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 The Hindu

Browse Source
This commit is contained in:
Kovid Goyal 2023-03-17 08:45:03 +05:30
parent 48ad093c4a
commit fbec3adb2c
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 27 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
recipes/hindu.recipe
View File

@ -1,7 +1,7 @@
import json import json
import re import re
from collections import defaultdict from collections import defaultdict
from datetime import date from datetime import datetime
from calibre.web.feeds.news import BasicNewsRecipe, classes from calibre.web.feeds.news import BasicNewsRecipe, classes
@ -10,10 +10,11 @@ def absurl(url):
url = 'https://www.thehindu.com' + url url = 'https://www.thehindu.com' + url
return url return url
local_edition = None
# Chennai is default edition, for other editions use 'th_hyderabad', 'th_bangalore', 'th_delhi', 'th_kolkata' etc # Chennai is default edition, for other editions use 'th_hyderabad', 'th_bangalore', 'th_delhi', 'th_kolkata' etc
local_edition = None
# For past editions, set date to, for example, '2023-01-28'
past_edition = None
class TheHindu(BasicNewsRecipe): class TheHindu(BasicNewsRecipe):
title = 'The Hindu' title = 'The Hindu'
@ -22,15 +23,18 @@ class TheHindu(BasicNewsRecipe):
no_stylesheets = True no_stylesheets = True
masthead_url = 'https://www.thehindu.com/theme/images/th-online/thehindu-logo.svg' masthead_url = 'https://www.thehindu.com/theme/images/th-online/thehindu-logo.svg'
remove_attributes = ['style', 'height', 'width'] remove_attributes = ['style', 'height', 'width']
extra_css = '.caption{font-size:small; text-align:center;}'\
'.author{font-size:small; font-weight:bold;}'\ extra_css = '''
'.subhead, .subhead_lead {font-weight:bold;}'\ .caption {font-size:small; text-align:center;}
'img {display:block; margin:0 auto;}' .author {font-size:small; font-weight:bold;}
.subhead, .subhead_lead {font-weight:bold;}
img {display:block; margin:0 auto;}
'''
ignore_duplicate_articles = {'url'} ignore_duplicate_articles = {'url'}
keep_only_tags = [ keep_only_tags = [
classes('article-section ') classes('article-section')
] ]
remove_tags = [ remove_tags = [
@ -44,12 +48,22 @@ class TheHindu(BasicNewsRecipe):
img['src'] = img['data-original'] img['src'] = img['data-original']
return soup return soup
def __init__(self, *args, **kwargs):
BasicNewsRecipe.__init__(self, *args, **kwargs)
if self.output_profile.short_name.startswith('kindle'):
if not past_edition:
self.title = 'The Hindu ' + datetime.today().strftime('%b %d, %Y')
def parse_index(self): def parse_index(self):
if local_edition: global local_edition
yr = str(date.today().year) if local_edition or past_edition:
mn = date.today().strftime('%m') if local_edition is None:
dy = date.today().strftime('%d') local_edition = 'th_chennai'
url = 'https://www.thehindu.com/todays-paper/' + yr + '-' + mn + '-' + dy + '/' + local_edition + '/' today = datetime.today().strftime('%Y-%m-%d')
if past_edition:
today = past_edition
self.log('Downloading past edition of', local_edition + ' from ' + today)
url = absurl('/todays-paper/' + today + '/' + local_edition + '/')
else: else:
url = 'https://www.thehindu.com/todays-paper/' url = 'https://www.thehindu.com/todays-paper/'
raw = self.index_to_soup(url, raw=True) raw = self.index_to_soup(url, raw=True)