What is unicode_literals used for?

I get a weird problem with __future__.unicode_literals in Python. Without importing unicode_literals I get the correct output:

# encoding: utf-8
# from __future__ import unicode_literals
name = 'helló wörld from example'
print name

But when I add the unicode_literals import:

# encoding: utf-8
from __future__ import unicode_literals
name = 'helló wörld from example'
print name

I got this error:

UnicodeEncodeError: 'ascii' codec can't encode character u'\xf3' in position 4: ordinal not in range(128)

Does unicode_literals encode every string as an utf-8? What should I do to override this error?


Your terminal or console is failing to let Python know it supports UTF-8.

Without the from __future__ import unicode_literals line, you are building a byte string that holds UTF-8 encoded bytes. With the string you are building a unicode string.

print has to treat these two values differently; a byte string is written to sys.stdout unchanged. A unicode string is encoded to bytes first, and Python consults sys.stdout.encoding for that. If your system doesn't correctly tell Python what codec it supports, the default is to use ASCII.

Your system failed to tell Python what codec to use; sys.stdout.encoding is set to ASCII, and encoding the unicode value to print failed.

You can verify this by manually encoding to UTF-8 when printing:

# encoding: utf-8
from __future__ import unicode_literals
name = 'helló wörld from example'
print name.encode('utf8')

and you can reproduce the issue by creating unicode literals without the from __future__ import statement too:

# encoding: utf-8
name = u'helló wörld from example'
print name

where u'..' is a unicode literal too.

Without details on what your environment is, it is hard to say what the solution is; this depends very much on the OS and console or terminal used.