- Installation
- Getting Started
- API Calls
- Advanced SDK Configuration & Polyfills
- Making telephony calls
- Call management using JavaScript
- SMS
- MMS
- Fax
- Page visibility
- Tracking network Requests And Responses
You will need to configure SDK instance with your unique application key & secret this is provided by RingCentral's developer relations team.
This instance will be used later on to perform calls to API.
You can also supply optional polyfills and dependency injections, please read the polyfills section of this readme.
SDK can be used in 3 environments:
-
Install the NPM package:
npm install @ringcentral/sdk --save
-
Require the SDK:
import {SDK} from "@ringcentral/sdk"; // op const SDK = require('@ringcentral/sdk').SDK; const rcsdk = new SDK({ server: SDK.server.sandbox, clientId: 'yourClientId', clientSecret: 'yourClientSecret', redirectUri: '' // optional, but is required for Implicit Grant and Authorization Code OAuth Flows (see below) });
Follow installation steps for NodeJS. Don't forget to add target: 'web'
to your webpack.config.js
to tell Webpack.
Pick the option that works best for you:
-
Use CDN Attention! Versions listed here may be outdated:
-
Download everything manually:
The SDK is represented by the global RingCentral constructor. Your application must create an instance of this object:
Add the following to your HTML:
<script type="text/javascript" src="path-to-scripts/promise.auto.js"></script>
<script type="text/javascript" src="path-to-scripts/fetch.umd.js"></script>
<script type="text/javascript" src="path-to-scripts/ringcentral.js"></script><!-- or ringcentral.min.js -->
<script type="text/javascript">
var rcsdk = new RingCentral.SDK({
server: RingCentral.SDK.server.sandbox,
clientId: 'yourClientId',
clientSecret: 'yourClientSecret',
redirectUri: '' // optional, but is required for Implicit Grant and Authorization Code OAuth Flows (see below)
});
</script>
// Add this to your RequireJS configuration file
require.config({
paths: {
'ringcentral': 'path-to-scripts/ringcentral' // or ringcentral.min
}
});
// Then you can use the SDK like any other AMD component
require(['ringcentral'], function(/** @type RingCentral.sdk */ ns) {
var rcsdk = new ns.SDK({
server: ns.SDK.server.sandbox,
clientId: 'yourClientId',
clientSecret: 'yourClientSecret',
redirectUri: '' // optional, but is required for Implicit Grant and Authorization Code OAuth Flows (see below)
});
});
Make sure that polyfills are added to the page before or together with the SDK. Make sure SDK is configured to use polyfills.
Read API documentation for more information.
Login is accomplished by calling the rcsdk.login()
method of the Platform object with certain parameters depending
on the type of auth flow. A Promise
instance is returned, resolved with an AJAX Response
object.
SDK provides a set of helper functions to assist with 3-legged OAuth flow.
You can pick one of two available setups:
- Top Frame — user temporarily navigates away from the app to RingCentral login pages
- Popup — RingCentral login pages are opened in a popup
Both setups support Authorization Code and Implicit Grant flows.
This setup is good when your app uses the entire browser window.
When login is required for your application, the app's login page needs to open a URL that can be generated by SDK:
var loginUrl = rcsdk.loginUrl({implicit: true}); // implicit parameter is optional, default false
window.location.assign(loginUrl); // or .replace()
Your application need to have a special landing page to handle OAuth Redirect URI. On this page you need to resume the flow:
var loginOptions = rcsdk.parseLoginRedirect(window.location.hash || window.location.search);
rcsdk.login(loginOptions).then(...);
The above mentioned flow assumes that login page will be rendered in the top frame, e.g. user will temporarily navigate away from the app to RingCentral login pages.
This setup is good when your app is rendered as a widget on a third-party sites.
If you would like to simply open RingCentral login pages in a popup, you may use the following short-hand in your app's login page:
var loginUrl = rcsdk.loginUrl({implicit: true}); // implicit parameter is optional, default false
rcsdk
.loginWindow({url: loginUrl}) // this method also allows to supply more options to control window position
.then(function (loginOptions){
return rcsdk.login(loginOptions);
})
.then(...)
.catch(...);
In this case your landing page (the one to which Redirect URI points) need to call the following code:
RingCentral.SDK.handleLoginRedirect();
If your app uses Implicit Grant you should not provide clientSecret
when creating SDK instance. The resulting login
information will not have refresh_token
, which means that every time when access_token
expires, app should navigate
users to app login page and re-initiate login flow. Also your app will need to handle situations with refreshError
events properly, for instance, enter auth flow automatically without user intervention.
For dynamic apps with subscriptions Authorization Code should work better because it provides an ability to refresh
tokens automatically without affecting the state of the app, but it requires to expose clientSecret
.
This flow to be used only for private backend apps that can protect user credentials. Client-side apps are not allowed to use this flow.
Login is accomplished by calling the rcsdk.login()
method of the Platform object with username, extension
(optional), and password as parameters.
rcsdk
.login({
username: '18001234567', // phone number in full format
extension: '', // leave blank if direct number is used
password: 'yourpassword'
})
.then(function(response) {
// your code here
})
.catch(function(e) {
alert(e.message || 'Server cannot authorize user');
});
Because the login process is asynchronous, you need to call the promise's then
method and pass your success handler as
the continuation function.
This function will be called once login has succeeded, which allows the application to then perform updates to the user interface, and then perform the next actions using the API to load account details for the user's account and such.
Login can, of course, fail - a user can enter the incorrect password or mistype their user name.
To handle cases where login fails, you can provide an error handler function in a call to the Promise's catch
method.
To keep this example simple, a simple JavaScript alert is being used. In a real application, you will want to provide
a good UX in your login form UI.
To check in your Application if the user is authenticated, you can call the loggedIn
method of the SDK:
rcsdk.loggedIn().then(function(status){ if (status) { ... } else { ... } });
Or you can call ensureLoggedIn
method which works the same way as loggedIn
but rejects promise on failure:
rcsdk.ensureLoggedIn().then(function(){ ... }).catch(function(){ ... });
The SDK takes care of the token lifecycle. It will refresh tokens for you automatically. It will also automatically pause and queue all new API requests while the token is being refreshed in order to prevent data loss or inconsistency between SDK instances in different tabs. Paused / queued API requests will then be automatically processed once the token has been refreshed. All apropriate events will be emitted during this process.
If you just need to check whether the user has a valid token, you can call the accessTokenValid
method:
rcsdk.platform().auth().accessTokenValid(); // returns Promise<boolean>
You can retrieve save and set back the auth information:
var authData = rcsdk.platform().auth().data();
rcsdk.platform().auth().setData(authData);
It can be useful on the server if SDK instances are created and disposed for every HTTP request.
Logging the user out is trivial - just call the logout
method on the SDK:
rcsdk.logout().then(...).catch(...);
The platform provides the following events:
loginSuccess
loginError
logoutSuccess
logoutError
refreshSuccess
refreshError
— application may listen to this error and show login pagerateLimitError
To listen on platform events, you should call the on
method of the platform object:
var platform = rcsdk.platform();
platform.on(platform.events.refreshError, function(e){
// do something, usually open a login page
});
The on
method accepts an event type as its first argument and a handler function as its second argument.
To perform an authenticated API call, you should use the one of the methods of the platform object:
rcsdk
.send({
method: 'PUT',
url: '/restapi/v1.0/account/~/extension/~',
query: {...},
headers: {...},
body: {...}
})
.then(function(apiResponse){
return apiResponse.json();
})
.then(function(json){
alert(json.name);
})
.catch(function(e){
if (e.response || e.request) {
var request = e.request;
var response = e.response;
alert('API error ' + e.message + ' for URL' + request.url + ' ' + rcsdk.error(response));
} else {
alert(e.message);
}
});
If your Promise
library supports global error handler it might be useful to log Requests and Responses there.
rcsdk.get('/restapi/v1.0/account/~/extension/~', {...query}).then(...);
rcsdk.post('/restapi/v1.0/account/~/extension/~', {...body}, {...query}).then(...);
rcsdk.put('/restapi/v1.0/account/~/extension/~', {...body}, {...query}).then(...);
rcsdk.patch('/restapi/v1.0/account/~/extension/~', {...body}, {...query}).then(...);
rcsdk.delete('/restapi/v1.0/account/~/extension/~', {...query}).then(...);
json()
— if response type is JSON returns a plain JS objecttext()
— returns text representationmultipart()
— forContent-Type: multipart/mixed
responses returns an array ofApiResponse
toMultipart()
— same asmultipart()
but when response is notContent-Type: multipart/mixed
it will return an array with one item instead of errorok()
—true
for all2xx
responses,false
otherwiseerror()
— error string if anyrequest()
— low-levelRequest
objectresponse()
— low-levelResponse
object
If you need to download a binary file from API (call recording, fax attachment), you can do it as follows:
var fs = require('fs');
// read as buffer
rcsdk.get('/restapi/v1.0/account/~/messages/foo/content').then(function(res) {
return res.response().buffer(); // we are accessing Node Fetch's Response
}).then(function(buffer) {
fs.writeFileSync('./octocat.png', buffer);
});
// read as stream
rcsdk.get('/restapi/v1.0/account/~/messages/foo/content').then(function(res) {
res.response().body.pipe(fs.createWriteStream('./octocat.png')); // we are accessing Node Fetch's Response
});
See more here https://github.com/bitinn/node-fetch#usage.
rcsdk.get('/restapi/v1.0/account/~/messages/foo/content').then(function(res) {
return res.response().blob(); // or arrayBuffer(), we are accessing WhatWG Fetch's Response
}).then(function(blob ){
var img = document.createElement('img');
img.src = URL.createObjectURL(blob);
document.getElementById('container').appendChild(img);
});
See more here https://developer.mozilla.org/en-US/docs/Web/API/Response.
In any case you always can just add token to known URL of resource and download it using whatever library you want or
use directly as <img src="..."/>
:
var url = rcsdk.createUrl('/restapi/v1.0/account/~/messages/foo/content', {addServer: true, addToken: true});
Platform class emits rateLimitError
if server returns 429
status. You may supply an option handleRateLimit
and
SDK will re-execute the failed request. Please keep in mind that your application should somehow delay/throttle other
subsequent requests in this case because otherwise all those requests will be postponed too.
You can use any of your favourite fetch()
and Promise
polyfills. SDK tries to get them from global scope every
time new instance is created.
In rare case when SDK will not detect globals automatically you can set them as follows:
window.Promise = whatever;
window.fetch = whatever;
window.Headers = whatever;
window.Request = whatever;
window.Response = whatever;
Also you can manually define SDK properties:
var rcsdk = new SDK({
localStorage: whatever,
PUBNUB: whatever,
Promise: whatever,
fetch: whatever,
Headers: whatever,
Request: whatever,
Response: whatever
});
But taking into account the nature of polyfills, it's better to keep them global as described before.
In the NodeJS it might be useful to replace simple built-in storage with something persistent:
var sdk = new SDK({
localStorage: whatever
});
SDK works with localStorage
as with a simple object.
If you need to use 2 or more RingCentral accounts simultaneously, you need to create an instance of SDK for each account
and provide some unique cachePrefix
to SDK constructor (otherwise instances will share authentication).
var rcsdk1 = new SDK({
cachePrefix: 'foo-'
});
var rcsdk2 = new SDK({
cachePrefix: 'bar-'
});
In RingCentral terminology, making telephony calls is named "RingOut".
This example demonstrates a way to create a flexible RingOut tracking procedure. This is the most complex example with maximum fine-tuning - it could be simplified to suit the business requirements.
