mirror of
https://github.com/searxng/searxng.git
synced 2024-11-22 20:17:45 +01:00
[fix] handle single closing element in HTMLTextExtractor
This commit is contained in:
parent
06186e72a9
commit
699fe60e50
@ -82,8 +82,12 @@ class HTMLTextExtractor(HTMLParser):
|
|||||||
self.tags.append(tag)
|
self.tags.append(tag)
|
||||||
|
|
||||||
def handle_endtag(self, tag):
|
def handle_endtag(self, tag):
|
||||||
|
if not self.tags:
|
||||||
|
return
|
||||||
|
|
||||||
if tag != self.tags[-1]:
|
if tag != self.tags[-1]:
|
||||||
raise Exception("invalid html")
|
raise Exception("invalid html")
|
||||||
|
|
||||||
self.tags.pop()
|
self.tags.pop()
|
||||||
|
|
||||||
def is_valid_tag(self):
|
def is_valid_tag(self):
|
||||||
|
Loading…
Reference in New Issue
Block a user