|
onlinenewsreader.net
Posts: 328
Karma: 10143
Join Date: Dec 2009
Location: Phoenix, AZ & Victoria, BC
Device: Kindle 3, Kindle Fire, IPad3, iPhone4, Playbook, HTC Inspire
|
Updated New York Times recipe
I have updated the New York Times recipe to include the Web edition of the newspaper. This edition is taken from section index pages linked from www.nytimes.com and includes a lot of current and past content. Unconstrained, the feed will generate a 25MB MOBI file. If you confine articles to 7 days old or newer, it's around 15MB. A more practical approach may be to identify the sections you are interested in (e.g., 'Sunday Magazine', 'Politics', etc.) in the includeSections variable and those will be the only sections processed. Look at the variable web_sections to see what sections are available.
I have also corrected some problems in the daily New York Times recipe and the Headlines recipe, so I am posting them all here. Note that which recipe you get is controlled by the recipe variables headlinesOnly and webEdition.
New York Times daily recipe which you get by setting headlinesOnly=False and webEdition=False
Code:
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
'''
nytimes.com
'''
import re, string, time
from calibre import entity_to_unicode, strftime
from datetime import timedelta, datetime, date
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, NavigableString, Comment, BeautifulStoneSoup
class NYTimes(BasicNewsRecipe):
# set headlinesOnly to True for the headlines-only version. If True, webEdition is ignored.
headlinesOnly = False
# set webEdition to True for the Web edition of the newspaper. Set oldest_article to the
# number of days old an article can be for inclusion. If oldest_article = 0 all articles
# will be included. Note: oldest_article is ignored if webEdition = False
webEdition = False
oldest_article = 7
# 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
# 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'),
(u'Week in Review',u'weekinreview')]
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 = True
else:
title='New York Times'
description = 'Today\'s New York Times'
needs_subscription = True
month_list = ['january','february','march','april','may','june','july','august','september','october','november','december']
def decode_us_date(self,datestr):
udate = datestr.strip().lower().split()
try:
m = self.month_list.index(udate[0])+1
except:
return date.today()
d = int(udate[1])
y = int(udate[2])
try:
d = date(y,m,d)
except:
d = date.today
return d
earliest_date = date.today() - timedelta(days=oldest_article)
__author__ = 'GRiker/Kovid Goyal/Nick Redding'
language = 'en'
requires_version = (0, 7, 5)
timefmt = ''
masthead_url = 'http://graphics8.nytimes.com/images/misc/nytlogo379x64.gif'
cover_margins = (18,18,'grey99')
remove_tags_before = dict(id='article')
remove_tags_after = dict(id='article')
remove_tags = [dict(attrs={'class':[
'articleFooter',
'articleTools',
'columnGroup doubleRule',
'columnGroup singleRule',
'columnGroup last',
'columnGroup last',
'doubleRule',
'dottedLine',
'entry-meta',
'entry-response module',
'icon enlargeThis',
'leftNavTabs',
'metaFootnote',
'module box nav',
'nextArticleLink',
'nextArticleLink clearfix',
'post-tools',
'relatedSearchesModule',
'side_tool',
'singleAd',
re.compile('^subNavigation'),
re.compile('^leaderboard'),
re.compile('^module'),
]}),
dict(id=[
'adxLeaderboard',
'adxSponLink',
'archive',
'articleExtras',
'articleInline',
'blog_sidebar',
'businessSearchBar',
'cCol',
'entertainmentSearchBar',
'footer',
'header',
'header_search',
'inlineBox',
'login',
'masthead',
'masthead-nav',
'memberTools',
'navigation',
'portfolioInline',
'readerReviews',
'readerReviewsCount',
'relatedArticles',
'relatedTopics',
'respond',
'side_search',
'side_index',
'side_tool',
'toolsRight',
]),
dict(name=['script', 'noscript', 'style','form','hr'])]
no_stylesheets = True
extra_css = '''
.articleHeadline { text-align: left; margin-top:0.5em; margin-bottom:0.25em; }
.credit { text-align: right; font-size: small; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.byline { text-align: left; font-size: small; line-height:1em; margin-top:10px; margin-left:0; margin-right:0; margin-bottom: 0; }
.dateline { text-align: left; font-size: small; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.kicker { font-size: small; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.timestamp { text-align: left; font-size: small; }
.caption { font-size: small; font-style:italic; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
a:link {text-decoration: none; }
.articleBody { }
.authorId {text-align: left; }
.image {text-align: center;}
.source {text-align: left; }'''
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 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 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"):
return True
if 'nytimes.com' not in url:
return True
if 'podcast' in url:
return True
if '/video/' 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
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()
if self.username is not None and self.password is not None:
br.open('http://www.nytimes.com/auth/login')
br.select_form(name='login')
br['USERID'] = self.username
br['PASSWORD'] = self.password
raw = br.submit().read()
if 'Please try again' in raw:
raise Exception('Your username and password are incorrect')
return br
def skip_ad_pages(self, soup):
# Skip ad pages served before actual article
skip_tag = soup.find(True, {'name':'skip'})
if skip_tag is not None:
self.log.warn("Found forwarding link: %s" % skip_tag.parent['href'])
url = 'http://www.nytimes.com' + re.sub(r'\?.*', '', skip_tag.parent['href'])
url += '?pagewanted=all'
self.log.warn("Skipping ad to article at '%s'" % url)
return self.index_to_soup(url, raw=True)
def get_cover_url(self):
cover = None
st = time.localtime()
year = str(st.tm_year)
month = "%.2d" % st.tm_mon
day = "%.2d" % st.tm_mday
cover = 'http://graphics8.nytimes.com/images/' + year + '/' + month +'/' + day +'/nytfrontpage/scan.jpg'
br = BasicNewsRecipe.get_browser()
try:
br.open(cover)
except:
self.log("\nCover unavailable")
cover = None
return cover
def short_title(self):
return self.title
def index_to_soup(self, url_or_raw, raw=False):
'''
OVERRIDE of class method
deals with various page encodings between index and articles
'''
def get_the_soup(docEncoding, url_or_raw, raw=False) :
if re.match(r'\w+://', url_or_raw):
f = self.browser.open(url_or_raw)
_raw = f.read()
f.close()
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:
_raw = _raw.decode(docEncoding, 'replace')
massage = list(BeautifulSoup.MARKUP_MASSAGE)
massage.append((re.compile(r'&(\S+?);'), lambda match: entity_to_unicode(match, encoding=self.encoding)))
return BeautifulSoup(_raw, markupMassage=massage)
# Entry point
soup = get_the_soup( self.encoding, url_or_raw )
contentType = soup.find(True,attrs={'http-equiv':'Content-Type'})
docEncoding = str(contentType)[str(contentType).find('charset=') + len('charset='):str(contentType).rfind('"')]
if docEncoding == '' :
docEncoding = self.encoding
if self.verbose > 2:
self.log( " document encoding: '%s'" % docEncoding)
if docEncoding != self.encoding :
soup = get_the_soup(docEncoding, url_or_raw)
return soup
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("&","&", 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
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 not self.articles.has_key(feed):
self.ans.append(feed)
self.articles[feed] = []
self.articles[feed].append(
dict(title=title, url=url, date=pubdate,
description=description, author=author,
content=''))
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
print 'Index URL: '+'http://www.nytimes.com/pages/'+index_url+'/index.html'
soup = self.index_to_soup('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', 'story', 'story headline','sectionHeader','headlinesOnly multiline flush']}):
if div['class'] in ['story', 'story headline'] :
self.handle_article(div)
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 self.articles.has_key(k)]
return self.filter_ans(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 self.articles.has_key(k)]
return self.filter_ans(self.ans)
def parse_headline_index(self):
soup = self.index_to_soup('http://www.nytimes.com/pages/todaysheadlines/')
# Fetch the content table
content_table = soup.find('table',{'id':'content'})
if content_table is None:
self.log("FATAL ERROR: CANNOT FIND CONTENT TABLE")
return None
# Within this table are <td id=".*Column.*"> entries, each containing one or more h6 tags which represent sections
for td_col in content_table.findAll('td', {'id' : re.compile('Column')}):
for div_sec in td_col.findAll('div',recursive=False):
for h6_sec_name in div_sec.findAll('h6',{'style' : re.compile('text-transform: *uppercase')}):
section_name = self.tag_to_string(h6_sec_name,use_alt=False)
section_name = re.sub(r'^ *$','',section_name)
if section_name == '':
continue
if self.includeSections != []:
if section_name not in self.includeSections:
print "SECTION NOT INCLUDED: ",section_name
continue
if section_name in self.excludeSections:
print "SECTION EXCLUDED: ",section_name
continue
section_name=string.capwords(section_name)
section_name = section_name.replace('Op-ed','Op-Ed')
section_name = section_name.replace('U.s.','U.S.')
section_name = section_name.replace('N.y.','N.Y.')
pubdate = strftime('%a, %d %b')
search_div = div_sec
for next_tag in h6_sec_name.findNextSiblings(True):
if next_tag.__class__.__name__ == 'Tag':
if next_tag.name == 'div':
search_div = next_tag
break
# Get the articles
for h3_item in search_div.findAll('h3'):
byline = h3_item.h6
if byline is not None:
author = self.tag_to_string(byline,usa_alt=False)
else:
author = ''
a = h3_item.find('a', href=True)
if not a:
continue
url = re.sub(r'\?.*', '', a['href'])
if self.exclude_url(url):
continue
url += '?pagewanted=all'
if self.filterDuplicates:
if url in self.url_list:
continue
self.url_list.append(url)
title = self.tag_to_string(a, use_alt=True).strip()
desc = h3_item.find('p')
if desc is not None:
description = self.tag_to_string(desc,use_alt=False)
else:
description = ''
if not self.articles.has_key(section_name):
self.ans.append(section_name)
self.articles[section_name] = []
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 self.articles.has_key(k)]
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):
paras = soup.findAll(True)
for para in paras:
aTags = para.findAll('a')
for a in aTags:
if a.img is None:
a.replaceWith(a.renderContents().decode('cp1252','replace'))
return soup
def preprocess_html(self, soup):
if self.webEdition & (self.oldest_article>0):
date_tag = soup.find(True,attrs={'class': ['dateline','date']})
if date_tag:
date_str = self.tag_to_string(date_tag,use_alt=False)
date_str = date_str.replace('Published:','')
date_items = date_str.split(',')
try:
datestring = date_items[0]+' '+date_items[1]
article_date = self.decode_us_date(datestring)
except:
article_date = date.today()
if article_date < self.earliest_date:
self.log("Skipping article dated %s" % date_str)
return None
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()
return self.strip_anchors(soup)
def postprocess_html(self,soup, True):
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
article_body = soup.find(True, {'id':'articleBody'})
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 <<<")
# Change captions to italic
for caption in soup.findAll(True, {'class':'caption'}) :
if caption and caption.contents[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)
# Change <nyt_headline> to <h2>
h1 = soup.find('h1')
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)
else:
# Blog entry - replace headline, remove <hr> tags
headline = soup.find('title')
if headline:
tag = Tag(soup, "h2")
tag['class'] = "headline"
tag.insert(0, self.fixChars(headline.contents[0]))
soup.insert(0, tag)
hrs = soup.findAll('hr')
for hr in hrs:
hr.extract()
# 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)
# 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)
divTag = soup.find('div',attrs={'id':'articleBody'})
if divTag:
divTag['class'] = divTag['id']
# 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)
return soup
New York Times Headlines which you get by setting headlinesOnly=True and webEdition=False
Code:
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
'''
nytimes.com
'''
import re, string, time
from calibre import entity_to_unicode, strftime
from datetime import timedelta, datetime, date
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, NavigableString, Comment, BeautifulStoneSoup
class NYTimes(BasicNewsRecipe):
# 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_article to the
# number of days old an article can be for inclusion. If oldest_article = 0 all articles
# will be included. Note: oldest_article is ignored if webEdition = False
webEdition = False
oldest_article = 7
# 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
# 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'),
(u'Week in Review',u'weekinreview')]
