
On Sun, Sep 1, 2013 at 3:58 AM, Artyom Kazak
Would this be an appropriate place to propose adding mapM_ (and then possibly mapM) to bytestring library?
Was it suggested before? If yes, why was it rejected?
For what it's worth, this is exported by lens in Data.ByteString.Lens.bytes (or Data.ByteString.{Lazy,Strict}.Lens.bytes). The version exported by lens by default traverses each chunk in a balanced tree shape rather than left-to-right (which is probably not relevant to most use cases, actually -- the result is the same). It also works around http://ghc.haskell.org/trac/ghc/ticket/7556, which you'll want to watch for if you're writing this yourself. Note that you can construct mapM_ from mapM as discussed in another libraries@ thread recently (or by using e.g. mapMOf_). Shachaf