GUI tests #37
GUI tests #37
Conversation
|
@maresb please note that this branch is based on the |
tomers
force-pushed
the
tshalev-gui-tests
branch
from
df06643 to
12ec427
Compare
|
@tomers, any ideas what's going wrong here? It appears that the Python process inside tox is crashing for some reason. |
I suspect this: https://pytest-qt.readthedocs.io/en/latest/troubleshooting.html#tox-invocationerror-without-further-information
|
|
Nice. Also, we don't need to be using tox. It may be less trouble to switch to raw |
tomers
force-pushed
the
tshalev-gui-tests
branch
5 times, most recently
from
3ae54e2 to
bee2eb1
Compare
|
@maresb I think this PR is ready to be merged :-) |
|
Woah, great work!!! Go ahead and merge. For future reference could you please add a comment here about the references you used for:
I don't know where they came from, and they may need to be adjusted in the future, so I'd just like to gather the info while it's still fresh in your head. |
|
I'm really impressed, this looked like a hard one, so I wasn't expecting it to get done today. Very nice!!! |
Just for good measure. This will allow use to add other OSes in the future, in case needed.
According to pytest-qt's documentation [1]: > It might happen that your tox run finishes abruptly without any useful information, e.g.: > ... > pytest-qt needs a DISPLAY to run, otherwise Qt calls abort() and the process crashes immediately. To mitigate this, I've followed instructions in [2]. [1] https://pytest-qt.readthedocs.io/en/latest/troubleshooting.html#tox-invocationerror-without-further-information [2] https://pytest-qt.readthedocs.io/en/latest/troubleshooting.html#github-actions-azure-pipelines-travis-ci-and-gitlab-ci-cd
The GUI is rather useless when the app is being run in the test environment, as there are no printers connected (there is no USB backend at all). This simple test verifies the most basic existence of selected widget, and their content. We will need to figure out how to test the app with a state that resembles actual printer. We might need to introduce test mode, in which all supported printers are shown in the device selector, so we can test various features for each.
The comment in the second commit gives the proper reference. I've committed some changes in other commit by mistake, and now I've arranged it properly, so all dependency handling is done in a single commit, with proper documentation. Thanks for your kind words! |
The GUI is rather useless when the app is being run in the test environment, as there are no printers connected (there is no USB backend at all).
This simple test verifies the most basic existence of selected widget, and their content.
We will need to figure out how to test the app with a state that resembles actual printer. We might need to introduce test mode, in which all supported printers are shown in the device selector, so we can test various features for each.