[SACVS] CVS: spamassassin/lib/Mail/SpamAssassin Conf.pm,1.258,1.259 EvalTests.pm,1.448,1.449 PerMsgStatus.pm,1.354,1.355

Update of /cvsroot/spamassassin/spamassassin/lib/Mail/SpamAssassin
In directory sc8-pr-cvs1:/tmp/cvs-serv18462/lib/Mail/SpamAssassin

Modified Files:
        Conf.pm EvalTests.pm PerMsgStatus.pm 
Log Message:
bug 2459: reorganize auto-whitelist implementation


Index: Conf.pm
===================================================================
RCS file: /cvsroot/spamassassin/spamassassin/lib/Mail/SpamAssassin/Conf.pm,v
retrieving revision 1.258
retrieving revision 1.259
diff -b -w -u -d -r1.258 -r1.259
--- Conf.pm     21 Sep 2003 22:17:55 -0000      1.258
+++ Conf.pm     23 Sep 2003 21:37:02 -0000      1.259
@@ -163,6 +163,8 @@
   # Mail::SpamAssassin module users who set that configuration setting,
   # to receive the correct values.
 
+  $self->{use_auto_whitelist} = 1;
+  $self->{auto_whitelist_factory} = "Mail::SpamAssassin::DBBasedAddrList";
   $self->{auto_whitelist_path} = "__userstate__/auto-whitelist";
   $self->{auto_whitelist_file_mode} = '0700';
   $self->{auto_whitelist_factor} = 0.5;
@@ -2284,6 +2286,35 @@
 
     if (/^dcc_options\s+([A-Z -]+)/) {
       $self->{dcc_options} = $1; next;
+    }
+
+=item use_auto_whitelist ( 0 | 1 )             (default: 1)
+
+Whether to use auto-whitelists.  Auto-whitelists track the long-term
+average score for each sender and then shift the score of new messages
+toward that long-term average.  This can increase or decrease the score
+for messages, depending on the long-term behavior of the particular
+correspondent.
+
+For more information about the auto-whitelist system, please look at the
+the C<Automatic Whitelist System> section of the README file.  The
+auto-whitelist is not intended as a general-purpose replacement for static
+whitelist entries added to your config files.
+
+=cut
+
+    if (/^use_auto_whitelist\s+(\d+)$/) {
+      $self->{use_auto_whitelist} = $1; next;
+    }
+
+=item auto_whitelist_factory module (default: 
Mail::SpamAssassin::DBBasedAddrList)
+
+Select alternative whitelist factory module.
+
+=cut
+
+    if (/^auto_whitelist_factory\s+(.*)$/) {
+      $self->{auto_whitelist_factory} = $1; next;
     }
 
 =item auto_whitelist_path /path/to/file        (default: 
~/.spamassassin/auto-whitelist)

Index: EvalTests.pm
===================================================================
RCS file: 
/cvsroot/spamassassin/spamassassin/lib/Mail/SpamAssassin/EvalTests.pm,v
retrieving revision 1.448
retrieving revision 1.449
diff -b -w -u -d -r1.448 -r1.449
--- EvalTests.pm        21 Sep 2003 22:34:18 -0000      1.448
+++ EvalTests.pm        23 Sep 2003 21:37:02 -0000      1.449
@@ -916,6 +916,95 @@
 
