用户定义的__mul__方法不是可交换的 [英] User defined __mul__ method is not commutative
问题描述
我编写了一个类来表示Python中的向量(作为练习),但是在扩展内置运算符时遇到了问题。
I wrote a class to represent vectors in Python (as an exercise) and I'm having problems with extending the built-in operators.
我定义了<向量类的code> __ mul __ 方法。问题在于,在表达式 x * y
中,解释程序调用x的 __ mul __
方法,而不是y。
I defined a __mul__
method for the vector class. The problem is that in the expression x * y
the interpreter calls the __mul__
method of x, not y.
所以 vector(1、2、3)* 2
返回一个向量<2、4、6>这应该;但是 2 * vector(1、2、3)
会创建TypeError,因为内置的int类不支持与用户定义的向量相乘。
So vector(1, 2, 3) * 2
returns a vector <2, 4, 6> just like it should; but 2 * vector(1, 2, 3)
creates a TypeError because the built-in int class does not support multiplication by my user-defined vectors.
我可以通过简单地编写一个新的乘法函数来解决这个问题
I could solve this problem by simply writing a new multiplication function
def multiply(a, b):
try:
return a * b
except TypeError:
return b * a
,但这需要重新定义我要与用户定义的类一起使用的每个函数。
but this would require redefining every function that I want to use with my user-defined classes.
有没有办法才能使内置函数正确处理此问题?
Is there a way to make the built-in function handle this correctly?
推荐答案
如果您想对不同类型使用可交换性 您需要实现 __ rmul __()
。如果实现,将被调用,就像所有 __ r * __()
特殊方法一样,如果操作否则会引发 TypeError
。注意参数已交换:
If you want commutativity for different types you need to implement __rmul__()
. If implemented, it is called, like all __r*__()
special methods, if the operation would otherwise raise a TypeError
. Beware that the arguments are swapped:
class Foo(object):
def __mul_(self, other):
''' multiply self with other, e.g. Foo() * 7 '''
def __rmul__(self, other):
''' multiply other with self, e.g. 7 * Foo() '''
这篇关于用户定义的__mul__方法不是可交换的的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!