mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-09-29 15:31:08 -04:00
374 lines
14 KiB
Python
374 lines
14 KiB
Python
#!/usr/bin/env python
|
|
# vim:fileencoding=utf-8
|
|
# License: GPLv3 Copyright: 2018, Kovid Goyal <kovid at kovidgoyal.net>
|
|
|
|
from __future__ import absolute_import, division, print_function, unicode_literals
|
|
|
|
import datetime
|
|
import json
|
|
import re
|
|
from pprint import pprint # noqa
|
|
|
|
from calibre import strftime
|
|
from calibre.ebooks.BeautifulSoup import Tag
|
|
from calibre.utils.date import strptime
|
|
from calibre.web.feeds.news import BasicNewsRecipe
|
|
|
|
use_wayback_machine = False
|
|
|
|
|
|
# The sections to download when downloading the web edition, comment out
|
|
# the section you are not interested in
|
|
web_sections = [
|
|
('World', 'world'),
|
|
('U.S.', 'us'),
|
|
('Politics', 'politics'),
|
|
('New York', 'nyregion'),
|
|
('Business', 'business'),
|
|
('Technology', 'technology'),
|
|
('Sports', 'sports'),
|
|
('Science', 'science'),
|
|
('Health', 'health'),
|
|
('Opinion', 'opinion'),
|
|
('Arts', 'arts'),
|
|
# ('Books', 'books'),
|
|
('Movies', 'movies'),
|
|
('Music', 'arts/music'),
|
|
('Television', 'arts/television'),
|
|
('Style', 'style'),
|
|
('Dining & Wine', 'dining'),
|
|
('Fashion & Style', 'fashion'),
|
|
('Home & Garden', 'garden'),
|
|
('Travel', 'travel'),
|
|
('Education', 'education'),
|
|
('Multimedia', 'multimedia'),
|
|
('Obituaries', 'obituaries'),
|
|
('Sunday Magazine', 'magazine')
|
|
]
|
|
url_date_pat = re.compile(r'/(2\d\d\d)/(\d\d)/(\d\d)/')
|
|
|
|
|
|
def date_from_url(url):
|
|
m = url_date_pat.search(url)
|
|
if m is not None:
|
|
return datetime.date(*map(int, m.groups()))
|
|
|
|
|
|
def format_date(d):
|
|
try:
|
|
return strftime(' [%a, %d %b %Y]', d)
|
|
except Exception:
|
|
return strftime(' [%Y/%m/%d]', d)
|
|
|
|
|
|
def classes(classes):
|
|
q = frozenset(classes.split(' '))
|
|
return dict(attrs={
|
|
'class': lambda x: x and frozenset(x.split()).intersection(q)})
|
|
|
|
|
|
def new_tag(soup, name, attrs=()):
|
|
impl = getattr(soup, 'new_tag', None)
|
|
if impl is not None:
|
|
return impl(name, attrs=dict(attrs))
|
|
return Tag(soup, name, attrs=attrs or None)
|
|
|
|
|
|
class NewYorkTimes(BasicNewsRecipe):
|
|
title = 'The New York Times'
|
|
description = (
|
|
'New York Times. Todays Paper '
|
|
'Use advanced menu to make changes to fetch Web Edition'
|
|
)
|
|
encoding = 'utf-8'
|
|
__author__ = 'Kovid Goyal'
|
|
language = 'en_US'
|
|
ignore_duplicate_articles = {'title', 'url'}
|
|
no_stylesheets = True
|
|
is_web_edition = False
|
|
oldest_web_edition_article = 7 # days
|
|
|
|
extra_css = """
|
|
.byl, .time { font-size:small; color:#202020; }
|
|
.cap { font-size:small; text-align:center; }
|
|
.cred { font-style:italic; font-size:small; }
|
|
em, blockquote { color: #202020; }
|
|
.sc { font-variant: small-caps; }
|
|
.lbl { font-size:small; color:#404040; }
|
|
img { display:block; margin:0 auto; }
|
|
"""
|
|
|
|
@property
|
|
def nyt_parser(self):
|
|
ans = getattr(self, '_nyt_parser', None)
|
|
if ans is None:
|
|
from calibre.live import load_module
