On Thu, Dec 15, 2011 at 9:13 AM, Gregory Crosswhite
<[email protected]>wrote:
> To quote Ross Paterson's proposals:
>
> instance Alternative [] where
> ...
> some [] = []
> some (x:xs) = repeat (repeat x)
>
> many [] = [[]]
> many (x:xs) = repeat (repeat x)
>
Isn't this instance conflating the ZipList notion and the nondeterministic
list notion?
> • some v = (:) <$> v <*> many v
> • many v = some v <|> pure []
>
Is there a motivation for writing the second law like this and not like
many v = pure [] <|> some v
other than "parsers should try longest match first"? Because apart from
that, I find the version with flipped arguments to <|> more natural (base
case first). Incidentally, it would lead to terminating definitions of
'some' and 'many' for lists:
ghci> take 5 . map (take 5) $ some [1,2]
[[1],[1,1],[1,1,1],[1,1,1,1],[1,1,1,1,1]]
ghci> take 5 . map (take 5) $ many [1,2]
[[],[1],[1,1],[1,1,1],[1,1,1,1]]
Sebastian
_______________________________________________
Haskell-Cafe mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/haskell-cafe
