Hi,
if you re-inject mails, there are two ways to do so. You could pipe
the SMTP data (the header and body part of the message) to the program
and put the addresses, you want the mail to be resent to, on the
commandline. If the program you call needs to spawn a shell, you must
carefully filter what you put on the commandline. If you use BSMTP
the addresses are part of the data you send on stdin to the called
program, therefore this is the recommended way.
This does not have any impact if you use spamc as filter to the
normal local_delivery director but if you use a special director
for filtering.
cu,
Jan
Martijn Grendelman wrote:
>>Martijn Grendelman wrote:
>>
>>
>>>Now I was thinking: wouldn't it be easier to just use transport_filter
>>
> on
>
>>>existing transports? I guess I must be missing something. Please
>>
> enlighten
>
>>>me :-)
>>>
>>>local_delivery:
>>> driver = appendfile
>>> delivery_date_add
>>> envelope_to_add
>>> file = /var/mail/$local_part
>>> return_path_add
>>> transport_filter = /usr/bin/spamc
>>>
>>>Would this work?
>>>
>>
>>Yes, this does work. It won't let you make any routing decisions based on
>
> the SA
>
>>results, though, for obvious reasons.
>
>
> Obvious indeed, but I don't need that. I just want spam messages tagged by
> SA, sending the spam to the bitbucket or not is a client responsibility IMO.
>
> There's one more thing I was wondering about. The transport solution that
> I've seen suggested usually contains "use_bsmtp = true" option. I quote
> dman: "This pipes the message to exim using the BSMTP (batched SMTP)
> protocol. This avoids any nasties with shell metacharacters in addresses".
> Is this something I have to worry about? I really don't want to enable that
> option for my local_delivery transport ;-)
>
> Best regards,
> Martijn.
>
>
>
> --
>
> ## List details at http://www.exim.org/mailman/listinfo/exim-users Exim details at http://www.exim.org/ ##
>
>
