README

serialize / unserialize plain into object structures. Hydrator parses the DocComments of public properties and instanciates the classes according to the definiton.

Examples:

• hydrator-usage-example.php

Installation:

composer install phore/hydrator

Basic Example

class UserData {
    /**
     * @var string
     */
    public $name;
    
    /**
    * Assoc Array 
    * @var array<string, ClassType2> 
    */
    public $map;
    
    /**
     * @var int
     */
    public $age;
}

$input = ["name"=>"bob", "age"=>37];

$userData = phore_hydrate($input, UserData::class);

assert( $userData instanceof UserData);

$userData is a UserData Object and all properties casted correctly to desired types specified in DocComments.

Recognized Annotations

• Simple types like string, int, bool, float, array
• Array types like string[], int[]...
• Object types OtherClass
• Arrays of Objects OtherClass[]
• Nullable properties type|null

Guide

Getters / Setters

On objects, hydrator will try to set property values in the following order:

1. If object has a set<PropertyName>($value)-Method it will use it first
2. If the property is public it will be set directly
3. If there is a __set($name, $value) method it will be used

Default Values

Default values will be applied if no data was found for the specific key

public $prop1 = []

Optional Properties

You can define a property as optional by adding |null to the DocBlock.

/**
 * @var SomeEntity1|null
 */
public $entity1;

If the input data was not found, the value will be null.

Filter input data before hydration

To ease backwards compatibility issues, the magick __hydrate() method is called to prefilter the input data before it is hydrated.

class Entity1 {
    public $p1;

    public function __hydrate(array $input) : array
    {
        // .. modify input to match the object ..
        return $input;
    }
}

Dealing with additional / undefined input data

By default, on undefined input keys, hydrator will throw an exception. You can toggle this behaviour