import json
import pickle
import random
import nltk
import numpy as np
import tflearn
import gradio as gr
import requests
import torch
import pandas as pd
from bs4 import BeautifulSoup
from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
from nltk.tokenize import word_tokenize
from nltk.stem.lancaster import LancasterStemmer
import os

# Ensure necessary NLTK resources are downloaded
nltk.download('punkt')

# Initialize the stemmer
stemmer = LancasterStemmer()

# Load intents.json
try:
    with open("intents.json") as file:
        data = json.load(file)
except FileNotFoundError:
    raise FileNotFoundError("Error: 'intents.json' file not found. Ensure it exists in the current directory.")

# Load preprocessed data from pickle
try:
    with open("data.pickle", "rb") as f:
        words, labels, training, output = pickle.load(f)
except FileNotFoundError:
    raise FileNotFoundError("Error: 'data.pickle' file not found. Ensure it exists and matches the model.")

# Build the model structure
net = tflearn.input_data(shape=[None, len(training[0])])
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, len(output[0]), activation="softmax")
net = tflearn.regression(net)

# Load the trained model
model = tflearn.DNN(net)
try:
    model.load("MentalHealthChatBotmodel.tflearn")
except FileNotFoundError:
    raise FileNotFoundError("Error: Trained model file 'MentalHealthChatBotmodel.tflearn' not found.")

# Function to process user input into a bag-of-words format
def bag_of_words(s, words):
    bag = [0 for _ in range(len(words))]
    s_words = word_tokenize(s)
    s_words = [stemmer.stem(word.lower()) for word in s_words if word.lower() in words]
    for se in s_words:
        for i, w in enumerate(words):
            if w == se:
                bag[i] = 1
    return np.array(bag)

# Chat function
def chat(message, history):
    history = history or []
    message = message.lower()
    
    try:
        # Predict the tag
        results = model.predict([bag_of_words(message, words)])
        results_index = np.argmax(results)
        tag = labels[results_index]

        # Match tag with intent and choose a random response
        for tg in data["intents"]:
            if tg['tag'] == tag:
                responses = tg['responses']
                response = random.choice(responses)
                break
        else:
            response = "I'm sorry, I didn't understand that. Could you please rephrase?"

    except Exception as e:
        response = f"An error occurred: {str(e)}"
    
    history.append((message, response))
    return history, history


# Sentiment analysis setup
tokenizer_sentiment = AutoTokenizer.from_pretrained("cardiffnlp/twitter-roberta-base-sentiment")
model_sentiment = AutoModelForSequenceClassification.from_pretrained("cardiffnlp/twitter-roberta-base-sentiment")

# Emotion detection setup
def load_emotion_model():
    tokenizer = AutoTokenizer.from_pretrained("j-hartmann/emotion-english-distilroberta-base")
    model = AutoModelForSequenceClassification.from_pretrained("j-hartmann/emotion-english-distilroberta-base")
    return tokenizer, model

tokenizer_emotion, model_emotion = load_emotion_model()

# Google Places API setup for wellness professionals
url = "https://maps.googleapis.com/maps/api/place/textsearch/json"
places_details_url = "https://maps.googleapis.com/maps/api/place/details/json"
api_key = os.getenv("GOOGLE_API_KEY")  # Use environment variable for security

# Function to get places data using Google Places API
def get_places_data(query, location, radius, api_key, next_page_token=None):
    params = {
        "query": query,
        "location": location,
        "radius": radius,
        "key": api_key
    }

    if next_page_token:
        params["pagetoken"] = next_page_token

    response = requests.get(url, params=params)
    if response.status_code == 200:
        return response.json()
    else:
        return None

# Web scraping function to get wellness professional data (alternative to API)
def scrape_wellness_professionals(query, location):
    # User-Agent header to simulate a browser request
    headers = {
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36"
    }
    
    search_url = f"https://www.google.com/search?q={query}+near+{location}"
    
    # Make a request to the search URL with headers
    response = requests.get(search_url, headers=headers)
    if response.status_code == 200:
        soup = BeautifulSoup(response.text, 'html.parser')
        
        # Parse and extract wellness professionals from the HTML
        wellness_data = []
        results = soup.find_all("div", class_="BVG0Nb")  # Adjust class based on the actual HTML structure
        for result in results:
            name = result.get_text()
            link = result.find("a")["href"] if result.find("a") else None
            wellness_data.append([name, link])
        
        return wellness_data
    else:
        return []

# Main function to fetch wellness professional data and display on map
def get_wellness_professionals(location):
    query = "therapist OR counselor OR mental health professional OR marriage and family therapist OR psychotherapist OR psychiatrist OR psychologist OR nutritionist OR wellness doctor OR holistic practitioner OR integrative medicine OR chiropractor OR naturopath"
    radius = 50000  # 50 km radius
    
    # Using Google Places API if available
    data = get_places_data(query, location, radius, api_key)

    if data:
        results = data.get('results', [])
        wellness_data = []
        for place in results:
            name = place.get("name")
            address = place.get("formatted_address")
            latitude = place.get("geometry", {}).get("location", {}).get("lat")
            longitude = place.get("geometry", {}).get("location", {}).get("lng")
            wellness_data.append([name, address, latitude, longitude])
        return wellness_data

    # Fallback to scraping if API is not available or fails
    return scrape_wellness_professionals(query, location)

# Gradio interface setup for user interaction
def user_interface(message, location, history):
    history, history = chat(message, history)
    
    # Sentiment analysis
    inputs = tokenizer_sentiment(message, return_tensors="pt")
    outputs = model_sentiment(**inputs)
    sentiment = ["Negative", "Neutral", "Positive"][torch.argmax(outputs.logits, dim=1).item()]
    
    # Emotion detection
    pipe = pipeline("text-classification", model=model_emotion, tokenizer=tokenizer_emotion)
    emotion_result = pipe(message)
    emotion = emotion_result[0]['label']
    
    # Get wellness professionals
    wellness_data = get_wellness_professionals(location)
    
    return history, history, sentiment, emotion, wellness_data

# Gradio chatbot interface
chatbot = gr.Chatbot(label="Mental Health Chatbot")
location_input = gr.Textbox(label="Enter your location (latitude,longitude)", placeholder="e.g., 21.3,-157.8")

# Gradio interface definition
demo = gr.Interface(
    user_interface,
    [gr.Textbox(label="Message"), location_input, "state"],
    [chatbot, "state", "text", "text", "json"],
    allow_flagging="never",
    title="Mental Health & Well-being Assistant"
)

# Launch Gradio interface
if __name__ == "__main__":
    demo.launch()