談談stream的執行原理

等你歸去來 發表於 2021-06-12

  害,別誤會,我這裡說的stream不是流式程式設計,不是大資料處理框架。我這裡說的是stream指的是jdk中的一個開發工具包stream. 該工具包在jdk8中出現,可以說已經是冷飯了,為何還要你說?只因各家一言,不算得自家理解,如若有空,何多聽一版又何妨。

  本篇主要從幾個方面講講:1. 我們常見的stream都有哪些? 2. stream包有哪些好處? 3. stream包的實現原理? 相信這些多少會解開大家的一些迷惑。

 

1. 我們常見的stream都有哪些? 

  stream直接翻譯為流。何謂流?我們最常見的,比如網路中的資料傳輸,即tcp/udp那一套東西,都是建立在二進位制流的基礎上的。用流來形容這些資料或檔案的傳輸,非常形象,因為資料總是源源不斷地從一端流向另一端,這是不流是什麼。只是,傳輸到另一端之後,我們再做解析,便有了資料或檔案之說。其實這說的,便是高層協議了。

  另說一個stream, 那就是jdk中的各種InputStream了,它用於讀取檔案資料,讀取byte資料,其實也是源源不斷將資料從一個裝置流入到另一裝置。jdk中有InputStream/OutputStream, 作為根基,其上則是各種 FileInputStream, FileOutputStream, FileReader, FileWriter,... 實際上,整個io包幾乎都是在圍繞流這個概念來展開的。可見,io是相當的重要啊。

  再說一stream, 則是對大資料的處理了,stream,即是實時資料處理的重要技術實現,因與實時二字吻合,恰好又類似於資料從一裝置流入另一裝置,且是實時的。所以,stream在大資料領域也是大放異彩啊!

  還有更多的流概念,更多的流實現,不必細說,也無法細說。單隻知道,流無處不在,非常重要。

  還有本文要議的stream包,到底是何生物,且看後續說來。

 

2. stream包有何好處?

  stream包,在java中是以一個工具包的形式存在,即你用則以,不用亦可。

  那麼,用它到底有何好處?好處主要有二: 1.可以減少冗餘程式碼的編寫;比如要寫一個過濾器則只需呼叫一filter()傳入處理邏輯即可; 2.可以很方便的利用一些隱藏的升級好處或者多核帶來的好處;(當然你可能用不上這些好處)

  說實話,這兩個功能,看起來實際沒有太多的誘惑力,但凡我們封裝幾個方法,供外部呼叫,不也可以達到同等效果?是了!如果你有這等造詣,能夠抽象出足夠通用的方法,供各方使用,那你不算大牛何人算?說到底,stream也就是高手封裝的工具包而已。

  來幾個應用例項,看看stream都如何使用的:

public class StreamUtilTest {

    @Test
    public void testArrayStream() {
        // 1. 過濾值;改變值;排序;
        Integer[] intArr = {1, 2, 3, 5, 22, 8, 5};
        List<Integer> iArrList = Arrays.stream(intArr)
                                .filter(r -> r < 20)
                                .map(r -> r + 1)
                                .sorted().collect(Collectors.toList());
        System.out.println("result:" + iArrList);

        String[] strArr = {"a1,a2", "q,y,h", "ddd,bb,n", null};
        // 2. 過濾陣列;拆分值;輸出;
        Arrays.stream(strArr).filter(Objects::nonNull)
                .flatMap(r -> Arrays.stream(r.split(",")))
                .forEach(System.out::println);
    }

    @Test
    public void testListStream() {
        List<String> list = new ArrayList<>();
        list.add("ab");
        list.add("ccc");
        list.add("ddd");
        // 3. 求list中的最大值
        Optional<String> maxStr = list.stream().max(Comparator.naturalOrder());
        System.out.println(maxStr);
    }
}

  害,不必糾結裡面乾的事情復不復雜,有沒有意義,只知道有這用法即可。 反正就當你會這麼用,即能解決這般問題。這也是我們高階語言使用必備技能,學會呼叫api.

  不過需要說明的,java中有一句老話,叫做萬事萬物皆物件。 但見上面的寫法,自然不太像物件。是了,這是lamda語法,雖說另一主題,但何妨在此處一題。但既然說到這,不妨來想想這lamda到底是何物?從某種角度來說,它可以看作是一種內部類,不過寫法不太一樣。但是當我們仔細觀察class檔案的變化情況時,發現它與內部類又不太一致,因為java的內部類會在class中生成$xx.class的類檔案,而lamda表示式卻不會。但是不管怎麼樣,它是可以使用內部類的表達方式獲得同樣的效果,只需將該類代入到其中,即可達到同樣的效果。

  但要細說lamda表示式,則可以反編譯下class檔案,可以見些許端倪。

# 呼叫lamda表示式示例
...
        59: invokestatic  #4                  // Method java/util/Arrays.stream:([Ljava/lang/Object;)Ljava/util/stream/Stream;
        62: invokedynamic #5,  0              // InvokeDynamic #0:test:()Ljava/util/function/Predicate;
        67: invokeinterface #6,  2            // InterfaceMethod java/util/stream/Stream.filter:(Ljava/util/function/Predicate;)Ljava/ut
il/stream/Stream;
        72: invokedynamic #7,  0              // InvokeDynamic #1:apply:()Ljava/util/function/Function;
        77: invokeinterface #8,  2            // InterfaceMethod java/util/stream/Stream.map:(Ljava/util/function/Function;)Ljava/util/s
tream/Stream;
...
# 常量池定義,實際是定義了lamda的實現方式為 #0 號方法
    #5 = InvokeDynamic      #0:#102       // #0:test:()Ljava/util/function/Predicate;
# lamda表示式的具體實現1示例
BootstrapMethods:
  0: #98 invokestatic java/lang/invoke/LambdaMetafactory.metafactory:(Ljava/lang/invoke/MethodHandles$Lookup;Ljava/lang/String;Ljava/lan
g/invoke/MethodType;Ljava/lang/invoke/MethodType;Ljava/lang/invoke/MethodHandle;Ljava/lang/invoke/MethodType;)Ljava/lang/invoke/CallSite
;
    Method arguments:
      #99 (Ljava/lang/Object;)Z
      // 此處為呼叫具體的實現方法
      #100 invokestatic com/my/test/common/util/StreamUtilTest.lambda$testArrayStream$0:(Ljava/lang/Integer;)Z
      #101 (Ljava/lang/Integer;)Z
  1: #98 invokestatic java/lang/invoke/LambdaMetafactory.metafactory:(Ljava/lang/invoke/MethodHandles$Lookup;Ljava/lang/String;Ljava/lang/invoke/MethodType;Ljava/lang/invoke/MethodType;Ljava/lang/invoke/MethodHandle;Ljava/lang/invoke/MethodType;)Ljava/lang/invoke/CallSite;
    Method arguments:
      #105 (Ljava/lang/Object;)Ljava/lang/Object;
      #106 invokestatic com/my/test/common/util/StreamUtilTest.lambda$testArrayStream$1:(Ljava/lang/Integer;)Ljava/lang/Integer;
      #107 (Ljava/lang/Integer;)Ljava/lang/Integer;

# lamda表示式具體實現2, 上一步的靜態呼叫
  private static boolean lambda$testArrayStream$0(java.lang.Integer);
    descriptor: (Ljava/lang/Integer;)Z
    flags: ACC_PRIVATE, ACC_STATIC, ACC_SYNTHETIC
    Code:
      stack=2, locals=1, args_size=1
         0: aload_0
         1: invokevirtual #48                 // Method java/lang/Integer.intValue:()I
         4: bipush        20
         6: if_icmpge     13
         9: iconst_1
        10: goto          14
        13: iconst_0
        14: ireturn
      LineNumberTable:
        line 16: 0
      LocalVariableTable:
        Start  Length  Slot  Name   Signature
            0      15     0     r   Ljava/lang/Integer;
      StackMapTable: number_of_entries = 2
        frame_type = 13 /* same */
        frame_type = 64 /* same_locals_1_stack_item */
          stack = [ int ]
    MethodParameters:
      Name                           Flags
      r                              synthetic

  害,往深了就不說了。單說這lamda表示式,並非使用內部類來實現的,而是使用內部靜態函式來做的,所以也叫函數語言程式設計呢。煩話休提。

  最後,再來看看,這stream包究竟有何神聖地方?其實,就是一個以一個 Stream 介面定義為核心展開的,且看如下:

談談stream的執行原理
/**
 * A sequence of elements supporting sequential and parallel aggregate
 * operations.  The following example illustrates an aggregate operation using
 * {@link Stream} and {@link IntStream}:
 *
 * <pre>{@code
 *     int sum = widgets.stream()
 *                      .filter(w -> w.getColor() == RED)
 *                      .mapToInt(w -> w.getWeight())
 *                      .sum();
 * }</pre>
 *
 * In this example, {@code widgets} is a {@code Collection<Widget>}.  We create
 * a stream of {@code Widget} objects via {@link Collection#stream Collection.stream()},
 * filter it to produce a stream containing only the red widgets, and then
 * transform it into a stream of {@code int} values representing the weight of
 * each red widget. Then this stream is summed to produce a total weight.
 *
 * <p>In addition to {@code Stream}, which is a stream of object references,
 * there are primitive specializations for {@link IntStream}, {@link LongStream},
 * and {@link DoubleStream}, all of which are referred to as "streams" and
 * conform to the characteristics and restrictions described here.
 *
 * <p>To perform a computation, stream
 * <a href="package-summary.html#StreamOps">operations</a> are composed into a
 * <em>stream pipeline</em>.  A stream pipeline consists of a source (which
 * might be an array, a collection, a generator function, an I/O channel,
 * etc), zero or more <em>intermediate operations</em> (which transform a
 * stream into another stream, such as {@link Stream#filter(Predicate)}), and a
 * <em>terminal operation</em> (which produces a result or side-effect, such
 * as {@link Stream#count()} or {@link Stream#forEach(Consumer)}).
 * Streams are lazy; computation on the source data is only performed when the
 * terminal operation is initiated, and source elements are consumed only
 * as needed.
 *
 * <p>Collections and streams, while bearing some superficial similarities,
 * have different goals.  Collections are primarily concerned with the efficient
 * management of, and access to, their elements.  By contrast, streams do not
 * provide a means to directly access or manipulate their elements, and are
 * instead concerned with declaratively describing their source and the
 * computational operations which will be performed in aggregate on that source.
 * However, if the provided stream operations do not offer the desired
 * functionality, the {@link #iterator()} and {@link #spliterator()} operations
 * can be used to perform a controlled traversal.
 *
 * <p>A stream pipeline, like the "widgets" example above, can be viewed as
 * a <em>query</em> on the stream source.  Unless the source was explicitly
 * designed for concurrent modification (such as a {@link ConcurrentHashMap}),
 * unpredictable or erroneous behavior may result from modifying the stream
 * source while it is being queried.
 *
 * <p>Most stream operations accept parameters that describe user-specified
 * behavior, such as the lambda expression {@code w -> w.getWeight()} passed to
 * {@code mapToInt} in the example above.  To preserve correct behavior,
 * these <em>behavioral parameters</em>:
 * <ul>
 * <li>must be <a href="package-summary.html#NonInterference">non-interfering</a>
 * (they do not modify the stream source); and</li>
 * <li>in most cases must be <a href="package-summary.html#Statelessness">stateless</a>
 * (their result should not depend on any state that might change during execution
 * of the stream pipeline).</li>
 * </ul>
 *
 * <p>Such parameters are always instances of a
 * <a href="../function/package-summary.html">functional interface</a> such
 * as {@link java.util.function.Function}, and are often lambda expressions or
 * method references.  Unless otherwise specified these parameters must be
 * <em>non-null</em>.
 *
 * <p>A stream should be operated on (invoking an intermediate or terminal stream
 * operation) only once.  This rules out, for example, "forked" streams, where
 * the same source feeds two or more pipelines, or multiple traversals of the
 * same stream.  A stream implementation may throw {@link IllegalStateException}
 * if it detects that the stream is being reused. However, since some stream
 * operations may return their receiver rather than a new stream object, it may
 * not be possible to detect reuse in all cases.
 *
 * <p>Streams have a {@link #close()} method and implement {@link AutoCloseable},
 * but nearly all stream instances do not actually need to be closed after use.
 * Generally, only streams whose source is an IO channel (such as those returned
 * by {@link Files#lines(Path, Charset)}) will require closing.  Most streams
 * are backed by collections, arrays, or generating functions, which require no
 * special resource management.  (If a stream does require closing, it can be
 * declared as a resource in a {@code try}-with-resources statement.)
 *
 * <p>Stream pipelines may execute either sequentially or in
 * <a href="package-summary.html#Parallelism">parallel</a>.  This
 * execution mode is a property of the stream.  Streams are created
 * with an initial choice of sequential or parallel execution.  (For example,
 * {@link Collection#stream() Collection.stream()} creates a sequential stream,
 * and {@link Collection#parallelStream() Collection.parallelStream()} creates
 * a parallel one.)  This choice of execution mode may be modified by the
 * {@link #sequential()} or {@link #parallel()} methods, and may be queried with
 * the {@link #isParallel()} method.
 *
 * @param <T> the type of the stream elements
 * @since 1.8
 * @see IntStream
 * @see LongStream
 * @see DoubleStream
 * @see <a href="package-summary.html">java.util.stream</a>
 */
public interface Stream<T> extends BaseStream<T, Stream<T>> {

    /**
     * Returns a stream consisting of the elements of this stream that match
     * the given predicate.
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param predicate a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                  <a href="package-summary.html#Statelessness">stateless</a>
     *                  predicate to apply to each element to determine if it
     *                  should be included
     * @return the new stream
     */
    Stream<T> filter(Predicate<? super T> predicate);

    /**
     * Returns a stream consisting of the results of applying the given
     * function to the elements of this stream.
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param <R> The element type of the new stream
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element
     * @return the new stream
     */
    <R> Stream<R> map(Function<? super T, ? extends R> mapper);

    /**
     * Returns an {@code IntStream} consisting of the results of applying the
     * given function to the elements of this stream.
     *
     * <p>This is an <a href="package-summary.html#StreamOps">
     *     intermediate operation</a>.
     *
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element
     * @return the new stream
     */
    IntStream mapToInt(ToIntFunction<? super T> mapper);

    /**
     * Returns a {@code LongStream} consisting of the results of applying the
     * given function to the elements of this stream.
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element
     * @return the new stream
     */
    LongStream mapToLong(ToLongFunction<? super T> mapper);

    /**
     * Returns a {@code DoubleStream} consisting of the results of applying the
     * given function to the elements of this stream.
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element
     * @return the new stream
     */
    DoubleStream mapToDouble(ToDoubleFunction<? super T> mapper);

    /**
     * Returns a stream consisting of the results of replacing each element of
     * this stream with the contents of a mapped stream produced by applying
     * the provided mapping function to each element.  Each mapped stream is
     * {@link java.util.stream.BaseStream#close() closed} after its contents
     * have been placed into this stream.  (If a mapped stream is {@code null}
     * an empty stream is used, instead.)
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @apiNote
     * The {@code flatMap()} operation has the effect of applying a one-to-many
     * transformation to the elements of the stream, and then flattening the
     * resulting elements into a new stream.
     *
     * <p><b>Examples.</b>
     *
     * <p>If {@code orders} is a stream of purchase orders, and each purchase
     * order contains a collection of line items, then the following produces a
     * stream containing all the line items in all the orders:
     * <pre>{@code
     *     orders.flatMap(order -> order.getLineItems().stream())...
     * }</pre>
     *
     * <p>If {@code path} is the path to a file, then the following produces a
     * stream of the {@code words} contained in that file:
     * <pre>{@code
     *     Stream<String> lines = Files.lines(path, StandardCharsets.UTF_8);
     *     Stream<String> words = lines.flatMap(line -> Stream.of(line.split(" +")));
     * }</pre>
     * The {@code mapper} function passed to {@code flatMap} splits a line,
     * using a simple regular expression, into an array of words, and then
     * creates a stream of words from that array.
     *
     * @param <R> The element type of the new stream
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element which produces a stream
     *               of new values
     * @return the new stream
     */
    <R> Stream<R> flatMap(Function<? super T, ? extends Stream<? extends R>> mapper);

    /**
     * Returns an {@code IntStream} consisting of the results of replacing each
     * element of this stream with the contents of a mapped stream produced by
     * applying the provided mapping function to each element.  Each mapped
     * stream is {@link java.util.stream.BaseStream#close() closed} after its
     * contents have been placed into this stream.  (If a mapped stream is
     * {@code null} an empty stream is used, instead.)
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element which produces a stream
     *               of new values
     * @return the new stream
     * @see #flatMap(Function)
     */
    IntStream flatMapToInt(Function<? super T, ? extends IntStream> mapper);

    /**
     * Returns an {@code LongStream} consisting of the results of replacing each
     * element of this stream with the contents of a mapped stream produced by
     * applying the provided mapping function to each element.  Each mapped
     * stream is {@link java.util.stream.BaseStream#close() closed} after its
     * contents have been placed into this stream.  (If a mapped stream is
     * {@code null} an empty stream is used, instead.)
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element which produces a stream
     *               of new values
     * @return the new stream
     * @see #flatMap(Function)
     */
    LongStream flatMapToLong(Function<? super T, ? extends LongStream> mapper);

    /**
     * Returns an {@code DoubleStream} consisting of the results of replacing
     * each element of this stream with the contents of a mapped stream produced
     * by applying the provided mapping function to each element.  Each mapped
     * stream is {@link java.util.stream.BaseStream#close() closed} after its
     * contents have placed been into this stream.  (If a mapped stream is
     * {@code null} an empty stream is used, instead.)
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * @param mapper a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *               <a href="package-summary.html#Statelessness">stateless</a>
     *               function to apply to each element which produces a stream
     *               of new values
     * @return the new stream
     * @see #flatMap(Function)
     */
    DoubleStream flatMapToDouble(Function<? super T, ? extends DoubleStream> mapper);

    /**
     * Returns a stream consisting of the distinct elements (according to
     * {@link Object#equals(Object)}) of this stream.
     *
     * <p>For ordered streams, the selection of distinct elements is stable
     * (for duplicated elements, the element appearing first in the encounter
     * order is preserved.)  For unordered streams, no stability guarantees
     * are made.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">stateful
     * intermediate operation</a>.
     *
     * @apiNote
     * Preserving stability for {@code distinct()} in parallel pipelines is
     * relatively expensive (requires that the operation act as a full barrier,
     * with substantial buffering overhead), and stability is often not needed.
     * Using an unordered stream source (such as {@link #generate(Supplier)})
     * or removing the ordering constraint with {@link #unordered()} may result
     * in significantly more efficient execution for {@code distinct()} in parallel
     * pipelines, if the semantics of your situation permit.  If consistency
     * with encounter order is required, and you are experiencing poor performance
     * or memory utilization with {@code distinct()} in parallel pipelines,
     * switching to sequential execution with {@link #sequential()} may improve
     * performance.
     *
     * @return the new stream
     */
    Stream<T> distinct();

    /**
     * Returns a stream consisting of the elements of this stream, sorted
     * according to natural order.  If the elements of this stream are not
     * {@code Comparable}, a {@code java.lang.ClassCastException} may be thrown
     * when the terminal operation is executed.
     *
     * <p>For ordered streams, the sort is stable.  For unordered streams, no
     * stability guarantees are made.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">stateful
     * intermediate operation</a>.
     *
     * @return the new stream
     */
    Stream<T> sorted();

    /**
     * Returns a stream consisting of the elements of this stream, sorted
     * according to the provided {@code Comparator}.
     *
     * <p>For ordered streams, the sort is stable.  For unordered streams, no
     * stability guarantees are made.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">stateful
     * intermediate operation</a>.
     *
     * @param comparator a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                   <a href="package-summary.html#Statelessness">stateless</a>
     *                   {@code Comparator} to be used to compare stream elements
     * @return the new stream
     */
    Stream<T> sorted(Comparator<? super T> comparator);

    /**
     * Returns a stream consisting of the elements of this stream, additionally
     * performing the provided action on each element as elements are consumed
     * from the resulting stream.
     *
     * <p>This is an <a href="package-summary.html#StreamOps">intermediate
     * operation</a>.
     *
     * <p>For parallel stream pipelines, the action may be called at
     * whatever time and in whatever thread the element is made available by the
     * upstream operation.  If the action modifies shared state,
     * it is responsible for providing the required synchronization.
     *
     * @apiNote This method exists mainly to support debugging, where you want
     * to see the elements as they flow past a certain point in a pipeline:
     * <pre>{@code
     *     Stream.of("one", "two", "three", "four")
     *         .filter(e -> e.length() > 3)
     *         .peek(e -> System.out.println("Filtered value: " + e))
     *         .map(String::toUpperCase)
     *         .peek(e -> System.out.println("Mapped value: " + e))
     *         .collect(Collectors.toList());
     * }</pre>
     *
     * @param action a <a href="package-summary.html#NonInterference">
     *                 non-interfering</a> action to perform on the elements as
     *                 they are consumed from the stream
     * @return the new stream
     */
    Stream<T> peek(Consumer<? super T> action);

    /**
     * Returns a stream consisting of the elements of this stream, truncated
     * to be no longer than {@code maxSize} in length.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">short-circuiting
     * stateful intermediate operation</a>.
     *
     * @apiNote
     * While {@code limit()} is generally a cheap operation on sequential
     * stream pipelines, it can be quite expensive on ordered parallel pipelines,
     * especially for large values of {@code maxSize}, since {@code limit(n)}
     * is constrained to return not just any <em>n</em> elements, but the
     * <em>first n</em> elements in the encounter order.  Using an unordered
     * stream source (such as {@link #generate(Supplier)}) or removing the
     * ordering constraint with {@link #unordered()} may result in significant
     * speedups of {@code limit()} in parallel pipelines, if the semantics of
     * your situation permit.  If consistency with encounter order is required,
     * and you are experiencing poor performance or memory utilization with
     * {@code limit()} in parallel pipelines, switching to sequential execution
     * with {@link #sequential()} may improve performance.
     *
     * @param maxSize the number of elements the stream should be limited to
     * @return the new stream
     * @throws IllegalArgumentException if {@code maxSize} is negative
     */
    Stream<T> limit(long maxSize);

    /**
     * Returns a stream consisting of the remaining elements of this stream
     * after discarding the first {@code n} elements of the stream.
     * If this stream contains fewer than {@code n} elements then an
     * empty stream will be returned.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">stateful
     * intermediate operation</a>.
     *
     * @apiNote
     * While {@code skip()} is generally a cheap operation on sequential
     * stream pipelines, it can be quite expensive on ordered parallel pipelines,
     * especially for large values of {@code n}, since {@code skip(n)}
     * is constrained to skip not just any <em>n</em> elements, but the
     * <em>first n</em> elements in the encounter order.  Using an unordered
     * stream source (such as {@link #generate(Supplier)}) or removing the
     * ordering constraint with {@link #unordered()} may result in significant
     * speedups of {@code skip()} in parallel pipelines, if the semantics of
     * your situation permit.  If consistency with encounter order is required,
     * and you are experiencing poor performance or memory utilization with
     * {@code skip()} in parallel pipelines, switching to sequential execution
     * with {@link #sequential()} may improve performance.
     *
     * @param n the number of leading elements to skip
     * @return the new stream
     * @throws IllegalArgumentException if {@code n} is negative
     */
    Stream<T> skip(long n);

    /**
     * Performs an action for each element of this stream.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * <p>The behavior of this operation is explicitly nondeterministic.
     * For parallel stream pipelines, this operation does <em>not</em>
     * guarantee to respect the encounter order of the stream, as doing so
     * would sacrifice the benefit of parallelism.  For any given element, the
     * action may be performed at whatever time and in whatever thread the
     * library chooses.  If the action accesses shared state, it is
     * responsible for providing the required synchronization.
     *
     * @param action a <a href="package-summary.html#NonInterference">
     *               non-interfering</a> action to perform on the elements
     */
    void forEach(Consumer<? super T> action);

