As for wildcard import PEP8 <https://www.python.org/dev/peps/pep-0008/#imports> does not favor it and nor do I for the reasons listed in PEP8 and those that Didi listed. As for having a single file with all imports that would be imported by conftest, I vote against it because this is similar in a way to importing with wildcards - a suite will import unused stuff creating all the same unfavorable effects. As for importing directly inside the test code I also vote against it, because - as Didi noted above - there is no single point of access for imports, posing a risk for collisions, problem to override, and need to update in several places. So by process of elimination what is left AFAIS is to concentrate all the imports in conftest. Thanks On Mon, Jun 21, 2021 at 7:03 PM Yedidyah Bar David <didi@redhat.com> wrote:
Hi all,
We have several different styles of where/how to import fixtures in ovirt-system-tests:
- import directly inside the test code - import in conftest.py - import in fixtures modules - For all of above, both 'import *' and importing specific fixtures
I think we should try to agree on a specific style and then follow it.
One drawback of importing directly in test/fixtures code is that it's then impossible to override them in conftest.py.
A drawback of importing '*' and/or doing this in conftest.py is that you might inadvertently import more than you want, or this might happen eventually (after more stuff are added), that this makes it harder to find what uses what, and that it risks unintended collisions in names - as opposed to intended overrides.
A related issue is having to update many places if you add/change something.
If there is some kind of "best practices" document somewhere that people are happy with, perhaps we should follow it. Otherwise, we should come up with our own.
Personally I think I'd like to have a single file with all the imports, of specific fixtures (not '*'), and import this file from conftest.py of all the suites. Didn't actually try this and no idea what complications it might bring.
Comments/ideas/opinions/decisions are welcome :-)
Best regards, -- Didi _______________________________________________ Devel mailing list -- devel@ovirt.org To unsubscribe send an email to devel-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/devel@ovirt.org/message/VAOD5Q7WU2FC66...