Why is functools.partial not a subclass of types.FunctionType?

Question

In my code I need to detect if a variable is a function or not and preform some actions on it.

Everything went well until I now created a partial function using functools and suddenly some of my tests fail:

import types
import functools

def f(s):
    print(s)

l = lambda s: print(s)

pf = functools.partial(f, 'Hello World')
pl = functools.partial(l, 'Hello World')
test_f = isinstance(f, types.FunctionType) # True
test_l = isinstance(l, types.FunctionType) # True
test_pf = isinstance(pf, types.FunctionType) # False
test_pl = isinstance(pl, types.FunctionType) # False

Why is there a difference between those? Both varieties are callable... Even more importantly, how can I detect if some variable is a function or not even when it's a partial function if I can't use types.FunctionType?

Answer 1

functools.partial is a class with a __call__ method, it says in the documents:

Return a new partial object which when called will behave like func

(bold emphasis added by me)

We can confirm this in our Python REPL:

>>> from functools import partial
>>> add_one = partial(sum, 1)
>>> type(add_one)
<class 'functools.partial'>

A Python equivalent would be something like this:

class Partial:

    def __init__(self, func, *args, **kwargs):
        self.func = func
        self.args = args
        self.kwargs = kwargs

    def __call__(self, *args, **kwargs):
        return self.func(*self.args, *args, **self.kwargs, **kwargs)

So it creates a simple wrapper object around the function, and an object like this simply isn't a function. types.FunctionType only works on actual functions.

What you're looking for is a way to check if an object is callable, for that you can use the built-in callable function:

>>> callable(sum)
True
>>> callable(3)
False
>>> callable(functools.partial(sum, 1))
True