Solved

# Efficient way of defining error codes

Posted on 2010-09-14
522 Views
I am currently working on a big existing software architecture in C++. It has error codes in the header file defined something like this:

static const int ERROR_ABC -1;
static const int ERROR_PQR -2;
static const int ERROR_XYZ -3;
.
.
.
.

Since -1, -2, etc values are clashing with some numerical results after the code was modified recently, I have been asked to come up with a more efficient method to handle error codes. Please give your suggestions (preferably with examples) to solve this problem. One idea I got was to change -1, -2 etc in the header file to something bigger like 100001, 100002, etc. Not sure if this is really a good way.
0
Question by:sukhoi35
• 3
• 2
• 2
• +2

LVL 53

Assisted Solution

Infinity08 earned 100 total points
It depends on the range of acceptable values. If you use the same variable to store the result or the error code, then the error codes have to be outside of the range of valid values (obviously). Which ones you choose is up to you - one is not more efficient than the other.

A few notes though :

(a) why not use enums for defining error codes ?
(b) in cases where the valid values clash with the defined error codes, why not keep the error code separate from the result ? In other words, let the function return an error code (0 for success), and pass the result back via a function parameter.
0

LVL 32

Assisted Solution

phoffric earned 300 total points
How about:    enum Err_Code { ERROR_ABC = 0xFFAA, ERROR_PQR = 0xFFBB, ERROR_XYZ = 0xFFCC } err;>> Since -1, -2, etc values are clashing with some numerical results Not sure I understand how you got error codes mixed up with other numerical results, but maybe this extreme example will help.
``````class ERROR_CODE {

public:

enum Err_Code {

ERROR_ABC = 0xFFAA,

ERROR_PQR = 0xFFBB,

ERROR_XYZ = 0xFFCC

} err;

};

int main() {

ERROR_CODE::Err_Code err_status;

err_status = ERROR_CODE::ERROR_ABC;

}
``````
0

LVL 84

Expert Comment

what results are they clashing with, and how are those results obtained?
0

LVL 12

Assisted Solution

w00te earned 100 total points
Make your errors into an enumeration, and start the enumeration values equal to whatever high number you want.  After explicitly definign the first enumeration value, the rest of the enumerations will progress from there by default.

enum ErrorCodes
{
ERROR_ABC = 100001,
ERROR_PQR,
ERROR_XYZ
};

When you want to print the codes, you can just std::cout<<ERROR_ABC; or the containing variable.

-w00te
0

LVL 32

Accepted Solution

phoffric earned 300 total points
Revision of code:1) Used static2) Added namespace exampleThe idea is to limit what assignment operations can be made, and hopefully avoid numeric clashes.
``````namespace ERR_CODE {

static enum Err_Code { ERROR_ABC = 0xFFAA, ERROR_PQR = 0xFFBB, ERROR_XYZ = 0xFFCC };

}

class ERROR_CODE {

public:

static enum Err_Code { ERROR_ABC = 0xFFAA, ERROR_PQR = 0xFFBB, ERROR_XYZ = 0xFFCC };

};

int main() {

ERROR_CODE::Err_Code err_status;

ERR_CODE::Err_Code   err_code_stat;

err_status = ERROR_CODE::ERROR_ABC;

err_code_stat = ERR_CODE::ERROR_ABC;

int val = err_code_stat;  // ok, but not desirable

err_code_stat = val;      // compiler error (good)

val = err_status;         // ok, but not desirable

err_status = val;         // compiler error (good)

}
``````
0

Author Comment

Hi Experts,
Thank you very much for all your responses. The main reason for me to do this little refactoring of error handling is that the current code is kind of old and some changes were done to fit a customer requirement.

static const int ERROR_ABC -1;
static const int ERROR_PQR -2;
static const int ERROR_XYZ -3;

void func()
{
nRet = callFunc();

if(nRet==ERROR_ABC)
return nRet;
else if(nRet==ERROR_PQR)
return nRet;
else
return nRet;
}

In the earlier code callFunc() returned a min value of 0 only. Because of some unplanned modifications that were done, now the minimum value that it can return is -1 which incidentally equals ERROR_ABC.

Will award points and close the question soon
0

LVL 53

Expert Comment

The example code you gave does not make much sense. You're always returning nRet, so you might as well have left out all checks :

int func() {
return callFunc();
}

which still seems unnecessary. But I assume a few things got lost when you posted that code here.

>> Because of some unplanned modifications that were done, now the minimum value that it can return is -1

0

Author Comment

Yes you are right, I have not really pasted the original code here (since I am not supposed to) but typed it myself in a hurry. so, many things are not there above.

I anyway got the solution, so am closing the question. Thanks Everyone  :)
0

Author Closing Comment

Thanks!
0

## Featured Post

Preface I don't like visual development tools that are supposed to write a program for me. Even if it is Xcode and I can use Interface Builder. Yes, it is a perfect tool and has helped me a lot, mainly, in the beginning, when my programs were smallâ€¦
Go is an acronym of golang, is a programming language developed Google in 2007. Go is a new language that is mostly in the C family, with significant input from Pascal/Modula/Oberon family. Hence Go arisen as low-level language with fast compilationâ€¦
The viewer will learn how to user default arguments when defining functions. This method of defining functions will be contrasted with the non-default-argument of defining functions.
The viewer will be introduced to the member functions push_back and pop_back of the vector class. The video will teach the difference between the two as well as how to use each one along with its functionality.