一 C++ 中 string與wstring互轉

方法一：

string WideToMutilByte(const wstring& _src)
{
int nBufSize = WideCharToMultiByte(GetACP(), 0, _src.c_str(),-1, NULL, 0, 0, FALSE);

char *szBuf = new char[nBufSize];

WideCharToMultiByte(GetACP(), 0, _src.c_str(),-1, szBuf, nBufSize, 0, FALSE);

string strRet(szBuf);

delete []szBuf;
szBuf = NULL;

return strRet;
}

wstring MutilByteToWide(const string& _src)
{
//計算字符串 string 轉成 wchar_t 之后占用的內存字節數
int nBufSize = MultiByteToWideChar(GetACP(),0,_src.c_str(),-1,NULL,0);

//為 wsbuf 分配內存 BufSize 個字節
wchar_t *wsBuf = new wchar_t[nBufSize];

//轉化為 unicode 的 WideString
MultiByteToWideChar(GetACP(),0,_src.c_str(),-1,wsBuf,nBufSize);

wstring wstrRet(wsBuf);

delete []wsBuf;
wsBuf = NULL;

return wstrRet;
}

 

轉載：csdn

這篇文章里，我將給出幾種C++ std::string和std::wstring相互轉換的轉換方法。
 
第一種方法：調用WideCharToMultiByte()和MultiByteToWideChar()，代碼如下（關于詳細的解釋，可以參考《windows核心編程》）：
 

#include <string>
#include <windows.h>
using namespace std;
//Converting a WChar string to a Ansi string
std::string WChar2Ansi(LPCWSTR pwszSrc)
{
         int nLen = WideCharToMultiByte(CP_ACP, 0, pwszSrc, -1, NULL, 0, NULL, NULL);
 
         if (nLen<= 0) return std::string("");
 
         char* pszDst = new char[nLen];
         if (NULL == pszDst) return std::string("");
 
         WideCharToMultiByte(CP_ACP, 0, pwszSrc, -1, pszDst, nLen, NULL, NULL);
         pszDst[nLen -1] = 0;
 
         std::string strTemp(pszDst);
         delete [] pszDst;
 
         return strTemp;
}

 
string ws2s(wstring& inputws)
{
        return WChar2Ansi(inputws.c_str());
}

 

 
//Converting a Ansi string to WChar string

std::wstring Ansi2WChar(LPCSTR pszSrc, int nLen)
 
{
    int nSize = MultiByteToWideChar(CP_ACP, 0, (LPCSTR)pszSrc, nLen, 0, 0);
    if(nSize <= 0) return NULL;
 
         WCHAR *pwszDst = new WCHAR[nSize+1];
    if( NULL == pwszDst) return NULL;
 
    MultiByteToWideChar(CP_ACP, 0,(LPCSTR)pszSrc, nLen, pwszDst, nSize);
    pwszDst[nSize] = 0;
 
    if( pwszDst[0] == 0xFEFF)                    // skip Oxfeff
        for(int i = 0; i < nSize; i ++)
                            pwszDst[i] = pwszDst[i+1];
 
    wstring wcharString(pwszDst);
         delete pwszDst;
 
    return wcharString;
}

 
std::wstring s2ws(const string& s)
{
     return Ansi2WChar(s.c_str(),s.size());
}

 
 
第二種方法：采用ATL封裝_bstr_t的過渡：（注，_bstr_是Microsoft Specific的，所以下面代碼可以在VS2005通過，無移植性）；

#include <string>
#include <comutil.h>
using namespace std;
#pragma comment(lib, "comsuppw.lib")
 
string ws2s(const wstring& ws);
wstring s2ws(const string& s);
 
string ws2s(const wstring& ws)
{
         _bstr_t t = ws.c_str();
         char* pchar = (char*)t;
         string result = pchar;
         return result;
}

 
wstring s2ws(const string& s)
{
         _bstr_t t = s.c_str();
         wchar_t* pwchar = (wchar_t*)t;
         wstring result = pwchar;
         return result;
}

