Using python's super() within a new Julia-defined Python class

[bensetterholm]

I need to define a Python class that inherits another Python class with Python's super() function in my Julia code. However, I don't know how to access super (presuming that is currently even possible).

As a MWE of my problem, here is an attempt to translate the following pure Python code (taken from https://realpython.com/python-super/) into its equivalent PythonCall representation.

Python:

class Rectangle:
    def __init__(self, length, width):
        self.length = length
        self.width = width

    def area(self):
        return self.length * self.width

    def perimeter(self):
        return 2 * self.length + 2 * self.width

class Square(Rectangle):
    def __init__(self, length):
        super().__init__(length, length)

Julia attempt:

using PythonCall

Rectangle = pytype("Rectangle", (), [
    "__module__" => "__main__",

    pyfunc(
        name = "__init__",
        function (self, length, width)
            self.length = length
            self.width = width
            return
        end
    ),

    pyfunc(
        name = "area",
        self -> self.length * self.width
    ),

    pyfunc(
        name = "perimeter",
        self -> 2 * self.length + 2 * self.width
    ),
])

Square = pytype("Square", (Rectangle,), [
    "__module__" => "__main__",

    pyfunc(
        name = "__init__",
        function (self, length)
            super().__init__(length, length)
            return
        end
    )
])

Note that Rectangle works exactly as expected. However Square does not work due to the undefined super(). I do not see anything like pysuper() defined in PythonCall either. Am I missing something, or is this currently not possible?

[cjdoris]

Not currently possible I'm afraid. You could do Rectangle.__init__(self, ...) instead.

[bensetterholm]

Thank you; this seems to do the trick! Are there any potential "gotchas" you are aware of to using this syntax instead of super? (I only write Python code when I absolutely have to, and don't necessarily have a good mental model of all of the nuances of Python classes.)

[cjdoris]

If your class only has one base class then I think both ways are equivalent, whereas super() can properly handle multiple bases by traversing the MRO properly.