Upload 7 files

app.py

@@ -0,0 +1,1485 @@
+# -*- coding: utf-8 -*-
+"""app.ipynb
+
+Automatically generated by Colab.
+
+Original file is located at
+    https://colab.research.google.com/drive/1e0kcqimC-jHT2w8MwJMSXxzS-wJh1jYS
+"""
+
+import json
+import math
+import os
+import random
+import heapq
+import re
+import sqlite3
+import gradio as gr
+import numpy as np
+import matplotlib.pyplot as plt
+from scipy.ndimage import label, distance_transform_edt
+from gtts import gTTS
+from pydub import AudioSegment
+import speech_recognition as sr
+import openai
+
+CATEGORIES_FILE = "categories.json"
+PRODUCTS_FILE = "sort_data.json"
+DB_PATH = "mall.db"
+# Smart Mall Assistant
+# Navigation + chatbot + recommender system merged into one Colab-ready script.
+
+
+# ============================================================
+# CONFIG
+# ============================================================
+CELL_SIZE = 0.2
+MEMORY_FILE = "navigation_memory.json"
+GRID_0_FILE = "floor_0_grid.npy"
+GRID_1_FILE = "floor_0_grid.npy"
+
+OPENROUTER_KEY =os.getenv("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")
+OPENROUTER_MODEL = "arcee-ai/trinity-large-preview:free"
+
+# ============================================================
+# DATA LOADING
+# ============================================================
+def load_json_file(path, default):
+    try:
+        with open(path, "r", encoding="utf-8") as f:
+            return json.load(f)
+    except Exception:
+        return default
+
+categories_data = load_json_file(CATEGORIES_FILE, [])
+products_data = load_json_file(PRODUCTS_FILE, [])
+
+# ============================================================
+# STORE / ROOM MAPPING
+# ============================================================
+ROOM_INFO = {
+    350: "Smart Devices Hub",
+    351: "Gaming & Accessories Hub",
+    352: "Computer Systems Hub",
+    353: "Mobile & Tablets Hub",
+    354: "Health & Personal Care",
+    355: "Bedroom",
+    356: "Living Room",
+    357: "Study/Office",
+    358: "Kitchen",
+    450: "Dining Room",
+    451: "Bathroom",
+    452: "Furnishings",
+    453: "Kitchen and Dining",
+    454: "Home Decor",
+    455: "Tools and utility",
+    456: "Lighting and Electricals",
+    457: "Cleaning and Bath",
+    458: "Pet and Gardening",
+}
+
+STORE_CLUSTER_ROOM = {
+    "Smart Devices Hub": 350,
+    "Gaming & Accessories Hub": 351,
+    "Computer Systems Hub": 352,
+    "Mobile & Tablets Hub": 353,
+    "Health & Personal Care": 354,
+    "Bedroom": 355,
+    "Living Room": 356,
+    "Study/Office": 357,
+    "Kitchen": 358,
+    "Dining Room": 450,
+    "Bathroom": 451,
+    "Furnishings": 452,
+    "Kitchen and Dining": 453,
+    "Home Decor": 454,
+    "Tools and utility": 455,
+    "Lighting and Electricals": 456,
+    "Cleaning and Bath": 457,
+    "Pet and Gardening": 458,
+}
+
+STORE_CLUSTER = {
+    "Audio": "Smart Devices Hub",
+    "Cameras": "Smart Devices Hub",
+    "Smart Home Automation": "Smart Devices Hub",
+    "Smart Wearables": "Smart Devices Hub",
+    "Gaming": "Gaming & Accessories Hub",
+    "Laptop Accessories": "Gaming & Accessories Hub",
+    "Computer Peripherals": "Computer Systems Hub",
+    "Laptop & Desktop": "Computer Systems Hub",
+    "Storage": "Computer Systems Hub",
+    "Mobiles": "Mobile & Tablets Hub",
+    "Tablets": "Mobile & Tablets Hub",
+}
+
+SUB_TO_CATEGORY = {}
+for type_block in categories_data:
+    for cat in type_block.get("categories", []):
+        for sub in cat.get("subCategories", []):
+            SUB_TO_CATEGORY[sub] = cat["category"]
+
+NAME_TO_ROOM = {name.lower(): room for room, name in ROOM_INFO.items()}
+
+# ============================================================
+# GRID / CENTROIDS
+# ============================================================
+grid_0 = np.load(GRID_0_FILE)
+grid_1 = np.load(GRID_1_FILE)
+GRID_SHAPE = grid_0.shape
+
+room_centroids_f3_grid = {
+    350: (70, 465),
+    351: (70, 435),
+    352: (70, 395),
+    353: (70, 360),
+    354: (70, 320),
+    355: (70, 285),
+    356: (70, 215),
+    357: (70, 160),
+    358: (70, 85),
+}
+
+room_numbers_f4 = list(range(450, 459))
+room_centroids_f4_grid = {}
+for old, new in zip(room_centroids_f3_grid.keys(), room_numbers_f4):
+    room_centroids_f4_grid[new] = room_centroids_f3_grid[old]
+
+
+def apply_room_clearance():
+    for r, c in room_centroids_f3_grid.values():
+        for dr in [-1, 0, 1]:
+            for dc in [-1, 0, 1]:
+                rr, cc = r + dr, c + dc
+                if 0 <= rr < GRID_SHAPE[0] and 0 <= cc < GRID_SHAPE[1]:
+                    grid_0[rr, cc] = 0
+                    grid_1[rr, cc] = 0
+
+
+apply_room_clearance()
+
+
+def reset_grids():
+    global grid_0, grid_1, GRID_SHAPE
+    grid_0 = np.load(GRID_0_FILE)
+    grid_1 = np.load(GRID_1_FILE)
+    GRID_SHAPE = grid_0.shape
+    apply_room_clearance()
+
+# ============================================================
+# MEMORY LOG
+# ============================================================
+def load_memory():
+    try:
+        with open(MEMORY_FILE, "r", encoding="utf-8") as f:
+            return json.load(f)
+    except Exception:
+        return []
+
+
+def save_memory(memory):
+    with open(MEMORY_FILE, "w", encoding="utf-8") as f:
+        json.dump(memory, f, indent=4)
+
+
+user_memory = load_memory()
+navigation_counter = len(user_memory) + 1
+
+
+def save_navigation_mem(start_room, dest_room):
+    global navigation_counter
+    entry = {
+        "nav_id": f"N{navigation_counter}",
+        "start_room": start_room,
+        "dest_room": dest_room,
+    }
+    user_memory.append(entry)
+    save_memory(user_memory)
+    navigation_counter += 1
+    return entry
+
+# ============================================================
+# OPENROUTER CLIENT
+# ============================================================
+client = openai.OpenAI(
+    base_url="https://openrouter.ai/api/v1",
+    api_key=OPENROUTER_KEY,
+)
+
+SYSTEM_PROMPT = """
+You are an AI assistant for indoor navigation inside a shopping mall.
+
+Mall stores:
+350 Smart Devices Hub
+351 Gaming & Accessories Hub
+352 Computer Systems Hub
+353 Mobile & Tablets Hub
+354 Health & Personal Care
+355 Bedroom
+356 Living Room
+357 Study/Office
+358 Kitchen
+450 Dining Room
+451 Bathroom
+452 Furnishings
+453 Kitchen and Dining
+454 Home Decor
+455 Tools and utility
+456 Lighting and Electricals
+457 Cleaning and Bath
+458 Pet and Gardening
+
+Rules:
+- If the user asks about a store, answer with store information.
+- If the user asks for navigation, help them navigate.
+- If the user says they are hungry, suggest kitchen / dining / food-related stores.
+- Keep answers concise and helpful.
+"""
+
+# ============================================================
+# HELPERS / SESSION STATE
+# ============================================================
+def set_session_defaults(session):
+    session = session or {}
+    defaults = {
+        "username": None,
+        "gender": "other",
+        "low": 500,
+        "high": 5000,
+        "is_new": True,
+        "start_floor": None,
+        "start_xy": None,
+        "nav_mode": "Normal",
+        "recommended_room": None,
+        "nav_target_room": None,
+        "last_dest_room": None,
+        "last_referenced_room": None,
+        "selected_subcategory": None,
+        "accepted_store": False,
+    }
+    for k, v in defaults.items():
+        session.setdefault(k, v)
+    return session
+
+
+def get_floor(room):
+    if 350 <= room <= 358:
+        return 0
+    if 450 <= room <= 458:
+        return 1
+    return 0
+
+
+def get_centroid(room):
+    if room in room_centroids_f3_grid:
+        return room_centroids_f3_grid[room]
+    if room in room_centroids_f4_grid:
+        return room_centroids_f4_grid[room]
+    return None
+
+
+def clamp_point(x, y, grid):
+    r = int(max(0, min(grid.shape[0] - 1, round(float(x)))))
+    c = int(max(0, min(grid.shape[1] - 1, round(float(y)))))
+    return r, c
+
+
+def get_store_info(room):
+    info = {
+        350: "Smart devices, wearables, smart home products.",
+        351: "Gaming items, accessories, and related hardware.",
+        352: "Computers, laptops, storage, and peripherals.",
+        353: "Mobile phones, tablets, and accessories.",
+        354: "Health products and personal care items.",
+        355: "Bedroom furniture and decor.",
+        356: "Living room furniture and decor.",
+        357: "Office and study furniture and tools.",
+        358: "Kitchen furniture and kitchen essentials.",
+        450: "Dining room furniture and dining essentials.",
+        451: "Bathroom items and accessories.",
+        452: "Furnishings and home textiles.",
+        453: "Kitchen and dining accessories.",
+        454: "Home decor products.",
+        455: "Tools and utility items.",
+        456: "Lighting and electrical products.",
+        457: "Cleaning and bath products.",
+        458: "Pet and gardening products.",
+    }
+    name = ROOM_INFO.get(room)
+    if not name:
+        return None
+    return f"🏪 {name} (Room {room})\n📝 {info.get(room, 'Mall store information is available for this store.')}"
+
+def find_room_in_text(text):
+    text_l = (text or "").lower()
+    room_match = re.search(r"\b(3\d{2}|45[0-8])\b", text_l)
+    if room_match:
+        room = int(room_match.group(1))
+        if room in ROOM_INFO:
+            return room
+
+    for name, room in sorted(NAME_TO_ROOM.items(), key=lambda x: len(x[0]), reverse=True):
+        if name in text_l:
+            return room
+    return None
+
+
+def get_intro_message():
+    return (
+        "Hello 👋 Welcome to the Mall Navigation Assistant.\n"
+        "You can navigate to:\n"
+        + "\n".join([f"- {name} ({room})" for room, name in ROOM_INFO.items()])
+        + "\nIf you're hungry, you can go to the kitchen/dining or cafeteria-related stores."
+    )
+
+
+def local_chat_reply(user_text):
+    low = (user_text or "").lower().strip()
+    if not low:
+        return "Please type a message or ask about a store."
+    if any(word in low for word in ["hi", "hello", "hey", "good morning", "good evening", "how are you"]):
+        return "Hello 👋 Welcome to SmartMall. Tell me a store name, a room number, or where you want to go."
+    if any(word in low for word in ["thanks", "thank you"]):
+        return "You are welcome."
+
+    room = find_room_in_text(user_text)
+    if room is not None and any(k in low for k in ["what", "tell", "about", "info", "describe", "where"]):
+        info = get_store_info(room)
+        if info:
+            return info
+
+    return None
+
+
+# ============================================================
+# AUTH / DATABASE OPERATIONS
+# ============================================================
+def save_preference(username, preference):
+    if not preference or len(preference) != 3:
+        return
+    main, cat, sub = preference
+    cursor.execute(
+        """INSERT INTO user_preferences (username, main_category, category, subcategory)
+           VALUES (?, ?, ?, ?)""",
+        (username, main, cat, sub),
+    )
+    conn.commit()
+
+
+def entry_point():
+    print("\n=== Smart Mall System ===")
+    while True:
+        print("\n1) Signup")
+        print("2) Login")
+        choice = input("Enter choice: ").strip()
+        if choice == "1":
+            return "signup"
+        if choice == "2":
+            return "login"
+        print("Invalid input ❌")
+
+
+def signup():
+    print("\n=== SIGNUP ===")
+    while True:
+        username = input("Choose username: ").strip()
+        cursor.execute("SELECT username FROM users WHERE username=?", (username,))
+        if cursor.fetchone():
+            print("❌ Username exists, try again\n")
+        else:
+            break
+
+    password = input("Choose password: ").strip()
+    name = input("Enter name: ").strip()
+    age = int(input("Enter age: "))
+    gender = input("Enter gender (male/female): ").strip().lower()
+
+    auto = input("Use auto budget? (yes/no): ").strip().lower()
+    if auto == "no":
+        lower = float(input("Lower budget: "))
+        upper = float(input("Upper budget: "))
+    else:
+        lower, upper = 500, 5000
+
+    cursor.execute(
+        "INSERT INTO users VALUES (?, ?, ?, ?, ?, ?, ?)",
+        (username, password, name, age, gender, lower, upper),
+    )
+    conn.commit()
+
+    print("\n✅ Signup successful")
+    return username, gender, lower, upper, None
+
+
+def login():
+    print("\n=== LOGIN ===")
+    while True:
+        username = input("Username: ").strip()
+        password = input("Password: ").strip()
+
+        cursor.execute(
+            "SELECT * FROM users WHERE username=? AND password=?",
+            (username, password),
+        )
+        user = cursor.fetchone()
+
+        if user:
+            print("✅ Login successful")
+            break
+        print("❌ Invalid credentials, try again\n")
+
+    cursor.execute(
+        """SELECT main_category, category, subcategory
+           FROM user_preferences
+           WHERE username=?
+           ORDER BY id DESC LIMIT 1""",
+        (username,),
+    )
+    pref = cursor.fetchone()
+    preference = pref if pref else None
+
+    return username, user[4], user[5], user[6], preference
+
+
+def auth_system():
+    while True:
+        mode = entry_point()
+        if mode == "signup":
+            user_data = signup()
+            if user_data:
+                return user_data, True
+        elif mode == "login":
+            user_data = login()
+            if user_data:
+                return user_data, False
+        print("❌ Authentication failed, try again...\n")
+
+
+def get_ble_location():
+    print("\n📍 Enter your current location")
+    x = float(input("X coordinate: "))
+    y = float(input("Y coordinate: "))
+    return (x, y)
+
+
+def save_navigation(username, start_room, dest_room):
+    if isinstance(start_room, tuple):
+        start_room = f"{start_room[0]},{start_room[1]}"
+    cursor.execute(
+        "INSERT INTO history (username, start_point, end_store) VALUES (?, ?, ?)",
+        (username, start_room, str(dest_room)),
+    )
+    conn.commit()
+
+# ============================================================
+# STORE RECOMMENDATION
+# ============================================================
+def nearest_store(user_pos):
+    best_room = None
+    best_dist = float("inf")
+    for _, room in STORE_CLUSTER_ROOM.items():
+        centroid = get_centroid(room)
+        if centroid is None:
+            continue
+        dist = np.sqrt((user_pos[0] - centroid[0]) ** 2 + (user_pos[1] - centroid[1]) ** 2)
+        if dist < best_dist:
+            best_dist = dist
+            best_room = room
+    return best_room
+
+
+def most_visited_store(username):
+    cursor.execute(
+        """SELECT end_store, COUNT(*) as cnt
+           FROM history
+           WHERE username=?
