unicode.c

uclinux 下的vlc播放器源代码

            {
                free( utf_entry );
                goto error;
            }
            tab = newtab;
            tab[num++] = utf_entry;
        }
        vlc_closedir_wrapper( dir );

        if( compar != NULL )
            qsort( tab, num, sizeof( tab[0] ),
                   (int (*)( const void *, const void *))compar );

        *namelist = tab;
        return num;

    error:{
        unsigned i;

        for( i = 0; i < num; i++ )
            free( tab[i] );
        if( tab != NULL )
            free( tab );
        return -1;}
    }
}


static int utf8_statEx( const char *filename, void *buf,
                        vlc_bool_t deref )
{
#if defined (WIN32) || defined (UNDER_CE)
    /* retrieve Windows OS version */
    if( GetVersion() < 0x80000000 )
    {
        /* for Windows NT and above */
        wchar_t wpath[MAX_PATH + 1];

        if( !MultiByteToWideChar( CP_UTF8, 0, filename, -1, wpath, MAX_PATH ) )
        {
            errno = ENOENT;
            return -1;
        }
        wpath[MAX_PATH] = L'\0';

        return _wstati64( wpath, (struct _stati64 *)buf );
    }
#endif
#ifdef HAVE_SYS_STAT_H
    const char *local_name = ToLocale( filename );

    if( local_name != NULL )
    {
        int res = deref ? stat( local_name, (struct stat *)buf )
                       : lstat( local_name, (struct stat *)buf );
        LocaleFree( local_name );
        return res;
    }
    errno = ENOENT;
#endif
    return -1;
}


int utf8_stat( const char *filename, void *buf)
{
    return utf8_statEx( filename, buf, VLC_TRUE );
}

int utf8_lstat( const char *filename, void *buf)
{
    return utf8_statEx( filename, buf, VLC_FALSE );
}

/**
 * utf8_*printf: *printf with conversion from UTF-8 to local encoding
 */
static int utf8_vasprintf( char **str, const char *fmt, va_list ap )
{
    char *utf8;
    int res = vasprintf( &utf8, fmt, ap );
    if( res == -1 )
        return -1;

    *str = ToLocaleDup( utf8 );
    free( utf8 );
    return res;
}

int utf8_vfprintf( FILE *stream, const char *fmt, va_list ap )
{
    char *str;
    int res = utf8_vasprintf( &str, fmt, ap );
    if( res == -1 )
        return -1;

    fputs( str, stream );
    free( str );
    return res;
}

int utf8_fprintf( FILE *stream, const char *fmt, ... )
{
    va_list ap;
    int res;

    va_start( ap, fmt );
    res = utf8_vfprintf( stream, fmt, ap );
    va_end( ap );
    return res;
}


static char *CheckUTF8( char *str, char rep )
#define isutf8cont( c ) (((c) >= 0x80) && ((c) <= 0xBF)) 
{
    unsigned char *ptr, c;

    assert (str != NULL);

    ptr = (unsigned char *)str;
    while( (c = *ptr) != '\0' )
    {
        /* US-ASCII, 1 byte */
        if( c <= 0x7F )
            ptr++; /* OK */
        else
        /* 2 bytes */
        if( ( c >= 0xC2 ) && ( c <= 0xDF ) )
        {
            c = ptr[1];
            if( isutf8cont( c ) )
                ptr += 2; /* OK */
            else
                goto error;
        }
        else
        /* 3 bytes */
        if( c == 0xE0 )
        {
            c = ptr[1];
            if( ( c >= 0xA0 ) && ( c <= 0xBF ) )
            {
                c = ptr[2];
                if( isutf8cont( c ) )
                    ptr += 3; /* OK */
                else
                    goto error;
            }
            else
                goto error;
        }
        else
        if( ( ( c >= 0xE1 ) && ( c <= 0xEC ) ) || ( c == 0xEC )
         || ( c == 0xEE ) || ( c == 0xEF ) )
        {
            c = ptr[1];
            if( isutf8cont( c ) )
            {
                c = ptr[2];
                if( isutf8cont( c ) )
                    ptr += 3; /* OK */
                else
                    goto error;
            }
            else
                goto error;
        }
        else
        if( c == 0xED )
        {
            c = ptr[1];
            if( ( c >= 0x80 ) && ( c <= 0x9F ) )
            {
                c = ptr[2];
                if( isutf8cont( c ) )
                    ptr += 3; /* OK */
                else
                    goto error;
            }
            else
                goto error;
        }
        else
        /* 4 bytes */
        if( c == 0xF0 )
        {
            c = ptr[1];
            if( ( c >= 0x90 ) && ( c <= 0xBF ) )
            {
                c = ptr[2];
                if( isutf8cont( c ) )
                {
                    c = ptr[3];
                    if( isutf8cont( c ) )
                        ptr += 4; /* OK */
                    else
                        goto error;
                }
                else
                    goto error;
            }
            else
                goto error;
        }
        else
        if( ( c >= 0xF1 ) && ( c <= 0xF3 ) )
        {
            c = ptr[1];
            if( isutf8cont( c ) )
            {
                c = ptr[2];
                if( isutf8cont( c ) )
                {
                    c = ptr[3];
                    if( isutf8cont( c ) )
                        ptr += 4; /* OK */
                    goto error;
                }
                else
                    goto error;
            }
            else
                goto error;
        }
        else
        if( c == 0xF4 )
        {
            c = ptr[1];
            if( ( c >= 0x80 ) && ( c <= 0x8F ) )
            {
                c = ptr[2];
                if( isutf8cont( c ) )
                {
                    c = ptr[3];
                    if( isutf8cont( c ) )
                        ptr += 4; /* OK */
                    else
                        goto error;
                }
                else
                    goto error;
            }
            else
                goto error;
        }
        else
            goto error;

        continue;

error:
        if( rep == 0 )
            return NULL;
        *ptr++ = '?';
        str = NULL;
    }

    return str;
}

/**
 * EnsureUTF8: replaces invalid/overlong UTF-8 sequences with question marks
 * Note that it is not possible to convert from Latin-1 to UTF-8 on the fly,
 * so we don't try that, even though it would be less disruptive.
 *
 * @return str if it was valid UTF-8, NULL if not.
 */
char *EnsureUTF8( char *str )
{
    return CheckUTF8( str, '?' );
}


