Looks good overall so far!

I'm also biased against classes in python. See also the last sentence of the Zen of Python

😁 that's great I'm glad, I was afraid these changes would be too much. I've only left the 2 classes intest_base.py for now as it's a small file and didn't seem to make sense to split it yet.

I realized I might be getting myself into a continuous rebase party as new PR's with tests in the old files are merged, so I've cleaned it up a bit more and I think I want to look at fixtures in a later PR when I have more time, if this is in an OK shape?

@nejch Sorry totally missed this. Does one only see assignments and not reviewers in the feed?

@nejch I would love to merge this. Could you please rebase to latest master?

@nejch I would love to merge this. Could you please rebase to latest master?

Sorry@max-wittig I'm back from the dead/vacation/other distractions now 😁 I've rebased and rewritten the tests using responses now so it's more consistent with your runners example. Could you have another look to see if this is how you imaginedresponses-based tests?

Not sure about notifications, there's a separate Notification setting for PR Reviews so maybe you had them turned off?

We actually have less tests with this MR. I guess some got lost. Could yout take a look?@nejch

155 vs. 158

This MR:https://travis-ci.org/github/python-gitlab/python-gitlab/jobs/720445167
Master:https://travis-ci.org/github/python-gitlab/python-gitlab/jobs/720460657

We actually have less tests with this MR. I guess some got lost. Could yout take a look?@nejch

155 vs. 158

This MR:https://travis-ci.org/github/python-gitlab/python-gitlab/jobs/720445167
Master:https://travis-ci.org/github/python-gitlab/python-gitlab/jobs/720460657

Oof, I was weirded out by this until I realized that on master, 2 test cases (intest_groups.py) were run 3 times. They're duplicates because I had accidentally put them in a unittest base class (TestGroup) inherited by other test classes back when I was adding group export/import tests. 🤦
With pytest they're run only once, hence the lower number. If you excludetest_groups.py the number should be higher. But I checked all cases collected in master vs. this PR again and it should be OK otherwise.

Withtox -e py38 -- -v:

master:

gitlab/tests/objects/test_groups.py::TestGroup::test_create_group PASSED [ 75%]gitlab/tests/objects/test_groups.py::TestGroup::test_get_group PASSED    [ 75%]gitlab/tests/objects/test_groups.py::TestGroupExport::test_create_group PASSED [ 76%]gitlab/tests/objects/test_groups.py::TestGroupExport::test_create_group_export PASSED [ 76%]gitlab/tests/objects/test_groups.py::TestGroupExport::test_download_group_export PASSED [ 77%]gitlab/tests/objects/test_groups.py::TestGroupExport::test_get_group PASSED [ 78%]gitlab/tests/objects/test_groups.py::TestGroupExport::test_refresh_group_export_status SKIPPED [ 78%]gitlab/tests/objects/test_groups.py::TestGroupImport::test_create_group PASSED [ 79%]gitlab/tests/objects/test_groups.py::TestGroupImport::test_get_group PASSED [ 80%]gitlab/tests/objects/test_groups.py::TestGroupImport::test_import_group PASSED [ 80%]gitlab/tests/objects/test_groups.py::TestGroupImport::test_refresh_group_import_status SKIPPED [ 81%]

refactor:

gitlab/tests/objects/test_groups.py::test_get_group PASSED               [ 56%]gitlab/tests/objects/test_groups.py::test_create_group PASSED            [ 56%]gitlab/tests/objects/test_groups.py::test_create_group_export PASSED     [ 57%]gitlab/tests/objects/test_groups.py::test_refresh_group_export_status SKIPPED [ 57%]gitlab/tests/objects/test_groups.py::test_download_group_export PASSED   [ 58%]gitlab/tests/objects/test_groups.py::test_import_group PASSED            [ 58%]gitlab/tests/objects/test_groups.py::test_refresh_group_import_status SKIPPED [ 59%]

Thank you very much! Very nice!