-- | -- Module : Data.Edison.Seq.MyersStack -- Copyright : Copyright (c) 1998-1999, 2008 Chris Okasaki -- License : MIT; see COPYRIGHT file for terms and conditions -- -- Maintainer : robdockins AT fastmail DOT fm -- Stability : stable -- Portability : GHC, Hugs (MPTC and FD) -- -- Meyers Stacks. All operations are as listed in "Data.Edison.Seq" except -- the following: -- -- * lookup, inBounds, drop @O( min(i, log n) )@ -- -- * rhead*, size @O( log n )@ -- -- * subseq @O( min (i, log n) + len )@ -- -- /References:/ -- -- * Eugene Myers. \"An applicative random-access stack\". /Information -- Processing Letters/, 17(5):241-248, December 1983. module Data.Edison.Seq.MyersStack ( -- * Sequence Type Seq, -- instance of Sequence, Functor, Monad, MonadPlus -- * Sequence Operations empty,singleton,lcons,rcons,append,lview,lhead,ltail,rview,rhead,rtail, lheadM,ltailM,rheadM,rtailM, null,size,concat,reverse,reverseOnto,fromList,toList,map,concatMap, fold,fold',fold1,fold1',foldr,foldr',foldl,foldl',foldr1,foldr1',foldl1,foldl1', reducer,reducer',reducel,reducel',reduce1,reduce1', copy,inBounds,lookup,lookupM,lookupWithDefault,update,adjust, mapWithIndex,foldrWithIndex,foldrWithIndex',foldlWithIndex,foldlWithIndex', take,drop,splitAt,subseq,filter,partition,takeWhile,dropWhile,splitWhile, zip,zip3,zipWith,zipWith3,unzip,unzip3,unzipWith,unzipWith3, strict, strictWith, -- * Unit testing structuralInvariant, -- * Documentation moduleName ) where import Prelude hiding (concat,reverse,map,concatMap,foldr,foldl,foldr1,foldl1, filter,takeWhile,dropWhile,lookup,take,drop,splitAt, zip,zip3,zipWith,zipWith3,unzip,unzip3,null) import qualified Control.Applicative as App import qualified Data.Edison.Seq as S ( Sequence(..) ) import Data.Edison.Seq.Defaults import Control.Monad.Identity import Data.Monoid import Test.QuickCheck -- signatures for exported functions moduleName :: String empty :: Seq a singleton :: a -> Seq a lcons :: a -> Seq a -> Seq a rcons :: a -> Seq a -> Seq a append :: Seq a -> Seq a -> Seq a lview :: (Monad m) => Seq a -> m (a, Seq a) lhead :: Seq a -> a lheadM :: (Monad m) => Seq a -> m a ltail :: Seq a -> Seq a ltailM :: (Monad m) => Seq a -> m (Seq a) rview :: (Monad m) => Seq a -> m (a, Seq a) rhead :: Seq a -> a rheadM :: (Monad m) => Seq a -> m a rtail :: Seq a -> Seq a rtailM :: (Monad m) => Seq a -> m (Seq a) null :: Seq a -> Bool size :: Seq a -> Int concat :: Seq (Seq a) -> Seq a reverse :: Seq a -> Seq a reverseOnto :: Seq a -> Seq a -> Seq a fromList :: [a] -> Seq a toList :: Seq a -> [a] map :: (a -> b) -> Seq a -> Seq b concatMap :: (a -> Seq b) -> Seq a -> Seq b fold :: (a -> b -> b) -> b -> Seq a -> b fold' :: (a -> b -> b) -> b -> Seq a -> b fold1 :: (a -> a -> a) -> Seq a -> a fold1' :: (a -> a -> a) -> Seq a -> a foldr :: (a -> b -> b) -> b -> Seq a -> b foldl :: (b -> a -> b) -> b -> Seq a -> b foldr1 :: (a -> a -> a) -> Seq a -> a foldl1 :: (a -> a -> a) -> Seq a -> a reducer :: (a -> a -> a) -> a -> Seq a -> a reducel :: (a -> a -> a) -> a -> Seq a -> a reduce1 :: (a -> a -> a) -> Seq a -> a foldr' :: (a -> b -> b) -> b -> Seq a -> b foldl' :: (b -> a -> b) -> b -> Seq a -> b foldr1' :: (a -> a -> a) -> Seq a -> a foldl1' :: (a -> a -> a) -> Seq a -> a reducer' :: (a -> a -> a) -> a -> Seq a -> a reducel' :: (a -> a -> a) -> a -> Seq a -> a reduce1' :: (a -> a -> a) -> Seq a -> a copy :: Int -> a -> Seq a inBounds :: Int -> Seq a -> Bool lookup :: Int -> Seq a -> a lookupM :: (Monad m) => Int -> Seq a -> m a lookupWithDefault :: a -> Int -> Seq a -> a update :: Int -> a -> Seq a -> Seq a adjust :: (a -> a) -> Int -> Seq a -> Seq a mapWithIndex :: (Int -> a -> b) -> Seq a -> Seq b foldrWithIndex :: (Int -> a -> b -> b) -> b -> Seq a -> b foldlWithIndex :: (b -> Int -> a -> b) -> b -> Seq a -> b foldrWithIndex' :: (Int -> a -> b -> b) -> b -> Seq a -> b foldlWithIndex' :: (b -> Int -> a -> b) -> b -> Seq a -> b take :: Int -> Seq a -> Seq a drop :: Int -> Seq a -> Seq a splitAt :: Int -> Seq a -> (Seq a, Seq a) subseq :: Int -> Int -> Seq a -> Seq a filter :: (a -> Bool) -> Seq a -> Seq a partition :: (a -> Bool) -> Seq a -> (Seq a, Seq a) takeWhile :: (a -> Bool) -> Seq a -> Seq a dropWhile :: (a -> Bool) -> Seq a -> Seq a splitWhile :: (a -> Bool) -> Seq a -> (Seq a, Seq a) zip :: Seq a -> Seq b -> Seq (a,b) zip3 :: Seq a -> Seq b -> Seq c -> Seq (a,b,c) zipWith :: (a -> b -> c) -> Seq a -> Seq b -> Seq c zipWith3 :: (a -> b -> c -> d) -> Seq a -> Seq b -> Seq c -> Seq d unzip :: Seq (a,b) -> (Seq a, Seq b) unzip3 :: Seq (a,b,c) -> (Seq a, Seq b, Seq c) unzipWith :: (a -> b) -> (a -> c) -> Seq a -> (Seq b, Seq c) unzipWith3 :: (a -> b) -> (a -> c) -> (a -> d) -> Seq a -> (Seq b, Seq c, Seq d) strict :: Seq a -> Seq a strictWith :: (a -> b) -> Seq a -> Seq a structuralInvariant :: Seq a -> Bool moduleName = "Data.Edison.Seq.MyersStack" data Seq a = E | C !Int a (Seq a) (Seq a) -- what about strictness flags on tail and jump-tail? -- auxiliary function jump :: Seq t -> Seq t jump (C _ _ _ (C _ _ _ xs')) = xs' jump _ = error "MyersStack.jump: bug!" empty = E singleton x = C 1 x E E lcons x xs@(C i _ _ (C j _ _ xs')) | i == j = C (1 + i + j) x xs xs' lcons x xs = C 1 x xs xs lview E = fail "MyersStack.lview: empty sequence" lview (C _ x xs _) = return (x, xs) lhead E = error "MyersStack.lhead: empty sequence" lhead (C _ x _ _) = x lheadM E = fail "MyersStack.lheadM: empty sequence" lheadM (C _ x _ _) = return x ltail E = error "MyersStack.ltail: empty sequence" ltail (C _ _ xs _) = xs ltailM E = fail "MyersStack.ltailM: empty sequence" ltailM (C _ _ xs _) = return xs rview E = fail "MyersStack.rview: empty sequence" rview xs = return (rhead xs, rtail xs) rhead E = error "MyersStack.rhead: empty sequence" rhead (C _ x xs xs') = rh x xs xs' where rh _ _ (C _ y ys ys') = rh y ys ys' rh _ (C _ y ys ys') E = rh y ys ys' rh x E E = x rheadM E = fail "MyersStack.rheadM: empty sequence" rheadM (C _ x xs xs') = return (rh x xs xs') where rh _ _ (C _ y ys ys') = rh y ys ys' rh _ (C _ y ys ys') E = rh y ys ys' rh x E E = x rtail E = error "MyersStack.rtail: empty sequence" rtail (C _ x xs _) = rt x xs where rt _ E = E rt y (C _ x xs _) = lcons y (rt x xs) rtailM E = fail "MyersStack.rtailM: empty sequence" rtailM (C _ x xs _) = return (rt x xs) where rt _ E = E rt y (C _ x xs _) = lcons y (rt x xs) null E = True null _ = False size xs = go xs where go E = (0::Int) go (C j _ _ xs') = j + size xs' reverseOnto E ys = ys reverseOnto (C _ x xs _) ys = reverseOnto xs (lcons x ys) map _ E = E map f (C j x xs _') | j == 1 = C j (f x) ys ys | otherwise = C