Configuration

The caches module allows to setup cache configurations and then use them either using an alias or retrieving the config explicitly. To set the config, call caches.set_config:

caches.set_config(config)

Set (override) the default config for cache aliases from a dict-like structure. The structure is the following:

{
    'default': {
        'cache': "aiocache.SimpleMemoryCache",
        'serializer': {
            'class': "aiocache.serializers.StringSerializer"
        }
    },
    'redis_alt': {
        'cache': "aiocache.RedisCache",
        'endpoint': "127.0.0.10",
        'port': 6378,
        'serializer': {
            'class': "aiocache.serializers.PickleSerializer"
        },
        'plugins': [
            {'class': "aiocache.plugins.HitMissRatioPlugin"},
            {'class': "aiocache.plugins.TimingPlugin"}
        ]
    }
}

‘default’ key must always exist when passing a new config. Default configuration is:

{
    'default': {
        'cache': "aiocache.SimpleMemoryCache",
        'serializer': {
            'class': "aiocache.serializers.StringSerializer"
        }
    }
}

You can set your own classes there. The class params accept both str and class types.

All keys in the config are optional, if they are not passed the defaults for the specified class will be used.

If a config key already exists, it will be updated with the new values.

To retrieve a copy of the current config, you can use caches.get_config or caches.get_alias_config for an alias config.

Next snippet shows an example usage:

 1import asyncio
 2
 3from aiocache import caches, Cache
 4from aiocache.serializers import StringSerializer, PickleSerializer
 5
 6caches.set_config({
 7    'default': {
 8        'cache': "aiocache.SimpleMemoryCache",
 9        'serializer': {
10            'class': "aiocache.serializers.StringSerializer"
11        }
12    },
13    'redis_alt': {
14        'cache': "aiocache.RedisCache",
15        'endpoint': "127.0.0.1",
16        'port': 6379,
17        'timeout': 1,
18        'serializer': {
19            'class': "aiocache.serializers.PickleSerializer"
20        },
21        'plugins': [
22            {'class': "aiocache.plugins.HitMissRatioPlugin"},
23            {'class': "aiocache.plugins.TimingPlugin"}
24        ]
25    }
26})
27
28
29async def default_cache():
30    cache = caches.get('default')   # This always returns the same instance
31    await cache.set("key", "value")
32
33    assert await cache.get("key") == "value"
34    assert isinstance(cache, Cache.MEMORY)
35    assert isinstance(cache.serializer, StringSerializer)
36
37
38async def alt_cache():
39    # This generates a new instance every time! You can also use
40    # `caches.create("alt", namespace="test", etc...)` to override extra args
41    cache = caches.create("redis_alt")
42    await cache.set("key", "value")
43
44    assert await cache.get("key") == "value"
45    assert isinstance(cache, Cache.REDIS)
46    assert isinstance(cache.serializer, PickleSerializer)
47    assert len(cache.plugins) == 2
48    assert cache.endpoint == "127.0.0.1"
49    assert cache.timeout == 1
50    assert cache.port == 6379
51    await cache.close()
52
53
54async def test_alias():
55    await default_cache()
56    await alt_cache()
57
58    cache = Cache(Cache.REDIS)
59    await cache.delete("key")
60    await cache.close()
61
62    await caches.get("default").close()
63
64
65if __name__ == "__main__":
66    asyncio.run(test_alias())

When you do caches.get('alias_name'), the cache instance is built lazily the first time. Next accesses will return the same instance. If instead of reusing the same instance, you need a new one every time, use caches.create('alias_name'). One of the advantages of caches.create is that it accepts extra args that then are passed to the cache constructor. This way you can override args like namespace, endpoint, etc.

caches.add(alias: str, config: Dict[str, object]) None

Add a cache to the current config. If the key already exists, it will overwrite it:

>>> caches.add('default', {
        'cache': "aiocache.SimpleMemoryCache",
        'serializer': {
            'class': "aiocache.serializers.StringSerializer"
        }
    })
Parameters:
  • alias – The alias for the cache

  • config – Mapping containing the cache configuration

caches.get(alias: str) object

Retrieve cache identified by alias. Will return always the same instance

If the cache was not instantiated yet, it will do it lazily the first time this is called.

Parameters:

alias – str cache alias

Returns:

cache instance

caches.create(alias: str, **kwargs)

Create a new cache.

You can use kwargs to pass extra parameters to configure the cache.

Parameters:

alias – alias to pull configuration from

Returns:

New cache instance