Unit test for celery change_state wrapper

[o-nikolas]

If an older version of main is being used BaseExecutor.change_state may have fewer parameters, catch the TypeError that is raised in this situation instead of the previously caught AttributeError.

Add a unit test to cover this scenario

fixes #40011
fixes #39980
related #40012 - This PR solves the same issue, but didn't include tests. I spent some time today noodling around with a test case and it turned out to be stubborn. So I'm opening this PR. It can either be merged or used as inspiration for #40012

---

^ Add meaningful description above
Read the Pull Request Guidelines for more information.
In case of fundamental code changes, an 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 a newsfragment file, named {pr_number}.significant.rst or {issue_number}.significant.rst, in newsfragments.

[o-nikolas]

CC @eladkal @leakec

[leakec]

@o-nikolas I'm very unfamiliar with the airflow test case suite (I'm new to airflow in general), so I doubt I'll get much further in terms of writing a test without a significant time sink. Similarly, I wasn't sure if there were any cases where AttributeError could still be raised, it didn't look like that to me, but I left both in there just in case, i.e., #40012 catches both exception types rather than replacing AttributeError with TypeError.

If AttributeError will not be thrown and TypeError is the one that will always be raised here (looks like that to me), then this PR is a better fix than #40012, and it has a test case, which #40012 does not. If this test case captures everything (to me it looks like it does), then I think we should just close #40012 without merging and merge this one instead.

[o-nikolas]

@o-nikolas I'm very unfamiliar with the airflow test case suite (I'm new to airflow in general), so I doubt I'll get much further in terms of writing a test without a significant time sink. Similarly, I wasn't sure if there were any cases where AttributeError could still be raised, it didn't look like that to me, but I left both in there just in case, i.e., #40012 catches both exception types rather than replacing AttributeError with TypeError.

If AttributeError will not be thrown and TypeError is the one that will always be raised here (looks like that to me), then this PR is a better fix than #40012, and it has a test case, which #40012 does not. If this test case captures everything (to me it looks like it does), then I think we should just close #40012 without merging and merge this one instead.

It looks like @potiuk has merged it anyway so problem solved 😆

I'll rebase this PR, because it still adds a test case, which we can merge or not, let's see what the community thinks.

[leakec]

@o-nikolas I think adding the test case is definitely worth it. Also, would be great to get feedback on whether that exception should be

except TypeError:

or

except (AttributeError, TypeError):

If it's the former, we should merge what you have in.