Add cache_ok flag to sqlalchemy TypeDecorators.

[jkinkead]

This fixes #22647, which was marked closed due to lack of activity / feedback.

SQLAlchemy added the cache_ok flag in 1.4.14. If this is unset in a TypeDecorator subclass, it emits a log warning when the class is used. The fix is to set the value to True or False to suppress the warning.

I've set the value to True for the three TypeDecorator classes in the airflow.utils.sqlalchemy module, as these classes have no state and are safe to cache. Setting it to False would also suppress the warning, and leave behavior unchanged.

[boring-cyborg]

Congratulations on your first Pull Request and welcome to the Apache Airflow community! If you have any issues or are unsure about any anything please check our Contribution Guide (https://github.com/apache/airflow/blob/main/CONTRIBUTING.rst)
Here are some useful points:

• Pay attention to the quality of your code (flake8, mypy and type annotations). Our pre-commits will help you with that.
• In case of a new feature add useful documentation (in docstrings or in docs/ directory). Adding a new operator? Check this short guide Consider adding an example DAG that shows how users should use it.
• Consider using Breeze environment for testing locally, it’s a heavy docker but it ships with a working Airflow and a lot of integrations.
• Be patient and persistent. It might take some time to get a review or get the final approval from Committers.
• Please follow ASF Code of Conduct for all communication including (but not limited to) comments on Pull Requests, Mailing list and Slack.
• Be sure to read the Airflow Coding style.
  Apache Airflow is a community-driven project and together we are making it better 🚀.
  In case of doubts contact the developers at:
  Mailing List: dev@airflow.apache.org
  Slack: https://s.apache.org/airflow-slack

[jkinkead]

Fixes new issue #24504 .

[potiuk]

LGTM. But @ephraimbuddy @dstandish or @ashb co-review is needed

[github-actions]

The PR most likely needs to run full matrix of tests because it modifies parts of the core of Airflow. However, committers might decide to merge it quickly and take the risk. If they don't merge it quickly - please rebase it to the latest main at your convenience, or amend the last commit of the PR, and push it with --force-with-lease.

[uranusjr]

I think at least ExtendedJSON should use cache_ok = False since the value is not cachable? (Not sure about relativedelta.)

[jkinkead]

The cache_ok flag is confusing: It doesn't describe whether you can cache prepared values (the return values of process_bind_param), it describes whether you can use the instance of the type itself in caching the SQL statement the type is referenced in. Since these types are stateless and have static configuration, they're fine to use in the statement cache (they all prepare bind params identically).

See this section of the SQLAlchemy docs on how cache keys are generated.

Using a False here would exclude any statements with these columns from the SQLAlchemy statement cache, so it would have some performance implications.

[potiuk]

Makes sense . I rebased to see if it will be fine. also it might be necesary after #24819 is merged as we will be upgrading to sqlalchemy > 1.4.39 that will start emit this warning.

[potiuk]

Just in case - we should rebase this one after #24819 gets merged to test its behaviour anyway.

[boring-cyborg]

Awesome work, congrats on your first merged pull request!