You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@spamassassin.apache.org by qu...@apache.org on 2004/10/27 22:59:22 UTC
svn commit: rev 55734 - in spamassassin/trunk: lib/Mail/SpamAssassin rules
Author: quinlan
Date: Wed Oct 27 13:59:21 2004
New Revision: 55734
Modified:
spamassassin/trunk/lib/Mail/SpamAssassin/HTML.pm
spamassassin/trunk/rules/70_testing.cf
Log:
bug 3680: remove tests for empty anchors - very poor HTML ham results
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 13:59:21 2004
@@ -559,11 +559,6 @@
$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 $height = 0;
@@ -603,10 +598,6 @@
if ($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}++;
@@ -689,6 +680,7 @@
# text that is part of body and also stored separately
if (exists $self->{inside}{a} && $self->{inside}{a} > 0) {
+ # this doesn't worry about nested anchors
$self->{anchor}->[$self->{anchor_index}] .= $text;
}
if (exists $self->{inside}{title} && $self->{inside}{title} > 0) {
Modified: spamassassin/trunk/rules/70_testing.cf
==============================================================================
--- spamassassin/trunk/rules/70_testing.cf (original)
+++ spamassassin/trunk/rules/70_testing.cf Wed Oct 27 13:59:21 2004
@@ -191,10 +191,6 @@
endif # Mail::SpamAssassin::Plugin::URIDNSBL
# bug 3680: empty anchor, some others
-body T_HTML_LINK_EMPTY_1 eval:html_text_match('anchor', '^$')
-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