python-markupsafe 0.9.2-3build1 (i386 binary) in ubuntu natty

 MarkupSafe implements a unicode subclass that supports HTML strings:
 .
  >>> from markupsafe import Markup, escape
  >>> escape("<script>alert(document.cookie);</script>")
  Markup(u'&lt;script&gt;alert(document.cookie);&lt;/script&gt;')
  >>> tmpl = Markup("<em>%s</em>")
  >>> tmpl % "Peter > Lustig"
  Markup(u'<em>Peter &gt; Lustig</em>')
 .
 If you want to make an object unicode that is not yet unicode
 but don't want to lose the taint information, you can use the
 `soft_unicode` function:
 .
  >>> from markupsafe import soft_unicode
  >>> soft_unicode(42)
  u'42'
  >>> soft_unicode(Markup('foo'))
  Markup(u'foo')