add openlane scraper

This commit is contained in:
qananasikq
2026-04-21 23:49:56 +03:00
parent 7f72056289
commit 74f12da10a
32 changed files with 4713 additions and 0 deletions

View File

@@ -0,0 +1,379 @@
"""Runtime config: загрузка и применение runtime_config.json.
Позволяет менять параметры sync и фильтры машин без перезапуска сервиса.
Файл перечитывается при каждом запуске sync_listing.
"""
from __future__ import annotations
import json
import logging
from dataclasses import dataclass, field
from pathlib import Path
from typing import Any
logger = logging.getLogger("openlane_scraper.core.runtime_config")
@dataclass(slots=True)
class SyncConfig:
name: str | None = None
ids_initial_size: int | None = None
ids_next_size: int | None = None
ids_max_pages: int | None = None
condition_check_enabled: bool | None = None
lane: str = "openlane_marketplace"
only_new: bool = False
limit: int | None = None
@dataclass(slots=True)
class ListingConfig:
make: str | None = None
model: str | None = None
@dataclass(slots=True)
class RangeFilter:
min: int | None = None
max: int | None = None
@dataclass(slots=True)
class FlagFilters:
damaged_only: bool | None = None
run_and_drive: bool | None = None
@dataclass(slots=True)
class FieldFilters:
brands: list[str] = field(default_factory=list)
models: list[str] = field(default_factory=list)
years: list[int] = field(default_factory=list)
body_types: list[str] = field(default_factory=list)
colors: list[str] = field(default_factory=list)
drives: list[str] = field(default_factory=list)
gearboxes: list[str] = field(default_factory=list)
locations: list[str] = field(default_factory=list)
@dataclass(slots=True)
class Filters:
price: RangeFilter = field(default_factory=RangeFilter)
mileage: RangeFilter = field(default_factory=RangeFilter)
flags: FlagFilters = field(default_factory=FlagFilters)
include: FieldFilters = field(default_factory=FieldFilters)
exclude: FieldFilters = field(default_factory=FieldFilters)
# Legacy flat fields (backward compatibility).
brands: list[str] = field(default_factory=list)
models: list[str] = field(default_factory=list)
years: list[int] = field(default_factory=list)
body_types: list[str] = field(default_factory=list)
colors: list[str] = field(default_factory=list)
drives: list[str] = field(default_factory=list)
gearboxes: list[str] = field(default_factory=list)
locations: list[str] = field(default_factory=list)
exclude_brands: list[str] = field(default_factory=list)
exclude_models: list[str] = field(default_factory=list)
exclude_years: list[int] = field(default_factory=list)
exclude_body_types: list[str] = field(default_factory=list)
exclude_colors: list[str] = field(default_factory=list)
exclude_drives: list[str] = field(default_factory=list)
exclude_gearboxes: list[str] = field(default_factory=list)
exclude_locations: list[str] = field(default_factory=list)
@dataclass(slots=True)
class RuntimeConfig:
sync: SyncConfig = field(default_factory=SyncConfig)
listing: ListingConfig = field(default_factory=ListingConfig)
filters: Filters = field(default_factory=Filters)
def load_runtime_config(path: str | Path | None) -> RuntimeConfig:
"""Загрузить runtime_config.json. Если файл отсутствует — вернуть дефолты."""
