Sorry, I'm not familiar with Unicode. Just give me a sample showing the problem and I'll see what I can do.
EMG
Board index FiveWin for Harbour/xHarbour AscW() function
Page: 1 2 (25 posts)
From the beginning, we are going in the wrong direction. I am sorry for this deviation.
This is not conversion to and from UTF8 but of WideChar. (16bit char encoding).
Windows internally works with WideChar, i.e., 64 bit (little endian) characters.
In 8bit notation
"AB" is "4142" in hex
In 16bit notation
"AB" is "41004200" in hex
Regards
G. N. Rao.
Hyderabad, India
G. N. Rao.
Hyderabad, India
So, what's the point?
EMG
BIN2W( cWideString ) should give the value equivalent to AscW()
Regards
G. N. Rao.
Hyderabad, India
G. N. Rao.
Hyderabad, India
This small program demonstrates that BIN2W() works like AscW()
#include "fivewin.ch"
function Main()
local cUtf8, cWide, nAsc
cUtf8 := "అ"
? cUtf8
cWide := UTF8TOUTF16( cUtf8 )
nAsc := BIN2W( cWide )
? nAsc
? "Proof", HB_UTF8CHR( nAsc )
return nilRegards
G. N. Rao.
Hyderabad, India
G. N. Rao.
Hyderabad, India
So, what about HB_UTF8ASC()? Do we need it?
EMG
Enrico Maria Giordano wrote:So, what about HB_UTF8ASC()? Do we need it?
EMG
We need it.
This may not be what Mr. Natter is looking for but we need it in xHarbour.
Harbour has both HB_UTF8CHR() and HB_UTF8ASC().
But xHarbour has only HB_UTF8CHR() but not HB_UTF8ASC().
Addition of this function to xHarbour will be useful.
Regards
G. N. Rao.
Hyderabad, India
G. N. Rao.
Hyderabad, India
Done. Please try xHarbour build 10253.
EMG
Thank you.
FWH is now using HB_UTF8CHR(). There is no problem either with Harbour or xHarbour.
Till now there is no need to use HB_UTF8ASC().
But, if and when the need arises, what shall we do? There will be many users of older versions of xHarbour. Looks like we need to force them to upgrade xHarbour.
But what about xharbour.com users? They will get unresolved externals issue.
Regards
G. N. Rao.
Hyderabad, India
G. N. Rao.
Hyderabad, India
They can use this code:
EMG
#pragma BEGINDUMP
#include "error.ch"
#include "hbapierr.h"
static BOOL utf8tou16nextchar( UCHAR ucChar, int * n, USHORT * uc )
{
if( *n > 0 )
{
if( ( ucChar & 0xc0 ) != 0x80 )
return FALSE;
*uc = ( *uc << 6 ) | ( ucChar & 0x3f );
( *n )--;
return TRUE;
}
*n = 0;
*uc = ucChar;
if( ucChar >= 0xc0 )
{
if( ucChar < 0xe0 )
{
*uc &= 0x1f;
*n = 1;
}
else if( ucChar < 0xf0 )
{
*uc &= 0x0f;
*n = 2;
}
else if( ucChar < 0xf8 )
{
*uc &= 0x07;
*n = 3;
}
else if( ucChar < 0xfc )
{
*uc &= 0x03;
*n = 4;
}
else if( ucChar < 0xfe )
{
*uc &= 0x01;
*n = 5;
}
}
return TRUE;
}
HB_FUNC( HB_UTF8ASC )
{
const char * pszString = hb_parc( 1 );
if( pszString )
{
HB_SIZE nLen = hb_parclen( 1 );
USHORT wc = 0;
int n = 0;
while( nLen )
{
if( ! utf8tou16nextchar( ( unsigned char ) *pszString, &n, &wc ) )
break;
if( n == 0 )
break;
pszString++;
nLen--;
}
hb_retnint( wc );
}
else
hb_errRT_BASE_SubstR( EG_ARG, 3012, NULL, HB_ERR_FUNCNAME, HB_ERR_ARGS_BASEPARAMS );
}
#pragma ENDDUMPEMG
Page: 1 2 (25 posts)