Upload 13 files

.gitignore

@@ -0,0 +1 @@
+.env 

admin_dashboard.py

@@ -0,0 +1,182 @@
+import streamlit as st
+import pandas as pd
+import os
+import sys
+from dotenv import load_dotenv
+
+# Add src to path
+sys.path.append(os.getcwd())
+
+from src.rag_manager import RAGManager
+from src.sql_generator import SQLGenerator
+from src.db_connector import DatabaseConnector
+
+# --- 1. CONFIGURATION ---
+st.set_page_config(
+    page_title="NexusAI | Enterprise Data",
+    page_icon="✨",
+    layout="wide",
+    initial_sidebar_state="collapsed"
+)
+
+# Custom CSS
+st.markdown("""
+<style>
+    @import url('https://fonts.googleapis.com/css2?family=Inter:wght@400;600&display=swap');
+    html, body, [class*="css"] { font-family: 'Inter', sans-serif; }
+    .stApp { background-color: #0F1117; }
+    #MainMenu, footer, header { visibility: hidden; }
+    
+    .stChatMessage { background-color: transparent !important; border: none !important; }
+    
+    div[data-testid="stChatMessage"]:nth-child(odd) { flex-direction: row-reverse; }
+    div[data-testid="stChatMessage"]:nth-child(odd) .stMarkdown {
+        background-color: #2B2D31; color: #E0E0E0;
+        border-radius: 18px 18px 4px 18px; padding: 12px 20px;
+        text-align: right; margin-left: auto;
+    }
+    
+    div[data-testid="stChatMessage"]:nth-child(even) .stMarkdown {
+        background-color: transparent; color: #F0F0F0; padding-left: 10px;
+    }
+    
+    .stChatInput { position: fixed; bottom: 30px; width: 70% !important; left: 50%; transform: translateX(-50%); z-index: 1000; }
+    .stTextInput > div > div > input { background-color: #1E2128; color: white; border-radius: 24px; border: 1px solid #363B47; }
+    
+    div[data-testid="stDataFrame"] { background-color: #161920; border-radius: 10px; padding: 10px; border: 1px solid #30363D; }
+    section[data-testid="stSidebar"] { background-color: #0E1015; border-right: 1px solid #222; }
+</style>
+""", unsafe_allow_html=True)
+
+# --- 2. INITIALIZATION ---
+@st.cache_resource
+def get_core():
+    load_dotenv()
+    key = os.getenv("GEMINI_API_KEY")
+    return RAGManager(), SQLGenerator(key), DatabaseConnector()
+
+try:
+    rag, sql_gen, db = get_core()
+except Exception as e:
+    st.error(f"System Offline: {e}")
+    st.stop()
+
+# --- 3. SIDEBAR ---
+with st.sidebar:
+    st.markdown("## 🧠 NexusAI")
+    st.caption("Enterprise SQL Agent v2.0")
+    st.divider()
+    
+    if db:
+        st.success("🟢 Database Connected")
+    
+    st.markdown("### 📚 Quick Prompts")
+    prompts = [
+        "Top 5 employees by salary",
+        "Total sales revenue by Region",
+        "Show me products with low stock",
+        "Which department spends the most?"
+    ]
+    
+    for p in prompts:
+        if st.button(p, use_container_width=True):
+            st.session_state.last_prompt = p
+            
+    if st.button("🗑️ Clear Context", type="primary", use_container_width=True):
+        st.session_state.messages = []
+        st.rerun()
+
+# --- 4. MAIN INTERFACE ---
+if "messages" not in st.session_state:
+    st.session_state.messages = []
+
+if not st.session_state.messages:
+    st.markdown("""
+    <div style="text-align: center; margin-top: 100px;">
+        <h1 style="font-size: 3rem; background: -webkit-linear-gradient(#eee, #333); -webkit-background-clip: text; -webkit-text-fill-color: transparent;">
+            What can I help you analyze?
+        </h1>
+        <p style="color: #666;">Connect to your database and ask questions in plain English.</p>
+    </div>
+    """, unsafe_allow_html=True)
+
+for msg in st.session_state.messages:
+    with st.chat_message(msg["role"], avatar="👤" if msg["role"] == "user" else "✨"):
+        st.markdown(msg["content"])
+        
+        if "data" in msg:
+            # ✅ FIX: Switched to clean dataframe display
+            st.dataframe(msg["data"], hide_index=True)
+        if "chart" in msg:
+            st.bar_chart(msg["chart"])
+        if "sql" in msg:
+            with st.expander("🛠️ View Query Logic"):
+                st.code(msg["sql"], language="sql")
+
+# Handle Input
+user_input = st.chat_input("Ask anything...")
+
+if "last_prompt" in st.session_state and st.session_state.last_prompt:
+    user_input = st.session_state.last_prompt
+    st.session_state.last_prompt = None
+
+if user_input:
+    st.session_state.messages.append({"role": "user", "content": user_input})
+    with st.chat_message("user", avatar="👤"):
+        st.markdown(user_input)
+
+    with st.chat_message("assistant", avatar="✨"):
+        status_box = st.empty()
+        status_box.markdown("`⚡ analyzing...`")
+        
+        try:
+            tables = rag.get_relevant_tables(user_input)
+            context = "\n".join(tables)
+            
+            sql = sql_gen.generate_sql(user_input, context)
+            
+            results = db.execute_sql(sql)
+            status_box.empty()
+            
+            if not results:
+                response = "No data found matching that request."
+                st.markdown(response)
+                st.session_state.messages.append({"role": "assistant", "content": response, "sql": sql})
+            else:
+                df = pd.DataFrame(results)
+                df_clean = df.reset_index(drop=True)
+                
+                response = f"Found **{len(df)}** records."
+                st.markdown(response)
+                # ✅ FIX: Updated dataframe display
+                st.dataframe(df_clean, hide_index=True)
+                
+                chart_data = None
+                numeric_cols = df_clean.select_dtypes(include=['number']).columns
+                
+                if not numeric_cols.empty and len(df_clean) > 1:
+                    try:
+                        non_numeric = df_clean.select_dtypes(exclude=['number']).columns
+                        st.markdown("##### 📊 Trends")
+                        if not non_numeric.empty:
+                            x_axis = non_numeric[0]
+                            y_axis = numeric_cols[0]
+                            chart_data = df_clean.set_index(x_axis)[y_axis]
+                            st.bar_chart(chart_data, color="#7B61FF")
+                        else:
+                            chart_data = df_clean[numeric_cols[0]]
+                            st.bar_chart(chart_data, color="#7B61FF")
+                    except Exception:
+                        pass
+
+                st.session_state.messages.append({
+                    "role": "assistant",
+                    "content": response,
+                    "data": df_clean,
+                    "chart": chart_data,
+                    "sql": sql
+                })
+
+        except Exception as e:
+            status_box.empty()
+            st.error(f"Error: {e}")

