58 lines
2.0 KiB
Python
58 lines
2.0 KiB
Python
"""子站解析器注册表与商品页分派
|
|
|
|
乐天部分官方旗舰店的商品页会 302 跳出 item.rakuten.co.jp，落到各自独立的站点。
|
|
这里按落地域名把页面分派给对应解析器；落到未登记的站点时抛 OffIchibaRedirectError，
|
|
让上游能明确区分「站点不支持」与「被反爬拦截」，而不是白白重试。
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
from urllib.parse import urlsplit
|
|
|
|
from app.core import site
|
|
from app.core.errors import OffIchibaRedirectError, ScrapeParseError
|
|
from app.models.scrape import ItemDetailData
|
|
from app.parsers.state import PageValidator, require_state_marker
|
|
from app.parsers.subsites import biccamera, books, brandavenue
|
|
from app.parsers.subsites.base import SubsitePage, SubsiteParser
|
|
|
|
SUBSITE_PARSERS: dict[str, SubsiteParser] = {
|
|
module.HOST: SubsiteParser(
|
|
host=module.HOST,
|
|
source=module.SOURCE,
|
|
shop_name=module.SHOP_NAME,
|
|
validate=module.validate,
|
|
parse=module.parse,
|
|
)
|
|
for module in (books, brandavenue, biccamera)
|
|
}
|
|
|
|
|
|
def host_of(url: str) -> str:
|
|
return urlsplit(url).hostname or ""
|
|
|
|
|
|
def build_item_page_validator(requested_url: str) -> PageValidator:
|
|
"""构造商品页校验器：按落地域名选用对应的页面校验规则
|
|
|
|
落到未登记的站点时直接抛错——换 cookie 或上浏览器都改变不了页面归属。
|
|
"""
|
|
|
|
def validate(html: str, final_url: str) -> str | None:
|
|
host = host_of(final_url)
|
|
if host == site.ITEM_HOST:
|
|
return require_state_marker(html, final_url)
|
|
parser = SUBSITE_PARSERS.get(host)
|
|
if parser is None:
|
|
raise OffIchibaRedirectError(requested_url, final_url)
|
|
return parser.validate(html)
|
|
|
|
return validate
|
|
|
|
|
|
def parse_subsite_item(page: SubsitePage) -> ItemDetailData:
|
|
"""把子站页面交给对应解析器"""
|
|
parser = SUBSITE_PARSERS.get(host_of(page.final_url))
|
|
if parser is None:
|
|
raise ScrapeParseError(f"没有匹配的子站解析器：{page.final_url}")
|
|
return parser.parse(page)
|