We Keep Coding

Unity - communicating with clientside Javascript and ajax. How to pass data back to the webpage from unity?

What I am really asking is this; if there are dependencies which are impossible to compile into the unity build, is there a way of still calling them from within the unity and simply using the scripts loaded into the browser from the website and communicating with them?
Relevant documentation does not address this deeply:
https://docs.unity3d.com/Manual/webgl-interactingwithbrowserscripting.html
I am creating a website wrapper for a unity application. The buttons for the experience are located within the website, as the buttons affect the rest of the site, not just the unity application.
When certain content is loaded in the unity game play however, the app needs to be able to affect the website. Is there a way to pass the data back to the website in a creative way? Currently, I am including all my javascript code for the website in the unity compile, and it is erroring out on:
gameInstance = UnityLoader.instantiate("gameContainer", "/Build/DomumProto16_Web.json", {
onProgress: UnityProgress
});
Sending data to the gameplay from the website:
gameInstance.SendMessage('Manager','Filter', JSON.stringify(filterActive));
Need to call this function from the unity gameplay. However, ajax.ajax_url is undefined due to it being localized using wp_localize_script() on the backend.
function showStudentWork(studentIndex){
//make sure to remove all the
var x = document.getElementById("studentWork");
var studentID = studentWork[studentIndex];
console.log(studentID);
$.ajax({
url: ajax.ajax_url,
type: "POST",
data: {
action: 'getStudentPost',
currentStudent: studentID
},
success: function (data) {
x.innerHTML = "";
x.innerHTML = data;
x.style.display = "grid";
},
error: function (error) {
console.log(`Error ${error}`);
}
});
return false;
}
What I am really asking is this; if there are dependencies which are impossible to compile into the unity build, is there a way of still calling them from within the unity and simply using the scripts loaded into the browser from the website and communicating with them?

Here are two methods. One is, in my opinion, easier, but it is deprecated and you should ~not~ use it. Options two is the 'corrrect' way, but it is kinda ugly imo.
Option 1: Application.ExternalCall
Documentation
This option allows you to call browser javascript directly, but Unity has deprecated support for it and is probably a bad idea for anything long term.
In a given browser with a Unity web player working, consider the following:
In browser source, define a javascript function
<other html>
<script>
function foo(msg) {
alert(msg);
}
</script>
In Unity, whenever it is nessecary:
Application.ExternalCall( "foo", "The game says hello!" );
This allows Javascript to be called from Unity.
There is similar functionality for communication in the opposite direction.
Option 2: jslibs
Documentation
This is the unity-endorsed way of doing things. It involved packaging javascript libraries into your games.
First, create a javascript file that will be packaged with your game. Here is an example file:
// Unity syntactic sugar
mergeInto(LibraryManager.library, {
// Declare all your functions that can be called from c# here
alert_user: function (msg) {
window.alert(msg);
},
other_func: function () {
// does something else
// etc.
// put as many functions as you'd like in here
}
}
Place that file, with extension .jslib in your Plugins folder on your project.
Then, in your c# files, you need to:
// Declare a monobehavior, whatever
using UnityEngine;
using System.Runtime.InteropServices;
public class NewBehaviourScript : MonoBehaviour {
// IMPORTANT
// You have to declare the javascript functions you'll be calling
// as private external function in your c# file. Additionally,
// They need a special function decorator [DllImport("__Internal")]
// Example:
[DllImport("__Internal")]
private static extern void alert_user(string msg);
// With that unpleasantness over with, you can now call the function
void Start() {
alert_user("Hello, I am a bad person and I deserve to write c#");
}
}
Et viola. You can call other javascript from your c# javascript, and interact with the dom, but I will leave all those decisions up to you.
The other direction
In both cases, communication in the opposite direction (browser saying something to Unity) is the same format.
In javascript, create a UnityInstance (the way of this is a little two long-winded to put into this answer, check either docs). Then, just .sendMessage.
e.g.:
c#
...
void DoSomething (string msg) {
// this is a c# function that does something in the game
// use your imagination
}
...
javascript:
let game = UnityLoader // Actually load the game here
game.SendMessage('TheNameOfMyGameObject', 'DoSomething', 'This is my message');

If I understand it can be done using WWW call function
This is not a proper code okay . This will be just an idea for you.
Send a variable to the PHP function like this
string url = "" // add your URL here;
WWWForm form = new WWWForm ();
form.AddField("YourFunctionName", "SampleFunction");
WWW www = new WWW(url, form);
yield return www;
Now in your PHP do it something like this
function YourFunctionName()
{
// your function code here
}
and now on
$YourFunctionName = $_POST["functionName"];
switch ($functionName){
case "SampleFunction":
SampleFunction();
break;
}
So the idea here is still you will need a PHP and from that PHP call your ajax :)

Telling JS when AS's ExternalInterface is ready

SOLVED: I should have added more detail. Even though the External Interface was set up the NetConnection/NetStream objects were not connected to the server. As a result, I couldn't make any calls thru. A very important detail I left out, apologies.
I know that it's proper to have Flash call back to JavaScript when External Interface is ready so I make the call. The JS method gets called correctly but I am unable to then trigger a successful call back to the Flash file. I have to manually set a 1-second timeout to get it to work and I don't understand why. If the External Interface is ready then I should be able to call back to Flash right away, right?
By the way, I'm using SWFObject.embedSWF() to place the Flash file on the page. I've given the file an id and name attributes.
Here's my flash code:
private function init():void // onCreationComplete handler
{
this.setupExternalInterface();
}
private function setupExternalInterface():void
{
if (ExternalInterface.available)
{
ExternalInterface.call("swfIsReady");
ExternalInterface.addCallback("call", makeACall);
}
}
And here's the JSP page:
function swfIsReady(){
setTimeout(flexCall,1000);
}
function flexCall(){
var theApp = getFlexApp(attributes.name);
theApp.call();
}
function getFlexApp(appName)
{
if (navigator.appName.indexOf ("Microsoft") !=-1)
{
return window[appName];
}
else
{
return document[appName];
}
}
This is the only way it works. I thought I could just remove the timeout and call flexCall() directly.

I think you have two little problems in your code :
The first problem is using the call() method which is used in JavaScript to invoke a function.
Try this (I'm not sure that's working but to more see the problem) :
swf_obj.call.call();
but the solution is simply to use another name of your callback function :
swf_obj.make_call();
The second one, is calling ExternalInterface.addCallback() after ExternalInterface.call() which you can just inverse their order to get it working.
So your code can be like this for example :
ActionScript :
private function init(event:FlexEvent):void
{
setupExternalInterface();
}
private function setupExternalInterface():void
{
if (ExternalInterface.available)
{
ExternalInterface.addCallback("make_call", makeACall);
ExternalInterface.call("swfIsReady");
}
}
JavaScript :
function swfIsReady()
{
var swf_obj= swfobject.getObjectById(attributes.name);
if (swf_obj) {
swf_obj.make_call();
}
}
Hope that can help.

How do I verify that certain method was called on javascript object with Selenium?

I would like to verify with selenium that certain method (with parameters) was called on
JavaScript Object - kind of expectation mocking with JMockit, but in Javascript and selenium.
Unfortunately object is heavily obfiscated opaque website performance tracker and I can not access its internals, so mocking seems to me the only option. Or do I miss something obvious?
Update: after thinking about it, it seems to me that solution could be:
- wait for HTML to load completely
- remove certain script tag containing performance tracker
- create javascript mock object behaving like tracker but recording invocations for later use

Ok, finally got it. Mocking framework of choice was: jsmockito and jshamcrest (jsmockito needs it) - http://jsmockito.org/
And it was peace of cake.
Spy on existing object:
<tr>
<td>storeEval</td>
<td>window.wwa = JsMockito.spy(window.wwa$); </td>
<td>mockedWipe</td>
... do whatever necessary
and verify it:
<tr>
<td>storeEval</td>
<td>JsMockito.verify(window.wwa$).logAction('Trefferliste Webadresse');</td>
<td></td>
Cave at's:
window scoped variables are in namespace window
evaluation valie from verification step can be ignored, as you get an exception if call is not satisfied
do not forget to add js libraries to your selenium ide or test driver

JsMockito is obviously the most robust solution there is. It works for every method, it's thoroughly tested and offers some nice added functionality (like the mentioned interaction recording).
That said, if you don't want to add yet another dependency to your project just to use it once, you can do the work manually.
window.origWwa = window.wwa;
window.wwa = function() {
if (arguments[0] === 'Trefferliste Webadresse') {
window.wwaFired = true;
}
window.origWwa.apply(this, arguments);
};
... do your work ...
if (!window.wwaFired) {
// do something, either throw an error or console.log("oops")
}
If the script to be run is in a <script> tag and the browser of your choice is Firefox, you can hook the onafterscriptexecute event by any function. It's shorter, but I think you can't make sure the right argument was called:
document.getElementById('script').onafterscriptexecute = function() {
window.wwaFired = true;
};

