calibre/recipes/latimes.recipe
2020-10-17 21:42:16 +05:30

120 lines
4.3 KiB
Python

#!/usr/bin/env python
import re
from collections import defaultdict
from calibre.web.feeds.news import BasicNewsRecipe
DIR_COLLECTIONS = [['world'],
['nation'],
['politics'],
['opinion', 'op-ed', 'opinion-la', 'editorials',
'readersreact', 'topoftheticket', 'endorsements'],
['local', 'lanow', 'california', 'crime',
'abcarian', 'education', 'weather'],
['business', 'hollywood', 'technology'],
['sports'],
['entertainment', 'movies', 'music',
'tv', 'arts', 'gossip', 'envelope'],
['books'],
['food', 'jonathon-gold', 'dailydish'],
['health'],
['style', 'laaffairs', 'pets'],
['science', 'sciencenow'],
['home'],
['travel'],
['fashion']]
def classes(classes):
q = frozenset(classes.split(' '))
return dict(attrs={
'class': lambda x: x and frozenset(x.split()).intersection(q)})
def absurl(url):
if url.startswith('/'):
url = 'https://www.latimes.com' + url
return url
def what_section(url):
parts = url.split('/')
return parts[-4].capitalize()
class LATimes(BasicNewsRecipe):
title = 'Los Angeles Times'
__author__ = 'Jose Ortiz'
description = 'The Los Angeles Times is a leading source of news on Southern California, entertainment, movies, television, music, politics, business, health, technology, travel, sports, environment, economics, autos, jobs, real estate and other topics affecting California' # noqa
category = 'news, politics, USA, Los Angeles, world'
oldest_article = 1
max_articles_per_feed = 200
no_stylesheets = True
encoding = 'utf8'
use_embedded_content = False
compress_news_images = True
compress_news_images_auto_size = 5
language = 'en'
remove_empty_feeds = True
ignore_duplicate_articles = {'url'}
publication_type = 'newspaper'
cover_url = 'http://www.latimes.com/includes/sectionfronts/A1.pdf'
keep_only_tags = [
classes('headline page-lead-media authors published-date page-article-container'),
]
remove_tags= [
classes('google-dfp-ad-wrapper enhancement')
]
def parse_index(self):
index = 'https://www.latimes.com/'
pat = r'^https://www\.latimes\.com/[^/]+?/story/20\d{2}-\d{2}-\d{2}/\S+'
articles = self.find_articles(index, pat)
for collection in DIR_COLLECTIONS:
if self.test:
continue
topdir = collection.pop(0)
collection_index = index + topdir + '/'
articles += self.find_articles(collection_index, pat)
for subdir in collection:
sub_index = collection_index + subdir + '/'
articles += self.find_articles(sub_index, pat)
feeds = defaultdict(list)
for article in articles:
section = what_section(article['url'])
feeds[section].append(article)
return [(k, feeds[k]) for k in sorted(feeds)]
def preprocess_html(self, soup):
for img in soup.findAll('img', attrs={'data-src': True}):
if img.findParent('a', href='http://www.latimes.com/opinion/la-letter-to-the-editor-htmlstory.html') \
is img.parent and img['data-src'].endswith('/la-letter-to-the-editor'):
img.parent.extract()
else:
img['src'] = img['data-src']
return soup
def find_articles(self, index, pattern):
self.log('Downloading and parsing index: ', index)
self.log('Pattern: ', pattern)
try:
soup = self.index_to_soup(index)
except:
self.log('Failed to download ', index)
return []
if soup.main is not None:
alinks = soup.main.findAll('a', {'href': re.compile(pattern)})
else:
alinks = soup.findAll('a', {'href': re.compile(pattern)})
alinks = [a for a in alinks if len(
a.contents) == 1 and a.find(text=True, recursive=False)]
articles = [
{'title': self.tag_to_string(a), 'url': absurl(a['href'])} for a in alinks]
self.log('Found: ', len(articles), ' articles.\n')
return articles