I wrote my generic class "MyVector" based on std::valarray
//myvector.h
#ifndef MYVECTOR_H
#define MYVECTOR_H
#include <valarray>
template <typename T>
class MyVector
{
public:
MyVector(int size){larr.resize(size);}
MyVector<T>& operator=(const MyVector<T>& other)
{
this->larr=other.larr;
return *this;
}
MyVector<T> operator+ ( const MyVector<T>& rhs); //body in .cpp file
template <typename U>
MyVector<T> operator+ (const U& val); //body in .cpp file
protected:
std::valarray larr;
}
#endif
//myvector.cpp
/*****other code*****/
template <typename T>
MyVector<T> MyVector<T>::operator+ ( const MyVector<T>& rhs)
{
MyVector<T> lv; lv.larr = this->larr + rhs.larr;
return lv;
}
template <typename T>
template <typename U>
MyVector<T> MyVector<T>::operator+ (const U& val)
{
MyVector<T> lv; lv.larr = this->larr + static_cast<T> (val);
return lv;
}
/*****other code*****/
Then I try to write a derived class DataVector, with all the functions of MyVector, especially, all my overloaded operators.
#ifndef DATAVECTOR_H
#define DATAVECTOR_H
#include "myvector.h"
class dataVector : public MyVector<int>
{
public:
dataVector& operator=(const dataVector& other)
{
this->larr=other.larr;
return *this;
}
using MyVector<int>::operator=;
using MyVector<int>::operator+;
}
#endif
When I try to compile main.cpp,
//main.cpp
#include "datavector.h"
#include "myvector.h"
dataVector datav1(10);
dataVector datav2(10);
dataVector datav3(10);
//if i write:
datav1=datav1 + 10; //works (gmake and compiler gcc7.5 on ubuntu)
//if i write:
datav3=datav1 + datav2; //does not work (gmake and compiler gcc7.5 on ubuntu)
I obtain this compiler error:
myvector.cpp: In instantiation of ‘MyVector<T> MyVector<T>::operator+(const U&) [with U = dataVector; T = int]’:
myvector.cpp:xxx:yy: error: invalid static_cast from type ‘const dataVector’ to type ‘int’
MyVector<T> lv; lv.larr = this->larr + static_cast<T> (val);
If I use MyVector: MyVector3=MyVector1+MyVector2 works well.
Can anyone help me? I know this code isn't well writed, but I'm still learning. thank you.
The problem is that while when you add MyVector
to MyVector
, MyVector::operator+ ( const MyVector<T>& rhs)
matches, but when you add dataVector
to dataVector
, dataVector::operator+ (const U& val)
(inherited from MyVector
) becomes a better match as it accepts anything , not necessarily something covertible to T
. There are several possible solutions.
dataVector
a typedef, like using dataVector = MyVector<int>
, if suitable.operator+
accepting dataVector
; it may call the inherited operator using an explicit cast (like *this + (const MyVector<int> &)rhs
).MyVector::operator+ (const U& val)
to accept T
only, as suggested by Jarod42.MyVector::operator+ (const U& val)
to accept sane types only using SFINAE (like template <typename U, typename = typename std::enable_if<std::is_arithmetic<U>::value>::type>
, or maybe is_convertible_to
, or whatever appropriate).
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.