Add applyToGroupOther option to onIntersect with generator support and tests#522
Merged
tracygardner merged 6 commits intomainfrom Apr 11, 2026
Merged
Conversation
Deploying flockxr with
|
| Latest commit: |
6a60c3f
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://6949d2d3.flockxr.pages.dev |
| Branch Preview URL: | https://codex-add-on-collision-suppo.flockxr.pages.dev |
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 10 minutes and 59 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
📝 WalkthroughWalkthroughAdds group-aware intersection registration: Changes
Sequence Diagram(s)sequenceDiagram
participant Block as Blockly Block
participant Gen as Event Generator
participant Physics as flockPhysics
participant Scene as Scene Graph
participant Registry as ActionManager
Block->>Gen: generate onIntersect code (top-level)
Gen->>Physics: onIntersect(src, otherSeed, {trigger, callback, applyToGroupOther:true})
Physics->>Physics: compute canonical group root from otherSeed
alt Scene present
Physics->>Scene: query meshes matching group root
Scene-->>Physics: list of mesh names
Physics->>Physics: filter out src, dedupe names
loop for each match
Physics->>Physics: call onIntersect(src, match, {applyToGroupOther:false,...})
Physics->>Registry: register ExecuteCodeAction for pair
Registry-->>Physics: action registered
end
else Scene absent
Physics->>Physics: enqueue pending entry in flock.pendingIntersections
end
Note over Physics,Scene: When new meshes are announced, announceMeshReady processes pendingIntersections and wires callbacks for new, non-duplicate RHS meshes
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (2)
api/physics.js (1)
614-615: Deduplicate group-root parsing helper.
getGroupRootis now duplicated in multiple methods. Please move it to a shared module-level helper so grouping behavior can’t drift between trigger/intersection paths.Proposed refactor
+const getGroupRoot = (name) => + name.includes("__") ? name.split("__")[0] : name.split("_")[0]; + export const flockPhysics = { onTrigger( meshName, { trigger, callback, mode = "wait", applyToGroup = false }, ) { - const getGroupRoot = (name) => - name.includes("__") ? name.split("__")[0] : name.split("_")[0]; - const groupName = getGroupRoot(meshName);🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@api/physics.js` around lines 614 - 615, Extract the getGroupRoot implementation into a single module-level helper function (e.g., declare const getGroupRoot = (name) => ...) and replace all duplicated local copies in trigger and intersection-related functions with calls to this shared helper; update any files or scopes that currently reimplement getGroupRoot so they reference the module-level symbol, remove the duplicated definitions, and ensure behavior remains identical (splitting on "__" first, then "_" as in the original implementation).tests/physics.test.js (1)
220-226: Strengthen self-pair exclusion assertion.The test currently proves positive behavior for the sibling mesh, but it doesn’t directly assert that self-trigger does not fire. Add a self-trigger check before the sibling trigger.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/physics.test.js` around lines 220 - 226, Before invoking the sibling trigger, explicitly assert that triggering the source mesh itself does not increment the counter: call sourceMesh.actionManager.processTrigger with flock.BABYLON.ActionManager.OnIntersectionEnterTrigger and { mesh: sourceMesh } and assert count remains 0; then call sourceMesh.actionManager.processTrigger with { mesh: otherMesh } and assert count becomes 1 (using the existing expect(count).to.equal(1)), ensuring the self-trigger exclusion is tested directly.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@tests/physics.test.js`:
- Around line 161-168: The test is racing because flock.onIntersect is
asynchronous; modify the test to await the async registration by awaiting the
Promise returned by flock.onIntersect wherever it's called (e.g., the
onIntersect calls in the tests), so registration completes before calling
processTrigger; update both occurrences of flock.onIntersect (the one that
currently increments count and the other at the later occurrence) to use await
so callbacks are reliably registered before invoking processTrigger.
---
Nitpick comments:
In `@api/physics.js`:
- Around line 614-615: Extract the getGroupRoot implementation into a single
module-level helper function (e.g., declare const getGroupRoot = (name) => ...)
and replace all duplicated local copies in trigger and intersection-related
functions with calls to this shared helper; update any files or scopes that
currently reimplement getGroupRoot so they reference the module-level symbol,
remove the duplicated definitions, and ensure behavior remains identical
(splitting on "__" first, then "_" as in the original implementation).