api_server.py

@@ -0,0 +1,128 @@
+import sys
+import os
+import uvicorn
+import re
+
+# 🚨 FORCE PYTHON TO FIND THE 'src' FOLDER
+sys.path.append(os.path.dirname(os.path.abspath(__file__)))
+
+from fastapi import FastAPI, HTTPException
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+from typing import List, Optional
+
+# Import your custom modules
+from src.db_connector import Database
+from src.rag_manager import RAGSystem 
+from src.sql_generator import SQLGenerator
+
+app = FastAPI()
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+class ChatRequest(BaseModel):
+    question: str
+    history: Optional[List[dict]] = [] 
+
+# --- HELPER: CLEAN AI OUTPUT ---
+def clean_sql(sql_text: str) -> str:
+    """Removes markdown, 'sql' tags, and extra whitespace."""
+    if not sql_text:
+        return ""
+    # Remove markdown code blocks (```sql ... ```)
+    cleaned = re.sub(r"```sql|```", "", sql_text, flags=re.IGNORECASE).strip()
+    # Remove trailing semicolons for consistency (optional, depending on DB driver)
+    cleaned = cleaned.rstrip(';') 
+    return cleaned
+
+print("--- 🚀 SYSTEM STARTUP SEQUENCE ---")
+try:
+    print("   ...Connecting to Database")
+    db = Database()
+    print("   ✅ Database Connection: SUCCESS")
+    
+    print("   ...Initializing RAG System")
+    rag = RAGSystem(db)
+    print("   ✅ RAG System: ONLINE")
+    
+    print("   ...Loading AI Model")
+    generator = SQLGenerator()
+    print("   ✅ AI Model: LOADED")
+    
+except Exception as e:
+    print(f"   ❌ CRITICAL STARTUP ERROR: {e}")
+
+@app.post("/chat")
+def chat_endpoint(request: ChatRequest):
+    print(f"\n📨 NEW REQUEST: {request.question}") 
+    
+    try:
+        # 1. Get Context
+        context = rag.get_relevant_schema(request.question)
+        
+        # 2. Generate SQL
+        raw_sql, explanation, friendly_msg = generator.generate_sql(request.question, context, request.history)
+        
+        # 3. Clean and Validate SQL
+        # If the generator returned an error string directly (Error 1 fix)
+        if "Error:" in raw_sql or "Invalid Query" in raw_sql:
+            return {
+                "answer": [],
+                "sql": raw_sql,
+                "message": "I couldn't generate a safe query for that request. Try asking for specific data like 'Show me users' or 'List orders'.",
+                "follow_ups": ["Show top 10 rows from users", "Count total orders"]
+            }
+
+        cleaned_sql = clean_sql(raw_sql)
+        print(f"   🧹 Cleaned SQL: {cleaned_sql}")
+
+        # Safety check: Ensure it's a SELECT
+        if not cleaned_sql.upper().startswith("SELECT"):
+             return {
+                "answer": [],
+                "sql": cleaned_sql,
+                "message": "Security Alert: I can only perform READ (SELECT) operations.",
+                "follow_ups": []
+            }
+
+        # 4. Run Query (Error 2 fix)
+        try:
+            results = db.run_query(cleaned_sql)
+        except Exception as db_err:
+            # Catch MySQL Syntax errors specifically
+            print(f"   ⚠️ DB Error: {db_err}")
+            return {
+                "answer": [f"Error: {str(db_err)}"], # This puts the error in a safe list
+                "sql": cleaned_sql,
+                "message": "There was a syntax error in the generated SQL. I have displayed the error above.",
+                "follow_ups": []
+            }
+        
+        # 5. Generate Follow-ups
+        follow_ups = generator.generate_followup_questions(request.question, cleaned_sql)
+        
+        return {
+            "answer": results,
+            "sql": cleaned_sql,
+            "explanation": explanation,
+            "message": friendly_msg, 
+            "follow_ups": follow_ups
+        }
+
+    except Exception as e:
+        print(f"❌ General Processing Error: {e}")
+        return {
+            "answer": [],
+            "sql": "-- System Error",
+            "message": f"Critical Error: {str(e)}",
+            "follow_ups": []
+        }
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=8000)

app_ui.py

@@ -0,0 +1,182 @@
+import streamlit as st
+import pandas as pd
+import os
+import sys
+from dotenv import load_dotenv
+
+# Add src to path
+sys.path.append(os.getcwd())
+
+from src.rag_manager import RAGManager
+from src.sql_generator import SQLGenerator
+from src.db_connector import DatabaseConnector
+
+# --- 1. CONFIGURATION ---
+st.set_page_config(
+    page_title="NexusAI | Enterprise Data",
+    page_icon="✨",
+    layout="wide",
+    initial_sidebar_state="collapsed"
+)
+
+# Custom CSS
+st.markdown("""
+<style>
+    @import url('https://fonts.googleapis.com/css2?family=Inter:wght@400;600&display=swap');
+    html, body, [class*="css"] { font-family: 'Inter', sans-serif; }
+    .stApp { background-color: #0F1117; }
+    #MainMenu, footer, header { visibility: hidden; }
+    
+    .stChatMessage { background-color: transparent !important; border: none !important; }
+    
+    div[data-testid="stChatMessage"]:nth-child(odd) { flex-direction: row-reverse; }
+    div[data-testid="stChatMessage"]:nth-child(odd) .stMarkdown {
+        background-color: #2B2D31; color: #E0E0E0;
+        border-radius: 18px 18px 4px 18px; padding: 12px 20px;
+        text-align: right; margin-left: auto;
+    }
+    
+    div[data-testid="stChatMessage"]:nth-child(even) .stMarkdown {
+        background-color: transparent; color: #F0F0F0; padding-left: 10px;
+    }
+    
+    .stChatInput { position: fixed; bottom: 30px; width: 70% !important; left: 50%; transform: translateX(-50%); z-index: 1000; }
+    .stTextInput > div > div > input { background-color: #1E2128; color: white; border-radius: 24px; border: 1px solid #363B47; }
+    
+    div[data-testid="stDataFrame"] { background-color: #161920; border-radius: 10px; padding: 10px; border: 1px solid #30363D; }
+    section[data-testid="stSidebar"] { background-color: #0E1015; border-right: 1px solid #222; }
+</style>
+""", unsafe_allow_html=True)
+
+# --- 2. INITIALIZATION ---
+@st.cache_resource
+def get_core():
+    load_dotenv()
+    key = os.getenv("GEMINI_API_KEY")
+    return RAGManager(), SQLGenerator(key), DatabaseConnector()
+
+try:
+    rag, sql_gen, db = get_core()
+except Exception as e:
+    st.error(f"System Offline: {e}")
+    st.stop()
+
+# --- 3. SIDEBAR ---
+with st.sidebar:
+    st.markdown("## 🧠 NexusAI")
+    st.caption("Enterprise SQL Agent v2.0")
+    st.divider()
+    
+    if db:
+        st.success("🟢 Database Connected")
+    
+    st.markdown("### 📚 Quick Prompts")
+    prompts = [
+        "Top 5 employees by salary",
+        "Total sales revenue by Region",
+        "Show me products with low stock",
+        "Which department spends the most?"
+    ]
+    
+    for p in prompts:
+        if st.button(p, use_container_width=True):
+            st.session_state.last_prompt = p
+            
+    if st.button("🗑️ Clear Context", type="primary", use_container_width=True):
+        st.session_state.messages = []
+        st.rerun()
+
+# --- 4. MAIN INTERFACE ---
+if "messages" not in st.session_state:
+    st.session_state.messages = []
+
+if not st.session_state.messages:
+    st.markdown("""
+    <div style="text-align: center; margin-top: 100px;">
+        <h1 style="font-size: 3rem; background: -webkit-linear-gradient(#eee, #333); -webkit-background-clip: text; -webkit-text-fill-color: transparent;">
+            What can I help you analyze?
+        </h1>
+        <p style="color: #666;">Connect to your database and ask questions in plain English.</p>
+    </div>
+    """, unsafe_allow_html=True)
+
+for msg in st.session_state.messages:
+    with st.chat_message(msg["role"], avatar="👤" if msg["role"] == "user" else "✨"):
+        st.markdown(msg["content"])
+        
+        if "data" in msg:
+            # ✅ FIX: Switched to clean dataframe display
+            st.dataframe(msg["data"], hide_index=True)
+        if "chart" in msg:
+            st.bar_chart(msg["chart"])
+        if "sql" in msg:
+            with st.expander("🛠️ View Query Logic"):
+                st.code(msg["sql"], language="sql")
+
+# Handle Input
+user_input = st.chat_input("Ask anything...")
+
+if "last_prompt" in st.session_state and st.session_state.last_prompt:
+    user_input = st.session_state.last_prompt
+    st.session_state.last_prompt = None
+
+if user_input:
+    st.session_state.messages.append({"role": "user", "content": user_input})
+    with st.chat_message("user", avatar="👤"):
+        st.markdown(user_input)
+
+    with st.chat_message("assistant", avatar="✨"):
+        status_box = st.empty()
+        status_box.markdown("`⚡ analyzing...`")
+        
+        try:
+            tables = rag.get_relevant_tables(user_input)
+            context = "\n".join(tables)
+            
+            sql = sql_gen.generate_sql(user_input, context)
+            
+            results = db.execute_sql(sql)
+            status_box.empty()
+            
+            if not results:
+                response = "No data found matching that request."
+                st.markdown(response)
+                st.session_state.messages.append({"role": "assistant", "content": response, "sql": sql})
+            else:
+                df = pd.DataFrame(results)
+                df_clean = df.reset_index(drop=True)
+                
+                response = f"Found **{len(df)}** records."
+                st.markdown(response)
+                # ✅ FIX: Updated dataframe display
+                st.dataframe(df_clean, hide_index=True)
+                
+                chart_data = None
+                numeric_cols = df_clean.select_dtypes(include=['number']).columns
+                
+                if not numeric_cols.empty and len(df_clean) > 1:
+                    try:
+                        non_numeric = df_clean.select_dtypes(exclude=['number']).columns
+                        st.markdown("##### 📊 Trends")
+                        if not non_numeric.empty:
+                            x_axis = non_numeric[0]
+                            y_axis = numeric_cols[0]
+                            chart_data = df_clean.set_index(x_axis)[y_axis]
+                            st.bar_chart(chart_data, color="#7B61FF")
+                        else:
+                            chart_data = df_clean[numeric_cols[0]]
+                            st.bar_chart(chart_data, color="#7B61FF")
+                    except Exception:
+                        pass
+
+                st.session_state.messages.append({
+                    "role": "assistant",
+                    "content": response,
+                    "data": df_clean,
+                    "chart": chart_data,
+                    "sql": sql
+                })
+
+        except Exception as e:
+            status_box.empty()
+            st.error(f"Error: {e}")

check_schema.py

@@ -0,0 +1,20 @@
+import os
+from dotenv import load_dotenv
+from src.db_connector import Database
+
+# Load your actual database connection
+db = Database()
+
+print("\n--- 🔍 REAL DATABASE SCHEMA ---")
+try:
+    # Get all tables
+    tables = db.get_tables()
+    for table in tables:
+        print(f"\n📂 TABLE: {table}")
+        # Get columns for this table
+        columns = db.get_table_schema(table)
+        for col in columns:
+            print(f"   - {col}")
+    print("\n-------------------------------")
+except Exception as e:
+    print(f"❌ Error reading schema: {e}")

db.sql

@@ -0,0 +1,34 @@
+create database chatbot;
+use chatbot;
+
+CREATE USER 'bot_user'@'%' IDENTIFIED BY 'YourSecurePassword123!';
+GRANT SELECT ON chatbot.* TO 'bot_user'@'%';
+FLUSH PRIVILEGES;
+
+
+USE chatbot;
+
+CREATE TABLE employees (
+    id INT AUTO_INCREMENT PRIMARY KEY,
+    name VARCHAR(100),
+    department VARCHAR(50),
+    salary DECIMAL(10,2),
+    hire_date DATE
+);
+
+CREATE TABLE sales (
+    sale_id INT AUTO_INCREMENT PRIMARY KEY,
+    employee_id INT,
+    amount DECIMAL(10,2),
+    sale_date DATE,
+    FOREIGN KEY (employee_id) REFERENCES employees(id)
+);
+
+-- Insert a little dummy data
+INSERT INTO employees (name, department, salary, hire_date) VALUES 
+('Alice', 'Sales', 70000, '2023-01-15'),
+('Bob', 'Engineering', 90000, '2022-05-20');
+
+INSERT INTO sales (employee_id, amount, sale_date) VALUES 
+(1, 500.00, '2023-06-01'),
+(1, 1200.50, '2023-06-03');

db_connector.py

@@ -0,0 +1,84 @@
+import pymysql
+import os
+from dotenv import load_dotenv
+from urllib.parse import urlparse, unquote
+
+# ✅ FIX 1: Class renamed to 'Database' (matches api_server.py)
+class Database:
+    def __init__(self):
+        load_dotenv()
+        db_uri = os.getenv("DB_URI")
+        
+        if not db_uri:
+            raise ValueError("❌ DB_URI is missing from .env file")
+
+        parsed = urlparse(db_uri)
+        self.host = parsed.hostname
+        self.port = parsed.port or 3306
+        self.user = parsed.username
+        self.password = unquote(parsed.password)
+        self.db_name = parsed.path[1:]
+
+    def get_connection(self):
+        return pymysql.connect(
+            host=self.host,
+            user=self.user,
+            password=self.password,
+            database=self.db_name,
+            port=self.port,
+            cursorclass=pymysql.cursors.DictCursor
+        )
+
+    # ✅ FIX 2: Method renamed to 'run_query' (matches api_server.py)
+    def run_query(self, query):
+        conn = self.get_connection()
+        try:
+            with conn.cursor() as cursor:
+                cursor.execute(query)
+                return cursor.fetchall()
+        except Exception as e:
+            return [f"Error: {e}"]
+        finally:
+            conn.close()
+
+    def get_tables(self):
+        """Returns a list of all table names."""
+        conn = self.get_connection()
+        try:
+            with conn.cursor() as cursor:
+                cursor.execute("SHOW TABLES")
+                return [list(row.values())[0] for row in cursor.fetchall()]
+        except Exception as e:
+            return []
+        finally:
+            conn.close()
+
+    def get_table_schema(self, table_name):
+        """Returns column details for a specific table."""
+        conn = self.get_connection()
+        try:
+            with conn.cursor() as cursor:
+                cursor.execute(f"DESCRIBE {table_name}")
+                columns = []
+                for row in cursor.fetchall():
+                    columns.append(f"{row['Field']} ({row['Type']})")
+                return columns
+        except Exception:
+            return []
+        finally:
+            conn.close()
+
+    # ✅ FIX 3: Added 'get_schema()' (no args) for the RAG system
+    def get_schema(self):
+        """Generates a full text schema of the database for the AI."""
+        tables = self.get_tables()
+        schema_text = ""
+        
+        for table in tables:
+            columns = self.get_table_schema(table)
+            schema_text += f"Table: {table}\nColumns:\n"
+            for col in columns:
+                schema_text += f"  - {col}\n"
+            schema_text += "\n"
+            
+        return schema_text

