You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Often optional components come in loose groups, like mutating functions push!, delete!, empty! would often all be implemented for a mutable AbstractSet.
it would be good to be able to specify this group with just mutable, and get the tests and traits for all its components.
Nested NamedTuple is one simple way to do it - but it loses the posibility to have components in multiple overlapping groups. adding a groups object after optional may be better design. It would be a NamedTuple of tuples of Symbol.
The text was updated successfully, but these errors were encountered:
While I understand the purpose, in the spirit of simplicity I don't really like the idea of a groups named tuple key. I think the person defining the interface can handle overlapping groups themselves, by concatenating named tuples.
OK I'm starting to see what you mean as I take a look at #27
Maybe if we had inheritance we could use that instead of groups?
As in, ArrayInterface could inherit from ArrayModifyInterface and ArrayQueryInterface?
Often optional components come in loose groups, like mutating functions
push!
,delete!
,empty!
would often all be implemented for a mutableAbstractSet
.it would be good to be able to specify this group with just
mutable
, and get the tests and traits for all its components.Nested NamedTuple is one simple way to do it - but it loses the posibility to have components in multiple overlapping groups. adding a
groups
object afteroptional
may be better design. It would be aNamedTuple
of tuples ofSymbol
.The text was updated successfully, but these errors were encountered: