Effects: effect points on shape layers start at the layer centre (#227)

Shape and text layers have comp-sized effect bounds centred on their
origin, but effects measured positions from the origin: a point at its
default "layer centre" sat at the comp's bottom-right corner, so Twirl
on a centred shape changed nothing and Bulge only its bottom-right
quarter. As in After Effects, effects now work in effect space,
measured from the top-left of the layer's effect bounds.

The renderer moves the layer buffer into effect space for the whole
stack (Buf::rebase, CPU and GPU alike) and hands effects everything
else in it: masks, other layers' pixels for layer parameters (each in
its own effect space, so Set Matte and displacement maps see a whole
shape layer), the layer's own frames (layer_input, also what Puppet,
Roto Brush and the Layer panel read) and the comp camera
(CompScene::rebased). Layer bounds are now right for every effect, so
pin_rect and EffectEnv::bounds_origin go and layer_rect serves edge
pinning too. Adjustment layers still run on the comp below as it is.

effect_bounds / EvalCtx::effect_to_comp replace the repeated "comp size
when there is no source" code; the viewer's effect point handles and
eyedroppers, Puppet pins, the Layer panel (paint, Roto Brush) and the
tracker's Corner Pin output use them.

Project schema 2: opening a schema 1 project moves positions held by
effects on shape and text layers by half the comp size, so set points
keep their place (point parameters with keyframes, Paint paths and
clone positions, Puppet seeds and pins, Liquify and Roto Brush
strokes). Points still at their default become the layer centre;
expression controls, adjustment layers and expressions are left as
they are (docs/architecture.md). The disk cache's render version is
bumped.

Tests: default_effect_points_sit_at_the_centre_of_a_shape_layer and
layer_parameters_see_a_whole_shape_layer (render, both fail before),
effect_points_on_a_shape_layer (GPU vs CPU),
opening_a_schema_1_project_moves_shape_layer_effect_points_into_effect_space
(engine), and the effects migration tests.

Part of #227

Co-Authored-By: Claude Opus 5.5 <[email protected]>

docs/architecture.md

@@ -82,13 +82,46 @@ a file that a newer EffectCraft saved shows a warning: fields this version doesn
 dropped if it saves the project. Save, Save As, Save a Copy and Increment and Save keep the
 file's format (`.ecprojx` stays XML).
 
+`schema` changes when a saved value changes meaning; a project from an older schema is converted
+when it is opened (`effectcraft_engine::upgrade_effects`, which also refreshes effect instances
+from the registry) and saved with the current one, which older versions refuse to open rather than
+render wrongly. Schema 2 put effect positions on shape and text layers in effect space (§4,
+#227). Opening a schema 1 project moves what an effect on a shape or text layer (not an adjustment
+layer, whose effects see the comp below as before) holds as a position by half the comp size,
+so points that were set keep their place in the comp: point parameters with their keyframes, Paint
+stroke paths and clone positions, Puppet mesh seeds, pin rest points and positions, and Liquify
+and Roto Brush strokes. A point parameter still at its default and not animated is left alone:
+the old default sat at the comp's bottom-right corner, where most effects did nothing, and the
+same value is now the layer centre, as in After Effects. Expression controls (Point Control…)
+keep their values, because their meaning is up to the expressions that read them, and expressions
+are not rewritten, so an expression that computes an effect point on a shape or text layer from
+other values may need `+ [thisComp.width / 2, thisComp.height / 2]`. Effects that use the layer
+bounds (edge pinning, wipes, Offset, generators) now cover the whole comp-sized bounds instead of
+their bottom-right quarter. Tracking data of effects that analyse footage (Warp Stabilizer, 3D
+Camera Tracker, Mocha shape) is left as it is: footage has a source rectangle, whose effect space
+did not change.
+
 ## 4. Rendering
 
 `Renderer::comp_frame(comp, t)` walks the layers bottom to top. For each visible layer it renders
 the **source** (solid, footage frame, text, shape contents, or a nested comp), applies **masks**,
 then **effects** in order, then **layer styles**, then the **transform** into comp space (with motion blur sub-samples
 when enabled), the **track matte**, and finally **blends** into the accumulator. Adjustment layers
-apply their effects to the accumulator. Runs of 3D layers are composited per pixel through the
+apply their effects to the accumulator.
+
+Effects run in **effect space** (`effectcraft_render::effect_bounds`): layer pixels measured from
+the top-left of the layer's effect bounds, which are the source rectangle, or for layers without
+one (shape, text) a comp-sized rectangle centred on the layer's origin that their content
+surrounds. A shape layer's default effect point (the bounds' centre) is therefore its content's
+origin, as in After Effects. The renderer moves the layer buffer into effect space for the whole
+stack (`Buf::rebase`, on the CPU and the GPU alike) and hands effects everything else in it too:
+masks, other layers' pixels for layer parameters (each in its own effect space), the layer's own
+frames at other times (`Renderer::layer_input`, also what Puppet, Roto Brush and the Layer panel
+read) and the comp camera. Effect point handles, Puppet pins and the Layer panel map effect space
+to the comp with `EvalCtx::effect_to_comp`. Adjustment layers run their effects on the comp below
+as it is.
+
+Runs of 3D layers are composited per pixel through the
 active camera, with lights, shadows and depth of field (Classic 3D), or rasterised as meshes by
 the Advanced 3D renderer (below).