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

把获取到的文本中的\xa0替换成空格

Browse Source
This commit is contained in:
g1879 2020-11-03 17:26:28 +08:00
parent 126fd466b2
commit bbea8871a0
2 changed files with 6 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
DrissionPage/driver_element.py
View File

@ -591,11 +591,12 @@ class ElementsByXpath(object):
if self.mode == 'single':
try:
e = get_nodes(the_node, xpath_txt=self.xpath, type_txt='9')
return DriverElement(e, self.timeout) if isinstance(e, WebElement) else e
return DriverElement(e, self.timeout) if isinstance(e, WebElement) else unescape(e).replace('\xa0', ' ')
except JavascriptException: # 找不到目标时
return None
elif self.mode == 'all':
e = get_nodes(the_node, xpath_txt=self.xpath)
e = filter(lambda x: x != '\n', e) # 去除元素间换行符
e = map(lambda x: unescape(x).replace('\xa0', ' ') if isinstance(x, str) else x, e) # 替换空格
return list(map(lambda x: DriverElement(x, self.timeout) if isinstance(x, WebElement) else x, e))

8
DrissionPage/session_element.py
View File

@ -37,11 +37,10 @@ class SessionElement(DrissionElement):
return unescape(self._inner_ele.text).replace('\xa0', ' ')
def texts(self, text_node_only: bool = False) -> List[str]:
nodes = self.eles('xpath:./*/node()')
if text_node_only:
return [x for x in nodes if isinstance(x, str)]
return self.eles('xpath:./*/text()')
else:
return [x if isinstance(x, str) else x.text for x in nodes]
return [x if isinstance(x, str) else x.text for x in self.eles('xpath:./*/node()')]
@property
def html(self) -> str:
@ -274,9 +273,10 @@ def execute_session_find(page_or_ele: BaseParser,
if mode == 'single':
ele = ele[0] if ele else None
return SessionElement(ele) if isinstance(ele, Element) else ele
return SessionElement(ele) if isinstance(ele, Element) else unescape(ele).replace('\xa0', ' ')
elif mode == 'all':
ele = filter(lambda x: x != '\n', ele) # 去除元素间换行符
ele = map(lambda x: unescape(x).replace('\xa0', ' ') if isinstance(x, str) else x, ele) # 替换空格
return [SessionElement(e) if isinstance(e, Element) else e for e in ele]
except:
if show_errmsg: