Radix排序的字符串数组错误输出2个字？

Question

在这个程序中，任务是重新排列单词数组中给定的字符串以按长度排序，然后使用基数排序方法按字母顺序排序。例如，给定输入：

“bar”，“cat”，“apple”，“bug”，“cog”，“caper”，“antler”，“ankle”，“bear”

正确的输出应该是：bar bug cat cog bear ankle apple caper antler

到目前为止，我给出的目前代码是：bar bug cat cog bear ankle antler caper apple

我的基数排序算法究竟在哪里出错了？为什么“鹿茸”，长度最大的字符串不在输出的末尾？

这是我目前的代码：

import java.util.ArrayList;

public class RadixSort 
{
    public static void main(String[] args) 
    {
        String[] words = {"bar", "cat", "apple", "bug", "cog", "caper", "antler", "ankle", "bear"};

        System.out.print("Before Radix Sort: ");
        for (String h : words)
            System.out.print(h + " ");

        strRadixSort(words);

        System.out.print("\n\nAfter Radix Sort: ");
        for (String h : words)
            System.out.print(h + " ");

        System.out.print("\n\nWhat the output should look like: ");
        System.out.print("bar bug cat cog bear ankle apple caper antler");

    }

    public static void strRadixSort(String[] words) 
    {
      ArrayList<String>[] Letterbuckets = new ArrayList[26];

      for (int i = 0; i < Letterbuckets.length; i++) 
      {
          Letterbuckets[i] = new ArrayList<String>();
      }

      boolean exit = false;
      int index = 1;
      int x = 0;

      while (exit == false) 
      {  
        exit = true;

        for (String i : words) 
        {
          int curIndex= i.length() - index; //start at the last char in string i

          if(curIndex < 0)
          {
              break; //We are done evaluating this word, go to the next word
          }

          x = determineLetterBucket(i, curIndex);

          Letterbuckets[x].add(i);
          x = 0;

          if (exit == true && curIndex > 0) 
          {
              exit = false;
          }
        }

        //Rearranges the order of the words in the
        //word array and clears all letter buckets
        //to prepare for the next pass.
        int a = 0;

        for(int k = 0; k < 26; k++) 
        {
          for (String i : Letterbuckets[k]) 
          {
            words[a++] = i;
          }

          Letterbuckets[k].clear();
        }

        //Moves to the next least significant letter
        index += 1;
      }
    }

    //Determines which number bucket the letter should go into
    //e.g. In the string "abc" the char at index 2 is "c",
    //therefore this would mean this word would go into bucket 2
    //which can also be represented as the "c" bucket.
    public static int determineLetterBucket(String currentWord, int currentIndex)
    {
        int bucket = 0;
        int j = 0;

        for(int k = 97; k < 123; k++)
        {
            if(currentWord.charAt(currentIndex) == k)
                bucket = j;
            else
                j++;
        }

        return bucket;
    }
}