In `@tests/physics.test.js`:
- Around line 220-226: Before invoking the sibling trigger, explicitly assert
that triggering the source mesh itself does not increment the counter: call
sourceMesh.actionManager.processTrigger with
flock.BABYLON.ActionManager.OnIntersectionEnterTrigger and { mesh: sourceMesh }
and assert count remains 0; then call sourceMesh.actionManager.processTrigger
with { mesh: otherMesh } and assert count becomes 1 (using the existing
expect(count).to.equal(1)), ensuring the self-trigger exclusion is tested
directly.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: bf963dbf-b4bf-49f3-bb3f-acf6dabd7efd
📒 Files selected for processing (3)
api/physics.jsgenerators/generators-events.jstests/physics.test.js
Deploying flockdev with
|
| Latest commit: |
6a60c3f
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://f67a19aa.flockdev.pages.dev |
| Branch Preview URL: | https://codex-add-on-collision-suppo.flockdev.pages.dev |
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@api/physics.js`:
- Around line 614-649: The code derives groupName from the raw otherMeshName and
stores/queries pending entries under that raw key, which can mismatch sanitized
scene mesh names; before computing getGroupRoot(otherMeshName) and before
scanning flock.scene.meshes, canonicalize otherMeshName using the same
exact-or-normalized lookup used by flock.whenModelReady() (i.e., resolve the
alias or normalized name via the flock.whenModelReady lookup logic), then
compute groupName = getGroupRoot(canonicalName) and use that canonicalName when
registering/queueing into flock.pendingIntersections and when comparing against
flock.scene.meshes so announceMeshReady() and the scene scan will find the same
group key.
In `@flock.js`:
- Around line 2504-2531: The pending trigger entries lack the original target
metadata so announceMeshReady's replay logic always treats entries as non-group
and rebinds them to the announcing mesh; update the queuing in
flockPhysics.onTrigger (or wherever flock.pendingTriggers entries are pushed) to
store the original meshName and applyToGroup (e.g. include meshName and
applyToGroup in each queued object) and change the replay logic in
announceMeshReady/wherever you iterate flock.pendingTriggers to use that stored
meshName/applyToGroup: when applyToGroup is true, fan out to all matching meshes
as before, and when false, invoke flock.onTrigger using the stored meshName (not
the announcing meshName) with applyToGroup set to false to avoid recursion.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 8d8015fe-bfbc-416f-85fb-f39108f7dcc0
📒 Files selected for processing (3)
api/physics.jsflock.jstests/physics.test.js
✅ Files skipped from review due to trivial changes (1)
- tests/physics.test.js
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (1)
flock.js (1)
2518-2532: Register only the newly ready sibling here.This path rescans the whole group and replays
onTrigger()for every sibling on each readiness event. MirroringpendingIntersections.registeredOtherswith a per-pendingregisteredMeshesset would make this O(1) per new sibling and avoid duplicate registration attempts.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@flock.js` around lines 2518 - 2532, The current applyToGroup branch rescans the entire group and re-invokes flock.onTrigger for every sibling each time, causing duplicate attempts; change it to track which meshes have already been registered per pending entry by adding a Set (e.g., pending.registeredMeshes) to the pending object and, instead of computing matching = flock.scene.meshes.filter(...), only register the single newly-ready mesh (the one that triggered this readiness event) if its name is not in pending.registeredMeshes; add the mesh name to pending.registeredMeshes after calling flock.onTrigger and keep remaining.push(pending) as before to preserve future sibling registration semantics.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@flock.js`:
- Around line 2534-2538: The check for a target control uses direct root
children lookup and misses nested GUI controls; update the detection logic in
the block that computes targetExists to use
flock.scene?.UITexture?.getControlByName(targetMeshName) (falling back to
flock.scene?.getMeshByName(targetMeshName)) instead of
_rootContainer._children.some(...), mirroring the recursive approach used by
whenModelReady(), so pending trigger replay correctly finds nested controls such
as those inside flock.stackPanel.
In `@tests/events.test.js`:
- Around line 414-452: The test leaves applyToGroup registrations in
flock.pendingTriggers which causes order-dependent failures; add cleanup in the
test suite's afterEach() to reset flock.pendingTriggers (e.g., clear the
Map/array or assign a fresh empty container) so deferred callbacks registered
with applyToGroup are removed between tests; locate the afterEach() that
currently only disposes meshes and add a line to clear/reset
flock.pendingTriggers to ensure no live callbacks remain.
