Be hesitatnt to try this at home. This tool is very wobbly.

$ yarn add -D observations

// ./observations.js
import Observations from 'observations'

Observations.register({
  name: 'js-watcher',
  match: ['./**/*.js', './**/*.jsx'],
  ignore: './watch.config.js',

  hanleError(error) {
    // Apply whatever functionality you like when
    // a process encounters an error.
    console.log(`Got an error, dude!`)
  },

  handleOutput(output) {
    // Apply whatever functionality you like when
    // a managed process provides output.
    console.log(`Got output, dude!`)
  },

  handleChange(change) {
    // Apply whatever functionality you like when
    // an observed path changes, and then programatically
    // provide the next set of instructions.
    change.exec([
      'echo $VIRTUAL_HOST',
      'ls ./public/css',
      'webpack'
    ])
  }
})

$ node observations

API

match: string[] = './**/*'

An array of globs/patterns to watch.

name: string = ''

[optional] Not yet used for anything.

ignore: string[] = ''

[optional] An array of globs/patterns to ignore.

persistent: boolean = true

[optional] Whether or not the process should exit on its own or stay alive.

handleOutput: (output: string)

[optional] A function to execute to handle process output.

handleError: (error: string)

[optional] A function to execute when process throws an error.

handleChange: (change: { data: string, exec: (commands: string[]) })

[optional] A function to execute when a watched file changes. handleChange is provided with a change object that contains data explaining the change and an exec method that allows you to provide which commands to execute after your custom functionality has completed.