if not path:
return RuntimeConfig()
p = Path(path)
if not p.exists():
logger.debug("runtime_config not found at %s — using defaults", p)
return RuntimeConfig()
try:
raw = json.loads(p.read_text(encoding="utf-8"))
except (json.JSONDecodeError, OSError) as exc:
logger.warning("Failed to parse runtime_config %s: %s — using defaults", p, exc)
return RuntimeConfig()
cfg = RuntimeConfig()
sync_raw = raw.get("sync") or {}
if isinstance(sync_raw, dict):
cfg.sync.name = _opt_str(sync_raw.get("name"))
cfg.sync.ids_initial_size = _opt_int(sync_raw.get("ids_initial_size"))
cfg.sync.ids_next_size = _opt_int(sync_raw.get("ids_next_size"))
cfg.sync.ids_max_pages = _opt_int(sync_raw.get("ids_max_pages"))
cfg.sync.condition_check_enabled = _opt_bool(sync_raw.get("condition_check_enabled"))
cfg.sync.lane = _opt_str(sync_raw.get("lane")) or cfg.sync.lane
only_new = _opt_bool(sync_raw.get("only_new"))
if only_new is not None:
cfg.sync.only_new = only_new
cfg.sync.limit = _opt_int(sync_raw.get("limit"))
listing_raw = raw.get("listing") or {}
if isinstance(listing_raw, dict):
cfg.listing.make = _opt_str(listing_raw.get("make"))
cfg.listing.model = _opt_str(listing_raw.get("model"))
filters_raw = raw.get("filters") or {}
if isinstance(filters_raw, dict):
cfg.filters.price = _parse_range(filters_raw.get("price"))
cfg.filters.mileage = _parse_range(filters_raw.get("mileage"))
cfg.filters.flags = _parse_flags(filters_raw.get("flags"))
include_payload = filters_raw.get("include") if isinstance(filters_raw.get("include"), dict) else filters_raw
exclude_payload = (
filters_raw.get("exclude")
if isinstance(filters_raw.get("exclude"), dict)
else {
"brands": filters_raw.get("exclude_brands"),
"models": filters_raw.get("exclude_models"),
"years": filters_raw.get("exclude_years"),
"body_types": filters_raw.get("exclude_body_types"),
"colors": filters_raw.get("exclude_colors"),
"drives": filters_raw.get("exclude_drives"),
"gearboxes": filters_raw.get("exclude_gearboxes"),
"locations": filters_raw.get("exclude_locations"),
}
)
cfg.filters.include = _parse_fields(include_payload)
cfg.filters.exclude = _parse_fields(exclude_payload)
# Backward-compatible flat aliases.
cfg.filters.brands = list(cfg.filters.include.brands)
cfg.filters.models = list(cfg.filters.include.models)
cfg.filters.years = list(cfg.filters.include.years)
cfg.filters.body_types = list(cfg.filters.include.body_types)
cfg.filters.colors = list(cfg.filters.include.colors)
cfg.filters.drives = list(cfg.filters.include.drives)
cfg.filters.gearboxes = list(cfg.filters.include.gearboxes)
cfg.filters.locations = list(cfg.filters.include.locations)
cfg.filters.exclude_brands = list(cfg.filters.exclude.brands)
cfg.filters.exclude_models = list(cfg.filters.exclude.models)
cfg.filters.exclude_years = list(cfg.filters.exclude.years)
cfg.filters.exclude_body_types = list(cfg.filters.exclude.body_types)
cfg.filters.exclude_colors = list(cfg.filters.exclude.colors)
cfg.filters.exclude_drives = list(cfg.filters.exclude.drives)
cfg.filters.exclude_gearboxes = list(cfg.filters.exclude.gearboxes)
cfg.filters.exclude_locations = list(cfg.filters.exclude.locations)
logger.info(
"Loaded runtime_config from %s: lane=%s, limit=%s",
p,
cfg.sync.lane,
cfg.sync.limit,
)
return cfg
def apply_filters(records: list[dict], filters: Filters) -> list[dict]:
"""Применить runtime-фильтры к списку сырых записей из API."""
if not records:
return records
include = getattr(filters, "include", None)
exclude = getattr(filters, "exclude", None)
include_brands = tuple(getattr(filters, "brands", ()) or getattr(include, "brands", ()))
include_models = tuple(getattr(filters, "models", ()) or getattr(include, "models", ()))
include_years = tuple(getattr(filters, "years", ()) or getattr(include, "years", ()))
include_body_types = tuple(getattr(filters, "body_types", ()) or getattr(include, "body_types", ()))
exclude_brands = tuple(getattr(filters, "exclude_brands", ()) or getattr(exclude, "brands", ()))
exclude_models = tuple(getattr(filters, "exclude_models", ()) or getattr(exclude, "models", ()))
exclude_years = tuple(getattr(filters, "exclude_years", ()) or getattr(exclude, "years", ()))
exclude_body_types = tuple(getattr(filters, "exclude_body_types", ()) or getattr(exclude, "body_types", ()))
price_cfg = getattr(filters, "price", None)
mileage_cfg = getattr(filters, "mileage", None)
flags_cfg = getattr(filters, "flags", None)
price_min = getattr(price_cfg, "min", None)
price_max = getattr(price_cfg, "max", None)
mileage_min = getattr(mileage_cfg, "min", None)
mileage_max = getattr(mileage_cfg, "max", None)
damaged_only = getattr(flags_cfg, "damaged_only", None)
run_and_drive = getattr(flags_cfg, "run_and_drive", None)
result = records
if include_brands:
include_set = {s.casefold() for s in include_brands}
result = [r for r in result if _get_brand(r).casefold() in include_set]
if exclude_brands:
exclude_set = {s.casefold() for s in exclude_brands}
result = [r for r in result if _get_brand(r).casefold() not in exclude_set]
if include_models:
include_set = {s.casefold() for s in include_models}
result = [r for r in result if _get_model(r).casefold() in include_set]
if exclude_models:
exclude_set = {s.casefold() for s in exclude_models}
result = [r for r in result if _get_model(r).casefold() not in exclude_set]
if include_years:
years_set = set(include_years)
result = [r for r in result if _get_year(r) in years_set]
if exclude_years:
years_set = set(exclude_years)
result = [r for r in result if _get_year(r) not in years_set]
if include_body_types:
include_set = {s.casefold() for s in include_body_types}
result = [r for r in result if _get_body_type(r).casefold() in include_set]
if exclude_body_types:
exclude_set = {s.casefold() for s in exclude_body_types}
result = [r for r in result if _get_body_type(r).casefold() not in exclude_set]
if price_min is not None or price_max is not None:
filtered: list[dict] = []
for r in result:
price = _get_price(r)
if price is None:
filtered.append(r)
continue
if price_min is not None and price < price_min:
continue
if price_max is not None and price > price_max:
continue
filtered.append(r)
result = filtered
if mileage_min is not None or mileage_max is not None:
filtered = []
for r in result:
miles = _get_mileage(r)
if miles is None:
filtered.append(r)
continue
if mileage_min is not None and miles < mileage_min:
continue
if mileage_max is not None and miles > mileage_max:
continue
filtered.append(r)
result = filtered
if damaged_only is not None:
result = [r for r in result if bool(r.get("is_damaged")) is damaged_only]
if run_and_drive is not None:
result = [r for r in result if bool(r.get("run_and_drive")) is run_and_drive]
return result
def _parse_range(payload: Any) -> RangeFilter:
payload = payload if isinstance(payload, dict) else {}
return RangeFilter(min=_opt_int(payload.get("min")), max=_opt_int(payload.get("max")))
def _parse_flags(payload: Any) -> FlagFilters:
payload = payload if isinstance(payload, dict) else {}
return FlagFilters(
damaged_only=_opt_bool(payload.get("damaged_only")),
run_and_drive=_opt_bool(payload.get("run_and_drive")),
)
def _parse_fields(payload: Any) -> FieldFilters:
payload = payload if isinstance(payload, dict) else {}
return FieldFilters(
brands=_str_list(payload.get("brands")),
models=_str_list(payload.get("models")),
years=_int_list(payload.get("years")),
body_types=_str_list(payload.get("body_types")),
colors=_str_list(payload.get("colors")),
drives=_str_list(payload.get("drives")),
gearboxes=_str_list(payload.get("gearboxes")),
locations=_str_list(payload.get("locations")),
)
def _opt_int(value: Any) -> int | None:
if value is None:
return None
try:
return int(value)
except (ValueError, TypeError):
return None
def _opt_bool(value: Any) -> bool | None:
if value is None:
return None
if isinstance(value, bool):
return value
if isinstance(value, str):
normalized = value.strip().casefold()
if normalized in {"1", "true", "yes", "on"}:
return True
if normalized in {"0", "false", "no", "off"}:
return False
return None
def _opt_str(value: Any) -> str | None:
if value is None:
return None
s = str(value).strip()
return s or None
def _str_list(value: Any) -> list[str]:
if not isinstance(value, list):
return []
return [str(v).strip() for v in value if str(v).strip()]
def _int_list(value: Any) -> list[int]:
if not isinstance(value, list):
return []
result: list[int] = []
for v in value:
iv = _opt_int(v)
if iv is not None:
result.append(iv)
return result
def _get_brand(record: dict[str, Any]) -> str:
return str(record.get("make") or record.get("brand") or "")
def _get_model(record: dict[str, Any]) -> str:
return str(record.get("model") or "")
def _get_body_type(record: dict[str, Any]) -> str:
return str(record.get("body_type") or record.get("body_style") or record.get("vehicle_type") or "")
def _get_year(record: dict[str, Any]) -> int | None:
return _opt_int(record.get("year"))
def _get_price(record: dict[str, Any]) -> int | None:
for key in ("current_high_bid", "buy_now_price", "price", "current_bid", "sale_price"):
value = _opt_int(record.get(key))
if value is not None:
return value
return None
def _get_mileage(record: dict[str, Any]) -> int | None:
return _opt_int(record.get("odometer") or record.get("mileage"))