v0.5.0

• Added checker for multiline list comprehensions ban (#48). A new checker, readability, has been implemented in the PyLint plugin for Databricks, introducing a new code rule R8923 that flags multiline list comprehensions. This change aims to improve code readability by suggesting a rewrite as a for loop when list comprehensions span multiple lines. The checker can be used by adding databricks.labs.pylint.readability to the load-plugins configuration in the pylintrc or pyproject.toml file. The commit also includes updates to the README file, documenting the new checker and the R8923 check, along with examples and instructions for using and disabling it. Additionally, new test functions have been added to test_readability.py to demonstrate the ability to enforce a ban on multiline list comprehensions.
• Bump sigstore/gh-action-sigstore-python from 2.1.1 to 3.0.0 (#47). In version 3.0.0 of sigstore/gh-action-sigstore-python, several changes and bug fixes have been implemented. Notable updates include the addition of recursive globbing with **, and the removal of certain settings such as fulcio-url, rekor-url, ctfe, and rekor-root-pubkey. The output settings signature, certificate, and bundle have also been removed, and inputs are now parsed according to POSIX shell lexing rules. The release-signing-artifacts setting no longer causes a hard error when used under the incorrect event. Additionally, the default suffix has changed from .sigstore to .sigstore.json, and release-signing-artifacts now defaults to true. This version also supports CI runners that use PEP 668 to constrain global package prefixes.

Dependency updates:

• Bump sigstore/gh-action-sigstore-python from 2.1.1 to 3.0.0 (#47).

Contributors: @dependabot[bot], @nfx

v0.4.0

• Added checks for missing mock usage (#44). This change introduces new checks to a PyLint plugin for use with Databricks, specifically targeting the usage of mock objects in Python code. Two new checks, R8921 and R8922, have been developed to ensure the proper assignment and utilization of mock objects. R8921 checks for instances where a mock object is not assigned to a variable, suggesting an assignment to enable assertions. R8922 checks for situations where a mock object is used after creation, recommending an assertion, return value, or side effect for correct usage. To disable these checks on specific lines, users can add comments # pylint: disable=mock-no-assign or # pylint: disable=mock-no-usage. The commit also includes documentation and examples to guide users in implementing and disabling these checks, enhancing the overall quality and reliability of the code.

Contributors: @nfx

v0.3.0

• Added integration with eradicate to highlight dead code (#39). This release integrates the eradicate project, a tool for identifying dead code in Python, into the project's code analysis and testing workflow. This integration will help enforce dead code removal, improve code quality, and make debugging easier by identifying and highlighting dead code. The C8920 check in the Pylint plugin reports any commented out code, and it can be disabled on a specific line by adding # pylint: disable=dead-code at the end of it.

Contributors: @nfx

v0.2.0

• Added documentation on how to enable specific checkers and how to silence specific warnings (#37). The latest release introduces new documentation in the README file for enabling specific checkers and disabling specific warnings for the pylint plugin for Databricks. The plugin offers checkers such as databricks-airflow, databricks-dbutils, databricks-legacy, databricks-notebooks, spark, and mocking, each with unique checks and codes. Users can use these checkers by adding the corresponding module name to the load-plugins configuration in their pylintrc or pyproject.toml file. The release also explains how to disable certain checks on specific lines using a comment with the disable directive followed by the corresponding symbol. Additionally, it includes information on how to use the Databricks SDK instead of internal APIs, legacy CLI, and dbutils. The docs.py script has also been updated with the new instructions on how to enable or disable specific checkers and warnings.

Contributors: @nfx

v0.1.1

• Fixed ToC for mocking checker (#29). In this release, we have made significant improvements to the mocking checker in our open-source library, which is specifically designed for identifying common mistakes and issues in Spark code written in Python. We have added two new rules, R8918: explicit-dependency-required and R8919: obscure-mock, which respectively check for the requirement of explicit dependencies and the use of obscure mocks. This update enhances the accuracy and quality of the code review process for Spark code written in Python, ensuring that the code meets the highest standards of quality and reliability. Additionally, we have fixed an issue in the Table of Contents (ToC) for the mocking checker, making it easier for users to navigate and utilize this feature.
• Fixed project urls for PyPI (#32). In this release, the pyproject.toml configuration file for the Python project has been updated to fix project URLs for PyPI. Specifically, the URLs for the Issues and Source fields have been changed from "https://github.com/databrickslabs/pylint/issues" and "https://github.com/databrickslabs/pylint" to "https://github.com/databrickslabs/pylint-plugin/issues" and "https://github.com/databrickslabs/pylint-plugin", respectively. This modification ensures that users can correctly report issues and access the source code for the pylint-plugin project on GitHub. The rest of the file remains unchanged. This enhancement provides a seamless experience for users who want to contribute to the project or seek support for any issues they encounter.
• Updated README.md to fix pip install command (#30). In this release, we have updated the installation command for the project to fix the pip install process. Previously, users installed the project using the command "pip install pylint-plugin-for-databricks", but this has been changed to "pip install databricks-labs-pylint". This update is part of issue #30 and is intended to improve the installation experience for users. It is important to note that no new methods have been added and no existing functionality has been changed; only the installation command has been updated. Software engineers adopting this project should use the new command to install it.

Contributors: @nfx, @alcole

v0.1.0

• Added checks for mocking best practices (#27). A new mocking checker has been added to the pylint plugin for Databricks to improve the quality of unit tests by enforcing best practices in mocking. This checker introduces two rules: explicit-dependency-required and obscure-mock. The explicit-dependency-required rule encourages injecting dependencies through constructors instead of using mock.patch, promoting code readability, maintainability, and dependency inversion for unit testing. The obscure-mock rule recommends using create_autospec(ConcreteType) over MagicMock to create mock objects that behave like concrete classes, resulting in more robust and maintainable unit tests. These rules can be configured to specify package names containing project code. This addition helps developers create reliable, maintainable, and testable code by adhering to recommended mocking practices.

Contributors: @nfx

v0.0.1

Initial public release

Contributors: