Interface FlowOps<Out,Mat>
-
- All Known Subinterfaces:
FlowOpsMat<Out,Mat>
,SubFlow<Out,Mat,F,C>
public interface FlowOps<Out,Mat>
Scala API: Operations offered by Sources and Flows with a free output side: the DSL flows left-to-right only.INTERNAL API: this trait will be changed in binary-incompatible ways for classes that are derived from it! Do not implement this interface outside the Akka code base!
Binary compatibility is only maintained for callers of this trait’s interface.
-
-
Method Summary
All Methods Instance Methods Abstract Methods Deprecated Methods Modifier and Type Method Description <U,M>
FlowOps$plus$plus(Graph<SourceShape<U>,M> that)
FlowOps
addAttributes(Attributes attr)
<Agg,Emit>
FlowOpsaggregateWithBoundary(scala.Function0<Agg> allocate, scala.Function2<Agg,Out,scala.Tuple2<Agg,java.lang.Object>> aggregate, scala.Function1<Agg,Emit> harvest, scala.Option<scala.Tuple2<scala.Function1<Agg,java.lang.Object>,scala.concurrent.duration.FiniteDuration>> emitOnTimer)
Aggregate input elements into an arbitrary data structure that can be completed and emitted downstream when custom condition is met which can be triggered by aggregate or timer.FlowOps
alsoTo(Graph<SinkShape<Out>,?> that)
FlowOps
alsoToAll(scala.collection.immutable.Seq<Graph<SinkShape<Out>,?>> those)
<M> Graph<FlowShape<Out,Out>,M>
alsoToGraph(Graph<SinkShape<Out>,M> that)
<S> FlowOps
ask(int parallelism, ActorRef ref, Timeout timeout, scala.reflect.ClassTag<S> tag)
Use theask
pattern to send a request-reply message to the targetref
actor.<S> FlowOps
ask(ActorRef ref, Timeout timeout, scala.reflect.ClassTag<S> tag)
Use theask
pattern to send a request-reply message to the targetref
actor.FlowOps
async()
Put an asynchronous boundary around thisFlow
.FlowOps
backpressureTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the time between the emission of an element and the following downstream demand exceeds the provided timeout, the stream is failed with aBackpressureTimeoutException
.<S> FlowOps
batch(long max, scala.Function1<Out,S> seed, scala.Function2<S,Out,S> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by aggregating elements into batches until the subscriber is ready to accept them.<S> FlowOps
batchWeighted(long max, scala.Function1<Out,java.lang.Object> costFn, scala.Function1<Out,S> seed, scala.Function2<S,Out,S> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by aggregating elements into batches until the subscriber is ready to accept them.FlowOps
buffer(int size, OverflowStrategy overflowStrategy)
Adds a fixed size buffer in the flow that allows to store elements from a faster upstream until it becomes full.<T> FlowOps
collect(scala.PartialFunction<Out,T> pf)
Transform this stream by applying the given partial function to each of the elements on which the function is defined as they pass through this processing step.<T> FlowOps
collectType(scala.reflect.ClassTag<T> tag)
Transform this stream by testing the type of each of the elements on which the element is an instance of the provided type as they pass through this processing step.FlowOps
completionTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the completion of the stream does not happen until the provided timeout, the stream is failed with aCompletionTimeoutException
.<U,Mat2>
FlowOpsconcat(Graph<SourceShape<U>,Mat2> that)
<U> FlowOps
concatAllLazy(scala.collection.immutable.Seq<Graph<SourceShape<U>,?>> those)
<U,Mat2>
Graph<FlowShape<Out,U>,Mat2>concatGraph(Graph<SourceShape<U>,Mat2> that, boolean detached)
<U,Mat2>
FlowOpsconcatLazy(Graph<SourceShape<U>,Mat2> that)
<O2> FlowOps
conflate(scala.Function2<O2,O2,O2> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by conflating elements into a summary until the subscriber is ready to accept them.<S> FlowOps
conflateWithSeed(scala.Function1<Out,S> seed, scala.Function2<S,Out,S> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by conflating elements into a summary until the subscriber is ready to accept them.FlowOps
delay(scala.concurrent.duration.FiniteDuration of, DelayOverflowStrategy strategy)
Shifts elements emission in time by a specified amount.DelayOverflowStrategy
delay$default$2()
FlowOps
delayWith(scala.Function0<DelayStrategy<Out>> delayStrategySupplier, DelayOverflowStrategy overFlowStrategy)
Shifts elements emission in time by an amount individually determined through delay strategy a specified amount.FlowOps
detach()
Detaches upstream demand from downstream demand without detaching the stream rates; in other words acts like a buffer of size 1.FlowOps
divertTo(Graph<SinkShape<Out>,?> that, scala.Function1<Out,java.lang.Object> when)
<M> Graph<FlowShape<Out,Out>,M>
divertToGraph(Graph<SinkShape<Out>,M> that, scala.Function1<Out,java.lang.Object> when)
FlowOps
drop(long n)
Discard the given number of elements at the beginning of the stream.FlowOps
dropWhile(scala.Function1<Out,java.lang.Object> p)
Discard elements at the beginning of the stream while predicate is true.FlowOps
dropWithin(scala.concurrent.duration.FiniteDuration d)
Discard the elements received within the given duration at beginning of the stream.<U> FlowOps
expand(scala.Function1<Out,scala.collection.Iterator<U>> expander)
Allows a faster downstream to progress independently of a slower upstream by extrapolating elements from an older element until new element comes from the upstream.<U> FlowOps
extrapolate(scala.Function1<U,scala.collection.Iterator<U>> extrapolator, scala.Option<U> initial)
Allows a faster downstream to progress independent of a slower upstream.<U> scala.None$
extrapolate$default$2()
FlowOps
filter(scala.Function1<Out,java.lang.Object> p)
Only pass on those elements that satisfy the given predicate.FlowOps
filterNot(scala.Function1<Out,java.lang.Object> p)
Only pass on those elements that NOT satisfy the given predicate.<T,M>
FlowOpsflatMapConcat(scala.Function1<Out,Graph<SourceShape<T>,M>> f)
Transform each input element into aSource
of output elements that is then flattened into the output stream by concatenation, fully consuming one Source after the other.<T,M>
FlowOpsflatMapMerge(int breadth, scala.Function1<Out,Graph<SourceShape<T>,M>> f)
Transform each input element into aSource
of output elements that is then flattened into the output stream by merging, where at mostbreadth
substreams are being consumed at any given time.<Out2,Mat2>
FlowOpsflatMapPrefix(int n, scala.Function1<scala.collection.immutable.Seq<Out>,Flow<Out,Out2,Mat2>> f)
Takes up ton
elements from the stream (less thann
only if the upstream completes before emittingn
elements), then applyf
on these elements in order to obtain a flow, this flow is then materialized and the rest of the input is processed by this flow (similar to via).<T> FlowOps
fold(T zero, scala.Function2<T,Out,T> f)
Similar toscan
but only emits its result when the upstream completes, after which it also completes.<T> FlowOps
foldAsync(T zero, scala.Function2<T,Out,scala.concurrent.Future<T>> f)
Similar tofold
but with an asynchronous function.<K> SubFlow<Out,Mat,FlowOps,java.lang.Object>
groupBy(int maxSubstreams, scala.Function1<Out,K> f)
This operation demultiplexes the incoming stream into separate output streams, one for each element key.<K> SubFlow<Out,Mat,FlowOps,java.lang.Object>
groupBy(int maxSubstreams, scala.Function1<Out,K> f, boolean allowClosedSubstreamRecreation)
This operation demultiplexes the incoming stream into separate output streams, one for each element key.FlowOps
grouped(int n)
Chunk up this stream into groups of the given size, with the last group possibly smaller than requested due to end-of-stream.FlowOps
groupedWeighted(long minWeight, scala.Function1<Out,java.lang.Object> costFn)
Chunk up this stream into groups of elements that have a cumulative weight greater than or equal to theminWeight
, with the last group possibly smaller than requestedminWeight
due to end-of-stream.FlowOps
groupedWeightedWithin(long maxWeight, int maxNumber, scala.concurrent.duration.FiniteDuration d, scala.Function1<Out,java.lang.Object> costFn)
Chunk up this stream into groups of elements received within a time window, or limited by the weight and number of the elements, whatever happens first.FlowOps
groupedWeightedWithin(long maxWeight, scala.concurrent.duration.FiniteDuration d, scala.Function1<Out,java.lang.Object> costFn)
Chunk up this stream into groups of elements received within a time window, or limited by the weight of the elements, whatever happens first.FlowOps
groupedWithin(int n, scala.concurrent.duration.FiniteDuration d)
Chunk up this stream into groups of elements received within a time window, or limited by the given number of elements, whatever happens first.FlowOps
idleTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the time between two processed elements exceeds the provided timeout, the stream is failed with aStreamIdleTimeoutException
.FlowOps
initialDelay(scala.concurrent.duration.FiniteDuration delay)
Delays the initial element by the specified duration.FlowOps
initialTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the first element has not passed through this operator before the provided timeout, the stream is failed with aInitialTimeoutException
.<U> FlowOps
interleave(Graph<SourceShape<U>,?> that, int segmentSize)
<U> FlowOps
interleave(Graph<SourceShape<U>,?> that, int segmentSize, boolean eagerClose)
<U> FlowOps
interleaveAll(scala.collection.immutable.Seq<Graph<SourceShape<U>,?>> those, int segmentSize, boolean eagerClose)
<U,M>
Graph<FlowShape<Out,U>,M>interleaveGraph(Graph<SourceShape<U>,M> that, int segmentSize, boolean eagerClose)
<U,M>
booleaninterleaveGraph$default$3()
<U,Mat2>
FlowOpsinternalConcat(Graph<SourceShape<U>,Mat2> that, boolean detached)
<U> FlowOps
internalConcatAll(Graph<SourceShape<U>,?>[] those, boolean detached)
<T> FlowOps
intersperse(T inject)
Intersperses stream with provided element, similar to howscala.collection.immutable.List.mkString
injects a separator between a List's elements.<T> FlowOps
intersperse(T start, T inject, T end)
Intersperses stream with provided element, similar to howscala.collection.immutable.List.mkString
injects a separator between a List's elements.<U> FlowOps
keepAlive(scala.concurrent.duration.FiniteDuration maxIdle, scala.Function0<U> injectedElem)
Injects additional elements if upstream does not emit for a configured amount of time.FlowOps
limit(long max)
Ensure stream boundedness by limiting the number of elements from upstream.<T> FlowOps
limitWeighted(long max, scala.Function1<Out,java.lang.Object> costFn)
Ensure stream boundedness by evaluating the cost of incoming elements using a cost function.FlowOps
log(java.lang.String name, scala.Function1<Out,java.lang.Object> extract, LoggingAdapter log)
Logs elements flowing through the stream as well as completion and erroring.scala.Function1<Out,java.lang.Object>
log$default$2()
LoggingAdapter
log$default$3(java.lang.String name, scala.Function1<Out,java.lang.Object> extract)
FlowOps
logWithMarker(java.lang.String name, scala.Function1<Out,LogMarker> marker, scala.Function1<Out,java.lang.Object> extract, MarkerLoggingAdapter log)
Logs elements flowing through the stream as well as completion and erroring.scala.Function1<Out,java.lang.Object>
logWithMarker$default$3()
MarkerLoggingAdapter
logWithMarker$default$4(java.lang.String name, scala.Function1<Out,LogMarker> marker, scala.Function1<Out,java.lang.Object> extract)
<T> FlowOps
map(scala.Function1<Out,T> f)
Transform this stream by applying the given function to each of the elements as they pass through this processing step.<T> FlowOps
mapAsync(int parallelism, scala.Function1<Out,scala.concurrent.Future<T>> f)
Transform this stream by applying the given function to each of the elements as they pass through this processing step.<T> FlowOps
mapAsyncUnordered(int parallelism, scala.Function1<Out,scala.concurrent.Future<T>> f)
Transform this stream by applying the given function to each of the elements as they pass through this processing step.<T> FlowOps
mapConcat(scala.Function1<Out,scala.collection.IterableOnce<T>> f)
Transform each input element into anIterable
of output elements that is then flattened into the output stream.FlowOps
mapError(scala.PartialFunction<java.lang.Throwable,java.lang.Throwable> pf)
While similar to<T>recover(scala.PartialFunction<java.lang.Throwable,T>)
this operator can be used to transform an error signal to a different one *without* logging it as an error in the process.<R,T>
FlowOpsmapWithResource(scala.Function0<R> create, scala.Function2<R,Out,T> f, scala.Function1<R,scala.Option<T>> close)
Transform each stream element with the help of a resource.<U,M>
FlowOpsmerge(Graph<SourceShape<U>,M> that, boolean eagerComplete)
<U,M>
booleanmerge$default$2()
<U> FlowOps
mergeAll(scala.collection.immutable.Seq<Graph<SourceShape<U>,?>> those, boolean eagerComplete)
<U,M>
Graph<FlowShape<Out,U>,M>mergeGraph(Graph<SourceShape<U>,M> that, boolean eagerComplete)
<U,M>
FlowOpsmergeLatest(Graph<SourceShape<U>,M> that, boolean eagerComplete)
MergeLatest joins elements from N input streams into stream of lists of size N.<U,M>
booleanmergeLatest$default$2()
<U,M>
Graph<FlowShape<Out,scala.collection.immutable.Seq<U>>,M>mergeLatestGraph(Graph<SourceShape<U>,M> that, boolean eagerComplete)
<U,M>
FlowOpsmergePreferred(Graph<SourceShape<U>,M> that, boolean preferred, boolean eagerComplete)
Merge two sources.<U,M>
booleanmergePreferred$default$3()
<U,M>
Graph<FlowShape<Out,U>,M>mergePreferredGraph(Graph<SourceShape<U>,M> that, boolean preferred, boolean eagerComplete)
<U,M>
FlowOpsmergePrioritized(Graph<SourceShape<U>,M> that, int leftPriority, int rightPriority, boolean eagerComplete)
Merge two sources.<U,M>
booleanmergePrioritized$default$4()
<U,M>
Graph<FlowShape<Out,U>,M>mergePrioritizedGraph(Graph<SourceShape<U>,M> that, int leftPriority, int rightPriority, boolean eagerComplete)
<U,M>
FlowOpsmergeSorted(Graph<SourceShape<U>,M> that, scala.math.Ordering<U> ord)
<U,M>
Graph<FlowShape<Out,U>,M>mergeSortedGraph(Graph<SourceShape<U>,M> that, scala.math.Ordering<U> ord)
FlowOps
named(java.lang.String name)
<U,Mat2>
FlowOpsorElse(Graph<SourceShape<U>,Mat2> secondary)
Provides a secondary source that will be consumed if this stream completes without any elements passing by.<U,Mat2>
Graph<FlowShape<Out,U>,Mat2>orElseGraph(Graph<SourceShape<U>,Mat2> secondary)
<U> FlowOps
prefixAndTail(int n)
Takes up ton
elements from the stream (less thann
only if the upstream completes before emittingn
elements) and returns a pair containing a strict sequence of the taken element and a stream representing the remaining elements.<U,Mat2>
FlowOpsprepend(Graph<SourceShape<U>,Mat2> that)
<U,Mat2>
Graph<FlowShape<Out,U>,Mat2>prependGraph(Graph<SourceShape<U>,Mat2> that, boolean detached)
<U,Mat2>
FlowOpsprependLazy(Graph<SourceShape<U>,Mat2> that)
<T> FlowOps
recover(scala.PartialFunction<java.lang.Throwable,T> pf)
Recover allows to send last element on failure and gracefully complete the stream Since the underlying failure signal onError arrives out-of-band, it might jump over existing elements.<T> FlowOps
recoverWith(scala.PartialFunction<java.lang.Throwable,Graph<SourceShape<T>,NotUsed>> pf)
Deprecated.Use recoverWithRetries instead.<T> FlowOps
recoverWithRetries(int attempts, scala.PartialFunction<java.lang.Throwable,Graph<SourceShape<T>,NotUsed>> pf)
RecoverWithRetries allows to switch to alternative Source on flow failure.<T> FlowOps
reduce(scala.Function2<T,T,T> f)
Similar tofold
but uses first element as zero element.<T> FlowOps
scan(T zero, scala.Function2<T,Out,T> f)
Similar tofold
but is not a terminal operation, emits its current value which starts atzero
and then applies the current and next value to the given functionf
, emitting the next current value.<T> FlowOps
scanAsync(T zero, scala.Function2<T,Out,scala.concurrent.Future<T>> f)
Similar toscan
but with an asynchronous function, emits its current value which starts atzero
and then applies the current and next value to the given functionf
, emitting aFuture
that resolves to the next current value.FlowOps
sliding(int n, int step)
Apply a sliding window over the stream and return the windows as groups of elements, with the last group possibly smaller than requested due to end-of-stream.int
sliding$default$2()
SubFlow<Out,Mat,FlowOps,java.lang.Object>
splitAfter(SubstreamCancelStrategy substreamCancelStrategy, scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams.SubFlow<Out,Mat,FlowOps,java.lang.Object>
splitAfter(scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams.SubFlow<Out,Mat,FlowOps,java.lang.Object>
splitWhen(SubstreamCancelStrategy substreamCancelStrategy, scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams, always beginning a new one with the current element if the given predicate returns true for it.SubFlow<Out,Mat,FlowOps,java.lang.Object>
splitWhen(scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams, always beginning a new one with the current element if the given predicate returns true for it.<S,T>
FlowOpsstatefulMap(scala.Function0<S> create, scala.Function2<S,Out,scala.Tuple2<S,T>> f, scala.Function1<S,scala.Option<T>> onComplete)
Transform each stream element with the help of a state.<T> FlowOps
statefulMapConcat(scala.Function0<scala.Function1<Out,scala.collection.IterableOnce<T>>> f)
Transform each input element into anIterable
of output elements that is then flattened into the output stream.FlowOps
take(long n)
Terminate processing (and cancel the upstream publisher) after the given number of elements.FlowOps
takeWhile(scala.Function1<Out,java.lang.Object> p)
Terminate processing (and cancel the upstream publisher) after predicate returns false for the first time, Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.FlowOps
takeWhile(scala.Function1<Out,java.lang.Object> p, boolean inclusive)
Terminate processing (and cancel the upstream publisher) after predicate returns false for the first time, including the first failed element iff inclusive is true Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.FlowOps
takeWithin(scala.concurrent.duration.FiniteDuration d)
Terminate processing (and cancel the upstream publisher) after the given duration.FlowOps
throttle(int elements, scala.concurrent.duration.FiniteDuration per)
Sends elements downstream with speed limited toelements/per
.FlowOps
throttle(int elements, scala.concurrent.duration.FiniteDuration per, int maximumBurst, ThrottleMode mode)
Sends elements downstream with speed limited toelements/per
.FlowOps
throttle(int cost, scala.concurrent.duration.FiniteDuration per, int maximumBurst, scala.Function1<Out,java.lang.Object> costCalculation, ThrottleMode mode)
Sends elements downstream with speed limited tocost/per
.FlowOps
throttle(int cost, scala.concurrent.duration.FiniteDuration per, scala.Function1<Out,java.lang.Object> costCalculation)
Sends elements downstream with speed limited tocost/per
.FlowOps
throttleEven(int elements, scala.concurrent.duration.FiniteDuration per, ThrottleMode mode)
Deprecated.Use throttle without `maximumBurst` parameter instead.FlowOps
throttleEven(int cost, scala.concurrent.duration.FiniteDuration per, scala.Function1<Out,java.lang.Object> costCalculation, ThrottleMode mode)
Deprecated.Use throttle without `maximumBurst` parameter instead.<Mat2> java.lang.Object
to(Graph<SinkShape<Out>,Mat2> sink)
<T,Mat2>
FlowOpsvia(Graph<FlowShape<Out,T>,Mat2> flow)
FlowOps
watch(ActorRef ref)
The operator fails with anWatchedActorTerminatedException
if the target actor is terminated.FlowOps
wireTap(Graph<SinkShape<Out>,?> that)
FlowOps
wireTap(scala.Function1<Out,scala.runtime.BoxedUnit> f)
This is a simplified version ofwireTap(Sink)
that takes only a simple function.<M> Graph<FlowShape<Out,Out>,M>
wireTapGraph(Graph<SinkShape<Out>,M> that)
FlowOps
withAttributes(Attributes attr)
<U> FlowOps
zip(Graph<SourceShape<U>,?> that)
Combine the elements of current flow and the givenSource
into a stream of tuples.<U,A>
FlowOpszipAll(Graph<SourceShape<U>,?> that, A thisElem, U thatElem)
Combine the elements of current flow and the givenSource
into a stream of tuples.<U,A,Mat2>
Flow<Out,scala.Tuple2<A,U>,Mat2>zipAllFlow(Graph<SourceShape<U>,Mat2> that, A thisElem, U thatElem)
<U,M>
Graph<FlowShape<Out,scala.Tuple2<Out,U>>,M>zipGraph(Graph<SourceShape<U>,M> that)
<U> FlowOps
zipLatest(Graph<SourceShape<U>,?> that)
Combine the elements of 2 streams into a stream of tuples, picking always the latest element of each.<U,M>
Graph<FlowShape<Out,scala.Tuple2<Out,U>>,M>zipLatestGraph(Graph<SourceShape<U>,M> that)
<Out2,Out3>
FlowOpszipLatestWith(Graph<SourceShape<Out2>,?> that, boolean eagerComplete, scala.Function2<Out,Out2,Out3> combine)
Combine the elements of multiple streams into a stream of combined elements using a combiner function, picking always the latest of the elements of each source.<Out2,Out3>
FlowOpszipLatestWith(Graph<SourceShape<Out2>,?> that, scala.Function2<Out,Out2,Out3> combine)
Combine the elements of multiple streams into a stream of combined elements using a combiner function, picking always the latest of the elements of each source.<Out2,Out3,M>
Graph<FlowShape<Out,Out3>,M>zipLatestWithGraph(Graph<SourceShape<Out2>,M> that, boolean eagerComplete, scala.Function2<Out,Out2,Out3> combine)
<Out2,Out3,M>
Graph<FlowShape<Out,Out3>,M>zipLatestWithGraph(Graph<SourceShape<Out2>,M> that, scala.Function2<Out,Out2,Out3> combine)
<Out2,Out3>
FlowOpszipWith(Graph<SourceShape<Out2>,?> that, scala.Function2<Out,Out2,Out3> combine)
Put together the elements of current flow and the givenSource
into a stream of combined elements using a combiner function.<Out2,Out3,M>
Graph<FlowShape<Out,Out3>,M>zipWithGraph(Graph<SourceShape<Out2>,M> that, scala.Function2<Out,Out2,Out3> combine)
FlowOps
zipWithIndex()
Combine the elements of current flow into a stream of tuples consisting of all elements paired with their index.
-
-
-
Method Detail
-
$plus$plus
<U,M> FlowOps $plus$plus(Graph<SourceShape<U>,M> that)
Concatenates thisFlow
with the givenSource
so the first element emitted by that source is emitted after the last element of this flow.This is a shorthand for
<U,Mat2>concat(akka.stream.Graph<akka.stream.SourceShape<U>,Mat2>)
-
addAttributes
FlowOps addAttributes(Attributes attr)
-
aggregateWithBoundary
<Agg,Emit> FlowOps aggregateWithBoundary(scala.Function0<Agg> allocate, scala.Function2<Agg,Out,scala.Tuple2<Agg,java.lang.Object>> aggregate, scala.Function1<Agg,Emit> harvest, scala.Option<scala.Tuple2<scala.Function1<Agg,java.lang.Object>,scala.concurrent.duration.FiniteDuration>> emitOnTimer)
Aggregate input elements into an arbitrary data structure that can be completed and emitted downstream when custom condition is met which can be triggered by aggregate or timer. It can be thought of a more generalgroupedWeightedWithin(long,scala.concurrent.duration.FiniteDuration,scala.Function1<Out,java.lang.Object>)
.'''Emits when''' the aggregation function decides the aggregate is complete or the timer function returns true
'''Backpressures when''' downstream backpressures and the aggregate is complete
'''Completes when''' upstream completes and the last aggregate has been emitted downstream
'''Cancels when''' downstream cancels
- Parameters:
allocate
- allocate the initial data structure for aggregated elementsaggregate
- update the aggregated elements, return true if ready to emit after update.harvest
- this is invoked before emit within the current stage/operatoremitOnTimer
- decide whether the current aggregated elements can be emitted, the custom function is invoked on every interval
-
alsoTo
FlowOps alsoTo(Graph<SinkShape<Out>,?> that)
Attaches the givenSink
to thisSource
, meaning that elements that pass through will also be sent to theSink
.It is similar to
wireTap(scala.Function1<Out, scala.runtime.BoxedUnit>)
but will backpressure instead of dropping elements when the givenSink
is not ready.'''Emits when''' element is available and demand exists both from the Sink and the downstream.
'''Backpressures when''' downstream or Sink backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream or Sink cancels
-
alsoToAll
FlowOps alsoToAll(scala.collection.immutable.Seq<Graph<SinkShape<Out>,?>> those)
Attaches the givenSink
s to thisSource
, meaning that elements that pass through will also be sent to theSink
.It is similar to
wireTap(scala.Function1<Out, scala.runtime.BoxedUnit>)
but will backpressure instead of dropping elements when the givenSink
s is not ready.'''Emits when''' element is available and demand exists both from the Sinks and the downstream.
