Sequencing bind with pure functions

Question

I often find myself wanting to insert regular functions into a "binded" sequence. Like in this contrived example:

getLine >>= lift (map toUpper) >>= putStrLn

I need to define the lift function lift :: (a -> b) -> a -> m b to make this work. Problem is I don't know of such a function, and Hoogle doesn't seem to either. I find this odd since this makes totally sense to me.

Now, there are probably other ways to make this work, but I like the way point-free style code allows me to scan the line in one pass to figure out what is happening.

let lift f x = return (f x) in
getLine >>= lift (map toUpper) >>= putStrLn

My question boils down to this: am I missing something or how come there isn't a function like lift. My experience in Haskell is still very limited, so I am assuming that most people solve this in a different way. Can someone explain to me the idiomatic way of solving this.

Answer 1

There are three idiomatic ways.

1. Don't use bind; use the first hit on your Hoogle search instead:

   liftM (map toUpper) getLine >>= putStrLn
   

   There are a variety of alternative spellings of liftM, such as fmap or (<$>).

2. Inline the lift function you defined:

   getLine >>= return . map toUpper >>= putStrLn
   

3. Use the monad laws to fuse the last two binds in option 2:

   getLine >>= putStrLn . map toUpper
   

Answer 2

Use the Functor instance in such cases:

> import Data.Char
> import Data.Functor
> map toUpper <$> getLine >>= putStrLn
foo
FOO
>