Add the "tokenchars=" and "separators=" options, for customizing the set of characters considered to be token separators, to the unicode61 tokenizer.

FossilOrigin-Name: e56fb462aa1f11bb23303ae0dc62815c21e26a52