 ###########################################################################
 
+sub check_from_in_auto_whitelist {
+    my ($self) = @_;
+
+    return unless $self->{main}->{conf}->{use_auto_whitelist};
+
+    if (!defined $self->{main}->{pers_addr_list_factory}) {
+      my $factory;
+      return unless $self->{main}->{conf}->{auto_whitelist_factory};
+      my $type = $self->{main}->{conf}->{auto_whitelist_factory};
+      if ($type =~ /^([_A-Za-z0-9:]+)$/) {
+       $type = $1;
+      }
+      else {
+       warn "illegal auto_whitelist_factory setting\n";
+       return;
+      }
+      eval '
+        require '.$type.';
+        $factory = '.$type.'->new();
+      ';
+      if ($@) { warn $@; undef $factory; }
+      $self->{main}->set_persistent_address_list_factory($factory);
+    }
+
+    local $_ = lc $self->get('From:addr');
+    return 0 unless /\S/;
+
+    # find the earliest usable "originating IP".  ignore reserved nets
+    my $origip;
+    foreach my $rly (reverse (@{$self->{relays_trusted}}, 
@{$self->{relays_untrusted}}))
+    {
+      next if ($rly->{ip_is_reserved});
+      if ($rly->{ip}) {
+       $origip = $rly->{ip}; last;
+      }
+    }
+
+    # Create the AWL object, catching 'die's
+    my $whitelist;
+    my $evalok = eval {
+      $whitelist = Mail::SpamAssassin::AutoWhitelist->new($self->{main});
+
+      # check
+      my $meanscore = $whitelist->check_address($_, $origip);
+      my $delta = 0;
+
+      dbg("AWL active, pre-score: " . $self->{hits} . ", mean: " .
+         ($meanscore || 'undef') . ", IP: " . ($origip || 'undef'));
+
+      if (defined ($meanscore)) {
+        $delta = ($meanscore - $self->{hits}) * 
$self->{main}->{conf}->{auto_whitelist_factor};
+       $self->{tag_data}->{AWL} = sprintf("%2.1f",$delta);
+       # Save this for _AWL_ tag
+      }
+
+      # Update the AWL *before* adding the new score, otherwise
+      # early high-scoring messages are reinforced compared to
+      # later ones.  http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=159704
+      if (!$self->{disable_auto_learning}) {
+        $whitelist->add_score($self->{hits});
+      }
+
+      # current AWL score changes with each hit
+      for my $set (0..3) {
+       $self->{conf}->{scoreset}->[$set]->{"AWL"} = sprintf("%0.3f", $delta);
+      }
+
+      if ($delta != 0) {
+        $self->_handle_hit("AWL", $delta, "AWL: ",
+                          $self->{main}->{conf}->{descriptions}->{AWL});
+      }
+
+      dbg("Post AWL score: ".$self->{hits});
+      $whitelist->finish();
+      1;
+    };
+
+    if (!$evalok) {
+      dbg ("open of AWL file failed: $@");
+      # try an unlock, in case we got that far
+      eval { $whitelist->finish(); };
+    }
+
+    # test hit is above
+    return 0;
+}
+
+###########################################################################
+
 sub _check_whitelist_rcvd {
   my ($self, $list, $addr) = @_;
 

Index: PerMsgStatus.pm
===================================================================
RCS file: 
/cvsroot/spamassassin/spamassassin/lib/Mail/SpamAssassin/PerMsgStatus.pm,v
retrieving revision 1.354
retrieving revision 1.355
diff -b -w -u -d -r1.354 -r1.355
--- PerMsgStatus.pm     19 Sep 2003 04:02:40 -0000      1.354
+++ PerMsgStatus.pm     23 Sep 2003 21:37:02 -0000      1.355
@@ -182,12 +182,8 @@
     # auto-learning
     $self->learn();
 
-    # add points from Bayes, before adjusting the AWL
+    # add points from learning systems (Bayes and AWL)
     $self->{hits} += $self->{learned_hits};
-
-    # Do AWL tests last, since these need the score to have already been
-    # calculated
-    $self->do_awl_tests();
   }
 
   $self->delete_fulltext_tmpfile();
@@ -262,7 +258,7 @@
     foreach my $test ( @{$self->{test_names_hit}} ) {
       # ignore tests with 0 score in this scoreset or if the test is a 
learning or userconf test
       next if ( $self->{conf}->{scores}->{$test} == 0 );
-      next if ( exists $self->{conf}->{tflags}->{$test} && 
$self->{conf}->{tflags}->{$test} =~ /\b(?:learn|userconf)\b/ );
+      next if ( exists $self->{conf}->{tflags}->{$test} && 
$self->{conf}->{tflags}->{$test} =~ /\bnoautolearn\b/ );
 
       $hits += $self->{conf}->{scores}->{$test};
     }
@@ -2001,75 +1997,6 @@
 
 ###########################################################################
 
-sub do_awl_tests {
-    my($self) = @_;
-
-    return unless (defined $self->{main}->{pers_addr_list_factory});
-
-    local $_ = lc $self->get('From:addr');
-    return 0 unless /\S/;
-
-    # find the earliest usable "originating IP".  ignore reserved nets
-    my $origip;
-    foreach my $rly (reverse (@{$self->{relays_trusted}}, 
@{$self->{relays_untrusted}}))
-    {
-      next if ($rly->{ip_is_reserved});
-      if ($rly->{ip}) {
-       $origip = $rly->{ip}; last;
-      }
-    }
-
-    # Create the AWL object, catching 'die's
-    my $whitelist;
-    my $evalok = eval {
-      $whitelist = Mail::SpamAssassin::AutoWhitelist->new($self->{main});
-
-      # check
-      my $meanscore = $whitelist->check_address($_, $origip);
-      my $delta = 0;
-
-      dbg("AWL active, pre-score: ".$self->{hits}.", mean: 
".($meanscore||'undef').
-                          ", originating-ip: ".($origip||'undef'));
-
-      if(defined($meanscore))
-      {
-          $delta = ($meanscore - $self->{hits}) * 
$self->{main}->{conf}->{auto_whitelist_factor};
-         $self->{tag_data}->{AWL} = sprintf("%2.1f",$delta);
-         # Save this for _AWL_ tag
-      }
-
-      # Update the AWL *before* adding the new score, otherwise
-      # early high-scoring messages are reinforced compared to
-      # later ones.  See
-      # http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=159704
-      #
-      if (!$self->{disable_auto_learning}) {
-        $whitelist->add_score($self->{hits});
-      }
-
-      # current AWL score changes with each hit
-      for my $set (0..3) {
-       $self->{conf}->{scoreset}->[$set]->{"AWL"} = sprintf("%0.3f", $delta);
-      }
-
-      if ($delta != 0) {
-       $self->_handle_hit("AWL",$delta,"AWL: ","Auto-whitelist adjustment");
-      }
-
-      dbg("Post AWL score: ".$self->{hits});
-      $whitelist->finish();
-      1;
-    };
-
-    if (!$evalok) {
-      dbg ("open of AWL file failed: $@");
-      # try an unlock, in case we got that far
-      eval { $whitelist->finish(); };
-    }
-}
-
-###########################################################################
-
 sub do_meta_tests {
   my ($self) = @_;
   local ($_);
@@ -2199,13 +2126,15 @@
     # Score of 0, skip it.
     next unless ($self->{conf}->{scores}->{$rulename});
 
-    # If the rule is a net rule, and we're in a non-net enabled scoreset, skip 
it.
+    # If the rule is a net rule, and we're in a non-net scoreset, skip it.
     next if (exists $self->{conf}->{tflags}->{$rulename} &&
-      (($scoreset & 1) == 0) && $self->{conf}->{tflags}->{$rulename} =~ 
/\bnet\b/);
+            (($scoreset & 1) == 0) &&
+            $self->{conf}->{tflags}->{$rulename} =~ /\bnet\b/);
 
-    # If the rule is a learn rule, and we're in a non-learn enabled scoreset, 
skip it.
+    # If the rule is a bayes rule, and we're in a non-bayes scoreset, skip it.
     next if (exists $self->{conf}->{tflags}->{$rulename} &&
-      (($scoreset & 2) == 0) && $self->{conf}->{tflags}->{$rulename} =~ 
/\blearn\b/);
+            (($scoreset & 2) == 0) &&
+            $self->{conf}->{tflags}->{$rulename} =~ /\bbayes\b/);
 
     my $score = $self->{conf}{scores}{$rulename};
     my $result;
@@ -2309,9 +2238,8 @@
 
     my $tflags = $self->{conf}->{tflags}->{$rule}; $tflags ||= '';
 
-    # ignore 'learn' or 'userconf' rules, when considering score for
-    # Bayesian auto-learning
-    if ($tflags =~ /\b(?:learn|userconf)\b/i) {
+    # ignore 'noautolearn' rules when considering score for Bayes auto-learning
+    if ($tflags =~ /\bnoautolearn\b/i) {
       $self->{learned_hits} += $score;
     }
     else {



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf