compose

The Compose canvas renderer: Kanvas, a composable drawing a scene graph onto a transparent Compose canvas — every shape in CGA model space, uniformly fit-scaled and clipped to the scene's bounds — plus the pure conversion boundary between model space and canvas px, the Compose bridge for the shared paint vocabulary, and a draw-call tracing entry validating draw output without pixel rasterization.

Packages

fromcrackpot.kintsugi.compose

common

The composable and the conversion boundary. Kanvas draws the scene; cgaToCanvas / canvasToCga convert between CGA model space (centered, Y-up) and canvas px (top-left-origin, Y-down) — the module's single coordinate boundary; fitScale is the one draw-scale formula the draw pass and pointer-input conversion share; and worldBorder derives the circle infinite lines are clipped against.

fromcrackpot.kintsugi.compose.style

common

The Compose bridge for the scene graph's paint vocabulary: ShapeColor to Compose Color and back — channel-exact towards Compose, sRGB-converted and 8-bit-rounded back.

fromcrackpot.kintsugi.compose.tracing

common

Draw-call tracing: canvasToTrace runs styled shapes through the production draw pass and returns each recorded Compose canvas call as one normalized text line — geometry and style assertions with no pixel rasterization.