Thread: Removal of gotos

O_o

What? What does that have to do with using `break' or `continue'? What do they have to do with `switch'? Are you making some indirect association? ('break' -> `goto' <- 'case acase:')

[Edit]

I've got it now. You asked that because of the common "rule" that every `case' block must end with a `break'. It is a stupid rule and has nothing to do with the standard.

[/Edit]

Whatever.

If the situation is best illustrated by a `switch' statement, I naturally use a `switch' statement, but you'll have to explain how that necessitates `break' or `continue'. I've managed without them for some time.

Actually, come to think on it, most of my code is "library level" where a `switch' statement wouldn't be allowed for mechanical reasons (For example, in template code where you'd have to use a different construct simple because the `case' labels must be integer like.) or design reasons. (The related values are not constant/literal expressions.) Considering that, I suppose it would be safe to say I don't use `switch' statements.

Soma

Originally Posted by phantomotap

O_o

What? What does that have to do with using `break' or `continue'?

Switch's have cases which use "break." That's all he meant.

The main reason goto's are bad is because they produce spaghetti code. "break" and "continue" do not produce spaghetti code. They are quite uniform and generally not too hard to follow. Using a few here and there (such as for/while/if statements, switch case/break is perfectly fine) is fine (even using a small amount of goto's under certain circumstances has been argued, even though I've never encountered them).