Some constructors and operator= methods use iterators instead of const_iterators

[marcotama]

Class Vectors constructors expects a const Vectors&, but the iterator in the loop is declared Vectors::iterator, and my compiler complains. However, if I change it to Vectors::const_iterator, the complaint vanishes. These are lines 1171--1175 of the constructor at hand of ControlAlgorithm.h:

      Vectors(const Vectors<T>& that)
      {
        for (typename Vectors<T>::iterator iter = that.begin(); iter != that.end(); ++iter)
          vectors.push_back(*iter);
      }

I believe this also affects Vectors::operator=, though I am not sure because SWIG (which I am using to build a Python interface to this library) does not expose any operator= method to Python (because Python does not have any equivalent semantic). These are lines 1177--1186 defining such method:

      Vectors_<T>& operator=(const Vectors_<T>& that)
      {
        if (this != that)
        {
          vectors.clear();
          for (typename Vectors_<T>::iterator iter = that.begin(); iter != that.end(); ++iter)
            vectors.push_back(*iter);
        }
        return *this;
      }

Similarly, the same problem affects Ranges constructor and possible Ranges::operator=. Following are lines 328--343 of Mathema.h:

      Ranges_(const Ranges_<T>& that)
      {
        for (typename Ranges_<T>::const_iterator iter = that.begin(); iter != that.end(); ++iter)
          ranges.push_back(*iter);
      }

      Ranges_<T>& operator=(const Ranges_<T>& that)
      {
        if (this != that)
        {
          ranges.clear();
          for (typename Ranges_<T>::iterator iter = that.begin(); iter != that.end(); ++iter)
            ranges.push_back(*iter);
        }
        return *this;
      }

It is possible that other operator= functions are affected by this problem, but my compiler is not complaining because SWIG is not exposing any operator= for the same reason as above.

[marcotama]

My compiler is c++ (Debian 4.9.2-10) 4.9.2.