VR UI world to canvas without using a camera

Hi!
I've got some sources you can look at that do something very similar.
For converting form 3D raycast to UI canvas space, have a look at: https://github.com/Unity-Technologi...utSystem/Plugins/UI/TrackedDeviceRaycaster.cs

More specifically this function: RayIntersectsRectTransform
There, we get the 4 corners of the Canvas' RectTransform, and find where your ray intersects with that plane. Once we've got the intersection point, we just need to determine if that intersection point is within the 4 corners of the actual RectTransform.

Now, the second part of your question involves taking that intersected plane and passing it through the underlying EventSystem. This was a little bit tricky, but have a look at the TrackedPointerEventData in this file https://github.com/Unity-Technologi.../InputSystem/Plugins/UI/TrackedDeviceModel.cs
I created a new type of EventData, because I needed to pass the 3D data through to the Raycaster, because PointerEventData always assumed a screen-space 2D coordinate. This new type let's you pass it through the EventSystems' raycasting APIs. Important to note you need to set an invalid screenspace, so the other raycasters don't try to use the event. Then, once you've done your 3D raycasting, set the EventData's position property to the hit's screen space position using Camera.WorldToScreenSpace (see the TrackedDeviceRaycaster.cs file's PerformRaycast function).

This should cover some simple examples, but if you'd like a full solution, have a look at this folder: https://github.com/Unity-Technologi.../com.unity.inputsystem/InputSystem/Plugins/UI
It's got some other stuff in there, but provides a working 3D tracked device to UI EventSystem integration.

Hope this helps, ping back here for more info!

 

Aha, I think I understand now.
So you have, something like an Image UI element on a canvas that was clicked, and you want to know something like "The click at 20% X(left to right), and 70% Y(Top to bottom)" or something similar?