Refactor the two ways of adding into a single method

[liamhuber]

Which is itself refactored so the checks are a bit more readable. There was one behaviour change here, and that is that we now allow iterating on labels even during node-addition-by-attribute-assignment (of course, still only when strict_naming is deactivated). I.e.

from pyiron_contrib.workflow.workflow import Workflow

@Workflow.wrap_as.node("y")
def fnc(x=0): return x + 1

wf = Workflow("my_workflow")
wf.deactivate_strict_naming()

wf.baz = fnc(label="base")
wf.baz = fnc(label="zeroth")
wf.baz = fnc(label="first")

print(wf.nodes.keys())
>>> dict_keys(['baz', 'baz0', 'baz1'])

Test is updated accordingly.

[coveralls]

Pull Request Test Coverage Report for Build 5019152589

• 31 of 34 (91.18%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.05%) to 13.317%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
pyiron_contrib/workflow/workflow.py	31	34	91.18%

Totals
Change from base Build 4998433075:	0.05%
Covered Lines:	1655
Relevant Lines:	12428

---

💛 - Coveralls