This dependency is only needed if USING_STATSD is set, and so had
apparently been missed from our dependency list previously (since few
sites use that setting).
There is a hard to reproduce race condition causing these tests to
occasionally fail. We believe it is caused by switching to the home tab and
not properly waiting for all the messages to load; see Issue #1243. The
tests are for the following pathway (not a high priority to test):
1. User starts editing a message.
2. allow_message_editing is turned off for the realm (in this case, by the
user going to the admin page and turning it off).
3. User finishes editing the message and hits send.
Running `./manage.py email-mirror` used to fail on python 3
because twisted.mail.imap4 is not python 3 compatible.
Display a message informing the user that email-mirror is not
available on python 3 instead of failing with a traceback.
Also add tools/test-management to py3-backend.
runtornado unbuffers its output using
sys.stdout = os.fdopen(sys.stdout.fileno(), 'w', 0).
This is not python 3 compatible since we can't specify
buffering on a text stream in python 3. So use the '-u'
option of python when calling runtornado.py to make output
unbuffered.
There were a bunch of authorization and well-formedness checks in
zerver.lib.actions.do_update_message that I moved to
zerver.views.messages.update_message_backend.
Reason: by convention, functions in actions.py complete their actions;
error checking should be done outside the file when possible.
Fixes: #1150.
This is controlled through the admin tab and a new field in the Realms table.
Notes:
* The admin tab setting takes a value in minutes, whereas the backend stores it
in seconds.
* This setting is unused when allow_message_editing is false.
* There is some generosity in how the limit is enforced. For instance, if the
user sees the hovering edit button, we ensure they have at least 5 seconds to
click it, and if the user gets to the message edit form, we ensure they have
at least 10 seconds to make the edit, by relaxing the limit.
* This commit also includes a countdown timer in the message edit form.
Resolves#903.
Taiga's webhook integration would give output events in a random
order which caused test failures on python 3 (seems like python
3 is more prone to non-deterministic failures). Fix that by
sorting the outputs obtained from events before concatenating them.
Use ujson.dumps to render raw messages sent by the PagerDuty
integration instead of using pprint.pformat. pprint.pformat
gives different results on python 2 and 3.
Correctly encode and decode strings in convert_html_to_markdown.
It wasn't possible to use universal_newlines=True since
Popen.communicate doesn't encode/decode strings correctly on
python 2.
There is a hard to reproduce race condition causing these tests to
occasionally fail. We believe it is caused by switching to the home tab and
not properly waiting for all the messages to load; see Issue #1243. The
tests are for the following pathway (not a high priority to test):
1. User starts editing a message.
2. allow_message_editing is turned off for the realm (in this case, by the
user going to the admin page and turning it off).
3. User finishes editing the message and hits send.
Tim Abbott
Rishi Gupta
Eklavya Sharma
Steve Howell
Rishi Gupta
Alex Gaynor
Umair Khan
Steve Howell
Preston Hansen
Tomasz Kolek
Taranjeet Singh