Reputation: 4877
I would like to read a large text file that contains two columns of integers. For each consecutive stretch where the first column is constant, I want to collect the second column as a list. As an example, the following sample
1 2
1 1
1 2
4 3
4 8
1 5
8 2
8 2
8 7
should produce the lists
2,1,2
3,8
5
2,2,7
What is a proper way to achieve this in Scala?
In particular, it would be very nice to have a "lazy" solution, so that I can treat each list without having to first load the entire file into memory.
Upvotes: 1
Views: 1374
Reputation: 161
val nums = """(\d+) (\d+)""".r
val tuples = io.Source.fromFile("list.txt").getLines collect {
case nums(label, num) => (label.toInt -> num.toInt)
}
def toList(tuples: Iterator[(Int, Int)]): Iterator[Seq[Int]] = {
if(!tuples.hasNext) Iterator.empty
else {
val (label, num) = tuples.next
val (succ, rest) = tuples.span(_._1 == label)
Iterator(num :: succ.map(_._2).toList) ++ toList(rest)
}
}
toList(tuples)foreach(println)
Upvotes: 1
Reputation: 24423
The simplest way is to use scala.io.Source
to read the file line by line. With getLines
, you can retrieve an Iterator[String]
over which you can map to split the lines and convert them to ints like this:
val intPairs = Source.fromFile("/path/to/file").getLines.map { line =>
line.split(" ").take(2).map(_.toInt)
}
I leave the grouping of consecutive lines as an exercise for you.
Upvotes: 1