index.html

@@ -0,0 +1,430 @@
+<!DOCTYPE html>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>PlainSQL | Enterprise Data Assistant</title>
+    <script src="https://cdn.tailwindcss.com"></script>
+    <script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
+    <link href="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/themes/prism-tomorrow.min.css" rel="stylesheet" />
+    <script src="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/prism.min.js"></script>
+    <script src="https://cdnjs.cloudflare.com/ajax/libs/prism/1.29.0/components/prism-sql.min.js"></script>
+    <link href="https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600&family=JetBrains+Mono:wght@400;500&display=swap" rel="stylesheet">
+    
+    <script>
+        tailwind.config = {
+            theme: {
+                extend: {
+                    fontFamily: { sans: ['Inter', 'sans-serif'], mono: ['JetBrains Mono', 'monospace'] },
+                    colors: { 
+                        dark: { 900: '#0B0C10', 800: '#15171E', 700: '#1F222E' }, 
+                        brand: { 500: '#38bdf8', 400: '#0ea5e9' } 
+                    }
+                }
+            }
+        }
+    </script>
+    <style>
+        body { background-color: #0B0C10; color: #E2E8F0; }
+        .scrollbar-custom::-webkit-scrollbar { width: 8px; }
+        .scrollbar-custom::-webkit-scrollbar-track { background: #15171E; }
+        .scrollbar-custom::-webkit-scrollbar-thumb { background: #334155; border-radius: 4px; }
+        
+        .bubble-user { background: linear-gradient(135deg, #38bdf8 0%, #0284c7 100%); color: white; border-radius: 16px 16px 4px 16px; }
+        .bubble-ai { background: #1F222E; border: 1px solid #2D3142; border-radius: 16px 16px 16px 4px; }
+        
+        .custom-table th { background: #262A3B; color: #94A3B8; font-size: 0.75rem; padding: 12px; text-transform: uppercase; letter-spacing: 0.05em; }
+        .custom-table td { border-bottom: 1px solid #1F222E; color: #CBD5E1; padding: 12px; font-size: 0.9rem; }
+        .custom-table tr:last-child td { border-bottom: none; }
+        
+        @keyframes fadeIn { from { opacity: 0; transform: translateY(10px); } to { opacity: 1; transform: translateY(0); } }
+        .animate-fade { animation: fadeIn 0.3s ease-out forwards; }
+        .typing-cursor::after { content: '▋'; animation: blink 1s step-start infinite; color: #38bdf8; margin-left: 2px; }
+        @keyframes blink { 50% { opacity: 0; } }
+        
+        .scrollbar-hide::-webkit-scrollbar { display: none; }
+        .scrollbar-hide { -ms-overflow-style: none; scrollbar-width: none; }
+    </style>
+</head>
+<body class="flex h-screen overflow-hidden selection:bg-brand-500 selection:text-white relative font-sans">
+
+    <audio id="sound-welcome" src="https://assets.mixkit.co/active_storage/sfx/2568/2568-preview.mp3" preload="auto"></audio>
+    <audio id="sound-message" src="https://assets.mixkit.co/active_storage/sfx/2346/2346-preview.mp3" preload="auto"></audio>
+
+    <div id="splash-screen" class="fixed inset-0 z-[100] bg-dark-900 flex flex-col items-center justify-center transition-opacity duration-700">
+        <div class="text-center space-y-6 animate-fade">
+            <div class="w-20 h-20 bg-brand-500 rounded-2xl flex items-center justify-center shadow-2xl shadow-brand-500/50 mx-auto mb-6">
+                <svg class="w-10 h-10 text-white" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M4 7v10c0 2.21 3.58 4 8 4s8-1.79 8-4V7M4 7c0 2.21 3.58 4 8 4s8-1.79 8-4M4 7c0-2.21 3.58-4 8-4s8 1.79 8 4m0 5c0 2.21-3.58 4-8 4s-8-1.79-8-4"/></svg>
+            </div>
+            <h1 class="text-4xl font-bold text-white tracking-tight">Plain<span class="text-brand-400">SQL</span></h1>
+            <p class="text-gray-400 text-sm">Enterprise Text-to-SQL Engine</p>
+            <button onclick="enterSystem()" class="mt-8 px-8 py-3 bg-white text-dark-900 font-bold rounded-full hover:bg-gray-100 transition-all shadow-lg hover:scale-105 transform">
+                Initialize System
+            </button>
+        </div>
+    </div>
+
+    <div id="chart-modal" class="fixed inset-0 z-50 flex items-center justify-center bg-black/80 backdrop-blur-sm hidden opacity-0 transition-opacity duration-300">
+        <div class="bg-dark-800 border border-dark-700 w-full max-w-4xl h-[600px] rounded-2xl shadow-2xl flex flex-col p-6 transform scale-95 transition-transform duration-300" id="chart-content">
+            <div class="flex justify-between items-center mb-4">
+                <h3 class="text-lg font-semibold text-white">Data Visualization</h3>
+                <button onclick="closeChart()" class="text-gray-500 hover:text-white p-2 text-xl">✕</button>
+            </div>
+            <div class="flex-1 relative w-full h-full"><canvas id="myChart"></canvas></div>
+        </div>
+    </div>
+
+    <aside class="w-64 bg-dark-800 border-r border-dark-700 flex flex-col hidden md:flex">
+        <div class="p-6 flex items-center gap-3">
+            <div class="w-8 h-8 bg-brand-500 rounded-lg flex items-center justify-center shadow-lg">
+                <svg class="w-5 h-5 text-white" fill="none" stroke="currentColor" viewBox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M4 7v10c0 2.21 3.58 4 8 4s8-1.79 8-4V7M4 7c0 2.21 3.58 4 8 4s8-1.79 8-4M4 7c0-2.21 3.58-4 8-4s8 1.79 8 4m0 5c0 2.21-3.58 4-8 4s-8-1.79-8-4"/></svg>
+            </div>
+            <span class="font-bold text-xl tracking-tight">PlainSQL</span>
+        </div>
+        <nav class="flex-1 px-4 space-y-2 overflow-y-auto scrollbar-custom">
+            <button onclick="window.location.reload()" class="w-full flex items-center gap-3 px-4 py-3 bg-brand-500/10 text-brand-400 rounded-xl border border-brand-500/20 hover:bg-brand-500/20 transition-all group">
+                <svg class="w-5 h-5 group-hover:rotate-90 transition-transform" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M12 6v6m0 0v6m0-6h6m-6 0H6"/></svg>
+                <span class="text-sm font-medium">New Analysis</span>
+            </button>
+            <div class="mt-8">
+                <p class="px-4 text-xs font-semibold text-gray-500 uppercase tracking-wider mb-2">Recent Queries</p>
+                <div id="history-list" class="space-y-1"></div>
+            </div>
+        </nav>
+        <div class="p-4 border-t border-dark-700">
+             <div class="flex items-center gap-3 px-3 py-2 bg-green-500/10 rounded-lg border border-green-500/20">
+                <div class="w-2 h-2 bg-green-500 rounded-full animate-pulse"></div>
+                <div class="flex flex-col">
+                    <span class="text-[11px] font-bold text-green-400 uppercase">System Online</span>
+                    <span class="text-[10px] text-gray-500">Read-Only Mode</span>
+                </div>
+            </div>
+        </div>
+    </aside>
+
+    <main class="flex-1 flex flex-col relative bg-dark-900">
+        <div id="chat-box" class="flex-1 overflow-y-auto p-4 md:p-8 space-y-6 pb-80 scroll-smooth scrollbar-custom">
+            <div class="flex gap-4 max-w-3xl mx-auto animate-fade">
+                <div class="w-8 h-8 rounded-full bg-brand-500/20 flex items-center justify-center flex-shrink-0 border border-brand-500/30 text-brand-400">
+                    <svg class="w-5 h-5" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M13 10V3L4 14h7v7l9-11h-7z"/></svg>
+                </div>
+                <div class="bubble-ai p-5 shadow-sm">
+                    <p class="text-sm leading-relaxed text-gray-200">
+                        Hello. I am <strong>PlainSQL</strong>. I can access your database securely to fetch real-time insights.<br><br>
+                        <em>Try asking: "Show me top 5 employees by salary" or "List all active users."</em>
+                    </p>
+                </div>
+            </div>
+        </div>
+
+        <div class="absolute bottom-0 w-full p-4 md:p-6 bg-gradient-to-t from-dark-900 via-dark-900 to-transparent z-20">
+            <div class="max-w-3xl mx-auto">
+                <div id="suggestions" class="flex gap-2 mb-3 overflow-x-auto pb-1 scrollbar-hide"></div>
+                <form id="chat-form" class="relative group">
