> Actually from my tests it doesn't seem to get that far as it returns (at > least here) in line 347 after calling GetMethodBaseSignature() at line > 344. That happens when the MethodBase has already been imported; it's returning the MethodReference which it created before, rather than creating a new one. That GetMethodBaseSignature() call followed by GetMemberReference() is just trying to find the cached MethodReference. The point of interest happens later in the same method, when the MethodBase is imported the first time.
> I found that "IsGenericMethod" and "IsGenericMethodDefinition" are > both false, seems strange. Also "ContainsGenericParameters" is false. No, that's right. The *type* is generic, but the *method* (ctor in this case) is not. Those are two independent things. (And constructors are never generic methods.) Also, all generic parameters on the method Nullable`1<int32>..ctor(int32) are closed; see the Remarks section of that method's docs. > I didn't read the ECMA-335 yet and I'm still learning the Cecil source so > sorry if I'm getting it wrong, this is what I understand so far. No problem! If we say something incorrect, someone will correct us and we'll all learn from it. -- Keith --~--~---------~--~----~------------~-------~--~----~ -- mono-cecil -~----------~----~----~----~------~----~------~--~---
