fixtures-generator

Fixtures Generator

This project was created to solve the most common problems with fixtures that we faced while building tests at VoxFeed.

Objetives

• Let each test to have its own fixtures, to prevent sharing fixtures across the tests suites.
• Specify a schema for each fixture. To make sure everyone still valid.
• Customize fixtures as required by each test.
• Make easier to find tests with invalid fixtures. Just update the schema an find all broken fixtures.

Install

Use npm or yarn:

npm install -D fixtures-generator

yarn add -D fixtures-generator

Usage

1. Declare the schemas you need
2. Import the module and build a generator
3. Use the generator to create fixtures

1. Declare the schemas you need

• Schemas are declared using JSON Schema specification.
• Schemas need to be exported as a plain object, where:
  • Object keys are the names of the schemas, this is how you will tell the generator which schema to use.
  • Object values are the schema objects.
• It is recommended to keep each schema in a separated file and join them in an index.

Example of schemas object (./schemas.js):

const authorSchema = {
  type: 'object',
  properties: {
    id: { type: 'integer' },
    name: { type: 'string' }
  }
};

const bookSchema = {
  type: 'object',
  properties: {
    id: { type: 'integer' },
    authorId: { type: 'integer' },
    title: { type: 'string' }
  }
};

const schemas = {
  author: authorSchema,
  book: bookSchema
};

module.exports = schemas;

2. Import the module and build a generator

After schemas are declared you should build a generator instance. This instance is the one you can use across your tests to generate fixtures.

Example (./generate-fixtures.js)

const FixtureGenerator = require('fixtures-generator');
const schemas = require('./schemas');

const options = {
  optionalsProbability: 0.5
};

const generateFixtures = new FixtureGenerator(schemas, options);

module.exports = generateFixtures;

Available options are:

Option	Type	Description
optionalsProbability	Float	Number between 0 and 1 that specifies the percentage of non-required propoerties that will be included in randomly generated fixtures.

3. Use the generator to create fixtures

Finally, import the fixtures generator instance and create some fixtures.

Example:

const generateFixtures = require('./generate-fixtures');

const params = {
  type: 'author',
  quantity: 3
};

const authors = generateFixtures(params); // Will generate 3 author fixtures.

Required parameters for generateFixtures method are:

Parameter	Type	Description
type	String	The name of the fixture schema to be used to generate and validate fixture.
recipe	[Object]	The seed of the fixtures to be generated on the same order. Resulting fixtures will keep all properties sent here.
quantity	Number	Quantity of fixtures to be generated. If quantity is higher than the size of recipe the exceeding fixtures will be randomly generated; otherwise parameter is omitted.

It is possible to create random fixtures by just specifying the required quantity:

const fixtures = generateFixtures('invite', 3);

// Result:
[
  {
    _id: ObjectId('592f0dabfcf681e3d5ceeb6c'),
    otherProperties...
  },
  {
    _id: ObjectId('592f0dabfcf681e3d7defb9a'),
    otherProperties...
  },
  {
    _id: ObjectId('592f0dabfcf681e3d14cab00'),
    otherProperties...
  }
];

Also, it is possible to specify an objects array that will be part of generated fixtures.

const recipe = [
  {
    campaign: ObjectId('ffffffffffffffffffffffff'),
    username: 'manuel'
  },
  {
    _id: ObjectId('a1a1a1a1a1a1a1a1a1a1a1a1')
  }
];

const fixtures = generateFixtures('invite', recipe);

// Result:
[
  {
    _id: ObjectId('592f0dabfcf681e3d5ceeb69'),
    campaign: ObjectId('ffffffffffffffffffffffff'),
    username: 'manuel',
    otherProperties...
  },
  {
    _id: ObjectId('a1a1a1a1a1a1a1a1a1a1a1a1'),
    campaign: ObjectId('39583dabfcf681e3d5ceeb61'),
    username: 'whatever',
    otherProperties...
  }
];

Schemas validation

When a recipe is used to generate fixtures, all passed properties must be valid; otherwise an error is thrown.

const recipe = [
  {
    oldProperty: 'lalala, I am a property that should not exist',
  }
];

const fixtures = generateFixtures('invite', recipe);

// It will throw an error since "oldProperty" is not a valida property for "invite" type.

⚠️ In order make this module flexible, validation is only at keys level. That means, data types are NOT validated (for example: if property username is declared as string, it won't throw error if an integer is sent in recipe). It is developer's responsibility to send the correct data type.