What is the best method to compare two vectors of CString

Question

I am trying to find the most efficient, optimized and fastest way to compare to std vectors of CString. the strings in question are case-sensitive. I have tried using the == operator for the vector container but this sometimes return false positives. I mean for instance if one vector contains elements in the order (a,b,c) and the other has them in the order (b,c,a) the == operator will return false even thought they share the same data. Another thing is that it does not do case sensitive comparison.

I have thought of using a basic nested loops approach like this:

//Not Tested

BOOL bMatch = TRUE;
for(int i=0; i<Vec1.size();i++)
{
  if(!bMatch)
     break;
  int nComp=0;
  for(int j=0;j<Vec2.size();j++)
  {
     if(vec1[i].CompareNoCase(Vec2[j])==0)
        {
          //We have a match--check next item
          break;
        }
     else
        {
          nComp++;
          if(nComp == Vec2.size()-1)
             {
                 //Reached end of vector and no match found
                 //Vectors don't match
                 bMatch=FALSE;
             }
        }

  }
}

The above code is not tested and I am not sure if there is probably a better way to achieve such comparison without the need of using nested loops.

Would appreciate any advice or help...

Answer 1

if one vector contains elements in the order (a,b,c) and the other has them in the order (b,c,a) the == operator will return false even thought they share the same data.

Simply insert the data into two containers where the order does not matter and compare those:

std::vector<CString> vec1;
std::vector<CString> vec2;

// ...

std::multiset<CString> set1(vec1.begin(), vec1.end());
std::multiset<CString> set2(vec2.begin(), vec2.end());

bool equal_data = (set1 == set2);

If you want to ignore the case (which the code in your question seems to suggest), you can parameterize std::multiset and std::equal with an appropriate comparator:

struct compareNoCase
{
    bool operator()(const CString& a, const CString& b)
    {
        return a.CompareNoCase(b);
    }
};

std::vector<CString> vec1;
std::vector<CString> vec2;

// ...

std::multiset<CString> set1(vec1.begin(), vec1.end(), compareNoCase());
std::multiset<CString> set2(vec2.begin(), vec2.end(), compareNoCase());

bool equal_data = std::equal(set1.begin(), set1.end(),
                             set2.begin(),
                             compareNoCase());

The parameterization of std::multiset guarantees that "hello" and "HELLO" in the same vector are treated as one value, and the parameterization of std::equal guarantees this across the two vectors.

And finally, if you know that no element occurs twice in the same vector , you can use set instead of multiset . Note that it's probably better to work with a set or multiset right from the start.

Answer 2

if (a,b,c) and (b,c,a) are the same for you then vector is a bad choice, use std::set or std::multiset instead, and, as already said, compare them with std::equal and pass strcmp as the comparator argument. This answer is valid if by CString you mean C-style null-terminated char arrays. If CString means MFC CString, FredOverflow's answer is perfect.

Answer 3

首先使用std :: sort对它们进行排序，然后将它们与std :: equal进行比较。

Answer 4

Dont't use simple for loop. Instead you can use iterators to retrieve the elements from both vectors and then compare the values either using _tcscmp or wcscmp.