Hi, I found the following trouble report and solution in the mailman-jp user's list. http://mm.tkikuchi.net/pipermail/mmjp-users/2006-October/001882.html He reported that his mailman displays "unknown encoding" error and die as follows:
--------------------------------------------------------------------- File "/usr/local/mailman/pythonlib/email/Charset.py", line 283, in convert ).encode(self.output_codec, 'replace') LookupError: unknown encoding: iso-2022-jp ---------------------------------------------------------------------
They said this problem was actualized when any Japanese charatcter was parsed after "import japanese" on the combination of mailman-2.1.9 and python-2.5. Another person suggested to comment out "import japanese" in the following files: /usr/lib/mailman/cron/paths.py /usr/lib/mailman/tests/paths.py /usr/lib/mailman/scripts/paths.py /usr/lib/mailman/bin/paths.py Python-2.4 or later has CJKCodec module in the default, so it is applied if "japanese" module is not loaded explicitly and you can avoid the trouble. He estimated "japanese" module breaks something in the combination. (*) CJK = China, Japan, Korea Then, the reporter replied his trouble got resolved. Could you try this way to confirm whether it can be a solution for us, or not. Thanks, -- OSS Platform Development Division, NEC KaiGai Kohei <kaigai@ak.jp.nec.com>
KaiGai Kohei wrote:
They said this problem was actualized when any Japanese charatcter was parsed after "import japanese" on the combination of mailman-2.1.9 and python-2.5.
I am running a python 2.4, but indeed with mailman 2.1.9. It looks like all of Tsugikazu Shibata's emails made it through the list to my mailbox and to the (private) mbox archive just fine, but mailman somehow failed to store them in the web archive. Tsugikazu Shibata did not get his own emails back because mailman had "nodupes" set by default for every user. I have switched that off.
Another person suggested to comment out "import japanese" in the following files: /usr/lib/mailman/cron/paths.py /usr/lib/mailman/tests/paths.py /usr/lib/mailman/scripts/paths.py /usr/lib/mailman/bin/paths.py
I have commented out the "import japanese" lines in those files. I hope the web archiving now works right for all emails. kind regards, Rik van Riel -- Politics is the struggle between those who want to make their country the best in the world, and those who believe it already is. Each group calls the other unpatriotic.
On Mon, 03 Sep 2007 11:26:22 -0400, riel wrote:
KaiGai Kohei wrote:
They said this problem was actualized when any Japanese charatcter was parsed after "import japanese" on the combination of mailman-2.1.9 and python-2.5.
I am running a python 2.4, but indeed with mailman 2.1.9. It looks like all of Tsugikazu Shibata's emails made it through the list to my mailbox and to the (private) mbox archive just fine, but mailman somehow failed to store them in the web archive.
Tsugikazu Shibata did not get his own emails back because mailman had "nodupes" set by default for every user. I have switched that off.
Thank you for that, ありがとうございます。
Another person suggested to comment out "import japanese" in the following files: /usr/lib/mailman/cron/paths.py /usr/lib/mailman/tests/paths.py /usr/lib/mailman/scripts/paths.py /usr/lib/mailman/bin/paths.py
I have commented out the "import japanese" lines in those files. I hope the web archiving now works right for all emails.
This email is also test purpose. このメールはテストの目的でも出しています。 しばた
Tsugikazu Shibata wrote:
This email is also test purpose. このメールはテストの目的でも出しています。
It seems to work: http://lists.kernelnewbies.org/pipermail/jp-kernelnewbies/2007-September/000... Thank you, KaiGai Kohei! -- Politics is the struggle between those who want to make their country the best in the world, and those who believe it already is. Each group calls the other unpatriotic.
participants (3)
-
KaiGai Kohei -
Rik van Riel -
Tsugikazu Shibata