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=-1.9 required=5.0 tests=BAYES_00 autolearn=ham
	autolearn_force=no version=3.4.4
X-Google-Thread: 103376,34a625e98d6a8792
X-Google-Attributes: gid103376,public
X-Google-Language: ENGLISH,ASCII-7-bit
Path: 
 g2news1.google.com!news2.google.com!fu-berlin.de!uni-berlin.de!not-for-mail
From: "Nick Roberts" <nick.roberts@acm.org>
Newsgroups: comp.lang.ada
Subject: Re: The right way to handle this difference in Ada
Date: Thu, 22 Jul 2004 13:07:33 +0100
Message-ID: <opsbjfyv06p4pfvb@bram-2>
References: <cdo4ic$q8f$1@e3k.asi.ansaldo.it>
Mime-Version: 1.0
Content-Type: text/plain; format=flowed; delsp=yes; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
X-Trace: news.uni-berlin.de VLkZAWrGU6/ui1mzl+hdkQ7lw+ntGT59UTyjBfSo3a9J3TSsE=
User-Agent: Opera M2/7.51 (Win32, build 3798)
Xref: g2news1.google.com comp.lang.ada:2343
Date: 2004-07-22T13:07:33+01:00
List-Id: <comp.lang.ada>

On Thu, 22 Jul 2004 12:23:09 +0200, vic <ppp@ppp.it> wrote:

> Please consider this type:
>
> type unsigned_int_12_type is range 0..(2*12)-1;
>
> and these 2 variables of the above type:
>
> a: unsigned_int_12_type;
> b: unsigned_int_12_type;
>
> If the code must perform a difference between a and b, say:
>
> c := a-b
>
> 1) which should be the type of c? I think it should be
> something like:
>
> type signed_int_12_type is range -unsigned_int_12'last ..
> unsigned_int_12'last.
>
> Is this right?

I'm guessing that what you really require is to be able to
pack one or more 12-bit integers into a record. Is this
correct?

If it is, the solution to your problem is likely to be this:

    subtype My_Integer is Integer range 0..2**12-1;

    A, B: My_Integer;

    C: Integer range -My_Integer'Last..My_Integer'Last;

    ...

    C := A-B;

    ...

    type My_Record is
       record
          ...
          X, Y: My_Integer;
          ...
       end record;

    for My_Record use
       record
          ...
          X     at 5 range 0..11;
          Y     at 5 range 12..23;
          ...
       end record;

The idea is that you use a standard base type for the
arithmetic, declaring subtypes which restrict the range
as appropriate, and then tell the compiler explicitly
how to squeeze values into 12-bit parts of the record
type. With luck, the compiler will oblige (or tell you
why it cannot).

Des this help?

> 2) And I should rename a "-" operation which gets two signed_int_12_type  
> as operands and returns
> signed_int_12_type as result?

Hopefully the above technique will work for you, and
enable to avoid the need to declare your own arithmetic
operations.

-- 
Nick Roberts