Skip to content

HORIZON HASKELL Docs lts/ghc-10.0.x8fef28d2026-11-02Search packagesCtrl K

Theme: |

GHC 10.0.0.20260917 · lts/ghc-10.0.x · 8fef28d · 2026-11-02

Modulepqueue-1.7.0.0Haskell2010

Data.PQueue.Prio.Max

General purpose priority queue. Each element is associated with a key, and the priority queue supports viewing and extracting the element with the maximum key.

A worst-case bound is given for each operation. In some cases, an amortized bound is also specified; these bounds hold even in a persistent context.

This implementation is based on a binomial heap augmented with a global root.

We do not guarantee stable behavior. Ties are broken arbitrarily -- that is, if k1 <= k2 and k2 <= k1, then there are no guarantees about the relative order in which k1, k2, and their associated elements are returned. (Unlike Data.Map, we allow multiple elements with the same key.)

This implementation offers a number of methods of the form xxxU, where U stands for unordered. No guarantees whatsoever are made on the execution or traversal order of these functions.

Contents

Construction
Query
Maximum view
Traversal
Map
Fold
Traverse
Subsets
Indexed
Predicates
Filter
List operations
Conversion from lists
Conversion to lists
Unordered operations
Helper methods

Synopsis

Packagepqueue-1.7.0.0
Exports72
LanguageHaskell2010
LicenceBSD-3-Clause
SourceMax.hs

newtypenewtype MaxPQueue k a

A priority queue where values of type a are annotated with keys of type k. The queue supports extracting the element with maximum key.

Instances14FoldableWithIndex, FunctorWithIndex, TraversableWithIndex, Functor, Foldable, Traversable, …

Ord k => FoldableWithIndex k (MaxPQueue k)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
FunctorWithIndex k (MaxPQueue k)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Ord k => TraversableWithIndex k (MaxPQueue k)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Functor (MaxPQueue k)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Ord k => Foldable (MaxPQueue k)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Ord k => Traversable (MaxPQueue k)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Traverses in descending order. mapM is strictly accumulating like mapMWithKey.
(Ord k, Eq a) => Eq (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
(Data k, Data a, Ord k) => Data (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
(Ord k, Ord a) => Ord (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
(Ord k, Read k, Read a) => Read (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
(Ord k, Show k, Show a) => Show (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Ord k => Semigroup (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
Ord k => Monoid (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals
(NFData k, NFData a) => NFData (MaxPQueue k a)Defined in pqueue-1.7.0.0 · Data.PQueue.Prio.Max.Internals

Construction

valueempty :: MaxPQueue k a

O(1). Returns the empty priority queue.

valuesingleton :: k -> a -> MaxPQueue k a

O(1). Constructs a singleton priority queue.

valueinsert :: Ord k => k -> a -> MaxPQueue k a -> MaxPQueue k a

Amortized O(1), worst-case O(\log n). Inserts an element with the specified key into the queue.

valueunion :: Ord k => MaxPQueue k a -> MaxPQueue k a -> MaxPQueue k a

Amortized O(\log \min(n_1,n_2)), worst-case O(\log \max(n_1,n_2)). Returns the union of the two specified queues.

valueunions :: Ord k => [MaxPQueue k a] -> MaxPQueue k a

The union of a list of queues: (unions == List.foldl union empty).

Query

valuenull :: MaxPQueue k a -> Bool

O(1). Checks if this priority queue is empty.

valuesize :: MaxPQueue k a -> Int

O(1). Returns the size of this priority queue.

Maximum view

valuefindMax :: MaxPQueue k a -> (k, a)

O(1). The maximal (key, element) in the queue. Calls error if empty.

valuegetMax :: MaxPQueue k a -> Maybe (k, a)

O(1). The maximal (key, element) in the queue, if the queue is nonempty.

valuedeleteMax :: Ord k => MaxPQueue k a -> MaxPQueue k a

O(\log n). Delete and find the element with the maximum key. Calls error if empty.

valuedeleteFindMax :: Ord k => MaxPQueue k a -> ((k, a), MaxPQueue k a)

O(\log n). Delete and find the element with the maximum key. Calls error if empty.

valueadjustMax :: (a -> a) -> MaxPQueue k a -> MaxPQueue k a

O(1). Alter the value at the maximum key. If the queue is empty, does nothing.

