ChatGPT解决这个技术问题 Extra ChatGPT

When to use an interface instead of an abstract class and vice versa?

This may be a generic OOP question. I wanted to do a generic comparison between an interface and an abstract class on the basis of their usage.

When would one want to use an interface and when would one want to use an abstract class?

In additions to answers below this is a nice short list of where you might want to prefer interfaces and where you might not: When To Use Interfaces: msdn.microsoft.com/en-us/library/3b5b8ezk(v=vs.80).aspx
use abstract when you arent sure what the class is gonna do. use interface if you are.
I am interested to see how many developers, who don't work at Microsoft, define and use interfaces in their day to day development.

J
Jorge Córdoba

I wrote an article about that:

Abstract classes and interfaces

Summarizing:

When we talk about abstract classes we are defining characteristics of an object type; specifying what an object is.

When we talk about an interface and define capabilities that we promise to provide, we are talking about establishing a contract about what the object can do.


This was ery helpful: Interfaces do not express something like "a Doberman is a type of dog and every dog can walk" but more like "this thing can walk". Thank you
Alex's explanation below, re: the difference between only describing functions implemented versus also describing state stored, seems like a better answer to this question, because the differences aren't just philosophical.
Duncan Malashock, not really. Jorge's answer is the better one. Alex's answer focuses on mechanics, while Jorge's more on semantics.
I like the statement before your specified answer: Use abstract classes and inheritance if you can make the statement “A is a B”. Use interfaces if you can make the statement “A is capable of [doing] as”
Very nicely explained:-)
A
Alex

An abstract class can have shared state or functionality. An interface is only a promise to provide the state or functionality. A good abstract class will reduce the amount of code that has to be rewritten because it's functionality or state can be shared. The interface has no defined information to be shared


This, to me, is the best answer here and it is a shame it was not voted higher. Yes, there are philosophical differences between the two concepts, but the root point is that abstract classes ensure that all descendants share functionality/state, where an interface only ensures a common bond.
For example, an abstract base class is used for the template method design pattern, whereas an interface is used for the strategy design pattern.
I think Jorge's summary explains the primary though behind existence for both of them while Alex's answer is the difference in outcomes. I wish I could mark both as the right answers, but I still prefer Jorge's answer.
And Here is example with code.
To me, this "A good abstract class will reduce the amount of code that has to be rewritten because it's functionality or state can be shared." statement is the core of the answer.
P
Paul Hollingsworth

Personally, I almost never have the need to write abstract classes.

Most times I see abstract classes being (mis)used, it's because the author of the abstract class is using the "Template method" pattern.

The problem with "Template method" is that it's nearly always somewhat re-entrant - the "derived" class knows about not just the "abstract" method of its base class that it is implementing, but also about the public methods of the base class, even though most times it does not need to call them.

(Overly simplified) example:

abstract class QuickSorter
{
    public void Sort(object[] items)
    {
        // implementation code that somewhere along the way calls:
        bool less = compare(x,y);
        // ... more implementation code
    }
    abstract bool compare(object lhs, object rhs);
}

So here, the author of this class has written a generic algorithm and intends for people to use it by "specializing" it by providing their own "hooks" - in this case, a "compare" method.

So the intended usage is something like this:

class NameSorter : QuickSorter
{
    public bool compare(object lhs, object rhs)
    {
        // etc.
    }
}

The problem with this is that you've unduly coupled together two concepts:

A way of comparing two items (what item should go first) A method of sorting items (i.e. quicksort vs merge sort etc.)

In the above code, theoretically, the author of the "compare" method can re-entrantly call back into the superclass "Sort" method... even though in practise they will never want or need to do this.

The price you pay for this unneeded coupling is that it's hard to change the superclass, and in most OO languages, impossible to change it at runtime.

The alternative method is to use the "Strategy" design pattern instead:

interface IComparator
{
    bool compare(object lhs, object rhs);
}

class QuickSorter
{
    private readonly IComparator comparator;
    public QuickSorter(IComparator comparator)
    {
        this.comparator = comparator;
    }

    public void Sort(object[] items)
    {
        // usual code but call comparator.Compare();
    }
}

class NameComparator : IComparator
{
    bool compare(object lhs, object rhs)
    {
        // same code as before;
    }
}

So notice now: All we have are interfaces, and concrete implementations of those interfaces. In practise, you don't really need anything else to do a high level OO design.

To "hide" the fact that we've implemented "sorting of names" by using a "QuickSort" class and a "NameComparator", we might still write a factory method somewhere:

ISorter CreateNameSorter()
{
    return new QuickSorter(new NameComparator());
}

Any time you have an abstract class you can do this... even when there is a natural re-entrant relationship between the base and derived class, it usually pays to make them explicit.

