You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[Due for payment 2026-10-01] Use the Table component for the "Add existing expense" flow #100589
Add expense > Add existing expense (from a report) renders its expense list with SelectionList and applies no sort at all. Auth returns rows newest-first, but the response is merged into the Onyx transactions_ collection and read back out with getEligibleTransactionsToAdd, which does Object.values(transactions).filter(...) — the API ordering is discarded and display order is just Onyx key iteration order. AddExistingExpense.tsx then only filters (search text, status) before mapping to list items, so users see an effectively random ordering and it's hard to bulk-select the expenses they want.
A customer flagged this; the current workaround is to go to Expenses, filter by Unreported + Draft, and bulk-move to a report instead.
Proposal
Convert this view to the Table component (src/components/Table) instead of bolting a one-off sort onto SelectionList. Table already ships sortable column headers (Table.Header), a search/filter bar (Table.FilterBar), empty/no-results/loading states, and multi-select — so we get sorting for free and the list matches how Spend > Expenses behaves.
Default sort should be expense date descending, matching our default expense ordering elsewhere.
Files
src/pages/AddExistingExpense.tsx — swap SelectionList for Table, move the status filter into Table.FilterBar, keep the select-all and footer behavior.
src/pages/UnreportedExpenseListItem.tsx — becomes the row renderer / column definitions.
contributingGuides/TABLE.md — component docs; see src/components/Tables/* for existing adoptions.
Things to watch
Pagination: the page fetches 50 at a time via fetchUnreportedExpenses and onEndReached. Sorting client-side only sorts what's loaded, so confirm the sort still reads correctly as more pages arrive.
Preserve the existing mobile/native layout — check the table renders acceptably on small screens.
[Contributor] The offending PR and associated issue have been commented on, pointing out the bug it caused and why, so the author and reviewers can learn from the mistake.
Link to the comment on the PR:
Link to the comment on the Issue:
[Contributor] If the regression was CRITICAL (e.g. interrupts a core flow) A discussion in #expensify-open-source has been started about whether any other steps should be taken (e.g. updating the PR review checklist) in order to catch this type of bug sooner.
Link to discussion:
[Contributor] If it was decided to create a regression test for the bug, please propose the regression test steps using the template below to ensure the same bug will not reach production again.
Regression Test Proposal
Test:
Open a report/IOU with eligible unreported/draft expenses.
Go to Add expense > Add existing expense and verify expenses load, sorted by date descending.
Click Confirm with nothing selected → verify "Select at least one expense" appears.
Select expenses → click Confirm → verify no error and the flow proceeds.
Select 2-3 expenses, search, then clear the search → verify selections are preserved.
Apply Status filter → verify results filter correctly and selection clears as expected.
Sort by Date and Amount → verify asc/desc toggling works.
Scroll to paginate → verify more expenses load and skeleton appears while loading.
Resize to narrow/RHP width → verify no layout overflow/clipping.
Test offline → verify pending-action feedback renders correctly.
Confirm selected expenses are successfully added to the report.
@mallenexpensify@JS00001, could we reduce the payment cut a bit on this one? The only real regression was #102003, and #102021 is really an edge case, so I don't think reducing it to $75 is warranted.
I think $125 would be fair given the effort involved. $75 feels a bit brutal 😅. These PRs takes time to test, and I often find and fix other issues along the way too. For example, while converting the table, I also noticed the skeletons needed updating and fixed those as well. Getting everything right across different cases takes a fair bit of time and effort.
Would really appreciate it if we could reconsider the final amount. 🙏
Activity
shawnborton commentedon Sep 8, 2026
@Krishna2323 can handle implementation on this one, then we can get a C+ review.
JS00001 commentedon Sep 10, 2026
Krishna's working on this one
Krishna2323 commentedon Sep 11, 2026
@Expensify/design The PR is ready for initial testing:
Monosnap.screencast.2026-09-11.16-30-33.1.mp4
dannymcclain commentedon Sep 11, 2026
Video is looking great. Such a better experience than a random unsorted list! 😂
situchan commentedon Sep 11, 2026
I can review this as well
shawnborton commentedon Sep 11, 2026
Love it, looks great!
22 remaining items
situchan commentedon Oct 2, 2026
Contributor+ Checklist:
[Contributor] The offending PR and associated issue have been commented on, pointing out the bug it caused and why, so the author and reviewers can learn from the mistake.
Link to the comment on the PR:
Link to the comment on the Issue:
[Contributor] If the regression was CRITICAL (e.g. interrupts a core flow) A discussion in #expensify-open-source has been started about whether any other steps should be taken (e.g. updating the PR review checklist) in order to catch this type of bug sooner.
Link to discussion:
[Contributor] If it was decided to create a regression test for the bug, please propose the regression test steps using the template below to ensure the same bug will not reach production again.
Regression Test Proposal
Test:
Do we agree 👍 or 👎
situchan commentedon Oct 2, 2026
3 deploy blockers linked to this issue:
mallenexpensify commentedon Oct 3, 2026
Payment Summary
Contributor: @Krishna2323 due $75 via NewDot
Contributor+: @situchan due $75 via NewDot
Krishna2323 commentedon Oct 5, 2026
@mallenexpensify @JS00001, could we reduce the payment cut a bit on this one? The only real regression was #102003, and #102021 is really an edge case, so I don't think reducing it to $75 is warranted.
I think $125 would be fair given the effort involved. $75 feels a bit brutal 😅. These PRs takes time to test, and I often find and fix other issues along the way too. For example, while converting the table, I also noticed the skeletons needed updating and fixed those as well. Getting everything right across different cases takes a fair bit of time and effort.
Would really appreciate it if we could reconsider the final amount. 🙏
situchan commentedon Oct 5, 2026
Agree and I suggest to consider #102021 as separate issue as the similar production bug happens in other places.
JS00001 commentedon Oct 5, 2026
I agree with increasing this up to $125, cc @mallenexpensify
mallenexpensify commentedon Oct 5, 2026
Payment Summary
Contributor: @Krishna2323 due $125 via NewDot
Contributor+: @situchan due $125 via NewDot
laurenreidexpensify commentedon Oct 21, 2026
Approved @Krishna2323 $125 New Expensify
laurenreidexpensify commentedon Nov 11, 2026
Approved $125 @situchan New Expensify