Add a Field1 instance for GHC.Tuple.Unit#878
Conversation
| _1 f (Identity a) = Identity <$> f a | ||
|
|
||
| instance Field1 (Unit a) (Unit b) a b where | ||
| _1 f (Unit a) = Unit <$> f a |
There was a problem hiding this comment.
Or should this be lazy???? I think all the tuple instances should be strict, but I know others disagree...
There was a problem hiding this comment.
I'd vote to keep it the way you've written it, as it's consistent with the other Field1 instances that are currently defined. If others feel strongly about changing this convention, then I think it would be best to leave that to a separate issue.
There was a problem hiding this comment.
Other Field1 insances are inconsistent. See Pair f g a and (a, b).
There was a problem hiding this comment.
Oops, I'm not sure what I was thinking when I wrote https://github.com/ekmett/lens/pull/878/files#r306307359, as the tuple instances are indeed implemented using irrefutable pattern matches. Hrm. I don't personally have a strong opinion on this, but for the sake of avoiding this PR from being bogged down by bikeshedding, it might be wise to stick to the current convention for now and raise a separate issue about how to reconcile the strictness of the instances for tuples and Product.
There was a problem hiding this comment.
We have _1' for when you want to reduce leaks
| _1 f (Identity a) = Identity <$> f a | ||
|
|
||
| instance Field1 (Unit a) (Unit b) a b where | ||
| _1 f (Unit a) = Unit <$> f a |
There was a problem hiding this comment.
I'd vote to keep it the way you've written it, as it's consistent with the other Field1 instances that are currently defined. If others feel strongly about changing this convention, then I think it would be best to leave that to a separate issue.
| import Data.Proxy (Proxy (Proxy)) | ||
| import GHC.Generics ((:*:) (..), Generic (..), K1 (..), | ||
| M1 (..), U1 (..)) | ||
| import GHC.Tuple (Unit) |
There was a problem hiding this comment.
Based on the CI failures, I think this should be (Unit(..)).
|
Most |
|
|
|
We could depend on |
4 participants
No description provided.