(newb) getting two diffrent answares with the same func

hello ..

i can't understand why but the same code with small adjustements give me diffrent answares :
when i use the C type of work char* complement(char* , char *); all works nice.
but when i use cpp func char* complement(foo &object) i get strange answares.

Code:

---

#include <iostream>
#include <string>

using namespace std;
class foo{
        char* number;
        int size;
        public:
                foo();
                foo(char *value);
                ~foo();
        char* complement(foo &object);
        char* complement(char* , char *);
        };

foo::foo(){
        number=NULL;
        size=0;
}
foo::foo(char *value){
        size=strlen(value)+1;
        number=new char[size];
        strcpy(number,value);
}

foo::~foo(){
        if (number!=NULL)
                delete [] number;
}
/*
 this works fine  but it afaik its ain't oop

 char *foo::complement(foo &object){
    char result=complement(this->number,object.number);
    return result;
 }

*/

char* foo::complement(foo &object){
       
        int second_size=object.size;
        char *result=NULL;
        int result_size=(size>second_size) ?size+1: second_size+1 ;//1 more for '\0'
        result=new char [result_size];
        int biggest_size=(size>second_size) ?size : second_size ;
        int smallest_size=(size<second_size) ?size : second_size ;
        char m[12]={'0','9','8','7','6','5','4','3','2','1','0','\0'};
        int flag=0;
        if(second_size<=size)
                {
                        for (int i=smallest_size-1;i>-1;i--)
                        {
                       
                                result[i+biggest_size-smallest_size]=m[ object.number[i]-'0'+flag];
                                if (result[i+biggest_size-smallest_size]!='0')
                                        flag=1;
                        }
                        for (int i=biggest_size - smallest_size-1;i>-1;i--)
                        {
                                result[i]='9';
                        }
                }
       
        if (second_size > size)
                {
                        for (int i=second_size-1;i>-1;i--)
                        {
                                result[i]=m[ object.number[i]-'0'+flag];
                                if (result[i]!='0')
                                        flag=1;
                        }
                }
       
        result[result_size -1]='\0';
       
        return result;

}


char* foo::complement(char *first,char *second){
        int size=strlen(first);
        int second_size=strlen(second);
        char *result=NULL;
        int result_size=(size>second_size) ?size+1: second_size+1 ;//1 more for '\0'
        result=new char [result_size];
        int biggest_size=(size>second_size) ?size : second_size ;
        int smallest_size=(size<second_size) ?size : second_size ;
        char m[12]={'0','9','8','7','6','5','4','3','2','1','0','\0'};
        int flag=0;
        if(second_size<=size){
        for (int i=smallest_size-1;i>-1;i--)
                {
                        result[i+biggest_size-smallest_size]=m[ second[i]-'0'+flag];
                        if (result[i+biggest_size-smallest_size]!='0')
                                flag=1;
                }
        for (int i=biggest_size - smallest_size-1;i>-1;i--)
                {
                result[i]='9';
                }
        }
        if (second_size > size){
                for (int i=second_size-1;i>-1;i--)
                {
                        result[i]=m[ second[i]-'0'+flag];
                        if (result[i]!='0')
                                flag=1;
                }
        }
        result[result_size -1]='\0';
       
        return result;
}

int main(){
foo a("10");
foo b("950");

char *result;
result=a.complement("10","950");
cout<<result<<endl;
delete [] result;
result=a.complement(b);
cout<<result<<endl;
delete [] result;
}

---

im using g++ and running debian.
no errors or warnings :

$ g++ -o comp -Wall -Wextra -ansi -pedantic comp.cc
$

Yours truly

Is there a special reason why you have implemented the function twice ? You can pass both foo's strings to the function that takes two strings instead of debugging two functions :)

yes i prefer to work with objects (im studing cpp now).
the function with two string works gr8 but it was only written to show that almost the same code gives two diffrent answares.
i prefer to take the code written allmost in c (two strings) and to transfer it to object oriented.

for now i do this:

Code:

---

char* foo(foo object){
  char* result=complement(this ->number,object.number);
 return result;
}

---

but afaik it ain't OO programing

more is that i can't understand why there are two diffrent answeres

moment ..
could you please compile and run it on your system (maybe the fault isn't in program but with some other soft that runs on my pc )

Quote:

---

Originally Posted by jabka

moment ..
could you please compile and run it on your system (maybe the fault isn't in program but with some other soft that runs on my pc )

---

No it's not. Without even compiling your code, I can guarantee that you won't stumble on compiler bugs that easily.

Quote:

---

yes i prefer to work with objects (im studing cpp now).

---

If working with objects meant implementing code more than once, OOP would have been stillborn.


As I see it, the most likely error is the off-by-one: You treat the sizes in the char* and the object version the same, when they aren't: the sizes in the object version already include the \0.

> number=NULL;
> size=2;
At least make the values consistent with one another, say NULL and zero

> size=strlen(number)+1;
> number=new char[size];
> strcpy(number,value);
> number[size-1]='\0';
How can you do strlen() on it before you even initialise it?
You should be doing strlen on the value parameter.
Also, the assignment of a '\0' is a waste of time as the strcpy() will have already copied it.

Quote:

---

Originally Posted by Salem

> number=NULL;
> size=2;
At least make the values consistent with one another, say NULL and zero

> size=strlen(number)+1;

---

didn't see this .. (fixed it)
thnx

Quote:

---

Originally Posted by Salem

>
> number=new char[size];
> strcpy(number,value);
> number[size-1]='\0';

How can you do strlen() on it before you even initialise it?
You should be doing strlen on the value parameter.
Also, the assignment of a '\0' is a waste of time as the strcpy() will have already copied it.

---

thnx didn't thout that strcpy opies '\0' as weal.
but ..

still the same output :
050
049XXX

one of the strangest things is :
in the second excution (the Object version) the i get other answar 049 instead if 050

i have edited the code on the first paste to the correct form :
changed size=2 to size=2
removed '\0' at constructer

Well, it's called debugging, so now is the time to practice doing it.

Step through the code one line at a time, and verify (for example) that the array indices are as you expect. It's a pretty sure bet that you're stepping off the end of an array somewhere.

Also, consider writing a third function which both
char* foo::complement(char *first,char *second)
and
char* foo::complement(foo &object)
can call to do the bulk of the work, since the code is essentially the same. At least then you only have to debug it once rather than twice.