From 483d63e0ef8ff9a9ce1dcfe4d71d86227d35e616 Mon Sep 17 00:00:00 2001 From: rasokiwayami <273806828+rasokiwayami@users.noreply.github.com> Date: Mon, 7 Sep 2026 14:39:34 +0900 Subject: [PATCH] fix(markdown): preserve links in list items --- src/note_mcp/utils/html_to_markdown.py | 3 ++ tests/unit/test_html_to_markdown.py | 43 ++++++++++++++++++++++++++ 2 files changed, 46 insertions(+) diff --git a/src/note_mcp/utils/html_to_markdown.py b/src/note_mcp/utils/html_to_markdown.py index 17dece8..664dd43 100644 --- a/src/note_mcp/utils/html_to_markdown.py +++ b/src/note_mcp/utils/html_to_markdown.py @@ -313,6 +313,9 @@ def _convert_list(html_content: str, ordered: bool = False, indent_level: int = text = _OL_PATTERN.sub("", text) text = text.strip() + # Preserve supported inline elements before removing residual HTML tags. + text = _convert_inline_elements(text) + # Clean up any remaining HTML tags from text text = re.sub(r"<[^>]+>", "", text).strip() diff --git a/tests/unit/test_html_to_markdown.py b/tests/unit/test_html_to_markdown.py index 6c143f7..888b585 100644 --- a/tests/unit/test_html_to_markdown.py +++ b/tests/unit/test_html_to_markdown.py @@ -3,6 +3,8 @@ Tests for html_to_markdown function that converts note.com HTML format back to Markdown. """ +from bs4 import BeautifulSoup + from note_mcp.utils.html_to_markdown import html_to_markdown from note_mcp.utils.markdown_to_html import markdown_to_html @@ -102,6 +104,47 @@ def test_nested_list_conversion(self) -> None: for line in sub_items: assert line.startswith(" ") or line.startswith(" ") + def test_ordered_and_unordered_list_links_are_preserved(self) -> None: + """Test that links survive conversion from paragraph-wrapped list items.""" + for tag, marker in (("ul", "-"), ("ol", "1.")): + html = f"""<{tag}> +

  • Read the official docs.

    • + """ + + result = html_to_markdown(html) + + assert f"{marker} Read the [official docs](https://github.com/drillan/note-mcp)." in result + + def test_nested_list_links_preserve_formatted_labels(self) -> None: + """Test links and inline formatting in nested list items.""" + reference_url = "https://note.com/drillan/n/n7379c02632c9" + html = f"""""" + + result = html_to_markdown(html) + + assert "- Start with [**API docs**](https://github.com/drillan/note-mcp)." in result + assert f" 1. Then read the [*reference guide*]({reference_url})." in result + + def test_list_link_roundtrip_preserves_label_and_href(self) -> None: + """Test that a list link keeps its label and href through HTML and Markdown.""" + source_html = """
      +

    1. Consult the guide.

      2. +
    """ + + markdown = html_to_markdown(source_html) + roundtrip_html = markdown_to_html(markdown) + links = BeautifulSoup(roundtrip_html, "html.parser").select('li a[href="https://github.com/drillan/note-mcp"]') + + assert "1. Consult [**the guide**](https://github.com/drillan/note-mcp)." in markdown + assert len(links) == 1 + assert links[0].get_text() == "the guide" + # === Blockquote Tests === def test_blockquote_conversion(self) -> None: