Update Bloomberg Businessweek

This commit is contained in:
unkn0w7n 2024-07-15 11:14:03 +05:30
parent fe79fce828
commit dfc2ca24f4
2 changed files with 106 additions and 153 deletions

View File

@ -1,196 +1,141 @@
import json
import random
import time
from collections import defaultdict
from datetime import datetime, timedelta
from calibre.web.feeds.news import BasicNewsRecipe, classes, prefixed_classes
from html5_parser import parse
from calibre.ebooks.BeautifulSoup import BeautifulSoup
from calibre.web.feeds.news import BasicNewsRecipe
# https://www.bloomberg.com/magazine/businessweek/24_12
# Set past_edition to edition id, which is '24_12'.
past_edition = None
def get_contents(x):
if x == '':
return ''
otype = x.get('type', '')
if otype == 'text':
if 'attributes' in x:
if 'strong' in x['attributes']:
return '<strong>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</strong>'
if 'emphasis' in x['attributes']:
return '<em>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</em>'
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
return x.get('value', '') + ''.join(map(get_contents, x.get('content', '')))
otype = x.get('role', '')
if otype == 'p':
return '<p>' + ''.join(map(get_contents, x.get('parts', ''))) + '</p>'
elif otype == 'text':
if 'style' in x:
return '<' + x['style'] + '>' + ''.join(map(get_contents, x.get('parts', ''))) + '</' + x['style'] + '>'
return x.get('text', '') + ''.join(map(get_contents, x.get('parts', '')))
elif otype == 'br':
return '<br>'
elif otype == 'paragraph':
return '<p>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</p>'
elif otype == 'heading':
return '<h3>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</h3>'
elif otype == 'list':
return '<ul>' + ''.join(map(get_contents, x.get('content', ''))) + '</ul>'
elif otype == 'listItem':
return '<li>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</li>'
elif otype == 'quote':
return '<blockquote class="col">' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</blockquote>'
elif otype == 'media':
if x['subType'] == 'photo':
return '<div><div class="img"><img src="{}"></div><div class="cap">{} <span>{}</span></div></div>'.format(
x['data']['photo']['src'], x['data']['photo']['caption'], x['data']['photo']['credit'])
elif x['subType'] == 'chart':
if x['data'] and x['data']['chart']:
return '<div class="img"><img src="{}"></div>'.format(x['data']['chart']['fallback'])
elif otype == 'link':
if 'data' in x:
if 'href' in x['data']:
return '<a href="' + x['data']['href'] + '">' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</a>'
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
elif otype == 'entity':
if x['subType'] == 'story':
if x['data'] and x['data']['link'] and x['data']['link']['destination']:
if 'web' in x['data']['link']['destination']:
return '<a href="' + x['data']['link']['destination']['web'] + '">' + x.get('value', '') + ''.join(
map(get_contents, x.get('content', ''))) + '</a>'
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
elif otype in {'div', 'callout'}:
return '<div>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</div>'
elif not any(x == otype for x in ['', 'ad', 'inline-newsletter', 'tabularData']):
if any(b in x for b in ['value', 'content']):
return '<i>' + x.get('value', '') + ''.join(map(get_contents, x.get('content', ''))) + '</i>'
elif otype == 'anchor':
return '<span>' + ''.join(map(get_contents, x.get('parts', ''))) + '</span>'
elif otype == 'h3':
return '<h4>' + ''.join(map(get_contents, x.get('parts', ''))) + '</h4>'
elif otype == 'ul':
return '<ul>' + ''.join(map(get_contents, x.get('parts', ''))) + '</ul>'
elif otype == 'li':
return '<li>' + ''.join(map(get_contents, x.get('parts', ''))) + '</li>'
elif otype == 'webview':
return '<br>' + x['html'] + ''.join(map(get_contents, x.get('parts', '')))
elif otype == 'blockquote':
return '<blockquote>' + ''.join(map(get_contents, x.get('parts', ''))) + '</blockquote>'
elif otype in {'image', 'video'}:
return '<br><img src="{}"><div class="img">{}</div>\n'.format(
x['imageURLs']['default'], x['caption'] + '<i> ' + x['credit'] + '</i>'
)
elif otype in {'correction', 'disclaimer'}:
return '<p class="corr">' + ''.join(map(get_contents, x.get('parts', ''))) + '</p>'
elif not any(x == otype for x in ['', 'ad', 'inline-newsletter', 'tabularData']):
return '<i>' + ''.join(map(get_contents, x.get('parts', ''))) + '</i>'
return ''
class Bloomberg(BasicNewsRecipe):
title = u'Bloomberg Businessweek'
title = 'Bloomberg Businessweek'
language = 'en'
__author__ = 'unkn0wn'
no_stylesheets = True
use_embedded_content = False
remove_attributes = ['style', 'height', 'width']
encoding = 'utf-8'
ignore_duplicate_articles = {'url'}
resolve_internal_links = True
masthead_url = 'https://assets.bwbx.io/s3/javelin/public/hub/images/BW-Logo-Black-cc9035fbb3.svg'
description = (
'Bloomberg Businessweek helps global leaders stay ahead with insights and in-depth analysis on the people,'
' companies, events, and trends shaping today\'s complex, global economy.'