'''Backpressures when''' downstream or any of the
Sink
s backpressures'''Completes when''' upstream completes
'''Cancels when''' downstream or any of the
Sink
s cancels
-
ask
<S> FlowOps ask(ActorRef ref, Timeout timeout, scala.reflect.ClassTag<S> tag)
Use theask
pattern to send a request-reply message to the targetref
actor. If any of the asks times out it will fail the stream with aAskTimeoutException
.Do not forget to include the expected response type in the method call, like so:
flow.ask[ExpectedReply](ref)
otherwise
Nothing
will be assumed, which is most likely not what you want.Defaults to parallelism of 2 messages in flight, since while one ask message may be being worked on, the second one still be in the mailbox, so defaulting to sending the second one a bit earlier than when first ask has replied maintains a slightly healthier throughput.
Similar to the plain ask pattern, the target actor is allowed to reply with
akka.util.Status
. Anakka.util.Status#Failure
will cause the operator to fail with the cause carried in theFailure
message.The operator fails with an
WatchedActorTerminatedException
if the target actor is terminated.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the futures (in submission order) created by the ask pattern internally are completed
'''Backpressures when''' the number of futures reaches the configured parallelism and the downstream backpressures
'''Completes when''' upstream completes and all futures have been completed and all elements have been emitted
'''Fails when''' the passed in actor terminates, or a timeout is exceeded in any of the asks performed
'''Cancels when''' downstream cancels
-
ask
<S> FlowOps ask(int parallelism, ActorRef ref, Timeout timeout, scala.reflect.ClassTag<S> tag)
Use theask
pattern to send a request-reply message to the targetref
actor. If any of the asks times out it will fail the stream with aAskTimeoutException
.Do not forget to include the expected response type in the method call, like so:
flow.ask[ExpectedReply](parallelism = 4)(ref)
otherwise
Nothing
will be assumed, which is most likely not what you want.Parallelism limits the number of how many asks can be "in flight" at the same time. Please note that the elements emitted by this operator are in-order with regards to the asks being issued (i.e. same behaviour as mapAsync).
The operator fails with an
WatchedActorTerminatedException
if the target actor is terminated, or with anTimeoutException
in case the ask exceeds the timeout passed in.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the futures (in submission order) created by the ask pattern internally are completed
'''Backpressures when''' the number of futures reaches the configured parallelism and the downstream backpressures
'''Completes when''' upstream completes and all futures have been completed and all elements have been emitted
'''Fails when''' the passed in actor terminates, or a timeout is exceeded in any of the asks performed
'''Cancels when''' downstream cancels
-
async
FlowOps async()
Put an asynchronous boundary around thisFlow
.If this is a
SubFlow
(created e.g. bygroupBy
), this creates an asynchronous boundary around each materialized sub-flow, not the super-flow. That way, the super-flow will communicate with sub-flows asynchronously.
-
backpressureTimeout
FlowOps backpressureTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the time between the emission of an element and the following downstream demand exceeds the provided timeout, the stream is failed with aBackpressureTimeoutException
. The timeout is checked periodically, so the resolution of the check is one period (equals to timeout value).'''Emits when''' upstream emits an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or fails if timeout elapses between element emission and downstream demand.
'''Cancels when''' downstream cancels
-
batch
<S> FlowOps batch(long max, scala.Function1<Out,S> seed, scala.Function2<S,Out,S> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by aggregating elements into batches until the subscriber is ready to accept them. For example a batch step might store received elements in an array up to the allowed max limit if the upstream publisher is faster.This only rolls up elements if the upstream is faster, but if the downstream is faster it will not duplicate elements.
Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' downstream stops backpressuring and there is an aggregated element available
'''Backpressures when''' there are
max
batched elements and 1 pending element and downstream backpressures'''Completes when''' upstream completes and there is no batched/pending element waiting
'''Cancels when''' downstream cancels
See also
FlowOps.conflateWithSeed
,FlowOps.batchWeighted
- Parameters:
max
- maximum number of elements to batch before backpressuring upstream (must be positive non-zero)seed
- Provides the first state for a batched value using the first unconsumed element as a startaggregate
- Takes the currently batched value and the current pending element to produce a new aggregate
-
batchWeighted
<S> FlowOps batchWeighted(long max, scala.Function1<Out,java.lang.Object> costFn, scala.Function1<Out,S> seed, scala.Function2<S,Out,S> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by aggregating elements into batches until the subscriber is ready to accept them. For example a batch step might concatenateByteString
elements up to the allowed max limit if the upstream publisher is faster.This element only rolls up elements if the upstream is faster, but if the downstream is faster it will not duplicate elements.
Batching will apply for all elements, even if a single element cost is greater than the total allowed limit. In this case, previous batched elements will be emitted, then the "heavy" element will be emitted (after being applied with the
seed
function) without batching further elements with it, and then the rest of the incoming elements are batched.'''Emits when''' downstream stops backpressuring and there is a batched element available
'''Backpressures when''' there are
max
weighted batched elements + 1 pending element and downstream backpressures'''Completes when''' upstream completes and there is no batched/pending element waiting
'''Cancels when''' downstream cancels
See also
FlowOps.conflateWithSeed
,FlowOps.batch
- Parameters:
max
- maximum weight of elements to batch before backpressuring upstream (must be positive non-zero)costFn
- a function to compute a single element weightseed
- Provides the first state for a batched value using the first unconsumed element as a startaggregate
- Takes the currently batched value and the current pending element to produce a new batch
-
buffer
FlowOps buffer(int size, OverflowStrategy overflowStrategy)
Adds a fixed size buffer in the flow that allows to store elements from a faster upstream until it becomes full. Depending on the definedOverflowStrategy
it might drop elements or backpressure the upstream if there is no space available'''Emits when''' downstream stops backpressuring and there is a pending element in the buffer
'''Backpressures when''' downstream backpressures or depending on OverflowStrategy:
- Backpressure - backpressures when buffer is full
- DropHead, DropTail, DropBuffer - never backpressures
- Fail - fails the stream if buffer gets full
'''Completes when''' upstream completes and buffered elements have been drained
'''Cancels when''' downstream cancels
- Parameters:
size
- The size of the buffer in element countoverflowStrategy
- Strategy that is used when incoming elements cannot fit inside the buffer
-
collect
<T> FlowOps collect(scala.PartialFunction<Out,T> pf)
Transform this stream by applying the given partial function to each of the elements on which the function is defined as they pass through this processing step. Non-matching elements are filtered out.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the provided partial function is defined for the element
'''Backpressures when''' the partial function is defined for the element and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
collectType
<T> FlowOps collectType(scala.reflect.ClassTag<T> tag)
Transform this stream by testing the type of each of the elements on which the element is an instance of the provided type as they pass through this processing step.Non-matching elements are filtered out.
Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the element is an instance of the provided type
'''Backpressures when''' the element is an instance of the provided type and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
completionTimeout
FlowOps completionTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the completion of the stream does not happen until the provided timeout, the stream is failed with aCompletionTimeoutException
.'''Emits when''' upstream emits an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or fails if timeout elapses before upstream completes
'''Cancels when''' downstream cancels
-
concat
<U,Mat2> FlowOps concat(Graph<SourceShape<U>,Mat2> that)
Concatenate the givenSource
to thisFlow
, meaning that once this Flow’s input is exhausted and all result elements have been generated, the Source’s elements will be produced.Note that the
Source
is materialized together with this Flow and is "detached" meaning it will in effect behave as a one element buffer in front of both the sources, that eagerly demands an element on start (so it can not be combined withSource.lazy
to defer materialization ofthat
).The second source is then kept from producing elements by asserting back-pressure until its time comes.
When needing a concat operator that is not detached use
concatLazy(akka.stream.Graph<akka.stream.SourceShape<U>, Mat2>)
If this
Flow
gets upstream error - no elements from the givenSource
will be pulled.'''Emits when''' element is available from current stream or from the given
Source
when current is completed'''Backpressures when''' downstream backpressures
'''Completes when''' given
Source
completes'''Cancels when''' downstream cancels
-
concatAllLazy
<U> FlowOps concatAllLazy(scala.collection.immutable.Seq<Graph<SourceShape<U>,?>> those)
Concatenate the givenSource
s to thisFlow
, meaning that once this Flow’s input is exhausted and all result elements have been generated, theSource
s' elements will be produced.Note that the
Source
s are materialized together with this Flow. Iflazy
materialization is what is needed the operator can be combined with for exampleSource.lazySource
to defer materialization ofthat
until the time when this source completes.The second source is then kept from producing elements by asserting back-pressure until its time comes.
For a concat operator that is detached, use
concat(akka.stream.Graph<akka.stream.SourceShape<U>, Mat2>)
If this
Flow
gets upstream error - no elements from the givenSource
s will be pulled.'''Emits when''' element is available from current stream or from the given
Source
s when current is completed'''Backpressures when''' downstream backpressures
'''Completes when''' given all those
Source
s completes'''Cancels when''' downstream cancels
-
concatGraph
<U,Mat2> Graph<FlowShape<Out,U>,Mat2> concatGraph(Graph<SourceShape<U>,Mat2> that, boolean detached)
-
concatLazy
<U,Mat2> FlowOps concatLazy(Graph<SourceShape<U>,Mat2> that)
Concatenate the givenSource
to thisFlow
, meaning that once this Flow’s input is exhausted and all result elements have been generated, the Source’s elements will be produced.Note that the
Source
is materialized together with this Flow. Iflazy
materialization is what is needed the operator can be combined with for exampleSource.lazySource
to defer materialization ofthat
until the time when this source completes.The second source is then kept from producing elements by asserting back-pressure until its time comes.
For a concat operator that is detached, use
concat(akka.stream.Graph<akka.stream.SourceShape<U>, Mat2>)
If this
Flow
gets upstream error - no elements from the givenSource
will be pulled.'''Emits when''' element is available from current stream or from the given
Source
when current is completed'''Backpressures when''' downstream backpressures
'''Completes when''' given
Source
completes'''Cancels when''' downstream cancels
-
conflate
<O2> FlowOps conflate(scala.Function2<O2,O2,O2> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by conflating elements into a summary until the subscriber is ready to accept them. For example a conflate step might average incoming numbers if the upstream publisher is faster.This version of conflate does not change the output type of the stream. See
FlowOps.conflateWithSeed
for a more flexible version that can take a seed function and transform elements while rolling up.This element only rolls up elements if the upstream is faster, but if the downstream is faster it will not duplicate elements.
Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' downstream stops backpressuring and there is a conflated element available
'''Backpressures when''' never
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
- Parameters:
aggregate
- Takes the currently aggregated value and the current pending element to produce a new aggregateSee also
FlowOps.conflate
,FlowOps.limit
,FlowOps.limitWeighted
FlowOps.batch
FlowOps.batchWeighted
-
conflateWithSeed
<S> FlowOps conflateWithSeed(scala.Function1<Out,S> seed, scala.Function2<S,Out,S> aggregate)
Allows a faster upstream to progress independently of a slower subscriber by conflating elements into a summary until the subscriber is ready to accept them. For example a conflate step might average incoming numbers if the upstream publisher is faster.This version of conflate allows to derive a seed from the first element and change the aggregated type to be different than the input type. See
FlowOps.conflate
for a simpler version that does not change types.This element only rolls up elements if the upstream is faster, but if the downstream is faster it will not duplicate elements.
Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' downstream stops backpressuring and there is a conflated element available
'''Backpressures when''' never
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
- Parameters:
seed
- Provides the first state for a conflated value using the first unconsumed element as a startaggregate
- Takes the currently aggregated value and the current pending element to produce a new aggregateSee also
FlowOps.conflate
,FlowOps.limit
,FlowOps.limitWeighted
FlowOps.batch
FlowOps.batchWeighted
-
delay
FlowOps delay(scala.concurrent.duration.FiniteDuration of, DelayOverflowStrategy strategy)
Shifts elements emission in time by a specified amount. It allows to store elements in internal buffer while waiting for next element to be emitted. Depending on the definedDelayOverflowStrategy
it might drop elements or backpressure the upstream if there is no space available in the buffer.Delay precision is 10ms to avoid unnecessary timer scheduling cycles
Internal buffer has default capacity 16. You can set buffer size by calling
addAttributes(inputBuffer)
'''Emits when''' there is a pending element in the buffer and configured time for this element elapsed * EmitEarly - strategy do not wait to emit element if buffer is full
'''Backpressures when''' depending on OverflowStrategy * Backpressure - backpressures when buffer is full * DropHead, DropTail, DropBuffer - never backpressures * Fail - fails the stream if buffer gets full
'''Completes when''' upstream completes and buffered elements have been drained
'''Cancels when''' downstream cancels
- Parameters:
of
- time to shift all messagesstrategy
- Strategy that is used when incoming elements cannot fit inside the buffer
-
delay$default$2
DelayOverflowStrategy delay$default$2()
-
delayWith
FlowOps delayWith(scala.Function0<DelayStrategy<Out>> delayStrategySupplier, DelayOverflowStrategy overFlowStrategy)
Shifts elements emission in time by an amount individually determined through delay strategy a specified amount. It allows to store elements in internal buffer while waiting for next element to be emitted. Depending on the definedDelayOverflowStrategy
it might drop elements or backpressure the upstream if there is no space available in the buffer.It determines delay for each ongoing element invoking
DelayStrategy.nextDelay(elem: T): FiniteDuration
.Note that elements are not re-ordered: if an element is given a delay much shorter than its predecessor, it will still have to wait for the preceding element before being emitted. It is also important to notice that
scaladsl.DelayStrategy
can be stateful.Delay precision is 10ms to avoid unnecessary timer scheduling cycles.
Internal buffer has default capacity 16. You can set buffer size by calling
addAttributes(inputBuffer)
'''Emits when''' there is a pending element in the buffer and configured time for this element elapsed * EmitEarly - strategy do not wait to emit element if buffer is full
'''Backpressures when''' depending on OverflowStrategy * Backpressure - backpressures when buffer is full * DropHead, DropTail, DropBuffer - never backpressures * Fail - fails the stream if buffer gets full
'''Completes when''' upstream completes and buffered elements have been drained
'''Cancels when''' downstream cancels
- Parameters:
delayStrategySupplier
- creates newDelayStrategy
object for each materializationoverFlowStrategy
- Strategy that is used when incoming elements cannot fit inside the buffer
-
detach
FlowOps detach()
Detaches upstream demand from downstream demand without detaching the stream rates; in other words acts like a buffer of size 1.'''Emits when''' upstream emits an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
divertTo
FlowOps divertTo(Graph<SinkShape<Out>,?> that, scala.Function1<Out,java.lang.Object> when)
Attaches the givenSink
to thisFlow
, meaning that elements will be sent to theSink
instead of being passed through if the predicatewhen
returnstrue
.'''Emits when''' emits when an element is available from the input and the chosen output has demand
'''Backpressures when''' the currently chosen output back-pressures
'''Completes when''' upstream completes and no output is pending
'''Cancels when''' any of the downstreams cancel
-
divertToGraph
<M> Graph<FlowShape<Out,Out>,M> divertToGraph(Graph<SinkShape<Out>,M> that, scala.Function1<Out,java.lang.Object> when)
-
drop
FlowOps drop(long n)
Discard the given number of elements at the beginning of the stream. No elements will be dropped ifn
is zero or negative.'''Emits when''' the specified number of elements has been dropped already
'''Backpressures when''' the specified number of elements has been dropped and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
dropWhile
FlowOps dropWhile(scala.Function1<Out,java.lang.Object> p)
Discard elements at the beginning of the stream while predicate is true. All elements will be taken after predicate returns false first time.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' predicate returned false and for all following stream elements
'''Backpressures when''' predicate returned false and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
dropWithin
FlowOps dropWithin(scala.concurrent.duration.FiniteDuration d)
Discard the elements received within the given duration at beginning of the stream.'''Emits when''' the specified time elapsed and a new upstream element arrives
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
expand
<U> FlowOps expand(scala.Function1<Out,scala.collection.Iterator<U>> expander)
Allows a faster downstream to progress independently of a slower upstream by extrapolating elements from an older element until new element comes from the upstream. For example an expand step might repeat the last element for the subscriber until it receives an update from upstream.This element will never "drop" upstream elements as all elements go through at least one extrapolation step. This means that if the upstream is actually faster than the upstream it will be backpressured by the downstream subscriber.
Expand does not support
akka.stream.Supervision.Restart
andakka.stream.Supervision.Resume
. Exceptions from theseed
function will complete the stream with failure.'''Emits when''' downstream stops backpressuring
'''Backpressures when''' downstream backpressures or iterator runs empty
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
- Parameters:
expander
- Takes the current extrapolation state to produce an output element and the next extrapolation state.
-
extrapolate
<U> FlowOps extrapolate(scala.Function1<U,scala.collection.Iterator<U>> extrapolator, scala.Option<U> initial)
Allows a faster downstream to progress independent of a slower upstream.This is achieved by introducing "extrapolated" elements - based on those from upstream - whenever downstream signals demand.
Extrapolate does not support
akka.stream.Supervision.Restart
andakka.stream.Supervision.Resume
. Exceptions from theextrapolate
function will complete the stream with failure.'''Emits when''' downstream stops backpressuring, AND EITHER upstream emits OR initial element is present OR
extrapolate
is non-empty and applicable'''Backpressures when''' downstream backpressures or current
extrapolate
runs empty'''Completes when''' upstream completes and current
extrapolate
runs empty'''Cancels when''' downstream cancels
- Parameters:
extrapolator
- takes the current upstream element and provides a sequence of "extrapolated" elements based on the original, to be emitted in case downstream signals demand.initial
- the initial element to be emitted, in case upstream is able to stall the entire stream.
-
extrapolate$default$2
<U> scala.None$ extrapolate$default$2()
-
filter
FlowOps filter(scala.Function1<Out,java.lang.Object> p)
Only pass on those elements that satisfy the given predicate.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the given predicate returns true for the element
'''Backpressures when''' the given predicate returns true for the element and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
filterNot
FlowOps filterNot(scala.Function1<Out,java.lang.Object> p)
Only pass on those elements that NOT satisfy the given predicate.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the given predicate returns false for the element
'''Backpressures when''' the given predicate returns false for the element and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
flatMapConcat
<T,M> FlowOps flatMapConcat(scala.Function1<Out,Graph<SourceShape<T>,M>> f)
Transform each input element into aSource
of output elements that is then flattened into the output stream by concatenation, fully consuming one Source after the other.'''Emits when''' a currently consumed substream has an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes and all consumed substreams complete
'''Cancels when''' downstream cancels
-
flatMapMerge
<T,M> FlowOps flatMapMerge(int breadth, scala.Function1<Out,Graph<SourceShape<T>,M>> f)
Transform each input element into aSource
of output elements that is then flattened into the output stream by merging, where at mostbreadth
substreams are being consumed at any given time.'''Emits when''' a currently consumed substream has an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes and all consumed substreams complete
'''Cancels when''' downstream cancels
-
flatMapPrefix
<Out2,Mat2> FlowOps flatMapPrefix(int n, scala.Function1<scala.collection.immutable.Seq<Out>,Flow<Out,Out2,Mat2>> f)
Takes up ton
elements from the stream (less thann
only if the upstream completes before emittingn
elements), then applyf
on these elements in order to obtain a flow, this flow is then materialized and the rest of the input is processed by this flow (similar to via). This method returns a flow consuming the rest of the stream producing the materialized flow's output.'''Emits when''' the materialized flow emits. Notice the first
n
elements are buffered internally before materializing the flow and connecting it to the rest of the upstream - producing elements at its own discretion (might 'swallow' or multiply elements).'''Backpressures when''' the materialized flow backpressures
'''Completes when''' the materialized flow completes. If upstream completes before producing
n
elements,f
will be applied with the provided elements, the resulting flow will be materialized and signalled for upstream completion, it can then complete or continue to emit elements at its own discretion.'''Cancels when''' the materialized flow cancels. When downstream cancels before materialization of the nested flow, the operator's default behaviour is to cancel immediately, this behaviour can be controlled by setting the
Attributes.NestedMaterializationCancellationPolicy
attribute on the flow. When this attribute is configured to true, downstream cancellation is delayed until the nested flow's materialization which is then immediately cancelled (with the original cancellation cause).- Parameters:
n
- the number of elements to accumulate before materializing the downstream flow.f
- a function that produces the downstream flow based on the upstream's prefix.
-
fold
<T> FlowOps fold(T zero, scala.Function2<T,Out,T> f)
Similar toscan
but only emits its result when the upstream completes, after which it also completes. Applies the given function towards its current and next value, yielding the next current value.If the function
f
throws an exception and the supervision decision isakka.stream.Supervision.Restart
current value starts atzero
again the stream will continue.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.Note that the
zero
value must be immutable.'''Emits when''' upstream completes
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
See also
FlowOps.scan
-
foldAsync
<T> FlowOps foldAsync(T zero, scala.Function2<T,Out,scala.concurrent.Future<T>> f)
Similar tofold
but with an asynchronous function. Applies the given function towards its current and next value, yielding the next current value.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.If the function
f
returns a failure and the supervision decision isakka.stream.Supervision.Restart
current value starts atzero
again the stream will continue.Note that the
zero
value must be immutable.'''Emits when''' upstream completes
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
See also
FlowOps.fold
-
groupBy
<K> SubFlow<Out,Mat,FlowOps,java.lang.Object> groupBy(int maxSubstreams, scala.Function1<Out,K> f, boolean allowClosedSubstreamRecreation)
This operation demultiplexes the incoming stream into separate output streams, one for each element key. The key is computed for each element using the given function. When a new key is encountered for the first time a new substream is opened and subsequently fed with all elements belonging to that key.WARNING: If
allowClosedSubstreamRecreation
is set tofalse
(default behavior) the operator keeps track of all keys of streams that have already been closed. If you expect an infinite number of keys this can cause memory issues. Elements belonging to those keys are drained directly and not send to the substream.Note: If
allowClosedSubstreamRecreation
is set totrue
substream completion and incoming elements are subject to race-conditions. If elements arrive for a stream that is in the process of closing these elements might get lost.The object returned from this method is not a normal
Source
orFlow
, it is aSubFlow
. This means that after this operator all transformations are applied to all encountered substreams in the same fashion. Substream mode is exited either by closing the substream (i.e. connecting it to aSink
) or by merging the substreams back together; see theto
andmergeBack
methods onSubFlow
for more information.It is important to note that the substreams also propagate back-pressure as any other stream, which means that blocking one substream will block the
groupBy
operator itself—and thereby all substreams—once all internal or explicit buffers are filled.If the group by function
f
throws an exception and the supervision decision isakka.stream.Supervision.Stop
the stream and substreams will be completed with failure.If the group by function
f
throws an exception and the supervision decision isakka.stream.Supervision.Resume
orakka.stream.Supervision.Restart
the element is dropped and the stream and substreams continue.Function
f
MUST NOT returnnull
. This will throw exception and trigger supervision decision mechanism.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' an element for which the grouping function returns a group that has not yet been created. Emits the new group
'''Backpressures when''' there is an element pending for a group whose substream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels and all substreams cancel
- Parameters:
maxSubstreams
- configures the maximum number of substreams (keys) that are supported; if more distinct keys are encountered then the stream failsf
- computes the key for each elementallowClosedSubstreamRecreation
- enables recreation of already closed substreams if elements with their corresponding keys arrive after completion
-
groupBy
<K> SubFlow<Out,Mat,FlowOps,java.lang.Object> groupBy(int maxSubstreams, scala.Function1<Out,K> f)
This operation demultiplexes the incoming stream into separate output streams, one for each element key. The key is computed for each element using the given function. When a new key is encountered for the first time a new substream is opened and subsequently fed with all elements belonging to that key.WARNING: The operator keeps track of all keys of streams that have already been closed. If you expect an infinite number of keys this can cause memory issues. Elements belonging to those keys are drained directly and not send to the substream.
