GBK是Windows默认的汉字编码方式，又称国标。想在控制台或是控件上显示汉字使用这种编码方式最方便。
UTF-8是一种变长字节编码方式，把字节数边长，可以容纳更多的信息。
UTF-8编码方式：
当码长为1字节的时候，兼容ascii编码，格式为0xxxxxxx （x处表示有效位）
当码长为2字节的时候，格式为110xxxxx 10xxxxxx
当码长为3字节的时候，格式为1110xxxx 10xxxxxx 10xxxxxx
当码长为4字节的时候，格式为11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
...
来自：/haohulala/article/details/86600936
·UTF8编码的码长最大为6字节，1111110x 10...
·如果一个汉字的utf8编码为110aaaaa 10bbbbbb 则其对应的Unicode编码为aaaaabbbbbb，其他同理
所以我们就有了对于单个字符的UTF-8到Unicode的转换过程
int GetUTF8Length(unsigned char *sUTF8){ unsigned char _s0=sUTF8[0];if( _s0 >> 7 == 0b0 )return 1;if( _s0 >> 5 == 0b110 )return 2;if( _s0 >> 4 == 0b1110 )return 3;if( _s0 >> 3 == 0b11110 )return 4;if( _s0 >> 2 == 0b111110 )return 5;if( _s0 >> 1 == 0b1111110 )return 6;return 0;}unsigned int UTF8_To_Unicode(unsigned char *sUTF8){int l=GetUTF8Length(sUTF8);if(l==1) return sUTF8[0];if(l==2) return ((sUTF8[0]& 0b00011111)<<6) + (sUTF8[1]& 0b00111111);//0b110xxxxxif(l==3) return ((sUTF8[0]& 0b00001111)<<12) + ((sUTF8[1]& 0b00111111)<<6)+ (sUTF8[2]& 0b00111111);//0b1110xxxxif(l==4) return ((sUTF8[0]& 0b00000111)<<18)+ ((sUTF8[1]& 0b00111111)<<12)+ ((sUTF8[2]& 0b00111111)<<6)+ (sUTF8[3]& 0b00111111);//0b11110xxxif(l==5) return ((sUTF8[0]& 0b00000011)<<24)+ ((sUTF8[1]& 0b00111111)<<18)+ ((sUTF8[2]& 0b00111111)<<12)+ ((sUTF8[3]& 0b00111111)<<6)+ (sUTF8[4]& 0b00111111);//0b111110xxif(l==6) return ((sUTF8[0]& 0b00000001)<<30)+ ((sUTF8[1]& 0b00111111)<<24)+ ((sUTF8[2]& 0b00111111)<<18)+ ((sUTF8[3]& 0b00111111)<<12)+ ((sUTF8[4]& 0b00111111)<<6)+ (sUTF8[5]& 0b00111111);//0b1111110x}
我暂且还没想出一种好的算法把这几大坨if用一个简短的表达式表示，欢迎大家讨论。
至于Unicode到GBK的转换，c++有一些函数可以帮忙
char* Unicode_To_GBK(unsigned int cUni){//Unicode 16 BEchar* sGBK=new char[6];wsprintf(sGBK,"%wc",(wchar_t)cUni);return sGBK;}
下面是读取汉字源码
#include <Windows.h>#include<iostream>#include<cstdio>//#include<bitset>#include<cstring>using namespace std;//void binoutput(int n){ //二进制输出 //cout<<bitset<10>(n)<<"\n"; //}int GetUTF8Length(char *sUTF8){ unsigned char _s0=sUTF8[0];if( _s0 >> 7 == 0b0 )return 1;if( _s0 >> 5 == 0b110 )return 2;if( _s0 >> 4 == 0b1110 )return 3;if( _s0 >> 3 == 0b11110 )return 4;if( _s0 >> 2 == 0b111110 )return 5;if( _s0 >> 1 == 0b1111110 )return 6;return 0;}unsigned int UTF8_To_Unicode(unsigned char *sUTF8){int l=GetUTF8Length(reinterpret_cast<char *>(sUTF8));if(l==1) return sUTF8[0];if(l==2) return ((sUTF8[0]& 0b00011111)<<6) + (sUTF8[1]& 0b00111111);//0b110xxxxxif(l==3) return ((sUTF8[0]& 0b00001111)<<12) + ((sUTF8[1]& 0b00111111)<<6)+ (sUTF8[2]& 0b00111111);//0b1110xxxxif(l==4) return ((sUTF8[0]& 0b00000111)<<18)+ ((sUTF8[1]& 0b00111111)<<12)+ ((sUTF8[2]& 0b00111111)<<6)+ (sUTF8[3]& 0b00111111);//0b11110xxxif(l==5) return ((sUTF8[0]& 0b00000011)<<24)+ ((sUTF8[1]& 0b00111111)<<18)+ ((sUTF8[2]& 0b00111111)<<12)+ ((sUTF8[3]& 0b00111111)<<6)+ (sUTF8[4]& 0b00111111);//0b111110xxif(l==6) return ((sUTF8[0]& 0b00000001)<<30)+ ((sUTF8[1]& 0b00111111)<<24)+ ((sUTF8[2]& 0b00111111)<<18)+ ((sUTF8[3]& 0b00111111)<<12)+ ((sUTF8[4]& 0b00111111)<<6)+ (sUTF8[5]& 0b00111111);//0b1111110x}char* Unicode_To_GBK(unsigned int cUni){//Unicode 16 BEchar* sGBK=new char[6];wsprintf(sGBK,"%wc",(wchar_t)cUni);return sGBK;}char* UTF8_To_GBK(char *sUTF8){return Unicode_To_GBK(UTF8_To_Unicode(reinterpret_cast<unsigned char *>(sUTF8)));}int main(){/*啊 GBK:0xB0 0xA1Unicode-16 LE:0x4A 0x55Unicode-16 BE:0x55 0x4AUTF-8:0xE5 0x95 0x8A*/FILE *fp=fopen("hanzi.txt","r");char str[100]={0},*pstr,*endstr;while(!feof(fp)){fgets(str,100,fp);pstr=str;endstr=str+strlen(str);while(pstr < endstr){printf("%s",UTF8_To_GBK(pstr));pstr+= GetUTF8Length(pstr);}}fclose(fp);}
提一嘴reinterpret_cast，强制转换真是太爽了，当年被一堆cannot convert from typeto const type 要折磨疯了。
