fix mypy on VisionDataset #8134
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/8134
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (3 Unrelated Failures)As of commit 9ae3188 with merge base 83090c2 ( BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
| random_offset: int = 0, | ||
| ) -> None: | ||
| super().__init__(None, transform=transform, target_transform=target_transform) # type: ignore[arg-type] | ||
| super().__init__(transform=transform, target_transform=target_transform) |
There was a problem hiding this comment.
Driveby, since we no longer need to pass None after #8124.
| def __init__( | ||
| self, | ||
| root: Optional[str] = None, | ||
| root: str = None, # type: ignore[assignment] |
There was a problem hiding this comment.
can we still keep the Optional[str] type which is technically correct?
There was a problem hiding this comment.
We can't. If we use Optional[str] as annotation, mypy correctly infers that for the self.root attribute. And with that we are back at the issues we are currently seeing, i.e. all child datasets that perform anything with self.root would now need to check that self.root is in fact a str and not None.
Reviewed By: vmoens Differential Revision: D52539016 fbshipit-source-id: d1726730d1d432b32f1ab36699ec5e361be75652
3 participants
Supersedes #8128. Error caused by #8124.