The sequence of RingOut is as follows:
- Perform a POST with the RingOut data
- Poll the RingOut status (GET requests) every second or so
Please refer to the following example:
var timeout = null; // reference to timeout object
var ringout = {}; // this is the status object (lowercase)
/**
* @param {Error} e
*/
function handleError(e) {
console.error(e);
alert(e.message);
}
function create(unsavedRingout) {
rcsdk
.post('/restapi/v1.0/account/~/extension/~/ringout', unsavedRingout)
.then(function(response) { return response.json() })
.then(function(ringout) {
console.info('First status:', ringout.status.callStatus);
update();
})
.catch(handleError);
}
/**
* @param {function(number?)} next - callback that will be used to continue polling
* @param {number} delay - last used delay
*/
function update() {
clearTimeout(timeout);
setTimeout(function() {
if (ringout.status && ringout.status.callStatus !== 'InProgress') return;
rcsdk
.get(ringout.uri)
.then(function(response) { return response.json() })
.then(function(ringout) {
console.info('Current status:', ringout.status.callStatus);
update();
})
.catch(handleError);
}, 500);
}
/**
* To stop polling, call this at any time
*/
function hangUp() {
clearTimeout(timeout);
if (ringout.status && ringout.status.callStatus !== 'InProgress') {
rcsdk
.delete(ringout.uri)
.catch(handleError);
}
// Clean
ringout = {
from: {phoneNumber: ''},
to: {phoneNumber: ''},
callerId: {phoneNumber: ''}, // optional,
playPrompt: true // optional
};
}
/**
* Start the ringout procedure (may be called multiple times with different settings)
*/
create({
from: {phoneNumber: '16501111111'},
to: {phoneNumber: '18882222222'},
callerId: {phoneNumber: '18882222222'}, // optional,
playPrompt: true // optional
});
If you are integrating with a CRM or ERP system, use of the JavaScript SDK is highly recommended. Following is an example of a call management integration that includes monitoring of incoming calls and performing of RingOuts.
A call management integration usually consists of the following tasks:
- Track the telephony status
- View the list of active calls
- View the recent calls
First, you need to load the initial Presence status (you can use Underscore or Lodash to simplify things):
var accountPresence = {};
rcsdk
.get('/restapi/v1.0/account/~/extension/~/presence?detailedTelephonyState=true')
.then(function(response) { return response.json(); })
.then(function(data) {
_.extend(accountPresence, data);
})
.catch(function(e) {
alert('Load Presence Error: ' + e.message);
});
In the meantime, you can also set up Subscriptions (you can use Underscore or Lodash to simplify things):
var subscription = rcsdk.createSubscription().addEvents(['/restapi/v1.0/account/~/extension/~/presence?detailedTelephonyState=true']);
subscription.on(subscription.events.notification, function(msg) {
_.extend(accountPresence, msg);
});
subscription.register().then(function(response) {
alert('Success: Subscription is listening');
}).catch(function(e) {
alert('Subscription Error: ' + e.message);
});
return subscription;
rcsdk.get('/restapi/v1.0/account/~/extension/~/active-calls', {query: {page: 1, perPage: 10}})
.then(function(response) { return response.json(); })
.then(function(data) {
activeCalls = data.records;
})
.catch(function(e) {
alert('Active Calls Error: ' + e.message);
});
rcsdk.get('/restapi/v1.0/account/~/extension/~/call-log', {query: {page: 1, perPage: 10}})
.then(function(response) { return response.json(); })
.then(function(data) {
calls = data.records;
})
.catch(function(e) {
alert('Recent Calls Error: ' + e.message);
});
By default, the load request returns calls that were made during the last week. To alter the time frame, provide custom
query.dateTo
and query.dateFrom
properties.
In order to send an SMS using the API, simply make a POST request to /account/~/extension/~/sms
:
rcsdk.post('/restapi/v1.0/account/~/extension/~/sms', {
from: {phoneNumber:'+12223334444'}, // Your sms-enabled phone number
to: [
{phoneNumber:'+15556667777'} // Second party's phone number
],
text: 'Message content'
})
.then(function(response) { return response.json(); })
.then(function(json) {
alert('Success: ' + json.id);
})
.catch(function(e) {
alert('Error: ' + e.message);
});
Fax endpoint understands multipart/form-data
requests. First part must always be JSON-encoded information about the
fax. Other parts should have filename
defined in order to be correctly presented in Service Web.
