app.py

import streamlit as st
import pandas as pd
import os
from dotenv import load_dotenv
from transformers import AutoTokenizer, AutoModelForCausalLM
import logging

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

class LLMService:
    def __init__(self, db_path):
        self.db_path = db_path
        # Load tokenizer and model
        self.tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-72B-Instruct")
        self.model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-72B-Instruct")

    def convert_to_sql_query(self, natural_query):
        try:
            # Tokenize input
            inputs = self.tokenizer(f"Translate this to SQL: {natural_query}", return_tensors="pt")
            # Generate output
            outputs = self.model.generate(**inputs, max_length=512, num_beams=5)
            # Decode output
            sql_query = self.tokenizer.decode(outputs[0], skip_special_tokens=True)
            return {"success": True, "query": sql_query}
        except Exception as e:
            logger.error(f"Error generating SQL query: {e}")
            return {"success": False, "error": str(e)}

    def execute_query(self, sql_query):
        try:
            import sqlite3
            conn = sqlite3.connect(self.db_path)
            cursor = conn.cursor()
            cursor.execute(sql_query)
            results = cursor.fetchall()
            columns = [desc[0] for desc in cursor.description]
            conn.close()
            return {"success": True, "results": results, "columns": columns}
        except Exception as e:
            logger.error(f"Error executing SQL query: {e}")
            return {"success": False, "error": str(e)}

def main():
    st.title("Natural Language to SQL Query Converter")
    st.write("Enter your question about the database in natural language.")
    
    # Load environment variables
    load_dotenv()
    db_path = os.getenv("DB_PATH")
    
    if not db_path:
        st.error("Missing database path in environment variables.")
        logger.error("DB path not found in environment variables.")
        return
    
    # Initialize LLM Service
    try:
        llm_service = LLMService(db_path=db_path)
    except Exception as e:
        st.error(f"Error initializing service: {str(e)}")
        return

    # Input for natural language query
    natural_query = st.text_area("Enter your query", "Show me all albums by artist 'Queen'", height=100)

    if st.button("Generate and Execute Query"):
        if not natural_query.strip():
            st.warning("Please enter a valid query.")
            return

        # Convert to SQL
        with st.spinner("Generating SQL query..."):
            sql_result = llm_service.convert_to_sql_query(natural_query)

        if not sql_result["success"]:
            st.error(f"Error generating SQL query: {sql_result['error']}")
            return

        # Display generated SQL
        st.subheader("Generated SQL Query:")
        st.code(sql_result["query"], language="sql")

        # Execute query
        with st.spinner("Executing query..."):
            query_result = llm_service.execute_query(sql_result["query"])

        if not query_result["success"]:
            st.error(f"Error executing query: {query_result['error']}")
            return

        # Check if there are results
        if query_result["results"]:
            df = pd.DataFrame(query_result["results"], columns=query_result["columns"])

            # Create a collapsible DataFrame using Streamlit's expander
            with st.expander("Click to view query results as a DataFrame"):
                st.dataframe(df)

            # Extract product details from the JSON result and display them
            json_results = df.to_dict(orient='records')
            if "title" in json_results[0] and "images" in json_results[0] and "price" in json_results[0]:
                st.subheader("Product Details:")
                for product in json_results:
                    price = product.get("price", "N/A")
                    title = product.get("handle", "N/A")
                    src = product.get("src", "N/A")
                    
                    # Display product details in a neat format using columns for alignment
                    with st.container():
                        col1, col2, col3 = st.columns([1, 2, 3])  # Adjust column widths as needed
                        
                        with col1:
                            st.image(src, use_container_width=True)  # Display product image with container width
                        with col2:
                            st.write(f"**Price:** {price}")  # Display price
                            st.write(f"**Title:** {title}")  # Display title
                        with col3:
                            st.write(f"**Image Source:** [Link]( {src} )")  # Link to the image if needed
        else:
            st.info("No results found.")

if __name__ == "__main__":
    main()