Fix intermittent ruff I001 error for pathfinder _version import
#1460
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add # isort: skip comment to the _version import in cuda_pathfinder/cuda/pathfinder/__init__.py to prevent ruff from inconsistently reordering the import based on whether _version.py exists. Problem: Ruff's import sorting (I001) behavior is inconsistent when imports cannot be resolved. The _version.py file is generated by setuptools-scm during the build process, so its existence depends on build state: - When _version.py EXISTS (after a build): Ruff can resolve the import and wants to sort it alphabetically. Since _version comes after _headers alphabetically, ruff expects the import order: _dynamic_libs, _headers, _version. - When _version.py DOESN'T EXIST (after git clean -fdx or fresh clone): Ruff cannot resolve the import and may skip checking it entirely, especially if cached as "unresolvable". This leads to inconsistent behavior where the same code passes or fails depending on build state. Root Cause: Ruff caches import resolution results. If the cache has stale data where _version.py didn't exist, ruff may skip checking that import. When the cache is refreshed or _version.py exists, ruff resolves it and wants to sort it, triggering I001 errors inconsistently. Solution: Add # isort: skip comment to explicitly tell ruff to skip sorting this import, regardless of file existence or cache state. This ensures consistent behavior in all scenarios: - After builds (when _version.py exists) - After git clean (when _version.py doesn't exist) - With fresh clones - With stale ruff cache The import is placed after the other imports (where it logically belongs alphabetically) with the skip directive to prevent ruff from moving it.
Contributor
|
Auto-sync is disabled for ready for review pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
Collaborator
Author
|
pre-commit.ci run |
Collaborator
Author
|
/ok to test |
|
mdboom
approved these changes
Jan 12, 2026
Contributor
mdboom
left a comment
mdboom
left a comment
There was a problem hiding this comment.
LGTM. This is sure an annoying aspect of setuptools-scm.
github-actions bot
pushed a commit
that referenced
this pull request
Jan 13, 2026
Removed preview folders for the following PRs: - PR #1460
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add
# isort: skipcomment to the_versionimport incuda_pathfinder/cuda/pathfinder/__init__.pyto prevent ruff from inconsistently reordering the import based on whether_version.pyexists.Problem
Ruff's import sorting (I001) behavior is inconsistent when imports cannot be resolved. The
_version.pyfile is generated bysetuptools-scmduring the build process, so its existence depends on build state:When
_version.pyEXISTS (after a build):Ruff can resolve the import and wants to sort it alphabetically. Since
_versioncomes after_headersalphabetically, ruff expects the import order:_dynamic_libs,_headers,_version.When
_version.pyDOESN'T EXIST (aftergit clean -fdxor fresh clone):Ruff cannot resolve the import and may skip checking it entirely, especially if cached as "unresolvable". This leads to inconsistent behavior where the same code passes or fails depending on build state.
Root Cause
Ruff caches import resolution results. If the cache has stale data where
_version.pydidn't exist, ruff may skip checking that import. When the cache is refreshed or_version.pyexists, ruff resolves it and wants to sort it, triggering I001 errors inconsistently.Example (before this PR):
Solution
Add
# isort: skipcomment to explicitly tell ruff to skip sorting this import, regardless of file existence or cache state. This ensures consistent behavior in all scenarios:_version.pyexists)git clean(when_version.pydoesn't exist)The import is placed after the other imports (where it logically belongs alphabetically) with the skip directive to prevent ruff from moving it.
Changes
_versionimport from top of file to after_headersimports# isort: skipcomment to prevent ruff from reordering the importNote: This PR was extracted from PR #1454