nodejsamazingenerator

v1.70.60-stable

Published

3 years ago

Downloads

0High
0Medium
0Low

nodejs generator express mocha jwt passport sendgrid sequelize sequelize-cli models dao mvc restful migrations docs generate amazing

NodeJS amazinGenerator

Description: Powerfull project with some features that could be take as basic, this project has some structure based on generator from Diego Haz generator project. Also has some configuration files based on the juniorhq88 generator project. This project helps developers to build FAST a robust environment backend on javascript, using scrum with test based on user stories on the sprint that your project may have. Enjoy :)

Several changes about this both project, were maked by developer team. However many thanks to this both developers by creating both projects as open source code :)

This project has integrated this feautres:

JWT: For authentication standard
Passport: To define method through the app is authenticable
Sendgrid: For mailing from back-end side
ExpressJS: To handling routes fetching and there flow as using middlewares with next() function in order to multi fetch.
Sequelize ORM: To handle DB connection from the backend.
Sequelize CLI: To handle migrations, creation of models, or databases.
Babel (allows ES6): To use latest syntaxis of JS
ApiDoc Generator: from JSDoc conventions generate an HTML page to see it.
JSDoc: Conventions to write docs at JS files.
Mocha To test files inside test folder with extension .spec
Factorygirl For creating factories design pattern generator for test using fakers and random seeding (Auto generated !!).
Faker For generating fakes vals to build test using Factories.
Sequelize Auto: This is just working through hard coded package.json, must be integrated with SDK of this lib, to create models inside the folder of the entity scope (As DAO, RESTful, MVC pattern should be)
Bodymen: As middleware on ExpressJS fetchings to validate body data structure ( Check docs )
Body Parser: To admit urlencoded and JSON type on ExpressJS (Check docs)
CORS: To allow fetching through REST methods.
Morgan: For NodeJS Logging (it can let save on persistance file logs).
Dotenv: To let NodeJS use .env file as common frameworks use to work.

Installation

Crate a directory like myprojectdirectory and then enter it by following code (replace with your project name of course)

mkdir myprojectdirectory 
cd myprojectdirectory

Then init an npm project skipping all options with enter (this is just to get project)

npm init

Then install nodejsamazingenerator through npm like this. A little questionary will be displayed, fill your envs to make project work. If you are not sure yet about all of this vars. You just can keep default values, and when you were sure about those vars, run npm run init:project and questionary will be display again to re fill it. Remember that this questionary actually rewrite your .env but also apidoc.json and package.json files.

npm i nodejsamazingenerator

Then install dependencies through npm like this

npm install

Finally just run to deploy server. (This will follow .env variables)

npm run dev:rollback

Project already started installation

First just clone your project from your repo

git clone my-repo-url

Then install dependencies

npm install

Finally configure your environment vars

npm run init:project

Run minimal features test to create DB tables

npm run test:rollback

And finally test it

npm run test

Usage

Last update let developers build full api entity scope (check folder structure below at myentity folder inside api folder) At now we just support datatypes like STRING, INTEGER, FLOAT, DATE, BOOLEAN, JSON:

npm run model --name myentity --attr nombre:string,apellido:string,correo:string

Note that myentity should be as many models as your project has. Repeat as many times you need. Once you finished, run this command to create your database, and migrate your whole list of models created.

npm run dev:rollback

Generate the whole models documentation by running:

npm run docs

Other features

This project will allow you to build fast a robust backend NodeJS server. following good pattern design as DAO, RESTful and MVC. Commands to remember

Create database from config.js (See docs 4 detail):
```
npm run sequelize -- db:create --env development
```

Drop database from config.js:

npm run sequelize -- db:drop --env production

Create new migration (See docs 4 detail):

npm run sequelize -- migration:generate --name my-migration

Run migrations on database to create tables (See docs 4 detail):
```
npm run sequelize -- db:migrate --env test
```
Generate server Docs following JSdocs conventions:
```
npm run docs
```
Run your test inside test folder with .spec extension on Mocha Docs following JSdocs conventions:
```
npm run test
```

(On dev) Generate models from DB by changing at package.json at scripts last one change vars mydbname, myhost, mydbuser, mydbpassword.

"scripts": {
   ...,
   "generate-models-from-db": "sequelize-auto -o /src/api -d mydbname -h myhost -u mydbuser -p 5432 -x mydbpassword -e postgres"
 }

And then running

npm run model:auto

Start server with reseting all by environment:
for development
```
 npm run dev:rollback
```
for production
```
 npm run prod:rollback
```
for testing
```
 npm run test:rollback
```

Directory structure

Overview

You can customize the src and api directories.

backnodejs/
	├─ docs/
	├─ generator/
	│  ├─ templates/
	│  ├─ __init__.py
	│  ├─ generator.js
	│  ├─ listdirmaker.py
	│  └─ listdirmaker.pyc
	├─ config/
	│  └─ config.json
	├─ migrations/
	│  ├─ 20200130214923-create-user-migration.js
	│  ├─ 20200202225008-create-passwordreset-model.js
	│  └─ timestamp-create-mymodel-migration.js
	├─ seeders/
	│  ├─ 20200207140211-user-seed.js
	│  └─ timestamp-myentity-seed.js
	├─ src/
	│ 	├─ api/
	│ 	│  ├─ auth/
	│ 	│  │  ├─ controller.js
	│ 	│  │  └─  index.js
	│ 	│  ├─ user/
	│ 	│  │  ├─ user.controller.js
	│ 	│  │  ├─ user.dao.js
	│ 	│  │  ├─ user.helpers.js
	│ 	│  │  ├─ user.factory.js
	│ 	│  │  ├─ user.model.js
	│ 	│  │  └─ index.js
	│ 	│  ├─ passwordreset/
	│ 	│  │  ├─ passwordreset.controller.js
	│ 	│  │  ├─ passwordreset.dao.js
	│ 	│  │  ├─ passwordreset.helpers.js
	│ 	│  │  ├─ passwordreset.factory.js
	│ 	│  │  ├─ passwordreset.model.js
	│ 	│  │  └─ index.js
	│ 	│  ├─ myentity/
	│ 	│  │  ├─ myentity.controller.js
	│ 	│  │  ├─ myentity.dao.js
	│ 	│  │  ├─ myentity.helpers.js
	│ 	│  │  ├─ myentity.factory.js
	│ 	│  │  ├─ myentity.model.js
	│ 	│  │  └─ index.js
	│ 	│  └─ index.js
	│ 	├─ services/
	│ 	│  ├─ express/
	│ 	│  ├─ jwt/
	│ 	│  ├─ mongoose/
	│ 	│  ├─ passport/
	│ 	│  ├─ sendgrid/
	│ 	│  ├─ factorygirl/
	│ 	│  └─ sequelize/
	│ 	├─ utils/
	│ 	│  ├─ console_colors.js
	│ 	│  └─ error_handler.js
	│ 	├─ config.js
	│ 	├─ index.js
	│ 	└─ app.js
	├─ test/
	│  └─  sprint.spec.js
	├─ .babelrc
	├─ .editorconfig
	├─ .env
	├─ .env.example
	├─ .eslintrc
	├─ .gitignore
	├─ apidoc.json
	├─ DOCS.md
	├─ LICENSE
	├─ package.json
	└─ README.md

src/api/index.js: Here is where the API endpoints are added to express router. Each API has its own folder.
src/api/myentity/myentity.model.js: It defines the data attributes of the database model. Also declares a Queryman object to be used at validations on controllers. Finally defines the model with Sequelize SDK
src/api/myentity/myentity.controller.js: This is the API controller file. It defines the main router middlewares which use the API model.
src/api/myentity/myentity.helpers.js: This file saves functions usefull for data access queries as getting an average from numbers or something like that :)
src/api/myentity/myentity.factory.js: This file is as it name say, a factory of fake data to build in real time, by default the amount of fake data is 100. You can edit it by going to this file. Also can be use a test. However Factory is a good pattern design used by most of frameworks. So we decide to integrate it on the generated files.
src/api/myentity/myentity.dao.js: A multiple Data Object Model use must be here, example clearly is relationship queries.
src/api/myentity/index.js: This is the entry file of the API. It defines the routes using, along other middlewares (like session, validation etc.).
services/: Here you can put helpers, libraries and other types of modules which you want to use in your APIs.

Documentation

EXPRESSJS: This service has a configuration service at src/services/express/index.js file, which basically sets some need configuration at express instance. Mainly config at this file (to avoid writing whole file) are:

export  default (apiRoot, routes) => {
	const  app = express()	  

	/* istanbul ignore next */
	if (env === 'production' || env === 'development') {
		app.use(cors())
		app.use(compression())
		app.use(morgan('dev'))
	}  

	app.use(bodyParser.urlencoded({ extended:  false }))
	app.use(bodyParser.json())
	app.use(apiRoot, routes)
	app.use(queryErrorHandler())
	app.use(bodyErrorHandler())

	return  app

}

JWT: This service has a configuration service at src/services/jwt/index.js file, which basically sets sign method to be used at passport service. Mainly configuration of this file (to avoid writing whole file) is to sign with JWT.


     

/**
* With promisify we can change the original function return by a Promise response
*/

const  jwtSign = Promise.promisify(jwt.sign)

     

/**

* This method allows to sign in through jsonwebtoken lib
* @param  {*}  id id from the fetching user
* @param  {*}  options options to sign in
* @param  {*}  method method to sign in at platform, usually jwt secret method

*/

export  const  sign = (id, options, method = jwtSign) => {
	console.log('Signing in with JWT_SECRET = ', config.jwtSecret)
	return  method({ id }, config.jwtSecret, options)
}

PASSPORT: This service has a configuration service at src/services/passport/index.js file, this service is kindy more complex than others, Passport just let us define by use() function, ways of authenticate through passport API. As example, just to avoid write whole file (its kindy long), lets just take the token use, that works with the service of JWT(last one). To use one of this config, just use authenticate() method from passport, on one of your middlewares configuration at express routes :) as follow:

use('token')

/**
 * Passport configuration for using JWT
 */
passport.use('token', new JwtStrategy({
  secretOrKey: config.jwtSecret, /** Mainly by giving to it the JWT_SECRET from .env (check config.js) */
  /**
   * Then by extracting from request the 
   * access_token given after logging API  
   * (It could be at URL, at Body from request, or Header from request)
   */
  jwtFromRequest: function (req) {
    var token = null
    if (req) {
      req.rawHeaders.map((header, index) => {
        /**
         * NOTE: This generator sets JWT token at header (Proxy-Authorization Bearer token) format 
         */
        if (header == 'proxy-authorization') { token = (req.rawHeaders[(index + 1)]).split('Bearer ').join('') }
      })
    }
    return token
  }
  /**
   * At last, get the user and is added to request as req.user with
   * done function, that takes the profile info and attaches it on the request object so its available on your callback url as req.user.
   * ( Read  https://hackernoon.com/passportjs-the-confusing-parts-explained-edca874ebead )
   */
}, async ({ id }, done) => {
  const userLogged = await models.user.findOne({ where: { id: id } })
  if (userLogged) {
    done(null, userLogged)
    return null
  } else throw 'User id dont match at Databse registers'
}))

authenticate('token')

export  const  token = ({ required, roles = User.roles } = {}) => (req, res, next) =>
	passport.authenticate('token', { session:  false }, (err, user, info) => {
	if (err || (required && !user) || (required && !~roles.indexOf(user.role))) {
		return  res.status(401).end()
	}
	req.logIn(user, { session:  false }, (err) => {
		if (err) return  res.status(401).end()
		next()
	})
})(req, res, next)

RESPONSE: This service has a configuration service at src/services/response/index.js file, and it has basically some standard at working with fetchings. One of this function as example (to avoid writing whole file) is that executes double function as success( res )( entity ), where res is res express object of fetching, and entity could be as example an user getted from sequelize service.

