Adds period_type TEXT NOT NULL DEFAULT 'unknown' to macro_indicator and
widens the PRIMARY KEY to (indicator_type, region, obs_date, period_type).
Before: yearly aggregate ('год'→obs_date YYYY-01-01) and Q1 ('I квартал'→
same date) shared the same PK slot → ON CONFLICT DO UPDATE made them
overwrite each other despite the in-memory dedup fix in #1687.
After: each EmissRow carries period_type from _emiss_period_granularity
('year'/'quarter'/'month'); DB ON CONFLICT targets include it, so
yearly+Q1 rows genuinely coexist.
Non-EMISS sources (CBR, rosstat open-data, domrf) use period_type='unknown'
(literal in SQL) — they are disambiguated by obs_date already, so the
wider PK is backward-compatible. All ON CONFLICT clauses in
cbr_macro_sync.py and rosstat_macro_sync.py updated accordingly.
Migration: 163_emiss_pk_period_type.sql (idempotent BEGIN/COMMIT).
Tests: 49 passed (emiss + rosstat suite), ruff clean, py_compile OK.
scrape_objective.sync_objective_group инкрементировал n_requests и reports_ok
ДО HTTP-стрима client.stream_report(...). Любая ошибка стрима (ObjectiveAuthError,
ObjectiveAPIError — оба ⊂ RuntimeError ⊂ Exception, ijson-сбой, обрыв сети)
проглатывалась generic except Exception в lots_pf-ветке без инкремента
reports_failed и без re-raise. Внешний except (ObjectiveAuthError|APIError)
был НЕДОСТИЖИМ для lots_pf. _finish_run(status='done' if reports_failed==0)
помечал run УСПЕШНЫМ при полностью провалившемся главном 600МБ отчёте «Лоты».
objective_lots/history оставались stale молча, админка показывала last_run_at,
downstream supply-layers L2 видел 'done', rows_lots=0.
Patch (зеркалит corp_sum — там паттерн уже правильный):
- n_requests/reports_ok перенесены ИЗ pre-HTTP в after-success (после
успешного commit'а внутри `with stream_report(...)`).
- В except Exception lots_pf-ветке: n_requests += 1 + reports_failed += 1
+ logger.exception (с traceback). Graceful (no re-raise) сохранён —
sibling jobs продолжают, но статус run теперь честный.
- corp_sum ветка не тронута. _finish_run сигнатура не меняется.
5 новых регрессионных тестов: success keeps reports_ok=1, auth-error /
api-error / generic Exception в lots_pf инкрементируют reports_failed (не
ok), _finish_run status='failed' при reports_failed>0. ruff clean.
Closes#1220
Два concurrency-бага в scrape_kn:
1. Lock value="1" без owner-токена + TTL=30мин = заявленной длительности
sweep'а (нулевой запас). Sweep > TTL → lock истекает, второй sweep
стартует, первый при завершении безусловным r.delete сносит ЧУЖОЙ
lock → возможен третий параллельный.
2. SIGKILL/redeploy mid-sweep: finally не выполняется, lock живёт до
30 мин. worker_ready метит run 'zombie' и enqueue'ит resume_kn_run
БЕЗ countdown, который ловит lock_held и возвращает skipped без
retry → run теряется до недельного beat.
Patch:
- _region_lock: value = uuid4().hex; release через Lua check-and-delete
(`if GET == ARGV[1] then DEL else 0 end`) — не сносим чужой lock.
- _LOCK_TTL_SECONDS 30 → 45 мин (полтора max sweep duration, запас).
- resume_kn_run: при lock_held → raise self.retry(countdown=300,
max_retries=12) → ~час окно для подхвата вместо silent skip.
- scrape_kn_region (scheduled) намеренно остаётся skipped — beat
поднимет в следующий weekly tick (другая семантика).
11 новых юнит-тестов (token uniqueness, Lua guard, retry-not-skip,
existing behaviors). 16/16 scrape_kn тестов зелёные. ruff clean.
Closes#1216
В WAF-ветке `nspd_geo` worker спит до 240с (4 мин при consecutive_waf=8),
heartbeat коммитится ДО сна. Старый `cleanup_zombies` (`* * * * *`) с
порогом `INTERVAL '2 minutes'` для status IN ('running', 'paused')
гарантированно ре-enqueue'ил **живой** WAF-job:
- Два worker'а параллельно долбили забаненный NSPD-сервис, углубляя бан.
- SELECT pending без `FOR UPDATE SKIP LOCKED` → дубли запросов.
- Counter'ы done/failed двух инстансов затираются друг другом.
Бонус-баг: status='paused' (после 8 WAF подряд) воскресал через минуту,
аннулируя WAF-защиту.
Patch: разделил пороги через module-level константы и параметризованный
SQL (CAST(:x AS interval) per psycopg v3 rule):
- `_ZOMBIE_RUNNING_THRESHOLD = "6 minutes"` — выше max WAF backoff (4 мин).
- `_ZOMBIE_PAUSED_THRESHOLD = "30 minutes"` — реальная WAF-пауза.
Beat-расписание (1 мин tick) не тронуто. WAF-логика / heartbeat-in-loop
(вариант B) / FOR UPDATE SKIP LOCKED (вариант «идеальный») вне scope.
4 новых юнит-теста (test_nspd_geo.py): psycopg v3 guard, оба порога
биндятся, два предиката вместо одного, paused > running.
18/18 nspd_geo тестов зелёные. ruff clean.
