rebaseline.py: try to download images from Google Storage before skia-autogen

tools/rebaseline.py

 #!/usr/bin/python
 
 '''
 Copyright 2012 Google Inc.
 
 Use of this source code is governed by a BSD-style license that can be
 found in the LICENSE file.
 '''
 
 '''
@@ skipping 131 matching lines @@
     # for example, "file:one/two" resolves to the file ./one/two
     # (relative to current working dir)
     def _GetContentsOfUrl(self, url):
         file_prefix = 'file:'
         if url.startswith(file_prefix):
             filename = url[len(file_prefix):]
             return open(filename, 'r').read()
         else:
             return urllib2.urlopen(url).read()
 
+    # Returns a dictionary of actual results from actual-results.json file.
+    #
+    # The dictionary returned has this format:
+    # {
+    #  u'imageblur_565.png': [u'bitmap-64bitMD5', 3359963596899141322],
+    #  u'imageblur_8888.png': [u'bitmap-64bitMD5', 4217923806027861152],
+    #  u'shadertext3_8888.png': [u'bitmap-64bitMD5', 3713708307125704716]
+    # }
+    #
+    # If the JSON actual result summary file cannot be loaded, the behavior
+    # depends on self._missing_json_is_fatal:
+    # - if true: execution will halt with an exception
+    # - if false: we will log an error message but return an empty dictionary
+    #
+    # params:
+    #  json_url: URL pointing to a JSON actual result summary file
+    #  sections: a list of section names to include in the results, e.g.
+    #            [gm_json.JSONKEY_ACTUALRESULTS_FAILED,
+    #             gm_json.JSONKEY_ACTUALRESULTS_NOCOMPARISON] ;
+    #            if None, then include ALL sections.
+    def _GetActualResults(self, json_url, sections=None):

[epoger, 2013/06/12 15:58:46]

Patchset 3 extracts a new _GetActualResults() method out of
_GetFilesToRebaseline().  This will help us look up actual-result checksums from
the actual-results.json file later on in the CL.

+        try:
+            json_contents = self._GetContentsOfUrl(json_url)
+        except urllib2.HTTPError:
+            message = 'unable to load JSON summary URL %s' % json_url
+            if self._missing_json_is_fatal:
+                raise ValueError(message)
+            else:
+                print '# %s' % message
+                return {}
+
+        json_dict = gm_json.LoadFromString(json_contents)
+        results_to_return = {}
+        actual_results = json_dict[gm_json.JSONKEY_ACTUALRESULTS]
+        if not sections:
+            sections = actual_results.keys()
+        for section in sections:
+            section_results = actual_results[section]
+            if section_results:
+                results_to_return.update(section_results)
+        return results_to_return
+
     # Returns a list of files that require rebaselining.
     #
     # Note that this returns a list of FILES, like this:
     #  ['imageblur_565.png', 'xfermodes_pdf.png']
     # rather than a list of TESTS, like this:
     #  ['imageblur', 'xfermodes']
     #
-    # If the JSON actual result summary file cannot be loaded, the behavior
-    # depends on self._missing_json_is_fatal:
-    # - if true: execution will halt with an exception
-    # - if false: we will log an error message but return an empty list so we
-    #   go on to the next platform
-    #
     # params:
     #  json_url: URL pointing to a JSON actual result summary file
     #  add_new: if True, then return files listed in any of these sections:
     #            - JSONKEY_ACTUALRESULTS_FAILED
     #            - JSONKEY_ACTUALRESULTS_NOCOMPARISON
     #           if False, then return files listed in these sections:
     #            - JSONKEY_ACTUALRESULTS_FAILED
     #
     def _GetFilesToRebaseline(self, json_url, add_new):
         if self._dry_run:
             print ''
             print '#'
         print ('# Getting files to rebaseline from JSON summary URL %s ...'
                % json_url)
-        try:
-            json_contents = self._GetContentsOfUrl(json_url)
-        except urllib2.HTTPError:
-            message = 'unable to load JSON summary URL %s' % json_url
-            if self._missing_json_is_fatal:
-                raise ValueError(message)
-            else:
-                print '# %s' % message
-                return []
-
-        json_dict = gm_json.LoadFromString(json_contents)
-        actual_results = json_dict[gm_json.JSONKEY_ACTUALRESULTS]
         sections = [gm_json.JSONKEY_ACTUALRESULTS_FAILED]
         if add_new:
             sections.append(gm_json.JSONKEY_ACTUALRESULTS_NOCOMPARISON)
-
-        files_to_rebaseline = []
-        for section in sections:
-            section_results = actual_results[section]
-            if section_results:
-                files_to_rebaseline.extend(section_results.keys())
-
+        results_to_rebaseline = self._GetActualResults(json_url=json_url,
+                                                       sections=sections)
+        files_to_rebaseline = results_to_rebaseline.keys()
+        files_to_rebaseline.sort()
         print '# ... found files_to_rebaseline %s' % files_to_rebaseline
         if self._dry_run:
             print '#'
         return files_to_rebaseline
 
     # Rebaseline a single file.
     def _RebaselineOneFile(self, expectations_subdir, builder_name,
                            infilename, outfilename):
         if self._dry_run:
             print ''
@@ skipping 59 matching lines @@
                                     outfilename=outfilename)
 
     # Rebaseline all platforms/tests/types we specified in the constructor.
     def RebaselineAll(self):
         for subdir in self._subdirs:
             if not subdir in SUBDIR_MAPPING.keys():
                 raise Exception(('unrecognized platform subdir "%s"; ' +
                                  'should be one of %s') % (
                                      subdir, SUBDIR_MAPPING.keys()))
             builder_name = SUBDIR_MAPPING[subdir]
+            json_url = '/'.join([self._json_base_url,
+                                 subdir, builder_name, subdir,
+                                 self._json_filename])
+
             if self._tests:
                 for test in self._tests:
                     self._RebaselineOneTest(expectations_subdir=subdir,
                                             builder_name=builder_name,
                                             test=test)
             else:  # get the raw list of files that need rebaselining from JSON
-                json_url = '/'.join([self._json_base_url,
-                                     subdir, builder_name, subdir,
-                                     self._json_filename])
                 filenames = self._GetFilesToRebaseline(json_url=json_url,
                                                        add_new=self._add_new)
                 for filename in filenames:
                     outfilename = os.path.join(subdir, filename);
                     self._RebaselineOneFile(expectations_subdir=subdir,
                                             builder_name=builder_name,
                                             infilename=filename,
                                             outfilename=outfilename)
 
 # main...
@@ skipping 30 matching lines @@
                     '"--tests aaclip bigmatrix"; if unspecified, then all ' +
                     'failing tests (according to the actual-results.json ' +
                     'file) will be rebaselined.')
 args = parser.parse_args()
 rebaseliner = Rebaseliner(tests=args.tests, configs=args.configs,
                           subdirs=args.subdirs, dry_run=args.dry_run,
                           json_base_url=args.json_base_url,
                           json_filename=args.json_filename,
                           add_new=args.add_new)
 rebaseliner.RebaselineAll()