import os
import re
import streamlit as st
import google.generativeai as genai
from transformers import pipeline, AutoModelForSequenceClassification, AutoTokenizer
from db import get_entry_by_index

# Configure Gemini API key
GEMINI_API_KEY = os.getenv("gemini_api")
if GEMINI_API_KEY:
    genai.configure(api_key=GEMINI_API_KEY)
else:
    st.error("⚠️ Google API key is missing! Set it in Hugging Face Secrets.")

# Load pre-trained sentiment analysis model
MODEL_NAME = "cardiffnlp/twitter-roberta-base-sentiment"
try:
    tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
    model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
    sentiment_pipeline = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)
except Exception as e:
    st.error(f"❌ Error loading sentiment model: {e}")

# Load Topic Extraction Model
try:
    topic_pipeline = pipeline("zero-shot-classification", model="facebook/bart-large-mnli")
except Exception as e:
    st.error(f"❌ Error loading topic extraction model: {e}")

# Predefined topic labels for classification
TOPIC_LABELS = [
    "Technology", "Politics", "Business", "Sports", "Entertainment",
    "Health", "Science", "Education", "Finance", "Travel", "Food"
]

def analyze_sentiment(text):
    try:
        result = sentiment_pipeline(text)[0]
        label = result['label']
        score = result['score']
        mapping = {"LABEL_0": "Negative", "LABEL_1": "Neutral", "LABEL_2": "Positive"}
        return mapping.get(label, "Unknown"), score
    except Exception as e:
        return f"Error analyzing sentiment: {e}", None

def extract_topic(text):
    try:
        result = topic_pipeline(text, TOPIC_LABELS)
        top_topic = result["labels"][0]
        confidence = result["scores"][0]
        return top_topic, confidence
    except Exception as e:
        return f"Error extracting topic: {e}", None

# Detect queries like "data entry 1" or "entry 3" (case-insensitive)
def is_entry_query(prompt):
    pattern = r"(?:data entry|entry)\s*(\d+)"
    match = re.search(pattern, prompt, re.IGNORECASE)
    if match:
        # Convert to index (assuming user numbering starts at 1)
        index = int(match.group(1)) - 1
        return True, index
    return False, None

def chatbot_response(user_prompt):
    if not user_prompt:
        return None, None, None, None, None
    try:
        entry_query, index = is_entry_query(user_prompt)
        if entry_query:
            entry = get_entry_by_index(index)
            if entry is None:
                return "❌ No entry found for the requested index.", None, None, None, None
            entry_text = entry.get("text", "No text available.")
            # Fixed AI response for entry queries (as per instructions)
            ai_response_text = "Let's break down this tweet-like MongoDB entry:"
            # Analyze the entry's text
            sentiment_label, sentiment_confidence = analyze_sentiment(entry_text)
            topic_label, topic_confidence = extract_topic(entry_text)
            return ai_response_text, sentiment_label, sentiment_confidence, topic_label, topic_confidence
        else:
            # For non-entry queries, fallback to the generative model as usual.
            model_gen = genai.GenerativeModel("gemini-1.5-pro")
            ai_response = model_gen.generate_content(user_prompt)
            sentiment_label, sentiment_confidence = analyze_sentiment(user_prompt)
            topic_label, topic_confidence = extract_topic(user_prompt)
            return ai_response.text, sentiment_label, sentiment_confidence, topic_label, topic_confidence
    except Exception as e:
        return f"❌ Error: {e}", None, None, None, None