You cant do that, you need to include the file name as well. However, doing so is not trivial, since articles are downloaded asynchrously so you dont know what filename an article will have in advance.
What you should do is use a uuid as the anchor and then implement postprocess_book to insert the links.
|