* fix(vat): make the ruta drill-down reconcile with the figure it explains
get_vat_declaration_totals drops four classes of entry before summing: posted
closing entries, source_type 'vat_settlement', the two kontantmetod year-end
reversals, and anything shaped like a momsredovisning. The drill-down behind
each ruta filtered on company, status and date only.
So expanding a ruta listed verifikat that are not in the number it claims to
explain, and the panel shows no total that would reveal the mismatch. On
production, 322 posted/reversed entries carrying 26xx lines across 214
companies sit in those excluded classes.
A momsdeklaration is räkenskapsinformation under BFL 5 kap. and this
drill-down is what a consultant uses to substantiate a filed figure, so the
two have to agree exactly.
The exclusion CTEs are lifted verbatim from the figure rather than re-derived,
because any divergence reintroduces exactly this bug. The new pg test asserts
the equality for the whole account set at once, so editing one function and
not the other fails CI instead of silently misreporting.
opening_balance entries are deliberately kept: the figure exempts them from
its `shaped` set, which leaves their lines in the totals, so excluding them
here would break the equality in the other direction. That has its own test.
Verified the test catches the defect by reinstalling the old function body and
watching it fail with the real numbers (2611: drill-down 250/240 vs figure
0/200), then restoring.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* fix(vat): update the existing drill-down pg test to the new signature
get_vat_ruta_source_lines gained p_ruta_accounts / p_net_accounts, and
production-error-regressions.pg.test.ts still called the old 9-argument form,
so pg-real failed with 42883 "function does not exist". I had grepped app/,
lib/ and extensions/ for callers and not tests/.
Neither fixture in that paging test is settlement-shaped, so paging behaviour
is unchanged; the equality itself is covered by the new reconcile test.
Also documents, in the tool-pg reset script, that its blanket grant to `anon`
(which PostgREST requires) makes that database invalid for the pg-real suite:
~29 of those files assert least privilege and fail there even on unmodified
main. That cost a confusing local run.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
All 100 files in extensions/general/mcp-server/__tests__ fake supabase.
query-journal.test.ts says out loud that its query chain is "exercised by the
live MCP smoke test", and no such test exists in CI. So the PostgREST grammar
of 157 tools, every .select() column string, every resource embed, every
or=(...) form, is gated by nothing and fails first in production.
pg-real cannot cover this: it holds a pg Pool and writes SQL, and none of that
grammar is resolved by Postgres. It is resolved by PostgREST at request time.
Adds a tool-pg vitest project, a docker-compose stack, a reset script that
replays every migration the way the pg-real CI job does, and a CI job.
The first sweep covers 74 read tools and finds no malformed query, across 87
real requests. That number is honest rather than impressive: with an empty
argument set many tools bail before querying. Per-tool fixtures are what
deepen it, and this harness is what makes writing them worth the effort.
Includes a self-test that injects a bad column and asserts the harness detects
it. That is not ceremony. It caught this file passing green while exercising
nothing, twice: once locally where supabase-js prefixes /rest/v1 onto a bare
PostgREST that does not serve it, and once on CI where Node 20 has no native
WebSocket, so every client construction threw and was swallowed by the
per-tool catch as a domain refusal. The client is now built once outside that
catch, the proof-of-life assertion counts real requests instead of being
trivially satisfiable, and realtime gets an inert transport.
Also excludes .next from all three vitest projects. These projects override
vitest's default excludes, so a local `npm run build` leaves a traced copy of
the repo that gets collected as a second set of test files.
Co-authored-by: Jakob Wennberg <311770904+jakobwennberg-oss@users.noreply.github.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Jakob Wennberg