openvidu-insecure-angular

Check it on GitHub

A client-side only application built with Angular framwork.

If it is the first time you use OpenVidu, it is higly recommended to start with openvidu-hello-world tutorial, as this app is no more than an extension of it with some new features and sytles.

This is the Angular version of openvidu-insecure-js. Try it if you plan to use Angular framework for your frontend.

Understanding this tutorial

OpenVidu is composed by the three modules displayed on the image above in its insecure version.

  • openvidu-browser: NPM package for your Angular app. It allows you to manage your video-calls straight away from your clients
  • openvidu-server: Java application that controls Kurento Media Server
  • Kurento Media Server: server that handles low level operations of media flow transmissions
Tutorial's name includes "insecure" word because this application has no backend and therefore it has no control over the users. Typically you don't want such application in production environments. When you feel comfortable with the client-side of OpenVidu, add your own server or follow one of our super simple secure tutorials.

Running this tutorial

1) Clone the repo and checkout to version v1.1.0:

git clone https://github.com/OpenVidu/openvidu-tutorials.git
cd openvidu-tutorials && git checkout v1.1.0

2) You will need angular-cli to serve the Angular frontend. You can install it with the following command:

npm install -g @angular/cli

3) Run the tutorial:

cd openvidu-insecure-angular
npm install
ng serve

4) openvidu-server and Kurento Media Server must be up and running in your development machine. The easiest way is running this Docker container which wraps both of them (you will need Docker CE):

docker run -p 8443:8443 --rm -e KMS_STUN_IP=stun.l.google.com -e KMS_STUN_PORT=19302 -e openvidu.secret=MY_SECRET openvidu/openvidu-server-kms:1.1.0

5) Go to localhost:4200 to test the app once the server is running. The first time you use the docker container, an alert message will suggest you accept the self-signed certificate of openvidu-server when you first try to join a video-call.


To learn some tips to develop with OpenVidu, check this FAQ

If you are using Windows, read this FAQ to properly run the tutorial

Understanding the code

This is an Angular project generated with angular-cli, and therefore you will see lots of configuration files and other stuff that doesn't really matter to us. We will focus on the following files under src/app/ folder:

  • app.component.ts: AppComponent, main component of the app. It contains the functionalities for joining a video-call and for handling the video-calls themselves.
  • app.component.html: HTML for AppComponent.
  • app.component.css: CSS for AppComponent.
  • stream.component.css: StreamComponent, auxiliary component to manage Stream objects on our own. It wraps the final HTML <video> which will display the video of its Stream property, as well as the user's nickname in a <p> element.

Let's see how app.component.ts uses NPM package openvidu-browser:


First line imports the necessary objects from openvidu-browser:

import { OpenVidu, Session, Stream } from 'openvidu-browser';

app.component.ts declares the following properties:

// OpenVidu objects
OV: OpenVidu;
session: Session;

// Streams to feed StreamComponent's
remoteStreams: Stream[] = [];
localStream: Stream;

// Join form
sessionId: string;
token: string;

// Main video of the page, will be 'localStream' or one of the 'remoteStreams',
// updated by an Output event of StreamComponent children
@Input() mainVideoStream: Stream;

OpenVidu object will allow us to get a Session object, which is declared just after it. remoteStreams array will store the active streams of other users in the video-call and localStream will be your own local webcam stream. Finally, sessionId and token params simply represent the video-call and your participant's nickname, as you will see in a moment.


Whenever a user clicks on the submit input defined in app.component.html, joinSession() method is called:

// --- 1) Get an OpenVidu object and init a session with a sessionId ---

// Init OpenVidu object
this.OV = new OpenVidu();

// We will join the video-call "sessionId". As there's no server, this parameter must start with the URL of
// OpenVidu Server (with secure websocket protocol: "wss://") and must include the OpenVidu secret at the end
this.session = this.OV.initSession('wss://' + location.hostname + ':8443/' + this.sessionId + '?secret=MY_SECRET');

Session's identifiers must begin with the URL where openvidu-server listens, so they can connect through WebSocket to it. It is necessary to explicitly set this URL in the param when using a pure frontend web. Since we are in a local sample app, OV.initSession will finally receive wss://localhost:8443/ as its openvidu-server URL. sessionId is the distinctive portion of the session identifier and allows OpenVidu to differentiate sessions from each other. In this case, this parameter is retrieved from HTML input <input class="form-control" type="text" id="sessionId" ... >, which may be filled by the user. Finally, '?secret=MY_SECRET' string allows us to connect to OpenVidu directly from the browser, without a server side.

WARNING: this is only for demos and developing environments. Do NOT include your secret in production. Check this FAQ to learn more.
// --- 2) Specify the actions when events take place ---

// On every new Stream received...
this.session.on('streamCreated', (event) => {

    // Add the new stream to 'remoteStreams' array
    this.remoteStreams.push(event.stream);

    // Subscribe to the Stream to receive it. Second parameter is an empty string 
    // so OpenVidu doesn't create an HTML video by its own 
    this.session.subscribe(event.stream, '');
});

// On every Stream destroyed...
this.session.on('streamDestroyed', (event) => {

    // Avoid OpenVidu trying to remove the HTML video element
    event.preventDefault();

    // Remove the stream from 'remoteStreams' array
    this.deleteRemoteStream(event.stream);
});

Here we subscribe to the Session events that interest us. As we are using Angular framework, a good approach will be treating each Stream as a component, contained in a StreamComponent. Thus, we need to store each new stream we received in an array (remoteStreams), and we must remove from it every deleted stream whenever it is necessary. To achieve this, we use the following events:

  • streamCreated: for each new Stream received by OpenVidu, we store it in our remoteStreams array and immediately subscribe to it so we can receive its video (empty string as second parameter, so OpenVidu doesn't create an HTML video on its own). HTML template of AppComponent will show the new video, as it contains an ngFor directive which will create a new StreamComponent for each Stream object stored in the array:

    <div *ngFor="let s of this.remoteStreams" class="stream-container col-md-6 col-xs-6">
      <stream-component [stream]="s" (mainVideoStream)="getMainVideoStream($event)"></stream-component>
    </div>
    
  • streamDestroyed: for each Stream that has been destroyed (which means a user has left the video-call), we remove it from remoteStreams array, so Angular will automatically delete the required StreamComponent from HTML. We call event.preventDefault() to cancel OpenVidu default behaviour towards streamDestroyed event, which is the deletion of the previously created HTML video element on streamCreated event. Because we are handling the video elements by ourselves taking advantage of Angular capabilities, we tell OpenVidu not to create them on streamCreated and not to delete them on streamDestroyed, by passing an empty string as second parameter on Session.subscribe() method on streamCreated and by calling event.preventDefault() on streamDestroyed.


Finally connect to the session and publish your webcam:


// --- 3) Connect to the session ---

// First param irrelevant if your app has no server-side. Second param will be received by every user
// in Stream.connection.data property, which will be appended to DOM as the user's nickname
this.session.connect(null, '{"clientData": "' + this.userName + '"}', (error) => {

    // If connection successful, initialize a publisher and publish to the session
    if (!error) {

        // --- 4) Get your own camera stream with the desired resolution ---

        // Both audio and video will be active. Second parameter is an empty string
        // so OpenVidu doesn't create an HTML video by its own
        let publisher = this.OV.initPublisher('', {
            audio: true,
            video: true,
            quality: 'MEDIUM'
        });

        // Store your webcam stream in 'localStream' object
        this.localStream = publisher.stream;
        // Set the main video in the page to display our webcam
        this.mainVideoStream = this.localStream;

        // --- 5) Publish your stream ---

        this.session.publish(publisher);

    } else {
        console.log('There was an error connecting to the session:', error.code, error.message);
    }
});

In session.connect method: first param is irrelevant when you don't have a backend (it is the user's token). Remember videoElementCreated event, when we added the user's nickname to the HTML? Well, second parameter is the actual value you will receive in Stream.connection.data property. So in this case it is a JSON formatted string with a "clientData" tag with "token" value, which is retrieved from HTML input <input type="text" id="participantId" required> (filled by the user and also reused for the first token param).

In the callback of Session.connect method, we check the connection has been succesful (error value must be null) and right after that we get a Publisher object with both audio and video activated and MEDIUM quality. We then store our local Stream (contained in Publisher.stream object) in localStream, make our main video display our stream and publish the Publisher object through Session.publish() method. The rest of users will receive our Stream object and will execute their streamCreated event.

With regard to our local Stream, AppComponent's HTML template has also one StreamComponent declaration ready to show our own webcam as we did with remote streams:

<div *ngIf="this.localStream" class="stream-container col-md-6 col-xs-6">
    <stream-component [stream]="this.localStream" (mainVideoStream)="getMainVideoStream($event)"></stream-component>
</div>

Last point worth considering is the implementation of StreamComponent. As we are handling Stream objects by ourselves (task which usually is taken care by OpenVidu), and because the URL of Stream objects takes some time to get its final value as the WebRTC negotiation takes place, we must listen to any change in stream @Input property. We do so getting the HTMLVideoElement from our view on ngAfterViewInit() hook method (attribute videoElement), and then listening to ngDoCheck(). This allows us to update videoElement.srcObject value, which is the ultimate property that indicates our<video> element where to receive the media stream. If we didn't do this, the Stream object will update its srcObject property, but our StreamComponent would keep the same initial videoElement.srcObject value. This ensures that all our StreamComponent's will properly display all the videos in the video-call using the correct srcOjbect value.

getNickNameTag() method feeds the view of StreamComponent with the nickName of the user. Remember session.connect method and its second param? It can be now found at stream.connection.data, so every user will receive the nickName of others.

videoClicked() tells our AppComponent parent that the user has clicked on certain video, and that the main view should update the main video stream.

ngAfterViewInit() { // Get HTMLVideoElement from the view
    this.videoElement = this.elementRef.nativeElement;
}

ngDoCheck() { // Detect any change in 'stream' property (specifically in its 'srcObject' property)
    if (this.videoElement && (this.videoElement.srcObject !== this.stream.getVideoSrcObject())) {
        this.videoElement.srcObject = this.stream.getVideoSrcObject();
    }
}

getNicknameTag() { // Gets the nickName of the user
    return JSON.parse(this.stream.connection.data).clientData;
}

videoClicked() { // Triggers event for the parent component to update its view
    this.mainVideoStream.next(this.stream);
}

Leaving the session:

Whenever we want a user to leave the session, we just need to call session.disconnect method:

leaveSession() {

    // --- 6) Leave the session by calling 'disconnect' method over the Session object ---
    if (this.session) { this.session.disconnect(); };

    ...
}