LibHub

Libraries.io minimalistic GitHub client.

Features

• Dead simple API
• Promise based
• Support for caching + conditional requests (optional)
• Support for pagination using link headers

Requirements

• Node.js v4.2.x

Getting Started

  const Client = require('libhub')
  let github = new Client(token)

  // GET
  github.get(`/repos/librariesio/libhub/issues`)
  .then( (issues) => {
    console.log(issues)
  })

  // POST
  github.post(`/repos/librariesio/libhub/issues`, {}, issueBody)
  .then( (issue) => {
    console.log(issue)
  })

Available methods: get, post, patch, put, delete

Error handling

  // 3xx/4xx/5xx errors
  github.get(`/repos/librariesio/${invalidId}`)
  .then( (repo) => {
    // This won't happen
  })
  .catch( (err) => {
    console.log(err.statusCode) // 404
  })

Caching

The Client constructor takes an optional cache object that should have a get and a set method. Both should return a Promise.

  const Client = require('libhub')
  const InMemoryCache = require('./test/cache')
  let github = new Client(token, { cache: InMemoryCache })

  github.get(`/repos/librariesio/libhub/issues`)
  .then( (issues) => {
    console.log(issues)
  })

Pagination

  github.get(`/events`, { allPages: true })
  .then( (events) => {
    console.log(events.length) // All the user public events ~300
  })

Why another GitHub client?

Well, simplicity. Most GitHub clients I used implement unnecessary abstractions. For example, octonode. This means that you need to check GitHub API documentation and Octonode documentation to perform any action. This gets boring pretty quickly.

With LibHub, you just need to lookup the resource URL and use it. Also, ES6 interpolation.

  github.get(`/repos/${owner}/${repo}/branches/${branch}`)

License

MIT