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 = True
else:
title='New York Times'
description = 'Today\'s New York Times'
needs_subscription = True
month_list = ['january','february','march','april','may','june','july','august','september','october','november','december']
def decode_us_date(self,datestr):
udate = datestr.strip().lower().split()
try:
m = self.month_list.index(udate[0])+1
except:
return date.today()
d = int(udate[1])
y = int(udate[2])
try:
d = date(y,m,d)
except:
d = date.today
return d
earliest_date = date.today() - timedelta(days=oldest_article)
__author__ = 'GRiker/Kovid Goyal/Nick Redding'
language = 'en'
requires_version = (0, 7, 5)
timefmt = ''
masthead_url = 'http://graphics8.nytimes.com/images/misc/nytlogo379x64.gif'
cover_margins = (18,18,'grey99')
remove_tags_before = dict(id='article')
remove_tags_after = dict(id='article')
remove_tags = [dict(attrs={'class':[
'articleFooter',
'articleTools',
'columnGroup doubleRule',
'columnGroup singleRule',
'columnGroup last',
'columnGroup last',
'doubleRule',
'dottedLine',
'entry-meta',
'entry-response module',
'icon enlargeThis',
'leftNavTabs',
'metaFootnote',
'module box nav',
'nextArticleLink',
'nextArticleLink clearfix',
'post-tools',
'relatedSearchesModule',
'side_tool',
'singleAd',
re.compile('^subNavigation'),
re.compile('^leaderboard'),
re.compile('^module'),
]}),
dict(id=[
'adxLeaderboard',
'adxSponLink',
'archive',
'articleExtras',
'articleInline',
'blog_sidebar',
'businessSearchBar',
'cCol',
'entertainmentSearchBar',
'footer',
'header',
'header_search',
'inlineBox',
'login',
'masthead',
'masthead-nav',
'memberTools',
'navigation',
'portfolioInline',
'readerReviews',
'readerReviewsCount',
'relatedArticles',
'relatedTopics',
'respond',
'side_search',
'side_index',
'side_tool',
'toolsRight',
]),
dict(name=['script', 'noscript', 'style','form','hr'])]
no_stylesheets = True
extra_css = '''
.articleHeadline { text-align: left; margin-top:0.5em; margin-bottom:0.25em; }
.credit { text-align: right; font-size: small; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.byline { text-align: left; font-size: small; line-height:1em; margin-top:10px; margin-left:0; margin-right:0; margin-bottom: 0; }
.dateline { text-align: left; font-size: small; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.kicker { font-size: small; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.timestamp { text-align: left; font-size: small; }
.caption { font-size: small; font-style:italic; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
a:link {text-decoration: none; }
.articleBody { }
.authorId {text-align: left; }
.image {text-align: center;}
.source {text-align: left; }'''
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 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 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"):
return True
if 'nytimes.com' not in url:
return True
if 'podcast' in url:
return True
if '/video/' 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
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()
if self.username is not None and self.password is not None:
br.open('http://www.nytimes.com/auth/login')
br.select_form(name='login')
br['USERID'] = self.username
br['PASSWORD'] = self.password
raw = br.submit().read()
if 'Please try again' in raw:
raise Exception('Your username and password are incorrect')
return br
def skip_ad_pages(self, soup):
# Skip ad pages served before actual article
skip_tag = soup.find(True, {'name':'skip'})
if skip_tag is not None:
self.log.warn("Found forwarding link: %s" % skip_tag.parent['href'])
url = 'http://www.nytimes.com' + re.sub(r'\?.*', '', skip_tag.parent['href'])