    /**
     * Performs an action for each element of this stream, in the encounter
     * order of the stream if the stream has a defined encounter order.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * <p>This operation processes the elements one at a time, in encounter
     * order if one exists.  Performing the action for one element
     * <a href="../concurrent/package-summary.html#MemoryVisibility"><i>happens-before</i></a>
     * performing the action for subsequent elements, but for any given element,
     * the action may be performed in whatever thread the library chooses.
     *
     * @param action a <a href="package-summary.html#NonInterference">
     *               non-interfering</a> action to perform on the elements
     * @see #forEach(Consumer)
     */
    void forEachOrdered(Consumer<? super T> action);

    /**
     * Returns an array containing the elements of this stream.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @return an array containing the elements of this stream
     */
    Object[] toArray();

    /**
     * Returns an array containing the elements of this stream, using the
     * provided {@code generator} function to allocate the returned array, as
     * well as any additional arrays that might be required for a partitioned
     * execution or for resizing.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @apiNote
     * The generator function takes an integer, which is the size of the
     * desired array, and produces an array of the desired size.  This can be
     * concisely expressed with an array constructor reference:
     * <pre>{@code
     *     Person[] men = people.stream()
     *                          .filter(p -> p.getGender() == MALE)
     *                          .toArray(Person[]::new);
     * }</pre>
     *
     * @param <A> the element type of the resulting array
     * @param generator a function which produces a new array of the desired
     *                  type and the provided length
     * @return an array containing the elements in this stream
     * @throws ArrayStoreException if the runtime type of the array returned
     * from the array generator is not a supertype of the runtime type of every
     * element in this stream
     */
    <A> A[] toArray(IntFunction<A[]> generator);

    /**
     * Performs a <a href="package-summary.html#Reduction">reduction</a> on the
     * elements of this stream, using the provided identity value and an
     * <a href="package-summary.html#Associativity">associative</a>
     * accumulation function, and returns the reduced value.  This is equivalent
     * to:
     * <pre>{@code
     *     T result = identity;
     *     for (T element : this stream)
     *         result = accumulator.apply(result, element)
     *     return result;
     * }</pre>
     *
     * but is not constrained to execute sequentially.
     *
     * <p>The {@code identity} value must be an identity for the accumulator
     * function. This means that for all {@code t},
     * {@code accumulator.apply(identity, t)} is equal to {@code t}.
     * The {@code accumulator} function must be an
     * <a href="package-summary.html#Associativity">associative</a> function.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @apiNote Sum, min, max, average, and string concatenation are all special
     * cases of reduction. Summing a stream of numbers can be expressed as:
     *
     * <pre>{@code
     *     Integer sum = integers.reduce(0, (a, b) -> a+b);
     * }</pre>
     *
     * or:
     *
     * <pre>{@code
     *     Integer sum = integers.reduce(0, Integer::sum);
     * }</pre>
     *
     * <p>While this may seem a more roundabout way to perform an aggregation
     * compared to simply mutating a running total in a loop, reduction
     * operations parallelize more gracefully, without needing additional
     * synchronization and with greatly reduced risk of data races.
     *
     * @param identity the identity value for the accumulating function
     * @param accumulator an <a href="package-summary.html#Associativity">associative</a>,
     *                    <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                    <a href="package-summary.html#Statelessness">stateless</a>
     *                    function for combining two values
     * @return the result of the reduction
     */
    T reduce(T identity, BinaryOperator<T> accumulator);

    /**
     * Performs a <a href="package-summary.html#Reduction">reduction</a> on the
     * elements of this stream, using an
     * <a href="package-summary.html#Associativity">associative</a> accumulation
     * function, and returns an {@code Optional} describing the reduced value,
     * if any. This is equivalent to:
     * <pre>{@code
     *     boolean foundAny = false;
     *     T result = null;
     *     for (T element : this stream) {
     *         if (!foundAny) {
     *             foundAny = true;
     *             result = element;
     *         }
     *         else
     *             result = accumulator.apply(result, element);
     *     }
     *     return foundAny ? Optional.of(result) : Optional.empty();
     * }</pre>
     *
     * but is not constrained to execute sequentially.
     *
     * <p>The {@code accumulator} function must be an
     * <a href="package-summary.html#Associativity">associative</a> function.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @param accumulator an <a href="package-summary.html#Associativity">associative</a>,
     *                    <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                    <a href="package-summary.html#Statelessness">stateless</a>
     *                    function for combining two values
     * @return an {@link Optional} describing the result of the reduction
     * @throws NullPointerException if the result of the reduction is null
     * @see #reduce(Object, BinaryOperator)
     * @see #min(Comparator)
     * @see #max(Comparator)
     */
    Optional<T> reduce(BinaryOperator<T> accumulator);

    /**
     * Performs a <a href="package-summary.html#Reduction">reduction</a> on the
     * elements of this stream, using the provided identity, accumulation and
     * combining functions.  This is equivalent to:
     * <pre>{@code
     *     U result = identity;
     *     for (T element : this stream)
     *         result = accumulator.apply(result, element)
     *     return result;
     * }</pre>
     *
     * but is not constrained to execute sequentially.
     *
     * <p>The {@code identity} value must be an identity for the combiner
     * function.  This means that for all {@code u}, {@code combiner(identity, u)}
     * is equal to {@code u}.  Additionally, the {@code combiner} function
     * must be compatible with the {@code accumulator} function; for all
     * {@code u} and {@code t}, the following must hold:
     * <pre>{@code
     *     combiner.apply(u, accumulator.apply(identity, t)) == accumulator.apply(u, t)
     * }</pre>
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @apiNote Many reductions using this form can be represented more simply
     * by an explicit combination of {@code map} and {@code reduce} operations.
     * The {@code accumulator} function acts as a fused mapper and accumulator,
     * which can sometimes be more efficient than separate mapping and reduction,
     * such as when knowing the previously reduced value allows you to avoid
     * some computation.
     *
     * @param <U> The type of the result
     * @param identity the identity value for the combiner function
     * @param accumulator an <a href="package-summary.html#Associativity">associative</a>,
     *                    <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                    <a href="package-summary.html#Statelessness">stateless</a>
     *                    function for incorporating an additional element into a result
     * @param combiner an <a href="package-summary.html#Associativity">associative</a>,
     *                    <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                    <a href="package-summary.html#Statelessness">stateless</a>
     *                    function for combining two values, which must be
     *                    compatible with the accumulator function
     * @return the result of the reduction
     * @see #reduce(BinaryOperator)
     * @see #reduce(Object, BinaryOperator)
     */
    <U> U reduce(U identity,
                 BiFunction<U, ? super T, U> accumulator,
                 BinaryOperator<U> combiner);

    /**
     * Performs a <a href="package-summary.html#MutableReduction">mutable
     * reduction</a> operation on the elements of this stream.  A mutable
     * reduction is one in which the reduced value is a mutable result container,
     * such as an {@code ArrayList}, and elements are incorporated by updating
     * the state of the result rather than by replacing the result.  This
     * produces a result equivalent to:
     * <pre>{@code
     *     R result = supplier.get();
     *     for (T element : this stream)
     *         accumulator.accept(result, element);
     *     return result;
     * }</pre>
     *
     * <p>Like {@link #reduce(Object, BinaryOperator)}, {@code collect} operations
     * can be parallelized without requiring additional synchronization.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @apiNote There are many existing classes in the JDK whose signatures are
     * well-suited for use with method references as arguments to {@code collect()}.
     * For example, the following will accumulate strings into an {@code ArrayList}:
     * <pre>{@code
     *     List<String> asList = stringStream.collect(ArrayList::new, ArrayList::add,
     *                                                ArrayList::addAll);
     * }</pre>
     *
     * <p>The following will take a stream of strings and concatenates them into a
     * single string:
     * <pre>{@code
     *     String concat = stringStream.collect(StringBuilder::new, StringBuilder::append,
     *                                          StringBuilder::append)
     *                                 .toString();
     * }</pre>
     *
     * @param <R> type of the result
     * @param supplier a function that creates a new result container. For a
     *                 parallel execution, this function may be called
     *                 multiple times and must return a fresh value each time.
     * @param accumulator an <a href="package-summary.html#Associativity">associative</a>,
     *                    <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                    <a href="package-summary.html#Statelessness">stateless</a>
     *                    function for incorporating an additional element into a result
     * @param combiner an <a href="package-summary.html#Associativity">associative</a>,
     *                    <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                    <a href="package-summary.html#Statelessness">stateless</a>
     *                    function for combining two values, which must be
     *                    compatible with the accumulator function
     * @return the result of the reduction
     */
    <R> R collect(Supplier<R> supplier,
                  BiConsumer<R, ? super T> accumulator,
                  BiConsumer<R, R> combiner);

    /**
     * Performs a <a href="package-summary.html#MutableReduction">mutable
     * reduction</a> operation on the elements of this stream using a
     * {@code Collector}.  A {@code Collector}
     * encapsulates the functions used as arguments to
     * {@link #collect(Supplier, BiConsumer, BiConsumer)}, allowing for reuse of
     * collection strategies and composition of collect operations such as
     * multiple-level grouping or partitioning.
     *
     * <p>If the stream is parallel, and the {@code Collector}
     * is {@link Collector.Characteristics#CONCURRENT concurrent}, and
     * either the stream is unordered or the collector is
     * {@link Collector.Characteristics#UNORDERED unordered},
     * then a concurrent reduction will be performed (see {@link Collector} for
     * details on concurrent reduction.)
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * <p>When executed in parallel, multiple intermediate results may be
     * instantiated, populated, and merged so as to maintain isolation of
     * mutable data structures.  Therefore, even when executed in parallel
     * with non-thread-safe data structures (such as {@code ArrayList}), no
     * additional synchronization is needed for a parallel reduction.
     *
     * @apiNote
     * The following will accumulate strings into an ArrayList:
     * <pre>{@code
     *     List<String> asList = stringStream.collect(Collectors.toList());
     * }</pre>
     *
     * <p>The following will classify {@code Person} objects by city:
     * <pre>{@code
     *     Map<String, List<Person>> peopleByCity
     *         = personStream.collect(Collectors.groupingBy(Person::getCity));
     * }</pre>
     *
     * <p>The following will classify {@code Person} objects by state and city,
     * cascading two {@code Collector}s together:
     * <pre>{@code
     *     Map<String, Map<String, List<Person>>> peopleByStateAndCity
     *         = personStream.collect(Collectors.groupingBy(Person::getState,
     *                                                      Collectors.groupingBy(Person::getCity)));
     * }</pre>
     *
     * @param <R> the type of the result
     * @param <A> the intermediate accumulation type of the {@code Collector}
     * @param collector the {@code Collector} describing the reduction
     * @return the result of the reduction
     * @see #collect(Supplier, BiConsumer, BiConsumer)
     * @see Collectors
     */
    <R, A> R collect(Collector<? super T, A, R> collector);

    /**
     * Returns the minimum element of this stream according to the provided
     * {@code Comparator}.  This is a special case of a
     * <a href="package-summary.html#Reduction">reduction</a>.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal operation</a>.
     *
     * @param comparator a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                   <a href="package-summary.html#Statelessness">stateless</a>
     *                   {@code Comparator} to compare elements of this stream
     * @return an {@code Optional} describing the minimum element of this stream,
     * or an empty {@code Optional} if the stream is empty
     * @throws NullPointerException if the minimum element is null
     */
    Optional<T> min(Comparator<? super T> comparator);

    /**
     * Returns the maximum element of this stream according to the provided
     * {@code Comparator}.  This is a special case of a
     * <a href="package-summary.html#Reduction">reduction</a>.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal
     * operation</a>.
     *
     * @param comparator a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                   <a href="package-summary.html#Statelessness">stateless</a>
     *                   {@code Comparator} to compare elements of this stream
     * @return an {@code Optional} describing the maximum element of this stream,
     * or an empty {@code Optional} if the stream is empty
     * @throws NullPointerException if the maximum element is null
     */
    Optional<T> max(Comparator<? super T> comparator);

    /**
     * Returns the count of elements in this stream.  This is a special case of
     * a <a href="package-summary.html#Reduction">reduction</a> and is
     * equivalent to:
     * <pre>{@code
     *     return mapToLong(e -> 1L).sum();
     * }</pre>
     *
     * <p>This is a <a href="package-summary.html#StreamOps">terminal operation</a>.
     *
     * @return the count of elements in this stream
     */
    long count();

    /**
     * Returns whether any elements of this stream match the provided
     * predicate.  May not evaluate the predicate on all elements if not
     * necessary for determining the result.  If the stream is empty then
     * {@code false} is returned and the predicate is not evaluated.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">short-circuiting
     * terminal operation</a>.
     *
     * @apiNote
     * This method evaluates the <em>existential quantification</em> of the
     * predicate over the elements of the stream (for some x P(x)).
     *
     * @param predicate a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                  <a href="package-summary.html#Statelessness">stateless</a>
     *                  predicate to apply to elements of this stream
     * @return {@code true} if any elements of the stream match the provided
     * predicate, otherwise {@code false}
     */
    boolean anyMatch(Predicate<? super T> predicate);

    /**
     * Returns whether all elements of this stream match the provided predicate.
     * May not evaluate the predicate on all elements if not necessary for
     * determining the result.  If the stream is empty then {@code true} is
     * returned and the predicate is not evaluated.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">short-circuiting
     * terminal operation</a>.
     *
     * @apiNote
     * This method evaluates the <em>universal quantification</em> of the
     * predicate over the elements of the stream (for all x P(x)).  If the
     * stream is empty, the quantification is said to be <em>vacuously
     * satisfied</em> and is always {@code true} (regardless of P(x)).
     *
     * @param predicate a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                  <a href="package-summary.html#Statelessness">stateless</a>
     *                  predicate to apply to elements of this stream
     * @return {@code true} if either all elements of the stream match the
     * provided predicate or the stream is empty, otherwise {@code false}
     */
    boolean allMatch(Predicate<? super T> predicate);

    /**
     * Returns whether no elements of this stream match the provided predicate.
     * May not evaluate the predicate on all elements if not necessary for
     * determining the result.  If the stream is empty then {@code true} is
     * returned and the predicate is not evaluated.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">short-circuiting
     * terminal operation</a>.
     *
     * @apiNote
     * This method evaluates the <em>universal quantification</em> of the
     * negated predicate over the elements of the stream (for all x ~P(x)).  If
     * the stream is empty, the quantification is said to be vacuously satisfied
     * and is always {@code true}, regardless of P(x).
     *
     * @param predicate a <a href="package-summary.html#NonInterference">non-interfering</a>,
     *                  <a href="package-summary.html#Statelessness">stateless</a>
     *                  predicate to apply to elements of this stream
     * @return {@code true} if either no elements of the stream match the
     * provided predicate or the stream is empty, otherwise {@code false}
     */
    boolean noneMatch(Predicate<? super T> predicate);

    /**
     * Returns an {@link Optional} describing the first element of this stream,
     * or an empty {@code Optional} if the stream is empty.  If the stream has
     * no encounter order, then any element may be returned.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">short-circuiting
     * terminal operation</a>.
     *
     * @return an {@code Optional} describing the first element of this stream,
     * or an empty {@code Optional} if the stream is empty
     * @throws NullPointerException if the element selected is null
     */
    Optional<T> findFirst();

    /**
     * Returns an {@link Optional} describing some element of the stream, or an
     * empty {@code Optional} if the stream is empty.
     *
     * <p>This is a <a href="package-summary.html#StreamOps">short-circuiting
     * terminal operation</a>.
     *
     * <p>The behavior of this operation is explicitly nondeterministic; it is
     * free to select any element in the stream.  This is to allow for maximal
     * performance in parallel operations; the cost is that multiple invocations
     * on the same source may not return the same result.  (If a stable result
     * is desired, use {@link #findFirst()} instead.)
     *
     * @return an {@code Optional} describing some element of this stream, or an
     * empty {@code Optional} if the stream is empty
     * @throws NullPointerException if the element selected is null
     * @see #findFirst()
     */
    Optional<T> findAny();

    // Static factories

    /**
     * Returns a builder for a {@code Stream}.
     *
     * @param <T> type of elements
     * @return a stream builder
     */
    public static<T> Builder<T> builder() {
        return new Streams.StreamBuilderImpl<>();
    }

