Fix Java function API methods for flatMap and mapPartitions to require producing only an Iterator, not Iterable. Also fix DStream.flatMap to require a function producing TraversableOnce only, not Traversable.

Author: srowen

core/src/main/java/org/apache/spark/api/java/function/CoGroupFunction.java

@@ -25,5 +25,5 @@
  * Datasets.
  */
 public interface CoGroupFunction<K, V1, V2, R> extends Serializable {
-  Iterable<R> call(K key, Iterator<V1> left, Iterator<V2> right) throws Exception;
+  Iterator<R> call(K key, Iterator<V1> left, Iterator<V2> right) throws Exception;
 }

core/src/main/java/org/apache/spark/api/java/function/DoubleFlatMapFunction.java

@@ -18,10 +18,11 @@
 package org.apache.spark.api.java.function;
 
 import java.io.Serializable;
+import java.util.Iterator;
 
 /**
  * A function that returns zero or more records of type Double from each input record.
  */
 public interface DoubleFlatMapFunction<T> extends Serializable {
-  public Iterable<Double> call(T t) throws Exception;
+  Iterator<Double> call(T t) throws Exception;
 }

core/src/main/java/org/apache/spark/api/java/function/FlatMapFunction.java

@@ -18,10 +18,11 @@
 package org.apache.spark.api.java.function;
 
 import java.io.Serializable;
+import java.util.Iterator;
 
 /**
  * A function that returns zero or more output records from each input record.
  */
 public interface FlatMapFunction<T, R> extends Serializable {
-  Iterable<R> call(T t) throws Exception;
+  Iterator<R> call(T t) throws Exception;
 }

core/src/main/java/org/apache/spark/api/java/function/FlatMapFunction2.java

@@ -18,10 +18,11 @@
 package org.apache.spark.api.java.function;
 
 import java.io.Serializable;
+import java.util.Iterator;
 
 /**
  * A function that takes two inputs and returns zero or more output records.
  */
 public interface FlatMapFunction2<T1, T2, R> extends Serializable {
-  Iterable<R> call(T1 t1, T2 t2) throws Exception;
+  Iterator<R> call(T1 t1, T2 t2) throws Exception;
 }

core/src/main/java/org/apache/spark/api/java/function/FlatMapGroupsFunction.java

@@ -24,5 +24,5 @@
  * A function that returns zero or more output records from each grouping key and its values.
  */
 public interface FlatMapGroupsFunction<K, V, R> extends Serializable {
-  Iterable<R> call(K key, Iterator<V> values) throws Exception;
+  Iterator<R> call(K key, Iterator<V> values) throws Exception;
 }

core/src/main/java/org/apache/spark/api/java/function/MapPartitionsFunction.java

@@ -24,5 +24,5 @@
  * Base interface for function used in Dataset's mapPartitions.
  */
 public interface MapPartitionsFunction<T, U> extends Serializable {
-  Iterable<U> call(Iterator<T> input) throws Exception;
+  Iterator<U> call(Iterator<T> input) throws Exception;
 }

core/src/main/java/org/apache/spark/api/java/function/PairFlatMapFunction.java

@@ -18,6 +18,7 @@
 package org.apache.spark.api.java.function;
 
 import java.io.Serializable;
+import java.util.Iterator;
 
 import scala.Tuple2;
 
@@ -26,5 +27,5 @@
  * key-value pairs are represented as scala.Tuple2 objects.
  */
 public interface PairFlatMapFunction<T, K, V> extends Serializable {
-  public Iterable<Tuple2<K, V>> call(T t) throws Exception;
+  Iterator<Tuple2<K, V>> call(T t) throws Exception;
 }

core/src/main/scala/org/apache/spark/api/java/JavaRDDLike.scala

@@ -121,7 +121,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
    *  RDD, and then flattening the results.
    */
   def flatMap[U](f: FlatMapFunction[T, U]): JavaRDD[U] = {
-    def fn: (T) => Iterable[U] = (x: T) => f.call(x).asScala
+    def fn: (T) => Iterator[U] = (x: T) => f.call(x).asScala
     JavaRDD.fromRDD(rdd.flatMap(fn)(fakeClassTag[U]))(fakeClassTag[U])
   }
 
@@ -130,7 +130,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
    *  RDD, and then flattening the results.
    */
   def flatMapToDouble(f: DoubleFlatMapFunction[T]): JavaDoubleRDD = {
-    def fn: (T) => Iterable[jl.Double] = (x: T) => f.call(x).asScala
+    def fn: (T) => Iterator[jl.Double] = (x: T) => f.call(x).asScala
     new JavaDoubleRDD(rdd.flatMap(fn).map((x: jl.Double) => x.doubleValue()))
   }
 
@@ -139,7 +139,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
    *  RDD, and then flattening the results.
    */
   def flatMapToPair[K2, V2](f: PairFlatMapFunction[T, K2, V2]): JavaPairRDD[K2, V2] = {
-    def fn: (T) => Iterable[(K2, V2)] = (x: T) => f.call(x).asScala
+    def fn: (T) => Iterator[(K2, V2)] = (x: T) => f.call(x).asScala
     def cm: ClassTag[(K2, V2)] = implicitly[ClassTag[(K2, V2)]]
     JavaPairRDD.fromRDD(rdd.flatMap(fn)(cm))(fakeClassTag[K2], fakeClassTag[V2])
   }
@@ -149,7 +149,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
    */
   def mapPartitions[U](f: FlatMapFunction[java.util.Iterator[T], U]): JavaRDD[U] = {
     def fn: (Iterator[T]) => Iterator[U] = {
-      (x: Iterator[T]) => f.call(x.asJava).iterator().asScala
+      (x: Iterator[T]) => f.call(x.asJava).asScala
     }
     JavaRDD.fromRDD(rdd.mapPartitions(fn)(fakeClassTag[U]))(fakeClassTag[U])
   }
@@ -160,7 +160,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
   def mapPartitions[U](f: FlatMapFunction[java.util.Iterator[T], U],
       preservesPartitioning: Boolean): JavaRDD[U] = {
     def fn: (Iterator[T]) => Iterator[U] = {
-      (x: Iterator[T]) => f.call(x.asJava).iterator().asScala
+      (x: Iterator[T]) => f.call(x.asJava).asScala
     }
     JavaRDD.fromRDD(
       rdd.mapPartitions(fn, preservesPartitioning)(fakeClassTag[U]))(fakeClassTag[U])
@@ -171,7 +171,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
    */
   def mapPartitionsToDouble(f: DoubleFlatMapFunction[java.util.Iterator[T]]): JavaDoubleRDD = {
     def fn: (Iterator[T]) => Iterator[jl.Double] = {
-      (x: Iterator[T]) => f.call(x.asJava).iterator().asScala
+      (x: Iterator[T]) => f.call(x.asJava).asScala
     }
     new JavaDoubleRDD(rdd.mapPartitions(fn).map((x: jl.Double) => x.doubleValue()))
   }
@@ -182,7 +182,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
   def mapPartitionsToPair[K2, V2](f: PairFlatMapFunction[java.util.Iterator[T], K2, V2]):
   JavaPairRDD[K2, V2] = {
     def fn: (Iterator[T]) => Iterator[(K2, V2)] = {
-      (x: Iterator[T]) => f.call(x.asJava).iterator().asScala
+      (x: Iterator[T]) => f.call(x.asJava).asScala
     }
     JavaPairRDD.fromRDD(rdd.mapPartitions(fn))(fakeClassTag[K2], fakeClassTag[V2])
   }
@@ -193,7 +193,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
   def mapPartitionsToDouble(f: DoubleFlatMapFunction[java.util.Iterator[T]],
       preservesPartitioning: Boolean): JavaDoubleRDD = {
     def fn: (Iterator[T]) => Iterator[jl.Double] = {
-      (x: Iterator[T]) => f.call(x.asJava).iterator().asScala
+      (x: Iterator[T]) => f.call(x.asJava).asScala
     }
     new JavaDoubleRDD(rdd.mapPartitions(fn, preservesPartitioning)
       .map(x => x.doubleValue()))
@@ -205,7 +205,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
   def mapPartitionsToPair[K2, V2](f: PairFlatMapFunction[java.util.Iterator[T], K2, V2],
       preservesPartitioning: Boolean): JavaPairRDD[K2, V2] = {
     def fn: (Iterator[T]) => Iterator[(K2, V2)] = {
-      (x: Iterator[T]) => f.call(x.asJava).iterator().asScala
+      (x: Iterator[T]) => f.call(x.asJava).asScala
     }
     JavaPairRDD.fromRDD(
       rdd.mapPartitions(fn, preservesPartitioning))(fakeClassTag[K2], fakeClassTag[V2])
@@ -290,7 +290,7 @@ trait JavaRDDLike[T, This <: JavaRDDLike[T, This]] extends Serializable {
       other: JavaRDDLike[U, _],
       f: FlatMapFunction2[java.util.Iterator[T], java.util.Iterator[U], V]): JavaRDD[V] = {
     def fn: (Iterator[T], Iterator[U]) => Iterator[V] = {
-      (x: Iterator[T], y: Iterator[U]) => f.call(x.asJava, y.asJava).iterator().asScala
+      (x: Iterator[T], y: Iterator[U]) => f.call(x.asJava, y.asJava).asScala
     }
     JavaRDD.fromRDD(
       rdd.zipPartitions(other.rdd)(fn)(other.classTag, fakeClassTag[V]))(fakeClassTag[V])

core/src/test/java/org/apache/spark/JavaAPISuite.java

@@ -880,8 +880,8 @@ public void flatMap() {
       "The quick brown fox jumps over the lazy dog."));
     JavaRDD<String> words = rdd.flatMap(new FlatMapFunction<String, String>() {
       @Override
-      public Iterable<String> call(String x) {
-        return Arrays.asList(x.split(" "));
+      public Iterator<String> call(String x) {
+        return Arrays.asList(x.split(" ")).iterator();
       }
     });
     Assert.assertEquals("Hello", words.first());
@@ -890,12 +890,12 @@ public Iterable<String> call(String x) {
     JavaPairRDD<String, String> pairsRDD = rdd.flatMapToPair(
       new PairFlatMapFunction<String, String, String>() {
         @Override
-        public Iterable<Tuple2<String, String>> call(String s) {
+        public Iterator<Tuple2<String, String>> call(String s) {
           List<Tuple2<String, String>> pairs = new LinkedList<>();
           for (String word : s.split(" ")) {
             pairs.add(new Tuple2<>(word, word));
           }
-          return pairs;
+          return pairs.iterator();
         }
       }
     );
@@ -904,12 +904,12 @@ public Iterable<Tuple2<String, String>> call(String s) {
 
     JavaDoubleRDD doubles = rdd.flatMapToDouble(new DoubleFlatMapFunction<String>() {
       @Override
-      public Iterable<Double> call(String s) {
+      public Iterator<Double> call(String s) {
         List<Double> lengths = new LinkedList<>();
         for (String word : s.split(" ")) {
           lengths.add((double) word.length());
         }
-        return lengths;
+        return lengths.iterator();
       }
     });
     Assert.assertEquals(5.0, doubles.first(), 0.01);
@@ -930,8 +930,8 @@ public void mapsFromPairsToPairs() {
     JavaPairRDD<String, Integer> swapped = pairRDD.flatMapToPair(
       new PairFlatMapFunction<Tuple2<Integer, String>, String, Integer>() {
         @Override
-        public Iterable<Tuple2<String, Integer>> call(Tuple2<Integer, String> item) {
-          return Collections.singletonList(item.swap());
+        public Iterator<Tuple2<String, Integer>> call(Tuple2<Integer, String> item) {
+          return Collections.singletonList(item.swap()).iterator();
         }
       });
     swapped.collect();
@@ -951,12 +951,12 @@ public void mapPartitions() {
     JavaRDD<Integer> partitionSums = rdd.mapPartitions(
       new FlatMapFunction<Iterator<Integer>, Integer>() {
         @Override
-        public Iterable<Integer> call(Iterator<Integer> iter) {
+        public Iterator<Integer> call(Iterator<Integer> iter) {
           int sum = 0;
           while (iter.hasNext()) {
             sum += iter.next();
           }
-          return Collections.singletonList(sum);
+          return Collections.singletonList(sum).iterator();
         }
     });
     Assert.assertEquals("[3, 7]", partitionSums.collect().toString());
@@ -1367,8 +1367,8 @@ public void zipPartitions() {
     FlatMapFunction2<Iterator<Integer>, Iterator<String>, Integer> sizesFn =
       new FlatMapFunction2<Iterator<Integer>, Iterator<String>, Integer>() {
         @Override
-        public Iterable<Integer> call(Iterator<Integer> i, Iterator<String> s) {
-          return Arrays.asList(Iterators.size(i), Iterators.size(s));
+        public Iterator<Integer> call(Iterator<Integer> i, Iterator<String> s) {
+          return Arrays.asList(Iterators.size(i), Iterators.size(s)).iterator();
         }
       };
 

docs/streaming-programming-guide.md

@@ -165,8 +165,8 @@ space into words.
 // Split each line into words
 JavaDStream<String> words = lines.flatMap(
   new FlatMapFunction<String, String>() {
-    @Override public Iterable<String> call(String x) {
-      return Arrays.asList(x.split(" "));
+    @Override public Iterator<String> call(String x) {
+      return Arrays.asList(x.split(" ")).iterator();
     }
   });
 {% endhighlight %}