On Tue, Aug 19, 2003 at 04:58:41PM -0500,
Rick Cooper <rick@???> is thought to have said:
> I just read your first message again, make that !def: if you are looking for
> missing message-id, sorry was looking at the condition and not the desire
(adding exim-users back in)
Actually this is odd.
With {def:h_message-id:} and the X-MailScanner match it results in a
rejection whether or not there's a Message-ID header.
With {!def:h_message-id:} and the X-MailScanner match it results in an
accept whether or not there's a Message-ID header.
Oh. I see why. It's because Exim has already inserted it's own Message-ID
header before the DATA ACL fires. So that makes testing for existance of
Message-ID: impossible. Philip is this a bug? Or at the very least a place
where the documentation needs to be updated?
Although I suppose it's easy enough to work around:
deny condition = ${if and{{match{$h_message-id:}{<E$message_id@$primary_hostname>}} \
{match{$h_x-mailscanner:}{Found to be clean}}} \
{true}{false}}
log_message = X-MailScanner header with no Message-ID. Probably \
W32.Sobig.F.
this seems to work for me now.
--
--------------------------------------------------------------------
Tabor J. Wells twells@???
Fsck It! Just another victim of the ambient morality
