feat(db): mig 152 — dedupe identical sequencing_rule clones (5 archived) (t-paliad-321 / m/paliad#144 follow-up)
Mig 151 (t-paliad-319) archived 5 of 6 duplicate procedural_events for
"Mängelbeseitigung / Zahlung" and reparented their sequencing_rules
onto the canonical PE. The 6 sequencing_rules themselves were left
active — and they are byte-for-byte clones (proceeding_type_id=NULL,
rule_code=NULL, duration 14d, primary_party=NULL, condition_expr=NULL,
…). The admin shows six indistinguishable rows for one legal concept.
This migration archives 5 of 6, keeping the row with the
lexicographically lowest UUID as canonical.
Pre-write verification (Supabase MCP, 2026-05-26):
- Exactly 1 clone-group surfaces under the full-signature query
(procedural_event_id, proceeding_type_id, rule_code, duration_*,
primary_party, condition_expr::text, trigger_event_id, alt_*,
anchor_alt, combine_op, parent_id, is_spawn, spawn_*):
6 "Mängelbeseitigung / Zahlung" rows.
- 0 paliad.deadlines reference any of the 5 to-be-archived rows
(verified via deadlines.sequencing_rule_id JOIN; rule_id column
was dropped in mig 140 / Slice B.4).
- Other name-duplicates (Antrag auf Patentänderung×4, Beginn des
Hauptsacheverfahrens×2, Berufungsbegründung-R.220.1×2,
Berufungsschrift-R.220.1×2) do NOT collapse under this signature —
their proceeding_type_id / rule_code / duration / primary_party
differ. Legitimately distinct rules per proceeding. This mig
leaves them alone.
Migration shape (mirrors mig 151):
1. Build dedupe mapping (duplicate_id → canonical_id) into a
ROW_NUMBER() OVER (PARTITION BY full-signature ORDER BY
created_at, id::text) TEMP table.
2. PRE NOTICE: surface every clone-group with its canonical + dups
so the deploy log shows what's about to be touched (m may want
to spot-check).
3. Snapshot the duplicates into paliad.sequencing_rules_pre_152
(precedent pre_091/093/095/098/140/151).
4. Reparent paliad.deadlines.sequencing_rule_id duplicate → canonical
BEFORE archiving (defensive no-op today).
5. set_config('paliad.audit_reason', …) — defensive; sequencing_rules
has no audit trigger yet (mig 151 §scope verified), but a future
trigger would inherit the reason automatically.
6. UPDATE sequencing_rules SET is_active=false,
lifecycle_state='archived' WHERE id IN dups.
7. POST assertions: expected archive count met, zero clone groups
remaining in active+published, zero live deadlines pointing at
an archived sequencing_rule. RAISE EXCEPTION on any mismatch.
Down: best-effort revert (flips archived → published from snapshot).
Doesn't undo the deadlines reparent (live data didn't need one;
snapshot doesn't carry pre-state of deadlines).
Build + vet clean. TestMigrations_NoDuplicateSlot passes.
This commit is contained in:
@@ -0,0 +1,17 @@
|
||||
-- 152_dedupe_identical_sequencing_rule_clones (down) — t-paliad-321
|
||||
--
|
||||
-- Best-effort revert from paliad.sequencing_rules_pre_152. Flips the
|
||||
-- archived rows back to is_active=true / lifecycle_state='published'.
|
||||
-- Does NOT undo the deadlines.sequencing_rule_id reparent — that would
|
||||
-- require remembering the previous pointer per row, which the snapshot
|
||||
-- on sequencing_rules doesn't carry. In live data the reparent was a
|
||||
-- no-op (zero deadlines pointed at duplicates), so this is fine.
|
||||
|
||||
UPDATE paliad.sequencing_rules sr
|
||||
SET is_active = true,
|
||||
lifecycle_state = 'published',
|
||||
updated_at = now()
|
||||
FROM paliad.sequencing_rules_pre_152 snap
|
||||
WHERE sr.id = snap.id;
|
||||
|
||||
DROP TABLE IF EXISTS paliad.sequencing_rules_pre_152;
|
||||
@@ -0,0 +1,240 @@
|
||||
-- 152_dedupe_identical_sequencing_rule_clones — t-paliad-321 / m/paliad#144 follow-up
|
||||
--
|
||||
-- Purpose: mig 151 archived 5 of 6 duplicate procedural_events for
|
||||
-- "Mängelbeseitigung / Zahlung" and reparented their sequencing_rules
|
||||
-- onto the canonical PE. The 6 sequencing_rules themselves remained
|
||||
-- active. Because every one of them is a byte-for-byte clone (same
|
||||
-- proceeding_type_id=NULL, rule_code=NULL, duration 14d, primary_party=NULL,
|
||||
-- everything else NULL, lifecycle_state='published') and only sequence_order
|
||||
-- differs, the admin shows six indistinguishable rows for one legal
|
||||
-- concept. This mig archives 5 of the 6 keeping the lexicographically
|
||||
-- lowest UUID as canonical.
