do not skip span for newline textnodes otherwise compute_new_comment_positions() will return bad results for pandoc texts.
# django helpers collections
from django.conf import settings
from django.core.urlresolvers import reverse
def absolute_reverse(view_id, args):
return settings.SITE_URL + reverse(view_id, args=args)