February 25th contest results

I simply can't find time to judge this entry-- sorry. However, we do have one judge (Sang-Drax) that did. Seeing how there were only 5 contestants, I don't see this being a major problem.

Here it is:

Quote:

---

Throughout the entries, I missed the support for "const std::string" and the prevention of 'evil' or pointless code.

const string s = "12345";
cout << Mid(s,1,2);
Mid("12345",1,2) = "abc";

Here are the results:

===================
=Sebastian McKelvy=
===================

Fidelity 0/10
--------
Basic test incorrect : 234
First test incorrect : 123
Second test incorrect : 123
Third test incorrect : abc

I also had to comment away these lines in my test code:
//s2 = Mid("0123456789",1,4);
//Mid(s1, 1, 1) = Mid(s4,1);

Efficiency 0/10
----------

The code is very short, but doesn't do much right

Elegance 1/5
--------

No commenting
The solution is not very elegant.


Portability 5/5
-----------

No problems


-== TOTAL: 6 ==-











==================
=Travis Dane=
==================


Fidelity 1/10
--------

Didn't provide a function, which was required.
Didn't understand what Mid()=Mid() should do.

Basic functionality doesn't work, the program outputs "hel " instead
of "hel"

He has an operator= though.


Efficiency 2/10
----------

Short and simple code, but lacking functionality.


Elegance 2/5
--------

Nice commenting
Not very good solution with the (non-static) function inside the class

Portability 4.5/5
-----------

He used conio.h, but only in the main() function.



-== TOTAL: 9.5 ==-















==============
=Dorian Boyce=
==============


Fidelity 7/10
--------

Doesn't work with C-strings:
s2 = Mid("0123456789",1,4);

Unable to handle constant strings


Efficiency 8/10
----------

Could've used member functions instead of for-loops in the = operator No member initialization list Manual copying of the string in operator= instead of non-reference argument


Elegance 2/5
--------

Constructor is public
No private copy-constructor (to prevent use outside the Mid() function) No commenting

Good solution

Portability 5/5
-----------

No problems

-== TOTAL: 23 ==-












===============
= CodePlug =
===============


Fidelity 8/10
--------

Compiled and succeeded all tests (on MS VC++.NET only -- see
portability) Great!

No support for const std::string though
Evil code did compile.

Efficiency 10/10
----------

Very short code!


Elegance 3/5
--------

Well commented!
C-style pointer casts
Global variables
Constructor is public (btw, why didn't he call the class Mid to get rid of the redundant functions?)


Portability 3/5
-----------

Code is not standard -- problem with reference class members when copying Doesn't compile with CodeWarrior Doesn't compile with Comeau C++ -- The absolutely best compiler when it comes to standard compliance. http://www.comeaucomputing.com/tryitout/


-== TOTAL: 24 ==-












======
=DavT=
======

Fidelity 8/10
--------

Compiled and succeeded all tests (on MS VC++.NET only -- see
portability) Great!

No support for const std::string though
Evil code did compile.

Efficiency 9/10
----------

Short code!

Elegance 4/5
--------

Well commented!
C-style pointer casts (unnecessary, too)


Portability 3/5
-----------

Code is not standard -- problem with reference class members when copying Doesn't compile with CodeWarrior Doesn't compile with Comeau C++ -- The absolutely best compiler when it comes to standard compliance. http://www.comeaucomputing.com/tryitout/


-== TOTAL: 24 ==-

---

It's a tie! DavT and CodePlug are the winners. I'll attach the entries, as well as mine and Sang-Drax's attempt at the contest.

Note: There will be a speed coding contest tommorow. Details to come.

Hmmm. Seeing as this was a C++ only contest, i dont think it was very fair to test C-style strings. For the other stuff, I was just being lazy :/

Anyways, congrats to the winners... i guess. (damnit i never win ):(

wheres the test code?

I believe this is what Sang-Drax was using:

Quote:

---

Originally posted by Sang-drax
You can try your functions using this test program.
It tests various syntactic constructions and the resulting strings.

Code:

---

int main()
{
    using namespace std;
     
    bool correct = true;
    string s1,s2; 
 
    //Basic test
    s2 = "0123456789";
    string s3 = Mid(s2,1,4);
    s1 = Mid(s2,1,4);
    s2 = Mid("0123456789",1,4);
    if (s1 != "1234" or s2 != "1234" or s3 != "1234")
        cout << "Basic test incorrect : " << s1, correct=false;
   
   
    //First test
    s1 = "abcdefg";
    Mid(s1, 3, 0) = "123";
    if (s1 != "abc123defg")
        cout << "First test incorrect : " << s1, correct=false;
       
       
    //Second test
    s1 = "abcdefg";
    s2 = "123";
    Mid(s1, 3, s2.size()) = s2;
    if (s1 != "abc123g")
        cout << "Second test incorrect : " << s1, correct=false;

    //Third test
    s1 = "abc";
    s2 = "123";
    Mid(s1, 1, 1) = Mid(s2,1);
    if (s1 != "a23c")
        cout << "Third test incorrect : " << s1, correct=false;   
   
    if (correct)
        cout << "All tests correct!";
}

---

---

That version of my code is an early version. I modified it later to prevent the code discussed in the other thread, which I still believe shouldn't compile.

Good work DavT and CodePlug! I liked the inheritance-solution.

Quote:

---

Originally posted by *ClownPimp*
Hmmm. Seeing as this was a C++ only contest, i dont think it was very fair to test C-style strings.

---

IMO, it was.
String constants in source code are still const char*, not std::string.
String functions must consider that string constants may be passed as arguments.