Re: #4189: Add (<.>) operator (generalizing (.) to Functor)

On Mon, 2 Aug 2010, David Menendez wrote:

On Sun, Aug 1, 2010 at 9:52 AM, Maciej Marcin Piechotka wrote:

...

The proposal is to add (<.>) function to Data.Functor/Control.Applicative: (<.>) :: (b -> c) -> (a -> f b) -> a -> f c f <.> g = fmap f . g -- (<.>) = (.) . fmap

In intend it is related to <$> in the same way as (.) is related to $: (a . b . c) d = a $ b $ c $ d (a <.> b <.> c) d = a <$> b <$> c <$> d

I'm not convinced. "fmap f . g" isn't that much longer than "f <.> g" and requires no new combinators.

'f' and 'g' might be infix expressions. Depending on the precedence we had to compare "fmap (f) . g" with "f <.> g" or "fmap (f) . (g)" with "f <.> g" .

I'd argue that "fmap f . fmap g . h" is better style, since it's obvious that this should be rewritten as "fmap (f . g) . h". In the example above, "a <$> b <$> c <$> d" is best transformed to "a . b . c <$> d".

I am also happy with fmap f . fmap g . h and a . b . c <$> d .