A strange issue indeed, this exact same python code is running on
another server with sendmail and works fine, as you can see from the
testmilter.py log:
[15:19:07] Connect from 127.0.0.1:37841 (localhost) with family: 4
[15:19:07] MAIL FROM: m...@example.org
[15:19:08] RCPT TO: g...@example.org
[15:19:08] RCPT TO: b...@example.org
[15:19:08] EOB
[15:19:08] bad_emails: ['b...@example.org']
[15:19:08] Close called. QID: rA8KJ7Tw009274
Even the python-virtualenv it uses libraries from is identical, copied
from one server to the other with rsync. Python and OS versions differ
slightly (redhat/python 2.7.5/sendmail vs ubuntu/python 2.7.3/postfix),
but I wouldn't think that would make the difference. python-libmilter
only lives in the above mentioned virtualenv that is identical.
I went ahead and ran:
tcpdump -w /tmp/log.txt -i lo -s 0 host 127.0.0.1 and port 5001
on both servers to see what was actually sent back and forth from
sendmail/postfix to the python milter, and discovered packets from
sendmail contain:
{rcpt_addr}b...@example.org
while packets from postfix contain:
{rcpt_addr}b...@example.org
And both of them end up sending this snippet later in the transaction:
R<b...@example.org>
So is it possible your milter implementation pulls the email from this
later bit when it *should* pull it from the first bit?
I'm running short on ideas here. :/
On 11/08/2013 02:46 PM, Wietse Venema wrote:
> moparisthebest:
>> ##### try 2, should be filtered but isn't
>> # smtp email addressed to 'g...@example.org b...@example.org'
>>
> ...
>> # milter log
>> [14:06:56] Connect from 127.0.0.1:48029 (localhost) with family: 4
>> [14:06:56] MAIL FROM: m...@example.org
>> [14:06:56] RCPT TO: g...@example.org
>> [14:06:56] RCPT TO: b...@example.org
>> [14:06:56] EOB
>> [14:06:56] bad_emails: ['b...@example.org']
>> [14:06:56] Close called. QID: None
>
> In my milter logging, the recipient address case is preserved:
>
> SMTP session:
>
> mail from:<>
> 250 2.1.0 Ok
> rcpt to:<x...@porcupine.org>
> 250 2.1.5 Ok
> rcpt to:<x...@porcupine.org>
> 250 2.1.5 Ok
> rcpt to:<x...@porcupine.org>
> 250 2.1.5 Ok
>
> Milter logging:
>
> test_rcpt "<x...@porcupine.org>"
> test_rcpt "<x...@porcupine.org>"
> test_rcpt "<x...@porcupine.org>"
>
> Therefore your milter library is folding case.
>
> Wietse
>
