Mailing List Archive

Author: jm
Date: Thu Feb 26 20:42:47 2004
New Revision: 6895

Added:
incubator/spamassassin/trunk/rules/25_uribl.cf
Modified:
incubator/spamassassin/trunk/MANIFEST
incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Conf.pm
incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Dns.pm
incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgMetadata.pm
incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PerMsgStatus.pm
incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Plugin/Test.pm
incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PluginHandler.pm
incubator/spamassassin/trunk/masses/mass-check
Log:
bug 1375: do DNSBL lookups on URLs, implemented as plugin Mail::SpamAssassin::Plugin::URIDNSBL

Modified: incubator/spamassassin/trunk/MANIFEST
==============================================================================
--- incubator/spamassassin/trunk/MANIFEST (original)
+++ incubator/spamassassin/trunk/MANIFEST Thu Feb 26 20:42:47 2004
@@ -308,3 +308,4 @@
tools/test_extract
tools/triplets.pl
lib/Mail/SpamAssassin/Constants.pm
+rules/25_uribl.cf

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Conf.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Conf.pm (original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Conf.pm Thu Feb 26 20:42:47 2004
@@ -1935,17 +1935,6 @@
}

###########################################################################
-
- if ($self->{main}->call_plugins ("parse_config", {
- line => $_,
- user_config => $scoresonly
- }))
- {
- # a plugin dealt with it successfully.
- next;
- }
-
-###########################################################################
# SECURITY: no eval'd code should be loaded before this line.
#
if ($scoresonly && !$self->{allow_user_rules}) { goto failed_line; }
@@ -2756,6 +2745,22 @@
###########################################################################

failed_line:
+
+ # last ditch: try to see if the plugins know what to do with it
+ if ($self->{main}->call_plugins ("parse_config", {
+ key => $key,
+ value => $value,
+ line => $_,
+ conf => $self,
+ user_config => $scoresonly
+ }))
+ {
+ # a plugin dealt with it successfully.
+ next;
+ }
+
+###########################################################################
+
my $msg = "Failed to parse line in SpamAssassin configuration, ".
"skipping: $_";


Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Dns.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Dns.pm (original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Dns.pm Thu Feb 26 20:42:47 2004
@@ -250,6 +250,9 @@
push(@left, $query);
}
}
+
+ $self->{main}->call_plugins ("check_tick", { permsgstatus => $self });
+
last unless @left;
last if time >= $timeout;
@waiting = @left;
@@ -283,6 +286,7 @@
$self->{tag_data}->{RBL} .= "<$dnsuri>" .
" [" . join(", ", @{ $answers }) . "]\n";
}
+
chomp $self->{tag_data}->{RBL} if defined $self->{tag_data}->{RBL};
}


Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgMetadata.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgMetadata.pm (original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/MsgMetadata.pm Thu Feb 26 20:42:47 2004
@@ -49,6 +49,7 @@

package Mail::SpamAssassin::MsgMetadata;
use strict;
+use bytes;

use Mail::SpamAssassin;
use Mail::SpamAssassin::Received;

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PerMsgStatus.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PerMsgStatus.pm (original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PerMsgStatus.pm Thu Feb 26 20:42:47 2004
@@ -143,6 +143,8 @@
# do head tests
$self->do_head_tests();

+ $self->{main}->call_plugins ("check_tick", { permsgstatus => $self });
+
# do body tests with decoded portions
{
my $decoded = $self->get_decoded_stripped_body_text_array();
@@ -156,6 +158,7 @@
$self->do_body_eval_tests($decoded);
undef $decoded;
}
+ $self->{main}->call_plugins ("check_tick", { permsgstatus => $self });

# do rawbody tests with raw text portions
{
@@ -166,6 +169,7 @@
$self->do_body_uri_tests($bodytext);
undef $bodytext;
}
+ $self->{main}->call_plugins ("check_tick", { permsgstatus => $self });

# and do full tests: first with entire, full, undecoded message
# use get_all_headers instead of
@@ -184,6 +188,7 @@

# finish the DNS results
$self->rbl_finish();
+ $self->{main}->call_plugins ("check_post_dnsbl", { permsgstatus => $self });

# Do meta rules second-to-last
$self->do_meta_tests();
@@ -2020,6 +2025,19 @@
# the clearing of the test state is now inlined as:
#
# $self->{test_log_msgs} = (); # clear test state
+#
+# except for this public API for plugin use:
+
+=item $status->clear_test_state()
+
+Clear test state, including test log messages from C<$status->test_log()>.
+
+=cut
+
+sub clear_test_state {
+ my ($self) = @_;
+ $self->{test_log_msgs} = ();
+}

sub _handle_hit {
my ($self, $rule, $score, $area, $desc) = @_;

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Plugin/Test.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Plugin/Test.pm (original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Plugin/Test.pm Thu Feb 26 20:42:47 2004
@@ -10,6 +10,9 @@
package Mail::SpamAssassin::Plugin::Test;

use Mail::SpamAssassin::Plugin;
+use strict;
+use bytes;
+
use vars qw(@ISA);
@ISA = qw(Mail::SpamAssassin::Plugin);


Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PluginHandler.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PluginHandler.pm (original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/PluginHandler.pm Thu Feb 26 20:42:47 2004
@@ -69,6 +69,7 @@
else {
dbg ("plugin: loading $package from \@INC");
$ret = eval qq{ require $package; };
+ $path = "(from \@INC)";
}

if (!$ret) {
@@ -103,13 +104,13 @@
foreach my $plugin (@{$self->{plugins}}) {
$plugin->{_inhibit_further_callbacks} = 0;

- dbg ("plugin: calling $subname on $plugin");
my $methodref = $plugin->can ($subname);

if (defined $methodref) {
eval {
$ret = &$methodref ($plugin, @_);
};
+ if ($ret) { dbg ("plugin: ${plugin}->${subname} => $ret"); }
}

if ($plugin->{_inhibit_further_callbacks}) {

Modified: incubator/spamassassin/trunk/masses/mass-check
==============================================================================
--- incubator/spamassassin/trunk/masses/mass-check (original)
+++ incubator/spamassassin/trunk/masses/mass-check Thu Feb 26 20:42:47 2004
@@ -35,6 +35,7 @@
log options
-o write all logs to stdout
--loghits log the text hit for patterns (useful for debugging)
+ --loguris log the URIs found
--hamlog=log use <log> as ham log ('ham.log' is default)
--spamlog=log use <log> as spam log ('spam.log' is default)

@@ -70,7 +71,7 @@
$opt_debug $opt_format $opt_hamlog $opt_head $opt_loghits
$opt_mid $opt_mh $opt_ms $opt_net $opt_nosort $opt_progress
$opt_showdots $opt_spamlog $opt_tail $opt_rules $opt_restart
- $opt_after $opt_rewrite);
+ $opt_loguris $opt_after $opt_rewrite);

use FindBin;
use lib "$FindBin::Bin/../lib";
@@ -91,7 +92,7 @@
GetOptions("c=s", "f=s", "j=i", "n", "o", "all", "bayes", "debug",
"hamlog=s", "head=i", "loghits", "mh", "mid", "ms", "net",
"progress", "rewrite:s", "showdots", "spamlog=s", "tail=i",
- "rules=s", "restart=i", "after=s",
+ "rules=s", "restart=i", "after=s", "loguris",
"dir" => sub { $opt_format = "dir"; },
"file" => sub { $opt_format = "file"; },
"mbox" => sub { $opt_format = "mbox"; },
@@ -251,13 +252,23 @@
$ma = $new_ma;
}

- my $status = $spamtest->check($ma);
+ # log-uris support
+ my $status;
+ my @uris;
+ if ($opt_loguris) {
+ my $pms = Mail::SpamAssassin::PerMsgStatus->new ($spamtest, $ma);
+ @uris = $pms->get_uri_list ();
+ $pms->finish();
+
+ } else {
+ $status = $spamtest->check($ma);
+ }

my @extra;
if (defined($time)) {
push(@extra, "time=$time");
}
- if (defined $status->{bayes_score}) {
+ if ($status && defined $status->{bayes_score}) {
push(@extra, "bayes=" . $status->{bayes_score});
}
if ($opt_mid) {
@@ -276,10 +287,22 @@
push(@extra, "mid=$mid");
}

- my $yorn = $status->is_spam() ? 'Y' : '.';
- my $score = $status->get_score();
- my $tests = join(",", sort(grep(length,$status->get_names_of_tests_hit(),$status->get_names_of_subtests_hit())));
- my $extra = join(",", @extra);
+ my $yorn;
+ my $score;
+ my $tests;
+ my $extra;
+
+ if ($opt_loguris) {
+ $yorn = '.';
+ $score = 0;
+ $tests = join (" ", sort @uris);
+ $extra = '';
+ } else {
+ $yorn = $status->is_spam() ? 'Y' : '.';
+ $score = $status->get_score();
+ $tests = join(",", sort(grep(length,$status->get_names_of_tests_hit(),$status->get_names_of_subtests_hit())));
+ $extra = join(",", @extra);
+ }

if (defined $opt_rewrite) {
open(REWRITE, "> " . ($opt_rewrite ? $opt_rewrite : "/tmp/out"));
@@ -310,7 +333,7 @@
}
}

- $status->finish();
+ if (defined $status) { $status->finish(); }
$ma->finish();
undef $ma; # clean 'em up
undef $status;

Added: incubator/spamassassin/trunk/rules/25_uribl.cf
==============================================================================
--- (empty file)
+++ incubator/spamassassin/trunk/rules/25_uribl.cf Thu Feb 26 20:42:47 2004
@@ -0,0 +1,23 @@
+
+loadplugin Mail::SpamAssassin::Plugin::URIDNSBL
+
+# note that this plugin defines a new config setting, 'uridnsbl',
+# which lists the zones to look up in advance. The rules will
+# not hit unless each rule has a corresponding 'uridnsbl' line.
+
+# URI-DNSBL lookups can take a *maximum* of this many seconds past the
+# normal DNSBL lookups.
+uridnsbl_timeout 2
+
+uridnsbl URIBL_SBLXBL sbl-xbl.spamhaus.org. TXT
+header URIBL_SBLXBL eval:check_uridnsbl('URIBL_SBLXBL')
+describe URIBL_SBLXBL Contains a URL listed in the SBL/XBL blocklist
+
+uridnsbl URIBL_DSBL list.dsbl.org. TXT
+header URIBL_DSBL eval:check_uridnsbl('URIBL_DSBL')
+describe URIBL_DSBL Contains a URL listed in the DSBL blocklist
+
+uridnsbl URIBL_SORBS dnsbl.sorbs.net. A
+header URIBL_SORBS eval:check_uridnsbl('URIBL_SORBS')
+describe URIBL_SORBS Contains a URL listed in the SORBS blocklist
+