Neuron Powered Motorization

    mongo-unit
    TypeScript icon, indicating that this package has built-in type declarations

    3.2.0 • Public • Published

    mongo-unit

    This library is done to simplify creation of integration tests for node.js application with Mongo DB. I starts local mongodb process using mongodb-memory-server library, and it work in "InMemory" mode, which improve performance of your tests.

    How to use it

    Since running embedded mongo takes some time, we need to config mocha to wait until mongo is up and running, the easiest way to do it add --delay flag for mocha and handle async flow, like this

    //package.json
    ...
    "scripts": {
            ...
            "test": "mocha ./**/*.spec.js --delay",
            ...
        },
    ...
    

    add init.spec.js

    //init.spec.js
    const mongoUnit = require('mongo-unit')
    
    mongoUnit.start().then(() => {
      console.log('fake mongo is started: ', mongoUnit.getUrl())
      process.env.DATABASE_URL = mongoUnit.getUrl() // this var process.env.DATABASE_URL = will keep link to fake mongo
      run() // this line start mocha tests
    })
    
    after(() => {
      const dao = require('./dao')
      console.log('stop')
      dao.close()
      return mongoUnit.stop()
    })

    Code under test

    //dao.js
    const mongoose = require('mongoose')
    
    const userSchema = new mongoose.Schema({
      name: String,
    })
    const taskSchema = new mongoose.Schema({
      userId: String,
      task: String,
    })
    
    module.exports = {
      init: () => mongoose.connect(process.env.DATABASE_URL),
      close: () => mongoose.disconnect(),
      User: mongoose.model('user', userSchema),
      Task: mongoose.model('task', taskSchema),
    }

    Test data

    {
      "users": [
        {
          "_id": "56d9bf92f9be48771d6fe5b1",
          "name": "test"
        },
        {
          "_id": "56d9bf92f9be48771d6fe5b2",
          "name": "John"
        }
      ],
      "tasks": [
        {
          "userId": "56d9bf92f9be48771d6fe5b1",
          "task": "do stuff"
        },
        {
          "userId": "56d9bf92f9be48771d6fe5b1",
          "task": "fix stuff"
        }
      ]
    }

    Mocha test

    //dao.spec.js
    const {expect} = require('chai')
    const mongoUnit = require('mongo-unit')
    const testData = require('./data.json')
    const daoUT = require('./dao')
    describe('dao', ()=>{
      
        before(() => daoUT.init())
        beforeEach(() => mongoUnit.load(testData))
        afterEach(() => mongoUnit.drop())
      
        it('should find all users', () => {
          return daoUT.User.find()
            .then(users => {
              expect(users.length).to.equal(2)
              expect(users[0].name).to.equal('test')
            })
        })
      
        it('should find all tasks for user 1', () => {
          return daoUT.User.find()
            .then(users => users[0])
            .then(user=>daoUT.Task.find({userId: user._id}))
            .then(tasks => {
              expect(tasks.length).to.equal(2)
              expect(tasks[0].task).to.equal('do stuff')
            })
        })
      })

    General info

    CircleCI

    I was inspired by dbUnit library, which is very popular in java world.

    There is alternative library for mocking Mongo: mockgoose

    Requirements

    It works on Node.js 12+

    Installation

    npm install -D mongo-unit

    Run test in docker example

    docker build -t mongo-unit . docker run -it mongo-unit

    API

    start(opts)

    It starts mongod on one of available port and returns Promise with URL to connect to this db. opts is optional params, you can specify some command line params for mongod:

    • opts.port - preferable mongo db port, default: 27017
    • opts.dbName - name of test db, default: test
    • opts.dbpath - db path, default: <node_modules/mongo-unit>\.mongo-unit
    • opts.verbose - enable debug logs, default: false
    • opts.version - specify which version of mongo to download. e.g. version=v3.6
    • opts.useReplicateSet - specify we want to start the memory server as a MongoMemoryReplSet needed to support transactions, default: false

    mongodb-memory-server can also be configured via environment variables. For instance, export MONGOMS_DEBUG=1will activate debug logging in the memory server. See their docs for more information.

    stop()

    It stops mongod process

    getUrl()

    Syncronius API returns URL to connect to test db, if test DB is not started it thows an Exception

    load(data)

    Inserts given data (like below) DB collections, returns Promise.

    {
      "collectionName1":[
        {"field1":"value1"},
        {"field2":"value2"}
      ],
      "collectionName2":[
        {"field3":"value3"},
        {"field4":"value4"}
      ]
    }

    clean(data)

    Clear collections based on given data (data format is the same), returns Promise.

    drop()

    Drops test DB, returns Promise.

    initDb(data)

    helper function, load db data into mongo

    dropDb()

    helper function, clear all db data from mongo

    Install

    npm i mongo-unit

    DownloadsWeekly Downloads

    3,172

    Version

    3.2.0

    License

    MIT

    Unpacked Size

    315 MB

    Total Files

    54

    Last publish

    Collaborators

    • mikhail.angelov