openpencil/tests/engine/render/canvas/effects/ordering.test.ts

98 lines
2.8 KiB
TypeScript
Raw Normal View History

import { describe, expect, mock, test } from 'bun:test'
import type { SceneGraph } from '@open-pencil/scene-graph'
import { createDefaultNode } from '@open-pencil/scene-graph/node-defaults'
import { renderShapeUncached } from '#core/canvas/scene'
test: typecheck the test suites and fix what that found (#896) * build: typecheck the test suites Tests were in no TypeScript program: no tsconfig included tests/** or packages/*/tests/**, and bun strips types without checking them, so a fixture could drop a required field and keep passing until something read it. @types/bun moves to the root because it was installed per package only, and #cli-tests/* joins the paths the root config already carries. * test: fix the type errors the test suites were hiding Typechecking the tests turned up 1123 errors. Most were ordinary strictness, but some were real: `NodeChange` bound to Figma's plugin typings rather than the Kiwi codec in thirteen .fig tests, materializeInstance was called with six arguments against five so the blobs and source children were dropped, CanvasKit pixels were written to a plain object that never reached WASM, and assertions were made through accessors that do not exist, so they asserted nothing. Fixtures that had quietly lost a required field now carry it, nullable results are narrowed through the existing expectDefined helper rather than assumed, and stand-ins for CanvasKit and the editor go through one named helper instead of an unexplained cast at each site. No test was deleted, skipped, or weakened, and no `any`, non-null assertion, or ts-expect-error was introduced. * docs: record what typechecking the tests established Pins the app program's global types with an assertion rather than a note, since an unpinned types list lets any root @types package decide which platform src/** is judged against. The two environment faults that look like code regressions — Vite's dependency pre-bundle outliving a package rebuild, and heavy .fig suites failing under load — go to the development docs, where an explanation belongs. * fix: align @types/bun and keep node types resolvable when extended The root manifest declared a newer @types/bun than every package, which check:monorepo rejects, and pinning the app program's types left them unresolvable from a config that extends this one out of tree. * fix: fail the test typecheck when the compiler itself fails The gate matched diagnostics by substring, so a compiler or config failure that named no test file printed a pass while having checked nothing. Diagnostics are now split by whether they name a file: an unscoped one is the run failing and stops the gate, a test file's is a finding, and a source file's stays out by design. Also drops the parameter planComponentConstruction never read, and makes the inner-shadow verification script exit non-zero when it renders no image instead of logging and succeeding. * chore: merge master into tests-typecheck
2026-10-05 12:42:38 +00:00
import { asCanvas } from '../helpers'
import { createMockCanvas, createMockRenderer } from './helpers'
describe('Renderer effect ordering (Behavioral)', () => {
test('drop shadow renders before fills', () => {
const r = createMockRenderer()
const canvas = createMockCanvas()
const node = createDefaultNode(() => 'node1', 'RECTANGLE', {
width: 100,
height: 100,
effects: [
{
type: 'DROP_SHADOW',
visible: true,
color: { r: 0, g: 0, b: 0, a: 0.5 },
offset: { x: 0, y: 0 },
radius: 10,
spread: 0
}
],
fills: [{ visible: true, type: 'SOLID', color: { r: 1, g: 0, b: 0, a: 1 }, opacity: 1 }],
strokes: []
})
const graph: Partial<SceneGraph> = {
getNode: mock(() => node)
}
const callOrder: string[] = []
r.renderEffects = mock((_c, _n, _r, _h, pass) => {
callOrder.push(`renderEffects:${pass}`)
})
r.drawNodeFill = mock(() => {
callOrder.push('drawNodeFill')
})
test: typecheck the test suites and fix what that found (#896) * build: typecheck the test suites Tests were in no TypeScript program: no tsconfig included tests/** or packages/*/tests/**, and bun strips types without checking them, so a fixture could drop a required field and keep passing until something read it. @types/bun moves to the root because it was installed per package only, and #cli-tests/* joins the paths the root config already carries. * test: fix the type errors the test suites were hiding Typechecking the tests turned up 1123 errors. Most were ordinary strictness, but some were real: `NodeChange` bound to Figma's plugin typings rather than the Kiwi codec in thirteen .fig tests, materializeInstance was called with six arguments against five so the blobs and source children were dropped, CanvasKit pixels were written