🐍

Create Object In Python

Table of contents

new vs init
What is Interning in python?

new vs init

new is the first step of instance creation. It's called first and is responsible for returning a new instance of your class. In contrast, init doesn't return anything; it's only responsible for initializing the instance after it's been created.

Python

class a:
    def __new__(cls, *args, **kwargs):
        print(cls.__name__)
        print(args)
        print(kwargs)
        obj = object.__new__(cls)
        print(id(obj))
        return obj
    
    def __init__(self):
        print(id(self))
m = a()
l = a(123,123)

Output

a
()
{}
4381665568
4381665568
/////////
a
(123, 123)
{}
4381668976
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: a.__init__() takes 1 positional argument but 3 were given

What is Interning in python?

Re-using objects of equal value on-demand instead of creating new objects.

We can use this technique by overriding new. Now we compare memory consumption in normal implementation and in using intern technique.

Normal implementation:

python

class Check:
    def __init__(self, file, rank) -> None:
        if file not in range(1,9) :
            raise ValueError('out of range')
        if rank not in 'abcdefgh': 
            raise ValueError('out of range')
        
        self._file = file
        self._rank = rank
def start():
    return {
        '1' : Check(1, 'a'),
        '2' : Check(2, 'a'),
        '3' : Check(3, 'a'),
        '4' : Check(4, 'a'),
        '5' : Check(5, 'a'),
        '6' : Check(6, 'a'),
        '7' : Check(7, 'a'),
        '8' : Check(8, 'a'),
        '9' : Check(1, 'b'),
        '10' : Check(2, 'b'),
        '12' : Check(3, 'b'),
        '13' : Check(4, 'b'),
        '14' : Check(5, 'b'),
        '15' : Check(6, 'b'),
        '16' : Check(7, 'b'),
        '17' : Check(8, 'b'),
        '18' : Check(1, 'c'),
        '19' : Check(2, 'c'),
    }
def main():
    import tracemalloc
    tracemalloc.start()
    board = [start() for i in range(10000)]
    _, peak = tracemalloc.get_traced_memory()
    print(peak/1000, 'kb')
if __name__ == '__main__':
    main()

output

33845.742 kb

Using interning technique and override new function:

python

class Check:
    _intern = {}
    def __new__(cls, file, rank):
        if file not in range(1,9) :
            raise ValueError('out of range')
        if rank not in 'abcdefgh': 
            raise ValueError('out of range')
        key = (file, rank)
        if key not in cls._intern:
            obj = object.__new__(cls)
            obj._file = file
            obj._rank = rank
            cls._intern[key] = obj
        return cls._intern[key]
def start():
    return {
        '1' : Check(1, 'a'),
        '2' : Check(2, 'a'),
        '3' : Check(3, 'a'),
        '4' : Check(4, 'a'),
        '5' : Check(5, 'a'),
        '6' : Check(6, 'a'),
        '7' : Check(7, 'a'),
        '8' : Check(8, 'a'),
        '9' : Check(1, 'b'),
        '10' : Check(2, 'b'),
        '12' : Check(3, 'b'),
        '13' : Check(4, 'b'),
        '14' : Check(5, 'b'),
        '15' : Check(6, 'b'),
        '16' : Check(7, 'b'),
        '17' : Check(8, 'b'),
        '18' : Check(1, 'c'),
        '19' : Check(2, 'c'),
    }
def main():
    import tracemalloc
    tracemalloc.start()
    board = [start() for i in range(10000)]
    _, peak = tracemalloc.get_traced_memory()
    print(peak/1000, 'kb')
if __name__ == '__main__':
    main()

output

6489.795 kb

with this technique we can reduce our memory consumption.

Edit this page