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

SessionElement的attr()删除对#和?开头相对路径的处理

Browse Source
This commit is contained in:
g1879 2020-11-07 15:46:35 +08:00
parent 80c8dab34c
commit e439c1f679
1 changed files with 8 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
DrissionPage/session_element.py
View File

@ -231,23 +231,18 @@ class SessionElement(DrissionElement):
:return: 属性值文本，没有该属性返回None :return: 属性值文本，没有该属性返回None
""" """
try: try:
# 获取href属性时返回绝对url
if attr == 'href': if attr == 'href':
# 如直接获取attr只能获取相对地址
link = self.inner_ele.get('href') link = self.inner_ele.get('href')
# 若链接为js或邮件，直接返回
if link.lower().startswith(('javascript:', 'mailto:')): if link.lower().startswith(('javascript:', 'mailto:')):
return link return link
elif link.startswith('#'):
if '#' in self.page.url: # 其它情况直接返回绝对url
return re.sub(r'#.*', link, self.page.url)
else:
return f'{self.page.url}{link}'
elif link.startswith('?'): # 避免当相对url以?开头时丢失参数的bug TODO:测试是否还存在
if '?' in self.page.url:
return re.sub(r'\?.*', link, self.page.url)
else:
return f'{self.page.url}{link}'
else: else:
return self._make_absolute(link) return self._make_absolute(link)
elif attr == 'src': elif attr == 'src':
return self._make_absolute(self.inner_ele.get('src')) return self._make_absolute(self.inner_ele.get('src'))
elif attr == 'text': elif attr == 'text':
@ -293,12 +288,13 @@ def execute_session_find(page_or_ele,
if mode not in ['single', 'all']: if mode not in ['single', 'all']:
raise ValueError(f"Argument mode can only be 'single' or 'all', not '{mode}'.") raise ValueError(f"Argument mode can only be 'single' or 'all', not '{mode}'.")
# 根据传入对象类型获取页面对象和lxml元素对象
if isinstance(page_or_ele, SessionElement): if isinstance(page_or_ele, SessionElement):
page = page_or_ele.page page = page_or_ele.page
page_or_ele = page_or_ele.inner_ele page_or_ele = page_or_ele.inner_ele
else: # 传入的是SessionPage对象 else: # 传入的是SessionPage对象
page = page_or_ele page = page_or_ele
page_or_ele = get_HtmlElement(page_or_ele.response.text) page_or_ele = HTML(page_or_ele.response.text)
try: try:
# 用lxml内置方法获取lxml的元素对象列表 # 用lxml内置方法获取lxml的元素对象列表
@ -328,8 +324,3 @@ def execute_session_find(page_or_ele,
except SelectorSyntaxError: except SelectorSyntaxError:
raise SyntaxError('Invalid css selector syntax.', loc) raise SyntaxError('Invalid css selector syntax.', loc)
def get_HtmlElement(html: str) -> _Element:
"""从html文本生成lxml的元素对象"""
return HTML(html)