如果你选用的是jQuery作为你的js开发基础库,那么你应该认真的阅读一下这篇文章,至少会帮你提升你对jQuery的认识与掌握。先贴上原文,有空翻译一下:
原文:http://www.tvidesign.co.uk/blog/improve-your-jquery-25-excellent-tips.aspx
mprove your jQuery - 25 excellent tips
jQuery is awesome. I've been using it for about a year now and although I was impressed to begin with I'm liking it more and more the longer I use it and the more I find out about it's inner workings.
I'm no jQuery expert. I don't claim to be, so if there are mistakes in this article then feel free to correct me or make suggestions for improvements.
I'd call myself an "intermediate" jQuery user and I thought some others out there could benefit from all the little tips, tricks and techniques I've learned over the past year. The article also ended up being a lot longer than I thought it was going to be so I'll start with a table of contents so you can skip to the bits you're interested in.
Google have been hosting several JavaScript libraries for a while now on Google Code and there are several advantages to loading it from them instead of from your server. It saves on bandwidth, it'll load very quickly from Google's CDN and most importantly it'll already be cached if the user has visited a site which delivers it from Google Code.
This makes a lot of sense. How many sites out there are serving up identical copies of jQuery that aren't getting cached? It's easy to do too...
Or, you can just include a direct reference like this...
Full instructions here
Not just a jQuery tip, there are some great cheat sheets out there for most languages. It's handy having every function on a printable A4 sheet for reference and luckily these guys have produced a couple of nice ones..
http://www.gscottolson.com/weblog/2008/01/11/jquery-cheat-sheet/
http://colorcharge.com/jquery/
OK, a general JavaScript tip here. But any big project that uses lots of jQuery probably uses lots of plugins (this site uses easing, localScroll, lightbox and preload) so it's usually applicable.
Browsers can't load scripts concurrently (well, most can't, yet), which means that if you've got several scripts downloading one at a time then you're really slowing down the loading of your page. So, assuming the scrips are being loaded on every page then you should consider combining them into one long script before deploying.
Some of the plugins will already be minified, but you should consider packing your scripts and any that aren't already. It only takes a few seconds. I'm personally a fan of Packer by Dean Edwards
If you haven't already installed Firebug then you really should. Aside from many other useful features such as allowing you to inspect http traffic and find problems with your CSS it has excellent logging commands that allow you to easily debug your scripts.
Here's a full explanation of all of it's features
My favourite features are "console.info", which you can use to just dump messages and variables to the screen without having to use alert boxes and "console.time" which allows you to easily set up a timer to wrap a bunch of code and see how long it takes. They're all really easy to use too...
In this instance I've deliberately written some very inefficient code! In the next few tips I'll show you how we can use the timer to show some improvements which can be made.
jQuery selectors are awesome. They make selecting any element on the page incredibly simple, but internally they have to do a fair amount of work and if you go mad with them you might find things starting to get pretty slow.
If you're selecting the same element time and time again (in a loop for example) then you can just select it once and keep it in memory while you manipulate it to your heart's content. Take the following example where we add items to an unordered list using a loop.
That takes 1066 milliseconds on my PC in Firefox 3 (imagine how long it would IE6!), which is pretty slow in JavaScript terms. Now take a look at the following code where we use the selector just once.
That only takes 224 milliseconds, more than 4x faster, just by moving one line of code.
We can make the code from the previous tip even faster by cutting down on the number of times we insert into the DOM. DOM insertion operations like .append() .prepend() .after() and .wrap() are relatively costly and performing lots of them can really slow things down.
All we need to do is use string concatenation to build the list and then use a single function to add them to your unordered list like .html() is much quicker. Take the following example...
On my PC that takes 216 milliseconds , just over a 1/5th of a second, but if we build the list items as a string first and use the HTML method to do the insert, like this....
That takes 185 milliseconds, not much quicker but that's another 31 milliseconds off the time.
OK, don't ask me why this one works (I'm sure a more experienced coder will explain).
In our last example we inserted 1000 list items into an unordered list using the .html() method. If we had have wrapped them in the UL tag before doing the insert and inserted the completed UL into another tag (a DIV) then we're effectively only inserting 1 tag, not 1000, which seems to be much quicker. Like this...
The time is now only 19 milliseconds, a massive improvement, 50x faster than our first example.
jQuery makes selecting DOM elements using classes as easy as selecting elements by ID used to be, so it's tempting to use classes much more liberally than before. It's still much better to select by ID though because jQuery uses the browser's native method (getElementByID) to do this and doesn't have to do any of it's own DOM traversal, which is much faster. How much faster? Let's find out.
I'll use the previous example and adapt it so each LI we create has a unique class added to it. Then I'll loop through and select each one once.
Just as I thought my browser had hung, it finished, in 5066 milliseconds (over 5 seconds). So i modified the code to give each item an ID instead of a class and then selected them using the ID.
This time it only took 61 milliseconds. Nearly 100x faster.
By default, when you use a selector such as $('.myDiv') the whole of the DOM will be traversed, which depending on the page could be expensive.
The jQuery function takes a second parameter when performing a selection.
jQuery( expression, context )
By providing a context to the selector, you give it an element to start searching within so that it doesn't have to traverse the whole of the DOM.
To demonstrate this, let's take the first block of code from the tip above. It creates an unordered list with 1000 items, each with an individual class. It then loops through and selects each item once. You'll remember that when selecting by class it took just over 5 seconds to select all 1000 of them using this selector.
I then added a context so that it was only running the selector inside the unordered list, like this...
It still took 3818 milliseconds because it's still horribly inefficient, but that's more than a 25% speed increase by making a small modification to a selector.
One of the coolest things about jQuery is it's ability to chain method calls together. So, for example, if you want to switch the class on an element.
If you're anything like me then you probably learned that in your first 5 minutes of reading about jQuery but it goes further than that. Firstly, it still works across line breaks (because jQuery = JavaScript), which means you can write neat code like this...
Making a habit of using chaining automatically helps you to cut down on your selector use too.
But it goes further than that. Let's say that you want to perform several functions on an element but one of the first functions changes the element in some way, like this...
We've selected a table, drilled down to find cells with a class of "firstColumn" and coloured them in red.
Let's say we now want to colour all the cells with a class of "lastColumn" blue. Because we've used the find() funciton we've filtered out all the cells that don't have a class of "firstColumn" so we need to use the selector again to get the table element and we can't continue chaining, right? Luckily jQuery has an end() function which actually reverts back to the previous unaltered selection so you can carry on chaining, like this...
It's also easier than you might think to write your own jQuery function which can chain. All you have to do is write a function which modifies an element and returns it.
How easy was that?
When I first started using jQuery I loved the fact that it was easy to use the pre-defined animations like slideDown() and fadeIn() to get some really cool effects incredibly easy. It's easy to take things further though because jQuery's animate() method is very easy to use and very powerful. In fact, is you look at the jQuery source code you'll see that internally those methods are just shortcuts which use the animate() function.
The animate() method simply takes any CSS style and smoothly transitions it from one value to another. So, you can change the width, height, opacity, background-color, top, left, margin, color, font-size, anything you want.
This is how easy it is to animate all your menu items grow to 100 pixels high when you roll over them.
Unlike other jQuery functions, animations are automatically queued, so if you want to run a second animation once the first is finished then just call the animate method twice, no callback necessary.
If you want the animations to happen concurrently then just put both styles in the params object of a single call, like this...
You can animate any property that's numeric. You can also download plugins to help you animate properties that aren't, like colors and background colors
jQuery makes it easier than ever to attach events to elements in the DOM unobtrusively, which is great, but adding too many events is inefficient. Event delegation allows you to add less events to achieve the same result in many situations. The best way to illustrate this is with an example...
A simple function which turns cells in a table red when you click on them. Let's say that you've got a grid with 10 columns and 50 rows though, that's 500 events bound. Wouldn't it be neater if we could just attach a single event to the table and when the table is clicked have the event handler work out which cell was clicked before turning it red?
Well that's exactly what event delegation is and it's easy to implement...
'e' contains information about the event, including the target element that actually received the click. All we have to do is inspect it to see which cell was actually clicked. Much neater.
Event delegation has another benefit. Normally, When you bind a handler to a collection of elements it gets attached to those elements and those elements only. If you add new elements to the DOM which would have been matched by the selector then they don't have the event handler bound to them (are you following me?) then nothing will happen.
When using event delegation you can add as many matching elements to the DOM as you like after the event is bound and they work too.
This is the most basic way of storing information about a block of html. jQuery is great at manipulating elements based upon their classes, so if you need to store information about the state of an element then why not add an extra class to store it?
Here's an example. We want to create an expanding menu. When you click the button we want the panel to slideDown() if it's currently closed, or slideUp() if it's currently open. We'll start with the HTML
Very simple! We've just added an extra class to the wrapper div which serves no other purpose other than to tell us the state of the item. So all we need is a click event handler which performs slideUp() or slideDown() on the corresponding panel when the button is clicked.
That's a very simple example, but you can add extra classes for storing all sorts of information about an element or HTML fragment.
However, in all but simple cases it's probably better to use the next tip.
It's not very well documented for some reason but jQuery has an internal data() method which can be used to store information in key/value pairs against any DOM element. Storing a piece of data is as simple as this...
We can amend the example from the previous tip. We'll use the same HTML (with the "expanded" class removed) and use the data() function instead.
