Files
accounted/extensions/general/enable-banking/lib/date-suggestions.ts
T
MattssonandClaude Fable 5 a47ba9fede fix(enable-banking): stop renewal history floods (gap-fill default + backfill reconciliation) (#1590)
* chore(ci): guard bedrock-sdk against automated version bumps

The 2026-07 prod outage (empty Bedrock streams breaking invoice OCR and
the assistant) came from an unreviewed @anthropic-ai/bedrock-sdk 0.32.0
bump. The package is exact-pinned to 0.29.1, but nothing stopped an
automated PR from proposing the bump again. Add a dependabot config in
security-updates-only posture (open-pull-requests-limit: 0) with an
ignore for bedrock-sdk >=0.30.0 so neither scheduled nor security
updates can reintroduce it silently.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(enable-banking): stop renewal history floods (gap-fill default + backfill reconciliation)

Renewing a bank connection walks the same pending_selection -> active
flow as a first connect, and a fresh consent often makes the bank
release history the first connect never delivered. Two gaps turned that
into a flood of falsely 'unhandled' rows over already-bookkept periods
(11 companies, ~600 rows in prod):

- The picker defaulted every renewal to the fiscal-year lookback. It now
  probes the connection's newest imported transaction and defaults a
  renewal to 'continue where the last fetch stopped' (7-day overlap,
  absorbed by external_id dedup), with an .attn warning when a longer
  lookback re-requests already-fetched periods.
- The inline initial backfill ran without the SIE-overlap guard that the
  manual /sync route and the cron both apply. It now suppresses
  auto-categorization on overlap and runs the same unattended-threshold
  reconciliation sweep, scoped per ledger account via
  resolveCashAccountScope instead of the pooled unscoped form
  (#1290/#1298), and surfaces the linked count as auto_matched in the
  sync summary.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(enable-banking): renewal guard survives mid-open refetch; harden sweep scope and window

Skeptic pass findings on the renewal-flood guard:

- REFUTED: the settings panel's visibility refetch (near-certain in a
  BankID reconnect) hands the open picker a fresh accounts identity; the
  pre-existing reset effect then wiped the gap-fill state while the
  probe effect never re-ran, silently stranding the renewal back on the
  fiscal-year default with no warning. The probe now keys its state by
  connectionId and shares the reset's triggers via an accounts dep, so
  wipe and re-probe always pair up.
- The sweep skips accounts whose cash_accounts row did not resolve
  (found: false) instead of degrading to the pooled currency-only form
  (#1290 write shape), which could otherwise follow a same-request
  mirror-upsert failure.
- The sweep window opens at the oldest booking date the bank actually
  returned: over-returning ASPSPs ingest rows outside the requested
  window, which the sweep would otherwise never examine.
- resolveGapFillStart clamps to the backend's 365-day lookback floor so
  the radio never promises a start date the backfill cannot honor.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(enable-banking): address PR #1590 review findings

- zizmor: add a 7-day cooldown to the dependabot npm entry.
- CodeRabbit: clear the event bus in the accounts-route beforeEach (repo
  test convention); surface probe query failures in AccountPickerDialog
  so a failed probe cannot read as a first connect and silently restore
  the fiscal-year default; build the sweep's ledger-account list with a
  string filter instead of a nullish fallback so the pooled scope path
  is structurally unreachable.
- Swedish compliance review: document at the sweep site that linking
  writes bank-feed metadata only, never journal tables, with the
  opening-balance link trigger and unlinkReconciliation reversibility
  spelled out.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-08-13 16:50:24 +02:00

121 lines
5.2 KiB
TypeScript

import type { CompanySettings } from '@/types'
import { getCurrentFiscalYearStart } from '@/lib/company/fiscal-year'
export interface BookedCoverage {
/** Entry date of the company's latest posted verifikat. */
lastBookedDate: string
/**
* Day after lastBookedDate (the earliest sync start that cannot overlap
* booked entries), clamped to today (UTC) so the backend accepts it.
*/
suggestedStartDate: string
}
/**
* Turn the latest posted verifikat date into a "start syncing from" suggestion.
*
* Issue #917: this used to be derived from sie_imports.fiscal_year_end, which
* is the fiscal PERIOD end, not how far the bookkeeping actually reaches. For
* a company whose SIE covered an extended first year (2025-10-01 to 2026-12-31)
* but whose entries stopped in May, the old value suggested a start date past
* every unbooked transaction. Returns null when there is nothing booked: no
* suggestion beats a misleading one.
*/
export function resolveBookedCoverage(
lastPostedEntryDate: string | null | undefined,
today: Date = new Date(),
): BookedCoverage | null {
if (!lastPostedEntryDate) return null
// Pin the math to UTC so the day-after arithmetic is timezone-independent.
const d = new Date(lastPostedEntryDate + 'T00:00:00Z')
d.setUTCDate(d.getUTCDate() + 1)
const dayAfter = d.toISOString().split('T')[0]
// The backend PATCH handler (index.ts) rejects initial_lookback_from_date
// unless Date.now() is past the date's UTC midnight, so the newest date it
// accepts is the current UTC date. A company whose latest verifikat is
// dated today (plausible at initial bank activation) would otherwise get
// tomorrow suggested here and a 400 when saving. Clamp to today; ISO date
// strings compare correctly as plain strings.
const todayUtc = today.toISOString().split('T')[0]
return {
lastBookedDate: lastPostedEntryDate,
suggestedStartDate: dayAfter <= todayUtc ? dayAfter : todayUtc,
}
}
export interface GapFillSuggestion {
/** Date of the newest transaction this connection has already imported. */
latestImportedDate: string
/**
* Suggested sync start: latestImportedDate minus GAP_FILL_OVERLAP_DAYS,
* clamped to today (UTC) so the backend accepts it.
*/
suggestedStartDate: string
}
/**
* Overlap requested before the newest already-imported row. The external_id
* dedup makes re-imported rows no-ops, so the overlap costs nothing, and it
* catches transactions the bank booked late around the boundary.
*/
export const GAP_FILL_OVERLAP_DAYS = 7
/**
* Turn the newest transaction a connection has already imported into a
* "continue where the last fetch stopped" suggestion for RENEWALS.
*
* A reconnect walks the same pending_selection → active flow as a first
* connect, and a fresh consent often makes the bank release history the first
* connect never delivered. Re-requesting a long lookback then floods the inbox
* with rows over already-bookkept periods (the 2026-08 renewal flood), so a
* renewal should default to fetching only the gap since the last import.
* Returns null when the connection has never imported anything: a first
* connect has no gap to fill.
*/
export function resolveGapFillStart(
latestImportedDate: string | null | undefined,
today: Date = new Date(),
): GapFillSuggestion | null {
if (!latestImportedDate) return null
const d = new Date(latestImportedDate + 'T00:00:00Z')
if (!Number.isFinite(d.getTime())) return null
d.setUTCDate(d.getUTCDate() - GAP_FILL_OVERLAP_DAYS)
let start = d.toISOString().split('T')[0]
// The backend clamps every lookback to 365 days. A renewal staler than that
// must show the date the backfill will actually start from, not promise a
// gap it cannot fill (the >90-day helper already points at SIE/file import
// for older history).
const floor = new Date(today.getTime())
floor.setUTCDate(floor.getUTCDate() - 365)
const floorUtc = floor.toISOString().split('T')[0]
if (start < floorUtc) start = floorUtc
// The backend PATCH handler rejects initial_lookback_from_date unless it is
// strictly in the past; today (UTC) is the newest value it accepts. A
// latestImportedDate in the future can only come from bad bank data: clamp
// rather than propagate it.
const todayUtc = today.toISOString().split('T')[0]
return {
latestImportedDate,
suggestedStartDate: start <= todayUtc ? start : todayUtc,
}
}
/**
* Resolve the start of the current fiscal year, preferring the actual
* fiscal_periods row that contains today over the recurring
* fiscal_year_start_month setting.
*
* Issue #917: the recurring setting cannot represent an extended or shortened
* first fiscal year (e.g. 2025-10-01 to 2026-12-31 for a company that later
* runs calendar years), so deriving from it alone returned 2026-01-01 where
* the real start was 2025-10-01. The period row is authoritative when it
* exists; the setting remains the fallback for companies without period rows.
*/
export function resolveFiscalYearStart(
currentPeriodStart: string | null | undefined,
settings: Pick<CompanySettings, 'fiscal_year_start_month' | 'entity_type'> | null | undefined,
today: Date = new Date(),
): string {
return currentPeriodStart || getCurrentFiscalYearStart(settings, today)
}