Test framework, stage 3: llui's tests in Catch2 by RyeMutt · Pull Request #471 · AlchemyViewer/Alchemy

RyeMutt · 2026-10-14T03:27:28Z

Description

The sixth wave of stage 3 of the test framework's move from TUT to Catch2 (TM23): llui's tests. 117 TUT files with 1,541 tests become three Catch2 binaries. Each test file still runs in a process of its own.

Binary	Files	Notes
`llui_tests`	113	no GL: templates, text and XUI below the view, and every widget above LLView over one headless UI
`llui_gl_tests`	3	the text view, a surface and a comparison drawn into llrender's hidden window and read back
`lluicolortable_test`	1	alone: each of its cases loads colour documents whose names stay in the table for every case after

Two gaps that one binary found. Each TUT executable registered only the widgets its own file happened to link. One binary registers them all, as the viewer does.

f3c0eec, xui_notes.xml was missing thirteen of Alchemy's widgets. The Library and the generated xui.xsd show its one-sentence notes, and these widgets had none. alxuischema's "every tag says what it is for" now sees them.
2899a2a, XUI Studio's window was too narrow for its panes.
- The window: it was 1620 wide, but its panes are authored for 1680. The canvas got 684 pixels instead of 744.
- The row under the canvas: it needed 748 pixels, so its last align buttons sat past its edge.
- The fix: the window opens at the width its panes add up to, and Reload is 54 wide like its neighbours, so the row is 744.
- Why TUT missed it: alxuifill's TUT build left out two of the row's controls, so the row fit.

The conversion:

afb2ae3, the conversion: tut2catch.py's output, with the names the idiom pass chose, given through --names. It marked two sites for review, both in code this Mac does not compile, and resolved them by hand:
- alkeymap's off-Mac Control-Alt check;
- alvimkeymap's off-Mac "nothing changed".
b9f739b, the idiom pass, by fifteen agents:
- Names: 1,512 of 1,541 changed. 1,102 had commas, which a Catch2 test spec reads as "or".
- Messages: 10,675 to 8,350, with 134 CAPTUREs.
- Helpers: 35 asserting helpers are named require_*.
- Labels: a label before a CHECK that calls one of those helpers is now a scoped INFO. The helper's own passing assertion took the unscoped label.
- Messages filled by a later call: where a call fills a message's text, the call's result is now kept and checked after the message, so the check is unchanged.
- Strength: 55 CHECKs whose subject the next lines read through are REQUIREs, and 1,251 REQUIREs that guarded nothing are CHECKs.
No case and no assertion was added, removed or moved.
747e770, a production fix in Linden's LLFocusMgr.
- What it keeps: the chain of the focused view and its ancestors. On the next change of focus it compares that chain with the new one, and tells the views not shared that they lost focus.
- The bug: two paths left handles in that chain:
  - focus taken without callbacks, which five destructors do (LLDragHandle, LLFolderViewItem, LLLayoutPanel, LLResizeHandle, the media focus);
  - a view that died holding focus.
- Impact: dead handles stay at the chain's root end, so no later change shared an ancestor with the old focus. Every view above it was told it lost focus it still had. A popover told that closes.
- Found: llfocusmgr's cases left such handles, and 22 cases failed after them in random orders.
- Test: two new cases, each failing without one of the two changes.
283c3d0, run setup, teardown and guard checks for this wave.
- Run setup: the headless UI and the colour table are made before the first case.
- After every case: popovers are escaped, focus, mouse capture and the top control are let go, the clipboard is emptied, and the views that asked to die are deleted.
- Guard checks: the floater view, the menu container, the key names, the root view, and the colour table's colours.
- Ends: the headless UI and the shared GL context end in the main's teardown, not at static destruction.
- Skin colours: they are loaded once for the process, through one accessor that tells the guard.
- Tested: I broke a case for each check on purpose, and each check named the case responsible.

Related Issues

Please link to a relevant GitHub issue for additional context.

Issue Link: none. This follows #468 (TM22).

