--- lib/Mail/SpamAssassin/BayesStore/MySQL.pm	(revision 1138970)

+++ lib/Mail/SpamAssassin/BayesStore/MySQL.pm	(working copy)

@@ -840,14 +840,28 @@

       return 0;

     }

+    # With ON DUPLICATE KEY UPDATE, the affected-rows value per row is 1 if

+    # the row is inserted as a new row and 2 if an existing row is updated.

+    #

+    # Due to a MySQL server bug a value of 3 can be seen.

+    # See: http://bugs.mysql.com/bug.php?id=46675

+    #   When executing the INSERT ... ON DUPLICATE KEY UPDATE statement

+    #   and checking the rows return count:

+    #   mysql_client_found_rows = 0: The second INSERT returns a row count

+    #                                of 2 in all MySQL versions.

+    #   mysql_client_found_rows = 1: The second INSERT returns this row count:

+    #     Before MySQL 5.1.20: 2

+    #     MySQL 5.1.20: undef on Mac OS X, 139775481 on Linux (garbage?)

+    #     MySQL 5.1.21 and up: 3

+    #

     my $num_rows = $rc;

     $sth->finish();

-    if ($num_rows == 1 || $num_rows == 2) {

+    if ($num_rows == 1 || $num_rows == 2 || $num_rows == 3) {

       my $token_count_update = '';

-      $token_count_update = "token_count = token_count + 1," if ($num_rows == 1);

+      $token_count_update = "token_count = token_count + 1," if $num_rows == 1;

       $sql = "UPDATE bayes_vars SET

                      $token_count_update

                      newest_token_age = GREATEST(newest_token_age, ?),

@@ -872,7 +886,11 @@

     }

     else {

       # $num_rows was not what we expected

-      dbg("bayes: _put_token: Updated an unexpected number of rows.");

+      my $token_displ = $token;

+      $token_displ =~ s/(.)/sprintf('%02x',ord($1))/egs;

+      dbg("bayes: _put_token: Updated an unexpected number of rows: %s, ".

+          "id: %s, token (hex): %s",

+          $num_rows, $self->{_userid}, $token_displ);

       $self->{_dbh}->rollback();

       return 0;

     }

@@ -987,8 +1005,24 @@

       else {

 	my $num_rows = $rc;

-	$need_atime_update_p = 1 if ($num_rows == 1 || $num_rows == 2);

-	$new_tokens++ if ($num_rows == 1);

+        # With ON DUPLICATE KEY UPDATE, the affected-rows value per row is 1 if

+        # the row is inserted as a new row and 2 if an existing row is updated.

+        # But see MySQL bug (as above): http://bugs.mysql.com/bug.php?id=46675

+

+        if ($num_rows == 1) {

+          $new_tokens++;

+          $need_atime_update_p = 1;

+        } elsif ($num_rows == 2 || $num_rows == 3) {

+          $need_atime_update_p = 1;

+        } else {

+          # $num_rows was not what we expected

+          my $token_displ = $token;

+          $token_displ =~ s/(.)/sprintf('%02x',ord($1))/egs;

+          dbg("bayes: _put_tokens: Updated an unexpected number of rows: %s, ".

+              "id: %s, token (hex): %s",

+              $num_rows, $self->{_userid}, $token_displ);

+          $error_p = 1;

+        }

       }

     }

@@ -1026,10 +1060,10 @@

       }

     }

     else {

-      # $num_rows was not what we expected

-      dbg("bayes: _put_tokens: Updated an unexpected number of rows.");

-      $self->{_dbh}->rollback();

-      return 0;

+      info("bayes: _put_tokens: no atime updates needed?  Num of tokens: %d",

+           scalar keys %{$tokens});

+#     $self->{_dbh}->rollback();

+#     return 0;

     }

   }

$FreeBSD: head/mail/p5-Mail-SpamAssassin/files/patch-bug6655 340725 2014-01-22 17:40:44Z mat $

https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6655

--- lib/Mail/SpamAssassin/Util.pm	2011-06-06 19:59:17.000000000 -0400

+++ lib/Mail/SpamAssassin/Util.pm	2011-08-26 17:12:19.000000000 -0400

@@ -1025,6 +1024,8 @@

     return;

   }

+  opendir(my $dh, $tmpdir) || die "Could not open $tmpdir: $!";

+  closedir $dh;

   my ($reportfile, $tmpfile);

   my $umask = umask 077;

@@ -1052,7 +1053,10 @@

     # ensure the file handle is not semi-open in some way

     if ($tmpfile) {

-      close $tmpfile  or info("error closing $reportfile: $!");

+      if (! close $tmpfile) {

+       info("error closing $reportfile: $!");

+       $tmpfile=undef;

+      }

     }

   }

--- sa-update.raw    2011-06-24 13:38:50.000000000 -0400

+++ sa-update.raw    2011-08-29 09:38:50.000000000 -0400

@@ -677,9 +677,9 @@

   # Write the content out to a temp file for GPG/Archive::Tar interaction

   dbg("channel: populating temp content file");

-  open(TMP, ">$content_file") || die "fatal: can't write to content temp file $content_file: $!\n";

+  open(TMP, ">$content_file") || die "fatal: couldn't create content temp file $content_file: $!\n";

   binmode TMP;

-  print TMP $content;

+  print TMP $content || die "fatal: can't write to content temp file $content_file: $!\n";

   close(TMP);

   # to sign  : gpg -bas file

@@ -695,7 +695,7 @@

       die "fatal: couldn't create temp file for GPG signature: $!\n";

     }

     binmode $tfh;

-    print $tfh $GPG;

+    print $tfh $GPG || die "fatal: can't write temp file for GPG signature: $!\n";

     close($tfh);

     dbg("gpg: calling gpg");

--- lib/Mail/SpamAssassin/Plugin/DCC.pm	2011-06-06 19:59:17.000000000 -0400

+++ lib/Mail/SpamAssassin/Plugin/DCC.pm	2011-11-26 07:22:36.000000000 -0500

@@ -15,6 +15,20 @@

 # limitations under the License.

 # </@LICENSE>

+# Changes since SpamAssassin 3.3.2:

+#   support for DCC learning.  See dcc_learn_score.

+#   deal with orphan dccifd sockets

+#   use `cdcc -q` to not stall waiting to find a DCC server when deciding

+#     whether DCC checks are enabled

+#   use dccproc -Q or dccifd query if a pre-existing X-DCC header shows

+#     the message has already been reported

+#   dccproc now uses -w /var/dcc/whiteclnt so it acts more like dccifd

+#   warn about the use of ancient versions of dccproc and dccifd

+#   turn off dccifd greylisting

+#   query instead of reporting mail messages that contain X-DCC headers and

+#     and so has probably already been reported

+#   try harder to find dccproc and cdcc when not explicitly configured

+

 =head1 NAME

 Mail::SpamAssassin::Plugin::DCC - perform DCC check of messages

@@ -30,30 +44,31 @@

 The DCC or Distributed Checksum Clearinghouse is a system of servers

 collecting and counting checksums of millions of mail messages.

-TheSpamAssassin.pm counts can be used by SpamAssassin to detect and

-reject or filter spam.

-

-Because simplistic checksums of spam can be easily defeated, the main

-DCC checksums are fuzzy and ignore aspects of messages.  The fuzzy

-checksums are changed as spam evolves.

+The counts can be used by SpamAssassin to detect and filter spam.

-Note that DCC is disabled by default in C<init.pre> because it is not

-open source.  See the DCC license for more details.

+See http://www.dcc-servers.net/dcc/ for more information about DCC.

-See http://www.rhyolite.com/anti-spam/dcc/ for more information about

-DCC.

+Note that DCC is disabled by default in C<v310.pre> because its use requires

+software that is not distributed with SpamAssassin and that has license

+restrictions for certain commercial uses.

+See the DCC license at http://www.dcc-servers.net/dcc/LICENSE for details.

+

+Enable it by uncommenting the "loadplugin Mail::SpamAssassin::Plugin::DCC"

+confdir/v310.pre or by adding this line to your local.pre.  It might also

+be necessary to install a DCC package, port, rpm, or equivalent from your

+operating system distributor or a tarball from the primary DCC source

+at http://www.dcc-servers.net/dcc/#download

+See also http://www.dcc-servers.net/dcc/INSTALL.html

 =head1 TAGS

 The following tags are added to the set, available for use in reports,

 header fields, other plugins, etc.:

-  _DCCB_    DCC server ID in a response

-  _DCCR_    response from DCC - header field body in X-DCC-*-Metrics

-  _DCCREP_  response from DCC - DCC reputation in percents (0..100)

-

-Tag _DCCREP_ provides a nonempty value only with commercial DCC systems.

-This is the percentage of spam vs. ham sent from the first untrusted relay.

+  _DCCB_    DCC server ID in X-DCC-*-Metrics header field name

+  _DCCR_    X-DCC-*-Metrics header field body

+  _DCCREP_  DCC Reputation or percent bulk mail (0..100) from

+	      commercial DCC software

 =cut

@@ -75,8 +90,6 @@

 use vars qw(@ISA);

 @ISA = qw(Mail::SpamAssassin::Plugin);

-use vars qw($have_inet6);

-

 sub new {

   my $class = shift;

   my $mailsaobject = shift;

@@ -87,7 +100,7 @@

   # are network tests enabled?

   if ($mailsaobject->{local_tests_only}) {

-    $self->{dcc_disabled} = 1;

+    $self->{use_dcc} = 0;

     dbg("dcc: local tests only, disabling DCC");

   }

   else {

@@ -128,20 +141,23 @@

 =item dcc_fuz2_max NUMBER

-This option sets how often a message's body/fuz1/fuz2 checksum must have been

-reported to the DCC server before SpamAssassin will consider the DCC check as

-matched.

-

-As nearly all DCC clients are auto-reporting these checksums, you should set

-this to a relatively high value, e.g. C<999999> (this is DCC's MANY count).

+Sets how often a message's body/fuz1/fuz2 checksum must have been reported

+to the DCC server before SpamAssassin will consider the DCC check hit.

+C<999999> is DCC's MANY count.

 The default is C<999999> for all these options.

 =item dcc_rep_percent NUMBER

-Only commercial DCC systems provide DCC reputation information. This is the

-percentage of spam vs. ham sent from the first untrusted relay.  It will hit

-on new spam from spam sources.  Default is C<90>.

+Only the commercial DCC software provides DCC Reputations.  A DCC Reputation

+is the percentage of bulk mail received from the last untrusted relay in the

+path taken by a mail message as measured by all commercial DCC installations.

+See http://www.rhyolite.com/dcc/reputations.html

+You C<must> whitelist your trusted relays or MX servers with MX or

+MXDCC lines in /var/dcc/whiteclnt as described in the main DCC man page

+to avoid seeing your own MX servers as sources of bulk mail.

+See http://www.dcc-servers.net/dcc/dcc-tree/dcc.html#White-and-Blacklists

+The default is C<90>.

 =cut

@@ -189,13 +205,9 @@

 =item dcc_home STRING

 This option tells SpamAssassin where to find the dcc homedir.

-If not given, it will try to get dcc to specify one, and if that fails it

-will try dcc's own default homedir of '/var/dcc'.

-If C<dcc_path> is not specified, it will default to looking in

-C<dcc_home/bin> for dcc client instead of relying on SpamAssassin to find it

-in the current PATH.  If it isn't found there, it will look in the current

-PATH. If a C<dccifd> socket is found in C<dcc_home> or specified explicitly,

-it will use that interface instead of C<dccproc>.

+If not specified, try to use the locally configured directory

+from the C<cdcc homedir> command.

+Try /var/dcc if that command fails.

 =cut

@@ -205,7 +217,7 @@

     type => $Mail::SpamAssassin::Conf::CONF_TYPE_STRING,

     code => sub {

       my ($self, $key, $value, $line) = @_;

-      if (!defined $value || !length $value) {

+      if (!defined $value || $value eq '') {

 	return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;

       }

       $value = untaint_file_path($value);

@@ -223,14 +235,16 @@

 =item dcc_dccifd_path STRING

-This option tells SpamAssassin where to find the dccifd socket. If

-C<dcc_dccifd_path> is not specified, it will default to looking for a socket

-named C<dccifd> in a directory C<dcc_home>.  The C<dcc_dccifd_path> can be

-a Unix socket name (absolute path), or an INET socket specification in a form

-C<[host]:port> or C<host:port>, where a host can be an IPv4 or IPv6 address

-or a host name, and port is a TCP port number. In case of an IPv6 address the

-brackets are required syntax. If a C<dccifd> socket is found, the plugin will

-use it instead of C<dccproc>.

+This option tells SpamAssassin where to find the dccifd socket instead

+of a local Unix socket named C<dccifd> in the C<dcc_home> directory.

+If a socket is specified or found, use it instead of C<dccproc>.

+

+If specifed, C<dcc_dccifd_path> is the absolute path of local Unix socket

+or an INET socket specified as C<[Host]:Port> or C<Host:Port>.

+Host can be an IPv4 or IPv6 address or a host name

+Port is a TCP port number. The brackets are required for an IPv6 address.

+

+The default is C<undef>.

 =cut

@@ -240,45 +254,60 @@

     type => $Mail::SpamAssassin::Conf::CONF_TYPE_STRING,

     code => sub {

       my ($self, $key, $value, $line) = @_;

-      $value = ''  if !defined $value;

-      $self->{dcc_dccifd_path_raw} = $value;  # for logging purposes

-      undef $self->{dcc_dccifd_host};

-      undef $self->{dcc_dccifd_port};

-      undef $self->{dcc_dccifd_socket};

-      local($1,$2,$3);

-      if ($value eq '') {

+

+      if (!defined $value || $value eq '') {

 	return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;

-      } elsif ($value =~ m{^ (?: \[ ([^\]]*) \] | ([^:]*) ) : ([^:]*) \z}sx) {

-        # "[host]:port" or "host:port", where a host can be an IPv4 or IPv6

-        # address or a host name, and port is a TCP port number or service name

-        my $host = defined $1 ? $1 : $2;

-        my $port = $3;

-        $self->{dcc_dccifd_host} = untaint_var($host);

-        $self->{dcc_dccifd_port} = untaint_var($port);

-        dbg("config: dcc_dccifd_path set to [%s]:%s", $host,$port);

-      } else {  # assume a unix socket

+      }

+

+      local($1,$2,$3);

+      if ($value =~ m{^ (?: \[ ([^\]]*) \] | ([^:]*) ) : ([^:]*) \z}sx) {

+	my $host = untaint_var(defined $1 ? $1 : $2);

+	my $port = untaint_var($3);

+	if (!$host) {

+	  info("config: missing or bad host name in dcc_dccifd_path '$value'");

+	  return $Mail::SpamAssassin::Conf::INVALID_VALUE;

+	}

+	if (!$port || $port !~ /^\d+\z/ || $port < 1 || $port > 65535) {

+	  info("config: bad TCP port number in dcc_dccifd_path '$value'");

+	  return $Mail::SpamAssassin::Conf::INVALID_VALUE;

+	}

+

+	$self->{dcc_dccifd_host} = $host;

+	$self->{dcc_dccifd_port} = $port;

+	if ($host !~ /^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\z/) {

+	  # remember to try IPv6 if we can with a host name or non-IPv4 address

+	  $self->{dcc_dccifd_IPv6} = eval { require IO::Socket::INET6 };

+	}

+	dbg("config: dcc_dccifd_path set to [%s]:%s", $host, $port);

+

+      } else {

+	# assume a unix socket

         if ($value !~ m{^/}) {

-          info("config: dcc_dccifd_path should be an absolute socket path");

+	  info("config: dcc_dccifd_path '$value' is not an absolute path");

         # return $Mail::SpamAssassin::Conf::INVALID_VALUE;  # abort or accept?

         }

         $value = untaint_file_path($value);

-      # test disabled, dccifd may not yet be running at spamd startup time

-      # if (!-S $value) {

-      #   info("config: dcc_dccifd_path '$value' isn't a local socket");

-      #   return $Mail::SpamAssassin::Conf::INVALID_VALUE;

-      # }

+

         $self->{dcc_dccifd_socket} = $value;

         dbg("config: dcc_dccifd_path set to local socket %s", $value);

+	dbg("dcc: dcc_dccifd_path set to local socket %s", $value);

       }

+

+      $self->{dcc_dccifd_path_raw} = $value;

     }

   });

 =item dcc_path STRING

-This option tells SpamAssassin specifically where to find the C<dccproc>

-client instead of relying on SpamAssassin to find it in the current PATH.

-Note that if I<taint mode> is enabled in the Perl interpreter, you should

-use this, as the current PATH will have been cleared.

+Where to find the C<dccproc> client program instead of relying on SpamAssassin

+to find it in the current PATH or C<dcc_home/bin>. This must often be set,

+because the current PATH is cleared by I<taint mode> in the Perl interpreter,

+

+If a C<dccifd> socket is found in C<dcc_home> or specified explicitly

+with C<dcc_dccifd_path>, use the C<dccifd(8)> interface instead of C<dccproc>.

+

+The default is C<undef>.

+

 =cut

@@ -289,12 +318,12 @@

     type => $Mail::SpamAssassin::Conf::CONF_TYPE_STRING,

     code => sub {

       my ($self, $key, $value, $line) = @_;

-      if (!defined $value || !length $value) {

+      if (!defined $value || $value eq '') {

 	return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;

       }

       $value = untaint_file_path($value);

       if (!-x $value) {

-	info("config: dcc_path '$value' isn't an executable");

+	info("config: dcc_path '$value' is not executable");

 	return $Mail::SpamAssassin::Conf::INVALID_VALUE;

       }

@@ -304,7 +333,7 @@

 =item dcc_options options

-Specify additional options to the dccproc(8) command. Please note that only

+Specify additional options to the dccproc(8) command.  Only

 characters in the range [0-9A-Za-z ,._/-] are allowed for security reasons.

 The default is C<undef>.

@@ -319,6 +348,7 @@

     code => sub {

       my ($self, $key, $value, $line) = @_;

       if ($value !~ m{^([0-9A-Za-z ,._/-]+)$}) {

+	info("config: dcc_options '$value' contains impermissible characters");

 	return $Mail::SpamAssassin::Conf::INVALID_VALUE;

       }

       $self->{dcc_options} = $1;

@@ -327,8 +357,9 @@

 =item dccifd_options options

-Specify additional options to send to the dccifd(8) daemon. Please note that only

-characters in the range [0-9A-Za-z ,._/-] are allowed for security reasons.

