| 1 |
# <@LICENSE>
|
| 2 |
# Licensed to the Apache Software Foundation (ASF) under one or more
|
| 3 |
# contributor license agreements. See the NOTICE file distributed with
|
| 4 |
# this work for additional information regarding copyright ownership.
|
| 5 |
# The ASF licenses this file to you under the Apache License, Version 2.0
|
| 6 |
# (the "License"); you may not use this file except in compliance with
|
| 7 |
# the License. You may obtain a copy of the License at:
|
| 8 |
#
|
| 9 |
# http://www.apache.org/licenses/LICENSE-2.0
|
| 10 |
#
|
| 11 |
# Unless required by applicable law or agreed to in writing, software
|
| 12 |
# distributed under the License is distributed on an "AS IS" BASIS,
|
| 13 |
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
| 14 |
# See the License for the specific language governing permissions and
|
| 15 |
# limitations under the License.
|
| 16 |
# </@LICENSE>
|
| 17 |
|
| 18 |
=head1 NAME
|
| 19 |
|
| 20 |
URIDNSBL - look up URLs against DNS blocklists
|
| 21 |
|
| 22 |
=head1 SYNOPSIS
|
| 23 |
|
| 24 |
loadplugin Mail::SpamAssassin::Plugin::URIDNSBL
|
| 25 |
uridnsbl URIBL_SBLXBL sbl-xbl.spamhaus.org. TXT
|
| 26 |
|
| 27 |
=head1 DESCRIPTION
|
| 28 |
|
| 29 |
This works by analysing message text and HTML for URLs, extracting the
|
| 30 |
domain names from those, querying their NS records in DNS, resolving
|
| 31 |
the hostnames used therein, and querying various DNS blocklists for
|
| 32 |
those IP addresses. This is quite effective.
|
| 33 |
|
| 34 |
=head1 USER SETTINGS
|
| 35 |
|
| 36 |
=over 4
|
| 37 |
|
| 38 |
=item uridnsbl_skip_domain domain1 domain2 ...
|
| 39 |
|
| 40 |
Specify a domain, or a number of domains, which should be skipped for the
|
| 41 |
URIBL checks. This is very useful to specify very common domains which are
|
| 42 |
not going to be listed in URIBLs.
|
| 43 |
|
| 44 |
=back
|
| 45 |
|
| 46 |
=head1 RULE DEFINITIONS AND PRIVILEGED SETTINGS
|
| 47 |
|
| 48 |
=over 4
|
| 49 |
|
| 50 |
=item uridnsbl NAME_OF_RULE dnsbl_zone lookuptype
|
| 51 |
|
| 52 |
Specify a lookup. C<NAME_OF_RULE> is the name of the rule to be
|
| 53 |
used, C<dnsbl_zone> is the zone to look up IPs in, and C<lookuptype>
|
| 54 |
is the type of lookup (B<TXT> or B<A>). Note that you must also
|
| 55 |
define a body-eval rule calling C<check_uridnsbl()> to use this.
|
| 56 |
|
| 57 |
Example:
|
| 58 |
|
| 59 |
uridnsbl URIBL_SBLXBL sbl-xbl.spamhaus.org. TXT
|
| 60 |
body URIBL_SBLXBL eval:check_uridnsbl('URIBL_SBLXBL')
|
| 61 |
describe URIBL_SBLXBL Contains a URL listed in the SBL/XBL blocklist
|
| 62 |
|
| 63 |
=item urirhsbl NAME_OF_RULE rhsbl_zone lookuptype
|
| 64 |
|
| 65 |
Specify a RHSBL-style domain lookup. C<NAME_OF_RULE> is the name of the rule
|
| 66 |
to be used, C<rhsbl_zone> is the zone to look up domain names in, and
|
| 67 |
C<lookuptype> is the type of lookup (B<TXT> or B<A>). Note that you must also
|
| 68 |
define a body-eval rule calling C<check_uridnsbl()> to use this.
|
| 69 |
|
| 70 |
An RHSBL zone is one where the domain name is looked up, as a string; e.g. a
|
| 71 |
URI using the domain C<foo.com> will cause a lookup of
|
| 72 |
C<foo.com.uriblzone.net>. Note that hostnames are stripped from the domain
|
| 73 |
used in the URIBL lookup, so the domain C<foo.bar.com> will look up
|
| 74 |
C<bar.com.uriblzone.net>, and C<foo.bar.co.uk> will look up
|
| 75 |
C<bar.co.uk.uriblzone.net>.
|
| 76 |
|
| 77 |
If an URI consists of an IP address instead of a hostname, the IP address is
|
| 78 |
looked up (using the standard reversed quads method) in each C<rhsbl_zone>.
|
| 79 |
|
| 80 |
Example:
|
| 81 |
|
| 82 |
urirhsbl URIBL_RHSBL rhsbl.example.org. TXT
|
| 83 |
|
| 84 |
=item urirhssub NAME_OF_RULE rhsbl_zone lookuptype subtest
|
| 85 |
|
| 86 |
Specify a RHSBL-style domain lookup with a sub-test. C<NAME_OF_RULE> is the
|
| 87 |
name of the rule to be used, C<rhsbl_zone> is the zone to look up domain names
|
| 88 |
in, and C<lookuptype> is the type of lookup (B<TXT> or B<A>).
|
| 89 |
|
| 90 |
C<subtest> is the sub-test to run against the returned data. The sub-test may
|
| 91 |
either be an IPv4 dotted address for RHSBLs that return multiple A records or a
|
| 92 |
non-negative decimal number to specify a bitmask for RHSBLs that return a
|
| 93 |
single A record containing a bitmask of results.
|
| 94 |
|
| 95 |
Note that, as with C<urirhsbl>, you must also define a body-eval rule calling
|
| 96 |
C<check_uridnsbl()> to use this.
|
| 97 |
|
| 98 |
Example:
|
| 99 |
|
| 100 |
urirhssub URIBL_RHSBL_4 rhsbl.example.org. A 127.0.0.4
|
| 101 |
urirhssub URIBL_RHSBL_8 rhsbl.example.org. A 8
|
| 102 |
|
| 103 |
=back
|
| 104 |
|
| 105 |
=head1 ADMINISTRATOR SETTINGS
|
| 106 |
|
| 107 |
=over 4
|
| 108 |
|
| 109 |
=item uridnsbl_max_domains N (default: 20)
|
| 110 |
|
| 111 |
The maximum number of domains to look up.
|
| 112 |
|
| 113 |
=back
|
| 114 |
|
| 115 |
=head1 NOTES
|
| 116 |
|
| 117 |
The C<uridnsbl_timeout> option has been obsoleted by the C<rbl_timeout>
|
| 118 |
option. See the C<Mail::SpamAssassin::Conf> POD for details on C<rbl_timeout>.
|
| 119 |
|
| 120 |
=cut
|
| 121 |
|
| 122 |
package Mail::SpamAssassin::Plugin::URIDNSBL;
|
| 123 |
|
| 124 |
use Mail::SpamAssassin::Plugin;
|
| 125 |
use Mail::SpamAssassin::Constants qw(:ip);
|
| 126 |
use Mail::SpamAssassin::Util;
|
| 127 |
use Mail::SpamAssassin::Logger;
|
| 128 |
use strict;
|
| 129 |
use warnings;
|
| 130 |
use bytes;
|
| 131 |
use re 'taint';
|
| 132 |
|
| 133 |
use vars qw(@ISA);
|
| 134 |
@ISA = qw(Mail::SpamAssassin::Plugin);
|
| 135 |
|
| 136 |
use constant LOG_COMPLETION_TIMES => 0;
|
| 137 |
|
| 138 |
# constructor
|
| 139 |
sub new {
|
| 140 |
my $class = shift;
|
| 141 |
my $samain = shift;
|
| 142 |
|
| 143 |
# some boilerplate...
|
| 144 |
$class = ref($class) || $class;
|
| 145 |
my $self = $class->SUPER::new($samain);
|
| 146 |
bless ($self, $class);
|
| 147 |
|
| 148 |
# this can be effectively global, at least in each process, safely
|
| 149 |
|
| 150 |
$self->{finished} = { };
|
| 151 |
|
| 152 |
$self->register_eval_rule ("check_uridnsbl");
|
| 153 |
$self->set_config($samain->{conf});
|
| 154 |
|
| 155 |
return $self;
|
| 156 |
}
|
| 157 |
|
| 158 |
# this is just a placeholder; in fact the results are dealt with later
|
| 159 |
sub check_uridnsbl {
|
| 160 |
return 0;
|
| 161 |
}
|
| 162 |
|
| 163 |
# ---------------------------------------------------------------------------
|
| 164 |
|
| 165 |
# once the metadata is parsed, we can access the URI list. So start off
|
| 166 |
# the lookups here!
|
| 167 |
sub parsed_metadata {
|
| 168 |
my ($self, $opts) = @_;
|
| 169 |
my $scanner = $opts->{permsgstatus};
|
| 170 |
|
| 171 |
if (!$scanner->is_dns_available()) {
|
| 172 |
$self->{dns_not_available} = 1;
|
| 173 |
return;
|
| 174 |
} else {
|
| 175 |
# due to re-testing dns may become available after being unavailable
|
| 176 |
# DOS: I don't think dns_not_available is even used anymore
|
| 177 |
$self->{dns_not_available} = 0;
|
| 178 |
}
|
| 179 |
|
| 180 |
$scanner->{'uridnsbl_activerules'} = { };
|
| 181 |
$scanner->{'uridnsbl_hits'} = { };
|
| 182 |
$scanner->{'uridnsbl_seen_domain'} = { };
|
| 183 |
|
| 184 |
# only hit DNSBLs for active rules (defined and score != 0)
|
| 185 |
$scanner->{'uridnsbl_active_rules_rhsbl'} = { };
|
| 186 |
$scanner->{'uridnsbl_active_rules_revipbl'} = { };
|
| 187 |
|
| 188 |
foreach my $rulename (keys %{$scanner->{conf}->{uridnsbls}}) {
|
| 189 |
next unless ($scanner->{conf}->is_rule_active('body_evals',$rulename));
|
| 190 |
|
| 191 |
my $rulecf = $scanner->{conf}->{uridnsbls}->{$rulename};
|
| 192 |
if ($rulecf->{is_rhsbl}) {
|
| 193 |
$scanner->{uridnsbl_active_rules_rhsbl}->{$rulename} = 1;
|
| 194 |
} else {
|
| 195 |
$scanner->{uridnsbl_active_rules_revipbl}->{$rulename} = 1;
|
| 196 |
}
|
| 197 |
}
|
| 198 |
|
| 199 |
# get all domains in message
|
| 200 |
|
| 201 |
# don't keep dereferencing this
|
| 202 |
my $skip_domains = $scanner->{main}->{conf}->{uridnsbl_skip_domains};
|
| 203 |
|
| 204 |
# list of hashes to use in order
|
| 205 |
my @uri_ordered;
|
| 206 |
|
| 207 |
# Generate the full list of html-parsed domains.
|
| 208 |
my $uris = $scanner->get_uri_detail_list();
|
| 209 |
|
| 210 |
# go from uri => info to uri_ordered
|
| 211 |
# 0: a
|
| 212 |
# 1: form
|
| 213 |
# 2: img
|
| 214 |
# 3: !a_empty
|
| 215 |
# 4: parsed
|
| 216 |
# 5: a_empty
|
| 217 |
while (my($uri, $info) = each %{$uris}) {
|
| 218 |
# we want to skip mailto: uris
|
| 219 |
next if ($uri =~ /^mailto:/);
|
| 220 |
|
| 221 |
# no domains were found via this uri, so skip
|
| 222 |
next unless ($info->{domains});
|
| 223 |
|
| 224 |
my $entry = 3;
|
| 225 |
|
| 226 |
if ($info->{types}->{a}) {
|
| 227 |
$entry = 5;
|
| 228 |
|
| 229 |
# determine a vs a_empty
|
| 230 |
foreach my $at (@{$info->{anchor_text}}) {
|
| 231 |
if (length $at) {
|
| 232 |
$entry = 0;
|
| 233 |
last;
|
| 234 |
}
|
| 235 |
}
|
| 236 |
}
|
| 237 |
elsif ($info->{types}->{form}) {
|
| 238 |
$entry = 1;
|
| 239 |
}
|
| 240 |
elsif ($info->{types}->{img}) {
|
| 241 |
$entry = 2;
|
| 242 |
}
|
| 243 |
elsif ($info->{types}->{parsed} && (keys %{$info->{types}} == 1)) {
|
| 244 |
$entry = 4;
|
| 245 |
}
|
| 246 |
|
| 247 |
# take the usable domains and add to the ordered list
|
| 248 |
foreach ( keys %{ $info->{domains} } ) {
|
| 249 |
if (exists $skip_domains->{$_}) {
|
| 250 |
dbg("uridnsbl: domain $_ in skip list");
|
| 251 |
next;
|
| 252 |
}
|
| 253 |
$uri_ordered[$entry]->{$_} = 1;
|
| 254 |
}
|
| 255 |
}
|
| 256 |
|
| 257 |
# at this point, @uri_ordered is an ordered array of uri hashes
|
| 258 |
|
| 259 |
my %domlist;
|
| 260 |
my $umd = $scanner->{main}->{conf}->{uridnsbl_max_domains};
|
| 261 |
while (keys %domlist < $umd && @uri_ordered) {
|
| 262 |
my $array = shift @uri_ordered;
|
| 263 |
next unless $array;
|
| 264 |
|
| 265 |
# run through and find the new domains in this grouping
|
| 266 |
my @domains = grep(!$domlist{$_}, keys %{$array});
|
| 267 |
next unless @domains;
|
| 268 |
|
| 269 |
# the new domains are all useful, just add them in
|
| 270 |
if (keys(%domlist) + @domains <= $umd) {
|
| 271 |
foreach (@domains) {
|
| 272 |
$domlist{$_} = 1;
|
| 273 |
}
|
| 274 |
}
|
| 275 |
else {
|
| 276 |
# trim down to a limited number - pick randomly
|
| 277 |
my $i;
|
| 278 |
while (@domains && keys %domlist < $umd) {
|
| 279 |
my $r = int rand (scalar @domains);
|
| 280 |
$domlist{splice (@domains, $r, 1)} = 1;
|
| 281 |
}
|
| 282 |
}
|
| 283 |
}
|
| 284 |
|
| 285 |
# and query
|
| 286 |
dbg("uridnsbl: domains to query: ".join(' ',keys %domlist));
|
| 287 |
foreach my $dom (keys %domlist) {
|
| 288 |
$self->query_domain ($scanner, $dom);
|
| 289 |
}
|
| 290 |
|
| 291 |
return 1;
|
| 292 |
}
|
| 293 |
|
| 294 |
sub set_config {
|
| 295 |
my($self, $conf) = @_;
|
| 296 |
my @cmds;
|
| 297 |
|
| 298 |
push(@cmds, {
|
| 299 |
setting => 'uridnsbl_max_domains',
|
| 300 |
is_admin => 1,
|
| 301 |
default => 20,
|
| 302 |
type => $Mail::SpamAssassin::Conf::CONF_TYPE_NUMERIC,
|
| 303 |
});
|
| 304 |
|
| 305 |
push (@cmds, {
|
| 306 |
setting => 'uridnsbl',
|
| 307 |
is_priv => 1,
|
| 308 |
code => sub {
|
| 309 |
my ($self, $key, $value, $line) = @_;
|
| 310 |
if ($value =~ /^(\S+)\s+(\S+)\s+(\S+)$/) {
|
| 311 |
my $rulename = $1;
|
| 312 |
my $zone = $2;
|
| 313 |
my $type = $3;
|
| 314 |
$self->{uridnsbls}->{$rulename} = {
|
| 315 |
zone => $zone, type => $type,
|
| 316 |
is_rhsbl => 0
|
| 317 |
};
|
| 318 |
}
|
| 319 |
elsif ($value =~ /^$/) {
|
| 320 |
return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;
|
| 321 |
}
|
| 322 |
else {
|
| 323 |
return $Mail::SpamAssassin::Conf::INVALID_VALUE;
|
| 324 |
}
|
| 325 |
}
|
| 326 |
});
|
| 327 |
|
| 328 |
push (@cmds, {
|
| 329 |
setting => 'urirhsbl',
|
| 330 |
is_priv => 1,
|
| 331 |
code => sub {
|
| 332 |
my ($self, $key, $value, $line) = @_;
|
| 333 |
if ($value =~ /^(\S+)\s+(\S+)\s+(\S+)$/) {
|
| 334 |
my $rulename = $1;
|
| 335 |
my $zone = $2;
|
| 336 |
my $type = $3;
|
| 337 |
$self->{uridnsbls}->{$rulename} = {
|
| 338 |
zone => $zone, type => $type,
|
| 339 |
is_rhsbl => 1
|
| 340 |
};
|
| 341 |
}
|
| 342 |
elsif ($value =~ /^$/) {
|
| 343 |
return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;
|
| 344 |
}
|
| 345 |
else {
|
| 346 |
return $Mail::SpamAssassin::Conf::INVALID_VALUE;
|
| 347 |
}
|
| 348 |
}
|
| 349 |
});
|
| 350 |
|
| 351 |
push (@cmds, {
|
| 352 |
setting => 'urirhssub',
|
| 353 |
is_priv => 1,
|
| 354 |
code => sub {
|
| 355 |
my ($self, $key, $value, $line) = @_;
|
| 356 |
if ($value =~ /^(\S+)\s+(\S+)\s+(\S+)\s+(\d{1,10}|\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3})$/) {
|
| 357 |
my $rulename = $1;
|
| 358 |
my $zone = $2;
|
| 359 |
my $type = $3;
|
| 360 |
my $subrule = $4;
|
| 361 |
$self->{uridnsbls}->{$rulename} = {
|
| 362 |
zone => $zone, type => $type,
|
| 363 |
is_rhsbl => 1, is_subrule => 1
|
| 364 |
};
|
| 365 |
$self->{uridnsbl_subs}->{$zone} ||= { };
|
| 366 |
push (@{$self->{uridnsbl_subs}->{$zone}->{$subrule}->{rulenames}}, $rulename);
|
| 367 |
}
|
| 368 |
elsif ($value =~ /^$/) {
|
| 369 |
return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;
|
| 370 |
}
|
| 371 |
else {
|
| 372 |
return $Mail::SpamAssassin::Conf::INVALID_VALUE;
|
| 373 |
}
|
| 374 |
}
|
| 375 |
});
|
| 376 |
|
| 377 |
push (@cmds, {
|
| 378 |
setting => 'uridnsbl_skip_domain',
|
| 379 |
default => {},
|
| 380 |
code => sub {
|
| 381 |
my ($self, $key, $value, $line) = @_;
|
| 382 |
if ($value =~ /^$/) {
|
| 383 |
return $Mail::SpamAssassin::Conf::MISSING_REQUIRED_VALUE;
|
| 384 |
}
|
| 385 |
foreach my $domain (split(/\s+/, $value)) {
|
| 386 |
$self->{uridnsbl_skip_domains}->{lc $domain} = 1;
|
| 387 |
}
|
| 388 |
}
|
| 389 |
});
|
| 390 |
|
| 391 |
# obsolete
|
| 392 |
push(@cmds, {
|
| 393 |
setting => 'uridnsbl_timeout',
|
| 394 |
code => sub {
|
| 395 |
# not a lint_warn(), since it's pretty harmless and we don't want
|
| 396 |
# to break stuff like sa-update
|
| 397 |
warn("config: 'uridnsbl_timeout' is obsolete, use 'rbl_timeout' instead");
|
| 398 |
return 0;
|
| 399 |
}
|
| 400 |
});
|
| 401 |
|
| 402 |
$conf->{parser}->register_commands(\@cmds);
|
| 403 |
}
|
| 404 |
|
| 405 |
# ---------------------------------------------------------------------------
|
| 406 |
|
| 407 |
sub query_domain {
|
| 408 |
my ($self, $scanner, $dom) = @_;
|
| 409 |
|
| 410 |
#warn "uridnsbl: domain $dom\n";
|
| 411 |
#return;
|
| 412 |
|
| 413 |
$dom = lc $dom;
|
| 414 |
return if $scanner->{uridnsbl_seen_domain}->{$dom};
|
| 415 |
$scanner->{uridnsbl_seen_domain}->{$dom} = 1;
|
| 416 |
$self->log_dns_result("querying domain $dom");
|
| 417 |
|
| 418 |
my $obj = { dom => $dom };
|
| 419 |
|
| 420 |
my $single_dnsbl = 0;
|
| 421 |
if ($dom =~ /^\d+\.\d+\.\d+\.\d+$/) {
|
| 422 |
my $IPV4_ADDRESS = IPV4_ADDRESS;
|
| 423 |
my $IP_PRIVATE = IP_PRIVATE;
|
| 424 |
# only look up the IP if it is public and valid
|
| 425 |
if ($dom =~ /^$IPV4_ADDRESS$/ && $dom !~ /^$IP_PRIVATE$/) {
|
| 426 |
$self->lookup_dnsbl_for_ip($scanner, $obj, $dom);
|
| 427 |
# and check the IP in RHSBLs too
|
| 428 |
if ($dom =~ /^(\d+)\.(\d+)\.(\d+)\.(\d+)$/) {
|
| 429 |
$dom = "$4.$3.$2.$1";
|
| 430 |
$single_dnsbl = 1;
|
| 431 |
}
|
| 432 |
}
|
| 433 |
}
|
| 434 |
else {
|
| 435 |
$single_dnsbl = 1;
|
| 436 |
}
|
| 437 |
|
| 438 |
if ($single_dnsbl) {
|
| 439 |
# look up the domain in the RHSBL subset
|
| 440 |
my $cf = $scanner->{uridnsbl_active_rules_rhsbl};
|
| 441 |
foreach my $rulename (keys %{$cf}) {
|
| 442 |
my $rulecf = $scanner->{conf}->{uridnsbls}->{$rulename};
|
| 443 |
$self->lookup_single_dnsbl($scanner, $obj, $rulename,
|
| 444 |
$dom, $rulecf->{zone}, $rulecf->{type});
|
| 445 |
|
| 446 |
# see comment below
|
| 447 |
$scanner->register_async_rule_start($rulename);
|
| 448 |
}
|
| 449 |
|
| 450 |
# perform NS, A lookups to look up the domain in the non-RHSBL subset
|
| 451 |
if ($dom !~ /^\d+\.\d+\.\d+\.\d+$/) {
|
| 452 |
$self->lookup_domain_ns($scanner, $obj, $dom);
|
| 453 |
}
|
| 454 |
}
|
| 455 |
|
| 456 |
# note that these rules are now underway. important: unless the
|
| 457 |
# rule hits, in the current design, these will not be considered
|
| 458 |
# "finished" until harvest_dnsbl_queries() completes
|
| 459 |
my $cf = $scanner->{uridnsbl_active_rules_revipbl};
|
| 460 |
foreach my $rulename (keys %{$cf}) {
|
| 461 |
$scanner->register_async_rule_start($rulename);
|
| 462 |
}
|
| 463 |
}
|
| 464 |
|
| 465 |
# ---------------------------------------------------------------------------
|
| 466 |
|
| 467 |
sub lookup_domain_ns {
|
| 468 |
my ($self, $scanner, $obj, $dom) = @_;
|
| 469 |
|
| 470 |
my $key = "NS:".$dom;
|
| 471 |
return if $scanner->{async}->get_lookup($key);
|
| 472 |
|
| 473 |
# dig $dom ns
|
| 474 |
my $ent = $self->start_lookup($scanner, $dom, 'NS',
|
| 475 |
$self->res_bgsend($scanner, $dom, 'NS', $key),
|
| 476 |
$key);
|
| 477 |
$ent->{obj} = $obj;
|
| 478 |
}
|
| 479 |
|
| 480 |
sub complete_ns_lookup {
|
| 481 |
my ($self, $scanner, $ent, $dom) = @_;
|
| 482 |
|
| 483 |
my $packet = $ent->{response_packet};
|
| 484 |
my @answer = !defined $packet ? () : $packet->answer;
|
| 485 |
|
| 486 |
my $IPV4_ADDRESS = IPV4_ADDRESS;
|
| 487 |
my $IP_PRIVATE = IP_PRIVATE;
|
| 488 |
|
| 489 |
foreach my $rr (@answer) {
|
| 490 |
my $str = $rr->string;
|
| 491 |
next unless (defined($str) && defined($dom));
|
| 492 |
$self->log_dns_result ("NSs for $dom: $str");
|
| 493 |
|
| 494 |
if ($str =~ /IN\s+NS\s+(\S+)/) {
|
| 495 |
my $nsmatch = $1;
|
| 496 |
|
| 497 |
if ($nsmatch =~ /^\d+\.\d+\.\d+\.\d+\.?$/) {
|
| 498 |
$nsmatch =~ s/\.$//;
|
| 499 |
# only look up the IP if it is public and valid
|
| 500 |
if ($nsmatch =~ /^$IPV4_ADDRESS$/ && $nsmatch !~ /^$IP_PRIVATE$/) {
|
| 501 |
$self->lookup_dnsbl_for_ip($scanner, $ent->{obj}, $nsmatch);
|
| 502 |
}
|
| 503 |
}
|
| 504 |
else {
|
| 505 |
$self->lookup_a_record($scanner, $ent->{obj}, $nsmatch);
|
| 506 |
}
|
| 507 |
}
|
| 508 |
}
|
| 509 |
}
|
| 510 |
|
| 511 |
# ---------------------------------------------------------------------------
|
| 512 |
|
| 513 |
sub lookup_a_record {
|
| 514 |
my ($self, $scanner, $obj, $hname) = @_;
|
| 515 |
|
| 516 |
my $key = "A:".$hname;
|
| 517 |
return if $scanner->{async}->get_lookup($key);
|
| 518 |
|
| 519 |
# dig $hname a
|
| 520 |
my $ent = $self->start_lookup($scanner, $hname, 'A',
|
| 521 |
$self->res_bgsend($scanner, $hname, 'A', $key),
|
| 522 |
$key);
|
| 523 |
$ent->{obj} = $obj;
|
| 524 |
}
|
| 525 |
|
| 526 |
sub complete_a_lookup {
|
| 527 |
my ($self, $scanner, $ent, $hname) = @_;
|
| 528 |
|
| 529 |
my $packet = $ent->{response_packet};
|
| 530 |
my @answer = !defined $packet ? () : $packet->answer;
|
| 531 |
foreach my $rr (@answer) {
|
| 532 |
my $str = $rr->string;
|
| 533 |
$self->log_dns_result ("A for NS $hname: $str");
|
| 534 |
|
| 535 |
if ($str =~ /IN\s+A\s+(\S+)/) {
|
| 536 |
$self->lookup_dnsbl_for_ip($scanner, $ent->{obj}, $1);
|
| 537 |
}
|
| 538 |
}
|
| 539 |
}
|
| 540 |
|
| 541 |
# ---------------------------------------------------------------------------
|
| 542 |
|
| 543 |
sub lookup_dnsbl_for_ip {
|
| 544 |
my ($self, $scanner, $obj, $ip) = @_;
|
| 545 |
|
| 546 |
$ip =~ /^(\d+)\.(\d+)\.(\d+)\.(\d+)$/;
|
| 547 |
my $revip = "$4.$3.$2.$1";
|
| 548 |
|
| 549 |
my $cf = $scanner->{uridnsbl_active_rules_revipbl};
|
| 550 |
foreach my $rulename (keys %{$cf}) {
|
| 551 |
my $rulecf = $scanner->{conf}->{uridnsbls}->{$rulename};
|
| 552 |
$self->lookup_single_dnsbl($scanner, $obj, $rulename,
|
| 553 |
$revip, $rulecf->{zone}, $rulecf->{type});
|
| 554 |
}
|
| 555 |
}
|
| 556 |
|
| 557 |
sub lookup_single_dnsbl {
|
| 558 |
my ($self, $scanner, $obj, $rulename, $lookupstr, $dnsbl, $qtype) = @_;
|
| 559 |
|
| 560 |
my $key = "DNSBL:".$dnsbl.":".$lookupstr;
|
| 561 |
return if $scanner->{async}->get_lookup($key);
|
| 562 |
my $item = $lookupstr.".".$dnsbl;
|
| 563 |
|
| 564 |
# dig $ip txt
|
| 565 |
my $ent = $self->start_lookup($scanner, $item, 'DNSBL',
|
| 566 |
$self->res_bgsend($scanner, $item, $qtype, $key),
|
| 567 |
$key);
|
| 568 |
$ent->{obj} = $obj;
|
| 569 |
$ent->{rulename} = $rulename;
|
| 570 |
$ent->{zone} = $dnsbl;
|
| 571 |
}
|
| 572 |
|
| 573 |
sub complete_dnsbl_lookup {
|
| 574 |
my ($self, $scanner, $ent, $dnsblip) = @_;
|
| 575 |
|
| 576 |
my $conf = $scanner->{conf};
|
| 577 |
my @subtests;
|
| 578 |
my $rulename = $ent->{rulename};
|
| 579 |
my $rulecf = $conf->{uridnsbls}->{$rulename};
|
| 580 |
|
| 581 |
my $packet = $ent->{response_packet};
|
| 582 |
my @answer = !defined $packet ? () : $packet->answer;
|
| 583 |
|
| 584 |
my $uridnsbl_subs = $conf->{uridnsbl_subs}->{$ent->{zone}};
|
| 585 |
foreach my $rr (@answer)
|
| 586 |
{
|
| 587 |
next if ($rr->type ne 'A' && $rr->type ne 'TXT');
|
| 588 |
|
| 589 |
my $rdatastr = $rr->rdatastr;
|
| 590 |
my $dom = $ent->{obj}->{dom};
|
| 591 |
|
| 592 |
if (!$rulecf->{is_subrule}) {
|
| 593 |
# this zone is a simple rule, not a set of subrules
|
| 594 |
# skip any A record that isn't on 127/8
|
| 595 |
if ($rr->type eq 'A' && $rr->rdatastr !~ /^127\./) {
|
| 596 |
warn("uridnsbl: bogus rr for domain=$dom, rule=$rulename, id=" .
|
| 597 |
$packet->header->id." rr=".$rr->string);
|
| 598 |
next;
|
| 599 |
}
|
| 600 |
$self->got_dnsbl_hit($scanner, $ent, $rdatastr, $dom, $rulename);
|
| 601 |
}
|
| 602 |
else {
|
| 603 |
foreach my $subtest (keys (%{$uridnsbl_subs}))
|
| 604 |
{
|
| 605 |
if ($subtest eq $rdatastr) {
|
| 606 |
foreach my $subrulename (@{$uridnsbl_subs->{$subtest}->{rulenames}}) {
|
| 607 |
$self->got_dnsbl_hit($scanner, $ent, $rdatastr, $dom, $subrulename);
|
| 608 |
}
|
| 609 |
}
|
| 610 |
# bitmask
|
| 611 |
elsif ($subtest =~ /^\d+$/) {
|
| 612 |
if ($rdatastr =~ m/^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}$/ &&
|
| 613 |
Mail::SpamAssassin::Util::my_inet_aton($rdatastr) & $subtest)
|
| 614 |
{
|
| 615 |
foreach my $subrulename (@{$uridnsbl_subs->{$subtest}->{rulenames}}) {
|
| 616 |
$self->got_dnsbl_hit($scanner, $ent, $rdatastr, $dom, $subrulename);
|
| 617 |
}
|
| 618 |
}
|
| 619 |
}
|
| 620 |
}
|
| 621 |
}
|
| 622 |
}
|
| 623 |
}
|
| 624 |
|
| 625 |
sub got_dnsbl_hit {
|
| 626 |
my ($self, $scanner, $ent, $str, $dom, $rulename) = @_;
|
| 627 |
|
| 628 |
$str =~ s/\s+/ /gs; # long whitespace => short
|
| 629 |
dbg("uridnsbl: domain \"$dom\" listed ($rulename): $str");
|
| 630 |
|
| 631 |
if (!defined $scanner->{uridnsbl_hits}->{$rulename}) {
|
| 632 |
$scanner->{uridnsbl_hits}->{$rulename} = { };
|
| 633 |
};
|
| 634 |
$scanner->{uridnsbl_hits}->{$rulename}->{$dom} = 1;
|
| 635 |
|
| 636 |
if ($scanner->{uridnsbl_active_rules_revipbl}->{$rulename}
|
| 637 |
|| $scanner->{uridnsbl_active_rules_rhsbl}->{$rulename})
|
| 638 |
{
|
| 639 |
# TODO: this needs to handle multiple domain hits per rule
|
| 640 |
$scanner->clear_test_state();
|
| 641 |
my $uris = join (' ', keys %{$scanner->{uridnsbl_hits}->{$rulename}});
|
| 642 |
$scanner->test_log ("URIs: $uris");
|
| 643 |
$scanner->got_hit ($rulename, "");
|
| 644 |
|
| 645 |
# note that this rule has completed (since it got at least 1 hit)
|
| 646 |
$scanner->register_async_rule_finish($rulename);
|
| 647 |
}
|
| 648 |
}
|
| 649 |
|
| 650 |
# ---------------------------------------------------------------------------
|
| 651 |
|
| 652 |
sub start_lookup {
|
| 653 |
my ($self, $scanner, $zone, $type, $id, $key) = @_;
|
| 654 |
|
| 655 |
my $ent = {
|
| 656 |
key => $key,
|
| 657 |
zone => $zone, # serves to fetch other per-zone settings
|
| 658 |
type => "URI-".$type,
|
| 659 |
id => $id,
|
| 660 |
completed_callback => sub {
|
| 661 |
my $ent = shift;
|
| 662 |
if (defined $ent->{response_packet}) { # not aborted or empty
|
| 663 |
$self->completed_lookup_callback ($scanner, $ent);
|
| 664 |
}
|
| 665 |
}
|
| 666 |
};
|
| 667 |
$scanner->{async}->start_lookup($ent);
|
| 668 |
return $ent;
|
| 669 |
}
|
| 670 |
|
| 671 |
sub completed_lookup_callback {
|
| 672 |
my ($self, $scanner, $ent) = @_;
|
| 673 |
my $type = $ent->{type};
|
| 674 |
my $key = $ent->{key};
|
| 675 |
$key =~ /:(\S+?)$/; my $val = $1;
|
| 676 |
|
| 677 |
if ($type eq 'URI-NS') {
|
| 678 |
$self->complete_ns_lookup ($scanner, $ent, $val);
|
| 679 |
}
|
| 680 |
elsif ($type eq 'URI-A') {
|
| 681 |
$self->complete_a_lookup ($scanner, $ent, $val);
|
| 682 |
}
|
| 683 |
elsif ($type eq 'URI-DNSBL') {
|
| 684 |
$self->complete_dnsbl_lookup ($scanner, $ent, $val);
|
| 685 |
}
|
| 686 |
}
|
| 687 |
|
| 688 |
# ---------------------------------------------------------------------------
|
| 689 |
|
| 690 |
sub res_bgsend {
|
| 691 |
my ($self, $scanner, $host, $type, $key) = @_;
|
| 692 |
|
| 693 |
return $self->{main}->{resolver}->bgsend($host, $type, undef, sub {
|
| 694 |
my ($pkt, $id, $timestamp) = @_;
|
| 695 |
$scanner->{async}->set_response_packet($id, $pkt, $key, $timestamp);
|
| 696 |
});
|
| 697 |
}
|
| 698 |
|
| 699 |
sub log_dns_result {
|
| 700 |
#my $self = shift;
|
| 701 |
#Mail::SpamAssassin::dbg("uridnsbl: ".join (' ', @_));
|
| 702 |
}
|
| 703 |
|
| 704 |
# ---------------------------------------------------------------------------
|
| 705 |
|
| 706 |
1;
|
Parent Directory
| 
Revision Log