valueadjustMaxA

:: Applicative f
=> a -> f a
-> MaxPQueue k a
-> f (MaxPQueue k a)

O(1) per operation. Alter the value at the maximum key in an Applicative context. If the queue is empty, does nothing.

valueadjustMaxWithKey :: (k -> a -> a) -> MaxPQueue k a -> MaxPQueue k a

O(1). Alter the value at the maximum key. If the queue is empty, does nothing.

valueadjustMaxWithKeyA

:: Applicative f
=> k -> a -> f a
-> MaxPQueue k a
-> f (MaxPQueue k a)

O(1) per operation. Alter the value at the maximum key in an Applicative context. If the queue is empty, does nothing.

valueupdateMax :: Ord k => (a -> Maybe a) -> MaxPQueue k a -> MaxPQueue k a

O(\log n). (Actually O(1) if there's no deletion.) Update the value at the maximum key. If the queue is empty, does nothing.

valueupdateMaxA

:: (Applicative f, Ord k)
=> a -> f (Maybe a)
-> MaxPQueue k a
-> f (MaxPQueue k a)

O(\log n) per operation. (Actually O(1) if there's no deletion.) Update the value at the maximum key in an Applicative context. If the queue is empty, does nothing.

valueupdateMaxWithKey

:: Ord k
=> k -> a -> Maybe a
-> MaxPQueue k a
-> MaxPQueue k a

O(\log n). (Actually O(1) if there's no deletion.) Update the value at the maximum key. If the queue is empty, does nothing.

valueupdateMaxWithKeyA

:: (Applicative f, Ord k)
=> k -> a -> f (Maybe a)
-> MaxPQueue k a
-> f (MaxPQueue k a)

O(\log n) per operation. (Actually O(1) if there's no deletion.) Update the value at the maximum key in an Applicative context. If the queue is empty, does nothing.

valuemaxView :: Ord k => MaxPQueue k a -> Maybe (a, MaxPQueue k a)

O(\log n). Retrieves the value associated with the maximum key of the queue, and the queue stripped of that element, or Nothing if passed an empty queue.

valuemaxViewWithKey :: Ord k => MaxPQueue k a -> Maybe ((k, a), MaxPQueue k a)

O(\log n). Retrieves the maximal (key, value) pair of the map, and the map stripped of that element, or Nothing if passed an empty map.

Traversal

Map

valuemap :: (a -> b) -> MaxPQueue k a -> MaxPQueue k b

O(n). Map a function over all values in the queue.

valuemapWithKey :: (k -> a -> b) -> MaxPQueue k a -> MaxPQueue k b

O(n). Map a function over all values in the queue.

valuemapKeys :: Ord k' => (k -> k') -> MaxPQueue k a -> MaxPQueue k' a

O(n). Map a function over all values in the queue.

valuemapKeysMonotonic :: (k -> k') -> MaxPQueue k a -> MaxPQueue k' a

O(n). mapKeysMonotonic f q == mapKeys f q, but only works when f is (weakly) monotonic (meaning that x <= y implies f x <= f y). The precondition is not checked. This function has better performance than mapKeys.

Note: if the given function returns bottom for any of the keys in the queue, then the portion of the queue which is bottom is unspecified.

Fold

valuefoldrWithKey :: Ord k => (k -> a -> b -> b) -> b -> MaxPQueue k a -> b

O(n \log n). Fold the keys and values in the map, such that foldrWithKey f z q == List.foldr (uncurry f) z (toDescList q).

If you do not care about the traversal order, consider using foldrWithKeyU.

valuefoldlWithKey :: Ord k => (b -> k -> a -> b) -> b -> MaxPQueue k a -> b

O(n \log n). Fold the keys and values in the map, such that foldlWithKey f z q == List.foldl (uncurry . f) z (toDescList q).

If you do not care about the traversal order, consider using foldlWithKeyU.

Traverse

valuetraverseWithKey

:: (Ord k, Applicative f)
=> k -> a -> f b
-> MaxPQueue k a
-> f (MaxPQueue k b)

O(n \log n). Traverses the elements of the queue in descending order by key. (traverseWithKey f q == fromDescList $ traverse (uncurry f) (toDescList q))

If you do not care about the order of the traversal, consider using traverseWithKeyU.

