Why is it impossible to override a getter-only property and add a setter?

Question

Why is the following C# code not allowed:

public abstract class BaseClass
{
    public abstract int Bar { get;}
}

public class ConcreteClass : BaseClass
{
    public override int Bar
    {
        get { return 0; }
        set {}
    }
}

CS0546 'ConcreteClass.Bar.set': cannot override because 'BaseClass.Bar' does not have an overridable set accessor

Answer 1

I think the main reason is simply that the syntax is too explicit for this to work any other way. This code:

public override int MyProperty { get { ... } set { ... } }

is quite explicit that both the get and the set are overrides. There is no set in the base class, so the compiler complains. Just like you can't override a method that's not defined in the base class, you can't override a setter either.

You might say that the compiler should guess your intention and only apply the override to the method that can be overridden (i.e. the getter in this case), but this goes against one of the C# design principles - that the compiler must not guess your intentions, because it may guess wrong without you knowing.

I think the following syntax might do nicely, but as Eric Lippert keeps saying, implementing even a minor feature like this is still a major amount of effort...

public int MyProperty
{
    override get { ... } // not valid C#
    set { ... }
}

or, for autoimplemented properties,

public int MyProperty { override get; set; } // not valid C#

Answer 2

It's possible.

tl;dr– You can override a get-only method with a setter if you want. It's basically just:

1. Create a new property that has both a get and a set using the same name.

2. override the prior get to alias the new get.

This enables us to override properties with get/set even if they lacked a setter in their base definition.

---

Situation: Pre-existing get-only property.

You have some class structure that you can't modify. Maybe it's just one class, or it's a pre-existing inheritance tree. Whatever the case, you want to add a set method to a property, but can't.

public abstract class A                     // Pre-existing class; can't modify
{
    public abstract int X { get; }          // You want a setter, but can't add it.
}
public class B : A                          // Pre-existing class; can't modify
{
    public override int X { get { return 0; } }
}

---

Problem: Can't override the get-only with get/set.

You want to override with a get/set property, but it won't compile.

public class C : B
{
    private int _x;
    public override int X
    {
        get { return _x; }
        set { _x = value; }   //  Won't compile
    }
}

---

Solution: Use an abstract intermediate layer.

While you can't directly override with a get/set property, you can:

1. Create a new get/set property with the same name.

2. override the old get method with an accessor to the new get method to ensure consistency.

So, first you write the abstract intermediate layer:

public abstract class C : B
{
    //  Seal off the old getter.  From now on, its only job
    //  is to alias the new getter in the base classes.
    public sealed override int X { get { return this.XGetter; }  }
    protected abstract int XGetter { get; }
}

Then, you write the class that wouldn't compile earlier. It'll compile this time because you're not actually override'ing the get-only property; instead, you're replacing it using the new keyword.

public class D : C
{
    private int _x;
    public new virtual int X
    {
        get { return this._x; }
        set { this._x = value; }
    }

    //  Ensure base classes (A,B,C) use the new get method.
    protected sealed override int XGetter { get { return this.X; } }
}

---

Result: Everything works!

var d = new D();

var a = d as A;
var b = d as B;
var c = d as C;

Print(a.X);      // Prints "0", the default value of an int.
Print(b.X);      // Prints "0", the default value of an int.
Print(c.X);      // Prints "0", the default value of an int.
Print(d.X);      // Prints "0", the default value of an int.

// a.X = 7;      // Won't compile: A.X doesn't have a setter.
// b.X = 7;      // Won't compile: B.X doesn't have a setter.
// c.X = 7;      // Won't compile: C.X doesn't have a setter.
d.X = 7;         // Compiles, because D.X does have a setter.

Print(a.X);      // Prints "7", because 7 was set through D.X.
Print(b.X);      // Prints "7", because 7 was set through D.X.
Print(c.X);      // Prints "7", because 7 was set through D.X.
Print(d.X);      // Prints "7", because 7 was set through D.X.

---

Discussion.

This method allows you to add set methods to get-only properties. You can also use it to do stuff like:

1. Change any property into a get-only, set-only, or get-and-set property, regardless of what it was in a base class.

2. Change the return type of a method in derived classes.

The main drawbacks are that there's more coding to do and an extra abstract class in the inheritance tree. This can be a bit annoying with constructors that take parameters because those have to be copy/pasted in the intermediate layer.

