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 Fronda

Browse Source
This commit is contained in:
Kovid Goyal 2013-04-09 09:23:19 +05:30
commit 03249c2549
2 changed files with 12 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
recipes/fronda.recipe
View File

@ -23,6 +23,7 @@ class Fronda(BasicNewsRecipe):
extra_css = ''' extra_css = '''
h1 {font-size:150%} h1 {font-size:150%}
.body {text-align:left;} .body {text-align:left;}
div#featured-image {font-style:italic; font-size:70%}
''' '''
earliest_date = date.today() - timedelta(days=oldest_article) earliest_date = date.today() - timedelta(days=oldest_article)
@ -55,7 +56,10 @@ class Fronda(BasicNewsRecipe):
articles = {} articles = {}
for url, genName in genres: for url, genName in genres:
try:
soup = self.index_to_soup('http://www.fronda.pl/c/'+ url) soup = self.index_to_soup('http://www.fronda.pl/c/'+ url)
except:
continue
articles[genName] = [] articles[genName] = []
for item in soup.findAll('li'): for item in soup.findAll('li'):
article_h = item.find('h2') article_h = item.find('h2')
@ -77,16 +81,15 @@ class Fronda(BasicNewsRecipe):
] ]
remove_tags = [ remove_tags = [
dict(name='div', attrs={'class':['related-articles', dict(name='div', attrs={'class':['related-articles','button right','pagination','related-articles content']}),
'button right',
'pagination']}),
dict(name='h3', attrs={'class':'block-header article comments'}), dict(name='h3', attrs={'class':'block-header article comments'}),
dict(name='ul', attrs={'class':'comment-list'}), dict(name='ul', attrs={'class':['comment-list','category','tag-list']}),
dict(name='ul', attrs={'class':'category'}),
dict(name='ul', attrs={'class':'tag-list'}),
dict(name='p', attrs={'id':'comments-disclaimer'}), dict(name='p', attrs={'id':'comments-disclaimer'}),
dict(name='div', attrs={'style':'text-align: left; margin-bottom: 15px;'}), dict(name='div', attrs={'style':'text-align: left; margin-bottom: 15px;'}),
dict(name='div', attrs={'style':'text-align: left; margin-top: 15px; margin-bottom: 30px;'}), dict(name='div', attrs={'style':'text-align: left; margin-top: 15px; margin-bottom: 30px;'}),
dict(name='div', attrs={'class':'related-articles content'}), dict(name='div', attrs={'id':'comment-form'}),
dict(name='div', attrs={'id':'comment-form'}) dict(name='span', attrs={'class':'separator'})
] ]
preprocess_regexps = [
(re.compile(r'komentarzy: .*?</h6>', re.IGNORECASE | re.DOTALL | re.M ), lambda match: '</h6>')]

BIN
recipes/icons/newsweek_polska.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 905 B