arabian_business.recipe

__license__ = 'GPL v3'
__copyright__ = '2011, Darko Miletic <darko.miletic at gmail.com>'
'''
www.arabianbusiness.com
'''

from calibre.web.feeds.news import BasicNewsRecipe


class Arabian_Business(BasicNewsRecipe):
    title = 'Arabian Business'
    __author__ = 'Darko Miletic'
    description = 'Comprehensive Guide to Middle East Business & Gulf Industry News including,Banking & Finance,Construction,Energy,Media & Marketing,Real Estate,Transportation,Travel,Technology,Politics,Healthcare,Lifestyle,Jobs & UAE guide.Top Gulf & Dubai Business News.'  # noqa
    publisher = 'Arabian Business Publishing Ltd.'
    oldest_article = 2
    max_articles_per_feed = 200
    no_stylesheets = True
    encoding = 'utf8'
    use_embedded_content = False
    language = 'en'
    remove_empty_feeds = True
    publication_type = 'newsportal'
    masthead_url = 'http://www.arabianbusiness.com/skins/ab.main/gfx/arabianbusiness_logo_sm.gif'
    extra_css             = """
                               body{font-family: Georgia,serif }
                               img{margin-bottom: 0.4em; margin-top: 0.4em; display:block}
                               .byline,.dateline{font-size: small; display: inline; font-weight: bold}
                               ul{list-style: none outside none;}
                            """

    conversion_options = {
        'comment': description, 'publisher': publisher, 'language': language
    }

    remove_tags_before = dict(attrs={'id': 'article-title'})
    remove_tags = [
        dict(name=['meta', 'link', 'base', 'iframe', 'embed', 'object']), dict(
            attrs={'class': 'printfooter'})
    ]
    remove_attributes = ['lang']

    feeds = [

    (u'Africa', u'http://www.arabianbusiness.com/world/Africa/?service=rss'),
    (u'Americas', u'http://www.arabianbusiness.com/world/americas/?service=rss'),
    (u'Asia Pacific', u'http://www.arabianbusiness.com/world/asia-pacific/?service=rss'),
    (u'Europe', u'http://www.arabianbusiness.com/world/europe/?service=rss'),
    (u'Middle East', u'http://www.arabianbusiness.com/world/middle-east/?service=rss'),
    (u'South Asia', u'http://www.arabianbusiness.com/world/south-asia/?service=rss'),
    (u'Banking & Finance', u'http://www.arabianbusiness.com/industries/banking-finance/?service=rss'),
    (u'Construction', u'http://www.arabianbusiness.com/industries/construction/?service=rss'),
    (u'Education', u'http://www.arabianbusiness.com/industries/education/?service=rss'),
    (u'Energy', u'http://www.arabianbusiness.com/industries/energy/?service=rss'),
    (u'Healthcare', u'http://www.arabianbusiness.com/industries/healthcare/?service=rss'),
    (u'Media', u'http://www.arabianbusiness.com/industries/media/?service=rss'),
    (u'Real Estate', u'http://www.arabianbusiness.com/industries/real-estate/?service=rss'),
    (u'Retail', u'http://www.arabianbusiness.com/industries/retail/?service=rss'),
    (u'Technology', u'http://www.arabianbusiness.com/industries/technology/?service=rss'),
    (u'Transport', u'http://www.arabianbusiness.com/industries/transport/?service=rss'),
    (u'Travel', u'http://www.arabianbusiness.com/industries/travel-hospitality/?service=rss'),
    (u'Equities', u'http://www.arabianbusiness.com/markets/equities/?service=rss'),
    (u'Commodities', u'http://www.arabianbusiness.com/markets/commodities/?service=rss'),
    (u'Currencies', u'http://www.arabianbusiness.com/markets/currencies/?service=rss'),
    (u'Market Data', u'http://www.arabianbusiness.com/markets/market-data/?service=rss'),
    (u'Comment', u'http://www.arabianbusiness.com/opinion/comment/?service=rss'),
    (u'Think Tank', u'http://www.arabianbusiness.com/opinion/think-tank/?service=rss'),
    (u'Arts', u'http://www.arabianbusiness.com/lifestyle/arts/?service=rss'),
    (u'Cars', u'http://www.arabianbusiness.com/lifestyle/cars/?service=rss'),
    (u'Food', u'http://www.arabianbusiness.com/lifestyle/food/?service=rss'),
    (u'Sport', u'http://www.arabianbusiness.com/lifestyle/sport/?service=rss')
    ]

    def print_version(self, url):
        return url + '?service=printer&page='

    def preprocess_html(self, soup):
        for item in soup.findAll(style=True):
            del item['style']
        for alink in soup.findAll('a'):
            if alink.string is not None:
                tstr = alink.string
                alink.replaceWith(tstr)
        return soup