Reci is a social recipe-sharing app where users can post recipes, review what they’ve cooked, and follow friends to stay updated on what’s hot in their kitchens.
This repo is strucutred as a mono-repo, with our frontend living under frontend/, and our backend living in backend/. The frontend is built using React, while the backend is built with Express. Both can be started with npm run dev, but see the startup instructions for more details. This is where the end-to-end application code that implements all the claimed features is.
Of special note, the backend pulls the SQL instructions it needs at runtime from the backend/sql/ folder. This folder is where all SQL lives for our project, including our seeding code.
Instead of having a single test-sample.sql, for each of our features listed in the Milestone 1 report, you can find their query in the backend/sql/ folder, and their corresponding expected output in backend/sql/test-sample-out/ such as backend/sql/test-sample-out/filterRecipes.out, as well as query performance data in backend/sql/test-production-out such as backend/sql/test-production-out/filterRecipes.log
You can find the corresponding expected output for production data in backend/sql/test-production-out such as backend/sql/test-production-out/filterRecipes.out as well as query performance in backend/sql/test-production-out such as backend/sql/test-production-out/filterRecipes.log
Every time the app starts up, we automatically delete the existing database instance (by doing docker compose down -v which deletes it's volume), create a new database container (by doing docker compose up), automatically run backend/sql/init.sql to create tables, constraints, stored procedures, and triggers, and finally run backend/sql/seed.sql to seed the database with toy data.
This process is made to ensure a consistent developer environement, making it fast to iterate on our SQL schemas.
Our production data generation script also lives in backend/sql/seed/, and you can read more about it in the README located there. This is the code for downloading/scraping/transforming real data that we have written.
For instructions to generate our production data, read backend/sql/seed/README.md. We consider production data as a dataset with 100k - 1 million total entries.
This is how to run our working database-driven application.
- Installed docker
- Install npm
- From the root of the repo, run
npm install - Create a
.envby copying the contents ofexample.env
As a note: the JWT secret is generated with openssl rand -base64 32, but we use the same secret in dev for consistency.
- Make sure docker is running
- Run
npm run dev. It will start the docker container for the database, npm install in FE and BE, startup the dev apps, and automatically reset and seed the database for you!
If you run npm run dev and notice you get an error containing error: password authentication failed for user "postgres", then there is likely an issue with the port you are using.
To solve the problem, navigate to docker-compose.yml and change the line containing the two ports "5433:5432" to something else, ex: "5432:5432". You will then need to change the port in the backend\src\db.ts to whatever new port you decided to use, ex: port: 5432.
Then, go through the setup steps again and it should work!
We've worked extensively on our backend systems with a long suite of fleshed out SQL queries and endpoints. Current features include:
- Fully working SQL database with dynamic loading of files
- Signing up and into accounts, and authentication with JWT cookies
- CRUD (Create Read Update Delete) operations for recipies, accounts
- Searching for recipies
- Reading recipies
- Reading followers, following, account information
- Frontend views for the feed, recipe search, settings, profile pages, and recipe views to read the recipes
- Reading recipe reviews, with support for adding images
- Settings page, login
- Fancy: a dynamic and personalized landing page feed with signed in and annonymous views
- Fancy: using triggers to automatically save recipes that you review
- Fancy: database level following constraints (you can't follow yourself, or follow the same person multiple times)
- User profile page with added statistics, followers, following, your recipies and reviews
- Your persronalized cookbook page (frontend) with all of your authored and saved recipes
- Add recipe and review modals
- Improved frontend navigation through hyperlinks in more places
- Fancy: advanced indexing to accelerate queries
- Fancy: materialized view for anonymous feed to greatly accelerate traffic
