Logical const

[Max Samukha]

On 11/29/2010 11:39 PM, Steven Schveighoffer wrote:
> On Mon, 29 Nov 2010 16:22:55 -0500, Max Samukha <spambox at d-coding.com>
> wrote:
>
>> On 11/29/2010 08:58 PM, Steven Schveighoffer wrote:
>>> Most likely not. How do you say that the 'draw' function switches the
>>> widget to a different parameterized type? With const, you can just slap
>>> a const on the end of the function.
>>>
>>> Here is some example of what I mean:
>>>
>>> class Widget
>>> {
>>> mutable Window location;
>>> int x, y, width, height;
>>> void draw() const { location.drawRectangle(x, y, width, height); }
>>> }
>>>
>>> where Window.drawRectangle is not a const function.
>>>
>>> -Steve
>>
>> I think we could try to devise a library solution. For example, the
>> problem could be solved with a templated struct + "alias this",
>> provided bugs in "alias this" are fixed. The following code works even
>> with the current implementation (dmd 2.050):
>>
>> struct Mutable(T)
>> {
>> T value;
>>
>> @property ref Unqual!T mutable() const
>> {
>> auto p = cast(Unqual!T*)&value;
>> return *p;
>> }
>> alias mutable this;
>>
>> ref opAssign(U)(auto ref U v) const
>> {
>> return mutable() = v;
>> }
>> }
>>
>> class Window
>> {
>> void drawRectangle(int x, int y, int width, int height)
>> {
>> writeln("Draw");
>> }
>> }
>>
>> class C
>> {
>> int x, y, width, height;
>> Mutable!Window location;
>>
>> this()
>> {
>> location = new Window;
>> }
>>
>> void draw() const { location.drawRectangle(x, y, width, height); }
>> }
>>
>> void main()
>> {
>> const C c = new C;
>> c.draw();
>> }
>>
>> (opAssign shouldn't be necessary. The compiler should properly forward
>> the assignment operation to the value returned by "mutable")
>
> I agree, a library solution would be ideal.
>
> Except the language says what results from your code (casting away const
> and then mutating) is undefined behavior. This means all bets are off,
> it can crash your program. I'm unsure how the compiler could take that
> route, but that's what's in the spec.
>
> Maybe because of the way const works, it never really is undefined, but
> there will always be that loophole.
>
> -Steve

Fair point.