[Python] Python DevOps in Neovim: Tooling with mini.nvim [closed]

I am setting up a Python development environment in Neovim and aiming for a comprehensive setup including static checking, linting, auto-formatting, autocompletion, warnings, testing and debugging. I'm specifically looking for recommendations on tools that integrate well with mini.nvim.

I've watched two YouTube videos that provide different approaches, and I'm looking for guidance on the best combination of tools, especially considering potential overlaps and more modern alternatives.

Here's a summary of the tools and configurations discussed in each video:

Video 1: "Neovim as a Python IDE" (https://youtu.be/4BnVeOUeZxc)


This video primarily uses NvChad as a base configuration and focuses on the following tools:

• Code Suggestions & Autocomplete & Warnings:
  • Pyright (Language Server Protocol - LSP) for code suggestions, autocompletion, and warnings.
  • Mentioned a fork: DetachHead/basedpyright for type checking improvements and Pylance features.
• Static Type Checking, Linting, and Auto-formatting:
  • null-ls (now none-ls) plugin to integrate non-LSP tooling.
  • Mypy for static type checking.
  • Ruff for Python linting.
  • Black for auto-formatting.
• Debugging:
  • nvim-dap (Debug Adapter Protocol) with nvim-dap-python and debugpy.
  • nvim-dap-ui for visual debugging.
  • Tree-sitter for Python syntax understanding.

Video 2: "Neovim Python Development with Kickstart" (https://youtu.be/IobijoroGE0)


This video uses the Kickstart starter configuration and highlights different tools or approaches for similar functionalities:

• Auto-formatting and Real-time Linting:
  • none-ls (fork of null-ls) and none-ls-extras.
  • Ruff for fast Python formatting and linting (replaces Black, Flake8, isort).
  • Prettier for JSON, YAML, and Markdown formatting.
  • shellformat for shell scripts.
• LSP Integration (Autocompletion, Hover, Go to Definition):
  • pylsp (Python Language Server) was chosen over Pyright due to past memory issues.
• Debugger Setup:
  • nvim-treesitter for Python.
  • nvim-dap-python for debugging.

My Question:


Given these two sets of recommendations, what are the most effective and compatible tools I should include in my Neovim setup, specifically with mini.nvim, to achieve a robust Python development environment with all the desired DevOps features (static checking, linting, auto-formatting, autocompletion, warnings, and debugging)?

I'm particularly interested in:

• Recommendations for a single, comprehensive LSP (e.g., Pyright vs. pylsp, or the basedpyright fork).
• The best approach for linting and formatting (e.g., relying solely on Ruff, or combining it with Mypy and Black).
• Any mini.nvim-specific considerations or plugins that simplify the integration of these tools.

Any insights or recommended configurations would be greatly appreciated!

Continue reading...