bot-backend
03e43f12b5
- yandex_detail.py: добавить save_detail_enrichment(db, listing_id, enrichment) → bool; UPDATE listings COALESCE по 17 полям (rooms/area_m2/floor/total_floors/address/ description/repair_state/publish_date/views_total_yandex/publish_date_relative/ agency_name/agency_founded_year/agency_objects_count/metro_stations/photo_urls/ newbuilding_url/newbuilding_id); metro/photos → jsonb; detail_enriched_at = NOW() - app/tasks/yandex_detail_backfill.py: новый task run_yandex_detail_backfill; snapshot SELECT source='yandex' AND detail_enriched_at IS NULL; YandexDetailScraper async context manager; consecutive None abort (max_consecutive_blocks=5); budget guard; run lifecycle update_heartbeat/mark_done/mark_failed - scheduler.py: trigger_yandex_detail_backfill_run + elif source=='yandex_detail_backfill' в scheduler_loop; window 12-15 UTC (после avito_detail 09-12 UTC) - migration 113: ADD COLUMN IF NOT EXISTS photo_urls jsonb, newbuilding_url text, newbuilding_id text; seed scrape_schedules yandex_detail_backfill enabled=true - tests: 10 тестов (8 backfill + 2 save_detail_enrichment); 1819 passed 0 failed Closes #1553
357 lines
12 KiB
Python
357 lines
12 KiB
Python
"""Tests for app.tasks.yandex_detail_backfill and save_detail_enrichment."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
import os
|
|
import sys
|
|
from unittest.mock import AsyncMock, MagicMock, patch
|
|
|
|
os.environ.setdefault("DATABASE_URL", "postgresql+psycopg://test:test@localhost:5432/test")
|
|
|
|
_wp_mock = MagicMock()
|
|
sys.modules.setdefault("weasyprint", _wp_mock)
|
|
|
|
import pytest # noqa: E402
|
|
|
|
from app.tasks.yandex_detail_backfill import ( # noqa: E402
|
|
YandexDetailBackfillResult,
|
|
run_yandex_detail_backfill,
|
|
)
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Path constants for patching
|
|
# ---------------------------------------------------------------------------
|
|
|
|
_FETCH = "app.tasks.yandex_detail_backfill.YandexDetailScraper"
|
|
_SAVE = "app.tasks.yandex_detail_backfill.save_detail_enrichment"
|
|
_RUNS = "app.tasks.yandex_detail_backfill.runs_mod"
|
|
_SLEEP = "app.tasks.yandex_detail_backfill.asyncio.sleep"
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Helpers
|
|
# ---------------------------------------------------------------------------
|
|
|
|
|
|
def _make_snapshot(n: int) -> list[dict]:
|
|
return [
|
|
{"id": i + 1, "source_url": f"https://realty.yandex.ru/offer/{i + 1}/"} for i in range(n)
|
|
]
|
|
|
|
|
|
def _mock_db(snapshot: list[dict]) -> MagicMock:
|
|
"""Fake Session: first execute() returns snapshot via .mappings().all()."""
|
|
db = MagicMock()
|
|
sel = MagicMock()
|
|
sel.mappings.return_value.all.return_value = snapshot
|
|
db.execute.return_value = sel
|
|
return db
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Tests: run_yandex_detail_backfill
|
|
# ---------------------------------------------------------------------------
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_empty_snapshot_marks_done() -> None:
|
|
"""Empty snapshot -> mark_done immediately, no fetch calls."""
|
|
db = _mock_db([])
|
|
runs = MagicMock()
|
|
mock_scraper_cls = MagicMock()
|
|
mock_scraper_inst = AsyncMock()
|
|
mock_scraper_cls.return_value.__aenter__ = AsyncMock(return_value=mock_scraper_inst)
|
|
mock_scraper_cls.return_value.__aexit__ = AsyncMock(return_value=None)
|
|
|
|
with (
|
|
patch(_FETCH, mock_scraper_cls),
|
|
patch(_RUNS, runs),
|
|
):
|
|
result = await run_yandex_detail_backfill(
|
|
db, run_id=1, params={"batch_size": 10, "budget_sec": 60}
|
|
)
|
|
|
|
assert isinstance(result, YandexDetailBackfillResult)
|
|
assert result.attempted == 0
|
|
assert result.enriched == 0
|
|
mock_scraper_inst.fetch_detail.assert_not_called()
|
|
runs.mark_done.assert_called_once()
|
|
runs.mark_failed.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_processes_snapshot_to_completion() -> None:
|
|
"""3 listings -> all fetched and enriched, mark_done called."""
