Add checkpoint logic and tests

This commit is contained in:
qananasikq
2026-04-15 11:10:32 +03:00
parent d9111be2d2
commit acbb045b6e
8 changed files with 432 additions and 175 deletions

View File

@@ -1,6 +1,7 @@
from __future__ import annotations
import unittest
from types import SimpleNamespace
from unittest.mock import MagicMock
from iaai_scraper.core.config import Settings
@@ -169,6 +170,50 @@ class TestScraperSync(unittest.TestCase):
self.assertTrue(IAAIScraper._is_protection_or_network_error(RuntimeError("captcha challenge")))
self.assertFalse(IAAIScraper._is_protection_or_network_error(RuntimeError("plain validation error")))
def test_recover_empty_listing_page_reload_recovers_links(self) -> None:
scraper = self._make_scraper()
page = MagicMock()
page_result = SimpleNamespace(
vehicle_links=[SimpleNamespace(href="https://www.iaai.com/VehicleDetail/123~US")],
)
scraper.listing_collector.collect_current_page = MagicMock(return_value=page_result)
all_raw_urls: list[str] = []
seen_urls: set[str] = set()
recovered_result, recovered_urls = scraper._recover_empty_listing_page(
page,
page_number=5,
all_raw_urls=all_raw_urls,
seen_urls=seen_urls,
)
page.reload.assert_called_once()
self.assertIs(recovered_result, page_result)
self.assertEqual(recovered_urls, ["https://www.iaai.com/VehicleDetail/123~US"])
self.assertEqual(all_raw_urls, ["https://www.iaai.com/VehicleDetail/123~US"])
def test_recover_empty_listing_page_reopens_listing_for_first_page(self) -> None:
scraper = self._make_scraper()
page = MagicMock()
page_result = SimpleNamespace(
vehicle_links=[SimpleNamespace(href="https://www.iaai.com/VehicleDetail/456~US")],
)
scraper.listing_collector.open_cars_listing = MagicMock()
scraper.listing_collector.collect_current_page = MagicMock(return_value=page_result)
recovered_result, recovered_urls = scraper._recover_empty_listing_page(
page,
page_number=1,
all_raw_urls=[],
seen_urls=set(),
)
scraper.listing_collector.open_cars_listing.assert_called_once_with(page)
page.reload.assert_not_called()
self.assertIs(recovered_result, page_result)
self.assertEqual(recovered_urls, ["https://www.iaai.com/VehicleDetail/456~US"])
if __name__ == "__main__":
unittest.main()