 
第三種方法：使用CRT庫的mbstowcs()函數和wcstombs()函數，平臺無關，需設定locale。

#include <string>
#include <locale.h>
using namespace std;
string ws2s(const wstring& ws)
{
         string curLocale = setlocale(LC_ALL, NULL);        // curLocale = "C";
 
         setlocale(LC_ALL, "chs");
 
         const wchar_t* _Source = ws.c_str();
         size_t _Dsize = 2 * ws.size() + 1;
         char *_Dest = new char[_Dsize];
         memset(_Dest,0,_Dsize);
         wcstombs(_Dest,_Source,_Dsize);
         string result = _Dest;
         delete []_Dest;
 
         setlocale(LC_ALL, curLocale.c_str());
 
         return result;
}

 
wstring s2ws(const string& s)
{
         setlocale(LC_ALL, "chs");
 
         const char* _Source = s.c_str();
         size_t _Dsize = s.size() + 1;
         wchar_t *_Dest = new wchar_t[_Dsize];
         wmemset(_Dest, 0, _Dsize);
         mbstowcs(_Dest,_Source,_Dsize);
         wstring result = _Dest;
         delete []_Dest;
 
         setlocale(LC_ALL, "C");
 
         return result;
}

二 utf8.utf16.utf32的相互轉化

可以參考Unicode.org 上有ConvertUTF.c和ConvertUTF.h （下載地址：http://www.unicode.org/Public/PROGRAMS/CVTUTF/）

實現文件ConvertUTF.c：（.h省）
/**//*
 * Copyright 2001-2004 Unicode, Inc.
 *
 * Disclaimer
 *
 * This source code is provided as is by Unicode, Inc. No claims are
 * made as to fitness for any particular purpose. No warranties of any
 * kind are expressed or implied. The recipient agrees to determine
 * applicability of information provided. If this file has been
 * purchased on magnetic or optical media from Unicode, Inc., the
 * sole remedy for any claim will be exchange of defective media
 * within 90 days of receipt.
 *
 * Limitations on Rights to Redistribute This Code
 *
 * Unicode, Inc. hereby grants the right to freely use the information
 * supplied in this file in the creation of products supporting the
 * Unicode Standard, and to make copies of this file in any form
 * for internal or external distribution as long as this notice
 * remains attached.
 */

/**//* ---------------------------------------------------------------------

    Conversions between UTF32, UTF-16, and UTF-8. Source code file.
    Author: Mark E. Davis, 1994.
    Rev History: Rick McGowan, fixes & updates May 2001.
    Sept 2001: fixed const & error conditions per
    mods suggested by S. Parent & A. Lillich.
    June 2002: Tim Dodd added detection and handling of incomplete
    source sequences, enhanced error detection, added casts
    to eliminate compiler warnings.
    July 2003: slight mods to back out aggressive FFFE detection.
    Jan 2004: updated switches in from-UTF8 conversions.
    Oct 2004: updated to use UNI_MAX_LEGAL_UTF32 in UTF-32 conversions.

    See the header file "ConvertUTF.h" for complete documentation.

------------------------------------------------------------------------ */

#include "ConvertUTF.h"
#ifdef CVTUTF_DEBUG
#include <stdio.h>
#endif

static const int halfShift  = 10; /**//* used for shifting by 10 bits */

static const UTF32 halfBase = 0x0010000UL;
static const UTF32 halfMask = 0x3FFUL;

#define UNI_SUR_HIGH_START  (UTF32)0xD800
#define UNI_SUR_HIGH_END    (UTF32)0xDBFF
#define UNI_SUR_LOW_START   (UTF32)0xDC00
#define UNI_SUR_LOW_END     (UTF32)0xDFFF
#define false       0
#define true        1

