In my project, there are a few lines in a MyClass.h
file
class MyClass{
public:
....
#ifndef __MAKECINT__
std::vector<Status_e(MyClass::*)(TupleInfo & info)> m_processObjects; //!
#endif
....
}
//The Status_e is an enum type
//The TupleInfo is a struct
I can't understand what the usage of this *
in above code snippet. And in the MyClass.cxx
file, the m_processObjects
used as this:
for (unsigned int f = 0; f < m_processObjects.size(); ++f) {
status = (this->*m_processObjects[f])( m_tuples[i] );
if ( status == FAILURE ) {
return EL::StatusCode::FAILURE;
}
....
}
....
I have never heard about the usage like this->*blabla
, but this code snippet works. So what it means?
A pointer variable points to a data type (like int ) of the same type, and is created with the * operator.
The * in expression means "the value stored in the address" so if c is a pointer to some char, then *c is the specific char. char *(*s); meaning that s is a pointer to a pointer to char, so s doesn't hold the address of a char, but the address of variable that hold the address of a char.
In computer programming, the dereference operator or indirection operator, sometimes denoted by " * " (i.e. an asterisk), is a unary operator (i.e. one with a single operand) found in C-like languages that include pointer variables.
* is the indirection operator in C and C++. Whenever it is used, it indicates that the variable next to it is a pointer containing the address of another variable. Indirection operator is also the "value stored at address" operator.
::*
denotes a Pointer to member.
With the surrounding code it's actually a Pointer to member function.
Status_e(MyClass::*)(TupleInfo & info)
is a member function of class MyClass
, returning Status_e
, and having one parameter TupleInfo&
. (The argument name info
is useless here but obviously silently ignored by the compiler.)
The other snippet in OP's question shows how to call it:
status = (this->*m_processObjects[f])(m_tuples[i]);
Storing a method pointer would look like this:
std::vector<Status_e(MyClass::*)(TupleInfo & info)> m_processObjects;
...
m_processObjects.push_back(&MyClass::aMethod);
Of course, the signature of MyClass::aMethod
must match.
A simplified sample to demonstrate it:
#include <iostream>
#include <vector>
class Test {
private:
std::vector<int(Test::*)(const char*)> _tblTestFuncs;
public:
Test()
{
_tblTestFuncs.push_back(&Test::func1);
_tblTestFuncs.push_back(&Test::func2);
_tblTestFuncs.push_back(&Test::func3);
}
int func1(const char *caller) { std::cout << "Test::func1 called from '"<< caller << "': "; return 1; }
int func2(const char *caller) { std::cout << "Test::func2 called from '"<< caller << "': "; return 2; }
int func3(const char *caller) { std::cout << "Test::func3 called from '"<< caller << "': "; return 3; }
void call()
{
for (size_t i = 0, n = _tblTestFuncs.size(); i < n; ++i) {
int result = (this->*_tblTestFuncs[i])("Test::call()");
std::cout << "Return: " << result << '\n';
}
}
};
int main()
{
Test test;
// test method vector in main()
std::vector<int(Test::*)(const char*)> tblTestFuncs;
tblTestFuncs.push_back(&Test::func1);
tblTestFuncs.push_back(&Test::func2);
tblTestFuncs.push_back(&Test::func3);
for (size_t i = 0, n = tblTestFuncs.size(); i < n; ++i) {
int result = (test.*tblTestFuncs[i])("main()");
std::cout << "Return: " << result << '\n';
}
// test method vector in Test
test.call();
// done
return 0;
}
Output:
Test::func1 called from 'main()': Return: 1
Test::func2 called from 'main()': Return: 2
Test::func3 called from 'main()': Return: 3
Test::func1 called from 'Test::call()': Return: 1
Test::func2 called from 'Test::call()': Return: 2
Test::func3 called from 'Test::call()': Return: 3
Live Demo on coliru.com
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With