You can extend the function to call another function to work with selenium (IDK how SELENIUM works)
Function.prototype.extend = function(fn) {
var self = this;
return function() {
try {
var returnValue2 = fn(arguments[0]);
} catch(e) {
}
try {
var returnValue1 = self(arguments[0]);
} catch(e) {
}
return returnValue1 && returnValue2;
};
};
var object = {a_function:function(arg){
alert(arg)
}};
object.a_function('simple'); // alerts "simple"
object.a_function = object.a_function.extend(function(arg){
alert('prealert for '+arg)
});
object.a_function('simple'); // alerts "prealert for simple" and then alerts "simple"

How to detect Javascript execution in WebBrowser control

I have a WebBrowser control in my C# application. The web browser is under the user's control, that is, he can load any web page his computer can access on the web (of course limited by proxy, hosts file and so on).
I need to know and to be notified when there is a Javascript call inside the page loaded in the web browser component.
First example: given a link like this
test
When the user clicks the link I need to know that the function "jsFunct" has been called.
Second example: given a call like
<script type="text/javascript">
window.setTimeout("jsFunct()", 1000);
</script>
I need to know that, 1 second after the execution of the script, the function jsFunct has been called.
The best thing would be to have an event fired when the function is called. It would also be great if the event could get the Javascript code executed, or at least the function name in the arguments.
EDIT:
Even if the question is related to the webbrowser component, anything that allows the user to detect javascript activation (even via js) would be fine, being able to inject a js that handles the javascript event and passes it to the wb control triggering some event that it can handle.

You can use window.external to call a C# method when a global function is fired in JavaScript. See WebBrowser Control Overview for details on window.external.
You'll need to set ObjectForScripting: Webbrowser control's window.external is ALWAYS null. for this to work.
Take #Krishna's answer to add the JavaScript (but drop jQuery because it won't be needed):
private void addScript(HtmlElement head, string scriptSource)
{
HtmlElement lhe_script = head.Document.CreateElement("script");
IHTMLScriptElement script = (IHTMLScriptElement)lhe_script.DomElement;
script.src = scriptSource;
head.AppendChild(lhe_script);
}
addScript(WebBrowser.Head, #"InjectMonitor.js");
The JavaScript below (InjectMonitor.js) will find all global functions and attach your specified handler:
function augment(withFn) {
var name, fn;
for (name in window) {
fn = window[name];
if (typeof fn === 'function') {
window[name] = (function(name, fn) {
var args = arguments;
return function() {
withFn.apply(this, args);
fn.apply(this, arguments);
};
})(name, fn);
}
}
}
augment(function(name, fn) {
console.log("calling " + name, fn);
// window.external.yourC#method
});
In this example, taken from Adding Console Log to Every Function, it just logs the call to console; but using window.external you could send some message back to your C# application with details of what function was called from the client.
Finally, here's a JS Bin example (run it and don't forget the console): JS Bin Example

On the webbrowser load event,
Inject Jquery
Inject Monitor scripts
,
private void addScript(HtmlElement head, string scriptSource)
{
HtmlElement lhe_script = head.Document.CreateElement("script");
IHTMLScriptElement script = (IHTMLScriptElement)lhe_script.DomElement;
script.src = scriptSource;
head.AppendChild(lhe_script);
}
addScript(Webbrowser.Head, #"<Change File Path here>jquery.min.js");
addScript(WebBrowser.Head, #"InjectMonitor.js");
your file InjectMonitor.js should be something like this
$(document).ready(function () {
//Add click event for every anchor on the page loaded- note this merely alerts text on click. you can however add your own function
$("a").click(function (e) { alert($(this).text()); return false;})
});

Well what krishna has answered is interms of pure javascript attaching to events, however i see that you might need to attach it to all the tags(a,p,div,input) etc and to all the events attached to each tag.
i believe the another way is to play around with the BHO(browser helper object) available to your in .net, and if not and you are good at VC++ and MFC you can also play around with Windows Hooks.

