All they are is passed through the tokeniser to ensure they lex as valid D (not parse). http://d-programming-language.org/lex.html A naked slash is not a valid token on its own, AFAIK, so it shouldn't work. But the other two are fine.