From 7f9adca2d633c171f83809469dd04e062cf1cc8d Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Thu, 6 Jan 2011 09:35:50 -0700 Subject: [PATCH] The Baltimore Sun by Josh Hall --- resources/recipes/baltimore_sun.recipe | 186 +++++++++++++++++++++++++ 1 file changed, 186 insertions(+) create mode 100644 resources/recipes/baltimore_sun.recipe diff --git a/resources/recipes/baltimore_sun.recipe b/resources/recipes/baltimore_sun.recipe new file mode 100644 index 0000000000..ac6906a5e6 --- /dev/null +++ b/resources/recipes/baltimore_sun.recipe @@ -0,0 +1,186 @@ +from __future__ import with_statement +__license__ = 'GPL 3' +__copyright__ = 'Original 2009, Kovid Goyal ' +__copyright__= 'Modified 2011, Josh Hall ' +__docformat__ = 'restructuredtext en' + +''' +www.baltimoresun.com +''' + +from calibre.web.feeds.news import BasicNewsRecipe + +class BaltimoreSun(BasicNewsRecipe): + + title = 'The Baltimore Sun' + __author__ = 'Josh Hall' + description = 'Politics, local and business news from Baltimore' + language = 'en' + oldest_article = 1 + max_articles_per_feed = 100 + remove_empty_feeds = True + use_embedded_content = False + no_stylesheets = True + remove_javascript = True + #masthead_url = 'http://www.baltimoresun.com/images/thirdpartylogo.gif' + + remove_tags_before = dict(name='div', attrs={'class':['story', 'entry']}) + remove_tags_after = [ + {'class':['photo_article',]}, + dict(name='div', attrs={'class':'shirttail-promo right clearfix'}), + ] + + keep_only_tags = [dict(name='div', attrs={'class':["story","entry-asset asset hentry"]}), + dict(name='div', attrs={'id':["pagebody","story","maincontentcontainer"]}), + ] + + + remove_tags = [{'id':["moduleArticleTools","content-bottom","rail","articleRelates module","toolSet","relatedrailcontent","div-wrapper","beta","atp-comments","footer","article-promo"]}, + {'class':["entry-footer-left","entry-footer-right","shirttail-promo right clearfix","clearfix","relatedTitle","articleRelates module","asset-footer","tools","comments","featurePromo","featurePromo fp-topjobs brownBackground","clearfix fullSpan brownBackground","curvedContent","toppaginate","module","module-header","module-content"]}, + dict(name='font',attrs={'id':["cr-other-headlines"]}), + dict(name=['iframe']), + ] + extra_css = ''' + h1{font-family:Arial,Helvetica,sans-serif; font-weight:bold;font-size:large;} + h2{font-family:Arial,Helvetica,sans-serif; font-weight:normal;font-size:small;} + .byline {font-family:Arial,Helvetica,sans-serif; font-size:xx-small;} + .date {font-family:Arial,Helvetica,sans-serif; font-size:xx-small;} + p{font-family:Arial,Helvetica,sans-serif;font-size:small;} + .copyright {font-family:Arial,Helvetica,sans-serif;font-size:xx-small;text-align:center} + .story{font-family:Arial,Helvetica,sans-serif;font-size:small;} + .entry-asset asset hentry{font-family:Arial,Helvetica,sans-serif;font-size:small;} + .pagebody{font-family:Arial,Helvetica,sans-serif;font-size:small;} + .maincontentcontainer{font-family:Arial,Helvetica,sans-serif;font-size:small;} + .story-body{font-family:Arial,Helvetica,sans-serif;font-size:small;} + body{font-family:Helvetica,Arial,sans-serif;font-size:small;} + ''' + feeds = [ + (u'Top Headlines', u'http://www.baltimoresun.com/rss2.0.xml'), + (u'Breaking News', u'http://www.baltimoresun.com/news/breaking/rss2.0.xml'), + (u'Top Maryland', u'http://www.baltimoresun.com/news/maryland/rss2.0.xml'), + #(u'Anne Arundel County', u'http://www.baltimoresun.com/news/maryland/anne-arundel/rss2.0.xml'), + (u'Baltimore City', u'http://www.baltimoresun.com/news/maryland/baltimore-city/rss2.0.xml'), + #(u'Baltimore County', u'http://www.baltimoresun.com/news/maryland/baltimore-county/rss2.0.xml'), + #(u'Carroll County', u'http://www.baltimoresun.com/news/maryland/carroll/rss2.0.xml'), + #(u'Harford County', u'http://www.baltimoresun.com/news/maryland/harford/rss2.0.xml'), + #(u'Howard County', u'http://www.baltimoresun.com/news/maryland/howard/rss2.0.xml'), + (u'Education', u'http://www.baltimoresun.com/news/education/rss2.0.xml'), + #(u'Obituaries', u'http://www.baltimoresun.com/news/obituaries/rss2.0.xml'), + (u'Local Politics', u'http://www.baltimoresun.com/news/maryland/politics/rss2.0.xml'), + (u'Weather', u'http://www.baltimoresun.com/news/weather/rss2.0.xml'), + #(u'Traffic', u'http://www.baltimoresun.com/features/commuting/rss2.0.xml'), + (u'Nation/world', u'http://feeds.chicagotribune.com/chicagotribune/news/nationworld/'), + (u'Weird News', u'http://www.baltimoresun.com/news/offbeat/rss2.0.xml'), + + + (u'Top Sports', u'http://www.baltimoresun.com/sports/rss2.0.xml'), + (u'Orioles/Baseball', u'http://www.baltimoresun.com/sports/orioles/rss2.0.xml'), + (u'Ravens/Football', u'http://www.baltimoresun.com/sports/ravens/rss2.0.xml'), + #(u'Terps', u'http://www.baltimoresun.com/sports/terps/rss2.0.xml'), + #(u'College Football', u'http://www.baltimoresun.com/sports/college/football/rss2.0.xml'), + #(u'Lacrosse', u'http://www.baltimoresun.com/sports/college/lacrosse/rss2.0.xml'), + #(u'Horse Racing', u'http://www.baltimoresun.com/sports/horse-racing/rss2.0.xml'), + #(u'Golf', u'http://www.baltimoresun.com/sports/golf/rss2.0.xml'), + #(u'NBA', u'http://www.baltimoresun.com/sports/nba/rss2.0.xml'), + #(u'High School', u'http://www.baltimoresun.com/sports/high-school/rss2.0.xml'), + #(u'Outdoors', u'http://www.baltimoresun.com/sports/outdoors/rss2.0.xml'), + + (u'Celebrity News', u'http://www.baltimoresun.com/entertainment/celebrities/rss2.0.xml'), + (u'Arts & Theater', u'http://www.baltimoresun.com/entertainment/arts/rss2.0.xml'), + (u'Movies', u'http://www.baltimoresun.com/entertainment/movies/rss2.0.xml'), + (u'Music & Nightlife', u'http://www.baltimoresun.com/entertainment/music/rss2.0.xml'), + (u'Restaurants & Food', u'http://www.baltimoresun.com/entertainment/dining/rss2.0.xml'), + (u'TV/Media', u'http://www.baltimoresun.com/entertainment/tv/rss2.0.xml'), + + (u'Health&Wellness', u'http://www.baltimoresun.com/health/rss2.0.xml'), + (u'Home & Garden', u'http://www.baltimoresun.com/features/home-garden/rss2.0.xml'), + (u'Living Green', u'http://www.baltimoresun.com/features/green/rss2.0.xml'), + (u'Parenting', u'http://www.baltimoresun.com/features/parenting/rss2.0.xml'), + (u'Fashion', u'http://www.baltimoresun.com/features/fashion/rss2.0.xml'), + (u'Travel', u'http://www.baltimoresun.com/travel/rss2.0.xml'), + (u'Faith', u'http://www.baltimoresun.com/features/faith/rss2.0.xml'), + + (u'Top Business', u'http://www.baltimoresun.com/business/rss2.0.xml'), + (u'Technology', u'http://www.baltimoresun.com/business/technology/rss2.0.xml'), + (u'Personal finance', u'http://www.baltimoresun.com/business/money/rss2.0.xml'), + (u'Real Estate', u'http://www.baltimoresun.com/classified/realestate/rss2.0.xml'), + (u'Jobs', u'http://www.baltimoresun.com/classified/jobs/rss2.0.xml'), + (u'DIY', u'http://www.baltimoresun.com/features/do-it-yourself/rss2.0.xml'), + (u'Consumer Safety', u'http://www.baltimoresun.com/business/consumer-safety/rss2.0.xml'), + (u'Investing', u'http://www.baltimoresun.com/business/money/rss2.0.xml'), + + (u'Sun Editorials', u'http://www.baltimoresun.com/news/opinion/editorial/rss2.0.xml'), + (u'Op/Ed', u'http://www.baltimoresun.com/news/opinion/oped/rss2.0.xml'), + (u'Readers Respond', u'http://www.baltimoresun.com/news/opinion/readersrespond/'), + + (u'Kevin Cowherd', 'http://www.baltimoresun.com/sports/bal-columnist-cowherd,0,6829726.columnist-rss2.0.xml'), + (u'Jay Hancock', u'http://www.baltimoresun.com/business/money/bal-columnist-hancock,0,6673611.columnist-rss2.0.xml'), + (u'Jacques Kelly', u'http://www.baltimoresun.com/news/maryland/bal-columnist-kelly,0,1154701.columnist-rss2.0.xml'), + (u'Marta H. Mossburg', u'http://www.baltimoresun.com/news/opinion/oped/bal-columnist-mossburg,0,7982155.columnist-rss2.0.xml'), + (u'Mike Preston', u'http://www.baltimoresun.com/sports/bal-columnist-preston,0,6169796.columnist-rss2.0.xml'), + (u'Susan Reimer', u'http://www.baltimoresun.com/news/opinion/bal-columnist-reimer,0,162466.columnist-rss2.0.xml'), + (u'Dan Rodricks', u'http://www.baltimoresun.com/news/maryland/bal-columnist-rodricks,0,7089843.columnist-rss2.0.xml'), + (u'Thomas F. Schaller', u'http://www.baltimoresun.com/news/opinion/columnists/bal-columnist-schaller,0,897397.columnist-rss2.0.xml'), + (u'Peter Schmuck', u'http://www.baltimoresun.com/sports/bal-columnist-schmuck,0,7485088.columnist-rss2.0.xml'), + (u'Ron Smith', u'http://www.baltimoresun.com/news/opinion/bal-columnist-ronsmith,0,3964803.columnist-rss2.0.xml'), + + (u'Baltimore Crime Beat', u'http://weblogs.baltimoresun.com/news/crime/blog/index.xml'), + (u'Getting There', u'http://weblogs.baltimoresun.com/news/traffic/index.xml'), + (u'InsideEd', u'http://weblogs.baltimoresun.com/news/education/blog/index.xml'), + (u'Maryland Politics', u'http://weblogs.baltimoresun.com/news/local/politics/index.xml'), + (u'Maryland Weather', u'http://weblogs.marylandweather.com/index.xml'), + (u'Second Opinion', u'http://weblogs.baltimoresun.com/news/opinion/index.xml'), + (u'You Dont Say', u'http://weblogs.baltimoresun.com/news/mcintyre/blog/index.xml'), + + (u'BaltTech', u'http://weblogs.baltimoresun.com/news/technology/index.xml'), + (u'Consuming Interests', u'http://weblogs.baltimoresun.com/business/consuminginterests/blog/index.xml'), + (u'Jay Hancocks Blog', u'http://weblogs.baltimoresun.com/business/hancock/blog/index.xml'), + (u'The Real Estate Wonk', u'http://weblogs.baltimoresun.com/business/realestate/blog/index.xml'), + + (u'Clef Notes', 'http://weblogs.baltimoresun.com/entertainment/classicalmusic/index.xml'), + (u'Dining at Large', u'http://weblogs.baltimoresun.com/entertainment/dining/reviews/blog/index.xml'), + (u'Midnight Sun', u'http://weblogs.baltimoresun.com/entertainment/midnight_sun/blog/index.xml'), + (u'Mike Sragow Gets Reel', u'http://weblogs.baltimoresun.com/entertainment/movies/blog/index.xml'), + (u'Read Street', u'http://weblogs.baltimoresun.com/entertainment/books/blog/index.xml'), + (u'Reality Check', u'http://weblogs.baltimoresun.com/entertainment/realitycheck/blog/index.xml'), + (u'Z on TV', u'http://weblogs.baltimoresun.com/entertainment/zontv/index.xml'), + + (u'BMore Green', u'http://weblogs.baltimoresun.com/features/green/index.xml'), + (u'Charm City Moms', u'http://weblogs.baltimoresun.com/features/baltimoremomblog/index.xml'), + (u'Exercists', u'http://weblogs.baltimoresun.com/health/fitness/index.xml'), + (u'Garden Variety', 'http://weblogs.baltimoresun.com/features/gardening/index.xml'), + #(u'In Good Faith', u'http://weblogs.baltimoresun.com/news/faith/index.xml'), + (u'Picture of Health', u'http://weblogs.baltimoresun.com/health/index.xml'), + (u'Unleashed', u'http://weblogs.baltimoresun.com/features/mutts/blog/index.xml'), + + #(u'Faceoff', u'http://weblogs.baltimoresun.com/sports/lacrosse/blog/index.xml'), + #(u'MMA Stomping Grounds', u'http://weblogs.baltimoresun.com/sports/mma/blog/index.xml'), + (u'Orioles Insider', u'http://weblogs.baltimoresun.com/sports/orioles/blog/index.xml'), + #(u'Outdoors Girl', u'http://weblogs.baltimoresun.com/sports/outdoors/blog/index.xml'), + (u'Ravens Insider', u'http://weblogs.baltimoresun.com/sports/ravens/blog/index.xml'), + #(u'Recruiting Report', u'http://weblogs.baltimoresun.com/sports/college/recruiting/index.xml'), + #(u'Ring Posts', u'http://weblogs.baltimoresun.com/sports/wrestling/blog/index.xml'), + (u'The Schmuck Stops Here', u'http://weblogs.baltimoresun.com/sports/schmuck/index.xml'), + (u'Toy Department', u'http://weblogs.baltimoresun.com/sports/thetoydepartment/index.xml'), + #(u'Tracking the Terps', u'http://weblogs.baltimoresun.com/sports/college/maryland_terps/blog/index.xml'), + #(u'Varsity Letters', u'http://weblogs.baltimoresun.com/sports/highschool/varsityletters/index.xml'), + (u'Virtual Vensanity', u'http://weblogs.baltimoresun.com/entertainment/bthesite/vensel/index.xml'), + + ] + + + def get_article_url(self, article): + print article.get('feedburner_origlink', article.get('guid', article.get('link'))) + return article.get('feedburner_origlink', article.get('guid', article.get('link'))) + + + def postprocess_html(self, soup, first_fetch): + for t in soup.findAll(['table', 'tr', 'td']): + t.name = 'div' + + for tag in soup.findAll('form', dict(attrs={'name':["comments_form"]})): + tag.extract() + for tag in soup.findAll('font', dict(attrs={'id':["cr-other-headlines"]})): + tag.extract() + + return soup