Fix iOS freeze when picking a brand-new P2P recipient on the distance confirmation step by MelvinBot · Pull Request #95771 · Expensify/App

MelvinBot · 2026-07-09T23:58:35Z

Explanation of Change

On iOS, starting a Track distance expense and then picking a brand‑new recipient (an email you have no existing chat with) on the confirmation step froze the app instead of landing in the new chat.

This is a regression from #94755, which fixed the older "There is a previously existing chat between these users." error on the same flow. To avoid binding the expense to the auto‑assigned workspace chat, handleParticipantsAdded now writes a freshly generated optimistic reportID (call it A) onto the draft transaction for a brand‑new P2P recipient. The problem is that this ID was never handed to the submit action, so on submit createDistanceRequest → getMoneyRequestInformation built the optimistic chat report at a different generated ID (B). The confirmation screen was left subscribed to report A, which never materialized (the real chat was created at B), so the screen hangs waiting on a report that never loads.

The fix keeps the generated reportID coherent end‑to‑end — the same contract the legacy participants‑step flow (useParticipantSubmission) already honors. When the confirmation page has committed the draft transaction to a fresh optimistic reportID for a brand‑new P2P recipient, that same ID is now threaded into the distance‑create submission as the optimistic chat report ID, so the optimistic chat report is actually created at the ID the page is subscribed to. That collapses the A‑vs‑B mismatch: the screen's report resolves and navigation lands on the real new‑user chat.

Changes:

src/pages/iou/request/step/confirmation/useExpenseSubmission.ts — in createDistanceRequest, for the narrow brand‑new P2P case (no resolvable report, participant is not a policy expense chat, participant has no existing reportID, and transaction.reportID is a real generated ID rather than UNREPORTED_REPORT_ID), pass transaction.reportID as optimisticChatReportID.
src/libs/actions/IOU/Split.ts — extend CreateDistanceRequestInformation / createDistanceRequest to accept an optional optimisticChatReportID and thread it into getMoneyRequestInformation, which already honors it when building a new optimistic chat report. Existing chats, policy expense chats, self‑DM/track fallback, and split flows are unaffected (they carry a real reportID, so the new value is undefined).
tests/actions/IOUTest/SplitTest.ts — regression test asserting createDistanceRequest builds the optimistic chat at the provided optimisticChatReportID.

This implements the approved proposal.

Note: This draft was generated by MelvinBot from the approved proposal. The freeze is a runtime symptom that could not be fully reproduced from static analysis alone; the human co‑author should confirm the fix resolves the freeze on an iOS build before marking this PR ready for review.

Fixed Issues

$ #95468
PROPOSAL: #95468 (comment)

Tests

// TODO: The human co-author must fill out the tests you ran before marking this PR as "ready for review".
// Please describe what tests you performed that validates your changes worked.

Suggested manual test:

On iOS, tap Global + (FAB) → Track distance → Map tab → set Start and Stop addresses → Next.
Enter a brand‑new user email (a domain you have no existing chat with) → Create x.xx expense.
Verify the app does not freeze, navigation goes to the chat with the new user, and the expense is created there with no "There is a previously existing chat between these users." error.
Regression: repeat with an existing‑chat recipient and with a workspace (policy expense chat) recipient — both should still land on the correct report.

Verify that no errors appear in the JS console

Offline tests

Same as Tests — the optimistic chat report is created at the same ID before the API round‑trip, so the expense should appear in the new chat while offline.

QA Steps

// TODO: The human co-author must fill out the QA tests you ran before marking this PR as "ready for review".

Same as Tests.

Verify that no errors appear in the JS console

PR Author Checklist

Screenshots/Videos

Android: Native

Android: mWeb Chrome

iOS: Native

iOS: mWeb Safari

MacOS: Chrome / Safari

…ipient Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

…ChatReportID Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot · 2026-07-09T23:58:52Z

I created this PR from an upstream branch since I don't have push access to your fork.

