selenium/thesis-file

1import pytesseract2from PIL import Image3from pdf2image import convert_from_path4 5MAX_WIDTH = 1200  # reduce image width to save memory6 7 8def preprocess_image(img: Image.Image) -> Image.Image:9    """Resize image while maintaining aspect ratio to reduce memory usage."""10    if img.width > MAX_WIDTH:11        ratio = MAX_WIDTH / img.width12        new_height = int(img.height * ratio)13        img = img.resize((MAX_WIDTH, new_height))14    return img15 16 17def extract_text_from_image(path: str) -> str:18    """Extract text from a single image safely with timeout."""19    try:20        img = Image.open(path)21        img = preprocess_image(img)22        text = pytesseract.image_to_string(23            img, timeout=60)  # 60s max per image24        return text25    except pytesseract.TesseractError as e:26        return f'OCR error: {str(e)}'27    except Exception as e:28        return f'OCR error: {str(e)}'29 30 31def extract_text_from_pdf(path: str) -> str:32    """Extract text from a PDF safely, processing pages one by one."""33    try:34        pages = convert_from_path(path)35        text = ''36        for page in pages:37            page = preprocess_image(page)38            text += pytesseract.image_to_string(page, timeout=60)39        return text40    except pytesseract.TesseractError as e:41        return f'OCR error: {str(e)}'42    except Exception as e:43        return f'OCR error: {str(e)}'44