PDFs are notoriously hard to convert to anything else. Imagine a web page where every single character is positioned absolutely relative to the `body` element:

PDFs are notoriously hard to convert to anything else. Imagine a web page where every single character is positioned absolutely relative to the `body` element: This is how PDF layout works. It's easier when the PDF also contains HTML tags. So your approach may actually work though it's kinda silly. Pandoc[1] is a better solution. [1] https://pandoc.org/ https://pandoc.org/