[DGD] Scanf vs explode

[Neil McBride]

> I fashioned my creator() function in driver.c after the way the sscanf's
> were
> originally done, due to the way the directories have changed, it requires
> quite a few of them depending on which one is true, I was wondering if it
> would be more efficient to do a single explode, and a switch() instead.

So, uh, I couldn't sleep.  I was wondering about some code efficiency too
but never really looked into how to find a good way to test coding styles
for efficiency, until I couldn't sleep .....

In my insomniac driven boredom, I put the following little file together. 
Somewhat imaginatively, I called it test.c:

void t1() {
  int i, j;

  for(i = 0; i < 10; i++) {

    j = status()[23];
    if(i == 0) {
      this_user()->message(i + "\n");
    } else if(i == 1) {
      this_user()->message(i + "\n");
    } else if(i == 2) {
      this_user()->message(i + "\n");
    } else if(i == 3) {
      this_user()->message(i + "\n");
    } else if(i == 4) {
      this_user()->message(i + "\n");
    } else if(i == 5) {
      this_user()->message(i + "\n");
    } else if(i == 6) {
      this_user()->message(i + "\n");
    } else if(i == 7) {
      this_user()->message(i + "\n");
    } else if(i == 8) {
      this_user()->message(i + "\n");
    } else if(i == 9) {
      this_user()->message(i + "\n");
    }
    j -= status()[23];
    this_user()->message("Ticks used: " + j +"\n");
  }
}

void t2() {
  int i, j;

  for(i = 0; i < 10; i++) {

    j = status()[23];
    switch(i) {
      case 0:
        this_user()->message(i + "\n");
      break;
      case 1:
        this_user()->message(i + "\n");
      break;
      case 2:
        this_user()->message(i + "\n");
        this_user()->message(i + "\n");
      break;
      case 3:
        this_user()->message(i + "\n");
      break;
      case 4:
        this_user()->message(i + "\n");
      break;
      case 5:
        this_user()->message(i + "\n");
      break;
      case 6:
        this_user()->message(i + "\n");
      break;
      case 7:
        this_user()->message(i + "\n");
      break;
      case 8:
        this_user()->message(i + "\n");
      break;
      case 9:
        this_user()->message(i + "\n");
      break;
    }
    j -= status()[23];
    this_user()->message("Ticks used: " + j +"\n");
  }
}

void t3() {
  int i, j;

  for(i = 0; i < 10; i++) {

    j = status()[23];
    this_user()->message(((i == 0) ? i : (i == 1) ? i : (i == 2) ? i : (i
== 3) ? i : (i == 4) ? i : (i == 5) ? i : (i == 6) ? i : (
i == 7) ? i : (i == 8) ? i : (i == 9) ? i : i) + "\n");
    j -= status()[23];
    this_user()->message("Ticks used: " + j +"\n");
  }
}

Oh yeah, element 23 status() results is ST_TICKS or the remaining ticks. 
Was the only way I could think of at 2am to get a good indication.  Any
other way? :)  Another thing, it's obvious the if/then and switch
statements can be removed in the given example and just send the user his
message, but that's not the point now is it?

Then, the moment of compiling .....

> compile test.c
$33 = </usr/sarak/sys/test>
> code $33->t1();
0
Ticks used: 303
1
Ticks used: 307
2
Ticks used: 311
3
Ticks used: 315
4
Ticks used: 319
5
Ticks used: 323
6
Ticks used: 327
7
Ticks used: 331
8
Ticks used: 335
9
Ticks used: 338
$34 = nil
> code $33->t2();
0
Ticks used: 303
1
Ticks used: 303
2
Ticks used: 303
3
Ticks used: 303
4
Ticks used: 303
5
Ticks used: 303
6
Ticks used: 303
7
Ticks used: 303
8
Ticks used: 303
9
Ticks used: 303
$35 = nil
> code $33->t3();
0
Ticks used: 303
1
Ticks used: 307
2
Ticks used: 311
3
Ticks used: 315
4
Ticks used: 319
5
Ticks used: 323
6
Ticks used: 327
7
Ticks used: 331
8
Ticks used: 335
9
Ticks used: 339

The verdict: Switch wins hands down, even after the first iteration :)

I hope you've all enjoyed this little adventure into what goes on when I
can't sleep :)

*bow*
_________________________________________________________________
List config page:  http://list.imaginary.com/mailman/listinfo/dgd