[Mimedefang] Validate users before scanning?

[John Nemeth]

On Oct 9, 10:51am, Jan Pieter Cornet wrote:
} On Thu, May 19, 2005 at 09:39:33AM -0400, David F. Skoll wrote:
} > > Sendmail has two macros available: ${nbadrcpts} and ${nrcpts}, that
} > 
} > Well, it's mildly useful to know ${nbadrcpts} and ${nrcpts}, but you
} > really need to know *which* recipient addresses are invalid, especially
} > if you're using stream_by_recipient.  So being able to access those macros
} > wouldn't be a complete solution.
} 
} Actually, it would be a solution :) If you see that ${nrcpts} has
} incremented, then you know the PREVIOUS recipient that filter_recipient
} saw, was accepted by sendmail.

     Yes, but there is no guarantee that the same slave will process
all of the recepients for a given message.  You would have to save
state information in a file in the message directory.  It could get
ugly, but it does sound doable.

} Of course, this means that you'll have to check ${nrcpts} at every
} xxfi_envrcpt() milter callback, and also at the xxfi_eom() callback.
} 
}-- End of excerpt from Jan Pieter Cornet