F*ckingNode

Abstract

We as humans have dedicated years of hard work to a goal: fixing Node.

We created NPM, PNPM, YARN, added ES6, async/await, fs.promises, even TS with --experimental-strip-types, we changed the order and made Deno, then we got hungry and made Bun, then we made Deno 2... Everything has been getting us closer to a better Node, and now we're one step closer to fixing NodeJS.

F*ckingNode, a CLI utility to automate cleanup of a Node user's worst nightmare: node_modules/.

(yes i'm calling it like that and i'm shipping that to production, don't question me)

By the way, while the name implies NodeJS-only support, Deno and Bun are coming soon!

Check the v2 branch, we're working on support for non-Node projects, and also other stuff like custom settings for the CLI and more!

Features

Cleaner

Automates cleaning of each NodeJS project you add by using your manager's built in features, so a single command makes everything cleaner. It will detect your lockfile (e.g. "pnpm-lock.yaml", "package-lock.json") to know whether to use npm, pnpm, or yarn (currently supported managers).

• fuckingnode clean - does the obvious.
• fuckingnode clean --update - does the obvious + updates your deps.
• fuckingnode clean --verbose - does the obvious with some extra logs.

--flags can be mixed to use more features at once.

clean can take an intensity level, either "normal", "hard", or "maxim"

clean normal, clean hard, clean maxim, or just clean (defaults to normal).

The higher the level, the more space you'll recover, but the slower the process will be.

This is what each level does:

• Normal: Runs default prune / autoclean and dedupe commands.
• Hard: Does the previous + cleans the entire cache (pnpm store prune / yarn cache clean / npm cache clean --force). Note: While pnpm will only purge unused packages, npm will clear the entire cache, making it go slower.
• Maxim: This does not run any cleanup command. It will simply remove the node_modules/ folder of each project. Slowest (plus you'll need to reinstall deps), hence it will ask for confirmation before starting. The good thing is that you'll probably get many GB of storage back, so this is actually useful if your drive is almost full.

Manager

F*ckingNode keeps a list of all paths to the projects it should clean - it's you who has to maintain it. Keep in mind paths should point to the root, where you have /package.json, /lockfile¹, and /node_modules.

• fuckingnode manager list - lists all projects.
• fuckingnode manager list --ignored - lists all ignored projects.
• fuckingnode manager add <path> - adds a project to list.
• fuckingnode manager remove <path> - removes a project from list.
• fuckingnode manager ignore <path> - ignores a project (won't be cleaned or updated).
• fuckingnode manager revive <path> - stops ignoring a project.

<path> refers to a path, either an absolute one (C:\Users\me\project), relative one (../project), or the --self flag which will use the Current Working Directory.

Best practice is to run fuckingnode manager add --self after creating a Node project from your CLI.

Others

• fuckingnode migrate <path> <"pnpm" | "npm" | "yarn"> - migrates a project to the specified package manager (basically removes lockfile, node_modules, and reinstalls with the selected package manager). For now it relies on the specified package manager's ability to understand other lockfiles to ensure version compatibility. No issues should occur.
• fuckingnode settings schedule <h> <d> - schedules automated cleaning. h indicates what hour of the day should the cleaning be done (number from 0 to 23). d indicates the day interval, e.g. 3 = every 3 days (number, or an * for daily cleaning).
• fuckingnode --help, fuckingnode --version, and fuckingnode self-update - all do the obvious.

And that's it for now.

Tip

If for any reason you want to skip a project without removing it from your list (idk, maybe you temporarily don't want anyone to update deps for that specific project), create an empty .fknodeignore file in the root of the project and F*ckingNode will ignore it, or just run fuckingnode ignore --self from the root of the project.

---

Installation

Microsoft Windows

1. Download the installer from the GitHub releases page. You'll see "INSTALLER" on the filename, there's just one.
2. Run it.
3. You're done! It should have automatically installed. The fuckingnode CLI command should now work out of the box.

Other systems (mac & Linux)

1. Download the program from the GitHub releases page. macOS and Linux have support for both x84_64 and ARM.
2. Place your downloaded file anywhere, like /scripts or /home/USER/my-scripts.
3. Add the path to the binary to your system's path environment variable.
4. You're done! The fuckingnode command will now work from your terminal.

Here's how to add the path to your path env variable, in case you didn't know:

# open your Bash config file with nano (or your preferred editor)
nano ~/.bashrc         # Linux
nano ~/.bash_profile   # macOS

# paste this
export PATH="$PATH:/home/USER/my-scripts/fuckingnode" # keep '$PATH' and replace the rest (/home...) with the actual path to wherever you saved fuckingnode. It's recommended that you keep the name like that, "fuckingnode" with lowercase.

# save with CTRL + O, ENTER, and CTRL + X
# then, reload your config
source ~/.bashrc          # Linux
source ~/.bash_profile    # macOS

Compiling from source

1. Install Deno 2.
2. Open this project from the root.
3. Run deno task compile.
4. An executable for each platform will be created at dist/. Run the executable for your platform (file names are specific to each platform).

---

Hope those motherf*ckers don't annoy you again! And hey, if you find any issue with the program, just open up an issue (or make a PR which would be awesome 😄).

Cya!

Footnotes

1. npm, pnpm, and yarn are supported, each one having it's own lockfile. ↩