How to optimise strcmp c-string comparison

Hi guys I have written some code and it works fine, the only problem is I think is not that mobile, I mean it looks big for what it does. Basically I have three names in c-string and I like to arrange them alphabetically, but I am not allowed to use arrays.
So the question is how to optimise my code? Thanks

Code:

#include<iostream>
#include<cstring>

using namespace std;

int main( )
{

	char name1[16] = {'\0'},
		name2[16] = {'\0'}, 
		name3[16] = {'\0'};

	cout <<"The program will display entered names in alphabetical order.\n";
	cout <<"Enter three names pressing the Return key after each name\n";
	cin >> name1 >> name2 >> name3;
	cout <<"The alphabetical order of the name is:\n";
	if(strcmp(name1, name2) < 0 && strcmp(name1, name3) < 0)        
	{
		if(strcmp(name2, name3) < 0)
			cout << name1 << endl << name2 << endl <<name3 << endl;
		else 
			cout << name1 << endl << name3 << endl << name2 << endl;
	}
	else if(strcmp(name2, name1) < 0 && strcmp(name2, name3) < 0)
	{
		if(strcmp(name1, name3) < 0)
			cout << name2 << endl << name1 << endl << name3 << endl;
		else 
			cout << name2 << endl << name3 << endl << name1 << endl;
	}
	else if(strcmp(name3, name1) < 0 && strcmp(name3, name2) < 0)
	{
		if(strcmp(name1, name2) < 0)
			cout << name3 << endl << name1 << endl << name2 << endl;
		else 
			cout << name3 << endl << name2 << endl << name1 << endl;
	}
	else if(strcmp(name1, name2) == 0 && strcmp(name2, name3) == 0)
	{
		cout << name1 << endl << name2 << endl <<name3 << endl;
	}
	else if(strcmp(name1, name2) == 0)
	{
		if(strcmp(name2, name3) < 0)
			cout << name1 << endl << name2 << endl <<name3 << endl;
		else 
			cout << name3 << endl << name1 << endl << name2 << endl;
	}
	else if(strcmp(name1, name3) == 0)
	{
		if(strcmp(name1, name2) < 0)
			cout << name1 << endl << name3 << endl << name2 << endl;
		else 
			cout << name2 << endl << name1 << endl << name3 << endl;
	}
	else if(strcmp(name2, name3) == 0)
	{
		if(strcmp(name2, name1) < 0)
			cout << name2 << endl << name3 << endl << name1 << endl;
		else
			cout << name1 << endl << name2 << endl <<name3 << endl;
	}
	return 0;
}

So the question is how to optimise my code? Thanks

Stop calling strcmp so much?

But then how I will do the comparison

1) You could store the results.

2) Why don't you just sort the names? You can do that in place using bubblesort.

Save the results of some of the comparisons, and use their meaning to determine the order. The logic is only a little different.

Note: This is a stupid idea but I am bored. Don't do this.

Code:

int bits = 0;
if (strcmp(name1, name2) < 0) bits++;
bits <<= 1;
if (strcmp(name1, name3) < 0) bits++;
bits <<= 1;
if (strcmp(name2, name3) < 0) bits++;

switch(bits) {
    case 7:
        cout << name1 << endl << name2 << endl <<name3 << endl;
    break;
    case 6:
        cout << name1 << endl << name3 << endl <<name2 << endl;
    break;
    case 4:
        cout << name3 << endl << name1 << endl <<name2 << endl;
    break;
    case 3:
        cout << name2 << endl << name1 << endl <<name3 << endl;
    break;
    case 1:
        cout << name2 << endl << name3 << endl <<name1 << endl;
    break;
    case 0:
        cout << name3 << endl << name2 << endl <<name1 << endl;
    break;
}

No really, don't do this.

Hi tabstop, thanks for the replay, I am tiring to understand the algorithm you used, I see the incremation and how it changes if it meets the if statement , but can you explain with couple of words, specially if the names are Ana, Boni, Roni it increments to 7…?? Please explain… this is not the bubble sort algorithm is it?

Originally Posted by tabstop

Note: This is a stupid idea but I am bored. Don't do this.

Code:

int bits = 0;
if (strcmp(name1, name2) < 0) bits++;
bits <<= 1;
if (strcmp(name1, name3) < 0) bits++;
bits <<= 1;
if (strcmp(name2, name3) < 0) bits++;

switch(bits) {
    case 7:
        cout << name1 << endl << name2 << endl <<name3 << endl;
    break;
    case 6:
        cout << name1 << endl << name3 << endl <<name2 << endl;
    break;
    case 4:
        cout << name3 << endl << name1 << endl <<name2 << endl;
    break;
    case 3:
        cout << name2 << endl << name1 << endl <<name3 << endl;
    break;
    case 1:
        cout << name2 << endl << name3 << endl <<name1 << endl;
    break;
    case 0:
        cout << name3 << endl << name2 << endl <<name1 << endl;
    break;
}

No really, don't do this.

Stupid, maybe..yet somehow ingenious.

Originally Posted by Svetlin

Hi tabstop, thanks for the replay, I am tiring to understand the algorithm you used, I see the incremation and how it changes if it meets the if statement , but can you explain with couple of words, specially if the names are Ana, Boni, Roni it increments to 7…?? Please explain… this is not the bubble sort algorithm is it?

No, this is not bubble sort. If you want to follow along, you should look at what << does (hint: it's not just for output).

Originally Posted by Bubba

Stop calling strcmp so much?

Originally Posted by Svetlin

But then how I will do the comparison

Use properties of strcmp(). With a bit of logic, it is possible to work out ways to call strcmp() less.

For example (assuming a, b are valid strings) then it is guaranteed

1) strcmp(a,b) will always return the same result, for a given pair of a and b.

2) (strcmp(a,b) > 0) == (strcmp(b,a) < 0)

There are other properties too if you want to work with three strings ..... tabstop actually used some of them

Ok now I understand what you are doing tabstop, you are using binary positions to calculate the possible choices…

Why use strcmp in the first place? Why use C-style strings?