+                    <div class="absolute inset-0 bg-brand-500/20 rounded-2xl blur-lg group-hover:bg-brand-500/30 transition-all opacity-0 group-hover:opacity-100"></div>
+                    <input type="text" id="user-input" 
+                        class="relative w-full bg-dark-800 text-white border border-dark-700 rounded-2xl py-4 pl-5 pr-14 focus:outline-none focus:border-brand-500 focus:ring-1 focus:ring-brand-500 transition-all placeholder-gray-500 shadow-xl" 
+                        placeholder="Ask a question in plain English..." autocomplete="off">
+                    <button type="submit" class="absolute right-2 top-2 p-2 bg-brand-500 hover:bg-brand-400 text-white rounded-xl transition-all shadow-lg active:scale-95">
+                        <svg class="w-5 h-5" fill="none" stroke="currentColor" viewBox="0 0 24 24"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M14 5l7 7m0 0l-7 7m7-7H3"/></svg>
+                    </button>
+                </form>
+                <div class="text-center mt-2">
+                    <p class="text-[10px] text-gray-600">AI Generated SQL can be inaccurate. Always verify important data.</p>
+                </div>
+            </div>
+        </div>
+    </main>
+
+    <script>
+        const API_URL = "http://127.0.0.1:8000/chat"; 
+        
+        const form = document.getElementById('chat-form');
+        const input = document.getElementById('user-input');
+        const chatBox = document.getElementById('chat-box');
+        const suggestionsBox = document.getElementById('suggestions');
+        const historyList = document.getElementById('history-list');
+        const soundWelcome = document.getElementById('sound-welcome');
+        const soundMessage = document.getElementById('sound-message');
+        const splashScreen = document.getElementById('splash-screen');
+        const modal = document.getElementById('chart-modal');
+        const modalContent = document.getElementById('chart-content');
+        
+        let chartInstance = null;
+        let conversationHistory = [];
+
+        function unlockAudio() {
+            soundMessage.volume = 0;
+            soundMessage.play().then(() => {
+                soundMessage.pause();
+                soundMessage.currentTime = 0;
+            }).catch(() => {});
+        }
+
+        function playIncomingSound() {
+            soundMessage.volume = 0.4;
+            soundMessage.currentTime = 0;
+            soundMessage.play().catch(e => console.warn("Audio blocked:", e));
+        }
+
+        function enterSystem() {
+            soundWelcome.volume = 0.5;
+            soundWelcome.play().catch(e => console.log("Init Audio Error:", e));
+            splashScreen.style.opacity = '0';
+            setTimeout(() => { splashScreen.style.display = 'none'; }, 700);
+            setTimeout(() => input.focus(), 800);
+        }
+
+        function checkGreeting(text) {
+            const t = text.toLowerCase();
+            const greetings = ['hello', 'hi', 'hey', 'good morning', 'good afternoon', 'hola'];
+            
+            if (greetings.some(g => t === g || t.startsWith(g + ' '))) {
+                return "Hello! 👋 I am **PlainSQL**, your data assistant. I'm here to help you query your database without writing code.<br><br>You can ask me things like: <em>'Show me top 5 employees by salary'</em> or <em>'List all active users.'</em>";
+            }
+            return null;
+        }
+
+        // FIX 2: IMPROVED SCROLL FUNCTION
+        // Uses setTimeout to ensure DOM is fully rendered before scrolling
+        function scrollToBottom() { 
+            setTimeout(() => {
+                chatBox.scrollTop = chatBox.scrollHeight;
+            }, 50);
+        }
+
+        async function typeText(element, text) {
+            element.classList.add('typing-cursor');
+            return new Promise(resolve => {
+                setTimeout(() => {
+                    element.innerHTML = text;
+                    element.classList.remove('typing-cursor');
+                    resolve();
+                }, 300 + Math.random() * 500);
+            });
+        }
+
+        function showLoading() {
+            const id = 'loading-' + Date.now();
+            const html = `
+                <div id="${id}" class="flex gap-4 max-w-3xl mx-auto animate-fade">
+                    <div class="w-8 h-8 rounded-full bg-brand-500/20 flex items-center justify-center text-brand-400">
+                        <svg class="w-4 h-4 animate-spin" fill="none" viewBox="0 0 24 24"><circle class="opacity-25" cx="12" cy="12" r="10" stroke="currentColor" stroke-width="4"></circle><path class="opacity-75" fill="currentColor" d="M4 12a8 8 0 018-8V0C5.373 0 0 5.373 0 12h4zm2 5.291A7.962 7.962 0 014 12H0c0 3.042 1.135 5.824 3 7.938l3-2.647z"></path></svg>
+                    </div>
+                    <div class="bubble-ai p-4 flex gap-2 items-center">
+                        <span class="text-xs text-gray-400">Analyzing database schema...</span>
+                    </div>
+                </div>`;
+            chatBox.insertAdjacentHTML('beforeend', html);
+            scrollToBottom();
+            return id;
+        }
+
+        async function appendMessage(role, content, isHtml = false) {
+            const wrapper = document.createElement('div');
+            wrapper.className = `flex gap-4 max-w-3xl mx-auto animate-fade ${role === 'user' ? 'justify-end' : ''}`;
+            
+            const aiIcon = `<div class="w-8 h-8 rounded-full bg-brand-500/20 flex items-center justify-center flex-shrink-0 border border-brand-500/30 text-brand-400"><svg class="w-5 h-5" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M13 10V3L4 14h7v7l9-11h-7z"/></svg></div>`;
+            const userIcon = `<div class="w-8 h-8 rounded-full bg-gray-700 flex items-center justify-center flex-shrink-0 text-gray-300"><svg class="w-5 h-5" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M16 7a4 4 0 11-8 0 4 4 0 018 0zM12 14a7 7 0 00-7 7h14a7 7 0 00-7-7z"/></svg></div>`;
+            
+            const bubble = document.createElement('div');
+            bubble.className = role === 'user' ? 'bubble-user p-4 max-w-[85%] shadow-lg text-sm' : 'bubble-ai p-5 max-w-[90%] shadow-sm w-full text-sm leading-relaxed';
+
+            if (role === 'user') {
+                bubble.textContent = content;
+                wrapper.innerHTML = bubble.outerHTML + userIcon;
+            } else {
+                wrapper.innerHTML = aiIcon + bubble.outerHTML;
+            }
+            chatBox.appendChild(wrapper);
+            
+            if (role === 'ai') {
+                playIncomingSound();
+                if (isHtml) {
+                    wrapper.querySelector('.bubble-ai').innerHTML = content;
+                    Prism.highlightAll(); 
+                } else {
+                    await typeText(wrapper.querySelector('.bubble-ai'), content);
+                }
+            }
+            scrollToBottom();
+        }
+
+        function renderSuggestions(questions) {
+            suggestionsBox.innerHTML = '';
+            if (!questions) return;
+            questions.forEach((q, i) => {
+                const btn = document.createElement('button');
+                btn.className = "whitespace-nowrap px-4 py-1.5 bg-dark-800 border border-dark-700 hover:border-brand-500 hover:text-brand-400 text-gray-400 text-xs font-medium rounded-full transition-all animate-fade";
+                btn.style.animationDelay = `${i * 0.1}s`;
+                btn.textContent = q;
+                btn.onclick = () => { input.value = q; form.dispatchEvent(new Event('submit')); };
+                suggestionsBox.appendChild(btn);
+            });
+            scrollToBottom();
+        }
+
+        function addToHistory(question) {
+            const btn = document.createElement('button');
+            btn.className = "w-full text-left px-4 py-2 text-xs text-gray-400 hover:text-white hover:bg-dark-700 rounded-lg transition-colors truncate animate-fade";
+            btn.textContent = question;
+            btn.onclick = () => { input.value = question; form.dispatchEvent(new Event('submit')); };
+            historyList.prepend(btn);
+        }
+
+        // FIX 3: EXPORT CSV FUNCTION
+        function downloadCSV(dataString) {
