Function pdf_to_bytes

src/datasets/features/pdf.py:22–27

Convert a pdfplumber.pdf.PDF object to bytes.

(pdf: "pdfplumber.pdf.PDF")

def pdf_to_bytes(pdf: "pdfplumber.pdf.PDF") -> bytes:
    """Convert a pdfplumber.pdf.PDF object to bytes."""
    with BytesIO() as buffer:
        for page in pdf.pages:
            buffer.write(page.pdf.stream)
        return buffer.getvalue()


@dataclass