[Findit] Fix blame for GIT and uptake bug fix.

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 os
+import Queue
+import threading
 import time
 
 from common import utils
 from result import Result
 
 
 INFINITY = float('inf')
 
+THREAD_NUMBER = 5

[aarya, 2014/08/28 01:55:12]

s/THREAD_NUMBER/MAX_THREADS or something like that.

[stgao, 2014/08/28 04:24:38]

Done.

+TASK_QUEUE = None
+
+
+def Worker():
+  global TASK_QUEUE
+  while True:
+    task_id, function, args, kwargs, result_semaphore = TASK_QUEUE.get()
+    print threading.current_thread().name, ' is running task ', task_id

[aarya, 2014/08/28 01:55:12]

Why print ? Won't this show up in findit results which we don't want.

[stgao, 2014/08/28 04:24:37]

Removed.
Sorry. That's for debugging.

+    try:
+      function(*args, **kwargs)
+    except:
+      pass
+    finally:
+      # Signal one task is done in case of exception.
+      result_semaphore.release()
+
+
+def RunTasks(tasks):
+  """Run given tasks. Not thread-safe: no concurrent calls of this function.
+
+  Return after all tasks were completed. A task is a dict as below:
+    {
+      'function': the function to call,
+      'args': the positional argument to pass to the function,
+      'kwargs': the key-value arguments to pass to the function,
+    }
+  """
+  if not tasks:
+    return
+
+  global TASK_QUEUE
+  if not TASK_QUEUE:
+    TASK_QUEUE = Queue.Queue()
+    for index in range(THREAD_NUMBER):
+      thread = threading.Thread(target=Worker, name='worker_%s' % index)
+      # Set as daemon, so no join is needed.
+      thread.daemon = True
+      thread.start()
+      print thread.name, 'is started'

[aarya, 2014/08/28 01:55:12]

probably need to remove print to prevent from not showing up in output.

[stgao, 2014/08/28 04:24:38]

Removed.

+
+  result_semaphore = threading.Semaphore(0)
+
+  # Push task to task queue for execution.
+  task_id = 0
+  for task in tasks:
+    task_id += 1

[aarya, 2014/08/28 01:55:12]

task_id is not needed. even in debugging, you can just print task['function'],
etc.

+    TASK_QUEUE.put(
+        (task_id, task['function'], task.get('args', []),
+         task.get('kwargs', {}), result_semaphore))
+
+  # Wait until all tasks to be executed.
+  for task in tasks:

[aarya, 2014/08/28 01:55:12]

probably for _ in tasks: 

[stgao, 2014/08/28 04:24:37]

Done.

+    result_semaphore.acquire()
 
 def GetRepositoryType(revision_number):
   """Returns the repository type of this revision number.
 
   Args:
     revision_number: A revision number or git hash.
 
   Returns:
     'git' or 'svn', depending on the revision_number.
   """
@@ skipping 197 matching lines @@
 
   Returns:
     None if the data retrieval fails, or the raw data.
   """
   count = 0
   while True:
     count += 1
     # Retrieves data from URL.
     try:
       status_code, data = utils.GetHttpClient().Get(url, timeout=timeout)
-    except IOError:
+    except IOError as e:
       status_code = -1
       data = None
 
     if status_code == 200:
       return data
 
     if count < retries:
       # If retrieval fails, try after sleep_time second.
       time.sleep(sleep_time)
     else:
@@ skipping 118 matching lines @@
   Returns:
     A string representation of the list.
   """
   return ', '.join(map(str, items))
 
 
 def PrettifyFrameInfo(frame_indices, functions):
   """Return a string to represent the frames with functions."""
   frames = []
   for frame_index, function in zip(frame_indices, functions):
-    frames.append('frame #%s, function "%s"' % (frame_index, function))
+    frames.append('frame #%s, "%s"' % (frame_index, function.split('(')[0]))
   return '; '.join(frames)
 
 
 def PrettifyFiles(file_list):
   """Returns a string representation of a list of file names.
 
   Args:
     file_list: A list of tuple, (file_name, file_url).
   Returns:
     A string representation of file names with their urls.
@@ skipping 118 matching lines @@
     review_url = None
     reviewers = None
     line_content = blame.line_content
     message = blame.message
 
     result = Result(suspected_cl, revision_url, component_name, author, reason,
                     review_url, reviewers, line_content, message)
     result_list.append(result)
 
   return result_list