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,LOTS_OF_MONEY
	autolearn=ham autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,89d280850f5d8df
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 2002-07-21 18:17:50 PST
Path: archiver1.google.com!postnews1.google.com!not-for-mail
From: kcarron@belcan.com (Karen Carron)
Newsgroups: comp.lang.ada
Subject: Re: Passing Unconstrained Arrays from FORTRAN to an Ada95 subunit
Date: 21 Jul 2002 18:17:50 -0700
Organization: http://groups.google.com/
Message-ID: <5489a352.0207211717.cd9e66f@posting.google.com>
References: <5489a352.0207210544.3d50d423@posting.google.com>
 <sDC_8.81853$Wt3.54908@rwcrnsc53>
NNTP-Posting-Host: 205.133.202.67
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Trace: posting.google.com 1027300670 7345 127.0.0.1 (22 Jul 2002 01:17:50
 GMT)
X-Complaints-To: groups-abuse@google.com
NNTP-Posting-Date: 22 Jul 2002 01:17:50 GMT
Xref: archiver1.google.com comp.lang.ada:27289
Date: 2002-07-22T01:17:50+00:00
List-Id: <comp.lang.ada>

tmoran@acm.org wrote in message news:<sDC_8.81853$Wt3.54908@rwcrnsc53>...
> > > What happens if you pass a constrained array, ie,
> > >    type int_array_type is array(integer range 1.. 10) of integer;
> > Yes, this works, but the routine with the unconstrained array is called
> > by many other routines which pass it varying sizes of arrays.
>   One technique used in interfacing to C is to declare
>     type int_array_type is array(positive) of integer;
> That is, declare the array type large enough for any actual instance
> and then make sure in the Ada routine that you don't index beyond
> the actual data.  It used to be common practice in Fortran to pass
> the length of the array as a separate parameter, eg
>   procedure something(a,n)
>   integer n,a(n)
> which you could write in Ada as
>   procedure something(a : in out int_array_type;
>                       n : in positive);
> Does modern Fortran pass the length implicitly?  Apparently your Ada
> compiler doesn't think so.

Thanks for your suggestions.  I got it to work by constraining the
array to the subtype positive.  So type int_array_type is now:

type int_array_type (positive) of integer;

but it performed the same as before when I constrained int_array_type
to integer (!!!!), although it did get rid of the compiler warnings. 
I was thinking it worked because the range was now constrained, but
that appears to not be the case; not only must it be constrained, but
the lower bounds must be 1.