Singleton pattern in python - Fri, Nov 11, 2022

Singleton pattern in python

This snippet contains the implementation of a singleton pattern in python:

class Singleton:

    __instance = None

    def __init__(self):
        print("Singleton.__init__, do nothing")     # This method gets called multiple times

    def __new__(cls):
        if not cls.__instance:
            cls._instance = super(Singleton, cls).__new__(cls)
            # Initialize the instance here instead of __init__ to avoid re-initialization
        return cls.__instance

if __name__ == '__main__':
    s1 = Singleton()
    s2 = Singleton()
    print(s1 is s2)     # True

    # Print the memory address of the two instances
    print(hex(id(s1)))
    print(hex(id(s2)))

First take a look at the class Singleton. The key point here is that whenever a class is instantiated the method __new__ is called to create the object. In the singleton above we override this method and check whether an instance has already been created or not. If no instance has yet been created we create one and store it as a class variable. The next time someone tries to instantiate the class, the same class instance is returned instead of a new one. The output of running this code should look like this:

Singleton.__init__, do nothing
Singleton.__init__, do nothing
True
0x7f9a6dc9fcd0
0x7f9a6dc9fcd0

You might have noticed that __init__ has been called twice despite the instance being the same. That’s why initialization code for this singleton should not be put in the __init__ method but rather be done after instance creation (see comment in line 30). Things might get even a bit more complicated when the singleton inherits from another class.
A full example with inheritance can be found here .

Back to Home