feat(switch-db): durable guard + validator for switch/compat quality

- trg_compat_speed_guard: BEFORE INSERT/UPDATE on compatibility forces
  algorithmic (spec_match/community) rows whose transceiver is faster than the
  switch max port speed to status=incompatible. Curated rows never touched.
  Makes the "no physically-impossible compatibility" invariant structural, so
  it survives a scraper regression.
- check_switch_quality(): on-demand proof; CI gate is
  `select bool_and(pass) from check_switch_quality() where enforced` (must be
  true). Data-source gaps (coding coverage, speed<=0, datasheet) tracked as
  enforced=false so they do not fail the gate before the data exists.
This commit is contained in:
root 2026-07-06 06:58:25 +00:00
parent 98d0946975
commit 479a0dbe56

View File

@ -0,0 +1,98 @@
-- sql/126-switch-quality-enforcement.sql
-- Durable enforcement + on-demand validation for the switch/compatibility DB.
-- Complements sql/125 (which cleaned the data): makes the "no physically-impossible
-- compatibility rows" invariant structural, so it survives a scraper regression.
--
-- Design (same "no false positives" rule as 125):
-- * The guard NEVER touches curated rows (vendor_matrix/vendor_compat/tested/datasheet)
-- -- those are trusted and, on understated switches, are the source of truth.
-- * For algorithmic rows (spec_match/community), if the transceiver is faster than the
-- switch's fastest known port (stored max_speed_gbps OR ports_config-derived), the row
-- is FORCED to status='incompatible' rather than dropped -- it stays recorded but is
-- excluded by every consumer that filters status='compatible' (e.g. queries.ts:302).
-- * Forcing (not RETURN NULL) means the guard never conflicts with the 125 flagging
-- UPDATE and never silently loses data. Idempotent; safe to re-run.
--
-- Depends on sql/125 (fn_ports_max_gbps + the v_* views). Apply after 125.
-- ── Guard: keep physically-impossible algorithmic matches out of 'compatible' ──
create or replace function compat_speed_guard() returns trigger
language plpgsql as $$
declare
sw_ceiling numeric;
tx_speed numeric;
begin
-- Trust curated/vendor/tested sources; they define reality on understated switches.
if NEW.verification_method in ('vendor_matrix','vendor_compat','tested','datasheet') then
return NEW;
end if;
if NEW.status is distinct from 'compatible' then
return NEW; -- only guard rows claiming compatibility
end if;
select greatest(coalesce(s.max_speed_gbps,0), coalesce(fn_ports_max_gbps(s.ports_config),0))
into sw_ceiling
from switches s where s.id = NEW.switch_id;
select t.speed_gbps into tx_speed
from transceivers t where t.id = NEW.transceiver_id;
if sw_ceiling is not null and sw_ceiling > 0
and tx_speed is not null and tx_speed > sw_ceiling then
NEW.status := 'incompatible';
if NEW.notes is null or NEW.notes = '' then
NEW.notes := 'auto(126): transceiver speed exceeds switch max port speed';
elsif position('auto(12' in NEW.notes) = 0 then
NEW.notes := NEW.notes || ' | auto(126): transceiver speed exceeds switch max port speed';
end if;
end if;
return NEW;
end $$;
comment on function compat_speed_guard() is
'BEFORE INSERT/UPDATE guard on compatibility: forces algorithmic (non-curated) rows whose transceiver is faster than the switch max port speed to status=incompatible. Never alters curated rows.';
drop trigger if exists trg_compat_speed_guard on compatibility;
create trigger trg_compat_speed_guard
before insert or update on compatibility
for each row execute function compat_speed_guard();
-- ── Validator: prove switch-DB quality on demand (CI / cron) ──────────────────
-- Hard invariants have enforced=true; data-source gaps (coverage, datasheet) are
-- reported but enforced=false so they don't fail the gate before the data exists.
-- CI gate: select bool_and(pass) from check_switch_quality() where enforced; -- must be true
create or replace function check_switch_quality()
returns table(check_name text, detail text, pass boolean, enforced boolean)
language sql stable as $$
select 'no_impossible_spec_match_compat',
(select count(*) from v_compat_speed_errors)::text || ' impossible rows',
(select count(*) from v_compat_speed_errors) = 0,
true
union all
select 'coding_coverage_100G_ge_50pct',
coalesce((select coverage_pct from v_coding_coverage where tier = 100), 0)::text || '%',
coalesce((select coverage_pct from v_coding_coverage where tier = 100), 0) >= 50,
false
union all
select 'transceivers_speed_positive',
(select count(*) from v_transceiver_speed_errors)::text || ' with speed<=0',
(select count(*) from v_transceiver_speed_errors) = 0,
false
union all
select 'switches_have_max_speed',
(select count(*) from v_switch_completeness where max_speed_gbps is null)::text || ' missing',
(select count(*) from v_switch_completeness where max_speed_gbps is null) = 0,
false
union all
select 'no_understated_switch_ports',
(select count(*) from v_switch_understated_ports)::text || ' understated',
(select count(*) from v_switch_understated_ports) = 0,
false;
$$;
comment on function check_switch_quality() is
'Switch-DB quality proof. select bool_and(pass) from check_switch_quality() where enforced; must be true (no physically-impossible compatibility rows). enforced=false rows track datasheet/data-source gaps.';
insert into _migrations (filename, applied_at)
select '126-switch-quality-enforcement.sql', now()
where not exists (select 1 from _migrations where filename = '126-switch-quality-enforcement.sql');