If you are working in a strict monad, consider using mapMWithKey.

valuemapMWithKey

:: (Ord k, Monad m)
=> k -> a -> m b
-> MaxPQueue k a
-> m (MaxPQueue k b)

A strictly accumulating version of traverseWithKey. This works well in IO and strict State, and is likely what you want for other "strict" monads, where ⊥ >>= pure () = ⊥.

Subsets

Indexed

valuetake :: Ord k => Int -> MaxPQueue k a -> [(k, a)]

O(k \log n). Takes the first k (key, value) pairs in the queue, or the first n if k >= n. (take k q == take k (toDescList q))

valuedrop :: Ord k => Int -> MaxPQueue k a -> MaxPQueue k a

O(k \log n). Deletes the first k (key, value) pairs in the queue, or returns an empty queue if k >= n.

valuesplitAt :: Ord k => Int -> MaxPQueue k a -> ([(k, a)], MaxPQueue k a)

O(k \log n). Equivalent to (take k q, drop k q).

Predicates

valuetakeWhile :: Ord k => (a -> Bool) -> MaxPQueue k a -> [(k, a)]

Takes the longest possible prefix of elements satisfying the predicate. (takeWhile p q == takeWhile (p . snd) (toDescList q))

valuetakeWhileWithKey :: Ord k => (k -> a -> Bool) -> MaxPQueue k a -> [(k, a)]

Takes the longest possible prefix of elements satisfying the predicate. (takeWhile p q == takeWhile (uncurry p) (toDescList q))

valuedropWhile :: Ord k => (a -> Bool) -> MaxPQueue k a -> MaxPQueue k a

Removes the longest possible prefix of elements satisfying the predicate.

valuedropWhileWithKey

:: Ord k
=> k -> a -> Bool
-> MaxPQueue k a
-> MaxPQueue k a

Removes the longest possible prefix of elements satisfying the predicate.

valuespan :: Ord k => (a -> Bool) -> MaxPQueue k a -> ([(k, a)], MaxPQueue k a)

Equivalent to (takeWhile p q, dropWhile p q).

valuespanWithKey

:: Ord k
=> k -> a -> Bool
-> MaxPQueue k a
-> ([(k, a)], MaxPQueue k a)

Equivalent to spanWithKey (k a -> not (p k a)) q.

valuebreak :: Ord k => (a -> Bool) -> MaxPQueue k a -> ([(k, a)], MaxPQueue k a)

Equivalent to span (not . p).

valuebreakWithKey

:: Ord k
=> k -> a -> Bool
-> MaxPQueue k a
-> ([(k, a)], MaxPQueue k a)

Equivalent to spanWithKey (k a -> not (p k a)) q.

Filter

valuefilter :: Ord k => (a -> Bool) -> MaxPQueue k a -> MaxPQueue k a

O(n). Filter all values that satisfy the predicate.

valuefilterWithKey :: Ord k => (k -> a -> Bool) -> MaxPQueue k a -> MaxPQueue k a

O(n). Filter all values that satisfy the predicate.

valuepartition

:: Ord k
=> a -> Bool
-> MaxPQueue k a
-> (MaxPQueue k a, MaxPQueue k a)

O(n). Partition the queue according to a predicate. The first queue contains all elements which satisfy the predicate, the second all elements that fail the predicate.

valuepartitionWithKey

:: Ord k
=> k -> a -> Bool
-> MaxPQueue k a
-> (MaxPQueue k a, MaxPQueue k a)

O(n). Partition the queue according to a predicate. The first queue contains all elements which satisfy the predicate, the second all elements that fail the predicate.

valuemapMaybe :: Ord k => (a -> Maybe b) -> MaxPQueue k a -> MaxPQueue k b

O(n). Map values and collect the Just results.

valuemapMaybeWithKey

:: Ord k
=> k -> a -> Maybe b
-> MaxPQueue k a
-> MaxPQueue k b

O(n). Map values and collect the Just results.

valuemapEither

:: Ord k
=> a -> Either b c
-> MaxPQueue k a
-> (MaxPQueue k b, MaxPQueue k c)

O(n). Map values and separate the Left and Right results.

valuemapEitherWithKey

:: Ord k
=> k -> a -> Either b c
-> MaxPQueue k a
-> (MaxPQueue k b, MaxPQueue k c)

O(n). Map values and separate the Left and Right results.

