PDFs are binary files, but they actually keep the http links in text format,
which makes it easy to scan and detect links within them...

It would be a good feature to be able to do that and download the child
documents, even if the links within the download PDF could maybe not be
changed to point to the local downloaded files...