import openai
import os
import gradio as gr
from dotenv import load_dotenv
import io
from PIL import Image

# Load environment variables (where your OpenAI key will be stored)
load_dotenv()

# Load the OpenAI API key from environment variables and strip any trailing newlines or spaces
openai.api_key = os.getenv("OPENAI_API_KEY").strip()

# Function to analyze the ad image using GPT-4-turbo's multimodal capabilities
def analyze_ad(image):
    # Convert the PIL image to bytes
    image_bytes = io.BytesIO()
    image.save(image_bytes, format='PNG')
    image_bytes = image_bytes.getvalue()

    # Placeholder for ad text extracted from the image (since GPT-4-turbo doesn't support image-to-text natively)
    # In a real scenario, you could integrate OCR to extract text or pass the image to a vision model.
    ad_text = "This is a placeholder for text extracted from the image."

    # Prompt for the marketing persona and scoring rubric
    prompt = f"""
    Analyze this advertisement and generate a marketing persona. The text in the ad says:
    '{ad_text}'
    Then, provide a score (out of 10) for each of the following:

    1. Relevance to Target Audience: Is the ad appealing to the intended demographic?
    2. Emotional Engagement: Does the ad evoke the right emotional response?
    3. Brand Consistency: Does the ad align with the brand’s voice and values?
    4. Creativity: How unique or innovative is the ad's design and text approach?
    5. Persuasiveness: Does the ad motivate action, such as clicking or purchasing?

    Provide the persona description and the scores in table form with a final score.
    """

    # Send the prompt to GPT-4-turbo for analysis
    response = openai.ChatCompletion.create(
        model="gpt-4-turbo",  # Use GPT-4 turbo model
        messages=[
            {"role": "system", "content": "You are a marketing expert analyzing an advertisement."},
            {"role": "user", "content": prompt}
        ],
        temperature=0.7,
        max_tokens=500
    )

    # Extract the response text from the API output
    result = response['choices'][0]['message']['content'].strip()

    # Return the result for display
    return result

# Function to process the image and run the analysis
def upload_and_analyze(image):
    # Pass the uploaded image to the analyze_ad function
    result = analyze_ad(image)
    
    return result

# Gradio interface for Hugging Face deployment
iface = gr.Interface(
    fn=upload_and_analyze,
    inputs=gr.Image(type="pil", label="Upload Advertisement Image"),  # Use type="pil" for the image input
    outputs=gr.Textbox(label="Marketing Persona and Ad Analysis"),
    title="Advertisement Persona and Scoring Analyzer",
    description="Upload an advertisement image, and the app will generate marketing personas and evaluate the ad based on Relevance, Emotional Engagement, Brand Consistency, Creativity, and Persuasiveness."
)

# Launch the app
if __name__ == "__main__":
    iface.launch()