gramps/LlamaCasty
0
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Preview: promote selected layer for hit testing

Browse Source 
Temporarily raise the selected preview item and prioritize it in the custom hit test without changing scene order. Refresh integration-test expectations for Brand logo layers and explicit thumbnail assets.

WPF references: https://learn.microsoft.com/en-us/dotnet/api/system.windows.controls.panel.zindex?view=windowsdesktop-10.0 https://learn.microsoft.com/en-us/dotnet/desktop/wpf/graphics-multimedia/hit-testing-in-the-visual-layer
This commit is contained in:
gramps
2026-10-28 13:09:15 -07:00
parent 7da287244b
commit 769c616706
13 changed files with 326 additions and 79 deletions
Download Patch File Download Diff File Expand all files Collapse all files
MyMistakes.md
+16
View File
@@ -775,6 +775,22 @@ with `#Name`) + panel properties (`IsHitTestVisible/Opacity/children/actual size
`camera 'test-camera' failed` are test noise, not DB state — to check pollution, query
the DB directly (`python3 sqlite3`, `SELECT DeviceId FROM Webcam`), not the log.
## Deterministic test for a selected layer above an overlap (2026-10-28)
When preview rendering uses WPF z-order but dragging uses an additional custom hit-test, test both
contracts without injecting a physical cursor: in a real `MainWindow` on the `RealApp` STA, overlap
two scene elements, select the lower one through the layer list, assert `Panel.GetZIndex` raises its
container, assert `window.InputHitTest(point)` resolves the selected visual, and call the same
production hit-test method used by the drag handler to assert it chooses that selected model. Clear
selection and assert both hit paths return the original top layer and the scene's element order is
unchanged. WPF documents that input hit testing returns the input element at a point and that visual
hit-test results follow z-order: <https://learn.microsoft.com/en-us/dotnet/desktop/wpf/graphics-multimedia/hit-testing-in-the-visual-layer>.
Do not use `SetCursorPos`/`mouse_event` for this specific assertion: the test passed alone but was
intermittent in the full suite, with a cursor move landing at the canvas edge (`X=1900`) rather than
the intended offset. Physical injection remains appropriate only when the behavior under test is the
actual OS mouse-routing/drag path (as in the layer-list reorder test).
## A "known failure" label without a recorded cause = a bug on life support
(2026-09-01, the audio triple-take) One line — `_delayedMix` (nullable, added by TASK 22,