|
||||
--
|
||||
-- Scope verified live before write (Supabase MCP, 2026-05-26):
|
||||
-- * Exactly 1 clone-group surfaces by the full-signature query
|
||||
-- below: 6 "Mängelbeseitigung / Zahlung" sequencing_rules with
|
||||
-- all-NULL discriminators and (duration_value=14, duration_unit='days').
|
||||
-- * 0 paliad.deadlines reference the 5 to-be-archived rows
|
||||
-- (verified via deadlines.sequencing_rule_id JOIN; the column
|
||||
-- formerly named deadlines.rule_id was dropped in mig 140 / B.4).
|
||||
-- * Other name-groups in the live corpus — "Antrag auf
|
||||
-- Patentänderung"×4, "Beginn des Hauptsacheverfahrens"×2,
|
||||
-- "Berufungsbegründung-R.220.1"×2, "Berufungsschrift-R.220.1"×2 —
|
||||
-- do NOT collapse under this signature because their
|
||||
-- proceeding_type_id / rule_code / duration / primary_party
|
||||
-- differ. They are legitimately distinct rules per proceeding;
|
||||
-- this mig leaves them alone.
|
||||
--
|
||||
-- Hard constraints honoured (mirrors mig 151):
|
||||
-- * No deletions. Archived rows flip to is_active=false +
|
||||
-- lifecycle_state='archived'. Rows stay in the table for audit.
|
||||
-- * Reparent paliad.deadlines.sequencing_rule_id duplicate →
|
||||
-- canonical BEFORE archiving, so no live deadline keeps pointing
|
||||
-- at an archived sequencing_rule. (deadlines.rule_id column
|
||||
-- dropped in mig 140; the back-link lives on sequencing_rule_id
|
||||
-- now — same UUID semantics.)
|
||||
-- * Snapshot the affected rows into paliad.sequencing_rules_pre_152
|
||||
-- in the same TX, mirroring precedent (migs 091/093/095/098/140/151).
|
||||
-- * set_config('paliad.audit_reason') is defensively called even
|
||||
-- though no audit trigger fires on sequencing_rules today (mig 151
|
||||
-- §comments documented this). Future audit trigger would inherit
|
||||
-- the reason automatically.
|
||||
--
|
||||
-- Generic-shape rationale: the audit query below uses the FULL
|
||||
-- signature paliadin specified — procedural_event_id, proceeding_type_id,
|
||||
-- rule_code, duration_value, duration_unit, primary_party, condition_expr,
|
||||
-- trigger_event_id, alt_*, anchor_alt, combine_op, parent_id, is_spawn,
|
||||
-- spawn_*. A NOTICE surfaces every group BEFORE the archive step so an
|
||||
-- operator running the deploy logs sees what's about to be touched.
|
||||
-- If new groups appear after future seeds, this mig is safe to re-run
|
||||
-- conceptually (it would archive any new clones) but only fires once
|
||||
-- via the applied_migrations protocol.
