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

Recipe: add cover_url to The Hindu

Browse Source 
Gets today's front page image displayed in Hindu ePaper website.

Also add some styles to the lead image and caption.
This commit is contained in:
Shiva Prasad 2021-06-13 20:56:18 +05:30 committed by GitHub
parent 7d3fc0caba
commit 5df3b224a6
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 12 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
recipes/hindu.recipe
View File

@ -15,12 +15,15 @@ def classes(classes):
class TheHindu(BasicNewsRecipe):
title = u'The Hindu'
language = 'en_IN'
epaper_url = 'https://epaper.thehindu.com'
oldest_article = 1
__author__ = 'Kovid Goyal'
max_articles_per_feed = 100
no_stylesheets = True
remove_attributes = ['style']
extra_css = '.lead-img-cont { text-align: center; } ' \
'.lead-img-caption { font-size: small; font-style: italic; }'
ignore_duplicate_articles = {'title', 'url'}
keep_only_tags = [
@ -29,6 +32,15 @@ class TheHindu(BasicNewsRecipe):
dict(id=lambda x: x and x.startswith('content-body-')),
]
def get_browser(self):
br = BasicNewsRecipe.get_browser(self)
br.addheaders += [('Referer', self.epaper_url)] # needed for fetching cover
return br
def get_cover_url(self):
url = self.index_to_soup(self.epaper_url + '/Login/DefaultImage', raw=True)
return url.replace(br'\\', b'/').decode('utf-8')[1:-1]
def preprocess_html(self, soup):
img = soup.find('img', attrs={'class': 'lead-img'})
try: