簡體   English   中英

以更實用的方式對目錄進行分組

[英]Grouping of directories in a more functional way

有人能給我一些關於如何將這個Scala代碼(Java命令方式)轉換為更實用的方法的見解嗎?

它是將目錄分組,總共4GB,將目錄/文件細分為4組,每組限制為1GB,將每個組添加到列表中。

object MyMapper {
  def main(args: Array[String]) {
    var map: Map[String, Long] = Map[String, Long]();
    map += ("a" -> 5);
    map += ("b" -> 5);
    map += ("c" -> 3);
    map += ("d" -> 2);
    map += ("e" -> 4);
    map += ("f" -> 2);
    map += ("g" -> 1);

    limitMapper(map)
  }

  def limitMapper(map: Map[String, Long]): Unit = {
    var offset = 0L;

    var result = Vector[Vector[String]]()
    var list = Vector[String]()
    val iterator: Iterator[(String, Long)] = map.toVector.iterator
    val LIMIT: Int = 10
    while (iterator.hasNext) {
      val (path, size) = iterator.next()
      println((path, size))

      offset += size
      if (offset > LIMIT) {
        offset = size;
        result = result :+ list
        list = Vector[String]()
      }

      list = list :+ path
    }

    //add remaining paths
    if (list.size > 0) {
      result = result :+ list
      list = Vector[String]()
    }

    println(result)
    println(s"LIMIT = $LIMIT")


  }
}


Output:
(e,4)
(f,2)
(a,5)
(b,5)
(g,1)
(c,3)
(d,2)
Vector(Vector(e, f), Vector(a, b), Vector(g, c, d))

我不確定這是不是最好的方法。 但我會做類似的事情:

val map = ...
val limit = 1073741824L

val res = map.foldLeft(Vector(Vector[String]())){ case (result, (path, size)) => 
  if(result.last.map(map(_)).sum + size > limit) {
    result :+ Vector(path)
  } else result.updated(result.size - 1, result.last :+ path)
}

這是一個可能的解決方案:

val (_, lastList, otherLists) = map.foldLeft(0L, List[String](), List[List[String]]()) {
  case ((soFar, newList, oldLists), (path, size)) =>
    if (soFar + size > limit)
      (size, List(path), newList :: oldLists)
    else
      (soFar + size, path :: newList, oldLists)
}
lastList :: otherLists

暫無
暫無

聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.

 
粵ICP備18138465號  © 2020-2024 STACKOOM.COM