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,FREEMAIL_FROM
	autolearn=unavailable autolearn_force=no version=3.4.4
X-Received: by 2002:a02:4dc9:: with SMTP id t70mr2654161jad.0.1547572927982;
        Tue, 15 Jan 2019 09:22:07 -0800 (PST)
X-Received: by 2002:a9d:2c22:: with SMTP id f31mr77673otb.4.1547572927839;
 Tue, 15 Jan 2019 09:22:07 -0800 (PST)
Path: 
 eternal-september.org!reader01.eternal-september.org!feeder.eternal-september.org!news.gegeweb.eu!gegeweb.org!usenet-fr.net!proxad.net!feeder1-2.proxad.net!209.85.166.216.MISMATCH!q69no79978itb.0!news-out.google.com!v71ni113ita.0!nntp.google.com!q69no79976itb.0!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail
Newsgroups: comp.lang.ada
Date: Tue, 15 Jan 2019 09:22:07 -0800 (PST)
In-Reply-To: <420cf9d0-fec9-4ab5-b5a8-ce8dced55a0d@googlegroups.com>
Complaints-To: groups-abuse@google.com
Injection-Info: glegroupsg2000goo.googlegroups.com; posting-host=146.5.2.29;
 posting-account=lJ3JNwoAAAAQfH3VV9vttJLkThaxtTfC
NNTP-Posting-Host: 146.5.2.29
References: <3fd9efd3-2ac5-4711-a1bc-fb75c57d7ffa@googlegroups.com>
 <2e4fb7e1-f649-43f6-95ec-61edf5a85732@googlegroups.com>
 <09955674-4095-410d-a688-1ed463ea6b2d@googlegroups.com>
 <420cf9d0-fec9-4ab5-b5a8-ce8dced55a0d@googlegroups.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <6c5eed8b-e736-42ac-9f43-983516c6db0c@googlegroups.com>
Subject: Re: New tool "gprexec", basically "make with project file"
From: Shark8 <onewingedshark@gmail.com>
Injection-Date: Tue, 15 Jan 2019 17:22:07 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Xref: reader01.eternal-september.org comp.lang.ada:55284
Date: 2019-01-15T09:22:07-08:00
List-Id: <comp.lang.ada>

On Tuesday, January 15, 2019 at 9:41:04 AM UTC-7, Shark8 wrote:
>=20
> Package PROJECT_NAME

Sorry, I accidentally submitted the form while composing my example... whic=
h is here:

Package PROJECT_NAME is

   Type Archetectures is ( x86, x86_64, ARM, SPARC, MIPS_V );
   Type Node_Type is (Storage, Processing);
   Type Partition_Type is (Active, Passive);
   Type Compilation_Parameters is record
     CPUs    : Natural :=3D 0;	-- Use as many cores as available.
	 Symbols : Boolean :=3D True;	-- Don't strip symbols.
	 Target  : Archetectures;
	 --...
   end record;
  =20
   Type Partition( Params : Compilation_Parameters; Style : Partition_Type =
) is record
     null; --... Other DSA paramters.
   end record;
  =20
   Type Node( Style : Node_type ) is record
     Archetecture : Archetectures;
     case Style is
	   when Storage =3D>    null; --...=20
	   when Processing =3D> null; --...
	 end case;
   end record;
  =20
   Generic
     Params : Compilation_Parameters;
   Procedure Compile;
  =20
   --- CONCEPTUAL GENERIC PACKAGE
   Generic
     Partitions : Array (Positive range <>) of not null access Partition;
   Package Compiler is
     Procedure Execute;
   End Compiler;
  =20
   --- CONCEPTUAL BODY FOR COMPILER.
   Package Body Compiler is
     Procedure Execute is
	 Begin
	   For P of Partitions loop
	     declare
		   Procedure Make is new Compile( P.Params );
		 begin
		   Make;
		 end;
	   End loop;
	 End Execute;
   End Compiler;
  =20
  =20
End PROJECT_NAME;

Now, obviously there would have to be standardization -- and it would proba=
bly work better if "Archetectures" were a parameter to PROJECT_NAME -- beca=
use if all config-packages were generic we could "nest" dependencies:

Generic
  Type STANDARD_PARAM is limited private; -- "Configuration standard param"
  with Package P1 is new Project_1( STANDARD_PARAM );
  with Package P2 is new Project_2( STANDARD_PARAM );
  -- P3 depends on P1&2
  with Package P3 is new Project_3( STANDARD_PARAM, P1, P2 );
Package Project_4 is
 -- ... STANDARD STRUCTURE.
End Project_4;

Now, all of that is operating with the idea of using Ada as a config-langua=
ge, which is doable, but perhaps a bit ugly... It might be a bit better to =
sit down, think about configurations (esp. in the presence of DSA) and deve=
lop an Ada-like language for that. (Perhaps in conjunction with a new Ada I=
R similar to DIANA, such that this configuration-description "compiles down=
 to" the proper generic-nodes which can then be interpreted by the compiler=
 as the configuration[s] to use; or processed by tools to inter-operate wit=
h current tools [ie IR =E2=86=92 (GPR_File, Gnatdist_Configuration_File) fo=
r GNAT].)