JonnyIncognito commented on a change in pull request #6210: [AIRFLOW-5567] BaseAsyncOperator URL: https://github.com/apache/airflow/pull/6210#discussion_r338510991
########## File path: airflow/models/base_async_operator.py ########## @@ -0,0 +1,161 @@ +# -*- coding: utf-8 -*- +# +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +""" +Base Asynchronous Operator for kicking off a long running +operations and polling for completion with reschedule mode. +""" + +from abc import abstractmethod +from typing import Dict, List, Optional, Union + +from airflow.models import SkipMixin, TaskReschedule +from airflow.models.xcom import XCOM_EXTERNAL_RESOURCE_ID_KEY +from airflow.sensors.base_sensor_operator import BaseSensorOperator +from airflow.utils.decorators import apply_defaults + + Review comment: I agree with trying to put the behaviour into BaseOperator. As it stands, the use-cases are: 1. Only operate (`BaseOperator` <- e.g. `EmrAddStepsOperator`) 1. Only sense (`BaseSensorOperator` <- e.g. `EmrStepSensor`) 1. Do both (to be determined) 1. Do none (`BaseOperator` <- e.g. `DummyOperator`) Currently 1, 2 and 4 are implemented as variations of execute(); 2 is the odd one out in having a special class, as would 3 with the current proposal. If BaseOperator had a default execute() that runs two phases, the other behaviours can be achieved by optionally implementing either of the phases. It's a typical problem using class hierarchies for behaviours, that it's hard to mix and match. I was actually thinking about whether we could use traits to keep the behaviours cleanly separated into their own implementations. But it's probably overkill with so few. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
