在长文件列表中查找 3 个最近修改的文件

Question

我有一个文件列表，我想对它进行排序并提取上次修改的前 3 个文件。

约束：由于下游应用程序的兼容性问题，我无法使用 Java 7

我目前的选择

解决方案1

File[] files = directory.listFiles();    
Arrays.sort(files, new Comparator<File>(){
    public int compare(File f1, File f2)
    {
        return Long.valueOf(f1.lastModified()).compareTo(f2.lastModified());
    } });

解决方案2

public static void sortFilesDesc(File[] files) {
  Arrays.sort(files, new Comparator() {
    public int compare(Object o1, Object o2) {
      if ((File)o1).lastModified().compareTo((File)o2).lastModified()) {
        return -1;
      } else if (((File) o1).lastModified() < ((File) o2).lastModified()) {
        return +1;
      } else {
        return 0;
      }
    }
  });
}

问题

以上两种解决方案需要更多的时间来执行和内存。我的文件列表包含大约 300 个 tar 文件，每个文件大小为 200MB。所以它消耗更多的时间和内存。

有没有办法有效地处理这个问题？

每个比较操作都使用一个内存较高的文件对象，有没有办法释放内存并有效地处理它？

Answer 1

You can do it much faster.

Arrays.sort(...) uses "quick sort", which takes ~ n * ln(n) operations.

This example takes only one iteration trough the whole array, which is ~ n operations.

public static void sortFilesDesc(File[] files) {        
    File firstMostRecent = null;
    File secondMostRecent = null;
    File thirdMostRecent = null;
    for (File file : files) {
        if ((firstMostRecent == null)
                || (firstMostRecent.lastModified() < file.lastModified())) {
            thirdMostRecent = secondMostRecent;
            secondMostRecent = firstMostRecent;             
            firstMostRecent = file;
        } else if ((secondMostRecent == null)
                || (secondMostRecent.lastModified() < file.lastModified())) {
            thirdMostRecent = secondMostRecent;
            secondMostRecent = file;
        } else if ((thirdMostRecent == null)
                || (thirdMostRecent.lastModified() < file.lastModified())) {
            thirdMostRecent = file;
        }
    }
} 

On small numbers of files you won't see much difference, but even for tens of files the difference will be significant, for bigger numbers - dramatic.

The code to check the algorithm (please put in a correct files structure):

package com.hk.basicjava.clasload.tests2;

import java.io.File;
import java.util.Date;


class MyFile extends File {

    private long time = 0; 

    public MyFile(String name, long timeMills) {
        super(name);
        time = timeMills;
    }
    @Override
    public long lastModified() {
        return time;
    }
}

public class Files {

    /**
     * @param args
     */
    public static void main(String[] args) {

        File[] files = new File[5]; 
        files[0] = new MyFile("File1", new Date(2013,1,15, 7,0).getTime());
        files[1] = new MyFile("File2", new Date(2013,1,15, 7,40).getTime());
        files[2] = new MyFile("File3", new Date(2013,1,15, 5,0).getTime());
        files[3] = new MyFile("File4", new Date(2013,1,15, 10,0).getTime());
        files[4] = new MyFile("File5", new Date(2013,1,15, 4,0).getTime());
        sortFilesDesc(files);
    }

    public static void sortFilesDesc(File[] files) {        
        File firstMostRecent = null;
        File secondMostRecent = null;
        File thirdMostRecent = null;
        for (File file : files) {
            if ((firstMostRecent == null)
                    || (firstMostRecent.lastModified() < file.lastModified())) {
                thirdMostRecent = secondMostRecent;
                secondMostRecent = firstMostRecent;             
                firstMostRecent = file;
            } else if ((secondMostRecent == null)
                    || (secondMostRecent.lastModified() < file.lastModified())) {
                thirdMostRecent = secondMostRecent;
                secondMostRecent = file;
            } else if ((thirdMostRecent == null)
                    || (thirdMostRecent.lastModified() < file.lastModified())) {
                thirdMostRecent = file;
            }
        }
        System.out.println("firstMostRecent : " + firstMostRecent.getName());
        System.out.println("secondMostRecent : " + secondMostRecent.getName());
        System.out.println("thirdMostRecent : " + thirdMostRecent.getName());
    } 

}