Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 10:44:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Live Mint: Add support for Saturday special edition

Browse Source
This commit is contained in:
Kovid Goyal 2022-02-20 13:19:31 +05:30
parent 03cac41da7
commit a4e55ce1e6
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 59 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

90
recipes/livemint.recipe
View File

@ -1,10 +1,14 @@
#!/usr/bin/env python
from calibre.web.feeds.news import BasicNewsRecipe, classes
from datetime import date
is_saturday = date.today().weekday() == 5
class LiveMint(BasicNewsRecipe):
title = u'Live Mint'
description = 'Financial News from India.'
language = 'en_IN'
__author__ = 'Krittika Goyal'
oldest_article = 1 # days
@ -14,40 +18,64 @@ class LiveMint(BasicNewsRecipe):
no_stylesheets = True
remove_attributes = ['style', 'height', 'width']
keep_only_tags = [
dict(name='h1'),
dict(name='picture'),
dict(name='figcaption'),
classes('articleInfo FirstEle summary highlights paywall'),
]
remove_tags = [
classes(
'trendingSimilarHeight moreNews mobAppDownload label msgError msgOk'
)
]
if is_saturday:
keep_only_tags = [
dict(name='h1'),
dict(name='h2', attrs={'id':'story-summary-0'}),
dict(name='picture'),
dict(name='div', attrs={'class':'innerBanCaption'}),
dict(name='div', attrs={'id':'date-display-before-content'}),
dict(name='div', attrs={'class':'storyContent'}),
]
remove_tags = [
classes(
'sidebarAdv similarStoriesClass moreFromSecClass'
)
]
feeds = [
('News', 'https://lifestyle.livemint.com/rss/news'),
('Food','https://lifestyle.livemint.com/rss/food'),
('Fashion','https://lifestyle.livemint.com/rss/fashion'),
('How to Lounge','https://lifestyle.livemint.com/rss/how-to-lounge'),
('Smart Living','https://lifestyle.livemint.com/rss/smart-living'),
]
else:
keep_only_tags = [
dict(name='h1'),
dict(name='picture'),
dict(name='figcaption'),
classes('articleInfo FirstEle summary highlights paywall'),
]
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'),
]
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'),
('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']
if is_saturday:
for img in soup.findAll('img', attrs={'data-img': True}):
img['src'] = img['data-img']
return soup