How to make same instance with same arguments but different for different arguments in python

Question

class A():
    def __init__(self, fn):
        self.fn = fn

a1 = A('a')        
a2 = A('a') 
b = A('b') 

print (a1==a2)
print (a1==b)

result should be True for first but False for second. I knew some way to implement singleton in Python. But all of them only generate one instance for every call. How do we link the __new__ method to __init__?

define `def __eq__(self,other): return self.fn == other.fn` and the `==` operator is redefined to match your needs. — Jean-François Fabre, Aug 01 '17 at 21:19

score 3 · Accepted Answer · answered Aug 01 '17 at 21:19

3

There is no reason to think about singletons here and no reason to do anything with __new__. If you want two instances to be considered equal based on some condition, then you need to define __eq__.

def __eq__(self, other):
    return isinstance(other, A) and self.fn == other.fn

(Note, fn is usually used as a holder for functions; you should think of another attribute name.)

answered Aug 01 '17 at 21:19

Daniel Roseman

588,541
66
880
895

I was going to answer that (except for the `isinstance` part). Maybe you should add that the `==` should be redefined if `fn` and `self.fn` can be 2 different class instances bearing the same values (and on, and on) – Jean-François Fabre Aug 01 '17 at 21:21
also `isinstance(other, A)` should probably be `type(other) == A` to restrict to actual `A` objects, not children – Jean-François Fabre Aug 01 '17 at 21:22
Thank you for you help. How about if I want to make a1 is a2 equal True? – Tiancheng Liu Aug 02 '17 at 13:35

score 0 · Answer 2 · answered Aug 01 '17 at 21:31

It is not a singleton you're looking for. A singleton is when your class can produce only one instance at the time, so that you will have A('a') is A('b') True.

What can be deduced from your example is you're looking to define __eq__, which has been explained here. Note that if you want to test equality for custom objects, you may have to define __hash__, which is explained here

How to make same instance with same arguments but different for different arguments in python

2 Answers2