АР
Size: a a a
АР
FT
val b=key.foreach(s=>s.filterNot(c=>c=='.'))
b.foreach(x=>println(x))
FT
val b=key.foreach(s=>s.filterNot(c=>c=='.'))
b.foreach(x=>println(x))
АР
АР
АР
АР
АР
GP
GP
GP
АР
val inputPath="alice.txt"
//val outputPath="alice-wordpunctcount"
val textFile = sc.textFile(inputPath)
val sorted = textFile.flatMap(line => line.split(" ")).
filter(x=> x.contains(",")||x.contains(".")||x.contains("?")||x.contains("!")||x.contains("-")||x.contains("`")||x.contains("\'")||x.contains(";")||x.contains(":")||x.contains("\"")).
map(word => (word, 1)).
reduceByKey(_ + _).
map(x => x.swap).
sortByKey(false).
map(x => x.swap)
val key=sorted.keys
key.foreach(s=>s.filterNot(c=>c=='.'))
key.foreach(x=>println(x))
АР
АР
FT
OI
АЖ
АЖ