to a plain object that never reached WASM, and assertions were made through accessors that do not exist, so they asserted nothing. Fixtures that had quietly lost a required field now carry it, nullable results are narrowed through the existing expectDefined helper rather than assumed, and stand-ins for CanvasKit and the editor go through one named helper instead of an unexplained cast at each site. No test was deleted, skipped, or weakened, and no `any`, non-null assertion, or ts-expect-error was introduced. * docs: record what typechecking the tests established Pins the app program's global types with an assertion rather than a note, since an unpinned types list lets any root @types package decide which platform src/** is judged against. The two environment faults that look like code regressions — Vite's dependency pre-bundle outliving a package rebuild, and heavy .fig suites failing under load — go to the development docs, where an explanation belongs. * fix: align @types/bun and keep node types resolvable when extended The root manifest declared a newer @types/bun than every package, which check:monorepo rejects, and pinning the app program's types left them unresolvable from a config that extends this one out of tree. * fix: fail the test typecheck when the compiler itself fails The gate matched diagnostics by substring, so a compiler or config failure that named no test file printed a pass while having checked nothing. Diagnostics are now split by whether they name a file: an unscoped one is the run failing and stops the gate, a test file's is a finding, and a source file's stays out by design. Also drops the parameter planComponentConstruction never read, and makes the inner-shadow verification script exit non-zero when it renders no image instead of logging and succeeding. * chore: merge master into tests-typecheck
2026-10-05 12:42:38 +00:00
renderShapeUncached(r, asCanvas(canvas), node, graph as SceneGraph)
expect(callOrder).toEqual(['renderEffects:behind', 'drawNodeFill', 'renderEffects:front'])
})
test('inner shadow and blur render after strokes', () => {
const r = createMockRenderer()
const canvas = createMockCanvas()
const node = createDefaultNode(() => 'node1', 'RECTANGLE', {
width: 100,
height: 100,
fills: [],
childIds: [],
effects: [
{
type: 'INNER_SHADOW',
visible: true,
color: { r: 0, g: 0, b: 0, a: 0.5 },
offset: { x: 0, y: 0 },
radius: 10,
spread: 0
}
],
fix(core): draw gradient and image strokes as the paint they are (#868) * fix(core): draw gradient and image strokes as the paint they are A stroke carried the paint vocabulary already, but nothing read it: the .fig reader sent every stroke paint through resolvedPaintColor, which returns black for a gradient or image, the renderer set a flat color on strokePaint, and the writer emitted a SOLID paint. Strokes now go through the same conversion fills do in both directions, and applyGradientFill and applyImageFill take the target Paint so a stroke reuses the fill shader path instead of growing a second one. forVisibleStrokes is the single place every stroke draw passes through, so the shader is set and cleared there rather than threaded through each draw helper. Closes #797 for rendering and .fig; authoring a gradient stroke from the stroke panel is still to come. * feat(app): author gradient and image strokes from the stroke panel StrokeSection opened a solid-only colour picker and synthesised a fake fill for the swatch, so a stroke could never be anything but one flat colour. It now opens FillPicker like the fill panel does, and applyStrokePaint keeps the stroke's weight, align, cap, join and dashes across a paint change. Completes #797. * fix(core): let a gradient stroke reach vector outlines and arrowheads A vector stroke draws its outline as a filled shape with fillPaint, a dashed one strokes the path, and arrowheads are filled shapes of their own; each cleared the shader first, so a gradient or image stroke on a vector drew black. The stroke pass now configures both paints and owns clearing them, and those helpers keep what it set. Resolve each gradient stop against the stroke's own colour binding rather than the stop's position, which looked up another stroke's. Reported in review of #868. * fix(core): release the shaders a paint no longer owns Every gradient and image shader was handed to a paint and then leaked: the paint takes its own reference, so the caller's handle has to go or WASM memory grows with each redraw. Only the diamond branch did this. A gradient stroke now configures two paints, which doubled the leak. Reported in review of #868. * test(render): model a shader handle the caller deletes The pattern shader double returned a plain string, so deleting the handle the paint no longer owns threw instead of passing.
2026-10-04 13:25:45 +00:00
strokes: [
test: typecheck the test suites and fix what that found (#896) * build: typecheck the test suites Tests were in no TypeScript program: no tsconfig included tests/** or packages/*/tests/**, and bun strips types without checking them, so a fixture could drop a required field and keep passing until something read it. @types/bun moves to the root because it was installed per package only, and #cli-tests/* joins the paths the root config already carries. * test: fix the type errors the test suites were hiding Typechecking the tests turned up 1123 errors. Most were ordinary strictness, but some were real: `NodeChange` bound to Figma's plugin typings rather than the Kiwi codec in thirteen .fig tests, materializeInstance was called with six arguments against five so the blobs and source children were dropped, CanvasKit pixels were written to a plain object that never reached WASM, and assertions were made through accessors that do not exist, so they asserted nothing. Fixtures that had quietly lost a required field now carry it, nullable results are narrowed through the existing expectDefined helper rather than assumed, and stand-ins for CanvasKit and the editor go through one named helper instead of an unexplained cast at each site. No test was deleted, skipped, or weakened, and no `any`, non-null assertion, or ts-expect-error was introduced. * docs: record what typechecking the tests established Pins the app program's global types with an assertion rather than a note, since an unpinned types list lets any root @types package decide which platform src/** is judged against. The two environment faults that look like code regressions — Vite's dependency pre-bundle outliving a package rebuild, and heavy .fig suites failing under load — go to the development docs, where an explanation belongs. * fix: align @types/bun and keep node types resolvable when extended The root manifest declared a newer @types/bun than every package, which check:monorepo rejects, and pinning the app program's types left them unresolvable from a config that extends this one out of tree. * fix: fail the test typecheck when the compiler itself fails The gate matched diagnostics by substring, so a compiler or config failure that named no test file printed a pass while having checked nothing. Diagnostics are now split by whether they name a file: an unscoped one is the run failing and stops the gate, a test file's is a finding, and a source file's stays out by design. Also drops the parameter planComponentConstruction never read, and makes the inner-shadow verification script exit non-zero when it renders no image instead of logging and succeeding. * chore: merge master into tests-typecheck
2026-10-05 12:42:38 +00:00
{
type: 'SOLID',
visible: true,
weight: 1,
opacity: 1,
color: { r: 0, g: 0, b: 0, a: 1 },
align: 'CENTER'
}
fix(core): draw gradient and image strokes as the paint they are (#868) * fix(core): draw gradient and image strokes as the paint they are A stroke carried the paint vocabulary already, but nothing read it: the .fig reader sent every stroke paint through resolvedPaintColor, which returns black for a gradient or image, the renderer set a flat color on strokePaint, and the writer emitted a SOLID paint. Strokes now go through the same conversion fills do in both directions, and applyGradientFill and applyImageFill take the target Paint so a stroke reuses the fill shader path instead of growing a second one. forVisibleStrokes is the single place every stroke draw passes through, so the shader is set and cleared there rather than threaded through each draw helper. Closes #797 for rendering and .fig; authoring a gradient stroke from the stroke panel is still to come. * feat(app): author gradient and image strokes from the stroke panel StrokeSection opened a solid-only colour picker and synthesised a fake fill for the swatch, so a stroke could never be anything but one flat colour. It now opens FillPicker like the fill panel does, and applyStrokePaint keeps the stroke's weight, align, cap, join and dashes across a paint change. Completes #797. * fix(core): let a gradient stroke reach vector outlines and arrowheads A vector stroke draws its outline as a filled shape with fillPaint, a dashed one strokes the path, and arrowheads are filled shapes of their own; each cleared the shader first, so a gradient or image stroke on a vector drew black. The stroke pass now configures both paints and owns clearing them, and those helpers keep what it set. Resolve each gradient stop against the stroke's own colour binding rather than the stop's position, which looked up another stroke's. Reported in review of #868. * fix(core): release the shaders a paint no longer owns Every gradient and image shader was handed to a paint and then leaked: the paint takes its own reference, so the caller's handle has to go or WASM memory grows with each redraw. Only the diamond branch did this. A gradient stroke now configures two paints, which doubled the leak. Reported in review of #868. * test(render): model a shader handle the caller deletes The pattern shader double returned a plain string, so deleting the handle the paint no longer owns threw instead of passing.
2026-10-04 13:25:45 +00:00
]
})
const graph: Partial<SceneGraph> = {
getNode: mock(() => node)
}
const callOrder: string[] = []
r.renderEffects = mock((_c, _n, _r, _h, pass) => {
callOrder.push(`renderEffects:${pass}`)
})
r.drawStrokeWithAlign = mock(() => {
callOrder.push('drawStrokeWithAlign')
})
test: typecheck the test suites and fix what that found (#896) * build: typecheck the test suites Tests were in no TypeScript program: no tsconfig included tests/** or packages/*/tests/**, and bun strips types without checking them, so a fixture could drop a required field and keep passing until something read it. @types/bun moves to the root because it was installed per package only, and #cli-tests/* joins the paths the root config already carries. * test: fix the type errors the test suites were hiding Typechecking the tests turned up 1123 errors. Most were ordinary strictness, but some were real: `NodeChange` bound to Figma's plugin typings rather than the Kiwi codec in thirteen .fig tests, materializeInstance was called with six arguments against five so the blobs and source children were dropped, CanvasKit pixels were written to a plain object that never reached WASM, and assertions were made through accessors that do not exist, so they asserted nothing. Fixtures that had quietly lost a required field now carry it, nullable results are narrowed through the existing expectDefined helper rather than assumed, and stand-ins for CanvasKit and the editor go through one named helper instead of an unexplained cast at each site. No test was deleted, skipped, or weakened, and no `any`, non-null assertion, or ts-expect-error was introduced. * docs: record what typechecking the tests established Pins the app program's global types with an assertion rather than a note, since an unpinned types list lets any root @types package decide which platform src/** is judged against. The two environment faults that look like code regressions — Vite's dependency pre-bundle outliving a package rebuild, and heavy .fig suites failing under load — go to the development docs, where an explanation belongs. * fix: align @types/bun and keep node types resolvable when extended The root manifest declared a newer @types/bun than every package, which check:monorepo rejects, and pinning the app program's types left them unresolvable from a config that extends this one out of tree. * fix: fail the test typecheck when the compiler itself fails The gate matched diagnostics by substring, so a compiler or config failure that named no test file printed a pass while having checked nothing. Diagnostics are now split by whether they name a file: an unscoped one is the run failing and stops the gate, a test file's is a finding, and a source file's stays out by design. Also drops the parameter planComponentConstruction never read, and makes the inner-shadow verification script exit non-zero when it renders no image instead of logging and succeeding. * chore: merge master into tests-typecheck
2026-10-05 12:42:38 +00:00
renderShapeUncached(r, asCanvas(canvas), node, graph as SceneGraph)
const strokeIdx = callOrder.indexOf('drawStrokeWithAlign')
const frontIdx = callOrder.indexOf('renderEffects:front')
expect(strokeIdx).toBeGreaterThan(-1)
expect(frontIdx).toBeGreaterThan(-1)
expect(strokeIdx).toBeLessThan(frontIdx)
})
})