C++ std::string 包含中文

最新推荐文章于 2024-08-07 20:45:23 发布

清风徐来

最新推荐文章于 2024-08-07 20:45:23 发布

阅读量1.4w

点赞数 9

分类专栏： C++

本文链接：https://blog.csdn.net/qq_30722721/article/details/102628769

版权

C++ 专栏收录该内容

50 篇文章 6 订阅

订阅专栏

环境：VS2015

场景描述：std::string 包含中文时，字符串截取有问题，如下所示：

int main()
{
    std::vector<std::string> spliteList;

    string value = "类型A123";
    for (int i = 0; i < value.size(); ++i)
    {
        spliteList.push_back(value.substr(i, 1));
    }
}

解决办法：std::string 转为 std::wstring后处理

#include <Windows.h>
#include <tchar.h>
#include <vector>
#include <string>

#pragma warning(disable:4996)
using namespace std;

std::string WStringToString(const std::wstring& ws)
{
    std::string strLocale = setlocale(LC_ALL, "");
    const wchar_t* wchSrc = ws.c_str();
    size_t nDestSize = wcstombs(NULL, wchSrc, 0) + 1;
    char *chDest = new char[nDestSize];
    memset(chDest, 0, nDestSize);
    wcstombs(chDest, wchSrc, nDestSize);
    std::string strResult = chDest;
    delete[]chDest;
    setlocale(LC_ALL, strLocale.c_str());
    return strResult;
}

std::wstring StringToWString(const std::string &str)
{
    std::wstring wContext = L"";
    int len = MultiByteToWideChar(CP_ACP, 0, str.c_str(), str.size(), NULL, 0);
    WCHAR* buffer = new WCHAR[len + 1];
    MultiByteToWideChar(CP_ACP, 0, str.c_str(), str.size(), buffer, len);
    buffer[len] = '\0';
    wContext.append(buffer);
    delete[] buffer;

    return wContext;
}

std::vector<std::string> SpiteStringCharacter(std::string context)
{
    std::vector<std::string> res;

    std::wstring wContext = StringToWString(context);
    for (int i = 0; i < wContext.length(); ++i)
    {
        std::wstring tmp = wContext.substr(i, 1);
        res.push_back(WStringToString(tmp));
    }

    return res;
}

bool IsChineseChar(std::wstring value)
{
    if (value.size() == 1)
    {
        unsigned char* pCh = (unsigned char*)&value[0];
        if (((*pCh >= 0) && (*pCh <= 0xff)) && (*(pCh + 1) >= 0x4e && *(pCh + 1) <= 0x9f))
        {
            return true;
        }
    }

    return false;
}

int GetStringChineseCharCount(std::string context)
{
    std::wstring wContext = StringToWString(context);

    int chineseCharCount = 0;
    for (int i = 0; i < wContext.length(); ++i)
    {
        if (IsChineseChar(wContext.substr(i, 1)))
        {
            ++chineseCharCount;
        }
    }

    return chineseCharCount;
}

int main()
{
    std::string value = "类型A123";
    std::vector<std::string> spliteList = SpiteStringCharacter(value);
    int count = GetStringChineseCharCount(value);

    return 0;
}

结果如下：