commit be007ad68cf45d05bb907d7e68b1e70ee53753da
author Ahmet Altay <aaltay@gmail.com> Tue Aug 06 10:34:56 2019 -0700
committer GitHub <noreply@github.com> Tue Aug 06 10:34:56 2019 -0700
tree eef54b5033478381867489e910de3186538ac8d4
parent a35de2f987e22f320e4be67adb40487464d85b35
parent 387b041e61d10d6b1f37b01171105f0d34fb178e

Merge pull request #9137 from tvalentyn/cp_9094

[BEAM-7838] LTS backport: Default to PiplelineState.UNKNOWN when job state received via v1beta3 cannot be recognized. 

sdks/python/apache_beam/runners/dataflow/dataflow_runner.py

diff --git a/sdks/python/apache_beam/runners/dataflow/dataflow_runner.py b/sdks/python/apache_beam/runners/dataflow/dataflow_runner.py
index 9f84305..9e42d82 100644
--- a/sdks/python/apache_beam/runners/dataflow/dataflow_runner.py
+++ b/sdks/python/apache_beam/runners/dataflow/dataflow_runner.py
@@ -1078,9 +1078,9 @@
   def _get_job_state(self):
     values_enum = dataflow_api.Job.CurrentStateValueValuesEnum
 
-    # TODO: Move this table to a another location.
-    # Ordered by the enum values.
-    api_jobstate_map = {
+    # Ordered by the enum values. Values that may be introduced in
+    # future versions of Dataflow API are considered UNKNOWN by the SDK.
+    api_jobstate_map = defaultdict(lambda: PipelineState.UNKNOWN, {
         values_enum.JOB_STATE_UNKNOWN: PipelineState.UNKNOWN,
         values_enum.JOB_STATE_STOPPED: PipelineState.STOPPED,
         values_enum.JOB_STATE_RUNNING: PipelineState.RUNNING,
@@ -1092,7 +1092,7 @@
         values_enum.JOB_STATE_DRAINED: PipelineState.DRAINED,
         values_enum.JOB_STATE_PENDING: PipelineState.PENDING,
         values_enum.JOB_STATE_CANCELLING: PipelineState.CANCELLING,
-    }
+    })
 
     return (api_jobstate_map[self._job.currentState] if self._job.currentState
             else PipelineState.UNKNOWN)