0x100101 / lab.nvim

Prototyping Tools for Neovim

👩‍🔬 lab.nvim

⚠️ Heads up: lab.nvim is very early in development and is currently in a proof of concept state. Right now lab.nvim is more-or-less stable and usable, but in any case, consider this pre-release beta software.

Goal

Provide a collection of unique prototyping tools for neovim. These tools should be extremely configurable and modular in nature.

Features

• Code Runner
• Quick Data

✳️ Code Runner

• The foundational feature for lab.nvim is a code runner with real-time, inline feedback. (Inspired by runjs, quokka and others.)
• The code runner currently supports JavaScript, Typescript, Python, and Lua with additional language support planned.
• The goal of the code runner isn't to be a full-fledged debugger, rather it aims to provide a simple rapid feedback mechanism that can be useful while working on prototyping tasks.

js_code_runner.mp4

Commands

Command	Action
Lab code run	Run or resume the code runner on the current file.
Lab code stop	Stop the code runner on the current file.
Lab code panel	Show the code runner info buffer.
Lab code config	Show the code runner config for the current file.

Note: that the run command is also automatically invoked each time you save changes to a file that is currently active.

Languages

Language	Supported
JS / TS	console, debugger, error
Python	print, breakpoint, error
Lua	print, error

✳️ Quick Data

• The quick data feature allows you to quickly insert fake data while prototyping.
• The implementation works by providing a dynamic snippet source to nvim-cmp.
• This feature currently supports fakerjs as a pre-configured data source.

quick_data.mp4

Configuration

• Add the source to your nvim-cmp setup:

sources = cmp.config.sources({
  { name = 'lab.quick_data', keyword_length = 4 }
}, 

• This feature is enabled by default but can be disabled:

require('lab').setup {
  quick_data = {
    enabled = false,
  }
}

Requirements

• neovim >= 0.7.2
• plenary.nvim
• node >= 16.10.0

Optional Requirements

• Python 3 (Python code runner)
• Lua 5.4 (Lua code runner)
• nvim-cmp (Quick data snippets)

Example Setup

Important: Notice the post install hook. Lab.nvim has a few internal node dependencies that should be installed. See: package.json

Packer

return require('packer').startup(function(use)
  use { '0x100101/lab.nvim', run = 'cd js && npm ci', requires = { 'nvim-lua/plenary.nvim' } }
end)

Vim Plug

Plug 'nvim-lua/plenary.nvim'
Plug '0x100101/lab.nvim', { 'do': 'cd js && npm ci' }

Setup options (defualts illustrated below)

require('lab').setup {
  code_runner = {
    enabled = true,
  },
  quick_data = {
    enabled = true,
  }
}

Key Mappings

nnoremap <F4> :Lab code stop<CR>
nnoremap <F5> :Lab code run<CR>
nnoremap <F6> :Lab code panel<CR>

📝 Roadmap

• Refactor the JavaScript code runner:
  • The code runner is more-or-less ad hoc / proof of concept, and should be re-factored in to 2 or 3 independent systems. Once this is complete the code will be more maintainable and tests can be added.
• Split out the server from the plugin and break out the server features in to modular, attachable components:
  • Version 1.0 should not be bundled with the node back-end, instead the user should install the back-end and additional components independently, as needed. (similar to language servers).
• Expand the configuration options to provide for more fine grained control over how everything works.
• Add a check health system.
• Investigate what options are available for adding tests to the front-end (neovim plugin)
• Improve documentation:
  • Add additional documentation regarding esbuild and the interesting features it exposes.