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

Update Mac World and Maximum PC

Browse Source
This commit is contained in:
Kovid Goyal 2012-09-23 17:10:08 +05:30
parent e719fb48fa
commit 984374825f
2 changed files with 33 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
recipes/mac_world.recipe
View File

@ -34,20 +34,21 @@ class macWorld(BasicNewsRecipe):
remove_javascript = True remove_javascript = True
no_stylesheets = True no_stylesheets = True
auto_cleanup = True
keep_only_tags = [ #keep_only_tags = [
dict(name='div', attrs={'id':'content'}) #dict(name='div', attrs={'id':'content'})
] #]
remove_tags = [ #remove_tags = [
{'class':['toolBar','mac_tags','toolBar btmTools','textAds']}, #{'class':['toolBar','mac_tags','toolBar btmTools','textAds']},
dict(name='p', attrs={'class':'breadcrumbs'}), #dict(name='p', attrs={'class':'breadcrumbs'}),
dict(id=['breadcrumb','sidebar','comments','topContentWrapper', #dict(id=['breadcrumb','sidebar','comments','topContentWrapper',
'rightColumn', 'aboveFootPromo', 'storyCarousel']), #'rightColumn', 'aboveFootPromo', 'storyCarousel']),
{'class':lambda x: x and ('tools' in x or 'toolBar' #{'class':lambda x: x and ('tools' in x or 'toolBar'
in x)} #in x)}
] #]
feeds = [ feeds = [
(u'MacWorld Headlines', u'http://rss.macworld.com/macworld/news'), (u'MacWorld Headlines', u'http://rss.macworld.com/macworld/news'),
@ -82,3 +83,4 @@ class macWorld(BasicNewsRecipe):
.articleInfo {color:#4D4D4D;font-family:Arial,Helvetica,sans-serif;font-size:10px; font-size-adjust:none; font-stretch:normal; font-style:bold; font-variant:normal; font-weight:bold; line-height:10px; text-decoration:none;} .articleInfo {color:#4D4D4D;font-family:Arial,Helvetica,sans-serif;font-size:10px; font-size-adjust:none; font-stretch:normal; font-style:bold; font-variant:normal; font-weight:bold; line-height:10px; text-decoration:none;}
img {align:left;} img {align:left;}
''' '''

41
recipes/maximum_pc.recipe
View File

@ -1,4 +1,3 @@
from calibre.ptempfile import PersistentTemporaryFile
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
class AdvancedUserRecipe1276930924(BasicNewsRecipe): class AdvancedUserRecipe1276930924(BasicNewsRecipe):
@ -14,30 +13,30 @@ class AdvancedUserRecipe1276930924(BasicNewsRecipe):
use_embedded_content = False use_embedded_content = False
no_stylesheets = True no_stylesheets = True
language = 'en' language = 'en'
temp_files = [] auto_cleanup = True
articles_are_obfuscated = True feeds = [#(u'News', u'http://www.maximumpc.com/articles/all/feed'),
feeds = [(u'News', u'http://www.maximumpc.com/articles/4/feed'), (u'News', u'http://www.maximumpc.com/articles/4/feed'),
(u'Reviews', u'http://www.maximumpc.com/articles/40/feed'), (u'Reviews', u'http://www.maximumpc.com/articles/40/feed'),
(u'Editors Blog', u'http://www.maximumpc.com/articles/6/feed'), (u'Editors Blog', u'http://www.maximumpc.com/articles/6/feed'),
(u'How-to', u'http://www.maximumpc.com/articles/32/feed'), (u'How-to', u'http://www.maximumpc.com/articles/32/feed'),
(u'Features', u'http://www.maximumpc.com/articles/31/feed'), (u'Features', u'http://www.maximumpc.com/articles/31/feed'),
(u'From the Magazine', u'http://www.maximumpc.com/articles/72/feed') (u'From the Magazine', u'http://www.maximumpc.com/articles/72/feed')
] ]
keep_only_tags = [ #keep_only_tags = [
dict(name='div', attrs={'class':['print-title','article_body']}), #dict(name='div', attrs={'class':['print-title','article_body']}),
] #]
remove_tags = [ #remove_tags = [
dict(name='div', attrs={'class':'comments-tags-actions'}), #dict(name='div', attrs={'class':'comments-tags-actions'}),
] #]
remove_tags_before = dict(name='div', attrs={'class':'print-title'}) #remove_tags_before = dict(name='div', attrs={'class':'print-title'})
remove_tags_after = dict(name='div', attrs={'class':'meta-content'}) #remove_tags_after = dict(name='div', attrs={'class':'meta-content'})
def get_obfuscated_article(self, url): #def get_obfuscated_article(self, url):
br = self.get_browser() #br = self.get_browser()
br.open(url) #br.open(url)
response = br.follow_link(url_regex = r'/print/[0-9]+', nr = 0) #response = br.follow_link(url_regex = r'/print/[0-9]+', nr = 0)
html = response.read() #html = response.read()
self.temp_files.append(PersistentTemporaryFile('_fa.html')) #self.temp_files.append(PersistentTemporaryFile('_fa.html'))
self.temp_files[-1].write(html) #self.temp_files[-1].write(html)
self.temp_files[-1].close() #self.temp_files[-1].close()
return self.temp_files[-1].name #return self.temp_files[-1].name