npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

laihui

v0.0.1

Published

A composable, middleware-driven HTTP client built on the Fetch API

Readme

laihui

A composable, middleware-driven HTTP client built on the Fetch API.

import { createFetch } from 'laihui'

const api = createFetch('https://api.example.com')

// Middleware pipeline (onion model)
api.use(async (ctx, next) => {
  console.time(ctx.request.url)
  await next()
  console.timeEnd(ctx.request.url)
})

// GET with type inference
const user = await api.get('/users/1').json<User>()

// POST with JSON body
const post = await api.post('/posts', {
  body: { title: 'Hello', content: 'World' },
}).json()

// Extend instance
const admin = api.extend({
  headers: { Authorization: 'Bearer token' },
})

Features

  • Middleware pipeline — Koa-style onion model, not flat hooks
  • Zero dependencies in core
  • Universal — Browser, Node 20+, Bun, Deno, Cloudflare Workers, Edge Runtime
  • TypeScript — Strict mode, generics on .json<T>()
  • Tree-shakable — ESM only, optional features via sub-path imports
  • Immutableextend() never mutates

Install

pnpm add laihui

Usage

Basic

const api = createFetch('https://api.example.com')

const data = await api.get('/users').json<User>()

Headers

const api = createFetch('https://api.example.com').extend({
  headers: { Authorization: 'Bearer token' },
})

await api.get('/users', {
  headers: { 'X-Request-ID': 'abc' },
})

Methods

api.get(url, options?)
api.post(url, options?)
api.put(url, options?)
api.patch(url, options?)
api.delete(url, options?)
api.head(url, options?)
api.options(url, options?)
api.trace(url, options?)

Response

const res = api.get('/users')
await res              // → Response
await res.json<T>()    // → T
await res.text()       // → string
await res.blob()       // → Blob
await res.bytes()      // → Uint8Array
await res.stream()     // → ReadableStream

Middleware

Use .use(fn) where fn is (ctx, next) => Promise<void>:

api.use(async (ctx, next) => {
  console.log('→', ctx.request.method, ctx.request.url)
  await next()
  console.log('←', ctx.response?.status)
})

Share state between middlewares via ctx.state:

api.use(async (ctx, next) => {
  ctx.state.user = await getUser()
  await next()
})

Built-in Middleware

Timeout

import { timeout } from 'laihui/middleware/timeout'

api.use(timeout({ timeout: 5000 }))

Retry

import { retry } from 'laihui/middleware/retry'

api.use(retry({
  limit: 3,
  delay: 1000,
  backoff: 'exponential',
  jitter: true,
  statusCodes: [408, 413, 429, 500, 502, 503, 504],
}))

Deduplication

import { dedupe } from 'laihui/middleware/dedupe'

api.use(dedupe())

Auth / Token Refresh

import { auth } from 'laihui/middleware/auth-refresh'

api.use(auth({
  getToken: () => localStorage.getItem('token'),
  refreshToken: async () => {
    const newToken = await fetch('/refresh').then(r => r.text())
    localStorage.setItem('token', newToken)
    return newToken
  },
}))

Error Handling

import { HTTPError, TimeoutError, RetryError } from 'laihui'

try {
  await api.get('/users').json()
} catch (error) {
  if (error instanceof HTTPError) {
    console.error(error.response.status, error.response.statusText)
  } else if (error instanceof TimeoutError) {
    console.error('Request timed out')
  } else if (error instanceof RetryError) {
    console.error('Retry failed after', error.attempts, 'attempts')
  }
}

Extend

const api = createFetch('https://api.example.com')
const admin = api.extend({
  headers: { Authorization: 'Bearer admin' },
})
// api is not mutated — immutable

Pagination

import 'laihui/patch/pagination'

interface Page {
  id: number
  title: string
}

const api = createFetch('https://api.example.com')

const pages = api.pagination<Page>('/api/pages', {
  mode: 'page',        // 'page' | 'cursor' | 'offset'
  size: 20,
  params: { status: 'published' },
})

const p1 = await pages.next()
// { items: Page[], total: 100, next: true, prev: false, page: 1, size: 20 }

const p2 = await pages.next()    // page 2
const back = await pages.prev()   // page 1
const jump = await pages.index(5) // page 5 (page mode only)
const again = await pages.reload()

Polling

import 'laihui/patch/poll'

const poll = api.poll<{ status: string }>('/jobs/1', {
  interval: 2000,
  condition: (data) => data.status === 'completed',
})

poll.on('data', (data) => console.log(data))
poll.on('done', (result) => console.log('Finished', result))
poll.on('error', (err) => console.error(err))

// Manual fetch
const result = await poll.next()

// Auto-polling
poll.start()
poll.stop()

Download

import 'laihui/patch/download'

const task = api.download('/files/report.pdf', {
  onProgress: ({ transferred, total, percent }) => {
    console.log(`${Math.round(percent * 100)}%`)
  },
})

// Awaitable — resolves to Uint8Array (Node) or Blob (browser)
const data = await task

// Or with lifecycle controls
task.on('progress', (p) => console.log(p.percent))
task.on('done', (data) => console.log('done', data))
task.on('error', (err) => console.error(err))
task.pause()      // aborts current request, keeps downloaded bytes
task.resume()     // resumes via Range / If-Range from where it stopped
task.abort()      // cancels

Downloads run through the middleware pipeline. Pause/resume use HTTP Range requests (with If-Range validation); if the server ignores Range and returns 200, the download restarts from scratch.

