Odd type syntax used with isinstance()

Question

I was recently tasked with migrating a bunch of existing python 2.7 scripts to python 3.x, and noticed a statement in one of the scripts that I don't quite understand:

if isinstance(some_variable, (list,)):
   # ...

At first I thought the (,) syntax might be a way to express a partially typed tuple, but that doesn't seem to be the case:

>>> isinstance(([],123), (list,))
False

Looking at the calling code, I found that the only argument values ever passed as some_variable were lists, so I tested:

>>> isinstance([], (list,))
True

So why would you use (list,) instead of just list with isinstance()?

I suspect that there might be other (list-like?) types for which the call to isinstance() would return True, but I don't have the requisite python experience to think of what it be :)

Answer 1

This is (list,) tuple with one element and like this list and for your example :

>>> isinstance(([],123), (list,))
False 

>>> isinstance(([],123), (tuple,))
True

>>> isinstance(([],123), tuple)
True

if you want check isinstance do with map like below:

>>> list(map(lambda x : isinstance(x , (list,)), ([],123)))
[True, False]

>>> all(map(lambda x : isinstance(x , (list,)), ([],123)))
False

>>> any(map(lambda x : isinstance(x , (list,)), ([],123)))
True

Answer 2

isinstance can get a type OR a tuple of types as the second argument.

(list,) is a tuple with 1 entry only. In the case that you have only one type to compare against you dont have to use tuple