Take this example:

- for idiom :

List list = Arrays.asList(“hello”, “world”);
for (String s : list) {
System.out.println(s);
}
String name = “World”;
String str = “Hello, ” + name;
System.out.println(str);

- while idom:

List list = Arrays.asList(“hello”, “world”);
Iterator it = list.iterator();
while (it.hasNext()) {
String s = it.next();
System.out.println(s);
}
String name = “World”;
String str = “Hello, ” + name;
System.out.println(str);

In that exemple, the generated bycode will differ, the while idom will require more memory on its stack frame for local variables than the for idom. The for idom uses 3 memory addresses and the while uses 4, because it cannot reuse the slot that was used for “it”.

It’s really a small difference though, the most important advantage of the for idom is really at compile time, that you you can safely reuse the “it” name for future variables and have it isolated from the one used by the control loop.

javap -c -classpath ~/project test.TestIterator

The -c does disassembly (shows bytecode). Otherwise, it will print method signatures by default. There are other flags as help if you run javap -help.

The tool I actually used is the fantastic ASM ByteCode Outline plugin for Eclipse. It lets you look at the bytecode for a selected method and can even show you the code to generate that bytecode using ASM, which is fairly indispensable when working at Terracotta. :)