import gradio as gr
import pickle as pk
import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
import os

# Get the absolute path of the current directory
current_dir = os.path.dirname(os.path.abspath(__file__))
model_path = os.path.join(current_dir, 'model.pkl')
scaler_path = os.path.join(current_dir, 'scaler.pkl')

def load_model_and_vectorizer():
    try:
        if not os.path.exists(model_path) or not os.path.exists(scaler_path):
            return None, None, f"Error: Model files not found at:\n{model_path}\n{scaler_path}"
        
        model = pk.load(open(model_path, 'rb'))
        vectorizer = pk.load(open(scaler_path, 'rb'))
        return model, vectorizer, None
    except Exception as e:
        return None, None, f"Error loading model: {str(e)}"

# Load the model and vectorizer
model, vectorizer, error = load_model_and_vectorizer()

def predict_sentiment(review):
    if error:
        return error
    
    if not review:
        return "Please enter a review first!"
    
    try:
        # Vectorize the text
        review_vector = vectorizer.transform([review])
        # Make prediction
        result = model.predict(review_vector)[0]
        
        if result == 0:
            return "Negative Review 😞"
        else:
            return "Positive Review 😊"
    except Exception as e:
        return f"Error during prediction: {str(e)}"

# Modern and aesthetic CSS adapted from the dashboard
css = """
body {
    background: linear-gradient(180deg, #2c3e50 0%, #3498db 100%);
    margin: 0;
    font-family: 'Roboto', -apple-system, BlinkMacSystemFont, sans-serif;
    color: #ecf0f1;
}

.gradio-container {
    max-width: 800px;
    margin: 3rem auto;
    background: rgba(44, 62, 80, 0.9);
    border-radius: 20px;
    padding: 2.5rem;
    box-shadow: 0 12px 40px rgba(0, 0, 0, 0.3);
    backdrop-filter: blur(10px);
    transition: transform 0.3s ease;
}

.gradio-container:hover {
    transform: translateY(-5px);
}

.header {
    text-align: center;
    margin-bottom: 2rem;
    animation: fadeIn 1s ease-in-out;
}

.header h1 {
    font-size: 2.4rem;
    background: linear-gradient(45deg, #e74c3c, #f1c40f);
    -webkit-background-clip: text;
    -webkit-text-fill-color: transparent;
    margin: 0;
    font-weight: 700;
}

.header p {
    font-size: 1.2rem;
    color: #bdc3c7;
    margin-top: 0.5rem;
    font-weight: 300;
}

.review-input {
    background: rgba(255, 255, 255, 0.1);
    border: 2px solid #4ECDC4;
    border-radius: 12px;
    padding: 1.5rem;
    font-size: 1.1rem;
    color: #ecf0f1;
    width: 100%;
    box-sizing: border-box;
    box-shadow: 0 4px 15px rgba(0, 0, 0, 0.2);
    transition: all 0.3s ease;
}

.review-input:focus {
    border-color: #f1c40f;
    box-shadow: 0 6px 20px rgba(241, 196, 15, 0.3);
    outline: none;
    transform: scale(1.01);
}

.review-input::placeholder {
    color: #95a5a6;
    font-style: italic;
}

.submit-btn {
    background: linear-gradient(45deg, #FF6B6B, #4ECDC4);
    color: white;
    border: none;
    padding: 1rem 3rem;
    border-radius: 50px;
    font-size: 1.1rem;
    font-weight: 600;
    cursor: pointer;
    margin: 1.5rem auto;
    display: block;
    transition: all 0.3s ease;
    box-shadow: 0 6px 20px rgba(78, 205, 196, 0.4);
}

.submit-btn:hover {
    background: linear-gradient(45deg, #4ECDC4, #FF6B6B);
    transform: translateY(-3px) scale(1.05);
    box-shadow: 0 10px 25px rgba(78, 205, 196, 0.5);
}

.submit-btn:active {
    transform: translateY(0);
    box-shadow: 0 4px 15px rgba(78, 205, 196, 0.3);
}

.output-container {
    margin-top: 2rem;
    padding: 1.5rem;
    background: rgba(255, 255, 255, 0.05);
    border-radius: 15px;
    text-align: center;
    min-height: 80px;
    font-size: 1.3rem;
    color: #f1c40f;
    box-shadow: 0 4px 15px rgba(0, 0, 0, 0.2);
    animation: slideUp 0.5s ease-out;
}

.error-message {
    background: rgba(231, 76, 60, 0.2);
    color: #e74c3c;
    padding: 1.2rem;
    border-radius: 12px;
    margin-bottom: 1.5rem;
    border-left: 5px solid #e74c3c;
    font-weight: 500;
    animation: shake 0.4s ease-in-out;
}

@keyframes fadeIn {
    from { opacity: 0; transform: translateY(-10px); }
    to { opacity: 1; transform: translateY(0); }
}

@keyframes slideUp {
    from { opacity: 0; transform: translateY(20px); }
    to { opacity: 1; transform: translateY(0); }
}

@keyframes shake {
    0%, 100% { transform: translateX(0); }
    20%, 60% { transform: translateX(-8px); }
    40%, 80% { transform: translateX(8px); }
}

@media (max-width: 600px) {
    .gradio-container {
        margin: 1rem;
        padding: 1.5rem;
    }
    
    .header h1 {
        font-size: 2rem;
    }
    
    .review-input {
        padding: 1rem;
    }
    
    .submit-btn {
        padding: 0.8rem 2rem;
    }
}
"""

# Create the Gradio interface
with gr.Blocks(
    title="Sentiment Analysis on Traveler Reviews",
    theme=gr.themes.Default(
        primary_hue="teal",
        secondary_hue="pink",
        neutral_hue="gray",
        radius_size="lg",
    ),
    css=css
) as demo:
    gr.Markdown("""
    <div class="header">
        <h1> Sentiment Analysis on Traveler Reviews</h1>
        <p>Type your review and let our AI uncover its vibe!</p>
    </div>
    """)
    
    if error:
        gr.Markdown(f"""
        <div class="error-message">
            ⚠️ {error}
        </div>
        """)
    
    review_input = gr.Textbox(
        placeholder="Share your thoughts about the your experience...",
        lines=5,
        elem_classes="review-input",
        show_label=False
    )
    
    predict_btn = gr.Button(
        "Analyze Sentiment",
        elem_classes="submit-btn"
    )
    
    output = gr.Textbox(
        placeholder="The sentiment will appear here...",
        elem_classes="output-container",
        show_label=False,
        interactive=False
    )
    
    predict_btn.click(
        fn=predict_sentiment,
        inputs=review_input,
        outputs=output
    )

if __name__ == "__main__":
    demo.launch(share=True)