|
|
self._nyt_parser = ans = load_module('calibre.web.site_parsers.nytimes')
|
|
return ans
|
|
|
|
def get_nyt_page(self, url, skip_wayback=False):
|
|
if use_wayback_machine and not skip_wayback:
|
|
from calibre import browser
|
|
return self.nyt_parser.download_url(url, browser())
|
|
return self.index_to_soup(url, raw=True)
|
|
|
|
def preprocess_raw_html(self, raw_html, url):
|
|
return self.nyt_parser.extract_html(self.index_to_soup(raw_html), url)
|
|
|
|
articles_are_obfuscated = use_wayback_machine
|
|
|
|
if use_wayback_machine:
|
|
def get_obfuscated_article(self, url):
|
|
from calibre.ptempfile import PersistentTemporaryFile
|
|
with PersistentTemporaryFile() as tf:
|
|
tf.write(self.get_nyt_page(url))
|
|
return tf.name
|
|
|
|
recipe_specific_options = {
|
|
'web': {
|
|
'short': 'Type in yes, if you want Web Edition',
|
|
'default': 'Todays Paper'
|
|
},
|
|
'days': {
|
|
'short': 'Oldest article to download from this news source. In days ',
|
|
'long': 'For example, 1, gives you articles from the past 24 hours\n(Works only for Web_Edition)',
|
|
'default': str(oldest_web_edition_article)
|
|
},
|
|
'date': {
|
|
'short': 'The date of the edition to download (YYYY/MM/DD format)\nUsed to fetch past editions of NYT newspaper',
|
|
'long': 'For example, 2024/07/16'
|
|
},
|
|
'res': {
|
|
'short': (
|
|
'For hi-res images, select a resolution from the following\noptions: '
|
|
'popup, jumbo, mobileMasterAt3x, superJumbo'
|
|
),
|
|
'long': (
|
|
'This is useful for non e-ink devices, and for a lower file size\nthan '
|
|
'the default, use mediumThreeByTwo440, mediumThreeByTwo225, articleInline.'
|
|
),
|
|
},
|
|
'comp': {
|
|
'short': 'Compress News Images?',
|
|
'long': 'enter yes',
|
|
'default': 'no'
|
|
}
|
|
}
|
|
|
|
def __init__(self, *args, **kwargs):
|
|
BasicNewsRecipe.__init__(self, *args, **kwargs)
|
|
c = self.recipe_specific_options.get('comp')
|
|
d = self.recipe_specific_options.get('days')
|
|
w = self.recipe_specific_options.get('web')
|
|
if w and isinstance(w, str):
|
|
if w == 'yes':
|
|
self.is_web_edition = True
|
|
if d and isinstance(d, str):
|
|
self.oldest_web_edition_article = float(d)
|
|
if c and isinstance(c, str):
|
|
if c.lower() == 'yes':
|
|
self.compress_news_images = True
|
|
|
|
def read_todays_paper(self):
|
|
INDEX = 'https://www.nytimes.com/section/todayspaper'
|
|
# INDEX = 'file:///t/raw.html'
|
|
d = self.recipe_specific_options.get('date')
|
|
if d and isinstance(d, str):
|
|
INDEX = 'https://www.nytimes.com/issue/todayspaper/' + d + '/todays-new-york-times'
|
|
return self.index_to_soup(self.get_nyt_page(INDEX, skip_wayback=True))
|
|
|
|
def read_nyt_metadata(self):
|
|
soup = self.read_todays_paper()
|
|
pdate = soup.find('meta', attrs={'name':'pdate', 'content': True})['content']
|
|
date = strptime(pdate, '%Y%m%d', assume_utc=False, as_utc=False)
|
|
self.cover_url = 'https://static01.nyt.com/images/{}/nytfrontpage/scan.jpg'.format(date.strftime('%Y/%m/%d'))
|
|
self.timefmt = strftime(' [%d %b, %Y]', date)
|
|
return soup
|
|
|
|
def parse_todays_page(self):
|
|
soup = self.read_nyt_metadata()
