Parsing of literate files fails because of special character (#)
A sharp sign (#) in the documentation part of a literate haskell file causes the parser to fail. As I understand it, any line that doesn't start with a greater-than sign should be ignored. Note: this error is only triggered when the sharp is the first character on the line.
#+
> module Empty where
The previous code leads to the following error:
with ghc-7.8.3: Bug.lhs:1:2: lexical error at character '+'[[BR]] with ghc-8.0.1: Bug.lhs:1:1: error: parse error on input ‘#+’