My Recipe aims to be more than an regular receipe searching website, but a website with tools to provide more insight on the ingredients used for cooking a dish.
Website URL: https://bookncode.com (Currently Closed)

• Technologies
• Architecture
• Database Schema
• Main Features
• Features Demo
  • Home Page
  • Recipe Search
  • Nutrient Level Search
  • Vegetable and Fruit Price Search
  • User Recipes Management
  • User Basic Functions
• APIs Documentation
• Contact

• Node.js / Express.js
• Web Crawler (cheerio)
• JWT Auth.
• CI / CD (Jenkins)
• SSL Certificate (Let's Encrypt)
• Error Handling: Identify type of error and return error message to frontend with corresponding status

• HTML
• CSS
• JavaScript
• AJAX

• MySQL
• Redis (Cache)

• EC2
• S3
• CloudFront

• HTTP & HTTPS
• Domain Name System(DNS)
• NGINX

• Unit Test: Jest
• Load Test: Artillery

• Facebook Login API
• MVC design pattern
• Git / GitHub

• NGINX redirects 443 port requests from clients to corresponding ports
• Result searching process:
  • Search Redis (cache) for matching result and return the result if found
  • If no matching result search MySQL DB for result and return result if found
  • If no matching result in MySQL DB, the app will trigger web crawling and return the found result after storing it into the database.

• Recipe Search
  • Search using dish names
  • Search by inputing ingredients you have
• Nutrient Level Search
  • Renders up to 10 most relevant food and provide nutrient level for each upon search
• Vegetable & Fruit Price Search
  • Return weekly wholesale & retail price from crawler
  • Users can trace food price will be updated automatically weekly
• Member System
  • Supports Facebook Login
  • Upload / Update / Delete Recipes

• Popular recipes are stored in cache (TTL 6hrs) and updated periodically through fetching most liked recipes from databse
• Quick search function using dish name

• Refer to Architecture for Dish Name Search & Ingredients Search process
• Database will grow as more users search for different dishes
• Hot Keywords are rendered by selecting the top 6 most searched keywords recorded in database

• Select 10 most relevant result from 2000+ data and renders nutrient detail on click

• Member only function
• Upon search the application will crawl for wholesale, retail price and relevent result of the searched item and store in Redis
• Users can choose to add items into trace list and the application will update and compare the prices weekly
• Update schedules are achieved thorugh utilizing node-schedule module
• This function uses Redis as database, because information are regularly updated (optimize efficiency)

• Users can upload / update and delete recipes through profile
• Images of recipes are uploaded to AWS S3 and read through AWS CloudFront to optimise load speed

• Users can add recipe to their favorite and manage favorites in their profile
• Users can also like recipes which is then used to render popular recipes
• Profile page also allow users to change basic information and password

For details on APIs request and response, click the link below:

Link: APIs Doc.