src/app.py

# import streamlit as st
# import torch
# import pandas as pd
# import numpy as np
# from pathlib import Path
# import sys
# import plotly.express as px
# import plotly.graph_objects as go
# from transformers import BertTokenizer
# import nltk

# # Download required NLTK data
# try:
#     nltk.data.find('tokenizers/punkt')
# except LookupError:
#     nltk.download('punkt')
# try:
#     nltk.data.find('corpora/stopwords')
# except LookupError:
#     nltk.download('stopwords')
# try:
#     nltk.data.find('tokenizers/punkt_tab')
# except LookupError:
#     nltk.download('punkt_tab')
# try:
#     nltk.data.find('corpora/wordnet')
# except LookupError:
#     nltk.download('wordnet')

# # Add project root to Python path
# project_root = Path(__file__).parent.parent
# sys.path.append(str(project_root))

# from src.models.hybrid_model import HybridFakeNewsDetector
# from src.config.config import *
# from src.data.preprocessor import TextPreprocessor

# # Page config is set in main app.py

# @st.cache_resource
# def load_model_and_tokenizer():
#     """Load the model and tokenizer (cached)."""
#     # Initialize model
#     model = HybridFakeNewsDetector(
#         bert_model_name=BERT_MODEL_NAME,
#         lstm_hidden_size=LSTM_HIDDEN_SIZE,
#         lstm_num_layers=LSTM_NUM_LAYERS,
#         dropout_rate=DROPOUT_RATE
#     )
    
#     # Load trained weights
#     state_dict = torch.load(SAVED_MODELS_DIR / "final_model.pt", map_location=torch.device('cpu'))
    
#     # Filter out unexpected keys
#     model_state_dict = model.state_dict()
#     filtered_state_dict = {k: v for k, v in state_dict.items() if k in model_state_dict}
    
#     # Load the filtered state dict
#     model.load_state_dict(filtered_state_dict, strict=False)
#     model.eval()
    
#     # Initialize tokenizer
#     tokenizer = BertTokenizer.from_pretrained(BERT_MODEL_NAME)
    
#     return model, tokenizer

# @st.cache_resource
# def get_preprocessor():
#     """Get the text preprocessor (cached)."""
#     return TextPreprocessor()

# def predict_news(text):
#     """Predict if the given news is fake or real."""
#     # Get model, tokenizer, and preprocessor from cache
#     model, tokenizer = load_model_and_tokenizer()
#     preprocessor = get_preprocessor()
    
#     # Preprocess text
#     processed_text = preprocessor.preprocess_text(text)
    
#     # Tokenize
#     encoding = tokenizer.encode_plus(
#         processed_text,
#         add_special_tokens=True,
#         max_length=MAX_SEQUENCE_LENGTH,
#         padding='max_length',
#         truncation=True,
#         return_attention_mask=True,
#         return_tensors='pt'
#     )
    
#     # Get prediction
#     with torch.no_grad():
#         outputs = model(
#             encoding['input_ids'],
#             encoding['attention_mask']
#         )
#         probabilities = torch.softmax(outputs['logits'], dim=1)
#         prediction = torch.argmax(outputs['logits'], dim=1)
#         attention_weights = outputs['attention_weights']
    
#     # Convert attention weights to numpy and get the first sequence
#     attention_weights_np = attention_weights[0].cpu().numpy()
    
#     return {
#         'prediction': prediction.item(),
#         'label': 'FAKE' if prediction.item() == 1 else 'REAL',
#         'confidence': torch.max(probabilities, dim=1)[0].item(),
#         'probabilities': {
#             'REAL': probabilities[0][0].item(),
#             'FAKE': probabilities[0][1].item()
#         },
#         'attention_weights': attention_weights_np
#     }

# def plot_confidence(probabilities):
#     """Plot prediction confidence."""
#     fig = go.Figure(data=[
#         go.Bar(
#             x=list(probabilities.keys()),
#             y=list(probabilities.values()),
#             text=[f'{p:.2%}' for p in probabilities.values()],
#             textposition='auto',
#         )
#     ])
    
#     fig.update_layout(
#         title='Prediction Confidence',
#         xaxis_title='Class',
#         yaxis_title='Probability',
#         yaxis_range=[0, 1]
#     )
    
#     return fig

# def plot_attention(text, attention_weights):
#     """Plot attention weights."""
#     tokens = text.split()
#     attention_weights = attention_weights[:len(tokens)]  # Truncate to match tokens
    
