Re: PR java/19870 (Part 2/2): Handle private static methods accessed across nested classes

[Ranjit Mathew <rmathew at gmail dot com>]

On 18 Jun 2005 23:02:46 -0600, Tom Tromey <tromey@redhat.com> wrote:
> >>>>> "Ranjit" == Ranjit Mathew <rmathew@gmail.com> writes:
> 
> Ranjit>   FAIL: 15.12.3-explicit-constructor-9
> 
> Ranjit> The latter confuses me (but both Jikes and javac 1.5.0.03
> Ranjit> compile it without errors). I'd be grateful if someone could
> Ranjit> explain exactly why it should *not* be an error.
> 
> The test case:
> 
>     class T15123ec9 {
>         private int m() { return 1; }
>         T15123ec9(int i) {}
>         class Sub extends T15123ec9 {
>             Sub() {
>                 super(m()); // m is not inherited, so it is the enclosing m
>             }
>         }
>     }

Thank you for pasting the testcase. I ought to have done it
myself and I apologise for the oversight.


> If m() were not private, the call would be to Sub.m(), which would be
> invalid because it would be a use of 'this' in a static context.

What static context?


> However, m() is not inherited.  So, the call is to
> 'T15123ec9.this.m()', which is valid as it doesn't refer to the object
> being constructed.

Yes, now I see it. I got a bit confused by the fact that
both 'this$0' and 'this' are instances of T15123ec9 (I
thought them to be the *same* instance...).

Thanks for the explanation.

Ranjit.

-- 
Ranjit Mathew      Email: rmathew AT gmail DOT com

Bangalore, INDIA.    Web: http://ranjitmathew.hostingzero.com/