gendesign/tradein-mvp/backend/tests/scrapers/test_avito_anti_bot.py
lekss361 5bf650c843 fix(tradein): route sweep shared-session (SERP+detail) через mobile proxy (#623)
Корень банов city-sweep: pipeline создаёт собственную shared AsyncSession и
присваивает её scraper._cffi, ПЕРЕЗАТИРАЯ проксированную сессию из
AvitoScraper.__aenter__ (#788). Через эту же сессию идёт и SERP, и detail-энричмент
(fetch_detail), и houses. Сессия создавалась без proxies → detail-страницы летели
с datacenter-IP → HTTP 429 → весь run mark_banned (scrape_runs id=9: «Avito detail
HTTP 429 …»). Единичный /admin/scrape работал, т.к. зовёт fetch_around напрямую
(проксированный _cffi из __aenter__), а не pipeline.

- scrape_pipeline.py: helper _avito_proxies() (из settings.avito_proxy_url) →
  proxies= в обе shared AsyncSession (run_avito_pipeline + run_avito_city_sweep).
  Пусто/нет env → прямое подключение (dev no-op). n1/yandex не трогаем (им прокси
  не нужен, не жжём мобильный трафик).
- test: _avito_proxies None без env / {http,https} при заданном URL.

Проверено: 18 passed (anti-bot + serp_date). ruff clean. Требует deploy (cron
city-sweep каждые 2ч на проде продолжит банить detail-429 до выката).

Refs #623, #788, #562.
2026-05-30 21:02:26 +03:00

432 lines
16 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

"""Test anti-bot hardening: blocked exceptions + early abort."""
from unittest.mock import AsyncMock, MagicMock, call, patch
import pytest
from app.services.scrapers.avito_exceptions import (
AvitoBlockedError,
AvitoContentBlockedError,
AvitoRateLimitedError,
)
# ── fetch_detail tests ──────────────────────────────────────────
@pytest.mark.asyncio
async def test_fetch_detail_raises_blocked_on_403() -> None:
"""fetch_detail должен raise AvitoBlockedError на HTTP 403."""
from app.services.scrapers.avito_detail import fetch_detail
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 403
mock_session.get = AsyncMock(return_value=mock_response)
with pytest.raises(AvitoBlockedError, match="403"):
await fetch_detail("/ekaterinburg/kvartiry/test-123", cffi_session=mock_session)
@pytest.mark.asyncio
async def test_fetch_detail_raises_rate_limited_on_429() -> None:
"""fetch_detail должен raise AvitoRateLimitedError на HTTP 429."""
from app.services.scrapers.avito_detail import fetch_detail
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 429
mock_session.get = AsyncMock(return_value=mock_response)
with pytest.raises(AvitoRateLimitedError, match="429"):
await fetch_detail("/ekaterinburg/kvartiry/test-123", cffi_session=mock_session)
@pytest.mark.asyncio
async def test_fetch_detail_raises_value_error_on_other_non_200() -> None:
"""fetch_detail должен raise ValueError на прочие non-200 (не 403/429)."""
from app.services.scrapers.avito_detail import fetch_detail
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 503
mock_session.get = AsyncMock(return_value=mock_response)
with pytest.raises(ValueError, match="503"):
await fetch_detail("/ekaterinburg/kvartiry/test-123", cffi_session=mock_session)
# ── SERP (avito.py) tests ───────────────────────────────────────
@pytest.mark.asyncio
async def test_fetch_around_raises_content_blocked_on_page1_empty_html() -> None:
"""fetch_around должен raise AvitoContentBlockedError на HTTP 200 + 0 карточек на странице 1."""
from app.services.scrapers.avito import AvitoScraper
scraper = AvitoScraper()
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 200
# Garbage HTML: нет data-marker="item" → _parse_html вернёт []
mock_response.text = "<html><body><p>captcha challenge</p></body></html>"
mock_session.get = AsyncMock(return_value=mock_response)
scraper._cffi = mock_session
with pytest.raises(AvitoContentBlockedError, match="content-block"):
await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
@pytest.mark.asyncio
async def test_fetch_around_content_blocked_is_subtype_of_blocked() -> None:
"""AvitoContentBlockedError должен быть поймать как AvitoBlockedError (наследование)."""