+Specify additional options to send to the dccifd daemon with

+the ASCII protocol described on the dccifd(8) man page.

+Only characters in the range [0-9A-Za-z ,._/-] are allowed for security reasons.

 The default is C<undef>.

@@ -342,265 +373,306 @@

     code => sub {

       my ($self, $key, $value, $line) = @_;

       if ($value !~ m{^([0-9A-Za-z ,._/-]+)$}) {

+	info("config: dccifd_options '$value' contains impermissible characters");

 	return $Mail::SpamAssassin::Conf::INVALID_VALUE;

       }

       $self->{dccifd_options} = $1;

     }

   });

+=item dcc_learn_score n		(default: undef)

+

+Report messages with total scores this much larger than the

+SpamAssassin spam threshold to DCC as spam.

+

+=cut

+

+  push (@cmds, {

+    setting => 'dcc_learn_score',

+    is_admin => 1,

+    default => undef,

+    type => $Mail::SpamAssassin::Conf::CONF_TYPE_NUMERIC,

+  });

+

   $conf->{parser}->register_commands(\@cmds);

 }

+

+

+

+sub ck_dir {

+  my ($self, $dir, $tgt, $src) = @_;

+

+  $dir = untaint_file_path($dir);

+  if (!stat($dir)) {

+    my $dir_errno = 0+$!;

+    if ($dir_errno == ENOENT) {

+      dbg("dcc: $tgt $dir from $src does not exist");

+    } else {

+      dbg("dcc: $tgt $dir from $src is not accessible: $!");

+    }

+    return;

+  }

+  if (!-d _) {

+    dbg("dcc: $tgt $dir from $src is not a directory");

+    return;

+  }

+

+  $self->{main}->{conf}->{$tgt} = $dir;

+  dbg("dcc: use '$tgt $dir' from $src");

+}

+

 sub find_dcc_home {

   my ($self) = @_;

+  my $dcc_libexec;

+

+  # just once

+  return if defined $self->{dcc_version};

+  $self->{dcc_version} = '?';

   my $conf = $self->{main}->{conf};

-  return if !$conf->{use_dcc};

-  my $dcchome = $conf->{dcc_home} || '';

-  # If we're not given the DCC homedir, try getting DCC to tell us it.

-  # If that fails, try the DCC default homedir of '/var/dcc'.

-  if ($dcchome eq '') {

+  # Get the DCC software version for talking to dccifd and formating the

+  # dccifd options and the built-in DCC homedir.  Use -q to prevent delays.

+  my $cdcc_home;

+  my $cdcc = $self->dcc_pgm_path('cdcc');

+  my $cmd = '-qV homedir libexecdir';

+  if ($cdcc && open(CDCC, "$cdcc $cmd 2>&1 |")) {

+    my $cdcc_output = do { local $/ = undef; <CDCC> };

+    close CDCC;

-    my $cdcc = Mail::SpamAssassin::Util::find_executable_in_env_path('cdcc');

+    $cdcc_output =~ s/\n/ /g;		# everything in 1 line for debugging

+    dbg("dcc: `%s %s` reports '%s'", $cdcc, $cmd, $cdcc_output);

+    $self->{dcc_version} = ($cdcc_output =~ /^(\d+\.\d+\.\d+)/) ? $1 : '';

+    $cdcc_home = ($cdcc_output =~ /\s+homedir=(\S+)/) ? $1 : '';

+    if ($cdcc_output =~ /\s+libexecdir=(\S+)/) {

+      $self->ck_dir($1, 'dcc_libexec', 'cdcc');

+    }

+  }

-    my $cdcc_home = '';

-    if ($cdcc && -x $cdcc && open(CDCC, "$cdcc homedir 2>&1|")) {

-      dbg("dcc: dcc_home not set, querying cdcc utility");

-      $cdcc_home = <CDCC> || '';

-      close CDCC;

+  # without a home, try the homedir from cdcc

+  if (!$conf->{dcc_home} && $cdcc_home) {

+    $self->ck_dir($cdcc_home, 'dcc_home', 'cdcc');

+  }

+  # finally fall back to /var/dcc

+  if (!$conf->{dcc_home}) {

+    $self->ck_dir($conf->{dcc_home} = '/var/dcc', 'dcc_home', 'default')

+  }

-      chomp $cdcc_home;

-      $cdcc_home =~ s/\s+homedir=//;

-      dbg("dcc: cdcc reports homedir as '%s'", $cdcc_home);

-    }

-

-    # try first with whatever the cdcc utility reported

-    my $cdcc_home_errno = 0;

-    if ($cdcc_home eq '') {

-      $cdcc_home_errno = ENOENT;

-    } elsif (!stat($cdcc_home)) {

-      $cdcc_home_errno = 0+$!;

-    }

-    if ($cdcc_home_errno == ENOENT) {

-      # no such file

-    } elsif ($cdcc_home_errno != 0) {

-      dbg("dcc: cdcc reported homedir $cdcc_home is not accessible: $!");

-    } elsif (!-d _) {

-      dbg("dcc: cdcc reported homedir $cdcc_home is not a directory");

-    } else {  # ok

-      dbg("dcc: cdcc reported homedir $cdcc_home exists, using it");

-      $dcchome = untaint_var($cdcc_home);

-    }

-

-    # try falling back to /var/dcc

-    if ($dcchome eq '') {

-      my $var_dcc_errno = stat('/var/dcc') ? 0 : 0+$!;

-      if ($var_dcc_errno == ENOENT) {

-        # no such file

-      } elsif ($var_dcc_errno != 0) {

-        dbg("dcc: dcc_home not set and dcc default homedir /var/dcc ".

-            "is not accessible: $!");

-      } elsif (!-d _) {

-        dbg("dcc: dcc_home not set and dcc default homedir /var/dcc ".

-            "is not a directory");

-      } else {  # ok

-        dbg("dcc: dcc_home not set but dcc default homedir /var/dcc exists, ".

-            "using it");

-        $dcchome = '/var/dcc';

+  # fall back to $conf->{dcc_home}/libexec or /var/dcc/libexec for dccsight

+  if (!$conf->{dcc_libexec}) {

+    $self->ck_dir($conf->{dcc_home} . '/libexec', 'dcc_libexec', 'dcc_home');

       }

+  if (!$conf->{dcc_libexec}) {

+    $self->ck_dir('/var/dcc/libexec', 'dcc_libexec', 'dcc_home');

     }

-    if ($dcchome eq '') {

-      dbg("dcc: unable to get homedir from cdcc ".

-          "and the dcc default homedir was not found");

-    }

-

-    # Remember found homedir path

-    dbg("dcc: using '%s' as DCC homedir", $dcchome);

-    $conf->{dcc_home} = $dcchome;

+  # format options for dccifd

+  my $opts = ($conf->{dccifd_options} || '') . "\n";

+  if ($self->{dcc_version} =~ /\d+\.(\d+)\.(\d+)$/ &&

+      ($1 < 3 || ($1 == 3 && $2 < 123))) {

+    if ($1 < 3 || ($1 == 3 && $2 < 50)) {

+      info("dcc: DCC version $self->{dcc_version} is years old, ".

+           "obsolete, and likely to cause problems.  ".

+           "See http://www.dcc-servers.net/dcc/old-versions.html");

+    }

+    $self->{dccifd_lookup_options} = "header " . $opts;

+    $self->{dccifd_report_options} = "header spam " . $opts;

+  } else {

+    # dccifd after version 1.2.123 understands "cksums" and "no-grey"

+    $self->{dccifd_lookup_options} = "cksums grey-off " . $opts;

+    $self->{dccifd_report_options} = "header spam grey-off " . $opts;

   }

 }

-sub is_dccifd_available {

-  my ($self) = @_;

-

+sub dcc_pgm_path {

+  my ($self, $pgm) = @_;

+  my $pgmpath;

   my $conf = $self->{main}->{conf};

-  $self->{dccifd_available} = 0;

-  if (!$conf->{use_dcc}) {

-    dbg("dcc: dccifd is not available: use_dcc is false");

-  } elsif (defined $conf->{dcc_dccifd_host}) {

-    dbg("dcc: dccifd inet socket chosen: [%s]:%s",

-        $conf->{dcc_dccifd_host}, $conf->{dcc_dccifd_port});

-    $self->{dccifd_available} = 1;

-  } else {

-    my $sockpath = $conf->{dcc_dccifd_socket};

-    my $dcchome = $conf->{dcc_home};

-    if (defined $sockpath) {

-      dbg("dcc: dccifd local socket chosen: %s", $sockpath);

-    } elsif (defined $conf->{dcc_dccifd_path_raw}) {

-      # avoid falling back to defaults if explicitly provided but wrong

-    } elsif (defined $dcchome && $dcchome ne '' && -S "$dcchome/dccifd") {

-      $sockpath = "$dcchome/dccifd";

-      $conf->{dcc_dccifd_socket} = $sockpath;

-      dbg("dcc: dccifd default local socket chosen: %s", $sockpath);

+  $pgmpath = $conf->{dcc_path};

