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: 103376,5556ca7de188d5ef,start
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 2001-12-11 14:22:36 PST
Path: 
 archiver1.google.com!news1.google.com!newsfeed.stanford.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-FFM2.ecrc.net!news.iks-jena.de!lutz
From: lutz@iks-jena.de (Lutz Donnerhacke)
Newsgroups: comp.lang.ada
Subject: Multidimensional array vs. array of array
Date: Tue, 11 Dec 2001 22:22:00 +0000 (UTC)
Organization: IKS GmbH Jena
Message-ID: <slrna1d1o8.7nl.lutz@belenus.iks-jena.de>
NNTP-Posting-Host: belenus.iks-jena.de
X-Trace: branwen.iks-jena.de 1008109320 28877 217.17.192.34 (11 Dec 2001
 22:22:00 GMT)
X-Complaints-To: usenet@iks-jena.de
NNTP-Posting-Date: Tue, 11 Dec 2001 22:22:00 +0000 (UTC)
User-Agent: slrn/0.9.6.3 (Linux)
Xref: archiver1.google.com comp.lang.ada:17787
Date: 2001-12-11T22:22:00+00:00
List-Id: <comp.lang.ada>

How does a skilled Ada programmer define a return type containing a
multidimensional array? Of course, the simple solution works:

  type Equations is array (Rows range <>, Columns range <>) of Number;
  type Workspace (row : Rows; col : Columns) is record
    eqs : Equations (Rows'First .. row, Columns'First .. col);
    ... some more stuff ...
  end record;
  function Init_Workspace return Workspace;

But this approach fails miserably on extracting a single row from the result
as well as on construction the result in a recursive function.

I do not found a way to define this type as an array of arrays despite
writing a generic package over the maximum values:
  generic
     type Columns is range <>;
  package
     type Equation  is arry (Columns) of Number;
     type Equations is array (Rows range <>) of Equation;
     type Workspace (row : Rows) is record
        eqs : Equations (Rows'First .. row);
        ... some more stuff ...
     end record;
     function Init_Workspace return Workspace;
  end package;

But this approach does not work for a situation where determining the
resulting constraints is equivalent to simulating the Init_Workspace
recursion. This should be avoided: Error-prone, duplicated and slightly
modified code.

Any idea?