C/C++中ASCII与Unicode字符串相互转换

时间:2023-02-04 20:20:10

1、ASCII to Unicode

函数:wcstombs(VC6)、wcstombs_s

实例:

// crt_wcstombs_s.c
// This example converts a wide character
// string to a multibyte character string.
#include <stdio.h>
#include <stdlib.h>
#include <assert.h>

#define BUFFER_SIZE 100

int main( void )
{
    size_t   i;
    char      *pMBBuffer = (char *)malloc( BUFFER_SIZE );
    wchar_t*pWCBuffer = L"Hello, world.";

    printf( "Convert wide-character string:\n" );

    // Conversion
    wcstombs_s(&i, pMBBuffer, (size_t)BUFFER_SIZE,
               pWCBuffer, (size_t)BUFFER_SIZE );

    // Output
    printf("   Characters converted: %u\n", i);
    printf("    Multibyte character: %s\n\n",
     pMBBuffer );

    // Free multibyte character buffer
    if (pMBBuffer)
    {
    free(pMBBuffer);
    }
}

 

2、Unicode to ASCII

函数:mbstowcs

 

// crt_mbstowcs.c
// compile with: /W3
// illustrates the behavior of the mbstowcs function

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

int main( void )
{
    size_t size;
    int nChar = 2; // number of characters to convert
    int requiredSize;

    unsigned char    *pmbnull  = NULL;
    unsigned char    *pmbhello = NULL;
    char* localeInfo;
   
    wchar_t *pwchello = L"\x3042\x3043"; // 2 Hiragana characters
    wchar_t *pwc;

    /* Enable the Japanese locale and codepage */
    localeInfo = setlocale(LC_ALL, "Japanese_Japan.932");
    printf("Locale information set to %s\n", localeInfo);
   
    printf( "Convert to multibyte string:\n" );

    requiredSize = wcstombs( NULL, pwchello, 0); // C4996
    // Note: wcstombs is deprecated; consider using wcstombs_s
    printf("  Required Size: %d\n", requiredSize);

    /* Add one to leave room for the null terminator. */
    pmbhello = (unsigned char *)malloc( requiredSize + 1);
    if (! pmbhello)
    {
        printf("Memory allocation failure.\n");
        return 1;
    }
    size = wcstombs( pmbhello, pwchello, requiredSize + 1); // C4996
    // Note: wcstombs is deprecated; consider using wcstombs_s
    if (size == (size_t) (-1))
    {
        printf("Couldn't convert string. Code page 932 may"
                " not be available.\n");
        return 1;
    }
    printf( "  Number of bytes written to multibyte string: %u\n",
            (unsigned int) size );
    printf( "  Hex values of the " );
    printf( " multibyte characters: %#.2x %#.2x %#.2x %#.2x\n",
            pmbhello[0], pmbhello[1], pmbhello[2], pmbhello[3] );
    printf( "  Codepage 932 uses 0x81 to 0x9f as lead bytes.\n\n");

    printf( "Convert back to wide-character string:\n" );

    /* Assume we don't know the length of the multibyte string.
     Get the required size in characters, and allocate enough space. */

    requiredSize = mbstowcs(NULL, pmbhello, 0); // C4996
    /* Add one to leave room for the NULL terminator */
    pwc = (wchar_t *)malloc( (requiredSize + 1) * sizeof( wchar_t ));
    if (! pwc)
    {
        printf("Memory allocation failure.\n");
        return 1;
    }
    size = mbstowcs( pwc, pmbhello, requiredSize + 1); // C4996
    if (size == (size_t) (-1))
    {
       printf("Couldn't convert string--invalid multibyte character.\n");
    }
    printf( "  Characters converted: %u\n", (unsigned int)size );
    printf( "  Hex value of first 2" );
    printf( " wide characters: %#.4x %#.4x\n\n", pwc[0], pwc[1] );
    free(pwc);
    free(pmbhello);
}