main work by @drewhk with contributions from @2m and @rkuhn
This work uncovered many well-hidden bugs in existing Stages, in
particular StatefulStage. These were hidden by the behavior of
OneBoundedInterpreter that normally behaves more orderly than it
guarantees in general, especially with respect to the timeliness of
delivery of upstream termination signals; the bugs were then that
internal state was not flushed when onComplete arrived “too early”.
* also make factories more consistent by only offering
FlowGraph.create()
* also remove secondary (edge-based) FlowGraph.Builder DSL
* also improve naming for conversions from Graph to
Source/Flow/BidiFlow/Sink
- explicitly provide Unit values and place parens around tuple creation
- remove structural type usage in TestUtils
- fix Java double-casts
- use unused Java values by asserting their non-nullness
- work around inability to place case class in trait (scripted test)
The remaining warnings about using private types in public methods are
bogus as reported in https://issues.scala-lang.org/browse/SI-9490.
* Gives Inlets and Outlets a `carbonCopy` method and switches to allocate them via `apply`
* Removes 4 Array allocations per FanIn and uses a bitmasked array instead
* Makes the FlattenStrategy.concat instance a singleton
- ActorProcessor terminated eagerly when ConcatAll had just taken up a
new input stream but not yet received onSubscribe for it
- The ActorProcessor eagerly shuts itself down upon onError and that
cannot be changed without completely reworking the Pump, so I opted
for just tracking the outstanding substreamSubscribers that have not
yet seen OnSubscribe and making them cancel properly when that arrives
(possibly later).
* found one bug of too early actor stop and thereby missing cancel of upstream,
in fan-in tests ""work with one immediately failed and one nonempty publisher"
These are synchronous implementations, because we need to be Java 6
compatible while developing on 2.3.x. However asynchronous
implementations using AsynchronousFileChannel will come soon for JDK7
users.
+ ActorPublisher/Subscriber now manage stopping of the actor
+ added documentation on configuring dispatcher for File IO
+ properly handle if source file does not exist
+ file sink / source come with default io dispatcher
> verified no actors are leaking
> exceptions are caught and onErrored properly
+ moved files to akka.stream.io
+ Added OutputStreamSink and InputStreamSource
* Remove optional attributes parameter in favor of withAttributes
* Overload explosion when trying to add attributes to ZipWith.
* Aligned with Flow and Source.
* Remove name parameter (no overloads), naming is performed using `.withAttributes` or the new
convenience `.named`. Those adds the OperationAttribute.Name and also change the name of the
shape Inlet and Outlet.
* Remove Source/Sink parameter list for 0 parameter methods,
this allows usage of `Sink.head` instead of `Sink.head()`
Changed rules:
* 1.9 Always onSubscribe prior to any other signals
* 1.9 NullPointerException if subscriber is null
* 3.17 Long overflow, effectively unbounded instead of onError
Fixed some more things:
* fixed some FIXME
* Long drop and take
* memory leaks in tck tests, use BeforeClass to create ActorSystem
use AfterClass to shutdown ActorSystem
* more tck tests
* don't emit OnComplete when substream is cancelled
* work around for memory leak in PrefixAndTail
* add supervion for stages and built in ops run by interpreter
* add supervision for mapAsync and mapAsyncUnordered
* add supervision to groupBy and splitWhen
* reference doc for scala and java
* FlowMaterializer is now the actor independent interface
* ActorFlowMaterializer is the actor based interface
* MaterializerSettings renamed to ActorFlowMaterializerSettings
* impl.ActorBasedFlowMaterializer renamed to impl.ActorFlowMaterializerImpl
* Optimizations included in ActorFlowMaterializerSettings
* Note that http is using FlowMaterializer in api, but I suspect that it
will currently only run with a ActorFlowMaterializer
* StreamTcp and Http extensions now return Flows and Sources that can be materialized later
* Flow can now be completed with another flow to be turned into a runnable flow
- Switches from using size-of-1/2 Vectors to using Lists
- Fixes an issue where processorForNode wouldn't use the dispatcher form the settings
- Adds a dedicated Collect fusion op
- Adds various simplifications to ActorBasedFlowMaterializer
- Adds FIXMEs where appropriate
- Switches `grouped` to use a VectorBuilder
- Adds support for `scan`
- ActorBasedFlowMaterializer now uses Iterator instead of head+tail decomp on Seqs
- Identity and Completed Transformers are now cached
- Adds dedicated AstNodes for all combinators
- Adds a hook-in point for fusion in `ActorBasedFlowMaterializer`
- Adds support for `Operate` an AstNode with a function that create a fusing.Op
- Adds experimental and slow optimizer as a PoC
- Adds verification that Ast.Fused does not exist in optimizer input
* Fold Tap/Drain into Source/Sink
* Create Source/Sink helpers to create all Sources/Sinks
* Make concrete Source/Sink implementations private[scaladsl2]
* runWith(drainWithKey) and runWith(tapWithKey) returning the
materialized drain/tap
* remove toPublisher, toSubscriber, and friends, in favor of using
runWith
* MaterializedMap is the return type for both flow and graph
* Source[T] return type for Source.apply methods
* Cleaned up type signatures
* Changed names of Source/Sink/ProcessorFlow to Tap/Drain/Pipe
* Introduced new Source/Sink/Flow
* Made all Pipe types package private
* Changed the TypeGraph DSL to work on Flow/Sink/Source
