---

mad-utils

Collection of utilities I keep repeatedly rewriting across projects.

---

Examples - most useful methods

(see full docs in lower sections for more details)

cap1LowerRest :: (string) => string

Capitalize first letter, convert rest to lowercase.

cap1LowerRest('aSdF'); // => 'Asdf'

capitalize :: (string) => string

Capitalize first letter. Leave the rest as-is.

capitalize('asdf'); // => 'Asdf'
capitalize('asDf'); // => 'AsDf'

eliminateWhitespace :: (string) => string

eliminateWhitespace('    asdf 123    ff    '); // => 'asdf123ff'

switchExpr :: ((cond: any, valueToReturnIfCondTruthy: V)*, defaultValue?: W) => V | W | never;

Function-based switch statement. Takes 2 or more args.

• Args 1, 3, 5, 7, etc. are the conditions, and 2, 4, 6, etc. their corresponding return values.
  • On hitting a truthy odd-numbered arg, switchExpr returns the next (even-numbered) arg, then exits.
  • If none are truthy, the default value is returned (the last argument - which must be an odd-numbered arg).
    • If no default value is present after all conditions returned false, throws an error.

Examples:

switchExpr(true, 'val1');                // Output: 'val1'

switchExpr(false, 'val1', 'defaultVal'); // Output: 'defaultVal'

switchExpr(
    false, 'v1',
    true, 'v2',
    'defaultReturnVal');                 // Output: 'v2'

switchExpr(undefined, 'v1', '', 'v2');   // Throws Error

first :: (Array<T|any>) => T|any;

first([1,2,3]); => 1

last :: (Array<T|any>) => T|any;

last([1,2,3]); => 3

matchAny :: (any[]) => (any) => boolean

Search array for value. Returns true if array contains value. Uses simple JSON.stringify for comparison.

matchAny([1, 2, 3])(2);
// => true

matchAny(['a', 6, [1, 2, 3], 'gr'])([1, 2, 3]);
// => true

uuid :: () => string

uuid(); // => 5A42CCCF-6B10-488B-B957-4D1E5A113DA4
uuid(); // => 38259F99-73D5-4EE1-B11F-5D33CE8AD2C6

get :: (Object, string, any?) => any

Safely get the item at the given object path.

const obj = {a: {b: {c: 'value'}}};
get(obj, 'a.b.c');                    // => 'value'
get(obj, 'a.b.zzz', 'default value'); // => 'default value'

isBoolean :: (any) => boolean

isBoolean(true); // => true
isBoolean(false); // => true
isBoolean({}); // => false

isFalse :: (any) => boolean

isFalse(true); // => false
isFalse(false); // => true
isFalse('ok'); // => false

isInt :: (any) => boolean

isInt(5); // => true
isInt(-10); // => true
isInt(1.6); // => false
isInt('okok'); // => false

pushIfNew :: (T[], T) => T[]

const arr = [1, 2, 3]

pushIfNew(3); // => [1, 2, 3]
pushIfNew(4); // => [1, 2, 3, 4]

console.log(arr); // => [1, 2, 3, 4]

repeatChars :: (string, number) => string

repeatChars('a', 5); // => 'aaaaa'
repeatChars('aa', 5); // => 'aaaaaaaaaa'

getLangFromURLPathname ::

Signature:

<T = ('en' | 'fr')>(
    string? = location.pathname,
    T[]?    = ['en', 'fr']
    T?      = 'en'
) => string

TODO confirm getLangFromURLPathname signature

Example usage:

// With URL http://example.com/auth/fr/ok:
getLangFromURLPathname(); // => 'fr'

// With URL http://example.com/en/auth/ok:
getLangFromURLPathname(); // => 'en'

// With URL given as param:
getLangFromURLPathname(`http://example.com/auth/sp/ok`, [`en`, `fr`, `sp`]); // => 'sp'

parseQueryParams :: (queryParamsString?: string = window.location.search) => Object

// With URL http://example.com/home?hello=everyone&gr=argh:
parseQueryParams(); // => { hello: 'everyone', gr: 'argh' }

---

Installation

• npm:

npm install --save mad-utils

• yarn:

yarn add mad-utils

---

• NOTE: the documentation is an extreme work-in-progress.
• Recent versions have considerably changed the design, API, and even structure from earlier ones.
  • Considerably more functions are available
  • Existing functions have been massively changed (mostly to be more robust & less surprising);
  • The library has been split into 3 parts:
    • node
    • browser
    • isomorphic/shared
      • consumed by default, and also used by both the node and browser submodules.
  • A few exports have been deprecated (such as the parseDate function and ParsedDate type)
    • Mostly due to irrelevance (items were taken from my own projects).
  • (The docs still remain mostly up to date)

---

Sub-modules

• Broken into 3 sub-modules: node, browser, and isomorphic.
• The node and browser sub-modules each include the entirety of the isomorphic sub-module.

Isomorphic

Importing isomorphic (default) submodule

// Import all namespaces, functions, types, etc. from isomorphic submodule
import {m_} from 'mad-utils';

// Import isomorphic namespaces
import {array, json, enum, number, object, query, string, types} from 'mad-utils';

// Import individual isomorphic functions, types, classes, etc.
import {first, isNumberLike, parseQueryParams, castToNum, stringToEnumVal} from 'mad-utils';

• All modules exported from mad-utils provide everything in the isomorphic module.

Isomorphic namespaces

• array
• date
• enum
• error
• func / functionUtils
• json
• locale
• number
• object
• query
• search
• string
• types
• validation

All of the above namespaces are also importable from the NodeJS and Browser modules.

NodeJS submodule

• Isomorphic exports, plus exports that will only work in a Node environment, such as:

  • Anything using the NodeJS core API.
  • Anything requiring file handling.
  • Anything based on DOM-free unit testing.
  • Anything intended for use with (or relying on) on a browser-unfriendly library:
    • e.g. Express, Mocha, Chai.

• Will generally crash your application if imported into the browser.

Importing node sub-module

// Import all namespaces, functions, types, etc. from node & isomorphic submodules
import {m_} from 'mad-utils/lib/node';

// Import node (and isomorphic) namespaces
import {file, test, middleware, webpackUtils, nodeError, date} from 'mad-utils/lib/node';

// Import individual node (and isomorphic) functions, types, classes, etc.
import {
    isDir,
    wasRunAsScript,
    replaceInFile,
    getJsFilesInDir,
    globalActivateCleanStack,
    useMiddlewareInProductionOnly,
    third,
    thirdLast,
    splitLines,
    composeExpressMiddlewares,
    isNonMinFile,
    eliminateWhitespace
} from 'mad-utils/lib/node';

Node-specific namespaces

• file
• middleware
• nodeError
• test
• webpackUtils
• types (expanded to include both isomorphic and Node-specific types)

Browser submodule

• Exports that will only work in a browser environment, or one with a mocked DOM (e.g. JSDom)
• Generally causes errors to throw if used in Node without a special environment set up
  • e.g. JSDom, or inclusion of various window mocks/polyfills

Importing browser submodule

// Import all namespaces, functions, types, etc. from browser submodule
import {m_} from 'mad-utils/lib/browser';

// Import namespaces from browser (and isomorphic) submodules
import {dom, types} from 'mad-utils/lib/node';

// Import individual browser (and isomorphic) functions, types, classes, etc.
import {
    browserVersion,
    browserEngineVersion,
    getUserAgentString,
    assignFrozenClone
} from 'mad-utils/lib/node';

Browser namespaces

• dom
• types (expanded to include both isomorphic and Browser-specific types)

---

Functions, by namespace

More import notes

If using a high-level import (mUtils, m_, __), you can access functions either via their namespaces or directory. E.g.

mUtils.search.replaceAll
mUtils.replaceAll
__.number.isInt
__.isInt
m_.date.isLeapYear
m_.isLeapYear
m_.array.secondLast
m_.secondLast
...etc...

mUtils, __, and m_ are 'full collection' exports

You can also get them like this if you hate named imports:

import madUtils from 'mad-utils';
const h = madUtils.m_;

Namespace strategy

Inclusive, overlapping namespace strategy used.

Namespaces treated more like keywords than parent types.

• Many functions are included in more than 1 namespace.

The main philosophy behind this API design is to make common functions maximally available.

• Repeatedly checking sifthing through namespaces trying to remember where a function lives is annoying.
• However, having 100s of functions together in a giant namespace with no other form of organization available is also annoying.
• I opted for a compromise, where everything was included in a giant namespace, while also including smaller "sub-namespaces".
  • This also has import advantages, since you can opt to pull in as much or as little as you need on each reference to mad-utils, without having to import whole namespaces and pluck individual functions off.

Common types [WIP]

NumLike

Either a number, or a string that can be parsed to a number

StrOrNever

Either a string, or 'Never' (indicating an error threw in the function)

---

---

---

Namespace contents

---

---

Namespace: array (isomorphic)

Get items from array by position

first :: (T[]) => T

Return first item in given array

first(['a', 'b', 'c', 'd']);  // => 'a'

second :: (T[]) => T

Return second item in given array

second(['a', 'b', 'c', 'd']);  // => 'b'

third :: (T[]) => T

Return third item in given array

third(['a', 'b', 'c', 'd']);  // => 'c'

last :: (T[]) => T

Return last item in given array

last(['a', 'b', 'c', 'd']);  // => 'd'

secondLast :: (T[]) => T

Return secondLast item in given array

secondLast(['a', 'b', 'c', 'd']);  // => 'c'

thirdLast :: (T[]) => T

Return thirdLast item in given array

thirdLast(['a', 'b', 'c', 'd']);  // => 'b'

first2 :: (T[]) => T[]

Return first 2 items in given array

first2(['a', 'b', 'c', 'd']);  // => ['a', 'b']

first3 :: (T[]) => T[]

Return first 3 items in given array

first3(['a', 'b', 'c', 'd']);  // => ['a', 'b', 'c']

last2 :: (T[]) => T[]

Return last 2 items in given array

last2(['a', 'b', 'c', 'd']);  // => ['c', 'd']

last3 :: (T[]) => T[]

Return last 3 items in given array

last3(['a', 'b', 'c', 'd']);  // => ['b', 'c', 'd']

firstN :: (T[], number) => T[]

Return first 'n' number of items from given array

firstN(['a', 'b', 'c', 'd'], 2);  // => ['a', 'b']

lastN :: (T[], Int) => T[]

Return last 'n' items from given array. Return full array if too many items requested.

lastN(['a', 'b', 'c', 'd'], 2);  // => ['c', 'd']

Create array

arrayN :: (length: Int) => undefined[]

Create empty array of given length (integer).

arrayN(3);  // => [undefined, undefined, undefined]

splitLines :: (string, {preserveEmptyLines: false}) => string[]

Split large multiline string into array where each line is an item. Removes blank lines by default, unless preserveEmptyLines option is set to true.

splitLines(
    'first line' +
    '\n ' +
    'third line' +
    '\n',
    'fourth line'
);
// => ['first line', ' ', 'third line', 'fourth line']

splitLines(`
    first line

    second line`,
    {preserveEmptyLines: true}
);
// => ['', 'first line', '', 'second line']

Exclude items from array by position

withoutFirst :: (T[] | string) => T[] | string

Remove first character or array item.

withoutFirst([1, 2, 3, 4, 5]) // => [2, 3, 4, 5]
withoutFirst('hello');  // => 'ello'

withoutFirst2 :: (T[] | str) => T[] | str

Remove first 2 characters or array items.

withoutFirst2([1, 2, 3, 4, 5]); // => [3, 4, 5]
withoutFirst2('abcdef'); // => 'cdef'

withoutFirst3 :: (T[] | string) => T[] | string

Remove first 3 characters or array items.

withoutFirst3([1, 2, 3, 4, 5]); // => [4, 5]

withoutLast :: (T[] | string) => T[] | string

Remove last character or array item.

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

withoutLast2 :: (T[] | string) => T[] | string

Remove last 2 characters or array items.

withoutLast2([1, 2, 3, 4, 5]); // => [1, 2, 3]

withoutLast3 :: (T[] | string) => T[] | string

Remove last 3 characters or array items.

withoutLast3([1, 2, 3, 4, 5]); // => [1, 2]

withoutFirstN :: (T[]|str, number) => T[] | string

Remove first N characters or array items.

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

withoutLastN :: (T[] | string, number) => T[] | string

Remove last N characters or array items.

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

Array typechecking

isArray :: (T[] | T) => boolean

True if item is an array

isArray([]); // => true

class CustomArray extends Array { }

isArray(new CustomArray()); // => true

Add to or subtract from array

removeMatches :: (any[], any[] | any) => any[]

NON-MUTATIVE. PERFORMANCE-INTENSIVE.

Return new array with all items in arr2OrItem removed from array1. If array2 is not an array, remove matching item from array1.

removeMatches([1, 2, 3], 2); // => [1, 3]

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

rmAllFalsy

(arr: any[]) => arr[]

• Return new array with all falsy values in the given array eliminated.
• NON-MUTATIVE

Examples:

rmAllFalsy([1, 2, 3, false, null, 4, 0, 'asdf', '', 'ok', undefined, 'fine']);
// => [1, 2, 3, 4, 'asdf', 'ok']

Array searching

matchAny: (haystack: T[]) => (needle: T) => boolean

Search an array for a value.

• Returns true if array haystack contains needle.

Note that it uses simple JSON.stringify for array and object comparison

• use something else if deep comparisons are required.

Sane behaviour for matching against null, undefined, NaN, etc.

• e.g. NaN matched against an array with NaN returns true

Curried.

Examples:

matchAny([1, 2, 3])(2);
// => true

matchAny(['a', 6, [1, 2, 3], 'gr'])([1, 2, 3]);
// => true

matchAny(['a', 6, null, 'last'])(null);
// => true

---

Namespace: date (isomorphic)

[TYPE] NumRange0To6

Shorthand for any number between 0 and 6

[CONSTANT] defaultTimestampFormat :: string

String that creates a timestamp in a nice, human-readable format when passed to MomentJS. YYYY/MM/DD : hh:mm:ss

Examples:

console.log(defaultTimestampFormat);
// => `YYYY/MM/DD : HH:mm:ss`;

isLeapYear :: (year: NumLike) => boolean

Returns true if given year is a leap year. Accepts integers, strings that can be converted to integers, and arrays with a single item, where said item is an integer or string convertable to an integer. Any other input will throw.

Examples:

isLeapYear(2004); // => true
isLeapYear(2003); // => false

convertDayOfWeekNumToString :: (day: 0..6, abbrev: boolean) => string | never

Converts numeric day of the week to string day of the week. e.g. 0 -> 'Sunday', 6 -> 'Saturday' Args:

• day: number from 0 to 6 for conversion
• abbrev: If true, return the shorthand day names (e.g. 'Mon' vs. 'Monday'). Default: false.

Examples:

convertDayOfWeekNumToString(5); // => 'Friday'
convertDayOfWeekNumToString(2, true); // => 'Tues'

now

(timeFormat?: string) => string

• Get the current date, formatted for display in the stream of Express logs to the CLI.
• Args:
  • timeFormat?: Optional momentJS timestamp format e.g. MM/DD::hh:mm:ss
    • More info at https://momentjs.com/docs/#/parsing/string-format/

Examples:

now(); // => 2017/05/28 : 02:51:39
now(`YYYY/MM hh:mm`); // => 2017/02 02:51

isDateLike (exported from types-iso - see below)

---

Namespace: dom (browser)

getUserAgentString :: () => string

Return raw and unparsed browser user agent string (convenience function)

Example:

getUserAgentString();
// => "Mozilla/4.0 (Macintosh; Intel Mac OS X 7_12_6) AppleWebKit/501.16 (KHTML, like Gecko) Chrome/50.0.1010.99 Safari/210.22"

osName :: () => string

Extract name of current user's OS (operating system) from browser user agent string. (Note: Memoizes result - i.e. 1st call to function stores result; all future calls reference stored result).

Example:

osName(); // => "Mac OS"

osNameSnakeCase :: () => string

Extract name of OS from browser user agent string, & convert it to snake_case. (Note: memoizes result)

Example:

osNameSnakeCase(); // => "mac_os"

browserName :: () => string

Extract name of current browser from browser user agent string. (Note: memoizes result)

Example:

browserName(); // => "Firefox"

browserEngineName :: () => string

Extract name of current browser's rendering engine from browser user agent string. (Note: memoizes result)

Example:

browserEngineName(); // => "Webkit"

osVersion

Extract version of current OS from browser user agent string. (Note: memoizes result)

Example:

osVersion(); // => "15.9.1"

browserVersion :: () => string

Extract version of current browser from browser user agent string. (Note: memoizes result)

