Hi, I'm working with a "C" shared library that returns a pointer to char. It's all ok when not using special characters such as "?" or accents, but when I get these, they are replaced by "?" signs.

For example, if I receive an "?", the debugger shows this:

lStr := 'A'#209'oranza';

The correct string is:

lStr := 'A?oranza';

What can I do to show accented characters?.

Regards,

It's a known problem and already reported. here is no fix available yet. The debugger does can not yet display this.

Actually it is 2 problems.
1) is the "?"
2) the other is #209 etc

Well, I can live with the debugger problem, but is there a solution for the "?" sign end displaying in my app?.

If you get an ansi string, you must convert it to utf8 before giving it to the LCL

need to find the exact function

label.caption :- ansitoutf8(s) // assuming it is ansi in the system codepage

Yes, I've tried that, but it shows a box with an X inside where the ? should appear...

My codepage is es_AR.UTF-8.

u:unicodestring;
u:=Putf8Char(buf); // or PAnsiChar depends on what your C DLL returns
label.caption:=u; // implicit cast

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20130409/dd34b626/attachment.html>

No, I've tried all kinds of combinations of ansitoutf8, unicodestring, etc, but nothing works, all returns "?" or the box with the X in the place of "?" or accents.

Impertinent question really, but do you know for certain that the font assigned to your label has glyphs corresponding to each unicode entity your string contains?

I'm finding some weird results.

If I execute my program from an XFCE4-Terminal, I get the diamond with a "?" inside, but if I redirect the output of my program to vim (executed from the same xfce4-terminal) I can see the "?" (ASCII 209). Please see the attached screenshot, the upper part shows the console output, the lower part shows vim running on the same console.

So, aparently the text is ASCII and not UTF8 as I thought.

I hope this data helps to find a solution...

i was wondering the same thing but couldn't get it to phrase properly... the font table is what i was having trouble with expressing... thanks for clearly pointing out another possible problem area for UTF-8 glyph rendering... hopefully this will be documented as a possible pitfall in the UTF functions' documentation or basic explanation of operations...

I really don't know, I'm using the same font as my Terminal (where the ? is shown correctly). Apart from it, I checked with other different fonts, and Charsets, but got the same result.

maybe your terminal is having problems? have you tried another terminal package in place of FCEwhateveritwas?

El 09/04/2013 19:35, Leonardo M. Ram? escribi?:

Hello,

Maybe you have an ANSI string inside a string type declared as UTF8 coded, so try this:

var
   ansivar: ansistring;
begin
   ansivar:=lStr; //Your lStr
   SetCodePage(RawByteString(ansivar),CP_ACP,false);
   //Display the ansivar instead lStr to check.
end;

Try using ISO_8859_1ToUTF8 from the unit lconvencoding.pas

Or another routine from this unit, depending on the encoding.

Felipe Monteiro de Carvalho wrote

Try using ISO_8859_1ToUTF8 from the unit lconvencoding.pas Or another routine from this unit, depending on the encoding.

Great!, that worked. Thanks Felipe!.