I'm trying to do some deep metaprogramming magic in D2, and an issue that I
can't seem to get around is that I can't find a way to get a tuple of
struct/class fields, or something similarly useful, with reference semantics.  Ex:

struct S {
    uint foo = 1;
}

void main() {
    S s;
    foreach(element; s.tupleof)
         element = 2;
    writeln(s.foo);  //Still 1.
}

The obvious thing, which I already tried was:

foreach(ref element; s.tupleof)  //Doesn't compile.

Is there a way to get a tuple representation of a class/struct such that I can change the fields of the class/struct through the tuple representation, or should I file this as an enhancement?

On Thu, 9 Oct 2008 01:49:36 +0000 (UTC), dsimcha <dsimcha@yahoo.com>
wrote:

>I'm trying to do some deep metaprogramming magic in D2, and an issue that I
>can't seem to get around is that I can't find a way to get a tuple of
>struct/class fields, or something similarly useful, with reference semantics.  Ex:
>
>struct S {
>    uint foo = 1;
>}
>
>void main() {
>    S s;
>    foreach(element; s.tupleof)
>         element = 2;
>    writeln(s.foo);  //Still 1.
>}
>
>The obvious thing, which I already tried was:
>
>foreach(ref element; s.tupleof)  //Doesn't compile.
>
>Is there a way to get a tuple representation of a class/struct such that I can change the fields of the class/struct through the tuple representation, or should I file this as an enhancement?

Possible workaround:

struct S {
    uint foo = 1;
}

void main() {
    S s;
    foreach(i, element; s.tupleof)
         s.tupleof[i] = 2;

    writeln(s.foo);
}

But you file it as enhancement anyway.

Reply to Max,

> struct S {
> uint foo = 1;
> }
> void main() {
> S s;
> foreach(i, element; s.tupleof)
> s.tupleof[i] = 2;
> writeln(s.foo);
> }

That's what I have ended up doing every time I have needed this.