List operations

Conversion from lists

valuefromList :: Ord k => [(k, a)] -> MaxPQueue k a

O(n). Build a priority queue from the list of (key, value) pairs.

valuefromAscList :: [(k, a)] -> MaxPQueue k a

O(n). Build a priority queue from an ascending list of (key, value) pairs. The precondition is not checked.

valuefromDescList :: [(k, a)] -> MaxPQueue k a

O(n). Build a priority queue from a descending list of (key, value) pairs. The precondition is not checked.

Conversion to lists

valuekeys :: Ord k => MaxPQueue k a -> [k]

O(n \log n). Return all keys of the queue in descending order.

valueelems :: Ord k => MaxPQueue k a -> [a]

O(n \log n). Return all elements of the queue in descending order by key.

valueassocs :: Ord k => MaxPQueue k a -> [(k, a)]

O(n \log n). Equivalent to toDescList.

valuetoAscList :: Ord k => MaxPQueue k a -> [(k, a)]

O(n \log n). Return all (key, value) pairs in ascending order by key.

valuetoDescList :: Ord k => MaxPQueue k a -> [(k, a)]

O(n \log n). Return all (key, value) pairs in descending order by key.

valuetoList :: Ord k => MaxPQueue k a -> [(k, a)]

O(n \log n). Equivalent to toDescList.

If the traversal order is irrelevant, consider using toListU.

Unordered operations

valuefoldrU :: (a -> b -> b) -> b -> MaxPQueue k a -> b

O(n). An unordered right fold over the elements of the queue, in no particular order.

valuefoldrWithKeyU :: (k -> a -> b -> b) -> b -> MaxPQueue k a -> b

O(n). An unordered right fold over the elements of the queue, in no particular order.

valuefoldMapWithKeyU :: Monoid m => (k -> a -> m) -> MaxPQueue k a -> m

O(n). An unordered monoidal fold over the elements of the queue, in no particular order.

valuefoldlU :: (b -> a -> b) -> b -> MaxPQueue k a -> b

O(n). An unordered left fold over the elements of the queue, in no particular order. This is rarely what you want; foldrU and foldlU' are more likely to perform well.

valuefoldlU' :: (b -> a -> b) -> b -> MaxPQueue k a -> b

O(n). An unordered strict left fold over the elements of the queue, in no particular order.

valuefoldlWithKeyU :: (b -> k -> a -> b) -> b -> MaxPQueue k a -> b

O(n). An unordered left fold over the elements of the queue, in no particular order. This is rarely what you want; foldrWithKeyU and foldlWithKeyU' are more likely to perform well.

valuefoldlWithKeyU' :: (b -> k -> a -> b) -> b -> MaxPQueue k a -> b

O(n). An unordered left fold over the elements of the queue, in no particular order.

valuetraverseU

:: Applicative f
=> a -> f b
-> MaxPQueue k a
-> f (MaxPQueue k b)

O(n). An unordered traversal over a priority queue, in no particular order. While there is no guarantee in which order the elements are traversed, the resulting priority queue will be perfectly valid.

valuetraverseWithKeyU

:: Applicative f
=> k -> a -> f b
-> MaxPQueue k a
-> f (MaxPQueue k b)

O(n). An unordered traversal over a priority queue, in no particular order. While there is no guarantee in which order the elements are traversed, the resulting priority queue will be perfectly valid.

valuekeysU :: MaxPQueue k a -> [k]

O(n). Return all keys of the queue in no particular order.

valueelemsU :: MaxPQueue k a -> [a]

O(n). Return all elements of the queue in no particular order.

valueassocsU :: MaxPQueue k a -> [(k, a)]

O(n). Equivalent to toListU.

valuetoListU :: MaxPQueue k a -> [(k, a)]

O(n). Returns all (key, value) pairs in the queue in no particular order.

Helper methods

valueseqSpine :: MaxPQueue k a -> b -> b

Deprecated. This function is no longer necessary or useful.

O(\log n). seqSpine q r forces the spine of q and returns r.

Note: The spine of a MaxPQueue is stored somewhat lazily. In earlier versions of this package, some operations could produce chains of thunks along the spine, occasionally necessitating manual forcing. Now, all operations are careful to force enough to avoid this problem.