calibre/recipes/nytimes.recipe
2015-01-23 19:08:21 +05:30

1208 lines
58 KiB
Python
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

#!/usr/bin/env python2
# -*- coding: utf-8 -*-
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
'''
nytimes.com
'''
import re, string, time
from calibre import strftime
from datetime import timedelta, date
from time import sleep
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, BeautifulStoneSoup
class NYTimes(BasicNewsRecipe):
recursions=1 # set this to zero to omit Related articles lists
match_regexps=[r'/[12][0-9][0-9][0-9]/[0-9]+/'] # speeds up processing by preventing index page links from being followed
# set getTechBlogs to True to include the technology blogs
# set tech_oldest_article to control article age
# set tech_max_articles_per_feed to control article count
getTechBlogs = True
remove_empty_feeds = True
tech_oldest_article = 14
tech_max_articles_per_feed = 25
# set getPopularArticles to False if you don't want the Most E-mailed and Most Viewed articles
# otherwise you will get up to 20 of the most popular e-mailed and viewed articles (in each category)
getPopularArticles = True
popularPeriod = '1' # set this to the number of days to include in the measurement
# e.g. 7 will get the most popular measured over the last 7 days
# and 30 will get the most popular measured over 30 days.
# you still only get up to 20 articles in each category
# set headlinesOnly to True for the headlines-only version. If True, webEdition is ignored.
headlinesOnly = True
# set webEdition to True for the Web edition of the newspaper. Set oldest_web_article to the
# number of days old an article can be for inclusion. If oldest_web_article = None all articles
# will be included. Note: oldest_web_article is ignored if webEdition = False
webEdition = False
oldest_web_article = None
# download higher resolution images than the small thumbnails typically included in the article
# the down side of having large beautiful images is the file size is much larger, on the order of 7MB per paper
useHighResImages = True
compress_news_images = True
compress_news_images_auto_size = 5
# replace paid Kindle Version: the name will be changed to "The New York Times" to cause
# previous paid versions of the new york times to best sent to the back issues folder on the kindle
replaceKindleVersion = False
# includeSections: List of sections to include. If empty, all sections found will be included.
# Otherwise, only the sections named will be included. For example,
#
# includeSections = ['Politics','Sports']
#
# would cause only the Politics and Sports sections to be included.
includeSections = [] # by default, all sections included
# excludeSections: List of sections to exclude. If empty, all sections found will be included.
# Otherwise, the sections named will be excluded. For example,
#
# excludeSections = ['Politics','Sports']
#
# would cause the Politics and Sports sections to be excluded. This parameter can be used
# in conjuction with includeSections although in most cases using one or the other, but
# not both, is sufficient.
excludeSections = []
# one_picture_per_article specifies that calibre should only use the first image
# from an article (if one exists). If one_picture_per_article = True, the image
# will be moved to a location between the headline and the byline.
# If one_picture_per_article = False, all images from the article will be included
# and shown in their original location.
one_picture_per_article = False
# The maximum number of articles that will be downloaded
max_articles_per_feed = 100
use_embedded_content = False
# Whether to omit duplicates of articles (typically arsing when articles are indexed in
# more than one section). If True, only the first occurance will be downloaded.
filterDuplicates = True
# Sections to collect for the Web edition.
# Delete any you don't want, or use includeSections or excludeSections
web_sections = [(u'World',u'world'),
(u'U.S.',u'national'),
(u'Politics',u'politics'),
(u'New York',u'nyregion'),
(u'Business','business'),
(u'Technology',u'technology'),
(u'Sports',u'sports'),
(u'Science',u'science'),
(u'Health',u'health'),
(u'Opinion',u'opinion'),
(u'Arts',u'arts'),
(u'Books',u'books'),
(u'Movies',u'movies'),
(u'Music',u'arts/music'),
(u'Television',u'arts/television'),
(u'Style',u'style'),
(u'Dining & Wine',u'dining'),
(u'Fashion & Style',u'fashion'),
(u'Home & Garden',u'garden'),
(u'Travel',u'travel'),
('Education',u'education'),
('Multimedia',u'multimedia'),
(u'Obituaries',u'obituaries'),
(u'Sunday Magazine',u'magazine')
]
tech_feeds = [
(u'Tech - News', u'http://pogue.blogs.nytimes.com/feed/'),
(u'Tech - Bits', u'http://bits.blogs.nytimes.com/feed/'),
(u'Tech - Gadgetwise', u'http://gadgetwise.blogs.nytimes.com/feed/'),
(u'Tech - Open', u'http://open.blogs.nytimes.com/feed/')
]
if headlinesOnly:
title='New York Times Headlines'
description = 'Headlines from the New York Times'
needs_subscription = False
elif webEdition:
title='New York Times (Web)'
description = 'New York Times on the Web'
needs_subscription = False
elif replaceKindleVersion:
title='The New York Times'
description = 'Today\'s New York Times'
needs_subscription = False
else:
title='New York Times'
description = 'Today\'s New York Times'
needs_subscription = False
def decode_url_date(self,url):
urlitems = url.split('/')
try:
d = date(int(urlitems[3]),int(urlitems[4]),int(urlitems[5]))
except:
try:
d = date(int(urlitems[4]),int(urlitems[5]),int(urlitems[6]))
except:
return None
return d
if oldest_web_article is None:
earliest_date = date.today()
else:
earliest_date = date.today() - timedelta(days=oldest_web_article)
oldest_article = 365 # by default, a long time ago
__author__ = 'GRiker/Kovid Goyal/Nick Redding'
language = 'en'
requires_version = (0, 7, 5)
encoding = 'utf-8'
timefmt = ''
# simultaneous_downloads = 1 # no longer required to deal with ads
cover_margins = (18,18,'grey99')
keep_only_tags = dict(id=['article', 'story', 'content'])
remove_tags = [
dict(attrs={'class':[
'articleFooter',
'articleTools',
'rfd', 'story-footer-links', 'page-footer',
'columnGroup singleRule',
'columnGroup last',
'columnGroup last',
'doubleRule',
'dottedLine',
'entry-meta',
'entry-response module',
'leftNavTabs',
'metaFootnote',
'inside-story',
'module box nav',
'nextArticleLink',
'nextArticleLink clearfix',
'post-tools',
'relatedSearchesModule',
'side_tool',
'singleAd',
'postCategory column',
'refer tagRefer', # added for bits blog post
'entry entry-utility', # added for DealBook
'entry-tags', # added for DealBook
'footer promos clearfix', # added for DealBook
'footer links clearfix', # added for DealBook
'tabsContainer', # added for other blog downloads
'column lastColumn', # added for other blog downloads
'pageHeaderWithLabel', # added for other gadgetwise downloads
'column two', # added for other blog downloads
'column two last', # added for other blog downloads
'column three', # added for other blog downloads
'column three last', # added for other blog downloads
'column four', # added for other blog downloads
'column four last', # added for other blog downloads
'column last', # added for other blog downloads
'entry entry-related',
'subNavigation tabContent active', # caucus blog navigation
'mediaOverlay slideshow',
'wideThumb',
'video', # added 02-11-2011
'videoHeader', # added 02-11-2011
'articleInlineVideoHolder', # added 02-11-2011
'assetCompanionAd',
'nytint-sectionHeader',
re.compile('^subNavigation'),
re.compile('^leaderboard'),
re.compile('^module'),
re.compile('commentCount'),
'lede-container',
'credit'
]}),
dict(attrs={'class':lambda x: x and 'related-coverage-marginalia' in x.split()}),
dict(attrs={'class':lambda x: x and 'interactive' in x.split()}),
dict(attrs={'class':lambda x: x and 'sharetools' in x.split()}),
dict(attrs={'class':lambda x: x and 'ad' in x.split()}),
dict(name='div', attrs={'class':re.compile('toolsList')}), # bits
dict(name='div', attrs={'class':re.compile('postNavigation')}), # bits
dict(name='div', attrs={'class':'tweet'}),
dict(name='span', attrs={'class':'commentCount meta'}),
dict(name='div', attrs={'id':'header'}),
dict(name='div', attrs={'id':re.compile('commentsContainer')}), # bits, pogue, gadgetwise, open
dict(name='ul', attrs={'class':re.compile('entry-tools')}), # pogue, gadgetwise
dict(name='div', attrs={'class':re.compile('nocontent')}), # pogue, gadgetwise
dict(name='div', attrs={'id':re.compile('respond')}), # open
dict(name='div', attrs={'class':re.compile('entry-tags')}), # pogue
dict(id=[
'adxLeaderboard',
'pagelinks',
'adxSponLink',
'anchoredAd_module',
'anchoredAd_spot',
'archive',
'articleExtras',
'articleInline',
'blog_sidebar',
'businessSearchBar',
'cCol',
'entertainmentSearchBar',
'footer',
'header',
'header_search',
'inlineBox',
'login',
'masthead',
'masthead-nav',
'memberTools',
'navigation', 'navigation-ghost', 'navigation-modal', 'navigation-edge',
'page-footer',
'portfolioInline',
'readerReviews',
'readerReviewsCount',
'relatedArticles',
'relatedTopics',
'respond',
'ribbon',
'side_search',
'side_index',
'side_tool',
'toolsRight',
'skybox', # added for DealBook
'TopAd', # added for DealBook
'related-content', # added for DealBook
'whats-next',
]),
dict(name=['script', 'noscript', 'style','form','hr', 'button', 'meta', 'footer'])]
no_stylesheets = True
extra_css = '''
.articleHeadline { text-align: left; margin-top:0.5em; margin-bottom:0.25em; }
.credit { font-weight: normal; text-align: right; font-size:
50%; line-height:1em; margin-top:5px; margin-left:0;
margin-right:0; margin-bottom: 0; }
.byline { text-align: left; font-size: 50%; line-height:1em; margin-top:10px; margin-left:0; margin-right:0; margin-bottom: 0; }
.dateline { text-align: left; font-size: 50%; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.kicker { font-size: 50%; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.timestamp { font-weight: normal; text-align: left; font-size: 50%; }
.caption { font-size: 50%; font-style:italic; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
a:link {text-decoration: none; }
.date{font-size: 50%; }
.update{font-size: 50%; }
.articleBody { }
.authorId {text-align: left; font-size: 50%; }
.image {text-align: center;}
.aside {color:blue;margin:0px 0px 0px 0px; padding: 0px 0px 0px 0px; font-size:100%;}
.asidenote {color:blue;margin:0px 0px 0px 0px; padding: 0px 0px 0px 0px; font-size:100%;font-weight:bold;}
.source {text-align: left; font-size: x-small; }'''
articles = {}
key = None
ans = []
url_list = []
def filter_ans(self, ans) :
total_article_count = 0
idx = 0
idx_max = len(ans)-1
while idx <= idx_max:
if self.includeSections != []:
if ans[idx][0] not in self.includeSections:
print "SECTION NOT INCLUDED: ",ans[idx][0]
del ans[idx]
idx_max = idx_max-1
continue
if ans[idx][0] in self.excludeSections:
print "SECTION EXCLUDED: ",ans[idx][0]
del ans[idx]
idx_max = idx_max-1
continue
if True: # self.verbose
self.log("Section %s: %d articles" % (ans[idx][0], len(ans[idx][1])))
for article in ans[idx][1]:
total_article_count += 1
if True: # self.verbose
self.log("\t%-40.40s... \t%-60.60s..." % (article['title'].encode('cp1252','replace'),
article['url'].encode('cp1252','replace')))
idx = idx+1
self.log("Queued %d articles" % total_article_count)
return ans
def exclude_url(self,url):
if not url.startswith("http"):
return True
if not url.endswith(".html") and 'dealbook.nytimes.com' not in url: # added for DealBook
return True
if 'nytimes.com' not in url:
return True
if 'podcast' in url:
return True
if '/video/' in url:
return True
if '/multimedia/' in url:
return True
if '/slideshow/' in url:
return True
if '/magazine/index' in url:
return True
if '/interactive/' in url:
return True
if '/reference/' in url:
return True
if '/premium/' in url:
return True
if '#comment' in url:
return True
if '#postComment' in url:
return True
if '#postcomment' in url:
return True
if re.search('/\d\d\d\d/\d\d/\d\d/',url) is None:
print("NO DATE IN "+url)
return True
return False
def fixChars(self,string):
# Replace lsquo (\x91)
fixed = re.sub("\x91","",string)
# Replace rsquo (\x92)
fixed = re.sub("\x92","",fixed)
# Replace ldquo (\x93)
fixed = re.sub("\x93","",fixed)
# Replace rdquo (\x94)
fixed = re.sub("\x94","",fixed)
# Replace ndash (\x96)
fixed = re.sub("\x96","",fixed)
# Replace mdash (\x97)
fixed = re.sub("\x97","",fixed)
return fixed
def get_browser(self):
br = BasicNewsRecipe.get_browser(self)
return br
cover_tag = 'NY_NYT'
def get_cover_url(self):
from datetime import timedelta, date
cover = 'http://webmedia.newseum.org/newseum-multimedia/dfp/jpg'+str(date.today().day)+'/lg/'+self.cover_tag+'.jpg'
br = BasicNewsRecipe.get_browser(self)
daysback=1
try:
br.open(cover)
except:
while daysback<7:
cover = 'http://webmedia.newseum.org/newseum-multimedia/dfp/jpg'+str((date.today() - timedelta(days=daysback)).day)+'/lg/'+self.cover_tag+'.jpg'
br = BasicNewsRecipe.get_browser(self)
try:
br.open(cover)
except:
daysback = daysback+1
continue
break
if daysback==7:
self.log("\nCover unavailable")
cover = None
return cover
masthead_url = 'http://graphics8.nytimes.com/images/misc/nytlogo379x64.gif'
def short_title(self):
return self.title
def article_to_soup(self, url_or_raw, raw=False):
from contextlib import closing
import copy
from calibre.ebooks.chardet import xml_to_unicode
print("ARTICLE_TO_SOUP "+url_or_raw)
if re.match(r'\w+://', url_or_raw):
br = self.clone_browser(self.browser)
open_func = getattr(br, 'open_novisit', br.open)
with closing(open_func(url_or_raw)) as f:
_raw = f.read()
if not _raw:
raise RuntimeError('Could not fetch index from %s'%url_or_raw)
else:
_raw = url_or_raw
if raw:
return _raw
if not isinstance(_raw, unicode) and self.encoding:
if callable(self.encoding):
_raw = self.encoding(_raw)
else:
_raw = _raw.decode(self.encoding, 'replace')
nmassage = copy.copy(BeautifulSoup.MARKUP_MASSAGE)
nmassage.extend(self.preprocess_regexps)
nmassage += [(re.compile(r'<!DOCTYPE .+?>', re.DOTALL), lambda m: '')]
# Some websites have buggy doctype declarations that mess up beautifulsoup
# Remove comments as they can leave detritus when extracting tags leaves
# multiple nested comments
nmassage.append((re.compile(r'<!--.*?-->', re.DOTALL), lambda m: ''))
usrc = xml_to_unicode(_raw, self.verbose, strip_encoding_pats=True)[0]
usrc = self.preprocess_raw_html(usrc, url_or_raw)
return BeautifulSoup(usrc, markupMassage=nmassage)
def massageNCXText(self, description):
# Kindle TOC descriptions won't render certain characters
if description:
massaged = unicode(BeautifulStoneSoup(description, convertEntities=BeautifulStoneSoup.HTML_ENTITIES))
# Replace '&' with '&'
massaged = re.sub("&#038;","&", massaged)
massaged = re.sub("&amp;","&", massaged)
return self.fixChars(massaged)
else:
return description
def feed_title(self,div):
return ''.join(div.findAll(text=True, recursive=True)).strip()
def handle_article(self,div):
thumbnail = div.find('div','thumbnail')
if thumbnail:
thumbnail.extract()
a = div.find('a', href=True)
if not a:
return
url = re.sub(r'\?.*', '', a['href'])
if self.exclude_url(url):
return
url += '?pagewanted=all'
if self.filterDuplicates:
if url in self.url_list:
return
if self.webEdition:
date_tag = self.decode_url_date(url)
if date_tag is not None:
if self.oldest_web_article is not None:
if date_tag < self.earliest_date:
self.log("Skipping article %s" % url)
return
