Closed ImmemorConsultrixContrarie closed 3 years ago
Regarding the "dirty hacks," yes, I think achieving parity between the two would be nice.
That said, I think I would prefer not seeing #[doc(hidden)]
used to hide traits. Frunk has always been pretty open with its traits, even when they are things we might prefer to be implementation details. Seeing a trait bound in the docs that you can't click on can be kind of frustrating when there's no clue what it is for. So I would suggest adding a note to it explaining precisely why it needs to exist (IIUC, Foldr with a single unborrowed Fn
s) and discouraging users from using it directly.
(there is the downside of them appearing in search results. If only there were a #[doc(no_search)]
!)
By the way, since this PR is highly likely to get merged and is a breaking change, you may want to add a bunch of other breaking changes to the patch until you push the new version, @lloydmeta.
There was a check failing due to a new clippy recommendation, which I fixed on master
, merging into this branch to see if it passes..
@ImmemorConsultrixContrarie do you want to add a short for the change in this PR in https://github.com/lloydmeta/frunk/blob/master/CHANGELOG.md ? If not, I can do it later.
Closes #171.
Unresolved question: do we want dirty hacks to completely unify
foldl
andfoldr
arguments?