Open-MSS / MSS / 7696139441
70%

DEFAULT BRANCH: develop
Ran
Jobs 1
Files 103
Run time 8s
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 69.106% (-0.3%) from 69.405%
7696139441

push

github

web-flow 
Remove test retries in CI (#2153)

* Do not retry failed tests in CI

Tests that do not consistently pass are broken tests and should be
fixed. Failed tests should not be retried to avoid accidentally adding
new broken tests.

* Remove some race conditions in tests

Most Qt related actions are asynchronous by design. Just doing them and
then asserting the result therefore is a race condition. Having a qWait
or processEvents call is also not enough, since it is not guaranteed
that they will ensure the Qt action to have happened. A qWait will also
always be either too long or too short.

The pytest-qt plugin instead provides multiple different wait methods
that wait until a given event has happened (e.g. a signal was emitted or
a assertion callback no longer errors) or a timeout is reached, giving
the Qt event loop time to do its thing in between the checks.  These
methods do not suffer from the same race conditions and should be used
instead.

* Increase sio.sleep duration

There was an issue in CI where it was possible that the get_messages
call for the current timestamp returned both messages instead of none.
Maybe this was because not enough time elapsed between creating the
messages and getting them.

* Add MSCOLAB_SSO_DIR to generated test config

This should fix a bug in CI where it was possible that this path was
still in the home directory instead of in a temporary one.

* Skip problematic tests

* Give Qt some time to handle its events

In these tests in test_mscolab_admin_window.py it is possible that a
late signal triggers a call to render_new_permission, at which point the
mscolab_server_url is unset, which in turn leads to a TypeError in the
tests teardown. Adding a qWait at the ends of these tests should reduce
the likelihood of this issue, but won't eliminate it completely. I have
no idea what to wait for instead to fix this properly though.

13618 of 19706 relevant lines covered (69.11%)

0.69 hits per line

Jobs
ID Job ID Ran Files Coverage
1 7696139441.1 0
69.11
Source Files on build 7696139441
Detailed source file information is not available for this build.