Add support to link multiple PDF files to a single (BibTex) record / post-process PDFs

[GerHobbelt]

As we import encrypted, mangled or otherwise tough-to-use-directly PDFs into Qiqqa, we should be able to add transformation processes a la https://github.com/GerHobbelt/qiqqa-revengin qpdf scripts and other customizable fixes) which take a PDF as input, apply a transformation and output another PDF which technically contains the same 'human information' but hashes to a different checksum.

In an ideal world, both these PDFs (original and transformed derivative) should be filed in the same slot in Qiqqa.

This is very much related to #7 and #12.

Examples of PDF nasties:

• Fairchild/ON-Semi datasheets trigger PDF content not visible (white/empty view)  #7 + add ability to analyze the Librarie(s) and list all broken/b0rked/damaged PDFs in there, then select them and apply user-chosen activity #12 as they apparently are print-protected and causing Qiqqa to render entirely blank pages rather than actual content.
• XeteX-produced papers are a horror - not sure about the XeTeX bit, but I surely have run into many PDF papers which were clearly produced with some TeX variant which caused the qiqqa OCR action to fail miserably no matter what I tried. I had to apply another OCR package with some forced image-to-text settings to turn those buggers into Qiqqa-readable texts. 😠
• anything with 'banner pages' which are spam/clutter, particularly ones which have odd page size/width (hello, Farnell) and throwing Qiqqa off guard (thumbs up -- or is it down? -- for some Korean Uni's) and unable to pop up a decent title thanks to crufty lead obscuring the real meat in there.
• ... others i cannot recall off the top of my skull right now ...