[stubtest] support @type_check_only decorator
#16422
Conversation
for more information, see https://pre-commit.ci
Co-authored-by: Alex Waygood <Alex.Waygood@Gmail.com>
test-data/unit/check-classes.test
Outdated
There was a problem hiding this comment.
Not sure I fully understand what the point is of adding the tests you've added in this file, since they all seem to be asserting undesirable behaviour
There was a problem hiding this comment.
It is not undersirable, it just a status-quo. It is expected to be working like this for now. We don't have any other expected behavior for now.
There was a problem hiding this comment.
Yes, and the status quo is undesirable -- ideally, mypy would fully support @type_check_only. If somebody came along and implemented the feature in the future, I think it would be highly confusing if existing tests started failing as a result of the feature being implemented
There was a problem hiding this comment.
Ok, we now know that these tests work and can add them later.
I am actually interested in working on this further :)
This comment has been minimized.
This comment has been minimized.
There was a problem hiding this comment.
Do we need the changes to the fixtures anymore, after the latest push?
There was a problem hiding this comment.
It is typing-full, it won't hurt.
This comment has been minimized.
This comment has been minimized.
|
Diff from mypy_primer, showing the effect of this PR on open source code: discord.py (https://github.com/Rapptz/discord.py): typechecking got 1.07x slower (148.4s -> 159.0s)
(Performance measurements are based on a single noisy sample)
|
There are several
TODOitems for the future (not this PR):@type_check_only@overloaded functions. But, how? There are two options: we can treat individual overload cases as@type_check_onlyor we can treat the whole func. Sincetypesheddoes not have any examples of this, I prefer to defer this discussion to somewhere else and support this when we decideRefs #15146