import requests
from bs4 import BeautifulSoup
import pandas as pd
import gradio as gr
from groq import Groq

# Step 1: Scrape free courses from Analytics Vidhya
def fetch_free_courses():
    url = "https://courses.analyticsvidhya.com/pages/all-free-courses"
    response = requests.get(url)
    soup = BeautifulSoup(response.content, 'html.parser')

    courses_data = []
    
    # Extract course details
    for card in soup.select('header.course-card__img-container'):
        image_element = card.find('img', class_='course-card__img')
        
        if image_element:
            title = image_element.get('alt')
            img_url = image_element.get('src')
            
            link = card.find_previous('a')
            if link:
                course_link = link.get('href')
                if not course_link.startswith('http'):
                    course_link = 'https://courses.analyticsvidhya.com' + course_link

                courses_data.append({
                    'title': title,
                    'image_url': img_url,
                    'course_link': course_link
                })
    return courses_data

courses = fetch_free_courses()

# Step 2: Load data into a DataFrame
df = pd.DataFrame(courses)

client = Groq()

# Course search function using Groq
def course_recommendation(query):
    try:
        print(f"Search query: {query}")
        print(f"Total available courses: {len(df)}")

        # Prompt construction for Groq
        prompt = f"""
        Based on the query: "{query}",
        Rank the courses below based on relevance (0 to 1), with 1 being highly relevant.
        Filter out courses with relevance scores below 0.5.
        
        Courses:
        {df['title'].to_string(index=False)}
        """

        print("Sending query to Groq for recommendation...")
        # Sending the request to Groq for results
        response = client.chat.completions.create(
            model="mixtral-8x7b-32768",
            messages=[
                {"role": "system", "content": "You are a course recommendation assistant."},
                {"role": "user", "content": prompt}
            ],
            temperature=0.3,
            max_tokens=800
        )
        print("Response received from Groq.")

        # Parse the Groq response
        recommended_courses = []
        content = response.choices[0].message.content
        print("Groq's response:\n", content)

        for line in content.split('\n'):
            if line.startswith('Title:'):
                course_title = line.split('Title:')[1].strip()
            elif line.startswith('Relevance:'):
                score = float(line.split('Relevance:')[1].strip())
                if score >= 0.5:
                    matching_course = df[df['title'] == course_title]
                    if not matching_course.empty:
                        course_data = matching_course.iloc[0]
                        recommended_courses.append({
                            'title': course_title,
                            'image_url': course_data['image_url'],
                            'course_link': course_data['course_link'],
                            'score': score
                        })
        
        return sorted(recommended_courses, key=lambda x: x['score'], reverse=True)[:10]

    except Exception as e:
        print(f"Error during course search: {e}")
        return []

# Gradio function to search and display courses
def gradio_search_interface(query):
    results = course_recommendation(query)
    
    if results:
        html_output = '<div class="results-section">'
        for course in results:
            html_output += f"""
            <div class="course-item">
                <img src="{course['image_url']}" alt="{course['title']}" class="course-thumbnail"/>
                <div class="course-details">
                    <h4>{course['title']}</h4>
                    <p>Relevance: {round(course['score'] * 100, 2)}%</p>
                    <a href="{course['course_link']}" target="_blank" class="course-link-button">Explore Course</a>
                </div>
            </div>"""
        html_output += '</div>'
        return html_output
    else:
        return '<p class="no-courses-message">No matching courses found. Try another search.</p>'

# Custom CSS to make the interface attractive
custom_css = """
body {
    background-color: #eaeef3;
    font-family: 'Montserrat', sans-serif;
}
.results-section {
    display: flex;
    flex-wrap: wrap;
    gap: 20px;
}
.course-item {
    background-color: white;
    border-radius: 12px;
    box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
    overflow: hidden;
    width: 48%;
    transition: transform 0.3s ease;
}
.course-item:hover {
    transform: translateY(-10px);
}
.course-thumbnail {
    width: 100%;
    height: 160px;
    object-fit: cover;
}
.course-details {
    padding: 15px;
    text-align: center;
}
.course-details h4 {
    font-size: 18px;
    color: #333;
    margin: 10px 0;
}
.course-details p {
    color: #555;
    font-size: 14px;
}
.course-link-button {
    display: inline-block;
    background-color: #ff5733;
    color: white;
    padding: 8px 16px;
    text-decoration: none;
    border-radius: 6px;
    margin-top: 10px;
}
.course-link-button:hover {
    background-color: #c44524;
}
.no-courses-message {
    text-align: center;
    color: #777;
    font-size: 16px;
}
"""

# Setting up the Gradio interface
iface = gr.Interface(
    fn=gradio_search_interface,
    inputs=gr.Textbox(label="Search for a course", placeholder="e.g., Python for data analysis, ML basics"),
    outputs=gr.HTML(label="Course Results"),
    title="Analytics Vidhya Course Finder",
    description="Discover the best free courses from Analytics Vidhya tailored to your query.",
    theme="compact",
    css=custom_css,
    examples=[
        ["Data Science for Beginners"],
        ["Python Programming"],
        ["Advanced Machine Learning"],
        ["Business Analytics"],
    ]
)

# Run the Gradio interface
if __name__ == "__main__":
    iface.launch()