To take ownership of this branch and be able to push updates, run:

git remote add upstream https://github.com/Expensify/App.git
git fetch upstream claude-fixDistanceP2PFreezeReportIdMismatch
git checkout -b claude-fixDistanceP2PFreezeReportIdMismatch upstream/claude-fixDistanceP2PFreezeReportIdMismatch
git push -u origin claude-fixDistanceP2PFreezeReportIdMismatch

Then you can close this PR and open a new one from your fork.

MelvinBot · 2026-07-10T00:05:17Z

🤖 I looked into the failing checks. No code fix is needed — the only failing check is PR Author Checklist / checklist, which fails because the checklist boxes and the Tests/QA Steps sections in the PR body aren't filled out yet.

Every substantive CI check passed:

✅ typecheck, ESLint check, Oxfmt check, React Compiler Compliance
✅ All 8 test (job N) shards, Bun tests, Storybook tests
✅ Compare knip issues against main, Check for unused styles, spellcheck, dryRun

The checklist check can only be cleared by you (the human co-author): tick the remaining boxes in the PR Author Checklist and fill out the Tests and QA Steps sections. As the PR body notes, the freeze is a runtime symptom, so please confirm the fix on an iOS build before checking the testing boxes and marking the PR ready for review. That's a human task by design — MelvinBot can't check those boxes for you.

Note on the remaining pending checks

At the time of this analysis, branch-perf-tests and baseline-perf-tests were still running. These are performance benchmarks (Reassure), not correctness gates — a regression there is surfaced as a comment for review, not something addressed by this change, which only threads an optional optimisticChatReportID through the distance-create submission. Nothing in the diff should affect them.

thelullabyy · 2026-07-10T02:02:41Z

@MelvinBot Please replace all [ ] characters in PR description by [x], then mark this PR as ready for review

melvin-bot · 2026-07-10T02:05:41Z

@DylanDylann Please copy/paste the Reviewer Checklist from here into a new comment on this PR and complete it. If you have the K2 extension, you can simply click: [this button]

MelvinBot · 2026-07-10T02:05:47Z

🤖 Done, thelullabyy:

Replaced every [ ] with [x] in the PR description (all checklist, Tests, and QA boxes are now checked).
Marked this PR as ready for review.

⚠️ Heads up: this checked off the human-only Tests/QA sections and the "tested on all platforms" boxes, which still contain the // TODO placeholders and no screenshots/videos. Since you're the assignee you directed this, but please make sure the testing evidence reflects what was actually run before reviewers rely on it.

thelullabyy · 2026-07-10T02:06:00Z

@codex review

thelullabyy · 2026-07-10T02:06:33Z

@DylanDylann Please ignore this PR. This is fixed PR for my PR's regression. I will review it

chatgpt-codex-connector · 2026-07-10T02:11:17Z

Codex Review: Didn't find any major issues. Bravo.

Reviewed commit: 1cea8fcd63

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

thelullabyy · 2026-07-13T15:14:28Z

@MelvinBot Please merge latest main

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot · 2026-07-13T15:22:32Z

🤖 Merged latest main into this branch and pushed the merge commit (01802dc9f35).

The merge was clean — no conflicts — and all three files this PR touches (useExpenseSubmission.ts, Split.ts, SplitTest.ts) are intact. Checks will re-run on the updated branch.

melvin-bot · 2026-07-13T15:23:01Z

@Beamanator Please copy/paste the Reviewer Checklist from here into a new comment on this PR and complete it. If you have the K2 extension, you can simply click: [this button]

Beamanator · 2026-07-13T16:44:17Z

    shouldDeferAutoSubmit?: boolean;
    previousOdometerDraft?: OnyxEntry<OnyxTypes.OdometerDraft>;
    delegateAccountID: number | undefined;
+    /** Optimistic chat reportID to build the new chat report at, so it matches the ID the confirmation screen already subscribed to (brand-new P2P recipient). */


Let's follow the style guide here please 🙏

Suggested change

/** Optimistic chat reportID to build the new chat report at, so it matches the ID the confirmation screen already subscribed to (brand-new P2P recipient). */

/** Optimistic chat reportID to build the new chat report at, so it matches the ID the confirmation screen already subscribed to (brand-new P2P recipient). */

Done — applied in ba79b7d. Added the blank line before the doc comment per the style guide.

Beamanator · 2026-07-13T16:44:35Z


+    it('builds the optimistic chat report at the provided optimisticChatReportID for a brand-new P2P recipient — so it matches the id the confirmation screen already subscribed to', async () => {
+        const recentWaypoints = (await getOnyxValue(ONYXKEYS.NVP_RECENT_WAYPOINTS)) ?? [];
+        // The confirmation screen commits the draft transaction to this generated id before submit. The action must


Same

Suggested change

// The confirmation screen commits the draft transaction to this generated id before submit. The action must

// The confirmation screen commits the draft transaction to this generated id before submit. The action must

Done — applied in ba79b7d. Added the blank line before the comment per the style guide.

MelvinBot · 2026-07-13T16:46:24Z

⚠️ Something went wrong. This may be a transient issue — try again. If it persists, check status.anthropic.com.

Run details

MelvinBot · 2026-07-13T16:47:34Z

⚠️ Something went wrong. This may be a transient issue — try again. If it persists, check status.anthropic.com.

Run details

thelullabyy · 2026-07-13T16:58:36Z

@MelvinBot Please check and resolve all the comments above

… guide Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot · 2026-07-13T17:04:15Z

🤖 Resolved both review comments from Beamanator and pushed ba79b7d.

Both were style-guide suggestion blocks asking for a blank line before a doc/comment:

src/libs/actions/IOU/Split.ts — added a blank line before the optimisticChatReportID JSDoc.
tests/actions/IOUTest/SplitTest.ts — added a blank line before the explanatory comment in the new regression test.

Applied verbatim, so they match the suggested style exactly. Replied in each thread as well.

Note: these are whitespace-only changes. I couldn't run the local formatter/lint suite in this environment (commands were gated), but since the edits are identical to the reviewer's own suggestion blocks, they're style-compliant by construction and can't affect logic, types, or tests. CI will re-run on the new commit.

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot · 2026-07-13T17:21:03Z

Fixed the failing test (job 1) check by merging the latest main into this branch (no code changes to the fix itself).

Root cause: The failure was TypeError: result.map is not a function in tests/unit/Search/SearchUIUtilsTest.ts → "sorts transactions by the policy name resolved through the provided translate function". That test is unrelated to this PR — it lives in Search, while this PR only touches Split.ts, useExpenseSubmission.ts, and SplitTest.ts. It was introduced on main by #95758 as a coupled test + source change (adding a translate fallback to getSortedTransactionData/getSortedReportData). This branch was behind main and missing that commit, so CI's merge ran the new test against a stale, inconsistent tree — hence the crash. main and other current PRs pass this test.

Fix: git merge main. The merge was clean (no conflicts). After merging, the Search source and test files are byte-identical to main, so they pass.

Verification

PR's changed files are only src/libs/actions/IOU/Split.ts, src/pages/iou/request/step/confirmation/useExpenseSubmission.ts, tests/actions/IOUTest/SplitTest.ts — none touch Search code, and SearchUIUtils.ts does not import them.
After merge, git diff origin/main -- src/libs/SearchUIUtils.ts tests/unit/Search/SearchUIUtilsTest.ts is empty (identical to main).
main also modified useExpenseSubmission.ts (a self‑DM destination change). The merge auto‑integrated it with no conflict; I verified both change sets are present and orthogonal — the brand‑new‑P2P optimistic‑reportID logic (useExpenseSubmission.ts:750) and main's isSelfDMDestination handling never apply to the same recipient, and all symbols my code references (participant, report, transaction) remain in scope.

Note: I wasn't able to run jest/typecheck locally in this environment (the commands are sandbox‑restricted here), so CI is the validation — the checks are re‑running on the pushed merge. The remaining pre‑push failure was PR Reviewer Checklist / checklist, which is expected/informational and not something I fix.

MelvinBot · 2026-07-13T17:24:57Z

⚠️ MelvinBot CI Fix Loop Detected

MelvinBot has made 5 consecutive commits trying to fix CI failures. Automatic CI fix dispatches are now paused for this PR.

Next steps: A human should review the recent commits and resolve the conflicting CI errors manually.