|
|
snapshot = _make_snapshot(3)
|
|
db = _mock_db(snapshot)
|
|
runs = MagicMock()
|
|
mock_enrichment = MagicMock()
|
|
|
|
mock_scraper_cls = MagicMock()
|
|
mock_scraper_inst = AsyncMock()
|
|
mock_scraper_inst.fetch_detail = AsyncMock(return_value=mock_enrichment)
|
|
mock_scraper_cls.return_value.__aenter__ = AsyncMock(return_value=mock_scraper_inst)
|
|
mock_scraper_cls.return_value.__aexit__ = AsyncMock(return_value=None)
|
|
|
|
with (
|
|
patch(_FETCH, mock_scraper_cls),
|
|
patch(_RUNS, runs),
|
|
patch(_SAVE, return_value=True),
|
|
patch(_SLEEP, new_callable=AsyncMock),
|
|
):
|
|
result = await run_yandex_detail_backfill(
|
|
db, run_id=2, params={"batch_size": 10, "budget_sec": 3600}
|
|
)
|
|
|
|
assert result.attempted == 3
|
|
assert result.enriched == 3
|
|
assert result.failed == 0
|
|
assert mock_scraper_inst.fetch_detail.call_count == 3
|
|
runs.mark_done.assert_called_once()
|
|
runs.mark_failed.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_none_abort_after_max_consecutive() -> None:
|
|
"""5 consecutive None results -> abort, mark_done (NOT mark_failed)."""
|
|
snapshot = _make_snapshot(10)
|
|
db = _mock_db(snapshot)
|
|
runs = MagicMock()
|
|
|
|
mock_scraper_cls = MagicMock()
|
|
mock_scraper_inst = AsyncMock()
|
|
mock_scraper_inst.fetch_detail = AsyncMock(return_value=None)
|
|
mock_scraper_cls.return_value.__aenter__ = AsyncMock(return_value=mock_scraper_inst)
|
|
mock_scraper_cls.return_value.__aexit__ = AsyncMock(return_value=None)
|
|
|
|
with (
|
|
patch(_FETCH, mock_scraper_cls),
|
|
patch(_RUNS, runs),
|
|
patch(_SLEEP, new_callable=AsyncMock),
|
|
):
|
|
result = await run_yandex_detail_backfill(
|
|
db,
|
|
run_id=3,
|
|
params={"batch_size": 10, "budget_sec": 3600, "max_consecutive_blocks": 5},
|
|
)
|
|
|
|
assert result.attempted == 5
|
|
assert result.failed == 5
|
|
assert result.enriched == 0
|
|
runs.mark_done.assert_called_once()
|
|
runs.mark_failed.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_none_resets_on_success() -> None:
|
|
"""1 None + 1 success + 1 None -> consecutive resets, no abort after 1 success."""
|
|
snapshot = _make_snapshot(3)
|
|
db = _mock_db(snapshot)
|
|
runs = MagicMock()
|
|
mock_enrichment = MagicMock()
|
|
|
|
mock_scraper_cls = MagicMock()
|
|
mock_scraper_inst = AsyncMock()
|
|
mock_scraper_inst.fetch_detail = AsyncMock(side_effect=[None, mock_enrichment, None])
|
|
mock_scraper_cls.return_value.__aenter__ = AsyncMock(return_value=mock_scraper_inst)
|
|
mock_scraper_cls.return_value.__aexit__ = AsyncMock(return_value=None)
|
|
|
|
with (
|
|
patch(_FETCH, mock_scraper_cls),
|
|
patch(_RUNS, runs),
|
|
patch(_SAVE, return_value=True),
|
|
patch(_SLEEP, new_callable=AsyncMock),
|
|
):
|
|
result = await run_yandex_detail_backfill(
|
|
db,
|
|
run_id=4,
|
|
params={"batch_size": 10, "budget_sec": 3600, "max_consecutive_blocks": 5},
|
|
)
|
|
|
|
assert result.attempted == 3
|
|
assert result.enriched == 1
|
|
assert result.failed == 2
|
|
runs.mark_done.assert_called_once()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_budget_guard_stops_loop() -> None:
|
|
"""Budget expired before first listing -> fetch_detail not called."""
