update sync

This commit is contained in:
qananasikq
2026-04-24 18:00:09 +03:00
parent 8c441b2aec
commit 1453eee83b
7 changed files with 221 additions and 46 deletions

View File

@@ -105,6 +105,8 @@ class HumanPaceConfig:
@dataclass(slots=True)
class ListingConfig:
cars_url: str = _env_str("IAAI_CARS_LISTING_URL", "https://www.iaai.com/Vehiclelisting/Cars")
filtered_search_url: str | None = _env_optional_str("IAAI_FILTERED_SEARCH_URL")
filtered_search_urls_raw: str | None = _env_optional_str("IAAI_FILTERED_SEARCH_URLS")
max_pages_per_run: int = _env_int("IAAI_MAX_PAGES_PER_RUN", 9999)
max_vehicles_per_run: int = _env_int("IAAI_MAX_VEHICLES_PER_RUN", 50000)
page_link_limit: int = _env_int("IAAI_PAGE_LINK_LIMIT", 500)
@@ -120,6 +122,26 @@ class ListingConfig:
listing_segments_json: str = _env_str("IAAI_LISTING_SEGMENTS", "")
fast_segment_year_splits: bool = _env_bool("IAAI_FAST_SEGMENT_YEAR_SPLITS", True)
@property
def filtered_search_urls(self) -> list[str]:
urls: list[str] = []
if self.filtered_search_url and self.filtered_search_url.strip():
urls.append(self.filtered_search_url.strip())
if self.filtered_search_urls_raw and self.filtered_search_urls_raw.strip():
raw = self.filtered_search_urls_raw.strip()
try:
parsed = json.loads(raw)
except (json.JSONDecodeError, ValueError):
parsed = None
if isinstance(parsed, list):
candidates = [str(item or "").strip() for item in parsed]
else:
candidates = [part.strip() for part in raw.replace("\n", ",").split(",")]
for candidate in candidates:
if candidate and candidate not in urls:
urls.append(candidate)
return urls
# Список брендов IAAI для автоматической сегментации.
# Покрывает >99% автомобилей на сайте. Порядок: от крупных к мелким.
@@ -280,6 +302,8 @@ class ScrapingProfileConfig:
http_first: bool = _env_bool("IAAI_HTTP_FIRST", True)
browser_fallback_enabled: bool = _env_bool("IAAI_BROWSER_FALLBACK_ENABLED", True)
anonymous_bootstrap_enabled: bool = _env_bool("IAAI_ANONYMOUS_BOOTSTRAP_ENABLED", True)
verbose_http_logs: bool = _env_bool("IAAI_VERBOSE_HTTP_LOGS", False)
verbose_progress_logs: bool = _env_bool("IAAI_VERBOSE_PROGRESS_LOGS", False)
challenge_refresh_enabled: bool = _env_bool("IAAI_CHALLENGE_REFRESH_ENABLED", True)
challenge_refresh_attempts: int = _env_int("IAAI_CHALLENGE_REFRESH_ATTEMPTS", 3)
listing_post_attempts: int = _env_int("IAAI_LISTING_POST_ATTEMPTS", 4)