Hi friends,

I'm trying to create a function to replace accented characters by single characters, e.g:

? = a
? = e
? = c
? = n

I'm trying to use OnKeyDown and OnUTF8KeyPress, but is not working, so I made some tests and saw that the problem is in the component:

Code:
---------------------------------
procedure TForm1.Button1Click(Sender: TObject);
begin
  Edit2.Text := Edit1.text[1]
end;
---------------------------------
When: Edit1.Text = 'A', Edit2.Text = 'A';
When: Edit1.Text = '?', Edit2.Text = '', empty;

What can this be?

I started a thread here:
http://www.lazarus.freepascal.org/index.php/topic,8816.msg42665.html#msg42665

Silvio Clecio wrote

I'm trying to create a function to replace accented characters by single characters, e.g:

Maybe you should show how your code for this is.

The easiest way is making a large if block.

There is no problem here, Lazarus encodes strings using UTF-8. This code is simply completely flawed. The [] operator refers to bytes of the string, so you are only taking the first byte of the string and assigning it to the text of the TEdit, which obviously only works for single byte characters, but UTF-8 chars usually have 1, 2 or 3 bytes.

I suggest that you read more on UTF-8: http://en.wikipedia.org/wiki/UTF-8

Hello Silvio!

On Sun, Mar 7, 2010 at 7:29 PM, Silvio Clecio wrote: > I'm trying to create a function to replace accented characters by single > characters, e.g:

Try the UTF8-Tools units:

http://wiki.lazarus.freepascal.org/Theodp

At the end of the page there is an example "Case demo" on replacing accented characters.

Best regards!
-Marco

Hi Felipe

I know this link on the wiki, but it did not really help you solve the problem in TCustomEdit.

See this demo, I explain it better how to reproduce the error:

http://groups.google.com.br/group/lazarus-brasil/web/be_bug_tests.zip?hl=pt-BR

Thanks.

Hello Marco, thanks for the response :)

I'll test it, I give a return in a moment.

Will send the demos that I'm trying to do, please see how the error appears to be in TCustomEdit.

Sending (WithoutAccent-pt_BR) in a moment...

Hello Lazarus-List,

Sunday, March 7, 2010, 11:57:45 PM, you wrote:

SC> I know this link on the wiki, but it did not really help you solve the SC> problem in TCustomEdit.
SC> See this demo, I explain it better how to reproduce the error: SC> http://groups.google.com.br/group/lazarus-brasil/web/be_bug_tests.zip?hl=pt-BR

So your problem is that UTF8Key is not "realtime" modificable as "Key" in KeyPress event ?

As UTF8Key appears as "var" parameter it looks like a bug to me.

Meanwhile you can use KeyPress to perform the task converting key to UTF8, perform the changes needed and them convert back to ansi mode. This way when UTF8KeyPress accepts changed UTF8Key (if it should do it) you should simply change the routine position and remove the 2 conversion lines.

Marco, I tested now, is really an very good package :)

You have some example to replace the text in real time using this package?

Please see this, I'm trying to do and only works right with OnChange:

http://lazarus-brasil.googlegroups.com/web/be_bug_tests_II.zip (or: http://bit.ly/aknuUm)

I need to avoid the OnChange; up of the 500 characters this function slows.

Thanks again :)

You have some example to replace the text in real time using this package?

Sorry, I don't have an example like that, but the example listed on the page seems simple enough:

 s := TUTF8Scanner.Create(Memo1.text);
 s.FindChars := '???';
 repeat
   case s.FindIndex(s.Next) of
 {?} 0: s.Replace('oe');
 {?} 1: s.Replace('ae');
 {?} 2: s.Replace('ue');
   end;
 until s.Done;
 Memo1.Text := s.UTF8String;
 s.free;

Hope it helps!
-Marco

Hellow JoshyFun,

JoshyFun wrote

So your problem is that UTF8Key is not "realtime" modificable as "Key" in KeyPress event ? As UTF8Key appears as "var" parameter it looks like a bug to me.

Yes. For me too. And here em my PC, OnKeyPress works normal only for Windows (XP). :(

In a very simple test, I'm trying this:

[code]
const
CWithAccent: array [0..46] of string = ('?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?',
'?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?', '?',

    '?', '?', '?', '?', '?', '?', '?', '?', '?', '?');
  CWithoutAccent: array [0..46] of string = ('a', 'a', 'a', 'a', 'a',

'e', 'e',
'e', 'e', 'i', 'i', 'i', 'i', 'o', 'o', 'o', 'o', 'o', 'u', 'u', 'u', 'u',
'A', 'A', 'A', 'A', 'A', 'E', 'E', 'E', 'E', 'I', 'I', 'I', 'O', 'O', 'O',

    'O', 'O', 'U', 'U', 'U', 'U', 'c', 'C', 'n', 'N');

uses
  StrUtils;
procedure TForm1.Edit1KeyPress(Sender: TObject; var Key: char);
begin
  Key := StringsReplace(AnsiToUtf8(Key), CWithAccent, CWithoutAccent,
[rfReplaceAll])[1];
end;
[/code]
(Please, see well here: http://pastebin.com/qmu80j74)

But unfortunately is working only on Windows. :(

Hello Marco, no problems friend :)

I'm trying with TUTF8Scanner + OnKeyPress, and TUTF8Scanner + OnUTF8KeyPress, but not works. :(

Another test that fails:

procedure TForm1.Edit1UTF8KeyPress(Sender: TObject; var UTF8Key: TUTF8Char);
begin
  UTF8Key := '?';
end;

Appear the caracters of the current typing in Edit, and not '?'. :(

It seems to be a bug in OnUTF8KeyPress, the similar behaviour was reported at Russian Free Pascal forum. Would you please create a bug report?

Vladimir Zhirov wrote

It seems to be a bug in OnUTF8KeyPress, the similar behaviour was reported at Russian Free Pascal forum. Would you please create a bug report?

Hi friend,

Yes. Please see here:

http://bugs.freepascal.org/view.php?id=15932

Chances are that this code works.

procedure TForm1.Edit1Change(Sender: TObject);
var
  s :string;
  p :integer;
begin
  s := edit1.text;
  while utf8pos('?', s) > 0 do
  begin
    p := utf8pos('?', s);
    utf8delete(s, p, utf8length('?'));
    utf8insert('a', s, p);
  end;
  edit1.selstart := 100;
  edit1.text := s;
end;

2010/3/8 Silvio Clecio

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20100309/548459a1/attachment.html>

uses lclproc;

2010/3/9 Ant?nio

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20100309/e9425ec5/attachment.html>

Friend, thanks for the dedication in helping me :)

Your function works for these characters that you presented. And thank you again for your efforts to find a solution. At the moment I gave up of the On(UTF8) KeyPress, I will use OnChange.

I was avoiding OnChange that it becomes slow with more than 500 characters, but I found a way to solve with the help of you. See a full example with the solution here:

<http://www.silvioprog.com.br/downloads> File: "WithoutAccent-pt_BR.zip".

In my country we have many problems with accented characters in Lazarus, I to do an article (in my language) showing how to reduce the gaps especially in DB, and publish here:

<http://blog.silvioprog.com.br>

Thanks to everyone who contributed in this question ;D