PathJS is a lightweight, client-side routing library that allows you to create "single page" applications.
- Lightweight
- Tested (tests available in the
./tests
directory) - Supports the
onhashchange
method - Supports paramaterized routes
- Supports optional route components (dynamic routes)
- Supports Aspect Oriented Programming
- Compatible with all major browsers (Tested on Firefox 3.6, Firefox 4.0, Chrome 9, Opera 11, IE8)
- Allows you to define root routes, and rescue methods
- Independant of all third party libraries, but plays nice with all of them
Explanation
PathJS allows you to bind methods to specific routes, providing you with the ability to dynamically change the content of your web page. A route can be any string prepended with a hash, such as:
#/my/first/route
#!/hashbang/route
#kittens
Binding Routes
PathJS provides the Path
object. This is the root of the library, and is your gateway into route-defining heaven. You can define your routes like so:
// Use an anonymous function
Path.map("#/my/first/route").to(function(){
alert("Hello, World!");
});
// Or define one and use it
function hello_world(){
alert("Hello, World!");
}
Path.map("#/kittens").to(hello_world);
Aspect Oriented Programming
In addition to defining methods that will be executed when a route is activated, you can define methods that will be called before a route
is activated, and after a route is left. This can be done via the enter
and exit
methods, respectively. They work exactly the same as
the to
method:
//Let's add an 'enter' method to one of our routes
Path.map("#/my/first/route").enter(function(){
alert("Enter, minions!");
});
// You can also chain the methods together
Path.map("#!/hashbang/route").enter(fade_in).to(function(){
alert("Method chaining is great!");
}).exit(fade_out);
Before Filters and Execution Halting
In some cases, you may want to perform multiple actions before an action is taken, and depending on their results, cancel the action altogether. For this reason, PathJS supports multiple 'enter' actions, which can be assigned individually or as an array:
Path.map("#/my/first/route").enter(function(){
alert("First one!");
}).enter([
function(){
alert("Second!");
return false;
},
function(){
alert("And third!");
}
]);
Path.map("#/my/first/route").to(function(){
alert("Action!")
});
The methods are executed first-in-first-out, and if any of them returns false, the execution chain is immediately halted. In the example above, the third 'enter' method, as well as the actual action, will never get called, because the second one explicitly returns false.
Route Parameters
What good would a routing system be if it didn't allow you to use parameters? If you provide a route that contains a :token
, that token
will match anything, as long as the rest of the route matches as well. You can access the parameters inside your methods via the this.params
object:
Path.map("#/users/:name").to(function{
alert("Username: " + this.params['name']);
});
The above route will match any of the following href
s:
#/users/mike
#/users/27
Optional/Dynamic Routes
You can define a route that has optional components by wrapping the non-mandatory components in parentheses. Inside your bound method(s), any params that were not provided will come back as undefined.
Path.map("#/users(/:user_id)").to(function(){
var user_id = this.params["user_id"] || "Set a default here!";
});
The above route will match both of the following:
#/users // Your "user_id" parameter will be undefined.
#/users/7 // Your "user_id" parameter will be set to "7".
Root Route
If a user were to land on your page without a route defined, you can force them to use a root route. This route will be automatically selected on page load:
Path.root("#/home");
Rescue Method
If a route somehow ended up in your system without being properly bound to an action, you can specify a "rescue" method that will be called. This lets you provide instant user feedback if they click an undefined route:
Path.rescue(function(){
alert("404: Route Not Found");
});
Automatic Dispatching
If a user gets to your page with an already defined route (for example, the click a referral link with the href
of "www.yoursite.com/media#download"),
PathJS will automatically find and execute the appropriate route methods.
Listen Carefully
You can define routes all day long, but if you don't tell us to listen for them, nothing's going to happen. Once you've got your routes defined, start the listener up by simply typing:
Path.listen();
You should always wrap your Path.listen()
statements in some form of "Document Ready" method. This prevents errors when users come
to your site with a predefined route. Without knowing the DOM is completely done loading, that route will be executed, and may try to
perform operations it won't yet have the ability to do.
- Optimizations
- Adding support for HTML5 push/pop state
To make a pull request, please do the following:
- Mention what specific version of PathJS you were using when you encountered the issue/added the feature. This can be accessed by doing
Path.version
in a debugger console - Provide a pastie or gist that demonstrates the bug/feature
- Make sure you update the test suite with your changes. All tests must pass
- Make sure to update the minified version of the source
- Do not modify the
Path.version
attribute. I will modify that manually when merging the request
This code is still under development, and as such, minor revisions may break compatibility with earlier versions of the library. Please keep this in mind when using PathJS.
Copyright (c) 2011 Mike Trpcic, released under the MIT license.