test_experimental_markdown_syntax_text_splitter_split_lines_on_multi_files() — langchain Function Reference

Source Code

libs/text-splitters/tests/unit_tests/test_text_splitters.py lines 2045–2170

def test_experimental_markdown_syntax_text_splitter_split_lines_on_multi_files() -> (
    None
):
    """Test ExperimentalMarkdownSyntaxTextSplitter split lines on multiple files.

    Test experimental markdown syntax splitter split on each line called consecutively
    on two files.
    """
    markdown_splitter = ExperimentalMarkdownSyntaxTextSplitter(return_each_line=True)
    output = []
    for experimental_markdown_document in EXPERIMENTAL_MARKDOWN_DOCUMENTS:
        output += markdown_splitter.split_text(experimental_markdown_document)
    expected_output = [
        Document(
            page_content="Content for header 1 from Document 1",
            metadata={"Header 1": "My Header 1 From Document 1"},
        ),
        Document(
            page_content="Content for header 2 from Document 1",
            metadata={
                "Header 1": "My Header 1 From Document 1",
                "Header 2": "Header 2 From Document 1",
            },
        ),
        Document(
            page_content="```python",
            metadata={
                "Code": "python",
                "Header 1": "My Header 1 From Document 1",
                "Header 2": "Header 2 From Document 1",
            },
        ),
        Document(
            page_content="def func_definition():",
            metadata={
                "Code": "python",
                "Header 1": "My Header 1 From Document 1",
                "Header 2": "Header 2 From Document 1",
            },
        ),
        Document(
            page_content="   print('Keep the whitespace consistent')",
            metadata={
                "Code": "python",
                "Header 1": "My Header 1 From Document 1",
                "Header 2": "Header 2 From Document 1",
            },
        ),
        Document(
            page_content="```",
            metadata={
                "Code": "python",
                "Header 1": "My Header 1 From Document 1",
                "Header 2": "Header 2 From Document 1",
            },
        ),
        Document(
            page_content="We should also split on the horizontal line",
            metadata={"Header 1": "Header 1 again From Document 1"},
        ),
        Document(
            page_content="This will be a new doc but with the same header metadata",
            metadata={"Header 1": "Header 1 again From Document 1"},
        ),
        Document(
            page_content="And it includes a new paragraph",
            metadata={"Header 1": "Header 1 again From Document 1"},
        ),
        Document(
            page_content="Content for header 1 from Document 2",
            metadata={"Header 1": "My Header 1 From Document 2"},
        ),
        Document(
            page_content="Content for header 2 from Document 2",
            metadata={
                "Header 1": "My Header 1 From Document 2",
                "Header 2": "Header 2 From Document 2",
            },
        ),
        Document(
            page_content="```python",