Question

I have URLs like below that need to be recognized by the PHP code. Based on what the URL is, data needs to be shown:

www.example.com/music/username/popular
http://www.example.com/music/username/recent/
http://example.com/music/username/favorites/ignore_this /*Ignore everything after favorites*/
http://www.example.com/music/2011/05/02 /*Shows all music uploaded on this date*/
www.example.com/groups
http://www.example.com/groups/jazz
http://example.com/places/japan/?param=ignore_this /*Ignore everything after japan*/
www.example.com/search/rock/

The first URL should show a user’s popular music. www.example.com/groups should list all public groups. And so on..

• http:// is optional
• / at the end is optional
• If anything (like groups) is entered in UPPERCASE, it should be converted to lowercase

What is the best way to recognize these URLs, using regex and create a Switch case? Example snippet would be great.

Answer 1

Here is the system I use (it’s OOP, but you can change easily if you don’t like classes).

if($this->request->uriMatch('#^/$#'))   //simplest regexp, no substring is matched
    $this->home();  //show the Home page
elseif($this->request->uriMatch('#^/news/(\d+)\.html$#')) //matches a number!
    $this->newsItem($this->request->uri(0),0); // calls newsItem() function and passes the first (0th) matched substring (in our case it's number) to it as an argument
elseif($this->request->uriMatch('#^/news_(\d{4})_(\d{1,2})\.html$#')) //matches 2 numbers
    $this->newsList(0,$this->request->uri(0),$this->request->uri(1)); //passes both numbers to function newsList()
elseif($this->request->uriMatch('#^/products/latest(?:-(\d+))?\.html$#')) //may match one number, or may not match anything
    $this->products('latest',$this->request->uri(0,1)); //if matched, passes the matched number, if not: passes "1" (as default value)
elseif($this->request->uriMatch('#^/products/(\d+)(?:-(\d+))?\.html$#')) //may match 1 or 2 numbers, this is a mix of previous 2 cases :)
    $this->products($this->request->uri(0),$this->request->uri(1,1));
else    //if nothing was matched, then 404!
    $this->response->redirect('/404.html');

Note that (?: ) regexp is a non-matching subpattern, so it doesn’t affect anything.

One example for your the case you provided:

if($this->request->uriMatch('#^/music/([a-z0-9]+)/favorites/?#i'))

? means the last / may not exist. Note that there is no $ sign at the end, which means everything after favorites will be ignored. The i modifier (after #) means the text case is not important.

$this->request is an instance of class Request, here it is:

class Request{
    private $uri;   //this holds the URI
    private $uriArray;  //this will hold the matched substrings of the URI according to our REGEXPs
    public function __construct(){
        // initializes URI, it doesn't contain http:// and the domain!
        $this->uri = $_SERVER['REQUEST_URI'];
    }
    public function uriMatch($regex){
        // parses URL according to REGEX
        $b = preg_match($regex, $this->uri, $this->uriArray); // $b is false, if the URL was not matched
        if($b==1)   //if $b is not false, uriArray contains the URL AND the matched substrings (http://am.php.net/manual/en/function.preg-match.php).
            array_shift($this->uriArray); // we are removing the first element (which is the URL), we need only matched substrings

        return $b==1; //returns true if and only if the URL was matched!
    }

    public function uri($n, $default=false){
        //returns n-th matched substring, or $default, if it was not set
        // ... one can add some error handling here
        return isset($this->uriArray[$n]) ? $this->uriArray[$n] : $default;
    }
}