kyanite

A light weight pure functional library with single type utility functions and it only depends on itself.

Contents

• View the Changelog
• View the Documentation
• Philosophy
• Key Features
• How To
• Credit

Philosophy

The goal for Kyanite is to be stripped down, light weight, and intuitive. The idea is to be performant and easy to use in a functional setting, making it simple to build reusable functions in your code base with powerful and organized algorithmic pipes. Ultimately, Kyanite’s continued growth and improvement will support the growth and improvement of the JavaScript community

Key Features

• Purely Functional - This was a main focus for the project. I wanted it to be an easy to use, functional system while also being completely pure by making use of piping and transducers to boost performance.
• Single type utility functions - Theoretically, all of the functionality is based around accepting a single data type, doing what it does, and giving you back a result, thus making it reliable, stable, and lightweight.
• Everything is curried! Setup static in one spot and then pass the rest of the dynamic data in later.
• Data last ideology

How To

With a CDN

<!-- It is recommended to replace the @latest with a strict version number for production -->
<script src="https://cdn.jsdelivr.net/npm/kyanite@latest/dist/kyanite.min.js"></script>
<!-- To use the debug friendly kyanite simply remove .min from the filename -->
<script src="https://cdn.jsdelivr.net/npm/kyanite@latest/dist/kyanite.js"></script>
<script>
  const K = kyanite

  K.isEmpty({})
</script>

Or installing with npm i kyanite and then:

Standard module system

// This will use the module path in the package.json (src/index.js)
import * as K from 'kyanite'
// However if you want to grab the dev version
import * as K from 'kyanite/dist/kyanite.js'

Common JS

// For the prod minified version
const K = require('kyanite')
// For the dev/debug friendly version
const K = require('kyanite/dist/kyanite.js')

Local copy script tag

<!-- To use the prod minified version -->
<script src="/path/to/dist/kyanite.min.js"></script>
<!-- To use the dev/debug friendly version -->
<script src="/path/to/dist/kyanite.js"></script>
<script>
  const K = kyanite

  K.isEmpty({})
</script>

Credit

A lot of the if not most of the inpiration for this library came from 2 libraries I follow closely, Primarily most of it stems from:

• foreword by Abstract Tools which is a very nice and easy to use library developed by a close friend. This is where a lot of functionality, AND the idea of a pure single data type system came from I can't recommend it enough.
• Ramdajs by Ramda a large and fairly handy library where the original idea sparked

Changelog

v0.11.0 (0.11.1)

Breaking Changes

• factors output has changed to be inclusive, it also works with negative numbers now
  • Example: factors(-36) // => [1, 2, 3, 4, 6, 9, 12, 18, 36]
  • You can easily get negative factors with negate and map
    • e.g: map(negate, factors(-36)) // => [-1, -2, -3, -4, -6, -9, -12, -18, -36]
• empty (and isEmpty until its removed) will now throw a type error for unsupported types
  • empty supports the same types as count as well as null and undefined
  • e.g: empty(1) // => TypeError: Unsupported type: Number
• juxt functionality changes. It's now a 2 param curried function that takes arrays for data
  • e.g: juxt([Math.min, Math.max], [3, 4, 9, -3]) // => [-3, 9]
  • Also moved it to be in the array category
• ap re written to be a proper S combinator function
  • e.g: ap(x => y => x + y, z => z * 2, 2) // => 6
• Removed amend function since it's really just a limited Object.assign no real use for it

New

• reduce now supports iterable types like Map and Set
• Added inc Function which increases a number by 1
• Added dec Function which decreases a number by 1
• Added isZero Function which checks if the number passed in is equal to the number zero
• Added size Function which takes a map or set data type and returns its size
  • Note: It is currently categorized under Function which may change in the future as I add more Map|Set friendly functions
• Added count Function which takes any collection based data and counts the values within
• Added reduced Function which should be used with reduce or reduceRight as a short circuit for the function (see improved)
• has is now generic for Array, String, Object, Map, and Set data types
• Added within Function which acts like between but is exclusive of the numbers provided

Improved

• isEmpty now supports Maps and Sets
• Slight increase in partition performance
• Slight increase in omit performance
• Added import example onto each function, so it's visible within the examples
• factors had some typos in its docs
• Cleaned up always documentation
• Cleaned up identity documentation examples
• Cleaned up type to be more lightweight
• Completely re wrote reduce and reduceRight
  • No breaking changes
  • Both functions now support reduced which will short circuit out of the iteration and return the value
  • This is a good strategy for performance boosts
• reduce, reduceRight, and reduced are all using the @@transducer protocol
• Moved every, some, reject, filter, and find to use this new flow giving them decent performance improvements
• Tweaked height function for a tiny performance boost