I am a little confused about that.
Doesn’t Processing use OpenGL as the basis for its P2D and P3D renderers?
In the shader tutorial is explained for example which steps are taken to render a Pshape as shader.
The actual gsls file I see as a code running at real time.
How must I understand pre-calculated?
It may not be useful to think of this shader with respect to PShape though?! From the Processing side this is just drawing a full screen quad (rectangle), like all the ShaderToy shaders. As you say, everything is GLSL code running at real-time - quite literally in the sense that every pixel in every frame of the animation is calculated mathematically based on its x,y coordinates and the time in seconds. If you change how you update the uniform for time every frame, you change the animation rate / run it backwards / make it stationary.
I must admit the logic involved in the “reversal” of the calculation - from plotting a shape to working backwards from the pixel coordinates - scrambles my brain a bit!