CELE-107 Selection sync across widgets

[dvcorreia]

Acceptance criteria:

• If I select a neuron, the selection should be reflected across widgets of the same workspace, for instance if I click in the graph ADA, the same neuron should light up/ be selected in the 3d viewer. For the EM viewer instead this neuron should be the only one visible for the segmentations, hide all the others.
• The selection can be performed from any widget, graph, 3d viewer or EM viewer.

[aranega]

All looks good in a general way :), I only ask for few modifications on some parts of the code as it would help to gain in clarity.

[dvcorreia]

Review changes done! Thanks @aranega 🙏

[aranega]

All looks good :) thanks @dvcorreia ! 🙏

[aranega]

I converted again the PR to draft because of a small bug that I forgot is not tackled (about selecting a neuron in the 2D graph viewer, then clearing the selection in EM, and not having the selection style removed from the 2D viewer).

I'm working on it.

[aranega]

@dvcorreia @ddelpiano I finally found the way to have something logical when a class neuron is selected in the 2D graph, and the various interactions with the 3D/EM viewers. It wasn't easy to find a first solution that would fit some scenarios.

[aranega]

@dvcorreia @ddelpiano I finally was able to fix the last issue with the synchronizers. Now all should be just fine 🙏

[ddelpiano]

@aranega @dvcorreia could one of you bring the latest development inside this branch please? There are changes that are obsolete now (e.g. the threeDsettings.ts changes are coming from the last PRs that Salam opened and she fixed a bunch of issues with those). Thanks!

[aranega]

@ddelpiano the branch is already up to date with develop, I modified the threeDsettings.ts file as there was warning in the console. Three is not considering string for colors but numbers. As the string wasn't recognize, it was interpreted by three/fiber as pure white (0xFFFFFF).

[dvcorreia]

Also confirmed that is up to date with develop

[dvcorreia]

@ddelpiano, I tested this and seems to work quite well. The issues we had with the neuron selected highlight in the 2D view seemed to have been solved, thanks @aranega :)