/**//* --------------------------------------------------------------------- */

ConversionResult ConvertUTF32toUTF16 (
    const UTF32** sourceStart, const UTF32* sourceEnd,
    UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags) {
    ConversionResult result = conversionOK;
    const UTF32* source = *sourceStart;
    UTF16* target = *targetStart;
    while (source < sourceEnd) {
    UTF32 ch;
    if (target >= targetEnd) {
        result = targetExhausted; break;
    }
    ch = *source++;
    if (ch <= UNI_MAX_BMP) { /**//* Target is a character <= 0xFFFF */
        /**//* UTF-16 surrogate values are illegal in UTF-32; 0xffff or 0xfffe are both reserved values */
        if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
        if (flags == strictConversion) {
            --source; /**//* return to the illegal value itself */
            result = sourceIllegal;
            break;
        } else {
            *target++ = UNI_REPLACEMENT_CHAR;
        }
        } else {
        *target++ = (UTF16)ch; /**//* normal case */
        }
    } else if (ch > UNI_MAX_LEGAL_UTF32) {
        if (flags == strictConversion) {
        result = sourceIllegal;
        } else {
        *target++ = UNI_REPLACEMENT_CHAR;
        }
    } else {
        /**//* target is a character in range 0xFFFF - 0x10FFFF. */
        if (target + 1 >= targetEnd) {
        --source; /**//* Back up source pointer! */
        result = targetExhausted; break;
        }
        ch -= halfBase;
        *target++ = (UTF16)((ch >> halfShift) + UNI_SUR_HIGH_START);
        *target++ = (UTF16)((ch & halfMask) + UNI_SUR_LOW_START);
    }
    }
    *sourceStart = source;
    *targetStart = target;
    return result;
}

/**//* --------------------------------------------------------------------- */

ConversionResult ConvertUTF16toUTF32 (
    const UTF16** sourceStart, const UTF16* sourceEnd,
    UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags) {
    ConversionResult result = conversionOK;
    const UTF16* source = *sourceStart;
    UTF32* target = *targetStart;
    UTF32 ch, ch2;
    while (source < sourceEnd) {
    const UTF16* oldSource = source; /**//*  In case we have to back up because of target overflow. */
    ch = *source++;
    /**//* If we have a surrogate pair, convert to UTF32 first. */
    if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_HIGH_END) {
        /**//* If the 16 bits following the high surrogate are in the source buffer */
        if (source < sourceEnd) {
        ch2 = *source;
        /**//* If it's a low surrogate, convert to UTF32. */
        if (ch2 >= UNI_SUR_LOW_START && ch2 <= UNI_SUR_LOW_END) {
            ch = ((ch - UNI_SUR_HIGH_START) << halfShift)
            + (ch2 - UNI_SUR_LOW_START) + halfBase;
            ++source;
        } else if (flags == strictConversion) { /**//* it's an unpaired high surrogate */
            --source; /**//* return to the illegal value itself */
            result = sourceIllegal;
            break;
        }
        } else { /**//* We don't have the 16 bits following the high surrogate. */
        --source; /**//* return to the high surrogate */
        result = sourceExhausted;
        break;
        }
    } else if (flags == strictConversion) {
        /**//* UTF-16 surrogate values are illegal in UTF-32 */
        if (ch >= UNI_SUR_LOW_START && ch <= UNI_SUR_LOW_END) {
        --source; /**//* return to the illegal value itself */
        result = sourceIllegal;
        break;
        }
    }
    if (target >= targetEnd) {
        source = oldSource; /**//* Back up source pointer! */
        result = targetExhausted; break;
    }
    *target++ = ch;
    }
    *sourceStart = source;
    *targetStart = target;
#ifdef CVTUTF_DEBUG
if (result == sourceIllegal) {
    fprintf(stderr, "ConvertUTF16toUTF32 illegal seq 0x%04x,%04x\n", ch, ch2);
    fflush(stderr);
}
#endif
    return result;
}

