spurious test failures regarding LPModerate and XMLRPCRunner
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Launchpad itself |
Fix Released
|
Undecided
|
Martin Pool |
Bug Description
Running './bin/test -t mail' in a recent branch of Launchpad produces the following spurious failures:
https:/
mbp@lp-lucid% ./bin/test -t incoming
Test-module import failures:
Module: lp.services.
Traceback (most recent call last):
File "/home/
module = original_
File "/home/
from Mailman.Handlers import LPModerate
ImportError: cannot import name LPModerate
Module: lp.services.
Traceback (most recent call last):
File "/home/
module = original_
File "/home/
from lp.services.
File "/home/
module = original_
File "/home/
from Mailman.Queue import XMLRPCRunner
ImportError: cannot import name XMLRPCRunner
This made me miss some other actually important failures in testing lp:~mbp/launchpad/dkim
Related branches
- Barry Warsaw (community): Approve
- Gavin Panella (community): Approve
-
Diff: 23 lines (+6/-1)1 file modifiedbuildmailman.py (+6/-1)
affects: | launchpad → launchpad-registry |
tags: | added: qa-untestable |
Changed in launchpad: | |
status: | Fix Committed → Fix Released |
Changed in launchpad: | |
milestone: | none → 11.01 |
The problem seems to be that lp.services. mailman. monkeypatches. monkey_ patch() has not been run. Despite the name, this doesn't monkey patch in the usual Python sense by tweaking in-memory structures, but rather it patches a mailman source tree.
This is called by ./buildmailman.py and also lp/services/ mailman/ runmailman. py. I guess my tree should have had the former run, but it has not.
I think the reason behind _that_ may be that 'make compile' has some declared dependencies but they may not be enough to make sure it runs. I wonder how I ever got a mailman tree that wasn't patched though.