Consider adding support for PDF inspection using peepdf

[Lewis Rosenthal]

It would be useful to have the ability to scan PDFs for malicious content, with an option to scan before opening. This will become more important when we ultimately have JS support in Poppler, of course, but for now, like any other type of virus or malware scanning, it is mainly to ensure that we don't pass along bad stuff to others.

​peepdf is a set of Python modules (with their own dependencies) to provide in-depth analysis of PDF documents. It would probably be advisable if peepdf and its dependencies (​pyv8 and ​pylibemu) were added to the repo, here, and we utilize those modules. I have not thought through the implementation of this, however, so that will require some more careful planning.