    /**
     * Returns an empty sequential {@code Stream}.
     *
     * @param <T> the type of stream elements
     * @return an empty sequential stream
     */
    public static<T> Stream<T> empty() {
        return StreamSupport.stream(Spliterators.<T>emptySpliterator(), false);
    }

    /**
     * Returns a sequential {@code Stream} containing a single element.
     *
     * @param t the single element
     * @param <T> the type of stream elements
     * @return a singleton sequential stream
     */
    public static<T> Stream<T> of(T t) {
        return StreamSupport.stream(new Streams.StreamBuilderImpl<>(t), false);
    }

    /**
     * Returns a sequential ordered stream whose elements are the specified values.
     *
     * @param <T> the type of stream elements
     * @param values the elements of the new stream
     * @return the new stream
     */
    @SafeVarargs
    @SuppressWarnings("varargs") // Creating a stream from an array is safe
    public static<T> Stream<T> of(T... values) {
        return Arrays.stream(values);
    }

    /**
     * Returns an infinite sequential ordered {@code Stream} produced by iterative
     * application of a function {@code f} to an initial element {@code seed},
     * producing a {@code Stream} consisting of {@code seed}, {@code f(seed)},
     * {@code f(f(seed))}, etc.
     *
     * <p>The first element (position {@code 0}) in the {@code Stream} will be
     * the provided {@code seed}.  For {@code n > 0}, the element at position
     * {@code n}, will be the result of applying the function {@code f} to the
     * element at position {@code n - 1}.
     *
     * @param <T> the type of stream elements
     * @param seed the initial element
     * @param f a function to be applied to to the previous element to produce
     *          a new element
     * @return a new sequential {@code Stream}
     */
    public static<T> Stream<T> iterate(final T seed, final UnaryOperator<T> f) {
        Objects.requireNonNull(f);
        final Iterator<T> iterator = new Iterator<T>() {
            @SuppressWarnings("unchecked")
            T t = (T) Streams.NONE;

            @Override
            public boolean hasNext() {
                return true;
            }

            @Override
            public T next() {
                return t = (t == Streams.NONE) ? seed : f.apply(t);
            }
        };
        return StreamSupport.stream(Spliterators.spliteratorUnknownSize(
                iterator,
                Spliterator.ORDERED | Spliterator.IMMUTABLE), false);
    }

    /**
     * Returns an infinite sequential unordered stream where each element is
     * generated by the provided {@code Supplier}.  This is suitable for
     * generating constant streams, streams of random elements, etc.
     *
     * @param <T> the type of stream elements
     * @param s the {@code Supplier} of generated elements
     * @return a new infinite sequential unordered {@code Stream}
     */
    public static<T> Stream<T> generate(Supplier<T> s) {
        Objects.requireNonNull(s);
        return StreamSupport.stream(
                new StreamSpliterators.InfiniteSupplyingSpliterator.OfRef<>(Long.MAX_VALUE, s), false);
    }

    /**
     * Creates a lazily concatenated stream whose elements are all the
     * elements of the first stream followed by all the elements of the
     * second stream.  The resulting stream is ordered if both
     * of the input streams are ordered, and parallel if either of the input
     * streams is parallel.  When the resulting stream is closed, the close
     * handlers for both input streams are invoked.
     *
     * @implNote
     * Use caution when constructing streams from repeated concatenation.
     * Accessing an element of a deeply concatenated stream can result in deep
     * call chains, or even {@code StackOverflowException}.
     *
     * @param <T> The type of stream elements
     * @param a the first stream
     * @param b the second stream
     * @return the concatenation of the two input streams
     */
    public static <T> Stream<T> concat(Stream<? extends T> a, Stream<? extends T> b) {
        Objects.requireNonNull(a);
        Objects.requireNonNull(b);

        @SuppressWarnings("unchecked")
        Spliterator<T> split = new Streams.ConcatSpliterator.OfRef<>(
                (Spliterator<T>) a.spliterator(), (Spliterator<T>) b.spliterator());
        Stream<T> stream = StreamSupport.stream(split, a.isParallel() || b.isParallel());
        return stream.onClose(Streams.composedClose(a, b));
    }

    /**
     * A mutable builder for a {@code Stream}.  This allows the creation of a
     * {@code Stream} by generating elements individually and adding them to the
     * {@code Builder} (without the copying overhead that comes from using
     * an {@code ArrayList} as a temporary buffer.)
     *
     * <p>A stream builder has a lifecycle, which starts in a building
     * phase, during which elements can be added, and then transitions to a built
     * phase, after which elements may not be added.  The built phase begins
     * when the {@link #build()} method is called, which creates an ordered
     * {@code Stream} whose elements are the elements that were added to the stream
     * builder, in the order they were added.
     *
     * @param <T> the type of stream elements
     * @see Stream#builder()
     * @since 1.8
     */
    public interface Builder<T> extends Consumer<T> {

        /**
         * Adds an element to the stream being built.
         *
         * @throws IllegalStateException if the builder has already transitioned to
         * the built state
         */
        @Override
        void accept(T t);

        /**
         * Adds an element to the stream being built.
         *
         * @implSpec
         * The default implementation behaves as if:
         * <pre>{@code
         *     accept(t)
         *     return this;
         * }</pre>
         *
         * @param t the element to add
         * @return {@code this} builder
         * @throws IllegalStateException if the builder has already transitioned to
         * the built state
         */
        default Builder<T> add(T t) {
            accept(t);
            return this;
        }

        /**
         * Builds the stream, transitioning this builder to the built state.
         * An {@code IllegalStateException} is thrown if there are further attempts
         * to operate on the builder after it has entered the built state.
         *
         * @return the built stream
         * @throws IllegalStateException if the builder has already transitioned to
         * the built state
         */
        Stream<T> build();

    }
}
View Code

  只是,這介面中定義的引數,都是些經過特殊定義的介面,即函式式介面,即預設只需實現一個方法即可介面類定義。

 

3. stream包的具體實現?

  如上一節,我們已知stream中主要依賴於許多的介面定義。既然是介面,那就必然無法直接呼叫,須要有與之對應的實現方可呼叫。所以,我們需要有特定的場景,才可以來談stream 的實現問題。

  所以,我們先以相對簡單的 Integer 的流轉化與處理過程,一探stream究竟。

     // java.util.Arrays#stream(T[])
    /**
     * Returns a sequential {@link Stream} with the specified array as its
     * source.
     *
     * @param <T> The type of the array elements
     * @param array The array, assumed to be unmodified during use
     * @return a {@code Stream} for the array
     * @since 1.8
     */
    public static <T> Stream<T> stream(T[] array) {
        return stream(array, 0, array.length);
    }
    // java.util.Arrays#stream(T[], int, int)
    /**
     * Returns a sequential {@link Stream} with the specified range of the
     * specified array as its source.
     *
     * @param <T> the type of the array elements
     * @param array the array, assumed to be unmodified during use
     * @param startInclusive the first index to cover, inclusive
     * @param endExclusive index immediately past the last index to cover
     * @return a {@code Stream} for the array range
     * @throws ArrayIndexOutOfBoundsException if {@code startInclusive} is
     *         negative, {@code endExclusive} is less than
     *         {@code startInclusive}, or {@code endExclusive} is greater than
     *         the array size
     * @since 1.8
     */
    public static <T> Stream<T> stream(T[] array, int startInclusive, int endExclusive) {
        // 構造 iterator, 帶入 StreamSupport 中
        return StreamSupport.stream(spliterator(array, startInclusive, endExclusive), false);
    }

