Type signature of readIntP seems redundant?
Current type signature of readIntP is: readIntP :: Num a => a -> (Char -> Bool) -> (Char -> Int) -> ReadP a This seems kinda redundant. It inputs two functions, one checking whether a character is valid digit, and the other assigning an integer to the digit. Shouldn't we merge them together to one function? Like this: import Control.Monad import Data.Foldable import Data.Maybe import Text.ParserCombinators.ReadP readIntP :: Num a => a -> (Char -> Maybe Int) -> ReadP a readIntP base valDigit = do maybeNs <- manyTill (fmap (fmap fromIntegral . valDigit) get) $ do str <- fmap (fmap valDigit) look case str of [] -> return () (Nothing:_) -> return () _ -> pfail let n = foldl' (\l r -> case l of Nothing -> r Just l1 -> fmap (base * l1 +) r ) Nothing maybeNs guard (isJust n) return (fromJust n)
readIntP :: Num a => a -> (Char -> Bool) -> (Char -> Int) -> ReadP a
Dunno, but one could guess that people find this function more convenient to apply, in the spirit of readIntP x (('0' <=) .&&. (<= '9')) ((- ord '0') . ord) One could always add your function, under a different name. They easily convert into each other, using a function like boolToMaybe :: Bool -> a -> Maybe a boolToMaybe True a = Just a boolToMaybe False _ = Nothing On 2020-06-06 01:04, Dannyu NDos wrote:
Current type signature of readIntP is:
readIntP :: Num a => a -> (Char -> Bool) -> (Char -> Int) -> ReadP a
This seems kinda redundant. It inputs two functions, one checking whether a character is valid digit, and the other assigning an integer to the digit.
Shouldn't we merge them together to one function? Like this:
importControl.Monad importData.Foldable importData.Maybe importText.ParserCombinators.ReadP
readIntP:: Numa=> a-> (Char-> MaybeInt) -> ReadPa readIntP base valDigit = do maybeNs <- manyTill (fmap (fmap fromIntegral . valDigit) get) $ do str <- fmap (fmap valDigit) look casestr of []-> return () (Nothing:_) -> return () _ -> pfail letn = foldl' (\l r -> casel of Nothing -> r Just l1 -> fmap (base * l1 +) r ) Nothing maybeNs guard (isJust n) return (fromJust n)
_______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries
I would guess history has a lot to do with it. The whole readP mechanism is
pretty old, and many of our modern conventions hadn't really taken hold yet.
On Mon, Jun 8, 2020, 2:54 PM Andreas Abel
readIntP :: Num a => a -> (Char -> Bool) -> (Char -> Int) -> ReadP a
Dunno, but one could guess that people find this function more convenient to apply, in the spirit of
readIntP x (('0' <=) .&&. (<= '9')) ((- ord '0') . ord)
One could always add your function, under a different name.
They easily convert into each other, using a function like
boolToMaybe :: Bool -> a -> Maybe a boolToMaybe True a = Just a boolToMaybe False _ = Nothing
On 2020-06-06 01:04, Dannyu NDos wrote:
Current type signature of readIntP is:
readIntP :: Num a => a -> (Char -> Bool) -> (Char -> Int) -> ReadP a
This seems kinda redundant. It inputs two functions, one checking whether a character is valid digit, and the other assigning an integer to the digit.
Shouldn't we merge them together to one function? Like this:
importControl.Monad importData.Foldable importData.Maybe importText.ParserCombinators.ReadP
readIntP:: Numa=> a-> (Char-> MaybeInt) -> ReadPa readIntP base valDigit = do maybeNs <- manyTill (fmap (fmap fromIntegral . valDigit) get) $ do str <- fmap (fmap valDigit) look casestr of []-> return () (Nothing:_) -> return () _ -> pfail letn = foldl' (\l r -> casel of Nothing -> r Just l1 -> fmap (base * l1 +) r ) Nothing maybeNs guard (isJust n) return (fromJust n)
_______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries
_______________________________________________ Libraries mailing list Libraries@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries
participants (3)
-
Andreas Abel -
Dannyu NDos -
David Feuer