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=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,f4e1ae22e5bbdd9f
X-Google-Attributes: gid103376,public
From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: Generic in out object parameters
Date: 1997/04/19
Message-ID: <dewar.861457164@merv>#1/1
X-Deja-AN: 235962191
References: <Forum.861207636.14373.petera@woodlands>
 <33561FCC.C38@gsfc.nasa.gov>
Organization: New York University
Newsgroups: comp.lang.ada
Date: 1997-04-19T00:00:00+00:00
List-Id: <comp.lang.ada>


Stephen says

<<   generic
      Default_Expecting : in STRING;
      Next_Arg : in out NATURAL; -- index of next argument to read
   procedure Get_String
        (Item : out Ada.Strings.Unbounded.Unbounded_String;
         Expecting : in String := Default_Expecting);
>>

As an example of using in out parameters in generics.

Basically what is going on here is that the generic mechanism is being
used for partial parametrization. This is particularly convenient in
the case of in out parameters, since it is not possible to provide
defaults for in out parameters. Essentially the above generic defition
and an associated instantiation of it provides default in out parameters.

But i wonder if it is convincing. At least in this case? It seems
extremely unlikely that one would instantiation Get_String more than
once in a program, so why not just make it a package in which the
index is a local variable. Well I suppose you could say something
about task safety .. still ...