Modern browsers have FormData
class which could be used for sending faxes.
var body = {
to: [{phoneNumber: '123'}], // see all available options on Developer Portal
faxResolution: 'High'
},
formData = new FormData();
// This is the mandatory part, the name and type should always be as follows
formData.append('json', new File([JSON.stringify(body)], 'request.json', {type: 'application/json'}));
// Find the input[type=file] field on the page
var fileField = document.getElementById('input-type-file-field');
// Iterate through all currently selected files
for (var i = 0, file; file = fileField.files[i]; ++i) {
formData.append('attachment', file); // you can also use file.name instead of 'attachment'
}
// To send a plain text
formData.append('attachment', new File(['some plain text'], 'text.txt', {type: 'application/octet-stream'}));
// Send the fax
rcsdk.post('/restapi/v1.0/account/~/extension/~/fax', formData);
As similar to the fax endpoint, MMS
understands multipart/form-data
requests. In order to send an MMS using the API, simply make a POST request to /account/~/extension/~/sms
:
var body = {
from: {phoneNumber: '+12223334444'}, //// Your mms-enabled phone number
to: [{phoneNumber: '123'}], // see all available options on Developer Portal
faxResolution: 'High'
},
formData = new FormData();
// This is the mandatory part, the name and type should always be as follows
formData.append('json', new File([JSON.stringify(body)], 'request.json', {type: 'application/json'}));
// Find the input[type=file] field on the page
var fileField = document.getElementById('input-type-file-field');
// Iterate through all currently selected files
for (var i = 0, file; file = fileField.files[i]; ++i) {
formData.append('attachment', file); // you can also use file.name instead of 'attachment'
}
// Send the mms
rcsdk.post('/restapi/v1.0/account/~/extension/~/sms', formData);
In order to identify the MMS-Enabled phone numbers on an extension, simply make a GET request to /account/~/extension/~/phone-number
var mmsEnabledNumbers = [];
rcsdk
.get('/restapi/v1.0/account/~/extension/~/phone-number', {'perPage': 'max'})
.then(function(response) { return response.json(); })
.then(function(data) {
var phoneNumbers = data.records;
for (var i = 0; i < phoneNumbers.length; i++ ) {
if (phoneNumbers[i].features.indexOf("MmsSender") != -1 ) {
mmsEnabledNumbers.push(phoneNumbers[i].phoneNumber);
}
}
})
.catch(function(e) {
alert('MMS Enabled Phone Number Population Error:\n\n' + e.message);
});
Further reading:
SDK is capable of sending FormData
objects created by form-data module.
First, you will need to install it:
npm install form-data
Then you can build your fax, but keep in mind that FormData API in NodeJS is slightly different from the browser:
var FormData = require('form-data'),
body = {
to: [{phoneNumber: '123'}], // see all available options on Developer Portal
faxResolution: 'High'
},
formData = new FormData();
// This is the mandatory part, the name and type should always be as follows
formData.append('json', Buffer(JSON.stringify(body)), {filename: 'request.json', contentType: 'application/json'});
// To send a plain text
formData.append('attachment', Buffer('some plain text'), {filename: 'text.txt', contentType: 'text/plain'});
// To send a file from file system
formData.append('attachment', require('fs').createReadStream('/foo/bar.jpg'));
// Send the fax
rcsdk.post('/restapi/v1.0/account/~/extension/~/fax', formData);
Further reading:
You can use any of the libraries that work with the Page Visibility API, such as visibility.js.
This allows tracking the visibility of the page/tab/window/frame so that the application can react accordingly. Following are some actions that the application may wish to take whenever it becomes visible:
- Check authentication
- Reload/resync time-sensitive information from the server
- Send heartbeats to the server
Another usage is to reduce the number of Call Log or Messages reloads when the application is not visible. The SDK does not require that any such optimizations be implemented in the application, but it is considered good practice.
You can set up tracking for all network requests (for instance, to log them somewhere) by obtaining a Client
object
and registering observers on its various events:
var client = rcsdk.client();
client.on(client.events.beforeRequest, function(apiResponse) {}); // apiResponse does not have response at this point
client.on(client.events.requestSuccess, function(apiResponse) {});
client.on(client.events.requestError, function(apiError) {});