Generating duplicate expansion dividers

Currently, because we rely on the set name for checking for which sets to generate expansion dividers for, in languages where the set name is not always unique (e.g. in dutch both "alchemy" and "alchemy extras" are named "De Alchemisten"), the code generates an extra divider. 

The way we choose which sets to include could really use a rewrite. 