Hi Andrus,

If you have VFP 9 then: STRCONV(cExpression, nConversionSetting [, nRegionalIdentifier [, nRegionalIDType]])

nRegionalIDType it's a flag if nRegionalIdentifier is Locale ID, codepage or FontCharSet. (modified)

 cres='?'
 ON ERROR note
 cres = STRCONV(CHR(0xc3)+CHR(0x83)+ CHR(0xc2)+CHR(0x9c) ,11 , 1257, 1)
 ON ERROR

IF cres#'?'
  ? i, cres
  ENDIF

ENDFOR

MartinaJ

>>>Returns question mark.
>>>Result should be Ü character, single byte 0xDC. Those 4 bytes per character are returned in WinHttpRequest.ResponseText. How to get 0xDC result ?
>>
>>Possibly the result you are getting is double width text, so you may need to do strconv() twice - first utf8 to double width, then that to 1257, I think the parameter is 2. I think I had cases when I had to do that (in my case, to 1250).
>
>Strconv() Parameter 2 does not accept code page. It requires locale id. Baltic locale id is not documented.
>
>Tried
>
>

>FOR i=1 TO 99999
>  cres='?'
>  ON ERROR note
>  cres = STRCONV( strconv(CHR(0xc3)+CHR(0x83)+ CHR(0xc2)+CHR(0x9c) ,11) ,2,i)
>  ON ERROR
>
>IF cres#'?'
>  ? i, cres
>  ENDIF
>
>ENDFOR
>

>
>This does not return any value.
>How to use your suggestion?