A simplified way to generate massive mock data based on a schema, using the awesome fake/random data generators like (FakerJs, ChanceJs, CasualJs and RandExpJs), all in one tool to generate your fake data for testing.
Now the library has been migrated 100% to typescript typing are included.
You can test online here: https://danibram.github.io/mocker-data-generator/
Install the module with:
npm install mocker-data-generator
Import it
var mocker = require('mocker-data-generator').default (vainilla way)
or
import mocker from 'mocker-data-generator' (ES6 or Typescript way)Then use it:
var user = {
firstName: {
faker: 'name.firstName'
},
lastName: {
faker: 'name.lastName'
},
country: {
faker: 'address.country'
},
createdAt: {
faker: 'date.past'
},
username: {
function: function() {
return (
this.object.lastName.substring(0, 5) +
this.object.firstName.substring(0, 3) +
Math.floor(Math.random() * 10)
)
}
}
}
var group = {
description: {
faker: 'lorem.paragraph'
},
users: [
{
function: function() {
return this.faker.random.arrayElement(this.db.user).username
},
length: 10,
fixedLength: false
}
]
}
var conditionalField = {
type: {
values: ['HOUSE', 'CAR', 'MOTORBIKE']
},
'object.type=="HOUSE",location': {
faker: 'address.city'
},
'object.type=="CAR"||object.type=="MOTORBIKE",speed': {
faker: 'random.number'
}
}
// Using traditional callback Style
mocker()
.schema('user', user, 2)
.schema('group', group, 2)
.schema('conditionalField', conditionalField, 2)
.build(function(error, data) {
if (error) {
throw error
}
console.log(util.inspect(data, { depth: 10 }))
// This returns an object
// {
// user:[array of users],
// group: [array of groups],
// conditionalField: [array of conditionalFields]
// }
})
// Using promises
mocker()
.schema('user', user, 2)
.schema('group', group, 2)
.schema('conditionalField', conditionalField, 2)
.build()
.then(
data => {
console.log(util.inspect(data, { depth: 10 }))
// This returns an object
// {
// user:[array of users],
// group: [array of groups],
// conditionalField: [array of conditionalFields]
// }
},
err => console.error(err)
)
// Synchronously
// This returns an object
// {
// user:[array of users],
// group: [array of groups],
// conditionalField: [array of conditionalFields]
// }
var data = mocker()
.schema('user', user, 2)
.schema('group', group, 2)
.schema('conditionalField', conditionalField, 2)
.buildSync()
console.log(util.inspect(data, { depth: 10 }))Data generation goes with model based composed by generators, the generators can have access to the data generated and to the entity generated. Generators run synchronously, take care of the related entities!!
-
schema(name, schema, generationType): Add a new schema, you must specify this params:
- name (String): Name of the schema.
- schema (JSON): The schema you define
- generationType (integer or JSON): In this field you specify how you will generate this schema. 3 ways:
- Integer to specify how many objects of this schema you want.
- JSON with this object
{max: '<maximunValues>'}you can also optionally pass min{max: '<maximunValues>', min: '<minimumValues>', this will generate a range of objects of this schema, between (0 and max) or (min and max) randomly. - JSON with this object
{uniqueField: '<yourUniqueField>'}this means that this field () is an array and you want to generate entities with this unique values
-
reset(): Clean the internal DB.
-
restart(): Clean the internal DB and all the schemas inside.
-
build(callback): This methods start to produce the data and wrap it to the callback function, the callback funtion have 2 parameters, error and data generated.
- buildSync(): Synchronous version of
build(callback). Returns generated data or throws an error.
- Normal string key: indicates the key.
- Commaseparated string key: indicates that there is a conditional, before the comma you must specify a conditional (you have all level fields generated in this moment), then you must specify the field if the conditional is true see the example.
-
static: For fixed fields
{ static: 'hello im fixed field' }
-
self: get himself object, and evaluate the string, so you can get calculated fields.
- eval (Optional): Also now you can pass, eval to true, to use like like in versions < 2.6.0
{ self: 'id' } //will get the id of the generated entity { self: 'id', eval: true } // will get the first user id
-
db: get the db, and evaluate the string, so you can access to this entities.
- eval (Optional): Also now you can pass, fast to true, eval to true, to use like like in versions < 2.6.0
{ db: 'user[0].id' } // will get the first user id { db: 'user[0].id', eval: true } // will get the first user id
-
eval: evaluate the current string, remember that i inject all the awesome methods, faker, chance, casual, randexp, and also the db and object methods. With this eval field, you must pass an exactly JSON syntax:
{ eval: 'object.id' } // OR { eval: 'db.user[0]' } // OR { eval: 'faker.lorem.words()' }
-
hasOne: You can pass 2 parameters:
-
hasOne: the name of the related entity, get one random.
-
get (Optional): String that will be evaluated over the random related entity.
-
eval (Optional): Only affects if get is passed, the get param only support dotted paths, with eval=true you can use an eval string, this impacts on the performance
{ hasOne: 'user' } // this populate the field with one random user // OR: { hasOne: 'user', get: 'id' } // this populate the field with one id of a random user // OR: { hasOne: 'user', get: 'id', eval: true } // this populate the field with one id of a random user with eval string
-
-
hasMany: You can pass 4 parameters:
-
hasMany: the name of the related entity, get one random.
-
amount (Optional): Fixed number of related entities to get.
-
min (Optional): Minimum entities to get, buy default is 1, if you want the chance to have empty arrays please specify min to 0.
-
max (Optional): Maximum entities to get.
-
get (Optional): String that will be evaluated over the random related entity.
-
eval (Optional): Get will only support dotted paths, with eval= true you can get from an evaluable string
-
unique (Optional): hasMany will get unique values from the entity (Make sure that you have many unique data in the source)
// In this case we will get 1 user (hasMany) { hasMany: 'user' } // this populate the field with one random user // OR: { hasMany: 'user', amount: 1, //optional } // In this case we will get 1 (amount) user (hasMany) // OR: { hasMany: 'user', max: 3 // optional } // In this case we will get as max 3 (max) users (hasMany) // OR: { hasMany: 'user', min: 1 //optional max: 3 //optional } // In this case we will get bettween min 1 (min) and max 3 (max) users (hasMany) // OR: { hasMany: 'user', get: 'id' } // In this case we will get the id (get) from 1 random user (hasMany)
-
-
incrementalId: For incremental numeric ids, pass the start number to increment. If you put incrementalId = true it takes from 0 the ids.
{ incrementalId: 0 }
-
funcion: No params are passed, only context (
this), in this you have{db, object, faker, chance}, and you can use faker or chance functions, object (the specified model), db (actual data generated){ function: function(){ // this.db // this.object // this.faker // this.chance // this.casual return yourValue } } //OR: { function(){ // this.db // this.object // this.faker // this.chance // this.casual return yourValue } }
-
faker: you can use directly faker functions like: (note that, db (actual entities generated), object (actual entity generated) are injected), you must pass an exactly JSON syntax, now also the multilang is supported by the property locale (Thanks @sleicht for the inspiration. By default I take English locale. This are the locales supported: https://github.com/marak/Faker.js/#localization).
- eval (Optional): You can use like in versions < 2.6.0, su with this true, it will turn faker field string into an evaluable string, also loosing speed
{ faker: 'lorem.words' } // Run faker.lorem.words() { faker: 'lorem.words()' } // Run faker.lorem.words() { faker: 'lorem.words(1)' } // Run faker.lorem.words(1) { faker: 'integer({"min": 1, "max": 10})' } // Run faker.lorem.words(1) and take the first { faker: 'random.arrayElement(db.users)' } // Run faker.arrayElement over a generated user entity { faker: 'random.arrayElement(db.users)["userId"]' } // Run faker.arrayElement over a generated user entity and take the userId only { faker: 'address.streetAddress', locale: 'zh_CN' } // Got 711 蔡 街 { faker: 'address.streetAddress' } // Got 5036 Daniel Village { faker: 'address.streetAddress', eval: true } // Got 5036 Daniel Village
-
chance: you can use directly chance functions, you can do: (note that, db (actual entities generated), object (actual entity generated) are injected), you must pass an exactly JSON syntax:
- eval (Optional): You can use like in versions < 2.6.0, su with this true, it will turn chance field string into an evaluable string, also loosing speed
{ chance: 'integer' } // Run chance.integer() { chance: 'integer()' } // Run chance.integer() { chance: 'integer({"min": 1, "max": 10})' } // Run chance.integer({"min": 1, "max": 10}) { chance: 'street_suffixes()[0]["name"]' } // Run chance.street_suffixes() takes first result and the name inside { chance: 'street_suffixes()[0]["name"]', eval: true } // Run chance.street_suffixes() takes first result and the name inside
-
casual: you can use directly use casualJs functions, you can do: (note that, db (actual entities generated), object (actual entity generated) are injected), you must pass an exactly JSON syntax:
- eval (Optional): You can use like in versions < 2.6.0, su with this true, it will turn casual field string into an evaluable string, also loosing speed
{ casual: 'country' } { chance: 'array_of_digits()' } { casual: 'array_of_digits(3)[0]', eval: true } { casual: 'array_of_digits(3)[0]', eval: true }
-
randexp: pass a regexp string to use randexp generator.
{ randexp: /hello+ (world|to you)/ }
-
[Array]: you can pass an array that indicates an array of data you can create, passing in the first field the generator (function, faker, or array(not Tested)), and in the second field pass a config object (length, fixedLentgh)
-
length: to know how many values
-
fixedLength (Optional): true to create always same amount of values in the array, false to generate a random number between 0 and 'length' value. False by default.
-
concat (Optional): An stringuified array ex: '[object.id, db.users.id]'. This should be an evaluable string to concat with the array that are generating. Also takes in mind that if you have a fixedLength, should not increase the length.
-
strictConcat (Optional): true to remove duplicates in the concatenated string array, when it is calculated. False by default.
[{ // Any generator // Faker faker: 'random.arrayElement(db.users).userId' // Chance chance: 'integer' // Function that has included index, length and self that refers at the actual array generation function: function (index, length, self){ return /**/ } // Array config length: 10, fixedLength: true // Concat concat: '[db.users[0].userId, db.users[1].userId]' strictConcat: true }]
-
- [virtual]: Boolean, if you pass this option, this mean that this field will not appear at the output entity. But you can use during the generation.
{
// Any generator
// Faker
faker: 'random.arrayElement(db.users)[userId]'
// Chance
chance: 'integer'
// Static
static: 'any static field'
// Function
function: function (){ return /**/ }
// With the virtual option
virtual: true
}Initialize mocker with the config, and then generate any entity with promises style, use generate function that accepts the name of the model and the amount of data to generate. Like the example:
mocker()
.schema('user', user, 2)
.schema('group', group, 2)
.schema('conditionalField', conditionalField, 2)
.build(function(err, data) {
console.log(util.inspect(data, { depth: 10 }))
// This returns an object
// {
// user:[array of users],
// group: [array of groups],
// conditionalField: [array of conditionalFields]
// }
})You can also pass instead of the number, an object with the a config, from now {uniqueField}. If this field exists tells to the generator that instead of init a fixed length of data, generate an amount of data depending of the values of the field you will specify. You have 2 way to deal with this, check the examples See the output of this example:
//
// First way, using an 'values' embedded object
//
var cat = {
name: {
values: ['txuri', 'pitxi', 'kitty']
}
}
var m = mocker()
.schema('cat', cat, 10)
.schema('cat2', cat, { uniqueField: 'name' })
.build(function(err, data) {
console.log(util.inspect(data, { depth: 10 }))
})
//
// Second way, without 'values' embedded.
//
var cat = {
name: ['txuri', 'pitxi', 'kitty']
}
var m = mocker()
.schema('cat', cat, 10)
.schema('cat2', cat, { uniqueField: 'name' })
.build(function(err, data) {
console.log(util.inspect(data, { depth: 10 }))
})In version >= 2.6.0, eval option was introduced to run mocker-data-generator like olders versions, so by default is running without eval: faker, chance, casual, hasMany, hasOne, db and self. This means that this methods loose habilities, when eval is not passed, but this are the speed results with eval active (old way) and without (new way)
faker eval old: 0.969ms
faker now: 0.215ms
chance eval old: 0.559ms
chance now: 0.099ms
casual eval old: 0.360ms
casual now: 0.026ms
You can visit the repo url here: https://github.com/danibram/mocker-api-tester
Or visit the api directly: https://mocker-api.herokuapp.com/
Run npm install;npm run dev to watch the project, webpack compile the code automatically. Run npm build to build the normal and minified version.
json-schema-faker is awesome and works really nice, but i need a simplified and fast way to generate mock data for my projects, so i created this.
I couldn't do this without this awesome libraries, so thanks to all:
- Faker: [https://github.com/Marak/faker.js]
- Chance: [https://github.com/victorquinn/chancejs]
- Casual: [https://github.com/boo1ean/casual]
- RandExpJs: [https://github.com/fent/randexp.js]
- typescript-starter: [https://github.com/bitjson/typescript-starter]
Licensed under the MIT license. 2017