/**
 * IsUTF8: checks whether a string is a valid UTF-8 byte sequence.
 *
 * @param str nul-terminated string to be checked
 *
 * @return str if it was valid UTF-8, NULL if not.
 */
const char *IsUTF8( const char *str )
{
    return CheckUTF8( (char *)str, 0 );
}


/**
 * UTF32toUTF8(): converts an array from UTF-32 (host byte order)
 * to UTF-8.
 *
 * @param src the UTF-32 table to be converted
 * @param len the number of code points to be converted from src
 * (ie. the number of uint32_t in the table pointed to by src)
 * @param newlen an optional pointer. If not NULL, *newlen will
 * contain the total number of bytes written.
 *
 * @return the result of the conversion (must be free'd())
 * or NULL on error (in that case, *newlen is undefined).
 */
static char *
UTF32toUTF8( const uint32_t *src, size_t len, size_t *newlen )
{
    char *res, *out;

    /* allocate memory */
    out = res = (char *)malloc( 6 * len );
    if( res == NULL )
        return NULL;

    while( len > 0 )
    {
        uint32_t uv = *src++;
        len--;

        if( uv < 0x80 )
        {
            *out++ = uv;
            continue;
        }
        else
        if( uv < 0x800 )
        {
            *out++ = (( uv >>  6)         | 0xc0);
            *out++ = (( uv        & 0x3f) | 0x80);
            continue;
        }
        else
        if( uv < 0x10000 )
        {
            *out++ = (( uv >> 12)         | 0xe0);
            *out++ = (((uv >>  6) & 0x3f) | 0x80);
            *out++ = (( uv        & 0x3f) | 0x80);
            continue;
        }
        else
        if( uv < 0x110000 )
        {
            *out++ = (( uv >> 18)         | 0xf0);
            *out++ = (((uv >> 12) & 0x3f) | 0x80);
            *out++ = (((uv >>  6) & 0x3f) | 0x80);
            *out++ = (( uv        & 0x3f) | 0x80);
            continue;
        }
        else
        {
            free( res );
            return NULL;
        }
    }
    len = out - res;
    res = realloc( res, len );
    if( newlen != NULL )
        *newlen = len;
    return res;
}

/**
 * FromUTF32(): converts an UTF-32 string to UTF-8.
 *
 * @param src UTF-32 bytes sequence, aligned on a 32-bits boundary.
 *
 * @return the result of the conversion (must be free()'d),
 * or NULL in case of error.
 */
char *FromUTF32( const uint32_t *src )
{
    const uint32_t *in;
    size_t len;

    /* determine the size of the string */
    for( len = 1, in = src; *in; len++ )
        in++;

    return UTF32toUTF8( src, len, NULL );
}

/**
 * UTF16toUTF8: converts UTF-16 (host byte order) to UTF-8
 *
 * @param src UTF-16 bytes sequence, aligned on a 16-bits boundary
 * @param len number of uint16_t to convert
 */
static char *
UTF16toUTF8( const uint16_t *in, size_t len, size_t *newlen )
{
    char *res, *out;

    /* allocate memory */
    out = res = (char *)malloc( 3 * len );
    if( res == NULL )
        return NULL;

    while( len > 0 )
    {
        uint32_t uv = *in;

        in++;
        len--;

        if( uv < 0x80 )
        {
            *out++ = uv;
            continue;
        }
        if( uv < 0x800 )
        {
            *out++ = (( uv >>  6)         | 0xc0);
            *out++ = (( uv        & 0x3f) | 0x80);
            continue;
        }
        if( (uv >= 0xd800) && (uv < 0xdbff) )
        {   /* surrogates */
            uint16_t low = GetWBE( in );
            in++;
            len--;

            if( (low < 0xdc00) || (low >= 0xdfff) )
            {
                *out++ = '?'; /* Malformed surrogate */
                continue;
            }
            else
                uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000;
        }
        if( uv < 0x10000 )
        {
            *out++ = (( uv >> 12)         | 0xe0);
            *out++ = (((uv >>  6) & 0x3f) | 0x80);
            *out++ = (( uv        & 0x3f) | 0x80);
            continue;
        }
        else
        {
            *out++ = (( uv >> 18)         | 0xf0);
            *out++ = (((uv >> 12) & 0x3f) | 0x80);
            *out++ = (((uv >>  6) & 0x3f) | 0x80);
            *out++ = (( uv        & 0x3f) | 0x80);
            continue;
        }
    }
    len = out - res;
    res = realloc( res, len );
    if( newlen != NULL )
        *newlen = len;
    return res;
}


/**
 * FromUTF16(): converts an UTF-16 string to UTF-8.
 *
 * @param src UTF-16 bytes sequence, aligned on a 16-bits boundary.
 *
 * @return the result of the conversion (must be free()'d),
 * or NULL in case of error.
 */
char *FromUTF16( const uint16_t *src )
{
    const uint16_t *in;
    size_t len;

    /* determine the size of the string */
    for( len = 1, in = src; *in; len++ )
        in++;

    return UTF16toUTF8( src, len, NULL );
}