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

...

Browse Source
This commit is contained in:
Kovid Goyal 2022-05-02 15:21:01 +05:30
parent 6476d29ab3
commit 6901b92b2e
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 14 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
recipes/financial_times.recipe
View File

@ -16,6 +16,7 @@ class ft(BasicNewsRecipe):
remove_empty_feeds = True
ignore_duplicate_articles = {'url'}
remove_attributes = ['style', 'width', 'height']
masthead_url = 'https://im.ft-static.com/m/img/masthead_main.jpg'
def get_cover_url(self):
soup = self.index_to_soup(
@ -35,7 +36,7 @@ class ft(BasicNewsRecipe):
('Climate', 'https://www.ft.com/climate-capital?format=rss'),
('Opinion', 'https://www.ft.com/opinion?format=rss'),
('Life & Arts', 'https://www.ft.com/life-arts?format=rss'),
('how to spend it', 'https://www.ft.com/htsi?format=rss'),
('How to spend it', 'https://www.ft.com/htsi?format=rss'),
]
def preprocess_raw_html(self, raw, *a):
@ -50,17 +51,25 @@ class ft(BasicNewsRecipe):
title = data['headline']
body = data['articleBody']
body = body.replace('\n\n', '<p>')
# remove embedded image links
body = re.sub(r'\[https://\S+?\]', '', body)
author = ''
if 'author' in data:
try:
author = data['author']['name']
except TypeError:
author = ' and '.join(x['name'] for x in data['author'])
image = desc = ''
image = desc = title_image_url = ''
if data.get('image'):
image = '<p><img src="{}">'.format(data['image']['url'])
title_image_url = data['image']['url']
image = '<p><img src="{}">'.format(title_image_url)
# embedded image links
def insert_image(m):
url = m.group()[1:-1]
if url == title_image_url:
return ''
return '<p><img src="{}">'.format(url)
body = re.sub(r'\[https://\S+?\]', insert_image, body)
if data.get('description'):
desc = '<h2>' + data['description'] + '</h2>'
html = '<html><body><h1>' + title + '</h1>' + desc + '<h3>' + author + '</h3>' + image + '<p>' + body