1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
|
package spark.rdd
import java.text.SimpleDateFormat
import java.util.Date
import org.apache.hadoop.conf.Configuration
import org.apache.hadoop.io.Writable
import org.apache.hadoop.mapreduce._
import spark.{Dependency, RDD, SerializableWritable, SparkContext, Split, TaskContext}
private[spark]
class NewHadoopSplit(rddId: Int, val index: Int, @transient rawSplit: InputSplit with Writable)
extends Split {
val serializableHadoopSplit = new SerializableWritable(rawSplit)
override def hashCode(): Int = (41 * (41 + rddId) + index)
}
class NewHadoopRDD[K, V](
sc : SparkContext,
inputFormatClass: Class[_ <: InputFormat[K, V]],
keyClass: Class[K],
valueClass: Class[V],
@transient conf: Configuration)
extends RDD[(K, V)](sc, Nil)
with HadoopMapReduceUtil {
// A Hadoop Configuration can be about 10 KB, which is pretty big, so broadcast it
val confBroadcast = sc.broadcast(new SerializableWritable(conf))
// private val serializableConf = new SerializableWritable(conf)
private val jobtrackerId: String = {
val formatter = new SimpleDateFormat("yyyyMMddHHmm")
formatter.format(new Date())
}
@transient private val jobId = new JobID(jobtrackerId, id)
@transient private val splits_ : Array[Split] = {
val inputFormat = inputFormatClass.newInstance
val jobContext = newJobContext(conf, jobId)
val rawSplits = inputFormat.getSplits(jobContext).toArray
val result = new Array[Split](rawSplits.size)
for (i <- 0 until rawSplits.size) {
result(i) = new NewHadoopSplit(id, i, rawSplits(i).asInstanceOf[InputSplit with Writable])
}
result
}
override def getSplits = splits_
override def compute(theSplit: Split, context: TaskContext) = new Iterator[(K, V)] {
val split = theSplit.asInstanceOf[NewHadoopSplit]
val conf = confBroadcast.value.value
val attemptId = new TaskAttemptID(jobtrackerId, id, true, split.index, 0)
val hadoopAttemptContext = newTaskAttemptContext(conf, attemptId)
val format = inputFormatClass.newInstance
val reader = format.createRecordReader(
split.serializableHadoopSplit.value, hadoopAttemptContext)
reader.initialize(split.serializableHadoopSplit.value, hadoopAttemptContext)
// Register an on-task-completion callback to close the input stream.
context.addOnCompleteCallback(() => reader.close())
var havePair = false
var finished = false
override def hasNext: Boolean = {
if (!finished && !havePair) {
finished = !reader.nextKeyValue
havePair = !finished
}
!finished
}
override def next: (K, V) = {
if (!hasNext) {
throw new java.util.NoSuchElementException("End of stream")
}
havePair = false
return (reader.getCurrentKey, reader.getCurrentValue)
}
}
override def getPreferredLocations(split: Split) = {
val theSplit = split.asInstanceOf[NewHadoopSplit]
theSplit.serializableHadoopSplit.value.getLocations.filter(_ != "localhost")
}
}
|
