Socket
Socket
Sign inDemoInstall

cmd-ts

Package Overview
Dependencies
11
Maintainers
1
Versions
30
Alerts
File Explorer

Advanced tools

Install Socket

Detect and block malicious and high-risk dependencies

Install

    cmd-ts

> 💻 A type-driven command line argument parser, with awesome error reporting 🤤


Version published
Weekly downloads
71K
decreased by-11.88%
Maintainers
1
Install size
500 kB
Created
Weekly downloads
 

Readme

Source

cmd-ts

💻 A type-driven command line argument parser, with awesome error reporting 🤤

Not all command line arguments are strings, but for some reason, our CLI parsers force us to use strings everywhere. 🤔 cmd-ts is a fully-fledged command line argument parser, influenced by Rust's clap and structopt:

🤩 Awesome autocomplete, awesome safeness

🎭 Decode your own custom types from strings with logic and context-aware error handling

🌲 Nested subcommands, composable API

Basic usage

import { command, run, string, number, positional, option } from 'cmd-ts';

const cmd = command({
  name: 'my-command',
  description: 'print something to the screen',
  version: '1.0.0',
  args: {
    number: positional({ type: number, displayName: 'num' }),
    message: option({
      long: 'greeting',
      type: string,
    }),
  },
  handler: (args) => {
    args.message; // string
    args.number; // number
    console.log(args);
  },
});

run(cmd, process.argv.slice(2));
command(arguments)

Creates a CLI command.

Decoding custom types from strings

Not all command line arguments are strings. You sometimes want integers, UUIDs, file paths, directories, globs...

Note: this section describes the ReadStream type, implemented in ./src/example/test-types.ts

Let's say we're about to write a cat clone. We want to accept a file to read into stdout. A simple example would be something like:

// my-app.ts

import { command, run, positional, string } from 'cmd-ts';

const app = command({
  /// name: ...,
  args: {
    file: positional({ type: string, displayName: 'file' }),
  },
  handler: ({ file }) => {
    // read the file to the screen
    fs.createReadStream(file).pipe(stdout);
  },
});

// parse arguments
run(app, process.argv.slice(2));

That works okay. But we can do better. In which ways?

  • Error handling is out of the command line argument parser context, and in userland, making things less consistent and pretty.
  • It shows we lack composability and encapsulation — and we miss a way to distribute shared "command line" behavior.

What if we had a way to get a Stream out of the parser, instead of a plain string? This is where cmd-ts gets its power from, custom type decoding:

// ReadStream.ts

import { Type } from 'cmd-ts';
import fs from 'fs';

// Type<string, Stream> reads as "A type from `string` to `Stream`"
const ReadStream: Type<string, Stream> = {
  async from(str) {
    if (!fs.existsSync(str)) {
      // Here is our error handling!
      throw new Error('File not found');
    }

    return fs.createReadStream(str);
  },
};

Now we can use (and share) this type and always get a Stream, instead of carrying the implementation detail around:

// my-app.ts

import { command, run, positional } from 'cmd-ts';

const app = command({
  // name: ...,
  args: {
    stream: positional({ type: ReadStream, displayName: 'file' }),
  },
  handler: ({ stream }) => stream.pipe(process.stdout),
});

// parse arguments
run(app, process.argv.slice(2));

Encapsulating runtime behaviour and safe type conversions can help us with awesome user experience:

  • We can throw an error when the file is not found
  • We can try to parse the string as a URI and check if the protocol is HTTP, if so - make an HTTP request and return the body stream
  • We can see if the string is -, and when it happens, return process.stdin like many Unix applications

And the best thing about it — everything is encapsulated to an easily tested type definition, which can be easily shared and reused. Take a look at io-ts-types, for instance, which has types like DateFromISOString, NumberFromString and more, which is something we can totally do.

Inspiration

This project was previously called clio-ts, because it was based on io-ts. This is no longer the case, because I want to reduce the dependency count and mental overhead. I might have a function to migrate types between the two.

FAQs

Last updated on 17 Jul 2023

Did you know?

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install

Related posts

SocketSocket SOC 2 Logo

Product

  • Package Alerts
  • Integrations
  • Docs
  • Pricing
  • FAQ
  • Roadmap

Stay in touch

Get open source security insights delivered straight into your inbox.


  • Terms
  • Privacy
  • Security

Made with ⚡️ by Socket Inc