It's possible to provide a document that looks like and URL but isn't #1
Description
The checks in #add_page don't correctly check that an URL is the only thing in the document before assuming that the file should be downloaded. It's possible that this may lead to accidental execution of code on the command line.