Unity3D, a C# editor and few 3d meshes of your choice.
We want to see how you tackle a technical problem. We also want to see your skills to produce efficient and maintainable code over time. This exercise might seem simple and the expected behaviour could be achieved in a single function but we are also interested in how you might structure your code and classes considering this feature may be the beginning of a larger one destined to evolve in time.
The exercise consists in developing a camera focus on objects in the scene. Let's consider in that problem that we have a single camera in the scene using the perspective projection mode (the projection property on the camera should be set to perspective, see here : https://docs.unity3d.com/Manual/class-Camera.html) and you placed some items in the scene (you can use Unity pre-integrated 3D objects such as Sphere, Cube etc. for simplicity).
When you play the application, your script should position the camera correctly and smoothly (using Vector3.Lerp for instance) so that all objects are rendered entirely (no cropped objects in the camera rendering) while preserving its rotation, its vertical and horizontal field of view.
The only non visible objects should be the ones occluded by others.
The script should work no matter the initial camera orientation, the number of objects, their position and scale in the scene.
This is not mandatory, add this feature if you feel like it.
Make this works also for an orthographic projection camera (the projection property on the camera should be set to orthographic, see here : https://docs.unity3d.com/Manual/class-Camera.html).