It's hard to find answers to this problem when you don't know exactly how to describe it...
What is the most idiomatic way to deal with a fairly deep (but fixed) nested set of tests that you want to run sequentially but terminate as soon as the first one comes up with a successful result?
Instead of the following
Option 1: (results in way too much indentation)
def make_decision():
results = ... some code or function that returns a list or None
if results:
decision = random.choice(results)
else:
results = ... other code or function
if results:
decision = random.choice(results)
else:
results = ... other code or function
if results:
decision = random.choice(results)
else:
results = ... other code or function
if results:
decision = random.choice(results)
...etc.
else:
decision = None
print(decision)
return decision
Option 2
Another option is to return from the function early but I'm not sure it's good practice to have so many returns scattered about and in this case I would prefer to do one final thing at the end (e.g. print(decision)) before returning:
def make_decision():
results = ... some code or function that returns a list or None
if results:
return random.choice(results)
results = ... other code or function
if results:
return random.choice(results)
results = ... other code or function
if results:
return random.choice(results)
...etc.
return None
Option 3
Finally I could make each test a separate function as described here and call them iteratively, provided each test function has the same set of arguments.
def test1(args):
...
def test2(args):
...
def test3(args):
...
def make_decision():
decision = None
for test in [test1, test2, test3, ...]:
results = test(args)
if results:
decision = random.choice(results)
break
print(decision)
return decision
This looks the best but I hadn't planned to make a function for every test and some tests can be done with the same function but with different arguments and some are just one-liners whereas others are multiple lines. So then would I have to build a list of functions and arguments before starting the loop? Or make a list of partial functions?
Any better suggestions welcome (before I go ahead with option 3 above)
UPDATE 2018-07-21:
A future potential option
Unbeknown to me as I was wrestling with this problem, PEP 572 was approved (and Mr van Rossum resigned as a consequence). If this PEP is implemented the following solution will also be possible I think:
def make_decision():
if (results := ... some code or function) is not None:
decision = random.choice(results)
elif (results := ... some code or function) is not None:
decision = random.choice(results)
...etc.
else:
decision = None
return decision
Sorry, this is obvious now that re-look at it:
def make_decision():
results = ... some code or function that returns a list or None
if results is None:
results = ... other code or function
if results is None:
results = ... other code or function
...etc.
if results is None:
decision = None
else:
decision = random.choice(results)
print(decision)
return decision