Hello,
I have string with ? and I want to delete all symbols to ?. When I delete symbol ? returns #176. How can I delete this symbol? Exampe: s := '49?45 ';

After: delete (s, 1, 2); --> '?45 '
After: delete (s, 1, 1); or delete (s, pos ('?', s), 1); -->

#176'45 '

I want to reseive only '45 '.
Regards,
Kamen

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20131010/73844d0b/attachment.html>

delete (s, pos ('?', s), length('?'));

Mattias

Kamen Kamen wrote

Hello, I have string with ? and I want to delete all symbols to ?. When I delete symbol ? returns #176. How can I delete this symbol?

What is the encoding of this string ? UTF8 ?

Delete uses a byte count, not a char count. If your string is encoded in UTF-8, then you will probably get the behaviour you see.

To my knowledge, there is currently no routine in the RTL that does this correct.

Michael.

Most probably '?' is a multi-byte UTF-8 symbol. You can declare it as a
  const Deg: UTF8string = '?';
and obtain the part past it using
   NextPos := Pos(Deg,s)+ Length(Deg);
   Tail := Copy(s, NextPos, Length(s));

When working with strings containing non-ASCII characters you should forget about Char and character (byte) based functions. Use strings instead, even for single letters, and the string based functions (Pos, StringReplace...).

DoDi

And/or use the UTF8 functions in unit LazUtf8.

Mattias

As Lazarus works with UTF8, and the string functions like "pos" and "delete" count bytes rather than (visible) characters, you can't use the visible character position for these functions.

-Michael

Michael Schnell wrote

As Lazarus works with UTF8, and the string functions like "pos" and "delete" count bytes rather than (visible) characters, you can't use the visible character position for these functions.

As Mattias pointed out: in this case Use Utf8Delete(). It is in LazUtf8 unit.

Bart

As Lazarus works with UTF8, and the string functions like "pos" and "delete" count bytes rather than (visible) characters, you can't use the visible character position for these functions.

The "visible character position" is an amazingly complex topic. Gladfully most programmers don't need them for their daily work. You can use Delete and Pos for UTF8.

Mattias

Thanks, Mattias! Now is OK.

-------- ?????????? ????? --------
??: Mattias Gaertner
???????: Re: [Lazarus] String with ?
??:
????????? ??: ?????????, 2013, ???????? 10 11:57:04 EEST

On Thu, 10 Oct 2013 11:52:24 +0300 (EEST)

delete (s, pos ('?', s), length('?'));

Mattias

Note: UTF8Delete works in code points, not in visible characters. The "visible characters" depend on font and rendering engine.

Mattias

Thanks, Bart! Where I can find help for

LazUtf8 unit? I was looking here: http://www.freepascal.org/docs.var with no success.

-------- ?????????? ????? --------
??: Bart
???????: Re: [Lazarus] String with ?
??: Lazarus mailing list
????????? ??: ?????????, 2013, ???????? 10 12:38:39 EEST

As Mattias pointed out: in this case Use Utf8Delete().

It is in LazUtf8 unit.

Bart

Thanks, Mattias!

-------- ?????????? ????? --------
??: Mattias Gaertner
???????: Re: [Lazarus] String with ?
??:
????????? ??: ?????????, 2013, ???????? 10 13:24:00 EEST

On Thu, 10 Oct 2013 11:38:39 +0200

Note: UTF8Delete works in code points, not in visible characters.

The "visible characters" depend on font and rendering engine.

Mattias