In consonance with the concepts that I expressed in the beginning, my objective here is to demonstrate how the strategy pattern can help to create a data validation system capable of verifying different types of user data.

Bearing in mind this important requisite, the first step that I'm going to take will consist of building a validation strategy selector (naturally, represented by a PHP class) which will determine what validation strategy must be applied, in accordance with the type of data submitted by a user.

That being said, the validation strategy selector looks like this:

// define 'ValidationStrategySelector' class
class ValidationStrategySelector{
   private $strategy=NULL;
   public function __construct($strategy){
     if($strategy!='alpha'&&$strategy!='alphanum'&&$strategy!
='number'&&$strategy!='email'){
       throw new Exception('Invalid parameter for validation
strategy!');
     }
     switch($strategy){
       case "alpha": 
         $this->strategy=new StrategyAlphabetic();
         break;
       case "alphanum": 
         $this->strategy=new StrategyAlphanumeric();
         break;
       case "number": 
         $this->strategy=new StrategyNumber();
         break;
       case "email": 
         $this->strategy=new StrategyEmail();
         break;
     }
   }
   public function validateData($inputData){
     if(!$this->strategy->validateData($inputData)){
       return 'Input data is not valid!';
     }
     return 'Input data is OK!';
   }
}

As you can see, the above strategy selector accepts the type of validation strategy, that is the "$strategy" parameter, as its unique input parameter. Next, according to the value of this variable, a specific strategy class is properly instantiated to validate a huge variety of data types, including alphabetic and alphanumeric data, as well as numbers and email addresses.

Actually, the logic that stands behind the strategy pattern is demonstrated by the other method that belongs to the previous class, in this case the one called "validateData()," since it takes care of using the proper strategy to validate a specific type of data. So far, the signature of the above class look really simple, right?

Okay, at this stage I'm sure that you understand how the prior validation strategy selector works. Nevertheless, as you saw, the class in question creates four different strategy objects in turn, in accordance with the type of data being verified.

Therefore, it's obviously necessary to show the respective signatures for these strategy classes, so you can see more clearly how they're capable of performing different validation on user-supplied data.

As you may have guessed, defining all these new classes is something that will be covered in the following section. So please click on the link that appears below and keep reading.