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
_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
`_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
`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
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
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
- 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.
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
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
Add the LLM prose-composition path for the parcel-forecast chat, layered
over the deterministic Step-1 fallback which stays the safety net.
- chat/tools.py: 5 read-only section tools (exec_summary, product_recommendation,
forecast, risks, scenarios) — pure slices of the loaded report dict, no DB/
recompute, graceful on missing sections. market_now (raw analyze blob) and meta
are deliberately NOT exposed -> highest-PII data cannot reach the LLM.
- chat/safe_payload.py: the §19 gate — single place that builds the outbound
SafePayload from a section-aggregate allowlist; honors is_confidential hard-block.
- chat/orchestrator.py: manual tool-call loop with call-cap/termination, real
grounded_in provenance; any LLMResult.ok=False (disabled/timeout/rate_limited/
redaction_refused/call_cap/provider_error/empty) degrades to the deterministic answer.
- llm/prompts.py: versioned chat_system@v1 — answer only from sections, never
fabricate numbers, advisory tone, decline out-of-scope.
- api/v1/chat.py: branch on settings.llm_enabled; sync complete bridged via
run_in_threadpool. Default-off -> deterministic path, no provider built.
- Tests: fake provider only (no network), planted-secret redaction-boundary +
per-reason fallback + call-cap + numbers-from-report coverage.
Refs #957
Step 1 of #957. Answers parcel questions by reading the already-persisted
§22 SiteFinderReport (latest_run_for, schema "1.0", read-only) and returning
templated RU answers with engine numbers verbatim (§16, never fabricated).
Intent routing (explicit or RU keyword match) -> per-section renderers;
graceful on partial/missing sections and pending (no run / DB error) without
500s. Works with llm_enabled=False (llm_used always False); LLM composition
is Step 2. Mounted off /api/v1/chat so rbac_guard auto-requires an authed
known user.
Refs #957
Upgrade the §25.3 cannibalization index from a same-class-competitor proxy to
true own-portfolio overlap: score the recommended candidate segment against the
developer's own portfolio (get_own_portfolio, #1169 PR1) across axes —
audience/class (ordinal distance), price ₽/м² (interval overlap), unit-mix
(L1 similarity), timing (half-life decay) — geo-weighted by haversine proximity
to the parcel, aggregated by geo-weighted soft-max (the strongest nearby
cannibalizer dominates, not a diluting mean). Empty portfolio -> labelled proxy
fallback, confidence forced low, never presented as the true index.
Pure scoring fns unit-tested without DB; None-not-0 on missing axes; thin/
only-current portfolio -> low confidence + §26 note; deterministic (sorted
tie-breaks, no RNG). class+price+geo active now; unit-mix+timing plumbed via
optional params for a follow-up that wires them from the orchestrator horizon.
ruff + mypy clean; 151 special-index tests pass (964 forecasting dir, no regr).
Refs #1169
Foundation PR: unified "our projects" source the §25.3 overlap engine (PR2)
will consume. Two origins normalized to OwnProject (class/timing/price/unit-mix):
- current <- domrf_kn_objects filtered by settings.own_developer_ids (numeric
prefix of composite dev_id; empty -> [] graceful, no DB hit, no hardcoded id)
- future <- new manual-entry own_planned_project entity (migration 148)
Adds OWN_DEVELOPER_IDS config (comma-sep -> list[int], default []),
own_planned_project table (range/unit_mix CHECKs via IMMUTABLE helper, generated
geom), /api/v1/own-projects CRUD (created_by from X-Authenticated-User), and
get_own_portfolio(db). Per-source graceful degradation; psycopg-v3 CAST clean.
Does not touch special_indices.py or parcels.py (out of scope).
Refs #1169
demand_index used a fixed clamp01(velocity/50); on a live prod refresh
all 8 ЕКБ districts sold ≥50/mo so it saturated to 1.0 everywhere — zero
discrimination between districts. Redesign to mirror infra_index:
normalize each district's unit_velocity against the city reference (MAX
district velocity per refresh run), so demand always discriminates and
self-calibrates as the market grows (no magic constant to rot).
