Add executorlib container

[niklassiemer]

closes #584

[github-actions]

👈 Launch a binder notebook on branch pyiron/pyiron_workflow/executorlib01

[liamhuber]

@niklassiemer, I guess this

 ======================================================================
ERROR: test_executors (integration.test_workflow.TestWorkflow.test_executors)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/pyiron_workflow/pyiron_workflow/tests/integration/test_workflow.py", line 152, in test_executors
    Workflow.create.ExecutorlibExecutor,
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
AttributeError: 'Creator' object has no attribute 'ExecutorlibExecutor'

----------------------------------------------------------------------

Just needs to be replaced with the relevant Workflow.create.excutorlib... counterpart, then we're good to go.

I suggest adopting one of the two suggestions I made to tighten up the interface, but if you hate both I can live with it as-is and otherwise once the tests are passing I'm happy here. Thanks for taking care of it!

[niklassiemer]

Thanks for the feedback! I was actually using the tests here to figure out what needs to be changed in the tests xD and I was thinking to add __all__ Executors in a loop. However, I like the class attributes more :)

Indeed, I would like to have pyiron_workflow aligned to executorlib asap to do a new release and allow the cmti environment to use the newest base / atomistics/workflow stack.

#Edit
I had not seen you second review comment at the mobile at first... I love it :D

[niklassiemer]

Funny enough all my code will be gone and you did the full work after all xD

[niklassiemer]

I should not try these things on mobile xD do we want to maintain downwards compatibility to older versions of executorlib? Otherwise I would need to target the executorlib update branch I suppose :)

[niklassiemer]

@liamhuber I do not know why the coveralls/codacy is failing - is it related to some hickups (I saw a comment of Jan on another PR but I do not remember where and I am not sure if this is related).
From my perspective this is good to go otherwise. Also the API change is rather small (....create.ExecutorlibExecutor -> ....create.executorlib.Executor or ....create.executorlib.SingleNodeExecutor; depending on the version of executorlib).
If you are happy, I would like a merge and a timely new release, such that I can get this into the cmti environment on Monday.