-
grouped
FlowOps grouped(int n)
Chunk up this stream into groups of the given size, with the last group possibly smaller than requested due to end-of-stream.n
must be positive, otherwise IllegalArgumentException is thrown.'''Emits when''' the specified number of elements have been accumulated or upstream completed
'''Backpressures when''' a group has been assembled and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
groupedWeighted
FlowOps groupedWeighted(long minWeight, scala.Function1<Out,java.lang.Object> costFn)
Chunk up this stream into groups of elements that have a cumulative weight greater than or equal to theminWeight
, with the last group possibly smaller than requestedminWeight
due to end-of-stream.minWeight
must be positive, otherwise IllegalArgumentException is thrown.costFn
must return a non-negative result for all inputs, otherwise the stage will fail with an IllegalArgumentException.'''Emits when''' the cumulative weight of elements is greater than or equal to the
minWeight
or upstream completed'''Backpressures when''' a buffered group weighs more than
minWeight
and downstream backpressures'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
groupedWeightedWithin
FlowOps groupedWeightedWithin(long maxWeight, scala.concurrent.duration.FiniteDuration d, scala.Function1<Out,java.lang.Object> costFn)
Chunk up this stream into groups of elements received within a time window, or limited by the weight of the elements, whatever happens first. Empty groups will not be emitted if no elements are received from upstream. The last group before end-of-stream will contain the buffered elements since the previously emitted group.maxWeight
must be positive, andd
must be greater than 0 seconds, otherwise IllegalArgumentException is thrown.'''Emits when''' the configured time elapses since the last group has been emitted or weight limit reached
'''Backpressures when''' downstream backpressures, and buffered group (+ pending element) weighs more than
maxWeight
'''Completes when''' upstream completes (emits last group)
'''Cancels when''' downstream completes
-
groupedWeightedWithin
FlowOps groupedWeightedWithin(long maxWeight, int maxNumber, scala.concurrent.duration.FiniteDuration d, scala.Function1<Out,java.lang.Object> costFn)
Chunk up this stream into groups of elements received within a time window, or limited by the weight and number of the elements, whatever happens first. Empty groups will not be emitted if no elements are received from upstream. The last group before end-of-stream will contain the buffered elements since the previously emitted group.maxWeight
must be positive,maxNumber
must be positive, andd
must be greater than 0 seconds, otherwise IllegalArgumentException is thrown.'''Emits when''' the configured time elapses since the last group has been emitted or weight limit reached
'''Backpressures when''' downstream backpressures, and buffered group (+ pending element) weighs more than
maxWeight
or has more thanmaxNumber
elements'''Completes when''' upstream completes (emits last group)
'''Cancels when''' downstream completes
-
groupedWithin
FlowOps groupedWithin(int n, scala.concurrent.duration.FiniteDuration d)
Chunk up this stream into groups of elements received within a time window, or limited by the given number of elements, whatever happens first. Empty groups will not be emitted if no elements are received from upstream. The last group before end-of-stream will contain the buffered elements since the previously emitted group.n
must be positive, andd
must be greater than 0 seconds, otherwise IllegalArgumentException is thrown.'''Emits when''' the configured time elapses since the last group has been emitted or
n
elements is buffered'''Backpressures when''' downstream backpressures, and there are
n+1
buffered elements'''Completes when''' upstream completes (emits last group)
'''Cancels when''' downstream completes
-
idleTimeout
FlowOps idleTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the time between two processed elements exceeds the provided timeout, the stream is failed with aStreamIdleTimeoutException
. The timeout is checked periodically, so the resolution of the check is one period (equals to timeout value).'''Emits when''' upstream emits an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or fails if timeout elapses between two emitted elements
'''Cancels when''' downstream cancels
-
initialDelay
FlowOps initialDelay(scala.concurrent.duration.FiniteDuration delay)
Delays the initial element by the specified duration.'''Emits when''' upstream emits an element if the initial delay is already elapsed
'''Backpressures when''' downstream backpressures or initial delay is not yet elapsed
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
initialTimeout
FlowOps initialTimeout(scala.concurrent.duration.FiniteDuration timeout)
If the first element has not passed through this operator before the provided timeout, the stream is failed with aInitialTimeoutException
.'''Emits when''' upstream emits an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or fails if timeout elapses before first element arrives
'''Cancels when''' downstream cancels
-
interleave
<U> FlowOps interleave(Graph<SourceShape<U>,?> that, int segmentSize)
Interleave is a deterministic merge of the givenSource
with elements of thisFlow
. It first emitssegmentSize
number of elements from this flow to downstream, then - same amount forthat
source, then repeat process.Example:
Source(List(1, 2, 3)).interleave(List(4, 5, 6, 7), 2) // 1, 2, 4, 5, 3, 6, 7
After one of upstreams is complete then all the rest elements will be emitted from the second one
If it gets error from one of upstreams - stream completes with failure.
'''Emits when''' element is available from the currently consumed upstream
'''Backpressures when''' downstream backpressures. Signal to current upstream, switch to next upstream when received
segmentSize
elements'''Completes when''' the
Flow
and givenSource
completes'''Cancels when''' downstream cancels
-
interleave
<U> FlowOps interleave(Graph<SourceShape<U>,?> that, int segmentSize, boolean eagerClose)
Interleave is a deterministic merge of the givenSource
with elements of thisFlow
. It first emitssegmentSize
number of elements from this flow to downstream, then - same amount forthat
source, then repeat process.If eagerClose is false and one of the upstreams complete the elements from the other upstream will continue passing through the interleave operator. If eagerClose is true and one of the upstream complete interleave will cancel the other upstream and complete itself.
If it gets error from one of upstreams - stream completes with failure.
'''Emits when''' element is available from the currently consumed upstream
'''Backpressures when''' downstream backpressures. Signal to current upstream, switch to next upstream when received
segmentSize
elements'''Completes when''' the
Flow
and givenSource
completes'''Cancels when''' downstream cancels
-
interleaveAll
<U> FlowOps interleaveAll(scala.collection.immutable.Seq<Graph<SourceShape<U>,?>> those, int segmentSize, boolean eagerClose)
Interleave is a deterministic merge of the givenSource
s with elements of thisFlow
. It first emitssegmentSize
number of elements from this flow to downstream, then - same amount forthat
source, then repeat process.If eagerClose is false and one of the upstreams complete the elements from the other upstream will continue passing through the interleave operator. If eagerClose is true and one of the upstream complete interleave will cancel the other upstream and complete itself.
If it gets error from one of upstreams - stream completes with failure.
'''Emits when''' element is available from the currently consumed upstream
'''Backpressures when''' downstream backpressures. Signal to current upstream, switch to next upstream when received
segmentSize
elements'''Completes when''' the
Flow
and givenSource
completes'''Cancels when''' downstream cancels
-
interleaveGraph
<U,M> Graph<FlowShape<Out,U>,M> interleaveGraph(Graph<SourceShape<U>,M> that, int segmentSize, boolean eagerClose)
-
interleaveGraph$default$3
<U,M> boolean interleaveGraph$default$3()
-
internalConcat
<U,Mat2> FlowOps internalConcat(Graph<SourceShape<U>,Mat2> that, boolean detached)
-
internalConcatAll
<U> FlowOps internalConcatAll(Graph<SourceShape<U>,?>[] those, boolean detached)
-
intersperse
<T> FlowOps intersperse(T start, T inject, T end)
Intersperses stream with provided element, similar to howscala.collection.immutable.List.mkString
injects a separator between a List's elements.Additionally can inject start and end marker elements to stream.
Examples:
val nums = Source(List(1,2,3)).map(_.toString) nums.intersperse(",") // 1 , 2 , 3 nums.intersperse("[", ",", "]") // [ 1 , 2 , 3 ]
In case you want to only prepend or only append an element (yet still use the
intercept
feature to inject a separator between elements, you may want to use the following pattern instead of the 3-argument version of intersperse (SeeSource.concat
for semantics details):Source.single(">> ") ++ Source(List("1", "2", "3")).intersperse(",") Source(List("1", "2", "3")).intersperse(",") ++ Source.single("END")
'''Emits when''' upstream emits (or before with the
start
element if provided)'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
intersperse
<T> FlowOps intersperse(T inject)
Intersperses stream with provided element, similar to howscala.collection.immutable.List.mkString
injects a separator between a List's elements.Additionally can inject start and end marker elements to stream.
Examples:
val nums = Source(List(1,2,3)).map(_.toString) nums.intersperse(",") // 1 , 2 , 3 nums.intersperse("[", ",", "]") // [ 1 , 2 , 3 ]
'''Emits when''' upstream emits (or before with the
start
element if provided)'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
keepAlive
<U> FlowOps keepAlive(scala.concurrent.duration.FiniteDuration maxIdle, scala.Function0<U> injectedElem)
Injects additional elements if upstream does not emit for a configured amount of time. In other words, this operator attempts to maintains a base rate of emitted elements towards the downstream.If the downstream backpressures then no element is injected until downstream demand arrives. Injected elements do not accumulate during this period.
Upstream elements are always preferred over injected elements.
'''Emits when''' upstream emits an element or if the upstream was idle for the configured period
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
limit
FlowOps limit(long max)
Ensure stream boundedness by limiting the number of elements from upstream. If the number of incoming elements exceeds max, it will signal upstream failureStreamLimitException
downstream.Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.
'''Emits when''' upstream emits and the number of emitted elements has not reached max
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes and the number of emitted elements has not reached max
'''Errors when''' the total number of incoming element exceeds max
'''Cancels when''' downstream cancels
See also
FlowOps.take
,FlowOps.takeWithin
,FlowOps.takeWhile
-
limitWeighted
<T> FlowOps limitWeighted(long max, scala.Function1<Out,java.lang.Object> costFn)
Ensure stream boundedness by evaluating the cost of incoming elements using a cost function. Exactly how many elements will be allowed to travel downstream depends on the evaluated cost of each element. If the accumulated cost exceeds max, it will signal upstream failureStreamLimitException
downstream.Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.
Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' upstream emits and the accumulated cost has not reached max
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes and the number of emitted elements has not reached max
'''Errors when''' when the accumulated cost exceeds max
'''Cancels when''' downstream cancels
See also
FlowOps.take
,FlowOps.takeWithin
,FlowOps.takeWhile
-
log
FlowOps log(java.lang.String name, scala.Function1<Out,java.lang.Object> extract, LoggingAdapter log)
Logs elements flowing through the stream as well as completion and erroring.By default element and completion signals are logged on debug level, and errors are logged on Error level. This can be adjusted according to your needs by providing a custom
Attributes.LogLevels
attribute on the given Flow:Uses implicit
LoggingAdapter
if available, otherwise uses an internally created one, which usesakka.event.Logging(materializer.system, materializer)
as its source (use this class to configure slf4j loggers).Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the mapping function returns an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
log$default$2
scala.Function1<Out,java.lang.Object> log$default$2()
-
log$default$3
LoggingAdapter log$default$3(java.lang.String name, scala.Function1<Out,java.lang.Object> extract)
-
logWithMarker
FlowOps logWithMarker(java.lang.String name, scala.Function1<Out,LogMarker> marker, scala.Function1<Out,java.lang.Object> extract, MarkerLoggingAdapter log)
Logs elements flowing through the stream as well as completion and erroring.By default element and completion signals are logged on debug level, and errors are logged on Error level. This can be adjusted according to your needs by providing a custom
Attributes.LogLevels
attribute on the given Flow:Uses implicit
MarkerLoggingAdapter
if available, otherwise uses an internally created one, which usesakka.event.Logging.withMarker(materializer.system, materializer)
as its source (use this class to configure slf4j loggers).Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the mapping function returns an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
logWithMarker$default$3
scala.Function1<Out,java.lang.Object> logWithMarker$default$3()
-
logWithMarker$default$4
MarkerLoggingAdapter logWithMarker$default$4(java.lang.String name, scala.Function1<Out,LogMarker> marker, scala.Function1<Out,java.lang.Object> extract)
-
map
<T> FlowOps map(scala.Function1<Out,T> f)
Transform this stream by applying the given function to each of the elements as they pass through this processing step.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the mapping function returns an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
mapAsync
<T> FlowOps mapAsync(int parallelism, scala.Function1<Out,scala.concurrent.Future<T>> f)
Transform this stream by applying the given function to each of the elements as they pass through this processing step. The function returns aFuture
and the value of that future will be emitted downstream. The number of Futures that shall run in parallel is given as the first argument tomapAsync
. These Futures may complete in any order, but the elements that are emitted downstream are in the same order as received from upstream.
If the function
f
throws an exception or if theFuture
is completed with failure and the supervision decision isakka.stream.Supervision.Stop
the stream will be completed with failure.If the function
f
throws an exception or if theFuture
is completed with failure and the supervision decision isakka.stream.Supervision.Resume
orakka.stream.Supervision.Restart
the element is dropped and the stream continues.If the
Future
is completed withnull
, it is ignored and the next element is processed.The function
f
is always invoked on the elements in the order they arrive.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the Future returned by the provided function finishes for the next element in sequence
'''Backpressures when''' the number of futures reaches the configured parallelism and the downstream backpressures or the first future is not completed
'''Completes when''' upstream completes and all futures have been completed and all elements have been emitted
'''Cancels when''' downstream cancels
-
mapAsyncUnordered
<T> FlowOps mapAsyncUnordered(int parallelism, scala.Function1<Out,scala.concurrent.Future<T>> f)
Transform this stream by applying the given function to each of the elements as they pass through this processing step. The function returns aFuture
and the value of that future will be emitted downstream. The number of Futures that shall run in parallel is given as the first argument tomapAsyncUnordered
. Each processed element will be emitted downstream as soon as it is ready, i.e. it is possible that the elements are not emitted downstream in the same order as received from upstream.
If the function
f
throws an exception or if theFuture
is completed with failure and the supervision decision isakka.stream.Supervision.Stop
the stream will be completed with failure.If the function
f
throws an exception or if theFuture
is completed with failure and the supervision decision isakka.stream.Supervision.Resume
orakka.stream.Supervision.Restart
the element is dropped and the stream continues.If the
Future
is completed withnull
, it is ignored and the next element is processed.The function
f
is always invoked on the elements in the order they arrive (even though the result of the futures returned byf
might be emitted in a different order).Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' any of the Futures returned by the provided function complete
'''Backpressures when''' the number of futures reaches the configured parallelism and the downstream backpressures
'''Completes when''' upstream completes and all futures have been completed and all elements have been emitted
'''Cancels when''' downstream cancels
-
mapConcat
<T> FlowOps mapConcat(scala.Function1<Out,scala.collection.IterableOnce<T>> f)
Transform each input element into anIterable
of output elements that is then flattened into the output stream.The returned
Iterable
MUST NOT containnull
values, as they are illegal as stream elements - according to the Reactive Streams specification.'''Emits when''' the mapping function returns an element or there are still remaining elements from the previously calculated collection
'''Backpressures when''' downstream backpressures or there are still remaining elements from the previously calculated collection
'''Completes when''' upstream completes and all remaining elements have been emitted
'''Cancels when''' downstream cancels
-
mapError
FlowOps mapError(scala.PartialFunction<java.lang.Throwable,java.lang.Throwable> pf)
While similar to<T>recover(scala.PartialFunction<java.lang.Throwable,T>)
this operator can be used to transform an error signal to a different one *without* logging it as an error in the process. So in that sense it is NOT exactly equivalent torecover(t => throw t2)
since recover would log thet2
error.Since the underlying failure signal onError arrives out-of-band, it might jump over existing elements. This operator can recover the failure signal, but not the skipped elements, which will be dropped.
Similarly to
<T>recover(scala.PartialFunction<java.lang.Throwable,T>)
throwing an exception insidemapError
_will_ be logged.'''Emits when''' element is available from the upstream or upstream is failed and pf returns an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or upstream failed with exception pf can handle
'''Cancels when''' downstream cancels
-
mapWithResource
<R,T> FlowOps mapWithResource(scala.Function0<R> create, scala.Function2<R,Out,T> f, scala.Function1<R,scala.Option<T>> close)
Transform each stream element with the help of a resource.The resource creation function is invoked once when the stream is materialized and the returned resource is passed to the mapping function for mapping the first element. The mapping function returns a mapped element to emit downstream. The returned
T
MUST NOT benull
as it is illegal as stream element - according to the Reactive Streams specification.The
close
function is called only once when the upstream or downstream finishes or fails. You can do some clean-up here, and if the returned value is not empty, it will be emitted to the downstream if available, otherwise the value will be dropped.Early completion can be done with combination of the
takeWhile(scala.Function1<Out,java.lang.Object>)
operator.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.You can configure the default dispatcher for this Source by changing the
akka.stream.materializer.blocking-io-dispatcher
or set it for a given Source by usingActorAttributes
.'''Emits when''' the mapping function returns an element and downstream is ready to consume it
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
- Parameters:
create
- function that creates the resourcef
- function that transforms the upstream element and the resource to output elementclose
- function that closes the resource, optionally outputting a last element
-
merge
<U,M> FlowOps merge(Graph<SourceShape<U>,M> that, boolean eagerComplete)
Merge the givenSource
to thisFlow
, taking elements as they arrive from input streams, picking randomly when several elements ready.'''Emits when''' one of the inputs has an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' all upstreams complete (eagerComplete=false) or one upstream completes (eagerComplete=true), default value is
false
'''Cancels when''' downstream cancels
-
merge$default$2
<U,M> boolean merge$default$2()
-
mergeAll
<U> FlowOps mergeAll(scala.collection.immutable.Seq<Graph<SourceShape<U>,?>> those, boolean eagerComplete)
Merge the givenSource
s to thisFlow
, taking elements as they arrive from input streams, picking randomly when several elements ready.'''Emits when''' one of the inputs has an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' all upstreams complete (eagerComplete=false) or one upstream completes (eagerComplete=true), default value is
false
'''Cancels when''' downstream cancels
-
mergeGraph
<U,M> Graph<FlowShape<Out,U>,M> mergeGraph(Graph<SourceShape<U>,M> that, boolean eagerComplete)
-
mergeLatest
<U,M> FlowOps mergeLatest(Graph<SourceShape<U>,M> that, boolean eagerComplete)
MergeLatest joins elements from N input streams into stream of lists of size N. i-th element in list is the latest emitted element from i-th input stream. MergeLatest emits list for each element emitted from some input stream, but only after each input stream emitted at least one element.'''Emits when''' an element is available from some input and each input emits at least one element from stream start
'''Completes when''' all upstreams complete (eagerClose=false) or one upstream completes (eagerClose=true)
-
mergeLatest$default$2
<U,M> boolean mergeLatest$default$2()
-
mergeLatestGraph
<U,M> Graph<FlowShape<Out,scala.collection.immutable.Seq<U>>,M> mergeLatestGraph(Graph<SourceShape<U>,M> that, boolean eagerComplete)
-
mergePreferred
<U,M> FlowOps mergePreferred(Graph<SourceShape<U>,M> that, boolean preferred, boolean eagerComplete)
Merge two sources. Prefer one source if both sources have elements ready.'''emits''' when one of the inputs has an element available. If multiple have elements available, prefer the 'right' one when 'preferred' is 'true', or the 'left' one when 'preferred' is 'false'.
'''backpressures''' when downstream backpressures
'''completes''' when all upstreams complete (This behavior is changeable to completing when any upstream completes by setting
eagerComplete=true
.)
-
mergePreferred$default$3
<U,M> boolean mergePreferred$default$3()
-
mergePreferredGraph
<U,M> Graph<FlowShape<Out,U>,M> mergePreferredGraph(Graph<SourceShape<U>,M> that, boolean preferred, boolean eagerComplete)
-
mergePrioritized
<U,M> FlowOps mergePrioritized(Graph<SourceShape<U>,M> that, int leftPriority, int rightPriority, boolean eagerComplete)
Merge two sources. Prefer the sources depending on the 'priority' parameters.'''emits''' when one of the inputs has an element available, preferring inputs based on the 'priority' parameters if both have elements available
'''backpressures''' when downstream backpressures
'''completes''' when both upstreams complete (This behavior is changeable to completing when any upstream completes by setting
eagerComplete=true
.)
-
mergePrioritized$default$4
<U,M> boolean mergePrioritized$default$4()
-
mergePrioritizedGraph
<U,M> Graph<FlowShape<Out,U>,M> mergePrioritizedGraph(Graph<SourceShape<U>,M> that, int leftPriority, int rightPriority, boolean eagerComplete)
-
mergeSorted
<U,M> FlowOps mergeSorted(Graph<SourceShape<U>,M> that, scala.math.Ordering<U> ord)
Merge the givenSource
to thisFlow
, taking elements as they arrive from input streams, picking always the smallest of the available elements (waiting for one element from each side to be available). This means that possible contiguity of the input streams is not exploited to avoid waiting for elements, this merge will block when one of the inputs does not have more elements (and does not complete).'''Emits when''' all of the inputs have an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' all upstreams complete
'''Cancels when''' downstream cancels
-
mergeSortedGraph
<U,M> Graph<FlowShape<Out,U>,M> mergeSortedGraph(Graph<SourceShape<U>,M> that, scala.math.Ordering<U> ord)
-
named
FlowOps named(java.lang.String name)
-
orElse
<U,Mat2> FlowOps orElse(Graph<SourceShape<U>,Mat2> secondary)
Provides a secondary source that will be consumed if this stream completes without any elements passing by. As soon as the first element comes through this stream, the alternative will be cancelled.Note that this Flow will be materialized together with the
Source
and just kept from producing elements by asserting back-pressure until its time comes or it gets cancelled.On errors the operator is failed regardless of source of the error.
'''Emits when''' element is available from first stream or first stream closed without emitting any elements and an element is available from the second stream
'''Backpressures when''' downstream backpressures
'''Completes when''' the primary stream completes after emitting at least one element, when the primary stream completes without emitting and the secondary stream already has completed or when the secondary stream completes
'''Cancels when''' downstream cancels and additionally the alternative is cancelled as soon as an element passes by from this stream.
-
orElseGraph
<U,Mat2> Graph<FlowShape<Out,U>,Mat2> orElseGraph(Graph<SourceShape<U>,Mat2> secondary)
-
prefixAndTail
<U> FlowOps prefixAndTail(int n)
Takes up ton
elements from the stream (less thann
only if the upstream completes before emittingn
elements) and returns a pair containing a strict sequence of the taken element and a stream representing the remaining elements. If ''n'' is zero or negative, then this will return a pair of an empty collection and a stream containing the whole upstream unchanged.In case of an upstream error, depending on the current state - the master stream signals the error if less than
n
elements has been seen, and therefore the substream has not yet been emitted - the tail substream signals the error after the prefix and tail has been emitted by the main stream (at that point the main stream has already completed)'''Emits when''' the configured number of prefix elements are available. Emits this prefix, and the rest as a substream
'''Backpressures when''' downstream backpressures or substream backpressures
'''Completes when''' prefix elements have been consumed and substream has been consumed
'''Cancels when''' downstream cancels or substream cancels
-
prepend
<U,Mat2> FlowOps prepend(Graph<SourceShape<U>,Mat2> that)
Prepend the givenSource
to thisFlow
, meaning that before elements are generated from this Flow, the Source's elements will be produced until it is exhausted, at which point Flow elements will start being produced.Note that the
Source
is materialized together with this Flow and is "detached" meaning in effect behave as a one element buffer in front of both the sources, that eagerly demands an element on start (so it can not be combined withSource.lazy
to defer materialization ofthat
).This flow will then be kept from producing elements by asserting back-pressure until its time comes.
When needing a prepend operator that is not detached use
prependLazy(akka.stream.Graph<akka.stream.SourceShape<U>, Mat2>)
'''Emits when''' element is available from the given
Source
or from current stream when theSource
is completed'''Backpressures when''' downstream backpressures
'''Completes when''' this
Flow
completes'''Cancels when''' downstream cancels
-
prependGraph
<U,Mat2> Graph<FlowShape<Out,U>,Mat2> prependGraph(Graph<SourceShape<U>,Mat2> that, boolean detached)
-
prependLazy
<U,Mat2> FlowOps prependLazy(Graph<SourceShape<U>,Mat2> that)
Prepend the givenSource
to thisFlow
, meaning that before elements are generated from this Flow, the Source's elements will be produced until it is exhausted, at which point Flow elements will start being produced.Note that the
Source
is materialized together with this Flow and will then be kept from producing elements by asserting back-pressure until its time comes.When needing a prepend operator that is also detached use
prepend(akka.stream.Graph<akka.stream.SourceShape<U>, Mat2>)
If the given
Source
gets upstream error - no elements from thisFlow
will be pulled.'''Emits when''' element is available from the given
Source
or from current stream when theSource
is completed'''Backpressures when''' downstream backpressures
'''Completes when''' this
Flow
completes'''Cancels when''' downstream cancels
-
recover
<T> FlowOps recover(scala.PartialFunction<java.lang.Throwable,T> pf)
Recover allows to send last element on failure and gracefully complete the stream Since the underlying failure signal onError arrives out-of-band, it might jump over existing elements. This operator can recover the failure signal, but not the skipped elements, which will be dropped.Throwing an exception inside
recover
_will_ be logged on ERROR level automatically.'''Emits when''' element is available from the upstream or upstream is failed and pf returns an element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or upstream failed with exception pf can handle
'''Cancels when''' downstream cancels
-
recoverWith
<T> FlowOps recoverWith(scala.PartialFunction<java.lang.Throwable,Graph<SourceShape<T>,NotUsed>> pf)
Deprecated.Use recoverWithRetries instead. Since 2.4.4.RecoverWith allows to switch to alternative Source on flow failure. It will stay in effect after a failure has been recovered so that each time there is a failure it is fed into thepf
and a new Source may be materialized.Since the underlying failure signal onError arrives out-of-band, it might jump over existing elements. This operator can recover the failure signal, but not the skipped elements, which will be dropped.
Throwing an exception inside
recoverWith
_will_ be logged on ERROR level automatically.'''Emits when''' element is available from the upstream or upstream is failed and element is available from alternative Source
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or upstream failed with exception pf can handle
'''Cancels when''' downstream cancels
-
recoverWithRetries
<T> FlowOps recoverWithRetries(int attempts, scala.PartialFunction<java.lang.Throwable,Graph<SourceShape<T>,NotUsed>> pf)
RecoverWithRetries allows to switch to alternative Source on flow failure. It will stay in effect after a failure has been recovered up toattempts
number of times so that each time there is a failure it is fed into thepf
and a new Source may be materialized. Note that if you pass in 0, this won't attempt to recover at all.A negative
attempts
number is interpreted as "infinite", which results in the exact same behavior asrecoverWith
.Since the underlying failure signal onError arrives out-of-band, it might jump over existing elements. This operator can recover the failure signal, but not the skipped elements, which will be dropped.
