import streamlit as st
from textblob import TextBlob
import nltk
from nltk.sentiment.vader import SentimentIntensityAnalyzer
from transformers import pipeline

# Ensure necessary NLTK data is downloaded
nltk.download('vader_lexicon')

# Sidebar Navigation
st.sidebar.title("🧠 Mental Health Analyzer")
page = st.sidebar.radio("🌟 Navigate", ["❤️ Importance of Mental Health", "🧩 ML-Based Analysis", "🤖 Transformer Tips"])

# Page 1: Importance of Mental Health
if page == "❤️ Importance of Mental Health":
    st.title("🌼 Why Mental Health Matters")
    st.write("""
Mental health is crucial for overall well-being. It influences how we think, feel, and act. 🧠💬
Good mental health can:
- 🌟 Improve productivity and performance
- 🤝 Enhance relationships
- 💪 Increase self-esteem and confidence
- 🧘 Help manage emotions and stress

Remember, it's okay to seek help and talk about your feelings. 🤗
    """)

# Page 2: ML-Based Analysis
elif page == "🧩 ML-Based Analysis":
    st.title("🧪 Analyze Your Mental State")
    user_input = st.text_area("✍️ Share your current thoughts or feelings:")
    if st.button("🔍 Analyze"):
        if user_input:
            st.subheader("📈 Analysis Results")

            # TextBlob Sentiment
            blob = TextBlob(user_input)
            polarity = blob.sentiment.polarity
            subjectivity = blob.sentiment.subjectivity
            st.write(f"**🧪 TextBlob Polarity:** {polarity:.2f}")
            st.write(f"**🔬 TextBlob Subjectivity:** {subjectivity:.2f}")

            # VADER Sentiment
            sid = SentimentIntensityAnalyzer()
            vader_scores = sid.polarity_scores(user_input)
            st.write(f"**⚡ VADER Compound Score:** {vader_scores['compound']:.2f}")

            # Emotion Detection with Transformers
            emotion_pipe = pipeline("text-classification", model="bhadresh-savani/distilbert-base-uncased-emotion", top_k=2)
            emotions = emotion_pipe(user_input)
            st.write("**🎭 Emotion Predictions:**")
            for inner_list in emotions:
             for emo in inner_list:
                st.write(f"- {emo['label']} (Score: {emo['score']:.2f})")

            # Tips based on polarity
            st.subheader("🌈 Tips to Improve Your Mood")
            if polarity < -0.2 or vader_scores['compound'] < -0.2:
                st.write("- 🧘‍♀️ Practice mindfulness and deep breathing exercises.")
                st.write("- 👥 Talk to a trusted friend or family member.")
                st.write("- 🚶 Engage in light physical activity, like a short walk.")
            elif polarity > 0.2 and vader_scores['compound'] > 0.2:
                st.write("- 📖 Keep up with activities that make you happy.")
                st.write("- ✍️ Share your positive feelings in a gratitude journal.")
                st.write("- 🤗 Continue connecting with supportive people.")
            else:
                st.write("- 🧩 Take breaks and reflect on your feelings.")
                st.write("- 📝 Try journaling or meditation to gain clarity.")
                st.write("- 🛌 Maintain a balanced routine of work and rest.")
        else:
            st.warning("⚠️ Please enter some text to analyze.")

# Page 3: Transformer-Based Tips Generator
else:
    st.title("🤖 Get Personalized Tips")
    user_input = st.text_area("📝 Describe how you're feeling or what you need tips on:")
    if st.button("🚀 Generate Tips"):
        if user_input:
            gen = pipeline("text-generation", model="gpt2", max_length=150)
            prompt = f"As a supportive mental health assistant, provide tips for someone feeling: {user_input}\nTips:"
            result = gen(prompt, num_return_sequences=1)[0]['generated_text']
            st.write(result.replace(prompt, "").strip())
        else:
            st.warning("⚠️ Please share your thoughts to get tips.")