* Start to get rid of pylint; it's more trouble than it's worth and pyflakes

  does a pretty good job anyway.
* Remove master.get_lock_data() now that flufl.lock 2.1 provides the same
  detailed information.
* Add WatcherState.none to indicate that the master is not running.
* Instrument master_state() and acquire_lock_1() for testing, and add
  unittests.
* LBYL for 'bin/mailman start' so that the error message when the master is
  already running happens in the foreground process and is more user friendly.
* Add 'bin/mailman status' to provide master queue runner status on the
  command line.

1 files changed, 37 insertions, 0 deletions

diff --git a/src/mailman/commands/docs/status.txt b/src/mailman/commands/docs/status.txt
new file mode 100644
index 000000000..6deb7fdc0
--- /dev/null
+++ b/src/mailman/commands/docs/status.txt
@@ -0,0 +1,37 @@
+==============
+Getting status
+==============
+
+The status of the Mailman master process can be queried from the command line.
+It's clear at this point that nothing is running.
+::
+
+    >>> from mailman.commands.cli_status import Status
+    >>> status = Status()
+
+    >>> class FakeArgs:
+    ...     pass
+
+The status is printed to stdout and a status code is returned.
+
+    >>> status.process(FakeArgs)
+    GNU Mailman is not running
+    0
+
+We can simulate the master queue runner starting up by acquiring its lock.
+
+    >>> from flufl.lock import Lock
+    >>> lock = Lock(config.LOCK_FILE)
+    >>> lock.lock()
+
+Getting the status confirms that the master queue runner is running.
+
+    >>> status.process(FakeArgs)
+    GNU Mailman is running (master pid: ...
+
+We shutdown the master queue runner, and confirm the status.
+
+    >>> lock.unlock()
+    >>> status.process(FakeArgs)
+    GNU Mailman is not running
+    0