from app.services.scrapers.avito import AvitoScraper
scraper = AvitoScraper()
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 200
mock_response.text = "<html><body></body></html>"
mock_session.get = AsyncMock(return_value=mock_response)
scraper._cffi = mock_session
with pytest.raises(AvitoBlockedError):
await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
@pytest.mark.asyncio
async def test_fetch_around_page_gt1_empty_returns_gracefully() -> None:
"""fetch_around на page>1 с 0 карточек должен тихо завершить пагинацию (не raise)."""
from app.services.scrapers.avito import AvitoScraper
scraper = AvitoScraper()
mock_session = AsyncMock()
# Страница 1: нормальный HTML с одной карточкой
valid_html = (
"<html><body>"
'<div data-marker="item" data-item-id="123456789">'
'<a data-marker="item-title" href="/ekaterinburg/kvartiry/test-123456789">'
"2-к. квартира, 54,3 м², 5/9 эт."
"</a>"
'<meta itemprop="price" content="4500000"/>'
"</div>"
"</body></html>"
)
# Страница 2: пустой HTML → end of pagination
empty_html = "<html><body><p>no listings</p></body></html>"
page1_resp = MagicMock()
page1_resp.status_code = 200
page1_resp.text = valid_html
page2_resp = MagicMock()
page2_resp.status_code = 200
page2_resp.text = empty_html
mock_session.get = AsyncMock(side_effect=[page1_resp, page2_resp])
scraper._cffi = mock_session
with patch("app.services.scrapers.avito.AvitoScraper.sleep_between_requests", AsyncMock()):
result = await scraper.fetch_around(56.84, 60.60, 1000, pages=2)
# Должен вернуть лоты со страницы 1, не raise на странице 2
assert len(result) >= 1
@pytest.mark.asyncio
async def test_fetch_around_raises_blocked_on_403() -> None:
"""AvitoScraper.fetch_around должен raise AvitoBlockedError на HTTP 403."""
from app.services.scrapers.avito import AvitoScraper
scraper = AvitoScraper()
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 403
mock_session.get = AsyncMock(return_value=mock_response)
scraper._cffi = mock_session
with pytest.raises(AvitoBlockedError):
await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
@pytest.mark.asyncio
async def test_fetch_around_raises_rate_limited_on_429() -> None:
"""AvitoScraper.fetch_around должен raise AvitoRateLimitedError на HTTP 429."""
from app.services.scrapers.avito import AvitoScraper
scraper = AvitoScraper()
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 429
mock_session.get = AsyncMock(return_value=mock_response)
scraper._cffi = mock_session
with pytest.raises(AvitoRateLimitedError):
await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
# ── Pipeline abort tests ────────────────────────────────────────
@pytest.mark.asyncio
async def test_pipeline_aborts_after_3_consecutive_blocks() -> None:
"""Pipeline должен прервать sweep после 3 consecutive 403 (re-raise)."""
from app.services.scrape_pipeline import run_avito_pipeline
mock_db = MagicMock()
mock_scraper = MagicMock()
mock_scraper.__aenter__ = AsyncMock(return_value=mock_scraper)
mock_scraper.__aexit__ = AsyncMock(return_value=None)
mock_scraper.fetch_around = AsyncMock(return_value=[])
mock_scraper._cffi = None
mock_row = MagicMock()
mock_row.__getitem__ = lambda self, key: (
"https://www.avito.ru/ekaterinburg/kvartiry/test-1" if key == "source_url" else None
)
mock_mappings = MagicMock()
mock_mappings.all.return_value = [mock_row, mock_row, mock_row, mock_row]
mock_execute = MagicMock()
mock_execute.mappings.return_value = mock_mappings
mock_db.execute.return_value = mock_execute
block_error = AvitoBlockedError("HTTP 403")
with (
patch("app.services.scrape_pipeline.AvitoScraper", return_value=mock_scraper),
patch(
"app.services.scrape_pipeline.fetch_detail",
AsyncMock(side_effect=block_error),
),
patch("app.services.scrape_pipeline.asyncio.sleep", AsyncMock()),
patch(
"curl_cffi.requests.AsyncSession",
return_value=AsyncMock(__aenter__=AsyncMock(), __aexit__=AsyncMock()),
),
):
with pytest.raises(AvitoBlockedError):
await run_avito_pipeline(
mock_db,
56.8,
60.6,
1000,
enrich_houses=False,
enrich_detail_top_n=4,
)
@pytest.mark.asyncio
async def test_pipeline_sleep_between_detail_requests() -> None:
"""Pipeline должен вызывать asyncio.sleep между каждым detail request."""
from app.services.scrape_pipeline import run_avito_pipeline
mock_db = MagicMock()
mock_scraper = MagicMock()
mock_scraper.__aenter__ = AsyncMock(return_value=mock_scraper)
mock_scraper.__aexit__ = AsyncMock(return_value=None)
mock_scraper.fetch_around = AsyncMock(return_value=[])
mock_scraper._cffi = None
mock_row = MagicMock()
mock_row.__getitem__ = lambda self, key: (
"https://www.avito.ru/ekaterinburg/kvartiry/test-1" if key == "source_url" else None
)
mock_mappings = MagicMock()
mock_mappings.all.return_value = [mock_row, mock_row, mock_row]
mock_execute = MagicMock()
mock_execute.mappings.return_value = mock_mappings
mock_db.execute.return_value = mock_execute
mock_enrichment = MagicMock()
sleep_mock = AsyncMock()
with (
patch("app.services.scrape_pipeline.AvitoScraper", return_value=mock_scraper),
patch(
"app.services.scrape_pipeline.fetch_detail",
AsyncMock(return_value=mock_enrichment),
),
patch("app.services.scrape_pipeline.save_detail_enrichment", return_value=True),
patch("app.services.scrape_pipeline.asyncio.sleep", sleep_mock),
patch(
"curl_cffi.requests.AsyncSession",
return_value=AsyncMock(__aenter__=AsyncMock(), __aexit__=AsyncMock()),
),
patch("app.services.scrape_pipeline.random.uniform", return_value=1.0),
):
await run_avito_pipeline(
mock_db,
56.8,
60.6,
1000,
enrich_houses=False,
enrich_detail_top_n=3,
request_delay_sec=7.0,
)
assert sleep_mock.call_count == 2
for c in sleep_mock.call_args_list:
assert c == call(7.0)
@pytest.mark.asyncio
async def test_city_sweep_marks_banned_on_block() -> None:
"""City sweep mark'ит run как banned при AvitoBlockedError.
Migration 015 задокументировал 'banned' как 'Avito вернул 403/captcha' — именно этот сценарий.
"""
from app.services.scrape_pipeline import run_avito_city_sweep
mock_db = MagicMock()
mock_runs = MagicMock()
mock_runs.is_cancelled.return_value = False
mock_runs.update_heartbeat = MagicMock()
mock_runs.mark_banned = MagicMock()
mock_runs.mark_done = MagicMock()
block_error = AvitoBlockedError("IP blocked at page=1")
with (
patch("app.services.scrape_pipeline.scrape_runs", mock_runs),
patch(
"app.services.scrape_pipeline.run_avito_pipeline",
AsyncMock(side_effect=block_error),
),
patch(
"curl_cffi.requests.AsyncSession",
return_value=AsyncMock(
__aenter__=AsyncMock(return_value=AsyncMock()),
__aexit__=AsyncMock(return_value=None),
),
),
):
result = await run_avito_city_sweep(
mock_db,
run_id=42,
anchors=[(56.84, 60.60, "Центр"), (56.79, 60.53, "ЮЗ")],
)
assert mock_runs.mark_banned.called
assert not mock_runs.mark_done.called
assert result.anchors_done == 1
# ── #623 firewall-детект + ротация IP ────────────────────────────
_CARD_HTML = (
"<html><body>"
'<div data-marker="item" data-item-id="123456789">'
'<a data-marker="item-title" href="/ekaterinburg/kvartiry/test-123456789">'
"2-к. квартира, 54,3 м², 5/9 эт."
