1. Introduction

Design patterns play a pivotal role in creating robust, maintainable, and scalable code. One such pattern that stands out for its versatility and usefulness is the proxy pattern.

In this tutorial, we’ll delve into the proxy pattern, exploring its definition, use cases, and implementation in Kotlin.

2. Understanding the Proxy Pattern

The proxy pattern is a structural design pattern that provides a surrogate or placeholder for another object to control access to it. This surrogate allows for the addition of behavior before, after, or around the actual object’s method calls. The proxy pattern is widely used in scenarios where we need to control access, manage resources, or add functionality without modifying the existing code.

3. Variations to the Proxy Pattern

Let’s explore several variations of the proxy pattern.

3.1. Virtual Proxies

Virtual proxies are placeholders for objects that are computationally expensive to create. By using a virtual proxy, we can delay the creation of the actual object until it is required, optimizing performance. This is especially useful when dealing with large datasets or resource-intensive operations:

interface RealObject {
    fun performOperation()
}
class RealObjectImpl : RealObject {
    override fun performOperation() {
        println("RealObject performing operation")
    }
}
class VirtualProxy : RealObject {
    private val realbject by lazy { RealObjectImpl() }
    override fun performOperation() {
        realObject.performOperation()
    }
}

In this code, the VirtualProxy class implements the RealObject interface, acting as a proxy for the RealObjectImpl. The performOperation() method of the VirtualProxy class delegates the actual operation to the real object. The proxy also uses lazy() initialization to defer the creation of RealObjectImpl until performOperation() is called the first time.

3.2. Protection Proxies

Protection proxies control access to sensitive or critical components by enforcing access restrictions. They ensure that only authorized clients can interact with the actual object, providing an additional layer of security. Systems that need to restrict operations based on user roles or permissions will commonly rely on a pattern like this:

interface SensitiveObject {
    fun access()
}
class SensitiveObjectImpl : SensitiveObject {
    override fun access() {
        println("SensitiveObject accessed")
    }
}
class ProtectionProxy(private val userRole: String) : SensitiveObject {
    private val realObject: SensitiveObjectImpl = SensitiveObjectImpl()
    override fun access() {
        if (userRole == "admin") {
            realObject.access()
        } else {
            println("Access denied. Insufficient privileges.")
        }
    }
}

In this code, the ProtectionProxy acts as a proxy for the SensitiveObjectImpl and adds an access control mechanism.

 3.3. Logging Proxies

Logging proxies enable the tracking and monitoring of method calls on an object. By intercepting calls, a logging proxy can log relevant information such as method names, parameters, and return values. This is valuable for debugging, performance monitoring, and auditing:

interface ObjectToLog {
    fun operation()
}
class RealObjectToLog : ObjectToLog {
    override fun operation() {
        println("RealObjectToLog performing operation")
    }
}
class LoggingProxy(private val realObject: RealObjectToLog) : ObjectToLog {
    override fun operation() {
        println("Logging: Before operation")
        realObject.operation()
        println("Logging: After operation")
    }
}

In this code, an interface ObjectToLog declares the method operation(). Furthermore, the RealObjectToLog class implements ObjectToLog. The LoggingProxy class acts as a proxy, by both implementing ObjectToLog and by wrapping RealObjectToLog to intercept the operation() call. Before and after the actual method execution, this proxy prints log messages.

3.4. Remote Proxies

Remote proxies act as local representations of objects that reside in a different address space such as on a remote server. This allows for transparent communication between components, making it easier to work with distributed systems:

interface RemoteObject {
    fun performRemoteOperation()
}
class RemoteObjectImpl : RemoteObject {
    override fun performRemoteOperation() {
        println("RemoteObject performing remote operation on the server")
    }
}
class RemoteProxy(private val serverAddress: String) : RemoteObject {
    override fun performRemoteOperation() {
        println("Proxy: Initiating remote communication with server at $serverAddress")
        val remoteObject = RemoteObjectImpl()
        remoteObject.performRemoteOperation()

        println("Proxy: Remote communication complete")
    }
}
class Client(private val remoteObject: RemoteObject) {
    fun executeRemoteOperation() {
        println("Client: Performing operation through remote proxy")
        remoteObject.performRemoteOperation()
    }
}
class Server(private val remoteObject: RemoteObject) {
    fun startServer() {
        println("Server: Server started")
    }
}

In this example, the Client uses the RemoteProxy to initiate a remote operation on the server-side RemoteObject. The communication between the client and server is abstracted by the RemoteProxy, making it transparent for the client to work with the remote object as if it were a local object:

 

In the proxy pattern, an interface is established for the RealSubject and Proxy, enabling seamless substitution. This common interface ensures that the proxy, which implements the same interface, can be passed to any client expecting the real service object.

