fictional

fictional

Generate fake data deterministically from a given input

import { word } from 'fictional'

word('id-1')
// => 'Raệkoshi'

word('id-2')
// => 'Tąme'

word('id-1')
// => 'Raệkoshi'

import { words, shape } from 'fictional'

const user = shape({
  name: words.options({ capitalize: 'all' })
})

user('id-1')
// => { name: 'Nɨ Nohakemá Yuꝁaina' }

user('id-2')
// => { name: 'Kaᶄevamu Hacḣimo' }

user('id-1')
// => { name: 'Nɨ Nohakemá Yuꝁaina' }

• Why
• Overview
  • Makers
  • Composition
  • Options
  • Currying
  • Security
• API Reference
  • Primaries
    • int()
    • bool()
    • float()
    • dateString()
    • char()
    • word()
    • words()
    • sentence()
    • paragraph()
  • Composition
    • join()
    • oneOf()
    • someOf()
    • times()
    • tuple()
    • shape()
    • oneOfWeighted()
• Install & Use

# Why? How is this different to faker?

Libraries like faker or chance allow you to seed a psuedo-random number generator (PRNG), such that the same sequence of values will be generated every time. If that is all you need, those libraries are for you.

Sometimes though, instead of needing to obtain the same sequence of generated values every time, from some set of identifiers, you need to obtain the same mapping to generated values every time. This is where fictional comes in.

For example, when generating api response data in tests, you need to rely on each field always mapping to the same generated value every test run. To illustrate, you could use fictional to generate some user entity in a test:

import { word, shape, int } from 'fictional'

const name = shape({
  first: word,
  last: word
})

const user = shape({
  id: int,
  name
})

user('id-1')
// => { id: 4641209466322491, name: { first: 'Raeniƴo', last: 'Nḁchi' } }

To some extent, there are ways of achieving similar results with libraries like faker, but we haven't found ways that do not have practical limitations:

• It is possible to simply seed the PRNG for every identifier, and then use it to generate only a single value. This seems to be a misuse of these libraries though: there is an up-front cost to seeding these PRNGs that can be expensive if done for each and every value to be generated. Here are benchmarks that point to this up-front cost.
• You can generate a sequence of N values, hash identifiers to some integer smaller than N, then simply use that as an index to lookup a value in the sequence. This can even be done lazily. Still, you're now limiting the uniqueness of the values to N. The larger N is, the larger the cost of keeping these sequences in memory, or the more computationally expensive it is if you do not hold onto the sequences in memory. The smaller N is, the less unique your generated values are.

# API Overview

# Makers

Fictional provides functions that take in some identifying value as input, and generate a corresponding output value. These functions are called makers.

import { word } from 'fictional'

// `word` is a maker
word('id-1')
// => 'Raệkoshi'

The given input can be any JSON-serializable value. For any two calls to the same maker function, provided the input given in each call serializes down to the same value, the same output will be returned. Makers work statelessly, so for the same input, the same value will be returned regardless of the enviornment, process, call ordering, or any other external factors.

Note that unlike JSON.stringify(), object property ordering is not considered.

import { word } from 'fictional'

word({
  a: 21,
  b: 23
})
// => 'Kī̀nchihani'

word({
  b: 23,
  a: 21
})
// => 'Kī̀nchihani'

# Composition

const streetAddress = join(' ', [
  int.options({
    min: 1,
    max: 200
  }),
  word,
  oneOf(['Drive', 'Street', 'Avenue'])
])

streetAddress('id-1')
// => '82 Y̊oke Drive'

streetAddress('id-2')
// => '132 Ẏukaikeki Street'

Some makers take in identifying value as the only required argument and return. These kinds of makers are described in the docs as primary makers. word() is an example of such a maker.

However, sometimes the data you need generated requires a combination of different makers. Fictional provides functions for doing this: they take in an identifying value and makers as arguments, and compose these makers in some way to produce a corresponding output. These kinds of makers are described in the docs as composition makers. join() (shown above) is an example of a such a maker.

