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

Update Indian Express

Browse Source
This commit is contained in:
Kovid Goyal 2022-04-04 14:53:53 +05:30
parent 575e8e7e63
commit 7944a8a022
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 15 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
recipes/indian_express.recipe
View File

@ -14,11 +14,13 @@ class IndianExpress(BasicNewsRecipe):
oldest_article = 1 # days
max_articles_per_feed = 25
encoding = 'utf-8'
masthead_url = 'https://indianexpress.com/wp-content/themes/indianexpress/images/indian-express-logo-n.svg'
no_stylesheets = True
use_embedded_content = False
remove_attributes = ['style','height','width']
ignore_duplicate_articles = {'url'}
extra_css = '#storycenterbyline {font-size:small};'
keep_only_tags = [
classes('heading-part full-details')
@ -29,7 +31,10 @@ class IndianExpress(BasicNewsRecipe):
dict(name='a', attrs={'href':'https://indianexpress.com/section/explained/?utm_source=newbanner'}),
dict(name='img', attrs={'src':'https://images.indianexpress.com/2021/06/opinion-button-300-ie.jpeg'}),
dict(name='a', attrs={'href':'https://indianexpress.com/section/opinion/?utm_source=newbanner'}),
classes('share-social appstext story-tags ie-int-campign-ad ie-breadcrumb custom_read_button unitimg copyright')
classes(
'share-social appstext ie-int-campign-ad ie-breadcrumb custom_read_button unitimg copyright'
' storytags pdsc-related-modify news-guard'
)
]
feeds = [
('Front Page', 'https://indianexpress.com/print/front-page/feed/'),
@ -41,9 +46,18 @@ class IndianExpress(BasicNewsRecipe):
('Delhi Confidential', 'https://indianexpress.com/section/delhi-confidential/feed'),
('Economy', 'https://indianexpress.com/print/economy/feed'),
('Express Network', 'https://indianexpress.com/print/express-network/'),
('Research', 'https://indianexpress.com/section/research/feed/'),
('UPSC-CSE Key','https://indianexpress.com/section/upsc-current-affairs/feed/'),
('World','https://indianexpress.com/section/world/feed/'),
('Business', 'https://indianexpress.com/section/business/feed/'),
# Want to add more? go-to:https://indianexpress.com/syndication/
]
def get_cover_url(self):
soup = self.index_to_soup('https://www.magzter.com/IN/The-Indian-Express-Ltd./The-Indian-Express-Mumbai/Newspaper/')
for citem in soup.findAll('meta', content=lambda s: s and s.endswith('view/3.jpg')):
return citem['content']
def preprocess_html(self, soup):
for img in soup.findAll('img'):
noscript = img.findParent('noscript')