National GEographic by gagsays

recipes/natgeo.recipe

@@ -0,0 +1,71 @@
+#!/usr/bin/env  python
+# -*- coding: utf-8 -*-
+__license__   = 'GPL v3'
+__copyright__ = '2011, gagsays <gagsays at gmail dot com>'
+'''
+nationalgeographic.com
+'''
+from calibre.web.feeds.news import BasicNewsRecipe
+class NatGeo(BasicNewsRecipe):
+    title          = u'National Geographic'
+    description = 'Daily news articles from The National Geographic'
+    language = 'en'
+    oldest_article = 20
+    max_articles_per_feed = 25
+    encoding              = 'utf8'
+    publisher              = 'nationalgeographic.com'
+    category               = 'science, nat geo'
+    __author__           = 'gagsays'
+    masthead_url        = 'http://s.ngeo.com/wpf/sites/themes/global/i/presentation/ng_logo_small.png'
+    description           = 'Inspiring people to care about the planet since 1888'
+    timefmt = ' [%a, %d %b, %Y]'
+    no_stylesheets        = True
+    use_embedded_content  = False
+
+    extra_css = '''
+                body {color: #000000;font-size: medium;}
+                h1 {color: #222222; font-size: large; font-weight:lighter; text-decoration:none; text-align: center;font-family:Georgia,Times New Roman,Times,serif;}
+          h2 {color: #454545; font-size: small; font-weight:lighter; text-decoration:none; text-align: justify; font-style:italic;font-family :Georgia,Times New Roman,Times,serif;}
+                h3 {color: #555555; font-size: small; font-style:italic; margin-top: 10px;}
+                img{margin-bottom: 0.25em;display:block;margin-left: auto;margin-right: auto;}
+                a:link,a,.a,href {text-decoration: none;color: #000000;}
+                .caption{color: #000000;font-size: xx-small;text-align: justify;font-weight:normal;}
+                .credit{color: #555555;font-size: xx-small;text-align: left;font-weight:lighter;}
+                p.author,p.publication{color: #000000;font-size: xx-small;text-align: left;display:inline;}
+                p.publication_time{color: #000000;font-size: xx-small;text-align: right;text-decoration: underline;}
+                p {margin-bottom: 0;}
+                p + p {text-indent: 1.5em;margin-top: 0;}
+                .hidden{display:none;}
+                #page_head{text-transform:uppercase;}
+               '''
+
+    def parse_feeds (self):
+      feeds = BasicNewsRecipe.parse_feeds(self)
+      for feed in feeds:
+        for article in feed.articles[:]:
+          if 'Presented' in article.title or 'Pictures' in article.title:
+            feed.articles.remove(article)
+      return feeds
+
+    def preprocess_html(self, soup):
+        for alink in soup.findAll('a'):
+            if alink.string is not None:
+               tstr = alink.string
+               alink.replaceWith(tstr)
+        return soup
+
+    remove_tags_before = dict(id='page_head')
+    keep_only_tags = [
+         dict(name='div',attrs={'id':['page_head','content_mainA']})
+    ]
+    remove_tags_after = [
+         dict(name='div',attrs={'class':['article_text','promo_collection']})
+    ]
+    remove_tags    = [
+                               dict(name='div', attrs={'class':['aside','primary full_width']})
+                               ,dict(name='div', attrs={'id':['header_search','navigation_mainB_wrap']})
+                                ]
+    feeds = [
+                    (u'Daily News', u'http://feeds.nationalgeographic.com/ng/News/News_Main')
+                    ]
+