|
|
script = soup.findAll('script', text=lambda x: x and 'window.__preloadedData' in x)[0]
|
|
script = type(u'')(script)
|
|
json_data = script[script.find('{'):script.rfind(';')].strip().rstrip(';')
|
|
data = json.loads(json_data.replace(':undefined', ':null'))['initialState']
|
|
containers, sections = {}, {}
|
|
article_map = {}
|
|
gc_pat = re.compile(r'groupings.(\d+).containers.(\d+)')
|
|
pat = re.compile(r'groupings.(\d+).containers.(\d+).relations.(\d+)')
|
|
for key in data:
|
|
if 'Article' in key:
|
|
adata = data[key]
|
|
if adata.get('__typename') == 'Article':
|
|
url = adata.get('url')
|
|
summary = adata.get('summary')
|
|
headline = adata.get('headline')
|
|
if url and headline:
|
|
title = data[headline['id']]['default']
|
|
article_map[adata['id']] = {
|
|
'title': title, 'url': url, 'description': summary or ''}
|
|
elif 'Legacy' in key:
|
|
sdata = data[key]
|
|
tname = sdata.get('__typename')
|
|
if tname == 'LegacyCollectionContainer':
|
|
m = gc_pat.search(key)
|
|
containers[int(m.group(2))] = sdata['label'] or sdata['name']
|
|
elif tname == 'LegacyCollectionRelation':
|
|
m = pat.search(key)
|
|
grouping, container, relation = map(int, m.groups())
|
|
asset = sdata['asset']
|
|
if asset and asset['typename'] == 'Article' and grouping == 0:
|
|
if container not in sections:
|
|
sections[container] = []
|
|
sections[container].append(asset['id'].split(':', 1)[1])
|
|
|
|
feeds = []
|
|
for container_num in sorted(containers):
|
|
section_title = containers[container_num]
|
|
if container_num in sections:
|
|
articles = sections[container_num]
|
|
if articles:
|
|
feeds.append((section_title, []))
|
|
for artid in articles:
|
|
if artid in article_map:
|
|
art = article_map[artid]
|
|
feeds[-1][1].append(art)
|
|
|
|
def skey(x):
|
|
name = x[0].strip()
|
|
if name == 'The Front Page':
|
|
return 0, ''
|
|
return 1, name.lower()
|
|
feeds.sort(key=skey)
|
|
for section, articles in feeds:
|
|
self.log('\n' + section)
|
|
for article in articles:
|
|
self.log(article['title'] + ' - ' + article['url'])
|
|
# raise SystemExit(1)
|
|
return feeds
|
|
|
|
def parse_article_group(self, container):
|
|
for li in container.findAll('li'):
|
|
article = li.find('article')
|
|
if article is None:
|
|
a = li.find('a', href=True)
|
|
if a is not None:
|
|
title = self.tag_to_string(li.find(['h3', 'h2'])).strip()
|
|
paras = li.findAll('p')
|
|
if not title:
|
|
title = self.tag_to_string(paras[0]).strip()
|
|
if not title:
|
|
raise ValueError('No title found in article')
|
|
url = a['href']
|
|
if url.startswith('/'):
|
|
url = 'https://www.nytimes.com' + url
|
|
desc = ''
|
|
if len(paras) > 0:
|
|
desc = self.tag_to_string(paras[-1])
|
|
date = ''
|
|
d = date_from_url(url)
|
|
if d is not None:
|
|
date = format_date(d)
|
|
today = datetime.date.today()
|
|
delta = today - d
|
|
if delta.days > self.oldest_web_edition_article:
|
|
self.log.debug('\tSkipping article', title, 'as it is too old')
|
|
continue
|
|
yield {'title': title, 'url': url, 'description': desc, 'date': date}
|
|
continue
|
|
h2 = article.find(['h2', 'h3'])
|
|
if h2 is not None:
|
|
