一个几乎相同的问题出现了Code Jam问题，这里有一个解决方案:

http://code.google.com/codejam/contest/dashboard?c=186264#s=a&a=1

下面用一个例子总结一下这个方法:

34722641

A.将数字序列分成两部分，使右边的部分尽可能长，同时保持递减顺序:

34722 641

(如果整个数字是递减的，就没有比这个数字更大的数字了。)

在这一点上，你知道没有从左边开始的更大的数了，因为右边的剩余数字已经尽可能大了。

责任。选择第一个序列的最后一位:

3472(2) 641

B.2。找出第二个序列中比它大的最小的数字:

3472(2) 6(4)1

你要做的就是找到左边可能的最小增量。

B.3。交换:

3472(2) 6(4)1
->
3472(4) 6(2)1
->
34724 621

C.将第二个序列按递增顺序排序:

34724 126

d .完成了!

34724126

你把这个数字分开，这样你就知道没有更大的数字具有相同的左边部分，你把左边部分增加了尽可能小的量，你让剩下的右边部分尽可能小，所以你可以确保这个新数字是用相同的数字集合可以得到的最小的更大的数字。

I didn't know anything about the brute force algorithm when answering this question, so I approached it from another angle. I decided to search the entire range of possible solutions that this number could possibly be rearranged into, starting from the number_given+1 up to the max number available (999 for a 3 digit number, 9999 for 4 digits, etc.). I did this kind of like finding a palindrome with words, by sorting the numbers of each solution and comparing it to the sorted number given as the parameter. I then simply returned the first solution in the array of solutions, as this would be the next possible value.

下面是我的Ruby代码:

def PermutationStep(num)

    a = []
    (num.to_s.length).times { a.push("9") }
    max_num = a.join('').to_i
    verify = num.to_s.split('').sort
    matches = ((num+1)..max_num).select {|n| n.to_s.split('').sort == verify }

    if matches.length < 1
      return -1
    else
      matches[0]
    end
end

这是个很有趣的问题。

这是我的java版本。在我检查其他贡献者的评论之前，从弄清楚模式到完全完成代码，我花了大约3个小时。很高兴看到我的想法和别人一样。

O (n)的解决方案。老实说，如果时间只有15分钟，并且要求在白板上完成完整的代码，我将会失败。

以下是我的解决方案的一些有趣点:

避免任何排序。 完全避免字符串操作 实现O(logN)空间复杂度

我在代码中添加了详细注释，并在每个步骤中添加了大O。

  public int findNextBiggestNumber(int input  )   {
    //take 1358642 as input for example.
    //Step 1: split the whole number to a list for individual digital   1358642->[2,4,6,8,5,3,1]
    // this step is O(n)
    int digitalLevel=input;

    List<Integer> orgNumbersList=new ArrayList<Integer>()   ;

    do {
        Integer nInt = new Integer(digitalLevel % 10);
        orgNumbersList.add(nInt);

        digitalLevel=(int) (digitalLevel/10  )  ;


    } while( digitalLevel >0)    ;
    int len= orgNumbersList.size();
    int [] orgNumbers=new int[len]  ;
    for(int i=0;i<len;i++){
        orgNumbers[i ]  =  orgNumbersList.get(i).intValue();
    }
    //step 2 find the first digital less than the digital right to it
    // this step is O(n)


    int firstLessPointer=1;
    while(firstLessPointer<len&&(orgNumbers[firstLessPointer]>orgNumbers[ firstLessPointer-1 ])){
        firstLessPointer++;
    }
     if(firstLessPointer==len-1&&orgNumbers[len-1]>=orgNumbers[len-2]){
         //all number is in sorted order like 4321, no answer for it, return original
         return input;
     }

    //when step 2 step finished, firstLessPointer  pointing to number 5

     //step 3 fristLessPointer found, need to find  to  first number less than it  from low digital in the number
    //This step is O(n)
    int justBiggerPointer=  0 ;

    while(justBiggerPointer<firstLessPointer&& orgNumbers[justBiggerPointer]<orgNumbers[firstLessPointer]){
        justBiggerPointer++;
    }
    //when step 3 finished, justBiggerPointer  pointing to 6

    //step 4 swap the elements  of justBiggerPointer and firstLessPointer .
    // This  is O(1) operation   for swap

   int tmp=  orgNumbers[firstLessPointer] ;

    orgNumbers[firstLessPointer]=  orgNumbers[justBiggerPointer]  ;
     orgNumbers[justBiggerPointer]=tmp ;


     // when step 4 finished, the list looks like        [2,4,5,8,6,3,1]    the digital in the list before
     // firstLessPointer is already sorted in our previous operation
     // we can return result from this list  but  in a differrent way
    int result=0;
    int i=0;
    int lowPointer=firstLessPointer;
    //the following pick number from list from  the position just before firstLessPointer, here is 8 -> 5 -> 4 -> 2
    //This Operation is O(n)
    while(lowPointer>0)        {
        result+= orgNumbers[--lowPointer]* Math.pow(10,i);
        i++;
    }
    //the following pick number from list   from position firstLessPointer
    //This Operation is O(n)
    while(firstLessPointer<len)        {
        result+= orgNumbers[firstLessPointer++ ]* Math.pow(10,i);
        i++;
    }
     return  result;

}

下面是在Intellj中运行的结果:

959879532-->959892357
1358642-->1362458
1234567-->1234576
77654321-->77654321
38276-->38627
47-->74

至少，这里有几个基于字符串的暴力解决方案的例子，你应该能够马上想到:

38276中的数字排序为23678

38627排序的数字列表是23678

蛮力增量，排序和比较

沿着蛮力解决方案将转换为字符串 然后用这些数字强行找出所有可能的数字。

从它们中创建int，把它们放在一个列表中并排序， 获取目标条目之后的下一个条目。

如果你花了30分钟在这个问题上，却没有想出一个蛮力的方法，我也不会雇用你。

在商业世界中，一个不优雅、缓慢和笨拙但能完成工作的解决方案总是比没有解决方案更有价值，事实上，这几乎描述了所有不优雅、缓慢和笨拙的商业软件。

 private static int GetNextHigherNumber(int num)
        {
            //given 38276 return 38627

            string numberstring = num.ToString();

            char[] sNum = numberstring.ToCharArray();

            for (int i = sNum.Length - 1; i > 0; i--)
            {
                for (int j = i - 1; j > 0; j--)
                {
                    if (sNum[i] > sNum[j])
                    {
                        for (int x = i; x > j; x--)
                        {
                            char chr = sNum[x]; 
                            sNum[x] = sNum[x - 1];
                            sNum[x - 1] = chr;
                        }

                        i = 0;
                        break;
                    }
                }
            }

            numberstring = string.Empty;
            for(int x= 0 ; x<sNum.Length;x++)
            {
                numberstring += sNum[x].ToString();
            }

            return Convert.ToInt32(numberstring);
        }

我们需要找到最右边的0位，后面是1，然后将最右边的0位翻转为1。

例如，我们的输入是487，也就是二进制的111100111。

我们把后面有1的0往右翻转最多

所以我们得到 111101111

但是现在我们多了一个1，少了一个0，所以我们减少了右边1的个数 位增加1，并将0位的no增加1，得到

111101011 -二进制491

int getNextNumber(int input)
{
    int flipPosition=0;
    int trailingZeros=0;
    int trailingOnes=0;
    int copy = input;

    //count trailing zeros
    while(copy != 0 && (copy&1) == 0 )
    {
        ++trailingZeros;

        //test next bit
        copy = copy >> 1;
    }

    //count trailing ones
    while(copy != 0 && (copy&1) == 1 )
    {
        ++trailingOnes;

        //test next bit
        copy = copy >> 1;
    }

    //if we have no 1's (i.e input is 0) we cannot form another pattern with 
    //the same number of 1's which will increment the input, or if we have leading consecutive
    //ones followed by consecutive 0's up to the maximum bit size of a int
    //we cannot increase the input whilst preserving the original no of 0's and
    //1's in the bit pattern
    if(trailingZeros + trailingOnes  == 0 || trailingZeros + trailingOnes == 31)
        return -1;

    //flip first 0 followed by a 1 found from the right of the bit pattern
    flipPosition = trailingZeros + trailingOnes+1;
    input |= 1<<(trailingZeros+trailingOnes);

    //clear fields to the right of the flip position
    int mask = ~0 << (trailingZeros+trailingOnes);
    input &= mask;

    //insert a bit pattern to the right of the flip position that will contain
    //one less 1 to compensate for the bit we switched from 0 to 1
    int insert = flipPosition-1;
    input |= insert;

    return input;
}