|
|
snapshot = _make_snapshot(5)
|
|
db = _mock_db(snapshot)
|
|
runs = MagicMock()
|
|
|
|
mock_scraper_cls = MagicMock()
|
|
mock_scraper_inst = AsyncMock()
|
|
mock_scraper_cls.return_value.__aenter__ = AsyncMock(return_value=mock_scraper_inst)
|
|
mock_scraper_cls.return_value.__aexit__ = AsyncMock(return_value=None)
|
|
|
|
mono_values = iter([0.0, 999.0, 999.0])
|
|
with (
|
|
patch(_FETCH, mock_scraper_cls),
|
|
patch(_RUNS, runs),
|
|
patch("app.tasks.yandex_detail_backfill.time.monotonic", side_effect=mono_values),
|
|
):
|
|
await run_yandex_detail_backfill(db, run_id=5, params={"batch_size": 5, "budget_sec": 1})
|
|
|
|
mock_scraper_inst.fetch_detail.assert_not_called()
|
|
runs.mark_done.assert_called_once()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_top_level_exception_marks_failed() -> None:
|
|
"""db.execute raises -> mark_failed called, exception re-raised."""
|
|
db = MagicMock()
|
|
db.execute.side_effect = RuntimeError("DB connection lost")
|
|
runs = MagicMock()
|
|
|
|
with (
|
|
patch(_RUNS, runs),
|
|
):
|
|
with pytest.raises(RuntimeError, match="DB connection lost"):
|
|
await run_yandex_detail_backfill(
|
|
db, run_id=6, params={"batch_size": 5, "budget_sec": 60}
|
|
)
|
|
|
|
runs.mark_failed.assert_called_once()
|
|
runs.mark_done.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_backfill_fetch_exception_continues() -> None:
|
|
"""RuntimeError on one listing -> failed++, db.rollback(), loop continues for next."""
|
|
snapshot = _make_snapshot(2)
|
|
db = _mock_db(snapshot)
|
|
runs = MagicMock()
|
|
mock_enrichment = MagicMock()
|
|
|
|
mock_scraper_cls = MagicMock()
|
|
mock_scraper_inst = AsyncMock()
|
|
mock_scraper_inst.fetch_detail = AsyncMock(
|
|
side_effect=[RuntimeError("parse error"), mock_enrichment]
|
|
)
|
|
mock_scraper_cls.return_value.__aenter__ = AsyncMock(return_value=mock_scraper_inst)
|
|
mock_scraper_cls.return_value.__aexit__ = AsyncMock(return_value=None)
|
|
|
|
with (
|
|
patch(_FETCH, mock_scraper_cls),
|
|
patch(_RUNS, runs),
|
|
patch(_SAVE, return_value=True),
|
|
patch(_SLEEP, new_callable=AsyncMock),
|
|
):
|
|
result = await run_yandex_detail_backfill(
|
|
db, run_id=7, params={"batch_size": 10, "budget_sec": 3600}
|
|
)
|
|
|
|
assert result.failed == 1
|
|
assert result.enriched == 1
|
|
assert result.attempted == 2
|
|
db.rollback.assert_called()
|
|
runs.mark_done.assert_called_once()
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Tests: save_detail_enrichment
|
|
# ---------------------------------------------------------------------------
|
|
|
|
|
|
def test_save_detail_enrichment_maps_fields_to_update() -> None:
|
|
"""save_detail_enrichment calls db.execute with UPDATE and commits."""
