* fix(tasks): make detail-panel add-subtask work in the Planner (#8617)
The detail panel and context menu delegated "Add subtask" to
AddSubtaskInputService.requestOpen(), a signal consumed only by the
<task> row that renders the parent. The Planner renders tasks as
<planner-task>, so the request was dropped and nothing happened
(regression from #8423; Today view still worked).
- task-detail-panel now hosts its own inline <add-subtask-input>
(works in every view; also fixes the input opening behind the
bottom panel on mobile). It controls the sub-task section's
expansion via a signal and focuses the input after the expand
animation completes; animates in/out with [@expandFade].
- task-detail-item gains expandedChange/afterExpand outputs so the
panel can control/observe the Material expansion panel.
- context menu addSubTask() reverts to direct addSubTaskTo() (a
transient menu has no place to host the inline draft).
Adds a Planner e2e repro and updates the affected unit specs.
* fix(tasks): address multi-review findings for #8617 add-subtask
- Focus the inline input via a deferred timeout in onSubTasksAfterExpand:
with animations disabled Material fires afterExpand synchronously within
the same CD pass, before the addSubtaskInput viewChild is committed, so
the first collapsed→expand "Add subtask" click left the input unfocused.
- Reset isSubTasksExpanded on task switch so the sub-task section doesn't
stay sticky-expanded across tasks (the panel instance is reused).
- Return focus to the "Add subtask" button when the draft is closed via
Escape, instead of dropping focus to <body>.
- Refresh the now-stale AddSubtaskInputService doc comment.
- Assert the draft input is focused in the Planner e2e (the focus path was
previously uncovered).
* fix(tasks): keep context-menu add-subtask on the inline-draft bus
The earlier context-menu change to addSubTaskTo() was unnecessary: the
context menu's "Add sub-task" entry is gated behind isAdvancedControls,
which only the <task> row enables. planner-task and schedule-event leave
it false, so the entry is hidden there — meaning the menu action is only
ever reachable from a rendered <task> row, where requestOpen() works.
Reverting restores the v18.12 inline-draft UX for that path and shrinks
the diff. (#8617 was only ever reachable via the detail panel, which the
self-hosted input fix already covers.)
* test(tasks): cover add-subtask focus with animations disabled
Guards the deferred-focus fix: with animations disabled Material fires
the expansion panel's afterExpand synchronously, before the panel's
add-subtask-input viewChild is committed. Verified this test fails
without the setTimeout deferral and passes with it.
* fix(tasks): keep subtask input open after add
* fix(tasks): keep subtask input open after add
* perf(tasks): avoid task row effect reactivity
* fix(tasks): address subtask input review
* feat(tasks): ease in the inline subtask input instead of popping
* fix(tasks): consume inline subtask-input request to prevent stale focus-steal
Address multi-review findings on the inline subtask input:
- The open request was held in a signal that was never cleared, so a task row re-created with the same id (e.g. navigating away from a project and back) re-ran its open effect on init and re-opened the input, stealing focus with no user action. Reset via consume() once the row acts on it.
- Drop the redundant requestId counter (a fresh request value already re-fires the effect); request payload is now just the parentId string.
- Add an aria-label to the input (placeholder is not an accessible name).
- _commit() returns void (its boolean result was unused).
* feat(tasks): animate the inline subtask input out and tighten its top gap
- Use the expandFade animation (enter + leave) so the input collapses out instead of vanishing. Caveat: when the input is the only thing in .sub-tasks (adding the first subtask, then cancelling without creating any), the enclosing @if collapses in the same tick and Angular skips the child leave animation, so it's instant in that one case.
- Reduce the input's top margin from --s-half (4px) to --s-quarter (2px).
* feat(tasks): refocus the task row when the subtask draft is cancelled with Escape
The inline subtask input's 'closed' output now reports why it closed ('escape' vs 'blur'). On Escape (a keyboard cancel) the host task calls focusSelf() so keyboard navigation continues from that row; on blur it doesn't, since focus already moved elsewhere. focusSelf() is a no-op on touch.
Covered by unit tests (reason emitted, host refocus only on escape) and an e2e assertion that the task row is focused after Escape.
* feat(tasks): return focus to the task the subtask draft was opened from
Escape previously refocused the parent row that hosts the input. Capture the originating task (which may be a subtask) when opening the draft — before focus moves into the input and the parent row claims focusedTaskId — and restore that on Escape. Falls back to the host row when no origin was captured; no-op on touch.
Focus-by-id prefers the last #t-<id> instance (the detail side-panel one) to match the existing inline-edit focus convention when a task renders twice.
Covered by unit tests and an e2e proving focus returns to the originating subtask, not its parent.
* test(tasks): update subtask e2e for the inline draft input flow
The add-subtask UX changed from 'press a -> empty subtask title focused for edit' to an inline draft input (type + Enter to commit, stays open for rapid entry). Update every e2e site that added subtasks via the old textarea selector:
- WorkViewPage.addSubTask helper: wait for .e2e-add-subtask-input, fill + Enter, then Escape to close the draft for a clean post-state (fixes simple-subtask, add-to-today, drag-task-into-subtask, finish-day-quick-history, and the sync specs that use the helper).
- add-subtask-with-detail-panel-open.spec.ts: assert the draft input opens focused from panel/main-list focus; rewrite the multi-subtask case to repeated Enter (the new rapid-entry path).
- supersync-archive-subtasks + supersync-lww-conflict: same inline-draft flow (not runnable in-sandbox; fixed by analogy).
Verified locally: detail-panel-open 3/3, simple-subtask, add-to-today 5/5, drag-into-subtask, finish-day-quick-history all green.
---------
Co-authored-by: Johannes Millan <johannes.millan@gmail.com>
- schedule-locale-ng0701.spec: mock CalendarEventActionsService so the
real DI chain (IssueService -> SnackService -> LOCAL_ACTIONS -> Actions)
is not constructed. Its missing ScannedActionsSubject provider threw in
afterAll (NG0201) and aborted the entire Karma run at ~2951/10838.
- add-subtask-with-detail-panel-open.spec: wait for the panel's deferred
auto-focus to land before re-focusing the main-list row, and re-apply
focus on each retry via toPass() to fix a focus-race flake under load.
- jira-api.service.spec: pin navigator.onLine=true for the auto-import
pagination tests so they assert pagination, not the runner's network
state (offline CI tripped the "Jira Offline" guard in _sendRequest$).
* build: drop unused elevate.exe from Windows build #8135
elevate.exe is bundled by electron-builder's NSIS target solely so
electron-updater can apply privileged per-machine updates. We ship no
in-app auto-updater (electron-updater is not a dependency; the autoUpdater
block in electron/start-app.ts is commented out), so the binary is unused
dead weight and a frequent AV false-positive. Set packElevateHelper: false
to stop bundling it; safe because perMachine is not true.
* fix(tasks): add subtask via "a" shortcut when detail panel is open
With the detail panel open, pressing "a" created a sub-task but did not
focus it for editing.
Two paths were broken:
- Focus inside the panel: the panel auto-focuses a detail item on open, so
focus is not on a <task> row and the global task-shortcut handler drops
the shortcut. The panel now handles taskAddSubTask in its keydown listener
(stopPropagation avoids a double-add when focus is on an in-panel row).
- Focus on the main-list task row: addSubTaskTo -> focusTaskById targeted the
last #t-<id> copy, which is the in-panel copy inside the collapsed (hidden)
sub-task section and cannot take focus. focusTaskById now falls back to the
last focusable copy, so the new sub-task title is focused for editing
(the panel follows focus to the new sub-task).
* feat(history): merge Quick History and Worklog into one view
- single /history route; legacy worklog & quick-history redirect to it
- one History menu entry; navigate-to-task targets history
- show per-day work start-end as its own column in the day table
- show enabled habits/simple-counters (icon + tooltip) in the expanded day
- extract HistoryTaskRowComponent for archived task rows
- remove WorklogComponent and QuickHistoryComponent
* fix(history): restore quick history behavior
* refactor(history): collapse to a single unified view
- remove the Worklog/Quick History toggle; show one full all-time
history tree (year -> month -> week -> day)
- drop the quick-history mode, significance filter and ?view= param
- extract shared HistoryDayMetaComponent; reuse HistoryTaskRowComponent
in the Daily Summary "This Week" widget
- convert project colors + dateStr auto-expand to signals
- update e2e specs to the single-view selectors
* refactor(history): remove dead quick-history code and i18n keys
After merging Quick History into the unified History view, the
quick-history data pipeline had no consumers. Remove it:
- _quickHistoryData$, quickHistoryWeeks$, _loadQuickHistoryForWorkContext
- map-archive-to-worklog-weeks util (+ tz spec) and WorklogYearsWithWeeks
- orphaned F.QUICK_HISTORY, MH.QUICK_HISTORY, MH.WORKLOG translation keys
Keep WorklogWeekSimple (base of the still-used WorklogWeek).