url += '?pagewanted=all'
self.log.warn("Skipping ad to article at '%s'" % url)
return self.index_to_soup(url, raw=True)
def get_cover_url(self):
cover = None
st = time.localtime()
year = str(st.tm_year)
month = "%.2d" % st.tm_mon
day = "%.2d" % st.tm_mday
cover = 'http://graphics8.nytimes.com/images/' + year + '/' + month +'/' + day +'/nytfrontpage/scan.jpg'
br = BasicNewsRecipe.get_browser()
try:
br.open(cover)
except:
self.log("\nCover unavailable")
cover = None
return cover
def short_title(self):
return self.title
def index_to_soup(self, url_or_raw, raw=False):
'''
OVERRIDE of class method
deals with various page encodings between index and articles
'''
def get_the_soup(docEncoding, url_or_raw, raw=False) :
if re.match(r'\w+://', url_or_raw):
f = self.browser.open(url_or_raw)
_raw = f.read()
f.close()
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:
_raw = _raw.decode(docEncoding, 'replace')
massage = list(BeautifulSoup.MARKUP_MASSAGE)
massage.append((re.compile(r'&(\S+?);'), lambda match: entity_to_unicode(match, encoding=self.encoding)))
return BeautifulSoup(_raw, markupMassage=massage)
# Entry point
soup = get_the_soup( self.encoding, url_or_raw )
contentType = soup.find(True,attrs={'http-equiv':'Content-Type'})
docEncoding = str(contentType)[str(contentType).find('charset=') + len('charset='):str(contentType).rfind('"')]
if docEncoding == '' :
docEncoding = self.encoding
if self.verbose > 2:
self.log( " document encoding: '%s'" % docEncoding)
if docEncoding != self.encoding :
soup = get_the_soup(docEncoding, url_or_raw)
return soup
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("&","&", 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
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 not self.articles.has_key(feed):
self.ans.append(feed)
self.articles[feed] = []
self.articles[feed].append(
dict(title=title, url=url, date=pubdate,
description=description, author=author,
content=''))
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
print 'Index URL: '+'http://www.nytimes.com/pages/'+index_url+'/index.html'
soup = self.index_to_soup('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', 'story', 'story headline','sectionHeader','headlinesOnly multiline flush']}):
if div['class'] in ['story', 'story headline'] :
self.handle_article(div)
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 self.articles.has_key(k)]
return self.filter_ans(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 self.articles.has_key(k)]
return self.filter_ans(self.ans)
def parse_headline_index(self):
soup = self.index_to_soup('http://www.nytimes.com/pages/todaysheadlines/')
# Fetch the content table
content_table = soup.find('table',{'id':'content'})
if content_table is None:
self.log("FATAL ERROR: CANNOT FIND CONTENT TABLE")
return None
# Within this table are <td id=".*Column.*"> entries, each containing one or more h6 tags which represent sections
for td_col in content_table.findAll('td', {'id' : re.compile('Column')}):
for div_sec in td_col.findAll('div',recursive=False):
for h6_sec_name in div_sec.findAll('h6',{'style' : re.compile('text-transform: *uppercase')}):
section_name = self.tag_to_string(h6_sec_name,use_alt=False)
section_name = re.sub(r'^ *$','',section_name)
if section_name == '':
continue
if self.includeSections != []:
if section_name not in self.includeSections:
print "SECTION NOT INCLUDED: ",section_name
continue
if section_name in self.excludeSections:
print "SECTION EXCLUDED: ",section_name
continue
section_name=string.capwords(section_name)
section_name = section_name.replace('Op-ed','Op-Ed')
section_name = section_name.replace('U.s.','U.S.')
section_name = section_name.replace('N.y.','N.Y.')
