Home

Awesome

UserIn 2.0 · License

UserIn is an NodeJS Express middleware to build Authorization Servers that support OAuth 2.0 workflows and integrate with Identity Providers (e.g., Google, Facebook, GitHub). Its openid mode exposes an API that complies to the OpenID Connect specification. With UserIn, the OAuth 2.0/OpenID Connect flows are abstracted so that developers focus only on implementing basic CRUD operations (e.g., get user by username and password, insert token's claims object) using the backend storage of their choice.

To ease testing, UserIn ships with a utility that allows to export a collection.json to Postman.

UserIn is designed to expose web APIs that support two different flow types:

(1) Other OAuth 2.0 authorization flows that do not require a consent page are the password, client_credentials grant type flows. Those flows are generally used for programmatic access.

Table of contents

Getting started

Creating a UserIn Authorization Server consists in creating a UserInStrategy class (which must inherit from the Strategy class) and then registering that class with the UserIn middleware. That UserInStrategy class must implement specific methods based on how many UserIn features must be supported. UserIn replaces the burden of implementing OAuth 2.0 logic with simple CRUD implementations.

Install UserIn:

npm i userin

If you need to support authentication using Facebook, install the Facebook passport (more about other providers in the Setting up an identity provider section):

npm i passport-facebook

const express = require('express')
const app = express()
const { UserIn, Strategy, Postman } = require('userin')
const Facebook = require('passport-facebook')

class YourStrategy extends Strategy {
	constructor(config) {
		super(config)
		this.name = 'yourstrategyname',

		// loginsignup mode (login & signup are not part of the OAuth 2.0 spec)
		// ====================================================================
		// 		Implement those seven methods if you need to support the 'loginsignup' 
		// 		mode (i.e., allowing users to login/signup with their username and password only)
		this.create_end_user = (root, { user }, context) => { /* Implement your logic here */ }
		this.get_end_user = (root, { user }, context) => { /* Implement your logic here */ }
		this.generate_access_token = (root, { claims }, context) => { /* Implement your logic here */ }
		this.generate_refresh_token = (root, { claims }, context) => { /* Implement your logic here */ }
		this.get_refresh_token_claims = (root, { token }, context) => { /* Implement your logic here */ }
		this.get_access_token_claims = (root, { token }, context) => { /* Implement your logic here */ }
		this.delete_refresh_token = (root, { token }, context) => { /* Implement your logic here */ }

		// loginsignupfip mode  (login & signup are not part of the OAuth 2.0 spec, though using an FIP to get an access_token requires interacting with OAuth 2.0. workflows)
		// ===============================================================================================
		// 		Add those four methods to the above seven if you also need to support login and signup with Identity 
		// 		Providers such as Facebook, Google, ...
		this.create_fip_user = (root, { strategy, user }, context) => { /* Implement your logic here */ }
		this.get_fip_user = (root, { strategy, user }, context) => { /* Implement your logic here */ }
		this.generate_authorization_code = (root, { claims }, context) => { /* Implement your logic here */ }
		this.get_authorization_code_claims = (root, { token }, context) => { /* Implement your logic here */ }

		// openid mode (OAuth 2.0 with OpenID support)
		// ===========================================
		// 		Add those thirteen methods to the following eight if you need to support all the OpenID Connect
		// 		APIs which allow third-parties to use your APIs:
		// 			1. 'generate_access_token',
		// 			2. 'generate_authorization_code',
		// 			3. 'generate_refresh_token',
		// 			4. 'get_end_user', 
		// 			5. 'get_authorization_code_claims',
		// 			6. 'get_refresh_token_claims'
		// 			7. 'get_access_token_claims'
		// 			8. 'delete_refresh_token'
		this.get_identity_claims = (root, { user_id, scopes }, context) => { /* Implement your logic here */ }
		this.get_client = (root, { client_id, client_secret }, context) => { /* Implement your logic here */ }
		this.get_id_token_claims = (root, { token }, context) => { /* Implement your logic here */ }
		this.get_auth_request_claims = (root, { token }, context) => { /* Implement your logic here */ }
		this.get_auth_consent_claims = (root, { token }, context) => { /* Implement your logic here */ }
		this.generate_id_token = (root, { claims }, context) => { /* Implement your logic here */ }
		this.generate_auth_request_code = (root, { claims }, context) => { /* Implement your logic here */ }
		this.generate_auth_consent_code = (root, { claims }, context) => { /* Implement your logic here */ }
		this.get_claims_supported = (root) => { /* Implement your logic here */ }
		this.get_scopes_supported = (root) => { /* Implement your logic here */ }
		this.link_client_to_user = (root) => { /* Implement your logic here */ }
		// Those two OpenID event handlers are optional. If they are not implemented, the UserIn middleware uses default
		// values instead:
		// 	For 'get_jwks' UserIn uses an empty array.
		// 	For 'get_grant_types_supported' UserIn uses this array: ['password', 'client_credentials', 'authorization_code', 'refresh_token']
		this.get_jwks = (root) => { /* Implement your logic here */ }
		this.get_grant_types_supported = (root) => { /* Implement your logic here */ }

		// IMPORTANT NOTE: The above event handlers support both synchronous and asynchronous implementations. Both the 
		// following are correct:
		// 		this.generate_access_token = (root, { claims }, context) => { /* Implement your logic here */ }
		// 		or 
		// 		this.generate_access_token = async (root, { claims }, context) => { /* Implement your await logic here */ }
	}
}

const userin = new UserIn({
	Strategy: YourStrategy,
	modes:['loginsignup', 'loginsignupfip', 'openid'], // You have to define at least one of those three values.
	config: {
		baseUrl: 'http://localhost:3330',
		consentPage: 'https://your-domain.com/consent-page', 	// only required when modes contains 'openid'.
		tokenExpiry: {
			access_token: 3600,
			id_token: 3600, 					// only required when modes contains 'openid'.
			code: 30 							// only required when modes contains 'loginsignupfip' or 'openid'.
		}
	}
})

// [Optional] This code requires that an app is registered with 
// Facebook first. For more details about this topic, please refer 
// to the "Setting up an identity provider" section. 
userIn.use(Facebook, {
	scopes: ['public_profile'],
	profileFields: ['id', 'displayName', 'photos', 'email', 'first_name', 'middle_name', 'last_name']
})

// [Optional] Example of how to listen to events and even modify their response. 
userIn.on('generate_access_token', (root, payload, context) => {
	console.log(`'generate_access_token' event fired. Payload:`)
	console.log(payload)
	console.log('Previous handler response:')
	console.log(root)
	console.log('Current context:')
	console.log(context)
})

// [Optional] Exposes an extra 'v1/postman/collection.json' endpoint
// More about this topic in the Integration testing section.
userIn.use(new Postman('userin-my-app'))

app.use(userIn)

app.listen(3330)

The list of exposed endpoints is detailed under the Endpoints section. That list is also discoverable via the following endpoints: All the endpoints that the UserIn middleware exposes are discoverable at the following two endpoints:

Auth modes

The idea behind those modes is to add new non-OAuth 2.0 web APIs to complement the OAuth 2.0 specification. Indeed, the challenges that OAuth 2.0 aim to fix are not related to securing your apps using your own web APIs. OAuth 2.0 is designed to let third-parties use your APIs on behalf of your users. But as a software engineer, you often need to perform both (usually starting with the challenge to secure your APIs to power your apps). UserIn's aim is to offer an implementation strategy that is progressive. A usual progression would be:

  1. Allow your users to login or create a new account using username and password. UserIn calls this the loginsignup mode.
  2. Add support for login or create new account with Identity Providers (e.g., Facebook, Google). UserIn calls this the loginsignupfip mode.
  3. Allow third-parties to use your API. UserIn calls this the openid mode.

