problem: pdfs containing pictures are big and very slow to display.
Better have some program that splits pages and distributing the letters accordingly. I've seen some programs for this splitting, but have not tried them
I would think of something like this:
http://www.unixuser.org/~euske/pytho...ner/index.html
extracting letter location, then using pycairo to make new (half) pages ..
maybe easier with pypdf?
http://pybrary.net/pyPdf/