Exercises in programming style with Scala - scala

Question

I've started to read the "Exercises in programming style" book recently and one of the tasks there is to implement each programming style in a language of your choice. I decided to go with Scala (I'm fairly new to it) and I'm already stuck with the first "good old school" style. The constraints are:
Very small amount of primary memory, typically orders of magnitude smaller than the data that needs to be processed/generated. (The example sets the limit to 1024 cells)
No labels -- i.e. no variable names or tagged memory addresses. All we have is memory that is addressable with numbers.
Original example (which reads a file line by line and counts the words) is in Python and goes like this:
data = []
data.append([]) # data[1] is line (max 80 characters)
data.append(None) # data[2] is index of the start_char of word
data.append(0) # data[3] is index on characters, i = 0
data.append(False) # data[4] is flag indicating if word was found
data.append('') # data[5] is the word
data.append('') # data[6] is word,NNNN
data.append(0) # data[7] is frequency
...
f = open(sys.argv[1])
# Loop over input file's lines
while True:
data[1] = [f.readline()]
...
So we see there are some variables (f and data) but the main idea is to keep it to a minimum and use python array as a bunch of "memory addresses".
Is it even possible to implement old-school-programming style (no variable names or tagged memory addresses) in Scala? Specifically is there a way to avoid "line" variable when reading file content?
for (line <- Source.fromFile("example.txt").getLines) {
println(line.toUpperCase)
}
Reading the file content into an array similar to the original example doesn't work because it's doesn't have an extractor (value data is not a case class, nor does it have an unapply/unapplySeq member).
P.S. I'm very well aware of the fact that the whole task is probably a 5-liner in Scala but that's not the point.

Answer 1

Sure you can abstain from introducing variables besides the data-array (and solve the problem imperative-style). Simply put everything into your array instead of assigning it to a local variable.
Obviously, the code will be a nightmare, because the array won't be typed and you won't have any meaningful names for any of your data, but I'm assuming that's what you're aiming for with this exercise.
import scala.io.Source
/**
* data 0 : file as line iterator
* data 1 : index of first unused data cell
* data 2 : current line
* data 3 : index of the first letter of the current word
* data 4 : index of the last letter of the current word
* data 5 : current word
* data 6 : temp index to find already initialized words
* data 7 : flag: Word found
* data 8, 10, 12, ... words
* data 9, 11, 13, ... frequencies
*/
object GoodOldSchool {
def main(args: Array[String]): Unit = {
val data: Array[Any] = new Array[Any](1024)
data(0) = Source.fromFile(args(0)).getLines()
data(1) = 8 // first free cell
while (data(0).asInstanceOf[Iterator[String]].hasNext) {
data(2) = data(0).asInstanceOf[Iterator[String]].next()
data(3) = 0 // index first letter of current word
data(4) = 0 // index last letter of current word
// find index last letter of current word
while (data(4).asInstanceOf[Int] < data(2).asInstanceOf[String].length) {
// find the next space (we ignore punctuation)
while (data(4).asInstanceOf[Int] < data(2).asInstanceOf[String].length && data(2).asInstanceOf[String].charAt(data(4).asInstanceOf[Int]) != ' ') {
data(4) = data(4).asInstanceOf[Int] + 1
}
data(5) = data(2).asInstanceOf[String].substring(data(3).asInstanceOf[Int], data(4).asInstanceOf[Int]) // current word
data(6) = 8 // cell index
data(7) = false // word already found
8 until data(1).asInstanceOf[Int] by 2 foreach { _ =>
// Here, we do a case-sensitive word comparison
if (data(5) == data(data(6).asInstanceOf[Int])) {
data(data(6).asInstanceOf[Int] + 1) = data(data(6).asInstanceOf[Int] + 1).asInstanceOf[Int] + 1 // increment frequency
data(7) = true
}
data(6) = data(6).asInstanceOf[Int] + 2
}
if (data(7) == false) {
// create new frequency, because word was not discovered before
data(data(1).asInstanceOf[Int]) = data(5) // set word
data(data(1).asInstanceOf[Int] + 1) = 1 // set frequency
data(1) = data(1).asInstanceOf[Int] + 2 // used up two cells, update index of next free cell
}
// move to next word
data(3) = data(4).asInstanceOf[Int] + 1
data(4) = data(3)
}
}
data foreach println // let's have a look at our result
}
}

Answer 2

To get the total words count in the given file, the following scala code can be used:
Source.fromFile("example.txt")
.getLines.map { line => line.trim.split(" ").length}
.reduceLeft { _ + _ }