VS2010中C++实现编码判断，转换

最新推荐文章于 2025-05-29 20:29:32 发布

songyachao

最新推荐文章于 2025-05-29 20:29:32 发布

阅读量2.1k

点赞数

CC 4.0 BY-SA版权

分类专栏： C/C++ 个人总结文章标签： C++ 编码 VC VS2010

本文链接：https://blog.youkuaiyun.com/songyachao/article/details/51897704

个人总结同时被 2 个专栏收录

15 篇文章

订阅专栏

C/C++

5 篇文章

订阅专栏

本文介绍了一个用于检测文本编码格式并实现不同编码间转换的方法。通过使用COM组件IMultiLanguage3来检测输入文本的编码类型，如UTF-8或GBK，并提供了从UTF-8到Unicode再转为GB2312的具体实现。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

//------------------------------判断编码格式------------------------------------------------
int CMainFrame::checkCodeType(char* pText,DWORD nLength)
{
CoInitialize(NULL);
{
IMultiLanguage3Ptr pML(CLSID_CMultiLanguage, NULL, CLSCTX_INPROC);
// char data[] = { 0xD6, 0xD0, 0xCE, 0xC4, 0xB1, 0xE0, 0xC2, 0xEB };
char* database = new char[nLength];
database=pText;
int isize = sizeof(database);
DetectEncodingInfo result[32];
int result_count = sizeof(result) / sizeof(result[0]);
HRESULT hr = pML->DetectInputCodepage(MLDETECTCP_NONE, 0, database, &isize, result, &result_count);
if (!SUCCEEDED(hr))
{
fprintf(stderr, "Failed with 0x%x\n", hr);
CoUninitialize();
return hr;
}
for (int i = 0; i < result_count; i++)
{
WCHAR desc[100] = {0};
pML->GetCodePageDescription(result[i].nCodePage, result[i].nLangID, desc, 100);
if(desc[0]=='U')
{
cFlag=2; //utf-8
return cFlag;
}else if(desc[0]=='C')
{
cFlag=1; //gbk
return cFlag;
}
}
}
CoUninitialize();
return cFlag;
}
//--------------------------------------------------------
void CMainFrame::UTF_8ToUnicode(wchar_t* pOut,char *pText)
{
char* uchar = (char *)pOut;

uchar[1] = ((pText[0] & 0x0F) << 4) + ((pText[1] >> 2) & 0x0F);

uchar[0] = ((pText[1] & 0x03) << 6) + (pText[2] & 0x3F);

return;

}

void CMainFrame::UnicodeToGB2312(char* pOut,wchar_t uData)
{
WideCharToMultiByte(CP_ACP,NULL,&uData,1,pOut,sizeof(wchar_t),NULL,NULL);
return;
}
void CMainFrame::UTF_8ToGB2312(char* pOut, char *pText, DWORD pLen)
{
//char * newBuf = new char[pLen];
char * newBuf =pOut;

char Ctemp[2]; //4
memset(Ctemp,0,2);//4
int i =0;
int j = 0;
while(i < pLen)
{
if(pText <0)
{
newBuf[j++] = pText[i++];
}
else
{
WCHAR Wtemp;
UTF_8ToUnicode(&Wtemp,pText + i);
UnicodeToGB2312(Ctemp,Wtemp);
newBuf[j] = Ctemp[0];
newBuf[j + 1] = Ctemp[1];
i += 3;
j += 2;
}

}
newBuf[j] =' ';
pOut = newBuf;
// delete []newBuf;
return;
}