float sub-class to alter intake and __str__ behaviour

H Doucet :

I've sub-classed float to alter its __str__() method to end with a symbol (in my case €).

The input is filtered to remove a symbol (in my case €).

class Euro(float):
    def __new__(cls, value):
        v = ''.join([_ for  _ in value if _ != '€' ])
        return super(Euro, cls).__new__(cls, v)

    def __str__(self):
        return f'{self} €'

But when I print, I get a recursive print error.

g = Euro('123.23 €')
print (g) 

Error:

Euro.py, line  __str__
    return f'{self} €'

 [Previous line repeated 329 more times]

RecursionError: maximum recursion depth exceeded

chepner :

Don't use inheritance; a Euro is not a kind of float, and money should never be represented using floats due to the imprecision of the floating-point approximations of real numbers.

Instead, use composition to store an attribute representing the number of Euros, using something like decimal.Decimal to represent the euros-and-cents exactly.

from decimal import Decimal


class Euro:
    # In accordance with how to use super properly,
    # always accept and pass on unrecognized keyword arguments.
    def __init__(self, value, **kwargs):
        super().__init__(**kwargs)

        self.value = Decimal(value.strip('€'))

    def __str__(self):
        return f'{self.value} €'