Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Improved Suedduetsche Zeitung

Browse Source
This commit is contained in:
Kovid Goyal 2011-04-16 06:46:00 -06:00
parent 6dcc6ae9f1
commit bb75ec2aaf
1 changed files with 28 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

44
recipes/sueddeutschezeitung.recipe
View File

@ -1,4 +1,3 @@
__license__ = 'GPL v3' __license__ = 'GPL v3'
__copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>' __copyright__ = '2010, Darko Miletic <darko.miletic at gmail.com>'
''' '''
@ -19,11 +18,11 @@ class SueddeutcheZeitung(BasicNewsRecipe):
encoding = 'cp1252' encoding = 'cp1252'
needs_subscription = True needs_subscription = True
remove_empty_feeds = True remove_empty_feeds = True
delay = 2 delay = 1
PREFIX = 'http://www.sueddeutsche.de' PREFIX = 'http://www.sueddeutsche.de'
INDEX = PREFIX + '/app/epaper/textversion/' INDEX = PREFIX + '/app/epaper/textversion/'
use_embedded_content = False use_embedded_content = False
masthead_url = 'http://pix.sueddeutsche.de/img/layout/header/logo.gif' masthead_url = 'http://pix.sueddeutsche.de/img/layout/header/SZ_solo288x31.gif'
language = 'de' language = 'de'
publication_type = 'newspaper' publication_type = 'newspaper'
extra_css = ' body{font-family: Arial,Helvetica,sans-serif} ' extra_css = ' body{font-family: Arial,Helvetica,sans-serif} '
@ -36,7 +35,7 @@ class SueddeutcheZeitung(BasicNewsRecipe):
, 'linearize_tables' : True , 'linearize_tables' : True
} }
remove_attributes = ['height','width'] remove_attributes = ['height','width','style']
def get_browser(self): def get_browser(self):
br = BasicNewsRecipe.get_browser() br = BasicNewsRecipe.get_browser()
@ -50,24 +49,37 @@ class SueddeutcheZeitung(BasicNewsRecipe):
remove_tags =[ remove_tags =[
dict(attrs={'class':'hidePrint'}) dict(attrs={'class':'hidePrint'})
,dict(name=['link','object','embed','base','iframe']) ,dict(name=['link','object','embed','base','iframe','br'])
] ]
keep_only_tags = [dict(attrs={'class':'artikelBox'})] keep_only_tags = [dict(attrs={'class':'artikelBox'})]
remove_tags_before = dict(attrs={'class':'artikelTitel'}) remove_tags_before = dict(attrs={'class':'artikelTitel'})
remove_tags_after = dict(attrs={'class':'author'}) remove_tags_after = dict(attrs={'class':'author'})
feeds = [ feeds = [
(u'Politik' , INDEX + 'Politik/' ) (u'Politik' , INDEX + 'Politik/' )
,(u'Seite drei' , INDEX + 'Seite+drei/' ) ,(u'Seite drei' , INDEX + 'Seite+drei/' )
,(u'Meinungsseite', INDEX + 'Meinungsseite/') ,(u'Meinungsseite' , INDEX + 'Meinungsseite/')
,(u'Wissen' , INDEX + 'Wissen/' ) ,(u'Wissen' , INDEX + 'Wissen/' )
,(u'Panorama' , INDEX + 'Panorama/' ) ,(u'Panorama' , INDEX + 'Panorama/' )
,(u'Feuilleton' , INDEX + 'Feuilleton/' ) ,(u'Feuilleton' , INDEX + 'Feuilleton/' )
,(u'Medien' , INDEX + 'Medien/' ) ,(u'Medien' , INDEX + 'Medien/' )
,(u'Wirtschaft' , INDEX + 'Wirtschaft/' ) ,(u'Wirtschaft' , INDEX + 'Wirtschaft/' )
,(u'Sport' , INDEX + 'Sport/' ) ,(u'Sport' , INDEX + 'Sport/' )
,(u'Bayern' , INDEX + 'Bayern/' ) ,(u'Bayern' , INDEX + 'Bayern/' )
,(u'Muenchen' , INDEX + 'M%FCnchen/' ) ,(u'Muenchen' , INDEX + 'M%FCnchen/' )
,(u'Muenchen City' , INDEX + 'M%FCnchen+City/' )
,(u'Jetzt.de' , INDEX + 'Jetzt.de/' )
,(u'Reise' , INDEX + 'Reise/' )
,(u'SZ Extra' , INDEX + 'SZ+Extra/' )
,(u'Wochenende' , INDEX + 'SZ+am+Wochenende/' )
,(u'Stellen-Markt' , INDEX + 'Stellen-Markt/')
,(u'Motormarkt' , INDEX + 'Motormarkt/')
,(u'Immobilien-Markt', INDEX + 'Immobilien-Markt/')
,(u'Thema' , INDEX + 'Thema/' )
,(u'Forum' , INDEX + 'Forum/' )
,(u'Leute' , INDEX + 'Leute/' )
,(u'Jugend' , INDEX + 'Jugend/' )
,(u'Beilage' , INDEX + 'Beilage/' )
] ]
def parse_index(self): def parse_index(self):