Commit graph

291 commits

Author SHA1 Message Date
14f3ef2019 fix(week-review): backend-аудит v2 — 82 фиксов (#1660)
All checks were successful
Deploy / build-worker (push) Successful in 2m47s
Deploy / deploy (push) Successful in 1m20s
Deploy / changes (push) Successful in 9s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m52s
Co-authored-by: bot-backend <bot-backend@gendsgn.local>
Co-committed-by: bot-backend <bot-backend@gendsgn.local>
2026-06-17 17:13:38 +00:00
479f01ae1a Merge remote-tracking branch 'forgejo/main' into HEAD
Some checks failed
CI / changes (push) Successful in 7s
CI / changes (pull_request) Successful in 8s
CI / frontend-tests (push) Successful in 1m7s
CI / openapi-codegen-check (push) Failing after 2m9s
CI / frontend-tests (pull_request) Successful in 1m4s
CI / openapi-codegen-check (pull_request) Failing after 1m37s
CI / backend-tests (pull_request) Failing after 9m7s
CI / backend-tests (push) Failing after 9m11s
# Conflicts:
#	backend/app/services/site_finder/gate_verdict.py
2026-06-15 21:07:05 +03:00
86e9ea2937 fix(week-review): автофиксы код-ревью — 169 issue (label «week ревью 1»)
Многоагентный аудит + имплементация: один воркер на файл, точечные правки.
Верификация: py_compile (47/47 .py) + tsc --noEmit (0 ошибок). Unit-тесты
не прогонялись (окружение не поднято: rollup native dep / нет pytest-venv).

Полностью исправлено (169): #1336, #1337, #1339, #1340, #1341, #1342, #1343, #1345, #1346, #1348, #1349, #1350, #1351, #1354, #1356, #1358, #1359, #1360, #1362, #1364, #1365, #1366, #1367, #1368, #1369, #1370, #1371, #1372, #1373, #1374, #1375, #1376, #1377, #1378, #1379, #1380, #1381, #1382, #1384, #1385, #1386, #1387, #1388, #1389, #1390, #1391, #1392, #1394, #1395, #1396, #1397, #1399, #1400, #1401, #1402, #1403, #1404, #1408, #1409, #1410, #1411, #1412, #1413, #1414, #1415, #1416, #1417, #1418, #1420, #1423, #1425, #1426, #1427, #1428, #1429, #1430, #1431, #1432, #1433, #1434, #1435, #1437, #1438, #1439, #1440, #1441, #1442, #1443, #1444, #1445, #1446, #1447, #1448, #1449, #1450, #1451, #1452, #1453, #1454, #1455, #1456, #1457, #1458, #1459, #1460, #1461, #1462, #1463, #1464, #1465, #1466, #1467, #1468, #1469, #1471, #1472, #1473, #1474, #1476, #1478, #1479, #1481, #1482, #1483, #1484, #1485, #1487, #1488, #1489, #1490, #1491, #1492, #1493, #1494, #1495, #1496, #1497, #1499, #1500, #1501, #1502, #1504, #1505, #1506, #1507, #1510, #1514, #1515, #1516, #1517, #1518, #1519, #1521, #1522, #1523, #1524, #1525, #1526, #1527, #1528, #1529, #1531, #1532, #1533, #1534, #1535, #1536, #1537, #1538

Частично (9, in-file часть, остаток cross-file): #1361, #1419, #1422, #1424, #1470, #1475, #1477, #1480, #1498
Требуют cross-file (3, не тронуты): #1338, #1363, #1421
Пропущено (1): #1539

Не входило в партию: 22 needs-Leha issue (нужны решения владельца).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-15 20:21:11 +05:00
b405ae291d feat(site-finder): additive objective_lots price fallback for competitors (#307 OBJ-3) (#1332)
All checks were successful
Deploy / changes (push) Successful in 7s
Deploy / build-backend (push) Successful in 1m57s
Deploy / build-worker (push) Successful in 2m51s
Deploy / build-frontend (push) Successful in 3m6s
Deploy / deploy (push) Successful in 1m12s
2026-06-14 16:10:02 +00:00
72e0b050f9 feat(site-finder): POI saturation per capita по районам ЕКБ (#42) (#1329)
All checks were successful
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m38s
Deploy / build-worker (push) Successful in 2m52s
Deploy / deploy (push) Successful in 1m12s
2026-06-14 15:05:22 +00:00
e5e0ccb283 feat(site-finder): сетевые обременения из ЗОУИТ (#1070, Refs #103) (#1328)
All checks were successful
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m36s
Deploy / build-worker (push) Successful in 2m40s
Deploy / deploy (push) Successful in 1m12s
2026-06-14 14:49:32 +00:00
5507bbc1a4 feat(site-finder): domrf→cad geom-match + parking_ratio конкурентов (#96) (#1327)
Some checks failed
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / build-backend (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / changes (push) Has been cancelled
2026-06-14 14:49:28 +00:00
2c22c3f7ea feat(site-finder): per-building помещения/машино-места + parking_ratio (on-demand MVP, #96)
All checks were successful
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 6s
CI / changes (push) Successful in 8s
CI / frontend-tests (pull_request) Has been skipped
CI / openapi-codegen-check (push) Successful in 1m59s
CI / openapi-codegen-check (pull_request) Successful in 1m58s
CI / backend-tests (push) Successful in 8m48s
CI / backend-tests (pull_request) Successful in 8m48s
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 4m58s
Deploy / build-worker (push) Successful in 5m46s
Deploy / deploy (push) Successful in 1m30s
Оживляет мёртвый foundation list_objects_in_building (#168 Q3-deferred). Fix реального
блокера: NSPDOptions не парсил objdocId (camelCase из NSPD search) → objdoc_id всегда None,
метод нельзя было вызвать. AliasChoices(objdoc_id, objdocId) + registers.

premises_lookup.get_building_premises(cad_num) — резолв objdoc → list_objects_in_building,
graceful (WAF/сеть/not-found → None). parking_ratio = машино-места/помещения (None при 0
помещений, 0.0 при реальном отсутствии паркинга). Verified live 66:41:0106036:183.

MVP on-demand (без bulk/схемы/prod-записей). parking_ratio готов, но НЕ wired в analyze:
конкуренты приходят из ДОМ.РФ без cad_num → нужен domrf↔cad_buildings geom-match (отдельная
задача). Сервис примет cad_num как только matching появится. 11 тестов, ruff clean.

Refs #96
2026-06-14 18:48:16 +05:00
50555821fb feat(site_finder): routing POI decay (#41) + OSM utility distances (#44) + ЗОУИТ geom (#255bk)
Some checks failed
CI / changes (push) Successful in 7s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 7s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 7m42s
CI / backend-tests (pull_request) Successful in 7m40s
Deploy / changes (push) Has been cancelled
Deploy / build-backend (push) Has been cancelled
Deploy / build-worker (push) Has been cancelled
Deploy / build-frontend (push) Has been cancelled
Deploy / deploy (push) Has been cancelled
#41: per-category routing-radius decay через reusable ORS /matrix client (foot-walk
время + per-category пороги + decay-curves), opt-in ?decay=routing, graceful straight-line
fallback (analyze не 500 при ORS down) + length-guard durations==destinations.
#44: +5 OSM-категорий (transformer/gas/water/sewerage/heat) via Overpass nwr;
nearest_{water_main,substation,gas,heat}_m в analyze.utilities.
#255bk: ST_AsGeoJSON(CAST(geom AS geometry)) в _get_cad_zouit_overlaps → geom_geojson.

Closes #41
Closes #44
2026-06-13 22:31:56 +05:00
658d724075 feat(analytics): velocity-anomaly alerts (#17) + ARN ДДУ price indicator (#99)
Some checks failed
CI / changes (push) Successful in 8s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 8s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 7m37s
CI / backend-tests (pull_request) Successful in 7m42s
Deploy / deploy (push) Blocked by required conditions
Deploy / build-backend (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / build-frontend (push) Blocked by required conditions
Deploy / changes (push) Has been cancelled
#17: detect_velocity_anomalies + GET /analytics/velocity-alerts — z-score drop-detection
на domrf_kn_sale_graph (double-gate z≤-2 AND drop≤-30%, starvation-guards). Snapshot=richest
(не MAX — prod новейшие частичные), lookback anchored на latest report_month (scrape лаг ~4мес).
Prod: ЖК Центральный Парк -69%, ~14ms.

#99: mv_ddu_price_indicator (миграция 152) + POST /market/ddu-indicator — ARN-mirror
ценовой индекс per quarter×area_bucket из rosreestr_deals (ДДУ регион 66). MVP: subject-level,
period Q, window 2025-Q2+, methods 1/2 (basis/previous, prev_period_value honesty). Q1-2026
headline 1.0185 vs ARN 1.03 (±5%). Method 3 blocked (нет pre-2025-Q2 данных) — задокументировано.

Closes #17
Closes #99
2026-06-13 22:04:49 +05:00
8242e51b61 feat(generative): exporters (dxf/pdf) + generative test suite (#54 #55 #56)
Some checks failed
CI / frontend-tests (push) Has been skipped
CI / changes (push) Successful in 10s
CI / changes (pull_request) Successful in 7s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 7m39s
CI / backend-tests (pull_request) Successful in 7m39s
Deploy / deploy (push) Has been cancelled
Deploy / build-frontend (push) Has been cancelled
Deploy / build-backend (push) Has been cancelled
Deploy / build-worker (push) Has been cancelled
Deploy / changes (push) Successful in 6s
Дополняет движок: exporters/{dxf,pdf}.py (ezdxf + WeasyPrint lazy) + 5 тест-модулей
(geometry/placement/teap_financial/exporters/api). Не вошли в предыдущий commit
(untracked-директории).
2026-06-13 21:33:20 +05:00
3b79533f9f feat(generative): движок концепций Stage 1a/1b/1c (#54 #55 #56)
Some checks failed
CI / changes (push) Successful in 7s
CI / frontend-tests (push) Has been skipped
CI / backend-tests (push) Has been cancelled
Заменяет generative stubs детерминированным end-to-end пайплайном:
- 1a geometry: WGS84-parse, метрическая AEQD-репроекция, setback-буфер, placement grid
- 1b placement: greedy section-fill + STRtree коллизии (full-gap), 3 стратегии, coverage cap
- 1c teap/financial: площади/квартиры/FAR/паркинг; выручка/затраты/маржа + IRR-proxy
- exporters: ezdxf DXF (геометрия) + WeasyPrint ТЭП/фин PDF (lazy import)
generate() заменил generate_stub в route (422 на degenerate). mypy-strict + ruff clean,
31 generative-тест + full suite 3078 passed. ConceptVariant заполняется реальными числами.

Closes #54
Closes #55
Closes #56
2026-06-13 21:32:32 +05:00
1883271908 fix(cadastre): NSPD 500-retry + per-layer skip (#252) + disable dead pzz-beat (#259) + geom backfill (#200)
Some checks failed
CI / changes (push) Successful in 7s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 6s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m43s
CI / backend-tests (pull_request) Successful in 6m41s
Deploy / build-backend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / build-frontend (push) Blocked by required conditions
Deploy / changes (push) Has been cancelled
#252: NspdBulkServerError на HTTP 5xx + WMS <ServiceException>-body → quarter-level
autoretry; _grid_walk_category флагует layer_failed только если ВСЕ cells 5xx (none ok),
network-errors исключены — пустой/флапающий слой не ложно-скипается. harvest_meta.layer_X_failed.

#259: pzz-sync-monthly beat отключён (PKK6 deprecated, pzz_zones_ekb=0, заменён
zone_regulation_cache+NSPD dumps). logger.exception→warning (стоп GlitchTip BACKEND-1B),
raise сохранён для admin-endpoint.

#200: geom_unavailable флаг (миграция 151) + backfill_parcel_geom task — grid-walk
layer 36368 per quarter лечит 930/964 NULL-geom участков (NSPD search вернул центроиды),
остаток marks unavailable (не ретраить вечно). SAVEPOINT per-cell, idempotent.

Closes #252
Closes #259
Closes #200
2026-06-13 20:45:05 +05:00
84d08940c6 fix(analytics): свежая ипотечная ставка + recommend last_updated + drop dead route (#236 #237 #251)
All checks were successful
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Successful in 3m9s
Deploy / build-backend (push) Successful in 4m12s
Deploy / build-worker (push) Successful in 5m31s
Deploy / deploy (push) Successful in 1m19s
#236: _current_mortgage_rate сортировал cbr_mortgage_series по TEXT-полю period
лексикографически ('Январь'>'Февраль') → возвращал stale-точку. Сортировка по
parsed (year, month) + regex-guard на malformed period-строки. Остаётся на
subsidized weighted series (НЕ domrf market — иная метрика). Verified prod:
Февраль 2026 (было Январь 2026).

#237: scope.last_updated = MAX(snapshot_date) domrf_kn_objects в recommend_mix +
подпись «данные на <дата>» в recommend page.

#251: удалён never-implemented GET /parcels/{parcel_id} (501) + ParcelDetail схема
(frontend bare 1-сегментный route не зовёт).

Closes #236
Closes #237
Closes #251
2026-06-13 15:22:22 +00:00
3cf9fad683 fix(backend): экранировать Excel formula-injection (#1244) + увести chat-чтение с event loop (#1245)
Some checks failed
CI / changes (push) Successful in 6s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 7s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m38s
CI / backend-tests (pull_request) Successful in 6m37s
Deploy / build-backend (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Has been cancelled
#1244 (security): внешние/скрейпинг-строки (comm_name из DOM.РФ, headline/usp_text)
с ведущим = + - @ \t \r писались как есть → openpyxl сохранял как формулы
(data_type='f'), исполнялись при открытии в Excel/LibreOffice. _sanitize_formula
префиксует такие строки апострофом (OWASP CSV-injection escape); числа/даты/bool
не трогаются. _write_kv labels тоже санитизируются. Подтверждено на openpyxl 3.1.5.

#1245 (concurrency): async ask() вызывал sync get_report_for_chat() (sync SQLAlchemy
тянет крупный JSONB §22-отчёт) напрямую — блокировал event loop, в отличие от
LLM-ветки (run_in_threadpool). Обёрнуто в run_in_threadpool.

Closes #1244
Closes #1245
2026-06-13 18:10:21 +05:00
b211183940 feat(site-finder): детерминированная атрибуция застройщика в analyze (#1088)
All checks were successful
CI / changes (pull_request) Successful in 6s
CI / changes (push) Successful in 8s
CI / frontend-tests (push) Has been skipped
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m33s
CI / backend-tests (pull_request) Successful in 6m33s
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m44s
Deploy / build-worker (push) Successful in 5m24s
Deploy / deploy (push) Successful in 1m30s
Сервис get_developer_attribution поверх fn_developer_for_parcel (миграция 149):
топ-1 застройщик участка + track-record (РНС/РВЭ + домрф) + nearby_developers.
Дедуп по норм-ИНН (лучший match_method первым). Wire в analyze additive без флага
(чистый DB-резолвер по индексам, graceful → None). Beat-refresh developer_registry
ежемесячно 05:30 МСК после ekburg-permits. Pydantic-схемы + 18 тестов.

Источники: ekburg_construction_permits (РНС, ИНН) ⋈ domrf_kn_objects по
нормализованному ИНН + spatial/quarter fallback. 68% domrf-застройщиков имеют РНС.
EXPLAIN: point-lookup 0.06ms, резолвер ~30ms (functional/GIST индексы в миграции).

Closes #1088
2026-06-13 16:13:45 +05:00
eada9a25ac fix(scrapers): izyatie OCR — '6б:' → '66:' для кад-номеров МСК-66 (#1228) 2026-06-13 15:03:59 +05:00
367ed890f1 fix(objective-sync): log seed failure вместо bare except pass (#1226) 2026-06-13 15:03:34 +05:00
49d26b4696 fix(poi_loader): timedelta(days=730) вместо replace(year=-2) → leap-year 29 февраля safe (#1232) 2026-06-13 15:03:08 +05:00
c3c6e2cef1 fix(okn): expand registry coverage + load detail-card name/address (#1159)
okn_objects_sync собирал 428 ОКН из ~798 ЕГРОКН и сохранял raw_props={id}
без названия/адреса — analyze показывал 'ОКН рядом' без указания КАКОГО.

Coverage: расширил _ADDRESS_PASSES с 2 до 11 подстрок (Екатеринбург,
г. Екатеринбург, 7 районов ЕКБ, 2 формы Свердловской обл.). _PASSES =
декартово (address × category). first-wins дедуп сохраняет высшую категорию.

Detail: fetch_okn_detail(source_id) — GET Show/Show?id=, BeautifulSoup
парсит HTML (dl/dt/dd и tr/td), извлекает name, address, protection_category,
object_kind, dating. _enrich_with_detail сливает в raw_props с graceful
failover (cap=1500, exception per-object не валит прогон).

Lookup: parcel_okn_objects возвращает name/address из raw_props->>'name'.
Existing ird_analyze тесты не сломаны.

49 okn-related тестов passed, ruff clean.

Closes #1159
2026-06-13 15:02:50 +05:00
2e428405d2 fix(site-finder): расширить except в parcel_ird_overlaps до DataError (#1095)
parcel_ird_overlaps ловил только (OperationalError, ProgrammingError).
При D9b-wiring в analyze malformed WKT в ST_GeomFromText давал PostGIS
ERROR → SQLAlchemy DataError, который пробивал try/except → analyze
падал вместо graceful-degrade.

- ird_overlay_lookup: + DataError в imports и в except tuple.
- quarter_dump_lookup: симметричное расширение в _get_engineering_*.
- test_ird_overlay_lookup: parametrized test_graceful_when_db_fails
  покрывает 3 класса (Operational/Programming + DataError для WKT).

Closes #1095
2026-06-13 15:02:50 +05:00
8a30238564 fix(forecasting): propagate confounded flag DemandSupplyForecast → §15 (#1222)
DemandSupplyForecast.as_dict() не эмитил 'confounded'/'is_confounded_window',
report_assembler._confounded() всегда возвращал False и §15 confounded_window
factor в compute_report_confidence был мёртв: 48-мес окна, пересекающие
2024-07-01 шок никогда не тянули report confidence к 'low' и шок не назывался
в rationale.

Patch: добавлено confounded: bool в DemandSupplyForecast (от §9.5 macro_coef
OR §9.6 rate_sensitivity), exposed в as_dict(). _confounded() уже использовал
.get() defensively — блокер был в producer'е.

+3 теста: contract на real DemandSupplyForecast.as_dict(), end-to-end
assemble_report → confounded_window factor surfaces at level=low, weakest-link
тянет overall к 'low'. 61 report_assembler + 1034 forecasting тестов зелёные.

Closes #1222
2026-06-13 15:02:50 +05:00
703d3905b8 fix(site-finder): normalize supply room_bucket vocabulary to velocity side (#1229)
best_layouts._SUPPLY_BATCH_SQL эмитил {studio,euro-1,euro-2,1,2,3,4+},
а _INLINE_VELOCITY_SQL читает {студия,1,2,3,4+} из
objective_corpus_room_month (prod check: 'euro-*' rows отсутствуют).

Эффект: rooms=2 + area<50 уходили в euro-1/euro-2 supply-стороной →
выпадали из знаменателя bucket '2' → sold_pct_of_supply двушек
завышен, is_oversold ложно True. (rb='euro-*') dead lookups в supply_map.

Patch: убраны euro-* WHEN в supply CASE. SF-08 euro-биннинг отложен
до момента когда velocity-сторона начнёт его отдавать. +2 regression
теста (bucket match, string guard). 35 best_layouts тестов зелёные.

Closes #1229
2026-06-13 15:02:50 +05:00
780dd5640c fix(analytics): _district_market_saturation latest-snapshot filter (#1221)
domrf_kn_objects/sales_agg хранят ~3 snapshot'а на obj_id (weekly UPSERT,
retention=none). JOIN без latest-фильтра считал каждый ЖК N раз:
sold_median тянулся старыми perc → sat_factor занижен, гейт n>=5
проходил уже при 2 ЖК (n=6), sat_n завышен ~×N в scope. Сданные ЖК
просачивались через старые снапшоты со статусом 'Строящиеся'.

Patch: AND a.snapshot_date = (SELECT MAX(snapshot_date) FROM ...) для
обеих таблиц (зеркало #1210/#1212). COUNT(*) → COUNT(DISTINCT a.obj_id)
defensive. 29 recommend_mix + 8 saturation тестов зелёные.

Closes #1221
2026-06-13 15:02:50 +05:00
d694e735cd fix(site-finder): per-row SAVEPOINT в noise_loader (#1231)
Loop UPSERT в sync_noise_sources_to_db без begin_nested: один замкнутый
3-точечный natural=water way [A,B,A] даёт POLYGON((A,B,A)), PostGIS
отвергает (< 4 точек в ring) → outer tx rollback + raise → весь weekly
noise/water/utility sync падает, тот же way отравляет каждый прогон.

- Оборачиваем каждый UPSERT в `with db.begin_nested():` + per-row
  try/except → logger.warning + skipped++ (канон pzz_loader.py:111).
- В _way_to_polygon_wkt проверяем итоговое кольцо ≥ 4 точек (fail-safe).
- Outer except: добавлен logger.exception для видимости.

Closes #1231
2026-06-13 15:02:50 +05:00
285e8f974a fix(sf): ekb_ppt_tep post-фильтры + врезка в analyze (#1136)
All checks were successful
Deploy / changes (push) Successful in 6s
Deploy / build-backend (push) Successful in 2m12s
Deploy / build-frontend (push) Has been skipped
Deploy / build-worker (push) Successful in 4m34s
Deploy / deploy (push) Successful in 1m11s
Post-фильтры парсера (по образцу ppt2018_22823):
- Табл.11: отбрасываем строки без zone_name (преамбульные артефакты
  merged-cells давали 2 фантома при 12 реальных зонах).
- Табл.13: dedup по (phase, composition, zone, area_ha) + фильтры
  короткого-без-зоны и нумерационного шума (~364 строк → ~30-40 реальных).

Analyze-врезка: новый ppt_tep_lookup.parcel_ppt_tep — JOIN
planning_projects ⋈ ekb_ppt_tep по doc_ref↔source_key/doc_full_name
(best-effort, без FK). Wired в build_ird_analyze_block рядом с
planning_projects/krt_requisites — DB-источники, graceful.

Seed-URL остаётся placeholder с # VERIFY (ingest пропускает с WARNING).

Tests: 27 в parser (3 новых), 7 lookup, 22 wiring (2 новых).
All ruff/syntax green.

Closes #1136
2026-06-13 09:32:52 +00:00
a5ff281bad fix(admin-jobs): UPSERT в PUT /admin/jobs/settings — не терять изменения (#1223)
All checks were successful
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m37s
Deploy / build-worker (push) Successful in 2m46s
Deploy / deploy (push) Successful in 1m19s
UPDATE без проверки rowcount затрагивал 0 строк для несуществующего
job_type; get_one возвращал hardcoded _fallback. PUT отвечал 200,
админ считал настройку сохранённой — но она терялась (новый job_type
до seed-миграции, опечатка в path).

Patch: update() проверяет result.rowcount: при 0 делает INSERT с
переданными колонками (непереданные → дефолты таблицы из м.81).
INSERT использует ON CONFLICT (job_type) DO NOTHING + повторный
UPDATE как защита от гонки. Поведение для существующих строк не
изменилось. 8 новых тестов (6 service + 2 API).

Closes #1223
2026-06-13 08:36:23 +00:00
28c567a34e fix(scoring): correct price_feasibility reason — market rate, not subsidized (#1225)
All checks were successful
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m35s
Deploy / build-worker (push) Successful in 2m35s
Deploy / deploy (push) Successful in 1m21s
ProductScore.reason для price_feasibility (§14.2) называл основу платежа
'субсид. ставка', но compute_affordability с #981 DoD перешёл на рыночный
прокси key_rate + 4.5 п.п. (~19%, rate_kind='key_rate_proxy'); субсидированный
путь не вызывается. Метка делала рыночный платёж в §22-отчёте похожим на
льготный и противоречила affordability.degraded_reason в том же payload.

Текстовый фикс: 'рыночная ставка key_rate + спред, §7.9' — число платежа
было корректно, advisory-описание теперь совпадает.

Closes #1225
2026-06-13 08:20:30 +00:00
10df430cac fix(forecasting): _price_overlap zero-width discontinuity at §25.3 (#1224)
Some checks failed
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 8s
Deploy / build-worker (push) Has been cancelled
Deploy / build-backend (push) Has been cancelled
Degenerate price band (own_min==own_max или c_lo==c_hi, оба разрешены
CHECK 148 и Pydantic) внутри другой вилки возвращали 0.0 вместо 1.0:
фильтр w>0 выкидывал нулевую ширину → 0/positive=0.0. Это рвало
докстринг 'полное накрытие узкого = 1.0' и давало разрыв:
[148k,152k]→1.0 vs [150k,150k]→0.0, занижая среднее каннибализации.

Patch: вырожденные ширины обрабатываются ДО нормирования.
lo<=hi → точка внутри другой вилки → 1.0, вне → 0.0. +inf-обе-премиум
ветка перенесена в начало (избежать inf-inf=nan). +7 новых тестов в
TestPriceOverlap. 220 special_indices тестов зелёные.

Closes #1224
2026-06-13 08:10:46 +00:00
e71224ef54 fix(scrapers): commit outer tx в domrf_catalog scrape_catalog_batch (#1227)
Some checks are pending
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Waiting to run
Deploy / build-backend (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / build-frontend (push) Blocked by required conditions
begin_nested() лишь RELEASE'ит SAVEPOINT — без db.commit() все UPDATE
цен откатывались при db.close() в Celery task. Зеркалит правильный
паттерн scrape_catalog_objects (domrf_catalog_object.py:460): try-
commit/except-rollback в конце функции. 50 catalog тестов зелёные.

Closes #1227
2026-06-13 08:06:15 +00:00
0c62d0b0c3 fix(site-finder): склеить разорванный regex в _PLANNING_NO_MATCH_SQL (#1230)
Some checks failed
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 7s
Deploy / build-backend (push) Has been cancelled
Deploy / build-worker (push) Has been cancelled
CI / changes (pull_request) Successful in 7s
CI / changes (push) Successful in 8s
CI / backend-tests (pull_request) Has been skipped
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Has been skipped
CI / frontend-tests (push) Has been skipped
Литерал regex был разорван переносом строки внутри одной пары '...':
ветки 'Донбасской\\n' и 'Лумумбы\\n' содержали NL+8 пробелов
и никогда не матчили однострочные planning_projects.full_name (PG ~ —
POSIX ARE без (?x)). Geom-only КРТ-fallback молча терял ППТ по этим двум
топонимам.

Patch: вынес паттерн в _KRT_TOPONYM_REGEX через Python adjacent-literal
concat, склеил в SQL через +. Все 17 веток в одной строке.
9 krt_lookup тестов зелёные.

Closes #1230
2026-06-13 12:43:28 +05:00
c48f45c48f fix(forecasting): resolve admin district to micros в _query_artificial_demand (#1205)
Some checks failed
Deploy / build-backend (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Has been cancelled
objective_lots.district хранит МИКРО-вокабуляр ('Уралмаш', 'ЖБИ', ...).
_query_artificial_demand фильтровал сырым АДМИН-именем ('Кировский' с
forecast.py:123) → ol.district='Кировский' = 0 строк → n_sold=0 → §25.5
Artificial Demand 'unavailable' с ложной причиной «нет проданных лотов»
в каждом district-scoped отчёте. Тот же класс бага, что #1211 в
_price_sensitivity.

Patch: импорт resolve_objective_districts + замена сырого
`ol.district = CAST(:district AS text)` на зеркальный паттерн
sales_series._SOURCE_B_SQL / market_metrics._SALES_WINDOW_SQL:
  (CAST(:has_district AS boolean) IS FALSE
   OR ol.district = ANY(CAST(:districts AS text[])))

Сигнатура _query_artificial_demand / _build_artificial_demand НЕ меняется
— caller остаётся admin-aware на входе.

+5 новых тестов (TestArtificialDemandDistrictResolution: резолвер вызван,
микро в bind, n_sold>0 после фикса), 6 обновлённых SQL-тестов. 21 passed
artificial_demand + 1030 forecasting тестов зелёные. ruff clean.

Closes #1205
2026-06-13 07:34:54 +00:00
0fffc535cd fix(analytics): rolling 6mo/6mo окна в _district_velocity_trend (#1203)
All checks were successful
Deploy / changes (push) Successful in 7s
Deploy / build-backend (push) Successful in 1m35s
Deploy / build-frontend (push) Has been skipped
Deploy / build-worker (push) Successful in 2m37s
Deploy / deploy (push) Successful in 1m13s
_district_velocity_trend сравнивал recent vs prior, но окна были
ЗАХАРДКОЖЕНЫ датами 2025:
  recent = report_month >= 2025-07-01           (без верхней границы!)
  prior  = report_month BETWEEN 2025-01-01 AND 2025-06-30

objective_corpus_room_month обновляется еженедельно (objective_etl,
period_month без капа). На 2026-06 recent-окно ≈ 11-12 месяцев vs
prior=6 → при плоском рынке ratio ≈ 1.83, дальше навсегда упирается
в clamp(_, 2.0) → trend_factor=2.0 → macro_velocity_mult ×2 →
bucket_velocity ×2 → months_to_sellout занижен в ~1.8-2x.
Искажение растёт с каждым месяцем; trend-сигнал мёртв.

Patch: rolling 6mo/6mo окна относительно NOW():
  recent = report_month >= NOW() - INTERVAL '6 months'
  prior  = report_month >= NOW() - INTERVAL '12 months'
           AND report_month < NOW() - INTERVAL '6 months'

Равные окна, plain rolling, без хардкода. 29/29 velocity_trend +
recommend_mix тестов зелёные. ruff clean.

Closes #1203
2026-06-13 06:24:21 +00:00
ddbd924b02 fix(competitors): _AVG_PRICE_SQL — фильтр latest snapshot (#1210)
Some checks failed
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 5s
Deploy / build-backend (push) Successful in 1m41s
Deploy / build-frontend (push) Has been skipped
Deploy / build-worker (push) Has been cancelled
domrf_kn_flats версионируется (UNIQUE(id, snapshot_date), м.50), scraper
UPSERT per snapshot — то же что для domrf_kn_objects (которое в L3 supply
после #1212 берём только latest). _AVG_PRICE_SQL фильтра snapshot_date НЕ
имел → AVG усреднял ИСТОРИЮ цен (stale на растущем рынке) → UI-поле
Competitor.avg_price_per_m2 + вход _price_similarity получали устаревшую
цену. COUNT '%прод%' множил sold ×N снапшотов → raw_sold/flat_count кратно
завышен → попадал в гард-нейтраль 0.5 или искажал stage_at_horizon как
×N-завышенный sold_pct.

Patch: WHERE f.snapshot_date = (SELECT MAX(snapshot_date) FROM domrf_kn_flats).
Зеркало паттерна best_layouts._SUPPLY_BATCH_SQL и _COMPETITORS_SQL DISTINCT ON
(уже было latest). 51/51 competitors-тестов зелёные.

Closes #1210
2026-06-13 06:17:26 +00:00
5826812816 fix(domrf_kn): детерминированный sha256-fallback для flat_id (#1208)
Some checks failed
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 6s
Deploy / build-worker (push) Has been cancelled
Deploy / build-backend (push) Has been cancelled
Fallback flat_id в _norm_flat использовал abs(hash(elem)) % (2**63 - 1).
В CPython hash(str) РАНДОМИЗИРОВАН per-process (PYTHONHASHSEED нигде в
репо не зафиксирован — uvicorn/celery не выставляют его). Эффект:

- При resume упавшего sweep (resume_kn_run в новом процессе воркера) до
  10 объектов после checkpoint перечитываются. Квартиры без flatId
  получают ДРУГОЙ hash-id → ON CONFLICT (id, snapshot_date) не
  срабатывает → дубли строк одной квартиры в одном snapshot.
- То же при повторном прогоне за ту же дату — каждый раз новый id.
- Дубли инфлируют все агрегации (units_sold/price медианы, supply_layers).

Patch:
- abs(hash(elem)) → int.from_bytes(sha256(elem)[:8], "big") % (2**63 - 1).
- sha256 стабилен между процессами/перезапусками. 8 байт → BIGINT-fit.

6 новых юнит-тестов (test_domrf_kn_normalize.py): formula matches sha256,
stable across calls, distinct elems→distinct ids, flatId wins over
fallback, no-id→None, BIGINT-fit. ruff clean.

Closes #1208
2026-06-13 06:09:54 +00:00
c29f4aef57 fix(llm): cap серверного Retry-After через _MAX_BACKOFF_S=30s (#1209)
Some checks are pending
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / build-backend (push) Blocked by required conditions
Deploy / build-worker (push) Blocked by required conditions
Deploy / changes (push) Successful in 7s
В complete-loop'е min(...,30) применялся ТОЛЬКО к exponential backoff'у
(else-ветка). Серверный Retry-After уходил в time.sleep как есть.
_parse_retry_after принимает любое число секунд ("86400".isdigit() → True),
а provider'ы шлют Retry-After до 86400с при quota-exhaustion (OpenAI,
CDN-503). Эффект: time.sleep блокирует поток anyio-threadpool на часы
(до 48ч при llm_max_retries=2). Async-консьюмер (chat.py) мостится через
run_in_threadpool — поток держит токен пула (~40 потоков, общий с sync
Depends(get_db)) и DB-сессию → пул исчерпывается → стопор приложения.

Patch: вынес кап на module-level _MAX_BACKOFF_S=30s, применяю к ОБЕИМ
веткам (Retry-After И exp.backoff). raw_wait логируется отдельно для
наблюдаемости (видно когда провайдер просил больше).

Новый юнит-тест test_rate_limited_retry_after_capped: provider шлёт
retry_after=86400 → time.sleep вызывается с 30 (не 86400). 14/14
client тестов + 55/55 LLM-suite зелёные. ruff clean.

Closes #1209
2026-06-13 06:07:46 +00:00
4c2f19ace0 fix(market-metrics): resolve admin→micros for _price_sensitivity (#1211)
Some checks failed
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 5s
Deploy / build-backend (push) Has been cancelled
Deploy / build-worker (push) Has been cancelled
_price_sensitivity передавал сырое admin-имя ('Кировский') в _elasticity_coef,
который фильтрует objective_corpus_room_month.district по МИКРО-вокабуляру
(Втузгородок, ЖБИ, …) → регрессия получала 0 точек → всегда FALLBACK_ELASTICITY.
§9.2 district-level эластичность молча НЕ считалась в /analyze-пути (только
'Академический' совпадал в обоих вокабулярах случайно).

