From bd51d08f8823c7a1ff08e052f403dfce5c64c39b Mon Sep 17 00:00:00 2001 From: Kovid Goyal Date: Sun, 6 Dec 2009 13:41:29 -0700 Subject: [PATCH] New recipes for The Huffington Post and Glenn Beck by kwetal --- resources/recipes/glennbeck.recipe | 97 +++++++++++++++++ resources/recipes/huffingtonpost.recipe | 134 ++++++++++++++++++++++++ 2 files changed, 231 insertions(+) create mode 100644 resources/recipes/glennbeck.recipe create mode 100644 resources/recipes/huffingtonpost.recipe diff --git a/resources/recipes/glennbeck.recipe b/resources/recipes/glennbeck.recipe new file mode 100644 index 0000000000..09f54b2d6f --- /dev/null +++ b/resources/recipes/glennbeck.recipe @@ -0,0 +1,97 @@ +from calibre.web.feeds.news import BasicNewsRecipe +from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, Comment + +class GlennBeckRecipe(BasicNewsRecipe): + __license__ = 'GPL v3' + __author__ = 'kwetal' + language = 'en' + version = 1 + + title = u'Glenn Beck' + publisher = u'Premiere Radio Networks' + category = u'News, Opinion' + description = u'The fusion of entertainment and enlightenment' + + oldest_article = 7 + max_articles_per_feed = 100 + + no_stylesheets = True + remove_javascript = True + use_embedded_content = False + + feeds = [(u'Glenn Beck', u'http://feeds.feedburner.com/GlennBeckArticles')] + + def preprocess_html(self, soup): + # Their html is horribly broken; if we search for the div that has the content BeatifulSoup returns the div with only the headline and no content. + # This is due to illegal nesting of tags. So we do it the hard way. + + # We can find this one, and we don't want it. + div = soup.find('div', attrs = {'id': 'extraInfo'}) + if div: + div.extract() + + # Don't want these either. + iframes = soup.findAll('iframe') + [iframe.extract() for iframe in iframes] + + # Get empty document. + freshSoup = self.getFreshSoup() + + # This is the broken div; but we can find the headline. + newsDiv = soup.find('div', attrs = {'class': 'news-detail'}) + if newsDiv: + if newsDiv.h1: + freshSoup.body.append(newsDiv.h1) + + # The content is wrapped in

tags, most of the time anyway. + counter = 0 + for p in soup.findAll('p'): + if p.get('class') == 'smalltextwhite': + # But we don't want this one. + continue + + freshSoup.body.append(p) + counter += 1 + + # Debugging block + #h3 = Tag(freshSoup, 'h3') + #h3.append('First counter: ' + str(counter)) + #freshSoup.body.insert(0, h3) + + # In some articles the content is not wrapped in

tags. In that case the counter is low. + # 2 is the magic number that seems to work. + if counter <= 2: + # So they are playing hard-to-get: first throw out all comments. + comments = soup.findAll(text = lambda text: isinstance(text, Comment)) + [comment.extract() for comment in comments] + + # Find all unwrapped strings. + for txt in soup.findAll(text = True): + raw = txt.strip() + # Debugging line + #para.append(raw + '(parent: ' + txt.parent.name + '; length: ' + str(len(raw)) + '; start: ' + raw[0:4] + ')') + + if (txt.parent.name == 'body' and len(raw) > 0) and not (len(raw) == 6 and raw == ' '): + # This is our content; ignore the rest. + para = Tag(freshSoup, 'p') + para.append(raw) + freshSoup.body.append(para) + counter += 1 + + # Now if the counter is still 0 or 1 they did something completely different and we still have an empty article. In a last attempt, add the whole content div, just in case. + if counter < 2: + freshSoup.body.append(newsDiv) + + # Debugging block + #h3 = Tag(freshSoup, 'h3') + #h3.append('Second counter: ' + str(counter)) + #freshSoup.body.insert(1, h3) + + return freshSoup + + def getFreshSoup(self, title = None): + if title: + return BeautifulSoup('' + str(title) + '') + else: + return BeautifulSoup('') + diff --git a/resources/recipes/huffingtonpost.recipe b/resources/recipes/huffingtonpost.recipe new file mode 100644 index 0000000000..fc5ba26f1c --- /dev/null +++ b/resources/recipes/huffingtonpost.recipe @@ -0,0 +1,134 @@ +from calibre.web.feeds.news import BasicNewsRecipe +from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag, Comment +import re + +class HuffingtonPostRecipe(BasicNewsRecipe): + __license__ = 'GPL v3' + __author__ = 'kwetal' + language = 'en' + version = 1 + + title = u'The Huffington Post' + publisher = u'huffingtonpost.com' + category = u'News, Politics' + description = u'Political Blog' + + oldest_article = 1.5 + max_articles_per_feed = 100 + use_embedded_content = False + + no_stylesheets = True + remove_javascript = True + # Seems to work best, but YMMV + simultaneous_downloads = 1 + + feeds = [] + feeds.append((u'Latest News', u'http://feeds.huffingtonpost.com/huffingtonpost/LatestNews')) + # Works, but appears to be a subset of the politics-blog feed + #feeds.append((u'Politics', u'http://www.huffingtonpost.com/feeds/verticals/politics/index.xml')) + feeds.append((u'Politics', u'http://www.huffingtonpost.com/feeds/verticals/politics/blog.xml')) + # Does not work + #feeds.append((u'Politics: News', u'http://www.huffingtonpost.com/feeds/verticals/politics/news.xml')) + # Works, but appears to be a subset of the media-blog feed + #feeds.append((u'Media', u'http://www.huffingtonpost.com/feeds/verticals/media/index.xml')) + feeds.append((u'Media', u'http://www.huffingtonpost.com/feeds/verticals/media/blog.xml')) + # Does not work + #feeds.append((u'Media: News', u'http://www.huffingtonpost.com/feeds/verticals/media/news.xml')) + # Works, but appears to be a subset of the business-blog feed + #feeds.append((u'Business', u'http://www.huffingtonpost.com/feeds/verticals/business/index.xml')) + feeds.append((u'Business', u'http://www.huffingtonpost.com/feeds/verticals/business/blog.xml')) + # Does not work + #feeds.append((u'Business: News', u'http://www.huffingtonpost.com/feeds/verticals/business/news.xml')) + feeds.append((u'Original Reporting', u'http://www.huffingtonpost.com/tag/huffpolitics/feed')) + feeds.append((u'Original Posts', u'http://www.huffingtonpost.com/feeds/original_posts/index.xml')) + + keep_only_tags = [] + # For reporters posts + keep_only_tags.append(dict(name = 'div', attrs = {'class' : 'about_reporter_name'})) + keep_only_tags.append(dict(name = 'div', attrs = {'class' : 'entry'})) + # For blog posts + keep_only_tags.append(dict(name = 'div', attrs = {'id' : 'blog_author_info'})) + keep_only_tags.append(dict(name = 'div', attrs = {'id' : 'blog_title'})) + + remove_tags = [] + remove_tags.append(dict(name = 'div', attrs = {'class' : 'contin_below'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'adver_cont_below'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'blogger_menu_content'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'chicklets'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'google-searcG-blogp'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'forma_email'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'new_share_module'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'cse-branding-right'})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'clear'})) + remove_tags.append(dict(name = 'div', attrs = {'style' : re.compile('clear:both;*')})) + remove_tags.append(dict(name = 'div', attrs = {'class' : re.compile('facebookvote_reaction_blog.*')})) + remove_tags.append(dict(name = 'div', attrs = {'class' : re.compile('read_more.*')})) + remove_tags.append(dict(name = 'div', attrs = {'class' : re.compile('facebookvote_v2.*')})) + remove_tags.append(dict(name = 'div', attrs = {'class' : re.compile('facebookvote_reaction.*')})) + remove_tags.append(dict(name = 'div', attrs = {'class' : 'slideshow_poll'})) + + remove_attributes = ['style'] + + extra_css = 'a[href] {color: blue; text-decoration: none; cursor: pointer;}' + + def preprocess_html(self, soup): + # Condens the soup. + soup = self.unComment(self.nukeHead(soup)) + + # Don't want the picture of the author + blogAuthor = soup.find('div', attrs = {'id': 'blog_author_info'}) + if blogAuthor: + for img in blogAuthor.findAll('img'): + img.extract() + + byline = soup.find('h2') + if byline: + h2 = Tag(soup, 'h2') + raw = self.tag_to_string(byline) + h2.append(raw) + byline.replaceWith(h2) + else: + byline = soup.find('div', attrs = {'class': re.compile('about_*reporter_*name')}) + if byline: + h2 = Tag(soup, 'h2') + raw = self.tag_to_string(byline) + h2.append(raw.strip()) + byline.replaceWith(h2) + + headline = soup.find('h1') + if headline: + h1 = Tag(soup, 'h1') + raw = self.tag_to_string(headline) + h1.append(raw) + headline.replaceWith(h1) + + return soup + + def postprocess_html(self, soup, first): + # Get rid of those pesky
tags + html = re.sub(r'\n
\n', '', str(soup)) + newSoup = BeautifulSoup(html) + + return newSoup + + def nukeHead(self, soup): + titleStr = '' + newHead = Tag(soup, 'head') + newTitle = Tag(soup, 'title') + newHead.append(newTitle) + head = soup.head + if head: + title = head.title + if title: + titleStr = self.tag_to_string(title) + newTitle.append(titleStr) + head.replaceWith(newHead) + else: + soup.insert(0, newHead) + return soup + + def unComment(self, soup): + comments = soup.findAll(text = lambda text: isinstance(text, Comment)) + [comment.extract() for comment in comments] + return soup +