[snuziale]

Summary

Hardens the apollo-wind DataTable (and EditableCell). An audit found 30+ defects, several of which silently showed users wrong data. This PR fixes them, closes the accessibility gaps, and adds the props consumers were working around. The public API is additive only.

Demo

Side-by-side recording, main on the left and this branch on the right, same props on both sides (11 scenes, 92 s).

data-table-before-after.mp4

Before and after

Scenario	Before (main)	After (this PR)
showPagination={false} with 25 rows	Only the first 10 rows render, the rest are unreachable	All 25 render
Sort a text column on a table of 10 rows or fewer	Case-sensitive: Apple, Date, Elderberry, apple pie, banana	Case-insensitive, natural number order
A row is removed while another is selected	Selection is tied to position and slides onto the next record	With getRowId, selection stays on the record
Edit a cell on page 2	Table jumps back to page 1, focus falls to <body>	Same page, focus returns to the cell
Parent re-renders while a cell is being edited	Editor closes and the text is lost	Edit survives
Clear a number cell and commit	NaN is committed	Previous value kept (or null with meta.nullable), min/max clamped
Unparseable date in an editable date column	The table throws RangeError	Value shown as is
globalFilterFn when the first row has null or object values	Typing filters nothing	Predicate runs once per row
searchKey on a number column	Nothing matches	Case-insensitive "contains" match
Resize a column from the keyboard	Handle takes focus but ignores keys	Arrows, Shift+arrows, Home/End, Enter or double-click to reset
Date-only string such as 2024-12-15	Shown a day early west of Greenwich	Shown as that calendar day

Other fixes

• Sorted headers expose aria-sort.
• Resize handles are named per column, have an orientation and sane value bounds.
• Row checkboxes and editable cells have distinct accessible names.
• No max-width: 9007199254740991px inline style on every cell, and an explicit size: 150 is honoured.
• maxBodyHeight: header and body stay aligned with classic (non-overlay) scrollbars.
• The toolbar and pager wrap on narrow screens instead of overflowing.
• pageSize prop changes after mount are respected. Empty data reads "Page 1 of 1".
• The column menu shows header text instead of raw column ids.
• DataTableColumnHeader supports Shift+click multi-sort and renders plain text for non-sortable columns.
• The date editor opens on the month of the current value.
• Four em dashes removed from story copy.

New props

• className, getRowId, loading, emptyState, onRowClick
• strings (Partial<DataTableStrings>): every built-in string, following the existing strings convention in this package
• pageSizeOptions, pagination, onPaginationChange, manualPagination, rowCount
• searchValue, onSearchChange, showSelectionCount
• Column meta.label and meta.nullable

New exports: DataTableStrings, DEFAULT_DATA_TABLE_STRINGS. Three new stories: "Editing with paging and selection", "Loading, empty state and row click", "Localization".

Behaviour changes to review

No prop was removed or retyped, but these are visible to existing consumers:

Change	Who it affects
Accessible names: "Select row" is now "Select row 1", "Edit name" is now "Edit Name, row 1", "Toggle" is now "Active, row 1", "Resize column" is now "Resize Name column"	Tests or automation that query the old names
"n of m row(s) selected" is hidden until there is a select column, controlled selection, or a selected row (showSelectionCount overrides)	Tables showing the count with nothing selectable
Column menu shows header text instead of the raw id	Tests reading menu items
showPagination={false} renders all rows	Anyone relying on the truncation to 10
Text sort order on small tables	Row order
The page no longer resets to 1 when data changes	Consumers who swap in a whole new dataset and expected a reset
The select column is 48px wide and not resizable (was 150px)	Resizable tables are 102px narrower
With maxBodyHeight and classic scrollbars, the scrollbar gutter is always reserved	A blank strip when rows do not overflow

The commit is typed feat without a BREAKING CHANGE footer. If the accessible-name changes should count as breaking for semantic-release, that footer needs adding before merge.

Regression and performance checks

• The original, unmodified test files run against the new code: 51 of 54 pass. The 3 failures are the first three rows of the table above.
• Old and new were rendered side by side across 28 feature combinations and diffed element by element. Row content and order are identical in all 28.
• No other package in the monorepo imports this DataTable.

Old vs new on identical data, interleaved runs, median, in Chromium against the Storybook dev build (absolute numbers are inflated by React dev mode):

Case	Old (ms)	New (ms)	Change
Mount, 50 rows x 8 columns	31.2	31.4	+1%
Mount, 500 rows on one page	318.2	288.5	-9%
Mount, 1,000 rows on one page	709.7	711.1	0%
Mount, 500 rows, maxBodyHeight	348.6	279.5	-20%
Mount, 20,000 rows, page of 50	50.5	49.6	-2%
Sort text column, 20,000 rows	168.7	106.6	-37%
Sort number column, 20,000 rows	50.0	50.9	+2%
searchKey keystroke, 20,000 rows	36.0	34.7	-4%
globalFilterFn keystroke, no match	17.0	7.3	-57%
Toggle one row, 500 visible	148.8	142.0	-5%
Resize step, 500 visible	119.6	121.7	+2%

Differences within about 5% are noise. Gzipped size of the two source files: data-table 2.8 KB to 5.6 KB, editable-cell 1.5 KB to 1.8 KB.

Not in this PR

• Row virtualization or memoized rows. With 500 rows on screen every interaction re-renders all of them, in both the old and new code.
• Drag-and-drop rows and expandable rows. Those two stories still use the plain Table.
• In maxBodyHeight mode, columns without an explicit size still share width equally.

Test plan

• data-table.test.tsx and editable-cell.test.tsx: 130 tests, run under a US timezone
• Full apollo-wind suite: 3513 tests pass with NODE_OPTIONS=--no-experimental-webstorage (without it, form-plugins.test.tsx fails on Node 26, a file this PR does not touch)
• tsc --noEmit, biome lint, build
• Driven by hand in Chromium: every editor type, paging, delete-selected, page-size picker, search, sort, resize by mouse and keyboard, sticky body scroll with forced classic scrollbars, loading and empty states, French strings, 375px viewport
• Review the behaviour changes above and decide on the semantic-release footer

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.

[github-actions]

Dependency License Review

• ✅ 1958 package(s) scanned
• ✅ No license issues found
• ⚠️ 2 package(s) excluded (see details below)

License distribution

License	Packages
MIT	1722
ISC	91
Apache-2.0	57
BSD-3-Clause	28
BSD-2-Clause	23
BlueOak-1.0.0	8
MPL-2.0	4
CC0-1.0	4
MIT-0	3
MIT OR Apache-2.0	2
(MIT OR Apache-2.0)	2
Unlicense	2
LGPL-3.0-or-later	1
Python-2.0	1
CC-BY-4.0	1
(MPL-2.0 OR Apache-2.0)	1
Unknown	1
Artistic-2.0	1
(WTFPL OR MIT)	1
(BSD-2-Clause OR MIT OR Apache-2.0)	1
CC-BY-3.0	1
0BSD	1
(MIT OR CC0-1.0)	1
MIT AND ISC	1

Excluded packages

Package	Version	License	Reason
@img/sharp-libvips-linux-x64	1.3.4	LGPL-3.0-or-later	LGPL pre-built binary, not linked
khroma	2.1.0	Unknown	MIT per GitHub repo, missing license field in package.json

[github-actions]

Storybook visual diff

⚠️ Visual changes detected: 33 changed, 6 added, 1 errored (of 70 compared, 30 unchanged). View report

Baseline is the deployed main Storybook, so changes merged to main after this branch was last updated can also appear here. Logs

Updated (PT): Nov 12, 2026, 12:20:22 PM

[github-actions]

📊 Coverage + size by package

Per-package coverage and bundle size on this PR. New-line coverage = of the source lines this PR adds or changes, the % hit by tests.

Package	Coverage	New-line coverage	Packed (gzip)	Unpacked	vs main
@uipath/apollo-core	—	—	42.02 MB	50.17 MB	±0
@uipath/apollo-react	59.7%	—	2.49 MB	14.02 MB	+96 B
@uipath/apollo-ui-icons	—	—	2.86 MB	6.97 MB	±0
@uipath/apollo-vertex	—	—	—	—	—
@uipath/apollo-wind	82.4%	96.9% (248/256)	908.5 KB	5.11 MB	+10.9 KB
@uipath/ap-chat	85.8%	—	40.48 MB	46.10 MB	±0

_{"Coverage" is each package's own coverage.include scope (e.g. apollo-core instruments only scripts/). "Packed"/"Unpacked" come from npm pack --dry-run and only cover built packages — "—" means not measured this run (package not affected / not built). "vs main" is the packed (gzipped) delta against the last successful main build (the package-sizes artifact from the Release workflow); "—" there means no main baseline was available this run. The baseline is main's latest build, not this PR's exact merge-base, so it includes any drift since the branch diverged. Packages with no vitest config are omitted.}