Awesome
<div style="text-align:center"><img src="ressources/visual_assets/Sniprun_transparent.png" /></div> <div align="center"><p> <a href="https://github.com/michaelb/sniprun/releases/latest"> <img alt="Latest release" src="https://img.shields.io/github/v/release/michaelb/sniprun" /> </a> <a href="https://github.com/michaelb/sniprun/actions"> <img alt="CI build" src="https://github.com/michaelb/sniprun/workflows/Rust/badge.svg" /> </a> <a href="https://github.com/michaelb/sniprun/releases"> <img alt="Total downloads" src="https://img.shields.io/github/downloads/michaelb/sniprun/total" /> </a> <a href="https://github.com/michaelb/sniprun/pulse"> <img alt="Last commit" src="https://img.shields.io/github/last-commit/michaelb/sniprun"/> </a> </p> </div>Introduction
Sniprun is a code runner plugin for neovim written in Lua and Rust. It aims to provide stupidly fast partial code testing for interpreted and compiled languages . Sniprun blurs the line between standard save/run workflow, jupyter-like notebook, and REPL/interpreters.
</br>TLDR: Plug 'michaelb/sniprun', {'do': 'sh install.sh'}, :SnipRun, :'<,'>SnipRun, :SnipInfo
Installation, configuration, ...
See installation instructions , configuration tips , usage explanations and much more useful information on the WIKI.
Demos
Send to Sniprun snippets of any language
A very simple example (in C), play the .gif and look in the command area:
The result can be returned in multiple (even at the same time) ways:
| Classic | Virtual Text |
|---|---|
 |  |
| Temporary Floating Window | Terminal |
 |  |
| Notification | API |
 |  |
REPL-like behavior is available for some languages
Python, Julia, Lua, JavaScript & Typescript (via deno), Clojure, R, Mathematica, Sage, coming soon for many other interpreted (and compiled) languages. With REPL-like behavior ,you can run code dependent on previously executed code, just like in a REPL, from within your favorite editor.
Features
Sniprun is a way to quickly run small snippets of code, on the fly, and iterate very quickly and conveniently. To learn a language, to quickly experiment with new features (not yet embedded in classes or a whole project etc...), or to develop simple code pipelines (like a machine learning exercise) that fit in a unique file, sniprun is probably the best plugin out there.
As a matter of proof, Sniprun :
- Officially supports all these languages (compiled & interpreted), and virtually any language
- can create and connect to REPLs in order to present an interactive and playful interface
- can run things like GUI plots, networks requests or even Ansible playbooks
- doesn't require even one line of configuration by default (but can be customized up to the tiniest things)
- can run code from a part of a file which isn't complete / contains errors
- can automatically fetch (in some languages) the
importsnecessary for your code snippet - can run live (at every keystroke)
- lends itself to easy mappings and Vim motions
- has an API (for running code, and displaying results)
- has many result display modes that can be enabled at the same time, and for different output status if wanted
- supports literate programming in Markdown, Orgmode and Neorg
Known limitations
Due to its nature, Sniprun may have trouble with programs that :
- Mess with standard output / stderr
- Need to read from stdin
- Access files; sniprun does not run in a virtual environment, it accesses files just like your own code do, but since it does not run the whole program , something might go wrong. Relative paths may cause issues, as the current working directory for sniprun will be somewhere in ~/.cache/sniprun, and relative imports may miss.
- No support for Windows
- NixOS, MacOS users have to compile sniprun locally. Sniprun has not been tested on other Unixes (besides Linux itself, of course)
Changelog
It's been quite a journey already! For history fans, see the full changelog.
Contributing
Sniprun has been made contributor-friendly (see CONTRIBUTING.md), so it's relatively easy to create / fix interpreters for any language. But any (constructive) issue, discussion, or doc Pull Request is a welcome form of contribution !