Godot regression test project

This repository contains Godot project which is tested in offical Godot repository(3.2 branch)

It aims to be complex project, which will allow to find crashes, leaks and invalid memory writes before PR is merged.

Sadly it can't find automatically any logic errors.

Reproduce bugs

If CI find bug, you can easily without much effort check which scene cause problems.

Each scene is independent of the other, so it is easy to disable some for testing.

To create minimal test scene:

• Look at the console output - there is printed info about current used scene
• If you run Start.tscn scene(which run other scenes in queue) then you can modify list of scenes which needs to be run in Autoload.gd by commenting records in alone_steps or all_in_one array(just like in picture)

How it works?

Autoload

There is only one scene which is loaded when project starts - "Autoload.gd". It handle exiting project after selected number of seconds

When opening any scene, automatically time to exit is set.

If running projet with e.g. this parameters

godot 20 -v

Then time is set to 20 seconds so it means that if scenes is 10 (EACH in alone_steps array + one for ALL scenes in all_in_one array), then each scene will be show for 2 seconds

Checking All scenes at once

There are two scenes which opens all scenes:

• All.tscn - opens all scenes at once
• Start.tscn - opens each scene one by one

AIO

This are scenes which only opens once, because there is no need to open it more times(no scripts or only with _ready function).

Other Scenes

Each other scenes checks specific types of nodes like lights, rendering, physics, text or reparenting.

Contributions

Contributions are welcome.

For now there is no requirements to format code.

New functionalities(e.g. physics checks) should be done in different folders(cleaner view to resources)

Epilepsy Warning

Due using by project a lot of functions from each type of Node, screen may flicker, images and objects may change randomly color and size which may lead some users to health problems.