🌍 Semi-Automatic AI-Powered CSV Translation Tool

A web-based platform for semi-automatic CSV translation powered by AI (MarianMT) and enhanced with human validation.

---

📑 Table of Contents

• Features
• Screenshots
• Tech Stack
• Architecture
• Quick Start
• Documentation
• Linting & Formatting
• Testing
• Contact

---

✨ Features

Category	Description
📁 File Handling	Upload CSV files, select rows/columns, validate content
🤖 Smart Translation	AI-based translation (MarianMT via Rust), error highlighting, Redis caching
🔐 Authentication	JWT/OAuth2 login, MongoDB-based session persistence
⬇️ Export	Download translated CSV files

---

🖼️ Screenshots

Home Page

Sign Dialog

Dashboard Page

Tips Dialog

Translation Example

---

🛠️ Tech Stack

Layer	Technologies
Frontend	Next.js, React, TypeScript, Tailwind CSS, shadcn/ui
Backend	Django, Django REST Framework, Python, JWT, OAuth2
AI/ML	Rust, PyO3, MarianMT, Hugging Face
Data	MongoDB, Redis
Infrastructure	Docker, GitHub Actions, Postman

---

🧱 Architecture

System architecture of the CSV Translation Tool

---

🚀 Quick Start

Prerequisites

• Python 3.12
• Docker
• Rust
• Node.js (16+)
• npm (bundled with Node.js)

Setup Instructions

1. Clone the repository

   git clone https://github.com/Kerciu/csv-translation-tool.git
   cd csv-translation-tool

2. Ensure Docker Desktop is running

3. Build and run the application

   docker-compose build
   docker-compose up

4. Access the app Navigate to http://localhost:3000 Use test_data/example.csv to try it out!

---

📚 Documentation

🧠 General Overview

• Project documentation available here

🧪 API Documentation

• Swagger available at http://127.0.0.1:8000/swagger/ (when app is running)

🛠️ Building with Sphinx (Python)

1. Install development dependencies

   pip install -r requirements-dev.txt

2. Generate HTML docs From inside server/docs:

   Linux:

   make html
   xdg-open build/html/index.html

   Windows:

   .\make.bat html
   Start-Process .\build\html\index.html

---

🧹 Linting & Formatting

Tools Used

• Python: black, isort, flake8
• React: ESLint, Prettier
• Rust: clippy, rustfmt
• Pre-commit: For unified code quality enforcement

How to Run

Python + React

pip install -r requirements-dev.txt
pre-commit install
pre-commit run --all-files

Rust

cd model
cargo fmt
cargo clippy

---

🧪 Testing

Python

1. Run the app:

   docker-compose up

2. Open shell inside the backend container:

   docker exec -it translator-server bash

3. Run tests:

   python manage.py test

Rust

1. Navigate to model directory:

   cd model

2. Ensure Redis is running on port 6379, then:

   cargo test

---

📬 Contact

For questions, feedback, or collaboration:

Authors

• Kacper Górski • Kerciu • [email protected]
• Szymon Kamiński • RudyKarpus • [email protected]