UTF-8 Validation

UTF-8 Validation

题目链接：https://leetcode.com/problems...

这道题关键是搞懂题目意思。

UTF-8

1 byte: characters from 0 to 127 == ASCII
2 bytes: characters from 127 to 2047
3 bytes: characters from 2048 to 65535
4 bytes: characters from 65536 to 1112064

The leading bits tell: the length of the bytes

• 1 byte: the 1st bit is 0

• 2 bytes:

  1. 1st byte: start with "110"

  2. 2nd byte: start with "10"

• 3 bytes:

  1. 1st byte: start with "1110"

  2. 2nd byte: start with "10"

  3. 3rd byte: start with "10"

• 4 bytes:

  1. 1st byte: start with "11110"

  2. 2nd byte: start with "10"

  3. 3rd byte: start with "10"

  4. 4th byte: start with "10"

思路及代码

知道意思之后，这道题就很简单了。
一个loop，每次分三步来做，loop invariant是每次data[i]都是first byte of 新的character

1. 统计data[i]后8位里面，从前开始有多少个1，用变量ones来保存，其中ones可能的值只有0, 2, 3, 4

2. 从 data[i+1] 开始检查，后八位中的前两位是否为'10'，一共检查ones - 1

3. 更新i的值为 i + ones

public class Solution {
    public boolean validUtf8(int[] data) {
        /* 1. check how many '1's = ones
         * 2. check (i + 1, i + ones - 1) for '10'
         * 3. update i = i + ones
         * valid ones: 0, 2, 3, 4
         */
        int i = 0;
        while(i < data.length) {
            // 1. find ones
            int ones = 0;
            while(((data[i] >> (7 - ones)) & 1) == 1) {
                ones++;
            }
            // invalid ones 
            if(ones == 1 || ones > 4) return false;
            // 2. check 1s
            i++;
            while(ones-- > 1) {
                if(i >= data.length || ((data[i] >> 6) & 3) != 2) return false;
                // 3. update i
                i++;
            }
        }
        
        return true;
    }
}

Advantage of UTF-8

• implement Unicode: encode different symbols(Chinese...)

• web pages are often coded in UTF-8, XML, JSON

• only use binary representation: 0 and 1

• endianness independent

Disadvantage of UTF-8

• space: use more bytes, larger

• time: calculate