/**//* --------------------------------------------------------------------- */

/**//*
 * Index into the table below with the first byte of a UTF-8 sequence to
 * get the number of trailing bytes that are supposed to follow it.
 * Note that *legal* UTF-8 values can't have 4 or 5-bytes. The table is
 * left as-is for anyone who may want to do such conversion, which was
 * allowed in earlier algorithms.
 */
static const char trailingBytesForUTF8[256] = {
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
    1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
    2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5
};

/**//*
 * Magic values subtracted from a buffer value during UTF8 conversion.
 * This table contains as many values as there might be trailing bytes
 * in a UTF-8 sequence.
 */
static const UTF32 offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL,
             0x03C82080UL, 0xFA082080UL, 0x82082080UL };

/**//*
 * Once the bits are split out into bytes of UTF-8, this is a mask OR-ed
 * into the first byte, depending on how many bytes follow.  There are
 * as many entries in this table as there are UTF-8 sequence types.
 * (I.e., one byte sequence, two byte etc.). Remember that sequencs
 * for *legal* UTF-8 will be 4 or fewer bytes total.
 */
static const UTF8 firstByteMark[7] = { 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC };

/**//* --------------------------------------------------------------------- */

/**//* The interface converts a whole buffer to avoid function-call overhead.
 * Constants have been gathered. Loops & conditionals have been removed as
 * much as possible for efficiency, in favor of drop-through switches.
 * (See "Note A" at the bottom of the file for equivalent code.)
 * If your compiler supports it, the "isLegalUTF8" call can be turned
 * into an inline function.
 */

/**//* --------------------------------------------------------------------- */

ConversionResult ConvertUTF16toUTF8 (
    const UTF16** sourceStart, const UTF16* sourceEnd,
    UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags) {
    ConversionResult result = conversionOK;
    const UTF16* source = *sourceStart;
    UTF8* target = *targetStart;
    while (source < sourceEnd) {
    UTF32 ch;
    unsigned short bytesToWrite = 0;
    const UTF32 byteMask = 0xBF;
    const UTF32 byteMark = 0x80;
    const UTF16* oldSource = source; /**//* In case we have to back up because of target overflow. */
    ch = *source++;
    /**//* If we have a surrogate pair, convert to UTF32 first. */
    if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_HIGH_END) {
        /**//* If the 16 bits following the high surrogate are in the source buffer */
        if (source < sourceEnd) {
        UTF32 ch2 = *source;
        /**//* If it's a low surrogate, convert to UTF32. */
        if (ch2 >= UNI_SUR_LOW_START && ch2 <= UNI_SUR_LOW_END) {
            ch = ((ch - UNI_SUR_HIGH_START) << halfShift)
            + (ch2 - UNI_SUR_LOW_START) + halfBase;
            ++source;
        } else if (flags == strictConversion) { /**//* it's an unpaired high surrogate */
            --source; /**//* return to the illegal value itself */
            result = sourceIllegal;
            break;
        }
        } else { /**//* We don't have the 16 bits following the high surrogate. */
        --source; /**//* return to the high surrogate */
        result = sourceExhausted;
        break;
        }
    } else if (flags == strictConversion) {
        /**//* UTF-16 surrogate values are illegal in UTF-32 */
        if (ch >= UNI_SUR_LOW_START && ch <= UNI_SUR_LOW_END) {
        --source; /**//* return to the illegal value itself */
        result = sourceIllegal;
        break;
        }
    }
    /**//* Figure out how many bytes the result will require */
    if (ch < (UTF32)0x80) {         bytesToWrite = 1;
    } else if (ch < (UTF32)0x800) {     bytesToWrite = 2;
    } else if (ch < (UTF32)0x10000) {   bytesToWrite = 3;
    } else if (ch < (UTF32)0x110000) {  bytesToWrite = 4;
    } else {                bytesToWrite = 3;
                        ch = UNI_REPLACEMENT_CHAR;
    }

    target += bytesToWrite;
    if (target > targetEnd) {
        source = oldSource; /**//* Back up source pointer! */
        target -= bytesToWrite; result = targetExhausted; break;
    }
    switch (bytesToWrite) { /**//* note: everything falls through. */
        case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
        case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
        case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
        case 1: *--target =  (UTF8)(ch | firstByteMark[bytesToWrite]);
    }
    target += bytesToWrite;
    }
    *sourceStart = source;
    *targetStart = target;
    return result;
}