else:
self.log("Skipping article %s" % url)
return
self.url_list.append(url)
title = self.tag_to_string(a, use_alt=True).strip()
description = ''
pubdate = strftime('%a, %d %b')
summary = div.find(True, attrs={'class':'summary'})
if summary:
description = self.tag_to_string(summary, use_alt=False)
author = ''
authorAttribution = div.find(True, attrs={'class':'byline'})
if authorAttribution:
author = self.tag_to_string(authorAttribution, use_alt=False)
else:
authorAttribution = div.find(True, attrs={'class':'byline'})
if authorAttribution:
author = self.tag_to_string(authorAttribution, use_alt=False)
feed = self.key if self.key is not None else 'Uncategorized'
if feed not in self.articles:
self.ans.append(feed)
self.articles[feed] = []
self.articles[feed].append(
dict(title=title, url=url, date=pubdate,
description=description, author=author,
content=''))
def get_popular_articles(self,ans):
if self.getPopularArticles:
popular_articles = {}
key_list = []
def handleh3(h3tag):
try:
url = h3tag.a['href']
except:
return ('','','','')
url = re.sub(r'\?.*', '', url)
if self.exclude_url(url):
return ('','','','')
url += '?pagewanted=all'
title = self.tag_to_string(h3tag.a,False)
h6tag = h3tag.findNextSibling('h6')
if h6tag is not None:
author = self.tag_to_string(h6tag,False)
else:
author = ''
ptag = h3tag.findNextSibling('p')
if ptag is not None:
desc = self.tag_to_string(ptag,False)
else:
desc = ''
return(title,url,author,desc)
have_emailed = False
emailed_soup = self.index_to_soup('http://www.nytimes.com/most-popular-emailed?period='+self.popularPeriod)
for h3tag in emailed_soup.findAll('h3'):
(title,url,author,desc) = handleh3(h3tag)
if url=='':
continue
if not have_emailed:
key_list.append('Most E-Mailed')
popular_articles['Most E-Mailed'] = []
have_emailed = True
popular_articles['Most E-Mailed'].append(
dict(title=title, url=url, date=strftime('%a, %d %b'),
description=desc, author=author,
content=''))
have_viewed = False
viewed_soup = self.index_to_soup('http://www.nytimes.com/most-popular-viewed?period='+self.popularPeriod)
for h3tag in viewed_soup.findAll('h3'):
(title,url,author,desc) = handleh3(h3tag)
if url=='':
continue
if not have_viewed:
key_list.append('Most Viewed')
popular_articles['Most Viewed'] = []
have_viewed = True
popular_articles['Most Viewed'].append(
dict(title=title, url=url, date=strftime('%a, %d %b'),
description=desc, author=author,
content=''))
viewed_ans = [(k, popular_articles[k]) for k in key_list if k in popular_articles]
for x in viewed_ans:
ans.append(x)
return ans
def get_tech_feeds(self,ans):
if self.getTechBlogs:
tech_articles = {}
key_list = []
save_oldest_article = self.oldest_article
save_max_articles_per_feed = self.max_articles_per_feed
self.oldest_article = self.tech_oldest_article
self.max_articles_per_feed = self.tech_max_articles_per_feed
self.feeds = self.tech_feeds
tech = self.parse_feeds()
self.oldest_article = save_oldest_article
self.max_articles_per_feed = save_max_articles_per_feed
self.feeds = None
for f in tech:
key_list.append(f.title)
tech_articles[f.title] = []
for a in f.articles:
tech_articles[f.title].append(
dict(title=a.title, url=a.url.partition('?')[0], date=a.date,
description=a.summary, author=a.author,
content=a.content))
tech_ans = [(k, tech_articles[k]) for k in key_list if k in tech_articles]
for x in tech_ans:
ans.append(x)
return ans
def parse_web_edition(self):
for (sec_title,index_url) in self.web_sections:
if self.includeSections != []:
if sec_title not in self.includeSections:
print "SECTION NOT INCLUDED: ",sec_title
continue
if sec_title in self.excludeSections:
print "SECTION EXCLUDED: ",sec_title
continue
try:
soup = self.index_to_soup('http://www.nytimes.com/pages/'+index_url+'/index.html')
except:
continue
print 'Index URL: '+'http://www.nytimes.com/pages/'+index_url+'/index.html'
self.key = sec_title
# Find each article
for div in soup.findAll(True,
attrs={'class':['section-headline', 'ledeStory', 'story', 'story headline','sectionHeader','headlinesOnly multiline flush']}):
if div['class'] in ['story', 'story headline', 'storyHeader'] :
self.handle_article(div)
elif div['class'] == 'ledeStory':
divsub = div.find('div','storyHeader')
if divsub is not None:
self.handle_article(divsub)
ulrefer = div.find('ul','refer')
if ulrefer is not None:
for lidiv in ulrefer.findAll('li'):
self.handle_article(lidiv)
elif div['class'] == 'headlinesOnly multiline flush':
for lidiv in div.findAll('li'):
self.handle_article(lidiv)
self.ans = [(k, self.articles[k]) for k in self.ans if k in self.articles]
return self.filter_ans(self.get_tech_feeds(self.get_popular_articles(self.ans)))
def parse_todays_index(self):
soup = self.index_to_soup('http://www.nytimes.com/pages/todayspaper/index.html')
skipping = False
# Find each article
for div in soup.findAll(True,
attrs={'class':['section-headline', 'story', 'story headline','sectionHeader','headlinesOnly multiline flush']}):
if div['class'] in ['section-headline','sectionHeader']:
self.key = string.capwords(self.feed_title(div))
self.key = self.key.replace('Op-ed','Op-Ed')
self.key = self.key.replace('U.s.','U.S.')
self.key = self.key.replace('N.y.','N.Y.')
skipping = False
if self.includeSections != []:
if self.key not in self.includeSections:
print "SECTION NOT INCLUDED: ",self.key
skipping = True
if self.key in self.excludeSections:
print "SECTION EXCLUDED: ",self.key
skipping = True
elif div['class'] in ['story', 'story headline'] :
if not skipping:
self.handle_article(div)
elif div['class'] == 'headlinesOnly multiline flush':
for lidiv in div.findAll('li'):
if not skipping:
self.handle_article(lidiv)
self.ans = [(k, self.articles[k]) for k in self.ans if k in self.articles]
return self.filter_ans(self.get_tech_feeds(self.get_popular_articles(self.ans)))
def parse_headline_index(self):
soup = self.index_to_soup('http://www.nytimes.com/pages/todaysheadlines/')
section_name='Unknown Section'
pubdate = strftime('%a, %d %b')
for td_col in soup.findAll('td'):
h6_sec_name = td_col.find('h6')
if h6_sec_name is not None:
new_section_name = self.tag_to_string(h6_sec_name,use_alt=False)
new_section_name = re.sub(r'^ *$','',new_section_name)
if new_section_name == '':
continue
section_name = new_section_name
continue
atag = td_col.find('a')
if atag is not None:
h4tag = None
for h4tag in atag.findNextSiblings('h4'):
break
if h4tag is None:
continue
author = self.tag_to_string(h4tag,use_alt=False)
try:
url = re.sub(r'\?.*', '', atag['href'])
except:
continue
if self.exclude_url(url):
continue
if '?' in url:
url += '&pagewanted=all'
else:
url += '?pagewanted=all'
if self.filterDuplicates:
if url in self.url_list:
continue
self.url_list.append(url)
title = self.tag_to_string(atag, use_alt=False).strip()
desc = atag.parent.find('p')
if desc is not None:
description = self.tag_to_string(desc,use_alt=False)
else:
description = ''
if section_name not in self.articles:
self.ans.append(section_name)
self.articles[section_name] = []
print('Title '+title+' author '+author)
self.articles[section_name].append(dict(title=title, url=url, date=pubdate, description=description, author=author, content=''))
self.ans = [(k, self.articles[k]) for k in self.ans if k in self.articles]
return self.filter_ans(self.ans)
def parse_index(self):
if self.headlinesOnly:
return self.parse_headline_index()
elif self.webEdition:
return self.parse_web_edition()
else:
return self.parse_todays_index()
def strip_anchors(self,soup,kill_all=False):
paras = soup.findAll(True)
for para in paras:
aTags = para.findAll('a')
for a in aTags:
if a.img is None:
if kill_all or (self.recursions==0):
a.replaceWith(self.tag_to_string(a,False))
else:
if 'href' in a:
if a['href'].startswith('http://www.nytimes'):
if not a['href'].endswith('pagewanted=all'):
url = re.sub(r'\?.*', '', a['href'])
if self.exclude_url(url):
a.replaceWith(self.tag_to_string(a,False))
else:
a['href'] = url+'?pagewanted=all'
elif not (a['href'].startswith('http://pogue') or
a['href'].startswith('http://bits') or
a['href'].startswith('http://travel') or
a['href'].startswith('http://business') or
a['href'].startswith('http://tech') or
a['href'].startswith('http://health') or
a['href'].startswith('http://dealbook') or
a['href'].startswith('http://open')):
a.replaceWith(self.tag_to_string(a,False))
return soup
def handle_tags(self,soup):
try:
print("HANDLE TAGS: TITLE = "+self.tag_to_string(soup.title))
except:
print("HANDLE TAGS: NO TITLE")
if soup is None:
print("ERROR: handle_tags received NoneType")
return None
## print("HANDLING AD FORWARD:")
# print(soup)
if self.keep_only_tags:
body = Tag(soup, 'body')
try:
if isinstance(self.keep_only_tags, dict):
self.keep_only_tags = [self.keep_only_tags]
for spec in self.keep_only_tags:
for tag in soup.find('body').findAll(**spec):
body.insert(len(body.contents), tag)
soup.find('body').replaceWith(body)
except AttributeError: # soup has no body element
pass
def remove_beyond(tag, next):
while tag is not None and getattr(tag, 'name', None) != 'body':
after = getattr(tag, next)
while after is not None:
ns = getattr(tag, next)
after.extract()
after = ns
tag = tag.parent
if self.remove_tags_after is not None:
rt = [self.remove_tags_after] if isinstance(self.remove_tags_after, dict) else self.remove_tags_after
for spec in rt:
tag = soup.find(**spec)
remove_beyond(tag, 'nextSibling')
if self.remove_tags_before is not None:
tag = soup.find(**self.remove_tags_before)
remove_beyond(tag, 'previousSibling')
for kwds in self.remove_tags:
for tag in soup.findAll(**kwds):
tag.extract()
