We Keep Coding

How to cancel a wasm process from within a webworker

I have a wasm process (compiled from c++) that processes data inside a web application. Let's say the necessary code looks like this:
std::vector<JSONObject> data
for (size_t i = 0; i < data.size(); i++)
{
process_data(data[i]);
if (i % 1000 == 0) {
bool is_cancelled = check_if_cancelled();
if (is_cancelled) {
break;
}
}
}
This code basically "runs/processes a query" similar to a SQL query interface:
However, queries may take several minutes to run/process and at any given time the user may cancel their query. The cancellation process would occur in the normal javascript/web application, outside of the service Worker running the wasm.
My question then is what would be an example of how we could know that the user has clicked the 'cancel' button and communicate it to the wasm process so that knows the process has been cancelled so it can exit? Using the worker.terminate() is not an option, as we need to keep all the loaded data for that worker and cannot just kill that worker (it needs to stay alive with its stored data, so another query can be run...).
What would be an example way to communicate here between the javascript and worker/wasm/c++ application so that we can know when to exit, and how to do it properly?
Additionally, let us suppose a typical query takes 60s to run and processes 500MB of data in-browser using cpp/wasm.
Update: I think there are the following possible solutions here based on some research (and the initial answers/comments below) with some feedback on them:
Use two workers, with one worker storing the data and another worker processing the data. In this way the processing-worker can be terminated, and the data will always remain. Feasible? Not really, as it would take way too much time to copy over ~ 500MB of data to the webworker whenever it starts. This could have been done (previously) using SharedArrayBuffer, but its support is now quite limited/nonexistent due to some security concerns. Too bad, as this seems like by far the best solution if it were supported...
Use a single worker using Emterpreter and using emscripten_sleep_with_yield. Feasible? No, destroys performance when using Emterpreter (mentioned in the docs above), and slows down all queries by about 4-6x.
Always run a second worker and in the UI just display the most recent. Feasible? No, would probably run into quite a few OOM errors if it's not a shared data structure and the data size is 500MB x 2 = 1GB (500MB seems to be a large though acceptable size when running in a modern desktop browser/computer).
Use an API call to a server to store the status and check whether the query is cancelled or not. Feasible? Yes, though it seems quite heavy-handed to long-poll with network requests every second from every running query.
Use an incremental-parsing approach where only a row at a time is parsed. Feasible? Yes, but also would require a tremendous amount of re-writing the parsing functions so that every function supports this (the actual data parsing is handled in several functions -- filter, search, calculate, group by, sort, etc. etc.
Use IndexedDB and store the state in javascript. Allocate a chunk of memory in WASM, then return its pointer to JavaScript. Then read database there and fill the pointer. Then process your data in C++. Feasible? Not sure, though this seems like the best solution if it can be implemented.
[Anything else?]
In the bounty then I was wondering three things:
If the above six analyses seem generally valid?
Are there other (perhaps better) approaches I'm missing?
Would anyone be able to show a very basic example of doing #6 -- seems like that would be the best solution if it's possible and works cross-browser.

For Chrome (only) you may use shared memory (shared buffer as memory). And raise a flag in memory when you want to halt. Not a big fan of this solution (is complex and is supported only in chrome). It also depends on how your query works, and if there are places where the lengthy query can check the flag.
Instead you should probably call the c++ function multiple times (e.g. for each query) and check if you should halt after each call (just send a message to the worker to halt).
What I mean by multiple time is make the query in stages (multiple function cals for a single query). It may not be applicable in your case.
Regardless, AFAIK there is no way to send a signal to a Webassembly execution (e.g. Linux kill). Therefore, you'll have to wait for the operation to finish in order to complete the cancellation.
I'm attaching a code snippet that may explain this idea.
worker.js:
... init webassembly
onmessage = function(q) {
// query received from main thread.
const result = ... call webassembly(q);
postMessage(result);
}
main.js:
const worker = new Worker("worker.js");
const cancel = false;
const processing = false;
worker.onmessage(function(r) {
// when worker has finished processing the query.
// r is the results of the processing.
processing = false;
if (cancel === true) {
// processing is done, but result is not required.
// instead of showing the results, update that the query was canceled.
cancel = false;
... update UI "cancled".
return;
}
... update UI "results r".
}
function onCancel() {
// Occurs when user clicks on the cancel button.
if (cancel) {
// sanity test - prevent this in UI.
throw "already cancelling";
}
cancel = true;
... update UI "canceling".
}
function onQuery(q) {
if (processing === true) {
// sanity test - prevent this in UI.
throw "already processing";
}
processing = true;
// Send the query to the worker.
// When the worker receives the message it will process the query via webassembly.
worker.postMessage(q);
}
An idea from user experience perspective:
You may create ~two workers. This will take twice the memory, but will allow you to "cancel" "immediately" once. (it will just mean that in the backend the 2nd worker will run the next query, and when the 1st finishes the cancellation, cancellation will again become immediate).

Shared Thread
Since the worker and the C++ function that it called share the same thread, the worker will also be blocked until the C++ loop is finished, and won't be able to handle any incoming messages. I think the a solid option would minimize the amount of time that the thread is blocked by instead initializing one iteration at a time from the main application.
It would look something like this.
main.js -> worker.js -> C++ function -> worker.js -> main.js
Breaking up the Loop
Below, C++ has a variable initialized at 0, which will be incremented at each loop iteration and stored in memory.
C++ function then performs one iteration of the loop, increments the variable to keep track of loop position, and immediately breaks.
int x;
x = 0; // initialized counter at 0
std::vector<JSONObject> data
for (size_t i = x; i < data.size(); i++)
{
process_data(data[i]);
x++ // increment counter
break; // stop function until told to iterate again starting at x
}
Then you should be able to post a message to the web worker, which then sends a message to main.js that the thread is no longer blocked.
Canceling the Operation
From this point, main.js knows that the web worker thread is no longer blocked, and can decide whether or not to tell the web worker to execute the C++ function again (with the C++ variable keeping track of the loop increment in memory.)
let continueOperation = true
// here you can set to false at any time since the thread is not blocked here
worker.expensiveThreadBlockingFunction()
// results in one iteration of the loop being iterated until message is received below
worker.onmessage = function(e) {
if (continueOperation) {
worker.expensiveThreadBlockingFunction()
// execute worker function again, ultimately continuing the increment in C++
} {
return false
// or send message to worker to reset C++ counter to prepare for next execution
}
}
Continuing the Operation
Assuming all is well, and the user has not cancelled the operation, the loop should continue until finished. Keep in mind you should also send a distinct message for whether the loop has completed, or needs to continue, so you don't keep blocking the worker thread.

Calling Javascript functions in async timing

I'm working on a canvas application that uses 2 web-workers to calculate some frames. You can see below how each worker sends a specific frame back to the main thread.
On the main thread I have an event listener for each worker. They listen for the frame, and when they receive it, they send it to a function called drawFrameToCanvas
This seems simple enough, but my problem is these webworkers can send their data out of order. So worker2 can finish before worker1
The way I thought to fix this problem is like this:
var desiredFrame = 0;
function drawFrameToCanvas(frameData, frameNumber){
if(frameNumber != desiredFrame){ //this is the wrong frame
//wait until it's the correct frame
setTimeout(function(){ drawFrameToCanvas(frame, frameNumber) }, 10);
}
}
However by the time frame 0 is finished, frame 1 has been requested to be drawn several times. So this didn't work. I then thought about making an object that would store the frame data and have the drawFrameToCanvas function call for this object.
However I couldn't think of a good way to do this without storing all the frames in memory as this object would always be referenced and never cleared for garbage collection.
Is there a better way to call javascript functions in these circumstances?

You can create queues for the results, and remove the data from the queues as they're processed to avoid the memory issues you mentioned. Something like this:
var worker1Results = [];
var worker2Results = [];
worker1.onmessage = function(e){
worker1Results.push(e.data);
performActionWhenBothDone();
}
worker2.onmessage = function(e){
worker2Results.push(e.data);
performActionWhenBothDone();
}
function performActionWhenBothDone(){
if (worker1Results.length && worker2Results.length) {
performAction(worker1Results.shift(), worker2Results.shift());
}
}
function performAction(worker1Result, worker2Result){
//drawFrameToCanvas code
}

simultaneous runs of two JS webworkers: one gets stuck

I'm working on a closed system web application to aid companies in their everyday online commerce chores. That means on the one hand that it won't be open to the public, on the other: it will have to deal with large amounts of data while maintaining a fluent work experience.
This is why I turned to web workers in JS to run all sorts of database access and data loading in the background.
My understanding is, that not only the main UI/main JS remains uninterrupted but also the different web workers run without hindering each other.
I now have the following setup:
mainJS: function statusCheck which runs on pageload:
function statusCheck() {
if(typeof(w__statusCheck) == "undefined") {
var w__statusCheck = new Worker("...statusCheck.js");
w__statusCheck.postMessage("go");
w__statusCheck.onmessage = function(e) {
var message = JSON.parse(e.data);
if(message.text!=undefined) displayMessage(message.text);
}
}
statusCheck.js which is the worker simply goes like this:
function checkStatus() {
console.log("statusCheck started");
// I will leave standard parts out:
// creating and testing the ajax variable against different browsers
ajaxRequest.onreadystatechange = function() {
if(ajaxRequest.readyState == 4) {
self.postMessage(ajaxRequest.responseText);
var timer;
timer = self.setTimeout(function(){
checkStatus();
}, 1000);
}
}
ajaxRequest.open("GET", "...worker_statusCheck.php", true);
ajaxRequest.send(null);
}
this.onmessage = function(e){
checkStatus();
};
As you can see, this restarts itself every second (for now). The intervall might be longer in production.
worker_statusCheck.php simply gets different things from the database and knits them into a JSON object which gives me the system status.
This works beautifully.
Now I have another worker which is supposed to get initiated by a click on a link to effectively call some php to perform actions:
mainJS loadWorker
function loadWorker(url="") {
console.log("loadWorker started");
if(url!="") {
var uniqueID = "XXX" // creating a random ID based on timestamp and Math.random()
if(typeof(window[uniqueID]) == "undefined") {
var variables = { ajaxURL: url };
window[uniqueID] = new Worker("....loadWorker.js");
window[uniqueID].postMessage(JSON.stringify(variables));
window[uniqueID].onmessage = function(e) {
var message = JSON.parse(e.data);
if(message["success"]!=undefined) {
variables["close"] = "yes";
window[uniqueID].postMessage(JSON.stringify(variables));
}
}
}
With every click on a certain link this gets called, creates a uniquely named worker, runs it, receives the data and tells the worker to close().
The php again does its thing and writes a progress update in the DB after each step of the lengthy procedure. These progress updates I fetch from the DB with the above repeating statusCheck.
Now, I can see the entries in the DB with timestamp, so I know they get written each at their time.
So, both workers do their job and run reliably. But I have noticed, that whenever I initiate the manual (randomly named) worker the statusCheck actually stops performing. It just gets stuck... I was able to confirm this with console output from both workers. So it's not the main JS that seems stuck, but the statusCheck actually pauses... and resumes when loadWorker is done.
Am I missing something fundamental here? Any insight would be appreciated since I'm new to this concept of web workers.
Thanx :)

Your question lacks resources to truly figure out what exactly goes wrong. I can concur that two web workers can operate at the same time, even with synchronous operations. I tested this for both for loops and sync XHR requests.
There are multiple things I would recommend though.
First - unless you're processing the data with some CPU heavy algorithm, web workers are waste of time. XHR requests do not block main thread (unless you explicitly ask them to).
In statusCheck() you declare var w__statusCheck which means a local variable. Therefore it will always be null as seen from outer scope. It might get garbage-collected once no code is running in the worker.
Do not use XMLHttpRequest.onreadystatechange. Use onload and onerror.
Random unique ID's for variables are almost always wrong. If you need to store the worker refference at all, either give it a reasonable name (eg. the url it's supposed to load) or use incremental id.
Do NOT stringify data that you post to web worker. It's already done for you by the browser, possibly in more optimal manner. Converting the data to something is a single most common stupid thing people do with web workers.
Also when posting question, at least make sure the code makes some sense. In your post curly braces do not match.

Alright.. I figured it out:
I was looking in all the wrong places. Turns out, I had initialized my php session in all the php scripts which are called by the workers. And my two parallel workers both called one. So the session file was locked by the first php script and the second had to wait until it was back open again. It was not the workers or the JS being hindered, it was the php.
I now took out the session initialization from my statusCheck.php and it works like a charm. I will keep it in those others that handle the user input responses because there it actually makes sense: user clicks on button "compile data XY" which is run by the worker and takes a while. Impatient as he is he already clicks the next button "show this data"... and due to the locked session file I have sort of a neat queue for those actions. :)
I still will take above recommendations to heart and see to it to improve my code. :)

What is the best way to check internet connection

I made a CMS which during operation pulls large amounts of data.
CMS is made in PHP, MySQL, jQuery, Bootstrap and use AJAX.
The problem is if you lose your internet connection can cause problems on displaying and scrolling.
I would love if there is a good way to show the error and blocks all functions on the site when there is no internet connection. When the connection is established it should be all function allowed on the site.
Thanks!
(Sorry for my bad English.)

If you are using jQuery, you can just hook on the global error handler and lock up your application when an error occurs. The lock up screen could simply ask to try again.
$( document ).ajaxError(function() {
// lock your UI here
});
Also, once the UI is locked, you could execute a function that would ping your server in an Exponential Backoff fashion and automatically unlock the application on network restore.
Locking your app can easily be done with jQuery's blockUI plugin.
Example
(function ($) {
var locked = false;
var errorRetryCount = 0;
var blockUiOptions = { message: "Oops! Could not reach the server!" };
// change this function to adjust the exponential backoff delay
function backoff(n) {
return Math.pow(2, n) * 100;
}
$(function () {
$( document ).ajaxError(function () {
var req = this;
errorRetryCount += 1;
if (!locked) {
locked = true;
$.blockUI(blockUiOptions);
}
// retry to send the request...
setTimeout(function () { $.ajax(req); }, backoff(errorRetryCount));
}).ajaxSuccess(function () {
locked && $.unblockUI();
locked = false;
errorRetryCount = 0;
});
});
})(jQuery);
Note: You may not want to retry indefinitely your request upon network failure, and would want to quit retrying at some point. Since this is out of the scope of this question, I'll leave it as it is. However, you may take a look at this related question, which may help you sort this part out.

If you're using jQuery already, you could create a simple ajax call to your server, and if it fails within a couple of seconds, either your server or the clients internet connection is down.
Something like this:
setInterval(function() {
$.ajax({
url: "https://cms.example.com/ping",
})
.fail(function( data ) {
alert('Connection lost?');
// remember do to something smart which shows the error just once
// instead of every five seconds. Increasing the interval every
// time it fails seems a good start.
});
}, 5*1000);

Using plain JavaScript and simple code:
window.navigator.onLine ? 'on' : 'off'
It supports by almost every browser, please check Can I use

edit: re-read your question and misunderstood my first pass through so this wouldn't be valid for continuous monitoring... but i'll leave it here anyways as it may be useful for someone else.
i would suggest loading a small js file that adds a class to an element of your page and then checking if that class is applied after the fact... assuming you are using jQuery
file on the remote server loaded into your page after jQuery via script tag
$('html').addClass('connected');
local code
if($('html').hasClass('connected')) {
// connected
} else {
// not connected
}