- normalize_demand(velocity, *, city_reference_velocity), pure + graceful
(None stays None; reference<=0 -> honest 0.0, no ZeroDivisionError)
- refresh_locations now two-pass: collect velocities (one
compute_market_metrics per district, no O(n^2)), derive city reference,
normalize + upsert; SAVEPOINT-per-row and counters preserved
- remove _DEMAND_SATURATION_UPM constant; log city_reference_velocity
- tests: rewrite demand normalization + add end-to-end city-relative
suite incl. discrimination regression guarding the all-1.0 prod bug
Refs #948
Promote district to a first-class `location` entity (ТЗ §8.2), ADDITIVE — no
district->FK refactor. New `location` table keyed by district_name (joinable by
string), carrying 4 normalized [0,1] indices (NULL when no data, never 0):
- infra: _district_poi_score / _city_avg_poi_score (per-district POI aggregate)
- competition: market_metrics.overstock_index (available/stuck competing supply
— orthogonal to demand; NOT sell-through, which is market-heat correlated w/ demand)
- demand: market_metrics.unit_velocity (saturating /50)
- future_supply: future_supply_pressure.index (passthrough, already 0..1)
- data/sql/146_location.sql: idempotent table + UNIQUE(district_name) + range
CHECK + centroid GIST
- services/site_finder/locations.py: compute_location_indices (reuses forecast
per-district fns) + refresh_locations (SAVEPOINT per-row, CAST, ON CONFLICT)
- workers/tasks/location_refresh.py + beat (Mon 07:00 MSK, after supply-layers)
- api/v1/locations.py: read-only GET list + GET by name (analyst+admin via rbac,
frontend pilot-gated)
- tests: 34 (normalization 0..1/null, competition⊥demand orthogonality, idempotent
upsert, read API list/by-name/404)
Part of #948 (Part A insight shipped #1164).
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.
Cold §22 forecast measured ~215-233s on prod: §9.x layers re-execute the same
horizon/segment-invariant DB loads with identical args hundreds of times per
report (profiled: get_competitors x69, market_metrics x124, get_monthly_macro
x290). Add a per-report ContextVar cache (forecast_cache(), opened once in the
orchestrator) + @cached(key_builder) on the expensive §9.x loaders so each
unique load runs ONCE and reuses the same frozen, read-only instance.
Output is byte-identical (memoized producers are frozen dataclasses / read-only
Pydantic, callers never mutate; cache is per-report, discarded on exit; no-op
outside the report build). No concurrency, no signature changes.
- forecast_request_cache.py: ContextVar cache + cached() decorator (no-op
outside context, reentrant, _MISS sentinel for cached None)
- @cached on competitors/future_supply/market_metrics/macro_series/
sales_series/macro_coefficient/demand_normalization/regression loaders
- orchestrator: wrap build_site_finder_report in forecast_cache()
- 58 tests: key discrimination (call-counting regression guard), no-op-outside,
per-report isolation, reentrancy, frozen-producer canary, amplification proof
(real get_monthly_macro xN->1)
code-reviewer APPROVE (keys correct, mutation-safe, output identical). 1265
forecast/cache tests green. No new deps. Refs #1129.
render_report_docx (python-docx) mirrors report_md/report_pdf section order &
content, reuses report_pdf pure helpers (DRY), graceful on thin/empty report.
Widen /forecast/export format Literal to include docx → Word attachment.
Add python-docx dep + regenerate uv.lock (uv sync --frozen passes). Part of #959.
render_report_telegram_summary (pure, no new deps, DRY-reuses report_pdf str
helpers) + `tg` format on GET /{cad}/forecast/export → inline text/plain snippet
(no attachment, copy-paste-ready). md/json unchanged; no-run 404, bad format 422.
Graceful on thin/empty reports. Part of EPIC #959.
render_report_markdown (pure, no new runtime deps) reuses report_pdf's str
helpers (DRY), + GET /{cad_num}/forecast/export?format=md|json. No forecast run
→ 404; graceful on thin/empty reports; GFM-safe table escaping. PDF/XLSX already
existed; this adds the cheapest no-dep formats. Part of EPIC #959.
