alienpavlov/webrtc-test-suite

Capability testing and Tools for WebRTC 📹 🎤 🔬

Real life Capability testing and utilities for WebRTC.

Testing webRTC capabilities by trying to use them. Not a feature detector. A capability tester.Also packs some utilities to make developer's live easier while making webRTC tools.

🚀 Test Your Browser

What Can It do

• Test basic feature support.
• Test If getUserMedia Actually works.
• Test if the browser and internet is capable of RTCPeerConnection
• Test if the internet Speed is good enough for WebRTC streaming.
• Handle API differences for getUserMedia, adding stream to DOM.
• Provide utility functions for webRTC media application.

Installation:

Using a package manager:

To install the package use this command.

yarn add webrtc-test-suite

Then you can import it in your code like this:

import * as _rtc from "webrtc-test-suite";

Including the JS file directly:

Add this to your HTML file:

<script src="https://unpkg.com/webrtc-test-suite@2.1.2/dist/index.js"></script>

You will get a global object called: _rtc. And you can access all the functionalities from that object.

How to use

This tool comes with a lot of capability test and utility functions. You can use them to create WebRTC enabled application and positively determine feature support. All the functions are described below.

Please include the webRTC Adapter package in your project. This plugin tries to cover most of the variations of the API but adapter covers almost all of it.

Functions that return a promise has a silent version that does not reject the promise on error. Instead returns null. Good for working with async-await.

Functions that accepts the verbose (Boolean) argument, will generate logs in the console if verbose is set to true. Default is false.

Function that accepts a timeout, that will automatically reject if the internal request does not fulfill before time. setting the timeout to 0 will disable timeout. The default timeout for peer connection related functions is 30000ms (30 seconds) and for media capture is 60000ms (1 minute).

0. checkFeatureSupport:

Please note: Feaeture detection is not the primary objective of this tool, Detecting if the feature actually works is the primary objective. Feature detection is provided just as an additional tool

checkFeatureSupport([verbose = false]) // Returns result object.

This is the newest addition to this tool in version 1.2.9. This checks for the feature support in the browser. (e.g: if the browser supports HTML5 video and Audio elements or RTCPeerConnection). Much like Modernizr. This wasn't primarily intended to be in this package since there's already tool like Modernizr that does this job really well. But since this detection is intended to be used internally, and it's always good to have one less dependency. It returns an output like this:

{
        video : {
            basic    : true
        },
        audio : {
            basic    : true,
            webAudio : true
        },
        rtcPeerConnection : true,
        rtcDataChannel    : false,
        getUserMedia      : "prefix-webkit",
        getDisplayMedia   : false
    }

Checks available:

Check	Meaning
video.basic	Basic HTML5 Video Support
audio.basic	Basic HTML5 Audio Support
audio.webAudio	Support for Web Audio API
rtcPeerConnection	Support for RTCPeerConnection API
rtcDataChannel	Support for RTC Data Channel API
getUserMedia	Support for the Audio Video Capture
getDisplayMedia	Support for Screen Capture

All the options can have these values:

Value	Meaning
false	Unsupported
"old"	Supported, but with older version of the API
"prefix-webkit"	Supported with webkit prefix
"prefix-moz"	Supported with moz prefix

1. checkMediaCapture and checkMediaCaptureSilent:

checkMediaCapture(constraints, [verbose = false,getStream = false, timeout = 60000]); // Returns Promise

Example Use:

_rtc.checkMediaCapture({audio: true, video: true})
    .then(()=>console.log("Could capture media stream"))
    .catch(()=>console.error("Could not capture media stream"));

This function takes MediaTrackConstraints as argument, calls getUserMedia API with those constraints, retrieves the Media stream, Checks if audio and video stream is active and according to the constraints provided. Then automatically stops the media capture and returns the result. If getStream is set to true, the mediaStream is not stopped, it's returned instead, on success.

2. checkPeerConnection and checkPeerConnectionSilent:

checkPeerConnection(RTCConfiguration, [verbose = false, timeout = 30000]) // Returns Promise

Example Use:

_rtc.checkPeerConnection({})
    .then(()=>console.log("Peer connection works"))
    .catch(()=>console.log("Peer connection does not work"));

This function takes RTCConfiguration as argument Creates two RTCPeerConnection with the provided RTCConfiguration and creates a data channel between those two. Check if data transfer is possible between the two RTCPeerConnection instances. And returns the results.

Tip: If you want to test your STUN (relay) server, pass iceTransportPolict: "relay" (See Documentation) with your RTCConfiguration. This will force the two PeerConnection to communicate through the relay server.

3. checkRelayPerformance and checkRelayPerformanceSilent:

checkRelayPerformance(RTCConfiguration, [verbose = false, timeout = 30000]) // Returns Promise

Example Use:

_rtc.checkRelayPerformance(rtcConfig)
    .then(()=>console.log("Peer connection works"))
    .catch(()=>console.log("Peer connection does not work"));

This function takes RTCConfiguration as argument. Which is mandatory.It also requires the user to supply at least one TURN server configuration, Against which the performance will be measured.This creates two RTCPeerConnection with the provided RTCConfiguration and creates a data channel between those two. Then transfers a relatively large random data between the two over the relay server, measures how much time it took to transfer the data. And returns the results.

Sample Output:

{
    elapsed : 305, // ms
    speed   : 5.8  // mbps
}

If you are willing to run this test, you can checkPeerConnection. Also it shows an estimated result of what was observed at that time.

4. checkInternetSpeed and checkInternetSpeedSilent:

checkInternetSpeed("probe/file.url", [verbose]) // Returns Promise

Example use:

_rtc.checkInternetSpeed("https://example-file.com/file.jpg")
    .then(speed=>console.log(`Your speed is ${speed}mbps`))
    .catch(()=>console.log("could not test Internet Speed"))

This function takes a file URL (Give at least >1mb for better results), somewhere in the web (better if it's in the same server as your TURN server), downloads the file and observes the download speed. This function makes use of the fetch API, so won't work with browsers that doesn't have fetch support (you can use a polyfill). This function returns the internet speed in mbps.

One might Argue, Internet speed is not part of WebRTC. Well, If you don't have a decent internet connection WebRTC applications might now work. And it's always good to know if it will work or not. Hence, this function got a place here.

Make sure the file you supplied isn't too large >2mb and also make sure the file is CORS enabled (has access-control-allow-origin header);

Please note: The users's actual internet speed and speed measured here can be different. This measures internet speed between the user's computer and the server the file was in and can be affected by a lot of factors. In my testing roughly ~1mbps speed was enough for a smooth video call.

Also, transfer rate over WebRTC can be significantly different from the measured internet speed, as it may or may not involve the server and may use a different protocol.

This function can only give and estimate of the observed download speed. Upload speed is not measured.

5. countDevies and countDeviesSilent:

countDevies([verbose = false]) // Returns Promise

Example use:

_rtc.countDevices()
    .then(result=>console.log(`You have ${result.audio.in} audio input devices`)
    .catch(()=>console.log("device count failed"));

This function counts all the audio video input output devices available(sort of).It returns an object like this:

{
    audio   : {in: 0, out: 0},
    video   : {in: 0, out: 0},
    unknown : 0
}

A point to note here: The current API does not give a count of video output devices In some cases doesn't give count for audio output devices too, so these counts will be 0 most of the time. The value is put there just for aesthetics. Besides, if you can see the output on your display, You definitely have at least one video output, so nothing to freak out 🤞🏼

Utilities

These functions are internally used and are exposed to make RTC application development easier.

6. getUserMedia and getUserMediaSilent:

getUserMedia(constraints, [verbose]) // Returns Promise

Example use:

_rtc.getUserMedia(MediaTrackConstraints)
    .then(stream=>{
        document.querySelector(".video").srcObject = stream;
    })
    .catch(()=>console.log("could not get media stream"));

If you are tired of handling different versions of getUserMedia, webkitGetUserMedia and the latest mediaDevices.getUserMedia, this function handles it for you. No matter what version of the API your browser supports, this function will call that version of the API and returns a promise with your media stream (or error).

this function takes MediaTrackConstraints as argument.

7. createRTCPeerConnection:

createRTCPeerConnection(peerConfiguration)

This function creates a new RTCPeerConnection instance. It handles the API variations of RTCPeerConnection, webkitRTCPeerConnection and mozRTCPeerConnection. Returns null if none are supported.

Misc Utility Functions:

This tool also comes with some utility functions for the app developer's convenience. The functions were made for internal use of the tool and then provided for the end user.

The utils object:

Smaller utilities are in the utils object. Mostly related to webRTC.

1. utils.flat:

Returns Promise

this function takes a promise as argument and returns another promise. If the source promise is resolved, this functions's promise resolves with the result. If the source promise is rejected. This function's promise resoves with null. If you are working with async-await this will save you a lot of try-catch block. This works with any types of promises. This function is used to generate all the silent versions of this tool.

2. utils.dom:

This consists of two functions. Both are used to attach and detatch media stream to a dom element (video or audio tag). Why you may ask, because, new implementation has the srcObject property, older browsers had to use createObjectURI function to convert the stream into a Object URI. this will handle the variation for you. Both functions return the DOM element.

1. utils.dom.addStreamToDOM(domElement, stream):

Adds a media stream to a DOM element. Example use:

// Let's assume we got a media stream called `stream`
let domElement = document.querySelector("video.test");
_rtc.utils.dom.addStreamToDOM(domElement,stream);

2. utils.dom.addStreamToDOM(domElement, stream):

Removes any media astream from DOM element. Example use:

// Let's assume we got a media stream called `stream`
let domElement = document.querySelector("video.test");
_rtc.utils.dom.removeStreamToDOM(domElement);

3. utils.stream:

this object only contains one utility function now. This namespace is kept to add more functions later.

1. utils.stream.stopMediaStream and utils.stream.stopMediaStreamSilent:

This function takes a media Stream and stops all the tracks associated with it. This also releases the input devices. This is the IDEAL way to stop a media stream once you are done with it.

this does not return a promise but the silent version is there to automatically handle any errors that may appear (since they are mostly non important)

Example use:

// For example, if we have a media stream named `stream`.
_rtc.utils.stream.stopMediaStreamSilent(stream);

---

This package is released under the MIT license, feel free to contribute.

made with 🖤 and JavaScript by Anam Ahmed.