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

Updated MIT Technology Review

Browse Source
This commit is contained in:
Kovid Goyal 2011-01-12 09:21:20 -07:00
parent 71a732da42
commit d92c03de7d
1 changed files with 22 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
resources/recipes/technology_review.recipe
View File

@ -35,7 +35,6 @@ class TechnologyReview(BasicNewsRecipe):
def get_article_url(self, article): def get_article_url(self, article):
return article.get('guid', article.get('id', None)) return article.get('guid', article.get('id', None))
def print_version(self, url): def print_version(self, url):
baseurl='http://www.technologyreview.com/printer_friendly_article.aspx?id=' baseurl='http://www.technologyreview.com/printer_friendly_article.aspx?id='
split1 = string.split(url,"/") split1 = string.split(url,"/")
@ -43,3 +42,25 @@ class TechnologyReview(BasicNewsRecipe):
split2= string.split(xxx,"/") split2= string.split(xxx,"/")
s = baseurl + split2[0] s = baseurl + split2[0]
return s return s
def postprocess_html(self,soup, True):
#remove picture
headerhtml = soup.find(True, {'class':'header'})
headerhtml.replaceWith("")
#remove close button
closehtml = soup.find(True, {'class':'close'})
closehtml.replaceWith("")
#remove banner advertisement
bannerhtml = soup.find(True, {'class':'bannerad'})
bannerhtml.replaceWith("")
#thanks kiklop74! This code removes all links from the text
for alink in soup.findAll('a'):
if alink.string is not None:
tstr = alink.string
alink.replaceWith(tstr)
return soup