non-const `Color::operator[]` should return mutable reference

[azeey]

Environment

• OS Version: Ubuntu 22.04
• Source or binary build? main, e0193a5

Description

• Expected behavior: To be able to mutate a component of Color using operator[]
• Actual behavior: Since it returns a value, any mutation is not applied to the component in Color
  

  gz-math/include/gz/math/Color.hh

  Line 162 in e0193a5

  public: float operator[](const unsigned int _index);

[Avisheet]

The issue you're describing is related to the operator[] overload for the Color class. The problem is that the operator is declared to return a float instead of a reference to a float. As a result, when you use the operator to access a component of the color and attempt to mutate it, the mutation doesn't apply to the actual color instance.

To fix this issue, you should modify the declaration of the operator[] to return a reference to a float. Here's the corrected declaration:

public: float& operator[](const unsigned int _index);

This modification allows you to use the operator to both access and mutate the individual components of the Color class.

I hope by these changes in the code you will be able to mutate a component of color using operator

[aagrawal05]

Hi, is this issue still open? If it hasn't been implemented I would be happy to open the pull request.

[Avisheet]

Hi @aagrawal05 , I am currently working on this issue. I'm getting some run time errors while running test cases. You can check those details here.
It would be appreciated if you can add some value to that .

[nunoguedelha]

Hi @Avisheet @azeey , is anyone currently working on this issue? I would like to contribute to gz-math repo, have free time and started looking into this good first issue. Please let me know if you still need help on this.

[azeey]

Hi @nunoguedelha, thanks for volunteering to take on this. #579 is relevant, but it was closed due to inactivity. I will assign this to you now.

[nunoguedelha]

Hi @azeey , sorry for the late reply, I'll be pushing the code shortly today or tomorrow. I'll base my branch off of main as this breaks API.
I remember that an operator overload was mentioned in the previous PR, but if I'm not mistaken, function overloading with different return types but same parameter list is not possible in C++, so the solution here would be to replace the current operator[] function definition with the float& return type. Do you agree?