[AIRFLOW-6428] Fix import path for airflow.utils.dates.days_ago in Example DAGs

[kaxil]

Import the days_ago function directly from airflow.utils.dates in all our Example DAGs.

Currently, without the entry in init.py, IDEs show that it could not find the reference to dates in most of our example_dags as they contain airflow.utils.dates.days_ago(1) and hence if you try to find the reference for days_ago function or dates modules it can't find it.

---

• Description above provides context of the change
• Commit message contains [AIRFLOW-6428]
• Unit tests coverage for changes (not needed for documentation changes)
• Commits follow "How to write a good git commit message"
• Relevant documentation is updated including usage instructions.
• I will engage committers as explained in Contribution Workflow Example.

In case of fundamental code change, Airflow Improvement Proposal (AIP) is needed.
In case of a new dependency, check compliance with the ASF 3rd Party License Policy.
In case of backwards incompatible changes please leave a note in UPDATING.md.
Read the Pull Request Guidelines for more information.

[codecov-io]

Codecov Report

Merging #7007 into master will decrease coverage by 0.28%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #7007      +/-   ##
==========================================
- Coverage   84.85%   84.56%   -0.29%     
==========================================
  Files         679      680       +1     
  Lines       38542    38543       +1     
==========================================
- Hits        32703    32594     -109     
- Misses       5839     5949     +110

Impacted Files	Coverage Δ
airflow/utils/__init__.py	100% <100%> (ø)
airflow/kubernetes/volume_mount.py	44.44% <0%> (-55.56%)	⬇️
airflow/kubernetes/volume.py	52.94% <0%> (-47.06%)	⬇️
airflow/kubernetes/pod_launcher.py	45.25% <0%> (-46.72%)	⬇️
airflow/kubernetes/refresh_config.py	50.98% <0%> (-23.53%)	⬇️
...rflow/contrib/operators/kubernetes_pod_operator.py	78.75% <0%> (-20%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update aa90753...267995a. Read the comment docs.

[potiuk]

Why don't we do it "differently" ...

I am not sure why we are using those imports in this form? Do we have something that holds us back from changing all "days_ago" imports to the form that is much more pythonic (IMHO)?

Adding anything to __init__.py inside the application creates unnecessary dependencies. Anybody using "airflow.utils.somethingelse" will add an implicit dependency to "airflow.utils.dates" if we add dates to __init__.py - even if it is not used directly. This adds unnecessary dependencies (and leads to circular dependencies).

I think most of our __init__.py should be empty (or removed if we go to implicit python3 packages). I believe adding anything to __init__.py makes only sense if we provide a reusable library with one package - when if you import it, you should have access to all exposed functions.

I am happy to discuss it though, as we might have different understanding - and maybe we should expose all "exposable" classes from unit in this way as part of the "official airflow interface" (but I still think import should be from airflow.utils.xxx import yyy anyway).

Just to summary - we have two options:

1. Import airflow and then rely on the init packages

import airflow

and then using

airflow.utils.dates.days_ago(2)

2. Import the function directly (much better IMHO).

from airflow.utils.dates import days_ago

Option 1 (with importing the whole 'airflow').

Option 2: (wiht importing only the function we need)

WDYT @kaxil?

[potiuk]

Just wanted to discuss general import approach in this case.

[kaxil]

Just wanted to discuss general import approach in this case.

I agree the 2nd option sounds lot better, I have changed the PR and updated all the example DAGs by importing just the function from module instead of init.py.

[potiuk]

👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍 👍

[potiuk]

Needs rebase @kaxil :(.

BTW. I think we need a "Needs rebase" probot.

[potiuk]

Hmm https://github.com/tibdex/autorebase - but it seems a bit abandoned in favour of autosquash.