)
simultaneous_downloads = 1
remove_empty_feeds = True
extra_css = '''
.auth {font-size:small; font-weight:bold;}
.time, .chart, .css--lede-byline, .css--lede-timestamp {font-size:small;}
.subhead, .cap span {font-style:italic; color:#404040;}
em, .col {color:#202020;}
.cat {font-size:small; color:gray;}
.news-figure-caption-text, .cap, .img, .css--caption-outer-wrapper {font-size:small; text-align:center;}
.auth { font-size:small; font-weight:bold; }
.subhead, .cap span { font-style:italic; color:#202020; }
em, blockquote { color:#202020; }
.cat { font-size:small; color:gray; }
.img, .news-figure-caption-text { font-size:small; text-align:center; }
.corr { font-size:small; font-style:italic; color:#404040; }
.chart { font-size:small; }
.news-figure-credit {font-size:small; text-align:center; color:#202020;}
'''
remove_tags = [
dict(name=['button', 'svg']),
dict(name='div', attrs={'id':['bb-that', 'bb-nav']}),
classes('twitter-logo bb-global-footer __sticky__audio__bar__portal__ css--social-wrapper-outer')
]
def get_browser(self, *a, **kw):
kw['user_agent'] = 'Mozilla/5.0 (Windows NT 10.0; rv:109.0) Gecko/20100101 Firefox/119.0'
br = BasicNewsRecipe.get_browser(self, *a, **kw)
br.set_handle_redirect(False)
return br
def parse_index(self):
soup = self.index_to_soup('https://www.bloomberg.com/businessweek')
bw = soup.find('a', href=lambda x: x and x.startswith('/magazine/businessweek/'))
edition = 'https://www.bloomberg.com' + bw['href']
self.log('Downloading ', edition)
self.cover_url = bw.find('img')['src'].replace('25x19', '600x800')
soup = self.index_to_soup(edition)
if timefmt := soup.find(**prefixed_classes('styles_TableOfContentsTitle__')):
self.timefmt = ' [' + (self.tag_to_string(timefmt).replace(' Issue', '')).strip() + ']'
inx = 'https://cdn-mobapi.bloomberg.com'
sec = self.index_to_soup(inx + '/wssmobile/v1/bw/news/list?limit=1', raw=True)
id = json.loads(sec)['magazines'][0]['id']
if past_edition:
id = past_edition
edit = self.index_to_soup(inx + '/wssmobile/v1/bw/news/week/' + id, raw=True)
d = json.loads(edit)
self.timefmt = ' [' + d['date'] + ']'
self.cover_url = d['image']['thumbUrl']
feeds_dict = defaultdict(list)
feeds = []
sec = ''
toc = soup.find('section', attrs={'id':'toc-archive-businessweek'})
for div in toc.findAll(**prefixed_classes('MagazinePageMagazineArchive_itemContainer__')):
h3 = div.find(**prefixed_classes('MagazinePageMagazineArchive_itemSection__'))
if h3 and h3.text:
sec = self.tag_to_string(h3)
self.log(sec)
a = div.find(**prefixed_classes('MagazinePageMagazineArchive_storyLink__'))
url = a['href']
if url.startswith('http') is False:
url = 'https://www.bloomberg.com' + a['href']
title = self.tag_to_string(a)
byl = div.find(**prefixed_classes('Byline_phoenix__'))
desc = self.tag_to_string(byl)
self.log('\t', title, '\n\t', desc, '\n\t\t', url)
feeds_dict[sec].append({"title": title, "url": url, "description": desc})
return [(sec, articles) for sec, articles in feeds_dict.items()]
for i in d['modules']:
section = i['title']
self.log(section)
articles = []
for x in i['articles']:
title = x['title']
url = inx + '/wssmobile/v1/stories/' + x['id']
self.log('\t', title)
articles.append({'title': title, 'url': url})
feeds.append((section, articles))
return feeds
def preprocess_raw_html(self, raw, *a):
root = parse(raw)
m = root.xpath('//script[@data-component-props="ArticleBody"]')
if not m:
m = root.xpath('//script[@data-component-props="FeatureBody"]')
if not m:
m2 = root.xpath('//script[@id="__NEXT_DATA__"]')
if not m2:
return raw
if m:
data = json.loads(m[0].text)
data = data['story']
data = json.loads(raw)
else:
data = json.loads(m2[0].text)
if 'story' in data['props']['pageProps']:
data = data['props']['pageProps']['story']
else:
return raw
title = '<h1>' + data['headline'] + '</h1>'
title = '<h1 title="{}">'.format(data['longURL']) + data['title'] + '</h1>'
cat = subhead = lede = auth = caption = ''
if 'primaryCategory' in data and data['primaryCategory'] is not None:
cat = '<p class="cat">' + data['primaryCategory'] + '</p>'
cat = '<div class="cat">' + data['primaryCategory'] + '</div>'
if len(data['abstract']) != 0 and len(data['abstract']) == 2:
subhead = '<div class="subhead"><p>' + data['abstract'][0] + '</p><p>' + data['abstract'][1] + '</p></div>'
else:
if 'summary' in data:
if 'abstract' in data and data['abstract'] and data['abstract'] is not None:
subhead = '<div class="subhead"><ul><li>' + '</li><li>'.join([x for x in data['abstract']]) + '</li></ul></div>'
elif 'summary' in data and data['summary']:
subhead = '<div class="subhead"><p>' + data['summary'] + '</p></div>'
if 'byline' in data and data['byline'] is not None:
auth = '<div><span class="auth">' + data['byline']\
+ '</span> | <span class="time">' + data['publishedAt'][:-14] + '</span></div>'
dt = datetime.fromtimestamp(data['updatedAt'] + time.timezone)
auth = '<p class="auth">' + 'By ' + data['byline'] + ' | Updated on ' + dt.strftime('%b %d, %Y at %I:%M %p') + '</p>'
if 'ledeImageUrl' in data and data['ledeImageUrl'] is not None:
lede = '<p id="img"><img src="{}">'.format(data['ledeImageUrl'])
if 'ledeImage' in data and data['ledeImage'] is not None:
x = data['ledeImage']
lede = '<br><img src="{}"><div class="img">{}</div>\n'.format(
x['imageURLs']['default'], x['caption'] + '<i> ' + x['credit'] + '</i>'
)
if 'ledeDescription' in data and data['ledeDescription'] is not None:
caption = '<span class="cap">' + data['ledeDescription'] + '</span>'
else:
if 'lede' in data and data['lede'] is not None:
if 'alt' in data['lede'] and data['lede']['alt'] is not None:
caption = '<span class="cap">' + data['lede']['alt'] + '</span>'
if m:
time.sleep(3)
body = data['body']
else:
body = ''
body_data = data['body']['content']
if data.get('type', '') == 'interactive':
body += '<p><em>' + 'This is an interactive article, which is supposed to be read in a browser.' + '</p></em>'
body_data = data['components']
for x in body_data:
body += get_contents(x)
pause = random.choice((5, 6, 7, 8, 9))
self.log('Delay: ', pause, ' seconds')
time.sleep(pause)
return '<html><body>' + cat + title + subhead + auth + lede + caption + '<div>' + body + '</div></body></html>'
html = '<html><body>' + cat + title + subhead + auth + lede + caption + '<div>' + body + '</div></body></html>'
return BeautifulSoup(html).prettify()
def preprocess_html(self, soup):
for h3 in soup.findAll('h3'):
h3.name = 'h4'
for icon in soup.findAll('img', attrs={'class':'video-player__play-icon'}):
icon.decompose()
for div in soup.findAll('div', attrs={'class':'chart'}):
@ -205,3 +150,9 @@ class Bloomberg(BasicNewsRecipe):
for img in soup.findAll('img', attrs={'src':lambda x: x and x.endswith(('-1x-1.jpg', '-1x-1.png'))}):
img['src'] = img['src'].replace('-1x-1', '750x-1')
return soup
def populate_article_metadata(self, article, soup, first):
article.url = soup.find('h1')['title']
article.summary = self.tag_to_string(soup.find('div', attrs={'class':'subhead'}))
article.text_summary = article.summary

View File

@ -19,7 +19,7 @@ def get_contents(x):
elif otype == 'br':
return '<br>'
elif otype == 'anchor':
return '<em>' + ''.join(map(get_contents, x.get('parts', ''))) + '</em>'
return '<span>' + ''.join(map(get_contents, x.get('parts', ''))) + '</span>'
elif otype == 'h3':
return '<h4>' + ''.join(map(get_contents, x.get('parts', ''))) + '</h4>'
elif otype == 'ul':
@ -30,11 +30,11 @@ def get_contents(x):
return '<br>' + x['html'] + ''.join(map(get_contents, x.get('parts', '')))
elif otype == 'blockquote':
return '<blockquote>' + ''.join(map(get_contents, x.get('parts', ''))) + '</blockquote>'
elif otype == ('image' or 'video'):
elif otype in {'image', 'video'}:
return '<br><img src="{}"><div class="img">{}</div>\n'.format(
x['imageURLs']['default'], x['caption'] + '<i> ' + x['credit'] + '</i>'
)
elif otype == ('correction' or 'disclaimer'):
elif otype in {'correction', 'disclaimer'}:
return '<p class="corr">' + ''.join(map(get_contents, x.get('parts', ''))) + '</p>'
elif not any(x == otype for x in ['', 'ad', 'inline-newsletter', 'tabularData']):
@ -44,7 +44,7 @@ def get_contents(x):
class Bloomberg(BasicNewsRecipe):
title = u'Bloomberg'
language = 'en'
language = 'en_US'
__author__ = 'unkn0wn'
no_stylesheets = True
remove_attributes = ['style', 'height', 'width']
@ -89,7 +89,7 @@ class Bloomberg(BasicNewsRecipe):
for arts in json.loads(art_soup)['modules']:
if arts['stories']:
for x in arts['stories']:
if x.get('type', '') == 'article':
if x.get('type', '') in {'article', 'interactive'}:
dt = datetime.fromtimestamp(x['published'] + time.timezone)
if (datetime.now() - dt) > timedelta(self.oldest_article):
continue
@ -109,7 +109,7 @@ class Bloomberg(BasicNewsRecipe):
cat = subhead = lede = auth = caption = ''
if 'primaryCategory' in data and data['primaryCategory'] is not None:
cat = '<p class="cat">' + data['primaryCategory'] + '</p>'
cat = '<div class="cat">' + data['primaryCategory'] + '</div>'
if 'abstract' in data and data['abstract'] and data['abstract'] is not None:
subhead = '<div class="subhead"><ul><li>' + '</li><li>'.join([x for x in data['abstract']]) + '</li></ul></div>'
@ -127,6 +127,8 @@ class Bloomberg(BasicNewsRecipe):
)
body = ''
if data.get('type', '') == 'interactive':
body += '<p><em>' + 'This is an interactive article, which is supposed to be read in a browser.' + '</p></em>'
body_data = data['components']
for x in body_data:
body += get_contents(x)