C / C++ / MFC

12-Jun-09 3:40

Hello,

previously i have posted in wrong forum.
Problem is - User can input any number, program need to take decision based on inputs.

Let suppose input is stored in variable n;
based on value of n some function will be called.
program logic can be -

if(n > 1 && n < 10 )
callfun1();
if(n > 11 && n < 20 )
callfun2();
if(n > 25 && n < 30 )
callfun3();
if(n > 33 && n < 38 )
callfun4();
if(n > 1 && n < 10 )
callfun5();
.
.
.
and so on
if(n > minlimit && n < maxlimit )
callfunX();

Is there any easy way to avoid if else chain to do similar work.
Or is there any way to change the limits in if condition at some central place something using #define.

Please provide your inputs/help to help me finding other ways to optimize programming.

Thank
Vikas

vicky

harold aptroot12-Jun-09 4:05

harold aptroot

12-Jun-09 4:05

You could arrange them in such a way that they effectively do a binary search. That wouldn't bring the total number of tests down but it would run faster (singe every time execution flows through it it can only take 1 path, so O(log n) steps instead of O(n))

Or you could fill a giant array with function pointers for O(1) time.

Or maybe you could change the functions to not need to be selected anymore (using funny bit magic if necessary) but whether that's possible depends on the function of course.

KarstenK12-Jun-09 4:21

KarstenK

12-Jun-09 4:21

a little help is the use of the keyword else Blush | :O

if(n > 1 && n < 10 )
{
callfun1();
}
else if(n > 11 && n < 20 )
....

and sort it so that the most likely cases coming first.

AFAIK the compiler resolves switch in a similar way.

Best is the use of enums. Cool | :cool:

Press F1 for help or google it.
Greetings from Germany

Repost

led mike12-Jun-09 4:23

12-Jun-09 4:23

You ignored the responses you got on this question the first time. Reposting is rude. Perhaps you can explain why you ignored the previous responses and still reposted the same question.

Re: Repost

vikasvds14-Jun-09 21:48

14-Jun-09 21:48

I am very sorry for my re-post.

I have re-posted my question because at other discussion thread i was getting responses in terms of C# and .Net, but i need response in terms of pure C.

I have not done it intentionally, because of web page movement by mistake i've clicked at C# discussion link and posted the question instead of C/C++ category.

Thank
Vikas

vicky

Re: Repost

led mike15-Jun-09 5:10

15-Jun-09 5:10

vikasvds wrote:
I am very sorry for my re-post.

Ok, but have you begun to understand that OOD/OOP has solutions to writing conditional code?

molesworth12-Jun-09 4:30

molesworth

12-Jun-09 4:30

You probably need to take a step back, and think about what you're trying to achieve, and lay it out in a more logical pattern. Perhaps you can convert the value 'n' into a unique parameter for a switch statement.

There's also nothing wrong with using "else" so I'm not sure why you'd want to avoid it...

There are three kinds of people in the world - those who can count and those who can't...

led mike12-Jun-09 4:35

12-Jun-09 4:35

molesworth wrote:
There's also nothing wrong with using "else" so I'm not sure why you'd want to avoid it...

Here are some quotes from Kent Beck in his book Implementation Patterns[^]

If/then and switch statements are the simplest form of instance –specific behavior….
The more paths through a program the less likely the program is to be correct….The proliferation of conditionals reduces reliability….This problem is compounded when conditionals are duplicated.

These problems can all be eliminated by converting the conditional logic to messages, either with subclasses or delegation.

molesworth12-Jun-09 6:19

molesworth

12-Jun-09 6:19

led mike wrote:
The more paths through a program the less likely the program is to be correct

Aye, that's very true, and anything that can reduce complexity will also reduce the likelihood of bugs, and improve readability and, more importantly, testability.

However, sometimes there's no way to improve on an if/else construct, and when used sensibly they're the right tools for the job. When you get into a morass of nested if/elseif/elseif... blocks, then you certainly do introduce problems.

Sheesh! Next thing you'll be telling me "goto" is a bad idea Smile | :)

There are three kinds of people in the world - those who can count and those who can't...

CPallini12-Jun-09 6:42

12-Jun-09 6:42

I wouldn't subclass a if.
Big Grin | :-D

If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.

This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong.
-- Iain Clarke

[My articles]

led mike12-Jun-09 8:23

12-Jun-09 8:23

Why not? It's simple

class MyIf : public if
{
};

IT'S FRIDAY!!!!! Beer | [beer]

CPallini12-Jun-09 22:39

12-Jun-09 22:39

And now saturday here Beer | [beer]

:pizza:

:whisky:

led mike15-Jun-09 5:08

15-Jun-09 5:08

Wow, why would someone vote this a '1'? OMG | :OMG:

CPallini15-Jun-09 8:39

15-Jun-09 8:39

~~I suppose it is a personal attack against pizza , so, basically, ignorance...~~

[added]

led mike wrote:
Wow, why would someone vote this a '1'?

...when is available a really idiotic post to target [^]? Laugh | :laugh:

[/added]

Lutosław12-Jun-09 8:31

Lutosław

12-Jun-09 8:31

molesworth wrote:

There's also nothing wrong with using "else" so I'm not sure why you'd want to avoid it...

Adding elses to this code would dramatically change it's flow. Note that for n in a range (1,10) all callfuncs would be called.

Greetings - Jacek Gajek

CPallini12-Jun-09 5:46

12-Jun-09 5:46

Have you something personal against if-else?

vikasvds15-Jun-09 2:32

15-Jun-09 2:32

CPallini wrote:
Have you something personal against if-else?

Yes, i don't like if-else
I hate them Dead | X|

vicky

CPallini15-Jun-09 2:59

15-Jun-09 2:59

Try

#define WHATEVER_YOU_LIKE_AS_IF     if
#define WHATEVER_YOU_LIKE_AS_ELSE   else

and then, for instance

WHATEVER_YOU_LIKE_AS_IF (i>0 && i<10)
//..
WHATEVER_YOU_LIKE_AS_ELSE WHATEVER_YOU_LIKE_AS_IF (i>11 && i<20)
//..
//...

Arun Singh K13-Jun-09 1:18

Arun Singh K

13-Jun-09 1:18

You can try something like this

int func1()
{
return 0;
}
int func2()
{
return 0;
}

typedef int (*funcptr)();
struct condition
{
int low;
int high;
funcptr func;
};

condition arr[] = {
{ 10, 20, func1 },
{ 20, 30, func2 } // Can add more data here
};

int n=17;

for (int i=0;i<2;++i)
{
if (n > arr[i].low && n < arr[i].high)
{
arr[i].func();
// If you want you can break here
}
}

-Arun

vikasvds14-Jun-09 21:30

14-Jun-09 21:30

Thank you very much all of your response,

The approach Arun has suggested is i think the way i was looking for as best and fasted approach.

Thanks Arun to help me suggesting this approach, it's really nice and definitely fastest method uses full potential of C/C++ language.

Thanks again.
Vikas

vicky

vikasvds14-Jun-09 21:32