webkitpy: Reworking test chunking for sharding.

third_party/WebKit/Tools/Scripts/webkitpy/layout_tests/controllers/manager.py

 # Copyright (C) 2010 Google Inc. All rights reserved.
 # Copyright (C) 2010 Gabor Rapcsanyi (rgabor@inf.u-szeged.hu), University of Szeged
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are
 # met:
 #
 #     * Redistributions of source code must retain the above copyright
 # notice, this list of conditions and the following disclaimer.
 #     * Redistributions in binary form must reproduce the above
@@ skipping 78 matching lines @@
         self._results_directory = self._port.results_directory()
         self._finder = LayoutTestFinder(self._port, self._options)
         self._runner = LayoutTestRunner(self._options, self._port, self._printer, self._results_directory, self._test_is_slow)
 
     def run(self, args):
         """Run the tests and return a RunDetails object with the results."""
         start_time = time.time()
         self._printer.write_update("Collecting tests ...")
         running_all_tests = False
         try:
-            paths, test_names, running_all_tests = self._collect_tests(args)
+            paths, all_test_names, running_all_tests = self._collect_tests(args)
         except IOError:
             # This is raised if --test-list doesn't exist
             return test_run_results.RunDetails(exit_code=test_run_results.NO_TESTS_EXIT_STATUS)
 
+        # Create a sorted list of test files so the subset chunk,
+        # if used, contains alphabetically consecutive tests.
+        if self._options.order == 'natural':
+            all_test_names.sort(key=self._port.test_key)
+        elif self._options.order == 'random':
+            all_test_names.sort()
+            random.Random(self._options.seed).shuffle(all_test_names)
+
+        test_names, tests_in_other_chunks = self._finder.split_into_chunks(all_test_names)
+
         self._printer.write_update("Parsing expectations ...")
         self._expectations = test_expectations.TestExpectations(self._port, test_names)
 
         tests_to_run, tests_to_skip = self._prepare_lists(paths, test_names)
-        self._printer.print_found(len(test_names), len(tests_to_run), self._options.repeat_each, self._options.iterations)
+
+        self._expectations.remove_tests(tests_in_other_chunks)
+
+        self._printer.print_found(
+            len(all_test_names), len(test_names), len(tests_to_run),
+            self._options.repeat_each, self._options.iterations)
 
         # Check to make sure we're not skipping every test.
         if not tests_to_run:
             _log.critical('No tests to run.')
             return test_run_results.RunDetails(exit_code=test_run_results.NO_TESTS_EXIT_STATUS)
 
         exit_code = self._set_up_run(tests_to_run)
         if exit_code:
             return test_run_results.RunDetails(exit_code=exit_code)
 
@@ skipping 116 matching lines @@
     def _http_tests(self, test_names):
         return set(test for test in test_names if self._is_http_test(test))
 
     def _is_perf_test(self, test):
         return self.PERF_SUBDIR == test or (self.PERF_SUBDIR + self._port.TEST_PATH_SEPARATOR) in test
 
     def _prepare_lists(self, paths, test_names):
         tests_to_skip = self._finder.skip_tests(paths, test_names, self._expectations, self._http_tests(test_names))
         tests_to_run = [test for test in test_names if test not in tests_to_skip]
 
-        if not tests_to_run:
-            return tests_to_run, tests_to_skip
-
-        # Create a sorted list of test files so the subset chunk,
-        # if used, contains alphabetically consecutive tests.
-        if self._options.order == 'natural':
-            tests_to_run.sort(key=self._port.test_key)
-        elif self._options.order == 'random':
-            tests_to_run.sort()
-            random.Random(self._options.seed).shuffle(tests_to_run)
-
-        tests_to_run, tests_in_other_chunks = self._finder.split_into_chunks(tests_to_run)
-        self._expectations.add_extra_skipped_tests(tests_in_other_chunks)
-        tests_to_skip.update(tests_in_other_chunks)
-
         return tests_to_run, tests_to_skip
 
     def _test_input_for_file(self, test_file):
         return TestInput(test_file,
                          self._options.slow_time_out_ms if self._test_is_slow(test_file) else self._options.time_out_ms,
                          self._test_requires_lock(test_file),
                          should_add_missing_baselines=(self._options.new_test_results and
                                                        not self._test_is_expected_missing(test_file)))
 
     def _test_requires_lock(self, test_file):
@@ skipping 268 matching lines @@
 
         stats = {}
         for result in initial_results.results_by_name.values():
             if result.type != test_expectations.SKIP:
                 stats[result.test_name] = {'results': (_worker_number(result.worker_name), result.test_number, result.pid, int(
                     result.test_run_time * 1000), int(result.total_run_time * 1000))}
         stats_trie = {}
         for name, value in stats.iteritems():
             json_results_generator.add_path_to_trie(name, value, stats_trie)
         return stats_trie