I need to know if someone can explain s:='Sér'; ShowMessage(s.Length.ToString)

[sergio.garcia]

Shouldn't the result of the programming code below be 3?
var
  s:string;
begin
  s:='Sér';  ShowMessage(s.Length.ToString);
end;

[Warfley]

The length of the String is not the length in (printed) characters, but the length in bytes. And in UTF-8 one character can be represented by up to 6 bytes (even though currently only up to 4 bytes are used, but it may be extended afterwards). E.g. '$' is 1 byte long, '£' is two bytes and '€' is three bytes long. Generally only 7-bit ASCII characters are also 1 byte in unicode (i.e. only the "american" alphabet).

And even the with the utf-8 character length, not all characters are printable, e.g. the é, while there is a unicode char to represent this, it can also be a combination of ´+e using character combination.

So general rule of thumb, when dealing with unicode, you can't tell from the length of a string how much characters will be printed

[paweld]

therefore, if you know that there may be non-ASCII characters in the string use the UTF8Length function.

Code: Pascal  [Select][+][-]

1. uses  
2.   LazUTF8;  
3.    
4. var
5.   s: String;
6. begin
7.   s := 'Sér';    
8.   ShowMessage(UTF8Length(s).ToString);
9. end;

[Remy Lebeau]

And in UTF-8 one character can be represented by up to 6 bytes (even though currently only up to 4 bytes are used, but it may be extended afterwards).

The original UTF-8 spec allowed for up to 6 bytes (technically, it could be unlimited!), but RFC 3629 restricts UTF-8 to 4 bytes max for compatibility with UTF-16, which physically can't exceed codepoint U+10FFFF.  I don't see them ever extending UTF-8 beyond that, unless Unicode itself eventually grows beyond U+10FFFF and consequently has to deprecate UTF-16.