fix(seed): repair qa seed so main compiles (actual_end_date + AuditCtx) (PMS-379) #288

Merged

vas2000-work merged 2 commits from fix/qa-seed-build-actual-end-date-audit-ctx into main

2026-06-17 02:08:44 +02:00

vas2000-work commented

2026-06-17 01:55:47 +02:00

Owner

main has not compiled since the PMS-361 (actual_end_date on CreateProjectRequest) and the audit-sweep merges (trailing ctx: &AuditCtx on create_project/phase/task/task_status + time_tracking create_work_type/create_time_entry). The QA seed (src/modules/seed/qa.rs) was never updated, so cargo check --all-targets failed with 8 errors (E0063 x2 + E0061 x6), which is why check.yml has been red (PMS-379, run #1755 on d6d16df1).

Fix: thread the existing AuditCtx::system(...) through the seed call sites, add one for the two helper fns via tenant.get(), and set actual_end_date: None on both project specs. No behaviour change.

Verified in-container (just pre-commit): fmt + clippy -D warnings + check --all-targets + unit tests + doc tests all green.

Fixes PMS-379.

main has not compiled since the PMS-361 (actual_end_date on CreateProjectRequest) and the audit-sweep merges (trailing `ctx: &AuditCtx` on create_project/phase/task/task_status + time_tracking create_work_type/create_time_entry). The QA seed (`src/modules/seed/qa.rs`) was never updated, so `cargo check --all-targets` failed with 8 errors (E0063 x2 + E0061 x6), which is why check.yml has been red (PMS-379, run #1755 on d6d16df1). Fix: thread the existing `AuditCtx::system(...)` through the seed call sites, add one for the two helper fns via `tenant.get()`, and set `actual_end_date: None` on both project specs. No behaviour change. Verified in-container (`just pre-commit`): fmt + clippy -D warnings + check --all-targets + unit tests + doc tests all green. Fixes PMS-379.

vas2000-work added 1 commit

2026-06-17 01:55:47 +02:00

fix(seed): repair qa seed for actual_end_date + AuditCtx create signatures

Check / fmt + clippy + compile + unit/doc tests (pull_request) Failing after 9s

Details

E2E / Playwright against staging (pull_request) Successful in 1m53s

Details

Integration / integration tests (pull_request) Failing after 1m58s

Details

5869e1f8e3

main stopped compiling: the QA seed (src/modules/seed/qa.rs) was not
updated when two changes landed:

- PMS-361 added the required `actual_end_date` field to
  CreateProjectRequest, but the two qa_project_specs literals omitted it
  (E0063).
- The "emit Create audit rows for every create path" sweep added a
  trailing `ctx: &AuditCtx` to create_project / create_project_phase /
  create_task / create_task_status and to time_tracking
  create_work_type / create_time_entry, but the seed call sites still
  passed the old arity (E0061 x6).

Thread the existing `AuditCtx::system(...)` through the seed() call sites
(a ctx is already built there), add one for the two helper fns that lack
it (via tenant.get()), and set actual_end_date: None on both project
specs. No behaviour change; restores `cargo check --all-targets`.

#PMS-379

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

vas2000-work added 1 commit

2026-06-17 02:01:30 +02:00

fix(migrations): renumber duplicate 051 prefixes to 052/053

E2E / Playwright against staging (pull_request) Successful in 1m8s

Details

Check / fmt + clippy + compile + unit/doc tests (pull_request) Successful in 2m26s

Details

Integration / integration tests (pull_request) Successful in 6m41s

Details

Create release / Create release from merged PR (pull_request) Has been skipped

Details

2bc8995e15

Three migrations collided on prefix 051 (parallel PRs: PMS-315 mileage,
PMS-373 payments, PMS-339 tax-rate), which trips the check.yml
"Migration prefix uniqueness" gate (sqlx keys _sqlx_migrations on the
numeric prefix). Keep the earliest-merged one (051_mileage_entries) as
051 - it is the only one that could already be in a deployed ledger -
and bump the two later, never-deployed ones:
  051_payments_amount_positive -> 052
  051_tax_rate_widen_column    -> 053
All three are independent (new table / CHECK / column widen), so order
is irrelevant. Prefixes are now unique.

#PMS-379

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>