Maybe this has been posted before, but I don't recall anybody managing to isolate this into a handful of lines of code, so I may as well post it now. (apologies if it has already been posted)

a.d:
	import b;

	class A
	{
	    B[] b;
	}

b.d:
	import a;

	class B : A
	{
	}


D:\temp\dtest>dmd a.d b.d
b.d: class B forward reference of base class A

 -- andy

On Mon, 19 Jan 2004 18:45:19 -0800, Andy Friesen wrote:

> Maybe this has been posted before, but I don't recall anybody managing to isolate this into a handful of lines of code, so I may as well post it now. (apologies if it has already been posted)
> 
> a.d:
> 	import b;
> 
> 	class A
> 	{
> 	    B[] b;
> 	}
> 
> b.d:
> 	import a;
> 
> 	class B : A
> 	{
> 	}
> 
> 
> D:\temp\dtest>dmd a.d b.d
> b.d: class B forward reference of base class A
> 
>   -- andy

I was going to suggest:
#################
module A;

class A
{
 	private import B;
    B b;
}
#################
module B;

import A;

class B : A
{
}
#################
but that kinda segfaults the compiler (0.78)
even if we remove the "module" statments.

which if funny because the test I done before
works fine:
#################
class A
{
	private import std.string;

	void a()
	{
		printf("A.a std.string.toString(1)== %.*s\n",std.string.toString(1));
	}
}

#################
private import ImpA;

class B
{
	private import std.string;
	void b()
	{
		printf("B.b std.string.toString(2)== %.*s\n",std.string.toString(2));
	}
}

void main()
{
	B b = new B;
	b.b();
}

#################

Ant

On Mon, 19 Jan 2004 22:08:38 -0500, Ant wrote:

ok forget the second test.
I guess it was to test something else.
I have a directory full of these simple tests...

Ant

On Mon, 19 Jan 2004 22:11:06 -0500, Ant wrote:

> On Mon, 19 Jan 2004 22:08:38 -0500, Ant wrote:
> 
> ok forget the second test.
> I guess it was to test something else.
> I have a directory full of these simple tests...
> 
> Ant

OK I went back to that second test.

If class B : A
still compiles which shows the point I was trying to make.

if class B : A (again) and remove the import std.string
from class B the it still compiles and ouputs
"B.b std.string.toString(2)== 2"

that shows that the "private" in "private import std.string" inside the body definition of A is not respected in B

Ant

Ant wrote:
> On Mon, 19 Jan 2004 22:11:06 -0500, Ant wrote:
> 
> 
>>On Mon, 19 Jan 2004 22:08:38 -0500, Ant wrote:
>>
>>ok forget the second test.
>>I guess it was to test something else.
>>I have a directory full of these simple tests...
>>
>>Ant
> 
> 
> OK I went back to that second test.
> 
> If class B : A
> still compiles which shows the point I was trying to make.
> 
> if class B : A (again) and remove the import std.string
> from class B the it still compiles and ouputs "B.b std.string.toString(2)== 2"
> 
> that shows that the "private" in "private import std.string"
> inside the body definition of A is not respected in B
> 
> Ant
> 

I also reported on this with minimal examples some time ago.

Disrespect of the 'private' keyword for modules seemed to be
the main problem.

Lars Ivar Igesund