---
Nitpick comments:
In `@flock.js`:
- Around line 2518-2532: The current applyToGroup branch rescans the entire
group and re-invokes flock.onTrigger for every sibling each time, causing
duplicate attempts; change it to track which meshes have already been registered
per pending entry by adding a Set (e.g., pending.registeredMeshes) to the
pending object and, instead of computing matching =
flock.scene.meshes.filter(...), only register the single newly-ready mesh (the
one that triggered this readiness event) if its name is not in
pending.registeredMeshes; add the mesh name to pending.registeredMeshes after
calling flock.onTrigger and keep remaining.push(pending) as before to preserve
future sibling registration semantics.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: fd91ab49-7655-48c1-8358-d793d00f2357
📒 Files selected for processing (4)
api/physics.jsflock.jstests/events.test.jstests/physics.test.js
✅ Files skipped from review due to trivial changes (1)
- tests/physics.test.js
🚧 Files skipped from review as they are similar to previous changes (1)
- api/physics.js
Comment on lines
+414
to
+452
| it("replays pending group trigger across siblings when applyToGroup is true", async function () { | ||
| const first = "lategroup_1"; | ||
| const second = "lategroup_2"; | ||
|
|
||
| let count = 0; | ||
| flock.onTrigger(first, { | ||
| trigger: "OnPickTrigger", | ||
| callback: () => count++, | ||
| applyToGroup: true, | ||
| }); | ||
|
|
||
| await flock.createBox(first, { | ||
| width: 1, | ||
| height: 1, | ||
| depth: 1, | ||
| position: [0, 0, 0], | ||
| }); | ||
| await flock.createBox(second, { | ||
| width: 1, | ||
| height: 1, | ||
| depth: 1, | ||
| position: [2, 0, 0], | ||
| }); | ||
| meshIds.push(first, second); | ||
|
|
||
| const mesh1 = flock.scene.getMeshByName(first); | ||
| const mesh2 = flock.scene.getMeshByName(second); | ||
| expect(mesh1).to.exist; | ||
| expect(mesh2).to.exist; | ||
|
|
||
| mesh1.actionManager?.processTrigger( | ||
| flock.BABYLON.ActionManager.OnPickTrigger, | ||
| ); | ||
| mesh2.actionManager?.processTrigger( | ||
| flock.BABYLON.ActionManager.OnPickTrigger, | ||
| ); | ||
|
|
||
| expect(count).to.equal(2); | ||
| }); |
There was a problem hiding this comment.
Reset deferred trigger state after this replay test.
applyToGroup: true registrations stay in flock.pendingTriggers by design, so this test leaves a live callback behind. afterEach() only disposes meshes, which makes later tests order-dependent if they reuse the same group root or the same shared flock instance.
🧹 Suggested cleanup in afterEach()
afterEach(function () {
// Clear any custom events registered during tests
if (flock.events) {
Object.keys(flock.events).forEach((key) => delete flock.events[key]);
}
+ flock.pendingTriggers = new Map();
+ flock.pendingIntersections = new Map();
// Dispose any meshes created during tests
meshIds.forEach((id) => {🧰 Tools
🪛 GitHub Actions: Prettier
[warning] Prettier reports formatting issue (listed under 'warn').
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@tests/events.test.js` around lines 414 - 452, The test leaves applyToGroup
registrations in flock.pendingTriggers which causes order-dependent failures;
add cleanup in the test suite's afterEach() to reset flock.pendingTriggers
(e.g., clear the Map/array or assign a fresh empty container) so deferred
callbacks registered with applyToGroup are removed between tests; locate the
afterEach() that currently only disposes meshes and add a line to clear/reset
flock.pendingTriggers to ensure no live callbacks remain.
tracygardner
deleted the
codex/add-on-collision-support-for-matching-meshes
branch
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Motivation
Description
applyToGroupOtheroption toflock.onIntersectto expand theotherMeshNameinto all matching scene meshes sharing the same group root and register intersections for each match, skipping the original source when appropriate, and preventing recursive expansion by resetting the flag for recursive calls.getGroupRoothelper that derives group roots by splitting the mesh name on__or_and using the first segment.on_collisionblock to emitapplyToGroupOther: truefor top-level blocks so generated code automatically expands right-hand groups at runtime.Testing
should register intersections for all matching right-hand group meshesandshould skip self-pair when expanding right-hand collision group, and all tests passed.Codex Task
Summary by CodeRabbit
New Features
Improvements
Tests