Question

Possible Duplicate:
Who actually last decide what is the Generic Type?

As my question title says can anyone please explain me what is bounds (upper bound, lower bound, exact bound) and how do they play role in type inference with examples?

Consider simple code :

void func<T> ( T firstparam , T secondparam) { }

and caller call it

func( 23 , 23.23 ); 

What are the bounds while type inference process happens and how they are used by inference process?. Should my presented example is way trivial and will not have any upper bounds etc, please include your own example which will represent the idea.

Answer 1

In your examples there are no bounds in type inference. If you call it like

 func( 23 , 23.23 ); 

Then candidates for types would be int and double. Int can be casted implicitly to double but not vise-versa, so the fixed type for T is double. This has nothing to do with bounds, but type compatibility.

But however, if you call it like

func( new object() , 23.23 );

Then the upper bound for T will be object, the lower bound will be double. In such case fixed type for T will be object.

Eric Lippert describe what bounds are and why they separate lower, upper and exact bounds at blogpost on How do we ensure that method type inference terminates?

Jon Skeet describe in detail the process of type inference at 9.4.3 Two-phase type inference in his wonderful book C# in Depth. Please read carefully his description for type inference for listing 9.11