Skip to content

Pipes are a pattern of function composition. They enable you to do several transformations to your data in 1 step.

License

Notifications You must be signed in to change notification settings

graficos/pipe-js

Repository files navigation

Pipe JS

Unit Tests

Installation

npm i @graficos/pipe-js

Intro

Pipes are a pattern of function composition. They enable you to do several transformations to a value in 1 step.

You pass a number of functions and a parameter that will be consumed by these functions, so you can make some transformations to this value.

How are they useful?

For example, let's imagine this scenario, where you're applying the result of a function to another function:

var result = f(g(x))

I bet you have seen it multiple times. And something even more scary:

var result = f(g(h(i(j(x))))) // 🤔

How are you supposed to read or even debug that? This is where pipes are useful:

// The first example would be the same as:
var result = pipe(g, f)(x)

// and the second example:
var result = pipe(j, i, h, g, f)(x)

You'll notice that with pipe, we read the transformations left-to-right (which was the same as "from the inside to the outside" in the initial examples), which improves the readability.

Don't be scared about the (...)(x) syntax, that's how you call "a function that returns a function", which is what pipe is. It's also called a "Higher Order Function", because you pass functions as parameters.

How to debug pipes

If you need to see what's happening on each step, you can create what we usually call a tap function ("tap", "pipe", ...see the correlation? ;)). Imagine we have the following example:

import { pipe } from '@graficos/pipe-js'

// `formatText` is the combination of other "transformations" applied sequentially.
const formatText = pipe(removeHTMLTags, formatAuthorInfo, addFooterText)

const myFormattedText = formatText(sourceText)

We could debug each step with the tap function:

import { pipe } from '@graficos/pipe-js'

const tap = (currentValue) => {
  console.log(currentValue)
  return currentValue // remember, all your functions must return the value to be passed to the next function
}

// `formatText` is the combination of other "transformations" applied sequentially.
const formatText = pipe(
  removeHTMLTags,
  tap,
  formatAuthorInfo,
  tap,
  addFooterText
)

const myFormattedText = formatText(sourceText)
console.log({ myFormattedText })

What if my functions are asynchronous?

That's why pipe-js comes also with an asyncPipe method (tree-shakeable if you don't use it).

Example:

import { asyncPipe } from '@graficos/pipe-js'

const someAsyncFunction = (value) =>
  new Promise((resolve) => resolve(value + 2))
const someOtherAsyncOperation = (value) =>
  new Promise((resolve) => resolve(value * 3))

const result = await asyncPipe(someAsyncFunction, someOtherAsyncOperation)(1) // -> result === 9

Just remember that just like pipe always returned a function, asyncPipe will always return a Promise. You, you need to uwrap the value that it'll return with async/await or using .then().

Examples

You can see some tests to see more examples in /src (the *.spec.js files).

Another example would be "composable derived data". Sometimes they are called selectors or "computed properties":

const addHello = (value) => `Hello, ${value}`
const toUpperCase = (value) => String.prototype.toUpperCase.call(value)

const person = {
  name: 'Paul',
  getGreet() {
    return pipe(addHello, toUpperCase)(this.name)
  },
}

console.log(person.getGreet()) // -> HELLO, PAUL
person.name = 'Hello'
console.log(person.getGreet()) // -> HELLO, HELLO

In this example, any time name changes, the derived state will be updated (as it is a function invoked everytime with the current this.name).

About the library

  • It's done in JS and using JSDoc, so you'll have proper autocompletion in your editor.

  • You can use it both in the browser and in node, as it is exported as UMD, ESM and CJS.

Read more about "pipes":

About

Pipes are a pattern of function composition. They enable you to do several transformations to your data in 1 step.

Topics

Resources

License

Stars

Watchers

Forks

Sponsor this project

 

Packages

No packages published