The impossible inline Javascript delay/sleep

There is a JavaScript function, of which I have zero control of the code, which calls a function that I wrote. My function uses DOM to generate an iFrame, defines it's src and then appends it to another DOM element. However, before my function returns, and thus allows continued execution of the containing function, it is imperative that the iFrame be fully loaded.
Here are the things that I have tried and why they do not work :
1. The SetTimeout option :
99.999% of the time, this is THE answer. As a matter of fact, in the past decade that I have been mentoring in JavaScript, I have always insisted that code could always be refactored to use this option, and never believed a scenario existed where that was not the case. Well, I finally found one! The problem is that because my function is being called inline, if the very next line is executed before my iFrame finishes loading, it totally neuters my script, and since the moment my script completes, the external script continues. A callback of sorts will not work
2. The "Do nothing" loop :This option you use while(//iFrame is not loaded){//do nothing}. In theory this would not return until the frame is loaded. The problem is that since this hogs all the resources, the iFrame never loads. This trick, although horribly unprofessional, dirty etc. will work when you just need an inline delay, but since I require an external thread to complete, it will not.In FF, after a few seconds, it pauses the script and an alert pops up stating that there is an unresponsive script. While that alert is up, the iFrame is able to load, and then my function is able to return, but having the browser frozen for 10 seconds, and then requiring the user to correctly dismiss an error is a no go.
3. The model dialogue :
I was inspired by the fact that the FF popup allowed the iFrame to load while halting the execution of the function, and thinking about it, I realized that it is because the modal dialogue, is a way of halting execution yet allowing other threads to continue! Brilliant, so I decided to try other modal options. Things like alert() work beautifully! When it pops up, even if only up for 1/10th of a second, the iFrame is able to complete, and all works great. And just in case the 1/10 of a second is not sufficient, I can put the model dialogue in the while loop from solution 2, and it would ensure that the iFrame is loaded in time. Sweet right? Except for the fact that I now have to pop up a very unprofessional dialogue for the user to dismiss in order to run my script. I fought with myself about this cost/benefit of this action, but then I encountered a scenario where my code was called 10 times on a single page! Having to dismiss 10 alerts before acessing a page?! That reminds me of the late 90s script kiddie pages, and is NOT an option.
4. A gazillion other delay script out there:There are about 10 jQuery delay or sleep functions, some of them actually quite cleverly developed, but none worked. A few prototype options, and again, none I found could do it! A dozen or so other libraries and frameworks claimed they had what I needed, but alas they all conspired to give me false hope.
I am convinced that since a built in model dialogue can halt execution, while allowing other threads to continue, there must be some code accessible way to do the same thing with out user input.
The Code is literally thousands upon thousands of lines and is proprietary, so I wrote this little example of the problem for you to work with. It is important to note the ONLY code you are able to change is in the onlyThingYouCanChange function
Test File :
<html>
<head>
</head>
</html>
<body>
<div id='iFrameHolder'></div>
<script type='text/javascript'>
function unChangeableFunction()
{
new_iFrame = onlyThingYouCanChange(document.getElementById('iFrameHolder'));
new_iFrame_doc = (new_iFrame.contentWindow || new_iFrame.contentDocument);
if(new_iFrame_doc.document)new_iFrame_doc=new_iFrame_doc.document;
new_iFrame_body = new_iFrame_doc.body;
if(new_iFrame_body.innerHTML != 'Loaded?')
{
//The world explodes!!!
alert('you just blew up the world! Way to go!');
}
else
{
alert('wow, you did it! Way to go!');
}
}
var iFrameLoaded = false;
function onlyThingYouCanChange(objectToAppendIFrameTo)
{
iFrameLoaded = false;
iframe=document.createElement('iframe');
iframe.onload = new Function('iFrameLoaded = true');
iframe.src = 'blank_frame.html'; //Must use an HTML doc on the server because there is a very specific DOM structure that must be maintained.
objectToAppendIFrameTo.appendChild(iframe);
var it = 0;
while(!iFrameLoaded) //I put the limit on here so you don't
{
//If I was able to put some sort of delay here that paused the exicution of the script, but did not halt all other browser threads, and did not require user interaction we'd be golden!
//alert('test'); //This would work if it did not require user interaction!
}
return iframe;
}
unChangeableFunction();
</script>
</body>
blank_frame.html :
<html>
<head>
</head>
<body style='margin:0px'>Loaded?</body>
</html>
HERE IS THE ANSWER I MADE FROM COMBINING IDEAS FROM RESPONDERS! YOU GUYS ROCK!
new source of the function I was allowed to change :
function onlyThingYouCanChange(objectToAppendIFrameTo)
{
iFrameLoaded = false;
iframe=document.createElement('iframe');
iframe.onload = new Function('iFrameLoaded = true');
iframe.src = 'blank_frame.html'; //Must use an HTML doc on the server because there is a very specific DOM structure that must be maintained.
objectToAppendIFrameTo.appendChild(iframe);
var it = 0;
while(!iFrameLoaded) //I put the limit on here so you don't
{
if (window.XMLHttpRequest)
{
AJAX=new XMLHttpRequest();
}
else
{
AJAX=new ActiveXObject("Microsoft.XMLHTTP");
}
if (AJAX)
{
AJAX.open("GET", 'slow_page.php', false);
AJAX.send(null);
}
else
{
alert('something is wrong with AJAX!');
}
//If I was able to put some sort of delay here that paused the exicution of the script, but did not halt all other browser threads, and did not require user interaction we'd be golden!
//alert('test'); //This would work if it did not require user interaction!
}
return iframe;
}
slow_page.php :
<?
usleep(100000);//sleep for 1/10th of a second, to allow iFrame time to load without DOSing our own server!
?>
I do want to note that I stated that there was nothing outside of that function that I could change, and adding the php page did violate that "rule" but in may case I was able to do that. If I were not able to do that, I could have called blank_frame.html instead of slow_page.php, and it should have only ever needed to call it once (so 2 times per frame load) assuming that it responded in an identical amount of time as the iFrame load. If for some reason the iFrame load was slower, it might call it 2ce (a total of 3 calls to the server)

Yeah, the fact that javascript is single threaded really bites you here. You can use a synchronous ajax call to a purposefully slow page to emulate a sleep, but you aren't going to get the results you want. Why don't you just make sure that your IFrame is loaded before unchangeable function is called?

