You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Mar 29, 2024. It is now read-only.
Before we can handle pure python interactors, we can enhance current C++ interactors with a python call mechanism.
Here is the simple modification I made with @anlambert MouseLassoNodesSelector. This allows me to root the call of the selection event through a python code, which itself can filter whatever it wants in return (especially in that case, the selected nodes have viewSelection set to True).
I believe we may generalize the case, and want some mechanism to pipe some python script after interaction, which allows us to refine the behavior of our interactors on the fly and really accelerate prototyping. Two classes of selectors will be particularly useful, group selectors-types (e.g. lasso) and node information type.
It may be done as a registration system, like with plugins, instead a python script would register to interactors and be executed after they have been applied. Or it can be the other way around, in the parameters of the interactor, we could specify an optional python plugin to run.
Here is a dummy example of the Default Python Selector, depending on the graph name, it routes the right plugin to apply, doubles the size of the selected nodes, then deselect them.
Before we can handle pure python interactors, we can enhance current C++ interactors with a python call mechanism.
Here is the simple modification I made with @anlambert MouseLassoNodesSelector. This allows me to root the call of the selection event through a python code, which itself can filter whatever it wants in return (especially in that case, the selected nodes have
viewSelection
set toTrue
).I believe we may generalize the case, and want some mechanism to pipe some python script after interaction, which allows us to refine the behavior of our interactors on the fly and really accelerate prototyping. Two classes of selectors will be particularly useful, group selectors-types (e.g. lasso) and node information type.
It may be done as a registration system, like with plugins, instead a python script would register to interactors and be executed after they have been applied. Or it can be the other way around, in the parameters of the interactor, we could specify an optional python plugin to run.
Here is a dummy example of the Default Python Selector, depending on the graph name, it routes the right plugin to apply, doubles the size of the selected nodes, then deselect them.
The text was updated successfully, but these errors were encountered: