import re
from calibre.web.feeds.news import BasicNewsRecipe

class TimesOfIndia(BasicNewsRecipe):
    title          = u'Times of India'
    language       = 'en_IN'
    __author__     = 'Kovid Goyal'
    oldest_article = 1 #days
    max_articles_per_feed = 25

    no_stylesheets = True
    keep_only_tags = [{'class':['maintable12', 'prttabl']}]
    remove_tags = [
            dict(style=lambda x: x and 'float' in x),
            {'class':['prvnxtbg', 'footbdrin', 'bcclftr']},
    ]

    feeds          = [
('Top Stories',
 'http://timesofindia.indiatimes.com/rssfeedstopstories.cms'),
('India',
 'http://timesofindia.indiatimes.com/rssfeeds/-2128936835.cms'),
('World',
 'http://timesofindia.indiatimes.com/rssfeeds/296589292.cms'),
('Mumbai',
 'http://timesofindia.indiatimes.com/rssfeeds/-2128838597.cms'),
('Entertainment',
 'http://timesofindia.indiatimes.com/rssfeeds/1081479906.cms'),
('Cricket',
 'http://timesofindia.indiatimes.com/rssfeeds/4719161.cms'),
('Sunday TOI',
 'http://timesofindia.indiatimes.com/rssfeeds/1945062111.cms'),
('Life and Style',
 'http://timesofindia.indiatimes.com/rssfeeds/2886704.cms'),
('Business',
 'http://timesofindia.indiatimes.com/rssfeeds/1898055.cms'),
('Mad Mad World',
 'http://timesofindia.indiatimes.com/rssfeeds/2178430.cms'),
('Most Read',
 'http://timesofindia.indiatimes.com/rssfeedmostread.cms')
]

    def get_article_url(self, article):
        url = BasicNewsRecipe.get_article_url(self, article)
        if '/0Ltimesofindia' in url:
            url = url.partition('/0L')[-1]
            url = url.replace('0B', '.').replace('0N', '.com').replace('0C',
                    '/').replace('0E', '-')
            url = 'http://' + url.rpartition('/')[0]
            match = re.search(r'/([0-9a-zA-Z]+?)\.cms', url)
            if match is not None:
                num = match.group(1)
                num = re.sub(r'[^0-9]', '', num)
                return ('http://timesofindia.indiatimes.com/articleshow/%s.cms?prtpage=1' %
                    num)
        else:
            cms = re.search(r'/(\d+)\.cms', url)
            if cms is not None:
                return ('http://timesofindia.indiatimes.com/articleshow/%s.cms?prtpage=1' %
                    cms.group(1))

        return url


    def preprocess_html(self, soup):
        return soup