Sunday, December 22, 2024
2 changes · master
Resolved issues and error corrections
This change increases the allowed time for a web desktop unit test that was occasionally failing because the limit was too tight. It helps reduce random test failures and keeps development checks more stable while a deeper investigation is pending.
Original PR description
The timeout is close to the real value leading to random failures. Increasing it waiting for a proper investigation.
Fixes small issues in the updated Trial Balance report so fiscal-year calculations include the correct start date and reports can be grouped in more flexible ways. This helps accounting users get accurate balances and use partner-based groupings without errors.
Original PR description
Minor bugfixes found when testing the new refactored trial balance.
1. Fix condition to be `date >= fiscalyear_start` rather than `date > fiscalyear_start` when filtering out AMLs in P&L accounts before the start of the fiscal year.
2. Enable groupings by other fields than account first; for example, grouping by partner_id then account_id should work:
a. Instead of attempting to de-mogrify the complete grouping key retrieved from Postgres, we instead get its component columns separately. This allows us to gracefully handle the partner_id column being NULL.
b. We only compute the `is_unaffected_earnings` column when it's needed, that is, when we're doing a grouping by account_or_unaff_id.
task-3748928