Compare commits

..

9 Commits

Author SHA1 Message Date
qananasikq
9c1756a959 Add vehicle color parsing 2026-08-03 21:05:45 +03:00
qananasikq
72e78de1d7 Update model translations 2026-08-03 21:04:19 +03:00
qananasikq
6bdbbe396c Remove temporary cars_100.txt 2026-07-01 14:44:34 +03:00
qananasikq
efbbf4a067 Update model translations 2026-07-01 14:42:35 +03:00
qananasikq
1c5c30e31e fix encar id mapping and docker build stability 2026-04-21 19:35:12 +03:00
qananasikq
eba3750f15 speed up cars count query 2026-04-17 23:53:54 +03:00
qananasikq
49de1d6f53 log create_tables errors 2026-04-17 23:53:53 +03:00
qananasikq
7f74793812 add shard page cap 2026-04-17 23:53:53 +03:00
qananasikq
53eb323287 fix batch pool leak 2026-04-17 23:53:10 +03:00
8 changed files with 888 additions and 42 deletions

BIN
.gitignore vendored

Binary file not shown.

View File

@@ -1,19 +1,22 @@
FROM python:3.12-slim
ENV PYTHONDONTWRITEBYTECODE=1 \
PYTHONUNBUFFERED=1
PYTHONUNBUFFERED=1 \
PIP_DISABLE_PIP_VERSION_CHECK=1 \
PIP_DEFAULT_TIMEOUT=300
RUN groupadd --system app && useradd --system --gid app --create-home app
WORKDIR /app
COPY pyproject.toml uv.lock ./
RUN pip install --no-cache-dir uv \
RUN pip install --no-cache-dir --retries 10 uv \
&& uv export --format requirements-txt --no-dev --no-hashes --no-emit-project --frozen -o /tmp/requirements.txt \
&& pip install --no-cache-dir -r /tmp/requirements.txt
&& pip install --no-cache-dir --retries 10 -r /tmp/requirements.txt
COPY . .
RUN pip install --no-cache-dir -e .
RUN pip install --no-cache-dir --retries 10 "setuptools>=68"
RUN pip install --no-cache-dir --no-build-isolation -e .
RUN python -m compileall -q encar_scraper
RUN chmod +x entrypoint.sh
RUN chown -R app:app /app

View File

@@ -26,6 +26,9 @@ from .translations import (
FUEL_TYPE_MAP,
GEARBOX_MAP_KO,
MODEL_BODY_TYPE_MAP,
MODEL_EXACT_TRANSLATIONS,
MODEL_PHRASE_TRANSLATIONS,
MODEL_TOKEN_TRANSLATIONS,
MODEL_TRANSLATIONS,
SELL_TYPE_MAP,
expand_allowed_brands,
@@ -42,6 +45,7 @@ BATCH_VEHICLES_CHUNK_SIZE = 20
ENCAR_DETAIL_URL_TEMPLATE = "https://www.encar.com/dc/dc_cardetailview.do?carid={vehicle_id}"
ENCAR_IMAGE_BASE = "https://ci.encar.com"
VEHICLE_ID_RE = re.compile(r"(?:carid|vehicleId)=?(\d+)")
NUMERIC_ID_RE = re.compile(r"(\d{6,})")
PHOTO_VEHICLE_ID_RE = re.compile(r"/(\d+)_\d+\.(?:jpg|jpeg|png|webp)$", re.IGNORECASE)
# Максимальный номер фото для пробинга и допустимые промахи подряд
@@ -155,30 +159,24 @@ class RuntimeFilterSpec:
@dataclass(slots=True)
class EncarMapper:
def map_to_car_record(self, vehicle_url: str, payload: dict[str, Any], probe_all_photos: bool = False) -> CarRecord:
vehicle_id = self._extract_vehicle_id(vehicle_url) or str(
payload.get("Id") or payload.get("vehicleId") or ""
)
vehicle_id = self._resolve_vehicle_id(vehicle_url, payload)
origin_id = f"encar:{vehicle_id}"
parser_id = f"encar:{vehicle_id}"
origin_url = vehicle_url or ENCAR_DETAIL_URL_TEMPLATE.format(vehicle_id=vehicle_id)
url_vehicle_id = self._extract_vehicle_id(vehicle_url)
if vehicle_url and url_vehicle_id == vehicle_id:
origin_url = vehicle_url
else:
origin_url = ENCAR_DETAIL_URL_TEMPLATE.format(vehicle_id=vehicle_id)
brand = self._as_str(payload.get("Manufacturer") or payload.get("Brand") or "UNKNOWN")
# Перевод бренда на русский
brand = BRAND_TRANSLATIONS.get(brand, brand)
# Model + Badge + BadgeDetail = полное название
# Нужна только базовая модель (без trim/sub-trim)
model_raw = self._as_str(payload.get("Model") or "UNKNOWN")
badge = self._as_str(payload.get("Badge") or "")
badge_detail = self._as_str(payload.get("BadgeDetail") or "")
# Полное описание модели: "Model / Badge BadgeDetail"
model_parts = [model_raw]
if badge:
model_parts.append(badge)
if badge_detail and badge_detail not in badge:
model_parts.append(badge_detail)
model = " / ".join(model_parts) if len(model_parts) > 1 else model_parts[0]
model = self._translate_text(model)
model = self._translate_model(model_raw)
year = self._to_year(payload.get("FormYear") or payload.get("Year"))
# Price в 만원 (10 000 KRW) → переводим в KRW
@@ -283,6 +281,66 @@ class EncarMapper:
match = VEHICLE_ID_RE.search(vehicle_url)
return match.group(1) if match else None
def _resolve_vehicle_id(self, vehicle_url: str | None, payload: dict[str, Any]) -> str:
# 1) Из URL (carid=123...)
from_url = self._extract_vehicle_id(vehicle_url)
if from_url:
return from_url
# 2) Из явных полей payload
for key in ("Id", "id", "vehicleId", "VehicleId", "carId", "carid"):
value = payload.get(key)
extracted = self._extract_numeric_id(value)
if extracted:
return extracted
# 3) Из фото-пути (например /.../41854374_001.jpg)
photos = payload.get("Photos") or []
if isinstance(photos, list):
for raw in photos:
if not isinstance(raw, dict):
continue
location = raw.get("location") or raw.get("Location") or ""
extracted = self._extract_numeric_id(location)
if extracted:
return extracted
# 4) Фолбэк: любой id-like ключ в payload
fallback = self._find_any_numeric_id(payload)
if fallback:
return fallback
raise ValueError("Cannot resolve numeric Encar vehicle id")
def _extract_numeric_id(self, value: Any) -> str | None:
if value is None:
return None
if isinstance(value, int):
return str(value)
text = self._as_str(value)
if not text:
return None
match = NUMERIC_ID_RE.search(text)
return match.group(1) if match else None
def _find_any_numeric_id(self, payload: dict[str, Any]) -> str | None:
stack: list[Any] = [payload]
while stack:
current = stack.pop()
if isinstance(current, dict):
for key, value in current.items():
if isinstance(key, str) and "id" in key.lower():
extracted = self._extract_numeric_id(value)
if extracted:
return extracted
if isinstance(value, (dict, list)):
stack.append(value)
elif isinstance(current, list):
for item in current:
if isinstance(item, (dict, list)):
stack.append(item)
return None
def _build_slug(self, brand: str, model: str, vehicle_id: str) -> str:
raw = re.sub(r"[^a-z0-9]+", "-", f"{brand}-{model}-{vehicle_id}".lower()).strip("-")
return raw or f"encar-{vehicle_id}"
@@ -297,12 +355,85 @@ class EncarMapper:
translated = re.sub(r"\s+", " ", translated).strip()
return translated
def _has_korean(self, text: str) -> bool:
return bool(re.search(r"[가-힣]", text))
def _normalize_model_text(self, text: str) -> str:
normalized = self._as_str(text)
if not normalized:
return normalized
normalized = re.sub(r"[\(\[(【][^)\])】]*[\)\])】]?", " ", normalized)
normalized = normalized.replace("·", " ").replace("/", " ")
normalized = re.sub(r"(?<=[A-Za-z])-클래스", " 클래스", normalized)
normalized = re.sub(r"(?<=[0-9])시리즈", "시리즈", normalized)
normalized = re.sub(r"\s+", " ", normalized).strip()
return normalized
def _clean_base_model_name(self, text: str) -> str:
model = self._as_str(text)
if not model:
return model
marketing_prefix = r"(?:The\s+All\s+New|All\s+New|The\s+New|Very\s+New|New)"
previous = None
while previous != model:
previous = model
model = re.sub(rf"^\s*{marketing_prefix}\s+", "", model, flags=re.IGNORECASE)
model = re.sub(r"\b\d+(?:st|nd|rd|th)\s+gen\b", " ", model, flags=re.IGNORECASE)
model = re.sub(r"\bgen\b", " ", model, flags=re.IGNORECASE)
model = re.sub(r"\s+", " ", model).strip(" -")
return model
def _translate_model(self, text: str) -> str:
original = self._normalize_model_text(text)
if not original or not self._has_korean(original):
return self._clean_base_model_name(original)
exact = MODEL_EXACT_TRANSLATIONS.get(original)
if exact:
return self._clean_base_model_name(exact)
translated = original
combined = {**MODEL_EXACT_TRANSLATIONS, **MODEL_PHRASE_TRANSLATIONS, **MODEL_TRANSLATIONS}
for source, target in sorted(combined.items(), key=lambda item: len(item[0]), reverse=True):
if source in translated:
pattern = rf"(?<![가-힣]){re.escape(source)}(?![가-힣])"
translated = re.sub(pattern, target, translated)
for source, target in sorted(MODEL_TOKEN_TRANSLATIONS.items(), key=lambda item: len(item[0]), reverse=True):
translated = re.sub(rf"(?<![A-Za-z0-9가-힣]){re.escape(source)}(?![A-Za-z0-9가-힣])", target, translated)
translated = re.sub(r"(?<=\d)시리즈", " Series", translated)
translated = re.sub(r"(?<=[A-Za-z])-?클래스", "-Class", translated)
translated = re.sub(r"\b([A-Z]{1,4}) Class\b", r"\1-Class", translated)
translated = re.sub(r"[가-힣]+", " ", translated)
translated = re.sub(r"\(\s*\)", " ", translated)
translated = re.sub(r"\s*-\s*(?=$|\))", "", translated)
translated = re.sub(r"\s+([),])", r"\1", translated)
translated = re.sub(r"\s+", " ", translated).strip(" -")
return self._clean_base_model_name(translated)
def _translate_color(self, color_raw: str) -> str:
value = self._as_str(color_raw)
value = re.sub(r"\s+", "", self._as_str(color_raw))
if not value:
return "other"
return COLOR_TRANSLATIONS.get(value, self._translate_text(value).lower() or "other")
def apply_batch_color(self, record: CarRecord, vehicle_data: dict[str, Any]) -> None:
spec = vehicle_data.get("spec")
if not isinstance(spec, dict):
return
color_raw = self._as_str(spec.get("colorName") or spec.get("customColor") or "")
if color_raw:
record.color = self._translate_color(color_raw)
def _to_float(self, value: Any) -> float | None:
if value is None:
return None
@@ -1378,25 +1509,33 @@ class EncarScraper:
for i in range(0, len(tasks), BATCH_VEHICLES_CHUNK_SIZE):
chunks.append(tasks[i:i + BATCH_VEHICLES_CHUNK_SIZE])
def _fetch_chunk(chunk: list[tuple[int, str]]) -> list[tuple[int, list[ImageRecord]]]:
"""Запрашивает фото для чанка машин одним batch-запросом."""
def _fetch_chunk(chunk: list[tuple[int, str]]) -> list[tuple[int, list[ImageRecord], dict[str, Any]]]:
ids_str = ",".join(vid for _, vid in chunk)
path = f"/v1/readside/vehicles?vehicleIds={ids_str}&include=PHOTOS"
for attempt in range(2):
path = f"/v1/readside/vehicles?vehicleIds={ids_str}&include=SPEC,PHOTOS"
for attempt in range(3):
try:
resp = pool.request("GET", path, timeout=15, retries=False)
if resp.status == 200:
if not resp.data:
return [(ri, []) for ri, _ in chunk]
return [(ri, [], {}) for ri, _ in chunk]
vehicles = json.loads(resp.data.decode("utf-8", errors="ignore"))
# Маппим по vehicleId, а не по индексу — API может пропускать удалённые ID
id_to_rec: dict[str, int] = {vid: ri for ri, vid in chunk}
result: list[tuple[int, list[ImageRecord]]] = []
for vehicle_data in vehicles:
result: list[tuple[int, list[ImageRecord], dict[str, Any]]] = []
matched_record_indexes: set[int] = set()
for vehicle_position, vehicle_data in enumerate(vehicles):
vid = str(vehicle_data.get("vehicleId") or vehicle_data.get("id") or "")
rec_idx = None
if vehicle_position < len(chunk):
positional_rec_idx = chunk[vehicle_position][0]
if positional_rec_idx not in matched_record_indexes:
rec_idx = positional_rec_idx
if rec_idx is None:
rec_idx = id_to_rec.get(vid)
if rec_idx is None:
continue
if rec_idx in matched_record_indexes:
continue
matched_record_indexes.add(rec_idx)
photos_raw = vehicle_data.get("photos") or []
images: list[ImageRecord] = []
seen_paths: set[str] = set()
@@ -1414,28 +1553,28 @@ class EncarScraper:
order_index=order,
))
images.sort(key=lambda x: x.order_index)
result.append((rec_idx, images))
result.append((rec_idx, images, vehicle_data))
return result
if resp.status in (429, 503):
time.sleep(1 + attempt * 2)
continue
return [(ri, []) for ri, _ in chunk]
return [(ri, [], {}) for ri, _ in chunk]
except json.JSONDecodeError:
return [(ri, []) for ri, _ in chunk]
return [(ri, [], {}) for ri, _ in chunk]
except Exception as exc:
logger.debug("Batch photo chunk error (attempt %d): %s", attempt + 1, exc)
if attempt < 1:
if attempt < 2:
time.sleep(1)
return [(ri, []) for ri, _ in chunk]
return [(ri, [], {}) for ri, _ in chunk]
results: dict[int, list[ImageRecord]] = {}
results: dict[int, tuple[list[ImageRecord], dict[str, Any]]] = {}
with ThreadPoolExecutor(max_workers=30) as executor:
futures = {executor.submit(_fetch_chunk, ch): ch for ch in chunks}
for future in as_completed(futures):
try:
for rec_idx, images in future.result():
if images:
results[rec_idx] = images
for rec_idx, images, vehicle_data in future.result():
if vehicle_data:
results[rec_idx] = (images, vehicle_data)
except Exception as exc:
logger.warning("Batch photo future failed: %s", exc)
@@ -1444,7 +1583,10 @@ class EncarScraper:
fallback_count = 0
for rec_idx, item in probe_tasks:
if rec_idx in results:
records[rec_idx].images = results[rec_idx]
images, vehicle_data = results[rec_idx]
if images:
records[rec_idx].images = images
self.car_mapper.apply_batch_color(records[rec_idx], vehicle_data)
api_ok += 1
else:
fallback_count += 1

View File

@@ -0,0 +1,326 @@
"""Korean -> English translation tables for EnCar (partner-provided basis).
Источник переводов от партнёра. Этот файл — эталон: бренды и модели берём
отсюда в первую очередь. Держим секционные словари плюс один плоский для
простых поисков.
"""
from __future__ import annotations
ENCAR_KO_EN_TABLES: dict[str, dict[str, str]] = {
"makers": {
"포르쉐": "Porsche",
"메르세데스-벤츠": "Mercedes-Benz",
"메르세데스 벤츠": "Mercedes-Benz",
"벤츠": "Mercedes-Benz",
"메르세데스-AMG": "Mercedes-AMG",
"메르세데스 AMG": "Mercedes-AMG",
"AMG": "AMG",
"BMW": "BMW",
"아우디": "Audi",
"폭스바겐": "Volkswagen",
"렉서스": "Lexus",
"토요타": "Toyota",
"닛산": "Nissan",
"인피니티": "Infiniti",
"혼다": "Honda",
"마쯔다": "Mazda",
"스바루": "Subaru",
"미쓰비시": "Mitsubishi",
"랜드로버": "Land Rover",
"재규어": "Jaguar",
"지프": "Jeep",
"볼보": "Volvo",
"푸조": "Peugeot",
"르노": "Renault",
"미니": "MINI",
"로터스": "Lotus",
"벤틀리": "Bentley",
"롤스로이스": "Rolls-Royce",
"람보르기니": "Lamborghini",
"페라리": "Ferrari",
"맥라렌": "McLaren",
"애스턴마틴": "Aston Martin",
"애스톤마틴": "Aston Martin",
"마세라티": "Maserati",
"코닉세그": "Koenigsegg",
"파가니": "Pagani",
"부가티": "Bugatti",
"캐딜락": "Cadillac",
"쉐보레": "Chevrolet",
"포드": "Ford",
"링컨": "Lincoln",
"크라이슬러": "Chrysler",
"닷지": "Dodge",
"": "Ram",
"테슬라": "Tesla",
"제네시스": "Genesis",
"현대": "Hyundai",
"기아": "Kia",
"KG모빌리티": "KG Mobility",
"쌍용": "SsangYong",
"알파로메오": "Alfa Romeo",
"피아트": "Fiat",
"시트로엥": "Citroen",
"스마트": "Smart",
"폴스타": "Polestar",
"리비안": "Rivian",
"루시드": "Lucid",
"BYD": "BYD",
"지리": "Geely",
"홍치": "Hongqi",
"이스즈": "Isuzu",
},
"models": {
"718 박스터": "718 Boxster",
"718 카이맨": "718 Cayman",
"911": "911",
"911 카브리올레": "911 Cabriolet",
"카이엔": "Cayenne",
"카이엔 쿠페": "Cayenne Coupe",
"파나메라": "Panamera",
"파나메라 스포츠 투리스모": "Panamera Sport Turismo",
"마칸": "Macan",
"타이칸": "Taycan",
"타이칸 크로스 투리스모": "Taycan Cross Turismo",
"918 스파이더": "918 Spyder",
"카레라 GT": "Carrera GT",
"718 스파이더": "718 Spyder",
"박스터": "Boxster",
"카이맨": "Cayman",
"4.0 스파이더": "4.0 Spyder",
"C 클래스": "C-Class",
"E 클래스": "E-Class",
"S 클래스": "S-Class",
"A 클래스": "A-Class",
"B 클래스": "B-Class",
"CLA": "CLA",
"CLS": "CLS",
"SL": "SL",
"SLK": "SLK",
"G 클래스": "G-Class",
"GLA": "GLA",
"GLB": "GLB",
"GLC": "GLC",
"GLE": "GLE",
"GLS": "GLS",
"AMG GT": "AMG GT",
"AMG GT 4도어 쿠페": "AMG GT 4-Door Coupe",
"V 클래스": "V-Class",
"1 시리즈": "1 Series",
"2 시리즈": "2 Series",
"3 시리즈": "3 Series",
"4 시리즈": "4 Series",
"5 시리즈": "5 Series",
"6 시리즈": "6 Series",
"7 시리즈": "7 Series",
"8 시리즈": "8 Series",
"M2": "M2",
"M3": "M3",
"M4": "M4",
"M5": "M5",
"M8": "M8",
"X1": "X1",
"X2": "X2",
"X3": "X3",
"X4": "X4",
"X5": "X5",
"X6": "X6",
"X7": "X7",
"Z4": "Z4",
"A3": "A3",
"A4": "A4",
"A5": "A5",
"A6": "A6",
"A7": "A7",
"A8": "A8",
"Q3": "Q3",
"Q5": "Q5",
"Q7": "Q7",
"Q8": "Q8",
"R8": "R8",
"TT": "TT",
"고스트": "Ghost",
"팬텀": "Phantom",
"레이스": "Wraith",
"": "Dawn",
"컬리넌": "Cullinan",
"스펙터": "Spectre",
"벤테이가": "Bentayga",
"컨티넨탈 GT": "Continental GT",
"플라잉스퍼": "Flying Spur",
"우라칸": "Huracan",
"아벤타도르": "Aventador",
"우루스": "Urus",
"레부엘토": "Revuelto",
"로마": "Roma",
"포르토피노": "Portofino",
"SF90 스트라달레": "SF90 Stradale",
"F8 트리뷰토": "F8 Tributo",
"콰트로포르테": "Quattroporte",
"기블리": "Ghibli",
"르반떼": "Levante",
"그란투리스모": "GranTurismo",
"반퀴시": "Vanquish",
"밴티지": "Vantage",
"DB11": "DB11",
"DB12": "DB12",
"DBS": "DBS",
},
"body_types": {
"세단": "Sedan",
"쿠페": "Coupe",
"컨버터블": "Convertible",
"카브리올레": "Cabriolet",
"오픈카": "Open Top",
"해치백": "Hatchback",
"왜건": "Wagon",
"스테이션왜건": "Station Wagon",
"SUV": "SUV",
"크로스오버": "Crossover",
"픽업트럭": "Pickup Truck",
"": "Van",
"미니밴": "Minivan",
"트럭": "Truck",
"쿠페형 SUV": "Coupe SUV",
"기타": "Other",
},
"fuel_types": {
"가솔린": "Gasoline",
"디젤": "Diesel",
"경유": "Diesel",
"하이브리드": "Hybrid",
"가솔린 하이브리드": "Gasoline Hybrid",
"플러그인 하이브리드": "Plug-in Hybrid",
"전기": "Electric",
"전기차": "Electric",
"수소": "Hydrogen",
"LPG": "LPG",
"LNG": "LNG",
"가솔린+전기": "Gasoline + Electric",
"가솔린 + 전기": "Gasoline + Electric",
},
"transmissions": {
"수동": "Manual",
"오토": "Automatic",
"자동": "Automatic",
"자동변속": "Automatic",
"세미오토": "Semi-automatic",
"CVT": "CVT",
"듀얼클러치": "Dual-clutch",
},
"sell_types": {
"일반": "General",
"리스": "Lease",
"렌트": "Rent",
"할부": "Installment",
},
"lease_types": {
"운용리스": "Operating Lease",
"금융리스": "Financial Lease",
},
"regions": {
"서울": "Seoul",
"부산": "Busan",
"인천": "Incheon",
"대구": "Daegu",
"대전": "Daejeon",
"광주": "Gwangju",
"울산": "Ulsan",
"세종": "Sejong",
"경기": "Gyeonggi",
"강원": "Gangwon",
"충북": "Chungbuk",
"충남": "Chungnam",
"전북": "Jeonbuk",
"전남": "Jeonnam",
"경북": "Gyeongbuk",
"경남": "Gyeongnam",
"제주": "Jeju",
},
"dealer_companies": {
"(주)기억": "Gieok Co., Ltd.",
"(주)네바퀴": "Four Wheels Co., Ltd.",
"(주)에이스타모터스": "A-Star Motors Co., Ltd.",
"(주)우진 오토모빌": "Woojin Automobile Co., Ltd.",
"(주)지온컴퍼니": "Zion Company Co., Ltd.",
"(주)플로시모터스": "Flossy Motors Co., Ltd.",
"㈜오토핸즈": "Auto Hands Co., Ltd.",
"GD카": "GD Car",
"MK모터스": "MK Motors",
"꿈차": "Dream Car",
"마이카4": "My Car 4",
"블루오토": "Blue Auto",
"세영모빌리티(주)": "Seyoung Mobility Co., Ltd.",
"스마일자동차": "Smile Motors",
"오토링크": "Auto Link",
"오토핸즈": "Auto Hands",
"우리들모터스": "Woorideul Motors",
"주식회사 모터바겐": "Motorwagen Co., Ltd.",
"주식회사 아이오토스": "iAutos Co., Ltd.",
"주식회사이들컴퍼니": "Ideul Company Co., Ltd.",
"카메이트": "Car Mate",
"카톡모터스": "Katok Motors",
"카피디": "Car PD",
"포르쉐 인증 중고차 센터 성수": "Porsche Certified Used Car Center Seongsu",
"포르쉐인증중고차센터성수": "Porsche Certified Used Car Center Seongsu",
},
"dealer_names": {
"고준용": "Go Jun-yong",
"공민중": "Gong Min-jung",
"김경진": "Kim Gyeong-jin",
"김명진": "Kim Myeong-jin",
"김민수": "Kim Min-su",
"김성언": "Kim Seong-eon",
"김성진": "Kim Seong-jin",
"김영균": "Kim Yeong-gyun",
"김완섭": "Kim Wan-seop",
"김현진": "Kim Hyeon-jin",
"김형진": "Kim Hyeong-jin",
"김호영": "Kim Ho-yeong",
"문명기": "Moon Myeong-gi",
"박원석": "Park Won-seok",
"박재현": "Park Jae-hyeon",
"방지윤": "Bang Ji-yun",
"송슬기": "Song Seul-gi",
"신영철": "Shin Yeong-cheol",
"신호진": "Shin Ho-jin",
"양시언": "Yang Si-eon",
"양태규": "Yang Tae-gyu",
"윤지완": "Yoon Ji-wan",
"이경민": "Lee Gyeong-min",
"이재욱": "Lee Jae-uk",
"임종현": "Lim Jong-hyeon",
"임형욱": "Lim Hyeong-uk",
"전태준": "Jeon Tae-jun",
"정상필": "Jeong Sang-pil",
"조경식": "Jo Gyeong-sik",
"조승재": "Jo Seung-jae",
"조환성": "Jo Hwan-seong",
"차상훈": "Cha Sang-hun",
"최경욱": "Choi Gyeong-uk",
"하상열": "Ha Sang-yeol",
"허지훈": "Heo Ji-hun",
},
}
def _build_flat_map() -> dict[str, str]:
result: dict[str, str] = {}
for section in ENCAR_KO_EN_TABLES.values():
result.update(section)
return result
ENCAR_KO_EN_TRANSLATIONS: dict[str, str] = _build_flat_map()
def translate_encar_ko_to_en(value: str, section: str | None = None) -> str:
"""Перевод корейского термина Encar на английский.
Если задан `section`, используется только словарь этой секции. Иначе —
плоский объединённый словарь.
"""
if section is not None:
return ENCAR_KO_EN_TABLES.get(section, {}).get(value, value)
return ENCAR_KO_EN_TRANSLATIONS.get(value, value)

View File

@@ -1,6 +1,10 @@
# Словари переводов корейского текста русский / английский.
# Используется в encar.py для маппинга данных Encar API.
import re
from .partner_dictionary import ENCAR_KO_EN_TABLES as _PARTNER_TABLES
# Топливо
@@ -176,6 +180,9 @@ BRAND_TRANSLATIONS = {
"기타 수입차": "Other Imported",
}
for _ko, _en in _PARTNER_TABLES["makers"].items():
BRAND_TRANSLATIONS.setdefault(_ko, _en)
# Бренды: английский → алиасы (корейский + русский)
@@ -277,6 +284,226 @@ def expand_allowed_brands(brands: set[str] | None) -> set[str] | None:
# Модели / комплектации: корейский → английский
MODEL_EXACT_TRANSLATIONS = {
"1시리즈": "1 Series",
"2시리즈": "2 Series",
"3시리즈": "3 Series",
"4시리즈": "4 Series",
"5시리즈": "5 Series",
"6시리즈": "6 Series",
"7시리즈": "7 Series",
"8시리즈": "8 Series",
"A-클래스": "A-Class",
"B-클래스": "B-Class",
"C-클래스": "C-Class",
"E-클래스": "E-Class",
"S-클래스": "S-Class",
"SL-클래스": "SL-Class",
"SLK-클래스": "SLK-Class",
"GLA-클래스": "GLA-Class",
"GLB-클래스": "GLB-Class",
"GLC-클래스": "GLC-Class",
"GLE-클래스": "GLE-Class",
"GLS-클래스": "GLS-Class",
"V-클래스": "V-Class",
"팰리세이드": "Palisade",
"싼타페": "Santa Fe",
"투싼": "Tucson",
"아반떼": "Avante",
"쏘나타": "Sonata",
"쏘나타 디 엣지": "Sonata The Edge",
"그랜저": "Grandeur",
"아이오닉5": "Ioniq 5",
"아이오닉6": "Ioniq 6",
"아이오닉": "Ioniq",
"넥쏘": "Nexo",
"벨로스터": "Veloster",
"코나": "Kona",
"캐스퍼": "Casper",
"베뉴": "Venue",
"스타렉스": "Starex",
"스타리아": "Staria",
"아슬란": "Aslan",
"베라크루즈": "Veracruz",
"카니발": "Carnival",
"쏘렌토": "Sorento",
"스포티지": "Sportage",
"에스컬레이드": "Escalade",
"프리미어": "Premier",
"1세대": "1st gen",
"2세대": "2nd gen",
"3세대": "3rd gen",
"4세대": "4th gen",
"5세대": "5th gen",
"6세대": "6th gen",
"7세대": "7th gen",
"8세대": "8th gen",
"모닝": "Morning",
"레이": "Ray",
"셀토스": "Seltos",
"니로": "Niro",
"스팅어": "Stinger",
"쏘울": "Soul",
"모하비": "Mohave",
"프라이드": "Pride",
"스토닉": "Stonic",
"카렌스": "Carens",
"제네시스": "Genesis",
"스파크": "Spark",
"말리부": "Malibu",
"트랙스": "Trax",
"크루즈": "Cruze",
"올란도": "Orlando",
"트레일블레이저": "Trailblazer",
"콜로라도": "Colorado",
"트래버스": "Traverse",
"캡티바": "Captiva",
"이쿼녹스": "Equinox",
"아베오": "Aveo",
"볼트": "Volt",
"임팔라": "Impala",
"맥스크루즈": "Max Cruz",
"렉스턴 스포츠 칸": "Rexton Sport Khan",
"렉스턴 스포츠": "Rexton Sport",
"렉스턴": "Rexton",
"토레스": "Torres",
"티볼리 에어": "Tivoli Air",
"티볼리 아머": "Tivoli Armor",
"티볼리": "Tivoli",
"코란도 투리스모": "Korando Turismo",
"코란도 스포츠": "Korando Sport",
"코란도": "Korando",
"체어맨": "Chairman",
"오피러스": "Opirus",
"익스플로러": "Explorer",
"머스탱": "Mustang",
"카이엔": "Cayenne",
"카이엔 쿠페": "Cayenne Coupe",
"파나메라": "Panamera",
"파나메라 스포츠 투리스모": "Panamera Sport Turismo",
"마칸": "Macan",
"타이칸": "Taycan",
"타이칸 크로스 투리스모": "Taycan Cross Turismo",
"918 스파이더": "918 Spyder",
"718 스파이더": "718 Spyder",
"4.0 스파이더": "4.0 Spyder",
"박스터": "Boxster",
"카이맨": "Cayman",
"911 카브리올레": "911 Cabriolet",
"카레라": "Carrera",
"티구안": "Tiguan",
"골프": "Golf",
"제타": "Jetta",
"투아렉": "Touareg",
"파사트": "Passat",
"아테온": "Arteon",
"쿠퍼": "Cooper",
"컨트리맨": "Countryman",
"클럽맨": "Clubman",
"디스커버리": "Discovery",
"레인지로버": "Range Rover",
"이보크": "Evoque",
"벨라": "Velar",
"디펜더": "Defender",
"랭글러": "Wrangler",
"체로키": "Cherokee",
"레니게이드": "Renegade",