+            try {
+                const data = JSON.parse(decodeURIComponent(dataString));
+                if (!data || !data.length) return;
+                
+                const headers = Object.keys(data[0]);
+                const csvRows = [];
+                csvRows.push(headers.join(','));
+                
+                for (const row of data) {
+                    const values = headers.map(header => {
+                        const escaped = ('' + row[header]).replace(/"/g, '\\"');
+                        return `"${escaped}"`;
+                    });
+                    csvRows.push(values.join(','));
+                }
+                
+                const blob = new Blob([csvRows.join('\n')], { type: 'text/csv' });
+                const url = window.URL.createObjectURL(blob);
+                const a = document.createElement('a');
+                a.setAttribute('hidden', '');
+                a.setAttribute('href', url);
+                a.setAttribute('download', 'data_export.csv');
+                document.body.appendChild(a);
+                a.click();
+                document.body.removeChild(a);
+            } catch (e) {
+                console.error("Export failed", e);
+            }
+        }
+
+        form.addEventListener('submit', async (e) => {
+            e.preventDefault();
+            const question = input.value.trim();
+            if (!question) return;
+
+            unlockAudio();
+            suggestionsBox.innerHTML = ''; 
+            input.value = '';
+
+            await appendMessage('user', question);
+            addToHistory(question);
+
+            const greetingResponse = checkGreeting(question);
+            if (greetingResponse) {
+                setTimeout(() => appendMessage('ai', greetingResponse, true), 500);
+                return;
+            }
+
+            const loadingId = showLoading();
+
+            try {
+                const res = await fetch(`${API_URL}?ts=${Date.now()}`, {
+                    method: 'POST',
+                    headers: { 'Content-Type': 'application/json' },
+                    body: JSON.stringify({ question, history: conversationHistory }) 
+                });
+
+                if (!res.ok) throw new Error("Backend connection failed.");
+
+                const data = await res.json();
+                document.getElementById(loadingId).remove();
+
+                if (data.sql && !data.sql.includes("Error")) {
+                    conversationHistory.push({ user: question, sql: data.sql });
+                    if(conversationHistory.length > 5) conversationHistory.shift();
+                }
+
+                let content = "";
+                if (data.message) content += `<div class="mb-4">${data.message}</div>`;
+
+                if (Array.isArray(data.answer) && data.answer.length > 0) {
+                    const firstRow = data.answer[0];
+                    if (typeof firstRow === 'string' && (firstRow.toLowerCase().includes("error"))) {
+                        content += `<div class="p-4 bg-red-500/10 border border-red-500/20 rounded-lg text-red-400 font-mono text-xs mb-3">⚠️ ${firstRow}</div>`;
+                    } else {
+                        const headers = Object.keys(firstRow);
+                        const dataStr = encodeURIComponent(JSON.stringify(data.answer));
+                        
+                        content += `
+                            <div class="overflow-hidden rounded-xl border border-dark-700 shadow-xl mb-3 bg-[#15171E]">
+                                <div class="overflow-x-auto">
+                                    <table class="w-full text-left custom-table">
+                                        <thead><tr>${headers.map(h => `<th>${h}</th>`).join('')}</tr></thead>
+                                        <tbody>${data.answer.map(row => `<tr>${headers.map(h => `<td>${row[h]}</td>`).join('')}</tr>`).join('')}</tbody>
+                                    </table>
+                                </div>
+                            </div>
+                            <div class="flex gap-2 mb-4">
+                                <button onclick="openChart('${dataStr}')" class="flex items-center gap-2 px-3 py-2 bg-brand-500 text-white text-xs font-bold rounded-lg hover:bg-brand-400 transition-colors">
+                                    <svg class="w-4 h-4" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M11 3.055A9.001 9.001 0 1020.945 13H11V3.055z"/></svg>
+                                    Visualize
+                                </button>
+                                <button onclick="downloadCSV('${dataStr}')" class="flex items-center gap-2 px-3 py-2 bg-dark-700 border border-dark-600 text-gray-300 text-xs font-bold rounded-lg hover:bg-dark-600 transition-colors">
+                                    <svg class="w-4 h-4" fill="none" viewBox="0 0 24 24" stroke="currentColor"><path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M4 16v1a3 3 0 003 3h10a3 3 0 003-3v-1m-4-4l-4 4m0 0l-4-4m4 4V4"/></svg>
+                                    Export CSV
+                                </button>
+                            </div>
+                        `;
+                    }
+                } else if (Array.isArray(data.answer) && data.answer.length === 0) {
+                    content += `<div class="p-4 bg-yellow-500/5 border border-yellow-500/10 rounded-lg text-yellow-500/80 text-xs mb-3">No records found matching query.</div>`;
+                }
+
+                if (data.sql) {
+                    content += `
+                        <div class="relative group mt-2">
+                            <div class="absolute -top-3 left-3 bg-dark-700 px-2 text-[10px] text-gray-400 rounded border border-dark-600">Generated SQL</div>
+                            <pre class="!m-0 !p-4 !bg-[#0d1117] !text-xs rounded-xl border border-dark-700/50"><code class="language-sql">${data.sql}</code></pre>
+                        </div>`;
+                }
+
+                if (content) await appendMessage('ai', content, true);
+                if (data.follow_ups) renderSuggestions(data.follow_ups);
+
+            } catch (err) {
+                document.getElementById(loadingId)?.remove();
+                await appendMessage('ai', `<div class="p-4 bg-red-900/20 border border-red-500/30 rounded-lg text-red-400 text-sm">Error: ${err.message}</div>`, true);
+            }
+        });
+
+        function openChart(dataString) {
+            const data = JSON.parse(decodeURIComponent(dataString));
+            const headers = Object.keys(data[0]);
+            let labelKey = headers.find(h => isNaN(data[0][h])) || headers[0];
+            let valueKey = headers.find(h => !isNaN(data[0][h]) && h !== labelKey) || headers[1];
+            const labels = data.map(row => row[labelKey]);
+            const values = data.map(row => row[valueKey]);
+
+            modal.classList.remove('hidden');
+            setTimeout(() => { modal.classList.remove('opacity-0'); modalContent.classList.remove('scale-95'); modalContent.classList.add('scale-100'); }, 10);
+            if (chartInstance) chartInstance.destroy();
+
+            const ctx = document.getElementById('myChart').getContext('2d');
+            chartInstance = new Chart(ctx, {
+                type: labels.length > 8 ? 'bar' : 'doughnut',
+                data: {
+                    labels: labels,
+                    datasets: [{
+                        label: valueKey.toUpperCase(),
+                        data: values,
+                        backgroundColor: ['#38bdf8', '#a855f7', '#ec4899', '#22c55e', '#eab308'],
+                        borderColor: '#15171E', borderWidth: 2
+                    }]
+                },
+                options: { responsive: true, maintainAspectRatio: false, plugins: { legend: { position: 'bottom', labels: { color: '#94A3B8' } } }, scales: { x: { display: false }, y: { display: false } } }
+            });
+        }
+
+        function closeChart() {
+            modal.classList.add('opacity-0'); modalContent.classList.remove('scale-100'); modalContent.classList.add('scale-95'); setTimeout(() => { modal.classList.add('hidden'); }, 300);
+        }
+    </script>
+</body>
+</html>

