[Findit] Strip platform from step_name before triggering try job.

appengine/findit/waterfall/swarming_tasks_to_try_job_pipeline.py

 # Copyright 2016 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.
 
 from pipeline_wrapper import BasePipeline
 from waterfall.process_swarming_task_result_pipeline import (
     ProcessSwarmingTaskResultPipeline)
 from waterfall.run_try_job_for_reliable_failure_pipeline import (
     RunTryJobForReliableFailurePipeline)
 from waterfall.trigger_swarming_task_pipeline import TriggerSwarmingTaskPipeline
@@ skipping 14 matching lines @@
     # The format would be [('step1', {'flaky_tests': ['test1', ..], ..}), ..]
     classified_tests_by_step = []
 
     if try_job_type == TryJobType.TEST:
       for step_name, tests in targeted_tests.iteritems():
         if not tests:  # Skip non-swarming tests.
           continue
         task_id = yield TriggerSwarmingTaskPipeline(
             master_name, builder_name, build_number, step_name, tests)
         step_future = yield ProcessSwarmingTaskResultPipeline(
-            master_name, builder_name, build_number, step_name, task_id)
+            master_name, builder_name, build_number, step_name,
+            task_id)

[stgao, 2016/03/10 23:50:28]

Could be reverted?

[chanli, 2016/03/11 00:49:41]

Missed... Done

         classified_tests_by_step.append(step_future)
 
    # Waits until classified_tests_by_step are ready.
     yield RunTryJobForReliableFailurePipeline(
         master_name, builder_name, build_number, good_revision,
         bad_revision, blame_list, try_job_type, compile_targets, targeted_tests,
         *classified_tests_by_step)