---

Bonus: You can change the property's return-type.

As a bonus, you can also change the return type if you want.

• If the base definition was get-only, then you can use a more-derived return type.

• If the base definition was set-only, then you can use a less-derived return type.

• If the base definition was already get/set, then:

  • you can use a more-derived return type if you make it set-only;

  • you can use a less-derived return type if you make it get-only.

In all cases, you can keep the same return type if you want.

Answer 3

I stumbled across the very same problem today and I think I have a very valid reason for wanting this.

First I'd like to argue that having a get-only property doesn't necessarily translate into read-only. I interpret it as "From this interface/abstract class you can get this value", that doesn't mean that some implementation of that interface/abstract class won't need the user/program to set this value explicitly. Abstract classes serve the purpose of implementing part of the needed functionality. I see absolutely no reason why an inherited class couldn't add a setter without violating any contracts.

The following is a simplified example of what I needed today. I ended up having to add a setter in my interface just to get around this. The reason for adding the setter and not adding, say, a SetProp method is that one particular implementation of the interface used DataContract/DataMember for serialization of Prop, which would have been made needlessly complicated if I had to add another property just for the purpose of serialization.

interface ITest
{
    // Other stuff
    string Prop { get; }
}

// Implements other stuff
abstract class ATest : ITest
{
    abstract public string Prop { get; }
}

// This implementation of ITest needs the user to set the value of Prop
class BTest : ATest
{
    string foo = "BTest";
    public override string Prop
    {
        get { return foo; }
        set { foo = value; } // Not allowed. 'BTest.Prop.set': cannot override because 'ATest.Prop' does not have an overridable set accessor
    }
}

// This implementation of ITest generates the value for Prop itself
class CTest : ATest
{
    string foo = "CTest";
    public override string Prop
    {
        get { return foo; }
        // set; // Not needed
    }
}

I know this is just a "my 2 cents" post, but I feel with the original poster and trying to rationalize that this is a good thing seems odd to me, especially considering that the same limitations doesn't apply when inheriting directly from an interface.

Also the mention about using new instead of override does not apply here, it simply doesn't work and even if it did it wouldn't give you the result wanted, namely a virtual getter as described by the interface.

Answer 4

I agree that not being able to override a getter in a derived type is an anti-pattern. Read-Only specifies lack of implementation, not a contract of a pure functional (implied by the top vote answer).

I suspect Microsoft had this limitation either because the same misconception was promoted, or perhaps because of simplifying grammar; though, now that scope can be applied to get or set individually, perhaps we can hope override can be too.

The misconception indicated by the top vote answer, that a read-only property should somehow be more "pure" than a read/write property is ridiculous. Simply look at many common read only properties in the framework; the value is not a constant / purely functional; for example, DateTime.Now is read-only, but anything but a pure functional value. An attempt to 'cache' a value of a read only property assuming it will return the same value next time is risky.

In any case, I've used one of the following strategies to overcome this limitation; both are less than perfect, but will allow you to limp beyond this language deficiency:

   class BaseType
   {
      public virtual T LastRequest { get {...} }
   }

   class DerivedTypeStrategy1
   {
      /// get or set the value returned by the LastRequest property.
      public bool T LastRequestValue { get; set; }

      public override T LastRequest { get { return LastRequestValue; } }
   }

   class DerivedTypeStrategy2
   {
      /// set the value returned by the LastRequest property.
      public bool SetLastRequest( T value ) { this._x = value; }

      public override T LastRequest { get { return _x; } }

      private bool _x;
   }

Answer 5

You could perhaps go around the problem by creating a new property:

public new int Bar 
{            
    get { return 0; }
    set {}        
}

int IBase.Bar { 
  get { return Bar; }
}

Answer 6

I can understand all your points, but effectively, C# 3.0's automatic properties get useless in that case.

You can't do anything like that:

public class ConcreteClass : BaseClass
{
    public override int Bar
    {
        get;
        private set;
    }
}

IMO, C# should not restrict such scenarios. It's the responsibility of the developer to use it accordingly.

Answer 7

The problem is that for whatever reason Microsoft decided that there should be three distinct types of properties: read-only, write-only, and read-write, only one of which may exist with a given signature in a given context; properties may only be overridden by identically-declared properties. To do what you want it would be necessary to create two properties with the same name and signature--one of which was read-only, and one of which was read-write.

Personally, I wish that the whole concept of "properties" could be abolished, except that property-ish syntax could be used as syntactic sugar to call "get" and "set" methods. This would not only facilitate the 'add set' option, but would also allow for 'get' to return a different type from 'set'. While such an ability wouldn't be used terribly often, it could sometimes be useful to have a 'get' method return a wrapper object while the 'set' could accept either a wrapper or actual data.

Answer 8

Here is a work-around in order to achieve this using Reflection:

var UpdatedGiftItem = // object value to update;

foreach (var proInfo in UpdatedGiftItem.GetType().GetProperties())
{
    var updatedValue = proInfo.GetValue(UpdatedGiftItem, null);
    var targetpropInfo = this.GiftItem.GetType().GetProperty(proInfo.Name);
    targetpropInfo.SetValue(this.GiftItem, updatedValue,null);
}

This way we can set object value on a property that is readonly. Might not work in all the scenarios though!

Answer 9

You should alter your question title to either detail that your question is solely in regards to overriding an abstract property, or that your question is in regards to generally overriding a class's get-only property.

---

If the former (overriding an abstract property)

That code is useless. A base class alone shouldn't tell you that you're forced to override a Get-Only property (Perhaps an Interface). A base class provides common functionality which may require specific input from an implementing class. Therefore, the common functionality may make calls to abstract properties or methods. In the given case, the common functionality methods should be asking for you to override an abstract method such as:

public int GetBar(){}

But if you have no control over that, and the functionality of the base class reads from its own public property (weird), then just do this:

public abstract class BaseClass
{
    public abstract int Bar { get; }
}

public class ConcreteClass : BaseClass
{
    private int _bar;
    public override int Bar
    {
        get { return _bar; }
    }
    public void SetBar(int value)
    {
        _bar = value;
    }
}

I want to point out the (weird) comment: I would say a best-practice is for a class to not use its own public properties, but to use its private/protected fields when they exist. So this is a better pattern:

public abstract class BaseClass {
    protected int _bar;
    public int Bar { get { return _bar; } }
    protected void DoBaseStuff()
    {
        SetBar();
        //Do something with _bar;
    }
    protected abstract void SetBar();
}

public class ConcreteClass : BaseClass {
    protected override void SetBar() { _bar = 5; }
}

---

If the latter (overriding a class's get-only property)

Every non-abstract property has a setter. Otherwise it's useless and you shouldn't care to use it. Microsoft doesn't have to allow you to do what you want. Reason being: the setter exists in some form or another, and you can accomplish what you want Veerryy easily.

The base class, or any class where you can read a property with {get;}, has SOME sort of exposed setter for that property. The metadata will look like this:

public abstract class BaseClass
{
    public int Bar { get; }
}

But the implementation will have two ends of the spectrum of complexity:

Least Complex:

public abstract class BaseClass
{
    private int _bar;
    public int Bar { 
        get{
            return _bar;
        }}
    public void SetBar(int value) { _bar = value; }
}

Most Complex:

public abstract class BaseClass
{
    private int _foo;
    private int _baz;
    private int _wtf;
    private int _kthx;
    private int _lawl;

    public int Bar
    {
        get { return _foo * _baz + _kthx; }
    }
    public bool TryDoSomethingBaz(MyEnum whatever, int input)
    {
        switch (whatever)
        {
            case MyEnum.lol:
                _baz = _lawl + input;
                return true;
            case MyEnum.wtf:
                _baz = _wtf * input;
                break;
        }
        return false;
    }
    public void TryBlowThingsUp(DateTime when)
    {
        //Some Crazy Madeup Code
        _kthx = DaysSinceEaster(when);
    }
    public int DaysSinceEaster(DateTime when)
    {
        return 2; //<-- calculations
    }
}
public enum MyEnum
{
    lol,
    wtf,
}

My point being, either way, you have the setter exposed. In your case, you may want to override int Bar because you don't want the base class to handle it, don't have access to review how it's handling it, or were tasked to hax some code real quick'n'dirty against your will.

In both Latter and Former (Conclusion)

Long-Story Short: It isn't necessary for Microsoft to change anything. You can choose how your implementing class is set up and, sans the constructor, use all or none of the base class.

Answer 10

Solution for only a small subset of use cases, but nevertheless: in C# 6.0 "readonly" setter is automatically added for overridden getter-only properties.

public abstract class BaseClass
{
    public abstract int Bar { get; }
}

public class ConcreteClass : BaseClass
{
    public override int Bar { get; }

    public ConcreteClass(int bar)
    {
        Bar = bar;
    }
}

Answer 11

This is not impossible. You simply have to use the "new" keyword in your property. For example,

namespace {
    public class Base {
        private int _baseProperty = 0;

        public virtual int BaseProperty {
            get {
                return _baseProperty;
            }
        }

    }

    public class Test : Base {
        private int _testBaseProperty = 5;

        public new int BaseProperty {
            get {
                return _testBaseProperty;
            }
            set {
                _testBaseProperty = value;
            }
        }
    }
}

It appears as if this approach satisfies both sides of this discussion. Using "new" breaks the contract between the base class implementation and the subclass implementation. This is necessary when a Class can have multiple contracts (either via interface or base class).

Hope this helps

Answer 12

Because that would break the concept of encapsulation and implementation hiding. Consider the case when you create a class, ship it, and then the consumer of your class makes himself able to set a property for which you originally provide a getter only. It would effectively disrupt any invariants of your class which you can depend on in your implementation.

Answer 13

A read-only property in the base class indicates that this property represents a value that can always be determined from within the class (for example an enum value matching the (db-)context of an object). So the responsibillity of determining the value stays within the class.

Adding a setter would cause an awkward issue here: A validation error should occur if you set the value to anything else than the single possible value it already has.

Rules often have exceptions, though. It is very well possible that for example in one derived class the context narrows the possible enum values down to 3 out of 10, yet the user of this object still needs to decide which one is correct. The derived class needs to delegate the responsibillity of determining the value to the user of this object. Important to realize is that the user of this object should be well aware of this exception and assume the responsibillity to set the correct value.

My solution in these kind of situations would be to leave the property read-only and add a new read-write property to the derived class to support the exception. The override of the original property will simply return the value of the new property. The new property can have a proper name indicating the context of this exception properly.

This also supports the valid remark: "make it as hard as possible for misunderstandings to crop up" by Gishu.

Answer 14

Because a class that has a read-only property (no setter) probably has a good reason for it. There might not be any underlying datastore, for example. Allowing you to create a setter breaks the contract set forth by the class. It's just bad OOP.

Answer 15

Because at the IL level, a read/write property translates into two (getter and setter) methods.

When overriding, you have to keep supporting the underlying interface. If you could add a setter, you would effectively be adding a new method, which would remain invisible to the outside world, as far as your classes' interface was concerned.

True, adding a new method would not be breaking compatibility per se, but since it would remain hidden, decision to disallow this makes perfect sense.

Answer 16

Because the writer of Baseclass has explicitly declared that Bar has to be a read-only property. It doesn't make sense for derivations to break this contract and make it read-write.

I'm with Microsoft on this one.
Let's say I'm a new programmer who has been told to code against the Baseclass derivation. i write something that assumes that Bar cannot be written to (since the Baseclass explicitly states that it is a get only property). Now with your derivation, my code may break. e.g.

public class BarProvider
{ BaseClass _source;
  Bar _currentBar;

  public void setSource(BaseClass b)
  {
    _source = b;
    _currentBar = b.Bar;
  }

  public Bar getBar()
  { return _currentBar;  }
}

Since Bar cannot be set as per the BaseClass interface, BarProvider assumes that caching is a safe thing to do - Since Bar cannot be modified. But if set was possible in a derivation, this class could be serving stale values if someone modified the _source object's Bar property externally. The point being 'Be Open, avoid doing sneaky things and surprising people'

Update: Ilya Ryzhenkov asks 'Why don't interfaces play by the same rules then?' Hmm.. this gets muddier as I think about it.
An interface is a contract that says 'expect an implementation to have a read property named Bar.' Personally I'm much less likely to make that assumption of read-only if I saw an Interface. When i see a get-only property on an interface, I read it as 'Any implementation would expose this attribute Bar'... on a base-class it clicks as 'Bar is a read-only property'. Of course technically you're not breaking the contract.. you're doing more. So you're right in a sense.. I'd close by saying 'make it as hard as possible for misunderstandings to crop up'.