Commit Graph

5470 Commits

Author SHA1 Message Date
Maurice 1f89a13f92 fix(assistant): send a blind change back only when its arguments do not fit its schema
Sending back every change called before its group was loaded cost a step
even when the model had the arguments right, and a valid change was no
longer held as the card the user decides on. A blind change is now held as
always when its arguments are of the shape the schema asks for, and sent
back with the schema loaded only when one is missing or of the wrong kind,
such as a number or a list written as text (#2967).
2026-11-05 15:05:58 +01:00
Maurice 93a23f0343 refactor: keep the audit follow-ups within the size ratchets
Eight files had grown past the line counts their ratchets hold, so code
moves to where it belongs instead of the baselines going up:

- a settle-up transfer is booked by applyTransfer in the money helpers
- whether an expense still carries its derived category is
  isDerivedCostCategory in @trek/shared, next to the category map
- a list is deleted with the images it named by deleteWithImages
- the explore POIs and their world-copy tracking are PoiMarkers
- the journey delete tests get a file of their own, and two stale
  migration-era comments in the journey service shrink to one line each
2026-11-05 14:36:25 +01:00
Maurice ce69ae99a1 chore: keep the audit follow-ups within the lint and format ratchets
The bringers list moves next to the other bringer helpers so the card file
exports components only, three tests read without non-null assertions or
any, and three files take the formatting their checks expect.
2026-11-05 14:36:24 +01:00
Maurice c7d5fc12c6 fix(packing): show who brings a shared item in a card with their pictures, and the others as a +N circle
The row's badges said who brings an item, who it is shared with and who
takes care of it in the browser's own tooltip. They now use TREK's, and
the bringers one lists everyone with their picture on a line of their
own. The others show as a +N circle tucked under the first picture.
Tooltip takes an optional content for tips that hold more than words.
2026-11-05 14:36:24 +01:00
Maurice b9c53d692b fix(map): draw points of interest on the world copy the view is on
POIs fetched for a view west of the trip frame's seam were drawn one world
away, off screen.
2026-11-05 14:36:24 +01:00
Maurice 13376d560f fix(settings): let road trip tiles, medium dates and poll deadlines follow the date format
Road trip day tiles still wrote numeric dates from the clock setting and
dates in the medium style kept the language's numbers. A poll deadline
whose time was cleared was read as UTC midnight.
2026-11-05 14:36:24 +01:00
Maurice 3a8191fc07 fix(costs): keep display-currency payments settled, count them as typed in the plan and the PDF, and file every booking type under its Costs key
A display-currency payment whose halves pointed opposite ways no longer
stayed settled (#1445), the plan sidebar and the PDF total still went
through the trip currency, and a zero-decimal display currency was not
rounded. Every booking type now maps to one Costs key on every path; a
cost the old map filed under 'other' still moves with a type change.
2026-11-05 14:36:23 +01:00
Maurice c0f7f81e23 fix(tours): offer waypoints as places only where a trip takes them, and say when every point was there already
The waypoints switch showed in every Tours import dialog; it now shows in
the trip's import only. Confirming it when every point was already in the
trip closed the dialog without a word.
2026-11-05 14:36:23 +01:00
Maurice 7743e4da16 fix(managed): read the wanderer as off on managed, keep the replica alerts mutable, and drop dead guards 2026-11-05 14:36:23 +01:00
Maurice 887f5d8c84 fix(deps): restore the vite 8.3.5 security floor
Re-applying the reverted dependency update left vite below the patched
version.
2026-11-05 14:36:23 +01:00
Maurice 2ec1199490 fix(place-contrib): re-check consent inside the catch-up step and never share withdrawn history again
The catch-up read its state outside its transaction and could revive a
backfill a withdraw had just ended; switching a field off and on after a
withdrawal shared the withdrawn history again. The catch-up now resolves
the country the way a save does and reads one field map.
2026-11-05 14:36:22 +01:00
Maurice 9dd58fe32f fix(storage): release a deleted journey's photos and a deleted list's images, and keep one bad key from stopping a category
Deleting a journey left its uploaded photos behind for good, and deleting
a list queued neither its cover nor its place images. An object with a key
the store refuses no longer fails every night's pass, and the admin view
keeps the counts of a category skipped for a running backup.
2026-11-05 14:36:22 +01:00
Maurice 402a02a302 fix(docker): say what the volume-root check leaves alone, and skip the walk for a custom fsGroup
The entrypoint comment, the Helm chart README, the wiki and the ssrfGuard
note now describe what the shipped check does.
2026-11-05 14:36:22 +01:00
Maurice 7b5a05afee fix(planner): keep timed stops in their local zone when flights take their clock slots
The zoned flights traded clock times among themselves while stops kept
theirs, so a dinner timed after a flight landed could be drawn before it.
A timed stop near one of the day's zoned endpoints now joins the swap in
that endpoint's zone.
2026-11-05 14:36:22 +01:00
Maurice c3357ebea4 fix(assistant): send a blind write back with its schema, pair unrun calls with their answers, and give the wordless-answer recovery its own nudge
A write from a group the turn had not loaded was held as a change card with
schema-less arguments and failed on apply (#2967); it now goes back to the
model with the group loaded. Calls answered as not run are now listed on
the assistant turn too, so a thread replayed on another provider has no
orphan tool results. The blind-call hint only goes with the blind call.
2026-11-05 14:36:21 +01:00
Maurice 91f6430762 test(collections): read a row's +N without waiting for one 2026-11-05 14:18:04 +01:00
Maurice 6ee540d5e4 test(collections): hover the +N of a row that has a category beside its labels 2026-11-05 13:53:57 +01:00
Maurice 99cdcfcd09 test(collections): cover fitting a row's labels and folding the map's label strip 2026-11-05 13:31:25 +01:00
Maurice fd11944ed6 test(collections): count the labels a row shows, and give the three-viewport spec room to clean up 2026-11-05 13:29:27 +01:00
Maurice 82c6a3cd5f fix(collections): name a hidden label's class without a leading space prettier strips 2026-11-05 13:10:15 +01:00
Maurice fbafa270c4 fix(collections): fit a row's labels without setting state in an effect, and bring them back when the row grows
The fitting ran in a layout effect and measured chips the row had already
squeezed, so a row that got wider kept its "+N". Every label now stays in
the row, the ones without room kept out of sight, and the observer that
sees the row change size fits them from each chip's full width. The name
takes the width it needs on a wide list and the labels fill the rest.
2026-11-05 13:01:49 +01:00
Maurice 1c54d7d6c1 fix(collections): give place rows two fixed lines and fit labels to the room
The first fix let a row wrap freely, which on a list beside the map put
the status on a line of its own and broke every row differently. A row is
now laid out the same way each time: on a list narrower than 760px the
name, address and status take the first line and the category and labels
one line below; a wider list keeps them on one line. The row shows as
many labels as it has room for, category first, and a "+N" chip names
the rest in its tooltip, so a wide screen shows every label. Chips that
cut a name short say it in full in their tooltip.

The label strip over the map hid the labels that did not fit behind a
scroll with no scrollbar. It now folds into a "Labels" button that opens
them as a list when the bar has no room for all of them, and unfolds
again when the room is back. The filter row stays on one line on a
desktop, a dropdown at its default keeping only its icon when narrow, and
the select bar's actions keep their icons with the words in a tooltip
when the bar is too narrow for them.
2026-11-05 12:46:16 +01:00
Maurice d2c7d83256 fix(collections): keep a place's name readable next to long labels
A place row was a three-column grid whose last column held the labels,
the category and the status, none of which could shrink. With two or
three long labels the name got what was left, often one letter or
nothing, and the status pill ran out of the card.

The row now wraps: the name keeps at least 11rem, and when the labels,
category and status do not fit beside it they go to a line of their own
under it and wrap there. A label or category too long for its chip ends
in an ellipsis instead of being cut mid-letter. Kept tours use the same
row and get the same layout.

Closes #2981
2026-11-05 12:33:05 +01:00
Maurice 91630dd811 fix(planner): let a timed booking stay where it is dropped when only other bookings stand against its time
A booking dropped against another booking's time went back to the nearest
slot its own time allowed. With a 06:30 flight listed below the day's
untimed stops, every slot among them was against the flight, so a 13:00
transit dropped between two stops went back behind the flight on every
drop, and the flight's legs cannot be dragged out of the way.

The dropped row now stays where it was dropped and the other timed
bookings it puts out of order move next to it, the earlier ones ahead of
it and the later ones behind. Timed stops and notes never move: against
one of those the row is still seated by its time as before (#2631). The
phone's move applies the same rule before it refuses a move.

