Files
paliad/internal/db/migrations/140_drop_deadline_rules.up.sql
mAi 1129baba7a
Some checks failed
Paliad CI gate / build (push) Has been cancelled
Paliad CI gate / test-go (push) Has been cancelled
Paliad CI gate / deploy (push) Has been cancelled
feat(db,services): Slice B.4 destructive drop — paliad.deadline_rules retired, INSTEAD OF triggers on view route writes (mig 140, t-paliad-305 / m/paliad#93)
Drops the legacy paliad.deadline_rules table after 3 weeks of dual-write
shadowing (mig 136 → B.2 dual-write → B.3 read cutover via view). The
new tables — paliad.procedural_events, paliad.sequencing_rules,
paliad.legal_sources — are the sole source of truth from this commit
forward.

Pre-flip drift verified clean against prod:
  deadline_rules=231 == sequencing_rules=231 == procedural_events=231
  legal_sources=87
  missing_sr=0, orphaned_sr=0, mismatched_lifecycle=0

* internal/db/migrations/140_drop_deadline_rules.up.sql (new) —
  Single TX, audit-first:
  1. CREATE TABLE paliad.deadline_rules_pre_140 AS TABLE paliad.deadline_rules
     (precedent migs 091/093/095/098 — snapshot in same TX as destructive op).
  2. Final reconciliation UPDATE on paliad.deadlines (no-op when
     drift is already 0; defensive against last-minute writes).
  3. DROP TRIGGER deadline_rules_audit_aiud.
  4. Re-point FKs to sequencing_rules:
     - paliad.appointments.deadline_rule_id → paliad.sequencing_rules(id)
     - paliad.deadline_rule_backfill_orphans.resolved_rule_id → paliad.sequencing_rules(id)
     (the id values are identical — sr.id inherited dr.id at mig 136.)
  5. DROP COLUMN paliad.deadlines.rule_id.
  6. DROP TABLE paliad.deadline_rules.
  7. CREATE INSTEAD OF INSERT + INSTEAD OF UPDATE triggers on
     paliad.deadline_rules_unified. Triggers route writes into the
     three new tables in the same TX, preserving the legacy column
     shape on the wire so RuleEditorService SQL only needs a
     table-name swap, not a structural rewrite. Synthetic-code mint
     expression is byte-identical to mig 136 + the B.2 dual-write
     helper. POST assertions confirm the table is gone, the column
     is gone, and the snapshot matches.

  Trigger design notes (1:N caveat documented in-trigger):
  - PE identity columns (code, name, name_en, description, event_kind,
    primary_party_default, legal_source_id, concept_id) mirror from
    the writing sequencing-rule.
  - PE lifecycle columns (lifecycle_state, published_at, is_active)
    deliberately do NOT mirror — a draft sequencing-rule cloned from
    a published source shares the source's PE; we don't want the
    clone's 'draft' lifecycle to leak back onto the source's PE.
    Practical bound today (1:1 corpus); explicit comment in-trigger
    for the eventual 1:N pattern.

* internal/db/migrations/140_drop_deadline_rules.down.sql (new) —
  Best-effort restore from the snapshot. Triggers / indexes /
  CHECK constraints from historical migrations are NOT replayed;
  operator must reapply 078/079/091/095/098/122/128/134/135 to
  bring the restored table to working shape. The down path is for
  catastrophic recovery, not casual revert.

* internal/services/rule_editor_service.go —
  Six syncDualWriteFromDeadlineRule(...) calls removed (the
  INSTEAD OF triggers now do the fan-out). Five
  INSERT/UPDATE paliad.deadline_rules statements (Create,
  UpdateDraft, CloneAsDraft INSERT+SELECT, Publish, peer-archive,
  flipLifecycle) renamed to paliad.deadline_rules_unified —
  trigger handles the routing.

* internal/services/rule_editor_orphans.go — ResolveOrphan no
  longer writes deadlines.rule_id (column dropped). Sets
  sequencing_rule_id directly + derives procedural_event_id from
  the matching sequencing_rules row in the same UPDATE statement.

* internal/services/deadline_service.go — deadlineColumns now
  lists sequencing_rule_id (Deadline.RuleID still binds to it via
  the db tag rename below). Update path's appendSet("rule_id",…)
  flipped to appendSet("sequencing_rule_id",…) and post-write
  derivation moved to the renamed syncDeadlineProceduralEventID
  helper.

* internal/services/projection_service.go,
  internal/services/submission_vars.go — `WHERE rule_id = $X`
  reads on paliad.deadlines flipped to sequencing_rule_id.

* internal/models/models.go — Deadline.RuleID db tag changed from
  "rule_id" to "sequencing_rule_id". Field name + JSON name kept
  for backward compat with the frontend and existing Go callers;
  semantic value is identical (same UUID).

* internal/services/dual_write.go — Massively trimmed.
  Removed: syncDualWriteFromDeadlineRule, syncDeadlineDualLinks,
  CheckDualWriteDrift, DualWriteDriftReport, HasDrift,
  StartDualWriteDriftCheckLoop. All referenced
  paliad.deadline_rules which no longer exists.
  Kept (renamed): syncDeadlineProceduralEventID — derives
  procedural_event_id from sequencing_rule_id after any
  DeadlineService.Update that touched the back-link.

* cmd/server/main.go — Removed the StartDualWriteDriftCheckLoop
  bootstrap call (and its `time` import that only that call
  needed). Comment notes the retirement.

* internal/services/dual_write_test.go — Removed the final
  CheckDualWriteDrift assertion in
  TestDualWrite_RuleEditorLifecycle (function deleted). The
  per-step asserts against procedural_events / sequencing_rules
  / legal_sources cover the same contract by direct query.

Hard rules followed:
- Audit-first: snapshot precedes destructive ops in the same TX.
- No silent data loss: pre-drop drift was zero; snapshot captures
  the final state; FK re-points use identical UUIDs.
- INSTEAD OF triggers documented in mig 140 — single source of
  truth for the legacy→new mapping.
- Down migration is honest about its scope (catastrophic recovery
  only).

Build + vet clean. TestMigrations_NoDuplicateSlot passes. Live-DB
tests skipped (no TEST_DATABASE_URL in this env) — they'll exercise
the full mig 140 + INSTEAD OF triggers in CI.
2026-05-26 19:53:24 +02:00

335 lines
16 KiB
PL/PgSQL

-- 140_drop_deadline_rules — Slice B.4 destructive drop (t-paliad-305 / m/paliad#93)
--
-- HARD STOPS:
-- * Audit-first: snapshot paliad.deadline_rules → paliad.deadline_rules_pre_140
-- in the SAME TRANSACTION as the DROP, per m's snapshot policy
-- (precedent migs 091/093/095/098). The whole .up.sql runs inside a
-- single transaction because the migration runner wraps it; if any
-- statement fails, the snapshot CREATE TABLE rolls back with the
-- destructive DROP.
-- * No data loss: paliad.deadline_rules has been a write-side shadow
-- since B.3 (B.2 dual-write keeps sequencing_rules + procedural_events
-- + legal_sources current). Drift verified clean before this slice
-- (deadline_rules=231, sequencing_rules=231, 0 mismatches across
-- counts/FKs/lifecycle/is_active).
--
-- What this migration does:
-- 1. Snapshot deadline_rules → deadline_rules_pre_140 (preserves audit
-- trail of the table's final state for forensic + revert paths).
-- 2. Final reconciliation: catch any deadlines whose
-- sequencing_rule_id/procedural_event_id columns drifted from the
-- legacy rule_id (no live drift today — defensive).
-- 3. Drop the audit trigger on deadline_rules (it can't fire on a
-- gone table; the trigger function itself stays for the historical
-- paliad.deadline_rule_audit reads).
-- 4. Re-point FKs that currently target deadline_rules.id over to
-- sequencing_rules.id. The id values are identical (sequencing_rules
-- inherited deadline_rules.id during mig 136 backfill), so no data
-- migration is needed — just the constraint swap. Affects:
-- - paliad.appointments.deadline_rule_id
-- - paliad.deadline_rule_backfill_orphans.resolved_rule_id
-- 5. Drop paliad.deadlines.rule_id column. Per design §5.4 step 16:
-- "DROP COLUMN paliad.deadlines.rule_id (keep rule_code +
-- custom_rule_text as the human-readable denormalized columns —
-- they're the safety net for orphaned deadlines per t-paliad-258)."
-- The new sequencing_rule_id + procedural_event_id columns from
-- mig 136 are the FK back-links from B.4 forward.
-- 6. DROP TABLE paliad.deadline_rules.
-- 7. INSTEAD OF triggers on paliad.deadline_rules_unified that route
-- INSERTs/UPDATEs to the underlying sr+pe+ls tables. Lets the
-- RuleEditorService keep its existing SQL shape (one INSERT, one
-- UPDATE per write method) with only a table-name swap. The
-- triggers project the legacy column shape back to the three new
-- tables exactly as the dual-write helper did in B.2.
--
-- Down: best-effort restore from the snapshot. The original triggers,
-- indexes, and FKs are NOT recreated — operator must replay historical
-- migrations 078/079/091/095/098/122 to bring the table back to a
-- working shape. The down path is for catastrophic recovery, not casual
-- revert.
-- ---------------------------------------------------------------
-- 1. Snapshot — must precede the destructive ops (same TX).
-- ---------------------------------------------------------------
CREATE TABLE paliad.deadline_rules_pre_140 AS TABLE paliad.deadline_rules;
COMMENT ON TABLE paliad.deadline_rules_pre_140 IS
'Snapshot of paliad.deadline_rules taken in mig 140 (Slice B.4, '
't-paliad-305) before the destructive DROP. Mirrors precedent '
'pre_091/093/095/098. Read-only forensic + revert source.';
-- ---------------------------------------------------------------
-- 2. Final reconciliation — should be a no-op (drift was 0 going
-- into this slice). Belt-and-braces against a write that snuck
-- in between drift-check and this migration.
-- ---------------------------------------------------------------
UPDATE paliad.deadlines d
SET sequencing_rule_id = d.rule_id,
procedural_event_id = sr.procedural_event_id
FROM paliad.sequencing_rules sr
WHERE sr.id = d.rule_id
AND d.rule_id IS NOT NULL
AND (d.sequencing_rule_id IS DISTINCT FROM d.rule_id
OR d.procedural_event_id IS DISTINCT FROM sr.procedural_event_id);
-- ---------------------------------------------------------------
-- 3. Drop the deadline_rules audit trigger. The trigger function
-- (paliad.deadline_rule_audit_trigger) stays defined for any
-- historical references; mig 079 created it.
-- ---------------------------------------------------------------
DROP TRIGGER IF EXISTS deadline_rules_audit_aiud ON paliad.deadline_rules;
-- ---------------------------------------------------------------
-- 4. Re-point FKs from deadline_rules → sequencing_rules.
-- ---------------------------------------------------------------
ALTER TABLE paliad.appointments
DROP CONSTRAINT IF EXISTS appointments_deadline_rule_id_fkey;
ALTER TABLE paliad.appointments
ADD CONSTRAINT appointments_deadline_rule_id_fkey
FOREIGN KEY (deadline_rule_id) REFERENCES paliad.sequencing_rules(id);
ALTER TABLE paliad.deadline_rule_backfill_orphans
DROP CONSTRAINT IF EXISTS deadline_rule_backfill_orphans_resolved_rule_id_fkey;
ALTER TABLE paliad.deadline_rule_backfill_orphans
ADD CONSTRAINT deadline_rule_backfill_orphans_resolved_rule_id_fkey
FOREIGN KEY (resolved_rule_id) REFERENCES paliad.sequencing_rules(id);
-- Drop the deadlines→deadline_rules FK before we drop the column.
ALTER TABLE paliad.deadlines
DROP CONSTRAINT IF EXISTS fristen_rule_id_fkey;
-- ---------------------------------------------------------------
-- 5. Drop paliad.deadlines.rule_id (column + remaining indexes).
-- ---------------------------------------------------------------
ALTER TABLE paliad.deadlines
DROP COLUMN IF EXISTS rule_id;
-- ---------------------------------------------------------------
-- 6. DROP TABLE paliad.deadline_rules. Now that:
-- - dependent FKs are re-pointed to sequencing_rules,
-- - the audit trigger is dropped,
-- - deadlines.rule_id is gone,
-- nothing references the table anymore. The self-FKs
-- (deadline_rules.parent_id, .draft_of) drop with the table.
-- ---------------------------------------------------------------
DROP TABLE paliad.deadline_rules;
-- ---------------------------------------------------------------
-- 7. INSTEAD OF triggers on the view — routes writes to sr+pe+ls.
-- ---------------------------------------------------------------
CREATE OR REPLACE FUNCTION paliad.deadline_rules_unified_insert_trigger()
RETURNS TRIGGER LANGUAGE plpgsql AS $fn$
DECLARE
v_legal_source_id uuid;
v_pe_id uuid;
v_code text;
BEGIN
-- legal_sources upsert (no-op if NEW.legal_source is NULL)
IF NEW.legal_source IS NOT NULL THEN
INSERT INTO paliad.legal_sources (citation, jurisdiction)
VALUES (NEW.legal_source,
COALESCE(NULLIF(split_part(NEW.legal_source, '.', 1), ''), 'other'))
ON CONFLICT (citation) DO NOTHING;
SELECT id INTO v_legal_source_id
FROM paliad.legal_sources
WHERE citation = NEW.legal_source;
END IF;
-- Mint synthetic code when submission_code is NULL — same recipe
-- as mig 136 + B.2 dual-write helper. Stays byte-identical.
v_code := COALESCE(NEW.submission_code,
'null.' || substring(replace(NEW.id::text, '-', ''), 1, 8));
-- procedural_events upsert. ON CONFLICT (code) deliberately leaves
-- lifecycle_state / published_at / is_active alone — those track
-- the procedural-event concept's own lifecycle, not the inserting
-- sequencing-rule's lifecycle (e.g. a CloneAsDraft of a published
-- rule creates a draft sr that shares the published PE; the PE
-- should stay 'published'). Identity columns DO update so an
-- admin editing a draft's name still flips the lawyer-visible
-- label (1:1 today; revisit when 1:N becomes a real pattern).
INSERT INTO paliad.procedural_events
(code, name, name_en, description, event_kind, primary_party_default,
legal_source_id, concept_id, lifecycle_state, published_at, is_active)
VALUES
(v_code, NEW.name, NEW.name_en, NEW.description, NEW.event_type,
NEW.primary_party, v_legal_source_id, NEW.concept_id,
COALESCE(NEW.lifecycle_state, 'draft'), NEW.published_at,
COALESCE(NEW.is_active, true))
ON CONFLICT (code) DO UPDATE SET
name = EXCLUDED.name,
name_en = EXCLUDED.name_en,
description = EXCLUDED.description,
event_kind = EXCLUDED.event_kind,
primary_party_default = EXCLUDED.primary_party_default,
legal_source_id = EXCLUDED.legal_source_id,
concept_id = EXCLUDED.concept_id,
-- lifecycle_state / published_at / is_active deliberately omitted
updated_at = now()
RETURNING id INTO v_pe_id;
-- sequencing_rules insert. id is the caller-supplied NEW.id so
-- existing FK back-links (deadlines.sequencing_rule_id) resolve.
INSERT INTO paliad.sequencing_rules
(id, procedural_event_id, proceeding_type_id, parent_id, trigger_event_id,
duration_value, duration_unit, timing,
alt_duration_value, alt_duration_unit, alt_rule_code, anchor_alt,
combine_op, condition_expr, primary_party, sequence_order,
is_spawn, spawn_label, spawn_proceeding_type_id,
is_bilateral, is_court_set, priority,
rule_code, rule_codes, deadline_notes, deadline_notes_en,
choices_offered, applies_to_target,
lifecycle_state, draft_of, published_at, is_active,
created_at, updated_at)
VALUES
(NEW.id, v_pe_id, NEW.proceeding_type_id, NEW.parent_id, NEW.trigger_event_id,
COALESCE(NEW.duration_value, 0), COALESCE(NEW.duration_unit, 'months'),
COALESCE(NEW.timing, 'after'),
NEW.alt_duration_value, NEW.alt_duration_unit, NEW.alt_rule_code, NEW.anchor_alt,
NEW.combine_op, NEW.condition_expr, NEW.primary_party,
COALESCE(NEW.sequence_order, 0),
COALESCE(NEW.is_spawn, false), NEW.spawn_label, NEW.spawn_proceeding_type_id,
COALESCE(NEW.is_bilateral, false), COALESCE(NEW.is_court_set, false),
COALESCE(NEW.priority, 'mandatory'),
