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

Fix CNN website recipe

Browse Source 
* Add style
	* Add CNN logo
	* Fix problem with image gallery
This commit is contained in:
Aurélien Chabot 2011-10-16 13:39:28 +02:00
parent a24ba3ff59
commit 5660b08fbb
1 changed files with 23 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
recipes/cnn.recipe
View File

@ -22,6 +22,14 @@ class CNN(BasicNewsRecipe):
#match_regexps = [r'http://sportsillustrated.cnn.com/.*/[1-9].html'] #match_regexps = [r'http://sportsillustrated.cnn.com/.*/[1-9].html']
max_articles_per_feed = 25 max_articles_per_feed = 25
extra_css = '''
h1 {font-size:xx-large; font-family:Arial,Helvetica,sans-serif;}
.cnn_story_author, .cnn_stryathrtmp {font-size:xx-small; color:#4D4D4D; font-family:Arial,Helvetica,sans-serif;}
.cnn_strycaptiontxt, .cnnArticleGalleryPhotoContainer {font-size:xx-small; color:#4D4D4D; font-family:Arial,Helvetica,sans-serif;}
.cnn_strycbftrtxt, .cnnEditorialNote {font-size:xx-small; color:#4D4D4D; font-family:Arial,Helvetica,sans-serif;}
.cnn_strycntntlft {font-size:medium; font-family:Arial,Helvetica,sans-serif;}
'''
preprocess_regexps = [ preprocess_regexps = [
(re.compile(r'<!--\[if.*if\]-->', re.DOTALL), lambda m: ''), (re.compile(r'<!--\[if.*if\]-->', re.DOTALL), lambda m: ''),
(re.compile(r'<script.*?</script>', re.DOTALL), lambda m: ''), (re.compile(r'<script.*?</script>', re.DOTALL), lambda m: ''),
@ -32,7 +40,12 @@ class CNN(BasicNewsRecipe):
remove_tags = [ remove_tags = [
{'class':['cnn_strybtntools', 'cnn_strylftcntnt', {'class':['cnn_strybtntools', 'cnn_strylftcntnt',
'cnn_strybtntools', 'cnn_strybtntoolsbttm', 'cnn_strybtmcntnt', 'cnn_strybtntools', 'cnn_strybtntoolsbttm', 'cnn_strybtmcntnt',
'cnn_strycntntrgt', 'hed_side', 'foot']}, 'cnn_strycntntrgt', 'hed_side', 'foot', 'cnn_strylftcntnt cnn_strylftcexpbx']},
{'class':['cnn_html_media_title_new', 'cnn_html_media_title_new cnn_html_media_title_none',
'cnnArticleGalleryCaptionControlText', 'articleGalleryNavContainer']},
{'id':['articleGalleryNav00JumpPrev', 'articleGalleryNav00Prev',
'articleGalleryNav00Next', 'articleGalleryNav00JumpNext']},
{'style':['display:none']},
dict(id=['ie_column']), dict(id=['ie_column']),
] ]
@ -58,3 +71,12 @@ class CNN(BasicNewsRecipe):
ans = BasicNewsRecipe.get_article_url(self, article) ans = BasicNewsRecipe.get_article_url(self, article)
return ans.partition('?')[0] return ans.partition('?')[0]
def get_masthead_url(self):
masthead = 'http://i.cdn.turner.com/cnn/.element/img/3.0/global/header/intl/hdr-globe-central.gif'
br = BasicNewsRecipe.get_browser()
try:
br.open(masthead)
except:
self.log("\nCover unavailable")
masthead = None
return masthead