On 2013-09-19, 10:06, Kagamin wrote:
On Wednesday, 11 September 2013 at 20:20:13 UTC, Simen Kjaeraas wrote:
On 2013-09-11, 20:29, Andrej Mitrovic wrote:
On 9/11/13, Kagamin <[email protected]> wrote:
I'd say, strong handles shouldn't act as pointers (and shouldn't
contain pointers), so null shouldn't work.
"NULL" is already used in a ton of WinAPI C/C++ code and MSDN
documentation, it would be a major pain in the ass to have to use e.g.
HWND(0) instead.
How's about
enum NULL = HANDLE(0);
Honestly I hate that, and am convinced the correct way to fix this is
for D to add opImplicitCastFrom.
Or multiple alias this:
struct _NULL
{
HANDLE zero;
void* ptr;
alias zero this;
alias ptr this;
}
If you want opaque handles, that does not solve the problem. What would
solve the problem is something like this:
struct DECLARE_HANDLE(int line = __LINE__, string file = __FILE__T) {
private void* payload;
private this(void* value) {
payload = value;
}
ref DECLARE_HANDLE opAssign(typeof(null) value) {
payload = null;
return this;
}
DECLARE_HANDLE opImplicitCastFrom(typeof(null) value) {
return DECLARE_HANDLE(null);
}
}
alias DECLARE_HANDLE!(__LINE__, __FILE__) HANDLE; // BUG11074
alias DECLARE_HANDLE!(__LINE__, __FILE__) HWND; // BUG11074
void test() {
void* voidPointer;
HANDLE handle;
HWND window;
SomeWindowsFunctionTakingAHandle(null); // No problem!
SomeWindowsFunctionTakingAHandle(handle); // No problem!
SomeWindowsFunctionTakingAHandle(voidPointer); // Does not compile.
SomeWindowsFunctionTakingAHandle(window); // Does not compile.
}
--
Simen
