from transformers import TrOCRProcessor, VisionEncoderDecoderModel
from PIL import Image
import torch
import re

# Load TrOCR model and processor once
processor = TrOCRProcessor.from_pretrained("microsoft/trocr-base-handwritten")
model = VisionEncoderDecoderModel.from_pretrained("microsoft/trocr-base-handwritten")

def extract_weight(image):
    try:
        # OCR Inference
        pixel_values = processor(images=image, return_tensors="pt").pixel_values
        generated_ids = model.generate(pixel_values)
        text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0].strip()
        print("OCR Output:", text)

        # Pattern to detect weight with optional decimal and unit (g or kg)
        match = re.search(r'(\d{1,5}(?:\.\d{1,3})?)\s*(kg|g)', text.lower())

        if match:
            value = match.group(1)
            unit = match.group(2)
            return f"{value} {unit}"
        else:
            return "No valid weight found"
    except Exception as e:
        return f"Error: {str(e)}"