Function test_load_from_bytesio_lazy

tests/test_parse.py:697–722

Test loading PDF from BytesIO with lazy=True.

()

def test_load_from_bytesio_lazy():
    """Test loading PDF from BytesIO with lazy=True."""
    filename = "tests/data/regression/table_of_contents_01.pdf"

    # Read file into BytesIO
    with open(filename, "rb") as file:
        file_content = file.read()
    bytes_io = BytesIO(file_content)

    parser = DoclingPdfParser(loglevel="fatal")

    # Load from BytesIO
    pdf_doc_bytesio = parser.load(path_or_stream=bytes_io, lazy=True)

    # Load from path for comparison
    pdf_doc_path = parser.load(path_or_stream=filename, lazy=True)

    # Both should have same number of pages
    assert pdf_doc_bytesio.number_of_pages() == pdf_doc_path.number_of_pages()

    # Load all pages and compare
    pdf_doc_bytesio.load_all_pages()
    pdf_doc_path.load_all_pages()

    # Pages should be identical
    assert pdf_doc_bytesio._pages == pdf_doc_path._pages


def test_load_from_bytesio_eager():