Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 18:54:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix #4768 (Jerusalem Post Failed Feed)

Browse Source
This commit is contained in:
Kovid Goyal 2010-02-08 21:17:35 -07:00
parent 3b28deb26e
commit 17c0f31808
1 changed files with 10 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
resources/recipes/jpost.recipe
View File

@ -10,22 +10,19 @@ class JerusalemPost(BasicNewsRecipe):
__author__ = 'Kovid Goyal' __author__ = 'Kovid Goyal'
max_articles_per_feed = 10 max_articles_per_feed = 10
no_stylesheets = True no_stylesheets = True
remove_tags_before = {'class':'byline'} remove_tags_before = {'class':'jp-grid-content'}
remove_tags = [ remove_tags_after = {'id':'body_val'}
{'class':['artAdBlock clearboth', 'tbartop', 'divdot_vrttbox',
'slideshow']},
dict(id=['artFontButtons', 'artRelatedBlock']),
]
remove_tags_after = {'id':'artTxtBlock'}
feeds = [ ('Front Page', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1123495333346'), feeds = [ ('Front Page', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1123495333346'),
('Israel News', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1178443463156'), ('Israel News', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1178443463156'),
('Middle East News', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1123495333498'), ('Middle East News', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1123495333498'),
('International News', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1178443463144'), ('International News', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1178443463144'),
('Editorials', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1123495333211'), ('Editorials', 'http://www.jpost.com/servlet/Satellite?pagename=JPost/Page/RSS&cid=1123495333211'),
] ]
def postprocess_html(self, soup, first): def preprocess_html(self, soup):
for tag in soup.findAll(name=['table', 'tr', 'td']): for x in soup.findAll(name=['form', 'input']):
tag.name = 'div' x.name = 'div'
return soup for x in soup.findAll('body', style=True):
del x['style']
return soup