You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Robert Menschel <Ro...@Menschel.net> on 2004/02/18 16:00:03 UTC
Re[2]: Another spammer sign to catch
Hello Bob,
Tuesday, February 17, 2004, 7:29:10 AM, you wrote:
>> There is at least one evil mailer that uses random lowercase words in
>> the X-mailer field. I have a rule to match these; it's part of an
>> upcoming "randoms" ruleset...
BA> I've found low FPs looking for "/^X-Mailer: [a-z ]*/"
BA> Would the following work?:
BA> header T_LCASED_XMAILER X-Mailer =~ /^[a-z ]*$/
BA> describe T_LCASED_XMAILER X-Mailer contains only lowercase words
BA> score T_LCASED_XMAILER 0.5
T_LCASED_XMAILER -- 34193s/5681h of 100794 corpus (82099s/18695h) 02/17/04
Hits 30% of my ham.
Most of the matches logged by mass-check show:
> # T_LCASED_XMAILER=""
in other words, NO x-mailer.
Include a test to exclude those, and you may have a good test.
Tuesday, February 17, 2004, 7:53:09 AM, Pierre wrote:
PT> Here's my current version -- note that the "evil" ones have a
PT> space after each word, even the last one:
PT> header RANDMAILER X-Mailer =~ /^([a-z]{4,15} ){1,5}$/
PT> describe RANDMAILER random words in X-Mailer field
PT> score RANDMAILER 2.0
RANDMAILER -- 1267s/0h of 100794 corpus (82099s/18695h) 02/17/04
hits 1.5% of my spam, and no ham. Works for me!
Bob Menschel
Re[3]: Another spammer sign to catch
Posted by Robert Menschel <Ro...@Menschel.net>.
Hello Brent,
Wednesday, February 18, 2004, 8:25:18 AM, you wrote:
BJN> On Wed, 18 Feb 2004, Robert Menschel <Ro...@Menschel.net> wrote:
>> BA> header T_LCASED_XMAILER X-Mailer =~ /^[a-z ]*$/
>> BA> describe T_LCASED_XMAILER X-Mailer contains only lowercase words
>> BA> score T_LCASED_XMAILER 0.5
>>
>> T_LCASED_XMAILER -- 34193s/5681h of 100794 corpus (82099s/18695h) 02/17/04
>>
>> Hits 30% of my ham.
BJN> You know, this has been discussed a lot on the list already (search the
BJN> archive for "X-Mailer"); why reinvent the wheel? Here's the rule I'm
BJN> using now, based on one that was posted weeks ago.
BJN> # X-Mailer handling
BJN> # Note that one person reported FP on "X-Mailer: webmail.delfi.lt"
BJN> # so we include "." after "^A-Z0-9"
BJN> header BCS_XMAILERBOGUS X-Mailer =~ /^[a-z][^A-Z0-9.]*$/
BJN> describe BCS_XMAILERBOGUS X-Mailer has NO uppercase letters, numbers, dots
BJN> score BCS_XMAILERBOGUS 2.0
Works OK. My results:
BCS_XMAILERBOGUS -- 2014s/2h of 100794 corpus (82099s/18695h) 02/18/04
Ham Matches:
# BCS_XMAILERBOGUS="a script\n" [Found in email from Macromedia.com, re: Dreamweaver]
# BCS_XMAILERBOGUS="ehouse\n" [Found in email from gmcard.com, re: GM brand credit card, to card holder]
Bob Menschel
Re[2]: Another spammer sign to catch
Posted by "Brent J. Nordquist" <b-...@bethel.edu>.
On Wed, 18 Feb 2004, Robert Menschel <Ro...@Menschel.net> wrote:
> BA> header T_LCASED_XMAILER X-Mailer =~ /^[a-z ]*$/
> BA> describe T_LCASED_XMAILER X-Mailer contains only lowercase words
> BA> score T_LCASED_XMAILER 0.5
>
> T_LCASED_XMAILER -- 34193s/5681h of 100794 corpus (82099s/18695h) 02/17/04
>
> Hits 30% of my ham.
You know, this has been discussed a lot on the list already (search the
archive for "X-Mailer"); why reinvent the wheel? Here's the rule I'm
using now, based on one that was posted weeks ago.
# X-Mailer handling
# Note that one person reported FP on "X-Mailer: webmail.delfi.lt"
# so we include "." after "^A-Z0-9"
header BCS_XMAILERBOGUS X-Mailer =~ /^[a-z][^A-Z0-9.]*$/
describe BCS_XMAILERBOGUS X-Mailer has NO uppercase letters, numbers, dots
score BCS_XMAILERBOGUS 2.0
--
Brent J. Nordquist <b-...@bethel.edu> N0BJN
Other contact information: http://kepler.acns.bethel.edu/~bjn/contact.html