On 07/28/2017 01:13 AM, Ben Finney wrote:
Ethan Furman writes:
class X(Enum):
Falsey = 0
Truthy = 1
Fakey = 2
def __bool__(self):
return bool(self.value)
I am surprised this is not already the behaviour of an Enum class,
without overriding the ‘__bool__’ method.
What would be a good reason not to have this behaviour by default for
‘Enum.__bool__’? (i.e. if this were reported as a bug on the ‘enum.Enum’
implementation, what would be good reasons not to fix it?)
It matches the docs. ;)
https://docs.python.org/3/library/enum.html#boolean-value-of-enum-classes-and-members
Enum members that are mixed with non-Enum types (such as int, str, etc.)
are evaluated according to the mixed-in type’s rules; otherwise, all
members evaluate as True. To make your own Enum’s boolean evaluation
depend on the member’s value add the following to your class:
Enum classes always evaluate as True.
The rationale is in PEP 435, in the functional API section.
The reason for defaulting to 1 as the starting number and not 0 is that 0 is
False
in a boolean sense, but enum members all evaluate to True .
Which still begs the question of: why? According to memory, and as Rustom guessed, an Enum member is a thing, and all
things in Python default to being True. If `bool(thing) == False` is a desirable characteristic then extra steps must
be taken. Either:
- subclass a type that already has that characteristic, such as int or float, or
- add your own __bool__ method (__nonzero__ in Python 2)
Looked at another way: The value attribute is just that -- an attribute. Depending on the use-case that attribute can
be irrelevant (which is why we have auto(), etc.), so by default Enum does not use the .value attribute in figuring out
if a member is something vs nothing (or truthy vs falsey).
--
~Ethan~
--
https://mail.python.org/mailman/listinfo/python-list
