Delphi - check if a Unicode character occurs in a set of characters?

user_18728572 0
Viewed 318

This code works good with Delphi-7 (until Delphi had Unicode support):

Value := edit1.Text[1];
    
if Value in ['м', 'ж'] then ...

'м', 'ж' - cyrillic symbols

But this construction doesn't work with Unicode charachter.
I try a lot of things, but they are doesn't work.
I also tried changing the value types to "Char" and "AnsiChar".

Doesn't work:

const
  MySet : set of WideChar = [WideChar('м'), WideChar('ж')];
begin
  Value := edit1.Text[1];
  if Value in MySet then ...

Doesn't work:

if AnsiChar(Value) in ['м', 'ж'] then ...

Doesn't work:

if CharInSet(Value, ['м', 'ж']) then ...

But this works good:

  if (Value = 'м') or (Value = 'ж') then ...

Whether there is an opportunity to check up UNICODE character by use of a SET in the modern versions of Delphi?
Or should we check each character individually?

My Delphi version is 10.4 update 2 Community Edition

delphi-7 unicode delphi
2 Answers

A Delphi set type can only handle a maximum of 256 values, so it cannot be used for handling Unicode characters. For handling Unicode, the System.Character unit provides various methods and helpers.

For this particular case, there is an IsInArray() character helper you can use. Instead of declaring a set of characters, you will need to declare an array of characters:

var
  ch: Char;
  a: array of Char;
  s: string;
begin
  a := ['м', 'ж'];
  s := 'abcж';
  for ch in s do
    if ch.IsInArray(a) then ...
end;

Note: Delphi XE7 introduced additional language support for initializing and working with dynamic arrays, and square brackets can also be used for simpler array initialization. In the context of above example, ['м', 'ж'] is not a set, but an array of wide characters.

user_42672440

check if a Unicode character occurs in a set of characters?

Do you mean a Delphi set?

In general, it is impossible to have a set of X where the base type X has more than 256 possible distinct values. So set of Byte is fine, but set of Word isn't possible. Since there are 256 * 256 distinct wide character values, it is therefore impossible to have a set of wide characters. (If this were indeed possible, a variable of such a set type would be 8 kB in size. That would be an unusually large variable.)

Since there is no such thing as "Delphi set of Unicode characters", the question "How to see if a character belongs to a Delphi set of Unicode characters" doesn't make sense.

Or do you simply mean a mathematical set?

If so, of course this is possible, but you cannot use a Delphi set to represent the mathematical set of characters. Instead, you need to use some other data type. One possibility is a simple array, if you don't mind its O(n) characteristics.

user_2828480
Related