Example:

browserVersion(); // => "64.1.5284.259"

browserEngineVersion :: () => string

Extract version of current browser's rendering engine from browser's user agent string. (Note: memoizes result)

Example:

browserEngineVersion(); // => "530.12"

---

Namespace: enum (isomorphic)

enumToStringArray

• WIP documentation

enumValToString

• WIP documentation

stringToEnumVal

• WIP documentation

isNumericEnumItem

• WIP documentation

isIndexEnumItem

• WIP documentation

isDataEnumItem

• WIP documentation

---

Namespace: error (isomorphic)

DecoratorError

• WIP documentation

scrubStackTrace

• WIP documentation

---

Namespace: error (node)

globalActivateCleanStack :: () => void

Remove pointless stacktrace items (node core). Modify the stacktrace length to be unlimited. Effects get applied globally immediately on running the function.

• Affects error handling behaviour for the entire life of the Node process this was run in.

Examples:

globalActivateCleanStack();
// /\-- This is literally the only way to use it.

---

Namespace: file (node)

isDir :: (fileOrDirPath: string) => boolean

Returns true if inode (aka file, directory, socket, etc.) at absolute path is a directory.

isDir(path.join(rootPath, 'node_modules')); // => isTrue

wasRunAsScript :: () => boolean

Must always be called like this: wasRunAsScript(path.basename(__filename)). WARNING: has some edge cases (Fixing them is a WIP TODO):

• Will (incorrectly) return true if the current file has the same name as the file that launched the process.
  • e.g. if process was launched by [project root]/server/index.js, and wasRunAsScript is run in [project root]/server/api/index.js, it will return true.
• Will not work for filenames with certain characters, such as (, ), [, ], and certain other special regex chars (. and - are OK).

Example (in some-file.js, with process launched via node some-file.js):

wasRunAsScript(path.basename(__filename)); // => true

pathFromRoot

WIP documentation

replaceInFile

WIP documentation

getJsFilesInDir

WIP documentation

isFileInDir

WIP documentation

isNonMinFile

WIP documentation

endsInDotJs :: (string) => boolean

True if the given string (generally a path) ends in .js.

Example:

endsInDotJs(`asdf.js`); // => true

getBaseFilenameFromPath

WIP documentation

---

Namespace: function (isomorphic)

switchExpr :: (...(cond: any, retValIfCondTru: V), def?: W) => V | W | never;

Function-based switch statement.

For each pair of args:

• the 1st arg is a condition that passes if truthy.
• the 2nd arg is the value returned if the condition passes.

If no conditions pass, and there was:

• ...an odd number of arguments given, then the final arg given to the function is returned.
• ...an even number of arguments given, an error is thrown.

Examples:

switchExpr(true, 'val1');                                // => 'val1'
switchExpr(true, 'val1', 'defaultVal');                  // => 'val1'
switchExpr(false, 'val1', 'defaultVal');                 // => 'defaultVal'
switchExpr(false, 'v1', 'condition1', 'v2');             // => 'v2'
switchExpr(false, 'v1', null, 'v2', 'defaultReturnVal'); // => 'v2'
switchExpr(false, 'v1', null, 'v2');                     // => [throws error]
switchExpr(false, 'v1');                                 // => [throws error]

let size = 'large';

switchExpr(
    size === 'tiny',   8,
    size === 'small',  10,
    size === 'medium', 12,
    size === 'large',  16,
    size === 'huge',   20,
                       12
);
// => 16

General syntax:

switchExpr(
    COND1, val_returned_if_COND1_truthy,
    COND2, val_returned_if_COND2_truthy,
    ...,
    defaultReturnVal
)

loopN :: (number, (...args) => T) => T[]

Run given function the given number of times.

Return results of all runs of the function as an array containing all N return vals.

Examples:

loopN(2, () => 'return_value'); // => ['return_value', 'return_value']

let i = 0;
loopN(10, () => i++); // => [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
console.log(i)        // => 10

loop2 :: ((...args) => T) => T[]

Run given function twice, and return results of both runs of the function as an array.

Example:

loop2(() => 'return_value'); // => ['return_value', 'return_value']

loop3, loop4, loop5

See loop2 above, but run the associated number of times

• e.g. loop4 runs 4 the function 4X instead of twice

Examples:

loop3(() => 'ret_val'); // => ['ret_val', 'ret_val', 'ret_val']
loop4(() => 'ret_val'); // => ['ret_val', 'ret_val', 'ret_val', 'ret_val']
loop5(() => 'ret_val'); // => ['ret_val', 'ret_val', 'ret_val', 'ret_val', 'ret_val']

let i = 0;
loop5(() => i++); // => [0, 1, 2, 3, 4]
console.log(i)    // => 5

getFnAsArr :: (fn: Function) => string[]

Return a function's source code in nicely spaced array format.

Examples:

getFnAsArr(() => 'ok')
// => [ 'function () { return \'ok\'; }' ]

function testFn() {
    console.log('log 1');
    return 'output';
}
getFnAsArr(testFn);
// => [ 'function testFn() {',
//      '    console.log(\'log 1\');',
//      '    return \'output\';',
//      '}']

---

Namespace: json (isomorphic)

jsonStringifyWFuncs :: (Object) => string

Stringify, while keeping the functions in position by pre-converting them to strings.

Example:

jsonStringifyWFuncs({a: 123, b: 'asdf', c: () => 'asdf'})
// =>
    '{"a":123,"b":"asdf","c":"function () { return \'asdf\'; }"}'

WIP documentation

---

Namespace: locale (isomorphic)

commonLangsObj :: Record<string, string>

• Object containing a set of common languages and their common ID codes
• e.g. {af: 'Afrikaans', en: 'English', ...}

commonLangAbbrevs :: string[]

• Array of common abbreviations for the most common languages
• e.g. ['af', 'en', ...]

commonLangNames :: string[]

• Array of the names of the most common languages
• e.g. ['Afrikaans', 'English', ...]

canadaLangsObj :: Record<string, string>

• Object mapping Canada's official languages to their abbreviations
• {en: English, fr: French}

canadaLangAbbrevs :: string[]

• Array of the abbreviations of Canada's official languages
• ['en', 'fr']

canadaLangNames :: string[]

• Array of the names of Canada's official languages
• ['English', 'French']

englishVariants

• Array of variants of English, by locale (codes)
• ['english', 'en', 'en_ca', 'en_us', ...]

frenchVariants :: string[]

• Array of variants of French, by locale (codes)
• ['french', 'fr', 'fr_fr', 'fr_ca', ...]

---

Namespace: middleware (node)

useMiddlewareInProdOnly

WIP documentation

composeExpressMiddlewares

WIP documentation

---

Namespace: number (isomorphic)

isInteger :: (any) => boolean

Returns true if given value is an integer.

Examples:

isInteger(1);      // => true
isInteger(232);    // => true
isInteger(82.12);  // => false
isInteger('232');  // => false
isInteger('asdf'); // => false

(Alias: isInt)

isIntegerLike :: (any) => boolean

Returns true if given value is an integer-like string or integer.

Examples:

isIntegerLike(232);    // => true
isIntegerLike('232');  // => true
isIntegerLike('asdf'); // => false
isInteger(82.12);      // => false

isNumberLike :: (any) => boolean

Returns true if given value is a number-like string or number.

Examples:

isNumberLike(1);         // => true
isNumberLike(9267.84);   // => true
isNumberLike('1');       // => true
isNumberLike('9267.84'); // => true
isNumberLike('1.2');     // => true
isNumberLike('.2');      // => true

isNumberLike('1.2.2');   // => false
isNumberLike('asdf');    // => false
isNumberLike(true);      // => false

(Alias: isNumLike)

uuid () => string

Generate a UUID, in format e.g. 3A0BF2C7-3077-45A0-97ED-EC5F13F127F1.

Examples:

uuid(); // => 'F6779B17-8CD1-409B-A2AA-1FE80CB86654'
uuid(); // => 'B574571F-097A-4ADB-837C-DCE8472C3314'

---

Namespace: object (isomorphic)

get :: (obj: Object, path: string, default: any?) => any

Safely get the item at the given object path.

Arguments:

• 1st: object to get from.
• 2nd: object path to get value from.
• 3rd: default value (if no value found at given path).

Examples:

const obj = {a: {b: {c: 'value'}}};

get(obj, 'a.b.c'); // => 'value'

get(obj, 'a.b.zzz', 'default value'); // => 'default value'

assignFrozenClone :: (Object, ...Object[]) => Readonly