svn commit: rev 6436 - incubator/spamassassin/trunk/lib/Mail/SpamAssassin

[fe...@apache.org]

Author: felicity
Date: Mon Feb  2 16:30:30 2004
New Revision: 6436

Modified:
   incubator/spamassassin/trunk/lib/Mail/SpamAssassin/EvalTests.pm
   incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgContainer.pm
Log:
don't need MsgContainer::as_string() anymore.  instead of rendering as HTML any text* attachment, only do 'text' and 'text/plain'.  fixes some FPs for MPART_ALT_DIFF, etc.

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/EvalTests.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/EvalTests.pm	(original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/EvalTests.pm	Mon Feb  2 16:30:30 2004
@@ -3605,11 +3605,13 @@
       if ( $type eq 'text/html' ) {
         foreach my $w ( grep(/\w/,split(/\s+/,$rnd)) ) {
           next if ( $w =~ /^URI:/ );
+	  #dbg("HTML: $w");
           $html{$w}++;
         }
       }
       else {
         foreach my $w ( grep(/\w/,split(/\s+/,$rnd)) ) {
+	  #dbg("TEXT: $w");
           $text{$w}++;
         }
       }
@@ -3621,6 +3623,8 @@
     while( my($k,$v) = each %text ) {
       delete $html{$k} if ( exists $html{$k} && $html{$k}-$text{$k} < 1 );
     }
+
+    #map { dbg("LEFT: $_") } keys %html;
 
     my $diff = scalar(keys %html)/$orig*100;
     $self->{madiff} = $diff if ( $diff > $self->{madiff} );

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgContainer.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgContainer.pm	(original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgContainer.pm	Mon Feb  2 16:30:30 2004
@@ -265,10 +265,12 @@
     my $text = $self->decode();
     my $raw = length($text);
 
-    # render text/html always, or any other text part as text/html based
+    # render text/html always, or any other text|text/plain part as text/html based
     # on a heuristic which simulates a certain common mail client
-    if ( $raw > 0 && ($self->{'type'} =~ m@^text/html\b@i ||
-        ($text =~ m/^(.{0,18}?<(?:$Mail::SpamAssassin::HTML::re_start)(?:\s.{0,18}?)?>)/ois &&
+    if ( $raw > 0 && (
+        $self->{'type'} =~ m@^text/html\b@i || (
+        $self->{'type'} =~ m@^text(?:$|/plain)@i &&
+	  $text =~ m/^(.{0,18}?<(?:$Mail::SpamAssassin::HTML::re_start)(?:\s.{0,18}?)?>)/ois &&
 	  _html_near_start($1))
         )
        ) {
@@ -491,15 +493,6 @@
 sub get_pristine_body {
   my ($self) = @_;
   return $self->{pristine_body};
-}
-
-=item as_string()
-
-=cut
-
-sub as_string {
-  my ($self) = @_;
-  return $self->get_all_headers(1) . "\n" . $self->{pristine_body};
 }
 
 =item ignore()