+  if (defined $pgmpath && $pgmpath ne '') {

+    # accept explicit setting for dccproc

+    return $pgmpath if $pgm eq 'dccproc';

+    # try adapting it for cdcc and everything else

+    if ($pgmpath =~ s{[^/]+\z}{$pgm}s) {

+      $pgmpath = untaint_file_path($pgmpath);

+      if (-x $pgmpath) {

+        dbg("dcc: dcc_pgm_path, found %s in dcc_path: %s", $pgm,$pgmpath);

+        return $pgmpath;

     }

-    if (defined $sockpath && -S $sockpath && -w _ && -r _) {

-      $self->{dccifd_available} = 1;

-    } elsif (!defined $conf->{dcc_dccifd_path_raw}) {

-      dbg("dcc: dccifd is not available: no r/w dccifd socket found");

-    } else {

-      dbg("dcc: dccifd is not available: no r/w dccifd socket found: %s",

-          $conf->{dcc_dccifd_path_raw});

     }

   }

-  return $self->{dccifd_available};

+  $pgmpath = Mail::SpamAssassin::Util::find_executable_in_env_path($pgm);

+  if (defined $pgmpath) {

+    dbg("dcc: dcc_pgm_path, found %s in env.path: %s", $pgm,$pgmpath);

+    return $pgmpath;

+  }

+

+  # try dcc_home/bin, dcc_libexec, and some desperate last attempts

+  foreach my $dir ($conf->{dcc_home}.'/bin',  $conf->{dcc_libexec},

+                   '/usr/local/bin', '/usr/local/dcc', '/var/dcc') {

+    $pgmpath = $dir . '/' . $pgm;

+    if (-x $pgmpath) {

+      dbg("dcc: dcc_pgm_path, found %s in %s: %s", $pgm,$dir,$pgmpath);

+      return $pgmpath;

+    }

+  }

+

+  return;

 }

-sub is_dccproc_available {

+sub is_dccifd_available {

   my ($self) = @_;

   my $conf = $self->{main}->{conf};

-  $self->{dccproc_available} = 0;

+  # dccifd remains available until it breaks

+  return $self->{dccifd_available} if $self->{dccifd_available};

-  if (!$conf->{use_dcc}) {

-    dbg("dcc: dccproc is not available: use_dcc is false");

-    return 0;

+  # deal with configured INET socket

+  if (defined $conf->{dcc_dccifd_host}) {

+    dbg("dcc: dccifd is available via INET socket [%s]:%s",

+	$conf->{dcc_dccifd_host}, $conf->{dcc_dccifd_port});

+    return ($self->{dccifd_available} = 1);

   }

-  my $dcchome = $conf->{dcc_home} || '';

-  my $dccproc = $conf->{dcc_path} || '';

-  if ($dccproc eq '' && ($dcchome ne '' && -x "$dcchome/bin/dccproc")) {

-    $dccproc = "$dcchome/bin/dccproc";

+  # the first time here, compute a default local socket based on DCC home

+  # from self->find_dcc_home() called elsewhere

+  my $sockpath = $conf->{dcc_dccifd_socket};

+  if (!$sockpath) {

+      if ($conf->{dcc_dccifd_path_raw}) {

+	$sockpath = $conf->{dcc_dccifd_path_raw};

+      } else {

+	$sockpath = "$conf->{dcc_home}/dccifd";

   }

-  if ($dccproc eq '') {

-    $dccproc = Mail::SpamAssassin::Util::find_executable_in_env_path('dccproc');

+      $conf->{dcc_dccifd_socket} = $sockpath;

   }

-  unless (defined $dccproc && $dccproc ne '' && -x $dccproc) {

-    dbg("dcc: dccproc is not available: no dccproc executable found");

-    return 0;

-  }

+  # check the socket every time because it can appear and disappear

+  return ($self->{dccifd_available} = 1) if (-S $sockpath && -w _ && -r _);

-  # remember any found dccproc

+  dbg("dcc: dccifd is not available; no r/w socket at %s", $sockpath);

+  return ($self->{dccifd_available} = 0);

+}

+

+sub is_dccproc_available {

+  my ($self) = @_;

+  my $conf = $self->{main}->{conf};

+

+  # dccproc remains (un)available so check only once

+  return $self->{dccproc_available} if  defined $self->{dccproc_available};

+

+  my $dccproc = $conf->{dcc_path};

+  if (!defined $dccproc || $dccproc eq '') {

+    $dccproc = $self->dcc_pgm_path('dccproc');

   $conf->{dcc_path} = $dccproc;

+    if (!$dccproc || ! -x $dccproc) {

+      dbg("dcc: dccproc is not available: no dccproc executable found");

+      return ($self->{dccproc_available} = 0);

+    }

+  }

-  dbg("dcc: dccproc is available: %s", $conf->{dcc_path});

-  $self->{dccproc_available} = 1;

-  return 1;

+  dbg("dcc: %s is available", $conf->{dcc_path});

+  return ($self->{dccproc_available} = 1);

 }

 sub dccifd_connect {

-  my($self) = @_;

+  my($self, $tag) = @_;

   my $conf = $self->{main}->{conf};

   my $sockpath = $conf->{dcc_dccifd_socket};

-  my $host = $conf->{dcc_dccifd_host};

-  my $port = $conf->{dcc_dccifd_port};

   my $sock;

+

   if (defined $sockpath) {

-    dbg("dcc: connecting to a local socket %s", $sockpath);

-    $sock = IO::Socket::UNIX->new(

-              Type => SOCK_STREAM, Peer => $sockpath);

-    $sock or die "dcc: failed to connect to a socket $sockpath: $!\n";

-  } elsif (defined $host) {

-    my $specified_path = $conf->{dcc_dccifd_path_raw};

-    if ($host eq '') {

-      die "dcc: empty host specification: $specified_path\n";

-    }

-    if (!defined $port || $port !~ /^\d+\z/ || $port < 1 || $port > 65535) {

-      die "dcc: bad TCP port number: $specified_path\n";

-    }

-    my $is_inet4 = $host =~ /^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\z/;

-    if ($is_inet4) {  # inet4 socket (IPv4 address)

-      dbg("dcc: connecting to inet4 socket [%s]:%s", $host,$port);

-      $sock = IO::Socket::INET->new(

-                Proto => 'tcp', PeerAddr => $host, PeerPort => $port);

-    } else {

-      if (!defined $have_inet6) {

-        $have_inet6 = eval { require IO::Socket::INET6 };

-        $have_inet6 = 0  if !defined $have_inet6;

+    $sock = IO::Socket::UNIX->new(Type => SOCK_STREAM, Peer => $sockpath);

+    if ($sock) {

+      dbg("$tag connected to local socket %s", $sockpath);

+      return $sock;

       }

-      if (!$have_inet6) {  # fallback to an inet4 socket (IPv4)

-        dbg("dcc: connecting(2) to inet4 socket [%s]:%s", $host,$port);

-        $sock = IO::Socket::INET->new(

-                  Proto => 'tcp', PeerAddr => $host, PeerPort => $port);

-      } else {  # inet6 socket (IPv6) or a host name

-        dbg("dcc: connecting to inet6 socket [%s]:%s", $host,$port);

+    $self->{dccifd_available} = 0;

+    info("$tag failed to connect to local socket $sockpath");

+    return $sock

+  }

+

+  # must be TCP/IP

+  my $host = $conf->{dcc_dccifd_host};

+  my $port = $conf->{dcc_dccifd_port};

+

+  if ($conf->{dcc_dccifd_IPv6}) {

+    # try IPv6 if we can with a host name or non-IPv4 address

+    dbg("$tag connecting to inet6 socket [%s]:%s", $host,$port);

         $sock = IO::Socket::INET6->new(

                   Proto => 'tcp', PeerAddr => $host, PeerPort => $port);

+    # fall back to IPv4 if that failed

       }

+  if (!$sock) {

+    dbg("$tag connecting to inet4 socket [%s]:%s", $host, $port);

+    $sock = IO::Socket::INET->new(

+		Proto => 'tcp', PeerAddr => $host, PeerPort => $port);

     }

-    $sock or die "dcc: failed to connect to [$host]:$port : $!\n";

-  } else {

-    die "dcc: dccifd socket not provided: $conf->{dcc_dccifd_path_raw}\n";

-  }

+

+  info("failed to connect to [$host]:$port : $!") if !$sock;

   return $sock;

 }

+# check for dccifd every time in case enough uses of dccproc starts dccifd

 sub get_dcc_interface {

   my ($self) = @_;

+  my $conf = $self->{main}->{conf};

-  if ($self->is_dccifd_available()) {

-    $self->{dcc_interface} = "dccifd";

-    $self->{dcc_disabled} = 0;

-  }

-  elsif ($self->is_dccproc_available()) {

-    $self->{dcc_interface} = "dccproc";

-    $self->{dcc_disabled} = 0;

+  if (!$conf->{use_dcc}) {

+    $self->{dcc_disabled} = 1;

+    return;

   }

-  else {

-    dbg("dcc: dccifd and dccproc are not available, disabling DCC");

-    $self->{dcc_interface} = "none";

+

+  $self->find_dcc_home();

+  if (!$self->is_dccifd_available() && !$self->is_dccproc_available()) {

+    dbg("dcc: dccifd and dccproc are not available");

     $self->{dcc_disabled} = 1;

   }

+

+  $self->{dcc_disabled} = 0;

 }

 sub dcc_query {

-  my ($self, $permsgstatus, $full) = @_;

+  my ($self, $permsgstatus, $fulltext) = @_;

   $permsgstatus->{dcc_checked} = 1;

+  if (!$self->{main}->{conf}->{use_dcc}) {

+    dbg("dcc: DCC is not available: use_dcc is 0");

+    return;

+  }

+

   # initialize valid tags

   $permsgstatus->{tag_data}->{DCCB} = "";

   $permsgstatus->{tag_data}->{DCCR} = "";

   $permsgstatus->{tag_data}->{DCCREP} = "";

-  # short-circuit if there's already a X-DCC header with value of

-  # "bulk" from an upstream DCC check

-  if ($permsgstatus->get('ALL') =~

-      /^(X-DCC-([^:]{1,80})?-?Metrics:.*bulk.*)$/m) {

-    $permsgstatus->{dcc_response} = $1;

+  if ($$fulltext eq '') {

+    dbg("dcc: empty message; skipping dcc check");

     return;

   }

-  my $timer = $self->{main}->time_method("check_dcc");

+  if ($permsgstatus->get('ALL') =~ /^(X-DCC-.*-Metrics:.*)$/m) {

+    $permsgstatus->{dcc_raw_x_dcc} = $1;

+    # short-circuit if there is already a X-DCC header with value of

+    # "bulk" from an upstream DCC check

+    # require "bulk" because then at least one body checksum will be "many"

+    # and so we know the X-DCC header is not forged by spammers

+    return if $permsgstatus->{dcc_raw_x_dcc} =~ / bulk /;

+  }

-  $self->find_dcc_home();

+  my $timer = $self->{main}->time_method("check_dcc");

   $self->get_dcc_interface();

-  my $result;

-  if ($self->{dcc_disabled}) {

-    $result = 0;

-  } elsif ($$full eq '') {

-    dbg("dcc: empty message, skipping dcc check");

-    $result = 0;

-  } elsif ($self->{dccifd_available}) {

-    my $client = $permsgstatus->{relays_external}->[0]->{ip};

-    my $clientname = $permsgstatus->{relays_external}->[0]->{rdns};

-    my $helo = $permsgstatus->{relays_external}->[0]->{helo} || "";

-    if ($client) {

-      $client = $client . "\r" . $clientname  if $clientname;

-    } else {

-      $client = "0.0.0.0";

-    }

-    $self->dccifd_lookup($permsgstatus, $full, $client, $clientname, $helo);

-  } else {

-    my $client = $permsgstatus->{relays_external}->[0]->{ip};

-    $self->dccproc_lookup($permsgstatus, $full, $client);

-  }

+  return if $self->{dcc_disabled};

+

+  my $envelope = $permsgstatus->{relays_external}->[0];

+  ($permsgstatus->{dcc_raw_x_dcc},

+   $permsgstatus->{dcc_cksums}) = $self->ask_dcc("dcc:", $permsgstatus,

+						 $fulltext, $envelope);

 }

 sub check_dcc {

@@ -609,28 +681,27 @@

   $self->dcc_query($permsgstatus, $full)  if !$permsgstatus->{dcc_checked};

-  my $response = $permsgstatus->{dcc_response};

-  return 0  if !defined $response || $response eq '';

+  my $x_dcc = $permsgstatus->{dcc_raw_x_dcc};

+  return 0  if !defined $x_dcc || $x_dcc eq '';

-  local($1,$2);

-  if ($response =~ /^X-DCC-(.*)-Metrics: (.*)$/) {

-    $permsgstatus->{tag_data}->{DCCB} = $1;

-    $permsgstatus->{tag_data}->{DCCR} = $2;

+  if ($x_dcc =~ /^X-DCC-(.*)-Metrics: (.*)$/) {

+    $permsgstatus->set_tag('DCCB', $1);

+    $permsgstatus->set_tag('DCCR', $2);

   }

-  $response =~ s/many/999999/ig;

-  $response =~ s/ok\d?/0/ig;

+  $x_dcc =~ s/many/999999/ig;

+  $x_dcc =~ s/ok\d?/0/ig;

   my %count = (body => 0, fuz1 => 0, fuz2 => 0, rep => 0);

-  if ($response =~ /\bBody=(\d+)/) {

+  if ($x_dcc =~ /\bBody=(\d+)/) {

     $count{body} = $1+0;

   }

-  if ($response =~ /\bFuz1=(\d+)/) {

+  if ($x_dcc =~ /\bFuz1=(\d+)/) {

     $count{fuz1} = $1+0;

   }

-  if ($response =~ /\bFuz2=(\d+)/) {

+  if ($x_dcc =~ /\bFuz2=(\d+)/) {

     $count{fuz2} = $1+0;

   }

-  if ($response =~ /\brep=(\d+)/) {

+  if ($x_dcc =~ /\brep=(\d+)/) {

     $count{rep}  = $1+0;

   }

   if ($count{body} >= $conf->{dcc_body_max} ||

@@ -651,185 +722,185 @@

 }

 sub check_dcc_reputation_range {

-  my ($self, $permsgstatus, $full, $min, $max) = @_;

-  $self->dcc_query($permsgstatus, $full)  if !$permsgstatus->{dcc_checked};

+  my ($self, $permsgstatus, $fulltext, $min, $max) = @_;

+

+  # this is called several times per message, so parse the X-DCC header once

+  my $dcc_rep = $permsgstatus->{dcc_rep};

+  if (!defined $dcc_rep) {

+    $self->dcc_query($permsgstatus, $fulltext)  if !$permsgstatus->{dcc_checked};

+    my $x_dcc = $permsgstatus->{dcc_raw_x_dcc};

+    if (defined $x_dcc && $x_dcc =~ /\brep=(\d+)/) {

+      $dcc_rep = $1+0;

+      $permsgstatus->set_tag('DCCREP', $dcc_rep);

+    } else {

+      $dcc_rep = -1;

+    }

+    $permsgstatus->{dcc_rep} = $dcc_rep;

+  }

-  my $response = $permsgstatus->{dcc_response};

-  return 0  if !defined $response || $response eq '';

+  # no X-DCC header or no reputation in the X-DCC header, perhaps for lack

+  # of data in the DCC Reputation server

+  return 0 if $dcc_rep < 0;

+  # cover the entire range of reputations if not told otherwise

   $min = 0   if !defined $min;

-  $max = 999 if !defined $max;

+  $max = 100 if !defined $max;

-  local $1;

-  my $dcc_rep;

-  $dcc_rep = $1+0  if defined $response && $response =~ /\brep=(\d+)/;

-  if (defined $dcc_rep) {

-    $dcc_rep = int($dcc_rep);  # just in case, rule ranges are integer percents

     my $result = $dcc_rep >= $min && $dcc_rep <= $max ? 1 : 0;

     dbg("dcc: dcc_rep %s, min %s, max %s => result=%s",

         $dcc_rep, $min, $max, $result?'YES':'no');

-    $permsgstatus->{tag_data}->{DCCREP} = $dcc_rep;

-    return $dcc_rep >= $min && $dcc_rep <= $max ? 1 : 0;

+  return $result;

+}

+

+# get the X-DCC header line and save the checksums from dccifd or dccproc

+sub parse_dcc_response {

+  my ($self, $resp) = @_;

+  my ($raw_x_dcc, $cksums);

+

+  # The first line is the header we want.  It uses SMTP folded whitespace

+  # if it is long.  The folded whitespace is always a single \t.

+  chomp($raw_x_dcc = shift @$resp);

+  my $v;

+  while (($v = shift @$resp) && $v =~ s/^\t(.+)\s*\n/ $1/) {

+    $raw_x_dcc .= $v;

+  }

+

+  # skip the "reported:" line between the X-DCC header and any checksums

+  # remove ':' to avoid a bug in versions 1.3.115 - 1.3.122 in dccsight

+  # with the length of "Message-ID:"

+  $cksums = '';

+  while (($v = shift @$resp) && $v =~ s/^([^:]*):/$1/) {

+    $cksums .= $v;

   }

-  return 0;

+

+  return ($raw_x_dcc, $cksums);

 }

-sub dccifd_lookup {

-  my ($self, $permsgstatus, $fulltext, $client, $clientname, $helo) = @_;

+sub ask_dcc {

+  my ($self, $tag, $permsgstatus, $fulltext, $envelope) = @_;

   my $conf = $self->{main}->{conf};

-  my $response;

-  my $left;

-  my $right;

-  my $timeout = $conf->{dcc_timeout};

-  my $opts = $conf->{dccifd_options};

-  my @opts = !defined $opts ? () : split(' ',$opts);

+  my ($pgm, $err, $sock, $pid, @resp);

+  my ($client, $clientname, $helo, $opts);

   $permsgstatus->enter_helper_run_mode();

+  my $timeout = $conf->{dcc_timeout};

   my $timer = Mail::SpamAssassin::Timeout->new(

            { secs => $timeout, deadline => $permsgstatus->{master_deadline} });

-  my $err = $timer->run_and_catch(sub {

+  $err = $timer->run_and_catch(sub {

     local $SIG{PIPE} = sub { die "__brokenpipe__ignore__\n" };

-    my $sock = $self->dccifd_connect();

-    $sock or die "dcc: failed to connect to a dccifd socket";

-

-    # send the options and other parameters to the daemon

-    $sock->print("header " . join(" ",@opts) . "\n")

-                                 or die "dcc: failed write";  # options

-    $sock->print($client . "\n") or die "dcc: failed write";  # client

-    $sock->print($helo . "\n")   or die "dcc: failed write";  # HELO value

-    $sock->print("\n")           or die "dcc: failed write";  # sender

-    $sock->print("unknown\r\n")  or die "dcc: failed write";  # recipients

-    $sock->print("\n")           or die "dcc: failed write";  # recipients

-

-    $sock->print($$fulltext)     or die "dcc: failed write";

-

-    $sock->shutdown(1) or die "dcc: failed socket shutdown: $!";

-

-    $sock->getline()   or die "dcc: failed read status";

-    $sock->getline()   or die "dcc: failed read multistatus";

+    # prefer dccifd to dccproc

+    if ($self->{dccifd_available}) {

+      $pgm = 'dccifd';

-    my @null = $sock->getlines();

-    if (!@null) {

-      # no facility prefix on this

-      die "dcc: failed to read header\n";

-    }

+      $sock = $self->dccifd_connect($tag);

+      if (!$sock) {

+	$self->{dccifd_available} = 0;

+	die("dccproc not available") if (!$self->is_dccproc_available());

-    # the first line will be the header we want to look at

-    chomp($response = shift @null);

-    # but newer versions of DCC fold the header if it's too long...

-    while (my $v = shift @null) {

-      last unless ($v =~ s/^\s+/ /);  # if this line wasn't folded, stop

-      chomp $v;

-      $response .= $v;

+	# fall back on dccproc if the socket is an orphan from

+	# a killed dccifd daemon or some other obvious (no timeout) problem

+	dbg("$tag fall back on dccproc");

     }

-

-    dbg("dcc: dccifd got response: %s", $response);

-

-  });

-

-  $permsgstatus->leave_helper_run_mode();

-

-  if ($timer->timed_out()) {

-    dbg("dcc: dccifd check timed out after $timeout secs.");

-    return;

   }

-  if ($err) {

-    chomp $err;

-    warn("dcc: dccifd -> check skipped: $err\n");

-    return;

-  }

+    if ($self->{dccifd_available}) {

-  if (!defined $response || $response !~ /^X-DCC/) {

-    dbg("dcc: dccifd check failed - no X-DCC returned: %s", $response);

-    return;

+      # send the options and other parameters to the daemon

+      $client = $envelope->{ip};

+      $clientname = $envelope->{rdns};

+      if (!defined $client) {

+	$client = '';

+      } else {

+	$client .= ("\r" . $clientname) if defined $clientname;

   }

+      $helo = $envelope->{helo} || '';

+      if ($tag ne "dcc:") {

+	$opts = $self->{dccifd_report_options}

+      } else {

+	$opts = $self->{dccifd_lookup_options};

+	# only query if there is an X-DCC header

+	$opts =~ s/grey-off/& query/ if defined $permsgstatus->{dcc_raw_x_dcc};

+      }

+      $sock->print($opts)	   or die "failed write options\n";

+      $sock->print($client . "\n") or die "failed write SMTP client\n";

+      $sock->print($helo . "\n")   or die "failed write HELO value\n";

+      $sock->print("\n")	   or die "failed write sender\n";

+      $sock->print("unknown\n\n")  or die "failed write 1 recipient\n";

+      $sock->print($$fulltext)     or die "failed write mail message\n";

+      $sock->shutdown(1) or die "failed socket shutdown: $!";

-  $response =~ s/[ \t]\z//;  # strip trailing whitespace

-  $permsgstatus->{dcc_response} = $response;

-}

+      $sock->getline()   or die "failed read status\n";

+      $sock->getline()   or die "failed read multistatus\n";

-sub dccproc_lookup {

-  my ($self, $permsgstatus, $fulltext, $client) = @_;

-  my $conf = $self->{main}->{conf};

-  my $response;

-  my %count = (body => 0, fuz1 => 0, fuz2 => 0, rep => 0);

-  my $timeout = $conf->{dcc_timeout};

+      @resp = $sock->getlines();

+      die "failed to read dccifd response\n" if !@resp;

-  $permsgstatus->enter_helper_run_mode();

-

-  # use a temp file here -- open2() is unreliable, buffering-wise, under spamd

+    } else {

+      $pgm = 'dccproc';

+      # use a temp file -- open2() is unreliable, buffering-wise, under spamd

+      # first ensure that we do not hit a stray file from some other filter.

+      $permsgstatus->delete_fulltext_tmpfile();

   my $tmpf = $permsgstatus->create_fulltext_tmpfile($fulltext);

-  my $pid;

-

-  my $timer = Mail::SpamAssassin::Timeout->new(

-           { secs => $timeout, deadline => $permsgstatus->{master_deadline} });

-  my $err = $timer->run_and_catch(sub {

-

-    local $SIG{PIPE} = sub { die "__brokenpipe__ignore__\n" };

-    # note: not really tainted, this came from system configuration file

-    my $path = untaint_file_path($conf->{dcc_path});

-

-    my $opts = $conf->{dcc_options};

+      my $path = $conf->{dcc_path};

+      $opts = $conf->{dcc_options};

     my @opts = !defined $opts ? () : split(' ',$opts);

     untaint_var(\@opts);

+      unshift(@opts, '-w', 'whiteclnt');

+      $client = $envelope->{ip};

+      if ($client) {

+	unshift(@opts, '-a', untaint_var($client));

+      } else {

+	# get external relay IP address from Received: header if not available

+	unshift(@opts, '-R');

+      }

+      if ($tag eq "dcc:") {

+	# query instead of report if there is an X-DCC header from upstream

+	unshift(@opts, '-Q') if defined $permsgstatus->{dcc_raw_x_dcc};

+      } else {

+	# learn or report spam

+	unshift(@opts, '-t', 'many');

+      }

-    unshift(@opts, "-a",

-            untaint_var($client))  if defined $client && $client ne '';

-

-    dbg("dcc: opening pipe: %s",

-         join(' ', $path, "-H", "-x", "0", @opts, "< $tmpf"));

+      dbg("$tag opening pipe to %s",

+	  join(' ', $path, "-C", "-x", "0", @opts, "<$tmpf"));

     $pid = Mail::SpamAssassin::Util::helper_app_pipe_open(*DCC,

-             $tmpf, 1, $path, "-H", "-x", "0", @opts);

+		$tmpf, 1, $path, "-C", "-x", "0", @opts);

     $pid or die "$!\n";

     # read+split avoids a Perl I/O bug (Bug 5985)

     my($inbuf,$nread,$resp); $resp = '';

     while ( $nread=read(DCC,$inbuf,8192) ) { $resp .= $inbuf }

     defined $nread  or die "error reading from pipe: $!";

-    my @null = split(/^/m, $resp, -1);  undef $resp;

+      @resp = split(/^/m, $resp, -1);  undef $resp;

     my $errno = 0;  close DCC or $errno = $!;

     proc_status_ok($?,$errno)

-      or info("dcc: [%s] finished: %s", $pid, exit_status_str($?,$errno));

-

-    if (!@null) {

-      # no facility prefix on this

-      die "failed to read header\n";

-    }

+	  or info("$tag [%s] finished: %s", $pid, exit_status_str($?,$errno));

-    # the first line will be the header we want to look at

-    chomp($response = shift @null);

-    # but newer versions of DCC fold the header if it's too long...

-    while (my $v = shift @null) {

-      last unless ($v =~ s/^\s+/ /);  # if this line wasn't folded, stop

-      chomp $v;

-      $response .= $v;

+      die "failed to read X-DCC header from dccproc\n" if !@resp;

     }

-

-    unless (defined($response)) {

-      # no facility prefix on this

-      die "no response\n";	# yes, this is possible

-    }

-

-    dbg("dcc: got response: %s", $response);

-

   });

+  if ($pgm eq 'dccproc') {

   if (defined(fileno(*DCC))) {  # still open

     if ($pid) {

-      if (kill('TERM',$pid)) { dbg("dcc: killed stale helper [$pid]") }

-      else { dbg("dcc: killing helper application [$pid] failed: $!") }

+	if (kill('TERM',$pid)) {

+	  dbg("$tag killed stale dccproc process [$pid]")

+	} else {

+	  dbg("$tag killing dccproc process [$pid] failed: $!")

