Return to Answer

class Add
{
public:
  Add(int x, int y) : m_x(x), m_y(y) {}
  operator int() { return m_x + m_y; }
  void x(int x) { m_x = x; }
  void y(int y) { m_y = y; }
  int print(Print& outp) {
    int res;
    res = outp.print(F("Add(x="));
    res += outp.print(m_x);
    res += outp.print(F(",y="));
    res += outp.print(m_y);
    res += outp.print(F(")"));
    return res;
  }
private:
  int m_x, m_y;
};

Add func(1,2);

void setup() {
  Serial.begin(57600);

  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);

  func.x(12);
  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);

  func.y(13);
  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);
}

void loop() {
}

class Add
{
public:
  Add(int x, int y) : m_x(x), m_y(y) {}
  operator int() { return m_x + m_y; }
  void x(int x) { m_x = x; }
  void y(int y) { m_y = y; }
  int print(Print& outp) {
    int res;
    res = outp.print(F("Add(x="));
    res += outp.print(m_x);
    res += outp.print(F(",y="));
    res += outp.print(m_y);
    res += outp.print(F(")"));
    return res;
  }
private:
  int m_x, m_y;
};

Add func(1,2);

void setup() {
  Serial.begin(9600);

  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);

  func.x(12);
  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);

  func.y(13);
  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);
}

void loop() {
}

(Update)

Below is a rewrite of @VE7JRO Add function class using cast operator to perform the evaluation:

class Add
{
public:
  Add(int x, int y) : m_x(x), m_y(y) {}
  operator int() { return m_x + m_y; }
  void x(int x) { m_x = x; }
  void y(int y) { m_y = y; }
  int print(Print& outp) {
    int res;
    res = outp.print(F("Add(x="));
    res += outp.print(m_x);
    res += outp.print(F(",y="));
    res += outp.print(m_y);
    res += outp.print(F(")"));
    return res;
  }
private:
  int m_x, m_y;
};

Add func(1,2);

void setup() {
  Serial.begin(57600);

  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);

  func.x(12);
  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);

  func.y(13);
  func.print(Serial);
  Serial.print(F("="));
  Serial.println(func);
}

void loop() {
}

This is (unfortunately) not an Arduino question. It is actually about the syntax of C++ which is a rather complex language.

The mechanism you have discovered is called default arguments. The function prototype or definition may contain the default values of the arguments. As this is a list the default values are restricted from right to left. It is not possible to use a default parameter for argument to the left of a given argument. The order is strictly right to left.

f(int x = 0, int y = 1, int z = 2)

f(1,2,3) binds x = 1, y = 2, z = 3
f(1,2) equals f(1,2,2) binds x = 1, y = 2, z = 2
f(1) equals f(1,1,2) binds x = 1, y = 1, z = 2
f() equals f(0,1,2) binds x = 0, y = 1, z = 2

There is a syntax for struct initialization (aka aggregate initialization) in C++ that allows naming of struct field values. This type of naming is not allowed for function calls.

Please read https://en.cppreference.com/w/cpp/language/default_arguments and https://en.cppreference.com/w/cpp/language/aggregate_initializationfor more details. And a C++ book.

There are languages that allow default arguments and named parameters e.g. C#, Python, Ruby and Smalltalk. Please see https://en.wikipedia.org/wiki/Named_parameter.

Cheers!

This is (unfortunately) not an Arduino question. It is actually about the syntax of C++ which is a rather complex language.

The mechanism you have discovered is called default arguments. The function prototype or definition may contain the default values of the arguments. As this is a list the default values are restricted from right to left. It is not possible to use a default parameter for argument to the left of a given argument. The order is strictly right to left.

f(int x = 0, int y = 1, int z = 2)

f(1,2,3) binds x = 1, y = 2, z = 3
f(1,2) equals f(1,2,2) binds x = 1, y = 2, z = 2
f(1) equals f(1,1,2) binds x = 1, y = 1, z = 2
f() equals f(0,1,2) binds x = 0, y = 1, z = 2

There is a syntax for struct initialization (aka aggregate initialization) in C++ that allows naming of struct field values. This type of naming is not allowed for function calls.

Please read https://en.cppreference.com/w/cpp/language/default_arguments, https://en.cppreference.com/w/cpp/language/aggregate_initialization and https://en.cppreference.com/w/cpp/utility/functional/bind for more details. And a good C++ book.

There are languages that allow default arguments and named parameters e.g. C#, Python, Ruby and Smalltalk. Please see https://en.wikipedia.org/wiki/Named_parameter.

Cheers!