tatiana opened a new pull request, #39670:
URL: https://github.com/apache/airflow/pull/39670

   <!--
    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.
    -->
   
   Closes: #39486
   
   Valid DAGs that worked in Airflow 2.8.x  and had tasks with outlets with 
specific URIs, such as 
`Dataset("postgres://postgres:5432/postgres.dbt.stg_customers")`, stopped 
working in Airflow 2.9.0 & Airflow 2.9.1, after #37005 was merged.
   
   This was a breaking change in an Airflow minor version. We should avoid this.
   
   Airflow < 3.0 should raise a warning, and from Airflow 3.0, we can make 
errors by default. We can have a feature flag to allow users who want to see 
this in advance to enable errors in Airflow 2. x, but this should not be the 
default behaviour.
   
   The DAGs should continue working on Airflow 2.x minor/micro releases without 
errors (unless the user opts in via configuration).
   
   By running the following DAG with `apache-airflow==2.9.1` and 
`apache-airflow-providers-postgres==5.11.0`, as an example:
   ```
   from datetime import datetime
   
   from airflow import DAG
   from airflow.datasets import Dataset
   from airflow.operators.empty import EmptyOperator
   
   
   
   with DAG(dag_id='empty_operator_example', start_date=datetime(2022, 1, 1), 
schedule_interval=None) as dag:
   
       task1 = EmptyOperator(
           task_id='empty_task1',
           dag=dag,
           
outlets=[Dataset("postgres://postgres:5432/postgres.dbt.stg_customers")]
       )
   
       task2 = EmptyOperator(
           task_id='empty_task2',
           dag=dag
       )
   
       task1 >> task2
   ```
   
   Causes to the exception:
   ```
   Broken DAG: [/usr/local/airflow/dags/example_issue.py]
   Traceback (most recent call last):
     File 
"/usr/local/lib/python3.11/site-packages/airflow/datasets/__init__.py", line 
81, in _sanitize_uri
       parsed = normalizer(parsed)
                ^^^^^^^^^^^^^^^^^^
     File 
"/usr/local/lib/python3.11/site-packages/airflow/providers/postgres/datasets/postgres.py",
 line 34, in sanitize_uri
       raise ValueError("URI format postgres:// must contain database, schema, 
and table names")
   ValueError: URI format postgres:// must contain database, schema, and table 
names
   ```
   
   This PR introduces the following:
   
   1. A boolean configuration within `[core],` named 
`strict_dataset_uri_validation,` which should be `False` by default.
   
   2. When this configuration is `False,` Airflow should raise a warning saying:
   ```
   From Airflow 3, Airflow will be more strict with Dataset URIs, and the URI 
xx will no longer be valid. Please, follow the expected standard as documented 
in XX.
   ```
   
   3. If this configuration is `True,` Airflow should raise the exception, as 
it does now in Airflow 2.9.0 and 2.9.1
   
   4. From Airflow 3.0, we change this configuration to be True by default.
   
   


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to