More flexible support for comments in Text.Parsec.Token about parsec

[mrkkrp]

@paf31, can you tell me more about this issue? My current understanding is that you don't want Parsec to ignore comments that start with -- | instead of --, so you can parse doc-strings specially.

We are working of Parsec's fork called Megaparsec and we have rewritten 70 % of code already. Currently we are working on complete QuickCheck test suite. If you're interested, see our repo. You can see what's already done in CHANGELOG.md file and in README.md (although the latter describes our ultimate goals, more than currently implemented). You can also check issue tracker if you have time.

So, we will soon get to token parsing where we will implement support for languages where indentation matters. While we're at it, we would like to know what else users of Parsec "fight" and what they would like to see improved. Since PureScript is a language, its implementation certainly needs to deal with parsing. Could you describe what Megaparsec could implement so Purescript would (hypotetically) benefit from it?