Fix: вызываем resolve_objective_districts() в _price_sensitivity и передаём
список микро через новый kwarg districts=[…] в _elasticity_coef. Резолвер
None ('не определён' / нет чистых алиасов) → пустой список → EKB-wide
регрессия. _elasticity_coef расширен с back-compat: districts=None →
legacy путь по district_name (другой caller в analytics_queries —
отдельный bug class, вне scope).

5 новых юнит-тестов TestPriceSensitivityDistrictResolution: admin→micros в
SQL bind, None→EKB-wide, regression preserved post-resolve, graceful.
76/76 market_metrics + 156/156 elasticity/sensitivity тестов зелёные.
ruff + psycopg v3 grep clean.

Closes #1211
2026-06-13 06:02:29 +00:00
bc2d393b05 fix(market_metrics): disambiguate ROLLUP grand-total via GROUPING() (#1214)
Some checks failed
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 7s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m36s
Deploy / build-worker (push) Has been cancelled
_SALES_WINDOW_SQL делал GROUP BY ROLLUP (rooms_int), rooms_int nullable
(ETL пишет NULL для «неопределённого типа», sales_series.py:399 явно
обрабатывает None). Проданный лот с rooms_int IS NULL даёт ДВЕ строки
rooms_int IS NULL (NULL-группа + grand-total итог), неразличимые в
Python (оба if r["rooms_int"] is None).

MixedAggregate-план PG16 эмитит grand-total ПЕРВЫМ (среди hash-строк),
NULL-группа после → loop затирает units_total частичным счётом (живой
тест на PG16: 2000 → 200). Эффект: unit_velocity / absorption_rate
занижены, months_of_supply завышен → base_pace в demand_supply_forecast
неверный (recommendation.py:586) → reports/scoring врёт.

Patch:
- SQL: добавить GROUPING(rooms_int) AS is_total (=1 для grand-total).
- Python: ветвить по is_total, NULL-комнатную группу класть в
  by_room['unknown'] (отдельный бакет), аккумулировать через +=
  вместо assign (защита от будущих NULL-вариантов).
- Тесты: моки получили "is_total" поле (1 для grand-total, 0 иначе).

71/71 market_metrics тестов зелёные. ruff clean.

Closes #1214
2026-06-13 05:57:19 +00:00
8f3e461959 fix(competitors): _ACTIVE_STATUSES = русские значения как у домрф (#1213)
Some checks failed
Deploy / changes (push) Successful in 7s
Deploy / build-backend (push) Successful in 1m27s
Deploy / build-frontend (push) Has been skipped
Deploy / build-worker (push) Successful in 2m59s
Deploy / deploy (push) Has been cancelled
_ACTIVE_STATUSES = frozenset({"sales", "construction"}) — английский словарь
никогда не совпадал с domrf_kn_objects.site_status, который scraper берёт
СЫРЫМ из siteStatus дом.рф (domrf_kn.py:316). Реальные prod-значения
русские: «Строящиеся»/«Сданные».

Прод-аудит:
- data/sql/105_add_sales_started_flag.sql фильтрует по 'Строящиеся' (~1322 строки).
- partial index 66_indexes_recommend.sql использует те же.
- analytics_queries.py, MarketTab.tsx, CompetitorTable.tsx — все на русских.

Эффект: у ВСЕХ Competitor в POST /parcels/{cad}/competitors is_active=False
и CompetitorsSummary.active_count=0 при любых данных — типизированный
контракт систематически врал.

Patch: _ACTIVE_STATUSES = frozenset({"Строящиеся"}). Заодно обновил два
unit-теста которые кодировали баг (использовали "sales"/"construction"
в моках, тестировали логику против сломанного словаря). Теперь моки
матчат реальную prod-форму.

51/51 competitors-тестов зелёные. ruff clean.

Closes #1213
2026-06-13 05:53:16 +00:00
d587b9e199 fix(supply_layers): L3 future-supply берёт ПОСЛЕДНИЙ снапшот → потом фильтр (#1212)
Some checks failed
Deploy / build-frontend (push) Blocked by required conditions
Deploy / deploy (push) Blocked by required conditions
Deploy / changes (push) Successful in 10s
Deploy / build-worker (push) Has been cancelled
Deploy / build-backend (push) Has been cancelled
_L3_FUTURE_SQL применял volatile-фильтры (ready_dt > horizon, free_flats
<= threshold) в WHERE ВНУТРИ CTE с DISTINCT ON (obj_id) ORDER BY
snapshot_date DESC. Это значит фильтр применялся ДО DISTINCT ON →
бралась «последний снапшот, ПРОШЕДШИЙ фильтр», а не последний в принципе.

Эффект на проде: объект, когда-то бывший «объявлен, не продаётся»
(free_flats=0/NULL), остаётся в L3 future-supply даже после открытия
продаж — свежий снапшот с free_flats=180 отфильтрован, взят старый
с free_flats=0. Двойной счёт с L1/L2 (open + future одного объекта),
stale flat_count/ready_dt в supply_layers и форсайте.

Patch: разделил фильтры на стабильные (в CTE: region_cd, district_name)
и volatile (во внешнем WHERE после DISTINCT ON: ready_dt, free_flats).
Зеркало паттерна L2-CTE (там volatile уже снаружи). Семантика теперь
матчит docstring: «свежий снапшот, затем фильтр».

62/62 supply_layers тестов зелёные. ruff clean. SQL psycopg v3
(CAST(:x AS interval)) уже корректен.

Closes #1212
2026-06-13 05:53:01 +00:00
f12c03588d fix(llm): close §19 redaction regex gaps for bare phone / CAPS name / bare SNILS (#1207)
All checks were successful
Deploy / changes (push) Successful in 7s
Deploy / build-frontend (push) Has been skipped
CI / frontend-tests (push) Has been skipped
CI / frontend-tests (pull_request) Has been skipped
Deploy / build-worker (push) Successful in 2m54s
CI / changes (push) Successful in 8s
CI / changes (pull_request) Successful in 6s
CI / backend-tests (push) Successful in 6m22s
CI / backend-tests (pull_request) Successful in 6m25s
Deploy / build-backend (push) Successful in 1m46s
Deploy / deploy (push) Successful in 1m10s
`_PHONE_RE` требовал префикс +7/8 + хотя бы один разделитель → copy-paste форма
«79221234567»/«89221234567» (самая частая) уходила в OpenAI как есть. `_FULLNAME_RE`
требовал строчные после первой буквы → стандарт ЕГРН/паспортных выписок «ИВАНОВ
ИВАН ИВАНОВИЧ» не матчил. СНИЛС без разделителей не покрывался вообще (`_SNILS_RE`
ждёт формат «NNN-NNN-NNN NN»). Для chat-пути scrub — единственный барьер перед
OpenAI (SafePayload.text без allowlist'а), при llm_enabled=True PII покидает РФ
(нарушение §19/152-ФЗ).

Patch:
- _PHONE_BARE_RE: (?<!\d)[78]\d{10}(?!\d) — 11-значные с префиксом 7/8.
- _SNILS_BARE_RE: (?<!\d)\d{11}(?!\d) — любые 11-значные без разделителей.
- _FULLNAME_RE: альтернатива в каждом слове — Titlecase | CAPS-2+.
- Конфликт phone-bare vs SNILS-bare разрешён порядком в _PII_PATTERNS: phone
  раньше SNILS (7/8-префикс семантически точнее).
- INN (ровно 10/12) с SNILS-bare не пересекается по длине.