rag_manager.py

@@ -0,0 +1,66 @@
+import chromadb
+from src.db_connector import Database
+
+class RAGSystem:
+    def __init__(self, db_instance=None):
+        # ✅ FIX: Accept the DB connection passed from the server
+        if db_instance:
+            self.db = db_instance
+        else:
+            self.db = Database()
+        
+        # Initialize ChromaDB
+        print("   ...Connecting to ChromaDB")
+        self.client = chromadb.PersistentClient(path="./chroma_db")
+        self.collection = self.client.get_or_create_collection(name="schema_knowledge")
+        
+        # Refresh memory
+        self._index_schema()
+
+    def _index_schema(self):
+        """Reads the database structure and saves it to ChromaDB."""
+        try:
+            tables = self.db.get_tables()
+            
+            if self.collection.count() > 0:
+                existing_ids = self.collection.get()['ids']
+                if existing_ids:
+                    self.collection.delete(ids=existing_ids)
+
+            for table in tables:
+                columns = self.db.get_table_schema(table)
+                col_list = []
+                for col in columns:
+                    if isinstance(col, dict):
+                        col_list.append(f"{col['name']} ({col['type']})")
+                    else:
+                        col_list.append(str(col))
+                
+                schema_text = f"Table: {table}\nColumns: {', '.join(col_list)}"
+                self.collection.add(
+                    documents=[schema_text],
+                    metadatas=[{"table": table}],
+                    ids=[table]
+                )
+            print(f"   ✅ RAG System: Indexed {len(tables)} tables.")
+            
+        except Exception as e:
+            print(f"   ⚠️ RAG Indexing Warning: {e}")
+
+    def get_relevant_schema(self, question):
+        try:
+            results = self.collection.query(query_texts=[question], n_results=3)
+            if results['documents']:
+                return "\n\n".join(results['documents'][0])
+            return ""
+        except Exception:
+            return self._get_full_schema_fallback()
+
+    def _get_full_schema_fallback(self):
+        tables = self.db.get_tables()
+        schema = []
+        for table in tables:
+            cols = self.db.get_table_schema(table)
+            col_list = [c['name'] if isinstance(c, dict) else str(c) for c in cols]
+            schema.append(f"Table: {table}\nColumns: {', '.join(col_list)}")
+        return "\n\n".join(schema)

requirements.txt

@@ -0,0 +1,21 @@
+# Core Framework
+langchain==0.3.0
+langchain-community
+langchain-core
+
+# Google Gemini Integration
+langchain-google-genai
+google-generativeai
+
+# Database & Vector Store
+sqlalchemy
+pymysql
+chromadb
+langchain-chroma
+
+# Utilities
+python-dotenv
+streamlit
+
+# Critical Version Fixes
+numpy<2.0.0

