calibre/recipes/biz_portal.recipe

from calibre.web.feeds.news import BasicNewsRecipe

class AdvancedUserRecipe1283848012(BasicNewsRecipe):
    description   = 'This is a recipe of BizPortal.co.il.'
    cover_url      = 'http://www.bizportal.co.il/shukhahon/images/bizportal.jpg'
    title          = u'BizPortal'
    language              = 'he'
    __author__ = 'marbs'
    extra_css='img {max-width:100%;} body{direction: rtl;},title{direction: rtl; } ,article_description{direction: rtl; }, a.article{direction: rtl; } ,calibre_feed_description{direction: rtl; }'
    simultaneous_downloads = 5
    remove_javascript     = True
    timefmt        = '[%a, %d %b, %Y]'
    remove_empty_feeds = True
    oldest_article = 1
    max_articles_per_feed = 100
    remove_attributes = ['width']
    simultaneous_downloads = 5
  #  keep_only_tags =dict(name='div', attrs={'id':'articleContainer'})
    remove_tags = [dict(name='img', attrs={'scr':['images/bizlogo_nl.gif']})]
    max_articles_per_feed = 100
    #preprocess_regexps = [
  #      (re.compile(r'<p>&nbsp;</p>', re.DOTALL|re.IGNORECASE), lambda match: '')
#        ]


    feeds          = [(u'חדשות שוק ההון', u'http://www.bizportal.co.il/shukhahon/messRssUTF2.xml'),
                           (u'חדשות וול סטריט בעברית', u'http://www.bizportal.co.il/shukhahon/images/bizportal.jpg'),
                           (u'שיווק ופרסום', u'http://www.bizportal.co.il/shukhahon/messRssUTF145.xml'),
                           (u'משפט', u'http://www.bizportal.co.il/shukhahon/messRssUTF3.xml'),
                           (u'ניתוח טכני', u'http://www.bizportal.co.il/shukhahon/messRssUTF5.xml'),
                           (u'דיני עבודה ושכר', u'http://www.bizportal.co.il/shukhahon/messRssUTF6.xml'),
                           (u'מיסוי', u'http://www.bizportal.co.il/shukhahon/messRssUTF7.xml'),
                           (u'טאבו', u'http://www.bizportal.co.il/shukhahon/messRssUTF8.xml'),
                           (u'נדל"ן', u'http://www.bizportal.co.il/shukhahon/messRssUTF160.xml'),
                              ]

    def print_version(self, url):
        split1 = url.split("=")
        print_url = 'http://www.bizportal.co.il/web/webnew/shukhahon/biznews02print.shtml?mid=' + split1[1]
        return print_url