CODEWITHSUNDEEP
pythonpython-3.xsuper
Cedric H.
Cedric H.

Reputation: 8298

Using super() with a single argument in Python 3

I read quite a lot of material about the correct use of super with and without argument(s) but there is still something that I do not understand.

Let's consider the following example:

class A:
    def foo(self):
        pass

class B(A):
    pass

This is the part I understand:

super(B, B())  # <bound method A.foo of <__main__.B object at ...>>

which makes sense, as the two argument version of super, with the second argument being an object returns a bound object.

Let's say that I want to access the function in A, not the bound method. I would try this:

super(B).foo  # AttributeError: 'super' object has no attribute 'foo'

Which is the part that I do not understand.

Trying to "trick" super in the following way works though:

super(B, B).foo  # <function __main__.A.foo(self)>

What am I missing?

Upvotes: 2

Views: 90

Answers (1)

chepner
chepner

Reputation: 531215

super returns a proxy to some object. To figure out which object you want to proxy, you need to see how you would get a reference to the function without using super.

If a is an instance of A, then a.foo gives you the return value of A.__dict__['foo'].__get__(a, A), which is an instance of method. To get the function itself, you use A.foo, which is the return value of A.__dict__['foo'].__get__(None, A).

This suggests you need a proxy for the class, not the instance of the class, which is why you pass A, rather than an instance of A, as the second argument to super.

Upvotes: 1

Related Questions