| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- bin/check_dbs because there are no more pickles
- bin/convert because in MM3 everything will use $-strings and we'll do the
conversion on import of the old list data
- bin/mmshell because setuptools provides its own equivalent
- bin/qrunner and bin/mailmanctl because the files were empty
There are a bunch of scripts still left in bin/ which will eventually get
removed. I'm leaving them for now because they're either helpers generally
tangential to Mailman (msgfmt, po2template, pygettext, templ2pot, transcheck),
or are using old interfaces that will go away soon (clone_member, convert,
discard, fix_url, list_admins, remove_members, reset_pw, sync_members).
Also moved bin/cleanarch into Mailman/bin/cleanarch.py and updated it to use
optparse.
Also added a small patch to genalias to standardize its help printing.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
configuration file. While the full conversion is not yet complete, everything
that seems to be required to run mailmanctl, qrunner, rmlist, and newlist have
been updated.
Basically, modules should no longer import mm_cfg, but instead they should
import Mailman.configuration.config. The latter is an object that's
guaranteed to exist, but not guaranteed to be initialized until some top-level
script calls config.load(). The latter should be called with the argument to
-C/--config which is a new convention the above scripts have been given.
In most cases, where mm_cfg.<variable> is used config.<variable> can be used,
but the exceptions are where the default value must be available before
config.load() is called. Sometimes you can import Mailman.Default and get the
variable from there, but other times the code has to be changed to work around
this limitation. Take each on a case-by-case basis.
Note that the various directories calculated from VAR_PREFIX, EXEC_PREFIX, and
PREFIX are now calculated in config.py, not in Defaults.py. This way a
configuration file can override the base directories and everything should
work correctly.
Other changes here include:
- mailmanctl, qrunner, and update are switched to optparse and $-strings, and
changed to the mmshell architecture
- An etc directory has been added to /usr/local/mailman and a
mailman.cfg.sample file is installed there. Sites should now edit an
etc/mailman.cfg file to do their configurations, although the mm_cfg file is
still honored. The formats of the two files are identical.
- list_lists is given the -C/--config option
- Some coding style fixes in bin/update, but not extensive
- Get rid of nested scope hacks in qrunner.py
- A start on getting EmailBase tests working (specifically test_message),
although not yet complete.
|
| |
|
|
|
|
|
|
|
|
|
| |
I didn't update cron/mailpasswds because that is going away. cron/disabled
and cron/gate_news are only minimally tested.
- Instead of using the RotatingFileHandler for our logs, use our own
ReopenableFileHandler, which only minimally derives from FileHandler and
implements a reopen() method. I think it's generally better to leave file
rotation to external tools such as logrotate.
- Remove some Python 2.1 compatibility stuff.
- Ignore .mo files in messages/vi/LC_MESSAGES
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
slot, since argv[0] is used by the Python interpreter to define its
library search paths. Depending on various and sundry environmental
issues ($PATH, --prefix, --exec-prefix, etc.) it can still work fine,
but this guarantees the right libraries will be found for the version
of Python being used.
Bug fix candidate.
|
| |
|
|
| |
the site list doesn't exist.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
mailmanctl -u, you might not be able to completely delete a list
through the web. This is because the archives are written with the
group of the mailmanctl script, which can be changed if that's not
root, but ttw list deletion will try to delete the archive as group
mailman. If the users don't match (and they usually won't),
permission crash.
This just includes a warning and some helpful information in the
docstring.
|
| |
|
|
|
|
| |
and cron scripts. When code/status == 0, there's no error (it's
likely --help output) so send that to stdout. Otherwise, it's an
error and the output goes to stderr.
|
| |
|
|
| |
since the default is fine.
|
| | |
|
| |
|
|
| |
Closes SF bug #599544 by Tokio Kikuchi.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
of numeric ids. Initial idea and patch by Todd Vierling, fleshed out
by Barry.
Specific changes here:
check_privs(): since we only have the MAILMAN_USER and MAILMAN_GROUP
names here, we need to convert them to their numeric ids via
getgrnam()/getpwnam(). We need the ids because setgid() and setuid()
want ids.
|
| |
|
|
|
|
| |
10). If a qrunner exits with something other than SIGINT, its restart
counter gets incremented and if that is greater than MAX_RESTARTS,
that qrunner won't be restarted.
|
| | |
|
| |
|
|
|
|
|
| |
needs to be prepended with the LOCK_DIR otherwise the unlink will
fail.
Also fixed a typo in the docstring.
|
| |
|
|
|
|
|
| |
the useful data is encoded in the last path component. Otherwise the
calculation gets messed up if the lock file is something like
/usr/local/mailman-2.1/locks/...
|
| |
|
|
| |
signal number from the os.wait() status value.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
flag so that when it's garbage collected, the lock files won't be
removed. Since they're the same as the forced lock's files, we'd be
removing the files at the wrong time.
acquire_lock(): Rearrange the return values so the logic is a little
easier to follow. Move the "return lock" into the try stanza.
|
| |
|
|
|
| |
to run as a mailmanctl subproc (subtly, but usefully changing it's
exit semantics).
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
does the right thing! The sub-qrunners are exec'd now from
bin/qrunner using some new command line options, so killing and
(auto-)restarting them will cleanly reload any changed modules.
Also, the lock is acquired in the foreground so you don't get ugly
error messages if another master qrunner is already running.
Finally, the separate lock-refresher process is gone. Too hard to
implement correct lock ownership transfer semantics with it (it was
more complicated than the simple pass-thru to the child). Now, lock
refresh is implemented by a once-a-day alarm signal in the master
qrunner process watcher.
Also implemented the `reopen' command which causes all the log files
to be re-opened (very useful if you're rotating log files!).
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
into the script's #! line -- based on --with-python settings, without
losing CVS revision history.
#! /usr/bin/env python
becomes
#! @PYTHON@
which gets substituted for in configure.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
stale lock (by matching hostname and pid) it will delete the lock
files before trying to reclaim them.
Also added -q/--quiet to suppress status messages (but not error
messages, which always go to stderr).
sigterm_handler(): We need to catch SIGTERM since that's what init
will send us when changing run levels. SIGTERM just kills all the
sub-qrunners with SIGINT (i.e. no restart).
|
| |
|
|
| |
mailmanctl is doing.
|
| |
|
|
| |
master-qrunner lock could not be acquired.
|
| |
|
|
|
|
|
| |
(concatenated strings are harder for some languages to translate
correctly), and use the usage() method when the check fails.
main(): Add -u / --run-as-user to skip the check_privs() test.
|
| |
|
|
|
| |
complains if we aren't running as the right user, gets called from main().
Should we check for the availability of os.(g|s)et(u|g)id ?
|
| |
|