Closes#1215
Closes the last #946 gap (СМР). Fetches «Индексы цен производителей на
строительную продукцию по РФ» from the OPEN Rosstat xlsx
(rosstat.gov.ru/storage/mediabank/Invest_ind_stroitel_MM-YYYY.xlsx),
parses the monthly «к предыдущему месяцу» section → macro_indicator
(indicator_type='construction_price_index', region='rf', monthly, %).
The reCAPTCHA-gated fedstat dataGrid.do path is DELIBERATELY NOT used —
the open Rosstat source needs no captcha. EMISS id=31108 documented as a
prod-only SDMX alternative (not wired).
URL is date-derived with bounded look-back (filename changes monthly;
_fetch_construction_latest walks current month back to 6, first 200 wins;
year-boundary handled, exhaustion raises, 404 breaks fast — no hang).
Wired into rosstat_macro_sync with per-source try/except (can't break
demography/income) + SAVEPOINT-per-row. None-not-0 on blank months.
33 tests (incl look-back + year-boundary + graceful), ruff clean.
Note: no construction-cost channel in macro_coefficient yet (demand-side
only) — series is ingested/available for future cost-side use + display.
Refs #946
fedstat ИПЦ is reCAPTCHA-blocked; CBR publishes inflation openly. Add
fetch_inflation + parse_inflation_xlsx (CBR UniDbQuery DownloadExcel/132934,
monthly % г/г, region=rf, source=cbr) to cbr_macro.py; upsert
indicator_type=inflation_yoy via the existing cbr_macro_sync task (per-series
guard, SAVEPOINT-per-row, CAST not ::, ON CONFLICT on the PK).
Surface inflation_yoy in MonthlyMacro (frozen, carry-forward) and ACTIVATE the
reserved §9.5 inflation channel (macro_coefficient f_inflation: level-vs-4%-target
nudge, non-positive to avoid double-counting f_rate, excluded from
_RATE_DRIVEN_FACTORS). Channel was DEGRADED (no data) -> now BACKED + consumed;
_CONF_HIGH_MIN_BACKED 4->5. Deterministic (§16/§26); renorm claims the reserved
0.08 slice as designed. Live-verified (2026-04 5.58%); 194 macro + 902 forecasting
tests green. No migration, no new deps.
Refs #946.
build_site_finder_report (§22) takes ~30-180s → runs in a background Celery task,
not inline on the sync /analyze endpoint.
- repository: latest_run_for gains keyword-only schema_version (default None keeps
v_analysis_runs_latest behavior, backward-compat); when given, reads base
analysis_runs filtered by schema_version ORDER BY created_at DESC LIMIT 1 — fetches
the latest analyze-1.0 site-analysis run even when newer 1.0 (§22) rows exist on top
(index-served via 127's (cad_num, created_at DESC)).
- new workers/tasks/forecast.py::forecast_site_finder_report: reads latest analyze-1.0,
runs the §22 orchestrator, persists SiteFinderReport.as_dict() as a 1.0 run via
persist_analysis_run. Graceful: no base run / compute error → logger + return None
(worker not crashed). time_limit=900/soft=840 (no global limit). Registered in include.
Prod-confirmed: analyze-1.0 result carries the full analyze dict (competitors+district)
→ orchestrator input valid. Endpoint trigger (3b-ii) + §9.x untouched. 943 tests pass;
code-review APPROVE (contracts verified vs real as_dict(); status done→complete normalized,
no IntegrityError). Refs #994#961.
REAL code-level cause (prod-confirmed, 2 parcels: "Geometry type (Polygon) does
not match column type (MultiPolygon)") of single-contour parcels stuck "fetching"
— complements migrations 129/130 (which fixed the cad_parcels NOT NULL chain).
F1: _save_parcel/_save_quarter now ST_Multi() the geom so single-contour Polygons
fit the MultiPolygon columns (migr 93/58), mirroring the bulk_harvest paths.
F2: _save_building rewritten to the wide cad_buildings schema (migr 92):
quarter_cad_number (was quarter_cad_num → UndefinedColumn), _safe_int(floors)
(range strings → NULL), source='nspd' NOT NULL.
F3: upsert_features wraps each feature in begin_nested() SAVEPOINT so one bad
feature no longer aborts the whole quarter snapshot (backend.md rule; mirrors
grid-walk). New tests/workers/test_nspd_geo.py + savepoint tests. 89 passed.
REOPENED. L3 future-supply rows are computed per (district_name, dev_group_name)
but dev_group_name was never a key column — only embedded in method text. With
complex_id/obj_class NULL for L3, every dev_group of a district collapsed to one
upsert key → ~95.6% loss. Ground-truth (Академический, prod): should be 13,808
units / 15 dev_groups / 54 objects; only 1 row / 607 units survived.
Migration 128: ADD COLUMN supply_layers.dev_group_name TEXT + rebuild
uq_supply_layers_logical to (layer, district_name, complex_id, obj_class,
dev_group_name, source, snapshot_date) NULLS NOT DISTINCT (L1/L2 dev_group_name
NULL stays transparent → their dedup unchanged; L3 distinct groups no longer
collapse). Dry-run-verified vs prod catalog (applies clean, ROLLBACK clean).
Worker: SupplyLayerRow gains dev_group_name (L1/L2=None, L3=group); _UPSERT_SQL
adds it to INSERT/VALUES (CAST(:dev_group_name AS text)) + ON CONFLICT (key col,
not in DO UPDATE SET). Service+worker regression tests assert same-district/
different-dev_group → distinct keys (no collapse). 234 supply tests pass.
Deploy applies migration before container restart; collapsed data self-heals on
next supply_layers_refresh. Verification = prod re-measure post-deploy.
Refs #970
