Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 02:34:06 -04:00
Code Issues Packages Projects Releases Wiki Activity

MobileBulgaria by Martin Tsanchev

Browse Source
This commit is contained in:
Kovid Goyal 2012-12-20 17:03:14 +05:30
parent ea50efb4f5
commit 992bf4b423
2 changed files with 35 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
recipes/empire_magazine.recipe
View File

@ -5,6 +5,7 @@ class AdvancedUserRecipe1341650280(BasicNewsRecipe):
title = u'Empire Magazine' title = u'Empire Magazine'
description = 'Author D.Asbury. Film articles from Empire Mag. ' description = 'Author D.Asbury. Film articles from Empire Mag. '
language = 'en'
__author__ = 'Dave Asbury' __author__ = 'Dave Asbury'
# last updated 7/7/12 # last updated 7/7/12
remove_empty_feeds = True remove_empty_feeds = True
@ -15,7 +16,7 @@ class AdvancedUserRecipe1341650280(BasicNewsRecipe):
cover_url = 'http://www.empireonline.com/images/magazine/cover.jpg' cover_url = 'http://www.empireonline.com/images/magazine/cover.jpg'
conversion_options = { conversion_options = {
'linearize_tables' : True, 'linearize_tables' : True,
} }
#auto_cleanup = True #auto_cleanup = True
preprocess_regexps = [ preprocess_regexps = [
(re.compile(r'<a href="http://twitter.com/share.*?</a>', re.IGNORECASE | re.DOTALL), lambda match: ''), (re.compile(r'<a href="http://twitter.com/share.*?</a>', re.IGNORECASE | re.DOTALL), lambda match: ''),
@ -32,20 +33,20 @@ class AdvancedUserRecipe1341650280(BasicNewsRecipe):
(re.compile(r'<!-- USER REVIEWS: START -->.*?<!-- USER REVIEWS: END -->', re.IGNORECASE | re.DOTALL), lambda match: '<!-- USER REVIEWS: START --><!-- USER REVIEWS: END -->'), (re.compile(r'<!-- USER REVIEWS: START -->.*?<!-- USER REVIEWS: END -->', re.IGNORECASE | re.DOTALL), lambda match: '<!-- USER REVIEWS: START --><!-- USER REVIEWS: END -->'),
(re.compile(r'Advertisement', re.IGNORECASE | re.DOTALL), lambda match: ''), (re.compile(r'Advertisement', re.IGNORECASE | re.DOTALL), lambda match: ''),
(re.compile(r'<a name="haveyoursay".*?now to have your say.', re.IGNORECASE | re.DOTALL), lambda match: ''), (re.compile(r'<a name="haveyoursay".*?now to have your say.', re.IGNORECASE | re.DOTALL), lambda match: ''),
] ]
keep_only_tags = [ keep_only_tags = [
# dict(name='h1'), # dict(name='h1'),
# dict(attrs={'class' : 'mediumblack'}), # dict(attrs={'class' : 'mediumblack'}),
] ]
remove_tags = [dict(name='td', attrs={'width':'200', 'valign' : 'top'}), remove_tags = [dict(name='td', attrs={'width':'200', 'valign' : 'top'}),
dict(name='b'), dict(name='b'),
dict(name='a',attrs={'name' : 'haveyoursay'}), dict(name='a',attrs={'name' : 'haveyoursay'}),
dict(attrs={'class' : 'newslink'}), dict(attrs={'class' : 'newslink'}),
] ]
feeds = [(u'News', u'http://feed43.com/7338478755673147.xml'), feeds = [(u'News', u'http://feed43.com/7338478755673147.xml'),
(u'Recent Features',u'http://feed43.com/4346347750304760.xml'), (u'Recent Features',u'http://feed43.com/4346347750304760.xml'),
(u'Interviews',u'http://feed43.com/3418350077724081.xml'), (u'Interviews',u'http://feed43.com/3418350077724081.xml'),
(u'Film Reviews',u'http://feed43.com/2643703076510627.xml'), (u'Film Reviews',u'http://feed43.com/2643703076510627.xml'),
] ]

27
recipes/mobile_bulgaria.recipe Normal file
View File

@ -0,0 +1,27 @@
from calibre.web.feeds.news import BasicNewsRecipe
class AdvancedUserRecipe1329123365(BasicNewsRecipe):
title = u'Mobilebulgaria.com'
__author__ = 'M3 Web'
description = 'The biggest Bulgarian site covering mobile consumer electronics. Offers detailed reviews, popular discussion forum, shop and platform for selling new and second hand phones and gadgets.'
category = 'News, Reviews, Offers, Forum'
oldest_article = 45
max_articles_per_feed = 10
language = 'bg'
encoding = 'windows-1251'
no_stylesheets = False
remove_javascript = True
keep_only_tags = [dict(name='div', attrs={'class':'bigblock'}),
dict(name='div', attrs={'class':'verybigblock'}),
dict(name='table', attrs={'class':'obiaviresults'}),
dict(name='div', attrs={'class':'forumblock'}),
dict(name='div', attrs={'class':'forumblock_b1'}),
dict(name='div', attrs={'class':'block2_2colswrap'})]
feeds = [(u'News', u'http://www.mobilebulgaria.com/rss_full.php'),
(u'Reviews', u'http://www.mobilebulgaria.com/rss_reviews.php'),
(u'Offers', u'http://www.mobilebulgaria.com/obiavi/rss.php'),
(u'Forum', u'http://www.mobilebulgaria.com/rss_forum_last10.php')]
extra_css = '''
#gallery1 div{display: block; float: left; margin: 0 10px 10px 0;} '''