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,f34f2cf6eacd2bc8
X-Google-NewGroupId: yes
X-Google-Attributes: gida07f3367d7,domainid0,public,usenet
X-Google-Language: ENGLISH,ASCII-7-bit
Received: by 10.68.30.202 with SMTP id u10mr3573276pbh.1.1319296608008;
        Sat, 22 Oct 2011 08:16:48 -0700 (PDT)
Path: 
 d5ni43758pbc.0!nntp.google.com!news1.google.com!newsfeed2.dallas1.level3.net!news.level3.com!bloom-beacon.mit.edu!newsswitch.lcs.mit.edu!nntp.TheWorld.com!not-for-mail
From: Robert A Duff <bobduff@shell01.TheWorld.com>
Newsgroups: comp.lang.ada
Subject: Re: Private constant of tagged limited controlled type
Date: Sat, 22 Oct 2011 11:16:47 -0400
Organization: The World Public Access UNIX, Brookline, MA
Message-ID: <wcclisd9huo.fsf@shell01.TheWorld.com>
References: <da20b9c1-54ff-4f1b-802b-9ddd5b2a84ff@u9g2000vby.googlegroups.com>
 <j7trtm$q5t$1@tornado.tornevall.net>
NNTP-Posting-Host: shell01.theworld.com
Mime-Version: 1.0
X-Trace: pcls6.std.com 1319296607 12752 192.74.137.71 (22 Oct 2011 15:16:47
 GMT)
X-Complaints-To: abuse@TheWorld.com
NNTP-Posting-Date: Sat, 22 Oct 2011 15:16:47 +0000 (UTC)
User-Agent: Gnus/5.1008 (Gnus v5.10.8) Emacs/21.3 (irix)
Cancel-Lock: sha1:jNlgV10gEp9nBxgZZxhMBXqGxQY=
Xref: news1.google.com comp.lang.ada:18669
Content-Type: text/plain; charset=us-ascii
Date: 2011-10-22T11:16:47-04:00
List-Id: <comp.lang.ada>

Jeffrey Carter <spam.jrcarter.not@spam.not.acm.org> writes:

> On 10/21/2011 11:14 PM, Malaise wrote:
>> The following complains that "initialization of limited object
>> requires aggregate or function call":
>
> Right.
>
>> with Ada.Finalization;
>> package Pack is
>>    type T is tagged limited private;
>>    C : constant T;
>> private
>>    type T is new Ada.Finalization.Limited_Controlled with record
>>      I : Integer := 0;
>>    end record;
>>    V : T;
>>    C : constant T := V;

That's illegal, because you're trying to copy a limited object.
What if T had a self-referential component?
If the above were allowed, C.Self would incorrectly point
to V, not C.  What if T contained a protected object?
It makes no sense to copy locks.

>> end Pack;
>> Aggregate seems impossible because Limited_Controlled is abstract and
>> anyway T is very complex in fact.
>
> I suggest you read ARM 4.3.1 and 4.3.2. You can write an aggregate of the form
>
> (Ada.Finalization.Limited_Controlled with others => <>)

Right, or:

    C : constant T := (Ada.Finalization.Limited_Controlled with I => 123)

>> A function (returning tagged type) has to be visible and anyway cannot
>> return V (in Ada 2005).
>
> This is an unusual claim. The extended return statement was introduced
> specifically for returning limited types from a function. You ought to
> read ARM 6.5. You could write
>
> return X : T;

Right.  Or it could say

    return (Ada.Finalization.Limited_Controlled with others => <>);

But the comment, "cannot return V" is correct -- no matter whether
V is local or global.

> to return a value of T with all components set to their default values.

I also suggest reading this:

http://www.adacore.com/2007/09/24/ada-gem-10/

and also the following 2 gems, which I wrote to explain
how limited types work in Ada 2005.

- Bob