Checklist

I have provided a clear title and detailed description for this pull request.
If useful, I have included media such as screenshots and video to show off my changes.
I have tested the changes locally and verified they work as intended.
All new and existing tests pass.
Code follows the project's style guidelines.
Documentation has been updated if needed.
Any dependent changes have been merged and published in downstream modules.
I have reviewed the contributing guidelines.

Additional Notes

Tested on macOS arm64, RelWithDebInfo:
- the full build is clean;
- all 462 tests of ctest -LE manual pass, and so do all 17 pending combined runs, llui's two among them;
- each llui binary, run whole, passes 10 random orders and 20 repeats with no state guard report;
- no anonymous type is used with a name-keyed registry in two files of one binary.
Per-case mode: with AL_TEST_PER_CASE, 1,543 case entries, and 1,542 pass. The other, aldiffview_gl's "range bands draw in the same number of batches", needs an earlier case of its file to have drawn first. That was already true under TUT; Test framework: fix the old test bugs in llui's tests #472 fixes it.
Parity: tut_parity.py holds every file to the TUT executables' recorded results: 1,541 tests, all ok.
- At the conversion commit: it holds plain for 115 files, and under --recounted for the two resolved by hand.
- Never-run sites: 4 assertion sites were reviewed as never running in a passing case on this host:
  - three are reached only when a mismatch is found;
  - one of altextview's runs only where a wide character's drawn width does not land near half a column. In this host's fonts it does.
- After the idiom pass: it holds under --recounted for all 117, every case's count unchanged.
Not compiled here: the off-Mac branches of alkeymap and alvimkeymap.
Old test bugs the idiom pass found are listed in b9f739b's message and left as they were. Test framework: fix the old test bugs in llui's tests #472 fixes them.

🤖 Generated with Claude Code

xui_notes.xml gives each tag the one sentence the Library and the generated xui.xsd show. Thirteen of Alchemy's widgets had none: angle_dial, choice_list, code_editor, color_table_panel, corner_field, diff_bar, diff_view, find_bar, image_field, offset_pad, output_view, pane_list and text_view. Each now has one, in the file's sorted run of Alchemy's own tags. alxuischema's "every tag says what it is for" did not see them: its TUT executable linked only the parts of llui its own file reached, and a widget registers itself from its own object, so the schema it read had none of these thirteen in it. In one binary with the rest of llui's tests, as in the viewer, every widget registers, and both alxuischema cases that hold the notes to the schema failed on them. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

… fits Its window was 1620 wide, but its three panes are authored at 400, 744 and 500, which with the bars between them come to a stack 1664 wide: a window of 1680. The navigator and the inspector keep their widths, so the canvas got what was left, 684. The row of controls under it is laid out from the left at fixed widths and needed 748, more than even the 744 it was authored in: its last align buttons sat past its edge. The window, its header bar, its stack and its status line are now the width the panes were authored for, so the canvas opens at 744; and Reload is 54 wide, as Show and Hide beside it are, so the row is 744 too. alxuifill's "every stack fills what it is in" did not see it. Its TUT executable linked only the parts of llui its own file reached, and a widget registers from its own object, so the row was built without the canvas_zoom spinner and the canvas_backdrop menu button, 102 pixels narrower. In one binary with the rest of llui's tests, as in the viewer, the spinner is there and the case failed; with the menu button linked in as well, the row now ends at the canvas's edge. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

TM23, the sixth wave of stage 3: llui's 117 TUT files, 1,541 tests, become two Catch2 binaries, each file still run in a process of its own: - llui_tests, 114 files with no GL: templates tested whole in their headers, text and XUI below the view, a file of llui's each, what comparing two texts is made of, and every widget above LLView, built over one headless UI on the source tree's skins and fonts; - llui_gl_tests, the three that draw into llrender's hidden window and read the pixels back: the text view, a surface and its tab strip, and a comparison. LLUI_TEST_APP_DIR is defined for every file of each. The fakes llui asks the viewer for are a SUPPORT object of both, and the benchmarks keep linking them as they did. The files are tut2catch.py's output, with the names the idiom pass that follows chose. It marked two sites TM-CHECK, both in code this Mac does not compile, resolved by hand: alkeymap's off-Mac check that no binding is Control-Alt with a key that types (an ensure() whose message is built), and alvimkeymap's off-Mac "nothing changed" (an ensure_equals() over strings). tut_parity.py, against the TUT executables' recorded results (1,541 tests, all ok), holds plain for 115 files and under --recounted for the two hand-resolved ones, with four assertion sites reviewed as never running in a passing case: three reached only to fail, inside an if that finds a mismatch (alcodeeditor, allinetable, alsyntaxhighlighter), and one of altextview's that runs only where a wide character's drawn width does not land near half a column, as it does in this host's fonts. All 119 entries pass. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

