Skip to content

Latest commit

 

History

History
355 lines (296 loc) · 15.8 KB

README.md

File metadata and controls

355 lines (296 loc) · 15.8 KB

OpenTok Archiving Sample for Node

This is a simple demo app that shows how you can use the OpenTok Node SDK to archive (or record) Sessions, list archives that have been created, download the recordings, and delete the recordings.

Running the App

First, download the dependencies using npm in this directory.

$ npm install

Next, add your own API Key and API Secret to the environment variables. There are a few ways to do this but the simplest would be to do it right in your shell.

$ export API_KEY=0000000
$ export API_SECRET=abcdef1234567890abcdef01234567890abcdef

Finally, start the app using node

$ node index.js

Visit http://localhost:3000 in your browser. You can now create new archives (either as a host or as a participant) and also play archives that have already been created.

Walkthrough

This demo application uses the same frameworks and libraries as the HelloWorld sample. If you have not already gotten familiar with the code in that project, consider doing so before continuing.

The explanations below are separated by page. Each section will focus on a route handler within the main application (index.js).

Creating Archives – Host View

Start by visiting the host page at http://localhost:3000/host and using the application to record an archive. Your browser will first ask you to approve permission to use the camera and microphone. Once you've accepted, your image will appear inside the section titled 'Host'. To start recording the video stream, press the 'Start Archiving' button. Once archiving has begun the button will turn green and change to 'Stop Archiving'. You should also see a red blinking indicator that you are being recorded. Wave and say hello! Stop archiving when you are done.

Next we will see how the host view is implemented on the server. The route handler for this page is shown below:

app.get('/host', function(req, res) {
  var sessionId = app.get('sessionId'),
      // generate a fresh token for this client
      token = opentok.generateToken(sessionId, { role: 'moderator' });

  res.render('host.ejs', {
    apiKey: apiKey,
    sessionId: sessionId,
    token: token
  });
});

If you've completed the HelloWorld walkthrough, this should look familiar. This handler simply generates the three strings that the client (JavaScript) needs to connect to the session: apiKey, sessionId and token. After the user has connected to the session, they press the 'Start Archiving' button, which sends an XHR (or Ajax) request to the http://localhost:3000/start URL. The route handler for this URL is shown below:

app.post('/start', function(req, res) {
  var hasAudio = (req.param('hasAudio') !== undefined);
  var hasVideo = (req.param('hasVideo') !== undefined);
  var outputMode = req.param('outputMode');
  var archiveOptions = {
    name: 'Node Archiving Sample App',
    hasAudio: hasAudio,
    hasVideo: hasVideo,
    outputMode: outputMode,
  };
  if (outputMode === 'composed') {
    startOptions.layout = { type: 'horizontalPresentation' };
  }
  opentok.startArchive(app.get('sessionId'), archiveOptions, function(err, archive) {
    if (err) return res.send(500,
      'Could not start archive for session '+sessionId+'. error='+err.message
    );
    res.json(archive);
  });
});

In this handler, the startArchive() method of the opentok instance is called with the sessionId for the session that needs to be archived. In this case, as in the HelloWorld sample app, there is only one session created and it is used here and for the participant view. This will trigger the recording to begin.

The optional second argument is for options. The name is stored with the archive and can be read later. The hasAudio, hasVideo, outputMode, values are read from the request body; these define whether the archive will record audio and video, and whether it will record streams individually or to a single file composed of all streams. See the "Changing Archive Layout" section below for information on the layout option.

The last argument is the callback for the result of this asynchronous function. The callback signature follows the common node pattern of using the first argument fo an error if one occurred, otherwise the second parameter is an Archive object. As long as there is no error, a response is sent back to the client's XHR request with the JSON representation of the archive. The client is also listening for the archiveStarted event, and uses that event to change the 'Start Archiving' button to show 'Stop Archiving' instead. When the user presses the button this time, another XHR request is sent to the http://localhost:3000/stop/:archiveId URL where :archiveId represents the ID the client receives in the 'archiveStarted' event. The route handler for this request is shown below:

app.get('/stop/:archiveId', function(req, res) {
  var archiveId = req.param('archiveId');
  opentok.stopArchive(archiveId, function(err, archive) {
    if (err) return res.send(500, 'Could not stop archive '+archiveId+'. error='+err.message);
    res.json(archive);
  });
});

This handler is very similar to the previous one. Instead of calling the startArchive() method, the stopArchive() method is called. This method takes an archiveId as its parameter, which is different for each time a session starts recording. But the client has sent this to the server as part of the URL, so the req.param('archiveId') expression is used to retrieve it.

Now you have understood the three main routes that are used to create the Host experience of creating an archive. Much of the functionality is done in the client with JavaScript. That code can be found in the public/js/host.js file. Read about the OpenTok.js JavaScript library to learn more.

Creating Archives - Participant View

With the host view still open and publishing, open an additional window or tab and navigate to http://localhost:3000/participant and allow the browser to use your camera and microphone. Once again, start archiving in the host view. Back in the participant view, notice that the red blinking indicator has been shown so that the participant knows his video is being recorded. Now stop the archiving in the host view. Notice that the indicator has gone away in the participant view too.

Creating this view on the server is as simple as the HelloWorld sample application. See the code for the route handler below:

app.get('/participant', function(req, res) {
  var sessionId = app.get('sessionId'),
      // generate a fresh token for this client
      token = opentok.generateToken(sessionId, { role: 'moderator' });

  res.render('participant.ejs', {
    apiKey: apiKey,
    sessionId: sessionId,
    token: token
  });
});

Since this view has no further interactivity with buttons, this is all that is needed for a client that is participating in an archived session. Once again, much of the functionality is implemented in the client, in code that can be found in the public/js/participant.js file.

Changing Archive Layout

Note: Changing archive layout is only available for composed archives, and setting the layout is not required. By default, composed archives use the "best fit" layout. For more information, see the OpenTok developer guide for Customizing the video layout for composed archives.

When you create a composed archive (when the outputMode is set to 'composed), we set the layout property of the options object passed into OpenTok.startArchive() to 'horizontalPresentation'. This sets the initial layout type of the archive. 'horizontalPresentation' is one of the predefined layout types for composed archives.

For composed archives, you can change the layout dynamically. The host view includes a Toggle layout button. This toggles the layout of the streams between a horizontal and vertical presentation. When you click this button, the host client switches makes an HTTP POST request to the '/archive/:archiveId/layout' endpoint:

app.post('/archive/:archiveId/layout', function (req, res) {
  var archiveId = req.param('archiveId');
  var type = req.body.type;
  app.set('layout', type);
  opentok.setArchiveLayout(archiveId, type, null, function (err) {
    if (err) return res.send(500, 'Could not set layout ' + type + '. error=' + err.message);
    res.send(200, 'OK');
  });
});

This calls the OpenTok.setArchiveLayout() method of the OpenTok Node.js SDK, setting the archive layout to the layout type defined in the POST request's body. The layout type will either be set to horizontalPresentation or verticalPresentation, which are two of the predefined layout types for OpenTok composed archives.

Also, in the host view, you can click any stream to set it to be the focus stream in the archive layout. (Click outside of the mute audio icon.) Doing so sends an HTTP POST request to the /focus endpoint:

app.post('/focus', function (req, res) {
  var otherStreams = req.body.otherStreams;
  var focusStreamId = req.body.focus;
  var classListArray = [];
  if (otherStreams) {
    var i; 
    for (i = 0; i < otherStreams.length; i++) {
      classListArray.push({
        id: otherStreams[i],
        layoutClassList: [],
      });
    }
  }
  classListArray.push({
    id: focusStreamId,
    layoutClassList: ['focus'],
  });
  app.set('focusStreamId', focusStreamId);
  opentok.setStreamClassLists(app.get('sessionId'), classListArray, function (err) {
    if (err) return res.send(500, 'Could not set class lists. Error:' + err.message);
    return res.send(200, 'OK');
  });
});

The body of the POST request includes the stream ID of the "focus" stream and an array of other stream IDs in the session. The server-side method that handles the POST requests assembles a classListArray array, based on these stream IDs:

[
  {
    "id": "6ad90229-df4f-4849-8974-5d675727c8b5",
    "layoutClassList": []
  },
  {
    "id": "aef616a5-769c-43e9-96d2-221edb986cbf",
    "layoutClassList": []
  },
  {
    "id": "db9f2372-7564-4b38-9bb2-d6ba4249fe63",
    "layoutClassList": ["focus"]
  }
]