+           GROUP BY end_store
+           ORDER BY cnt DESC""",
+        (username,),
+    )
+    result = cursor.fetchone()
+    return int(result[0]) if result else None
+
+
+# ============================================================
+# PRODUCT RECOMMENDER
+# ============================================================
+def clean_price(price_str):
+    digits = "".join(filter(str.isdigit, str(price_str)))
+    return int(digits) if digits else 0
+
+
+def get_categories_for_type(type_name):
+    for t in categories_data:
+        if t.get("type") == type_name:
+            return [c.get("category") for c in t.get("categories", [])]
+    return []
+
+
+def get_subcategories(type_name, category_name):
+    for t in categories_data:
+        if t.get("type") == type_name:
+            for c in t.get("categories", []):
+                if c.get("category") == category_name:
+                    return c.get("subCategories", [])
+    return []
+
+
+def resolve_store_from_category(parent_category):
+    store_name = STORE_CLUSTER.get(parent_category, parent_category)
+    room = STORE_CLUSTER_ROOM.get(store_name)
+    return store_name, room
+
+
+def category_flow():
+    print("\n🧭 Category Selection")
+
+    print("\nMain categories:")
+    for i, c in enumerate(categories_data):
+        print(i, c["type"])
+    main = int(input("Select main category: "))
+    main_cat = categories_data[main]
+
+    print("\nSub categories:")
+    for i, c in enumerate(main_cat["categories"]):
+        print(i, c["category"])
+    sub = int(input("Select category: "))
+    sub_cat = main_cat["categories"][sub]
+
+    print("\nSub-sub categories:")
+    for i, c in enumerate(sub_cat["subCategories"]):
+        print(i, c)
+    sub_sub = int(input("Select subcategory: "))
+
+    selected_subcategory = sub_cat["subCategories"][sub_sub]
+    if selected_subcategory is None:
+        print("❌ No subcategory selected")
+        return None, None, None, None
+
+    main_category = main_cat["type"]
+    parent_category = sub_cat["category"]
+    _, end_store = resolve_store_from_category(parent_category)
+    return selected_subcategory, end_store, parent_category, main_category
+
+
+def get_budget(default_low, default_high):
+    print("\n💰 Budget Selection")
+    choice = input("Use auto budget? (yes/no): ").strip().lower()
+    if choice == "no":
+        lower = float(input("Enter lower budget: "))
+        upper = float(input("Enter upper budget: "))
+    else:
+        lower, upper = default_low, default_high
+    return lower, upper
+
+
+def get_products(subcategory):
+    if not subcategory:
+        return []
+
+    result = []
+    subcategory = str(subcategory).strip().lower()
+
+    for type_block in products_data:
+        for cat in type_block.get("items", []):
+            cat_name = cat.get("category")
+            for sub in cat.get("items", []):
+                sub_name = sub.get("subCategory")
+                if cat_name and cat_name.lower() == subcategory:
+                    result.extend(sub.get("items", []))
+                elif sub_name and sub_name.lower() == subcategory:
+                    result.extend(sub.get("items", []))
+    return result
+
+
+def filter_by_budget(products, low, high):
+    filtered = []
+    for p in products:
+        price = clean_price(p.get("Price", 0))
+        if low <= price <= high:
+            item = dict(p)
+            item["price_int"] = price
+            filtered.append(item)
+    return filtered
+
+
+def gender_filter(products, gender):
+    if gender not in ["male", "female"]:
+        return products
+
+    keywords = {
+        "male": ["gaming", "tool", "power", "sports", "fitness"],
+        "female": ["beauty", "hair", "skin", "cosmetic", "makeup", "fashion"],
+    }
+
+    preferred, others = [], []
+    for p in products:
+        name = str(p.get("Name", "")).lower()
+        if any(k in name for k in keywords[gender]):
+            preferred.append(p)
+        else:
+            others.append(p)
+    return preferred + others
+
+
+def add_promotions(products):
+    result = []
+    for p in products:
+        discount = random.randint(10, 50)
+        original = clean_price(p.get("Price", 0))
+        discounted = int(original * (1 - discount / 100))
+        result.append({
+            "Name": p.get("Name", "Unknown Product"),
+            "Brand": p.get("Brand", "N/A"),
+            "Original": f"{original}",
+            "Discounted": f"{discounted}",
+            "Discount": discount,
+            "Rating": p.get("Rating", 0),
+            "Reviews": p.get("No of Reviews", 0),
+        })
+    return result
+
+
+def recommend_for_user(selected_subcategory, gender, low_budget, high_budget):
+    products = get_products(selected_subcategory)
+    if not products:
+        return {"store": "Unknown", "room": None, "products": []}
+
+    products = filter_by_budget(products, low_budget, high_budget)
+    if not products:
+        products = get_products(selected_subcategory)
+
+    products = gender_filter(products, gender)
+    products = sorted(products, key=lambda x: x.get("Rating", 0), reverse=True)
+    top_products = add_promotions(products[:5])
+
+    parent_category = SUB_TO_CATEGORY.get(selected_subcategory)
+    store_name = STORE_CLUSTER.get(parent_category, parent_category) if parent_category else None
+    room = STORE_CLUSTER_ROOM.get(store_name) if store_name else None
+
+    return {"store": store_name or "Unknown", "room": room, "products": top_products}
+
+# ============================================================
+# NAVIGATION / A*
+# ============================================================
+def find_nearest_free(grid, r, c):
+    free = (grid == 0)
+    _, ind = distance_transform_edt(~free, return_indices=True)
+    nr, nc = ind[:, r, c]
+    return int(nr), int(nc)
+
+
+def extract_stairs(grid):
+    binary = (grid == 2).astype(int)
+    labeled, num = label(binary)
+    return labeled, num
+
+
+stairs_0, n0 = extract_stairs(grid_0)
+stairs_1, n1 = extract_stairs(grid_1)
+stairs = {}
+stair_cells = {0: {}, 1: {}}
+for i in range(1, min(n0, n1) + 1):
+    sid = f"S{i}"
+    stairs[sid] = {"floors": [0, 1]}
+    for r, c in np.argwhere(stairs_0 == i):
+        stair_cells[0][(int(r), int(c))] = sid
+    for r, c in np.argwhere(stairs_1 == i):
+        stair_cells[1][(int(r), int(c))] = sid
+
+elevator_cells = {0: set(), 1: set()}
+for f in [0, 1]:
+    grid = grid_0 if f == 0 else grid_1
+    coords = np.argwhere(grid == 3)
+    for r, c in coords:
+        elevator_cells[f].add((int(r), int(c)))
+
+fire_active = False
+fire_room = None
+fire_step = 0
+fire_center = None
+crowded_active = False
+crowded_room = None
+crowded_center = None
+
+
+def mark_danger(grid, centroid, radius=20):
+    r0, c0 = centroid
+    for i in range(grid.shape[0]):
+        for j in range(grid.shape[1]):
+            dist = np.sqrt((i - r0) ** 2 + (j - c0) ** 2)
+            if dist < radius:
+                grid[i, j] = 1
+
+
+def mark_crowd(grid, centroid, room_radius=15, corridor_radius=40):
+    r0, c0 = centroid
+    for i in range(grid.shape[0]):
+        for j in range(grid.shape[1]):
+            dist = np.sqrt((i - r0) ** 2 + (j - c0) ** 2)
+            if dist < room_radius:
+                grid[i, j] = 5
+            elif dist < corridor_radius and grid[i, j] == 0:
+                grid[i, j] = 6
+
+
+def heuristic(a, b):
+    return abs(a[1] - b[1]) + abs(a[2] - b[2]) + abs(a[0] - b[0]) * 10
+
+
+def neighbors(node, mode):
+    floor, r, c = node
+    grid = grid_0 if floor == 0 else grid_1
+    moves = [(-1, 0), (1, 0), (0, -1), (0, 1)]
+    result = []
+
+    for dr, dc in moves:
+        nr = r + dr
+        nc = c + dc
+        if 0 <= nr < grid.shape[0] and 0 <= nc < grid.shape[1]:
+            cell = grid[nr, nc]
+            if cell == 1:
+                continue
+
+            cost = 1
+
+            if fire_active and fire_center is not None:
+                fr, fc = fire_center
+                dist = np.sqrt((nr - fr) ** 2 + (nc - fc) ** 2)
+                if dist < 120:
+                    cost += (120 - dist) * 0.3
+
+            if cell == 4:
+                cost += 8
+
+            if mode == "Crowded":
+                if cell == 5:
+                    cost += 200
+                elif cell == 6:
+                    cost += 40
+
+            if mode == "Special Needs" and cell == 4:
+                cost += 100
+
+            result.append(((floor, int(nr), int(nc)), cost))
+
+    is_stair = (r, c) in stair_cells[floor]
+    is_elevator = (r, c) in elevator_cells[floor]
+
+    if mode == "Special Needs":
+        if is_elevator:
+            for f in [0, 1]:
+                if f != floor:
+                    result.append(((f, r, c), 10))
+    else:
+        if is_stair:
+            sid = stair_cells[floor][(r, c)]
+            cost = 15 if mode != "Crowded" else 40
+            if fire_active and fire_center is not None:
+                fr, fc = fire_center
+                dist = np.sqrt((r - fr) ** 2 + (c - fc) ** 2)
+                if dist < 150:
+                    cost += 200
+            for f in stairs[sid]["floors"]:
+                if f != floor:
+                    result.append(((f, r, c), cost))
+
+    return result
+
+
+def astar(start, goal, mode="Normal"):
+    open_set = []
+    heapq.heappush(open_set, (0, start))
+    came = {}
+    g = {start: 0}
+
+    while open_set:
+        _, cur = heapq.heappop(open_set)
+        if cur == goal:
+            path = []
+            while cur in came:
+                path.append(cur)
+                cur = came[cur]
+            path.append(start)
+            return path[::-1]
+
+        for nxt, cost in neighbors(cur, mode):
+            ng = g[cur] + cost
+            if nxt not in g or ng < g[nxt]:
+                g[nxt] = ng
+                heapq.heappush(open_set, (ng + heuristic(nxt, goal), nxt))
+                came[nxt] = cur
+    return None
+
+
+def path_to_instructions(path):
+    if not path:
+        return "No path found."
