github-basic

Basic https interface to GitHub. Intended so it's easy to code using the official documentation. It also includes helpers for streaming paged results, and doing all the necessary actions to automatically submit pull requests.

Installation

npm install github-basic

API

github(options)

var github = require('github-basic');
var client = github({version: 3});

Create a new GitHub client with a set of options:

• version: (required) you should set this to the version number of the API you want to make requests against (e.g. {version: 3})
• auth: (default: null) '<my oauth token>' or {username: 'my user', password: 'my password', otp: 'my optional 2 factor authentication code'} to authenticate your requests
• cache: (default: null) proivde a caching mechanism for "get" requests. Can be 'memory', 'file' or a custom cache (only "file" will work with the "sync" option)
• sync: (default: false) set this to true and you will get a fully synchronous GitHub client, just use the results of functions directly, no need for promises or callbacks.

client.method(path, query[, callback])

• client.get(path, query[, callback])
• client.head(path, query[, callback])
• client.delete(path, query[, callback])
• client.post(path, query[, callback])
• client.patch(path, query[, callback])
• client.put(path, query[, callback])

Make an API request and parse the response as JSON. For get, head and delete requests, the query will be added as a query string to the end of the url. For other methods, it will be used as a body. You can optionally specify that parts of the query should instead be used as part of the url by adding :qs-name segments to the url. e.g.

client.get('/users/:user/gists', {user: 'ForbesLindesay'});

If the request is part of a paged request, you can use res.getNext(), res.getPrev(), res.getLast() and res.getFirst() to request other pages. Note that not all of these methods will always exist (e.g. there is no res.getNext() if you already have the last page).

client.methodBuffer(path, query[, callback])

• client.getBuffer(path, query[, callback])
• client.headBuffer(path, query[, callback])
• client.deleteBuffer(path, query[, callback])
• client.postBuffer(path, query[, callback])
• client.patchBuffer(path, query[, callback])
• client.putBuffer(path, query[, callback])

Make an API request and return the response as a Buffer. For get, head and delete requests, the query will be added as a query string to the end of the url. For other methods, it will be used as a body. You can optionally specify that parts of the query should instead be used as part of the url by adding :qs-name segments to the url. e.g.

client.getBuffer('/users/:user/gists', {user: 'ForbesLindesay'});

client.exists(user, repo[, callback])

Returns true if :user/:repo exists, and false if requesting the repo url returns an error.

client.fork(user, repo, options[, callback])

Forks the repo github.com/:user/:repo to the authenticated user and waits until the fork operation completes. To fork to an organization, just pass an organization string in the options object.

N.B. forking will currently appear successful even if the target repo already exists.

client.branch(user, repo, from, to[, callback])

Creates a new branch in github.com/:user/:repo using from as the source branch and to as the new branch name.

client.commit(user, repo, commit, options[, callback])

Commits a set of changes to github.com/:user/:repo. It only supports updating text files.

commit:

An object with:

FileUpdate:

An object with:

options:

An (optonal) object with:

client.pull(from, to, message [, callback])

Creates a pull request from from to to.

from:

An object with:

to:

An object with:

message:

Either:

or:

client.getStream(path, query)

Only works in async mode (default)

Sometimes the easiest way to handle GitHub's paginated results is to treat them as a stream. This method isn't (currently) clever enough to do streaming JSON parsing of the response, but it will keep requesting more pages as needed and it works properly with back pressure so as to not request more pages than are needed:

var github = require('github-basic')
var Stringifier = require('newline-json').Stringifier
 
var client = github({version: 3})
 
//stream all of ForbesLindesay's repos
client.getStream('/users/:user/repos', {user: 'ForbesLindesay'})
  .pipe(new Stringifier())
  .pipe(process.stdout)

Usage

var github = require('github-basic')
var client = github({version: 3})
 
//get all 'ForbesLindesay's gists in the last year
 
var since = new Date()
since.setUTCFullYear(since.getUTCFullYear() - 1)
 
// using callbacks
 
client.get('/users/:user/gists', {user: 'ForbesLindesay', since: since}, function (err, res) {
  if (err) throw err;
  console.dir(res)
})
 
// or
 
client.get('/users/ForbesLindesay/gists', {since: since}, function (err, res) {
  if (err) throw err;
  console.dir(res)
})
 
// using promises
 
client.get('/users/:user/gists', {user: 'ForbesLindesay', since: since}).done(function (res) {
  console.dir(res)
})
 
//or
 
client.get( '/users/ForbesLindesay/gists', {since: since}).done(function (res) {
  console.dir(res)
})
 
// getting raw github data
 
client.getBuffer('https://raw.githubusercontent.com/:owner/:repo/master/README.md', {
  owner: 'ForbesLindesay',
  repo: 'github-basic'
}).done(function (res) {
  process.stdout.write(res)
})

License

MIT