mirror of
git://github.com/kovidgoyal/calibre.git
synced 2025-12-21 20:46:17 +01:00
Updated Wall Street Journal recipe
This commit is contained in:
parent
e0e6ddb6cf
commit
5faeaba6c2
1 changed files with 5 additions and 1 deletions
|
|
@ -53,6 +53,10 @@ def get_browser(self):
|
|||
def postprocess_html(self, soup, first):
|
||||
for tag in soup.findAll(name=['table', 'tr', 'td']):
|
||||
tag.name = 'div'
|
||||
|
||||
for tag in soup.findAll('div', dict(id=["articleImage_1", "articleImage_2", "articleImage_3", "articleImage_4", "articleImage_5", "articleImage_6", "articleImage_7"])):
|
||||
tag.extract()
|
||||
|
||||
return soup
|
||||
|
||||
def get_article_url(self, article):
|
||||
|
|
@ -70,7 +74,7 @@ def cleanup(self):
|
|||
#('Most Emailed - Month', 'http://online.wsj.com/xml/rss/3_7254.xml'),
|
||||
(' Most Viewed - Day', 'http://online.wsj.com/xml/rss/3_7198.xml'),
|
||||
(' Most Viewed - Week', 'http://online.wsj.com/xml/rss/3_7251.xml'),
|
||||
# ('Most Viewed - Month', 'http://online.wsj.com/xml/rss/3_7252.xml'),
|
||||
#('Most Viewed - Month', 'http://online.wsj.com/xml/rss/3_7252.xml'),
|
||||
('Today\'s Newspaper - Page One', 'http://online.wsj.com/xml/rss/3_7205.xml'),
|
||||
('Today\'s Newspaper - Marketplace', 'http://online.wsj.com/xml/rss/3_7206.xml'),
|
||||
('Today\'s Newspaper - Money & Investing', 'http://online.wsj.com/xml/rss/3_7207.xml'),
|
||||
|
|
|
|||
Loading…
Reference in a new issue