Since Python is an interpreted language, it's easy to introduce errors which even a reasonably thorough testing procedure wouldn't detect.įor example, consider this simple refactoring: pep8-naming, which checks names against the PEP8 naming conventions.Ĭode that passes Flake8 is free of a range of issues that would otherwise not show up until later in the development cycle, which saves a lot of time.flake8-requirements, which checks for missing or unused dependencies, and.flake8-docstrings, which uses pydocstyle to enforce documentation presence and style,.

flake8-comprehensions, which checks for issues around comprehensions, generators, and initialisation,.

flake8-bugbear, which has rules that are considered too subjective for the main Flake8 distribution,.

There is an ecosystem of plugins which add further checks or integrate other tools, notably: Step 1: catching bugs and anti-patternsįlake8 is a Python linter: a tool which checks Python code for errors and evaluates it for quality and style. This post outlines a process for gradually introducing such tooling into an established project. Modern tooling can free you from having to think about all sorts of menial tasks, letting you write better code with less effort.

If you're reading this in 2021 (or later) you should probably use poetry instead of pip and configure your tools with pyproject.toml instead of setup.cfg, but otherwise this is still pretty good material.Īs a software developer, your time, focus, and mental stamina are precious resources. Updated : Just fixed a few typos and made some minor updates.

Step 1: catching bugs and anti-patterns.

Writing better Python code, automatically

Visual studio code python pyre