Closes #2983
2026-11-05 12:15:53 +01:00
Isaias Tavares 944ab3d29a docs(db): drop a historical note from the reservation join row docstring
"unchanged from the legacy split" describes how the code got here rather
than what it does, and the fold it names is stated in the same sentence.
2026-11-05 01:47:28 -03:00
Isaias Tavares dc4972977c test(probe): pin the presence label the sampler matches
The probe samples a presenceSet pair's flag as true by matching the
tuple label "present" by name, so a rename in the repository helper
would silently stop every partial update from reaching its SQL. The
label is now exported from arg-samples.ts, and a probe test reads the
real presence-set.ts and fails if no named tuple member carries it.
2026-11-05 01:47:28 -03:00
Isaias Tavares 2318308674 fix(db): fail plugin-data erasure on errors other than a missing table
deletePluginUserData swallowed every error from its three DELETEs, so a
lost connection or a refused statement on Postgres looked like success
and left the user's plugin rows behind. It now skips only a table that
does not exist, the case its comment described (a slimmed-down schema),
and rethrows anything else. The missing-table test moves to
isMissingTableError in migration-utils, which already decided this for
the migrations: the ORM maps SQLite's "no such table" and Postgres's
42P01 to TableNotFoundException. The docstring keeps the intent and drops
the plan history.
2026-11-05 01:47:28 -03:00
Isaias Tavares 1551f6d474 refactor(db): group duplicate emails by pushing into each group
The duplicate-email report rebuilt each group's array with a spread on
every row. Pushing into the group that the map already holds says the
same thing without the copies.
2026-11-05 01:47:28 -03:00
Isaias Tavares 783ad5c0f1 refactor(places): read the list's category and tag ids through one helper
The list filter turned its two query-string ids with a destructured
[...].map(), and the null-means-no-match convention (a non-canonical id
matches no place) was carried by a comment and two checks on the
result. placeListIdFilters in places.helpers.ts owns that convention and
returns null when the list must be empty, so the service reads as three
lines. The MCP tool passes the same strings to the same service method,
so it is unchanged.
2026-11-05 01:47:28 -03:00
Isaias Tavares c0bae97a14 refactor(db): comment the qualified upsert targets the same way
RouteUsageDaily and GoogleApiUsage both qualify the target row's columns in
ON CONFLICT ... DO UPDATE, because Postgres reads a bare name there as
ambiguous between the target and excluded (42702). Both now carry the same
comment above the upsert, and GoogleApiUsage's one-column chain sits on one
line like its neighbours. No change to the statements.
2026-11-05 01:47:28 -03:00
Isaias Tavares e82968d66c refactor(db): name the reservation-to-stay join condition once
The TEXT-against-INTEGER join between reservations.accommodation_id and
day_accommodations.id was spelled out eight times across Reservations and
DayAccommodations, each with its own cast and comment. accommodationIdMatches
in _shared/ is the one place that spells it, and it documents why the cast is
there: SQLite coerces the bare comparison and Postgres refuses it. The call
sites now read as what they test. The SQL is unchanged.

The size baseline is lowered to the new line counts.
2026-11-05 01:47:28 -03:00
Isaias Tavares 98533f3dbe refactor(db): type the database settings and inject one engine
The settings for the database were one flat record whose url and pool size
were meaningful only for Postgres, so every reader had to re-check the URL.
They are now a discriminated union per engine (DbEngineSettings): the
Postgres arm carries its URL and a pool size already capped at the ceiling,
and the SQLite arm carries nothing else. The env schema uses the same ceiling
constant, so the boot check and the derivation cannot drift apart.

The engine was resolved from the environment by every caller that needed it
(the lifecycle, the unit of work, the backup adapters), each building its own
instance. The engine is now provided once per container by
DatabaseEngineModule and injected everywhere, so the state an engine keeps
(Postgres's reopen hook) is the same wherever it is used. resolveEngine stays
for the MikroORM CLI config and the DI factory, and nothing else calls it.

The engine's reopen no longer carries a `reopenSwapsConnection` flag. The
engine itself decides what a reopen needs: SQLite closes and reconnects the
ORM's connection before the schema bootstrap (its handle is swapped under
Kysely's cache), and Postgres only migrates, because a restore replaces the
contents of the database the pool already reaches. The Postgres pool is never
closed on a reopen, since ending it waits for connections lent out during the
restore.

The backup ports answer every method the same way. verify() now throws a
DomainError for a refused archive on both engines, and the restore path maps
it back to the same { success: false, error, status } result, so the HTTP
response of a refused restore (status and body) is unchanged. The restore
step that swaps the database and then always reopens it is shared in
nest/backup/database-swap.ts, with the error-description helper in
nest/common, instead of two copies that could drift.
2026-11-05 01:04:21 -03:00
Isaias Tavares 1c3dc93716 refactor(db): keep the SQLite-only helpers and the Postgres number parsers explicit
createStandaloneOrm builds the SQLite schema snapshot, so it now refuses a
config that names another driver instead of silently building a snapshot of
the wrong engine. runSchemaBootstrap takes its snapshot flag as a required
argument: the reopen after a restore is the only caller that skips the copy,
and an omitted flag should not be able to pick a default for it.

The Postgres driver's number parsers say which values they refuse: NaN and
the infinities are reported as not finite instead of as beyond 2^53, which
was the wrong reason. The driver's header comment is folded into the constant
docs it described.
2026-11-05 01:02:04 -03:00
Isaias Tavares b8c9184dff test(db): move the backup, boot-restore and bootstrap tests onto the fake-tool helper
The backup, boot-restore and bootstrap tests stopped hand-writing their
stand-in tools and use tests/helpers/fake-pg-tools.ts, so one definition of
what a stand-in logs and answers covers all four files. The assertions about
adapter behaviour are unchanged.

PGBACKUP-002's refusals become it.each, each keeping its own message check:
a missing archive keeps its exact status test (PGBACKUP-002a), and the
accepted dump is PGBACKUP-002. Five refusals are covered in all: missing,
SQLite, newer, no history, and not a dump; the prompt counted six.

boot-restore hands its tables to the psql stand-in directly rather than
through a file the test writes after the plan is built. The bootstrap test
moves its Logger spy into beforeEach.
2026-11-05 00:42:11 -03:00
Isaias Tavares f3e6480606 test(db): drive the Postgres tool tests from one fake-tool helper
Four test files hand-wrote the stand-in pg_dump, pg_restore and psql
scripts, each in its own shape, and read their arguments by position ($4,
$6, args[3], slice(0, 5)). A change to what the tools are handed would have
meant finding every one of them.

tests/helpers/fake-pg-tools.ts installs the stand-ins: each logs its call,
and what it does is a named snippet (answersSchema, answersTables,
writesFile, writesRestrictedScript, listsMigrations, failsWith, touches,
copiesFileEndingWith). The output file is found by its --file= argument.
readToolCall and flagValues read a call back by name.

pg-tools.test.ts moves onto it. Argument checks use arrayContaining and
named flags; order is asserted only where it is the subject (the clear
script runs before the load, `--` precedes the dump path). PGTOOLS-007 and
PGTOOLS-009 split their scenarios into it.each, keeping their IDs as
suffixes.
2026-11-05 00:42:11 -03:00
Isaias Tavares 5503db4883 refactor(db): generate the Postgres baseline in one pass, check it in two modes
--check migrated a temporary SQLite database twice: once inside the
generation and again to read the triggers for the port check. The generation
now returns the triggers it already read, so a check is one migration.

main is split into writeMode and checkMode, and the frozen-or-not decision
lives in one place, checkBaseline. A baseline whose `through` is the newest
migration must match the entities and port every SQLite trigger; a baseline
frozen behind a newer migration is only checked by its manifest. The old
code made the same decision twice, once for the match and once for the
triggers, and the two could drift apart.

The migrated-database read is split into listing the objects and reading
them, so the temporary file is removed in one place.