"컴패스": "Compass",
"루비콘": "Rubicon",
"에비에이터": "Aviator",
"르반떼": "Levante",
"기블리": "Ghibli",
"어코드": "Accord",
"캠리": "Camry",
"프리우스": "Prius",
"타스만": "Tasman",
"토러스": "Taurus",
"카마로": "Camaro",
"비틀": "Beetle",
"e-트론": "e-tron",
"글래디에이터": "Gladiator",
"노틸러스": "Nautilus",
"시로코": "Scirocco",
"아틀라스": "Atlas",
"알파드": "Alphard",
"익스페디션": "Expedition",
"익스프레스Van": "Express Van",
"짐니": "Jimny",
"쿠가": "Kuga",
"클리오": "Clio",
"파일럿": "Pilot",
"콰트로포르테": "Quattroporte",
"고스트": "Ghost",
"팬텀": "Phantom",
"레이스": "Wraith",
"": "Dawn",
"컬리넌": "Cullinan",
"스펙터": "Spectre",
"벤테이가": "Bentayga",
"컨티넨탈 GT": "Continental GT",
"플라잉스퍼": "Flying Spur",
"우라칸": "Huracan",
"아벤타도르": "Aventador",
"우루스": "Urus",
"레부엘토": "Revuelto",
"로마": "Roma",
"포르토피노": "Portofino",
"SF90 스트라달레": "SF90 Stradale",
"F8 트리뷰토": "F8 Tributo",
"그란투리스모": "GranTurismo",
"반퀴시": "Vanquish",
"밴티지": "Vantage",
"AMG GT 4도어 쿠페": "AMG GT 4-Door Coupe",
}
def _register_partner_models() -> None:
for _ko, _en in _PARTNER_TABLES["models"].items():
MODEL_EXACT_TRANSLATIONS.setdefault(_ko, _en)
_compact = re.sub(r"(?<=\d)\s+시리즈", "시리즈", _ko)
_compact = re.sub(r"(?<=[A-Za-z])\s+클래스", "-클래스", _compact)
if _compact != _ko:
MODEL_EXACT_TRANSLATIONS.setdefault(_compact, _en)
_register_partner_models()
MODEL_PHRASE_TRANSLATIONS = {
"올 뉴": "All New",
"더 뉴": "New",
"디 올 뉴": "The All New",
"베리 뉴": "Very New",
"베리": "Very",
"뷰티풀": "Beautiful",
"넥스트": "Next",
"": "The ",
"디 엣지": "The Edge",
"액티브 투어러": "Active Tourer",
"투어러": "Tourer",
"그란쿠페": "GranCoupe",
"그란 쿠페": "GranCoupe",
"스포츠 칸": "Sport Khan",
"스포츠": "Sport",
"하이브리드": "Hybrid",
"일렉트릭": "Electric",
"네오": "Neo",
"트랜스폼": "Transform",
"프라임": "Prime",
"마스터": "Master",
"어반": "Urban",
"볼드": "Bold",
"아머": "Armor",
"에어": "Air",
}
MODEL_TOKEN_TRANSLATIONS = {
"시리즈": "Series",
"클래스": "Class",
"세대": "gen",
"": "New",
"": "All",
"": "Khan",
}
MODEL_TRANSLATIONS = {
# --- Топливо / тип двигателя ---
"가솔린": "Gasoline",
@@ -551,6 +778,20 @@ MODEL_TRANSLATIONS = {
"그란투리스모": "GranTurismo",
"반퀴시": "Vanquish",
"밴티지": "Vantage",
"시에라": "Sierra",
"시빅": "Civic",
"레인저": "Ranger",
"콜레오스": "Koleos",
"그랑 콜레오스": "Grand Koleos",
"페이톤": "Phaeton",
"티록": "T-Roc",
"큐브": "Cube",
"크라운": "Crown",
"라이즈": "Rise",
"시로코": "Scirocco",
"이오스": "Eos",
"비틀": "Beetle",
"폴로": "Polo",
# --- Служебные / категории ---
"(세부등급": "(sub-trim",
"없음)": "none)",
@@ -570,23 +811,25 @@ MODEL_TRANSLATIONS = {
}
# Цвета: корейский → английский
COLOR_TRANSLATIONS = {
"흰색": "white",
"검정색": "black",
"회색": "gray",
"은색": "silver",
"은회색": "silver-gray",
"진회색": "dark gray",
"쥐색": "mouse gray",
"밝은 회색": "light gray",
"청색": "blue",
"파란색": "blue",
"진청색": "dark blue",
"밝은 청색": "light blue",
"하늘색": "sky blue",
"녹색": "green",
"초록색": "green",
"진녹색": "dark green",
"밝은 녹색": "light green",
"담녹색": "light green",
"적색": "red",
"빨간색": "red",
"진적색": "dark red",
@@ -600,6 +843,11 @@ COLOR_TRANSLATIONS = {
"보라색": "purple",
"분홍색": "pink",
"금색": "gold",
"연금색": "light gold",
"명은색": "bright silver",
"은하색": "galaxy silver",
"자주색": "burgundy",
"진주색": "pearl",
"청금색": "blue-gold",
"청은색": "blue-silver",
"흑청색": "black-blue",

