blob: 48b5018dddbdd1326a7e7ff7bda771bc44c76ce0 (
plain) (
blame)
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
|
package spark.util
/** Provides a basic/boilerplate Iterator implementation. */
private[spark] abstract class NextIterator[U] extends Iterator[U] {
private var gotNext = false
private var nextValue: U = _
private var closed = false
protected var finished = false
/**
* Method for subclasses to implement to provide the next element.
*
* If no next element is available, the subclass should set `finished`
* to `true` and may return any value (it will be ignored).
*
* This convention is required because `null` may be a valid value,
* and using `Option` seems like it might create unnecessary Some/None
* instances, given some iterators might be called in a tight loop.
*
* @return U, or set 'finished' when done
*/
protected def getNext(): U
/**
* Method for subclasses to implement when all elements have been successfully
* iterated, and the iteration is done.
*
* <b>Note:</b> `NextIterator` cannot guarantee that `close` will be
* called because it has no control over what happens when an exception
* happens in the user code that is calling hasNext/next.
*
* Ideally you should have another try/catch, as in HadoopRDD, that
* ensures any resources are closed should iteration fail.
*/
protected def close()
/**
* Calls the subclass-defined close method, but only once.
*
* Usually calling `close` multiple times should be fine, but historically
* there have been issues with some InputFormats throwing exceptions.
*/
def closeIfNeeded() {
if (!closed) {
close()
closed = true
}
}
override def hasNext: Boolean = {
if (!finished) {
if (!gotNext) {
nextValue = getNext()
if (finished) {
closeIfNeeded()
}
gotNext = true
}
}
!finished
}
override def next(): U = {
if (!hasNext) {
throw new NoSuchElementException("End of stream")
}
gotNext = false
nextValue
}
}
|
