Backup_Repos/FanFicFare
1
0
Fork 0
mirror of https://github.com/JimmXinu/FanFicFare.git synced 2026-05-08 12:36:11 +02:00
Code Issues Projects Releases Wiki Activity

Re-soup in make_soup instead of utf8FromSoup.

Browse source
This commit is contained in:
Jim Miller 2015-01-04 17:07:30 -06:00
parent a3d5f9f5fb
commit 53f87344de
1 changed files with 4 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
fanficdownloader/adapters/base_adapter.py
View file

@ -500,9 +500,6 @@ class BaseSiteAdapter(Configurable):
if not fetch:
fetch=self._fetchUrlRaw
# re-soup because BS4/html5lib is more forgiving that way.
soup = self.make_soup(unicode(soup))
acceptable_attributes = ['href','name','class','id']
if self.getConfig("keep_style_attr"):
acceptable_attributes.append('style')
@ -572,7 +569,10 @@ class BaseSiteAdapter(Configurable):
Convenience method for getting a bs4 soup. Older and
non-updated adapters call the included bs3 library themselves.
'''
return bs4.BeautifulSoup(data,'html5lib')
## soup and re-soup because BS4/html5lib is more forgiving of
## incorrectly nested tags that way.
soup = bs4.BeautifulSoup(data,'html5lib')
return bs4.BeautifulSoup(unicode(soup),'html5lib')
def cachedfetch(realfetch,cache,url):
if url in cache: