asked on

binary_search with LessThanComparable prototype: compilation failure!

Ah hello.

Please consider the following code, which some may see as familiar due to my earlier questions:

#include <algorithm>
#include <vector>
#include <string>
#include <iostream>

using namespace std;

class Human
{
public:
	Human(std::string strName, int nAge) : m_strName(strName), m_nAge(nAge) {}
	std::string GetName() const { return m_strName; }
	int GetAge() const { return m_nAge; }

	bool operator<(const Human& rhs) const { return m_strName < rhs.GetName(); }
	bool operator<(const std::string& rhs) const { return m_strName < rhs; }

protected:
	std::string m_strName;	
	int m_nAge;
};

int main(int argc, char * argv[])
{
	std::vector<Human> vec;
	vec.push_back(Human("Joe", 20));
	vec.push_back(Human("Mary", 21));
	vec.push_back(Human("Arnold", 44));
	vec.push_back(Human("Calterine", 21));
	vec.push_back(Human("Calterine", 45));

	sort(vec.begin(), vec.end());
	std::vector<Human>::iterator iter = vec.begin(), iter2 = vec.end();

	bool found = binary_search( vec.begin(), vec.end(), "Mary" );

	return 0;
}

Open in new window

I was previously using the version of binary_search which took a predicate as the fourth argument (see http://www.sgi.com/tech/stl/binary_search.html - the second prototype) but after reading more about LessThanComparable I thought I would investigate this.

On VS 2005, I get the compilation error

binary '<' : no operator found which takes a right-hand operand of type 'Human' (or there is no acceptable conversion)
could be 'built-in C++ operator<(const char [5], const char [5])'
while trying to match the argument list '(const char [5], Human)'

Open in new window

on the call to binary_search.

How can I use binary_search in this manner?

TIA

SOLUTION

Zoppo

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

mrwad99

ASKER

Hi ZOPPO, good to see you are still active on EE after all these years :)

Yes, I tried that (that is, I had the global operator and the member operator: two global operators of (const std::string& lhs, const Human& rhs) and (const Human& lhs, const std::string& rhs) is the same and removes the need for the member version) and it does indeed work. I figured it would, since defining the global operator< is essentially the version of binary_search() which accepts a StrictWeakOrdering object, but wanted to see if there was any way I could do away with this altogether and define everything as members of the Human class...I guess there isn't, since we, when it boils down, need an operator<(Human&) within std::string to make this work, I believe...

Please correct me if this is wrong...

As a second question related to this, the documentation for StrictWeakOrdering at http://www.sgi.com/tech/stl/StrictWeakOrdering.html is very clear on what it should return:

"A Strict Weak Ordering is a Binary Predicate that compares two objects, returning true if the first precedes the second"

However, the documentation for LessThanComparable (http://www.sgi.com/tech/stl/LessThanComparable.html) does not have anything of the sort from what I can see. So, how do I know if I am supposed to (in the example I have given) code

bool operator<(const Human& rhs) const { return m_strName < rhs.GetName(); }

Open in new window

bool operator<(const Human& rhs) const { return rhs.GetName() < m_strName ; }

Open in new window

Can anyone advise on this?

SOLUTION

Zoppo

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

ASKER CERTIFIED SOLUTION

sarabande

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

mrwad99

ASKER

Yes Sara I was, but thanks for participating anyway.

mrwad99

ASKER

Thanks ZOPPO!

mrwad99

ASKER

Umm I accepted the wrong answer there; many apologies ZOPPO - I can request CS to get it changed if you want...?

Zoppo

No problem at all :o)

Sara's comment even is a possible solution, so it's not wrong to accept it ...

Have a nice day,

best regards,

ZOPPO

sarabande

Sara's comment even is a possible solution, so it's not wrong to accept it ...

you are very kind, ZOPPO.

thanks to both of you.

Sara