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-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,13703514e3723cbe
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 2001-08-13 00:34:06 PST
Path: 
 archiver1.google.com!newsfeed.google.com!newsfeed.stanford.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.direct.ca!look.ca!newshub2.rdc1.sfba.home.com!news.home.com!news1.rdc1.sfba.home.com.POSTED!not-for-mail
From: tmoran@acm.org
Newsgroups: comp.lang.ada
Subject: Re: C-style 'union' in Ada?
References: <9l7qon$d78$1@nntp9.atl.mindspring.net>
X-Newsreader: Tom's custom newsreader
Message-ID: <OpLd7.59206$Kd7.33701592@news1.rdc1.sfba.home.com>
Date: Mon, 13 Aug 2001 07:34:06 GMT
NNTP-Posting-Host: 24.7.82.199
X-Complaints-To: abuse@home.net
X-Trace: news1.rdc1.sfba.home.com 997688046 24.7.82.199 (Mon,
 13 Aug 2001 00:34:06 PDT)
NNTP-Posting-Date: Mon, 13 Aug 2001 00:34:06 PDT
Organization: Excite@Home - The Leader in Broadband http://home.com/faster
Xref: archiver1.google.com comp.lang.ada:11832
Date: 2001-08-13T07:34:06+00:00
List-Id: <comp.lang.ada>

>record that I want to hold a pointer to a buffer which is either a String
>or Wide_String, ...
>...  In the package I'm designing, I don't care whether it is a
>String or Wide_String, I will just pass it to the file I/O packages
>(stream, in this case).  In C, I would:
  How *do* you tell which it is?  Or do you really truly neither know
nor care?  If so, then the pointer itself is the abstraction
level - it's a black box of bits.  It could just as well be an
integer, or an array of bytes, whatever.
  If you do have a way to tell, can you incorporate that as a
discriminant in a variant record?  eg.

  type string_desc(Is_Wide : Boolean) is record
    length : Interfaces.C.Unsigned_Long;
    case Is_Wide is
      when True =>
        ansi_string : String_Desc;
      when False =>
        wide_string : Wide_String_Desc;
    end case;
  end record;

of course
struct
    {
    ...
    string_desc    file_name;
    ...
    } some_struct;
will in turn need a discriminant to pass along:
  type some_struct(Uses_Wide_String : Boolean) is record
    ...
    file_name : string_desc(Is_Wide => Uses_Wide_String);
    ...
  end record;

It's possible that tagged types would be appropriate.  An abstract
parent type with primitive procedures declared for those operations
that treat an object of the type as a black box, and abstract
primitives for those operations that *must* be declared separately
depending on the child type being ansi_string or wide_string.