[Tutor] simple unicode question

[Ben Finney]

Danny Yoo <dyoo at hashcollision.org> writes:

> Hi Albert,
>
> Just following up: I consider the 2 (or 3) arg form of unicode() to be
> a design flaw in the Standard Library.  It's redundant because the
> bytes type already has a decode() method:
>
>     https://docs.python.org/3/library/stdtypes.html#bytes.decode
>
>
> So I would personally write convert() like this:
>
> ######################################
> def convert(data):
>     if isinstance(data, float):
>         return unicode(data)
>     if isinstance(data, bytes):
>         return data.decode("utf-8")
>     raise ValueError("Unexpected data", data)
> ######################################

Since the type of ‘data’ is the only thing which determines whether an
error is raised, the error message might be improved by stating the
type::

    def convert(data):
        if isinstance(data, float):
            return unicode(data)
        if isinstance(data, bytes):
            return data.decode("utf-8")
        raise ValueError(
                "Unexpected type {type} for data {data!r}".format(
                    type=type(data), data=data))

which makes the errors more self-explanatory, IMO::

    >>> convert(17.3)
    '17.3'

    >>> convert(b'\x53\x70\x61\x6d')
    'Spam'

    >>> convert(12)
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
      File "<stdin>", line 9, in convert
    ValueError: Unexpected type <class 'int'> for data 12

    >>> convert('Spam')
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
      File "<stdin>", line 9, in convert
    ValueError: Unexpected type <class 'str'> for data 'Spam'

-- 
 \       “Anyone who puts a small gloss on [a] fundamental technology, |
  `\          calls it proprietary, and then tries to keep others from |
_o__)           building on it, is a thief.” —Tim O'Reilly, 2000-01-25 |
Ben Finney