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=unavailable autolearn_force=no version=3.4.4
Path: 
 eternal-september.org!reader01.eternal-september.org!reader02.eternal-september.org!news.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Natasha Kerensikova <lithiumcat@instinctive.eu>
Newsgroups: comp.lang.ada
Subject: Re: Can anyone help with GNAT.Perfect_Hash_Generators ? (Possible
 memory corruption)
Date: Fri, 9 Sep 2016 06:04:30 -0000 (UTC)
Organization: A noiseless patient Spider
Message-ID: <slrnnt4k7g.1cq.lithiumcat@nat.rebma.instinctive.eu>
References: <slrnnsot2o.1cq.lithiumcat@nat.rebma.instinctive.eu>
 <397dd8cb-2afc-43cd-972d-3b1a5a90cd5d@googlegroups.com>
 <slrnnsu608.1cq.lithiumcat@nat.rebma.instinctive.eu>
 <877fao24da.fsf@mid.deneb.enyo.de> <lyoa40rb84.fsf@pushface.org>
 <0b79fce9-b898-4a1c-91fb-ca41e87b4dd5@googlegroups.com>
 <lyr38us4qk.fsf@pushface.org>
 <08e0581a-30f1-4cc5-972c-2d6c539d9df6@googlegroups.com>
 <nqscvn$j2h$1@dont-email.me>
Injection-Date: Fri, 9 Sep 2016 06:04:30 -0000 (UTC)
Injection-Info: mx02.eternal-september.org;
 posting-host="eab84d932a0f4c9d4606240766f0f5e7";
	logging-data="31304"; mail-complaints-to="abuse@eternal-september.org";
	posting-account="U2FsdGVkX189+qQkk7zU3aAjo9V93UkN"
User-Agent: slrn/1.0.2 (FreeBSD)
Cancel-Lock: sha1:ss8c9oEW2cthYGkp82gvfElpEo4=
Xref: news.eternal-september.org comp.lang.ada:31735
Date: 2016-09-09T06:04:30+00:00
List-Id: <comp.lang.ada>

Hello,

On 2016-09-08, Jeffrey R. Carter <spam.jrcarter.not@spam.not.acm.org> wrote:
> On 09/08/2016 11:03 AM, Anh Vo wrote:
>> On Thursday, September 8, 2016 at 10:04:06 AM UTC-7, Simon Wright wrote:
>>> Anh Vo <anhvofrcaus@gmail.com> writes:
>>>
>>>> Then, what would be the proper fix?
>>>
>>> To what? If you mean the apparent bug in GNAT.Perfect_Hash_Generators
>> 
>> Yes.
>
> Clearly it needs to have the assumption, that all Strings have a lower bound of
> 1, eliminated.
>

I thought it would be easier to make the assumption internally true
instead.

Since the first thing it does with the input is to copy it internally
using the function New_Word, I would propose a fix along the line of
replacing it from

   function New_Word (S : String) return Word_Type is
   begin
      return new String'(S);
   end New_Word;

to

   function New_Word (S : String) return Word_Type is
      Result : Word_Type := new String (1 .. S'Length);
   begin
      Result.all := S;
      return Result;
   end New_Word;


That way there is no need to gather a deep understanding of the
algorithm to hunt all the places that rely on 'First = 1.
And I don't see any loss of performance or readability with this change,
in case anyone is worried about it.

However I haven't found the time yet to seriously test such a fix.


Natasha