One final thought: All we've done above is "compose" a "NameSorting" function by using a "QuickSort" function and a "NameComparison" function... in a functional programming language, this style of programming becomes even more natural, with less code.


Just because you can use Abstract classes or Template Method pattern doesn't mean you need to avoid them. Strategy Pattern is a different pattern for a different situation as in this example, but there are lots of examples where a template pattern is mucho more suitable than Strategy.
Well, in my experience I never run into them (situations where template method is preferable)... or rarely anyway. And that's all "abstract" is - language support for the "template method" design pattern.
Ok, I used it once for an expert system where the process was something like, get 1. FillTheParameters, 2. Make The Vector Product between them, 3. For Each Pair compute result, 4. Join the results, where steps 1 and 3 where delegated and 2 and 4 implemented in the base class.
I find almost any use of abstract classes harder to understand. Thinking in terms of boxes that communicate with each other instead of inheritance relationships is easier (for me)... But I also agree that current OO languages force too much boilerplate... Functional will be the way to go over OO
The example of misuse is pretty trivial. It rarely boils down to such nice stripped functionality like compare. Much more common are situations where there is some default functionality that the derived classes either replace or extend (and in the latter case it's perfectly valid to call the base class function). In your example there is no default functionality, so usage of abstract class has no justification.
G
Graham

If you are looking at java as OOP language,

"interface does not provide method implementation" is no longer valid with Java 8 launch. Now java provides implementation in interface for default methods.

In simple terms, I would like to use

interface: To implement a contract by multiple unrelated objects. It provides "HAS A" capability.

abstract class: To implement the same or different behaviour among multiple related objects. It establishes "IS A" relation.

Oracle website provides key differences between interface and abstract class.

Consider using abstract classes if :

You want to share code among several closely related classes. You expect that classes that extend your abstract class have many common methods or fields, or require access modifiers other than public (such as protected and private). You want to declare non-static or non-final fields.

Consider using interfaces if :

You expect that unrelated classes would implement your interface. For example,many unrelated objects can implement Serializable interface. You want to specify the behaviour of a particular data type, but not concerned about who implements its behaviour. You want to take advantage of multiple inheritance of type.

Example:

Abstract class ( IS A relation)

Reader is an abstract class.

BufferedReader is a Reader

FileReader is a Reader

FileReader and BufferedReader are used for common purpose : Reading data, and they are related through Reader class.

Interface ( HAS A capability )

Serializable is an interface.

Assume that you have two classes in your application, which are implementing Serializable interface

Employee implements Serializable

Game implements Serializable

Here you can't establish any relation through Serializable interface between Employee and Game, which are meant for different purpose. Both are capable of Serializing the state and the comparasion ends there.

Have a look at these posts :

How should I have explained the difference between an Interface and an Abstract class?


best answer I think.
Every human learns better with examples. Very good answer. Thanks!
D
Dherik

My two cents:

An interface basically defines a contract, that any implementing class must adhere to(implement the interface members). It does not contain any code.

On the other hand, an abstract class can contain code, and there might be some methods marked as abstract which an inheriting class must implement.

The rare situations I've used abstract classes is when i have some default functionality that the inheriting class might not be interesting in overriding, in say an abstract base class, that some specialized classes inherit from.

Example(a very rudimentary one!):Consider a base class called Customer which has abstract methods like CalculatePayment(), CalculateRewardPoints() and some non-abstract methods like GetName(), SavePaymentDetails().

Specialized classes like RegularCustomer and GoldCustomer will inherit from the Customer base class and implement their own CalculatePayment() and CalculateRewardPoints() method logic, but re-use the GetName() and SavePaymentDetails() methods.

You can add more functionality to an abstract class(non abstract methods that is) without affecting child classes which were using an older version. Whereas adding methods to an interface would affect all classes implementing it as they would now need to implement the newly added interface members.

An abstract class with all abstract members would be similar to an interface.


+1 for "You can add more functionality to an abstract class(non abstract methods that is) without affecting child classes which were using an older version. Whereas adding methods to an interface would affect all classes implementing it as they would now need to implement the newly added interface members."
interfaces can have "default" methods so having no method impl in interfaces is wrong idea. "Parent-to-Child" IS-A relationship is the key here. Also, "Shared attributes" vs. "Shared Properties". e.g. Dog iS-A Animal. But a dog can also "Walk"
s
sunwukung

OK, having just "grokked" this myself - here it is in layman's terms (feel free to correct me if I am wrong) - I know this topic is oooooold, but someone else might stumble across it one day...

Abstract classes allow you to create a blueprint, and allow you to additionally CONSTRUCT (implement) properties and methods you want ALL its descendants to possess.

An interface on the other hand only allows you to declare that you want properties and/or methods with a given name to exist in all classes that implement it - but doesn't specify how you should implement it. Also, a class can implement MANY interfaces, but can only extend ONE Abstract class. An Interface is more of a high level architectural tool (which becomes clearer if you start to grasp design patterns) - an Abstract has a foot in both camps and can perform some of the dirty work too.

Why use one over the other? The former allows for a more concrete definition of descendants - the latter allows for greater polymorphism. This last point is important to the end user/coder, who can utilise this information to implement the A.P.I(nterface) in a variety of combinations/shapes to suit their needs.

I think this was the "lightbulb" moment for me - think about interfaces less from the author's perpective and more from that of any coder coming later in the chain who is adding implementation to a project, or extending an API.


to build upon this: An object that implements an interface takes on it's TYPE. This is crucial. So, you can pass different variations of the interface to a class, but refer to them (and their methods) WITH THE TYPE NAME OF THE INTERFACE. Thus, you eliminate the need for a switch or if/else loop. Try this tutorial on the subject - it demonstrates the use of an interface via the Strategy Pattern. phpfreaks.com/tutorial/design-patterns---strategy-and-bridge/…
I totally agree about your lightbulb moment: "A.P.I(nterface) in a variety of combinations/shapes to suit their needs" ! Very very good point to make.
A
Aamir

When to do what is a very simple thing if you have the concept clear in your mind.

Abstract classes can be Derived whereas Interfaces can be Implemented. There is some difference between the two. When you derive an Abstract class, the relationship between the derived class and the base class is 'is a' relationship. e.g., a Dog is an Animal, a Sheep is an Animal which means that a Derived class is inheriting some properties from the base class.

Whereas for implementation of interfaces, the relationship is "can be". e.g., a Dog can be a spy dog. A dog can be a circus dog. A dog can be a race dog. Which means that you implement certain methods to acquire something.

I hope I am clear.


Your second examples can be still an "Is A" relationship. A race dog Is A dog
A
Andrew Barber

1.If you are creating something that provides common functionality to unrelated classes, use an interface.

2.If you are creating something for objects that are closely related in a hierarchy, use an abstract class.


S
Satya

When to prefer an abstract class over interface?

If one plans on updating a base class throughout the life of a program/project, it is best to allow that the base class be an abstract class If one is trying to build a backbone for objects that are closely related in a hierarchy, it is highly beneficial to use an abstract class

When to prefer an interface over abstract class?

If one is not dealing with a massive hierarchical type of framework, interfaces would be a great choice Because multiple inheritance is not supported with abstract classes(diamond problem), interfaces can save the day


The same type of thinking which made me search for a simple answer to the question.
FWIW, I really love this answer.
佚名

I wrote an article of when to use an abstract class and when to use an interface. There is a lot more of a difference between them other than "one IS-A... and one CAN-DO...". To me, those are canned answers. I mention a few reasons when to use either of them. Hope it helps.

http://codeofdoom.com/wordpress/2009/02/12/learn-this-when-to-use-an-abstract-class-and-an-interface/


M
Mazhar

Classes may inherit from only one base class, so if you want to use abstract classes to provide polymorphism to a group of classes, they must all inherit from that class. Abstract classes may also provide members that have already been implemented. Therefore, you can ensure a certain amount of identical functionality with an abstract class, but cannot with an interface.

Here are some recommendations to help you to decide whether to use an interface or an abstract class to provide polymorphism for your components.

If you anticipate creating multiple versions of your component, create an abstract class. Abstract classes provide a simple and easy way to version your components. By updating the base class, all inheriting classes are automatically updated with the change. Interfaces, on the other hand, cannot be changed once created in that way. If a new version of an interface is required, you must create a whole new interface.

If the functionality you are creating will be useful across a wide range of disparate objects, use an interface. Abstract classes should be used primarily for objects that are closely related, whereas interfaces are best suited for providing common functionality to unrelated classes.

If you are designing small, concise bits of functionality, use interfaces. If you are designing large functional units, use an abstract class.

If you want to provide common, implemented functionality among all implementations of your component, use an abstract class. Abstract classes allow you to partially implement your class, whereas interfaces contain no implementation for any members.

Copied from:
http://msdn.microsoft.com/en-us/library/scsyfw1d%28v=vs.71%29.aspx


There's nothing in UML that precludes multiple class inheritance. Multiple inheritance is determined by a programming language, not by UML. For example, multiple class inheritance isn't allowed in in Java and C#, for but is allowed in C++.
@BobRodes: There are a number of features that object-oriented frameworks can provide in various combinations, but not in all combinations. Generalized multiple inheritance precludes some other useful combinations of features including the ability to cast a reference directly to any parent type of the actual instance or any interface type supported thereby and the ability to independently compile base types and derived types and join them at run time.
@supercat Yours is a good explanation of some of the problems that result from using multiple inheritance. Nevertheless, there is nothing in UML that precludes multiple class inheritance in a diagram. I was responding to the above "Classes may inherit from only one base class..." which isn't quite so.
@BobRodes: The question was tagged Java. Java includes the indicated features, and thus is limited to forms of multiple inheritance that can't produce a "deadly diamond" (though in fact the way they've implemented default interface implementations makes the deadly diamond possible).
@supercat Oh, ok. I don't usually look at java tags, so at the time I wrote that I at least thought I was commenting on a UML answer. In any case, I agree with your comment.
D
Domn Werner

I think the most succinct way of putting it is the following:

Shared properties => abstract class. Shared functionality => interface.

And to put it less succinctly...

Abstract Class Example:

public abstract class BaseAnimal
{
    public int NumberOfLegs { get; set; }

    protected BaseAnimal(int numberOfLegs)
    {
        NumberOfLegs = numberOfLegs;
    }
}

public class Dog : BaseAnimal
{
    public Dog() : base(4) { }
}

public class Human : BaseAnimal 
{
    public Human() : base(2) { }
}

Since animals have a shared property - number of legs in this case - it makes sense to make an abstract class containing this shared property. This also allows us to write common code that operates on that property. For example:

public static int CountAllLegs(List<BaseAnimal> animals)
{
    int legCount = 0;
    foreach (BaseAnimal animal in animals)
    {
        legCount += animal.NumberOfLegs;
    }
    return legCount;
}

Interface Example:

public interface IMakeSound
{
    void MakeSound();
}

public class Car : IMakeSound
{
    public void MakeSound() => Console.WriteLine("Vroom!");
}

public class Vuvuzela : IMakeSound
{
    public void MakeSound() => Console.WriteLine("VZZZZZZZZZZZZZ!");        
}

Note here that Vuvuzelas and Cars are completely different things, but they have shared functionality: making a sound. Thus, an interface makes sense here. Further, it will allow programmers to group things that make sounds together under a common interface -- IMakeSound in this case. With this design, you could write the following code:

List<IMakeSound> soundMakers = new List<ImakeSound>();
soundMakers.Add(new Car());
soundMakers.Add(new Vuvuzela());
soundMakers.Add(new Car());
soundMakers.Add(new Vuvuzela());
soundMakers.Add(new Vuvuzela());

foreach (IMakeSound soundMaker in soundMakers)
{
    soundMaker.MakeSound();
}

Can you tell what that would output?

Lastly, you can combine the two.

Combined Example:

public interface IMakeSound
{
    void MakeSound();
}

public abstract class BaseAnimal : IMakeSound
{
    public int NumberOfLegs { get; set; }

    protected BaseAnimal(int numberOfLegs)
    {
        NumberOfLegs = numberOfLegs;
    }

    public abstract void MakeSound();
}

public class Cat : BaseAnimal
{
    public Cat() : base(4) { }

    public override void MakeSound() => Console.WriteLine("Meow!");
}

public class Human : BaseAnimal 
{
    public Human() : base(2) { }

    public override void MakeSound() => Console.WriteLine("Hello, world!");
}

Here, we're requiring all BaseAnimals make a sound, but we don't know its implementation yet. In such a case, we can abstract the interface implementation and delegate its implementation to its subclasses.

One last point, remember how in the abstract class example we were able to operate on the shared properties of different objects and in the interface example we were able to invoke the shared functionality of different objects? In this last example, we could do both.


A
Abhijeet Ashok Muneshwar

Consider using abstract classes if any of these statements apply to your situation:

You want to share code among several closely related classes. You expect that classes that extend your abstract class have many common methods or fields or require access modifiers other than public (such as protected and private). You want to declare non-static or non-final fields. This enables you to define methods that can access and modify the state of the object to which they belong.

Consider using interfaces if any of these statements apply to your situation:

You expect that unrelated classes would implement your interface. For example, the interfaces Comparable and Cloneable are implemented by many unrelated classes. You want to specify the behavior of a particular data type, but not concerned about who implements its behavior. You want to take advantage of multiple inheritances.

Source


S
Sebastian

Use an abstract class if you want to provide some basic implementations.


Thanks Sebastian. But What if I don't need to have a basic implementation? Wont an abstract class and interface be the same then if this is the only difference between them? Why is there a difference?
Because some languages don't have interfaces - C++.
N
Nick Fortescue

The answers vary between languages. For example, in Java a class can implement (inherit from) multiple interfaces but only inherit from one abstract class. So interfaces give you more flexibility. But this is not true in C++.


R
Ravindra babu

For me, I would go with interfaces in many cases. But I prefer abstract classes in some cases.

Classes in OO generaly refers to implementation. I use abstract classes when I want to force some implementation details to the childs else I go with interfaces.

Of course, abstract classes are useful not only in forcing implementation but also in sharing some specific details among many related classes.


P
Peter Miehle

in java you can inherit from one (abstract) class to "provide" functionality and you can implement many interfaces to "ensure" functionality


lil' hint: if you want to inherit from an abstract class and an interface, be sure, that the abstract class implements the interface
M
Mesh

This can be a very difficult call to make...

One pointer I can give: An object can implement many interfaces, whilst an object can only inherit one base class( in a modern OO language like c#, I know C++ has multiple inheritance - but isn't that frowned upon?)


Multiple inheritence allows for Mixin's to be implemented seemlessly, well written Mixin's are a breeze to work with but very hard to come by and difficult to write without falling short somewhere. Mixin's are pretty cool as a whole though IMO.
Actually i didn't, multiple inheritence is indeed a sure debate sparker between us geeks i see absolutely no reason to downvote. In fact i upvoted your answer.
The only point i tried to make is that ways to Mixin in languages with Single Inheritence is possible too (C#, PHP, javascript) but trough hacky behaviour or tacky syntax. I love Mixin's when they work but i'm still undecisive on wheter to from upon multiple inheritence or not.
This answer is more of a syntactical difference than a design difference. I think he is asking for a design difference
a
annakata

Purely on the basis of inheritance, you would use an Abstract where you're defining clearly descendant, abstract relationships (i.e. animal->cat) and/or require inheritance of virtual or non-public properties, especially shared state (which Interfaces cannot support).

You should try and favour composition (via dependency injection) over inheritance where you can though, and note that Interfaces being contracts support unit-testing, separation of concerns and (language varying) multiple inheritance in a way Abstracts cannot.


D
Dmitri Nesteruk

One interesting location where interfaces fare better than abstract classes is when you need to add extra functionality to a group of (related or unrelated) objects. If you cannot give them a base abstract class (e.g., they are sealed or already have a parent), you can give them a dummy (empty) interface instead, and then simply write extension methods for that interface.


A
Ayoub Boumzebra

The short answer: An abstract class allows you to create functionality that subclasses can implement or override. An interface only allows you to define functionality, not implement it. And whereas a class can extend only one abstract class, it can take advantage of multiple interfaces.


G
Gerrie Schenck

An abstract class can have implementations.

An interface doesn't have implementations, it simply defines a kind of contract.

There can also be some language-dependent differences: for example C# does not have multiple inheritance, but multiple interfaces can be implemented in a class.


When you say, "a kind of contract", do you mean like in web services?
Technically speaking, web services do not work with interfaces. With contract I mean the user of an object knows which methods are present on that object. For example an interface IMouse will have a Move method, and a left and right mouse button event.
A
Arslan Ali

Basic thumb rule is: For "Nouns" use Abstract class and for "Verbs" use interface

E.g: car is an abstract class and drive, we can make it an interface.


This doesn't make sense, we can also put the functionality of drive in the car - that is an abstract class.
s
satheesh

If we have an implementation that will be the same for all the derived classes and at that time it is better to use the abstract class over an interface. when we have an interface, we can move our implementation to any class that implements interface. In abstract class, it avoids code duplication and share the implementation for all derived class. The interfaces allow to develop loosely coupled systems which helps for better testing.


M
Mohammadreza

Both are contract for class definition:

Conclusion 1: Both intent is object generalization

In defining abstract classes, they can have default implementation also.

Conclusion 2: distinguish is in Behavioral generalization Design

In utilizing abstract classes, classes can inherit from just one abstract class

Conclusion 3: abstract class has limitation in utilizing. It means limitation in Behavioral generalization.

Final Conclusion-When to use which: Distinguish is in behavioral generalization level

In designing behavioral of classes, If functionality is just conceptually limited among determined classes or in other word, is share among determined class, use abstract class. but if functionality is more general than determined classes or we can/want add functionality to other classes, use interface as a contract.