Notice that until you reach the third step, you actually do not need OAuth 2.0 or OpenID.

loginsignup mode

This is the simplest group of flows to implement. It only supports login and signup with username and password. Generates short-lived access_token, and optionally long-lived refresh_token upon successfull authentication. Use it to let your users login and signup to your platform using a username and password only.

loginsignup strategy requirements

loginsignupfip mode

Supports login and signup with username/password and Federated Identity Providers (e.g., Facebook, Google). Generates short-lived access_token, short-lived authorization code, and optionally long-lived refresh_token upon successfull authentication. This mode is a superset of the loginsignup mode. Use it to let your users login and signup to your platform using a username and password as well as one or many FIPs.

loginsignupfip strategy requirements

This mode is a superset of loginsignup.

openid mode

Supports login (no signup) using any the OpenID Connect flows (Authorization code, Implicit, Credentials and Password). Generates short-lived access_token, short-lived authorization code, short-lived id_token, and optionally long-lived refresh_token upon successfull authentication. Use it to let others systems access your platform. OpenID Connect and OAuth 2.0 powers the following use cases:

openid strategy requirements

Endpoints

The number of endpoints exposed by UserIn depends on its modes. UserIn supports three modes which can be combined together:

  1. loginsignup: Non-OAuth 2.0 compliant set of APIs that powers an Authorization Server that can exchange your user's username and password with an access_token, and a refresh_token. Those tokens allow your Apps to safely access your platform's API.
  2. loginsignupfip: Same as the loginsignup mode with the extra ability to use an identity provider (e.g., Facebook) to access the tokens.
  3. openid: OAuth 2.0 and OpenID Connect compliant set of APIs that powers an Authorization Server that support multiple flows to exchange your user's username and password with various tokens. The difference between this mode and the previous two is that your user is making that exchange request within the context of a third-party system which is uniquely identify by its client_id. That third-party system must be registered on your platform before your user can use your APIs within that context. Contrary to the first two modes, OAuth 2.0 make it possible to restrict which APIs can be used by combining the client_id with scopes. OAuth 2.0 and OpenID are not designed to support creating accounts, which explain why UserIn supports the first two modes above. The purpose of OAuth 2.0 is to let third-party systems registered on your platform with specific scopes to leverage some or all of your APIs to enhance the experience of a subset of their users that also have an account on your platform.

By default, UserIn exposes the following web APIs:

PathnameModeMethodTypeDescription
/v1/.well-known/configurationAllGETNot OAuth 2.0Discovery metadata JSON about all web API.
/v1/postman/collection.jsonAllGETNot OAuth 2.0Postman collection 2.0 definition to create a Postman client. This endpoint is not toggled by default. To toggle it, please refer to the Publishing a Postman collection as a web link section.
/v1/loginloginsignup & loginsignupfipPOSTNot OAuth 2.0Lets user log in.
/v1/signuploginsignup & loginsignupfipPOSTNot OAuth 2.0Lets user sign up.
/oauth2/v1/tokenAllPOSTOAuth 2.0Gets one or many tokens (e.g., access_token, refresh_token, id_token).
/oauth2/v1/revokeAllPOSTOAuth 2.0Revokes a refresh_token.
/oauth2/v1/.well-known/openid-configurationopenidGETOAuth 2.0Discovery metadata JSON about OpenID web API only.
/oauth2/v1/authorizeopenidGETOAuth 2.0Redirects to your platform's consent page to prompt user to authorize a third-party to access their resources.
/oauth2/v1/authorizeconsentopenidGETNon-OAuth 2.0Processes the consent page's response. Though this is technically not part of the OAuth 2.0 specification, this API is what allows UserIn to implement the full OAuth 2.0 Authorization Code flow. That's why this API is still labelled as OAuth 2.0.
/oauth2/v1/introspectopenidPOSTOAuth 2.0Introspects a token (e.g., access_token, refresh_token, id_token).
/oauth2/v1/userinfoopenidGETOAuth 2.0Returns user's profile based on the claims associated with the access_token.
/oauth2/v1/certsopenidGETOAuth 2.0Array of public JWK keys used to verify id_tokens.

Additionally, for each identity provider installed on UserIn, the following new endpoint is added (this example uses Facebook):

PathnameModeMethodTypeDescription
/v1/facebook/authorizeloginsignupfipGETNot OAuth 2.0Redirects to Facebook consent page.

To learn more about setting up identity providers, please refer to the next section.

/.well-known/configuration

/login

Doc under construction...

/signup

Doc under construction...

/token

refresh_token grant type

authorization_code grant type

password grant type

client_credentials grant type

/revoke

/.well-known/openid-configuration

/authorize

/authorizeconsent

/introspect

/userinfo

/certs

Doc under construction...

/<IdP>/authorize

Doc under construction...

Events and event handlers

Events overview

UserIn behaviors are managed via events and event handlers. Out-of-the-box, UserIn does not define any handlers to respond to those events. As a software engineer, this is your job to implement those event handlers in adequation with your business logic. The following list represents all the events that can be triggered during an authentication or authorization flow, but worry not, you are not forced to implement them all. You only have to implement the event handlers based on the type of authentication and authorization flow you wish to support.

  1. create_end_user
  2. create_fip_user
  3. generate_access_token
  4. generate_authorization_code
  5. generate_id_token
  6. generate_refresh_token
  7. generate_auth_request_code
  8. generate_auth_consent_code
  9. get_access_token_claims
  10. get_authorization_code_claims
  11. get_auth_request_claims
  12. get_auth_consent_claims
  13. get_client
  14. get_config
  15. get_end_user
  16. get_fip_user
  17. get_id_token_claims
  18. get_identity_claims
  19. get_refresh_token_claims
  20. get_jwks
  21. get_claims_supported
  22. get_scopes_supported
  23. get_grant_types_supported
  24. delete_refresh_token
  25. link_client_to_user
  26. get_config: Automatically implemented.

Each of those events trigger a chain of event handlers. By default, only one handler is configured in that chain (the one that you should have implemented in your UserIn Strategy). UserIn exposes an on API that allows to add more handlers for each event as shown in this example:

userIn.on('generate_access_token', (root, payload, context) => {
	console.log(`'generate_access_token' event fired. Payload:`)
	console.log(payload)
	console.log('Previous handler response:')
	console.log(root)
	console.log('Current context:')
	console.log(context)
})

root is the response returned by the previous event handler. If your handler does not return anything, root is passed to the next handler. The code above is similar to this:

userIn.on('generate_access_token', (root, payload, context) => {
	console.log(`'generate_access_token' event fired. Payload:`)
	console.log(payload)
	console.log('Previous handler response:')
	console.log(root)
	console.log('Current context:')
	console.log(context)

	return root
})

If, on the other hand, your handler returns a response, that response overrides root.

Event APIs

create_end_user

Example of that logic encapsulated in a create_end_user.js:

const { error: { wrapErrors } } = require('puffy')
const services = require('../services')

/**
 * Creates new user.
 * 
 * @param  {Object} 	root					Previous handler's response. Occurs when there are multiple handlers defined for the same event. 
 * @param  {String}		payload.user.username		
 * @param  {String}		payload.user.password		
 * @param  {String}		payload.user...			More properties
 * @param  {Object}		context					Strategy's configuration
 * 
 * @return {Object}		user					This object should always defined the following properties at a minimum.
 * @return {Object}		user.id					String ot number
 */
const handler = async (root, { user }, { repos }) => {
	// Note: The following assertions have already been checked by UserIn so this function 
	// does not need to check these again:
	// 	- 'user' is truthy. 
	// 	- 'username' is truthy
	// 	- 'password' is truthy
	// 	- 'username' does not exist already

	const errorMsg = 'Failed to create end user'

	// 1. Verify password minimal requirements
	const { valid, reason } = services.password.strongEnough(user.password)

	if (!valid)
		throw new Error(`${errorMsg}. The password is not strong enought. ${reason}`)

	const [newUserErrors, newUser] = await repos.user.insert(user)
	if (newUserErrors)
		throw wrapErrors(errorMsg, newUserErrors)

	return newUser
}

module.exports = handler

create_fip_user

Doc under construction...

generate_access_token

Example of that logic encapsulated in a generate_access_token.js:

const { error: { wrapErrors } } = require('puffy')
const tokenManager = require('../tokenManager')

/**
 * Generates a new access_token. 
 * 
 * @param  {Object} 	root				Previous handler's response. Occurs when there are multiple handlers defined for the same event. 
 * @param  {Object}		payload.claims
 * @param  {String}		payload.state		This optional value is not strictly necessary, but it could help set some context based on your own requirements.
 * @param  {Object}		context				Strategy's configuration
 * 
 * @return {String}		token
 */
const handler = async (root, { claims, state }, { repos }) => {
	// Note: The following assertions have already been checked by UserIn so this function 
	// does not need to check these again:
	// 	- 'claims' is truthy and is an object
	// 
	// This function is expected to behave following the specification described at 
	// https://github.com/nicolasdao/userin#access_token-requirements
	
	const [errors, token] = await tokenManager(repos)('access_token').create(claims)
	if (errors)
		throw wrapErrors('Failed to create access_token', errors)

	return token
}

module.exports = handler

generate_authorization_code

Doc under construction...

generate_id_token

Doc under construction...

generate_refresh_token

Example of that logic encapsulated in a generate_refresh_token.js:

const { error: { wrapErrors } } = require('puffy')
const tokenManager = require('../tokenManager')

/**
 * Generates a new refresh_token. 
 * 
 * @param  {Object} 	root				Previous handler's response. Occurs when there are multiple handlers defined for the same event. 
 * @param  {Object}		payload.claims
 * @param  {String}		payload.state		This optional value is not strictly necessary, but it could help set some context based on your own requirements.
 * @param  {Object}		context				Strategy's configuration
 * 
 * @return {String}		token
 */
const handler = async (root, { claims, state }, { repos }) => {
	// Note: The following assertions have already been checked by UserIn so this function 
	// does not need to check these again:
	// 	- 'claims' is truthy and is an object
	// 
	// This function is expected to behave following the specification described at 
	// https://github.com/nicolasdao/userin#refresh_token-requirements
	
	const [errors, token] = await tokenManager(repos)('refresh_token').create(claims)
	if (errors)
		throw wrapErrors('Failed to create refresh_token', errors)

	return token
}

module.exports = handler

generate_auth_request_code

Doc under construction...

generate_auth_consent_code

Doc under construction...

get_access_token_claims

Doc under construction...

get_authorization_code_claims

Doc under construction...

get_auth_request_claims

Doc under construction...

get_auth_consent_claims

Doc under construction...

get_client

/**
 * Gets the client's audiences, scopes and auth_methods.  
 *  
 * @param  {Object} 	root					Previous handler's response. Occurs when there are multiple handlers defined for the same event. 
 * @param  {String}		payload.client_id
 * @param  {String}		payload.client_secret	Optional. If specified, this method should validate the client_secret.
 * @param  {Object}		context					Strategy's configuration
 * 
 * @return {[String]}	output.audiences		Client's audiences.	
 * @return {[String]}	output.scopes			Client's scopes.	
 * @return {[String]}	output.auth_methods		Client's auth_methods.	
 * @return {[String]}	output.redirect_uris	Client's allowed redirect URIs
 */
const handler = (root, { client_id, client_secret }, context) => {
	const client = context.repos.client.find(x => x.client_id == client_id)
	
	if (!client)
		return null

	if (client_secret && client.client_secret != client_secret)
		throw new Error('Unauthorized access')

	return {
		audiences: client.audiences || [],
		scopes: client.scopes || [],
		auth_methods: client.auth_methods || [],
		redirect_uris: client.redirect_uris || []
	}
}

get_config

This method is not required to support any workflow. Instead, it can be used as a utility to add custom features.

/**
 * Gets the strategy's configuration object. 
 * 
 * @param  {Object} 	root							Previous handler's response. Occurs when there 
 *                              						are multiple handlers defined for the same event. 
 * @return {String}		output.iss		
 * @return {Number}		output.expiry.id_token			
 * @return {Number}		output.expiry.access_token		
 * @return {Number}		output.expiry.refresh_token		
 * @return {Number}		output.expiry.code	
 */
const get_config = (root) => {
	console.log('get_config fired')
	console.log('Previous handler response:')
	console.log(root)
	
	return {
		iss: 'https://userin.com',
		expiry: {
			id_token: 3600,
			access_token: 3600,
			code: 30
		}
	}
}

get_end_user

Example of that logic encapsulated in a get_end_user.js:

const { error:{ InvalidCredentialsError } } = require('userin')
const { error: { wrapErrors } } = require('puffy')
const services = require('../services')

/**
 * Gets the user ID and optionnaly its associated client_ids if the 'openid' is supported.
 * If the username does not exist, a null value must be returned. However, the 'password' is optional. 
 * If the 'password' is provided, it must be verified. If the verification fails, an error of type 
 * InvalidCredentialsError must be thrown (const { error:{ InvalidCredentialsError } } = require('userin'))
 * 
 * @param  {Object} 	root					Previous handler's response. Occurs when there are multiple handlers defined for the same event. 
 * @param  {String}		payload.user.username
 * @param  {String}		payload.user.password
 * @param  {String}		payload.user...			More properties
 * @param  {String}		payload.client_id		Optional. Might be useful for logging or other custom business logic.
 * @param  {String}		payload.state			Optional. Might be useful for logging or other custom business logic.
 * @param  {Object}		context					Strategy's configuration
 * 
 * @return {Object}		user					This object should always defined the following properties at a minimum.
 * @return {Object}		user.id					String ot number
 * @return {[Object]}	user.client_ids		
 */
const handler = async (root, { user, client_id, state }, { repos }) => {
	// Note: The following assertions have already been checked by UserIn so this function 
	// does not need to check these again:
	// 	- 'user' is truthy. 
	// 	- 'user.username' is truthy
	// 
	// This function is expected to behave as follow:
	// 	- If the 'username' does not exist, a null value must be returned. 
	// 	- The 'password' is optional. 
	// 	- If the 'password' is provided, it must be verified. If the verification fails, an error of type 
	// 	InvalidCredentialsError must be thrown (const { error:{ InvalidCredentialsError } } = require('userin'))

	const errorMsg = 'Failed to get end user'

	const [confirmedUserErrors, confirmedUser] = await repos.user.find({ where:{ email:user.username } })
	if (confirmedUserErrors)
		throw wrapErrors(errorMsg, confirmedUserErrors)

	if (!confirmedUser)
		return null

	if (user.password) {
		const eMsg = `${errorMsg}. Invalid username or password.`
		const saltedPassword = confirmedUser.password
		if (!saltedPassword || !confirmedUser.salt)
			throw new InvalidCredentialsError(eMsg)

		const valid = services.password.verify({ 
			password:user.password, 
			salt:confirmedUser.salt, 
			hashedSaltedPassword:confirmedUser.password 
		})

		if (!valid)
			throw new InvalidCredentialsError(eMsg) 
	}

	return {
		id: confirmedUser.id,
		client_ids:[]
	}
}

module.exports = handler

get_fip_user

Doc under construction...

