This is a super crappy script I wrote that parses through Greek PDFs and, poorly, calculates the most common 750 words. It somewhat ignores articles, θα, να, and some other word fragments, but sucks at it. It was enough for my purposes so maybe it will help someone else.