On Aug 26, 2009, at 1:51 PM, Mike Stump wrote:
> On Aug 25, 2009, at 7:58 PM, Douglas Gregor wrote:
>> Checking that this is a reference to a virtual method is necessary,
>> but not sufficient. We also need to check that the name was not
>> qualified, e.g.,
>
> Fixed by putting in the code, commented out, but with a FIXME.
>
> struct test12_A {
> virtual void foo0() { }
> virtual void foo();
> } *test12_pa;
>
> void foo() {
> test12_pa->test12_A::foo();
> }
>
> -ast-prints as:
>
> void foo() {
> test12_pa->foo();
> }
>
> :-(
That's kinda why I said:
"This will probably require some more AST support, to keep track of
the nested-
name-specifier in a MemberExpr."
It's not terribly complicated to add that AST support, since Sema
already has all of the right information.
- Doug
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
