blackteay/DrissionPage
1
0
Fork 0
mirror of https://gitee.com/g1879/DrissionPage.git synced 2024-12-10 04:00:23 +08:00
Code Issues Projects Releases Wiki Activity

修改获取href的代码，避免当相对URL以?开头时requests-html丢失参数的bug

Browse Source
This commit is contained in:
g1879 2020-06-02 00:03:43 +08:00
parent d317e0330e
commit a31ed1d354
1 changed files with 8 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
DrissionPage/session_element.py
View File

@ -11,7 +11,6 @@ from typing import Union, List
from requests_html import Element, BaseParser from requests_html import Element, BaseParser
from .common import DrissionElement, get_loc_from_str, translate_loc_to_xpath from .common import DrissionElement, get_loc_from_str, translate_loc_to_xpath
from urllib.parse import urlparse, urljoin
class SessionElement(DrissionElement): class SessionElement(DrissionElement):
@ -105,14 +104,15 @@ class SessionElement(DrissionElement):
"""获取属性值""" """获取属性值"""
try: try:
if attr == 'href': if attr == 'href':
# TODO: 须测试
# 如直接获取attr只能获取相对地址 # 如直接获取attr只能获取相对地址
link = self._inner_ele.attrs['href'] link = self._inner_ele.attrs['href']
parsed = urlparse(link) if link.startswith('?'): # 避免当相对URL以?开头时requests-html丢失参数的bug
if not parsed.netloc: if '?' in self.inner_ele.url:
return urljoin(self._inner_ele.url, link) return re.sub(r'\?.*', link, self.inner_ele.url)
if not parsed.scheme: else:
return urljoin(urlparse(self._inner_ele.url).scheme, link) return f'{self.inner_ele.url}{link}'
else:
for link in self._inner_ele.absolute_links:
return link return link
elif attr == 'class': elif attr == 'class':
class_str = '' class_str = ''