DEPRECATED: This library is too large to keep in one repository. I have moved to a multi-repository solution: github.com/unctionjs

@unction/*

A set of very useful function. These functions are bound by these principles:

1. All functions will have type annotations.
2. All functions will only take a single argument.
3. All functions are curried.
4. All inner functions will be named based on the outer function and it's argument name.
5. Functions that deal with key values will take KeyChains (a list of keys).
6. Functions that deal with types will take their string name form.
7. Functions that mutate the original value, though rare, will have a suffix of M.
8. Functions that take or return promises will have a suffix of P.
9. Functions that can work on one type of Iterable can work on another type, covering:

• List (Array, Set, WeakSet)
• Record (object, Map, WeakMap)
• String
• Stream (xstream)

using

Each function is a unique package with it's own source, transpiliation, and tests. To get a function simply install:

$ npm install --save-dev @unction/hammer@latest

and then import:

import hammer from "@unction/hammer"

@unction/allObjectP

({[key: KeyType]: any | Promise}) => Promise<{[key: KeyType]: any}>

This takes an object where the values are probably promises and returns a promise that has that same object but with the resolved values.

Here's a good example of this function in use:

function signUp (attributes, slug) {
  return function thunk (dispatch, getState, {client}) {
    return allObjectP({
      loading: startLoading(slug),
      session: pushSession(attributes, client)
    })
      .then(({session}) => {
        return allObjectP({
          merge: mergeResource(session),
          current: storeCurrent(session.id),
          account: pullAccount(session.relationship.account.data.id, client),
        })
      })
      .then(({account}) => {
        return {
          merge: mergeResource(account),
          current: storeCurrent(account.id),
        }
      })
      .then(() => stopLoading(slug))
      .then(() => dispatch({type: "signUp"}))
      .catch(logger.error.bind(error))
  }
}

If we use allP or Promise.all we're getting an array back, but that's annoying to destructure. The allObjectP function gives us the concurrency we want with a named interface for the returned resolutions.

@unction/allP

Array<any | Promise> -> Promise<Array>

A port of the Promise.all() function.

Credit: @keithamus

@unction/always

any -> any -> any

Always returns the value given when called

always(1)() // 1
always(1)(0) // 1

@unction/append

any -> Array -> Array

Takes a value and puts it at the end of the given list

append(4)([5]) // => [5, 4]

@unction/appendM

any -> Array -> Array

Takes an array and an item and returns the combination of both, appended.

NOTE: This mutates the array

const data = [1, 2, 3]

appendM(4)(data)

Would return:

[1, 2, 3, 4]

@unction/applicator

(ValueType -> any) -> any

Takes a function and a value and applies that function to that value.

applicator(inc)(1) // 1

@unction/applicators

Iterable<ValueType -> ValueType> -> Iterable -> Array

Takes a list of functions and a list of values and applies the values to the functions.

applicators([
  recordfrom(["point", "x"]),
  recordfrom(["point", "y"]),
  recordfrom(["point", "z"]),
])([
  40.453,
  2.2,
  423.0,
])

returns

[
  {point: {x: 40.453}},
  {point: {y: 2.2}},
  {point: {z: 423.0}},
]

applicators({
  x: inc,
  y: dec
})({
  x: -1,
  y: 1
})

returns

{
  x: 0,
  y: 0
}

@unction/arrayify

any -> [any] | Array

Takes a value and turns it into an array of that value, unless the value is already an array.

arrayify("a")

returns

["a"]

arrayify(["a"])

returns

["a"]

@unction/aside

Array<(any => any)> -> any -> any

Use this function to do thing aside of a main pipeline.

export default function generateGraph () {
  return pipe(
    aside([format, log]),
    processData
  )
}

@unction/attach

KeyType -> ValueType -> IterableType -> IterableType

A polymorphic way to attach a value to an iterable

attach("hello")("world")({}) // => {hello: "world"}

@unction/cascadingKeyChain

Array -> IterableType -> ValueType

Cascades through multiple keychains in order to arrive at a value. Null keys are replaced with the previous keychain's value.

cascadingKeyChain(
  [
    ["ephemeral", "current", "session"],
    ["resources", "sessions", null, "relationships", "account", "data", "id"],
    ["resources", "accounts", null, "attributes", "name"]
  ]
)(
  {
    ephemeral: {current: {session: "1"}},
    resources: {
      sessions: {
        1: {
          id: "1",
          relationships: {account: {data: {id: "2"}}},
        },
      },
      accounts: {
        2: {
          id: "2",
          attributes: {name: "Kurtis Rainbolt-Greene"},
        },
      },
    },
  }
)

returns

"Kurtis Rainbolt-Greene"

@unction/catchP

(any -> any) -> Promise -> Promise

A port of the Promise.prototype.catch() function.

Credit: @keithamus

@unction/compact

Array -> Array

Takes a collection (Array or Object) and returns a copy of that value without null or undefined values.

avatarUrls // => [null, "/1.jpg", null, "/3.jpg"]
compact(avatarUrls)  // => ["/1.jpg", "/3.jpg"]

head(users) // {"avatar": null, "name": "Kurtis Rainbolt-Greene"}
compact(head(users)) // {"name": "Kurtis Rainbolt-Greene"}

@unction/computedProp

(IterableType -> any) -> Keychain -> IterableType -> IterableType

Given an object this function will return that object but with a new property, where the value is computed. The computation is given the object you'll be copying.

const computer = ({id, attributes: {username}}) => `${username}#${id}`
const key = "tag"
const payload = {
  id: "1",
  attributes: {
    username: "krainboltgreene"
  }
}

computedProp(computer)(key)(payload)

Would return:

{
  id: "1",
  tag: "krainboltgreene#1",
  attributes: {
    username: "krainboltgreene"
  }
}

const multiKey = ["attributes", "tag"]

computedProp(computer)(key)(payload)

Would return:

{
  id: "1",
  attributes: {
    tag: "krainboltgreene#1",
    username: "krainboltgreene"
  }
}

@unction/couple

any -> any -> [any, any]

Takes any value and then any value and returns an array containing those values.

couple(4)(5) // => [4, 5]

@unction/domEvents

DOMEventsConfigurationType -> DOMEventNameType -> DOMStreamType -> DOMEventStreamType

Takes a configuration, an event name, and a DOM source and returns an observable of that event type

domEvents({})("click")(DOM)

returns

--click--click--click-->

@unction/domEventsMany

domEventsManyConfigurationType -> (string | Array) -> DOMEventStreamType

Takes many events or * and returns an observable of those events

domEventsMany({})(["click", "input"])(DOM)

returns

--click--input--input--click--input

domEventsMany({})("*")(DOM)

returns

--click--input--hover--change--click-->

@unction/endsWith

string -> string -> boolean

Determines if a given subset of text is at the end of another set of text.

endsWith("!")("Hello, world!") // true

@unction/everyP

(Array<any | Promise>) -> Promise<[ResolvedPromisesType, RejectedPromisesType]>

Returns both resolved and rejected promises as distinct lists.

@unction/flattenTree

((any -> any) -> IterableType -> IterableType) -> (any -> any) -> number -> IterableType -> IterableType

Takes a tree and creates a single object where the root keys are conjoined nested keys.

flattenTree({
  data: {
    profile: {
      name: "Kurtis Rainbolt-Greene"
      age: 24
    },
    metadata: {
      interval: "10s"
    },
    location: "http://api.example.com/profiles/24"
  }
})

Would return:

{
  "data-profile-name": "Kurtis Rainbolt-Greene",
  "data-profile-age": 24,
  "data-interval": "10s",
  "data-location": "Kurtis Rainbolt-Greene"
}

@unction/flip

(any -> any) -> any -> any -> any

Flips a function's first and second arguments.

flip(key)({aaa: "1"})("aaa") // "1"

@unction/forEach

(any -> KeyType -> any) -> IterableType -> IterableType

Takes any kind of iterable object and figures out the best way to iterate over it.

forEach((x) => y)([])
forEach((x) => y)(new Map)
forEach((x) => y)({})

@unction/fresh

IterableType -> IterableType

Takes a iterable and returns an empty fresh version of that iterable.

fresh({aaa: "aaa"}) // {}
fresh(["aaa"]) // []
fresh({}) // {}
fresh([]) // []

@unction/hammer

KeyType -> IterableType -> IterableType

Use this to de-nest a nested object.

const payload = {
  id: 1
  attributes: {
    name: "Kurtis Rainbolt-Greene",
    age: 26
  }
}

hammer("attributes")(payload)

Which returns:

{
  id: 1,
  name: "Kurtis Rainbolt-Greene",
  age: 26
}

@unction/ifThenElse

PredicateType -> (any -> any) -> (any -> any) -> any

Based on a predicate it passes the value to a consequent or alternative function

ifThenElse(isEven)(toString)(toFloat)(1) // 1.0
ifThenElse(isEven)(toString)(toFloat)(2) // "2"

@unction/isArray

mixed -> boolean

Takes any value and then any value and returns an array containing those values.

isArray([]) // => true
isArray({}) // => false
isArray("") // => false

@unction/isIterable

mixed -> boolean

Takes any value and then any value and returns an array containing those values.

isIterable({}) // => true
isIterable([]) // => true
isIterable("") // => true
isIterable(1) // => false

@unction/isNil

mixed -> boolean

Determines if a value is not a value.

isNil(null) // true
isNil(undefined) // true
isNil(0) // false
isNil("") // false
isNil([]) // false
isNil({}) // false

@unction/isObject

mixed -> boolean

Takes a value and determines if it's an object.

isObject({}) // => true
isObject([]) // => false
isObject("") // => false

@unction/isPopulated

mixed -> boolean

Allows you to check if a iterable has any items.

isPopulated([1]) // true
isPopulated({a: 'b'}) // true
isPopulated({}) // false
isPopulated([]) // false
isPopulated("") // false
isPopulated("a") // true

@unction/isPresent

mixed -> boolean

This lets you know if it's a non-null, non-undefined value.

isPresent('x') // true
isPresent([]) // true
isPresent(null) // false
isPresent(undefined) // false

@unction/isType

string -> mixed -> boolean

Takes any value and then any value and returns an array containing those values.

isType("Object")({}) // => true
isType("Array")([]) // => true
isType("String")("") // => true

@unction/itself

any -> any

Always returns the value given when calling.

itself(1) // 1
itself(1) // 1

@unction/key

KeyType -> any -> ValueType

Returns the value of a specific key on an iterable. If no key is found it returns undefined. If the second argument isn't an iterable we return undefined, to allow for graceful failure.

key("aaa")({aaa: "1"}) // "1"
key("bbb")({aaa: "1"}) // undefined
key("bbb")(undefined) // undefined
key(0)(["aaa"]) // "aaa"

@unction/keyChain

KeyChainType -> TreeType -> ValueType

Takes a chain of keys and a tree, traversing down and reaching the last value. If any part of the chain is undefined or not an object the result will always be undefined.

keyChain(["aaa", "bbb", "ccc"])({aaa: {bbb: {ccc: "1"}}}) // "1"
keyChain(["aaa", "ddd", "ccc"])({aaa: {bbb: {ccc: "1"}}}) // undefined

@unction/lacksText

(string | RegExp) -> string -> boolean

Determines if a set of text does not have a subset of text.

const data = "I love pies!"
const lacksBestFood = lacksText("pizza")

lacksBestFood(data) // false

const data = "I love pies!"
const lacksPuncation = lacksText(/\!|\?|\./)

lacksPuncation(data) // false

@unction/mapKeys

(KeyType -> KeyType) -> IterableType -> IterableType

Like ramda's map, but instead of the value it maps over keys.

const attributes = {
  name: "Kurtis Rainbolt-Greene",
  createdAt: new Date()
}

mapKeys(kebab)(attributes)

Would return:

{
  name: "Kurtis Rainbolt-Greene",
  "created-at": new Date()
}

@unction/mapKeysWithValueKey

(ValueType => KeyType => KeyType) -> IterableType -> IterableType

Map over keys with the context of the value and key.

const attributes = {
  name: "Kurtis Rainbolt-Greene",
  createdAt: new Date()
}

mapKeys((value) => (key) => )(attributes)

Would return:

{
  name: "Kurtis Rainbolt-Greene",
  "created-at": new Date()
}

@unction/mapValues

(any -> any) -> IterableType -> IterableType

A pretty standard mapValues(), but with enforced unary currying.

mapValues(
  (value) => value + 1
)(
  [1, 2, 3]
)

Which will return:

[2, 3, 4]

@unction/mapValuesWithValueKey

(any => KeyType => any) -> IterableType -> IterableType

Just like map, but gives back the index argument (as an integer, not a string if array)

@unction/mergeAllLeft

Array -> IterableType

Merges a list of iterables (of the same type) into a single iterable.

mergeAllLeft([["0"], ["1"], ["2"]]) // ["0", "1", "2"]
mergeAllLeft([{aaa: "aaa"}, {bbb: "bbb"}, {ccc: "ccc"}]) // {aaa: "aaa", bbb: "bbb", ccc: "ccc",}

@unction/mergeAllRight

Array<IterableType> -> IterableType

Merges a list of iterables (of the same type) into a single iterable.

mergeAllRight([["0"], ["1"], ["2"]]) // ["0", "1", "2"]
mergeAllRight([{aaa: "aaa"}, {bbb: "bbb"}, {ccc: "ccc"}]) // {aaa: "aaa", bbb: "bbb", ccc: "ccc",}

@unction/mergeDeepLeft

IterableType -> IterableType -> IterableType

Recursively merges two objects/arrays. Merges objects with merge and arras with concat. Prefers left. THAT IS ALL.

const left = {
  alpha: "1"
}
const right = {
  beta: "2"
}

mergeDeepLeft(left)(right)
{
  alpha: "1",
  beta: "2"
}

const left = {
  alpha: {
    alpha1: "1"
  }
}
const right = {
  beta: {
    beta1: "1"
  }
}

mergeDeepLeft(left)(right)
{
  alpha: {
    alpha1: "1"
  },
  beta: {
    beta1: "1"
  }
}

const left = {
  alpha: [
    "1"
  ]
}
const right = {
  alpha: [
    "1"
  ]
}

mergeDeepLeft(left)(right)
{
  alpha: [
    "1",
    "1"
  ]
}

@unction/mergeDeepRight

IterableType -> IterableType -> IterableType

Recursively merges two objects/arrays. Merges objects with merge and arras with concat. Prefers right. THAT IS ALL.

const left = {
  alpha: "1"
}
const right = {
  beta: "2"
}

mergeDeepRight(left)(right)
{
  alpha: "1",
  beta: "2"
}

const left = {
  alpha: {
    alpha1: "1"
  }
}
const right = {
  beta: {
    beta1: "1"
  }
}

mergeDeepRight(left)(right)
{
  alpha: {
    alpha1: "1"
  },
  beta: {
    beta1: "1"
  }
}

const left = {
  alpha: [
    "1"
  ]
}
const right = {
  alpha: [
    "1"
  ]
}

mergeDeepRight(left)(right)
{
  alpha: [
    "1",
    "1"
  ]
}

@unction/mergeLeft

IterableType -> IterableType -> IterableType

Merges two iterables, preferring left.

const left = {
  alpha: "1",
  beta: "1"
}
const right = {
  beta: "2",
  zeta: "3"
}

mergeLeft(left)(right)

Which returns:

{
  alpha: "1",
  beta: "1",
  zeta: "3"
}

@unction/mergeRight

IterableType -> IterableType -> IterableType

mergeRights two iterables.

const left = {
  alpha: "1",
  beta: "1"
}
const right = {
  beta: "2",
  zeta: "3"
}

mergeRight(left)(right)

Which returns:

{
  alpha: "1"
  beta: "2",
  zeta: "3"
}

@unction/mergeWith

(ValueType -> ValueType -> any) -> IterableType -> IterableType -> IterableType

Merges two iterables and uses a provided function to handle conflicts. The function is given the the left value and the right value.

const left = {
  alpha: "0",
  beta: "1",
  zeta: "3"
}
const right = {
  alpha: "0",
  beta: "2",
  zeta: "3"
}

mergeWith((l) => (r) => l+r)(left)(right)

Which returns:

{
  alpha: "0",
  beta: "12",
  zeta: "3"
}

@unction/mergeWithKey

(IterableType -> IterableType -> KeyType -> any) -> IterableType -> IterableType -> IterableType

Merges two iterables and uses a provided function to handle conflicts. The function is given the key, the left value, and the right value.

const left = {
  beta: "1"
}
const right = {
  beta: "2"
}

mergeWith((key, leftValue, rightValue) => key+leftValue+rightValue)(left)(right)

Which returns:

{
  beta: "beta12"
}

@unction/nestedApply

((any -> any) -> IterableType -> IterableType) -> (any -> any) -> number -> IterableType -> IterableType

Takes a function (the application) that takes function(s) (later referred to as the inner) and value(s) (map(), forEach(), find()), a function (the inner) that will be applied to a value(s), and finally a number (depth) to apply that applicator around the inner.

In the below example we want to take two sets of records and index them by id:

const records = [
  [
    {
      id: "a1",
      type: "commercial",
    },
    {
      id: "a2",
      type: "commercial",
    }
  ],
  [
    {
      id: "b1",
      type: "residential",
    },
    {
      id: "b2",
      type: "residential",
    }
  ]
]

Normally we'd just do mapValues(indexBy(key("id"))), however we can make this easier and dynamic:

const nestedIndexById = nestedApply(mapValues)(indexBy(key("id")))(1)

nestedIndexById(records)

And the result:

[
  {
    a1: {
      id: "a1",
      type: "commercial",
    },
    a2: {
      id: "a2",
      type: "commercial",
    },
  },
  {
    b1: {
      id: "b1",
      type: "residential",
    },
    b2: {
      id: "b2",
      type: "residential",
    },
  },
]

@unction/optimisticP

Array<any | Promise> -> Promise<Array>

Will take an array of promises and returns a promise of only the resolved promises.

@unction/pairsKeys

Array<[string, any]> -> Array

Takes an array that looks like a list of pairs (key, values) and returns all the keys.

Lets say you have this data:

const data = {
  a: 1,
  b: 2,
  c: 3,
}

And you turn it into pairs:

const pairings = toPairs(data)

You would end up with this:

[
  ['a', 1],
  ['b', 2],
  ['c', 3],
]

Now you just want the keys:

pairsKeys(pairings)

You would get the following:

[
  'a',
  'b',
  'c',
]

@unction/pairsValues

Array<[KeyType, any]> -> Array

Takes an array that looks like a list of pairs (key, values) and returns all the values.

Lets say you have this data:

const data = {
  a: 1,
  b: 2,
  c: 3,
}

And you turn it into pairs:

const pairings = toPairs(data)

You would end up with this:

[
  ['a', 1],
  ['b', 2],
  ['c', 3],
]

Now you just want the keys:

pairsValues(pairings)

You would get the following:

[
  1,
  2,
  3,
]

@unction/pluck

KeyChainType -> IterableType -> Array

Given a keychain and records return the values at the keychain for each record.

pluck(
  ["attributes", "name"]
)(
  [
    {
      id: "1",
      attributes: {
        name: "Kurtis",
        age: 29,
        height: "5'10\"",
      },
    },
    {
      id: "2",
      attributes: {
        name: "Chris",
        age: 29,
        height: "5'8\"",
      },
    },
  ]
)

Which will return:

[
  "Kurtis",
  "Chris"
]

@unction/plucks

Array -> IterableType -> Array

Given keychain and records, return the values at the keychain for each record.

plucks(
  [
    ["attributes", "name"],
    ["attributes", "age"],
    ["attributes", "friends"],
    ["id"]
  ]
)(
  [
    {
      id: "1",
      attributes: {
        name: "Kurtis",
        age: 29,
        height: "5'10\"",
      },
    },
    {
      id: "2",
      attributes: {
        name: "Chris",
        age: 29,
        height: "5'8\"",
      },
    },
  ]
)

Which will return:

[
  ["Kurtis", 29, null, "1"],
  ["Chris", 29, null, "2"]
]

@unction/prepend

any -> Array -> Array

Takes a value and puts it at the end of the given list

prepend(4)([5]) // => [4, 5]

@unction/recordFrom

KeyChain -> any -> IterableType

Given a keychain and a value it creates an object that has keys based on the keychain.

recordFrom(["key", "subkey"])("value")

Which returns:

{
  key: {
    subkey: "value"
  }
}

@unction/reduceValues

(AccumulatedType -> ValueType -> AccumulatedType) -> InitialType -> IterableType -> AccumulatedType

A pretty standard reduceValues(), but where the 𝑓() is unary curries.

reduceValues(
  (accumulation) => (current) => `${accumulation}/${current}`
)(
  "~"
)(
  ["Users", "krainboltgreene", "Code"]
)

Which will return:

"~/Users/krainboltgreene/Code"

@unction/reduceWithValueKey

(AccumulatedType -> ValueType -> KeyType -> AccumulatedType) -> InitialType -> IterableType -> AccumulatedType

A pretty standard reduceWithValueKey(), but where the 𝑓() is unary curries.

reduceWithValueKey(
  (accumulation) => (current) => (key) => `${accumulation}/${current}:${key}`
)(
  "~"
)(
  ["Users", "krainboltgreene", "Code"]
)

Which will return:

"~/Users:0/krainboltgreene:1/Code:2"

@unction/rejectP

any -> Promise

A port of the Promise.reject() function.

Credit: @keithamus

@unction/replacewhen

PredicateType -> ValueType -> IterableType

Replaces values in an iterable with another value based on a predicate.

replaceWhen(isEven)(null)([1, 2, 3]) // [1, null, 3]

@unction/resolveP

any -> Promise

A port of the Promise.resolve() function.

Credit: @keithamus

@unction/sample

(string | Array) -> (string | any)

Takes an Array or string and randomly one element to return.

users() // => [{"id": 1, "name": "Kurtis Rainbolt-Greene"}, {"id": 2, "name": "Angela Englund"}]

sample(users()) // => {"id": 2, "name": "Angela Englund"}

sample(users()) // => {"id": 2, "name": "Angela Englund"}

sample(users()) // => {"id": 1, "name": "Kurtis Rainbolt-Greene"}

@unction/sampleSize

number -> (string | Array) -> (string | any)

Takes an Array or string and randomly picks n elements to return, but never the same one.

users() // => [{"id": 1, "name": "Kurtis Rainbolt-Greene"}, {"id": 2, "name": "Angela Englund"}]

sample(1)(users()) // => [{"id": 2, "name": "Angela Englund"}]

sample(2)(users()) // => [{"id": 2, "name": "Angela Englund"}, {"id": 1, "name": "Kurtis Rainbolt-Greene"}]

@unction/shuffle

(string | Array) -> (string | Array)

Takes an Array and returns an Array with the same content, but in a random order.

users()

Would return:

[{"id": 1, "name": "Kurtis Rainbolt-Greene"}, {"id": 2, "name": "Angela Englund"}, {"id": 3, "name": "Joshua Benitez"}]

shuffle(users())

Would return:

[{"id": 1, "name": "Kurtis Rainbolt-Greene"}, {"id": 3, "name": "Joshua Benitez"}, {"id": 2, "name": "Angela Englund"}]

shuffle(users())

Would return:

[{"id": 3, "name": "Joshua Benitez"}, {"id": 1, "name": "Kurtis Rainbolt-Greene"}, {"id": 2, "name": "Angela Englund"}]

shuffle(users())

Would return:

[{"id": 2, "name": "Angela Englund"}, {"id": 3, "name": "Joshua Benitez"}, {"id": 1, "name": "Kurtis Rainbolt-Greene"}]

@unction/splat

(ValueType -> any) -> Array -> any

Takes a function and a list of values and recursively applies the value to the functions.

splat((a) => (b) => a + b)([1, 2]) // 3

@unction/startsWith

string -> string -> boolean

Determines if a given subset of text is at the start of another set of text.

startsWith("Hello")("Hello, world!") // true

@unction/thenCatchP

(any -> any) -> (any -> any) -> Promise -> Promise

A port of the Promise.prototype.then() function, but with the extra catch argument.

Credit: @keithamus

@unction/thenP

(any -> any) -> Promise -> Promise

A port of the Promise.prototype.then() function.

Credit: @keithamus

@unction/thrush

any -> (any -> any) -> any

One of the fantasy birds: it takes a value, a function, and then applies that value to as the first argument to that function.

thrush(0)((value) => `${value}`) // "0"

@unction/treeify

Array<(any -> IterableType -> IterableType)> -> Array -> IterableType

This takes a list of functions (the folders) and an array of objects or an object of objects (the collection) to create a tree. Each function in the list of folders will in some way return a new object. All of the objects produced are then turned into a final tree.

const collection = [
  {
    id: "a1",
    type: "resources",
    attributes: {
      version: "v1",
      namespace: "accounts",
    }
  },
  {
    id: "a2",
    type: "resources",
    attributes: {
      version: "v1",
      namespace: "accounts",
    }
  },
  {
    id: "b1",
    type: "resources",
    attributes: {
      version: "v1",
      namespace: "profiles",
    }
  },
  {
    id: "b1",
    type: "resources",
    attributes: {
      version: "v2",
      namespace: "profiles",
    }
  }
]

The order goes from outer layer to deeper layer, so in this case the outer level properties will be based on key("type"), and the deepest layer properties will be based on key("id").

const functions = [
  groupBy(key("type")),
  groupBy(keyChain(["attributes", "namespace"])),
  groupBy(keyChain(["attributes", "version"])),
  indexBy(key("id")),
]

treeify(functions)(collection)

The resulting object looks like this:

{
  resources: {
    accounts: {
      v1: {
        a1: {
          id: "a1",
          type: "resources",
          attributes: {
            version: "v1",
            namespace: "accounts",
          }
        },
        a2: {
          id: "a2",
          type: "resources",
          attributes: {
            version: "v1",
            namespace: "accounts",
          }
        }
      }
    },
    profiles: {
      v1: {
        b1: {
          id: "b1",
          type: "resources",
          attributes: {
            version: "v1",
            namespace: "profiles",
          }
        }
      },
      v2: {
        b1: {
          id: "b1",
          type: "resources",
          attributes: {
            version: "v2",
            namespace: "profiles",
          }
        }
      }
    }
  }
}

@unction/type

any -> string

Returns the type name of the value provided.

type("a") // "String"
type(1) // "Number"
type({}) // "Object"
type([]) // "Array"
type(true) // "Boolean"
type(null) // "null"
type(undefined) // "undefined"

@unction/upTo

number -> Array

Just takes a maximum and produces an array of 1 to that number.

@unction/withoutKeyRecursive

KeyType -> IterableType -> IterableType

Returns a copy of an iterable without a key, no matter how deep the tree.

withoutKeyRecursive("__abstraction__")(
  {
    id: "1",
    name: "Kurtis Rainbolt-Greene",
    attributes: {
      version: "v1",
      namespace: "accounts",
      __abstraction__: {errors: []},
    },
    __abstraction__: {errors: []},
  }
)

Which will return:

{
  id: "1",
  name: "Kurtis Rainbolt-Greene",
  attributes: {
    version: "v1",
    namespace: "accounts",
  },
}

@unction/zip

IterableType -> IterableType -> IterableType

Takes two iterables and merges them together, combining their values into an array

zip([1, 2, 3])([4, 5, 6])

returns

[[1, 4], [2, 5], [3, 6]]

zip({x: 1, y: 2, z: 0})({x: 0, y: 0, z: 0})

returns

{x: [1, 0], y: [2, 0], z: [0, 0]}