test_html_splitter_keep_separator_start() — langchain Function Reference

Source Code

libs/text-splitters/tests/unit_tests/test_text_splitters.py lines 3889–3915

def test_html_splitter_keep_separator_start() -> None:
    """Test HTML splitting with keep_separator="start"."""
    html_content = """
    <h1>Section 1</h1>
    <p>This is some text. This is some other text.</p>
    """
    with suppress_langchain_beta_warning():
        splitter = HTMLSemanticPreservingSplitter(
            headers_to_split_on=[("h1", "Header 1")],
            max_chunk_size=10,
            separators=[". "],
            keep_separator="start",
        )
    documents = splitter.split_text(html_content)

    expected = [
        Document(
            page_content="This is some text",
            metadata={"Header 1": "Section 1"},
        ),
        Document(
            page_content=". This is some other text.",
            metadata={"Header 1": "Section 1"},
        ),
    ]

    assert documents == expected