fp-ts-reduce

A experimental code snippets to achieve the purpose of transducers using fp-ts.

Motivation

I want a transduce like function that can be used for Objects.

The well known functional programing library with JavaScrit, Ramda has already implemented transduce. But, ramda's tranduce can be used only for arrays.

With fp-ts, we can use map, filter, reduce not only for arrays, but also for objects. But, fp-ts doesn't have transduce.

Basic example

This example is made to look like the example from Ramda in Transducers

// TEO ( Tuple Either Option ) with fp-ts
import { TEO as _ } from '../src' // not published as npm module
import { readonlyArray as A, readonlyRecord as RR } from 'fp-ts'
import { flow } from 'fp-ts/function'

// ramda
import * as R from 'ramda'

// util
const isEven = (n: number) => n % 2 === 0
const double = (n: number) => n * 2

// data
const arr = [1, 2, 3, 4, 5, 6]
const obj = { a: 1, b: 2, c: 3, d: 4, e: 5, f: 6 }
let result: readonly number[]

// ramda
const doubleEvens_Ramda = R.compose(R.filter(isEven), R.map(double))
result = R.into([], doubleEvens_Ramda, arr)
console.log(result) // [4, 8, 12]

// TEO.ts ( array -> array )
const intoNumberArrays = _.reduce(A.Foldable, A.getMonoid<number>())
const doubleEvens = intoNumberArrays(flow(_.filter(isEven), _.map(flow(double, A.of))))
result = doubleEvens(arr)
console.log(result) // [4, 8, 12]

// TEO.ts ( object -> array )
const obj2NumberArrays = _.reduce(RR.Foldable, A.getMonoid<number>())
const doubleEvens_obj = obj2NumberArrays(flow(_.filter(isEven), _.map(flow(double, A.of))))
result = doubleEvens_obj(obj)
console.log(result) // [4, 8, 12]