"</a>"
'<meta itemprop="price" content="4500000"/>'
"</div>"
"</body></html>"
)
# Firewall-интерстициал Avito (HTTP 200, не 403) — IP-блок datacenter.
_FIREWALL_HTML = (
"<html><head><title>Доступ ограничен: проблема с IP</title></head>"
'<body><div class="firewall-container">...</div></body></html>'
)
@pytest.mark.asyncio
async def test_fetch_around_raises_blocked_on_firewall_html_200() -> None:
"""HTTP 200 + firewall-страница «Доступ ограничен» → AvitoBlockedError (#623).
Avito банит datacenter-IP отдачей firewall-интерстициала с кодом 200 (не 403) —
раньше это молча проходило как «0 cards / content-block», теперь явный IP-блок.
"""
from app.services.scrapers.avito import AvitoScraper
scraper = AvitoScraper()
mock_session = AsyncMock()
mock_response = MagicMock()
mock_response.status_code = 200
mock_response.text = _FIREWALL_HTML
mock_session.get = AsyncMock(return_value=mock_response)
scraper._cffi = mock_session
with pytest.raises(AvitoBlockedError, match="firewall"):
await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
@pytest.mark.asyncio
async def test_fetch_around_rotates_ip_then_succeeds(monkeypatch: pytest.MonkeyPatch) -> None:
"""При firewall + заданной changeip-ссылке: ротация IP и повтор страницы (#623)."""
from app.services.scrapers import avito as avito_mod
from app.services.scrapers.avito import AvitoScraper
monkeypatch.setattr(
avito_mod.settings, "avito_proxy_rotate_url", "https://changeip.example/?proxy_key=x"
)
monkeypatch.setattr(avito_mod.settings, "avito_proxy_max_rotations", 1)
scraper = AvitoScraper()
fw = MagicMock()
fw.status_code = 200
fw.text = _FIREWALL_HTML
ok = MagicMock()
ok.status_code = 200
ok.text = _CARD_HTML
mock_session = AsyncMock()
mock_session.get = AsyncMock(side_effect=[fw, ok])
scraper._cffi = mock_session
with (
patch.object(AvitoScraper, "_rotate_ip", AsyncMock(return_value=True)) as rot,
patch.object(AvitoScraper, "sleep_between_requests", AsyncMock()),
):
result = await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
rot.assert_awaited_once()
assert len(result) == 1
assert mock_session.get.await_count == 2 # firewall → rotate → retry OK
@pytest.mark.asyncio
async def test_fetch_around_blocked_when_rotation_exhausted(
monkeypatch: pytest.MonkeyPatch,
) -> None:
"""Если ротация не сняла бан (max_rotations исчерпан) → AvitoBlockedError (#623)."""
from app.services.scrapers import avito as avito_mod
from app.services.scrapers.avito import AvitoScraper
monkeypatch.setattr(
avito_mod.settings, "avito_proxy_rotate_url", "https://changeip.example/?proxy_key=x"
)
monkeypatch.setattr(avito_mod.settings, "avito_proxy_max_rotations", 1)
scraper = AvitoScraper()
fw = MagicMock()
fw.status_code = 200
fw.text = _FIREWALL_HTML
mock_session = AsyncMock()
mock_session.get = AsyncMock(return_value=fw) # всегда firewall
scraper._cffi = mock_session
with patch.object(AvitoScraper, "_rotate_ip", AsyncMock(return_value=True)) as rot:
with pytest.raises(AvitoBlockedError, match="firewall"):
await scraper.fetch_around(56.84, 60.60, 1000, pages=1)
assert rot.await_count == 1 # одна попытка ротации, потом сдаёмся
# ── #623 sweep-pipeline proxy (SERP+detail через mobile proxy) ───
def test_avito_proxies_helper(monkeypatch: pytest.MonkeyPatch) -> None:
"""_avito_proxies: None без env, dict {http,https} при заданном AVITO_PROXY_URL.
Регрессия: sweep создаёт shared-сессию, через неё идут и SERP, и detail —
без прокси detail летит с datacenter-IP → 429 → run banned (#623 followup).
"""
from app.services import scrape_pipeline as sp
monkeypatch.setattr(sp.settings, "avito_proxy_url", None)
assert sp._avito_proxies() is None
url = "http://u:p@mproxy.site:16928"
monkeypatch.setattr(sp.settings, "avito_proxy_url", url)
assert sp._avito_proxies() == {"http": url, "https": url}