calibre/resources/recipes/nytimes.recipe

534 lines
22 KiB
Python

#!/usr/bin/env python
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
'''
nytimes.com
'''
import re
import time
from calibre import entity_to_unicode
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, NavigableString, \
Comment, BeautifulStoneSoup
class NYTimes(BasicNewsRecipe):
title = 'New York Times Top Stories'
__author__ = 'GRiker'
language = 'en'
description = 'Top Stories from the New York Times'
# List of sections typically included in Top Stories. Use a keyword from the
# right column in the excludeSectionKeywords[] list to skip downloading that section
sections = {
'arts' : 'Arts',
'business' : 'Business',
'diningwine' : 'Dining & Wine',
'editorials' : 'Editorials',
'health' : 'Health',
'magazine' : 'Magazine',
'mediaadvertising' : 'Media & Advertising',
'newyorkregion' : 'New York/Region',
'oped' : 'Op-Ed',
'politics' : 'Politics',
'science' : 'Science',
'sports' : 'Sports',
'technology' : 'Technology',
'topstories' : 'Top Stories',
'travel' : 'Travel',
'us' : 'U.S.',
'world' : 'World'
}
# Add section keywords from the right column above to skip that section
# For example, to skip sections containing the word 'Sports' or 'Dining', use:
# excludeSectionKeywords = ['Sports', 'Dining']
# Fetch only Business and Technology
# excludeSectionKeywords = ['Arts','Dining','Editorials','Health','Magazine','Media','Region','Op-Ed','Politics','Science','Sports','Top Stories','Travel','U.S.','World']
# Fetch only Top Stories
# excludeSectionKeywords = ['Arts','Business','Dining','Editorials','Health','Magazine','Media','Region','Op-Ed','Politics','Science','Sports','Technology','Travel','U.S.','World']
# By default, no sections are skipped.
excludeSectionKeywords = []
# 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 = True
# The maximum number of articles that will be downloaded
max_articles_per_feed = 40
timefmt = ''
needs_subscription = True
masthead_url = 'http://graphics8.nytimes.com/images/misc/nytlogo379x64.gif'
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',
'icon enlargeThis',
'leftNavTabs',
'module box nav',
'nextArticleLink',
'nextArticleLink clearfix',
'post-tools',
'relatedSearchesModule',
'side_tool',
'singleAd',
'subNavigation tabContent active',
'subNavigation tabContent active clearfix',
]}),
dict(id=[
'adxLeaderboard',
'archive',
'articleExtras',
'articleInline',
'blog_sidebar',
'businessSearchBar',
'cCol',
'entertainmentSearchBar',
'footer',
'header',
'header_search',
'login',
'masthead',
'masthead-nav',
'memberTools',
'navigation',
'portfolioInline',
'relatedArticles',
'side_search',
'side_index',
'side_tool',
'toolsRight',
]),
dict(name=['script', 'noscript', 'style'])]
no_stylesheets = True
extra_css = '.headline {text-align: left;}\n \
.byline {font-family: monospace; \
text-align: left; \
margin-top: 0px; \
margin-bottom: 0px;}\n \
.dateline {font-size: small; \
margin-top: 0px; \
margin-bottom: 0px;}\n \
.timestamp {font-size: small; \
margin-top: 0px; \
margin-bottom: 0px;}\n \
.source {text-align: left;}\n \
.image {text-align: center;}\n \
.credit {text-align: right; \
font-size: small; \
margin-top: 0px; \
margin-bottom: 0px;}\n \
.articleBody {text-align: left;}\n \
.authorId {text-align: left; \
font-style: italic;}\n '
def dump_ans(self, ans) :
total_article_count = 0
for section in ans :
if self.verbose:
self.log("section %s: %d articles" % (section[0], len(section[1])) )
for article in section[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')))
self.log( "Queued %d articles" % total_article_count )
def fixChars(self,string):
# Replace lsquo (\x91)
fixed = re.sub("\x91","&#8216;",string)
# Replace rsquo (\x92)
fixed = re.sub("\x92","&#8217;",fixed)
# Replace ldquo (\x93)
fixed = re.sub("\x93","&#8220;",fixed)
# Replace rdquo (\x94)
fixed = re.sub("\x94","&#8221;",fixed)
# Replace ndash (\x96)
fixed = re.sub("\x96","&#8211;",fixed)
# Replace mdash (\x97)
fixed = re.sub("\x97","&#8212;",fixed)
return fixed
def get_browser(self):
br = BasicNewsRecipe.get_browser()
if self.username is not None and self.password is not None:
try:
br.open('http://www.nytimes.com/auth/login')
br.select_form(name='login')
br['USERID'] = self.username
br['PASSWORD'] = self.password
br.submit()
except:
self.log("\nFailed to login")
return br
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 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
print "index_to_soup()"
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 '&#38;'
massaged = re.sub("&","&#38;", massaged)
return self.fixChars(massaged)
else:
return description
def parse_index(self):
articles = {}
ans = []
feed = key = 'All Top Stories'
articles[key] = []
ans.append(key)
self.log("Scanning 1 section ...")
soup = self.index_to_soup('http://www.nytimes.com/pages/todaysheadlines/')
# Fetch the outer table
table = soup.find('table')
previousTable = table
# Find the deepest table containing the stories
while True :
table = table.find('table')
if table.find(text=re.compile('top stories start')) :
previousTable = table
continue
else :
table = previousTable
break
# There are multiple subtables, find the one containing the stories
for block in table.findAll('table') :
if block.find(text=re.compile('top stories start')) :
table = block
break
else :
continue
# Again there are multiple subtables, find the one containing the stories
for storyblock in table.findAll('table') :
if storyblock.find(text=re.compile('top stories start')) :
break
else :
continue
skipThisSection = False
todays_article_count = 0
# Within this table are <font face="times new roman, times, san serif"> entries
self.log("Fetching feed Top Stories")
for tr in storyblock.findAllNext('tr'):
if tr.find('span') is not None :
sectionblock = tr.find(True, attrs={'face':['times new roman, times,sans serif',
'times new roman,times, sans serif',
'times new roman, times, sans serif']})
section = None
bylines = []
descriptions = []
pubdate = None
# Get the Section title
for (x,i) in enumerate(sectionblock.contents) :
skipThisSection = False
# Extract the section title
if ('Comment' in str(i.__class__)) :
if 'start(name=' in i :
section = i[i.find('=')+1:-2]
if not self.sections.has_key(section) :
skipThisSection = True
break
# Check for excluded section
if len(self.excludeSectionKeywords):
key = self.sections[section]
excluded = re.compile('|'.join(self.excludeSectionKeywords))
if excluded.search(key) or articles.has_key(key):
skipThisSection = True
break
# Get the bylines and descriptions
if not skipThisSection :
lines = sectionblock.contents
contentStrings = []
for line in lines:
if not isinstance(line, Comment) and line.strip and line.strip() > "":
contentStrings.append(line.strip())
# Gather the byline/description pairs
bylines = []
descriptions = []
for contentString in contentStrings:
if contentString[0:3] == 'By ' and contentString[3].isupper() :
bylines.append(contentString)
else:
descriptions.append(contentString)
# Fetch the article titles and URLs
articleCount = len(sectionblock.findAll('span'))
todays_article_count += articleCount
for (i,span) in enumerate(sectionblock.findAll(attrs={'class':'headlineWrapper'})) :
a = span.find('a', href=True)
url = re.sub(r'\?.*', '', a['href'])
url += '?pagewanted=all'
title = self.tag_to_string(a, use_alt=True)
# prepend the section name
title = self.sections[section] + " &middot; " + title
if not isinstance(title, unicode):
title = title.decode('utf-8', 'replace')
# Allow for unattributed, undescribed entries "Editor's Note"
if i >= len(descriptions) :
description = None
else :
description = descriptions[i]
if len(bylines) == articleCount :
author = bylines[i]
else :
author = None
# Check for duplicates
duplicateFound = False
if len(articles[feed]) > 1:
for article in articles[feed] :
if url == article['url'] :
duplicateFound = True
break
if duplicateFound:
# Continue fetching, don't add this article
todays_article_count -= 1
continue
if not articles.has_key(feed):
articles[feed] = []
articles[feed].append(
dict(title=title, url=url, date=pubdate,
description=description, author=author, content=''))
# self.log("Queuing %d articles from %s" % (todays_article_count, "Top Stories"))
ans = self.sort_index_by(ans, {'Top Stories':-1})
ans = [(key, articles[key]) for key in ans if articles.has_key(key)]
self.dump_ans(ans)
return ans
def preprocess_html(self, soup):
# Skip ad pages served before actual article
skip_tag = soup.find(True, {'name':'skip'})
if skip_tag is not None:
self.log.error("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.error("Skipping ad to article at '%s'" % url)
soup = self.index_to_soup(url)
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 after headline
cgFirst = soup.find(True, {'class':'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 class="kicker" to <h3>
kicker = soup.find(True, {'class':'kicker'})
if kicker and kicker.contents[0]:
h3Tag = Tag(soup, "h3")
h3Tag.insert(0, self.fixChars(self.tag_to_string(kicker,
use_alt=False)))
kicker.replaceWith(h3Tag)
# Change captions to italic -1
for caption in soup.findAll(True, {'class':'caption'}) :
if caption and caption.contents[0]:
emTag = Tag(soup, "em")
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]
emTag.insert(0, c)
hrTag = Tag(soup, 'hr')
#hrTag['style'] = "margin-top:0em;margin-bottom:0em"
emTag.insert(1, hrTag)
caption.replaceWith(emTag)
# 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)
# Synthesize a section header
dsk = soup.find('meta', attrs={'name':'dsk'})
if dsk and dsk.has_key('content'):
hTag = Tag(soup,'h3')
hTag['class'] = 'section'
hTag.insert(0,NavigableString(dsk['content']))
articleTag = soup.find(True, attrs={'id':'article'})
if articleTag:
articleTag.insert(0,hTag)
# Add class="articleBody" to <div> so we can format with CSS
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
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