Pre-commit Hooks for Python Devs

Episode Deep Dive

Guest Introduction and Background

Stefanie Molin is a software engineer at Bloomberg who comes from a data analytics background, first working in R before moving over to Python. She has authored a second edition of Hands-On Data Analysis with Pandas, contributed significantly to the NumPy doc ecosystem, and has been a frequent conference speaker at events like EuroPython and PyCon. Stefanie’s focus on automated code quality checks led her to develop and share various pre-commit hooks, including a specialized one for NumPy docstrings and another for stripping EXIF metadata from images.

What to Know If You're New to Python

• Check out our course Up and Running with Git: While you don’t need deep Python expertise to use pre-commit hooks, having basic Git and Python knowledge will help you configure them effectively.

Key Points and Takeaways

1. Why Pre-Commit Hooks Matter
   They run checks on your code before it’s committed, saving time by catching formatting or logic issues early. This approach ensures team-wide code consistency and reduces friction in code reviews.

   • pre-commit.com
   • Git Hooks Documentation (git-scm.com)

2. Getting Started with Pre-Commit
   Installing the pre-commit Python package is straightforward. Once installed, you add a .pre-commit-config.yaml file to your repo, and run pre-commit install to wire everything up.

   • pre-commit on PyPI
   • Stefanie’s blog / website

3. Configuration and Workflow
   A typical .pre-commit-config.yaml specifies the Git repository hosting the hooks, their IDs, and any additional arguments. This file is checked into source control, so everyone on the project shares the same setup.

   • Example tools mentioned:
     • Ruff (GitHub)
     • Black (GitHub)

4. Creating Your Own Pre-Commit Hooks
   Stefanie outlined a four-step approach: write your core logic, build a CLI with argparse or similar, make it installable, and define it in a pre-commit-hooks.yaml. This pattern allows for easy sharing and installation.

   • NumPy doc pre-commit hook (in NumPy doc repo)
   • Pre-commit Hooks Repo

5. Automation vs. Checks
   If a hook can automatically fix an issue—like adding missing trailing commas—implement that rather than just failing. Reducing busywork keeps developers from bypassing the hooks.

   • Tools with auto-fix features:
     • Black
     • Ruff (--fix flag)

6. EXIF Stripper Example
   Stefanie created a hook that removes sensitive or unnecessary metadata from images before they’re committed. This prevents accidental data leaks (GPS coordinates, photographer details) and also reduces file size.

   • Pillow (Python Imaging Library)

7. Docstring Validation and NumPy Doc
   Contributing to scikit-learn sprints introduced Stefanie to the importance of docstring style and consistency. Her NumPy doc pre-commit hook automatically checks for trailing spaces and adherence to NumPy-style docstrings.

   • scikit-learn.org
   • Hands-On Data Analysis with Pandas

8. Continuous Integration and Pre-Commit
   While pre-commit gives immediate local feedback, CI ensures your entire team consistently meets code quality standards. Setting up the same checks in CI builds keeps you from having surprises at merge time.

   • Example: Combine GitHub Actions with pre-commit to run the same hooks in pull requests.

Interesting Quotes and Stories

• EXIF data realization: "I remember discovering the camera details, location, and the photographer’s name were all in the image. I didn’t want that on my blog!" – Stefanie Molin
• Auto-fix perspective: "If you can fix it, fix it. Don’t just warn about it." – Stefanie Molin on saving developers from repetitive tasks.

Learning Resources

• Stefanie’s Website: stefaniemolin.dev for blog posts, tutorials, and future talks.
• NumPy Doc: numpydoc.readthedocs.io for style guidelines and docstring best practices.
• Pre-commit: Pre-commit official site and GitHub repo.
• Hands-On Data Analysis with Pandas by Stefanie Molin (Second Edition) for in-depth data wrangling techniques.

Overall Takeaway

Implementing and customizing pre-commit hooks in your Python workflow can have a massive impact on code consistency and developer efficiency. From validating docstrings to stripping sensitive metadata, these checks enforce best practices at the earliest stage—saving time, preventing errors, and allowing teams to focus on what truly matters: writing great code.