Repository navigation
Conversation
Bumps the npm-minor-and-patch group with 5 updates: | Package | From | To | | --- | --- | --- | | [@types/jszip](https://github.com/Stuk/jszip) | `3.4.0` | `3.4.1` | | [@types/react](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/react) | `19.2.17` | `19.2.18` | | [@types/react-dom](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/react-dom) | `19.2.3` | `19.2.4` | | [@vitejs/plugin-react](https://github.com/vitejs/vite-plugin-react/tree/HEAD/packages/plugin-react) | `6.0.4` | `6.0.5` | | [vite](https://github.com/vitejs/vite/tree/HEAD/packages/vite) | `8.1.5` | `8.2.0` | Updates `@types/jszip` from 3.4.0 to 3.4.1 - [Changelog](https://github.com/Stuk/jszip/blob/main/CHANGES.md) - [Commits](https://github.com/Stuk/jszip/commits) Updates `@types/react` from 19.2.17 to 19.2.18 - [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases) - [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/react) Updates `@types/react-dom` from 19.2.3 to 19.2.4 - [Release notes](https://github.com/DefinitelyTyped/DefinitelyTyped/releases) - [Commits](https://github.com/DefinitelyTyped/DefinitelyTyped/commits/HEAD/types/react-dom) Updates `@vitejs/plugin-react` from 6.0.4 to 6.0.5 - [Release notes](https://github.com/vitejs/vite-plugin-react/releases) - [Changelog](https://github.com/vitejs/vite-plugin-react/blob/main/packages/plugin-react/CHANGELOG.md) - [Commits](https://github.com/vitejs/vite-plugin-react/commits/plugin-react@6.0.5/packages/plugin-react) Updates `vite` from 8.1.5 to 8.2.0 - [Release notes](https://github.com/vitejs/vite/releases) - [Changelog](https://github.com/vitejs/vite/blob/main/packages/vite/CHANGELOG.md) - [Commits](https://github.com/vitejs/vite/commits/create-vite@8.2.0/packages/vite) --- updated-dependencies: - dependency-name: "@types/jszip" dependency-version: 3.4.1 dependency-type: direct:development update-type: version-update:semver-patch dependency-group: npm-minor-and-patch - dependency-name: "@types/react" dependency-version: 19.2.18 dependency-type: direct:development update-type: version-update:semver-patch dependency-group: npm-minor-and-patch - dependency-name: "@types/react-dom" dependency-version: 19.2.4 dependency-type: direct:development update-type: version-update:semver-patch dependency-group: npm-minor-and-patch - dependency-name: "@vitejs/plugin-react" dependency-version: 6.0.5 dependency-type: direct:development update-type: version-update:semver-patch dependency-group: npm-minor-and-patch - dependency-name: vite dependency-version: 8.2.0 dependency-type: direct:development update-type: version-update:semver-minor dependency-group: npm-minor-and-patch ... Signed-off-by: dependabot[bot] <support@github.com>
- Refactored SQL queries in the `pack_words` and `predictions` methods to improve execution speed and reduce overhead. - Removed redundant operations to streamline data retrieval processes.
Drive relative moves from an authoritative position for the gesture so Windows touch absolute cursor placement cannot yank the pointer onto the finger. Co-authored-by: Cursor <cursoragent@cursor.com>
- Added a new module for cursor highlighting specific to Windows. - Updated mouse handling functions to utilize absolute cursor positioning, improving gesture responsiveness. - Enhanced trackpad gesture functions to maintain cursor visibility and state during interactions. - Refactored mouse event handling to support additional mouse event flags for better compatibility.
Map short stationary taps to left click and a second tap in the double-click window to double-click, matching real trackpad behavior.
fix: stabilize touch trackpad cursor and add tap-to-click
…minor-and-patch-48aa4d6334 chore(deps-dev): bump the npm-minor-and-patch group with 5 updates
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Refactor collapsed sizing with CollapsedFabCount enum and shared FAB_HOVER_SLACK constant synced between Rust and TS. Window dimensions now include hover headroom; CollapsedFab uses reduced scale and min sizes. Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
ReachPanel is Windows-only. Remove stub modules, macOS CI jobs, and cross-platform fallbacks; add compile_error guard for non-Windows builds. Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
…nly guardrails Pass the main window scale factor into compute_window_layout, commit the windows-only Cursor rule, run CI on PRs targeting dev, and document the platform policy in CONTRIBUTING. Co-authored-by: Cursor <cursoragent@cursor.com>
…eyboard Toolbar button collapses back to the FAB stack (including transparent mode) and suppresses auto-show until external focus is lost once. Co-authored-by: Cursor <cursoragent@cursor.com>
feat: mini mode layout fixes and Windows-only platform
|
Warning Review limit reached
Next review available in: 34 minutes You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: ⛔ Files ignored due to path filters (1)
📒 Files selected for processing (4)
📝 WalkthroughWalkthroughReachPanel is now Windows-only. The change adds Mini Mode, monitor-aware layouts, touch and mouse layout profiles, native focus tracking, cursor highlighting, improved trackpad gestures, transparent keyboard controls, and Windows-only release configuration. ChangesWindows-only platform and release contract
Mini Mode and input interaction
Estimated code review effort: 5 (Critical) | ~120 minutes Sequence Diagram(s)sequenceDiagram
participant App
participant AppStore
participant FocusTarget
participant Tauri
participant MiniModeShell
App->>AppStore: initialize and detect Mini Mode
AppStore->>Tauri: query monitors and apply layout
FocusTarget-->>AppStore: publish input-focus-changed
AppStore->>Tauri: animate keyboard visibility
AppStore->>MiniModeShell: update Mini Mode state
Possibly related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✨ Finishing Touches 💡 2📝 Generate docstrings 💡
⚔️ Resolve merge conflicts 💡
🧪 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: 16
Note
Due to the large number of review comments, Critical, Major severity comments were prioritized as inline comments.
🟡 Minor comments (6)
companion/.expo/README.md-11-14 (1)
11-14: 📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick winRemove committed Expo local state.
companion/.expo/README.mdstates that.expois machine-specific and must not be shared. Remove both files from Git. Add an ignore rule forcompanion/.expo/if one does not already exist.
companion/.expo/README.md#L11-L14: remove the committed Expo-generated README.companion/.expo/devices.json#L1-L3: remove the committed development-session state file.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@companion/.expo/README.md` around lines 11 - 14, Remove the committed files companion/.expo/README.md (lines 11-14) and companion/.expo/devices.json (lines 1-3) from Git, and add the companion/.expo/ ignore rule to the relevant .gitignore if it is not already present.src/i18n/it.ts-30-42 (1)
30-42: 📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick winMini Mode feature name uses two capitalizations in the new translations. Both files introduce the feature name in title case in the
miniModekey and in lower case in theminiModeCollapsekey. The shared root cause is a missing capitalization rule for the feature name across the new Mini Mode strings. Pick one form per language and apply it to every Mini Mode string, including any future keys.
src/i18n/it.ts#L30-L42: changeminiModeCollapseto use "Modalità Mini", or changeminiModeto lower case, so both keys agree.src/i18n/pt.ts#L30-L42: changeminiModeCollapseto use "Modo Mini", or changeminiModeto lower case, so both keys agree.Check the remaining locale files in the same way, because they were changed in the same commit.
#!/bin/bash # Compare Mini Mode string capitalization across all locale files. fd -t f -e ts . src/i18n --exec rg -n -A2 'miniMode' {}🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/i18n/it.ts` around lines 30 - 42, Standardize the Mini Mode feature-name capitalization across all new locale strings: in src/i18n/it.ts lines 30-42, make miniMode and miniModeCollapse consistently use “Modalità Mini”; in src/i18n/pt.ts lines 30-42, make them consistently use “Modo Mini”. Review the remaining locale files for the same miniMode and miniModeCollapse inconsistency and apply each language’s chosen capitalization consistently to all Mini Mode keys.src/lib/layoutProfiles.test.ts-25-31 (1)
25-31: 📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick winMake the mouse-layout assertion able to fail.
Line 30 uses
?? 0.28and then expects0.28. The assertion passes whenmouseLayoutis undefined and when it equals0.28. A regression wherepersistLayoutForKind(…, "touch")also writesmouseLayoutwould only be caught if the written value differed from0.28, and here it would not.
basehas nomouseLayout, so assert its absence directly.💚 Proposed fix
const stored = persistLayoutForKind(touchApplied, "touch"); expect(stored.touchLayout?.inputRowRightRatio).toBe(0.4); - expect(stored.mouseLayout?.inputRowRightRatio ?? 0.28).toBe(0.28); + expect(stored.mouseLayout).toBeUndefined();🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/lib/layoutProfiles.test.ts` around lines 25 - 31, Update the mouse-layout assertion in the test around persistLayoutForKind to assert that stored.mouseLayout is absent, since base does not define one. Remove the nullish fallback and preserve the existing touch-layout value assertion.src/stores/appStore.ts-889-919 (1)
889-919: 🚀 Performance & Scalability | 🟡 Minor | ⚡ Quick winDebounce pointer-kind profile switches
The capture-phase
windowpointerdownlistener callshandlePointerInputEventwithout debouncing. Rapid touch-to-mouse alternation can start overlapping persistence and layout-sync operations. Add a short debounce beforesetPointerInputKind.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/stores/appStore.ts` around lines 889 - 919, Debounce pointer-kind changes initiated by handlePointerInputEvent before invoking setPointerInputKind, so rapid alternating pointerdown events coalesce into one update and avoid overlapping persistence/layout synchronization. Keep the existing no-op behavior for the current kind and preserve setPointerInputKind’s persistence and window-sync logic.src/lib/pointerDrag.ts-69-70 (1)
69-70: 🩺 Stability & Availability | 🟡 Minor | ⚡ Quick winMove the
optionsRefwrite out of render.React can replay or discard a render, which can leave
optionsRef.currentpointing to callbacks from an uncommitted render. Update the ref in a layout effect.useEffectEventcannot replace this callback because Effect Events cannot be called from event handlers.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/lib/pointerDrag.ts` around lines 69 - 70, Move the optionsRef.current assignment out of the render path and update it in a layout effect within the pointer-drag hook, ensuring callbacks come from the committed render before event handlers use them. Keep optionsRef initialized with options and retain the existing event-handler access through optionsRef.Source: Linters/SAST tools
src/stores/appStore.ts-82-126 (1)
82-126: 🩺 Stability & Availability | 🟡 Minor | ⚡ Quick winDrain
pendingMiniLayoutSyncintoggleCollapsed'sfinallyblock. During settings transitions,AppShellcan renderCollapsedFabbeforeminiModeActiveis cleared. A Mini Mode sync can queue duringtoggleCollapsed, but thisfinallyblock does not drain it.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/stores/appStore.ts` around lines 82 - 126, Update toggleCollapsed’s finally block to drain pendingMiniLayoutSync after the transition completes, handling any queued Mini Mode layout sync using the existing syncMiniModeWindowLayout flow. Preserve the current cleanup behavior and ensure the queued request is cleared and executed when appropriate.
