Repository navigation
[Due for payment 2025-05-27] [$500 AND $250] Investigate codebase for potential endless loops #60889
Description
If you haven’t already, check out our contributing guidelines for onboarding and email contributors@expensify.com to request to join our Slack channel!
Version Number:
Reproducible in staging?:
Reproducible in production?:
If this was caught on HybridApp, is this reproducible on New Expensify Standalone?:
If this was caught during regression testing, add the test name, ID and link from TestRail:
Email or phone of affected tester (no customers):
Logs: https://stackoverflow.com/c/expensify/questions/4856
Expensify/Expensify Issue URL:
Issue reported by: @JmillsExpensify
Slack conversation (hyperlinked to channel name): https://expensify.slack.com/archives/C05LX9D6E07/p1745490519573089
Action Performed:
Break down in numbered steps
- Go to Reports
- Search some query
Expected Result:
Describe what you think should've happened
User gets valid reponse
Actual Result:
Describe what actually happened
The browser just hangs as there is some js loop
Workaround:
Can the user still use Expensify without this being fixed? Have you informed them of the workaround?
Platforms:
Select the officially supported platforms where the issue was reproduced:
- Android: Standalone
- Android: HybridApp
- Android: mWeb Chrome
- iOS: Standalone
- iOS: HybridApp
- iOS: mWeb Safari
- Windows: Chrome
- MacOS: Chrome / Safari
- MacOS: Desktop
Platforms Tested:
On which of our officially supported platforms was this issue tested:- Android: Standalone
- Android: HybridApp
- Android: mWeb Chrome
- iOS: Standalone
- iOS: HybridApp
- iOS: mWeb Safari
- Windows: Chrome
- MacOS: Chrome / Safari
- MacOS: Desktop
Activity
melvin-bot commentedon Apr 25, 2025
Triggered auto assignment to @alexpensify (
Bug), see https://stackoverflow.com/c/expensify/questions/14418 for more details. Please add this bug to a GH project, as outlined in the SO.melvin-bot commentedon Apr 25, 2025
Triggered auto assignment to @aldo-expensify (
AutoAssignerNewDotQuality)mountiny commentedon Apr 25, 2025
@aldo-expensify Check out the slack thread, the idea here is that someone from Callstack will jump on this and try to investigate the codebase around search for some possible endless loops
sosek108 commentedon Apr 25, 2025
Hi, I'm Hubert from Callstack and I would like to take a look at this issue.
aldo-expensify commentedon Apr 28, 2025
👋 @sosek108 how is this going?
sosek108 commentedon Apr 28, 2025
On slack conversation we diagnosed that potential cause for problem is
SuffixUkkonenTreealgorithm. There is no endless loop, but a lot of calls for finite loopsI have proposal for minor fixes to that algorithm,
sosek108 commentedon Apr 28, 2025
🚨 Edited by proposal-police: This proposal was edited at 2025-04-28 13:06:15 UTC.
Proposal
Please re-state the problem that we are trying to solve in this issue.
The convertToBase26 function incorrectly divides by 32 using a bitwise shift (>> 5) instead of properly dividing by 26. This causes wrong base-26 conversions for many numbers, especially above 26, and leads to incorrect results. Additionally, the caching mechanism incorrectly stores results under the wrong key due to modifying the input num inside the function.
What is the root cause of that problem?
The root cause is twofold:
Misuse of bitwise shifting (>> 5) under the false assumption that it mimics division by 26, while it actually divides by 32.
Caching the result after modifying the input parameter, resulting in wrong cache keys and cache misses.
What changes do you think we should make in order to solve the problem?
Replace the bitwise shift (num >>= 5) with correct integer division (Math.floor(num / 26)).
Preserve the original num value in a separate variable before any mutation, and use the original number as the key when caching the result.
Slightly improve readability by avoiding mutation of the input parameter directly.
What specific scenarios should we cover in automated tests to prevent reintroducing this issue in the future?
Validate that numbers around 26, 32, 52, 53, and 63 correctly convert to base-26 arrays.
Add tests for small numbers (e.g., 1, 2, 25, 26, 27) to ensure basic behavior.
Include tests with numbers where division by 26 and division by 32 would differ to specifically catch this kind of error
Add a test ensuring that negative inputs throw the correct error message.
These tests ensure correctness of logic for both typical cases and edge cases, and will immediately fail if future changes accidentally introduce wrong division logic again.
What alternative solutions did you explore? (Optional)
N/A
65 remaining items
mountiny commentedon May 21, 2025
Ok yeah so we have $500 to @ZhenjaHorbach and $250 to @jayeshmangwani
alexpensify commentedon May 21, 2025
Ok, we will use those numbers on the payment date of 2025-05-27.
[-][Due for payment 2025-05-27] [Due for payment 2025-05-22] [Due for payment 2025-05-14] Investigate codebase for potential endless loops[/-][+][Due for payment 2025-05-27] Investigate codebase for potential endless loops[/+]melvin-bot commentedon May 27, 2025
Payment Summary
Upwork Job
BugZero Checklist (@alexpensify)
[-][Due for payment 2025-05-27] Investigate codebase for potential endless loops[/-][+][Due for payment 2025-05-27] [$500] Investigate codebase for potential endless loops[/+]melvin-bot commentedon May 28, 2025
alexpensify commentedon May 28, 2025
Payment Summary
Contributor: @jayeshmangwani due $250 via NewDot - please submit a request
Contributor: @ZhenjaHorbach due $500 via NewDot - please submit a request
jayeshmangwani commentedon May 28, 2025
@alexpensify Please update the payment summary - $250 is due for me.
alexpensify commentedon May 28, 2025
@jayeshmangwani - done! Thanks for the update.
[-][Due for payment 2025-05-27] [$500] Investigate codebase for potential endless loops[/-][+][Due for payment 2025-05-27] [$500 + $250] Investigate codebase for potential endless loops[/+][-][Due for payment 2025-05-27] [$500 + $250] Investigate codebase for potential endless loops[/-][+][Due for payment 2025-05-27] [$500 AND $250] Investigate codebase for potential endless loops[/+]garrettmknight commentedon Jun 5, 2025
$250 approved for @jayeshmangwani
ZhenjaHorbach commentedon Jun 19, 2025
Requested in ND
garrettmknight commentedon Jun 26, 2025
$250 approved for @ZhenjaHorbach