I have read the definition of “UnderlyingSystemType” namely that it “Indicates the type provided by the common language runtime that represents this type.”
There is a related link on SO at When does the UnderlyingSystemType differ from the current Type instance but I cannot tell from the answers if it is actually at all possible to have an object whose type would be different from the UnderlyingSytemType.
I recently learned about CLS compliance, and that unsigned ints are not CLS compliant. And I was really hopeful that maybe that would do it, that the non-CLS compliant types might have a different underlying type, but that is not the case.
For what it’s worth, the code I used to test is:
Byte t01 = 1;
SByte t02 = 1;
Int16 t03 = 1;
UInt16 t04 = 1;
Int32 t05 = 1;
UInt32 t06 = 1;
Int64 t07 = 1;
UInt64 t08 = 1;
Single t09 = 1;
Double t10 = 1;
Decimal t11 = 1;
Console.WriteLine(t01.GetType().Equals(t01.GetType().UnderlyingSystemType));
Console.WriteLine(t02.GetType().Equals(t02.GetType().UnderlyingSystemType));
Console.WriteLine(t03.GetType().Equals(t03.GetType().UnderlyingSystemType));
Console.WriteLine(t04.GetType().Equals(t04.GetType().UnderlyingSystemType));
Console.WriteLine(t05.GetType().Equals(t05.GetType().UnderlyingSystemType));
Console.WriteLine(t06.GetType().Equals(t06.GetType().UnderlyingSystemType));
Console.WriteLine(t07.GetType().Equals(t07.GetType().UnderlyingSystemType));
Console.WriteLine(t08.GetType().Equals(t08.GetType().UnderlyingSystemType));
Console.WriteLine(t09.GetType().Equals(t09.GetType().UnderlyingSystemType));
Console.WriteLine(t10.GetType().Equals(t10.GetType().UnderlyingSystemType));
Console.WriteLine(t11.GetType().Equals(t11.GetType().UnderlyingSystemType));
When run, I just get a bunch of trues.
My question is is there a situation in which an underlying system type of an object can be different from its type? And what is the purpose of this distinction, is it merely to allow the definition of hypothetical types that cannot be instantiated? I can’t even create a new Type using the new keyword. And all of the properties of Type are get-only, so I’m lost as to what this feature does. Is the distinction useful in other languages, maybe?
Typeis an abstract class. The most common implementation you’ll see isRuntimeType, which is what objects typically are, but anyone can create an implementation ofType.RuntimeType‘sUnderlyingSystemTypewill just return the sameRuntimeType. As far as I’ve seen, this is only really significant if you have a method that takes aTypeor construct such a type locally, not if you get an object and callGetType. Here’s an example to help you understand: