Reputation: 2143
I recently learned about operator overloading in python and I would like to know if the following is possible.
Consider the following hypothetical/contrived class.
class My_Num(object):
def __init__(self, val):
self.val = val
def __add__(self, other_num):
if isinstance(other_num, My_Num):
return self.val + other_num.val
else:
return self.val + other_num
I know that the way that's written above, I can do things like this
n1 = My_Num(1)
n2 = My_Num(2)
n3 = 3
print n1 + n2
print n1 + n3
and those will work as expected. I also know that the way it's currently written I can't do this
n1 = My_Num(1)
n2 = 2
print n2 + n1
Is there any way around this? I know this example is contrived but I have an application in which it would be very useful if when I did operator overloading, the class for which I define the operator can appear on the right hand side of operator. Is this possible in python?
Upvotes: 20
Views: 6591
Reputation: 151157
Yes. For example, there is __radd__
. Also, there are none for __le__()
, __ge__()
, etc., but as Joel Cornett rightly observes, if you define only __lt__
, a > b
calls the __lt__
function of b
, which provides a workaround.
>>> class My_Num(object):
... def __init__(self, val):
... self.val = val
... def __radd__(self, other_num):
... if isinstance(other_num, My_Num):
... return self.val + other_num.val
... else:
... return self.val + other_num
...
>>> n1 = My_Num(1)
>>> n2 = 3
>>>
>>> print n2 + n1
4
>>> print n1 + n2
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: unsupported operand type(s) for +: 'My_Num' and 'int'
Note that in at least some cases it's reasonable to do something like this:
>>> class My_Num(object):
... def __init__(self, val):
... self.val = val
... def __add__(self, other_num):
... if isinstance(other_num, My_Num):
... return self.val + other_num.val
... else:
... return self.val + other_num
... __radd__ = __add__
Upvotes: 17
Reputation: 60210
You have to overload the __radd__
method (right-side addition). Your function should look pretty much the same as your __add__
method, e.g.:
def __radd__(self, other):
return self.val + other.val
Upvotes: 2