Reintroduce PytestReturnNotNoneWarning #13495
Conversation
psf-chronographer
bot
added
the
bot:chronographer:provided
|
Do we consider this a bugfix that is OK to enter in the next patch release, or should we wait for |
|
We should backport as we correct a regression |
Since this warning is meant to be permanent, added proper documentation to the `assert` section in the docs. Fixes pytest-dev#13477
nicoddemus
force-pushed
the
13477-none-warning
branch
from
3c8dff2 to
1e1c68d
Compare
nicoddemus
added
the
backport 8.4.x
|
I'm getting the same failure on Testing |
|
Blocked by #13497, will wait for us to fix that before merging this. |
|
The related issue is possibly something larger to fix so id prefer we dont block on it |
| ) | ||
| def test_foo(a, b, result): | ||
| assert foo(a, b) == result | ||
There was a problem hiding this comment.
We should maybe have some blurb on why you might want to return values?
There was a problem hiding this comment.
I believe this will only confuse readers. The official pytest documentation should continue to advise against returning anything from test functions. If plugins wish to implement their functionality using return values, they can describe their rationale in their own documentation.
There was a problem hiding this comment.
This looks good -- thanks! Apologies if this comment is off the mark, but do we need to separately handle the await case (#11372) that was the original rationale for changing the warning to an error?
If I understand correctly your question, no. The state of affairs will be:
|
Backport to 8.4.x: 💚 backport PR created✅ Backport PR branch: Backported as #13527 🤖 @patchback |
Since this warning is meant to be permanent, added proper documentation to the `assert` section in the docs. Fixes #13477 (cherry picked from commit 53f05c4) Co-authored-by: Bruno Oliveira <[email protected]>
Since this warning is meant to be permanent, added proper documentation to the
assertsection in the docs.Fixes #13477