NEW.rule_code, NEW.rule_codes, NEW.deadline_notes, NEW.deadline_notes_en,
NEW.choices_offered, NEW.applies_to_target,
COALESCE(NEW.lifecycle_state, 'draft'), NEW.draft_of,
NEW.published_at, COALESCE(NEW.is_active, true),
COALESCE(NEW.created_at, now()), COALESCE(NEW.updated_at, now()));
RETURN NEW;
END $fn$;
CREATE TRIGGER deadline_rules_unified_insert
INSTEAD OF INSERT ON paliad.deadline_rules_unified
FOR EACH ROW EXECUTE FUNCTION paliad.deadline_rules_unified_insert_trigger();
CREATE OR REPLACE FUNCTION paliad.deadline_rules_unified_update_trigger()
RETURNS TRIGGER LANGUAGE plpgsql AS $fn$
DECLARE
v_legal_source_id uuid;
v_code text;
BEGIN
-- legal_sources upsert (only if NEW.legal_source is non-NULL).
-- A change FROM non-NULL TO NULL clears legal_source_id on the
-- procedural_event below — same shape as mig 136 / B.2 behaviour.
IF NEW.legal_source IS NOT NULL THEN
INSERT INTO paliad.legal_sources (citation, jurisdiction)
VALUES (NEW.legal_source,
COALESCE(NULLIF(split_part(NEW.legal_source, '.', 1), ''), 'other'))
ON CONFLICT (citation) DO NOTHING;
SELECT id INTO v_legal_source_id
FROM paliad.legal_sources
WHERE citation = NEW.legal_source;
END IF;
v_code := COALESCE(NEW.submission_code,
'null.' || substring(replace(NEW.id::text, '-', ''), 1, 8));
-- Update procedural_events keyed by the existing PE link on
-- sequencing_rules. lifecycle_state / published_at / is_active on
-- PE are NOT mirrored from the per-sequencing-rule UPDATE — see
-- the INSERT trigger comment for the rationale (a draft sr that
-- shares its PE with a published peer must not flip the PE to
-- draft). Identity columns DO mirror so editing name/code from
-- the admin UI continues to reach the lawyer-visible label.
UPDATE paliad.procedural_events
SET code = v_code,
name = NEW.name,
name_en = NEW.name_en,
description = NEW.description,
event_kind = NEW.event_type,
primary_party_default = NEW.primary_party,
legal_source_id = v_legal_source_id,
concept_id = NEW.concept_id,
updated_at = now()
WHERE id = (SELECT procedural_event_id
FROM paliad.sequencing_rules
WHERE id = NEW.id);
-- Update sequencing_rules (1:1 by id).
UPDATE paliad.sequencing_rules
SET proceeding_type_id = NEW.proceeding_type_id,
parent_id = NEW.parent_id,
trigger_event_id = NEW.trigger_event_id,
duration_value = NEW.duration_value,
duration_unit = NEW.duration_unit,
timing = NEW.timing,
alt_duration_value = NEW.alt_duration_value,
alt_duration_unit = NEW.alt_duration_unit,
alt_rule_code = NEW.alt_rule_code,
anchor_alt = NEW.anchor_alt,
combine_op = NEW.combine_op,
condition_expr = NEW.condition_expr,
primary_party = NEW.primary_party,
sequence_order = NEW.sequence_order,
is_spawn = NEW.is_spawn,
spawn_label = NEW.spawn_label,
spawn_proceeding_type_id = NEW.spawn_proceeding_type_id,
is_bilateral = NEW.is_bilateral,
is_court_set = NEW.is_court_set,
priority = NEW.priority,
rule_code = NEW.rule_code,
rule_codes = NEW.rule_codes,
deadline_notes = NEW.deadline_notes,
deadline_notes_en = NEW.deadline_notes_en,
choices_offered = NEW.choices_offered,
applies_to_target = NEW.applies_to_target,
lifecycle_state = NEW.lifecycle_state,
draft_of = NEW.draft_of,
published_at = NEW.published_at,
is_active = NEW.is_active,
updated_at = now()
WHERE id = NEW.id;
RETURN NEW;
END $fn$;
CREATE TRIGGER deadline_rules_unified_update
INSTEAD OF UPDATE ON paliad.deadline_rules_unified
FOR EACH ROW EXECUTE FUNCTION paliad.deadline_rules_unified_update_trigger();
-- ---------------------------------------------------------------
-- 8. POST assertions.
-- ---------------------------------------------------------------
DO $$
DECLARE
v_snapshot_count int;
v_view_count int;
v_dr_table_exists int;
v_rule_id_col int;
BEGIN
SELECT COUNT(*) INTO v_snapshot_count FROM paliad.deadline_rules_pre_140;
SELECT COUNT(*) INTO v_view_count FROM paliad.deadline_rules_unified;
IF v_snapshot_count <> v_view_count THEN
RAISE EXCEPTION '[mig 140] FAILED POST: snapshot has % rows, view has % rows — drift between final state and snapshot',
v_snapshot_count, v_view_count;
END IF;
SELECT COUNT(*) INTO v_dr_table_exists
FROM information_schema.tables
WHERE table_schema = 'paliad' AND table_name = 'deadline_rules';
IF v_dr_table_exists > 0 THEN
RAISE EXCEPTION '[mig 140] FAILED POST: paliad.deadline_rules table still exists after DROP';
END IF;
SELECT COUNT(*) INTO v_rule_id_col
FROM information_schema.columns
WHERE table_schema = 'paliad' AND table_name = 'deadlines' AND column_name = 'rule_id';
IF v_rule_id_col > 0 THEN
RAISE EXCEPTION '[mig 140] FAILED POST: paliad.deadlines.rule_id column still exists after DROP';
END IF;
RAISE NOTICE '[mig 140] OK — deadline_rules dropped, snapshot=% rows, view=% rows, INSTEAD OF triggers active',
v_snapshot_count, v_view_count;
END $$;