Streams

Scala's streams are lazy lists, an infinite sequence of elements. For example, just like we zip two lists together, we could zip two streams too:

scala> def zip(p: Stream[Int], q: Stream[Int]): Stream[(Int, Int)] = 
     | (p.head, q.head) #:: zip(p.tail, q.tail) 
zip: (p: Stream[Int], q: Stream[Int])Stream[(Int, Int)] 
 
scala> val r = zip(Stream.from(9), Stream.from(10)) 
r: Stream[(Int, Int)] = Stream((9,10), ?) 

As you can see, the type of r variable is another stream. Also, the stream has a head element, a pair (9,10), and a function (also called a thunk). The function (in this case, zip) is called to produce successive elements.

As you can see, the tail is a question, meaning it is not yet computed; therefore, it is not yet known.

Why are streams called lazy? Note the definition of zip, which is defined recursively. There is no condition related to termination; the calls just recur. A recursive call is not evaluated right away but only when it is needed.

The idea is to force...