Closed aryoda closed 1 year ago
I have added both warnings to the unit test exclusion list since the goal of the unit test is to test only BiT CLI (not the GUI parts which are related to Qt5 and are triggered if the e.g. the system tray plugin is installed).
In the future a refactoring of BiT CLI und Qt (GUI) should be used to improve the unit test isolation to get rid of such side effects.
Also note: A non-english system language may possibly lead to a failed unit test again if Qt5 uses it (not tested)...
I have...
Great!
First I was confused about the workflow and tried to find out where that commit belongs to. ;) Now I get it that you created an extra public upstream branch. IMHO can be merged into "dev".
Also note: A non-english system language may possibly lead to a failed unit test again if Qt5 uses it (not tested)...
To my knowledge because of that pythons unittests by default ignoring system language. So it is always English (or untranslated).
@buhtz
tried to find out where that commit belongs to
yes, I didn't mention that I was using the Github UI directly to create and link the branch in this repo (and delete it again after merging).
IMHO can be merged into "dev".
I'd like to wait for feed-back from the ARCH community who reported this issue and merge then (via PR if possible)
using git PKGBUILD backintime-git still fails on testing: make: *** [Makefile:570: unittest] Error 1
Traceback (most recent call last):
File "/home/joe/.cache/yay/backintime-git/src/backintime/common/test/test_backintime.py", line 176, in test_local_snapshot_is_successful
self.assertRegex(filtered_log_output, re.compile(r'''INFO: Lock
AssertionError: Regex didn't match: 'INFO: Lock\nINFO: Take a new snapshot. Profile: 1 Main profile\nINFO: Call rsync to take the snapshot\nINFO: Save config file\nINFO: Save permissions\nINFO: Create info file\nINFO: Unlock' not found in 'INFO: Lock\nINFO: Take a new snapshot. Profile: 1 Main profile\nINFO: Call rsync to take the snapshot\nINFO: Save config file\nINFO: Save permissions\nWarning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.\nINFO: Create info file\nINFO: Unlock'
using git PKGBUILD backintime-git still fails on testing
Yes, the PKGBUILD does not yet use my fix (commit f056d90) because it is not yet merged.
I will merge it ASAP since the warning message to be ignored is the same...
Update: Pull request created (see below)
The fix is merged into the dev
branch now: https://github.com/bit-team/backintime/commit/66d248fe8f2bb7e1c3189a14d1f73eb54949ef2a
Tried to test it - but I still get the error.
I changed to the latest dev version, installed qt5-wayland
, exported export QT_QPA_PLATFORM=wayland
.
Results are in the log with checks for the variables set and a cat of the file CHANGES
to show that it is the dev
version used for the build.
One of three(?) failing messages:
======================================================================
FAIL: test_local_snapshot_is_successful (test.test_backintime.TestBackInTime.test_local_snapshot_is_successful)
From BIT initialization through snapshot
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/tobiasb/.cache/yay/backintime/src/backintime-dev/common/test/test_backintime.py", line 178, in test_local_snapshot_is_successful
self.assertRegex(filtered_log_output, re.compile(r'''INFO: Lock
AssertionError: Regex didn't match: 'INFO: Lock\nINFO: Take a new snapshot. Profile: 1 Main profile\nINFO: Call rsync to take the snapshot\nINFO: Save config file\nINFO: Save permissions\nINFO: Create info file\nINFO: Unlock' not found in 'INFO: Lock\nINFO: Take a new snapshot. Profile: 1 Main profile\nINFO: Call rsync to take the snapshot\nINFO: Save config file\nINFO: Save permissions\nINFO: Create info file\nWarning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.\nINFO: Unlock'
----------------------------------------------------------------------
Ran 3 tests in 2.844s
Should that work in the way I set up everything? Am I missing something?
@topas-rec THX for your details, I think I have found the problem and will fix the fix.
Edit: The reason is that the additional output of Qt5 that shall be ignored, is prefixed with Warning:
and I forgot to include this in the exclusion list (which is checked via startswith
, not "contains"):
\nWarning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.\n
I was able to build 1.4.1 without problems. Than you for fixing this.
The user @topas-rec reported in https://github.com/bit-team/backintime/issues/1421#issuecomment-1509761121 a failing unit test on ARCH Linux caused by this additional (not expected) warning in the actual output:
Furthermore this warning occurs if the optional package
qt5-wayland
is not installed:System info:
Details
See also
https://aur.archlinux.org/pkgbase/backintime#comment-911755