Thread overview
[dmd-internals] Module.semantic is not virtual
July 27
Hi all,
  Class Module contains the member functions semantic(), semantic2, and
semantic3() that are not virtual. They have the same name however as the
Dsymbol virtual methods. Because Module inherits from Dsymbol, it is
confusing to reuse the names. Clang warns about Module's functions hiding
the virtual methods of the parent class.

What was the intent here? Should Module's semantic functions override the base class's functions (but then why the *final* Package::semantic()?), or should they be separate function and just be renamed to something more clear?

thanks,
  Johan


July 30
No-one knows?

On Wed, Jul 27, 2016 at 8:50 PM, Johan Engelen <jbc.engelen@gmail.com> wrote:

> Hi all,
>   Class Module contains the member functions semantic(), semantic2, and
> semantic3() that are not virtual. They have the same name however as the
> Dsymbol virtual methods. Because Module inherits from Dsymbol, it is
> confusing to reuse the names. Clang warns about Module's functions hiding
> the virtual methods of the parent class.
>
> What was the intent here? Should Module's semantic functions override the base class's functions (but then why the *final* Package::semantic()?), or should they be separate function and just be renamed to something more clear?
>
> thanks,
>   Johan
>
>


July 31
They're just separate functions that happen to use the same name. They're never going to be called through a reference typed as Dsymbol (I hope).

On Sun, Jul 31, 2016 at 4:26 AM, Johan Engelen via dmd-internals <dmd-internals@puremagic.com> wrote:
> No-one knows?
>
> On Wed, Jul 27, 2016 at 8:50 PM, Johan Engelen <jbc.engelen@gmail.com> wrote:
>>
>> Hi all,
>>   Class Module contains the member functions semantic(), semantic2, and
>> semantic3() that are not virtual. They have the same name however as the
>> Dsymbol virtual methods. Because Module inherits from Dsymbol, it is
>> confusing to reuse the names. Clang warns about Module's functions hiding
>> the virtual methods of the parent class.
>>
>> What was the intent here? Should Module's semantic functions override the base class's functions (but then why the *final* Package::semantic()?), or should they be separate function and just be renamed to something more clear?
>>
>> thanks,
>>   Johan
>>
>
>
> _______________________________________________
> dmd-internals mailing list
> dmd-internals@puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-internals
_______________________________________________
dmd-internals mailing list
dmd-internals@puremagic.com
http://lists.puremagic.com/mailman/listinfo/dmd-internals
July 31
On 31 Jul 2016, at 5:18, Daniel Murphy via dmd-internals wrote:
> They're just separate functions that happen to use the same name. They're never going to be called through a reference typed as Dsymbol (I hope).

Yep. We should just rename Module.semantic.

 — David
_______________________________________________
dmd-internals mailing list
dmd-internals@puremagic.com
http://lists.puremagic.com/mailman/listinfo/dmd-internals