j (f x) ys (jump ys) where ys = map f xs fold = foldr fold' f = foldl' (flip f) fold1 = fold1UsingFold fold1' = fold1'UsingFold' foldr _ e E = e foldr f e (C _ x xs _) = f x (foldr f e xs) foldr' _ e E = e foldr' f e (C _ x xs _) = f x $! (foldr' f e xs) foldl _ e E = e foldl f e (C _ x xs _) = foldl f (f e x) xs foldl' _ e E = e foldl' f e (C _ x xs _) = e `seq` foldl' f (f e x) xs foldr1 _ E = error "MyersStack.foldr1: empty sequence" foldr1 f (C _ x xs _) = fr x xs where fr y E = y fr y (C _ x xs _) = f y (fr x xs) foldr1' _ E = error "MyersStack.foldr1': empty sequence" foldr1' f (C _ x xs _) = fr x xs where fr y E = y fr y (C _ x xs _) = f y $! (fr x xs) foldl1 _ E = error "MyersStack.foldl1: empty sequence" foldl1 f (C _ x xs _) = foldl f x xs foldl1' _ E = error "MyersStack.foldl1': empty sequence" foldl1' f (C _ x xs _ ) = foldl' f x xs inBounds i xs = inb xs i where inb E _ = False inb (C j _ _ xs') i | i < j = (i >= 0) | otherwise = inb xs' (i - j) lookup i xs = runIdentity (lookupM i xs) lookupM i xs = look xs i where look E _ = fail "MyersStack.lookup: bad subscript" look (C j x xs xs') i | i >= j = look xs' (i - j) | i > 0 = look xs (i - 1) | i == 0 = return x | otherwise = nothing nothing = fail "MyersStack.lookup: not found" lookupWithDefault d i xs = look xs i where look E _ = d look (C j x xs xs') i | i >= j = look xs' (i - j) | i > 0 = look xs (i - 1) | i == 0 = x | otherwise = d update i y xs = upd i xs where upd _ E = E upd 0 (C j _ xs xs') = C j y xs xs' upd i (C j x xs _) | j == 1 = C j x ys ys | otherwise = C j x ys (jump ys) where ys = upd (i - 1) xs adjust f i xs = adj i xs where adj _ E = E adj 0 (C j x xs xs') = C j (f x) xs xs' adj i (C j x xs _) | j == 1 = C j x ys ys | otherwise = C j x ys (jump ys) where ys = adj (i - (1::Int)) xs drop n xs = drp n xs where drp n xs | n <= 0 = xs drp _ E = E drp n (C j _ xs xs') | n < j = drp (n - 1) xs | otherwise = drp (n - j) xs' unzip E = (E, E) unzip (C j (x,y) ps _') | j == 1 = (C j x xs xs, C j y ys ys) | otherwise = (C j x xs (jump xs), C j y ys (jump ys)) where (xs,ys) = unzip ps unzip3 E = (E, E, E) unzip3 (C j (x,y,z) ts _') | j == 1 = (C j x xs xs, C j y ys ys, C j z zs zs) | otherwise = (C j x xs (jump xs), C j y ys (jump ys), C j z zs (jump zs)) where (xs,ys,zs) = unzip3 ts unzipWith _ _ E = (E, E) unzipWith f g (C j x xs _) | j == 1 = (C j (f x) as as, C j (g x) bs bs) | otherwise = (C j (f x) as (jump as), C j (g x) bs (jump bs)) where (as,bs) = unzipWith f g xs unzipWith3 _ _ _ E = (E, E, E) unzipWith3 f g h (C j x xs _) | j == 1 = (C j (f x) as as, C j (g x) bs bs, C j (h x) cs cs) | otherwise = (C j (f x) as (jump as), C j (g x) bs (jump bs), C j (h x) cs (jump cs)) where (as,bs,cs) = unzipWith3 f g h xs strict s@E = s strict s@(C _ _ xs _) = strict xs `seq` s strictWith _ s@E = s strictWith f s@(C _ x xs _) = f x `seq` strictWith f xs `seq` s -- the remaining functions all use defaults rcons = rconsUsingFoldr append = appendUsingFoldr concat = concatUsingFoldr reverse = reverseUsingReverseOnto fromList = fromListUsingCons toList = toListUsingFoldr concatMap = concatMapUsingFoldr reducer = reducerUsingReduce1 reducer' = reducer'UsingReduce1' reducel = reducelUsingReduce1 reducel' = reducel'UsingReduce1' reduce1 = reduce1UsingLists reduce1' = reduce1'UsingLists