#     # Ensure attention weights are in the correct format
#     if isinstance(attention_weights, (list, np.ndarray)):
#         attention_weights = np.array(attention_weights).flatten()
    
#     # Format weights for display
#     formatted_weights = [f'{float(w):.2f}' for w in attention_weights]
    
#     fig = go.Figure(data=[
#         go.Bar(
#             x=tokens,
#             y=attention_weights,
#             text=formatted_weights,
#             textposition='auto',
#         )
#     ])
    
#     fig.update_layout(
#         title='Attention Weights',
#         xaxis_title='Tokens',
#         yaxis_title='Attention Weight',
#         xaxis_tickangle=45
#     )
    
#     return fig

# def main():
#     st.title("📰 Fake News Detection System")
#     st.write("""
#     This application uses a hybrid deep learning model (BERT + BiLSTM + Attention) 
#     to detect fake news articles. Enter a news article below to analyze it.
#     """)
    
#     # Sidebar
#     st.sidebar.title("About")
#     st.sidebar.info("""
    
#     The model combines:
#     - BERT for contextual embeddings
#     - BiLSTM for sequence modeling
#     - Attention mechanism for interpretability
#     """)
    
#     # Main content
#     st.header("News Analysis")
    
#     # Text input
#     news_text = st.text_area(
#         "Enter the news article to analyze:",
#         height=200,
#         placeholder="Paste your news article here..."
#     )
    
#     if st.button("Analyze"):
#         if news_text:
#             with st.spinner("Analyzing the news article..."):
#                 # Get prediction
#                 result = predict_news(news_text)
                
#                 # Display result
#                 col1, col2 = st.columns(2)
                
#                 with col1:
#                     st.subheader("Prediction")
#                     if result['label'] == 'FAKE':
#                         st.error(f"🔴 This news is likely FAKE (Confidence: {result['confidence']:.2%})")
#                     else:
#                         st.success(f"🟢 This news is likely REAL (Confidence: {result['confidence']:.2%})")
                
#                 with col2:
#                     st.subheader("Confidence Scores")
#                     st.plotly_chart(plot_confidence(result['probabilities']), use_container_width=True)
                
#                 # Show attention visualization
#                 st.subheader("Attention Analysis")
#                 st.write("""
#                 The attention weights show which parts of the text the model focused on 
#                 while making its prediction. Higher weights indicate more important tokens.
#                 """)
#                 st.plotly_chart(plot_attention(news_text, result['attention_weights']), use_container_width=True)
                
#                 # Show model explanation
#                 st.subheader("Model Explanation")
#                 if result['label'] == 'FAKE':
#                     st.write("""
#                     The model identified this as fake news based on:
#                     - Linguistic patterns typical of fake news
#                     - Inconsistencies in the content
#                     - Attention weights on suspicious phrases
#                     """)
#                 else:
#                     st.write("""
#                     The model identified this as real news based on:
#                     - Credible language patterns
#                     - Consistent information
#                     - Attention weights on factual statements
#                     """)
#         else:
#             st.warning("Please enter a news article to analyze.")

# if __name__ == "__main__":
#     main() 

import streamlit as st
import torch
import pandas as pd
import numpy as np
from pathlib import Path
import sys
import plotly.express as px
import plotly.graph_objects as go
from transformers import BertTokenizer
import nltk

# Download required NLTK data
try:
    nltk.data.find('tokenizers/punkt')
except LookupError:
    nltk.download('punkt')
try:
    nltk.data.find('corpora/stopwords')
except LookupError:
    nltk.download('stopwords')
try:
    nltk.data.find('tokenizers/punkt_tab')
except LookupError:
    nltk.download('punkt_tab')
try:
    nltk.data.find('corpora/wordnet')
except LookupError:
    nltk.download('wordnet')

# Add project root to Python path
project_root = Path(__file__).parent.parent
sys.path.append(str(project_root))

from src.models.hybrid_model import HybridFakeNewsDetector
from src.config.config import *
from src.data.preprocessor import TextPreprocessor

# Set page config
st.set_page_config(
    page_title="TrueCheck - AI Fake News Detector",
    page_icon="🔍",
    layout="wide",
    initial_sidebar_state="collapsed"
)

# Custom CSS for modern styling
st.markdown("""
<style>
    /* Import Google Fonts */
    @import url('https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600;700&display=swap');
    
    /* Global Styles */
    .main {
        padding: 0;
    }
    
    .stApp {
        font-family: 'Inter', sans-serif;
        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
        min-height: 100vh;
    }
    
    /* Hide Streamlit elements */
    #MainMenu {visibility: hidden;}
    footer {visibility: hidden;}
    .stDeployButton {display: none;}
    header {visibility: hidden;}
    
    /* Hero Section */
    .hero-container {
        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
        padding: 4rem 2rem;
        text-align: center;
        color: white;
        margin-bottom: 2rem;
    }
    
    .hero-title {
        font-size: 4rem;
        font-weight: 700;
        margin-bottom: 1rem;
        text-shadow: 2px 2px 4px rgba(0,0,0,0.3);
        background: linear-gradient(45deg, #fff, #e0e7ff);
        -webkit-background-clip: text;
        -webkit-text-fill-color: transparent;
        background-clip: text;
    }
    
    .hero-subtitle {
        font-size: 1.3rem;
        font-weight: 400;
        margin-bottom: 2rem;
        opacity: 0.9;
        max-width: 600px;
        margin-left: auto;
        margin-right: auto;
        line-height: 1.6;
    }
    
    /* Features Section */
    .features-container {
        background: white;
        padding: 3rem 2rem;
        margin: 2rem 0;
        border-radius: 20px;
        box-shadow: 0 20px 40px rgba(0,0,0,0.1);
    }
    
    .features-grid {
        display: grid;
        grid-template-columns: repeat(auto-fit, minmax(300px, 1fr));
        gap: 2rem;
        margin-top: 2rem;
    }
    
    .feature-card {
        background: linear-gradient(135deg, #f8fafc 0%, #e2e8f0 100%);
        padding: 2rem;
        border-radius: 16px;
        text-align: center;
        transition: transform 0.3s ease, box-shadow 0.3s ease;
        border: 1px solid #e2e8f0;
    }
    
    .feature-card:hover {
        transform: translateY(-10px);
        box-shadow: 0 20px 40px rgba(0,0,0,0.15);
    }
    
    .feature-icon {
        font-size: 3rem;
        margin-bottom: 1rem;
        display: block;
    }
    
    .feature-title {
        font-size: 1.2rem;
        font-weight: 600;
        color: #1e293b;
        margin-bottom: 0.5rem;
    }
    
    .feature-description {
        color: #64748b;
        line-height: 1.5;
        font-size: 0.95rem;
    }
    
    /* Main Content Section */
    .main-content {
        background: white;
        padding: 3rem;
        border-radius: 20px;
        box-shadow: 0 20px 40px rgba(0,0,0,0.1);
        margin: 2rem 0;
    }
    
    .section-title {
        font-size: 2.5rem;
        font-weight: 700;
        text-align: center;
        color: #1e293b;
        margin-bottom: 1rem;
    }
    
    .section-description {
        text-align: center;
        color: #64748b;
        font-size: 1.1rem;
        margin-bottom: 2rem;
        max-width: 600px;
        margin-left: auto;
        margin-right: auto;
        line-height: 1.6;
    }
    
    /* Input Section */
    .stTextArea > div > div > textarea {
        border-radius: 12px;
        border: 2px solid #e2e8f0;
        padding: 1rem;
        font-size: 1rem;
        transition: border-color 0.3s ease;
        font-family: 'Inter', sans-serif;
    }
    
    .stTextArea > div > div > textarea:focus {
        border-color: #667eea;
        box-shadow: 0 0 0 3px rgba(102, 126, 234, 0.1);
    }
    
    /* Button Styling */
    .stButton > button {
        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
        color: white;
        border: none;
        border-radius: 12px;
        padding: 0.75rem 2rem;
        font-size: 1.1rem;
        font-weight: 600;
        font-family: 'Inter', sans-serif;
        transition: all 0.3s ease;
        box-shadow: 0 4px 15px rgba(102, 126, 234, 0.4);
        width: 100%;
    }
    
    .stButton > button:hover {
        transform: translateY(-2px);
        box-shadow: 0 8px 25px rgba(102, 126, 234, 0.6);
    }
    
    /* Results Section */
    .result-card {
        background: linear-gradient(135deg, #f8fafc 0%, #e2e8f0 100%);
        padding: 2rem;
        border-radius: 16px;
        margin: 1rem 0;
        box-shadow: 0 4px 15px rgba(0,0,0,0.1);
    }
    
    .success-message {
        background: linear-gradient(135deg, #dcfce7 0%, #bbf7d0 100%);
        color: #166534;
        padding: 1rem 1.5rem;
        border-radius: 12px;
        border-left: 4px solid #22c55e;
        font-weight: 500;
        margin: 1rem 0;
    }
    
    .error-message {
        background: linear-gradient(135deg, #fef2f2 0%, #fecaca 100%);
        color: #991b1b;
        padding: 1rem 1.5rem;
        border-radius: 12px;
        border-left: 4px solid #ef4444;
        font-weight: 500;
        margin: 1rem 0;
    }
    
    /* Footer */
    .footer {
        background: linear-gradient(135deg, #1e293b 0%, #334155 100%);
        color: white;
        padding: 3rem 2rem 2rem;
        text-align: center;
        margin-top: 4rem;
    }
    
    .footer-content {
        max-width: 1200px;
        margin: 0 auto;
    }
    
    .footer-title {
        font-size: 1.5rem;
        font-weight: 600;
        margin-bottom: 1rem;
    }
    
    .footer-text {
        color: #94a3b8;
        margin-bottom: 2rem;
        line-height: 1.6;
    }
    
    .footer-links {
        display: flex;
        justify-content: center;
        gap: 2rem;
        margin-bottom: 2rem;
    }
    
    .footer-link {
        color: #94a3b8;
        text-decoration: none;
        transition: color 0.3s ease;
    }
    
    .footer-link:hover {
        color: white;
    }
    
    .footer-bottom {
        border-top: 1px solid #475569;
        padding-top: 2rem;
        color: #94a3b8;
        font-size: 0.9rem;
    }
    
    /* Responsive Design */
    @media (max-width: 768px) {
        .hero-title {
            font-size: 3rem;
        }
        
        .features-grid {
            grid-template-columns: 1fr;
        }
        
        .main-content {
            padding: 2rem;
        }
        
        .footer-links {
            flex-direction: column;
            gap: 1rem;
        }
    }
</style>
""", unsafe_allow_html=True)

@st.cache_resource
def load_model_and_tokenizer():
    """Load the model and tokenizer (cached)."""
    model = HybridFakeNewsDetector(
        bert_model_name=BERT_MODEL_NAME,
        lstm_hidden_size=LSTM_HIDDEN_SIZE,
        lstm_num_layers=LSTM_NUM_LAYERS,
        dropout_rate=DROPOUT_RATE
    )
    state_dict = torch.load(SAVED_MODELS_DIR / "final_model.pt", map_location=torch.device('cpu'))
    model_state_dict = model.state_dict()
    filtered_state_dict = {k: v for k, v in state_dict.items() if k in model_state_dict}
    model.load_state_dict(filtered_state_dict, strict=False)
    model.eval()
    tokenizer = BertTokenizer.from_pretrained(BERT_MODEL_NAME)
    return model, tokenizer

@st.cache_resource
def get_preprocessor():
    """Get the text preprocessor (cached)."""
    return TextPreprocessor()

def predict_news(text):
    """Predict if the given news is fake or real."""
    model, tokenizer = load_model_and_tokenizer()
    preprocessor = get_preprocessor()
    processed_text = preprocessor.preprocess_text(text)
    encoding = tokenizer.encode_plus(
        processed_text,
        add_special_tokens=True,
        max_length=MAX_SEQUENCE_LENGTH,
        padding='max_length',
        truncation=True,
        return_attention_mask=True,
        return_tensors='pt'
    )
    with torch.no_grad():
        outputs = model(
            encoding['input_ids'],
            encoding['attention_mask']
        )
        probabilities = torch.softmax(outputs['logits'], dim=1)
        prediction = torch.argmax(outputs['logits'], dim=1)
        attention_weights = outputs['attention_weights']
    attention_weights_np = attention_weights[0].cpu().numpy()
    return {
        'prediction': prediction.item(),
        'label': 'FAKE' if prediction.item() == 1 else 'REAL',
        'confidence': torch.max(probabilities, dim=1)[0].item(),
        'probabilities': {
            'REAL': probabilities[0][0].item(),
            'FAKE': probabilities[0][1].item()
        },
        'attention_weights': attention_weights_np
    }

def plot_confidence(probabilities):
    """Plot prediction confidence."""
    fig = go.Figure(data=[
        go.Bar(
            x=list(probabilities.keys()),
            y=list(probabilities.values()),
            text=[f'{p:.2%}' for p in probabilities.values()],
            textposition='auto',
            marker_color=['#22c55e', '#ef4444'],
            marker_line_color='rgba(0,0,0,0.1)',
            marker_line_width=1
        )
    ])
    fig.update_layout(
        title={
            'text': 'Prediction Confidence',
            'x': 0.5,
            'xanchor': 'center',
            'font': {'size': 18, 'family': 'Inter'}
        },
        xaxis_title='Class',
        yaxis_title='Probability',
        yaxis_range=[0, 1],
        template='plotly_white',
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font={'family': 'Inter'}
    )
    return fig

def plot_attention(text, attention_weights):
    """Plot attention weights."""
    tokens = text.split()
    attention_weights = attention_weights[:len(tokens)]
    if isinstance(attention_weights, (list, np.ndarray)):
        attention_weights = np.array(attention_weights).flatten()
    formatted_weights = [f'{float(w):.2f}' for w in attention_weights]
    
    # Create color scale based on attention weights
    colors = ['rgba(102, 126, 234, ' + str(0.3 + 0.7 * (w / max(attention_weights))) + ')' 
              for w in attention_weights]
    
    fig = go.Figure(data=[
        go.Bar(
            x=tokens,
            y=attention_weights,
            text=formatted_weights,
            textposition='auto',
            marker_color=colors,
            marker_line_color='rgba(102, 126, 234, 0.8)',
            marker_line_width=1
        )
    ])
    fig.update_layout(
        title={
            'text': 'Attention Weights Analysis',
            'x': 0.5,
            'xanchor': 'center',
            'font': {'size': 18, 'family': 'Inter'}
        },
        xaxis_title='Tokens',
        yaxis_title='Attention Weight',
        xaxis_tickangle=45,
        template='plotly_white',
        plot_bgcolor='rgba(0,0,0,0)',
        paper_bgcolor='rgba(0,0,0,0)',
        font={'family': 'Inter'}
    )
    return fig

def main():
    # Hero Section
    st.markdown("""
    <div class="hero-container">
        <h1 class="hero-title">🔍 TrueCheck</h1>
        <p class="hero-subtitle">
            Advanced AI-powered fake news detection using cutting-edge deep learning technology. 
            Get instant, accurate analysis of news articles with our hybrid BERT-BiLSTM model.
        </p>
    </div>
    """, unsafe_allow_html=True)

    # Features Section
    st.markdown("""
    <div class="features-container">
        <h2 style="text-align: center; font-size: 2rem; font-weight: 700; color: #1e293b; margin-bottom: 1rem;">
            Why Choose TrueCheck?
        </h2>
        <p style="text-align: center; color: #64748b; font-size: 1.1rem; margin-bottom: 2rem;">
            Our advanced AI model combines multiple technologies for superior accuracy
        </p>
        <div class="features-grid">
            <div class="feature-card">
                <span class="feature-icon">🤖</span>
                <h3 class="feature-title">BERT Technology</h3>
                <p class="feature-description">
                    Utilizes state-of-the-art BERT transformer for deep contextual understanding of news content
                </p>
            </div>
            <div class="feature-card">
                <span class="feature-icon">🧠</span>
                <h3 class="feature-title">BiLSTM Processing</h3>
                <p class="feature-description">
                    Bidirectional LSTM networks capture sequential patterns and dependencies in text structure
                </p>
            </div>
            <div class="feature-card">
                <span class="feature-icon">👁️</span>
                <h3 class="feature-title">Attention Mechanism</h3>
                <p class="feature-description">
                    Advanced attention layers provide interpretable insights into model decision-making process
                </p>
            </div>
        </div>
    </div>
    """, unsafe_allow_html=True)

    # Main Content Section
    st.markdown("""
    <div class="main-content">
        <h2 class="section-title">Analyze News Article</h2>
        <p class="section-description">
            Paste any news article below and our AI will analyze it for authenticity. 
            Get detailed insights including confidence scores and attention analysis.
        </p>
    </div>
    """, unsafe_allow_html=True)

    # Input Section
    col1, col2, col3 = st.columns([1, 3, 1])
    with col2:
        news_text = st.text_area(
            "",
            height=200,
            placeholder="📰 Paste your news article here for analysis...",
            key="news_input"
        )
        
        analyze_button = st.button("🔍 Analyze Article", key="analyze_button")

    if analyze_button:
        if news_text:
            with st.spinner("🤖 Analyzing the news article..."):
                result = predict_news(news_text)
                
