Custom backends for cachetools

cachetools is a popular library for Python that provides a number of caching tools. In addition to providing decorators to easily memoize function and method calls it provides implementations for several common cache invalidation algorithms. While the cache decorators can take an arbitrary MutableMapping as the cache backend, the cache classes currently only work with an in-memory dictionary stored as a name-mangled __data attribute. Although there is an issue logged for this, the maintainer doesn’t seem to be interested in changing this. Of course it is always fine for an open source maintainer to choose which issues to address and which to leave as is but I often want to be able to use a cache backed by something other than an in-memory dict.

Luckily there is an easy workaround by consulting the Python documentation on private name mangling. Here is a simple example:

from cachetools import Cache


class MyBackend(dict):
    def __repr__(self):
        return "hi there, I'm a custom backend"


class MyCache(Cache):
    def __init__(self):
        super().__init__(maxsize=10)
        self._Cache__data = MyBackend()


cache = MyCache()
cache["foo"] = "foo"
cache["bar"] = "bar"
print(cache["foo"])
print(cache["bar"])
print(cache)

Running this script results in the following output:

foo
bar
MyCache(hi there, I'm a custom backend, maxsize=10, currsize=2)