memento.redis — org.clojars.roklenarcic/memento-redis 1.0.24

Settings for Redis cache.

Besides there, the cache also supports these settings from base Memento:

• key-fn
• ret-fn
• evt-fn
• ttl
• fade
• ret-ex-fn

anon-key clj

Specifies how Memento Segment ID is processed by default KeysGenerator, when ID is a function object. Available settings: :empty, :stringify, defaults to :empty.

If :empty, then any function object ID is replaced by empty string. If :stringify, then any function object ID is replaced by (str function)

cache clj

Value for :memento.core/type setting for redis cache

conn clj

Redis connection setting. It can be:

• a Carmine connection map
• a IDeref object (Var, Atom) which will contain a Carmine connection map, it will be derefed at each use.
• a 0-arg fn that will return a Carmine connection map, it will be evaluated at each use

2 of 3 here will be evaluated each lookup (or other operation).

hit-detect? clj

Cache setting, if set to true, any returned IObj that was cached in Redis will have meta key :memento.redis/cached? with value true

keygen clj

Specify the way the keys are generated. Value is an instance of KeysGenerator protocol.

If not specified, the default keys generator will be used.

name clj

Cache name. It is a part of Redis keys' names.

It is useful if you want to cache the same function in two different caches and have them not collide.

It defaults to empty string.

nuke!! clj

(nuke!! memoized-function)

(nuke!! conn keys-generator)

Remove all entries created by Memento.

The parameter is a memoized function.

The actual database being nuked (connection and database) and key naming schema (KeysGenerator) is determined based on the settings on the memoized function. For most applications pointing this at any memoized function will have the same result, but if you're using multiple different databases or KeysGenerator instances, then only the corresponding database/keys schema is wiped.

sha1 clj

(sha1 o)

Function that returns a byte array with SHA1 hash of the Nippy serialized form of the object. Used in conjunction with key-fn to shorten long keys.

Since serialized function arguments are part of the Redis keys, this can lead to functions with large arguments using up a lot of RAM in Redis just due to keys.

You can write a key-fn that returns a hash of the actual function arguments (or a particular long argument).

This is a convenience function to achieve that.

Note that nippy already does lz4 compression on keys over 8kb.

sha256 clj

(sha256 o)

Function that returns a byte array with SHA-256 hash of the Nippy serialized form of the object. Used in conjunction with key-fn to shorten long keys.

Since serialized function arguments are part of the Redis keys, this can lead to functions with large arguments using up a lot of RAM in Redis just due to keys.

You can write a key-fn that returns a hash of the actual function arguments (or a particular long argument).

This is a convenience function to achieve that.

Note that nippy already does lz4 compression on keys over 8kb.

ttl-fn clj

Cache and function bind setting. Value is a function (fn [mount-config key value]). It allows that you specify a function that decides TTL on per entry basis. It should return a duration, if nothing is returned, the ttl/fade setting is used.