diff --git a/core/src/main/scala/org/apache/spark/SparkContext.scala b/core/src/main/scala/org/apache/spark/SparkContext.scala index b23accbbb9410..28a865c0ad3b5 100644 --- a/core/src/main/scala/org/apache/spark/SparkContext.scala +++ b/core/src/main/scala/org/apache/spark/SparkContext.scala @@ -37,6 +37,7 @@ import org.apache.mesos.MesosNativeLibrary import org.apache.spark.broadcast.Broadcast import org.apache.spark.deploy.{LocalSparkCluster, SparkHadoopUtil} +import org.apache.spark.input.WholeTextFileInputFormat import org.apache.spark.partial.{ApproximateEvaluator, PartialResult} import org.apache.spark.rdd._ import org.apache.spark.scheduler._ @@ -371,6 +372,39 @@ class SparkContext( minSplits).map(pair => pair._2.toString) } + /** + * Read a directory of text files from HDFS, a local file system (available on all nodes), or any + * Hadoop-supported file system URI. Each file is read as a single record and returned in a + * key-value pair, where the key is the path of each file, the value is the content of each file. + * + *
For example, if you have the following files: + * {{{ + * hdfs://a-hdfs-path/part-00000 + * hdfs://a-hdfs-path/part-00001 + * ... + * hdfs://a-hdfs-path/part-nnnnn + * }}} + * + * Do `val rdd = sparkContext.wholeTextFile("hdfs://a-hdfs-path")`, + * + *
then `rdd` contains + * {{{ + * (a-hdfs-path/part-00000, its content) + * (a-hdfs-path/part-00001, its content) + * ... + * (a-hdfs-path/part-nnnnn, its content) + * }}} + * + * @note Small files are perferred, large file is also allowable, but may cause bad performance. + */ + def wholeTextFiles(path: String): RDD[(String, String)] = { + newAPIHadoopFile( + path, + classOf[WholeTextFileInputFormat], + classOf[String], + classOf[String]) + } + /** * Get an RDD for a Hadoop-readable dataset from a Hadoop JobConf given its InputFormat and other * necessary info (e.g. file name for a filesystem-based dataset, table name for HyperTable), diff --git a/core/src/main/scala/org/apache/spark/api/java/JavaSparkContext.scala b/core/src/main/scala/org/apache/spark/api/java/JavaSparkContext.scala index e531a57aced31..6cbdeac58d5e2 100644 --- a/core/src/main/scala/org/apache/spark/api/java/JavaSparkContext.scala +++ b/core/src/main/scala/org/apache/spark/api/java/JavaSparkContext.scala @@ -154,6 +154,34 @@ class JavaSparkContext(val sc: SparkContext) extends JavaSparkContextVarargsWork */ def textFile(path: String, minSplits: Int): JavaRDD[String] = sc.textFile(path, minSplits) + /** + * Read a directory of text files from HDFS, a local file system (available on all nodes), or any + * Hadoop-supported file system URI. Each file is read as a single record and returned in a + * key-value pair, where the key is the path of each file, the value is the content of each file. + * + *
For example, if you have the following files:
+ * {{{
+ * hdfs://a-hdfs-path/part-00000
+ * hdfs://a-hdfs-path/part-00001
+ * ...
+ * hdfs://a-hdfs-path/part-nnnnn
+ * }}}
+ *
+ * Do `JavaPairRDD then `rdd` contains
+ * {{{
+ * (a-hdfs-path/part-00000, its content)
+ * (a-hdfs-path/part-00001, its content)
+ * ...
+ * (a-hdfs-path/part-nnnnn, its content)
+ * }}}
+ *
+ * @note Small files are perferred, large file is also allowable, but may cause bad performance.
+ */
+ def wholeTextFiles(path: String): JavaPairRDD[String, String] =
+ new JavaPairRDD(sc.wholeTextFiles(path))
+
/** Get an RDD for a Hadoop SequenceFile with given key and value types.
*
* '''Note:''' Because Hadoop's RecordReader class re-uses the same Writable object for each
diff --git a/core/src/main/scala/org/apache/spark/input/WholeTextFileInputFormat.scala b/core/src/main/scala/org/apache/spark/input/WholeTextFileInputFormat.scala
new file mode 100644
index 0000000000000..4887fb6b84eb2
--- /dev/null
+++ b/core/src/main/scala/org/apache/spark/input/WholeTextFileInputFormat.scala
@@ -0,0 +1,47 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.input
+
+import org.apache.hadoop.fs.Path
+import org.apache.hadoop.mapreduce.InputSplit
+import org.apache.hadoop.mapreduce.JobContext
+import org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat
+import org.apache.hadoop.mapreduce.RecordReader
+import org.apache.hadoop.mapreduce.TaskAttemptContext
+import org.apache.hadoop.mapreduce.lib.input.CombineFileRecordReader
+import org.apache.hadoop.mapreduce.lib.input.CombineFileSplit
+
+/**
+ * A [[org.apache.hadoop.mapreduce.lib.input.CombineFileInputFormat CombineFileInputFormat]] for
+ * reading whole text files. Each file is read as key-value pair, where the key is the file path and
+ * the value is the entire content of file.
+ */
+
+private[spark] class WholeTextFileInputFormat extends CombineFileInputFormat[String, String] {
+ override protected def isSplitable(context: JobContext, file: Path): Boolean = false
+
+ override def createRecordReader(
+ split: InputSplit,
+ context: TaskAttemptContext): RecordReader[String, String] = {
+
+ new CombineFileRecordReader[String, String](
+ split.asInstanceOf[CombineFileSplit],
+ context,
+ classOf[WholeTextFileRecordReader])
+ }
+}
diff --git a/core/src/main/scala/org/apache/spark/input/WholeTextFileRecordReader.scala b/core/src/main/scala/org/apache/spark/input/WholeTextFileRecordReader.scala
new file mode 100644
index 0000000000000..c3dabd2e79995
--- /dev/null
+++ b/core/src/main/scala/org/apache/spark/input/WholeTextFileRecordReader.scala
@@ -0,0 +1,72 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.input
+
+import com.google.common.io.{ByteStreams, Closeables}
+
+import org.apache.hadoop.io.Text
+import org.apache.hadoop.mapreduce.InputSplit
+import org.apache.hadoop.mapreduce.lib.input.CombineFileSplit
+import org.apache.hadoop.mapreduce.RecordReader
+import org.apache.hadoop.mapreduce.TaskAttemptContext
+
+/**
+ * A [[org.apache.hadoop.mapreduce.RecordReader RecordReader]] for reading a single whole text file
+ * out in a key-value pair, where the key is the file path and the value is the entire content of
+ * the file.
+ */
+private[spark] class WholeTextFileRecordReader(
+ split: CombineFileSplit,
+ context: TaskAttemptContext,
+ index: Integer)
+ extends RecordReader[String, String] {
+
+ private val path = split.getPath(index)
+ private val fs = path.getFileSystem(context.getConfiguration)
+
+ // True means the current file has been processed, then skip it.
+ private var processed = false
+
+ private val key = path.toString
+ private var value: String = null
+
+ override def initialize(split: InputSplit, context: TaskAttemptContext) = {}
+
+ override def close() = {}
+
+ override def getProgress = if (processed) 1.0f else 0.0f
+
+ override def getCurrentKey = key
+
+ override def getCurrentValue = value
+
+ override def nextKeyValue = {
+ if (!processed) {
+ val fileIn = fs.open(path)
+ val innerBuffer = ByteStreams.toByteArray(fileIn)
+
+ value = new Text(innerBuffer).toString
+ Closeables.close(fileIn, false)
+
+ processed = true
+ true
+ } else {
+ false
+ }
+ }
+}
diff --git a/core/src/test/java/org/apache/spark/JavaAPISuite.java b/core/src/test/java/org/apache/spark/JavaAPISuite.java
index c6b65c7348ae0..2372f2d9924a1 100644
--- a/core/src/test/java/org/apache/spark/JavaAPISuite.java
+++ b/core/src/test/java/org/apache/spark/JavaAPISuite.java
@@ -17,9 +17,7 @@
package org.apache.spark;
-import java.io.File;
-import java.io.IOException;
-import java.io.Serializable;
+import java.io.*;
import java.util.*;
import scala.Tuple2;
@@ -599,6 +597,32 @@ public void textFiles() throws IOException {
Assert.assertEquals(expected, readRDD.collect());
}
+ @Test
+ public void wholeTextFiles() throws IOException {
+ byte[] content1 = "spark is easy to use.\n".getBytes();
+ byte[] content2 = "spark is also easy to use.\n".getBytes();
+
+ File tempDir = Files.createTempDir();
+ String tempDirName = tempDir.getAbsolutePath();
+ DataOutputStream ds = new DataOutputStream(new FileOutputStream(tempDirName + "/part-00000"));
+ ds.write(content1);
+ ds.close();
+ ds = new DataOutputStream(new FileOutputStream(tempDirName + "/part-00001"));
+ ds.write(content2);
+ ds.close();
+
+ HashMap