+	}

     }

     my $errno = 0;  close(DCC) or $errno = $!;

-    proc_status_ok($?,$errno)

-      or info("dcc: [%s] terminated: %s", $pid, exit_status_str($?,$errno));

+      proc_status_ok($?,$errno) or info("$tag [%s] dccproc terminated: %s",

+					$pid, exit_status_str($?,$errno));

+    }

   }

+

   $permsgstatus->leave_helper_run_mode();

   if ($timer->timed_out()) {

@@ -833,204 +904,182 @@

   $permsgstatus->leave_helper_run_mode();

   if ($timer->timed_out()) {

-    dbg("dcc: check timed out after $timeout seconds");

-    return;

+    dbg("$tag $pgm timed out after $timeout seconds");

+    return (undef, undef);

   }

   if ($err) {

     chomp $err;

-    if ($err eq "__brokenpipe__ignore__") {

-      dbg("dcc: check failed: broken pipe");

-    } elsif ($err eq "no response") {

-      dbg("dcc: check failed: no response");

-    } else {

-      warn("dcc: check failed: $err\n");

-    }

-    return;

+    info("$tag $pgm failed: $err\n");

+    return (undef, undef);

   }

-  if (!defined($response) || $response !~ /^X-DCC/) {

-    $response ||= '';

-    dbg("dcc: check failed: no X-DCC returned (did you create a map file?): %s", $response);

-    return;

+  my ($raw_x_dcc, $cksums) = $self->parse_dcc_response(\@resp);

+  if (!defined $raw_x_dcc || $raw_x_dcc !~ /^X-DCC/) {

+    info("$tag instead of X-DCC header, $pgm returned '%s'", $raw_x_dcc);

+    return (undef, undef);

   }

-

-  $permsgstatus->{dcc_response} = $response;

+  dbg("$tag %s responded with '%s'", $pgm, $raw_x_dcc);

+  return ($raw_x_dcc, $cksums);

 }

-# only supports dccproc right now

-sub plugin_report {

+# tell DCC server that the message is spam according to SpamAssassin

+sub check_post_learn {

   my ($self, $options) = @_;

-  return if $options->{report}->{options}->{dont_report_to_dcc};

-  $self->get_dcc_interface();

-  return if $self->{dcc_disabled};

-

-  # get the metadata from the message so we can pass the external relay information

-  $options->{msg}->extract_message_metadata($options->{report}->{main});

-  my $client = $options->{msg}->{metadata}->{relays_external}->[0]->{ip};

-  if ($self->{dccifd_available}) {

-    my $clientname = $options->{msg}->{metadata}->{relays_external}->[0]->{rdns};

-    my $helo = $options->{msg}->{metadata}->{relays_external}->[0]->{helo} || "";

-    if ($client) {

-      if ($clientname) {

-        $client = $client . "\r" . $clientname;

-      }

-    } else {

-      $client = "0.0.0.0";

-    }

-    if ($self->dccifd_report($options, $options->{text}, $client, $helo)) {

-      $options->{report}->{report_available} = 1;

-      info("reporter: spam reported to DCC");

-      $options->{report}->{report_return} = 1;

+  # learn only if allowed

+  return if $self->{learn_disabled};

+  my $conf = $self->{main}->{conf};

+  if (!$conf->{use_dcc}) {

+    $self->{learn_disabled} = 1;

+    return;

     }

-    else {

-      info("reporter: could not report spam to DCC via dccifd");

+  my $learn_score = $conf->{dcc_learn_score};

+  if (!defined $learn_score || $learn_score eq '') {

+    dbg("dcc: DCC learning not enabled by dcc_learn_score");

+    $self->{learn_disabled} = 1;

+    return;

     }

-  } else {

-    # use temporary file: open2() is unreliable due to buffering under spamd

-    my $tmpf = $options->{report}->create_fulltext_tmpfile($options->{text});

-    if ($self->dcc_report($options, $tmpf, $client)) {

-      $options->{report}->{report_available} = 1;

-      info("reporter: spam reported to DCC");

-      $options->{report}->{report_return} = 1;

+  # and if SpamAssassin concluded that the message is spam

+  # worse than our threshold

+  my $permsgstatus = $options->{permsgstatus};

+  if ($permsgstatus->is_spam()) {

+    my $score = $permsgstatus->get_score();

+    my $required_score = $permsgstatus->get_required_score();

+    if ($score < $required_score + $learn_score) {

+      dbg("dcc: score=%d required_score=%d dcc_learn_score=%d",

+	  $score, $required_score, $learn_score);

+      return;

     }

-    else {

-      info("reporter: could not report spam to DCC via dccproc");

     }

-    $options->{report}->delete_fulltext_tmpfile();

+

+  # and if we checked the message

+  return if (!defined $permsgstatus->{dcc_raw_x_dcc});

+

+  # and if the DCC server thinks it was not spam

+  if ($permsgstatus->{dcc_raw_x_dcc} !~ /\b(Body|Fuz1|Fuz2)=\d/) {

+    dbg("dcc: already known as spam; no need to learn");

+    return;

   }

+

+  # dccsight is faster than dccifd or dccproc if we have checksums,

+  #   which we do not have with dccifd before 1.3.123

+  my $old_cksums = $permsgstatus->{dcc_cksums};

+  return if ($old_cksums && $self->dccsight_learn($permsgstatus, $old_cksums));

+

+  # Fall back on dccifd or dccproc without saved checksums or dccsight.

+  # get_dcc_interface() was called when the message was checked

+

+  # is getting the full text this way kosher?  Is get_pristine() public?

+  my $fulltext = $permsgstatus->{msg}->get_pristine();

+  my $envelope = $permsgstatus->{relays_external}->[0];

+  my ($raw_x_dcc, $cksums) = $self->ask_dcc("dcc: learn:", $permsgstatus,

+					    \$fulltext, $envelope);

+  dbg("dcc: learned as spam") if defined $raw_x_dcc;

 }

-sub dccifd_report {

-  my ($self, $options, $fulltext, $client, $helo) = @_;

-  my $conf = $self->{main}->{conf};

-  my $timeout = $conf->{dcc_timeout};

-  # instead of header use whatever the report option is

-  my $opts = $conf->{dccifd_options};

-  my @opts = !defined $opts ? () : split(' ',$opts);

+sub dccsight_learn {

+  my ($self, $permsgstatus, $old_cksums) = @_;

+  my ($raw_x_dcc, $new_cksums);

+

+  return 0 if !$old_cksums;

+

+  my $dccsight = $self->dcc_pgm_path('dccsight');

+  if (!$dccsight) {

+    info("dcc: cannot find dccsight") if $dccsight eq '';

+    return 0;

+  }

-  $options->{report}->enter_helper_run_mode();

-  my $timer = Mail::SpamAssassin::Timeout->new({ secs => $timeout });

+  $permsgstatus->enter_helper_run_mode();

-  my $err = $timer->run_and_catch(sub {

+  # use a temp file here -- open2() is unreliable, buffering-wise, under spamd

+  # ensure that we do not hit a stray file from some other filter.

+  $permsgstatus->delete_fulltext_tmpfile();

+  my $tmpf = $permsgstatus->create_fulltext_tmpfile(\$old_cksums);

+  my $pid;

+  my $timeout = $self->{main}->{conf}->{dcc_timeout};

+  my $timer = Mail::SpamAssassin::Timeout->new(

+	   { secs => $timeout, deadline => $permsgstatus->{master_deadline} });

+  my $err = $timer->run_and_catch(sub {

     local $SIG{PIPE} = sub { die "__brokenpipe__ignore__\n" };

-    my $sock = $self->dccifd_connect();

-    $sock or die "report: failed to connect to a dccifd socket";

+    dbg("dcc: opening pipe to %s",

+	join(' ', $dccsight, "-t", "many", "<$tmpf"));

-    # send the options and other parameters to the daemon

-    $sock->print("spam " . join(" ",@opts) . "\n")

-      or die "report: dccifd failed write"; # options

-    $sock->print($client . "\n")

-      or die "report: dccifd failed write"; # client

-    $sock->print($helo . "\n")

-      or die "report: dccifd failed write"; # HELO value

-    $sock->print("\n")

-      or die "report: dccifd failed write"; # sender

-    $sock->print("unknown\r\n")

-      or die "report: dccifd failed write"; # recipients

-    $sock->print("\n")

-      or die "report: dccifd failed write"; # recipients

+    $pid = Mail::SpamAssassin::Util::helper_app_pipe_open(*DCC,

+	    $tmpf, 1, $dccsight, "-t", "many");

+    $pid or die "$!\n";

-    $sock->print($$fulltext) or die "report: dccifd failed write";

+    # read+split avoids a Perl I/O bug (Bug 5985)

+    my($inbuf,$nread,$resp); $resp = '';

+    while ( $nread=read(DCC,$inbuf,8192) ) { $resp .= $inbuf }

+    defined $nread  or die "error reading from pipe: $!";

+    my @resp = split(/^/m, $resp, -1);  undef $resp;

-    $sock->shutdown(1) or die "report: dccifd failed socket shutdown: $!";

+    my $errno = 0;  close DCC or $errno = $!;

+    proc_status_ok($?,$errno)

+	  or info("dcc: [%s] finished: %s", $pid, exit_status_str($?,$errno));

-    $sock->getline() or die "report: dccifd failed read status";

-    $sock->getline() or die "report: dccifd failed read multistatus";

+    die "dcc: failed to read learning response\n" if !@resp;

-    my @ignored = $sock->getlines();

+    ($raw_x_dcc, $new_cksums) = $self->parse_dcc_response(\@resp);

   });

-  $options->{report}->leave_helper_run_mode();

+  if (defined(fileno(*DCC))) {	  # still open

+    if ($pid) {

+      if (kill('TERM',$pid)) {

+	dbg("dcc: killed stale dccsight process [$pid]")

+      } else {

+	dbg("dcc: killing stale dccsight process [$pid] failed: $!") }

+    }

+    my $errno = 0;  close(DCC) or $errno = $!;

+    proc_status_ok($?,$errno) or info("dcc: dccsight [%s] terminated: %s",

+				      $pid, exit_status_str($?,$errno));

+  }

+  $permsgstatus->delete_fulltext_tmpfile();

+  $permsgstatus->leave_helper_run_mode();

   if ($timer->timed_out()) {

-    dbg("reporter: DCC report via dccifd timed out after $timeout secs.");

+    dbg("dcc: dccsight timed out after $timeout seconds");

     return 0;

   }

   if ($err) {

     chomp $err;

-    if ($err eq "__brokenpipe__ignore__") {

-      dbg("reporter: DCC report via dccifd failed: broken pipe");

-    } else {

-      warn("reporter: DCC report via dccifd failed: $err\n");

-    }

+    info("dcc: dccsight failed: $err\n");

     return 0;

   }

+  if ($raw_x_dcc) {

+    dbg("dcc: learned response: %s", $raw_x_dcc);

   return 1;

-}

-

-sub dcc_report {

-  my ($self, $options, $tmpf, $client) = @_;

-  my $conf = $self->{main}->{conf};

-  my $timeout = $options->{report}->{conf}->{dcc_timeout};

-

-  # note: not really tainted, this came from system configuration file

-  my $path = untaint_file_path($options->{report}->{conf}->{dcc_path});

-  my $opts = $conf->{dcc_options};

-  my @opts = !defined $opts ? () : split(' ',$opts);

-  untaint_var(\@opts);

-

-  # get the metadata from the message so we can pass the external relay info

-

-  unshift(@opts, "-a",

-          untaint_var($client))  if defined $client && $client ne '';

-

-  my $timer = Mail::SpamAssassin::Timeout->new({ secs => $timeout });

-

-  $options->{report}->enter_helper_run_mode();

-  my $err = $timer->run_and_catch(sub {

-

-    local $SIG{PIPE} = sub { die "__brokenpipe__ignore__\n" };

-

-    dbg("report: opening pipe: %s",

-        join(' ', $path, "-H", "-t", "many", "-x", "0", @opts, "< $tmpf"));

-

-    my $pid = Mail::SpamAssassin::Util::helper_app_pipe_open(*DCC,

-                $tmpf, 1, $path, "-H", "-t", "many", "-x", "0", @opts);

-    $pid or die "$!\n";

+  }

-    my($inbuf,$nread,$nread_all); $nread_all = 0;

-    # response is ignored, just check its existence

-    while ( $nread=read(DCC,$inbuf,8192) ) { $nread_all += $nread }

-    defined $nread  or die "error reading from pipe: $!";

+  return 0;

+}

-    dbg("dcc: empty response")  if $nread_all < 1;

+sub plugin_report {

+  my ($self, $options) = @_;

-    my $errno = 0;  close DCC or $errno = $!;

-    # closing a pipe also waits for the process executing on the pipe to

-    # complete, no need to explicitly call waitpid

-    # my $child_stat = waitpid($pid,0) > 0 ? $? : undef;

-    proc_status_ok($?,$errno)

-      or die "dcc: reporter error: ".exit_status_str($?,$errno)."\n";

-  });

-  $options->{report}->leave_helper_run_mode();

+  return if $options->{report}->{options}->{dont_report_to_dcc};

+  $self->get_dcc_interface();

+  return if $self->{dcc_disabled};

-  if ($timer->timed_out()) {

-    dbg("reporter: DCC report via dccproc timed out after $timeout seconds");

-    return 0;

-  }

+  # get the metadata from the message so we can report the external relay

+  $options->{msg}->extract_message_metadata($options->{report}->{main});

+  my $envelope = $options->{msg}->{metadata}->{relays_external}->[0];

+  my ($raw_x_dcc, $cksums) = $self->ask_dcc("reporter:", $options->{report},

+					    $options->{text}, $envelope);

-  if ($err) {

-    chomp $err;

-    if ($err eq "__brokenpipe__ignore__") {

-      dbg("reporter: DCC report via dccproc failed: broken pipe");

+  if (defined $raw_x_dcc) {

+    $options->{report}->{report_available} = 1;

+    info("reporter: spam reported to DCC");

+    $options->{report}->{report_return} = 1;

     } else {

-      warn("reporter: DCC report via dccproc failed: $err\n");

+    info("reporter: could not report spam to DCC");

     }

-    return 0;

-  }

-

-  return 1;

 }

 1;

-

-=back

-

-=cut

--- lib/Mail/SpamAssassin/Logger/Syslog.pm	2012/05/14 16:28:23	1338277

+++ lib/Mail/SpamAssassin/Logger/Syslog.pm	2012/05/14 16:31:09	1338278

@@ -167,17 +167,21 @@

   }

   $msg = $timestamp . ' ' . $msg  if $timestamp ne '';

-  # important: do not call syslog() from the SIGCHLD handler

-  # child_handler().   otherwise we can get into a loop if syslog()

-  # forks a process -- as it does in syslog-ng apparently! (bug 3625)

-  $Mail::SpamAssassin::Logger::LOG_SA{INHIBIT_LOGGING_IN_SIGCHLD_HANDLER} = 1;

+# no longer needed since a patch to bug 6745:

+# # important: do not call syslog() from the SIGCHLD handler

+# # child_handler().   otherwise we can get into a loop if syslog()

+# # forks a process -- as it does in syslog-ng apparently! (bug 3625)

+# $Mail::SpamAssassin::Logger::LOG_SA{INHIBIT_LOGGING_IN_SIGCHLD_HANDLER} = 1;

+

   my $eval_stat;

   eval {

     syslog($level, "%s", $msg); 1;

   } or do {

     $eval_stat = $@ ne '' ? $@ : "errno=$!";  chomp $eval_stat;

   };

-  $Mail::SpamAssassin::Logger::LOG_SA{INHIBIT_LOGGING_IN_SIGCHLD_HANDLER} = 0;

+

+# no longer needed since a patch to bug 6745:

+# $Mail::SpamAssassin::Logger::LOG_SA{INHIBIT_LOGGING_IN_SIGCHLD_HANDLER} = 0;

   if (defined $eval_stat) {

     if ($self->check_syslog_sigpipe($msg)) {

--- spamd/spamd.raw	2012/05/14 16:28:23	1338277

+++ spamd/spamd.raw	2012/05/14 16:31:09	1338278

@@ -589,6 +589,7 @@

 my $timeout_child;        # processing timeout (headers->finish), 0=no timeout

 my $clients_per_child;    # number of clients each child should process

 my %children;             # current children

+my @children_exited;

 if ( defined $opt{'max-children'} ) {

   $childlimit = $opt{'max-children'};

@@ -1033,6 +1034,8 @@

 # child_handler()  if !$scaling || am_running_on_windows();

   child_handler();  # it doesn't hurt to call child_handler unconditionally

+  child_cleaner();

+

   do_sighup_restart()  if defined $got_sighup;

   for (my $i = keys %children; $i < $childlimit; $i++) {

@@ -2523,7 +2526,8 @@

   my ($sig) = @_;

   # do NOT call syslog here unless the child's pid is in our list of known

-  # children.  This is due to syslog-ng brokenness -- bugs 3625, 4237.

+  # children.  This is due to syslog-ng brokenness -- bugs 3625, 4237;

+  # see also bug 6745.

   # clean up any children which have exited

   for (;;) {

@@ -2534,12 +2538,23 @@

     #

     my $pid = waitpid(-1, WNOHANG);

     last if !$pid || $pid == -1;

-    my $child_stat = $?;

+    push(@children_exited, [$pid, $?, $sig, time]);

+  }

-    if (!defined $children{$pid}) {

-      # ignore this child; we didn't realise we'd forked it. bug 4237

-      next;

-    }

+  $SIG{CHLD} = \&child_handler;    # reset as necessary, should be at end

+}

+

+# takes care of dead children, as noted by a child_handler()

+# called in a main program flow (not from a signal handler)

+#

+sub child_cleaner {

+  while (@children_exited) {

+    my $tuple = shift(@children_exited);

+    next if !$tuple;  # just in case

+    my($pid, $child_stat, $sig, $timestamp) = @$tuple;

+

+    # ignore this child if we didn't realise we'd forked it. bug 4237

+    next if !defined $children{$pid};

     # remove them from our child listing

     delete $children{$pid};

@@ -2550,15 +2565,10 @@

       my $sock = $backchannel->get_socket_for_child($pid);

       if ($sock) { $sock->close(); }

     }

-

-    unless ($Mail::SpamAssassin::Logger::LOG_SA{INHIBIT_LOGGING_IN_SIGCHLD_HANDLER}) {

-      info("spamd: handled cleanup of child pid [%s]%s: %s",

-           $pid, (defined $sig ? " due to SIG$sig" : ""),

-           exit_status_str($child_stat,0));

-    }

+    info("spamd: handled cleanup of child pid [%s]%s: %s",

+         $pid, (defined $sig ? " due to SIG$sig" : ""),

+         exit_status_str($child_stat,0));

   }

-

-  $SIG{CHLD} = \&child_handler;    # reset as necessary, should be at end

 }

 sub restart_handler {

@unexec rm -rf /var/spool/spamd