break or return?

[serkan sendur]

hi guys, i wonder if there is technically any difference between using break or return in order to quit a loop without reaching its condition?

[Nick Evan]

You're talking about c++?
The answer is yes:
break will only break out of the 'most nested loop' . For example:

int main()
{
    for (int i = 0; i < 10; i++){
        for (int j = 0; j < 10; j++){
            if (j == 5) 
                break;
            std::cout << "inner=" << j <<"\n";
        }
        std::cout << "outer=" << i<<"\n";
    }
    return 0;
}

will print

inner=0
inner=1
inner=2
inner=3
inner=4
[B]outer=0[/B]
inner=0
inner=1
//etc etc etc

As you can see, the break statement only breaks the inner loop when it reaches 4 ( <5 ). Then it start all over because the outer loop isn't done yet.

If you replace the previous code with:

int main()
{
    for (int i = 0; i < 10; i++){
        for (int j = 0; j < 10; j++){
            if (j == 5) 
                return 0;
            std::cout << "j=" << j <<"\n";
        }
        std::cout << "i=" << i<<"\n";
    }
    return 0;
}

the code will print:

inner=0
inner=1
inner=2
inner=3
inner=4

And it's done. You've used return, which means that it should return from the current function. In this case the function is 'main' so it quits the program.

[MosaicFuneral]

break; Just jumps, after an evalution. return(x); Will probably store the value were ever it feels appropriate, then tells the OS to return back to the called procedure.

[Salem]

> then tells the OS to return back to the called procedure.
The OS isn't involved in such things.

[serkan sendur]

thanks nick, they will make the same effect if there is only one level of loop, not nested, as a conclusion..

[Nick Evan]

thanks niek, they will make the same effect if there is only one level of loop, not nested, as a conclusion..

No. Break will jump out of a loop, return will jump out of a function. So only if your function exists out of only a loop, will they have they same effect.

[serkan sendur]

No. Break will jump out of a loop, return will jump out of a function. So only if your function exists out of only a loop, will they have they same effect.

yeah that was what i wanted to say.

[kbshibukumar]

However, having multiple exit points for a function makes it difficult to maintain. There is a greater chance that you'll forget to do the clean ups (which will be usually at the end of the function).

[Narue]

>However, having multiple exit points for a function makes it difficult to maintain.
Forcing a single exit point where it doesn't make sense also makes the function difficult to maintain due to added unnatural complexity.

>There is a greater chance that you'll forget to do the clean ups
>(which will be usually at the end of the function).
In well designed code this isn't a problem. Most (if not all) of your "clean ups" should be handled in destructors anyway.

[kbshibukumar]

>However, having multiple exit points for a function makes it difficult to maintain.
Forcing a single exit point where it doesn't make sense also makes the function difficult to maintain due to added unnatural complexity.

>There is a greater chance that you'll forget to do the clean ups
>(which will be usually at the end of the function).
In well designed code this isn't a problem. Most (if not all) of your "clean ups" should be handled in destructors anyway.

Do you have destructor for a function to clean up the locals inside?

[Narue]

>Do you have destructor for a function to clean up the locals inside?
Functions don't have destructors, but you can get the same effect by wrapping resources that need to be cleaned up within a class. This is called RAII, and it's a key concept in good C++ design.