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-Thread: 103376,64b29dfa2220a59f
X-Google-Attributes: gid103376,public,usenet
X-Google-Language: ENGLISH,UTF8
Path: 
 g2news2.google.com!news3.google.com!fu-berlin.de!newsfeed.arcor.de!newsspool2.arcor-online.net!news.arcor.de.POSTED!not-for-mail
Newsgroups: comp.lang.ada
Subject: Re: Reserve_Capacity for Unbounded_String?
From: Georg Bauhaus <rm.tsoh+bauhaus@maps.futureapps.de>
In-Reply-To: <46A70064.8090407@obry.net>
References: <1185134043.892012.217560@n2g2000hse.googlegroups.com>
	 <1185203238.701948.307410@m37g2000prh.googlegroups.com>
	 <f83ish$p8o$1@jacob-sparre.dk> <46A5B0FE.3060008@obry.net>
	 <f85i0u$jsl$1@jacob-sparre.dk> <46A69136.3010803@obry.net>
	 <f868qn$qac$1@jacob-sparre.dk>  <46A70064.8090407@obry.net>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Message-ID: <1185357317.24786.39.camel@kartoffel>
Mime-Version: 1.0
X-Mailer: Evolution 2.10.1 
Date: Wed, 25 Jul 2007 11:55:17 +0200
Organization: Arcor
NNTP-Posting-Date: 25 Jul 2007 11:54:50 CEST
NNTP-Posting-Host: d5eb704f.newsspool1.arcor-online.net
X-Trace: 
 DXC=3^g0hmj<_GT=>bdbdS?M0Yic==]BZ:af^4Fo<]lROoRQFl8W>\BH3YRC7L==5KlGV\N[W<lHX`=BT\6BhD5G6\i\hPgoa:V1n^]
X-Complaints-To: usenet-abuse@arcor.de
Xref: g2news2.google.com comp.lang.ada:1163
Date: 2007-07-25T11:54:50+02:00
List-Id: <comp.lang.ada>

On Wed, 2007-07-25 at 09:48 +0200, Pascal Obry wrote:
> Randy Brukardt a =C3=A9crit :
> > And now I've wasted far too much time on this worthless topic.
>=20
> If you think so! GNAT has implemented that as in a real application the
> code was running more than 10 times faster. If you think you have wasted
> far too much time, no problem. I'm not a Janus/Ada customer anyway!

In fact, there are many special string-related subprograms
in GNAT that arguably demonstrate the influences of effective
marketing on technical properties of compilers. E.g. there are
variations of

   function Str_Concat_3 (S1, S2, S3 : String) return String;
   --  Concatenate three strings and return resulting string

hidden in System.String_Ops_Concat_3 and used in algorithmically (!)
optimizing two "&"s in the compiler.

Likewise,  "... use of & removed for efficiency reasons"
in package body Ada.Strings.Fixed. (See also
Ada.Strings.Unbounded.Aux.Get_String for the unbounded case.)

Two conclusion seem to follow from this:

1) that GNAT authors agree with Randy when he says that you
have to be careful with "&"-centric algorithms, but

2) that AdaCore has better salespeople because customers are
more satisfied when the vendor has already done the "right" thing
for them. ;-)


Try this:

procedure Strtest is
   S: constant String :=3D "A" & Argument(1) & "B";
     -- occurences of "&" --> GNAT special case
begin
   Text_IO.put_line(S);
end Strtest;


Then, using GNAT, you get

$ nm strtest.o
...
00000019 T _ada_strtest
...
         U system__string_ops_concat_3__str_concat_3  <--
$