@trpc-limiter/core

The core of tRPC limiter.

Current Adapters

- Upstash / Redis
- Memory Store

Custom Adapter - Create

``ts
import { defineTRPCLimiter } from '@trpc-limiter/core'
import { MemoryStore } from './store'

export const createTRPCStoreLimiter = defineTRPCLimiter({
store: (opts) => new MemoryStore(opts), // this store will be created when defining the middleware
isBlocked: async (store, fingerPrint, opts) => {
// inferred store from above
const { totalHits, resetTime } = await store.increment(fingerPrint)
if (totalHits > opts.max) {
// this will be inferred and will be used as the first paramter of onLimit
return Math.ceil((resetTime.getTime() - Date.now()) / 1000)
}
return null // if request should not be blocked, return null
},
})
`

Custom Adapter - Use

`ts
import { initTRPC, TRPCError } from '@trpc/server'
import {
createTRPCStoreLimiter,
defaultFingerPrint,
} from '@trpc-limiter/memory'

// trpc context type
type IContext = {
req: Request
}

export const root = initTRPC.context().create()

const limiter = createTRPCStoreLimiter({
fingerprint: (ctx, _input) => defaultFingerPrint(ctx.req), // return the ip from the request
windowMs: 20000,
// hitInfo is inferred from the return type of isBlocked, its a number in this case
message: (hitInfo) => Too many requests, please try again later. ${hitInfo},
max: 15,
onLimit: (hitInfo, _ctx, fingerprint) => {
console.log(hitInfo, fingerprint)
throw new TRPCError({
code: 'INTERNAL_SERVER_ERROR',
message: 'Too many requests unique',
})
},
})

export const rateLimitedProcedure = root.procedure.use(limiter)
`

defaultFingerPrint

The defaultFingerPrint function is exported from every adapter, it basically tries and fetch the IP from the request supporting both NextJS & SolidStart (Fetch).

`ts
import {
createTRPCStoreLimiter,
defaultFingerPrint,
} from '@trpc-limiter/memory'

// or

import {
createTRPCUpstashLimiter,
defaultFingerPrint,
} from '@trpc-limiter/upstash'

// or

import { createTrpcRedisLimiter, defaultFingerPrint } from '@trpc-limiter/redis'
``