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 indian_express.recipe

Browse Source
This commit is contained in:
unkn0w7n 2024-10-29 12:10:43 +05:30
parent 68851263a4
commit a7925b7d2e
1 changed files with 11 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
recipes/indian_express.recipe
View File

@ -99,16 +99,22 @@ class IndianExpress(BasicNewsRecipe):
def articles_from_soup(self, soup):
ans = []
div = soup.find('div', attrs={'class':['nation', 'o-opin']})
for art in div.findAll(attrs={'class':['articles', 'o-opin-article']}):
div = soup.find('div', attrs={'class': ['nation', 'o-opin', 'myie-nation']})
for art in div.findAll(
attrs={'class': ['articles', 'o-opin-article', 'myie-articles']}
):
for a in art.findAll('a', href=True):
if not a.find('img') and not ('/profile/' in a['href'] or '/agency/' in a['href']):
if not a.find('img') and not any(
x in a['href'] for x in ['/profile/', '/agency/', '/section/']
):
url = a['href']
title = self.tag_to_string(a)
desc = ''
if p:= art.find('p'):
if p := art.find('p'):
desc = self.tag_to_string(p)
if da := art.find('div', attrs={'class':['date', 'o-opin-date']}):
if da := art.find(
'div', attrs={'class': ['date', 'o-opin-date', 'my-time']}
):
date = parse_date(self.tag_to_string(da)).replace(tzinfo=None)
today = datetime.now()
if (today - date) > timedelta(self.oldest_article):