Can anyone point out the flaw?

[a_aniq]

This is just a code written for my amusement purposes in my freetime. This code is for SPOJ's problem on Adding Reversed Numbers (code: ADDREV). But, I can't get where the program is wrong, you can test it after inputting second test case (the first test case works just fine). I am not some professional programmer, if it's a silly mistake, then also please explain it properly. I can trim the output using conditional modulo and division at the end phase, getting the right result. But where does this logic go wrong? The code goes as follows-

#include <stdio.h>
#include <stdlib.h>
int digits(int a){
    int k=1;
    int x=a;
    while (a%10==0){
        k--;
        a/=10;
    }
    while(x>9){
        x=x/10;
        k++;
    }
    return k;
}
int main() {
	int n,j=1;
	scanf("%d",&n);
	int *aa,*ba,*ca;
	while(n--) {
        int a,b,c,d,e,x,m=0;
        scanf("%d %d",&a,&b);
        c=digits(a);
        d=digits(b);
        if(c>d)
            e=c;
        else
            e=d;
        while (a%10==0&&b%10==0){
            a/=10;
            b/=10;
        }
        aa=calloc(e,sizeof(int));
        ba=calloc(e,sizeof(int));
        ca=calloc(e,sizeof(int));
        while(c--){
            aa[c]=a%10;
            a/=10;
        }
        while(d--){
            ba[d]=b%10;
            b/=10;
        }
        int r=0;
        for(x=0;x<e;x++){
            ca[x]=aa[x]+ba[x]+r;
            if(ca[e-1]==10)
                ca[e-1]=1;
            if(ca[e-1]>10)
                ca[e-1]=ca[e-1]/10+ca[e-1]%10*10;
            else if(ca[x]>9){
                ca[x]=ca[x]%10;
                r=1;
            }
            else r=0;
        }
        x=0;
        while(ca[x]==0)
            x++;
        int f=e-1;
        while(e--){
                if(ca[f]>9&&e==f){
                    m=m+ca[f];
                    j*=100;
                    e--;
                }
                m=m+ca[e]*j;
                j*=10;
                if(x==e)
                    break;
        }
        printf("%d\n",m);
    }
	return 0;
}

[leppyr64]

I am far from a professional coder too. I am a problem solver however.

You seem to indicate that there is a test case that you have tried that is failing. What is it?

[a_aniq]

In compiled Spoj classical problem solutions if you input 1 in the first line, it asks for the first test case. If you input 'n' then it asks for 'n' test cases, printing out the result after each case.
You can take any test case, if you have entered 'n'>1 in the first line of input
For example take the default input given in the SPOJ Problem page
Enter as input (First input denotes no. of test cases, then each test case consists of two inputs)
3
24
1
4358
754
305
794
Correct Output
34
1998
1
Received Output (to be corrected)
34
199800
1000000

[leppyr64]

        for(x=0;x<e;x++){
            ca[x]=aa[x]+ba[x]+r;
            if(ca[e-1]==10)
                ca[e-1]=1;
            if(ca[e-1]>10)
                ca[e-1]=ca[e-1]/10+ca[e-1]%10*10;
            else if(ca[x]>9){
                ca[x]=ca[x]%10;
                r=1;
            }
            else r=0;
        }

Why are you referencing ca[e-1]?

[a_aniq]

Well, considering that you have understood the logic and math behind my code completely, I am moving forward.
Ok, then I'm removing tit bits referencing to ca[e-1] and posting the modified code again

#include <stdio.h>
#include <stdlib.h>
int digits(int a){
    int k=1;
    int x=a;
    while (a%10==0){
        k--;
        a/=10;
    }
    while(x>9){
        x=x/10;
        k++;
    }
    return k;
}
int main() {
	int n,j=1;
	scanf("%d",&n);
	while(n--) {
        int a=0,b=0,c=0,d=0,e=0,x=0,m=0,r=0;
        scanf("%d %d",&a,&b);
        c=digits(a);
        d=digits(b);
        if(c>d)
            e=c;
        else
            e=d;
        while (a%10==0&&b%10==0){
            a/=10;
            b/=10;
        }
        int *aa=calloc(e,sizeof(int));
        int *ba=calloc(e,sizeof(int));
        int *ca=calloc(e,sizeof(int));
        while(c--){
            aa[c]=a%10;
            a/=10;
        }
        while(d--){
            ba[d]=b%10;
            b/=10;
        }
        for(x=0;x<e;x++){
            ca[x]=aa[x]+ba[x]+r;
            if(ca[x]>9){
                ca[x]=ca[x]%10;
                r=1;
            }
            else r=0;
        }
        x=0;
        while(ca[x]==0)
            x++;
        int f=e-1;
        while(e--){
                m=m+ca[e]*j;
                j*=10;
                if(x==e)
                    break;
        }
        printf("%d\n",m);
    }
	return 0;
}

Well, now try giving these inputs:
2
405
794
425
796
Correct Outputs
1001
1221
Received Outputs through this program
100
122000
Does it clear your doubt?

[a_aniq]

I just realized that for a test case such as (405,794) my program in my first post gave me an output of 101.
I have corrected it. Actually the problem was 01 would be saved as 1 in the array block.
I am posting the corrected code below (Another conditional statement). But still the main problem due to which I started this forum still persists.

#include <stdio.h>
#include <stdlib.h>
int digits(int a){
    int k=1;
    int x=a;
    while (a%10==0){
        k--;
        a/=10;
    }
    while(x>9){
        x=x/10;
        k++;
    }
    return k;
}
int main() {
	int n,j=1;
	scanf("%d",&n);
	while(n--) {
        int a=0,b=0,c=0,d=0,e=0,x=0,m=0,r=0;
        scanf("%d %d",&a,&b);
        c=digits(a);
        d=digits(b);
        if(c>d)
            e=c;
        else
            e=d;
        while (a%10==0&&b%10==0){
            a/=10;
            b/=10;
        }
        int *aa=calloc(e,sizeof(int));
        int *ba=calloc(e,sizeof(int));
        int *ca=calloc(e,sizeof(int));
        while(c--){
            aa[c]=a%10;
            a/=10;
        }
        while(d--){
            ba[d]=b%10;
            b/=10;
        }
        for(x=0;x<e;x++){
            ca[x]=aa[x]+ba[x]+r;
            if(ca[e-1]==10)
                ca[e-1]=-1;
            else if(ca[e-1]>10)
                ca[e-1]=ca[e-1]/10+ca[e-1]%10*10;
            else if(ca[x]>9){
                ca[x]=ca[x]%10;
                r=1;
            }
            else r=0;
        }
        x=0;
        while(ca[x]==0)
            x++;
        int f=e-1;
        while(e--){
                if(ca[f]>9&&e==f){
                    m=m+ca[f];
                    j*=100;
                    e--;
                }
                else if(ca[f]==-1&&e==f){
                    m=m+1;
                    j*=100;
                    e--;
                }
                m=m+ca[e]*j;
                j*=10;
                if(x==e)
                    break;
        }
        printf("%d\n",m);
    }
	return 0;
}

[leppyr64]

In the latest code that you have posted you have the ca[e-1] references back in. Please fix it and repost yor code.

I cannot 100% understand your logic because you get WA. If your logic were 100% accurate and your code were 100% accurate then your code would be AC. If your logic is 100% correct then errors is code cause your WA. If your logic is not correct then your code cannot cause an AC.

All that is left for me is to point out to you inconsistencies that might reflect an issue in your code or logic.

A small side note:
This problem is written with fairly small values for a and b. Your algorithm would be required for large values of a and b (1000 digits?) but is overly complex for this specific problem. I am more than happy to continue to help you debug it but just wanted to let you know.

[a_aniq]

Firstly, I have submitted a very simple code for this problem by coding a function which reverses the number, and I got AC.
Secondly, I coded this for my own amusement, not to submit. I develop codes using my own logic and debug them, I love doing it. But, I failed to do so in this case, and that's why I posted it here.
This code works well in other compilers (including ideone), but when submitting it to SPOJ it gives SIGSEGV instead of giving wrong answer.
And now for the logic of my code:
Suppose you are dealing with 495 & 704 in input.
I can't understand what's wrong with referencing ca[e-1] which you are complaining about, I am just referencing the last element of the array. But I'm gonna try writing a program which doesn't reference it.
Then starting from hundredths place-
4+7=11;
Carry 1 onto tenths place;
9+0+1=10;
Again carry 1 onto ones place;
5+4+1=10
Again carry 1 onto zeroths place and add it to a 0;
Summing it up,
495
704
1001
If the no. of digits are different, then-
Let's consider 4358 & 794 as input
then we have to consider them as 4358 &7940,
Similarly, starting from the hundredths place, and progressing as in the previous example, we get-
4358
7540
1998
Side note:
Anyways, many many thanks for giving my first post this much attention, although it's not that much essential, I just couldn't understand what was wrong with this code, that's that. You can leave it if you don't feel like debugging it. I thought this forum would be much more active and fun with numerous members commenting and such, but alas.

[leppyr64]

First, changed the dimensions on aa,ba, and ca to e+1.

I changed this block:

        for(x=0;x<e;x++){
            ca[x]=aa[x]+ba[x]+r;
            if(ca[e-1]==10)
                ca[e-1]=-1;
            else if(ca[e-1]>10)
                ca[e-1]=ca[e-1]/10+ca[e-1]%10*10;
            else if(ca[x]>9){
                ca[x]=ca[x]%10;
                r=1;
            }
            else r=0;
        }

To this:

		for(x=0;x<e;x++){
			ca[x]=aa[x]+ba[x]+r;
			r = ca[x]/10;
			ca[x] %= 10;
		}

Then I changed this block (I'm not sure what you were doing in this block)

        int f=e-1;
        while(e--){
                if(ca[f]>9&&e==f){
                    m=m+ca[f];
                    j*=100;
                    e--;
                }
                else if(ca[f]==-1&&e==f){
                    m=m+1;
                    j*=100;
                    e--;
                }
                m=m+ca[e]*j;
                j*=10;
                if(x==e)
                    break;
        }

To this:

		while(x < e)
			m = m*10 + ca[x++];

The forum is only busy when people have questions about problems. The site itself is quite active. I'm not associated with the site, I'm just a user who moderates the forum. I find it quite fun. If you have any other questions, feel free to post.