/**//* --------------------------------------------------------------------- */

/**//*
 * Utility routine to tell whether a sequence of bytes is legal UTF-8.
 * This must be called with the length pre-determined by the first byte.
 * If not calling this from ConvertUTF8to*, then the length can be set by:
 *  length = trailingBytesForUTF8[*source]+1;
 * and the sequence is illegal right away if there aren't that many bytes
 * available.
 * If presented with a length > 4, this returns false.  The Unicode
 * definition of UTF-8 goes up to 4-byte sequences.
 */

static Boolean isLegalUTF8(const UTF8 *source, int length) {
    UTF8 a;
    const UTF8 *srcptr = source+length;
    switch (length) {
    default: return false;
    /**//* Everything else falls through when "true" */
    case 4: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
    case 3: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return false;
    case 2: if ((a = (*--srcptr)) > 0xBF) return false;

    switch (*source) {
        /**//* no fall-through in this inner switch */
        case 0xE0: if (a < 0xA0) return false; break;
        case 0xED: if (a > 0x9F) return false; break;
        case 0xF0: if (a < 0x90) return false; break;
        case 0xF4: if (a > 0x8F) return false; break;
        default:   if (a < 0x80) return false;
    }

    case 1: if (*source >= 0x80 && *source < 0xC2) return false;
    }
    if (*source > 0xF4) return false;
    return true;
}

/**//* --------------------------------------------------------------------- */

/**//*
 * Exported function to return whether a UTF-8 sequence is legal or not.
 * This is not used here; it's just exported.
 */
Boolean isLegalUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd) {
    int length = trailingBytesForUTF8[*source]+1;
    if (source+length > sourceEnd) {
    return false;
    }
    return isLegalUTF8(source, length);
}

/**//* --------------------------------------------------------------------- */

ConversionResult ConvertUTF8toUTF16 (
    const UTF8** sourceStart, const UTF8* sourceEnd,
    UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags) {
    ConversionResult result = conversionOK;
    const UTF8* source = *sourceStart;
    UTF16* target = *targetStart;
    while (source < sourceEnd) {
    UTF32 ch = 0;
    unsigned short extraBytesToRead = trailingBytesForUTF8[*source];
    if (source + extraBytesToRead >= sourceEnd) {
        result = sourceExhausted; break;
    }
    /**//* Do this check whether lenient or strict */
    if (! isLegalUTF8(source, extraBytesToRead+1)) {
        result = sourceIllegal;
        break;
    }
    /**//*
     * The cases all fall through. See "Note A" below.
     */
    switch (extraBytesToRead) {
        case 5: ch += *source++; ch <<= 6; /**//* remember, illegal UTF-8 */
        case 4: ch += *source++; ch <<= 6; /**//* remember, illegal UTF-8 */
        case 3: ch += *source++; ch <<= 6;
        case 2: ch += *source++; ch <<= 6;
        case 1: ch += *source++; ch <<= 6;
        case 0: ch += *source++;
    }
    ch -= offsetsFromUTF8[extraBytesToRead];

