GB2312编码转换为UTF8编码

源程序：

#include <stdio.h>
#include <windows.h>


//
// Unicode 转换成UTF-8 
//
void UnicodeToUTF_8(char *pOut, const WCHAR wch)
{
	// 注意 WCHAR高低字的顺序,低字节在前，高字节在后
	char *pchar = (char *)&wch;
	
	pOut[0] = (0xE0 | ((pchar[1] & 0xF0) >> 4));
	pOut[1] = (0x80 | ((pchar[1] & 0x0F) << 2)) + ((pchar[0] & 0xC0) >> 6);
	pOut[2] = (0x80 | (pchar[0] & 0x3F));
	
	return;
}

//
// GB2312 转换成　Unicode
//
inline WCHAR Gb2312ToUnicode(const char *gbBuffer)
{
	WCHAR wch;
	::MultiByteToWideChar(CP_ACP, MB_PRECOMPOSED, gbBuffer, 2, &wch, 1);
	return wch;
}

//
//	GB2312 转为 UTF-8
//
int GB2312ToUTF_8(char *strOut, const char *pText, int nLen = -1)
{
	char buf[4];
	if (nLen == -1)
		nLen = strlen(pText);
	
	memset(buf,0,4);
	
	int i = 0;
	int k = 0;
	while (i < nLen)
	{
		//如果是英文直接复制就可以
		if( *(pText + i) >= 0)
		{
			//strOut += pText[i];
			strOut[k++] = pText[i];
			i++;
		}
		else
		{
			WCHAR wch;
			wch = Gb2312ToUnicode(pText + i);
			UnicodeToUTF_8(buf, wch);
			
// 			strOut += buf[0];
// 			strOut += buf[1];
// 			strOut += buf[2];
			strOut[k++] = buf[0];
			strOut[k++] = buf[1];
			strOut[k++] = buf[2];
			
			i += 2;
		}
	}
	
	return k;
}

//
// GB2312编码类型的文件可以通过EditPlus工具创建
//	UTF8编码类型的文件可以通过EditPlus工具验证查看
//
void main()
{

	//
	// 从一个GB2312的文件中读出来，写入到UTF8中去！
	//
	FILE *fp = fopen("gb2312.txt", "rb");
	FILE *fp2 = fopen("utf8.txt", "wb");
	char bufIn[1024] = {0};
	char bufOut[1024] = {0};
	int nFileLen = 0;
	if(fp)
	{
		//获得GB2312的文件的长度
		fseek(fp, 0L, SEEK_END);
		nFileLen = ftell(fp);

		fseek(fp, 0L, SEEK_SET);
		fread(bufIn, 1, nFileLen, fp);

		//
		// 转换
		//
		int k = GB2312ToUTF_8(bufOut, bufIn, nFileLen);
		
		//
		// 存成一个UTF8文件
		//
		if (fp2)
		{
			fwrite(bufOut, 1, k, fp2);
			fclose(fp2);
		}

	}
	fclose(fp);
}