On Wed, 12 Jul 2006, Evan Laforge wrote:

Also, my version of the "split but don't drop the delimiter" is breakAll, since it's like a recursive break.

words/unwords and lines/unlines have one argument, so a 2 argument fielts/unfields would break that "convention". Maybe:

fields = csv `separateWith` "," csv = fields `joinWith` "," -- equivalent to concatIntersperse

I may be confused about your point, but I think the way we get split functions down to one parameter is by partial application - commaSplit = splitBy (== ',') fields = commaSplit csv Hence I would much rather have the split condition be the first parameter -- the infix notation looks good, but it will need a "flip" to get the parameters in the right order. Donn Cave, donn@drizzle.com

Quoth "Jared Updike" : (... quoting me) |> Hence I would much rather have the split condition be the first |> parameter -- the infix notation looks good, but it will need a |> "flip" to get the parameters in the right order. | | This also goes along with join in python, i.e. | | "\n".join(xs) | ==> | "\n" `join` xs | in Haskell (where join sep = concat . intersperse sep ) Suit yourself! Since we seem to be in agreement on the basic point, I won't go into what I think about "\n".join(xs). Donn Cave, donn@drizzle.com