New recipe for Press by Darko Miletic

src/calibre/ebooks/epub/from_html.py

@@ -310,6 +310,8 @@ def create_cover_image(src, dest, screen_size, rescale_cover=True):
 def process_title_page(mi, filelist, htmlfilemap, opts, tdir):
     old_title_page = None
     f = lambda x : os.path.normcase(os.path.normpath(x))
+    if not isinstance(mi.cover, basestring):
+        mi.cover = None
     if mi.cover:
         if f(filelist[0].path) == f(mi.cover):
             old_title_page = htmlfilemap[filelist[0].path]

src/calibre/web/feeds/recipes/__init__.py

@@ -29,7 +29,7 @@ recipe_modules = ['recipe_' + r for r in (
            'jb_online', 'estadao', 'o_globo', 'vijesti', 'elmundo', 'the_oz',
            'honoluluadvertiser', 'starbulletin', 'exiled', 'indy_star', 'dna',
            'pobjeda', 'chicago_breaking_news', 'glasgow_herald', 'linuxdevices',
-           'hindu', 'cincinnati_enquirer', 'physics_world',
+           'hindu', 'cincinnati_enquirer', 'physics_world', 'pressonline',
           )]
 
 import re, imp, inspect, time, os

src/calibre/web/feeds/recipes/recipe_pressonline.py

@@ -0,0 +1,67 @@
+#!/usr/bin/env  python
+
+__license__   = 'GPL v3'
+__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
+
+'''
+pressonline.rs
+'''
+
+import re
+from calibre.web.feeds.recipes import BasicNewsRecipe
+
+class PressOnline(BasicNewsRecipe):
+    title                 = 'Press Online'
+    __author__            = 'Darko Miletic'
+    description           = 'Press Online portal dnevnih novina Press.Najnovije vesti iz Srbije i sveta,Sport,Dzet Set,Politika,Hronika,Komenteri,Zabava,Slike,Video,Horoskop,Nagradne igre,Kvizovi,Igrice'
+    publisher             = 'Press Publishing group'
+    category              = 'news, politics, Serbia'    
+    oldest_article        = 2
+    max_articles_per_feed = 100
+    no_stylesheets        = True
+    encoding              = 'utf8'
+    use_embedded_content  = True
+    cover_url             = 'http://www.pressonline.rs/img/logo.gif'
+    language              = _('Serbian')
+
+    extra_css = '@font-face {font-family: "serif1";src:url(res:///opt/sony/ebook/FONT/tt0011m_.ttf)} body{font-family: serif1, serif} .article_description{font-family: serif1, serif}'
+    
+    html2lrf_options = [
+                          '--comment', description
+                        , '--category', category
+                        , '--publisher', publisher
+                        ]
+    
+    html2epub_options = 'publisher="' + publisher + '"\ncomments="' + description + '"\ntags="' + category + '"' 
+     
+    preprocess_regexps = [(re.compile(u'\u0110'), lambda match: u'\u00D0')]
+
+    feeds = [ 
+               (u'Vesti Dana'      , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=37')
+              ,(u'Politika'        , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=29')
+              ,(u'U Fokusu'        , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=33')
+              ,(u'Globus'          , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=40')
+              ,(u'Komentar Dana'   , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=62')
+              ,(u'Hronika'         , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=39')
+              ,(u'Regioni'         , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=56')
+              ,(u'Republika Srpska', u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=51')
+              ,(u'Beograd'         , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=43')
+              ,(u'Dzet-Set Svet'   , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=41')
+              ,(u'Lifestyle'       , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=42')
+              ,(u'Sport'           , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=44')
+              ,(u'Press Magazine'  , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=63')
+              ,(u'Lola'            , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=70')
+              ,(u'Duplerica'       , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=72')
+              ,(u'Presspedia'      , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=80')
+              ,(u'Kolumne'         , u'http://www.pressonline.rs/page/stories/sr.html?view=rss&sectionId=57')              
+            ]
+
+    def preprocess_html(self, soup):
+        soup.html['xml:lang'] = 'sr-Latn-RS'
+        soup.html['lang']     = 'sr-Latn-RS'
+        mtag = '<meta http-equiv="Content-Language" content="sr-Latn-RS"/>\n<meta http-equiv="Content-Type" content="text/html; charset=utf-8">'
+        soup.head.insert(0,mtag)
+        img = soup.find('img')
+        if img:
+           del img['align']
+        return soup