Hello mailman-developers !
I have for a long time been annoyed by the fact that Mailman adds
more than one copy of the mailing-list-prefix to the Subject line in
replies sent to a list. This can occur if the Subject line contains
non-ASCII characters (this has been reported in the bug-database
as bug #230400: "German Umlaute mess up subject-line and body").
It looks something like (if you can view this in your mailers :-)):
Subject: Re: [Test] Re: [Test] räkmsörgås
The reason seems to be that Mailman has the following "algorithm" to
decide what to do:
- add list-name-prefix (if it isn't already there)
instead of the more correct:
- MIME-decode the subject
- add list-name-prefix (if it isn't already there)
- MIME-reencode the subject
If no MIME-decoding is done, Mailman can fail to detect the precense
of a mailing-list-prefix, since the prefix can be "buried" in the
MIME-encoded stuff.
I have made a patch to "CookHeaders.py" that fixes this problem, and
implements the second alternative above (see the attached patch).
Could something like this be added to future versions of Mailman ??
(I don't know if I have made the change the right way (I'm not a
Python programmer)).
/Johan Holmberg
*** CookHeaders.py.orig Fri Feb 16 13:40:19 2001
--- CookHeaders.py Wed Apr 11 14:00:17 2001
***************
*** 20,28 ****
--- 20,52 ----
import string
import re
import urlparse
+ import mimify
+
from Mailman import mm_cfg
+ #----------------------------------------------------------------------
+ # This function MIME-decodes the subject before deciding if a
+ # prefix should be added. IF we change the subject we have to
+ # "reencode" it before sending it.
+ #
+ # This is a "hack" made by Johan Holmberg <[EMAIL PROTECTED]>.
+ # Seems to work, but I'm no Python programmer (prefers Perl/Ruby :-))
+ # Use it at your own risk.
+ #
+
+ def add_prefix(prefix,subject):
+ if not prefix: return subject
+ text = mimify.mime_decode_header(subject)
+ if re.search(re.escape(prefix), text, re.I):
+ return subject
+ else:
+ text = prefix + subject
+ text = mimify.mime_encode_header(text)
+ return text
+
+ #----------------------------------------------------------------------
+
def process(mlist, msg, msgdata):
# Mark the message as dirty so that its text will be forced to disk next
***************
*** 48,55 ****
# we purposefully leave no space b/w prefix and subject!
if not subject:
msg['Subject'] = prefix + '(no subject)'
! elif prefix and not re.search(re.escape(prefix), subject, re.I):
! msg['Subject'] = prefix + subject
#
# get rid of duplicate headers
del msg['sender']
--- 72,79 ----
# we purposefully leave no space b/w prefix and subject!
if not subject:
msg['Subject'] = prefix + '(no subject)'
! elif prefix:
! msg['Subject'] = add_prefix(prefix,subject)
#
# get rid of duplicate headers
del msg['sender']
