[Haskell-cafe] point-free ADT pattern matching ?

[Vo Minh Thu]

2010/7/15 Alexey Karakulov <ankarakulov at gmail.com>:
> I wonder if pattern matching could be less verbose. Maybe this sounds weird,
> but here is example of what I mean:
>
>> type A = (Int, String)
>>
>> f :: String -> A -> A
>> f s (i,s') = (i, s ++ s')
>>
>> data B = B Int String deriving Show
>>
>>g :: String -> B -> B
>>g s (B i s') = B i $ s ++ s'
>
> Types A/B and functions f/g are quite similar: (x :: A) or (x :: B) means
> that x contains some integer and string values, and f/g functions take some
> string and prepend it to the string part of x. The code for f and g has the
> same level of verbosity, but -- ta-dah! -- we can use arrows and define f in
> a highly laconic manner:
>
>> import Control.Arrow
>> f' :: String -> A -> A
>> f' = second . (++)
>
> So my queastion is how I could define (g' :: String -> B -> B) in the same
> way.

I guess it is short because you make use of second... so you can
define second' for your B data type, or make B an instance of Arrow.

Cheers,
Thu