ioredis-cache v2.0.3
ioredis-cache
Compact redis cache using ioredis.
Features
- Store/retrieve/expire cache
- Compatible with redis
incr,hincrby,... - Cache multiple objects at once
- Delete caches by key pattern
- Take advantage of the redis hash map for memory efficiency
Installation
npm install ioredis-cache yarn add ioredis-cacheUsage
const Redis = require("ioredis")
const Cache = require("ioredis-cache")
const cache = new Cache(new Redis())
const value = await cache.cache("key", async () => ({ foo: "bar" }))Notes
Depends on your case, you may want to disconnect redis after using:
cache.redis.quit()APIs
#constructor
constructor(args: Redis | CacheOptions)
You can initialize the cache from an ioredis object or with options.
interface CacheOptions {
redis: object | Redis
parser: Parser
}
interface Parser { // Parse/unparse object when store in redis. Default is JSON
parse: (text: string) => any
stringify: (value: any) => string
}#cache
cache(key: string, callback: async (key: string) => any, ttl?: number): Promise<any>
Get the cached value of the key from redis. If the key does not exist, get the value from the callback, store it in redis and return the value.
If the callback return undefined, it will not be cached. null is cached normally.
If ttl is set, the key will expire after ttl second(s).
const getPost = (id) =>
cache(`post-${id}`, () => db.Post.getPostWithId(id), 60 * 60)#getCache
getCache(key: string): Promise<any>
Get the cached value of the key from redis. If the key does not exist, return undefined.
#setCache
setCache(key: string, value: any, ttl?: number)
Store the value to the cache key.
If ttl is set, the key will expire after ttl second(s).
#manyCache
manyCache(keys: string[], callback: async (ids: string[]) => { [id: string]: any } | any[], prefix?: string, ttl?: number)
Get cached values for a list of keys from redis. The uncached keys will be passed to the callback to get the corresponding values. These values will be stored in redis. Returns combined array of both cached and uncached values.
If prefix is set, it will be preend to the key when get/set in redis. (The keys which are passed to the callback still don't have the prefix)
This is useful if some cached ids were invalidated, you can recache multiple values at once. For example:
// invalidate cache of the changed post
db.Post.onChange(post => cache.deleteCache(`post-${post.id}`))
// this function queries all posts that are uncached at once
const getPosts = (ids) =>
manyCache(ids, async (uncachedIds) => {
const posts = await db.Post.queryAllPostsWithIds(uncachedIds)
return posts.reduce((map, post) => Object.assign(map, { [post.id]: post }), {})
}, prefix: 'post-', 60 * 60)#getManyCache
getManyCache(keys: string[]): Promise<any[]>
Return cached values for a list of keys from redis. If a key does not exist, the value will be undefined.
#setManyCache
setManyCache(keys: string[], ttl?: number)
Store the values with the corresponding keys in redis.
If ttl is set, the keys will expire after ttl second(s).
#deleteCache
deleteCache(...keys: string[]): Promise<number>
Delete the cached keys from redis and return the deleted number
#deletePattern
deletePattern(pattern: string, batch: number = 100)
Scan all keys and delete the keys that matched the pattern:
deletePattern('post:*')Remove all cached keys start with "post:*".
#acquire
acquire(key: string, amount: number, fn: (current: number) => any, float: boolean = false)
Increase the value of key by amount and pass it to fn function. If fn raises any exception, decrease the value of the key by amount.
By default, amount is treate as an integer. Set float = true if amount is a float.
#hashCache
hashCache(key: string, id: string, callback: () => any): Promise<any>
Get the cached id of the key from redis. If the id or key does not exist, get the value from the callback, store it in redis and return the value.
This function is similar to #cache, but stores value in redis hash for better memory effeciency. You can remove the hash by simply deleting the cache key. This is better than using #deletePattern if there are a lot of keys to be deleted. The only downside is you can't set the ttl for each id but only for the hash.
#getHashCache
getHashCache(key: string, id: string): Promise<any>
Return the cached id of the key from redis. If the id / key does not exist, return undefined.
#setHashCache
setHashCache(key: string, value: any)
Store the value to the id of key
#hashManyCache
hashManyCache(key: string, ids: string[], callback: (ids: string[]) => { [id: string]: any } | any[]): Promise<any[]>
Get cached values for the id array of the key from redis. The uncached ids will be passed to the callback to get the corresponding values. These values will be stored in redis. Returns a combined array of both cached and uncached values.
It is similar to #manyCache, but use redis hash to store data.
// invalidate cache of the changed post
db.Post.onChange(post => cache.deleteHashCache('post', post.id))
// this function queries all posts that are uncached at once
const getPosts = (ids) =>
hashManyCache('post', ids, async (uncachedIds) => {
const posts = await db.Post.queryAllPostsWithIds(uncachedIds)
return posts.reduce((map, post) => Object.assign(map, { [post.id]: post }), {})
})#getHashManyCache
getHashManyCache(key: string, ids: string[]): Promise<any[]>
Return cached values for the id array of the key from redis. If the id / key does not exist, the value will be undefined.
#setHashManyCache
setHashManyCache(key: string, valueMap: { [id: string]: any })
Store the values with the corresponding ids of the key in redis
#deleteHashCache
deleteHashCache(key: string, ...ids: string[]): Promise<number>
Delete the cached ids of the key from redis and return the deleted number
#hashAcquire
hashAcquire(key: string, id: string, amount: number, fn: (current: number) => any, float: boolean = false)
Increase the value of id in key by amount and pass it to fn function. If fn raises any exception, decrease the value of id in key by amount.
By default, amount is treated as an integer. Set float = true if amount is a float.
Run tests
Make sure the redis is running at localhost:6379
Run the following command:
yarn test