Upload

import 'laihui/patch/upload'

const task = api.upload('/files', file, {
  onProgress: ({ transferred, total, percent }) => {
    console.log(`${Math.round(percent * 100)}%`)
  },
})

const response = await task  // resolves to Response

task.on('progress', (p) => console.log(p.percent))
task.abort()

Transport selection (transport: 'auto' | 'fetch' | 'xhr', default 'auto'):

  • 'auto' — feature-detects support for request streams. If supported, uploads with progress via fetch + duplex: 'half'. If not supported and an XHR fallback is provided (browser only), uses XHR. Otherwise it warns and degrades to a progress-less upload.
  • 'fetch' — force fetch streaming.
  • 'xhr' — force XHR (requires the optional fallback transport).

The XHR transport is an opt-in module to keep the core pure-fetch and tree-shakable:

import { xhrTransport } from 'laihui/patch/upload/xhr'

api.upload('/files', file, {
  transport: 'auto',
  fallback: xhrTransport,   // enables progress in HTTP/1.1 / Safari
  onProgress: (p) => console.log(p.percent),
})

Resource (CRUD)

import 'laihui/patch/resource'

const users = api.resource('/users')

// GET /users
const list = await users.list().json()

// GET /users/42
const user = await users.get(42).json()

// POST /users
const created = await users.create({ name: 'Alice' }).json()

// PUT /users/7
const updated = await users.update(7, { name: 'Bob' }).json()

// DELETE /users/99
await users.remove(99)

API

createFetch(baseURL?: string)

Returns a FetchInstance.

FetchInstance

| Member | Description | |--------|-------------| | .get(url, options?) | GET request | | .post(url, options?) | POST request | | .put(url, options?) | PUT request | | .patch(url, options?) | PATCH request | | .delete(url, options?) | DELETE request | | .head(url, options?) | HEAD request | | .options(url, options?) | OPTIONS request | | .trace(url, options?) | TRACE request | | .use(middleware) | Register a middleware function | | .extend(options) | Create a new immutable instance | | .pagination(url, config?, options?) | Pagination helper (requires import 'laihui/patch/pagination') | | .poll(url, config?) | Polling helper (requires import 'laihui/patch/poll') | | .resource(url, config?) | CRUD helper (requires import 'laihui/patch/resource') | | .download(url, config?) | Download task helper (requires import 'laihui/patch/download') | | .upload(url, file, config?) | Upload task helper (requires import 'laihui/patch/upload') |

FetchOptions

Extends RequestInit (omitting signal, body).

| Option | Type | Description | |--------|------|-------------| | baseURL | string | Base URL for relative URL resolution | | headers | HeadersInit | Default request headers | | json | boolean | Default true. When true, body is treated as a JsonValue, automatically JSON.stringify-ed and sets Content-Type: application/json. When false, body is treated as raw BodyInit \| null | | body | JsonValue \| BodyInit \| null | Request body. Type is narrowed based on json: when json is true (default), expected as JsonValue; when json: false, expected as BodyInit \| null | | signal | AbortSignal \| AbortSignal[] | Single or multiple abort signals (merged automatically) | | params | URLSearchParamsInit (string \| URLSearchParams \| Iterable<[string, string \| number \| boolean]> \| Record<string, string \| number \| boolean>) | Query parameters appended to URL (serialized via URLSearchParams; a raw string is used as-is). For complex serialization, pass the final value yourself, e.g. params: qs.stringify(obj) / new URLSearchParams(obj) / new Map([...]) |

FetchResponse

Thenable (awaitable to Response directly) with typed helpers:

interface FetchResponse extends PromiseLike<Response> {
  json: <T = unknown>() => Promise<T>
  text: () => Promise<string>
  blob: () => Promise<Blob>
  stream: () => Promise<ReadableStream<Uint8Array>>
  bytes: () => Promise<Uint8Array>
  clone: () => Promise<Response>
}

The response body can only be consumed once (same as the native Fetch API). To read it multiple times, clone before reading:

const res = await api.get('/users').clone()
const a = await res.json()
const b = await res.clone().text()

Error Classes

| Class | Extends | Description | |-------|---------|-------------| | FetchError | Error | Base error with .request and .cause | | HTTPError | FetchError | Non-ok response, has .response | | AbortError | FetchError | Request was aborted | | TimeoutError | AbortError | Request timed out | | RetryError | FetchError | Retries exhausted, has .attempts | | NetworkError | FetchError | Network failure, has optional .code |

License

MIT