get_id_token_claims

Doc under construction...

get_identity_claims

Doc under construction...

get_refresh_token_claims

Example of that logic encapsulated in a get_refresh_token_claims.js:

const { error: { wrapErrors } } = require('puffy')
const tokenManager = require('../tokenManager')

/**
 * Gets the refresh_token's claims
 * 
 * @param  {Object} 	root				Previous handler's response. Occurs when there are multiple handlers defined for the same event. 
 * @param  {Object}		payload.token
 * @param  {Object}		context				Strategy's configuration
 * 
 * @return {Object}		claims				This object should always defined the following properties at a minimum.
 * @return {String}		claims.iss			
 * @return {Object}		claims.sub			String or number
 * @return {String}		claims.aud
 * @return {Number}		claims.exp
 * @return {Number}		claims.iat
 * @return {Object}		claims.client_id	String or number
 * @return {String}		claims.scope
 */
const handler = async (root, { token }, { repos }) => {
	// Note: The following assertions have already been checked by UserIn so this function 
	// does not need to check these again:
	// 	- 'token' is truthy and is a string
	// 
	// This function is expected to behave following the specification described at 
	// https://github.com/nicolasdao/userin#refresh_token-requirements
	
	const [errors, refresh_token] = await tokenManager(repos)('refresh_token').getClaims(token)
	if (errors)
		throw wrapErrors('Failed to create refresh_token', errors)

	return refresh_token
}

module.exports = handler

get_jwks

Doc under construction...

get_claims_supported

Doc under construction...

get_scopes_supported

Doc under construction...

get_grant_types_supported

Doc under construction...

delete_refresh_token

Doc under construction...

link_client_to_user

Doc under construction...

OpenID Connect tokens & authorization code requirements

If you're implementing a UserIn strategy that supports the openid mode, then you must generate your tokens and authorization code following strict requirements.

id_token requirements

An id_token must:

  1. Be cryptographically signed on your server so you can validate that it has indeed been issued by you and that it has not been tampered when you get it back.
  2. Be short-lived. It cannot be valid for more than an hour after being issued.
  3. Be a JWT containing at a minimum the following claims (to be compliant to the OIDC specification):
    • iss: Issuer Identifier for the Issuer of the response. The iss value is a case sensitive URL using the https scheme that contains scheme, host, and optionally, port number and path components and no query or fragment components.
    • sub: Subject Identifier. That the ID of the token owner (i.e., the ID you use in your own system to do a user lookup). It must not exceed 255 ASCII characters in length. The sub value is a case sensitive string.
    • aud: Audience(s) that this ID Token is intended for. It MUST contain the OAuth 2.0 client_id of the Relying Party as an audience value. It MAY also contain identifiers for other audiences. In the general case, the aud value is an array of case sensitive strings. In the common special case when there is one audience, the aud value MAY be a single case sensitive string.
    • exp: Expiration time on or after which the ID Token MUST NOT be accepted for processing. The processing of this parameter requires that the current date/time MUST be before the expiration date/time listed in the value. Implementers MAY provide for some small leeway, usually no more than a few minutes, to account for clock skew. Its value is a JSON number representing the number of seconds from 1970-01-01T0:0:0Z as measured in UTC until the date/time. See RFC 3339 [RFC3339] for details regarding date/times in general and UTC in particular.
    • iat: Time at which the JWT was issued. Its value is a JSON number representing the number of seconds from 1970-01-01T0:0:0Z as measured in UTC until the date/time.

The sub must be unique per iss. You can use the iss + sub to uniquely identify users.

This JWT can also contain more reserved OIDC fields defined here. The optional other fields are referred as claims (e.g., name, family_name, given_name). OIDC defines a series of standard claims associated with each scopes. The number of claims contained in the id_token depends on the scopes passed to the request. You are not limited to the standard claims. You can create your own scopes and associate whatever custom claims to each scope.

Here is an example from an Okta JWT id_token:

{
	"iss": "https://micah.okta.com/oauth2/aus2yrcz7aMrmDAKZ1t7",
	"sub": "00u2yulup4eWbOttd1t7",
	"aud": "0oa2yrbf35Vcbom491t7",
	"exp": 1501535822,
	"iat": 1501532222,
	"name": "Okta OIDC Fun",
	"locale": "en-US",
	"email": "okta_oidc_fun@okta.com",
	"ver": 1,
	"jti": "ID.Zx8EclaZmhSckGHOCRzOci2OaduksmERymi9-ad7ML4",
	"amr": [
		"pwd"
	],
	"idp": "00o1zyyqo9bpRehCw1t7",
	"nonce": "c96fa468-ca1b-46f0-8974-546f23f9ee6f",
	"preferred_username": "okta_oidc_fun@okta.com",
	"given_name": "Okta OIDC",
	"family_name": "Fun",
	"zoneinfo": "America/Los_Angeles",
	"updated_at": 1499922371,
	"email_verified": true,
	"auth_time": 1501528157
}

(1) Though the OAuth2 documentation specifies that the aud MUST contain the client_id of the resources that can accept it, most concrete implementation use URIs. For example, let's say that your token can only access the following two APIs: https://api.example.com and https://api.otherexample.com/somepath. In that case, the aud value associated with your token would be: "aud":"https://api.example.com https://api.otherexample.com".

access_token requirements

An access_token must:

  1. Be cryptographically signed on your server so you can validate that it has indeed been issued by you and that it has not been tampered when you get it back.
  2. Be short-lived. It cannot be valid for more than an hour after being issued.
  3. Be able to be associated with:
    • The client_id that made the original request.
    • The owner's identity, whether that owner is a user or a service account.
    • The scopes it was generated it from.
    • The audience(s) that can be accessed by this token(1).

Criteria 2 and 3 can be achieved by encoding this token using a JWT, but this is not required by the OAuth2 specification. If you decide to use JWT to implement your access tokens, the standard approach is to include some of the following standard OIDC claims:

Even if you choose to not use a JWT, those claims above should be associated with the access token in one way or the other.

The sub must be unique per iss. You can use the iss + sub to uniquely identify users.

Here is an example from an Okta JWT access_token:

{
	"iss": "https://micah.okta.com/oauth2/aus2yrcz7aMrmDAKZ1t7",
	"sub": "okta_oidc_fun@okta.com",
	"exp": 1501531801,
	"iat": 1501528201,
	"scope": "openid email",
	"client_id": "0oa2yrbf35Vcbom491t7",
	"aud": "test",
	"token_type": "Bearer",
	"active": true,
	"username": "okta_oidc_fun@okta.com",
	"jti": "AT.upPJqU-Ism6Fwt5Fpl8AhNAdoUeuMsEgJ_VxJ3WJ1hk",
	"uid": "00u2yulup4eWbOttd1t7"
}

To learn more about the various strategies used to generate access tokens and maintain their state, please refer to this article called OAuth Access Token Implementation.