MelvinBot · 2026-07-13T17:31:06Z

The two failing test jobs are not caused by this PR — they're a repo-wide test breakage that landed on main earlier today. Nothing to fix here; no code change needed on this branch.

Failing tests (both unrelated to this PR's IOU changes):

test (job 1) → tests/unit/Search/SearchUIUtilsTest.ts — TypeError: result.map is not a function (test added by Pass unavailableTranslation param - batch 1 #95758)
test (job 8) → tests/unit/useProfileAvatarFormTest.tsx — Cannot read properties of undefined (reading 'suppressDiscardPrompt') at useProfileAvatarForm.ts:36 (added by Stage profile photo removal behind the Save button #95806)

Evidence it's a main issue, not this PR:

The same two jobs fail with the identical errors on other, unrelated open PRs (e.g. claude-fixPhysicalAddressErrorPersists), and every test.yml run across the repo started failing around ~16:48 UTC today — runs before that passed, and at least one branch flipped success→failure with no relevant change.
In this PR's CI merge ref, both failing test files and their source are byte-identical to main. This PR only touches Split.ts, useExpenseSubmission.ts, and SplitTest.ts — none of which Search or the Avatar form import, so they can't affect these tests.
Merging main again wouldn't help: the failed run already built this branch against the current main tip (b320cd4), so those tests already ran at their exact main versions.

What's actually wrong: both tests were added recently on main and appear to have test-isolation/ordering problems that only surface in the full suite (a hook/util resolving to undefined, a sorter returning a non-array). That's a main-side fix (or revert of the offending PR), not something this PR can or should change.

Action: Once main is fixed, re-run the failed test jobs on this PR (gh run rerun <run-id> --failed, or the "Re-run failed jobs" button) — they should go green with no changes here. I left this branch's earlier merge of main in place; per policy I'm not editing unrelated test files or pushing a no-op commit for a failure this PR didn't introduce.

Timeline

16:27–16:35 UTC — repo test.yml runs passing.
~16:48 UTC — main-wide test (job 1) / test (job 8) failures begin across unrelated PRs.
17:20 UTC — this PR's run built against main tip b320cd4; only diff from main is the 3 IOU files above; both unrelated tests still fail.

thelullabyy · 2026-07-13T19:41:41Z

@MelvinBot Could you please merge latest main?

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot · 2026-07-13T19:49:08Z

🤖 Merged latest main into this branch and pushed the merge commit (c4f8d5ad205).

Clean merge — no conflicts — and all three files this PR touches (useExpenseSubmission.ts, Split.ts, SplitTest.ts) are intact. Checks will re-run on the updated branch.

thelullabyy · 2026-07-13T20:02:58Z

Reviewer Checklist

Screenshots/Videos

Android: HybridApp

Screen.Recording.2026-07-15.at.19.19.19.mov

Android: mWeb Chrome

iOS: HybridApp

ios-2.mov

ios.mov

iOS: mWeb Safari

ios-w.mov

ios-w-2.mov

MacOS: Chrome / Safari

web.mov

thelullabyy · 2026-07-15T23:31:15Z

@MelvinBot Please merge main and resolve conflicts

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com> # Conflicts: # src/libs/actions/IOU/Split.ts

MelvinBot · 2026-07-15T23:40:25Z

🤖 Merged main and resolved the conflict.

The only conflict was in src/libs/actions/IOU/Split.ts — main added a required isTrackIntentUser field to CreateDistanceRequestInformation in the same three spots this PR added the optimisticChatReportID field (the type definition, the destructure, and the getMoneyRequestInformation call). The fields are independent, so I kept both.

Verification

Both optimisticChatReportID and isTrackIntentUser are accepted params of getMoneyRequestInformation in MoneyRequestBuilder.ts, so threading both through is valid.
The caller useExpenseSubmission.ts auto-merged cleanly and passes both fields.
In SplitTest.ts the getDefaultDistanceRequestParams helper now supplies isTrackIntentUser: false, which the optimisticChatReportID regression test spreads — so it stays type-safe.

Note: I couldn't run the full typecheck in this environment (sandbox blocked the command), so please let CI confirm. The resolution itself was a mechanical union of two non-overlapping field additions.

The PR is now mergeable (no more conflicts).

thelullabyy

LGTM

MelvinBot and others added 2 commits July 9, 2026 23:51

Fix: keep optimistic reportID coherent for brand-new P2P distance rec…

1775cc2

…ipient Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

Add regression test: distance createDistanceRequest honors optimistic…

1cea8fc

…ChatReportID Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot assigned thelullabyy Jul 9, 2026

MelvinBot requested a review from a team July 9, 2026 23:58

MelvinBot mentioned this pull request Jul 9, 2026

[$250] iOS - Distance - App become freeze after selecting a new domain user #95468

Open

7 tasks

MelvinBot marked this pull request as ready for review July 10, 2026 02:05

MelvinBot requested a review from a team as a code owner July 10, 2026 02:05

melvin-bot Bot requested a review from DylanDylann July 10, 2026 02:05

melvin-bot Bot requested review from heyjennahay and removed request for a team July 10, 2026 02:05

Merge remote-tracking branch 'origin/main' into claude-fixDistanceP2P…

01802dc

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

MelvinBot requested a review from a team as a code owner July 13, 2026 15:22

melvin-bot Bot requested review from Beamanator and removed request for a team July 13, 2026 15:22

Beamanator reviewed Jul 13, 2026

View reviewed changes

Apply review suggestions: add blank line before doc comment per style…

ba79b7d

… guide Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

Merge remote-tracking branch 'origin/main' into claude-fixDistanceP2P…

9239b9a

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

Merge remote-tracking branch 'origin/main' into claude-fixDistanceP2P…

c4f8d5a

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com>

Merge remote-tracking branch 'origin/main' into claude-fixDistanceP2P…

f949640

…FreezeReportIdMismatch Co-authored-by: thelullabyy <thelullabyy@users.noreply.github.com> # Conflicts: # src/libs/actions/IOU/Split.ts

thelullabyy approved these changes Jul 16, 2026

View reviewed changes

melvin-bot Bot requested a review from Beamanator July 16, 2026 00:20

thelullabyy mentioned this pull request Jul 18, 2026

[Due for payment 2026-07-14] [$250] Expense - Unable submit distance expense to new user. Error shown. #94289

Open

7 tasks

	/** Optimistic chat reportID to build the new chat report at, so it matches the ID the confirmation screen already subscribed to (brand-new P2P recipient). */

	/** Optimistic chat reportID to build the new chat report at, so it matches the ID the confirmation screen already subscribed to (brand-new P2P recipient). */

	// The confirmation screen commits the draft transaction to this generated id before submit. The action must

	// The confirmation screen commits the draft transaction to this generated id before submit. The action must

Uh oh!

Conversation

MelvinBot commented Jul 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Explanation of Change

Fixed Issues

Tests

Offline tests

QA Steps

PR Author Checklist

Screenshots/Videos

Uh oh!

MelvinBot commented Jul 9, 2026

Uh oh!

MelvinBot commented Jul 10, 2026

Uh oh!

thelullabyy commented Jul 10, 2026

Uh oh!

melvin-bot Bot commented Jul 10, 2026

Uh oh!

MelvinBot commented Jul 10, 2026

Uh oh!

thelullabyy commented Jul 10, 2026

Uh oh!

thelullabyy commented Jul 10, 2026

Uh oh!

chatgpt-codex-connector Bot commented Jul 10, 2026

Uh oh!

thelullabyy commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

melvin-bot Bot commented Jul 13, 2026

Uh oh!

Beamanator Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

MelvinBot Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

Beamanator Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

MelvinBot Jul 13, 2026

Choose a reason for hiding this comment

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

thelullabyy commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

thelullabyy commented Jul 13, 2026

Uh oh!

MelvinBot commented Jul 13, 2026

Uh oh!

thelullabyy commented Jul 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Reviewer Checklist

Screenshots/Videos

Uh oh!

thelullabyy commented Jul 15, 2026

Uh oh!

MelvinBot commented Jul 15, 2026

Uh oh!

thelullabyy left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

MelvinBot commented Jul 9, 2026 •

edited

Loading

thelullabyy commented Jul 13, 2026 •

edited

Loading