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 Live Mint

Browse Source
This commit is contained in:
Kovid Goyal 2022-02-17 22:41:31 +05:30
parent d2baa161c7
commit bbfa51852c
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 42 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

63
recipes/livemint.recipe
View File

@ -1,4 +1,6 @@
from calibre.web.feeds.news import BasicNewsRecipe #!/usr/bin/env python
from calibre.web.feeds.news import BasicNewsRecipe, classes
class LiveMint(BasicNewsRecipe): class LiveMint(BasicNewsRecipe):
@ -7,26 +9,45 @@ class LiveMint(BasicNewsRecipe):
__author__ = 'Krittika Goyal' __author__ = 'Krittika Goyal'
oldest_article = 1 # days oldest_article = 1 # days
max_articles_per_feed = 50 max_articles_per_feed = 50
encoding = 'utf-8'
use_embedded_content = False use_embedded_content = False
no_stylesheets = True no_stylesheets = True
auto_cleanup = True remove_attributes = ['style', 'height', 'width']
feeds = [
('Companies','https://www.livemint.com/rss/companies'), keep_only_tags = [
('Opinion','https://www.livemint.com/rss/opinion'), dict(name='h1'),
('Money','https://www.livemint.com/rss/money'), dict(name='picture'),
('Politics','https://www.livemint.com/rss/politics'), dict(name='figcaption'),
('Science','https://www.livemint.com/rss/science'), classes('articleInfo FirstEle summary highlights paywall'),
('Industry','https://www.livemint.com/rss/industry'),
('Lounge','https://www.livemint.com/rss/lounge'),
('Education','https://www.livemint.com/rss/education'),
('Sports','https://www.livemint.com/rss/sports'),
('Technology','https://www.livemint.com/rss/technology'),
('News','https://www.livemint.com/rss/news'),
('Mutual Funds','https://www.livemint.com/rss/Mutual Funds'),
('Markets','https://www.livemint.com/rss/markets'),
('AI','https://www.livemint.com/rss/AI'),
('Insurance','https://www.livemint.com/rss/insurance'),
('Budget','https://www.livemint.com/rss/budget'),
('Elections','https://www.livemint.com/rss/elections'),
] ]
remove_tags = [
classes(
'trendingSimilarHeight moreNews mobAppDownload label msgError msgOk'
)
]
feeds = [
('Companies', 'https://www.livemint.com/rss/companies'),
('Opinion', 'https://www.livemint.com/rss/opinion'),
('Money', 'https://www.livemint.com/rss/money'),
('Economy', 'https://www.livemint.com/rss/economy/'),
('Politics', 'https://www.livemint.com/rss/politics'),
('Science', 'https://www.livemint.com/rss/science'),
('Industry', 'https://www.livemint.com/rss/industry'),
('Lounge', 'https://www.livemint.com/rss/lounge'),
('Education', 'https://www.livemint.com/rss/education'),
('Sports', 'https://www.livemint.com/rss/sports'),
('Technology', 'https://www.livemint.com/rss/technology'),
('News', 'https://www.livemint.com/rss/news'),
('Mutual Funds', 'https://www.livemint.com/rss/Mutual Funds'),
('Markets', 'https://www.livemint.com/rss/markets'),
('AI', 'https://www.livemint.com/rss/AI'),
('Insurance', 'https://www.livemint.com/rss/insurance'),
('Budget', 'https://www.livemint.com/rss/budget'),
('Elections', 'https://www.livemint.com/rss/elections'),
]
def preprocess_html(self, soup):
for img in soup.findAll('img', attrs={'data-src': True}):
img['src'] = img['data-src']
return soup