QuerySUTRA 0.2.0

SUTRA: Structured-Unstructured-Text-Retrieval-Architecture | AI-powered entity extraction from PDF/DOCX/TXT

🚀 QuerySUTRA v0.1.3

Structured-Unstructured-Text-Retrieval-Architecture

Natural Language to SQL with Cloud Export | PDF, DOCX, TXT Support

A comprehensive Python library that converts natural language questions into SQL queries, with support for multiple file formats and cloud database export.

---

✨ Key Features

✅ Natural Language to SQL - Ask questions in plain English
✅ Multiple Formats - CSV, Excel, JSON, SQL, PDF, DOCX, TXT, DataFrame
✅ Cloud Export - MySQL, PostgreSQL (local & cloud)
✅ Direct SQL - No API cost option
✅ Auto Visualization - Plotly/Matplotlib charts
✅ Interactive Mode - Ask user for visualization choice
✅ Complete Backup - Export to SQLite, JSON, Excel
✅ Jupyter Ready - Perfect for notebooks

---

📦 Installation

# Basic installation
pip install QuerySUTRA

# With MySQL support
pip install QuerySUTRA[mysql]

# With PostgreSQL support
pip install QuerySUTRA[postgres]

# With all database support
pip install QuerySUTRA[all]

---

🎯 Quick Start

from sutra import SUTRA

# Initialize with OpenAI API key
sutra = SUTRA(api_key="your-openai-key")

# Upload any format
sutra.upload("data.csv")      # CSV
sutra.upload("report.pdf")    # PDF ✨
sutra.upload("doc.docx")      # Word ✨
sutra.upload("data.xlsx")     # Excel
sutra.upload(dataframe)       # DataFrame

# Query with natural language
result = sutra.ask("What are the top 5 products?", viz=True)
print(result.data)

# Export to cloud
sutra.save_to_mysql("localhost", "root", "pass", "mydb")
sutra.save_to_postgres("host", "user", "pass", "db")

# Complete backup
sutra.backup()

---

📄 Supported File Formats

Format	Extension	Example
CSV	.csv	sutra.upload("data.csv")
Excel	.xlsx, .xls	sutra.upload("data.xlsx")
JSON	.json	sutra.upload("data.json")
SQL	.sql	sutra.upload("schema.sql")
PDF	.pdf	sutra.upload("report.pdf") ✨
Word	.docx	sutra.upload("document.docx") ✨
Text	.txt	sutra.upload("data.txt") ✨
DataFrame	pd.DataFrame	sutra.upload(df, name="sales")

---

🔥 New in v0.1.3

1. PDF Support

# Upload PDF files
sutra.upload("annual_report.pdf")

# Query the content
result = sutra.ask("What are the key findings in this report?")
print(result.data)

2. Word Document Support

# Upload DOCX files with tables
sutra.upload("sales_report.docx")

# Query the data
result = sutra.ask("Show me sales by region", viz=True)

3. Cloud Database Export

MySQL (Local or Cloud)

# Local MySQL
sutra.save_to_mysql("localhost", "root", "password", "mydb")

# AWS RDS MySQL
sutra.save_to_mysql(
    host="mydb.xxxx.us-east-1.rds.amazonaws.com",
    user="admin",
    password="cloudpass",
    database="production"
)

# Google Cloud SQL
sutra.save_to_mysql(
    host="35.123.456.789",
    user="admin",
    password="pass",
    database="mydb"
)

PostgreSQL (Local or Cloud)

# Local PostgreSQL
sutra.save_to_postgres("localhost", "postgres", "password", "mydb")

# Heroku PostgreSQL
sutra.save_to_postgres(
    host="ec2-xxx.compute-1.amazonaws.com",
    user="user",
    password="pass",
    database="dbname"
)

# AWS RDS PostgreSQL
sutra.save_to_postgres(
    host="mydb.xxxx.us-west-2.rds.amazonaws.com",
    user="admin",
    password="pass",
    database="prod"
)

4. Complete Export & Backup

# Export entire database
sutra.export_db("backup.db", format="sqlite")
sutra.export_db("dump.sql", format="sql")
sutra.export_db("data.json", format="json")
sutra.export_db("data.xlsx", format="excel")

# Export schema only
sutra.save_schema("schema.sql", format="sql")
sutra.save_schema("schema.json", format="json")
sutra.save_schema("schema.md", format="markdown")

# Complete backup (creates 3 files)
sutra.backup()  # Creates .db, .sql, .json files with timestamp

---

📖 Complete Examples

Example 1: PDF Analysis

from sutra import SUTRA

sutra = SUTRA(api_key="your-openai-key")

# Upload PDF
sutra.upload("financial_report.pdf")

# View extracted data
sutra.peek(n=10)

# Query the content
result = sutra.ask("What are the total revenues?")
print(result.data)

# Visualize
result = sutra.ask("Show revenue by quarter", viz=True)

Example 2: Multi-Format Analysis

sutra = SUTRA(api_key="your-key")

# Upload multiple formats
sutra.upload("sales.csv")
sutra.upload("report.docx")
sutra.upload("data.xlsx")

# List all tables
print(sutra.tables())

# Query across data
result = sutra.ask("What are total sales?")
print(result.data)

Example 3: Cloud Deployment

# Analyze in Colab/Jupyter
sutra = SUTRA(api_key="your-key")
sutra.upload("local_analysis.csv")

# Query and analyze
result = sutra.ask("Show top performers", viz=True)

# Deploy to production MySQL
sutra.save_to_mysql(
    host="production.mysql.com",
    user="admin",
    password="prod_password",
    database="analytics_db"
)

# Backup everything
sutra.backup("/backups")

Example 4: Direct SQL (No API Cost)

# Execute SQL directly - FREE!
result = sutra.sql("""
    SELECT region, 
           SUM(sales) as total_sales,
           AVG(sales) as avg_sales
    FROM sales_data 
    GROUP BY region
    ORDER BY total_sales DESC
""")

print(result.data)

Example 5: Interactive Mode

# Ask user for visualization preference
result = sutra.interactive("What are sales trends?")
# Prompts: "Do you want visualization? (yes/no):"

if result.success:
    print(result.data)

---

🛠️ API Reference

Initialization

sutra = SUTRA(api_key="your-openai-key", db="sutra.db")

Upload Data

sutra.upload(data, name="table_name")
# data = file path (str) or DataFrame

View Database

sutra.tables()          # List all tables
sutra.schema()          # Show database schema
sutra.peek(n=10)       # Preview data

Query Data

# Direct SQL (no API cost)
result = sutra.sql("SELECT * FROM table", viz=False)

# Natural language (uses API)
result = sutra.ask("question", viz=False)

# Interactive (prompts user)
result = sutra.interactive("question")

Export & Backup

# Export results
sutra.export(dataframe, "output.csv", format="csv")

# Export database
sutra.export_db("backup.db", format="sqlite")

# Save to cloud
sutra.save_to_mysql(host, user, password, database)
sutra.save_to_postgres(host, user, password, database)

# Complete backup
sutra.backup("/backup/path")

QueryResult Object

result.success   # bool - query succeeded
result.sql       # str - generated SQL
result.data      # DataFrame - query results
result.viz       # figure - visualization (if viz=True)
result.error     # str - error message (if failed)

---

💡 Use Cases

Data Analysis

sutra.upload("sales_data.csv")
result = sutra.ask("What products have declining sales?", viz=True)

Document Processing

sutra.upload("contract.pdf")
result = sutra.ask("What are the key terms and dates?")

Multi-Source Integration

sutra.upload("sales.csv")
sutra.upload("inventory.xlsx")
sutra.upload("report.docx")
result = sutra.ask("Combine all data sources")

Cloud Migration

# Local analysis
sutra.upload("data.csv")
result = sutra.ask("Analyze trends")

# Deploy to cloud
sutra.save_to_postgres("cloud-db.com", "user", "pass", "prod")

---

🎨 Features Comparison

Feature	Available	Cost
CSV/Excel/JSON Upload	✅	Free
PDF Upload	✅	Free
DOCX Upload	✅	Free
Direct SQL Queries	✅	Free
Natural Language Queries	✅	~$0.001/query
Visualization	✅	Free
MySQL Export	✅	Free
PostgreSQL Export	✅	Free
Backup & Export	✅	Free

---

💰 Cost Optimization

# FREE - Direct SQL (no API calls)
result = sutra.sql("SELECT * FROM data WHERE sales > 1000")

# PAID - Natural language (uses OpenAI API)
result = sutra.ask("Show products with sales over 1000")

# Tip: Use direct SQL when you know the query!

---

🧪 Testing

# Install
pip install QuerySUTRA

# Test
python -c "from sutra import SUTRA; print('✅ Success!')"

---

📚 Documentation

• Full Guide: See SUTRA_Complete_Guide.ipynb
• Publishing: See PUBLISHING_GUIDE.md
• Examples: See complete_example.py

---

🤝 Contributing

Contributions welcome! The main code is in sutra/sutra.py - a single, well-documented file.

---

📄 License

MIT License - Free to use in your projects!

---

🏆 Why QuerySUTRA?

• SUTRA = Structured-Unstructured-Text-Retrieval-Architecture
• Single-file design for simplicity
• Production-ready with error handling
• Cloud-native with export capabilities
• Comprehensive format support (PDF, DOCX, CSV, Excel, JSON)
• Cost-effective with free SQL mode

---

🌟 Credits

Author: Aditya Batta
Version: 0.1.3
License: MIT

---

📞 Support

• Issues: GitHub Issues
• PyPI: https://pypi.org/project/QuerySUTRA/

---

Made with ❤️ for data analysts and developers

Start analyzing with natural language today! 🚀