Thread: switch question

>> Is that smart to do? Would you pros do that?
In some situations, yes. You have to be very careful, though, because it is an easy thing to miss. You should make it obvious that you purposefully didn't put in the break, and you should make it obvious that the code under the next case counts for multiple cases. For example, notice how the comments help indicate what you are doing so that when you or somebody else comes back to maintain or update the code, there is less chance of confusion.

Code:

switch (val)
{
    case 1:
    case 2:
    case 3:
        code();
        moreCode();
        break;

    case 4:
        doStuff();
        doOtherStuff();
        // fall through on purpose!

    // case 4:
    case 5:
        doMoreStuff();
        break;
}

If you can do something in a manner that is more clear than the above, you should choose that version instead. Even with the comments I don't like the above code.

Lookup "Duff's device" if you want to see switch/case "abuse"

Forgetting to place a break after each case will only cause it to "fall through" to the next one anyway. I always make sure I end my cases with a break, it makes good sense, unless of course the idea of the program is to test the weakness in a switch statement that is

Sometimes I forget to end the case before WM_DESTROY with a break, so my program just exits when the that case is activated. First time when this happened I was quite confused.