(1) Though the OAuth2 documentation specifies that the aud MUST contain the client_id of the resources that can accept it, most concrete implementation use URIs. For example, let's say that your token can only access the following two APIs: https://api.example.com and https://api.otherexample.com/somepath. In that case, the aud value associated with your token would be: "aud":"https://api.example.com https://api.otherexample.com".

refresh_token requirements

A refresh_token must:

  1. Be protected against tampering(1).
  2. Be long-lived. Theoretically, this token could live forever or you could add an expiry date far away from their creation date. It is entirely up to you to decide how long you want those tokens to exist. However, what you may want to support is the ability to revoke them.
  3. Be associated with the context of the original request that created that refresh token(2). That context must include at a minimum:
    • The client_id so only the same client ID can exchange that refresh token for another articfact.
    • If the refresh token does not last forever, the creation date or the expiry date must be associated with the refresh token so it can be invalidated if it has expired.
    • The scopes that this refresh token was originally requested for. When the refresh token is used to acquire another token, those scopes need to be checked against the client_id to make sure they are still accessible.
    • The audience(s) that this refresh token was originally requested for. When the refresh token is used to acquire another token, this audience(s) need to be checked against the client_id to make sure they are still accessible.

(1) It is entirely up to you to decide how to protect your refresh token against tampering. (2) It is entirely up to you to decide how to persist that context between requests. Because it is highly desirable to invalidate refresh tokens, a natural solution is to persist refresh tokens (incl. persisting their associated context) in your own database.

Authorization code requirements

A code must:

  1. Be protected against tampering(1).
  2. Be very short-lived. OAuth2 recommends to set the duraction time between 30 and 60 seconds. The longest duration allowed by OAuth2 is 10 minutes.
  3. Be associated with the context of the original request that created that code(1). That context must include at a minimum:
    • The client_id so only the same client ID can exchange that code for another articfact later.
    • The creation date or the expiry date so the code can be invalidated if it has expired (recommended duraction is 30 to 60 seconds, max. 10 minutes).
    • The scopes that this code can associate with a token. Certain flows(e.g., refresh token scenario when the Authorization Code flow is used) rely on those scopes.
    • The audience(s) that this code can associate with a token.

(1) It is entirely up to you to decide how to protect your code against tampering. Two suggestions are listed below:

  1. Encrypt the context in the code. Because this context is not meant to be shareable with other third-parties, a simple AES encryption should do. When the code comes back later, it can be verified to ensure no tampering occured. Decrypting thde code allows to retrieve the context.
  2. Use a random unique identifier to generate the code and use it a key to store the context in a secured persistent storage. When the code comes back, that context can be extracted from that persistent storage using a simple lookup.

Setting up an identity provider

UserIn supports both Passport strategies and native OpenID providers via their .well-known/openid-configuration discovery endpoint (e.g., https://accounts.google.com/.well-known/openid-configuration). In both cases, an app must be registered with each identity provider. The annex of this document details the steps to set this up for some of the most popular provider in the Registering an application with an Identity Provider section.

Using Passport

Example of npm Passport packages:

The next example uses Facebook:

const { UserIn } = require('userin')
const Facebook = require('passport-facebook')
const YourStrategy = require('./src/YourStrategy.js')

const userin = new UserIn({
	Strategy: YourStrategy,
	modes:['loginsignupfip', 'openid'], // You have to define at least one of those three values.
	config: {
		baseUrl: 'http://localhost:3330',
		consentPage: 'https://your-domain.com/consent-page', 	// only required when modes contains 'openid'.
		tokenExpiry: {
			access_token: 3600,
			id_token: 3600, 					// only required when modes contains 'openid'.
			code: 30 							// only required when modes contains 'loginsignupfip' or 'openid'.
		}
	}
})

userIn.use(Facebook, {
	clientID: '12234',
	clientSecret: '54332432',
	scopes: ['public_profile'],
	profileFields: ['id', 'displayName', 'photos', 'email', 'first_name', 'middle_name', 'last_name']
})

NOTES:

Using an OpenID discovery endpoint

To this day (Oct. 2020), Google is the only major player to have adopted OpenID. The others have implemented specialized version of OAuth 2.0 (Facebook has rolled out their own implementation of OpenID Connect called Facebook Connect).

const { UserIn } = require('userin')
const YourStrategy = require('./src/YourStrategy.js')

const userin = new UserIn({
	Strategy: YourStrategy,
	modes:['loginsignupfip', 'openid'], // You have to define at least one of those three values.
	config: {
		baseUrl: 'http://localhost:3330',
		consentPage: 'https://your-domain.com/consent-page', 	// only required when modes contains 'openid'.
		tokenExpiry: {
			access_token: 3600,
			id_token: 3600, 					// only required when modes contains 'openid'.
			code: 30 							// only required when modes contains 'loginsignupfip' or 'openid'.
		}
	}
})

userIn.use({
	name:'google',
	client_id: '12234',
	client_secret: '54332432',
	discovery: 'https://accounts.google.com/.well-known/openid-configuration',
	scopes:['profile', 'email']
})

NOTES:

Implementation guidelines

Because OAuth 2.0 flows are not stateless we recommend to implement your UserIn strategy using dependency injection. This will greatly help with unit testing.

Creating a UserIn Strategy class

Doc under construction...

Unit testing

Testing a UserIn Strategy class

UserIn ships with a suite of Mocha unit tests. To test your own strategy:

  1. Install mocha and chai:
npm i -D mocha chai
  1. Create a new test folder in your project root directory.
  2. Under that test folder, create a new strategy.js (or whatever name you see fit), and paste code similar to the following:
const { testSuite } = require('userin')
const { YourStrategyClass } = require('../src/yourStrategy.js')

const options = { skip:'' } // Does not skip any test.

// To test a stragegy in 'loginsignup' mode, the following minimum config is required.
const config = {
	tokenExpiry: {
		access_token: 3600
	}
}

// The required stub's properties are (change the values to your own stub):
const stub = {
	user: {
		username: 'valid@example.com', // Valid username in your own stub data.
		password: '123456' // Valid password in your own stub data.
	},
	newUserPassword: 'd32def32feq' // Add the password that will be used to test new users
}

testSuite.testLoginSignup(YourStrategyClass, config, stub, options)
  1. Add a new test script in your package.json:
	"scripts": {
		"test": "mocha --exit"
	}
  1. Run the test:
npm test

testSuite API

The testSuite API exposes four different test suite, one for each mode + one that combines all the modes. Each test suite uses the same signature:

  1. testLoginSignup function
  2. testLoginSignupFIP function
  3. testOpenId function
  4. testAll function

The signature is (YourStrategyClass: UserInStrategy, config: Object, stub: Object[, options: Object]) where:

testLoginSignup function

Runs the following tests:

const { testSuite } = require('userin')
const { YourStrategyClass } = require('../src/yourStrategy.js')

