使用itoa转换为以3为基数的字符串。去掉最后一个小调，转换回10进制。

// Note: itoa is non-standard but actual implementations
// don't seem to handle negative when base != 10.
int div3(int i) {
    char str[42];
    sprintf(str, "%d", INT_MIN); // Put minus sign at str[0]
    if (i>0)                     // Remove sign if positive
        str[0] = ' ';
    itoa(abs(i), &str[1], 3);    // Put ternary absolute value starting at str[1]
    str[strlen(&str[1])] = '\0'; // Drop last digit
    return strtol(str, NULL, 3); // Read back result
}

这是Python中的，基本上，字符串比较和一个状态机。

def divide_by_3(input):
  to_do = {}
  enque_index = 0
  zero_to_9 = (0, 1, 2, 3, 4, 5, 6, 7, 8, 9)
  leave_over = 0
  for left_over in (0, 1, 2):
    for digit in zero_to_9:
      # left_over, digit => enque, leave_over
      to_do[(left_over, digit)] = (zero_to_9[enque_index], leave_over)
      if leave_over == 0:
        leave_over = 1
      elif leave_over == 1:
        leave_over = 2
      elif leave_over == 2 and enque_index != 9:
        leave_over = 0
        enque_index = (1, 2, 3, 4, 5, 6, 7, 8, 9)[enque_index]
  answer_q = []
  left_over = 0
  digits = list(str(input))
  if digits[0] == "-":
    answer_q.append("-")
  digits = digits[1:]
  for digit in digits:
    enque, left_over = to_do[(left_over, int(digit))]
    if enque or len(answer_q):
      answer_q.append(enque)
  answer = 0
  if len(answer_q):
    answer = int("".join([str(a) for a in answer_q]))
  return answer

第一:

x/3 = (x/4) / (1-1/4)

然后求x/(1 - y)

x/(1-1/y)
  = x * (1+y) / (1-y^2)
  = x * (1+y) * (1+y^2) / (1-y^4)
  = ...
  = x * (1+y) * (1+y^2) * (1+y^4) * ... * (1+y^(2^i)) / (1-y^(2^(i+i))
  = x * (1+y) * (1+y^2) * (1+y^4) * ... * (1+y^(2^i))

y = 1/4:

int div3(int x) {
    x <<= 6;    // need more precise
    x += x>>2;  // x = x * (1+(1/2)^2)
    x += x>>4;  // x = x * (1+(1/2)^4)
    x += x>>8;  // x = x * (1+(1/2)^8)
    x += x>>16; // x = x * (1+(1/2)^16)
    return (x+1)>>8; // as (1-(1/2)^32) very near 1,
                     // we plus 1 instead of div (1-(1/2)^32)
}

虽然它使用了+，但有人已经实现了按位操作的add。

3以2为底等于11。

所以只要做长除法(就像中学那样)，以2 × 11为底。以2为底比以10为底更简单。

对于从最有效位开始的每个位位:

判断prefix是否小于11。

如果它是输出0。

如果不是输出1，则替换前缀位进行适当的更改。只有三种情况:

 11xxx ->    xxx    (ie 3 - 3 = 0)
100xxx ->   1xxx    (ie 4 - 3 = 1)
101xxx ->  10xxx    (ie 5 - 3 = 2)

所有其他前缀都不可达。

重复到最低位，你就完成了。

很有趣的是，没有人回答一个泛泛的划分:

/* For the given integer find the position of MSB */
int find_msb_loc(unsigned int n)
{
    if (n == 0)
        return 0;

    int loc = sizeof(n)  * 8 - 1;
    while (!(n & (1 << loc)))
        loc--;
    return loc;
}


/* Assume both a and b to be positive, return a/b */
int divide_bitwise(const unsigned int a, const unsigned int b)
{
    int int_size = sizeof(unsigned int) * 8;
    int b_msb_loc = find_msb_loc(b);

    int d = 0; // dividend
    int r = 0; // reminder
    int t_a = a;
    int t_a_msb_loc = find_msb_loc(t_a);
    int t_b = b << (t_a_msb_loc - b_msb_loc);

    int i;
    for(i = t_a_msb_loc; i >= b_msb_loc; i--)  {
        if (t_a > t_b) {
            d = (d << 1) | 0x1;
            t_a -= t_b; // Not a bitwise operatiion
            t_b = t_b >> 1;
         }
        else if (t_a == t_b) {
            d = (d << 1) | 0x1;
            t_a = 0;
        }
        else { // t_a < t_b
            d = d << 1;
            t_b = t_b >> 1;
        }
    }

    r = t_a;
    printf("==> %d %d\n", d, r);
    return d;
}

按位加法已经在其中一个答案中给出，所以跳过它。

使用cblas，作为OS X加速框架的一部分。

[02:31:59] [william@relativity ~]$ cat div3.c
#import <stdio.h>
#import <Accelerate/Accelerate.h>

int main() {
    float multiplicand = 123456.0;
    float multiplier = 0.333333;
    printf("%f * %f == ", multiplicand, multiplier);
    cblas_sscal(1, multiplier, &multiplicand, 1);
    printf("%f\n", multiplicand);
}

[02:32:07] [william@relativity ~]$ clang div3.c -framework Accelerate -o div3 && ./div3
123456.000000 * 0.333333 == 41151.957031