Path: ...!weretis.net!feeder8.news.weretis.net!pasdenom.info!.POSTED.2a01:e0a:21:ea80:16a1:7d26:987e:eea2!not-for-mail
From: yamo' <yamo@beurdin.invalid>
Newsgroups: fr.comp.usenet.serveurs
Subject: Re: Champ cancel-lock multiligne
Date: Thu, 19 May 2022 13:16:09 +0200
Organization: <http://pasdenom.info/news.html>
Message-ID: <t658tq$8tj$1@rasp.pasdenom.info>
References: <t5orgk$oqk$1@dont-email.me> <t62gbk$hka$1@rasp.pasdenom.info>
 <t62hif$jce$1@shakotay.alphanet.ch> <t62hua$lf6$1@shakotay.alphanet.ch>
 <t62jeo$r8a$1@rasp.pasdenom.info> <t62o37$ead$1@shakotay.alphanet.ch>
 <t64l64$p4t$1@rasp.pasdenom.info> <t64seh$trg$1@shakotay.alphanet.ch>
 <t652iq$p67$1@rasp.pasdenom.info> <t65508$ur6$1@rasp.pasdenom.info>
 <t6562v$95b$1@shakotay.alphanet.ch> <t6581c$4ad$3@rasp.pasdenom.info>
 <t65869$nvm$1@dont-email.me> <t6589q$4ad$5@rasp.pasdenom.info>
 <t658m4$lci$1@shakotay.alphanet.ch>
Reply-To: yamo@groumpf.org
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Thu, 19 May 2022 11:16:10 -0000 (UTC)
Injection-Info: rasp.pasdenom.info; posting-account="stephane"; posting-host="2a01:e0a:21:ea80:16a1:7d26:987e:eea2";
	logging-data="9139"; mail-complaints-to="abuse@pasdenom.info"
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Firefox/68.0 SeaMonkey/2.53.12
Cancel-Lock: sha256:uHDukqn6IL35eOZTeRjgwnxgt1hLNcrjGqGlxZOZjzg=
In-Reply-To: <t658m4$lci$1@shakotay.alphanet.ch>
X-Seamonkey:  <https://www.seamonkey-project.org/>
Bytes: 3766
Lines: 74

Marc SCHAEFER a tapoté le 19/05/2022 13:12:
> yamo' <yamo@beurdin.invalid> wrote:
>> Tu peux.
> 
> Le diff ressemble à quoi?
> 
> J'avais fait une erreur de nom de variable dans une des premières
> versions du patch, et comme le script n'est pas en use strict; use
> warnings; c'était invisible.
> 
> Depuis, avec la version:
> 
>    https://nnrp.alphanet.ch/newsgroup/fr.comp.usenet.serveurs/%3Ct5u0k3$5op$1@shakotay.alphanet.ch%3E
> 
> Le bug était d'utiliser une variable inexistante $art plutôt que
> $headers dans le for ou le foreach.
> 
> Pas de souci me semble-t-il désormais.
> 


J'ai :

sub verify_cancel($$$) {
   my $r_hdr = shift || die;
   my $target = shift;
   my $descr = shift;

   my $headers = INN::head($target);
   if (!$headers){
#          saveart('log.cancel',"$descr of non-existing ID $target");
#          return undef;
           return "$descr of non-existing ID";
   }
   my %headers;
#   for my $line(split(/\s*\n/, $headers))    {
   my $previous;
   foreach my $line (split(/\s*\n/, $headers)) {
      if ($line =~ m/^([[:alnum:]-]+):\s+(.*)/) {
                $previous = $1;
      } elsif ($line =~ m/^\s+(.*)/) {
          if (defined($previous)) {
             $headers{$previous} .= " " . $1;
          }
          else {
             INN::syslog('notice', 'garbled line: ' . $line);
          }
      }
   }
   my $lock = $headers{'Cancel-Lock'};
   if (defined($lock)) {
      my $key = $r_hdr->{'Cancel-Key'} || return "$descr of $target
without Cancel-Key";
      return verify_cancel_key($key, $lock, $target);
   } else {
    # saveart('log.cancel',"$descr of $target without Cancel-Key and
Cancel-Lock");
    # -thh
    # no cancel-lock: go ahead and cancel anyway!
    # You refuse the cancel here...
    # local stuff...
    INN::syslog("info", "No Cancel-Key for $descr $target");
        saveart('bad.cancel','No Cancel-Key');
        return reject('No Cancel-Key');
    #return "$descr of $target without Cancel-key";
    # INN::cancel($target);
   }

   return undef;
}

	

-- 
Stéphane