Problem Statement

When sorting records in a table by a particular column or field, a stable sort will always retain the relative order of records that have the same key.

In this table of countries and cities, a stable sort on the second column, the cities, would keep the   US Birmingham   above the   UK Birmingham. (Although an unstable sort might, in this case, place the   US Birmingham   above the   UK Birmingham,   a stable sort routine would guarantee it). Similarly, stable sorting on just the first column would generate UK London as the first item and US Birmingham as the last item   (since the order of the elements having the same first word –   UK or US   – would be maintained).

(This Wikipedia table shows the stability of some common sort routines).

Let's start with the solution:

scala> val list = List((1, 'c'), (1, 'b'), (2, 'a'))
list: List[(Int, Char)] = List((1,c), (1,b), (2,a))

scala> val srt1 = list.sortWith(_._2 < _._2)
srt1: List[(Int, Char)] = List((2,a), (1,b), (1,c))

scala> val srt2 = srt1.sortBy(_._1) // Ordering[Int] is implicitly defined
srt2: List[(Int, Char)] = List((1,b), (1,c), (2,a))

scala> val cities = """
     | |UK  London
     | |US  New York
     | |US  Birmingham
     | |UK  Birmingham
     | |""".stripMargin.lines.filterNot(_ isEmpty).toSeq
cities: Seq[String] = ArrayBuffer(UK  London, US  New York, US  Birmingham, UK  Birmingham)

scala> cities.sortBy(_ substring 4)
res47: Seq[String] = ArrayBuffer(US  Birmingham, UK  Birmingham, UK  London, US  New York)