🧹 Nitpick comments (11)
src-tauri/src/lib.rs (1)
310-336: 📐 Maintainability & Code Quality | 🔵 Trivial | 🏗️ Heavy liftGroup the layout flags into a struct.
apply_window_layoutandanimate_window_layoutnow take nine positional parameters. Six of them areboolorf32. The call sites pass literal sequences such asfalse, false, false, height_ratio, false, false, .... A swapped argument compiles without an error and produces wrong window geometry.Pass a single
LayoutOptionsstruct instead. This also makes thecompute_window_layoutcall site self-documenting.♻️ Sketch of the struct-based signature
#[derive(Clone, Copy)] pub struct LayoutOptions { pub monitor_id: u32, pub collapsed: bool, pub collapsed_dictation: bool, pub collapsed_settings: bool, pub height_ratio: f32, pub mini_mode: bool, pub mini_keyboard_visible: bool, pub mini_keyboard_height_ratio: f32, } impl LayoutOptions { pub fn expanded(monitor_id: u32, height_ratio: f32) -> Self { Self { monitor_id, collapsed: false, collapsed_dictation: false, collapsed_settings: false, height_ratio, mini_mode: false, mini_keyboard_visible: false, mini_keyboard_height_ratio: window::MINI_KEYBOARD_HEIGHT_RATIO, } } } async fn apply_window_layout(app: &tauri::AppHandle, opts: LayoutOptions) -> Result<(), String> { /* … */ }The
cmd_move_window_to_monitorcall site then reads:apply_window_layout(&app, LayoutOptions::expanded(monitor_id, height_ratio)).awaitAlso applies to: 469-480, 515-519
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src-tauri/src/lib.rs` around lines 310 - 336, Introduce a shared LayoutOptions struct containing monitor_id and all layout flags and ratios, plus the expanded constructor shown by the existing conventions. Update apply_window_layout, animate_window_layout, compute_window_layout, and every affected call site to accept or pass LayoutOptions instead of positional parameters, using named fields to make layout intent explicit and preserving current behavior.vitest.config.ts (1)
1-7: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick winRegister Testing Library cleanup. Vitest globals are disabled, and no setup file calls
cleanup. Addcleanup()to the existingafterEachso eachrenderHookunmounts and removes its subscriptions.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@vitest.config.ts` around lines 1 - 7, Update the Vitest setup referenced by the test configuration to import Testing Library’s cleanup and invoke cleanup() from the existing afterEach hook. Ensure the setup runs for every test so renderHook instances unmount and remove subscriptions, while preserving the current Node environment configuration.src-tauri/src/input/focus_target.rs (2)
240-243: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low valueRemove the redundant prefix check.
lower.starts_with("chrome_widgetwin")can never be reached. Every string with that prefix already matcheschrome_.♻️ Proposed simplification
fn is_chromium_class(class: &str) -> bool { let lower = class.to_ascii_lowercase(); - lower.starts_with("chrome_") || lower.starts_with("chrome_widgetwin") + lower.starts_with("chrome_") }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src-tauri/src/input/focus_target.rs` around lines 240 - 243, Remove the redundant lower.starts_with("chrome_widgetwin") condition from is_chromium_class, leaving only the chrome_ prefix check while preserving the existing case-insensitive behavior.
418-441: 🚀 Performance & Scalability | 🔵 Trivial | 🏗️ Heavy liftUse one long-lived worker thread instead of a new thread per burst.
Each burst spawns a thread. The thread initializes COM, creates a
CUIAutomationinstance in thread-local storage, and never callsCoUninitialize. Focus events are frequent, so this repeats thread creation and COM/UIA object creation, and the cached automation object is discarded at every thread exit.A single dedicated worker thread with a condition variable keeps the UIA instance alive and removes the busy/dirty handshake.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src-tauri/src/input/focus_target.rs` around lines 418 - 441, The schedule_reevaluate_input_focus flow should use one long-lived dedicated worker instead of spawning a thread per burst. Move COM initialization and CUIAutomation creation into that worker’s thread-local setup, keep the UIA instance alive while waiting on a condition variable, and replace the REEVAL_DIRTY/REEVAL_WORKER_BUSY handshake with condition-variable signaling so focus events wake the worker for reevaluation.src-tauri/src/input/cursor_highlight.rs (1)
146-153: 🚀 Performance & Scalability | 🔵 Trivial | 💤 Low valueSkip redundant
MoveWindowcalls while idle.The loop calls
position_windowevery 16 ms while the ring is visible, even whenlast_posdid not change. Each call invalidates the window and queues a repaint. Track the last applied position and move only on change.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src-tauri/src/input/cursor_highlight.rs` around lines 146 - 153, Update the visibility loop around position_window to track the last position applied to the window and call position_window only when the current last_pos differs from it. Preserve the existing hide behavior and visible-position handling, while updating the tracked position after each actual move.src-tauri/src/input/mouse.rs (1)
17-34: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low valueRemove the duplicated default state definition.
The manual
Defaultimpl and the static initializer both describe the same initial state. They can diverge when a field is added. TheDefaultimpl also appears unused, because the static uses a const initializer.Derive
Defaultand keep the const initializer, or drop the impl.♻️ Proposed simplification
+#[derive(Default)] struct TrackpadGestureState { active: bool, last_good: Option<(i32, i32)>, } -impl Default for TrackpadGestureState { - fn default() -> Self { - Self { - active: false, - last_good: None, - } - } -} - static TRACKPAD_GESTURE: Mutex<TrackpadGestureState> = Mutex::new(TrackpadGestureState { active: false, last_good: None, });🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src-tauri/src/input/mouse.rs` around lines 17 - 34, Remove the manual Default implementation for TrackpadGestureState and retain the static TRACKPAD_GESTURE initializer, or derive Default if a default constructor is needed elsewhere. Ensure only one authoritative default-state definition remains and preserve the current active=false and last_good=None values.src/components/layout/ResizableSplitPane.tsx (1)
109-122: 📐 Maintainability & Code Quality | 🔵 Trivial | 💤 Low valueSet the drag state only after the hook accepts the gesture.
onSplitterPointerDownwritesdragStateRef.currentbefore it callssplitterDrag.onPointerDown.usePointerDragreturns early whenenabledisfalse, so noonEndruns and the ref keeps a stale value.The splitter renders only when
!sizedPaneCollapsed, so the path is unreachable today. The order still couples this component to the hook's internal guard. Add the same guard here.♻️ Proposed guard
const onSplitterPointerDown = (event: React.PointerEvent<HTMLDivElement>) => { + if (sizedPaneCollapsed) return; const container = containerRef.current; if (!container) return;🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/components/layout/ResizableSplitPane.tsx` around lines 109 - 122, Update onSplitterPointerDown to check the same enabled condition used by splitterDrag before assigning dragStateRef.current, and return without starting the gesture when disabled. Keep the container and width validation, then invoke splitterDrag.onPointerDown before or alongside setting drag state only after the gesture is accepted.src/lib/pointerDrag.test.ts (1)
28-78: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick winAdd coverage for the remaining
usePointerDragbranches.The suite covers the move path and the active-state transition. Three behaviors in the hook are untested:
enabled: falsemust skiponPointerDownand must not set the active state.pointercancelmust end the drag in the same way aspointerup.onEndmust run once with the terminating event.These branches guard drag state that can otherwise leak across gestures.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/lib/pointerDrag.test.ts` around lines 28 - 78, Extend the usePointerDrag tests with coverage for the disabled option, asserting enabled: false skips onPointerDown and leaves pointerDragActive false; add a pointercancel test verifying it ends the drag like pointerup; and verify onEnd is called exactly once with the terminating event. Use the existing renderHook setup, __resetPointerDragActiveForTests, and window event dispatch patterns.src/lib/layoutProfiles.ts (1)
3-33: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick winDerive the snapshot functions from
LAYOUT_PARTIAL_KEYS.The three layout fields are listed three times: in
LAYOUT_PARTIAL_KEYS, insnapshotFromSettings, and inapplyLayoutSnapshot. If a fourth layout field is added later and one list is missed, that field is silently dropped from the touch or mouse profile with no type error.Drive all three from the single key list.
♻️ Proposed refactor
const LAYOUT_PARTIAL_KEYS = [ "sectionStack", "inputRowRightRatio", "windowHeightRatio", ] as const satisfies ReadonlyArray<keyof AppSettings>; export function snapshotFromSettings(settings: AppSettings): LayoutSnapshot { - return { - sectionStack: settings.sectionStack, - inputRowRightRatio: settings.inputRowRightRatio, - windowHeightRatio: settings.windowHeightRatio, - }; + const snap: LayoutSnapshot = {}; + for (const key of LAYOUT_PARTIAL_KEYS) { + if (settings[key] !== undefined) { + Object.assign(snap, { [key]: settings[key] }); + } + } + return snap; } export function applyLayoutSnapshot( settings: AppSettings, snap: LayoutSnapshot, ): AppSettings { - return { - ...settings, - ...(snap.sectionStack !== undefined - ? { sectionStack: snap.sectionStack } - : {}), - ...(snap.inputRowRightRatio !== undefined - ? { inputRowRightRatio: snap.inputRowRightRatio } - : {}), - ...(snap.windowHeightRatio !== undefined - ? { windowHeightRatio: snap.windowHeightRatio } - : {}), - }; + const next = { ...settings }; + for (const key of LAYOUT_PARTIAL_KEYS) { + if (snap[key] !== undefined) { + Object.assign(next, { [key]: snap[key] }); + } + } + return next; }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/lib/layoutProfiles.ts` around lines 3 - 33, Refactor snapshotFromSettings and applyLayoutSnapshot to iterate over LAYOUT_PARTIAL_KEYS instead of explicitly naming each layout field, preserving undefined-snapshot values and the existing AppSettings merge behavior. Ensure adding a key to LAYOUT_PARTIAL_KEYS automatically includes it in both snapshot creation and application.src/lib/layoutProfiles.test.ts (1)
58-83: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick winAdd round-trip coverage for
sectionStack.Every test exercises only
inputRowRightRatio.sectionStackis the largest snapshot field and the only object-valued one, so a shallow-copy or reference-sharing regression between the touch and mouse profiles would not be caught. Add one switch test that stores differentsectionStackvalues per pointer kind and asserts they stay separate.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/lib/layoutProfiles.test.ts` around lines 58 - 83, Extend the switchPointerInputKindLayout test coverage with distinct object-valued sectionStack values for mouse and touch profiles, then assert switching to touch restores its sectionStack and switching back restores the mouse value independently. Ensure the assertions verify value separation rather than only inputRowRightRatio.src-tauri/src/window/mod.rs (1)
283-288: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick winAssert the DPI effect, not a loose lower bound.
collapsed_window_includes_hover_and_dpi_slackpasses withdpi_scale = 1.0as well, so it does not prove that the DPI slack is applied. Compare the result against the same count at scale 1.0.💚 Proposed test strengthening
#[test] fn collapsed_window_includes_hover_and_dpi_slack() { - let (width, height) = compute_collapsed_dimensions(CollapsedFabCount::Two, 1.5); - assert!(width >= 76); - assert!(height >= 76); + let (base_w, base_h) = compute_collapsed_dimensions(CollapsedFabCount::Two, 1.0); + let (width, height) = compute_collapsed_dimensions(CollapsedFabCount::Two, 1.5); + let extra = (FAB_HOVER_SLACK as f32 * 1.5).ceil() as u32 - FAB_HOVER_SLACK; + assert_eq!(width, base_w + extra); + assert_eq!(height, base_h + extra); }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src-tauri/src/window/mod.rs` around lines 283 - 288, Strengthen collapsed_window_includes_hover_and_dpi_slack by computing collapsed dimensions for CollapsedFabCount::Two at both DPI scales 1.5 and 1.0, then assert that the scaled result is larger than the baseline for the relevant dimensions. Retain the existing minimum-size checks if needed, but ensure the test directly verifies the DPI effect.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@package.json`:
- Around line 35-43: Align the Vite and Vitest dependency versions in
package.json by either upgrading vitest to a published release that supports
Vite 8 or pinning vite to a major supported by vitest 3.2.4. Preserve the
existing `@vitejs/plugin-react` compatibility and ensure the resulting dependency
range has no unsupported peer-version mismatch.
In `@src-tauri/src/icons.rs`:
- Around line 26-31: Update the cache-write flow around fs::rename so that a
rename failure re-checks whether cache_file now exists and returns its path when
valid. Only remove the temporary file and return None if the destination is
still unavailable, preserving the existing success path.
In `@src-tauri/src/input/cursor_highlight.rs`:
- Around line 114-157: Add a Windows message pump inside each iteration of
run_loop, after processing commands or before sleeping, by repeatedly calling
PeekMessageW for the overlay thread’s queue and passing retrieved messages
through TranslateMessage and DispatchMessageW. Preserve the existing command
handling, visibility updates, and shutdown behavior while ensuring all queued
window messages are removed and dispatched.
In `@src-tauri/src/input/focus_target.rs`:
- Around line 73-77: Update init so it no longer calls reevaluate_input_focus
directly on the caller thread; instead, schedule the initial evaluation through
the existing deferred worker mechanism while preserving hook installation and
app-handle setup order.
In `@src-tauri/src/lib.rs`:
- Around line 1028-1032: Update cmd_open_windows_settings to validate the
requested ms-settings page against a strict allowlist of supported settings
URIs, rejecting all other values before launching. Replace the cmd.exe
invocation with the registered tauri_plugin_opener or a Windows-only
ShellExecuteW call, preserving the existing error propagation and success
behavior.
In `@src-tauri/src/stt/mod.rs`:
- Around line 67-86: Update start_dictation to handle an existing
router().active backend before starting the selected engine: either reject the
request while active or stop the current backend first. Ensure the previous
backend cannot remain running when active is replaced, and preserve consistent
router state across the WinRt and Groq branches.
In `@src-tauri/src/window/mod.rs`:
- Around line 86-105: Update collapsed_fab_stack_height and
compute_collapsed_dimensions so COLLAPSED_SIZE, COLLAPSED_PAD, and
COLLAPSED_FAB_GAP are all scaled by dpi_scale before calculating the stack width
and height. Preserve the existing minimum scale behavior and continue including
the separately computed hover slack.
In `@src/components/mouse/Trackpad.tsx`:
- Around line 213-217: Update the movement handling around the dx/dy
calculations to accumulate fractional motion in subPixel.current before
rounding, preserving remainders across pointer events instead of discarding
them. Reset subPixel.current to { x: 0, y: 0 } in both onPointerDown and
endGesture so each gesture starts without stale movement.
- Around line 186-199: Update the cmd_trackpad_gesture_begin failure handler in
Trackpad.tsx to call pollError() after resetting dragging.current and
gestureReady.current for the active generation. Preserve the existing
stale-generation behavior and ensure the backend error is surfaced to the user
like the other trackpad actions.
- Around line 242-290: Use the existing gestureGen in endGesture to associate
flushThenEnd and endBackend with the gesture being ended. Capture the generation
before starting teardown, and abort any retry, cursor flush, backend end, or old
handleTap callback when gestureGen.current no longer matches, allowing a new
onPointerDown gesture to own rafId and backend cleanup.
In `@src/components/settings/SettingsPanel.tsx`:
- Around line 465-474: Update the Mini Mode control around the miniModeOverride
setting to expose all three states: null for Auto, true for forced On, and false
for forced Off. Replace the current boolean-only ToggleRow behavior with an
explicit Off option or tri-state control, ensuring user selection writes false
when disabling and preserves null for Auto.
In `@src/lib/miniMode.ts`:
- Around line 85-88: Update isMiniModeEligible to reject an empty monitors
array, returning true only when there is exactly one monitor or
isMirroredSetup(monitors) is true. Add a regression test asserting
isMiniModeEligible([]) returns false.
In `@src/lib/pointerDrag.ts`:
- Around line 92-105: Update src/lib/pointerDrag.ts lines 92-105 in
onPointerDown to return when activeRef.current is already set, preventing
concurrent pointers from orphaning the active gesture. Add unmount cleanup at
src/lib/pointerDrag.ts lines 107-113 to remove the pointermove, pointerup, and
pointercancel listeners, clear activeRef.current, and call
setPointerDragActive(false) when a drag is active.
- Around line 107-113: Add a useEffect cleanup in usePointerDrag that detects an
in-progress drag during unmount and invokes endDrag with the appropriate
event-free handling, ensuring all window listeners are removed and
activeDragCount is decremented. Import useEffect as needed, while preserving the
existing pointer event handlers.