I'm sure you'll agree this is much neater. For more information about data() and removeData(), see this page...
jQuery internals
jQuery has loads of built-in selectors for selecting elements by ID, class, tag, attribute and many more. But what do you do when you need to select elements based upon something else and jQuery doesn't have a selector?
Well, one answer would be to add classes to the elements from the start and use those to select them, but it turns out that it's not hard to extend jQuery to add new selectors.
The best way to demonstrate is with an example.
The first block of code creates a custom selector which finds any element that is more than 100 pixels tall. The second block just uses it to add a click handler to all those elements.
I won't go into any more detail here but you can imagine how powerful this is and if you search google for "custom jquery selector" you'll find loads of great examples.
The title might not make a lot of sense but this tip can potentially neaten up your code, reduce the weight and download time of your page and help your SEO. Take the following HTML for example...
That's an example of how a form might be marked up, modified slightly for illustrative purposes. I'm sure you'll agree it's pretty ugly and if you had a long form you'd end up with a fairly long ugly page. It's be nicer if you could just put this in your HTML.
All you have to do is a bit of jQuery manipulation to add all the ugly HTML back in. Like this...
It's not always advisable to do this, you'll get a bit of a flash as the page loads, but in certain situations where you've got a lot of repeated HTML it can really reduce your page weight and the SEO benefits of reducing all your repeated extraneous markup should be obvious.
Another way to speed up your page loads and neaten up the HTML that search spiders see is to lazy load whole chunks of it using an AJAX request after the rest of the page has loaded. The user can get browsing right away and spiders only see the content you want them to index.
We've used this technique on our own site. Those purple buttons at the top of the page drop down 3 forms, directions and a google map, which was doubling the size of our pages. So, we just put all that HTML in a static page and use the load() function to load it in once the DOM was ready. Like this...
I wouldn't use this everywhere. You have to consider the trade offs here. You're making extra requests to the server and portions of your page might not be available to the user right away, but used correctly it can be a great optimization technique.
jQuery isn't just about flash effects. The creator has exposed some really useful methods which fill a few gaps in JavaScript's repertoire.
http://docs.jquery.com/Utilities
In particular, browser support for certain common array functions is patchy (IE7 doesn't even have an indexOf() method!). Jquery has methods for iterating, filtering, cloning, merging and removing duplicates from Arrays.
Other common functions that are difficult in Javascript include getting the selected item in a drop down list. In plain old JavaScript you'd have to get the <select> element using getElementByID, get the child elements as an array and iterate through them checking whether each one was selected or not. jQuery makes it easy...
It's worth spending some time looking through the jQuery documentation on the main site and having a nose around some of the lesser known functions.
Most javascript frameworks make use of the $ symbol as a shorthand and this can cause clashes when trying to use more than one framework on the same page. Luckily there's a simple solution. The .noconflict() function gives control of the $ back and allows you to set your own variable name, like this...
This is another one of those problems that doesn't seem to be as well documented as it should be (not when I went looking anyway) and it's a fairly common requirement when building photo galleries, carousels etc, but it's fairly easy.
All you have to do is use the .load() method on an IMG element and put a callback function in it. The following example changes the "src" attribute of an image tag to load a new image and attaches a simple load function.
You should find that the alert is called as soon as the image is loaded.
jQuery is constantly improving and John Resig, it's creator, always seems to be in search of ways to improve performance.
jQuery is currently on version 1.2.6 but John has already revealed that he's working on a new selector engine called Sizzle, which may apparently improve selector speeds in Firefox by up to 4x. So, it pays to keep up to date.
You don't need to check if an element exists on the page before you manipulate it because jQuery will will simply do nothing if you try to select something and it isn't in the DOM. But when you do need to check if anything has been selected, or how many items have been selected you can use the length property.
Simple, but not obvious.
I learned this tip from Karl Swedberg whose excellent books I used to learn jQuery.
He recently left a comment on one of my previous articles about this technique and the basics are as follows...
Firstly, as soon as jQuery has loaded you use it to add a "JS" class to your HTML tag.
Because that only happens when javascript is enabled you can use it to add CSS styles which only work if the user has JavaScript switched on, like this...
So, what this means is that we can hide content when JavaScript is switched on and then use jQuery to show it when necessary (e.g. by collapsing some panels and expanding them when the user clicks on them), while those with JavaScript off (and search engine spiders) see all of the content as it's not hidden. I'll be using this one a lot in the future.
To read his full article click here.
This should be an obvious one but maybe not. if you have a habit of doing this...
... and then attaching an event handler like this...
... it'll probably work fine until you use it on a long page, at which point you'll notice that the # is causing it to jump to the top of the page when your click event is triggered.
All you have to do to prevent this default behaviour, or indeed any default behaviour on any event handler is to add "return false;" to your handler, like this...
A small tip this one but you can save a few characters by using shorthand for the $(document).ready function.
Instead of this...
You can do this...
If you enjoyed this article please digg it: