Java迭代器的怪异行为

Question

I've encounter a weird problem when using Java iterators.

In a certain function a receive an iterable object called filelist , and I perform the following :

System.out.println("First iteration:");
for(Text t : filelist) System.out.println(t);
System.out.println("Second iteration:");
for(Text t : filelist) System.out.println(t);

and the output is:

First iteration:

file2.txt

file1.txt

file1.txt

Second iteration:

filelist is of type Iterable<Text> . I'm working with the Hadoop map/reduce framework.

So my question is, why is file list empty in the second loop, when I didn't change it in the first one?

Answer 1

The Iterable might return the iterator that might be non-reiterable. Nothing in it's interface says it can't. So in your case it seems that Iterable is designed by library authors to be only used once. If you want to iterate second time, you need to store values in your own structure. Ie you can first do

List<Text> myList = new ArrayList(filelist)