80 lines
2.9 KiB
Python
80 lines
2.9 KiB
Python
"""乐天店铺页 __INITIAL_STATE__ → ShopDetailData
|
|
|
|
店铺首页（www.rakuten.co.jp/{店铺代码}/）与搜索页一样把状态内联在
|
|
`window.__INITIAL_STATE__` 里，店铺信息在顶层的 `shop` 节点下，结构比商品
|
|
详情页里的 shop.information 更完整（多出招牌图、39ショップ标记、休息日）。
|
|
|
|
店铺 logo 不在 state 里，只出现在页面的 ld+json AggregateRating 微数据中。
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
import re
|
|
from typing import Any
|
|
|
|
from app.core import site
|
|
from app.core.errors import ScrapeParseError
|
|
from app.models.scrape import ShopDetailData
|
|
from app.utils.coerce import as_dict, as_float, as_int, as_list, as_str
|
|
|
|
_LD_JSON_RE = re.compile(
|
|
r'<script[^>]*type="application/ld\+json"[^>]*>(.*?)</script>', re.S
|
|
)
|
|
|
|
|
|
def _parse_logo_url(html: str) -> str:
|
|
"""从 ld+json 的 AggregateRating.itemReviewed 里取店铺 logo"""
|
|
for match in _LD_JSON_RE.finditer(html):
|
|
try:
|
|
data = json.loads(match.group(1))
|
|
except ValueError:
|
|
continue
|
|
if not isinstance(data, dict) or data.get("@type") != "AggregateRating":
|
|
continue
|
|
reviewed = data.get("itemReviewed")
|
|
if isinstance(reviewed, dict):
|
|
return as_str(reviewed.get("logo"))
|
|
return ""
|
|
|
|
|
|
def parse_shop_detail(
|
|
state: dict[str, Any], *, shop_code: str, html: str = ""
|
|
) -> ShopDetailData:
|
|
"""把店铺页状态解析为商家详情
|
|
|
|
Raises:
|
|
ScrapeParseError: 状态中不存在 shop 节点
|
|
"""
|
|
shop = state.get("shop")
|
|
if not isinstance(shop, dict) or not shop:
|
|
raise ScrapeParseError("店铺页缺少 shop 节点")
|
|
|
|
review = as_dict(shop.get("shopReview"))
|
|
resolved_code = as_str(shop.get("shopCode")) or shop_code
|
|
# state 里的 shopUrl 是 http 的且不带尾斜杠，统一按规范形式给出
|
|
shop_url = f"{site.WWW_BASE_URL}{resolved_code}/" if resolved_code else ""
|
|
|
|
holidays = [
|
|
as_str(day)
|
|
for day in as_list(as_dict(as_dict(shop.get("shopCalendar")).get("eventDates")).get("holiday"))
|
|
if as_str(day)
|
|
]
|
|
|
|
return ShopDetailData(
|
|
shop_id=as_int(shop.get("shopId")) or None,
|
|
shop_code=resolved_code,
|
|
shop_name=as_str(shop.get("shopName")),
|
|
shop_url=shop_url,
|
|
introduction=as_str(shop.get("shopIntroduction")),
|
|
signboard_url=as_str(shop.get("signboardUrl")),
|
|
logo_url=_parse_logo_url(html),
|
|
review_score=as_float(review.get("average")),
|
|
review_count=as_int(review.get("total")),
|
|
# 评价数过少时站点不展示评分，此时 average 不可信
|
|
review_displayed=bool(review.get("didMeetDisplayRequirements")),
|
|
is_39_shop=bool(shop.get("is39Shop")),
|
|
age_verification_required=bool(shop.get("ageVerificationFlag")),
|
|
status=as_int(shop.get("status")) if shop.get("status") is not None else None,
|
|
holidays=holidays,
|
|
)
|