|
|
from app.services.scrapers.yandex_detail import (
|
|
DetailEnrichment,
|
|
MetroStation,
|
|
save_detail_enrichment,
|
|
)
|
|
|
|
enrichment = DetailEnrichment(
|
|
offer_id="12345",
|
|
source_url="https://realty.yandex.ru/offer/12345/",
|
|
rooms=2,
|
|
area_m2=54.5,
|
|
floor=3,
|
|
total_floors=9,
|
|
address="Екатеринбург, ул. Ленина, 10",
|
|
description="Хорошая квартира",
|
|
repair_state="standard",
|
|
publish_date=None,
|
|
views_total=42,
|
|
publish_date_relative="вчера",
|
|
agency_name="Агентство «Тест»",
|
|
agency_founded_year=2005,
|
|
agency_objects_count=150,
|
|
metro_stations=[MetroStation(name="Чкаловская", walk_min=11)],
|
|
photo_urls=["https://example.com/1.jpg", "https://example.com/2.jpg"],
|
|
newbuilding_url="https://realty.yandex.ru/kupit/novostrojka/test-12345/",
|
|
newbuilding_id="12345",
|
|
)
|
|
|
|
db = MagicMock()
|
|
result_mock = MagicMock()
|
|
result_mock.rowcount = 1
|
|
db.execute.return_value = result_mock
|
|
|
|
saved = save_detail_enrichment(db, listing_id=99, e=enrichment)
|
|
|
|
assert saved is True
|
|
db.execute.assert_called_once()
|
|
db.commit.assert_called_once()
|
|
|
|
# Verify the SQL call contains metro_stations and photo_urls as JSON
|
|
call_args = db.execute.call_args
|
|
params = call_args[0][1] # positional second arg to execute()
|
|
assert params["rooms"] == 2
|
|
assert params["area_m2"] == 54.5
|
|
assert params["agency_name"] == "Агентство «Тест»"
|
|
|
|
# metro_stations must be JSON-serialised
|
|
metro_val = params["metro_stations"]
|
|
assert metro_val is not None
|
|
parsed_metro = json.loads(metro_val)
|
|
assert parsed_metro[0]["name"] == "Чкаловская"
|
|
assert parsed_metro[0]["walk_min"] == 11
|
|
|
|
# photo_urls must be JSON-serialised
|
|
photo_val = params["photo_urls"]
|
|
assert photo_val is not None
|
|
parsed_photos = json.loads(photo_val)
|
|
assert len(parsed_photos) == 2
|
|
assert parsed_photos[0] == "https://example.com/1.jpg"
|
|
|
|
|
|
def test_save_detail_enrichment_empty_metro_and_photos() -> None:
|
|
"""Empty metro_stations and photo_urls -> NULL passed for both jsonb columns."""
|
|
from app.services.scrapers.yandex_detail import DetailEnrichment, save_detail_enrichment
|
|
|
|
enrichment = DetailEnrichment(
|
|
offer_id="99",
|
|
source_url="https://realty.yandex.ru/offer/99/",
|
|
)
|
|
|
|
db = MagicMock()
|
|
result_mock = MagicMock()
|
|
result_mock.rowcount = 1
|
|
db.execute.return_value = result_mock
|
|
|
|
save_detail_enrichment(db, listing_id=1, e=enrichment)
|
|
|
|
params = db.execute.call_args[0][1]
|
|
assert params["metro_stations"] is None
|
|
assert params["photo_urls"] is None
|
|
|
|
|
|
def test_save_detail_enrichment_rowcount_zero_returns_false() -> None:
|
|
"""rowcount=0 (listing_id not found) -> returns False."""
|
|
from app.services.scrapers.yandex_detail import DetailEnrichment, save_detail_enrichment
|
|
|
|
enrichment = DetailEnrichment(
|
|
offer_id="404",
|
|
source_url="https://realty.yandex.ru/offer/404/",
|
|
)
|
|
|
|
db = MagicMock()
|
|
result_mock = MagicMock()
|
|
result_mock.rowcount = 0
|
|
db.execute.return_value = result_mock
|
|
|
|
saved = save_detail_enrichment(db, listing_id=404, e=enrichment)
|
|
assert saved is False
|