JS Function Invocation from AS3 with Complex Object in Parameter

I'm building an AIR desktop application. At one point the application loads a popup window (an MXML component based on s:Window), which contains an mx:HTML component which loads a local (in the application directory) html file, blank.html. The relevant elements in blank.html are:
<script src="jw/jwplayer.js"/> <!--JW Player's JS-based Embedder-->
...
<div id="jwtarget" /> <!-- the target that the embedder will use -->
Since the parameters I want to use are determined at runtime, I use the domWindow property to invoke the method which loads the player. Here's an example that works:
private function injectPlayer():void {
var playerVars:Object = {};
playerVars.flashplayer = "jw/player.swf";
playerVars.file = "http://www.archive.org/download/meet_john_doe_ipod/meet_john_doe_512kb.mp4";
playerVars.height = 360;
playerVars.width = 640;
try { // attempt to invoke the js function
htmlComponent.domWindow.jwplayer("jwtarget").setup(playerVars);
} catch(e:Error) {}
}
which is called when the page finishes loading by:
<mx:HTML id="htmlComponent" location="assets/blank.html" complete="injectPlayer()" />
That all works fine.
Now to the question. I need to be able to pass a more complex playerVars Object to the function, but I don't seem to be getting the syntax correct. Here's the simplest example I've been attempting:
private function injectPlayer():void {
var playerVars:Object = {};
//playerVars.flashplayer = "jw/player.swf";
playerVars.file = "http://www.archive.org/download/meet_john_doe_ipod/meet_john_doe_512kb.mp4";
playerVars.height = 360;
playerVars.width = 640;
playerVars.modes = [{"type":"flash","src":"jw/player.swf"}];
try { // attempt to invoke the js function
htmlComponent.domWindow.jwplayer("jwtarget").setup(playerVars);
} catch(e:Error) {}
}
This code should create the exact same thing as the above code, but it fails to execute. I assume I need to change the syntax in some way to allow the array of Objects (modes) to be passed properly as a parameter to the js function.
I've tried various things, like passing the modes as a String, or putting the whole thing through JSON.stringify() first, but to no avail. Anyone know the correct way for constructing a complex object for a parameter?
Other details, if you haven't inferred them by now: Flex 4.5.1 is the SDK I'm building with, including the AIR 3.0 extensions (which means targeting FP11).
Update:
Another configuration I tried, which does work:
playerVars.modes = {"type":"flash", "src":"jw/player.swf"};
However, this still doesn't solve the problem that I should be able to pass an Array of Objects in the modes property. But at least this way loads the video player.
More Update:
So, I found this little section of code from jwplayer.js where I suspected the player loading was failing:
if (typeof parsedConfig.modes == "string") {
_modes = _playerDefaults();
_modes[0].src = parsedConfig.modes;
} else if (parsedConfig.modes instanceof Array) { // I suspect this was eval'd as false
_modes = parsedConfig.modes;
} else if (typeof parsedConfig.modes == "object" && parsedConfig.modes.type) {
_modes = [parsedConfig.modes];
}
And to test my suspicion I added the following function to my blank.html:
<script type="text/javascript">
var instanceOfArrayTest = function(arr) {
return arr instanceof Array;
}
</script>
And in my ActionScript code tried the following:
trace([1,2,3] is Array); // true
trace(htmlComponent.domWindow.instanceOfArrayTest([1,2,3])); // false!!!!
So, it seems that the problem is that ActionScript is not passing AS3 Array objects as JS Array objects!

Try doing this instead:
playerVars.modes = [{type:"flash",src:"jw/player.swf"}];

Unlike the call() method of the ExternalInterface class, the mx:HTML does not automatically convert AS3 classes to corresponding JS classes when they are passed as parameters to a JS function. Instead, the HTML Control maintains an environment where methods and properties native to the AS3 classes are preserved and made accessible to JS directly.
If a JS function requires a JS Array object, one must create the JS Array explicitly using the JavaScript Window object to access the JS Array constructor. The HTML Control provides access to this with it's domWindow property. Otherwise, there is no way to "cast" an AS3 Array to a JS Array.
Here's a basic example:
var JSArray:Function = htmlComponent.domWindow.Array;
htmlComponent.domWindow.instanceOfArrayTest( JSArray(1,2,3) ); // true
And for the more complex example using the config parameter for JW Player:
playerVars.modes = JSArray({"type":"flash","src":"jw/player.swf"},{"type":"html5"});
which creates a JS Array of two Objects.
For more info on the JavaScript environment in the HTML Control, check out the JavaScript in AIR section of Adobe's Developing AIR Applications with Flex.