].firstChild.nodeValue
 
+    (revision_info, _) = self.ParseChangelog(component, revision, revision)
+    message = revision_info[int(revision)]['message']
+
     # Return the parsed information.
     revision_url = url_map['revision_url'] % int(revision)
-    return (line_content, revision, author, revision_url)
+    return (line_content, revision, author, revision_url, message)