return soup
def preprocess_html(self, soup):
#print(strftime("%H:%M:%S")+" -- PREPROCESS TITLE="+self.tag_to_string(soup.title))
skip_tag = soup.find(True, {'name':'skip'})
if skip_tag is not None:
#url = 'http://www.nytimes.com' + re.sub(r'\?.*', '', skip_tag.parent['href'])
url = 'http://www.nytimes.com' + skip_tag.parent['href']
#url += '?pagewanted=all'
self.log.warn("Skipping ad to article at '%s'" % url)
sleep(5)
soup = self.handle_tags(self.article_to_soup(url))
# check if the article is from one of the tech blogs
blog=soup.find('div',attrs={'id':['pogue','bits','gadgetwise','open']})
if blog is not None:
old_body = soup.find('body')
new_body=Tag(soup,'body')
new_body.append(soup.find('div',attrs={'id':'content'}))
new_body.find('div',attrs={'id':'content'})['id']='blogcontent' # identify for postprocess_html
old_body.replaceWith(new_body)
for divr in soup.findAll('div',attrs={'class':re.compile('w190 right')}):
if divr.find(text=re.compile('Sign up')):
divr.extract()
divr = soup.find('div',attrs={'class':re.compile('^relatedArticlesModule')})
if divr is not None:
print("PROCESSING RELATED: "+self.tag_to_string(soup.title,False))
# handle related articles
rlist = []
ul = divr.find('ul')
if ul is not None:
for li in ul.findAll('li'):
atag = li.find('a')
if atag is not None:
if atag['href'].startswith('http://pogue') or atag['href'].startswith('http://bits') or \
atag['href'].startswith('http://open'):
atag.find(text=True).replaceWith(self.massageNCXText(self.tag_to_string(atag,False)))
rlist.append(atag)
divr.extract()
if rlist != []:
asidediv = Tag(soup,'div',[('class','aside')])
if soup.find('hr') is None:
asidediv.append(Tag(soup,'hr'))
h4 = Tag(soup,'h4',[('class','asidenote')])
h4.insert(0,"Related Posts")
asidediv.append(h4)
ul = Tag(soup,'ul')
for r in rlist:
li = Tag(soup,'li',[('class','aside')])
r['class'] = 'aside'
li.append(r)
ul.append(li)
asidediv.append(ul)
asidediv.append(Tag(soup,'hr'))
smain = soup.find('body')
smain.append(asidediv)
else:
print("CANNOT FIND RELATED: "+self.tag_to_string(soup.title,False))
for atag in soup.findAll('a'):
img = atag.find('img')
if img is not None:
atag.replaceWith(img)
elif 'href' not in atag:
atag.replaceWith(atag.renderContents().decode('cp1252','replace'))
elif not (atag['href'].startswith('http://www.nytimes') or atag['href'].startswith('http://pogue') or
atag['href'].startswith('http://bits') or atag['href'].startswith('http://open')):
atag.replaceWith(atag.renderContents().decode('cp1252','replace'))
hdr = soup.find('address')
if hdr is not None:
hdr.name='span'
for span_credit in soup.findAll('span','credit'):
sp = Tag(soup,'span')
span_credit.replaceWith(sp)
sp.append(Tag(soup,'br'))
sp.append(span_credit)
sp.append(Tag(soup,'br'))
else: # nytimes article
related = [] # these will be the related articles
first_outer = None # first related outer tag
first_related = None # first related tag
for outerdiv in soup.findAll(attrs={'class': re.compile('articleInline runaroundLeft')}):
for rdiv in soup.findAll('div','columnGroup doubleRule'):
if rdiv.find('h3') is not None:
if self.tag_to_string(rdiv.h3,False).startswith('Related'):
rdiv.h3.find(text=True).replaceWith("Related articles")
rdiv.h3['class'] = 'asidenote'
for litag in rdiv.findAll('li'):
if litag.find('a') is not None:
if litag.find('a')['href'].startswith('http://www.nytimes.com'):
url = re.sub(r'\?.*', '', litag.find('a')['href'])
litag.find('a')['href'] = url+'?pagewanted=all'
litag.extract()
related.append(litag)
if first_related is None:
first_related = rdiv
first_outer = outerdiv
else:
litag.extract()
for h6tag in rdiv.findAll('h6'):
if h6tag.find('a') is not None:
if h6tag.find('a')['href'].startswith('http://www.nytimes.com'):
url = re.sub(r'\?.*', '', h6tag.find('a')['href'])
h6tag.find('a')['href'] = url+'?pagewanted=all'
h6tag.extract()
related.append(h6tag)
if first_related is None:
first_related = rdiv
first_outer = outerdiv
else:
h6tag.extract()
if related != []:
for r in related:
if r.h6: # don't want the anchor inside a h6 tag
r.h6.replaceWith(r.h6.a)
first_related.ul.append(r)
first_related.insert(0,Tag(soup,'hr'))
first_related.append(Tag(soup,'hr'))
first_related['class'] = 'aside'
first_outer.replaceWith(first_related) # replace the outer tag with the related tag
for rdiv in soup.findAll(attrs={'class': re.compile('articleInline runaroundLeft')}):
rdiv.extract()