    /**
     * Returns a {@link Spliterator} covering the specified range of the
     * specified array.
     *
     * <p>The spliterator reports {@link Spliterator#SIZED},
     * {@link Spliterator#SUBSIZED}, {@link Spliterator#ORDERED}, and
     * {@link Spliterator#IMMUTABLE}.
     *
     * @param <T> type of elements
     * @param array the array, assumed to be unmodified during use
     * @param startInclusive the first index to cover, inclusive
     * @param endExclusive index immediately past the last index to cover
     * @return a spliterator for the array elements
     * @throws ArrayIndexOutOfBoundsException if {@code startInclusive} is
     *         negative, {@code endExclusive} is less than
     *         {@code startInclusive}, or {@code endExclusive} is greater than
     *         the array size
     * @since 1.8
     */
    public static <T> Spliterator<T> spliterator(T[] array, int startInclusive, int endExclusive) {
        return Spliterators.spliterator(array, startInclusive, endExclusive,
                                        Spliterator.ORDERED | Spliterator.IMMUTABLE);
    }
    // java.util.stream.StreamSupport#stream(java.util.Spliterator<T>, boolean)
    /**
     * Creates a new sequential or parallel {@code Stream} from a
     * {@code Spliterator}.
     *
     * <p>The spliterator is only traversed, split, or queried for estimated
     * size after the terminal operation of the stream pipeline commences.
     *
     * <p>It is strongly recommended the spliterator report a characteristic of
     * {@code IMMUTABLE} or {@code CONCURRENT}, or be
     * <a href="../Spliterator.html#binding">late-binding</a>.  Otherwise,
     * {@link #stream(java.util.function.Supplier, int, boolean)} should be used
     * to reduce the scope of potential interference with the source.  See
     * <a href="package-summary.html#NonInterference">Non-Interference</a> for
     * more details.
     *
     * @param <T> the type of stream elements
     * @param spliterator a {@code Spliterator} describing the stream elements
     * @param parallel if {@code true} then the returned stream is a parallel
     *        stream; if {@code false} the returned stream is a sequential
     *        stream.
     * @return a new sequential or parallel {@code Stream}
     */
    public static <T> Stream<T> stream(Spliterator<T> spliterator, boolean parallel) {
        Objects.requireNonNull(spliterator);
        return new ReferencePipeline.Head<>(spliterator,
                                            StreamOpFlag.fromCharacteristics(spliterator),
                                            parallel);
    }
        // java.util.stream.ReferencePipeline.Head#Head(java.util.Spliterator<?>, int, boolean)
        /**
         * Constructor for the source stage of a Stream.
         *
         * @param source {@code Spliterator} describing the stream source
         * @param sourceFlags the source flags for the stream source, described
         *                    in {@link StreamOpFlag}
         */
        Head(Spliterator<?> source,
             int sourceFlags, boolean parallel) {
            super(source, sourceFlags, parallel);
        }
    // java.util.stream.ReferencePipeline#ReferencePipeline(java.util.Spliterator<?>, int, boolean)
    /**
     * Constructor for the head of a stream pipeline.
     *
     * @param source {@code Spliterator} describing the stream source
     * @param sourceFlags The source flags for the stream source, described in
     *        {@link StreamOpFlag}
     * @param parallel {@code true} if the pipeline is parallel
     */
    ReferencePipeline(Spliterator<?> source,
                      int sourceFlags, boolean parallel) {
        super(source, sourceFlags, parallel);
    }
    // java.util.stream.AbstractPipeline#AbstractPipeline(java.util.Spliterator<?>, int, boolean)
    /**
     * Constructor for the head of a stream pipeline.
     *
     * @param source {@code Spliterator} describing the stream source
     * @param sourceFlags the source flags for the stream source, described in
     * {@link StreamOpFlag}
     * @param parallel {@code true} if the pipeline is parallel
     */
    AbstractPipeline(Spliterator<?> source,
                     int sourceFlags, boolean parallel) {
        this.previousStage = null;
        this.sourceSpliterator = source;
        this.sourceStage = this;
        this.sourceOrOpFlags = sourceFlags & StreamOpFlag.STREAM_MASK;
        // The following is an optimization of:
        // StreamOpFlag.combineOpFlags(sourceOrOpFlags, StreamOpFlag.INITIAL_OPS_VALUE);
        this.combinedFlags = (~(sourceOrOpFlags << 1)) & StreamOpFlag.INITIAL_OPS_VALUE;
        this.depth = 0;
        this.parallel = parallel;
    }
    如上,就返回了一 Stream 的具體例項,即是 ReferencePipeline.Head 的例項。故而,之後的每個stream操作如 filter,map,foreach方法,都盡在該 head 中進行實現了。一瞅便知。
    // java.util.stream.ReferencePipeline#filter
    @Override
    public final Stream<P_OUT> filter(Predicate<? super P_OUT> predicate) {
        Objects.requireNonNull(predicate);
        // 只返回了一個 StreamlessOp例項
        return new StatelessOp<P_OUT, P_OUT>(this, StreamShape.REFERENCE,
                                     StreamOpFlag.NOT_SIZED) {
            @Override
            Sink<P_OUT> opWrapSink(int flags, Sink<P_OUT> sink) {
                return new Sink.ChainedReference<P_OUT, P_OUT>(sink) {
                    @Override
                    public void begin(long size) {
                        downstream.begin(-1);
                    }

                    @Override
                    public void accept(P_OUT u) {
                        // 在必要時候呼叫 test() 方法即可
                        // 當test返回 true 時,該元素被保留傳入下一級呼叫中,此即filter的語義
                        if (predicate.test(u))
                            downstream.accept(u);
                    }
                };
            }
        };
    }
    // java.util.stream.ReferencePipeline#map
    @Override
    @SuppressWarnings("unchecked")
    public final <R> Stream<R> map(Function<? super P_OUT, ? extends R> mapper) {
        Objects.requireNonNull(mapper);
        // 同樣,僅返回一個 StatelessOp 的例項
        return new StatelessOp<P_OUT, R>(this, StreamShape.REFERENCE,
                                     StreamOpFlag.NOT_SORTED | StreamOpFlag.NOT_DISTINCT) {
            @Override
            Sink<P_OUT> opWrapSink(int flags, Sink<R> sink) {
                return new Sink.ChainedReference<P_OUT, R>(sink) {
                    @Override
                    public void accept(P_OUT u) {
                        // 同樣,在必要的時候呼叫 apply 方法
                        // 即 map 的語義為 每個元素都會呼叫該方法
                        downstream.accept(mapper.apply(u));
                    }
                };
            }
        };
    }
    @Override
    public final <R> Stream<R> flatMap(Function<? super P_OUT, ? extends Stream<? extends R>> mapper) {
        Objects.requireNonNull(mapper);
        // We can do better than this, by polling cancellationRequested when stream is infinite
        return new StatelessOp<P_OUT, R>(this, StreamShape.REFERENCE,
                                     StreamOpFlag.NOT_SORTED | StreamOpFlag.NOT_DISTINCT | StreamOpFlag.NOT_SIZED) {
            @Override
            Sink<P_OUT> opWrapSink(int flags, Sink<R> sink) {
                return new Sink.ChainedReference<P_OUT, R>(sink) {
                    @Override
                    public void begin(long size) {
                        downstream.begin(-1);
                    }

                    @Override
                    public void accept(P_OUT u) {
                        // flatmap 語義,所得結果,依次往下傳輸
                        try (Stream<? extends R> result = mapper.apply(u)) {
                            // We can do better that this too; optimize for depth=0 case and just grab spliterator and forEach it
                            if (result != null)
                                result.sequential().forEach(downstream);
                        }
                    }
                };
            }
        };
    }

  如上,幾個方法呼叫下來,我們基本都可以看到,都是一個個的 StatelessOp 的例項的返回,但都沒有觸發真正的計算。那麼,真正計算又要到幾時呢?相信有些其他知識面的你,定然會想到,在合適的時候再來觸發真正的運算操作。當資料結構不會發生本質的變化時,這種平衡就是存在的。只是在一些關鍵時候,才會觸發運算。這為後續進行平行計算或者效能優化提供了可能。

  那麼,stream包中,哪些運算是作為真正的觸發行為呢?至少 collect(), foreach(), reduce() 是會進行觸發的。 這些優化手段,不知和其他框架實現,誰先誰後,誰主誰從。反正,總是好的想法。在其他地方,也許叫許多運算元。

  我們以collect()探查如何使用這stream的威力?