The proxy holds a reference to the real subject. This reference allows the proxy to access the functionality of the original class. Moreover, the proxy class controls access to the real subject and may handle its creation and deletion. This design ensures that clients interact with the proxy, providing a layer of indirection that can manage the instantiation and behavior of the real subject.

4. Implementing the Proxy Pattern in Kotlin

Now that we’ve seen some examples, we’ll dive deeper into a specific implementation of the proxy pattern in Kotlin.

4.1. Example Without the Proxy Pattern

First, let’s look at some code to load and display an Image without the proxy pattern:

interface Image {
    fun display(): Unit
}
class RealImage(private val filename: String) : Image {
    init {
        loadFromDisk()
    }
    private fun loadFromDisk() {
        println("Loading image: $filename")
    }
    override fun display() {
        println("Displaying image: $filename")
    }
}

In this scenario, the RealImage class is responsible for loading and displaying the image. There’s no intermediate proxy class to control access or add any functionality.

Furthermore, in the absence of the proxy pattern, the client code interacts directly with the RealImage class. The image loading and displaying happens immediately when the display() method is called, without any intermediary steps or additional control.

4.2. Example With the Proxy Pattern

Conversely, let’s explore how to implement this solution using the proxy pattern. Specifically, using the same Image interface, our objective is to establish a proxy that effectively manages access to the underlying image object:

interface Image {
    fun display(): Unit
}
class RealImage(private val filename: String) : Image {
    init {
        loadFromDisk()
    }
    private fun loadFromDisk() {
        println("Loading image: $filename")
    }
    override fun display() {
        println("Displaying image: $filename")
    }
}
class ProxyImage(private val filename: String) : Image {
    private var realImage: RealImage? = null
    override fun display() {
        if (realImage == null) {
            realImage = RealImage(filename)
        }
        realImage?.display()
    }
}

In this example, the ProxyImage class acts as a proxy for the RealImage class. The actual image loading and displaying happens inside the RealImage class, but the ProxyImage class controls loading the image by delaying it until the client invokes the display() method. Additionally, ProxyImage only loads the image once, no matter how many times the client invokes the display() method.

5. Advantages of Using the Proxy Pattern

Let’s look at some advantages of the proxy pattern:

  • Proxies provide a way to control access to the actual object, allowing for additional checks, logging, or security measures. This is beneficial in scenarios where fine-grained control over method invocations is required.
  • Proxies enable the addition of new functionality before, after, or around the method calls of the actual object. This facilitates the implementation of features such as logging, caching, or access control without modifying the core functionality of the object.
  • The proxy pattern provides a means of managing the underlying resources that it wraps. For example, the proxy in this article managed the lazy loading of the resources controlled by the proxy.

6. Disadvantages of Using the Proxy Pattern

Let’s also go over some of the disadvantages of the proxy pattern:

  • Introducing proxies can lead to increased code complexity, especially when multiple proxy types are involved.
  • When dealing with multithreaded applications, synchronization issues may arise. If multiple threads attempt to access the real object simultaneously, the proxy needs to handle synchronization to ensure thread safety.
  • The use of proxies may lead to tight coupling between the client code and the proxy, making the system less flexible to changes. If changes are made to the real object or the proxy, clients relying on these proxies might need to be modified accordingly.

7. Conclusion

The proxy pattern in Kotlin is a powerful tool for managing object access, enhancing functionality, and optimizing resource usage. By employing proxies, developers can create more modular, maintainable, and extensible code. Whether it’s for lazy loading, access control, or logging, the proxy pattern proves to be a valuable asset.

As always, the full implementation of these examples is available over on GitHub.

Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments