Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-31 14:33:54 -04:00
Code Issues Packages Projects Releases Wiki Activity

...

Browse Source
This commit is contained in:
unkn0wn 2025-07-24 11:38:54 +05:30 committed by GitHub
parent fc42e81b6d
commit d353c5afe6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 6 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
recipes/wash_post.recipe
View File

@ -93,7 +93,11 @@ class TheWashingtonPost(BasicNewsRecipe):
('Commanders', 'http://feeds.washingtonpost.com/rss/sports/redskins'), ('Commanders', 'http://feeds.washingtonpost.com/rss/sports/redskins'),
] ]
def preprocess_raw_html(self, raw, *a): def preprocess_raw_html(self, raw, url):
if '/interactive/' in url:
return ('<html><body><article><h1>' + root.xpath('//h1')[0].text + '</h1><em>'
'This article is supposed to be read in a browser.'
'</em></article></body></html>')
root = parse(raw) root = parse(raw)
m = root.xpath('//script[@id="__NEXT_DATA__"]') m = root.xpath('//script[@id="__NEXT_DATA__"]')
@ -102,10 +106,7 @@ class TheWashingtonPost(BasicNewsRecipe):
text = data.get('label', {}).get('basic', {}).get('text', '') text = data.get('label', {}).get('basic', {}).get('text', '')
label = f'<p class="time">{text}</p>' if text else '' label = f'<p class="time">{text}</p>' if text else ''
if data.get('headlines'):
title = '<h1>' + data['headlines']['basic'] + '</h1>' title = '<h1>' + data['headlines']['basic'] + '</h1>'
elif data.get('metadata'):
title = '<h1>' + data['metadata']['headlines']['basic'] + '</h1>'
subhead = '<p class="subt">' + data['description'].get('basic', '') + '</h3>' subhead = '<p class="subt">' + data['description'].get('basic', '') + '</h3>'
promo_img = '' promo_img = ''