Streams - Java 8 Lambdas - page 17

There are several problems with this approach, though. It involves a lot of boilerplate code
that needs to be written every time you want to iterate over the collection. It's also hard to
write a parallel version of this for loop. You would need to rewrite every for loop individu-
ally in order to make them operate in parallel.

Finally, the code here doesn't fluently convey the intent of the programmer. The boilerplate
for loop structure obscures meaning; to understand anything we must read though the body
of the loop. For a single for loop, doing this isn't too bad, but when you have a large code
base full of them it becomes a burden (especially with nested loops).

Looking under the covers a little bit, the for loop is actually syntactic sugar that wraps up
the iteration and hides it. It's worth taking a moment to look at what's going on under the
hood here. The first step in this process is a call to the iterator method, which creates a
new Iterator object in order to control the iteration process. We call this external iteration .
The iteration then proceeds by explicitly calling the hasNext and next methods on this
Iterator . Example 3-2 demonstrates the expanded code in full, and Figure 3-1 shows the
pattern of method calls that happen.

Example 3-2. Counting London-based artists using an iterator
int int count = 0 ;

Iterator < Artist > iterator = allArtists . iterator ();
while

while ( iterator . hasNext ()) {
Artist artist = iterator . next ();
iif ( artist . isFrom ( "London" )) {
count ++;

}

}