pubdate = strftime('%a, %d %b')
search_div = div_sec
for next_tag in h6_sec_name.findNextSiblings(True):
if next_tag.__class__.__name__ == 'Tag':
if next_tag.name == 'div':
search_div = next_tag
break
# Get the articles
for h3_item in search_div.findAll('h3'):
byline = h3_item.h6
if byline is not None:
author = self.tag_to_string(byline,usa_alt=False)
else:
author = ''
a = h3_item.find('a', href=True)
if not a:
continue
url = re.sub(r'\?.*', '', a['href'])
if self.exclude_url(url):
continue
url += '?pagewanted=all'
if self.filterDuplicates:
if url in self.url_list:
continue
self.url_list.append(url)
title = self.tag_to_string(a, use_alt=True).strip()
desc = h3_item.find('p')
if desc is not None:
description = self.tag_to_string(desc,use_alt=False)
else:
description = ''
if not self.articles.has_key(section_name):
self.ans.append(section_name)
self.articles[section_name] = []
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 self.articles.has_key(k)]
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):
paras = soup.findAll(True)
for para in paras:
aTags = para.findAll('a')
for a in aTags:
if a.img is None:
a.replaceWith(a.renderContents().decode('cp1252','replace'))
return soup
def preprocess_html(self, soup):
if self.webEdition & (self.oldest_article>0):
date_tag = soup.find(True,attrs={'class': ['dateline','date']})
if date_tag:
date_str = self.tag_to_string(date_tag,use_alt=False)
date_str = date_str.replace('Published:','')
date_items = date_str.split(',')
try:
datestring = date_items[0]+' '+date_items[1]
article_date = self.decode_us_date(datestring)
except:
article_date = date.today()
if article_date < self.earliest_date:
self.log("Skipping article dated %s" % date_str)
return None
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()
return self.strip_anchors(soup)
def postprocess_html(self,soup, True):
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
article_body = soup.find(True, {'id':'articleBody'})
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 <<<")
# Change captions to italic
for caption in soup.findAll(True, {'class':'caption'}) :
if caption and caption.contents[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)
# Change <nyt_headline> to <h2>
h1 = soup.find('h1')
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)
else:
# Blog entry - replace headline, remove <hr> tags
headline = soup.find('title')
if headline:
tag = Tag(soup, "h2")
tag['class'] = "headline"
tag.insert(0, self.fixChars(headline.contents[0]))
soup.insert(0, tag)
hrs = soup.findAll('hr')
for hr in hrs:
hr.extract()
# 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)
# 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)
divTag = soup.find('div',attrs={'id':'articleBody'})
if divTag:
divTag['class'] = divTag['id']
# 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)
return soup
New York Times Web Edition which you get by setting headlinesOnly=False and webEdition=True. This recipe has oldest_article=7. You can increase or decrease the size of the download by increasing or decreasing this parameter.
Code:
#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
'''
nytimes.com
'''
import re, string, time
from calibre import entity_to_unicode, strftime
from datetime import timedelta, datetime, date
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, NavigableString, Comment, BeautifulStoneSoup
class NYTimes(BasicNewsRecipe):
# set headlinesOnly to True for the headlines-only version. If True, webEdition is ignored.
headlinesOnly = False
# set webEdition to True for the Web edition of the newspaper. Set oldest_article to the
# number of days old an article can be for inclusion. If oldest_article = 0 all articles
# will be included. Note: oldest_article is ignored if webEdition = False
webEdition = True
oldest_article = 7
# 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
# 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'),
(u'Week in Review',u'weekinreview')]
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 = True
else:
title='New York Times'
description = 'Today\'s New York Times'
needs_subscription = True
month_list = ['january','february','march','april','may','june','july','august','september','october','november','december']
def decode_us_date(self,datestr):
udate = datestr.strip().lower().split()
try:
m = self.month_list.index(udate[0])+1
except:
return date.today()
d = int(udate[1])
y = int(udate[2])
try:
d = date(y,m,d)
except:
d = date.today
return d
earliest_date = date.today() - timedelta(days=oldest_article)
__author__ = 'GRiker/Kovid Goyal/Nick Redding'
language = 'en'
requires_version = (0, 7, 5)
timefmt = ''
masthead_url = 'http://graphics8.nytimes.com/images/misc/nytlogo379x64.gif'
cover_margins = (18,18,'grey99')
remove_tags_before = dict(id='article')
remove_tags_after = dict(id='article')
remove_tags = [dict(attrs={'class':[
'articleFooter',
'articleTools',
'columnGroup doubleRule',
'columnGroup singleRule',
'columnGroup last',
'columnGroup last',
'doubleRule',
'dottedLine',
'entry-meta',
'entry-response module',
'icon enlargeThis',
'leftNavTabs',
'metaFootnote',
'module box nav',
'nextArticleLink',
'nextArticleLink clearfix',
'post-tools',
'relatedSearchesModule',
'side_tool',
'singleAd',
re.compile('^subNavigation'),
re.compile('^leaderboard'),
re.compile('^module'),
]}),
dict(id=[
'adxLeaderboard',
'adxSponLink',
'archive',
'articleExtras',
'articleInline',
'blog_sidebar',
'businessSearchBar',
'cCol',
'entertainmentSearchBar',
'footer',
'header',
'header_search',
'inlineBox',
'login',
'masthead',
'masthead-nav',
'memberTools',
'navigation',
'portfolioInline',
'readerReviews',
'readerReviewsCount',
'relatedArticles',
'relatedTopics',
'respond',
'side_search',
'side_index',
'side_tool',
'toolsRight',
]),
dict(name=['script', 'noscript', 'style','form','hr'])]
no_stylesheets = True
extra_css = '''
.articleHeadline { text-align: left; margin-top:0.5em; margin-bottom:0.25em; }
.credit { text-align: right; font-size: small; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.byline { text-align: left; font-size: small; line-height:1em; margin-top:10px; margin-left:0; margin-right:0; margin-bottom: 0; }
.dateline { text-align: left; font-size: small; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.kicker { font-size: small; line-height:1em;margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
.timestamp { text-align: left; font-size: small; }
.caption { font-size: small; font-style:italic; line-height:1em; margin-top:5px; margin-left:0; margin-right:0; margin-bottom: 0; }
a:link {text-decoration: none; }
.articleBody { }
.authorId {text-align: left; }
.image {text-align: center;}
.source {text-align: left; }'''
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 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 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"):
return True
if 'nytimes.com' not in url:
return True
if 'podcast' in url:
return True
if '/video/' 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
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()
if self.username is not None and self.password is not None:
br.open('http://www.nytimes.com/auth/login')
br.select_form(name='login')
br['USERID'] = self.username
br['PASSWORD'] = self.password
raw = br.submit().read()
if 'Please try again' in raw:
raise Exception('Your username and password are incorrect')
return br
def skip_ad_pages(self, soup):
# Skip ad pages served before actual article
skip_tag = soup.find(True, {'name':'skip'})
if skip_tag is not None:
self.log.warn("Found forwarding link: %s" % skip_tag.parent['href'])
url = 'http://www.nytimes.com' + re.sub(r'\?.*', '', skip_tag.parent['href'])
url += '?pagewanted=all'
self.log.warn("Skipping ad to article at '%s'" % url)
return self.index_to_soup(url, raw=True)
def get_cover_url(self):
cover = None
st = time.localtime()
year = str(st.tm_year)
month = "%.2d" % st.tm_mon
day = "%.2d" % st.tm_mday
cover = 'http://graphics8.nytimes.com/images/' + year + '/' + month +'/' + day +'/nytfrontpage/scan.jpg'
br = BasicNewsRecipe.get_browser()
try:
br.open(cover)
except:
self.log("\nCover unavailable")
cover = None
return cover
def short_title(self):
return self.title
def index_to_soup(self, url_or_raw, raw=False):
'''
OVERRIDE of class method
deals with various page encodings between index and articles
'''
def get_the_soup(docEncoding, url_or_raw, raw=False) :
if re.match(r'\w+://', url_or_raw):
f = self.browser.open(url_or_raw)
_raw = f.read()
f.close()
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:
_raw = _raw.decode(docEncoding, 'replace')
massage = list(BeautifulSoup.MARKUP_MASSAGE)
massage.append((re.compile(r'&(\S+?);'), lambda match: entity_to_unicode(match, encoding=self.encoding)))
return BeautifulSoup(_raw, markupMassage=massage)
# Entry point
soup = get_the_soup( self.encoding, url_or_raw )
contentType = soup.find(True,attrs={'http-equiv':'Content-Type'})
docEncoding = str(contentType)[str(contentType).find('charset=') + len('charset='):str(contentType).rfind('"')]
if docEncoding == '' :
docEncoding = self.encoding
if self.verbose > 2:
self.log( " document encoding: '%s'" % docEncoding)
if docEncoding != self.encoding :
soup = get_the_soup(docEncoding, url_or_raw)
return soup
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("&","&", 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
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 not self.articles.has_key(feed):
self.ans.append(feed)
self.articles[feed] = []
self.articles[feed].append(
dict(title=title, url=url, date=pubdate,
description=description, author=author,
content=''))
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
print 'Index URL: '+'http://www.nytimes.com/pages/'+index_url+'/index.html'
soup = self.index_to_soup('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', 'story', 'story headline','sectionHeader','headlinesOnly multiline flush']}):
if div['class'] in ['story', 'story headline'] :
self.handle_article(div)
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 self.articles.has_key(k)]
return self.filter_ans(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 self.articles.has_key(k)]
return self.filter_ans(self.ans)
def parse_headline_index(self):
soup = self.index_to_soup('http://www.nytimes.com/pages/todaysheadlines/')
# Fetch the content table
content_table = soup.find('table',{'id':'content'})
if content_table is None:
self.log("FATAL ERROR: CANNOT FIND CONTENT TABLE")
return None
# Within this table are <td id=".*Column.*"> entries, each containing one or more h6 tags which represent sections
for td_col in content_table.findAll('td', {'id' : re.compile('Column')}):
for div_sec in td_col.findAll('div',recursive=False):
for h6_sec_name in div_sec.findAll('h6',{'style' : re.compile('text-transform: *uppercase')}):
section_name = self.tag_to_string(h6_sec_name,use_alt=False)
section_name = re.sub(r'^ *$','',section_name)
if section_name == '':
continue
if self.includeSections != []:
if section_name not in self.includeSections:
print "SECTION NOT INCLUDED: ",section_name
continue
if section_name in self.excludeSections:
print "SECTION EXCLUDED: ",section_name
continue
section_name=string.capwords(section_name)
section_name = section_name.replace('Op-ed','Op-Ed')
section_name = section_name.replace('U.s.','U.S.')
section_name = section_name.replace('N.y.','N.Y.')
pubdate = strftime('%a, %d %b')
search_div = div_sec
for next_tag in h6_sec_name.findNextSiblings(True):
if next_tag.__class__.__name__ == 'Tag':
if next_tag.name == 'div':
search_div = next_tag
break
# Get the articles
for h3_item in search_div.findAll('h3'):
byline = h3_item.h6
if byline is not None:
author = self.tag_to_string(byline,usa_alt=False)
else:
author = ''
a = h3_item.find('a', href=True)
if not a:
continue
url = re.sub(r'\?.*', '', a['href'])
if self.exclude_url(url):
continue
url += '?pagewanted=all'
if self.filterDuplicates:
if url in self.url_list:
continue
self.url_list.append(url)
title = self.tag_to_string(a, use_alt=True).strip()
desc = h3_item.find('p')
if desc is not None:
description = self.tag_to_string(desc,use_alt=False)
else:
description = ''
if not self.articles.has_key(section_name):
self.ans.append(section_name)
self.articles[section_name] = []
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 self.articles.has_key(k)]
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):
paras = soup.findAll(True)
for para in paras:
aTags = para.findAll('a')
for a in aTags:
if a.img is None:
a.replaceWith(a.renderContents().decode('cp1252','replace'))
return soup
def preprocess_html(self, soup):
if self.webEdition & (self.oldest_article>0):
date_tag = soup.find(True,attrs={'class': ['dateline','date']})
if date_tag:
date_str = self.tag_to_string(date_tag,use_alt=False)
date_str = date_str.replace('Published:','')
date_items = date_str.split(',')
try:
datestring = date_items[0]+' '+date_items[1]
article_date = self.decode_us_date(datestring)
except:
article_date = date.today()
if article_date < self.earliest_date:
self.log("Skipping article dated %s" % date_str)
return None
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()
return self.strip_anchors(soup)
def postprocess_html(self,soup, True):
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
article_body = soup.find(True, {'id':'articleBody'})
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 <<<")
# Change captions to italic
for caption in soup.findAll(True, {'class':'caption'}) :
if caption and caption.contents[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)
# Change <nyt_headline> to <h2>
h1 = soup.find('h1')
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)
else:
# Blog entry - replace headline, remove <hr> tags
headline = soup.find('title')
if headline:
tag = Tag(soup, "h2")
tag['class'] = "headline"
tag.insert(0, self.fixChars(headline.contents[0]))
soup.insert(0, tag)
hrs = soup.findAll('hr')
for hr in hrs:
hr.extract()
# 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)
# 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)
divTag = soup.find('div',attrs={'id':'articleBody'})
if divTag:
divTag['class'] = divTag['id']
# 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)
return soup
|