// Use the 'option' value to control which test is run. By default, all tests are run.
// Valid test names are: 'all', 'strategy', 'login', 'signup'
// 
// const options = { skip:'all' } // Skips all tests in this suite.
// const options = { skip:'login' } // Skips the 'login' test in this suite.
// const options = { skip:['login', 'signup'] } // Skips the 'login' and 'signup' tests in this suite.
// const options = { only:'login' } // Only run the 'login' test in this suite.
// const options = { only:['login', 'signup'] } // Only run the 'login' and 'signup' tests in this suite.
const options = { skip:'', showResults:['login.handler.09,10'] } 	// Does not skip any test and show the results of:
																	// - Test 'login.handler.09'
																	// - Test 'login.handler.10'

// To test a stragegy in 'loginsignup' mode, the following minimum config is required.
const config = {
	tokenExpiry: {
		access_token: 3600
	}
}

// The required stub's properties are (change the values to your own stub):
const stub = {
	user: {
		id: 1,
		username: 'valid@example.com', // Valid username in your own stub data.
		password: '123456' // Valid password in your own stub data.
	},
	newUserPassword: 'd32def32feq' // Add the password that will be used to test new users
}

testSuite.testLoginSignup(YourStrategyClass, config, stub, options)

testLoginSignupFIP function

Runs the following tests:

const { testSuite } = require('userin')
const { YourStrategyClass } = require('../src/yourStrategy.js')

// Use the 'option' value to control which test is run. By default, all tests are run. 
// Valid test names are: 'all', 'strategy', 'login', 'signup', 'fiploginsignup'
// 
// const options = { skip:'all' } // Skips all tests in this suite.
// const options = { skip:'login' } // Skips the 'login' test in this suite.
// const options = { skip:['login', 'signup'] } // Skips the 'login' and 'signup' tests in this suite.
// const options = { only:'login' } // Only run the 'login' test in this suite.
// const options = { only:['login', 'signup'] } // Only run the 'login' and 'signup' tests in this suite.
const options = { skip:'' } // Does not skip any test.

// To test a stragegy in 'loginsignupfip' mode, the following minimum config is required.
const config = {
	tokenExpiry: {
		access_token: 3600,
		code: 30
	}
}

// The required stub's properties are (change the values to your own stub):
const stub = {
	user: {
		id: 1,
		username: 'valid@example.com', // Valid username in your own stub data.
		password: '123456' // Valid password in your own stub data.
	},
	newUserPassword: 'd32def32feq', // Add the password that will be used to test new users
	fipUser: { // this user should be different from the one above.
		id: '1N7fr2yt', // ID of the user in the identity provider plaftform
		fipName: 'facebook', // Identity provider's name
		userId: 2 // ID of the user on your platform
	}
}

testSuite.testLoginSignupFIP(YourStrategyClass, config, stub, options)

testOpenId function

Runs the following tests:

const { testSuite } = require('userin')
const { YourStrategyClass } = require('../src/yourStrategy.js')

// Use the 'option' value to control which test is run. By default, all tests are run. 
// Valid test names are: 'all', 'strategy', 'introspect', 'token', 'userinfo'
// 
// const options = { skip:'all' } // Skips all tests in this suite.
// const options = { skip:'introspect' } // Skips the 'introspect' test in this suite.
// const options = { skip:['introspect', 'token'] } // Skips the 'introspect' and 'token' tests in this suite.
// const options = { only:'introspect' } // Only run the 'introspect' test in this suite.
// const options = { only:['introspect', 'token'] } // Only run the 'introspect' and 'token' tests in this suite.
const options = { skip:'' } // Does not skip any test.

// To test a stragegy in 'openid' mode, the following minimum config is required.
const config = {
	openid: {
		iss: 'https://www.userin.com',
		tokenExpiry: {
			id_token: 3600,
			access_token: 3600,
			code: 30
		}
	}
}

// The required stub's properties are (change the values to your own stub):
const stub = {
	client: { 
		id: 'client_with_at_least_one_user_and_no_auth_methods',
		secret: '98765', 
		aud: 'https://private-api@mycompany.com',
		user: { 
			id: 1,
			username: 'valid@example.com', // Valid username in your own stub data.
			password: '123456' // Valid password in your own stub data.
			claimStubs: [{ // Define the identity claims you want to support here and fill the value for the 'valid@example.com' user.
				scope:'profile',
				claims: {
					given_name: 'Nic',
					family_name: 'Dao',
					zoneinfo: 'Australia/Sydney'
				}
			}, {
				scope:'email',
				claims: {
					email: 'nic@cloudlessconsulting.com',
					email_verified: true
				}
			}, {
				scope:'phone',
				claims: {
					phone: '+61432567890',
					phone_number_verified: false
				}
			}, {
				scope:'address',
				claims: {
					address: 'Castle in the shed'
				}
			}]
		}
	},
	altClient: { 
		id: 'another_client_with_no_auth_methods', 
		secret: '3751245'
	},
	privateClient: { 
		// this client must have its 'auth_methods' set to ['client_secret_basic'], ['client_secret_post'] or 
		// ['client_secret_basic', 'client_secret_post']
		id: 'yet_another_client_with_auth_methods', 
		secret: '3751245'
	}
}

testSuite.testOpenId(YourStrategyClass, config, stub, options)

testAll function

This test function tests all the previous three tests at once. Use it if you have created a UserIn Strategy class that imlements all the event handlers. The signature is the same as for the other tests. Merge all the stubs from the previous tests into a single stub object.

Dependency injection

The test suite supports inversion of control via dependency injection. All the event handlers supports the same signature:

(root: Object, payload: Object, context: Object).

For example:

YourStrategyClass.prototype.get_end_user = (root, { user }, context) => {
	const existingUser = USER_STORE.find(x => x.email == user.username)
	if (!existingUser)
		return null
	if (user.password && existingUser.password != user.password)
		throw new Error('Incorrect username or password')

	const client_ids = USER_TO_CLIENT_STORE.filter(x => x.user_id == existingUser.id).map(x => x.client_id)

	return {
		id: existingUser.id,
		client_ids
	}
}

This example shows that get_end_user depends on the USER_STORE and USER_TO_CLIENT_STORE to function. Those would typically be connectors that can perform IO queries to your backend storage. This code is not properly designed to support unit testing, especially if you are tryng to test inserts. To solve this problem, the best practice is to inject those dependencies from the outside.

This is one the purpose of the context object. The context object is the config object passed to the YourStrategyClass instance:

const { testSuite } = require('userin')
const { YourStrategyClass } = require('../src/yourStrategy.js')

// To test a stragegy in 'loginsignup' mode, the following minimum config is required.
const config = {
	tokenExpiry: {
		access_token: 3600
	},
	repos: {
		user: {
			find: (userId)
		}
	}
}

// The required stub's properties are (change the values to your own stub):
const stub = {
	user: {
		username: 'valid@example.com', // Valid username in your own stub data.
		password: '123456' // Valid password in your own stub data.
	},
}

testSuite.testLoginSignup(YourStrategyClass, config, stub, options)

In this example, let's modified the config as follow:

const config = {
	tokenExpiry: {
		access_token: 3600
	},
	repos: {
		user: USER_STORE,
		userToClient: USER_TO_CLIENT_STORE
	}
}

With this change, the get_end_user can be rewritten as follow:

YourStrategyClass.prototype.get_end_user = (root, { user }, context) => {
	const existingUser = context.repos.user.find(x => x.email == user.username)
	if (!existingUser)
		return null
	if (user.password && existingUser.password != user.password)
		throw new Error('Incorrect username or password')

	const client_ids = context.repos.userToClient.filter(x => x.user_id == existingUser.id).map(x => x.client_id)

	return {
		id: existingUser.id,
		client_ids
	}
}

This design pattern is called dependency injection. It allows to replace the behaviors from the outside. The following snippet shows how to inject dependencies in the UserIn middleware rather than on the Strategy:

const { someDependency } = require('../src/dependencies')
const userIn = new UserIn({
	Strategy: MockStrategy,
	modes:['loginsignup', 'loginsignupfip', 'openid'], // You have to define at least one of those three values.
	config: {
		baseUrl: 'http://localhost:3330',
		consentPage: 'https://your-domain.com/consent-page',
		tokenExpiry: {
			access_token: 3600,
			id_token: 3600,
			code: 30
		}
		someDependency
	}
})

Integration testing

Exporting the API to Postman

UserIn can publish its API documentation using Postman Collection v2.1. There are two ways to export a Postman collection:

  1. Publish a new web endpoint at {{YOUR_DOMAIN}}/v1/postman/collection.json and use that link in Postman to import that collection.
  2. Export the collection in a local file and then import that file in Postman.