* docs(history): reflect Quick History merge into unified History
The quick-history and worklog routes are now legacy aliases for the
single unified History view; drop the stale 'current-year mode/toggle'
framing. Concepts-note prose flagged for human review.
* feat(history): improve a11y and i18n of the unified history view
Accessibility:
- week table uses <thead> + <th scope=col>; icon-only work-times
header gets a translated aria-label
- expandable day row: real <button> disclosure control carrying
aria-expanded/aria-controls (keyboard + AT path) instead of a
role=button <tr>; row keeps a pointer click
- archived-task title is now a <button> (was a non-focusable <span>)
i18n: translate previously hardcoded aria-labels (export/restore) and
the week-range tooltip; reuse VIEW_TASK_DETAILS/RESTORE_TASK_FROM_ARCHIVE,
add EXPORT_DATA + WEEK_RANGE_TOOLTIP keys.
Polish (carried in): convert worklogData$ to a toSignal signal (drop
AsyncPipe + dead animations), relocate the shared row to
features/worklog/worklog-task-row (WorklogTaskRowComponent) to break a
worklog<->history cycle, make template-unused services private, tighten
projectColor input type. SCSS focus rings use focus-ring tokens.
e2e: locate task titles via td.title button after the span->button change.
* feat(tasks): allow dragging tasks into subtask lists
* fix(tasks): support childless subtask drop targets
* fix(tasks): support dragging subtasks back to main list
* fix(tasks): preserve nested subtask sorting
* fix(tasks): address subtask drag review feedback
* fix(tasks): address drag conversion review
* test: skip onboarding in migration fresh-start e2e
* fix(tasks): harden convertToSubTask guards and dedupe tag cleanup
Address review findings on the drag-to-subtask feature (#7905):
- Keep the section and crud meta-reducer guards in lock-step via a shared
canApplyConvertToSubTask(). Previously the section reducer stripped a task
from its section even when the crud reducer rejected the convert (missing
target parent or self-target), leaving the task top-level yet dropped from
section ordering on a replayed/concurrent op.
- Reject nesting under a target that is itself a subtask. The UI renders only
two levels, so deeper nesting would orphan the task and leave the
grandparent's time aggregation stale.
- Tighten op-log payload validation to require string taskId/targetParentId.
- Dedupe the "remove task ids from all tags" logic shared by
convertToSubTask/deleteTask/deleteTasks into removeTasksFromAllTags().
- Collapse the tri-state afterTaskId positioning in handleConvertToMainTask
(undefined and null both prepend via moveItemAfterAnchor).
- Name the DragPointer type in DropListService.
Adds reducer specs for the rejected-target-parent cases.
* fix(tasks): remove empty subtask drop-target for childless parents
The dashed empty sub-task drop zone that appeared under childless parent
tasks during a drag was unwanted UI. Remove it along with its supporting
machinery (the subTaskDropCandidate signal in ScheduleExternalDragService,
the pointerdown candidate-arming in TaskListComponent, the
isEmptySubTaskDropTargetMounted computed, and the related template/SCSS).
Consequence: a task can now be nested by dragging only onto a parent that
already has a subtask list. Dragging a subtask back out to a main list
(convertToMainTask) and nesting into an existing subtask list
(convertToSubTask) are unchanged.
* fix(tasks): reliably convert subtasks dragged to the top-level list
Two issues prevented dragging a subtask out to the top-level list to
convert it into a main task:
- CDK only caches a sibling drop-list's geometry when its enterPredicate
passes at drag start (_startReceiving). The pointer is always over the
source subtask list then, so the top-level list was never cached and
conversion silently failed until an unrelated parent drag warmed it.
Open a one-microtask accept window at subtask drag start so CDK caches
the top-level lists' geometry; the pointer guard resumes afterwards.
- An expanded neighbour's subtask list "caught" the drag in the dead-band
just above the next parent (sibling order resolves subtask lists before
the top-level list), silently re-parenting the subtask instead of
converting it, and growing/sticking once entered. Treat only an actual
subtask row as "inside" a foreign subtask list; its trailing padding now
falls through to the top-level list for conversion. The source list
still blocks anywhere, so in-list sorting is unaffected.
* fix(tasks): use midpoint-crossing sort for drag preview placement
CDK's SingleAxisSortStrategy swaps siblings as soon as the cursor enters
any part of their clientRect. For task lists whose item rects span the
parent's full element (header + its expanded subtask list), that swap
displaces the drop preview far below the cursor — landing it past the
target's subtasks instead of where the user is pointing.
Monkey-patch the strategy's `_getItemIndexFromPointerPosition` with a
relative-position midpoint rule (the pattern used by dnd-kit and
react-beautiful-dnd): a sibling only swaps with the dragged item once
the cursor crosses into the half of the sibling that's on the dragged
item's approach side. `enter()` keeps CDK's first-inside semantics so
initial placement still lands somewhere sensible.
Also extend `_pointerSubTaskList()` with a `.sub-tasks` wrapper fallback
so the leading strip between a parent's header and its first subtask
routes the drag into that SUB list at index 0 instead of falling through
to the top-level list and converting at the parent's slot.
* test(tasks): align drag predicate spec data
* test(tasks): reconcile createMockDrop calls after combining drag PRs
Combining the two #7905 PR lineages crossed a stale signature:
the midpoint-sort commit added enterPredicate cases calling the old
3-arg createMockDrop(modelId, filteredTasks, listId), while keilogic's
spec-alignment commit narrowed it to (modelId, listId) since
enterPredicate never reads filteredTasks. Drop the vestigial arrays
from the three affected cases so they match the simplified helper.
* fix(tasks): widen first/last subtask drop targets
Dragging a task to the first/last position of a subtask list was hard:
the midpoint sort patch gives each slot a half-row trigger, and the
two end slots have no neighbouring row to borrow the other half from,
so they were only half of the edge row. The existing "easier dragging"
padding lived on :host, outside the `.task-list-inner` cdkDropList, so
it never grew the CDK hit-rect.
Move that padding inside the sub-task drop list so the strip above the
first and below the last subtask is part of the drop rect. CDK's
enter() then drops a task arriving there as the first/last child
(SingleAxisSortStrategy._shouldEnterAsFirstChild). The matching :host
padding is dropped for sub-lists so the visible gap stays compact.
* refactor(tasks): apply multi-review follow-ups to subtask drag
Hardening and fixes surfaced by the multi-agent review of the
drag-into-subtask feature:
- Scope the CDK midpoint sort patch to VERTICAL lists. It mutates the
shared SingleAxisSortStrategy prototype app-wide, so horizontal lists
(boards, issue panel) were silently swept into the new hit-test;
gating to vertical keeps them on CDK's stock behaviour and sidesteps
the right-to-left index-inversion question. Update the spec to assert
the horizontal fall-back.
- Keep first-child re-parent in a sub-list's leading pad. The drop-target
CSS fix moved the SUB list's top padding inside `.task-list-inner`, so
part of the header→first-row strip now lives in the drop rect; report
the leading pad as a row (re-parent) while the trailing pad stays a
convert-to-main dead-band. Doc comment updated to match.
- Memoise the per-pointer subtask-list hit-test in DropListService so the
several enterPredicate calls CDK fires per pointer move reuse one
document.elementFromPoint.
- Tighten convertToSubTask op-log validation to isValidEntityId (rejects
'' / 'undefined' / 'null') instead of a bare typeof string check.
- Log ids, not full task objects, in the drop handler (titles/notes must
not reach the exportable log).
- Drop the dead _resetMidpointSortPatchForTests export; note why the
seeded drag pointer is inert after a plain tap.
- Add direct unit tests for the canConvert/canApply guard pair.
* test(tasks): cover drop()-to-convert dispatch and the drag e2e
Close the two coverage gaps flagged in review:
- Unit: drive the public drop() handler with CdkDragDrop events and
assert the resulting convertToSubTask / convertToMainTask dispatch,
including the newIds→afterTaskId placement math (after-anchor,
first-slot null anchor) and the isDone flag for the DONE list. This
was previously untested — only _move() and the reducer were.
- E2E: a real CDK drag of a top-level task onto a subtask row, asserting
it converts to a subtask. Uses the manual stepped-mouse gesture (CDK
ignores HTML5 dragTo), matching work-view/sections.spec.ts. Verified
green 3× in a row.
* fix(tasks): cut drop snap-back flicker by emitting default list sync
On drop, CDK tears down its preview/placeholder but never moves the real
DOM node (the list is NgRx-driven), so the un-moved task is briefly
revealed at its old slot until the store re-render lands. customizeUndone
Tasks() pushed *every* work-view list emission through
observeOn(animationFrameScheduler), adding a guaranteed extra frame to
that re-render and widening the visible snap-back.
Keep the frame-defer only for the customized (sort/group/filter) path —
it does heavier work and is driven by CD-bound signals, where a sync emit
can re-enter change detection. The default path is store-driven only, so
emit it on the same tick: the list re-renders without the extra frame and
the dropped task lands in place with far less (ideally no) snap-back.
task-view-customizer spec (43) green; drag-into-subtask e2e green.
* fix(tasks): scope CDK midpoint sort patch to task-list instances
Patch the dragged list's own SingleAxisSortStrategy instance instead of
the shared prototype. registerDropList is only called by task lists, so
the prototype mutation silently changed the hit-test of every other
vertical CDK list (planner, notes, boards, tree-dnd). Shadowing the
method per instance keeps the midpoint-crossing rule scoped to task
lists and leaves all other lists on CDK's stock behaviour.
The CDK-rename guard and 'this' binding are unchanged; the global
isPatched flag is dropped since each instance now patches independently.
* fix(sync): validate afterTaskId in convertToSubTask payload
The convertToSubTask op-log validation branch accepted any payload with
valid taskId/targetParentId, ignoring afterTaskId. Require it to be a
string or null (the action type's contract) so a crafted/malformed sync
payload is rejected at the validation boundary rather than treated as a
not-found anchor downstream.
* docs(tasks): clarify customizer frame-defer and drag-to-done intent
Comment-only. Correct the task-view-customizer note: the customized
path stays on the animation-frame scheduler to batch the signal-driven
emission burst on context switch (commit fddedf3fa6), not because a sync
emit re-enters change detection (the consumer is toSignal). Document
that dragging a subtask onto the DONE list intentionally converts it to
a main task done today, even outside the Today context.
* fix(tasks): keep midpoint sort patch alive across CDK strategy recreation
The previous commit scoped the patch by shadowing the method on the
strategy *instance*. That regressed subtask drag 100% of the time: CDK
rebuilds the SingleAxisSortStrategy on every drag start
(DropListRef.withOrientation runs in its beforeStarted hook), so the
instance shadow was discarded before the first pointer move and task
lists fell back to CDK's stock first-inside hit-test — the exact tall-
item preview misplacement the patch exists to prevent.
Patch the shared prototype instead (survives recreation) but apply the
midpoint rule only when the strategy's container is a registered task
list, delegating to CDK's original otherwise. This keeps the scoping
goal (planner/notes/boards/tree-dnd stay on stock behaviour) without the
recreation fragility. Extract the scoping dispatcher as a pure function
and unit-test it.
---------
Co-authored-by: kei <keletrh@gmail.com>
The done-toggle was refactored from an inline SVG with class="done-toggle"
into a standalone Angular component <done-toggle>. Update all e2e selectors
from '.done-toggle' (class) to 'done-toggle' (element) to match.
- Replace drag handle with SVG done-toggle circle with checkmark draw-on animation
- Improve mobile UX: cdkDragStartDelay for touch, bottom-sheet context menu
- Move issue/repeat indicators from drag handle to tag-list as chips
- Add isFinishDayEnabled config toggle
- Reorder context menu: add sub task before duplicate, deadline as own entry
- Fix context menu not opening on swipe left
- Fix race condition in done toggle animation timeout
Update test expectations in map-archive-to-worklog.spec.ts and E2E tests
to match the new alphabetical sorting order for worklog entries.
The sorting feature sorts tasks alphabetically while keeping subtasks
grouped with their parent tasks. This changes the expected order in tests:
- Parent tasks now sorted alphabetically (MT1 before PT1)
- Subtasks sorted alphabetically within their group (SUB_B before SUB_C)
- Remove all console.log statements from E2E test files
- Replace console.error with throw new Error for proper error handling
- Remove console.warn statements
- Fix unused variable linting errors in plugin tests