copy = copyUsingLists mapWithIndex = mapWithIndexUsingLists foldrWithIndex = foldrWithIndexUsingLists foldrWithIndex' = foldrWithIndex'UsingLists foldlWithIndex = foldlWithIndexUsingLists foldlWithIndex' = foldlWithIndex'UsingLists take = takeUsingLists splitAt = splitAtDefault filter = filterUsingFoldr partition = partitionUsingFoldr subseq = subseqDefault takeWhile = takeWhileUsingLview dropWhile = dropWhileUsingLview splitWhile = splitWhileUsingLview -- for zips, could optimize by calculating which one is shorter and -- retaining its shape zip = zipUsingLists zip3 = zip3UsingLists zipWith = zipWithUsingLists zipWith3 = zipWith3UsingLists -- FIXME what are the structural invariants? structuralInvariant = const True -- instances instance S.Sequence Seq where {lcons = lcons; rcons = rcons; lview = lview; lhead = lhead; ltail = ltail; lheadM = lheadM; ltailM = ltailM; rheadM = rheadM; rtailM = rtailM; rview = rview; rhead = rhead; rtail = rtail; null = null; size = size; concat = concat; reverse = reverse; reverseOnto = reverseOnto; fromList = fromList; toList = toList; fold = fold; fold' = fold'; fold1 = fold1; fold1' = fold1'; foldr = foldr; foldr' = foldr'; foldl = foldl; foldl' = foldl'; foldr1 = foldr1; foldr1' = foldr1'; foldl1 = foldl1; foldl1' = foldl1'; reducer = reducer; reducer' = reducer'; reducel = reducel; reducel' = reducel'; reduce1 = reduce1; reduce1' = reduce1'; copy = copy; inBounds = inBounds; lookup = lookup; lookupM = lookupM; lookupWithDefault = lookupWithDefault; update = update; adjust = adjust; mapWithIndex = mapWithIndex; foldrWithIndex = foldrWithIndex; foldrWithIndex' = foldrWithIndex'; foldlWithIndex = foldlWithIndex; foldlWithIndex' = foldlWithIndex'; take = take; drop = drop; splitAt = splitAt; subseq = subseq; filter = filter; partition = partition; takeWhile = takeWhile; dropWhile = dropWhile; splitWhile = splitWhile; zip = zip; zip3 = zip3; zipWith = zipWith; zipWith3 = zipWith3; unzip = unzip; unzip3 = unzip3; unzipWith = unzipWith; unzipWith3 = unzipWith3; strict = strict; strictWith = strictWith; structuralInvariant = structuralInvariant; instanceName _ = moduleName} instance Functor Seq where fmap = map instance App.Alternative Seq where empty = empty (<|>) = append instance App.Applicative Seq where pure = return x <*> y = do x' <- x y' <- y return (x' y') instance Monad Seq where return = singleton xs >>= k = concatMap k xs instance MonadPlus Seq where mplus = append mzero = empty instance Eq a => Eq (Seq a) where xs == ys = (size xs == size ys) && (toList xs == toList ys) instance Ord a => Ord (Seq a) where compare = defaultCompare instance Show a => Show (Seq a) where showsPrec = showsPrecUsingToList instance Read a => Read (Seq a) where readsPrec = readsPrecUsingFromList instance Arbitrary a => Arbitrary (Seq a) where arbitrary = do xs <- arbitrary return (fromList xs) instance CoArbitrary a => CoArbitrary (Seq a) where coarbitrary xs = coarbitrary (toList xs) instance Monoid (Seq a) where mempty = empty mappend = append ------------- {- questions: - any benefit to E | C1 x xs | CJ Int# x xs xs' - any benefit to length instead of delta? - any benefit to delta not counting x (i.e., base 0 instead of base 1)? I don't believe any will do any better, except possibly the first -}