from flask import Flask, request, jsonify
import os
import tempfile
import fitz  # PyMuPDF
from paddleocr import PaddleOCR
from PIL import Image

app = Flask(__name__)

# Initialize PaddleOCR once (for English)
ocr = PaddleOCR(lang='en')

@app.route('/')
def index():
    return jsonify({"message": "Welcome to the PaddleOCR + PyMuPDF PDF Extractor API!"})

@app.route('/extract', methods=['POST'])
def extract_text():
    if 'pdf' not in request.files:
        return jsonify({"error": "No PDF file uploaded"}), 400

    pdf_file = request.files['pdf']

    # Save PDF temporarily
    temp_dir = tempfile.mkdtemp()
    pdf_path = os.path.join(temp_dir, pdf_file.filename)
    pdf_file.save(pdf_path)

    doc = fitz.open(pdf_path)
    extracted_data = []

    for page_number, page in enumerate(doc, start=1):
        # Convert PDF page to image
        pix = page.get_pixmap(dpi=200)
        img_path = os.path.join(temp_dir, f"page_{page_number}.png")
        pix.save(img_path)

        # Perform OCR
        result = ocr.ocr(img_path, cls=True)
        page_text = ""
        for line in result:
            for word_info in line:
                text = word_info[1][0]
                page_text += text + " "

        extracted_data.append({
            "page": page_number,
            "text": page_text.strip()
        })

    return jsonify({
        "status": "success",
        "pages": extracted_data
    })

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=7860)