diff --git a/resources/recipes/new_scientist.recipe b/resources/recipes/new_scientist.recipe index 1727a926ed..b40be458bc 100644 --- a/resources/recipes/new_scientist.recipe +++ b/resources/recipes/new_scientist.recipe @@ -32,15 +32,16 @@ class NewScientist(BasicNewsRecipe): } preprocess_regexps = [(re.compile(r'.*?', re.DOTALL|re.IGNORECASE),lambda match: '')] - keep_only_tags = [dict(name='div', attrs={'id':['pgtop','maincol','nsblgposts','hldgalcols']})] + keep_only_tags = [dict(name='div', attrs={'id':['pgtop','maincol','blgmaincol','nsblgposts','hldgalcols']})] remove_tags = [ dict(name='div' , attrs={'class':['hldBd','adline','pnl','infotext' ]}) - ,dict(name='div' , attrs={'id' :['compnl','artIssueInfo','artTools']}) + ,dict(name='div' , attrs={'id' :['compnl','artIssueInfo','artTools','comments','blgsocial']}) ,dict(name='p' , attrs={'class':['marker','infotext' ]}) ,dict(name='meta' , attrs={'name' :'description' }) + ,dict(name='a' , attrs={'rel' :'tag' }) ] - remove_tags_after = dict(attrs={'class':'nbpcopy'}) + remove_tags_after = dict(attrs={'class':['nbpcopy','comments']}) remove_attributes = ['height','width'] feeds = [