使用ROT13的codecs.decode中可能存在错误

Question

Python 3.5.1 on Ubuntu

>>> from codecs import decode
>>> s = 'string'
>>> b = b'bytes'
>>> decode(b, 'utf8')
'bytes'
>>> decode(s, 'utf8')
Traceback (most recent call last):
  File "/usr/lib/python3.5/encodings/utf_8.py", line 16, in decode
    return codecs.utf_8_decode(input, errors, True)
TypeError: a bytes-like object is required, not 'str'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: decoding with 'utf8' codec failed (TypeError: a bytes-like object is required, not 'str')

so far everything behaves as expected. But when I try to use ROT13 encoding I get:

>>> decode(s, 'rot13')
'fgevat'
>>> decode(b, 'rot13')
Traceback (most recent call last):
  File "/usr/lib/python3.5/encodings/rot_13.py", line 18, in decode
    return (input.translate(rot13_map), len(input))
TypeError: a bytes-like object is required, not 'dict'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: decoding with 'rot13' codec failed (TypeError: a bytes-like object is required, not 'dict')

As @snakecharmerb points out ROT13 en/decoding is only supposed to work on strings.
The exception Message though, is still wrong since it states, that a bytes-like object was expected even when a bytes-like object was actually passed, and mentions some dictionary the user did not create.

Answer 1

Regarding the ROT13 codec, the Python 3.5 docs state :

The following codec provides a text transform: a str to str mapping. It is not supported by str.encode() (which only produces bytes output).

that is, you can only pass unicode strings ( str objects) when encoding to ROT13 and you will only get str objects back.

This is different from Python 2.x, when ROT13 was treated the same as other codecs. ROT13 was not ported to Python 3 initially, because ROT13 does not encode a unicode string to bytes - it just swaps letters around. It as reinstated, as a text transform in Python 3.2 and 3.4. The full story is in this bug report .