setup_full_db.py

@@ -0,0 +1,105 @@
+import pymysql
+import random
+from faker import Faker
+from dotenv import load_dotenv
+import os
+from urllib.parse import urlparse, unquote  # <--- Added 'unquote' here
+
+# --- CONFIGURATION ---
+NUM_EMPLOYEES = 50
+NUM_CUSTOMERS = 100
+NUM_PRODUCTS = 20
+NUM_SALES = 1000
+
+# Load credentials
+load_dotenv()
+db_uri = os.getenv("DB_URI")
+
+# Parse URI
+parsed = urlparse(db_uri)
+username = parsed.username
+# FIX: 'unquote' converts 'Lalit%40851' back to 'Lalit@851'
+password = unquote(parsed.password) 
+host = parsed.hostname
+port = parsed.port
+# FIX: Handle cases where path is empty or just slash
+dbname = parsed.path[1:] if parsed.path else "chatbot"
+
+print(f"--- 🏭 INITIALIZING BUSINESS SIMULATOR for DB: {dbname} ---")
+
+try:
+    # Connect without selecting a DB first (to create it if missing)
+    conn = pymysql.connect(host=host, user=username, password=password, port=port)
+    cursor = conn.cursor()
+
+    # 1. CREATE DATABASE & TABLES
+    print("1. Rebuilding Schema...")
+    cursor.execute(f"CREATE DATABASE IF NOT EXISTS {dbname}")
+    cursor.execute(f"USE {dbname}")
+
+    cursor.execute("SET FOREIGN_KEY_CHECKS = 0")
+    for t in ["sales", "employees", "products", "customers", "departments"]:
+        cursor.execute(f"DROP TABLE IF EXISTS {t}")
+    cursor.execute("SET FOREIGN_KEY_CHECKS = 1")
+
+    queries = [
+        "CREATE TABLE departments (id INT AUTO_INCREMENT PRIMARY KEY, name VARCHAR(50) UNIQUE, budget DECIMAL(15, 2), location VARCHAR(100))",
+        "CREATE TABLE employees (id INT AUTO_INCREMENT PRIMARY KEY, name VARCHAR(100), email VARCHAR(100), department_id INT, role VARCHAR(50), salary DECIMAL(10, 2), hire_date DATE, FOREIGN KEY (department_id) REFERENCES departments(id))",
+        "CREATE TABLE products (id INT AUTO_INCREMENT PRIMARY KEY, name VARCHAR(100), category VARCHAR(50), price DECIMAL(10, 2), stock_quantity INT)",
+        "CREATE TABLE customers (id INT AUTO_INCREMENT PRIMARY KEY, name VARCHAR(100), company VARCHAR(100), region VARCHAR(50), join_date DATE)",
+        "CREATE TABLE sales (id INT AUTO_INCREMENT PRIMARY KEY, employee_id INT, customer_id INT, product_id INT, quantity INT, total_amount DECIMAL(10, 2), sale_date DATE, FOREIGN KEY (employee_id) REFERENCES employees(id), FOREIGN KEY (customer_id) REFERENCES customers(id), FOREIGN KEY (product_id) REFERENCES products(id))"
+    ]
+
+    for q in queries:
+        cursor.execute(q)
+
+    # 2. GENERATE DATA
+    fake = Faker()
+    print("2. Manufacturing Data...")
+
+    # Departments
+    depts = ["Sales", "Engineering", "HR", "Marketing", "Executive"]
+    dept_ids = []
+    for d in depts:
+        cursor.execute("INSERT INTO departments (name, budget, location) VALUES (%s, %s, %s)", (d, random.randint(50000, 1000000), fake.city()))
+        dept_ids.append(cursor.lastrowid)
+
+    # Employees
+    emp_ids = []
+    roles = ["Manager", "Associate", "Analyst", "Director", "Intern"]
+    for _ in range(NUM_EMPLOYEES):
+        cursor.execute("INSERT INTO employees (name, email, department_id, role, salary, hire_date) VALUES (%s, %s, %s, %s, %s, %s)", 
+                       (fake.name(), fake.email(), random.choice(dept_ids), random.choice(roles), random.randint(40000, 150000), fake.date_between(start_date='-5y', end_date='today')))
+        emp_ids.append(cursor.lastrowid)
+
+    # Products
+    prod_ids = []
+    for _ in range(NUM_PRODUCTS):
+        cursor.execute("INSERT INTO products (name, category, price, stock_quantity) VALUES (%s, %s, %s, %s)", 
+                       (fake.bs().title(), random.choice(["Software", "Hardware", "Service"]), round(random.uniform(50, 5000), 2), random.randint(0, 500)))
+        prod_ids.append(cursor.lastrowid)
+
+    # Customers
+    cust_ids = []
+    for _ in range(NUM_CUSTOMERS):
+        cursor.execute("INSERT INTO customers (name, company, region, join_date) VALUES (%s, %s, %s, %s)", 
+                       (fake.name(), fake.company(), random.choice(["North America", "Europe", "Asia", "South America"]), fake.date_between(start_date='-3y', end_date='today')))
+        cust_ids.append(cursor.lastrowid)
+
+    # Sales
+    print(f"   -> Generating {NUM_SALES} sales transactions...")
+    for _ in range(NUM_SALES):
+        prod = random.choice(prod_ids)
+        cursor.execute("SELECT price FROM products WHERE id=%s", (prod,))
+        price = cursor.fetchone()[0]
+        qty = random.randint(1, 10)
+        cursor.execute("INSERT INTO sales (employee_id, customer_id, product_id, quantity, total_amount, sale_date) VALUES (%s, %s, %s, %s, %s, %s)", 
+                       (random.choice(emp_ids), random.choice(cust_ids), prod, qty, price * qty, fake.date_between(start_date='-1y', end_date='today')))
+
+    conn.commit()
+    conn.close()
+    print("✅ DONE! Database is populated.")
+
+except Exception as e:
+    print(f"\n❌ CRITICAL ERROR: {e}")
+    print("Double check your password in .env!")

sql_generator.py

@@ -0,0 +1,76 @@
+import os
+import re
+import json
+from huggingface_hub import InferenceClient
+from dotenv import load_dotenv
+
+class SQLGenerator:
+    def __init__(self, api_key=None):
+        load_dotenv()
+        self.api_token = os.getenv("HUGGINGFACEHUB_API_TOKEN")
+        self.repo_id = "Qwen/Qwen2.5-Coder-32B-Instruct"
+        self.client = InferenceClient(token=self.api_token, timeout=25.0)
+
+    def generate_followup_questions(self, question, sql_query):
+        return ["Visualize this result", "Export as CSV", "Compare with last year"]
+
+    def generate_sql(self, question, context, history=None):
+        if history is None: history = []
+
+        forbidden = ["DROP", "DELETE", "UPDATE", "INSERT", "ALTER", "TRUNCATE", "GRANT"]
+        if any(word in question.upper() for word in forbidden):
+             return "SELECT 'Error: Blocked by Safety Layer' as status", "Safety Alert", "I cannot execute commands that modify data."
+
+        history_text = ""
+        if history:
+            history_text = "PREVIOUS CONVERSATION:\n" + "\n".join([f"User: {h['user']}\nSQL: {h['sql']}" for h in history[-2:]])
+
+        system_prompt = f"""You are an elite SQL Expert.
+Schema:
+{context}
+
+{history_text}
+
+Rules:
+1. Output JSON: {{ "sql": "SELECT ...", "message": "Friendly text", "explanation": "Brief summary" }}
+2. Query MUST be Read-Only (SELECT).
+3. Do not include markdown formatting like ```json.
+"""
+        messages = [{"role": "system", "content": system_prompt}, {"role": "user", "content": question}]
+
+        try:
+            print(f"   ⚡ Generating SQL...")
+            response = self.client.chat_completion(messages=messages, model=self.repo_id, max_tokens=1024, temperature=0.1)
+            raw_text = response.choices[0].message.content
+            
+            sql_query = ""
+            message = "Here is the data."
+            explanation = "Query generated successfully."
+
+            try:
+                clean_json = re.sub(r"```json|```", "", raw_text).strip()
+                data = json.loads(clean_json)
+                sql_query = data.get("sql", "")
+                message = data.get("message", message)
+                explanation = data.get("explanation", explanation)
+            except:
+                match = re.search(r"(SELECT[\s\S]+?;)", raw_text, re.IGNORECASE)
+                if match: sql_query = match.group(1)
+
+            sql_query = sql_query.strip().replace("\n", " ")
+            if sql_query and not sql_query.endswith(";"): sql_query += ";"
+            
+            # ✅ FIX: Strip comments and whitespace before validation
+            clean_check = re.sub(r"/\*.*?\*/|--.*?\n", "", sql_query, flags=re.DOTALL).strip().upper()
+
+            # ✅ FIX: Allow SELECT or WITH clauses
+            if not clean_check.startswith("SELECT") and not clean_check.startswith("WITH"):
+                print(f"   ⚠️ Invalid SQL Blocked: {sql_query}")
+                return "SELECT 'Error: Invalid Query Type (Non-SELECT)' as status", "Safety Error", "I can only perform read-only operations."
+
+            return sql_query, explanation, message
+
+        except Exception as e:
+            print(f"   ❌ Model Error: {e}")
+            safe_e = str(e).replace("'", "").replace('"', "")
+            return f"SELECT 'Error: {safe_e}' as status", "System Error", "An unexpected error occurred."