title = self.tag_to_string(h2)
|
|
a = h2.find('a', href=True)
|
|
if a is not None:
|
|
url = a['href']
|
|
if url.startswith('/'):
|
|
url = 'https://www.nytimes.com' + url
|
|
desc = ''
|
|
p = h2.findNextSibling('p')
|
|
if p is not None:
|
|
desc = self.tag_to_string(p)
|
|
date = ''
|
|
d = date_from_url(url)
|
|
if d is not None:
|
|
date = format_date(d)
|
|
today = datetime.date.today()
|
|
delta = today - d
|
|
if delta.days > self.oldest_web_edition_article:
|
|
self.log.debug('\tSkipping article', title, 'as it is too old')
|
|
continue
|
|
yield {'title': title, 'url': url, 'description': desc, 'date': date}
|
|
|
|
def parse_web_section(self, soup, slug):
|
|
|
|
def log(article):
|
|
self.log('\t', article['title'] + article['date'], ':', article['url'])
|
|
if article.get('description'):
|
|
self.log('\t\t', article['description'])
|
|
|
|
cid = slug.split('/')[-1]
|
|
if cid == 'dining':
|
|
cid = 'food'
|
|
try:
|
|
container = soup.find(id='collection-{}'.format(cid)).find('section')
|
|
except AttributeError:
|
|
container = None
|
|
if container is None:
|
|
raise ValueError('Failed to find articles container for slug: {}'.format(slug))
|
|
for ol in container.findAll('ol'):
|
|
for article in self.parse_article_group(ol):
|
|
log(article)
|
|
yield article
|
|
|
|
def parse_web_sections(self):
|
|
self.read_nyt_metadata()
|
|
feeds = []
|
|
for section_title, slug in web_sections:
|
|
url = 'https://www.nytimes.com/section/' + slug
|
|
try:
|
|
soup = self.index_to_soup(self.get_nyt_page(url))
|
|
except Exception:
|
|
self.log.error('Failed to download section:', url)
|
|
continue
|
|
self.log('Found section:', section_title)
|
|
articles = list(self.parse_web_section(soup, slug))
|
|
if articles:
|
|
feeds.append((section_title, articles))
|
|
if self.test and len(feeds) >= self.test[0]:
|
|
break
|
|
return feeds
|
|
|
|
def parse_index(self):
|
|
# return [('All articles', [
|
|
# {'title': 'XXXXX', 'url': 'https://www.nytimes.com/2020/11/27/world/americas/coronavirus-migrants-venezuela.html'},
|
|
# ])]
|
|
if self.is_web_edition:
|
|
return self.parse_web_sections()
|
|
return self.parse_todays_page()
|
|
|
|
def get_browser(self, *args, **kwargs):
|
|
kwargs['user_agent'] = 'Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)'
|
|
br = BasicNewsRecipe.get_browser(self, *args, **kwargs)
|
|
br.addheaders += [
|
|
('Referer', 'https://www.google.com/'),
|
|
('X-Forwarded-For', '66.249.66.1')
|
|
]
|
|
return br
|
|
|
|
def preprocess_html(self, soup):
|
|
w = self.recipe_specific_options.get('res')
|
|
if w and isinstance(w, str):
|
|
res = '-' + w
|
|
for img in soup.findAll('img', attrs={'src':True}):
|
|
if '-article' in img['src']:
|
|
ext = img['src'].split('?')[0].split('.')[-1]
|
|
img['src'] = img['src'].rsplit('-article', 1)[0] + res + '.' + ext
|
|
for c in soup.findAll('div', attrs={'class':'cap'}):
|
|
for p in c.findAll(['p', 'div']):
|
|
p.name = 'span'
|
|
return soup
|
|
|
|
def get_article_url(self, article):
|
|
url = BasicNewsRecipe.get_article_url(self, article)
|
|
if not re.search(r'/video/|/athletic/', url):
|
|
return url
|