NB This is extremely hacky, and I wouldn't use it in any real-world situation. Among other potential issues, given sufficient traffic you could end up DDOSing yourself.
You could create sleep functionality by making non-asynchronous (A)JAX calls. In some older browsers this may freeze everything, but at least it won't require any kind of user response.
while (!iFrameLoaded)
{
if (XMLHTTPRequest) {
var request = new XMLHttpRequest();
} else {
var request = new ActiveXObject("Microsoft.XMLHTTP");
}
request.open('GET', 'anyoldfile.htm', false);
request.send();
// check if the iframe is loaded and set iFrameLoaded
}

What you really need is an event to be fired when the iFrame content has loaded. This is actually really easy because the page inside the iFrame has its own events and it can access scripts on the parent page. You will need to be able to change the contents of the iFrame though.
In your iFrame, you'll need this piece of code
// Use whichever DOMReady function you like, or window.onload would work
window.addEventListener('DOMContentLoaded', function() {
if (parent.window.myFunction) {
parent.window.myFunction();
}
}, false);
Then in your parent page, make a function called "myFunction" and put all the scripts you need to fire in there. This should work every time.
Edit: To get this to work you really need two functions. I'm assuming that's really not an option so we'll hack the one function to contain two functions and call the right part when we need it to.
function onlyThingYouCanChange(stringOrObject) {
function createIFrame(objectToAppendIFrameTo) {
// This comment represents all the code that appends your iFrame
}
function onIFrameReady() {
// This comment represents all the stuff you want to happen when the iFrame is ready
}
// The bones of it
if (stringOrObject === "iFrameLoaded") {
onIFrameReady();
} else {
createIFrame(stringOrObject);
}
}
The script in the iFrame should now be changed to something like this:
// Use whichever DOMReady function you like, or window.onload would work
window.addEventListener('DOMContentLoaded', function() {
if (parent.window.onlyThingYouCanChange) {
parent.window.onlyThingYouCanChange('iFrameLoaded');
}
}, false);
I haven't tested it, but in theory that should do it

A stupefyingly simple ;-} answer using XPCOM:
// Get instance of the XPCOM thread manager.
var threadManager=Components.classes['#mozilla.org/thread-manager;1'].getService(
Components.interfaces.nsIThreadManager);
// Release current thread.
function doThread() {threadManager.currentThread.processNextEvent(false);};
// Event enabled delay, time in ms.
function delay(time) {
var end;
var start=Date.now();
do {
end=Date.now();
doThread();
} while ((end-start) <= time);
}
Works in recent version of Firefox. Sorry no hope for Explorer!

A recursive function might help out in this case. just call the function until a global variable indicates that the frame is loaded
var iFrameStarted = false; //you need two global vars
var iFrameLoaded = false;
function onlyThingYouCanChange(objectToAppendIFrameTo)
{
if (iFrameLoaded=false) // if the frame has loaded then you are done. skip everything and return iframe
{ if (iFrameStarted = false) //otherwise start the frame if it has not been
{
iFrameStarted = true;
iframe=document.createElement('iframe');
iframe.onload = new Function('iFrameLoaded = true');
iframe.src = 'blank_frame.html'; //Must use an HTML doc on the server because there is a very specific DOM structure
objectToAppendIFrameTo.appendChild(iframe);
var it = 0;
for (i=0;i<10000;i++) {} //slow down execution so you are not recursing yourself to death
onlyThingYouCanChange(objectToAppendIFrameTo); //start the recursion process
}
else //the frame has been started so continue recursion until the frame loaded
{
for (i=0;i<10000;i++) {} //slow down execution so you are not recursing yourself to death
onlyThingYouCanChange(objectToAppendIFrameTo); recursively call your function until the frame is loaded
}
}
return iframe; //you only get here when all the recursions are finished
}

Why can you not modify the base code? For example, it could be fairly simple to change the core function from
function unChangeableFunction()
{
new_iFrame = onlyThingYouCanChange(document.getElementById('iFrameHolder'));
new_iFrame_doc = (new_iFrame.contentWindow || new_iFrame.contentDocument);
if(new_iFrame_doc.document)new_iFrame_doc=new_iFrame_doc.document;
new_iFrame_body = new_iFrame_doc.body;
if(new_iFrame_body.innerHTML != 'Loaded?')
{
//The world explodes!!!
alert('you just blew up the world! Way to go!');
}
else
{
alert('wow, you did it! Way to go!');
}
}
To something like this:
function unChangeableFunction()
{
var new_iFrame = onlyThingYouCanChange(document.getElementById('iFrameHolder'));
new_iFrame.onload = function()
{
new_iFrame_doc = (new_iFrame.contentWindow || new_iFrame.contentDocument);
if(new_iFrame_doc.document)new_iFrame_doc=new_iFrame_doc.document;
new_iFrame_body = new_iFrame_doc.body;
if(new_iFrame_body.innerHTML != 'Loaded?')
{
//The world explodes!!!
alert('you just blew up the world! Way to go!');
}
else
{
alert('wow, you did it! Way to go!');
}
};
}
If that doesn't work for you, how about a transparent modification of the original code? Compile it with Javascript Strands and use the built-in futures support to handle this. Note that Javascript 1.7 also supports continuations, but would require changing the code manually to use them.

Another solution that may not be applicable, depending on how much you have simplified the original code. You could set an onload handler, then throw an error, then call unChangeableFunction in your onload handler:
function onlyThingYouCanChange(objectToAppendIFrameTo)
{
// using global variable func_called
if (!func_called) {
func_called = true;
var iframe=document.createElement('iframe');
iframe.src = 'blank_frame.html';
iframe.id = 'myIframe';
iframe.onload = function() {
unChangeableFunction();
};
objectToAppendIFrameTo.appendChild(iframe);
throw new Error('not an error');
} else {
return document.getElementById('myIframe');
}
}
This function (like unChangeableFunction) will be called twice: once in the first instance, then again when the onload handler is triggered. The two different pathways reflect this.
Again, this is hacky, and a definite abuse of JS's error functionality.

you can use cookie and setTimeout like that:
in blank_frame.html add a script:
<script type="text/javascript">
function deleteCookie(cookie_name)
{
var cookie_date=new Date();
cookie_date.setTime(cookie_date.getTime()-1);
document.cookie=cookie_name+="=;expires="+cookie_date.toGMTString();
}
function setCookie(name,value,expires,path,domain,secure){
document.cookie=name+"="+escape(value)+((expires)?"; expires="+expires.toGMTString():"")+((path)?"; path="+path:"")+((domain)?"; domain="+domain:"")+((secure)?"; secure":"");
}
window.onload=function(){
setCookie('iframe_loaded','yes',false,'/',false,false);
}
</script>
Basically you're adding a cookie iframe_loaded with value yes.
IMO it's better to remove the cookie as you need to do the same if you'll reload the page.
You can as well set the domain in setCookie function call.
Now in main file we'll use setTimeout with function that will check if the cookie exists, if it does then the function will return iframe like in your code:
function onlyThingYouCanChange(objectToAppendIFrameTo)
{
function get_cookie(cookie_name){
var results = document.cookie.match('(^|;) ?'+cookie_name+'=([^;]*)(;|$)');
return results?unescape(results[2]):null;
}
function deleteCookie(cookie_name){
var cookie_date=new Date();
cookie_date.setTime(cookie_date.getTime()-1);
document.cookie=cookie_name+="=;expires="+cookie_date.toGMTString();
}
iFrameLoaded = false;
iframe=document.createElement('iframe');
iframe.onload = new Function('iFrameLoaded = true');
iframe.src = 'blank_frame.html'; //Must use an HTML doc on the server because there is a very specific DOM structure that must be maintained.
objectToAppendIFrameTo.appendChild(iframe);
var it = 0;
function checkiframe(){
if(get_cookie('iframe_loaded')=="yes"){
alert('iframe loaded');
deleteCookie('iframe_loaded');
return iframe;
}else{
setTimeout(checkiframe,1000);
}
}
checkiframe();
}
As a failsafe cookie is being deleted in this file as well.
Hopefully that will give you something to work with :)
Cheers
G.