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 Bloomberg Business Week

Browse Source
This commit is contained in:
Kovid Goyal 2022-10-30 17:41:55 +05:30
parent 73637baaa9
commit 4c08d56111
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 11 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
recipes/bloomberg-business-week.recipe
View File

@ -14,7 +14,7 @@ class Bloomberg(BasicNewsRecipe):
ignore_duplicate_articles = {'url'} ignore_duplicate_articles = {'url'}
resolve_internal_links = True resolve_internal_links = True
masthead_url = 'https://assets.bwbx.io/s3/javelin/public/hub/images/BW-Logo-Black-cc9035fbb3.svg' masthead_url = 'https://assets.bwbx.io/s3/javelin/public/hub/images/BW-Logo-Black-cc9035fbb3.svg'
delay = 1.5 delay = 2
extra_css = ''' extra_css = '''
#auth {font-size:small; font-weight:bold;} #auth {font-size:small; font-weight:bold;}
#time {font-size:small;} #time {font-size:small;}
@ -23,6 +23,11 @@ class Bloomberg(BasicNewsRecipe):
.news-figure-credit {font-size:small; text-align:center; color:#202020;} .news-figure-credit {font-size:small; text-align:center; color:#202020;}
''' '''
remove_tags = [
dict(name='div', attrs={'id':['bb-that', 'bb-nav']}),
classes('twitter-logo bb-global-footer')
]
def get_browser(self): def get_browser(self):
br = browser() br = browser()
br.set_handle_redirect(False) br.set_handle_redirect(False)
@ -47,7 +52,9 @@ class Bloomberg(BasicNewsRecipe):
articles = [] articles = []
for art in div.findAll('article'): for art in div.findAll('article'):
a = art.find('a', **classes('story-list-story__info__headline-link')) a = art.find('a', **classes('story-list-story__info__headline-link'))
url = 'https://www.bloomberg.com' + a['href'] url = a['href']
if url.startswith('http') is False:
url = 'https://www.bloomberg.com' + a['href']
title = self.tag_to_string(a) title = self.tag_to_string(a)
desc = '' desc = ''
sum = art.find(**classes('story-list-story__info__summary')) sum = art.find(**classes('story-list-story__info__summary'))
@ -66,7 +73,8 @@ class Bloomberg(BasicNewsRecipe):
m = re.search('data-component-props="ArticleBody">', raw) m = re.search('data-component-props="ArticleBody">', raw)
if not m: if not m:
m = re.search('data-component-props="FeatureBody">', raw) m = re.search('data-component-props="FeatureBody">', raw)
if not m:
return raw
raw = raw[m.start():] raw = raw[m.start():]
raw = raw.split('>', 1)[1] raw = raw.split('>', 1)[1]
data = json.JSONDecoder().raw_decode(raw)[0] data = json.JSONDecoder().raw_decode(raw)[0]