Really, that wouldn't be too bad to impliment. The hardest part would be getting the text into the database, which could be fairly well automated on existing ebooks, just being a bitch on books that are not available as ebooks. Even still with their tech, they can scan and OCR the text on those without too much hassle.
|