refactor mobile.de parser, fix country mapping, update README

This commit is contained in:
qananasikq
2026-05-04 18:12:10 +03:00
parent 800867b4e6
commit 6dc08c15f0
3 changed files with 2 additions and 117 deletions

View File

@@ -1,14 +0,0 @@
class ScraperError(Exception):
"""Базовое исключение скрапера."""
class AntiBotDetectedError(ScraperError):
"""Вызывается, когда сайт блокирует автоматизацию."""
class SiteStructureChangedError(ScraperError):
"""Вызывается, когда структура страницы изменилась и данных не хватает."""
class ListingResumeError(ScraperError):
"""Вызывается, когда resume по checkpoint больше недостижим."""

View File

@@ -1,53 +0,0 @@
import logging
import random
import time
from collections.abc import Callable
from functools import wraps
from typing import Any
from playwright.sync_api import Error, TimeoutError as PlaywrightTimeoutError
from .exceptions import AntiBotDetectedError
logger = logging.getLogger("MOBILEDE_scraper.retry")
# Типы исключений, при которых retry имеет смысл.
RETRYABLE_EXCEPTIONS = (
PlaywrightTimeoutError,
Error,
ConnectionError,
OSError,
TimeoutError,
AntiBotDetectedError,
)
def retryable(
max_attempts: int,
delay_seconds: float = 2.5,
backoff_multiplier: float = 2.0,
jitter_seconds: float = 0.0,
) -> Callable[[Callable[..., Any]], Callable[..., Any]]:
def decorator(func: Callable[..., Any]) -> Callable[..., Any]:
@wraps(func)
def wrapper(*args: Any, **kwargs: Any) -> Any:
last_error: Exception | None = None
for attempt in range(1, max_attempts + 1):
try:
return func(*args, **kwargs)
except RETRYABLE_EXCEPTIONS as exc:
last_error = exc
logger.warning("%s failed on attempt %s/%s: %s", func.__name__, attempt, max_attempts, exc)
if attempt < max_attempts:
sleep_for = delay_seconds * (backoff_multiplier ** (attempt - 1))
if jitter_seconds > 0:
sleep_for += random.uniform(0, jitter_seconds)
logger.debug("Retrying %s in %.2fs", func.__name__, sleep_for)
time.sleep(sleep_for)
if last_error is not None:
raise last_error
raise RuntimeError("Retry wrapper failed without a captured exception")
return wrapper
return decorator

View File

@@ -1,7 +1,6 @@
import json import json
import re
from pathlib import Path from pathlib import Path
from typing import Any, Callable, Iterable from typing import Any
def save_to_json(data: Any, filename: str | Path) -> None: def save_to_json(data: Any, filename: str | Path) -> None:
@@ -10,21 +9,7 @@ def save_to_json(data: Any, filename: str | Path) -> None:
path.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8") path.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
def first_non_empty(values: Iterable[Any]) -> Any | None:
for value in values:
if value not in (None, "", [], {}, ()):
return value
return None
# Регулярные выражения для VIN, lot и price.
VIN_RE = re.compile(r"\b([A-HJ-NPR-Z0-9]{17})\b", re.IGNORECASE)
LOT_RE = re.compile(r"\b(\d{7,10})\b")
PRICE_RE = re.compile(r"\$\s?([\d,]+(?:\.\d{1,2})?)")
def deep_find_key(obj, target_keys: set[str], max_depth: int = 64, _depth: int = 0) -> list: def deep_find_key(obj, target_keys: set[str], max_depth: int = 64, _depth: int = 0) -> list:
# Рекурсивно ищет значения по набору ключей в произвольном JSON-дереве.
found = [] found = []
if _depth >= max_depth: if _depth >= max_depth:
return found return found
@@ -37,36 +22,3 @@ def deep_find_key(obj, target_keys: set[str], max_depth: int = 64, _depth: int =
for item in obj: for item in obj:
found.extend(deep_find_key(item, target_keys, max_depth=max_depth, _depth=_depth + 1)) found.extend(deep_find_key(item, target_keys, max_depth=max_depth, _depth=_depth + 1))
return found return found
def deep_find_all_keys(
payloads: list,
field_map: dict[str, set[str]],
max_depth: int = 64,
) -> dict[str, list]:
"""Извлекает все нужные поля за один проход по JSON."""
# Готовим обратную карту: нормализованный ключ -> имя поля.
reverse: dict[str, str] = {}
for field_name, keys in field_map.items():
for k in keys:
reverse[k.lower()] = field_name
result: dict[str, list] = {f: [] for f in field_map}
def _recurse(obj: Any, depth: int) -> None:
if depth >= max_depth:
return
if isinstance(obj, dict):
for k, v in obj.items():
field = reverse.get(k.lower())
if field is not None:
result[field].append(v)
_recurse(v, depth + 1)
elif isinstance(obj, list):
for item in obj:
_recurse(item, depth + 1)
for payload in payloads:
_recurse(payload, 0)
return result