In Python 3 str is the type for unicode-enabled strings, while bytes is the type for sequences of raw bytes.

type("f") == type(u"f")  # True, <class 'str'>
type(b"f")               # <class 'bytes'>

In Python 2 a casual string was a sequence of raw bytes by default and the unicode string was every string with “u” prefix.

type("f") == type(b"f")  # True, <type 'str'>
type(u"f")               # <type 'unicode'>

Unicode to bytes

Unicode strings can be converted to bytes with .encode(encoding).

Python 3

>>> "£13.55".encode('utf8')
b'\\xc2\\xa313.55'
>>> "£13.55".encode('utf16')
b'\\xff\\xfe\\xa3\\x001\\x003\\x00.\\x005\\x005\\x00'

Python 2

in py2 the default console encoding is sys.getdefaultencoding() == 'ascii' and not utf-8 as in py3, therefore printing it as in the previous example is not directly possible.

>>> print type(u"£13.55".encode('utf8'))
<type 'str'>
>>> print u"£13.55".encode('utf8')
SyntaxError: Non-ASCII character '\\xc2' in...

# with encoding set inside a file

# -*- coding: utf-8 -*-
>>> print u"£13.55".encode('utf8')
£13.55

If the encoding can’t handle the string, a UnicodeEncodeError is raised:

>>> "£13.55".encode('ascii')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
UnicodeEncodeError: 'ascii' codec can't encode character '\\xa3' in position 0: ordinal not in range(128)

Bytes to unicode

Bytes can be converted to unicode strings with .decode(encoding).

A sequence of bytes can only be converted into a unicode string via the appropriate encoding!

>>> b'\\xc2\\xa313.55'.decode('utf8')
'£13.55'

If the encoding can’t handle the string, a UnicodeDecodeError is raised:

>>> b'\\xc2\\xa313.55'.decode('utf16')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/csaftoiu/csaftoiu-github/yahoo-groups-backup/.virtualenv/bin/../lib/python3.5/encodings/utf_16.py", line 16, in decode
    return codecs.utf_16_decode(input, errors, True)
UnicodeDecodeError: 'utf-16-le' codec can't decode byte 0x35 in position 6: truncated data