I know this question's been asked before (e.g., What is the difference between the bridge pattern and the strategy pattern?).
However, could someone please explain, using clear-cut examples, what the difference is and in what sorts of cases one must be selected over the other? Less conceptual theory, more practical "real-life" scenarios would be appreciated.
The difference simply lies in that they solve different problems: The State pattern deals with what (state or type) an object is (in) -- it encapsulates state-dependent behavior, whereas. the Strategy pattern deals with how an object performs a certain task -- it encapsulates an algorithm.
But, the key difference is that Strategy Pattern is about modifying a behaviour of a context in runtime using strategies, while Template Method Pattern is about following a skeleton implementation of an algorithm and modifying its behaviour by overriding methods of the skeleton class in the subclasses.
Bridge is used when we need to decouple an abstraction from its implementation so that the two can vary independently. This type of design pattern comes under structural pattern as this pattern decouples implementation class and abstract class by providing a bridge structure between them.
The strategy pattern allows you to change the implementation of something used at runtime. The decorator pattern allows you augment (or add to) existing functionality with additional functionality at run time.
The Bridge Pattern makes a distinction between an abstraction and an implementation in such a way that the two can vary independently. I will use the example from
Patterns in Java, Volume 1: A Catalog of Reusable Design Patterns Illustrated with UML, Second Edition
You need to provide classes that access physical sensors such as found in scales, speed measuring devices etc. Each sensor produces a number but the number could mean different things. For the scale it could mean the weight and for the speed measuring device it may mean speed.
So you can start by creating a Sensor abstract class to represent the commonality between all sensors and various subclasses for the different types of sensors. This is a robust design allowing you to provide many more types of sensors in the future.
Now suppose that sensors are provided by different manufacturers. You will have to create a heirarchy of sensor classes for manufacturer X and another for manufacturer Y. The problem now is that the clients would need to know the difference between the manufacturers. And if you decide to support a third manufacturer...?
The solution is to provide the main abstraction heirarchy, ie. the Sensor abstract class and sub classes such as SpeedSensor and WeightSensor and so on. Then provide the interface (Bridge) that will exist between the abstraction and the implementation. So there will be a SensorInterface, WeightSensorInterface and SpeedSensorInterface, which dictates the interface that each concrete sensor class must provide. The abstraction does not know about the implementation, rather it knows about the interface. Finally, you can create an concreate implementation for each manufacturer. That is, XSensor, XWeightSensor and XSpeedSensor, YSensor, YSpeedSensor and YWeightSensor.
Clients depend only on the abstraction but any implementation could be plugged in. So in this setup, the abstraction could be changed without changing any of the concrete classes, and the implementation could be changed without worrying about the abstraction.
As you can see this describes a way to structure your classes.
The Strategy on the other hand is concerned with changing the behaviour of an object at run time. I like to use the example of a game with a character that possesses several different types of weapons. The character can attack but the behaviour of attack depends on the weapon that the character is holding at the time, and this cannot be known at compile time.
So you make the weapon behaviour pluggable and inject it into the character as needed. Hence a behavioral pattern.
These two patterns solve different problems. The strategy is concerned with making algorithms interchangeable while the Bridge is concerned with decoupling the abstraction from the inplementation so that you can provide multiple implementations for the same abstraction. That is, the bridge is concerned with entire structures.
Here are a few links that might be useful:
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With