                # Results Section
                st.markdown('<div class="main-content">', unsafe_allow_html=True)
                
                col1, col2 = st.columns([1, 1], gap="large")
                
                with col1:
                    st.markdown("### 📊 Prediction Result")
                    if result['label'] == 'FAKE':
                        st.markdown(f'''
                        <div class="error-message">
                            🔴 <strong>FAKE NEWS DETECTED</strong><br>
                            Confidence: {result["confidence"]:.2%}
                        </div>
                        ''', unsafe_allow_html=True)
                    else:
                        st.markdown(f'''
                        <div class="success-message">
                            🟢 <strong>AUTHENTIC NEWS</strong><br>
                            Confidence: {result["confidence"]:.2%}
                        </div>
                        ''', unsafe_allow_html=True)
                
                with col2:
                    st.markdown("### 📈 Confidence Breakdown")
                    st.plotly_chart(plot_confidence(result['probabilities']), use_container_width=True)
                
                st.markdown("### 🎯 Attention Analysis")
                st.markdown("""
                <p style="color: #64748b; text-align: center; margin-bottom: 2rem;">
                    The visualization below shows which words our AI model focused on while making its prediction. 
                    Darker colors indicate higher attention weights.
                </p>
                """, unsafe_allow_html=True)
                st.plotly_chart(plot_attention(news_text, result['attention_weights']), use_container_width=True)
                
                st.markdown("### 🔍 Detailed Analysis")
                if result['label'] == 'FAKE':
                    st.markdown("""
                    <div class="result-card">
                        <h4 style="color: #ef4444; margin-bottom: 1rem;">⚠️ Fake News Indicators</h4>
                        <ul style="color: #64748b; line-height: 1.8;">
                            <li><strong>Linguistic Patterns:</strong> The model detected language patterns commonly associated with misinformation</li>
                            <li><strong>Content Inconsistencies:</strong> Identified potential factual inconsistencies or misleading statements</li>
                            <li><strong>Attention Analysis:</strong> High attention weights on suspicious phrases and emotionally charged language</li>
                            <li><strong>Structural Analysis:</strong> Text structure and flow patterns typical of fabricated content</li>
                        </ul>
                        <p style="color: #7c3aed; font-weight: 500; margin-top: 1rem;">
                            💡 <strong>Recommendation:</strong> Verify this information through multiple reliable sources before sharing.
                        </p>
                    </div>
                    """, unsafe_allow_html=True)
                else:
                    st.markdown("""
                    <div class="result-card">
                        <h4 style="color: #22c55e; margin-bottom: 1rem;">✅ Authentic News Indicators</h4>
                        <ul style="color: #64748b; line-height: 1.8;">
                            <li><strong>Credible Language:</strong> Professional journalistic writing style and balanced reporting tone</li>
                            <li><strong>Factual Consistency:</strong> Information appears coherent and factually consistent</li>
                            <li><strong>Attention Analysis:</strong> Model focused on factual statements and objective reporting</li>
                            <li><strong>Structural Integrity:</strong> Well-structured content following standard news article format</li>
                        </ul>
                        <p style="color: #7c3aed; font-weight: 500; margin-top: 1rem;">
                            💡 <strong>Note:</strong> While likely authentic, always cross-reference important news from multiple sources.
                        </p>
                    </div>
                    """, unsafe_allow_html=True)
                
                st.markdown('</div>', unsafe_allow_html=True)
        else:
            st.markdown('''
            <div class="main-content">
                <div class="error-message" style="text-align: center;">
                    ⚠️ Please enter a news article to analyze
                </div>
            </div>
            ''', unsafe_allow_html=True)

    # Footer
    st.markdown("""
    <div class="footer">
        <div class="footer-content">
            <h3 class="footer-title">TrueCheck AI</h3>
            <p class="footer-text">
                Empowering users with AI-driven news verification technology. 
                Built with advanced deep learning models for accurate fake news detection.
            </p>
            <div class="footer-links">
                <a href="#" class="footer-link">About</a>
                <a href="#" class="footer-link">How It Works</a>
                <a href="#" class="footer-link">Privacy Policy</a>
                <a href="#" class="footer-link">Contact</a>
            </div>
            <div class="footer-bottom">
                <p>&copy; 2025 TrueCheck AI. Built with ❤️ using Streamlit, BERT, and PyTorch.</p>
                <p>Disclaimer: This tool provides AI-based analysis. Always verify important information through multiple sources.</p>
            </div>
        </div>
    </div>
    """, unsafe_allow_html=True)

if __name__ == "__main__":
    main()