python - 如何区分类内和类外使用 __setattr__ 设置的属性?

标签 python magic-methods setattr

__setattr__() 中有什么方法可以区分来自类或子/继承类内部的属性集,以及来自当前类或子类外部的属性集?

我想从“外部”更改设置属性的工作方式,在我制作模块的情况下,我希望用户在设置属性时与在类内部设置属性时具有不同的逻辑。

例如:
i.x = 5 当从类中调用时通常应该分配 5 并且 i 是它的一个实例,但是当从另一个类中调用时它应该减去 5 而不是设置为 5。

最佳答案

有点低级,但你可以使用 inspect 模块:

import inspect

class A:

    def __init__(self):
        self.__x = 0

    @property
    def x(self):
        return self.__x

    @x.setter
    def x(self, value):
        f = inspect.currentframe()
        if 'self' in f.f_back.f_locals and issubclass(type(f.f_back.f_locals['self']), A):
            print('Called from class!')
            self.__x = -value
        else:
            print('Called from outside!')
            self.__x = value

    def fn(self):
        print('Calling A.x from inside:')
        self.x = 10

class B(A):
    def __init__(self):
        super().__init__()

    def fn2(self):
        print('Calling B.x from inside:')
        self.x = 15

a = A()
print("A.x after init:", a.x)
print('Calling A.x from outside')
a.x = 10
print("A.x called from the outside:", a.x)
a.fn()
print("A.x called from the inside:", a.x)

b = B()
print("B.x after init:", b.x)
print('Calling B.x from outside')
b.x = 20
print("B.x called from the outside:", b.x)
b.fn2()
print("B.x called from the inside:", b.x)

打印:

A.x after init: 0
Calling A.x from outside
Called from outside!
A.x called from the outside: 10
Calling A.x from inside:
Called from class!
A.x called from the inside: -10
B.x after init: 0
Calling B.x from outside
Called from outside!
B.x called from the outside: 20
Calling B.x from inside:
Called from class!
B.x called from the inside: -15

关于python - 如何区分类内和类外使用 __setattr__ 设置的属性?,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/56550263/

相关文章:

python - 更改 Tensorflow 中新分配变量的形状

python - 在定义类时在 python 中设置具有给定名称的类属性

php 捕捉 array_push 的魔术方法

Python 类 % 运算符重载

android - 如何在 android 文件上设置扩展用户属性?

python - 在 UI 元素上使用 setattr()

python - 了解 Python Web 应用程序部署

python - 在python中访问内存地址

python - 使用 multiprocessing -Pool- 和 -sklearn-,代码运行但核心不显示任何工作

python - 如何使 math 和 numpy 模块中的现有函数支持用户定义的对象?