Exploring Code Reviews in Python and Automating the Process

What goes into a code review in Python? Is there a difference in how a large organization practices code review compared to a smaller one? What do you do if you’re a solo developer? This week on the show, Brendan Maginnis and Nick Thapen from Sourcery return to talk about code review and automated code assistance.

Nick and Brendan discuss their personal experiences with code reviews. From working together at a large company to starting Sourcery as a pair of developers to growing a team, they’ve experienced how various organizations review code.

We discuss the importance of keeping code reviews positive and maintaining developer momentum. Nick and Brendan share techniques for automating organizational code rules and retaining institutional developer knowledge.

They’ve been busy developing new features for Sourcery, moving beyond the current refactoring features. Their team is leveraging LLMs to build an automated code assistant that can perform code reviews, write tests, and answer questions about your codebase. The goal is for it to act as a personalized computer pair programmer.

Course Spotlight: Using the bpython Enhanced REPL

In this video course, you’ll learn about bpython, an alternative Python REPL that brings code suggestions and many other IDE-like features to the terminal. Once you discover how much bpython can improve your productivity, you’ll never want to return to using the vanilla Python REPL again.

Topics:

• 00:00:00 – Introduction
• 00:02:41 – Why discuss code reviewing?
• 00:04:16 – What were your first code reviews like?
• 00:06:31 – Developing positive techniques as a team
• 00:14:07 – Time between reviews
• 00:17:04 – Will the techniques vary based on the organization size?
• 00:18:43 – Things to standardize and remove from code review
• 00:20:06 – Video Course Spotlight
• 00:22:06 – Creating a tool that goes beyond refactoring
• 00:23:42 – Do newer models provide better context for code review?
• 00:25:52 – Working toward an automated pair programmer
• 00:33:42 – Creating rules for an organization’s code standards
• 00:42:55 – Comparing completion tools with code assistance
• 00:48:18 – Working on an existing codebase vs a new project
• 00:52:11 – Current state of Sourcery
• 00:56:19 – What are you excited about in the world of Python?
• 01:00:12 – What do you want to learn next?
• 01:03:41 – How can people follow your work online?
• 01:05:17 – Thanks and goodbye

Show Links:

• Sourcery 1000x Faster Code Review
• Sourcery - GitHub Bot Install
• Sourcery - VSCode Extension - Visual Studio Marketplace
• Sourcery - IntelliJ IDEs Plugin
• Episode #55: Getting Started With Refactoring Your Python Code – The Real Python Podcast
• How to do a code review - eng-practices - Google GitHub
• Generating Code without Generating Technical Debt?
• Episode #117: Measuring Python Code Quality, Simplicity, and Maintainability – The Real Python Podcast
• Writing Custom Rules - Sourcery Documentation
• Nuitka the Python Compiler - Documentation
• Episode #154: Targeting WebAssembly Platforms & Distilling a Minimum Viable Python – The Real Python Podcast
• Rust Programming Language
• Speedsolving the cube : easy-to-follow, step-by-step instructions for many popular 3-D puzzles in SearchWorks catalog
• My Arcade Releases the Atari Gamestation Pro with 200+ Games

Level up your Python skills with our expert-led courses:

• Using the bpython Enhanced REPL
• Testing Your Code With pytest
• Python Basics Exercises: Strings and String Methods

Support the podcast & join our community of Pythonistas