@octokit/routes

Octokit routes

machine-readable, always up-to-date GitHub REST API route specifications

Downloads

• All routes: octokit.github.io/routes/index.json
• All routes for a scope, e.g. repos: octokit.github.io/routes/routes/repos.json
• A single route, e.g. GET /repos/:owner/:repo: octokit.github.io/routes/routes/repos/get.json

Example

Example route definition

{
  "name": "Lock an issue",
  "enabledForApps": true,
  "method": "PUT",
  "path": "/repos/:owner/:repo/issues/:number/lock",
  "params": [
    {
      "name": "lock_reason",
      "type": "string",
      "description": "The reason for locking the issue or pull request conversation. Lock will fail if you don't use one of these reasons:  \n\\* `off-topic`  \n\\* `too heated`  \n\\* `resolved`  \n\\* `spam`",
      "required": false
    }
  ],
  "description": "Users with push access can lock an issue or pull request's conversation.\n\nNote that, if you choose not to pass any parameters, you'll need to set `Content-Length` to zero when calling out to this endpoint. For more information, see \"[HTTP verbs](/v3/#http-verbs).\"",
  "documentationUrl": "https://developer.github.com/v3/issues/#lock-an-issue"
}

Usage as Node module

const ROUTES = require('@octokit/routes')

returns an object with keys being the route scopes such as activity, issues, repositories, etc (one for navigation header in the sidebar at https://developer.github.com/v3/).

The value for each scope is an array of REST API endpoint specification.

If you don’t need all routes definitions, you can require a scope or a specific route definition instead

const REPO_ROUTES = require('@octokit/routes/routes/repos')
const GET_REPO_ROUTE = require('@octokit/routes/routes/repos/get')

How it works

This package updates itself using a daily cronjob running on Travis. All routes/*.json files as well as index.js is generated by bin/octokit-rest-routes.js. Run node bin/octokit-rest-routes.js --usage for instructions.

The update script is scraping GitHub’s REST API documentation pages and extracts the meta information using cheerio and a ton of regular expressions :)

For simpler local testing and tracking of changes all loaded pages are cached in the cache/ folder.

1. Find documentation pages

• Index page cached in cache/v3/index.html
• Result cached in cache/pages.json

Opens https://developer.github.com/v3/, find all documentation page URLs in the side bar navigation.

2. On each documentation page, finds sections

• Documentation pages cached in cache/v3/*/index.html, e.g. cache/v3/repos/index.html
• Documentation sub pages cached in cache/v3/*/*/index.html, e.g. cache/v3/repos/branches/index.html
• All sections found on pages are cached next to the index.html files, e.g. cache/v3/repos/sections.json
• HTML of sections are stored next to the index.html files, e.g. cache/v3/repos/create.html

Loads HTML of each documentation page, finds sections in page.

3. In each section, finds endpoints

• Some sections don’t have endpoints, such as Notifications Reasons
• Some sections have multiple endpoints, see #3

Loads HTML of documentation page section. Turns it into routes/*.json file. In some cases the HTML cannot be turned into an endpoint using the implemented patterns. For these cases custom overrides are defined.

LICENSE

MIT