    if (target >= targetEnd) {
        source -= (extraBytesToRead+1); /**//* Back up source pointer! */
        result = targetExhausted; break;
    }
    if (ch <= UNI_MAX_BMP) { /**//* Target is a character <= 0xFFFF */
        /**//* UTF-16 surrogate values are illegal in UTF-32 */
        if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
        if (flags == strictConversion) {
            source -= (extraBytesToRead+1); /**//* return to the illegal value itself */
            result = sourceIllegal;
            break;
        } else {
            *target++ = UNI_REPLACEMENT_CHAR;
        }
        } else {
        *target++ = (UTF16)ch; /**//* normal case */
        }
    } else if (ch > UNI_MAX_UTF16) {
        if (flags == strictConversion) {
        result = sourceIllegal;
        source -= (extraBytesToRead+1); /**//* return to the start */
        break; /**//* Bail out; shouldn't continue */
        } else {
        *target++ = UNI_REPLACEMENT_CHAR;
        }
    } else {
        /**//* target is a character in range 0xFFFF - 0x10FFFF. */
        if (target + 1 >= targetEnd) {
        source -= (extraBytesToRead+1); /**//* Back up source pointer! */
        result = targetExhausted; break;
        }
        ch -= halfBase;
        *target++ = (UTF16)((ch >> halfShift) + UNI_SUR_HIGH_START);
        *target++ = (UTF16)((ch & halfMask) + UNI_SUR_LOW_START);
    }
    }
    *sourceStart = source;
    *targetStart = target;
    return result;
}

/**//* --------------------------------------------------------------------- */

ConversionResult ConvertUTF32toUTF8 (
    const UTF32** sourceStart, const UTF32* sourceEnd,
    UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags) {
    ConversionResult result = conversionOK;
    const UTF32* source = *sourceStart;
    UTF8* target = *targetStart;
    while (source < sourceEnd) {
    UTF32 ch;
    unsigned short bytesToWrite = 0;
    const UTF32 byteMask = 0xBF;
    const UTF32 byteMark = 0x80;
    ch = *source++;
    if (flags == strictConversion ) {
        /**//* UTF-16 surrogate values are illegal in UTF-32 */
        if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
        --source; /**//* return to the illegal value itself */
        result = sourceIllegal;
        break;
        }
    }
    /**//*
     * Figure out how many bytes the result will require. Turn any
     * illegally large UTF32 things (> Plane 17) into replacement chars.
     */
    if (ch < (UTF32)0x80) {         bytesToWrite = 1;
    } else if (ch < (UTF32)0x800) {     bytesToWrite = 2;
    } else if (ch < (UTF32)0x10000) {   bytesToWrite = 3;
    } else if (ch <= UNI_MAX_LEGAL_UTF32) {  bytesToWrite = 4;
    } else {                bytesToWrite = 3;
                        ch = UNI_REPLACEMENT_CHAR;
                        result = sourceIllegal;
    }
   
    target += bytesToWrite;
    if (target > targetEnd) {
        --source; /**//* Back up source pointer! */
        target -= bytesToWrite; result = targetExhausted; break;
    }
    switch (bytesToWrite) { /**//* note: everything falls through. */
        case 4: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
        case 3: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
        case 2: *--target = (UTF8)((ch | byteMark) & byteMask); ch >>= 6;
        case 1: *--target = (UTF8) (ch | firstByteMark[bytesToWrite]);
    }
    target += bytesToWrite;
    }
    *sourceStart = source;
    *targetStart = target;
    return result;
}

/**//* --------------------------------------------------------------------- */

ConversionResult ConvertUTF8toUTF32 (
    const UTF8** sourceStart, const UTF8* sourceEnd,
    UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags) {
    ConversionResult result = conversionOK;
    const UTF8* source = *sourceStart;
    UTF32* target = *targetStart;
    while (source < sourceEnd) {
    UTF32 ch = 0;
    unsigned short extraBytesToRead = trailingBytesForUTF8[*source];
    if (source + extraBytesToRead >= sourceEnd) {
        result = sourceExhausted; break;
    }
    /**//* Do this check whether lenient or strict */
    if (! isLegalUTF8(source, extraBytesToRead+1)) {
        result = sourceIllegal;
        break;
    }
    /**//*
     * The cases all fall through. See "Note A" below.
     */
    switch (extraBytesToRead) {
        case 5: ch += *source++; ch <<= 6;
        case 4: ch += *source++; ch <<= 6;
        case 3: ch += *source++; ch <<= 6;
        case 2: ch += *source++; ch <<= 6;
        case 1: ch += *source++; ch <<= 6;
        case 0: ch += *source++;
    }
    ch -= offsetsFromUTF8[extraBytesToRead];

