def get_task()

in api-server/tesazure/backends/batch/__init__.py [0:0]
66 lines of code
32 McCabe index (conditional complexity)

    def get_task(self, task_id):
        try:
            batch_job = self.batch_client.job.get(task_id)
        except azbatch.models.BatchErrorException as e:
            if e.error.code == "JobNotFound":
                return False
            raise

        tes_task = tesmodels.TesTask()
        tes_task.creation_time = batch_job.creation_time

        tes_log = tesmodels.TaskLog()
        tes_log.start_time = batch_job.creation_time

        if batch_job.state == azbatch.models.JobState.completed:
            tes_log.end_time = batch_job.state_transition_time

        # Default state inheritance, with 'active' as QUEUED unless we get more detailed into from tasks
        state_map = {
            azbatch.models.JobState.active: tesmodels.TaskStatus.QUEUED,
            azbatch.models.JobState.completed: tesmodels.TaskStatus.COMPLETE,
            azbatch.models.JobState.deleting: tesmodels.TaskStatus.CANCELED,
            azbatch.models.JobState.disabled: tesmodels.TaskStatus.PAUSED,
            azbatch.models.JobState.disabling: tesmodels.TaskStatus.PAUSED,
            azbatch.models.JobState.enabling: tesmodels.TaskStatus.PAUSED,
            azbatch.models.JobState.terminating: tesmodels.TaskStatus.RUNNING,
        }
        tes_task.state = state_map.get(batch_job.state, tesmodels.TaskStatus.UNKNOWN)

        # Check prep and release task status, but only if they exist
        if any([batch_job.job_preparation_task, batch_job.job_release_task]):
            for task_info in self.batch_client.job.list_preparation_and_release_task_status(batch_job.id):
                if task_info.job_preparation_task_execution_info:
                    # System error if file marshalling failed
                    if task_info.job_preparation_task_execution_info.result == azbatch.models.TaskExecutionResult.failure:
                        tes_task.state = tesmodels.TaskStatus.SYSTEM_ERROR
                    # Initializing when files are being downloaded
                if task_info.job_preparation_task_execution_info.state == azbatch.models.JobPreparationTaskState.running:
                    tes_task.state = tesmodels.TaskStatus.INITIALIZING

                if task_info.job_release_task_execution_info:
                    # System error if file marshalling failed
                    if task_info.job_release_task_execution_info.result == azbatch.models.TaskExecutionResult.failure:
                        tes_task.state = tesmodels.TaskStatus.SYSTEM_ERROR
                    # Active if files are being uploaded
                    if task_info.job_release_task_execution_info.state == azbatch.models.JobReleaseTaskState.running:
                        tes_task.state = tesmodels.TaskStatus.RUNNING

        # Check status against batch task summary
        if batch_job.execution_info.terminate_reason == azbatch.models.TaskExecutionResult.failure:
            tes_task.state = tesmodels.TaskStatus.EXECUTOR_ERROR

        if tes_task.state == tesmodels.TaskStatus.UNKNOWN:
            task_status = self.batch_client.job.get_task_counts(batch_job.id)
            if task_status.failed:
                tes_task.state = tesmodels.TaskStatus.EXECUTOR_ERROR
            elif task_status.running == 0 and task_status.active > 0:
                tes_task.state = tesmodels.TaskStatus.QUEUED

        # Get the executor logs (each batch task)
        for batch_task in self.batch_client.task.list(batch_job.id):
            if batch_task.execution_info.result == azbatch.models.TaskExecutionResult.failure:
                # Above assumes EXECUTOR_ERROR on any task failure. Since we are responsible for orchestrating Batch,
                # Batch USER_ERROR should be a TES-user-facing SYSTEM_ERROR regardless of if batch_task.execution_info.failure_info.category is
                # azbatch.models.ErrorCategory.server_error or azbatch.models.ErrorCategory.user_error.
                tes_task.state = tesmodels.TaskStatus.SYSTEM_ERROR

            # Executor logs
            executor_log = tesmodels.ExecutorLog(start_time=batch_task.creation_time, end_time=batch_task.state_transition_time)
            executor_log.exit_code = batch_task.execution_info.exit_code

            # Add output streams to executor logs (only available if job is completed)
            if batch_job.state == azbatch.models.JobState.completed:
                with _handle_batch_error(should_raise=False):
                    output = self.batch_client.file.get_from_task(batch_job.id, batch_task.id, 'stdout.txt')
                    for data in output:
                        executor_log.stdout += data.decode('utf-8')

                with _handle_batch_error(should_raise=False):
                    output = self.batch_client.file.get_from_task(batch_job.id, batch_task.id, 'stderr.txt')
                    for data in output:
                        executor_log.stderr += data.decode('utf-8')

            tes_log.logs = tes_log.logs + [executor_log]

        # FIXME: tes_task is not loaded from db so outputs is empty here
        for output in tes_task.outputs:
            output_log = tesmodels.OutputFileLog()
            output_log.path = output.path
            output_log.url = output.url
            output_log.size_bytes = 0  # TODO
            tes_log.outputs = tes_log.outputs + [output_log]

        tes_task.logs = [tes_log]
        return tes_task