sink

Documentation

Index

• type Sink
• type TODO
  • func (_ *TODO) Begin(size int)
  • func (_ *TODO) CancellationRequested() bool
  • func (_ *TODO) End()

Types

type Sink

type Sink interface {
	consumer.Consumer

	/**
	 * Resets the sink state to receive a fresh data set.  This must be called
	 * before sending any data to the sink.  After calling {@link #end()},
	 * you may call this method to reset the sink for another calculation.
	 * @param size The exact size of the data to be pushed Downstream, if
	 * known or {@code -1} if unknown or infinite.
	 *
	 * <p>Prior to this call, the sink must be in the initial state, and after
	 * this call it is in the active state.
	 */
	Begin(size int)

	/**
	 * Indicates that all elements have been pushed.  If the {@code Sink} is
	 * stateful, it should send any stored state Downstream at this time, and
	 * should clear any accumulated state (and associated resources).
	 *
	 * <p>Prior to this call, the sink must be in the active state, and after
	 * this call it is returned to the initial state.
	 */
	End()

	/**
	 * Indicates that this {@code Sink} does not wish to receive any more data.
	 *
	 * @implSpec The default implementation always returns false.
	 *
	 * @return true if cancellation is requested
	 */
	CancellationRequested() bool
}

*

• An extension of {@link Consumer} used to conduct values through the stages of
• a stream pipeline, with additional methods to manage size information,
• control flow, etc. Before calling the {@code accept()} method on a
• {@code Sink} for the first time, you must first call the {@code begin()}
• method to inform it that data is coming (optionally informing the sink how
• much data is coming), and after all data has been sent, you must call the
• {@code end()} method. After calling {@code end()}, you should not call
• {@code accept()} without again calling {@code begin()}. {@code Sink} also
• offers a mechanism by which the sink can cooperatively signal that it does
• not wish to receive any more data (the {@code cancellationRequested()}
• method), which a source can poll before sending more data to the
• {@code Sink}. *
• <p>A sink may be in one of two states: an initial state and an active state.
• It starts out in the initial state; the {@code begin()} method transitions
• it to the active state, and the {@code end()} method transitions it back into
• the initial state, where it can be re-used. Data-accepting methods (such as
• {@code accept()} are only valid in the active state. *
• @apiNote
• A stream pipeline consists of a source, zero or more intermediate stages
• (such as filtering or mapping), and a terminal stage, such as reduction or
• for-each. For concreteness, consider the pipeline: *
• <pre>{@code
• int longestStringLengthStartingWithA
• = strings.stream()
• .filter(s -> s.startsWith("A"))
• .mapToInt(String::length)
• .max();
• }</pre> *
• <p>Here, we have three stages, filtering, mapping, and reducing. The
• filtering stage consumes strings and emits a subset of those strings; the
• mapping stage consumes strings and emits ints; the reduction stage consumes
• those ints and computes the maximal value. *
• <p>A {@code Sink} instance is used to represent each stage of this pipeline,
• whether the stage accepts objects, ints, longs, or doubles. Sink has entry
• points for {@code accept(Object)}, {@code accept(int)}, etc, so that we do
• not need a specialized interface for each primitive specialization. (It
• might be called a "kitchen sink" for this omnivorous tendency.) The entry
• point to the pipeline is the {@code Sink} for the filtering stage, which
• sends some elements "Downstream" -- into the {@code Sink} for the mapping
• stage, which in turn sends integral values Downstream into the {@code Sink}
• for the reduction stage. The {@code Sink} implementations associated with a
• given stage is expected to know the data type for the next stage, and call
• the correct {@code accept} method on its Downstream {@code Sink}. Similarly,
• each stage must implement the correct {@code accept} method corresponding to
• the data type it accepts. *
• <p>The specialized subtypes such as {@link Sink.OfInt} override
• {@code accept(Object)} to call the appropriate primitive specialization of
• {@code accept}, implement the appropriate primitive specialization of
• {@code Consumer}, and re-abstract the appropriate primitive specialization of
• {@code accept}. *
• <p>The chaining subtypes such as {@link ChainedInt} not only implement
• {@code Sink.OfInt}, but also maintain a {@code Downstream} field which
• represents the Downstream {@code Sink}, and implement the methods
• {@code begin()}, {@code end()}, and {@code cancellationRequested()} to
• delegate to the Downstream {@code Sink}. Most implementations of
• intermediate operations will use these chaining wrappers. For example, the
• mapping stage in the above example would look like: *
• <pre>{@code
• IntSink is = new Sink.AbstractChainedReferenceSink<U>(sink) {
• public void accept(U u) {
• Downstream.accept(mapper.applyAsInt(u));
• }
• };
• }</pre> *
• <p>Here, we implement {@code Sink.AbstractChainedReferenceSink<U>}, meaning that we expect
• to receive elements of type {@code U} as input, and pass the Downstream sink
• to the constructor. Because the next stage expects to receive integers, we
• must call the {@code accept(int)} method when emitting values to the Downstream.
• The {@code accept()} method applies the mapping function from {@code U} to
• {@code int} and passes the resulting value to the Downstream {@code Sink}. *
• @param <T> type of elements for value streams
• @since 1.8

type TODO

type TODO struct {
	consumer.TODO
}

func (*TODO) Begin

func (_ *TODO) Begin(size int)

func (*TODO) CancellationRequested

func (_ *TODO) CancellationRequested() bool

func (*TODO) End

func (_ *TODO) End()