pictuga/morss
1
1
Fork 0
Code Issues Pull Requests Actions Packages Releases Activity

readabilite: add tags to black list

Browse Source
This commit is contained in:
pictuga
2017-03-24 21:50:26 -10:00
parent 91da0f36dc
commit bfefa8d599
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
morss/readabilite.py
View File

@@ -40,14 +40,14 @@ regex_bad = re.compile('|'.join(['robots-nocontent', 'combx', 'comment',
'shoutbox', 'sidebar', 'sponsor', 'ad-', 'agegate', 'pagination',
'pager', 'popup', 'tweet', 'twitter', 'com-', 'sharing', 'share', 'social',
'contact', 'footnote', 'masthead', 'media', 'meta', 'outbrain', 'promo',
'related', 'scroll', 'shoutbox', 'sidebar', 'sponsor', 'shopping', 'tags',
'related', 'scroll', 'shoutbox', 'shopping', 'tags',
'tool', 'widget', 'hide']), re.I)
regex_good = re.compile('|'.join(['and', 'article', 'body', 'column',
'main', 'shadow', 'content', 'entry', 'hentry', 'main', 'page',
'pagination', 'post', 'text', 'blog', 'story', 'par', 'editorial']), re.I)
tags_junk = ['script', 'head', 'iframe', 'object', 'noscript', 'param', 'embed', 'layer', 'applet', 'style', 'form', 'input', 'textarea']
tags_junk = ['script', 'head', 'iframe', 'object', 'noscript', 'param', 'embed', 'layer', 'applet', 'style', 'form', 'input', 'textarea', 'button']
attributes_fine = ['title', 'src', 'href', 'type', 'name', 'for', 'value']