Remove black and bandit from pre-commit in favor of ruff

PGijsbers commented 3 weeks ago

It's just fewer dependencies. There might be small disparities between bandit and ruff, but it seems simpler to use just one tool. (and it's faster)

Summary by Sourcery

Replace 'black' and 'bandit' with 'ruff' in pre-commit configuration to reduce dependencies and improve performance. Enhance type handling in conversion functions and simplify union type annotations. Clean up unused code and comments, and update test cases for improved readability and compliance with linting rules.

Enhancements:

Replace usage of 'dict' and 'list' with 'Mapping' and 'Iterable' in conversion functions for better type handling.
Simplify union type annotations in Pydantic models using the '|' operator.
Remove unused code and comments across various modules to improve code clarity.
Use 'http.client.OK' instead of hardcoded status codes in test assertions for better readability.
Refactor global variable usage in database setup functions to include 'noqa' comments for linting compliance.

Build:

Remove 'black' and 'bandit' from pre-commit configuration in favor of using 'ruff' for linting and formatting.

Tests:

Update test cases to use 'http.client.OK' for status code assertions.
Modify test logic to use 'next(iter(...))' for accessing elements in sets or lists.

sourcery-ai[bot] commented 3 weeks ago

Reviewer's Guide by Sourcery

This pull request removes black and bandit from pre-commit hooks in favor of using ruff for both linting and formatting. It also includes various code improvements and type hint updates across multiple files.

File-Level Changes

Change	Details	Files
Replace black and bandit with ruff in pre-commit configuration	Remove bandit hook Remove black hook Update ruff hook to include formatting Add ruff-format hook	`.pre-commit-config.yaml`
Update type hints and improve code quality	Replace specific type checks with more generic ones (e.g., dict to Mapping, list to Iterable) Use union syntax (	) instead of Union[] Remove redundant nested_int_to_str function Update function annotations and disable some linter warnings Replace httpx.codes.OK with http.client.OK	`src/core/conversions.py` `src/schemas/datasets/dcat.py` `src/routers/openml/datasets.py` `src/routers/openml/tasks.py` `tests/routers/openml/study_test.py` `tests/routers/openml/flows_test.py`
Improve code organization and imports	Move imports to TYPE_CHECKING block Remove commented-out code Update import statements	`src/schemas/datasets/mldcat_ap.py` `src/schemas/datasets/openml.py` `src/schemas/flows.py`
Address linter warnings and improve code quality	Add noqa comments to suppress specific linter warnings Rename _get_for_datasets to get_for_datasets Update global variable declarations Simplify list creation syntax	`src/database/qualities.py` `src/database/setup.py` `tests/routers/openml/migration/datasets_migration_test.py` `src/database/users.py` `src/routers/openml/tasktype.py`

Sequence Diagram

No sequence diagram generated.

Tips

- Trigger a new Sourcery review by commenting `@sourcery-ai review` on the pull request. - Continue your discussion with Sourcery by replying directly to review comments. - You can change your review settings at any time by accessing your [dashboard](https://app.sourcery.ai): - Enable or disable the Sourcery-generated pull request summary or reviewer's guide; - Change the review language; - You can always [contact us](mailto:support@sourcery.ai) if you have any questions or feedback.

PGijsbers commented 3 weeks ago

Nevermind. I am stupid, or tired, or both.

openml / server-api