pycaches 0.0.7

Python caching made easy

A bunch of caches.

Features

✓ Ease of use cache decorator

✓ Support for non-Hashable keys (dictionaries, lists, sets)

✓ Different cache replacement policies (random, LRU)

✓ Time-based item expiration

□ Cache hit/miss statistics

□ Rich configuration, sane defaults

□ Optional persistency

Installation

Recommended way is to use poetry:

poetry install pycaches

But you also can install library with pip:

pip install pycaches

Usage

cache decorator

from pycaches import cache


@cache()
def example():
    print("Hi, I will be called once!")


example()  # Prints "Hi, I will be called once!"
example()  # Is not called

import time

from pycaches import cache


@cache()
def long_computation(x):
    print("Performing long computation...")
    time.sleep(1)
    return x + 1


long_computation(5)  # Sleeps for 1 second and returns 6
long_computation(5)  # Immediately returns 6

long_computation(6)  # Sleeps for 1 second and returns 7
long_computation(6)  # Immediately returns 7
long_computation(6)  # And again

Cache class

import time
from datetime import timedelta

from pycaches import Cache


cache = Cache()
cache.save("a", 1)
cache.save("b", 2)
cache.save("c", 3, expire_in=timedelta(seconds=10))

cache.has("c")  # returns True
cache.get("a")  # returns 1

time.sleep(10)
cache.has("c")  # False
cache.get("c")  # raises KeyError

Different cache replacement policies

from pycaches import Cache
from pycaches.policies import LRU

"""
LRU stands for Least Recently Used.
So LRU policy removes Least Recently Used item from cache
if it's size exceed max_items.
"""


cache = Cache(max_items=2, replacement_policy=LRU())
cache.save("a", 1)
cache.save("b", 2)
cache.save("c", 3)

cache.has("a")  # returns False
cache.has("b")  # returns True

cache.save("d", 4)

cache.has("b")  # returns False

Disable deepcopy for keys

from pycaches import cache

"""
Cache class and cache decorator accepts `copy_keys` argument.
If you can garantee that keys will not change even if they are mutable,
you may set it to `True` to speed things up.
"""


@cache(copy_keys=False)
def faster_caching(x):
    return x


faster_caching({1, 2, 3})  # returns {1, 2, 3}

Contribution

Just clone repository, make your changes and create a pull request.

Do not forget to make sure code quality is high: run linters, typecheckers, check code coverage, etc. You can do it all with make:

1. make lint: pylint and pycodestyle
2. make typecheck: mypy
3. make test: pytest
4. make coverage: pytest with pytest-cov
5. make quality: radon
6. make build: setup.py

And just make or make all to run all these targets.