app.py

import gradio as gr
import openai
import base64
import io
from PIL import Image
import fitz  # PyMuPDF
import os

# Load API key
openai.api_key = os.getenv("OPENAI_API_KEY")

# Prompt for extraction
prompt = """
You are analyzing a medical document or an application form from a patient. 
Extract the following fields as JSON:
- Position applied for
- Office/Ministry
- Duty station
- First name(s)
- Surname
- Date of birth
- Gender
- Citizenship
- Postal Address
- Residential Address
- Email
- Phone number (mobile)
"""

def process_pdf(pdf_file):
    # pdf_file is already bytes when using gr.File(type="binary")
    doc = fitz.open(stream=pdf_file, filetype="pdf")
    results = []

    for page_num in range(len(doc)):
        page = doc.load_page(page_num)
        pix = page.get_pixmap(dpi=200)  # Use 150-200 DPI for balance

        # Convert to PIL Image
        image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)

        # Convert to base64 JPEG
        buffered = io.BytesIO()
        image.save(buffered, format="JPEG")
        base64_image = base64.b64encode(buffered.getvalue()).decode()

        # Send to GPT-4o
        response = openai.chat.completions.create(
            model="gpt-4o",
            messages=[
                {"role": "user", "content": [
                    {"type": "text", "text": prompt},
                    {"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{base64_image}"}}
                ]}
            ],
            max_tokens=1000
        )

        results.append(response.choices[0].message.content.strip())

    return "\n\n---\n\n".join(results)

# Gradio UI
demo = gr.Interface(
    fn=process_pdf,
    inputs=gr.File(type="binary", label="Upload PDF Form"),
    outputs="textbox",
    title="Healthelic Form Data Extractor (PDF Scanner) - OpenAI GPT-4o",
    description="Upload a scanned medical form in PDF format to extract key fields using GPT-4o vision model."
)

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