def test_valid(self):
assert_false(is_unicode(random_bytes))
assert_true(is_unicode(unicode_string))
assert_false(is_unicode(False))
assert_false(is_unicode(5))
assert_false(is_unicode(None))
assert_false(is_unicode([]))
assert_false(is_unicode(()))
assert_false(is_unicode({}))
assert_false(is_unicode(object))
def to_utf8_if_unicode(value):
"""
Converts an argument to a UTF-8 encoded byte string if the argument
is a Unicode string.
:param value:
The value that will be UTF-8 encoded if it is a string.
:returns:
UTF-8 encoded byte string if the argument is a Unicode string; otherwise
the value is returned unchanged.
"""
return unicode_to_utf8(value) if is_unicode(value) else value
def unicode_to_utf8(value):
"""
Converts a string argument to a UTF-8 encoded byte string if it is a
Unicode string.
:param value:
If already a byte string or None, it is returned unchanged.
Otherwise it must be a Unicode string and is encoded as UTF-8.
"""
if value is None or is_bytes(value):
return value
assert is_unicode(value)
return value.encode("utf-8")
def bytes_to_unicode(value, encoding="utf-8"):
"""
Converts bytes to a Unicode string decoding it according to the encoding
specified.
:param value:
If already a Unicode string or None, it is returned unchanged.
Otherwise it must be a byte string.
:param encoding:
The encoding used to decode bytes. Defaults to UTF-8
"""
if value is None or is_unicode(value):
return value
assert is_bytes(value)
return value.decode(encoding)