rubico

🏞 a shallow river in northeastern Italy, just south of Ravenna

Motivation

You are suddenly dropped into a world where all people write code in assembly.

Language Hierarchy

There is no "High Level Language", only "Assembly Language". There is no C, just ASM. There are no variables, only registers. You are in charge of managing all memory in your programs: moving data from register to register, pushing and poping data on the hardware supported stack.

How would you write a webserver, or a database? How long would that take? How much longer would it take you to do whatever it is that you are currently doing?

We need not stay here any longer.

...interdimensional warp...

Welcome back to reality, where the world is rife with programming languages above assembly

How did this come to be? Why would anyone not want to spend their day to day in assembly? According to an answer thread on stack overflow,

ASM has poor legibility and isn't really maintainable compared to higher-level languages.

[Assembly] takes more code to do the same thing as in a high-level languge, and there is a direct correlation between lines of code and bugs.

Another take from wikipedia

In contrast to low-level programming languages, [high-level programming languages] may use natural language elements, be easier to use, or may automate (or even hide entirely) significant areas of computing systems (e.g. memory management), making the process of developing a program simpler and more understandable than when using a lower-level language.

Perhaps the abundance of higher level languages comes down to readability versus performance

First code for correctness, then for clarity (the two are often connected, of course!). Finally, and only if you have real empirical evidence that you actually need to, you can look at optimizing.

IMO the obvious readable version first, until performance is measured and a faster version is required.

I would go for readability first.

Great, looks like people are for readability, so where does that leave us? And what does any of this have to do with rubico?

Consider these two samples of JavaScript code. Both execute an asynchronous function for every element of an array

Promise.all(array.map(doAsyncThing)) // vanilla JavaScript

map(doAsyncThing)(array) // rubico

It looks like you can write a little less to do the same thing with rubico. Is the rubico version more readable? I'd say it's up for debate.

What if we wanted to do multiple asynchronous things in parallel for every item of the array?

Promise.all([
  Promise.all(array.map(doAsyncThingA)),
  Promise.all(array.map(doAsyncThingB)),
  Promise.all(array.map(doAsyncThingC)),
]) // vanilla JavaScript

map(fork([
  doAsyncThingA,
  doAsyncThingB,
  doAsyncThingC,
]))(array) // rubico

It looks like vanilla JavaScript has four more Promise.all statements, and two more map keywords. rubico, on the other hand, has one map and one fork. Simpler? Hold your horses.

What if we now want to do another async thing per item of each of the responses?

Promise.all([
  Promise.all(array.map(doAsyncThingA).then(
    arrayA => Promise.all(arrayA.map(doAsyncThingAA))
  )),
  Promise.all(array.map(doAsyncThingB).then(
    arrayB => Promise.all(arrayB.map(doAsyncThingBB))
  )),
  Promise.all(array.map(doAsyncThingC).then(
    arrayC => Promise.all(arrayC.map(doAsyncThingCC))
  )),
]) // vanilla JavaScript

map(fork([
  pipe([doAsyncThingA, map(doAsyncThingAA)]),
  pipe([doAsyncThingB, map(doAsyncThingBB)]),
  pipe([doAsyncThingC, map(doAsyncThingCC)]),
]))(array) // rubico

I think it's safe to say that rubico is more expressive here.

Back to assembly. You could compare what rubico does for JavaScript to what C does for assembly.

In contrast to assembly, C uses natural language elements, is easier to use, and automates (but doesn't hide entirely) memory management. C makes the process of developing a program simpler and more understandable than when using assembly.

In contrast to vanilla JavaScript, rubico automates (hides entirely) the cruft surrounding Promises. rubico makes the process of developing a program simpler and more understandable than when using vanilla JavaScript.

I should also mention that when you use rubico, you get the benefits of the functional programming paradigm (but not the confusion) for free.

At your leisure, motivations 1, 2, and 3 for functional programming.

Introduction

Installation

Documentation

Transducers

Examples

rubico resolves the Promise.all of three Promises

  1. simplify asynchronous programming
  2. enable functional programming
  3. free transducers

programs written with rubico follow a functional style, otherwise known as data last.

This is data first

[1, 2, 3, 4, 5].map(number => number * 2) // => [2, 4, 6, 8, 10]

This is data last

map(number => number * 2)([1, 2, 3, 4, 5]) // => [2, 4, 6, 8, 10]

Data last saves you brain power when you name things

const xyz = async x => {
  const resultOfFoo = await foo(x)
  const resultOfBar = await bar(resultOfFoo)
  const resultOfBaz = await baz(resultOfBar)
  return resultOfBaz
} // data first

const xyz = pipe([foo, bar, baz]) // data last

Installation

with npm

npm i rubico

with deno

import {
  pipe, fork, assign, tap, tryCatch, switchCase,
  map, filter, reduce, transform,
  any, all, and, or, not,
  eq, gt, lt, gte, lte,
  get, pick, omit,
} from 'https://deno.land/x/rubico/rubico.js'

browser script, global rubico

<script src="https://unpkg.com/rubico/index.js" crossorigin></script>

browser module

<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <title>Your Title Here</title>
    <script type="module">
import {
  pipe, fork, assign, tap, tryCatch, switchCase,
  map, filter, reduce, transform,
  any, all, and, or, not,
  eq, gt, lt, gte, lte,
  get, pick, omit,
} from 'https://deno.land/x/rubico/rubico.js'

// your code here
    </script>
</head>
<body></body>
</html>

System Requirements

  • minimum node version: 10.3

Documentation

rubico aims to hit the sweet spot between expressivity and interface surface area. There are 23 functions at the moment; this number is not expected to go up much more or at all. Instead, some methods will have property functions that represent the same signature (i.e. map vs map.series) but exhibit differences in behavior (i.e. map executes in parallel while map.series executes in series).

[series] and [parallel] are tags to denote the asynchronous behavior of methods that accept multiple functions.

  • [series]: execute functions one at a time. If order is not implied, it is left to the implementation. (i.e. iterating an Object)
  • [parallel]: execute functions in parallel.

All higher order functions accept sync or async functions; if all provided functions are synchronous, the entire execution is synchronous.

function composition

  • pipe [series] - chain functions together
  • tap - spy on data
  • tryCatch [series] - try a function, catch with another
  • switchCase [series] - control flow

function + data composition

  • fork [parallel] - multiply data by functions
    • fork.series [series]
  • assign [parallel] - set properties on data by functions

data transformation

  • map [parallel] - apply a function to data
    • map.pool [parallel] - map with asynchronous limit
    • map.withIndex [parallel] - map with index
    • map.series [series]
    • map.sleep [series] - map.series with delay in between todo
  • filter [parallel] - exclude data by predicate
    • filter.withIndex [parallel] - filter with index
  • reduce [series] - execute data transformation (powerful)
  • transform [series] - execute data transformation (convenient)

predicate composition

  • any [parallel] - is function of any data truthy?
  • all [parallel] - is function of all data truthy?
  • and [parallel] - any functions of data truthy?
  • or [parallel] - all functions of data truthy?
  • not - not(equals)(x) is !equals(x)

comparison

  • eq [parallel] - left equals right?
  • gt [parallel] - left > right?
  • lt [parallel] - left < right?
  • gte [parallel] - left >= right?
  • lte [parallel] - left <= right?

property + index access

  • get - access a value by path or index

data composition

  • pick - only allow provided properties
    • pick.range - only allow provided range todo
  • omit - exclude provided properties

pipe

chains functions from left to right in series

y = pipe(functions)(x)

functions is an array of functions

x is anything

if x is a function, pipe chains functions from right to left, see transducers

y is the output of running x through the chain of functions

y is wrapped in a Promise if any of the following are true:

  • any function of functions is asynchronous
pipe([
  x => x + ' ',
  x => x + 'world',
])('hello') // => 'hello world'

pipe([
  async x => x + ' ',
  x => x + 'world',
])('hello') // => Promise { 'hello world' }

back to documentation

fork

parallelizes functions with data, retaining functions' type and shape

y = fork(functions)(x)

functions is an array of functions or an object of functions

all functions of functions are run concurrently

x is anything

if functions is an array, y is functions.map(f => f(x))

if functions is an object, y is an object of entries key: f(x) for entry key: f of functions

y is wrapped in a Promise if any of the following are true:

  • any function of functions is asynchronous
fork([
  x => x + 'world',
  x => x + 'mom'
])('hello') // => ['hello world', 'hello mom']

fork([
  x => x + 'world',
  async x => x + 'mom'
])('hello') // => Promise { ['hello world', 'hello mom'] }

fork({
  a: x => x + 'world',
  b: x => x + 'mom',
})('hello') // => { a: 'hello world', b: 'hello mom' }

fork({
  a: x => x + 'world',
  b: async x => x + 'mom',
})('hello') // => Promise { { a: 'hello world', b: 'hello mom' } }

fork.series executes functions with data in series, retaining functions' type and shape

y = fork.series(functions)(x)

back to documentation

assign

parallelizes functions with data, merging output into data

y = assign(functions)(x)

functions is an object of functions

all functions of functions are run concurrently

x is an object

output is an object of entries key: f(x) for entry key: f of functions

y is output merged into x

y is wrapped in a Promise if any of the following are true:

  • any function of functions is asynchronous
assign({
  hi: x => 'hi ' + x,
  bye: x => 'bye ' + x,
})({ name: 'Ed' }) // => { name: 'Ed', hi: 'hi Ed', bye: 'bye Ed' }

assign({
  async hi: x => 'hi ' + x,
  bye: x => 'bye ' + x,
})({ name: 'Ed' }) // => Promise { { name: 'Ed', hi: 'hi Ed', bye: 'bye Ed' } }

assign({
  name: () => 'not Ed',
})({ name: 'Ed' }) // => { name: 'not Ed' }

back to documentation

tap

calls a function with data, returning data

y = tap(f)(x)

x is anything

f is a function that expects one argument x

y is x

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous

if x is a function, y is a transduced reducing function, see transducers

y = tap(f)(x); reduced = reduce(y)(z)

reduce is reduce,

z is an iterable, async iterable, or object

zi is an element of z

f is a function that expects one argument zi

reduced is equivalent to reduce(x)(z)

tap(
  console.log, // > 'hey'
)('hey') // => 'hey'

const asyncConsoleLog = async x => console.log(x)

tap(
  asyncConsoleLog, // > 'hey'
)('hey') // => Promise { 'hey' }

const concat = (y, xi) => y.concat([xi])

reduce(
  tap(console.log)(concat), // > 1 2 3 4 5
  [],
)([1, 2, 3, 4, 5]) // => [1, 2, 3, 4, 5]

back to documentation

tryCatch

tries a function with data, catches with another function

y = tryCatch(f, g)(x)

f is a function that expects one argument x

g is a function that expects two arguments err and x

x is anything

err is a value potentially thrown by f(x)

if f(x) throws err, y is g(err, x), else y is f(x)

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous
  • f is synchronous, g is asynchronous, and f(x) threw
const onError = (e, x) => `${x} is invalid: ${e.message}`

tryCatch(
  x => x,
  onError,
)('hello') // => 'hello'

const throwGoodbye = () => { throw new Error('goodbye') }

tryCatch(
  throwGoodbye,
  onError,
)('hello') // => 'hello is invalid: goodbye'

const rejectWithGoodbye = () => Promise.reject(new Error('goodbye'))

tryCatch(
  rejectWithGoodbye,
  onError,
)('hello') // => Promise { 'hello is invalid: goodbye' }

back to documentation

switchCase

an if, else if, else construct for functions

y = switchCase(functions)(x)

x is anything

functions is an array of functions

given

  • predicate if functions if1, if2, ..., ifN
  • corresponding do functions do1, do2, ..., doN
  • an else function elseDo

functions is an array of functions

[
  if1, do1,
  if2, do2,
  ..., ...,
  elseDo,
]

switchCase evaluates functions in functions from left to right

y is the first do(x) whose corresponding if(x) is truthy

y is wrapped in a Promise if any of the following are true:

  • any evaluated functions are asynchronous
const isOdd = x => x % 2 === 1

switchCase([
  isOdd, () => 'odd',
  () => 'even',
])(1) // => 'odd'

switchCase([
  async isOdd, () => 'odd',
  () => 'even',
])(1) // => Promise { 'odd' }

back to documentation

map

applies a function to each element of data in parallel, retaining data type and shape

y = map(f)(x)

x is an iterable, an async iterable, an object, or a function

xi is an element of x

f is a function that expects one argument xi

y is of type and shape x with f applied to each element, with some exceptions:

  • if x is an async iterable but not a built-in type, y is a generated async iterable
  • if x is an iterable but not a built-in type, y is a generated iterable
  • if x is an iterable but not a built-in type and f is asynchronous, y is an iterable of promises

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous and x is not an async iterable

if x is a function, y is a transduced reducing function, see transducers

y = map(f)(x); reduced = reduce(y)(z)

reduce is reduce,

z is an iterable, async iterable, or object

zi is an element of z

f is a function that expects one argument zi

x is a reducing function that expects two arguments y and f(zi)

reduced is equivalent to reduce(x)(map(f)(z))

const square = x => x ** 2

map(
  square,
)([1, 2, 3, 4, 5]) // => [1, 4, 9, 16, 25]

const asyncSquare = async x => x ** 2

map(
  asyncSquare,
)([1, 2, 3, 4, 5]) // => Promise { [1, 4, 9, 16, 25] }

map(
  Math.abs,
)(new Set([-2, -1, 0, 1, 2])) // => { Set { 0, 1, 2 } }

const double = ([k, v]) => [k + k, v + v]

map(
  double,
)(new Map([['a', 1], ['b', 2]])) // => Map { 'aa' => 2, 'bb' => 4 }

map(
  byte => byte + 1,
)(new Uint8Array([97, 98, 99])) // Uint8Array [ 98, 99, 100 ]

map(
  word => word + 'z',
)({ a: 'lol', b: 'cat' }) // => { a: 'lolz', b: 'catz' }

map.pool Apply a function to every element of data in parallel with limited concurrency

y = map.pool(size, f)(x)

map.withIndex Apply a function to every element of data in parallel with index and reference to data

y = map.withIndex(f)(x); yi = f(xi, i, x)

map.series Apply a function to every element of data in series

y = map.series(f)(x)

map.seriesWithIndex Apply a function to every element of data in series with index and reference to data

 = map.seriesWithIndex(f)(x); yi = f(xi, i, x)

back to documentation

filter

filters elements out of data in parallel based on provided predicate

y = filter(f)(x)

x is an iterable, an async iterable, an object, or a function

xi is an element of x

f is a function that expects one argument xi

y is of type and shape x with elements xi where f(xi) is truthy, with some exceptions:

  • if x is an async iterable but not a built-in type, y is a generated async iterable
  • if x is an iterable but not a built-in type, y is a generated iterable
  • if x is an iterable but not a bulit-in type and f is asynchronous, filter will throw a TypeError

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous and x is not an async iterable

if x is a function, y is a transduced reducing function, see transducers

y = filter(f)(x); reduced = reduce(y)(z)

reduce is reduce,

z is an iterable, async iterable, or object

zi is an element of z

f is a function that expects one argument zi

x is a reducing function that expects two arguments y and zi

reduced is equivalent to reduce(x)(filter(f)(z))

const isOdd = x => x % 2 === 1
filter(
  isOdd,
)([1, 2, 3, 4, 5]) // => [1, 3, 5]

const asyncIsOdd = async x => x % 2 === 1
filter(
  asyncIsOdd,
)([1, 2, 3, 4, 5]) // => Promise { [1, 3, 5] }

filter(
  letter => letter !== 'y',
)('yoyoyo') // => 'ooo'

const abcSet = new Set(['a', 'b', 'c'])
filter(
  letter => abcSet.has(letter),
)(new Set(['a', 'z'])) // => Set { 'a' }

filter(
  ([key, value]) => key === value,
)(new Map([[0, 1], [1, 1], [2, 1]])) // => { Map { 1 => 1 } }

filter(
  bigint => bigint <= 3n,
)(new BigInt64Array([1n, 2n, 3n, 4n, 5n])) // => BigInt64Array [1n, 2n, 3n]

filter(
  value => value === 1,
)({ a: 1, b: 2, c: 3 }) // => { a: 1 }

filter.withIndex Filter, but with each predicate called with index and reference to data

y = filter.withIndex(f)(x); yi = f(xi, i, x)

back to documentation

reduce

transforms data in series according to provided reducing function and initial value

y = reduce(f, x0)(x)

x is an iterable, an async iterable, or an object

xi is an element of x

f is a reducing function that expects two arguments y and xi

x0 is optional, and if provided:

  • y starts as x0
  • iteration begins with the first element of x

if x0 is not provided:

  • y starts as the first element of x
  • iteration begins with the second element of x

y is f(y, xi) for each successive xi

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous
  • x is an async iterable
const add = (y, xi) => y + xi

reduce(
  add,
)([1, 2, 3, 4, 5]) // => 15

reduce(
  add, 100,
)([1, 2, 3, 4, 5]) // => 115

const asyncAdd = async (y, xi) => y + xi

reduce(
  asyncAdd,
)([1, 2, 3, 4, 5]) // => Promise { 15 }

const asyncNumbersGeneratedIterable = (async function*() {
  for (let i = 0; i < 5; i++) { yield i + 1 }
})() // generated async iterable that yields 1 2 3 4 5

const concat = (y, xi) => y.concat([xi])

reduce(
  concat, [],
)(asyncNumbersGeneratedIterable) // => Promise { [1, 2, 3, 4, 5] }

reduce(
  concat, [],
)({ a: 1, b: 1, c: 1, d: 1, e: 1 }) // => [1, 2, 3, 4, 5]

back to documentation

transform

transforms data in series according to provided transducer and initial value

y = transform(f, x0)(x)

x is an iterable, an async iterable, or an object

f is a transducer, see transducers

x0 is null, an array, a string, a set, a map, a typed array, or a writable

y is x transformed with f into x0

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous
  • x is an async iterable

in the following examples, map is map

const square = x => x ** 2

transform(map(
  square,
), null)([1, 2, 3, 4, 5]) // => null

const asyncSquare = async x => x ** 2

transform(map(
  asyncSquare,
), [])([1, 2, 3, 4, 5]) // => Promise { [1, 4, 9, 16, 25] }

transform(map(
  square,
), '')([1, 2, 3, 4, 5]) // => '1491625'

transform(map(
  square,
), new Set())([1, 2, 3, 4, 5]) // => Set { 1, 4, 9, 16, 25 }

transform(map(
  number => [number, square(number)],
), new Map())([1, 2, 3, 4, 5]) // => Map { 1 => 1, 2 => 4, 3 => 9, 4 => 16, 5 => 25 }

const charToByte = x => x.charCodeAt(0)

transform(map(
  square,
), new Uint8Array())([1, 2, 3, 4, 5]), // => Uint8Array [1, 4, 9, 16, 25]

const asyncNumbersGeneratedIterable = (async function*() {
  for (let i = 0; i < 5; i++) { yield i + 1 }
})() // generated async iterable that yields 1 2 3 4 5

transform(map(
  square,
), process.stdout)(asyncNumbersGeneratedIterable) // > 1 4 9 16 25
// => Promise { process.stdout }

back to documentation

any

applies a function to each element of data parallel, returns true if any evaluations truthy

y = any(f)(x)

x is an iterable or an object

xi is an element of x

f is a function that expects one argument xi

y is true if all f(xi) are truthy, false otherwise

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous
const isOdd = x => x % 2 === 1

any(
  isOdd,
)([1, 2, 3, 4, 5]) // => true

const asyncIsOdd = async x => x % 2 === 1

any(
  asyncIsOdd,
)([1, 2, 3, 4, 5]) // => Promise { true }

any(
  isOdd,
)({ b: 2, d: 4 }) // => false

back to documentation

all

applies a function to each element of data in parallel, returns true if all evaluations truthy

y = all(f)(x)

x is an iterable or an object

xi is an element of x

f is a function that expects one argument xi

y is true if any f(xi) are truthy, false otherwise

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous
const isOdd = x => x % 2 === 1

all(
  isOdd,
)([1, 2, 3, 4, 5]) // => false

const asyncIsOdd = async x => x % 2 === 1

all(
  asyncIsOdd,
)([1, 2, 3, 4, 5]) // => Promise { false }

all(
  isOdd,
)({ a: 1, c: 3 }) // => true

back to documentation

and

applies each function of functions in parallel to data, returns true if all evaluations truthy

y = and(functions)(x)

x is anything

functions is an array of functions

f is a function of functions

y is true if all f(x) are truthy, false otherwise

y is wrapped in a Promise if any of the following are true:

  • any f is asynchronous
const isOdd = x => x % 2 === 1

const asyncIsOdd = async x => x % 2 === 1

const lessThan3 = x => x < 3

and([
  isOdd,
  lessThan3,
])(1) // => true

and([
  asyncIsOdd,
  lessThan3,
])(1) // => Promise { true }

and([
  isOdd,
  lessThan3,
])(2) // => false

back to documentation

or

applies each function of functions in parallel to data, returns true if any evaluations truthy

y = or(functions)(x)

x is anything

functions is an array of functions

f is a function of functions

y is true if any f(x) are truthy, false otherwise

y is wrapped in a Promise if any of the following are true:

  • any f is asynchronous
const isOdd = x => x % 2 === 1

const asyncIsOdd = async x => x % 2 === 1

const lessThan3 = x => x < 3

or([
  isOdd,
  lessThan3,
])(5) // => true

or([
  asyncIsOdd,
  lessThan3,
])(5) // => Promise { true }

or([
  isOdd,
  lessThan3,
])(6) // => false

back to documentation

not

applies a function to data, logically inverting the result

y = not(f)(x)

x is anything

f is a function that expects one argument x

y is true if f(x) is falsy, false otherwise

y is wrapped in a Promise if any of the following are true:

  • f is asynchronous
const isOdd = x => x % 2 === 1

const asyncIsOdd = async x => x % 2 === 1

not(
  isOdd,
)(2) // => true

not(
  asyncIsOdd,
)(2) // => Promise { true }

not(
  isOdd,
)(3) // => false

back to documentation

eq

tests left strictly equals right

y = eq(left, right)(x)

x is anything

left is a non-function value or a function that expects one argument x

right is a non-function value or a function that expects one argument x

leftCompare is left if left is a non-function value, else left(x)

rightCompare is right if right is a non-function value, else right(x)

y is true if leftCompare strictly equals rightCompare, false otherwise

y is wrapped in a Promise if any of the following are true:

  • left is asynchronous
  • right is asynchronous
const square = x => x ** 2

const asyncSquare = async x => x ** 2

eq(
  square,
  1,
)(1) // => true

eq(
  asyncSquare,
  1,
)(1) // => Promise { true }

eq(
  square,
  asyncSquare,
)(1) // => Promise { true }

eq(
  1,
  square,
)(2) // => false

eq(1, 1)() // => true
eq(0, 1)() // => false

back to documentation

gt

tests left greater than right

y = gt(left, right)(x)

x is anything

left is a non-function value or a function that expects one argument x

right is a non-function value or a function that expects one argument x

leftCompare is left if left is a non-function value, else left(x)

rightCompare is right if right is a non-function value, else right(x)

y is true if leftCompare is greater than rightCompare

y is wrapped in a Promise if any of the following are true:

  • left is asynchronous
  • right is asynchronous
gt(
  x => x,
  10,
)(11) // => true

gt(
  async x => x,
  10,
)(11) // => Promise { true }

gt(
  x => x,
  10,
)(9) // => false

gt(2, 1)() // => true
gt(1, 1)() // => false
gt(0, 1)() // => false

back to documentation

lt

tests left less than right

y = lt(left, right)(x)

x is anything

left is a non-function value or a function that expects one argument x

right is a non-function value or a function that expects one argument x

leftCompare is left if left is a non-function value, else left(x)

rightCompare is right if right is a non-function value, else right(x)

y is true if leftCompare is less than rightCompare

y is wrapped in a Promise if any of the following are true:

  • left is asynchronous
  • right is asynchronous
lt(
  x => x,
  10,
)(9) // => true

lt(
  async x => x,
  10,
)(9) // => Promise { true }

lt(
  x => x,
  10,
)(11) // => false

lt(0, 1)() // => true
lt(1, 1)() // => false
lt(2, 1)() // => false

back to documentation

gte

tests left greater than or equal right

y = gte(left, right)(x)

x is anything

left is a non-function value or a function that expects one argument x

right is a non-function value or a function that expects one argument x

leftCompare is left if left is a non-function value, else left(x)

rightCompare is right if right is a non-function value, else right(x)

y is true if leftCompare is greater than or equal to rightCompare

y is wrapped in a Promise if any of the following are true:

  • left is asynchronous
  • right is asynchronous
gte(
  x => x,
  10,
)(11) // => true

gte(
  async x => x,
  10,
)(11) // => Promise { true }

gte(
  x => x,
  10,
)(9) // => false

gte(2, 1)() // => true
gte(1, 1)() // => true
gte(0, 1)() // => false

back to documentation

lte

tests left less than or equal right

y = lte(left, right)(x)

x is anything

left is a non-function value or a function that expects one argument x

right is a non-function value or a function that expects one argument x

leftCompare is left if left is a non-function value, else left(x)

rightCompare is right if right is a non-function value, else right(x)

y is true if leftCompare is less than or equal to rightCompare

y is wrapped in a Promise if any of the following are true:

  • left is asynchronous
  • right is asynchronous
lte(
  x => x,
  10,
)(9) // => true

lte(
  async x => x,
  10,
)(9) // => Promise { true }

lte(
  x => x,
  10,
)(11) // => false

lte(0, 1)() // => true
lte(1, 1)() // => true
lte(2, 1)() // => false

back to documentation

get

accesses a property by path

y = get(path, defaultValue)(x)

x is an object

path is a number, string, a dot-delimited string, or an array

defaultValue is optional; if not provided, it is undefined

y depends on path:

  • if path is a number or string, y is x[path]
  • if path is a dot-delimited string 'p.a...t.h', y is x['p']['a']...['t']['h']
  • if path is an array ['p', 'a', ..., 't', 'h'], y is x['p']['a']...['t']['h']
  • if path is not found in x, y is defaultValue
get('a')({ a: 1, b: 2 }) // => 1

get('a')({}) // => undefined

get('a', 10)({}) // => 10

get(0)(['hello', 'world']) // => 'hello'

get('a.b.c')({ a: { b: { c: 'hey' } } }) // => 'hey'

get([0, 'user', 'id'])([
  { user: { id: '1' } },
  { user: { id: '2' } },
]) // => '1'

back to documentation

pick

constructs a new object from data composed of the provided properties

y = pick(properties)(x)

x is an object

properties is an array of strings

y is an object composed of all properties enumerated in properties and defined in x

pick(['a', 'b'])({ a: 1, b: 2, c: 3 }) // => { a: 1, b: 2 }

pick(['d'])({ a: 1, b: 2, c: 3 }) // => {}

back to documentation

omit

constructs a new object from data without the provided properties

y = omit(properties)(x)

x is an object

properties is an array of strings

y is an object composed of every property in x except for those enumerated in properties

omit(['a', 'b'])({ a: 1, b: 2, c: 3 }) // => { c: 3 }

omit(['d'])({ a: 1, b: 2, c: 3 }) // => { a: 1, b: 2, c: 3 }

Transducers

Transducers enable us to wrangle very large or infinite streams of data in a
composable and memory efficient way. Say you had veryBigData in an array

veryBigData = [...]
veryBigFilteredData = veryBigData.filter(datum => datum.isBig === true)
veryBigProcessedData = veryBigFilteredData.map(memoryIntensiveProcess)
console.log(veryBigProcessedData)

The above is not very memory efficient because of the intermediate arrays veryBigFilteredData
and veryBigProcessedData. We're also logging out a large quantity of data at once to the console.

With rubico, you could express the above transformation as a single pass
without incurring a memory penalty

veryBigData = [...]
transform(pipe([
  filter(datum => datum.isBig === true),
  map(memoryIntensiveProcess),
]), process.stdout)(veryBigData)

In this case, pipe([filter(...), map(...)]) is a transducer, and we're writing each datum
to the console via process.stdout. transform consumes our pipe([filter(...), map(...)])
transducer and supplies it with veryBigData.

Behind the scenes, transform is calling reduce with a reducing function suitable for writing
to process.stdout converted from the transducer pipe([filter(...), map(...)])

reducer is an alias for reducing function, very much the same as the one supplied to reduce

y = reduce(reducer)(x)

A reducer takes two arguments: an aggregate y and an iterative value xi.
It can be something like (y, xi) => doSomethingWith(y, xi)

A transducer is a function that takes a reducer and returns another reducer

transducer = reducer => (y, xi) => reducer(doSomethingWith(y, xi))

The transducer above, when passed a reducer, returns another reducer that will do something
with y and xi, then pass it to the input reducer

We can create a chained reducer by passing a reducer to a chain of transducers

Imagine dominos falling over. The reducer you pass to a chain of transducers is called last.
Because of this implementation detail,

if x is a function, pipe chains functions from right to left

You can use pipe to construct chains of transducers. Pipe will read left to right in all cases.

There are two other functions you'll need to get started with transducers, map and filter.

given x is a reducer, f is a mapping function; map(f)(x) is a transduced reducer
that applies f to each element in the final transform pipeline.

given x is a reducer, f is a predicate function; filter(f)(x) is a transduced reducer
that filters each element in the final transform pipeline based on f

The following transformations isOdd, square, and squaredOdds are used as transducers

const concat = (y, xi) => y.concat([xi])

const isOdd = filter(x => x % 2 === 1)

transform(isOdd, [])([1, 2, 3, 4, 5]) // => [1, 3, 5]
reduce(
  isOdd(concat),
  [],
)([1, 2, 3, 4, 5]) // => [1, 3, 5]

const square = map(x => x ** 2)

transform(square, [])([1, 2, 3, 4, 5]) // => [1, 4, 9, 16, 25]
reduce(
  square(concat),
  [],
)([1, 2, 3, 4, 5]) // => [1, 4, 9, 16, 25]

const squaredOdds = pipe([isOdd, square])

transform(squaredOdds, [])([1, 2, 3, 4, 5]) // => [1, 9, 25]
reduce(
  squaredOdds(concat),
  [],
)([1, 2, 3, 4, 5]) // => [1, 9, 25]

The following transformations isOdd, square, and squaredOdds are not used as transducers

const isOdd = filter(x => x % 2 === 1)

isOdd([1, 2, 3, 4, 5]) // => [1, 3, 5]

const square = map(x => x ** 2)

square([1, 2, 3, 4, 5]) // => [1, 4, 9, 16, 25]

const squaredOdds = pipe([isOdd, square])

squaredOdds([1, 2, 3, 4, 5]) // => [1, 9, 25]

Examples

A webserver using map, transform, and https://deno.land/std/http serve

import { serve } from "https://deno.land/std/http/server.ts";
import { map, transform } from "https://deno.land/x/rubico/rubico.js"
const s = serve({ port: 8001 });
console.log("http://localhost:8001/");
transform(map(req => {
  req.respond({ body: "Hello World\n" });
}), null)(s);

A server with middleware

import { serve } from 'https://deno.land/std/http/server.ts'
import {
  pipe, fork, assign, tap, tryCatch, switchCase,
  map, filter, reduce, transform,
  any, all, and, or, not,
  eq, gt, lt, gte, lte,
  get, pick, omit,
} from 'https://deno.land/x/rubico/rubico.js'

const join = delim => x => x.join(delim)

const addServerTime = req => {
  req.serverTime = (new Date()).toJSON()
  return req
}

const traceRequest = pipe([
  fork([
    pipe([get('serverTime'), x => '[' + x + ']']),
    get('method'),
    get('url'),
  ]),
  join(' '),
  console.log,
])

const respondWithHelloWorld = req => {
  req.respond({ body: 'Hello World\n' })
}

const respondWithServerTime = req => {
  req.respond({ body: `The server time is ${req.serverTime}\n` })
}

const respondWithNotFound = req => {
  req.respond({ body: 'Not Found\n' })
}

const route = switchCase([
  eq('/', get('url')), respondWithHelloWorld,
  eq('/time', get('url')), respondWithServerTime,
  respondWithNotFound,
])

const onRequest = pipe([
  addServerTime,
  tap(traceRequest),
  route,
])

const s = serve({ port: 8001 })
console.log('http://localhost:8001/')
transform(map(onRequest), null)(s)