- PHP 8.0 or higher
Hook routes (a combination of one or more HTTP methods and a pattern) using $router->match(method(s), pattern, function):
$router->match('GET|POST', 'pattern', function() { … });Router supports GET, POST, PUT, PATCH, DELETE, HEAD, and OPTIONS HTTP request methods. Pass in a single request method, or multiple request methods separated by |.
When a route matches against the current URL (e.g. $_SERVER['REQUEST_URI']), the attached route handling function will be executed. The route handling function must be a callable. Only the first route matched will be handled. When no matching route is found, a 404 handler will be executed.
Shorthands for single request methods are provided:
$router->get('pattern', function() { /* ... */ });
$router->post('pattern', function() { /* ... */ });
$router->put('pattern', function() { /* ... */ });
$router->delete('pattern', function() { /* ... */ });
$router->options('pattern', function() { /* ... */ });
$router->patch('pattern', function() { /* ... */ });You can use this shorthand for a route that can be accessed using any method:
$router->all('pattern', function() { /* ... */ });Route Patterns can be static or dynamic:
- Static Route Patterns contain no dynamic parts and must match exactly against the
pathpart of the current URL. - Dynamic Route Patterns contain dynamic parts that can vary per request. The varying parts are named subpatterns and are defined using either Perl-compatible regular expressions (PCRE) or by using placeholders
A static route pattern is a regular string representing a URI. It will be compared directly against the path part of the current URL.
Examples:
/about/contact
Usage Examples:
// This route handling function will only be executed when visiting '/anime'
$router->get('/users', function() {
echo 'Anime Page Contents';
});This type of Route Patterns contain dynamic parts which can vary per request. The varying parts are named subpatterns and are defined using regular expressions.
Examples:
/users/(\d+)/profile/(\w+)
Commonly used PCRE-based subpatterns within Dynamic Route Patterns are:
\d+= One or more digits (0-9)\w+= One or more word characters (a-z 0-9 _)[a-z0-9_-]+= One or more word characters (a-z 0-9 _) and the dash (-).*= Any character (including/), zero or more[^/]+= Any character but/, one or more
Note: The PHP PCRE Cheat Sheet might come in handy.
The subpatterns defined in Dynamic PCRE-based Route Patterns are converted to parameters which are passed into the route handling function. Prerequisite is that these subpatterns need to be defined as parenthesized subpatterns, which means that they should be wrapped between parens:
// Bad Routing
$router->get('/hello/\w+', function($name) {
echo 'Hello ' . htmlentities($name);
});
// Good Routing
$router->get('/hello/(\w+)', function($name) {
echo 'Hello ' . htmlentities($name);
});Note: The leading / at the very beginning of a route pattern is not mandatory, but is recommended.
When multiple subpatterns are defined, the resulting route handling parameters are passed into the route handling function in the order they are defined in:
$router->get('/anime/(\d+)/photos/(\d+)', function($animeId, $photoId) {
echo 'Anime #' . $movieId . ', photo #' . $photoId;
});This type of Route Patterns are the same as Dynamic PCRE-based Route Patterns, but with one difference: they don't use regexes to do the pattern matching but they use the more easy placeholders instead. Placeholders are strings surrounded by curly braces, e.g. {name}. You don't need to add parens around placeholders.
Examples:
/anime/{id}/profile/{username}
Placeholders are easier to use than PRCEs, but offer you less control as they internally get translated to a PRCE that matches any character (.*).
$router->get('/anime/{animeId}/photos/{photoId}', function($animeId, $photoId) {
echo 'Anime #' . $animeId . ', photo #' . $photoId;
});Note: the name of the placeholder does not need to match with the name of the parameter that is passed into the route handling function:
$router->get('/anime/{foo}/photos/{bar}', function($animeId, $photoId) {
echo 'Anime #' . $animeId . ', photo #' . $photoId;
});Route subpatterns can be made optional by making the subpatterns optional by adding a ? after them. Think of blog URLs in the form of /blog(/year)(/month)(/day)(/slug):
$router->get(
'/blog(/\d+)?(/\d+)?(/\d+)?(/[a-z0-9_-]+)?',
function($year = null, $month = null, $day = null, $slug = null) {
if (!$year) { echo 'Blog overview'; return; }
if (!$month) { echo 'Blog year overview'; return; }
if (!$day) { echo 'Blog month overview'; return; }
if (!$slug) { echo 'Blog day overview'; return; }
echo 'Blogpost ' . htmlentities($slug) . ' detail';
}
);The code snippet above responds to the URLs /blog, /blog/year, /blog/year/month, /blog/year/month/day, and /blog/year/month/day/slug.
Note: With optional parameters it is important that the leading / of the subpatterns is put inside the subpattern itself. Don't forget to set default values for the optional parameters.
The code snipped above unfortunately also responds to URLs like /blog/foo and states that the overview needs to be shown - which is incorrect. Optional subpatterns can be made successive by extending the parenthesized subpatterns so that they contain the other optional subpatterns: The pattern should resemble /blog(/year(/month(/day(/slug)))) instead of the previous /blog(/year)(/month)(/day)(/slug):
$router->get('/blog(/\d+(/\d+(/\d+(/[a-z0-9_-]+)?)?)?)?', function($year = null, $month = null, $day = null, $slug = null) {
// ...
});Note: It is highly recommended to always define successive optional parameters.
To make things complete use quantifiers to require the correct amount of numbers in the URL:
$router->get('/blog(/\d{4}(/\d{2}(/\d{2}(/[a-z0-9_-]+)?)?)?)?', function($year = null, $month = null, $day = null, $slug = null) {
// ...
});Use $router->mount($baseroute, $fn) to mount a collection of routes onto a subroute pattern. The subroute pattern is prefixed onto all following routes defined in the scope. e.g. Mounting a callback $fn onto /anime will prefix /anime onto all following routes.
$router->mount('/anime', function() use ($router) {
// will result in '/anime/'
$router->get('/', function() {
echo 'anime overview';
});
// will result in '/anime/id'
$router->get('/(\d+)', function($id) {
echo 'anime id ' . htmlentities($id);
});
});Nesting of subroutes is possible, just define a second $router->mount() in the callable that's already contained within a preceding $router->mount().
We can route to the class action like so:
$router->get('/(\d+)', '\App\Controllers\AnimeController@showDetails');When a request matches the specified route URI, the showDetails method on the AnimeController class will be executed. The defined route parameters will be passed to the class method.
The method can be static (recommended) or non-static (not-recommended). In case of a non-static method, a new instance of the class will be created.
If most/all of your handling classes are in one and the same namespace, you can set the default namespace to use on your router instance via setNamespace()
$router->setNamespace('\App\Controllers');
$router->get('/users/(\d+)', 'UserController@showProfile');The default 404 handler sets a 404 status code and exits. You can override this default 404 handler by using $router->set404(callable);
$router->set404(function() {
header('HTTP/1.1 404 Not Found');
// ... do something special here
});You can also define multiple custom routes e.x. you want to define an /api route, you can print a custom 404 page:
$router->set404('/api(/.*)?', function() {
header('HTTP/1.1 404 Not Found');
header('Content-Type: application/json');
$jsonArray = array();
$jsonArray['status'] = "404";
$jsonArray['status_text'] = "route not defined";
echo json_encode($jsonArray);
});Also supported are Class@Method callables:
$router->set404('\App\Controllers\ErrorController@_404');The 404 handler will be executed when no route pattern was matched to the current URL.
💡 You can also manually trigger the 404 handler by calling $router->trigger404()
$router->get('/anime/([a-z0-9-]+)', function($id) use ($router) {
if (!Anime::exists($id)) {
$router->trigger404();
return;
}
// …
});Router supports Before Route Middlewares, which are executed before the route handling is processed.
Like route handling functions, you hook a handling function to a combination of one or more HTTP request methods and a specific route pattern.
$router->before('GET|POST', '/admin/.*', function() {
if (!isset($_SESSION['user'])) {
header('location: /auth/login');
exit();
}
});Unlike route handling functions, more than one before route middleware is executed when more than one route match is found.
Before route middlewares are route specific. Using a general route pattern (viz. all URLs), they can become Before Router Middlewares (in other projects sometimes referred to as before app middlewares) which are always executed, no matter what the requested URL is.
$router->before('GET', '/.*', function() {
// ... this will always be executed
});Run one (1) middleware function, name the After Router Middleware (in other projects sometimes referred to as after app middlewares) after the routing was processed. Just pass it along the $this->router->resolve() function in Daishi\Foundation\Application. The run callback is route independent.
$this->router->resolve(function() { … });Note: If the route handling function has exit()ed the resolve callback won't be run.
Use X-HTTP-Method-Override to override the HTTP Request Method. Only works when the original Request Method is POST. Allowed values for X-HTTP-Method-Override are PUT, DELETE, or PATCH.
There's no such thing as $_PUT in PHP. One must fake it:
$router->put('/anime/(\d+)', function($id) {
// Fake $_PUT
$_PUT = array();
parse_str(file_get_contents('php://input'), $_PUT);
// ...
});When making HEAD requests all output will be buffered to prevent any content trickling into the response body, as defined in RFC2616 (Hypertext Transfer Protocol -- HTTP/1.1):
The HEAD method is identical to GET except that the server MUST NOT return a message-body in the response. The metainformation contained in the HTTP headers in response to a HEAD request SHOULD be identical to the information sent in response to a GET request. This method can be used for obtaining metainformation about the entity implied by the request without transferring the entity-body itself. This method is often used for testing hypertext links for validity, accessibility, and recent modification.
To achieve this, router will internally re-route HEAD requests to their equivalent GET request and automatically suppress all output.