GerindT/reSearch

Welcome to ReSearch a website for AI research papers inspired by ResearchGate and ArXiv

Welcome to ReSearch a website for AI research papers inspired by ResearchGate and ArXiv

Technologies used:

Frontend:

• React js bootstrapped by Vite
• Tailwind Css with Flowbite
• Eslint and Prettier

Bootraped by vite webpack

Backend:

• PHP
• Mysql
• Apache
• Docker for deployment

All Bootstrap by Xampp in local development

Features:

All typical CRUD features like:

• User creation (authentication),deletion,modification and information
• Comment creation,deletion,modification and information
• Paper creation,deletion,information but cannot modify for integrity purposies
• Users have 3 roles, normal users,superusers(moderators),admin
• Admin can asign superusers and have access to the dashboard and all the features superusers have
• Superusers can verify papers in the categories they are assign at, they cannot create other superusers
• Users can create new papers and comment, they can favorite papers and modify their accounts
• Search by title and category that is choosen, pages for erros,success and loading to have consistency of data

API Endpoints:

There are two type of verifications being done, one from the frontend from react with Javascript and givin the correct headers making it nearly impossible to traspass and two being done by the backend when the data is recived in the api endpoint. All verifications are done in react whereas in php some are done to give double security.

Authentication is done with the build in php password_hash() function and password_verify() which use the build in PASSWORD_BCRYPT to specify the salt and hash of crypto.

File upload is used for img and paper (pdf) using move_uploaded_file() function of php to get a url that can be used by the frontend.

Enpoints are based on RPC Json (mostly action based since php worked better this way) and a bit of REST api structure.

• /index.php:
  • GET -> Gets the user info if there is a session in place or not, and all the categories as well
  • POST -> Seperated by actions:
    • Register: Create new user with password hashing
    • Login: Create a session if the user info are correctly send
    • Update: Change user settings, mainly the avatar using the $\_FILES global var
  • DELETE-> Destroy the session when user sign's out
• /posts.php:
  • GET -> Gets all the post with all the needed data, key thing here are string concatination and using joins
  • POST -> Add a new post to the posts table, keeping in mind the file uploading of a PDF paper (for deployment it can range until 5mb)
• /singlePost.php:
  • GET -> Gets a single post with all the data like in posts.php
  • PUT -> Seperated by actions:
    • updateVerification: Used in the moment a admin or superuser updates the status of a paper into a verified one (this goes only one way)
    • toggleFavorite: Used when a user wants to favorite a paper (works as a toggle you can favorite and unfavorite)
  • DELETE -> Deletes a single post given the id, the user is verified before the request is done
• /comments.php:
  • POST -> Creates a single comment for a given paper
  • PUT -> Update a single comment for a given paper
  • DELETE -> Delete a single comment for a given paper
• /dashboard.php:
  • GET -> Get all the users and their categories if they have
  • POST -> Add the superusers in the user table and the categoriestouser if they are added, if not remove their superuser status from the user table

Running localy:

Frontend

Prerequisite Node v14 and up

Run command:

• npm i

• npm run dev

Fronted endpoint localhost:5173

Backend:

Start XAMPP:

• Import sql file in the phpmyadmin dashboard

• Put the api folder in the htdocs folder

Api endpoint localhost:80/api/

Deployment:

Frontend

Netlify using netlify cli

• npm run build

• ntl deploy --prod

Backend:

• Dockerfile to run the php and apache server hosted in render.com

• Remote Mysql server running of oracle free instance

Enjoy 🥳

PS:Cors is enabled and can be accessed in the headers.