gendesign/tradein-mvp/backend/tests/tasks
bot-backend 81fc5e4014
All checks were successful
CI / changes (pull_request) Successful in 6s
CI / backend-tests (pull_request) Has been skipped
CI / frontend-tests (pull_request) Has been skipped
CI / openapi-codegen-check (pull_request) Has been skipped
fix(tradein/avito): detect HTTP-200 soft-block as block + coords-priority backfill ordering (#1950, #1967)
Под нагрузкой Avito отдаёт HTTP 200 с общей витриной ("Авито — Объявления
на сайте Авито", без item-view) вместо detail-страницы — мягкий анти-бот
deflect. _is_firewall_page (3 IP-block маркера) её не ловил → parse падал
ValueError("Cannot extract item_id") → generic failed, blocked=0, IP не
ротировался, enriched каскадно падал к 0 (прод runs 431→437: 115→0).

FIX 2: _is_detail_soft_block() (нет item-view + generic <title> в первых 16KB)
OR-ится в is_firewall в обоих путях fetch_detail (curl + browser) → soft-block
проходит существующую 403/firewall reconnect-машинерию (свежий backconnect
exit-IP) либо после MAX_RETRIES → AvitoBlockedError → loop-ротация.

FIX 1: snapshot ORDER BY (lat IS NULL) DESC, scraped_at DESC NULLS LAST —
сперва листинги без координат (detail-страница даёт координаты здания, #1967).

Tests: test_avito_detail_soft_block.py (reconnect-escape, exhausted->blocked,
unit-detector) + lat-priority assert в snapshot-тесте.
2026-06-28 10:51:18 +03:00
..
__init__.py fix(tradein): geocode backfill + remove Avito exclusion from estimator (#490) 2026-05-23 20:03:48 +00:00
test_avito_detail_backfill.py fix(tradein/avito): detect HTTP-200 soft-block as block + coords-priority backfill ordering (#1950, #1967) 2026-06-28 10:51:18 +03:00
test_backfill_listings_coords_geoportal.py feat(tradein): бэкфилл координат листингов из ekb_geoportal_buildings (street+house match) 2026-06-20 17:08:46 +03:00
test_cadastral_geo_match.py fix(cadastral): single-statement match UPDATE — OFFSET chunk skipped rows 2026-06-18 10:28:19 +03:00
test_geocode_missing.py fix(tradein): wire nightly geocode backfill для listings в in-app scheduler 2026-06-16 10:01:49 +03:00
test_newbuilding_enrich_backfill.py feat(tradein-scheduler): nightly newbuilding-enrichment schedule (#973) 2026-06-13 16:24:03 +05:00
test_yandex_address_backfill.py fix(tradein): yandex address-enrich — city-agnostic title regex (спутники ЕКБ) (#875) 2026-05-31 09:08:29 +00:00
test_yandex_detail_backfill.py fix(tradein): yandex detail backfill — заменить fetch_detail на curl_cffi+proxy 2026-06-16 14:05:22 +03:00
test_yandex_newbuilding_sweep.py fix(tradein): yandex sweep — guard ext_id=NULL перед fetch_jk + cleanup httpx (#974 review M1/M3) 2026-06-15 18:41:08 +03:00