问题
I'm dynamically adding <script> tags to a page's <head>, and I'd like to be able to tell whether the loading failed in some way -- a 404, a script error in the loaded script, whatever.
In Firefox, this works:
var script_tag = document.createElement('script');
script_tag.setAttribute('type', 'text/javascript');
script_tag.setAttribute('src', 'http://fail.org/nonexistant.js');
script_tag.onerror = function() { alert("Loading failed!"); }
document.getElementsByTagName('head')[0].appendChild(script_tag);
However, this doesn't work in IE or Safari.
Does anyone know of a way to make this work in browsers other than Firefox?
(I don't think a solution that requires placing special code within the .js files is a good one. It's inelegant and inflexible.)
回答1:
There is no error event for the script tag. You can tell when it is successful, and assume that it has not loaded after a timeout:
<script type="text/javascript" onload="loaded=1" src="....js"></script>
回答2:
If you only care about html5 browsers you can use error event (since this is only for error handling, it should be ok to only support this on next gen browsers for KISS IMHO).
From the spec:
If the src attribute's value is the empty string or if it could not be resolved, then the user agent must queue a task to fire a simple event named error at the element, and abort these steps.
~
If the load resulted in an error (for example a DNS error, or an HTTP 404 error) Executing the script block must just consist of firing a simple event named error at the element.
This means you don't have to do any error prone polling and can combine it with async and defer attribute to make sure the script is not blocking page rendering:
The defer attribute may be specified even if the async attribute is specified, to cause legacy Web browsers that only support defer (and not async) to fall back to the defer behavior instead of the synchronous blocking behavior that is the default.
More on http://www.w3.org/TR/html5/scripting-1.html#script
回答3:
The script from Erwinus works great, but isn't very clearly coded. I took the liberty to clean it up and decipher what it was doing. I've made these changes:
- Meaningful variable names
- Use of
prototype. require()uses an argument variable- No
alert()messages are returned by default - Fixed some syntax errors and scope issues I was getting
Thanks again to Erwinus, the functionality itself is spot on.
function ScriptLoader() {
}
ScriptLoader.prototype = {
timer: function (times, // number of times to try
delay, // delay per try
delayMore, // extra delay per try (additional to delay)
test, // called each try, timer stops if this returns true
failure, // called on failure
result // used internally, shouldn't be passed
) {
var me = this;
if (times == -1 || times > 0) {
setTimeout(function () {
result = (test()) ? 1 : 0;
me.timer((result) ? 0 : (times > 0) ? --times : times, delay + ((delayMore) ? delayMore : 0), delayMore, test, failure, result);
}, (result || delay < 0) ? 0.1 : delay);
} else if (typeof failure == 'function') {
setTimeout(failure, 1);
}
},
addEvent: function (el, eventName, eventFunc) {
if (typeof el != 'object') {
return false;
}
if (el.addEventListener) {
el.addEventListener(eventName, eventFunc, false);
return true;
}
if (el.attachEvent) {
el.attachEvent("on" + eventName, eventFunc);
return true;
}
return false;
},
// add script to dom
require: function (url, args) {
var me = this;
args = args || {};
var scriptTag = document.createElement('script');
var headTag = document.getElementsByTagName('head')[0];
if (!headTag) {
return false;
}
setTimeout(function () {
var f = (typeof args.success == 'function') ? args.success : function () {
};
args.failure = (typeof args.failure == 'function') ? args.failure : function () {
};
var fail = function () {
if (!scriptTag.__es) {
scriptTag.__es = true;
scriptTag.id = 'failed';
args.failure(scriptTag);
}
};
scriptTag.onload = function () {
scriptTag.id = 'loaded';
f(scriptTag);
};
scriptTag.type = 'text/javascript';
scriptTag.async = (typeof args.async == 'boolean') ? args.async : false;
scriptTag.charset = 'utf-8';
me.__es = false;
me.addEvent(scriptTag, 'error', fail); // when supported
// when error event is not supported fall back to timer
me.timer(15, 1000, 0, function () {
return (scriptTag.id == 'loaded');
}, function () {
if (scriptTag.id != 'loaded') {
fail();
}
});
scriptTag.src = url;
setTimeout(function () {
try {
headTag.appendChild(scriptTag);
} catch (e) {
fail();
}
}, 1);
}, (typeof args.delay == 'number') ? args.delay : 1);
return true;
}
};
$(document).ready(function () {
var loader = new ScriptLoader();
loader.require('resources/templates.js', {
async: true, success: function () {
alert('loaded');
}, failure: function () {
alert('NOT loaded');
}
});
});
回答4:
I know this is an old thread but I got a nice solution to you (I think). It's copied from an class of mine, that handles all AJAX stuff.
When the script cannot be loaded, it set an error handler but when the error handler is not supported, it falls back to a timer that checks for errors for 15 seconds.
function jsLoader()
{
var o = this;
// simple unstopable repeat timer, when t=-1 means endless, when function f() returns true it can be stopped
o.timer = function(t, i, d, f, fend, b)
{
if( t == -1 || t > 0 )
{
setTimeout(function() {
b=(f()) ? 1 : 0;
o.timer((b) ? 0 : (t>0) ? --t : t, i+((d) ? d : 0), d, f, fend,b );
}, (b || i < 0) ? 0.1 : i);
}
else if(typeof fend == 'function')
{
setTimeout(fend, 1);
}
};
o.addEvent = function(el, eventName, eventFunc)
{
if(typeof el != 'object')
{
return false;
}
if(el.addEventListener)
{
el.addEventListener (eventName, eventFunc, false);
return true;
}
if(el.attachEvent)
{
el.attachEvent("on" + eventName, eventFunc);
return true;
}
return false;
};
// add script to dom
o.require = function(s, delay, baSync, fCallback, fErr)
{
var oo = document.createElement('script'),
oHead = document.getElementsByTagName('head')[0];
if(!oHead)
{
return false;
}
setTimeout( function() {
var f = (typeof fCallback == 'function') ? fCallback : function(){};
fErr = (typeof fErr == 'function') ? fErr : function(){
alert('require: Cannot load resource -'+s);
},
fe = function(){
if(!oo.__es)
{
oo.__es = true;
oo.id = 'failed';
fErr(oo);
}
};
oo.onload = function() {
oo.id = 'loaded';
f(oo);
};
oo.type = 'text/javascript';
oo.async = (typeof baSync == 'boolean') ? baSync : false;
oo.charset = 'utf-8';
o.__es = false;
o.addEvent( oo, 'error', fe ); // when supported
// when error event is not supported fall back to timer
o.timer(15, 1000, 0, function() {
return (oo.id == 'loaded');
}, function(){
if(oo.id != 'loaded'){
fe();
}
});
oo.src = s;
setTimeout(function() {
try{
oHead.appendChild(oo);
}catch(e){
fe();
}
},1);
}, (typeof delay == 'number') ? delay : 1);
return true;
};
}
$(document).ready( function()
{
var ol = new jsLoader();
ol.require('myscript.js', 800, true, function(){
alert('loaded');
}, function() {
alert('NOT loaded');
});
});
回答5:
To check if the javascript in nonexistant.js returned no error you have to add a variable inside http://fail.org/nonexistant.js like var isExecuted = true; and then check if it exists when the script tag is loaded.
However if you only want to check that the nonexistant.js returned without a 404 (meaning it exists), you can try with a isLoaded variable ...
var isExecuted = false;
var isLoaded = false;
script_tag.onload = script_tag.onreadystatechange = function() {
if(!this.readyState ||
this.readyState == "loaded" || this.readyState == "complete") {
// script successfully loaded
isLoaded = true;
if(isExecuted) // no error
}
}
This will cover both cases.
回答6:
my working clean solution (2017)
function loaderScript(scriptUrl){
return new Promise(function (res, rej) {
let script = document.createElement('script');
script.src = scriptUrl;
script.type = 'text/javascript';
script.onError = rej;
script.async = true;
script.onload = res;
script.addEventListener('error',rej);
script.addEventListener('load',res);
document.head.appendChild(script);
})
}
As Martin pointed, used like that:
loaderScript("myscript.js")
.then(() => { console.log("loaded"); })
.catch(() => { console.log("error"); });
回答7:
I hope this doesn't get downvoted, because in special circumstances it is the most reliable way to solve the problem. Any time the server allows you to get a Javascript resource using CORS (http://en.wikipedia.org/wiki/Cross-origin_resource_sharing), you have a rich array of options to do so.
Using XMLHttpRequest to fetch resources will work across all modern browsers, including IE. Since you are looking to load Javascript, you have Javascript available to you in the first place. You can track the progress using the readyState (http://en.wikipedia.org/wiki/XMLHttpRequest#The_onreadystatechange_event_listener). Finally, once you receive the content of the file, you can execute it with eval ( ). Yes, I said eval -- because security-wise it is no different from loading the script normally. In fact, a similar technique is suggested by John Resig to have nicer tags (http://ejohn.org/blog/degrading-script-tags/).