    if (target >= targetEnd) {
        source -= (extraBytesToRead+1); /**//* Back up the source pointer! */
        result = targetExhausted; break;
    }
    if (ch <= UNI_MAX_LEGAL_UTF32) {
        /**//*
         * UTF-16 surrogate values are illegal in UTF-32, and anything
         * over Plane 17 (> 0x10FFFF) is illegal.
         */
        if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) {
        if (flags == strictConversion) {
            source -= (extraBytesToRead+1); /**//* return to the illegal value itself */
            result = sourceIllegal;
            break;
        } else {
            *target++ = UNI_REPLACEMENT_CHAR;
        }
        } else {
        *target++ = ch;
        }
    } else { /**//* i.e., ch > UNI_MAX_LEGAL_UTF32 */
        result = sourceIllegal;
        *target++ = UNI_REPLACEMENT_CHAR;
    }
    }
    *sourceStart = source;
    *targetStart = target;
    return result;
}

/**//* ---------------------------------------------------------------------

    Note A.
    The fall-through switches in UTF-8 reading code save a
    temp variable, some decrements & conditionals.  The switches
    are equivalent to the following loop:
    {
        int tmpBytesToRead = extraBytesToRead+1;
        do {
        ch += *source++;
        --tmpBytesToRead;
        if (tmpBytesToRead) ch <<= 6;
        } while (tmpBytesToRead > 0);
    }
    In UTF-8 writing code, the switches on "bytesToWrite" are
    similarly unrolled loops.

   --------------------------------------------------------------------- */

 

三 C++ 的字符串與C#的轉化

1）將system::String 轉化為C++的string：
// convert_system_string.cpp
// compile with: /clr
#include <string>
#include <iostream>
using namespace std;
using namespace System;

void MarshalString ( String ^ s, string& os ) {
   using namespace Runtime::InteropServices;
   const char* chars =
      (const char*)(Marshal::StringToHGlobalAnsi(s)).ToPointer();
   os = chars;
   Marshal::FreeHGlobal(IntPtr((void*)chars));
}

void MarshalString ( String ^ s, wstring& os ) {
   using namespace Runtime::InteropServices;
   const wchar_t* chars =
      (const wchar_t*)(Marshal::StringToHGlobalUni(s)).ToPointer();
   os = chars;
   Marshal::FreeHGlobal(IntPtr((void*)chars));
}

int main() {
   string a = "test";
   wstring b = L"test2";
   String ^ c = gcnew String("abcd");

   cout << a << endl;
   MarshalString(c, a);
   c = "efgh";
   MarshalString(c, b);
   cout << a << endl;
   wcout << b << endl;
}

2）將System::String轉化為char*或w_char*
// convert_string_to_wchar.cpp
// compile with: /clr
#include < stdio.h >
#include < stdlib.h >
#include < vcclr.h >

using namespace System;

int main() {
   String ^str = "Hello";

   // Pin memory so GC can't move it while native function is called
   pin_ptr<const wchar_t> wch = PtrToStringChars(str);
   printf_s("%S\n", wch);

   // Conversion to char* :
   // Can just convert wchar_t* to char* using one of the
   // conversion functions such as:
   // WideCharToMultiByte()
   // wcstombs_s()
   //  etc
   size_t convertedChars = 0;
   size_t  sizeInBytes = ((str->Length + 1) * 2);
   errno_t err = 0;
   char    *ch = (char *)malloc(sizeInBytes);

   err = wcstombs_s(&convertedChars,
                    ch, sizeInBytes,
                    wch, sizeInBytes);
   if (err != 0)
      printf_s("wcstombs_s  failed!\n");

    printf_s("%s\n", ch);
}