Catching an integer overflow exception in FreePascal

Question

I'm trying this simple program to catch exceptions.

The division by zero exception is successfully caught. However, I cannot seem to catch the integer overflow.

I understand that a SmallInt type integer would overflow after +32767 since the largest signed value = 2^15 - 1 (i.e., +32767).

Here's the code:

Program ss;

Var

  Price, Qty: SmallInt;
  Rate: Real;

Begin
  Try

    write('Enter Total price '); Readln(Price);
    Write('Enter qty '); Readln(Qty);

    Rate := Price / Qty;
    Writeln('Single item = ', Rate:0:2);

  except
    On EDivByZero Do Writeln('qty must be a valid num above 0');
    On EIntOverflow Do Writeln('too large')
  end;
End.

However, it appears that the overflow exception is not caught. Rather, the overflowing value is taken as a negative number and the code continues as if nothing is wrong.

Here's the output:

Enter Total price 32768

Enter qty 1

Single item = -32768.00

Press any key to return to IDE

Am I doing something wrong in the code? What is the correct method of catching the integer overflow in FreePascal?

Thanks!

I'm not an expert in Free Pascal but my guess is that the overflow exception is only triggered as a result of a calculation. Dividing by 1 doesn't change the value so it doesn't generate an exception. Try adding or subtracting or multiplying values that are in range but the result of the operation is out of range. For example x := 32767; x := x+1 — Stuart
– Stuart, Commented Jan 7, 2016 at 5:20
@Stuart Thanks for your input. I tried multiplying 17000 by 2 and storing it in a SmallInt. It gave the output as -31536. So in other words, it overflows but not run into the exception-handling code. — itsols
– itsols, Commented Jan 7, 2016 at 5:42
Why don't you try the code I suggested. 17000 * 2 will probably be performed as integer multiplication, and when you store the result in a smallint, it is then truncated to a smallint Truncating an integer is not a calculation and so it might not generate a integer overflow exception. — Stuart
– Stuart, Commented Jan 7, 2016 at 6:20
@Stuart For your example, the output is -32768 (which is the overflowing value) and no exception is triggered. I recall some older language (I think it was VB6 but I cannot confirm this right now) giving a similar result. Thanks again. — itsols
– itsols, Commented Jan 7, 2016 at 9:44

Marco van de Voort · Accepted Answer · 2016-01-07 10:01:51Z

1

You must include unit sysutils, as it contains the code that changes the deepest kind of runtime errors into catchable language exceptions

answered Jan 7, 2016 at 10:01

Marco van de Voort

26.5k6 gold badges62 silver badges95 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

itsols Over a year ago

Thanks Marco. Apart from including this unit, is there anything I need to do? I've included it and it works without any change.

itsols Over a year ago

For what it's worth, my compiler version shows 2.6.2-8 on the Help > About menu

Collectives™ on Stack Overflow

Catching an integer overflow exception in FreePascal

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related