The Maybe
container is used when a series of computations
could return None
at any point.
Maybe
consist of two types: Some
and Nothing
.
We have a convenient method to create different Maybe
types
based on just a single value:
>>> from returns.maybe import Maybe
>>> str(Maybe.from_value(1))
'<Some: 1>'
>>> str(Maybe.from_value(None))
'<Nothing>'
It might be very useful for complex operations like the following one:
>>> from attr import dataclass
>>> from typing import Optional
>>> from returns.maybe import Maybe
>>> @dataclass
... class Address(object):
... street: Optional[str]
>>> @dataclass
... class User(object):
... address: Optional[Address]
>>> @dataclass
... class Order(object):
... user: Optional[User]
>>> def get_street_address(order: Order) -> Maybe[str]:
... return Maybe.from_value(order.user).map(
... lambda user: user.address,
... ).map(
... lambda address: address.street,
... )
>>> with_address = Order(User(Address('Some street')))
>>> empty_user = Order(None)
>>> empty_address = Order(User(None))
>>> empty_street = Order(User(Address(None)))
>>> str(get_street_address(with_address)) # all fields are not None
'<Some: Some street>'
>>> str(get_street_address(empty_user))
'<Nothing>'
>>> str(get_street_address(empty_address))
'<Nothing>'
>>> str(get_street_address(empty_street))
'<Nothing>'
One may ask: “How is that different to the Optional[]
type?”
That’s a really good question!
Consider the same code to get the street name
without Maybe
and using raw Optional
values:
order: Order # some existing Order instance
street: Optional[str] = None
if order.user is not None:
if order.user.address is not None:
street = order.user.address.street
It looks way uglier and can grow even more uglier and complex when new logic will be introduced.
Typing will only work correctly if our mypy plugin is used. This happens due to mypy issue.
Sometimes we have to deal with functions
that dears to return Optional
values!
We have to work with it the carefully
and write if x is not None:
everywhere.
Luckily, we have your back! maybe
function decorates
any other function that returns Optional
and converts it to return Maybe
instead:
>>> from typing import Optional
>>> from returns.maybe import Maybe, maybe
>>> @maybe
... def number(num: int) -> Optional[int]:
... if num > 0:
... return num
... return None
>>> result: Maybe[int] = number(1)
>>> str(result)
'<Some: 1>'
When working with regular Python,
you might need regular Optional[a]
values.
You can easily get one from your Maybe
container at any point in time:
>>> from returns.maybe import Maybe
>>> assert Maybe.from_value(1).value_or(None) == 1
>>> assert Maybe.from_value(None).value_or(None) is None
As you can see, revealed type of .value_or(None)
is Optional[a]
.
Use it a fallback.
Let’s say you have this dict
: {'a': 1, 'b': None}
And you want to get Maybe[int]
values by string keys from there.
When trying both existing key 'b'
and missing key 'c'
you will end up with Nothing
for both values.
But, they are different! You might need to know exactly which case you are dealing with.
In this case, it is better to switch to Result
type.
Let’s see how to model this real-life situation:
>>> from returns.result import Success, Failure, safe
>>> source = {'a': 1, 'b': None}
>>> md = safe(lambda key: source[key])
>>> assert md('a') == Success(1)
>>> assert md('b') == Success(None)
>>> # Is: Failure(KeyError('c'))
>>> assert md('c').failure().args == ('c',)
This way you can tell the difference
between empty values (None
) and missing keys.
You can always use returns.converters.result_to_maybe()
to convert Result
to Maybe
.
See the original issue about Some(None) for more details and the full history.
We do have IOResult
, but we don’t have IOMaybe
. Why?
Because when dealing with IO
there are a lot of possible errors.
And Maybe
represents just None
and the value.
It is not useful for IO
related tasks.
So, use Result
instead, which can represent what happened to your IO
.
You can convert Maybe
to Result
and back again with special Converters.
Well, because Maybe
only has a single type argument: _ValueType
.
And all these method implies that we also has _ErrorType
.
We used to have them. There were several issues:
If we leave their signature untouched (with the explicit None
error type)
then we would have to write functions that always ignore the passed argument.
It is a bit ugly!
If we change the signature of the passed function to have zero arguments, then we would have a lot of problems with typing. Because now different types would require different callback functions for the same methods!
We didn’t like both options and dropped these methods in some early release.
Now, Maybe
has returns.maybe.Maybe.or_else_call()
method to call
a passed callback function with zero argument on failed container:
>>> from returns.maybe import Some, Nothing
>>> assert Some(1).or_else_call(lambda: 2) == 1
>>> assert Nothing.or_else_call(lambda: 2) == 2
This method is unique to Maybe
container.
Maybe
(inner_value)[source]¶Bases: returns.primitives.container.BaseContainer
, typing.Generic
Represents a result of a series of computations that can return None
.
An alternative to using exceptions or constant is None
checks.
Maybe
is an abstract type and should not be instantiated directly.
Instead use Some
and Nothing
.
success_type
¶Success type that is used to represent the successful computation.
inner_value (+_ValueType) –
alias of _Some
failure_type
¶Failure type that is used to represent the failed computation.
inner_value (None
) –
alias of _Nothing
map
(function)[source]¶Composes successful container with a pure function.
>>> from returns.maybe import Some, Nothing
>>> def mappable(string: str) -> str:
... return string + 'b'
>>> assert Some('a').map(mappable) == Some('ab')
>>> assert Nothing.map(mappable) == Nothing
function (Callable
[[+_ValueType], Optional
[~_NewValueType]]) –
Maybe
[~_NewValueType]
apply
(function)[source]¶Calls a wrapped function in a container on this container.
>>> from returns.maybe import Some, Nothing
>>> def appliable(string: str) -> str:
... return string + 'b'
>>> assert Some('a').apply(Some(appliable)) == Some('ab')
>>> assert Some('a').apply(Nothing) == Nothing
>>> assert Nothing.apply(Some(appliable)) == Nothing
>>> assert Nothing.apply(Nothing) == Nothing
bind
(function)[source]¶Composes successful container with a function that returns a container.
>>> from returns.maybe import Nothing, Maybe, Some
>>> def bindable(string: str) -> Maybe[str]:
... return Some(string + 'b')
>>> assert Some('a').bind(bindable) == Some('ab')
>>> assert Nothing.bind(bindable) == Nothing
value_or
(default_value)[source]¶Get value from successful container or default value from failed one.
>>> from returns.maybe import Nothing, Some
>>> assert Some(0).value_or(1) == 0
>>> assert Nothing.value_or(1) == 1
default_value (~_NewValueType) –
Union
[+_ValueType, ~_NewValueType]
or_else_call
(function)[source]¶Get value from successful container or default value from failed one.
Really close to value_or()
but works with lazy values.
This method is unique to Maybe
container, because other containers
do have .rescue
, .alt
, .fix
methods.
But, Maybe
does not.
Instead, it has this method to execute some function if called on a failed container:
>>> from returns.maybe import Some, Nothing
>>> assert Some(1).or_else_call(lambda: 2) == 1
>>> assert Nothing.or_else_call(lambda: 2) == 2
It might be useful to work with exceptions as well:
>>> def fallback() -> NoReturn:
... raise ValueError('Nothing!')
>>> Nothing.or_else_call(fallback)
Traceback (most recent call last):
...
ValueError: Nothing!
function (Callable
[[], ~_NewValueType]) –
Union
[+_ValueType, ~_NewValueType]
unwrap
()[source]¶Get value from successful container or raise exception for failed one.
>>> from returns.maybe import Nothing, Some
>>> assert Some(1).unwrap() == 1
>>> Nothing.unwrap()
Traceback (most recent call last):
...
returns.primitives.exceptions.UnwrapFailedError
+_ValueType
failure
()[source]¶Get failed value from failed container or raise exception from success.
>>> from returns.maybe import Nothing, Some
>>> assert Nothing.failure() is None
>>> Some(1).failure()
Traceback (most recent call last):
...
returns.primitives.exceptions.UnwrapFailedError
None
from_value
(inner_value)[source]¶Creates new instance of Maybe
container based on a value.
>>> from returns.maybe import Maybe, Some, Nothing
>>> assert Maybe.from_value(1) == Some(1)
>>> assert Maybe.from_value(None) == Nothing
inner_value (Optional
[+_ValueType]) –
Maybe
[+_ValueType]
from_iterable
(inner_value)[source]¶Transforms an iterable of Maybe
containers into a single container.
>>> from returns.maybe import Maybe, Some, Nothing
>>> assert Maybe.from_iterable([
... Some(1),
... Some(2),
... ]) == Some((1, 2))
>>> assert Maybe.from_iterable([
... Some(1),
... Nothing,
... ]) == Nothing
>>> assert Maybe.from_iterable([
... Nothing,
... Some(1),
... ]) == Nothing
Some
(inner_value)[source]¶Public unit function of protected _Some
type.
Can return Nothing
for passed None
argument.
Because Some(None)
does not make sence.
>>> from returns.maybe import Some
>>> str(Some(1))
'<Some: 1>'
>>> str(Some(None))
'<Nothing>'
inner_value (Optional
[+_ValueType]) –
Maybe
[+_ValueType]
Nothing
: returns.maybe.Maybe[NoReturn] = <returns.maybe._Nothing object>¶Public unit value of protected _Nothing
type.
maybe
(function)[source]¶Decorator to convert None
-returning function to Maybe
container.
This decorator works with sync functions only. Example:
>>> from typing import Optional
>>> from returns.maybe import Nothing, Some, maybe
>>> @maybe
... def might_be_none(arg: int) -> Optional[int]:
... if arg == 0:
... return None
... return 1 / arg
>>> assert might_be_none(0) == Nothing
>>> assert might_be_none(1) == Some(1.0)
Requires our mypy plugin.
function (Callable
[…, Optional
[+_ValueType]]) –
Callable
[…, Maybe
[+_ValueType]]