scrubber.allowed_tags = set((
'a',
'abbr',
'acronym',
'b',
'bdo',
'big',
'blockquote',
'br',
'center',
'cite',
'code',
'dd',
'del',
'dfn',
'div',
'dl',
'dt',
'em',
'h1',
'h2',
'h3',
'h4',
'h5',
'h6',
'hr',
'i',
'img',
'ins',
'kbd',
'li',
'ol',
'param',
'pre',
'p',
'q',
's',
'samp',
'small',
'span',
'strike',
'strong',
'sub',
'sup',
'table',
'tbody',
'td',
'th',
'thead',
'tr',
'tt',
'ul',
'u',
'var',
'wbr',
))
import os
from scrubber import Scrubber
import BeautifulSoup
#import BeautifulSoup
# initialise the scrubber! all this stuff is overriding scrubber defaults so hack it to bits if you want!
scrubber = Scrubber(autolink=False)
scrubber.allowed_tags = set((
'a', 'abbr', 'acronym', 'b', 'bdo', 'big', 'blockquote', 'br',
'center', 'cite', 'code',
'dd', 'del', 'dfn', 'div', 'dl', 'dt', 'em',
'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'hr', 'i', 'img', 'ins',
'kbd', 'li', 'ol', 'param', 'pre', 'p', 'q',
's', 'samp', 'small', 'span', 'strike', 'strong', 'sub', 'sup',
'table', 'tbody', 'td', 'th', 'thead', 'tr', 'tt', 'ul', 'u',
'var', 'wbr',
))
scrubber.disallowed_tags_save_content = set((
'blink', 'body', 'html','font',
))
scrubber.allowed_attributes = set((
'align', 'alt', 'border', 'cite', 'dir',
'height', 'href', 'src', 'title', 'type', 'width',
'face', 'size', # font tags
'flashvars', # Not sure about flashvars - if any harm can come from it
'classid', # FF needs the classid on object tags for flash
'name', 'value', 'quality', 'data', 'scale', # for flash embed param tags, could limit to just param if this is harmful
'salign', 'align', 'wmode',
)) # Bad attributes: 'allowscriptaccess', 'xmlns', 'target'
scrubber.normalized_tag_replacements = {'b': 'strong', 'i': 'em'}
# any giveaway classes the definately identify a footer.
