This took me a while to figure out:

Let’s define a normal string which should be evaluated as unicode later

s=’at\u0026t’

string to unicode convert function doesn’t help in this case:
>>> unicode(s)
u’at\\u0026t’

Instead we need to decode string with unicode-escape encoding:

>>> s=’at\u0026t’
>>> type(s)

>>> s.decode( ‘unicode-escape’ )
u’at&t’

Tags: