Question

I have a class defined as

class Conditional_Singleton(object):
    def __init__(self, condition=None):
        self.initialization = some_func_of(condition)

Since the “initialization” step really takes a lot of time to run and the output could be huge, I would really like that the class not be recreated over and over again given the same input, but do so once the input changes. Any ideas?

Answer 1

class Conditional_Singleton(object):
    saved_insances = {}
    def __new__(cls, condition = None):
        if not condition in cls.saved_instances:
            cls.saved_instances[condition] = super(
                Conditional_Singleton, cls).__new__(cls)
        return cls.saved_instances[condition]
    def __init__(self, condition = None):
        if not hasattr(self, 'initialization'):
            self.initialization = some_func_of(condition)
        

This will save the dict of created instances in the class itself and automatically reuse them if the same condition is passed.

Edit: Just saw your comment. You need to find a good way to hash condition for this to work, so unless it derives from a class that implements __hash__, you need to implement it yourself.

Edit 2: Using __new__ instead of __init__ for the singleton.