Publishing a Postman collection as a web link

Use this API:

userIn.use(new Postman('your-collection-name'))

The full example looks like this:

const express = require('express')
const app = express()
const Facebook = require('passport-facebook')
const { UserIn, Postman } = require('userin')
const YourStrategy = require('./src/YourStrategy')

const userIn = new UserIn({
	Strategy: YourStrategy,
	modes:['loginsignupfip', 'openid'], // You have to define at least one of those three values.
	config: {
		baseUrl: 'http://localhost:3330',
		consentPage: 'https://your-domain.com/consent-page', 	// only required when modes contains 'openid'.
		tokenExpiry: {
			access_token: 3600,
			id_token: 3600, 					// only required when modes contains 'openid'.
			code: 30 							// only required when modes contains 'loginsignupfip' or 'openid'.
		}
	}
})

userIn.use(Facebook, {
	scopes: ['public_profile'],
	profileFields: ['id', 'displayName', 'photos', 'email', 'first_name', 'middle_name', 'last_name']
})

userIn.use({
	name:'google',
	discovery: 'https://accounts.google.com/.well-known/openid-configuration',
	scopes:['profile', 'email']
})

userIn.use(new Postman('userin-my-app'))

app.use(userIn)
app.listen(3330, () => console.log('UserIn listening on https://localhost:3330'))

Export a Postman collection in a local file

Once the UserIn instance has been created and configured, use the Postman utility as follow:

Postman.export({
	userIn,
	name: 'userin-my-app',
	path: './postman-collection.json'
})

The full example looks like this:

const express = require('express')
const app = express()
const Facebook = require('passport-facebook')
const { UserIn, Postman } = require('userin')
const YourStrategy = require('./src/YourStrategy')

const userIn = new UserIn({
	Strategy: YourStrategy,
	modes:['loginsignupfip', 'openid'], // You have to define at least one of those three values.
	config: {
		baseUrl: 'http://localhost:3330',
		consentPage: 'https://your-domain.com/consent-page', 	// only required when modes contains 'openid'.
		tokenExpiry: {
			access_token: 3600,
			id_token: 3600, 					// only required when modes contains 'openid'.
			code: 30 							// only required when modes contains 'loginsignupfip' or 'openid'.
		}
	}
})

userIn.use(Facebook, {
	scopes: ['public_profile'],
	profileFields: ['id', 'displayName', 'photos', 'email', 'first_name', 'middle_name', 'last_name']
})

userIn.use({
	name:'google',
	discovery: 'https://accounts.google.com/.well-known/openid-configuration',
	scopes:['profile', 'email']
})

Postman.export({
	userIn,
	name: 'userin-my-app',
	path: './postman-collection.json'
})

app.use(userIn)
app.listen(3330, () => console.log('UserIn listening on https://localhost:3330'))

When this code is executed, a postman-collection.json file is autogenerated. Use Postman to import the collection using this file.

Authorization code flow implementation

  1. User-agent browses to the /oauth2/v1/authorize URL passing in the query parameters the following properties:
    • client_id [required]
    • response_type [required]
    • redirect_uri [required]
    • scope [optional]
    • state [optional]
    • code_challenge [optional]
    • code_challenge_method [optional]
  2. UserIn validates those parameters. Upon successful validation, UserIn:
    1. Creates an opaque code that can be exchanged later for those parameters above.
    2. Redirects the user-agent to the consent page including the opaque code in the query parameters.
  3. The backend behind the consent page (which is outside of UserIn control) authenticates the user (using probably the UserIn /login or signup web APIs) and prompt the user to confirm access to their resources by the third-party (aka the client_id). Upon agreeing, the consent page backend is expected to behave as follow:
    1. Creates an new opaque short-lived consentcode that UserIn can exchange for a user_id, a username and the original code from step 2.1.
    2. Redirects the user-agent to the UserIn /v1/authorizeconsent endpoint including the consentcode in the query parameters.
  4. UserIn:
    1. Exchanges the consentcode for an object contaiing the following properties:
      • user_id
      • username
      • code: That's the original code created in step 2.1.
      • exp: A timestamp representing the number of seconds since epoch. That timestamp represents the expiration date after which this consentcode is not valid anymore.
    2. Exchanges the code for the original parameters from step 1.
    3. Creates an explicit link between the client_id and the user.id so that client_id can retrieve tokens without forcing the user to go throught the consent page again.
    4. Generates the tokens based on the response_type value from step 1.
    5. Redirects the user-agent to the redirect_uri URL from step 1 including in the tokens in the query parameters.

Flexible flows

Flexible flows are those who leverage the UserIn APIs built to support the OAuth 2.0 flows but do not obey to the strict OAuth 2.0 specification.

Generally speaking, those flows exist to power the non-third-party use cases, i.e., the ones where your API powers your own Apps directly (e.g., signing up with username and password). Those flows do not need any client_id (which exist to identity a third-party). UserIn's value proposition is to leverage the existing OAuth 2.0 APIs to support both standard (requires a client_id and maybe a client_secret too) and non-standard flows. To deliver this value, UserIn uses this simple approach. When users are authorized via your platform's consent page (OAuth 2.0 flow), then tokens (including the authorization code) are linked to a client_id. All subsequent flows involving those tokens require a client_id. On the other hand, When users login or signup via your login/signup page (non-OAuth 2.0 flow), then no client_id is associated with the generated tokens, and therefore the client_id is not required.

Login/Signup flow using an third-party Identity Provider

This is the case where a user wish to use an identity provider such as Facebook to login or signup to your platform. Behind the scene, UserIn interacts with the identity provider's OAuth 2.0 authorization code flow to make this happen, but this next web API is not part of the OAuth 2.0 specification.

GET https://YOUR_DOMAIN/v1/google/authorize?
    response_type=code&
    redirect_uri=https://YOUR_DOMAIN/v1/google/authorizecallback&
    scope=profile&
    mode=signup

Notice that this HTTP GET is similar to the OAuth 2.0 /authorize request used in the Authorization Code flow except:

Contributing - Developer notes

Contribution guidelines

Coming soon...

