test_html_splitter_preserve_nested_in_paragraph() — langchain Function Reference

Source Code

libs/text-splitters/tests/unit_tests/test_text_splitters.py lines 3550–3569

def test_html_splitter_preserve_nested_in_paragraph() -> None:
    """Test preserving deeply nested elements (code inside paragraph).

    tests the case where a preserved element (<code>) is nested
    inside a non-container element (<p>)
    """
    html_content = "<p>before <code>KEEP_THIS</code> after</p>"
    with suppress_langchain_beta_warning():
        splitter = HTMLSemanticPreservingSplitter(
            headers_to_split_on=[],
            elements_to_preserve=["code"],
        )
    documents = splitter.split_text(html_content)

    assert len(documents) == 1
    content = documents[0].page_content
    # All text should be preserved
    assert "before" in content
    assert "KEEP_THIS" in content
    assert "after" in content