Python继承默认值被覆盖

Question

我试图在python中使用继承，但有一个问题，即当没有给出famil的值时，名为famil的字段会被前一个实例的输入覆盖。你知道吗

 class Person:
    def __init__(self, famil=[], name="Jim"):
        self._family = famil
        self._name = name

    def get_family(self):
        return self._family

    def get_age(self):
        return self._age


class Woman(Person):
    def __init__(self, family=[], name="Jane"):
        print(family)
        Person.__init__(self, family, name)

    def add_family(self, name):
        self._family += [name]


if __name__ == "__main__":
    a = Woman()
    a.add_family("john")
    print(a.get_family())
    b = Woman()
    print(b.get_family())
    print(a.get_family())

输出：

[]
['john']
['john']
['john']
['john']

预期产量：

[]
['john']
[]
[]
['john']

这让我很困惑，因为我试图学习继承，我认为a和b应该相互独立。你知道吗

Answer 1

正如在注释中提到的，这是一个可变默认参数的问题，在Python中工作不好。另见https://docs.python-guide.org/writing/gotchas/

最好的处理方法是使默认参数不可变，如果没有提供，则指定可变的默认值。例如：

class Person:
    def __init__(self, family=None, name="Jim"):
        self._family = family or []
        self._name = name

class Woman(Person):
    def __init__(self, family=None, name="Jane"):
        super().__init__(self, family or [], name)