Let's say I have 2 interfaces defined like so:
public interface ISkuItem
{
public string SKU { get; set; }
}
public interface ICartItem : ISkuItem
{
public int Quantity { get; set; }
public bool IsDiscountable { get; set; }
}
When I go to implement the interface in C#, VS produces the following templated code:
public class CartItem : ICartItem
{
#region ICartItem Members
public int Quantity { get {...} set {...} }
public bool IsDiscountable { get {...} set {...} }
#endregion
#region ISkuItem Members
public string SKU { get {...} set {...} }
#endregion
}
In VB.NET, the same class is built out like so:
Public Class CartItem
Implements ICartItem
Public Property IsDiscountable As Boolean Implements ICartItem.IsDiscountable
'GET SET'
End Property
Public Property Quantity As Integer Implements ICartItem.Quantity
'GET SET'
End Property
Public Property SKU As String Implements ISkuItem.SKU
'GET SET'
End Property
End Class
VB.NET explicitly requires you to add Implements IInterfaceName.PropertyName
after each property that gets implemented whereas C# simply uses region
s to indicate which properties and methods belong to the interface.
Interestingly in VB.NET, on the SKU
property, I can specify either Implements ISkuItem.SKU
or Implements ICartItem.SKU
. Although the template built by VS defaults to ISkuItem
, I can also specify ICartItem
if I want. Oddly, because C# only uses region
s to block out inherited properties, it seems that I can't explicitly specify the implementing interface of SKU
in C# like I can in VB.NET.
My question is: Is there any importance behind being able to specify one interface or another to implement properties in VB.NET, and if so, is there a way to mimic this functionality in C#? Furthermore, what is the effect of specifying one interface over another when implementing properties?
I think the other answers are actually a little off the mark here.
In the example you've posted, one interface inherits from the other. This simply means that it offers the same members as its base, plus some additional members.
These are not two independent interfaces that happen to expose members with the same name. ICartItem.SKU
is the same thing as ISkuItem.SKU
. That ICartItem
inherits from ISkuItem
simply means that ISkuItem
, as an interface, represents a subset of the functionality offered by ICartItem
.
Consider this code:
class CartItem : ICartItem
{
public int Quantity { get; set; }
public bool IsDiscountable { get; set; }
string ISkuItem.SKU
{
get { return "ISkuItem"; }
set { throw new NotSupportedException(); }
}
string ICartItem.SKU
{
get { return "ICartItem"; }
set { throw new NotSupportedException(); }
}
}
This class will not compile. You cannot define ICartItem.SKU
explicitly in this case, because ICartItem.SKU
is just ISkuItem.SKU
. There's no "other" SKU
property to define.
So, to answer your questions directly:
Is there any importance behind being
able to specify one interface or
another to implement properites in
VB.NET?
When they are separate, unrelated interfaces: yes.
As others have pointed out, you can provide different implementations for different interfaces' members sharing a common name.
But when one interface inherits from the other: no.
It doesn't matter because they're the same thing.
What is the effect of specifying one
interface over another when
implementing properites?
Again, if they're unrelated interfaces, it has the effect already discussed by others: providing different implementations for the two interfaces. But if one derives from the other, it has no effect.
Yes this is importance, this called Explicit and Implicit interface implementation.
In C# you can do this by prefixing method name with an interface name, like that:
public class CartItem : ICartItem, ISkuItem
{
#region ICartItem Members
public int Quantity { get {...} set {...} }
public bool IsDiscountable { get {...} set {...} }
#endregion
#region ISkuItem Members
public string ISkuItem.SKU { get {...} set {...} } //like this
public string ICartItem.SKU { get {...} set {...} } //like this
#endregion
}
Yes, you can implement different functionality behind each interface. Assume both interfaces have the same signature. Depending on which interface you cast your implementation to will control which interface is executed.
... C# Example of explict interfaces ...
public interface ITest1 { string Get(); }
public interface ITest2 { string Get(); }
// new is just to get rid of a compiler warning
public interface ITest3 : ITest1, ITest2 { new string Get(); }
public class MyTest : ITest1, ITest2
{
public string Get() { return "local"; }
string ITest1.Get() { return "hello"; }
string ITest2.Get() { return "world"; }
string ITest3.Get() { return "hi"; }
}
class Program
{
static void Main(string[] args)
{
var mytest = new MyTest();
// note that if mytest.Get() does not exist if all of the
// interfaces are explicit
var v0 = mytest.Get(); //local
var v1 = ((ITest1)mytest).Get(); //hello
var v2 = ((ITest2)mytest).Get(); //world
var v3 = ((ITest3)mytest).Get(); //hi
}
}
... Similar Code in VB.Net ...
Module Module1
Sub Main()
Dim myinstance = New MyTest()
Dim v0 = myinstance.DoWork() 'local
'By the way... note that the following methods are called
'by the interface signature and not the defind method name
'in the class
Dim v1 = DirectCast(myinstance, ITest1).DoWork() 'hello
Dim v2 = DirectCast(myinstance, ITest2).DoWork() 'world
Dim v3 = DirectCast(myinstance, ITest3).DoWork() 'hi
End Sub
End Module
Public Interface ITest1
Function DoWork() As String
End Interface
Public Interface ITest2
Function DoWork() As String
End Interface
Public Interface ITest3
Inherits ITest1
Inherits ITest2
Shadows Function DoWork() As String
End Interface
Public Class MyTest
Implements ITest3
'Implements ITest1
'Implements ITest2
Public Function DoWork() As String
Return "local"
End Function
Private Function DoWork1() As String Implements ITest1.DoWork
Return "hello"
End Function
Private Function DoWork2() As String Implements ITest2.DoWork
Return "world"
End Function
Private Function DoWork3() As String Implements ITest3.DoWork
Return "hi"
End Function
End Class