Manifest parsing uses a type guard, isBaselineManifest, instead of casts
after a check, and PGBASE-004 pins what it refuses.
2026-11-05 00:36:42 -03:00
Isaias Tavares 33a18f29dd refactor(db): split the Postgres schema adjustments by concern
schema-adjust.ts mixed the statement splitter, the trigger ports and the
column adjustments, and every caller repeated `meta.props as EntityProperty[]
as ColumnMeta[]` to reach them. Now:

- sql-split.ts holds splitStatements. Its scope is written down: `;` outside
  quotes, '/" quoted text and the $$ body, and the session-setting lines it
  drops. A tagged dollar quote ($tag$) is refused with an error instead of
  being split inside, which the splitter could not do correctly.
- postgres-triggers.ts holds the hand-written trigger ports, which the
  baseline and the parity test read by name.
- schema-adjust.ts holds the column adjustments. applyPostgresColumnAdjustments
  (renamed from adjustColumns, which hid that it mutates its input) takes an
  entity's props directly: ColumnMeta now reads customType as unknown, so
  the entity's property metadata satisfies it without a cast. Each adjustment
  is a small function that returns whether it changed the column.

The timestamp check is now `instanceof DbTimestampType` rather than a compare
on the constructor's name, which would have broken silently on a rename.
2026-11-05 00:32:18 -03:00
Isaias Tavares d498fe2bea refactor(db): read the Postgres bootstrap as inspect, create or verify, migrate
bootstrapPostgresSchema did three jobs in one 35-line body, and the safety
net's label came from an `as string` cast on the last executed migration.

It now reads in three steps: inspect what the database holds, then create
the schema from the baseline (createFromBaseline) or verify the recorded
history against this build (refuseNewerHistory), then run the pending
migrations. The label is read from the list with `.at(-1)` and only when
there is one: a database the baseline just created has no history to label
its copy by, which is the case the old `created` flag tested.

The deps keep baseline as a thunk on purpose: the committed baseline is read
only on the create path, so a database already at its head never touches it.
2026-11-05 00:23:08 -03:00
Isaias Tavares 0e93df378f refactor(db): define the "TREK's tables" catalog query once
The bootstrap's currentTables and the client tools' pgTableNames each wrote
the same information_schema query for the current schema, so a change to
which tables count as TREK's would have had to land twice. It is now
CURRENT_SCHEMA_TABLES_SQL in dialect/catalog.ts, beside the other catalog
statements, and both read it.

pgTableNames now lists its names in name order too, as the bootstrap
already did. Nothing reads that order for meaning; the boot-restore refusal
names the first few tables, and that list was already sorted.
2026-11-05 00:21:55 -03:00
Isaias Tavares 1c73ac4c35 refactor(db): split the Postgres client-tool module by concern
pg-tools.ts held six concerns in 366 lines: tool resolution and process
running, the libpq connection environment, the CVE-2025-8714 version policy,
the CLEAR-SCHEMA SQL and the operations themselves. Each now has its own
module, and pg-tools.ts keeps only the operations the backup and the boot
restore call, so their import paths do not change.

- pg-process.ts: PgTool, resolvePgTool, pgConnectionEnv, PgToolMissingError
  and run (stderr is summarized there, as before).
- restore-safety.ts: hasRestrictFix and the refusal. The minimum versions in
  the error are built from the RESTRICT_SINCE table, so the message cannot
  drift from the policy.
- clear-schema.ts: the SQL that empties the schema before a load, unchanged.

The version floor also no longer keeps its own OLDEST_MAJOR and
LAST_MAJOR_WITHOUT_IT constants: both are read from the table. Behaviour is
unchanged. The literal /dev/null is os.devNull, and the COPY parsing of a
dump's migration history is its own function.
2026-11-05 00:20:22 -03:00
Isaias Tavares ae005fc582 fix(db): check the restore script for TREK's tables line by line
pgRestore read the whole pg_restore --file script into one string to find
TREK's migration table before clearing the schema. The script carries every
COPY row of the dump, so a large database made the read exceed V8's maximum
string length before the check ran, and the restore failed with an
unhelpful error instead of loading or refusing.

The check now streams the script with readline and stops at the first
matching line. The match itself is unchanged: the same CREATE TABLE line for
the same quoted schema, so a dump without TREK's history is still refused
before anything is cleared. The regex's escaping is a named helper.

