Note also that typeclasses are open, so ghc is not allowed to say that there is no instance of Num for lists there; it will happily infer a type which requires such an instance, and only when it needs to firm down to concrete types at some point will it notice that there's no such instance in scope. (I think some such instances do exist, in fact, in various programs.) Just to make things more interesting, numeric literals are not sufficient to make it think otherwise because of the implicit fromIntegral / fromRational, which a Num instance for lists would need to supply at the appropriate type. FlexibleContexts is because the Haskell standard is extremely pedantic about the form that typeclass instances and contexts may take; I think if you had written it as (Num ([] a)) it would have passed without an extension. On the one hand, it's something of an irritation; on the other, it *does* help to catch thinkos like the above. -- brandon s allbery kf8nh Sent with Sparrow (http://www.sparrowmailapp.com/?sig) On Friday, February 15, 2013 at 2:33 AM, sheng chen wrote:
Hi,
I was puzzled by the following little program.
sum' [] = [] sum' (x:xs) = x + sum' xs
I thought the GHC type checker will report a type error. However, the type checker accepts this program and gives the type
Num [a] => [[a]] -> [a]
When I add type annotation to the program
sum' :: Num [a] => [[a]] -> [a] sum' [] = [] sum' (x:xs) = x + sum' xs
The GHC asks me to add FlexibleContexts language extension.
I would appreciate explanation on this issue.
Thanks, Sheng _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org (mailto:Haskell-Cafe@haskell.org) http://www.haskell.org/mailman/listinfo/haskell-cafe