#79 test(silver): guard against silent fiscal-calendar fallback

[Anonymous]

Originally created by: grynn-in

Closes [#77].

Why

silver_gl_entries derives fiscal_year/period from a fiscal-calendar join, falling back to the posting date when the join misses (#65). The fallback is correct for calendar fiscal years but silently wrong for offset ones — and a missed join is what put every row in FY0/P0 before [#65]/#76. Both fixes are silent, so a future re-break wouldn't surface.

What

A singular test that fails when an entity with GL activity points at a fiscal calendar not loaded in silver_fiscal_periods — the precise miss condition. Scoped to active entities so inert seed rows (see [#78]) don't trip it.

Verification

dbt test --select assert_fiscal_calendar_is_loaded  ->  PASS

AMHQ/AMUS/AMDE all map to Standard, which is loaded. Flipping any active entity back to a non-existent calendar would now fail the build instead of quietly falling back.

[Anonymous]

Ticket changed by: grynn-in

• status: open --> closed