DataLoom

A web-based GUI for data wrangling — manage and transform tabular datasets (CSV, TSV, JSON, XLSX, Parquet) through a graphical interface powered by pandas, without writing code.

Supported File Formats

• CSV (.csv)
• TSV (.tsv)
• JSON (.json)
• XLSX (.xlsx)
• Parquet (.parquet)

Features

• Upload and manage datasets in CSV, TSV, JSON, XLSX, and Parquet formats through a graphical interface
• Apply pandas-powered transformations: filter, sort, pivot, deduplicate, and more
• Inline cell editing and row/column management
• Checkpoint system — save and revert dataset states
• Full action history tracking via change logs

Prerequisites

• Node.js >= 18
• Python 3.12+
• uv (Python package manager)
• PostgreSQL

Getting Started

Backend

cd dataloom-backend
cp .env.example .env          # Configure DB credentials
uv sync
uv run uvicorn app.main:app --reload --port 4200

Frontend

cd dataloom-frontend
npm install
npm run dev

Service	Port
Frontend	3200
Backend	4200

Running Tests

# Backend
cd dataloom-backend && uv run pytest

# Frontend
cd dataloom-frontend && npm run test

Project Structure

dataloom/
  dataloom-backend/    # Python FastAPI server
  dataloom-frontend/   # React + Vite SPA

Contributing

See CONTRIBUTING.md for guidelines.

License

Apache 2.0

Author

Oshan Mudannayake

For questions or queries about this project, please reach out via email.