View File

@@ -1,5 +1,5 @@
[build-system]
requires = ["setuptools>=68", "wheel"]
requires = ["setuptools>=68"]
build-backend = "setuptools.build_meta"
[project]
@@ -23,6 +23,7 @@ dependencies = [
[project.optional-dependencies]
dev = [
"httpx>=0.28.0",
"pytest>=8.3.0",
"pytest-cov>=5.0.0",
]

View File

@@ -27,7 +27,7 @@ class TestEncarMapper(unittest.TestCase):
self.assertEqual(record.origin_id, "encar:41421262")
self.assertEqual(record.parser_id, "encar:41421262")
self.assertEqual(record.brand, "Hyundai")
self.assertEqual(record.model, "G90 (RS4)")
self.assertEqual(record.model, "G90")
self.assertEqual(record.year, 2023)
self.assertEqual(record.price, 85900000) # 8590 만원 × 10000
self.assertEqual(record.mileage, 37923)
@@ -40,10 +40,133 @@ class TestEncarMapper(unittest.TestCase):
url = ENCAR_DETAIL_URL_TEMPLATE.format(vehicle_id="41421262")
self.assertEqual(EncarMapper()._extract_vehicle_id(url), "41421262")
def test_map_to_car_record_uses_base_model_without_trim(self):
payload = {
"Id": 41854374,
"Manufacturer": "현대",
"Model": "Genesis",
"Badge": "BH330 Grand",
"BadgeDetail": "Prime팩",
"FormYear": "2011",
"Price": 1200,
"Mileage": 100000,
"Photos": [
{"location": "/carpicture01/pic4185/41854374_001.jpg", "ordering": 1},
],
}
record = EncarMapper().map_to_car_record(
ENCAR_DETAIL_URL_TEMPLATE.format(vehicle_id="41854374"),
payload,
)
self.assertEqual(record.brand, "Hyundai")
self.assertEqual(record.model, "Genesis")
self.assertEqual(record.origin_id, "encar:41854374")
def test_map_to_car_record_resolves_numeric_id_from_photos(self):
payload = {
"Id": "car-BOqLLkjkgbpStlVJEJvbRG",
"Manufacturer": "Kia",
"Model": "K3",
"FormYear": "2018",
"Price": 990,
"Mileage": 75000,
"Photos": [
{"location": "/carpicture01/pic4187/41878507_001.jpg", "ordering": 1},
],
}
record = EncarMapper().map_to_car_record("", payload)
self.assertEqual(record.origin_id, "encar:41878507")
self.assertEqual(record.parser_id, "encar:41878507")
self.assertIn("carid=41878507", record.origin_url)
def test_translate_model_keeps_english_text(self):
mapper = EncarMapper()
self.assertEqual(mapper._translate_model("All New Carnival"), "Carnival")
self.assertEqual(mapper._translate_model("New Kia Ray"), "Kia Ray")
self.assertEqual(mapper._translate_model("K5 Hybrid 3rd gen"), "K5 Hybrid")
self.assertEqual(mapper._translate_model("New K5 3rd gen"), "K5")
self.assertEqual(mapper._translate_model("New Tiguan"), "Tiguan")
self.assertEqual(mapper._translate_model("New QM6"), "QM6")
self.assertEqual(mapper._translate_model("Discovery Sport 2nd gen"), "Discovery Sport")
def test_translate_model_handles_korean_fragments(self):
mapper = EncarMapper()
cases = {
"5시리즈 (G30)": "5 Series",
"S-클래스 W223": "S-Class W223",
"Sonata 디 엣지(DN8)": "Sonata The Edge",
"렉스턴 Sport 칸": "Rexton Sport Khan",
"티볼리 Air": "Tivoli Air",
"코란도 투리스모": "Korando Turismo",
"Kia Carnival 4세대": "Kia Carnival",
"Cadillac 에스컬레이드 5세대": "Cadillac Escalade",
"K7 프리미어": "K7 Premier",
}
for raw, expected in cases.items():
with self.subTest(raw=raw):
self.assertEqual(mapper._translate_model(raw), expected)
def test_translate_model_drops_untranslated_korean(self):
mapper = EncarMapper()
cases = {
"슈팅브레이크": "",
"The New G70 슈팅브레이크": "G70",
"G70 (슈팅브레이크)": "G70",
}
for raw, expected in cases.items():
with self.subTest(raw=raw):
self.assertEqual(mapper._translate_model(raw), expected)
def test_translate_color_normalizes_api_spacing(self):
mapper = EncarMapper()
self.assertEqual(mapper._translate_color("흰색"), "white")
self.assertEqual(mapper._translate_color("흰 색"), "white")
self.assertEqual(mapper._translate_color("검정색"), "black")
self.assertEqual(mapper._translate_color("진주색"), "pearl")
self.assertEqual(mapper._translate_color("쥐색"), "mouse gray")
self.assertEqual(mapper._translate_color("은회색"), "silver-gray")
self.assertEqual(mapper._translate_color("하늘색"), "sky blue")
self.assertEqual(mapper._translate_color("자주색"), "burgundy")
def test_apply_batch_color_uses_spec_color_name(self):
payload = {
"Id": 42456458,
"Manufacturer": "Lincoln",
"Model": "Corsair",
"FormYear": "2022",
"Price": 2690,
"Mileage": 26600,
}
mapper = EncarMapper()
record = mapper.map_to_car_record(
ENCAR_DETAIL_URL_TEMPLATE.format(vehicle_id="42456458"),
payload,
)
mapper.apply_batch_color(record, {"spec": {"colorName": "흰 색"}})
self.assertEqual(record.brand, "Lincoln")
self.assertEqual(record.model, "Corsair")
self.assertEqual(record.year, 2022)
self.assertEqual(record.price, 26900000)
self.assertEqual(record.mileage, 26600)
self.assertEqual(record.color, "white")
class TestEncarScraper(unittest.TestCase):
def test_collect_listing_limits_results(self):
scraper = EncarScraper()
scraper._fetch_listing_page = MagicMock(side_effect=[
{"SearchResults": [{"Id": 1}, {"Id": 2}]},
])

View File

@@ -195,11 +195,14 @@ class TestAPILoad(unittest.TestCase):
persistence: PersistenceService = app.state.persistence
persistence.upsert_cars_batch([_make_car(i, images=4) for i in range(200)])
cls.app = app
cls.client = TestClient(app)
@classmethod
def tearDownClass(cls):
import os
cls.client.close()
cls.app.state.persistence.engine.dispose()
if os.path.exists("test_api_load.db"):
os.remove("test_api_load.db")