Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

...

Browse Source
This commit is contained in:
Kovid Goyal 2013-01-11 12:41:52 +05:30
parent 0b40e8a864
commit 4c09d25062
1 changed files with 3 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
recipes/bwmagazine2.recipe
View File

@ -11,16 +11,15 @@ class BusinessWeekMagazine(BasicNewsRecipe):
category = 'news' category = 'news'
encoding = 'UTF-8' encoding = 'UTF-8'
keep_only_tags = [ keep_only_tags = [
dict(name='div', attrs={'id':'article_body_container'}), dict(name='div', attrs={'id':'article_body_container'}),
] ]
remove_tags = [dict(name='ui'),dict(name='li')] remove_tags = [dict(name='ui'),dict(name='li'),dict(name='div', attrs={'id':['share-email']})]
no_javascript = True no_javascript = True
no_stylesheets = True no_stylesheets = True
cover_url = 'http://images.businessweek.com/mz/covers/current_120x160.jpg' cover_url = 'http://images.businessweek.com/mz/covers/current_120x160.jpg'
def parse_index(self): def parse_index(self):
#Go to the issue #Go to the issue
soup = self.index_to_soup('http://www.businessweek.com/magazine/news/articles/business_news.htm') soup = self.index_to_soup('http://www.businessweek.com/magazine/news/articles/business_news.htm')
@ -47,7 +46,6 @@ class BusinessWeekMagazine(BasicNewsRecipe):
if section_title not in feeds: if section_title not in feeds:
feeds[section_title] = [] feeds[section_title] = []
feeds[section_title] += articles feeds[section_title] += articles
div1 = soup.find ('div', attrs={'class':'column center'}) div1 = soup.find ('div', attrs={'class':'column center'})
section_title = '' section_title = ''
for div in div1.findAll('h5'): for div in div1.findAll('h5'):