Understanding parentage in environment diagrams

[Clunk_Head]

I'm confused about defining the parent of a function as the frame in which the function is defined as it pertains to a HOF that defines a function and then drops out of scope.

def foo(x):
    print(f'foo x: {x}')
    return x

def bar(f):
    print(f'bar f: {f}')
    def bat(y):
        print(f'bat f(y): {f(y)}')
        return y
    return bat

new_func = bar(foo)
# f1    bar         parent = global
#       f = foo
#       return bat
# Doesn't f1 cease to exist here? Out of scope

new_func(5)
# f2    new_func    parent = f1
#       y = 5
#       return y
# If f1 is out of scope, how can the parent of f2 be f1?

[deanhystad]

I think you need a better example.

def Outer(x):
    def Inner(y):
        return y**x

    return Inner


squares = Outer(2)
cubes = Outer(3)
print(squares, cubes)
for i in range(5):
    print(i, squares(i), cubes(i))

Output:
<function Outer.<locals>.Inner at 0x000001C6CDF6C280> <function Outer.<locals>.Inner at 0x000001C6CE27C280>
0 0 0
1 1 1
2 4 8
3 9 27
4 16 64

squares and cubes each reference a function closure. The difference between a function and a function closure is that the closure also contains all the variables in the enclosed scope. This is how Inner() remembers x==2 when called using squares, and x==3 when called using cubes. Like other objects in Python, the function closures get garbage collected when their reference count goes to zero.