How does the javascript preloading work?

问题:

I don't want to know a way to preload images, I found much on the net, but I want to know how it works. How is javascript able to preload images? I mean, I tried a snippet from here, and even if it works, it doesn't seem to preload images.

When I check firebug, I can see that the image is loaded twice, once while the preloading, another time when displaying it!

To improve this code I'd like to know how it works.

Here is what i do:

 function preload(arrayOfImages) {
        $(arrayOfImages).each(function(){
            $('<img/>')[0].src = this;
            //(new Image()).src = this;
            alert(this +'  &&   ' + i++);

        });
    }

then i do something like that:

preloader = function() {
    preload(myImages);
}

$(document).ready(preloader);

Here is how i display/add the image :

$("li.works").click(function() {
             $("#viewer").children().empty();
             $('#viewer').children().append('<img src=\'images/ref/'+this.firstChild.id+'.jpg\' alt="'+this.firstChild.id+'" \/>')
             $("#viewer").children().fadeIn();

回答1:

Your basic Javascript preloader does this:

var image = new Image();
image.src = '/path/to/the/image.jpg';

The way it works is simply by creating a new Image object and setting the src of it, the browser is going to go grab the image. We're not adding this particular image to the browser, but when the time comes to show the image in the page via whatever method we have setup, the browser will already have it in its cache and will not go fetch it again. I can't really tell you why whatever you have isn't working this way without looking at the code, though.

One interesting gotcha that is discussed in this question is what happens when you have an array of images and try preloading them all by using the same Image object:

var images = ['image1.jpg','image2.jpg'];
var image = new Image();
for(var x = 0; x < images.length; x++) {
    image.src = images[x];
}

This will only preload the last image as the rest will not have time to preload before the loop comes around again to change the source of the object. View an example of this. You should be able to instantly see the second image once you click on the button, but the first one will have to load as it didn't get a chance to preload when you try to view it.

As such, the proper way to do many at once would be:

var images = ['image1.jpg','image2.jpg'];
for(var x = 0; x < images.length; x++) {
    var image = new Image();
    image.src = images[x];
}

回答2:

Javascript preloading works by taking advantage of the caching mechanism used by browsers.

The basic idea is that once a resource is downloaded, it is stored for a period of time locally on the client machine so that the browser doesn't have to retrieve the resource again from across the net, the next time it is required for display/use by the browser.

Your code is probably working just fine and you're just misinterpeting what Fire Bug is displaying.

To test this theory just hit www.google.com with a clean cache. I.e. clear your download history first.

The first time through everything will likely have a status of 200 OK. Meaning your browser requested the resource and the server sent it. If you look at the bottom on the Fire Bug window it will says how big the page was say 195Kb and how much of that was pulled from cache. In this case 0Kb.

Then reload the same page without clearing your cache, and you will still see the same number of requests in FireBug.

The reason for this is simple enough. The page hasn't changed and still needs all the same resources it needed before.

What is different is that for the majority of these requests the server returned a 304 Not Modified Status, so the browser checked it's cache to see if it already had the resource stored locally, which in this case it did from the previous page load. So the browser just pulled the resource from the local cache.

If you look at the bottom of the Fire Bug window you will see that page size is still the same (195Kb) but that the majority of it, in my case 188Kb, was pulled locally from cache.

So the cache did work and the second time i hit Google I saved 188Kb of download.

I'm sure you will find the same thing with preloading your images. The request is still made but if the server returns a status of 304 then you will see that the image is in fact just pulled from local cache and not the net.

So with caching, the advantage is NOT that you kill off all future resource requests, i.e. a Uri lookup is still made to the net but rather that if possible the browser will pull from the local cache to satisify the need for the content, rather than run around the net looking for it.

回答3:

You may be confused by the concept of "preloading". If you have a bunch of images in your HTML with <img src="...">, they cannot be preloaded with Javascript, they just load with the page.

Preloading images with Javascript is about loading images not already in the document source, then displaying them later. They are loaded after the page has rendered for the first time. They are preloaded in order to eliminate/minimize loading time when it comes to making them appear, for example when changing an image on mouse rollover.

For most applications, it is usually better practice to use "CSS sprites" as a form of preloading, in lieu of Javascript. SO should have a ton of questions on this.

回答4:

It just involves making a new DOM image object and setting the src attribute. Nothing clever and AFAIK, it has always worked for me.

Is it possible the second "load" firebug is showing you is it loading it from cache?

回答5:

The index on the loop is only looking at the first image. Change it to use the index:

function preload(arrayOfImages) {
    $(arrayOfImages).each(function(i){   // Note the argument
        $('<img/>')[i].src = this;    // Note the i
        //(new Image()).src = this;
        alert(this +'  &&   ' + i++);

    });
}

Edit: In retrospect, this was wrong and I can see you're trying to create image elements. I don't understand why the index is there at all, there need not be an index. I think the function should look like this:

function preload(arrayOfImages) {
    $(arrayOfImages).each(function () {
        $('<img/>').attr('src', this);
    });
}    

And to instantiate it, why not just do this:

$(function () {    // Equivalent to $(document).ready()
    preload(myImages);
});

JavaScript image preloading works because when a DOM element that contains an image is created, the image is downloaded and cached. Even if another request is made when the image is actually rendered from the HTML, the server will send back a 304 (not changed), and the browser will simply load the image from its cache.

Paolo suggests using the following notation to create an image object:

var image = new Image();

While this will work, the DOM-compliant way of doing this is:

var image = document.createElement('img');
image.setAttribute('src', 'path/to/image.jpg');

Which is the way it is being done in the script, except it's using jQuery's HTML string literal syntax to do it. Additionally, most modern browsers offer compatibility with the Image() constructor by simply calling DOM-standard methods. For example, if you open up the Google Chrome JavaScript console and type Image, this is what you'll get:

function Image() {
    return document.createElementNS('http://www.w3.org/1999/xhtml', 'img');
}

Chrome merely uses the native DOM methods to create an image element.