Throwing an exception inside
recoverWithRetries
_will_ be logged on ERROR level automatically.'''Emits when''' element is available from the upstream or upstream is failed and element is available from alternative Source
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or upstream failed with exception pf can handle
'''Cancels when''' downstream cancels
- Parameters:
attempts
- Maximum number of retries or -1 to retry indefinitelypf
- Receives the failure cause and returns the new Source to be materialized if any
-
reduce
<T> FlowOps reduce(scala.Function2<T,T,T> f)
Similar tofold
but uses first element as zero element. Applies the given function towards its current and next value, yielding the next current value.If the stream is empty (i.e. completes before signalling any elements), the reduce operator will fail its downstream with a
NoSuchElementException
, which is semantically in-line with that Scala's standard library collections do in such situations.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' upstream completes
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
See also
FlowOps.fold
-
scan
<T> FlowOps scan(T zero, scala.Function2<T,Out,T> f)
Similar tofold
but is not a terminal operation, emits its current value which starts atzero
and then applies the current and next value to the given functionf
, emitting the next current value.If the function
f
throws an exception and the supervision decision isakka.stream.Supervision.Restart
current value starts atzero
again the stream will continue.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.Note that the
zero
value must be immutable.'''Emits when''' the function scanning the element returns a new element
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
See also
FlowOps.scanAsync
-
scanAsync
<T> FlowOps scanAsync(T zero, scala.Function2<T,Out,scala.concurrent.Future<T>> f)
Similar toscan
but with an asynchronous function, emits its current value which starts atzero
and then applies the current and next value to the given functionf
, emitting aFuture
that resolves to the next current value.If the function
f
throws an exception and the supervision decision isakka.stream.Supervision.Restart
current value starts atzero
again the stream will continue.If the function
f
throws an exception and the supervision decision isakka.stream.Supervision.Resume
current value starts at the previous current value, or zero when it doesn't have one, and the stream will continue.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.Note that the
zero
value must be immutable.'''Emits when''' the future returned by
f
completes'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes and the last future returned by
f
completes'''Cancels when''' downstream cancels
See also
FlowOps.scan
-
sliding
FlowOps sliding(int n, int step)
Apply a sliding window over the stream and return the windows as groups of elements, with the last group possibly smaller than requested due to end-of-stream.n
must be positive, otherwise IllegalArgumentException is thrown.step
must be positive, otherwise IllegalArgumentException is thrown.'''Emits when''' enough elements have been collected within the window or upstream completed
'''Backpressures when''' a window has been assembled and downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
sliding$default$2
int sliding$default$2()
-
splitAfter
SubFlow<Out,Mat,FlowOps,java.lang.Object> splitAfter(SubstreamCancelStrategy substreamCancelStrategy, scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams. It *ends* the current substream when the predicate is true. This means that for the following series of predicate values, three substreams will be produced with lengths 2, 2, and 3:false, true, // elements go into first substream false, true, // elements go into second substream false, false, true // elements go into third substream
The object returned from this method is not a normal
Source
orFlow
, it is aSubFlow
. This means that after this operator all transformations are applied to all encountered substreams in the same fashion. Substream mode is exited either by closing the substream (i.e. connecting it to aSink
) or by merging the substreams back together; see theto
andmergeBack
methods onSubFlow
for more information.It is important to note that the substreams also propagate back-pressure as any other stream, which means that blocking one substream will block the
splitAfter
operator itself—and thereby all substreams—once all internal or explicit buffers are filled.If the split predicate
p
throws an exception and the supervision decision isakka.stream.Supervision.Stop
the stream and substreams will be completed with failure.If the split predicate
p
throws an exception and the supervision decision isakka.stream.Supervision.Resume
orakka.stream.Supervision.Restart
the element is dropped and the stream and substreams continue.'''Emits when''' an element passes through. When the provided predicate is true it emits the element and opens a new substream for subsequent element
'''Backpressures when''' there is an element pending for the next substream, but the previous is not fully consumed yet, or the substream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels and substreams cancel on
SubstreamCancelStrategy.drain
, downstream cancels or any substream cancels onSubstreamCancelStrategy.propagate
See also
FlowOps.splitWhen
.
-
splitAfter
SubFlow<Out,Mat,FlowOps,java.lang.Object> splitAfter(scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams. It *ends* the current substream when the predicate is true.
-
splitWhen
SubFlow<Out,Mat,FlowOps,java.lang.Object> splitWhen(SubstreamCancelStrategy substreamCancelStrategy, scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams, always beginning a new one with the current element if the given predicate returns true for it. This means that for the following series of predicate values, three substreams will be produced with lengths 1, 2, and 3:false, // element goes into first substream true, false, // elements go into second substream true, false, false // elements go into third substream
In case the *first* element of the stream matches the predicate, the first substream emitted by splitWhen will start from that element. For example:
true, false, false // first substream starts from the split-by element true, false // subsequent substreams operate the same way
The object returned from this method is not a normal
Source
orFlow
, it is aSubFlow
. This means that after this operator all transformations are applied to all encountered substreams in the same fashion. Substream mode is exited either by closing the substream (i.e. connecting it to aSink
) or by merging the substreams back together; see theto
andmergeBack
methods onSubFlow
for more information.It is important to note that the substreams also propagate back-pressure as any other stream, which means that blocking one substream will block the
splitWhen
operator itself—and thereby all substreams—once all internal or explicit buffers are filled.If the split predicate
p
throws an exception and the supervision decision isakka.stream.Supervision.Stop
the stream and substreams will be completed with failure.If the split predicate
p
throws an exception and the supervision decision isakka.stream.Supervision.Resume
orakka.stream.Supervision.Restart
the element is dropped and the stream and substreams continue.'''Emits when''' an element for which the provided predicate is true, opening and emitting a new substream for subsequent element
'''Backpressures when''' there is an element pending for the next substream, but the previous is not fully consumed yet, or the substream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels and substreams cancel on
SubstreamCancelStrategy.drain
, downstream cancels or any substream cancels onSubstreamCancelStrategy.propagate
See also
FlowOps.splitAfter
.
-
splitWhen
SubFlow<Out,Mat,FlowOps,java.lang.Object> splitWhen(scala.Function1<Out,java.lang.Object> p)
This operation applies the given predicate to all incoming elements and emits them to a stream of output streams, always beginning a new one with the current element if the given predicate returns true for it.
-
statefulMap
<S,T> FlowOps statefulMap(scala.Function0<S> create, scala.Function2<S,Out,scala.Tuple2<S,T>> f, scala.Function1<S,scala.Option<T>> onComplete)
Transform each stream element with the help of a state.The state creation function is invoked once when the stream is materialized and the returned state is passed to the mapping function for mapping the first element. The mapping function returns a mapped element to emit downstream and a state to pass to the next mapping function. The state can be the same for each mapping return, be a new immutable state but it is also safe to use a mutable state. The returned
T
MUST NOT benull
as it is illegal as stream element - according to the Reactive Streams specification. Anull
state is not allowed and will fail the stream.For stateless variant see
FlowOps.map
.The
onComplete
function is called only once when the upstream or downstream finished, You can do some clean-up here, and if the returned value is not empty, it will be emitted to the downstream if available, otherwise the value will be dropped.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the mapping function returns an element and downstream is ready to consume it
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
- Parameters:
create
- a function that creates the initial statef
- a function that transforms the upstream element and the state into a pair of next state and output elementonComplete
- a function that transforms the ongoing state into an optional output element
-
statefulMapConcat
<T> FlowOps statefulMapConcat(scala.Function0<scala.Function1<Out,scala.collection.IterableOnce<T>>> f)
Transform each input element into anIterable
of output elements that is then flattened into the output stream. The transformation is meant to be stateful, which is enabled by creating the transformation function anew for every materialization — the returned function will typically close over mutable objects to store state between invocations. For the stateless variant seeFlowOps.mapConcat
.The returned
Iterable
MUST NOT containnull
values, as they are illegal as stream elements - according to the Reactive Streams specification.Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the mapping function returns an element or there are still remaining elements from the previously calculated collection
'''Backpressures when''' downstream backpressures or there are still remaining elements from the previously calculated collection
'''Completes when''' upstream completes and all remaining elements has been emitted
'''Cancels when''' downstream cancels
See also
FlowOps.mapConcat
-
take
FlowOps take(long n)
Terminate processing (and cancel the upstream publisher) after the given number of elements. Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.The stream will be completed without producing any elements if
n
is zero or negative.'''Emits when''' the specified number of elements to take has not yet been reached
'''Backpressures when''' downstream backpressures
'''Completes when''' the defined number of elements has been taken or upstream completes
'''Cancels when''' the defined number of elements has been taken or downstream cancels
See also
FlowOps.limit
,FlowOps.limitWeighted
-
takeWhile
FlowOps takeWhile(scala.Function1<Out,java.lang.Object> p)
Terminate processing (and cancel the upstream publisher) after predicate returns false for the first time, Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.The stream will be completed without producing any elements if predicate is false for the first stream element.
'''Emits when''' the predicate is true
'''Backpressures when''' downstream backpressures
'''Completes when''' predicate returned false (or 1 after predicate returns false if
inclusive
or upstream completes'''Cancels when''' predicate returned false or downstream cancels
See also
FlowOps.limit
,FlowOps.limitWeighted
-
takeWhile
FlowOps takeWhile(scala.Function1<Out,java.lang.Object> p, boolean inclusive)
Terminate processing (and cancel the upstream publisher) after predicate returns false for the first time, including the first failed element iff inclusive is true Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.The stream will be completed without producing any elements if predicate is false for the first stream element.
Adheres to the
ActorAttributes.SupervisionStrategy
attribute.'''Emits when''' the predicate is true
'''Backpressures when''' downstream backpressures
'''Completes when''' predicate returned false (or 1 after predicate returns false if
inclusive
or upstream completes'''Cancels when''' predicate returned false or downstream cancels
See also
FlowOps.limit
,FlowOps.limitWeighted
-
takeWithin
FlowOps takeWithin(scala.concurrent.duration.FiniteDuration d)
Terminate processing (and cancel the upstream publisher) after the given duration. Due to input buffering some elements may have been requested from upstream publishers that will then not be processed downstream of this step.Note that this can be combined with
take(long)
to limit the number of elements within the duration.'''Emits when''' an upstream element arrives
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes or timer fires
'''Cancels when''' downstream cancels or timer fires
-
throttle
FlowOps throttle(int elements, scala.concurrent.duration.FiniteDuration per)
Sends elements downstream with speed limited toelements/per
. In other words, this operator set the maximum rate for emitting messages. This operator works for streams where all elements have the same cost or length.Throttle implements the token bucket model. There is a bucket with a given token capacity (burst size). Tokens drops into the bucket at a given rate and can be
spared
for later use up to bucket capacity to allow some burstiness. Whenever stream wants to send an element, it takes as many tokens from the bucket as element costs. If there isn't any, throttle waits until the bucket accumulates enough tokens. Elements that costs more than the allowed burst will be delayed proportionally to their cost minus available tokens, meeting the target rate. Bucket is full when stream just materialized and started.The burst size is calculated based on the given rate (
cost/per
) as 0.1 * rate, for example: - rate < 20/second => burst size 1 - rate 20/second => burst size 2 - rate 100/second => burst size 10 - rate 200/second => burst size 20The throttle
mode
isakka.stream.ThrottleMode.Shaping
, which makes pauses before emitting messages to meet throttle rate.'''Emits when''' upstream emits an element and configured time per each element elapsed
'''Backpressures when''' downstream backpressures or the incoming rate is higher than the speed limit
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
throttle
FlowOps throttle(int elements, scala.concurrent.duration.FiniteDuration per, int maximumBurst, ThrottleMode mode)
Sends elements downstream with speed limited toelements/per
. In other words, this operator set the maximum rate for emitting messages. This operator works for streams where all elements have the same cost or length.Throttle implements the token bucket model. There is a bucket with a given token capacity (burst size or maximumBurst). Tokens drops into the bucket at a given rate and can be
spared
for later use up to bucket capacity to allow some burstiness. Whenever stream wants to send an element, it takes as many tokens from the bucket as element costs. If there isn't any, throttle waits until the bucket accumulates enough tokens. Elements that costs more than the allowed burst will be delayed proportionally to their cost minus available tokens, meeting the target rate. Bucket is full when stream just materialized and started.Parameter
mode
manages behavior when upstream is faster than throttle rate: -akka.stream.ThrottleMode.Shaping
makes pauses before emitting messages to meet throttle rate -akka.stream.ThrottleMode.Enforcing
fails with exception when upstream is faster than throttle rate. Enforcing cannot emit elements that cost more than the maximumBurstIt is recommended to use non-zero burst sizes as they improve both performance and throttling precision by allowing the implementation to avoid using the scheduler when input rates fall below the enforced limit and to reduce most of the inaccuracy caused by the scheduler resolution (which is in the range of milliseconds).