+
+    instructions = []
+    last_floor, last_r, last_c = path[0]
+    current_dir = None
+    dist = 0
+
+    def add_instruction(direction, distance):
+        if distance > 0:
+            instructions.append(f"Move {direction} {distance * CELL_SIZE:.1f}m.")
+
+    for i in range(1, len(path)):
+        f, r, c = path[i]
+        if f != last_floor:
+            add_instruction(current_dir, dist)
+            instructions.append(f"Take stairs from floor {last_floor + 3} to floor {f + 3}.")
+            current_dir = None
+            dist = 0
+            last_floor = f
+
+        dr = r - last_r
+        dc = c - last_c
+        if abs(dr) > abs(dc):
+            direction = "down" if dr > 0 else "up"
+            step = abs(dr)
+        else:
+            direction = "right" if dc > 0 else "left"
+            step = abs(dc)
+
+        if direction == current_dir:
+            dist += step
+        else:
+            add_instruction(current_dir, dist)
+            current_dir = direction
+            dist = step
+
+        last_r, last_c = r, c
+
+    add_instruction(current_dir, dist)
+    instructions.append("You have arrived at your destination room.")
+    return ". ".join(instructions)
+
+
+def plot_static_path(path):
+    fig, axs = plt.subplots(1, 2, figsize=(18, 6))
+    grids = [grid_0, grid_1]
+
+    for i, grid in enumerate(grids):
+        axs[i].imshow(grid, cmap="gray_r", origin="upper")
+        stairs_pos = np.where(grid == 2)
+        axs[i].scatter(stairs_pos[1], stairs_pos[0], s=40, label="Stairs")
+
+        xs = [p[2] for p in path if p[0] == i]
+        ys = [p[1] for p in path if p[0] == i]
+        if xs and ys:
+            axs[i].plot(xs, ys, linewidth=3, label="Path")
+            axs[i].scatter(xs[0], ys[0], s=80, label="Start")
+            axs[i].scatter(xs[-1], ys[-1], s=80, label="End")
+
+        axs[i].set_title(f"Floor {i + 3}")
+        axs[i].legend()
+
+    img_path = "/content/static_path.png"
+    plt.savefig(img_path)
+    plt.close()
+    return img_path
+
+
+def navigate(start_floor, start_xy, dest_room, mode="Normal"):
+    if start_xy is None or dest_room is None:
+        return None, "❌ Missing start or destination."
+
+    sf = int(start_floor)
+    sr, sc = clamp_point(start_xy[0], start_xy[1], grid_0 if sf == 0 else grid_1)
+    sr, sc = find_nearest_free(grid_0 if sf == 0 else grid_1, sr, sc)
+
+    df = get_floor(dest_room)
+    dc = get_centroid(dest_room)
+    if dc is None:
+        return None, "❌ Could not find room centroid."
+
+    gr, gc = clamp_point(dc[0], dc[1], grid_0 if df == 0 else grid_1)
+    gr, gc = find_nearest_free(grid_0 if df == 0 else grid_1, gr, gc)
+
+    path = astar((sf, sr, sc), (df, gr, gc), mode)
+    instructions = path_to_instructions(path)
+    img_path = plot_static_path(path) if path else None
+    return img_path, instructions
+
+# ============================================================
+# CHATBOT LOGIC
+# ============================================================
+def chatbot_response_fallback(text):
+    local = local_chat_reply(text)
+    if local is not None:
+        return local
+    try:
+        response = client.chat.completions.create(
+            model=OPENROUTER_MODEL,
+            messages=[
+                {"role": "system", "content": SYSTEM_PROMPT},
+                {"role": "user", "content": text},
+            ],
+        )
+        content = response.choices[0].message.content
+        if content:
+            return content
+    except Exception:
+        pass
+    return "I am here to help with mall navigation and store information. Try a store name like Electronics Store or room 351."
+
+
+def extract_rooms(user_text, history):
+    lower = (user_text or "").lower().strip()
+    room = find_room_in_text(user_text)
+
+    if room is not None:
+        if any(k in lower for k in ["what is", "tell me", "about", "info", "describe", "where is"]):
+            return "chat", None, room, False
+        if any(k in lower for k in ["take me", "guide me", "go there", "navigate", "go to", "show me"]):
+            return "navigation", None, room, True
+        return "navigation", None, room, False
+
+    if any(phrase in lower for phrase in ["take me there", "guide me there", "go there", "navigate me there"]):
+        return "navigation", None, None, True
+
+    conversation = ""
+    for u, b in history[-6:]:
+        if u:
+            conversation += f"User: {u}\n"
+        if b:
+            conversation += f"Assistant: {b}\n"
+
+    prompt = f"""
+You are an AI assistant for indoor navigation inside a shopping mall.
+
+Extract navigation intent from the conversation.
+
+Return JSON only with:
+{{
+  "intent": "navigation" or "chat",
+  "start_room": null,
+  "dest_room": null,
+  "confirmed": true or false
+}}
+
+Conversation:
+{conversation}
+
+User message:
+{user_text}
+"""
+
+    try:
+        response = client.chat.completions.create(
+            model=OPENROUTER_MODEL,
+            messages=[{"role": "user", "content": prompt}],
+            response_format={"type": "json_object"},
+        )
+        data = json.loads(response.choices[0].message.content)
+        intent = data.get("intent", "chat")
+        start_room = data.get("start_room")
+        dest_room = data.get("dest_room")
+        confirmed = data.get("confirmed", False)
+        if start_room:
+            start_room = int(start_room)
+        if dest_room:
+            dest_room = int(dest_room)
+        return intent, start_room, dest_room, confirmed
+    except Exception:
+        return "chat", None, None, False
+
+
+def chatbot_respond(user_text, history, session):
+    session = set_session_defaults(session)
+    history = history or []
+    if not history:
+        history.append(("", get_intro_message()))
+
+    lower = (user_text or "").lower().strip()
+    local = local_chat_reply(user_text)
+    if local is not None and not any(k in lower for k in ["take me", "guide me", "go there", "navigate", "go to", "show me"]):
+        history.append((user_text, local))
+        try:
+            gTTS(local).save("/content/voice.mp3")
+        except Exception:
+            pass
+        return history, None, "/content/voice.mp3", session
+
+    intent, _, dest_room, confirmed = extract_rooms(user_text, history)
+
+    info_room = find_room_in_text(user_text)
+    if info_room is not None and any(k in lower for k in ["what", "tell", "about", "info", "describe", "where"]):
+        info = get_store_info(info_room)
+        if info:
+            bot = f"{info}\n\nWould you like navigation to {ROOM_INFO[info_room]} (Room {info_room})?"
+            session["last_referenced_room"] = info_room
+            session["last_dest_room"] = info_room
+            history.append((user_text, bot))
+            try:
+                gTTS(bot).save("/content/voice.mp3")
+            except Exception:
+                pass
+            return history, None, "/content/voice.mp3", session
+
+    if intent == "navigation" and dest_room is not None:
+        store_name = ROOM_INFO.get(dest_room, f"Room {dest_room}")
+        if not confirmed:
+            bot = f"I can guide you to {store_name} (Room {dest_room}). Do you want navigation?"
+            history.append((user_text, bot))
+            session["last_referenced_room"] = dest_room
+            session["last_dest_room"] = dest_room
+            session["nav_target_room"] = dest_room
+            try:
+                gTTS(bot).save("/content/voice.mp3")
+            except Exception:
+                pass
+            return history, None, "/content/voice.mp3", session
+
+        if session.get("start_floor") is None or session.get("start_xy") is None:
+            bot = "Please set your starting location first in the Shop & Navigate tab."
+            history.append((user_text, bot))
+            try:
+                gTTS(bot).save("/content/voice.mp3")
+            except Exception:
+                pass
+            return history, None, "/content/voice.mp3", session
+
+        img_path, instructions = navigate(session["start_floor"], session["start_xy"], dest_room, session.get("nav_mode", "Normal"))
+        save_navigation(session.get("username", "guest"), (session["start_floor"], session["start_xy"]), dest_room)
+        save_navigation_mem(f"floor={session['start_floor']},x={session['start_xy'][0]},y={session['start_xy'][1]}", dest_room)
+
+        bot = f"🗺️ Navigating to {store_name} (Room {dest_room})\n\n{instructions}"
+        history.append((user_text, bot))
+        session["last_dest_room"] = dest_room
+        session["nav_target_room"] = dest_room
+        try:
+            gTTS(f"Navigating to {store_name}. {instructions.replace(chr(10), ' ')}").save("/content/voice.mp3")
+        except Exception:
+            pass
+        return history, img_path, "/content/voice.mp3", session
+
+    if any(phrase in lower for phrase in ["take me there", "guide me there", "go there", "navigate me there"]):
+        dest = (
+            session.get("last_dest_room")
+            or session.get("nav_target_room")
+            or session.get("last_referenced_room")
+            or session.get("recommended_room")
+        )
+        if dest is not None:
+            store_name = ROOM_INFO.get(dest, f"Room {dest}")
+            if session.get("start_floor") is None or session.get("start_xy") is None:
+                bot = f"I can take you to {store_name}, but I need your starting location first in the Shop & Navigate tab."
+                history.append((user_text, bot))
+                try:
+                    gTTS(bot).save("/content/voice.mp3")
+                except Exception:
+                    pass
+                return history, None, "/content/voice.mp3", session
+
+            img_path, instructions = navigate(session["start_floor"], session["start_xy"], dest, session.get("nav_mode", "Normal"))
+            save_navigation(session.get("username", "guest"), (session["start_floor"], session["start_xy"]), dest)
+            save_navigation_mem(f"floor={session['start_floor']},x={session['start_xy'][0]},y={session['start_xy'][1]}", dest)
+
+            bot = f"🗺️ Navigating to {store_name} (Room {dest})\n\n{instructions}"
+            history.append((user_text, bot))
+            session["last_dest_room"] = dest
+            session["nav_target_room"] = dest
+            try:
+                gTTS(f"Navigating to {store_name}. {instructions.replace(chr(10), ' ')}").save("/content/voice.mp3")
+            except Exception:
+                pass
+            return history, img_path, "/content/voice.mp3", session
+
+    bot = chatbot_response_fallback(user_text)
+    history.append((user_text, bot))
+    try:
+        gTTS(bot).save("/content/voice.mp3")
+    except Exception:
+        pass
+    return history, None, "/content/voice.mp3", session
+
+
+def speech_to_text(audio):
+    if audio is None:
+        return ""
+    try:
+        sound = AudioSegment.from_file(audio)
+        wav = "/content/temp.wav"
+        sound.export(wav, format="wav")
+        r = sr.Recognizer()
+        with sr.AudioFile(wav) as src:
+            data = r.record(src)
+        return r.recognize_google(data)
+    except Exception:
+        return ""
+
+# ============================================================
+# UI HELPERS
+# ============================================================
+def format_products_html(promos, store_name, room):
+    if not promos:
+        return "<div style='color:#ff6b6b;padding:20px;text-align:center;font-family:monospace'>❌ No products found in your budget range.</div>"
+
+    cards = ""
+    for p in promos:
+        stars = "⭐" * int(round(float(p.get("Rating", 0))))
+        cards += f"""
+        <div style="background:linear-gradient(135deg,#1a1a2e,#16213e);border:1px solid #00d4ff33;
+                    border-radius:12px;padding:16px;margin-bottom:12px;">
+          <div style="display:flex;justify-content:space-between;align-items:flex-start;gap:10px;">
+            <div style="flex:1;">
+              <div style="font-size:11px;color:#888;font-family:monospace;margin-bottom:4px">{p.get('Brand','N/A')}</div>
+              <div style="color:#e0e0e0;font-size:13px;font-family:monospace;line-height:1.4">{p.get('Name','Unknown')}</div>
+              <div style="margin-top:8px;font-size:12px;color:#ffd700">{stars} {p.get('Rating',0)} ({int(p.get('Reviews',0)):,} reviews)</div>
+            </div>
+            <div style="text-align:right;flex-shrink:0;">
+              <div style="color:#ff6b6b;font-size:12px;text-decoration:line-through;font-family:monospace">{p.get('Original','0')} EGP</div>
+              <div style="color:#00ff88;font-size:16px;font-weight:bold;font-family:monospace">{p.get('Discounted','0')} EGP</div>
+              <div style="background:#ff4500;color:white;border-radius:20px;padding:3px 10px;
+                          font-size:11px;font-weight:bold;margin-top:4px;font-family:monospace">{p.get('Discount',0)}% OFF 🔥</div>
+            </div>
+          </div>
+        </div>"""
+
+    return f"""
+    <div style="font-family:monospace;max-height:500px;overflow-y:auto;padding:4px;">
+      <div style="color:#00d4ff;font-size:14px;font-weight:bold;margin-bottom:12px;
+                  padding-bottom:8px;border-bottom:1px solid #00d4ff33;">
+        🏪 {store_name} · Room {room}
+      </div>
+      {cards}
+    </div>"""
+
+
+def set_location_ui(x, y, floor_choice, session):
+    session = set_session_defaults(session)
+    if x is None or y is None:
+        return "<div style='color:#ff6b6b;font-family:monospace'>❌ Please enter coordinates.</div>", session
+    session["start_floor"] = 0 if str(floor_choice) == "3" else 1
+    session["start_xy"] = (float(x), float(y))
+    session["start_room"] = session["start_xy"]
+    return f"<div style='color:#00ff88;font-family:monospace'>✅ Location saved: floor {int(floor_choice)}, x={x}, y={y}</div>", session
+
+
+def recommend_store_for_session(session):
+    session = set_session_defaults(session)
+    if not session.get("username"):
+        return "<div style='color:#ff6b6b;font-family:monospace'>❌ Please login first.</div>", session
+    if session.get("start_xy") is None or session.get("start_floor") is None:
+        return "<div style='color:#ff6b6b;font-family:monospace'>❌ Please enter your current location first.</div>", session
+
+    if session.get("is_new", True):
+        room = nearest_store(session["start_xy"])
+        msg = "🆕 New user → Nearest store recommended"
+    else:
+        room = most_visited_store(session["username"])
+        if room is None:
+            room = nearest_store(session["start_xy"])
+        msg = "🔁 Returning user → Most visited store recommended"
+
+    store_name = ROOM_INFO.get(room, f"Room {room}")
+    session["recommended_room"] = room
+    session["nav_target_room"] = room
+    session["last_referenced_room"] = room
+    session["last_dest_room"] = room
+
+    html = f"""
+    <div style="background:linear-gradient(135deg,#1a1a2e,#0d2137);border:1px solid #00d4ff55;
+                border-radius:12px;padding:20px;font-family:monospace;">
+      <div style="color:#888;font-size:11px;letter-spacing:2px">{msg}</div>
+      <div style="color:#00d4ff;font-size:22px;font-weight:bold;margin:8px 0">{store_name}</div>
+      <div style="color:#555;font-size:12px">Room {room}</div>
+    </div>"""
+    return html, session
+
+
+def update_cat(type_name):
+    return gr.update(choices=get_categories_for_type(type_name), value=None)
+
+
+def update_sub(type_name, cat_name):
+    return gr.update(choices=get_subcategories(type_name, cat_name), value=None)
+
+
+def handle_agreement(choice, session):
+    session = set_session_defaults(session)
+    session["accepted_store"] = (choice or "").startswith("✅")
+    types = [t["type"] for t in categories_data]
+    if session["accepted_store"]:
+        return (
+            gr.update(visible=False),
+            gr.update(visible=False),
+            gr.update(visible=False),
+            gr.update(choices=types, visible=True),
+        )
+    return (
+        gr.update(choices=types, visible=True),
+        gr.update(visible=True),
+        gr.update(visible=True),
+        gr.update(choices=types, visible=True),
+    )
+
+
+def do_navigate(choice, type_name, cat_name, sub_name, nav_mode_val, danger_room, session):
+    session = set_session_defaults(session)
+    if not session.get("username"):
+        return "❌ Please login first.", None
+    if session.get("start_xy") is None or session.get("start_floor") is None:
+        return "❌ Please set your location first.", None
+
+    reset_grids()
+
+    if "Yes" in (choice or ""):
+        dest_room = session.get("recommended_room")
+        if dest_room is None:
+            return "❌ Please get a recommendation first.", None
+    else:
+        if not sub_name:
+            return "❌ Please select a sub-category.", None
+        parent = SUB_TO_CATEGORY.get(sub_name, cat_name)
+        store = STORE_CLUSTER.get(parent, parent)
+        dest_room = STORE_CLUSTER_ROOM.get(store)
+        if dest_room is None:
+            return "❌ Could not map the selected category to a store.", None
+
+    session["dest_room"] = dest_room
+    session["nav_mode"] = nav_mode_val
+
+    room_val = None
+    try:
+        if danger_room is not None and str(danger_room).strip() != "":
+            room_val = int(float(danger_room))
+    except Exception:
+        room_val = None
+
+    global fire_active, fire_room, fire_step, fire_center
+    global crowded_active, crowded_room, crowded_center
+
+    fire_active = False
+    crowded_active = False
+    fire_room = None
+    crowded_room = None
+    fire_center = None
+    crowded_center = None
+
+    if nav_mode_val == "Fire" and room_val is not None and room_val in ROOM_INFO:
+        fire_active = True
+        fire_room = room_val
+        fire_center = get_centroid(room_val)
+        if fire_center is not None:
+            fire_step += 1
+            radius = 20 + fire_step * 6
+            if get_floor(room_val) == 0:
+                mark_danger(grid_0, fire_center, radius)
+            else:
+                mark_danger(grid_1, fire_center, radius)
+
+    if nav_mode_val == "Crowded" and room_val is not None and room_val in ROOM_INFO:
+        crowded_active = True
+        crowded_room = room_val
+        crowded_center = get_centroid(room_val)
+        if crowded_center is not None:
+            if get_floor(room_val) == 0:
+                mark_crowd(grid_0, crowded_center)
+            else:
+                mark_crowd(grid_1, crowded_center)
+
+    save_navigation(session["username"], (session["start_floor"], session["start_xy"]), dest_room)
+    save_navigation_mem((session["start_floor"], session["start_xy"]), dest_room)
+
+    img_path, instructions = navigate(session["start_floor"], session["start_xy"], dest_room, nav_mode_val)
+    return instructions, img_path
+
+
+def do_get_products(type_name, cat_name, sub_name, session):
+    session = set_session_defaults(session)
+    if not session.get("username"):
+        return "<div style='color:#ff6b6b;font-family:monospace'>❌ Please login first.</div>"
+    if not sub_name:
+        return "<div style='color:#ff6b6b;font-family:monospace'>❌ Please select a sub-category.</div>"
+
+    if type_name and cat_name and sub_name:
+        save_preference(session["username"], (type_name, cat_name, sub_name))
+
+    gender = session.get("gender", "other")
+    low = session.get("low", 500)
+    high = session.get("high", 5000)
+    result = recommend_for_user(sub_name, gender, low, high)
+    session["selected_subcategory"] = sub_name
+    return format_products_html(result["products"], result["store"], result["room"])
+
+# ============================================================
+# MAIN GRADIO APP
+# ============================================================
+with gr.Blocks(
+    theme=gr.themes.Base(primary_hue="cyan", secondary_hue="blue", neutral_hue="slate"),
+    css="""
+    body { background: #0a0a14 !important; }
+    .gradio-container { background: #0a0a14 !important; max-width: 1200px !important; }
+    .tab-nav button { font-family: monospace; letter-spacing: 2px; font-size: 12px; }
+    .panel { background: #111827; border: 1px solid #1e3a5f; border-radius: 12px; padding: 20px; }
+    h1, h2, h3 { font-family: monospace !important; letter-spacing: 3px !important; }
+    footer { display: none !important; }
+    """,
+    title="SmartMall AI System",
+) as demo:
+
+    session_state = gr.State({})
+
+    gr.HTML("""
+    <div style="text-align:center;padding:24px 0 8px;font-family:monospace;">
+      <div style="font-size:28px;font-weight:900;letter-spacing:6px;
+                  background:linear-gradient(90deg,#00d4ff,#00ff88,#ff4466);
+                  -webkit-background-clip:text;-webkit-text-fill-color:transparent">
+        SMART MALL AI
+      </div>
+      <div style="color:#555;font-size:11px;letter-spacing:4px;margin-top:6px">
+        NAVIGATION · RECOMMENDATIONS · ASSISTANT
+      </div>
+    </div>
+    """)
+
+    with gr.Tabs():
+        # ----------------------------------------------------
+        # TAB 1 — AUTH
+        # ----------------------------------------------------
+        with gr.Tab("🔐 LOGIN / SIGNUP"):
+            gr.HTML("<div style='color:#00d4ff;font-family:monospace;font-size:13px;margin-bottom:16px'>Enter your credentials to begin</div>")
+            with gr.Row():
+                with gr.Column(scale=1):
+                    auth_username = gr.Textbox(label="Username", placeholder="your_username")
+                    auth_password = gr.Textbox(label="Password", type="password", placeholder="••••••••")
+                    auth_name = gr.Textbox(label="Full Name (signup only)", placeholder="John Doe")
+                    auth_age = gr.Number(label="Age (signup only)", value=25)
+                    auth_gender = gr.Dropdown(["male", "female", "other"], label="Gender (signup only)", value="male")
+                    auth_budget_l = gr.Number(label="Min Budget EGP (signup only)", value=500)
+                    auth_budget_h = gr.Number(label="Max Budget EGP (signup only)", value=5000)
+
+                    with gr.Row():
+                        signup_btn = gr.Button("SIGNUP", variant="primary")
+                        login_btn = gr.Button("LOGIN", variant="secondary")
+                    auth_status = gr.HTML()
+
+            def do_signup(username, password, name, age, gender, low, high, session):
+                cursor.execute("SELECT username FROM users WHERE username=?", (username,))
+                if cursor.fetchone():
+                    return "<div style='color:#ff6b6b;font-family:monospace'>❌ Username already exists.</div>", session
+                cursor.execute(
+                    "INSERT INTO users VALUES (?,?,?,?,?,?,?)",
+                    (username, password, name, int(age), gender, low, high),
+                )
+                conn.commit()
+                session = set_session_defaults(session)
+                session.update({"username": username, "gender": gender, "low": low, "high": high, "is_new": True})
+                return f"<div style='color:#00ff88;font-family:monospace'>✅ Welcome, {name}! Account created.</div>", session
+
+            def do_login(username, password, session):
+                cursor.execute("SELECT * FROM users WHERE username=? AND password=?", (username, password))
+                user = cursor.fetchone()
+                if not user:
+                    return "<div style='color:#ff6b6b;font-family:monospace'>❌ Invalid credentials.</div>", session
+                session = set_session_defaults(session)
+                session.update({"username": user[0], "gender": user[4], "low": user[5], "high": user[6], "is_new": False})
+                return f"<div style='color:#00ff88;font-family:monospace'>✅ Welcome back, {user[2]}!</div>", session
+
+            signup_btn.click(do_signup, [auth_username, auth_password, auth_name, auth_age, auth_gender, auth_budget_l, auth_budget_h, session_state], [auth_status, session_state])
+            login_btn.click(do_login, [auth_username, auth_password, session_state], [auth_status, session_state])
+
+        # ----------------------------------------------------
+        # TAB 2 — RECOMMENDATION + NAVIGATION
+        # ----------------------------------------------------
+        with gr.Tab("🏪 SHOP & NAVIGATE"):
+            gr.HTML("<div style='color:#00d4ff;font-family:monospace;font-size:13px;margin-bottom:16px'>Get store recommendations and navigate</div>")
+
+            gr.HTML("<div style='color:#888;font-family:monospace;font-size:11px;margin-bottom:8px'>STEP 0 — SET YOUR LOCATION</div>")
+            with gr.Row():
+                loc_x = gr.Number(label="X Coordinate")
+                loc_y = gr.Number(label="Y Coordinate")
+                loc_floor = gr.Dropdown(["3", "4"], label="Floor", value="3")
+                set_loc_btn = gr.Button("📍 Set Location")
+            loc_status = gr.HTML()
+
+            with gr.Row():
+                with gr.Column(scale=1):
+                    gr.HTML("<div style='color:#888;font-family:monospace;font-size:11px;margin-bottom:8px'>STEP 1 — GET RECOMMENDATION</div>")
+                    recommend_btn = gr.Button("🎯 Get Store Recommendation", variant="primary")
+                    rec_result_html = gr.HTML()
+                    rec_agree = gr.Radio(["✅ Yes, navigate me there", "❌ No, I'll choose my own"], label="Accept recommendation?")
+
+                    gr.HTML("<div style='color:#888;font-family:monospace;font-size:11px;margin-top:20px;margin-bottom:8px'>STEP 2 — CHOOSE CATEGORY (for navigation if rejected, and for products)</div>")
+                    cat_type = gr.Dropdown([t["type"] for t in categories_data], label="Main Category")
+                    cat_cat = gr.Dropdown([], label="Category")
+                    cat_sub = gr.Dropdown([], label="Sub-Category")
+                    nav_mode = gr.Dropdown(["Normal", "Fire", "Crowded", "Special Needs"], label="Navigation Mode", value="Normal")
+                    danger_room = gr.Textbox(label="Fire/Crowded Room Number", placeholder="Example: 350")
+                    navigate_btn = gr.Button("🗺️ Start Navigation", variant="primary")
+
+                with gr.Column(scale=1):
+                    nav_instructions = gr.Textbox(label="Navigation Instructions", lines=10, interactive=False)
+                    nav_image = gr.Image(label="Mall Map", type="filepath")
+
+            gr.HTML("<div style='color:#888;font-family:monospace;font-size:11px;margin-top:20px;margin-bottom:8px'>STEP 3 — PRODUCT RECOMMENDATIONS</div>")
+            with gr.Row():
+                prod_type = gr.Dropdown([], label="Category Type")
+                prod_cat = gr.Dropdown([], label="Category")
+                prod_sub = gr.Dropdown([], label="Sub-Category")
+            get_products_btn = gr.Button("🛍️ Get Product Recommendations", variant="primary")
+            products_html = gr.HTML()
+
+            set_loc_btn.click(set_location_ui, [loc_x, loc_y, loc_floor, session_state], [loc_status, session_state])
+            recommend_btn.click(recommend_store_for_session, [session_state], [rec_result_html, session_state])
+            rec_agree.change(handle_agreement, [rec_agree, session_state], [cat_type, cat_cat, cat_sub, prod_type])
+            cat_type.change(update_cat, [cat_type], [cat_cat])
+            cat_cat.change(update_sub, [cat_type, cat_cat], [cat_sub])
+            prod_type.change(update_cat, [prod_type], [prod_cat])
+            prod_cat.change(update_sub, [prod_type, prod_cat], [prod_sub])
+
+            navigate_btn.click(
+                do_navigate,
+                [rec_agree, cat_type, cat_cat, cat_sub, nav_mode, danger_room, session_state],
+                [nav_instructions, nav_image],
+            )
+
+            get_products_btn.click(do_get_products, [prod_type, prod_cat, prod_sub, session_state], [products_html])
+
+            def init_dropdowns():
+                types = [t["type"] for t in categories_data]
+                return gr.update(choices=types), gr.update(choices=types)
+
+            demo.load(init_dropdowns, [], [prod_type, cat_type])
+
+        # ----------------------------------------------------
+        # TAB 3 — CHATBOT
+        # ----------------------------------------------------
+        with gr.Tab("🤖 AI ASSISTANT"):
+            gr.HTML("<div style='color:#00d4ff;font-family:monospace;font-size:13px;margin-bottom:16px'>Chat with the mall AI assistant</div>")
+
+            chatbot_widget = gr.Chatbot(
+                label="SmartMall Assistant",
+                height=480,
+                value=[("", "👋 Welcome to SmartMall! I can help you navigate, find products, or answer any questions. How can I help?")],
+            )
+
+            with gr.Row():
+                chat_msg = gr.Textbox(placeholder="Ask me anything about the mall...", label="Message", scale=4)
+                chat_mic = gr.Audio(source="microphone", type="filepath", label="🎤", scale=1)
+
+            chat_audio_out = gr.Audio(label="Voice Response")
+            chat_img_out = gr.Image(label="Navigation Map", type="filepath")
+
+            def chat_respond(msg, history, session):
+                if not msg.strip():
+                    return history, None, None, session
+                history, img, audio, session = chatbot_respond(msg, history, session)
+                return history, img, audio, session
+
+            def voice_chat(audio, history, session):
+                text = speech_to_text(audio)
+                if not text:
+                    history.append(("🎤 [voice]", "Sorry, I couldn't understand the audio."))
+                    return history, None, None, session
+                return chat_respond(text, history, session)
+
+            chat_msg.submit(chat_respond, [chat_msg, chatbot_widget, session_state], [chatbot_widget, chat_img_out, chat_audio_out, session_state])
+            chat_mic.change(voice_chat, [chat_mic, chatbot_widget, session_state], [chatbot_widget, chat_img_out, chat_audio_out, session_state])
+
+if __name__ == "__main__":
+    demo.launch()

categories.json

@@ -0,0 +1,318 @@
+[
+    {
+        "type": "Electronics",
+        "categories": [
+            {
+                "category": "Audio",
+                "subCategories": [
+                    "Bluetooth Headphones",
+                    "Wired Headphones",
+                    "True Wireless Earbuds",
+                    "Bluetooth Speakers",
+                    "Soundbars",
+                    "Home Theatres",
+                    "TV Streaming Device"
+                ]
+            },
+            {
+                "category": "Cameras",
+                "subCategories": [
+                    "DSLR & Mirrorless",
+                    "Sports & action",
+                    "Point & Shoot",
+                    "Instant Cameras",
+                    "Camcorders",
+                    "Camera tripods",
+                    "Camera Lenses",
+                    "Drone",
+                    "Flashes",
+                    "Gimbals",
+                    "Binoculars"
+                ]
+            },
+            {
+                "category": "Computer Peripherals",
+                "subCategories": [
+                    "Printers",
+                    "Monitors",
+                    "Projectors",
+                    "Portable projectors",
+                    "Ink Catridges",
+                    "Ink Bottles",
+                    "Receipt Printers",
+                    "Lamination Machines",
+                    "Note Counting Machines",
+                    "Barcode scanners",
+                    "Currency Detectors"
+                ]
+            },
+            {
+                "category": "Gaming",
+                "subCategories": [
+                    "Gaming Consoles",
+                    "Gaming Mouse",
+                    "Gaming Keyboards",
+                    "Gamepads",
+                    "Games",
+                    "Gaming Mousepads",
+                    "Controllers",
+                    "Gaming Components"
+                ]
+            },
+            {
+                "category": "Health & Personal Care",
+                "subCategories": [
+                    "Trimmers",
+                    "Hair Straighteners",
+                    "Hair Dryers",
+                    "Epilators",
+                    "Glucometers And Accessories",
+                    "Blood Pressure Monitors",
+                    "Digital Thermometers",
+                    "Weighing Scales",
+                    "Massagers",
+                    "Nebulizers",
+                    "Vaporizers"
+                ]
+            },
+            {
+                "category": "Laptop Accessories",
+                "subCategories": [
+                    "Mouse",
+                    "Laptop Keyboards",
+                    "Router",
+                    "Data Cards",
+                    "UPS",
+                    "USB Gadgets",
+                    "Security Software",
+                    "Laptop Battery",
+                    "Laptop Adapter",
+                    "Wireless USB Adapter",
+                    "Processor"
+                ]
+            },
+            {
+                "category": "Laptop & Desktop",
+                "subCategories": [
+                    "Laptops",
+                    "Gaming Laptops",
+                    "Desktop PCs",
+                    "All in One PCs",
+                    "Mini PCs",
+                    "Tower PCs",
+                    "PC Finder"
+                ]
+            },
+            {
+                "category": "Smart Home Automation",
+                "subCategories": [
+                    "Smart Assistants",
+                    "Smart Lights",
+                    "Smart Cameras",
+                    "Smart Switches",
+                    "Smart Door Locks",
+                    "Sensors & Alarms"
+                ]
+            },
+            {
+                "category": "Smart Wearables",
+                "subCategories": [
+                    "Smart Watches",
+                    "Smart Bands",
+                    "Smart Glasses"
+                ]
+            },
+            {
+                "category": "Storage",
+                "subCategories": [
+                    "Mobile Memory Card",
+                    "Computer Storage Pendrive",
+                    "Mobile Storage Pendrive",
+                    "External Hard Drive",
+                    "Internal Hard Drive"
+                ]
+            },
+            {
+                "category": "Tabelts",
+                "subCategories": ["Wifi", "Wifi + Cellular"]
+            },
+            {
+                "category": "Mobiles",
+                "subCategories": [
+                    "Mi mobiles",
+                    "Realme mobiles",
+                    "Samsung mobiles",
+                    "Infinix mobiles",
+                    "OPPO mobiles",
+                    "Apple mobiles",
+                    "Vivo mobiles",
+                    "Honor mobiles",
+                    "Asus mobiles"
+                ]
+            }
+        ]
+    },
+    {
+        "type": "Furniture",
+        "categories": [
+            {
+                "category": "Bedroom",
+                "subCategories": [
+                    "Beds",
+                    "Mattresses",
+                    "Wardrobes",
+                    "Bedding",
+                    "Side tables",
+                    "Mirrors",
+                    "Pillows"
+                ]
+            },
+            {
+                "category": "Living Room",
+                "subCategories": [
+                    "Sofa",
+                    "Coffee table",
+                    "TV Stand",
+                    "cabinets",
+                    "Curtains",
+                    "Lighting"
+                ]
+            },
+            {
+                "category": "Study/Office",
+                "subCategories": [
+                    "Desk",
+                    "Laptop tables",
+                    "Office chair",
+                    "Drawer units",
+                    "Shelves"
+                ]
+            },
+            {
+                "category": "Kitchen",
+                "subCategories": [
+                    "Kitchen chairs",
+                    "Kitchen trolleys",
+                    "Kitchen appliances",
+                    "Kitchen sinks",
+                    "Kitchen taps"
+                ]
+            },
+            {
+                "category": "Dining Room",
+                "subCategories": [
+                    "Dining tables",
+                    "Dining chairs",
+                    "Dining sets",
+                    "Cutlery units"
+                ]
+            },
+            {
+                "category": "Bathroom",
+                "subCategories": [
+                    "Vanities",
+                    "Showers",
+                    "Bathroom storage",
+                    "Taps",
+                    "Sinks"
+                ]
+            }
+        ]
+    },
+    {
+        "type": "Household",
+        "categories": [
+            {
+                "category": "Furnishings",
+                "subCategories": [
+                    "Bed Linens",
+                    "Bedsheets",
+                    "Blankets",
+                    "Curtains",
+                    "Bath linen",
+                    "Pillows",
+                    "Kitchen linen"
+                ]
+            },
+            {
+                "category": "Kitchen and Dining",
+                "subCategories": [
+                    "Cookware",
+                    "Lunchboxes, bottles",
+                    "Knives,Choppers",
+                    "Gas stove",
+                    "Dinnerware",
+                    "Kitchen storage",
+                    "Bakeware"
+                ]
+            },
+            {
+                "category": "Home Decor",
+                "subCategories": [
+                    "Lighting",
+                    "Wallpapers",
+                    "Paintings and posters",
+                    "Clocks",
+                    "Decoratives",
+                    "Vases",
+                    "Home fragrances",
+                    "Photo frames",
+                    "Wall decor"
+                ]
+            },
+            {
+                "category": "Tools and utility",
+                "subCategories": [
+                    "Hand tools",
+                    "Power tools",
+                    "Measuring tools",
+                    "Home storage",
+                    "Umbrellas",
+                    "Lock and security",
+                    "Paint equipment"
+                ]
+            },
+            {
+                "category": "Lighting and Electricals",
+                "subCategories": [
+                    "Bulbs",
+                    "Emergency lights",
+                    "Torches",
+                    "Tube lights",
+                    "Extension cords",
+                    "Batteries",
+                    "Solar lights"
+                ]
+            },
+            {
+                "category": "Cleaning and Bath",
+                "subCategories": [
+                    "Mops",
+                    "Cleaning Supplies",
+                    "Air freshners",
+                    "Household supplies",
+                    "Liquid Detergents",
+                    "Taps and Faucets",
+                    "Bathroom shelves",
+                    "Shower heads",
+                    "Bathroom Accesories"
+                ]
+            },
+            {
+                "category": "Pet and Gardening",
+                "subCategories": [
+                    "Plant seeds",
+                    "Manure",
+                    "Pots",
+                    "Garden tools set",
+                    "Watering equipment",
+                    "Dog essentials",
+                    "Cat essentials",
+                    "Fish and aquatic",
+                    "Pet hygiene",
+                    "Pet toys"
+                ]
+            }
+        ]
+    }
+]

floor_0_grid.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5371819d353037135d8ec5f47937c8857c7e02c17e3c4eda3e0bdb12dc8876d
+size 42581

requirements.txt

@@ -0,0 +1,9 @@
+gradio==3.39.0
+gradio-client==0.3.0
+numpy
+scipy
+matplotlib
+openai
+gTTS
+SpeechRecognition
+pydub