In `@src/lib/toolWindows.ts`:
- Around line 57-71: Update getMonitors so failures from
invoke("cmd_list_monitors") are caught and return an empty MonitorInfo[] instead
of rejecting. Preserve the existing successful query behavior, allowing
openToolWindow and its positioning fallback to continue when no monitors are
available.
In `@src/stores/appStore.ts`:
- Around line 845-861: Update the miniModeSettingChanged condition in the store
update flow to depend only on miniModeOverride, excluding miniModeTransparent.
Preserve the existing refreshMiniModeState behavior for override changes and
leave transparency updates to the CSS path without triggering window layout
animation.
---
Minor comments:
In `@companion/.expo/README.md`:
- Around line 11-14: Remove the committed files companion/.expo/README.md (lines
11-14) and companion/.expo/devices.json (lines 1-3) from Git, and add the
companion/.expo/ ignore rule to the relevant .gitignore if it is not already
present.
In `@src/i18n/it.ts`:
- Around line 30-42: Standardize the Mini Mode feature-name capitalization
across all new locale strings: in src/i18n/it.ts lines 30-42, make miniMode and
miniModeCollapse consistently use “Modalità Mini”; in src/i18n/pt.ts lines
30-42, make them consistently use “Modo Mini”. Review the remaining locale files
for the same miniMode and miniModeCollapse inconsistency and apply each
language’s chosen capitalization consistently to all Mini Mode keys.
In `@src/lib/layoutProfiles.test.ts`:
- Around line 25-31: Update the mouse-layout assertion in the test around
persistLayoutForKind to assert that stored.mouseLayout is absent, since base
does not define one. Remove the nullish fallback and preserve the existing
touch-layout value assertion.
In `@src/lib/pointerDrag.ts`:
- Around line 69-70: Move the optionsRef.current assignment out of the render
path and update it in a layout effect within the pointer-drag hook, ensuring
callbacks come from the committed render before event handlers use them. Keep
optionsRef initialized with options and retain the existing event-handler access
through optionsRef.
In `@src/stores/appStore.ts`:
- Around line 889-919: Debounce pointer-kind changes initiated by
handlePointerInputEvent before invoking setPointerInputKind, so rapid
alternating pointerdown events coalesce into one update and avoid overlapping
persistence/layout synchronization. Keep the existing no-op behavior for the
current kind and preserve setPointerInputKind’s persistence and window-sync
logic.
- Around line 82-126: Update toggleCollapsed’s finally block to drain
pendingMiniLayoutSync after the transition completes, handling any queued Mini
Mode layout sync using the existing syncMiniModeWindowLayout flow. Preserve the
current cleanup behavior and ensure the queued request is cleared and executed
when appropriate.
---
Nitpick comments:
In `@src-tauri/src/input/cursor_highlight.rs`:
- Around line 146-153: Update the visibility loop around position_window to
track the last position applied to the window and call position_window only when
the current last_pos differs from it. Preserve the existing hide behavior and
visible-position handling, while updating the tracked position after each actual
move.
In `@src-tauri/src/input/focus_target.rs`:
- Around line 240-243: Remove the redundant
lower.starts_with("chrome_widgetwin") condition from is_chromium_class, leaving
only the chrome_ prefix check while preserving the existing case-insensitive
behavior.
- Around line 418-441: The schedule_reevaluate_input_focus flow should use one
long-lived dedicated worker instead of spawning a thread per burst. Move COM
initialization and CUIAutomation creation into that worker’s thread-local setup,
keep the UIA instance alive while waiting on a condition variable, and replace
the REEVAL_DIRTY/REEVAL_WORKER_BUSY handshake with condition-variable signaling
so focus events wake the worker for reevaluation.
In `@src-tauri/src/input/mouse.rs`:
- Around line 17-34: Remove the manual Default implementation for
TrackpadGestureState and retain the static TRACKPAD_GESTURE initializer, or
derive Default if a default constructor is needed elsewhere. Ensure only one
authoritative default-state definition remains and preserve the current
active=false and last_good=None values.
In `@src-tauri/src/lib.rs`:
- Around line 310-336: Introduce a shared LayoutOptions struct containing
monitor_id and all layout flags and ratios, plus the expanded constructor shown
by the existing conventions. Update apply_window_layout, animate_window_layout,
compute_window_layout, and every affected call site to accept or pass
LayoutOptions instead of positional parameters, using named fields to make
layout intent explicit and preserving current behavior.
In `@src-tauri/src/window/mod.rs`:
- Around line 283-288: Strengthen collapsed_window_includes_hover_and_dpi_slack
by computing collapsed dimensions for CollapsedFabCount::Two at both DPI scales
1.5 and 1.0, then assert that the scaled result is larger than the baseline for
the relevant dimensions. Retain the existing minimum-size checks if needed, but
ensure the test directly verifies the DPI effect.
In `@src/components/layout/ResizableSplitPane.tsx`:
- Around line 109-122: Update onSplitterPointerDown to check the same enabled
condition used by splitterDrag before assigning dragStateRef.current, and return
without starting the gesture when disabled. Keep the container and width
validation, then invoke splitterDrag.onPointerDown before or alongside setting
drag state only after the gesture is accepted.
In `@src/lib/layoutProfiles.test.ts`:
- Around line 58-83: Extend the switchPointerInputKindLayout test coverage with
distinct object-valued sectionStack values for mouse and touch profiles, then
assert switching to touch restores its sectionStack and switching back restores
the mouse value independently. Ensure the assertions verify value separation
rather than only inputRowRightRatio.
In `@src/lib/layoutProfiles.ts`:
- Around line 3-33: Refactor snapshotFromSettings and applyLayoutSnapshot to
iterate over LAYOUT_PARTIAL_KEYS instead of explicitly naming each layout field,
preserving undefined-snapshot values and the existing AppSettings merge
behavior. Ensure adding a key to LAYOUT_PARTIAL_KEYS automatically includes it
in both snapshot creation and application.
In `@src/lib/pointerDrag.test.ts`:
- Around line 28-78: Extend the usePointerDrag tests with coverage for the
disabled option, asserting enabled: false skips onPointerDown and leaves
pointerDragActive false; add a pointercancel test verifying it ends the drag
like pointerup; and verify onEnd is called exactly once with the terminating
event. Use the existing renderHook setup, __resetPointerDragActiveForTests, and
window event dispatch patterns.
In `@vitest.config.ts`:
- Around line 1-7: Update the Vitest setup referenced by the test configuration
to import Testing Library’s cleanup and invoke cleanup() from the existing
afterEach hook. Ensure the setup runs for every test so renderHook instances
unmount and remove subscriptions, while preserving the current Node environment
configuration.
🪄 Autofix
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 Plus
Run ID: 5409213e-5508-4c51-a944-ab6452dc83c9
⛔ Files ignored due to path filters (1)
package-lock.jsonis excluded by!**/package-lock.json
📒 Files selected for processing (63)
.cursor/mcp.json.cursor/rules/windows-only.mdc.github/scripts/generate-latest-json.mjs.github/workflows/build.yml.github/workflows/release.yml.gitignoreCONTRIBUTING.mdREADME.mdcompanion/.expo/README.mdcompanion/.expo/devices.jsondocs/accessibility-requirements.mddocs/superpowers/plans/2026-08-11-mini-mode-layout-fixes.mdpackage.jsonsrc-tauri/src/icons.rssrc-tauri/src/input/cursor_highlight.rssrc-tauri/src/input/focus_target.rssrc-tauri/src/input/mod.rssrc-tauri/src/input/mouse.rssrc-tauri/src/input/stub.rssrc-tauri/src/installed_apps.rssrc-tauri/src/lib.rssrc-tauri/src/stt/mod.rssrc-tauri/src/stt/network.rssrc-tauri/src/stt/stub.rssrc-tauri/src/tts/mod.rssrc-tauri/src/tts/stub.rssrc-tauri/src/window/mod.rssrc-tauri/src/window/stub.rssrc-tauri/src/window/windows.rssrc-tauri/tauri.conf.jsonsrc/App.tsxsrc/components/common/ModeToggle.tsxsrc/components/common/SectionIcons.tsxsrc/components/common/SuggestionsBar.tsxsrc/components/keyboard/DictationButton.tsxsrc/components/keyboard/KeyButton.tsxsrc/components/keyboard/Keyboard.tsxsrc/components/keyboard/KeyboardSection.tsxsrc/components/layout/AppShell.tsxsrc/components/layout/CollapsedFab.tsxsrc/components/layout/FloatingSection.tsxsrc/components/layout/MiniModeShell.tsxsrc/components/layout/ResizableSplitPane.tsxsrc/components/layout/SectionCanvas.tsxsrc/components/mouse/Trackpad.tsxsrc/components/settings/SettingsPanel.tsxsrc/i18n/de.tssrc/i18n/el.tssrc/i18n/en.tssrc/i18n/es.tssrc/i18n/fr.tssrc/i18n/it.tssrc/i18n/pt.tssrc/lib/layoutProfiles.test.tssrc/lib/layoutProfiles.tssrc/lib/miniMode.test.tssrc/lib/miniMode.tssrc/lib/pointerDrag.test.tssrc/lib/pointerDrag.tssrc/lib/toolWindows.tssrc/lib/types.tssrc/stores/appStore.tsvitest.config.ts
💤 Files with no reviewable changes (6)
- src-tauri/src/tts/stub.rs
- src-tauri/src/window/stub.rs
- .github/scripts/generate-latest-json.mjs
- src-tauri/src/stt/stub.rs
- src-tauri/src/input/stub.rs
- src-tauri/src/tts/mod.rs
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
fix: address PR review findings across Mini Mode input and tooling
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@src-tauri/src/db/mod.rs`:
- Around line 1050-1067: Close or otherwise release the database connection
owned by db before calling std::fs::remove_dir_all(dir) in this test cleanup
block. Ensure all SQLite handles, including the connection guarded by db.conn,
are dropped before removal, and handle the cleanup result according to the
existing test conventions instead of ignoring a Windows failure.
🪄 Autofix
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 Plus
Run ID: 2dad0ace-94aa-4e30-8173-1dd19c8b60ed
⛔ Files ignored due to path filters (1)
package-lock.jsonis excluded by!**/package-lock.json
📒 Files selected for processing (31)
.gitignorepackage.jsonsrc-tauri/src/db/mod.rssrc-tauri/src/icons.rssrc-tauri/src/input/cursor_highlight.rssrc-tauri/src/input/focus_target.rssrc-tauri/src/input/mouse.rssrc-tauri/src/lib.rssrc-tauri/src/stt/mod.rssrc-tauri/src/window/mod.rssrc/App.tsxsrc/components/layout/ResizableSplitPane.tsxsrc/components/mouse/Trackpad.tsxsrc/components/settings/SettingsPanel.tsxsrc/i18n/de.tssrc/i18n/el.tssrc/i18n/en.tssrc/i18n/es.tssrc/i18n/fr.tssrc/i18n/it.tssrc/i18n/pt.tssrc/lib/layoutProfiles.test.tssrc/lib/layoutProfiles.tssrc/lib/miniMode.test.tssrc/lib/miniMode.tssrc/lib/pointerDrag.test.tssrc/lib/pointerDrag.tssrc/lib/toolWindows.tssrc/stores/appStore.tsvitest.config.tsvitest.setup.ts
💤 Files with no reviewable changes (1)
- src-tauri/src/input/mouse.rs
🚧 Files skipped from review as they are similar to previous changes (23)
- src/i18n/en.ts
- vitest.config.ts
- src/i18n/el.ts
- src/i18n/es.ts
- src/i18n/fr.ts
- src/App.tsx
- .gitignore
- src/lib/layoutProfiles.test.ts
- src/i18n/pt.ts
- src/components/layout/ResizableSplitPane.tsx
- src/i18n/de.ts
- src/lib/layoutProfiles.ts
- src/lib/toolWindows.ts
- src/components/settings/SettingsPanel.tsx
- src-tauri/src/icons.rs
- src/i18n/it.ts
- src-tauri/src/stt/mod.rs
- src/lib/pointerDrag.ts
- src/components/mouse/Trackpad.tsx
- src/lib/miniMode.ts
- package.json
- src-tauri/src/input/focus_target.rs
- src-tauri/src/window/mod.rs
Comment on lines
+1050
to
+1067
| let conn = db.conn.lock().expect("lock"); | ||
| let cats: i64 = conn | ||
| .query_row( | ||
| "SELECT COUNT(*) FROM phrase_categories WHERE profile_id = 'orphan'", | ||
| [], | ||
| |r| r.get(0), | ||
| ) | ||
| .expect("count categories"); | ||
| assert!(cats >= 2, "expected Basic Needs + Emergency categories"); | ||
| let phrases: i64 = conn | ||
| .query_row( | ||
| "SELECT COUNT(*) FROM phrases WHERE profile_id = 'orphan' AND language = 'el'", | ||
| [], | ||
| |r| r.get(0), | ||
| ) | ||
| .expect("count phrases"); | ||
| assert!(phrases > 0, "expected locale phrases to be seeded"); | ||
| let _ = std::fs::remove_dir_all(dir); |
There was a problem hiding this comment.
🩺 Stability & Availability | 🟡 Minor | ⚡ Quick win
Close the database before removing the temporary directory.
db still owns the SQLite connection at Line 1067. Windows keeps reach-panel.db locked, so remove_dir_all(dir) fails. The ignored result leaves a temporary directory after each test run.
Proposed fix
assert!(phrases > 0, "expected locale phrases to be seeded");
- let _ = std::fs::remove_dir_all(dir);
+ drop(conn);
+ drop(db);
+ std::fs::remove_dir_all(dir).expect("remove temp db directory");As per coding guidelines, ReachPanel targets Windows only.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| let conn = db.conn.lock().expect("lock"); | |
| let cats: i64 = conn | |
| .query_row( | |
| "SELECT COUNT(*) FROM phrase_categories WHERE profile_id = 'orphan'", | |
| [], | |
| |r| r.get(0), | |
| ) | |
| .expect("count categories"); | |
| assert!(cats >= 2, "expected Basic Needs + Emergency categories"); | |
| let phrases: i64 = conn | |
| .query_row( | |
| "SELECT COUNT(*) FROM phrases WHERE profile_id = 'orphan' AND language = 'el'", | |
| [], | |
| |r| r.get(0), | |
| ) | |
| .expect("count phrases"); | |
| assert!(phrases > 0, "expected locale phrases to be seeded"); | |
| let _ = std::fs::remove_dir_all(dir); | |
| let conn = db.conn.lock().expect("lock"); | |
| let cats: i64 = conn | |
| .query_row( | |
| "SELECT COUNT(*) FROM phrase_categories WHERE profile_id = 'orphan'", | |
| [], | |
| |r| r.get(0), | |
| ) | |
| .expect("count categories"); | |
| assert!(cats >= 2, "expected Basic Needs + Emergency categories"); | |
| let phrases: i64 = conn | |
| .query_row( | |
| "SELECT COUNT(*) FROM phrases WHERE profile_id = 'orphan' AND language = 'el'", | |
| [], | |
| |r| r.get(0), | |
| ) | |
| .expect("count phrases"); | |
| assert!(phrases > 0, "expected locale phrases to be seeded"); | |
| drop(conn); | |
| drop(db); | |
| std::fs::remove_dir_all(dir).expect("remove temp db directory"); |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@src-tauri/src/db/mod.rs` around lines 1050 - 1067, Close or otherwise release
the database connection owned by db before calling std::fs::remove_dir_all(dir)
in this test cleanup block. Ensure all SQLite handles, including the connection
guarded by db.conn, are dropped before removal, and handle the cleanup result
according to the existing test conventions instead of ignoring a Windows
failure.
Sources: Coding guidelines, Learnings
…icts. Keep Mini Mode toolbar and predictions off/enable UX; apply theme-aware suggestion chip colors from main. Co-authored-by: Cursor <cursoragent@cursor.com>
This was referenced Aug 12, 2026
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.
Summary
Promote
devtomain, including:devCommit message
feat: promote mini mode and Windows-only layout fixes to mainRelated issue
N/A
Accessibility impact
Mini Mode makes the keyboard appear when typing in other apps and collapse to large FABs when not — better for single-monitor / mirrored accessibility setups. Focus detection now covers modern Windows text fields (Search, browsers, WinUI), not only classic Edit controls.
Test plan
npm run tauri devlocally on Windowsnsis/msi) still look correctScreenshots (if UI changed)
See mini-mode work in #101 if needed; this PR is the
dev→mainpromotion.Summary by CodeRabbit