Thread: const correctness gone wrong I suspect...

  1. #1
    Registered User
    Join Date
    Jun 2011
    Posts
    2

    const correctness gone wrong I suspect...

    Hi, I've been having a few headaches trying to get my code to compile. I feel like I've tried every way of changing "const"-ness, but still can't get this to compile with g++ in a cygwin environment.

    Would you be so kind to tell me where on earth I am going wrong? Below is the code and the compiler error. The error talks about "qualifiers" which I assume mean there's a "const" missing somewhere, but for the life of me, I can't figure out where.

    Code:
    #include <iostream>
    
    using namespace std;
    
    class Test
    {
       public:
         Test() : val(0) {}
         Test(int v) : val(v) {}
    
         int getVal() { return val; }
    
         bool operator<(const Test& rhs) {
           if (rhs.val < val) return false;
           else return true;
         }
    
       private:
         int val;
    };
    
    
    template <class T>
    inline const T& mymax(const T& a, const T& b)
    {
       if (a < b) return b;
       else return a;
    }
    
    int main(int argc, char* argv)
    {
       int a1 = 10, a2 = 20;
       const Test t1(10); 
       const Test t2(20);
    
       int result = mymax(a1,a2);
       const Test result2 = mymax(t1,t2);
    
       cout << "Largest value of ints is: " << result << endl;
       cout << "Largest value of Test(ints) is: " << result2.getVal() << endl;
       
    }
    The compiler generates this error (g++):

    Code:
    maxTamplate.cpp: In function `const T& mymax(const T&, const T&) [with T = Test]`:
    maxTemplate.cpp:37:    instantiated from here
    maxTemplate.cpp:26: error: passing `const Test' as `this' argument of `bool Test::operator<(const Test&)' discards qualifiers
    I've no idea what this message is trying to tell me

  2. #2
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Location
    Inside my computer
    Posts
    24,654
    Code:
         int getVal() const { return val; }
    
         bool operator<(const Test& rhs) const {
           if (rhs.val < val) return false;
           else return true;
         }
    The later will actually fix the problem. Remember that any getter--any function that does not change any class member variables--should be const.
    The error is essentially trying to say that you are calling a non-const function (operator <) on a const object.
    Last edited by Elysia; 06-26-2011 at 10:11 AM.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  3. #3
    Registered User
    Join Date
    Jun 2011
    Posts
    2
    Many thanks Elysia. That's cleared up something I've read and vaguely know about, but have never implemented before. I appreciate
    your help.

  4. #4
    C++ Witch laserlight's Avatar
    Join Date
    Oct 2003
    Location
    Singapore
    Posts
    28,413
    Incidentally, instead of writing:
    Code:
    if (rhs.val < val) return false;
    else return true;
    you can write:
    Code:
    return rhs.val >= val;
    Quote Originally Posted by Bjarne Stroustrup (2000-10-14)
    I get maybe two dozen requests for help with some sort of programming or design problem every day. Most have more sense than to send me hundreds of lines of code. If they do, I ask them to find the smallest example that exhibits the problem and send me that. Mostly, they then find the error themselves. "Finding the smallest program that demonstrates the error" is a powerful debugging tool.
    Look up a C++ Reference and learn How To Ask Questions The Smart Way

  5. #5
    Algorithm Dissector iMalc's Avatar
    Join Date
    Dec 2005
    Location
    New Zealand
    Posts
    6,318
    Quote Originally Posted by laserlight View Post
    Incidentally, instead of writing:
    Code:
    if (rhs.val < val) return false;
    else return true;
    you can write:
    Code:
    return rhs.val >= val;
    Which as it happens means that this classes less-than operator is incorrect, and thus cannot be used to store the item in a std::set for example. The problem is that for two equal items a and b, a<b and b<a would both return true. This is illogical and causes std::set to barf.
    I suggest changing it to this:
    Code:
           return val < rhs.val;
    My homepage
    Advice: Take only as directed - If symptoms persist, please see your debugger

    Linus Torvalds: "But it clearly is the only right way. The fact that everybody else does it some other way only means that they are wrong"

  6. #6
    Registered User
    Join Date
    Aug 2010
    Location
    Poland
    Posts
    733
    Quote Originally Posted by matwsij View Post
    Code:
    template <class T>
    inline const T& mymax(const T& a, const T& b)
    {
       if (a < b) return b;
       else return a;
    }
    a and/or b may be temporary objects and (an invalid) reference to an already destroyed object will be returned. Return by value.

  7. #7
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Location
    Inside my computer
    Posts
    24,654
    I believe it is quite safe. A small experiment seems to confirm this:
    Code:
    #include <iostream>
    
    class XTest
    {
    public:
    	XTest(int n): m_n(n) { std::cout << "Constructing " << n << "...\n"; }
    	~XTest() { std::cout << "Destructing " << m_n << "...\n"; m_n = 0xFFFFFFFF; }
    	bool operator < (const XTest& rhs) const { return m_n < rhs.m_n; }
    	int m_n;
    };
    
    template<typename T>
    const T& max(const T& lhs, const T& rhs)
    {
    	if (lhs < rhs)
    		return rhs;
    	else
    		return lhs;
    }
    
    int main()
    {
    	XTest t1(1), t2(2), t3(3);
    	std::cout << "Max of t1 and t2 is: " << max(t1, t2).m_n << "\n";
    	std::cout << "Max of t3 and t4 is: " << max(t3, XTest(4)).m_n << "\n";
    }
    (Yes, I know this is neither good practice nor portable; but then again, this is merely a quick and dirty test with the MSVC compiler; your mileage may vary.)
    Output:

    Constructing 1...
    Constructing 2...
    Constructing 3...
    Max of t1 and t2 is: 2
    Constructing 4...
    Max of t3 and t4 is: 4
    Destructing 4...
    Destructing 3...
    Destructing 2...
    Destructing 1...

    We clearly see that the lifetime of t4 is extended to after the statement.
    I cannot be certain, though, as I am unsure exactly what the standard mandates about this.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  8. #8
    Lurking whiteflags's Avatar
    Join Date
    Apr 2006
    Location
    United States
    Posts
    9,612
    The standard library uses the same signature for max as Elysia's code does and the only difference between Elysia's code and the code in question is use of inlining.

  9. #9
    Algorithm Dissector iMalc's Avatar
    Join Date
    Dec 2005
    Location
    New Zealand
    Posts
    6,318
    See section 12.2 point 5 of the C++0x standard final draft
    In short, it is fine.
    My homepage
    Advice: Take only as directed - If symptoms persist, please see your debugger

    Linus Torvalds: "But it clearly is the only right way. The fact that everybody else does it some other way only means that they are wrong"

  10. #10
    Registered User
    Join Date
    Aug 2010
    Location
    Poland
    Posts
    733
    o_O

    So it is correct, good to know.

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Program crashes I suspect variable n.
    By Witch in forum C Programming
    Replies: 7
    Last Post: 04-01-2010, 03:31 AM
  2. deep const correctness
    By robwhit in forum C Programming
    Replies: 20
    Last Post: 12-12-2008, 08:51 PM
  3. Suspect my CMOS battery is dying
    By Mario F. in forum Tech Board
    Replies: 13
    Last Post: 05-01-2008, 07:06 AM
  4. const correctness and static member variables
    By cunnus88 in forum C++ Programming
    Replies: 2
    Last Post: 02-21-2008, 05:26 PM