C#

try
{
//...
}
catch (NullReferenceException ex)
{
    Console.WriteLine(ex.Source);
}
catch (ArgumentException ex)
{
    Console.WriteLine(ex.ParamName);
}

try
{
//...
}
//Not sure this is legal in Java, it isn't in c#, even without the "ex"s
catch (NullReferenceException ex, ArgumentException ex)
{
   // this would fail anyway as NullReferenceException does not have a paramname property.
   Console.WriteLine(ex.ParamName);
}

private static void Main(string[] args)
{
    Exception ex = null;
    try
    {
        throw new InvalidOperationException("oops");
    }
    catch (InvalidOperationException myEx)
    {
        ex = myEx;
        goto Something;
    }
    catch (InvalidTimeZoneException myEx)
    {
        ex = myEx;
        goto Something;
    }
    catch (Exception myEx)
    {
        Console.WriteLine("Exception: " + myEx.ToString());
    }

    return;
Something:
    Console.WriteLine(ex.ToString());
    return;

}

private static void Main(string[] args)
{
    try
    {
        throw new InvalidOperationException("oops");
    }
    catch (Exception ex)
    {
        switch (ex.GetType().ToString())
        {
            case "System.InvalidOperationException":
                goto case "System.InvalidTimeZoneException";
            case "System.InvalidTimeZoneException":
                Console.WriteLine(ex.ToString());
                return;
            case "System.Exception":
                Console.WriteLine("Exception: " + ex.ToString());
                break;
            default:
                Console.WriteLine("Hmmm.");
                break;
        }
    }
}

private static void Main(string[] args)
{
    try
    {
        throw new InvalidOperationException("oops");
    }
    catch (Exception ex)
    {
        switch (ex.GetType().ToString())
        {
            case "System.InvalidOperationException":
            case "System.InvalidTimeZoneException":
                Console.WriteLine(ex.ToString());
                return;
            case "System.Exception":
                Console.WriteLine("Exception: " + ex.ToString());
                break;
            default:
                Console.WriteLine("Hmmm.");
                break;
        }
    }
}

blackblizzard wrote:

Couldn't you just do this?

blackblizzard wrote:

GetType() is computationally expensive

blackblizzard wrote:

I'll just use a method.

aspdotnetdev wrote:

Hmm, I thought C# complained when you let it fall through like that, but it seems the compiler only complains when you have something under the first case above and no "break". That is, it expects that all the logic be placed on the bottom-most of the cases you fall through. Good to know... thanks!

aspdotnetdev wrote:

Not sure if that's true. I know Reflection.Emit and dynamically building types and things of that sort are expensive, but I'm not sure GetType is all that expensive, as it is pretty much known at compile time. As far as derived classes, that shouldn't be any more expensive than other inherited method call. Depends on how .NET implements GetType under the hood, but I tend to think it would use the faster technique. The string comparison done during the switch statement could slow things down marginally, but when you are dealing with exception cases, you typically don't need to worry about speed (assuming you are not using exceptions to control the typical flow of your program).

DX Roster wrote:

first of all use POP control protocol for send mail because it provide facility to send mail as well as store mail..

eyalbi007 wrote:

I'll add myself as recipient (regular or CC or BCC), but it's a bit dirty...