From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,45a9122ddf5fcf5
X-Google-Attributes: gid103376,public
From: mazzanti@iei.pi.cnr.it (Franco Mazzanti)
Subject: Re: Rules for Representation of Subtypes
Date: 1996/10/03
Message-ID: <mazzanti-0310961743320001@131.114.200.115>#1/1
X-Deja-AN: 186957238
organization: IEI-CNR
newsgroups: comp.lang.ada
Date: 1996-10-03T00:00:00+00:00
List-Id: <comp.lang.ada>


Robert A Duff wrote:

> >generic
> >   type Source(<>) is limited private;
> >   type Target is (<>);
> >function Checked_Scalar_Conversion (S:Source) return Target; 
> >
> >
> >with Unchecked_Conversion;
> >function Checked_Scalar_Conversion (S : Source) return Target is
> >   type My_Rec is record
> >      Scalar : Target;
> 
> That's illegal, since Target has unknown discrims.  But other than that,
> this approach will work.
> 
> But, you don't really need to go to all that trouble.  If you just make
> sure your integer types match the hardware (like -2**31..2**31-1 or
> whatever), then unchecked conv of integers will tend to work just fine.
>  ...
>  ...
> 
> - Bob

illegal? isn't Target required to be just a discrete type [RM 12.5.2(2)]?
If I am wrong, is this a GNAT bug (since it compiles without problem)?

Notice that all this trouble might be needed if I want my program to be
as far as possible implementation-independent, or if I want to look for
same mechanically verifiable (as far as possible implementation-independent)
safe coding guidelines (without going to disallow all unchecked conversions).

Franco