import gradio as gr
from openai import OpenAI
import os

ACCESS_TOKEN = os.getenv("HF_TOKEN")

client = OpenAI(
    base_url="https://api-inference.huggingface.co/v1/", 
    api_key=ACCESS_TOKEN,
)

*def* *generate_study_material*(
    topic, 
    difficulty, 
    question_type, 
    focus_areas, 
    anxiety_level,
    num_questions
):
    # Customize prompt based on anxiety level and learning focus
    anxiety_prompts = {
        "High": "Create a gradual, confidence-building set of questions. Start with easier concepts and progressively increase difficulty. Include encouraging notes.",
        "Medium": "Balance challenge with achievability. Include hints for tougher questions and positive reinforcement.",
        "Low": "Focus on comprehensive concept testing while maintaining an encouraging tone."
    }
    
    focus_prompt = {
        "concept_understanding": "Emphasize questions that test deep understanding rather than memorization.",
        "problem_solving": "Include scenario-based questions that require analytical thinking.",
        "quick_recall": "Focus on key definitions and fundamental concepts.",
        "practical_application": "Create questions based on real-world applications."
    }

    base_prompt = f"""
    Act as an expert educational psychologist and subject matter expert creating an exam preparation guide.
    Topic: {topic}
    Difficulty: {difficulty}
    Question Type: {question_type}
    Number of Questions: {num_questions}
    
    Special Considerations:
    - Anxiety Level: {anxiety_level}
    {anxiety_prompts[anxiety_level]}
    
    - Learning Focus: {focus_areas}
    {focus_prompt[focus_areas]}
    Generate questions following these guidelines:
    1. Start with a brief confidence-building message
    2. Include clear, unambiguous questions
    3. Provide detailed explanations for each answer
    4. Add study tips relevant to the topic
    5. Include a "Remember" section with key points
    
    Format:
    - For Multiple Choice: Include 4 options with explanations for each
    - For Short Answer: Provide structure hints and model answers
    - For Descriptive: Break down marking criteria and include outline points
    
    Additional Requirements:
    - Include think-aloud strategies for problem-solving
    - Add time management suggestions
    - Highlight common misconceptions to avoid
    - End with a positive reinforcement message
    """

    *try*:
        messages = [
            {"role": "system", "content": "You are an expert educational content generator."},
            {"role": "user", "content": base_prompt}
        ]
        
        response = client.chat.completions.create(
            model="Qwen/QwQ-32B-Preview",
            messages=messages,
            max_tokens=2048,
            temperature=0.7,
            top_p=0.9
        )
        
        *return* response.choices[0].message.content
    
    *except* *Exception* *as* e:
        *return* f"An error occurred: {*str*(e)}\nPlease try again with different parameters."

*def* *create_interface*():
    *with* gr.Blocks(theme=gr.themes.Soft(primary_hue="blue")) *as* iface:
        gr.Markdown("""
        # <div align="center"><strong>📚 Exam Preparation Assistant</strong></div>
        
        Welcome to your personalized exam preparation assistant! This tool is designed to help you:
        - Build confidence through practiced learning
        - Understand concepts deeply
        - Reduce exam anxiety through structured practice
        
        Remember: Every practice session brings you closer to mastery! 🌟
        """)
        
        *with* gr.Row():
            *with* gr.Column():
                topic = gr.Textbox(
                    label="Topic or Subject",
                    placeholder="Enter the topic you want to study (e.g., 'Python Lists and Tuples', 'Chemical Bonding')",
                    lines=2
                )
                
                difficulty = gr.Radio(
                    choices=["Beginner", "Intermediate", "Advanced"],
                    label="Difficulty Level",
                    value="Intermediate",
                    info="Choose based on your current understanding"
                )
                
                question_type = gr.Radio(
                    choices=["Multiple Choice", "Short Answer", "Descriptive"],
                    label="Question Type",
                    value="Multiple Choice",
                    info="Select the format that best helps your learning"
                )
                
                focus_areas = gr.Radio(
                    choices=[
                        "concept_understanding",
                        "problem_solving",
                        "quick_recall",
                        "practical_application"
                    ],
                    label="Learning Focus",
                    value="concept_understanding",
                    info="What aspect do you want to improve?"
                )
                
                anxiety_level = gr.Radio(
                    choices=["High", "Medium", "Low"],
                    label="Current Anxiety Level",
                    value="Medium",
                    info="This helps us adjust the difficulty progression"
                )
                
                num_questions = gr.Slider(
                    minimum=3,
                    maximum=10,
                    value=5,
                    step=1,
                    label="Number of Questions"
                )
                
                submit_btn = gr.Button(
                    "Generate Study Material",
                    variant="primary"
                )

            *with* gr.Column():
                output = gr.Textbox(
                    label="Your Personalized Study Material",
                    lines=20,
                    show_copy_button=True
                )
        
        
        submit_btn.click(
            *fn*=generate_study_material,
            inputs=[
                topic,
                difficulty,
                question_type,
                focus_areas,
                anxiety_level,
                num_questions
            ],
            outputs=output
        )
    
    *return* iface

*if* **name** == "__main__":
    iface = create_interface()
    iface.launch()