From: marciant@earthlink.net
Subject: Re: A function that cannot be called?
Date: Fri, 19 Oct 2018 11:22:51 -0700 (PDT)
Date: 2018-10-19T11:22:51-07:00 [thread overview]
Message-ID: <f58ce6ac-1737-4d5a-8d4e-14f7d2940711@googlegroups.com> (raw)
In-Reply-To: <0a680981-7b14-4b72-b51a-b06ab7822d54@googlegroups.com>
On Friday, October 19, 2018 at 2:19:07 PM UTC-4, marc...@earthlink.net wrote:
> On Thursday, October 18, 2018 at 8:14:32 AM UTC-4, G.B. wrote:
> > There is a type declaration, I rememberd, in Ada, that
> > does not allow any object of the type to be declared.
> > I think the type should be along the lines below. Is
> > this correct?
> >
> > Then, a function that nominally returns objects of
> > this type should be impossible to call. Is this right,
> > too?
> >
> > generic
> > type T(<>) is limited private; -- gen. arg. type
> >
> > package What is
> >
> > type Void (<>) is private;
> >
> > function Impossible (X : T) return Void;
> >
> > private
> >
> > package Inner is
> > type Void (<>) is private;
> > private
> > type Void is record
> > null;
> > end record;
> > end Inner;
> >
> > type Void is new Inner.Void;
> >
> > end What;
> >
> > (If current Ada could still pass by reference in both call
> > direction and return direction, then I'd have added *limited*
> > to the Voids' declarations. Without it, a body of Impossible
> > is closer to becoming possible by using Convention => Ada
> > together with an 'Address clause on a Result object. I think.)
>
> Before Ada 95 or Ada 2005 it would have been a package
> spec that had the following (with P1,P2,etc. being valid types):
>
> type Not_Objects_Just_for_Parameters is limited private;
> function Return_A_Corresponding_Value(p1:t1;p2:t2;etc...)
> return Not_Objects_Just_for_Parameters;
> procedure Force_Call_Of_Function_1
> (Function_Call:Not_Objects_Just_for_Parameters;
> Other_Param3:p3;etc...);
> procedure Force_Call_Of_Function_2
> (Function_Call:Not_Objects_Just_for_Parameters;
> Other_Param4:p4;etc...);
>
> I miss that way of forcing/ensuring new execution of some
> code as a prerequisite of use of a subprogram. ;-(
> Being able to hang on to the result of such a function
> by using renames wrecks this usage.
Oh, I think maybe (<>) had to be used also on the type definition
just like you had.
next prev parent reply other threads:[~2018-10-19 18:22 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-18 12:14 A function that cannot be called? G.B.
2018-10-18 15:33 ` Stefan.Lucks
2018-10-18 20:21 ` G.B.
2018-10-18 20:57 ` Niklas Holsti
2018-10-19 7:15 ` Dmitry A. Kazakov
2018-10-19 13:55 ` G.B.
2018-10-19 15:31 ` Dmitry A. Kazakov
2018-10-18 17:03 ` AdaMagica
2018-10-18 19:36 ` G.B.
2018-10-18 21:30 ` Randy Brukardt
2018-10-19 14:00 ` G.B.
2018-10-19 15:39 ` Dmitry A. Kazakov
2018-10-20 1:34 ` Randy Brukardt
2018-10-20 9:14 ` G.B.
2018-10-20 11:13 ` Simon Wright
2018-10-20 14:11 ` Dmitry A. Kazakov
2018-10-21 9:25 ` G.B.
2018-10-21 9:07 ` G.B.
2018-10-21 9:51 ` Dmitry A. Kazakov
2018-10-21 10:57 ` Niklas Holsti
2018-10-21 18:00 ` Simon Wright
2018-10-19 8:48 ` AdaMagica
2018-10-19 11:15 ` G.B.
2018-10-19 17:06 ` AdaMagica
2018-10-19 19:57 ` G.B.
2018-10-19 22:06 ` Jere
2018-10-21 10:14 ` G.B.
2018-10-21 11:30 ` Egil H H
2018-10-23 11:45 ` G.B.
2018-10-23 14:35 ` Jere
2018-10-23 14:57 ` Dmitry A. Kazakov
2018-10-23 17:49 ` G.B.
2018-10-23 19:25 ` Dmitry A. Kazakov
2018-10-24 7:35 ` G.B.
2018-10-24 8:14 ` Dmitry A. Kazakov
2018-10-19 18:19 ` marciant
2018-10-19 18:22 ` marciant [this message]
2018-10-20 1:36 ` Randy Brukardt
2018-10-20 2:54 ` marciant
2018-10-19 20:25 ` Shark8
2018-10-19 23:28 ` marciant
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox