Right now, the code that reads in the dictionary can choke up on formatting issues in the dictionary, like extra spaces:

Should potentially be an easy fix with the new/upcoming "ContentCoder" engine that I've written, but will have to test implementation once I've got time to get things added in (could be a little while!).