Yeezy Cache provides simple, light, function result caching (memoization). By wrapping a function in the cache
decorator you can avoid repeatedly reaching out to slower external services (such as a database or external API) or performing intensive processing—assuming your functions are pure or deterministic, of course, and therefore can be cached.
First install Yeezy:
npm install --save yeezy-cache
# or
yarn add yeezy-cache
Next you'll need to configure Yeezy for your use-case (you only need to do this once in your project).
import { InMemoryStorage, configure } from "yeezy-cache"
configure({
storage: new InMemoryStorage(),
expiration: 60
})
The only requirement for memoization is that a function must determine its output solely based on input—that is, the output of a function is the derivative of its input. To do so, simply wrap your function in the cache
decorator.
import { cache } from "yeezy-cache"
const increment(input) => input + 1
export default cache(increment)
That's it! The first time you call your cached version of increment()
the result will be calculated and cached. All subsequent calls, until the expiration, will be retrieved from your Storage.
Some options can be set at the global level. The storage
option informs Yeezy where to store your cached items. The expiration
option sets the default amount of time (in seconds) to cache results.
import { configure, InMemoryStorage } from "yeezy-cache"
configure({
storage: new InMemoryStorage(),
expiration: 60
})
The cache
decorator also has optional parameters that can configure storage
and expiration
on a per-function cache basis.
import { cache } from "yeezy-cache"
import { RedisStorage } from "yeezy-cache-redis"
export default cache({
storage: RedisStorage(),
expiration: 120
})(increment)
Yeezy tries to be as simple as possible to use. However, if you cache two functions with the same name and same number and types of arguments collisions will occur. By default Yeezy hashes the function name and arguments to produce a key which references the function's cached result. Therefore, in these edge-cases, it is required you use the key
option to prevent collisions.
import { cache } from "yeezy-cache"
export default cache({
key: 'plus-one-increment'
})(increment)
Yeezy ships with an InMemoryStorage
object for experimentation, but also includes an interface (IStorage
) for extending to any caching mechanism of choice.