Conversation
| create_resource_server(app, use_cache, use_cache) | ||
| url = "/user" | ||
|
|
||
| # case 1 |
There was a problem hiding this comment.
Suggestion: Instead of having case 1... case 5 all in one test, use pytest.mark.parametrize() to separate these into individual test cases.
There was a problem hiding this comment.
I like pytest parametrize, but in that particular case pytest it a bit overkill to me. There are 3 vars to parametrize (url, error, error_description) for a 4 lines test. Anyway, I think both solutions would be fine.
I did not put a lot of factorization effort into OAuth1, but feel free to open a PR if you want to work on this.
azmeuk
changed the title
azmeuk
force-pushed
the
pytest-paradigm
branch
from
7b345a4 to
765867c
Compare
azmeuk
force-pushed
the
pytest-paradigm
branch
from
765867c to
273a85e
Compare
|
You don't need to migrate tests for jose module. Since we are going to remove jose module from authlib, we will also remove its tests. |
I know but that was a really quick thing to do. |
3 participants
What kind of change does this PR introduce?
This migrates all the tests from the unittest paradigm to the pytest paradigm:
load_configmethods.This makes the unit tests a lot more legible, and a bit shorter. There are still 748 tests as before.
Follow-up to #746
Checklist
prek.pragma: no cover