    // java.util.stream.ReferencePipeline#collect(java.util.stream.Collector<? super P_OUT,A,R>)
    @Override
    @SuppressWarnings("unchecked")
    public final <R, A> R collect(Collector<? super P_OUT, A, R> collector) {
        A container;
        // 即分並行與序列
        if (isParallel()
                && (collector.characteristics().contains(Collector.Characteristics.CONCURRENT))
                && (!isOrdered() || collector.characteristics().contains(Collector.Characteristics.UNORDERED))) {
            container = collector.supplier().get();
            BiConsumer<A, ? super P_OUT> accumulator = collector.accumulator();
            forEach(u -> accumulator.accept(container, u));
        }
        else {
            // 序列執行
            container = evaluate(ReduceOps.makeRef(collector));
        }
        return collector.characteristics().contains(Collector.Characteristics.IDENTITY_FINISH)
               ? (R) container
               : collector.finisher().apply(container);
    }

    /**
     * Constructs a {@code TerminalOp} that implements a mutable reduce on
     * reference values.
     *
     * @param <T> the type of the input elements
     * @param <I> the type of the intermediate reduction result
     * @param collector a {@code Collector} defining the reduction
     * @return a {@code ReduceOp} implementing the reduction
     */
    public static <T, I> TerminalOp<T, I>
    makeRef(Collector<? super T, I, ?> collector) {
        Supplier<I> supplier = Objects.requireNonNull(collector).supplier();
        BiConsumer<I, ? super T> accumulator = collector.accumulator();
        BinaryOperator<I> combiner = collector.combiner();
        class ReducingSink extends Box<I>
                implements AccumulatingSink<T, I, ReducingSink> {
            @Override
            public void begin(long size) {
                state = supplier.get();
            }

            @Override
            public void accept(T t) {
                accumulator.accept(state, t);
            }

            @Override
            public void combine(ReducingSink other) {
                state = combiner.apply(state, other.state);
            }
        }
        // 返回ReuceOp
        return new ReduceOp<T, I, ReducingSink>(StreamShape.REFERENCE) {
            @Override
            public ReducingSink makeSink() {
                return new ReducingSink();
            }

            @Override
            public int getOpFlags() {
                return collector.characteristics().contains(Collector.Characteristics.UNORDERED)
                       ? StreamOpFlag.NOT_ORDERED
                       : 0;
            }
        };
    }

    // 運算一系列任務
    /**
     * Evaluate the pipeline with a terminal operation to produce a result.
     *
     * @param <R> the type of result
     * @param terminalOp the terminal operation to be applied to the pipeline.
     * @return the result
     */
    final <R> R evaluate(TerminalOp<E_OUT, R> terminalOp) {
        assert getOutputShape() == terminalOp.inputShape();
        if (linkedOrConsumed)
            throw new IllegalStateException(MSG_STREAM_LINKED);
        linkedOrConsumed = true;

        return isParallel()
               ? terminalOp.evaluateParallel(this, sourceSpliterator(terminalOp.getOpFlags()))
               : terminalOp.evaluateSequential(this, sourceSpliterator(terminalOp.getOpFlags()));
    }

        // java.util.stream.ReduceOps.ReduceOp#evaluateSequential
        @Override
        public <P_IN> R evaluateSequential(PipelineHelper<T> helper,
                                           Spliterator<P_IN> spliterator) {
            return helper.wrapAndCopyInto(makeSink(), spliterator).get();
        }

    // java.util.stream.AbstractPipeline#wrapAndCopyInto
    @Override
    final <P_IN, S extends Sink<E_OUT>> S wrapAndCopyInto(S sink, Spliterator<P_IN> spliterator) {
        copyInto(wrapSink(Objects.requireNonNull(sink)), spliterator);
        return sink;
    }

    // java.util.stream.AbstractPipeline#wrapSink
    @Override
    @SuppressWarnings("unchecked")
    final <P_IN> Sink<P_IN> wrapSink(Sink<E_OUT> sink) {
        Objects.requireNonNull(sink);
        // 基本是按照倒序來排的
        for ( @SuppressWarnings("rawtypes") AbstractPipeline p=AbstractPipeline.this; p.depth > 0; p=p.previousStage) {
            // 一層層包裝運算元
            sink = p.opWrapSink(p.previousStage.combinedFlags, sink);
        }
        return (Sink<P_IN>) sink;
    }

    // java.util.stream.AbstractPipeline#copyInto
    @Override
    final <P_IN> void copyInto(Sink<P_IN> wrappedSink, Spliterator<P_IN> spliterator) {
        Objects.requireNonNull(wrappedSink);
        // 依次呼叫 begin, foreach, end 方法
        if (!StreamOpFlag.SHORT_CIRCUIT.isKnown(getStreamAndOpFlags())) {
            wrappedSink.begin(spliterator.getExactSizeIfKnown());
            // 每個元素依次迭代, 一層層退出來
            spliterator.forEachRemaining(wrappedSink);
            wrappedSink.end();
        }
        else {
            copyIntoWithCancel(wrappedSink, spliterator);
        }
    }

        // java.util.Spliterators.ArraySpliterator#forEachRemaining
        @SuppressWarnings("unchecked")
        @Override
        public void forEachRemaining(Consumer<? super T> action) {
            Object[] a; int i, hi; // hoist accesses and checks from loop
            if (action == null)
                throw new NullPointerException();
            if ((a = array).length >= (hi = fence) &&
                (i = index) >= 0 && i < (index = hi)) {
                do { action.accept((T)a[i]); } while (++i < hi);
            }
        }

  可見,該stream包的實現中,大量使用了包裝器模式,責任鏈模式,模板方法模式,以及在必要的節點再進行統一的運算觸發。且在必要的時候開啟平行計算,為上層應用帶了各種可能。在使用起來極其簡單的同時,又兼顧了效能。(我說的不是通常的效能,比如我自己寫幾個簡單的filter豈不效能更好?)其中,以上這些簡便的寫法,得益於lamda語法的支援,以及幾個簡單的函式式介面定義。比如 Consumer, Function... 它們都被定義在java.util.function包下面。

談談stream的執行原理
@FunctionalInterface
public interface Consumer<T> {

    /**
     * Performs this operation on the given argument.
     *
     * @param t the input argument
     */
    void accept(T t);

    /**
     * Returns a composed {@code Consumer} that performs, in sequence, this
     * operation followed by the {@code after} operation. If performing either
     * operation throws an exception, it is relayed to the caller of the
     * composed operation.  If performing this operation throws an exception,
     * the {@code after} operation will not be performed.
     *
     * @param after the operation to perform after this operation
     * @return a composed {@code Consumer} that performs in sequence this
     * operation followed by the {@code after} operation
     * @throws NullPointerException if {@code after} is null
     */
    default Consumer<T> andThen(Consumer<? super T> after) {
        Objects.requireNonNull(after);
        return (T t) -> { accept(t); after.accept(t); };
    }
}
@FunctionalInterface
public interface Function<T, R> {

    /**
     * Applies this function to the given argument.
     *
     * @param t the function argument
     * @return the function result
     */
    R apply(T t);

    /**
     * Returns a composed function that first applies the {@code before}
     * function to its input, and then applies this function to the result.
     * If evaluation of either function throws an exception, it is relayed to
     * the caller of the composed function.
     *
     * @param <V> the type of input to the {@code before} function, and to the
     *           composed function
     * @param before the function to apply before this function is applied
     * @return a composed function that first applies the {@code before}
     * function and then applies this function
     * @throws NullPointerException if before is null
     *
     * @see #andThen(Function)
     */
    default <V> Function<V, R> compose(Function<? super V, ? extends T> before) {
        Objects.requireNonNull(before);
        return (V v) -> apply(before.apply(v));
    }

    /**
     * Returns a composed function that first applies this function to
     * its input, and then applies the {@code after} function to the result.
     * If evaluation of either function throws an exception, it is relayed to
     * the caller of the composed function.
     *
     * @param <V> the type of output of the {@code after} function, and of the
     *           composed function
     * @param after the function to apply after this function is applied
     * @return a composed function that first applies this function and then
     * applies the {@code after} function
     * @throws NullPointerException if after is null
     *
     * @see #compose(Function)
     */
    default <V> Function<T, V> andThen(Function<? super R, ? extends V> after) {
        Objects.requireNonNull(after);
        return (T t) -> after.apply(apply(t));
    }

    /**
     * Returns a function that always returns its input argument.
     *
     * @param <T> the type of the input and output objects to the function
     * @return a function that always returns its input argument
     */
    static <T> Function<T, T> identity() {
        return t -> t;
    }
}

@FunctionalInterface
public interface Supplier<T> {

    /**
     * Gets a result.
     *
     * @return a result
     */
    T get();
}
View Code

 

相關文章