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 Live Mint

Browse Source
This commit is contained in:
Kovid Goyal 2022-08-18 17:50:25 +05:30
parent 7fdf70f733
commit 81639f749a
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 20 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
recipes/livemint.recipe
View File

@ -1,8 +1,11 @@
#!/usr/bin/env python #!/usr/bin/env python
from calibre.web.feeds.news import BasicNewsRecipe, classes import json
import re
from datetime import date from datetime import date
from calibre.web.feeds.news import BasicNewsRecipe, classes
is_saturday = date.today().weekday() == 5 is_saturday = date.today().weekday() == 5
@ -46,10 +49,12 @@ class LiveMint(BasicNewsRecipe):
('Smart Living','https://lifestyle.livemint.com/rss/smart-living'), ('Smart Living','https://lifestyle.livemint.com/rss/smart-living'),
] ]
else: else:
# some wsj articles wont load
extra_css = ''' extra_css = '''
#img-cap {font-size:small; text-align:center;} #img-cap {font-size:small; text-align:center;}
#auth-info {font-size:small; text-align:center;} #auth-info {font-size:small; text-align:center;}
.highlights {font-style:italic;} .highlights {font-style:italic;}
.summary{font-style:italic; color:#404040;}
''' '''
cover_url = 'https://epsfs.hindustantimes.com/MINT/2022/04/05/Delhi/Delhi/5_01/1ec7ad14_01_mr.jpg' cover_url = 'https://epsfs.hindustantimes.com/MINT/2022/04/05/Delhi/Delhi/5_01/1ec7ad14_01_mr.jpg'
@ -85,6 +90,20 @@ class LiveMint(BasicNewsRecipe):
('Elections', 'https://www.livemint.com/rss/elections'), ('Elections', 'https://www.livemint.com/rss/elections'),
] ]
def preprocess_raw_html(self, raw, *a):
if '<script>var wsjFlag=true;</script>' in raw:
m = re.search(r'type="application/ld\+json">[^<]+?"@type": "NewsArticle"', raw)
raw1 = raw[m.start():]
raw1 = raw1.split('>', 1)[1].strip()
data = json.JSONDecoder().raw_decode(raw1)[0]
value = data['hasPart']['value']
body = data['articleBody'] + '</p> <p>' + re.sub(r'([a-z]\.|[0-9]\.)([A-Z])', r'\1 <p> \2', value)
body = '<div class="FirstEle"> <p>' + body + '</p> </div>'
raw = re.sub(r'<div class="FirstEle">([^}]*)</div>', body, raw)
return raw
else:
return raw
def preprocess_html(self, soup): def preprocess_html(self, soup):
for span in soup.findAll('figcaption'): for span in soup.findAll('figcaption'):
span['id'] = 'img-cap' span['id'] = 'img-cap'