| 1 |
#!/usr/bin/env python |
| 2 |
# vim:fileencoding=utf-8 |
| 3 |
# License: GPLv3 Copyright: 2018, Kovid Goyal <kovid at kovidgoyal.net> |
| 4 |
|
| 5 |
import datetime |
| 6 |
import json |
| 7 |
import re |
| 8 |
from pprint import pprint |
| 9 |
|
| 10 |
from calibre import strftime |
| 11 |
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag |
| 12 |
from calibre.utils.date import strptime |
| 13 |
from calibre.web.feeds.news import BasicNewsRecipe |
| 14 |
from polyglot.functools import lru_cache |
| 15 |
|
| 16 |
is_web_edition = False |
| 17 |
use_wayback_machine = False |
| 18 |
|
| 19 |
# This is an Apollo persisted query hash which you can get |
| 20 |
# from looking at the XHR requests made by: https://www.nytimes.com/section/todayspaper |
| 21 |
# or by https://www.nytimes.com/section/world |
| 22 |
persistedQuery = '1f99120a11e94dd62a9474f68ee1255537ee3cf7eac20a0377819edb2fa1fef7' |
| 23 |
|
| 24 |
# The sections to download when downloading the web edition, comment out |
| 25 |
# the section you are not interested in |
| 26 |
web_sections = [ |
| 27 |
'world', |
| 28 |
'us', |
| 29 |
'politics', |
| 30 |
'nyregion', |
| 31 |
'business', |
| 32 |
'technology', |
| 33 |
'sports', |
| 34 |
'science', |
| 35 |
'health', |
| 36 |
'opinion', |
| 37 |
'arts', |
| 38 |
'books', |
| 39 |
'movies', |
| 40 |
'arts/music', |
| 41 |
'arts/television', |
| 42 |
'style', |
| 43 |
'food', |
| 44 |
'fashion', |
| 45 |
'travel', |
| 46 |
'education', |
| 47 |
'multimedia', |
| 48 |
'obituaries', |
| 49 |
'magazine', |
| 50 |
] |
| 51 |
# web_sections = [ 'business' ] |
| 52 |
url_date_pat = re.compile(r'/(2\d\d\d)/(\d\d)/(\d\d)/') |
| 53 |
|
| 54 |
|
| 55 |
def date_from_url(url): |
| 56 |
m = url_date_pat.search(url) |
| 57 |
if m is not None: |
| 58 |
return datetime.date(*map(int, m.groups())) |
| 59 |
|
| 60 |
|
| 61 |
def format_date(d): |
| 62 |
try: |
| 63 |
return strftime(' [%a, %d %b %Y]', d) |
| 64 |
except Exception: |
| 65 |
return strftime(' [%Y/%m/%d]', d) |
| 66 |
|
| 67 |
|
| 68 |
def classes(classes): |
| 69 |
q = frozenset(classes.split(' ')) |
| 70 |
return dict(attrs={'class': lambda x: x and frozenset(x.split()).intersection(q)}) |
| 71 |
|
| 72 |
|
| 73 |
def new_tag(soup, name, attrs=()): |
| 74 |
impl = getattr(soup, 'new_tag', None) |
| 75 |
if impl is not None: |
| 76 |
return impl(name, attrs=dict(attrs)) |
| 77 |
return Tag(soup, name, attrs=attrs or None) |
| 78 |
|
| 79 |
|
| 80 |
def absolutize_href(href): |
| 81 |
if not href.startswith('http'): |
| 82 |
href = 'https://www.nytimes.com/' + href.lstrip('/') |
| 83 |
return href |
| 84 |
|
| 85 |
|
| 86 |
@lru_cache(2) |
| 87 |
def parser_module(): |
| 88 |
from calibre.live import load_module |
| 89 |
|
| 90 |
return load_module('calibre.web.site_parsers.nytimes') |
| 91 |
|
| 92 |
|
| 93 |
class NewYorkTimes(BasicNewsRecipe): |
| 94 |
if is_web_edition: |
| 95 |
title = 'The New York Times (Web)' |
| 96 |
description = ( |
| 97 |
'New York Times (Web). You can edit the recipe to remove sections you are not interested in. ' |
| 98 |
'Use advanced menu to make changes to fetch Todays Paper' |
| 99 |
) |
| 100 |
else: |
| 101 |
title = 'The New York Times' |
| 102 |
description = 'New York Times. Todays Paper Use advanced menu to make changes to fetch Web Edition' |
| 103 |
encoding = 'utf-8' |
| 104 |
__author__ = 'Kovid Goyal' |
| 105 |
language = 'en_US' |
| 106 |
ignore_duplicate_articles = {'title', 'url'} |
| 107 |
no_stylesheets = True |
| 108 |
oldest_web_edition_article = 7 # days |
| 109 |
browser_type = 'webengine' |
| 110 |
|
| 111 |
extra_css = ''' |
| 112 |
.byl, .time { font-size:small; color:#202020; } |
| 113 |
.cap { font-size:small; text-align:center; } |
| 114 |
.cred { font-style:italic; font-size:small; } |
| 115 |
em, blockquote { color: #202020; } |
| 116 |
.sc { font-variant: small-caps; } |
| 117 |
.lbl { font-size:small; color:#404040; } |
| 118 |
img { display:block; margin:0 auto; } |
| 119 |
''' |
| 120 |
|
| 121 |
@property |
| 122 |
def nyt_parser(self): |
| 123 |
return parser_module() |
| 124 |
|
| 125 |
def get_nyt_page(self, url, skip_wayback=False): |
| 126 |
if use_wayback_machine and not skip_wayback: |
| 127 |
from calibre import browser |
| 128 |
|
| 129 |
return self.nyt_parser.download_url(url, browser()) |
| 130 |
return self.index_to_soup(url, raw=True) |
| 131 |
|
| 132 |
def preprocess_raw_html(self, raw_html, url): |
| 133 |
return self.nyt_parser.extract_html(self.index_to_soup(raw_html), url) |
| 134 |
|
| 135 |
articles_are_obfuscated = use_wayback_machine |
| 136 |
|
| 137 |
if use_wayback_machine: |
| 138 |
|
| 139 |
def get_obfuscated_article(self, url): |
| 140 |
from calibre.ptempfile import PersistentTemporaryFile |
| 141 |
|
| 142 |
with PersistentTemporaryFile() as tf: |
| 143 |
tf.write(self.get_nyt_page(url)) |
| 144 |
return tf.name |
| 145 |
|
| 146 |
recipe_specific_options = { |
| 147 |
'web': { |
| 148 |
'short': 'Type in yes, if you want ' + ('Todays Paper' if is_web_edition else 'Web Edition'), |
| 149 |
'default': 'Web Edition' if is_web_edition else 'Todays Paper', |
| 150 |
}, |
| 151 |
'days': { |
| 152 |
'short': 'Oldest article to download from this news source. In days ', |
| 153 |
'long': 'For example, 1, gives you articles from the past 24 hours\n(Works only for Web_Edition)', |
| 154 |
'default': str(oldest_web_edition_article), |
| 155 |
}, |
| 156 |
'date': { |
| 157 |
'short': 'The date of the edition to download (YYYY/MM/DD format)\nUsed to fetch past editions of NYT newspaper', |
| 158 |
'long': 'For example, 2024/07/16', |
| 159 |
}, |
| 160 |
'res': { |
| 161 |
'short': ('For hi-res images, select a resolution from the following\noptions: popup, jumbo, mobileMasterAt3x, superJumbo'), |
| 162 |
'long': ( |
| 163 |
'This is useful for non e-ink devices, and for a lower file size\nthan ' |
| 164 |
'the default, use mediumThreeByTwo440, mediumThreeByTwo225, articleInline.' |
| 165 |
), |
| 166 |
}, |
| 167 |
'comp': {'short': 'Compress News Images?', 'long': 'enter yes', 'default': 'no'}, |
| 168 |
} |
| 169 |
|
| 170 |
def __init__(self, *args, **kwargs): |
| 171 |
BasicNewsRecipe.__init__(self, *args, **kwargs) |
| 172 |
c = self.recipe_specific_options.get('comp') |
| 173 |
d = self.recipe_specific_options.get('days') |
| 174 |
w = self.recipe_specific_options.get('web') |
| 175 |
self.is_web_edition = is_web_edition |
| 176 |
if w and isinstance(w, str): |
| 177 |
if w == 'yes': |
| 178 |
self.is_web_edition = not is_web_edition |
| 179 |
if d and isinstance(d, str): |
| 180 |
self.oldest_web_edition_article = float(d) |
| 181 |
if c and isinstance(c, str): |
| 182 |
if c.lower() == 'yes': |
| 183 |
self.compress_news_images = True |
| 184 |
|
| 185 |
def todays_paper_url(self): |
| 186 |
pdate = self.recipe_specific_options.get('date') |
| 187 |
if pdate and isinstance(pdate, str): |
| 188 |
return 'https://www.nytimes.com/issue/todayspaper/{}/todays-new-york-times'.format(pdate) |
| 189 |
return 'https://www.nytimes.com/section/todayspaper' |
| 190 |
|
| 191 |
def parse_todays_page(self): |
| 192 |
url = self.todays_paper_url() |
| 193 |
soup = self.index_to_soup(url) |
| 194 |
return parse_todays_page(soup) |
| 195 |
|
| 196 |
def parse_web_sections(self): |
| 197 |
feeds = [] |
| 198 |
for slug in web_sections: |
| 199 |
url = 'https://www.nytimes.com/section/' + slug |
| 200 |
self.log('Download section index:', url) |
| 201 |
soup = self.index_to_soup(url) |
| 202 |
# with open('/t/raw.html', 'w') as f: |
| 203 |
# f.write(str(soup)) |
| 204 |
section_title, articles = parse_web_section(soup) |
| 205 |
self.log('Section:', section_title) |
| 206 |
if articles: |
| 207 |
feeds.append((section_title, articles)) |
| 208 |
for a in articles: |
| 209 |
self.log('\t', a['title'], a['url']) |
| 210 |
else: |
| 211 |
self.log(' No articles found in section:', section_title) |
| 212 |
if self.test and len(feeds) >= self.test[0]: |
| 213 |
break |
| 214 |
return feeds |
| 215 |
|
| 216 |
def parse_index(self): |
| 217 |
# return [('All articles', [ |
| 218 |
# {'title': 'XXXXX', 'url': 'https://www.nytimes.com/2020/11/27/world/americas/coronavirus-migrants-venezuela.html'}, |
| 219 |
# ])] |
| 220 |
date, feeds = self.parse_todays_page() |
| 221 |
pdate = date.strftime('%Y/%m/%d') |
| 222 |
self.cover_url = 'https://static01.nyt.com/images/{}/nytfrontpage/scan.jpg'.format(pdate) |
| 223 |
self.timefmt = strftime(' [%d %b, %Y]', date) |
| 224 |
if self.is_web_edition: |
| 225 |
return self.parse_web_sections() |
| 226 |
for s, articles in feeds: |
| 227 |
self.log('Section:', s) |
| 228 |
for a in articles: |
| 229 |
self.log('\t', a['title'], a['url']) |
| 230 |
return feeds |
| 231 |
|
| 232 |
def get_browser(self, *args, **kwargs): |
| 233 |
# kwargs['user_agent'] = 'Mozilla/5.0 (compatible; archive.org_bot; Wayback Machine Live Record; +http://archive.org/details/archive.org_bot)' |
| 234 |
br = BasicNewsRecipe.get_browser(self, *args, **kwargs) |
| 235 |
return br |
| 236 |
|
| 237 |
def preprocess_html(self, soup): |
| 238 |
w = self.recipe_specific_options.get('res') |
| 239 |
if w and isinstance(w, str): |
| 240 |
res = '-' + w |
| 241 |
for img in soup.findAll('img', attrs={'src': True}): |
| 242 |
if '-article' in img['src']: |
| 243 |
ext = img['src'].split('?')[0].split('.')[-1] |
| 244 |
img['src'] = img['src'].rsplit('-article', 1)[0] + res + '.' + ext |
| 245 |
for c in soup.findAll('div', attrs={'class': 'cap'}): |
| 246 |
for p in c.findAll(['p', 'div']): |
| 247 |
p.name = 'span' |
| 248 |
return soup |
| 249 |
|
| 250 |
def get_article_url(self, article): |
| 251 |
url = BasicNewsRecipe.get_article_url(self, article) |
| 252 |
if not re.search(r'/video/|/athletic/|/card/', url): |
| 253 |
return url |
| 254 |
self.log('\tSkipping ', url) |
| 255 |
|
| 256 |
|
| 257 |
def preloaded_data(soup): |
| 258 |
candidates = soup.find_all('script', string=lambda x: x and 'window.__preloadedData' in x) |
| 259 |
script = candidates[0] |
| 260 |
script = str(script) |
| 261 |
raw = script[script.find('{') : script.rfind(';')].strip().rstrip(';') # } |
| 262 |
raw = parser_module().clean_js_json(raw) |
| 263 |
# with open('/t/raw.json', 'w') as f: |
| 264 |
# f.write(raw) |
| 265 |
return json.JSONDecoder(strict=False).raw_decode(raw)[0]['initialState'] |
| 266 |
|
| 267 |
|
| 268 |
def asset_to_article(asset): |
| 269 |
title = asset['headline']['default'] |
| 270 |
return {'title': title, 'url': asset['url'], 'description': asset['summary']} |
| 271 |
|
| 272 |
|
| 273 |
def parse_web_section(soup): |
| 274 |
data = preloaded_data(soup) |
| 275 |
article_map = {} |
| 276 |
for k, v in data.items(): |
| 277 |
if v['__typename'] == 'Article': |
| 278 |
article_map[k] = asset_to_article(v) |
| 279 |
articles = [] |
| 280 |
for k, v in data['ROOT_QUERY'].items(): |
| 281 |
if k.startswith('workOrLocation'): |
| 282 |
c = data[v['__ref']] |
| 283 |
section_title = c['name'] |
| 284 |
for k, v in c['collectionsPage'].items(): |
| 285 |
if k.startswith('stream'): |
| 286 |
for k, v in v.items(): |
| 287 |
if k.startswith('edges'): |
| 288 |
for q in v: |
| 289 |
r = q['node']['__ref'] |
| 290 |
if r.startswith('Article:'): |
| 291 |
articles.append(article_map[r]) |
| 292 |
if not articles: |
| 293 |
for c in c['collectionsPage']['embeddedCollections']: |
| 294 |
for e in c['stream']['edges']: |
| 295 |
for k, v in e.items(): |
| 296 |
if k.startswith('node'): |
| 297 |
articles.append(article_map[v['__ref']]) |
| 298 |
return section_title, articles |
| 299 |
|
| 300 |
|
| 301 |
def parse_todays_page(soup): |
| 302 |
m = soup.find('meta', attrs={'name': 'nyt-collection:uri'})['content'].split('/') |
| 303 |
pdate = strptime('{}/{}/{}'.format(*m[-4:-1]), '%Y/%m/%d', assume_utc=False, as_utc=False) |
| 304 |
article_map = {} |
| 305 |
data = preloaded_data(soup) |
| 306 |
for k, v in data.items(): |
| 307 |
if v['__typename'] == 'Article': |
| 308 |
article_map[k] = asset_to_article(v) |
| 309 |
feeds = [] |
| 310 |
for k, v in data['ROOT_QUERY'].items(): |
| 311 |
if k.startswith('workOrLocation'): |
| 312 |
for g in data[v['__ref']]['groupings']: |
| 313 |
for c in g['containers']: |
| 314 |
articles = [] |
| 315 |
for r in c['relations']: |
| 316 |
ref = r['asset']['__ref'] |
| 317 |
if ref in article_map: |
| 318 |
articles.append(article_map[ref]) |
| 319 |
if articles: |
| 320 |
feeds.append((c['label'], articles)) |
| 321 |
return pdate, feeds |
| 322 |
|
| 323 |
|
| 324 |
if __name__ == '__main__': |
| 325 |
import sys |
| 326 |
|
| 327 |
with open(sys.argv[-1]) as f: |
| 328 |
html = f.read() |
| 329 |
soup = BeautifulSoup(html) |
| 330 |
if is_web_edition: |
| 331 |
section_title, articles = parse_web_section(soup) |
| 332 |
print(section_title) |
| 333 |
pprint(articles) |
| 334 |
else: |
| 335 |
pdate, feeds = parse_todays_page(soup) |
| 336 |
print(pdate) |
| 337 |
pprint(feeds) |