This method also lets you separate the loading from the eval, and execute functions before and after the eval happens. It becomes very useful when loading scripts in parallel but evaluating them one after the other -- something browsers can do easily when you place the tags in HTML, but don't let you by adding scripts at run-time with Javascript.
CORS is also preferable to JSONP for loading scripts (http://en.wikipedia.org/wiki/XMLHttpRequest#Cross-domain_requests). However, if you are developing your own third-party widgets to be embedded in other sites, you should actually load the Javascript files from your own domain in your own iframe (again, using AJAX)
In short:
Try to see if you can load the resource using AJAX GET
Use eval after it has successfully loaded
To improve it:
Check out the cache-control headers being sent
Look into otherwise caching the content in localStorage, if you need to
Check out Resig's "degrading javascript" for cleaner code
Check out require.js
回答8:
This trick worked for me, although I admit that this is probably not the best way to solve this problem. Instead of trying this, you should see why the javascripts aren't loading. Try keeping a local copy of the script in your server, etc. or check with the third party vendor from where you are trying to download the script.
Anyways, so here's the workaround: 1) Initialize a variable to false 2) Set it to true when the javascript loads (using the onload attribute) 3) check if the variable is true or false once the HTML body has loaded
<html>
<head>
<script>
var scriptLoaded = false;
function checkScriptLoaded() {
if (scriptLoaded) {
// do something here
} else {
// do something else here!
}
}
</script>
<script src="http://some-external-script.js" onload="scriptLoaded=true;" />
</head>
<body onload="checkScriptLoaded()">
<p>My Test Page!</p>
</body>
</html>
回答9:
Here is another JQuery-based solution without any timers:
<script type="text/javascript">
function loadScript(url, onsuccess, onerror) {
$.get(url)
.done(function() {
// File/url exists
console.log("JS Loader: file exists, executing $.getScript "+url)
$.getScript(url, function() {
if (onsuccess) {
console.log("JS Loader: Ok, loaded. Calling onsuccess() for " + url);
onsuccess();
console.log("JS Loader: done with onsuccess() for " + url);
} else {
console.log("JS Loader: Ok, loaded, no onsuccess() callback " + url)
}
});
}).fail(function() {
// File/url does not exist
if (onerror) {
console.error("JS Loader: probably 404 not found. Not calling $.getScript. Calling onerror() for " + url);
onerror();
console.error("JS Loader: done with onerror() for " + url);
} else {
console.error("JS Loader: probably 404 not found. Not calling $.getScript. No onerror() callback " + url);
}
});
}
</script>
Thanks to: https://stackoverflow.com/a/14691735/1243926
Sample usage (original sample from JQuery getScript documentation):
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<title>jQuery.getScript demo</title>
<style>
.block {
background-color: blue;
width: 150px;
height: 70px;
margin: 10px;
}
</style>
<script src="http://code.jquery.com/jquery-1.9.1.js"></script>
</head>
<body>
<button id="go">» Run</button>
<div class="block"></div>
<script>
function loadScript(url, onsuccess, onerror) {
$.get(url)
.done(function() {
// File/url exists
console.log("JS Loader: file exists, executing $.getScript "+url)
$.getScript(url, function() {
if (onsuccess) {
console.log("JS Loader: Ok, loaded. Calling onsuccess() for " + url);
onsuccess();
console.log("JS Loader: done with onsuccess() for " + url);
} else {
console.log("JS Loader: Ok, loaded, no onsuccess() callback " + url)
}
});
}).fail(function() {
// File/url does not exist
if (onerror) {
console.error("JS Loader: probably 404 not found. Not calling $.getScript. Calling onerror() for " + url);
onerror();
console.error("JS Loader: done with onerror() for " + url);
} else {
console.error("JS Loader: probably 404 not found. Not calling $.getScript. No onerror() callback " + url);
}
});
}
loadScript("https://raw.github.com/jquery/jquery-color/master/jquery.color.js", function() {
console.log("loaded jquery-color");
$( "#go" ).click(function() {
$( ".block" )
.animate({
backgroundColor: "rgb(255, 180, 180)"
}, 1000 )
.delay( 500 )
.animate({
backgroundColor: "olive"
}, 1000 )
.delay( 500 )
.animate({
backgroundColor: "#00f"
}, 1000 );
});
}, function() { console.error("Cannot load jquery-color"); });
</script>
</body>
</html>
回答10:
The reason it doesn't work in Safari is because you're using attribute syntax. This will work fine though:
script_tag.addEventListener('error', function(){/*...*/}, true);
...except in IE.
If you want to check the script executed successfully, just set a variable using that script and check for it being set in the outer code.
回答11:
This can be done safely using promises
function loadScript(src) {
return new Promise(function(resolve, reject) {
let script = document.createElement('script');
script.src = src;
script.onload = () => resolve(script);
script.onerror = () => reject(new Error("Script load error: " + src));
document.head.append(script);
});
}
and use like this
let promise = loadScript("https://cdnjs.cloudflare.com/ajax/libs/lodash.js/3.2.0/lodash.js");
promise.then(
script => alert(`${script.src} is loaded!`),
error => alert(`Error: ${error.message}`)
);
回答12:
It was proposed to set a timeout and then assume load failure after a timeout.
setTimeout(fireCustomOnerror, 4000);
The problem with that approach is that the assumption is based on chance. After your timeout expires, the request is still pending. The request for the pending script may load, even after the programmer assumed that load won't happen.
If the request could be canceled, then the program could wait for a period, then cancel the request.
回答13:
This is how I used a promise to detect loading errors that are emited on the window object:
<script type='module'>
window.addEventListener('error', function(error) {
let url = error.filename
url = url.substring(0, (url.indexOf("#") == -1) ? url.length : url.indexOf("#"));
url = url.substring(0, (url.indexOf("?") == -1) ? url.length : url.indexOf("?"));
url = url.substring(url.lastIndexOf("/") + 1, url.length);
window.scriptLoadReject && window.scriptLoadReject[url] && window.scriptLoadReject[url](error);
}, true);
window.boot=function boot() {
const t=document.createElement('script');
t.id='index.mjs';
t.type='module';
new Promise((resolve, reject) => {
window.scriptLoadReject = window.scriptLoadReject || {};
window.scriptLoadReject[t.id] = reject;
t.addEventListener('error', reject);
t.addEventListener('load', resolve); // Careful load is sometimes called even if errors prevent your script from running! This promise is only meant to catch errors while loading the file.
}).catch((value) => {
document.body.innerHTML='Error loading ' + t.id + '! Please reload this webpage.<br/>If this error persists, please try again later.<div><br/>' + t.id + ':' + value.lineno + ':' + value.colno + '<br/>' + (value && value.message);
});
t.src='./index.mjs'+'?'+new Date().getTime();
document.head.appendChild(t);
};
</script>
<script nomodule>document.body.innerHTML='This website needs ES6 Modules!<br/>Please enable ES6 Modules and then reload this webpage.';</script>
</head>
<body onload="boot()" style="margin: 0;border: 0;padding: 0;text-align: center;">
<noscript>This website needs JavaScript!<br/>Please enable JavaScript and then reload this webpage.</noscript>
回答14:
Well, the only way I can think of doing everything you want is pretty ugly. First perform an AJAX call to retrieve the Javascript file contents. When this completes you can check the status code to decide if this was successful or not. Then take the responseText from the xhr object and wrap it in a try/catch, dynamically create a script tag, and for IE you can set the text property of the script tag to the JS text, in all other browsers you should be able to append a text node with the contents to script tag. If there's any code that expects a script tag to actually contain the src location of the file, this won't work, but it should be fine for most situations.
回答15:
onerror Event
*Update August 2017: onerror is fired by Chrome and Firefox. onload is fired by Internet Explorer. Edge fires neither onerror nor onload. I wouldnt use this method but it could work in some cases. See also
<link> onerror do not work in IE
*
Definition and Usage The onerror event is triggered if an error occurs while loading an external file (e.g. a document or an image).
Tip: When used on audio/video media, related events that occurs when there is some kind of disturbance to the media loading process, are:
- onabort
- onemptied
- onstalled
- onsuspend
In HTML:
element onerror="myScript">
In JavaScript, using the addEventListener() method:
object.addEventListener("error", myScript);
Note: The addEventListener() method is not supported in Internet Explorer 8 and earlier versions.
Example Execute a JavaScript if an error occurs when loading an image:
img src="image.gif" onerror="myFunction()">
来源：https://stackoverflow.com/questions/538745/how-to-tell-if-a-script-tag-failed-to-load