I am thinking about opening a bug with the following code:

struct S
{
    @disable this();

    static void opCall()
    {}
}

void main()
{}

Error: struct deneme.S static opCall is hidden by constructors and can never be called

Which seems to be due to the following change:


https://github.com/D-Programming-Language/dmd/commit/79ae211e71cf0937523010e39f7f0981e9550904

What do you think?

Ali

On Friday, 30 January 2015 at 22:41:35 UTC, Ali Çehreli wrote:
> I am thinking about opening a bug with the following code:
>
> struct S
> {
>     @disable this();
>
>     static void opCall()
>     {}
> }
>
> void main()
> {}
>
> Error: struct deneme.S static opCall is hidden by constructors and can never be called
>
> Which seems to be due to the following change:
>
>
> https://github.com/D-Programming-Language/dmd/commit/79ae211e71cf0937523010e39f7f0981e9550904
>
> What do you think?
>
> Ali

It should only be an error when static opCall() cant be distinguishable from this.

---
struct S
{
    @disable this();
    static string opCall(){return "yo mister White";}
}
void main()
{}
---

is distinguishable (by return type) but cant be compiled.
You're right there's a problem.

On 01/30/2015 03:19 PM, BBaz wrote:

> It should only be an error when static opCall() cant be distinguishable
> from this.
>
> ---
> struct S
> {
>      @disable this();
>      static string opCall(){return "yo mister White";}
> }
> void main()
> {}
> ---
>
> is distinguishable (by return type) but cant be compiled.

I agree that this is a problem but return types are not parts of function signatures; so return types do not help "distinguish" functions. Besides, constructors don't have return types; so it is a little bit of a stretch to compare them to functions. :)

> You're right there's a problem.

Thank you. Filed:

  https://issues.dlang.org/show_bug.cgi?id=14087

Ali

> "distinguish"

Yes, I know this a strange word. But it seems to be a valid one:
http://www.collinsdictionary.com/dictionary/english/distinguish?showCookiePolicy=true

"distinguishable" is ok as well.

On 01/30/2015 04:30 PM, BBaz wrote:
>> "distinguish"
>
> Yes, I know this a strange word. But it seems to be a valid one:
> http://www.collinsdictionary.com/dictionary/english/distinguish?showCookiePolicy=true
>
>
> "distinguishable" is ok as well.
>

Sorry, I did not mean to emphasize "distinguish" over "distinguishable". I was quoting you as although I understood what you said, I am not used to hearing that word used in function matching. However, "match" does appear in the spec: :)

  http://dlang.org/function.html#function-overloading

Ali