Convert Kindle HTML notes to Markdown

notes_parse.py

from bs4 import BeautifulSoup
import argparse


def clean_text(section) -> str:
    return section.get_text().lstrip().replace('\n', '').replace('\t', '')


def parse_chapter(text):
    # There is NO chapter number, but there is A page number. Example ->
    # Highlight(<span class="highlight_orange">orange</span>) - Page xxvii · Location 381
    if '>' not in text and 'Page' in text:
        page = text.split('Page')[1].split('·')[0].strip()
        return None, page

    # There is A chapter number, but there is NO page number. Example ->
    # Highlight(<span class="highlight_orange">orange</span>) - I > Location 38
    # Note: BS parses the <span>
    if '>' in text and 'Page' not in text:
        chapter = text.split('>')[0].strip()
        chapter = chapter.split('Highlight(orange) -')[1].strip()
        return chapter, None
    # There is A chapter number, and there is A page number. Example ->
    # Highlight(<span class="highlight_orange">orange</span>) - I.1 Enter the two Bishops, [the Archbishop] of Canterbury and [the Bishop of] Ely. > Page 7 · Location 778
    if '>' in text and 'Page' in text:
        chapter = text.split('Highlight(orange) -')[1]\
                      .split('>')[0].strip()
        page = text.split('Page')[1].split('·')[0].strip()
        return chapter, page

    # There is NO chapter number, and there is NO page number
    return None, None
def remove_brackets(text) -> str:
    """Brackets are used for Wikilinks and conflict with Obsidian"""
    return text.replace('[', '').replace(']', '')


def convert_notes(inputf) -> str:
    mdf = ""
    with open(inputf, 'r') as htmlf:
        soup = BeautifulSoup(htmlf.read(), 'html.parser')

        mdf += f"## Reflection\n"
        mdf += f"## Notes\n"
        # Find first section heading
        start_node = soup.find('div', {'class': 'sectionHeading'})
        mdf += f"### {clean_text(start_node)}\n"
        # Notes must be parsed sequentially like so, bc Amazon does not use children elements.
        is_note = False
        chapter = ''
        page = None
        current = start_node
        while current.next_sibling:
            current = current.next_sibling
            if not current.name or not current.attrs or 'class' not in current.attrs:
                continue
            class_name = current['class'][0]

            # Indent notes to make them a child element
            if class_name == 'noteText' and is_note:
                mdf += f"\t+ **Note**: {clean_text(current)}\n"
                continue
            if class_name == 'noteText':
                if page and page != '':
                    mdf += f"+ *Page {page}*: {clean_text(current)}\n"
                # Not all books have page numbers. Some only use "Location".
                else:
                    mdf += f"+ {clean_text(current)}\n"
                continue
            # Notes added to highlights begin with "Note".
            # Setting `is_note` allows us to make the note text a sublist to the highlight.
            elif class_name == 'noteHeading':
                text = clean_text(current)
                is_note = text.startswith('Note - ')

                if not is_note:
                    new_chapter, page = parse_chapter(text)
                    if new_chapter is not None and new_chapter != chapter:
                        mdf += f"#### {new_chapter}\n"
                    # # Has not happened yet, but I want to know if it does.
                    # if new_chapter is None and chapter is not None:
                    #     mdf += f"#### NO CHAPTER\n"
                    chapter = new_chapter
                continue
            elif class_name == 'sectionHeading':
                mdf += f"### {clean_text(current)}\n"
                continue
    return remove_brackets(mdf)


def parse_args():
    parser = argparse.ArgumentParser(description="Process filename")
    parser.add_argument("f", type=str, help="The input filename")
    return parser.parse_args().f


if __name__ == '__main__':
    print(convert_notes(parse_args()))