http://bugzilla.spamassassin.org/show_bug.cgi?id=3600
------- Additional Comments From [EMAIL PROTECTED] 2004-11-05 09:29 -------
It's a feature of Exim, not a problem. Exim allows to configure the received
headers in many ways. So you'll find different header formats on many
distributions and even more in different installations.
I send you a mail which passes exim with this configuration line:
#
# Received Headerzeilen
#
received_header_text = Received: by $primary_hostname (Exim $version_number)\n\
\ ${if def:sender_rcvhost {from $sender_rcvhost\n }}\
${if def:received_for {for $received_for${if def:received_protocol { }}}}\
${if def:received_protocol {with $received_protocol}}\
${if def:received_for {\n }{${if def:received_protocol {\n
}}}}\
id <$message_id>
which generates something like this:
Received: by Mail.CIS.FU-Berlin.DE (Exim 4.42)
from mail.zedat.fu-berlin.de ([130.133.1.48])
with esmtp
id <1CQ6eG-01ACOt-W2>; Fri, 05 Nov 2004 17:09:13 +0100
As far as I understand, this is fully legal and the parser in SpamAssassin
should not depend on the order of the keywords (by...from versus from...by).
Instead it should parse the header after separating the different values by the
token names.
Exim is a very flexible MTA, others may follow. The regular expressions in
SpamAssassin are not as flexible yet. A real token parser might help.
Heiko
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
