📊 CSV SQL Playground (Privacy-Friendly Streamlit App)

A lightweight SQL interface built with Streamlit + DuckDB that lets you upload CSV files, register them as DuckDB tables, and run SQL queries in your browser.

🛡️ Designed with privacy in mind — intended to evolve toward full client-side persistence using browser storage.

---

🚀 Live Demo 👉 Try it on Streamlit Cloud

🚀 Features

• 📁 Upload CSV files directly in the browser
• 🧃 Register CSVs as DuckDB tables (locally)
• 🔍 Preview table data
• ✍️ Write and run custom SQL queries on uploaded tables
• 📤 Export query results as downloadable CSVs
• 🧠 Session-based state (no login required)

---

🛠️ Roadmap / TODO

• Integrate streamlit-js-eval
  Store uploaded CSVs in browser localStorage as base64
  ➕ Enables user-side privacy and persistence across sessions

• Delete / Rename Tables
  Add UI to remove or rename registered DuckDB tables

• Integrate OpenAI
  Use natural language prompts to generate SQL queries

• HTML Print View
  Render query results in a clean printable format

• Jupyter-like Cells via eval()
  Allow users to write and execute SQL or markdown-like blocks inline

---

🧰 Tech Stack

• 🐍 Python 3.11+
• ⚡ Streamlit
• 🦆 DuckDB
• 🐼 Pandas
• 🍪 streamlit-js-eval (planned)

---

⚙️ Setup

pip install -r requirements.txt
streamlit run main.py