svn commit: rev 55677 - in spamassassin/trunk: lib/Mail/SpamAssassin rules

[qu...@apache.org]

Author: quinlan
Date: Wed Oct 27 00:56:04 2004
New Revision: 55677

Modified:
   spamassassin/trunk/lib/Mail/SpamAssassin/HTML.pm
   spamassassin/trunk/rules/70_testing.cf
Log:
bug 3680: empty anchor
URIBL test for Jeff
some other anchor rules


Modified: spamassassin/trunk/lib/Mail/SpamAssassin/HTML.pm
==============================================================================
--- spamassassin/trunk/lib/Mail/SpamAssassin/HTML.pm	(original)
+++ spamassassin/trunk/lib/Mail/SpamAssassin/HTML.pm	Wed Oct 27 00:56:04 2004
@@ -559,6 +559,11 @@
       $self->examine_text_style($1, $2);
     }
   }
+  # image inside of an anchor
+  # XXX - test code
+  if ($tag eq "img" && exists $self->{inside}{a} && $self->{inside}{a} > 0) {
+    $self->{anchor}->[$self->{anchor_index}] .= " <img> ";
+  }
   if ($tag eq "img" && exists $attr->{width} && exists $attr->{height}) {
     my $width = 0;
     my $height = 0;
@@ -598,6 +603,10 @@
   if ($tag eq "a") {
     $self->{anchor_index}++;
     $self->{anchor}->[$self->{anchor_index}] = "";
+    # XXX - test code
+    if (!exists $attr->{href} || $attr->{href} !~ /\S/) {
+      $self->put_results(t_anchor_empty_href => 1);
+    }
   }
   if ($tag eq "title") {
     $self->{title_index}++;

Modified: spamassassin/trunk/rules/70_testing.cf
==============================================================================
--- spamassassin/trunk/rules/70_testing.cf	(original)
+++ spamassassin/trunk/rules/70_testing.cf	Wed Oct 27 00:56:04 2004
@@ -181,3 +181,20 @@
 # try \b around a rule deleted at some point
 body T_OPTING_IN_CAPS		/\b(?-i:O)pt.?(?-i:I)n\b/i
 
+# testing for Jeff, do not use!!!
+ifplugin Mail::SpamAssassin::Plugin::URIDNSBL
+urirhsbl URIBL_MP fraud.rhs.mailpolice.com.   A
+body     URIBL_MP eval:check_uridnsbl('URIBL_MP')
+describe URIBL_MP URI's domain appears in MailPolice fraud list
+tflags   URIBL_MP net
+score    URIBL_MP 2.0
+endif	# Mail::SpamAssassin::Plugin::URIDNSBL
+
+# bug 3680: empty anchor, some others
+body T_HTML_LINK_EMPTY_1	eval:html_text_match('anchor', '^$')
+body T_HTML_LINK_EMPTY_2	eval:html_text_match('anchor', '^\s*$')
+body T_HTML_LINK_PUSH_HERE	eval:html_text_match('anchor', '(?i)(?:push|go|cl[1l]ck)\s*(?:here|this)')
+body T_HTML_LINK_OPT		eval:html_text_match('anchor', '(?i)opt.?(?:out|in)')
+body T_HTML_LINK_OPT_OUT	eval:html_text_match('anchor', '(?i)opt.?out')
+body T_HTML_LINK_EMPTY_HREF  eval:html_test('t_anchor_empty_href')
+describe T_HTML_LINK_EMPTY_HREF  HTML anchor has empty or missing HREF value