8 новых юнит-тестов покрывают все три дыры + конфликт-резолюцию. 24/24 redaction
тестов + 55/55 LLM-suite зелёные.

Closes #1207
2026-06-13 09:56:58 +05:00
31c5974c72 fix(forecasting): нормализовать dict-district в build_site_finder_report (#1130 follow-up)
All checks were successful
CI / changes (push) Successful in 7s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 7s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m26s
CI / backend-tests (pull_request) Successful in 6m22s
Deploy / build-backend (push) Successful in 1m45s
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-worker (push) Successful in 2m42s
Deploy / deploy (push) Successful in 1m12s
`analyze["district"]` в этой кодовой базе — dict вида
{"district_name": "Верх-Исетский", "dist_to_center": 0.0, "median_price_per_m2": ...}.
Штатный caller (`workers/tasks/forecast.py:123`) явно извлекает `district_name`:
`district = row.district or analyze["district"]["district_name"]`. Но новые callers
(тесты, расширения чата, ad-hoc эндпоинты) легко передают сырой dict без знания этой
конвенции — тогда внутри §9.x-слоёв compute_market_metrics(district=<dict>) падает
с TypeError: unhashable type: 'dict' в forecast_request_cache.wrapper,
`_safe_call` это проглатывает → секции future_market.forecasts_by_horizon=[] и
scenarios.by_scenario={} тихо остаются пустыми (silent degrade, не 500).

Добавлен `_normalize_district(district)` — pure-нормализация на входе оркестратора:
  - str → как есть;
  - None → None;
  - dict с district_name (непустая строка) → извлекаем;
  - dict без district_name / с пустым / неподдерживаемый тип → None + logger.warning.

7 unit-тестов в test_orchestrator.py::TestNormalizeDistrict (все варианты входов).
Не меняет поведение штатного caller'а (str → str), только защищает от случайных
dict-callers.

Discovered through: #1130 Phase A (мой первый тестовый скрипт со скормленным
сырым `analyze["district"]` dict выдал forecasts.n_horizons=0 + 15 TypeError'ов
в _safe_call). Закрывает чип task_4a4aa3bb.

Refs #1130
2026-06-13 09:24:55 +05:00
49b85ab1d6 perf(site-finder): air-quality TTL-кэш + neighbors_summary 2→1 SQL (#1130 Phase B)
All checks were successful
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 6s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m28s
CI / backend-tests (pull_request) Successful in 6m24s
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m37s
CI / changes (push) Successful in 6s
Deploy / build-worker (push) Successful in 3m3s
Deploy / deploy (push) Successful in 1m24s
Phase B продолжает hot-path latency cut'ы analyze_parcel поверх Phase A (PR #1194):

1. get_air_quality_cached в weather_cache (третий TTL-слот рядом с forecast/climate):
   - hit-TTL 1h (hourly bucket Open-Meteo), negative-TTL 5min.
   - timeout 5s → 2s.
   - single-flight под per-cache threading.Lock.
   - Снимает 0.26с на cache-hit и до 5с при DNS-fail (тот же антипаттерн,
     что был у forecast/climate до Phase A).

2. _neighbors_summary: 2 SQL → 1 statement.
   - WITH neighbors AS (... LIMIT 30), overlaps AS (... LIMIT 5)
   - + COALESCE(json_agg(row_to_json(...)), '[]'::json) в основном SELECT.
   - Один round-trip вместо двух (~-47 ms на каждый analyze).
   - Семантика идентична: те же ST_DWithin(geom, point, 100m) для neighbors,
     ST_Intersects + ST_Area для overlaps. Anti-fakes guards (_COST_PER_M2_MIN/MAX,
     float() try/except, >50 m² overlap) сохранены.

Формат возвращаемых dict не меняется (frontend контракт). 5 новых юнит-тестов в
test_weather_cache.py (TestAirQualityCache: hot-hit, negative, empty-hourly,
separate-from-weather, ttl-expires). 7 файлов API-тестов обновили patch-цель
_fetch_air_quality_sync → get_air_quality_cached.

Refs #1130
2026-06-12 22:42:32 +05:00
97ff0c9e81 perf(site-finder): TTL-кэш Open-Meteo в analyze hot-path (#1130 Phase A)
All checks were successful
CI / frontend-tests (pull_request) Has been skipped
CI / changes (push) Successful in 7s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 6s
CI / backend-tests (push) Successful in 6m28s
CI / backend-tests (pull_request) Successful in 6m26s
Deploy / build-frontend (push) Has been skipped
Deploy / changes (push) Successful in 6s
Deploy / build-backend (push) Successful in 1m41s
Deploy / build-worker (push) Successful in 4m51s
Deploy / deploy (push) Successful in 1m24s
Два внешних HTTP-вызова в analyze (_fetch_weather_sync + _fetch_seasonal_weather_sync)
сидели без кэша. На проде в private network с restricted egress DNS до *.open-meteo.com
фейлит и каждый replay висит полный timeout (5s + 15s = до 20с лишних на analyze).
Профиль cProfile одного replay: 2.38с в httpx + 1.73с в SSL recv.

Вынес логику в app/services/weather_cache.py:
- TTL hit: forecast 6h, climate normals 7d. Negative-cache: 5min (DNS-fail не повторяет
  timeout на каждый analyze, восстановление подхватывается через ~5 минут).
- Ключ — округлённые до 0.01° (~1 км) координаты.
- Single-flight: per-cache threading.Lock + check-then-fetch-then-store.
- Тайм-ауты сокращены: forecast 5s→2s, climate 15s→3s.

Контракт для caller'а не изменился: None по-прежнему допустим (env_ok-флаг в
_build_environmental_score). 7 файлов API-тестов обновили patch-цели на новые имена
get_weather_cached / get_seasonal_weather_cached. 11 новых юнит-тестов в
tests/services/test_weather_cache.py (single-flight через threading.Barrier(16), TTL
expire через monkeypatch _now, изоляция forecast vs climate).

Refs #1130
2026-06-12 19:58:07 +05:00
04da30b57a fix(sf): psql stdin-bootstrap пароля gendesign_reader + COALESCE-protect внешних id (review fixup, #976)
All checks were successful
CI / changes (pull_request) Successful in 6s
CI / changes (push) Successful in 7s
CI / frontend-tests (push) Has been skipped
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m23s
CI / backend-tests (pull_request) Successful in 6m21s
- Создан tradein-mvp/ops/db-bootstrap/set_gendesign_reader_password.sql:
  set_config GUC + format('%L') + \o /dev/null — зеркальный паттерн
  ops/db-bootstrap/set_tradein_fdw_password.sql. Пароль не попадает
  в stdout/logs (leak-protected).
- deploy-tradein.yml: шаг bootstrap переписан на stdin-redirect
  (< ops/db-bootstrap/set_gendesign_reader_password.sql) вместо сломанного
  -c "ALTER ROLE ... PASSWORD :'TRADEIN_READER_PASSWORD'", которое давало
  syntax error at ':' (psql не интерполирует :'var' внутри -c строки).
- newbuilding_crossload.py ON CONFLICT DO UPDATE: yandex_jk_id и
  cian_internal_house_id → COALESCE(EXCLUDED.col, table.col), чтобы
  cian-запись не затирала yandex_jk_id NULL'ом и наоборот.
- Тест test_upsert_sql_coalesce_external_ids проверяет наличие COALESCE
  в DO UPDATE секции для обоих id.
2026-06-12 11:26:40 +03:00
1a4778d006 feat(sf): cross-load ETL tradein→gendesign newbuilding_listings — вариант A (#976)
- docker-compose.prod.yml: backend + worker добавлены в сеть gendesign_shared
  (нужен прямой TCP к tradein-postgres для psycopg ETL)
- tradein-mvp/data/sql/101_gendesign_reader_role.sql: роль gendesign_reader
  (LOGIN, SELECT на houses/house_sources/houses_price_dynamics/
   house_reliability_checks/house_reviews; пароль через TRADEIN_READER_PASSWORD
   bootstrap в deploy-tradein.yml)
- .forgejo/workflows/deploy-tradein.yml: bootstrap ALTER ROLE gendesign_reader
  PASSWORD из env после миграций; warn (не exit 1) если переменная не задана
- backend/app/core/config.py: settings.tradein_database_url (env TRADEIN_DATABASE_URL)
- backend/app/services/etl/newbuilding_crossload.py: run_crossload() — psycopg
  server-side cursor, batch 500, UPSERT ON CONFLICT (source, ext_house_id),
  SAVEPOINT per-row; disabled-режим если url пуст
- backend/app/workers/tasks/etl_newbuilding_crossload.py: Celery task
- backend/app/workers/beat_schedule.py: nightly 03:30 МСК (00:30 UTC)
- backend/app/api/v1/admin_scrape.py: POST /admin/scrape/newbuilding-crossload
- backend/tests/services/test_newbuilding_crossload.py: 10 тестов без реальной БД

needs-human: TRADEIN_READER_PASSWORD → tradein .env.runtime;
             TRADEIN_DATABASE_URL → gendesign backend/.env.runtime
2026-06-12 11:18:47 +03:00
c9720a6212 feat(forecasting): activate §25.3 cannibalization unit-mix axis (4/4), gated (#1169)
All checks were successful
Deploy / build-worker (push) Successful in 3m1s
CI / changes (push) Successful in 6s
CI / frontend-tests (push) Has been skipped
CI / changes (pull_request) Successful in 5s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (push) Successful in 6m32s
CI / backend-tests (pull_request) Successful in 6m24s
Deploy / changes (push) Successful in 7s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 2m1s
Deploy / deploy (push) Successful in 1m25s
Feed candidate_unit_mix into _build_cannibalization (mirrors how #1169 fed
candidate_release_month from the launch window), completing §25.3 to all 4 axes:
class + price + timing + unit-mix, plus geo weight.

- candidate mix from recommend_mix "buckets[].share_pct" (same rule-based
  квартирография as §22 product_tz), extracted + normalized to {bucket: share}.
- _canonical_room_bucket folds recommend_mix RU labels ("1-к 30-45", "80+ м²")
  and manual own_planned_project Latin keys ("1k") into one room-count space —
  without it the L1 similarity would silently be 0 (disjoint keys).
- recommend_mix is HEAVY, so it's GATED: derived only when the own-portfolio has
  >=1 project with a non-empty unit_mix; get_own_portfolio fetched once in
  compute_special_indices and threaded into _build_cannibalization (no double
  fetch). With OWN_DEVELOPER_IDS unset (portfolio empty) → zero added cost on the
  hot §22 report path.
- Graceful (recommend_mix None/empty/raises → axis excluded, None-not-0),
  deterministic. Unit-mix only fires for manual-future own-projects with a mix
  (domrf-current carry unit_mix=None) — expected narrowness, documented.

205 tests; ruff + mypy clean. Scope: special_indices.py + test only; no deps.

Refs #1169
2026-06-09 16:09:29 +05:00
645b3c14a3 fix(chat): clean квартирография rendering in chat_system prompt (v2, #957)
All checks were successful
CI / frontend-tests (push) Has been skipped
CI / changes (push) Successful in 6s
CI / frontend-tests (pull_request) Has been skipped
CI / changes (pull_request) Successful in 6s
CI / backend-tests (pull_request) Successful in 6m24s
Deploy / build-frontend (push) Has been skipped
Deploy / build-worker (push) Successful in 3m5s
Deploy / deploy (push) Successful in 1m49s
CI / backend-tests (push) Successful in 6m24s
Deploy / changes (push) Successful in 5s
Deploy / build-backend (push) Successful in 2m8s
Live chat rendered the recommended unit-mix as a garbled rank-hyphen-type
list ("2-1-к", "3-2-к" = "rank 2: 1-к"). Add rule 6 to chat_system@v2:
list квартирография/segments by clean TYPE (студия / 1-к / 2-к / 3-к /
80+ м²) + share %, no ordinal-number-hyphen-type concatenation. Bumped
prompt version 1->2 (versioned-prompt convention). Test locks v2 + the rule.

Refs #957
2026-06-09 15:07:14 +05:00
cd48a095c0 feat(forecasting): activate §25.3 cannibalization timing axis from launch window (#1169)
Some checks failed
Deploy / build-worker (push) Successful in 2m32s
Deploy / changes (push) Successful in 6s
Deploy / build-frontend (push) Has been skipped
Deploy / build-backend (push) Successful in 1m32s
Deploy / deploy (push) Has been cancelled
Derive candidate_release_month = report-as-of (date.today()) + §25.1 Launch
Window peak-deficit horizon, threaded into _build_cannibalization so the timing
overlap axis activates against own-portfolio release_month (near-in-time own
projects raise cannibalization risk). Launch Window now computed once in
compute_special_indices and reused (no double-compute). Launch Window
unavailable -> candidate_release_month None -> timing axis gracefully excluded
(None-not-0); cannibalization still scores on class/price/geo. Adds stdlib
_add_months helper (year-boundary safe, no new dep). Deterministic. 168 tests.

§25.3 now: class+price+timing+geo active; unit-mix remains phase-2.

Refs #1169
2026-06-09 07:10:59 +00:00
f9e045028c feat(macro): land construction price index (СМР) via open Rosstat xlsx (#946)
All checks were successful
CI / changes (push) Successful in 6s
CI / changes (pull_request) Successful in 7s
CI / frontend-tests (pull_request) Has been skipped
CI / backend-tests (pull_request) Successful in 6m27s
Deploy / build-frontend (push) Has been skipped
CI / frontend-tests (push) Has been skipped
CI / backend-tests (push) Successful in 6m29s
Deploy / changes (push) Successful in 6s
Deploy / build-backend (push) Successful in 1m43s
Deploy / build-worker (push) Successful in 3m3s
Deploy / deploy (push) Successful in 1m15s
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
2026-06-09 11:28:05 +05:00