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 1. Build the dedupe mapping (duplicate_id → canonical_id) into a
|
||||
-- TEMP table used by every subsequent step.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
CREATE TEMP TABLE tmp_sr_dedupe ON COMMIT DROP AS
|
||||
WITH ranked AS (
|
||||
SELECT
|
||||
id, procedural_event_id, proceeding_type_id, rule_code,
|
||||
duration_value, duration_unit, primary_party,
|
||||
condition_expr, trigger_event_id, alt_duration_value,
|
||||
alt_duration_unit, alt_rule_code, anchor_alt, combine_op,
|
||||
parent_id, is_spawn, spawn_label, spawn_proceeding_type_id,
|
||||
created_at,
|
||||
ROW_NUMBER() OVER (
|
||||
PARTITION BY
|
||||
procedural_event_id, proceeding_type_id, rule_code,
|
||||
duration_value, duration_unit, primary_party,
|
||||
condition_expr::text, trigger_event_id,
|
||||
alt_duration_value, alt_duration_unit, alt_rule_code,
|
||||
anchor_alt, combine_op, parent_id, is_spawn, spawn_label,
|
||||
spawn_proceeding_type_id
|
||||
ORDER BY created_at, id::text
|
||||
) AS rn,
|
||||
COUNT(*) OVER (
|
||||
PARTITION BY
|
||||
procedural_event_id, proceeding_type_id, rule_code,
|
||||
duration_value, duration_unit, primary_party,
|
||||
condition_expr::text, trigger_event_id,
|
||||
alt_duration_value, alt_duration_unit, alt_rule_code,
|
||||
anchor_alt, combine_op, parent_id, is_spawn, spawn_label,
|
||||
spawn_proceeding_type_id
|
||||
) AS grp_size
|
||||
FROM paliad.sequencing_rules
|
||||
WHERE is_active = true
|
||||
AND lifecycle_state = 'published'
|
||||
)
|
||||
SELECT
|
||||
r.id AS duplicate_id,
|
||||
canon.id AS canonical_id,
|
||||
r.procedural_event_id,
|
||||
(SELECT name FROM paliad.procedural_events WHERE id = r.procedural_event_id) AS pe_name
|
||||
FROM ranked r
|
||||
JOIN ranked canon
|
||||
ON canon.procedural_event_id IS NOT DISTINCT FROM r.procedural_event_id
|
||||
AND canon.proceeding_type_id IS NOT DISTINCT FROM r.proceeding_type_id
|
||||
AND canon.rule_code IS NOT DISTINCT FROM r.rule_code
|
||||
AND canon.duration_value IS NOT DISTINCT FROM r.duration_value
|
||||
AND canon.duration_unit IS NOT DISTINCT FROM r.duration_unit
|
||||
AND canon.primary_party IS NOT DISTINCT FROM r.primary_party
|
||||
AND canon.condition_expr::text IS NOT DISTINCT FROM r.condition_expr::text
|
||||
AND canon.trigger_event_id IS NOT DISTINCT FROM r.trigger_event_id
|
||||
AND canon.alt_duration_value IS NOT DISTINCT FROM r.alt_duration_value
|
||||
AND canon.alt_duration_unit IS NOT DISTINCT FROM r.alt_duration_unit
|
||||
AND canon.alt_rule_code IS NOT DISTINCT FROM r.alt_rule_code
|
||||
AND canon.anchor_alt IS NOT DISTINCT FROM r.anchor_alt
|
||||
AND canon.combine_op IS NOT DISTINCT FROM r.combine_op
|
||||
AND canon.parent_id IS NOT DISTINCT FROM r.parent_id
|
||||
AND canon.is_spawn IS NOT DISTINCT FROM r.is_spawn
|
||||
AND canon.spawn_label IS NOT DISTINCT FROM r.spawn_label
|
||||
AND canon.spawn_proceeding_type_id IS NOT DISTINCT FROM r.spawn_proceeding_type_id
|
||||
AND canon.rn = 1
|
||||
WHERE r.rn > 1 AND r.grp_size > 1;
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 2. Surface every clone-group as a NOTICE before archiving.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
DO $$
|
||||
DECLARE
|
||||
rec record;
|
||||
total_to_archive int;
|
||||
BEGIN
|
||||
SELECT COUNT(*) INTO total_to_archive FROM tmp_sr_dedupe;
|
||||
RAISE NOTICE '[mig 152] PRE: % sequencing_rules row(s) will be archived', total_to_archive;
|
||||
FOR rec IN
|
||||
SELECT pe_name, canonical_id, COUNT(*) AS dup_count, array_agg(duplicate_id::text ORDER BY duplicate_id::text) AS dup_ids
|
||||
FROM tmp_sr_dedupe
|
||||
GROUP BY pe_name, canonical_id
|
||||
ORDER BY pe_name
|
||||
LOOP
|
||||
RAISE NOTICE '[mig 152] % canonical=% duplicates=% ids=%',
|
||||
rec.pe_name, rec.canonical_id, rec.dup_count, rec.dup_ids;
|
||||
END LOOP;
|
||||
END $$;
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 3. Snapshot the rows about to be archived (only the duplicates;
|
||||
-- the canonicals stay in the live table). Matches precedent.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
CREATE TABLE paliad.sequencing_rules_pre_152 AS
|
||||
SELECT sr.*
|
||||
FROM paliad.sequencing_rules sr
|
||||
JOIN tmp_sr_dedupe d ON d.duplicate_id = sr.id;
|
||||
|
||||
COMMENT ON TABLE paliad.sequencing_rules_pre_152 IS
|
||||
'Snapshot of paliad.sequencing_rules rows archived by mig 152 '
|
||||
'(identical clones — Mängelbeseitigung / Zahlung × 5). Mirrors '
|
||||
'precedent pre_091/093/095/098/140/151. Read-only revert source. '
|
||||
't-paliad-321 / m/paliad#144 follow-up.';
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 4. Reparent paliad.deadlines.sequencing_rule_id duplicate → canonical
|
||||
-- BEFORE archiving. Today's live data has 0 deadlines pointing at
|
||||
-- any duplicate, but the statement is safe + defensive against a
|
||||
-- race between drift-check and apply.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
UPDATE paliad.deadlines d
|
||||
SET sequencing_rule_id = m.canonical_id,
|
||||
procedural_event_id = (SELECT procedural_event_id
|
||||
FROM paliad.sequencing_rules
|
||||
WHERE id = m.canonical_id),
|
||||
updated_at = now()
|
||||
FROM tmp_sr_dedupe m
|
||||
WHERE d.sequencing_rule_id = m.duplicate_id;
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 5. Defensive audit-reason. Sequencing_rules has no audit trigger
|
||||
-- today (mig 151 §scope verified), but set_config is transactional
|
||||
-- and a future audit trigger inherits the reason automatically.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
SELECT set_config('paliad.audit_reason',
|
||||
'mig 152: archive identical sequencing_rule clones (mig 151 follow-up; t-paliad-321)',
|
||||
true);
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 6. Archive the duplicates.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
UPDATE paliad.sequencing_rules
|
||||
SET is_active = false,
|
||||
lifecycle_state = 'archived',
|
||||
updated_at = now()
|
||||
WHERE id IN (SELECT duplicate_id FROM tmp_sr_dedupe);
|
||||
|
||||
-- ----------------------------------------------------------------
|
||||
-- 7. POST assertions.
|
||||
-- ----------------------------------------------------------------
|
||||
|
||||
DO $$
|
||||
DECLARE
|
||||
v_archived int;
|
||||
v_remaining_dupes int;
|
||||
v_orphan_deadlines int;
|
||||
BEGIN
|
||||
-- a. Did the expected number of rows get archived?
|
||||
SELECT COUNT(*) INTO v_archived
|
||||
FROM paliad.sequencing_rules
|
||||
WHERE id IN (SELECT duplicate_id FROM tmp_sr_dedupe)
|
||||
AND lifecycle_state = 'archived'
|
||||
AND is_active = false;
|
||||
IF v_archived <> (SELECT COUNT(*) FROM tmp_sr_dedupe) THEN
|
||||
RAISE EXCEPTION '[mig 152] FAILED POST: expected % rows archived, got %',
|
||||
(SELECT COUNT(*) FROM tmp_sr_dedupe), v_archived;
|
||||
END IF;
|
||||
|
||||
-- b. No clone group of size > 1 should remain in active+published.
|
||||
SELECT COUNT(*) INTO v_remaining_dupes FROM (
|
||||
SELECT 1
|
||||
FROM paliad.sequencing_rules
|
||||
WHERE is_active = true AND lifecycle_state = 'published'
|
||||
GROUP BY procedural_event_id, proceeding_type_id, rule_code,
|
||||
duration_value, duration_unit, primary_party,
|
||||
condition_expr::text, trigger_event_id,
|
||||
alt_duration_value, alt_duration_unit, alt_rule_code,
|
||||
anchor_alt, combine_op, parent_id, is_spawn, spawn_label,
|
||||
spawn_proceeding_type_id
|
||||
HAVING COUNT(*) > 1
|
||||
) g;
|
||||
IF v_remaining_dupes > 0 THEN
|
||||
RAISE EXCEPTION '[mig 152] FAILED POST: % clone group(s) still active+published after archive', v_remaining_dupes;
|
||||
END IF;
|
||||
|
||||
-- c. No deadline points at an archived sequencing_rule.
|
||||
SELECT COUNT(*) INTO v_orphan_deadlines
|
||||
FROM paliad.deadlines d
|
||||
JOIN paliad.sequencing_rules sr ON sr.id = d.sequencing_rule_id
|
||||
WHERE sr.lifecycle_state = 'archived';
|
||||
IF v_orphan_deadlines > 0 THEN
|
||||
RAISE EXCEPTION '[mig 152] FAILED POST: % live deadline(s) still point at an archived sequencing_rule', v_orphan_deadlines;
|
||||
END IF;
|
||||
|
||||
RAISE NOTICE '[mig 152] OK — archived=%, remaining clone groups=0, orphan deadlines=0',
|
||||
v_archived;
|
||||
END $$;
|
||||
Reference in New Issue
Block a user