深入理解字符串编码与转换-优快云博客

本文链接：https://blog.youkuaiyun.com/ryfdizuo/article/details/36897043

多字节和宽字符

C++中string / char* ，wstring / wchar_t*

C++测试

window下面

char* cName = "北京市";
// 多字节转化成宽字符字符串！
unsigned short wsName[50] = {0};
int wideCharCount = MultiByteToWideChar(CP_ACP, 0, (LPSTR)cName, -1, NULL, 0) - 1;
MultiByteToWideChar(CP_ACP, 0, (LPSTR)cName, -1, (LPWSTR)wsName, wideCharCount + 1);

for (int i=0; i<wideCharCount; i++)
{
	printf("%d ", wsName[i]);
}printf("\n");

输出

21271 20140 24066

Linux下面

测试代码如下：

#include <stdlib.h>
#include <stdio.h>
#include <string.h>

#include <locale.h>
#include <iostream>
#include <string>
using namespace std;

void multibyte_to_widechar_test();
void read_file(const char* fname);
void dump_uchar(unsigned char ch);

int main()
{
    multibyte_to_widechar_test();
    read_file("chs");

    printf("any key pressed to exit...\n");
    getchar();
    
    return 0;
}

void multibyte_to_widechar_test()
{
    typedef string str_t;
    str_t cur_loc = setlocale(LC_ALL, NULL);
    printf("cur_locale = %s\n", cur_loc.c_str());   
    setlocale(LC_ALL, "zh_CN.GBK"); 
    
    char mb_buf[100];
    strcpy(mb_buf, "北京市");
    int mbstr_len = strlen(mb_buf);
    
    wchar_t* wcstr = NULL; 
    int wcstr_len = mbstowcs(wcstr, mb_buf, 0) + 1; 

    printf("mb_len = %d, wc_len = %d\n", mbstr_len, wcstr_len);

    wcstr = new wchar_t[wcstr_len];
    int ret = mbstowcs(wcstr, mb_buf, mbstr_len);

    if (ret <= 0)
    {
        printf("转化失败\n");
    }
    else    
    {
        printf("转化成功\n");
        // wsprintf(L"%ls\n", wcstr);

        printf("view1 =====\n");
        
        for (int i=0; i<wcstr_len - 1; i++) 
        {       
            int code = (int)wcstr[i];
            printf("%d\t", code);
        } printf("\n");
        
        printf("view2 =====\n");
        for (int i=0; i<wcstr_len - 1; i++) 
        {       
            int code = (int)wcstr[i];
            dump_uchar( (unsigned char)(code/256) );
            dump_uchar( (unsigned char)(code%256) );
        } printf("\n");
        
    }

    setlocale(LC_ALL, cur_loc.c_str());

}

void dump_uchar(unsigned char ch)
{
    const char* str = "0123456789abcdef";

    printf("0x%c%c\t", str[ch/16], str[ch%16]);
}

void read_file(const char* fname)
{
    FILE* fp = fopen(fname, "r");

    if (!fp)
    {
        return;
    }

    printf("===============\n");

    char buffer[100] = {0};
    fgets(buffer, 100, fp);
    printf("%s", buffer);

    printf("view1 =========== \n");
    int len = strlen(buffer) - 1;
    for (int i=0; i<len; i++)
    {
        dump_uchar((unsigned char)buffer[i]);
    }printf("\n");

    printf("view2 =========== \n");
    for (int i=0; i<len; i+=2)
    {
        unsigned char down = (unsigned char)buffer[i];
        unsigned char high = (unsigned char)buffer[i+1];
        printf("%d ", (high<<8)|down);
    } printf("\n");

    fclose(fp);
}

multibyte_to_widechar_test函数将多字节编码转化成unicode编码。然后输出unicode串内容。

read_file尝试读取文件中字符串编码内容。

chs通过vi直接生成，内容为”北京市“，,/base_profile中设置如下：

export LC_ALL="zh_CN.GBK"

所以chs文件的编码默认是gbk。

g++ test.cpp -o app_test，然后运行输出：

root@h10-xx-xx-xx:~/peteryfren/cpp/encode_app> ./app_test 
cur_locale = C
mb_len = 6, wc_len = 4
转化成功
view1 =====
21271   20140   24066
view2 =====
0x53    0x17    0x4e    0xac    0x5e    0x02
===============
北京市
view1 =========== 
0xb1    0xb1    0xbe    0xa9    0xca    0xd0
view2 =========== 
45489 43454 53450 
any key pressed to exit...

“北京市”的unicode编码值与window上输出一致。“北京市”的gbk2312编码为45489,43454,53450。

同时linux vi创建的文件编码为gbk，与base_profile中设置一致。

BTW linux下通过iconv将utf-8编码文件转化成unicode：

iconv -f UTF-8 -t GBK test.txt -o pp.txt

python2.7测试

>>> s = u'北京市'
>>> s
u'\u5317\u4eac\u5e02'
>>> gbks = '北京市'
>>> gbks
'\xb1\xb1\xbe\xa9\xca\xd0'
>>> s.encode('utf-8')
'\xe5\x8c\x97\xe4\xba\xac\xe5\xb8\x82'

2.7下面加u表示unicode编码，不加u使用了gbk编码。

python3.3下面不能输出字符串的字节码，

>>s

等价于，

>>print(s)