[ty] Improve invalid-type-form diagnostic where a module-literal type is used in a type expression and the module has a member which would be valid in a type expression
#18244
Conversation
AlexWaygood
added
ty
...ests_for_invalid_ty…_-_Diagnostics_for_comm…_-_Module-literal_used_…_(652fec4fd4a6c63a).snap
Outdated
Show resolved
Hide resolved
|
|
you have to leave some diagnostic improvements to contributors ;) |
…pe is used in a type expression and the module has a member which would be valid in a type expression
AlexWaygood
force-pushed
the
alex/datetime-dot-datetime-diag
branch
from
fdf2569 to
bc9d7f8
Compare
There was a problem hiding this comment.
Not something directly related to this PR, but these info lines are inconsistent in their capitalizations and punctuations.
There was a problem hiding this comment.
And I'd argue that the message is a bit awkwardly phrased. How about this?
Did you mean to use its member
Image.Imageinstead?
There was a problem hiding this comment.
That's a nice shortening, thanks!
There was a problem hiding this comment.
Not something directly related to this PR, but these
infolines are inconsistent in their capitalizations and punctuations.
(I agree that this is inconsistent, but I don't plan to address it in this PR since it predates this PR)
Summary
It's fairly common in Python do to something like this:
when you actually meant to do this:
This PR adds subdiagnostics to hint to the user what the cause of the
invalid-type-formerror might be in cases like this.Closes astral-sh/ty#373
Test Plan
Snapshots