Re: Custom 'Float' class. Am I right here?

[Terry Reedy]

Prashant Saxena wrote:

import sys

class Float(float):
"""
Custom float datatype with addtional attributes.
"""

def __new__(self,
value=0.0, #default value
name='', # string
& nbsp; range=(0.0, 1.0) # tuple
)

try:
self.name = name
self.range = range
self.pos = (1, 1)
return float.__new__(s elf, value)
except:
print ('Invalid value : Float = %s %s' % (str(value),
sys.exc_info()[:2]))
def setpos(self, value):
self.pos = value
def getpos(self):
return self.pos*2.0
p = property(getpos , setpos)

myFloat = Float(0.23)
print myFloat.pos

I am trying to create a custom 'float' datatype by subtyping default
'float'. There are few things I would like to know here:

1. How to make 'name' & 'range' only readable attributes.

A property with a working get and set that raises an exception.

2. The property 'p' is not working as it has to. What's wrong here?

Don't know.

3. I would be heavily instancing this class, and I won't be creating any
new attribute on instance. Is there any way to optimize? __slots__?

I believe this is what slots is for.