kicker_tag = soup.find(attrs={'class':'kicker'})
if kicker_tag: # remove Op_Ed author head shots
tagline = self.tag_to_string(kicker_tag)
if tagline=='Op-Ed Columnist':
img_div = soup.find('div','inlineImage module')
if img_div:
img_div.extract()
if self.useHighResImages:
try:
# open up all the "Enlarge this Image" pop-ups and download the full resolution jpegs
enlargeThisList = soup.findAll('div',{'class':'icon enlargeThis'})
if enlargeThisList:
for popupref in enlargeThisList:
popupreflink = popupref.find('a')
if popupreflink:
reflinkstring = str(popupreflink['href'])
refstart = reflinkstring.find("javascript:pop_me_up2('") + len("javascript:pop_me_up2('")
refend = reflinkstring.find(".html", refstart) + len(".html")
reflinkstring = reflinkstring[refstart:refend]
popuppage = self.browser.open(reflinkstring)
popuphtml = popuppage.read()
popuppage.close()
if popuphtml:
st = time.localtime()
year = str(st.tm_year)
month = "%.2d" % st.tm_mon
day = "%.2d" % st.tm_mday
imgstartpos = popuphtml.find('http://graphics8.nytimes.com/images/' + year + '/' + month +'/' + day +'/') + \
len('http://graphics8.nytimes.com/images/' + year + '/' + month +'/' + day +'/')
highResImageLink = 'http://graphics8.nytimes.com/images/' + year + '/' + \
month +'/' + day +'/' + popuphtml[imgstartpos:popuphtml.find('.jpg',imgstartpos)+4]
popupSoup = BeautifulSoup(popuphtml)
highResTag = popupSoup.find('img', {'src':highResImageLink})
if highResTag:
try:
newWidth = highResTag['width']
newHeight = highResTag['height']
imageTag = popupref.parent.find("img")
except:
self.log("Error: finding width and height of img")
popupref.extract()
if imageTag:
try:
imageTag['src'] = highResImageLink
imageTag['width'] = newWidth
imageTag['height'] = newHeight
except:
self.log("Error setting the src width and height parameters")
except Exception:
self.log("Error pulling high resolution images")
try:
# in case pulling images failed, delete the enlarge this text
enlargeThisList = soup.findAll('div',{'class':'icon enlargeThis'})
if enlargeThisList:
for popupref in enlargeThisList:
popupref.extract()
except:
self.log("Error removing Enlarge this text")
return self.strip_anchors(soup,False)
def postprocess_html(self,soup,first_fetch):
if not first_fetch: # remove Related links
for aside in soup.findAll('div','aside'):
aside.extract()
soup = self.strip_anchors(soup,True)
#print("RECURSIVE: "+self.tag_to_string(soup.title))
if soup.find('div',attrs={'id':'blogcontent'}) is None:
if first_fetch:
aside = soup.find('div','aside')
if aside is not None: # move the related list to the end of the article
art = soup.find('div',attrs={'id':'article'})
if art is None:
art = soup.find('div',attrs={'class':'article'})
if art is not None:
art.append(aside)
try:
if self.one_picture_per_article:
# Remove all images after first
largeImg = soup.find(True, {'class':'articleSpanImage'})
inlineImgs = soup.findAll(True, {'class':'inlineImage module'})
if largeImg:
for inlineImg in inlineImgs:
inlineImg.extract()
else:
if inlineImgs:
firstImg = inlineImgs[0]
for inlineImg in inlineImgs[1:]:
inlineImg.extract()
# Move firstImg before article body
cgFirst = soup.find(True, {'class':re.compile('columnGroup *first')})
if cgFirst:
# Strip all sibling NavigableStrings: noise
navstrings = cgFirst.findAll(text=True, recursive=False)
[ns.extract() for ns in navstrings]
headline_found = False
tag = cgFirst.find(True)
insertLoc = 0
while True:
insertLoc += 1
if hasattr(tag,'class') and tag['class'] == 'articleHeadline':
headline_found = True
break
tag = tag.nextSibling
if not tag:
headline_found = False
break
if headline_found:
cgFirst.insert(insertLoc,firstImg)
else:
self.log(">>> No class:'columnGroup first' found <<<")
except:
self.log("ERROR: One picture per article in postprocess_html")
try:
# Change captions to italic
for caption in soup.findAll(True, {'class':'caption'}) :
if caption and len(caption) > 0:
cTag = Tag(soup, "p", [("class", "caption")])
c = self.fixChars(self.tag_to_string(caption,use_alt=False)).strip()
mp_off = c.find("More Photos")
if mp_off >= 0:
c = c[:mp_off]
cTag.insert(0, c)
caption.replaceWith(cTag)
except:
self.log("ERROR: Problem in change captions to italic")
try:
# Change <nyt_headline> to <h2>
h1 = soup.find('h1')
blogheadline = str(h1) # added for dealbook
if h1:
headline = h1.find("nyt_headline")
if headline:
tag = Tag(soup, "h2")
tag['class'] = "headline"
