rspec/rules/S5890/python/rule.adoc

== Why is this an issue?

Type hints can be used to communicate the intended type of a given variable. These are not enforced at runtime and not respecting them might not necessarily lead to runtime errors.


It is however confusing and could lead to maintainability issues.


=== Noncompliant code example

[source,python]
----
def my_function():
    my_int: int = "string"  # Noncompliant
----


=== Compliant solution

[source,python]
----
def my_function():
    my_str: str = "string" 
----


ifdef::env-github,rspecator-view[]

'''
== Implementation Specification
(visible only on this page)

=== Message

* If the asignment value is not ``++None++`` then 
** Primary: Assign to "XXX" a value of type "YYY" instead of "ZZZ" or update the type hint of "XXX".
** Secondary: Type hint.
* If the assignment value is ``++None++``
** Primary: Replace the type hint "XXX" with "Optional[XXX]" or don't assign "None" to "YYY"
** Secondary: Type hint.


=== Highlighting

* Primary: The assigned value
* Secondary: The type hint


'''
== Comments And Links
(visible only on this page)

=== on 12 May 2021, 00:45:18 Mike Heyns wrote:
This rule does not appear to be working for https://docs.python.org/3/library/dataclasses.html#init-only-variables[Dataclass Init-Only variables]:


----
@dataclass
class Book:
    name: str     
    condition: InitVar[str] = ''
----

____Assign to "condition" a value of type "InitVar[str]" instead of "str" or update its type hint.____


In fact, most examples including the official documentation use ``++None++`` as the default for init-only arguments. Attempting to use ``++InitVar[Optional[str]]++`` has the same effect as above.

endif::env-github,rspecator-view[]