Unit tests

The logTestErrors API

Almost all unit tests use the custom logTestErrors API. This API's purpose is to capture explicit error logs to display them when the developer uses the verbose mode. This API leverages UserIn's functional error handling style.

const { logTestErrors } = require('./_core')

const verbose = true
const logTest = logTestErrors()

it('Should fail when something bad happens.', done => {
	const logE = logTest(done)

	logE.run(co(function *() {
		// Functional error handling style where the output is always an arrat where the first element is an array of errors
		// and the second is the exppected result. 
		const [errors, result] = yield someFunction()

		// Log errors to support the  verbose mode. 
		logE.push(errors)

		// Run the usual assertions
		assert.isOk(errors, '01')
		assert.isOk(errors.length, '02')
		assert.isOk(errors.some(e => e.message && e.message.indexOf('Missing \'get_client\' handler') >= 0), '03')
		done()
	}))
})

FAQ

How to use UserIn in Postman?

Please refer to Exporting the API to Postman.

How to deal with Facebook restriction to HTTPS redirect only when testing locally?

The easiest solution is to use ngrok which can expose a web server running on your local machine to the internet via both HTTP and HTTPS. For example, your UserIn server that is locally accessible via http://localhost:3330 will be available publicly via https://2e6c759d16cf.ngrok.io. Add that URL to the allowlist in your Facebook App and then update the {{base_url}} in Postman to use that new URL.

When is the client_secret required?

https://developer.okta.com/blog/2019/08/22/okta-authjs-pkce https://auth0.com/docs/flows/authorization-code-flow-with-proof-key-for-code-exchange-pkce

Annex

Jargon and concepts

Grant types

Grant types are labels used in the /token API to determine how the provided credentials must be exchanged with tokens. OAuth 2.0 supports the following grant types:

ABAC PBAC CBAC & Co

Main ref at https://en.wikipedia.org/wiki/Attribute-based_access_control#Other_models

ABAC stands for Attribute-based access control. This is strictly equivalent to PBAC (Policy-Based Access Control used in IAM) and CBAC (Claims-Based Access Control used the Microsoft landscape).

Historically, access control models have included mandatory access control (MAC), discretionary access control (DAC), and more recently role-based access control (RBAC). These access control models are user-centric and do not take into account additional parameters such as resource information, the relationship between the user (the requesting entity) and the resource, and dynamic information e.g. time of the day or user IP. ABAC tries to address this by defining access control based on attributes which describe the requesting entity (the user), the targeted object or resource, the desired action (view, edit, delete...), and environmental or contextual information. This is why access control is said to be attribute-based.

Registering an application with an Identity Provider

Facebook

Goal

Facebook App set up steps

  1. Browse to https://developers.facebook.com/.
  2. In the top menu, expand My Apps, click on the Add New App link and fill up the form. Once submitted, you'll be redirected to your new app dashboard page.
  3. Optionally, create a test version of your new app (recommended) to ease testing during the development stage. In the top left corner, click on the app name to expand the menu. At the bottom of that menu, click on the Create Test App button.
  4. Get the Client ID and the Client Secret. The Client ID can be found easily at the top of each page. The Client Secret is under Settings/Basic section in the left menu.
  5. Add valid OAuth redirect URI:
    1. In the left menu, expand the Facebook Login tab and click on Settings.
    2. Add your authorized redirect URI under Valid OAuth Redirect URIs.

IMPORTANT NOTE: Facebook only allows HTTPS redirect URIs. This can make local development on localhost challenging. We recommend to use the ngrok to overcome this limitation. This utility offers a free plan that allows to expose your localhost to the web and uses HTTPS.

Troubleshooting - Can't Load URL: The domain of this URL isn't included in the app's domains

This error happens when you've stopped testing in dev mode (i.e., using localhost) and you've either forgot to proceed to step 5 above.

<img src="https://user-images.githubusercontent.com/3425269/89261172-f72faf00-d670-11ea-8fec-3078b07491dd.png" width="400px">

The error message above should appear at the following URL: https://www.facebook.com/v3.2/dialog/oauth?response_type=code&redirect_uri=LONG_ENCODED_URL&scope=public_profile

To fix this issue:

Google

Goal

Google App set up steps

  1. Sign in to your Google Cloud console at https://console.cloud.google.com.
  2. Choose the project tied to your app, or create a new one.
  3. Once your project is created/selected, expand the left menu and select APIs & Services / Credentials
  4. In the Credentials page, select the Credentials tab, click on the Create credentials button and select OAuth Client ID. Fill up the form:
    1. Name: This is not really important and will not be displayed to your user. You can leave the default.
    2. Authorized JavaScript origins: This is optional but highly recommended before going live.
    3. Authorized redirect URIs: This is required.
  5. After completing the step above, a confirmation screen pops up. Copy the client ID (i.e., the Client ID) and the client secret (i.e., the Client Secret).
  6. In the Credentials page, select the OAuth consent screen tab. Fill up the form depending on your requirements. Make sure you update the Application name to your App name so that your App users see that name in the consent screen. You can also add your brand in the consent screen by uploading your App logo. Don't forget to click the Save button at the bottom to apply your changes.

LinkedIn

Goal

LinkedIn App set up steps

  1. Sign in to your LinkedIn account and then browse to https://www.linkedin.com/developers/apps to either create a new App or access any existing ones. For the sake of this tutorial, the next steps only focus on creating a new App.
  2. In the top right corner of the My Apps page, click on the Create app button.
  3. Fill up the form and then click the Create app button at the bottom.
  4. Once the the App is created, you are redirected to the App's page. In that page, select the Auth tab and copy the Client ID (i.e., the Client ID) and the Client secret (i.e., the Client Secret).
  5. Still in the Auth tab, under the OAuth 2.0 settings section, enter the redirect URI.

Troubleshooting - LinkedIn - Bummer, something went wrong

This error happens when you've either forgot to proceed to step 5 above or made a mistake in that step.

<img src="https://user-images.githubusercontent.com/3425269/89270120-3c0e1280-d67e-11ea-8545-a8b1664b70fc.png" width="400px">

The error message above should appear at the following URL: https://www.linkedin.com/oauth/v2/authorization?response_type=LONG_ENCODED_URLcode&redirect_uri=&scope=r_liteprofile%20r_emailaddress%20w_member_social&client_id=123456

To fix this issue:

GitHub

Goal

GitHub App set up steps

  1. Sign in to your Github account and then browse to https://github.com/settings/apps to either create a new App or access any existing ones. For the sake of this tutorial, the next steps only focus on creating a new App.
  2. In the top right corner of the GitHub Apps page, click on the New GitHub App button.
  3. Fill up the form and then click the Create GitHub App button at the bottom. The most important field to fill is the User authorization callback URL. Enter the redirect URI.

WARNING: Up until August 2020, there is a bug in the GitHub consent page if the redirect uri is not configured properly. If it is not, UserIn won't be able to request the consent page. It will look like the browser is blocked spinning forever, waiting for a response. NOTE: The App creation form forces you to enter a Homepage URL and a Webhook URL. If you don't have any, that's not important. Just enter random URIs (e.g., Homepage URL: https://leavemealone.com Webhook URL: https://leavemealone.com)

  1. Once the the App is created, you are redirected to the App's page. In that page, copy the Client ID (i.e., the Client ID) and the Client secret (i.e., the Client Secret).

Troubleshooting - GitHub consent page is not reachable. Browser stays stuck after request consent page

Up until August 2020, this seems to be a GitHub bug. The expected behavior is to reach an error page with a diagnostic and some recommendations. Instead, the browser stays stuck soinning forever. As of August 2020, this issue is most likely due to a misconfigured redirect URI in your GitHub app. Please refer to step 3 above.

References