The idiom pass over TM23's conversion, by fifteen agents on copies, merged back onto it. Names, messages and assertion strength only: every case and every assertion is where the conversion put it. - Names: 1,512 of 1,541 changed. Most were set_test_name sentences, and 1,102 had commas, which a Catch2 test spec reads as "or"; many ran past a hundred characters. Where a shorter name dropped a why, a comment above the case keeps it. - Messages: 10,675 to 8,350, with 134 CAPTUREs. Most of those removed restated their check; most that remain say which step, line or key a failure is at. - 35 asserting helpers are named require_*: require_key, require_typed, require_url, require_same_layout, require_lexed and the like. A label before a CHECK that calls one is a scoped INFO around it: the helper's own passing assertion took an unscoped one. - Messages built from what a later call fills (an error string) come after the call: its result is kept and checked, the same value. - 55 CHECKs whose subject the next lines index or follow are REQUIREs, and 1,251 REQUIREs that guarded nothing are CHECKs. - alkeymap's and alvimkeymap's namespace tut, kept only for the code the converter could not compile, is an anonymous namespace. tut_parity.py holds all 117 files under --recounted, every case's count unchanged, with the conversion's four sites reviewed as never running here. Old test bugs the agents found, left as they were: - reads with no guard, where a failure would crash rather than fail: a findChild(), find(), getParentByType(), tabButton() or optional used unchecked, and [0], front() or back() of what nothing sized, in many widget tests (quickopen, specimenlist, studiofloater, propertygrid, tabcontainer, dockpanel, popover, jumpbar, codeeditor, textview, textlayout, undostack, snippetsession, diffmodel, alxui*); - checks that cannot fail: CHECK(true) in alvimkeymap; checks after a getChild() that makes a dummy, after the check that implies them, or over the test's own expected model (altextview_gl's wave coverage); alxuiservice's "|| true"; alpanelist's fold check before any callback is set; altextundo's forgotten-steps case, where nothing is forgotten; - checks of something other or less than they say: alsyntaxgrammars' and alsyntaxhighlighter's load messages built before the loads; altextdiff checks minimal against patience twice where histogram was meant; llemojidictionary's gender check passes by variant order; llurlentry's "multicase" inputs are lower case; comments promising checks that are not there (aldiffview_gl, alcolortablepanel, lltabcontainer, propertygrid, settingrow and others); - state a failure leaks: altextview_gl's sResolutionGeneration, never put back; signal connections over locals that a teardown signal could reach; altextsearch's two threads write one bool unsynchronised; - leftovers: stale comments (alviewtype, alxuiedit, alpanelist's comparator), and cases with no ui.ok() skip. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…focus LLFocusMgr keeps the chain of the focused view and its ancestors, and on the next change of focus drops the ancestors the old focus shares with the new, then tells the rest of the old chain it lost focus. Two ways left handles to views that were gone, or no longer focused, in that chain: - removeKeyboardFocusWithoutCallback, which LLDragHandle, LLFolderViewItem, LLLayoutPanel, LLResizeHandle and the media focus call from their destructors, cleared the focus but kept the chain; - the walk down the old chain told and dropped only the views still alive, so a view that died holding focus kept its dead handle there. Handles stuck at the chain's root end match no ancestor of the new focus, so every change after was found to share none: every view above the old focus was told it lost focus it still had, as were the views above the one whose focus was taken without callbacks. A popover told so closes, as an editor that lost focus would. Focus taken without callbacks now leaves no chain, and dead handles are dropped before the chains are compared. Two new llfocusmgr cases fail without each change. In one process with the rest of llui's tests, llfocusmgr's cases left such handles, and popovers, Go to Line and the studio floater's questions failed after them in random orders. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…the state guard watches what llui keeps llui_test_checks, a listener in llui's test binaries: - before the first case, makes the headless UI, as the viewer has it before it builds a view (in llui_gl_tests on llrender's shared context, with the fonts' glyphs in textures), and the colour table, loaded by none; - after every case, clears what a frame of the viewer clears between one thing and the next: escapes any popover left up, lets go of the keyboard, the mouse and the top, empties the clipboard, and deletes the views that asked to die; - gives the state guard, after every case, the floater view, LLMenuGL's menu container, LLKeyboard's key names, the UI's root view, and the colour table's colours, the loaded and the user's, each with its value: not its generation, which a colour set and set back still moves. Each binary takes llrender's and llcommon's checks too. I broke a case each way, and each check named it; and a case that left focus and the clipboard full left the next case nothing of either. The headless UI and the shared GL context end with the main, after the last case, rather than at static destruction: HeadlessUI::get() hands its end to the teardown in a Catch2 binary, and puts back the process's directory object, and uiGL() is llrender's sharedHeadlessGL(). Making the UI is the state the guard compares with from then on. The plan made the UI at run start, so that no case is the one that made it; the 52 files that never ask for it pass with it there. Three tests read the skin's colours, each loading them once for its process behind a static of its own; ll_test::loadSkinColours() loads them once for the process and tells the guard, since the table's names cannot be taken back out. lluicolortable's cases each load documents of their own, whose names stay for every case after, so it is a binary of its own, lluicolortable_test, without llui's checks, as the plan allowed. llui_tests, llui_gl_tests and lluicolortable_test, each run whole, pass 10 random orders and 20 repeats with no guard report. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

coderabbitai · 2026-10-14T03:27:42Z

Important

Review skipped

Too many files!

This PR contains 126 files, which is 26 over the limit of 100.

To get a review, reduce the PR to 100 files or fewer by splitting it into smaller PRs or changing its base branch.

Upgrade to a paid plan to raise the limit.

This review couldn't start because sufficient usage credits or metered capacity aren't available. Add credits or update usage-based reviews in the billing tab, then retry.

⚙️ Run configuration

Configuration used: Repository UI
Review profile: CHILL
Plan: Advanced
Run ID: 99deb780-ddc2-4690-baab-a26890511a99

📥 Commits

Reviewing files that changed from the base of the PR and between 9649a04 and 0376800.

📒 Files selected for processing: 126

indra/cmake/ALAddTestBinary.cmake
indra/llui/CMakeLists.txt
indra/llui/llfocusmgr.cpp
indra/llui/tests/alanchoredranges_test.cpp
indra/llui/tests/alangledial_test.cpp
indra/llui/tests/albracketindex_test.cpp
indra/llui/tests/alcanvasview_test.cpp
indra/llui/tests/alchangepeek_test.cpp
indra/llui/tests/alchoicelist_test.cpp
indra/llui/tests/alcodecards_test.cpp
indra/llui/tests/alcodeeditor_test.cpp
indra/llui/tests/alcolorfield_test.cpp
indra/llui/tests/alcolorpicker_test.cpp
indra/llui/tests/alcolorsheet_test.cpp
indra/llui/tests/alcolortablepanel_test.cpp
indra/llui/tests/alcompletionmodel_test.cpp
indra/llui/tests/alcornerfield_test.cpp
indra/llui/tests/aldifflexer_test.cpp
indra/llui/tests/aldiffmerge_test.cpp
indra/llui/tests/aldiffmodel_test.cpp
indra/llui/tests/aldiffview_gl_test.cpp
indra/llui/tests/aldiffview_test.cpp
indra/llui/tests/aldockpanel_test.cpp
indra/llui/tests/aldraggesture_test.cpp
indra/llui/tests/alemptystate_test.cpp
indra/llui/tests/alfenwicktree_test.cpp
indra/llui/tests/alfindbar_test.cpp
indra/llui/tests/alfixlistmodel_test.cpp
indra/llui/tests/alflagsfield_test.cpp
indra/llui/tests/alfolderfilter_test.cpp
indra/llui/tests/alfoldmodel_test.cpp
indra/llui/tests/alfollowscontrol_test.cpp
indra/llui/tests/alfontfield_test.cpp
indra/llui/tests/alfuzzymatch_test.cpp
indra/llui/tests/alheadlessui_fixture.h
indra/llui/tests/alhistorylist_test.cpp
indra/llui/tests/alimagefield_test.cpp
indra/llui/tests/aljumpbar_test.cpp
indra/llui/tests/alkeycapture_test.cpp
indra/llui/tests/alkeymap_test.cpp
indra/llui/tests/allinetable_test.cpp
indra/llui/tests/almaxima_test.cpp
indra/llui/tests/almenuslot_test.cpp
indra/llui/tests/alnotecarditems_test.cpp
indra/llui/tests/aloffsetpad_test.cpp
indra/llui/tests/aloutputview_test.cpp
indra/llui/tests/alpanefolds_test.cpp
indra/llui/tests/alpanelist_test.cpp
indra/llui/tests/alplace_test.cpp
indra/llui/tests/alpopover_test.cpp
indra/llui/tests/alprefixsums_test.cpp
indra/llui/tests/alpropertygrid_test.cpp
indra/llui/tests/alquickopen_test.cpp
indra/llui/tests/alscopebar_test.cpp
indra/llui/tests/alsettingrow_test.cpp
indra/llui/tests/alsnippetsession_test.cpp
indra/llui/tests/alspecimenlist_test.cpp
indra/llui/tests/alstringmatch_test.cpp
indra/llui/tests/alstudiofloater_test.cpp
indra/llui/tests/alsurface_gl_test.cpp
indra/llui/tests/alsurface_test.cpp
indra/llui/tests/alsyntaxgrammars_test.cpp
indra/llui/tests/alsyntaxhighlighter_test.cpp
indra/llui/tests/altabstrip_test.cpp
indra/llui/tests/altextcarets_test.cpp
indra/llui/tests/altextdiff_test.cpp
indra/llui/tests/altextdocument_test.cpp
indra/llui/tests/altextediting_test.cpp
indra/llui/tests/altextfind_test.cpp
indra/llui/tests/altextgotoline_test.cpp
indra/llui/tests/altextindent_test.cpp
indra/llui/tests/altextlayout_test.cpp
indra/llui/tests/altextsearch_test.cpp
indra/llui/tests/altextspelling_test.cpp
indra/llui/tests/altextundo_test.cpp
indra/llui/tests/altextview_gl_test.cpp
indra/llui/tests/altextview_test.cpp
indra/llui/tests/aluigl_fixture.h
indra/llui/tests/alundostack_test.cpp
indra/llui/tests/alunifieddiff_test.cpp
indra/llui/tests/alviewtype_test.cpp
indra/llui/tests/alvimkeymap_test.cpp
indra/llui/tests/alvimmappings_test.cpp
indra/llui/tests/alvimpattern_test.cpp
indra/llui/tests/alvimregisters_test.cpp
indra/llui/tests/alxuicatalog_test.cpp
indra/llui/tests/alxuidiagnostics_test.cpp
indra/llui/tests/alxuidocuments_test.cpp
indra/llui/tests/alxuiedit_test.cpp
indra/llui/tests/alxuifields_test.cpp
indra/llui/tests/alxuifill_test.cpp
indra/llui/tests/alxuifindings_test.cpp
indra/llui/tests/alxuilint_test.cpp
indra/llui/tests/alxuinotes_test.cpp
indra/llui/tests/alxuioverlay_test.cpp
indra/llui/tests/alxuischema_test.cpp
indra/llui/tests/alxuiselection_test.cpp
indra/llui/tests/alxuiservice_test.cpp
indra/llui/tests/alxuishellbuild_test.cpp
indra/llui/tests/alxuisourcemap_test.cpp
indra/llui/tests/alxuitranslate_test.cpp
indra/llui/tests/alxuitreemodel_test.cpp
indra/llui/tests/llaccordionctrltab_test.cpp
indra/llui/tests/llchatautocompletemodel_test.cpp
indra/llui/tests/llcombobox_test.cpp
indra/llui/tests/llemojidictionary_test.cpp
indra/llui/tests/llfloater_test.cpp
indra/llui/tests/llfocusmgr_test.cpp
indra/llui/tests/lllayoutstack_test.cpp
indra/llui/tests/llradiogroup_test.cpp
indra/llui/tests/llresizebar_test.cpp
indra/llui/tests/llspinctrl_test.cpp
indra/llui/tests/llstatbar_test.cpp
indra/llui/tests/lltabcontainer_test.cpp
indra/llui/tests/lltextparser_test.cpp
indra/llui/tests/llui_test_checks.cpp
indra/llui/tests/lluicolortable_test.cpp
indra/llui/tests/lluistring_test.cpp
indra/llui/tests/llurlentry_test.cpp
indra/llui/tests/llurlmatch_test.cpp
indra/llui/tests/llurlregistry_test.cpp
indra/llui/tests/llview_test.cpp
indra/llui/tests/llxuiparser_test.cpp
indra/newview/skins/default/xui/en/floater_xui_studio.xml
indra/newview/skins/default/xui/en/xui_notes.xml
indra/test/CMakeLists.txt

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

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.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

AL_TEST_CATCH2 was defined on each test binary and not on the OBJECT libraries of checks it links, so a fixture the tests share with their checks had two inline bodies in one binary. HeadlessUI::get() keeps the process's one UI in a function's static: the Catch2 body in `made`, which the harness's teardown ends, and the TUT body in `ui`. Apple's linker kept the checks' TUT body for every call, so the Catch2 one never ran; GCC and MSVC inlined each where it was called, and the run setup's UI and the cases' were two. On Linux the second failed on a settings group the first had named, and on Windows every case that asked for the UI crashed: 63 of llui_tests' files and llui_gl_tests on each CI row. The shared GL context's accessor had the same split. The harness carries the define now, as a usage requirement of al_test_support, so whatever links it is built for Catch2: the binaries and their check libraries alike. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

<binary>/stray_cases runs the binary with a filter that matches nothing, to find the cases no file's entry runs. A run starts its listeners whatever it matches, and llui_gl_tests' checks make the shared GL context before the first case, so on a host without one -- the Windows rows, which build GL tests disabled -- the stray run crashed where every file's own entry was skipped. It is disabled with them now; llrender's GL binaries, whose checks make nothing, lose a run that could only pass. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

RyeMutt and others added 6 commits October 13, 2026 19:11

RyeMutt mentioned this pull request Oct 14, 2026

Test framework: fix the old test bugs in llui's tests #472

Merged

4 of 9 tasks

github-actions Bot added c/cpp llui cmake viewer skins labels Oct 14, 2026

RyeMutt added this pull request to stack #473 October 14, 2026 03:28

RyeMutt and others added 2 commits October 14, 2026 01:02

This was referenced Oct 14, 2026

Test framework, stage 3: newview's tests in Catch2 #474

Merged

Test framework: fix the old test bugs in newview's tests #475

Merged

Test framework, stage 4: the benchmarks in Catch2 #478

Merged

RyeMutt merged commit 3efd6ff into develop Oct 14, 2026
21 checks passed

RyeMutt deleted the rye/tests-tm23 branch October 14, 2026 10:53

RyeMutt mentioned this pull request Oct 14, 2026

Test framework, stage 5: TUT removed #479

Merged

5 of 9 tasks

Navigation Menu

Repository navigation

Test framework, stage 3: llui's tests in Catch2 - #471

Test framework, stage 3: llui's tests in Catch2#471
RyeMutt merged 8 commits into
developfrom
rye/tests-tm23

RyeMutt commented Oct 14, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented Oct 14, 2026 •

edited

Loading

Review skipped

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

RyeMutt commented Oct 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Related Issues

Checklist

Additional Notes

Uh oh!

coderabbitai Bot commented Oct 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

RyeMutt commented Oct 14, 2026 •

edited

Loading

coderabbitai Bot commented Oct 14, 2026 •

edited

Loading