In the example above, a maker returning fictitious street addresses is formed by using join() to compose int, word(), and other composing maker, oneOf.

Under the hood, composition makers re-hash the identifying value each time a maker is given as input is used. This ensures that a unique value is generated for each maker provided, while still keeping the result deterministic. In the example below, the tuple() maker ensures that each word in the returned array has a different value.

tuple('id-1', [word, word])
// => [ 'Ṱayu', 'Y̊oke' ]

// this is roughly the same as doing
word(hash('id-1')), word(hash(hash('id-1')))

# Options

Many makers accept an options object as an argument for configuring how the generated output looks:

int('id-1')
// => 656963231996220

int('id-1', {
  min: 1,
  max: 99
})
// => 7

As a convenience, it is also possible to extend these makers to use specific options by using the .options() api:

const newInt = int.options({
  min: 1,
  max: 99
})

newInt('id-1')
// => 7

newInt('id-2')
// => 57

.options() returns a new function that will call the original maker function with the given arguments. It is still possible to provide options when calling the returned function. In this case, these options will override any options given to .options():

const newInt = int.options({
  min: 1,
  max: 99
})

newInt('id-1', { max: 3 })
// => 1

.options() can also be called on the returned function, to further extend the maker:

const newInt = int.options({ min: 1 }).options({ max: 99 })

newInt('id-1')
// => 7

newInt('id-2')
// => 57

# Currying

Composition makers take in more than one required argument. If the identifying input value is not given as an argument (one less than the required arguments is provided), then a new function will be returned. This function will take an identifying input value as its only argument, and call the original composition maker with both this argument and the other required arguments initially given. This limited form of currying can be convienent for composing makers:

const companyName = join(' ', [word, oneOf(['Incorporated', 'Systems'])])

companyName('id-1')
// => 'Ṱayu Systems'

companyName('id-2')
// => 'Moʋa Systems'

# Security

Under the hood, Fictional uses a keyed hash function called SipHash in order to map input values to output values deterministically.

Out of the box, fictional uses a hardcoded key. If it is important that no information about the input values can be inferred, then you'll need to generate and use your own key instead:

// Store this key somewhere safe
const key = hash.generateKey('aSY3k#uf^dHlj12@')

hash.setKey(key)

# API Reference

# Primaries

# int(input[, options])

Takes in an identifying input value and returns an integer.

int('id-23')
// => 2211849950287729

options

• min=1 and max=Number.MAX_SAFE_INTEGER: the minimum and maximum possible values for returned numbers

int('id-2', {
  min: 2,
  max: 99
})
// => 15

# bool(id)

Takes in an identifying input value and returns a boolean.

bool('id-23')
// => true

# float(id[, options])

Takes in an identifying input value and returns a number value with both a whole and decimal segment.

float('id-23')
// => 4139094247650.859

options

• min=1 and max=Number.MAX_SAFE_INTEGER: the minimum and maximum possible values for returned numbers

float('id-2', {
  min: 2,
  max: 99
})
// => 2.54

# dateString(id[, options])

Takes in an identifying input value and returns a string representing a date in ISO 8601 format.

dateString('id-23')
// => '1989-02-18T02:01:32.000Z'

options

• minYear=1980 and maxYear=2019: the minimum and maximum possible year values for returned dates

dateString('id-2', {
  minYear: 1980,
  maxYear: 2089
})
// => '2003-06-14T18:06:24.000Z'

# char(input)

Takes in an identifying input value and returns a string with a single character.

char('id-23')
// => 'B'

The generated character will be an alphanumeric: lower and upper case ASCII letters and digits 0 to 9. Alternative character ranges are listed below. To choose your own range of characters, see char.inRanges().

Predefined character ranges

char.ascii('id-2')
// => '_'

char.digit('id-3')
// => '0'

Fictional ships with makers for a predefined set of character ranges. Similar to char(), these makers take in only an identifying input value as an argument and return a string with a single character. The following ranges are available:

• char.ascii: Any ASCII character

• char.digit: Characters for numbers 0 to 9

• char.alphanumeric (alias: char): lower and upper case ASCII letters and digits 0 to 9

• char.letter (alias: char.asciiLetter): Lower and upper case ASCII letters

• char.lower (alias: asciiLower): Lower case ASCII letters

• char.upper (alias char.asciiUpper): Upper case ASCII letters

• char.unicode: Any character from the ASCII and Latin-1 Supplement unicode blocks

• char.unicodeLetter: Lower and upper case letters from the ASCII and Latin-1 Supplement unicode blocks

• char.unicodeLower: Lower case letters from the ASCII and Latin-1 Supplement unicode blocks

• char.unicodeUpper: Upper case letters from the ASCII and Latin-1 Supplement unicode blocks

• char.latin1: Any character from the Latin-1 Supplement unicode block

• char.latin1Letter: Lower and upper case Latin-1 Supplement letters

• char.latin1Lower: Lower case Latin-1 Supplement letters

• char.latin1Upper: Upper case Latin-1 Supplement letters

# char.inRanges(ranges)

Takes in an array of [min, max] pairs, where min and max are integers specifying the minimum and maximum possible Unicode code point values for a desired range of characters, and returns a maker function that will return characters in those given ranges.

const symbols = char.inRanges([
  // misc symbols
  [0x2600, 0x26ff],

  // emoticons
  [0x1f600, 0x1f64f]
])

symbols('id-1')
// => '⚜'

char.inRanges is designed to allow characters in the ranges given to all have a similar likelihood of being returned.

To allow for composition, each item in the array of ranges can also be a pre-defined character range, or another character range defined using char.inRanges():

const misc = char.inRanges([[0x2600, 0x26ff]])
const emoticons = char.inRanges([[0x1f600, 0x1f64f]])
const letterOrSymbol = char.inRanges([misc, emoticons, char.letter])

letterOrSymbol('id-2')
// => '😑'

# word(id[, options])

Takes in an identifying input value and returns a string value resembling a fictitious word.

word('id-23')
// => 'Muvẩ'

options

• capitalize=true: whether or not the word should start with an upper case letter
• unicode=true: whether or not the string should contain non-ascii unicode characters. If true is given, each returned word will always contain a single unicode character. If false is given, each returned word will never contain non-ascii characters. If a value between 0 and 1 is given, that value will represent the probability of a returned value containing a single unicode character.
• minSyllables=2 and maxSyllables=4: the minimum and maximum possible number of syllables that returned words will contain

word('id-2', {
  minSyllables: 1,
  maxSyllables: 6,
  unicode: 0.382
})
// => 'Ța'

# words(id[, options])

Takes in an identifying input value and returns a string value resembling fictitious words.

words('id-23')
// => 'Kiṇmihy tǟhyme meȟy'

options

• min=2 and max=3: the minimum and maximum possible number of words that returned strings will contain.
• capitalize='first': whether or not the words should start with upper case letters. If true or 'all' is given, each string returned will start with an upper case letter in each word. If 'first' is given, for each string returned, only the first word will start with an upper case letter. If false is given, each string returned will always contain only lower case letters.
• unicode=true: whether or not the string should contain non-ascii unicode characters. If true is given, each returned word will always contain a single unicode character. If false is given, each returned word will never contain non-ascii characters. If a value between 0 and 1 is given, that value will represent the probability of a returned value containing a single unicode character.
• minSyllables=1 and maxSyllables=4: the minimum and maximum possible number of syllables that returned words will contain

words('id-2', {
  min: 5,
  max: 8,
  unicode: 0.618,
  capitalize: 'all'
})
// => 'Nṓvishikai Șoma Ke Hyvicea Mï Mȅki'

# sentence(id[, options])

Takes in an identifying input value and returns a string value resembling a sentence of fictitious words.

sentence('id-23')
// => 'Ha vakina kachi hǎ nirae sohyme, kamukin vamusomi kinỳomuyo some nokai.'

options

• minClauses=1 and maxClauses=2: the minimum and maximum possible number of clauses that a returned sentence will contain.
• minWords=5 and maxWords=8: the minimum and maximum possible number of words that each clause will contain.
• unicode=0.382: whether or not the string should contain non-ascii unicode characters. If true is given, each returned word will always contain a single unicode character. If false is given, each returned word will never contain non-ascii characters. If a value between 0 and 1 is given, that value will represent the probability of a returned value containing a single unicode character.
• minSyllables=1 and maxSyllables=4: the minimum and maximum possible number of syllables that returned words will contain

sentence('id-2', {
  minClauses: 2,
  maxClauses: 3,
  minWords: 2,
  maxWords: 3,
  unicode: 0.9
})
// => 'Ranọmu ceameyu, m̃ashia manohasộ.'

# paragraph(id[, options])

Takes in an identifying input value and returns a string value resembling a paragraph of fictitious words.

paragraph('id-23')
// => 'Maketaki ceaki mø kiashi kesokaimu raemaḉhi ta, rayuta ha namimeki hǟkiyu raeshichi raemiraechi. Vaketayo vį̃ nimu hykaika hyki kihahykaī̀ vakeyoke hysomi. Sonaȟyha chiko kāso socea kime cḣihyma. Kinva kinmemami mïko no shimemoa. Ṃu shicħi chi nöhakoyu ṁichi nimome nichikaį̃chi yuahyki, racea hasomu ʋi mehakino mi mắceamu soka shihykami. Ḳo makemira nỉceamo ni kechiyu kinmiraechi ka vamo. Cea chimo hakesohý yo yohy kaishimo yovamu kinmenino.'

options

• minSentences=3 and minSentences=7: the minimum and maximum possible number of sentences that a returned paragraph will contain.
• minClauses=1 and maxClauses=2: the minimum and maximum possible number of clauses that each sentence will contain.
• minWords=5 and maxWords=8: the minimum and maximum possible number of words that each clause will contain.
• unicode=0.382: whether or not the string should contain non-ascii unicode characters. If true is given, each returned word will always contain a single unicode character. If false is given, each returned word will never contain non-ascii characters. If a value between 0 and 1 is given, that value will represent the probability of a returned value containing a single unicode character.
• minSyllables=1 and maxSyllables=4: the minimum and maximum possible number of syllables that returned words will contain

paragraph('id-2', {
  minSentences: 2,
  minSentences: 3,
  unicode: 0.9
})
// => 'Kḛkin kinkeviṁu rahysḣi kî ᶄesora. Ŷomu ceasokịrae noỹu misoƙishi chį̃, ṛasomeva raeṿi mḛ rahỹvame shikaiḱeko kinḱeyu. Kiyomi nö hamōchira kinceasoha raćhima minokeshi. Mume hasōkin niʋakinke kehakiȧ ṁo kiäyuni čhiahy ceaḳe, tãhymimu mukiņ hǻ vacħi yǘmonahy vimümo.'

# Composition

# join(input, joiner, values)

Takes in an identifying input value and an array of makers as values, calls each with a unique identifying input, and joins the results with the given joiner.

join('id-23', ' ', [word, oneOf(['Street', 'Drive'])])
// => 'Naceakõmu Street'

If an item in the value array is not a function, that value will be used as-is:

join('id-2', ' ', [word, 'Drive'])
// => 'Moʋa Drive'

joiner can also be a function, in which case it will be called with the results of resolving each item in values as input:

join('id-3', ([a, b, c]) => `${a}-${b} ${c}`, [word, word, word])
// => 'Namī̀nocea-Viḱaikemu Rẫhyshi'

If any of the items in values resolves to a nested array, that array will be flattened (regardless of nesting depth):

join('id-2', '', [char.letter, times(3, char.alphanumeric)])
// => 'rKuL'

# oneOf(input, values)

Takes in an identifying input value and an array of values, and returns an item in values that corresponds to that input:

oneOf('id-23', ['red', 'green', 'blue'])
// => 'red'

If an item in values is a maker, that maker will be called and the result will be returned:

oneOf('id-2', [int, word, char])
// => 'Mốta'

# someOf(input, range, values)

Takes in an identifying input value and an array of values, repeatedly picks items from that array a number of times within the given range. Each item will be picked no more than once.

someOf('id-23', [1, 2], ['red', 'green', 'blue'])
// => []

As shown above, range can be a tuple array of the minimum and maximum possible number of items that can be picked.

It can also be given as a number, in which case exactly that number of items will be picked:

someOf('id-2', 2, ['red', 'green', 'blue'])
// => [ 'blue', 'red' ]

If an item in values is a maker, that maker will be called and the result will be returned:

someOf('id-3', [1, 2], [int, word, char])
// => []

# times(input, range, maker)

Takes in an identifying input value and a maker, calls that maker repeatedly (each time with a unique input) for a number of times within the given range, and returns the results as an array:

times('id-23', [4, 5], word)
// => [ 'Kaiḣaso', 'Niávi', 'Nȫta', 'Muḿemu' ]

As shown above, range can be a tuple array of the minimum and maximum possible number of times the maker should be called. It can also be given as a number, in which case the given maker will be called exactly that number of times:

times('id-2', 2, word)
// => [ 'Chį̃meta', 'Shiȟavi' ]

# tuple(input, values)

Takes in an identifying input value and an array of makers as values, calls each with a unique identifying input, and returns the array of results.

tuple('id-23', [char, char])
// => [ '7', '3' ]

If an item in the value array is not a function, that value will be used as-is:

tuple('id-2', [char, '!'])
// => [ 'f', '!' ]

# shape(input, properties)

Takes in an identifying input value and an object of makers as properties, calls each property's value with a unique identifying input, and returns results as an object.

shape('id-23', {
  firstName: word,
  lastName: word
})
// => { firstName: 'Yomuńiyu', lastName: 'Ÿumemo' }

If an item in the properties object is not a function, that value will be used as-is:

shape('id-23', {
  name: join(' ', [word, word]),
  active: true
})
// => { name: 'Soraēhyrae Muṃeni', active: true }

# oneOfWeighted(id, values)

Takes in an identifying input value and a value array of consisting of [probability, value] pairs, and returns one of one of the values in that array. The likelihood of a particular value being returned will correspond to the probability given for it, where probability is a number between 0 and 1.

oneOfWeighted('id-23', [
  [0.9, 'red'],
  [0.05, 'green'],
  [0.05, 'blue']
])
// => 'red'

If an item in values is a maker, that maker will be called and the result will be returned:

oneOfWeighted('id-2', [
  [0.9, word],
  [0.05, char],
  [0.05, int]
])
// => 'Vikoķaita'

For each [probability, value] pair in the array of values, if the given probability is not a number, that probability will be considered unassigned. All items with unassigned probabilities will receive an equal share of the remaining probability after accounting for all items with assigned probabilities (all items for which a number value was given for their probability). In the example below, 'green' and 'blue' will both have a probability of 0.4 of being returned ((1 - 0.2) / 2).

oneOfWeighted('id-23', [
  [0.2, 'red'],
  [null, 'green'],
  [null, 'blue']
])
// => 'green'

# Install & Use

You can use fictional as the npm package fictional:

npm i -D fictional  # chances are you want it as a devDependency
# or
yarn add -D fictional

Fictional can be used in both es-module-aware and commonjs bundlers/environments.

// es module
import { word } from 'fictional'

// or alternatively
import word from 'fictional/word'

// commonjs
const { word } = require('fictional')

// or alternatively
const word = require('fictional/word')

It can also be used a <script>:

<script
  crossorigin
  src="https://unpkg.com/fictional/dist/umd/fictional.js"
></script>

<script>
  fictional.word('some-identifier')
</script>