PGTOOLS-005b pins that the script is never read whole.
2026-11-05 00:13:04 -03:00
Isaias Tavares 293cdbc353 fix(assistant): record an unrun change as failed before the stream ends
A decision whose change never ran was written as failed only after the response stream had closed, so a client reading the thread again at once could still find it approved. On the in-memory SQLite database the write won the race; the Postgres lane lost it (ASTE2E-014). The failure is now recorded before the close, in both decision paths, and the stream still closes when that write fails. ASTE2E-026 slows the write down to pin it.
2026-11-04 23:56:09 -03:00
Isaias Tavares 63a2c9db00 docs(db): drop a migration count and a planning reference from comments
The baseline's count went stale with the first migration after the cut, and the lane test pointed at a document outside the repository.
2026-11-04 23:40:13 -03:00
Isaias Tavares e848f343b5 refactor(db): keep the reservations join in its repository
The joined read moved out to reservation-join.ts as a free function, unlike the other files beside the repositories, which hold row types and leave the statements to the repository. joinedQuery() is back in ReservationsRepository with the cast; the row types it answers with move to reservation-rows.ts, which depends on nothing but the Kysely tables.
2026-11-04 23:31:03 -03:00
Isaias Tavares 4325c145ce feat(db): cut the Postgres baseline at the newest migration
The two migrations that landed on dev since the cut were written for SQLite: the longitude wrap subtracts a comparison from an integer, which Postgres refuses. No Postgres database exists yet, so the baseline now stands for both, and a new Postgres install starts with Journey and Collections on, as a new SQLite one does.
2026-11-04 22:54:14 -03:00
Isaias Tavares 53bd8505cd fix(db): read the current schema without SQLite's concatenation spelling
lint:dialect counts every || in src/db as SQLite-only, the psql query that names TREK's schema included. concat_ws gives the same tab-separated line.
2026-11-04 22:54:14 -03:00
Isaias Tavares 298f0409d5 docs(db): record the Supabase checks in the Postgres page
The checklist ran against a Supabase project over the direct connection. The page now says what it found: the CA to verify the server with, the IPv6-only direct host, the postgres role creating the collation, backups holding only TREK's schema, and the latency of a distant region.
2026-11-04 22:46:18 -03:00
Isaias Tavares e0013bc077 docs(db): document running TREK on Postgres, with compose and Helm support
- wiki/Database-Postgres.md: requirements (PostgreSQL 14 to 17, an empty
  database of its own, a role that is not a superuser, ICU), configuration
  and SSL, first boot, upgrades and the .pgdump copy, backups and restore,
  limitations (no SQLite to Postgres move yet, one replica), notes for
  Supabase, and the manual checklist still to run against it.
- docker-compose.postgres.yml: TREK beside a postgres:17 service.
- Helm: database.url (written into the chart's Secret) or
  database.existingSecret supplies TREK_DB_URL; replicas stay at one, now
  for the state TREK keeps in memory.
- Environment variables, backups, updating, troubleshooting, the Docker
  Compose and Helm install pages, the sidebar and the README point at it.
2026-11-04 22:45:32 -03:00
Isaias Tavares 526437926a test(db): run the suites that boot the app on Postgres in their own lane
npm run test:pg --workspace=server (vitest.pg.config.ts) runs every suite
that boots the app through buildApp() on the real database module against
Postgres, each test file on its own copy of one template database the
engine's bootstrap migrates once per run. TREK_TEST_DB_URL names the server;
without it the lane fails instead of passing on nothing. The CI job
'Server: Postgres Tests' runs it on a postgres:17 service and is part of
ci-ok.

tests/postgres/ holds what only Postgres needs proving:
- PARITY-PG: MikroORM's comparator finds nothing between the bootstrapped
  database and the entities. What it reports that changes nothing (key
  columns it would make nullable, defaults and checks Postgres spells its
  own way, the hand-ported triggers) is proven so: the spelling statements
  run in a rolled-back transaction and every default and check must read
  the same after, and every ported trigger must exist.
- the write lock: concurrent read-then-insert units of work never see the
  same count, and counts are numbers; a search ignores case.

tours-hike and tours-plan-types stay on SQLite only: they move a faked
setTimeout on until a tool call ends, which a database over the network
does not finish within.
2026-11-04 22:45:32 -03:00
Isaias Tavares 3f1a786313 fix(db): keep the Postgres pool across a restore
The reopen after a restore closed the ORM's pool and waited for it to end.
Ending a pg pool waits for every connection it has lent out, and one lent
to a job while the restore ran was never handed back, so the restore
request never returned and every later query failed on a destroyed driver.
A Postgres restore replaces the contents of the same database, so the pool
now stays and the reopen only migrates; SQLite still reconnects around its
swapped file.
2026-11-04 22:45:31 -03:00