New script that generates two JSON files with tokenization data about snippets (does not run on Travis). Data can be used for visualizations and possibly creating a complexity index for the snippets.