WARNING: Be aware that throttle is using scheduler to slow down the stream. This scheduler has minimal time of triggering next push. Consequently it will slow down the stream as it has minimal pause for emitting. This can happen in case burst is 0 and speed is higher than 30 events per second. You need to increase the
maximumBurst
if elements arrive with small interval (30 milliseconds or less). Use the overloadedthrottle
method withoutmaximumBurst
parameter to automatically calculate themaximumBurst
based on the given rate (cost/per
). In other words the throttler always enforces the rate limit whenmaximumBurst
parameter is given, but in certain cases (mostly due to limited scheduler resolution) it enforces a tighter bound than what was prescribed.'''Emits when''' upstream emits an element and configured time per each element elapsed
'''Backpressures when''' downstream backpressures or the incoming rate is higher than the speed limit
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
throttle
FlowOps throttle(int cost, scala.concurrent.duration.FiniteDuration per, scala.Function1<Out,java.lang.Object> costCalculation)
Sends elements downstream with speed limited tocost/per
. Cost is calculating for each element individually by callingcalculateCost
function. This operator works for streams when elements have different cost(length). Streams ofByteString
for example.Throttle implements the token bucket model. There is a bucket with a given token capacity (burst size). Tokens drops into the bucket at a given rate and can be
spared
for later use up to bucket capacity to allow some burstiness. Whenever stream wants to send an element, it takes as many tokens from the bucket as element costs. If there isn't any, throttle waits until the bucket accumulates enough tokens. Elements that costs more than the allowed burst will be delayed proportionally to their cost minus available tokens, meeting the target rate. Bucket is full when stream just materialized and started.The burst size is calculated based on the given rate (
cost/per
) as 0.1 * rate, for example: - rate < 20/second => burst size 1 - rate 20/second => burst size 2 - rate 100/second => burst size 10 - rate 200/second => burst size 20The throttle
mode
isakka.stream.ThrottleMode.Shaping
, which makes pauses before emitting messages to meet throttle rate.'''Emits when''' upstream emits an element and configured time per each element elapsed
'''Backpressures when''' downstream backpressures or the incoming rate is higher than the speed limit
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
throttle
FlowOps throttle(int cost, scala.concurrent.duration.FiniteDuration per, int maximumBurst, scala.Function1<Out,java.lang.Object> costCalculation, ThrottleMode mode)
Sends elements downstream with speed limited tocost/per
. Cost is calculating for each element individually by callingcalculateCost
function. This operator works for streams when elements have different cost(length). Streams ofByteString
for example.Throttle implements the token bucket model. There is a bucket with a given token capacity (burst size or maximumBurst). Tokens drops into the bucket at a given rate and can be
spared
for later use up to bucket capacity to allow some burstiness. Whenever stream wants to send an element, it takes as many tokens from the bucket as element costs. If there isn't any, throttle waits until the bucket accumulates enough tokens. Elements that costs more than the allowed burst will be delayed proportionally to their cost minus available tokens, meeting the target rate. Bucket is full when stream just materialized and started.Parameter
mode
manages behavior when upstream is faster than throttle rate: -akka.stream.ThrottleMode.Shaping
makes pauses before emitting messages to meet throttle rate -akka.stream.ThrottleMode.Enforcing
fails with exception when upstream is faster than throttle rate. Enforcing cannot emit elements that cost more than the maximumBurstIt is recommended to use non-zero burst sizes as they improve both performance and throttling precision by allowing the implementation to avoid using the scheduler when input rates fall below the enforced limit and to reduce most of the inaccuracy caused by the scheduler resolution (which is in the range of milliseconds).
WARNING: Be aware that throttle is using scheduler to slow down the stream. This scheduler has minimal time of triggering next push. Consequently it will slow down the stream as it has minimal pause for emitting. This can happen in case burst is 0 and speed is higher than 30 events per second. You need to increase the
maximumBurst
if elements arrive with small interval (30 milliseconds or less). Use the overloadedthrottle
method withoutmaximumBurst
parameter to automatically calculate themaximumBurst
based on the given rate (cost/per
). In other words the throttler always enforces the rate limit whenmaximumBurst
parameter is given, but in certain cases (mostly due to limited scheduler resolution) it enforces a tighter bound than what was prescribed.'''Emits when''' upstream emits an element and configured time per each element elapsed
'''Backpressures when''' downstream backpressures or the incoming rate is higher than the speed limit
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
throttleEven
FlowOps throttleEven(int elements, scala.concurrent.duration.FiniteDuration per, ThrottleMode mode)
Deprecated.Use throttle without `maximumBurst` parameter instead. Since 2.5.12.This is a simplified version of throttle that spreads events evenly across the given time interval. throttleEven using best effort approach to meet throttle rate.Use this operator when you need just slow down a stream without worrying about exact amount of time between events.
If you want to be sure that no time interval has no more than specified number of events you need to use
throttle(int,scala.concurrent.duration.FiniteDuration)
with maximumBurst attribute.
-
throttleEven
FlowOps throttleEven(int cost, scala.concurrent.duration.FiniteDuration per, scala.Function1<Out,java.lang.Object> costCalculation, ThrottleMode mode)
Deprecated.Use throttle without `maximumBurst` parameter instead. Since 2.5.12.This is a simplified version of throttle that spreads events evenly across the given time interval.Use this operator when you need just slow down a stream without worrying about exact amount of time between events.
If you want to be sure that no time interval has no more than specified number of events you need to use
throttle(int,scala.concurrent.duration.FiniteDuration)
with maximumBurst attribute.
-
to
<Mat2> java.lang.Object to(Graph<SinkShape<Out>,Mat2> sink)
Connect thisFlow
to aSink
, concatenating the processing steps of both.+------------------------------+ | Resulting Sink[In, Mat] | | | | +------+ +------+ | | | | | | | In ~~> | flow | ~~Out~~> | sink | | | | Mat| | M| | | +------+ +------+ | +------------------------------+
The materialized value of the combined
Sink
will be the materialized value of the current flow (ignoring the given Sink’s value), use {@link Flow#toMat[Mat2* toMat} if a different strategy is needed.See also
FlowOpsMat.toMat
when access to materialized values of the parameter is needed.
-
watch
FlowOps watch(ActorRef ref)
The operator fails with anWatchedActorTerminatedException
if the target actor is terminated.'''Emits when''' upstream emits
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Fails when''' the watched actor terminates
'''Cancels when''' downstream cancels
-
wireTap
FlowOps wireTap(scala.Function1<Out,scala.runtime.BoxedUnit> f)
This is a simplified version ofwireTap(Sink)
that takes only a simple function. Elements will be passed into this "side channel" function, and any of its results will be ignored.If the wire-tap operation is slow (it backpressures), elements that would've been sent to it will be dropped instead. It is similar to
alsoTo(akka.stream.Graph<akka.stream.SinkShape<Out>, ?>)
which does backpressure instead of dropping elements.This operation is useful for inspecting the passed through element, usually by means of side-effecting operations (such as
println
, or emitting metrics), for each element without having to modify it.For logging signals (elements, completion, error) consider using the
log(java.lang.String,scala.Function1<Out,java.lang.Object>,akka.event.LoggingAdapter)
operator instead, along with appropriateActorAttributes.logLevels
.'''Emits when''' upstream emits an element; the same element will be passed to the attached function, as well as to the downstream operator
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
wireTap
FlowOps wireTap(Graph<SinkShape<Out>,?> that)
Attaches the givenSink
to thisFlow
as a wire tap, meaning that elements that pass through will also be sent to the wire-tap Sink, without the latter affecting the mainline flow. If the wire-tap Sink backpressures, elements that would've been sent to it will be dropped instead.It is similar to
alsoTo(akka.stream.Graph<akka.stream.SinkShape<Out>, ?>)
which does backpressure instead of dropping elements.'''Emits when''' element is available and demand exists from the downstream; the element will also be sent to the wire-tap Sink if there is demand.
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
withAttributes
FlowOps withAttributes(Attributes attr)
-
zip
<U> FlowOps zip(Graph<SourceShape<U>,?> that)
Combine the elements of current flow and the givenSource
into a stream of tuples.'''Emits when''' all of the inputs have an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' any upstream completes
'''Cancels when''' downstream cancels
-
zipAll
<U,A> FlowOps zipAll(Graph<SourceShape<U>,?> that, A thisElem, U thatElem)
Combine the elements of current flow and the givenSource
into a stream of tuples.'''Emits when''' at first emits when both inputs emit, and then as long as any input emits (coupled to the default value of the completed input).
'''Backpressures when''' downstream backpressures
'''Completes when''' all upstream completes
'''Cancels when''' downstream cancels
-
zipAllFlow
<U,A,Mat2> Flow<Out,scala.Tuple2<A,U>,Mat2> zipAllFlow(Graph<SourceShape<U>,Mat2> that, A thisElem, U thatElem)
-
zipGraph
<U,M> Graph<FlowShape<Out,scala.Tuple2<Out,U>>,M> zipGraph(Graph<SourceShape<U>,M> that)
-
zipLatest
<U> FlowOps zipLatest(Graph<SourceShape<U>,?> that)
Combine the elements of 2 streams into a stream of tuples, picking always the latest element of each.A
ZipLatest
has aleft
and aright
input port and oneout
port.No element is emitted until at least one element from each Source becomes available.
'''Emits when''' all of the inputs have at least an element available, and then each time an element becomes available on either of the inputs
'''Backpressures when''' downstream backpressures
'''Completes when''' any upstream completes
'''Cancels when''' downstream cancels
-
zipLatestGraph
<U,M> Graph<FlowShape<Out,scala.Tuple2<Out,U>>,M> zipLatestGraph(Graph<SourceShape<U>,M> that)
-
zipLatestWith
<Out2,Out3> FlowOps zipLatestWith(Graph<SourceShape<Out2>,?> that, scala.Function2<Out,Out2,Out3> combine)
Combine the elements of multiple streams into a stream of combined elements using a combiner function, picking always the latest of the elements of each source.No element is emitted until at least one element from each Source becomes available. Whenever a new element appears, the zipping function is invoked with a tuple containing the new element and the other last seen elements.
'''Emits when''' all of the inputs have at least an element available, and then each time an element becomes available on either of the inputs
'''Backpressures when''' downstream backpressures
'''Completes when''' any of the upstreams completes
'''Cancels when''' downstream cancels
-
zipLatestWith
<Out2,Out3> FlowOps zipLatestWith(Graph<SourceShape<Out2>,?> that, boolean eagerComplete, scala.Function2<Out,Out2,Out3> combine)
Combine the elements of multiple streams into a stream of combined elements using a combiner function, picking always the latest of the elements of each source.No element is emitted until at least one element from each Source becomes available. Whenever a new element appears, the zipping function is invoked with a tuple containing the new element and the other last seen elements.
'''Emits when''' all of the inputs have at least an element available, and then each time an element becomes available on either of the inputs
'''Backpressures when''' downstream backpressures
'''Completes when''' any upstream completes if
eagerComplete
is enabled or wait for all upstreams to complete'''Cancels when''' downstream cancels
-
zipLatestWithGraph
<Out2,Out3,M> Graph<FlowShape<Out,Out3>,M> zipLatestWithGraph(Graph<SourceShape<Out2>,M> that, scala.Function2<Out,Out2,Out3> combine)
-
zipLatestWithGraph
<Out2,Out3,M> Graph<FlowShape<Out,Out3>,M> zipLatestWithGraph(Graph<SourceShape<Out2>,M> that, boolean eagerComplete, scala.Function2<Out,Out2,Out3> combine)
-
zipWith
<Out2,Out3> FlowOps zipWith(Graph<SourceShape<Out2>,?> that, scala.Function2<Out,Out2,Out3> combine)
Put together the elements of current flow and the givenSource
into a stream of combined elements using a combiner function.'''Emits when''' all of the inputs have an element available
'''Backpressures when''' downstream backpressures
'''Completes when''' any upstream completes
'''Cancels when''' downstream cancels
-
zipWithGraph
<Out2,Out3,M> Graph<FlowShape<Out,Out3>,M> zipWithGraph(Graph<SourceShape<Out2>,M> that, scala.Function2<Out,Out2,Out3> combine)
-
zipWithIndex
FlowOps zipWithIndex()
Combine the elements of current flow into a stream of tuples consisting of all elements paired with their index. Indices start at 0.'''Emits when''' upstream emits an element and is paired with their index
'''Backpressures when''' downstream backpressures
'''Completes when''' upstream completes
'''Cancels when''' downstream cancels
-
-