[Haskell-cafe] lexing nested comments in GHC/Parser/Lexer.x

[Benjamin Redelings]

Hi,

I notice that in Lexer.x for GHC nested comments are parsed using a 
special function, with the comment "nested comments require traversing 
by hand, they can't be parsed using regular expressions."

While I see that nested comments can't be parsed using regular 
expressions, the lexer has a state stack, and so it seems like nested 
comments could be parsed by pushing a new "nested comment" state onto 
the lexer state every time a "{-" is encountered.  But perhaps this is 
not done because its not clear how to make the entire comment into a 
single token that way?  This might be a naive question -- I don't have 
that much experience with lexers.

Thanks!

-BenRI