/**
* This methods builds a json response based on the original response, giving
* to it a status and also the result of the fetching as an object entity
* @param  {*}  res receives the res from a fetching request
* @param  {*}  status Code with status of the response for a fetching request
* */
export  const  success = (res, status) => (entity) => {
	if (entity) {
		res.status(status || 200).json(entity)
	}
	return  null
}

SENDGRID: This service has a configuration service at src/services/sendgrid/index.js file, and it has basically config to use mailing with sendgrid.

import  sendgridMail  from  '@sendgrid/mail'
import { sendgridKey, defaultEmail } from  '../../config'
     
/**
* With the SENDGRID_KEY from .env we load
*/
sendgridMail.setApiKey(sendgridKey)  

/**
* This method receives the email from is going to be sended the email, the destination email
* Subject of the email indeed, and the content. This can be as HTML string sure
* @param  {*}  param0 Receives an object with basic data to send an email
*/

export  const  sendMail = ({
	fromEmail = defaultEmail,
	toEmail,
	subject,
	content
	}) => {
		const  msg = {
		to:  toEmail,
		from:  fromEmail,
		subject,
		html:  content
		}
		return  sendgridMail.send(msg)
	}
}

An example of using this service is password reset controller as follow code

/**
 * This method creates a new passwordreset by using bodymen and other usefull functions from the response service, check it ;)
 * @param {*} param0 params from the request, getting it with bodymen
 * @param {*} res res object to return to fetcher
 * @param {*} next next function from Express to go through the next middleware on the fetch
 */
import { sendMail } from '../../services/sendgrid'
import models from '../../services/sequelize'
import consoleColors from '../../utils/console_colors'	
import { errorHandler } from '../../utils/error_handler'
import { success, notFound } from '../../services/response'
export const create = async ({ bodymen: { body } }, res, next) => {
  console.log(consoleColors.successConsole, '[SUCCESS] Master and Bodyman middlewares check completed succesfully')
  try {
    const userFetcher = await models.user.findOne({ where: { email: body.email } })

    if (userFetcher) {
      const newPasswordReset = await models.passwordreset.build({ user_id: userFetcher.id, rest_token: uid(32) })
      newPasswordReset.save()

      if (newPasswordReset) {
        const link = `${body.link.replace(/\/$/, '')}/${newPasswordReset.rest_token}`
        const mail = await sendMail({ toEmail: userFetcher.email, subject: ' Password Reset', content: htmlEmailContent(userFetcher.username, link) })
        if (mail) {
          success(res)({ msj: 'Email sent with status ' + mail[0].statusCode })
        } else throw 'Could not send email'
      } else throw 'Could not generate your password reset ticket'
    } else throw 'User does not exist.'
  } catch (err) {
    errorHandler(err)
    notFound(res)({ msj: err })
  }
}

SEQUELIZE: This service has a configuration service at src/services/sequelize/index.js file, and it is based on SequelizeCLI config file generator. To avoid writting whole file it basically works, loading whole models list from project, in Sequelize object cache, which allow to better perfomance at making queries with it.

Instanced class of Sequelize for NodeJS


const  sequelize = new  Sequelize(config.postgresql.uri, {
	logging:  config.postgresql.db_log === 'true',
	dialect:  'postgres',
	host:  config.postgresql.db_host,
	port:  config.postgresql.db_port,
	define: {
		timestamps:  false  /** Must be true */
	}
})

Recursive function to find models.js


const  searchModels = (target, result_array = []) => {
	fs.readdirSync(target, { withFileTypes:  true }).map(element  => {
		if (element.isDirectory()) {
			const  directoryPath = path.join(target, element.name)
			searchModels(directoryPath, result_array)
		} else {
			const  isModel = (element.name.indexOf('.model') !== -1)
			if (isModel) {
			result_array.push(path.join(target, element.name))
			}
		}
	})
}

then using it with a defined path as inside api older

const  target = path.join(__dirname, '../../api')
searchModels(target, models)
models.map((modelPath) => {
	const  model = sequelize.import(modelPath)
	db[model.name] = model
	if ('associate'  in  db[model.name]) {
		db[model.name].associate(db)
	}
})

SEQUELIZE-CLI: This lib can be executed by using npx executable from nodeJS as follow


npx sequelize-cli

But this way does not allow developers to use ES6 features, so by running custom command npm run sequelize -- will execute SequelizeCLI but with babel rules to read ES6 modules features.

Which will be used for commands to make migrations, or create them if needed. Set some seeders and other stuff. As example something prettier basic as create the db without getting into the console from DB API...

npm run sequelize -- db:create --env test

Then to drop the db just do

npm run sequelize -- db:drop --env test

Rembember also to check .env file with your vars as:

PROJECT_NAME=myprojectname
DB_HOST=mydbhost
DB_USER=mydbuser
DB_PASSWORD=mydbpassword
DB_PORT=mydbport
DB_NAME_DEV=mydbdevname
DB_NAME_TEST=mydbtestname
DB_NAME_PROD=mydbprodname
SERVER_IP=myserverip
SERVER_PORT=myserverport
DEFAULT_EMAIL_ALIAS=myserveremailalias
SENDGRID_KEY=mysendgridkey
DB_LOG=mydblog
JWT_SECRET=myjwtsecret
MASTER_KEY=mymasterkey

For creating migrations for example we can use command like this (remember use - instead of _):

npm run sequelize -- migration:generate --name my-migration

This will create your migration with a timestamp in the name just to set a flow that would be follow by CLI API.

So, by using this we can now create a model with our format RESTful design pattern, where the model example is mymodelname you could replace as you want


npm run model --name mymodelname --attr name:string,location:string,email:string

Then the model will be created at the path src/api/mymodelname/mymodelname.model.js and the migration will be stored at migrations folder as timestamp-create-mymodelname-model.js:

'use strict'

import { userAttributes } from '../src/api/user/user.model'

/**
* With userAttributes we can spread whole attr model config
*/
module.exports = (() => {
return {
	up: (sequelize, DataTypes) => {
	return sequelize.createTable('users', {
		...userAttributes(DataTypes)
	})
	},
	down: (queryInterface, Sequelize) => {
	return queryInterface.dropTable('users')
	}
}
})()

Then at the model src/api/mymodelname/mymodelname.model.js a bodyman Schema will be generated to use it as middleware on ExpressJS fetchs, also will be declared the modelAttribute function which should return an Schema from Sequelize standard, and finally the definition of the model from Sequelize SDK as follow example:

/* jshint indent: 2 */

import bcrypt from 'bcrypt'
import { errorHandler } from '../../utils/error_handler'

const roles = [
'superadmin',
'admin',
'user'
]

/**
* Queryman data Schema for validation middleware
*
*/
export const userDataSchema = {
email: {
	type: String,
	required: true
},
password: {
	type: String,
	required: true
},

role: {
	type: String,
	required: true,
	enum: roles
},
username: {
	type: String,
	required: true
}
}

/**
* This method build usefull attr to be declared in many places as it could be necessary (migrations example)
* @param {*} DataTypes From Sequelize instance
*/
export const userAttributes = (DataTypes) => {
return {
	id: {
	type: DataTypes.INTEGER,
	allowNull: false,
	primaryKey: true,
	autoIncrement: true
	},
	email: {
	type: DataTypes.STRING,
	allowNull: false
	},
	password: {
	type: DataTypes.STRING,
	allowNull: false
	},
	role: {
	type: DataTypes.STRING(256),
	allowNull: false
	},
	username: {
	type: DataTypes.STRING,
	allowNull: false
	}
}
}

/**
* Sequelize model SDK for NodeJS with SequelizeCLI standard 
*/
export default (sequelize, DataTypes) => {
return sequelize.define('users', {
	...userAttributes(DataTypes)
}, {
	tableName: 'users',
	hooks: {
	/**
	* hooks methods from models recognize some
	* lifecycle hooks from the user model component
	* to execute some custom 'scripts' on them
	* (Check https://sequelize.org/master/manual/hooks.html)
	*/
	beforeSave: async (user) => {
		try {
		const hashedPass = await bcrypt.hash(user.password, 9)
		user.password = hashedPass
		} catch (err) {
		errorHandler(err)
		}
	}
	}
})
}

Now lets guess that we were working without Sequelize-CLI and did create models from the scratch by ourselfs. So then, lets create a migration for example a already created src/api/user/user.model.js with command:

npm run sequelize migration:generate --name user-migration

Checks file generated as timestamp-create-user-migration.js.js inside of it, is just skeleton of the migration, lets fill it with the copy of user.model.js like this:

'use strict'

import { userAttributes } from '../src/api/user/user.model'

/**
* With userAttributes we can spread whole attr model config
*/
module.exports = (() => {
return {
	up: (sequelize, DataTypes) => {
	return sequelize.createTable('users', {
		...userAttributes(DataTypes)
	})
	},
	down: (queryInterface, Sequelize) => {
	return queryInterface.dropTable('users')
	}
}
})()

Now running this command will migrate all your migrations...

npm run sequelize -- db:migrate --env test

And if you are looking to restart the whole environment, and then start your backend app just run it like this to development

npm run dev:rollback

Bodyman: This lib allows expressJS to filter if data comes as should be on our fetch, declaring as a Schema from Sequelize model, but with other syntaxis :

attribute: {
    type: String,
    match: /^\S+@\S+\.\S+$/,
    required: true,
    unique: true,
    trim: true,
    lowercase: true,
    minlength: 6,
    enum: roles
  }

Following last example, check how a user should be declared with bodyman, for example at user.model.js below of sequelize declaration:

export const userDataSchema = {
  deleted_at: {
    type: Date,
    required: false
  },
  email: {
    type: String,
    required: true
  },
  password: {
    type: String,
    required: true
  },
  username: {
    type: String,
    required: true
  }
}

Then at route config we can validate this by following code, which one configurate as middlewares, master passport method authentication (for authorized apps just fetching API), then bodyman through body which executes the schema validate() method, to check if the definition that we gave is correct, finally just using create() method for create the model indeed:

import { middleware as query } from 'querymen'
import { master } from '../../services/passport'
const { email, password, username } from './user.model.js'
const { create } from './user.controller.js

router.post('/',
  master(),
  body({ email, password, username }),
  create)

Api Docs: This file will be autogenerated. This lib allow us to create a HTML document with the documentation of our API, using JSDoc standard at our code (Please document your code, remember that code is readed by other humans). Check the apidoc.json file, should look like this (Check more config at https://apidocjs.com/#configuration):
```
{
    "name": "myprojectname",
    "version": "0.0.1",
    "description": "mydescription",
    "title": "mydoctittle",
    "url" : "http://server-host:serverport"
}
```
With it, by running npm run docs will generate the doc, and open it at your default web browser.

FactoryGirl with Faker: ON DEV so this is not include on generator yet but there are examples on how to build a factory using the new service at services/factorygirl/index.js, adding a test inside tests folder, and running npm run test:rollback you can import this factory definition with the Sequelize driver loaded and make your tests !! :)

/** Factory lib for JS with adapters for Mongoose ODM or Sequelize ORM */
import factoryGirl from 'factory-girl'
/** Usefull functions from JS */
import bluebird from 'bluebird'

/** with Factory, we bascially get factory instance, then to allow asyn/await features we promosify the instance */
const factory = new factoryGirl.Factory().promisify(bluebird)

/** This case we use Sequelize */
const adapter = new factoryGirl.SequelizeAdapter()

/** Set adapter and export instance as default */
factory.setAdapter(adapter)
export default factory

Then a factory looks like follow:

import factory from '../../services/factorygirl'
import models from '../../services/sequelize'
import faker from 'faker'

/**
 * For generating fakers strategies check the API (https://www.npmjs.com/package/faker)
 * Description: Generator use as default follow: 
 *  DataTypes.STRING: faker.lorem.sentence()
 *  DataTypes.INTEGER: faker.random.number()
 *  DataTypes.DATE: Date.now()
 * Rembember that is correct as well to build object with params to change this default
 * by building factory as factory.build('passwordreset', { some_ttr: some_other_val }, callback)
 *
 */
factory.define('passwordreset', models.passwordreset, {
  user_id: () => faker.random.number(),
  rest_token: () => faker.lorem.sentence(),
  created_at: () => Date.now()
})

export default factory

TODO list

Follow list is a pending goals list to do (PR are welcome)

Generate test files for mocha inside test folder with .spec.js extension
Add --mode flag to command whit enum likes 'single' for, actual working of generator, and 'massive' followed by a models.json that will be added in future versions of this package.
Add to Readme.md a section of examples, which could have images with how to fetch generated API, through Postman, insomnia or Advanced Restful client as a fetching tool
Add validations to command when user dont use it as he should.

Published

Vulnerabilities

Links

Maintainers

Keywords

Readme