Add presubmit check for copyright notice

PRESUBMIT.py

 # Copyright (c) 2013 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.
 
 
 """Top-level presubmit script for Skia.
 
 See http://dev.chromium.org/developers/how-tos/depottools/presubmit-scripts
 for more details about the presubmit API built into gcl.
 """
@@ skipping 100 matching lines @@
   results = []
   if failing_files:
     results.append(
         output_api.PresubmitError(
             'The following files have #if or #ifdef before includes:\n%s\n\n'
             'See skbug.com/3362 for why this should be fixed.' %
                 '\n'.join(failing_files)))
   return results
 
 
+def _CopyrightChecks(input_api, output_api, source_file_filter=None):
+  results = []
+  year_pattern = r'\d{4}'
+  year_range_pattern = r'%s(-%s)?' % (year_pattern, year_pattern)
+  years_pattern = r'%s(,%s)*,?' % (year_range_pattern, year_range_pattern)
+  copyright_pattern = (
+      r'Copyright (\([cC]\) )?%s %s' % (years_pattern, '%s'))
+  owners = [
+    'ARM Ltd.',
+    'Code Aurora Forum. All rights reserved.',
+    'Google,? Inc\.?( All rights reserved.)?',
+    'Intel Inc.',
+    'Lua.org, PUC-Rio',
+    'Skia',
+    'The Android Open Source Project',
+    'The Chromium Authors. All rights reserved.',
+    'The Native Client Authors. All rights reserved.',

[borenet, 2015/03/23 14:38:05]

I just pulled these from existing copyright notices.  Are these all valid? 
Should I just check that there's a copyright notice at all, rather than having
an owner whitelist?  For that matter, there are a few files which have no
copyright owner, ie. "Copyright 2014".  Is that valid?

[mtklein, 2015/03/23 14:49:50]

Let's both exclude third_party ('Lua.org, PUC-Rio') and allow any non-empty
copyright holder ([^\w]+) ?

[borenet, 2015/03/23 17:20:35]

Done.

+  ]
+
+  for affected_file in input_api.AffectedSourceFiles(source_file_filter):
+    contents = input_api.ReadFile(affected_file, 'rb')
+    found = False
+    for owner in owners:
+      if re.search(copyright_pattern % owner, contents):
+        found = True
+        break
+    if not found:
+      results.append(output_api.PresubmitError(
+          '%s is missing a correct copyright header.' % affected_file))
+  return results
+
+
 def _CommonChecks(input_api, output_api):
   """Presubmit checks common to upload and commit."""
   results = []
   sources = lambda x: (x.LocalPath().endswith('.h') or
                        x.LocalPath().endswith('.gypi') or
                        x.LocalPath().endswith('.gyp') or
                        x.LocalPath().endswith('.py') or
                        x.LocalPath().endswith('.sh') or
                        x.LocalPath().endswith('.cpp'))
   results.extend(
       _CheckChangeHasEol(
           input_api, output_api, source_file_filter=sources))
   results.extend(_PythonChecks(input_api, output_api))
   results.extend(_IfDefChecks(input_api, output_api))
+  results.extend(_CopyrightChecks(input_api, output_api,
+                                  source_file_filter=sources))
   return results
 
 
 def CheckChangeOnUpload(input_api, output_api):
   """Presubmit checks for the change on upload.
 
   The following are the presubmit checks:
   * Check change has one and only one EOL.
   """
   results = []
@@ skipping 235 matching lines @@
     state and an error if it is in 'Closed' state.
   """
   results = []
   results.extend(_CommonChecks(input_api, output_api))
   results.extend(
       _CheckTreeStatus(input_api, output_api, json_url=(
           SKIA_TREE_STATUS_URL + '/banner-status?format=json')))
   results.extend(_CheckLGTMsForPublicAPI(input_api, output_api))
   results.extend(_CheckOwnerIsInAuthorsFile(input_api, output_api))
   return results