Re: Using CNI on inner classes

[Andrew Haley <aph at redhat dot com>]

Bryce McKinlay wrote:
> On Thu, Aug 13, 2009 at 10:10 AM, Andrew Haley<aph@redhat.com> wrote:
>
>> If the method is declared private, it really will be private, even
>> to cni.
>>
>> If you really need to breach privacy, you can do this in Test$Inner.h:
>>
>> class Test$Inner : public ::java::lang::Object
>> {
>>  friend class ::Test;
>>
>> but I can't really see the point of private methods in an inner class
>> that you actually intend to be called from the enclosing class.
>
> In Java the private methods are accessible to the enclosing class.

Ah yes, that rule, the one that requires all the "synthetic access$0"
business.  I just did a web search to find out the reason for this
rather odd rule, and I found "... inner classes can access all members
of the declaring class, even private members. In fact, the inner class
itself is said to be a member of the class; therefore, following the
rules of object-oriented engineering, it should have access to all
members of the class."
http://onjava.com/pub/a/onjava/excerpt/HardcoreJava_chap06/index.html

> So, ideally gcjh would add friend declarations, like the above,
> automatically to inner classes to approximate the Java semantics.

That's a nice idea.  There's no reason why not; I just think no-one
ever thought of it.

> But, the easiest fix here is just to make the private method
> package-private instead.

I think so.

Andrew.