This is passed in as the classListArray parameter of the OpenTok.setStreamClassLists() method of the OpenTok Node.js SDK:

opentok.setStreamClassLists(app.get('sessionId'), classListArray, function (err) {
  if (err) return res.send(500, 'Could not set class lists. Error:' + err.message);
  return res.send(200, 'OK');
});

This sets one stream to have the focus class, which causes it to be the large stream displayed in the composed archive. (This is the behavior of the horizontalPresentation and verticalPresentation layout types.) To see this effect, you should open the host and participant pages on different computers (using different cameras). Or, if you have multiple cameras connected to your machine, you can use one camera for publishing from the host, and use another for the participant. Or, if you are using a laptop with an external monitor, you can load the host page with the laptop closed (no camera) and open the participant page with the laptop open.

The host client page also uses OpenTok signaling to notify other clients when the layout type and focus stream changes, and they then update the local display of streams in the HTML DOM accordingly. However, this is not necessary. The layout of the composed archive is unrelated to the layout of streams in the web clients.

When you playback the composed archive, the layout type and focus stream changes, based on calls to the OpenTok.setArchiveLayout() and OpenTok.setStreamClassLists() methods during the recording.

Past Archives

Start by visiting the history page at http://localhost:3000/history. You will see a table that displays all the archives created with your API Key. If there are more than five, the older ones can be seen by clicking the "Older →" link. If you click on the name of an archive, your browser will start downloading the archive file. If you click the "Delete" link in the end of the row for any archive, that archive will be deleted and no longer available. Some basic information like when the archive was created, how long it is, and its status is also shown. You should see the archives you created in the previous sections here.

We begin to see how this page is created by looking at the route handler for this URL:

app.get('/history', function(req, res) {
  var page = req.param('page') || 1,
      offset = (page - 1) * 5;
  opentok.listArchives({ offset: offset, count: 5 }, function(err, archives, count) {
    if (err) return res.send(500, 'Could not list archives. error=' + err.message);
    res.render('history.ejs', {
      archives: archives,
      showPrevious: page > 1 ? ('/history?page='+(page-1)) : null,
      showNext: (count > offset + 5) ? ('/history?page='+(page+1)) : null
    });
  });
});

This view is paginated so that we don't potentially show hundreds of rows on the table, which would be difficult for the user to navigate. So this code starts by figuring out which page needs to be shown, where each page is a set of 5 archives. The page number is read from the request's query string parameters. The offset, which represents how many archives are being skipped, is always calculated as five times as many pages that are less than the current page, which is (page - 1) * 5. Now there is enough information to ask for a list of archives from OpenTok, which we do by calling the listArchives() method of the opentok instance. The first parameter is an optional object to specify a count and offset. If we are not at the first page, we can pass the view a string that contains the relative URL for the previous page. Similarly, we can also include one for the next page. Now the application renders the view using that information and the partial list of archives.

At this point the template file views/history.ejs handles looping over the array of archives and outputting the proper information for each column in the table. It also places a link to the download and delete routes around the archive's name and its delete button, respectively.

The code for the download route handler is shown below:

app.get('/download/:archiveId', function(req, res) {
  var archiveId = req.param('archiveId');
  opentok.getArchive(archiveId, function(err, archive) {
    if (err) return res.send(500, 'Could not get archive '+archiveId+'. error='+err.message);
    res.redirect(archive.url);
  });
});

The download URL for an archive is available as a property of an Archive instance. In order to get an instance to this archive, the getArchive() method of the opentok instance is used. The first parameter is required and it is the archiveId. We use the same technique as above to read that archiveId from the URL. The second parameter is a callback function, whose signature has arguments for the error object and the resulting archive instance. Lastly, we send a redirect response back to the browser with the archive's URL so the download begins.

The code for the delete route handler is shown below:

app.get('/delete/:archiveId', function(req, res) {
  var archiveId = req.param('archiveId');
  opentok.deleteArchive(archiveId, function(err) {
    if (err) return res.send(500, 'Could not stop archive '+archiveId+'. error='+err.message);
    res.redirect('/history');
  });
});

Once again the archiveId is retrieved from the URL of the request. This value is then passed to the deleteArchive() method of the opentok instance. The callback only has an argument for an error in case one occurred. Now that the archive has been deleted, a redirect response back to the first page of the history is sent back to the browser.

That completes the walkthrough for this Archiving sample application. Feel free to continue to use this application to browse the archives created for your API Key.