Skip to content
Create a gist now

Instantly share code, notes, and snippets.

Embed URL


Subversion checkout URL

You can clone with
Download ZIP
A CachedProperty descriptor; It receives an expensive getter function (& optional arguments for it). The __get__ runs the getter function with the object instance as the first argument (& optional arguments provided in __init__) and caches the result. The __set__ just changes the cache to the 'value', __delete__ removes the cache.
class CachedProperty(object):
A cached property; A bit specific, as
it runs the getter as an instance method
& forwards the instance & init args to it.
Uses python descriptors protocol
def __init__(self, getter, *args):
self.getter = getter
self.args = args or None
def __get__(self, instance, owner):
# If we don't have cache, go ahead and call the getter & cache its result
if not hasattr(self, '_cache'):
# This is the expensive getter call; The getter is provided with instance & optional args
result = self.getter(instance, *self.args) if self.args is not None else self.getter(instance)
# Cache the result in self
setattr(self, '_cache', result)
# We always return our cache
return getattr(self, '_cache')
def __set__(self, instance, value):
# Set the new value to the cache
setattr(self, '_cache', value)
def __delete__(self, instance):
# Remove cache
delattr(self, '_cache')
class SomeClass(object):
A, B, C = 1, 2, 3
def __init__(self, *args, **kwargs):
self.args = args or ()
self.kwargs = kwargs or {}
def has_arg(self, arg):
return arg in self.args
def expensive_getter_fn(x)
# An expensive computation here
return x
# Typical usage case: expensive getter & some args for it to receive
my_cached_prop_a = CachedProperty(expensive_getter_fn, SomeClass.A)
my_cached_prop_b = CachedProperty(expensive_getter_fn, SomeClass.B)
my_cached_prop_c = CachedProperty(expensive_getter_fn, SomeClass.C)
# Another way to use it: lambda getter that invokes some instance method
my_cached_prop_d = CachedProperty(lambda instance, arg: instance.has_arg, "some_arg")
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.