Python integer particularities in CPython

pyints.py

In [1]: a = 1

In [2]: b = 1

In [3]: a is b
Out[3]: True

In [4]: id(a)
Out[4]: 140720033629792

In [5]: id(a) == id(b) == id(1)
Out[5]: True

In [6]: # Let's check 0:

In [7]: a = 0

In [8]: b = 0

In [9]: id(a)
Out[9]: 140720033629760

In [11]: id(a) == id(b) == id(0)
Out[11]: True

In [12]: # Check further:

In [13]: c=256

In [14]: id (c)
Out[14]: 140720033637952

In [15]: id(c)-id(a) # That particular number..??
Out[15]: 8192

In [16]: # Mmmm.. wait a minute!

In [17]: import sys

In [18]: print(sys.int_info)
sys.int_info(bits_per_digit=30, sizeof_digit=4)

In [19]: # sizeof_digit = 4 means 4 bytes, so 32 bits!
    ...: # I get it!, now:
    ...: (id(c)-id(a))/(sys.int_info.sizeof_digit*8) # 8 bits = 1 byte
Out[19]: 256.0

In [20]: # WAAAA!. Check one more:
    ...: c = 257

In [21]: d = 257

In [22]: c is d
Out[22]: False

In [23]: id(c)
Out[23]: 140719783347216

In [24]: id(c) == id(d)
Out[24]: False

In [25]: (id(c)-id(a))/(sys.int_info.sizeof_digit*8) # 8 bits = 1 byte
Out[25]: -7821329.5

In [26]: #Uh... ok, not what anyone expect right?
    ...: #So, from -5 to 256 values are
    ...: #  preloaded in a contiguous table in memory (Integer singletons)
    ...: #This is done by implementation in CPython
    ...: #  for optimization purposes.

In [33]: # You can check more of this here:
    ...:  # For ints:
    ...: #  https://docs.python.org/2/c-api/int.html#c.PyInt_FromLong
    ...: #  https://docs.python.org/3.7/c-api/long.html?highlight=pylong#c.PyLong_FromLong