使用类名调用静态方法-好还是不好？

Question

I am getting more and more familiar with static methods and class methods and I know the difference between each. But I ran into a problem today where I would kind of like to reference both self and cls in a method.

The only way I know how to accomplish this is to make a normal class method (not with @classmethod , but simply with def ) and call the class with it's name explicitely like so:

class myClass:
    def __init__():
        self._ser.connect('COM5')

    def ask(self, message: str) -> str:
        return myClass.clean_output(self._ser.query(message))

    @staticmethod
    def clean_output(dirty_string: str):
        clean_string = dirty_string.strip().replace(chr(4))
        return clean_string

This example is an over-simplified version of the philosophy. I would like to call a clean or parse function on data I get back, like from a serial device. Is there any way to implement the ask method like so?:

    def ask(self, message: str) -> str:
        return cls.clean_output(self._ser.query(message))

Or is it ok that I'm calling it with myClass explicitly like that? If it is, when should programmers use @classmethod and when is it permissible to use the class name itself? Is using the @classmethod decorator only really needed when you expect subclassing to happen?

Answer 1

Just call the static method on self :

def ask(self, message: str) -> str:
    return self.clean_output(self._ser.query(message))

It is available there too.

Attributes on a class are always available on the instances too (provided there is no attribute on the instance itself with the same name masking it). Methods (be they regular, static or class methods) are no exception, they too are just attributes. Their binding behaviour doesn't matter here.