Extract the table of contents from a PDF file and save it as an OPML, e.g. for import into a mind map

pdf_toc_to_opml.py

# Requirements
# yattag==1.12.2
# PyPDF2==1.26.0

# Tested with Python 3.7.4 on macOS


import sys
from pathlib import Path

from PyPDF2 import PdfFileReader

from yattag import Doc
from yattag import indent

def dump_outline(doc, outline):

    # This is a rather convoluted function, due to the way
    # PyPDF2 returns the PDF outline.
    # If the parent node is located at position i, the child
    # nodes are stored as a list in position i+1

    curr_idx = 0

    while curr_idx < len(outline):

        try:
            if isinstance(outline[curr_idx +1], list):
                with doc.tag('outline', text=outline[curr_idx].title):
                    dump_outline(doc, outline[curr_idx+1])

                # Jump over child nodes
                curr_idx += 2
                continue
        except IndexError:
            pass

        # Use a self-closing tag, there are no child nodes
        doc.stag('outline', text=outline[curr_idx].title)
        curr_idx += 1

        
def main(pdf_filename):

    try:
        input_file = Path(pdf_filename)
        # The OPML file is saved in the same location as the PDF file
        output_file = input_file.with_suffix('.opml')

        pdf_doc = PdfFileReader(open(input_file, 'rb'))

        doc, tag, text, line = Doc().ttl()

        doc.asis('<?xml version="1.0" encoding="UTF-8"?>')
        with tag('opml', version='1.0'):
            with tag('head'):

                # If the document has no title, we'll use the 
                # PDF filename (without extension)
                
                title = pdf_doc.getDocumentInfo().title
                if title is None:
                    title = input_file.stem
                line('title', title)

            with tag('body'):
                dump_outline(doc, pdf_doc.outlines)

        with open(output_file, 'w') as f:
            f.write(indent(doc.getvalue()))

    except Exception as e:
        print(f'Error: {e!r}')


if __name__ == '__main__':
    # The PDF filename (incl path) is passed as an argument on the command line
    main(sys.argv[1])