proposed @noreturn attribute

[Nick Sabalausky (Abscissa) via Digitalmars-d]

On 07/08/2017 08:17 AM, Andrei Alexandrescu wrote:
> On 7/8/17 6:15 AM, Walter Bright wrote:
>> Has anyone a better idea? Does anyone want to write a DIP for this?
> 
> An attribute is fine. A more PL-minded possibility is to return a 
> specific type:
> 
> struct None
> {
>      @disable this();
>      @disable this(this);
>      @disable @property None init();
> }
> 
> None ThisFunctionExits();
> 
> The compiler detects (without having anything hardwired about the 
> particular type "None") that the type None is impossible to create and 
> copy/move from a function, and therefore decrees the function will never 
> return.

Pro:
- Having the indication "this doesn't return" syntactically be the 
return type makes a lot of sense. I like that a lot. (Though I'd have no 
objection to Walter's original suggestion either).

- It's one less built-in @attribute to scare people away with objections 
of "too many attributes"! (And admittedly, idiomatic D can end up with 
lots of attribtes all over the place.)

Con:
- Inferring that a type is meant to indicate "functions 'returning' this 
do not return", strikes me as very round-about, hackish, unintuitive and 
just unclean.

I really think this is one case where a dedicated compiler-understood 
symbol is not only fully justified but also much cleaner.

That would be my #1 top preference: Your suggestion here, but just make 
"NoReturn" (or whatever it'll be called) a special compiler-understood 
"type", not something inferred from the detailed of a library-provided type.

In any case, I'm definitely onboard with the idea of having SOME 
standard way to tell the compiler "this does not return". That would be 
nice to have.