This commit is contained in:
Kovid Goyal 2011-01-12 12:50:02 -07:00
parent f0447ea1cd
commit a9c671efaf

View File

@ -4,7 +4,6 @@ __copyright__ = '2010-2011, Darko Miletic <darko.miletic at gmail.com>'
msnbc.msn.com
'''
import re
from calibre.web.feeds.recipes import BasicNewsRecipe
class MsNBC(BasicNewsRecipe):
@ -19,7 +18,7 @@ class MsNBC(BasicNewsRecipe):
publisher = 'msnbc.com'
category = 'news, USA, world'
language = 'en'
extra_css = """
extra_css = """
body{ font-family: Georgia,Times,serif }
.hide{display: none}
.caption{font-family: Arial,sans-serif; font-size: x-small}
@ -44,7 +43,7 @@ class MsNBC(BasicNewsRecipe):
,dict(attrs={'class':['gl_headline','articleText','drawer-content Linear','v-center3','byline','textBodyBlack']})
]
remove_attributes=['property','lang','rel','xmlns:fb','xmlns:v','xmlns:dc','xmlns:dcmitype','xmlns:og','xmlns:media','xmlns:vcard','typeof','itemscope','itemtype','itemprop','about','type','size','width','height','onreadystatechange','data','border','hspace','vspace']
remove_tags = [
dict(name=['iframe','object','link','embed','meta','table'])
,dict(name='span', attrs={'class':['copyright','Linear copyright']})
@ -70,7 +69,7 @@ class MsNBC(BasicNewsRecipe):
if item.has_key('id') and item['id'].startswith('vine-'):
item.extract()
if item.has_key('class') and ( item['class'].startswith('ad') or item['class'].startswith('vine')):
item.extract()
item.extract()
for item in soup.body.findAll('img'):
if not item.has_key('alt'):
item['alt'] = 'image'
@ -83,6 +82,6 @@ class MsNBC(BasicNewsRecipe):
for alink in soup.findAll('a'):
if alink.string is not None:
tstr = alink.string
alink.replaceWith(tstr)
alink.replaceWith(tstr)
return soup