Changeset 65

Timestamp:

28.11.2006 21:01:04 (2 years ago)

Author:

decoder

Message:

New function focr_autosort_scanset: Enables the plugin to automatically adapt the order of the scansets to the current mail traffic.
Currently effective scansets are sorted to the beginning of the scanset list, to save scanner passes.

Files:

• trunk/devel/FuzzyOcr.cf (modified) (1 diff)
• trunk/devel/FuzzyOcr.pm (modified) (1 diff)
• trunk/devel/FuzzyOcr/Config.pm (modified) (1 diff)
• trunk/devel/FuzzyOcr/Scanset.pm (modified) (1 diff)

trunk/devel/FuzzyOcr.cf

@@ -68 +68 @@
-## This saves resources, but lowers the scores because not the best, but the first best scanset is taken as result.
-#focr_minimal_scanset 0.0
+#
+## This option is only used when focr_minimal_scanset is enabled. Basically, this counts the effectiveness
+## of a scanset on the current mail traffic and resorts the scansets with the most effective first.
+## This saves unnecessary scanner passes and saves resources. The default is 1.
+#focr_autosort_scanset 1
+#
+## This is a parameter for the focr_autosort_scanset function, and specifies the maximum value of the effectiveness
+## counter used in each scanset. If you increase this, it will take longer until the autosort function adapts to new
+## types of spam, setting it too low will lower the effectiveness of the function. Default is 10.
+#focr_autosort_buffer 10
-#
-#########

trunk/devel/FuzzyOcr.pm

@@ -658 +658 @@
-            if ($mcnt >= $conf->{focr_counts_required} and $conf->{focr_minimal_scanset}) {
-                debuglog("Scanset \"$scanlabel\" generates enough hits ($mcnt), skipping further scansets...");
+                if ($conf->{focr_autosort_scanset}) {
+                    foreach my $s (@$scansets) {
+                        if ($s->{label} eq $scanlabel) {
+                            if ($s->{hit_counter} < $conf->{focr_autosort_buffer}) {
+                                $s->{hit_counter} = $s->{hit_counter} + 1;
+                            }
+                        } else {
+                            if ($s->{hit_counter} > 0) {
+                                $s->{hit_counter} = $s->{hit_counter} - 1;
+                            }
+                        }
+                    }
+                    debuglog("Resorting scanset list...");
+                    @$scansets = sort { $b->{hit_counter} <=> $a->{hit_counter} } @$scansets;
+                }
-                last;
-            }
-            #push( @ocr_results, [@result] );
-            #push( @used_scansets, $scanset );
-        }
-        if ($conf->{focr_enable_image_hashing}) {

trunk/devel/FuzzyOcr/Config.pm

@@ -381 +381 @@
-            setting => 'focr_minimal_scanset',
-            default => 0,
+            type => $Mail::SpamAssassin::Conf::CONF_TYPE_BOOL
+        });
+    push (@cmds, {
+            setting => 'focr_autosort_scanset',
+            default => 1,
+            type => $Mail::SpamAssassin::Conf::CONF_TYPE_BOOL
+        });
+    push (@cmds, {
+            setting => 'focr_autosort_buffer',
+            default => 10,
-            type => $Mail::SpamAssassin::Conf::CONF_TYPE_BOOL
-        });

trunk/devel/FuzzyOcr/Scanset.pm

@@ -11 +11 @@
-        "preprocessors" => $preprocessors,
-        "ocr_command"   => $ocr_command,
-
+
+
-    }, $class;
-}