python-cachelock

The cachelock serves to ensure that your code block or function is executed one at a time using the cache as acquirer.

How to install:

pip install cachelock

or

pipenv install cachelock

How to use

You can use of two forms:

- Use the `Lock` class to your blocks;
- Or use the decorator `once` to block your functions or tasks;

With cachelock.Lock

Arguments:

with cachelock.Lock(
    key=...
    cache=...,
) ...

You should use the Lock class as with context:

with cachelock.Lock(key='foo'):
    ...your code ...

If it is already locked, the LockError will be raised.

With cachelock.once

Arguments:

@cachelock.once(
    key=...
    cache=...,
    raises_if_lock=False
)
...

With the decorator, you can guarantee the unique execution of some function. It also allows you to configure your key according to the arguments of the function.

@celery.task
@cachelock.once(key='foo-{arg_a}-{arg_b}')
def func(arg_a, arg_b):
    pass

That way, if it is locked, the function quits silently. You can also ask once to raise LockError:

@celery.task
@cachelock.once(
    key='foo-{arg_a}-{arg_b}',
    raise_if_lock=True
)
def func(arg_a, arg_b):
    pass

Integrate with Django

The cachelock checks whether django is installed, and if so, it uses django's own cache as aquirer. And if you wish, you can modify the default cache with DEFAULT_CACHELOCK_ALIAS configuration through django settings. The value must be an alias of some existing cache, by default it uses default

Remembering that by default, cachelock uses its own internal cache in memory if there is no integration with django.

Customizing your own cache

To work it is necessary that the implementation of the cache has the get, delete and set methods. Ex.:

class DummyCache:

    def set(self, key, value):
        pass

    def get(self, key, default=None):
        pass

    def delete(self, key):
        pass

cache = DummyCache()

use:

@cachelock.once(key=..., cache=cache)

or

with cachelock.Lock(key..., cache=cache) ...