tag.insert(0, self.fixChars(headline.contents[0]))
h1.replaceWith(tag)
elif blogheadline.find('entry-title'): # added for dealbook
tag = Tag(soup, "h2") # added for dealbook
tag['class'] = "headline" # added for dealbook
tag.insert(0, self.fixChars(h1.contents[0])) # added for dealbook
h1.replaceWith(tag) # added for dealbook
else:
# Blog entry - replace headline, remove <hr> tags - BCC I think this is no longer functional 1-18-2011
headline = soup.find('title')
if headline:
tag = Tag(soup, "h2")
tag['class'] = "headline"
tag.insert(0, self.fixChars(self.tag_to_string(headline,False)))
soup.insert(0, tag)
hrs = soup.findAll('hr')
for hr in hrs:
hr.extract()
except:
self.log("ERROR: Problem in Change <nyt_headline> to <h2>")
try:
# if this is from a blog (dealbook, fix the byline format
bylineauthor = soup.find('address',attrs={'class':'byline author vcard'})
if bylineauthor:
tag = Tag(soup, "h6")
tag['class'] = "byline"
tag.insert(0, self.fixChars(self.tag_to_string(bylineauthor,False)))
bylineauthor.replaceWith(tag)
except:
self.log("ERROR: fixing byline author format")
try:
# if this is a blog (dealbook) fix the credit style for the pictures
blogcredit = soup.find('div',attrs={'class':'credit'})
if blogcredit:
tag = Tag(soup, "h6")
tag['class'] = "credit"
tag.insert(0, self.fixChars(self.tag_to_string(blogcredit,False)))
blogcredit.replaceWith(tag)
except:
self.log("ERROR: fixing credit format")
try:
# Change <h1> to <h3> - used in editorial blogs
masthead = soup.find("h1")
if masthead:
# Nuke the href
if masthead.a:
del(masthead.a['href'])
tag = Tag(soup, "h3")
tag.insert(0, self.fixChars(masthead.contents[0]))
masthead.replaceWith(tag)
except:
self.log("ERROR: Problem in Change <h1> to <h3> - used in editorial blogs")
try:
# Change <span class="bold"> to <b>
for subhead in soup.findAll(True, {'class':'bold'}) :
if subhead.contents:
bTag = Tag(soup, "b")
bTag.insert(0, subhead.contents[0])
subhead.replaceWith(bTag)
except:
self.log("ERROR: Problem in Change <h1> to <h3> - used in editorial blogs")
try:
# remove the <strong> update tag
blogupdated = soup.find('span', {'class':'update'})
if blogupdated:
blogupdated.replaceWith("")
except:
self.log("ERROR: Removing strong tag")
try:
divTag = soup.find('div',attrs={'id':'articleBody'})
if divTag:
divTag['class'] = divTag['id']
except:
self.log("ERROR: Problem in soup.find(div,attrs={id:articleBody})")
try:
# Add class="authorId" to <div> so we can format with CSS
divTag = soup.find('div',attrs={'id':'authorId'})
if divTag and divTag.contents[0]:
tag = Tag(soup, "p")
tag['class'] = "authorId"
tag.insert(0, self.fixChars(self.tag_to_string(divTag.contents[0],
use_alt=False)))
divTag.replaceWith(tag)
except:
self.log("ERROR: Problem in Add class=authorId to <div> so we can format with CSS")
#print(strftime("%H:%M:%S")+" -- POSTPROCESS TITLE="+self.tag_to_string(soup.title))
return soup
def populate_article_metadata(self, article, soup, first):
if not first:
return
idxdiv = soup.find('div',attrs={'class':'articleSpanImage'})
if idxdiv is not None:
if idxdiv.img:
self.add_toc_thumbnail(article, re.sub(r'links\\link\d+\\','',idxdiv.img['src']))
else:
img = soup.find('body').find('img')
if img is not None:
self.add_toc_thumbnail(article, re.sub(r'links\\link\d+\\','',img['src']))
shortparagraph = ""
try:
if len(article.text_summary.strip()) == 0:
articlebodies = soup.findAll('div',attrs={'class':'articleBody'})
if articlebodies:
for articlebody in articlebodies:
if articlebody:
paras = articlebody.findAll('p')
for p in paras:
refparagraph = self.massageNCXText(self.tag_to_string(p,use_alt=False)).strip()
# account for blank paragraphs and short paragraphs by appending them to longer ones
if len(refparagraph) > 0:
if len(refparagraph) > 70: # approximately one line of text
newpara = shortparagraph + refparagraph
newparaDateline,newparaEm,newparaDesc = newpara.partition('&mdash;')
if newparaEm == '':
newparaDateline,newparaEm,newparaDesc = newpara.partition('')
if newparaEm == '':
newparaDesc = newparaDateline
article.summary = article.text_summary = newparaDesc.strip()
return
else:
shortparagraph = refparagraph + " "
if shortparagraph.strip().find(" ") == -1 and not shortparagraph.strip().endswith(":"):
shortparagraph = shortparagraph + "- "
else:
article.summary = article.text_summary = self.massageNCXText(article.text_summary)
except:
self.log("Error creating article descriptions")
return