On Sat, Dec 15, 2012 at 11:31:22AM -0800, Jonathan M Davis wrote:
> On Saturday, December 15, 2012 19:50:34 Iain Buclaw wrote:
> > On Saturday, 15 December 2012 at 18:38:29 UTC, H. S. Teoh wrote:
> > > With latest git dmd:
> > >   auto makeVoldemort(int x) {
> > >
> > >           struct Voldemort {
> > >
> > >                   @property int value() { return x; }
> > >
> > >           }
> > >           return Voldemort();
> > >
> > >   }
> > >   void main() {
> > >
> > >           auto v = makeVoldemort();
> > >           writeln(v.value);
> > >
> > >   }
> > >
> > > Compile error:
> > >   test.d(3): Error: function test.makeVoldemort.Voldemort.value
> > >
> > > cannot access frame of function test.makeVoldemort
> > >
> > > Changing 'struct' to 'class' works. Is this deliberate, or is it a
> > > bug?  It is certainly inconsistent with Walter's article on
> > > Voldemort types, which uses structs as examples.

[...]

> > Pretty certain it's deliberate.  No closure is created for nested
> > structs to access it's parent, complying with it's POD behaviour.
>
> static nested structs don't have access to their outer scopes.
> Non-static structs do. This reeks of a bug.

[...]

Found the reference in TDPL, §7.1.9 (p.263):

        Nested structs embed the magic "frame pointer" that allows them
        to access outer values such as a and b in the example above.
        [...] If you want to define a nested struct without that
        baggage, just prefix struct with static in the definition of
        Local, which makes Local a regular struct and consequently
        prevents it from accessing a and b.

Ironically enough, Andrei in the subsequent paragraph discourages the
use of such nested structs, whereas Walter's article promotes the use of
such Voldemort types as a "happy discovery". :)

Anyway, filed a bug:

        http://d.puremagic.com/issues/show_bug.cgi?id=9162